CA2286301A1 - The filamentous phage as a multivalent scaffold for coupling synthetic peptides for immunization - Google Patents

Abstract

The filamentous bacteriophage is a common immunogenic carrier for generating anti-peptide antibodies against recombinant peptides displayed on its surface. For immunizations, peptides are displayed as fusions to either the minor coat protein, pIII, or, more commonly the major coat protein, pVIII. Phage displaying a poorly expressed malarial peptide on pVIII
were compared to phage bearing the same peptide on pIII , and to phage that were chemically coupled to a synthetic version of the peptide using the crosslinker, sulfo-succinimidyl 4-(N-maleimido-methyl)cyclohexane-1-carboxylate (sulfo-SMCC) for their ability to induce an anti-peptide Ab response in mice. We also included in this comparison the same peptide displayed as a recombinant fusion to the maltose-binding protein (MBP) of Escherichia coli, as well as the synthetic counterpart as chemically crosslinked to MBP with sulfo-SMCC. The results demonstrate that only the chemically-crosslinked conjugates elicited a strong anti-peptide antibody response by ELISA, probably due to the higher number of peptides that were displayed by the phage and MBP conjugates as compared to their recombinant counterparts.
The amount of peptide that was covalently linked to the phage was increased approximately two-fold when a lysine residue was engineered near the N-terminus of mature pVIII. The ease of chemically conjugating synthetic peptides to phage, and of modifying pVIII by genetic engineering, suggests new roles for the filamentous phage in vaccine studies.

Description

The Filamentous Phage as a Multivalent Scaffold for Coupling Synthetic Peptides for Immunization Michael B. Zwick' and Jamie K. Scott'"
' Institute of Molecular Biology and Biochemistry, Simon Fraser University, 8888 University Drive, Burnaby, B.C., Canada, VSA 156. Tel: (604) 291-5658, Fax: (604) 291-5583, email:
[email protected] "Corresponding author Aug. 19, 1999 J .1 . . . ,.
Introduction The filamentous bacteriophage is a long flexible rod, approximately 1 um in length and 6.5 nm in diameter. A circular, single strand of DNA in the interior of the virus is encapsulated by five phage-encoded proteins. Four minor coat proteins are found in pairs at either end of the phage at five copies each (i.e., pIII and pVI at one end, and pVII and pIX at the opposite end). In particular, pIII is a 406 amino acid protein that projects as a knob-like structure from one tip of the phage. The bulk of the phage surface consists of the major coat protein (pVIII), an alpha-helical protein that coats the length of the phage in a shingled, or fish-scale-like pattern (Glucksman et al., 1992; Marvin et al., 1994). ??Figure of pha,~e showing recombinant fusion proteins??
Many groups have used the filamentous bacteriophage as an immunological carrier for the purpose of eliciting anti-peptide antibody (Ab) responses in animals. In all of these systems, the peptide is fused by recombinant methods to a coat protein and is thus displayed on the surface of the phage. The first to use a recombinant peptide displayed by the phage as an immunogen were de la Cruz et al. (1988 j. These authors used pIII-display in which only about 5 copies of peptide are displayed per phage. They found that a potent anti-phage Ab response was induced, but that only marginal titers of Ab were generated against the peptide. Soon thereafter, pVIII-display was pioneered for immunogenic purposes (Greenwood et al., 1991), and, because pVIII offers the possibility of having many more copies of recombinant peptide per phage, it was found that phage displaying recombinant peptides on pVIII could elicit a strong anti-peptide Ab 'a ' response in the absence of adjuvant. Further work by the same group (Willis et al., 1993) also showed that the Ab responses against these phage were T-cell dependent. At the same time, Minenkova et al. (1993) immunized rabbits with phage bearing a peptide from the HIV-1 p17 Gag protein on pVIII and found that Ab was elicited which recognized the target antigen.
Subsequently, there have been a multitude of studies in which random peptide libraries (Scott &
Smith, 1990) are screened with Abs that are specific for a target antigen, and the phage clones thus selected are then used to immunize animals (for examples see Motti et al., 1994; Zhong et al., 1994; Demangel et al., 1996; for reviews see Smith & Petrenko, 1997;
Zwick et al., 1998a).
The goal in these experiments is usually to elicit cross-reactive Abs that bind both the peptide-bearing phage clone and the target antigen.
For pVIII-display, a hybrid system is employed, in which the peptide:pVIII
fusion is supplemented with wild-type pVIII, either in traps, as in the "type 8+8"
system, or in cis, as in the "type 88" system used in this study (see Smith & Petrenko, 1997, for types of phage display systems). The resulting hybrid virion bears both wild type and the peptide:pVIII fusion on its surface. For some peptides, as much as 30-40% of the total pVIII on the phage coat may be recombinant, whereas for others only a few copies of peptide may be detected per phage (Malik et al., 1996/1998). We wanted to test the anti-peptide Ab response elicited in mice by phage bearing a pVIII-displayed peptide, designated NANP, corresponding to the major antigenic repeat of the malarial circumsporozoite protein. We used as a vector, f88-4 (Zhong et al., 1994), since our panel of random peptide libraries (Bonnycastle et al., 1996) are displayed by this phage vector. Before immunizing, we measured the copy number of the NANP peptide on the phage coat and found that it was especially low, even if the phage were amplified while inducing the ..- , _, ;i tac promoter of recombinant gene 8 (~0.4 % of wild type pVIII). It was later determined that the peptide copy-number was low in general for the f88-4 system (~0.4 - S % of wild type pVIII).
This makes such phage good for affinity-selection experiments, since low densities of recombinant peptide would allow for the selection of high affinity clones from a library.
However, we assumed that the low peptide copy number would make the phage a poor immunogen.
To solve the problem of low copy number of peptide on phage, we had the idea of using the phage, which have a repetitive multivalent array of pVIIIs, as a classical immunological earner to which many synthetic peptides could be coupled -- probably more so than for recombinant display on pVIII. We ofren desire to optimize, characterize and measure the affinity of synthetic peptide analogs of peptides derived from our libraries, as is commonly practiced, because such peptides are free of the confounding effects of multivalency and possible contributions from the phage coat to binding. Naturally, we were aware that conjugation may affect the affinity of the Ab for the peptide. It is also true, however, that at least in some cases, immobilization of a peptide at one end can induce a persistence in its overall structure that does not exist when the peptide is free in solution, as was shown by Jelinek et al.
(1997) for an HIV-derived peptide displayed on the phage coat. Given the advantages of free peptide model systems and the problem of low copy number of recombinant peptide, we decided to use the phage as a classical protein earner for coupling with synthetic peptide, and to compare this immunogen to phage bearing recombinant peptide on either pIII or pVIII.
There are several advantages of using the phage over other classical carrier proteins (e.g., tetanus toxoid, keyhole limpet hem.ocvanin, and ovalbumin). The phage are ''antigenically-S

