Ras-independent Activation of the Raf/MEK/ERK Pathway upon Calcium-induced Differentiation of Keratinocytes*

MAPKs are crucially involved in the regulation of growth and differentiation of a variety of cells. To elucidate the role of MAPKs in keratinocyte differentiation, activation of ERK, JNK, and p38 in response to stimulation with extracellular calcium was analyzed. We provide evidence that calcium-induced differentiation of keratinocytes is associated with rapid and transient activation of the Raf/MEK/ERK pathway. Stimulation of keratinocytes with extracellular calcium resulted in activation of Raf isozymes and their downstream effector ERK within 10–15 min, but did not increase JNK or p38 activity. Calcium-induced ERK activation differed in kinetics from mitogenic ERK activation by epidermal growth factor and could be modulated by alterations of intracellular calcium levels. Interestingly, calcium stimulation led to down-regulation of Ras activity at the same time that ERK activation was initiated. Expression of a dominant-negative mutant of Ras also did not significantly impair calcium-induced ERK activation, indicating that calcium-mediated ERK activation does not require active Ras. Despite the transient nature of ERK activation, calcium-induced expression of the cyclin-dependent kinase inhibitor p21/Cip1 and the differentiation marker involucrin was sensitive to MEK inhibition, which suggests a role for the Raf/MEK/ERK pathway in early stages of keratinocyte differentiation.

MAPKs are crucially involved in the regulation of growth and differentiation of a variety of cells. To elucidate the role of MAPKs in keratinocyte differentiation, activation of ERK, JNK, and p38 in response to stimulation with extracellular calcium was analyzed. We provide evidence that calcium-induced differentiation of keratinocytes is associated with rapid and transient activation of the Raf/MEK/ERK pathway. Stimulation of keratinocytes with extracellular calcium resulted in activation of Raf isozymes and their downstream effector ERK within 10 -15 min, but did not increase JNK or p38 activity. Calcium-induced ERK activation differed in kinetics from mitogenic ERK activation by epidermal growth factor and could be modulated by alterations of intracellular calcium levels. Interestingly, calcium stimulation led to down-regulation of Ras activity at the same time that ERK activation was initiated. Expression of a dominant-negative mutant of Ras also did not significantly impair calcium-induced ERK activation, indicating that calcium-mediated ERK activation does not require active Ras. Despite the transient nature of ERK activation, calcium-induced expression of the cyclin-dependent kinase inhibitor p21/Cip1 and the differentiation marker involucrin was sensitive to MEK inhibition, which suggests a role for the Raf/MEK/ERK pathway in early stages of keratinocyte differentiation.
The epidermis is a regenerative organ subject to permanent turnover throughout the life of an organism. In healthy skin, there is a balance between proliferation of mitotically active keratinocytes in the innermost basal layer and differentiation of post-mitotic cells within the suprabasal layers of the epithelium. Under certain pathological conditions such as tumorigenesis or in hyperproliferative diseases, this equilibrium is disturbed. Elucidation of the intracellular signaling pathways involved in regulation of these processes is therefore of great interest.
Several in vitro modulators of keratinocyte growth and differentiation have been described (1,2). Stimulators of prolifer-ation include epidermal growth factor (EGF) 1 (3) and transforming growth factor-␣ (4), whereas elevation of extracellular calcium levels induces many changes that also occur during the differentiation process in vivo. Switching extracellular calcium levels from 0.05 to Ͼ0.1 mM results in rapid and irreversible cell cycle arrest in keratinocytes (5,6). This is followed by increased expression of differentiation-associated proteins as well as induction of morphological changes characteristic of differentiation such as the establishment of intercellular desmosomal contacts (5,7). In addition to its well characterized effect on keratinocyte differentiation in vitro, calcium constitutes a gradient in epidermal layers (8,9), which also implies a function for calcium as regulator of keratinocyte growth and differentiation in vivo. Intracellularly, calcium has been shown to trigger phospholipase C␥ activation, thereby elevating cytosolic calcium levels (10,11). It also activates certain protein kinase C (PKC) isozymes (12) as well as tyrosine kinases such as Fyn (13); however, the downstream consequences of these events are not clear.
