Differentiation (1992) 49: 77-83
Differentiation Ontogeny, Neoplasia and Differentiation Therapy
0 Springer-Verlag 1992
Hormonal regulation of adult type keratin gene expression in larval epidermal cells of the frog Xenupus laevis Keiko Shimizu-Nishikawa and Leo Miller * Laboratory of Molecular Biology, Department of Biological Sciences, University of Illinois at Chicago, Chicago, Illinois, 60680, USA Accepted in revised form December 17, 1991
Abstract. Triiodothyronin (T,) is known to induce amphibian metamorphosis but other hormones such as glucocorticoids accelerate T, action. The increase in plasma concentration of both T, and glucocorticoids during metamorphic climax is correlated with the transformation of the epidermis from larval type (uncornified) to adult type (cornified). Previously we have shown that T3 induced adult-type 63 Kd keratin gene expression and cornification of the larval epidermis. In this study, we have examined the effects of T3 and hydrocortisone (HC) on the conversion of larval to adult epidermal cells in vitro. When larval epidermal cells were treated with both T, and HC, they had a synergistic effect on adulttype keratin synthesis (both 63 Kd and 49 Kd keratins) and epidermal cornification. The synergistic effect between T, and HC required a pretreatment with T3 for 3 days. During this time, addition of HC to cultures containing T, did not change the amount of 63 Kd keratin mRNA. Thus, HC did not reduce the lag time for epidermal cells to respond to T,. After 4days of hormone treatment, T, increased the amount of 63 Kd keratin mRNA 9-fold while T, and HC induced it 18-fold. When cultures were pretreated with T, for 3 days, a 1 day treatment with HC was sufficient to obtain the synergistic effect. Thus the induction of 63 Kd keratin gene expression by T, required a much longer lag (3 days) than the lag required for the synergistic action of T3 and HC (< 1 day). These results demonstrate for the first time that the synergistic effects of T, and glucocorticoid on metamorphosis occur at the mRNA level in a purified population of cells.
Introduction Amphibian metamorphosis, which is induced by the thryoid hormone triiodothyronine (T,), involves a dynamic remodeling of the body including cell death,
* To whom offprint requests should be sent
growth, and differentiation (For review, [26]). Observations of morphological changes during metamorphosis have demonstrated that exogenous adrenal corticoids accelerate thyroid hormone-induced changes in vivo [6, 121 and in tail cultures [ll]. The plasma concentration of glucocorticoids increase in parallel with T3 during metamorphosis [9] and inhibition of endogenous corticoid synthesis retards thyroxine-induced metamorphosis [lo]. These results suggest an important interplay between endogenous corticoids and T, during amphibian metamorphosis. However, the interaction of T3 and corticoids during metamorphosis has not been analyzed at the cellular or molecular levels. The Xenopus epidermis is a good system for studying the mechanism of hormone induced differentiation during metamorphosis. At metamorphosis, T, induces the epidermis to change from a larval type to an adult type. Larvae have an uncornified bilayered epidermis which will not differentiate until metamorphosis. On the other hand, adults have a stratified epidermis with terminally differentiated cornified cells in the outermost layer (For review, [ 51). Furthermore, epidermal keratins change from a larval type to an adult type during metamorphosis [3]. In larval skin explant cultures, cornification and high-level expression of the adult specific 63 Kd keratin gene is induced by T3 [ 15, 161. The mechanism by which T3 and steroid hormones affect gene expression has been studied extensively in adult mammalian cells. Hormone-receptor complexes bind to specific DNA sequences called hormone response elements and induce transcription of tissue specific genes (For review, [l, 41). Although hormone induced gene expression is immediate and reversible in adult mammals, the induction of 63 Kd keratin gene expression by T, during metamorphosis requires a long lag period and is irreversible. This suggests that the control of tissue specific gene expression during metamorphosis is different from that observed in adult mammalian cells [16]. Thus, it will be interesting to determine the mechanism by which T, and glucocorticoids interact during amphibian metamorphosis.
78
In this report, we have used cultures of purified epidermal cells [20] to investigate the effects of hydrocortisone (HC) on differentiation and gene expression. We demonstrate that T3 and HC have a synergistic effect on epidermal cornification and adult type keratin gene expression. We also show that the synergistic effect of T3 and HC on 63 Kd keratin gene expression does not occur until the gene is activated by a pretreatment with 7-3.
Methods Isolation of epidermal cells. Fertilized eggs of Xenopus luevis were grown in our laboratory. Tadpoles at stage 52-54 [17] were used for experiments. Epidermal cells were isolated as described previously [20]. Body skins, excluding the tail, were used for all experiments. The skins were digested at room temperature with 0.18% trypsin for 5 min and with 2.5 U/ml dispase for 30 min and then incubated for 18 h at 4" C. Cells were dissociated by pipetting the dispase-digested skins. Epidermal cells were purified by two separation steps including a continuous gradient of Percoll and a discontinuous gradient of bovine serum albumin and Percoll. Culture of isolated cells and skins. Purified epidermal cells were seeded at 1.3 x 105-1.8 x 105/cmZon fibronectin coated dishes. Epidermal cells or dissected skins were cultured with a 1 :3 mixture of Ham's F12 and Eagle's minimum essential medium (MEM) with nonessential amino acids without calcium chloride (Whittaker Bioproducts, Walkersville, Md., USA). The culture medium was supplemented with 5% Chelex- and charcoal-treated fetal calf serum and 20 pg/ml insulin. The calcium concentration of the medium was adjusted to 0.05 m M unless specified otherwise. The cultures were incubated at 25" C in a humidified chamber gassed with 5% C 0 2 and 95% air. Gel electrophoresis. Cultured epidermal cells, cultured skins, or dissected skins from tadpoles which were grown in hormone containing water were labeled for 3 h with 100 pCi/ml of Tran ["S] label (ICN Radiochemicals, Irvine, Calif., USA). Cytoskeletal fractions of the scraped cells or skins were extracted in low salt buffer and high saIt buffer containing Triton X-100 [15]. We previously identified these cytoskeletal proteins as keratins using biochemical and immunological techniques [3, 241. The cytoskeletal fractions were subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) [ 131 or two-dimensional electrophoresis with nonequilibrium pH gradient electrophoresis (NEPHGE) [21]. The gels were soaked in salicylic acid, dried, and exposed to X-ray film. R N A isolation. Total RNA of cultured epidermal cells was extracted with guanidium thiocyanate followed by centrifugation in cesium chloride solutions [22]. Slot blot analysis was performed as described by Sambrook et al. [22] and the manual for Genescreen Plus (New England Nuclear, Boston, Ma., USA). Denatured total RNA (2 pg) was blotted onto Genescreen Plus using a Minifold I1 apparatus (Schleicher & Schuell, Keene, N.H., USA). The filters were prehybridized for 4 h and then hybridized with [35P]CTPlabeled RNA probes (lo6 cpm/ml) for 18 h at 60" C. Prehybridization and hybridization buffer contained 50% formamide, 5 x standard saline citrate (SSC), 5 x Denhardt's solution, 100 pg/ml herring DNA, 40 m M sodium phosphate buffer and 1% SDS. The filters were washed once in 2 x SSC at room temperature and three times in 0.1 x SSC at 65" C and then exposed to X-ray film. Slot blots were quantitated by scanning the X-ray film with a Hoeffer GS 300 densitometer. All of the genes used in this study were cloned into Bluescribe [15]. pM7 contains the 3' /$turn and untranslated region of the Xenopus 63 Kd keratin cDNA, pUF164 [8]. pM11 contains Xenopus rDNA sequences from pXlrl1 [2].
