AU2362699A - Pneumococcal surface proteins, genes encoding these and expression of these or portions thereof - Google Patents

Description

S F Ref: 412594D1

AUSTRALIA

PATENTS ACT 1990 COMPLETE SPECIFICATION FOR A STANDARD PATENT

ORIGINAL

Name and Address of Applicant: Actual Inventor(s): Address for Service: Invention Title: UAB Research Foundation UAB Station Birmingham Alabama 35294 UNITED STATES OF AMERICA David E. Briles, Larry S. McDaniel, Edwin Swiatlo, Janet Yother, Marilyn J. Crain, Susan Holllngshead, Rebecca Tart and Alexis Brooks-Walter Spruson Ferguson, Patent Attorneys Level 33 St Martins Tower, 31 Market Street Sydney, New South Wales, 2000, Australia Pneumococcal Surface Proteins, Genes Encoding These and Expression of These or Portions Thereof The following statement is a full description of this invention, including the best method of performing it known to me/us:- 5845 1 Pneumococcal Surface Proteins, Genes Encoding These and Expression of These or Portions Thereof Field of the Invention This invention relates to pneumococcal genes, portions thereof, expression products therefrom and uses of such genes, portions and products; especially to genes of Streptococcus pneumoniae, the gene encoding pneumococcal surface protein A (PspA) (said gene being "pspA"), pspA-like genes, pneumococcal surface protein C (PspC) (said gene being "pspC"), portions of such genes, expression products therefrom, and the uses of such genes, portions thereof and expression products i [n:\libc]03437:DKM therefrom. Such uses include uses of the genes and portions thereof for obtaining expression products by recombinant techniques, a~s well as for detecting the presence of Streptococcus pneumonjae or strains thereof by detecting DNA thereof by hybridization or amplification PCR) and hybridization techniques obtaining DNA-containing sample, contacting same with genes or fragment under PCR, amplification and/or hybridization conditions, and detecting presence of or qj isolating hybrid or amplified product). The expression product uses include use in preparing antigenic, immunological or vaccine compositions, for eliciting antibodies, an immunological response (other than or additional to antibodies) or a protective response (including antibody or other immunological response by admiistrin composition to a suitable host) or, the a a expression product can be for use in detectingth presence of Streptococcus pneuznoniae by detecting a. antibodies to Streptococcus pneurnonhae protein(s) or 0 antibodies to a portion thereof in a host, by a obtaining an antibody-containing sample from a relevant a aa host, contacting the sample with expression product and detecting binding (for instance by having the product labeled); and, the antibodies generated by the af. aforementioned compositions are usfu in dianostic or detection kits or assays. Thus, the invention relates to varied copstosof matter and methods for use thereof.

BAC~G~y aT E N

ETION

Streptococcus pneumoniae is an important cause of otitis media, meningitis, bacteremia and pneumonia.

Despite the use of antibiotics and vaccines, the prevalence of pneumococcal infections has declined little over the last twenty-five years.

3 It is generally accepted that immunity to Streptococcus pneumoniae can be mediated by specific antibodies against the polysaccharide capsule of the pneumococcus. However, neonates and young children fail to make an immune response against polysaccharide antigens and can have repeated infections involving the same capsular serotype.

One approach to immunizing infants against a number of encapsulated bacteria is to conjugate the capsular polysaccharide antigens to protein to make them immunogenic. This approach has been successful, for example, with Haemophilus influenzae b (see U.S. Patent no. 4,496,538 to Gordon and U.S. Patent no. 4,673,574 to Anderson). However, there are over eighty known capsular serotypes of S. pneumoniae of which twenty-three account for most of the disease. For a pneumococcal polysaccharide-protein conjugate to be successful, the capsular types responsible for most pneumococcal infections would have to be made adequately immunogenic.

This approach may be difficult, because the twenty-three polysaccharides included in the presently-available vaccine are not all adequately immunogenic, even in adults.

:An alternative approach for protecting S' children, and also the elderly, from pneumococcal infection would be to identify protein antigens that could elicit protective immune responses. Such proteins may serve as a vaccine by themselves, may be used in conjunction with successful polysaccharide-protein conjugates, or as carriers for polysaccharides.

I've -0 McDaniel et al. J. Exp. Med. 160:386-397, 1984, relates to the production of hybridoma antibodies 3C that recognize cell surface polypeptide(s) on S.

pneumoniae and protection of mice from infection with certain strains of encapsulated pneumococci by such antibodies. This surface protein antigen has been termed "pneumococcal surface protein A" or PspA for short.

McDaniel et al. Microbial Pathogenesis 1:519-531, 1986, relates to studies on the characterization of the PspA. Considerable diversity in the PspA molecule in different strains was found, -as were differences in the epitopes recognized by different antibodies.

McDaniel et al. (III), J. Exp. Med. 165:381- 394, 1987, relates to immunization of X-linked immunodeficient (XID) mice with non-encapsulated pneumococci expressing PspA, but not isogenic pneumococci lacking PspA, protects mice from subsequent fatal infection with pneumococci.

it, McDaniel et al. Infect. Immun., 59:222- 228, 1991, relates to immunization of mice with a recombinant full length fragment of PspA that is able to elicit protection against pneumococcal strains of capsular types 6A and 3.

Crain et al, Infect.Immun., 56:3293-3299, 1990, relates to a rabbit antiserum that detects PspA in 100% 95) of clinical and laboratory isolates of strains of S. pneumoniae. When reacted with seven monoclonal antibodies to PspA, fifty-seven S. pneumoniae isolates exhibited thirty-one different patterns of reactivity.

4 The PspA protein type is independent of capsular type. It would seem that genetic mutation or exchange in the environment has allowed for the development of a large pool of strains which are highly diverse with respect to capsule, PspA, and possibly other molecules with variable structures. Variability of PspA's from different strains also is evident in their S molecular weights, which range from 67 to 99 kD. The observed differences are stably inherited and are not the result of protein degradation.

Immunization with a partially purified PspA from a recombinant X gtll clone, elicited protection against challenge with several S. pneumoniae strains representing different capsular and PspA types, as described in McDaniel et al. Infect. Zmmun. 59:222- 228, 1991. Although clones expressing PspA were constructed according to that paper, the product was insoluble and isolation from cell fragments following lysis was not possible.

While the protein is variable in structure between different pneumococcal strains, numerous crosso reactions exist between all PspA's, suggesting that sufficient common epitopes may be present to allow a single PspA or at least a small number of PspA's to elicit protection against a large number of S. pneumoniae strains.

i' In addition to the published literature specifically referred to above, the inventors, in conjunction with co-workers, have published further details concerning PspA's, as follows: 1. Abstracts of 89th Annual Meeting of the American Society for Microbiology, p. 125, item D-257, May 1989; 2. Abstracts of 90th Annual Meeting of the American Society for Microbiology, p. 98, item D-106, May 1990; 3. Abstracts of 3rd International

ASM

Conference on Streptococcal Genetics, p.

11, item 12, June 1990; 4. Talkington et al, Infect. Immun. 59:1285- 1289, 1991; Yother et al J. Bacteriol. 174:601- 609, 1992; and 6. Yother et al J. Bacteriol. 174:610- 618, 1992.

7. McDaniel et al Microbiol.

Pathogenesis, 13:261-268.

It would be useful to provide PspA or fragments thereof in compositions, including PspA's or fragments from varying strains in such compositions, to provide antigenic, immunological or vaccine compositions; and, it 6 is even further useful to show that the various strains can be grouped or typed, thereby providing a basis for cross-reactivities of PspA's or fragments thereof, and thus providing a means for determining which strains to represent in such compositions (as well as how to test for, detect or diagnose one strain from another).

Further, it would be advantageous to provide a pspA like gene or a pspC gene in certain strains, as well as primers (oligonucleotides) for identification of such a gene, as well as of conserved regions in that gene and in pspA; for instance, for detecting, determining, isolating, or diagnosing strains of S. pneumonia. These uses and advantages, it is believed, have not heretofore 0o been provided in the art.

OBJECTS AND SUMMARY OF THE INVENTION The invention provides an isolated amino acid molecule comprising residues 1 to 115, 1 to 260, 192 to 588, 192 to 299, or residues 192 to 260 of pneumococcal S" surface protein A of Streptococcus pneumoniae.

The invention further provides an isolated

DNA

S molecule comprising a fragment of a pneumococcal surface protein A gene of Streptococcus pneumoniae encoding the isolated amino acid molecule.

The invention also provides PCR primers or hybridization probes comprising the isolated

DNA

molecule.

The invention additionally provides an antigenic, vaccine or immunological composition 86 comprising the amino acid molecule.

The invention includes an isolatedDNA molecule comprising nucleotides 1 to 26, 1967 to 1990, 161 to 187, 1093 to 1117, or 1312 to 1331 or 1333 to 1355 of a pneumococcal surface protein A gene of Streptococcus 9o pneumoniae. The DNA molecule can be used as a PCR primer or hybridization probe; and therefore the invention comprehends a PCR primer or hybridization probe comprising the isolated DNA molecule.

The invention also includes an isolated

DNA

3j molecule comprising a fragment having homology with a portion of a pneumococcal surface protein A gene of Streptococcus pneumoniae. The DNA preferably is the

I

following (which include the portion having homology and restriction sites, and selection of other restriction sites or sequences for suc4 DNA is within the ambit of the skilled artisan from this disclosure): S'

CCGGATCCAGCTCCTGCACCAAAAAC

GCGCGTCGACGGCTTAAACCCATTCACCATTGG;

CCGGATCCTGAGCCAGAGCAGTTGGCTG;

CCGGATCCGCTCAAAGAGATTGATGAGTCTG;

GCGGATCCCGTAGCCAGTCAGTCTAAAGCTG

0

CTGAGTCGACTGGAGTTTCTGGAGCTGGAGC

CCGGATCCAGCTCCAGCTCCAGAAACTCCAG

GCGGATCCTTGACCAATATTTACGGAGGAGGC;

GTTTTTGGTGCAGGAGCTGG;

GCTATGGGCTACAGGTTG;

CCACCTGTAGCCATAGC;

CCGCATCCAGCGTGCCTATCTTAGGGGCTGGTT; and

GCAAGCTTATGATATAGAAATTTGTAAC

(thus, the invention broadly comprehends DNA homologous to portions of pspA; preferably further including S 0o restriction sequences).

These DNA molecules can be used as PCR primers or probes; and thus, the invention comprehends a primer or probe comprising and of these molecules.

The invention further still provides

PCR

S probe(s) which distinguishes between pspA and pspA-like nucleotide sequence, as well as PCR probe(s) which hybridizes to both pspA and pspA-like nucleotide sequences.

Additionally, the invention includes a PspA S extract prepared by a process comprising: growing pneumococci in a first medium containing choline chloride, eluting live pneumococci with a choline chloride containing salt solution, and growing the pneumococci in a second medium containing an alkanolamine and substantially no choline; as well as a PspA extract prepared by that process and further comprising purifying PspA by isolation on a choline-Sepharose affinity column.

These processes are also included in the invention.

An immunological composition comprising thses extracts is comprehended by the invention, as well as an immunological composition comprising the full length PspA.

A method for enhancing the immunogenicity of a PspA-containing immunological composition comprising, in said composition, the C-terminal portion of PspA, is additionally comprehended, as well.

An immunological composition comprising at least two PspAs. The latter immunological composition can have the PspAs from different groups or families; the S groups or families can be based on RFLP or sequence studies (see, Fig. 13).

Further, the invention provides an isolated amino acid molecule comprising pneumococcal surface protein C, Pspc, of Streptococcus pneumoniae having an S alpha-helical, proline rich and repeat regions, an isolated DNA molecule comprising a pneumcoccal surface protein C gene encoding the aforementioned Pspc, and primers and hybrization probes consisting essentially of the isolated DNA molecule.

Still further, an isolated amino acid molecule comprising pneumococcal surface protein C, PspC, of Streptococcus pneumoniae is provided, having an alphahelical, proline rich and repeat regions, having substantial homology with a protection eliciting region Q of PspA, and an isolated DNA molecule comprising a pneumcoccal surface protein C gene encoding the aforementioned PspC, and primers and hybrization probes consisting essentially of the isolated DNA molecule are provided by the present invention.

%0 Additionally, the present invention provides immunological compositions comprising Pspc.

These and other embodiments are disclosed or are obvious from the following detailed description.

BRIEF DESCRIPTION OF THE FIGURES nFigures 1 A and 1 B show: Evaluation of diested plasmid constructs. Fig. 1A: 1% agarose gel electrophoresis of plasmids isolated from transformed

E.

coli BL21(DE3) strains stained with ethidium bromide.

Lane 1: 1 kb DNA ladder (sizes noted in kb), lane 2: pRCT125; lane 3: pRC105, lane 4: DBL5 DspA insert, lane pRCT113, lane 6: BG9739 EyDA insert, lane 7: pRCT117, and lane 8: L81905 PESA insert. Fig. IB: Corresponding Southern blot of gel in Fig. 1A probed with full-length Rxl sEEA and hybridization detected as described in Example 1. The arrow indicates the 1.2 kb EspA digested inserts from plasmid constructs and the PCR-amplified o ESA fragments from the pneumococcal donor strains used in cloning.

Figure 2 shows: Evaluation of strain RCTiOS cell fractions containing truncated DBL5 PspA. Proteins from E. coli cell fractions were resolved by 10% SDS- PAGE, transferred to NC, and probed with MAb XiR278.

Lane 1: molecular weight markers (noted in kDa), lane 2: full-length, native DBL5 PspA, lane 3: uninduced cells, lanes 4-6: induced cells; 1 hr, 2 hr, and 3 hr of IPTG induction respectively, lane 7: periplasmic proteins, lane 8: cytoplasmic proteins, and lane 9: insoluble cell wall/membrane material.

Figure 3 shows: SDS-PAGE of R36A PspA (80 ng) column isolated from CDM-ET and an equal volume of an equivalent WG44.1 prep. Identical gels are shown stained with Bio-Rad silver kit or immunoblotted with PspA MAb XiR278(B). The PspA isolated from R36A shows the characteristic monomer (84 kDa) and dimer bands.

Figure 4 shows: Cell lysates of pneumococcal isolates MC27 and MC28 were subjected to SDS-PAGE and ce transferred to nitrocellulose for Western blotting with seven MAb to PspA. 7D2 detected a protein of 82 kDa in each isolate and XiR278 and 2A4 detected a protein of 190 kDa in each isolate. MAb Xi64, Xi126, 1A4 and SR4W4 were not reactive. Strains MC25 and MC26 yielded identical sC results.

Figure 5 (Figs. 5A and 5B) shows: Southern blot of Hind III digest of MC25-MC28 chromosomal

DNA

developed at a stringency greater than 95 percent.

A

digest of Rxl DNA was used as a comparison. The blot was probed with LSMpspA13/2, a full length Rxl probe (Fig. and LSMpspAl2/6 a 5' probe of Rxi pspA (Fig. The S same concentration of Rxl DNA was used in both panels, but the concentrations of MC25-MC28 DNA in Fig. SB were half that used in Fig. 5A to avoid detection of partial digests.

Figure 6 shows: RFLP of amplified pspA. PspA 'o from MC25 was amplified by PCR using 5' and 3' primers for pspA (LSM13 and LSM, respectively). The amplified DNA was digested with individual restriction endonucleases prior to electrophoresis and staining with ethidium bromide. Lane 1 BcllI, Lane 2 BAMHI, Lane 3 SBstNI, Lane 4 PstI, Lane 5 SacI, Lane 6 EcoRI, Lane 7 SmaI, Lane 8 KpnI.

Figure 7 shows: A depiction of PspA showing the relative location and orientation of the oligonucleotides.

Figure 8 shows: Derivatives of the S.

pneumoniae D39-Rxl family.

Figures 9 to 10 show: Electrophoresis of pspA or amplified pspA product with HhaI (Fig. Sau3AI (Fig. 2 ,Figure 11 shows: RFLP pattern of two isolates from six families.

Figure 12 shows: RFLP pattern of two isolates from six families (using products from amplification with SKH2 and LSM13).

3C Figure 13 shows: Sequence primarily in the Nterminal half of PspA.

Figure 14 shows: Cell lysates of pneumococcal isolates MC27 and MC28, subjected to SDS-PAGE and Western blotting with seven MAbs to PspA; 7D2 detected a protein 3 of 82 kDa in each isolate, and Xi278 and 2A4 detected a protein of 190 kDa in each isolate; MAbs Xi64, Xi126, 1A4 and SR4W4 were not reactive; strains MC25 and MC26 yielded identical results (not shown).

Figure 15A and 5B show: a Southern blot of Hind III digest of MC25-28 chromosomal DNA, using a digest of Rxl DNA as a comparison; the blot was probed with LSMpspA13/2, a full length Rxl probe and LSMpspAl2/6, a 5' probe of Rxl pspA the same concentration of Rxl DNA was used in both panels, but the concentrations of MC25-28 DNA in B were half that used in 'o A to avoid detection of partial digests.

Figures 15C and 15D show: the nucleotide sequences of primers LSM13, LSM2, LSM12 and LSM6, and that of probes LSMpspAl3/2 and LSMpspAl2/ 6 S Figure 16 shows: RFLP of amplified pspA, wherein PspA from MC25 was amplified by PCR using 5' and 3' primers for pspA (LSM13 and LSM2, respectively); the amplified DNA was digested with individual restriction endonucleases prior to electrophoresis and staining with S.ethidium bromide; Bel I was used in lane 1; BamH I was used in lane 2; BstN I was used in lane 3; Pst I was used in lane 4; Sac I was used in lane 5; EcoR I was used in lane 6; Sma I was used in lane 7; and Kpn I was used in lane 8, Figure 17 shows: position and orientation of oligonucleotides relative to domains encoded by pspA; numbers along the bottom of the Figure represent amino acids in the mature PspA polypeptide from strain Rxl, and arrows represent the relative position (not to scale) and orientation of oligonucleotides.

0 Figure 18 shows: a restriction map of the pZero vector.

Figure 19 shows: the nucleotide sequences of SKH2, LSM13, N192 and C588.

Figure 20 shows: a comparison of the structural motifs of PspA and Pspc; PspA has a smaller alpha-helical region, and does not contain the direct repeats within the alpha-helix (indicated by the dashed lines); the alpha-helical regions which are homologous between PspA and PspC are indicated by the dashed lines); the alphahelical regions which are homologous between PspA and PspC are indicated by the striped pattern; and PCR primers are indicated by the arrows.

Figure 21 shows: the amino acid and nucleotide sequence of PspC, wherein the putative -10 and regions are underlined, and the ribosomal binding site is in lower case.

ic Figure 22 shows: the Bestfit analysis of PspA and Pspc; percent identity is 69% and percent similarity is 77%; amino acids of PspA are one the bottom line (1- 588) and amino acids of Pspc are on the top line (249- 891), and a dashed line indicated identity.

Figure 23 shows: the coiled coil motif of the alpha-helix of PspC; amino acids that are not in the coiled coil motif are in the right column.

I

i 1 Figure 24 shows: the repeat regions of the Figure 25 shows: helical and proline region Figure 26 shows: of pspA.Rxl to pspC.D39.

Figure 27 shows: pspC.EF6797 and pspC.D39.

Figure 28 shows: Pspc of EF6797 and D39.

Figure 29 shows: PspC.D39 and PspA.Rxl.

DETAILED DEBCRIPTION a matrix plot comparison of alpha-helical region of Pspc.

the sequence of the alpha s of LXS532 (PspC.D39).

a comparison of nucleotides a BESTFIT analysis of the amino acid comparison of the amino acid comparison of Knowledge of and familiarity with the applications incorporated herein by reference is assumed; and, those applications disclose the sequence of pspA as well as certain portions thereof, and PspA and 1 compositions containing PspA.

As discussed above and in the following Examples, the invention relates to truncated PspA, e.g., PspA C-terminal to position 192 such as a.a. 192-588 ("BC100") 192-299 and 192-260 of PspA eliciting crossprotection, as well as to DNA encoding such truncated PspA (which amplify the coding for these amino acid regions homologous to most PspAs).

The invention further relates to a pspA-like gene, or a pspc gene and portions thereof probes, primers) which can hybridize thereto and/or amplify that gene, as well as to DNA molecules which hybridize to tc pspA, so that one can, by hybridization assay and/or amplification, ascertain the presence of a particular pneumococcal strain; and, the invention provides that a PspC can be produced by the pspA-like or pspC sequence (which PspC can be used like PspA).

,6 sIndeed, the invention further relates to oligonucleotide probes and/or primers which react with pspA and/or pspC of many, if not all, strains, so as to permit identification, detection or diagnosis of any pneumococcal strain, as well as to expression products of o such probes and/or primers, which can provide crossreactive epitopes of interest.

The repeat region of pspA and/or pspC is highly conserved such that the present invention provides oligonucleotide probes or primers to this region reactive with most, if not all strains, thereby providing diagnostic assays and a means for identifying epitopes of interest.

The invention demonstrates that the pspc gene is homologous to the pspA gene in the leader sequence, 3o first portion of the proline-rich region and in the repeat region; but, these genes differ in the second portion of their proline-rich regions and at the very 3' end of the gene encoding the 17 amino acid tail of PspA.

The product of the pspC gene is expected to lack a Cterminal tail, suggesting different anchoring than PspA.

Drug interference with functions such as surface binding of the coding for repeat regions of pspA and the pspc genes, or with the repeat regions of the expressioh products, is therefore a target for intervention of pneumococcal infection.

Further still, the invention provides evidence of additional pspA homologous sequences, in addition to pspA and the pspC sequence. The invention, as mentioned above, includes oligonucleotide probes or primers which distinguish between pspA and the pspC sequence, e.g., LSM1 and LSM2, useful for diagnostic detecting, or isolating purposes; and LSM1 and LSM10 or LSM1 and LSM7 which amplify a portion of the pspc gene, particularly the portion of that gene which encodes an antigenic, immunological or protective protein.

i 4The invention further relates to a method for 4 the isolation of native PspA by growth of pneumococci medium containing high concentrations of(about 0.9% to about preferably choline chloride, elation of live pneumococci with a salt solution containing choline chloride, about 1% about preferably 2% choline chloride, and growth of pneumococci in medium in which the choline in the medium has been almost or substantially completely replaced with a lower alkanolamine, preferably

C

2 alkanolamine, preferably

C

2 alkanolamine, preferably ethanolamine 0.0000005% to 0.0000015%, preferably 0.000001% choline chloride plus 0.02% to 0.04% alkanolamine (ethanolamine), preferably PspA from such pneumococci is then preferably isolated from a choline-sepharose affinity column, thereby providing 3o highly purified PspA. Such isolated and/or purified PspA is highly immunogenic and is useful in antigenic, immunological or vaccine composition.

Indeed, the growth media of the pneumococci grown in the presence of the alkanolamine (rather than 3 choline) contains PspA and is itself highly.immunogenic and therefore useful as an antigenic, immunological or vaccine composition; and, is rather inexpensive to produce. Per microgram of PspA, the PspA in the alkanolamine medium is much more protective than PspA isolated by other means, from extracts. Perhaps, without wishing to necessarily be bound by any one particular theory, there is a synergistic effect upon PspA by the other components present prior to isolation, or simply PspA is more protective (more antigenic) prior to isolation and/or purification (implying a possibility of some loss of activity from the step of isolation and/or purification).

The invention further relates to the N-terminal 115 amino acids of PspA, which is useful for compositions comprising an epitope of interest, immunological or vaccine compositions, as well as the DNA coding therefor, i which is useful in preparing these N-terminal amino acids by recombination, or for use as probes and/or primers for hybridization and/or amplification for identification, detection or diagnosis purposes.

The invention further demonstrates that there is a grouping among the pspA RFLP families. This provides a method of identifying families of different PspAs based on RFLP pattern of pspAs, as well as a means for obtaining diversity of PspAs in an antigenic, immunological or vaccine composition; and, a method of I characterizing clonotypes of PspA based on RFLP patterns of PspA. And, the invention thus provides oligonucleotides which permit amplification of most, a majority, if not all of S. pneumoniae and thereby permit RFLP analysis of a majority, if not all, S.

pneumoniae.

The invention also provides PspC, having an approximate molecular weight of 105 kD, with an estimated pI of 6.09, and comprising an alpha-helical region, followed by a proline-rich domain and repeat region.

A

major cross-protective region of PspA comprises the Cterminal third of the alpha-helical region (between residues 192 and 260 of PspA), which region accounts for the binding of 4 of 5 cross-protective MAb, and PspA fragments comprising this region can elicit crossprotective immunity in mice. Homology between Pspc and PspA begins at amino acid 148 of PspA, thus including the region from 192 to 299, and including the entire PspC sequence C-terminal of amino acid 486. Due to the substantial sequence homology between PspA and Pspc in a region comprising the epitopes of interest, known to be protection eliciting, PspC is likely to comprise epitopes of interest similar to those found in PspA. Antibodies specific for this region of PspA, between amino acids 148 and 299, should cross-react with Pspc, and thus afford protection by reacting with PspC and PspA.

*I Similarly, immunization with PspC would be expected to elicit antibodies cross-protective against PspA.

An epitope of interest is an antigen or immunogen or immunologically active fragment thereof from a pathogen or toxin of veterinary or human interest.

The present invention provides an immunogenic, immunological or vaccine composition containing the pneumococcal epitope of interest, and a pharmaceutically acceptable carrier or diluent. An immunological composition containing the pneumococcal epitope of interest, elicits an immunological response local or 0 A systemic. The response can, but need not be, protective.

Am immunogenic composition containing the pneumococcal epitope of interest, likewise elicits a local or systemic immunological response which can, but need not be, protective. A vaccine composition elicits a local or systemic protective response. Accordingly, the terms "immunological composition, and "immunogenic composition, include a "vaccine composition,, (as the two former terms can be protective compositions).

The invention therefore also provides a method of inducing an immunological response in a host mammal comprising administering to the host an immunogenic, immunological or vaccine composition comprising the I I I pneumococcal epitope of interest, and a pharmaceutically acceptable carrier or diluent.

The DNA encoding,the pneumococcal epitope of interest can be DNA which codes for full length PspA, PspC, or fragments thereof. A sequence which codes for a fragment of PspA or PspC can encode that portion of PspA or PspC which contains an epitope of interest, such as a protection-eliciting epitope of the protein.

Regions of PspA and PspC have been identified 0 from the Rxl strain of S. pneumoniae which not only contain protection-eliciting epitopes, but are also sufficiently cross-reactive with other PspAs from other S. pneumoniae strains so as to be suitable candidates for the region of PspA to be incorporated into a vaccine, immunological or immunogenic composition. Epitopic regions of PspA include residues 1 to 115, 1 to 314, 192 to 260 and 192 to 588. DNA encoding fragments of PspA can comprise DNA which codes for the aforementioned epitopic regions of PspA; or it can comprise DNA encoding .overlapping fragments of PspA, fragment 192 to 588 includes 192 to 260, and fragment 1 to 314 includes 1 to S 115 and 192 to 260.

As to epitopes of interest, one skilled in the art can determine an epitope of immunodominant region of Sa peptide or polypeptide and ergo the coding DNA therefor from the knowledge of the amino acid and corresponding DNA sequences of the peptide or polypeptide, as well as from the nature of particular amino acids size, charge, etc.) and the codon dictionary, without undue ^w experimentation.

A general method for determining which portions of a protein to use in an immunological composition focuses on the size and sequence of the antigen of interest. "In general, large proteins, because they have Smore potential determinants are better antigens than small ones. The more foreign an antigen, that is the less similar to self configurations which induce tolerance, the more effective it is in provoking an immune response." Ivan Roitt, Essential Immunologyv, 1988.

As to size, the skilled artisan can maximize the size of the protein encoded by the DNA sequence to be inserted into the viral vector (keeping in mind the packaging limitations of the vector). To minimize the DNA inserted while maximizing the size of the protein expressed, the DNA sequence can exclude introns (regions 'IC of a gene which are transcribed but which are subsequently excised from the primary RNA transcript).

At a minimum, the DNA sequence can code for a peptide at least 8 or 9 amino acids long. This is the minimum length that a peptide needs to be in order to a 'i stimulate a CD4+ T cell response (which recognizes virus infected cells or cancerous cells). A minimum peptide a a length of 13 to 25 amino acids is useful to stimulate a a, CD8+ T cell response (which recognizes special antigen presenting cells which have engulfed the pathogen). see Kendrew, supr-a. However, as these are minimum lengths, 4, 4 these peptides are likely to generate an immunological response, an antibody or T cell response; but, for a protective response (as from a vaccine composition) a longer peptide is preferred.

4.31 With respect to the sequence, the DNA sequence 4 lot.

#oil preferably encodes at least regions of the peptide that a a generate an antibody response or a T cell response. one method to determine T and B cell epitopes involves epitope mapping. The protein of interest "is fragmented lz, into overlapping peptides with proteolytic enzymes. The individual peptides are then tested for their ability to bind to an antibody elicited by the native protein or to induce T cell or B cell activation. This approach has been particularly useful in mapping T-cell epitopes since S the T cell recognizes short linear peptides- complexed with MHC molecules. The method is less effective for determining B-cell epitopes" since B cell epitopes are often not linear amino acid sequence but rather result from the tertiary structure of the folded three dimensional protein. Janis Kuby, Immunoloq, (1992) pp.

79-80.

SAnother method for determining an epitope of interest is to choose the regions of the protein that are hydrophilic. Hydrophilic residues are often on the surface of the protein and therefore often the regions of the protein which are accessible to the antibody. Janis O Kuby, Immunology, (1992) P. 81.

Yet another method for determining an epitope of interest is to perform an X-ray cyrstallographic analysis of the antigen (full length)-antibody complex.

Janis Kuby, Immunology, (1992) p. Still another method for choosing an epitope of interest which can generate a T cell response is to identify from the protein sequence potential HLA anchor binding motifs which are peptide sequences which are Sknown to be likely to bind to the MHC molecule.

.i e The peptide which is a putative epitope, to generate a T cell response, should be presented in a MHC complex. The peptide preferably contains appropriate anchor motifs for binding to the MHC molecules, and should bind with high enough affinity to generate an immune response. Factors which can be considered are: the HLA type of the patient (vertebrate, animal or human) expected to be immunized, the sequence of the protein, the presence of appropriate anchor motifs and the occurance of the peptide sequence in other vital cells.

An immune response is generated, in general, as follows: T cells recognize proteins only when the protein has been cleaved into smaller peptides and is presented in a complex called the "major histocompatability complex MHC" located on another cell's surface. There are two classes of MHC complexes class I and class II, and each class is made up of many different alleles. Different patients have different types of MHC complex alleles; they are said to have a 'different HLA type'.

Class I MHC complexes are found on virtually every cell and present peptides from proteins produced inside the cell. Thus, Class I MHC complexes are useful for killing cells which when infected by viruses or which have become cancerous and as the result of expression of an oncogene. T cells which have a protein called CD4 on their surface, bind to the MHC class I cells and secrete ic lymphokines. The lymphokines stimulate a response; cells arrive and kill the viral infected cell.

Class II MHC complexes are found only on antigen- presenting cells and are used to present peptides from circulating pathogens which have been endocytosed by the antigen- presenting cells. T cells which have a protein called CD8 bind to the MHC class

II

Scells and kill the cell by exocytosis of lytic granules.

Some guidelines in determining whether a protein is an epitopes of interest which will stimulate a i T cell response, include: Peptide length the peptide should be at least 8 or 9 amino acids long to fit into the MHC class I complex and at least 13-25 amino acids long to fit into a class II MHC complex. This length is a minimum for the peptide to bind to the MHC complex. It is preferred for the peptides to be longer than these S lengths because cells may cut the expressed peptides.

The peptide should contain an appropriate anchor motif S. which will enable it to bind to the various class I or class II molecules with high enough specificity to 0o generate an immune response (See Bocchia, M. et al, S eific Binding of Leukemia ncoene Fusion Protein Peptides to HLA Class I Molecule, Blood 85:2680-2684; Englehard, VH, Structure of -etides associated with class I and class II MHC molecules Ann. Rev. Immunol.

3- 12:181 (1994)). This can be done, without-undue experimentation, by comparing the sequence of the protein of interest with published structures of peptides associated with the MHC molecules. Protein epitopes recognized by T cell receptors are peptides generated by enzymatic degradation of the protein molecule and are prestnted on the cell surface in association with class I Sor class II MHC molecules.

Further, the skilled artisan can ascertain an epitope of interest by comparing the protein sequence with sequences listed in the protein data base. Regions of the protein which share little or no homology are Sbetter choices for being an epitope of that protein and are therefore useful in a vaccine or immunological composition. Regions which share great homology with widely found sequences present in vital cells should be avoided.

Even further, another method is simply to generate or express portions of a protein of interest, S- generate monoclonal antibodies to those portions of the protein of interest, and then ascertain whether those 4 antibodies inhibit growth in vitro of the pathogen from o which the from which the protein was derived. The skilled artisan can use the other guidelines set forth in this disclosure and in the art for generating or expressing portions of a protein of interest for analysis as to whether antibodies thereto inhibit growth in vitro.

For example, the skilled artisan can generate portions of a protein of interest by: selecting 8 to 9 or 13 to amino acid length portions of the protein, selecting hydrophilic regions, selecting portions shown to bind from X-ray data of the antigen (full length)-antibody Scomplex, selecting regions which differ in sequence from other proteins, selecting potential HLA anchor binding motifs, or any combination of these methods or other methods known in the art.

Epitopes recognized by antibodies are expressed on the surface of a protein. To determine-the regions of a protein most likely to stimulate an antibody response one skilled in the art can preferably perform an epitope map, using the general methods described above, or other mapping methods known in the art.

As can be seen from the foregoing, without undue experimentation, from this disclosure and the knowledge in the art, the skilled artisan can ascertain the amino acid and corresponding DNA sequence of an epitope of interest for obtaining a T cell, B cell and/or antibody response. In addition, reference is made to Gefter et al., U.S. Patent No. 5,019,384, issued May 28, .o 1991, and the documents it cites, incorporated herein by reference (Note especially the "Relevant Literature" section of this patent, and column 13 of this patent which discloses that: "A large number of epitopes have been defined for a wide variety of organisms of interest.

Of particular interest are those epitopes to which neutralizing antibodies are directed. Disclosures of such epitopes are in many of the references cited in the Relevant Literature section.") Further, the invention demonstrates that more 2.C than one serologically complementary PspA molecule can be in an antigenic, immunological or vaccine composition, so as to elicit better response, protection, for S instance, against a variety of strains of pneumococci; and, the invention provides a system of selecting PspAs j< for a multivalent composition which includes crossprotection evaluation so as to provide a maximally efficacious composition.

The determination of the amount of antigen, PspA or truncated portion thereof and optional 3c adjuvant in the inventive compositions and the preparation of those compositions can be in accordance with -standard techniques well known to those skilled in the pharmaceutical or veterinary arts. In particular, the amount of antigen and adjuvant in the inventive 3( compositions and the dosages administered are determined by techniques well known to those skilled in the medical or veterinary arts taking into consideration such factors as the particular antigen, the adjuvant (if present), the age, sex, weight, species and condition of the particular patient, and the route of,administration. For instance, dosages of particular PspA antigens for suitable hosts in T which an immunological response is desired, can be readily ascertained by those skilled in the art from this disclosure (see, the Examples), as is the amount of any adjuvant typically administered therewith. Thus, the skilled artisan can readily determine the amount of iO antigen and optional adjuvant in compositions and to be administered in methods of the invention. Typically, an adjuvant is commonly used as 0.001 to 50 wt% solution in phosphate buffered saline, and the antigen is present on the order of micrograms to milligrams, such as about 0.0001 to about 5 wt%, preferably about 0.0001 to about 1 wt%, most preferably about 0.0001 to about 0.05 wt% (see, S: Examples below or in applications cited herein).

Typically, however, the antigen is present in b an amount on the order of micrograms to milligrams, or about 0.001 to about 20 wt%, preferably about 0.01 to about 10 wt%, and most preferably about 0.05 to about wt% (see, Examples below).

Of course, for any composition to be t i administered to an animal or human, including the components thereof, and for any particular method of administration, it is preferred to determine therefor: toxicity, such as by determining the lethal dose (LD) and LDs in a suitable animal model rodent such as mouse; and, the dosage of the composition(s), l 3 concentration of components therein and timing of administering the composition(s), which elicit a suitable immunological response, such as by titrations of sera and analysis thereof for antibodies or antigens, by ELISA and/or RFFIT analysis. Such determinations do not require undue experimentation from the knowledge of the skilled artisan, this disclosure and the documents cited herein. And, the time for sequential administrations can be ascertained without undue experimentation.

Examples of compositions of the invention include liquid preparations for orifice, oral, nasal, anal, vaginal, peroral, intragastric, mucosal perlingual, alveolar, gingival, olfactory or respiratory mucosa) etc., administration such as suspensions, syrups or elixirs; and, preparations for parenteral, subcutaneous, intradermal, intramuscular or Sintravenous administration injectable administration), such as sterile suspensions or emulsions. Such compositions may be in admixture with a suitable carrier, diluent, or excipient such as sterile water, physiological saline, glucose or the like. The compositions can also be lyophilized. The compositions can contain auxiliary substances such as wetting or emulsifying agents, pH buffering agents, gelling or viscosity enhancing additives, preservatives, flavoring Si' agents, colors, and the like, depending upon the route of administration and the preparation desired. Standard texts, such as "REMINGTON'S PHARMACEUTICAL SCIENCE", 17th edition, 1985, incorporated herein by reference, may be consulted to prepare suitable preparations, without undue experimentation.

S pCompositions of the invention, are conveniently provided as liquid preparations, isotonic aqueous solutions, suspensions, emulsions or viscous compositions which may be buffered to a selected pH. If digestive tract absorption is preferred, compositions of the 3o invention can be in the "solid" form of pills, tablets, capsules, caplets and the like, including "solid" preparations which are time-released or which have a liquid filling, gelatin covered liquid, whereby the gelatin is dissolved in the stomach for delivery to the Sgut. If nasal or respiratory (mucosal) administration is desired, compositions may be in a form and dispensed by a squeeze spray dispenser, pump dispenser or aerosol dispenser. Aerosols are usually under pressure by means of a hydrocarbon. Pump dispensers can preferably dispense a metered dose or, a dose having a particular particle size.

SCompositions of the invention can contain pharmaceutically acceptable flavors and/or colors for rendering them more appealing, especially if they are administered orally. The viscous compositions may be in the form of gels, lotions, ointments, creams and the like rc and will typically contain a sufficient amount of a thickening agent so that the viscosity is from about 2500 to 6500 cps, although more viscous compositions, even up to 10,000 cps may be employed. Viscous compositions have a viscosity preferably of 2500 to 5000 cps, since above that range they become more difficult to administer.

