WO1994017180A1 - Nuclear localized transcription factor kinases - Google Patents

Abstract

New nuclear kinases that are regulated by signal transduction and that participate in phosphorylation cascades which regulate transcription and related methods for regulating transcription. The novel nuclear kinases play a vital role in gene expression, particularly with regard to the proper expression of genes involved in the avoidance of leukemia in humans. The kinase is (i) substantially exclusively intranuclearly localized; (ii) capable of autophosphorylation; (iii) selectively bindable with antibodies raised against GST-fsh and GST-RING3; (iv) of a molecular weight of from about 82.5 to about 92.7 kilodaltons; and (v) includes peptide sequences Asp-Ser-Asn-Pro-Asp-Glu-Ile-Glu-Ile-Asp-Phe-Glu-Thr-Leu-Lys-Pro-Thr-Thr-Leu and Ala-Val-His-Glu-Gln-Leu-Ala-Ala-Leu-Ser-Gln-Ala-Pro. Protein kinase activity was also discovered for fsh (Drosophila) and RING3 (human). RING3 is attributed with a role consistent with that of the novel kinases.

Description

NUCLEAR LOCALIZED TRANSCRIPTION FACTOR KINASES Background of the Invention

The study of how the phosphorylation of proteins by protein kinases regulates eukaryotic transcription has enjoyed rapid progress. It has long been known that many extracellular receptors transfer structural information from hormones and growth factors into phosphorylation information. A well understood and growing set of cytosolic kinases, including protein kinase A, protein kinase C, calcium

calmodulin-dependent kinases, oncogene-encoded kinases, casein kinase II, glycogen synthase kinase III and the

erk -encoded family of kinases, have been shown to contain biochemical signals destined for the nucleus (Czech et al.,1988; Bohmann, 1990; Blenis,1991; Jackson, 1992). In the nucleus, phosphorylation has been shown to affect both the DNA-binding activity and the transcriptional activation of several transcription factors, some positively and some negatively (Hunter and Karin,1992). Activated transcription of immediate early genes seems to be conducted primarily through phosphorylation.

However, efforts to date to delineate pathways of information flow from receptor to promoter have tended to focus on the physical extension of cytosolic signalling cascaodes into example, by

nucleus-cytosol partitioning of transcription factor substrates and nuclear forms of cytoso l i c kinases (Hunter and Karin, 1992). Cytosolic protein kinase C has Deen shown to phosphorylate the NFkB-IkB complex in the cytosol, leading to dissociation of the complex, whereupon

transcriptionally active NFkB partitions into the nucleus (Baeuerle and Baltimore, 1988a; b). Furthermore, a number of kinases that are activated by extracellular signals appear to translocate to the nuclear membrane (or to a nuclear fraction) upon mitogenic stimulation of the cell, notably protein kinase A (Nigg et al., 1985; Adams et al., 1991), protein kinase C (Leach et al., 1989),

microtubule-associated protein/myelin basic protein

(MAP/MBP) kinase and ribosomal S6 kinase (RSK) (Chen et al. 1992). Some kinases that phosphorylate nuclear targets seem to be activated in the cytosol, such as the cellular homolog of c-abl (Van Etten et al., 1989; Dikstein et al., 1992; Kipreos and Wang, 1992); whereas the partitioning behavior of others is either not known to be regulated, such as FER (Hao et al., 1991) and Spl kinase (Jackson et al., 1990); or is poorly understood, such as the transcription

complex-associated kinase that phosphorylates the C-termmal domain of RNA polymerase 11 (Feaver et al., 1991; Lu et al., 1992) and p33^cdk2, which is present in a complex with the transcription factor E2F, cyclin A and plO7 (Devoto et al.,

1992).

While these physical translocations are undoubtedly important features of signal transduction, they provide no suggestion that there might also exist signalling kinases that are downstream targets of cytosolic kinases but that are restricted solely to the nucleus. T he I nvention

New nuclear kinases that are regulated by signal transduction and that participate m phosphorylation cascades which regulate transcription and related methods for regulating transcription are described. The novel nuclear kinases play a vital role in gene expression, particularly with regard to the proper expression of genes involved in the avoidance of leukemia in humans. The kinase is (i) substantially exclusively intranuclearly localized; (ii) capable of autophosphorylation; (ii) selectively bindable with antibodies raised against GST-fsh and

GST-RING3; (iv) of a molecular weight of from about 82.5 to about 92.7 kilodaltons; and (v) includes peptide sequences Asp-Ser-Asn-Pro-Asp-Glu-Ile-Glu-Ile-Asp-Phe-Glu-Thr-Leu- Lys-Pro-Thr-Thr-Leu and Ala-Val-His-Glu-Gln-Leu-Ala-Ala- Leu-Ser-Gln-Ala-Pro. Protein kinase activity was also discovered for fsh (Drosophila) and RING3 (human). RING3 is attributed with a role consistent with that of the novel kinase.

In the process of arriving at the invention, nuclear extracts of cultured cell lines were screened for previously uninvestigated kinases that exhibit regulated activity and that might phosphorylate transcription factors. In order to increase the likelihood that such a kinase is important in transcription, it3 principal features were considered to include primarily (although not necessarily exclusively) nuclear localization, and responsiveness to drugs, hormones or conditions that are known independently to promote transcription in cultured cells.

A straightforward screen of nuclear extracts of Jurkat cells, performed by adding magnesium and - ³²

P ] ATP to the

denaturing gel elec phores and auto radiography did not reveal any obvious differences in phospnorylation between the proteins from ceiis stimulated with phorbol myristate acetate/phytohemagglutinin and controls. Too many

phosphorylated proteins obscured potential differences.

In accordance with the invention, this problem was addressed and overcome by nitrocellulose electroblotting gels of electrophoretically separated proteins, denaturing the blotted proteins in guanidine hydrochloride (e.g., 7M), renaturing them in buffer and then probing the blot with 3² P]ATP (Ferrell and Martin, 1989). This assay visualized only those kinases which could renature after gel

electrophoresis in sodium dodecyl sulfate and could

autophosphorylate.

of the Drawings

Figure 1A Aur phosphe

f a 90 kDa kinase in Jurkat cells is increased by forskolin treatment. Nuciear extracts from forskolin-treated and control Jurkat cells were fractionated on phosphocellulose columns. The columns were eluted stepwise with 0.3 M and 0.6 M sodium chloride in buffer. Eluted proteins were precipitated with

trichloroacetic acid, solubilized in sodium dodecyl sulfate, separated by gel electrophoresis, elctroblotted to

nitrocellulose, denatured in guanidme hydrochloride, renatured in buffer, and assayed by autophosphorylation in the presence of

P]ATP and magnesium. Autoradiographs of blots are shown. Unfractionated nuclear extracts (UF) are included for comparison.

Figure 1B: Time course of increased

autophosphorylation that is induced by forskolin. Jurkat cells were stimulated with 50 μM forskolin for 0, 3, 6, 9, 12, 15, and 30 minutes. Nuclear extracts of these cells were applied to phosphocellulose. Proteins were eluted with 0.3 M sodium chloride and assayed as described for Figure 1A.

Figure 1C: Treatments that affect 90 kDa

autophosphorylation in A431 cells. A431 cells were treated with epidermal growth factor (1 nM), calf serum (20%), forskolin (25 μM) or heat (42°C) for 30 minutes. Nuclear extracts of treated cells and control cells were applied to phosphoceliulose. Proteins were eluted with 1 M sodium chloride and assayed as described for Figure 1A. The 90 kDa kinase is identified with an arrow.

Figure 2: The 90 kDa Kinase activity is localized to the nucleus in HeLa cells. A

of HeLa cells was prepared in the presen

phatase inhibitors a nd centrifuged at 100,000 m g for one hour at 4ºC. Equal amounts of protein (50 ug) from this supernatant (the cell cytosol) and a HeLa nuclear extract were separated by gel electrophoresis and assayed as described for Figure 1A. The 90 kDa kinase is identified with an arrow.

Figure 3: The 90 kDa kinose activity HeLa cells has an apparent molecular weight in excess of 1,000,000 by size exclusion chromatography. Proteins in HeLa nuclear extract were separated by size exclusion chromatography on Sephacryl S-200 at physiological ionic strength and assayed as

described for Figure 1A. The void volume of the column and molecular weight standards (for size exclusion

chromatography) for immunoglobulin (158,000) and ovalbumin (44,000) are indicated above the relevant fractions, and molecular weight standards (for gel electrophoresis) are indicated to the left of the figure. The 90 kDa kinase is identified with an arrow.

Figure 4: Renatured 90 kDa kinase was tested for its ability to phosphorylate myelin basic protein. Strips of nitrocellulose upon which the renatured kinase was

immobilized were incubated in Eppendorf tubes with a

solution of myelin basic protein, radioactive ATP and magnesium in buffer. After one hour at 30°C, proteins were solubilized at 100°C in sodium dodecyi sulfate, separated by gel electrophoresis and visualized by Coomassie blue stain. An autoradiograph of phosphorylated myelin basic protein is shown in the figure. Three species of kinase were tested in triplicate: the activity in unfractionated HeLa nuclear extract (UF), the activity in a 0.3 M sodium chloride fraction from phosphocellulose-fractionated HeLa nuclear extract, and the

: in

sodium chloride

fraction of the same

Figure 5A: Fractionation of autophosphorylatmg 90 kDa kinase by phosphocellulose (P-11) chromatography. HeLa nuclear extract was applied to phosphoceliulose and eluted stepwise in 0.1 M increments with 0.2 M to 0.8 M sodium chloride in buffer. Eluted proteins were assayed as

described for figure 1A. Unfractionated nuclear extract is included for comparison at the left of the figure.

