| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
EPM project groups, Osaka R and D Laboratories, Sumitomo Electric Industries, Yokohama 244-8588, Japan; and
* Life Science Division, Lawrence Berkeley National Laboratory, Berkeley, California, USA
2Correspondence: EPM project groups, Osaka R and D Laboratories, Sumitomo Electric Industries LTD. 1, Taya-cho Sakae-ku, Yokohama 244-8588, Japan. E-mail: hirai-yohei{at}sei.co.jp
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: morphogenesis • hair cycle • organ culture assay • anagen induction • molecular dynamics
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. Studies have shown that epithelial–mesenchymal interactions control the growth, organization, and differentiation of hair follicles (2
3
4
5)
, affecting processes as diverse as hormonal regulation, energy consumption, blood vessel expansion, and immune stimulation (6
7
8
9
10
11)
. Molecules produced by dermal papilla fibroblasts have been a particular target of investigation, and these have been found to act both as positive (12
13
14
15
16)
and negative (17
, 18)
mediators of telogen-to-anagen progression. However, components of the extracellular matrix (2
, 19
20
21)
and systemic hormonal signals (22
, 23)
can also influence anagen induction, as can signaling molecules responsible for general developmental processes including bone morphogenic proteins, TGFß, Wnts, and sonic hedgehog (1
, 24
25
26
27)
.
The extracellular signaling molecule epimorphin is identical to syntaxin-2, a member of a cytoplasmic protein family involved in intracellular vesicle transport (28
29
30
31
32)
. Epimorphin/syntaxin-2 appears to be unique among members of the syntaxin family in that it can also be secreted from the cell or presented on the extracellular face of the cytoplasmic membrane, where it directs epithelial morphogenic processes in many organs, including lung (29
, 33)
, liver (34
, 35)
, skin (36)
, gallbladder (37)
, pancreas (38)
, intestine (39)
, and the mammary gland (40
, 41)
. In mammary epithelial cells, epimorphin was found to act through induction of the transcription factor CCAAT/enhancer binding protein-ß (C/EBPß) (41)
, which has also been identified in studies of hair follicle cycling (42)
. A role for epimorphin in embryonic hair follicle morphogenesis was initially implicated by its tissue distribution and inhibitory effects of an anti-epimorphin monoclonal antibody (29
, 43)
. Many subsequent investigations were stymied by the poor solubility of recombinant epimorphin protein (28
, 40
, 44)
, but we and others have recently developed techniques for preparation of more soluble forms of active epimorphin (38
, 41
, 45)
.
In most rodent strains, the growth cycle of dorsal hair follicles are initially synchronized; the second telogen stage begins around day 50 and may last 50 or more days (46
47
48
49)
. We used C57BL/6 mice to characterize the role of epimorphin in anagen induction of hair follicles. Using peptide mapping, we also identified an active peptide possessing the function of the parental protein, but only when modified with particular sulfhydryl-reactive compounds. We used the selective activity of the derivatives of the minimal epimorphin peptide as a basis for molecular dynamics studies to identify the structural determinants of anagen induction by epimorphin. These studies provide insight into a mesenchymal–epithelial signaling pathway relevant to hair follicle physiology that also possesses applications in the clinic.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, 50)
. Proteolytic fragments of H1 were obtained by treatment at 1 mg/mL with trypsin (10 µg/mL) or V8 protease (10 µg/mL) (Takara). After 12 h incubation, the proteases were inactivated at 100°C, and reactions were assessed by electrophoresis and HPLC. All synthetic peptides (Bex Corporation, Tokyo, Japan; Peptide Institute, Inc., Osaka, Japan) were dissolved in PBS and sterile filtered. The cysteine-containing peptides were allowed to self-dimerize or were sulfhydryl-blocked by reaction with N-ethylmaleimide (NEM), bismaleimidehexane (BMH), or the water-soluble BMH functional analog 1,8-bismaleimidotriethyleneglycol (BM[PEO]3) (Pierce, Rockford, IL, USA), then dialyzed against PBS. The products were analyzed by gel filtration chromatography using Superdex Peptide Pc 3.2/30 column (Smart Systems, Amersham Biosciences, Piscataway, NJ, USA). Differing relative concentrations of cross-linking agents resulted in different proportions of dimeric and monomeric products. The ratio of the modified monomer to the modified dimer in the product reaction could be roughly controlled by the relative concentration of the cross-linking agent; when the cross-linker was twofold more abundant than the peptide, the yield of monomer was >70%; if the cross-linker was present at 0.5x the molar concentration of the peptide, the dimeric form was the primary product (>90%).
