Introduction

Epidermal differentiation is a tightly controlled process that maintains the assembly of the epidermal barrier against the outside environment. The shed, terminally differentiated keratinocytes are replenished by a sequential action of two functionally distinct cell populations in the basal layer of the epidermis. The epidermal stem cells exist as clusters of infrequently dividing cells that give rise to transit amplifying daughter cells. The transit amplifying cell population comprises the majority of cells undergoing division at any given time in the epidermis. The progeny of the transit amplifying cells eventually withdraws from the cell cycle and undergoes remarkable morphological and biochemical changes, leading to the formation of anuclear and flattened cell envelopes (reviewed in Niemann and Watt, 2002; Blanpain and Fuchs, 2006; Watt et al., 2006).

The regulation of epidermal differentiation involves several interlinked signaling pathways. These include the Wnt/-catenin/T-cell factor pathway in the determination of epidermal cell lineage commitment (Gat et al., 1998; Niemann et al., 2002) and Sonic hedgehog that is implicated in hair follicle development together with Wnt signaling (St-Jacques et al., 1998; Silva-Vargas et al., 2005).

Another inherent feature of epidermal cells is an ability to respond rapidly to epidermal injury by a burst of cell proliferation and migration to regenerate the epidermis. Epidermal wound healing involves presentation of and signaling by growth factors and cytokines secreted by infiltrated immune cells, dermal fibroblasts, and activated keratinocytes (Werner and Grose, 2003). For example, signaling via fibroblast growth factor receptor 2-IIIb appears to be required for efficient wound healing, indicating a role for fibroblast growth factor (FGF), family members FGF-7, FGF-10, and FGF-22 in the induction of proliferation and migration of epidermal keratinocytes (Werner et al., 1994; Xia et al., 1999; Beyer et al., 2003).

Heparan sulfate proteoglycans (HSPGs) modulate the fate of several signaling molecules implicated in epidermal homeostasis and repair. These growth factors and cytokines include members of the FGF (Rapraeger et al., 1991; Guimond et al., 1993; Loo and Salmivirta, 2002), Wnt (Tsuda et al., 1999; Alexander et al., 2000; Baeg et al., 2001), transforming growth factor /bone morphogenetic protein (Lyon et al., 1997; Rider, 2006) families, as well as Sonic hedgehog (The et al., 1999; Rubin et al., 2002) and heparin-binding epidermal growth factor-like protein (Aviezer and Yayon, 1994). The interactions between oligosaccharide sequences in the glycosaminoglycan (GAG) side chains of HSPGs and growth factors can sequester the growth factors in the extracellular matrix, protect the growth factors against proteolysis, or act as a necessary component of the signaling complex (Gallagher, 2001; Nakata and Kimata, 2002).

Syndecans are cell-surface HSPGs that contain an ectodomain decorated with up to four long GAG chains, a transmembrane domain and a short, conserved cytoplasmic domain, that interacts with adapter proteins such as syntenins and calmodulin-associated serine/threonine kinase (CASK). Moreover, the cytoplasmic tail of at least syndecans 2 and 4 can bind directly to signaling proteins such as protein kinase C (see Bass and Humphries, 2002; Fears and Woods, 2006 for recent reviews). Syndecan-1 harbors both heparan and chondroitin sulfate side chains and is the most abundant family member in the epidermis. Syndecan-1 is strongly expressed in the suprabasal, differentiating epidermal cells, and only present in low levels in the basal cells comprising the proliferating cell populations (Sanderson et al., 1992; Inki et al., 1994b). However, the expression level of syndecan-1 in all cell layers of the epidermis is markedly upregulated upon wounding (Elenius et al., 1991). A role for syndecan-1 in the regulation of wound healing is underlined by investigation of corneal wound healing in syndecan-null mice (Stepp et al., 2002). The wounded syndecan-negative corneas failed to initiate a proliferative burst 24 hours after wounding. In addition, the migration of the corneal cells was hampered by the absence of syndecan-1 (Stepp et al., 2002).

To study the function of HSPGs in the regulation of epidermal differentiation and wound healing, we utilized human keratin 14 promoter to express human syndecan-1 in mouse epidermis. We report here that syndecan-1 overexpression can increase proliferation of newborn epidermal keratinocytes and change the expression of differentiation-related proteins. However, the transgene expression compromises normal epidermal regeneration during adult wound healing by reducing cell proliferation of the newly re-epithelialized epidermis.

Results

Generation of syndecan-1 transgenic mice

Syndecan-1 is the predominant cell-surface proteoglycan in the epidermis, where it is localized mostly to the surface of suprabasal spinous cells (Sanderson et al., 1992; Inki et al., 1994b). We confirmed that these observations are also valid in newborn mouse epidermis. Immunostaining of mouse syndecan-1 in the epidermis of newborn mice revealed very low expression in the basal epidermis compared to suprabasal cells (Figure 1a). Furthermore, we established that in the inbred C57 Bl/6J strain used in our studies, syndecan-1 expression is strongly upregulated throughout the hyperproliferative wound edge epidermis (Figure 1b). To investigate the biological role of restricted expression of syndecan-1 in the stem and transit amplifying (TA) cell compartments of the epidermis, we used a construct where the coding region of the human syndecan-1 gene is under the human K14 keratin promoter (Figure 1c) to generate transgenic mice. Six DNA-positive founder animals were detected and subsequently shown to transmit the transgene to the next generation. Two lines with equally high expression levels of the transgene were maintained for further studies.

Figure 1.

