SEARCH

SEARCH BY CITATION

Keywords:

  • amphibian metamorphosis;
  • differentiation;
  • epidermal basal cells;
  • epithelial–mesenchymal interactions;
  • fibroblasts

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Anuran larval skin undergoes a process of metamorphosis into preadult and adult skin. Basal skein, larval basal and adult basal cells are basement membrane-attaching cells in the larval, preadult and adult epidermis, respectively, and are identified as cells expressing genes of RLK (Rana larval keratin), both RLK and RAK (Rana adult keratin), and RAK. Larval to preadult skin conversion takes place in the histological entity called the skin transformation center (STC). The present study performed a cDNA subtractive gene screening on cDNA of the larval and the preadult skin, and cloned the secreted protein acidic and rich in cysteine (SPARC) gene as an upregulated gene in the larva to preadult skin conversion. RAK gene-positive basal skein cells and fibroblasts in and around the STC were weakly and strongly sparc-positive, respectively. Using sparc and rak, we redefined the STC and visualized it on a histological section as an approximately 150 µm-long region that contained about 20 rak-negative and weakly sparc-positive basal cells. Intense sparc expression was observed in basal skein cells, but not in larval basal cells, suggesting that SPARC acts as a suppressor of rak during epidermal differentiation. This suggestion was tested by investigating the effect of SPARC on cultured larval basal cells. We observed that SPARC suppressed the expression of rak, but not rlk.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

The anuran tadpole drastically remodels its tissues during metamorphosis from larval to adult type to adapt to a transition from an aquatic to terrestrial living environment (Gilbert & Frieden 1981; Yoshizato 1986, 1989, 1992, 1996; Shi 1999). The anuran adult skin is basically identical to the mammalian adult counterpart in its histology (Fox & Whitear 1986; Robinson & Heintzelman 1987; Yoshizato 1992; Izutsu et al. 1993). The basal cells on the basement membrane are mitotically active. Their descendants upwardly migrate and differentiate into spinous, granular and finally cornified cells, demonstrating that adult basal cells contain a population of epidermal stem cells (Yoshizato 1989, 1996). Rana adult basal cells express a unique gene, rak, the Rana adult keratin (RAK) gene (Suzuki et al. 2001). The larval epidermis is histologically quite different from the adult one and composed of three types of cells: apical, basal skein and suprabasal skein cells (Robinson & Heintzelman 1987; Izutsu et al. 1993; Yoshizato 1996; Utoh et al. 2000; Suzuki et al. 2001; Watanabe et al. 2001, 2002). Apical cells are located in the outermost layer and express rk8, the Rana keratin 8 gene (Suzuki et al. 2001). Skein cells uniquely contain Eberth's figure, huge cytoplasmic keratin bundles (Eberth 1866; Fox & Whitear 1986; Fox 1992). These cells are present on the basement membrane (basal skein cells) or suprabasally (suprabasal skein cells) depending on the metamorphic stage of the tadpole (Fox & Whitear 1986). Rana larval keratin (RLK) is a component of the keratin bundles and can therefore serve as a useful biochemical indicator of both types of skein cells (Suzuki et al. 2001). Its gene (rlk) was cloned and characterized.

Larval basal cells first emerge around stage III of Taylor and Kollros (TK) staging (Taylor & Kollros 1946; Tamakoshi et al. 1998; Utoh et al. 2000; Suzuki et al. 2001, 2002). Several cell type-specific marker genes and proteins have revealed that basal skein cells are precursor cells of larval basal cells that differentiate into adult basal cells (Suzuki et al. 2002). Basal skein cells are positive for genes of collagen α1 (I) (colα1+), rlk+, and negative for rak (rak). Larval basal cells are colα1, rlk+ and rak+ and adult basal cells are defined as colα1, rlk and rak+ cells. We also demonstrated the presence of rak+-basal skein cells, transit cells between basal skein and larval basal cells (Suzuki et al. 2002). The skin composed of basal skein cells and larval basal cells were designated as the larval and the preadult skin, respectively (Kawai et al. 1994; Tamakoshi et al. 1998). Tamakoshi et al. (1998) demonstrated that the larval to preadult skin conversion takes place in the skin transformation center (STC). The STC in Rana tadpole skin first appears at a dorsal lateral site and migrates over the body skin, except for the frontal mouth and tail region, transforming the larval into the preadult skin. The tail skin does not yield larval basal cells and later undergoes apoptosis, as shown by Nishikawa et al. (1989) and Yoshizato (1989 and 1992). These studies indicate that the differentiation potential of basal skein cells differs depending on region and determines the different metamorphic fate between body and tail skin.

