PKCδ Activation: A Divergence Point in the Signaling of Insulin and IGF-1—Induced Proliferation of Skin Keratinocytes

Materials. Tissue culture media and serum were purchased from Biological Industries (Beit HaEmek, Israel). Enhanced chemiluminescence was performed with a kit purchased from Bio-Rad (Israel). Polyclonal antibodies to Na⁺/K⁺ pump isoforms and monoclonal anti—p-tyr antibody were purchased from Upstate Biotechnology (Lake Placid, NY). Polyclonal and monoclonal antibodies to PKC isoforms were purchased from Santa Cruz (California, USA) and Transduction laboratories (Lexington, KY). Horseradish peroxidase—anti-rabbit and anti-mouse IgG were obtained from Bio-Rad (Israel). Leupeptin, aprotinin, phenylmethylsulfonyl fluoride (PMSF),dithiothreitol (DTT), Na-orthovanadate, and pepstain were purchased from Sigma Chemicals (St. Louis, MO). Insulin (recombinant human insulin [humulinR]) was purchased from Eli Lilly France SA (Fergersheim, France). IGF-1 was a gift from Cytolab (Israel).

Isolation and culture of murine keratinocytes. Primary keratinocytes were isolated from newborn BALB/C mice as described(25). Keratinocytes were cultured in Eagle's minimal essential medium containing 8% Chelex-(Chelex-100, Bio-Rad) treated fetal calf serum. To maintain a proliferative basal cell phenotype, the final Ca²⁺ concentration was adjusted to 0.05 mmol/l. Experiments were performed 5-7 days after plating.

Preparation of cell extracts and Western blot analysis. For crude membrane fractions, whole-cell lysates were prepared by scraping cells into phosphate-buffered saline (PBS) containing 10 μg/ml aprotinin, 10 μg/ml leupeptin, 2 μg/ml pepstatin, 1 mmol/l PMSF, 10 mmol/l EDTA, 200 μmol/l NaVO₄, and 10 mmol/l NaF. After homogenization and four freeze/thaw cycles, lysates were spun down at 4°C for 20 min in a microcentrifuge at maximal speed. The supernatant containing the soluble cytosol protein fraction was transferred to another tube. The pellet was resuspended in 250 μl PBS containing 1% Triton X-100 with protease and phosphatase inhibitors, incubated for 30 min at 4°C, and spun down in a microcentrifuge at maximal speed at 4°C. The supernatant contains the membrane fraction. Protein concentrations were measured using a modified Lowry assay (Protein Assay Kit;Bio-Rad). Western blot analysis of cellular protein fractions was carried out as described (26).

Preparation of cell lysates for immunoprecipitation. Culture dishes containing keratinocytes were washed with Ca²⁺/Mg²⁺—free PBS. Cells were mechanically detached in radioimmunoprecipitation assay (RIPA) buffer (50 mmol/l Tris HCl,pH 7.4, 150 mmol/l NaCl, 1 mmol/l EDTA, 10 mmol/l NaF, 1% Triton X-100, 0.1%SDS, and 1% Na deoxycholate) containing a cocktail of protease and phosphatase inhibitors (20 μg/ml leupeptin, 10 μg/ml aprotinin, 0.1 mmol/l PMSF, 1 mmol/l DTT, 200 μmol/l orthovanadate; and 2 μg/ml pepstatin). The preparation was centrifuged in a microcentrifuge at maximal speed for 20 min at 4°C. The supernatant was used for immunoprecipitation.

Immunoprecipitation. The lysate was precleared by mixing 0.3 ml of cell lysate with 25 μl of Protein A/G Sepharose (Santa Cruz, CA), and the suspension was rotated continuously for 30 min at 4°C. The preparation was then centrifuged at maximal speed at 4°C for 10 min, and 30 μl of A/G Sepharose was added to the supernatant along with specific polyclonal or monoclonal antibodies to the individual PKC isoforms (dilution 1:100). The samples were rotated overnight at 4°C. The suspension was then centrifuged at maximal speed for 10 min at 4°C, and the pellet was washed with RIPA buffer. The suspension was again centrifuged at 15,000g (4°C for 10 min) and washed four times in TBST. Sample buffer (0.5M Tris HCl, pH 6.8,10% SDS, 10% glycerol, 4% 2 β-mercaptoethanol, and 0.05% bromophenol blue) was added and the samples were boiled for 5 min and then subjected to SDS-PAGE.

