All procedures involving laboratory animals were evaluated and authorized by the Institutional Animal Care and Use Committee (IACUC). The whole procedure for samples collected was carried out in strict accordance with the protocol approved by the IACUC (permit number DK996).

10 skin samples from Yorkshire pigs born in the same period at different postnatal stages (P10, P18, P20, P30, P40, P45, P62, P75, P85, and P100) were obtained from precisely localized dorsal skin areas, as described previously (Paus et al., 1999).

Using a comprehensive guide for the recognition and classification of the distinct stages of the HF cycle, the HF cycle stages were identified based on the characteristics and gene expression of different periods. The morphologic criteria described should be used to determine the stages of a defined number of longitudinally cut HF per time point (e.g., 10 HF per time point) (Supplementary Figure S1). As described by Paus et al. (Paus et al., 1999) and Kloepper et al. (Kloepper et al., 2010), >2 TUNEL⁺ apoptotic cells in the hair bulb for early catagen; epithelial strand (ES), ball-shaped DP, hair club, germ capsule, thinner hair bulb and TUNEL⁺ apoptotic cells in ES for mid-catagen; thinner hair bulb matrix, brush-like hair club, shorter ES and apoptotic cells for late catagen; the second hair germ (HG) for telogen; onion-shaped DP, more hair bulb matrix, IRS, higher Ki67 ⁺ cell for anagen. Hematoxylin and eosin (H&E) staining were used to show the morphological characteristics of each stage and immunofluorescence staining was used to detect the expression of marker genes at different stages to distinguish the different stages of the hair cycle.

To further validate the periodization and to explore the expression of key genes for hair follicle cycle development at different periods of the pig HF cycle, we examined the expression of CD34, AE15, Versican, and K14 genes associated with HF cycle development at different stages (P18, P45, P62, P75, P85, and P100) of the pig HF cycle. Gene expression and localization are presented by immunofluorescence staining. The information related to antibody details for immunofluorescent staining are in Supplementary Table S1.

Several genes such as Wnt5a, Wnt10b (Reddy et al., 2001), SHH (Song et al., 1999), LHX2 (Kobielak et al., 2007), LEF1 (Lee and Tumbar, 2012), α-SMA have been suggested to play an important role in the hair cycle. Therefore, we examined the expression of these genes in the anagen (P18), catagen (P62) and telogen (P85) to explore their effects on different periods of pig HF cycle development. Gene expression was detected by RT-PCR. In addition, we examined the expression of two classical marker proteins of the hair follicle, SOX9 and K14, in different periods-the anagen (P18), catagen (P62), and telogen (P85) by Western blot. The antibody information details for Western blot are in Supplementary Table S1. The primer sequences are in Supplementary Table S2.

Skin tissue soaked in paraformaldehyde was cut into small pieces of approximately 1 × 3 mm along the direction of hair growth. The pieces were placed in an embedding box and soaked in formalin. Embedding boxes containing the tissues were rinsed in running water for 1 h, dehydrated in different gradients of ethanol (90%, 90%, 100%) for 1 h each time, then in 100% xylene (I→II→III) for 1 h each time. The paraffin wax was immersed in a thermostat at 55°C–65°C thrice for 1 h each time. The tissues were then embedded in paraffin. Paraffin blocks were cut into 2–3 µm slices using a slicer, spread in a 40°C–50°C water bath thermostat, fished out with a clean slide, and dried on a 37°C dryer to remove the moisture.

Paraffin sections were placed in a 60°C thermostat for 1 h and removed, washed thrice with xylene for 10 min each time, then washed thrice with anhydrous ethanol, stained with hematoxylin for 5 min, rinsed with running water for 10 min, stained with eosin for 5 min, washed briefly with alcohol, dehydrated with an ethanol gradient (70%, 90%, 100%) for 5 min each, washed thrice with xylene for 5 min each, and then covered with a coverslip using neutral resin.