Molecular weight markers are shown at the extreme left and an arrow indicates the average apparent molecular mass of the 90 kDa autophosphorylating species.

Figure 5B: Substrate-directed phosphorylation activity of the 90 kDa kinase is associated with the slower-migrating forms identified by autophosphorylation, not with the faster-migrating forms. The same fractions from figure 5A were renatured and used to phosphorylate meylin basic protein as described in figure 4.

Figure 6A: Autophosphorylated amino acids are serine and threonine, not tyrosine. HeLa nuclear extract was applied to phosphoceliulose and eluted with 0.3 M sodium chloride. The 90 kDa kinase was isolated, renatured as described for figure 1A, and autophosphorylated with

[ -³²P]ATP. The radioactive kinase was then hydrolyzed in

6N hydrochloric acid and its constituent radioactive

aminoacids were separated by electrophoresis on silica. An autoradiograph of the silica sheet is shown. Authentic phosphoaminoacid standards were run in the same sample and visualized with ninhydrin to identify the radioactive spots. Figure 6B: Myelin

in is phosphorylated only in serine. (Left side) Myelin b a is protein was

phosphorylated with renatured 90 kDa kinase as described for figure 4, then hydrolyzed and separated as described for figure 6A. (Right side) As a control for tyrosine

phosphorylation, myelin basic protein was also

phosphorylated in a separate reaction with recombinant p43^v-abl protein.

Figure 7: Immune sera to fsh and RING3

immunoprecipitate the 90 kDa autophosphorylation activity. The cDNAs for fsh and RING3 were obtained and fusions with glutathione-S-transferase (GST) were constructed. These constructs were expressed in E. coli, the fusion proteins were purified and used to immunize rabbits. The 90 kDa kinase was immunoprecipitated from HeLa nuclear extracts with these antisera and assayed by autophosphorylation as described for figure 1A. Pre-immune sera were not capable of immunoprecipitating this activity. Autoradiographs of 90 kDa kinase renatured from immunoprecipitates of preimmune and immune sera against RING3 (above) and fsh (below) are shown.

Figure 8: Immunofluorescence with antibodies against RING3 confi rm nuclear localization of 90 kDa kinase in HeLa cells . HeLa cells were grown on coverslips and fixed by standard methods. Coverslips were incubated with preimmune serum or with affinity-purified antibodies against GST fusion proteins of RING3 or fsh. Goat anti-rabbit secondary antibodies conjugated to fluorescein were used to visualize the primary antibodies. Description

In one aspect the invention relates to a method for regulating transcription of a gene of interest. The method comprises modulating the activity of a kinase that activates transcription of the gene and is (i) substantially

exclusively intranucleariy localized; (ii) capable of autophosphorylation; (iii) selectively bmdable with

antibodies raised against GST-fsh and GST-RING3; (iv) of a molecular weight of from about 82.5 to about 92.7

kilodaltons; and (v) includes peptide sequences

Asp-Ser-Asn-Pro-Asp-Glu-Ile-Glu-Ile-Asp-Phe-Glu-Thr-Leu- Lys-Pro-Thr-Thr-Leu and Ala-Val-His-Glu-Gln-Leu-Ala-Ala- Leu-Ser-Gln-Ala-Pro.

In this aspect, modulating the kinase activity can comprise stimulating the phosphorylation activity of the kinase, so as to effect either inducing transcription or preventing inhibition of transcription. More particularly, the kinase activates a transcription factor for initiating transcription of the gene of interest. Alternatively, modulating the kinase activity can comprise inhibiting the phosphorylation activity of the kinase so as to inhibit the transcription of the gene of interest. More particularly, the kinase activates a transcription factor for the gene of interest.

Another aspect of the invention provides a method for regulating the transcription of a gene of interest in a cell which comprises contacting said cell with a compound that specifically binds with a cell surface receptor so as to modulate by signal transduction the activity of a kinase that: (i) is substantially exclusively intranucleariy localized; (ii) is capable of autophosphorylation; (iii) is selectively bindable wit., antibodies raised against CST-fsh and OST-RING3; (iv) has a molecular weight of from about 82.5 to about 92.7 kilodaitons; and (v) includes peptide sequences Asp-Ser-Asn-Pro-Asp-Glu-Ile-Glu-Ile-Asp-Fhe-Clu- Thr-Leu-Lys-Pro-Thr-Thr-Leu and Ala-Val-His-Glu-Gln-Leu- Ala-Ala-Leu-Ser-Gln-Ala-Pro. In this aspect, the compound stimulates a cell surface receptor including those from the group consisting of peptide growth factor receptors, steroid and other hormone receptors, ion channel receptors,

serpentine cell surface receptors and immunophilins. This aspect also relates to stimulating the signal transduction function of said kinase by contacting said cell with

forskolin or analog or derivation thereof.

Also, lectin-binding receptors, antibody/antigen receptors, cell contact receptors that are otherwise

involved in self/non-self or immune recognition or that are encoded by the major histocompatability complex, cell adhesion receptors, steroid and other hormone receptors, chemotactic receptors and others.

In this aspect, regulating can comprise stimulating the phosphorylation activity of the kinase so as to either induce transcription or prevent inhibition of transcription of the gene of interest. Alternatively this aspect can include a method for inhibiting the phosphorylation activity of the kinase.

Another aspect of the invention provides a recombinant method of regulating transcription of a gene of interest. In this aspect, the transcription of a gene downstream or even upstream in a signal transduction pathway from the kinase of the invention is regulated by an expressable gene construct comprising a oter and a structural gene sequence for the kinase of the invention or a protein that affects the activity of the kinase that is introduced into the cell where transcription is to be regulated.

Another aspect of the invention relates to new nuclear kinases that are regulated by signal transduction and that participate in phosphorylation cascades which regulate transcription and related methods for regulating

transcription. The kinases are: (i) substantially- exclusively intranucleariy localized; (ii) capable of autophosphorylation; (iii) selectively bindable with antibodies raised against GST-fsh and GST-RING3; (iv) of a molecular weight of from about 82.5 to about 92.7

kilodaltons; and (v) include peptide sequences Asp-Ser- Asn-Pro-Asp-Glu-Ile-Glu-Ile-Asp-Phe-Glu-Thr-Leu-Lys-Pro- Thr-Thr-Leu and Ala-Val-His-Glu-Gln-Leu-Ala-Ala-Leu-Ser- Gln-Ala-Pro. Also contemplated are biologically active fragment derivatives, analogs, conjugates, and fusion proteins thereof.

As another aspect of the invention, protein kinase activity has now been discovered for fsh (Drosophila) and RING3 (human). RING3 is attributed with a role consistent with that of the novel kinases.

Another aspect of the invention relates to the

recognition of the relevance of the novel kinase(s) of the invention, fsh and RING3 to (i) the regulation or control of the cell cycle, (ii) control of cell differentiation, and (iii) control of cell growth and metabolism.

Regulated transcription of eukaryotic nuclear genes depends in part on the accurate processing of information about the extracellular

especially hormone, growth, factor, drug, or other ligand binding to plasma membrane (cell surface) receptors. This information is processed through multiple kinase cascades within the cell, ultimately giving rise to the activation of transcription factors in the nucleus. It is of vital interest to identity the kinases that play a role in this transduction of signals and to delineate the signalling networks involved.

In accordance with the invention it was reasoned that there must be many uncharacterized kinases that are mostly or entirely localized to the nucleus, that receive signals transduced from the cytosol and then go on to phosphorylate and activate nuclear components of the transcription

apparatus.

A previously uncharacterized autophosphorylation activity has now been identified in HeLa nuclear extract. Denaturing polyacrylamide gels of nuclear proteins were electroblotted to nitrocellulose, renatured and assayed for autophosphorylated bands by overlaying the blots with

(γ-³² P]ATP and magnesium ion. Autoradiography of the blots visualized a limited set of kinases, one of which was predominantly nuclear and which exhibited rapidly increased autophosphorylation in forskolin-stimulated Jurkat cells and serum and forskolin-stimulated A431 cells.

A forskolin analog with greater water solubility and activity approximately equal for forskolin was used in these studies. This analog is called

7β-deacetyl-7β (γ -N-methylpiperazino)-butyryl forskolin and is supplied by Calbiochem as product number 344273.

Because forskolin is a potent activator of adenylate

cyclase, other pharmacological agonists of adenylate cyclase, such as other

of

orskoin, are

expected to stimulate the activity of the 90 kDa kinase in a similar manner. By the same token, other drugs, hormones or growth factors that increase the intracellular activity of cyclic AMP-dependent protein kinase, which is the target of the cyclic AMP produced by adenylate cyclase, are also expected to stimulate the 90 kDa kinase. These other drugs or hormones include membrane-permeable derivatives of cyclic AMP, such as its dibutyryl ester or "caged" cyclic AMP;

isoproterenol, glucagon, epinephrine, bombesin or other agents that mobilize cyclic AMP in their responsive cell types; and other as yet uncharacterized factors in serum that mobilize cyclic AMP. In addition, agents that inhibit phosphodiesterase, which destroys the cyclic AMP, will increase intracellular cyclic AMP concentrations and by mobilizing cyclic AMP-dependent protein kinase, are expected to increase the activity of the 90 kDa kinase.