However, above that range, the compositions can approach solid or gelatin forms which are then easily administered as a swallowed pill for oral ingestion.

'Liquid preparations are normally easier to prepare than gels, other viscous compositions, and solid compositions. Additionally, liquid compositions are somewhat more convenient to administer, especially by injection or orally, to animals, children, particularly small children, and others who may have difficulty swallowing a pill, tablet, capsule or the like, or in multi-dose situations. Viscous compositions, on the other hand, can be formulated within the appropriate viscosity range to provide longer contact periods with S mucosa, such as the lining of the stomach or nasal oo mucosa.

Obviously, the choice of suitable carriers and other additives will depend on the exact route of administration and the nature of the particular dosage form, liquid dosage form whether the 3' composition is to be formulated into a solUtion, a suspension, gel or another liquid form], or solid dosage form whether the composition is to be formulated into a pill, tablet, capsule, caplet, time release form or liquid-filled form].

Solutions, suspensions and gels, normally contain a major amount of water (preferably purified water) in addition to the antigen, lipoprotein and optional adjuvant. Minor amounts of other ingredients such as pH adjusters a base such as NaOH), emulsifiers or dispersing agents, buffering agents, preservatives, wetting agents, jelling agents, i' methylcellulose), colors and/or flavors may also be present. The compositions can be isotonic, it can have the same osmotic pressure as blood and lacrimal fluid.

The desired isotonicity of the compositions of Sthis invention may be accomplished using sodium chloride, or other pharmaceutically acceptable agents such as S: dextrose, boric acid, sodium tartrate, propylene glycol or other inorganic or organic solutes. Sodium chloride Sis preferred particularly for buffers containing sodium o. ions.

Viscosity of the compositions may be maintained S at the selected level using a pharmaceutically acceptable thickening agent. Methylcellulose is preferred because it is readily and economically available and is easy to work with. Other suitable thickening agents include, for example, xanthan gum, carboxymethyl cellulose, hydroxypropyl cellulose, carbomer, and the like. The S preferred concentration of the thickener will depend upon the agent selected. The important point is to use an Sc amount which will achieve the selected viscosity.

Viscous compositions are normally prepared from solutions by the addition of such thickening agents.

A pharmaceutically acceptable preservative can be employed to increase the shelf-life of the 31 compositions. Benzyl alcohol may be suitable, although a variety of preservatives including, for example, parabens, thimerosal, chlorobutanol, or benzalkonium

I

chloride may also be employed. A suitable concentration of the preservative will be from 0.02% to 2% based on the total weight although there may be appreciable variation depending upon the agent selected.

Those skilled in the art will recognize that the components of the compositions must be selected to be chemically inert with respect to the PspA antigen and optional adjuvant. This will present no problem to those skilled in chemical and pharmaceutical principles, or 'o problems can be readily avoided by reference to standard texts or by simple experiments (not involving undue experimentation), from this disclosure and the documents cited herein.

The immunologically effective compositions of S this invention are prepared by mixing the ingredients following generally accepted procedures. For example the selected components may be simply mixed in a blender, or I other standard device to produce a concentrated mixture 4 which may then be adjusted to the final concentration and viscosity by the addition of water or thickening agent and possibly a buffer to control pH or an additional 4 4 solute to control tonicity. Generally the pH may be from about 3 to 7.5. Compositions can be administered in dosages and by techniques well known to those skilled in .2 the medical and veterinary arts taking into consideration such factors as the age, sex, weight, and condition of the particular patient or animal, and the composition form used for administration solid vs. liquid).

Dosages for humans or other mammals can be determined .C without undue experimentation by the skilled artisan, from this disclosure, the documents cited herein, the Examples below from the Examples involving mice).

Suitable regimes for initial administration and booster doses or for sequential administrations also are 3 variable, may include an initial administration followed by subsequent administrations; but nonetheless, may be

~I

ascertained by the skilled artisan, from this disclosure the documents cited herein, and the Examples below.

PCR techniques for amplifying sample DNA for diagnostic detection or assay methods are known from the 6 art cited herein and the documents cited herein (see Examples), as are hybridization techniques for such methods. And, without undue experimentation, the skilled artisan can use gene products and antibodies therefrom in diagnostic, detection or assay methods by procedures OD known in the art.

The following Examples are provided for illustration and are not to be considered a limitation of the invention.

EXAMPLES

1 6 EXAMPLE 1 TRUNCATED STREPTOCOCCUS PNEUMONIAE PspA MOLECULES ELICIT CROSS-PROTECTIVE

IMMUNITY

AGAINST PNEUMOCOCCAL

CHALLENGE

Since the isolation of S. pneumoniae from human S saliva in 1881 and its subsequent connection with lobar .0 pneumonia two years later, human disease resulting from pneumococcal infection has been associated with a significant degree of morbidity and mortality. A recent survey of urgently needed vaccines in the developing and developed world places an improved pneumococcal vaccine among the top three vaccine priorities of industrialized countries. The currently licensed vaccine is a 23 -valent composition of pneumococcal capsular polysaccharides that is only about 60% effective in the elderly and due to poor efficacy is not recommended for use in children 0 below two years of age. Furthermore the growing frequency of multi-drug resistant strains of S.

pneumoniae being isolated accentuates the need for a more effective vaccine to prevent pneumococcal infections.

The immunogenic nature of proteins makes them .3 prime targets for new vaccine strategies. Pneumococcal molecules being investigated as potential protein vaccine candidates include pneumolysis, neuraminidase, autolysin and PspA. All of these proteins are capable of eliciting immunity in mice resulting in extension of life and protection against death with challenge doses near the

LD

50 PspA is unique among these macromolecules in that is can elicit antibodies in animals that protect against S inoculums 100-fold greater than the LD 0 PspA is a surface-exposed protein with an apparent molecular weight of 67-99 kDa that is expressed by all clinically relevant S. pneumoniae strains examined to date. Though PspAs from different pneumococcal 0o strains are serologically variable, many PspA antibodies exhibit cross-reactivities with PspAs from unrelated strains. Upon active immunization with PspA, mice generate PspA antibodies that protect against subsequent challenge with diverse strains of S. pneumoniae. The K' immunogenic and protection-eliciting properties of PspA suggest that it may be a good candidate molecule for a Sprotein-based pneumococcal vaccine.

Four distinct domains of PspA have been .identified based on DNA sequence. They include a N- Sterminal highly charged alpha-helical region, a prolinerich 82 amino acid stretch, a C-terminal repeat segment Scomprised of ten 2 0-amino acid repeat sequences, and a 17-amino acid tail. A panel of MAbs to Rxl PspA have been produced and the binding sites of nine of these Mabs f were recently localized within the Rxl pspA sequence in the alpha-helical region. Five of the Rxl Mabs were protective in mice infected with a virulent pneumococcal strain, WU2. Four of these five protective antibodies were mapped to the distal third (amino acids 192-260) of S3 the alpha-helical domain of Rxl PspA.

Truncated PspAs containing amino acids 192-588 or 192-299, from pneumococcal strain Rxl were cloned and the recombinant proteins expressed and evaluated for their ability to elicit protection against subsequent 3 challenge with S. pneumoniae WU2. As with full-length Rxl PspA, both truncated PspAs containing the distal alpha-helical region protected mice against fatal WU2 I I pneumococcal infection. However, the recombinant PspA fragment extending from amino acid 192 to 588 was more immunogenic than the smaller fragment, probably due to its larger size. In addition, the protection elicited by the amino acid fragment 192-588 of Rxl was comparable to that elicited by full-length Rxl PspA. Therefore, crossprotective epitopes of other PspAs were also sought in the C-terminal two-thirds of the molecule. As discussed below, PspAs homologous to amino acids 192-588 of strain SRxl were amplified by PCR, cloned, and expressed in E.

coli. Then three recombinant PspAs, from capsule type 4 and 5 strains, were evaluated for their ability to confer cross-protection against challenge strains of variant capsular types. The data demonstrate that the truncated PspAs from capsular type 4 and 5 strains collectively protect against or early death caused by challenge with capsular type 4 and 5 parental strains as well as type 3, 6A, and 6B S. pneumoniae.

Bacterial strains and culture conditions. All o pneumococci were from the culture collection of this laboratory, and have been described (Yother, J. et al, Infect. Immun. 1982; 36: 184-188; Briles, et al., Infect. Immun. 1992; 60: 111-116; McDaniel, et al., Microb. Pathog. 1992; 13: 2 61- 2 69; and McDaniel, L.S, et Sal., In: Ferretti, J.J. et al., eds. Genetics of streptococci enterococci and lactococci. 1995; 283- 286), with the exception of clinical isolates TJ0893, 0922134 and BG8740. Pneumococcal strains TJ0893 and 0 922134 were recovered from the blood of a 43-year old c male and an elderly female, respectively. S. pneumoniae BG8743 is a blood isolate from an 8-month old infant.

Strains employed in this study included capsular type 3 (A66.3, EF10197, WU2), type 4 (BG9739, EF3296, EF5668, L81905), type 5 (DBL5), type 6A (DBL6A, EF6796), type 6B (BG7322, BG9163, DBL1), type 14 TJO893), type 19 (BG8090), and type 23 (0922134, BG8743). In addition, strain WG44.1, which expresses no detectable PspA, was

~I

employed in PspA-specific antibody analysis. All chemicals were purchased from Fisher Scientific, Fair Lawn, New Jersey unless indicate otherwise.

S. pneumoniae were grown in Todd Hewitt broth (Difco, Detroit, Michigan) supplemented with 5% yeast extract (Difco). Mid-exponential phase cultures were used for seeding inocula in Lactated Ringer's (Abbott laboratories, North Chicago, Illinois) for challenge studies. For pneumococcal strains used in challenge studies, inocula ranged from 2.8 to 3.8 log 10

CFU

(verified by dilution plating on blood agar). Plates were incubated overnight in a candle jar at 37 0

C.

E. coli DH1 and BL21(DE3) were cultured in LB medium Bacto-tryptone (Difco), 0.5% Bacto Yeast S(Difco), 0.5% NaC1, 0.1% dextrose). For the preparation of cell lysates, recombinant E. coil were grown in minimal E medium supplemented with 0.05 M thiamine, 0.2% i glucose, 0.1% casamino acids (Difco), and 50 mg/ml Skanamycin. Permanent bacterial stocks were stored at o 80 0 C in growth medium containing 10% glycerol.

Construction of Dlasmid-based strains. pET-9a (Novagen, Madison, Wisconsin) was used for cloning truncated pspA genes from fourteen S. pneumoniae strains: DBL6A, WU2, BG9739, EF5668, L81905, 0922134, "4 BG8090, BG8743, BG9163, DBL1, EF3296, EF6796, and EF10197 (Table pspA gene fragments, from fifteen strains were amplified by PCR using two primers provided by Connaught Laboratories, Swiftwater, Pennsylvania Primer N192- 5'GGAAGGCCATATGCTCAAAGAGATTGATGAGTCT3' and primer 3c C588 5'CCAAGGATCCTTAAACCCATTCACCATTGGC 3 were engineered with NdeI and BamHI restriction endonuclease sites, respectively. PCR-amplified gene products were digested with BamHI and NdeI, and ligated to linearized PET-9a digested likewise and further treated with 3 bacterial alkaline phosphatase United States Biochemical Corporation, Cleveland, Ohio) to prevent recircularization of the cut plasmid. Clones were first established in E. coli BL21(DE3) which contained a chromosomal copy of the T7 RNA polymerase gene under the control of an inducible lacUV5 promoter., E. coli DH1 cells were transformed by the method of Hanahan (Hanahan, D. J. Mol. Biol. 1983; 166: 557-580). Stable transformants were identified by screening on LB-kanamycin plates. Plasmid constructs, isolated from each of these strains, were electroporated (Electro Cell Manipulator 600, BTX Electroporation to System, San Diego, California) into E. coli BL21(DE3) and their respective strain designations are listed in Table 1. The pET-9a vector alone was introduced into E. coli BL21(DES) by electroporation to yield strain RCT125 (Table All plasmid constructs and PCR-amplified pspA gene fragments were evaluated by agarose gel electrophoresis (with 1 kb DNA ladder, Gibco

BRL,

Gaithersburg, Maryland). Next, Southern analysis was S. performed using LMpspAl, a previously described full- J. length pspA probe (McDaniel. L.S. et al., Microb. Pathog.

1992; 13: 261-269) random primed labeled with Sdigoxigenin-ll-dUTP (Genius System, Boehringer Mannheim, Indianapolis, Indiana). Hybridization was detected with chemiluminescent sheets according to the manufacturer's instructions (Schleicher Schuell, Keene, New 2 5 Hampshire).

Cell fractionation of recombinant E. co S strains. Multiple cell fractions from transformed

E.

coli were evaluated for the expression of truncated PspA molecules. Single colonies were inoculated into 3 ml LB 30 cultures containing kanamycin and grown overnight at 37C. Next, an 80 ml LB culture, inoculated with 1:100 dilution of the overnight culture, was grown at 37 0 C to mid-exponential phase (A,00 of ca. 0.5) and a 1 ml sample was harvested and resuspended (uninduced cells) prior to i induction with isopropylthiogalactoside (IPTG, 0.3 mM final concentration). Following 1, 2, and 3 hr of induction, 0.5 ml of cells were centrifuges, resuspended, and labeled induced cells. The remaining culture was divided into two aliquots, centrifuged (4000 x g, 10 min, DuPont Sorvall RC 5B Plus), and the supernatant discarded. One pellet was resuspended in 5 ml of 20 mM Tris-HC1 ph 7.4 200 mM NaC1, 1 mM (ethylenedinitrilo)tetraacetic acid disodium salt (EDTA) and frozen at -200C overnight. Cells were thawed at 65 0 C for 30 min, placed on ice, and sonicated for vive 10-sec pulses (0.4 relative output, Fisher Sonic Dismembrator, Dynatech p Laboratories, Inc. Chantilly, Virginia). Next, the material was centrifuged (9000 x g, 20 min) and the supernatant was designed the crude extract-cytoplasmic fraction. The pellet was resuspended in Tris-NaCl-EDTA buffer and labeled the insoluble cell well and membrane fraction. The other pellet, from the divided induced culture, was resuspended in 10 ml of 30 mM Tris-HC1 pH containing 20% sucrose and 1 mM EDTA and incubated at y_ room temperature for 10 min with agitation. Cells were S then centrifuged, the supernatant removed, and the pellet resuspended in 5 mM MgSO, (10 ml, 10 min, shaking 4 0

C

bath). This material was centrifuged and the supernatant was designated osmotic shock-periplasmic fraction. Cell fractions were evaluated by SDS-PAGE and immunoblot S analysis.

MAbs to PsDA. PspA-specific monoclonal antibodies (MAbs) XiR278 and 1A4 were used as previously described (Crain, M.J. et al., 1990, Infect. Immun.; 58: 3293-3299). MAb P50-92D9 was produced by immunization with DBL5 PspA. The PspA-specificity of MAb P50-92D9 was confirmed by Western Analysis by its reactivity with native PspAs from S. pneumoniae DBL5, BG9739, EF5668, and L81095 and its failure to recognize the PspA-control strain WG44.1.

SSDS-PAGE and immunoblot analysis. E. coli cell Sfractions containing recombinant PspA proteins and biotinylated molecular weight markers (low range, Bio- Rad, Richmond, California) were separated by sodium I I dodecyl sulfate-polyacrylamide Bethesda Research Laboratories, Gaithersburg, Maryland) gel electrophoresis (SDS-PAGE) by the method of Laemmli (Laemmli, U.K. Nature 1970; 227: 680-685). Samples were first boiled for 5 min in sample buffer containing 60 mM Tris pH 6.8, 1% 2-Bmercaptoethanol (Sigma, St. Louis, Missouri), 1% SDS, glycerol, and 0.01% bromophenol blue. Gels were subsequently transferred (1 hr, 100 volts) to nitrocellulose (0.45 mM pores, Millipore, Bedford, Massachusetts) as per the method of Towbin et al. Blots were blocked with 3% casein, 0.05% Tween 20 in 10 mM Tris, 0.1 M NaC1, pH 7.4 for 30 min prior to incubating with PspA-specific monoclonal antibodies diluted in PBST for 1 hr at 25 0 C. Next, the blot was washed 3 times with SPBST before incubating with alkaline phosphatase-labeled goat anti-mouse immunoglobulin (Southern Biotechnology Associates, Inc., Birmingham, Alabama) for 1 hr at Washes were performed as before and blots was developed with 0.5 mg/ml 5-bromo-4-chloro-3-indolyl phosphate and S 0.01% nitro blue tetrazolium (Sigma) first dissolved in 150 gl of dimethyl sulfoxide and then diluted in 1.5 M Tris-HC1 pH 8.8. Dot blots were analyzed similarly.

Lysate samples (2 were spotted on nitrocellulose filters (Millipore), allowed to dry, blocked, and detected as just described.

Preparation of cell lysates containing S recombinant PspA proteins. Transformed E. coli strains RCT105, -RCT113, RCT117, and RCT125 (Table 2) were grown in mid-exponential phase in minimal E medium before

IPTG

induction (2 mM final concentration, 2 hours, 37 0

C).

Cultures were harvested by centrifugation (10 min at 9000 x resuspended in Tris-acetate pH 6.9, and frozen at 0 C overnight. Samples were thawed at 65 0 C for 30 min, cooled on ice, and sonicated. Next the samples were "3 treated with 0.2 mM AEBSF (Calbiochem, La Jolla, California) at 37 0 C for 30 min and finally centrifuged to remove cell wall and membrane components. Dot blot analysis was performed using PspA-specific MAbs to validate the presence of recombinant, truncated PspA molecules in the lysates prior to their use as immunogens in mice. Unused lysate material was stored at -20 0 c s until subsequent immunizations were performed.

Mouse immunization and challense.

CBA/CAHN-

XID/J mice (Jackson Laboratories, Bar Harbor, Maine), 6- 12 weeks old, were employed for protection studies.

These mice carry a X-linked immunodeficiency that .0 prevents them from generating antibody to polysaccharide components, thus making them extremely susceptible to pneumococcal infection. Animals were immunized subcutaneously with cell lysates from E coli recombinant strains RCT105, RCT113, RCT117, and RCT125 (Table 2) in t' complete Freund's adjuvant for primary immunizations.

Secondary injections were administered in incomplete adjuvant and subsequent boosts in dHo. Immunized and nonimmunized mice (groups of 2 to 5 animals) were challenged with S. pneumoniae strains A66.3, BG7322, DBL6A, WU2, DBL5, BG9739, and L81905 intravenously (tail vein) to induce pneumococcal sepsis. Infected animals were monitored for 21 days and mice that survived the 3week evaluation period were designated protected against death and scored as surviving 22 days for statistical *S analysis. Protection that resulted in extension of life was calculated as a comparison between mean number of days to death for immunized versus pooled control mice (nonimmunized and RCT125 sham-immunized; total of 6-7 animals).

o Determination of PspA serum levels. Mice were bled retro-orbitally following the secondary boost and again prior to challenge. Representative mouse titers were evaluated by enzyme-linked immunorsorbent assay (ELISA) using native, parental PspAs isolated from pneumococcal strains DBL5, BG9739, and L81905. PspAs were immobilized on microtiter plates by incubating in NaHCO 0.5 M Na 2

CO

3 pH9.5 at 4 0 C overnight. Alkaline I f I phosphatase-labeled goat anti-mouse immunoglobulin (Southern Biotechnology Associates, Inc.) was used to detect mouse serum antibodies. Color development was with p-nitrophenyl phosphate (Sigma, 1 mg/ml) in 0.5 m SMgCL pH 9.8 with 10% diethanolamine and absorbance was read at 405 nm after a 30 min incubation. Reciprocal titers were calculated as the last dilution of antibody that registered an optical density value of 0.1. Sera from individual mice within a particular immunogen group were evaluated separately and then the respective titers from four mice per group were combined to obtain titer range (Table 3).

Statistics. The one-tailed Fisher exact and two sample rank tests were used to evaluate protection against death and extension of life in the mouse model.

Cloninq of truncated PsA genes. Using primers N192 and C588, truncated pspA genes from fifteen diverse pneumococcal strains representing eight different capsular types (Table 1) were amplified by PCR. Even S 2. o though variability exists in pspA genes from different strain, this result demonstrates that sufficient conservation exists between variant pspA genes to allow sequence amplification in all strains examined to date.

Successful pspA PCR-amplification extended to all capsule S types evaluated.

Fourteen of the amplified pspA genes were cloned and three clones containing truncated PspA molecules from pneumococcal strains DBL5, BG9739, and L81905 were further studies (Table To verify the *3 constructions, plasmids from recombinant E. coli strains (RCT105, RCT113, RCT117, and RCT125 (Table 2) were isolated, digested with NdeI and BAMHI restriction endonucleases, and electrophoresed in 1% agarose side-byside with the PCR products used in their respective constructions (Figure 1A). The digestion reaction was complete for pRCT105, while pRCT113 and pRCT117 digestions were incomplete (lanes 5 and 7, respectively).

This gel was denatured and DNA transferred to nylon for Southern analysis. Figure 1B depicts the corresponding Southern blot probed withfull-length RxlpspA DNA. Lane 1 contains pRCT125, digested vector alone, which does not react with the pneumococcal DNA-specific probe, as expected. The pspA-specific probe hybridized with the PCT products and the digested plasmid inserts (see arrow, Figure IB) as well as the partially undigested pRCT113 and pRCT117 (lane 5 and confirming successful cloning o of DBL5, BG9739, and L81905 pspA DNA. Constructions were similarly confirmed with the eleven additional recombinant strains containing truncated pspA genes from S. pneumoniae strains of different capsular and PspA types.

Expression of recombinant PSA in E. coli Bl21(De3). Transformed E. coli strains RCT105, RCT113, SRCT117, and RCT125 were cultured to mid-exponential phase prior to the addition of IPTG to induce expression of the cloned, truncated pspA gene in each strain. A cell RD fractionation experiment was performed to identify the location of recombinant PspA proteins in transformed

E.

coli strains. Samples representing uninduced cells, S. included cells (1 hr, 2 hr, and 3 hr time intervals), the periplasmic fraction, the cytoplasmic fraction, and I insoluble cell wall/membrane material were resolved by SDS-PAGE. Proteins were then transformed to nitrocellulose and Western analysis was performed using monoclonal antibodies specific for PspA epitopes.

Figure 2 reveals that both the cytoplasmic (lane 8) and the insoluble matter fractions (lane 9), from recombinant strain RCT 105, contain a protein of approximately 53.7 kDa that is recognized by MAb XiR278 that is not seen in the uninduced cell sample (lane 3).

This protein increases in quantity in direct correlation 3 with the length of IPTG induction (lanes 4-6; 1 hr, 2 hr, and 3 hr respectively). No truncated RCT105 PspA was found in the periplasmics fraction (lane which was expected since the pET-9a vector lacks a signal sequence that would be necessary for directing proteins to the periplasm. The observed molecular weight (ca. 53.5 kDa) is larger than the predicted molecular weight for the 1.2 Skb DBL5pspA gene product (43.6 kDa; Figure 1A, lane 4).

Like full-length Rxl PspA, the observed and predicted molecular weights for truncated PspAs do not agree precisely. In addition, immunoblot analysis was performed for recombinant E. coli strains RCT113, and to RCT117 (using MAbs 1A4 and P50-92D, respectively) and similar results were obtained, while no cell fractions from control strain RCT125 were recognized by MAb XiR278.

Evaluating the rotective capacity of recombinant truncated PspAs. The truncated PspA proteins from strains RCT113, RCT117, and RCT105 were expressed and analyzed for their ability to generate cross-protection against a battery of seven S. pneumoniae S strains. Control mice (non-immunized and RCT125 shamimmunized) and recombinant PspA-immunized mice were o challenged with mouse-virulent strains A66.3, BG7322, DBL6A, WU2, DBL5, BG9739, and L81905. Table 3 presents the day of death for each infected mouse.

R Immunization with truncated PspA from RCT113, RCT117, and RCT 1 05 conferred protection against death for all mice challenged with capsular type 3 strains (A66.3 and WU2 (Table The three truncated PspAs also provided significant protection against death with DBL6A, and BG7322 pneumococci (capsular types 6A and 6B, respectively). In addition, immunization with recombinant

RCT

1 13 PspA extended days to death in mice challenged with strains DBL5, BG9739, and L81905, while RCT117 PspA prolonged the lives of mice inoculated with BG9739 pneumococci (Table Truncated BG9739 PspA elicited protection against all challenge strains (100%) Sevaluated in this study, while recombinant L81905 and truncated PspAs conferred protection against death with 71% and 57% of S. pneumoniae challenge strains, respectively.

Anti-PspA antibody titers elicited by the three immunogens vary over approximately a 10-fold range (Table S The lowest antibody levels were elicited by RCT105 and this truncated PspA also elicited protection against the fewest number of challenge strains. RCT113 and RCT117 elicited three and nine time as much anti-PspA antibody, respectively. As expected, no antibody to PspA Swas detected in nonimmunized mice nor was specific-PspA antibody measured in mice immunized with the vector-only control strain (RCT125).

In summary, immunization with RCT113 and RCT117 PspAs.protected mice against fatal challenge with capsular type 3 and 6A strains and extended life for mice inoculated with type 4, 5, and 6B pneumococci. RCT105 PspA immunization protected against fatal infection with capsular type 3 and 6B strains and prolonged time to death for type 6A S. pneumoniae but offered not protection against type 4 and 5 strains. These data demonstrate that truncated PspAs from capsular type 4 and pneumococci collectively protect mice and ergo other hosts, such as humans, against or delay death caused by each of the seven challenge strains. In general, i however, more complete protection was observed against Sstrains of capsular type 3, 6A, and 6B than against type 4 and 5 S. pneumoniae.

PspA has been shown to be a protectioneliciting molecule of S. pneumoniae. Immunization with PspA has also been shown to be cross-protective, although eliciting more complete protection against certain strains than others. Thus, it is possible that a broadly protective PspA vaccine might need to contain PspAs of more than one pneumococcal strain. The distal third of 13 the alpha-helical region of PspA has been identified as a major protective region of PspA. Moreover, this region is presented in a very antigenic form when expressed with

~I~

the intact C-terminal half of the molecule. In this Example, the ability to use truncated PspA proteins homologous to the region of RXI PspA extending from amino acid residue 192 to the C-terminus at residue 588 is a demonstrated.

The C-terminal two-thirds of PspA was cloned from fourteen strains by PCR amplification of a gene fragment of the appropriate size (1.2 kb) which hybridized with full-length Rxl pspA. Successful

PCR

;o amplification extended to all capsule types analyzed.

Thus, the C-terminal two-third of PspA may be amplified from many, if not all, pneumococcal capsule types with Rxl pspA-specific primers. This technique is thus applicable to the development of antigenic immunological or vaccine compositions containing multiple PspA or fragments thereof.

Of these clones, three truncated PspA proteins were expressed and evaluated in mouse immunization studies to determine their ability to cross-protect against challenge with a variety of pneumococcal capsular types. All three recombinant PspAs elicited antibody reactive with their respective donor PspA and all three elicited protection against pneumococcal infection. Of the two truncated PspA proteins that elicited the highest S4s antibody responses, 100% and 71% of the challenge strains were protected. RCT105 PspA, which elicited the lowest titers of PspA-specific antibody, yielded protection against 57% of S. pneumoniae strains evaluated. With all truncated PspAs, significant levels of protection were S observed in four of the seven challenge strains. In fact, in all instances except for on (RCT105-immunized mice challenged with strain BG9739) the trend was for truncated PspA-immunization to elicit protection against pneumococcal challenge. These results demonstrate that S truncated Rxl PspA (amino acids 192-588) cross-protects mice against fatal S. pneumoniae WU2 challenge. More importantly, these data show that the homologous regions of -diverse PspAs demonstrate comparable cross-protective abilities.

Strains of capsular type 4 and 5 were more difficult to protect against than were type 3, 6A and 6B 56 pneumococcal strains, Serological differences in PspAs might affect cross-protection in some cases. Yet the difficulty in protecting against the type 4 and 5 strains used herein could not be explained on this basis, since the truncated PspA immunogens were cloned from the same ic three type 4 and 5 strains used for challenge. Both PspAs from the type 4 strains delayed death caused by one or both type 4 challenge strains but neither could prevent death caused by either type 4 pneumococcal strain. Moreover, the truncated PspA from the type strain DBL5 elicited protection against death or delayed death with strains of capsular types 3, 6A and 6B but failed to protect against infection with its donor strain or either type 4 challenge strain.

.:..There may be several reasons why the truncated ~PspAs from capsular type 4 and type 5 strains failed to protect against death even with their homologous donor

S.

pneuraoniae strains. one possibility is that the type 4 and 5 strains chosen for study are especially virulent in the. XID mouse model. XID mice fail to make antibodies to polysaccharides and are therefore extremely susceptible to pneumococcal infection with less than 100*CFU of most strains, including those of capsular type 3, 4, 5, 6A, and 6B. The increased mouse virulence of types 4 and is apparent from the fact that in immunologically normal Smice these strains have lower

LD

50 s and/or are more consistently f atal than strains of capsular types 3, 6A, or 6B.

Another possibility is that epitopes critical to protection-eliciting capacity with capsular type 4 and 53 strains are not present in the C-terminar: two-thirds of PspA (amino acids 192-588), the truncated fragments used for immunization. The critical epitopes for these strains may be located in the N-terminal two thirds of the alpha-helical region of their PspA molecules.

Finally, it is also possible that PspA may be less exposed on some S. pneumoniae strains than others.

Strain Rxl PspA amino acid sequence does not contain the cell wall attachment motif LPXTGX described by Schneewind et al. found in many gram-positive bacteria. Rather, PspA has a novel anchoring mechanism that is mediated by choline interactions between pneumococcal membraneassociated lipoteichoic acid and the repeat region in the C-terminus of the molecule. Electron micrographic examination has confirmed the localization of PspA on the pneumococcal surface and PspA-specific MAb data supports the accessibility of surface-exposed PspA. However, it is.not known whether S. pneumoniae strains differ substantially in the degree to which different PspA regions are exposed to the surrounding environment. Nor is it known if the quantity of PspA expressed on the :i bacterial cell surface differs widely between strains.

4 6 6 43 Table i. nspA recombinant strains categorized by prleumococcal capsular type.

Capsular Parent Strains Respective Type Recombinant Strains 3 WU2, EF10197 RCTl1i, RCT137 4 BG9739, EF5668 RCT1l3, RCT115 L81905, EF'3296 RCTl17, RCT133 RCT105 6A DBL6A, EF6796 RCT1O9, RCT135 6B BG9163, DBLl RCTl29, RCTl31 14 TJ0893* none* 19 BG8090 RCT121 23 0922134, BG8743 RCTll9, RCT123 Truncated PSPA amplified recently, not yet cloned Table 2. Description of recombinant strains used in evaluating the protection-eliciting capacity of truncated PspAs in mice.

RcmiatDescription Capsule Tp Strain of Parent PspA *.RCT 105 BL21(DE3) _.coli with pET-9a: RCT 113 BL21(DE3) .Qoji 4 with pET-9a:BG9739 RCT 117 BL21(DE3) L~ol with 4 pET-9a: L81905 RCT 125 BL21(DE3) 9 _cl with -ET-9a (vector only) Table 3. Evaluation of the protection elicited by truncated S nunaePspA molecules in mice by days to death post-challenge Challenge Strain [capsular type] log,, dose in CFU) ImmunIzIng ReclproWa A66.3: W12 BL6 BG7322 :]15 BG9739 181906 recornblnant afLtI-PSPA [type 31 [type 31 [typ SAI tp 6B] 1typ 5] [type 4) typ 41 ~PSPIPSP~A tfler (3.57) (3.24])3.1 (3.81) (3.56) (3.62) donor strain RCT113/BG9739 559D-50.300 0x>21' 4x>21§ 15.3x>21±t 12.13.t6.>21±t 3,3,4,5' 5.5.5,7' 15.6.8.8' FICT117/LS1905 5590-150,900o 4x>21' 0x>219 7,16.2x>21' 10.12,13, >21' 3,3,4.41 4.5,13, >21' 3,4.6.8 1860-16.770 4X>21' 4x>21§ 8,10.13,21t 4x>21t 2.2.2. >21 2.2.2.4 4,5,5S, RCT125/ 20-02 368> 21 2.3.>21 3.6,6.6 7.8.6,14 2.2.2.2 2.2,3.4.6 2.3.5.5 vector only none 0 2.2.2 2Z3 3.3.4 6.7.9 2.5 3,5 12.5 1 Animals surviving the 3-week evaluation period were sacrificed and days to death recorded as 21 days. For statistical analysis, _P values were calculated at 22 days for these fully protected mice.

Range of four sera per group of mice;, titers measured against native donor PspAs t P<.012 E<0.035 E<0.057 Note: One-tailed Fisher exact and two sample rank tests were used for statistical analysis.

EXAMPLE 2 LOCALIZATION OF PROTECTION-ELICITING EPITOPES AND PspA OF S. PNEUMONIAE This Example, the ability of PspA epitopes on two PspA fragments (amino acids 192-588 and 192-299) to elicit cross-protection against a panel of diverse pneumococci is demonstrated. Also, this Example identifies regions homologous to amino acids 192-299 of Rxl in 15 other diverse pneumococcal strains. The DNA encoding these regions was then amplified and cloned.

S The recombinant PspA fragments expressed were evaluated for their ability to elicit cross-protection against a panel of virulent pneumococci.

Bacterial strains and media conditions.

S.

pneumoniae strains were grown in Todd Hewitt broth with 0.5% yeast extract (THY) (both from Difco Laboratories, Detroit, Michigan) at 37 0 C or on blood agar plates containing 3% sheep blood at 37 0 C under reduced oxygen tension. E. coli strains were grown in Luria-Bertani medium or minimal E medium. Bacteria were stored at in growth medium supplemented with 10% glycerol.

E.

coli were transformed by the methods of Hanahan (Hanahan, D. J. Mol. Biol. 1983; 166: 557). Ampicillin (Ap) was Sused at a concentration of 100 gg/ml for E. Coli.

Construction of pIN-III-ompA3 and pMAL-based

E.

Coli strains. Recombinant plasmids pBC100 and pBAR416 that express and secrete pspA fragments from E. Coli were constructed with pIN-III-ompA3 as previously described (McDaniel, L.S. et al., Microb. Pathog. 1994; 1: 323).

The pMAL-p2 vector (New England Biolabs, c Protein Fusion Purification System, catalog #800) was used for cloning pspA gene fragments to amino acids 192- S 299 from strain Rxl and from 7 other S. pneumoniae strains: R36A, D39, A66, BF9739, DBL5, DBL6A, and LM100.

Amplification of the pspA gene fragments was done by the 3 polymerase chain reaction (PCR) as described previously (McDaniel, L.S. et al., Microb. Pathog. 1994; 17: 323) using primers 5'CCGGATCCGCTCAAAGAGATTGATGAGTCTG 3 [LSM4] and 5'CTGAGTCGACTGAGTTTCTGGAGCTGGAGC 3 [LMS6] made with BamHI and Sall restriction endonuclease sites, respectively. Primers were based on the sequence of Rxi PspA. PCR products and the pMAL vector were digested with BAMHI and Sall, and ligated together. Clones were transformed into E. Coli DH5a by the methods of Hanahan.

Stable transformants were selected on LB plates containing 100g/ml Ap. These clones were screened on LB plates containing 0.1 mM IPTG, 80 pg/ml X-gal and 100 gg/ml Ap and replica LB plates with 100 /g/ml Ap to according to the manufacturer's instructions. The strain designations for these constructs are listed in Table 6.

Positive clones were evaluated for the correct pspA gene fragment by agarose gel electrophoresis following plasmid isolation by the methods of Birnboim and Doly (Birnboim, H.C. et al., Nucl. Acids Res. 1979, z: 1513). Southern analysis was done as previously described using a fulllength pspA probe (McDaniel, L.S. et al., Microb. Pathog.

1994; 17: 323), randomly primed labeled with digoxigenin- I 11-dUTP (Genius System, Boehdinger Mannheim, Indianapolis, Indiana) and detected by chemiluminescence.

Exression of recombinant PsA rotein frac2ents. For induction of expression of strains BC100 and BAR416, bacteria were grown to an optical density of approximately 0.6 at 660 nm at 37 0 C in minimal media, and IPTG was added to a final concentration of 2 mM. The cells were incubated for an additional 2 hours at 37 0

C,

S harvested, and the periplasmic contents released by osmotic shock. For strains BAR36A, BAR39, BAR66, BAR5668, BAR9739, BARL5, BAR6A and BAR100, bacteria were o grown and induced as above except LB media 10 mM.

glucose was the culture medium. Proteins from these strains were purified over an amylose resin column according to the manufacturer's instructions (New England Biolabs, Protein Fusion Purification System, Catalog S#800). Briefly, amylose resin was poured into a 10 mL column and washed with column buffer. The diluted I I osmotic shock extract was loaded at a flow rate of approximately 1 mL/minute. The column was then washed again with column buffer and the fusion protein eluted off the column with column buffer containing 10 mM maltose. Lysates were stored at -20oC until further use.

Characterization of truncated Ps A proteins used for immunization. The truncated PspA molecules, controls and molecular weight markers (Bio-Rad, Richmond, CA) were electrophoresed in a 10% sodium dodecyl (SDS) io polyacrylamide gel and electroblotted onto nitrocellulose. Rabbit polyclonal anti-PspA serum and rabbit antimaltose binding protein were used as the primary antibodies to probe the blots.

A direct binding ELISA procedure was used to i' quantitatively confirm reactivities observed by immunoblotting. For all protein extracts, osmotic shock preparations were diluted to a concentration of 3 /g/ml in phosphate buffered saline (PBS), and 100 Il was added to the wells of Immulon 4 microtitration plates (Dynatech C.R Laboratories, Inc., Chantilly, VA). After blocking with 1.5% bovine serum albumin in PBS, unfractionated tissue S culture supernates of individual MAbs were titered in duplicated by three-fold serial dilution through seven wells and developed using an alkaline phosphatase-labeled goat anti-mouse immunoglobulin secondary antibody (Southern Biotech Associates, Birmingham, AL) and alkalinephosphatase substrate (Sigma, St. Louis,

MO).

