Results

The gene expression profile of the DP reflects the signalling environment in the base of the hair follicle during anagen

The DP and the DS comprise the contiguous follicular dermis, the outermost layer of the hair follicle. In the rat vibrissae as the DS is separated from the hair-follicle epithelium by a glassy membrane (Figure 1a), we hypothesized that any genes that were exclusively expressed in the DP during anagen would be expressed as the result of reciprocal interactions between the DP and the matrix epithelium. To identify these genes we performed microarray hybridization analyses comparing the expression of over 8000 genes in RNA from pooled freshly dissected DP and DS from male brown Norway rats. Twenty-three genes were expressed at higher levels in the DP based on a statistical analysis of the expression data (Table S1). Consistent with the hypothesis that genes were induced in the DP during anagen, no genes were consistently expressed at higher levels in the DS.

Figure 1.

Outlines of the dissections for hybridization and confirmation of selected hybridization results by RT-PCR. (A) Schematic of a hair follicle showing the major tissues and outlining the dissections used in the hybridization experiments. The horizontal line shows the level of transection below which DP and DS are obtained (gray tissues). ORS, outer root sheath; IRS, inner root sheath; HS, hair shaft; DP, dermal papilla; DS, dermal sheath. (B) Conformation of differential expression for selected genes upregulated in the DP. smooth muscle actin (AMSA), NCAM, and -actin were used as hair follicle mesenchyme and loading controls. (C) Expression of Igfbp2 in the mid-anagen vibrissa, showing both DP and epithelial staining. (D) Expression of ID3 in the mid-anagen vibrissa is restricted to the follicular mesenchyme and is increased in the tip of the DP. (E) Vimentin expression in the mid-anagen vibrissa. Scale bar, 100 m.

Full figure and legend (92K)

Over 50% of the induced in the anagen DP encoded-secreted factors. Protease nexin 1 (^{Yu et al, 1995}), a well-characterized DP marker, was induced as well as the fasciclin-like protein periostin, the SPARC-like gene (SC1/Ecm2), fibulin 5, fibronectin1, and versican, which are all involved in cell adhesion (^{Adams and Watt 1990;du Cros et al, 1995;Girard and Springer, 1995;Gillan et al, 2002;Schiemann et al, 2002}). Interestingly, the rat orthologue of Xenopus Wise, a modulator of WNT signalling (^{Itasaki et al, 2003}) was the most highly induced gene in the DP.

Other DP-induced genes modulated the major signalling pathways active in the hair follicle. The procollagenase-C-proteinase (Pcolce) enhancer protein enhances the ability of the tolloid-like protein BMP1 to inactivate BMP signalling (^{Hulmes et al, 1997}). The latent TGF-binding protein 1 (Ltbp1) binds TGF, inactivating it in the latent-associated complex (^{Taipale et al, 1994}), whereas the insulin growth factor-binding protein 2 (Igfbp2) performs a similar role with the insulin-like growth factors (^{Zhou and Bondy, 1992}). Interestingly, various members of the inhibitor of DNA-binding (Id) gene family were also induced in the DP.

Over 1300 genes were determined to be present in DP and DS but were not differentially expressed (Table S2). Although the vast majority of these genes were housekeeping/metabolic control genes, a number of them encoded proteins that are modulated by the DP-induced genes. For example, the insulin growth factors (IGF) 1 and 2 were expressed in both DP and DS. Some genes were notably absent, for example, the BMP inhibitor Noggin was present on the chip, but not detected. WNT 5a and 4 were, however, not present on the array, but we have detected expression in the DP and DS by RT-PCR previously (O'Shaughnessy et al, unpublished). Also, the well-described markers of follicular dermis, alkaline phosphatase, and smooth muscle actin were present in both the DP and DS. RT-PCR analysis confirmed differential expression of DP-induced genes (Figure 1b). Expression patterns of known genes that were DP induced or present in both DP and DS were compared. Igfbp2 showed expression in both the matrix cells and the DP but not the DS (Figure 1c), showing that expression in the DP did not preclude expression in the epithelial portion of the hair follicle. In contrast, ID3 was expressed at a higher level in the tip of the DP (Figure 1d). Vimentin, a non-differentially expressed gene was clearly restricted to the follicular dermis but was not differentially expressed (Figure 1e).

Interaction-dependent expression of a subset of genes in DP cells

To test whether the DP–matrix interaction was required for the continued expression of the induced genes at anagen, we performed RT-PCR analysis of serially passaged DP cells with the hypothesis that culture of the DP would eliminate the expression of the gene of interest if it was dependent on factors from the adjacent matrix cells. Furthermore, we would expect expression to be maintained throughout the initial explant stage, where the cultured DP cells still had access to the factors produced by the matrix cells, and then be eliminated after the first passage, where trypsin treatment would degrade any extracellular components in the culture system (Figure 2a). Consistent with this model, Wise expression was maintained during explant and then rapidly declined in subsequent passages, whereas expression of another gene induced in the DP during anagen, the Pcolce, was maintained over two passages (Figure 2b).

Figure 2.