Adenovirus constructs. The recombinant adenovirus vectors were constructed as described (27). The dominant negative mutant of mouse PKCδ was generated by the substitution of the lysine residue at the ATP-binding site with alanine(28). The mutant delta cDNA was cut from SRD expression vector with EcoR I and ligated into the pAxCA1w cosmid cassette to construct the Ax vector. The dominant negative activity of this gene was demonstrated by the abrogation of its autophosphorylation activity (29).

Transduction of keratinocytes with PKC isoform genes. The culture medium was aspirated and keratinocyte cultures were infected with the viral supernatant (29) containing PKCδ recombinant adenoviruses for 1 h. The cultures were then washed twice with low Ca²⁺-containing minimum essential medium (MEM) and refed. Cells were transferred 10-h postinfection to serum-free low Ca²⁺-containing MEM for 24 h. Keratinocytes from control and insulin-treated cultures were used for proliferation assays, ⁸⁶Rb uptake, or extracted and fractionated into cytosol and membrane fractions for immunoprecipitation and Western blotting.

PKC activity. Specific PKC activity was determined in freshly prepared immunoprecipitates from keratinocyte cultures after appropriate treatments. These lysates were prepared in RIPA buffer without NaF. Activity was measured with the use of the SignaTECT PKC assay system (Promega, Madison,WI) according to the manufacturer's instructions. PKCα pseudosubstrate was used as the substrate in these studies.

Cell proliferation. Cell proliferation was measured by[³H]thymidine incorporation in 24-well plates. Cells were pulsed with [³H]thymidine (1 μCi/ml) overnight. After incubation, cells were washed five times with PBS and 5% thrichloracetic acid was added to each well for 30 min. The solution was removed and cells were solubilized in 1%Triton X-100. The labeled thymidine incorporated into cells was counted in a ³H-window of a Tricarb liquid scintillation counter.

Na⁺/K⁺ pump activity.Na⁺/K⁺ pump activity was determined by the measurements of ouabain-sensitive uptake of ⁸⁶Rb by whole cells in 1 ml of K⁺-free PBS containing 2 mmol/l RbCl and 2.5 μCi of ⁸⁶Rb (30). Rb uptake was terminated after 15 min by aspiration of the medium, after which the cells were rinsed rapidly four times in cold 4°C K⁺-free PBS and solubilized in 1% Triton X-100. The cells from the dish were added to 3 ml H₂O in a scintillation vial. Samples were counted in a ³H-window of a Tricarb liquid scintillation counter. Rb-uptake specifically related to Na⁺/K⁺ pump activity was determined by subtraction of the counts per minute accumulated in the presence of 10^-4 mol/l ouabain from the uptake determined in the absence of the inhibitor.

Effects of insulin and IGF-1 on keratinocyte proliferation.Initially we wanted to characterize the mitogenic effects of both insulin and IGF-1 on skin keratinocytes. The ability of the hormones to induce keratinocyte proliferation was evaluated by measuring thymidine incorporation. As shown in Fig. 1A,both insulin and IGF-1 stimulated thymidine incorporation in a dose-dependent manner with maximal induction achieved at 10^-7 and 10^-8mol/l, respectively. At each concentration, the maximal stimulation by IGF-1 was greater than that by insulin. Interestingly, when both hormones were given together, their mitogenic effects were additive at all concentrations tested(Fig. 1B). These results suggest that insulin and IGF-1 regulate keratinocyte proliferation through distinct pathways.

Effects of insulin and IGF-1 on regulation of the Na⁺/K⁺ pump. We next attempted to identify the possible downstream elements that could serve as a divergence point in mediating insulin- and IGF-1—induced proliferation. We initially examined the effects of insulin and IGF-1 on Na⁺/K⁺ pump activity. The Na⁺/K⁺ pump is an established regulator of proliferation and differentiation of keratinocytes and is known to be regulated by insulin. Figure 2A demonstrates the effects of insulin and IGF-1 on Na⁺/K⁺ pump activity as measured by ouabain-sensitive ⁸⁶Rb uptake. As seen, insulin but not IGF-1 significantly increased pump activity.