The plates were read at 405 nm in a Dynatech plate reader after 25 minutes, and the 30% end point was calculated So for each antibody with each preparation.

Immunization and Protection Assays. Six to nine week old CBA/CAHN-XID/J (CBA/N) mice were obtained from the Jackson Laboratory, Bar Harbor, Maine.

CBA/N

mice carry an X-linked immunodeficiency trait, which renders them relatively unable to respond to polysaccharide antigens, but they do respond with normal levels of antibodies against protein antigens. Because of, the absence of antibodies reactive with the phosphocholine determinant of C-polysaccharide in their serum, the mice are highly ,susceptible to pneumococcal infection. Mice immunized with the BCI00 fragment were injected inguinally with antigen emulsified in CFA, giving an approximate dose of 3 Ag of protein per mouse.

Fourteen days later the mice were boosted intraperitoneally with 3 Ag of antigen diluted in Ringer's lactate without adjuvant. Control mice were io immunized following the same protocol with diluent and adjuvant, but no antigen. Mice immunized with the BAR416 fragment were injected with 0.2 ml at two sites in the sublinguinal area with antigen emulsified in CFA. The mice were boosted inguinally fourteen days later with 1i antigen emulsified in IFA and were boosted a second time fourteen days later intraperioneally with 0.2 ml of antigen diluted in Ringer's lactate without adjuvant.

Mice that were immunized with the homologues of RXlBAR41 6 were immunized as described above. The 2 control animals followed the same immunization protocol but received maltose binding protein (MBP) diluted 1:1 in S: CFA for their immunization and were also boosted with MBP.re also boosted with Serum analysis. Mice were retro-orbitally bled with a 75 gi heparinized microhematocrit capillary tube (Fisher Scientific) before the first immunization and then once approximately 2 hours before challenge with virulent pneumococci. The serum was analyzed for the presence of antibodies to PspA by an enzyme-linked c¢ immunosorbent assay (ELISA) using native full-length R36A PspA as coating antigen as previously described (McDaniel, L.S. Microb. Pathog. 1994; 17: 323).

Intravenous infection of mice. Pneumococcal cultures were grown to late log phase in THY.

3 Pneumococci were diluted to 104 CFU based on-the optical density at 420 nm into lactated Ringer's solution. Seven days following the last boost injection for each group, injection for each group, diluted pneumococci were injected intravenously (tail vein) in a volume of 0.2 ml and plated on blood agar plates to confirm the numbers of CFU per milliliter. The final challenge dose was approximately 50-100 times the

LD

50 of each pneumococcal strain listed in Tables 4-6.

The survival of the mice was followed for 21 days.

Animals remaining alive after 21 days were considered to have survived the challenge.

Statistical analysis. Statistical significance io of differences in days to death was calculated with the Wilcoxon two-sample rank test. Statistical significance of survival versus death was made using the Fisher exact test. In each case, groups of mice immunized with PspA containing preparations were compared to unimmunized 16 controls, or controls immunized with preparations lacking PSpA. One-tailed, rather than two-tailed, calculations were used since immunization with PspA or fragments of PspA has never been observed to cause a statistically significant decrease in resistance to infection.

;Coninq into pMAL vector. Using primers based *U on the sequence of Rxl PspA, LSM4 and LSM6, pspA gene S; fragments were amplified by PCR from fifteen out of fifteen pneumococcal strains examined. Seven of the S eleven gene fragments were cloned into pMAL-p2 and transformed into E. coli (Table The correct insert 4 U for each new clone was verified by agarose gel electrophoresis and Southern hybridization analysis.

Plasmids from recombinant E. coli strains BAR36A, BAR39, BAR66, BAR9739, BARLS, BAR6A and BAR100oo were isolated, 3o digested with BamHI and SalI restriction endonucleases and electrophoresed on a 0.7% TBE agarose gel. The gel was then denatured and the DNA transferred to a nylon membrane for southern hybridization. The blot was probed with full-length Rx. pspA DNA at high stringency Sconditions. The cloning of R36A, D39, A66, BG9739, DBL6A and LMI00 pspA DNA into pMal-p2 was confirmed by the recognition of all BamHI and SalI digested

DNA

inserts by the Rxi probe.

Expression and cnnformat-- n of truncated recombinant proteins. The transformed E. coli strains SBAR36A, BAR39, BAR66, BAR9739, BARL5, BAR6A and BAR100 were grown in LB media supplemented with 10 mM glucose and induced with 2 mM IPTG for expression of the truncated PspA protein fused with maltose binding protein. Transformed E. coli strains BCI00 and BAR416 '0 which express PspA fragments fused to the OmpA leader sequence in the pIN-III-ompA3 vector, were grown in minimal medium and induced with 2 mM IPTG for expression.

Both vectors, pIN-III-ompA3 and pMal-p2, are vectors in which fusion proteins are exported to the periplasmic 1 space. Therefore, an osmotic shock extract from the pMal-p2 containing bacteria was then run over an amylose column for purification and resolved by SDS-PAGE western blotting. The western blot of the protein extracts from BAR36A, BAR39, BAR66, BAR9739, BARL5, BAR6A and BAR100 were recognized by a rabbit polyclonal antibody made to strain BCI00 PspA. The apparent Mr of full-length PspA from WU2 is 91.5 kD. The Mr of maltose binding protein is 42 kD and the expected Mr for the PspA portion of the fusion is 12 kD. All extracts exhibited molecular weights that ranged from 54 to 80 kD. This range of molecular weights can be attributed to the variability of pspA among different pneumococcal strains. An ELISA, with plates coated with the various cloned fragments quantitatively confirmed the reactivities that were 3 0 observed in the western blots with all protein extracts.

[Protection and cross-protection a ainst fatal pneumococcal infection elicited by cloned PspA fragments.

CBA/N mice were immunized with the truncated PspA fragment encoded by pBCI00, which is composed of amino acids 192 to 588 of RXl PspA, and challenged with 13 different S. pneumoniae strains representing 7 different capsular types (Table With all 13 strains, the immunization resulted in protection from death or an extended time to death. With 10 of the strains the difference was statistically significant. With strains of capsular types 3, 6A, and 6B, all immunized mice were protected against death. Although there were fewer survivors in the case of capsular types 2, 4, and 5, the immunization with BCIOO resulted in significant increases in times to death.

The BC100 immunization studies made it clear IV that epitopes C-terminal to residue 192 could elicit cross-protection. The BAR416 fragment, which includes amino acids 192-299, could elicit protection from fatal infection with a single challenge strain WU2. This Example shows the ability of BAR416 immunization to protect against the 6 strains that had been best protected against by immunization with BC100.

Immunization with the BAR416 construct resulted in increased time to death for all 6 challenge strains examined (Table BAR416 provided significant protection against death with WU2, A66, BG7322 and EF6796 pneumococci (capsular types 3, 3, 6B and 6A S respectively). It also prolonged the lives of mice challenged with ATCC6303 and DBL6A pneumococci (capsular types 3 and 6A respectively). Serum from mice immunized with the BAR416 fragment yielded a geometric mean reciprocal anti-PspA ELISA titer to full-length Rxl PspA of 750. Mice immunized with BC100 had geometric mean reciprocal titers of close to 2000, while non-immunized mice had anti-PspA titers of The above data indicates that the BAR416 fragment from Rxl elicits adequate cross-reactive immunity to protect against diverse pneumococci and suggests that this region must be serologically conserved among PspAs. This hypothesis was confirmed by immunized with recombinant BAR416 homologous regions from the 7 different clones and then challenging with strain WU2 (Table All 7 immunogens elicited significant protection. PspA fragments from capsular types 2 and 22 and the rough R36A strain elicited complete protection against death with all challenged mice. All of the other immunogens were able to extend the time to death of all the mice with the median days to death being 21 days or >21 days. Serum from mice immunized with the BAR416 homologous fragments had anti-PspA reciprocal titers that ranged from 260 to 75,800 with an average of 5700 while control animals immunized with only maltose binding '0 protein had anti-PspA reciprocal titers of Antibody reactivities. All of the above immunization studies attest to the cross-reactivity of epitopes encoded by amino acids from position 192-299.

This region includes the C-terminal third of the aiS helical region and the first amino acids of the proline rich region. Other evidence that epitopes within this region are cross-reactive among different PspAs comes form the cross-reactivity of a panel of nine MAbs all of which were made by immunization with Rxl PspA. The epitopes of four of the antibodies in the panel reacted with epitopes mapping between amino acids 192-260. The epitopes of the other five MAbs in the panel map between amino acids 1 and 115 (McDaniel, et al., Microb.

Pathog. 1994; 17: 323). Each of these 9 MAbs were tested for its ability to react with 8 different PspAs in addition to Rxl. The 5 MAbs whose epitopes were located within the first 115 amino acids, reacted on average with only 1 other PspA. Three of the 5 in fact, did not react with any of the other 8 PspAs. In contrast the MAbs 3c whose epitopes map between 192 and 260 amino acids each S cross-reacted with an average of 4 of the 8 non-Rxl PspAs, and all of them reacted with at least two non-Rxl PspAs. Thus, based on this limited section of individual epitopes, it would appear that epitopes in the region AS from 192-260 amino acids are generally much more crossreactive than epitopes in the region from 1-115 amino acids.

The BC100 fragment of Rxl PspaA can elicit protection against the encapsulated type 3 strain WU2.

Although the PspAs of the two strains can be distinguished serologically they are also cross-reactive (Crain, et al., Infect. Immun. 1990; 58: 3293).

The earlier finding made it clear that epitopes crossprotective between Rxl and WU2 PspAs exist. The importance of cross-reactions in the region C-terminal to residue 192 is demonstrated in this Example where 13 i mouse virulent challenge strains have been used to elicit detectable protection against all of them. The first indication that epitopes C-terminal to position 192 might be able to elicit cross-protection came from our earlier study where we showed the MAbs Xi64, XiR278, XiR1323, and XiR1325, whose epitopes mapped between amino acids 192 and 260 of strain Rxl PspA, could protect against infection with WU2. Moreover, immunization with PspA S. fragments from 192-588 and 192-299 were able to elicit protection against infection against WU2. This Example So shows that the BC100 Rxl fragment (192-588) elicits significant protection against each of 13 different mouse virulent pneumococci, thereby firmly establishing the ability of epitopes C-terminal to position 192 to elicit a protective response. The observation that a fusion protein containing amino acids 192-299 fused C-terminal to maltose binding protein could also elicit crossprotection, permits the conclusion that epitopes in this 107 amino acid region of PspA are sufficient to elicit significant cross-protection against a number of 3C different strains.

Evidence that a comparable region of other PspAs is also able to elicit cross-protection cam from the studies where sequences homologous to the 192-299 region of Rxl PspA were made from 5 other PspAs. All Sof these fragments elicited significant protection against challenge with strain WU2. These data provide some suggestion for serologic differences in crossprotection elicited by the 192-299 region.

Based on present,evidence, without wishing to be bound by any one particular theory, it is submitted that the PspAs in strains D39, Rxl and R36A are identical. All of the 9 mice immunized with the 192-299 fragments from R36A and D39 survived challenge with WU2.

Only LM100, one of the non-R36A/D39 PspAs, protected as high a percentage of mice from WU2. The difference in i' survival elicited by the R36A/D39 PspAs and the non-Rxl related PspAs was statistically significant.

The data does indicate however, that all of the differences in protection against different strains are not due to differences in serologic cross-reactivity.

SBC100, which is made from Rxl, protected against death in 100% of the mice challenged with 7 different strains of S. pneumonia, but only delayed death with strain D39, which is thought to have the same PspA as strain Rxl.

Thus, some of the differences in cross-protection are So undoubtedly related to factors other than PspA crossreactivity. Whether such factors are related to differences in virulence of the different strains in the hypersuceptible Xid mouse, or differences in requirements for epitopes N-terminal to amino acid 192, or difference in the role of PspA in different strains is not yet S' known.

These results suggest that a vaccine containing only the recombinant PspA fragments homologous with Rxl amino acids 192-299 is effective against pneumococcal infection. Moreover, the results demonstrate that utility of PspA a.a. 192-299, a.a. 192-260 and DNA coding therefor, e.g. primers N192 or 588 (variants of LSM4 and LSM2) as useful for detecting the presence of pneumococciae by detecting presence of that which binds to the amino acid or to the DNA, or which is amplified by the DNA, by using that DNA as a hybridization probe, or as a PCR primer, or by using the amino acids in antibody-binding kits, assays or tests; and, the results demonstrate that a.a. 192-299 and a.a. 192-260 can be used to elicit antibodies for use in antibody-binding kits assays or tests.

e i e i a Table 4 Protection of mice by immunization with BCIOO from RxI PspA ChIlle BCIOO Immunogen Strain'nge Calnule PSPA alive Median dap type dead Survival alive D39 2 25 0/5 0% WL12 3 1 4/0 100% >21 ATCC6303 3 7 5/0 10D% A66 3 13 4/0 100% 21 EF10197 3 5/0 100% 21 EF5668 4 1 12 1/3 251% 9 EF3296 4 20 1/3 25% 5 L81905 4 23 1/3 2 ED% 4 BG9739 4 26 0/4 00/0 6.5 5 33 0/5 0% 5 0/4 BG7322 6 24 4/0 100% >21 D131-6A SA 19 1()0% EF6796 6A 1 4/0 P/O 100% >21 P/O 100% 1 >21 COn1rOI3 0 alive Median days dead survival alive 0/3 (NO 2 0/3 W/0 3 0/5 7 0/3 00/0 1 0/3 0*/0 2 0/3 (r/O 4 0/3

TIO

0/6 (NO 2 0/3 (r/O 2 0/3 0% 2 1/2 33.3% 6 0/3 0% 1 0/3 0%0 7

P

value, 0.02 0.002 0.004 0.03 0.02

N.S.

N.S.

0.02

N.S.

0.02 0.03 0.03 O-=03 Mice were challenged with approximately 103 CFU/mL of each strain §P values were based on comparison of days alive by a one-tailed Wilcoxon 2 sample-rank test Table 5 Protection of mice by immunization with BAR416 from Rx1 PspA BCIOO Immunogen BC0 lmunoe Challenge Slraln* Capsule SlIve dead Survtvaj Medlan days Bit" Ove aliveMedian daysp

P

V.1 I dead~ spe fdedave A 1n ure 1 2 4/1 80% >21 1 0.002 ATC33 37 2/31 40% 113 1 /4l .)ri*L A66 3 13 5/0 100% >21 0/5 2 0.004 BG7322 6 24 3/2 60% >21 0/4 7 0.02 EF6796 6A 1 3/2 60% >21 0/5. 0% 5 0.004 DBL316A 6A 19 0/5 0% 7 0/5 0%2 0.008 Note, mice were challenged with about 103 CFU of each strain.

§P values were based on Comparison of days alive by a one-tailed Wilcoxon 2 sample-rank test

U

Table 6 Protection of mice against S. pneumoniae WU2 by immunization with BAR416 Analogs of 7 PspAs Immunogen Parent Capsule PspA alive %Median days P value* Strain type type Itotal# Survival alive vs. MBPR BAR36A R36A 25 4/4 100% 21 0.002 BAR39 D39 2 25 5/5 100% 21 0.0008 BAR66 A66 3 13 7/8 88% 21 <0.0001 BAR9739 BG9739 4 26 5/8 63% 21 0.0002 DBL5 5 33 4/8 50% 21 0.03 BAR6A DBL6A 6A 19 3/8 60% >21 0.05 BAR100 LMV100 22 ND 5/5 100% >21 0.0008 I4BP -0/8 0% 2 P values were based on comparison of days alove by one-tailed Wilcoxon 2 sample-rank test Note, the PspA fragments used for immunization were cloned from products amplified with primers LSMV4 and LSMV6. In addition to the strains listed above, PCR reactions with LSMV4 and LSM6I amplified products of the appropriate size strains BG9163, WU2, L819051, EF6796, EF5668, BG7376, BG7322 and BG5-8A.

m U U U U 0 S U Table 7 Reactivity of MAbs with PspAs of Different Pneumnococci Donor of test PspA MAb mapping to 1-115 amino acids MAb mapping to 192-260 amino acids Strain Capsule PspA X11 26 XiRI224 XiR1526 XiR35 XiR1 6 XiR1323 X164 XiR27B XIR1325 Type lgG2b 1gm fgG2b lgG2a IgG2a 1gM 1gM IgG1 IgG2a Rtxl rough 2S 44 ATCC101813 3 3 +4 EF10197 3 18 4+ BG9739 4 28 18905 4 23 BG-5-8-A 06A/- BG9163 6B 21 LM100 22 NO WU2 3 1 4 Note, immunoblot analysis was carried out with the nine MAbs from this study from this study against a panel of nine different pneumococcal' strains. Rxl served as a positive control. The results are presented as (strong reaction), (weak, but clearly positive reation), (difficult to detect),.- (no reaction). The PspA of all strains gave a positive reaction with rabbit antiserum against PspA. N.D. means not determined. Mapping of epitopes was to fragments of Mtain Rxl PspA.

EXAMPLE 3 ISOLATION OF PspA AND TRUNCATED

FORMS

THEREOF AND IMMUNIZATION

THEREBY

PspA is attached,to the pneumococcal surface through a choline binding site on PspA. This allows for successful procedures for the isolation of FL-PspA. PspA can be released from the surface of pneumococci by elution with 2 percent choline chloride or by growth in a chemically defined medium (CDM) contained 1.2 percent CC (CDM-ET). Since CDM-ET superatants lack high F' concentrations of choline, the PspA released into them can be adsorbed to a choline (or choline analog) column and isolated be elution from the column with 2 percent choline chloride

(CC).

This Example describes the ability to obtain SPspA by these procedures, and the ability of PspA obtained by these procedures to elicit protection in mice against otherwise fatal pneumococcal sepsis. Native PspA from strains R36A, RX1, and WU2 was used because these strains have been used previously in studies of the S o ability of PspA to elicit protective immunity (see, e.g., Examples infra and supra). The first MAbs to PspA were made against PspA from strain R36A and the first cloned fragments of PspA and PspA mutants came from strain Rxl.

Strain Rxl was derived from strain R36A, which was in turn derived from the encapsulated type 2 strain, D39.

S PspAs from these three strains appears to be identical based on serologic and molecular weight analysis.

Molecular studies have shown no differences in the pspA genes of strains D39, Rxl, and R36A. The third strain 53 that provided PspA in this Example is the mouse virulent capsular type 3 strain WU2. Its PspA is highly crossreactive with that from R36A and Rxl, and immunization with Rxl and D39 PspA can protect against otherwise fatal infections with strain WU2.

3 S. pneumoniae Strains of S. pneumoniae used in this study have been described previously (Table Bacteria were grown in either Todd-Hewitt broth with 0.5 percent yeast extract (THY), or a chemically defined medium (CDM) described previously 32, 43. Serial passage of stock cultures was avoided. Strains were maintained frozen in THY 20 percent glycerol and cultured from a scraping of the frozen culture.

Recovery of PspA from pneumococci PspA is not found in the medium of growing pneumococci unless they have reached stationary phase and autolysis has commenced 36. To release PspA from jO pneumococci three procedures were used. In one approach were grow pneumococci 100 ml of THY and collect the cells by centrifugation at mid-log phase. The pellet was washed three time in lactated Ringer's solution (Abbot Lab. North Chicago, IL), suspended in a small volume of 2 f percent choline chloride in phosphase buffered saline (PBS) (pH incubated for 10 minutes at room temperature, and centrifuged to remove the whole Spneumococci. From immuboblots with anti Pspa MAb Xi126 48 S at serial dilutions of the original culture, the ,2 suspended pellet, and the supernatant, it was evident that this procedure released about half of the PspA originally present on the pneumococci. Analysis of silver stained polyacrylamide gels showed this S supernatant to contain proteins in addition to PspA 36 a The CDM used in the remaining two procedures was modified from that of Van der Rijn 43 For normal growth it contained 0.03% CC. To cause PspA to be released during bacterial growth, the pneumococci were grown in CDM containing 1.2 percent choline chloride (CDM-CC), or in CDM containing 0.03 percent ethanolamine and only 0.000,001 percent choline (CDM-ET). In media lacking a normal concentration of choline the F-antigen and C-polysaccharide contain phosphoethanolamine rather than phosphocholine 49 In CDM-CC and CDM-ET, PspA is released from the pneumococcal surface because of its inability to bind to the cholines in the lipoteichoic acids 36. In addition to releasing PspA from the pneumococcal surface, growth in CDM-CC or CDM-ET facilitates PspA isolation by its other effects on the cell wall. In these media pneumococci do not autolyse 49 thus permitting them to be grown into stationary phase to S maximize the yield of PspA. In these media septation does not occur and the pneumococci grown in long chains 36 49 As the pneumococci reach stationary phase they die cease making Psp^A, and rapidly settle out.

Preliminary studies, using serial dilution dot blots to io quantitate PspA, indicated that the production of PspA ceases at about the time the pneumococci begin to settle out, with the formation of visible strands of the condensed pneumococcal chains. When the pneumococci began to settle out, the medium was recovered by i2 centrifugation at 2900 x g for 20 minutes, and filtered with a low protein-binding filter (.45y Nalgene Tissue Culture Filter #158-0045).

For growth in CDM-CC or CDM-ET, the pneumococci were first adapted to the defined medium and then, in the ;c case of CDM-ET, to very low choline concentrations. To do this, strains were first inoculated into 1 part of THY and 9 parts of CDM medium containing 0.03 percent choline and 0.03 percent ethanolamine. After two subsequent subcultures in CDM containing 0.03 percent choline and 0.03 percent ethanoline (0.1 ml of culture 0.9 ml of pre-warmed fresh medium), the culture was used to inoculate CDM with only 0.003 percent choline (and 0.03 percent ethanolamine). These steps was repeated until the strain would grow in CDM-ET containing 0.000,001 .3c percent choline and 0.03 percent ethanolamine. It was critical that cultures be passed while in exponential growth phase (at about 10 7 CFU.ml). Even trace contamination of the medium by exogenous choline resulted in the failure of the PspA to be released from the 3~ pneumococcal surface 36 Thus, disposable plastic ware was used for the preparation of CDM-ET media and for growth of cultures. Once a strain was adapted to CDM-ET it was froze in 80 percent CDM-ET and 20 percent glycerol at 800C. When grown subsequently the strain was inoculated directly into the CDM-ET.

Isolation of native (full-length) PspA SPspA was isolated from the medium of cells grown in CDM-ET using choline-Sepharose prepared by conjugating choline to epoxy-activated Sepharose".

A

separate column was used for media from different strains to avoid cross-contamination of their different PspAs.

For isolation of PspA from clarified CDM-ET, we used a 0.6 ml bed volume of choline-Sepharose. The column bed was about 0.5 cm high and 1.4 cm in diameter. The flow rate during loading and washing was approximately 3 ml/min. After loading 300 ml CDM-ET supernatant, the column was washed 10 times with 3 ml volumes of 50 mM Tris acetate buffer, pH 6.9 containing 0.25 M NaCl (TAB).

The washed column was eluted with sequential 3 ml volumes of 2 percent CC in TAB. Protein eluted from the column was measured (Bio-Rad protein assay, Bio-Rad, Hercules, CA). The column was monitored by quantitative dot blot.

The loading material, washes, and the eluted material were dot blotted (1 Al) as undiluted, 1/4, 1/16, 1/64, 1/256, and 1/1024 on nitrocellulose. The membranes were S then blocked with 1 percent BSA in PBS, incubated for 1 2 hr with PspA-specific MAbs Xi126 or XiR278, and developed with biotinylated goat-anti-mouse Ig, alkaline phosphatase conjugated streptavidin (Southern Biotechnology Associates Inc. Birmingham, AL), and nitrobluetetrazolium substrate with 5-bromo 4-chloro-3o indoyl phosphate p-toluidine salt (Fisher Scientific, Norcross

GA)

17 The purity of eluted PspA was assessed by silver-stained (silver stain kit, Bio Rad, Hercules,

CA)

S SDS-PAGE gels run as described previously. Immunoblots of SDS-PAGE gels were developed with MAbs Xi126 and XiR2781 7 Isolation of 29 kDa PspA The 29 kDa fragment comprising the N-terminal 260 amino acids of PspA was produced in DH1 E. coli from pJY4306 31 37. An overnight culture of JY4306 was grown in 100 ml of Luria Broth (LB) containing 50g/ml ampicillin.

The culture was grown at 37 0 C in a shaker at 225 rpm.

This culture was used to inoculate 6 one liter cultures that were grown under the same conditions. When the culture O.D. at 600 nm reached 0.7, 12 grams of cells, as wo a wet paste, were harvested at 4 0 C at 12,000 xg. The pellet was washed in 10 volumes of 25 mM Tris pH 7.7 at 0°C and suspended in 600 ml of 20% sucrose, 25 mM Tris pH 7.7 with 10 mM ethylenediamine tetraacetic acid (EDTA) for 10 minutes. The cells were pelleted by centrifugation (8000 xg) and rapidly suspended in 900 ml of 1 percent sucrose with 1 mM Pefabloc SC hydrochloride (Boehringer Mannheim Corp., Indianapolis, IN.) at 0° C.

The suspension was pelleted at 8000 xg at 4°C. The precipitated from the periplasmic extract by 70 percent C saturated ammonium sulfate overnight at 4 0 C for minutes. The precipitated protein was resuspended in ml of 20 mM histidine 1 percent sucrose at pH 6.6 (HSB).

Insoluble materials were removed at 1,000 xg at 4 0 C for 10 minutes. The clarified material was dialyzed versus r HSB, passed through a 0.2gm filter and further purified on a 1 ml MonoQ HR 5/5 column (Pharmacia Biotech, Inc., Piscataway, equilibrated with HSB. The clarified material was loaded on the column at 1 ml/min, and the column was washed with 10 column volumes of HSB. The 3 column was then eluted with a gradient change to 5 mM NaCI per minute at a flow rate of 1 ml/min. As detected .by immuno blot with Xil26, SDS-PAGE and absorbance, PspA eluted as a single peak at approximately 0.27 to 0.30 M NaCI. By SDA-PAGE the material was approximately percent pure. The yield from 6 liters of culture was 2 mg (Bio-Rad protein assay) of recombinant PspA.

Growth of pneumococci for challenge Mice were challenged with log-phase pneumococci grown in THY. For challenge, the pneumococci were diluted directly into lactated Ringer's without prior washing or centrifugation. In inject the desired numbers of pneumococci, their concentration in lactated Ringer's solution was adjusted to an O.D. of about 0.2 at 420 nM (LKB Ultrospec III spectrophotometer). The number of pneumococci present was calculated at 5x10 8 CFU per o0 ml/O.D. and confirmed by colony counts (on blood agar) of serial dilutions of the inoculum.

Immunization, challenge, and bleeding of mice CBA/CAHN/XID/J (CB A/N) and BALB/cByJ (BALB/c) mice were purchased from Jackson Laboratory Bar Harbor, ME. Mice were given two injections two weeks apart and challenged i.v. two weeks later. Injections without CFA were given intrapertioneally in a 0.1 ml of Ringers.

Where indicated, the first injection was given in complete Freund's adjuvant (CFA) consisting of approximately a 1:1 emulsion of antigen solution and CFA oil (Difco, Detroit MI). Antigen in CFA was injected inguinally in 0.2 ml divided between the two hind legs.

All mice were boosted i/p. without adjuvant. When mice were injected with media supernatants or 2 percent choline chloride eluates of whole bacteria, the amounts of material injected were expressed as the volume of media from which the injected material was derived. For example, if the clarified medium from pneumococci grown in CDM-CC or CDM-ET was used for immunization without jo dilution or concentration, the dose was described as 100 pl. If the material was first diluted 1/10, or concentrated 10 fold, the dose was referred to as 10 or 1000 p1 respectively.

ELISA. for antibodies of PspA Specific modifications of previously reported ELISA conditions 7 are described. Microtitration plates (Nunc Maxisorp, P.G.C. Scientific, Gaithersburg MD.) were coated with undiluted supernatants of Rxl and WG44.1 pneumococci grown in CDM-ET or 1 percent BSA in PBS.

Mice were bled retro-orbitally (75 Al) in a heparanized capillary tube (Fisher Scientific, Fair Lawn, The blood was immediately diluted in 0.5 ml of one percent bovine serum albumin in PBS. The dilution of the resultant sera was 1/15 based on an average hematocrit of 47 percent. The sera were diluted in 7 three fold dilution in microtitration wells starting at 1/45. Mab ,o Xil26 was used as a positive control. The maximum reproducible O.D. observed with Xil26 was defined as "maximum The O.D. observed in the absence of immune sera or MAb was defined as "minimum O.D." Antibody titers were defined as the dilution that gives ,4 33 percent of maximum O.D. The binding to the Rxl CDM-ET coated plates was shown to be PspA-specific, since in no case did we observe 33 percent of maximum binding of immune sera or Xil26 on plates coated with WG44.1 CDM-ET or BSA.

Statistical analysis Unless otherwise indicated P values refer to comparisons using the Wilcoxin two-sample rank test to compare the numbers of days to death in different groups.

Mice alive at 21 days were assigned a value of 22 for the 6 sake of calculation. P values of >0.05 have been regarded as not significant. Since we have never observed immunization with PspA or other antigens to make pneumococci more susceptible to infection the P values have been calculated as single tailed tests. To determine what the P value would have been if a two tailed test had been used the values given should be multiplied by two. In some cases P values were given for comparisons of alive versus dead. These were always calculated using the Fisher exact test. All statistical Scalculations were carried out on a Macintosh computer using InStat(San Diego, CA). PspA is the major protection-eliciting component released from pneumococci grown in CDM-ET or CDM-CC, or released from conventionally grown pneumococci by elution with 2% CC.

PspA-containing preparations from pneumococci were able to protect mice from fatal sepsis following i.v. challenge with 3 x 10 3 (100 x LD 50 capsular type 3 S. pneumoniae (Table Comparable preparations from the strains unable to elicit protection. Regardless of the method of isolation the minimum protective dose was derived from pneumococci grown in from 10-30 /l of '0 medium. We also observed 9 that supernatants of log phase pneumococci grown in normal THY or CDM media could not elicit protection (data not shown). This finding is consistent with earlier studies 36 37 indicating the PspA is not normally released in quantity into the medium of growing pneumococci.

Isolated PspA can elicit protection against fatal infection Although PspA was necessary for these preparations to elicit protection it was possible that it did not act alone. Mice were thus, immunized with purified FL-PspA to address this question.

Isolation of FL-PspA from CDM-ET growth medium We isolated the FL-PspA from CDM-ET rather than from CDM-CC medium or a 2 percent choline chloride 16 elution of live cells, because the high levels of choline S, present in the latter solutions prevents adsorption of the PspA to the choline residues on the choline-Sepharose column. PspA for immunization was isolated from strain R36A, as the strain is non-encapsulated and the isolated So PspA could not be contaminated with capsular polysaccharide. As a control we have conducted mock isolations from WG44.1 since this strain has an inactivated pspA gene and produces no PspA. The results shown in Table 10 are typical of isolations from 300 ml .3 of CDM-ET medium from R36A grown pneumococci. We isolated 84 gg of PspA from 300 ml of medium, or about 280 Ag/liter. Based on the dot blot results this appears to be about 75% of the PspA in the original medium, and that CDM-ET from R36A cultures contains about 400 Ag/liter of PspA, or about,0.4 gg/ml.

No serologically detectable PspA was seen in the CDM-ET from WG44.1 cultures. More significantly there was undetectable protein recovered from the choline-Sepharose column after adsorption of CDM-ET from a WG44.1 culture, indicating that PspA is the only protein that could be isolated by this procedure.

0 o Moreover, by silver stained SDS PAGE gel the PspA isolated from R36A appeared to be homogenous (Figure 3).

Although autolysin can also be isolated on choline- Sepharose 20 5 0 we did not expect it to be isolated by this procedure since autolysin is not released from pneumococci grown in choline deficient medium 36 The immunologic purity of the isolated PspA was emphasized by the fact that immunization with it did not elicit any antibodies detectable on plates coated with CDM-ET supernatants of WG44.1.

Loading more than 300 ml on the 0.6 ml bed volume column did not result in an increased yield, which 1.1 suggested that the column capacity had been reached.

However, increasing the depth of the choline-Sepharose bed to greater than 0.5 cm, decreased the amount of PspA a eluted from the column, presumably because of nonspecific trapping of aggregates in the column matrix.

The elution buffer contains 50 mM Tris acetate 0.25 M NaCl and 2% choline chloride. Elution without added NaCl or with 1M Nacl resulted in lower yields. Elution with o less than 1% CC also reduced yields.

Immunization o fmice with purified R36A PspA For immunization we used only the first 3 ml i. fraction of the R36A column. Mice were immunized with two injections of 1, 0.1, or 0.01 pg o fR36A PspA, spaced 3c two weeks apart. As controls, some mice were inoculated with comparable dilutions of the first 3 ml fraction from the WG44.1 column. Purified FL-PspA elicited antibody to PspA at all doses regardless of whether CFA was used as an adjuvant (Table 11). In the absence of CFA the highest levels of antibody, were seen with the 1 gg dose of PspA. In the presence of CFA, however, the 0.1 pg Sdose was as immunogenic as the 1 pg dose.

To test the ability of the different doses of the different doses of PspA to elicit protection against challenge we infected the immunized mice with two capsular type 3 strains, WU2 and A66. Although both of jo these strains are able to kill highly susceptible

CBA/N

XID mice at challenge doses of less than 102, the A66 strain is several logs more virulent when BALB/c mice are used 47,52. The difference in virulence ofA66 and WU2, was partially compensated for by challenging the immunized CBA/N mice with lower doses of strain A66 than WU2.

After immunization of CBA/N mice with 1 and 0.1 gg doese of PspA we observed protection against WU2 challenge regardless of whether or not CFA was used as an adjuvant (Table At the lowest dose, 0.01 Ag PspA, 0most of the mice immunized with PspA CFA lived whereas most immunized with PspA alone did not; however, the difference was not statistically significant. When immunized mice were challenged with the more virulent strain A66 survivors were only observed among mice immunized with the 1 and 0.1 pg doses. There was slightly more protection against fatal A66 infection among mice immunized with CFA than without, but the difference was not statistically significant. When the two sample rank test was used to analyze the time to death of mice infected with A66 we observed a statistically significant delay in the time to death ineach immunized group as compared to the pooled controls.

The 29 kDa N-terminal fracrment of PspA can elicit S Protection a ainst infection when inected With CFA We have compared the immunogenicity, with and without CFA, of an isolated 29 kDa fragment composed of the first 260 amino acids of PspA. Unlike the case with FL-PspA, adjuvant ws required for the 29 kDa fragment to elicit a protective response. This was observed even though the immunizing doses of the 29 kDa antigen used were 10 and 30 pg/mouse, or about 100 and 300 times the minimum does of FL-PspA that can elicits protection in the absence of adjuvant.

Injection with CFA revealed the presence of additional protection eliciting antigen(s) in CDM-CC. and CDM-ET io growth medium but not in the 2 percent choline chloride eluates of live cells The observation that Freud's adjuvant could have such a major effect on the immunogenicity of the 29 kDa fragment (Table 12), prompted us to reexamine the immunogens described in Table 2 to determine if immunization with adjuvant might enhance protection elicited by PspA-containing preparations or provide evidence for protection eliciting antigens in addition to PspA. By using CFA with the primary injection, the does S. 0 of PspA-containing growth medium (CDM-CC and CDM-ET) S required to elicit protection was reduced fro 10-30 1l (Table 9) down to 1 to 3 pl (Table 13). When CFA was S. used as an adjuvant with CDM-CC and CDM-ET from PspA' strains WG44.1 and JY1119 we were able to elicit IG protective immune responses if material from 2100 or more of media were injected. Thus, although there were apparently some protection eliciting components other than PspA in CDC-CC and CDM-ET growth media, PspA remained the major protection eliciting component even in Jo the presence of adjuvant.

One of the media used for injection was CDM-ET in which JY2141 had been grown. This medium elicited protection against WU2 challenge even when injected at doses as low as 1 pl. It should be noted that although this strain does not make full-length PspA,-it secretes a truncated molecule comprising the first 115 amino acids of PspA into the growth medium. Thus, unlike CDM-ET from WG44.1 and JY1119, CDM-ET from JY2141 with 2 percent CC was relatively non-immunogenic even when emulsified with CFA. This result is consistent with the fact that the 115 amino acid N-terminal PspA fragment of JY2141 is not S surface attached 37 and would be expected to be washed away prior to the elution with 2 percent CC.

Extension of studies to BALB/c mice i.p. challen e route The studies above all involve i.v. challenge of CBA/N mice expressing with the XID genetic defect. The W> i.v. route, used in the present studies provides a relevant model for bacteremia and sepsis, but pneumococci have higher LD 50s when injected i.v. than i.p. CBA/N mice are hypersusceptible to pneumococcal infection because of the XIC defect. This genetic defect prevents them from ,i having circulating naturally occurring antibody to phosphocholine. The absence of these antibodies have been shown to make XID mice several logs more susceptible to pneumococci than isogenic mice lacking the immune defect. From the data in Table 14 it is clear, however, &o that immunization with PspA can protect against infection in mice lacking the XID defect even when the challenge is S, by the i.p. route. Thus, there is no reason to suspect ,1 .that the results presented are necessarily dependent on S: the use of the CBA/N XID mouse or the i.v. route.

PspA is highly immunogenic These studies provide the first quantitative S data on the amount of purified FL-PspA that is required to elicit protective immunity in mice. The isolated PspA for these studies was obtained by taking advantage of the ,30 fact that the C-terminal half of PspA binds to cell surface choline 36 The isolated FL-PspA was found to be highly immunogenic in the mouse. Only two injections of 100 ng of PspA in the absence of adjuvant were reqired to elicit protection against otherwise fatal sepsis with greater than 100 Ld 50 of capsulat type 3 S. pneumoniae.

When the first injection was given with adjuvant, doses as small as 10 ng could elicit protective response The potent immunogenicity of PspA, and the ability to isolate it on choline-Sehparose columns provides a demonstration for the possible use of PspA as a vaccine in humans.