Paraffin sections were removed from the oven at 60° for 1 h, washed three times with xylene for 10 min each time, washed with ethanol at 100%, 95%, and 80% concentrations for 10 min each, rinsed with running water for 5 min, washed once with phosphate-buffered saline (PBS; 15 min), and incubated with antigen restoration solution in the microwave oven for 5 min on high-power and then turned to low-power for 20 min. The sections were cooled for 1 h, washed thrice with PBS for 5 min/time, incubated with 5% sheep serum for 30 min, and incubated with primary antibody for 12 h at 4°C in a refrigerator. Paraffin sections were washed thrice with PBS, incubated with secondary antibody for 30 min, washed thrice with PBS, incubated with 4′,6-diamidino-2-phenylindole (DAPI, Solarbio, C0065) for 2 min, washed thrice with PBS, and then covered with an anti-fluorescence attenuation sealing solution. The samples were imaged using a TCS SP5 confocal microscope (Leica, Wetzlar, Germany). The numbers of cells and fluorescence intensity were analyzed using ImageJ (National Institute of Health, Bethesda, MD, United States) to quantify.

Paraffin sections were removed from the oven at 60°C for 1 h, washed thrice with xylene for 10 min each time, washed with anhydrous ethanol for 5 min, 90% ethanol for 2 min, 70% ethanol for 2 min, and deionized water for 2 min. Proteinase K without DNase was then added to the sample at 20ºC–37°C for 20 min. The TUNEL detection solution was added to the samples after washing with PBS thrice, and the samples were protected from light for 6 min. After washing thrice with PBS and adding DAPI, it was left for 2 min at room temperature. Finally, the sample was washed with PBS, and an anti-fluorescence attenuation sealing solution was added to the sample and observed under a confocal microscope.

Skin tissue was ground and put into a centrifuge tube containing TRIzol, to which 200 µL of chloroform was added, and centrifuged for 15 min (4°C, 12,000 r) after standing for 3 min at room temperature. The supernatant was then aspirated into a new tube, 500 µL of isopropanol was added, and the tube was left on ice for 20 min; after centrifugation (4°C, 12,000 r) the supernatant was poured off to retain the precipitate, which was washed twice using 1 mL of pre-cooled 75% alcohol, diluted with RNase-free double-distilled water (ddH₂O), and finally the RNA was checked for quality before storage at −80°C.

1) Total RNA Purification: Total RNA (1 μg) from the skin was added to 2 μL of 5×gDNA Eraser Buffer and 1 μL of gDNA Eraser (Takara Bio Inc., Tokyo, Japan) and then supplemented to 10 μL with RNase-free ddH₂O. Purified RNA was obtained via PCR as per a specific procedure. 2) Reverse transcription of RNA: The following mixture was prepared: purification RNA (10 μL) in (1), Prime Script RT Enzyme Mix (1.0 μL), RT Prime Mix (1.0 μL), 5× PrimeScript Buffer 2 (4.0 μL), RNase Free dH₂O (4.0 μL). The cDNA was obtained via PCR as per a specific procedure, and the cDNA was diluted 3-fold with RNase ddH₂O. 3) q-PCR: The following mixture was prepared: 2 μL of cDNA, 1 μL of PCR Reverse Primer (10 μM), 1 μL of PCR Forward Primer (10 μM), and 6 μL of SYBR Green I Master. The mRNA expression was determined via q-PCR using the Roche Light Cycler 480 (Roche Diagnostics International AG, Rotkreuz, Switzerland).