The dermal papilla (DP) explant culture model and RT-PCR of DP cells. (A) Rationale behind the DP cell culture model. Gray scale denotes the degree of interaction-dependent gene expression. The anagen DP in vivo have access to the matrix keratinocytes and the factors produced by these cells (arrows). Upon dissection and initial explant, the fresh DP structures still contain these matrix-derived factors and the explanted cells have access to them, maintaining interaction-dependent expression. Upon passaging, these factors are removed by the proteases used to detach the cells, thus reducing interaction-dependent gene expression. (B) RT-PCR analysis of selected genes in explant culture and subsequent passages.

Full figure and legend (17K)

Characterization of rWise and expression of transcripts in the hair follicle

Wise is a secreted cysteine knot-containing protein (Figure 3a), the Xenopus homologue of which is able to modulate signalling by the canonical and non-canonical WNT (^{Itasaki et al, 2003}). As WNT signalling is critical to both the development and the cycling of the hair follicle, we chose this gene for further analysis. Wise transcripts were expressed in the DP, the inner root sheath and strongly in the precortex of both anagen vibrissae (Figure 3b and sense probe, Figure 3c) and pelage follicles (Figure 3d). Expression in the DP in middle anagen in both the vibrissae and pelage follicle was relatively low compared with the precortical region. DP cells, however, clearly show signal with the antisense probe compared with sense controls (Figure 3e and f). A C-terminal Myc-tagged Wise was expressed in a perinuclear vesicular pattern characteristic of secreted proteins (Figure 3g), and was detected in the insoluble fraction of the Triton X-100 (TX) lysate by western blot, together providing evidence that the protein was present in secretory vesicles (Figure 3g). Interestingly, the protein was not detected in the supernatant. Possibly, this could be due to the loss of the epitope by post-translational modification, or, in common with a number of WNT signalling involved proteins (^{Willert et al, 2003}), Wise was not freely soluble.

Figure 3.

Primary sequence and expression of rWise. (A) Primary sequence of rat Wise, the black underline is the secretion motif, the black box is the cysteine knot region, and the gray box is the region to which the AK-1 antibody was raised. (B) Localization of Wise transcripts in the anagen vibrissa. O, outer root sheath; i, inner root sheath; dp, dermal papilla; Hs, hair shaft. (C) Sense control, m, melanin pigment. (D) Detail of dermal papilla cells from anti-sense (AS) and sense (S) hybridizations. (E) Localization of Wise transcripts in the anagen pelage follicle. ors, outer root sheath; ds, dermal sheath. In all cases, the dermal papilla is specified by dotted lines. (F) Expression of Myc-Tagged rWise in 3T3 cells. (G) Western blots of Triton X-100 soluble (s) and insoluble (i) fractions of lysates from WISE- and vector-transfected 3T3 cells incubated with either anti-myc or the AK-1 antibody. IVT, in vitro translation product of rWise RNA. Scale bars, 100 m (B–D) and 10 m (E, F).

Full figure and legend (109K)

An antibody was raised to the C-terminal 20 amino acids of Wise, AK1, and used to probe total lysates from untransfected cells. We detected proteins of 26 and 18 kDa, a similar size to Xenopus Wise with the 26 kDa protein corresponding to the Wise-A subunit of the Xenopus protein and 18 kDa potentially being the result of post-translational modification (Figure 3g and^{Itasaki et al, 2003}). In agreement with this, in vitro translation of the Wise RNA, in an assay where post-translational modification was not possible, produced a single protein product of 26 kDa (Figure 3g).

rWise enhances WNT signalling and indirectly inhibits BMP and TGF signalling

As the rat homologue of Wise was uncharacterized, we performed animal cap explants from Xenopus laevis embryos injected with rat Wise RNA. In caps, the pan-neural marker and WNT target gene NCAM (^{Sasai et al, 1994}) was induced, as was the cement gland marker, and marker of BMP inhibition, CG-13 (^{Sasai et al, 1994}), whereas the forebrain marker Otx-2 (^{Blitz and Cho, 1995}) was not induced (Figure 4a). In whole embryos, however, injection of Wise led to a marked increase in all anterior neural markers. This weak neutralization as well as a concomitant reduction in BMP transcripts is diagnostic of WNT signalling (^{Baker et al, 1999}). To confirm this, we performed RT-PCR on the dorsal and ventral marginal zone (DMZ and VMZ, respectively) of Wise injected early gastrula embryos (Figure 4b). In whole embryos and the DMZ, there was increased expression of the WNT target Siamois (^{Sasai et al, 1995}), whereas in the VMZ and whole embryos, BMP4 transcripts were reduced. Taken together, rWise had a comparable function to the Xenopus orthologue.

Figure 4.

Molecular analysis of Wise expression in Xenopus explants and NIH3T3 cells. (A) Ectodermal explants were grown to late tailbud stages from uninjected and USAG-1-injected embryos (500 pg into each of four blastomeres). Anterior markers were assayed by RT-PCR. (B) Dorsal (DMZ) and ventral marginal zone (VMZ) explants were removed from early gastrulae injected as in (A). Expression of the WNT target Siamois and BMP4 were assayed by RT-PCR. (C–F) Indirect immunofluoresence of Smad 5 (C, D), and Smads 2 and 3 (E, F) on Usag-transfected (C, E) and vector-transfected (D, F) NIH3T3 cells. Scale bar, 50 m. (G) Representative luciferase assay, performed in triplicate on transfected cells. Error bars are 2 standard deviation. (H) Northern blot analysis of Ltbp1 in transfected cells. -actin is the loading control.

