MOBIX, Institute for Molecular Biology and Biotechnology, McMaster University, Hamilton, Ontario, Canada

Correspondence to: Michael A Rudnicki, MOBIX Institute for Molecular Biology and Biotechnology, McMaster University, Life Science Building, Room 437, 1280 Main Street West, Hamilton, Ontario, Canada L8S 4K1

We have cloned and characterized a novel murine Ste20-related kinase designated SLK. SLK displays high homology to the Ste20-related kinase LOK, and is more distantly related to MST1 and 2, both Ste20-like kinases. In addition, SLK displays high homology to microtubule and nuclear associated protein (M-NAP) and AT1-46, both of unknown function. SLK is ubiquitously expressed as multiple mRNAs in tissues and cell lines and is downregulated by mitogen depletion in differentiating myoblasts. Biochemical characterization showed that SLK overexpression activates c-Jun amino-terminal kinase 1 (JNK1). However, in vitro kinase assays indicated that SLK was not activated in response to various growth factors or stress-inducing agents. Immunofluorescence studies revealed that SLK colocalized to distinct cytosolic domains, preferentially at the periphery of the cells. In addition, prolonged overexpression of SLK in cultured fibroblasts resulted in apoptosis as demonstrated by annexin-V and TUNEL staining. Our results suggest that SLK belongs to a new family of protein kinases, mediating activation of the stress response pathway through a novel signaling cascade.

Ste20 kinase; apoptosis; JNK1

Introduction

Cell growth and differentiation are regulated by complex processes involving a large number of signaling cascades. The activation, or inhibition of various pathways results in the expression of specific subsets of genes directly involved in proliferation or terminal differentiation (Davis, 1993; Fanger et al., 1997). Extracellular signals acting on growth factor receptors or G-protein coupled receptors transduce their effect through a kinase cascade resulting in the activation of mitogen-activated protein kinases (MAPK) (Davis, 1993). Stress-inducing agents as well as apoptotic triggers have been demonstrated to signal through stress-responsive pathways leading to the activation of stress-activated protein kinases such as c-Jun amino-terminal kinases (JNKs) and p38 (Fanger et al., 1997).

In yeast, the Ste20 serine/threonine protein kinase regulates a mitogen-activated protein kinase pathway consisting in Ste11 (MEK kinase), Ste7 (MEK) and Fus3/Kss1 (MAPK) protein kinases involved in the control of mating response (Zhao et al., 1995). Alpha factor binding to its G-protein coupled receptor results in the downstream activation of Ste20 by the complex released from the heterotrimeric G protein. This interaction results in translocation of Ste20 to the scaffolding protein Ste5, leading to the sequential activation of Ste11, Ste7 and Fus3/Kss1 (Leberer et al., 1997a). Ste20 was also demonstrated to bind the small GTPase Cdc42, however the Cdc42 binding domain of Ste20 has been shown to be dispensable for pheromone signaling in yeast (Leberer et al., 1997b).

Apoptosis is an active genetically controlled process triggered by various stimuli in different cells. External stimuli such as cytokines, UV irradiation and numerous drugs trigger apoptotic responses characterized by a series of morphological changes that include cytoplasmic shrinkage, chromatin condensation, membrane blebbing, DNA fragmentation and the formation of apoptotic bodies (Kerr et al., 1972; Raff et al., 1993; Steller, 1995). Several studies have implicated the catalytic activities of several kinases during apoptosis. The activation of JNKs in response to apoptotic triggers such as TNF- and Fas ligand has been demonstrated (Verjeij et al., 1996; Xia et al., 1995; Yang et al., 1997). The activity of ASK1, a MAPKKK, RIP and ZIP kinases have also been recently shown to mediate apoptosis in cultured cells (Ichijo et al., 1997; Kawai et al., 1998; Stanger et al., 1995).

We have cloned and characterized a murine protein kinase representing a new member of the Ste20 kinase family. Database searches revealed that this novel kinase was the murine homolog of the human Ste20-like kinase SLK. SLK is ubiquitously expressed and is downregulated upon mitogen depletion of myoblasts and fibroblasts cultures. Here we present biochemical and expression studies that indicate that SLK is part of a novel signaling pathway mediating activation of c-Jun terminal kinase1 (JNK1) and inducing apoptosis in cultured fibroblasts.

