Allan J. Pearson

Prolactin and Its Receptor Are Expressed in Murine Hair Follicle Epithelium, Show Hair Cycle-Dependent Expression, and Induce Catagen

Here, we provide the first study of prolactin (PRL) and prolactin receptor (PRLR) expression during the nonseasonal murine hair cycle, which is, in contrast to sheep, comparable with the human scalp and report that both PRL and PRLR are stringently restricted to the hair follicle epithelium and are strongly hair cycle-dependent. In addition we show that PRL exerts functional effects on anagen hair follicles in murine skin organ culture by down-regulation of proliferation in follicular keratinocytes. In telogen follicles, PRL-like immunoreactivity was detected in outer root sheath (ORS) keratinocytes. During early anagen (III to IV), the developing inner root sheath (IRS) and the surrounding ORS were positive for PRL. In later anagen stages, PRL could be detected in the proximal IRS and the inner layer of the ORS. The regressing (catagen) follicle showed a strong expression of PRL in the proximal ORS. In early anagen, PRLR immunoreactivity occurred in the distal part of the ORS around the developing IRS, and subsequently to a restricted area of the more distal ORS during later anagen stages and during early catagen. The dermal papilla (DP) stayed negative for both PRL and PRLR throughout the cycle. Telogen follicles showed only a very weak PRLR staining of ORS keratinocytes. The long-form PRLR transcript was shown by real-time polymerase chain reaction to be transiently down-regulated during early anagen, whereas PRL transcripts were up-regulated during mid anagen. Addition of PRL (400 ng/ml) to anagen hair follicles in murine skin organ culture for 72 hours induced premature catagen development in vitro along with a decline in the number of proliferating hair bulb keratinocytes. These data support the intriguing concept that PRL is generated locally in the hair follicle epithelium and acts directly in an autocrine or paracrine manner to modulate the hair cycle.

Expression patterns of keratin intermediate filament and keratin associated protein genes in wool follicles

The catalogue of hair keratin intermediate filaments (KIFs) and keratin-associated proteins (KAPs) present in wool follicles is incomplete. The full coding sequences for three novel sheep KIFs (KRT27, KRT35 and KRT38) and one KAP (KRTAP4-3) were established in this study. Spatial expression patterns of these and other genes (KRT31, KRT85, KRTAP6-1 and trichohyalin) were determined by in situ hybridisation in wool follicles at synchronised stages of growth. Transcription proceeded in the order: trichohyalin, KRT27, KRT85, KRT35, KRT31, KRT38, KRTAP6-1 and KRTAP4-3, as determined by increasing distance of their expression zones from the germinal matrix in anagen follicles. Expression became gradually more restricted to the lower follicle during follicle regression (catagen), and ceased during dormancy (telogen). Some genes (KRT27, KRT31, KRT85 and KRTAP6-1), but not others, were expressed in cortical cells forming the brush-end, indicating specific requirements for the formation of this anchoring structure. The resumption of keratin expression was observed only in later stages of follicle reactivation (proanagen). KIF expression patterns in primary wool follicles showed general resemblance to their human homologues but with some unique features. Consistent differences in localisation between primary and secondary wool follicles were observed. Asymmetrical expression of KRT27, KRT31, KRT35, KRT85 and trichohyalin genes in secondary follicles were associated with bulb deflection and follicle curvature, suggesting a role in the determination of follicle and fibre morphology.

Annotation of sheep keratin intermediate filament genes and their patterns of expression

Abstract:  Keratin IF (KRT) and keratin‐associated protein genes encode the majority of wool and hair proteins. We have identified cDNA sequences representing nine novel sheep KRT genes, increasing the known active genes from eight to 17, a number comparable to that in the human. However, the absence of KRT37 in the type I family and the discovery of type II KRT87 in sheep exemplify species‐specific compositional differences in hair KRT genes. Phylogenetic analysis of hair KRT genes within type I and type II families in the sheep, cattle and human genomes revealed a high degree of consistency in their sequence conservation and grouping. However, there were differences in the fibre compartmentalisation and keratinisation zones for the expression of six ovine KRT genes compared with their human orthologs. Transcripts of three genes (KRT40, KRT82 and KRT84) were only present in the fibre cuticle. KRT32, KRT35 and KRT85 were expressed in both the cuticle and the fibre cortex. The remaining 11 genes (KRT31, KRT33A, KRT33B, KRT34, KRT36, KRT38‐39, KRT81, KRT83 and KRT86‐87) were expressed only in the cortex. Species‐specific differences in the expressed keratin gene sets, their relative expression levels and compartmentalisation are discussed in the context of their underlying roles in wool and hair developmental programmes and the distinctive characteristics of the fibres produced.