Phosphodiesterase inhibitors include such, agents as

3-isobutyl-1-methylxanthine. Finally, DNA expression vectors that encode the catalytic subunit of protein kinase A will also increase the intracellular activity of protein kinase A and are thereby expected to increase the activity of the 90 kDa kinase.

The kinase activity resolved as a multiplet of average apparent molecular mass 90 kDa on SDS gels.

Phosphoceliulose chromatography fractionated the multiplet into forms of slightly slower and faster mobility. The forms of slightly slower mobility on SDS gels bound more weakly to phosphoceliulose than the forms of slightly faster mobility, probably due to hyperphosphorylation, and these exhibited both increased autophosphorylation and increased substrate-directed phosphorylation if they were isolated from cells that had been stimulated with serum or forskolin Analogous phenomena have beer. b served with pp70^56K (Prise et al., 1992) and pp90^rsK (Vik et al., 1990) S6 kinases, for which the higher phosphorylated forms exhibit both reduced mobility on SDS gels and increased substrate-directed phosphorylation. This pattern, in which increased

phosphorylation of the kinase reflects increased

substrate-directed activity, is observed also with protein kinase C (Mitchell, et al., 1989) and appears to define a subclass of intracellular signalling kinases.

This 90 kDa autophosphorylation activity appears to be ubiquitous, as it was observed in every mammalian cell line assayed. Its presence in yeast, Drosophila or plants has not yet been determined. The occurrence of well-resolved autophosphorylating kinases of the same apparent mobility in a wide variety of cells was consistent with the fact that the renaturation assay that was used selected for

independently refolding catalytic domains that do not require protein disulfide isomerase or chaperonin activity to regain function. Such stably folded kinases are likely to share certain structural features or folding pathways that are conserved in different organisms.

Autophosphorylation was dramatically reduced to a basal level when the extracts were prepared in the absence of phosphatase inhibitors or when extracts were warmed to 37°C in the presence of 5 mM magnesium ion, thereby eliminating differences between extracts prepared from control or stimulated cells, a frequently observed feature of

signal-transducing kinases (Kozma et al., 1990). Taken together, these observations pointed to the interpretation that the 90 kDa kinase is probably rapidly and reversibly phosphorylated and activated by an intracellular signal transduction pathway. These changes to autophosphorylation activity were remarkably stal a phenorenon that has been observed among certain kinase: involved in signal

transduction, such as RSK and MAP kinases (Ahn et al., 1990).

Consistent with a large body of work on "switch

kinases" and nuclear phosphorylation (Ahn and Krebs, 1990; Pelech et al., 1987; 1990) that has failed to detect

phosphotyrosine phosphorylation in the nucleus, the 90 kDa species contained no phosphotyrosine, and phosphorylefted its myelin basic protein substrate only on serine. Myelin basic protein is used frequently as an in vitro substrate for MAP kinases, and its suitability here suggested that the 90 kDa kinase might share some structural or functional

similarities with the members of the MAP kinase family or with the RSK family, which is immunologically related to the MAP kinases. However, the 90 kDa kinase did not

phosphorylate a range of substrates, including peptide substrates for S6 kinase, MAP kinase, casein kinase II, glycogen synthase kinase III, protein kinase C,

calcium-calmodulin-dependent protein kinase, or Raytide (a general substrate for tyrosine kinases). It did

phosphorylate peptide substrates for smooth muscle myosin light chain kinase and for protein kinase A and was active in the absence of cofactors. These substrate peptides have at least two hydrophobic and basic amino acids that are positioned two residues toward the N-terminus from a

phosphorylatable serine (Kemp and Pearson, 1990). Amino acid sequence analysis of the phosphorylation site and site-directed mutagenesis will provide more information about the substrate requirements of the kinase. The nuclear localization of the enzyme and its lack of immunoreactivity with rabbit polyclonal antibodies to protein kinase C, RSK or MAP kinases indicated that the 90 kDa kinase was unlike other previously characterized u nases.

This question of its identity was most effectively resolved by purification and microsequencing of the protein. The initial strategic choice to look for activities that were renaturable on nitrocellulose after electrophoresis in SDS and electroblotting proved to have an added benefit of disqualifying those kinases that rapidly lose activity during purification. The great stability of the activity made purification easier. Classical protein purification techniques yielded activity in sufficient purity and

quantity to obtain a microsequence of two internal peptides. One of these peptides produced a significant match with both the Drosophila fsh gene (DROFSHB) (Haynes et al., 1989) and with the human RING3 gene (HUMRING3) (Beck et al.,1991), whereas the other matched only RING3. This result implied that these genes encoded functional kinases, which has not been reported in the literature. Bacterially expressed cDNAs for these genes, however, failed to exhibit

autophosphorylation in vitro. Nevertheless, rabbit

polyclonal antibodies raised against two bacterially

expressed N-terminal fusions of the human and Drosophila cDNAs with glutathione-S-transferase were capable of

immunoprecipitating the original 90 kDa autophosphorylation activity from HeLa nuclear extract, whereas preimmune sera were not. Open reading frames for fsh identify gene

products of molecular weight 110,000 and 205,000 for 5.9 kb and 7.6 kb mRNA transcripts, respectively (Haynes, et al., 1989), which is at variance with the apparent molecular mass of the activity isolated from HeLa nuclear extract.

Post-translational proteolytic cleavage of the fsh gene products to give active kinase may account for this

discrepancy; such proteolytic activation of kinases is well exemplified by protein kinase C eavage to yield

tonstitutively active protein kin se M (Scha

et al., 1990). The open reading frame for RING3, on the other hand, identifies a gene product of molecular weight 83,000 (Beck et al., 1991), which could account for the apparent

molecular masses observed, especially for a phosphorylated gene product (Kozma et al., 1990).

Consistent with the apparently nuclear localization of the HeLa 90 kDa kinase, the RING3 gene possesses a putative nuclear localization sequence (KKKRK, at amino acid 508;

Beck et al., 1991); however, this sequence is absent from fsh and lies upstream from a C-terminal domain of homology (170 amino acids, 72% identity) that the two genes share (Beck et al., 1991). This region of homology contains some elements of a putative kinase catalytic domain.

Subdomain structure

The molecular organization of fsh and RING3 is

puzzling. A putative phosphate-binding motif GXXXXGK

(Moller and Amons, 1985) is located at amino acid residue 809 in fsh, and a putative ATP-binding motif GXGXXG (Hanks et al., 1988) is found at amino acid position 880 in fsh and 557 in RING3. Fifteen residues downstream from the last glycine of GXGXXG in fsh and 11 downstream in RING3 lies a putative catalytic lysine residue, which is in both cases two residues downstream from an alanine that is conserved in many kinases. This sequence in fsh (GAG SVG GVG GAG AAG GGN ASK), is most highly related to the corresponding subdomains of c-mos (GAG gfG sVy kAt yrG vpv AiK; Watson et al., 1982). A run of glutamates 43 residues downstream from the putative catalytic lysine in fsh and 13 downstream in RING3

identifies a putative subdomain III within a C-terminal region of significant sequence homology between fsh and RING3. Another motif, DFE, at position 997 in fsh and 640 in RING3, suggests subdomain VII ( Z FC). An interesting extended run of serines at the C-terminus of the open reading frame may function as an autophosphorylation or autoinhibitory domain (Banerjee, et al., 1990). Despite these sequence similarities, some of the conserved

subdomains that ordinarily occur downstream from the putative DFG motif are missing in fsh and RING3 because the open reading frame ends.

However, lack of certain consensus motifs does not mean that a particular open reading frame cannot encode a kinase (Wu et al., 1990; Maru and Witte, 1991). Strangely, in fact, putative subdomains occur in an appropriate order in the middle of the open reading frame of each gene, in a different region of significant homology between fsh and RING3. Here, a motif YHDIIKXPXXL suggests subdomain VIB; a motif APEF in fsh and AQEF in RING3 suggests subdomain VIII; a motif DWAMXRKL suggests domain IX; a motif YAKM in RING3 suggests subdomain X and HRLAEXXXXXXXXHEQLAA in RING3 suggests subdomain XI. This unusual overall structure is not incompatible with kinase catalytic activity and may have arisen through exon shuffling. Another signal-transducing kinase, S6 kinase II from chicken and Xenopus, possesses two catalytic domains that are homologous to protein kinase C and substrate recognition domains homologous to

phosphorylase kinase. This structure has also been proposed to arise from exon shuffling (Banerjee, et al., 1990). We have been unable to identify other kinases in the GenBank database that exhibit similar organization or that have regions of homology with fsh and RING3 that extend much beyond the individual subdomains discussed here. The failure of GST fusions of fsh and RING3 to show kinase activity in vitro makes sense in light of their unusual subdomain organization. In addition, a possible frameshift mutation in clone e5.8 that produced a truncated transcript would have similarly interrupted the coding region of the putative kinase before essential subdomains had been translated. Nevertheless, enough sequence remained for both fsh and RING3 fusions with GST to generate rabbit antibodies that were immunoreactive with the homologous nuclear kinase from HeLa cells. Translation of full-length fsh and RING3 and mutation of the putative catalytic lysine should unambiguously resolve the question of catalytic function.

