Department of Molecular Pathology, Box 172, The University of Texas MD Anderson Cancer Center, 1515 Holcombe Boulevard, Houston, Texas, TX 77030, USA

Correspondence to: Balraj Singh, Department of Molecular Pathology, Box 172, The University of Texas MD Anderson Cancer Center, 1515 Holcombe Boulevard, Houston, Texas, TX 77030, USA

^aCD Pham and VB Vuyyuru contributed equally to this work

The c-Mos serine/threonine protein kinase is an essential component of cytostatic factor (CSF), which is required for metaphase II arrest of eggs in vertebrates. Previously, we showed that c-Mos residue Ser-16 is phosphorylated in the ts110 Mo-MuSV-encoded Gag-Mos fusion protein. Here we provide evidence that Mos is phosphorylated at Ser-16 in transfected COS-1 cells. To investigate the role of this phosphorylation, Ser-16 was substituted with alanine or glutamic acid in full-length v-Mos (an Env-Mos fusion protein that contains 31 additional amino acids at the amino terminus of c-Mos), its mouse c-Mos equivalent version (v-Mos residues 32-374, hereafter referred to as Mos), and mouse c-Mos. Constructs expressing mutant versions of Mos were transfected into COS-1 and NIH3T3 cells in a transient and stable manner, respectively. Synthesis and proteolysis of Mos were evaluated by pulse-chase analysis of ³⁵S-methionine-labeled proteins. Our findings indicate that the S16A mutant of Mos was highly unstable. It accumulated to approximately 10% of the level of wild-type Mos or its S16E mutant. In addition, the S16A mutation but not the S16E mutation inhibited Mos interaction with a cellular protein, p35, suggesting that phosphorylation at Ser-16 may promote Mos interaction with p35. As expected from its destabilizing effect, the S16A mutation caused a dramatic decrease in the cellular transforming activity of Mos (determined by soft-agar colony-formation assay with the stably transfected NIH3T3 cells), which is known to correlate with its CSF function. Efficient ubiquitin-mediated proteolysis of c-Mos requires proline as the second residue from the amino-terminus. In contrast to Mos, neither the stability nor protein kinase activity of v-Mos (in which c-Mos residue Pro-2 becomes Pro-33) was affected by the S16A mutation. To provide further proof that, similar to c-Mos, the S16A mutant is recognized by the proteolysis system through Pro-2, we show that the effect of the S16A mutation is reversed by the Pro-2-Ala mutation. Thus, our results indicate that Ser-16 has an important role in the regulation of c-Mos and that phosphorylation at Ser-16 may inhibit proteolysis of c-Mos.

mos oncogene; protein phosphorylation; enzyme regulation

Introduction

The cellular mos gene encodes a germ cell-specific serine/threonine protein kinase (Maxwell and Arlinghaus, 1985) that is responsible for metaphase II arrest of unfertilized eggs in vertebrates (Sagata et al., 1989). Phosphorylation of c-Mos plays a key role in regulating its function (reviewed by Singh and Arlinghaus, 1997). Both c-Mos and v-Mos are most highly phosphorylated in M phase of the cell cycle (Liu et al., 1990; Nishizawa et al., 1992). c-Mos is highly unstable in oocytes before the start of meiotic divisions. Phosphorylation of Xenopus c-Mos at Ser-3 after germinal vesicle breakdown (GVBD) in the Met-Pro-Ser-Pro conserved sequence increases c-Mos stability by protecting it from the ubiquitin-mediated degradation machinery that recognizes the adjacent residue Pro-2 (Nishizawa et al., 1992, 1993). Phosphorylation at Ser-3 is postulated to interfere with the Pro-2 recognition by the ubiquitination system (Nishizawa et al., 1992). Regulation of Mos stability is critical for its normal physiological function (Nishizawa et al., 1992; Chen and Cooper, 1995).

In metaphase II-arrested unfertilized Xenopus eggs, c-Mos is more stable than in earlier stages of oocyte maturation and does not appear to turn over. Mechanisms in addition to Ser-3 phosphorylation may therefore be involved in stabilizing c-Mos at metaphase II; these may involve regulation by phosphorylation at additional sites and/or by protein-protein interactions. c-Mos is part of a large multiprotein complex known as cytostatic factor (CSF) (Sagata et al., 1989). The identity of other CSF components and the regulation of its assembly or disassembly are not yet understood, but regulation by c-Mos phosphorylation may well be one of the mechanisms.