A signaling pathway known to be crucially involved in the regulation of growth and differentiation in a variety of cells is the Raf/MEK/ERK kinase cascade. This pathway transduces signals from the extracellular compartment through an evolutionarily conserved intracellular signaling module that is composed of the serine/threonine kinase Raf, the dual specificity kinase MEK, and the nuclear shuttle kinase ERK (14,15). In pheochromocytoma PC12 cells, the best studied example for involvement of this pathway in differentiation, the duration of ERK activation determines which fate a cell adopts: transient activation of ERK by EGF leads to mitosis, whereas sustained activation by nerve growth factor (NGF) promotes differentiation of these cells into sympathetic-like neurons (15,16). Similarly, sustained activation of the Raf/MEK/ERK pathway is associated with differentiation of muscle cells (17). However, recent data argue against the notion that sustained activation of the Raf/MEK/ERK pathway is the only determinant decisive for induction of cellular differentiation and suggest that other signaling pathways may also play a role. For instance, inhibition of the sustained phase of ERK activation alone was not sufficient to block NGF-induced differentiation of PC12 cells * This work was supported by Grant LU 477/3-2 (to S. L. and M. G.) through Program "Molekulare Differenzierungsmechanismen von Epithelien" (SPP1028) from the Deutsche Forschungsgemeinschaft. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 1 The abbreviations used are: EGF, epidermal growth factor; PKC, protein kinase C; MEK, mitogen-activated protein kinase/extracellular signal-regulated kinase kinase; ERK, extracellular signal-regulated kinase; NGF, nerve growth factor; MAPK, mitogen-activated protein kinase; MBP, myelin basic protein; JNK, c-Jun N-terminal kinase; HA, hemagglutinin; GST, glutathione S-transferase; BAPTA-AM, 1,2-bis(2aminophenoxy)ethane-N,N,NЈ,NЈ-tetraacetic acid acetoxymethyl ester; FCS, fetal calf serum; EMEM, Eagle's minimal essential medium; PBS, phosphate-buffered saline; NHEK, normal human epidermal keratinocytes; GAP, GTPase-activating protein. (18). Furthermore, NGF stimulation of PC12 cells resulted in rapid and sustained activation of the p38 MAPK pathway (19,20), and inhibition of this pathway prevented NGF-induced neurite outgrowth in these cells (20). In muscle cells, a similar correlation between p38 activation and differentiation has been observed (21,22), which supports the view that the p38 pathway is also involved in the regulation of cellular differentiation.
Since the role of MAPK cascades (particularly the Raf/MEK/ ERK pathway) in the regulation of keratinocyte differentiation is still unclear, we compared the activation and function of these pathways upon calcium-induced differentiation of human keratinocytes. We show here that calcium-induced differentiation of human keratinocytes leads to a transient, peak-like activation of ERK, which differs in kinetics from EGF-induced ERK activation. In contrast to EGF, activation of ERK by calcium occurs through a Ras-independent pathway and appears to be mediated by accumulation of intracellular calcium. Despite the transient nature of ERK activation, the calciuminduced up-regulation of the cyclin-dependent kinase inhibitor p21/Cip1 and the differentiation marker involucrin is sensitive to MEK inhibition, which suggests a role for the Raf/MEK/ERK pathway in the early stages of keratinocyte differentiation.
Cell Culture and Plasmids-HaCaT cells were routinely grown in Dulbecco's modified Eagle's medium (Life Technologies, Inc.) containing 5% FCS (PAA Laboratories). For experiments, cells were passaged once in Eagle's minimal essential medium (EMEM) without calcium (Bio-Whittaker, Inc.) containing 5% Chelex-treated FCS and adjusted to 0.05 mM calcium (5) (referred to below as low calcium medium) and seeded in EMEM for all assays. For stimulation, calcium was directly pipetted into the medium from a stock solution of 250 mM CaCl 2 in PBS to a final concentration of 1.0 mM. In all inhibitor experiments, inhibitors were added to the medium 30 min before stimulation.
Primary normal human epidermal keratinocytes (NHEK) were isolated from neonatal foreskin as described (23) and grown in a 1:1 mixture of serum-free keratinocyte growth medium (Life Technologies, Inc.) and calcium-free EMEM. The final calcium concentration of the mixture medium was 0.05 mM.
The plasmids used for transient transfection experiments are all derived from the pcDNA3 mammalian expression vector. The HA-ERK2 construct used here has been described earlier (24). Ras N17 is a dominant-negative point mutant of Ras (25), and Raf C4 is a dominantnegative deletion mutant of c-Raf-1 (26) that has been fused to a HA tag (27).