Next, we examined the effects of insulin and IGF-1 on Na⁺/K⁺ pump protein isoform expression(Fig. 2B). Skin keratinocytes express the α₁, α₂,α₃, β₁, and β₂ isoforms of the Na⁺/K⁺ pump. After insulin stimulation, expression of the α₂ and α₃ but not theα₁ isoforms was increased as early as 30 min after stimulation (Fig. 2B). The elevated expression was maintained for up to 24 h(Fig. 2B). No change was observed in the protein expression of the β₁ orβ₂ subunits (results not shown). Consistent with the lack of effect of IGF-1 on Na⁺/K⁺ pump activity, this hormone did not affect protein expression of either the α(Fig. 2B) or β(data not shown) subunits. Interestingly, in contrast to the differential effects of insulin and IGF-1 on the Na⁺/K⁺ pump, both factors similarly activated other immediate downstream elements of the insulin- and IGF-1—signaling pathway. These included the phosphorylation and activation of IRS1, IRS2, MAPK, and PI3K (results not shown). Because the Na⁺/K⁺ pump activity plays a role in skin proliferation,we next wanted to determine whether the distinct regulation of Na⁺/K⁺ pump activity by insulin is associated with keratinocyte proliferation. Thus, we studied the effects of insulin and IGF-1 on keratinocyte proliferation in cells that were pretreated with ouabain, a specific inhibitor of the Na⁺/K⁺ pump. As shown in Fig. 3, ouabain(10^-4 mol/l) completely blocked insulin-induced thymidine incorporation. In contrast, the proliferative effects of IGF-1 were essentially unaffected by ouabain. Moreover, the addition of ouabain with both insulin and IGF-1 reduced the increase in thymidine incorporation to the level induced by IGF-1 alone. Thus, the ability of ouabain to block only the insulin-associated component of proliferation further suggests that insulin and IGF-1 use different signaling pathways to induce their respective proliferative effects.

Effects of insulin and IGF-1 on PKC isoform translocation and activity. PKC is another major signaling pathway, which mediates keratinocyte proliferation and differentiation(28,31,32)and was shown in other tissues to be regulated by insulin signaling(33,34,35). In skin, PKC isoforms α, δ, ϵ, η, and ζ are expressed (36). Because the activation of PKC isoforms is associated with their translocation to membrane fractions, we first examined the effects of insulin and IGF-1 on translocation of the various PKC isoforms from cytosol to the membrane. As seen in Fig. 4B, as early as 1 min after stimulation, insulin specifically induced translocation of PKCδ from the cytosol to the membrane fractions. Membrane expression of PKCδ was maintained for several hours after insulin stimulation. In contrast, IGF-1 reduced PKCδ expression in the membrane and increased its relative level of expression in the cytosol fraction. No change in distribution of the other PKC isoforms was seen after stimulation by either insulin of IGF-1 (Fig. 4A). Interestingly, whereas stimulation with epidermal growth factor (EGF) and high calcium concentrations induced tyrosine phosphorylation of PKCδ, neither insulin nor IGF-1 induced tyrosine phosphorylation of the PKCδ isoform(Fig. 4D). To determine if the differential regulation of PKCδ could be mediated by the Na⁺/K⁺ pump, we further analyzed the effects of ouabain on the expression and translocation of PKCδ. As seen in Fig.4C, ouabain, the Na⁺/K⁺ pump inhibitor, did not affect PKCδdistribution or expression in nonstimulated cells. Furthermore, ouabain did not interfere with insulin-induced translocation of PKCδ.

To determine whether the translocation of PKCδ is sufficient for its activation, we next measured kinase activity of PKC immunoprecipitates from the cytoplasmic and membrane fractions of insulin- and IGF-1—treated keratinocytes. As shown in Fig. 5, insulin but not IGF-1 increased activity of PKCδ in the membrane fraction. No elevation in PKCδ activity was observed in the cytoplasmic fraction. The insulin-induced activation was specific for PKCδ and no activation of PKCs α, ϵ, η, or ζ was observed for up to 30 min after insulin stimulation (not shown). Altogether,these results suggest selective PKCδ activation specifically by insulin but not by IGF-1 stimulation.