Moloney murine sarcoma virus (Mo-MuSV)-encoded v-Mos contains 31 amino acids added to the amino terminus of the entire c-Mos, which resulted from the fusion of a portion of viral env gene with the sequences upstream of c-mos open reading frame (van Beveren et al., 1981). As a result, the c-Mos residue Pro-2 became internalized in v-Mos; the v-Mos residue 2 is an alanine (van Beveren et al., 1981). Due to this key difference, v-Mos is significantly more stable than c-Mos. We have been using the Mo-MuSV cell transformation system for several years in an attempt to understand the role of phosphorylation in the regulation of Mos. We reported that the v-Mos residue Ser-47, which is equivalent to the c-Mos residue Ser-16, is phosphorylated by the M phase-promoting factor (MPF; cyclin B-Cdc2 kinase) in vitro (Bai et al., 1991). This residue is also phosphorylated in v-Mos-transformed cells arrested in mitosis. Matten et al. (1996) reported that the MAP kinase can also phosphorylate Ser-16 on MBP-Mos^xe in vitro although this site was phosphorylated poorly compared to other sites Ser-3, Ser-26 and Ser-158. We investigated by site-directed mutagenesis the role of Ser-16 and its phosphorylation in the regulation of mouse c-Mos function.

Results

Experimental system

In this study, we utilized the v-mos gene of Mo-MuSV strain 124. To produce a c-Mos equivalent version, we used an amino-terminal deletion mutant of v-Mos in which protein translation begins at a methionine equivalent to the first residue of c-Mos. This amino-terminal deletion mutant of v-Mos will be referred to as Mos to distinguish it from v-Mos and c-Mos. Although Mos contains some amino acid substitutions as compared to c-Mos, these substitutions (located in the protein kinase domain involving ten c-Mos residues) do not affect c-Mos protein kinase activity or its in vivo phosphorylation pattern (our unpublished results). In fact, v-Mos encoded by the HT1 strain of Mo-MuSV, which contains no substitutions in the c-Mos region (Seth and Vande Woude, 1985), behaves similarly to v-Mos encoded by Mo-MuSV 124 (Singh et al., 1988). Since we initially identified Ser-16 as a phosphorylation site on viral Gag-Mos, we first mutated this residue in v-Mos to investigate its functional significance. Further experiments reported below, using the Mos and c-Mos expression constructs, provided evidence that phosphorylation of Ser-16 may be important for stabilization of the c-Mos protein. The experiments were carried out using transfected somatic cells that lack any significant endogenous c-mos expression. This is particularly important for analysing the effect of mutations, in order to avoid the problems resulting from positive feedback regulation of c-Mos which may occur at the levels of c-Mos translation and c-Mos phosphorylation (Roy et al., 1996; Matten et al., 1996).

Identification of Ser-16 as the major phosphorylation site on Mos

Ser-16 and the following residue Pro-17 are evolutionarily conserved in c-Mos from frog to man. We have reported that Ser-16 is phosphorylated on v-Mos in ts110 Mo-MuSV-transformed NRK-6m2 cells (Bai et al., 1991). Identification of the phosphorylation site in vivo was based on the comigration with the in vitro phosphorylated synthetic peptide and manual peptide sequencing by Edman degradation reactions (Bai et al., 1991). Prior to determining the functional significance of this phosphorylation, we wanted to know whether this site is also phosphorylated on Mos. For this purpose, we expressed normal Mos and its S16E mutant (in which glutamic acid was substituted for Ser-16) in COS-1 cells by transient transfection. The proteins were metabolically labeled with ³²P-orthophosphate, immunoprecipitated with anti-Mos(332-343) and subjected to tryptic phosphopeptide mapping. Comparison of maps of the wild-type (WT) and the S16E mutant Mos showed a major phosphopeptide (spot # 1) present in WT but absent in the mutant (Figure 1). The position of this phosphopeptide on the two-dimensional map was similar to the predicted position of the tryptic phosphopeptide containing Ser-16, using methods described by Boyle et al. (1991). Migration during electrophoresis is predicted from the mass and net charge of the peptide at a given pH. R_f value of the peptide during chromatography is predicted by averaging the experimental R_f values of the constituent amino acids in a given chromatography buffer. The identity of the phosphopeptide was confirmed by observing its comigration in 2-D maps with the phosphorylated synthetic peptide containing c-Mos residues 14-22 (e.g., see Yang et al., 1996). This result provides strong evidence that Ser-16 is a major phosphorylation site in Mos. We were unable to obtain a tryptic map in case of the S16A mutant due to its extremely low level in the cells. Comparison of the maps of the WT Mos and its kinase-inactive (K90R) mutant showed that Ser-16 phosphorylation does not depend on Mos kinase activity (Figure 1). We have shown previously that v-Mos is phosphorylated predominantly at Ser-25 and poorly at Ser-16 in COS-1 cells (Yang et al., 1996).