Results

SLK, a Ste20-related kinase

During the course of a two hybrid screen using a myogenic regulatory factor (MyoD), one of several positive clones was found to encode the mouse homolog of human SLK, a Ste20-related kinase. Further experiments failed to validate this clone as a true interacting protein for the myogenic regulator. To allow further characterization of the kinase, a 1.6 kb partial cDNA insert was used to survey mRNA samples from differentiating myoblasts. Northern blot analysis revealed high level expression in myoblasts that was strongly downregulated following myotube formation (not shown). Therefore, to further investigate the role of the mouse kinase, a muscle lambda gt11 cDNA library was screened to isolate clones containing full length cDNAs. One clone (3E5) was found to contain a 5253 bp cDNA encoding full length protein (Figure 1). Clone 3E5 encoded a 1202 amino acid polypeptide and shared over 90% nucleotide identity with sequences from the guinea-pig (Itoh et al., 1997) human and rat Ste20-like kinases (SLK) recently deposited in the database, suggesting that 3E5 represents the murine homolog. Inspection of the 3' untranslated region (UTR) did not reveal the existence of a poly(A)⁺ tail or a consensus polyadenylation signal, suggesting that 3E5 bears a partial 3' UTR. However, several AU-rich motifs, previously implicated in growth factor-dependent mRNA turnover (Ross, 1996), are present in the 3' UTR.

Searches for homologous proteins revealed that the kinase catalytic domain was closely related to LOK and MST1/2, members of the Ste20 family of serine/threonine kinases. Interestingly, the central portion of 3E5 was found to be highly homologous to microtubule and nuclear associated protein (M-NAP). The remaining carboxy terminal region of the protein displayed high homology to AT1-46, a cDNA clone isolated from astrocyte mRNA. The function of both M-NAP and AT1-46 is currently unknown. Searches for known protein motifs using PSORT revealed putative nuclear localization signals at positions 15, 422 and 1149. However, nuclear localization was never observed following transfection and immunostaining of Myc epitope-tagged SLK (see Figure 7). A consensus SH3 binding site (P-X-X-P-X) was found at position 735 (Pawson and Scott, 1997), suggesting a potential interaction with SH3 domain-containing proteins. Other than a C-terminal region unusually rich in charged amino acids, database scans using Motif Finder failed to detect the presence of other consensus protein motifs.

Further analysis of the coding region revealed that SLK encoded an N-terminal serine/threonine kinase catalytic domain, with the signature sequence Gly-X-Gly-X-X-Gly identifying subdomain I at amino acid residue 41 (Figure 2a). The conserved lysine residue within the ATP binding site of subdomain II was found at amino acid residue 63. The catalytic core extended further, up to residue 282 and presented all the characteristic subdomains of serine/threonine kinases (Hanks and Hunter, 1995) (Figure 2a). Alignments and database scans revealed that SLK displayed 74% identity, in the kinase domain, to LOK, a novel kinase preferentially expressed in lymphocytes (Kuramochi et al., 1997) (Figure 2a). The SLK kinase domain was also found to be related to MST1 and MST2, both Ste20-like kinases (Creasy and Chernoff, 1995; Katoh et al., 1995; Schinkmann and Blenis, 1997). In the kinase subdomain VIII, characterizing kinase family members, SLK presented the Ste20 signature motif, suggesting that it represents a novel family member (Figure 2a). Further analysis showed that the central portion of SLK, from residues 339 - 947 (Figure 2b) shared 70% identity with microtubule and nuclear associated protein (M-NAP). The remainder of SLK, from residues 788 - 926, overlapping slightly with the M-NAP domain, displayed 63% identity to AT1-46 (Schaar et al., 1996) (Figure 2b). Interestingly, M-NAP and AT1-46 share significant homology in the overlapping region. A second region of homology (71%) to AT1-46 was observed from SLK residues 957 - 1171. Interestingly, these two AT1-46 domains were found to be 56% identical to the C-terminal region of LOK which also shares extensive identity to AT1-46 in its C-terminus (Figure 2b). These observations suggest that SLK and LOK may represent members of a new protein kinase family. Furthermore, the AT1-46 homology domain might be a novel protein motif likely to be important for LOK and SLK functions. We have termed this new motif the ATH domain, for AT1-46 homology domain.

Expression of SLK in tissues and cell lines

To gain insights into the role of SLK, adult human RNA samples were surveyed for SLK expression by Northern blotting. Analysis of poly(A)⁺ RNA from various tissues with a SLK-specific probe revealed the existence of at least three distinct isoforms of about 6, 7 and 8 kb (Figure 3a). All isoforms were expressed at similar levels with the exception that testes and colon tissues were found to express relatively higher levels of the 7 and 8 kb isoforms. Analysis of mouse tissues showed a similar pattern of expression. Interestingly, very low levels of SLK mRNAs were observed in liver samples (Figure 3b). RNA analysis of mouse cell lines showed that SLK was undetectable in undifferentiated P19 embryocarcinoma cells (Figure 3c), suggesting a role for SLK in more differentiated cell types. In contrast to murine tissue samples which predominantly expressed the 8 kb mRNA (Figure 3b), the other murine cell lines surveyed displayed similar levels of all three isoforms (Figure 3c).

