A total of 15 skin biopsies (Table 1) of 5 × 2 cm were obtained following the relevant legal and ethical guidelines, from anaesthetized female pigs of 45 - 50 kg body

Table 1. Cell isolation efficiency in porcine skin biopsies.

weight. Biopsies were stored in PBS or RPMI (Sigma) supplemented with 2% penicillin/streptomycin (P/S) and processed at a maximum 24 hours post-sacrifice, as previously described [10], with variations in the donor area and the disaggregating enzyme of choice. Biopsies were obtained from three different regions: lower leg, abdomen and back skin. To obtain cell suspensions, the tissue was cut into 1 - 2 mm pieces and incubated with trypsin/ EDTA (T/E, Sigma) or liberase DH (Roche). For protocol 1, lower leg skin was incubated with trypsin/EDTA overnight (O/N); for protocol 2, dorsal skin was incubated with T/E between 4 and 16 hours and for protocol 3 ventral skin was incubated with liberase between 24 - 48 hours. After the first digestion step, epidermis was removed from dermis and discarded. Dermis was then minced and incubated with the same enzyme as in the first step, at 37˚C. The sample was then centrifuged and the supernatant discarded. The remaining skin pieces were then mechanically dissociated and the cells obtained cultured in DMEM/F12 (Sigma) supplemented with methylcellulose (0.8%, R & D), B27 (2%, Gibco), FGF (40 ng/ml, Gibco), EGF (20 ng/ml, Austral Biologicals), P/S and glutamine (Sigma) at a cell density of 90,000 cells/ cm².

For cell differentiation, 2000 cells were plated per well onto coverslips covered with extracellular matrix secreted by the 804G cell line [11], on 24-well culture dishes. Cells were let to proliferate in basal medium (BM) consisting of DMEM/F12 supplemented with B27, P/S, glutamine and 1% FBS (Sigma) for 3-4 days. Then, cells were let to differentiate for 2 weeks in neural differentiation medium consisting of DMEM/F12 supplemented with P/S, glutamine, N2 (1%, Invitrogen), forskolin (5 uM, Sigma) and Heregulin β-1 (40 ng/ml, Peprotech).

Spheres were seeded in 12-mm diameter-coverslips previously treated to improve adherence as above. Two hours later, dermospheres were fixed with 4% paraformaldehyde for 15 minutes and permeabilised in 0.2% Triton X-100 and 6% FBS for 40 minutes. Differentiated cell cultures were fixed with 4% paraformaldehyde in PBS for 10 minutes and permeabilised in 0.5% Triton X- 100 for 10 minutes. Anti-nestin (Santa Cruz, polyclonal), vimentin (Sigma, clone v9), fibronectin (Sigma, polyclonal), SMA (Sigma, clone 1A4) and bIII tubulin (Abcam, polyclonal) primary antibodies were used. Goat anti rabbit Alexa 488 (H + L) and Goat anti mouse Alexa 488 (H + L) (Molecular Probes) were used as secondary antibodies. Hoechst 33258 (Sigma, 1 µg/ml) was used as nuclear counterstain and the slides were mounted on Vectashield (Vector laboratories).

To estimate percent cells differentiated to a given phenotype, 10 fields were counted per coverslip. Total cells were estimated through nuclear counts and cells in a given phenotype through counting of cells positive to the specific markers. Average of the 10 fields was calculated for each coverslip.

One-way ANOVA tests, with Scheffé corrections, were used to calculate p values between different protocols. IBM SPSS Statistics Version 20 was used.

In order to assess the best protocol for isolating dermal precursor spheres from adult porcine skin, we subjected 15 porcine skin biopsies (Table 1) to three different processing protocols. All protocols isolated SKPs as previously described [10], with variations in the donor area

and the disaggregating enzyme of choice (Figure 1). In protocol 1 and 2, Trypsin/EDTA (T/E) was used for skin disaggregation, while Liberase DH was used in protocol 3. With respect to donor area, lower leg (protocol 1, n = 3), dorsal (protocol 2, n = 7) and ventral skin (protocol 3, n = 5) were tested. The cell suspensions obtained were seeded at a fixed cell density of 90,000 cells/cm² in untreated culture plates, and grown in serum free proliferation medium consisting of DMEM/F12 supplemented with methylcellulose, B27, FGF, EGF, P/S and L-Glutamine (see Matherials and Methods). We first compared the number of cells per gram of tissue obtained in each protocol.

On average, protocol 3 yielded increased cellularity (Figure 1(b)) of the initial cell suspension, the difference to protocol 2 being statistically significant (p < 0.05).

However, it was more important to see if the obtained cells give rise to spheres, which are enriched in precursor cells. Protocol 1 yielded no sphere (0/3 biopsies). Protocol 2 generated spheres in 71.43% of the cases (5/7 biopsies). Cells obtained under protocol 3 gave rise to spheres in 100% of the cases (5/5 biopsies).

These results suggest that processing adult porcine ventral skin with Liberase DH generates SKP cultures in a reproducible manner, while ensuring maximum cellular yields are obtained per gram of tissue.

To characterize the porcine adult SKPs (pSKPs) obtained by using the protocol of choice (protocol 3), porcine dermospheres (Figure 2(a)) were analyzed by immunofluorescence to detect SKP markers nestin, fibronectin and vimentin [1-3]. As expected, porcine dermospheres expressed all three canonical markers (Figures 2(b)-(d)).

To verify their multipotent differentiation capacity, pSKPs were put into differentiation in basal medium (consisting of DMEM/F12 supplemented with B27, P/S, glutamine and 1% FBS) and in neural differentiation medium (consisting of DMEM/F12 supplemented with P/S, glutamine, N2, forskolin, and Heregulin β) for 14 days (see Matherials and Methods). Differentiated pSKPs expressed both mesodermal (SMA, 16.4%) and neural differentiation markers (bIII tubulin, 25.2%; Figure 2(e) and data not shown).