Generation of K14-HSyn transgenic mice. (a) Endogenous mouse syndecan is expressed most strongly in the spinous and granular cell layers of the epidermis. Mouse epidermis was stained with the rat mAb 281-2 against syndecan-1 core protein. The arrowhead in panels a, b, and d points at the dermal–epidermal junction. (b) Upon wounding, syndecan-1 expression is upregulated in the proliferating keratinocytes, but downregulated at the leading front of migrating epidermal sheet (^*). (c) The transgenic construct. Gray rectangles represent the promoter and 3'untranslated region of the human K14 keratin gene. The white rectangle represents the coding region of the human syndecan-1 full-length cDNA. K, KpnI; S, SacII; E, EcoRV; the EcoRV sites indicate the diagnostic fragment used in Southern blots. (d) Immunohistochemical detection of the transgene expression in the tail epidermis of a DNA-positive mouse. Human syndecan-1 was detected with mAb B.B4. (e) B.B4 staining of the skin of a control littermate. No specific staining can be detected. The arrowhead in panels d and e indicates the basal cell layer. (f) Detection of the transgene mRNA in the epidermis by Northern blotting. Human syndecan 1 mRNA (right hand panel) was present in the total RNA isolated from the skin of DNA-positive mice (Tg, middle lane), but absent in the skin of DNA-negative littermates (WT, outer lanes). Endogenous mouse syndecan-1 mRNA was present in all skin samples (left hand panel). Bar, 50 m.

Full figure and legend (99K)

The expression of human syndecan-1 in transgenic skin was confirmed by Northern blotting and immunohistochemical staining, using the mAb B.B4 specific for human syndecan-1 (Wijdenes et al., 1996). In the skin, strong specific staining was observed only in the basal layer of the epidermis of DNA-positive animals (Figure 1d) in all analyzed founder lines, indicating that the transgene was correctly expressed in the epidermal stem cells and basal transit amplifying keratinocytes only. No specific staining was observed in the skin of negative littermates (Figure 1e). Likewise, expression of human syndecan-1 was detected in total skin RNA of transgenic but not control littermate mice (Figure 1f). As the transgene expression was confined to the basal layer and the majority of endogenous syndecan-1 was present in the suprabasal epidermis, the cells expressing the transgene showed at least five-fold overexpression of the protein, compared to the endogenous gene.

Transgenic syndecan is a heparan and chondroitin sulfate proteoglycan

To verify whether the human syndecan detected in the basal epidermis of the transgenic animals is expressed as a proteoglycan, we purified diethylaminoethyl (DEAE)-bound proteoglycan fractions from the skin of transgenic mice and control littermates. Endogenous mouse syndecan-1 (from the suprabasal keratinocytes) was further separated from the total proteoglycan fraction by affinity to mAb 281-2 sepharose beads. Proteoglycan fractions from 281-2-bound and unbound material were subjected to gradient gel electrophoresis and detected by Western blotting (Figure 2a). Both the endogenous and transgenic syndecans run as a smear from 90 to 200 kDa, as expected, from proteoglycans carrying GAG side chains of heterogeneous size (Figure 2a). Endogenous syndecan was of equal size and approximately equally abundant in the epidermis of both DNA-positive and -negative mice. The mAb B.B4 detected proteoglycans only in the skin samples from the DNA-positive animals (Figure 2a). Furthermore, B.B4-reactive proteoglycans were only present in the fraction that was not bound to 281-2 sepharose beads (Figure 2a), which confirms that B.B4 does not cross-react with the endogenous mouse syndecan. The digestion of the proteoglycans by heparitinase or chondroitinase revealed that most side chains attached to the transgenic human syndecan core protein were relatively long heparan sulfate chains (Figure 2b). The digestion of the total proteoglycans fraction with heparitinase reduced the proteoglycans smear to a fuzzy band of about 90 kDa, whereas chondroitinase ABC digestion affected the size distribution of the proteoglycans only slightly, but reduced the intensity of the smear (Figure 2b). These results are consistent with the previous descriptions of syndecan-1 structure in cultured human keratinocytes (Sanderson et al., 1992).

Figure 2.

Transgenic syndecan-1 is expressed as a proteoglycan with attached heparan and chondroitin sulfate chains. (a) Western blot of proteoglycans purified from the skin of transgenic and control mice. Total proteoglycans from transgenic (tg) and control mouse (wt) skin were further fractioned by mAb 281-2 affinity beads to bound (b; endogenous mouse syndecan) and unbound supernatant (s; all other proteoglycans) fractions, separated on a gradient SDS-PAGE and blotted with either mAb 281-2 or B.B4 against human syndecan-1. Note that B.B4 recognizes a proteoglycan only in the TG skin, and does not cross-react with the endogenous, 281-2-bound syndecan-1. Lines indicate migration of 200 and 70 kDa molecular weight markers. (b) Transgenic syndecan-1 harbors heparan sulfate and chondroitin sulfate side chains. Purified proteoglycans from the skin of a transgenic mouse were incubated with the digestion buffer alone (control) or digested with heparitinase (H'ase) or chondroitinase ABC (C'ase) before separation in a 2–15% gradient SDS-PAGE and Western blotting with B.B4 mAb against human syndecan-1.

Full figure and legend (46K)

Epidermal hyperproliferation in the K14-HSyn transgenic mice

Histological analysis of skin from newborn F1 generation transgenic K14-HSyn mice revealed large areas of thickened epidermis (Figure 3). The transgenic epidermis appeared to have a normal cornified layer, occasional parakeratosis in the granular layer, and increased number of spinous cell layers (Figure 3b). The number of suprabasal cell layers was increased and flattening and anucleation of the cells appeared to be delayed (Figure 3). The extent of the hyperproliferation appeared to vary between different skin locations and between individual mice. Some lesions contained localized areas of remarkably thickened epidermis that appeared to involve some downgrowth into the dermis (Figure 3c).

Figure 3.

Hyperproliferation in newborn K14 Hsyn epidermis. (a) Hematoxylin–eosin-stained back skin section from 3 day-old-control mouse mice. (b) Hematoxylin–eosin stained back skin section K14 Hsyn transgenic, (c). K14 Hsyn transgenic (different founder line than in panel b). (d) PCNA staining of a control back skin section. (e) Quantitation of PCNA-positive nuclei. Percentages of PCNA-positive nuclei in basal and first suprabasal cell layers of control and transgenic mice (three each) were counted from digital microscope images. Error bars indicate standard deviation. The difference in the PCNA-positive nuclei was statistically significant (t-test, P<0.05). (f) PCNA staining of K14 Hsyn transgenic back skin section. Bar, 50 m.