Analysis of human tumor cell lines showed that the chronic myelogenous leukemia cell lines K-562 and the colon adenocarcinoma cell line SW480 expressed high levels of SLK mRNA, relative to normal blood lymphocytes and colon tissue, respectively (Figure 3d). Whether high levels of SLK correlates with these specific tumor phenotypes remains to be clarified. Other human tumor cell lines were found to express SLK at levels that were comparable to normal human tissues of the same origin.

To facilitate the biochemical characterization and to gain insights into SLK functions, anti-SLK antibodies were developed. Rabbit polyclonal antibodies were generated against a GST-SLK fusion protein encompassing residues 95 - 551 and used in Western blot analysis. Specific reactivity to murine SLK protein was observed at approximately 220 and 170 kDa in control 293 cells expressing full length SLK or a C-terminal truncation (SLK3'), respectively (Figure 4a). No signal was detected in control vector transfected 293 cells. The anti-SLK antibody was further validated in an immunoprecipitation-Western blot analysis. As shown in Figure 4a, specific SLK reactivity was observed against immunoprecipitated Myc-tagged SLK protein from 293 lysates using anti-myc antibodies, but not to vector-transfected control lysates, suggesting that our antibody reacts to murine SLK. The p220 SLK is relatively larger than the predicted molecular weight of 148 kDa, suggesting that SLK is subject to post-translational modifications (Figure 4a). In contrast to Northern blotting, Western analysis of C2C12 myoblast cultures revealed 6 SLK immunoreactive polypeptides (Figure 4b). Some of these may represent breakdown products, isoforms, or, alternatively, they might reflect post-translational modifications. Interestingly, SLK protein expression decreased as myogenic differentiation progressed (Figure 4b). A marked reduction was observed 3 days following the onset of differentiation, well after the upregulation of myogenin, a marker for terminal differentiation (Olson and Klein, 1994). Alternatively, the observed decrease in SLK expression could be due to growth factor depletion which was used to initiate differentiation. Furthermore, the 3' UTR of cDNA clone 3E5 displays several AU-rich motifs potentially mediating growth factor-dependent stability of SLK mRNA. To test whether the downregulation of SLK during C2C12 differentiation could be attributed to growth factor depletion, NIH3T3 cells were serum-starved for 24 h, stimulated with 10% serum and analysed for SLK expression by Western blotting. As shown in Figure 4c, serum starvation of 3T3 cells resulted in a rapid decrease in SLK expression. Following serum stimulation of 3T3 cultures for 24 h, a SLK reactive species of approximately 140 kDa was induced, likely representing one of several isoforms (Figure 4c). In contrast to C2C12, 3T3 cells displayed only three anti-SLK reactive proteins. With the exception of p220 SLK, these reactive species did not correspond to any of the ones detected in C2C12 cells, suggesting differential splicing or processing among cell types.

Modulation of SLK activity by extracellular stimuli

As shown in Figure 5a, bacterially expressed GST-SLK and immunoprecipitated Myc-SLK autophosphorylated and efficiently phosphorylated MBP and histone H1 in vitro. However, SLK did not phosphorylate the JNK and p38 substrates GST-Jun and GST-ATF2, respectively. Therefore, all subsequent assays were performed using the substrate histone H1. Interestingly, when compared to full length GST-SLK, immunoprecipitated Myc-SLK from transfected 293 cells displayed markedly higher autophosphorylation activity (Figure 5a), suggesting differential processing of the kinase in the two expression systems. Treatment of active GST-SLK (Figure 5b) or immunoprecipitated Myc-SLK (not shown) with calf intestinal phosphatase (CIAP) followed by kinase assay caused a marked decrease in GST-SLK activity, suggesting that its activity is regulated by phosphorylation. Furthermore these results indicate that SLK is active in a phosphorylated state.

Various growth factors, stress-inducing agents and apoptotic triggers have been demonstrated to affect the activity of several MAP kinases such as ERK1/2 as well as stress-activated kinases (Davis, 1993; Fanger et al., 1997; Hughes, 1995; Xia et al., 1995). To test whether SLK activity could be modulated by such factors, Myc epitope-tagged SLK was transfected into 293 cells followed by stimulation with various factors and stress inducing agents in the absence of serum (see Figure 6a). The activity of SLK following stimulation was evaluated by in vitro kinase assays. Interestingly, none of the factors tested were found to significantly modulate SLK kinase activity in overexpressing 293 cells (Figure 6a), suggesting that SLK is a component of a novel signaling pathway. Transfection and kinase assays using a kinase dead Myc-SLK showed no phosphorylation on histone H1, suggesting that the kinase activity observed with wild type Myc-SLK is not due to a co-precipitated kinase.

