Experimental animals were obtained from the Inner Mongolia Jinlai Animal Husbandry Technology (Hohhot, China). Skin samples were obtained from cashmere goat embryos at 45, 55, 65, and 75 days. These samples were stored at −80°C.

Based on the previously constructed skin transcriptome database of cashmere goats at different embryonic stages (45 days, 55 days, 65 days, and 75 days) [13], we analyzed the differentially expressed lncRNAs across these stages using the criteria of |log2foldchange|≥1 and p-value≤0.05 as selection thresholds. Previous research by our team has revealed that, at 45 days of embryonic development, the skin of cashmere goats possesses only a complete epidermal structure, lacking any hair follicle structures. Primary hair follicles initiate their development at 55 days of embryonic development and continue to extend towards the dermal layer as the fetus develops, ultimately resulting in the formation of a complete primary hair follicle structure before birth. The development of secondary hair follicles lags behind that of primary hair follicles, initiating their formation at 75 days of embryonic development and continuing to develop until a complete secondary hair follicle structure is formed before birth [7,8]. According to the characteristics of hair follicle development described above, embryonic day 55 was considered to be the key point for primary hair follicle morphogenesis, and embryonic day 75 was considered to be the key point for secondary hair follicle morphogenesis. We classified three comparison groups (d55 vs d45, d65 vs d45, and d65 vs d55) as Stage 1, which is associated with the growth and development of primary hair follicles. The other three comparison groups (d75 vs d45, d75 vs d55, and d75 vs d65) were classified as Stage 2, which is related to both primary and secondary hair follicle development. Subsequently, we took the intersection of Stage 1 (primary hair follicles) and Stage 2 (primary hair follicles+ secondary hair follicles) and excluded this intersection from Stage 2. The remaining lncRNAs in Stage 2 were considered important lncRNAs related to the morphogenesis of secondary hair follicles.

Dermal fibroblasts were cultured in a medium containing 10% fetal bovine serum. The culture conditions for all cell lines were maintained at 37°C with 5% CO₂. Lentiviral transfection was employed to deliver the target plasmid into dermal fibroblasts, and the target plasmid was prepared by Shanghai Hanheng Biotechnology (Shanghai, China). After transfection, puromycin was added to perform antibiotic selection.

LncLocator [14] predicted lncRNA ST6GALNAC3’s location. Cytoplasmic and nuclear RNA from dermal fibroblasts were extracted using the Cytoplasmic & Nuclear RNA Purification Kit (Norgen Biotek, Thorold, ON, Canada), followed by quantitative realtime reverse transcription polymerase chain reaction (qRT-PCR) to detect lncRNA ST6GALNAC3 expression in these locations.

Prepare the cells to the appropriate density. The cells were then treated with 4% paraformaldehyde for 20 min, prehybridization solution for 1 h, and then incubated with a probe mix overnight. Subsequently, the prehybridization solution was washed off, and blocking serum (BSA) was added. The cells were incubated with the BSA for 30 min at room temperature. Finally, the cells were incubated with DAPI staining solution for 8 min, washed to remove the staining solution, and observed under a fluorescence microscope.

Sequence information of lncRNA and mRNA was obtained using the previously constructed cashmere goat skin transcriptome database. The sequence information of miRNA was obtained from the miRBase database. In this study, miRanda [15] and TargetScan [16] software were used to predict the target miRNAs of lncRNA and target genes of miRNA, respectively. miRanda [15] is mainly based on the free binding energy between lncRNA and miRNA, while TargetScan [16] is primarily based on seed sequence to predict the targeting relationship between mRNA and miRNA. The sequence information of lncRNA and miRNA was input into the miRanda [15] software. When the threshold was less than −10, it indicated a targeted binding relationship between them (the smaller the threshold, the greater the possibility of interaction between them). The sequence information of mRNA and miRNA was input into the TargetScan [16] software. When the TargetScan threshold exceeded 50, it was inferred that there might be a targeted binding relationship between them (the higher the threshold, the greater the likelihood of interaction between them). Subsequently, the RNAhybrid software (v2.1.2) was employed to validate the lncRNA-miRNA and miRNA-mRNA targeting relationships identified earlier. When the binding free energy of a targeting relationship pair was less than −20 kcal/mol, we preliminarily believed that there was a strong binding ability between them (the lower the free binding energy, the stronger the binding ability).