,. .....,., . ,s homogeneous" in the sense that with many, many repeating copies of the same, small molecular-weight protein (i.e., pVIII), they will likely induce a restricted Ab response. In support of this, it was recently shown that the preponderance of pVIII-reactive Ab raised against the phage recognize only about the first 12 amino acids on pVIII (Kneissel et al., 1999). Another advantage of the phage is that the repetitive arrangement of pVIII on the phage coat would allow for multivalent display of synthetic peptides in the same context, as opposed to a large, complex protein that would display each peptide differently on various portions of the protein surface (the same advantage exists for recombinant peptide-display on pVIII). Other advantages of the phage are that they are easily produced and purified. They are easy to engineer by recombinant methods, so the "immunogenic landscape" of the phage can be tailored for a particular purpose.
Finally, the large size of the phage may lend to it adjuvant-like effects, and better approximates the size and composition of pathogens (especially viruses) than do the more commonly used protein carriers. Recently, the utility of the phage as an effective immunologic carrier was shown by Meola et al. (1995), who found that recombinant peptide displayed on pVIII
was the best mode of immunization when it was compared to the same peptide displayed on pIII, recombinant human H-ferritin, the hepatitis B virus core peptide, and as a synthetic multiple antigenic peptide (M.AP; Tam, 1988). The results of M~ola et al. (1995), and others clearly demonstrate that the phage is an excellent recombinant carrier for peptide immunizations. For all of the above reasons, we hypothesized that the phage would be an excellent choice as a classical immunological carrier.
On the other hand, there may be drawbacks of using the phage as a classical protein carrier, and thus other modes of peptide display may be preferable. For example, because of the predominance of the small sized pVIII on the phage coat, there may be a limited T-cell response against the phage. With only 50 residues, the number of good MHC class II
epitopes in pVIII is probably low, although the studies of Willis et al. (1993) clearly indicate that the immune response against bacteriophage, at least in BALB/c mice, is T-cell dependent.
Although too numerous to mention here, there are many different vehicles for peptide immunizations involving both recombinant and synthetic peptides. For recombinant display of peptide, some have used various proteins as carriers for a fused peptide. We chose to assess whether peptides displayed on monomeric maltose-binding protein (MBP; Zwick et al., 1998b) would serve as a suitable immunogen, as has been previously demonstrated for a peptide from the hepatitis B virus (Martineau et al., 1991/1992). In the case of synthetic peptide-carrier conjugates, a protein carrier is typically crosslinked to a large molar excess of peptide, thus creating a multivalent immunogen. We also included as an immunogen in this study the NANP peptide conjugated to MBP for comparison to its phage counterparts.
In order to couple peptides specifically to MBP and the phage, the water-soluble, heterobifunctional crosslinker sulfo-succinimidyl 4-(N-maleimidomethyl)cyclohexane-1-carboxylate (sulfo-SMCC) was used. There is a maleimide moiety on one end of sulfo~SMCC that reacts with free sulflrydryls to form a covalent bond, and a succinimidyl group on the opposite end that reacts with primary amines also to form a covalent linkage. Thus, any peptide containing a free sulfhydryl, in theory, can be coupled to every copy of pVIII that contains an exposed lysine residue, and the peptide be displayed in only one orientation. Accordingly, if a lysine residue were exposed on every copy of pVIII, thousands of nucleophilic, amine functionalities would be available for the covalent ligation of peptidic, or p. ,w o . . ~ . ..-.
'-.
haptenic molecules to produce a highly polyvalent immunogen. The first lysine residue in pVIII
appears at position 8, which is partially exposed on the phage surface (Armstrong & Perham 1983), but resides in a region that has alpha-helical structure so there are restrictions in mobility at this position, as opposed to residues 1 to S which are more disordered (Glucksman et al., 1992). As Lys8 may not be sufficiently exposed for efficient chemical coupling to bulky peptides, we decided to engineer a second lysine residue closer to the N-terminus of pVIII on fl phage. The resulting phage, designated fl-K, was compared to wild type phage for efficiency of peptide coupling, and both conjugates were used in immunization experiments.
Thus, in this study, we decided to test the immunogenicity of the synthetic version of the malarial peptide, using two different variants of the filamentous bacteriophage as carriers, and compare it to the immunogenicity of the recombinant counterpart fused to either pIII or pVIII. In addition, the same peptide was displayed as a recombinant fusion to MBP (Zwick et al., 1998b), and the synthetic version of the peptide was also chemically conjugated to MBP, in order to compare these MBP conjugates with their phage counterparts, and determine the best immunogen for eliciting anti-peptide Abs. We found that for a 24-mer peptide containing the major repeat sequence of the malarial circumsporozoite protein, the optimal anti-peptide response was achieved by challenpng mice with synthetic peptide-carrier conjugates rather than with recombinant display on either pllI, pVIII, or MBP. Both phage and MBP
appeared to be equally effective as carriers for synthetic peptide as the anti-peptide Ab response was similar in these conjugates. The use of chemical conjugation for peptide display on phage increases the possibilities for the phage as a carrier for immunizations, and could present new options for vaccine design in the future.