Full figure and legend (77K)

We also investigated downstream signalling from BMP4 and TGF in Wise-transfected NIH3T3 fibroblasts. Consistent with the experiments in Xenopus, Smad 5, a BMP target, was mostly excluded from the nuclei of transfected cells (figures compared with Figure 4d) whereas expression of the TGF/activin target Smads 2 and 3 was nuclear in both vector and Wise-transfected cells; however, there was a consistent increase in cytoplasmic Smads 2 and 3 in the Wise-expressing cells (Figure 4e and f). Reporter gene transcription from the TGF responsive promoter DE-Luc (^{Germain et al, 2000}) was reduced (p<0.05, paired t test), but there was no reduction from an activin-responsive promoter, ARE-Luc (Figure 4g) in the Wise-transfected cells. Since Ltbp1 expression was induced in the anagen DP and would only inhibit TGF signalling, we tested whether this could be a possible cause of the reduction of nuclear SMAD 2/3 in the 3T3 cells. There was a marked increase in Ltbp1 transcripts compared with untransfected controls (Figure 4h).

Dynamic expression of rWise during hair-follicle morphogenesis

In the 15.0 d post-coitum (dpc) rat snout, there was no pelage hair-follicle development, but the initial stages of vibrissae development were clear as thickenings of the single-cell layer epidermis. At this stage, the expression of Wise was widespread throughout the epidermis and dermis (Figure 5a), although at this stage clear determination of a dermal condensate was not possible. When the nuclei were overlaid in a merged image (Figure 5b), almost all of the Wise staining was associated with the cells. In the developing basal lamina, between the hair placode and the presumptive dermal condensate, there was, however, clear Wise positivity (Figure 5b, arrows).

Figure 5.

Expression of Wise in embryo snout hair follicles and the epidermis. (A) Wise expression in embryonic day 15 snout epidermis showing developing vibrissa. Dotted line shows the separation between dermis and epidermis. (B) Overlay of nuclear counterstain to show cells. Arrowheads show region of Wise expression in the separation between dermis and epidermis. (C) Wise expression in an adjacent region of snout epidermis containing well-developed vibrissa follicles. Dotted lines show dermal papillae. Filled arrowheads point to the high expressing precortical regions, and the empty arrowhead points to the non-expressing outer root sheath. (D) Wise expression in day 17 snout epidermis, showing two follicles in the hair peg stage of differentiation, and a more developed follicle. Large dashed lines denote the basal lamina, and dotted lines delineate the dermal condensate and dermal papilla. (E) Wise expression in E17 epidermis. Dotted line shows basal lamina. (F) Expression of Wise in adult epidermis. Scale bar, 100 m.

Full figure and legend (171K)

In 17.0 dpc embryos, the pelage hair follicles were in various stages of development. By the hair peg stage, there was expression in the DP and the hair peg itself (Figure 5d). At later stages of development, when the DP is surrounded by the epithelium, Wise expression was maintained in both the DP and the epithelium. Later stages of hair-follicle development were represented by the 17.0 dpc vibrissa follicles, which had initiated hair-shaft formation (Figure 5c). Wise expression was reduced slightly in the DP and was markedly increased in the precortex whereas still being absent in the outer root sheath and DS (Figure 5c). Dynamic changes in the epidermis were seen between the developing embryo and the adult. In the 17.0 dpc, embryo there was high expression in both the epidermis and dermis (Figure 5e), which by the adult had reduced significantly in the epidermis, restricted to the basal layer, as well as almost completely eliminated in the dermis (Figure 5f). During the same period, there was the expression of Wise in the outermost periderm layer at 17.0 dpc, whereas in the adult, the expression appeared to be associated with the cornified envelope (Figure 5e and f).

Dynamic expression of Wise during the hair cycle

We examined the expression of Wise in the vibrissae follicle since changes in expression would be more easily determined in the larger DP. In early anagen follicles prior to hair-shaft differentiation, the number of cells in the DP was at a maximum. Furthermore, Wise expression was at its highest, with continuous expression throughout the papilla space (Figure 6a). At this time, expression in the DP was very high compared with the DS (Figure 6a, arrowheads). Wise was also expressed throughout the matrix epithelial cells in the bulb. By mid-anagen (Figure 6b), when hair-shaft formation was fully underway, Wise was expressed at a significantly lower level; in the DP, with expression being largely confined to the cells themselves. There were, however, focal areas of high expression in the DP, notably at the tip of the papilla in proximity to the precortex, which also had very high levels of expression at this time. During catagen, the smaller DP did not express Wise, and the epithelial column near to the papilla expressed low levels of Wise (Figure 6c). Magnification of the DP showed high expression in DP cells and in the intercellular spaces at early anagen (Figure 6d). Expression fell both inside and outside the cell from middle anagen (Figure 6e), falling completely by catagen (Figure 6f). The other compartment that was notably Wise high expressing was the bulge region, comprising the lower epithelium of the very early anagen hair follicle (Figure 6g) and restricted to the outer root sheath cells adjacent to the erector pili muscle in the mid/late anagen hair follicle (Figure 6h).

Figure 6.

