Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Extracellular signal-regulated kinases (ERKs) are serine/ threonine kinases of the mitogen-activated protein kinase (MAPK) superfamily thought to play a key role in the transduction of mitogenic signals to the cell nucleus. The major pathway involved in activation of ERKs 1 and 2 appears to require the sequential activation of Ras, Raf-1, and MAPK/ERK kinase (MEK) (1). Alternative pathways for ERK activation, however, do exist. Chao and colleagues (10) investigated the role of Raf-1 in ERK activation employing a BALB/c 3T3 cell derivative stably transfected with a plasmid vector overexpressing a dominant-negative Raf-1. Despite the absence of functional Raf-1, epidermal growth factor (EGF) and phorbol ester were effective in activating ERK, demonstrating an alternative, Raf-1-independent pathway for ERK activation. Similarly, expression of Raf-1 antisense messenger RNA in normal rat kidney fibroblasts failed to block ERK activation following oncogenic and nononcogenic signals (11), and less than 20% of MEK-activating activity in EGF-treated Swiss 3T3 cells is attributable to Raf-1 (12).

Studies in cultured tracheal myocytes are consistent with the notion that ERK activation may occur independently of Raf-1 in airway smooth muscle. In bovine tracheal myocytes, pretreatment with either forskolin or histamine, both of which increase adenyl cyclase activity, significantly reduces both platelet-derived growth factor (PDGF)-induced and EGF-induced Raf-1 activation. However, ERK activity is not significantly affected, demonstrating that ERK activation may occur independently of Raf-1 (13). PDGF-induced ERK activation is also forskolin-insensitive (14) in rat tracheal myocytes, suggesting a Raf-1-independent ERK activation pathway in these cells. Because MEK1 activation appears to be required and sufficient for ERK activation in a variety of cell types, including bovine tracheal myocytes (5, 15, 16), it is likely that Raf-1-independent pathways to ERK activation involve alternative activators of MEK1. Activation of MEK1 has been demonstrated to require the phosphorylation of two serine codons, Ser²¹⁸ and Ser²²² (7). Several alternative MEK activators have been identified in other systems, including other members of the raf family of protein kinases, A-Raf and B-Raf (17- 19), MEK kinase-1 (MEKK1) (20, 21), Mos (22, 23), and Tpl-2 (24).

The present study was undertaken to identify Raf-1- independent upstream activation pathways of ERK in bovine tracheal myocytes. Our data suggest the presence of a novel PDGF-stimulated, forskolin-insensitive MEK1 kinase with a molecular mass of 40 to 45 kD.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Materials

Antihuman -smooth muscle actin, peroxidase-linked goat antirabbit immunoglobulin (Ig)G, peroxidase-linked goat antimouse IgG, myelin basic protein (MBP), and molecular weight standards for gel filtration chromatography were purchased from Sigma Chemical (St. Louis, MO). PDGF was obtained from Upstate Biotechnology (Lake Placid, NY). HiTrap Q ion exchange and Superdex 200 gel filtration columns were obtained from Pharmacia (Piscataway, NJ). For in vitro phosphorylation assays, recombinant inactive glutathione-S-transferase (GST)-MEK1, "wild-type" GST-MEK1, and GST-ERK2 were purchased from Upstate Biotechnology. [-³²P]adenosine triphosphate (ATP) and an enhanced chemiluminescence kit were obtained from DuPont/NEN Research Products (Wilmington, DE). For Western analyses, antibodies against Raf-1, A-Raf, B-Raf, and Mos were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Anti-phosphoMAPK was purchased from Promega (Madison, WI). Antisera against Mos and MEKK1 were provided by G. Vande Woude (National Cancer Institute, Frederick, MD) and D. Templeton (Case Western Reserve University, Cleveland, OH), respectively. For Mos immunoblots, bovine oocyte extracts were provided by L. Liu (University of Connecticut, Storrs, CT).