The dual-luciferase reporter system was used to verify the targeting relationship in 293T cells. Cells were co-transfected with chi-miR-24-3p mimic and either psiCHECK2-lncRNA ST6GALNAC3-WT/MUT or psiCHECK2-ID4-WT/MUT using LipoFiter (Hanheng, Shanghai, China). Luciferase activity was measured 48 h post-transfection (Promega, Madison, WI, USA).

Cell proliferation was assessed by CCK8 and EdU methods. CCK8 measured proliferation using a kit (Solarbio, Beijing, China), with OD read at 450 nm. The EdU assay was performed based on the instructions of the BeyoClick EdU-555 Cell Proliferation Detection Kit (Beyotime, Shanghai, China). The EdU assay calculates proliferation as the ratio of EdU-labeled cells to Hoechst-labeled cells.

Cell apoptosis was determined using the Annexin V-APC/PI Kit (Elabscience Biotechnology, Wuhan, China). A single-cell suspension was prepared and stained with reagents. After incubating in the dark for 20 min, cell apoptosis was detected by flow cytometry.

The cell cycle was determined by the DNA content quantification method (Solarbio). Cells were fixed in 70% ethanol, resuspended, and incubated at 4°C. After 24 h, RNase A was added, and cells were incubated at 37°C for 30 min. The PI staining solution was then added, and cells were incubated at 4°C in the dark for 30 min.

The cell scratch assay involved scraping a cell monolayer and capturing images at 0 h and 24 h post-scratch. Cells were washed with PBS, added fresh serum-free medium, and incubated. Cell migration rate was calculated as ([0 h area–24 h area]/0 h area)×100%.

Statistical analysis was conducted using SPSS 26.0 (IBM, Armonk, NY, USA) and GraphPad Prism 8.0 (GraphPad Software, San Diego, CA, USA). All data are presented as mean±SD from ≥3 independent experiments. Significance was determined by Student’s t-tests with p<0.05 or p<0.01.

To identify lncRNAs related to the morphogenesis of secondary hair follicles in cashmere goats, we screened and analyzed the transcriptome database of skin tissues from different embryonic stages that had been previously established [13]. Initially, we analyzed the differentially expressed lncRNAs across these stages using the criteria of p-value≤0.05 and |log2foldchange|≥1. Through differential comparison analysis, we identified a total of 1,209 lncRNAs that were differentially expressed in skin tissues at different embryonic stages. Previous research by our team has revealed that, at 45 days of embryonic development, the skin of cashmere goats possesses only a complete epidermal structure, lacking any hair follicle structures. Primary hair follicles initiate their development at 55 days of embryonic development and continue to extend towards the dermal layer as the fetus develops, ultimately resulting in the formation of a complete primary hair follicle structure before birth. The development of secondary hair follicles lags behind that of primary hair follicles, initiating their formation at 75 days of embryonic development and continuing to develop until a complete secondary hair follicle structure is formed before birth [7,8]. The morphogenesis of secondary hair follicles determines the number of secondary hair follicles in cashmere goats, which subsequently influences the production of cashmere [5]. Therefore, based on the characteristics of hair follicle morphological development in the embryonic period, we considered embryonic day 55 as the key time node for primary hair follicle morphogenesis and embryonic day 75 for secondary hair follicle morphogenesis. We designated three comparison groups (d55 vs d45, d65 vs d45, d65 vs d55) as Stage 1, which encompassed a total of 1,051 lncRNAs associated with the development of primary hair follicles. Additionally, we designated the other three comparison groups (d75 vs d45, d75 vs d55, d75 vs d65) as Stage 2, which encompassed a total of 903 lncRNAs related to the morphogenesis of both primary and secondary hair follicles. To identify lncRNAs specifically associated with secondary hair follicle morphogenesis, we first determined the intersection of lncRNAs present in both Stage 1 and Stage 2. These 745 overlapping lncRNAs are likely involved in processes common to both primary and secondary hair follicle development. By excluding these 745 lncRNAs from the 903 lncRNAs in Stage 2, we obtained a set of 158 lncRNAs that were considered important for the morphogenesis of secondary hair follicles (Figure 1A). Among these 158 lncRNAs, we discovered that lncRNA ST6GALNAC3 was highly expressed in samples at 45, 55, and 65 days of the embryonic period, while it was down-regulated at 75 days of the embryonic period. This lncRNA exhibits good biological reproducibility. The 75th day of the embryonic stage is a crucial period for the morphogenesis of secondary hair follicles in cashmere goats. Therefore, we further examined the expression of lncRNA ST6GALNAC3 in skin tissues at different periods using qRT-PCR. Consistent with the sequencing results, the expression of lncRNA ST6GALNAC3 was significantly lower at embryonic day 75 compared with other stages (Figure 1D). Thus, we speculate that this lncRNA may be related to the morphogenesis of secondary hair follicles.