Abbreviations Ab, antibody; sulfo-SMCC, sulfo-succinimidyl 4-(N-maleimidomethyl)cyclohexane-1-carboxylate; MAP, multiple antigenic peptide; MBP, maltose-binding protein; RF, replicative form; PEG, polyethylene glycol; PBS, phosphate buffered saline; TBS, Tris buffered saline; SDS-PAGE, solium dodecyl sulfate;
ELISA, enzyme-linked immunosorbent assay; ABTS, 2 2'-azino-bis(3-ethylbenzthiazoline-6-sulfonic acid; r, recombinant; s, synthetic; N, NANP peptide, NANPNVDP(NANP)3;
Materials and methods Materials The oligonucleotide Primer 1 has the sequence 5'-GCCGCTTTTGCGGGATC GTCCGAAGCT
TTNGMACCCTCAGCA GCGAAAGAC-3' (where N = A, C, G, T and M = C or A; GibcoBRL
Life Technolo;~ies. Inc., Gaithersbure. MD). The E. coli strain, K91, the f88-4 (Zhong et al., 1994) and the fUSEl7 (pIII-displayed NANP) vectors were all kindly supplied by G.P. Smith (University of Missouri-Columbia). ?? supplied the fl vector (Hill & Petersen, 1983).
The crosslinker, sulfo-SMCC was purchased from Pierce (Pierce, Rockford, IL).
The i, ~. ~..
monoclonal Ab Pf2A10 (Wirtz et al., 1987) was a gift from R. Wirtz (WRAIR, Washington, DC;
SmithKline Beecham and New York University). Female BALB/c mice (6 weeks old) were purchased from Charles River (St. Constant, Quebec).
fl -K Vector Construction fl phage (Hill & Petersen, 1982) was subjected to site-directed mutagenesis (Sambrook et al., 1989) using Primer 1: This oligonucleotide contains a codon with one degenerate, and one partially-degenerate position that encodes either a serine or an alanine residue. This partial degeneracy was included in the design of the primer in case of the unlikely event that if one of the residues prevented phage production, the other would be selected. Thus, 300 ng purified, fl ssDNA was mixed with 35 ng Primer 1 in 10 uL 10 mM Tris-HCl (pH 7.5), 2 mM
MgCl2 and 50 mM NaCI. The annealing mixture was heated at 70°C for 10 min, allowed to cool to room temperature over 45 min, and put on ice for 10 min. The annealing mixture was mixed, on ice, with 10 uL reaction mix containing 20 mM Tris-HCl (pH 7.5), 10 mM MgCl2, 1 mM
dithiothreitol, 1 X bovine serum albumin, 1 mM ATP, 0.25 mM dNTP, 4 units T7 DNA
polymerase (New England Biolabs Ltd. (NEB), Mississauga, ON), and 4 units T4 DNA ligase (NEB). The mixture was incubated for 5 min on ice, 5 min at room temperature, and then 60 min at 37 °C. The fill-in reaction was stopped by heating the mixture to 70 ° C for 10 min. The fill-in products were separated on a 0.8% agarose gel in 4X GBB buffer (168 mM Tris, 80 mM sodium acetate, 7.2 mM Na2EDTA and pH adjusted to 8.3 with glacial acetic acid). The presence of a band, approximately 6.4 kb in size, indicated the presence of replicative form (RF) circularized, fl dsDNA. An aliquot (1 uL) of the fill-in product was used to transform 100 uL of Escherichia coli (K91) cells following the heat shock method described in Sambrook et al.
(1989). The transformed cells were mixed with top agar (-, and maintained at 50°C), and quickly poured onto agar plates containing NZY medium. Single colonies were picked, amplified, and the RF
DNA was isolated and tested for the presence of the desired mutation by digestion with the restriction endonuclease HindIII (NEB) following the manufacturer's instructions. A single clone was selected that contained the HindIII site, and the correct sequence was verified by DNA
sequence analysis (Bonnycastle, 1998).
Large-scale preparation offl-Kphage Three liters of NZY medium (Smith & Scott, 1993) were inoculated with a single, well isolated colony of K91 cells infected with fl-K phage, ~~d the culture was shaken at 250 rpm for 21 h at 37°C. The culture was centrifuged at 4500 x g (rma,~ for 10 min at 4°C , and the supernatant spun again at 11 000 x g for 10 min at 4°C. The supernatant containing the phage was mixed with 0.1 S vol of PEG/IVaCI ( 16.7% polyethylene glycol (PEG 8000, Sigma, St. Louis, MO), 3.3 M NaCI) by inverting 100 times, and left to precipitate overnight at 4°C. The phage were centrifuged at 11 000 x g for -10 min at 4 °C, the supernatant removed, and the pellet was resupended in 100 mL PBS (23 m_'~f I~zaHzP04, 77 mM Na,HP04, 1 SO mM NaCI, pH
7.4). The resuspended phage were cleared by centrifugation at 20 000 x g for 10 min at 4°C, and then ~ ,, .. , .J
precipitated with PEG/NaCI as described. The phage were resuspended in 50 mL
PBS, cleared again, then solid CsCI (C-3139, Sigma) was added to a concentration of 31%
(w/v). The phage mixture was transferred to two 39-mL quick-seal polyallomer tubes (#342414, Beckman Instruments Inc., Palo Alto, CA), and centrifuged at 57 000 rpm in a 70Ti rotor on an ultracentrifuge (L8-80, Beckman) for 21 h at 5°C. A large, diffuse, bluish band containing the phage was removed from the tubes using a 18-gauge needle equipped with a 10-mL
syringe, and distributed between two 32-mL polycarbonate tubes (#355631, Beckman). The tubes were filled to the shoulder with PBS, and spun at 50 000 rpm for 4 h at 4°C in a 70Ti rotor. The phage pellets were resuspended in a total of 28 mL PBS, heated to 70°C for 30 min, and cleared by spinning at 20 000 x g for 20 min at 4 ° C. The supernatant was transferred to a fresh tube, and the phage concentration, determined by absorbance measurement,was 2.8 x 10'4 phage particles/mL (Day, 1969). The phage stock was checked by DNA gel electrophoresis, and a single band was observed. The sequence in the region of the mutagenesis was reconfirmed by DNA sequence analysis.
Production of f88, rN.~pIII, and r14'.pYlll phage The recombinant, NANP peptide-bearing phage (rN:pIII and rN:pVIII phage), and the parent vector f88-4 (Zhong et al., 1994; also referred to as f88 in the text), were amplified and purified by a method similar to that used for fl-K prage, except that, during cell growth the culture medium contained tetracycline (15 ug/uL) to sei_ect for E. coli cells infected with phage, which harbor tetracycline resistance genes.
Coupling synthetic peptide to fl-K and f88-4 phage About 2.5 x 10'3 physical particles of fl-K or f88-4 phage were mixed with 0.6 mg of sulfo-SMCC crosslinker (Pierce) in a total vc.lunle of 500 uL PBS. The final concentration of phage was 5 x 10'3 particles/mL, and was chosen because, at higher concentrations of phage (i.e., 101'' particles/mL or greater), the suspension becomes gel-like, and the addition of crosslinker causes significant precipitation. At ~ x 10'3 particles/mL, there was still a small amount of precipitate after mixing with sulfo-SMCC; however, samples more dilute than this were avoided since lower coupling efficiencies occur under these conditions. The mixtures of phage and sulfo-SMCC were rotated slowly at 37°C for 45 min, brought to 1 mL with PBS, and precipitated by mixing with 0.15 vol of PEG/NaCI and incubating on ice nor 15 min. The samples were spun at 15 000 x g for 15 min at 4°C, and all traces of the supernatants were removed. The pellets were resuspended in 1 mL PBS, and the phage were precipitated a second time with PEG/NaCI as before. The SMCC-activated phage w ere considerably more susceptible to PEG-precipitation than untreated phage, since the phage solutions immediately turned cloudy upon addition of PEG/h,TaCl, hence, extensive incubations at 4°C, and lor_g centrifugation times were unnecessary.
The activated phage were resuspended in 1 mL PBS, and mixed with 1.35 mg NANP
peptide (HZN-IVANPNVDPNANPNANPI~ ~-DD-Orn{bio}-C-CONHz; where HZN- represents a free N-:, >, . , . , terminus, -DD- was added as a spacer and to increase solubility, Orn{bio}
represents a biotinylated ornithine residue, and -CONH2 indicates that the C-terminus is blocked by amidation;
Alberta Peptide Institute, Edmonton, AB). At 2700 pVIII molecules per phage particle and a total of 2.5 x 10'3 phage, there are 112 nmol pVIII available for coupling.
Thus, 1.35 mg (S00 nmol) of the NANP peptide (mol. wt. 2701 g/mol) produces ~4.5 times molar excess of peptide over pVIII molecules. The mixtures of peptide and activated phage were rotated slowly overnight at 4°C, and the samples were purified twice by PEG-precipitation (as above) to remove free peptide. The final pellets, containing the synthetic NANP peptide-phage conjugates (sN-f1K and sN-f88), were resuspended in 900 uL PBS, and stored frozen at -20°C. Aliquots were removed and run on a 0.8% agarose gel in 4X GBB buffer, and DNA from the peptide-phage conjugates appeared as single bands that were identical in mobility to those of untreated phage.
Production of recombinant NANP peptide: ~LIBP fusion protein (rN.~MBP), and preparation of .rynthetic NANP peptide-MBP conjugate (sAr MBP) The rN:MBP fusion protein. as well as control MBP (i. e., recombinant. MBP
fused to an unrelated peptide sequence, YDVPD~-A, and purified under the same conditions as the rN:~P), were both produced from t~. pMal-X v~c.-~;ter iii E. coli, and purified as described previously (Zwick et al., 1998b). To prepare the synthetic sN-MBP, 2 mg purified control MBP
was mixed with 2 mg sulfo-SMCC in 300 uL PBS and rocked slowly for 40 min at 37°C. The sample was diluted to 2 mL with PBS, transferred to a Centricon-30 ultrafiltration device (Amicon, Inc., Beverly MA), and then washed 4 times in PBS at 4°C
following the manufacturer's instructions. The SMCC-activated MBP was divided, and approximately 1 mg was mixed with 1.35 mg synthetic NANP peptide in 185 uL PBS containing 5 mM
EDTA, whereas 0.8 mg SMCC-activated MBP was taken through the conjugation steps in 150 uL PBS
containing 5 mM EDTA, in the absence of peptide. The samples were rocked at 4°C for 2 h 15 min to allow coupling, and washed in PBS four times as befi;re. The samples were back-eluted from the centricon device in a final volume of 300 uL and were stored frozen at -20°C.
SDS-PAGE
Phage proteins were separated u~ ~ng a mod~riec~ SDS-PAGE system (pers. comm., R.N.
Perham) based on that of Schagger & von Jagow (1987). The stacking gel consisted of 16.6%
acrylamide, 0.17% bis, 19% glycerol, 1 M Tris-HCl (pH 8.3), and the separating gel consisted of 4.7% acrylamide, 0.048% bis, 0.72 M Tris-HCl (pH 8.3). The upper (cathode) chamber was almost completely filled with buffer containing 0.1 M Tris, 0.1 M Tricine (no pH adjustment necessary; cat#T-7911, Sigma, St. Louis, MO), and 0.1% SDS. The buffer in the lower (anode) chamber was 0.2 M Tris-HCl (pH 8.9~_ In pr~.parati~n for loading, phage samples were diluted in PBS and mixed with 1/3 volumes .1X gei loading buffer (8% SDS, 40% glycerol, 68 mM Tris-HCI, pH 6.8, 0.008% bromophenol bhne). All samples were boiled for 5 min immediately before loading. The MBP proteins were run on a 12% acrylamide gel according to the method of Sambrook et al. (1989). SDS-PAGs containing MBP samples were stained with Coomassie blue following the procedure of Sambrook et al. (1989), whereas the SDS-PAGs containing phage samples were silver-stained essentially following the method of Morrisey ( 1981 ).
Mouse immunizations The mice were immunized as summarized in Table ~. All injections were done intraperitoneally (i.p.), and diluted in a total of 100 uL PBS containing 1 mg/mL AdjuvaxTM
adjuvant (Alpha-Beta Technology, Worcester, MA). The mice were bled from the tail vein on Days 0, 14, and 28, just prior to injection. On Day 42, the mice were bled by cardiac puncture under COZ anaesthesia.
The blood samples were allowed to clot overnight at 4°C, and then centrifuged at 12 000 x g for min. The serum supernatants were trmsfei:ed to fresh microfuge tubes, and, for the tail-bleed samples, mixed with an equal volume of Tris buffered saline (TBS; 100 mM Tris-HCI, pH 7.5, 150 mM NaCI) containing 2% BSA and 0.04% sodium azide. For the serum samples from the final bleeds, sodium azide (Sigma) was added to a final concentration of 0.02%
(w/v). All of the serum samples were heated to 52 °C in a water bath for 20 min to inactivate complement, and stored at 4°C. The mice sera from within each group were pooled before use in the ELISA
experiments.
ELISA and KinExA