Localisation of insulin-like growth factor receptors in skin follicles of sheep (Ovis aries) and changes during an induced growth cycle.

Pelage growth cycles are regulated by circulating prolactin in many mammals, but the intercellular mediators of this signaling are unknown. Binding sites for insulin-like growth factors (IGFs) were examined in sheep skin to show changes in distribution and abundance of IGF receptors associated with a prolactin stimulus and the subsequent hair follicle growth cycle. Follicle cycles were induced in New Zealand Wiltshire ewes by a surge in plasma prolactin following a 4-month period of prolactin suppression with bromocriptine. Eight treated and three control sheep were slaughtered at intervals over 43 days during the follicle growth cycle. At 12-20 days after the elevation of prolactin, wool follicles passed through brief catagen and telogen phases, followed by a return to anagen. IGF binding sites were localized in skin sections by incubation with 125I-IGF-I or 125I-IGF-II. Displacement with competitive binding inhibitors (unlabeled IGF-I, IGF-II, des(1-3)IGF-I, des(1-6)IGF-II, or insulin) and affinity cross-linking showed that these binding sites were predominantly IGF type 1 and type 2 (mannose-6-phosphate) receptors. The radioligands bound especially to follicle germinal cells and prekeratinocytes. Increases in specific binding of both radioligands were observed after the rise in prolactin, but prior to anatomical changes in follicles associated with cessation of growth. For IGF-I, highest binding density was observed during catagen in the germinal matrix and dermal papilla cells. For IGF-II, peak density occurred during late anagen/early catagen in the germinal matrix and during telogen in the dermal papilla. These cycle associated changes in receptor availability suggest that IGF receptors are involved in control of the wool growth.

The Microanatomy, Cell Replication, and Keratin Gene Expression of Hair Follicles during a Photoperiod-lnduced Growth Cycle in Sheep

Exposure of New Zealand Wiltshire sheep to long days, following 24 weeks of short days, caused a synchronised out-of-season wool follicle growth cycle. Skin biopsies were collected at intervals between 3 and 30 days and follicles were examined by light microscopy in both transverse and longitudinal section to describe the regressive (catagen), resting (telogen) and regenerative (proanagen) stages of the induced growth cycle. Follicles were generally in the growing phase (anagen) during short day treatment but by day 20 after exposure to long day photoperiod. 16% of follicles were in late catagen. By day 52, all follicles were in various stages of catagen, telogen and proanagen. The progression through the cycle occurred more slowly, but was morphologically similar to follicle growth cycles reported in rodents and goats, induced by plucking or melatonin, respectively. Follicles in early catagen were rarely observed, possibly reflecting the brevity of this phase of the cycle. Late catagen follicles were distinguished by the presence of a brush end and an inner root sheath, the latter disappearing as follicles entered telogen. Immunocytochemistry of proliferating cell nuclear antigen provided evidence that mitotic activity in the follicle bulb ceased completely during the brief telogen phase. The simultaneous absence of type I intermediate filament keratin mRNA indicated that keratinocyte differentiation had also been interrupted. Cell proliferation was re-established in early proanagen prior to observable changes in the follicle microanatomy. The relatively synchronised follicle growth cycle induced by photoperiod manipulation represents a potentially useful model for the study of changes in follicle ultrastructure and the endocrine and biochemical regulation of seasonal hair growth patterns.

