TECHNICAL FIELD

The invention relates to a cell culture system. In particular, the cell culture system is for testing the behaviour of dermal papilla cells from hair follicles.

BACKGROUND ART

Hair grows from follicles embedded in the skin. Foetal development of hair follicles is mediated by a series of morphogenetic interactions between the ectoderm and underlying dermis (Hardy, 1992; Olivera-Martinez ef al., 2004). An early step in the process is the aggregation of dermal cells to form a condensate underneath an ectodermal placode. The condensate develops into a dermal papilla, a regulatory structure enclosed by the bulb of the mature, growing, hair follicle. Although epithelial tissue derived from the placode produces the hair fibre itself, the dermal component largely determines the type of follicle that develops (Sengel, 1990). When ectoderm and dermis from different body sites or species are surgically recombined, the follicles that develop are like those of the donor site for the dermis, rather than the ectoderm. The size of the mature dermal papilla is a major determinant of the size of the follicle and the hair fibre it produces (Elliott et al., 1999).

Interactions between epithelial and dermal tissues continue in adult life as follicles cycle between an active, hair producing state (called "anagen") and a quiescent phase ("telogen") (Hardy, 1992; Paus and Foitzik, 2004). As follicles regress into telogen, cells emigrate from the dermal papilla into the adjacent dermal sheath (Tobin et al., 2003a; Tobin et al., 2003b). The papilla becomes more compact and reduced in size; growth of the hair fibre stops because the papilla no longer stimulates the proliferation and differentiation of epithelial progenitor cells. As follicles re-enter the anagen phase, cells migrate from the dermal sheath back into the papilla (Tobin ef al., 2003a; Tobin et al., 2003b), a process that reiterates the initial aggregation of dermal cells in development. The papilla becomes enlarged, and the expression of paracrine signals which stimulate the epithelial cells is reactivated to begin production of a new fibre. The duration of the anagen phase determines the length of the hair fibre. Long hairs on the scalp are produced by follicles that remain in anagen for several years (Halloy ef al., 2000). The follicles that produce short, fine hairs on other body sites remain in anagen for only a few weeks (Seago and Ebling, 1985).

Disorders in human hair growth are driven by changes in follicle size (Jahoda, 1998; Whiting, 2001). In androgenic alopecia (common male-pattern baldness), follicles become miniaturised so as to produce increasingly short, fine fibres, instead of the long, thick fibres otherwise found on the scalp. Miniaturisation is associated with a reduction in papilla size and is coupled with the follicle cycling process. It is thought that miniaturisation is ultimately caused by the failure of dermal cells to migrate back into the papilla as a new anagen phase is initiated. Conversely, hypertrichosis, or excess hair growth, is caused by excessive enlargement of the papilla, follicle and hair. Controlling the size of the papilla is the key to managing hair growth disorders.

It would therefore be an advantage to have a method of isolating and culturing dermal papilla cells to provide an experimental model or assay for investigating cellular mechanisms in hair growth. Dermal papilla cells retain some ability to aggregate in culture (Inamatsu et al., 1998; Jahoda and Oliver, 1984; Messenger et a/., 1986; Rushan et al., 2007). However, previously described culture systems are not useful for this purpose because the extent of aggregation is variable and this behaviour is generally lost altogether within a few passages. The size of aggregates cannot meaningfully be quantified.

In contrast, a useful culturing system would engender the formation of well-developed, papilla- like aggregates, and so provide the advantage of quantifying the size and number of aggregates that form. It would thereby facilitate a better understanding of follicle neogenesis and hair growth disorders. The culturing system would provide the further advantage of allowing a range of compounds to be tested, to determine whether they alter papilla size. Such compounds would be useful, for example, for the treatment of human hair growth disorders or for altering animal fibre attributes.

It is an object of the present invention to address the foregoing problems or at least to provide the public with a useful choice.

All references, including any patents or patent applications cited in this specification are hereby incorporated by reference. No admission is made that any reference constitutes prior art. The discussion of the references states what their authors assert, and the applicants reserve the right to challenge the accuracy and pertinence of the cited documents.

It will be clearly understood that, although a number of prior art publications are referred to herein, this reference does not constitute an admission that any of these documents form part of the common general knowledge in the art, in New Zealand or in any other country.

It is acknowledged that the term 'comprise' may, under varying jurisdictions, be attributed with either an exclusive or an inclusive meaning. For the purpose of this specification, and unless otherwise noted, the term 'comprise' shall have an inclusive meaning - i.e. that it will be taken to mean an inclusion of not only the listed components it directly references, but also other non- specified components or elements. This rationale will also be used when the term 'comprised' or 'comprising' is used in relation to one or more steps in a method or process.

Further aspects and advantages of the present invention will become apparent from the ensuing description which is given by way of example only.

DISCLOSURE OF INVENTION

The inventors have found that ovine dermal papilla cells undergo more extensive, uniform and stable aggregation in culture compared with other species. These cells form numerous spheroid structures reminiscent of a papilla in vivo. The ovine cell culture system may be used to make discoveries concerning wool growth, and hair growth in other mammalian species, including humans. Culture conditions may also be altered to suit dermal papilla cells of other species if they are found to show persistent aggregative properties as the inventors have shown in sheep.

Throughout the specification the term "wool" should be taken as meaning the animal fibre produced by sheep which is homologous to human hair.

According to one aspect of the present invention there is provided a method of preparing a dermal papilla cell culture which assists the aggregative behaviour of the cells, including the steps of: a) collecting a wool bearing skin sample from a sheep; b) isolating a wool follicle from the surrounding dermis from the wool bearing skin sample; c) excising a dermal papilla from the isolated wool follicle; d) transferring the excised dermal papilla to a growth medium; and e) maintaining the excised dermal papilla in the growth medium to allow proliferation of dermal papilla cells.

Preferably, the skin sample in step a) is obtained from a body site which grows characteristically long, fine wool fibres, for example the lower neck region. However this should not be seen as limiting. Preferably, the sheep should be less than 18 months old.

Preferably, the skin sample is placed immediately in dissection medium and kept ice-cold.

Preferably, the dissection medium is Minimum Essential Medium (MEM) (containing Earle's salts) with 292 mg/l glutamine, 100 units/ml penicillin, 0.1 mg/ml streptomycin, 12.5 ng/ml amphotericin B, and 10% lamb serum. It will be appreciated that a person skilled in the art will know suitable methods and media for retaining cell viability following the excision of the skin samples.

It will be appreciated that a person skilled in the art will know suitable methods for sterile technique and the microdissection of the hair follicle and dermal papilla from the surrounding dermis in steps b) and c).

Preferably, step d) is conducted by transferring 6-8 dermal papillae to one 8.8 cm ² tissue culture dish containing 2 ml of growth medium. Preferably, the growth medium is: MEM with 292 mg/I glutamine, 100 units/ml penicillin, 0.1 mg/ml streptomycin, 12.5 ng/ml amphotericin B, and 20% lamb serum.

At step e), preferably the cultures are maintained at 37 ⁰ C, 5% CO ₂ and are incubated for 1 week, preferably without being disturbed. The medium may be then replenished twice-weekly.