Cell Culture and Harvesting of Cell Extracts

Bovine tracheal smooth-muscle cells were isolated as described previously (9, 13). Myocytes of passage number 4 or less were studied. Confluent cultures exhibited the typical "hill and valley" appearance under phase-contrast microscopy and showed specific immunostaining with anti- -smooth-muscle actin.

Cell cultures were grown to confluence in 100-mm plates, serum-starved for 24 h, and treated with PDGF (30 ng/ml for 10 min). To inhibit Raf-1 activity, cells were pretreated with forskolin (50 µM for 15 min). After treatment, cells were washed with cold phosphate-buffered saline (PBS) (150 mM NaCl, 0.1 M phosphate [pH 7.5]) and lysed with 0.4 ml of buffer A (consisting of 10 mM Tris-HCl [pH 7.4], 10 mM MgCl₂, 2 mM ethyleneglycol-bis-(-aminoethyl ether)-N,N'-tetraacetic acid [EGTA], 1 µg/ml aprotinin, 0.2 mM phenylmethylsulfonylfluoride [PMSF], 1 mM sodium fluoride, and 1% 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonate). Lysates were centrifuged (14,000 rpm for 10 min at 4°C), and the supernatants were resolved by ion exchange chromatography.

Western Blot Analyses

Whole-cell extracts or chromatographic fractions were resolved by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). Proteins were transferred to a nitrocellulose membrane using a semidry transfer apparatus (Hoefer, South San Francisco, CA). After Ponceau staining, the membrane was incubated with the relevant primary antibody. Signals were amplified using the appropriate peroxidase-conjugated secondary antibody, and visualized by enhanced chemiluminescence.

Chromatography

Anion exchange and gel filtration chromatography were performed on a Pharmacia fast protein liquid chromatography system. Clarified cell lysate was filtered and loaded on a HiTrap Q column (volume, 5.0 ml) equilibrated with 20 ml of buffer A. The column was washed with 10 ml of buffer A and developed with a 40-ml linear gradient from 0 to 1 M NaCl in buffer A at 0.5 ml/min. Fractions were assayed for MEK1 phosphorylation activity as described below.

To determine the apparent molecular mass of the MEK1 kinase, peak fractions of MEK1 phosphorylation activity were pooled and concentrated by centrifugation (5,000 rpm for 1 h at 4°C) using a Centricon microconcentrator (10,000 molecular weight cutoff; Amicon, Beverly, MA). After equilibration in a buffer containing 20 mM Tris-HCl (pH 7.4), 10 mM MgCl₂, 2 mM EGTA, 0.2 mM PMSF, and 150 mM NaCl, this material was chromatographed on a a Superdex 200 gel filtration column (volume, 25 ml). The flow rate was 1 ml/min, and 250-µl fractions were collected. The column was calibrated using the following molecular weight standards: relative molecular weight (M_r) 2,000,000, blue dextran 2,000; M_r 200,000, -amylase; M_r 150,000, alcohol dehydrogenase; M_r 66,000, albumin; M_r 29,000, carbonic anhydrase; and M_r 12,400, cytochrome C. Fractions were assayed for MEK1 phosphorylation activity as described later. Apparent molecular weight of the forskolin-insensitive, PDGF-stimulated MEK1 kinase was calculated from a calibration curve prepared by plotting the logarithms of the known molecular weights of protein standards versus their respective volume of elution volume/void volume (Ve/Vo) values.

MEK1 Phosphorylation Assay

In vitro MEK phosphorylation assays were performed as previously described (9). Aliquots of each fraction (20 µl) were incubated with approximately 3 µg of purified inactive MEK1 (MEK1-GST fusion protein, agarose-conjugated) in 20 µl of phosphorylation buffer (10 mM Tris-HCl, 0.5 mM dithiothreitol, 2 mM EGTA, and 2 mM MgCl₂). The kinase reaction was initiated by adding 5 µCi [-³²P]ATP to each mixture, and the incubation was continued for 20 min at 30°C. After washing the phosphorylated GST-MEK beads in cold PBS, sample buffer was added and the samples were resolved on a 10% SDS gel. Proteins were transferred to nitrocellulose by semidry blotting. After Ponceau staining, the membrane was exposed to film and the level of MEK1 phosphorylation measured by optical scanning. Although multiple phosphorylation bands were found (see RESULTS), we quantified the phosphorylation of only the major MEK1 bands, as determined by Ponceau staining. In almost all experiments, one 72-kD band was observed. On a few occasions, doublets were found.