A large body of published 17, 29, 37 as well as W unpublished evidence indicates that the major protection eliciting epitopes of PspA are located in the a-helical (N-terminal) half of the molecule. From the present studies, it is clear that immunization with N-terminal fragments containing the first 115 or 260 of the 288 to amino acid a-helical region are able to elicit protection when given with CFA. However, these fragments were not able to elicit protective responses without CFA. In the case of the both the 115 and 260 amino acid fragments, even immunization at 100 times the minimum dose that is C) immunogenic for FL-PspA failed to elicit a protective response. This result is consistent with previous results showing that a fragment composed of the Nterminal 245 amino acids 31, 37 could elicit protection against otherwise fatal pneumococcal infection of mice o. o when the immunization was given with CFA 32 In that study no immunization without CFA was attempted. Even though the C-terminal half of PspA may not contain major protection-eliciting epitopes it appears to contain S sequence important in the immunogenicity of the molecule 2• 5 as a whole, since the full length molecule elicited much S greater protection than the N-terminal fragments. The effect of the C terminal half on antigenicity may be in part that it doubles the size of the immunogen.

Molecules containing the C-terminal half of PspA may also So be especially immunogenic because they exhibit more extensive aggregation than is seen with fragments expressing only the a-helical region 38. Protein S" aggregates are known to generally be more antigenic and less tolerogenic than individual free molecules 54 PspA is the major protection eliciting component of our pneumococcal extracts Evidence that PspA is the major protection eliciting component of the CDM-ET, CDM-CC growth media and the two percent CC eluates was dependent on the use of mutant pneumococci that lacked the ability to produce FL-PspA. More than one pspA mutant strain was used to insure that the failure to elicit protection in the absence of FL-PspA was not a spurious result of non-PspA mutation blocking the production of some other antigen.

Strains WG44.2 and JY1119 contain identical deletions that include the 5' end of the pspA genes and extend about 3 kb upstream of pspA 37 WG44.1 is a mutant of the non-encapsulated strain Rxl and JY1119 was made by /f transforming capsular type 3 strain WU2 with the WG44.1 pspA mutation. In no case were preparations from WG4.1 and JY1119 as efficient at eliciting protection as those from the PspA+ strains. To rule out the possibility that protection elicited by preparations from the PspA+ 4 O strains was elicited by some non-PspA molecule also encoded by a 3kb deletion linked to the mutant pspA genes of WG44.1 and JY1119, we also used strains JY2141 and LM34 2 6 3 7 In these strains the Rxl pspA gene has been insertionally inactivated causing the production of N- Z terminal fragments of 115 and 245 amino acids .respectively. These strains have no other known S mutations. Although Rxl and R36A are closely related non-encapsulated strains, some of the studies included Rxl as the PspA+ control since it is the isogenic partner o to WG44.1, LM34, and JY2141. The N terminal fragments produced by JY2141 and LM34 lack the surface anchor and are secreted into the medium 36. Two percent CC eluates of JY2141 were non-protection eliciting even in the presence of adjuvant. In the absence of adjuvant, CDM-ET $3 from JY2141 was not protection-eliciting. tM34 was tested without CFA in only 3 mice, but gave results consistent with those obtained with Jy2141.

Anticapsular antibodies are known to be protective against pneumococcal infection 19 However, in these studies it is unlikely they account for any of the protection we attributed to PspA. Our challenge strain bore the type 3 capsular polysaccharide and our primary source of PspA was strain R36A, which is a spontaneous non-encapsulated mutant of a capsular type 2 strain 39,4*. The 36A strain has been recently demonstrated to lack detectable type 3 capsule on the Ssurface or in its cytoplasm 5 Furthermore, the CBA/N mice used in most of the studies are unable to make antibody responses type 3 polysaccharide 56 Non-PspA protection components The observaiton that CDM-CC and CDM-ET supernatants of WG44.1 could elicit protection when injected in large amounts with adjuvant, suggested that these supernatants contained at least trace amounts of non-PspA protection eliciting molecules. In the case of preparations containing PspA eluted from the surface of S. live washed pneumococci with 2 percent CC, there was no evidence for any protection eliciting components other than PspA, presumably because the protection-eliciting non-PspA proteins released into the media were removed by the previous washing step. The identity of the protection eliciting molecules in the WG44.1 supernatant are unknown. In this regard, it is of interest that unlike R36A, strain Rxl has been shown to contain a very small amount of cytoplasmic type 3 polysaccharide (but totally lacks surface type 3 polysaccharide 55). This difference from Rxl apparently came about through genetic manipulations in the contruction of Rxl from R36A 39 1 Thus, preparations made from Rxl or from its daughter S strains WG44.1, LM34, or JY2141, could potentially contain small amounts of capsular polysaccharide. For a SS' number of reasons however, it seems very unlikely that the non-PspA protection-eliciting material identified in these studies was type 3 capsular polysaccharide (expressed by the WU2 challenge strain: 1) growth of these strains was either in CDM-CC or CDM-ET, each of which prevent autolysin activity and lysis 57 that would be required to release the small amount of type 3 polysaccharide from the cytoplasm of the Rxl family strains; 2) CBA/N mice made protective responses to the non-PspA antigens, but express XID immune response deficiency which permits responses to proteins, but blocks antibody to most polysaccharides 46, including type 3 capsular polysaccharide 56; and 3) immunogenicity of the non-PspA component required CFA, an adjuvant known to stimulate T-dependent (protein) rather than T-independent (polysaccharide) antibody responses.

A number of non-PspA protection eliciting t3 pneumococcal proteins have been identified: pneumolysin, autolysin, neuraminidase, and PspA which are 52, 36.5, 107 and 37 kDa respectively 21,58,59,6o The non-PspA protection eliciting components reported here could be composed of a mixture of these and/or other nono identified proteins. Attempts to identify lambda clones producing non-PspA protection eliciting proteins as efficacious as PspA have not been successful 25 Isolation of PspA The protection capacity CDM-CC, CDM-ET and iSg material eluted from live cells with 2% CC were similar in. terms of the volume of the original culture from which the injected dose was derived. The major advantage of eluting the PspA from the surface of pneumococci with 2 percent CC is that the pneumococci may be grown in any 30 standard growth medium, and do not have to be first adapted to a defined medium. Moreover, concentration of S PspA can be accomplished by centrifugation of th epneumococci prior to the elution of the PspA. An advantage of using either CDM-CC and CDM-ET media was g- that these media prevented lysis and pneumococci could be grown into stationary phase without contaminating the preparations with cytoplasmic contents and membrane and wall components. A particular advantage of CDM-ET growth medium is that since it lacks high concentrations of choline of PspA contained in it can be adsorbed directly to a choline-Sepharose column for affinity purification.

6 One liter of CDM-ET growth medium contains about 400 .g of PspA, and we were able to isolate about 3/4 of it to very high purity. At 0.1 pg/dose, a liter of CDM-ET contains enough PspA to immunize about 4,000 mice or possibly 40-400 humans. Our present batch size ;o for a single column run is only 300 ml of CDM-ET. This could presumably be increased by increasing the amount of the adsorbent surface by increasing the diameter of the column. Using our present running buffer we have found that a choline-Sepharose resin depth of 0.5 cm was optimal; increases beyond 0.5 cm caused the overall yield to decrease rather than increase, even in the presence of larger loading of R36A CDM-ET.

4 i 4 I St

S.

S

pass. S S 5 0 *5 Table 8 Pneumococcal Strains A. C U apsul~e Type D3 9 R3 6A PspA Expressed full length full length Paten strain 2 non-encapsulated Patent Rxl non-encapsulated I full In"4 WG44. 1 non-encapsulated LM34 non-encapsulated I L ~Jk none Rxl Construction technique clinical isolate non-encapsulated mutuant derived from R3 6A aberrant insertion activation with pKSD3 00 insertional inactivation with pKSD300 insertional References 26,44 23,44,45 26,39,41 26,37 aa-1--245 of Rxla I Rxl JY2 141 jnon-encapsulatedi aa-l-115 of Rxl V Rxl JY119 A66 3 3 full length none WU2 !full length LM434 and LY2141 express fragments containing the first 245 and first 115 amino acids of Rxl PspA respectively.

Table 9 PspA is the major protection-eliciting component in antigen preparations made by three different methods

S.

S S 4 6 4 I. S Preparation Strain Dose as Median Alive: P versus (PapA volume of Days Dead controlsb Status) media in Alive 2% CC eluate R36A 1000 >21 from live (PspA*) 200 >21 cells 20 >21 2 1.5 0.2 all R36S >21 6:2 0.03 JY2141 1000 3,>21 1:1 (aal-115) 200 1 0:2 1 0:2 CDM-CC Rxl 100 >21 9:0 <0.0001 clarified (PspA-) 30 >21 2:0 medium 10 2 1:2 3 2 0:3 All 2,>21 12:6 0.0004 LM34 100 2,2,>21 1:2 WG44.1 100 2 0:9 (papA*) 30 2 0:3 2 0:3 4 2 0:3 WU2 1000 >21 3:0 0.05 (pSpA 100 >21 ALL >21 4:0 0.03 JY1119 1000 4 0:3 (pspA*) CDM-CC 100 2 0:2 CDM-ET R36A 100 >21 8:0 <0.0001 clarified (pspA 10 3,>21 5:5 0.004 medium 1 1.5 0.1 2 0:2 ALL >21 16:12 0.006 JY2141 100 1.5 0:2 (aal-115) 10 1.5 0:2 WG44.1 100 3 0:2 (papA 10 1.5 0:2 None Antigen dose is given as the volume of growth media from which Antigen dose is given as the volume of growth media from which the 0.1 ml of injected material was derive. Each mouse was injected twice i.p. with the indicate doe diluted as necessary in lactated Ringer's injection solution.

Controls used for statistical comparisons: 2% CC, all JY2141; CDM-CC Rxl, all WG44.1; CDM-CC WU2, JY2141 all JY2141.

I/_

C S Table 10 Isolation of PspA from 300.1 of CDM-ET media after growth of R36A or WG44.1 pneumococci' R36A W044.1 Fraction pg total pp max total pg total pa max protein/mi protein reciprocal dot blot protein proteinF reciprocal dot blot' unitsb d per/ml dot blot growth media 13.3 3,990 4 1200 13.7 4,110 <1 fall-through 13.6 4,080 1 300 13.5 4,050 <1 1st wash <1 <1 wash <1 <1 elution 11 26 78 256 770 <1 <1 elution #2 2 6 16 48 <1 <1 elution f3 <1 4 12 <1 <1 total eluted 84 830 The columns were loaded with 300 ml of clarified CDM-ET medium after the growth of R36A or WG44.1. The column was washed with 10 sequential 3 ml fractions of TBA. Elution was with TBA plus 2 percent CC.

S Total g protein or total dot blot units reflect the total in the 300 ml of the loading material or the 3 ml size of the eluted fractions.

MAb XiR278 was used in the immunoblots to detect PspA in dot blots.

Dot blot units were calculated as the reciprocal as the reciprocal dot titer times the volume in ml.

mra. S S

S

Table 11 Purified full-length PspA is able to elicit protection against fatal sepsis in nice Antigen Doze, Adjuvant or Diluent Anti-PapA titer b (Log mean Challenge with Challenge with 105" WtJ2 10O'2 A. 66 Alive: Median P vs.

Dead Days pooled Alive: Median P vs.

Dead Days pooled Alive controlac control- R36A (PspA+) 1 pg 0.1 0.01 WG44. I (PapA7) 1 pg 0.1 0.01 3600 p1 360 36 3600 p1 360 36 Ringer's Ringer's Ringer'a

CFA

CFA

CFA

Ringer's Ringer s Ringer' s

CFA

CFA

CFA

CFA

3.3 ±0.2 2.6 +0.2 2.7 0.2 3.5 0.2 3.6 +0.1 3.1 4-0.2 6 6 6 6 6 6 <1.6 5:0 4:0 1:4 5:0 5:0 4:1 n. d n.d n. d n.d n. d n.d 1:5 >21 >21 4 >21 >21 >21 n.d n. d n.d n.d n.d n.d 4 0.015 0.041 n. s.

0.027 0.013 0.015 2:3 1:4 0:5 3: 2 2: 3 0:5 1:4 0:5 0:5 0:5 1:4 0:5 n. d.

0.002 0. 0032 0. 0058 0.0012 0. 0012 0.0058 n.s .n.s n. s n.s n.es n. s saline pooled controls 6 n. d.

For comparison with the data in Table 2, it should be noted that the 1, 0.1 and 0.01 pg doses were derived from 3600, 360, and 36 p1 of R36A growth media. Equivalent dilutions of the PspA" eluate from strain WG44.1. were injected as controls. The amount of the WG44.1 preparations injected is listed as 3600, 360, and 36 p1 and corresponds to the volume original growth medium from which the doses of WG44.1 was prepared.

b Antibody values were expressed as reciprocal ELISA titer.

C P values calculated by the Wilcoxon two sample rank test. By Kruskal-Wallis nonparanietric ANOVA for the WU2 challenge was significant at P=0.01, for A66 significance was a P<0.000.

I

?r~ a a a a a a

I

r Table 12. The 29 kDa N-terminal fragment of adjuvant to elicit against WU2' Rxl PspA must be injected with ug 29 kDa Adjuvant or Median Days P versus PspA diluent Alive Alive: Dead noneb CFA >21 3 :0 0.0006 3 CFA >21 3: 0 Ringer's 2 0 3 3 Ringer's 2 1: 2 none CFA 2 0: 7 none Ringer's 2 0: 7 S The 29 kDa fragment comprises the first 260 amino acids of PspA.

b For the calculation of P values the 30ug and 3 ug data were pooled; mice immunized with PspA CFA were compared to CFA controls; mice immunized with PspA Ringer's were compared to controls immunized with Ringers's.

Only the statistically significant P values are shown. The calculated P value of PspA CFA versus CFA alone, was 0.0006 by both the Wilcoxon two sample rank test and the Fisher exact test.

82 Table 13 PspA is not the only protection eliciting molecule released from pneumococci by interference with binding to choline on the surface of pneumococci Strain Dose (as Median Alive: P values' Preparation (PspA volume Day Dead status) in ul) Alive P vs. all JY2141 2% CC eulate R36A 1000 >21 2 0 0.02 from live cells (PspA*) 200 >21 5 0 0.02 >21 5 :0 0.02 2 >21 5 :0 0.001 all R36A >21 17: 0 JY2141 1000 >21 2 0 (aa1- 115) 200 1 0: 2 1 0: 2 2 1 0: 2 all JY2141 1 2 6 P versus pooled cont. CDM-CC RxI 1000 >21 3: 0 0.002 clarified (PspA*) 100 >21 3 0 0.002 medium WU2 1000 >21 3: 0 0.002 CFA (PapA*) 100 >21 3 0 0.002 3 >21 3: 0 0.002 WG44.1 1000 >21 5: 1 0.0001 (PspA*) 100 2.5 2 4 0.002 JY1I9 1000 >21 3 :0 0.002 (PspA*) 100 21 3: 0 0.002 CDM-ET R36A 1000 >21 3: 1 0.004 clarified (PspA') 10 >21 4 0 0.004 medium 1 21 3: 1 0.004 CFA 0.2 2 0:4 JY2141 10 >21 2 :0 (aal1-115) 1 >21 all JY2141 >21 4:0 0.004 WG44.1 100 >21 (PspA') 10 2 0: 2 CDM-ET only CFA 2 0: 9 None none 1.5 0: 4 Pooled Controls' 2 0: 13 in Cases where there were not statistically significant results no P value was shown.

b'Pooled Conrols' refers to 'CDM-ET only' Data and *None' data.

V

Table 14 Immunication of BALB/c mice with isolated PspA elicits protection against S. Pneumoniae Antigen Adjuvant Challenge Days to Death P vlSafflst Source Dose' doren Log OFU Route R36A (PspA+) lug CFA 4 i.p. 2,>21,>2l,>2 0.06./0.03 WG44.l (PspA) 1O0ul CFA 4 i.p.1 None CFA 4 i.p. 2,3 2,2,2,4 R36A (PspA+) lug none 6 2,>2l,>21,>2 0.06./O.03 WG44.l (PspAN) lO0ul none 6 .1 None none 6 i.v. 5,7 2,2,2,3 Pooled i.v and imv. or i.p. 0.008/0.0007 i.p. results The lug dose of R36A PspA was Isolated from 100 ul of COM-ET medium. As a control mice were Injected with an corresponding volume of choline-column effluent from a mock isolation of PspA from the PspA* strain WG44.l. The dose of WG44.l material Is expressed as 100 ul since this Is the volume 0DM- ET from which the injected column effluent was derived.

bp values calculated by Wilcoxon two-sample rank test, TSR, or Fisher exact, FE versus pooled controls for each groupd. "Pooled controls" include data obtained with by injection of "WG44.1" and *none'. The and Lv. studies i EXAMPLE 4 EVIDENCE FOR SIMULTANEOUS EXPRESSION

OF

TWO PspAs From Southern blot analysis there has been an issue as to whether most isolates of S. pneumoniae has s two DNA sequences that hybridize with both 5' and 3' halves of Rxl pspA, or whether this is an artifact of Southern blot. When bacterial lysates have been examined by Western blot, the results have always been consistent with the production of a single PspA by each isolate.

i? This Example provides evidence for the first time that two PspAs of different apparent molecular weights and different serotypes can be simultaneously expressed by the same isolate.

Different PspAs frequently share cross-reactive (z epitopes, and an immune serum to one PspA was able to recognize PspAs on all pneumococci. In spite of these similarities, PspAs of different strains can generally be distinguished by their molecular weights and by their Sreactivity with a panel of PspA-specific monoclonal o antibodies (MAbs).

.4 A serotyping system for PspA has been developed which uses a panel of seven MAbs. PspA serotypes are Sdesignated based on the pattern of positive or negative reactivity in immunoblots with this panel of MAbs. Among f a panel of 57 independent isolates of 9 capsular groups/types, 31 PspA serotypes were observed. The large diversity of PspA was substantiated in a subsequent study of 51 capsular serotype 6B isolates from Alaska, provided by Alan Parkinson at the Arctic Investigations Laboratory -30 of the Centers for Disease Control and Prevention. Among these 51 capsular type 6B isolates were observed 22 different PspAs based on PspA serotype and molecular weight variations of PspA.

While most pneumococcal strains appear to have I3 two DNA sequences homologous with both the 5' and 3' halves of pspA, site-specific truncation mutations of Rxl have revealed that one these, pspA, encodes PspA. The other sequence has been provisionally designated as the pspA-like sequence. At present whether the pspA-like sequence makes a gene product is unknown. Evidence that the pspA and pspA-like genes are homologous but distinct groups of alleles comes from Southern blot analysis at high stringencies. Additional evidence that pspA and the pspA-like loci are distinct comes from studies using PCR primers that permit amplification of a single product approximately 2Kb in size from 70% of pneumococci. For the remaining 30% of pneumococci no amplification was r o observed with the primers used.

Evidence for two PspAs: When the strains of MC25-28 were examined with the panel of seven MAbs specific for different PspA epitopes, all four demonstrated the same patterns of Sreactivity (Fig. The MAbs XiR278 and 2A4 detected a PspA molecule with an apparent molecular weight of 190 KDa in each isolate. In accordance with the previous PspA serotyping system, the 190 KDa molecule was designated as PspA type 6 because of its reactivity with '2 XiR278 and 2A4, but none of the five other MAbs in the typing system. Each isolate also produced a second PspA molecule with an apparent molecular weight 82 KDa. The 82 KDs PspA in each isolate was detected only with the MAb 7D2 and was designated as type 34. No reactivity was s5 detected with MAbs Xil26, Xi64, 1A4, or SR4W4. The fact that. all four capsular 6B strains exhibit two PspAs, based on both molecular weights and PspA serotypes, suggested that they might be members of the same clone.

Simultaneous production of both PspAs: o0 Results from the colony immunobloting showed that both PspAs were present simultaneously in each colony of these isolates when grown in vitro. All colonies on each plate of the original culture, as well as all of the progeny colonies from a single colony, reacted with MAbs XiR278, 2A4, and 7D2.

Number of PsPA genes: One explanation for the second PspA molecule was that these strains contained an extra pspA gene.

Since most strains contain a pspA gene and a pspA-like gene it was expected that if an extra gene were present one might observe at least three pspA homologous loci in isolates MC25-28. In Hind III digests of MC25-28 each strain revealed a 7.7 and 3.6 Kb band when probed with plSMpspAl3/2 (Figure 5A). In comparison, when Rxl DNA was digested with Hind III and hybridized with plSMpspA13.2, homologous sequences were detected on 9.1 and 4.2 Kb fragments as expected from previous studies (Figure 5A). Results consistent with only two pspAhomologous genes in MC25-28 were also obtained with digestion using four additional enzymes (Table In previous studies it has been reported that probes for the 5' half of pspA (encoding the alphahelical half of the protein) bind the pspA-like sequence of most strains only at a stringency of around 90%. With chromosomal digests of MC25-28 we observed that the o Rxl probe of pLSMpspA12/6 bound both pspA homologous bands at a stringency of greater than 95 percent. The same probe bound only the pspA containing fragment Rxl at a stringency above 95 percent (Figure Further characterization of the pspA gene was done by RFLP analysis of PCR amplified pspA from each strain. Since previous studies indicated that individual strains yielded only one product, and since the amplification is carried out with primers based on a known pspA sequence, it seems likely that in each case 0S the amplified products represent the pspA rather than the pspA-like gene. When MC25-28 were subjected to this procedure, an amplified pspA product of 2.1 Kb was produced in each case. When digested with Hha 1 digest the sum of the fragments obtained with each enzyme was 3o approximately equal to the size of the 2.1 Kb amplified product (Figure These results suggest that the 2.1 Kb amplified DNA represents the amplified product of only a single DNA sequence. Rxl, by comparison, produced an amplified product of 2.0 Kb and five fragments of 0.76, 0.468, 0390, 0.349 and 0.120, when digested with Hha 1 as expected from its known pspA sequence.

6 The four isolates examined in this Example are the first in which two PspAs have unambiguously been observed. The interpretation that two PspAs are simultaneously expressed by a single pneumococcal isolate is based on the observation that bands of different io molecular weights were detected by different MAbs to PspA. Isolates used in this study were from a group originally selected for study by Brian Spratt because of their resistance to penicillin. It is very likely that all four of the isolates making two PspAs are related since they share PspA serotypes, amplified pspA RFLPs, chromosomal pspA RFLPs, capsule type, and resistance to penicillin.

The interpretation of studies presented here, showing the existence of two PspAs in the four strains 0o MC25-28, must be set in the context of what is know about the serology PspA as detected by Western blots. PspAs of different strains have been shown previously to exhibit apparent molecular weight sizes ranging from 60 to 200 KDa as detected by Western blots. At least part of this a• a difference in size is attributable to secondary structure. Even for the PspA of a single isolate, band of .several sizes are generally observed. Mutation and immunochemistry studies have demonstrated, however, that all of the different sized PspA band from Rxl are made by go a single gene capable of encoding a 69 KDa protein. The heterogeneity of band size on Western blots of PspA made by a single strain appears to be due to both degradation and polymerization.

PspA was originally defined by reciprocal 36' absorption studies demonstrating that a panel of MAbs to Rxl surface proteins each reacted with some protein and later by studies using Rxl and WU2 derivatives expressing various truncated forms of PspA. In both cases it was clear that each MAbs to the PspA of a given strain reacted with the same protein. Such detailed studies have not been done with each of the several hundred human S isolates. It is possible that with some isolates, reactivity of the MAbs with two PspAs may have gone unnoticed. This could have happened if all reactive antibodies detected both PspAs of the same isolate, or if the most prominent migration bands from each of the two ro PspAs co-migrated. With isolates MC25-28 the observation of .two PspAs was possible because clearly distinguishable bands of different molecular weights reacted preferentially with different MAbs.

Applicants favor the interpretation that isolates MC25-28 each make two PspAs, because an alternative possibility, namely, that the 190 KDa PspA detected by MAbs XiR278 and 2A4 might be a dimer of the 84 KDa monomer detected by MAb 7D2, if the epitopes recognized by the different MAbs were dependent on either A the dimeric or monomeric status of the protein, seems unlikely since whenever MAbs react with the PspA of a i* strain, they usually detect both the monomeric and the dimeric forms. No other isolates have been observed where some MAbs detected only the apparent dimeric form -5 of PspA while others detected only the monomeric form.

There could be several possible explanations for the failure to observe two PspAs produced by most strains. 1) All pneumococci might make two pspAs in culture, but MAbs generally recognize only one of them So (perhaps in this isolate there has been a recombination between pspa DNA and the pspA-like locus, thus allowing that locus to make a product detected by MAb to PspA).

2) All pneumococci can have two pspAs but the expression of one of them generally does not occur under in vitro growth conditions. 3) The pspA-like locus is normally a nonfunctional pseudogene sequence that for an unexplained reason has become functional in these isolates.

It seems unlikely that the expression of only a single PspA by most strains is the result of a phase shift that permits the expression of only the pspA or pspA-like gene at any one time, since many of the strains examined repeatedly and consistently produce the same PspA. In the case of strains MC25-28, the appearance of two PspAs is apparently not the result of a phase switch, since individual colonies produced both the type 6 and the type 34 PspAs.

o Presumably in these four strains, the second PspA protein is produced by the pspA-like DNA sequence.

At high stringency, the probe comprising the coding region of the alpha-helical half of PspA recognized both pspA homologous sequences of MC25-28 but not the pspAlike sequence of Rxl. This finding indicates that the pspA-like sequence of MC25-28 is more similar to the Rxl pspA sequence than is the Rxl pspA-like sequence. If the pspA-like sequence of these strains is more similar to pspA than most pspA-like sequences, it could explain why we were able to see the products of pspA-like genes of these strains with our MAbs. The finding of two families of PspAs made in vivo by pneumococci, allows for use of the second PspA in compositions, as well as the use of DNA primers or probes for the second gene for more conclusive detecting, determining or isolating of pneumococci.

Isolates and Bacterial Cell Culture: Pneumococcal isolates described in.these studies were cultured from patients in Barcelona, Spain *o (one adult at Bellvitge Hospital, and three children at San Juan de Dios) between 1986 and 1988 (Table These penicillin resistant pneumococci originally in the collection of Dr. Brian Spratt were shared with applicants by Dr. Alexander Tomasz at the Rockefeller Institute. Rxl is a rough pneumococcus used in previous studies, and it is the first isolate in which pspA was sequenced. Bacteria were grown in Todd-Hewitt broth with yeast extract or on blood agar plates overnight in a candle jar. Capsular serotype was confirmed by cell agglutination using Danish antisera (Statens Seruminstitut, Copenhagen, Denmark) as previously described. The isolates were subsequently typed as 6B by Quellung reaction, utilizing rabbit antisera against 6A or 6B capsule antigen prepared by Dr. Barry Gray.

Bacterial lysates: Cell lysates were prepared by incubating the 0o bacterial cell pellet with 0.1% sodium deoxycholate, 0.01% sodium dedecylsulfate (SDS), and 0.15 M sodium citrate, and then diluting the lysate in 0.5M Tris hydrochloride (pH 6.8) as previously described. Total pneumococcal protein in the lysates was quantitated by the bicinchonic acid method (BCA Protein Assay Reagent; Pierce.Chemical Company, Rockford,

IL).

PspA serotyping: Serotyping of PspA was performed according to previously published methods. Briefly, pneumococcal cell Slysates were subjected to SDS-PAGE, transferred to nitrocellulose membranes, and developed as Western blots using a panel of seven MAbs to PspA. PspA serotypes were assigned based on the particular combination of MAbs with which each PspA was reactive.

Colony Immunoblotting: A ten ml tube of Todd-Hewitt broth with yeast extract was inoculated with overnight growth of MC23 from a blood agar plate. The isolate was allowed to grow to a concentration of 10 7 cells/ml as determined by C0 an O.D. of 0.07 at 590nm. MC23 was serially diluted and spread-plated on blood agar plates to give approximately 100 cells per plate. The plates were allowed to grow overnight in a candle jar, and a single block agar plate with well-defined colonies was selected. Four oz nitrocellulose membranes were consecutively placed on the plate. Each membrane was lightly weighted and left in place for 5 minutes. In order to investigate the possibility of phase-variation between the two proteins detected on Western blots a single colony was picked from the plate, resuspended in ringers, and spread-plated onto a blood agar plate. The membranes were developed as a Western blots according to PspA serotyping methods.

Chromosomal DNA Preparation: Pneumococcal chromosomal DNA was prepared as in Example 9. The cells were harvested, washed, lysed, and digested with 0.5% (wt/vol) SDS and 100g/ml proteinase

K

io at 37 0 C for 1 hour. The cell wall debris, proteins, and polysccharides were complexed with 1% hexadecyl trimethyl ammonium bromide (CTAB) and 0.7M sodium chloride at 65 o c for 20 minutes, then extracted with chloroform/isoamyl alcohol. DNA was precipitated with 0.6 volumes Sisopropanol, washed, and resuspended in 10mM Tris-HCL, ImM EDTA, pH 8.0. DNA concentration was determined by spectrophotometric analysis at 260nm.

Probe preparation: and 3' oligonucleotide primers homologous o with nucleotides 1 to 26 and 1967 to 1990 of Rxl pspA (LSM 13 and LSM2, respectively) were used to amplify the full length pspA and construct probe LSMpspA13/2 from Rxl genomic DNA. 5' and 3' oligonucleotide primers homologous to nucleotides 161 to 187 and nucleotides 1093 to 1117 (LSM 12 and LSM 6, respectively) were used to amplify the variable alpha-helical region to construct probe LSMpspA12/6. PCR generated DNA was purified by i. Gene Clean (Bioll Inc., Vista, CA) and random primelabeled with digoxigenin-ll-dUTP using the Genius 1 Nonradioactive DNA Labeling and Detection Kit as described by the manufacturer (Boehringer Mannheim, Indianapolis,

IN).

DNA electrophoresis: For Southern blot analysis, approximately of chromosomal DNA was digested to completion with a single restriction endonuclease, (Hind III, Kpn 1, EcoR 1, Dra 1, or Pst 1) then electrophoresed on a 0.7% agarose gel for 16-18 hours at 35 volts. For PCR analysis, 5ul of product were incubated with a single restriction endonuclease,, (Bcl i, BamH 1, Pst 1, Sac 1, EcoR 1 Sma 1, and Kpn 1) then electrophoresed on a 1.3% agrose gel for 2-3 hours at 90 volts. In both case, 1 Kb DNA ladder was used for molecular weight makers

(BRL,

Gaithersburg, MD) and gels were stained with ethidium bromide for 10 minutes and photographed with a ruler.

Southern blot hybridization ,o The DNA in the gel was depurinated in 0.25N HCL for 10 minutes, denatured in 0.5M NaOH and 1.5M NsC1 for minutes, and neutralized in 0.5M Tric-HCl (pH 7.2), NaCl and ImM disodium EDTA for 30 minutes. DNA was transferred to a nylon membrane (Micron Separations

INC,

MA) using a POSIBLOT pressure blotter (Strategene, La Jolla, CA for 45 minutes and fixed by UV irradiation.

The membranes were prehybridized for 3 hours at 42oC in formamide, 5X SSC, 5X Denhardt solution, 25mM sodium phosphate (pH 0.5% SDS 3% (wt/vol) dextran sulfate O and 500/g/ml of denatured salmon containing formamide, 5X SSC, 1X Denhardt solution, 20mM sodium phosphate (pH 0.5% SDS, 3% dextran sulfate, 250xg/ml denatured sheared salmon sperm DNA and about of heat-denatured diogoxigenin-labeled probe

DNA.

After hybridization, the membranes were washed twice in 0.1% SDS and 2X SSC for 3 minutes at room temperature.

The membranes were washed twice to a final stringency of 0.1% SDS in 0.3X SSC at 65 0 C for 15 minutes. This 4* procedure yields a stringency greater than 95 percent.

0o The-membranes were developed using the Genius 1 Nonradioactive DNA Labeling and Detection Kit as described by the manufacturer (Boehringer Mannheim, Indianapolis, IN). To perform additional hybridization with other probes, the membranes were stripped in 0.2N NaOH/0.1%SDS at 40 0 C for 30 minutes and then washed twice in 2X SSC.

Polymerase Chain Reaction

(PCR):

and 3' primers homologous with the DNA encoding the N- and C-terminal ends of PspA (LSM13 and LSM2, respectively) were used in these experiments.

Amplifications were made using Tag DNA polymerase, MgC1 2 and 10X reaction buffer obtained from Promega (Madison, WI). DNA used for PCR was prepared using the method previously described in this paper. Reactions were conducted in 50ml volumes containing 0.2mM of each dNTP, and iml of each primer at a working concentration of io 50mM. MgC1, was used at an optimal concentration of 1.75mM with 0.25 units of Tag DNA polymerase. Ten to thirty ng of genomic DNA was added to each reaction tube.

The amplification reactions were performed in a thermal cycler Research, Inc.) using the following three step program. Step 1 consisted of a denaturing temperature of 94 0 C for 2 minutes. Step 2 consisted of 9 complete cycles of a denaturing temperature of 94 0 C for 1 minute, an annealing temperature of 50 0 C for 2 minutes, and an extension temperature of 72 0 C for 3 minutes. Step Sot 3 cycled for 19 times with a denaturing temperature of 94 0 C for 1-minute, an annealing temperature of 60 0 C for 2 minutes, and an extension temperature of 72 0 C for 3 minutes. At the end of the last cycle, the samples were held at 72 0 C for 5 minutes to ensure complete extension.

SB 9 and size estimation: 4( 4Fragment sizes in the molecular weight standard and in the Southern blot hybridization patterns were calculated from migration distances. The standard molecular sizes were fitted to a logarithmic regression model using Cricket Graph (Cricket Software, Malvern, PA). The molecular weights of the detected bands were S* estimated by entering the logarithmic line equation obtained by Cricket Graph into Microsoft Excel (Microsoft Corporation, Redmond, WA) in order to calculate molecular 3, weights based in migration distances observed in the Southern blot.

TABLE Restriction Enzyme Strains Examined Restriction Fragments MC26 MC27 MC28 RXI MC25-MC28 RX1 Hind III 7.7,3.6 9.1, 4.2 Kpn I 11.6, 10.6 10.6, 9.8 EcoR I 8.4, 7.6 7.8, 6.6 Drl++ 2.1,1.1 1.9,0.9 Pst I 14, 6.1 10.0,4.0 TABLE 16 Penicillin Resistant Capsular Serogroup 6 Strains from Spain Isolate Penicillin U a ite8 Hospital 1 1986 sputum Bell vitge MC26 4 1988 ear San Juan de Dios MC27 1 1988 ear San Juan de Dios MC28 2 1988 ?San Juan de Dios i EXAMPLE 5 SOUTHERN BLOT ANALYSIS OF pspAs AND FRAGMENTS OF pap In this example,, Applicants used oligonucleotides derived from the DNA sequence of pspA of 6 S. pneumoniae Rxl both as hybridization probes and as primers in the polymerase chain reaction to investigate the genetic variation and conservation of the different regions of pspA and pspA-like sequences. The probes used ranged in size from 17 to 33 bases and included sequences to representing the minus 35, the leader, the a-helical region, the proline-rich regions, the repeat regions, and the C-terminus. Applicants examined 18 different isolates representing 12 capsular and 9 PspA serotypes.

The proline-rich, repeat, and leader, regions were highly S conserved among pspA and pspA-like sequence.

In the previous Example, it was shown that strain Rxl and most other strains of S. pneumoniae had two homologous sequences that could hybridize with probes encoding the N terminal and C terminal halves of PspA.

.Z This conclusion that these were separate sequences was .supported by the fact that no matter which restriction enzymes was used there were always at least two (generally two sometimes three or four) restriction fragments of Rxl and most other strains hybridized with the pspA probes. When the genome of Rxl was digested with HindIII and hybridized with these, two pspA- S homologous sequences were found to be in 4.0 and 9.1 kb fragments. Using derivative of Rxl which had insertion mutations in pspA, it was possible to determine that the 3C 4.0 kb fragment contained the functional pspA sequence.

The pspA-homologous sequence included within the 9.1 kb band was referred to as the pspA-like sequence. Whether or not the pspA-like sequences makes a product is not rii know, and none has been identified in vitro. Since pspA- ,3 specific mutants can be difficult to produce in most strains, and exist for only a limited number of pneumococcal isolates, this Example identifies oligonucleotide probes that could distinguish between the pspA and pspA-like sequences.

The purpose of this Example was to further define both the conserved and variable regions of pspA, S and to determine whether the central proline-rich region is variable or conserved, and identify those domains of pspA that are most highly conserved in the pspA-like sequence (and ergo, provide oligonucleotides that can distinguish between the two). Oligonucleotides were used I° and are therefore useful as both hybridization probes and as primers for polymerase chain reaction (PCR) analysis.

Hybridization with oliqonucleotide probes The oligonucleotides used in this study were based on the previously determined sequence of Rxl PspA.

Their position and orientation relative to the structural domains of Rxl PspA are shown in Figure 7. The reactivity of these oligonucleotide probes with the pspA and pspA-like sequences was examined by hybridization with a HindIII digest of Rxl genomic DNA (Table 17). As 2o expected, each of the eight probes recognized the pspAcontaining 4.0 kb fragment of the HindIII digested Rxl DNA. Five of the 8 probes (LSM1, 2, 3, 7, and 12) could also recognize the pspA-like sequence of the 9.1 kb band at least at low stringency. At high stringency four of the probes (LSM2, 3, 4 and 5) were specific for the kb.

These 8 probes were used to screen HindIII digest of the DAN from 18 strains of S. pneumoniae at low and high stringency. For comparison to earlier studies 3 each of the strains was also screened using a full-length pspA probe. Table 23 illustrates the results obtained with each strain at high stringency. Table 18 summarizes the reactivities of the probes with the strains at high and low stringency. Strain Rxl is a laboratory .3 derivative of the clinical isolate, D39. The results obtained with both strains were identical. They are listed under a single heading in Table 23 and are counted as a single strain in Table 28. Although AC17 and AC94 are related clinical isolates, they have distinguishable pspAs and are listed separately. All of the other strains represent independent isolates.

SThe only strain not giving at least two pspAhomologous HindIII fragments was WU2. This observation was expected since WU2 was previously shown to have only one pspA-homologous sequence and to give only a single HindIII fragment that hybridizes with Rxl pspA. Even at i0 high stringency 6 of the 8 probes detected more than one fragment in at least one of the 18 strains Tables 18 and 23. Probes LSM7, 10 and 12 reacted with DNA from a majority of the strains and detected two fragments in over 59% of the strains they reacted with. In almost 6 every case the fragments detected by the oligonucleotide probes were identical in size to those detected by the full-length pspA probe. Moreover, the same pairs of fragments were frequently detected by probes from the 3' as well as the 5' half Rxl pspA. These results are 2-C consistent with earlier findings that the pairs of HindIII fragments from individual isolated generally include two separate but homologous sequences, rather than fragments of a single pspA gene.