Immune Complex Kinase Assays-Cells were grown in EMEM containing 0.05 mM calcium and 5% Chelex-treated FCS until they were 70 -80% confluent. Then cells were starved for 18 h in 0.5% FCS. After stimulation, cells were washed once with ice-cold PBS and lysed for 1 h at 4°C in Triton lysis buffer (20 mM Tris (pH 7.4), 137 mM NaCl, 10% (v/v) glycerol, 1% (v/v) Triton X-100, 2 mM EDTA, 50 mM sodium ␤-glycerophosphate, 20 mM sodium pyrophosphate, 1 mM Pefabloc, 5 g/ml aprotinin, 5 g/ml leupeptin, 5 mM benzamidine, and 1 mM sodium orthovanadate). Cell debris was removed by centrifugation at 15,000 ϫ g for 10 min. After quantification of the protein content in the supernatants (Bio-Rad protein assay), equal amounts of total lysates were incubated with 25 l of protein A or G agarose and 1 g/ml rabbit antiserum against ERK2, JNK1, or p38 for 2 h at 4°C. For Raf kinase assays, Raf kinases were precipitated with either a rabbit polyclonal antibody against c-Raf-1 or a mouse monoclonal antibody against B-Raf. For precipitation of transfected HA-tagged ERK, a mouse monoclonal antibody against HA (12CA5) was used. The precipitated kinases were washed twice with Triton lysis buffer containing 500 mM NaCl and twice with kinase buffer (10 mM MgCl 2 , 25 mM HEPES (pH 7.5), and 25 mM sodium ␤-glycerophosphate supplemented with 5 mM benzamidine, 0.5 mM dithiothreitol, and 1 mM sodium orthovanadate). Each sample was then incubated with MBP, GST-c-Jun-(1-135), or 3pK KϾM as substrate for ERK2, JNK1, and p38, respectively, in the presence of 100 M ATP, 5 M [␥-32 P]ATP, and kinase buffer in a volume of 20 l for 15 min at 30°C. Deviating from the standard procedure described above, for Raf kinase assays, immunoprecipitated kinases were washed twice with Triton lysis buffer containing 150 mM NaCl and twice with kinase buffer lacking dithiothreitol and containing 150 mM NaCl. Kinase reactions were then performed in the same kinase buffer supplemented with 0.5 mM dithiothreitol using dominant-negative MEK K97M as a substrate. Reactions were terminated by addition of 5ϫ Laemmli SDS sample buffer and boiling for 5 min. Proteins were separated by SDSpolyacrylamide gel electrophoresis and blotted onto nitrocellulose membranes (0.2 m; Schleicher & Schü ll). Kinase activities were routinely quantified with a Bio-Imaging BAS 2000 analyzer (Fuji) and visualized by autoradiography. In all cases, Western blot analysis was performed to confirm equal precipitation of ERK2, JNK1, p38, HA-ERK2, c-Raf-1, and B-Raf. Experiments were repeated at least twice.
Western Blotting-For the preparation of lysates, keratinocytes were washed once with ice-cold PBS and lysed in Triton lysis buffer on ice for 1 h. Cells were scraped into microcentrifuge tubes and pelleted by centrifugation at 15,000 ϫ g for 10 min. The supernatant was collected, and protein content was measured by the Bio-Rad protein assay. Equal protein amounts were separated by SDS-polyacrylamide gel electrophoresis and blotted onto nitrocellulose membranes. Proteins were then immunostained with the appropriate primary antibody and detected with either horseradish peroxidase-conjugated protein A (Amersham Pharmacia Biotech) or a horseradish peroxidase-coupled anti-mouse antibody (Roche Molecular Biochemicals) in a standard ECL reaction (Amersham Pharmacia Biotech) according to the manufacturer's instructions. Western blotting for qualitative analysis of calcium-induced proteins was performed three times with similar results.
Ras Activation Assay-Ras activation assays were performed as described (28). Cells were grown to 70% confluency and starved for 18 h in 0.5% FCS. After stimulation, cells were washed once with ice-cold PBS and lysed directly on the plate in Mg 2ϩ -containing lysis buffer (25 mM HEPES (pH 7.5), 150 mM NaCl, 10 mM MgCl 2 , 1 mM EDTA, 10% glycerol, 1% Nonidet P-40, 0.25% sodium deoxycholate, 1 mM sodium orthovanadate, 10 g/ml aprotinin, and 0.5 g/ml leupeptin) for 1 h at 4°C. Cells were scraped into microcentrifuge tubes and shock-frozen in liquid nitrogen to enhance lysis. Samples were thawed up again, and cell debris was removed by centrifugation at 15,000 ϫ g for 10 min. After measuring the protein contents (Bio-Rad protein assay), lysates were adjusted to equal protein levels and aliquoted into two portions of 500 g each of total cell lysate. One portion was frozen on liquid nitrogen and stored at Ϫ70°C for ERK2 kinase assays. The other portion was immediately subjected to precipitation with 20 g of the recombinant GST-fused Ras-binding domain of c-Raf-1 (amino acids 1-149) (28) freshly coupled to glutathione-Sepharose beads (Amersham Pharmacia Biotech) to pull down GTP-loaded Ras. After 1 h of precipitation at 4°C, the beads were washed three times with Mg 2ϩ -containing lysis buffer, and precipitates were separated by SDS-polyacrylamide gel electrophoresis using 14% SDS gels. Proteins were blotted onto polyvinylidene fluoride membranes (Millipore Corp.), and the amount of precipitated Ras-GTP was determined by Western blot analysis with an anti-pan Ras monoclonal antibody (Transduction Laboratories). Additionally, amounts of precipitated Ras-GTP were quantified by densitometric measurements using Tina 2.09d software (Raytest Isotopenmessgerä te GmbH). ERK2 kinase assays with the stored lysates were performed as described above.