Role of Ser-16 phosphorylation

To determine the role of Ser-16 phosphorylation, we generated Mos mutants in which serine was substituted with either alanine or glutamic acid, which would mimic unphosphorylated and phosphorylated serine, respectively, in terms of charge. The initial experiments were to determine whether the mutant proteins are expressed appropriately and whether they are immunoprecipitated with the Mos antibodies. When mutant Mos proteins expressed in transiently transfected COS-1 cells were metabolically labeled with ³⁵S-methionine for 30 min and immunoprecipitated with anti-Mos(6-24) and anti-Mos(332-343), some interesting findings emerged. Although anti-Mos(332-343) recognized WT and both mutant Mos proteins, anti-Mos(6-24) failed to recognize the S16A mutant (Figure 2). In contrast, anti-Mos(6-24) recognized the S16E Mos similar to WT Mos. The mutations lie in the middle of the dominant antigenic epitope in the synthetic peptide used to generate the Mos(6-24) antibody (NK Herzog, B Singh and RB Arlinghaus, unpublished observations) and, therefore, alanine substitution in Mos seemed to have destroyed the recognition of Mos by this antibody.

A more significant point suggested by the experiment shown in Figure 2 was that the S16A Mos was present at a lower level (approximately one fifth) compared to the S16E mutant or WT Mos; the latter two were present in comparable amounts (compare C-Term antibody precipitates in Figure 2). Another potentially important finding that needs further study to understand its significance, was that the S16A mutation significantly weakened Mos interaction with a cellular protein, p35 (Figure 2). The effect was specific, since the S16A mutation did not affect Mos interaction with other cellular proteins (e.g., p70 and p60). The S16E mutant of Mos also interacted efficiently with p35. The S16A mutation not only inhibited Mos interaction with p35 but also enhanced Mos interaction with two cellular proteins p30 and p25 (Figure 2). It appears that there is a competition between p35 and p30/p25 for binding with Mos; phosphorylation (or a negative charge) at position 16 shifts the balance in favor of p35 binding. We have recently identified p70 as Hsp70 (Liu et al., 1999); the identity of other Mos-interacting proteins is not known. Interactions of p35, p30 and p25 with Mos may be mediated by the amino-terminal sequences since these proteins were not coimmunoprecipitated with the amino terminal antibody, Mos(6-24) (compare N-Term and C-Term antibody lanes of Figure 2).

Next, we determined the effect of Ser-16 substitutions on the steady-state level of Mos by Western immunoblotting with the Mos(332-343) antibody. The Mos(6-24) antibody failed to detect the S16A mutant on the immunoblot (data not shown). In agreement with the immunoprecipitation results (Figure 2), the level of S16A mutant Mos was found to be reduced to one tenth that of WT Mos. The S16A mutation also caused an increase in the electrophoretic mobility of Mos during SDS - PAGE (Figure 2). In contrast, the S16E mutant behaved like WT Mos in terms of amount and electrophoretic mobility.

We attempted to determine the effect of Ser-16 substitutions on the protein kinase activity of Mos by an immune complex kinase assay. This required the use of Mos(6-24) antibody for the immunoprecipitation of Mos as other available antibodies inhibit Mos protein kinase activity (Singh et al., 1990). The experiment showed that the S16E mutant of Mos produced in transfected COS-1 cells had protein kinase activity similar to that of WT Mos; the S16E mutation did not affect the ability of Mos to autophosphorylate or phosphorylate its substrate MEK1 (data not shown). In this experiment, no kinase activity was observed in case of the S16A mutant because it was not immunoprecipitated with this antibody (see Figure 2).

To address whether mutating Ser-16 affects c-Mos similar to Mos, we generated the S16A and S16E mutants of mouse c-Mos and tested their effects on the amount of c-Mos in COS-1 cells and in the rabbit reticulocyte lysate system. We found that the S16A mutation but not the S16E mutation caused a reduction in the amount of c-Mos in both COS-1 cells and in vitro (Figure 3). Thus, the mutations affected c-Mos identically to Mos. c-Mos is known to phosphorylate MEK1 and thus activate the MAPK pathway. To obtain an indirect estimate of the protein kinase activity of c-Mos mutants, we analysed the status of MAPK phosphorylation. By Western blotting with a phospho-specific antibody (Promega, Madison, WI, USA) that specifically recognizes active MAPK, we found that c-Mos and its S16A and S16E mutants caused activation of MAPK in the reticulocyte lysate system which was roughly proportional to the amount of c-Mos (compare Figures 3b and c). We obtained similar results with the phospho-specific MEK1 antibody (data not shown). As a control, the MAPK pathway was not activated in the rabbit reticulocyte lysate system in mock-translation reaction (Figure 3) or upon translation of kinase-inactive Mos (data not shown). We conclude from these experiments that the S16A mutation does not significantly affect the protein kinase activity of c-Mos.

