Regulation of Human Involucrin Promoter Activity by Novel Protein Kinase C Isoforms*

Human involucrin (hINV) mRNA level and promoter activity increase when keratinocytes are treated with the differentiating agent, 12-O-tetradecanoylphorbol-13-acetate (TPA). This response is mediated via a p38 mitogen-activated protein kinase-dependent pathway that targets activator protein 1 (Efimova, T., LaCelle, P. T., Welter, J. F., and Eckert, R. L. (1998) J. Biol. Chem. 273, 24387–24395). In the present study we examine the role of various PKC isoforms in this regulation. Transfection of expression plasmids encoding the novel PKC isoforms δ, ε, and η increase hINV promoter activity. In contrast, neither conventional PKC isoforms (α, β, and γ) nor the atypical isoform (ζ) regulate promoter activity. Consistent with these observations, promoter activity is inhibited by the PKCδ-selective inhibitor, rottlerin, but not by Go-6976, an inhibitor of conventional PKC isoforms, and novel PKC isoform-dependent promoter activation is inhibited by dominant-negative PKCδ. This regulation appears to be physiologically important, as transfection of keratinocytes with PKCδ, -ε, or -η increases expression of the endogenous hINV gene. Synergistic promoter activation (≥100-fold) is observed when PKCε- or -η-transfected cells are treated with TPA. In contrast, the PKCδ-dependent response is more complex as either activation or inhibition is observed, depending upon PKCδ concentration.

Human involucrin (hINV) 1 is a marker of keratinocyte differentiation that is exclusively expressed in differentiated, suprabasal keratinocytes, both in vivo and in vitro (1)(2)(3)(4)(5). 12-O-Tetradecanoylphorbol-13-acetate (TPA), a keratinocytedifferentiating agent, is extensively used to induce keratinocyte differentiation. We have previously shown that TPA treatment of human keratinocytes increases hINV mRNA level and promoter activity. This increase is mediated via a Ras 3 MEKK1 3 MEK3 3 p38 signaling cascade. One target of this cascade is activator protein 1 (AP1) that binds an AP1-binding site, AP1-1, in the hINV proximal regulatory region (6 -9). A key question to be resolved is the identity of the kinase(s) that initiate this cascade and mediate the effects of TPA in normal human keratinocytes. The various isoforms of PKC are key candidates for this regulatory role.
The protein kinase C (PKC) family consists of at least 11 distinct serine/threonine protein kinases that are classified into three groups. The conventional/classical PKCs (cPKCs) are calcium-, phospholipid-, and diacylglycerol-dependent (␣, ␤I, ␤II, and ␥); the novel PKCs (nPKCs) are calcium-independent PKCs (␦, ⑀, , , and ); and the atypical PKCs (aPKCs) are calcium-and diacylglycerol-independent (, and ) (10 -12). The differences in cofactor requirements, tissue distribution, subcellular localization, and substrate specificity suggest distinct biological functions for each PKC isozyme (10,12,13). Epidermal keratinocytes express ␣, ␦, ⑀, , and isoforms (14 -18). As involucrin is a model for the study of gene expression in stratifying epithelia, it is important to identify which of these PKC isoforms are involved in regulation of the hINV gene.
In the present study we demonstrate that novel PKC isoforms ␦, ⑀, and , but not the conventional and atypical PKC forms, are involved in regulation of hINV gene expression.
Plasmids-We have previously published the structure of the hINV promoter construct pINV-241, which include nucleotides Ϫ241/Ϫ7 of the hINV promoter, linked to the luciferase reporter gene in pGL2-Basic (7). All positions are defined relative to the hINV gene transcription start site (20).