Microwells (EasyWash, Costar, Corning Inc., Corning, NY) were coated with 35 uL TBS
containing either 5 x 109 f88 particles, 200 ng MBP, or 1 ug streptavidin (Roche Diagnostics, Laval, Quebec), and gently rocked overnight at 4°C. For all the washing steps, 200 uL TBS
containing 0.1% Tween 20 (ICN Biomedicals Inc., Aurora, OH) were dispensed by an EL-403 plate washer (Bio-Tek Instruments, Inc., Winooski, VT); the plate was shaken for 5 sec on low setting, and set to stand S sec between each wash. The wells were washed twice, and 35 uL TBS
containing 10'z molecules of biotinylated NANP peptide was added to the aspirated wells containing immobilized streptavidin; all other wells received 200 uL TBS
containing 2% BSA
(Fraction V, Sigma). After incubating the microplate for 30 min at room temperature, wells containing peptide were topped with 200 uL TBS containing 2% BSA, and the plate was blocked for 1 h at 37°C. The wells were washed twice, and 35 uL pooled serum from each group (Table 1) diluted in TBS containing 1% BSA and 0.1% Tween 20 were added. The Abs were incubated in the appropriate wells for 2 h at room temperature. The wells were washed five times, and 35 uL goat anti-mouse IgG (H+L):horseradish-peroxidase conjugate (Pierce), diluted 1:500 in TBS
containing 0.1% Tween 20, was added. ABTS solution was prepared by mixing 3.07 ml 0.1 M
citric acid and 1.93 ml 0.2 M NazHPO~, and adding 2 mg of 2 2'-azino-bis(3-ethylbenzthiazoline-6-sulfonic acid) (Sigma) and 5 uL 30°/'~ (w/wi IIZOZ (BDH Inc., Toronto, ON). After 30 min at room temperature, the wells were «-as~ed six times and 35 uL freshly-prepared ABTS solution was added. After 45 min, the absorbance at 405 nm and 450 nm was measured on an EL 312e Bio-Kinetics Reader (Bio-Tek), and the results were reported as A4p5-A~9o.