Although the STC has been characterized histologically in detail (Tamakoshi et al. 1998; Utoh et al. 2000), no studies have identified the STC using molecular probes, visualized or located it on a histological section since its concept was first proposed (Tamakoshi et al. 1998). The molecular mechanism determining the differentiation potential of basal skein cells has also been poorly understood. As a first step in uncovering the genes that participate in larval to preadult skin conversion, we performed a cDNA subtractive gene screening (Amano 1998) using the region containing the preadult tadpole back skin as a tester, and the back skin that did not contain the preadult region as a driver. As a result, 123 cDNA clones were screened as upregulated in the tester skin. The present study characterized the secreted protein acidic and rich in cysteine (SPARC) gene, designated sparc, as being among them. The results indicated that SPARC expression is developmentally regulated in a unique pattern during skin metamorphosis and inhibits the expression of rak, but not rlk, in the cultured larval basal cells. Thus, SPARC is proposed to function as a suppressor of adult-keratin gene expression during anuran premetamorphic skin development. Utilizing sparc and rak as probes, we redefined the STC and were able to visualize and locate it on a histological section of larval skin.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Animals

Tadpoles and frogs of Rana catesbeiana were purchased from a local supplier (Hamamatsu Seibutsu Kyouzai, Shizuoka, Japan) and staged according to Taylor & Kollros (1946). The animals were anesthetized during operation by decreasing ambient temperature to below 4°C with ice. When necessary, tadpoles at stage X were intraperitoneally injected through their tail muscles with 3 × 10−10m of 3,5,3′-triiodothyronine (T3; Sigma, St Louis, MO, USA) per bodyweight (Kistler et al. 1975; Oofusa & Yoshizato 1991).

cDNA subtractive gene screening

Whole body skins of tadpoles between stages II and IV were collected from about 100 tadpoles and separated into two groups under a stereoscopic microscope (the tester and the driver). The tester sample consisted of back skin tissues that contained the preadult and the larval tissues surrounding the preadult region. The driver sample consisted of back skin tissues that did not contain the preadult region. Because the tester sample deposits hydroxyapatites in the dermis, it shows a whitish tint and therefore the preadult skin was easily and unequivocally distinguishable from the larval (Tamakoshi et al. 1998). Total RNA was isolated from skin samples with ISOGEN (Nippon Gene, Tokyo, Japan) and used for purifying mRNA with a polyA tract mRNA isolation system (Promega, Madison, WI, USA). 5′-suppression subtractive hybridization (Amano 1998) was performed by using a PCR-Select cDNA Subtraction Kit (BD Bioscience Clontech, Palo Alto, CA, USA). The amplified cDNA fragments were directly ligated into a pGEM T-easy vector system (Promega).

Quantification of mRNA

Total RNA was isolated from tadpole skins with ISOGEN and from cultured larval basal cells with TRIzol (Invitrogen, Carlsbad, CA, USA). 1 µg was converted to first strand cDNA with random hexamers using a Thermoscript RT system (Invitrogen) following the supplier's protocol. The primers of 5′-RACE and SP1 (Table 1) were used for the amplification of sparc. Polymerase chain reaction (PCR) for sparc was carried out as follows: at 94°C for 30 s, 63°C for 30 s, 72°C for 30 s. PCR for rlk, rak, and rpL8 was performed as reported by Sachs & Shi (2000) and Suzuki et al. (2002). PCR amplification was performed within a range in which a gene would be linearly amplified. The PCR products of rpL8, rlk, and rak were electrophoresed and visualized by staining with ethidium bromide. The brightness of each band was quantified with Kodak Digital Science 1D Image Analysis Software (Eastman Kodak, Rochester, NY, USA).