Western blot and immunohistochemistry
SDS-PAGE and Western blot analyses were carried out as described previously (28)
. Primary antibodies were polyclonal rabbit anti-epimorphin (28)
, used at 3 µg/mL, anti-C/EBPß gene products (Santa Cruz Biochemistry, Santa Cruz, CA; 1 µg/mL), and monoclonal antibody 27 (mAb27; 10 µg/mL). Secondary antibodies were peroxidase-linked anti-rabbit, mouse, and rat antibodies (Amersham, Little Chalfont, UK; 1/1000 dilution). Tissue localization experiments were as described previously (51)
using 5 µg/mL of biotinylated epimorphin fragments (or 2 µg/mL of biotinylated peptides) or 10 µg/mL mAb27, followed by FITC-labeled streptavidin or FITC-labeled anti-rat antibodies (both Amersham, used at 1/200 dilution). For localization of epimorphin, polyclonal antibodies were prepared in rats using biotin-labeled peptides pep 7*(I) and (II) immobilized to streptavidin Sepharose beads (Amersham/Pharmacia) as the immunogen. The specificity of these antibodies was tested by immunoprecipitation of epimorphin-containing cell lysates (40)
. These polyclonal antibodies were used at 1/300 for immunohistochemistry.
Isolation and identification of a specific marker to the growing hair follicles
To generate mAbs specific to the anagen developmental stage, surgically collected anagen hair follicles and anagen dorsal skin from C57BL/6 mice were treated with an extraction buffer containing 8M urea, 100 mM DTT, and 2% SDS at 37°C for 12 h. The extract was dialyzed against PBS, mixed with complete adjuvant (Cappel, Cochranville, PA, USA), and injected into 8-wk-old Lewis rats. Immunizations were repeated for 2 months, then splenocytes were isolated and fused with P3U1 as described (29)
. Positive hybridomas were selected by dot blot for specificity to anagen whisker hair follicle lysates obtained from 35-day-old female C57BL mice (Japan SLC). The isolated hybridoma clone was grown in serum-free GIT medium (Wako, Kyoto, Japan) and the supernatant was precipitated with 50% saturated ammonium sulfate. The precipitated monoclonal antibody, termed mAb27, was thoroughly dialyzed against PBS. Isolation of the cDNA encoding the mAb27 antigen was accomplished by screening an ExCell expression library (Invitrogen Corp., Carlsbad, CA, USA) containing cDNA isolated from anagen dorsal hair follicles. Hybridization screening of 240,000 plaques produced 30 positive clones that overlapped a single cDNA transcript with multiple redundancy. To confirm the isolated cDNA as that of the mAb27 antigen, a fragment of the recombinant protein was generated and purified as His-tagged form (20 kDa) and injected into rats. After two immunization boosts, anti-serum collected from the rats was compared with mAb27 by Western analysis and found both antigens to be identical.
Animal assays
For ex vivo culture assay, right and left upper lip skin fragments were removed from day 12 embryos of ICR mice (n=5) and placed separately on porous membranes (Nuclepore, Clifton, NJ, USA; SN110419). Each membrane was floated on medium as described previously (51)
. After incubation for 6 days with either test or control samples, tissues on the membrane were harvested and analyzed by Western blot.