Most preferably, the dermal papilla cells are harvested after they have proliferated to cover at least half of the culture surface.

Preferably, the cells are harvested by trypsinisation. The cells may then be seeded into a new culture vessel to allow continued proliferation. It will be appreciated that a person skilled in the art will know the suitable cell harvesting and seeding techniques.

Cells may be repeatedly harvested and seeded into new vessels (passaged) to permit ongoing proliferation.

Preferably, cells are seeded at a density of at least 10,000 cells per cm ² of growth surface.

Most preferably, the cells are allowed to become confluent and show preliminary signs of aggregation before being harvested. By this means, strains of cells may be selected for aggregative ability which is retained for at least 5 passages.

The dermal papilla cells may be stored frozen in a medium consisting of ovine serum and DMSO.

Preferably, a cryopreservation medium is 90% lamb serum and 10% DMSO. The vials containing 1 ml aliquots of cells may be stored frozen in liquid nitrogen until thawed for their subsequent use. The papilla cells may be frozen with an approximately cell density of 5,000,000 cells/ml. According to another aspect of the present invention, there is provided a method of preparing a hair growth assay comprising the steps of; a) seeding ovine dermal papilla cells into a suitable growth medium; b) incubating the cells to allow cell proliferation and show at least early stage aggregation; c) harvesting the cells; d) resuspending the cells in a suitable growth medium; e) seeding the cells into a suitable culture vessel.

Preferably, the ovine dermal papilla cells for the hair growth assay method may be obtained from the dermal papilla cell culture method substantially described above.

It will be appreciated that a person skilled in the art will know the suitable cell seeding techniques, allowing the dermal papilla cells to be seeded into a suitable growth medium in step a). The cells are thawed at 37 ⁰ C and seeded into a vessel with a suitable growth medium, according to standard mammalian cell culture techniques (Phelan, 2007).

Preferably the growth medium in step a) is MEM with 292 mg/L glutamine, 100 units/ml penicillin, 0.1 mg/ml streptomycin, 12.5 ng/ml amphotericin B, and 20% lamb serum, as described above.

Preferably, the cells in step b) are grown to confluence and show preliminary signs of aggregation when the cells begin to separate and form clusters.

Preferably, in step c) the cells are harvested and dissociated from the culture vessel and each other. Preferably, the cells are harvested by trypsinisation.

Preferably, in step d) the cells are initially resuspended at a concentration of at least 1 million cells per ml. Preferably, in step d) the cells are resuspended in MEM with 292 mg/l glutamine, 100 units/ml penicillin, 0.1 mg/ml streptomycin, 12.5 ng/ml amphotericin B, and 20% lamb serum.

Cells may then be diluted in medium such that the final seeding density will be 130,000-300,000 cells per cm ² of growth surface. Optionally, a test compound may be included in the diluent medium (see below). Preferably, the vessel in step e) is a tissue culture plate. For example, the tissue culture plate may be a 12-well plate. It will be appreciated that there are a large number of plate formats.and use of such plates should not be considered beyond the scope of the invention.

Preferably, in step e) the cells are seeded at a density of up to 300,000 cells per cm ² of growth surface. In one embodiment, the inventors have found the cells may be seeded at a density of 130,000 cells per cm ² to achieve particularly advantageous results.

It will be appreciated that the seeding of the cells may be undertaken by any suitable technique known by those skilled in the art.

Preferably, the plate contains a base medium into which cells are seeded.

Preferably, the base medium includes; MEM, with 292 mg/L glutamine, 100 units/ml penicillin, 0.1 mg/ml streptomycin, 12.5 ng/ml amphotericin B and 20% lamb serum.

Preferably, the total volume of base medium after seeding the cells is 3.0 ml per well in a 12- well plate. However, this should not be seen as limiting.

Preferably, one or more test compounds are added to the cells after step d) or e). It will be appreciated that there is a wide range of compounds that can be examined to test the effect on cell aggregation. For example, test compounds may include signalling molecules, particularly growth factors or inhibitors of growth factors, or known pharmaceuticals, such as minoxidil.

Culture vessels may be pre-coated with collagen, preferably rat type I collagen. Alternative well coatings include collagen of other sources and other extra cellular matrix proteins, such as human fibronectin and mouse laminin. It will be appreciated that surface coatings may affect cell attachment, proliferation and aggregate formation, such that different coatings or uncoated vessels might be used for different embodiments of the culture and assay system.

Following the seeding of the dermal papilla cells, the cells may be maintained for a set time period.

Preferably, the dermal papilla cells are maintained for approximately one to two weeks, during which time the dermal papilla cells form aggregates. Aggregates generally begin to form by 3 days after seeding. Preferably, the medium is replenished at regular time intervals. The volumes and frequencies of medium replenishment may be inter-related. For example, if the cells are maintained in a total volume of 1500 μl medium per well of a 12-we!l plate, then 750 μl spent medium may be removed and replaced with 750 μl fresh medium every second day. Alternatively, if the cells are maintained in a total volume of 3000 μl medium per well of a 12-well plate, then 1500 μl spent medium may be removed and replaced with 1500 μl fresh medium every third day.

Cell aggregation may be monitored at set time intervals, for example, 1-3 day intervals.

Once a suitable amount of time has elapsed for the cells to aggregate extensively, the culture may then be stained to enable the cell aggregates to be examined. It will be appreciated that a variety of fixation and staining techniques may be used to visualise cell structures and will be known to a skilled person. For example, the Van Gieson's stain procedure (Culling, 1957) provides suitable contrast of aggregated cells from those remaining in the background.

Preferably, the stained cultures are then digitally photographed to allow for computer assisted image analysis of aggregate size. It will be appreciated that other cell measurement techniques to allow quantitative analysis will be known to a skilled person.

In accordance with another aspect of the present invention there is provided a method for analysing images to measure the size of dermal papilla cell aggregates by the steps of:

a) staining the cells; b) creating a photograph image of the cells to be analysed; c) analysing the images.

The image analysis may be undertaken by subtraction of non-uniform background and thresholding, for example, using Analysis® pro software, provided by Soft Imaging Systems GmbH, Germany. The grey image may be converted into a binary image that best discriminates the objects (corresponding to cell aggregates) from the background. This image analysis method may take into account one or more of the following parameters: feret diameter, sphericity, and particle edge to centre (minimum).

Alternatively, aggregates may be located in images by an edge detection method. For example, the second derivative Difference of Gaussians method (Ma & Li, 1998) may be used to find object edges corresponding to local maxima of rates of change of image intensity. A region growth algorithm may be applied to define clusters of pixels which can then be outlined by a convex hull (corresponding to cell aggregates). For each convex hull, diameter parameters and aspect ratio may be calculated.

Preferably, a manual check is made of one or more of these values by comparing measurements with the appearance of associated aggregates in the image.

According to another aspect of the present invention there is provided a hair growth assay kitset which includes a vessel including ovine dermal papilla cells and a suitable medium.

In preferred embodiments the vessel may be in the form of a multi-welled plate in which each well contains dermal papilla cells and a suitable medium.