Simultaneously, we further examined the coding potential and cellular distribution of lncRNA ST6GALNAC3. To analyze the coding ability of lncRNA ST6GALNAC3, we utilized CPC [17] and CNCI [18] software. The results showed CPC [17] and CNCI [18] scores of 0.1 and −0.02, respectively, indicating that lncRNA ST6GALNAC3 lacks protein-coding capacity. LncLocator [14] software predicted that lncRNA ST6GALNAC3 is primarily located in the cytoplasm (Figure 1B). Subsequently, we utilized FISH and nuclear-cytoplasmic separation experiments to detect the distribution of lncRNA ST6GALNAC3 in dermal fibroblasts. The experimental results were consistent with the predictions, revealing that this lncRNA is mainly localized in the cytoplasm of dermal fibroblasts (Figures 1C, 1E).

The morphogenesis and development of hair follicles result from the continuous proliferation and differentiation of dermal fibroblasts and epithelial cells. Among them, dermal fibroblasts ultimately differentiate to form the dermal papilla structure, which serves as the signaling hub for hair follicle growth and regulates the development of the entire hair follicle. Therefore, we utilized lentiviral transfection to introduce lncRNA ST6GALNAC3-sh1/2/3 vectors into dermal fibroblasts and subsequently evaluated their interference efficiency. We found that the lncRNA ST6GALNAC3-sh3 vector exhibited the best interference efficiency, and thus selected it for subsequent cellular phenotypic experiments (Figure 2A).

Firstly, we utilized EdU assays and CCK8 methods to examine the impact of lncRNA ST6GALNAC3 interference on the proliferation of dermal fibroblasts. The results of the cell proliferation experiments indicated that lncRNA ST6GALNAC3-sh could increase the proportion of EdU-positive cells in dermal fibroblasts, and the proliferative capacity of these cells was also significantly enhanced compared to that of the control group (Figures 2B, 2C). Subsequently, flow cytometry was used to detect changes in the cell cycle and apoptosis, further investigating the role of lncRNA ST6GALNAC3 in promoting cell proliferation. The experimental results indicated that the knockdown of lncRNA ST6GALNAC3 significantly decreased the apoptotic rate of dermal fibroblasts while significantly increasing the proportion of cells in the S phase of the cell cycle (Figures 2E, 2F). Additionally, we further examined the expression of genes related to proliferation and apoptosis following lncRNA ST6GALNAC3 knockdown. qRT-PCR results demonstrated that the lncRNA ST6GALNAC3-sh could upregulate the expression of CCND1, CCNE1, CCND2, and PCNA (proliferation-related genes), while downregulating the expression of Bax and Casp9 (apoptosis-related genes) (Figure 2D). In summary, lncRNA ST6GALNAC3-sh can promote the proliferation of dermal fibroblasts, and this promotional effect may be mediated by increasing the proportion of cells in the S phase and enhancing the expression of cell proliferation-related genes. Dermal fibroblasts not only continuously proliferate and differentiate but also migrate further into the dermis of the skin, ultimately forming the dermal papilla structure at the base of the hair follicle. Therefore, this study utilized a cell scratch assay to further investigate the impact of lncRNA ST6GALNAC3 on cellular migration ability (Figure 2G). The results showed that the migration ability of dermal fibroblasts was significantly increased following transfection with the lncRNA ST6GALNAC3-sh. In conclusion, the lncRNA ST6GALNAC3-sh enhances the proliferation and migration ability of dermal fibroblasts, thereby regulating the formation of the dermal papilla.

