Protein Kinase C Regulates Keratinocyte Transglutaminase (TGK ) Gene Expression in Cultured Primary Mouse Epidermal Keratinocytes Induced to Terminally Differentiate by Calcium

Andrzej A. Dlugosz and Stuart H . Yuspa Laboracory of Cellular Carcinogenesis and Tumor Promotion, Division of Cancer Etiology, National Cancer Institute, Bethesda, Maryland, U.S.A.

During the final stage of epidermal differentiation, activation of keratinocyte transglutaminase results in covalent cross­linking of a variety of proteins to form highly protective cornified cell envelopes. We have studied the regulation of keratinocyte trans glutaminase (TGK ) gene expression in murine epidermal keratinocytes induced to terminally dif­ferentiate in vitro by increasing the level of extracellular Ca++ or treatment with the protein kinase C (PKC) activator 12-O-tetradecanoylphorbol-13-acetate (TPA). Raising extra­cell ular Ca ++ induces squamous differentiation of cultured ke­ratinocytes and elicits a concentration-dependent increase in expression of TGK mRN A; keratinocytes grown for 24 h in 0.12 mM Ca++ medium express -12 times as much TGK

mRNA as basal cells (grown in 0.05 mM Ca++ medium), whereas cultures exposed to 1.4 mM Ca++ express -17 times as much. TPA induces squamous differentiation and TGK

mRNA even in basal keratinocyte cultures grown in 0 .05 mM Ca++ medium, suggesting that expression of this differ­entiation marker is regulated by the PKC signaling pathway.

The barrier function of mammalian skin can be attrib­

. uted largely to the stratum corneum, a protective layer of terminally differentiated keratinocytes that pro­vides the interface between organism and external en­vironment. The stratum corneum is the most superfi­

cial of four cellular compartments in the epidermis, each expressing a unique pattern of keratinocyte differentiation markers. The kera­tin pairs K5 /K 14 and K 1 /K 1 0 are expressed in basal and spinous cells, respectively; loricrin, filaggrin, and involucrin are expressed in granular cells [1,2] . During the final stages of epidermal differen­tiation, epidermal transglutaminase generates €-(y-glutamyl) lysine bonds that cross-link substrates such as loricrin and involucrin to form a highly protective cornified cell envelope (reviewed in [3,4]) .

The analysis of keratinocytes isolated from mouse and human epidermis has identified Ca++ as a key regulator of terminal differ­entiation in vitro [5]. Keratinocytes require medium with a reduced extracellular Ca++ concentration (0.05 mM) to maintain a prolifer­ative, basal cell-like phenotype [6]. Raising Ca++ in the medium to

Manuscript received September 1, 1993; accepted for publication No­vember 12,1993.

Reprint requests co: Dr. Andrzej A. Dlugosz, LCCTP /NCI/NIH, Build­ing 37 /Room 3B25, 9000 Rockville Pike, Bethesda, MD 20892.

Abbreviation: TGK , keratinocyte trans glutaminase.

Induction of TGK mRNA in response to TPA treatment is transient, reaching a peak at 6 - 8 h and returning to baseline by 24 h. In contrast, elevation of TGK mRNA levels in re­sponse to Ca++ persists for at least 24 h . The increased abun­dance of TGK mRNA reflects increased transcription of the TGK gene, based on nuclear run-on analysis of Ca++ - and TPA-treated keratinocytes. Induction ofTGK mRNA by ei­ther TPA or Ca++ is blocked in the presence of cyclohexi­mide, suggesting that a PKC-dependent protein factor is re­quired for TGK gene expression in response to both stimuli. Furthermore, the accumulation ofTGK mRNA in keratino­cytes treated with TPA or Ca++ is blocked in cells treated with the PKC inhibitor GF 109203X or bryostatin. These results suggest that the induction ofTGK gene expression by Ca++ is dependent on PKC, providing further support for the hypothesis that PKC plays a central role in regulating the late stages of epidermal differentiation. Key words: transcription / phorbol ester / GF 109203X / bryostatin. ] Invest Dermatol 102:409-414, 1994

0.12 mM triggers stepwise changes in gene expression similar to those observed in epidermis; induction of the structural markers K1 and K10 is followed by the appearance ofloricrin and filaggrin [7J. Epidermal transglutaminase is also activated in 0.12 mM Ca++ me­dium; however, higher transglutaminase activity is detected when cells are exposed to medium containing 1.4 mM Ca++ [8]. Com­bined with these in vitro findings, the demonstration of a Ca++ gradient in mouse and human epidermis, with elevated Ca++ levels detected in differentiating cell layers [9 -11 J, suggests that Ca++ provides a physiologic signal for keratinocyte differentiation.

12-0-tetradecanoylphorbol-13-acetate (TPA) is a potent inducer of epidermal trans glutaminase activity and cornified envelopes both . i~1 vitro and ill vivo J8,12,13], indi.cating that pharmacologic activa­tIon of PKC can tngger the terml11al stage of keratinocyte differen­tiation. Furthermore, cultured keratinocytes grown in the presence of elevated extracellular Ca++ exhibit increased levels of inositol phosphates [14-16J, intracellular Ca++ [17], and diacylglycerol 116,18], suggesting that PKC is also involved in Ca++-mediated keratinocyte differentiation. Consistent with this hypothesis, Ca++-dependent accumulation ofloricrin and filaggrin is blocked in cells where PKC has been inactivated, whereas expression of mRNA encoding these markers is enhanced by PKC activators [19).