Tissue Culture, Cell Transfection, and Luciferase Assay-Normal human foreskin keratinocytes were cultured as described previously. Third passage keratinocytes were transfected in 35-mm diameter dishes when approximately 60% confluent. For transfection experiments, 4 l of Fugene-6 reagent was added to 96 l of KSFM and incubated at 25°C for 5 min. The mixture was then mixed with 2 g of involucrin promoter reporter plasmid or, for co-transfection experiments, with 1 g of involucrin reporter plasmid and 1 g of kinase expression plasmid. The mixture was incubated at 25°C for 15 min and then added directly to the cells in 2 ml of KSFM. In general, the final DNA concentration in all groups was adjusted to 2 g of DNA per 4 l of Fugene-6 reagent per 35-mm dish by addition of empty expression vector. However, in the dose-response experiments, the final DNA concentration was 4 g per 8 l of Fugene-6 per 35-mm dish. After 24 h the cells were treated with KSFM in the presence or absence of TPA and/or the indicated inhibitor. After an additional 24 h, the cells were washed with phosphate-buffered saline, dissolved in 250 l of cell culture lysis reagent (Promega), and harvested by scraping. Luciferase activity was assayed immediately using Promega luciferase assay kit and a Berthold luminometer. All assays were performed in triplicate, and each experiment was repeated a minimum of three times. Luciferase activity was normalized per g of protein as described previously. A plasmid expressing green fluorescent protein (CLONTECH) was used to monitor transfection efficiency as described (28).
Immunoblot Analysis-Cultured keratinocytes, grown in KSFM, were treated with or without 50 ng/ml TPA and/or indicated pharmacological agent (concentrations indicated in each experiment) for 24 h prior to preparation of total cell extracts. Equal quantities of protein were electrophoresed on denaturing polyacrylamide gels and transferred to nitrocellulose. The membranes were blocked and then incubated with the appropriate primary antibody followed by a goat antirabbit IgG secondary antibody. Secondary antibody binding was visualized using a chemiluminescent detection system (Amersham Pharmacia Biotech). Fig.  1A shows that treatment of keratinocytes with phorbol ester increases human involucrin protein levels. This TPA-dependent increase in endogenous gene expression can be inhibited by BIS-IM, an agent that inhibits all PKC isoforms. To identify the specific PKC isoforms responsible for this regulation, we co-transfected keratinocytes with pINV-241 involucrin promoter-luciferase reporter construct (6,7,9) and expression plas-mids encoding specific wild type PKC isoforms. The results, see Fig. 1B, indicate that novel PKC (nPKC) isoforms ␦, ⑀, and increase hINV promoter activity as efficiently as TPA treatment (Ͼ10-fold). In contrast, the conventional PKC (cPKC) isoforms ␣, ␤I, and ␥, and the atypical isoform, PKC, produce minimal changes. As shown in Fig. 1C, the novel isoform-dependent increase is also observed for the endogenous gene, suggesting that the regulation is physiologically relevant.

PKC Isoforms That Regulate hINV Promoter Activity-
Failure of the cPKC isoforms to regulate hINV promoter activity in keratinocytes could result from a failure of the enzymes to be expressed or because they are expressed in an inactive form. To address these concerns, we transfected keratinocytes with PKC␣, -␦, -⑀, and -, and we monitored for presence of the corresponding protein by immunoblot. As shown in Fig. 2A, each PKC isoform is expressed at a comparable level. Although not evident from the exposure shown here, PKC␣, -␦, -⑀, and -are also expressed in non-transfected keratinocytes. To ensure that the transfected cPKC isoforms are active, we transfected the c-fos promoter (27), which responds to phorbol ester-sensitive PKC isoforms (29 -32), with each cPKC isoform. Fig. 2B shows that PKC␣, -␤1, and -␥ strongly increase c-fos promoter activity, confirming activity of these enzymes in keratinocytes. Thus, the lack of hINV promoter activation by the cPKC isoforms is not due to low expression or lack of activity of the expressed enzymes.
Concentration-dependent Regulation of hINV Promoter Ac-

FIG. 1. Novel PKC isoforms activate hINV gene expression.