Full figure and legend (142K)

To analyze whether the thickened epidermis resulted from increased proliferation, we stained paraffin sections of transgenic and wild-type epidermis with an antibody against proliferating cell nuclear antigen (PCNA). The number of proliferating cells was significantly increased in the epidermis of transgenic mice and PCNA-positive cells were frequently seen in the first suprabasal cell layer, while less than 1% of the suprabasal cells were PCNA positive in the control epidermis (Figure 3d–f). A similar increase in proliferating cells was also seen by immunostaining the newborn mouse interfollicular epidermis with an antibody against Ki67 proliferation marker (Figure S1). Interestingly, no change in Ki67 immunoreactivity was observed in hair follicle outer root sheaths or hair bulbs (Figure S1). No difference was observed in the expression or distribution of two syndecan-binding growth factors, FGF-2 and FGF-7, indicating that the altered keratinocyte proliferation is not due to increased expression of these growth factors. FGF-2 expression was detected predominantly in basal keratinocytes, whereas FGF-7 was expressed at low levels in the dermis (Figure S2). These data are consistent with previous reports on the tissue distribution of these growths factors (Schulze-Osthoff et al., 1990; Werner et al., 1992; Kibe et al., 2000).

Expression of epidermal differentiation markers in K14-HSyn transgenic mice

Epidermal hyperproliferation, wound healing, and other epidermal cellular stress situations are usually accompanied by expression of K6 keratin (Fuchs, 1995). In the control animals, K6 was restricted to the hair follicles (Figure 4a), whereas in the hyperproliferative transgenic epidermis, strong K6 expression was seen in both hair follicle outer root sheath and the interfollicular epidermis (Figure 4b). During epidermal differentiation, expression of K10 replaces K14 in the suprabasal epidermis. Keratin 10 was expressed in the transgenic skin in a manner similar to control skin, beginning mostly in the first suprabasal cell layer (Figure 4c and d).

Figure 4.

Keratin and involucrin expression in the K14 HSYn transgenic epidermis. (a) Keratin 6 expression in control skin. Note that the staining is restricted to hair follicles. (b) Keratin 10 staining in the control epidermis. (c) Keratin 6 staining in the transgenic epidermis. Note the prominent staining of the interfollicular epidermis. (d) Keratin 10 staining in the transgenic epidermis. (e) Involucrin expression in control newborn mouse skin. (f) Involucrin expression in K14Hsyn transgenic mouse skin. Bar equals 50 m in all panels.

Full figure and legend (122K)

The progress of terminal differentiation was further evaluated by studying the expression of involucrin, an early precursor protein for cornified envelopes (Kalinin et al., 2002), in the transgenic and control epidermis. Involucrin expression commenced in the spinous layer and the antibody staining was strongest in the granular layer (Figure 4e). In the transgenic epidermis, strong involucrin staining was detected in the upper spinous and granular layers (Figure 4f), thus indicating that terminal differentiation takes place in the hyperproliferative transgenic epidermis. We used immunoblotting of epidermal cell protein extracts to further evaluate terminal differentiation of the transgenic keratinocytes (Figure 5). We also included in our analysis syndecan-1-null animals that have not been reported to display any phenotype in uninjured adult epidermis (Stepp et al., 2002). We found that involucrin expression was not changed in the transgenic epidermis, and blotting with an antibody against pro-filaggrin revealed the characteristic ladder of pro-filaggrin processing in both control and transgenic epidermis (Figure 5). The only observed sign of a delayed differentiation was relative absence of crosslinked loricrin in the transgenic epidermis. Transglutaminase crosslinking of loricrin in the epidermis resulted in distinct bands in control epidermis, whereas in transgenic epidermis, the majority of loricin was present as a single monomeric 30 kDa polypeptide (Figure 5). The sizes of crosslinked and monomeric loricrin species corresponded well to those observed in in vitro crosslinking assays (Candi et al., 2001). In contrast to K14 hSyn transgenic mice, no difference between syndecan-1 knock-out animals and wild-type animals was observed (Figure 5). The delay in loricrin processing did not affect the completion of the terminal differentiation program. Furthermore, an in situ assay for transglutaminases indicated that the combined activities of transglutaminases 1 and 3 were present in the upper spinous and granular layers of the transgenic epidermis (Figure S3).

Figure 5.

Expression of differentiation markers in epidermal protein extracts. Neonate wild-type (WT), transgenic (Tg) and syndecan-null (KO) epidermises (three mice each) were isolated by dispase treatment, solubilized in Laemmli's buffer, and analyzed by Western Blotting. The blots were probed with antibodies against pro-filaggrin (top panels), loricrin (middle panels), and involucrin (bottom panels). The ladder of bands in the filaggrin blot shows characteristic pattern of pro-filaggrin processing. M indicates monomeric loricrin and C indicates crosslinked loricrin bands.

Full figure and legend (41K)

K14-Hsyn expression interferes with wound re-epithelialization

Endogenous syndecan-1 is up-regulated over 15-fold in the proliferating and migrating keratinocytes at the wound edge (Elenius et al., 1991; Figure 1). Recently, Stepp and co-workers reported that lack of syndecan-1 inhibits keratinocyte migration and proliferation in corneal wounds (Stepp et al., 2002). To compare wound healing in K14 Hsyn mice and syndecan-1-null mice, we bred both genotypes in C57Bl/6J background for seven generations and performed punch biopsy wound healing experiment on 6- to 7-week old (telogen stage hair cycle) animals. At day 3 post-wounding, when control animals show extensive hyperproliferation at the wound edge and the knockout mice only slightly less pronounced activation, there was only a small increase in epidermal thickness at the K14 hSyn transgenic wound edge (Figure 6). The difference in keratinocyte proliferation was even more pronounced at day 7, when the control epidermis in the middle of the closed wound was stratified and hyperproliferative (Figure 6d and g), whereas the transgenic epidermis, although covering the wound, consisted of only a few layers of flattened keratinocytes that resembled cells normally seen in the upper spinous and granular layers of the epidermis (Figure 6e and h). It appeared that only proliferation, but not keratinocyte migration was affected, as no difference in wound closure was observed.