Transfections-HaCaT cells were seeded onto 10-cm plates at a density of 1.2 ϫ 10 6 cells. After 2 days, the medium was replaced, and cells were transfected with an expression construct for HA-ERK2 and empty vector, Ras N17, or Raf C4 using FuGENE 6 transfection reagent. 18 l of FuGENE 6 were added to 582 l of calcium-free EMEM without any additives and incubated for 5 min at room temperature. Subsequently, the mixture was added dropwise to the DNA; and after 15 min of incubation at room temperature, the DNA/FuGENE 6 mixture was added to the cells. After 6 h, the medium was replaced again, and cells were starved 24 h later. Kinase assays were then performed as described above.

Elevation of Extracellular Calcium Leads to a Rapid and
Transient Activation of the Raf/MEK/ERK Pathway-To evaluate whether calcium-induced differentiation of keratinocytes is associated with activation of MAPK pathways, the spontaneously immortalized human keratinocyte cell line HaCaT (29) was stimulated by raising extracellular calcium levels from 0.05 to 1.0 mM, and activation of ERK, JNK, and p38 was analyzed in vitro by immune complex kinase assays. Elevation of extracellular calcium did not influence either JNK or p38 activity (Fig. 1, B and C), but resulted in rapid and transient activation of ERK (Fig. 1A, upper panel). An increase in ERK activity became apparent 10 min after the calcium switch, reached its maximum after 15 min, and rapidly decreased to basal levels thereafter. Similarly, calcium stimulation of primary NHEK induced a transient ERK activation with an identical time course as in HaCaT cells (data not shown). In contrast to calcium, EGF triggered activation of ERK with different kinetics (Fig. 1A). EGF treatment resulted in an im-mediate increase in ERK activity as early as 2 min after stimulation, and kinase activity was still found to be significantly elevated after 2 h (Fig. 1A, lower panel).
Since calcium is not a classical activator of the Raf/MEK/ ERK pathway, we next addressed the question whether calcium mediates its effect on ERK activation through the common Raf/MEK module or via an alternative route. To investigate if MEK is an upstream component of the pathway triggered by calcium, we analyzed the effect of two different MEK-specific inhibitors, PD98059 (30) and U0126 (31), on calcium-induced ERK activation. Both compounds inhibited calcium-induced ERK activation in a concentration-dependent manner ( Fig. 2A). Similarly, PD98059 completely abolished calcium-induced phosphorylation of ERK as shown by Western blot analysis using a phospho-specific antibody against ERK1 and ERK2 (data not shown). These results indicate that the increased activity of ERK in response to calcium stimulation is mediated by activation of MEK.
To determine whether elevation of extracellular calcium also leads to activation of serine/threonine kinases of the Raf family in HaCaT cells, we performed immune complex kinase assays for B-Raf and c-Raf-1 using kinase-inactive MEK as a substrate for immunoprecipitated B-Raf or c-Raf-1, respectively. Shifting calcium levels from 0.05 to 1.0 mM caused a rapid and transient increase in both c-Raf-1 and B-Raf activities, which correlated well in kinetics with the observed ERK activation (Fig. 2B). Expression of a dominant-negative mutant of Raf (Raf C4) completely blocked activation of cotransfected HA-tagged ERK2 by calcium (Fig. 2C), suggesting that calcium-induced ERK activation is mediated by activation of Raf isoforms.