A, keratinocytes were treated for 24 h with 50 ng/ml TPA and/or 3 M BIS-IM. Total cell extracts were then prepared and assayed for hINV protein level by immunoblot. B, cultured human epidermal keratinocytes were transfected with the 1 g of pINV-241 reporter plasmid and PKC-encoding plasmids (1 g) or empty expression vector (EV). After 24 h, the cultures were treated with (ϩ) or without (Ϫ) 50 ng/ml TPA. At 48 h after transfection, the cells were harvested, and lysates were assayed for luciferase activity. C, keratinocytes were transfected with 4 g of PKC␦, -⑀, or -. At 24 h post-transfection, the cells were harvested for preparation of total cell extracts. Equal quantities of extract was electrophoresed on an 8% denaturing/reducing polyacrylamide gel, transferred to nitrocellulose, and incubated with hINV-specific antibody (19).

FIG. 2. Transfected PKC isoforms are expressed in keratinocytes.
A, cultured keratinocytes were transfected with expression plasmids encoding indicated PKC isozymes (ϩ) or empty expression vector (Ϫ). After 48 h, total cell extracts were prepared. Equal quantities of protein were electrophoresed, transferred to nitrocellulose, and immunoblotted with PKC isozyme-specific antibodies. The arrows indicate migration of the each respective PKC isoform. The molecular masses are indicated in kilodaltons. B, keratinocytes were transfected with c-fos promoter reporter plasmid in the presence of expression plasmids encoding the conventional ␣, ␤1, and ␥ PKC isozymes or empty expression vector (EV). After 48 h, total cell extracts were prepared and assayed for luciferase activity.
tivity by Individual PKC Isoforms-PKC-dependent responses can be concentration-dependent, and so we studied the effects of various concentrations of PKC expression plasmid on promoter activity. PKC␣ did not regulate promoter activity at any concentration examined (Fig. 3). Although not evident in this figure, because the responses are minimal, cPKC␤I and cPKC␥ slightly stimulated promoter activity at high plasmid concentrations, and promoter activity was increased a modest 2-fold by 0.25 g of PKC plasmid with no further increase at higher plasmid concentrations. In contrast to these minimal responses, the nPKC isoforms, -␦, -⑀, and -, produced similar dose-response curves and a 12-35-fold increase in promoter activity.
Dominant-negative PKC␦ Suppresses PKC␦-, -⑀-, and --dependent Promoter Activation-The previous experiment suggests that PKC␦ may be the primary PKC controlling hINV gene expression. However, additional nPKC isoforms may also have a role. To confirm a role for PKC␦ and to study the role of the other nPKC isoforms, we used a dominant-negative form of PKC␦, dn␦(KR), in which the ATP-binding site is mutated (26). We show, in Fig. 5A, that dn␦(KR) completely inhibits TPA-dependent promoter activation. Thus, ligand-dependent activation of the promoter is inhibited by dn␦(KR). In Fig. 5B we show that dn␦(KR) completely inhibits PKC␦-dependent promoter activation. However, it is interesting that dn␦(KR) also inhibits wild type PKC-and PKC⑀-dependent promoter activation, although less efficiently compared with the dn␦(KR)-dependent inhibition of PKC␦-driven activity. This result suggest caution in assigning the sole regulatory role to PKC␦.
Activation by PKC Isoforms in the Presence of TPA-The results presented above indicate that nPKC isoforms activate basal hINV promoter activity. To examine the effects of TPA on PKC isoform-dependent activation, keratinocytes were cotransfected with pINV-241 reporter vector and PKC expression plasmid and treated with TPA. As shown in Fig. 6, none of the classical PKC isoforms (␣, ␤1, and ␥) or the atypical isoform () altered the TPA-dependent response. However, PKC⑀ and PKC produced a dramatic superinduction of hINV promoter activity in the presence of TPA (100-fold activation). Unexpectedly, and in contrast to the PKC␦-dependent activation observed in the absence of TPA (Fig. 3), PKC␦ suppressed the TPA-dependent activation to TPA-nonstimulated levels.