According to another aspect of the present invention there is provided a culture of ovine dermal papilla cells obtained by the method of the present invention.

According to another aspect of the present invention there are provided isolated ovine dermal papillae.

It should be appreciated from the above description that the preferred embodiments will provide a model hair growth assay and materials therefor.

BRIEF DESCRIPTION OF DRAWINGS

Further aspects of the present invention will become apparent from the following description which is given by way of example only and with reference to the accompanying drawings in which:

Figure 1 : Scanning electron micrograph showing the three dimensional appearance of ovine dermal papilla cell aggregate reformed on the surface of a culture vessel. Scale bar = 100 μm.

Figure 2: Effect of serum type.

(A, B) Aggregation in 10 % foetal calf serum on rat tail collagen.

(C, D) Aggregation in 20 % lamb serum on rat tail collagen. (B) and (D) are higher power views of the boxed areas in (A) and (C) respectively. Scale bars = 1 mm.

Figure 3: Photographs illustrating effects of culture substrate. Stained cell aggregates are dark spots formed within 22 mm diameter circular wells of multiwell tissue culture plates. Cells were seeded onto one of five substrates: (A) uncoated polystyrene plate, (B) rat tail collagen prepared in-house and air dryed on the plate surface, and commercially available plates pre-coated with (C) rat collagen I, (D) human fibronectin, or (E) mouse laminin. Two representative wells are shown for each treatment with media containing 0 and 20 mM lithium chloride. Quantitative data for the full set of experiments are shown in Figure 4.

Figure 4: Effects of culture substrate.

Solid bars show average diameter of aggregates formed on substrates as described in Figure 3. Responses to the three lithium concentrations (0, 10 and 20 mM) are represented by bars of differing shades of grey. Error bars show high and low values of the diameter range.

Figure 5: Stages of aggregation of dermal papilla cells.

(A, B) Stage 2, separation of homogenous cell layer to expose substrate. Note elongated cells extending into exposed area.

(C, D) Stage 3, formation of dense ridges of tissue at edge of exposed areas. (E, F) Stage 4, ridges begin to break up to form chains of discrete aggregates. (E) One aggregate with bridges of dense tissue extending vertically. (F) Two adjacent aggregates, nearly separated.

(G, H) Stage 5, discrete aggregates continue to contract. (H) is denser than (G), and probably extends higher above the substrate.

(I, J) Radial, rather than parallel mode of contraction. (J) is more advanced and denser than (I).

(K, L) Formation of aggregates without exposure of substrate. Note the "grain" of the surrounding cell layer is often radial to the edges of aggregates, suggesting elongated cells under tension. Scale bar = 200μm

Figure 6: Extensive cell proliferation of primary ovine dermal papilla cells.

Three strains of cells were propagated from dermal papilla explants from two different sheep. The numbers of founder cells were unknown and cells harvested were unequal up until the fourth or fifth passage (non-joined points). One cell isolate was split into subconfluent (open circles) and confluent/ aggregating lineages (closed circles). Two further isolates grown as confluent/ aggregating cultures (grey triangles and diamonds) showed comparable growth rates. These latter two cultures are ongoing. All cell cultures retained the ability to aggregate until the last passage. Figure 7: Effect of lithium chloride.

(A-C) Aggregates formed in the presence of 10, 15 and 20 mM lithium chloride respectively. These cultures were stained with Giemsa. Scale bar = 1 mm.

(D) Dose-response curves showing the effect of lithium chloride on aggregate diameter. Six cell strains are shown, derived from five different animals. (StrainsDPC-13 and DPC-15 are from the same animal.) For each strain, aggregate sizes were normalised to a value of 1.0 in the absence of lithium chloride. Error bars = standard errors of the mean of aggregate diameter within the assay.

(E) Cell morphology in the absence of lithium chloride.

(F) Cell morphology in the presence of 40 mM lithium chloride. Scale bar = 200μm

Figure 8: Effect of SU5402.

Dose response curves for effect on aggregate size of the FGFR1 inhibitor, SU5402, in the absence or presence of 10 mM lithium chloride. The data are from a set of three experiments using cell strains derived from two different sheep. For each experiment, aggregate sizes were normalised to a value of 1.0 in the absence of SU5402 and lithium chloride. Error bars = standard errors of the mean of aggregate diameter within the assay.

Figure 9: Effect of dorsomorphin.

Dose response curves for effect on aggregate size of an inhibitor of the BMP type I receptor, dorsomophin, in the absence or presence of 10 mM lithium chloride. The data are from a set of three experiments using cell strains derived from two different sheep. For each experiment, aggregate sizes were normalised to a value of 1.0 in the absence of dorsomorphin and lithium chloride. Error bars = standard errors of the mean of aggregate diameter within the assay.

Figure 10: Effect of minoxidil sulphate.

(A) Dose response curves showing the effect of minoxidil sulphate on aggregate size in the absence and presence of 10 or 20 mM lithium chloride. For each experiment, aggregate sizes were normalised to a value of 1.0 in the absence of minoxidil sulphate and lithium chloride. Means of three repeated experiments are shown, each using a cell strain derived from a different sheep. Error bars = s.e.m.

(B) Dose response curves showing the effect of streptomycin inclusion in the culture media on the response to minoxidil. For a single cell strain, the same experiment was repeated in the presence and absence of streptomycin. Aggregate sizes were normalised to a value of 1.0 in the absence of minoxidil sulphate and lithium chloride. Error bars = standard errors of the mean of aggregate diameter within the assay. (C-D) Photomontages showing aggregation in the presence of 10 mM lithium chloride and 0 μM, 20 μM and 100 μM minoxidil sulphate respectively. Well diameter is 22.1mm.

BEST MODES FOR CARRYING OUT THE INVENTION

Methods

Tissue culture media and reagents

Tissue culture media and reagents were obtained from Invitrogen (Carlsbad, CA, USA) and

Sigma-Aldrich (St. Louis, MO, USA). Tissue culture plasticware was from Nunc (Roskilde,

Denmark) or Corning (Lowell, MA, USA). Minoxidil sulphate was from Sigma-Aldrich (St Louis,

MO, USA); lithium chloride from BDH ₁ now Merck (Whitehouse Station, NJ, USA).

Dorsomorphin and SU5402 were obtained from Calbiochem (La JoIIa, CA, USA).

Obtaining Tissue Cultures

Wool-bearing skin was collected from the head or neck of lambs, within 15 minutes of death, and placed in MEM, 100 units/ml penicillin, 0.1 mg/ml streptomycin, 12.5 ng/ml amphotericin B, 10% lamb serum, on ice. Microdissection of the skin was performed in the same medium at room temperature, under a stereomicroscope, in a sterile cabinet.

Follicles were cut from the surrounding dermis using microscissors. Dermal papillae were then excised from isolated follicles using 26 gauge hypodermic needles and fine watchmakers' forceps. To initiate cultures, 6-8 papillae were transferred to a 35 mm diameter dish containing 2 ml MEM, 100 units/ml penicillin, 0.1 mg/ml streptomycin, 12.5 ng/ml amphotericin B, 20% lamb serum. Cultures were maintained at 37 ⁰ C, 5% CO ₂ .