For the whole animal assay, 49- to 51-day-old female C57BL mice were purchased from Charles River, Yokohama, Japan (hair follicles had just entered the second telogen phase). Dorsal hairs were carefully removed with clippers (CLEA Japan, Tokyo, Japan); we did not use shaving razors or hair depilatory cream because these treatments can perturb the hair cycle (52)
. Mice were kept for 2 days (5 mice/cage) and all mice in the same cage received the same treatment as a group. For treatment with protein, 2–3 mm x 3 mm sections of dorsal skin were removed in a bilateral symmetric pattern and a mixture of 20 µL of 1 mg/mL sample (H1 and H12); 50 mg ointment (Macrogol 1500, Nikko, Giffu, Japan) was applied to one wound and control treatment (20 µL of 1 mg/mL H3 and 50 mg ointment) to the other. Ointment administration was carried out every 24 h until the wounds were completely healed. Right and left wounds usually healed in the same period (within 1 wk) without any considerable difference in the re-epithelialization process. After an additional 2 wk, dorsal skins of the mice were photographed. For statistical analysis, mice were treated using a nonwounding protocol. The samples (H1 and H12) were subcutaneously injected (0.1 mg protein or control in 0.1 mL PBS per day) and pigmented areas were scored for each animal (0%=0, 1–20%=1, 21–40%=2, 41–60%=3, 61–80%=4, >80%=5), since the melanogenesis of the skin can be used as an indicator of initiation of anagen (53)
. To test the response to small peptides applied by transepidermal absorption, samples (0.03 mg in 0.2 mL of 50% ethanol solution) were spread once a day; the pigmented areas were scored as above. Evaluation was carried out as double-blind tests (neither administrators nor scorers could identify the samples as test or control) at Inoue Experimental Animals, Ltd. For some experiments, dorsal skin was removed on day 40 and stained for histological markers, or prepared for Western and Northern blot analyses. For histological analyses, the paraffin sections (8 µm) were stained with hematoxylin (51)
. For Northern blot analyses, total RNA was isolated with TRIzol reagent (Gibco BRL) according to the instruction manual and 10 µg total RNA was applied to each electrophoretic lane. A cDNA clone of 0.9 kb containing a cDNA sequence for the mAb27 epitope, labeled using DIG-High Prime (Roche, Nutley, NJ, USA), was used as probe.
Molecular modeling
The homology model of H12 was constructed and visualized using the DeepView Swiss-PdbViewer v3.7 (http://www.expasy.org/spdbv). Molecular dynamics (MD) simulations were performed with HyperChem 7.0 (Hypercube) using the Amber96 force field (54)
. Molecules simulated included pep7 monomer, pep7 disulfide dimer, NEM-pep7, BMH-pep7, and BM[PEO]3-pep7. In each case, peptides were configured as 
-helices and derivatized appropriately, then solvated in periodic boundary conditions using the TIP3P water model (55)
. Initial geometry optimizations performed using the Polak-Ribiere conjugate gradient method to relieve local atomic clashes. MD simulations were run to 1 ns (using a 0.001 ps step size) and performed in duplicate from different starting conformations.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
and humans (43)
; here we used labeled epimorphin and specific anti-epimorphin antibodies to enable a more detailed examination of epimorphin expression and of the epimorphin target in the postnatal anagen stage of the hair follicle. Immunohistochemistry using a rat anti-epimorphin polyclonal antibody revealed staining in the dermal papilla (DP) and connective tissue sheath (CTS) surrounding the follicular epithelia of early and late anagen of adult mouse hair follicles (Fig. 1
a, b, i; cycle stage determined by morphology after sectioning, data not shown). The epimorphin target in hair follicles was identified by exposing hair follicle tissue sections to biotinylated epimorphin protein fragments H1 and H12. Although the spatial cellular arrangement of epithelial cell types is not as clear in dorsal hair follicles, we found that the soluble epimorphin fragment H12 bound to most of the epithelial compartments of early anagen whisker hair follicles but primarily to the inner root sheath (IRS) and the keratogenous area (KA) of late anagen follicles (Fig. 1e-g, j
). Analysis of catagen and telogen hair follicles revealed much less association with epimorphin antibody (Fig. 1d
) and H12 (Fig. 1h
), suggesting that the epimorphin signaling system is primarily active in anagen. Given that the DP and CTS regions are comprised of dense fibroblastic cell populations and are known to be key controllers of follicular cycling (56
57
58)
, these results suggested that epimorphin might act as a mesenchymal signal to induce or maintain the anagen cycle stage.