Table 1.  Primer Information
Name of primersSequence
5′ RACE5′-CTGACTAGGACATCCTCCAGCCAGTCG-3′
5′ RACE nested5′-TGTTCTCATCACGCTCATACAGGCTG-3′
3′ RACE5′-TCATGCCACTTCTTTGCCGCCAA-3′
3′ RACE nested5′-GTTCTGTGGTAATGACAGCGAAACCTA-3′
SP15′-TGCCGCCCAACCCCGAGAGA-3′
SP25′-GTCAGGCAGGTGGCGGGATCTG-3′

Cloning of bullfrog SPARC cDNA and in situ hybridization of its mRNA

Four oligonucleotide primers (5′-RACE, 3′-RACE, 5′-RACE nested, and 3′-RACE nested) were designed (Table 1) with the nucleotide sequence of a subtracted clone, designated B11. 5′-rapid amplification of cDNA ends (RACE) and 3′-RACE were performed with a Marathon cDNA amplification kit (BD Bioscience Clontech) using these primers and the tester cDNA as a template for PCR. PCR was carried out by using 5′-RACE, 3′-RACE, and adapter primer 1 in a Gene-amp PCR system 9600 (Perkin-Elmer, Foster City, CA, USA) with 1 cycle at 94°C for 30 s, 25 cycles at 94°C for 5 s and at 68°C for 3 min. An aliquot (5 µL) of the first PCR product was diluted in 245 µL of Tricine-ethylenediaminetetraacetic acid (EDTA) buffer, and used as a template for nested PCR. Nested PCR was carried out by using 5′-RACE nested, 3′-RACE nested, and adapter primer 2 in PCR system 9600 (Perkin-Elmer) with one cycle at 94°C for 30 s, 27 cycles at 94°C for 5 s and at 68°C for 3 min The nested PCR products were subcloned into pGEM T-easy Vector (Promega). Two additional primers, SP1 and SP2 (Table 1), were used to determine the nucleotide sequence of the 3′-RACE product. In situ hybridization was performed as reported by Ishii et al. (1997) and Asahina et al. (1999) using a cRNA probe synthesized from the 5′-RACE fragment of SPARC cDNA as a template. Sections were stained with hematoxylin and eosin (H&E) for histological examination.

Immunohistochemistry

Tissues were fixed overnight with NEOFIX (Merck, Darmstadt, Germany) at 4°C, dehydrated with a graded series of ethanol and xylene, and embedded in Tissue Prep 2 (Fisher Scientific, Fair Lawn, NJ, USA), sectioned at 5 µm. The tissues were rehydrated in PBS containing 0.1% Tween 20 (PBST), and boiled in Target Retrieval Solution (Dako Cytomation, Carpinteria, CA, USA) for 10 min in a microwave oven. The following procedures were done at room temperature. The sections were blocked with 10% normal goat sera for 30 min, incubated for 1 h with 1:1000 diluted antisera and mouse SPARC (LSL, Tokyo, Japan), washed 3 times with PBST for 5 min and incubated for 30 min with antirabbit IgG Alexa 594 or 488 (Molecular Probes, Eugene, OR, USA) that had been diluted 1000-fold with PBST. Western blotting showed that the mouse antisera used in this study had a specific cross-reactivity to bullfrog SPARC (data not shown).

Isolation and cultivation of larval basal cells

Larval basal cells were isolated from tadpoles at stage X and cultured in 35 mm diameter tissue culture dishes (BD, Franklin Lakes, NJ, USA; 106 cells per dish) or chamber-slides (Nalge Nunc, Rochester, NY, USA; 9 × 104 cells per slide) at 24°C in a humidified atmosphere of 5% CO2 and 95% air (Nishikawa & Yoshizato 1985, 1986). The dishes and slides had been precoated with human plasma fibronectin (Biogenesis, Poole, UK) by immersion in 1 mL of 70% of RPMI-1640 containing 22 µg/mL fibronectin for 45 min at room temperature. Cells in dishes and on slides were given 2 mL and 300 µL of culture media, respectively. The culture medium used consisted of 70% RPMI-1640 and 20 mmN-2-hydroxyeythlpiperazine-N′-2-ethanesulfonic acid (pH 7.4), 10 mm NaHCO3, 10% charcoal-treated fetal calf serum (Yoshizato et al. 1980), growth factors (insulin 20 µg/mL, EGF 10 ng/mL, transferrin 10 µg/mL), 100 IU/mL penicillin, and 100 µg/mL streptomycin. The effects of SPARC were examined by adding bovine SPARC to the cultures (osteonectin; Haematologic Technologies, Essex, VT, USA), 8 µg/mL for 24 h.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Subtractive screening and cloning of SPARC