Effects of Ser-16 substitutions on the biological activity of Mos

We had two reasons for choosing the soft-agar colony-formation assay, using stably transfected NIH3T3 cells, to determine the transforming activity of Mos mutants. First, the assay would tell us whether the S16A mutant has protein kinase activity in vivo. Second, results of this assay correlate with the CSF activity assay which involves assay of the metaphase arrest in a two-cell stage frog embryo (Freeman et al., 1990; Okazaki et al., 1991; Yew et al., 1991). As expected from the in vitro protein kinase assay, transfection with the S16E mutant and WT Mos resulted in a similar number of colonies in soft-agar (Figure 4, Table 1). Of interest is the finding that the S16A mutant also caused a low-level transformation of NIH3T3 cells. Although the number of soft-agar colonies was less (about one log lower than that of WT Mos), the S16A Mos mutant reproducibly gave colonies that were not observed in control NIH3T3 cells (Figure 4). The lower number of soft-agar colonies correlated well with the reduced amount of S16A mutant Mos protein present in transiently transfected COS-1 cells and in the rabbit reticulocyte system. Because of an extremely low amount of Mos protein present in stably transfected cells, WT or mutant Mos could not be detected (data not shown). It is well established that the transforming activity of Mos depends on and correlates with its protein kinase activity (Singh et al., 1986; Yew et al., 1991). As expected, the kinase-inactive K90R mutant of Mos behaved similar to mock-transfected NIH3T3 cells in the soft-agar colony formation assay (Table 1). We conclude from these experiments that the S16A mutant Mos has protein kinase activity. However, these results do not rule out the possibility that the S16A mutation affects the protein kinase activity of Mos.

Effect of the S16A mutation on Mos degradation

The results presented so far suggested that the S16A mutation enhances degradation and/or inhibits synthesis of Mos. Since c-Mos is degraded by the ubiquitin-mediated proteolysis system, we used the rabbit reticulocyte lysate in vitro translation system, which has been employed to study ubiquitin-mediated proteolysis of numerous proteins including c-Mos (Nishizawa et al., 1993). To determine whether the S16A mutation enhances Mos degradation, the S16A and WT Mos were synthesized in a rabbit reticulocyte lysate system in the presence of [³⁵S]methionine from equal amounts of in vitro synthesized capped RNAs. After 15 min, the translation reaction was supplemented with an excess of non-radioactive methionine (0.2 mM) in order to follow the degradation of already synthesized radioactive Mos. We found that the S16A mutant was degraded significantly faster than the WT Mos (Figure 5). While the amount of mutant Mos was greatly reduced by 45 min, the WT Mos was relatively unaffected. In these experiments, we always had a lower amount of the S16A mutant protein than the WT moss at the 0 min time point of chase, which corresponds to the time point at the end of 15 min radioactive labeling period. This is presumably caused by faster degradation and possibly less synthesis of mutant Mos during the 15-min labeling period. We carried out a similar pulse-chase analysis on mutant Mos expressed in transiently transfected COS-1 cells and obtained similar results. The S16E mutant and WT Mos were degraded at a similar slower rate than the S16A mutant (data not shown).

No effect of S16A mutation on v-Mos

The results presented above are consistent with a model in which phosphorylation of Ser-16 could play a role in stabilizing the Mos protein. This model predicts that unphosphorylated Ser-16 (or Ala-16), but not phosphorylated Ser-16 (or Glu-16) would be recognized efficiently by the ubiquitin mediated proteolysis system. Based on the `second codon rule' for the degradation of Mos, the other key residue which determines recognition of Mos by the ubiquitin system is a proline at position 2 (Nishizawa et al., 1992, 1993). Therefore, it was of interest to determine whether the S16A mutation (S47A in v-Mos) would destabilize v-Mos in which Pro-2 becomes Pro-33. The wild-type and mutant v-Mos proteins expressed in COS-1 cells were analysed for their steady-state levels and protein kinase activity. Interestingly, the alanine for serine substitution at position 47 of v-Mos had essentially no effect on its steady-state level as determined by Western immunoblotting with the Mos(332-343) antibody (Figure 6c). Similarly, the mutation did not affect the v-Mos protein kinase activity (Figure 6a). These results indicate that the recognition of Ser-16 by the proteolysis system also depends on the amino-terminal sequence of Mos and that, therefore, v-Mos stability may not be affected by this residue's phosphorylation status. As a separate point, the S47A mutant of v-Mos (indicated as S16A in Figure 6) was recognized efficiently by the Mos(6-24) [v-Mos(37-55)] antibodies in both Western blots (Figure 6b) and immunoprecipitation reactions (Figure 6a). Although it is not ruled out that the mutation could affect the antibody recognition of v-Mos to a low degree, a dramatic difference in the recognition of S47A v-Mos versus S16A Mos by anti-Mos(6-24) suggests a complex role of this region in the folding of the Mos protein.

P2A mutation reverses the effect of the S16A mutation

The main concern with the use of mutants such as the S16A mutant is that the mutation may cause such a fundamental change in the structure of the Mos protein that the mutant protein is recognized by the proteolysis system in a manner that is not relevant to c-Mos degradation. To address this concern, we created another point mutation, P2A, in the S16A mutant; the P2A mutation inhibits the proteolysis of c-Mos (Nishizawa et al., 1992). Interestingly, we found that the P2A-S16A mutant was stable similar to the P2A mutant (Figure 7a). This result strongly suggests that the S16A mutant, similar to normal c-Mos, is recognized by the proteolysis machinery through the Pro-2 residue of c-Mos. To address the noted concern by another approach, we added a 39-amino acid long unrelated sequence at the N-terminus of the S16A Mos mutant that changes the position of Pro-2 converting it into Pro-41. The resulting His-Mos fusion proteins contain the following 39-amino acid sequence at the N-terminus: MRGSHHHHHHGMASMTGGQQMGRDLYDDD-DKDRWIRPRG. As expected, similar to the P2A mutation, this manipulation also reversed the destabilizing effect of the S16A mutation (compare the level of WT and the S16A mutant of His-Mos in Figure 7b).