To specifically link insulin-induced PKCδ activation to insulin-induced keratinocyte proliferation we used rottlerin, a specific inhibitor of PKCδ, and studied its effects on insulin-induced proliferation. As seen in Fig. 6, rottlerin inhibited keratinocyte proliferation induced by insulin. In contrast, wortmanin, a PI3K inhibitor, did not have any effect on insulin induced proliferation. These results suggest that insulin-induced proliferation is independent of PI3K but is specifically linked to PKCδactivation.

To directly study the association between insulin-induced PKCδactivation and insulin-induced keratinocyte proliferation, we used recombinant PKC adenovirus constructs to overexpress both wild-type PKCδ(WTPKCδ) as well as a kinase-inactive dominant-negative PKCδ(DNPKCδ), which abrogates the endogenous PKCδ activity. Both constructs, as well as a PKCδ construct, were efficiently expressed in keratinocytes (Fig. 7A). Furthermore, overexpressing PKCδ and PKCα induced an increase in isoform-specific PKC activity several fold above control levels (Fig. 7B). Next, we followed the effects of overexpressing WTPKCδ and DNPKCδ on insulin-induced keratinocyte proliferation. As can be seen in Fig. 8A, overexpression of WTPKCδ without insulin treatment, but not overexpression of PKCα, increased thymidine incorporation. The increase was similar to the increase induced by insulin in control cells. Moreover, insulin could not further increase the upregulated proliferation of the WTPKCδ overexpressing cells. In contrast,stimulation by IGF-1 increased thymidine incorporation in a similar manner in both noninfected cells and in cells overexpressing WTPKCδ and PKCα(Fig. 8A). These results indicate that insulin, but not IGF-1, mediates proliferation of keratinocytes through a pathway involving PKCδ.

The direct involvement of PKCδ in insulin-induced proliferation was further proven by abrogating PKCδ activity. As seen in Fig. 8B, basal thymidine incorporation in cells overexpressing the DNPKCδ was slightly, but significantly, lower than that in noninfected cells. However, overexpression of DNPKCδ completely eliminated insulin-induced proliferation but did not affect IGF-1—induced proliferation. Moreover, the additive effects of insulin and IGF-1 were reduced to that of IGF-1 alone.

Finally, the specificity of PKCδ activation to the insulin-mediated pathway was analyzed by investigating the effects of DNPKCδ mutant on the mitogenic response to a variety of growth factors including the following:IGF-1, EGF, keratinocyte growth factor (KGF), endothelial cell growth factor(EcGF), and platelet-derived growth factor (PDGF). As seen in Fig. 9, the overexpression of DNPKCδ selectively eliminated the proliferative effects induced by insulin but did not block those of any of the other growth factors tested.

Effects of overexpressed WTPKCδ and DNPKCδ on insulin-induced ⁸⁶Rb uptake. Our results so far demonstrate that insulin-induced proliferation is selectively mediated by activation of PKCδ and is associated with stimulation of Na⁺/K⁺pump activity. To demonstrate that these effects are causally related, we examined effects of WTPKCδ or DNPKCδ on insulin-induced Na⁺/K⁺ pump activity. As can be seen in Fig. 10A,overexpression of WTPKCδ increased resting pump activity to a level similar to that induced by insulin. Insulin did not cause a further increase in pump activity in the cells overexpressing WTPKCδ. Furthermore,DNPKCδ significantly reduced resting pump activity and blocked the insulin-induced stimulation of the pump to a level lower than basal pump activity in control unstimulated cells. Finally, we examined the effects of WTPKCδ and DNPKCδ on the expression of Na⁺/K⁺ pump isoforms(Fig. 10B). Interestingly, whereas insulin induced the expression of α₂and α₃ isoforms, overexpression of PKCδ increased expression of α₂ similarly to insulin stimulation, and insulin could not further increase α₂ expression. In contrast, no change in α₃ isoform expression was observed and WTPKCδ did not interfere with insulin-induced expression ofα₃. Furthermore, abrogating PKCδ activation by overexpressing DNPKCδ completely inhibited insulin-inducedα₂ expression but had no effect on insulin-induced expression of the α₃ isoform (Fig. 10B).