In this report, we have used pharmacologic agents combined with changes in extracellular Ca++ to determine the involvement ofPKC

0022-202X/94/S07.00 Copyright © 1994 by The Society for Investigative Dermatology, Inc.

409

410 DLUGOSZ AND YUSPA

A. Ca2+ (mM)

& ~ ~ NB Cy c). .......

TGK

GAPDH

B. 25

c 20 <d:: .Q z .... a:: g 15 E"'O c

10 ::.::-<.!)"'O ~o 5 u..

0 0 .05 0 .12 1.4

Ca2+ (mM)

Figure 1. The steady-state level ofTGK mRNA is increased during Ca++­mediated keratinocyte differentiation . A) Primary keratinocytes grown as basal cells (in 0.05 mM Ca++ medium) were exposed to media with the indicated Ca++ concentration for 24 h to induce terminal differentiation. Poly(A:)+ RNA was analyzed by Northern blotting using TGK and GAPDH cDNA probes, as described in Materia ls and Methods. The TGK transcript is - 3.3 kb in length based on residual ribosomal RNA as ma,:kers for 5 kb and 2 kb. NB indicates RNA isolated from newborn mouse epidermis used as a positive control. Similar results were obtained in two additional experi­ments. B) Quantitation of Ca++-induced TGK mRNA expression by scan­ning densitometry . TG K mRNA at each Ca++ concentration was normalized to GAPDH and the fold-induction expressed relative to the level in basal cell cultures. Values represent data from three experiments ± SEM.

in regulating keratinocyte transglutaminase (TGK ) gene expression in llitro. Our findings indicate that activation of PKC is both neces­sary and sufficient to induce TGK gene expression in cultured ke­ratinocytes, strongly supporting the concept that late stages of epi­dermal differentiation are regulated through Ca++ -dependent activation of the PKC pathway.

MATERIALS AND METHODS

Cell Culture Primary keratinocytes were isolated from skin of newborn BALB/c mice as previously described [6] and cultured in 60-mm tissue-cul­ture dishes (Costar, Cambridge, MA). Cells were grown in Eagle's mini­mum essential medium (without Ca++ or Mg++) supplemented with 8% Ca++-depleted fetal calf serum [6] and 0.25% penicillin-streptomycin solu­tion (GIBCO, Grand Island, NY). The Ca++ concentra.tion in the medium was adjusted to specific levels by adding an appropriate vol ume of 280 mM CaCI2. Cells grown in medium with 0.05 mM Ca++ exhibited a basal cell-like phenotype; terminal differentiation was induced by exposure to medium with higher extracellular Ca++ concentrations , as described in the figure legends.

Northern Blot and Nuclear Run-On Analysis Total RNA was iso­lated from primary keratinocytes by lysis in 4 M guallidine isothiocyanate followed by ultracentrifugation through a cesium chloride gradient [20] . In some experiments, poly(A)+ RNA was isolated directly from cell Iysates as previously described [21] . Transcripts were separated by electrophoresis in a 1 % agarose/0.66 M formaldehyde gel, transferred to reinforced nitrocellu­lose (BA-S NC; Schleicher & Schuell, Keene, NH), and baked at 80°C in a

THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

A. o 0 .1 10 100 103 1()4 TPA (nM)

GAPDH

B. 10

8

6

4

2

O~-{fl--~I ~.~.~".~~~o~,,~o~.L~'~'~,,·~.~,.u·~,,~J~o ~o~J

o 0.1 10 100 103 104

TPA (nM )

Figure 2. TPA induces TGK mRNA expression. A) Keratinocyte cultures grown in 0.05 mM Ca++ medium were exposed to TPA at the indicated concentrations for 4 h . Total RNA was isolated and Northern blot analysis performed as described in Materia ls arId Methods. B) Scanning densitometry of Northern bIotin (A). The level ofTGK mRNA was normalized to GAPDH and changes in abundance expressed relative to the DMSO-treated control.

vacuum oven [7]. Conditions for Northern blotting were previously de­scribed [22], except that pre-hybridization and hybridization solutions con­tained 6 X sodium citrate/sodium chloride buffer (SSe), 5 X Denhardt's solution, 0.5% sodium dodecyl sulfate (SDS), 100 tlg/ml sheared salmon­sperm DNA, and 50% formam ide [23]. TG K mRNA was identified by hybridization to rat or human TGK cDNA fragments (- 2 kb) [24], kindly provided by Dr. Robert Rice (University of California, Davis, CAl. The glyceraldehyde-3-phosphate dehydrogenase (GAPDH) probe was a full­length rat eDNA in pUC18 [25] . Probes were labeled with 32P_CTP by nick-translation. Filters were routinely washed at a maximum stringency of 0.5 X SSC with 0.2% SDS, at 65°C. Transcript levels were quantified using a scanning laser densitometer and ImageQuant software (Molecular Dy­namics, Sunnyvale, CAl . The abundance ofTGK mRN A was normalized to GAPDH transcripts.