Extracellular Calcium Induces ERK Activation through a Ras-independent Pathway-A common mode of Raf activation by growth factors is the recruitment of Raf to the plasma membrane by activated Ras. We therefore measured Ras activity in response to both calcium and EGF stimulation employing a precipitation assay for active GTP-loaded Ras in which the GTP-(but not the GDP)-loaded protein is selectively precipitated with a GST fusion protein of the Ras-binding domain of c-Raf-1 (28). We observed a time-dependent down-regulation of Ras activity after elevation of extracellular calcium levels, whereas EGF stimulation resulted in a 3-4-fold increase in Ras activity (Fig. 3A). When the same lysates were analyzed for ERK2 activation, the EGF-induced up-regulation of Ras activity correlated with a pronounced ERK activation; however, no such correlation between Ras and ERK activity was found after calcium stimulation. Elevation of extracellular calcium rather resulted in increased ERK activation levels at the same time that down-regulation of Ras activity was observed (Fig. 3B). To confirm that calcium-induced ERK activation occurs independently of Ras activity, HaCaT cells were transfected either with a dominant-negative mutant of Ras (Ras N17) or empty vector, and activation of cotransfected HA-ERK2 was analyzed after calcium stimulation. Fig. 4A shows that although expression of dominant-negative Ras N17 reduced both basal and calciuminduced ERK2 activities compared with the vector controls, the relative increase in activity remained roughly the same. In contrast, the same concentrations of Ras N17 completely blocked EGF-induced activation of coexpressed HA-ERK (Fig.  4B). Taken together, these results indicate that calcium-mediated ERK activation occurs through a Ras-independent pathway, whereas EGF-induced ERK activation is clearly dependent on Ras activation.
Calcium-induced ERK Activation Is Mediated by Intracellular Calcium, but Appears to Be Independent of PKC Activation-Elevation of extracellular calcium leads to a rise of cytosolic calcium levels in keratinocytes (32). This appears to be Cells were harvested at the indicated time points, and subjected to immune complex kinase assays for ERK (A), JNK (B), or p38 (C) as described under "Experimental Procedures." As positive control, cells in B and C were stimulated with 0.5 mM arsenite as described previously (24). Equal loading of immunoprecipitated kinases was assessed by Western blotting. mediated through a phospholipase C/inositol 1,4,5-trisphosphate-dependent pathway that mobilizes calcium from intracellular stores (10, 11), but might additionally result from direct influx of extracellular calcium through voltage-independent channels (33).
To investigate whether reduction of intracellular free calcium affects ERK activity, we examined the effect of BAPTA-AM, a chelator of intracellular calcium that has previously been shown to inhibit differentiation of keratinocytes (34). Ha-CaT cells were pretreated with BAPTA-AM and incubated in high calcium medium (1.0 mM) for 15 min. Cells were then lysed and analyzed for ERK2 activity. Fig. 5A shows that preincubation with 30 M BAPTA-AM almost completely abrogated ERK activation induced by extracellular calcium, whereas concentrations up to 50 M BAPTA-AM did not affect EGF-induced ERK activation. These data indicate that ERK activation by extracellular calcium (but not by EGF) is dependent on the presence of free calcium within the cell.
If extracellular calcium exerts its effect on ERK activation by elevation of intracellular calcium levels, agents that artificially increase cytosolic calcium levels should also be capable of activating the Raf/MEK/ERK pathway. To evaluate this hypothesis, we stimulated HaCaT cells with 100 nM thapsigargin, which triggers calcium release from intracellular stores, and measured the kinetics of ERK activation by immune complex kinase assays. Thapsigargin treatment resulted in a transient increase in ERK activity similar to that observed after stimulation with extracellular calcium (Fig. 5B). This implies a model in which activation of the Raf/MEK/ERK pathway by extracellular calcium is transduced by an intracellular pathway that involves an intermediate step critically dependent on availability of free calcium within the cell.
Among the kinases known to be regulated by calcium are certain isoforms of the PKC family, namely the classical PKCs (␣, ␤1, ␤2, and ␥) (35). Since some PKC isoforms (most importantly, PKC␣ and PKC⑀) have been shown to play a role in

FIG. 2. Calcium-induced ERK activation is mediated by Raf isoforms and MEK.