MEK1 Kinase Assay

To determine whether the MEK1 phosphorylation we observed was associated with MEK1 activation, a coupled assay was performed in which activation of MEK1 was assessed by virtue of its ability to activate ERK, as measured by MBP phosphorylation. Aliquots of each fraction (10 µl) were incubated with or without approximately 2 µg of purified "wild-type" MEK1, 1 µg of ERK2 substrate, and a reaction buffer containing 20 mM Tris-HCl (pH 7.4), 10 mM MgCl₂, 2 mM MnCl₂, 0.2 mM EGTA, 50 mM "cold" ATP, 0.1 mM sodium vanadate, 10 mM p-nitrophenyl phosphate, and 1 mg/ml bovine serum albumin (20 min at 30°C). To assay ERK activity, 5 µg of MBP and 5 µCi [-³²P]ATP were added and the incubation was continued for 20 min at 30°C. The reaction was stopped by adding gel sample buffer. Samples were resolved on a 15% SDS gel and transferred to nitrocellulose by semidry blotting. The level of MEK phosphorylation was visualized by autoradiography.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Confirmation of Forskolin-Insensitive ERK Phosphorylation after PDGF Treatment

In a previous study, we demonstrated that inhibition of Raf-1 activity by forskolin or histamine does not significantly attenuate PDGF-induced ERK activation in bovine tracheal myocytes (13). To confirm this finding, we performed anti-ERK immunoblotting using a phosphospecific antibody (25) that recognizes ERKs only when phosphorylated at Thr183 and Tyr185, which are required for full enzymatic activity (26). Treatment with PDGF induced substantial phosphorylation of ERKs 1 and 2 (Figure 1). Pretreatment with forskolin had no apparent effect on ERK phosphorylation, implying the existence of alternate ERK activation pathways.

View larger version (18K): [in this window] [in a new window]   Figure 1.   Effect of forskolin (FSK) pretreatment on PDGF- induced ERK phosphorylation. Bovine tracheal myocytes were pretreated with forskolin (50 µM for 20 min) before stimulation with PDGF (30 ng/ml for 10 min). Cell proteins were resolved by SDS-PAGE, transferred to nitrocellulose, and probed with an anti-ERK phosphospecific antibody. Forskolin had no perceptible effect on ERK phosphorylation.

Identification of Forskolin-Insensitive MEK1 Phosphorylation

To identify other kinases that might phosphorylate MEK1, the upstream activator of ERK in these cells (9), we examined MEK phosphorylation activity in bovine tracheal myocytes treated with both PDGF and forskolin. Cell lysates were chromatographed on a HiTrap Q ion exchange column and fractions assayed for the ability to phosphorylate a recombinant inactive MEK1 in vitro (Figures 2a and 2b). In vitro MEK phosphorylation assays were performed as previously described (9). In all five experiments, MEK1 phosphorylation activity was detected in fractions 18 through 26 (80 to 160 mM NaCl), with the peak fraction eluting at a NaCl concentration of 140 mM. Lysates from cells treated only with forskolin exhibited only minimal MEK phosphorylation activity (Figure 2a).