I" vitro transcription reactions were performed as previously described [261 with modifications [19]. Briefly, nuclei were isolated from -1 X 108 cells and stored at - 70° C in 200 til of buffer containing 50 mM tris, pH 8.3, 5 mM MgCI2, 0.1 mM ethylenediaminetetraacetic acid (EDTA), and 40% glycerol. Nuclei were added to 200 III of 2 X reaction buffer that contained 10 mM tris, pH 8.0, 5 mM MgCI2' 0.3 M KCI, 200 U/ml RNAsin (Pro­mega, Madison, WI), 500 tiM ATP, CTP, and GTP, 10 tiM UTP, and 200 or 250 tlCi 32P_UTP (800 Ci/mmol, New England Nuclear [NEN], Boston, MA) . Nascent transcripts were elongated for 30 min at 30°C. Nuclei were lysed in 4 M guanidine isothiocyanate and total RNA isolated as described above. Ten micrograms of plasmid DNA (rat TGK and GAPDH, described above) was immobilized on nitrocellulose filters and hybridization to radio­labeled transcripts performed as previously described [19].

I25I-Epidermal Growth Factor (EGF) Binding Assay Binding ofl25I_ labeled EGF to primary ke ratinocytes was determined as previously de­scribed [27], with minor modifications. Following treatment, cells were washed twice with ice-cold binding buffer (Dulbecco's minimum essential medium with 50 mM N ,N-bis-[2-hydroxyethyl]-2-aminosulfonic acid [pH 7.4] and 1 mg/ml bovine serum albumin), then incubated with 1251_EGF (0 .1 ItCi/ml/well ; NEN) ± excess unlabeled EGF (1 Itg/ml, receptor grade; Collaborative Research, Bedford, MA) in 1 ml of binding buffer for 5 h on a bed of ice. Cultures were washed four times with ice-cold binding buffer and cells harvested in two 500 til volumes oflysis buffer (0.1 M tris, pH 7.4, 0.5% SDS,l mM EDTA). Radioactivity was determined by scintillation counting; non-specific binding was < 200 cpm in all groups.

VOL. ' 102. NO.4 APRIL 1994 PKC REGULATES TGK GENE EXPRESSION 411

TPA Ca2+ + TPA Treatment (hrs.) o 2 4 6 10 24 2 4 6 10 24 2 4 6 10 24

•

GAPDH

Ethidium Bromide

Figure 3. Time-course analysis ofTGK mRNA induction in primary keratinocytes exposed to Ca++, TPA. or Ca++ + TPA combined. TPA concentration was 100 nM. Total RNA was isolated at the start of the experiment and after the indicated treatment intervals. Based on densitometric analysis. maximum induction ofTGK mRNA in response to Ca++. TPA. and Ca++ + TPA was 8.8-fold (at 10 h). 179-fold (at 6 h) and 201-fold (at 6 h) . respectively. Similar results were obtained in an additional experiment with fewer timepoints.

Reagents TPA was obtained from LC Services (Woburn. MA) . bryostatin (bryostatin 1) was a gift from Dr. George Pettit (Arizona State Universi ry. Tempe. AZ). and the PKC inhibitor GF 109203X was kindly provided by Dr. Jorge Kirilovsky (Glaxo Pharmaceuticals, Les Ulis, France). TPA and GF 109203X stocks were in dimethylsulfoxide; bryostatin was in ethanol.

RESULTS

Ca++ and TPA Induce TGK mRNA in Cultured Keratino­cytes TGK gene expression was examined in primary mouse epi­dermal keratinocytes growing as basal cells (in 0.05 mM Ca++ me­dium) or induced to terminally differentiate in medium containing 0 .12 or 1.4 mM Ca++. Although expression of structural differen­tiation markers is restricted to extracellular Ca++ ranging from 0.1 to 0.3 mM [7], a squamous phenotype is triggered at all Ca++ concentrations 2: 0.1 mM Ca++ [B]. In primary epidermal keratino­cyte cultures, extracel lular Ca++ induces TGK mRNA in a concen­tration-dependent manner. Based on densitometric scanning of Northern blots, keratinocytes grown for 24 h in 0.12 mM Ca++ medium express -12 times as much TGK mRNA as basal cells (grown in 0.05 mM Ca++ medium), whereas cultures grown in 1.4 mM Ca++ express -17 times as much (Fig 1) . In individual experi­ments, TGK mRNA expression in 1.4 mM Ca++ medium was always higher (by 37% ± 13% SEM, n = 3) than in 0.12 mM Ca++ medium. 1.4 mM Ca++ medium was used in subsequent experi­ments to induce terminal differentiation, unless otherwise indi­cated.

Because PKC activators can substitute for Ca++ to induce erider­mal transglutaminase activity [2B], we examined the effects 0 TPA on TGK gene expression ill vitro. In this experiment, treatment of basal cell cultures with TPA for 4 h caused a maximal ninefold increase in the level of TGK mRNA with a 50% effective dose (EDso) of -10 nM (Fig 2). Similar to Ca++ (Fig IB), the level of TGK mRNA induction by TPA was variable between experiments, probably due to different baseline levels ofTGK mRNA in basal cell cultures. In four experiments, induction ofTGK mRNA following exposure to 100 nM TPA for 4 h varied from ninefold to 7B-fold (average = 46-fold ± 14, SEM). Combined, these results indicate that keratinocyte differentiation induced by either Ca++ or TPA is associated with increased expression ofTGK mRNA.