|
To test the role of epimorphin in hair follicle progression, symmetrical punches were made in dorsal skin of C57BL/6 mice and recombinant epimorphin H1 and H12 were applied to one wound site and a control treatment (H3, a carboxyl-terminal portion of epimorphin dispensable for the morphoregulatory function in many systems; refs 40
, 41
) to the other. In this system, melanogenesis is a definite indicator of anagen induction (1
, 53)
, and we found that anagen-driven melanogenesis occurred more rapidly in epimorphin-treated areas relative to controls (Fig. 2
A; no difference in wound healing was detected between the two treatment regimens). Statistical analyses were performed by introducing epimorphin using a subcutaneous injection protocol (59)
and showed that the smaller recombinant protein H1 produced a more consistent effect than the larger H12 protein (Fig. 2B
), perhaps due to more effective tissue penetration.
|
Identification of a minimal activity domain of epimorphin
Encouraged by the increased activity of the H1 fragment relative to H12, its larger counterpart, we wanted to identify the minimal peptide motif required for anagen induction. We found that digestion of H1 by different proteases had differential effects on epimorphin association with anagen hair follicle sections: trypsin-digested H1 effectively blocked association of biotinylated H1 with the IRS and KA, while V8-digested H1 lacked this activity (Fig. 3
). These results suggested that a specific peptide fragment generated by the tryptic digest could occupy the epimorphin target. Candidate peptide fragments of the tryptic digest (those not overlapped by V8-generated peptides) were chemically synthesized. Two of these peptides contained cysteine residues and were assayed as a mixture of homodimeric and monomeric forms (Fig. 4
A). In addition, the peptide fragment was prepared in a chemically modified form to prevent homodimerization through free sulfhydryls. These modifications included reaction with N-ethyl maleimide (NEM), bis-maleimidehexane (BMH), or BM[PEO]3, a water-soluble analog of BMH; BMH and BM[PEO]3 are homobifunctional cross-linking agents capable of producing a mixture of monomeric and dimeric reaction products depending on the relative ratio of peptide and cross-linker; modified peptides were found to be active for induction of anagen (Fig. 4B
).
|
|
The candidate peptides were tested using several different techniques. First, incubation of biotinylated H1 and the peptides with hair follicle tissue sections revealed that pep7, SIEQSCDQDE, if used immediately after solubilization, could block association of H1 with the IRS and KA (data not shown). Further evidence that pep7 represented the active domain of epimorphin was provided by the fact that recombinant H1 proteins lacking the pep7 sequence no longer associated with the target (data not shown).
To screen the candidate peptides for induction of anagen, we developed an organ culture assay with greater sensitivity than the whole mouse assay. First, we generated a monoclonal antibody by injecting rats with protein lysates of anagen hair follicles and screening the resultant hybridomas with the same lysates. This procedure led to the identification of mAb27, an antibody that recognized an anagen specific hair follicle protein that migrated at 
220 kDa (referred to as AHF; Fig. 5
A) and was expressed in the outer layers of anagen-stage adult hair follicles (Fig. 5B
) and throughout the embryonic follicular epithelium (data not shown). An expression library was screened to identify the mAb27 antigen and 30 clones were found to overlap a single sequence (Fig. 6
). That this sequence encoded the mAb27 antigen was verified by expressing the cDNA as a recombinant peptide (generating rAHF) and using rAHF as an immunogen; mAb27 and anti-rAHF polyclonal antisera both bound to a 220 kDa band (Fig. 5C
).