Most intriguing is the potential role that these genes may play in development and transformation. In Drosophila, mutations in the fsh locus exhibit a homeotic phenotype.

The gene is required at two stages in development: once during early embryogenesis, where there arises a maternal effect that cannot be rescued by sperm carrying the wild type allele (Forquignon, 1981), and again during pupation, where the effect is zygotic and the numbers of progeny with the mutant phenotype decreases with increasing dose of the wild type allele (Cans et al., 1980). These two periods are temperature-sensitive and at the restrictive temperature, the effect is lethal. At semi-permissive temperatures, heterozygous progeny can reach adulthood, but have a high frequency of organ deficiencies and transformations of the third thoracic segment to the second thoracic segment. This phenotype resembles transformations in the bithorax complex of genes, and indeed, mutations in certain bithorax genes such as trithorax (also called Regulator-of-bithorax) and Ultrabithorax are closely linked to fsh and show

increased severity with hypodosage of wild type fsh (Forquignon, 1981). Trithorax ( Genbank accession number M31617), which is required during embryogenesis and later, requires activation by fsh (Digan et al., 1986) and has a number of putative zinc fingers that indicate it is likely to be a transcription factor (Mazo et al., 1990). In light of this evidence, we propose that fsh is a kinase that phosphorylates trithorax, thereby activating trithorax- directed transcription of developmentally important genes. The N-terminal domain of the open reading frame of trithorax has several phosphorylation motifs of the type

(R/K) (R/K)XX(S/T), which can be recognized by protein kinase A and myosin light chain kinase (Kemp and Pearson, 1990). The substrate specificity of. the 90 kDa kinase described above is not inconsistent with an ability to phosphorylate trithorax.

The biological role of RING3 is less well understood. Its occurrence in the major histocompatibility complex among a set of recently characterized genes that do not appear to have a role in antigen processing suggests that if it is a nuclear kinase in human cells, there may be a human homolog of trithorax that has zinc fingers and that functions as its transcription factor substrate. In this regard, a recent report that maps the breakpoint in chromosomal

translocations of human chromosome 11 has identified an interrupted gene with significant homology to trithorax (Djabali et al., 1992). This t(4; 11) (q21;q23) reciprocal translocation is associated with acute lymphocytic leukemias in children, which may represent a biological consequence of loss-of-function of the human trithorax homolog. In

addition to putative zinc fingers, the homolog has "AT hook" motifs that may be involved in binding to the minor groove of DNA (Tkachuk et al., 1992) In-frame fusions of this human homolog, called the ALL-1 (Gu et al., 1992), to the AF-4 gene as a result of the reciprocal transiocation may give rise to a new protein with oncogenic transforming potential (Tkachuk et al., 1992; Gu et al., 1992).

Alternatively, deletion of important regulatory sites of phosphorylation in ALL-1 by the translocation may account in part for the oncogenic effects. In such a case, we would predict by analogy that loss-of-function mutations in RINC3 might be associated with leukemias as well. Further

characterization of the relationship between RINC3 and the trithorax homolog might illuminate the genomic organisation of RINC3 within the major histocompatibility locus and its expression in T-cells (Beck et al., 1991), as well as the emerging parallel relationships between the important players in Drosophila development and human cancer.

Example 1

Characterization of Transcription Factor Nuclear Kinases

Materials

Dextran sulfate (average molecular weight 500,000), insulin, glucagon, acidic fibroblast growth factor, reactive green-19 and Cibacron Blue 3GA (Type 3000) agarose were from Sigma. Sodium orthovanadate, Tris, trichloroacetic acid, MgCl₂ and NaOH were from Fisher. Disodium ATP and NP-40 were from Pharmacia LKB. Guanidine hydrochloride was from Fluka. Phytohemagglutinin-M, phorbol 12-myristate

13-acetate, calcium ionophore A23187, human interleukin-2, (70-deacetyl-7β-[-(γ-N-methylpiperazino]-butyryl)-forskolin dihydrochloride and rat brain protein kinase C were from Calbiochem. Wheat germ extract and rabbit reticulocyte extract systems for in vitro translation were from Promega. Raytide and recombinant p43 were from Oncogene Science (Uniondale, NY) . In vitro phosphorylation substrates were from Peninsula Laboratories (Belmont, CA). [α - and γ

-³² P]ATP were from New England Nuclear. Nitrocellulose membranes were from Micron Separations (Westboro, MA) and polyvinyl difluoride membranes (PVDF) were from Millipore.

Recombinant protein A-agarose was from Repligen (Cambridge,

MA). Phosphoceliulose resin (P-11) was from Whatman.

Polyacrylamide monomer, bisacrylamide and ammonium

persulfate were from National Diagnostics (Manville, NJ).

Cell culture media, calf serum, monoclonal antibody MAb 1.9 against rat brain protein kinase C and TEMED were from

GIBCO-BRL. Fetal bovine serum for Jurkat cell culture was from Whittaker Bioproducts (Walkersville, MD). Cell lines

(HeLa, ATCC CCL 2; A431, ATCC CRL 1555; Jurkat, ATCC TIB

152; HUT78, ATCC TIB 161; Mvl Lu. ATCC CCL64; CHO DUKX B1,

ATCC CRL 901 O; CV-1, ATCC CCL T O) and COS-7. ATCC CRL 1551)) were obtained from ATCC and cultured as advised Jurkat clone E6-1 (Weiss et al., 1984) and HUT78 (Gazdar et al., 1980) were obtained through the AIDS Research and Reference Reagent Program, Division of AIDS, NIAID, NIH. 3T3-L1 fibroblasts (ATCC CL 173) were the kind gift of Michael P. Czech, University of Massachusetts Medical Center, and were differentiated according to Rubin et al. (1978). An

Epstein-Barr virus-transformed human B-cell line (32D clone 13) (FitzGerald et al., 1991) was obtained from Joel

Greenberger, University of Massachusetts Medical Center. Human peripheral blood lymphocytes from normal, healthy volunteers, were the kind gift of John Sullivan, University of Massachusetts Medical Center. Rabbit polyclonal antisera to purified GST fusion proteins were prepared by Berkeley Antibody Company (Richmond, CA). Platelet-derived growth factor was the kind gift of Roger J. Davis, University of Massachusetts Medical Center. Rabbit anti-rat MAP kinase (erk-1 C-terminal 333-367 amino acids) polyclonal antibody was from Upstate Biotechnology (Lake Placid, NY) and

anti-RSK rabbit antibody was the kind gift of John Blenis, Harvard Medical School. HeLa nuclear extract was prepared with 1 mM sodium vanadate and 50 mM β-glycerophosphate after the method of Dignara et al. (1983). All other reagents were from Sigma.

Renaturation assay

Proteins were separated in polyacrylamide gels according to Laemmli (1970) and electroblotted to nitrocellulose or PVDF membranes. Blotting was performed in 25 mM Tris, 192 mM glycine, 10% methanol at 100 mA for one hour at room

temperature. The blots were incubated in 7 M guanidine HCl, 50 mM Tris-HCI pH 8.5,50 mM DTT and 2 mM EDTA for one hour at room temperature to denature blotted proteins, then transferred to ice-cold buffer containing 100 mM NaCl, 50 mM Tris-HCl, 2 mM DTT, 2mM EDTA and 0.1% NP-40. Proteins were renatured overnight at 4°C in this buffer. The blots were then blocked with 5% dextran sulfate in the same buffer for one hour at room temperature; then incubated for one hour at room temperature with 0.15 mCi/ml[γ -³²P]ATP (25 nM) in 30 mM Tris-HCl pH 8.0, 30 mM MgCl₂,2 mM DTT, 0.2 mM EDTA and

0.1% NP-40; and washed sequentially for ten minutes each with: 50 mM Tris-HCl pH 8.0, 0.1% NP-40 in 50 mM Tris-HCl pH

8.0,50 mM Tris-HCl pH 8.0, 1 M NaOH, 10% acetic acid and then water, after Ferrell and Martin (1989). The blots were then autoradiographed at -80°C.