To further explore the regulation ofTGK gene expression, kerati­nocytes were treated with Ca++ or TPA alone or with both agents combined. Total RNA was isolated at the start of the experiment and after 1, 2, 4, 6, 10. and 24 h of treatment. Based on densitome-

tric scanning of Northern blots in this experiment, TGK mRNA was increased a maximum 179-fold in TPA-treated cultures (at 6 h) and B.B-fold in Ca++-treated cultures (at 10 h). In cultures treated with Ca++ + TPA for 6 h, TGK transcripts were elevated to a maximum level 201 times greater than in basal cell cultures (Fig 3). Although the initial induction ofTGK mRNA was detected earlier in response to TPA than Ca++ (2 vs 4 h), TGK mRNA expression was reduced to baseline after 24 h exposure to TP A both in the presence or absence of Ca++ (Fig 3). In contrast, induction of TGK mRNA in response to Ca++ alone remained maximally elevated through 24 h (Fig 3).

TGK mRNA Expression is Regulated at the Transcriptional Level Nuclear run-on analysis was performed to determine whether Ca++ and TPA affect TGK gene expression at the transcrip­tionallevel. For these experiments, cultures were induced to differ­entiate in medium with 0.12 mM Ca++ rather than 1.4 mM Ca++ because of difficulties in the isolation of nuclei from keratinocytes grown at high extracellular Ca++ concentrations. The relative tran­scription rate of the TGK gene is increased during 0.12 mM Ca++­induced keratinocyte differentiation (Fig 4A). In differentiating keratinocytes exposed to TPA for 3 h, TGK transcription is induced to a greater extent (Fig 4B). These findings suggest that accumula­tion ofTGKmRNA in response to Ca++ orTPA is due at least in part to increased transcription of the TGK gene.

TGK mRNA Induction is Dependent on Protein Synthe­sis To further characterize the molecular regulation ofTGK gene expression, primary keratinocytes were exposed for 6 h to TPA or solvent, ± Ca++, in the presence or absence of20 ,ugj ml cyclohexi­mide. This concentration of cycloheximide has previously been shown to block protein synthesis by > 95% in cultured mouse epi­dermal keratinocytes [2B] . Accumulation ofTGK mRNA is blocked by cycloheximide in cu ltures treated with TPA, Ca++, or Ca++ + TPA (Fig 5) , suggesting that a protein factor is required for the induction ofTGK gene expression by both stimuli. GAPDH mRNA levels do not appear to be influenced by cycloheximide (Fig 5).

Induction of TGK mRNA by Ca++ or TPA Requires PKC The induction of TGK mRNA by TPA even in basal cell keratinocyte cultures suggested that expression of this marker in response to Ca++ occurs through activation of the PKC signaling pathway. To test this hypoth esis, TGK mRNA expression was ex­amined in cultures treated with the selective PKC inhibitor GF

412 DLUGOSZ AND YUSPA

A. pGEM

TGK

GAPDH

Control Ca2+

B. pGEM

TGK

GAPDH

Ca2+ TPA + Ca2+

Figure 4. Ca++ and TPA increase TGK gene transc.!j>tion: A) Primary keratinocytes were grown as basal cells 111 0.05 mM Ca medIUm (Colltrol) or induced to terminally differentiate for 24 h in 0.12 mM Ca++ medIUm (Ca2+). Simila.r results were obtained in an additional experiment. B) Keratl­nocytes grown in 0.12 mM Ca++ medium for 20 h were exposed to 100 nM TPA (TPA + Ca2+) or DMSO (Ca2+) for an additional 3 h. Nuclei were isolated. and run-on analysis performed as described in Materials alld Methods. pGEM rep~esents pGEM 7Zf(+) plasmid DNA used as a negative control. Two additional expenments produced sllrular results.

109203X [29]. The effect of this compound on inhibition of 1251_ EGF binding byTPA [30] was examined as a measure of its ability to block PKC-mediated events in cultured keratinocytes. GF 109203X blocked this response to TP A in a dose-dependent man­ner, with control levels of '25I-EGF binding achieved at ~ 1 .uM GF 109203X (Fig 6A), suggesting that this agent is an effective PKC inhibitor in cultured keratinocytes. GF 109203X also inhibited TPA- and Ca++ -mediated accumulation ofTGK mRNA (Fig 6B,C). Based on results of scanning densitometry, the approximate IC50s

CHX Ca2+

TPA + +

+ +

+ + +

+ + +

+ +

GAPDH

Figure 5. TGK mRNA induction by Ca++ or TPA is blocked i'.' t~e pres­ence of cycloheximide. Fresh medium ± 20 Jlg/ml cycloheXimide was added to basal cell cultures. After 30 min, Ca++, TPA (100 11M), or both agents were added to control and cycloheximide-t~eated .cultures. Total RNA was isolated after 6 h for Northern blot analYSIS. Simtiar results were obtained in an additional experiment.

THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

A. =8000

~ 16000 2 0 z 4000 ::> 0 ID u.

2000 (!)