Figure 6.

Wound healing in K14 Syn and syndecan-/- mice. (a–c) Hematoxylin–eosin staining of punch biopsy wound 3 days after wounding. Arrowheads indicate hyperproliferative epidermis at wound edge, and ^*indicates the migrating epidermal front. (a) Wild type, (b) K14-hSyn transgenic, (c) syndecan-1 null. (d–f) Hematoxylin–eosin stainings of 7-day-old punch biopsy wounds. Arrows indicate the wound edges. (d) Wild type, and (e) K14Hsyn transgenic mice. Note the thin re-epithelialized epidermis, (f) syndecan-1-null mouse. (g–i) High-magnification images of the new epidermis. (d) Wild type, (e) K14Hsyn transgenic, (f) syndecan-1 null mice. Bar in panels a–f equals 200 m, and in panels g–i equals 50 m.

Full figure and legend (160K)

The morphology of the syndecan-1-null wounds appeared to be more variable, but the re-epithelialized epidermis was usually less well organized and thinner than in controls (Figure 6f and i). In both transgenic and knockout wounds (day 7 time point), the granulation tissue appeared to be less well organized than in control wounds, indicating a possible role for syndecan-1 in the mediation of signals to either organize or remodel the newly laid dermis. The difference between syndecan-1 overexpressing epidermis and control or syndecan-1-null epidermis was confirmed by morphometric analysis. Measurements of the thickness of the re-epithelialized epidermis from three independent wounds for each genotype showed that the transgenic epidermis was significantly thinner than either wild type or syndecan knockout epidermis (measured as the area of epidermis in the middle of the wound from hematoxylin-eosin stained sections) (Figure 7a). Furthermore, cell proliferation, as visualized by Ki67-positive nuclei, was greatly reduced in the transgenic epidermis covering the wound (Figure 7b), compared to re-epithelialized wild-type wound (Figure 7c).

Figure 7.

Epidermal thickness and cell proliferation in re-epithelialized epidermis. (a) Morphometric measurement of the new epidermis. Mean and SEM of the transverse area of epidermis in the middle of the wound of hematoxylin–eosin stained paraffin sections from three independent wounds for each genotype. The difference between wild-type mice and transgenic mice is statistically significant (t-test, P<0.05). (b) K14 Hsyn transgenic mouse, Ki67 immunohistochemistry. Note the lack of cell proliferation in the epidermis covering the wound (arrow). (c) Wild type, Ki67 immunohistochemistry. Note cell proliferation in the epidermis covering the wound (arrow). Ki67-positive sebaceous gland nuclei are indicated with ^* in panels b and c. Bar equals 50 m in panels b and c.

Full figure and legend (67K)

Isolation of primary keratinocytes from transgenic and wild-type mice demonstrated that the decreased proliferation can also be observed in vitro. Transgenic keratinocytes grown on collagen-coated matrices proliferated very slowly compared to wild-type keratinocytes, and it proved impossible to establish long-term cultures from these mice (Figure S4).

Discussion

In order to study the role of HSPGs in the proliferation and differentiation of epidermal keratinocytes, we expressed human syndecan-1 in the basal layer of mouse epidermis that harbors epidermal stem cells and their rapidly proliferating progeny, transit amplifying cells. The transgene expression was restricted to the basal layer of the epidermis and the outer root sheath of hair follicles, as shown by immunohistochemical staining (Figure 1). Importantly, the transgene was expressed as a proteoglycan that, as expected for syndecan-1, carried both heparan sulfate and chondroitin sulfate side chains (Figure 2). This confirms the functionality of the transgene-derived protein as a proteoglycan. Neither the mRNA levels nor the GAG composition of the endogenous mouse syndecan-1 was affected, indicating that the transgene expression did not compromise expression of other proteoglycans.

Newborn transgenic mice exhibited hyperproliferative skin lesions that were characterized by suprabasal cell proliferation and increased number of suprabasal cell layers. The thickened transgenic epidermis expressed keratin 6, a hallmark of hyperproliferation, but the expression of markers for terminal differentiation, keratin 10 and involucrin, was not overtly disrupted. Thus, expression of syndecan-1 in basal keratinocytes can cause hyperproliferation, but does not prevent the commitment of the transit amplifying cells to differentiation. No difference was observed in involucrin expression levels or in pro-filaggrin processing. Only monomeric but not crosslinked loricrin was detected in the epidermal extracts of transgenic mice, which suggests a delay in epidermal differentiation (Figure 5). However, it should be noted that even though loricrin is a major component of cornified envelopes, it is not essential for skin homeostasis as loricrin-null mice eventually develop a normal skin and do not suffer from barrier defects (Koch et al., 2000).

During wound healing, syndecan-1 is upregulated in the wound edge keratinocytes (Elenius et al., 1991). The upregulation is, at least in part, achieved by activation of a far upstream enhancer element in syndecan-1 gene that in cultured keratinocytes is responsive to FGF and epidermal growth factor families of growth factors and is modulated by interactions with the extracellular matrix (Jaakkola et al., 1998a, 1998b; Määttä et al., 1999). We found that overexpression of syndecan-1 has a profound effect on cell proliferation in the newly re-epithelialized epidermis. This effect is clearly distinct from the phenotype of the newborn epidermis and suggests that the "activated" keratinocyte phenotype responsible for re-epithelialization is regulated differently from the multilayered newborn mouse epidermis.