Results Induction of 63 Kd keratin synthesis by T, or HC
We have previously shown that purified larval epidermal cells respond to T3 and synthesize the adult specific 63 Kd keratin [20]. As shown in Fig. 1, M (lane 2) and lo-* M (lane 3) T3 induced snythesis of the 63 Kd keratin in larval epidermal cells. 10-l' A4 T3 had no effect on 63 Kd keratin synthesis (data not shown). Unexpectedly 63 Kd keratin synthesis was also induced by HC alone. It is known that corticoids themselves do not induce metamorphosis [ I l , 121. 5 x lo-' A4 HC induced high level synthesis of the 63 Kd keratin (Fig. 1, lane 5). However, unlike T,, which induces high level 63 Kd keratin synthesis over two orders of magnitude, only restricted concentrations of HC induced high level expression. We regard the T,-induced 63 Kd protein and HC-induced 63 Kd protein as basic keratins because they react with monoclonal antibody AE3 [25] which recognizes all known basic keratins (data not shown). Since the induction of keratin synthesis by HC was unexpected, we examined the effect of HC on the kinetics of keratin synthesis in cultured epidermal cells, cultured skin and tadpoles. As shown in Fig. 2, HC induced 63 Kd keratin synthesis in cultured epidermal cells on the second day but not at any time in cultured skin or in vivo. These results suggest that the HC effect may be an artifact of the cell culture technique since HC had no effect by itself on skin organ culture or in vivo, where the normal histological organization of the skin is maintained. Furthermore, analysis of mRNA levels (see below) have shown that the unique response of cultured epidermal cells to HC is at the translational level. In contrast to HC, T, induced 63Kd keratin synthesis in both cell and organ culture as well as in vivo (Fig. 2). Synergistic effects of T, and HC on keratin synthesis Keratin synthesis in cultures. The combined effects of T 3 and HC on keratin synthesis was also examined in epidermal cell cultures. Combined T3 and HC induced 63 Kd keratin synthesis more than either T3 alone or HC alone (Fig. 3). Furthermore, 49 Kd keratin synthesis
1
2
3
4
5
6
Fig. 1. Effects of triiodothyronine (T3) and hydrocortisone (HC) on keratin synthesis in cultured epidermal cells. Epidermal cells were cultured with no hormone (lane I ) , lo-" T3 (lane 2), lo-' M T3 (lanes), 5 x lo-' M HC (lune4), 5 x lo-' M HC (lane 9, or 5 x M HC (lane 6 ) for 4 days. After labeling with [35S] amino acids, the cytoskeletal fractions were isolated, subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), and detected by fluorography. a, 63 Kd keratin; b, actin
1
2
3
4
5
6
7
8
9
A
-a
B
-a
-b C
---2
1
3
5
4
2days
3days
6
7
8
9
10
4days
-a
=t 1
2
3
4
Fig. 3. Combined effects of T, and HC on keratin synthesis in cultured epidermal cells. Epidermal cells were cultured with no hormone (lane i), lo-' M T, (lane 2), 5 x M HC (lane 3), or lo-' M T3 and 5 x M HC (lane 4 ) for 4 days. After labeling with [35S]amino acids, the cytoskeletal fractions were isolated, subjected to SDS-PAGE, and detected by fluorography. a, 63 Kd keratin; b, 49 Kd keratin; c, actin
was induced to a high level by T3 and HC (Fig. 3, lane 4). The 49 Kd keratin, as well as the 63 Kd keratin, are adult type keratins in Xenopus and both are initially expressed in the larval epidermis during metamorphosis [3]. The induction of 63 kD keratin synthesis by T, and HC was also reflected by changes in the amount of 63 Kd keratin as measured by Coomassie brilliant blue staining and immunoblotting with AE3 (data not shown). These results suggest that T, and HC have a synergistic action on adult keratin synthesis. To examine the combined effect of T3 and HC on keratin synthesis in more detail, two dimensional electrophoresis was performed. In agreement with the one dimensional electrophoresis data, 63 Kd and 49 Kd keratin synthesis were induced to high levels with T3 and HC (Fig. 4). Two dimensional electrophoretic patterns showed that the 63 Kd keratins induced by T,, HC, or T, plus HC migrated at exactly the same positions. As previously demonstrated by Hoffmann et al. [S], the 63 Kd keratins appeared as three spots. Therefore, our data suggest that T,-induced, HC-induced, and T, plus HC-induced keratins are the same polypeptides. Two dimensional electrophoretic patterns also showed that the 49 Kd keratin is an acidic keratin which corresponds to Xenopus keratin IV [7]. In contrast to induction of
11
12
5days
-
b
Fig. 2A-C. Kinetics of hormone induction of 63 Kd keratin synthesis. Epidermal cells (A), dissected skins (B), or tadpoles (C) were kept in media containing no hormone (lanes i, 4 , 7 , and fO), M T3 (lanes 2, 5, 8, and fl), or 5 x M HC (lanes 3, 6, 9, and 12) for the indicated periods. After labeling with amino acids, the cytoskeletal fractions were isolated, subjected to SDS-PAGE, and detected by fluorography. a, 63 Kd keratin; b, actin
adult type keratin synthesis by T 3 or T3 plus HC, synthesis of larval type keratins were decreased by T, and HC but not by T, alone (compare Fig. 4D with Fig. 4A and B). 63 Kd keratin synthesis in vivo. To determine whether HC acts synergistically with T, in vivo, keratin synthesis was examined in tadpoles which were grown in water containing hormones for 1-5days. The same amount of newly synthesized 63 Kd keratin was detected in both T,- and T, plus HC-treated tadpoles from the second day to fourth day of hormone treatment (Fig. 5, lanes 5, 6, 8, 9, 11, and 12). However, on the fifth day of hormone treatment, tadpoles treated with T, plus HC synthesized more 63 Kd keratin than tadpoles treated with only T,. These results demonstrate that HC works synergistically with T, on adult type keratin synthesis not only in vitro but also in vivo.