The KinExA immunoassay is an in-solution assay for Kd determination, and has been described in detail by Blake et al. (1997). The procedure used in this study was essentially that of Craig et al. ( 1998), with the modification that the secondary (detection) Ab was Cy5-conjugated goat anti-mouse F(ab')Z (Jackson Immunoresearch Laboratories Inc., West Grove, PA). The assay was conducted using the KinExA instrument (Sapidyne Instruments, Inc., Boise, ID), and the data was analyzed using the software provided by the manufacturer (Sapidyne).
Amino acid analysis All amino acid analyses were performzd at the Alberta Peptide Institute (U.
Alberta, Edmonton, AB).
Results Characterization of fl -K phage, and immunogen conjugates The nucleotide and amino acid sequences of the N-terminal region of mature pVIII, for both wild-type fl and the phage bearii~g the extra Lys residue near the N-temlinus of pVIII (fl-K) are shown in Fig. ~ . The insertion. of four residues into the N-terminal segment of pVIII was well tolerated by fl -K phage, since the phage yield was very high (~2 x 10' ' phage/L), and no '..
mutations were discovered by DNA sequencing in the insert region following large-scale amplification of the phage. Phage that were coupled to the NANP peptide with sulfo-SMCC, or treated with SMCC alone, were subjected to DNA-gel electrophoresis, and the DNA from the chemically-modified phage appeared as a single band that was identical in mobility to that of untreated phage, indicating that the ssDNA of the phage was not modified by the crosslinking (data not shown). When the same samples were subjected to SDS-PAGE, the effect of the

2~
crosslinker could be seen (Fig.. The appearance of multiple bands with a slower mobility than pVIII, indicated that the NANP peptide was successfully coupled to the major coat protein of fl-K. It appeared that the side-chain of Lys8 in the mature pVIII of f88-4 was su~ciently exposed to react with the succinimidyl moiety of sulfo-SMCC (the amino acid sequence of mature pVIII
for f88 is identical to fl in the region shown in Fig.l), or, that the less-reactive amino terminus of pVIII is participating. In the case of fl-K treated with SMCC alone, some crosslinking of pVIII
can be seen, which resulted in a ladder-like effect. Although sulfo-SMCC-activated phage are supposed to couple specifically to sulfliydryls, in their absence, the maleimide moiety of sulfo-SMCC is known to react with free primary amines at a 1000-fold lower rate than with sulfhydryls (Hermanson, 1996). Thus, for sulfo-SMCC-treated fl-K, the free E-amino group of the lysine residue near the N-terminus of pVIII likely crosslinked to itself in the absence of a Cys-containing peptide. The ladderin~ effect observed with SMCC-treated fl-K
does not occur with the f88 counterpart, although some dimer can be seen, indicating there is some adventitious crosslinking occurring as well, albeit to a lesser degree.
All of the chemically-coupled conjugates were examined by amino acid analysis in order to determine the number of NANP peptides displayed per carrier molecule (results shown in w, ..
Table 2). Table 2 clearly shows that the chemically coupling of synthetic peptide to a protein Garner results in a much higher number of peptides per carrier molecule than does recombinant display. The amino acid analysis of sN-f88 and sN-fl K, revealed that a large percentage of pVIIIs were coupled to the peptide on the phage (527% and 96114%, respectively). The enhancement of coupling of synthetic NANP to fl-K, above that of f88-4, can be attributed to the existence of the engineered Lys residue near the N-terminus of pVIII on fl-K
(see above). The peptide to carrier ratio for sN-MBP is also very high (318 to 1). The number of Lys residues in MBP is 35, so the coupling ratio of the NANP peptide to MBP must be near the limit.
As mentioned above, the recombinant display of the NANP peptide on the phage and MBP was of lower density. The copy number of the NANP peptide on the phage was determined visually by SDS-PAGE (Fig. and found to be very low (~16 copies per phage, assuming 3900 copies of pVIII for f88-4 phage, or 0.4% of the total pVIII).
Our experience with recombinant pep:pVIII expression (i.e-, with clones from f88-based vectors) has been that roughly 0.5%-5% of the major coat proteins are recombinant. Thus, the NANP
peptide is poorly expressed in our recombinant pVIII-display system, and may reflect the immunogenicity of other, poorly expressed pep:pVIII fasions. The copy number of peptides was assumed to be 5 per phage for rN:pIII, and 1 copy per hiBP molecule for rN:MBP (data summarized in Table 2).
The recombinant phage, rN:pIII and rN:pIII, and the recombinant MBP displaying the NANP peptide, rN:MBP, were all probed with the NANP-specific, monoclonal Ab Pf2A.10 (Wirtz et al., 1987), and all the recombinant constructs were found to be positive by ELISA (data not shown). Immobilized, synthetic N.AIv'p peptide also produced a strong signal by ELISA

J ~u probing with Pf2A.10 (data not shown). By KinExA analysis, the free NANP
peptide bound to Pf2A 10 with a Kd = 130 nM (Error bounds: Kd High = 173 nM, and Kd low = 81 nM).
Mouse immunizations Six groups of Balb/c mice (groups 1-6) were immunized with different recombinant, and chemically-coupled peptide-phage conjugates, and their sera were tested for anti-peptide and anti-carrier Ab responses by ELISA. 'Ve found from a previous dosage experiment (data not shown) that 20 ug of phage was sufficient to elicit an anti-peptide Ab response in BALB/c mice, as well as a strong anti-phage response; thus, the mice were injected with an equivalent of 25 ug of phage protein (see Table 1). In most cases, the mice were primed and boosted with the same immunogen. In two groups (3 and 4), however, we tried a heterologous immunization in which the priming and boosting immunogens were different (see Table 1). The purpose of the heterologous immunizations were to sew if the differences in pVIII would alter the immunogenicity of the phage such that, more of the Ab response would be directed toward the peptide than the phage coat in the boosting conjugate (see Discussion). The mice in groups 9 and 10 were immunized with SMCC-treated f88-4, and SMCC-treated fl-K, respectively. They were added as controls in case there was a large difference in the Ab response against the phage carrier due to the alteration of the phag~ coat by (i), the addition of the four amino acid residues in the pVIII of fl-K and, (ii), the modification caused by the crosslinker, sulfo-SMCC.