A, calcium-induced ERK activation is sensitive to MEK inhibition. HaCaT cells were stimulated with 1.0 mM calcium in the absence or presence of different concentrations of the MEK inhibitors PD98059 and U0126, and ERK2 activation was determined by immune complex kinase assays. The concentrations of PD98059 used in this experiment were 5, 10, and 20 M; U0126 was utilized at concentrations of 1, 5, and 10 M. Parallel experiments in which cells were treated with corresponding concentrations of Me 2 SO instead of PD98059 or U0126 showed no effect on calcium-induced ERK activation (data not shown). B, kinetics of c-Raf and B-Raf activation by calcium. HaCaT cells were stimulated with 1.0 mM calcium or 1 ng/ml EGF for the indicated time intervals. Cells were harvested, and the activity of immunoprecipitated c-Raf-1 or B-Raf was determined by immune complex kinase assays using dominant-negative MEK (dnMEK) as a substrate. Equal loading of the immunoprecipitated kinases in A and B was confirmed by Western blotting. C, calcium-induced ERK activation is dependent on Raf activation. HaCaT cells were transfected with dominant-negative Raf C4 or empty vector, and activation of cotransfected HA-ERK2 was determined by immune complex kinase assay. For the calcium-stimulated samples, cells were harvested 15 min after stimulation with 1.0 mM calcium. Equal immunoprecipitation of HA-ERK2 was confirmed by Western blotting.

FIG. 3. Calcium-induced ERK activation is associated with concomitant down-regulation of Ras activity.
HaCaT cells were stimulated with 1.0 mM calcium or 1 ng/ml EGF and harvested at the indicated time points. Protein contents in the lysates were adjusted to equal amounts and divided into two portions. One aliquot was subjected to a Ras activation assay as described under "Experimental Procedures"; the other portion was analyzed for ERK activity by immune complex kinase assay. A: upper panel, GTP-loaded active Ras was precipitated with GST-Raf-1 Ras-binding domain-(1-149) and quantified by Western blotting with an anti-pan Ras antibody (note that a longer exposure time was chosen to visualize calcium-induced downregulation of basal Ras activity). To ensure that differences in Ras GTP were not due to quantitative differences in total cellular Ras, an immunoblot for Ras is shown. Lower panel, shown is ERK activation as quantified by phosphorylation of myelin basic protein (MBP). Equal precipitation of ERK was confirmed by Western blotting. B: shown is the quantification of Ras activity (gray bars) and ERK activity (black bars) upon calcium or EGF stimulation. Values of two independent experiments are depicted as percent of basal activity (mean Ϯ S.E.). calcium-induced differentiation of mouse keratinocytes (12) and both isoforms are potent activators of c-Raf-1 in other cell types (36,37), these enzymes were candidate mediators of calcium-induced activation of the Raf/MEK/ERK pathway. We therefore analyzed the potential of three different PKC inhibitors to interfere with calcium-induced ERK activation. Neither the broad-range PKC inhibitors GF 109203X and Ro-31-8220, which both potently inhibit all classical PKC isoforms and PKC⑀, nor the more PKC␣-specific compound Ro-32-0432 significantly reduced calcium-induced ERK activation at concentrations that effectively blocked 12-O-tetradecanoylphorbol-13acetate-induced gene expression in A301 T-cells ( Fig. 5C and data not shown). These data argue against a prominent role of PKCs in calcium-induced ERK activation.
Calcium-induced Up-regulation of the Cyclin-dependent Kinase Inhibitor p21/Cip1 Is Sensitive to MEK Inhibition-Calcium stimulation resulted in a rapid ERK activation, which raises the possibility that the Raf/MEK/ERK pathway plays a role in the early stages of keratinocyte differentiation. An initial step in the differentiation process is cell cycle arrest (38). Among the marker molecules that can be used to analyze the proliferative status of cells are cyclin-dependent kinase inhibitor proteins (39). In mouse keratinocytes, elevation of extracellular calcium levels induces cell cycle arrest, which is associated with up-regulation of the cyclin-dependent kinase inhibitor p21/Cip1 (40). Similarly, calcium stimulation increases p21/Cip1 expression in human keratinocytes (41). Since several reports have implicated components of the Raf/MEK/ ERK pathway in cell cycle control in other cells (42), we analyzed whether inhibition of MEK activation is able to block calcium-induced up-regulation of p21/Cip1 protein expression in HaCaT cells. Fig. 6 shows that calcium-induced p21/Cip1 expression was completely abolished by PD98059, indicating that this process involves activation of the Raf/MEK/ERK pathway.