Numerous studies on lncRNAs have demonstrated that their intracellular localization is closely associated with their regulatory functions. Typically, lncRNAs located in the cytoplasm often indirectly regulate the expression of related genes by sponging miRNAs. Preliminary experimental results suggest that lncRNA ST6GALNAC3 is mainly located in the cytoplasm of cells. Consequently, we hypothesize that lncRNA ST6GALNAC3 may act as a ceRNA to regulate the expression of related genes. Firstly, we conducted bioinformatics analysis utilizing miRanda [15] and found potential binding sites between the lncRNA ST6GALNAC3 and chi-miR-24-3p. Furthermore, by employing the RNAhybrid software (v2.1.2), we acquired sequence information about the binding sites between lncRNA ST6GALNAC3 and chi-miR-24-3p. The binding free energy between the two entities was computed to be lower than −20 kcal/mol, which signifies a robust binding capacity (Figure 3C). At the same time, we predicted the target genes of chi-miR-24-3p and found that it has a potential binding capacity with 103 mRNAs. Figure 3A displays their regulatory network. Next, we did enrichment analysis on chi-miR-24-3p’s target genes (Figures 3B, 3E). The results showed these genes were mainly in pathways related to hair follicle development, like TGF-β, Wnt, and PI3K-AKT. This further indicates that lncRNA ST6GALNAC3 can act as a ceRNA to regulate the expression of genes related to hair follicle development, and then indirectly participate in the process of hair follicle development. Among these target genes, we selected the ID4 gene, which is enriched in the TGF-β signaling pathway, for subsequent experimental validation and further analysis. The prediction results from RNAhybrid (v2.1.2) indicate that the binding free energy between ID4 and chi-miR-24-3p is −28 kcal/mol, which is less than −20 kcal/mol, suggesting a strong binding capacity between them (Figure 3D). Subsequently, we employed qRT-PCR to examine the expression levels of lncRNA ST6GALNAC3 and ID4 in cell lines with chi-miR-24-3p knockdown/overexpression. The experimental results showed that both lncRNA ST6GALNAC3 and ID4 were highly expressed in the chi-miR-24-3p knockdown cell lines, whereas they were lowly expressed in the chi-miR-24-3p overexpression cell lines (Figure 3F). This phenomenon is consistent with the ceRNA hypothesis. Furthermore, we constructed wild-type and mutant dual-luciferase reporter plasmids for lncRNA ST6GALNAC3 and ID4-3’UTR, respectively, and employed the dual-luciferase reporter gene system to detect the targeting relationships between lncRNA ST6GALNAC3 and chi-miR-24-3p, as well as between the ID4-3’UTR and chi-miR-24-3p. The results showed that the chi-miR-24-3p mimic significantly decreased the luciferase activity of both the wild-type lncRNA ST6GALNAC3 reporter gene and the wild-type ID4-3’UTR reporter gene, whereas there was no significant change in the mutant versions, indicating that both lncRNA ST6GALNAC3 and ID4 can bind to chi-miR-24-3p (Figures 3G, 3H).