Primary cultures were left for 1 week without being disturbed, and were then given twice-weekly medium changes. Cells were harvested by trypsinisation and seeded into a 25 cm ² flask for the first passage, and into 10 cm dishes for subsequent passages. Cells were allowed to become confluent and show preliminary signs of aggregation before being passaged. Cells were frozen in 90% lamb serum, 10% DMSO at a density of 5,000,000 cells/ml. A vial of 1 ml was thawed and seeded into a single 10 cm dish for subsequent use.

Cell cultures were examined with the aid of a Zeiss Axiovert 40CFL microscope, and wells photographed using an Olympus SZ40 microscope and a Dage-MTI DC330E CCD camera or a Leica DFC290 camera with Leica Application Suite software (version V 2.5.0). Scanning electron microscope images of aggregates formed in culture were obtained by critical point drying, sputter coating with gold, and secondary electron imaging using a Jeol JSM 7000F Field Emission Gun instrument.

Optimised Culture Conditions

The effects of serum supplements on cell growth and aggregation were assessed by growing cells in 10 cm dishes in MEM with antibiotics (as above) supplemented with either 0, 5, 10, or 20% foetal bovine serum (FBS) ₁ or 0, 5, 10, or 20% lamb serum. Media additionally contained either 0, 10, or 20 mM lithium chloride. Aggregate diameter was measured according to the standard assay procedure (described below).

The effects of attachment substrates on cell growth and aggregation were assessed by growing cells in base medium with 20% lamb serum (described below). Cells were trypsinised and seeded into 12-well tissue culture plates with one of five surface coatings: uncoated polystyrene, collagen type I extracted from rat tail tendons, pre-coated rat collagen type I, human fibronectin, or mouse laminin. Pre-coated plates were from Becton Dickinson (Oxford, UK). Media contained 0, 10, or 20 mM lithium chloride. Within an experiment, each combination treatment was repeated in quadruplicate. Experiments were replicated using cells isolated from different sheep (uncoated plastic and precoated collagen I in duplicate; in-house collagen 1 , fibronectin and laminin in triplicate). Wells were photographed and assessed subjectively according to consistency of aggregate form, cell attachment and rate of aggregate formation, and size differential between lithium treatments. Aggregate diameter was measured according to the standard assay procedure (described below).

Dermal papilla cells were grown in vitro until replicative senescence to determine the persistence of the aggregative phenotype. Cells were isolated from ovine dermal papilla explants, propagated for 3-4 passages, and then split into two cultures. One culture was sequentially seeded at 4,000 to 7,000 cells/cm ² and passaged before the cells reached confluence. The other was seeded at 17,500 cells/cm ² so as to attach as a confluent monolayer and allowed to aggregate before each passage. An aggregation assay (as described below) was performed to determine the presence or absence of aggregate formation within two weeks at every third passage for subconfluent cells and at each passage for confluent/ aggregating cells. The experiment with aggregating cells was repeated for two additional cell isolates from different sheep.

Assay procedures To begin a standardised assay, cells were again grown to confluence and allowed to show preliminary signs of aggregation. Cells were harvested by trypsinisation, resuspended at approximately 1,333,000 cells per ml, and seeded into a 12-well plate at a density of up to 1,000,000 cells/well. Typically, the plate was pre-coated with rat type I collagen. The base medium was MEM, 292 mg/L glutamine, 100 units/ml penicillin, 0.1 mg/ml streptomycin, 12.5 ng/ml amphotericin B, 20% lamb serum, supplemented with the test compound of interest.

Four compounds with known pharmacologic effects were tested in the assay at concentration ranges that spanned published IC ₅₀ values for other tissue or cell types. Lithium chloride was added at concentrations of 0-40 mM (Rao et al. 2005). 3-[3-(2-carboxyethyl)-4-methylpyrrol-2- methylidenyl]-2-indolinone (SU5402) was added at concentrations of 0-40 μM (Mohammadi ef al. , 1997). 6-[4-(2-piperidin-1 -yl-ethoxy)-phenyl)]-3-pyridin-4-yl-pyrrazolo[1 ,5-a]-pyrimidine (dorsomorphin) was added at concentrations of 0-2.5 μM (Yu et al., 2008). Minoxidil sulphate was added at concentrations of 0-100 μM (Li ef al., 2001; Messenger & Rundegren, 2004). SU5402 and dorsomorphin were tested in the absence or presence of 10 mM lithium chloride. Minoxidil was tested with 0, 10 and 20 mM lithium chloride.

Cells were maintained in 1.5 ml of medium per well, and given a half-volume medium change every second day (i.e. 750 μl spent medium removed, 750 μl fresh medium added). Alternatively, the cells were maintained in 3.0 ml and given a half-volume change on the third day.

Aggregation was monitored at 1-3 day intervals. Experiments were stopped once numerous spherical aggregates were seen throughout the cultures, typically between 1 and 2 weeks after setting up the assay.

Cultures were then fixed and stained via the following steps: two brief washes with PBS; 90% ethanol, 5% acetic acid, 5% water for 20 min.; three brief washes with water; Van Gieson's solution (0.5 g acid fuchsin, 500 ml saturated picric acid, 0.5 ml concentrated hydrochloric acid) for 3 min.; four 2 min. washes with water; 50% ethanol for 5 min.; brief washes with 70% then 95% ethanol, two 3 min. washes in 100% ethanol.

An alternative stain was as follows: two brief washes with PBS; brief washes with 50% methanol, 50% PBS, then with 100% methanol; 100% methanol for 10 min.; Giemsa stain (Sigma-Aldrich, St Louis, MO, USA) for 4 min.; the Giemsa stain was then diluted with 4 volumes of water and removed; 4 brief washes with water; brief washes with 70% ethanol, then 95% ethanol, then 100% ethanol twice. Imaging

Stained cultures were photographed for image analysis using an Olympus SZ40 microscope with zoom set to 0.67x, 0.5x photoeyepiece or 0.38x camera coupler, camera and image capture software, as described above.

Aggregate size information was extracted from digital images using at least one of three methods: (i) digitally assisted manual measurement, (ii) image analysis by thresholding, and (iii) image analysis by edge detection.

(i) Manual measurements were performed with the aid of Image J software, version 1.36b (NIH, Bethesda, MA, USA). Images were inspected to exclude unusually shaped or elongated aggregates (aspect ratio greater than approximately 2). For a random sample of up to 20 remaining aggregates, two perpendicular measurements were made between the opposite edges of the stained central body of the aggregate. For each aggregate, the root mean square of the perpendicular measurements was taken as a diameter value.

(ii) Image analysis by thresholding was conducted using Analysis® pro software (Soft Imaging Systems GmbH, Mϋnster, Germany). Objects (aggregates) were discriminated from background by converting the image to greyscale, smoothing to create a background image, background removal, and threshold determination from the second derivative of the greyscale histogram. From each discrete object in the binary image, measurements were made of Feret diameter (minimum and maximum), grey value CV, sphericity, and particle edge to centre (minimum). Using these parameters, identified objects were accepted or rejected on size and shape criteria. For each object (single aggregate), the square root of the measured area was taken as a diameter value.