The differences in the frequency with which the Si oligonucleotides reacted with (at least one fragment) of the strains in the panel was significant at P 0.0001 by 2 x 8 chi square). When the oligonucleotides were compared in terms of their ability to react with both fragments of each strain the P value was also 0.0001.

o0 Table 18 gives the percentage of strains reactive with each probe, the percentage in which only one fragment was reactive, and the percentage in which two (or more) fragments were reactive.

sThe last column in Table 18 give the ratio of 3~ strains that showed one reactive HindIII fragment at high stringency divided by the total number of reactive strains. In this column values of 1 were obtained with probes that only reacted with one band in each reactive strain. Such probes are assumed to be those that are most specific for pspA. 'The lowest values were obtained with probes that generally see two bands in each strain.

S Such probes are assumed to be those that represent regions relatively conserved between the pspA and pspAlike sequences. At high stringency, probes LSM3 and LSM4 detected only a single HindIII fragment in the DAN of strains they reacted with. These findings suggested .o probes LSM3 and LSM4 were generally detecting alleles of pspA rather than the pspA-like sequence. The observation that the fragments detected by LSM3 or LSM4 were also detected by all of the other reactive probes, strengthened the conclusion that these probes generally ,6 detected the pspA rather than the pspA-like sequence.

WU2 has only one pspA-homologous DNA sequence and secretes a serologically detectable PspA. The fact that LSM3 reacts with the single HindIII fragment of WU2 is consistent with the interpretation that LSM3 detects the o pspA sequences. Sequences representing the second proline region (LSM1) and the C-terminus (LSM2) appeared Sto also be relatively specific for the pspA sequences since they were generally detected in only one of the SHindIII fragments of each strain.

Oligonucleotides, LSM12, and LSM10 detected the most conserved epitopes of pspA and generally reacted with both pspA-homologous fragments of each strain (Table 18). LSM7 was not quite as broadly cross-reactive but detected two PspAs in 41% of strains including almost of the strains it reacted with. Thus, sequences representing the leader, first proline region, and the repeat region appear to be relatively conserved not only within pspA but between the pspA and pspA-like sequences.

a. *LSM3, 4, and 5 reacted with the DNA from the smallest J fraction of strains of any oligonucleotide (29 percent), suggesting that the portion of pspA encoding 100 the a-helical region is the least conserved region of pspA.

With two strains BG85C and L81905, the oligonucleotides detected more than two HindIII fragments containing pspA-homologous sequences. Because of the small size of the oligonucleotide probes and the absence of HindIII restriction sites within any of them, it is very unlikely that these multiple fragments were the results of fragmentation of the target DNA within the probed regions. In almost every case the extra oligonucleotides were detected at high stringency by more than one oligonucleotide. These data strongly suggest that at least in these two strains there are 3 or 4 sequences homologous to at least portions of the pspA.

The probes most reactive with these additional sequences are those for the leader, the a-helical region and the proline rich region. The evidence for the existence of these additional pspA-related sequences was strengthened by results with BG58C and L81905 at low stringency where 2o the LSM3 (a-helical) primer picked up the extra 1.2 kb band of L81905 (in addition to the 3.6 kb band) and the LSM7 (proline-rich) primer picked up the extra 3.2 and 1.4 kb bands (in addition to the 3.6 kb band) of BG58C.

SAmplification of pspA The utility of these oligonucleotides as PCR primers was examined by determining if they could amplify fragments of pspA from the genomic DNA of different pneumococcal isolates. Applicants attempted to amplify pspAs from 14 .diverse strains of S. pneumoniae comprising 12-different capsular types using primers based on the Rxl pspA sequence. Applicants observed that the 3' primer LSM2, which is located at the 3' end of pspA, would amplify an apparent pspA sequence from each of the 14 pneumococcal strains when used in combination with 3S LSM1 located in the region of pspA encoding the prolinerich region (Table 19). LSM2 was also used in combination with four other 5' primers LSM1, 3, 7, 8 and 101 12. LSM8 is located 5' of the pspA start site (near the region).

If a predominant sequence of the expected length was amplified that could be detected on a Southern blot with a full-length pspA probe, we assumed that pspA gene of the amplified DNA had homologous sequences similar to those of the pspA primers used. Based on these criteria the primer representing the a-helical sequence was found to be less conserved than the primers to representing the leader, proline, and C-terminal sequences. These results were consistent with those observed for hybridization. The lowest frequency of amplification was observed with LSM8 which is from the Rxl sequence 5' of the pspA start site. This oligonucleotide was not used in the hybridization studies.

Further evidence for variability comes from differences in the sizes of the amplified pspA gene. The Example showed that when PCR primers LSM12 and LSM2 were -o used to amplify the entire coding region of PspA, PCR products from different pneumococcal isolates ranged in size from 1.9 and 2.3 kb (Table 20). The regions within pspA encoding the a-helical, proline-rich, and repeats were also amplified from the same isolates. As seen in Table 20, the variation in size of pspA appeared to come largely from variation in the size of pspA encoding encodes the a-helical region.

Using probes that consisted of approximately the 5' and 3' halves of pspA it has been determined that 9o the portion of pspA that encodes the a-helical regions is less conserved than the portion of pspA that encodes the C-terminal half of the molecule. This Example show using 4 oligonucleotide probes from within each half of the DNA encoding PspA. Since a larger number of smaller probes 6 were used, Applicants have been able to obtain a higher resolution picture of conserved and variable sequences within pspA and have also been able to identify regions 102 of likely differences and similarities between pspA and the pspA-like sequences.

The only strains in which the pspA gene has been identified by molecular mutations are Rxl, D39 and WU2. RXl and D39 apparently have identical pspA molecules that are the result of the common laboratory origin of these two strains. WU2 lacks the pspA-like gene. Thus, when most pneumococci are examined by Southern blotting using full length-pspA as a probe, it to is not possible to distinguish between the pspA and pspAlike loco, since both are readily detected. A major aim of these studies was to attempt to identify conserved and variable regions within the pspA and pspA-like loci.

A

related aim was to determine whether probes based on the Rxl pspA could be identified that would permit one to differentiate pspA from the pspA-like sequence. Ideally such probes would be based on relatively conserved portion of the pspA sequence that was quite different in the pspA-like sequence. A useful pspA specific probe c would be expected to identify the known Rxl and WU2 pspA genes and identify only a single HindIII fragment in most other strains. Two probes (LSM3 and LSM4) never reacted with more than one pspA-homologous sequence in any S' particular strain. Both of reacted with Rxl pspA and s LSM3 reacted with WU2 pspA. Each of these probes reacted with 4 of the other 15 strains. When these probes identified a band, however, the band was generally also detected by all other Rxl probes reactive with that strain's DNA. Additional evidence that the LSM3 and LSM4 S 05 were restricted to reactivity with pspA was that they reacted with the same bands in all three non-Rxl strains.

Each.probe identifies pspA in certain strains and even I*I when used in combination they recognized pspA in over percent of strains. Probes for the second proline-rich 3 region (LSM1) and the C-terminus of pspA (LSM2) generally, but not always, identified only one pspAhomologous sequence at high stringency. Collectively 103 LSM1, 2, 3, and 4 reacted with 16 of the 17 isolates and in each case revealed a consensus band recognized by most to all of the reactive pr6bes.

By making the assumption that in different strains the Rxl pspA probes are more likely to recognize pspA than the pspA-like sequences, it is possible to make some predictions about areas of conservation and variability within the pspA and pspA-like sequences.

When a probe detected only a single pspA-homologous sequence in an isolate, it was assumed that it was pspA.

If the probe detected two pspA-homologous sequences, it was assumed that it was reacting with both the pspA and pspA-like sequence. Thus, the approximate frequency with which a probe detects pspA can be read from Table 18 as the percent of strains where it detects at least one pspA-homologous band. The approximate frequency with which the probes detect the pspA-like sequence is the percent.of strains in which two or more pspA-homologous band are detected.

c .Using these assumptions the most variable portion of portion of the pspA gene was observed to be the -35 region and the portion encoding a-helical region.

The most conserved portion of pspA was found to be the repeat region, the leader and the proline rich region.

Although only one probe from the region was used, the high degree of conservation among the 10 repeats in the 4" Rxl sequence makes it likely that other probes for the repeat regions give similar results.

The portion of the pspA-like sequence most 3c similar to Rxl pspA was that encoding the leader sequence, the 5' portion of the proline rich region, and the repeat region, and those portions encoding the Nterminal end of the proline-rich and repeat regions. The repeat region of PspA has been shown to be involved in 3 the attachment to PspA to the pneumococcal surface. The conservation of the repeat region among both pspA and pspA-like genes suggests that if is PspA-like protein is 104 produced, that it may have a surface attachment mechanism similar to that of PspA. The need for a functional attachment site may explain the conservation of the repeat region. Moreover, the conservation in DNA encoding the repeat regions of the pspA and pspA-like genes suggests that the repeat regions may serve as a potential anti-pneumococcal drug target. The conservation in the leader sequence between pspA and the pspA-like sequence was also not surprising since similar S conservation has been reported for the leader sequence of other gram positive proteins, such as M protein of group A streptococci. It is noteworthily, however, that there is little evidence at the DNA level that the PspA lead is shared by many genes other than PspA and the possible ;M gene product of the pspA-like locus.

Although the region encoding the C-terminus of pspA (LSM12) or the 3' portion of the proline-rich sequence (LSM1) appear to be highly conserved within pspA genes, corresponding regions in the pspA-like sequences .Z are either lacking, or very distinct from those in pspA.

The reason for conservation at these sites is not 'I.S apparent. In the case of the PspA, its C-terminus does not appear to be necessary for attachment, since mutants lacking the C-terminal 49 amino acids are apparently as I tightly attached to the cell surface as those with the complete sequence. Whether these difference from pspA portends a subtle difference in the mechanism of attachment of proteins produced by these two sequences in unknown. If the C-terminal end of the pspA-like 30 sequence, or the 3' portion of the proline-rich sequence S. in the pspA-like sequence are as conserved within the pspA-like family of genes as it is within pspA, then this region of pspA and the pspA-like sequence serve as targets for the development of probes to distinguish between all pspA and pspA-like genes.

With two strains, some of the oligonucleotide probes identified more than two pspA-homologous 105 sequences. In the case of each of these strains, there was a predominant sequence recognized by almost all of the probes, and two or three additional sequences that were each recognized by at least two of the probes. One interpretation of the data is that there may be more than two pspA-homologous genes in some strains. The significance of such sequences is far from established.

It is of interest however, that although the additional sequences is far from established. It is of interest io however, that although the additional sequences share areas of homology with the leader, a-helical, and proline region, they exhibited no homology with the repeat region of the C-terminus of pspA. These sequences, thus, might serve as elements that can recombine with pspA and/or the ,i pspA-like sequences to generate sequence diversity.

Alternatively the sequences might produce molecules with very different C-terminal regions, and might not be surface attached. If these pspA-like sequences make products, however, they, like PspA, may be valuable as a component of a pneumococcal antigenic, immunological vaccine compositions.

SBacterial strains, growth conditions and isolation of chromosomal

DNA

S. pneumoniae strains used in this study are 4 listed in Table 5. Strains were grown in 100 ml of Todd- Hewitt broth with 0.5% yeast extract at 37 0 C to an q approximate density of 5x10 8 cells/ml. Following harvesting of the cells by centrifugation (2900xg, minutes), the DNA was isolated as previously described o and stored at 4 0 C in TE (10mM Tris, 1mM EDTA, pH Amplification of psA sequences Polymerase chain reaction (PCR) primers, which were also used as oligonucleotide probes in Southern hybridizations, were designed based on the sequence of 36 pspA from pneumococcal strain Rxl. These oligonucleotides were obtained from Oligos Etc.

(Wilsonville, OR) and are listed in Table 22.

106 PCRs were done with a MJ Research, Inc., Programmable Thermal Cycler (Watertown, MA) as previously described using approximately 10 ng of genomic pneumococcal DAN with appropriate 5' and 3' primer pair.

The sample was brought to a total volume of 50 sl containing a final concentration of 50mM KC1, 10mM Tris- HCl (PH 1.5 mM MgCl 2 0.001% gelatin, 0.5 mM each primer, 200mM of each deoxynucleotide triphosphate, and U of Tag DNA polymerase. Following overlaying of the o samples with 50 Al of mineral oil, the samples were denatured at 94 0 C for 2 minutes. Then the samples were subjected to 10 cycles consisting of 1 minute at 94 0 C, 2 minutes at 50 0 C, and 3 minutes at 72 0 C followed by another 20 cycles of 1 minute at 94 0 C, 2 minutes at 50 0

C,

and 3 minutes at 72 0 C followed by another 20 cycles of 1 minute at 94 0 C, 2 minutes at 60 0 C, and 3 minutes at 726C.

After all 30 cycles, the samples were held at 72 0 C for an additional 5 minutes prior to cooling to 4 0 C. The PCR products were analyzed by agarose gel electrophoresis.

o DNA hybridization analysis Approximately 5Ag of chromosomal DNA was digested with HindIII according to the manufacturer's instructions (Promega, Inc., Madison, WI). The digested DNA was electrophoresesed at 35 mV overnight in a 0.8% agarose gels and then vacuum-blotted onto Nytran membranes (Schleicher Schuell, Keene, NH).

Labeling of oligonucleotide with and detection of probe-target hybrids were both performed with the Genius System according to the manufacturer's c instructions (Mannheim, Indianapolis, IN). All hybridizations were done for 18 hours at 42 0 C without formamide. By assuming that 1% base-pair mismatching results in a 1°C decrease in Tm designations of "high" and "low" stringency were defined by salt concentration Sand temperature of post-hybridization washes. Homology between probe and target sequences was derived using calculated Tm the established method. High stringency is 107 defined as 90% or greater homology, arnd low stringency is 80-85% sequence homology.

108 Table 17. Hybridization of oligonucleotideswt HindIll restriction fragments of Rxl DNA.

Stringency Oligonucleotide Region Low High LSM12 Leader N.D. 4.0, 9.1 a-helix N.D. LSM3 a-helix 4.0, 9.1 LSM4 a-helix 4.0 LSM7 Proline 4.0, 9.1 4.0, 9.1 LSM1 Proline 4.0, 9.1 4.0, 9.1 Repeats N.D. 4.0, 9.1 LSM2 C-terminus 4.0, 9.1 Note. Values indicated are the sizes of restriction fragments expressed as kb.

*1 4 en 109 Table 18.

Summary of Hybridization at High and Low Stringency of 8 Oligonucleotides with Hindill Restriction Fragments of the 17 Pneumococcal Isolates Listed in Figure 2 Oligonucleotide Percent Percent Percent 1 band/ with ~1 band with 2 bands with 1 band 1 band Low High Low High Low High Low High LSM12 82 59 24 0.29 29 18 12 0.40 LSM3 65 35 41 0 24 35 0.36 1.00 LSM4 35 29 0 0 35 29 1.00 1.00 LSM7 94 71 71 41 24 29 0.25 0.42 LSM1 100 65 53 12 47 53 0.47 0.82 94 59 35 0.37 LSM2 88 53 41 12 47 41 0.53 0.78 Note, for all values listed all 17 strains were examined.

If no value is listed, then no strains were examined.

110 Table 19. Amplification of Pneumococcal Isolates using the Indicated 5' Prime Combination with the 3' Primer LSM2 at the 3' end of pspA Primer Region Nucleotide Position Amplified/ Percent Tested Amplified LSM8 -35 47 to 70 2/14 14 LSM12 leader 162 to 188 8/14 57 LSM3 a-helical 576 to 598 3/14 21 LSM7 proline 1093 to 1117 12/14 86 LSM1 proline 1312 to 1331 14/14 100 Note, by 2x5 chi square analysis the different primers amplified different amplified different frequencies of pspAs (P 0.0001). The tendency for there to be more amplification with the 3' most primers was significant at P <0.0001.

Si 111 Table 20. Size of amplified pspA fragments in kilobases pspA Primer Pairs number of pspAs Size Range S.D.

Region examined Full length LSM12 LSM2 9 1.9 2.3 0.4 0.17 a-helical LSM12 LSM6 6 1.1 1.5 0.4 0.17 Proline LSM7 LSM9 3 0.23 0 0 Repeats LSM1 LSM2 19 0.6- 0.65 0.05 0.01 Note: amplification was attempted with each set of primers on a panel of 19 different pspAs. Data is shown only for pspAs that could be amplified with the indicated primer pairs.

II

I I I I 112 Table 21 Pneumococcal Strains Strain Relevant characteristics WU2 Capsular type 3, PspA type 1 D39 Capsular type 2, PspA type R36A Nonencapsulated mutant of D39, PspA type Rx1 Nonencapsulated variant of R36A, PspA type Capsular type 5, PspA type 33

DB

L6 A Capsular type 6A, PspA type 19 A66 Capsular type 3, PspA type 13 AC94 Capsular type 9L, PspA type 0 AC17 Capsular type 9L, PspA type 0 Capsular type 9L, PspA type 0 AC107 Capsular type 9V, PspA type 0 AC100 Capsular type 9V, PspA type 0 AC140 Capsular type 9N, PspA type 18 D109-1B Capsular type 23, PspA type 12 BG9709 Capsular type 9, PspA type 0 BG58C Capsular type 6A, PspA type ND L81905 Capsular type 4, PspA type L81905 Capsular type 4, PspA type L82233 Capsular type 14, PspA type 0 L82006 Capsular type 1, PspA type 0 Table 22 PCR Primers Primer Sequence i(5' to 3') LSM 1

CCGGATCCAGCTCCTGCACCAAAC

*LSM2 GCGCGTCGACGGC1TAAACCCATTCACCATTGG LSM3 CCGGATCCTGAGCCAGAGCAGI-rGGCTG LSM4

CCGGATCCGCTCAAAGAGATTGATGAGTCTG

GCGGATCCCGTAGCCAGTCAGTCTAGCTG

LSM6 CTGAGTCGACTGGAG1T17CTGGAGCTGGAGC LSM7

CCGGATCCAGCTCCAGCTCCAGAAACTCCAG

LSM8 GCGGATCC1'TGACCAATAm7ACGGAGGAGGC LSM9 GTT7TTGGTGCAGGAGCTGG LSM 10

GCTATGGCTACAGGTTG

LSM1 1

CCACCTGTAGCCATAGC

LSM 12

CCGGATCCAGCGTGCCTATCUTAGGGGCTGGUT

LSM 13

GCAAGCTTATGATATAGAAATTGTAC

Table 23 Hybridization at high strength of eight different PspA probes with Hind III digests of 18 strains of Streptococcus pneumoniae Strain Probe -I 1 1 MIl/ WU2 085 BI S BL6Aj A,66 AC94 AC17 A CC 1A00O7 AC100 AC140 OC1og BG9709 BG58C LB 195 =3L82 006 L I -In L I I i 3.7,5.8 3.0.3.4 3.6.6.3 3.6.6.3 3.6.6.3 4.0,8.0 t1 LSM12 4.0,9.1 3.7.5.8 3.0,3.4 3.6.6.3 I I

ISMS

3.6,6.3 I I i f i LSM3 LSM4 4.0 4.0 38 f 6.3 4.0.8.0 4.0 3.0,.0 4.0 3.0.4.0 4.0 3.3,4.7 3.3,4.7 22 2.2.

9.6 2.22.

9.6 22 2.2 96 22 1.4.3.2Z 3.8 3.6 3.6 368 3.8 3.6.5.2 3.8 12 2-3.

3.6 3.6 3.6 2.3 3.7 8.2 1.3,3.7 3.? 3.7 4.3, 6.4 i f i i I I I LSM7 4.0.9.1 3.0.3.4 3.2.

I LSMI 4.0.9.1 3.7,5.8 3.2 36 I I I I 52: LSM1O 4.0.9.1 3.8 3.7 3.4 3.6, 3.6.3 3.2 3.6.6.3 4.0 4.0 3.3,4.7 2.2 3.2 f1.7 4 go 3.6 3.6,5.2 6.4 4. 3.

6 3.6 36,6.3 4.0 304...7I Note: All probes were tested versus HindIll digests of all strains. If no bands are listed none were detected. Strains Rxi and D39 gave identical results and are shown In a single column. The full name Of Strain AC109 is ACI1M1B

I

115 EXAMPLE 6 RESTRICTION FRAGMENT LENGTH

POLYMORPHISMS

OF pspA REVEALS

GROUPING

Pneumococcal surface A (PspA) is a protection eliciting protein of Streptococcus pneumoniae. The deduced amino acid sequence of PspA predicts three distinct domains; an a helical coiled-coil region, followed by two adjacent proline-rich regions, and ten amino acid repeats. Almost all PspA molecules are crossreactive with each other in variable degrees. However, iO using a panel of monoclonal antibodies specific for individual epitopes, this protein has been shown to exhibit considerable variability even within strains of the same capsular type. Oligonucleotide primers based on the sequence of pspA from S. pneumoniae Rxl were used to amplify the full-length pspA gene and the 5' portion of the gene including the a-helical and the proline-rich region. PCR-amplified product were digested with Hha I or Sau3A I to visualize restriction fragment length polymorphism of pspA. Although strains were collected a' from around the world and represented 21 different capsular types, isolates could be grouped into 17 families or subfamilies based on their RFLP pattern. The validity of this approach was confirmed by demonstrating that pspA of individual strains which are known to be clonally related were always found within a single pspA family.

Numerous techniques have been employed in epidemiological surveillance of pneumococci which include serotyping, ribotyping, pulsed field electrophoresis, c multilocus enzyme electrophoresis, penicillin-binding protein patterns, and DNA fingerprinting. Previous studies have also utilized the variability of pneumococcal surface protein A (PspA) to differentiate pneumococci. This protein, which can elicit protective 35 antipneumococcal antibodies, is a virulence factor found on all pneumococcal isolates. Although PspA molecules are commonly cross-reactive, they are seldom antigenically identical. This surface protein is the 116 most serologically diverse protein know on pneumococci; therefore, it is an excellent market to be used to follow individual strains. Variations in PspA and the DNA surrounding its structural gene have proven useful for 6 differentiation of S. pneumoniae.

When polyclonal sera are used to identify PspA, cross-reaction is observed between virtually all isolates. Conversely, when panels of monoclonal antibodies are used to compare PspA of independent ,o isolation they are almost always observed to express different combinations of PspA epitopes. A typing system based on this approach has limitations because it does not easily account for differences in monoclonal binding strength to different PspA molecules. Moreover, some strains are weakly reactive with individual monoclonal antibodies and may not always give consistent results.

A less ambiguous typing system that takes advantage of the diversity of PspA was therefore necessary to develop and was used to examine the cc clonality of strains. This method involves examination of the DNA within and adjacent to the pspA locus.

Southern hybridizations of pneumococcal chromosomal

DNA

S' digested with various endonucleases, such as Hind III, Dra I, or Kpn I, and probed with labeled pspA provided a 25 means to study the variability of the chromosome surrounding pspA. When genomic DNA is probed, the pspA and the pspA-like loci are revealed. In most digests the pspA probe hybridizes to 2-3 fragments and, digests of independent isolates were generally dissimilar.

0 ':Like the monoclonal typing system, the Southern hybridization procedure permitted the detection of clones of pneumococci. However, it did not provide a molecular approach for following pspA diversity. Many of the restriction sites defining the restriction fragment S3 length polymorphism (RFLP) were outside 6f the pspA gene, and it was difficult to differentiate the pspA gene from the pspA-like locus. In an effort to develop a system to 117 follow pspA diversity Applicants examined the RFLP of PCR-amplified pspA. Amplified pspA was digested with Sau3A I and Hha I, restriction enzymes with four base recognition sites. To evaluate the utility of this approach pspA from clinical and laboratory strains known to be clonally related as well as random isolates were examined.

Bacterial strains Derivatives of the S. pneumoniae D39-Rxl family :o were kindly provided by Rob Massure and Sanford Lacks (Figure Eight clinical isolates from Spain and four isolates from Hungary, a gift from Alexander Tomasz.

Seventy-five random clinical isolates from Alabams, Sweden, Alaska, and Canada were also studied.

PCR amplifications The oligonucleotide primers used in this study are listed in Table 24. Chromosomal DNA, which was isolated according to procedures described by Dillard et al., was used as template for the PCR reactions.

Amplification was accomplished in a 50 pl reaction containing approximately 50 ng template DNA, .25U Taq, AM of each primer, 175 AM MgC12, and 200 MM dNTP in a reaction buffer containing 10 AM Tris-HCl, pH 9.0, KC1, 0.1% Triton X-100, 0.01% wt/vol. gelatin. The X mixture was overlaid with mineral oil, and placed in a S'"DNA thermal cycler. The amplification program consisted on an initial denaturation step at 94 0 C, followed by 29 cycles opf 94 0 C for 1 min, 55 0 C for 2 min, and 72 0 C for 3 min. The final cycle included an incubation at 720C for o 5 min.

Restriction fragment analysis of PCR-amplified product Aliquots of the PCR mixtures were digested with Hha I or Sau3A I in a final volume of 2 0gl according to manufacturer's protocols. After digestion the DNA 3 fragments were electrophoresed on a 1.3%'TBE agarose gel and stained with ethidium bromide. Fragment sizes were estimated by comparison to a 1kb DNA ladder (Gibco BRL).

118 Because of the variability of pspA, and the fact that the entire pspA sequence is known for only one gene, it has not been possible to design primers which amplify pspA from 100% of pneumococcal strains. However, oligonucleotide primers, LSM2 and LSM1, can amplify an 800 bp region of the C-terminal end in 72 of the 72 stains tested. Based on hybridizations at different stringencies, this region was found to be relatively conserved in pneumococcal strains, and thus would not be S expected to be optimal for following restriction polymorphisms within the pspA molecule. LSM13 and LSM2, primers which amplify the full length pspA gene, can amplify pspA from approximately 79% 55/75 of the strains tested (Table Stability of amplified RFLP pattern within clonally related pneumococci To -determine the stability of pspA during long passages in vitro, we examined the RFLP pattern of the pspA gene of the derivatives of the S. pneumoniae D39-Rxl family. Rxl is an acapsular derivative of S. pneumoniae D39, the prototypical pneumococcal laboratory strain isolated by Avery in 1914. Throughout the 1900's spontaneous and chemical mutations have been introduced into D39 by different laboratories (Figure During this period unencapsulated strains were maintained in vitro, and D39 was passed both in vivo and in vitro passage. All the derivatives of D39, including Rxl, R6, RNC,. and R36A, produced a 1.9 kb fragment upon PCR amplification of full length pspA. All members of the *family exhibited the RFLP pattern. Digestion with Sau3A I of PCR amplified full length pspA revealed a .83, .58, .36 and a .27 kb fragment in all of the D39-rX1 derivatives of the family. Digesting the full length pspA with Hha I resulted in bands which were .76. .47, .39, .35, and .12 kb (Figure 9 or Table 26).

The stability of pspA polymorphism was also investigated using pneumococcal isolates which had 119 previously been shown to be clonally related by other criteria, including capsule type, antibiotic resistance, enzyme electromorph, and PspA serotype. Three sets of isolates, all of which were highly penicillin resistant, were collected from patients during an outbreak in Hungary and two separate outbreaks in Spain. PCR amplified full length pspA from the capsular type 19A pneumococcal strains from the outbreak in Hungary, DB18, DB19, DB20, and DB21, resulted in a band approximately o 2.0 kb. After digesting full length pspA with Hha I, four fragments were visualized., 89, .48, and .28 kb.

Digestion with Sau3A I yielded five fragments .880, .34, and .10kb. Capsule type 6B pneumococcal strains, DB1, DB2, DB3, and DB4, were obtained from an outbreak in Spain. Full length pspA from these strains were approximately 1.9 kb. Digestion of the PCRamplified fragment with Hhs I resulted in four fragments which were .83, .43, .33, and .28 kb. Sau3A I digestion yield a .88, .75, .34, and .10 kg fragments. DB6, DB8, ?o and DB9, which are capsular serotype 23F strains, were isolated from a second outbreak in Spain. DB6, DB8, and DB9 had an amplified pspA product which was 2.0 kb. Hha I digested fragments were .90, .52, .34, and .30 kb and Sau3A I fragments were .75, .52, .39, .22, .20, and kb in size (Figure 10). DB7 had a 19A capsular serotype i o and was not identical to DB6, DB8, and DB9. In the D39/RX1 family and in each of the three outbreak families the size of the fragments obtained from the Hha I and the Sau3A I digests totaled approximately 2.0 kb which is 0 expected if the amplified product represents a single pspA sequence.

Diversity of RFLP pattern of amplified pspA from random pneumococcal isolates PCR amplification of the pspA gene from random clinical pneumococcal isolates yielded full-length pspA ranging in size from 1.8 kb to 2.3 kb. RFLP analysis of PCR-derived pspA revealed two to six DNA I I i 120 fragments ranging in size from 100 bp to 1.9 kb depending on the strain. The calculated sum of the fragments never exceeded the size of the 'original amplified fragment.

Not all pneumococcal strains had a unique pspA, and some seemingly unrelated isolates from different geographical regions and different capsular types exhibited similar RFLP.patterns. Isolates were grouped into families based on the number of fragments produced by Hha I and Sau3A I digests and the relative size of these fragments.

Based on the RFLP patterns it was possible to identify 17 families with four of the families containing pairs of subfamilies. Within families all of the restriction fragments were essentially the same regardless which restriction enzyme was used. The subfamilies represent situations where two families share most but not all the restriction fragments. With certain strains an FRLP pattern was observed where detectable fragment size differed from the pattern of the established family by less than 100 bp. Since the Sdifferences were considered small compared to the differences in the fragment size and the number of fragments between families, they were not considered in •family designation. The RFLP pattern of two isolates S. from six of the families is pictured in Figure 11, Table S* 27. These families were completely independent of the capsular type or the protein type as identified by monoclonal antibodies (Table 28 and 29).

Previous DNA hybridization studies have demonstrated that the pspA gene of different isolates are 3 the most conserved in their 3' region of the gene and more variable in the 5' region of the gene. Thus, if seemed likely that the differences in the pspA families reflected primarily differences in the 5' end of the gene. To confirm this theory, the a helical and proline region of pspA was examined without the amino acid repeats. Nucleotide primers LSM13 and KSH2 were used to amplify this fragment which is approximately 1.6 kb.

Examination of this region of pspA afforded two things.

This primer pair permitted amplification of of the strains which is greater than the 75% of the strains which can be amplified with oligonucleotides which amplify the full length gene. Second, it allowed Applicants to examine if the original groupings which were based on the full length gene coincide with the fingerprint patterns obtained by looking at the 5' half 10 of the gene.

Figure 12 contains the same strains which were examined in Figure 11 but the PCR products were amplified with SKH2 and LSM13. The RFLP patterns obtained from digestion of the amplified a helical and proline rich If region confirms the original designated families.

However, these primers amplify a smaller portion of the psaA and therefore the difference is the families is not as dramatic as the RFLP patterns obtained from the RFLP pattern of the full length gene.

.2c The polymerase chain reaction has simplified the process of analyzing pspA gene and have provided a means of using pspA diversity to examine the epidemiology of S. pneumoniae. Because not all strains contained a unique fingerprint of pspA, RFLP patterns of pspA cannot be used alone to identify the clonality of a strain.

These results indicate the RFLP of PCR-amplified pspA from pneumococcal strains in conjunction with other techniques may be useful for identifying the clonal relatedness among pneumococcal isolates, and that this iC pattern is stable over long passages in vitro.

These findings suggests that the population of pspA is not as diverse as originally believed.

PCR-RFLP

0 of pspA may perhaps represent a relatively simplistic technique to quickly access the variability of the gene within a population. Further, these findings enable S techniques to diagnose. S. pneumoniae via PCR or 122 hybridization by primers on probes to regions of pspA common within groupings.

The sequence studies divide the known strains into several families based on sequence homologies.

Sequence data demonstrates that there have been extensive recombinations occurring in nature within pspA genes.

The net effect of the recombination is that the "families" identified by specific sequences differ depending upon which part of the pspA molecule is used 1o for analysis. "Families" or "grouping identified by the half of the alpha-helical region, the 3' half of the a-helical region and the proline rich region are each distinct and differ slightly from each other. In addition there is considerable evidence of other 1, diversity (including base substitute and deletions and insertions in the sequences) among otherwise closely related molecules.

This result indicates that it is expected that there will be a continuum of overlapping sequences of a o PspAs, rather than a discrete set of sequences.

The findings indicate that there is the greatest conservation of sequence in the 3' half of a ahelical region and in the immediate 5' tip. Because the diversity in the mid half of the a-helical region is 5" greater, this region is of little use in predicting i' cross-reactivity among vaccine components and challenge strains. Thus, the sequence of 3' half of the alphahelical region and the 5' tip of the coding sequence are likely to the critical sequence for predicting PspA SC cross-reactions and vaccine composition.

The sequence of the proline-rich region may not t" be particularly important to composition of a vaccine because this region has not been shown to be able to elicit cross-protection even though it is highly conserved. The reason for this is presumably because antibodies to epitopes in this region are not surface exposed.

123 Based on our present sequences of 27 diverse pspAs we-have found that there are 4 families of the 3' half of the a-helical region and 2-3 families of the very tip the a-helical region. Together these form 6 6 combinations of the 3' and 5' families. This approach therefore should permit us to identify a panel of pspAs with 3' and 5 helical sequences representative of the greatest number of different pspAs. See Fig. 13.

t* a a..

V V -V .6O@ V

V

V. V Relationship of Capsular type and RFLP family.

Table 29.

RELATIONSHIP BETWVEEN CAPSULAR TYPE AND RFLP FAMILY Capsule Type 12 3 4 5 5 6A 6B 7 8 9A 9L 9N 9V 10 11 12 13 14 15 19 22 23 31 3 3 1 1 2 1 2 1 2

G

H

K

KK

L

M

MM

1 1 2 1 1 1 1 2 4 1 1 1 2 2 Table 24. -Oiigonucleotides used in this study.

~tl 125 125 Designation Sequence 5'-3' LSM2 GCG CGI CGA CGG CUT AMA CCC ATT CAC CAT LSM1 COG GAT CCA GCT OCT LSM13 GCA AGCUTA TGA TAT AUT UG TMA C SKH2 OCA CAT ACC GUT TTC G liT CCA GCC Nucleotide position 1990 to 1967 1312 to 1331 I to 26 1333 to 1355J :1 4 I *4 64 a.

S

6** 6 S

S

126 Table 25. Amplification of pspA from a panel of 72 independent isolates* of S.

pneumoniae.

I

I a *Our strain collection contains several groups of isolates known to be previously to be clonal and collected for that purpose. The data reported in the table includes only representative isolate from such clonal groups.

4 49 .0 4 127 Table 26. Rxl-D39 derivatives ISOLATE SIZE OF HMA I DIGESTS SIZE OF Sau3A I DIGESTS (Kb) D39 .76, .47,.39,.35,.12 -83,.58,.36,.27 Rxl .76, .47,.39,.35,.12 .83,.58,.36,.27 R800 .76, .47, .39, .35, .12 .83,.58,.36,.27 R6 .76, .47, .39, .35, .12 .83,.58,.36,.27 R61 .76, .47, .39, .35, .12 .83,.58,8.6,.27 R6X .76, .47,.39,.35,.12 .83, .58, .36, .27 R36NC .76, .47, .39, .35, .12 .83, .58, .36, .27 R36A .76,.47,.39,.35,.12 .83,.58,.36,.27 Table 27. Strain information and family designation of independent isolates.

STRAIN CAPSULE PspA FAMILY SIZE OF Hha I SIZE OF Sau3A I TYPE TYPE FRAGMENTS

FRAGMENTS

BG9163 6B 21 C 1.55, .35 1.05, .35, .22 EF6796 6A 1 C 1.5,.35 1.05, .35, .22 EF5668 4 12 DD 1.25, .49, .32 1.0, .80, EF8616A 4 ND DD 1.25, .49, .32 1.0,.80,.35 EF3296 4 20 E 1.0, .40, .33 1.15, .50, .34 EF4135 4 ND E 1.0, .40, .33 1.15, .50, .34 BG7619 10 ND F 1.3, .40, .29, .10 .82, .76, BG7941 11 ND F 1.3, .40, .29, .10 .82, .76, BG7813 14 8 H 1.05, .70, .36 .90, .77, BG7736 8 ND H 1.05, .70, .36 .90,.77,.35 AC1 13 9A ND I 1.4, .34, .28 1.2, AC99 9V 5 1 1.4,.34,.28 1.2,.80

C

'C Ce *11I "so 0 Table 28. Relationship of RFLP family and PspA type.

RELATIONSHIP BETWEEN PSPA TYPE AND RFLP FAMILY PspA Type pspA 0 1 3 5 8 12 13 16 18 19 20 21 24 25 26 30 33 34 36 37 ND FAMILY A 1

B

C 2 1 14 D11 DD 2 F 114 FF13 H I3 12 21 4I 1

KK

1 3 MM1 129 EXAMPLE 7 ABILITY OF PspA IMMUNOGENS TO PROTECT AGAINST INDIVIDUAL CHALLENGE STRAINS CBA/N or BALB cJ mice were given 1 injection of Ag PspA in CFA, followed 2 weeks later by a boost S in saline, and challenged between 7 and 14 (average days post boost. Control mice were administered a similar immunization regimen, except that the immunization came from an isogeneic strain unable to make PspA. The PspA was either full length, isolated from io pneumococci or cloned full length or BC100 PspA, as little statistical significance has been seen in immunogenicity between full length PspA and BC100. The challenge doses ranged from about 10 3 to 104 pneunocci in inoculum, but in all cases the challenge was at least 100 times

LD

50 The results are shown in the following Tables 60, and the conclusions set forth therein.

From the data, it appears that an antigenic, immunological or vaccine composition can contain any two IC to.seven, preferably three to five PspA, PspAs from R36A and BG9739, alone, or combined with any or all of PspAs from Wu2, Ef5668, and DB15. Note that surprisingly WU2 PspA provided better protection against D39 that did R36a/Rxl/D39, and that also surprisingly PspA from Wu2 2 protected better against BG9739 than did PspA from BG9739. Combinations containing R36A, BG9739 and WU2 PspAs were most widely protective; and therefore, a '.preferred composition can contain any three PspA, preferably R36A, BG9739 and WU2. The data in this 3o Example shows that PspA from varying strains is protective, and that it is possible to formulate protective compositions using any PspA or any combination of the PspAs from the eight different PspAs employed in 'c the tests. Similarly, one can select PspAs on the basis 3- of the groupings in the previous Example.- Note additionally that each of PspA from R36A, BG9739, EF5668 and DBL5 are, from the data, good for use in compositions.