One concern regarding the effect of the S16A mutation on Mos - p35 interaction (Figure 2) was that the reduction in p35 coprecipitation may just reflect a decrease in Mos level. To address this, we investigated p35 association with the S16A mutant of (His)₆-Mos by coimmunoprecipitation (Figure 7c). The data showed that the S16A mutation resulted in reduction in p35 association without decreasing the level of (His)₆-Mos, indicating an important role of Ser-16 in mediating p35 association with Mos. Another point evident by the immunoprecipitation data in Figure 7c is that, similar to the results obtained with v-Mos (Liu et al., 1999) and the P2A mutant of Mos (data not shown), p30 and p25 did not associate with His-Mos. These results indicate that the N-terminus of c-Mos is important for mediating interaction with these proteins.

Discussion

The results presented here demonstrate that Ser-16 is a major phosphorylation site on Mos. The other major phosphorylation site is Ser-25, which is included in the tryptic phosphopeptide migrating at top left in the maps shown in Figure 1 (Yang et al., 1996; data not shown). Both serine residues are also phosphorylated on v-Mos, but Ser-25 is the major site and Ser-16 a minor site of phosphorylation (Yang et al., 1996). Our results support a model in which phosphorylation at Ser-16 serves to protect Mos from proteolysis. Phosphorylation at this site occurs very poorly on v-Mos and, according to our mutagenesis data, may not have any significant influence on v-Mos activity. Although our mutagenesis experiments provide strong evidence for the role of Ser-16 and its phosphorylation in regulating c-Mos, we have thus far been unable to specifically manipulate the phosphorylation status of Ser-16 and determine its effect on protein degradation.

Our results are analogous, in several ways, to the results obtained regarding the role of Ser-3 phosphorylation in regulating the stability of c-Mos. In both cases, the evidence for the role of phosphorylation is based on the identification of the phosphorylation site and on mutagenesis studies; direct correlation between phosphorylation and c-Mos stability is yet to be shown. It remains possible, in both cases, that the Ala mutants may be more unstable than the unphosphorylated c-Mos and that the Glu mutants may have different stability compared to phosphorylated c-Mos. What is important is that the enhanced degradation of c-Mos caused by either the S16A mutation (this study) or the S3A mutation (Nishizawa et al., 1992) does not occur in the presence of the P2A mutation. These results suggest that, similar to normal c-Mos, the mutations are recognized by the proteolysis machinery through the Pro-2 residue. Thus, overall evidence supports the model of positive regulation of c-Mos by phosphorylation at Ser-3 and Ser-16.

During Xenopus oocyte maturation, the amount of c-Mos protein present is regulated by a positive feedback regulation mechanism (Roy et al., 1996; Matten et al., 1996). Activation of the MAP kinase pathway by c-Mos kinase increases the amount of c-Mos protein by increasing its translation. Since c-Mos causes activation of the MAP kinase pathway and MPF (Posada et al., 1993; Huang and Ferrell Jr, 1996), phosphorylation of c-Mos at Ser-16 by MPF and/or MAP kinase may potentially represent a part of the mechanism involved in the positive-feedback regulation of c-Mos. Although our results support the role of Ser-16 phosphorylation in Mos stabilization, additional direct or indirect roles of this phosphorylation are not ruled out. As an example, more activation of the MAP kinase pathway by WT c-Mos than by the S16A mutant could in some way enhance c-Mos translation as suggested by others (Roy et al., 1996; Matten et al., 1996). One possibility that can be tested is that Ser-16 phosphorylation in the nascent polypeptide chains mediates some protein-protein interactions that may in turn determine the fate of c-Mos by an enhanced polypeptide elongation rate and/or reduced protein degradation. Our observation of difference between S16A and WT Mos in p35 association (Figures 2 and 7) is consistent with this interesting possibility. The identification of p35 will be important in understanding the mechanism of regulation of Mos by phosphorylation at Ser-16.

In Mos-transfected COS-1 cells, Mos is not phosphorylated at Ser-3 (Yang et al., 1998). Therefore, similar to the situation in immature oocytes, Mos is highly susceptible to degradation. This is the situation in which we studied the stabilizing effects of Ser-16 phosphorylation. It is possible that the stabilizing effect of Ser-16 phosphorylation synergizes with the stabilization through Ser-3 phosphorylation in metaphase II-arrested unfertilized eggs. Further studies are needed to determine the phosphorylation status of Ser-3 and Ser-16 at different stages of oocyte maturation in mice. To better understand the role of Ser-16 phosphorylation we must identify the protein kinase(s) that phosphorylate Ser-16 in vivo, and find ways to alter specifically the phosphorylation status of Ser-16. Our preliminary attempts at manipulating the phosphorylation state of Mos by using protein kinase inhibitors (roscovitine and PD98059) and -protein phosphatase have, thus far, not produced results.