";' • ~ 0

DMSO TPA

GF 109203X (I'M) - - 0.1 0.3 1 3 10

B. TPA 1

GF 109203X(I'M) 0 0 0.1 0.3 1 3 10

TGK

GAPDH

C. Ca" (

GF 109203X(I'M) 0 0 0.1 0.3 1 3 10

TGK

GAPDH

Figure 6. Induction ofTGK mRNA is blocked by the PKC inhibitor GF 109203X. A) GF 109203X blocks TPA-mediated inhibition of I25I-EGF binding. Primary keratinocytes grown in 12-well dishes were treated with the indicated concentration of GF 109203X for 45 min. 100 nM TPA or 0.1 % DMSO was added and binding of 12SI_EGF determined after a 2-h incubation, as described in Materials MId Methods. Triplicate cultures were assayed at each dose, with the SEM indicated by error bars. Similar results were obtained in an additional experiment. B) Basal cell cultures were treated with the indicated concentration of GF 109203X for 1 h, followed by exposure to 100 nM TPA for 6 h. Northern blot analysis was performed using total RN A as described in Materials alld Methods. C) Basal keratinocyte cultures were exposed to 1.4 mM Ca++ medium ± GF 109203X for 8 h. In two other experiments, GF 109203X blocked accumulation ofTGK mRNA in 0.12 mM Ca++ medium at a dose of 5 or 10 jiM.

for blocking TGK mRN A induced by either Ca++ or TP A are within the same order of magnitude.

As an additional approach to assess PKC's involvement in TGK gene expression, keratinocytes were pre-treated with 60 nM bryo­statin to functionally inactivate PKC. This agent selectively blocks a variery of PKC-mediated responses in cultured keratinocytcs [31-33]' Both TPA- and Ca++-mediated induction of TGK mRNA is blocked in keratinocytes pre-treated with bryostatin (Fig 7) . Similar to the results with GF 109203X (Fig 6), the approximate IC50s for blocking TGK gene expression in TPA- or Ca++ -treated cultures are within the same order of magnitude, based on densitometric analy­sis. Combined, the results of these experiments suggest that activa­tion of the PKC pathway is required for the induction ofTGK gene expression in response to either TP A or Ca++.

DISCUSSION

Keratinocyte differentiation can be induced in vitro by raising extra­cellular Ca++ from 0.05 to ~ 0.1 mM, and elevated Ca++ levels have been demonstrated in vivo in differentiating layers of epidermis, suggesting that this ion provides a physiologic signal regulating epidermal differentiation. Based 011 analysis ofTGK gene expression as a marker for keratinocyte differentiation, the results of this study suggest that PKC mediates the Ca++ signal for this aspect of kerati-

VOL: 102, NO.4 APRIL 1994

A. TPA

o 0.01 0.1 10 100 Bryostatin (nM)

GAPDH

B.

o 0.01 0.1 10 100 Bryostatin (nM)

GAPDH

Figure 7. Ind~ction ofTGK mRN ~ is blocked in keratinocytes where PKC has been inactivated usmg bryostatm. Basal cell cultures were pre-treated with the indicated concentrations ofbryostatin for 22 h, followed by expo­sure to 100 aM TPA for 6 h (A) or 1.4 mM Ca* medium for 23 h (B), ± bryostatin. 60 aM bryostatin blocked induction ofTGK mRNA by TP A or 0.12 mM Ca* medium in an additional experiment.

nocyte differentiation: 1) both Ca++ and TPA induce TGK

gene expression in cultured keratinocytes; 2) induction of TGK

mRNA by either Ca++ or TPA is dependent on a cycloheximide­sensitive factor; 3) pharmacologic inhibition of PKC blocks both Ca++ -and TPA-mediated induction of TGK gene expression. To­gether with recent findings implicating PKC in Ca++-mediated expression of the granular cell markers loricrin and filaggrin [19), these results strongly support the concept that PKC plays a funda­mental role in regulating the late stages of epidermal differentia­tion.

The involvement of Ca++ in regulating TGK gene expression differs from that for other keratinocyte-specific markers. In cul­tured mouse epidermal keratinocytes, spinous (Kl, KI0) and granu­lar (loricrin, filaggrin) markers are induced by extracellular Ca++ in a limited concentration range between 0.10 and 0.16 mM [7]. In medium with 1.4 mM Ca++, these structural markers are not in­duced [7], whereas expression ofTGK mRNA is maximally elevated (Fig 1), consistent with the predominantly squamous phenotype observed under these culture conditions. In light of these in vitro observations it is noteworthy that the highest levels of intracellular Ca++ have been localized to cells in the upper granular layer of both mouse and human epidermis [10,11]. Preferential expression of TGK at this stage of epidermal differentiation would favor cross­linking of pre-existing structural proteins into cornified cell enve­lopes, which are largely responsible for the stratum corneum's pro­tective function. A second distinction between the regulation of TGK and other keratinocyte markers is the ability to induce expres­sion of the former in 0.05 mM Ca++ medium by treating with TPA (Fig 2), suggesting that PKC activation is sufficient to induce ex­pression ofTGK . In contrast, expression of the granular cell markers loricrin and filaggrin appears to be dependent 011 a specific intracel­lular Ca++ level in addition to activation of the PKC pathway [19]. A direct requirement for Ca++ in regulating expression of keratino­eyte structural markers is supported by the recent demonstration of a Ca++ -responsive element in the 3' non-coding region of the human keratin 1 gene [34].

PKC REGULATES TGK GENE EXPRESSION 413

Although the role of Ca++ in triggering keratinocyte-specific gene expression is well established, the biochemical pathways regu­lating this complex process have not been fully elucidated. Induc­tion of epidermal trans glutaminase activity and cornified envelope formation by TPA, both irl vitro and in vivo [8], suggested that the terminal stage of keratinocyte differentiation is regulated by PKC. The observation that increased extracellular Ca++ is associated with elevated cellular diacylglycerol levels [16,18] provided a biochemi­cal link coupling the Ca++ signal for keratinocyte differentiation to the PKC signaling pathway. The ability to block Ca++-mediated TGK gene expression by inactivating PKC in mouse (Figs 6,7) as well as human [35] keratinocytes strongly suggests that Ca++ in­duces this differentiation marker through the PKC signaling path­way. Along with the upregulation ofTGK gene expression by PKC, the additional influence of other factors is likely to be important in determining the final level of expression. For example, whereas dexamethasone also induces TGK mRNA in cultured keratinocytes, retinoic acid blocks accumulation of TGK mRNA in response to TPA, Ca++, or dexamethasone [36].

