We used canine AD-MSCs that were supplied from Seung Hoon Lee (Department of Veterinary Clinical Sciences, College of Veterinary Medicine, Seoul National University); these AD-MSCs were previously characterized (18). The AD-MSCs were cultured in low-glucose Dulbecco’s modified eagle medium (DMEM, Gibco^®, Life Technologies, Grand Island, NY, USA) with 10% fetal bovine serum (FBS, Gibco^®, Life Technologies, Grand Island, NY, USA), 100 units/mL penicillin and 100 mg/mL streptomycin at 37.0 °C, in a 5% CO₂ incubator.

Full-thickness skin biopsies obtained from the flank region of healthy beagle dogs approximately 3 years old were immediately placed in phosphate-buffered saline (PBS). Under a dissecting microscope, the hair follicles were isolated using a No. 15 blade, watchmaker’s forceps, and 23G needles. Isolated hair follicles were then transferred to a fresh medium-containing dish for further isolation. After the hair cycle stage was evaluated by the position of DP and tip of hair shaft, DPs in late anagen were isolated (19). Five DPs were placed together in a single well of a 12-well cell-culture cluster, which contained 1.2 mL of AmnioMAX™-C100 complete medium (AmnioMAX™, Invitrogen, Carlsbad, CA, USA) supplemented with 10% inactivated and steroid-free canine serum at 37 °C, in a 5% CO₂ incubator (6, 8, 20-22). The culture medium was replaced every third day. Primary cultured DPCs were harvested when confluent and then passaged on every tenth day. All procedures were approved by the Seoul National University’s Institutional Animal Care and Use Committee (SNU-140729-5).

Canine AD-MSCs were subjected to a monolayer culture in DMEM supplemented with 10% FBS, 100 units/mL penicillin and 100 mg/mL streptomycin, until the cells occupied approximately 80% of the culture dish. Next, the culture medium was replaced with dermal papilla-forming medium (DPFM) (23) that contained 10 ng/mL hydrocortisone, 5 mL Insulin-Transferrin-Selenium liquid media (ITS, Gibco^®, Life Tecnologies, Grand Island, NY, USA), and 100 units/mL penicillin, 100 mg/mL streptomycin, and 20 ng/mL HGF (recombinant human HGF, Gibco^®, Life Tecnologies, NY, USA). The medium was changed every third day for three weeks, after which the culture was treated with the StemPro^® Accutase^® Cell Dissociation Reagent (Accutase^®, Gibco^®, Life Technologies, Grand Island, NY, USA) at concentrations ranging from 20 to 40 μL/cm² to detach the cells from the culture dish (detach-attach step). Cell distribution was changed and cell aggregation was observed 24 h after treatment with Accutase; the aggregated cell mass became suspended in the medium five to seven days after the treatment. The suspended cell aggregates were then isolated from the culture by centrifugation at 500 rpm for 3 min. All procedures were performed based on those of previous studies (4, 23).

Total RNA from all three cell types (canine AD-MSCs, DPCs and DPLTs) was isolated using Hybrid-R™ with RiboEx^® (GeneAll^®, Seoul, Korea) according to the manufacturer’s instructions. cDNA was then synthesized, using one microgram of the total RNA, with a PrimeScript™ RT-PCR Kit (Takara Bio Inc., Shiga, Japan) for RT-PCR processing according to the manufacturer’s instructions. Cycling conditions included initial denaturation for 10 min at 94 °C, followed by thirty-five 15-s cycles at 95 °C and 60-s cycles at 60 °C. The products were then analyzed by electrophoresis on a 1.6% agarose gel and subsequently visualized by the Redsafe™ Nucleic Acid Staining Solution (iNtRON Biotechnology, Gyeonggi-do, Korea). We tested well-known hair-inductive markers of canine DPCs, such as alkaline phosphatase detection (ALP), Wnt inhibitory factor 1 (WIF1), laminin, vimentin, versican, and prominin-1 (1, 8, 13, 15, 24). Glyceraldehyde 3-phosphate de hydrogenase was used as the reference gene.