Insulin and IGF-1 exert their mitogenic and metabolic effects in different tissues via distinct receptors(4,8). Both insulin and IGF-1 are essential for the growth and maintenance of several cell types including keratinocytes in culture and are essential components of the growth medium of these cells(9). However, skin is not considered to be a classic insulin responsive tissue, because glucose transport is not induced in response to acute insulin stimulation. Therefore,the effects of insulin in skin were mostly attributed to its ability to activate the closely related IGFR(1). We have previously shown that in keratinocytes, insulin and IGF-1 can both stimulate receptors and activate similar downstream effectors(37). However, the current study demonstrates that whereas both growth factors induce keratinocyte proliferation in a dose-dependent manner, each hormone exert its effects through distinct signaling pathways. Our initial indication for differential regulation of keratinocyte proliferation by insulin and IGF-1 was confirmed by our finding that these hormones had additive effects on keratinocyte proliferation when given together, at maximal proliferation-inducing concentrations for each hormone (Fig. 1A,1B). To identify the divergence point in insulin- and IGF-1—signaling pathway in regulation of keratinocyte proliferation, we investigated elements known to both regulate keratinocyte proliferation and to act as downstream effectors of insulin signaling. These studies revealed that insulin but not IGF-1 signaling is mediated by PKCδ and involves the stimulation of the Na⁺/K⁺ pump.

In this study, we determined that the Na⁺/K⁺ pump actively participates in transmitting insulin but not IGF-1 signals, leading to keratinocyte proliferation. Insulin-induced Na⁺/K⁺pump activation was associated with selective increases in expression of theα₂ and α₃ Na⁺/K⁺ pump subunit isoforms. The significant role of the Na⁺/K⁺pump in the insulin-signaling pathway was also confirmed pharmacologically by treatment of the cells with ouabain, a selective inhibitor of the Na⁺/K⁺ pump. Pretreatment of keratinocytes with ouabain completely blocked insulin-induced proliferation of keratinocytes but did not affect proliferation induced by IGF-1. Furthermore, in studies in which additive effects of insulin and IGF-1 were examined, ouabain inhibited only the insulin component and reduced proliferation to a level induced by stimulation with IGF-1 alone. These findings demonstrate the involvement of the Na⁺/K⁺ pump in mitogenic effects of insulin and further strengthen the idea that insulin and IGF-1 act via separate signaling pathways to induce keratinocyte proliferation.

Na⁺/K⁺ pump activity has been demonstrated to be regulated by a variety of hormones in different tissues(17). After pump activation,the Na⁺/K⁺ gradient provides the force for active transport of amino acids, phosphate, and glucose. Several studies have suggested the involvement of the Na⁺/K⁺ pump in regulation of cellular proliferation in variety of cell types(18,38,39,40,41). However, whereas the activation of the Na⁺/K⁺ pump was known to be an important target of insulin action(42,43),this is the first study that directly implicates specific regulation of the Na⁺/K⁺ pump in insulin-induced keratinocyte proliferation. The modulation of Na⁺/K⁺ pump activity is thought to be regulated by direct phosphorylation and dephosphorylation of Na⁺/K⁺ pump isoforms by protein kinases and protein phosphatases(44,45). Specifically, PKC phosphorylation of the α subunits of the Na⁺/K⁺ pump was shown to affect the activation state of the Na⁺/K⁺ pump in vitro and in vivo(19,46,47,48). However, the functional significance of the PKC-mediated changes in the phosphorylation state of the Na⁺/K⁺ pump has not been conclusively demonstrated. Furthermore, as the majority of the studies used nonspecific pharmacological activators and inhibitors of PKC, a specific PKC isoform or a distinct function for PKC could not be identified. In this study,we directly linked hormonal stimulation of the Na⁺/K⁺pump to specific activation of PKCδ leading to the induction of cellular proliferation. Activation of the Na⁺/K⁺ pump by overexpression of PKCδ and the fact that insulin could not further increase this effect indicate that a common pathway is involved. Moreover, the blockade of insulin-induced Na⁺/K⁺ pump activity by overexpression of a DNPKCδ mutant and the ability of ouabain, a specific pump inhibitor, to abolish the effects of insulin on proliferation without abrogating insulin-induced activation of PKCδ places the Na⁺/K⁺ pump downstream of insulin-mediated PKC activation. However, whereas insulin stimulation—induced expression of both α₂ and α₃ isoforms, insulin-induced PKCδ activation was only associated with changes in α₂expression. Induction of Na⁺/K⁺ pump activity and isoform expression has been linked to cell proliferation in different cell systems(49,50,51). However, this is the first report linking the insulin-induced proliferation with PKCδ- mediated induction of the Na⁺/K⁺ pump in keratinocytes. These results are in accordance with the existence of an ion gradient in skin in vivo and with the well-documented effects of Ca²⁺, K⁺, and Na⁺ ions on keratinocyte proliferation and differentiation(52,53,54). These observations are consistent with a role for both insulin and the Na⁺/K⁺ pump in keratinocyte proliferation and may explain the significance of insulin as an essential component of growth medium of cultured keratinocytes.