(iii) Image analysis by edge detection was conducted using a custom computer application written in Delphi 2009 for Windows. The second derivative Difference of Gaussians edge detection method was used with recommended parameters (Ma & Li, 1998) to find object edges corresponding to local maxima of rates of change of image intensity (Marr & Hildreth, 1980). The output image with its detected object edges was intensified via nonlinear sigma scaling contrast stretch. A region growth algorithm was applied to the contrast stretched image, using non-recursive two-pass raster scanning to form clusters of pixels which were then outlined by a convex hull (corresponding to a detected cell aggregate). For each convex hull, the maximum diameter, minimum width, and aspect ratio were calculated. Finally, for those convex hulls whose aspect ratio falls between specified limits, circularity was calculated, together with a linear measure of object size (the square root of the product of minimum width and maximum diameter).

For all three methods, a manual check was made of outlying values. Mean aggregate sizes in treated and untreated cell cultures were compared using two-tailed, heteroscedastic t-tests in Microsoft Excel.

Results

Aggregation of ovine dermal papilla cells

Cells grown from ovine dermal papilla explants in the culture conditions devised by the inventors frequently exhibit aggregative behaviour reminiscent of dermal condensation and papilla formation during follicle morphogenesis (Figure 1).

Optimum cell assays

A range of culture conditions were tested to optimise the aggregation of ovine papilla cells, with respect to both the number of aggregates that form and the three-dimensional structure of individual aggregates. The inventors found that the cells aggregate best when grown in medium supplemented with 20% lamb serum, rather than the more commonly used 10% foetal calf serum (Figure 2). Aggregation appeared to require high cell densities, and so cells were seeded at a minimum of 130,000 cells/cm ² in the standardised assay. Frequent medium changes were required to maintain the large number of cells in a relatively low volume.

The coating of the surface of the culture vessel with proteins or glycoproteins of the extracellular matrix modified the adhesion and aggregate formation of dermal papilla cells (Figures 3 and 4). Cells readily attached to uncoated plastic and aggregate formation was more rapid than with protein coatings. Maximal size differential was achieved with collagen I coatings (Figure 3B and 3C). Fibronectin and laminin both enhanced cell attachment but gave less consistent aggregations and size differences between lithium treatments. Laminin reduced aggregate size, reducing or reversing the sensitivity to variation in lithium concentration (Figure 3D and Figure 4). Therefore, for the standard assay, the preferred culture substrate was a coating of collagen I to facilitate controlled cell attachment at the chosen cell density.

Aggregation progressed continuously under these conditions, but five characteristic stages could be recognised (Figure 5). Before aggregation (stage 1), cells appeared as a relatively homogenous field. They were dense and completely covered the substrate, but still comprised a monolayer. The earliest sign of aggregation (stage 2) was a localised separation of the cell layer to expose a small area of the substrate (Figure 5A-B). A few elongated cells often stretched into the exposed area. Several such sites appeared at the same time in one culture well. Next (stage 3), the area of the exposed regions expanded and the cells at the boundaries appeared to pull more closely together to form dense ridges of tissue (Figure 5C-D). These were three-dimensional - several cell layers thick. At lower magnification, the ridges formed elongated stripes within the culture. As the ridges began to break up (stage 4), they sometimes formed a chain of broad aggregates connected by narrower bridges of dense cells (Figure 5E- F). Finally (stage 5), the connecting bridges disappeared, leaving discrete aggregates (Figure 5G-H). These continued to contract, ultimately forming spheroid structures. Not all cells incorporated into aggregates - monolayered cells at relatively low density usually surrounded even the most well developed aggregates. It appeared that cell proliferation continued in parallel with aggregation, leading to re-colonisation of exposed areas of substrate.

There was some variability in this progression. Aggregates often appeared to emerge from a field of cells without first forming an elongated ridge of tissue. In these cases, the lines of contraction appeared to be arranged radially, centred on the forming aggregate (Figure 51-J). In contrast, ridges appeared to result when the lines of contraction were parallel, perpendicular to the forming ridge. Typically, single wells included both radial and parallel modes of contraction, resulting in the appearance of both spots and stripes during the intermediate stages. The relative frequency of the radial and parallel modes varied between experiments. Furthermore, in some experiments aggregates formed with minimal exposure of the substrate. In these cases, elongated cells were often seen radial to the edges of the aggregates, suggesting the cells were under tension (Figure 5K-L).

Experiments also varied in the degree to which the aggregation was completed. In some cases, the process stopped at the fourth or even the third stage described above. There was also variation between cell strains isolated from different animals. A few strains never aggregated well enough to allow robust measurement of aggregate size. Of a total of 19 cell strains isolated to date from 12 sheep of New Zealand strongwool breeds (Romney, Romney x Dorset, and composite breed), 11 aggregated well and have been used in quantitative assays, five have initially aggregated but have not been tested in a standard assay, and three did not aggregate during the derivation process. In the useful cell strains, the aggregative behaviour was stable for at least five passages.

Three strains that were maintained in vitro for 22-27 passages (at least 80 cell doublings) continued to proliferate at a constant rate (Figure 6). All three retained the ability to aggregate and displayed dose dependant responses in aggregate size to lithium chloride. Both subconfiuent and high density cell cultures displayed robust aggregative ability until they began to senesce (as evinced by their variable morphology and slowing proliferation). Both subconfiuent and high density cultures of this strain senesced at approximately 70 cell doublings after they diverged at passage 4 (Figure 6). Between the initial outgrowth of cells from explants up to passage 4, there were an estimated 12-18 additional cell doublings. (This number could not be measured precisely, due to the undefined number of founder cells.) Thus, these primary cells continued to proliferate and aggregate for 80-90 cell doublings, indicating that their use in the standard assay at passage 5 or 6 is probably conservative.

Under our culture conditions, cell doubling times were 20 hours for cells passaged before confluence, and 42 hours for cells that were seeded at high density and allowed to aggregate before each passage (Figure 6). Two additional cell strains maintained as confluent cultures showed similar growth rates and also continued to aggregate. This retention of aggregative phenotype is a distinctive feature of dermal papilla cells derived from sheep, as compared to other species including humans (Inamatsu et al., 2006). Furthermore, the continued aggregate formation is not required for the maintenance of aggregative ability, as demonstrated by the cell lineage that were kept subconfluent for 70 cell doublings and yet cells were still capable of forming aggregates, even when they were becoming senescent.

In the standardised assay, cells are allowed to aggregate for 1-2 weeks. After this time, dense, spheroid aggregates appear throughout the culture. Van Gieson's solution (or Giemsa) is used to selectively stain the three-dimensional aggregates, but not the surrounding monolayer of cells. After photography, image analysis is used to quantify mean aggregate diameter and aggregate number. The size of these aggregates is similar to the size of wool follicle papillae in vivo.