To gain insights into potential downstream effectors of SLK, the relative kinase activity of JNK1 isoforms or the mitogen-activated protein kinases ERK1/2 was determined by immunoprecipitations and in vitro kinase assays following transient transfections of Myc-SLK. As shown in Figure 6c, JNK1 activity was found to be upregulated 3 - 5-fold relative to vector-transfected cells, suggesting that c-jun amino-terminal kinase1 (JNK1) is activated by SLK overexpression. However, JNK activation by SLK overexpression was about threefold less than that observed following UV irradiation (Figure 6c, lane 2). In contrast to JNK1, immunoprecipitated ERK1/2 were found to be inactive following SLK transfection, suggesting that components of the mitogen response pathways are not markedly affected by SLK overexpression (Figure 6c).

Cellular distribution and induction of apoptosis by SLK

Because our anti-SLK antibodies failed to detect antigen in fixed cells, a Myc epitope tag expression vector carrying full length SLK transfected into C2C12 myoblasts. To evaluate the cellular distribution of SLK, immunostaining was performed using 9E10 monoclonal antibodies on fixed cultures 16 h following transient transfection. Interestingly, 9E10-positive cells displayed high concentration of Myc-SLK protein in distinct cytosolic domains, predominantly at the periphery of the cells (Figure 7a). The distribution of these domains appeared suggestive of scaffold structures from which SLK signaling could occur. Transfection of the kinase dead mutant SLK^K63R did not alter the cellular distribution (Figure 7b). No reactivity was observed in vector-transfected cells (Figure 7c).

In order to further investigate the role of SLK, expression vectors were transfected into cultured cells to generate stable cell lines. However, our attempts at generating stably expressing cells were unsuccessful, suggesting that SLK overexpression is detrimental. To investigate this possibility, Swiss 3T3 cells were transiently transfected with HA-tagged SLK and expressing cells were analysed for morphological changes and the appearance of apoptotic markers 24 h following transfection. As shown in Figure 8, the vast majority of SLK expressing cells displayed extensive shrinkage and were found to be positive for both TUNEL (terminal dUTP nick end labeling; Figure 8a,b) and annexin-V (Figure 8c,d) apoptotic markers. Together with our inability to generate stably expressing cell lines, these results strongly suggest that SLK overexpression induces an apoptotic response in cultured cells.

To determine whether SLK kinase activity is required for the induction of apoptosis, an activated form of HA-tagged SLK (amino acids 1 - 373; HA-SLK^1-373) and the corresponding inactive version, HA-SLK^1-373K63R, were transfected into Swiss 3T3 fibroblasts and assayed for annexin V binding. Expression of active HA-SLK^1-373 resulted in rounding up of the cells and annexin V binding 14 - 16 h following transfection (Figure 8e,f). However, cells expressing the inactive form HA-SLK^1-373K63R appeared morphologically normal and were found to be negative for annexin V binding, suggesting that kinase activity is required for the induction of apoptosis by SLK^1-373 (Figure 8g,h). Interestingly, when vectors encoding full length wild type or inactive SLK were expressed, an apoptotic response was observed independently of kinase activity, suggesting that one or more domains other than the kinase region can mediate an apoptotic response (Sabourin and Rudnicki, unpublished).

Discussion

We have cloned and characterized a murine Ste20-related protein kinase designated SLK. Database searches revealed that the cloned cDNA encodes a protein kinase highly related to LOK, a Ste20-related protein kinase preferentially expressed in lymphocytes (Kuramochi et al., 1997). Analysis of the kinase domain shows that SLK is also related to MST1/2 (Creasy and Chernoff, 1995; Katoh et al., 1995; Schinkmann and Blenis, 1997) and more distantly related to Ste20, a yeast kinase involved in the pheromone response pathway (Zhao et al., 1995). Furthermore, SLK showed 70% identity to M-NAP and 63 and 71% identity to AT1-46 in two distinct carboxy terminal domains. Interestingly, LOK also display extensive homology to AT1-46 in its C-terminal domain, suggesting that AT1-46, SLK and LOK represent members of a new protein family. Furthermore, the AT1-46 homology (ATH) domain may represent a novel protein motif required for SLK and LOK function. Further analysis of SLK protein showed the presence of an SH3-binding motif and a coiled-coil structure in the C-terminal region. Northern blot analysis demonstrated that SLK is ubiquitously expressed in adult tissues and its expression appears to be restricted to more differentiated cell types.