The mouse sera were tested by ELISA for their reactivity with biotinylated NANP
peptide that was captured on immobilized streptavidin (results shown in Fig:.
Most striking, is that the the anti-peptide response was greatest with the chemically-conjugated peptide immunogens, and it did not matter significantly whether the carrier was f88, fl-K or MBP. At serum dilutions less than the 1:800 dilution shown in Fig., there was a detectable IgG response against NANP peptide for mice immunized with rN:pIII and rN:pVIII (i. e., O.D.
~0.2 at 1:50 serum dilution), indicating that pVIII display was not significantly better than pIII display despite having a 3-fold difference in the peptide to phage ratio. On the other hand, there was a significant ELISA signal against peptide at a 1:800 serum dilution for groups 1-4, 7 and 8. The order of serum-reactivity of these groups of mice with the NANP peptide, correlates with the number of peptides per kDa of carver as shown in the right-hand column of Table 2. Thus, our results support the strategy of increasing the copy number of peptide on a given carrier to achieve the maximum anti-peptide Ab response. Again, the strongest anti-peptide responses were from the synthetic peptide conjugates. Given that the results are from pooled sera, it was not seen how the responses varied within a group. Thus, it is difficult to make inferences about the smaller differences in ELISA signals such as amnong the groups receiving the different synthetic peptide conjugates. Nonetheless, it remains char that the anti-peptide response was best with the chemically-conjugated peptide immure~gens, and it did not matter significantly whether the carrier was f88, fl-K or MBP.
We questioned whether the lower reponse against the recombinant counterparts was due to a generally low response against the entire immunogen (perhaps the chemical coupling itself enhanced the immunogenicity of these immunogens). Thus, we tested the response against Garner and saw that the carrier responses were comparable, and thus the response against peptide S
relative to carrier was enhanced with synthetic conjugates. Fig. ~ shows the results of ELISAs comparing the IgG response against f88-4 phage for all the phage-immunized mice. Apart from some fluctuations in ELISA signals, most of the conjugates had similar reactivities toward f88-4 phage. This was an important result because it shows that even with high levels of crosslinking with peptide, a potent anti-phage Ab response is elicited, and that the Abs against either fl-K or f88 phage are largely cross-reactive with one another. This does not necessarily mean that the Ab specificities against pVIII in either phage are identical, as there could be very little similarity in the pVIII-specific Abs toward the ri~-o phage. (See Western blot results) However, a significant portion of the Ab response is directed toward the minor coat proteins (i.e., pIII, pVI, pVII and pIX), and these Abs will recognize both phage with identical affinity and specificity.
Fig~shows that pooled sera from group 9 had a significantly lower ELISA signal against phage than all the others. It is unclear as to why this is, since SMCC-f88 is less modified than sN-f88;
the latter gave a strong ELISA signal against f88-4 phage. Possibly, the group-size (4 mice) was not be representative of the average anti-phage Ab response with this immunogen. The carrier response was also checked for the mice immunized with the recombinant, and chemically-coupled peptide-MBP conjugates. The ELISA results showing the IgG response of pooled sera from mice immunized with sN-MBP (gzoup 7), rN:MBP (group 8) or S14ZCC-MBP
(group 11) are shown in Fig.~C As can be seen, the anti-MBP IgG response was greater for mice immunized with rN-MBP. This was not unexpected because the mice in group 8 received more protein (33 ug) than did groups 7 and 11 (10 ug), amd the latter two were immunized with chemically-modified MBP.

i '" ., , , . _ .,~
Discussion In this study, it was determined that the phage can be used as an effective classical carrier for the preparation of synthetic peptide conjugates, capable of eliciting good anti-peptide Ab responses in BALB/c mice. We have shown this also for two other peptides (data not shown), which, similar to the NANP peptide, conjugated to the fl-K phage at high density (more than one peptide per copy of pVIII, data not shown). We also showed that the potency of the anti-peptide Ab response elicited by a conjugate of the NANP peptide with MBP, matched that of the phage conjugates, all of which were superior to recombinant phage and MBP in eliciting an Ab response against peptide.
Hey et al. ( 1994) used what they termed "a two fusion partner system" for eliciting anti-peptide Abs in rabbits. Their results indicated that, in some cases, the peptide response could be increased relative to the carrier response if the same peptide were used to boost the animal but that a different carrier be used for the prime and boosts. In their study, however, T-cell help probably derived, at least in part, from the fusion peptides themselves, which were fairly large (39 to 96mers). Because the sera v~~e used for each group were pooled from 4 mice each, there lacks the replicates to do sensitive statistics. Thus, we can be confident of large differences, but we have reservations about smaller ones such as in the heterologous immunizations. With this in mind, there does appear to be slightly utter ELISA signals against peptide for serum from heterologously-immunized mice (groups 3 and 4), as compared to those that were immunized homologously (groups 1 and 2; see Fi~~. It must be pointed out, however, that the anti-phage Ab response appears to be stronger in the heterologous group as well (Fig.. It also appears that sN-f88 was somewhat better than sN-fl K at eliciting an anti-NANP
response in mice. This may suggest that the density of the NANP peptide on sN-fl K is too high, and that the optimal ratio of peptide to pVIII for eliciting a potent anti-peptide response in mice is less than 1 to 1. In any event, the fl -K phage may have had an enhancing effect on anti-peptide Ab production in the heterologous immunizations. Iin order to improve upon this effect, further mutagenesis of fl-K
would be necessary to reduce the cross-reactive response against phage proteins. The filamentous phage are particularly well-suited to mutagenesis, and heterologous immunization experiments with divergent phage clones may be a particularly attractive means of eliciting anti-peptide Ab responses, as well as examining the behaviour of the Ab response following heterologous challenge.
In summary, we found that the lack of a potent anti-peptide Ab response induced by phage displaying recombinant peptide on pIII, or by phage with a low copy number of recombinant peptide on pVIII, can be overcome by the multivalent display of synthetic peptide chemically-coupled to the phage coat proteins. Our results must be taken with some caution, however, if a synthetic peptide is derived from an eptitope mimic or "mimotope". If it is determined that a peptide derived from a phage-display library intimately depends on the phage coat for its activity, the synthetic version may be a poor candidate for immunizations, at least without further optimization. The addition of a Lys residue in fl-K, permitted superior levels of peptide crosslinking using the heterobifunctional crosslinker, sulfo-SMCC. The correlation beriveen the surface density of a peptide on an immunogen and the immunogenicity of the peptide may, however, have a limit. since serum from mice immunized with sN-f88 gave a stronger anti-peptide Ab response than did the serum against sN-fl K. Because phage are so ideally-suited to mutagenesis, further modifications to pVIII may be easily achieved. We chose to position the extra Lys residue in fl-K a few residues downstream of the leader peptidase cleavage site, since Lys residues positioned immediately adjacent to this site may inhibit the insertion of promature proteins in the inner membrane of E. coli (Andersson &
von Heijne, 1993). It may be possible, however, to shift the lysine residue introduced in fl-K even closer to the N-terminus of pVIII. Moreover, the pre-existing lysine residue in fl-K, corresponding to position 8 in fl, could be mutated to an arginine or another polar residue, to reduce side-reactions from occurring. The outer domains of pIII could be removed to lower the complexity and cross-reactivity in the Ab response against heterologous challenge with phage. The phage could also be genetically engineered to display a T-cell epitope, or another reactivity prior to the chemical conjugation step. Experiments involving the interplay between recombinant technology and chemical-conjugation should yield some interesting results for vaccine research in the future.
Acknowledgements We thank Scott Diguistini for technical assistance, and Loeki Van der Waal and the staff at the SFU Animal Care Facility for the care and handling of the mice used in this study. This work was supported by IJ .~ . . , References Andersson, H. and von Heijne, G. (1993) Position-specific Asp-Lys pairing can affect signal sequence function and membrane protein topology. J. Biol. Chem. 268, 21389-93.
Armstrong, J., Hewitt, J. A. & Perham, R. N. (1983). Chemical modification of the coat protein in bacteriophage fd and orientation of the virion during assembly and disassembly.
EMBOJ2(10), 1641-6.
Blake, D. A., Khosraviani, M., Pavlov, A. R. & Blake, R. C. (1997).
Characterization of a metal-specific monoclonal antibody. Immunol. Technol. Environ. Applic. 657, 49-60.
Berkowitz, S. A. & Day, L. A. (1980). Turbidity measurements in an analytical ultracentrifuge.
Determinations of mass per length for filamentous viruses fd, Xf, and Pf3.
Biochemistry 19(12), 2696-702.
Bonnycastle, L.L., Mehroke, J.S., Rashed, M., Gong, X. and Scott, J.K. (1996) Probing the basis of antibody reactivity with a panel of constrained peptide libraries displayed by filamentous phage. J. Mol. Biol. 2~8, 747-62.
Bonnycastle, L.L.C. (1998) DNA sequences obtained from filamentous bacteriophage virions by direct cycle sequencing. Anal. Biochem. 256, 140-142.
Craig, L., Sanschagrin, P. C., Rozek, A., Lackie, S., Kuhn, L. A. & Scott, J.
K. (1998). The Role of Structure in Antibody Cross-Reactivity Between Peptides and Folded Proteins. J. Mol.
Biol. 281, 183-201.
Day, L.A. (1969) Conformations of siagle-stranded DNA and coat protein in fd bacteriophage as revealed by ultraviolet absorption spectroscopy. J Mol Biol 39, 265-77.