130 A note about use of medians rather than averaces. Applicants have chosen to express data as median (a non-parametric'parameter) rather than averages because the times to death do not follow a normal distribution. In fact there are generally two peaks.

One is around day 3 or 6 when most of the mice die and the other is at 21 for mice that live. Thus, it becomes nonsensical to average values like 21 or 22 with values like 3 or 6. One mouse that lives out of 5 has a o tremendous effect on such an average but very little effect on the median. Thus, the median becomes the most robust estimator of time to death of most of the mice.

i* 4* *ee TABLE Relative ability of different PspAs to Protect against each challenge strains of S. pneumonia, (Summary Of statistically signiffint protection) Vaccine PspA a i t o n e c a s P s A p sM g3 A R l J 9 0 W U12 9 7 3 9 O 8 E F 3 2 9 6 E F 5 6 6 8 L B 1 9 0 S J S 5 0 i .3 J S 3 0 2 0 A ll I m m u n e b e so 979DBLS OBIBA pmlecl Strain type tye family K a b E DD b if

D

M39 2 25 K EF1 0197 3 18 A T C c 3 0 3 -3 7 a. L BG9 26 b D+ 0 0 0 EF684 12 DD 0 0+ 4+ DOLSA 6A 19 D 6G9163 6B 21 C+4 BG7322 B 24 W ftty al one.at '"hat 4-2M a a o-lt at 1 a I bo ld jte t 2 S t f o t o d c 2 or a ne tab a t a c t a0wa w e re a a i a e m m o a0 ex ep t indiatet in the I+ I footnote beo". "hdt "8Mmunw* rihowe i~tioiant protectont saat death but Widirs!u date c a=.fid not, the rofttt for 'ad imonun.* Is pratied undr beat protaction On the aomptbn that atwo mim wave done in each daot on O more of ttem vt i have exhibted aignificant Protectionl aainst death atelitcirl significantt Protection eigoiri death, 2: 50% protection from death s tatistically eignificmnt Protacton against deat 50% protect from death q stahtitcaly *qgnircent delay indeath;, protco fomdat s tatiatically signil'icant delay in death; <20 protectiont horn death, for aqgniticarot protection againet death 1xut not a ignilican datay it death) 4 Not altatimttoally delay intime~ to deatht; but at 50% protection from death 04 Not statitically delay bn time to death; but 1.5 day extensboan in medion time to dath or X 20% protection Fhorn death. 0. NoappareNt etansion in time, to death or protection from death.

7 TABLE 31 Relative ability of different PspAs to Protect against each challenge strains of S. prleumoniae (Expressed as Median days Alive post challenge) Va1ccine PspA cianng Cps PpA P"p R36A Rxi. JD908IWiJ2 JSIO2O/BG EF3296 EF5668 Le 1905 JS5O1O.3 jS0 JImun Al DB39 97 BSA control Strain type tye family K a b E DOb t 392 25 K 4.5 >21 4 2 1> 1> 1> 2 1> 2>2> 2>22 6 6 7 3 3 2 6 b 3 2 1 62 2 1 32 523 13 2 2 2 N e: Bod 1eots97 titcat sgi 1catexesinatlieatP< .5.2m1 fn dno>2001 4< .0;lag>21 dnoe P<0.2 Me i n t me o d at n at d a 2 a e iu lonM h r t e m d u a>o2iti1 >21r o a u s. I h se c s s th u b r shw AaeThCC coet6oth3eda.0n3es assIh aue7iv retas lsettote2aclae mdan>2ato1lvlussuha nict ta7 temeia s afwybewentw um esb I ti3 cse3an sinia1 d nth riia 3dle(>21) om xprmet wee Eemiaedpio9o6 1daspo4 inetin2 Tee0slite eso3o sumho eer ha1esls oldhveben2gnfcaty fece by 5r> th eafl tenlal sinc ve2 fe mic inet2wt h tan sdi hs tdehv vrbe bevdt i e hn1 15daspatc2llne.FrDtaiia pross2l mc>aiv21 h edo1ex3imns>er2ssmd ohae en>o2lte1prtctd and 90 fo the bak of caclain al au n miewr sindv3et>1 0 0* *9

O

TABLE 32 Ability of different PspAs to Protect Against Each Challenge strain of S. pneumoniae (Expressed as increase in survival time in days) (A denotes 2t 50% Immune mice alive) Challeng Cap 24sp Csp 1A, AD U 7.520 -1F329 AF6 L815 AS0. JS02 Al

A

R36A WU2 BG9739 EF26 F66 810 01 DBLA Allun est Straio~ d n ttrla y pjifen type fam ilytP .5 K af bd E.00 DD0 b D eoe.P( 0.2 n a.e '..eeftitda.3 w ent rop hred ~aindosntFa i~oacn~~mo D39. 2te.c 2t5 Kgia.... l 2.t5heet Adne 2 3joa A..ec a kict ta h wd. efaybran w ebe.h i.te 3ed4 TABLE 33 Relative ability of different PspAs to Protect against each challenge strains of S. pneumonee (expressed alive at 21 days post challenge) Vaccine PspA D93 2 26 K 38 60 13250 25 03 1 BG19703 4 23 b 10 0 3 40 0 14 0 14 0 8905 64 24 b 10 0 25 4 0 80 25 80 25 Sold. denotee statietiah prn.~n otection eganat deathi at P 0.05. Bold amlI foot. indcatee eionfican protction aosinet death of 0.02 s 0.05. Sold Io* font. indicoteamiont.,.nt protcion 89"t doath at P 0.0)2.

0 0 0 TABLE 34 Relative ability of different PspAs to Protect against each challenge strains of S. pneumoniae from death at 21 days post challenge) PspA Challenge Caps PspA psjp4 R36A, JD908/WU2 JSl1020/ EF3296 EF5668 181905 JS5010.3 JS3020 All Best Rxl, D39 BG9739 ___DBL5 OBISA Immune result Strain type tye family K a b E DDb 11 D -I D39 2 25 K 36 59 28 36 59 WU2 3 1 a 100 100 100 100 -F 100 100 100 100 100 A66 3 13 a 71 100 79 74 100 58 16 75 100 EF10197 3 18 M 100 80 90 100o ATCC6303 3 7 a 100 _0 0 100 BG9739 4 26 b 11 60 13 25 025 0 0 12 6 EF3296 4 20 E 25 2 0 10 0 0 8 EF5668 4 12 DD 14 18 56 34 100 34 56 -10 3 0 L81905 4 23 b 10 0 31 40 0 0 14 0 14 D131- 5 33 If 10 14 0 0 29 0 4 29 EF6796 6A 1 C 100 __100 100 DBL-6A GA 19 D 66 22 30 -4 58 22 .4 79 33 79 BG9163 613 21 C 86 7583 8 BG7322 GB 24 C 100 57 22 0 88 22 79 22 52 100 Bold. denote. attatimetty Sgnfiant rotection wasinat death at P 0.05. Bold 11 font, indicatee ignidficant Protection againat death at 0.02 s P 0.05. BSl INC* font. idictes wignifican, protection against death at P 0.02.

Protected has been contcted forW ut n. rain the cntrol mice.

Pooted 100 K 1% alve in krnowne alla in alitlf10. v* in contol. Thus. R Oh.we r any mim aOwe in the contrf a-nia, the catnotatad proteted' i leas tOwn the oheuved 0% alv. Noted in the Prone.ue tabte. The 0ompi nst to this No It 100% of im nuntIzd me,. had. Negative toanthers mean that imnmunized moae lived than did controf mica. PIne.. note that none of thee negative tnumbersare aigrnifcant evan tt'eeQh oe we uakng a one taeed tee,.

TABLE Recommended Immunogens to Protecion against the indicated ch~allenge strains Of S. Pfetrmsae Based on Protectioni scoe Based on median days alive and perent protected (numbers refer 10 preference as a vaccine strain with respect to the indicated challenge strain. 1 -best) PspA Challenge Caps PapA psp4 R36A. RxI. JD908/Wii2 JS102O/BG EF3296 EF5668 L8t905 081.5 OBISA 0C- 39 9739 JSSOI 0.3 JS3020 Strain te tye faily K a b 2 DD b It D 0D39 2 25 K 2 1 3 WU2 3 1 a 1 A66 3 13 a 2 1 2 2 1 0 EF10197 3 18 NI 1 2 ATCC6303 3 7 a I 8(G9739 4 26 b 3 1 2 3 3 2 0 0 EF3296 4 20 E 1 1 2

F

L.81905 4 23 b 2 0 1 1 0 00 0815 5 33 Il 2 30 1 EF6796 6A 1 C 1 DBl-6A 6A 19 D 2 0 3 0 2 0 0 BG9163 68 21 C 1 1 6(G7322 68 1 24 C 1 2 3713 3 Number of #1 Is 7 E E bol, denote att..ticafy uignificant protection .0sinutdoo t OP 0.05. Whe. "ore*thtan on* PsoA wareewquey peotectitre, ftw e vafi 0, were gon to eacht. Rcocm.dt,. w c O ba seadon day, to death wth protection diidig the.

fapecifty amortg tho.t we greater then 50% of rrcc. Iicd Io 21 d eye. '0 idcat test wers condjucted but conptered to the other PapA* tivi. orn. Is not tee,e,dsd.

Cotantsione: Statisticsy significanct ptetion egsmet deetu Mtd,> 50% pratectcort 11114 of the streine 79% Statitically elgnificomt protection againut det, 13114 gwak. 93% Statistoal significan~t extension, of fle itt 14114 at 100% of tr.inst.

III

137 TABLE 36 Best Choice for Vaccine Components as Of 95/8/27 Vaccine Component (cumulative strains protected) maximally protected Criterion Critrion3 PspA for each challenge strain R36A (7) 50% WU2 (10) 71% 8G9739* (11) EF5668 (12) D8L5 (13) 93% 6 ,DB L6A (14) 100%/ 869/ i J I I 2t#2 PspA for each challenge strain R36A (12) BG9739 (12) 86% 100% I I I Max score type score R36A (9) 64% WU2 (11) BG9739 (13)

DBLS

(14) I9 I fY O Max Increase in R36A WU2 BG9739 0815 Days alive (11) (13) (14) 64% 79% 92% 100% protected R36A WU2 DBIL5 EF5668 DBISA EF3296 (10) (11) (12) (13) (14) 50% 64% 79%- 86% 92'% 100%/ Theoretical mixture R36A BG9739 DBL5 EF3296 based on a few (10) (12) (13) (14) testable 64% 86% 92% 100%/ assumptions (see below) *This is not a unique combination. See table below.

TABLE 37 Combinations where all Challenge Strains have a Vaccine strain 17 C with a score of a: #2 Number of Combination Number Total #is Total PspAs in of #Is and Combination #1 strains #2s 2 R36A +BG9739 8 10 3 R36A +BG9739 11 15 3 R36A +WU2 11 15 21 0815 3 R36A +WU2 11 15 23 EF5668 3 R36A +WU2 11 15 22 0815 TBE 38 ci

S

138 Pooted Date for ProtaciOn against 039 by various PapAs; alive for each mnouse Log CFU 039 lAce Days to Davrniae RxIIRl35A 1JDOO fWU21 143 14.6 4Jca"J Ij_ C8A./N cSA/N 2.3.33~4 t- I .1 SALB/c 3.3x 21 E143 -0 ca E1 40 BC1 00

CSA/N

4.4.5.7,15 1 4~ 1 1 EF5868 AN Inrvnuns conitrol 11.2.2.2 1.1.2.2.3 1.1.2.3,4 2.2.2.4 2.8.3x>10 3.3.3.5,5 2.2.2 1.2.2,2.3 401. 6x2,3,3.4 1.1.2.2.2 4 x 11 x 2,3.3,4 1.1.2.2.2.2.6, 1.*1 .9x2,5x3.3 Sxl .1 6x2.6x3,4, 3x> 21 x4.6.6,6.7,1 5. 4.5.5.5. 21 15x> 21 CBA/N 2.4x 21 2 5 3x>2 E164 CWAN 2.2.3.2x >2 I. J. All All 2.3.3.3,4,4, 4.5,7.15 4x2. 15X3.

Ux4.5.7.1 S.

9x>21 2.5.3x 21 C. *e 0* C C I C

C.

TABLE 39 Pooled DOto for Protection egeinat D39 by vout PapAs Median Day* Alive alive:dad with corresponding Pivalues Exp. Log PMce Rx1IR30A 039 JD908 IWU2I EF5868 All Imfmne Control

CFU

D39 med ad ned a:d mod ad fred a:d med ad 143 4.5 CSAIN 2 0:5 2 nis E145 4.0 CBAfN 3 0:5 2 0:6 I nlie E028 BCG 5.93 BALB/c >21 3:1 2 04 029 nes T E143 3.0 CBAtN >21 3:2 3 nis nes E140O8C100 2.81 MBAIN 5 0:5 2 0:3 E169 2.7 M8AIN >21 4:1 >21 3:2 2 .018 .024 .018 n.s.

E154 2.a C8A/N 3 2:32 4 0:10 2 0:5 2 0:13 0008 ne.

ANl 4.5 9:15 >21 3:2 412.01 3:7 5 15:24 2 1:32 .0057 .001 .0008 .0045 n. v .034 0001 .0002 alive 38 60 30 38 E3 r59 28 38 1x11039 WU2 EF 8All irine controls -*0 S. *O 5 0 50. 0. TABLE

CFU

I

W1J2 IMICE FL-R36A I x BCIO0 Pooled Data for Protection against W1.12. by various PspAs Days to Death/immunogen JD108 JS1020 B(39739 EF5668 L81905 DSLS JS3020 control _(WU2) 1(8(39739) WcOO WcOO WcOO (DBL6A) 4- Dr. Ed, expt.

lots of prior expts.

I 4 E012 '3.0 JCB3A/N E028 6.01 BALB/c 1 5x>21 4- 4x> 21 0.05/n.s.

E084 1 3.1 1 CBA/N 5x 2 E125 3.57 CBA/N4x 2 4>2 Eb 13 3.4BA Nx 21__ x 1 E1229 3.98 CBA/N E172 3.98 CBA/N Ali 19x >21 4x>21 5x>21 8x>21 4x>21 8x>21 4x>21 All Immune 61x>21 4x>21 Sx> 21 4x>21 5x>21 4,6,6, 21 2,2,3,3,3, 21 1,5X2,3,4 1,1,2,2.2,3 5x2 5x3 6 x1 3 3 x2,20 x3, 4,4,4,6,6,>2 1 1 0 0* 6 TABLE 41 Data for Protection againt WU2 by various PspAs Exp. CFU MfCE Median days Alive Alive:Dead P value based on Alive:Dead FWU2 P value calculated compared to pooled controls (in this case 65 control mice) FL-R36A Rx1 JD1O8 JS1020 BG9739 EF5668 L81905 DBL5 JS3020 control B 8100 (WU2) (BG9739) bclOO cOO bc1OO (DBL6A) Dr. Ed, expt.

lots of prior expts. E012 J 3.0 CBA/N I15x>21 1x2,7x3,4 E028 I6.01 IBALB/c 4x>21 21 j J 0.05/n.s. E084 3.75' CBA/N 3x>~ 1r, E125 bc1OO

CBA/N

4x>21 4x>21 I4x>21 I I t I 1 4 4 1 1 1 El129

CBA/N

E140 3.43 CBA/N 4x>21 E143 3.0 CBA/N E144 3.9 CBA/N E172 :i3.98 CBA/N 1.2'2,2,3,3, 2,2,3.3,3,>21 2,2.2,2,3 1,5x2,3,4 1.1,2,2,2,3 5X2 5x3 >21 4:0 <.0001 >21 19:0 <.0001 >21 4:0 <.0001 21 5:0 <.0001 >21 8:0 <.0001 >21 4:0 <.0001 :>21 <.0001 >21 4:0 <.0001 5x>21 >21 .0001 100 4.4-4.

I

4

J

alive Iioo I AP 2 *@0 0 0 0000 00- WU2 Challenge IL days of death 9. median days of death alive: dead P value based on days to P value based on alive:dead Score alive j 1l 4. '4 1 I All immune 61x>21 >21 AIfl protected 100 2 <Anni Iv~ ji I- I 0o1 1 All controls 6x1 ,33x2,20 x3,4,4,4,6,6, >21 100 2 1:64 I "Oe 0 0 00 *0 00 see TABLE 42 Data for Protection against AMS. by various PspAs CFU Days to Death/imniunogen E xp. ASS MICE FL-R36A/ Rx1 JD908 JS1O2O BG9739 EF5668 181905 181905 JS501O.3 DBL5 JS3020 control D39 BC100 (WI1J2) (BG9739) bclOO FL WcOO FL bclOO (DBL6A) E169 2.60 CBA/N 5x>21 Sx>21 1,1.2,2, E152 2.78 CBA/N 4x>21 4x>21 4x>21 3x2,3,6, 11 1 i i1 1 1-6, >21 Eli 4 CBA/N 2,8,3x>22 3,4,4,2x 2#,44,5, 2, E143 3.0 CBA/ 1 4,4>1O 22 E143 3.0 BA/N4,4x 102, E140 3.43 CBA/N ___4x>21If_

I

E172 3.94 C8A/N 5x>21 2,2.2.

3_ 2,3,3 ,1,1 E145 CBA/N 13,4x 21 I I I T F f I I I I E121 4.16, CBA/N 3x3,2x4,5 x>21 t 1 t t t I I I IL I 3x3,2x4,1 3,14x>21 4x> 21 5x>21 2,8,3x 21 4x>21 4,4x >21 5x>21 4x >21 3,4,4,2x >21 4x 1- f. J f 4 1 1 1 median;

A:D

4 >21 21 :0 2,4,4.5, >21 1,22, 4 1 ,8x2,> 21 7x1 ,22x 2,3x3,4, 3x6,2x >21 2- I I I j J 1 1 *0 9 0S 0 0 9 *0 09

CFU

Exp. A66 MICE FR36A/ Rx1

JD

D9jBC100

(W

P values <0.0001 0.0002 <0.4 Mini Pools R36A/Rxl/WG44.1

JD~

>21> 5: P values <0C rank/a:d <0.0001 <01) Pooled Data for Protection against ASS. by various PspAs Days to Death/immunogen 908 JS1020 BG9739 EF5668 L81905 181905 JS5010.3 0815 JS3020 control U2) (BG9739) WcOO FL WcOO FIL bclOO (DBISA) 0001 0.004 0.0002 0.0006 <0.0001 0.0002 0.0025 0.0002 0.015 0001 0.0075 <0.0001 0.006 <0.0001 0.0001 n.s. 0.0001 n.s 908 BG9739 EF5668 L81905 DBL5 DBL6A Control 3,4,4,4,6x>2i 21>21 21 >21 >21 4 2 0 8:2 4:1 9:0 6:4 1:4k 2:36 001 0.0006 0.015 001 <0.0001 0.006 <:0.0001 0.0004 n s.

0BO 75 100 60 20 79 74 100 58 16 0 08 BG9739 EF5668 151905 DDL

DSA

Score alive 72 10 71 10 A66 challenge A66chalene JI R6A/, c/VG44.1 fJD9

S.

a. a.

TABLE 43 Data for Protection against EF1 0197. by various PspAs Exp. CFU Mice Days to Death/immunogen

EF

10197 RXI JS1020 L81905 JS3020 EFS668 JS5010.3 control B C QO (BG9739) (DBL6A)

FL

(DBLS)o E140 3.00 CBA/N 5x>21 MI BCG 2.70 CBA/N El 29 3.34 CBA/N 8,4x>23 This was a passive protection study. Its controls have been included to increase the numbers of control mice.

H-

U,

e.g C C 0S eg 0 go 0 g *g *e eg gg* Os TABLE 44 Pool of Pools for protection against EF10197 Group line Descriptionj la jRxl (E140)I Delay in death and/or survival days to death (median) 5x 211 P values etc.

0.017 vs lb t 1 .1 0.0013 vs 4b JS1020 (EI291 8 4v 0 all immune 0.0007 vs 3b 8. 9x 21 I l*

I

Survival alive:dead P values etc.

5:0 0.018 vs lb 0.0008 vs 4b 4:1 0.024 vs 3b 9:1 0.0002 vs 4b 0:3 0:8 0.0001 vs 4b AxI controls (El 401 2b -Ml BCG 2,2,2,2,2t_ 3b JJS1020 cont. (E129) all control (without Ml

BCG)

2,2.2,2,9 2,2,2,2,2,2,2,9

I

TABLE Summary of protection against EF10197 Immunogen alive:dead alive protected median DOD P time alive P alive:Dead Soore* RXl 5:0 100 100 21 0.017 0.018 JS1020 4:1 80 80 21 0.0007 0.024 all immune 9:1 90 90 >21 <0.0001 0.0002 all controls 0:8 0 0 2E q a a a a..

a.

eB 0* 0 *0 0 0.

00- SO 0 TABLE 46 TABLE 47 TABLE 48 statistically significant protection against death with >50% protected.

.0 eg 0 0 0 00 00 *o@o.

0 0 TABLE 49

CFU

Exp. BG9739 E140 2.76 E 104 2.89 E125 3.56 EM7 3.71 E124 3.76 E084 4.05 E144 4.09 All median a:d P rank P a:d Mice Pooled Data for Protection against B09739, by FIL PspAs Days to Deathifimmunogen R36A BC100 JD908 JS1020 bclOO0 EF329 EFS66 bcloo JsS0lO WcOO JS3020 control FIL (RxI) (WU2) (BG9739 (BG9739) 6 FIL 8 FL (181905) .3 FIL (DBL5) (DBLSA) 3,3,10,112,3 6,6,7,8,82,, 31 7 2222 22355 .4 3

ICBA/N

Xid iCBA/N BALB/c Xid 2,3,6.

2,11 2,3,3.6 2,1 5,5,5,7 4,5,13, 21 2,2,2,4 6,7,3 x >21 3,3, 10, 11 3,10 0:4 673x >21 >21 3:2 4x2,2x> 1 4 4X2,6,6,7 18.8,2X> 21 6 2:9 5.5,5,7 0:4 2,3,3,7, 2,3,3,7, >21 3 1:4 ,4 2,3,3,3 ,4 3 5,13 0:5 1:3 2,22 2.2,2,2, 9 7X2,3, 3.4 2 2 0:10 0:4 SX2,3,9 2 0:10 3,3,4,4,5 3,4,6,6,7 2,2.2,2,2 2,2,2,3,3 2lx2,7x2 ,3x4,3x5, 3x6,7' 2 0:38 TABLE Pooled Data for Protection against 8G9739, by WcONs and FL PspAs CFU Mice ____Days toDeath/immunogen____ Exp. BG9739 R36A BCIOO JD908 JS1020 bc1OO EF3296 EF5668 bcI 00 JS501 bcIOO JS3020 control FL (Rxl) (WU2) (8G9739 (BG9739) FL FL (1.81905) 0.3 FL (DBL-5) (DBL6A) (0515) E140 2.76 CBA/N 3,3,10.1 2,2,3 1 E104 2.89 Xid 6.6,7,8,8 2,2,2,3, 2,2,2,2. 2.2,3,5,5 4 3 E125 3.56 CBA/N 5.5,5,7 4,5,13, 2,2,2,4 3,3,4,4,5 6 E172 3.71 CBA/N 6,7,3 x 3,4,6,6,7 2-1 E124 3.76 Xid 2,22 2,2,2,2, 2,2,2,2,2 3 9 E084 4.05 BALB/c 4x2,2x IW E144 4.09 Xid 2.;3,3,7 2,3,3,3,22,33 10 4 Pole Dat for Prtcinaans-a79 yWosadF Fl. ooe Data) for2 Protectio agintG9739 byL FcLO and -195 FL3 FLPspB (Bs FLxp. O 8G97 979 R36A/xD3 WC10 JD0 S 0B739 EF3296 EF5668 bL8190 13501 bclOO JS 0 Contro All 2,4x3,6,1 0,11, >21 6,7,3x> 4x2,3x5,2x6,2x7,2x8,2 2,3,3,7, 2,3x3,4 4,5,13, l0x2,3,3,4,4 8x2,3,9 21x2,7x3 median days alive 3 >21 63351 alive:dead 1:8 3:2 2:13 1:4 0:5 1:3 0:140:0:3 P days alive 0.0096 0.000 0.0013 n.s. n.s. 0.0022 n.s. ns P alive:dead fl.S. 0.0008 n.s n.s. n.s. n.s. n.s. n.s. L Score 0+ 0 0 0 0 alive 11 60 13 25 0 0 0 0 0 protected 16013 25 0 0 0 0 0 BG9739 challenge R36A/Rxl/D39 BU G9739 EF3296 EF5668 DBL5 DBIS D3LA dCnt.

'0 0- 00* o 000 TABLE 51

I--

Pooled Iole Exp.

CFU

EF3296 Mice E84 1 j 3.99 f BALB/c El 40 El 04 El124 E172 I 2.92

CBA/N

3.11 CBA/N 3.94 fCBA/N 4.06

CBA/N

All Data for Protection against EF3296. by various PspAs Days to Death/immunogen______ Rxl BC100 JD908 WU2 JS 1020 JS5010.3FL JS3020 control (BG9739) (DBL5) (DBL6A) 4x2, 1492 3,4,6, 21 3,3,3 4,5,5,5,6 2,2,2,3,3 2,2,3,4,5 j2,2,2,3,4 1,1,2,2,2 1,1,2,2,2 1,1,2,2,2 3,4x6 6,>21 4,5 5x3,4,4x6 5- 5 4.5 2 2 2 1:3 1:4 1:9 0:9 0:10 0:27 0.0077 0.0094 n.s. n.s. n.s.

n.s. n.s. n.s. n.s. n.s. n.s.

0+ 0 0 25 20 10 0 0 0 median days to death alive:dead P-days to death P-alive:dead Score alive prot.

Best E39 chllng Rx D0.S12 S00.F S0.2 oto BCO WL1 .B979 (DL.

-DL

a 5* 0 0 00 a a a a S 000@ "ae a. a 0*s TABLE 52 Pooled Date For Protection against EFS6OS. by varpou wnd bcl OOs

CFU

EF5668 fAcs I Dayt to Daatirirrmnogen A38A RxI Scico JD908 JS1 02OfBG9 (WU21 1 7391 EF3290 EFS868 LS81905 JS5O1 0.3 aI

DIL-S

JS3020 D818A t f I f I I I CBAtN 5x: >10 I I I f I 1 1

CBA/N

4.8, 12, I 21 III___ 3.89 CBA/N 2,2.2.3 3.3.4, 2x>21 I t I 4 I I El 24

CSAIN

I I. i E145 CEBA/N 13.4.41 19 2,10 3x >19 Poo 3,3x4.0.12,1 8 2.2.2 2.10.

J2x >21 3. <21 3x>21 metndays live 2 <21 alive:dead 2:7 1:4 3:2 24 1'3.

xi>19 1 3, 2x> 21 13 2:3 n..

n.a.

0+ 40 34 3,3,3x >21 3.3.

3x 5.6 contro? 1.1.2,2.

2,4,6 1.3.6.7 3.3.34.9 2.3.34.

>21 Wx,304 3xB,7,9.

2x>21 3 2:21 9- 5x <21 3.3.4 2x>21 days alive fl f$l Ci 00101 P ative:dead jn. s.

Score 4 1 n.s.

0.027 001 5:.0 2:3 0.001n.s.

0.0002 n.m1.

0+ 100 40 100 34 <21 5 3:2 0:5 n.s.

0.027 n.s.

0+ 60 0 56 -10 %alive 22 25 60 %prot. 221 18 so 0 4 0 0 0 S 0 Oh 0 S 0S 00 Sos 00 0 0 EF58aB I RAftlD39 WU BG9739 EF3298 EF8 I oz IDeA Jcntl of protection against EF6796 Immunogen alive: dead alive protected median DOD P time alive P alive vs. dead RXl 4:0 100 100 21 0.029 0.029 E controls 0:3 0 j01 *S4 C S S S SS S S C S *S 55 55555 5 5 TABLE 53 p p. p 0. pa P p p p ppp p p p. pp ppppp p 4 a. a 0 a a.

*0 a a a S6 aC ee a C TAflLE 54 Exp. BG963_ Mice RxI El169 2.67 I CBA/N E14 3.14 fCBA/N Ell 29 E028 4.0

CBA/N

I

6.217 kIrnnxlogone

I

CBA/N

Pooled Date Median days mlive alivs:dead days alive P alive:desd 5x>24 n.

Rxl1/R36AID39 8,8x >21 21 8:1 0.0086 0.0045 89 86 RX1/R36AtD39 Data for Protection against BG9163 by various PspAs Days to Death/immunogen xI.BCG JS 1020 all immune control (BG9739 4,5,8,8, >24 12,4x >23 7,9,9,13, >23 3x>2 BG9739 all irrmun. control 1 2.4x 21 6. 1 2.1 2x> 21 4.5,7.8,8,9.9,12, 2x> 21 21 >21 4:1 12:2 2:8 0.0097 0.0027 0,047 0.0022 80 8620 75 83 0 BG9739 all irmne control alive Pont.

sctre BG9163 Challenge a -o *s 9.0 105 41 00 00 0 00* 0 v 0 0000 00 00 00. 00 a a TABLE Exp.

El 72 E140 E084 El104 E 124 El125 E144 All median alive: P tank P ad scor$

CFU

L81 905 2.45 3.11 3.88 3.5 3.5 3.8 _4.11

FACO

CBAIN

BALS

CBA/N

CEIA/N

R38A 3.3.5.6, 3.3.5.0, >21 1:4 Pooled Data for Protetion against L81905. by various EL.PspA* BCIOO JD908 (Axil CWUJ21 3,4,5,8.

8 0:5 0:5 .151020 97391 2.2.5x> 14 3.,8,8, 11 2.2.3,7.

8x>1, x>21 5:7 Days to Deathlirrsmunogori bclOO EF3298 EF5668 bclOO AS bcI 00 JS3020 control L81905 5010.3 IDSL51 106LBA1 3.3.4.4,4 2.2.2,3,3 1,802 3,3,3.2x 3.4.5.5, 2.4.4,4,5 a 5.8,8,8 6,0.0.2x 2.2,3,3. 3,4.8,8 4x2.4x3, 4.5,5.5 Ux2.3.3. 1.120x2.8x3.dx4.4x S S a a *0 S e 0 9 TABLE 56

R

TABLE 57 Pooled Dots for Protection agamnst DOLS by vious FI-PxpAv bclOOs Exp. CFU I 0 ays to Oehthkirvyinoge, R3A S~O S12 bIO EF560B WcOO JSS01O.3 b,:100 JS3020 control 899739 JS1 020 1819705 0815 DOLS DSLOA EW4 3.90 SALS/c x 8E1 4 0 3 .2 7 C S 4 5 5 ,2 1 2Exp2 E104 3.39 Xid 3.3.0. 7,7,15, 2.2.4,5, 2,4x3 >22 22, >22 El124 3.78 Xid 2.2,2.5. 5x2 1,1.2.2.2 E125 3.81 MBAIN 3,3.4.5 3,3,4,4 2.2,2. 5x2.5 4.3 XI 3,21 E144 41 XO 2.2.3.4.52 >10 4 5x2 total t Nam* of Pool R35AM/0I39 BG9739 EFS688 181905 0815 OUtA controls Pooled Data 4 x 3 .2x 4 21 8x2.4x3,4,5. 22.3.4. 3.3,4,4 6x2 5,7715,4x>21 7x2.4.6. 1.1.2110.403,5 21 4 median dys @live 4 3 3 3.5 6 2 2 sliwaed 1:9 2:12 0:4 0:4 4:10 0:10 0:32 P days alive <0.0001 0.0083 .041 0.001 0.0025 nas P elivo:dead n.s. fls n ft n a0.05 $core %eia1 40 0 29 0 0 protected 10 14 0 0 2 DBScaleg R38AJRxiIO39 BG9739 EFS688 181905 OI DB LSA controls This immunization was with cell eluted PspA. Note BALB/cJ mice were used. Also note 104 Challenge CFU.

0.00 D1315 days of median days alive:dead P value P value Score alive protected challenge death of death based on based on days to alive:dead death All immune 3.5 7:49 0.0001 _j 0.034 3. 63.

All 4 %ovil IV z n:-q-,a flR'~

II

I- I TABLE 58 TABLE 59 Pool of Pools for protection against EF6796 Desc ripo Delay in time to death and/or survival Protection against death days to death (median DOD) P values etc. alive:dead P values etc.

la Rx1 4x 21 0.029 4:0 0.029 l b RxI controls .i,i,i 0:3 o C ee CCC 0. e-- TABLE Exp.

El 71 Pooled Onto for Protection egmini 807322. by verious FL.P&PAt bclOO, BF fc Days to Deethrrnrnunogsn 732 D9/ xI JD98 oci00 EF298 EF658 WO Jbcl. OO JS3020 control 1130A BC100 (WU2) OG9739 L 81905 3 081.5 DBLOA I

DBLS

2.78 CBAN 10.15,3J CI 43 E152 3.14 3.11

CMAIN

CBA/N

MAIN

7.ex).1 4x>21 12. 13,1 6. 21 10. 12. 1 3,>21 E148 3.57 C8A/N 18,20,3 x>21 E169 3.94 CBAJN 5X>21 Immurigen, 38A/fRxl/039 Pools 18.20,12x>21 median days &live >21 slive:dead 9:0 P- days alive <0.0001 Palive:dead <0.0001 Avi.e 100 %protected 1o0 Score JD908 10.15.3 x>21 21 3.2 0.0007 12,13,1 6, 21 14.5 1:3 0.001 5.3xxO.

10 EF3296 5.3,xO.

10 0:5 as.

0 0 >21, >21.

>21.

>21 6.10.11.

11.19 01 21. >21 EF5688 7.8x >2 L81905 10,12, 13,> 21 L8.>21

DBLUA

4.81 2.2,3.4,7.7 3,8.6, 21 8.7.7.8.8,9 ,14 4,5.5,e, 2 2.5,5.6.7 Cat.

1,3X2.3.3.

4.45x5.7..

8.6x7,4x.

9,14,2x 2 2:3 0.004 n.s 80 25 57 22 >21 8:1 <0.000 <0.000 89 88 12.5 1:3 0.0013 n.s.

22 >21 4 :5 0.0002 0.0076 79 0.029 n-s.

22 6 +:t 11i 0* 0O a. 0*O 6e0O ho I I LL_ IDG7322 Ma"WV9 IL MAftl=B I3 J0908 IBG9739 E73296 1F Men L81905 DOOAC"t S OS S Se S 5 5 5 *S 55 55 5* .55 5. 5 5--

I

167 EXAMPLE 8 ABILITY OF PspA IMMUNOGENS TO PROTECT AGAINST INDIVIDU CHALLENGE

STRAINS

In Example 7 some of the capsular type 2, 4, and 5 strains were not completely protected from death by immunization. In these studies the BALB/cByJ mouse was used instead of the hypersusceptible, immunodeficient CBA/N mouse used for the Example 7 studies. With the BALB/cJ mouse it was observed that immunization with PspA was in fact able to protect against death with capsular o type 2, 4, and 5 pneumococci. This result is shown in the table below.

The data from Table 60A also demonstrates that a mixture of 4 5 full length PspAs was as effective, or more effective than immunization with a single PspA.

Table GOA. Days of death of BALB/cByj imice after immunization with monovalent and polyvalent vaccine.

Challenge Strains Immunogen strai C" PppA pr t e gosi D-th name type B gMn~of Chdkfw I mt R36A 4 S vl Mit IY2141

N

cld CFA I10.5pg ecb). CFA

CFA

D39 2 25 2 4.76 3,4x >21 3.4x >21 3 4 5 11 1 >21 4 WU2 3 I 2 4.8 4x 21 4x >21 6, 3x >21 3, 4, 2x >21 A66 3 13 4.7 3,3, >21, 2, 3x >21 2,2, 3, 4 2, 3,4,4 21 BG9739 4 26 1 4.07 4.4 7, 8x >21 3,8>2 1,5,6,6,9, 3,3,3,4,6, 3 8 x 14x 2 1 7, 7 2x 2 1 L.81905 4 23 1 6.90- 2,2,2,2,5, 2,6,8,9, to1,,t1,,2, lp 4 x2,3x3, 6.96 5, 6x 21 3, 4, 5, 2x 4, 21 4x 21 21 EF5668 4 12 4 6.10- 3 3 4 ,7x 3 x 3, 6x >21 4x 3, 4,4, 6, 3, 5X4, 6, 6.93 21 6, 21 >21 DBLS 4 33 2 3.30 7 14,Rx> 21 3 5 2 x 2, 2,2, 4,6 4, 5,5, 6,9 21 DBL6A 6A 19 1 4.34 6,9, 10, 11, 10,11,12,13, 3, 11, 11, 13, 8,9, 11, 21, 21 >21 16 >21 BG7322 6B 21 3.9 8, 8, 3x 21 5x 21 6, 6, 7, 8, 10 2, 5, 6, 8, 8 Note, amzbers scb its ft 21 indicate dut 8 raice lived greater d6n 21 days.

Noe,JY2 1 4 1 is a preparation from a strain that lacks PspA. None no immunization.

Note, mice were given two immunizations with PspA two weeks apart and challenged intravenously 2 weeks after the last immunization. The first immunization was given with complete Freund's adjuvant (CFA) subcutaneously, the seconid immunization was given intraperitoneally in saline.

14 valent vaccine mixture R36A, BG9739, EF5668, and DBL5 all E180 24 valent vaccine mixture R36A, BG9739, DBL5, EF3296, D39 and DBL6A valent vaccine mixture R36A, BG9739, DBL5, EF3296, EF5668 170 EXAMPLE 9 CHARACTERIZATION OF PspA EPITOPES

WITHIN

PNEUMOCOCCAL STRAINS MC25-28 The strains examined came from a group of 13 capsular serotype 6B strains which have been identified that are members of a multiresistant clone, having resistance to penicillin, chloramphenicol, tetracycline, and some have acquired resistance to erythromycin. The pneumococcal isolates described in the following studies (MC25-28) are members of this 6B clone. Although ro previously thought to be geographically restricted to Spain (unlike the widespread multiresistant Spanish serotype 23F clone), members of this clone have been shown to be responsible for an increase in resistance to penicillin in Iceland (Soares, et al., J. Infect.

id Dis. 1993; 168: 158-163).