Several isoforms of PKC, including α, δ, and η, have been shown to regulate growth and differentiation of skin keratinocytes(28,31,55). Our results provide further evidence for the role of PKCδ in keratinocyte proliferation. PKCδ is a unique isoform among the PKC family of proteins involved specifically in growth and maturation of various cell types (56). This isoform was shown to participate in apoptosis(57,58)differentiation(59,60)and cell-cycle retardation or arrest(61,62). However, PKCδ was also shown to be specifically regulated by stimulation of several growth factors including EGF, PDGF, and neurotransmitters, as well as by the mitogenic signal by v-src and the oncogenic form of c-Ha-ras(59,63,64,65,66). Changes in PKCδ regulation are usually associated with its translocation to membranal fractions, tyrosine phosphorylation of the enzyme, and activation or deactivation of its intrinsic kinase activity(60,64). In several of these studies, PKCδ tyrosine phosphorylation was associated with inhibition of PKCδ activity or degradation of the enzyme(64,65,67). In the current study, we found that insulin-induced PKCδ activity was not associated with induction of tyrosine phosphorylation. Rather, PKC activation was associated with translocation of the enzyme and stable expression of PKCδ in the membrane fraction for several hours. Because the phosphorylation level of PKCδ is thought to regulate its activity,enzyme stability, and/or substrate specificity(59,63,64,65,66,67),the functional significance of the unphosphorylated state of PKCδ in this study could be related to its effect on keratinocyte growth. In contrast to the effects of insulin, IGF-1 translocated PKCδ from the membrane to the cytosol but had no appreciable effect on PKC activity. The importance of this effect to the mitogenic action of IGF-1 is currently unclear. However,because mitogenic stimulation by EGF, KGF, PDGF, EcGF, or IGF-1 was not abrogated by the dominant negative mutant of PKCδ, insulin appears to be the primary activator of this PKC isoform in the regulation of keratinocyte proliferation.

The link between PKCδ and insulin signaling has also been established in several other systems. For example, we have recently shown that in muscle cultures, PKCδ mediates insulin-induced glucose transport(33,34). Similarly, in cells overexpressing the IR, insulin stimulation was shown to be associated with activation of PKCδ(68,69). Furthermore, the insulin stimulation was found to be specifically associated with activation of PKCδ(33,34,35,69). In addition, we have shown in this study that whereas insulin-induced proliferation of kertinocytes is mediated by PKCδ, this pathway was independent of PI3K, an important mediator of both insulin and IGF-1. Similar to the findings in this study, in a previous report we have found that in another model system of muscle myotubes, insulin-induced PKCδ activation was independent of PI3K activity(33,34). However, whereas in these studies insulin-mediated PKCδ activation has been linked to the metabolic effects of insulin, this is the first report linking PKCδ to insulin-mediated cell proliferation. In conclusion, this study shows for the first time that PKCδ, a multifunctional serine kinase, serves as a divergence point in transmitting insulin but not IGF-1 mitogenic signals. Future studies will be aimed at elucidating the role of insulin-induced PKCδ-mediated proliferation and its effects on the transmission of mitogenic signals by a variety of growth factors in skin keratinocytes.

This study was supported by a Focus Giving Grant from Johnson & Johnson and in part by the Sorrell Foundation and grants from the Israel Science Foundation founded by the Israel Academy of Sciences and Humanities, and by the Chief Scientist's Office of the Israel Ministry of Health. E.W. is a recipient a Career Development Award from the Juvenile Diabetes Foundation International. S.R.S. is the incumbent of the Louis Fisher Chair in Cellular Pathology, Bar Ilan University, Israel.