The difference between the effect of ectopic syndecan-1 expression on epidermal differentiation in newborn mice compared to adult resting skin and to wound healing is intriguing. A possible explanation is the role of a large number of heparin-binding growth factors and cytokines that are involved in wound healing, although they only play a small role in normal epidermal homeostasis (reviewed by Werner and Grose, 2003). Moreover, it is likely that a different set of growth factors is involved in the development of the skin, compared to repair of the epidermis in wound healing. For example, the expression of FGF-7 (keratinocyte growth factor) that binds to FGFR2IIIb receptor is upregulated over 100-fold upon wounding (Werner et al., 1992) and dominant-negative FGFRIIIb impairs wound healing (Werner et al., 1994). Subsequent biochemical analysis has revealed that both FGF-7 and the receptor bind heparan sulfate (LaRochelle et al., 1999). Crucially, at high concentrations, heparin inhibits FGF-7 but not FGF-1 activity (LaRochelle et al., 1999). Moreover, only a small fraction of heparin shows high affinity to FGF-7, as opposed to FGF-1 and FGF-2, and specifically supports FGF7-FGFRIIIb signaling (Luo et al., 2006). Thus, overexpression of syndecan-1 can change the effect of heparan sulfate from promotion to inhibition of FGF-7 signaling. Likewise, some signaling pathways that are crucial in skin development, such as wnt/-catenin pathway, are not active in the maintenance of the adult epidermal homeostasis (Posthaus et al., 2002).

Further evidence for the function of syndecan-1 in keratinocyte proliferation has been obtained from syndecan-1 knockout mice (Stepp et al., 2002) and from mice that overexpress syndecan-1 systemically under the cytomegalovirus (CMV) promoter (Elenius et al., 2004). Stepp et al. (2002) analyzed the proliferation of corneal keratinocytes before and after corneal debridement in syndecan-1 gene targeted mice. They found that syndecan-1 is required for induced proliferation and migration of corneal keratinocytes. The fact that both loss (Stepp et al., 2002) and overexpression (this study) of syndecan-1 can compromise inducible keratinocyte proliferation is consistent with the dual effect of heparan sulfate on and FGF-7 signaling and suggests that there exists an optimal proteoglycan expression level that regulates the proliferative potential of keratinocytes during wound healing.

Elenius et al. (2004) studied wound healing in mice that overexpress syndecan-1 in several tissues, including the skin and blood. The elevated levels of shed syndecan-1 ectodomain in these animals inhibit both re-epithelialization and remodeling of the newly laid matrix in dermis. Since the wound healing of wild-type skin grafted to overexpressing mice was also hampered, the effects were largely attributed to the systemic overexpression affecting also the dermis, blood vessels, and immune cells (Elenius et al., 2004). In this study, we have been able to pinpoint the keratinocyte-specific effects of syndecan-1 expression by using a transgene only expressed in basal keratinocytes. Notably, the studies conducted by us and by Elenius et al. (2004) indicate that syndecan-1 expressed either by keratinocytes themselves or by the other surrounding cell types can equally modify keratinocyte proliferation. Our studies also support the observations of Elenius et al. (2004) that organization of granulation tissue is influenced by syndecan-1 expression in skin. The changes in the granulation tissues seen in both studies and the increase of dermal cell proliferation observed in the newborn transgenic skin suggests that epidermal expression of syndecan-1 can influence signaling in dermal cells.

In conclusion, syndecan-1 is an important modulator of epidermal cell proliferation. Stepp et al. (2002) recently reported how syndecan-1 knockout mice display impaired proliferation and migration of corneal keratinocytes in eye abrasion wound healing. Here, we have demonstrated that overexpression of syndecan-1 in basal keratinocytes promotes keratinocyte proliferation during epidermal development, but restricts proliferation during wound healing. Taken together, these studies suggest a role for syndecan-1 in the control of the proliferative potential of epidermal transit amplifying cells.

Materials and Methods

Generation of transgenic mice and animal procedures

To target the expression of human syndecan-1 in the basal layer of the epidermis in mice, we subcloned the coding region of full-length human syndecan-1 cDNA (Mali et al., 1990) into an expression vector containing human keratin K14 promoter and 3'untranslated region (Vassar et al., 1989). The K14hSyn insert was purified for microinjection and founder mice were generated using standard transgenic methods. DNA-positive mice were identified by Southern blotting of tail-tip DNA. For Southern blotting, the genomic DNA was cut with EcoRV, electrophoresed, and transferred on Hybond N+ (GE Healthcare, Little Chalfont, Buckinghamshire, UK) filters and probed with full-length syndecan-1 cDNA or an EcoRV fragment (Figure 1c) encompassing part of the K14 promoter and part of the human syndecan-1 cDNA. K14Hsyn overexpressing mice and syndecan-1-null mice (Stepp et al., 2002) were bred to C57Bl/6J background for seven generations, to minimize any effect of variation of the genetic background in subsequent experiments. Animal work was approved by Animal Ethics Committees of Universities of Turku and Durham and covered by a UK Home Office Project license.

To study wound healing in overexpressing, gene targeted syndecan-1-null and C57Bl/6J wild-type control mice, 6 to 7-week old animals were shaved; 2 days later, only animals that were still in the resting phase of the hair follicle cycle were used in the experiments. Punch biopsy wounds measuring 3 mm were made on the back skin of anesthetized mice. At days 3, 5, 7, and 10 post-wounding, the animals were killed and wound tissue along with some surrounding skin was excised and processed for histology and immunohistochemistry.

Primary keratinocytes were isolated and grown in Defined Keratinocyte Growth Medium (Invitrogen, Paisley, UK) on collagen-coated culture dishes, as described earlier (Määttä et al., 2001). Cell proliferation was measured as the number of viable cells, as described (Long et al., 2006).

RNA isolation and Northern blotting

Total RNA was isolated from snap–frozen skin samples by homogenization in 4 M guanidinium isothiocyanate and subsequent ultracentrifugation, as described (Mali et al., 1990). A 10 g weight of total RNA was separated on a 1% agarose gel, in the presence of formaldehyde, transferred on Hybond-N membranes, and sequentially probed with human and mouse syndecan-1 cDNA probes.