There are several distinctions between TPA- and Ca++-induced TGK gene expression (Fig 3) that may be related to the way these two stimuli influence PKC: 1) TPA induces TGK mRNA more rapidly than Ca++. The slower induction of TGK mRNA in Ca++­treated cells may reflect slow accumulation of endogenous diacyl­glycerol with a corresponding delay in PKC activation [16]. In contrast, TPA directly activates PKC resulting in expression of cer­tain genes (e.g.,fos and jlln) within minutes, others (e.g., transin and collagenase) within hours [37]. 2) TGK mRNA is induced to a higher level by TPA than Ca++. Consistent with this result, phorbol esters are far more effective activators of PKC than physiologic agents (hormones, growth factors, etc.) that generate endogenous diacylglycerols via hydrolysis of membrane phospholipids. 3) De­spite the more rapid appearance and greater induction of TGK

mRNA in response to TPA, by 24 h transcript levels return to baseline, whereas in Ca++-treated keratinocytes TGK mRNA re­mains maximally elevated. Transient induction ofTGK mRNA by TPA may be due to PKC downregulation [38] resulting in a limited activation phase in TPA-treated cultures and brief induction ofTGK mRN A. Unlike TP A, Ca++ does not downregulate PKC in cultured keratinocytes* [39] , permitting prolonged activation of the PKC pathway in response to this signal. Relative to Ca++-treated cul­tures, the low level ofTGK mRNA in cultures exposed for 24 h to Ca++ + TPA s.upports the notion that PKC is required for pro­longed expression ofTGK mRNA in response to Ca++.

Inhibition of Ca++- and TPA-mediated TGK mRNA induction by cycloheximide (Fig 5) suggests that TGK gene expression is de­pendent on a protein factor(s) that is regulated by PKC. This result may explain findings reported in an earlier study in which TPA­mediated transglutaminase activity was blocked by cycloheximide [28]. Recent documentation of an AP-1 site upstream of the human TGK initiation codon [40] suggests that PKC may regulate tran-· scription of the TGK gene via Fos and Jun family members. Sup­porting this notion, TP A increases luciferase activity in cells trans­fee ted with a rabbit TGK promoter reporter construct, and this increase is blocked by bryostatin [41]. Additional studies are needed to determine whether PKC-mediated changes in transcript stability also contribute to the increased levels ofTGK mRNA seen in TPA­and Ca++-treated cultures.

TCIl PKC isoforms have been described [42J; five of these, PKC ex, <5, E, (, and 1'/, are expressed in cultured mouse epidermal keratino­cytes* [43] . Studies in several cell types indicate that different PKC isozymes have distinct functions [44,45]. The restricted expression of PKC 1'/ in granular cells suggests that this isoform is involved in regulating late stages of epidermal differentiation [46]. Experiments are currently underway to directly assess the role of individual PKC isozymes in regulating keratinocyte differentiation ill vitro.

• Denning MF, Dtugosz AA, Yuspa SH: Redistribution of specific protein kinase C isozymes accompany Ca2+ -induced keratinocyte differentiation (abstr) .) Invest Dermato/100:495, 1993.

414 DLUGOSZ AND YUSPA

The results of this study may have implications for the treatment of a variety of skin disorders. For example, marked thickening of the stratum corneum is frequently a prominent feature in psoriasis. This change may be secondary to the increased TGK mRNA levels [47] and transglutaminase activity [48] that have been reported in le­sional skin. Identification of PKC as a key regulator of TGK gene expression suggests that selective inhibitors of this signaling path­way may be useful in treating dermatoses characterized by hyper­keratosis.

We are gratiful to Dr. Robert Rice (UllillersilY of Calij"omia, Davis, CAY for providing TGK eDNAs, Dr. George Pettit (Arizofla Stale Ulliversity, Tempe, AZ) for bryostatill 1, Dr. Jorge Kirilovsky (Glaxo Laboratories, Les Ulis, Frallee) for GF 109203X, alld Dr. Ulrike Licltti for reviewillg the marlllscript. Dr. Dlugosz lVas supported, iI, part, by the Thomas B. Fitzpatrick-Kao Corporatiol' Research a IVa rd.