Synergistic effects of T, and HC on expression of 63 Kd keratin mRNA To investigate the mechanism of the synergistic effects of T, and HC in more detail, we examined the expression of 63 Kd keratin mRNA in cultured larval epidermal cells. Slot blots of total cell RNA were hybridized with pM7 (Fig. 6A), a cDNA subclone from the 3' region of the 63 Kd keratin gene. The specificity of pM7 for the 63 Kd keratin mRNA was demonstrated by Northern blots, RNase protection and hybrid selection followed by in vitro translation [15, 161. Blots were also hybridized with a probe for rRNA (pM11; Fig. 6B) to demonstrate that equal quantities of RNA were loaded in all slots. As shown in Fig. 6A, 63 Kd keratin mRNA was induced 9-fold by T, (slot 2) and 18-fold by T3 and HC (slot 4) on the fourth day of hormone treatment. However, compared to the low level expression of the 63 Kd keratin gene seen in control cultures, no induction was observed by HC alone (slot 3). These results and
80
Fig. 4A-D. Combined effects of T3 and HC on keratin synthesis in cultured epidermal cells. Epidermal cells were cultured in media containing 0.3 mM calcium with no hormone (A), lo-' M T, (B), 5 x lo-' M HC (C), or lo-' M T, and 5 x lo7 M HC (D) for 4 days. After labeling with amino acids, the cytoskeletal fractions were isolated and subjected to two-dimensional electrophoresis. First-dimension is nonequilibrium pH gradient electrophoresis (NEPHGE); from left (acidic) to right (basic). Second-dimension is SDS-PAGE. Proteins were detected by fluorography. a, 63 Kd keratin; b, 49 Kd keratin; c, actin
-----
1
3
2
4
lday
5
2days
6
7
8
3days
9 1 0
11
4days
Fig. 5. Combined effects of T3 and HC on 63 Kd keratin synthesis in vivo. Tadpoles were grown in water containing no hormone (lanes I , 4 , 7, fa,and 13), lo-' M T3 (lanes 2, 5, 8, If, nrzd I4), or lo-' M T, and 5 x M HC (lanes 3, 6 , Y, 12, and 15) for
A
1
2
3
4
Fig. 6A, 6. Induction of 63 Kd keratin mRNA with T3 and HC in cultured epidermal cells. Epidermal cells were cultured in media containing 0.3 m M calcium with no hormone (slot I), lo-' M T, (slot 2), 5 x lo-' M HC (slot 3), or lo-* M T, and 5 x lo-' M HC (slot 4 ) for 4 days. Total RNA from epidermal cells was hybridized with cRNA probes prepared from pM7 (A), a 3'-end subclone of 63 kD keratin cDNA, or pM11 (B), a subclone of pXlrl1 which detects rRNA
the keratin synthesis data suggest that induction of 63 Kd keratin synthesis by HC alone is restricted to cells in culture and occurs at the translational level. More importantly they also suggest that HC works synergisti-
12131415 5days
the indicated periods. Skins were dissected and then labeled with [35S] amino acids. The cytoskeletal fractions were isolated, subjected to SDS-PAGE, and detected by fluorography. a, 63 Kd keratin; b, actin
cally with T, at the mRNA level and that HC alone does not alter the basal level of 63 Kd keratin mRNA present in control cells. Unlike the induction of gene expression by T, or steroid hormones in adult mammalian cells, induction of 63 Kd keratin mRNA by T3 requires a long lag period [16]. We therefore tested whether HC can decrease this lag period. As shown in Fig. 7, no induction of 63 Kd keratin mRNA was observed on the second day of culture, with or without hormones (slots l , 2, and 3). However, on the third day of culture, 63 Kd keratin mRNA was induced twofold with T, or T3 plus HC (slots 5 and 6). This result indicates that HC does not change the lag period which is required for induction of 63 Kd keratin mRNA by T,. We also examined how long it takes to get synergism between T, and HC. Larval epidermal cells were treated with T, for 4 days. During this time, HC was added to the T,-containing cultures for various periods. As shown in Fig. 8, treatment with HC for 4 days (slot 2), the last 2 days (slot 3), or the last 1 day (slot 4) induced larger amounts of 63 Kd keratin mRNA than induced by T3 alone (slot 6). In comparison to treatment with T, alone (slot 6), addition of HC
81
A
B
1
2 3 4 5 6 u u 2days
3days
Fig. 7A, B. Kinetics of hormone induction of 63 Kd keratin mRNA in cultured epidermal cells. Epidermal cells were cultured in media containing 0.3 mM calcium with no hormone (slots 1 and 4), lo-* M T, (slots 2 and 5), or lo-* M T, and 5 x M HC (slots 3 and 6 ) for the indicated periods. Total RNA from epidermal cells was hybridized with cRNA probes prepared from pM7 (A), a 3'-end subclone of 63 Kd keratin cDNA, or pM11 (B), a subclone of pXlrl1 which detects rRNA
in the following experiments larval epidermal cells were cultured with high calcium media (0.3 mM). Cornified cells were observed on the second day in culture with T, and HC (Fig. 9D) but not until the third day in cultures treated only with T3 (Fig. 9B, F). Further, on the third day of culture, more cornified cells were observed in cultures treated with T, and HC (Fig. 9H) than with T, alone (Fig. 9F). Treatment with T, and HC also induced a change in the pattern of cornification. Cultures which were treated with T, and H C had sheetlike cornified cell layers whereas T, treated cultures had mainly single cornified cells. Cultures without hormones or with HC alone did not show any cornified cells (Fig. 9A, C, E, G). Thus, HC induced 63 Kd keratin synthesis in these cultures but did not induce cornification. When larval epidermal cells were cultured with low calcium media (0.05 mM), cornified cells were not produced even in the presence of T3 and HC (data not shown). Discussion
B
1
2
3
4
5
6
Fig. 8A, B. Kinetics of the synergistic effects of T, and HC on expression of 63 Kd keratin rnRNA. Epidermal cells were cultured in media containing 0.3 mM calcium with no hormone (slot 1) or lo-' M T, (slots 2, 3, 4 , 5, and 6)for 4 days. HC (final concentration 5 x lo-' M ) was added to the T,-containing cultures for 4 days (slot 2), the last 2 days (slot 3), the last 1 day (slot 4), or the last 6 h (slot 5). Total RNA from epidermal cells was hybridized with cRNA probes prepared from pM7 (A), a 3'-end subclone of 63 Kd keratin cDNA, or pM11 (B), a subclone of pXlrl1 which detects rRNA
to the T,-containing culture for the last 6 h (slot 5) didn't show a significant increase of 63 Kd keratin mRNA. Thus the synergistic effect of T3 and HC required a pretreatment with T3 for 3 days after which a 1 day treatment with HC was sufficient for the synergistic effect. During the 3 day pretreatment with T,, HC had no effect on 63 Kd keratin mRNA expression.