View larger version (57K): [in this window] [in a new window]   View larger version (16K): [in this window] [in a new window]   Figure 2.   Identification of forskolin-insensitive MEK1 phosphorylation. Cell cultures were grown to confluence, serum-starved, and treated with PDGF (30 ng/ml for 10 min). To inhibit Raf-1 activity, cells were pretreated with forskolin (50 µM for 15 min). Cell lysates were subjected to anion exchange fast protein liquid chromatography, and the MEK1 kinase activity of fractions was assessed by in vitro phosphorylation assay using a recombinant kinase-inactive MEK1 as a substrate. (a) Upper panel: representative MEK1 phosphorylation assay demonstrating activity in fractions 18 through 26 (80 to 160 mM). Lower panel: representative MEK phosphorylation assay from cells exposed to forskolin alone. (b) Chromatograph demonstrating MEK1 phosphorylation activity (solid line), NaCl concentration (dashed line), and protein concentration (faint, dotted line).

In vitro phosphorylation assays performed with GST- MEK1 substrate demonstrated the phosphorylation of multiple proteins. To determine whether these bands represented phosphorylated MEK1 or simply autophosphorylation, MEK1 phosphorylation HiTrap Q fractions holding peak MEK1 phosphorylation activity were re-assayed, this time in the absence of MEK1 substrate (Figure 3). We found minimal phosphorylation of proteins in the absence of the 72-kD GST-MEK1 substrate, confirming that the phosphorylation observed in the MEK1 phosphorylation assays was indeed that of MEK1 recombinant proteins. Conversely, the phosphorylation of two proteins of approximately 58 to 64 kD molecular weight in fractions 34 through 38 (Figure 2a) persisted in the absence of GST- MEK1 substrate (Figure 3), demonstrating that the phosphorylation observed in these fractions was not indicative of MEK1 phosphorylation activity.

View larger version (36K): [in this window] [in a new window]   Figure 3.   In vitro phosphorylation assay of peak fractions performed in the presence (left panel) or absence (right panel) of MEK1. To confirm that the apparent MEK1 phosphorylation we observed was not due to phosphorylation of an alternative protein of similar molecular weight, peak fractions 22 and 24 were examined for phosphorylation activity in the presence or absence of GST-MEK1 substrate. In the presence of MEK1, fractions 22 and 24 demonstrated MEK1 phosphorylation activity (left panel). However, in the absence of substrate, these fractions showed minimal phosphorylation (right panel). On the other hand, fractions 36 and 38 showed persistent phosphorylation of two approximately 60-kD proteins despite the absence of substrate, demonstrating that phosphorylation of these proteins did not constitute MEK1 phosphorylation activity.

Confirmation of MEK1 Kinase Activity in Fractions Containing MEK1 Phosphorylation Activity

The presence of multiple phosphorylated bands in the MEK1 in vitro phosphorylation assays raised the possibility that MEK1 can be phosphorylated on multiple sites in vitro, only some of which may be physiologic. To determine whether the phosphorylation of MEK1 observed in active fractions resulted in the activation of this enzyme, we evaluated the ability of relevant fractions to activate ERK, as measured by MBP phosphorylation. Fractions were incubated with MEK1, ERK2, and ATP, and ERK activity was then assayed by measuring the phosphorylation of MBP by radiolabeled ATP (Figure 4). Using this coupled assay, we found that fractions 20 through 26 induced substantial phosphorylation of MBP, whereas fractions containing no MEK phosphorylation activity showed only minimal MBP phosphorylation. No significant MBP phosphorylation was noted when MEK1 was excluded from the incubation mixture. These results suggest that the protein(s) eluting at 100 to 160 mM NaCl are capable of stimulating both MEK1 phosphorylation and activation, indicating the presence of a MEK kinase.

View larger version (39K): [in this window] [in a new window]   Figure 4.   MEK1 kinase activity assay. To test for MEK1 kinase activity, a coupled assay was performed in which the ability of relevant fractions to activate ERK was measured by MBP phosphorylation. First, aliquots of each fraction were incubated with MEK1, ERK2, and ATP. Activated ERK was then assayed by incubation with MBP and [-32P]ATP. Fractions 20 through 26 induced substantial phosphorylation of MBP, whereas fractions containing no MEK1 phosphorylation activity showed minimal MBP phosphorylation (left panel). No significant MBP phosphorylation was noted when MEK1 was excluded from the incubation mixture (right panel).