i i . . , , ., de la Cruz, V.F., Lal, A.A. and McCutchan, T.F. (1988) Immunogenicity and epitope mapping of foreign sequences via genetically engineered filamentous phage. J Biol Chem 263, 4318-22.
Demangel, C., Lafaye, P. & Mazie, J. C. (1996). Reproducing the immune response against the Plasmodium vivax merozoite surface protein 1 with mimotopes selected from a phage-displayed peptide library. Mol Immunol 33(11-12), 909-16.
Glucksrnan, M. J., Bhattacharjee, S. & Makowski, L. (1992). Three-dimensional structure of a cloning vector. X-ray diffraction studies of filamentous bacteriophage M13 at resolution. JMoI Biol 226(2), 455-70.
Greenwood, J., Willis, A. E. & Perham, R.. ( 1991 ). Multiple display of foreign peptides on a filamentous bacteriophage. Peptides from Plasmodium falciparum circumsporozoite protein as antigens. JMol Biol 220(4), 821-7.
Hermanson, G.T. (1996) Maleimides. In: Bioconjugate Techniques. Academic Press, San Diego, CA, p. 148.
Hey, A.W., Johnsen, J.L, Johansen, B. and Traavik, T. (1994) A two fusion partner system for raising antibodies against small immunogens expressed in bacteria. J Immunol Methods 173, 149-56.
Hill, D.F. and Petersen, G.B. (1982) Nucleotide sequence of bacteriophage fl DNA. J Virol 44, 32-46.
Jelinek, R., Terry, T. D., Gesell, J. J., Malik, P., Perham, R. N. & Opella, S. J. (1997). NMR
structure of the principal neutra3izing determinant of HIV-1 displayed in filamentous bacteriophage coat protein. J~lIn1 Biol 266(4), 649-55.

, ;, '.i Kneissel, S., Queitsch, L, Petersen, G., Behrsing, O., Micheel, B. & Dubel, S.
(1999). Epitope structures recognised by antibodies against the major coat protein (g8p) of filamentous bacteriophage fd (Inoviridae). JMoI Biol 288(1), 21-8.
Malik, P., Terry, T.D., Gowda, L.R., Langara, A., Petulcllov, S.A., Symmons, M.F., Welsh, L.C., Marvin, D.A. and Perham, R.N. (1996) Role of Capsid Structure and Membrane Protein Processing in Determining the Size and Copy Number of Peptides Displayed on the Major Coat Protein of Filamentous Bacteriophage. J. Mol. Biol. 260, 9-21.
Malik, P., Terry, T.D., Bellintani, F. and Perham, R.N. (1998) Factors limiting display of foreign peptides on the major coat protein of filamentous bacteriophage capsids and a potential role for leader peptidase. FEBS Lett 436, 263-6.
Martineau, P., Charbit, A., Leclerc, C., alerts, C., O'Callaghan, D. &
Hofnung, M. (1991). A
Genetic System to Elicit and Monitor Anti-Peptide Antibodies Without Peptide Synthesis. BiolTechnology 9, 170-172.
Martineau, P., Guillet, J.-G., Leclerc, C. & Hofnung, M. (1992). Expression of heterologous peptides at two permissive sites of the MaIE protein: antigenicity and immunogenicity of foreign B-cell and T-cell epitop~s. Gene 113, 35-46.
Marvin, D. A., Hale, R. D., Nave, C. & Citterich, M. H. (1994). Molecular models and structural comparisons of native and mutant class I filamentous bacteriophages Ff (fd, fl, M13), Ifl and IKe. JMoI Biol 235(1), 26~--86.
Meola, A., Delmastro, P., Monaci, P., Luzzago, A., Nicosia, A., Felici, F., Cortese, R. and Galfre, G. ( 1995) Derivation of vaccines from mimotopes. Immunologic properties of _. , ,. .. . . . . ,,3 human hepatitis B virus surface antigen mimotopes displayed on filamentous phage. J
Immunol 154, 3162-72.
Minenkova, O. O., Ilyichev, A. A., Kishchenko, G. P. & Petrenko, V. A. (1993).
Design of specific immunogens using filamentous phage as the carrier. Gene 128(1), 85-8.
Morrissey, J. H. (1981 ). Silver stain for proteins in polyacrylamide gels: a modified procedure with enhanced uniform sensitivity. Anal Biachem 117(2), 307-10.
Motti, C., Nuzzo, M., Meola, A., Galfre, G., Felici, F., Cortese, R., Nicosia, A. & Monaci, P.
(1994). Recognition by human sera and immunogenicity of HBsAg mimotopes selected from an M 13 phage display library. Gene 146(2), 191-8.
Sambrook, J., Fritsch, E.F. and Maniatis, T. (1989) Molecular Cloning: A
Laboratory Manual.
Cold Spring Harbor Laboratory Press, Plainview, New York.
Schagger, H. and von Jagow, G. (1987) Tricine-sodium dodecyl sulfate-polyacrylamide gel electrophoresis for the separation of proteins in the range from 1 to 100 kDa.
Anal Biochem 166, 368-79.
Scott, J. K. & Smith, G. P. (1990). Searching for Peptide Ligands with an Epitope Library.
Science 249, 386-390.
Smith, G. P. & Scott, J. K. (1993). Libraries of peptides and proteins displayed on filamentous phage. Methods Enrymol 217, 2'28-57.
Smith, G. P., and Petrenko, V. A. (1997) Phage Display Chem. Rev. 97, 391-410.
Tam. J.P. (1988) Synthetic peptide ~-ac~ine design: synthesis and properties of a high- density multiple antigenic peptide svste~. Proc Natl Acad Sci U S A 85, X409-13.