The following techniques were used to characterize the location of difference PspA epitopes: Bacterial cell culture Bacteria were grown in Todd-Hewitt broth with .o 0.5% yeast extract or on blood agar plates overnight at 37 0 C in a candle jar. Capsular serotype was confirmed by cell agglutination using Danish antisera (Statens Seruminstitut, Copenhagen, Denmark). The isolates were subtyped as 6B by Quellung reaction, utilizing rabbit S antisera against 6A or 6B capsule antigen.

Bacterial lysates Cell lysates were prepared by incubating the S. bacterial cell pellet with 0.1% sodium deoxycholate, 0.01% sodium dodecylsufate (SDS), and 0.15 M sodium S citrate, and then diluting the lysate in 0.5M Tris hydrochloride (pH Total pneumococcal protein in the lysates was quantitated by the bicinchoninic acid S method (BCA Protein Assay Reagent; Pierce Chemical Company, Rockford,

IL).

53 PsnA serotVping Pneumococcal cell lysates were subjected to S SDS-PAGE, transferred to nitrocellulose membranes, and developed as Western blots using a panel of seven MAbs to 171 PspA. PspA serotypes were assigned based on the particular combination of MAbs with which each PspA was reactive.

Colony immunoblotting A ten mL tube of Todd-Hewitt broth with yeast extract was inoculated with overnight growth of from a blood agar plate. The isolate was allowed to grow to a concentration of 107 cells/mL as determined by an O.D. of 0.07 at 590nm. MC25 was serially diluted and o spread-plated on blood agar plates to give approximately 100 cells per plate. The plates were allowed to grow overnight in a candle jar, and a single blood agar plate with well-defined colonies was selected. Four nitrocellulose membranes were consecutively placed on the 16 plate. Each membrane was lightly weighted and left in place for 5 min. In order to investigate the possibility of phase-variation between the two proteins detected on Western blots a single colony was picked from the plate, resuspended in ringer's solution, and spreadplated onto a go blood agar plate. The membranes were developed as Western blots according to PspA serotyping methods.

When the strains MC25-28 were examined with the panel of seven MAbs specific for different PspA epitopes, all four demonstrated the same patterns of reactivity <C (Fig. 14). The MAbs XiR278 and 2A4 detected a PspA molecule with an apparent molecular weight of 190 kDa in each isolate. In accordance with the PspA serotyping system, the 190 kDa molecule was designated as PspA type 6 because of its reactivity with XiR278 and 2A4, but none '3 So of the five other MAbs in the typing system. Each isolate also produced a second PspA molecule with an apparent molecular weight of 82 kDa. The 82 kDa PspA of -each isolate was detected only with the MAb 7D2 and was designated as type 34. No reactivity was detected with 5 MAbs Xi126, Xi64, 1A4, or SR4Wr. Results from the colony immunoblotting showed that both PspAs were present simultaneously in these isolates under in vitro growth 172 conditions. All colonies on the plate, as well as all of the progeny form a single colony, reacted with MAbs XiR278, 2A4, and 7D2.

EXAMPLE 10 SOUTHERN BLOT ANALYSIS OF CHROMOSOMAL

DNA

SISOLATED FROM PNEUMOCOCCAL STRAINS MC25-28 Pneumococcal chromosomal DNA was prepared by the Youderian method (Sheffield, et al., Biotechniques, 1992; 12: 836-839). Briefly, for a 500 ml culture in THY or THY with 1% choline, cells were ie centrifuged at 8000 rpm in GSA rotor for 30 minutes at 4 C. The supernatant was decanted, and the cells were washed with 1 to 2 volumes of sterile water to remove choline, if used. This step was only necessary when sodium deoxycholate was used. The wasted cells were i' centrifued twice a 8000 rpm in GSA rotor for 10 minutes.

Cells were resuspended in 3.5 ml TE buffer, containing 1% SDS or 1% sodium deoxycholate, and incubated at 37 0 C for minutes if sodium deoxycholate was used. If SDS was used, incubation at 37 0 C was not necessary. The cells 2o were incubated at 65 0 C for 15 minutes, and 1/5 volume of M potassium acetate was added, and the cell suspension was incubated for 30 minutes at 65 0

C.

The cells were placed on ice for 60 minutes, and centrifuged at 12,000 rpm in an SS-34 rotor for i2P minutes. The supernatant was transferred to a clean centrifuge tube, and 2 volumes of cold 95% ethanol was added. After mixing, DNA was spooled on to a glass pasteur pipet, and air dried. The DNA was resuspended in 4 ml TE, and 4.0 g cesium chloride was added. The solution was split into two aliquots in ultracentrifuge tubes, and the tubes were filled to their maximum capacity using 1.0 g/ml cesium chloride in TE. Before closing the tubes, 300 ml of 10 ug/ml ethidium bromide was added.

se The solution was centrifuged at-45,000 rpm overnight, or for 6 hours at 55,000 rpm. The chromosomal band was extracted using a gradient, at least 6 times with 1 volume'each salt-saturated isopropanol. The 173 aqueous phase was extracted by adding 2 volumes ethanol. The DNA came out of solution immediately, and it was spooled on to a pasteur pipet. The DNA pellet was washed by dipping the spooled DNA in 5 ml 70% ethanol.

The DNA was air dried, and resuspended in the desired volume of TE, 500 ul.

The cells were harvested, washed, lysed, and digested with 0.5% (st/vol) SDS and 100g/mL proteinase

K

at 37 0 C for 1 h. The cell wall debris, proteins, and polysaccharides were complexed with 1% hexadecyl trimethyl ammonium bromide (CTAB) and 0.7M sodium chloride at 650c for 20 min., and then extracted with chloroform/isoamyl alcohol. DNA was precipitated with 0.6 volumes isopropanol, washed, and resuspended in 'd Tris-HC1, imM EDTA, pH 8.0. DNA concentration was determined by spectrophotometric analysis at 260 nm (Meade, H.M. et al., J. Bacteriol 1982; 149: 114-122; Silhavy, T.J. et al., Experiments with Gene Fusion, Cold o Spring Harbor: Cold Spring Harbor Laboratory, 1984; and Murray, et al., Nucleic Acids Res. 1980; 8 4321- 4325).

Probe preparation and 3' oligonucleotide primers homologous with nucleotides to 26 and 1967 to 1990 of Rxl pspA (LSM13 and LSM2, respectively) were used to amplify the full length pspA and construct probe LSMpspAl3/2 from Rxl genomic DNA. 5' and 3' oligonucleotide primers -homologous to nucleotides 161 to 187 and nucleotides 1093 to 1117 (LSM12 and LSM6, respectively) were used to o amplify the variable a-helical region to construct probe LSMpspA12/6. PCR generated DNA was purified by Gene Clean (Bioll01 Inc., Vista, CA) and random prime-labeled with digoxigenin-l1-dUTP using the Genius 1 Nonradioactive DNA Labeling and Detection Kit as described by the manufacturer (Boehringer Mannheim, Indianapolis,

IN).

I I I 174 DNA electrophoresis For Southern blot analysis, approximately of chromosomal DNA was digested to completion with a single restriction endonuclease (Hind III, Kpn i, EcoRI, Dra I, or Pst then electrophoresed on a 0.7% agarose gel for 16-48 h at 35 volts. For PCR analysis, 5gL of product were incubated with a single restriction endonuclease (Bcl 1, BamH I, Bst I, Pst I, Sac I, EcoR I, Sma I, and Kpn then electrophoresed on a 1.3% agarose gel for 2-3 h at 90 volts. In both cases, 1 kb DNA ladder was used for molecular weight markers

(BRL,

Gaithersburg, MD), and gels were stained with ethidium bromide for 10 min and photographed with a ruler.

Southern blot hybridization The DNA in the gel was depurinated in 0.25N HC1 for 10 min, denatured in 0.5M NaOH and 1.5M NaCl for min, and neutralized in 0.5M Tris-HCl (pH 1.5M NaCl and ImM disodium EDTA for 30 min. DNA was transferred to S a nylon membrane (Micron Separations INC, MA) using a POSIBLOT pressure blotter (Stratagene, LaJolla, CA) for min and fixed by UV irradiation. The membranes were prehybridized for 3 h at 42 0 C in 50% formamide, 5X SSC, Denhardt solution, 25mM sodium phosphate (pH SDS, 3% (wt/vol) dextran sulfate and 500g/mL of denatured salmon sperm DNA. The membranes were then hybridized at 42 0 c for 18 h in a solution containing :i formamide, 5X SSC, 1X Denhardt solution, 20mM sodium phosphate (pH 0.5% SDS, 3% dextran sulfate, 2 50gg/mL denatured sheared salmon sperm DNA, and about O: 20ng of heat-denatured digoxigenin-labeled probe

DNA.

After hybridization, the membranes were washed twice in 0.1% SDS and 2X SSC for 3 min at room temperature. The membranes were washed twice to a final stringency of 0.1% SDS in 0.3X SSC at 65 0 C for 15 min. This procedure 3' yielded a stringency greater than 95 percent. The membranes were developed using the Genius 1 Nonradioactive DNA Labeling and Detection Kit as 175 described by the manufacturer (Boehringer Mannheim, Indianapolis, IN). To perform additional hybridization with other probes, the membranes were stripped in 0.2N NaOH/0.1% SDS at 40 0 C for 30 min and then washed twice in 1 2X SSC.

PCR

and 3' primers homologous with the DNA encoding the N- and C-terminal ends of PspA (LSM13 and LSM2, respectively) were used. Reactions were conducted S in 50gL volumes containing 0.2mM of each dNTP, and 1/L of each primer at a working concentration of 50mM. MgCl 2 was used at an optimal concentration of 1.75mM with 0.25 units of Taq DNA polymerase. Ten to thirty ng of genomic DNA was added to each reaction tube. The amplification S reactions were performed in a thermal cycler

(M.J.

Research, Inc.) using the following three step program: Step 1 consisted of a denaturing temperature of 94 0 C for 2 min; Step 2 consisted of 9 complete cycles of a denaturing temperature of 94 0 C for 1 min, an annealing temperature of 50 0 C for 2 min, and an extension temperature of 72 0 o for 3 min; Step 3 cycled for 19 times with a denaturing temperature 94 0 c for 1 min, an annealing temperature of 60°C for 2 min, and an extension temperature of 72 0 C for 3 min; and at the end of the last 4 cycle, the samples were held at 72 0 C for 5 min to ensure complete extension.

Band size estimation Fragment sizes in the molecular weight standard and in the Southern blot hybridization patterns were 3c calculated from migration distances. The standard S molecular sizes were fitted to a logarithmic regression model using Cricket Graph (Cricket Software, Malvern, PA). The molecular weights of the detected bands were estimated by entering the logarithmic line equation obtained by Cricket Graph into Microsoft Excel (Microsoft Corporation, Redmond, WA) in order to calculate molecular 176 weights based on migration distances observed in the Southern blot.

Since most strains contain a pspA gene and a pspC gene, it was expected that if an extra gene were present one might observe at least three pspA homologous loci in isolates MC25-28. In Hind III digests of MC25-28 each strain revealed 7.7 and 3.6 kb bands when probed with LSMpspAl3/2 (Figure 15A and 15C). In comparison, when Rxl DNA was digested with Hind III and hybridized ;o with LSMpspAl3/2, homologous sequences were detected on 9.1 and 4.2 kb fragments, as expected from previous studies with PspA (Figure 15A). Results consistent with two pspA-homologous genes in MC25-28 were obtained with two pspA-homologous genes in MC25-28 digested using four fS additional enzymes (Table 61).

Table 61. Chromosomal RFLPs with probe LSMpspA13/2 for isolates MC25-28 and Rxl Restriction Strains Examined Restriction Fragments Enzyme (sizes in kilobases) MC26 MC27 MC28 RX1 MC25-28 Rxl Hind III 7.7, 3.6 9.1, 4.2 Kpn I 11.6, 10.6 10.6, 9.8 EcoR I 8.4, 7.6 7.8, 6.6 Dra I 2.1, 1.1 1.9, 0.9 SPat I >14, 6.1 10.0, The four isolates examined are all members of a single clone of capsular type 6B pneumococci isolated from Spain. These four isolates are the first in which two PspAs have been observed, PspA and PspC, based on the observation that bands of different molecular weights were detected by different MAbs to PspA.

Mutation and immunochemistry studies have demonstrated that all of the different sized PspA bands from Rxl are made of a single gene capable of encoding a-69kDa 3; protein, supporting the assertion that two PspAs have been observed, PspA and PspC.

It has been observed that probes for the half of pspA (encoding the a-helical half of the protein) bind the pspC sequence of most strains only at a stringency of around 90%. With chromosomal digests of MC25-28, it was observed that the 5' Rxl probe LSMpspA12/6 (Figure 15D) bound two pspA homologous bands at even higher stringency. The same probe bound only the pspA containing fragment of Rxl at the higher stringency (Figure io Further characterization of the pspA gene was done by RFLP analysis of PCR amplified pspA from each strain. Since previous studies indicated that individual strains yielded only one product, and since the amplification was conducted with primers based on a known pspA sequence, it was assumed that the product amplified from each strain represented the pspA rather than the pspC gene. When MC25-28 were subjected to this procedure, an amplified pspA product of 2.1 kb was obtained from each of the four strains. When digested g° with Hha I, this fragment yielded bands of 1.1, 0.46, 0.21 and 0.19 kb for each of the four isolates. A single isolate, MC25, was analyzed with eight additional enzymes. Using each restriction enzyme, the sum of the fragments was always approximately equal to the size of whole pspA (Figure 16). These results suggested that the 2.1 kb amplified DNA represents the amplified product of only a single pspA gene. Rxl produced an amplified product of 2.0 kb and five fragments of 0.76. 0.468, 0.390, 0.349 and 0.120 kb when digested with Hha I as 3c expected from its known pspA sequence.

There are several possible explanations for the observation of PspA and PspC in these strains but not in other strains. All isolates might make PspA and PspC in culture, but MAbs generally recognize only PspA (perhaps, 3) in this isolate there has been a recombination between pspC DNA and the pspC locus, allowing that locus to make a product detected by MAb to PspA). All isolates can 178 have PspA and Pspc, but the expression of one of them generally does not occur under in vitro growth conditions. The pspc locus is normally a nonfunctional pseudogene sequence that, for an unexplained reason, has become functional in these isolates. Results from the colony immunoblotting of these isolates failed to show a detectable in vitro phase shift between either PspA type 6 (XIR278 and 2A4) or PspA type 34 (7D2) protein. This strengthens the second explanation, and suggests that the Ie second PspA is these isolates is due to the pspC gene not being turned off during in vitro growth conditions.

Presumably, in these four strains, the second PspA protein is provided by the pspc DNA sequence. At high stringency, the probe comprising the coding region of the a-helical half of PspA recognized both pspA homologous sequences of MC25-18, but not the pspc sequence of Rxl. The finding indicated that the pspc sequence of MC25-28 is more similar to the Rxl pspA sequence than the Rxl pspC sequence. If the pspC Wo sequence of these strains is more similar to pspA than most pspc sequences, it could explain why the products of pspC genes cannot generally be identified by MAbs.

EXAMPLE 11 IDENTIFICATION OF CONSERVED AND VARIABLE REGIONS OF pspA AND pspC SEQUENCES

OF

Z S. PNEUMONIAE The S. pneumoniae strains used in this study are listed in Table 62. The strains are human clinical isolates representing 12 capsular and 9 PspA serotypes.

All strains were grown at 37 0 C in 100ml of Todd-Hewitt S_ broth supplemented with 0.5% yeast extract to an S approximate density of 5 x 108 cells/ml. After harvesting of the cells be centrifugation (2900 g, 10min), the DNA was isolated, and stored at 4 0 c in TE (10 mM Tris, mM *EDTA, l e *e «e 179 Table 62. Streptococcus pneumoniae strains used.

Strain Relevant phenotype Reference WU2 D39 R3 6A RxI

DBLS

DBL6A A66 AC9 4 AC 17 AC4 0 AC107

ACIO

AC140 D109-1B BG9 709 L81905 L82233 L82006 capsular type 3, PspA type 1 Capsular type 2, PapA type 25 Nonencapsuated mutant of D39, papA type 25 Derivatitve of R36A, PspA type 25 Capsular type 5, PspA type 33 Capsular type 6A, PapA type 19 Capsular type 3, PapA type 13 Capsular type 9L, PepA type 0 Capsular type 9L, PspA type 0 Capsular type 9L, PspA type 0 Capsular type 9V, PspA type 0 Capsular type 9V, PapA type 0 Capsular type 9N, PapA type 18 Capsular type 23, PapA type 12 Capsular type 9, PspA type 0 Briles et al." 1981 Avery et al., 1944 Avery et al.: 1944 Shoemaker and Guild, 1974 Yother et al., 1986 Yother et al., 1986 Avery et al., 1944 Waltman et al. 1992 Waltman et al., 1992 Waltman et al., 1992 Waltman et al., 1992 Waltman et al., 1992 Waltman et al., 1992 McDaniel et al., 1992 McDaniel et al., 1992 Capsular type Capsular type Capsular type 4, PapA type 25 14, PapA type 0 1, PapA type 0 McDaniel McDaniel McDaniel et al., et al., et al., 1992 1992 1992 eq

C

C

4W

C

CC

C

Approximately 5Ag of chromnosomal DNA was digested with HindIII according to the manufacturer's instructions (Promega, Inc., Madison, WI) The digested DNA was subjected to electrophoresis at 35 mV overnight in 0.8% agarose gels and then vacuum-blotted onto Nytran® membranes (Schleicher Schuell, Keene, NH).

The oligonucleotides uses were based on the previously determined sequence of Rxl pspA. Their position and orientation relative to the structural domains of Rxl pspA are shown .in Figure 17. Labeling of oligonucleotides and detection of probe-target hybrids lfOt 180 were both performed with the Genius System® according to manufacturer's instructions (Boehringer-Mannhein, Indianapolis, IN). All hybridizations were done for 18 hours at 42 0 C without formamide. By assuming that 1% T base-pair mismatching results in a 1lC decrease in Tm arbitrary designations of "high" and "low" stringency were defined by salt concentration and temperature of post-hybridization washes. Homology between probe and target sequences was derived using calculated T by S established methods. High stringency is defined as and low stringency is 5 85% base-pair matching.

PCR primers, which were also used as oligonucleotide probes in Southern blotting and hybridizations, were designed based on the sequence of S pspA from pneumococcal strain Rxl. These oligonucleotides were synthesized by Oligos, Etc.

(Wilson, OR), and are listed in Table 63.

4* 181 Table 63. Oligonucleotide sequences.

Primer 5, 3, *LSM111

CCGGATCCAGCTCCTGCACCAAAAC

LSH2

GCGCGTCGACGCTTAACCCATTCACCATTGG

LSZ43

CCGGATCCTOAGCCAGAGCAGTTGGCTG

LSM4

CCGGATCCGCTCAAAGAGATTGATGAGTCTG

GCGGATCCCGTAGCCAGTAGTCTAGCTG

LSM6

CTGAGTCGACTGGAGTTTCTGAGCTGGAGC

LSM7

CCGGATCCAGCTCCAGCTCAGAIACTCCAG

LSM9

GTTTTTGGTGCAGGAGCTGG

LSM1 0

GCTATGGCTACAGGTTG

LSM12

CCGGATCCAGCGTGCCTATCTTAGGGGCTGGT

LSM1 12

GCGTCTACATRRCGGAO

PCR was done with an MJ Research, Inc., Programmable Thermal Cycler (Watertown, MA) using approximately 10 ng of genomic pneumococcal DNA as template with designated 5' and 3' primer pairs. The sample was brought to a total volume of 50 gil containing a final concentration of 50 mM KCl, 10 mM Tris-HCl (pH 1.5mM MgCl 2 0.01% gelatin, 0.5 gM of each primer, 200 /1M of each deoxynucleoside triphosphate, and 2.5 U of Tag' DNA polymerase. The samples were denatured at 940C for 2 minutes and subjected to 10 cycles consisting of: 1 min at 94 0 C, 2 min at 50 0 C, and 3 min at 720C, followed by 20 cycles of: 1 min at 94 0 C, 2 min at 600C, and 3 min at 720C. After 30 total cycles, the samples were held at 720C for an additional 5 min prior to cooling to 4 0

C.

The amplicons were then analyzed by agarose gel electrophoresis.

Oligonucleosides were used to probe HindlIl digests of DNA from 18 strains of S. pneuznoniae under conditions of low and high stringency. Each strain was also screened using a full-length pspA probe. Table 64 summarizes the results for each strain under conditions of high stringency. Strain Rxi is a laboratory derivative of the clinical isolate D39 and consequently, 182 both strains showed identical hybridization patterns and are a single column in Table 64.

J

Table 64. Summary of hybridization of OligonUcleotides with HfINDIII chromosomal restriction fragments.

Probe PoclI Wui DBI-5 D39 2 FL-RxIl 4.0.9. 3.8 3.7.5 l1. a LSM12 4.0.9. 3.6 3.7,5.

1 8 LSM3 4.0 3.8 LSM4 4.0 LSM7 4.09. 3.8 3 ISMI 1 4.0,9. 3.8 3.7.5.

1 1 a 4.0.9. 3.8 3.7 LSM2 4.0 0 3.7 0616

A

3.0.3, 4 3.0.3.

4 3.4 3.4 ABB AC94 38.4. 3.683 3 4.3 c 383 6.3 3 3.6 FAC

I?

3.8e.

3 3.01.0, 3 AC40 3.2.3.

6 3.2,3.

Strains ACIO AC10 AC14 081 B G7DG5B8 90 L 2 3 8 0 7 0 0 9 BG7 BG8L10L2]L20 9 c 36 3.2.3. 4.0.8. 3.0.4. 3.3.4. 3.3,4.7 1.4.3.2 3.6.5. 8.2.3. 4.3.6.4 6 0 0 7 3.6 2 7 04.0 3.3.4, 2.29.0 13. 3.3.

0 7 3.

3.2,3.

a 3.2 3.8.6. 3.2 3 3.6 -FurI-ength pspA of *train AxI.

Onumbers are size in kilobase peirs.

'no hybridization observed with corresponding probe.

i 184 The only strain which did not have more than one pspA-homologous HindIII fragment was WU2, which was previously shown using a full-length pspA probe. Even at high stringency, six of the eight probes detected more than one fragment in at least one of the 18 strains (Table 64). LSM7, 10 and 12 hybridized with two fragments in more than one-half of the strains, and the fragments detected by the oligonucleotide probes were identical in size to those detected by the full-length S pspA probe. Moreover, the same pairs of fragments were frequently detected by probes derived from the 3' as well as the 5' region of Rxl pspA. These results suggested that the HindIII fragments from different isolates include two separate but homologous sequences, rather I& than fragments of a single pspA gene. Based on the diversity of the hybridization patterns and the size of restriction fragments, it is clear that pspA and pspc sequences are highly diverse and that these loci have considerable sequence variability as determined by 0> location of HindIII recognition sites.

Oligonucleotides which hybridize with a single restriction fragments in each strain were assumed to be specific for pspA. At high stringency, LSM3 and LSM4 detected only a single HindIII fragment in the strains 3C with which they reacted. Restriction fragments containing homology to LSM3 or LSM4 were the same as those which hybridize with all of the other homologous probes. This suggested that LSM3 and LSM4 specifically detect pspA rather than the pspC sequence. That LSM3 o hybridizes with a single restriction fragment of WU2 further confirmed that this oligonucleotide is specific for pspA. Sequences from the portion of the gene encoding the second proline region (LSM111) and the Cterminus (LSM2) appeared to be relatively specific for pspA since they generally detect only one of the HindIII fragments of each strain.

I I 185 Oligonucleotides LSM12 and LSM10 were able to detect the most conserved epitopes of pspA and generally hybridize with multiple restriction fragments of each strain (Table 65). LSM7 was not as broadly crossreactive, but detected two pspAs in 41% of strains including almost 60% of the strains with which it reacts.

Thus, sequences representing the leader, first proline region, and the repeat region appear to be relatively conserved not only within pspA but between the pspA and pspC. sequences. LSM3, 4, and 5 hybridize with the smallest number of strains of any oligonucleotides (29-35 percent), suggesting that the a-helical domain is the least conserved region within pspA. In strains BG58C and L81905 oligonucleotides detect more than two HindIII fragments containing sequences with homology to pspA.

Because of the absence of HindIII restriction sites within any of the oligonucleotides it was unlikely that these .multiple fragments result from the digestion of chromosomal DNA within the target regions. Also, the 32 additional restriction fragments were detected at high stringency by more than one oligonucleotide. Possibly, in these two strains, there are three or four sequences with DNA homology to some portions of pspA. The probes most consistently reactive with these additional 2< sequences are those for the leader, the alpha-helical region, and the proline-rich region.

The oligonucleotides used as hybridization probes were also tested for their utility as primers in the polymerase chain reaction (PCR). Amplification of o pspA from 14 strains of S. pneumoniae comprising 12 different capsular types was attempted with the primers listed in Table 63. LSM2, derived from the 3' end of pspA, were able to amplify an apparent pspA sequence from each of 14 pneumococcal strains when used in combination S with LSM111, which is within the sequence of pspA encoding the proline-rich region. Combinations of LSM2 with primers upstream in pspA were variably successful in 186 amplifying sequences (Table 65). The lowest frequency of amplification was observed with LSM112 which was derived from the Rxl sequence 5' to the pspA start site. This oligonucleotide was not used in the hybridization studies. DNA fragments generated by PCR were blotted and hybridized with a full-length pspA probe to confirm homology to pspA.

Further evidence for variability at the pspA locus comes from the differences in the sizes of the amplified pspA gene. When PCR primers LSM12 and LSM2 were used to amplify the entire coding region of PspA, PCR products from different pneumococcal isolates ranged in size from 1.9 to 2.3 kbp. The regions of pspA which encode the a-helical, proline-rich, and repeat domains were amplified from corresponding strains and variation in pspA appears to come from sequences within the ahelical coding region.

Table 65. Amplification of pspA by PCR using the indicated oligonucleotides as 5' primers in combination with the 3' primer LSM2.

primer Domain Amplified/ Percent Tested Amplified LSM112 -35 (upstream) 2/14 14 LSM12 leader 8/14 57 LSM3 a-helical 3/14 21 LSM7 proline 12/14 86 LSM1I proline 14/14 100 These studies have provided a finer resolution

V.

map of the location of conserved and variable sequences 11: :.go within pspA. Additionally, regions of divergence and identity between pspA and the pspC sequences have been identified. This data confirmed serological studies, and demonstrated that pspA and pspC sequences are highly 2 187 variable at the DNA sequence level. The diversity of HindIII restriction fragment polymorphisms contained pspA and the pspc sequence supported earlier data using larger probes that detected extensive variability of the DNA in Sand around these sequences.

A useful pspA-specific DNA probe would identify Rxl and WU2 pspA genes, in which restriction maps are known, and would identify only a single restriction fragment in most strains. Two probes, LSM3 and LSM4, do S not hybridize with more than one HindIII restriction fragment in any strain of pneumococcus. Both of these oligonucleotides hybridize with Rxl pspA and LSM3 hybridizes with WU2 pspA. However, each of these probes hybridize with only four of the other 15 strains. When S these probes identify a fragment, however, it is generally also detected by all other Rxl-derived probes.

Oligonucleotides from the second proline-rich region (LSM111) and the C-terminus of pspA (LSM2) generally identify only one pspA-homologous sequence at high Sstrihgency. Collectively, LSM111, 2, 3 and 4 react with S16 of the 17 isolates and in each case revealed a consensus DNA fragment recognized by most or all of the oligonucleotide probes.

When an oligonucleotide probe detected only a a single DNA fragment it was presumed to be pspA. If the probe detected multiple fragments, it was-presumed to hybridize with pspA. If the probe detected multiple Sfragments, it was presumed to hybridize with pspA and the 188 pspC sequence. Based on these assumptions the most variable portion between pspA and pspC is the region immediately upstream from the -35 promoter region and that portion encoding the a-helical region. The most conserved portion between pspA and pspC was found to be the repeat region, the leader and the proline-rich region sequences. Although only one probe from within the repeat region was used, the high degree of conservation among the 10 repeats in the Rxl sequence makes it likely that other probes within the repeat sequences would give similar results.

The portion of Rxl pspA most similar to the pspC sequence was that encoding the leader peptide, the upstream portion of the proline-rich region, and the repeat region. The repeat region of PspA has been shown to be involved in the attachment of this protein to the pneumococcal cell surface. The conservation of the repeat region within pspc sequences suggests that if these loci encode a protein, it may have a similar .o functional attachment domain. The conservation of the e leader sequence between pspA and the pspC sequence was also not surprising since similar conservation has been reported for the leader sequence of other proteins from gram positive organisms, such as M protein of group

A

f streptococci (Haanes-Fritz, E. et al., Nucl. Acids Res.

1988; 16: 4667-4677).

In two strain, some oligonucleotide probes identified more than two pspA-homologous sequences. In I C 189 these strains, there was a predominant sequence recognized by almost all of the probes, and two or three additional sequences share homology with DNA encoding the leader, a-helical, and proline region, and they have no homology with sequences encoding the repeat region in the C-terminus of PspA. These sequences might serve as cassettes which can recombine with pspA and/or the pspC sequences to generate antigenic diversity.

Alternatively, the sequences might encode proteins with 00 very different C-terminal regions and might not be surface attached by the mechanism of PspA.

Oligonucleotides which hybridize with a single chromosomal DNA fragment were used as primers in PCR to examine the variability of domains within pspA. These results demonstrate that full-length pspA varies in size among strains of pneumococci, and that this variability is almost exclusively the result of sequences in the alpha-helix coding region.

EXAMPLE 12 CLONING OF PspC SChromosomal DNA from S. pneumoniae EF6796, S serotype 6A clinical isolate, was isolated by methods including purification through a cesium chloride S. gradient, as described in Example 8. The HindII-EcoRI fragment of EF6796 was cloned in modified pZero vector (Invitrogen, San Diego, CA) in which the Zeocinresistance cassette was replaced by a kanamycin cassette (shown in Figure 18). Recombinant plasmids were S electroporated into Escherichia coli TOPIOF' cells

[F'

190 {lacIqTetR} mcrA A(mrr-hsdRMS-mcrBC) 080acZAM15 AlacX74 deoR recAl araD139 A(ara-leu)7967 galU galK rpsL endAl nupG] (Invitrogen).

The 5' region of pspA.Rxl does not hybridize to pspc sequence at high stringencies by Southern analysis.

Utilizing both the full-length Rxl pspA probe, and a probe containing the sequence encoding a-helical region of PspA, it was possible to identify which DNA fragment contained pspA and which fragment contained the pspc ,c locus. The pspC locus and the pspA gene of EF6796 were mapped using restriction enzymes. After digestion of chromosomal DNA with HindIII, the pspc locus was localized to a fragment of approximately 6.8 kb.

Following a double digest with HindIII and EcoRI, the pspC locus was located in a 3.5 kb fragment. To obtain the intact pspC gene of EF6796, chromosomal DNA was digested with HindIII, separated by agarose gel electrophoresis, the region between 6 and 7.5 kb purified, and subsequently digested with EcoRI. This ^O digested DNA was analyzed by electrophoresis, and DNA fragments of 3.0 to 4.0 kb were purified (GeneClean, Biol01, Inc., Vista, CA). The size-fractionated DNA was then ligated in HindIII-EcoRI-digested pZero, and electroplated into E. coli TOP1OF' cells. Kanamycin- O( resistant transformants were screened by colony blots and probed with full-length pspA. A transformant, LXS200, contained a vector with a 3.5 kb insert which hybridized to pspA.

191 Sequencing of pspC in pLXS200 was completed using automated DNA sequencing on an ABI 377 (Applied Biosystems, Inc., PLACE). Sequence analyses were performed using the University Of Wisconsin Genetics Computer Group (GCG) programs supported by the Center for AIDS Research (P30 AI27767), MacVector 5.0, Sequencer 2.1, and DNA Strider programs. Sequence similarities of pspC were determined using the NCBI BLAST server. The coiled-coil structure predicted by pspC sequence was o analyzed using Matcher.

A gene probe for cloning the pspC locus Two oligonucleotide primers, N192 and C558 (shown in Figure 19), have been used previously to clone fragments homologous to the region of Rxl pspA encoding ;C amino acids 192-588 from various pneumococcal strains.

These primers are modifications (altered restriction sites) of LSM4 and LSM2 which were previously shown to amplify DNA encoding the C-terminal 396 amino acids of PspA.Rxl (Figure 17); this includes approximately 100 S amino acids of the a-helical region, the proline rich region, and the C-terminal choline-binding repeat region.

Using primers N192 and C558, a 1.2 kb fragment from strain EF6796 was amplified by PCR, and subsequently cloned in pET-9A (designated PRCT135). This insert was S then partially sequenced.

Independently, a larger pspA fragment from strain EF6796 was made using primers LSM13 and SKH2 192 (shown in Figure 19) for the purpose of direct sequencing of serologically diverse pspA genes.

The LSM13 and SKH2 primer pair result in the amplification of the 5' end of most pspA gene(s) encoding the .upstream promoter, the leader peptide, the a-helical, and the proline-rich regions (amino acid -15 to 450) (Figure 20). From the strain EF6796, the LSM13 and SKH2 primers amplified a 1.3 kb fragment (pspA.EF6796), which was sequenced. The sequence from pRCT135 and the S LSM13/SKH2 PCR-generated fragment pspA.EF6796 was not identical. The fragment obtained by PCR using primers LSM13 and SKH2 was designated pspA based on its location within the same chromosomal location as pspA.Rxl. The cloned fragment in pRCT135 was assumed to represent the sequence of the second gene locus, pspC, known to be present from Southern analysis. Both genes have significant similarity to the corresponding regions of the prototype pspA gene from strain Rxl. The second gene locus was called pspc, in recognition of its distinct O chromosomal location, not sequence differences from the prototype pspA gene.

Analysis of the nucleotide and amino acid sequenc of pspC EF6796 To test the hypothesis that pRCT135 represented S pspC of EF6796, and to further investigate pspC, the entire EF6796 pspC gene was cloned as a 3.4 kb HindIII- EcoR1 fragment forming pLXS200. DNA sequence of the pspC-containing clone pLXS200 revealed an open reading 193 fram of 2782 nucleotides based on the analysis of putative transcriptional and translation start and stop sites (Figure 21). The predicted open reading frame encodes a 105 kDa protein which has an estimated pi of c 6.09.

PspA.Rxl and PspC.EF6796 are similar in that they both contain an a-helical region followed by a proline-rich domain and repeat region (Figure However, there are several features of the amino acid S sequence of PspC which are quite distinct from PspA.

From comparisons at the nucleotide as well as the predicted amino acid sequence, it is apparent that the region of strong homology between PspC and PspA begins at amino acid 458 of PspC (amino acid 147 of PspA) and S extends to the C-terminus of both proteins (positions 899 and 588 respectively). The predicted amino acid sequence of.PspC.EF6796 and PspA.Rxl are 76% similar and 68% identical based on GCG Bestfit program for this region (Figure 22). The nucleotide sequence identity between Q 0 pspc and pspA is 87% for the same region. Eight bases upstream of the ATG start site is putative ribosomal binding site, TAGAAGGA. The proposed transcriptional start -35 (TATACA) and -10 (TATAGT) regions are located between 258 to 263 and 280 to 285, respectively (Figure 21). A potential transcriptional terminator occurs at a stem loop between nucleotides 3237 through 3287. The putative signal sequence of PspC is typical of other gram positive bacteria. This region consists of a charged 194 region followed by a hydrophobic core of amino acids.

A

potential cleavage site of the signal peptide occurs at amino acid 37 following the Val-His-Ala. The first amino acid of the mature protein is a Glu residue.

4 Other than features similar to all signal sequences, there is no homology in this region between pspA and pspC. This confirms that pspc is present in a separate chromosomal locus from that of pspA. The signal sequence and upstream region have striking similarity to to the similar regions of S. agalactiae 8 antigen (accession number X59771). The Bantigen of Group B streptococci is a cell surface receptor that binds IgA. Similarity to the bac gene ends with the start of the mature protein of PspC, and the nucleotides are 75% identical in this region. Thus, although pspC is in a very similar chromosomal locus to the B antigen, it is clearly a distinct protein.

The N-terminus of PspC is quite different from the N-terminus of PspA. Prediction of the secondary structure utilizing Chou-Fausman analysis (Chao, P.Y. et al., Adv. Enzymol. Relat Areas Mol. Biol. 1978: 47: 148), suggests that the structure of amino acids 16 to 589 of PspC is predominately a-helical. The Matcher program.was used to examine periodicity in the a-helical region of PspA. The characteristic seven residue Periodicity is maintained by having hydrophobic residues at the first and fourth positions (a and d) and hydrophobic residues at the remaining positions. The

S

I

195 coiled-coil region of the a-helix of Pspc (between amino acid 32 to 600) has three breaks in the heptad repeat (Figure 23). These disturbances in the 7 residue periodicity occur at amino acids 99 to 104, 224 to 267 and 346 to 350. The a-helical region of PspA has seven breaks in the motif, each break ranging from a few amino acids to 23 amino acids each. In contrast, the three breaks in the coiled-coil motif of PspC involve 5, 43 and 4 amino acids, respectively.

;c The sequence encoding the a-helical region of PspC contains two direct repeats 483 nucleotides (160 amino acids) long which are 88% percent identical at the nucleotide level. These repeats, which occur between nucleotides 562 to 1045 and nucleotides 1312 to 1795, are conserved both at the nucleotide and amino acid level (amino acids 188 to 348 and 438 to 598) (Figure 24).

PspA lacks evidence for any repeats this prominent within the a-helical region. These repeat regions could provide a mechanism for recombination that could alter the No. terminal half of the PspC molecule. Although repeat motifs are common in bacterial surface proteins, a direct repeat this large or separated by a large spacer region is novel. The evolutionary significance of this region is not known. A Blast search of the repeat region and 'I the 267 nucleotide bases between them revealed no sequence with significant homology at the nucleotide or amino acid level. However, one of the structural breaks in the coil-coiled region of PspC is the region between 196 the two repeats. Perhaps some deviation from coiled-coil structure between the two repeats is critical to maintain the a-helical structure.