Regions of preadult skin were excised from the back of tadpoles between stages II and IV. Most of these regions were of preadult skin; their peripheries, however, were of larval skin and were similarly excised. cDNA subtractive gene screening was performed using preadult skin as the tester, and larval skin as the driver samples. We randomly selected 123 clones (cDNA fragments, their average length being 213 bp) and determined their sequences. These clones were selected as possible coding flames and were subjected to the European Molecular Biology Laboratory's (EMBL) sequence identity test database. Eighty-six clones (70% of the total clones) were identified as known genes (data not shown). Among the known genes, 2 clones, 259 bp-long B-11 and 580 bp-long C-29, showed a high level of homology to the sparc gene of several species (bovine, mouse, human, chick and frog). 5′- and 3′-RACE on B-11 and C-29 amplified 622 bp and 1461 bp fragments, respectively, which covered the nucleotide sequence (1822 bp) containing start codon (ATG), stop codon (TAA), two poly(A) addition signals (AATAAA), and poly(A) tail. These sequence data were submitted to the DDBJ database under accession number AB116365. The deduced amino acid sequence showed a high similarity (>75%) to SPARC of several vertebrates (Fig. 1). The bullfrog SPARC contained all known sequences characteristic to it with the 15 cysteine residues present at the predicted sites and the two EF-hand motifs with calcium binding ability (Bassuk et al. 1993). In the present study, we focused on the SPARC gene for further analysis.

image

Figure 1. Deduced amino acid sequence of bullfrog secreted protein acidic and rich in cysteine (SPARC). Two putative EF-hand signatures (underlined region) are located at residues 183–193 and 276–287. Fifteen cysteine residues (*) are perfectly conserved as in other animal species. Bold numerals at the right end of each line indicate number of amino acid residues from the amino-terminal methionine. SS, signal sequence; AD, acidic domain; FD, follistatin-like domain; ED, extracellular Ca2+-binding domain.

Download figure to PowerPoint

Expression of sparc in relation to larva to preadult skin transformation

Skin tissues were removed from a central region of the back skin of tadpoles at stages II, IV, VI, XII, XVII and XXII. The preparation of the tissues from stages II and IV was identical to the tester and driver preparation, respectively. Similarly, skin tissues were isolated from matured adults. Total RNA was extracted from skin samples and subjected to reverse transcriptase (RT)–PCR (Fig. 2). SPARC mRNA levels at stage IV were much higher than at stage II, thus confirming the result of cDNA subtractive gene screening. Its expression markedly decreased at stage VI, the stage when the larval skin completed its transformation to preadult skin (Tamakoshi et al. 1998), then slightly increased again at stage XVII (the prometamorphic stage), and became higher at stage XXII (the metamorphic climax stage) and in the adult skin. The expression profile shown in Figure 2 suggested that SPARC is involved in both larva to preadult and preadult to adult skin transformation.

image

Figure 2. Expression of sparc gene mRNA during spontaneous metamorphosis. A central region of the back skin was isolated from tadpoles at the indicated stages and total RNA was isolated from the tissue. The skin samples at stages II and IV, and those at VI, XII, XVII and XXII were of larval skin, chimeric skin composed of both larval and preadult, and preadult skin, respectively. RNA (1 µg) was used as a template for reverse transcriptase-polymerase chain reaction (RT–PCR) of sparc and rpL8. The rpL8 gene was used as an internal control (Sachs & Shi 2000).

Download figure to PowerPoint

In situ hybridization of mRNA was performed on tadpole larval skin sections at stages II and preadult stage IV (Fig. 3). In the larval skin, the thin collagen lamella was directly attached to the basement membrane, and fibroblasts were sparsely scattered in primary connective tissues (p-ct) beneath the collagen lamella (Fig. 3A). Fibroblasts (Fig. 3C, arrows) and the basal skein cells (Fig. 3C, arrowheads) intensely expressed SPARC mRNA. In the preadult skin, fibroblasts often migrated into the collagen lamella that consequently became much thicker (Fig. 3B). The secondary connective tissue (s-ct) was developed between the thickened collagen lamella and the basement membrane. Basal skein cells were replaced with larval basal cells. These fibroblasts (Fig. 3D, arrows) and the larval basal cells (Fig. 3D, arrowheads) weakly expressed sparc mRNA. The sparc mRNA distributions from mRNA shown in Figure 3(C), gave out intense immunosignals of SPARC in larval skin (Fig. 4C) but not in preadult skin (Fig. 4D), as expected. Basal skein cells were heavily stained in their peripheral regions. However, fibroblasts were hardly stained in spite of intense in situ hybridization signals (Fig. 4C).