The cell and tissue extracts were prepared using the PRO-PREP™ protein extraction solution (iNtRON Biotechnology, Gyeonggi-do, Korea). Protein samples were run on a 10% sodium dodecyl sulfate (SDS)-polyacrylamide gel, transferred onto a nitrocellulose membrane, and incubated with an anti-human versican (VCAN) antibody (R&D systems, Minneapolis, MN, USA), anti-laminin antibody (ab11575, Abcam, Cambridge, MA, USA), and anti-vimentin antibody (ab8069, Abcam, Cambridge, MA, USA). An anti-beta actin antibody was used (sc-47778, Santa Cruz Biotechnologiy, Paso Robles, CA, USA) as an internal control. Blots were then incubated with a peroxidase-conjugated secondary antibody and visualized using a Novex^® ECL Chemiluminescent Substrate Reagent Kit (ECL, Novex^®, Grand Island, Life Technologies, NY, USA).

A student’s t-test was used for comparing the size between the canine DPCs and DPLTs. A value of p<0.05 was considered to be statistically significant. The statistical analyses were performed using GraphPad Prism version 6.00 for Windows (GraphPad Software, La Jolla, CA, USA).

The DPs were teardrop-shaped (Fig. 1a). Their approximate sizes were measured and compared with those of DPLTs (Fig. 1b). The mean actual canine DP maximal width was 265.8 ± 11.50 μm. The first outgrowth of dermal papilla cells presented after five days, and the cells had a flattened, polygonal morphology (Fig. 1c) with multiple cytoplasmic processes. We also observed the aggregative growth pattern in the multi-layer culture with the formation of the pseudo-papillae (Fig. 1d) and cells formed multilayered conglomerates on day 8.

After the cell detach-attach process, cell aggregates were formed within five days. The mean DPLT size was 296 ± 12.43 μm, and there was no statistical difference when compared to actual DP size (p = 0.1465) (Fig. 1e). Various factors affected DPLT formation; of these, cell inoculation density was one of the most prominent factors. We tried DPLT formation with different cell inoculation densities, from 2.0 x 10⁴ cells/cm² to 8.0 x 10⁴ cells/cm². It was found that the optimal inoculation cell density was from 6.0 x 10⁴ cells/cm² to 8.0 x 10⁴ cells/cm² (data not shown). Cell migration only occurred in densities lower than 2.0 x 10⁴ cells/cm². Even though significant aggregation was observed from a density of 4.0 x 10⁴ cells/cm², in the optimal density, the number of DPLTs was higher and the sizes were more regular than those of DPLTs in a density of 4.0 x 10⁴ cells/cm².

We performed RT-PCR and Western blotting to evaluate molecular similarities between DPLTs and DPs. Well-known hair-inductive markers, as previously mentioned, were assessed for evaluating the hair-inductive abilities of three groups on a genetic level. In addition, we assessed the quantitative protein expression level with laminin, vimentin and versican.

According to previous studies, ALP was used as an indicator for hair inductivity (6, 15), and versican was expressed in anagen hair follicles but absent in telogen hair follicles. This might support the idea that DP played an important role in anagen induction and maintenance (15, 24). WIF1, vimentin and laminin were related to signal transduction, cellular junction and cytoskeleton, and extracellular matrix and cell adhesion of DPs, respectively (12). In previous studies, WIF1 was considered to positively relate to in vivo murine hair-inductive capacity and was significantly over-represented in freshly isolated canine DPs (1, 6, 11, 13). Vimentin is known for the functions of promoting the proliferation of DPCs and increasing cell migration and growth factor expression and laminin plays an important role in promoting DP development and function during early hair morphogenesis (25).

RT-PCR analysis revealed that AD-MSCs, DPCs, and DPLTs expressed all markers as mentioned above (Fig. 2). In western blotting, all three groups showed similar versican expression level, which was reflected by their hair-inducing ability. However, we also observed that vimentin and laminin had a higher level of expression in DPCs and DPLTs compared to AD-MSCs (Fig. 3), which might mean that DPLTs have higher hair inductivity than AD-MSCs and the potential to be a viable substitute for actual DPs.