Calcium-induced Up-regulation of the Keratinocyte Differentiation Marker Involucrin Is Sensitive to Inhibition of MEK-To further analyze whether the Raf/MEK/ERK cascade is involved in the determination of the differentiation process, we investigated if inhibition of the pathway at the level of MEK interferes with calcium-induced up-regulation of the differentiation marker involucrin. Primary NHEK were seeded in low calcium medium supplemented with 0.05 mM calcium and then exposed to high calcium medium (1.0 mM) for 48 h in the presence or absence of 20 M PD98059 or 10 M U0126. Subsequently, cells were lysed, and the amount of involucrin was determined by Western blot analysis. Whereas calcium-induced up-regulation of involucrin was partially inhibited by PD98059, 10 M U0126 was sufficient to reduce involucrin expression to basal levels (Fig. 7A), which is consistent with the enhanced potential of U0126 to inhibit MEK activity (31). Reduction was not due to a general inhibition of protein synthesis or unequal loading since equal amounts of ERK2 were detected when the same blot was probed with an antibody for ERK2. Further control experiments in which NHEK were cultured for 48 h in high calcium Cells were then lysed, and ERK2 activation was assessed by immune complex kinase assay. Incubation of cells with the solvent Me 2 SO at corresponding concentrations did not alter ERK activity (data not shown). B, HaCaT cells were grown in low calcium medium, starved as described as described for A, and stimulated with 100 nM thapsigargin for the indicated times. Subsequently, cells were lysed, and ERK2 activity was examined. C, HaCaT cells were grown under low calcium conditions as described for A. Cells were either left untreated or incubated for 30 min with 3 M GF 109203X, 100 nM Ro-31-0432, or 1 M Ro-32-8220. Cells were stimulated with 1.0 mM calcium for 15 min, and ERK2 activity in the lysates was compared with that in unstimulated cells. Equal loading of immunoprecipitated ERK in A-C was assessed by Western blotting. medium containing equivalent amounts of the solvent Me 2 SO instead of the inhibitors revealed no inhibitory effect of Me 2 SO on involucrin expression (data not shown). In contrast to calcium, EGF stimulation of NHEK did not result in increased involucrin levels, indicating that ERK activation alone is not sufficient to induce its expression (Fig. 7B). This implies that appropriate timing of ERK activation and/or activation of additional pathways is required to augment involucrin expression. DISCUSSION In this report, the role of MAPK pathways in calcium-induced differentiation of human keratinocytes was examined. We observed that elevation of extracellular calcium levels resulted in a rapid and transient activation of ERK, but not JNK and p38. ERK activation occurred independently of Ras activity since expression of dominant-negative Ras N17 did not prevent calcium-induced ERK activation. Furthermore, calcium stimulation led to down-regulation of Ras activity in HaCaT cells, which is consistent with previous observations of Medema et al. (43) in primary keratinocytes. In their study, elevation of ex-tracellular calcium did not result in ERK activation, but inhibited EGF-induced ERK activation when preincubated. Interestingly, inhibition was transient and only effective within a narrow time window of calcium pretreatment. Since, in our experiments, calcium triggered a transient, peak-like ERK activation that was associated with concomitant down-regulation of Ras activity, it is conceivable that calcium stimulation renders the cell temporarily insensitive to mitogenic ERK activation, which may account for its inhibitory effect on EGF-induced ERK activation. It is tempting to speculate that this transient down-regulation of Ras activity may represent a kind of "safety mechanism" that prevents an immediate second wave of ERK activation by other stimuli that may disturb the cellular response to the calcium signal.
Although Ras activity is down-regulated after calcium treatment, the direct downstream effectors of Ras, Raf, and MEK become activated. It is still unclear whether calcium-induced Raf activation is mediated by GTPases different from Ras or by a GTPase-independent mechanism, and experiments are underway to address this question.
Both HaCaT cells and primary NHEK express the MEK activators c-Raf-1 and B-Raf ( Fig. 2 and data not shown). Since activation of both isoforms was observed upon calcium stimulation, we presently cannot decide which isoform mediates the response. In PC12 cells, sustained ERK activation induced by NGF was reported to be a result of B-Raf activation, which may reflect isozyme-specific control of proliferation and differentiation in neuronal cells (18,44). Although our experiments do not rule out that Raf isozymes have distinct functions in keratinocytes, the similar activation kinetics of both Raf isoforms rather argue against such a notion and predict a common mode of regulation for both isozymes in keratinocytes. On the other hand, the recent finding that c-Raf-1 mutant mice exhibited a defect in skin differentiation particularly emphasizes the role of c-Raf-1 in the regulation of keratinocyte growth and differentiation in vivo (45), even though B-Raf may also be required.