Immunoblotting of HiTrap Q Fractions for Known MEK Activators

HiTrap Q fractions were probed for known MEK activators by Western blot analysis. Specific antibodies for ERKs, Raf-1, Tpl-2, MEKK1, A-Raf, B-Raf, and Mos were employed. (Although ERKs are not thought to phosphorylate MEK in vivo, this phenomenon has been observed in vitro [7].) Immunoreactive ERK1, ERK2, Raf-1, Tpl-2, and MEKK1 were each present in lysates from bovine tracheal myocytes, though not in the precise fractions containing MEK1 phosphorylation activity (Figure 5). We were unable to detect A-Raf, B-Raf, or Mos convincingly in any of the HiTrap Q fractions (not shown).

View larger version (33K): [in this window] [in a new window]   Figure 5.   Immunoblotting for known MEK activators. HiTrap Q fractions were assayed for known MEK activators by Western blot analysis. In the experiments shown, immunoreactive phosphorylated ERK1 (44 kD) was detected in fraction 28, and phosphorylated ERK2 (42 kD) in fraction 30. Immunoreactive Raf-1 (74 kD) was detected in fraction 36. Immunoreactive Tpl-2 (56 kD) was detected in fraction 32. Immunoblotting for MEKK1 using an antibody against the C-terminal kinase domain revealed the presence of at least three major bands. For comparison, whole cell extracts are shown at the far left (Ext).

Apparent Molecular Mass of the Forskolin-Insensitive, PDGF-Stimulated MEK1 Kinase

To determine the apparent molecular mass of the forskolin-insensitive, PDGF-stimulated MEK1 kinase, peak fractions of MEK phosphorylation activity were pooled and concentrated, and the proteins were separated by Superdex 200 gel filtration chromatography (Figures 6a and 6b). MEK phosphorylation activity was detected in fractions 14 through 22 (peak fraction, 18). Molecular weight was calculated from a calibration curve prepared by plotting the logarithms of the known molecular weights of protein standards versus their respective volume of Ve/Vo values. The active fractions corresponded to an apparent molecular mass of 40 to 45 kD.

View larger version (50K): [in this window] [in a new window]   View larger version (13K): [in this window] [in a new window]   Figure 6.   Apparent molecular mass of the MEK1 kinase. Peak HiTrap Q fractions of MEK1 phosphorylation activity were pooled, concentrated, and resolved on a Superdex 200 gel filtration column. The column was calibrated using commercially available molecular-weight standards. Fractions were assayed by in vitro phosphorylation assay using a recombinant inactive MEK1 as a substrate. (a) MEK1 phosphorylation activity was detected in fractions 14 through 22 (peak fraction, 18). (b) The peak fraction of MEK1 kinase activity corresponded to a molecular mass of 40 kD.

Because of the similarities in molecular mass between the PDGF-stimulated MEK1 kinase and Mos, ERK1, and ERK2, proteins from pooled fractions holding peak kinase activity were separated by SDS-PAGE, and immunoblots were probed for these proteins. The presence of Mos protein was assessed using an antibody raised against a peptide based on the homologous nucleotide sequence of human, Xenopus, and mouse Mos (27). This antibody is capable of recognizing bovine Mos from matured oocytes (28). An extract of unfertilized bovine oocytes, in which this protein is abundant, was examined as a positive control (Figure 7). Mos was not present in the pooled active fractions but was identified in bovine oocytes, confirming that Mos was not responsible for the MEK1 kinase activity we observed. Active ERK1 and ERK2 were also absent from pooled active fractions (not shown).