Synergistic effects of T3 and HC on epidermal corn fieation Since cornification is a landmark of epidermal differentiation in adult frogs [5] and can be induced by T3 in larval skin explant cultures [16], we examined the combined effects of T, and HC on cornification of larval epidermal cells. Our previous results indicated that a calcium concentration of at least 0.15 m M was required for cornification of adult epidermal cells [24]. Therefore,
It has been known for almost 40 years that glucocorticoids accelerate T,-induced amphibian metamorphosis [6]. Since cell cultures were not previously used to analyze the interactions of T, and glucocorticoids, it was not known whether glucocorticoids act directly on target cells or indirectly through other tissues and organs. In this study, using cultures of purified larval epidermal cells, we have demonstrated that H C acts synergistically with T3 on gene expression during amphibian metamorphosis. Interestingly, this synergistic effect between T3 and HC on 63 Kd keratin gene expression required pretreatment of the cells with T, for 3 days and during this time HC had no effect. Our previous results demonstrated that expression of the 63 Kd keratin gene is regulated at two levels: a low level, basal expression independent of T,; and a T3induced high level expression [ 151. HC synergizes with T3 only after the gene is induced by T,; it does ot have any effect on the basal expression of the keratin gene. While it is clear that the synergistic action of HC and T, is at the mRNA level, further experiments will be necessary to determine whether the transcription rate and/or the stability of the 63 Kd keratin mRNA is changed. Two observations suggest that the synergistic effect cannot be due solely to the stabilization of mRNA. First, HC by itself does not change the basal level of 63 Kd keratin mRNA and thus it is unlikely that HC alters the stability of the T,-induced mRNA. Second, T3 plus H C also have a synergistic effect on the timing and extent of epidermal cell cornification, 49 Kd keratin synthesis and the disappearance of larval keratins. It seems unlikely that the multiple mRNAs required for these aspects of epidermal cell differentiation would all be controlled by mRNA stabilization. Previously, we demonstrated that expression of the 63 Kd keratin gene by T, showed a long latent period and suggested that regulatory factors, which arc induced directly by T,, arc involved in the expression of tissue-
82
Fig.9A-H. Effects of T3 and HC on cornification of cultured epidermal cells. Epidermal cells were cultured with no hormone (A and E), lo-* M T3 (B and F), 5 x M HC (C and G ) , or l o - * M T3 and 5 x M HC (D and H). Photographs were taken on the second (A, B, C, and D) or third (E, F, G, and H) day of culture
83
specific genes [16]. Similar lag times have been reported recently for the T,-induced down regulation of the Xenopus trypsin gene [23] and the induction of N-CAM [14]. In Rana catesbeiuna, inhibition of tail DNA synthesis by T, also requires a long lag period [18, 191. The data presented here demonstrate that HC does not reduce the lag time necessary for T3 to induce high level expression of the keratin gene. However, it is not clear whether HC acts indirectly (through the expression of regulatory factors) or directly on the 63 Kd keratin gene. It is likely that HC acts more directly than T3 on 63 Kd keratin gene expression because it acts more rapidly (< 1 day) than T, (3 days). To clarify this point, direct assay of cis-regulatory elements of the 63 Kd keratin gene will be necessary. While only T, can initiate amphibian metamorphosis, it is clear that other hormones such as glucocorticoids are also essential. As shown here, HC works synergistically with T3 during differentiation of the epidermis. HC not only increases T,-induced cornification and expression of the 49 Kd and 63 Kd keratins but it also promotes normal epidermal cell differentiation in vitro in other ways. During the transformation of the epidermis from larval to adult programs of differentiation, larval-type keratins are replaced by adult-type keratins [ 3 ] .T3 alone did not reduce larval type keratin synthesis, but T3 and HC together induced a drastic reduction in larval keratin synthesis (Fig. 4). Furthermore, T, and HC treated larval epidermal cells produced sheets of cornified cells as seen in vivo, while T3 by itself induced only scattered single cornified cells (Fig. 9). These results indicate that glucocorticoids play an integral part in skin differentiation and other events occurring during metamorphosis. Acknowledgements. We would like to thank Prof. S. Kikuyama, Prof. K. Yoshizato, and Dr. A. Nishikawa for helpful suggestions. We are grateful to Dr. W. Hoffmann for supplying the cDNA clone pUF164 and Dr. J. Doering for supplying the rDNA clone. This study was supported by a National Institutes of Health grant (HD 24438).
References 1. Beato M (1989) Gene regulation by steroid hormones. Cell 56: 335-344 2. Dawid IB, Wellauer PK (1976) A reinvestigation of 5’-3’ polarity in 40 S ribosomal RNA precursor of Xenopus lueuis. Cell 8 :443-448 3. Ellison TR, Mathisen PM, Miller L (1985) Development changes in keratin patterns during epidermal maturation. Dev Biol 112:329-337 4. Evans RM (1988) The steroid and thyroid hormone receptor super family. Science 240: 889-895 5. Fox H (1986) The skin of amphibia. In: Bereiter-Hahn J, Matoltsy AG, Richards KS (eds) Biology of the Integument, Vol. 2. Spring-Verlag, New York, pp 78-108 6. Frieden E, Nail B (1955) Biochemistry of amphibian metamorphosis. I. Enhancement of induced metamorphosis by glucocorticoids. Science 121: 37-39
7. Hoffmann W, Franz JK (1984) Amino acid sequence of the carboxy-terminal part of an acidic type I cytokeratin of molecular weight 51 000 from Xenopus luevis epidermis as predicted from the cDNA sequence. EMBO J 3: 1301-2306 8. Hoffmann W, Franz JK, Franke WW (1985) Amino acid microheterogeneity of basic (type 11) cytokeratins of Xenopus epidermis and evolutionary conservativity of helical and non-helical domains. J Mol Biol 184 :713-724 9. Jolivet Jaudet G, Leloup Hatey J (1984) Variations in aldosterone and corticosterone plasma levels during metamorphosis in Xenopus laevis tadpoles. Gen Comp Endocrinol 56: 59-65 10. Kikuyama S, Niki K, Mayumi M, Kawamura K (1982) Retardation of thyroxine-induced metamorphosis by Amphenone B in toad tadpoles. Endoclinol Jpn 29: 659-662 11. Kikuyama S, Niki K, Mayumi M, Shibayama R, Nishikawa M, Shintake N (1983) Studies on corticoid action on the toad tadpole tail in vitro. Gen Comp Endocrinol 5 2 : 395-399 12. Kobayashi H (1958) Effect of desoxycorticosterone acetate on metamorphosis induced by thyroxine in anuran tadpoles. Endocrinology 62: 371-377 13. Laemmli U K (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227: 680-685 14. Levi G, Broders F, Dunon D, Edelman GM, Thiery JP (1990) Thyroxine-dependent modulations of the expression of the neural cell adhesion molecule N-CAM during Xenopus laeuis metamorphosis. Development 109 :681-692 15. Mathisen PM, Miller L (1987) Thyroid hormone induction of keratin genes: a two-step activation of gene expression during development. Genes Dev I :1101-1 117 16. Mathisen PM, Miller L (1989) Thyroid hormone induces constitutive keratin gene expression during Xenopus laevis development. Mol Cell Biol 9:1823-1831 17. Nieukoop PD, Faber J (1967) In: Nieukoop PD, Faber J (eds) Normal table of Xenopus luevis (Daudin). North Holland Publishing 18. Nishikawa A, Yoshizato K (1986) Hormonal regulation of growth and life span of bullfrog tadpole tail epidermal cells cultured in vitro. J Exp Zoo1 237:221-230 29. Nishikawa A, Kaiho M, Yoshizato K (1989) Cell death in the anuran tadpole tail: Thyroid hormone induces keratinization and tail-specific growth inhibition of epidermal cells. Dev Biol 1311337-344 20. Nishikawa A, Shimizu-Nishikawa K, Miller L (1990) Isolation, characterization, and in vitro culture of larval and adult epidermal cells of the frog Xenopus luevis. In Vitro 26: 1 128-1 134 21. O’Farrell PZ, Goodman HM, O’Farrell PH (1977) High resolution two-dimensional electrophoresis of basic as well as acidic proteins. Cell 12: 1133-1342 22. Sambrook J, Fritsch EF, Maniatis T (1989) In: Sambrook J, Fritsch EF, Maniatis T (eds) Molecular Cloning; A Laboratory Manual, Cold Spring Harbor Laboratory Press, Cold Spring Harbor 23. Shi Y-B, Brown D D (1990) Developmental and thyroid hormone-dependant regulation of pancreatic genes in Xenopus h e vis. Genes Dev 4:1107-1113 24. Shimizu-Nishikawa K, Miller L (1991) Calcium regulation of epidermal cell differentiation in the frog Xenopus lueuis. J Exp Zoo1 (in press) 25. Sun T-T, Eichner R, Schermer A, Cooper D, Nelsen WG, Weiss RA (1984) Classification, expression, and possible mechanisms of evolution of mammalian epithelial keratins: A unifying model. In: Levine A, Topp W, Vande Wounde G, Watson JD (eds) Cancer Cells; Transformed Phenotype, vol. I. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, pp 169-176 26. Yoshizato K (1989) Biochemistry and cell biology of amphibian metamorphosis with a special emphasis on the mechanism of removal of larval organs. Tnt Rev Cytol 119:97-1 49
Differentiation Ontogeny, Neoplasia and Differentiation Therapy
0 Springer-Verlag 1992
Hormonal regulation of adult type keratin gene expression in larval epidermal cells of the frog Xenupus laevis Keiko Shimizu-Nishikawa and Leo Miller * Laboratory of Molecular Biology, Department of Biological Sciences, University of Illinois at Chicago, Chicago, Illinois, 60680, USA Accepted in revised form December 17, 1991
Abstract. Triiodothyronin (T,) is known to induce amphibian metamorphosis but other hormones such as glucocorticoids accelerate T, action. The increase in plasma concentration of both T, and glucocorticoids during metamorphic climax is correlated with the transformation of the epidermis from larval type (uncornified) to adult type (cornified). Previously we have shown that T3 induced adult-type 63 Kd keratin gene expression and cornification of the larval epidermis. In this study, we have examined the effects of T3 and hydrocortisone (HC) on the conversion of larval to adult epidermal cells in vitro. When larval epidermal cells were treated with both T, and HC, they had a synergistic effect on adulttype keratin synthesis (both 63 Kd and 49 Kd keratins) and epidermal cornification. The synergistic effect between T, and HC required a pretreatment with T3 for 3 days. During this time, addition of HC to cultures containing T, did not change the amount of 63 Kd keratin mRNA. Thus, HC did not reduce the lag time for epidermal cells to respond to T,. After 4days of hormone treatment, T, increased the amount of 63 Kd keratin mRNA 9-fold while T, and HC induced it 18-fold. When cultures were pretreated with T, for 3 days, a 1 day treatment with HC was sufficient to obtain the synergistic effect. Thus the induction of 63 Kd keratin gene expression by T, required a much longer lag (3 days) than the lag required for the synergistic action of T3 and HC (< 1 day). These results demonstrate for the first time that the synergistic effects of T, and glucocorticoid on metamorphosis occur at the mRNA level in a purified population of cells.