Our results imply that phosphorylation of Ser-16 may lead to conformational alterations in Mos. In this regard, our recent finding of Mos interaction with Pin1, a peptidyl-prolyl cis-trans isomerase (VB Vuyyuru, CD Pham, KP Lu and B Singh, manuscript in preparation), may be of considerable interest. Pin1 catalyzes isomerization of proline residues having a phosphorylated residue at their amino-termini (Yaffe et al., 1997). It will be of interest to see whether phosphorylation at Ser-16 promotes isomerization at Pro-17 and thus provide an explanation of the results presented here.

Materials and methods

Plasmids and site-directed mutagenesis

The Mo-MuSV 124 v-mos gene and its mutants were expressed in COS-1 and NIH3T3 cells from the SV40 late promoter by cloning into pJC119 vector (Hannink and Donoghue, 1985) or pEUK-C1 vector (Clontech, Palo Alto, CA, USA). Alanine was substituted for residue Ser-47 of v-Mos using T7-GEN in vitro mutagenesis kit (United States Biochemical, Cleveland, Ohio, USA). The procedure involved copying the v-mos template cloned into M13mp18RF vector using a v-mos oligonucleotide primer encoding the desired mutation. The S16E mutation in Mos was created by polymerase chain reactions (PCR) using oligonucleotide primers containing appropriate base substitutions. The mutations were confirmed by DNA sequencing. The wild-type v-Mos, Mos and their corresponding Ser-16 mutations were cloned into appropriate expression vectors.

To generate the S16A and S16E mutations in mouse c-Mos, we substituted a NsiI-ApaI fragment excised from pBluescript-mos-S16A and pBluescript-mos-S16E with the corresponding fragment from pBluescript-c-mos^mu. All the amino acid substitutions in Mos compared to mouse c-Mos are included in this fragment encoding c-Mos residues 66-343 (van Beveren et al., 1981). For expression in COS-1 cells, an XhoI fragment encoding entire c-Mos and the appropriate Ser-16 mutation was cloned into the XhoI site of pJC119 vector. The P2A and P2A-S16A mutants were generated by PCR using the sense primer encoding the P2A mutation and copying either the normal mos template or the S16A mos template. The His-Mos expression constructs were generated by cloning the entire mos gene with or without the S16A mutation into the XhoI site of the pJC119-His vector (Yang et al., 1998). The His-Mos protein contains 39 additional amino acids (including hexahistidine) at the N-terminus of Mos.

Mos protein analysis

The plasmids were transfected into COS-1 cells by the DEAE-dextran method (Sambrook et al., 1989). For stable expression in NIH3T3 cells, cotransfection with the pKJ1 neo vector (Adra et al., 1987) was carried out using LipofectAMINE reagent (GIBCO-BRL, Gaithersburg, MD, USA). Transfected cells were selected in the presence of geneticin (400 g/ml; GIBCO-BRL). Methods of metabolic labeling with ³⁵S-methionine or ³²P-orthophosphate, immunoprecipitation, Western immunoblotting, protein kinase assays involving autophosphorylation and transphosphorylation of MEK1 (using GST-MEK1 K97R kinase-inactive mutant), and tryptic phosphopeptide mapping have been described previously (Pham et al., 1995; Yang et al., 1996). ³⁵S-methionine-labeled COS-1 cells were lysed in a buffer containing 1% nonidet p40, as were the cells used for Mos protein kinase assays (Singh et al., 1988).

Mos synthesis and degradation in rabbit reticulocyte lysates

Capped RNAs encoding wild-type and mutant versions of Mos were synthesized from linearized pBluescript-mos DNA template using either T3 or T7 Cap-scribe kit (Boehringer Mannheim Corporation, IN, USA). In vitro translation was carried out in a rabbit reticulocyte lysate system (Promega, Madison, WI, USA) in the presence of [³⁵S]methionine or [³⁵S]cysteine (Amersham, Arlington Heights, IL, USA). To examine the degradation of ³⁵S-methionine-labeled Mos, an excess of non-radioactive methionine (0.2 mM final concentration) was added to the translation mixture after 15 min, aliquots were collected at various time points, acetone-precipitated and analysed by SDS-polyacrylamide gel electrophoresis followed by fluorography.

Soft-agar colony-formation assay

The transforming activity of Mos mutants was assayed by analysing the ability of stably transfected NIH3T3 cells to form colonies in soft-agar as described previously (Yang et al., 1998). Cells (1´10⁴) were seeded in 9.6-cm² wells of 6-well plates in triplicate. After incubation at 37°C for 14 days, colonies were stained with 0.1% p-iodonitrotetrazolium violet (Sigma Chemical Company, St. Louis, MO, USA) in phosphate-buffered saline (0.6 ml per well) overnight. Excess dye solution was removed the next day and colonies were photographed.