|
|
We used mAb27 and organ cultures of upper lip skin fragments from 12-day mouse embryos as the basis for an assay to evaluate the activity of the H1-derived peptides (Fig. 5D
). We found that treatment with the H1 fragment clearly up-regulated AHF and increased expression of C/EBPß gene products (Fig. 5E
). The C/EBPß transcription factors have been shown to mediate the effect of epimorphin in mammary epithelial cells (41)
. Surprisingly, we found that unmodified pep7 (present as a mixture of monomeric and dimeric forms) was much less active than some derivatized forms in inducing expression of AHF (data not shown) or for specific binding to hair follicle IRS and the KA (Fig. 4A
). pep7 modified with NEM (Fig. 5H
) showed no binding or AHF-inducing activity (data not shown). However, pep7 modified with BMH or BM[PEO]3 (but not samples containing cross-linker alone) were quite effective in the organ culture assay and showed specific binding to hair follicle IRS and the KA (Fig. 5F, G
). Apparently, reaction with the bifunctional cross-linker BMH or BM[PEO]3 produced a product with properties of the parental epimorphin molecule that were not present in the underivatized or NEM-treated pep7 samples.
Modified pep7 induces anagen when applied topically to telogen skin
Reaction of pep7 with BMH or BM[PEO]3 produced monomeric and dimeric forms, depending on the relative concentrations of peptide and cross-linker. These individual species were purified by HPLC; both showed the same binding pattern in sectioned hair follicles and were active in the organ culture assay (Fig. 4B
). The reduced molecular mass of pep7 derivatives (relative to H1) suggested that whole animal tests could be accomplished percutaneously, an approach less invasive than skin punch or subcutaneous injections. Since monomeric and dimeric forms were both found to be active in the ex vivo assay and for specific binding to the IRS and the KA (Fig. 5F, G
), we generated a standard preparation of pep7 reacted with cross-linker; when characterized, we found this standard preparation contained primarily a monomeric form but also some dimeric forms. This preparation was preferred over HPLC-purified monomeric or dimeric forms of pep7 derivatives because sufficient material could be generated for animal experiments.
We applied this modified pep7 preparation, referred to as EPMp7, topically to mouse skin to assess responsiveness to the modified peptide in vivo. The monomeric component of EPMp7 is 1.4 kDa; although many compounds >500 Da are unable to penetrate the epidermal barrier (60)
, the trans-appendageal pathway can allow larger molecules to reach the hair follicular components (61)
. For example, cyclosporin A, an 11 amino acid peptide, can dramatically influence the hair cycle when applied topically (7)
. In our studies, EPMp7 (100 µg/mL dissolved in 50% ethanol) or control vehicle was spread onto mouse dorsal skin every 24 h. This experiment was conducted three times, with 5 or 10 mice in each test or control group; in every experiment, EPMp7 treatments accelerated anagen-dependent pigmentation and subsequent induction of visible hair growth (Fig. 7
A, B). Histological examinations showed that epidermal and dermal layers in the skin treated with EPMp7 were thicker than the control and contained many growing hair follicles (Fig. 7Ac, d
). Up-regulation of AHF is apparent in mice treated with EPMp7 (Fig. 7C
).