Introduction Amphibian metamorphosis, which is induced by the thryoid hormone triiodothyronine (T,), involves a dynamic remodeling of the body including cell death,
* To whom offprint requests should be sent
growth, and differentiation (For review, [26]). Observations of morphological changes during metamorphosis have demonstrated that exogenous adrenal corticoids accelerate thyroid hormone-induced changes in vivo [6, 121 and in tail cultures [ll]. The plasma concentration of glucocorticoids increase in parallel with T3 during metamorphosis [9] and inhibition of endogenous corticoid synthesis retards thyroxine-induced metamorphosis [lo]. These results suggest an important interplay between endogenous corticoids and T, during amphibian metamorphosis. However, the interaction of T3 and corticoids during metamorphosis has not been analyzed at the cellular or molecular levels. The Xenopus epidermis is a good system for studying the mechanism of hormone induced differentiation during metamorphosis. At metamorphosis, T, induces the epidermis to change from a larval type to an adult type. Larvae have an uncornified bilayered epidermis which will not differentiate until metamorphosis. On the other hand, adults have a stratified epidermis with terminally differentiated cornified cells in the outermost layer (For review, [ 51). Furthermore, epidermal keratins change from a larval type to an adult type during metamorphosis [3]. In larval skin explant cultures, cornification and high-level expression of the adult specific 63 Kd keratin gene is induced by T3 [ 15, 161. The mechanism by which T3 and steroid hormones affect gene expression has been studied extensively in adult mammalian cells. Hormone-receptor complexes bind to specific DNA sequences called hormone response elements and induce transcription of tissue specific genes (For review, [l, 41). Although hormone induced gene expression is immediate and reversible in adult mammals, the induction of 63 Kd keratin gene expression by T, during metamorphosis requires a long lag period and is irreversible. This suggests that the control of tissue specific gene expression during metamorphosis is different from that observed in adult mammalian cells [16]. Thus, it will be interesting to determine the mechanism by which T, and glucocorticoids interact during amphibian metamorphosis.
78
In this report, we have used cultures of purified epidermal cells [20] to investigate the effects of hydrocortisone (HC) on differentiation and gene expression. We demonstrate that T3 and HC have a synergistic effect on epidermal cornification and adult type keratin gene expression. We also show that the synergistic effect of T3 and HC on 63 Kd keratin gene expression does not occur until the gene is activated by a pretreatment with 7-3.
Methods Isolation of epidermal cells. Fertilized eggs of Xenopus luevis were grown in our laboratory. Tadpoles at stage 52-54 [17] were used for experiments. Epidermal cells were isolated as described previously [20]. Body skins, excluding the tail, were used for all experiments. The skins were digested at room temperature with 0.18% trypsin for 5 min and with 2.5 U/ml dispase for 30 min and then incubated for 18 h at 4" C. Cells were dissociated by pipetting the dispase-digested skins. Epidermal cells were purified by two separation steps including a continuous gradient of Percoll and a discontinuous gradient of bovine serum albumin and Percoll. Culture of isolated cells and skins. Purified epidermal cells were seeded at 1.3 x 105-1.8 x 105/cmZon fibronectin coated dishes. Epidermal cells or dissected skins were cultured with a 1 :3 mixture of Ham's F12 and Eagle's minimum essential medium (MEM) with nonessential amino acids without calcium chloride (Whittaker Bioproducts, Walkersville, Md., USA). The culture medium was supplemented with 5% Chelex- and charcoal-treated fetal calf serum and 20 pg/ml insulin. The calcium concentration of the medium was adjusted to 0.05 m M unless specified otherwise. The cultures were incubated at 25" C in a humidified chamber gassed with 5% C 0 2 and 95% air. Gel electrophoresis. Cultured epidermal cells, cultured skins, or dissected skins from tadpoles which were grown in hormone containing water were labeled for 3 h with 100 pCi/ml of Tran ["S] label (ICN Radiochemicals, Irvine, Calif., USA). Cytoskeletal fractions of the scraped cells or skins were extracted in low salt buffer and high saIt buffer containing Triton X-100 [15]. We previously identified these cytoskeletal proteins as keratins using biochemical and immunological techniques [3, 241. The cytoskeletal fractions were subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) [ 131 or two-dimensional electrophoresis with nonequilibrium pH gradient electrophoresis (NEPHGE) [21]. The gels were soaked in salicylic acid, dried, and exposed to X-ray film. R N A isolation. Total RNA of cultured epidermal cells was extracted with guanidium thiocyanate followed by centrifugation in cesium chloride solutions [22]. Slot blot analysis was performed as described by Sambrook et al. [22] and the manual for Genescreen Plus (New England Nuclear, Boston, Ma., USA). Denatured total RNA (2 pg) was blotted onto Genescreen Plus using a Minifold I1 apparatus (Schleicher & Schuell, Keene, N.H., USA). The filters were prehybridized for 4 h and then hybridized with [35P]CTPlabeled RNA probes (lo6 cpm/ml) for 18 h at 60" C. Prehybridization and hybridization buffer contained 50% formamide, 5 x standard saline citrate (SSC), 5 x Denhardt's solution, 100 pg/ml herring DNA, 40 m M sodium phosphate buffer and 1% SDS. The filters were washed once in 2 x SSC at room temperature and three times in 0.1 x SSC at 65" C and then exposed to X-ray film. Slot blots were quantitated by scanning the X-ray film with a Hoeffer GS 300 densitometer. All of the genes used in this study were cloned into Bluescribe [15]. pM7 contains the 3' /$turn and untranslated region of the Xenopus 63 Kd keratin cDNA, pUF164 [8]. pM11 contains Xenopus rDNA sequences from pXlrl1 [2].