Subsequently, we constructed three ID4 interference vectors and transfected them into dermal fibroblasts with lentiviral transfection. The experiment found that ID4-sh2 had the best interference efficiency, so we chose this cell line for the next experiments (Figure 4A). Firstly, we used methods such as CCK8, EdU, and Annexin V-APC/PI double staining to investigate the effects of ID4 interference on the proliferation and apoptotic abilities of dermal fibroblasts. The experimental results showed that knocking down ID4 can promote the proliferation of dermal fibroblasts and inhibit their apoptosis (Figures 4B–4D). These results were similar to those obtained following lncRNA ST6GALNAC3 knockdown, in line with the ceRNA hypothesis. Furthermore, experimental results from cell cycle analysis indicated that knocking down ID4 increased the proportion of dermal fibroblasts in the S phase (Figure 4G). Additionally, qRT-PCR was employed to detect changes in the expression of marker genes related to cell proliferation and apoptosis in the cell lines. Our findings revealed that ID4-sh enhanced the expression of cell proliferation-related marker genes and anti-apoptosis marker genes, while simultaneously inhibiting the expression of apoptosis-promoting genes (Figure 4F). Experimental results from cell migration assays showed that the migration ability of dermal fibroblasts was significantly enhanced after transfection with ID4-sh (Figure 4E). Based on the aforementioned experimental results, knocking down ID4 promoted the proliferation and migration abilities of dermal fibroblasts. This promotional effect was achieved by increasing the proportion of cells in the S phase and upregulating the expression of cell proliferation-related genes. In addition, we noted that the effects of ID4-sh and lncRNA ST6GALNAC3-sh on the phenotypes of dermal fibroblasts were consistent, which phenomenon is in line with the ceRNA hypothesis.

Furthermore, to validate the lncRNA ST6GALNAC3-chi-miR-24-3p-ID4 regulatory network, we added chi-miR-24-3p inhibitor to both the lncRNA ST6GALNAC3-sh and ID4-sh cell lines and examined the resulting changes in cellular phenotypes. Previous experimental results demonstrated that both lncRNA ST6GALNAC3-sh and ID4-sh could enhance the proliferation and migration of dermal fibroblasts while inhibiting their apoptosis. After adding chi-miR-24-3p inhibitor to the lncRNA ST6GALNAC3-sh or ID4-sh cell lines, we observed a significant reduction in the proportion of EdU-positive cells in the co-transfected group compared to the singly transfected groups (Figure 5A). Simultaneously, the proportion of cells in the S phase and the expression levels of cell proliferation marker genes also significantly decreased. This indicates that the chi-miR-24-3p inhibitor can counteract the proliferation-promoting effects of lncRNA ST6GALNAC3-sh and ID4-sh on dermal fibroblasts (Figures 5B,6B). Next, we investigated cell apoptosis. The results of the apoptosis experiments demonstrated that, after adding chi-miR-24-3p inhibitor to the lncRNA ST6GALNAC3-sh and ID4-sh cell lines, there was a significant increase in the proportion of apoptotic cells, with the most pronounced increase observed in the proportion of early apoptotic cells (Figure 6A). The cell migration ability was evaluated through the cell scratch assay. As shown in Figure 5C, with the addition of chi-miR-24-3p inhibitor to the lncRNA ST6GALNAC3-sh and ID4-sh cell lines, the migration-promoting effect induced in these cell lines was subsequently counteracted.

The comprehensive results indicate that chi-miR-24-3p, as a common target miRNA for both lncRNA ST6GALNAC3 and ID4, can rescue the proliferation and migration-promoting effects caused by lncRNA ST6GALNAC3-sh and ID4-sh. This suggests the existence of an lncRNA ST6GALNAC3-chi-miR-24-3p-ID4 regulatory axis in dermal fibroblasts. Additionally, in dermal fibroblasts, lncRNA ST6GALNAC3 can act as a sponge for chi-miR-24-3p, regulating the expression of ID4, and subsequently influencing the proliferation and migration abilities of dermal fibroblasts, thereby modulating the morphogenesis of secondary hair follicles (Figure 7).