Activation of the Raf/MEK/ERK pathway by extracellular calcium appears to be mediated by a mechanism that triggers elevation of cytoplasmic calcium levels. In accordance with our observation that ERK activation by extracellular calcium is mediated by Raf isoforms, a previous study reported Raf-dependent activation of ERK by elevation of intracellular calcium levels in mouse 3T3 fibroblasts (46). Recently, a receptor for extracellular calcium has been cloned from both human and mouse keratinocytes that might provide the missing link between elevation of extracellular calcium and increased intracellular calcium levels (47,48). Mice deficient in the full-length form of this receptor show defective skin development, and keratinocytes of these mice do not respond with elevation of cytosolic calcium levels after stimulation with extracellular calcium (48). Interestingly, a homolog of this receptor is required for ERK activation in response to extracellular calcium in Rat1 fibroblasts (49). Although the underlying mechanism is not resolved yet, calcium receptors are putative candidates for the mediation of calcium-induced ERK activation in keratinocytes.
An early step in the differentiation process of keratinocytes is cell cycle arrest. Our observation that calcium-induced expression of the cyclin-dependent kinase inhibitor p21/Cip1 is sensitive to MEK inhibition provides a first hint that the Raf/ MEK/ERK pathway plays an important role in the initial switch from growth to differentiation. Whether activation of the Raf/MEK/ERK pathway is the actual trigger for calciuminduced cell cycle arrest has to await further experiments; however, such a view would be consistent with previous data from other cell types that have implicated this pathway in cell FIG. 6. Calcium-induced up-regulation of p21/Cip1 depends on MEK activation. HaCaT cells grown to 50 -70% confluency in low calcium medium were starved for 18 h in 0.5% FCS and then exposed for 24 h to high calcium medium (1.0 mM) in the presence or absence of 20 M PD98059. Cells were lysed, and p21/Cip1 protein was detected by Western blotting. Equal loading of total protein was confirmed by staining the same blot with antiserum against ERK2.
FIG. 7. Calcium-induced up-regulation of the differentiation marker involucrin is sensitive to MEK inhibition. A, primary NHEK were seeded in serum-free medium containing 0.05 mM calcium until they were 50% confluent. The cells were then either fed with fresh medium and grown for additional 48 h or incubated with fresh medium containing 1.0 mM calcium in the presence or absence of 10 M U0126 or 20 M PD98059. Thereafter, cells were lysed and analyzed for involucrin expression by Western blotting. B, NHEK were stimulated for 48 h with 50 ng/ml EGF or 1.0 mM calcium. Involucrin levels were compared with those in unstimulated NHEK by Western blot analysis. Equal loading of total protein in A and B was confirmed by staining the same blot with antiserum against ERK2. cycle control (42). Since transforming growth factor-␤-induced expression of p21/Cip1 is also dependent on MEK in keratinocytes (50), activation of the Raf/MEK/ERK pathway may be a common mechanism by which several stimuli induce cell cycle arrest in these cells.
Calcium-induced expression of the differentiation marker involucrin was sensitive to MEK inhibition. This suggests that activation of the Raf/MEK/ERK pathway directly or indirectly affects differentiation-specific gene expression. Since EGF stimulation did not result in increased involucrin expression in primary NHEK, we conclude that activation of the ERK pathway alone is not sufficient to induce involucrin synthesis. Although the different kinetics of ERK activation by calcium and EGF may account for the reciprocal effect of both stimuli on involucrin expression, we favor a model in which calcium triggers activation of multiple pathways that act in concert with the Raf/MEK/ERK pathway to induce the differentiation program in keratinocytes. An early event occurring upon calciuminduced differentiation is enhanced tyrosine phosphorylation (51). Particularly phosphorylation of a protein of ϳ60 kDa (p62) that associates with the Ras GTPase-activating protein (GAP) has attained much attention since it was reported to become phosphorylated after both calcium treatment of mouse keratinocytes (52) and disruption of integrin-mediated cell-matrix interactions (53). Down-regulation of surface integrins and loss of adhesiveness to the extracellular matrix are among the first changes to occur during the differentiation process in vivo (54), which suggests that phosphorylation of p62 in response to calcium stimulation may mimic the signal induced by integrin loss in vivo. p62 phosphorylation occurs as early as 5 min after calcium stimulation, but is not observed after stimulation with EGF (52). Interestingly, calcium-induced phosphorylation of p62 is associated with translocation of Ras GAP to the membrane. Since Ras GAPs catalyze the conversion of GTP-loaded active Ras to its inactive GDP-loaded form, translocation of Ras GAP to the membrane may be responsible for the calciuminduced down-regulation of Ras activity observed in our experiments, although it is presently unclear if translocation of Ras GAP is sufficient to activate its enzymatic activity. Taken together, our data provide evidence that the Raf/MEK/ERK pathway plays a crucial role in the regulation of keratinocyte growth and differentiation even though other pathways may be important in the modulation of the cellular response upon its activation.