For phosphorylation of protein or peptide substrates in solution, nitrocellulose strips of immobilized, renatured kinase were excised and placed in Eppendorf tubes with 30 mM Hepes pH 8.0, 30 mM MgCl₂ (or 1 mM MnCl₂), 2 mM DTT, 0.5 mM EDTA, 0.1% NP-40, 10 μM disodium ATP, 0.15 mCi/ml [γ-³²P]ATP and 1 mg/ml myelin basic protein or 1 mg/ml peptide

substrate in a final volume of 10 μl. Reactions were allowed to proceed for one hour at 30°C, whereupon protein phosphorylation reactions were quenched with SDS sample buffer and heated to 100°C. Phosphorylated protein samples were separated by SDS PAGE and visualized by Coomassie Blue staining and autoradiography. Peptide phosphorylation reactions were quenched with 4 volumes of ice-cold 10% phosphoric acid. Phosphorylated peptide samples were applied to 0.1 ml columns of phosphoceliulose that had been pre-equilibrated with 0.5% phosphoric acid. Columns were washed extensively with 0.5% phosphoric acid and

radioactivity that remained in the columns was quantified by Cerenkov counting. Phoschc ami noac id analysis

HeLa nuclear extract was fractionated on

phosphoceliulose and proteins that eluted at 0.3M and 1.0M NaCl were precipitated with TCA, washed with acetone, solubilized in SDS, separated on an 8% polyacrylamide gel, electrobiotted to PVDF, denatured in guanidine hydrochloride and renatured as described. PVDF strips were excised, placed in Eppendorf tubes and incubated with [γ-³² P]ATP without unlabelled ATP as described. The autophosphorylated kinases were then hydrolyzed in 6N HCl at 110ºC for one hour, according to Lewis et al. (1990). The acidic solution was diluted five-fold with water : methanol (50:50; v:v) and lyophilized twice. The lyophilization residue was dissolved in 30% formic acid and spiked with phosphoaminoacid

standards. Equal amounts of radioactivity (about 2000 cpm by Cerenkov counting) were applied to a cellulose thin layer plate. Phosphoaminoacids were separated by electrophoresis in glacial acetic acid : pyridine : water (10:1:189; v:v:v) at pH 3.5 for two hours at 1000V and visualized with

ninhydrin. Radioactive spots were visualized by

autoradiography.

Protein purification

HeLa nuclear extract (100 ml) was applied to a column of reactive green-19 resin (40 ml bed volume, equilibrated with buffer A (20 mM Hepes pH 7.0, 50 mM NaCl, 50 mM

β-glycerophosphate, 1 mM DTT, 0.2 mM EDTA, 0.02% NaN₃ and 0.1% NP-40). All steps were conducted on ice or at 4°C.

The flow-through was discarded and the column was washed extensively with buffer A and then eluted in batch with buffer B (Buffer A supplemented with 20 mM disodium ATP, 20 mM EDTA and 0.5 M NaCl, pH 7.0). Ammonium sulfate was added to the eluate to 50% (w/v) at 4°C over the course of an hour, whereupon the suspension was centruged at 10,000 x g for 45 minutes at 4ºC. The pellet was recovered dissolvel in buffer C (Buffer A supplemented with 10 mM MnCl₂, pH 8.0) and desalted on Sephadex G-25 that had been equilibrated with buffer C. The desalted protein was applied to a column of Cibacron Blue 3GA agarose (Type 3000) that had been equilibrated with buffer C. The flow-through was discarded, the column was washed extensively with buffer C, and eluted in batch with buffer D (Buffer A supplemented with 20 mM EDTA and 0.15 M NaCl, pH 8.0). The eluate was diluted 1:1 with buffer E (Buffer A with no added NaCl, pH 8.0) and applied to phosphoceliulose that had been equilibrated with buffer F (Buffer A at pH 8.0). The phosphoceliulose was washed extensively and eluted in batch with 0.6 M NaCl in buffer F. The eluate was precipitated with trichloroacetic acid (10% final), washed with acetone and solubilized in SDS sample buffer. Proteins were resolved by SDS PAGE in 8% polyacrylamide, blotted to nitrocellulose and visualized with Ponceau S. The band corresponding to the

autophosphorylation activity of the 90 kDa kinase was excised and digested with trypsin. Tryptic peptides were resolved by HPLC and microsequenced.

Results

Autophosphorylation of a 90 kDa range kinase multiplet

Jurkat cells were treated with 50 μM forskolin for 30 minutes and nuclear extracts of forskolin-stimulated or control cells were fractionated on phosphoceliulose.

Proteins were separated by polyacrylamide gel

electrophoresis in sodium dodecyl sulfate (SDS-PAGE), transferred to nitrocellulose, denatured in guanidine

hydrochloride and assayed after renaturation. The assay revealed one major and two minor autophosphorylating kinases. The major activity, a multiplet with an atparen

mobility of 90 kDa (spanning 34.5 - 92.7 kDa), was greater in extracts from forskolin-treated Jurkat cells than in extracts from control cells (Figure 1A). Phosphoceliulose chromatography resolved the multiplet into forms of slightly lower and higher apparent mobility, which we interpret to arise from differences in protein phosphorylation. The best separation and recovery of forms was obtained by eluting the column first with 0.3 M NaCl and then with 1.0 M NaCl.

Autophosphorylation of the species that eluted at 0.3 M NaCl increased sharply within 6 minutes of stimulation and gradually declined through 30 minutes (Figure 1B).

Autophosphorylation of the species that eluted at higher salt did not respond significantly to forskolin stimulation. If extracts were prepared without the phosphatase inhibitors sodium vanadate (1 mM) and β-glycerolphosphate (50 mM) (Yu et al.,1987), or if extracts were briefly warmed to 37°C in the presence of 5 mM magnesium chloride, the 90 kDa

autophosphorylation activity from stimulated cells was reduced to the levels in control cells, whereas other species (e.g., 60 kDa) were unchanged. Taken together, these results indicate that this kinase is reversibly phosphorylated and increases its autophosphorylation in response to an agonist of protein kinase A.

Other agents that affect Jurkat cell metabolism, such as phytohemagglutinin, phorbol myri state acetate, calcium ionophore A23187, serum and interleukin-2, alone or in combination, had little or no effect on the

autophosphorylation activity of the 90 kDa multiplet in

Jurkat cells or peripheral blood lymphocytes over one hour or over four days. Other conditions, such as herpesvirus infection, heat shock, or exposure to hydrogen peroxide, sodium periodate, socium vanadate or cadmium ion, were similarly ineffective

The 90 kDa kinase appears to be ubiquitous, as we have found well-resolved multiplets of similar apparent molecular mass in human epithelial carcinoma (HeLa), human epidermoid carcinoma (A431), human T-cell leukemia (Jurkat),

Epstein-Barr virus-transformed human B-cell (32D cl3), human T-cell lymphoma (HUT78), mouse embryo fibroblast and

adipocyte (3T3-L1), mink lung (MvlLu), Chinese hamster ovary (CHO), African green monkey kidney (CV-1 and COS-7) cell lines, although the activities differed in their

chromatographic behavior on phosphoceliulose and varied slightly in their apparent mobility. Depending on the cell type, up to ten additional autophosphorylating species, with apparent molecular masses as low as >20 kDa, were also detectable. It is not known whether these activities arise from transcripts of homologous genes, or whether they are unrelated to each other. In some cell lines, such as A431, an 84 kDa multiplet autophosphorylation activity (78.4 - 89.2 kDa) was both serum-inducible in serum-starved cells and forskolin-inducible, and required higher salt

concentrations to elute from phosphoceliulose (Figure lC). Neither insulin nor glucagon treatment of 3T3-L1 adipocytes increased this autophosphorylation activity, nor did

forskolin or platelet-derived growth factor treatment of CHO cells, nor did fibroblast growth factor treatment of 3T3-L1 fibroblasts, nor did epidermal growth factor or heat shock treatments of A431 cells (Figure IC). These changes to autophosphorylation activity were remarkably stable. Major differences in activity between kinase prepared from control or stimulated cells reproducibly survived cell lysis, fractionation, boiling in SDS, gel electrophoresis,

electroblotting, denaturation in guanidine hydrochloride, renaturatio n overnight in buffer at 4°C and exposure to

sodium hydroxide.

The blotted, renatured kinases were immobilized and no t in contact with potential protein substrates in solution, so only autophosphorylation was observable. These activities represent autophosphorylations, not merely binding of ATP to renatured proteins, because when blots were probed in parallel with [α-³²P]ATP and [ γ-³²P] ATP, only background radioactivity was associated with the membranes probed with [α-³² P]ATP, whereas discrete bands were revealed with the membranes probed with (γ-³² P] ATP. The renatured kinase required magnesium ion for activity. Blots were routinely incubated in 30 mM magnesium chloride; manganese ion, which was tested from 1 mM to 30 mM, was a poor substitute.

Physical properties of the 90 kDa kinase

The autophosphorylation activities present in nuclear and cytosolic extracts of HeLa cells were compared. A

homogenate of HeLa cells was prepared in the presence of phosphatase inhibitors as described above and centrifuged at 100,000 x g for one hour at 4°C. The protein concentration of the supernatant (S-100) was determined and equal amounts of protein from the S-100 and nuclear extract (also cleared of particulate residue) were subjected to SDS-PAGE.

Proteins were electrobiotted to nitrocellulose, renatured and probed with ( γ - ³²P]ATP. As shown in Figure 2, an autophosphorylation signal (91.0 - 98.8 kDa) in nuclear extract is virtually absent from cytosolic extract. Other autophosphorylation activities resolved on this blot, however, appear to be roughly equally represented in nuclear and cytosolic fractions (70.9, 64.3, 52.8 and 37.9 kDa).