To investigate further the nPKC␦ effect, we transfected keratinocytes with a fixed amount of hINV reporter plasmid (2 g) and increasing concentrations of nPKC␦ expression plasmid (0.25-2 g), and we treated with TPA (Fig. 7). PKC␦, at 0.25 g expression plasmids per dish, produced a strong promoter activation. A comparable increase was observed at lower PKC␦ concentrations (0.06 g of PKC␦ per dish, not shown)  indicating that very small concentrations of this isoform can activate transcription. Interestingly, a smaller increase is observed at intermediate plasmid concentrations, and inhibition is observed at high (2 g) plasmid levels. In contrast, nPKC⑀ and nPKC (Fig. 7) markedly enhance the TPA-induced hINV promoter activity at all concentrations tested.
One mechanism whereby PKC is inactivated is via degradation (10,35). We therefore determined whether PKC␦, ⑀ and levels change in response to TPA treatment. The results, shown in Fig. 8, indicate that PKC⑀ is markedly decreased by TPA treatment; PKC␦ is decreased by 50%,; and PKC is slightly increased. These results suggest that PKC level is not correlated with ability to drive TPA/PKC-dependent hINV promoter activation.

Involucrin Promoter Activity and Endogenous hINV Gene
Expression Are Regulated by Novel PKC Isoforms-We have previously presented evidence indirectly implicating PKC in the signal transduction pathway leading to hINV promoter activation (7,9). Evidence includes the finding that TPA increases hINV mRNA levels and promoter activity (7), and BIS-IM, a general PKC inhibitor, blocks these responses (9). The present studies were designed, in part, to identify which PKC isoform(s) are involved in this regulation. Keratinocytes express classical, novel, and atypical PKC isoforms, including cPKC␣, nPKC␦, nPKC⑀, nPKC, and aPKC (14 -18, 36). All of these forms, with the exception of PKC, can be activated by TPA (25). Because of their distinct pattern of expression, primary sequence, differing response to stimuli, and differing cofactor dependence, each PKC is expected to have a different function (10,13). Therefore, it is important to determine which isoforms regulate keratinocyte target gene expression and which pathways convey the regulatory signal. To address these issues, we expressed wild type PKC isoforms in normal human keratinocytes and monitored the effects on basal hINV promoter activity. These experiments identify the novel PKC isoforms ␦, ⑀, and as potent inducers of promoter activity. In contrast, atypical PKC, and the conventional PKC isoforms ␣, ␤1, and ␥, failed to regulate activity. The lack of activation by PKC␣, -␤1, and -␥ was not due to inadequate expression of these kinases, as each was detected at high level by immunoblot. Moreover, activity of these kinases was confirmed by demonstrating PKC␣, -␤1, and -␥-dependent regulation of the c-fos promoter. c-fos is known to be regulated by TPA-dependent PKC isoforms (29 -32). A regulatory role for the novel PKC isoforms is further supported by the finding that Go-6976, an inhibitor of conventional but not novel PKC isoforms (34), does not inhibit promoter activity.
The Role of PKC␦-An important role for PKC␦ is suggested by the finding that promoter activity is inhibited by concentrations of rottlerin (33) that selectively inhibit PKC␦. However, our results also support a role for PKC⑀ and PKC in two ways. First, transfection of keratinocytes with PKC⑀ or -activates hINV promoter activity and expression of the endogenous hINV gene. Second, dominant-negative PKC␦ inhibits PKC⑀and PKC-dependent activity. Dominant-negative mutants have been extensively utilized to map signal transduction cascades. These proteins function to inhibit the activity of the endogenous wild type enzymes by a variety of mechanisms. For PKCs, dominant-negative mutants have been constructed by mutating threonine phosphorylation sites in the activation loop of the kinase domain (37) or by mutating the site that binds ATP, a necessary cofactor for enzyme activity (26, 29, 38 -41).