Purified GST-SLK fusion protein showed that it autophosphorylates efficiently and that it can phosphorylate exogenous substrates. However, the JNK and p38 substrates c-Jun and ATF2, respectively, were not phosphorylated by SLK. Interestingly, alkaline phosphatase treatment of recombinant SLK lead to a substantial decrease in kinase activity in vitro, suggesting that its activity is regulated by phosphorylation. Immunoprecipitations and in vitro kinase assays following overexpression of SLK in 293 cells showed that it activated isoforms of JNK1. As for SLK, the Ste20-related kinase p65PAK also activates the stress response pathway (Fanger et al., 1997). Similarly, activation of the related kinases MST1/2 activates the p38 MAPK pathway (Creasy and Chernoff, 1995; Graves et al., 1998). However, treatment of transfected 293 cells with various growth factors and stress-inducing agents did not result in any significant changes in SLK activity, suggesting that it is part of a novel pathway. One possibility is that SLK possesses low kinase activity and is activated by autophosphorylation in vitro, making it difficult to observe any significant effects. Alternatively, SLK is constitutively active and is downregulated through a novel signaling pathway.

Immunolocalization of Myc epitope-tagged SLK protein revealed that it is localized predominantly to the periphery of C2C12 myoblasts. Interestingly, SLK was found to be localized to distinct cytosolic domains. Whether these represent larger signaling complexes involving SLK remains to be elucidated. One possibility is that the putative SH3-binding domain of SLK mediates interactions with other proteins of such complexes.

Consistent with JNK1 activation, our results indicate that overexpression of SLK induces apoptosis in cultured fibroblasts. Whether SLK-induced apoptosis is mediated through JNK1 activation or a novel pathway is unclear. Potential signaling pathways and the mechanisms that control SLK activity are currently under investigation.

We have cloned and partially characterized a murine Ste20-related kinase designated SLK. SLK was found to be ubiquitously expressed in adult tissues and differentiated cell types. Immunoprecipitated, as well as purified SLK fusion protein were shown to be active in an in vitro kinase assay. Although it was not activated by various stress-inducing agents, SLK was found to activate JNK1 and induce apoptosis. The identification of SLK substrates, regulatory molecules and interacting partners will provide further insights into the mechanisms underlying its regulation and the signaling pathways that it regulates.

Materials and methods

Cloning and analysis of SLK

Full length SLK (GenBank Accession number AF112855) was obtained from an adult mouse muscle gt11 cDNA library following plaque screening (Sambrook et al., 1989) using a partial SLK cDNA clone obtained through two hybrid screening. Full length cDNA inserts were subcloned into pBluescript and sequenced on an ABI automated sequencer. Homology searches were performed using NCBI Blast software and homologies are presented as per cent identities. The serine/threonine kinase subdomains were identified by alignment of the consensus amino acid sequence of the catalytic domain (Hanks and Hunter, 1995) to that of SLK. Multiple alignment analysis was performed using the MegAlign program from the DNAStar software package. Further analyses for consensus protein motifs were performed using PSORT, Motif Finder and PRO Site.

Cell culture and transfections

C2C12 cells were maintained in Dulbecco's modified Eagle medium supplemented with 15% fetal calf serum (FCS) and induced to differentiate in DMEM containing 2% horse serum. NIH3T3, P19 and 293 cells were grown in DMEM containing 10% FCS. 3T3 cells were starved in DMEM supplemented with 0.5% FCS and subsequently stimulated with growth medium where indicated. The cultures were transfected by Lipofectamine (Gibco-BRL) according to the manufacturer's instructions using 2 - 6 g of plasmid DNA.

SLK expression analysis

For expression analysis, total RNA from various adult mouse tissues and cell lines was prepared (Sambrook et al., 1989) and poly(A)⁺ RNA was isolated by one round of selection through oligo dT cellulose (Sambrook et al., 1989). A total of 4 g of poly(A)⁺ RNA for each cell line was subjected to Northern blot analysis using a SLK-specific probe. Analysis of human tissues and cell line RNA was performed by Northern hybridization of a SLK probe to Multiple Tissue Northerns (Clontech) containing 2 g of poly(A)⁺ RNA per lane. Human RNA filters were washed under reduced stringency (0.2´SSC/0.1% SDS/55°C). Analysis of SLK protein expression was performed by lysing cultures (150 mM NaCl, 50 mM Tris, pH 7.4, 1 mM EDTA, 1% Triton-X 100 and 1 g/ml of each aprotinin, pepstatin and leupeptin) and subjecting 20 g of total lysate to Western blot analysis using anti-SLK rabbit polyclonal antibody (Sambrook et al., 1989). Reactive proteins were detected by ECL (Amersham) using a goat anti-rabbit horse-radish peroxidase (HRP) labeled secondary antibody. SLK polyclonal antibodies were generated by immunization of New Zealand rabbits using purified GST-SLK^95-551, encompassing part of the kinase and M-NAP domains. Specific immunoreactivity to murine SLK protein was observed in 293 cells transfected with full length and truncated SLK expression vectors.