View larger version (49K): [in this window] [in a new window]   Figure 7.   Immunoblotting of pooled active fractions for Mos. HiTrap Q (lane 1) and Superdex 200 (lane 2) fractions holding peak kinase activity were pooled and immunoblots were probed using an antibody raised against a peptide based on the homologous nucleotide sequence of human, Xenopus, and mouse Mos (27). Bovine tracheal myocyte whole-cell extracts (Ext) and bovine oocytes (Oo) were also examined.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Although the major ERK activation pathway appears to involve sequential activation of Ras, Raf-1, and MEK (1- 9), alternative pathways for ERK activation exist (10). In tracheal myocytes, cyclic adenosine monophosphate (cAMP) accumulation significantly reduces PDGF-induced Raf-1 activation (13) but ERK activity is unaffected (13, 14), demonstrating that ERK activation may occur independently of Raf-1 in cultured airway smooth-muscle cells. Because MEK1 activation appears to be required and sufficient for ERK activation in bovine tracheal myocytes (9), we hypothesized that MEK1 is phosphorylated and activated in these cells by a serine/threonine kinase other than Raf-1. To test this hypothesis, we fractionated lysates from forskolin and PDGF-treated cells by anion exchange chromatography, and assessed the MEK1 kinase activity of individual fractions by in vitro phosphorylation assay. MEK1 phosphorylation activity was detected in fractions 18 through 26 (80 to 160 mM), with the peak fraction eluting at a NaCl concentration of 140 mM. Although Raf-1 was detected in the lysates of bovine tracheal myocytes, none was found in the active fractions, confirming the presence of a Raf-1- independent ERK activation pathway. Moreover, additional serine/threonine kinases known to activate MEK in other cell types (see below) were absent from active fractions, suggesting the presence of a novel PDGF-stimulated, forskolin-insensitive MEK1 kinase. Active fractions were further purified on a gel filtration column, and peak MEK phosphorylation activity was detected in a fraction corresponding to an apparent molecular mass of 40 to 45 kD.

Several alternative MEK activators have been identified in other systems. Raf-1-independent activation of ERKs may involve other members of the raf family of protein kinases, A-Raf and B-Raf. The 68-kD A-Raf has been demonstrated to phosphorylate and activate MEK1 in HeLa cells (17). Due to alternative splicing, the B-raf gene encodes two proteins of 95 and 68 kD. The 95-kD isoform of B-Raf appears to be a major activator of MEK in PC12 neuronal cells (18, 29) and NIH3T3 cells (19). Expression of B-Raf has been demonstrated to confer resistance to the inhibitory effect of cAMP on ERK activation in Raf-1 fibroblasts (30) and to be activated by cAMP in PC12 cells (31). Thus, there is ample evidence from other systems that B-Raf may activate MEK1 in a cAMP-insensitive manner. However, activation of B-Raf has been demonstrated to be cAMP-sensitive in NIH3T3 cells (19) and, under some conditions, in PC12 cells (29, 30, 32). In the present study, we were unable to detect A-Raf or B-Raf in bovine tracheal myocytes, consistent with previous work demonstrating a wide variation in tissue expression of these genes (33). Unlike the ubiquitously expressed C-raf (the gene encoding Raf-1), highest levels of A-raf transcripts are found in the urogenital system, whereas the highest levels of B-raf are found in neural cells and testes (18, 29). The absence of A-Raf or B-Raf expression, combined with the apparent molecular mass of the kinase as determined by gel filtration chromatography (40 to 45 kD), suggests that neither A-Raf nor B-Raf is responsible for the forskolin-insensitive MEK1 kinase activity observed following PDGF treatment of bovine tracheal myocytes.