Expression of Wise in the cycling vibrissae and pelage hair follicle. (A) Wise expression in the early anagen vibrissae. Arrowhead points to low expressing dermal sheath cells. (B) Wise expression in the mid-anagen vibrissa; arrowheads point to the high expressing regions of the precortex and the dermal papilla (DP). (C) Wise expression in the catagen vibrissae. Arrowheads point to the epithelial column. In all panels the DP is indicated by a dotted line and the red signal is the propidium iodide counterstain. Scale bar, 100 m. (D–F) Magnification of the dermal papilla region of the vibrissae. (D) Early anagen; (E) middle anagen; (F) catagen. Scale bar, 10 m. (G) Expression of Wise in the bulge in the middle anagen follicle. Bracket denotes the bulge region. (H) Expression of Wise in the bulge region near the club hair at anagen onset. Scale bar, 50 m. (I) Schematic of dynamic Wise expression during the hair cycle, gray scale denotes the level of expression.

Full figure and legend (182K)

Discussion

Gene expression and signalling pathways active in the anagen DP

In this paper, we have captured a glimpse of the specific gene expression pathways that are activated in response to interaction of the DP with the overlying matrix cells. It was therefore informative to examine the secreted factors expressed by the DP in the context of the known biology of the hair follicle. Pcolce1 was induced in the DP, and activates BMP1 (^{Hulmes et al, 1997}), the higher vertebrate homologue of the BMP inhibitor tolloid (^{Suzuki et al, 1996}). As Pcolce1 expression was maintained in culture, it is likely that inhibition of BMP signalling is an intrinsic property of DP cells. Ltbp1 was similarly upregulated in the DP during anagen, consistent with the role of TGF1 in the hair follicle of inducing catagen (^{Foitzik et al, 2000}). Ltbp1 may sequester any TGF into the latency-associated peptide, and in so doing prolong anagen. Paradoxically, Igfbp2, however, functions by binding IGF1, which has been reported to prolong anagen (^{Rudman et al, 1997}). This suggests there is a balance of opposing signalling pathways that control the length of the anagen phase of the hair cycle.

Many extracellular matrix (ECM) proteins are specifically induced in the DP. It has been proposed that one of the roles of the DP is the production of a specialized ECM (^{Matsuzaki and Yoshizato, 1998}). Our analyses clearly show this is the case, with increased synthesis of developmentally important matrix components such as the SPARC-like gene SC1, an important part of the stem cell niche in the hematopoietic system (^{Oritani and Kincade, 1998}). Fibulin 5, Osf2, versican, and fibronectin all bind integrins (^{Adams and Watt, 1990;du Cros et al, 1995;Girard and Springer, 1995;Gillan et al, 2002;Schiemann et al, 2002}), which would be important in the control of keratinocyte differentiation in the hair shaft.

One surprising finding from the array data was that many signalling molecules typically associated with the DP and hair follicle, namely BMP4, IGF1 and 2, and members of the WNT family (4 and 5a), were expressed in both the DP and DS, whereas it was the modulators of these pathways that were expressed in an interaction-dependent fashion in the anagen DP. It may be that the delicate fine-tuning of the signalling pathways that is necessary during the hair cycle is more easily achieved by varying the concentration of the modulators, rather than altering the absolute levels of the signalling protein itself.

Wise function in the DP, hair follicle, and interfollicular epidermis

Wise binds the WNT co-receptor, LRP6, and can either repress or activate WNT signalling (^{Itasaki et al, 2003}). Our experiments seem to show a similar function for rat Wise. From these experiments, it is, however, not clear whether in the hair follicle, Wise plays a repressive or activatory role in the hair follicle. In the topgal transgenic mouse (^{DasGupta and Fuchs, 1999}), the -catenin co-transcription factor LEF1 is active at two distinct phases in the hair-follicle cycle, firstly during bulge activation in telogen, and in the precortex in anagen. The expression of Wise we observe in the precortex region exactly agrees to the zone of LEF1 activation in the topgal mouse. Expression in the DP can be explained if we invoke the more recent WNT-signalling reporter mouse described by^{Maretto et al (2003)}. Here reporter gene activity in the anagen hair follicle was widespread in the DP and the matrix cells, which was exactly what we observed with Wise expression. LEF1 is unlikely, based on the above findings, to play a role in activating WNT signalling in the DP; however, there is evidence that other Tcf family members and -catenin are expressed in the DP (^{Ridanpaa et al, 2001}). Cultured DP cells that express WNT 4 and 5a, but are unable to induce follicle induction, do not express Wise, and hence cannot activate WNT signalling, providing a potential explanation for the recent results of^{Shimizu and Morgan (2004)}, who found that adding WNT to long-term cultured DP cells is unable to rescue their inductive ability.