For expression studies and immunolocalization experiments, HA- or Myc-tagged pcDNA3 (Invitrogen) expression vectors bearing full length, kinase dead or truncated SLK, were constructed using standard cloning procedures (Sambrook et al., 1989). The kinase inactive mutant was generated through site directed PCR-mediated mutagenesis. The SLK3' truncation (residues 1 - 950) was obtained by removing the last 263 amino acids through XhoI restriction enzyme digestion and religation (Sambrook et al., 1989). HA-SLK^1-373 was constructed by removing the cDNA portion encoding amino acids 374 - 1202 and religating. The kinase activating mutation (lysine to arginine at position 63) was introduced through PCR-base mutagenesis. Following transfection, the cultures were fixed for 10 min in 4% paraformaldehyde and SLK protein was detected using 9E10 or 12CA5 monoclonal antibodies in conjunction with FITC (or TRITC)-labeled secondary antibodies. Annexin-V (Boehringer) binding and TUNEL (Oncor) reactions were performed according to the manufacturer's instructions.

Immunoprecipitations and in vitro kinase assays

For growth factor and stress agent stimulation, transfected 293 cells were serum starved in DMEM for 1 h prior to exposure to the stimuli. The cells were then exposed to the agonists as indicated in the figure legends and lysed as described above. For SLK expression analysis, 20 g of total cell lysate was subjected to Western blotting with anti-SLK antibodies as described above. For in vitro kinase assays, 100 g of total cell lysate was immunoprecipitated with 1 g of 9E10 monoclonal antibodies and 20 l of G-protein sepharose (Pharmacia) for 2 h at 4°C. Immunoprecipitates were washed three times with NETN (50 mM Tris-HCl, pH 7.5, 150 mM NaCl, 1 mM EDTA, 0.1% Nonidet P-40) and once with kinase assay buffer (20 mM Tris-HCl, pH 7.5, 15 mM MgCl₂, 10 mM NaF, 10 mM -glycerophosphate, 1 mM sodium orthovanadate). Reactions (20 l) were initiated by the addition of 5 Ci of [-³²P]ATP and 3 g of histone H1. After a 30 min incubation at 30°C, reactions were terminated by the addition of 4´SDS sample buffer and 20 l aliquots were fractionated by 12% SDS - PAGE. Gels were stained, dried and exposed to X-ray films.

Recombinant GST-SLK fusion protein was purified using glutathione sepharose as described (Pharmacia) from 10 ml cultures. GST-SLK immobilized on beads was assayed directly for kinase activity on various substrates or treated with CIAP, then washed five times with NETN containing 10 mM NaF and 10 mM -glycerophosphate prior to kinase assays.

Genebank no AF112 855.

Acknowledgements

We thank Catherine Neville and Robert Korneluk for sequencing cDNAs. This work was supported by grants from the Medical Research Council of Canada and the National Institutes of Health to MA Rudnicki. LA Sabourin is supported by a post-doctoral fellowship from the Medical Research Council of Canada. MA Rudnicki is a Research Scientist of the Medicar Research Council of Canada.

Creasy CL and Chernoff J. (1995). Gene 167, 303-306. MEDLINE

Davis RJ. (1993). J. Biol. Chem. 268, 14553-14556. MEDLINE

Fanger GR, Gerwins P, Widmann C, Jarpe MB and Johnson GL. (1997). Curr. Opin. Genet. Dev. 7, 67-74. MEDLINE

Graves JD, Gotoh Y, Draves KE, Ambrose D, Han DK, Wright M, Chernoff J, Clark EA and Krebs EG. (1998). EMBO J. 17, 2224-2234. MEDLINE

Hanks SK and Hunter T. (1995). FASEB J. 9, 576-596. MEDLINE

Hughes DA. (1995). Semin. Cell. Biol. 6, 89-94. MEDLINE

Ichijo H, Nishida E, Irie K, ten Dijke P, Saitoh M, Moriguchi T, Takagi M, Matsumoto K, Miyazono K and Gotoh Y. (1997). Science 275, 90-94. Article MEDLINE

Itoh S, Kameda Y, Yamada E, Tsujikawa K, Mimura T and Kohama Y. (1997). Arch. Biochem. Biophys. 340, 201-207. MEDLINE