This result suggested that the 90 kDa kinase is

predominantly nuclear in localization. HeLa nuclear extract was fractionated on Seohacryl 3- 200 at physiological ionic strength, and proteins were resolved by PAGE, blotted and renatured. The 90 kDa multiplet autophosphorylation activity was excluded from the column along with activities of apparent molecular mass 51.0, 36.4 kDa and a multiplet at 59.2 - 61.3 kDa, whereas other activities of apparent molecular mass 68.6 and <20 kDa were included (Figure 3). This result indicates that under these conditions, the 90 kDa kinase multiplet associates with other proteins in a complex with an apparent molecular mass in excess of 1,500 kDa. The identity of these other proteins and biological significance of their association are unknown.

In vitro phosphorylation of protein and peptide substrates

Phosphocellulose-purified, renatured 90 kDa kinase was tested for its substrate specificity with several different protein and peptide substrates: S6 peptide (RRLSSRA; Pelech et al.,1986), MAP kinase substrate peptide (APRTPGGRR;

Sanghera et al., 1990), casein kinase II substrate

(RRREEETEEE; Kuenzel and Krebs,1985), protein kinase C serine-25 substrate (RFARKGSLRQKNV; House and Kemp, 1987), calmodulin-dependent kinase substrate (PLSRTLSVSS-NH2;

Pearson et al., 1985), smooth muscle myosin light chain kinase substrate (LLRPQRATSNVFS-NH₂; Kemp et al., 1983), Kemptide, a substrate for protein kinase A (LRRASLG;

,^,Kemptide", Kemp et al.,1977), Raytide (a tyrosine kinase substrate), histones H1, H2, H2B, H3 and myelin basic protein. The results are shown in Table I: Table I

In vitro phosphorylation of

peptides by renatured 90 kDa kinase

1 Phosphorylated with the catalytic subunit of protein kinase A.

2 Phosphorylated with recombinant p43^v-abl

The partially purified kinase phosphorylated the peptide substrates for myosin light chain kinase and protein kinase A, and myelin basic protein. Myelin basic protein is a useful in vitro substrate for MAP kinases, but the MAP kinase substrate peptide (which includes the aminoacid residues that MAP kinases ordinarily phosphorylate) was not phosphorylated here. The catalytic subunit of protein kinase A was used as a positive control for Kemptide

phosphorylation and recombinant p43 was used as a positive control for Raytide phosphorylation. These

suggest that the 90 kDa kinase has a narrow substrate specificity and with respect to its substrate preferences is unlike other known kinases.

HeLa nuclear extracts were fractionated on

phosphoceliulose as described. Proteins that eluted at 0.3 M and 1.0 M NaCl were blotted to nitrocellulose, renatured and tested for their ability to phosphorylate myelin basic protein that was supplied in solution with magnesium

chloride and [ -³²P]ATP. The kinase that eluted at the lower salt concentration, which is the inducible form in

Jurkat cells and which has a slightly slower apparent mobility (Figure 1A), showed greater substrate-directed phosphorylation activity than the species that eluted at the higher salt concentrations (Figure 4). This result suggests that induced autophosphorylation activity is linked to increased substrate-directed phosphorylation activity. To refine this result, HeLa nuclear extract was more narrowly fractionated on phosphoceliulose. Eluates were blotted to nitrocellulose and renatured either for autophosphorylation or for phosphorylation of myelin basic protein supplied in solution. Autophosphorylation of the slower mobility form was correlated with phosphorylation of myelin basic protein, whereas autophosphoryation of the faster mobility form was not (Figures 5A and 5B). We interpret this collection of results to mean that the 90 kDa kinase is phosphorylated by a kinase kinase to generate a hyperphosphorylated form that is the active participant in signal transduction.

Phosphoaminoacid analysis

The 90 kDa kinase was blotted to a polyvinyldifluoride membrane and autophosphorylated with [γ -³²P ]ATP. The immobilized protein was hydrolyted in 6N HCl, and the phosphoaminoacids were separved by electrophoresis on silica. Autoradiography revealed that radioactivity was associated in equal measure with ninhydrin-visuaiized standards for phosphoserine (pS) and phosphothreonine (pT). Mo radioactivity was associated with a phosphotyrosine (pY) standard (Figure 6A). Myelin basic protein that had been phosphorylated in vitro with renatured 90 kDa kinase was solubilized in SDS and resolved by PAGE. Phosphorylated myelin basic protein was visualized with Coomassie brilliant blue, excised from the gel and acid-hydrolyzed as before. Radioactivity was associated only with pS, not with pT or pY (Figure 6B). As a positive control for tyrosine

phosphorylation, recombinant p43 was incubated with myelin basic protein under the same conditions. In this case, pY alone was observed (Figure 6B).

Purification and microsequencing of the 90 kDa multiplet

The 90 kDa nuclear kinase shares some of the characteristics of signal-transducing kinases such as pp90

pp42^mapk/Pp44^mapk and protein kinase C (Chen et al., 1992; Chung et al., 1991; Leach et al., 1989). Western blots of partially purified 90 kDa kinase were probed with rabbit polyclonal antibodies to RSK, MAP kinase and protein kinase C. Protein kinase C from rat brain was used as a positive control. These antisera displayed no detectable

immunoreactivity with the partially purified kinase above background.

From these results, we suspected that this 90 kDa protein had not been studied previously. We therefore undertook to purify it and obtain microsequence information. HeLa nuclear extract was prepared in the presence of 50 mM β-glycerophosphate and 1 mM sodium vanadate. The

autophosphorylation activity provided the assay for purification. The partially p rified kinase was resolved by SDS-PAGE, biotted to nitrocellilose and digested with trypsin. Tryptic peptides were separated by HFLC and two peptides were microsequenced. The first of these

(DSNPDEIEIDFETLKPTTL) was used to probe the GenBank

database; a statistically significant match was obtained to the female sterile homeotic (fsh) gene of Drosophila

(DROFSHB; accession number M23222, Haynes et al, 1989).

There were 18/19 identities (95% match, with a S/T

substitution at residue 17). A statistically significant match was also obtained with the RING3 gene of humans

(HUMRING3; accession number M80613, Beck et al., 1991)

(17/19 identities (89% match) with a E/D substitution at position 5 in addition to the S/T substitution of fsh). The second peptide (AVHEQLAALSQAP ) gave a statistically

significant match to RING3 (12/13 identities (92%) with a A/G substitution at position 12), but did not match fsh.

The first peptide lies within the C-terminal region of highest homology between fsh and RING3, whereas the second peptide lies outside of this region, in the N-terminal direction (Beck et al., 1991). These results were

surprising because neither RING3 nor fsh have been reported to be kinases. However, inspection of the protein sequence of these genes reveals several correctly spaced amino acid motifs that are consistent with the genes encoding a protein kinase (Hanks et al., 1988; Hanks, S.,1991), although of unusual organization. It was important to determine whether the gene products exhibit kinase activity.

Clones that encode the cDNAs for fsh and RING3 were obtained from the authors of the original published reports. The cDNA for fsh was a set of overlapping clones, one of which, e5.16 (Haynes et al, 1989), included the putative catalytic region .of interest. This cDNA contained a site for Ncol that provided an

methionine at position 785 o f the open reading frame, almost 100 ammo acids upstream from the putative GXGXXG , and ran through the stop codon at amino acid 1107. The fragment was fused in-frame to the 5' untranslated region of the 3-globm gene and subjected to in vitro transcription/translation in both rabbit reticulocyte and wheat germ extracts as described. Products were separated by SDS-PAGE, blotted to

nitrocellulose, denatured, renatured and probed with

[ -32P]ATP as described above. No radioactivity was associated with renatured bands in excess of the signals seen in extracts primed with control viral mRNA. Several kinases in the translation reaction were capable of

renaturation by this assay and may have obscured signals from the kinase in question. The fsh fragment was also expressed as a histidine-tagged protein, a T7

antibody-tagged protein and as a glutathione-S-transferase (GST) fusion protein. The bacterially expressed fusion protein showed little or no activity. The same was true for fusion constructs of RING3, which had a convenient Ncol site to yield an in-frame methionine at amino acid 347 of the open reading frame, just over 200 residues upstream from the putative GXGXXG and continued through the run of serines to the stop codon at 755. The failure of these bacterially translated cDNAs to demonstrate kinase activity in vitro can probably be ascribed to the usual shortcomings with

post-translational processing in prokaryotes.

The GST fusion proteins with the fsh and RING3

fragments were purified from lysates of induced bacteria, purified on glutathione agarose and injected into rabbits for the production of antibody The fusion constructs were of the correct electrophore

on SDS gels. Immune sera were screened for the

immunoprecipitate the original kinase activity from HeLa nuclear extract Botn immune sera to GST-fsh and to GST-FINC3 immunoprecipitated the renaturable autophosphorylation activity with apparent moiecuiar weight 90 kDa, whereas none of the preimmune sera immunoprecipitated this activity. This result strongly supports the interpretation that fsh and RING3 are kinases, and that the microsequenced peptides are derived from a HeLa gene product that is homologous to the Drosophila and human T-cell cDNAs.

Rabbit immune sera were incubated with GST-agarose to remove antibodies against GST epitopes. Polyclonal

antibodies against RING3 and fsh epitopes were then purified by antigen affinity chromatography on Affigel columns of each purified GST fusion protein. Purified antibodies were assayed by immunoblot. Indirect immunofluorescence of HeLa cells with purified polyclonal antibody and with

fluorescein-conjugated second antibody shows that RING3 is localized to the nucleus (Figure 8). This result is in agreement with Figure 2, which suggested that the kinase is exclusively nuclear, and with the presence of a putative nuclear localization signal in the open reading frame of the RING3 cDNA. Antibody was excluded from nucleoli, consistent with patterns of nuclear staining for proteins that are involved in transcription. Purified antibodies against fsh epitopes showed punctate cytoplasmic staining and

perinuclear localization, but did not show typical nuclear staining.