Results Induction of 63 Kd keratin synthesis by T, or HC
We have previously shown that purified larval epidermal cells respond to T3 and synthesize the adult specific 63 Kd keratin [20]. As shown in Fig. 1, M (lane 2) and lo-* M (lane 3) T3 induced snythesis of the 63 Kd keratin in larval epidermal cells. 10-l' A4 T3 had no effect on 63 Kd keratin synthesis (data not shown). Unexpectedly 63 Kd keratin synthesis was also induced by HC alone. It is known that corticoids themselves do not induce metamorphosis [ I l , 121. 5 x lo-' A4 HC induced high level synthesis of the 63 Kd keratin (Fig. 1, lane 5). However, unlike T,, which induces high level 63 Kd keratin synthesis over two orders of magnitude, only restricted concentrations of HC induced high level expression. We regard the T,-induced 63 Kd protein and HC-induced 63 Kd protein as basic keratins because they react with monoclonal antibody AE3 [25] which recognizes all known basic keratins (data not shown). Since the induction of keratin synthesis by HC was unexpected, we examined the effect of HC on the kinetics of keratin synthesis in cultured epidermal cells, cultured skin and tadpoles. As shown in Fig. 2, HC induced 63 Kd keratin synthesis in cultured epidermal cells on the second day but not at any time in cultured skin or in vivo. These results suggest that the HC effect may be an artifact of the cell culture technique since HC had no effect by itself on skin organ culture or in vivo, where the normal histological organization of the skin is maintained. Furthermore, analysis of mRNA levels (see below) have shown that the unique response of cultured epidermal cells to HC is at the translational level. In contrast to HC, T, induced 63Kd keratin synthesis in both cell and organ culture as well as in vivo (Fig. 2). Synergistic effects of T, and HC on keratin synthesis Keratin synthesis in cultures. The combined effects of T 3 and HC on keratin synthesis was also examined in epidermal cell cultures. Combined T3 and HC induced 63 Kd keratin synthesis more than either T3 alone or HC alone (Fig. 3). Furthermore, 49 Kd keratin synthesis
1
2
3
4
5
6
Fig. 1. Effects of triiodothyronine (T3) and hydrocortisone (HC) on keratin synthesis in cultured epidermal cells. Epidermal cells were cultured with no hormone (lane I ) , lo-" T3 (lane 2), lo-' M T3 (lanes), 5 x lo-' M HC (lune4), 5 x lo-' M HC (lane 9, or 5 x M HC (lane 6 ) for 4 days. After labeling with [35S] amino acids, the cytoskeletal fractions were isolated, subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), and detected by fluorography. a, 63 Kd keratin; b, actin
1
2
3
4
5
6
7
8
9
A
-a
B
-a
-b C
---2
1
3
5
4
2days
3days
6
7
8
9
10
4days
-a
=t 1
2
3
4
Fig. 3. Combined effects of T, and HC on keratin synthesis in cultured epidermal cells. Epidermal cells were cultured with no hormone (lane i), lo-' M T, (lane 2), 5 x M HC (lane 3), or lo-' M T3 and 5 x M HC (lane 4 ) for 4 days. After labeling with [35S]amino acids, the cytoskeletal fractions were isolated, subjected to SDS-PAGE, and detected by fluorography. a, 63 Kd keratin; b, 49 Kd keratin; c, actin
was induced to a high level by T3 and HC (Fig. 3, lane 4). The 49 Kd keratin, as well as the 63 Kd keratin, are adult type keratins in Xenopus and both are initially expressed in the larval epidermis during metamorphosis [3]. The induction of 63 kD keratin synthesis by T, and HC was also reflected by changes in the amount of 63 Kd keratin as measured by Coomassie brilliant blue staining and immunoblotting with AE3 (data not shown). These results suggest that T, and HC have a synergistic action on adult keratin synthesis. To examine the combined effect of T3 and HC on keratin synthesis in more detail, two dimensional electrophoresis was performed. In agreement with the one dimensional electrophoresis data, 63 Kd and 49 Kd keratin synthesis were induced to high levels with T3 and HC (Fig. 4). Two dimensional electrophoretic patterns showed that the 63 Kd keratins induced by T,, HC, or T, plus HC migrated at exactly the same positions. As previously demonstrated by Hoffmann et al. [S], the 63 Kd keratins appeared as three spots. Therefore, our data suggest that T,-induced, HC-induced, and T, plus HC-induced keratins are the same polypeptides. Two dimensional electrophoretic patterns also showed that the 49 Kd keratin is an acidic keratin which corresponds to Xenopus keratin IV [7]. In contrast to induction of
11
12
5days
-
b
Fig. 2A-C. Kinetics of hormone induction of 63 Kd keratin synthesis. Epidermal cells (A), dissected skins (B), or tadpoles (C) were kept in media containing no hormone (lanes i, 4 , 7 , and fO), M T3 (lanes 2, 5, 8, and fl), or 5 x M HC (lanes 3, 6, 9, and 12) for the indicated periods. After labeling with amino acids, the cytoskeletal fractions were isolated, subjected to SDS-PAGE, and detected by fluorography. a, 63 Kd keratin; b, actin
adult type keratin synthesis by T 3 or T3 plus HC, synthesis of larval type keratins were decreased by T, and HC but not by T, alone (compare Fig. 4D with Fig. 4A and B). 63 Kd keratin synthesis in vivo. To determine whether HC acts synergistically with T, in vivo, keratin synthesis was examined in tadpoles which were grown in water containing hormones for 1-5days. The same amount of newly synthesized 63 Kd keratin was detected in both T,- and T, plus HC-treated tadpoles from the second day to fourth day of hormone treatment (Fig. 5, lanes 5, 6, 8, 9, 11, and 12). However, on the fifth day of hormone treatment, tadpoles treated with T, plus HC synthesized more 63 Kd keratin than tadpoles treated with only T,. These results demonstrate that HC works synergistically with T, on adult type keratin synthesis not only in vitro but also in vivo.