REFERENCES

1. Roop DR, Nakazawa H. Mehrel T. Cheng C, Chung S, Rothnagel JA, Steinert PM, Yuspa SH: Sequential changes in gene expression during epidermal differ­ennatlOn.ln: Rogers GE, Rels PJ. Ward KA, Marshall RC (cds.). The Biology oJ Wool otld Hair. C hapman and Hall, London. 1988. pp 311-324

2. Fuchs E: Epidermal differentiation: the bare essentials. J Cell Bioi 111 :2807 -2814, 1990

3. Polakowska RR, Goldsmith LA: The cell envelope and transglutaminases. In: GoldsmIth LA (cd.). Physiology, Biochelllistry, atld Molewlar Biology of Ihe Skltl . Oxford, New York, 1991, pp 168-201

4. Rice RH, Mehrpouyan M. O'Callahan W, Parenteau NL. Rubin AL: Keratino­cyte transglutaminase: differentiation marker and member of an extended fam­ily. El'ithel Cell Bioi 1:128- 137, 1992

5. Yuspa SH: Methods for the use of epidermal cell culture to study chemical carcin­ogens. In: Skerrow D, Skerrow C (cds.). Met/,ads itl Skitl Research. John Wiley and Sons, Sussex. 1985, pp 213-249

6. Hennings H, Michael D, Cheng C. Steinert P, Holbrook KA, Yuspa SH: Calcium regulation of growth and differentiation of mouse epidermal cells in culture. Cell 19:245-254. 1980

7. Yuspa SH, Kilkenny AE. Steinert PM. Roop DR: Expression of murine epidermal d,fferentiation markers IS tightly regulated by restricted extracellular calcium concentrations in vitro.J Cell Bioi 109:1207 -1217. 1989

8. Lichti U. Yuspa SH: Modulation of tissue and epidermal transglutaminases in mouse epidermal cells after treatment with 12-0-tetradecanoylphorbol-13-acetate and/or retinoic acid in vivo and in culture. Cotlcer Res 48:74 -81, 1988

9. Malmquist KC, Carlson LE. Forslind B, Roomans GM. Akselsson DR: Proton and electron microprobe analysis of human skin . Nucl [lIstr Methods Phys Res 3:611-617.1984

t O. Mcn~n GK, Grayson S, Eli~ P.M: lon,ic calcium reservoirs in mammalian cpidcr­Ims: ultrastructural localIzation by lon-capture cytochemistry.J [livest Dermatol 84:508-512,1985

11. Menon GK. Elias PM: Ultrastructural localization of calciuO) in psoriatiC and normal human epidermis. Arch Dermatol127:57 -63,1991

12. Yuspa SH. Ben T, Hennings H: The induction of epidermal trallsglutaminaseand terminal differentiatioll by tlllllor promoters in cultured epidermal cells. Carci-1I0gfllesis 4:1413-1418,1983

13. Jetten AM. George MA, Pettit GR, Herald CL, Rearick JI: Action of phorbol esters, bryostatins, and retinoic acid on cholesterol sulfate synthesis: rclation to the multistep process of differentiation in human epidermal kcratinocytes. ] bIVest DermatoI93 :108 - 115 , 1989

14. Tang W, Ziboh VA, IsscroffRR. Martinez D: Turnover of inositol phospholipids in cultured murine keratinocytes: possible involvement of inositol trisphos­phate in cellular differentiation.] bIVest Dermato/90:37 -43. 1988

15. Jaken S, Yuspa SH: Early signals for keratinocyte differentiation: role of Ca2+_ mediated inositol lipid metabolism in normal and neoplastic epidermal ce lls. Carcillog,tlesis 9:1033-1038,1988

16. Lee E, Yuspa SH: Changes in inositol phosphate metabolism arc associated with terminal differentiation and neoplasia in mouse keratinocytes. Carcitlogetlesis 12:1651 -1658. 1991

17. Kruszewski FH, Hennings H, Yuspa SH, Tucker RW: Regulation of intracellular free calcium in normal murine keratinocytes. Am J Pllysiol 261 :C767 - C773, 1991

18. Ziboh V A, IsseroffRR, Pandey R: Phospholipid metabolism in calcium-regulated differentiation in cultured murine keratinocytes. Biochem Biophys Res Comm'lII 122:1234-1240,1984

19. Dlugosz AA. Yuspa SH: Coordinate changes in gene expression which mark the spinous to granular cell transition in epidennis are regulated by protein kinase c.J Cell Bioi 120:217-225. 1993

20. ChirgwinJM, PrzybylaAE, MacDonald Rj, RutterWJ: Isolation of biologically active ribonucleic acid from sources enriched in ribonuclease. Biochemistry 18:5294-5299.1979

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

39.

40.

41.

42.

43 .

44.

45.

THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

Badley JE, Bishop GA, St. John T, Frelinger JA: A simple, rapid method for the purification of Poly A+ RNA. Biotecllllology 6: 114 - 116, 1988

Dlugosz AA, Yuspa SH: Staurosporine induces protein kinase C agonist effects and maturation of normal and neoplastic mouse keratinocytes itl vitro. Cancer Res 51 :4677 - 4684, 1991

Sambrook J, Fritsch EF, Maniatis T: MoleCIIlar Clollitlg: A Laboratory Ma/lllal. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, 1989

Phillips MA, Stewart BE, Qin Q, C hakravarty R. FloydEE,Jetten AM. Rice RH: Primary structure of keratinocyte transglutaminase. Proc Natl Acod Sci USA 87:9333-9337,1990

Fort p. Marty L, Piechaczyk M, cl Sabrouty S, Dani C,jeanteur p. BlanchardJM: Various rat adult tissues express only one major mRNA species from the glycer­aldehyde-3-phosphatc-dehydrogenase multigcnic family. Nucleic Acids Res 13:1431 - 1442. 1985

Greenberg ME, Ziff BE: Stimulation of 3T3 cells induces transcription of the c-fos proto-oncogene. Nature 311:433-438,1984