Lithium Chloride Experiments

The assay was used to investigate the effect of lithium chloride on papilla cell aggregation. Lithium chloride acts as an agonist of Wnt signalling (Rao et al., 2005; Williams ef al., 2004) or as an inhibitor of Src kinase signalling or inositol phosholipid signalling (Sarkar et al., 2005). Wnt signalling has been shown to regulate papilla cell behaviour in vivo (Shimizu and Morgan, 2004). The inventors hypothesised that lithium chloride would influence aggregation in vitro.

The results showed dose-dependent effects on the size of aggregates (Figure 7A-C). These results were repeated with six cell strains from five different animals. Quantitative analysis showed that aggregate size remained constant or increased up to 5-10 mM lithium chloride, and then decreased markedly at higher concentrations (Figure 7D). One cell strain produced smaller aggregates at low to moderate lithium chloride concentrations, perhaps reflecting genetic differences in the donor animals. At 40 mM lithium chloride, aggregates did not form at all, and the cells adopted a simple monolayered conformation with morphology different to that of untreated cells (Figure 7E-F). SU5402 and Dorsomorphin Experiments

The assay was used to investigate the effects on papilla cell aggregation of blockade of fibroblast growth factor (FGF) signalling with the FGFR1 inhibitor, SLJ5402, (Figure 8) and of bone morphogenetic protein (BMP) signalling with the BMP type I receptor inhibitor, dorsomorphin (Figure 9). These two signalling pathways have been strongly implicated in skin pattern formation and hair follicle morphogenesis (Jiang et al., 1999; Botchkarev & Paus, 2003; Sharov et a/., 2006). The inventors hypothesised that their inhibition would influence aggregation in vitro.

Both compounds induced a dose dependant reduction in aggregate size (Figures 8 and 9). This effect was additive with the size-reducing effect of 10 mM lithium chloride.

Minoxidil Sulphate Experiments

The inventors next investigated the effect of minoxidil sulphate on aggregation. The sulphate is an active metabolite of minoxidil (Messenger and Rundegren, 2004). In three cell strains, the inventors found little effect on the size of aggregates when cells were treated with minoxidil sulphate alone (Figure 10A). However, when cells were treated with minoxidil sulphate in combination with 10 or 20 mM lithium chloride, the inventors found that the minoxidil sulphate dose-dependently reduced the miniaturisation effect that was seen with lithium chloride alone. This result mimics the effect of topically-applied minoxidil in vivo, where it blocks the miniaturisation of follicles in balding skin, but has little effect on healthy follicles that are not undergoing miniaturisation. Although streptomycin has been reported to inhibit minoxidil in vitro (Sanders et a/., 1996), the inventors found that the inclusion of streptomycin in the culture medium had no effect on the response to minoxidil (Figure 10B). Minoxidil also affected the number of aggregates that formed per well (Figure 10C-E).

Discussion

The inventors showed that, under appropriate culture conditions, ovine dermal papilla cells aggregate to form three-dimensional papilla-like structures. The size of these structures is not random, but is consistent with the size of the papillae from which the cells are derived. The inventors further showed that the size of these aggregates can be modified by treating the cultures with bioactive compounds known to have similar effects on follicles in vivo. Thus, the culture system appears to reproduce the aggregation of dermal cells seen in vivo, providing a new experimental model for both follicle morphogenesis and for the alterations in follicle size seen in adult hair growth disorders. A similar culture system has been described by Young et al. (2008). Good spontaneous aggregation was seen for rat vibrissa dermal papilla cells grown on EVAL membrane at passage 3. These structures were also able to induce the formation of new follicles when implanted into the skin of immune-deficient mice. However no experiments were described in which the size of aggregates was modified by treatments, nor was such a response measured.

Others have also shown that cultured papilla cells are able to induce the development of new follicles when grafted into the skin (Jahoda et al., 1993). The effect is seen with adult tissues, even when cells are grafted at body sites that do not normally grow hair (Inamatsu et al., 1998; Reynolds and Jahoda, 1992). Like aggregation, the hair-inducing capacity of papilla cells is often lost after a few passages in culture. It is likely that the aggregation of papilla cells in vitro and their hair-inducing capability in vivo are interrelated; to induce a new hair follicle, cells would need to re-form a functional papilla to elicit the appropriate response in adjacent epithelial tissues (Home et al., 1986).

One means of maintaining the hair-inducing ability of papilla cells is to grow them in the presence of a source of Wnt protein (Kishimoto et al., 2000; Shimizu and Morgan, 2004). No effects on papilla cell aggregation were described in these reports. However, the hair-inducing ability of papilla cells is associated with expression of versican, a core protein of extracellular matrix proteoglycans (Kishimoto et al., 2000; Kishimoto et al., 1999; Shimizu and Morgan, 2004). In vivo, versican is expressed in the anagen papilla, but is down-regulated both in telogen and in the miniaturised follicles seen in alopecia (Soma ef al., 2005). Loss of versican expression in mouse vibrissa papilla cells is associated with loss of inductive ability (Kishimoto ■ et al., 1999). Versican is involved in cell adhesion, migration and extracellular matrix assembly, and so is a good candidate for a mechanistic role in papilla cell aggregation (Rahmani ef al., 2006; Wight, 2002).

No effects on papilla cell aggregation were described in these reports of Wnt-treated papilla cells. In contrast, the inventors found that lithium chloride, which mimics Wnt signalling, affects both the number and size of aggregates. These differences may relate to the species (mouse versus sheep) or follicle type (vibrissae versus pelage) from which the papilla cells were isolated. The inventors found that sheep vibrissa cells aggregate less well than cells from wool follicles. Mouse vibrissa papilla cells generally show little aggregation in culture after the first passage, whereas ovine wool follicle papilla cells show extensive aggregation even in the absence of lithium chloride. Thus, for the mouse cells, there may have been no baseline aggregation process on which the Wnt protein could act. In addition to Wnt, many other signalling molecules are involved in regulating hair follicle development (Botcharev & Paus, 2003). In particular, members of the FGF and BMP families of growth factors have been shown to influence the pattern formation of skin appendages and maintenance of dermal papilla cell phenotype (Rendl et al., 2008), and systems of interacting growth activators and antagonists in the skin have been proposed and modelled (Jiang et al., 1999; Michon et al., 2008). The inventors therefore sought to test the dermal papilla cell culture system using inhibitors of FGF and BMP pathways. Both SU5402 (an FGF receptor inhibitor) and dorsomorphin (a BMP receptor inhibitor) altered the measured size of dermal papilla cell aggregates at concentration ranges previously reported to have biological effects in other tissues. These results are consistent with FGF and BMP signals affecting dermal cell condensation both in vivo and in vitro, and indicate that some crucial cellular and molecular mechanisms are retained within the simplified cell culture system devised by the inventors.