The downgrowth of the developing follicle is the result of concomitant WNT signalling and a reduction in BMP signalling (^{Jamora et al, 2003}). This is exactly the effect of rWise expression in the Xenopus, so by extension, Wise is likely to have the same effect in the developing hair follicle, where expression is seen throughout the hair peg. Expression of Wise is widespread in the dermis and epidermis during development, which may be necessary for signalling by large number of WNT family members in the dermis or epidermis to occur during late embryonic development (^{Reddy et al, 2001}). If this is also the case in the adult, then in the follicle, WNT signalling is concentrated in the DP and the Matrix at different times, with WNT signalling in the DP being at a maximum just after the initial onset of anagen, and WNT signalling in the matrix and precortex at a maximum in middle anagen. WNT signalling is virtually absent at catagen. The time-lag in Wise expression in the DP, compared with the precortex, may be a necessary component of the timing of the hair cycle, particularly if, as was the case with Ltbp1, Wise expression caused some, if not all, of the anagen DP gene-specific changes we identified in the microarray screen. Expression of Wise in the bulge region in early anagen argues that Bulge activation causes Wise expression in the downgrowing hair follicle, which activates DP gene expression, which produces a signalling environment that allows for gene expression changes in the matrix to occur to allow hair growth. By this time in the hair cycle, Wise expression has fallen in the DP and hence, as a result, the gene expression changes in the precortex finish, preventing hair-shaft differentiation and eventually leading to catagen (Figure 6i).

Finally Wise is clearly expressed in a non-epithelial/non-mesenchymal population, in the vessels at the base of the hair follicle, and in the basement membrane of the developing hair follicle. There is little or no evidence of a role for WNT signalling in the function of the basement membrane; however, expression of Wise in the basement membrane would be consistent, and necessary for it to function in epithelial–mesenchymal interactions.

Gene expression changes in cultured DP cells

Compared with other gene expression analyses on the DP (^{Sleeman et al, 2000}) many of the same genes were expressed in the DP and cultured DP cells. Although differences between the two analyses were reflected by the fact that there were differences in methodology (EST sequencing vs array hybridization), there were still clear differences in the biology of the in vivo DP and cultured DP cells, even after a single passage. Although there may be trivial reasons for these differences, namely, the contamination of the DP cells with matrix keratinocytes or components of the vasculature, the fact that we detect very low levels of expression of Wise in the passage 1 DP cells would argue against an ectodermal contamination. An equally valid reason would be the culture environment of the DP cells lacking crucial components that the matrix cells provide. This is consistent with the reports that the inductive ability of DP cells can be maintained through many more passages than normal by co-implanting germinative epithelial cells (^{Reynolds and Jahoda, 1996}).

Materials and Methods

Microarray hybridization

DP and DS were obtained from Brown Norway rat vibrissae (Charles River, Wilmington). Thirty of each were pooled for a single experiment. 0.1 g RNA was extracted (RNeasy kit, Qiagen, Valencia, Spain), and poly-A+ RNA was selected using the Oligotex system (Qiagen). Second-strand cDNA was synthesized using the Superscript II kit (Invitrogen, Carlsbad, New Mexico) after the RNA was annealed with a T7 promoter-poly-T primer (Genset, Evry, France). Biotin-labelled cRNA was made from this cDNA (Enzo Diagnostics, Farmingdale, New York). The whole probe was hybridized to the RG34A rat genome chip (Affymetrix, Santa Clara, California) according to the manufacturers' specifications. Three DP hybridizations and two DS hybridzations allowed six crosswise comparisons, with the DS being the baseline in all analyses. Genes that were tagged as present and increased in all six analyses with a p-value of less than or equal to 0.01 by Mann–Whitney analysis, and two or more highly expressed in the DP in absolute terms compared with the DS, were regarded as differentially expressed.

RT-PCR analysis of vibrissal tissue

Differential expression was confirmed with another set of DP and DS by RT-PCR and comparison with loading controls. Comparisons were made between the DP and the matrix component liberated after collagen capsule inversion by RT-PCR. DP used in RT-PCR analysis of Wise expression were explant cultured undisturbed for 11 d, after which they could be serially passaged. Primers were as follows: NCAM F: ATTGTCACCATCATGGGGCT, R: GTGAGCTGCCTTGGATTTTC; -smooth muscle actin F: TCATTGGGATGGAGTCAGCG, R: CAGCTTCGTCGTATTCCTGT; -actin F: GTGTGACGTTGACATCCGTA, R: ACTCATCGTACTCCTGCTTG; rWise F: CATGCTTCCTCCTGCCATTCA, R: GGCTCCAGTACTTTGTTCC; protease nexin 1 F: CATCATCCCTCACATCAGTAC, R: GTCACTACTGCGGCTTTGGTC; SC1 F: ACCACCCCATTGAACTTCTC, R: CAAAGAAGCGCGTTATGCAG;CD81 F: ATGGGGGTGGAGGGCTGC, R: GTACACGGAGCTGTTCCGG; SPARC F: CCGAGAGTTCCCAGCATCAT, R: AGCTTGTGGCCCTTCTTGGT.

Immunofluorescence analysis rat embryo snout and vibrissa

Throughout this study, all experiments on animal tissue were performed after institutional approval by Columbia University, and Institutional and national guides for the care and use of laboratory animals were followed. Tissues were embedded in OCT and 8 m sections were fixed in 4% paraformaldehyde in phosphate-buffered saline (PBS), before blocking and permeabilizing in 0.2% fish skin gelatin (Sigma, St Louis, Missouri), 0.4% TX in PBS. Incubations of primary and secondary antibodies were performed in the same solution. The following antibodies and dilutions were used; Goat anti-Igfbp2 (Santa Cruz Biotechnology, Santa Cruz, California), 1/50; rabbit anti-ID3 (Santa Cruz Biotechnology), 1/100 and goat anti-vimentin (Santa Cruz Biotechnology), 1/100; AK-1 (anti-rat Wise), 1/5; swine anti-rabbit (Sigma) 1/100. Donkey anti-goat (Santa Cruz Biotechnology) 1/250. Images were taken using a Zeiss Axiophot microscope and the Axiovision imaging system (Zeiss, Thornwood, New York). Slides were counterstained in a 5 ng per mL solution of propidium iodide in PBS.