|
Molecular modeling of epimorphin
The differential activity of BMH/BM[PEO]3-reacted pep7 compared with pep7 reacted with NEM or to native pep7 presented a puzzle: why should these seemingly similar molecules show such divergent properties? To address this issue, we performed a series of molecular modeling and dynamics studies. We generated a homology model of H12. This region of epimorphin is highly similar (>60% amino acid identity) to a fragment of syntaxin-1A that has been the subject of several crystallographic and NMR structural studies (62
63
64)
. The structures depict an amino-terminal unstructured domain (residues 1-29), a bundle composed of three -helices (residues 30-63, 73-105, and 112-145), and a carboxyl-terminal unstructured tail (residues 145-188) (Fig. 8
A). In the resultant H12 homology model, the amino acid sequence corresponding to pep7 is located at the carboxyl-terminal end of helix b (Fig. 8B
), with Cys-101 oriented toward the center of the three-helix bundle (Fig. 8C
). The buried position of Cys-101 suggests that it does not participate in a disulfide linkage in the native protein, but may play a structural role by interacting noncovalently with one or both of helices a and c. We hypothesized that the difference between the active and inactive forms of pep7 might therefore be due to conformational differences induced by interaction of the peptide with the derivatizing agent. Consistent with this hypothesis, molecular dynamics simulations of monomeric pep7 derivatized with BM[PEO]3 or BMH suggested that the long chain of the derivatizing agent could adopt a position parallel to the amino-terminal half of the pep7 helix and that the terminal carboxylic acid of the modifying chain formed a salt bridge with the terminal amino group of the peptide (Fig. 8D
). In this conformation, the modifying chain packs against the peptide helix, burying a portion of the same helical surface that is in contact with the other two helices in the parental H12 molecule. Accordingly, we propose that the intramolecular interaction between pep7 and BMH or BM[PEO]3 could stabilize a native-like structure of the pep7 amino-terminal region, shifting the conformational equilibrium to favor the biologically active epitope. In support of the idea that the amino-terminal region of pep7 is responsible for the biological activity, preliminary mutagenesis experiments suggest that the three carboxyl-terminal amino acids of pep7 can be substituted without affecting induction of anagen (data not shown).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. Six distinct mechanisms involved in epithelial–mesenchymal interactions have been linked to hair follicle cycling (1)
: fibroblast growth factor, transforming growth factor-ß, sonic hedgehog, wnts, neurotrophins, and homeobox-involved pathways. To these six, we now add a seventh: epimorphin, a morphogen produced by follicular mesenchymal cells that binds to follicular epithelial cells (Figs. 1
and 5)
. Epimorphin has been most extensively characterized in mammary gland morphogenesis, where it is produced by stromal fibroblasts and myoepithelial cells, and stimulates branching and luminal morphogenesis of luminal epithelial cells (40)
. The mechanisms involved in these processes are still under investigation, but are known to involved intracellular signaling through the transcription factor C/EBPß (41)
and extracellular production of matrix-metalloproteinases (MMPs) (45)
. It is interesting that the IRS, a principal epimorphin binding site in the anagen hair follicle (Figs. 1
and 5)
, has been found to express C/EBPß and MMP-9 during anagen (42
, 65)
and that we detected up-regulation of C/EBPß in anagen hair follicles and in epimorphin-treated embryonic skin (Fig. 5A, E
). Such similarities may not be unexpected given the close physiological relationship between hair follicles and the mammary gland (66
, 67)
. However, epimorphin has also been implicated in morphogenic processes in lung, kidney, and pancreas, and the commonalities in mechanisms of epimorphin function in skin and mammary gland suggest that similar mechanisms may exist in these organs (29
, 30
, 38
, 44)
. Epimorphin might not trigger the proliferation of germinal matrix cells, because this molecule is not abundant in DP compared with CTS with relatively weak binding property to these matrix epithelia (Fig. 1)
.
Our identification of AHF as an anagen-specific hair follicle protein induced by epimorphin in our ex vivo assay allowed us to determine that specific derivatives of pep7 could recapitulate the activity of the larger epimorphin molecule. Sequence analysis of AHF (Fig. 6)
suggests a close relationship with trichohyalin, a major structural protein of IRS, and the medulla layer of the hair follicle (68
69
70)
. Other proteins with considerable sequence similarity to trichohyalin have been identified in the mouse (71)
, so it is possible that AHF is a distinct member of an emerging family of trichohyalin-like proteins. Preliminary experiments suggest that some of the factors known to induce anagen in vivo, including sonic hedgehog and hepatocyte growth factor, do not induce expression of AHF in the ex vivo assay (data not shown). However, follicular developmental processes in the day 12 embryo only partially reproduce anagen induction in the adult skin, so it is unclear whether expression of AHF in our assay is a selective response to epimorphin. In any case, many general models of anagen induction already exist (1)
, so a more selective assay may help dissect individual mechanisms. That trichohyalin has been implicated as a potential autoantigen in inflammatory alopecia areata (72
, 73)
suggests that the assay may also have applicability for studying broader aspects of hair follicle dysfunction.