Synergistic effects of T, and HC on expression of 63 Kd keratin mRNA To investigate the mechanism of the synergistic effects of T, and HC in more detail, we examined the expression of 63 Kd keratin mRNA in cultured larval epidermal cells. Slot blots of total cell RNA were hybridized with pM7 (Fig. 6A), a cDNA subclone from the 3' region of the 63 Kd keratin gene. The specificity of pM7 for the 63 Kd keratin mRNA was demonstrated by Northern blots, RNase protection and hybrid selection followed by in vitro translation [15, 161. Blots were also hybridized with a probe for rRNA (pM11; Fig. 6B) to demonstrate that equal quantities of RNA were loaded in all slots. As shown in Fig. 6A, 63 Kd keratin mRNA was induced 9-fold by T, (slot 2) and 18-fold by T3 and HC (slot 4) on the fourth day of hormone treatment. However, compared to the low level expression of the 63 Kd keratin gene seen in control cultures, no induction was observed by HC alone (slot 3). These results and
80
Fig. 4A-D. Combined effects of T3 and HC on keratin synthesis in cultured epidermal cells. Epidermal cells were cultured in media containing 0.3 mM calcium with no hormone (A), lo-' M T, (B), 5 x lo-' M HC (C), or lo-' M T, and 5 x lo7 M HC (D) for 4 days. After labeling with amino acids, the cytoskeletal fractions were isolated and subjected to two-dimensional electrophoresis. First-dimension is nonequilibrium pH gradient electrophoresis (NEPHGE); from left (acidic) to right (basic). Second-dimension is SDS-PAGE. Proteins were detected by fluorography. a, 63 Kd keratin; b, 49 Kd keratin; c, actin
-----
1
3
2
4
lday
5
2days
6
7
8
3days
9 1 0
11
4days
Fig. 5. Combined effects of T3 and HC on 63 Kd keratin synthesis in vivo. Tadpoles were grown in water containing no hormone (lanes I , 4 , 7, fa,and 13), lo-' M T3 (lanes 2, 5, 8, If, nrzd I4), or lo-' M T, and 5 x M HC (lanes 3, 6 , Y, 12, and 15) for
A
1
2
3
4
Fig. 6A, 6. Induction of 63 Kd keratin mRNA with T3 and HC in cultured epidermal cells. Epidermal cells were cultured in media containing 0.3 m M calcium with no hormone (slot I), lo-' M T, (slot 2), 5 x lo-' M HC (slot 3), or lo-* M T, and 5 x lo-' M HC (slot 4 ) for 4 days. Total RNA from epidermal cells was hybridized with cRNA probes prepared from pM7 (A), a 3'-end subclone of 63 kD keratin cDNA, or pM11 (B), a subclone of pXlrl1 which detects rRNA
the keratin synthesis data suggest that induction of 63 Kd keratin synthesis by HC alone is restricted to cells in culture and occurs at the translational level. More importantly they also suggest that HC works synergisti-
12131415 5days
the indicated periods. Skins were dissected and then labeled with [35S] amino acids. The cytoskeletal fractions were isolated, subjected to SDS-PAGE, and detected by fluorography. a, 63 Kd keratin; b, actin
cally with T, at the mRNA level and that HC alone does not alter the basal level of 63 Kd keratin mRNA present in control cells. Unlike the induction of gene expression by T, or steroid hormones in adult mammalian cells, induction of 63 Kd keratin mRNA by T3 requires a long lag period [16]. We therefore tested whether HC can decrease this lag period. As shown in Fig. 7, no induction of 63 Kd keratin mRNA was observed on the second day of culture, with or without hormones (slots l , 2, and 3). However, on the third day of culture, 63 Kd keratin mRNA was induced twofold with T, or T3 plus HC (slots 5 and 6). This result indicates that HC does not change the lag period which is required for induction of 63 Kd keratin mRNA by T,. We also examined how long it takes to get synergism between T, and HC. Larval epidermal cells were treated with T, for 4 days. During this time, HC was added to the T,-containing cultures for various periods. As shown in Fig. 8, treatment with HC for 4 days (slot 2), the last 2 days (slot 3), or the last 1 day (slot 4) induced larger amounts of 63 Kd keratin mRNA than induced by T3 alone (slot 6). In comparison to treatment with T, alone (slot 6), addition of HC
81
A
B
1
2 3 4 5 6 u u 2days
3days
Fig. 7A, B. Kinetics of hormone induction of 63 Kd keratin mRNA in cultured epidermal cells. Epidermal cells were cultured in media containing 0.3 mM calcium with no hormone (slots 1 and 4), lo-* M T, (slots 2 and 5), or lo-* M T, and 5 x M HC (slots 3 and 6 ) for the indicated periods. Total RNA from epidermal cells was hybridized with cRNA probes prepared from pM7 (A), a 3'-end subclone of 63 Kd keratin cDNA, or pM11 (B), a subclone of pXlrl1 which detects rRNA
in the following experiments larval epidermal cells were cultured with high calcium media (0.3 mM). Cornified cells were observed on the second day in culture with T, and HC (Fig. 9D) but not until the third day in cultures treated only with T3 (Fig. 9B, F). Further, on the third day of culture, more cornified cells were observed in cultures treated with T, and HC (Fig. 9H) than with T, alone (Fig. 9F). Treatment with T, and HC also induced a change in the pattern of cornification. Cultures which were treated with T, and H C had sheetlike cornified cell layers whereas T, treated cultures had mainly single cornified cells. Cultures without hormones or with HC alone did not show any cornified cells (Fig. 9A, C, E, G). Thus, HC induced 63 Kd keratin synthesis in these cultures but did not induce cornification. When larval epidermal cells were cultured with low calcium media (0.05 mM), cornified cells were not produced even in the presence of T3 and HC (data not shown). Discussion
B
1
2
3
4
5
6
Fig. 8A, B. Kinetics of the synergistic effects of T, and HC on expression of 63 Kd keratin rnRNA. Epidermal cells were cultured in media containing 0.3 mM calcium with no hormone (slot 1) or lo-' M T, (slots 2, 3, 4 , 5, and 6)for 4 days. HC (final concentration 5 x lo-' M ) was added to the T,-containing cultures for 4 days (slot 2), the last 2 days (slot 3), the last 1 day (slot 4), or the last 6 h (slot 5). Total RNA from epidermal cells was hybridized with cRNA probes prepared from pM7 (A), a 3'-end subclone of 63 Kd keratin cDNA, or pM11 (B), a subclone of pXlrl1 which detects rRNA
to the T,-containing culture for the last 6 h (slot 5) didn't show a significant increase of 63 Kd keratin mRNA. Thus the synergistic effect of T3 and HC required a pretreatment with T3 for 3 days after which a 1 day treatment with HC was sufficient for the synergistic effect. During the 3 day pretreatment with T,, HC had no effect on 63 Kd keratin mRNA expression.