The inventors also showed that the miniaturisation of aggregates induced by lithium chloride can be inhibited by co-treatment with minoxidil sulphate. Minoxidil alone had no effect - papilla cell aggregation is only sensitive to minoxidil sulphate in the presence of lithium chloride, when aggregates are miniaturised compared with untreated cells. These results closely parallel the effects of minoxidil on adult human hair follicles in vivo. Minoxidil is used to treat alopecia by inhibiting follicle miniaturisation (Messenger and Rundegren, 2004). Topical minoxidil inhibits the miniaturisation of affected follicles in alopecia, but has little effect on healthy, non- miniaturising follicles (Abell, 1988; Headington, 1987; Katz et al., 1987; Uno et al., 1987). Thus the novel culture system described here is a good model for the behaviour of adult human follicles. In vitro aggregation appears to reproduce the remodelling of the papilla that is seen in adult follicle cycling, and which is disrupted in hair growth disorders.

These results also shed light on the mode of action of minoxidil. The gross effects of minoxidil are well characterised through its use for treating alopecia (Messenger and Rundegren, 2004). Minoxidil increases the duration of the anagen phase of the cycle as well as enlarging follicles, thereby promoting the growth of longer, thicker hairs (Diani et al., 1992; Uno, 1991). However, the cellular and molecular mechanisms of minoxidil action remain unclear thirty years after its development for clinical use. Effects on follicle vascularisation or epithelial cells have been proposed, as well as direct effects on the papilla (Headington, 1987; Messenger and Rundegren, 2004).

The molecular mechanism of minoxidil action involves its activity as an opener of ATP-sensitive potassium channels (Buhl et al., 1992; Meisheri et al., 1988). These channels are widely- expressed in skin and other tissues, but it has recently been shown that minoxidil-sensitive isoforms are expressed in the papilla, and not in the epithelial tissue of hair follicles (Shorter et a/., 2008). Minoxidil stimulation results in the secretion of ATP into the extracellular space, where it is rapidly converted to adenosine. Direct effects of minoxidil and extracellular adenosine on cultured papilla cells have been shown (lino et al., 2007; Li et al., 2001). The cells respond by increasing expression of growth factors such as VEGF and FGF7, both of which are implicated in maintaining the anagen phase of the follicle cycle. The inventors' results suggest minoxidil also influences the aggregative behaviour of papilla cells to promote the maintenance of an anagen-typical structure. Because papilla remodelling is coupled with follicle cycling, it is quite reasonable that the aggregative behaviour of papilla cells is co- regulated with the expression of mitogenic signals that act on the epithelial cells which produce the hair fibre.

Aspects of the present invention have been described by way of example only and it should be appreciated that modifications and additions may be made thereto without departing from the scope thereof.

REFERENCES

Abell E (1988) Histologic response to topically applied minoxidil in male-pattern alopecia. Clinics in Dermatology 6:191-194.

Botchkarev VA, Paus R (2003) Molecular biology of hair morphogenesis: development and cycling. Journal of Experimental Zoology B 298: 164-80.

Buhl AE ₁ Waldon DJ, Conrad SJ, Mulholland MJ, Shull KL, Kubicek MF, et al. (1992) Potassium channel conductance: a mechanism affecting hair growth both in vitro and in vivo. The Journal of Investigative Dermatology 98:315-319.

Culling CFA. (1957) Handbook of Histopathological Technique. Butterworth & Co. Ltd, London.

Diani AR, Mulholland MJ, Shull KL, Kubicek MF, Johnson GA, Schostarez HJ, et al. (1992) Hair growth effects of oral administration of finasteride, a steroid 5 alpha-reductase inhibitor, alone and in combination with topical minoxidil in the balding stumptail macaque. The Journal of Clinical Endocrinology and Metabolism 74:345-350.

Elliott K, Stephenson TJ ₁ Messenger AG (1999) Differences in hair follicle dermal papilla volume are due to extracellular matrix volume and cell number: implications for the control of hair follicle size and androgen responses. The Journal of Investigative Dermatology 113:873-877.

Halloy J, Bernard BA ₁ Loussouam G, Goldbeter A (2000) Modeling the dynamics of human hair cycles by a follicular automaton. Proceedings of the National Academy of Sciences of the United States of America 97:8328-8333.

Hardy MH (1992) The secret life of the hair follicle. Trends in Genetics 8:55-61.

Headington JT (1987) Hair follicle biology and topical minoxidil: possible mechanisms of action. Dermatologica 175 Suppl 2:19-22.

Home KA, Jahoda CA, Oliver RF (1986) Whisker growth induced by implantation of cultured vibrissa dermal papilla cells in the adult rat. Journal of Embryology and Experimental Morphology 97: 111 -124. lino M, Ehama R, Nakazawa Y, Iwabuchi T, Ogo M, Tajima M, et al. (2007) Adenosine stimulates fibroblast growth factor-7 gene expression via adenosine A2b receptor signaling in dermal papilla cells. The Journal of Investigative Dermatology 127:1318-1325.

lnamatsu M, Matsuzaki T, Iwanari H, Yoshizato K (1998) Establishment of rat dermal papilla cell lines that sustain the potency to induce hair follicles from afollicular skin. The Journal of Investigative Denvatology 111 :767-775.

lnamatsu M, Tochio T, Makabe A, Endo T, Oomizu S, Kobayashi E ₁ et al. (2006) Embryonic dermal condensation and adult dermal papilla induce hair follicles in adult glabrous epidermis through different mechanisms. Development, Growth & Differentiation 48:73-86.

Jahoda CA (1998) Cellular and developmental aspects of androgenetic alopecia. Experimental Dermatology 7:235-248.

Jahoda CA, Oliver RF (1984) Vibrissa dermal papilla cell aggregative behaviour in vivo and in vitro. Journal of Embryology and Experimental Morphology 79:211-224.

Jahoda CA, Reynolds AJ, Oliver RF (1993) Induction of hair growth in ear wounds by cultured dermal papilla cells. The Journal of Investigative Dermatology 101:584-590.

Jiang TX, Jung HS, Widelitz RB, Chuong CM (1999) Self-organization of periodic patterns by dissociated feather mesenchymal cells and the regulation of size, number and spacing of primordia. Development 126:4997-5009.

Katz HI, Hien NT, Prawer SE, Goldman SJ (1987) Long-term efficacy of topical minoxidil in male pattern baldness. Journal of the American Academy of Dermatology 16:711-718.

Kishimoto J, Burgeson RE, Morgan BA (2000) Wnt signaling maintains the hair-inducing activity of the dermal papilla. Genes & Development 14:1181-1185.

Kishimoto J, Ehama R, Wu L, Jiang S, Jiang N, Burgeson RE (1999) Selective activation of the versican promoter by epithelial- mesenchymal interactions during hair follicle development. Proceedings of the National Academy of Sciences of the United States of America 96:7336- 7341.

Li M, Marubayashi A, Nakaya Y, Fukui K, Arase S (2001) Minoxidil-induced hair growth is mediated by adenosine in cultured dermal papilla cells: possible involvement of sulfonylurea receptor 2B as a target of minoxidil. The Journal of Investigative Dermatology 117:1594- 1600.

Ma S-D ₁ Li B-C. (1998) Derivative computation by multiscale filters. Image and Vision Computing 16:43-53.