. .. . . ~i Willis, A. E., Perham, R. N. & Wraith, D. (1993). Immunological properties of foreign peptides in multiple display on a filamentous bacteriophage. Gene 128, 79.
Wirtz, R. A., Zavala, F., Charoenvit, Y., Campbell, G. H., Burkot, T. R., Schneider, L, Esser, K.
M., Beaudoin, R. L. & Andre, R. G. (1987). Comparative testing of monoclonal antibodies against Plasmodium falciparum sporozoites for ELISA development.
Bull.
W.H.O. 65(1), 39-45.
Zhong, G., Smith, G.P., Berry, J. and Brunham, R.C. (1994) Conformational mimicry of a chlamydial neutralization epitope on filamentous phage. J. Biol. Chem. 269, 24188.
Zwick, M. B., Shen, J. & Scott, J. K. (1998a). Phage-displayed peptide libraries. Curr Opin Biotechnol 9(4), 427-36.
Zwick, M. B., Bonnycastle, L. L., Noren, K. A., Venturini, S., Leong, E., Barbas, C. F., 3rd, Noren, C. J. & Scott, J. K. (1998b). The maltose-binding protein as a scaffold for monovalent display of peptides derived from phage libraries. Anal Biochem 264( 1 ), 87-97.

;.
.. , ,,., a Table 1 Mouse Immunization Schedule Group Number Priming Immunogen Used for Dosage per of Mice Immunogen First (Day 14) and Injection' in Group (Day 0) Second (Day 28) Boosts Experimental groups:

1 4 sN-fl K sN-fl K 25 ug 2 4 sN-f88 sN-f88 25 ug

3 4 sN-f88 sN-f1K 25 ug

4 3 sN-fl K sN-f88 25 ug 4 rN:pIII rN:pIII 25 ug 6 4 rN:pVIII rN:pVIII 25 ug 7 4 sN-MBP sN-MBP 10 ug 8 4 rN:MBP rN:MBP 33 ug Control groups:

9 3 SMCC-f88 SMCC-f88 25 ug 3 SMCC-f1K SMCC-f1K 25 ug 11 3 SMCC-MBP SMCC-MBP 10 ug ' All of the mice were immunized intraperitoneally (i.p.) with 100 uL 1 mg/mL
Adjuvax adjuvant in PBS. All of the phage-immunized mice received 8.3 x 10" conjugated phage particles per injection, which is ~25 ug protein assuming that an O.D. at 269 nm of 30 is 6 mg/mL of phage protein.

Table 2 The number of NANP peptides displayed per carrier molecule for the recombinant and synthetic peptide immunogens Number of peptides displayed / coupled Name Description of Peptide-Carrier Conjugate eer p~u~ per phage'per lcDao f p carrier III or MBP

sN-f88 f88 vector crosslinked 0.52 2030 0.088 to synthetic NANP peptide sN-f1K f1K vector crosslinked 0.96 2590 0.16 to synthetic NANP peptide rN:pIII recombinant NANP peptide1 5 0.00022 displayed on pIII

rN:pVIII recombinant NANP peptide0.004 16 0.0007 displayed on pVIII

sN-MBP MBP crosslinked to synthetic31 - 0.60 NANP peptide rN:MBP recombinant NANP peptide1 - 0.019 displayed on MBP

a the number of copies is calculated assuming fl-K (6.4 kb) has 2700 pVIIIs and f88-4/rN:pVIII
(9.2 kb) and rN:pIII (9.2 kb) have 3900 pVIIIs.
b the molecular-weight of fl-K approximately 16 Mda (Berkowitz & Day, 1980), and thus f88-4 was calculated to be 23 Mda. The molecular weight of the MBPs produced from the pMal-X
vector is ~52 kDa.

~ 'd .r . .
Fig.? Cartoon illustration of the filamentous bacteriophage.
Fig.l. Engineering a Lys residue into the N-terminus of the mature, major coat protein (pVIII) of fl phage. A. A partial nucleotide and amino acid sequence of pVIII of fl phage (same as for fd and its derivative, f88-4 phage). The vertical line (~) indicates the leader peptidase cleavage site, and the wedge (") indicates the insertion site of the four residues in fl-K. The underlined sequence corresponds to the region of homology to Primerl (see Materials and Methods). B.
The analogous region in fl-K as is shown for fl in A. The four residues inserted in the N-terminal region of mature pVIII are shown between the asterisks (*). The unique HindIII site of fl-K is also shown (underlined).

Fig.',. Estimating the number of NANP peptide:pVIII fasions per phage particle. Lane 1: f88-4, 6x10'° phage particles, Lanes 2-9: Decreasing amounts of recombinant phage (3-fold diltuions).
Lane 2: 6x10'° particles. Lane 3: 2x10'° partciles. Lane 4:
6.7x109 particles. Lane 5: 2.2.x109 particles. Lane 6: 7.4x108 particles. Lane 7: 2.5x108 particles. Lane 8:
8.2x10' particles. Lane 9: 2.7x10' particles. Based on the band intensities of the wild-type pVIII
band in Lane 8 and the NANP peptide:pVIII band in Lane 3, one can wstimate that the expression of the recombinant band is 5 3-fold dilutions, or 243-fold Iower than that of wild-type pVIII.
Given a pVIII copy number of 3900 for f88-4 phage, the copy number of NANP peptide:pVIII per phage is ~16.
[Reprinted with permission CSHL?]

Fig.: SDS-PAGE analysis of phage proteins (A.) and MBP (B.) before and after treatment with sulfo-SMCC and peptide coupling. A. Lane l: f88, no treatment. Lane 2: fl-K, no treatment.
Lane 3: f88, treatment with sulfo-SMCC. Lane 4: fl-K, treatment with sulfo-SMCC. Lane S:
fl-K, after coupling with NANP peptide. Lane 6: f88, after coupling with NANP
peptide. B.
MBP
Fig.4. ELISA signals against immobilized NANP peptide for pooled sera from mice immunized with different carriers displaying the NANP peptide. The categories (bottom) indicate the ELISA signals of pooled serum from each group of mice (see Table ~, following the prime, first and second boosts (indicated on the right with the serum dilution used).
The results are the average from two separate experiments.
Fig.S. ELISA signals against immobilized f88-4 phage for pooled sera from mice immunized with different carriers displaying the NANP peptide. The results are the average from two separate experiments. The categories (bottom) indicate the ELISA signals of pooled serum from each group of mice (see Table ~, following the prime, first and second boosts (indicated on the right with the serum dilution used). The results are the average from two separate experiments.
Fig.6. ELISA signals against immobilized MBP for pooled sera from mice immunized with different carriers displaying NANP peptide. The results are the average from two separate experiments. The categories (bottom) indicate the ELISA signals of pooled serum from each group of mice (see Table ~, following the prime, first and second boosts (indicated on the right with the serum dilution used). The resu.~ts are the average from two separate experiments.

Claims