Katoh M, Hirai M, Sugimura T and Terada M. (1995). Oncogene 10, 1447-1451. MEDLINE

Kawai T, Matsumoto M, Takeda K, Sanjo H and Akira S. (1998). Mol. Cell. Biol. 18, 1642-1651. MEDLINE

Kerr JF, Wyllie AH and Currie AR. (1972). Br. J. Cancer 26, 239-257. MEDLINE

Kuramochi S, Moriguchi T, Kuida K, Endo J, Semba K, Nishida E and Karasuyama H. (1997). J. Biol. Chem. 272, 22679-22684. MEDLINE

Leberer E, Thomas DY and Whiteway M. (1997a). Curr. Opin. Genet. Dev. 7, 59-66. MEDLINE

Leberer E, Wu C, Leeuw T, Fourest-Lieuvin A, Segall JE and Thomas DY. (1997b). EMBO J. 16, 83-97. Article MEDLINE

Olson EN and Klein WH. (1994). Genes Dev. 8, 1-8. MEDLINE

Pawson T and Scott JD. (1997). Science 278, 2075-2080. Article MEDLINE

Raff MC, Barres BA, Burne JF, Coles HS, Ishizaki Y and Jacobson MD. (1993). Science 262, 695-700. MEDLINE

Ross J. (1996). Trends Genet. 12, 171-175. Article MEDLINE

Sambrook J, Fritsch EF and Maniatis T. (1989). Molecular Cloning: a Laboratory Manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor.,

Schaar DG, Varia MR, Elkabes S, Ramakrishnan L, Dreyfus CF and Black IB. (1996). Brain Res. 721, 217-228. MEDLINE

Schinkmann K and Blenis J. (1997). J. Biol. Chem. 272, 28695-28703. Article MEDLINE

Stanger BZ, Leder P, Lee TH, Kim E and Seed B. (1995). Cell 81, 513-523. MEDLINE

Steller H. (1995). Science 267, 1445-1449. MEDLINE

Verheij M, Bose R, Lin XH, Yao B, Jarvis WD, Grant S, Birrer MJ, Szabo E, Zon LI, Kyriakis JM, Haimovitz-Friedman A, Fuks Z and Kolesnick RN. (1996). Nature 380, 75-79. MEDLINE

Xia Z, Dickens M, Raingeaud J, Davis RJ and Greenberg ME. (1995). Science 270, 1326-1331. MEDLINE

Yang X, Khosravi-Far R, Chang HY and Baltimore D. (1997). Cell 89, 1067-1076. MEDLINE

Zhao ZS, Leung T, Manser E and Lim L. (1995). Mol. Cell. Biol. 15, 5246-5257. MEDLINE

Figure 1 Nucleotide and amino acid sequence of murine SLK. The N-terminal kinase domain is highlighted in black. The M-NAP and AT1-46 regions are shaded and boxed, respectively. The putative SH3-binding motif has been highlighted in white within the M-NAP domain

Figure 2 Sequence alignment of murine SLK (mSLK) with related proteins. (a) Alignment of mSLK kinase domain with that of the Ste20-related kinases human SLK (hSLK), LOK and MST. The kinase subdomains are numbered I - XI and conserved residues within the kinase domain are marked by an asterisk. The characteristic Ste20 motif in subdomain VIII is underlined. (b) Alignment of mSLK with Rat SLK, LOK, M-NAP and AT1-46. In both alignments, the shaded area represents identical amino acid residues. Numbers on the left indicate the amino acid residues

Figure 3 Northern blot analysis of tissues and cell lines. (a) 2 g of poly(A)⁺ RNA from various tissues was analysed for SLK expression and normalized for -actin mRNA levels. Northern analysis shows the presence of at least three SLK mRNA species of approximately 6, 7 and 8 kb in length (see arrows). (b) SLK mRNA expression in murine tissues. 4 g of poly(A)⁺ RNA from various murine tissues were surveyed for SLK mRNA levels. (c) SLK mRNA expression in murine cell lines. 4 g of poly(A)⁺ RNA from four murine cell lines were surveyed for SLK mRNA levels. In contrast to adult tissues, cultured cells expressed similar amounts of all three SLK mRNA isoforms. Interestingly, the embryocarcinoma cell line P19 showed no detectable levels of SLK mRNA. (d) Analysis of SLK expression in various human cell lines. Two g of poly(A)⁺ RNA from different human tumor cell lines (Clonetech MTN) were analysed for SLK expression. Relative to normal blood lymphocytes and colon tissue, respectively, high levels of tissue, respectively, high levels of SLK mRNA were observed in K562 cells, a chronic myelogenous leukemia cell line and SW480, a colon adenocarcinoma