Purification of proteoglycans

Whole trunk skins of 1-week-old transgenic mice and their DNA-negative control littermates were snap–frozen in liquid nitrogen. The skins were homogenized in phosphate-buffered saline containing 1 mM phenyl-methane-sulfonyl-fluoride (PMSF), 1% Nonidet P40, 0.1% deoxycholate, and 0.1% SDS, sonicated for 5 minutes in a bath sonicator, and rotated for 15 minutes at +4°C. Detergent-insoluble material was separated by centrifugation and proteoglycans were isolated from the supernatants by DEAE sephacel, as described previously (Salmivirta et al., 1991). Endogenous mouse syndecan-1 was further affinity purified from the DEAE eluate by mAb 281-2 linked to Sepharose CL-4B beads (Pharmacia). GAG content of endogenous and transgene derived syndecans was determined by digestion of the chains using 1 mU/ml heparitinase and/or 0.1 U/ml chondroitinase ABC (Seikagagu Kogyo, Tokyo, Japan), as described (Elenius et al., 1990).

Epidermal protein extracts and Western blotting

Isolated murine skin pieces were incubated in 1.2 U/ml dispase (Roche Molecular Biochemicals, Mannheim, Germany) overnight at 4°C. The epidermis was peeled away from the dermis, washed twice in phosphate-buffered saline (calcium/magnesium free), and snap–frozen in liquid nitrogen. Whole cell protein was extracted from the separated epidermis using 2 Laemmli's sample buffer supplemented with protease inhibitor cocktail (Roche, Mannheim, Germany). Protein concentration was determined using the BCA protein assay reagent kit (Pierce, Cramlington, UK). Immunoblotting was performed using the semi-dry blotting method, according to the manufacturer's instructions (Bio-Rad, Hertfordshire, UK). 12–25 g proteins were loaded and were separated on 4–12% Bis–Tris gradient gels (Invitrogen) and transferred onto Hybond nitrocellulose membranes (Amersham Pharmacia Biotech, Little Chalfont, Buckinghamshire, UK). The membranes were probed using a panel of primary antibodies (filaggrin polyclonal antibody diluted 1:400 (PRB-417P; Covance, Berkeley, CA), Loricrin polyclonal antibody diluted 1:500 (ab24722; ABCAM, Cambridge, UK), Involucrin polyclonal antibody diluted 1:1,000 (ab28057; ABCAM) for 1 hour at room temperature. Protein bands were visualized using the ECL Plus chemiluminescent detection reagent (Amersham Pharmacia Biotech) on luminescent image analyzer (LAS-1000plus; Fuji Photo Film, UK).

Purified proteoglycans (mAb 281-2-bound and unbound fractions) were separated in 2–15% gradient SDS-PAGE gels and blotted on Hybond N+ membranes. Human syndecan-1 was detected by mAb B.B4 (Serotec, Oxford, UK) and mouse syndecan-1 by mAb 281-2.

Histology and immunohistochemistry

Tissue samples were fixed in phosphate-buffered formaline and embedded in paraffin blocks. Deparaffinized sections (5 m thick) were cut and immunohistochemistry was performed using Vectastain ABC kit (Vector Laboratories, Orton Southgate, Peterborough, UK). The primary antibodies used were as follows: mAb B.B4 (Serotec) for human syndecan-1; rat mAb 281-2 for mouse syndecan-1; a rabbit polyclonal against mouse involucrin; MK6, rabbit polyclonal against mouse keratin 6; MK10 rabbit polyclonal against mouse keratin 10 (all from Covance); NCL-PCNA (Novocastra, Newcastle Upon Tyne, UK) against PCNA; NCL-L-Ki67-MM1 (Novocastra) against proliferation antigen Ki67; polyclonal anti bFGF (40–63) (Calbiochem, Merck Chemicals Ltd., Nottingham, UK); and polyclonal anti FGF-7 (C19) (SantaCruz, Autogen Bioclear UK Ltd., Wiltshire, UK). Transglutaminase activity was visualized in unfixed frozen sections by using fluorescently labeled monodansylcadaverine (Invitrogen) as substrate, as described (Raghunath et al., 1998).

PCNA-positive nuclei were scored from digital micrographs of stained sections (three sections each from mice from two different founder lines). Number of positive nuclei was compared by Student's t-test.

Epidermal thickness in the re-epithelialized 7-day old wounds was analyzed by measuring a defined area of the epidermis in pixels from digital photomicrographs (50 original magnification) by NIH Image program. An equally wide section from the middle of the wound of three independent wounds from each genotype was analyzed.

Conflict of Interest

The authors declare no conflict of interest.