FIG. 5. Dominant-negative PKC␦ inhibits novel PKC-dependent promoter activation.
A, keratinocytes were transfected with 2 g of pINV-241 in the presence (ϩ) or absence of (Ϫ) dn␦(KR) and treated with 50 ng/ml TPA. After 24 h, extracts were assayed for promoter activity. B, keratinocytes were transfected with 2 g of pINV-241 and 1 g of PKC␦, -, or -⑀ in the presence (ϩ) or absence (Ϫ) of 1 g of dominant-negative PKC␦ (dn␦(KR)). At 24 h after transfection extracts were prepared and assayed for luciferase activity. We have used a form of PKC␦ in which a conserved lysine at the ATP-binding site is converted to arginine to inactivate the enzyme (26). Our experiments show that dominant-negative PKC␦ inhibits PKC␦-dependent promoter activation, a result that is consistent with a role for PKC␦ in regulating hINV gene expression. However, albeit to a lesser extent, dnPKC␦ also inhibits PKC-and PKC⑀-dependent promoter activation. There are several possible mechanisms whereby dnPKC␦ could inhibit PKC-and PKC⑀-dependent responses. First, dnPKC␦ may titrate a kinase that is required for activation of all novel PKCs and, thereby, inhibit activity of all nPKC isoforms (i.e. dominant-negative PKC␦ may not specifically inhibit of PKC␦ in our system). Second, dnPKC␦ may interfere with chaperone "docking" proteins that may regulate the function of multiple PKC isoforms. Experiments that suggest these possibilities have been noted using activation-loop mutants of PKC (10, 37). The dnPKC␦ used in the present studies is an ATP-binding site mutant (26). ATP-binding site mutants may be more specific inhibitors of the corresponding wild type PKC isoform than activation-loop mutants; however, this has not been rigorously tested. Third, PKC␦, -⑀, and -may indirectly regulate the level/activity of each other by regulating expression of the corresponding genes. This interesting possibility is not unprecedented, as a recent study in mouse lymphoma cells shows that PKC␣ increases PKC␦ protein level by regulating PKC␦ mRNA level (42). Fourth, PKC␦, -⑀, and -may share a common substrate(s). Fifth, overexpression of individual PKC isoforms could lead to non-selective activation of downstream targets of other PKC isoforms. Thus, although our present studies strongly point to a role for PKC␦, it is likely that PKC⑀ andalso play an important role.
Function of PKCs in the Presence of the PKC Activator, Phorbol Ester-Diacylglycerol is a ligand that directly activates PKC isoforms (11). TPA is a stable diacylglycerol analog that mimics the effects of diacylglycerol and strongly activates PKC (43)(44)(45) and is a potent inducer of keratinocyte differentiation (46). Treating cultured keratinocytes with TPA increases cell differentiation, and this change is correlated with an increase in hINV mRNA and protein (7,9,47,48). To study the effects of TPA-dependent activation of individual PKC isoforms on hINV promoter activity, we transfected cells with PKC expression constructs and then treated with TPA. The results indicate a synergistic activation (Ն100-fold, Fig. 6) of promoter activity when PKC⑀-or --treated cells are incubated with TPA. This increase depends directly on the concentration of PKC⑀ orexpression plasmid transfected. In contrast, no potentiation was observed for PKC␣, -␤1, -␥, or -. PKC␦, however, caused synergistic promoter activation at moderate plasmid concentrations and inhibition at higher plasmid concentrations. It is not clear why the response to ␦ is biphasic; however, this result suggests that the PKC-dependent regulation is complex. One possible explanation is that PKC␦ levels are reduced by TPA treatment. However, immunoblot results suggest that PKC␦ levels are decreased by only 30 -50% in response to TPA treatment. In contrast, PKC⑀ levels are reduced substantially, and PKC levels are relatively unchanged. These results suggest that TPA-dependent regulation of PKC level does not explain the difference in activity. There are other possible explanations. For example, PKC␦ undergoes tyrosine phosphorylation in response to various stimuli, including epidermal growth factor, platelet-derived growth factor, transforming growth fac- Keratinocytes were treated with 50 ng/ml TPA for 24 h followed by preparation of nuclear extracts. Equivalent quantities of extract, layered based on protein concentration, were electrophoresed on an 8% acrylamide gel and transferred to nitrocellulose for detection using PKC⑀-, -␦-, and --specific antibodies. Binding of the primary antibody was detected by incubation with an appropriate secondary antibody, and binding was visualized using ECL technology.