Skin tissues were added to protein lysate (RIPA) and protease inhibitor (PMSF; RIPA: PMSF = 100:1; Solarbio, Beijing, China), and the supernatant after centrifugation (4°C, 12,000×g, 5 min) was obtained for further analysis. The protein concentration was quantitatively analyzed using the Beyotime BCA Protein Concentration Assay Kit (Beyotime Biotechnology, Shanghai, China). A 10% sodium dodecyl-sulfate polyacrylamide gel electrophoresis (SDS-PAGE) gel was prepared as per the Beyotime kit, following which the proteins were separated into different bands via SDS-PAGE. After electrophoresis, the samples were transferred to a polyvinylidene difluoride membrane and blocked with a 10× mixture of tris-buffered saline and Polysorbate 20 (TBST) + milk powder to prepare a 5% milk powder blocking solution used for 2–3 h. Samples were washed five times with 1× TBST (5 min each time), incubated with the antibody (overnight at 4°C), washed five times with 1× TBST (5 min each time), and incubated with the secondary antibody (2 h at room temperature). The samples were washed five times with TBST and then photographed in a protein developer with a specific protein developer solution. The results were analyzed using ImageJ (National Institute of Health, Bethesda, MD, United States) to quantify protein expression.

Anagen (<P45) and Early Catagen (P45-P62). Delineation of key time points in pig HF cycle based on morphological characteristics at different time points with marker gene detection (Figure 1; Supplementary Figure S1). Longitudinally cut HF at different days after birth revealed that the HF morphological structure has matured until 18 days after birth (P18), with the CTS, ORS, IRS, and hair shaft visible from the outer layer to the inner layer (Figure 1B). The SOX9 ⁺ ORS is visible (Figure 1C), the hair bulb was filled with a hair matrix of four to five cell layers, and the DP expanded the hair bulb completely. Ki67 ⁺ proliferating cells were detected in the hair matrix (Figure 1D; Supplementary Figure S2A). At P45, however, we found no Ki67 ⁺ cells in the hair matrix (Figure 1G_b), instead TUNEL⁺ apoptotic cells were detected above the DP (Figure 1H_c and Supplementary Figure S2B), although the morphological structure of the HF does not seem to change significantly (Figure 1F). The results indicate that the matrix cells show signs of regression at P45.

Mid-catagen (P62-P75). The appearance of epithelial strand, hair club, and more TUNEL⁺ cells characterize the mid-catagen. At P62, the development of hair club, epithelial strand (ES) and the germ capsule around the hair club were the key features of the mid-catagen (Figure 1J). Germ capsule formation led to the constriction of the developing epithelial strand between the DP and germ capsule. The brush-like hair club (Figure 1J_b) was surrounded by SOX9 ⁺ ORS cell layers (Figure 1K). The decreasing hair matrix and proximal ORS cells formed the epithelial strand (Figure 1J_e) between the compact ball-shaped DP (Figure 1J_a) and germ capsule (Figure 1J_c). Degraded matrix leads to a thinner hair bulb (Figure 1J_f). Furthermore, the degraded IRS displayed sharply demarcated borders (Figure 1J_d). In addition, TUNEL⁺ cells were detected in the ES (Figure 1L_g; Supplementary Figure S2D), and no Ki67⁺ cells were observed in the matrix (Figure 1_K). These indicate that the HF is in the middle catagen.

Late catagen (P75-P85). The late catagen is characterized by the shorter ES, thinner matrix (almost disappear), and prominent hair club. At P75, the results of hematoxylin and eosin staining showed that HF morphology was altered compared to that at P62 (Figure 1N). The depressed ES (Figure 1N_c) was shorter and surrounded by a thickened connective sheath (Figure 1N_e). The thinner matrix was almost disappeared (Figure 1N_a). The SOX9 ⁺ ORS cells (Figure 1O) surrounded the brush-like hair club, and the IRS was unclear. Immunofluorescence staining showed TUNEL⁺ apoptotic cells (Figure 1P_f) in the ORS and DP.

Telogen (P85-P100). The telogen HF is easily recognizable because of the secondary hair germ (SHG). At P85, the lower HF length was the shortest (Figures 2A, B). The SHG was formed by compact matrix (Figure 2B_b), and a compact ball-shaped DP was closely attached to the SHG (Figure 2B_c). Telogen HF did not display any part of the IRS (Figure 2B_d). In addition, no TUNEL⁺ cells were detected in the SHG (Figure 2D), and a small number of SOX9 ⁺ cells are present in the SHG (Figure 2C_e). The HF is in a state of relative silence in P85 with cellular inactivity.