Figure 4 Expression of SLK in transfected 293 cells, C2C12 myoblasts and NIH3T3 fibroblasts. (a) 293/pcDNA3, 293/SLK and 293/SLK3' represent antibody control lysates from vector-transfected 293 cells or cells transfected with a full length and a carboxy terminal truncated version of SLK, respectively. (a) Panel 2 shows specific reactivity of anti-SLK to immunoprecipitated (9E10; anti-myc) and immunoblotted (anti-SLK) Myc-SLK3' protein expressed in 293 cells. (b) Expression of SLK in C2C12 myoblasts. C2C12 cells in growth medium (time 0) were induced to differentiate in DMEM containing 2% horse serum (DM) and harvested at 1, 3 and 5 days following the induction of differentiation. Twenty g of total cell lysate was analysed for SLK expression and differentiation-specific markers by immunoblotting. A significant decrease in SLK protein levels is observed following the transfer of the cells to differentiation medium. (c) NIH3T3 cells were serum starved for 24 h and then stimulated with 10% fetal calf serum for 1 day and total cell lysates were analysed for SLK protein levels. A rapid decrease in SLK levels was observed in serum starved cells. However, following serum stimulation, SLK levels were up-regulated

Figure 5 In vitro kinase assay of purified or immunoprecipitated SLK. (a) Bacterially expressed full length SLK (lane A) or immunoprecipitated Myc-tagged SLK (lane B) were assayed for kinase activity on various substrates. Both showed activity on MBP and histone H1. No detectable activity was observed on GST-Jun or GST-ATF2. (b) Inactivation of SLK by calf intestinal alkaline phosphatase (CIAP) treatment. Purified recombinant GST-SLK protein was subjected to two and 20 units of CIAP for 30 min at 37°C, washed extensively and assayed for activity in the presence of phosphatase inhibitors as described in Materials and methods. Decreased kinase activity was observed following CIAP treatment

Figure 6 Effect of extracellular stimuli on SLK activity. (a) Immunoprecipitation and in vitro kinase assay of transfected Myc-SLK. Following transfection of 293 cells, the cultures were starved for 1 h in serum free medium and then stimulated with various agents for 15 min. The cells were then lysed and assayed for SLK activity on histone H1 in vitro. Lane 1: untransfected 293, lane 2: 10% FCS, lane 3: 300 nM TPA, lane 4: 100 ng/ml EGF, lane 5: non-starved, lane 6: 10 M A23187, lane 7: UV (200 J/m²), lane 8: 100 ng/ml TNF-, lane 9: 20 ng/ml IL1-, lane 10: unstimulated, lane 11: inactive Myc-SLK, lane 12: as lane 5. SLK activity was not markedly affected after exposure to the various stimuli. One representative from three independent experiments is shown. (b) Immunoblot analysis shows that the cultures expressed equivalent amounts of transfected SLK protein. (c) Activation of JNK1 by SLK. 293 cells were transiently transfected with Myc-SLK, or vector alone, and JNK1 or ERK1/2 was immunoprecipitated 18 h later using an anti-human JNK1 (Pharmingen) or ERK1/2 (Santa Cruz) antibody. Kinase activity was evaluated on GST-Jun (JNK) or MBP (ERK1/2) and compared to UV (JNK1), or EGF (ERK1/2)-treated 293 cells

Figure 7 Immunolocalization of SLK in C2C12 myoblasts. C2C12 cells were transiently transfected with a Myc-epitope tagged SLK vector and stained 16 h after transfection for expression using 9E10 supernatant. Distinct clusters of staining can be observed at the periphery of the cells for both the wild type (a) and SLK^K63R, a kinase inactive mutant form of SLK (b). No reactivity was observed in vector transfected cells (c). Photomicrographs are shown at 400´

Figure 8 Expression of apoptotic markers in SLK-expressing cells. Swiss 3T3 cells were transiently transfected with a HA-epitope tagged SLK vector and stained for SLK expression 24 h following transfection using 12CA5 monoclonal antibody. Primary antibodies were detected using FITC-labeled (a) or TRITC-labeled secondary antibodies (c). The cover slips were also subjected to TUNEL reactions using biotin/dUTP/avidin TRITC (b) or stained using FITC-annexin-V (d) prior to fixation. Colocalization of TUNEL (or Annexin-V binding) and SLK expression was observed in a high proportion of transfected cells. Wild type (HA-SLK^1-373; e and f) or mutant (HA-SLK^1-373K63R; g and h) SLK kinase region were transfected into Swiss 3T3 cells and stained with TRITC anti-HA (e and g) or FITC-annexin V (f and h). Photomicrograph are shown at 400´