image

Figure 3. Distribution of sparc mRNA in the larval and preadult skin. Sections were prepared from the back skin at stages II (larval skin; A,C) and IV (preadult skin; B,D). The sections were subjected to hematoxylin and eosin (H&E)-staining (A,B), and in situ hybridization of sparc mRNA (C,D). (A) The larval epidermis was placed on the thin collagen lamella (cl) under which primary connective tissues (p-ct) were present. Basal skein cells (bs; arrowheads) were attached to the basement membrane. Fibroblasts (fb; arrows) were present underneath the collagen lamella but did not invade it. (B) Basal skein cells were replaced with larval basal cells (lb; arrowheads). Secondary connective tissues (s-ct) developed between the epidermis and the thickened collagen lamella. Fibroblasts (arrows) were observed in s-ct, and both in and on the collagen lamella. (C) Basal skein cells (arrowheads) and fibroblasts underneath the collagen lamella (arrows) expressed sparc mRNA intensely in the larval skin. (D) Both larval basal cells (arrowheads) and fibroblasts both in and on the collagen lamella (arrows) expressed sparc mRNA although their levels were much lower than basal skein cells and fibroblasts in the larval skin. Bars, 50 µm.

Download figure to PowerPoint

image

Figure 4. Distribution of SPARC in larval and preadult skin. Tissue sections were prepared from the back skin at stage III (larval skin; A,C) and at stage IV (preadult skin; B,D). The sections were subjected to immunohistochemistry with anti-SPARC antisera. Sections were viewed through a phase contrast microscope (A,B) and a fluorescence microscope (C,D). Intense signals were seen on basal skein cells (bs; arrowheads) in the larval skin (C). No immune reactivity was seen on the larval basal cells (lb; arrows) in the preadult skin (D). cl, collagen lamella. Bar, 50 µm.

Download figure to PowerPoint

Suzuki et al. (2002) investigated expression patterns of several epidermal marker genes during early metamorphic stages and demonstrated that basal skein cells differentiated into rak+-basal skein cells in the STC. The rak+-basal skein cells then differentiated into larval basal cells in the preadult skin. To compare the expression pattern of rak with that of sparc in the STC, STC-containing sections were prepared from stage IV back skin and were subjected to in situ hybridization using rak and sparc probes (Fig. 5). H&E-staining on STC-containing sections showed that the collagen lamella was invaded by fibroblasts and that s-ct had not developed between the basement membrane and the collagen lamella (Fig. 5A). The collagen lamella becomes thinner toward the right side of the view, indicating that the larval to preadult skin transformation is in progression in the right side of Figure 5. rak+-basal skein cells were observed in the left side of Figure 5(B). The rak+-intensity decreased in the direction of the larval region and became null at a point approximately one-third from the left side (Fig. 5B). In the present study, we define the preadult region as the skin in which basal cells are rak+. Conversely, epidermal cells in the right hand side of Figure 5(C) intensely expressed sparc, highest around the center to an approximately 20-cell wide region and decreasing toward the larval region. Epidermal cells in the left hand side of Figure 5(C) expressed the gene at an extremely low level. This sparc expression profile strongly suggests that basal skein cells in the larval region are ‘activated’ in the region where the STC is approaching. The gene expression patterns of rak and sparc suggest that the downregulation of sparc correlates with the differentiation between the basal skein and the rak+-larval basal skein cells in the STC. Histology and expression profiles of rak and sparc led us to propose that the STC be redefined as a region where epidermal basal cells express sparc mRNA at a low level and are negative for rak. Assuming the STC to be a circle, its diameter was estimated to be approximately 20-cells wide, 150 µm from the region shown in Figure 5. Basal skein cells in the larval skin region become activated in the expression of sparc as the STC approaches their location. When they are incorporated into the STC, sparc expression decreases to a low level. The STC subsequently moves further towards the larval region and leaves these cells behind, causing them to turn on rak.