Synergistic effects of T3 and HC on epidermal corn fieation Since cornification is a landmark of epidermal differentiation in adult frogs [5] and can be induced by T3 in larval skin explant cultures [16], we examined the combined effects of T, and HC on cornification of larval epidermal cells. Our previous results indicated that a calcium concentration of at least 0.15 m M was required for cornification of adult epidermal cells [24]. Therefore,
It has been known for almost 40 years that glucocorticoids accelerate T,-induced amphibian metamorphosis [6]. Since cell cultures were not previously used to analyze the interactions of T, and glucocorticoids, it was not known whether glucocorticoids act directly on target cells or indirectly through other tissues and organs. In this study, using cultures of purified larval epidermal cells, we have demonstrated that H C acts synergistically with T3 on gene expression during amphibian metamorphosis. Interestingly, this synergistic effect between T3 and HC on 63 Kd keratin gene expression required pretreatment of the cells with T, for 3 days and during this time HC had no effect. Our previous results demonstrated that expression of the 63 Kd keratin gene is regulated at two levels: a low level, basal expression independent of T,; and a T3induced high level expression [ 151. HC synergizes with T3 only after the gene is induced by T,; it does ot have any effect on the basal expression of the keratin gene. While it is clear that the synergistic action of HC and T, is at the mRNA level, further experiments will be necessary to determine whether the transcription rate and/or the stability of the 63 Kd keratin mRNA is changed. Two observations suggest that the synergistic effect cannot be due solely to the stabilization of mRNA. First, HC by itself does not change the basal level of 63 Kd keratin mRNA and thus it is unlikely that HC alters the stability of the T,-induced mRNA. Second, T3 plus H C also have a synergistic effect on the timing and extent of epidermal cell cornification, 49 Kd keratin synthesis and the disappearance of larval keratins. It seems unlikely that the multiple mRNAs required for these aspects of epidermal cell differentiation would all be controlled by mRNA stabilization. Previously, we demonstrated that expression of the 63 Kd keratin gene by T, showed a long latent period and suggested that regulatory factors, which arc induced directly by T,, arc involved in the expression of tissue-
82
Fig.9A-H. Effects of T3 and HC on cornification of cultured epidermal cells. Epidermal cells were cultured with no hormone (A and E), lo-* M T3 (B and F), 5 x M HC (C and G ) , or l o - * M T3 and 5 x M HC (D and H). Photographs were taken on the second (A, B, C, and D) or third (E, F, G, and H) day of culture
83
specific genes [16]. Similar lag times have been reported recently for the T,-induced down regulation of the Xenopus trypsin gene [23] and the induction of N-CAM [14]. In Rana catesbeiuna, inhibition of tail DNA synthesis by T, also requires a long lag period [18, 191. The data presented here demonstrate that HC does not reduce the lag time necessary for T3 to induce high level expression of the keratin gene. However, it is not clear whether HC acts indirectly (through the expression of regulatory factors) or directly on the 63 Kd keratin gene. It is likely that HC acts more directly than T3 on 63 Kd keratin gene expression because it acts more rapidly (< 1 day) than T, (3 days). To clarify this point, direct assay of cis-regulatory elements of the 63 Kd keratin gene will be necessary. While only T, can initiate amphibian metamorphosis, it is clear that other hormones such as glucocorticoids are also essential. As shown here, HC works synergistically with T3 during differentiation of the epidermis. HC not only increases T,-induced cornification and expression of the 49 Kd and 63 Kd keratins but it also promotes normal epidermal cell differentiation in vitro in other ways. During the transformation of the epidermis from larval to adult programs of differentiation, larval-type keratins are replaced by adult-type keratins [ 3 ] .T3 alone did not reduce larval type keratin synthesis, but T3 and HC together induced a drastic reduction in larval keratin synthesis (Fig. 4). Furthermore, T, and HC treated larval epidermal cells produced sheets of cornified cells as seen in vivo, while T3 by itself induced only scattered single cornified cells (Fig. 9). These results indicate that glucocorticoids play an integral part in skin differentiation and other events occurring during metamorphosis. Acknowledgements. We would like to thank Prof. S. Kikuyama, Prof. K. Yoshizato, and Dr. A. Nishikawa for helpful suggestions. We are grateful to Dr. W. Hoffmann for supplying the cDNA clone pUF164 and Dr. J. Doering for supplying the rDNA clone. This study was supported by a National Institutes of Health grant (HD 24438).
References 1. Beato M (1989) Gene regulation by steroid hormones. Cell 56: 335-344 2. Dawid IB, Wellauer PK (1976) A reinvestigation of 5’-3’ polarity in 40 S ribosomal RNA precursor of Xenopus lueuis. Cell 8 :443-448 3. Ellison TR, Mathisen PM, Miller L (1985) Development changes in keratin patterns during epidermal maturation. Dev Biol 112:329-337 4. Evans RM (1988) The steroid and thyroid hormone receptor super family. Science 240: 889-895 5. Fox H (1986) The skin of amphibia. In: Bereiter-Hahn J, Matoltsy AG, Richards KS (eds) Biology of the Integument, Vol. 2. Spring-Verlag, New York, pp 78-108 6. Frieden E, Nail B (1955) Biochemistry of amphibian metamorphosis. I. Enhancement of induced metamorphosis by glucocorticoids. Science 121: 37-39
7. Hoffmann W, Franz JK (1984) Amino acid sequence of the carboxy-terminal part of an acidic type I cytokeratin of molecular weight 51 000 from Xenopus luevis epidermis as predicted from the cDNA sequence. EMBO J 3: 1301-2306 8. Hoffmann W, Franz JK, Franke WW (1985) Amino acid microheterogeneity of basic (type 11) cytokeratins of Xenopus epidermis and evolutionary conservativity of helical and non-helical domains. J Mol Biol 184 :713-724 9. Jolivet Jaudet G, Leloup Hatey J (1984) Variations in aldosterone and corticosterone plasma levels during metamorphosis in Xenopus laevis tadpoles. Gen Comp Endocrinol 56: 59-65 10. Kikuyama S, Niki K, Mayumi M, Kawamura K (1982) Retardation of thyroxine-induced metamorphosis by Amphenone B in toad tadpoles. Endoclinol Jpn 29: 659-662 11. Kikuyama S, Niki K, Mayumi M, Shibayama R, Nishikawa M, Shintake N (1983) Studies on corticoid action on the toad tadpole tail in vitro. Gen Comp Endocrinol 5 2 : 395-399 12. Kobayashi H (1958) Effect of desoxycorticosterone acetate on metamorphosis induced by thyroxine in anuran tadpoles. Endocrinology 62: 371-377 13. Laemmli U K (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227: 680-685 14. Levi G, Broders F, Dunon D, Edelman GM, Thiery JP (1990) Thyroxine-dependent modulations of the expression of the neural cell adhesion molecule N-CAM during Xenopus laeuis metamorphosis. Development 109 :681-692 15. Mathisen PM, Miller L (1987) Thyroid hormone induction of keratin genes: a two-step activation of gene expression during development. Genes Dev I :1101-1 117 16. Mathisen PM, Miller L (1989) Thyroid hormone induces constitutive keratin gene expression during Xenopus laevis development. Mol Cell Biol 9:1823-1831 17. Nieukoop PD, Faber J (1967) In: Nieukoop PD, Faber J (eds) Normal table of Xenopus luevis (Daudin). North Holland Publishing 18. Nishikawa A, Yoshizato K (1986) Hormonal regulation of growth and life span of bullfrog tadpole tail epidermal cells cultured in vitro. J Exp Zoo1 237:221-230 29. Nishikawa A, Kaiho M, Yoshizato K (1989) Cell death in the anuran tadpole tail: Thyroid hormone induces keratinization and tail-specific growth inhibition of epidermal cells. Dev Biol 1311337-344 20. Nishikawa A, Shimizu-Nishikawa K, Miller L (1990) Isolation, characterization, and in vitro culture of larval and adult epidermal cells of the frog Xenopus luevis. In Vitro 26: 1 128-1 134 21. O’Farrell PZ, Goodman HM, O’Farrell PH (1977) High resolution two-dimensional electrophoresis of basic as well as acidic proteins. Cell 12: 1133-1342 22. Sambrook J, Fritsch EF, Maniatis T (1989) In: Sambrook J, Fritsch EF, Maniatis T (eds) Molecular Cloning; A Laboratory Manual, Cold Spring Harbor Laboratory Press, Cold Spring Harbor 23. Shi Y-B, Brown D D (1990) Developmental and thyroid hormone-dependant regulation of pancreatic genes in Xenopus h e vis. Genes Dev 4:1107-1113 24. Shimizu-Nishikawa K, Miller L (1991) Calcium regulation of epidermal cell differentiation in the frog Xenopus lueuis. J Exp Zoo1 (in press) 25. Sun T-T, Eichner R, Schermer A, Cooper D, Nelsen WG, Weiss RA (1984) Classification, expression, and possible mechanisms of evolution of mammalian epithelial keratins: A unifying model. In: Levine A, Topp W, Vande Wounde G, Watson JD (eds) Cancer Cells; Transformed Phenotype, vol. I. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, pp 169-176 26. Yoshizato K (1989) Biochemistry and cell biology of amphibian metamorphosis with a special emphasis on the mechanism of removal of larval organs. Tnt Rev Cytol 119:97-1 49