Previous studies have shown that a major crossprotective region of PspA comprises the C-terminal 1/3 of the a-helical region (between residues 192 and 260 of PspA.Rxl). This region accounts for the binding of 4 of cross-protective immunity in mice. Homology between PspC and PspA begins at amino acid 148 of PspA, thus including the region from 192 299. The homology between PspA and the PspC includes the entire PspC sequence C-terminal of amino acid 486. Based on the fact that PspA and PspC are so similar in this region known to be protection-eliciting, PspC is also likely to be a 'f protection-eliciting molecule. Because of close sequence and conformational similarity of the proteins in this region, antibodies specific for the region of PspA between amino acid 148 and 299 should cross-react with PspC and thus afford protection by reacting with PspC and S PspA. Likewise, immunization with the PspC would be expected to elicit antibodies cross-protective against PspA. The differences between PspC of strain EF6796 and PspA of strain Rxl is no greater than the differences between many additional PspAs, which have been shown to 25 be highly cross-protective.

A proline-rich domain exists between amino acid Sj 590 to 652. The sequence, PAPAPEK, is repeated six times in this region. This region is very similar to the 197 proline-rich region of PspA.Rxl which contains the sequence PAPAP repeated eight times in two proline-rich regions. These two regions of PspA.Rxl are separated by 27 charge amino acids; no such spacer region is present in PspC.

Many cell surface proteins of other gram positive bacteria contain proline-rich regions. These are often associated with a domain of protein that is predicted to be near the cell wall murein layer when the ,0 protein is cell-associated. For example, in M proteins of S. pyogenes this domain contains both a Pro- and Glyrich regions. The fibronectin-binding protein of S.

pyogenes, S. dysgalactiae, and Staphylococcus aureus contains a proline-rich region with a three-residue lb periodicity (pro-charged-uncharged) that is not found in PspA or PspC. An M-like protein of S. equi contains a proline-rich region that is comprised of the tetrapeptide PEPK. This region lacks glycine normally found in the proline regions of M-proteins. The last proline repeat o region of this molecule is PAPAK, which is more similar to the proline-region of PspA and PspC than it is to Mproteins.

Proline-rich regions of gram positive bacterial proteins have been reported previously to transit the cell wall. The differences in proline-rich regions of proteins from diverse bacteria may reflect differences in protein function or possibly subtle differences in cell wall function. Proline-rich regions are thought to be 198 responsible for aberrant migration of these proteins through SDS-polyacrylamide gels.

The repeat region of PspC is a common motif found among several proteins in gram positive organisms.

d Autolysin of S. pneumoniae, toxins A and B of Clostridium difficile, glucosyltransferases from s. downei and S.

mutans, and CspA of C. acetobiltylicum all contain similar regions. In PspA these repeats are responsible for binding to the phosphatidylcholine of teichoic acid i' and lipoteichoic acid in cell wall of pneumococci.

However, bacterial proteins containing C-terminal repeats are secreted, which may imply either a lost or gained function. Although all of these proteins have similar repeat regions the similarity of the repeat regions of ,T PspA and PspC is much greater than that of PspC to the other proteins (Table 66).

Interestingly, PspC like PspA has a 17 amino acid partially hydrophobic tail. The function of this 17 amino acid region is unknown. In the case of PspA it has been shown that mutants lacking the tail bind the surface of pneumococci as well as PspAs in which the tail is expressed. Presently, it is now known whether PspC is attached to the cell surface or secreted.

PspA and PspC proteins both have a-helical 2" coiled-coil regions, proline-rich central regions, repeat regions, with a choline binding motifs, and the Cterminal 17 amino acid tail. PspA and PspC share three 199 regions of high sequence identity. One of these is a protection-eliciting region present within the a-helical domain. The other two regions are the prolinerich domain and a repeat domain shared with other choline S' binding proteins and thought to play a role in cell surface association. The similarity throughout most of the structure of the PspA and the PspC molecules raises the possibility that the two molecules may play at least slightly redundant functions. However, the fact that the '0 N-terminal half of the protein is not homologous to any of the a-helical sequence of PspA suggests the PspC and PspA may have evolved for at least somewhat different roles on the cell surface. One of the most striking differences between the two molecules is the single repeat in the a-helical region of Pspc. Although neither the exact function of PspA nor of PspC are known, the observation that a major cross-protective region of PspA is highly homologous with a similar region of Pspc, raises the possibility that both molecules are 2o protection-eliciting and elicit cross-protective antibodies.

The sequence similarity between the promoter region of the pspc gene and the bac gene from group B streptococci is very intriguing. It implies that an interspecies recombination event has occurred and, this interspecies recombination has contributed to the Sevolution of the pspC. The pspC gene thus has a chimeric structure, being partially like pspA and partially like 4 *i I I 200 the 6 antigen. In the latter case, all protein similarity is limited to the signal sequence. Similar interspecies recombination events have contributed to the evolution of the genes encoding penicillin binding protein.

Using analogous procedures, a second PspC sequence was isolated from strain D39 of S. pneumoniae.

Figures 25 to 29 show the sequence data of PspC from strain D39, complete from upstream of the promoter 'C through the proline-rich region. Strain D39 has the same genetic background as strains Rxl, from which pspA was sequenced. D39 and Rxl have the same pspc gene based on Southern blot analysis.

The alpha-helical encoding region of the D39 pspC gene is one third of the size of the homologous region from the EF6796 pspC gene. The proline-rich region of the D39 pspC gene was more similar to Rxl pspA than to EF6796 pspc. Even so, the two pspC genes were 86% identical at the nucleotide sequence, and 67% S identical at the amino acid level.

In the alpha-helical sequence of EF6797 pspc a strong repeat was observed. This was absent in the pspc sequence of D39. The D39 pspc sequence also lacks a leader sequence, found in the EF6797 pspC sequence.

This data strongly indicates that there is variability in the structure of pspc, similar to previous observations for pspA. In the case of pspC, however, the t

I

«t t I 201 extent of variability appears to be even greater than that which has been observed for pspA.

202 Table 66.

Having thus described in detail certain ir preferred embodiments of the present invention, it is to be understood that the invention defined by the appended claims is not to be limited by particular details set forth in the above description, as many apparent variations thereof are possible without departing from the spirit or scope thereof.

3** 0 203

REFERENCES

1. Alexander, Lock, Peeters,

C.C.A.M.,

Poolman, Andrew, Mitchell,

T.J.,

Hansman, Paton, Immunization of mice with pneumolysin toxoid confers a significant degreee of protection against at least nine serotypes of Streptococcus pneumoniae, Infection and Immunity 1994, 62:5683-5688.

2. Amsbaugh, Hansen, Prescott, Stashak, Barthold, Baker, Genetic control of the antibody response to type III pneumococcal polysaccharide in mice. I. Evidence that an X-linked gene plays a decisive role in determining responsiveness, J. Exp. Med. 1972, 136:931-949.

3. Anonymous, Pneumococcal polysaccharide vaccine,

MMWR

1981, 30:410-419.

4. Austrian, Pneumococcal Vaccine: Development and Prospects, Am. J. Med. 1979, 67:547-549.

Avery, McLeod, and McCarty, Studies on the chemical nature of the substance inducing transformation of pneumococcal types, Induction of transformation by a desoxyribonucleic acid fraction isolated from pneumococcus type III., J. Exp. Med.

1944, 79:137-158.

6. Avery, Goebel, Chemoimmunological studies of the soluble specific substance of pneumococcus. The isolation and properties of the acetyl polysaccharide of pneumococcus type 1 J.

Exp. Med. 1933, 58:731-755.

7. Badenes, and Dan E. Parfitt, Reducing background and interference on Southern Blots probed with nonradioactive chemiluminescent probes, BioTechniques 1994, 17:622-624.

S 8. Berry, Yother, Briles, Hansman

D.,

Paton, Reduced virulence of a defined pneumolysin-negative mutant of Streptococcus pneumoniae, Infect. Immun. 1989, 57:2037-42.

9. Berry, Lock, Hansman, Paton

J.C.,

Contribution of autolysin to virulence of Streptococcus pneumoniae, Infect Immun 1989, 57:2324-30.

Birnboim H.C. and J. Doly, A rapid alkaline extraction procedure for screening recombinant plasmid DNA, Nuc. Acids Res., 1979, 7:1513.

204 11. Briese, Hakenbeck, Interaction of the pneumococcal amidase with lipoteichoic acid and choline, 1985, 146:417-427.

12. Briles, J. Yother and L.S. McDaniel, Role of pneumococcal surface protein A in the virulence of Streptococcus pneumoniae, Rev. Infect. Dis. 1988, 10:S372-374.

13. Briles, Horowitz, McDaniel,

L.S.,

Benjamin, Jr., Claflin, Booker

C.L.,

Scott, Forman, C. Genetic control of susceptibility to pneumococcal infection, Curr. Top.

Microbiol. Immunol. 1986, 124:103-120.

14. Briles, Forman, Horowitz, Volanakis, Benjamin, Jr., McDaniel, Eldridge, Brooks, Antipneumococcal effects of Creactive protein and monoclonal antibodies to pneumococcal cell wall and capsular antigens, Infect. Immun. 1989, 57:1457-1464.

Briles, Nahm, Schroer, Davie,

J.,

Baker, Kearney, Barletta,

R.,

Antiphosphocholine antibodies found in normal mouse serum are protective against intravenous infection with type 3 Streptococcus pneumoniae, J. Exp. Med.

1981, 153:694-705.

16. Briles, M.J. Crain, B.M. Gray, C. Forman and J. Yother, A strong association between capsular type and mouse virulence among human isolates of Streptococcus pneumoniae, Infect. Immun. 1992, 60:111-116.

17. Briles Forman, Crain, Mouse antibody to phosphocholine can protect mice from infection with mouse-virulent human isolates of Streptococcus S-pneumoniae, Infect. Immun. 1992, 60:1957-62.

18. Brooks-Walter, McDaniel, Hollingshead Briles, D.E. Restriction fragment length polymorphisms of pspA of Streptococcus pneumoniae reveal a genetic polymorphism, Submitted.

19. Brooks-Walter A. and L.S. McDaniel, 1994, Unpublished data.

4 20. Cadoz J. Armand, F. Arminjon, Michel,

M.

Michel, F. Denis and G. Schiffman, A new 23 valent pneumococcal vaccine: immunogenicity and reactogenicity in adults, J. Biol. Stand. 1985, 13:261.

21. Chiu, Greenberg, Marcy, Wong, Chang, Chiu, Ward, Mucosal V 205 antibody responses in infants following immunization with Haemophilus influenzae, Pediatr. Res.

Abstracts 1994, 35:10A.

22. Cohen Parry alpha-helical coiled coils: more facts and better predictions, Science 1994, 236:488-9.

23. Cowan, M.J. Ammann, Wara, Howie,

V.M.

Schultz, Doyle, Kaplan, Pneumococcal polysaccharide immunization in infants and children, Pediatrics 1978, 62:721-727.

24. Crain Unpublished data.

Crain, Waltman, II, Turner, Yother, Talkington, McDaniel, Gray, Briles, Pneumococcal surface protein

A

(PspA) is serologically highly variable and is expressed by all clinically important capsular serotypes of Streptococcus pneumoniae, Infect.

Immun. 1990, 58:3293-3299.

26. Dagen, Melamed, Abramson, Piglansky, L., Greenberg, Mendelman, Bohidar, Ter- Minassian, Cvanovich, Lov, Rusk, C., Donnelly, Yagupsky, Effect of heptavalent pneumococcal-OMPC conjugate vaccine on nasopharyngeal carriage when administered during the 2nd year of life, Pediatr. Res. 1995, 3 7:172A.

27. Davis Boststein Roth In: A manual for genetic engineering advanced bacterial genetics, Cold Spring Harbor, NY (Cold Spring Harbor Laboratory Press) 1980.

28. Dillard, Yother, J. Genetic and molecular characterization of capsular polysaccharide biosynthesis in Streptococcus pneumoniae type 3, Mol. Microbiol. 1994, 12:959-972.

29. Douglas R.M. and H.B. Miles, Vaccination against Streptococcus pneumoniae in childhood: lack of demonstrable benefit in young Australian children, J. Infect. Dis. 1984, 149:861.

Douglas J.C. Paton, S.J. Duncan and D.J.

P. Hansman, Antibody response to pneumococcal vaccination in children younger than five years of age, J. Infect. Dis., 1983, 148:131.

31. Farley, King, Nair, Hines,

S.E.,

Tressler, Vink, Invasive pneumococcal disease among infected and uninfected children of mothers with immunodeficiency virus infection,

J.

Pediatr. 1994, 124:853-858.

206 32. Fattom, Vann, Szu, Sutton,

A.,

Bryla, Shiffman, Robbins, J.B. Schneerson, R. Synthesis and physiochemical and immunological characterization of Oneumococcus type 12F polysaccharide-diphtheria toxoid conjuates, Infect.

Immun. 1988, 56:2292-2298.

33. Fedson, Pneumococcal vaccination in the prevention of community-acquired pneumonia: an optimistic view of cost-effectiveness, Sem Resp.

Infect. 1993, 8:285-293.

34. Feldman, Munro Jeffery et al.

Pneumolysin induces the salient histologic features of pneumococcal infection in the rat lung in viv e Am. J. Respir. Cell. Mol. Biol. 1992, 5:416-23.

Filice, L.L. Van Etta, C.P. Darby and D.W.

Fraser, Bacteremia in Charleston County, South Carolina. Am. J. Epidemiol., 1986, 123:128.

36. Fischetti, Pancholi, Schneewind,

O.

Conservation of a hexapeptide sequence in the anchor region of surface proteins from gram-positive cocci, Mol. Microbiol., 1990, 4:1603-1605.

37. Forrester D.W. Jahigen and F.M. LaForce, Inefficacy of pneumococcal vaccine in a high-risk population, Am. J. Med. 83:425.

38. Garcia, Garcia, Lopez, Overproduction and rapid purifcation of the amidase of Streptococcus pneumoniae, Arch. Microbiol. 1987, 149:52-56.

39. Giebink Preventing pneumococcal disease in children: recommendations for using pneumococcal vaccine, Pediatr. Infect Dis., 1985, 4:343.

Giebink The microbiology of otitis media, Pediatr. Infect. Dis. 1989, 8:S18.

41. Gillespie Aspects of pneumococcal infection including bacterial virulence, host response and vaccination, J. Med. Microbiol., 1989, 28:237.

42. Gotschlich, Goldschneider, Lepow,

M.L.,

Gold, R. The immune response to bacterial polysaccharides in man, In: Antibodies in human S"diagnosis and therapy, (Ed. Haber, Krause,

R.M.)

Raven, New York, 1977, 391-402.

43. Gray Pneumococcal infection in an era of multiple antibiotic resistance, Adv. Ped. Inf. Dis.

1995; In press.

207 44. Haanes-Fritz, Kraus, Burdett, Dale, Beachey, and Cleary, Comparison of the leader sequences of four group A streptococcal

M

protein genes. Nucl. 'Acids. Res., 1988, 16:4667- 4677.

Hanahan D. Studies on transformation of Escherichia coli with plasmids, J. Mol. Biol. 1983, 166:557-80.

46. Kauppi, Eskola, Kathty, Influenza type b (Hib) conjugate vaccines induce mucosal IgAl and IgA2 antibody responses in infants and children, ICAAC Abstracts 1993, 33:174.

47. Kennedy, Derousse, Immunologic response of 12 -18 month old children to licensed pneumococcal polysaccharide vaccine primed with Streptococcus pneumoniae 19F conjugate vaccine, ICAAC 1994, 34th annual meeting, 236.

48. Laemmli UK. Cleavage of structural proteins during the assembly of the head of bacteriophage T4 Nature 1970, 227:680-5.

49. Langermann, S. Palaszynski, Burlein,

J.E.,

Koenig, Hanson, Briles, Stover, Protective humoral response against pneumococcal infection in mice elicited by recombinant Bacille Calmette-Gudin vaccines expressing PspA, J. Exp. Med. 1994, 180:2277-2286.

Leinonen A. Sakkinen, R. kalliokoski,

J.

Luotenen, M. Timonen and P.H. Mekela, Antibody response to 14-valent pneumococcal capsular polysaccharide vaccine in preschool age children, Pediatr. Infect. Dis. 1986, 5:39.

51. Lock, Patent, Hansman, Purification and immunologic characterization of neuraminidase produced by Streptococcus pneumoniae, Microbial Pathogenesis 1988, 4:33-43.

52. Lock, Hansman, Paton, Comparative efficacy of autolysin and pneumolysin as immunogens Sprotecting mice against infection by Streptococcus pneumoniae, Microbial Pathogenesis 1992, 12:137-143.

53. Lock, Paton, Hansman, Comparative efficacy of pneumococcal neuraminidase and pneumolysin as immunogens protective against Streptococcus pneumoniae, Microb. Pathog. 1988, 5:461-7.

54. Makela, M. Leinonen, J. Pukander and P. Karma *A study of the pneumococcal vaccine in prevention of 208 clinically acute attacks of recurrent otitis media, Rev. Infect. Dis. 1981, 3:S124.

McCarty, In: The iransforming principle, New York, Norton, 1985, 252.

56. McDaniel, Ralph, McDaniel, Briles, Localization of protection-eliciting epitopes on PspA of Streptococcus pneumoniae between amino acid residues 192 and 260, Microb. Pathog. 1994, 17:323-337.

57. McDaniel, Scott, Widenhofer, Carroll Briles, Analysis of a surface protein of Streptococcus pneumoniae recognized by protective monoclonal antibodies, Microb. Pathog. 1986, 1:519- 531.

58. McDaniel, Brooks-Walter, Briles

D.E.,

Swiatlo, E. Oligonucleotides identify conserved and variable regions of pspA and pspA-like sequences of Streptococcus pneumoniae, Mol. Microbiol., Submitted.

59. McDaniel L.S. and D.E. Briles, In: Monoclonal antibodies against bacteria, (Orlando, FL, Academic Press, Inc.), 1986, 143.

McDaniel, Yother, Vijayakumar,

M.,

McGarry, Guild, Briles, Use of insertional inactivation to facilitate studies of biological properties of pneumococcal surface protein A (PspA), J. Exp. Med. 1987, 165:381-394.

61. McDaniel, Sheffield, Delucchi, and Briles, D.E. (1991) PspA, a surface protein of Streptococcus pneumoniae, is capable of eliciting protection against pneumococci of more than one capsular type, Infect Immun 59:222-228.

62. McDaniel, Scott, Kearney, Briles, D.E. Monoclonal antibodies against protease sensitive pneumococcal antigens can protect mice Sfrom fatal infection with Streptococcus pneumoniae, J. Exp. Med. 1984, 160:386-397.

*A

63. McDaniel, McDaniel, In: Analysis of the gene encoding type 12 PspA of S. pneumoniae EF5668, (Ferretti Gilmore Klaenhammer

T.R.,

SBrown F. Genetics of Streptococci, Enterococci and Lactococci (Basel: Karger) 1995, 283-6.

S 64. McDaniel, Sheffield, Swiatlo,

E.,

Yother, J. Crain, M.J, and Briles, Molecular localization of variable and conserved regions of pspA, and identification of additional pspA-

~II

209 homologous sequences in Streptococcus pneumoniae, Microbial Pathogenesis, 1992, 13:261-269.

Meade, S.R. Long, C.B. Ruvkin, S.E. Brown and F.M. Ausubel, Physical and genetic characterization of symbiotic and auxotrophic mutants of Rhizobium melioti induced by transposon Tn5 mutagenesis,

J

Bacteriol 1982, 149:114-122.

66. Meinkoth, and Wahl, Hybridization of nucleic acids immobilized on solid supports, Anal. Biochem.

(1984) 138:267-284.

67. Mond, Lees, Snapper, T cellindependent antigens type 2, Ann. Rev. Immunol 1995, 13:655-692.

68. Mufson, Streptococcus pneumoniae. In: Principles and Practice of Infectious Diseases, Mandell Douglas Jr., Bennett eds.

(New York: Churchill Livingston) 1990, 1539-50.

69. Munoz, J.M. Musser, M. Crain, D.E. Briles,

A.

Marton, A.J. Parkinson, U. Sorensen and A. Tomasz, Geographic distribution of penicillin-resistent clones of Streptococcus pneumoniae: characterization by penicillin-binding protein profile, surface protein A typing, and multilocus enzyme analysis, Clinic. Infect. Dis. 1992, 15:112-118.

Murray, and W.F. Thompson, Rapid isolation of high molecular weight plant DNA, Nucl. Acids. Res.

1980, 8:4321-4325.

71. Musher Infections caused by Streptococcus pneumoniae: Clinical spectrum, pathogenesis, immunity, and treatment, Clin. Infect. Dis., 1992, 14:801.

72. Musher, Raizan, Weinstein, L. The effect of Listeria monocytogenes on resistance to pneumococcal infection. Soc. Exp. Bio. and Med., 1970, 135:557-560.

S 73. Nordenstam B. Anderson, D.E. Briles, J. Brooks, A. Oden, A. Svanborg and C.S. Eden, High antiphosphorylcholine antibody levels and mortality associated with pneumonia, Scand. J. Infect. Dis., 1990, 22:187.

74. Osborn, Munson, Separation of the inner (cytoplasmic) and outer membranes of gram negative bacteria, Methods Enzymol., 1974, 31A:642-653.

75. Paton, Pathogenesis of pneumococcal disease, 1993, 363-368.

i ai i i i 210 76. Paton, Lock, Lee, Li, J.P., Berry, Mitchell. Purification and immunogenicity of genetically obtained pneumolysin toxoids and their conjugation to Streptococcus pneumoniae type 19F polysaccharide, Infect. Immun.

1991, 59:2297-2304.

77. Paton, Lock, Hansman, Effect of immunization with pneumolysin on survival time of mice challenged with Streptococcus pneumoniae, Infect. Immun. 1983, 40:548-552.

78. Raven, Reciprocal capsular transformations of pneumococci, j. Bact. 1959, 77:296-309.

79. Rijn, MacLeod, McCarty, Studies on the chemical nature of the substance inducing transformation of pneumococcal types, Induction of transformation by a desoxyribonucleic acid fraction isolated from pneumococcus type III, J. Exp. Med.

1944, 79:137-158.

Riley, I.D. and R.M. Douglas, An epidemiologic approach to pneumococcal disease, Rev. Infect. Dis., 1981, 3:233.

81. Robbins, Austrian, Lee, Rastogi, Schiffman, Henrichsen, Makela,

P.H.,

Broome, Facklam, Tiesjema, Parke, Jr., Considerations for formulating the second-generation pneumococcal capsular polysaccharide vaccine with emphasis on the crossreactive types within groups, J. Infect Dis 1983, 148:1136-1159.

82. Roberts, Jeffery, Mitchell, Andrew, Boulnois, Feldman, Cole, P.J., Wilson, R. Effect of immunization with Freund's adjuvant and pneummolysin on histologic features of pneumococcal infection in the rat lung in vivo, S* Infect. Immun. 1992, 60:4969-4972.

83. Sampson, O'Connor, Stinson,

A.R.,

Tharpe, Russell, Cloning and nucleotide sequence analysis of psaA, the Streptococcus pneumoniae gene encoding a 37 -kilodalton protein homologous to previously reported Streptococcus sp.

adhesions, Infect. Immun. 1994, 62:319-24.

S 84. Scheewind, Fowler, Faull, Structure of cell wall anchor of cell surface proteins in Staphylococcus aureus, Science 1995, 268:103-106.

Schneewind Model Fischetti Sorting of protein A to the staphylococcal cell wall, Cell 1992, 70:267-81.

<ee e e e* *e 211 86. Schwartz, Gove, Lob-Lovit, Kirkwood, B.R. Potential interactions for the prevention of childhood pneumonia in developing countries: etiology of accute lower respiratory infections among young children in developing countries, Ped.

Infect. Dis., in Press.

87. Shapiro, Berg, Austrian, Schroeder Parcells, Margolis, Adair, R.K., Clemmens, J.D. Protective efficacy of polyvalent pneumococcal polysaccharide vaccine, N. Engl. Med., 1991, 325:1453-1460.

88. Sheffield, W.H. Benjamin and L.S. McDaniel.

Detection of DNA in Southern Blots by Chemiluminescence is a sensitive and rapid technique, Biotechniques 1992, 12:836-839.

89. Shoemaker, Guild, Destruction of low efficiency markers in a slow process occurring at a heteroduplex stage of transformation, Mol. Gen.

Genet. 1974, 128:283-290.

Siber, G.R. Pneumococcal disease: prospects for a new generation of vaccines, Science 1994, 265:1385- 1387.

91. Sihavy, M.L. Berman and L.W. Enquist, In: Experiments with gene fusions, (Cold Spring Harbor: Cold Springs Harbor Laboratory), 1984.

92. Smith, I. Krohn, G.T. Hermanson A.K. Malia, F.H. Gartner, M.D. Provenzano, E.K. Fujimoto,

N.M.

Goeke, B.J. Olson and D.C. Klenk, Measurement of Protein using Bicinchoninic Acid, Anal Biochem 1985, 150:76-85.

93. Smith, Guild, W.R. A plasmid in Streptococcus pneumoniae, J. Bacteriol. 1979, 137:735-739.

94. Southern, Detection of specific sequences among DNA fragments separated by gel electrophoresis, J. Mol. Biol. 1975, 98:503-517.

Stein, Thymus-independent and thymus-dependent responses to polysaccharide antigens, J. Infect. Dis. 1992, 162:S49.

S 96. Studier Moffatt Use of baceriophage T7 SRNA polymerase to direct selective high-level expression of cloned genes, J. Mol. Biol. 1986, 189: 113-30.' e D a 97. Talkington, Koenig, Russell, The 37 kDa a. ~protein of Streptococcus pneumoniae protects mice

I

212 against fatal challenge, American Society of Microbiology Abstracts 1992, 149.

98. Talkington, Voellinger, McDaniel,

L.S.,

Briles, Analysis of pneumococcal PspA microheterogeneity in SDS polyacrylamide gels and the association of PspA with the cell membrane, Microb. Pathog. 1992, 13:343-355.

99. Talkington, Crimmins, Voellinger,

D.C

Jother, Briles, A 4 3-kilodalton pneumococcal surface protein, PspA: isolation, protective abilities, and structural analysis of the amino-terminal sequence, Infect. Immun. 1991, 59:1285-1289.

100. Tart, McDaniel, Ralph, Briles, D.E. Truncated Streptocccus pneumoniae PspA molecules elicit cross-protective immunity against pneumococcal challenge in mice, J. Infect. Dis.

1995, In Press.

101. Tomasz, A. Surface components of Streptococcus pneumoniae. Rev. Infect. Dis. 1981, 3:190-211.

102. Tomasz, A. Biological consequences of the replacement of choline by ethanolamine in the cell wall of pneumococcus: chain formation loss of transformability, and loss of autolysis, Proc. Natl.

Acad. Sci. USA 1968, 59:86-93.

103. Towbin H, Staehelin T, Gordon Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications, PNAS 1979, 76:4350-4.

104. Tuomanen, Liu, Hengstler, Zak, O., Tomasz, The Induction of meningeal inflammation by components of the pneumococcal cell wall. 1985, 151, 859-868.

105. Tuomanen, Tomasz, Hengstler, Zak, O.

The relative role of bacterial cell wall and capsule in the induction of inflammation in pneumococcal meningitis, J. Infect. Dis. 1985, 151:535-540.

106. van de Rijn, Kessler, R.E. Growth characteristics of Group A Streptococci in a new ':chemically defined medium, Infec. Immun. 1980, 27:444-448.

107. Waltman, W.D. II, McDaniel, Gray, and Briles, Variation in the molecular weight of f, ~PspA (pneumococcal surface protein A) among Streptococcus pneumoniae, Microbial Pathogenesis 1990, 8:61-69.

213 108. Waltman, II, McDaniel, Andersson,

B.,

Bland, Gray, Svanborg-Eden, Briles, Protein serotyping of Streptococcus pneumoniae based on reactivity to six monoclonal antibodies, Microb. Pathog. 1988, 5:159-167.

109. Weigle, W.O. Immunological unresponsiveness In: Adv.

Immunol., (Ed. Dixon, Kunkel, Academic Press, New York, New York, 1973, 61-162.

110. Wicker L.S. and I. Scher, X-linked immune deficiency (Xid) of CBA/N mice, New York: Apringer-Verlag, 1986, 86.

111. Wright S.H. Sell, W.K. Vaughn, C. Andrews, K.B. McConnell and G. Schiffman, Clinical studies of pneumococcal vaccines in infants, II, Efficacy and effect on nasopharyngeal carriage, Rev. Infect. Dis.

1981, 3:S108.

112. Yother, McDaniel, Crain, Talkington, Briles, Pneumococcal surface protein

A:

Structural analysis and biological significance In: Genetics and Molecular Biology of Streptococci, Lactococci, and Enterococci, (ed. Dunny, G.M., Cleary, McKay, American Society for Microbiology, Washington, DC, 1991, 88-91.

113. Yother J. and D.E. Briles, Structural properties and evolutionary relationships of PspA, a surface protein of Streptococcus pneumoniae, as revealed by sequences analysis, J. Bact. 1992, 174:601-609.

114. Yother, McDaniel, and Briles,

D.E.,

Transformation of encapsulated Streptococcus pneumoniae, J. Bacteriol. 1986, 168:1463-1465.

115. Yother J, White JM. Novel surface attachment mechanism for the streptococcus pneumoniae protein PspA, J. Bact. 1994, 176:2976-85.

116. Yother, Handsome, Briles, D.E. Truncated Sforms of PspA that are secreted from Streptococcus pneumoniae and their use in functional studies and cloning of the pspA gene, J. Bact., 1992, 174, 610- 618.

117. Yother J, Forman C, Gray BM, Briles DE. Protection of mice from infection with Streptococcus pneumoniae by anti-phosphocholine antibody, Infect. Immun.

1982, 36:184-8.

118. Zar Biostatistical Analysis, 2nd Ed., Englewood Cliffs, NJ, Prentice-Hall, Inc., 1984, t718.

Claims (45)

1. An isolated amino acid molecule consisting of residues 1 to 115, 1 to 260, 192 to 588, 192 to 299, or residues 192 to 260 of pneumococcal surface protein A of Streptococcus pneumoniae.

2. An isolated DNA molecule consisting of a fragment of pneumococcal surface protein A gene of Streptococcus pneumonia encoding the isolated amino acid molecule of claim 1.

3. A PCR primer consisting essentially of the isolated DNA molecule of claim 2.

4. A hybridization probe consisting essentially of the isolated DNA molecule of claim 2. An immunological composition comprising the amino acid molecule of claim 1.

6. An isolated DNA molecule consisting of nucleotides 1 to 26, 1967 to 1990, 161 to 187, 1093 to 1117 or 1312 to 1331, or 1333 to 1355 of a pneumococcal surface protein A gene of Streptococcus pneumoniae.

7. A PCR primer consisting essentially of the isolated DNA molecule of claim 6.

8. A hybridization probe consisting essentially of the isolated DNA molecule of claim 6.

9. An isolated DNA molecule consisting of a fragment of a pneumococcal surface protein A gene of Steptococcus pneumoniae consisting of a nucleotide sequence to selected from CCGGATCCAGCTCCTGCACCAAAAAC .GCGCGTCGACGGCTTAAACCCATTCACCATTGG; CCGGATCCTGAGCCAGAGCAGTTGGCTG; CCGGATCCGCTCAAAGAGATTGATGAGTCTG; GCGGATCCCGTAGCCAGTCAGTCTAAAGCTG SCTGAGTCGACTGGAGTTTCTGGAGCTGGAGC CCGGATCCAGCTCCAGCTCCAGAAACTCCAG; GCGGATCCTTGACCAATATTTACAGGAGGC;AGG GTTTTTGGTGCAGGAGCTGG; S* GCTATGGGCTACAGGTTG; CCACCTGTAGCCATAGC; CCGCATCCAGCGTGCCTATCTTAGGGGCTGGTT; a GCAAGCTTATGATATAGAAATTTGTAAC. 215 A PCR primer consisting essentially of at least one isolated DNA molecule of claim 9.

11. A hybridization probe consisting essentially of at least one isolated DNA molecule of claim 9.

12. PCR probe(s) which distinguishes between pspA and pspA-like nucleotide sequences.

13. PCR probe(s) which hybridizes to both pspA and pspA-like nucleotide sequences.

14. A PspA extract prepared by a process comprising growing pneumococci in a first medium containing choline chloride, eluting live pneumococci with a choline chloride containing salt solution, and growing the pneumococci in a second medium containing an alkanolamine and substantially no choline. A PspA extract prepared by growing pneumococci in a first medium containing choline chloride, eluting live pneumococci with a choline chloride containing salt solution, growing the pneumococci in a second medium containing an alkanolamine and substantially no choline, and purifying PspA by isolation on a choline-Sepharose Saffinity column.

16. An immunological composition comprising the extract of claim 14. 216

17. An immunological composition comprising the extract of claim

18. An immunological composition comprising full length PspA.

19. A method for enhancing immunogenicity of a PspA-containing immunological composition comprising including in said composition the C-terminal portion of PspA. An immunological composition comprising at least two PspAs.

21. The immunological composition of claim wherein the PspAs are from different groups based on RFLP.

22. PCR amplification product from a primer as claimed in claims 3, 7, 10, 12 or 13.

23. An isolated DNA molecule consisting of a nucleotide sequence homologous to a portion of pspA.

24. An isolated amino acid molecule comprising pneumococcal surface protein C, PspC, of Streptococcus pneumoniae having alpha-helical, proline rich and repeat regions.

25. An isolated DNA molecule comprising a *pneumococcal surface protein C gene of S. pneumoniae encoding the isolated amino acid molecule of claim 24.

26. A PCR primer consisting essentially of the isolated DNA molecule of claim

27. A hybridization probe consisting essentially of the isolated DNA molecule of claim 217

28. An immunological composition comprising the amino acid molecule of,claim 24.

29. An isolated amino acid molecule of claim 24 having strong homology with pneumococcal surface protein A, PspA, of S. pneumoniae from amino acid 458 of PspC, corresponding to amino acid 147 of PspA, extending to a C-terminus of PspC and PspA. An isolated amino acid molecule of claim 24, further comprising a signal sequence consisting essentially of a charged region followed by a hydrophobic core of amino acids.

31. An isolated amino acid molecule of claim 24, wherein the alpha-helical region further comprises a seven residue periodicity and a coiled coil region having three breaks in a heptad repeat.

32. An isolated amino acid molecule comprising pneumococcal surface protein C, Pspc, of S. pneumoniae having alpha-helical, proline rich and repeat regions, wherein the alpha-helical region comprises a C-terminus having substantial homology with a protection-eliciting region of PspA.

33. An isolated DNA molecule comprising a S pneumococcal surface protein C gene of S. pneumoniae encoding the isolated amino acid molecule of claim 32. :34. A PCR primer consisting essentially of the isolated DNA molecule of claim 33. A hybridization probe consisting essentially of the isolated DNA molecule of claim 33. 4 1 1 218

36. An immunological composition comprising the amino acid molecule of claim 32.

37. An isolated amino acid molecule of claim 24, further comprising a 17 amino acid, partially hydrophobic tail.

38. An isolated amino acid molecule of claim 32, further comprising a 17 amino acid, partially hydrophobic tail.

39. An isolated amino acid molecule of claim 24, further comprising an epitope of interest. An isolated amino acid molecule of claim 32, further comprising an epitope of o1 interest.

41. An immunological composition comprising the amino acid molecule of claim 39.

42. An immunological composition comprising the amino acid molecule of claim

43. An isolated amino acid molecule, substantially as hereinbefore described with reference to any one of the Examples.

44. An isolated DNA molecule PCR primer or hybridization probe, substantially as hereinbefore described with reference to any one of the Examples. A PspA extract, substantially as hereinbefore described with reference to any one of the Examples.

46. A method for enhancing immunogenicity of a PspA-containing immunological composition, substantially as hereinbefore described with reference to any one of the Examples.

47. An immunological composition, substantially as hereinbefore described with reference to any one of the Examples.

48. A method of immunising a patient against infection with Streptococcus pneumoniae, comprising treating said patient with an isolated amino acid according to any one of claims 1, 14, 15, 24, 29 to 32, 37 to 40, 43 or 45 or an immunological composition according to any one of claims 5, 16 to 18, 20, 21, 36, 41, 42 or 47. 30 49. An isolated amino acid according to any one of claims 1, 14, 15, 24, 29 to 32, 37 to 40, 43 or 45 or an immunological composition according to any one of claims 5, 16 to 18, 20, 21, 36, 41, 42 or 47, when used to immunise a patient against infection with Streptococcus pneumoniae.

50. Use of an isolated amino acid according to any one of claims 1, 14, 15, 24, 29 to 32, 37 to 40, 43 or 45 or an immunological composition according to any one of claims 16 to 18, 20, 21, 36, 41, 42 or 47, in the manufacture of a-medicament for immunising Sa patient against infection with Streptococcus pneumoniae.

51. A method, isolated amino acid or use according to any one of claims 48 to substantially as hereinbefore described. 4ep [n:\libc]03437:ssd 219

52. A method of diagnosing the presence of Streptococcus pneumoniae comprising the use of a detection agent selected from: a DNA molecule, PCR primer or hybridization probe according to any one of claims 2 to 13, 22 23, 25 to 27, 33 to 35 or 44; an isolated amino acid according to any one of claims 1, 14, 15, 24, 29 to 32, 37 to 40, 43 or 45; or an immunological composition according to any one of claims 5, 16 to 18, 20, 21, 36, 41, 42 or 47.

53. A detection agent selected from: a DNA molecule, PCR primer or hybridization probe according to any one of claims 2 to 13, 22 23, 25 to 27, 33 to 35 or 44; an isolated amino acid according to any one of claims 1, 14, 15, 24, 29 to 32, 37 to lo 40, 43 or 45; or an immunological composition according to any one of claims 5, 16 to 18, 20, 21, 36, 41, 42 or 47; when used for diagnosing the presence of Streptococcus pneumoniae.

54. Use of a detection agent selected from: a DNA molecule, PCR primer or hybridization probe according to any one of claims 2 to 13, 22 23, 25 to 27, 33 to 35; an isolated amino acid according to any one of claims 1, 14, 15, 24, 29 to 32, 37 to 40, 43 or or an immunological composition according to any one of claims 5, 16 to 18, 20, 21, 36, 41, 42 or 47; in the manufacture of a composition for diagnosing the presence of Streptococcus pneumoniae. A method, detection agent or use according to any one of claims 52 to 54, wherein the diagnosis is that of infection of a patient with Streptococcus pneumoniae.

56. A method, detection agent or use according to any one of claims 52 to 54, substantially as hereinbefore described. Dated 22 January, 1999 UAB Research Foundation Patent Attorneys for the Applicant/Nominated Person fto. SPRUSON FERGUSON b0 9e