References

1. Alexander CM, Reichsman F, Hinkes MT, Lincecum J, Becker KA, Cumberledge S et al. (2000) Syndecan-1 is required for Wnt-1-induced mammary tumorigenesis in mice. Nat Genet 25:329–332 | Article | PubMed | ISI | ChemPort |
2. Aviezer D, Yayon A (1994) Heparin-dependent binding and autophosphorylation of epidermla growth factor (EGF) receptor by heparin-binding EGF-like growth factor but not by EGF. Proc Natl Acad Sci USA 91:12173–12177 | Article | PubMed | ChemPort |
3. Baeg GH, Lin X, Khare N, Baumgartner S, Perrimon N (2001) Heparan sulfate proteoglycans are critical for the organization of the extracellular distribution of Wingless. Development 128:87–94 | PubMed | ISI | ChemPort |
4. Bass MD, Humphries MJ (2002) Cytoplasmic interactions of syndecan-4 orchestrate adhesion receptor and growth factor receptor signaling. Biochem J 368:1–15 | Article | PubMed | ISI | ChemPort |
5. Beyer TA, Werner S, Dickson C, Grose R (2003) Fibroblast growth factor 22 and its potential role during skin development and repair. Exp Cell Res 287:228–236 | Article | PubMed | ISI | ChemPort |
6. Blanpain C, Fuchs E (2006) Epidermal stem cells of the skin. Annu Rev Cell Dev Biol 22:339–373 | Article | PubMed | ISI | ChemPort |
7. Candi E, Oddi S, Terrinoni A, Paradisi A, Ranalli M, Finazzi-Agro A et al. (2001) Transglutaminase 5 cross-links loricrin, involucrin, and small proline-rich proteins in vitro. J Biol Chem 276:35014–35023 | Article | PubMed | ISI | ChemPort |
8. Elenius K, Salmivirta M, Inki P, Mali M, Jalkanen M (1990) Binding of human syndecan to extracellular matrix proteins. J Biol Chem 265:17837–17843 | PubMed | ISI | ChemPort |
9. Elenius K, Vainio S, Laato M, Salmivirta M, Thesleff I, Jalkanen M (1991) Induced expression of syndecan in healing wounds. J Cell Biol 114:585–595 | Article | PubMed | ISI | ChemPort |
10. Elenius V, Gotte M, Reizes O, Elenius K, Bernfield M (2004) Inhibition by the soluble syndecan-1 ectodomains delays wound repair in mice overexpressing syndecan-1. J Biol Chem 279:41928–41935 | Article | PubMed | ISI | ChemPort |
11. Fears CY, Woods A (2006) The role of syndecans in disease and wound healing. Matrix Biol 25:443–456 | Article | PubMed | ISI | ChemPort |
12. Fuchs E (1995) Keratins in the skin. Annu Rev Cell Dev Biol 11:123–153 | Article | PubMed | ISI | ChemPort |
13. Gallagher J (2001) Heparan sulfate: growth control with a restricted sequence menu. J Clin Invest 108:357–361 | Article | PubMed | ISI | ChemPort |
14. Gat U, DasGupta R, Degenstein L, Fuchs E (1998) De novo hair follicle morphogenesis and hair tumors in mice expressing a truncated beta-catenin in skin. Cell 95:605–614 | Article | PubMed | ISI | ChemPort |
15. Guimond S, Maccarana M, Olwin BB, Lindahl U, Rapraeger AC (1993) Activating and inhibiting heparin sequences for FGF-2 (basic FGF). Distinct requirements for FGF-1, FGF-2 and FGF-4. J Biol Chem 268:23906–23914 | PubMed | ISI | ChemPort |
16. Inki P, Larjava H, Haapasalmi K, Miettinen HM, Grenman R, Jalkanen M (1994b) Expression of syndecan-1 is induced by differentiation and suppressed by malignant transformation of human keratinocytes. Eur J Cell Biol 63:43–51 | PubMed | ISI | ChemPort |
17. Jaakkola P, Kontusaari S, Kauppi T, Maatta A, Jalkanen M (1998a) Wound reepithelialization activates a growth factor-responsive enhancer in migrating keratinocytes. FASEB J 12:959–969 | PubMed | ISI | ChemPort |
18. Jaakkola P, Maatta A, Jalkanen M (1998b) The activation and composition of FiRE (an FGF-inducible response element) differ in a cell type- and growth factor-specific manner. Oncogene 17:1279–1286 | Article | PubMed | ISI | ChemPort |
19. Kalinin AE, Kajava AV, Steinert PM (2002) Epithelial barrier function: assembly and structural features of the cornified cell envelope. Bioessays 24:789–800 | Article | PubMed | ISI | ChemPort |
20. Kibe Y, Takenaka H, Kishimoto S (2000) Spatial and temporal expression of basic fibroblast growth factor protein during wound healing of rat skin. Br J Dermatol 143:720–727 | Article | PubMed | ISI | ChemPort |
21. Koch PJ, de Viragh PA, Scharer E, Bundman D, Longley MA, Bickenbach J et al. (2000) Lessons from loricrin-deficient mice: compensatory mechanisms maintaining skin barrier function in the absence of a major cornified envelope protein. J Cell Biol 151:389–400 | Article | PubMed | ISI | ChemPort |
22. LaRochelle WJ, Sakaguchi K, Atabey N, Cheon HG, Takagi Y, Kinaia T et al. (1999) Heparan sulfate proteoglycan modulates keratinocyte growth factor signaling through interaction with both ligand and receptor. Biochemistry 38:1765–1771 | Article | PubMed | ISI | ChemPort |
23. Long HA, Boczonadi V, McInroy L, Goldberg M, Maatta A (2006) Periplakin-dependent re-organisation of keratin cytoskeleton and loss of collective migration in keratin-8-downregulated epithelial sheets. J Cell Sci 119:5147–5159 | Article | PubMed | ISI | ChemPort |
24. Loo BM, Salmivirta M (2002) Heparin/heparan sulfate domains in binding and signaling of fibroblast growth factor 8b. J Biol Chem 277:32616–32623 | Article | PubMed | ISI | ChemPort |
25. Luo Y, Ye S, Kan M, McKeehan WL (2006) Control of fibroblast growth factor (FGF) 7- and FGF1-induced mitogenesis and downstream signaling by distinct heparin octasaccharide motifs. J Biol Chem 281:21052–21061 | Article | PubMed | ISI | ChemPort |
26. Lyon M, Rushton G, Gallagher JT (1997) The interaction of the transforming growth factor-betas with heparin/heparan sulfate is isoform-specific. J Biol Chem 272:18000–18006 | Article | PubMed | ISI | ChemPort |
27. Määttä A, DiColandrea T, Groot K, Watt FM (2001) Gene targeting of envoplakin, a cytoskeletal linker protein and precursor of the epidermal cornified envelope. Mol Cell Biol 21:7047–7053 | Article | PubMed | ISI | ChemPort |