Strickland JE, Jetten AM, Kawamura H, Yuspa SH: Interaction of epidermal growth fac tor WIth basal and dIfferentiating epidermal cell s of mice resistant and sensitive to carcinogenesis. Carei/lOgellesis 5:735 - 740, 1984

Yuspa SH, Ben T, Hennings H , Lichti U: Phorbol ester tumor promoters induce epIdermal transglutaminase activity. Biochelll Biophys Res Commutl 97:700-708,1980

Toullec D. Pianetti P, Coste H, Bellevergue P, Grand-Perret T, Ajakane M. Baudet V. Boissin P, Boursier E. Loriolle F, Duhamel L, Charon D Kiril~vsky J: Thebisindolylmaleimide GF 109203X is a potent and sclectiv~ IIllubltor of protelll kInase C. J Bioi Chelll 266:15771-15781. 1991

Lee LS, Weinstein IB: Tumor-promoting phorbol esters inhibit binding of epi­dermal growth factor to cellular receptors. ScietlCf 202:313- 315, 1978

Sako T , Yuspa SH. Herald CL. Pettit GR, Blumberg PM: Partial parallelism and partIal blockade by bryostatin 1 of effects of phorbol ester tumor promoters on primary mouse epidermal cel ls. Catlcer Res 47:5445-5450.1987

Pettit GR, Herald CL, Doubek DL. Herald DL: Isolation and structure ofbryo­statin l.J Am Chem Soc 104:6846-6848, 1982

Blumberg PM, Pettit GR, Warren BS, Szallasi A, Schuman LD, Sharkey NA, Nakakuma H,Dell'Aquila ML. de Vries DJ: The protein kinase C pathway in tumor promotion. Prog Clill Bioi Res 298:201-212. 1989

Huff CA, Yuspa SH. Rosenthal D: Identification of control elements 3' to the human keratin 1 gene that regulate cell type and differentiation-specific ex­pression.J Bioi Chelll 268:377-384, 1993

Yada Y, Polakowska RR. Okano Y, Nozawa Y: Protein kinase C-dependent expressIOn of type I transglu~aminase mRNA in ganglioside GQ'b- and cal­clum-stlmulatcd human keratl11ocytcs . Biael,em Biopllys Res ComnJlHl 190:688-694, 1993

Liew F. Yamanishi K: Regulation of transglutaminase 1 gene expression by 12-0-tetradecanoylphorbol -1 3-acetate, dexamethasone, and retinoic acid in cul­tured human keratinocytes. Exp Cell Res 202:310-315,1992

An~el P, Karin M: The role ofJun, Fos and the AP-l complex in cell-prolifera­tion and transformation. Biochi/II Biophys Acta Rev Callcer 1072:129-157 1991 '

Blumberg PM: Complexities of the protein kinase C pathway. Mol Carcitlog 4:339-344,1991

Matsui ~S, Chew SL, DeL~o VA: Protein kinase C in normal human epidermal keratmocytes dUTlng prolIferatIOn and calcium-induced differentiation.J III vest DermatoI99:565-571,1992

Yamanishi K, Inazawa J, Liew F, Nonomura K. Ariyama T, Yasuno H, Abe T, DOl H. Htrano J, Fukushima S: Structure of the gene for human transglutamin­ase I.J Bioi Clrern 267:17858-17863, 1992

Saunders NA, Bernacki SH,. Vollberg TM,Jetten AM: Regulation of transgluta­mlllase type I expressIOn 111 squamous differentiating rabbit tracheal epithelial cells and human epidermal keratinocytes: effects of retinoic acid and phorbol esters. Mol Elldocrillol7:387 -398 1993

NishizukaY: hltracellularsignaling by hydrolysis of phospholipids and activation of protem kInase C. SciCllce 258:607 -614. 1992

Dlugosz AA, Mischak H. Mushinski JF, Yuspa SH: Transcripts encoding protein kmase C alpha, delta. epsilon, zeta, and eta arc expressed in basal and differen­tlatlI~g mouse keratinocytcs in vitro and exhibit quantitative changes in neo­plastIC cells. Mol Carci/lOg 5:286-292, 1992

Otte AP~ Moon RT: Protein kinase C isozymes have distinct roles in neural 111ductlOn and competence in Xenopus. Cell 68:1021-1029. 1992

Gusovsky F, Gutkind JS: Selective effects of activation of protein kinase C iso­zymes on cyclic AMP accumulation. Mol Ph"rlllaco/39:124-129, 1991

46. Osad~ S. Hashim?to Y. Nomura S. Kohno Y, Chida K, Tajima 0, Kubo K, Akll110tO K. KOIzumi H, Kitamura Y, Suzuki K. Ohno S, Kuroki T: Predomi­na~t ex.pres~ion of ~1PKC e~a, . a Ca-indepcndent isofonn of protein kinase C in epIthelIal tissues, 111 aSSOCIatIOn with epithelial differentiation. Cell Crowth Differ 4:167-175, 1993

47.

48.

Schroeder WT, Thacher SM, Stewart-Galetka S Annarclla M Chema D Siciliano M~. Davies PJ. Tang HY. Sowa BA, D~vic M: Type I 'keratinocyt~ transglutamlllase: expreSSIon 111 human skin and psoriasis. ] Ir,Vest D ennatol 99:27 - 34, 1992

Esmann J . Voorhees JJ. Fisher GJ: Increased membrane-associated transgluta­mmase activIty 111 psonaSlS. Bioclrelll Biopllys Res Com 111/1 II 164:219-224 1989 •