Marr D, Hildreth E. (1980) Theory of Edge Detection. Proceedings of the Royal Society of London. Series B, Biological Sciences 207 1167:215-217.

Meisheri KD, Cipkus LA, Taylor CJ (1988) Mechanism of action of minoxidil sulfate-induced vasodilation: a role for increased K+ permeability. The Journal of Pharmacology and Experimental Therapeutics 245:751-760.

Messenger AG, Rundegren J (2004) Minoxidil: mechanisms of action on hair growth. The British Journal of Dermatology 150: 186-194.

Messenger AG, Senior HJ, Bleehen SS (1986) The in vitro properties of dermal papilla cell lines established from human hair follicles. The British Journal of Dermatology 114:425-430.

Mohammadi M, McMahon G, Sun L, Tang C, Hirth P, Yeh BK, Hubbard SR, Schlessinger J (1997) Structures of the tyrosine kinase domain of fibroblast growth factor receptor in complex with inhibitors. Science 276:955-960.

Michon F, Forest L, Collomb E, Demongeot J, Dhouailly D (2008) BMP2 and BMP7 play antagonistic roles in feather induction. Development 135:2797-805.

Olivera-Martinez I, Viallet JP, Michon F, Pearton DJ, Dhouailly D (2004) The different steps of skin formation in vertebrates. International Journal of Developmental Biology 48:107-115.

Paus R, Foitzik K (2004) In search of the "hair cycle clock": a guided tour. Differentiation; research in biological diversity 72:489-511.

Phelan MC (2007) UNIT 1.1 Basic Techniques in Mammalian Cell Tissue Culture. In Bonifacino JS, Dasso M, Harford JB ₁ Lippincott-Schwartz J, Yamada KM (eds.) Current Protocols in Cell Biology, John Wiley and Sons Inc.

Qiao J, Teumer JK, Philips, EJ (2007) Cell co-culture. Patent application number PCT/US2007/006790, Patent number WO 2007/109223 A2 Rahmani M ₁ Wong BW, Ang L, Cheung CC, Carthy JM, Walinski H, et al. (2006) Versican: signaling to transcriptional control pathways. Canadian Journal of Physiology and Pharmacology 84:77-92.

Rao AS, Kremenevskaja N, Resch J, Brabant G (2005) Lithium stimulates proliferation in cultured thyrocytes by activating Wnt/beta-catenin signalling. European journal of endocrinology / European Federation of Endocrine Societies 153:929-938.

Rendl M, Polak L, Fuchs E (2008) BMP signaling in dermal papilla cells is required for their hair follicle-inductive properties. Genes & Development 22:543-557.

Reynolds AJ, Jahoda CA (1992) Cultured dermal papilla cells induce follicle formation and hair growth by transdifferentiation of an adult epidermis. Development (Cambridge, England) 115:587-593.

Rushan X, Fei H, Zhirong M, Yu-Zhang W (2007) Identification of proteins involved in aggregation of human dermal papilla cells by proteomics. Journal of Dermatological Science 48:189-197.

Sanders DA, Fiddes I, Thompson DM, Philpott MP, Westgate GE, Kealey T (1996) In the absence of streptomycin, minoxidil potentiates the mitogenic effects of fetal calf serum, insulin-like growth factor 1, and platelet-derived growth factor on NIH 3T3 fibroblasts in a K+ channel-dependent fashion. The Journal of Investigative Dermatology 107:229-234.

Sarkar S, Floto RA, Berger Z, lmarisio S, Cordenier A, Pasco M, et al. (2005) Lithium induces autophagy by inhibiting inositol monophosphatase. Journal of Cell Biology 170:1101-1111.

Seago SV, Ebling FJ (1985) The hair cycle on the human thigh and upper arm. The British Journal of Dermatology 113:9-16.

Sengel P (1990) Pattern formation in skin development. International Journal of Developmental Biology 34:33-50.

Sharov AA, Sharova TY, Mardaryev AN, Tommasi di Vignano A, Atoyan R, Weiner L, et al. (2006) Bone morphogenetic protein signaling regulates the size of hair follicles and modulates the expression of cell cycle-associated genes. Proceedings of the National Academy of Sciences of the United States of America 103:18166-18171. Shimizu H ₁ Morgan BA (2004) Wnt signaling through the beta-catenin pathway is sufficient to maintain, but not restore, anagen-phase characteristics of dermal papilla cells. The Journal of Investigative Dermatology 122:239-245.

Shorter K, Farjo NP ₁ Picksley SM, Randall VA (2008) Human hair follicles contain two forms of ATP-sensitive potassium channels, only one of which is sensitive to minoxidil. Faseb Journal 22: 1725-1736.

Soma T, Tajima M, Kishimoto J (2005) Hair cycle-specific expression of versican in human hair follicles. Journal of Dermatological Science 39:147-154.

Teumer J, Cooley J (2005) Follicular cell implantation: an emerging cell therapy for hair loss. Seminars in Plastic Surgery 19:193-200.

Teumer JK, Qiao J, Lafont E, Wolowacz RG (2004) Cultivation of hair inductive cells. Patent application number PCT/GB2003/004949, Patent number WO 2004/044188 A1

Tobin DJ, Gunin A, Magerl M, Handijski B, Paus R (2003a) Plasticity and cytokinetic dynamics of the hair follicle mesenchyme: implications for hair growth control. The Journal of Investigative Dermatology 120:895-904.

Tobin DJ, Gunin A, Magerl M, Paus R (2003b) Plasticity and cytokinetic dynamics of the hair follicle mesenchyme during the hair growth cycle: implications for growth control and hair follicle transformations. The Journal of Investigative Dermatology Symposium proceedings / the Society for Investigative Dermatology, lnc 8:80-86.

Uno H (1991) Quantitative models for the study of hair growth in vivo. Annals of the New York Academy of Sciences 642: 107-124.

Uno H, Cappas A, Brigham P (1987) Action of topical minoxidil in the bald stump-tailed macaque. Journal of the American Academy of Deπvatology 16:657-668.

Whiting DA (2001) Possible mechanisms of miniaturization during androgenetic alopecia or pattern hair loss. Journal of the American Academy of Dermatology 45: S81 -86.

Wight TN (2002) Versican: a versatile extracellular matrix proteoglycan in cell biology. Current Opinion in Cell Biology 14:617-623. Williams R, Ryves WJ, Dalton EC, Eickholt B, Shaltiel G, Agam G, et al. (2004) A molecular cell biology of lithium. Biochemical Society Transactions 32:799-802.

Young T-H, Lee C-Y, Chiu H-C ₁ Hsu C-J, Lin S-J. (2008) Self-assembly of dermal papilla cells into inductive spheroidal microtissues on poly(ethulene-co-vinyl alcohol) membranes for hair follicle regeneration. Biomaterials 29:3521-3530.

Yu PB ₁ Hong CC, Sachidanandan C ₁ Babitt JL ₁ Deng DY, Hoyng SA, Lin HY, Bloch KD ₁

Peterson RT (2008) Dorsomorphin inhibits BMP signals required for embryogenesis and iron metabolism. Nature Chemical Biology 4:33-41.