28. Määttä A, Jaakkola P, Jalkanen M (1999) Extracellular matrix-dependent activation of syndecan-1 expression in keratinocyte growth factor-treated keratinocytes. J Biol Chem 274:9891–9898 | Article | PubMed |
29. Mali M, Jaakkola P, Arvilommi A-M, Jalkanen M (1990) Sequence of human syndecan-1 indicates a novel gene family of integral membrane proteoglycans. J Biol Chem 265:6884–6889 | PubMed | ISI | ChemPort |
30. Nakata H, Kimata K (2002) Heparan sulfate fine structure and specificity of proteoglycan functions. Biochim Biophys Acta 1573:312–318 | PubMed |
31. Niemann C, Owens DM, Huelsken J, Birchmeier W, Watt FM (2002) Expression of DeltaNLef1 in mouse epidermis results in differentiation of hair follicles into squamous epidermal cysts and formation of skin tumours. Development 129:95–109 | PubMed | ISI | ChemPort |
32. Niemann C, Watt FM (2002) Designer skin: lineage commitment in postnatal epidermis. Trends Cell Biol 12:185–192 | Article | PubMed | ISI | ChemPort |
33. Posthaus H, Williamson L, Baumann D, Kemler R, Caldelari R, Suter MM et al. (2002) Beta-catenin is not required for proliferation and differentiation of epidermal mouse keratinocytes. J Cell Sci 115:4587–4595 | Article | PubMed | ISI | ChemPort |
34. Raghunath M, Hennies HC, Velten F, Wiebe V, Steinert PM, Reis A et al. (1998) A novel in situ method for the detection of deficient transglutaminase activity in the skin. Arch Dermatol Res 290:621–627 | Article | PubMed | ISI | ChemPort |
35. Rapraeger AC, Krufka A, Olwin BB (1991) Requirement of heparan sulfate for bFGF-mediated fibroblast growth and myoblast differentiation. Science 252:1705–1708 | Article | PubMed | ISI | ChemPort |
36. Rider CC (2006) Heparin/heparan sulphate binding in the TGF-beta cytokine superfamily. Biochem Soc Trans 34:458–460 | Article | PubMed | ISI | ChemPort |
37. Rubin JB, Choi Y, Segal RA (2002) Cerebellar proteoglycans regulate sonic hedgehog responses during development. Development 129:2223–2232 | PubMed | ISI | ChemPort |
38. Salmivirta M, Elenius K, Vainio S, Hofer U, Chiquet-Ehrismann R, Thesleff I et al. (1991) Syndecan from embryonic tooth mesenchyme binds tenascin. J Biol Chem 266:7733–7739 | PubMed | ISI | ChemPort |
39. Sanderson RD, Hinkes MT, Bernfield M (1992) Syndecan-1, a cell-surface proteoglycan, changes in size and abundance when keratinocytes stratify. J Invest Dermatol 99:390–396 | Article | PubMed | ISI | ChemPort |
40. Schulze-Osthoff K, Goerdt S, Sorg C (1990) Expression of basic fibroblast growth factor (bFGF) in Kaposi's sarcoma: an immunohistologic study. J Invest Dermatol 95:238–240 | Article | PubMed | ChemPort |
41. Silva-Vargas V, Lo Celso C, Giangreco A, Ofstad T, Prowse DM, Braun KM et al. (2005) Beta-catenin and Hedgehog signal strength can specify number and location of hair follicles in adult epidermis without recruitment of bulge stem cells. Dev Cell 9:121–131 | Article | PubMed | ISI | ChemPort |
42. St-Jacques B, Dassule HR, Karavanova I, Botchkarev VA, Li J, Danielian PS et al. (1998) Sonic hedgehog signaling is essential for hair development. Curr Biol 8:1058–1068 | Article | PubMed | ISI | ChemPort |
43. Stepp MA, Gibson HE, Gala PH, Sta. Iglesia DD, Pajoonesh-Ganji A, Pal-Ghosh S et al. (2002) Defects in keratinocyte activation during wound healing in the syndecan-1 deficient mouse. J Cell Sci 115:4517–4531 | Article | PubMed | ISI | ChemPort |
44. The I, Bellaiche Y, Perrimon N (1999) Hedgehog movement is regulated through tout velu-dependent synthesis of a heparan sulfate proteoglycan. Mol Cell 4:633–639 | Article | PubMed | ISI | ChemPort |
45. Tsuda M, Kamimura K, Nakato H, Archer M, Staatz W, Fox B et al. (1999) The cell-surface proteoglycan Dally regulates Wingless signaling in Drosophila. Nature 400:276–280 | Article | PubMed | ISI | ChemPort |
46. Vassar R, Rosenberg M, Ross S, Tyner A, Fuchs E (1989) Tissue-specific and differentiation specific expression of a human K14 keratin gene in transgenic mice. Proc Natl Acad Sci USA 86:1563–1567 | Article | PubMed | ChemPort |
47. Watt FM, Lo Celso C, Silva-Vargas V (2006) Epidermal stem cells: an update. Curr Opin Genet Dev 16:518–524 | Article | PubMed | ISI | ChemPort |
48. Werner S, Grose R (2003) Regulation of wound healing by growth factors and cytokines. Physiol Rev 83:835–870 | PubMed | ISI | ChemPort |
49. Werner S, Peters KG, Longaker MT, Fuller-Pace F, Banda MJ, Williams LT (1992) Large induction of keratinocyte growth factor expression in the dermis during wound healing. Proc Natl Acad Sci USA 89:6896–6900 | Article | PubMed | ChemPort |
50. Werner S, Smola H, Liao X, Longaker MT, Krieg T, Hofschneider PH et al. (1994) The function of KGF in morphogenesis of epithelium and re-epithelialization of wounds. Science 266:819–822 | Article | PubMed | ISI | ChemPort |
51. Wijdenes J, Vooijs WC, Clement C, Post J, Morard F, Vita N et al. (1996) A plasmocyte selective monoclonal antibody (B-B4) recognizes syndecan-1. Br J Haematol 94:318–323 | Article | PubMed | ISI | ChemPort |
52. Xia YP, Zhao Y, Marcus J, Jiminez PA, Ruben SM, Moore PA et al. (1999) Effects of keratinocyte growth factor-2 (KGF-2) on wound healing in an ischemia-impaired rabbit ear model and scar formation. J Pathol 188:431–438 | Article | PubMed | ISI | ChemPort |

Acknowledgments

We would like to dedicate this paper to the memory of Dr Markku Salmivirta, who recently passed away. We would like to thank the transgenic core services at the University of Turku for the microinjections and Mrs Anni Kieksi for expert technical help. This project was supported by BBSRC and AICR (Association of International Cancer Research) grants to AM.

SUPPLEMENTARY MATERIAL

Figure S1. Expression of the proliferation marker Ki67 in the interfollicular epidermis and hair follicles.

Figure S2. Expression of FGF-2 and FGF-7 in transgenic and control skin.

Figure S3. Transglutaminase 1 and 3 activity in newborn mouse epidermis.

Figure S4. Growth of control and transgenic primary keratinocytes.