Claims

WHAT I S CLA I MED I S :

1. A method for regulating transcription of a gene of interest, which method comprises: modulating the activity of a kinase that activates transcription of the gene and is characterized as follows: (i) is substantially exclusively intranucleariy localized; (ii) is capable of autophosphorylation; (iii) is selectively bindable with antibodies raised against GST-fsh and

GST-RING3; (iv) having a molecular weight of from about 82.5 to about 92.7 kilodaltons; and (v) includes peptide

sequences

Asp-Ser-Asn-Pro-Asp-Glu-Ile-Glu-Ile-Asp-Phe-Glu-Thr-Leu- Lys-Pro-Thr-Thr-Leu and Ala-Val-His-Glu-Gln-Leu-Ala-Ala- Leu-Ser-Gln-Ala-Pro.

2. The method of Claim 1 wherein modulating the kinase activity comprises stimulating the phosphorylation activity of the kinase.

3. The method of Claim 2 wherein stimulating the kinase induces transcription of the gene of interest.

4. The method of Claim 3 wherein the kinase activates a transcription factor for initiating transcription of the gene of interest.

5. The method of Claim 2 wherein stimulating the kinase comprises preventing inhibition of transcription.

6. The method of Claim 1 wherein modulating the kinase activity comprises mhiciting the phosphorylation activity of the kinase.

7. The method of claim 6 wherein the

a transcription factor for the gene of interest

8. The method of Claim 1 which comprises regulating the ALL-1 gene.

9. A method for regulating the transcription of a gene of interest in a cell which comprises contacting said cell with a compound that specifically binds with a cell surface receptor so as to modulate by signal transduction the activity of a kinase characterized as follows: (i) is substantially exclusively intranucleariy localized; (ii) is capable of autophosphorylation; (iii) is selectively bindable with antibodies raised against GST-fsh and

GST-RING3; (iv) has a molecular weight of from about 82.5to about 92.7 kilodaltons; and (v) includes peptide sequences Asp-Ser-Asn-Pro-Asp-Glu-Ile-Glu-Ile-Asp-Phe-Glu-Thr-Leu-Lys- Pro-Thr-Thr-Leu and Ala-Val-His-Glu-Gln-Leu-Ala-Ala-Leu-Ser- Gln-Ala-Pro.

10. The method of Claim 9 wherein the compound stimulates a cell surface receptor of a type selected from the group consisting of peptide growth factor receptors, hormone receptors, ion channel receptors, serpentine cell surface receptors and immunophilins.

11. The method of Claim 9 wherein modulating the kinase activity comprises stimulating the phosphorylation activity of the kinase.

12. The method of Claim 11 wherein stimulating the kinase induces transcription of the gene interest.

13. The method of Claim 11 wherein stimulating the kinase comprises preventing inhibition of transcription

14. The method of Claim 9 wherein modulating the kinase activity comprises a method for inhibiting

phosphorylation activity of the kinase.

15. The method of Claim 9 which comprises stimulating the signal transduction function of said kinase by

contacting said cell with forskolin or an analog or

derivative thereof.

16. A kinase characterized as follows: (i) is substantially exclusively intranucleariy localized; (ii) is capable of autophosphorylation; (iii) is selectively

bindable with antibodies raised against GST-fsh and

GST-RING3; (iv) has a molecular weight of from about 82.5 to about 92.7 kilodaltons; and (v) includes peptide sequences Asp-Ser-Asn-Pro-Asp-Glu-Ile-Glu-Ile-Asp-Phe-Glu-Thr-Leu-Lys- Pro-Thr-Thr-Leu and Ala-Val-His-Glu-Gln-Leu-Ala-Ala-Leu-Ser- Gln-Ala-Pro.

References

Adams, S. R., Harootunian, A. T., Buechler, Y. J., Taylor, S. S., and Tsien, R. Y. (1991). Fluorescence ratio imaging of cyclic AMP in single cells. Nature 349, 694-697.

Ahn, N. G., Weiel, J. E., Chan, C. P., and Krebs, E. G.

(1990). Identification of multiple epidermal growth -factor- stimulated protein serine/threonine kinases from Swiss 3T3 cells. J. Biol. Chem. 265, 11487-11494.

Ahn, N. G., and Krebs, E. G. (1990). Evidence for an epidermal growth factor- stimulated protein kinase cascade in Swiss 3T3 cells. J. Biol. Chem. 265,11495-11501.

Anderson, N. G., Mailer, J. L., Tonks, N. K., and Sturgill, T. W. (1990). Requirement for integration of signals from two distinct phosphorylation pathways for activation of MAP kinase. Nature 343, 651-653.

Baeuerle, P. A., and Baltimore, D. (1988a). Activation of DNA-binding activity in an apparently cytoplasmic precursor of the NF-kB transcription factor. Cell 53, 211-217.

Baeuerle, P. A., and Baltimore, D. (1988b). IkB: a specific inhibitor of the NF-kB transcription factor. Science 242, 540-546.

Banerjee, P., Ahmad, M.F., Grove, J.R.,

insulin/mitogen-activated 70-kDa S6 protein kinase. Proc Nat'l. Acad. Sci. USA 87:8550-3554. Bec k S., Hanson, I., Kelly, A., Pappin, D. J. C . , and Trowsdale, J. (1991). A homologue of the Drosophila female sterile homeotic (fsh) gene in the class II region of the human MHC. DNA Seq. 2, 203-210.

Blenis, J. (1991). Growth-regulated signal transduction by the MAP kinases and RSKs. Cancer Cells 3, 445-449.

Bohmann, D. (1990). Transcription factor phosphorylation: a link between signal transduction and the regulation of gene expression. Cancer Calls 2, 337-344.

Boulton, T. G., Yancoupoulos, G. D., Gregory, J. S.,

Slaughter, C, Moomaw, C, Hsu, J., and Cobb, M. H. (1991). An insulin-stimulated protein kinase similar to yeast kinases involved in cell cycle control. Science 249, 64-67.

Chen, R.-H., Sarnecki, C, and Blenis, J. (1992). Nuclear localization and regulation of erk- and rsk-encoded protein kinases. Mol. Cell. Biol. 12, 915-927.

Chung, J., Pelech, S., and Blenis, J. (1991).

Mitogen-activated Swiss mouse 3T3 RSK kinase I and II are related to pp44 ^mpk from sea star oocytes and participate in the regulation of pp90^rsk activity. Proc. Nat'l. Acad.

Sci. USA 88, 4981-4985.

Cicirelli, M. F., Pelech, S. L., and Krebs, E. G. (1988). Insulin and progesterone activate a common synthetic

ribosomal protein S6 peptide kinase in Xenopus oocytes.

FE3S Lett. 241, 195-201.

Czech, M. P., Klarlund, J. K Yacaloff, K. A., Bradford, A. P., and Lewis, R. E. (1988) receptor signalling.

Ac ti va t i on o f mul tiple se ri ne ki nase s Chem 263 1 1017 - 1 1020 .

Devoto, S.H., Mudryj, M., Pines, J., Hunter, T. . and Nevins,

J.R. (1992). A cyclin A-protein kinase complex possesses sequence-specific DNA binding activity: p33^cdk2 is a

component of the E2F-cyclin A complex. Cell 68, 167-176.

Digan, M. E., Haynes, S. R., Mozer, B. A., Dawid, I. B., Forquignon, F., and Gans, M. (1986). Genetic and molecular analysis of fs(l)h, a maternal effect homeotic gene in

Drosophila. Dev. Biol. 114, 161-169.

Dignam, J. D., Lebovitz, R. M., and Roeder, R. G. (1983). Accurate transcription by RNA polymerase II in a soluble extract from isolated mammalian nuclei. Nucl. Acids Res. 11, 1475-1489.

Dikstein, R., Heffetz, D., Ben-Neriah, Y., and Shaul, Y.

(1992). c-Abl has a sequence-specific enhancer binding activity. Cell 69, 751-757.

Djabali, M., Selleri, L., Parry, P., Bower, M., Young, B. D., and Evans, G. A. (1992). A trithorax-like gene is interrupted by chromosome 11q23 translocations in acute leukemias. Nature Genet. 2, 113-118.

Feaver, W. J., Gileadi, O., Li, Y., and Kornberg, R. (1991). CTD kinase associated with yeast RNA polymerase II

initiation factor b. Cell 67, 1223-1230.

Farrell, J. E., Jr., and Martin, G. S. (1989) Thrombin stimulates the activities of multiple previously unidentified protein kinases in platelets. J

264, 20723-20729.

Forquignon, F. (1 981). A maternal effect mutation leading to deficiencies of organs and homeotic transformations in the adults of Drosophila. Wilhelm Roux's Arch. Dev. Biol. 190, 132-138.