MEKK was originally identified as the 78-kD mammalian form of the yeast protein kinases Byr1 and Ste7 (20). In this initial study, overexpression of MEKK in COS cells induced activation of MEK1. However, it was later demonstrated that induction of MEKK activation in HeLa (34) and NIH3T3 cells (35) activates Jun amino-terminal kinase (JNK, also known as stress-activated protein kinase) but not ERK, consistent with the notion that MEKK preferentially regulates the JNK pathway. Recently, it has been demonstrated that the MEKK1 gene actually encodes a 196-kD protein product (36). Smaller forms of MEKK1 may be the result of proteolytic cleavage by caspases (36) or translation initiation at downstream sites. These smaller proteins may retain kinase activity. In addition, three additional MEKK isoforms capable of regulating both ERK and JNK have been cloned. MEKK2, a 70-kD protein, and MEKK4, a 180-kD protein, each appear to activate JNK preferentially (39, 40). MEKK3, a 72-kD protein, is unique in that it appears to regulate ERKs preferentially (39). In our study, probing of anion exchange fractions with an antiserum against the C-terminal catalytic domain of MEKK1 (35) revealed at least three immunoreactive proteins with molecular weights ranging from 48 to 72 kD, consistent with the notion that MEKK1 may be enzymatically cleaved to smaller forms. It is unclear why we should find such cleavage products in cells not undergoing apoptosis, but the presence of 72-kD MEKK1 in nonapoptotic cells has been noted previously by many investigators. The smaller proteins eluted from the anion exchange column in the "flow-through" fractions, fractions that did not contain significant MEK1 phosphorylation activity. Therefore, it is unlikely that MEKK1 is a major activator of MEK1 in bovine tracheal myocytes.

It has recently been demonstrated that Tpl-2, a serine/ threonine kinase homologous to the yeast gene product Ste11, may phosphorylate and activate MEK1 and stress-activated protein kinase/ERK kinase-1, the upstream activator of JNK (24, 41). Because Tpl-2 is expressed primarily in spleen, thymus, liver, and lung (42), it is conceivable that Tpl-2 is responsible for the activation of MEK in bovine tracheal smooth muscle. However, as with Raf-1 and MEKK1, this protein was eluted at a higher salt concentration than the peak MEK phosphorylation activity.

Another candidate MEK activator is Mos (22), a serine/ threonine kinase whose expression is mainly restricted to testes and ovaries (33). In NIH3T3 cells, the c-mos gene product has been demonstrated to activate ERK in vivo via direct phosphorylation of MEK (23). Because of the similarity between the apparent molecular mass of the PDGF-stimulated MEK1 kinase we identified and Mos (39 kD), we probed immunoblots of cell extracts, individual fractions, and pooled fractions holding peak kinase activity for Mos protein using an antiserum raised against a peptide based on the homologous nucleotide sequence of human, Xenopus, and mouse Mos. We could not identify Mos in any of these samples, demonstrating that Mos is not the source of the MEK1 kinase activity in our system. However, Mos was detectable in both bovine and Xenopus oocytes, proving that the antibody we employed was capable of recognizing bovine Mos, if present, in our samples.

Despite the number of potential MEK activators addressed above, it is likely that additional kinases remain to be identified. Reuter and colleagues (19) reported a novel MEK kinase of 40 to 50 kD in NIH3T3 fibroblasts that eluted from a Mono Q ion exchange column at 10 to 80 mM NaCl. Pang and associates (43) found a nerve growth factor-stimulated MEK kinase in PC12 cells with an apparent molecular mass of 45 to 50 kD. The activity of the latter kinase, which in preliminary studies was also found in PDGF-stimulated NIH3T3 cells, eluted from a Mono Q ion exchange column at 160 mM NaCl and appeared to be independent of protein kinase C (PKC) and dependent on Ras. The PDGF-stimulated MEK1 kinase we identified in bovine tracheal myocytes eluted at a similar NaCl concentration (140 mM) and has a similar molecular weight (40 to 45 kD), suggesting that the two kinases are the same protein. Indeed, PDGF-induced ERK activation in bovine tracheal myocytes appears to be both PKC-independent and Ras-dependent (M. Hershenson, unpublished observations), suggesting that the two proteins have the same upstream activators. Further purification of this protein should provide insight into mitogen-activated signaling in airway smooth muscle.