The molecular mechanism of epimorphin induction of AHF remains to be clarified. That the expression pattern of AHF (Fig. 5B
) does not correspond to the binding pattern of epimorphin (Fig. 1e-g
) or its derivatives (Fig. 5G
) suggests that the relationship between epimorphin and AHF is likely to be indirect. One possibility is that epimorphin acts to induce AHF through the C/EBPß transcription factor. Another possibility is that spatiotemporal expression of epimorphin affects bone morphogenic proteins (BMP) and/or their inhibitor, Noggin, as the former is critical for cyto-differentiation of hair shaft cells and the latter for follicular outgrowth (74)
. This latter hypothesis is especially attractive because BMP4 has been shown to be a mediator for epimorphin action in another intestinal epithelial cells (39)
, and recombinant BMP4 led to up-regulation of AHF in our organ culture assay (data not shown).
The hair follicle is an organ that illustrates many general developmental properties, including differentiation, epithelial–mesenchymal signaling, pattern formation, cell and organ growth cycles, and tissue regression and regeneration (1)
; our studies suggest that epimorphin may play a role in many of these processes in hair follicles and the mammary gland. Accordingly, identification of a minimal structural determinant of epimorphin capable of inducing hair follicle cycling has implications beyond hair follicle biology and the obvious applications to the clinic. Small molecules have been used extensively as probes of biological processes, and our molecular dynamics studies suggest that even smaller active derivatives of pep7 might be constructed. Indeed, recent optimization of pep7 led to a construct that is active in anagen induction at lower than 1 µg/mL (data not shown). Such molecules could be especially useful for identifying the spatiotemporal expression of the epimorphin target in live animals, information relevant to hair follicle biology, and potentially for developmental processes in other organs where epimorphin signaling has been implicated, including mammary gland, lung, pancreas, and kidney. Investigations into this possibility are currently underway.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Received for publication April 30, 2003. Accepted for publication June 26, 2003.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  Y. Hirai, M. J. Bissell, and D. C. Radisky Extracellular localization of epimorphin/syntaxin-2 Blood, October 15, 2007; 110(8): 3082 - 3082. [Full Text] [PDF]
|
|
|
|
 |
|
 A. Fadloun, D. Kobi, J.-C. Pointud, A. K. Indra, M. Teletin, C. Bole-Feysot, B. Testoni, R. Mantovani, D. Metzger, G. Mengus, et al. The TFIID subunit TAF4 regulates keratinocyte proliferation and has cell-autonomous and non-cell-autonomous tumour suppressor activity in mouse epidermis Development, August 15, 2007; 134(16): 2947 - 2958. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. Hirai, C. M. Nelson, K. Yamazaki, K. Takebe, J. Przybylo, B. Madden, and D. C. Radisky Non-classical export of epimorphin and its adhesion to {alpha}v-integrin in regulation of epithelial morphogenesis J. Cell Sci., June 15, 2007; 120(12): 2032 - 2043. [Abstract] [Full Text] [PDF]
|
|
|
|
), and dimeric pep7B(II) (when pep7/BMH
2). These separate species were purified by gel filtration chromatography (left). Both species showed a binding pattern similar to H1 and H12 (right top) and were found to be active in the anagen hair follicle assay, as shown by induction of AHF protein (bottom, Western blot of AHF induced by pep7B(I) and pep7B(II) compared with PBS).