Acknowledgements

We thank Ralph Arlinghaus for valuable suggestions and critical reading of the manuscript. We also thank Linda Kimbrough for typing and Lore Feldman for editing the manuscript. This work was supported by grants R01 CA45125 and CA16672 (core) from the National Institutes of Health.

Adra CN, Boer PH and McBurney MW. (1987). Gene 60, 65-74. MEDLINE

Bai W, Singh B, Karshin WL, Shonk RA and Arlinghaus RB. (1991). Oncogene 6, 1715-1723. MEDLINE

Boyle WJ, van der Geer P and Hunter T. (1991). Methods Enzymol. 201, 110-149. MEDLINE

Chen M and Cooper JC. (1995). Mol. Cell. Biol. 15, 4727-4734. MEDLINE

Freeman RS, Kanki JP, Ballantyne SM, Pickham KM and Donoghue DJ. (1990). J. Cell Biol. 111, 533-541. MEDLINE

Hannink M and Donoghue DJ. (1985). Proc. Natl. Acad. Sci. USA 82, 7894-7898. MEDLINE

Huang CYF and Ferrell Jr JE. (1996). EMBO J. 15, 2169-2173. MEDLINE

Liu H, Vuyyuru VB, Pham CD, Yang Y and Singh B. (1999). Oncogene 18, in press.

Liu J, Singh B, Wlodek D and Arlinghaus RB. (1990). Oncogene 5, 171-178. MEDLINE

Matten WT, Copeland TD, Ahn NG and Vande Woude GF. (1996). Dev. Biol. 179, 485-492. MEDLINE

Maxwell SA and Arlinghaus RB. (1985). Virology 143, 321-333. MEDLINE

Nishizawa M, Okazaki K, Furuno N, Watanabe N and Sagata N. (1992). EMBO J. 11, 2433-2446. MEDLINE

Nishizawa M, Furono N, Okazaki K, Tanaka H, Ogawa Y and Sagata N. (1993). EMBO J. 12, 4021-4027. MEDLINE

Okazaki K, Furuno N, Watanabe N, Ikawa Y, Vande Woude GF and Sagata N. (1991). Jpn. J. Cancer Res. 82, 250-253. MEDLINE

Pham CD, Arlinghaus RB, Zheng C-F, Guan K-L and Singh B. (1995). Oncogene 10, 1683-1688. MEDLINE

Posada J, Yew N, Ahn NG, Vande Woude GF and Cooper JA. (1993). Mol. Cell. Biol. 13, 2546-2553. MEDLINE

Roy LM, Haccard O, Izumi T, Larres BG, Lewellyn AL and Maller JL. (1996). Oncogene 12, 2203-2211. MEDLINE

Sagata N, Watanabe N, Vande Woude GF and Ikawa Y. (1989). Nature 342, 512-518. MEDLINE

Sambrook J, Fritsch EF and Maniatis T. (1989). Molecular cloning. Cold Spring Harbour Laboratory Press. Cold Spring Harbor, N.Y.,

Seth A and Vande Woude GF. (1985). J. Virol. 56, 144-152. MEDLINE

Singh B, Al-Bagdadi F, Liu J and Arlinghaus RB. (1990). Virology 178, 535-542. MEDLINE

Singh B and Arlinghaus RB. (1997). Progress in Cell Cycle Research, Vol. 3. Meijer L, Guidet S and Philippe M. (eds.). Plenum Press: New York, USA, pp. 251-259.

Singh B, Hannink M, Donoghue DJ and Arlinghaus RB. (1986). J. Virol. 60, 1148-1152. MEDLINE

Singh B, Herzog NK, Liu J and Arlinghaus RB. (1988). Oncogene 3, 79-85.

van Beveren C, van Straaten F, Galleshaw JA and Verma IM. (1981). Cell 27, 97-108. MEDLINE

Yaffe MB, Schutkowski M, Shen M, Zhou XZ, Stukenberg PT, Rahfeld J-U, Xu J, Kuang J, Kirschner MW, Fischer G, Cantley LC and Lu KP. (1997). Science 278, 1957-1960. Article MEDLINE

Yang Y, Herrmann CH, Arlinghaus RB and Singh B. (1996). Mol. Cell. Biol. 16, 800-809. MEDLINE

Yang Y, Pham CD, Vuyyuru VB, Liu H, Arlinghaus RB and Singh B. (1998). J. Biol. Chem. 273, 15946-15953. MEDLINE

Yew N, Oskarsson M, Daar I, Blair DG and Vande Woude GF. (1991). Mol. Cell. Biol. 11, 604-610. MEDLINE