In situ hybridization

Sections were fixed as described above. Digoxygenin-labelled RNA probes were synthesized using the RNA synthesis kit (Roche, Basel, Switzerland) from the Wise PCR product cloned into pGemTeasy (Promega, Madison, Wisconsin). No fragmentation of probes was required. Slides were hybridized for 12 h at 60°C, before incubating with an alkaline phosphatase-conjugated sheep anti-digoxygenin antibody (1/2000, Roche) and subsequently incubated in NBT/BCIP solution (tablets from Sigma) until sufficient color development was observed. Sections were post-fixed before microscopic examination.

Analysis of Wise expression and BMP/TGF signalling in NIH3T3 cells

The full-length rat Wise cDNA was cloned into pcDNA3.1 myc-his (Invitrogen), and transfected in NIH3T3 cells using the lipofectamine plus system (Invitrogen). Cells were selected in Geneticin (G418, Invitrogen). Cells were fixed and permeabilized before antibody incubation. The following antibodies and dilutions were used; anti-SMAD 2/3, and anti-SMAD 5 goat polyclonals (Santa Cruz Biotechnology) both 1/100; anti-myc mouse monoclonal antibody (Sigma) 1/100.

For luciferase reporter assays, selected cells were additionally transfected with either DE-Luc or ARE-Luc (Both tifts from Caroline Hill, Cancer Research UK). Cells were lysed in the cell lysis buffer (Promega), according to manufacturers' instructions. Luciferase buffer (Promega) was added to the lysates and light production was measured. One microgram of total RNA from Wise and vector-transfected cells was loaded onto a formaldehyde gel and blotted. The blot was hybridized with a radiolabelled random hexamer-labelled probe of the Ltbp1 or -actin PCR product.

Xenopus embryo manipulations, injections, and RT-PCR analysis

Embryos were obtained from in vitro fertilizations, and cultured in 0.1 Barth's modified ringer solution (MMR). Animal cap explants were taken from mid-blastula embryos, and cultured until tailbud stages. Marginal zones were dissected but not cultured prior to RNA extraction. Five nanoliter injections were performed in 4% ficoll, and embryos cultured in 0.1 MMR/4% ficoll. Wise RNA was obtained by transcription using the mMessagemMachine kit (Ambion, Austin, Texas) from the T7 promoter of the cloned and tagged cDNA.

cDNA was synthesized from 10 embryos equivalent of explants or whole embryos. Each single RT-PCR reaction used a 0.5 embryo equivalent. EF1 was used as a loading control in all analyses. The following primers were previously described: EF1a, Sia (^{Blumberg et al, 1991}); NCAM, CG-13 (^{Sasai et al, 1994}); Otx2 (^{Blitz and Cho, 1995}); BMP4 (^{Dale et al, 1992}).

Western blots and in vitro translation

Protein extracts were made from NIH3T3 cells that were either untransfected, vector transfected, or Wise transfected. TX lysis buffer was used. Soluble and insoluble fractions of the TX lysates were loaded onto denaturing polyacrylamide gels and blotted. The following antibody concentration was used in western blots: anti-myc 9E10 1/1000 and AK-1 1/500. In vitro translation was performed using the full-length rWise RNA in a reticulocyte lysate with S³⁵ methionine (Amersham, Cambridge, UK). The resulting product was run on a gel, which was dried, and subjected to autoradiography.