image

Figure 5. Comparison of the expression pattern of rak and sparc in STC-containing tissues. Tissue sections of STC-containing skin regions were prepared from stage IV tadpole back skin, subjected to H&E-staining (A), and in situ hybridization using antisense probes of rak (B) and sparc (C). (A) Fibroblasts (oblique arrows) often invaded the collagen lamella (cl), but s-ct had not developed between the epidermis and the collagen lamella. The far right oblique arrow points to the last fibroblast present in the collagen lamella of the larval region. The STC is forward and to the right. (B) rak+-basal skein cells (arrowheads) appeared in the left side (preadult region), but not in the right side (larval region). We define the preadult region as the skin in which basal cells are rak+. (C) Intense signals of sparc mRNA in the epidermis (arrowheads) was observed in the right side but not in the left side. Fibroblasts (oblique arrows) were sparc-positive. Closed and open thick arrows represent the boundary between the rak+ and rak- regions, and the boundary between sparc++ and sparc+ regions, respectively. The larval region is defined as the skin in which basal cells are sparc++. Bar, 50 µm.

Download figure to PowerPoint

Expression of sparc in relation to preadult to adult skin transformation

The result shown in Figure 2 suggested involvement of SPARC in the transformation of preadult to adult skin. We examined the effect of thyroid hormone on sparc expression (Fig. 6). Stage X tadpoles were treated with T3 for up to 6 days. T3 upregulated sparc mRNA at 3 days and its activated state continued thereafter (Fig. 6A). H&E-staining on skin sections of these animals showed that the histology of the skin treated with T3 for 6 days was identical to that of the adult skin except for the absence of cornified layers (Fig. 6D). Both epidermal and mesenchymal cells were negative in hybridization signals in control tadpoles (Fig. 6C). Intense signals of SPARC mRNA were detected in adult basal cells and fibroblasts (Fig. 6E). Immunosignals of anti-SPARC antisera were observed in suprabasal cell layers of the skin treated with T3 for 6 days (Fig. 6F).

image

Figure 6. Effects of T3 on sparc mRNA expression in the preadult skin. Tadpoles at stage X were treated with T3 for up to 6 days. (A) Total RNA was extracted from the back skin of T3-treated tadpoles at days indicated and used as a template for RT–PCR of sparc and rpL8. (B–F) tissue sections were prepared from back skin of control (B,C) and experimental (treated with T3 for 6 days) tadpoles (D,E,F). The sections were subjected to H& E-staining (B,D), in situ hybridization of sparc mRNA (C,E), and immunohistochemistry of SPARC (F). T3 induced the skin conversion from preadult (B) to adult skin (D). There was no signal of sparc mRNA in the preadult skin (C). sparc mRNA was observed in adult basal cells (ab; arrowheads) and fibroblasts (fb; arrows, E). SPARC immunosignals were observed in suprabasal cell layers (F). ap, apical cell; sb, suprabasal skein cell; lb, larval basal cell; gr, granular cell; sp, spinous cell. A dashed line in (F) represents the location of basement membrane. Bar, 50 µm.

Download figure to PowerPoint

We also performed sparc in situ hybridization on mature adult skin (Fig. 7). SPARC mRNA was expressed in adult basal cells and fibroblasts, but not in suprabasal cells (Fig. 7B). Interestingly, immunohistochemistry revealed that SPARC proteins were present in suprabasal but not in basal cells (Fig. 7D).

image

Figure 7. SPARC gene expression in the adult skin. The sections were prepared from the back skin of matured adults. They were subjected to H&E-staining (A), in situ hybridization with antisense (B) and sense sparc cRNA probes (C), and immunostaining using anti-SPARC antisera (D). sparc mRNA were observed in adult basal cells (ab; arrowheads) and fibroblasts (fb; arrows, B). SPARC proteins were present in suprabasal cell layers except for cornified cell layers. No immuno-signals were detected in the basal cell layer (D). gr, granular cell; sp, spinous cell; co, cornified cell. A dashed line in (D) represents the location of basement membrane. Bar, 50 µm.