Next Anagen (>P100). In this stage, the secondary hair germ proliferated and thickened to wrap around the developing DP (Figure 2F_a), and the HF matrix increased to form a newly expanded hair bulb (Figure 2F_b). New IRS (Figure 2F_d) and the hair shaft (Figure 2F_c) reappeared. Intense Ki67 immunofluorescence in the hair matrix shows that the hair matrix is actively proliferating (Figure 2G_e; Supplementary Figure S2C), however, no apoptotic cells were detected (Figure 2H_f). The results showed that HF entered the new anagen phase at P100.

In the above results, we initially classified the different stages of the pig HF cycle. To further determine the accuracy of the periodization and to explore the role of key genes in the cyclic development of pig HF, we examined the expression of CD34, AE15, and Versican at different stages (P18, P45, P62, P75, P85, and P100) of the pig HF cycle (Figure 3). We found that the marker gene for the IRS, AE15, was significantly expressed in P18 compared to P45 (Figure 3_a, b; Supplementary Figure S2E), but not in P62 (Figure 3A_c), P75 (Figure 3A_d) and P85 (Figure 3A_e), further suggesting a gradual regression of the IRS during HF degeneration. Interestingly, we found that AE15 was expressed in the hair papilla and hair matrix during the new anagen phase (P100) (Figure 3A_f). CD34 is expressed in the ORS in anagen (Figure 3B_a, f, arrowed), almost disappears in catagen (Figure 3B_b, c, d), and is not expressed during the telogen (Figure 3B_e). The Versican is expressed at different stages of the pig HF cycle, but there are differences in the level and location of expression. For example, Versican is specifically expressed in the DP in the anagen (Figure 3C_a, f), but in catagen and telogen, it is expressed mainly in the ES (Figure 3C_c, d) or SHG (Figure 3C_e) rather than in the DP. In addition, K14, as a classical marker of the hair follicle, was expressed in the ORS and the hair matrix at every stage (Figure 3D), and K14 immunofluorescence was intense in the hair matrix of anagen hair bulb (Figure 3_a, f).

From these findings, we clarified the spatial expression of key genes associated with HF cycle development on pig hair follicles cycle. At the same time, these results help division of the pig HF cycle.

To further investigate the expression of important genes affecting anagen, catagen and telogen in the pig HF cycle, some important genes related to HF cyclic development, such as Wnt10b, Wnt5a, LEF1, SHH LHX2, and α-SMA were selected for mRNA expression assays. The expression of Wnt10b (Figure 4A) and Wnt5a (Figure 4B) are significantly decreased when HF entered the catagen, and the expression of LEF1 (Figure 4C), SHH (Figure 4D), and LHX2 (Figure 4E) decreased significantly in the catagen compared to the anagen.

The results of gene expression during the telogen and anagen phases revealed that the expression of Wnt10b, Wnt5a, LEF1, LHX2, and SHH was significantly higher in anagen than in telogen (Figure 4). However, there was no significant difference in the expression of these genes between the catagen and telogen. It suggests that these genes may regulate the transition of hair follicles from anagen to catagen phase, rather than continuing to maintain the degenerative process. Additionally, the expression of α-SMA (Figure 4F) gene was significantly higher in anagen or telogen than in catagen, which may be related to the agglutination of DPs.

Furthermore, as classical markers of the hair follicle, SOX9 and K14 play important roles in HF cycle, and we also examined and quantified their protein expression using Western-bolt (Supplementary Figure S3). The results showed that the protein expression of SOX9 and K14 was significantly higher in the anagen than in the catagen and telogen. The result also helps to illustrate the changes of HF matrix in the HF cycle.