FitzGerald , T., Santucci, M. A., Das, I., Kase, K ., Pierce, J.H., and Greenberger, J.S. (1991). The v-abl, c-fms, or v- myc oncogene induces gamma radiation resistance of

hematopoietic progenitor cell line 32D cl 3 at clinical low dose rate. Internal. J. Rad. Oncology 21, 1203-121 0.

Cans, M., Forquignon, F., and Masson, M. (1980). The role of dosage of the region 7D1-7D5-6 of the X chromosome in the production of homeotic transformations in Drosophila

melanogaster. Genetics 96, 887-902.

Gazdar, A. F., Carney, D. N., Bunn, P. A., Russell, E. K., Jaffe, E. S., Schechter, G. P., and Guccian, J. G. (1980). Mitogen requirements for the in vitro propagation of

cutaneous T-cell lymphomas. Blood 55, 409-417.

Gu, Y., Nakamura, T., Alder, H., Prasad, R., Canaani, O., Cimino, G., Croce, C. M., and Canaani, E. (1992). The t(4;11) chromosome translocation of human acute

leukemias fuses the ALL-1 gene, related to Drosophila trithorax, to the AF-4 gene. Cell 71, 701-708.

Hanks, S. K. (1991). Eukaryotic protein kinases. Curr.

Biol. 1, 369-383. Hanks, S. K., Quinn, A. M., and Hunter, T. (1988). The protein kinase family: Conserved features and deduced phyl-ogeny of the catalytic domains. Science 241, 42-52.

Hao, Q.-L., Ferris, D. K., White, G., Heisterkamp, N., and Groffen, J. (1991). Nuclear and cytoplasmic location of the FER tyrosine kinase. Mol. Cell. Biol. 11, 1180-1183.

Haynes, S. R., Mozer, B. A., Bhatia-Dey, N., and Dawfd, 1. B. (1989). The Drosophila fsh locus, a maternal effect homeotic gene, encodes apparent transmembrane proteins.

Dev. Biol. 134, 246-257.

House, C, and Kemp, B. E. (1987). Science 238,1726.

Hunter, T., and Karin, M. (1992). The regulation of

transcription by phosphorylation. Cell 70, 375-387.

Ingham, P. W. (1985). Genetic control of the spatial pattern of selector gene expression in Drosophila. Cold Spring Harbor symp. quant, biol. 50, 201-208.

Jackson, S. P. (1992). Regulating transcription factor activity by phosphorylation. Trends Cell Biol. 2, 104-108.

Jackson, S. P., MacDonald, J. J., Lees-Miller, S., and

Tjian, R. (1990). GC box-binding induces phosphorylation of Spl by a DNA-dependent protein kinase. Cell 63, 155-165.

Kemp, B. E., Pearson, R. B., and House, C. (1983). Role of basic residues in the phosphorylation of synthetic peptides by myosin light chain kinase. Proc. Nat'l. Acad. Sci. USA 80, 7471-7475. Kemp, B. E., and Pearson, R. B. (1990). Protein kinase recognition sequence motifs. Trends Biochem. Sci. 15, 342- 346.

Kemp, B. E., Graves, D. J., Benjamini, E., and Krebs, E. G. (1977). Role of multiple basic residues in determining the substrate specificity of cyclic AMP-dependent protein kinase. J. Biol. Chem. 252, 4888-4894.

Kipreos, E. T., and Wang, J. Y. J. (1992). Cell

cycle-regulated binding of c-Abl tyrosine kinase to DNA. Science 256, 382-385.

Kozma, S. C, Ferrari, S., Bassand, P., Siegmann, M., Totty, N., and Thomas, G. (1990). Cloning of the mitogen-activated S6 kinase from rat liver reveals an enzyme of the second messenger subfamily. Proc. Nat'l. Acad. Sci. USA 87,

7365-7369.

Kuenzel, E. A., and Krebs, E. G. (1985). A synthetic peptide substrate for casein kinase 11. Proc. Nat'l. Acad. Sci. USA 82,737-741.

Laemmli, U. K. (19707). Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227,680-685.

Leach, K. L., Powers, E. A., Rum, V. A., Jaken, S., and Kaufmann, S. (1989). Type 3 protein kinase C localization to the nuclear envelope of phorbol ester-treated NIH 3T3 cells. J. Cell Biol. 109, 685-695.

Lewis, R. E., Wu, G. P., MacDonald, R. C., and Czech, M. P (1990). Insulin-sensitive phosphorylation of serine 1293/1294 on the human insulin receptor by a tightly

associated serine kinase. J. Biol. Chem. 265, 947-954.

Lu, H., Zawel, L., Fisher, L., Egly, J.-M., and Reinberg, D.

(1992). Human general transcription factor IIH

phosphorylates the C-terminal domain of RNA polymerase II. Nature 358, 641-645.

Maru, Y., and Witte, O.N. (1991). The BCR gene encodes a novel serine/threonine kinase activity within a single exon. Cell 67, 459-468.

Mazo, A. M., Huang, D.-H., Mozer, B. A., and Dawid, I. B. (1990). The trithorax gene, a trans-acting regulator of the bithorax complex in Drosophila, encodes a protein with zinc-binding domains. Proc. Nat'l. Acad. Sci. USA 87, 2112- 2116.

Mitchell, F.E., Marais, R.M., and Parker, P.J. (1989). The phosphorylation of protein kinase C as a potential measure of activation. Biochem. J. 261, 131-136.

Moller, W., and Amons, R. (1985). Phosphate-binding

sequences in nucleotide-binding proteins. FEBS Lett. 186, 1-7.

Nigg, E. A., Hilz, H., Eppenberger, H. M., and Dutly, F.

(1985). Rapid and reversible translocation of the catalytic subunit of cAMP-dependent protein kinase type II from the Golgi complex to the nucleus. EMBO J. 4, 2801-2806.

Pearson, R. B., Woodgett, J. R., Cohen, P., and Kemp, B. E. (1985). Substrate specificity of a multifunctional calmodulin-dependent protein kinase. J . Biol. Chem. 260, 14471-14476.

Pelech, S. L., Olwin, B. B., and Krebs, E. G. (1986).

Fibroblast growth factor treatment of Swiss 3T3 calls activates a subunit S6 kinase that phosphorylates a

synthetic peptide substrate. Proc. Nat'l. Acad. Sci. USA 83, 5968-5972.

Pelech, S. L., Sanghera, J. S., and Daya-Makin, M. (1990). Protein kinase cascades in meiotic and mitotic cell cycle control. Biochem. Cell. Biol. 68, 1297-1330.

Pelech, S. L., Tinker, D. A., Chan, C. P., and Krebs, E. G. (1987). "Role of protein phosphorylation in growth factor signal transduction", In: Insulin-like growth factors and their receptors in the central nervous system. Mohan, K., Raizade, M., Phillips, M. I., and LeRoith, D., eds. Plenum, New York. pp. 27-46.

Pulverer, B. J., Kyriakis, J. M., Avruch, J., Nikdakaki, E., and Woodgett, J. R. (1991). Phosphorylation of c-Jun mediated by MAP kinases. Nature 353, 670-674.

Rubin, C. S., Hirsch, A., Fung, C., and Rosen, O. M. (1978). Development of hormone receptors and hormonal responsiveness in vitro. J. Biol. Chem. 253, 7570-7578.

Sanghera, J. S., Aebershold, R., Morrison, H. D., Bures, E. J., and Pelech, S. L. (1990). Identification of the sites in myelin basic protein that are phosphorylated by

meiosis-activated protein kinase p44 ^mpk . FEBS Left 273,

223-226. Schaap, D., Hsi an, J., Totty, N., and Parker, P . J. (1990). Proteolytic activation of protein kinase C-epsiion. Eur. J Biochem. 191 ; 431-435.

Tkachuk, D. C, Kohler, S., and Cleary, M. L. (1992).

Involvement of a homolog of Drosophila trithorax by 11q23 chromosomal translocations in acute leukemias. Cell 71, 691-700.

Van Etten, R. A., Jackson, P., and Baltimore, D. (1989). The mouse type IV c-abl gene product is a nuclear protein and activation of transforming ability is associated with cytoplasmic localization. Cell 58, 669-678.

Vik, T. A., Sweet, L. J., and Erikson, R. L. (1990).

Coinfection of insect cells with recombinant baculovirus expressing pp60^src results in the activation of a

serine-specific protein kinase pp90^rsk . Proc. Nat'l. Acad. Sci. USA 87, 2685-2689.

Watson el al. (1982). A human cellular homolog of viral oncogene v-mos. Proc. Nat'l. Acad. Sci. USA 79, 4078-4082.

Weiss, A.L., Wiskocil, R.L., and Stobo, J.D. (1984). The role of T3 surface molecules in the activation of human T cells: A two-stimulus requirement for IL 2

production reflects events occurring at a pretranslational level. J. Immunol. 133, 123-128.

Wu, J. Y., Zhou, Z., Judd, A., Cartwright, C. A., and

Robinson, W. S. (1990). The hepatitis B virus-encoded transcriptional transactivator hbx appears to be a novel protein serine/threonine kinase. Cell 63, 687-695. Yu, K.-T., Khala:, N., and Czech, M. P. (1987). Insulin stimulates a novel Mn²⁺ -dependent cytosolic serine kinase in rat adipocytes. J. Biol. Chem. 262, 16677-16685.