An In Vitro Model for the Morphogenesis of Hair Follicle Dermal Papillae

TO THE EDITOR Hair follicle size is correlated with the size of the dermal papilla (DP) (Ibrahim and Wright, 1982; Elliott et al., 1999). DP size is dynamically regulated during the follicle growth cycle. Cells emigrate from the DP during catagen and then repopulate it in anagen (Tobin et al., 2003; Chi et al., 2010). The follicle miniaturization seen in androgenetic alopecia is thought to be driven by dysfunctional DP cell (DPC) movement, whereas hypertrichosis is caused by excessive DPC migration or proliferation, producing an abnormally large DP (Jahoda, 1998). The mechanisms that determine DP size are poorly understood. An in vitro model for DP morphogenesis would facilitate their investigation. DPCs are well known to aggregate in culture, which is likely to be an expression of their morphogenetic behavior. However, DPC aggregation is variable and is typically lost after a period of culture ex vivo (Horne et al., 1986; Song et al., 2005). In contrast, we have found that ovine wool follicle DPCs exhibit particularly robust and stable aggregation. We have optimized culture conditions for these cells to establish a model for DP morphogenesis and a quantitative assay for aggregate size. Cultures of ovine DPCs were initiated by microdissecting papillae and explanting them in culture medium (Supplementary Methods online). The cells showed a fibroblastic morphology (Figure 1a), and formed whorl patterns on reaching confluence (Figure 1b). Localized variations in density then began to appear (Figure 1c and d). The highdensity patches continued to condense, eventually forming three-dimensional spheroids projecting up from the culture substrate (Figure 1e–g). The spheroids could be stained with Van Gieson’s solution (Figure 1h and i), allowing measurement of their size by image analysis. Alkaline phosphatase (AP) is expressed in the DP in vivo (Rendl et al., 2005). AP staining was observed in most ovine DPCs composing aggregates (Figure 1j–l). Versican, a proteoglycan also associated with DP formation (Kishimoto et al., 1999), was expressed only in aggregating cells (Figure 1m and p). Of the 19 cell strains from 12 sheep, 11 consistently aggregated as described above, for at least 5 passages. Among three strains (from two sheep), there was no loss of aggregative behavior before replicative senescence at 80.3–93.0 population doublings (22–27 passages, Supplementary Figure S1 online). Another 5 of the 19 strains aggregated in early passages but were not further tested. Three strains did not form discrete aggregates suitable for size determination. We established a standardized assay for quantifying the effect of bioactive compounds on aggregate size (Supplementary Methods online, Supplementary Figure S2 online). The addition of 10–30 mM lithium chloride (LiCl) induced a dose-dependent reduction in size (Figure 2a and b). Aggregation was abolished at a concentration of 40 mM. Dorsomorphin (a BMPR-1 and VEGFR-2 inhibitor) and SU5402 (a FGFR-1 and VEGFR-2 inhibitor) also reduced the aggregate size (Figure 2c and d). LiCl has pleiotropic effects on intracellular signaling, including action as a Wnt mimetic and an inhibitor of inositol phospholipid signaling (Chiu and Chuang, 2010). In light of the broad specificity of these compounds, the molecular mechanisms underlying aggregate miniaturization remain to be determined. The effects of additional compounds are shown in Supplementary Figure S3 online. The hair-loss drug, minoxidil, reversed the aggregate miniaturization induced by LiCl (Figure 2e). The effects of LiCl on aggregation persisted after it was removed from the cells (Supplementary Figure S4a online). We investigated the effect of LiCl pretreatment on the in vivo induction of hair follicles by ovine DPCs (Supplementary Figure S4b–f online). LiCl at a concentration of 40 mM blocked follicle induction in vivo, as well as aggregation in vitro (Supplementary Table S1 online). To our knowledge, a DPC culture system that permits quantitative measurement of aggregate size has not previously been reported. Although human and rodent DPCs exhibit aggregative behavior (Messenger, 1984; Horne et al., 1986; Song et al., 2005), the aggregates are less well formed, with less well-defined boundaries that preclude size measurement. Aggregation typically diminishes and then disappears as the cells continue to be propagated ex vivo. We found that robust aggregation of ovine DPCs continued after extensive growth in culture, allowing numerous assays to be performed with the same cell population. DPC aggregates expressed AP and versican, whereas monolayers did not. AP and versican are expressed in DP in vivo and are markers of follicleinducing activity (Kishimoto et al., 1999; Rendl et al., 2005). The localized expression of these markers LETTER TO THE EDITOR