References

1. Adams, JC, Watt, FM: Changes in keratinocyte adhesion during terminal differentiation: Reduction in fibronectin binding precedes alpha 5 beta 1 integrin loss from the cell surface. Cell 1990 63:425–435,  | Article | PubMed | ISI | ChemPort |
2. Baker, JC, Beddington, RS, Harland, RM: WNT signaling in Xenopus embryos inhibits bmp4 expression and activates neural development. Genes Dev 1999 13:3149–3159,  | Article | PubMed | ISI | ChemPort |
3. Blitz, IL, Cho, KW: Anterior neurectoderm is progressively induced during gastrulation: The role of the Xenopus homeobox gene orthodenticle. Development 1995 121:993–1004,  | PubMed | ISI | ChemPort |
4. Blumberg, B, Wright, CV, De Robertis, EM, Cho, KW: Organizer-specific homeobox genes in Xenopus laevis embryos. Science253:194–196, 1991
5. Botchkarev, VA, Botchkareva, NV, Roth, W, et al: Noggin is a mesenchymally derived stimulator of hair-follicle induction. Nat Cell Biol 1999 1:158–164, 10.1038/11078 | Article | PubMed | ISI | ChemPort |
6. Cotsarelis, G, Sun, TT, Lavker, RM: Label-retaining cells reside in the bulge area of pilosebaceous unit: Implications for follicular stem cells, hair cycle, and skin carcinogenesis. Cell 1990 61:1329–1337,  | Article | PubMed | ISI | ChemPort |
7. Dale, L, Howes, G, Price, BM, Smith, JC: Bone morphogenetic protein 4: A ventralizing factor in early Xenopus development. Development 1992 115:573–585,  | PubMed | ISI | ChemPort |
8. DasGupta, R, Fuchs, E: Multiple roles for activated LEF/TCF transcription complexes during hair follicle development and differentiation. Development 1999 126:4557–4568,  | PubMed | ISI | ChemPort |
9. du Cros, DL, LeBaron, RG, Couchman, JR: Association of versican with dermal matrices and its potential role in hair follicle development and cycling. J Invest Dermatol 1995 105:426–431, 10.1111/1523-1747.ep12321131 | Article | PubMed | ChemPort |
10. Foitzik, K, Lindner, G, Mueller-Roever, S, et al: Control of murine hair follicle regression (catagen) by TGF-beta1 in vivo. FASEB J 2000 14:752–760,  | PubMed | ISI | ChemPort |
11. Germain, S, Howell, M, Esslemont, GM, Hill, CS: Homeodomain and winged-helix transcription factors recruit activated Smads to distinct promoter elements via a common Smad interaction motif. Genes Dev 2000 14:435–451,  | PubMed | ISI | ChemPort |
12. Gillan, L, Matei, D, Fishman, DA, Gerbin, CS, Karlan, BY, Chang, DD: Periostin secreted by epithelial ovarian carcinoma is a ligand for alpha(V)beta(3) and alpha(V)beta(5) integrins and promotes cell motility. Cancer Res 2002 62:5358–5364,  | PubMed | ISI | ChemPort |
13. Girard, JP, Springer, TA: Cloning from purified high endothelial venule cells of hevin, a close relative of the antiadhesive extracellular matrix protein SPARC. Immunity 1995 2:113–123, 10.1016/1074-7613(95)90083-7 | Article | PubMed | ISI | ChemPort |
14. Huelsken, J, Vogel, R, Erdmann, B, Cotsarelis, G, Birchmeier, W: Beta-Catenin controls hair follicle morphogenesis and stem cell differentiation in the skin. Cell 2001 105:533–545,  | Article | PubMed | ISI | ChemPort |
15. Hulmes, DJ, Mould, AP, Kessler, E: The CUB domains of procollagen C-proteinase enhancer control collagen assembly solely by their effect on procollagen C-proteinase/bone morphogenetic protein-1. Matrix Biol 1997 16:41–45, 10.1016/S0945-053X(97)90115-3 | Article | PubMed | ISI | ChemPort |
16. Itasaki, N, Jones, CM, Mercurio, S, Rowe, A, Domingos, PM, Smith, JC, Krumlauf, R: Wise, a context dependent activator and inhibitor of Wnt signalling. Development 2003 130:4295–4305, 10.1242/dev.00674 | Article | PubMed | ISI | ChemPort |
17. Jahoda, CAB, Oliver, RF: Vibrissa dermal papilla cell aggregative behaviour in vivo and in vitro. J Embryol Exp Morphol 1984a 79:211–224,  | PubMed | ISI | ChemPort |
18. Jahoda, CAB, Oliver, RF: Changes in hair growth characteristics following the wounding of vibrissa follicles in the hooded rat. J Embryol Exp Morphol 1984b 83:81–93,  | ISI | ChemPort |
19. Jahoda, CAB, Reynolds, AJ: Hair follicle dermal sheath cells: Unsung participants in wound healing. Lancet 2001 358:1445–1448, 10.1016/S0140-6736(01)06532-1 | Article | PubMed | ISI | ChemPort |
20. Jamora, C, DasGupta, R, Kocieniewski, P, Fuchs, E: Links between signal transduction, transcription and adhesion in epithelial bud development. Nature 2003 422:317–322, 10.1038/nature01458 | Article | PubMed | ISI | ChemPort |
21. Jones, CM, Smith, JC: Establishment of a BMP-4 morphogen gradient by long-range inhibition. Dev Biol 1998 194:12–17, 10.1006/dbio.1997.8752 | Article | PubMed | ISI | ChemPort |
22. Kishimoto, J, Burgeson, RE, Morgan, BA: WNT signaling maintains the hair-inducing activity of the dermal papilla. Genes Dev 2000 14:1181–1185,  | PubMed | ISI | ChemPort |