Download figure to PowerPoint

SPARC suppresses the expression of rak, but not rlk, in the cultured larval basal cells

We tested the possibility that SPARC suppresses rak expression in larval epidermal cells. Effects of SPARC on the expression of rlk and rak were examined using cultured larval basal cells. Sage et al. (1989) reported that SPARC showed antispread activity on the cultured bovine aortic endothelial cells in a dose-dependent manner between 2 and 42 µg/mL. In the present study, SPARC showed antispread activity on Rana epidermal cells at 8 µg/mL, but not at 4 µg/mL (data not shown). Thus, we tested the effect of SPARC on the rak expression at 8 µg/mL. The larval basal cells were isolated from the back skin of tadpoles at stage X, cultured for 24 h with or without 8 µg/mL SPARC, and were then subjected to extraction of total RNA for determining mRNA of rak and rlk by RT–PCR (Fig. 8). SPARC significantly suppressed the expression of rak, whereas there was no effect on the expression of rlk.

image

Figure 8. Suppression of rak expression in the larval basal cells by SPARC. Larval basal cells were isolated from the back skin of tadpoles at stage X and cultured for 24 h in the presence (+) or absence (–) of 8 µg/mL SPARC. Total RNA was extracted and used as templates of RT–PCR for quantifying mRNA of rak, rlk, and rpL8. SPARC suppressed the expression of rak, but not rlk, in larval basal cells. The results are shown as the average with its standard error (n = 3). RT–PCR was performed twice and similar results were obtained. (*) statistically significant level against SPARC (–) at P < 0.01 obtained by Student's t-test.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Larval basal cells are defined as colα1, rlk+ and rak+, and adult basal cells are defined as colα1, rlk– andrak+. We reported that basal skein cells (colα1+, rlk+, rak) differentiate into first rak+-basal skein cells (colα1+, rlk+, rak+), and then larval basal cells (colα1, rlk+, rak+) during the early prometamorphic stage (Suzuki et al. 2002). The larval basal cells finally differentiate into adult basal cells (colα1, rlk, rak+) at the metamorphic climax stage. These serial epidermal transformations take place in the body region, but not in the tail region (Tamakoshi et al. 1998).

Previously, we reported the presence of STC in the bullfrog tadpole where the larval skin is transformed into the preadult skin (Tamakoshi et al. 1998). Fibroblasts in the STC migrate into the collagen lamella and develop s-ct between the collagen lamella and the basement membrane, whereas basal skein cells in the STC differentiate into rak+-basal skein cells, and then larval basal cells (Tamakoshi et al. 1998; Utoh et al. 2000; Suzuki et al. 2002). However, the STC has been hypothetical and its entity has not been characterized at a molecular level. In the present study, we examined the expression patterns of rak and sparc in the STC-containing region. The results indicated that the region is actually three separate regions: (i) sparc++ and rak; (ii) sparc+ and rak ; and (iii) sparc+ and rak+. In (i), sparc is expressed in basal skein cells, and fibroblasts start to migrate into collagen lamella. Thus, (i) is a larval edge of the STC-containing region. In (ii), basal skein cells greatly decrease sparc mRNA expression, while epidermal basal cells are not committed to express rak yet. These basal skein cells are supposed to be on the way to differentiating into rak+-basal skein cells. Thus, (ii) is a central region of the STC-containing region. In (iii), the basal skein cells have differentiated into rak+-basal skein cells. Therefore, (iii) is an adult edge of the STC-containing region. On the basis of these results, we propose to define STC as the sparc+ and rak--region.

The present study screened sparc as a gene that is upregulated in the STC-containing region in comparison to its expression in the larval region. The SPARC gene showed a unique expression pattern during skin metamorphosis. SPARC is also known as osteonectin (Termine et al. 1981) or BM-40 (Dziadek et al. 1986), and has been reported as a secreted glycoprotein (Yan & Sage 1999; Bradshaw & Sage 2001; Brekken & Sage 2001). Its expression is associated with a variety of biological phenomena such as development, cell turnover, wound healing and regeneration (Yan & Sage 1999; Bradshaw & Sage 2001; Brekken & Sage 2001). cRNA-interference or antibody-blocking of SPARC caused abnormal developments of Caenorhabditis elegans and Xenopus laevis (Purcell et al. 1993; Fitzgerald & Schwarzbauer 1998). SPARC expression was activated during the in vitro differentiation of mouse myoblasts, and anti-SPARC antibodies almost completely prevented their differentiation (Cho et al. 2000).