Figure 1 Tryptic phosphopeptide maps of Mos and its mutants. Mos wild-type (WT) or its mutants S16E and K90R (kinase-inactive) were expressed in COS-1 cells. Cells labeled with ³²P-orthophosphate (1 mCi/ml) for 2 h were lysed in RIPA buffer, Mos was immunoprecipitated with anti-Mos(332-343) and subjected to two-dimensional phosphopeptide mapping. The bottom right panel shows a schematic diagram of phosphopeptide spots and their identities. The phosphopeptide number 1 (marked to the right of the spot) is present in maps of K90R mutant and WT Mos but not the S16E mutant. A common spot (numbered 2) present in the top left of all maps corresponds to a phosphopeptide that includes Ser-25 phosphate (Yang et al., 1996). A minor phosphopeptide (numbered 3) may include an autophosphorylation site, as it is missing in the map of the kinase-inactive Mos. Identity of phosphopeptide numbered 4, which is present in all maps, is unknown at present

Figure 2 Effect of Ser-16 mutations on immunoprecipitation of Mos and associated proteins. Two days after transfection, COS-1 cells were labeled with ³⁵S-methionine for 30 min, lysed in a buffer containing 1% NP-40 and immunoprecipitated with anti-Mos(6-24) (N-Term) or anti-Mos(332-343) (C-Term). In addition to Mos, proteins coprecipitated with Mos are indicated by arrows. All these proteins also coprecipitated with Mos in anti-Mos (229-240) immunoprecipitates (not shown) confirming that they are Mos-associated proteins. As specificity controls, immunoprecipitates obtained with the antibodies pre-reacted with the cognate peptide antigens (peptide plus lanes) are included. Shown at left are the molecular-weight markers (Diversified Biotech, Boston, MA, USA)

Figure 3 Effect of Ser-16 mutations on steady-state levels and protein kinase activity of c-Mos. (a) Western blot detection of c-Mos and its S16A and S16E mutants in transfected COS-1 cells using the anti-c-Mos(229-240) antibody. Notice the faster migration of the S16A mutant similar to the S16A Mos mutant shown in Figure 2. (b) Relative amount of ³⁵S-cysteine-labeled c-Mos mutants after translation for 1 h in the rabbit reticulocyte lysate system. (c) Activation of the MAPK pathway by c-Mos. Equivalent amounts of rabbit reticulocyte lysate samples (10 l each) after 1 h translation (same as shown in b) were subjected to Western blot analysis with a phospho-specific MAPK antibody (Promega) which detects only the active forms of ERK1 and ERK2. The left unnumbered lane in b and c represents the translation without any exogenously added RNA. (d) Equivalent amount of ethidium bromide-stained RNAs (resolved on a agarose gel) encoding c-Mos and its mutants which were used for in vitro translation whose results are shown in b and c

Figure 4 Effect of Ser-16 mutations on the transforming activity of Mos. NIH3T3 cells stably transfected with wild-type (WT) Mos and S16A and S16E mutants were used to carry out the soft-agar colony-formation assay as described in Materials and methods. Mock-transfected NIH3T3 cells (top left) served as a negative control

Figure 5 Effect of the S16A mutation on Mos degradation in rabbit reticulocyte lysate. After synthesis of Mos (WT and the S16A mutant) for 15 min, its degradation was followed in the presence of excess non-radioactive methionine as described in Materials and methods. Chase time in minutes refers to the time after the addition of non-radioactive methionine. A part of WT Mos sample at 45 min time point (lane 11) was lost accidentally

Figure 6 The S16A mutation does not affect v-Mos protein level and kinase activity. (a) v-Mos and its version carrying the S16A mutation (S47A v-Mos) were produced by transfection in COS-1 cells and analysed by immune complex kinase assay using the Mos (6-24) antibody. Both autophosphorylation and trans-phosphorylation of GST-MEK1 were analysed. (b and c) Western blot detection of v-Mos and its S16A mutant with two separate Mos antibodies that are indicated at the bottom. Mock, mock-transfected COS-1 cells control. Relative intensities of background bands provide useful information by serving as gel-loading controls

Figure 7 The P2A mutation or the amino acid sequence addition at the N-terminus of Mos reverses the destabilizing effect of the S16A mutation. (a) The wild-type (WT) Mos and the indicated mutants were synthesized in the rabbit reticulocyte lysate system in the presence of [³⁵S]cysteine. Five l samples were analysed by SDS - PAGE followed by fluorography. (b) His-Mos and the S16A mutant of His-Mos were produced in transfected COS-1 cells and detected by Western blotting with the anti Mos(332-343) antibody. The left lane represents the mock-transfected cells. Seen below the His-Mos band are the artifacts of the enhanced chemiluminescent (ECL) procedure. (c) Co-immunoprecipitation analysis. His-Mos and its S16A mutant produced in COS-1 cells were metabolically labeled with ³⁵S-methionine and analysed by immunoprecipitation with the Mos(332-343) antibody as in Figure 2

 Comparative analysis of transforming efficiency of Mos mutants