23. Kulessa, H, Turk, G, Hogan, BL: Inhibition of Bmp signaling affects growth and differentiation in the anagen hair follicle. EMBO J 2000 19:6664–6674, 10.1093/emboj/19.24.6664 | Article | PubMed | ISI | ChemPort |
24. Maretto, S, Cordenonsi, M, Dupont, S, et al: Mapping WNT/beta-catenin signaling during mouse development and in colorectal tumors. Proc Natl Acad Sci USA 2003 100:3299–3304, 10.1073/pnas.0434590100 | Article | PubMed | ChemPort |
25. Matsuzaki, T, Yoshizato, K: Role of hair papilla cells on induction and regeneration processes of hair follicles. Wound Repair Regen 1998 6:524–530, 10.1046/j.1524-475X.1998.60605.x | Article | PubMed | ChemPort |
26. Niemann, C, Owens, DM, Hulsken, J, Birchmeier, W, Watt, FM: Expression of DeltaNLef1 in mouse epidermis results in differentiation of hair follicles into squamous epidermal cysts and formation of skin tumours. Development 2002 129:95–109,  | PubMed | ISI | ChemPort |
27. Oritani, K, Kincade, PW: Lymphopoiesis and matrix glycoprotein SC1/ECM2. Leuk Lymphoma 1998 32:1–7,  | PubMed | ISI | ChemPort |
28. Paus, R: Principles of hair cycle control. J Dermatol 1998 25:793–802,  | PubMed | ChemPort |
29. Reddy, S, Andl, T, Bagasra, A, Lu, MM, Epstein, DJ, Morrisey, EE, Millar, SE: Characterization of Wnt gene expression in developing and postnatal hair follicles and identification of Wnt5a as a target of Sonic hedgehog in hair follicle morphogenesis. Mech Dev 2001 107:69–82, 10.1016/S0925-4773(01)00452-X | Article | PubMed | ISI | ChemPort |
30. Reynolds, AJ, Jahoda, CA: Hair matrix germinative epidermal cells confer follicle-inducing capabilities on dermal sheath and high passage papilla cells. Development 1996 122:3085–3094,  | PubMed | ISI | ChemPort |
31. Ridanpaa, M, Fodde, R, Kielman, M: Dynamic expression and nuclear accumulation of beta-catenin during the development of hair follicle-derived structures. Mech Dev 2001 109:173–181, 10.1016/S0925-4773(01)00563-9 | Article | PubMed | ISI | ChemPort |
32. Rudman, SM, Philpott, MP, Thomas, GA, Kealey, T: The role of IGF-I in human skin and its appendage: Morphogen as well as mitogen?J Invest Dermatol 1997 109:770–777, 10.1111/1523-1747.ep12340934 | Article | PubMed | ISI | ChemPort |
33. Sasai, Y, Lu, B, Steinbeisser, H, De Robertis, EM: Regulation of neural induction by the Chd and Bmp-4 antagonistic patterning signals in Xenopus. Nature 1995 376:333–336, 10.1038/376333a0 | Article | PubMed | ISI | ChemPort |
34. Sasai, Y, Lu, B, Steinbeisser, H, Geissert, D, Gont, LK, De Robertis, EM: Xenopus chordin: A novel dorsalizing factor activated by organizer-specific homeobox genes. Cell 1994 79:779–790,  | Article | PubMed | ISI | ChemPort |
35. Schiemann, WP, Blobe, GC, Kalume, DE, Pandey, A, Lodish, HF: Context-specific effects of fibulin-5 (DANCE/EVEC) on cell proliferation, motility, and invasion. Fibulin-5 is induced by transforming growth factor-beta and affects protein kinase cascades. J Biol Chem 2002 277:27367–27377, 10.1074/jbc.M200148200 | Article | PubMed | ISI | ChemPort |
36. Shimizu, H, Morgan, BA: Wnt signaling through the -catenin pathway is sufficient to maintain, but not restore, anagen-phase characteristics of dermal papilla cells. J Invest dermatol 2004 122:239–245,  | Article | PubMed | ISI | ChemPort |
37. Sleeman, MA, Murison, JG, Strachan, L, et al: Gene expression in rat dermal papilla cells: Analysis of 2529 ESTs. Genomics 2000 69:214–224, 10.1006/geno.2000.6300 | Article | PubMed | ISI | ChemPort |
38. Suzuki, N, Labosky, PA, Furuta, Y, et al: Failure of ventral body wall closure in mouse embryos lacking a procollagen C-proteinase encoded by Bmp1, a mammalian gene related to Drosophila tolloid. Development 1996 122:3587–3595,  | PubMed | ISI | ChemPort |
39. Taipale, J, Miyazono, K, Heldin, CH, Keski-Oja, J: Latent transforming growth factor-beta 1 associates to fibroblast extracellular matrix via latent TGF-beta binding protein. J Cell Biol 1994 124:171–181, 10.1083/jcb.124.1.171 | Article | PubMed | ISI | ChemPort |
40. Willert, K, Brown, JD, Danenberg, E, et al: Wnt proteins are lipid-modified and can act as stem cell growth factors. Nature 2003 423:448–452, 10.1038/nature01611 | Article | PubMed | ISI | ChemPort |
41. Yu, DW, Yang, T, Sonoda, T, et al: Message of nexin 1, a serine protease inhibitor, is accumulated in the follicular papilla during anagen of the hair cycle. J Cell Sci 1995 108:3867–3874,  | PubMed | ISI | ChemPort |
42. Zhou, J, Bondy, C: Insulin-like growth factor-II and its binding proteins in placental development. Endocrinology 1992 131:1230–1240, 10.1210/en.131.3.1230 | Article | PubMed | ISI | ChemPort |

Acknowledgments

We thank C. J. Whitehouse for critical reading of the paper and Caroline Hill for DE-luc and ARE-luc. This work was funded in part by the Dermatology Foundation (R. O.) and the Steven and Michelle Kirsch Foundation (A. M. C. and C. A. B. J.).