In this study, we showed that sparc mRNA expression is modulated during the conversion in bullfrogs of both larval to preadult skin, and of preadult to adult skin. During the former process, the expression was downregulated. Conversely, during the latter process, the expression was upregulated. The SPARC gene was constitutively expressed after skin metamorphosis. SPARC mRNA was expressed in adult basal cells and SPARC protein was detected in the intercellular spaces of the suprabasal cell layers. A similar distribution pattern of sparc mRNA and its protein has been reported in human adult skin (Hunzelmann et al. 1998). Sodium n-butyrate-induced terminal differentiation of cultured human keratinocytes activated sparc mRNA expression (Ford et al. 1993). The T3-induced conversion of the preadult skin to its adult counterpart also caused sparc activation. These results indicate that sparc activation is generally important for the formation of adult skin and its maintenance.

Expression of RAK mRNA is terminated in suprabasal cell layers (Suzuki et al. 2001) where SPARC protein is deposited (the present study). We also showed that SPARC suppresses rak expression in cultured larval basal cells. Thus, we conclude that SPARC acts as a suppressor for rak expression in epidermal cells. The suppressive effect of SPARC on keratin gene expression in the cultured larval basal cells was observed for rak but not for rlk. Suzuki et al. (2002) demonstrated that preadult skin is capable of transforming into a histologically true adult skin in vitro in the presence of T3 and aldosterone. When preadult skin was cultured in T3-deficient media, the skin was transformed into a ‘pseudo’ adult skin whose suprabasal epidermal cells expressed rlk. In the present study, T3 was not included in media. Thus, the cells did not terminate rlk expression.

Epithelial–mesenchymal interaction plays important roles in larval to preadult skin conversion (Kawai et al. 1994). Recently, Utoh et al. (2003) demonstrated that platelet-derived growth factor A (PDGF-A) is one of the factors involved in the interaction in X. laevis skin metamorphosis. Both AG1296 (a potent inhibitor for PDGF signaling) and an excessive extracellular domain of PDGF receptor-α, inhibited the migration of fibroblasts into the collagen lamella during T3-induced in vitro skin conversion. Fibroblast migration into the collagen lamella is prerequisite in the development of s-ct, a mesenchyme unique to preadult and adult skin. In mammals, PDGF signaling plays important roles in epidermal differentiation and proliferation (Schatteman et al. 1992; Skobe & Fusenig 1998). Conversely, it has been reported that SPARC binds to PDGF-AB and BB, and inhibits the binding of these dimers to PDGF receptors (Brekken & Sage 2001). These results suggest that SPARC functions in the development of larval skin. SPARC might act as a suppressor for the PDGF-dependent migration of fibroblasts into the collagen lamella. Tamakoshi et al. (1998) reported that STC migrates to larval regions of body skin and converts them to preadult skin. SPARC might regulate the STC migration by suppressing PDGF signaling.

Previously we reported the expression profile of colα1 during larval to preadult skin conversion (Asahina et al. 1997; Utoh et al. 2000). Basal and rak+-basal skein cells are colα1+. Larval and adult basal cells are colα1. COLα1 gene is upregulated in fibroblasts that are actively participating in the formation of s-ct. The expression profile of sparc is similar to that of colα1, except for the rak+-basal skein cells that do not express sparc. Thus, it is likely that the expression of sparc and colα1 is controlled similarly during the development of larval skin. Transforming growth factor (TGF)-β1 is a potent stimulator of the synthesis of extracellular matrices including type I collagen (Reed et al. 1994) and is also known as an activator of sparc gene expression (Wrana et al. 1991; Brekken & Sage 2001). The expression of TGF-β1 gene is also modulated by SPARC (Brekken & Sage 2001). Therefore, TGF-β1 may function in larval to preadult skin conversion through interactive regulations with these proteins.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

This study was partly supported by a Grant-in-Aid for Scientific Research (A, no. 07404057) from the Ministry of Education, Culture, Sports, Science and Technology of Japan. We thank Dr Yusuke Watanabe, Dr Ken Oofusa and Dr Kinji Asahina for helpful discussions during this research.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References