tor-␣, carbachol, extracellular ATP or UTP, and hydrogen peroxide (35). Moreover, tyrosine kinases of the Src family phosphorylate PKC␦ in vitro, although the functional significance of this phosphorylation has not been clearly established (45). Several reports in keratinocytes suggest that phosphorylation of PKC␦ on tyrosine residues in the regulatory domain diminishes activity (49 -51), although studies in other systems report increased PKC activity following tyrosine phosphorylation (52-54)). Thus, high level overexpression of PKC␦ may result in tyrosine phosphorylation-dependent inactivation of PKC␦. However, although PKC␦ could be inactivated by phosphorylation, it is difficult to understand the unique plasmid concentration dependence of the inhibition. As noted above, PKC␦ is expressed at high levels in keratinocytes. It is possible that very high plasmid concentrations inhibit promoter activity by saturating the system with PKC␦ which "squelches" the response.
PKC Isoforms and Keratinocyte Function-Our results suggest that PKC␦, -⑀, and -regulate hINV gene expression. Takahashi et al. (55) showed that PKC␥ increases basal hINV promoter activity by 2-3-fold in the absence of TPA, whereas PKC␣ and -increase activity by 2-fold in TPA-treated cells. These studies differ from ours in that we do not observe PKC␣dependent regulation. Moreover, the magnitude of our responses are much larger. We attribute the different findings to the fact that we use normal human keratinocytes, whereas Takahashi et al. (55) used SV40-immortalized keratinocytes. Signal transduction is known to be altered in immortalized cell lines, and a significant amount of circumstantial evidence indicates that the involucrin gene is not always appropriately regulated in immortalized keratinocyte cell lines. For example, in cell lines, the level of hINV gene expression and the response to stimuli is significantly reduced compared with normal cells. 2 However, the studies in both cell types support a role for PKC as a regulator of hINV gene expression.
In epidermis, involucrin is expressed in the late spinous and granular layers but not in the basal layer (1,3,5). Type I transglutaminase is another marker of keratinocyte differentiation that displays a similar spatial and temporal pattern of expression (56,57). Thus, it is useful to compare mechanisms that regulate expression of the involucrin and transglutaminase type 1 (TG1) genes. Recent studies indicate that overexpression of the and ␦ PKC isoforms in human keratinocytes causes an increase in TG1-encoding mRNA. This is correlated with growth inhibition and morphological changes (58). In contrast, the ␣ and PKC isoforms do not regulate TG1 expression. Regulation in response to PKC⑀ was not studied. In addition, it has been reported that expression of exogenous PKC in a rat keratinocyte cell line efficiently induces TG1 transcription, but PKC␣, PKC␤II, PKC␥, and PKC did not regulate activity (59). Yuspa and co-workers (60) have shown that TPA blocks the calcium-dependent increase in K1 and K10 (spinous layer markers) and simultaneously increases filaggrin and loricrin expression (granular layer markers). This TPA-dependent response is blocked by bryostatin, a PKC inactivating agent or cycloheximide, a protein synthesis inhibitory agent. This suggests that PKC regulates genes in the transition from spinous to granular layers (60). These results are consistent with ours, as involucrin is predominantly a granular cell marker (2). Moreover, PKC is known to be localized in the epidermal granular layer (61). Thus, our studies suggest a role for novel PKCs as regulators of hINV gene expression.