1. Npytet/tet skin is fibrotic with hyperkeratosis
To interrogate whether skin pathologies are present in Npytet/tet mice, we performed histopathological evaluations on the dorsal skin of 22-week-old (i.e., pre-graying stage; [9]) and 35-week-old (i.e., post-graying stage; [9]) Npy+/+ and Npytet/tet mice (Figure 1). 22-week-old Npytet/tet skin appears to exhibit epidermal thickening with hyperkeratosis, compared to their Npy+/+ littermates, as evidenced by hematoxylin and eosin (H&E) stained sections (Figure 1A). Following this observation, we quantified the thickness of each layer of the skin (epidermis, dermis, and subcutaneous adipose tissue). Between Npy+/+ and Npytet/tet mice, no difference was found in the thickness of the whole skin, dermis, or subcutaneous adipose tissue, however, the epidermis of 22-week-old Npytet/tet skin is significantly thicker than that of Npy+/+ skin (Figure 1B). At 35 weeks old, hyperkeratosis (Figure 1C, D) in Npytet/tet skin persists, and fibrosis, evidenced by an increased abundance of pink collagen fibers, is now observed in the dermis. No differences between the genotypes are observed in the thickness of the dermis or subcutaneous adipose tissue at this timepoint. These findings indicate that chronic overexpression of NPY in the skin is sufficient to induce functional changes in epidermal keratinocytes and dermal fibroblasts that is pathological.
2. Npytet/tet skin is infiltrated by macrophages and T cells
Following the identification of dermal fibrosis and epidermal thickening in Npytet/tet skin, we questioned what mechanisms would contribute to these NPY-induced pathologies. Knowing that NPY has diverse functions on immune cells, including pro-inflammatory and anti-inflammatory effects [13, reviewed in (17), 30], we hypothesized that NPY can induce macrophage infiltration and can increase T cell proliferation in the skin [14, 31-34]. Accordingly, we assessed immunolabeled macrophages (F4/80) and T cells (CD3) to determine the average number of these immune cells within the skin of 22- and 35-week-old mice (Figures 2 and 3). Macrophages are significantly more abundant in the skin of 22-week-old Npytet/tet mice when compared to their Npy+/+ littermates (Figure 2A,B). This accumulation of macrophages is observed in the dermis and subcutaneous adipose tissue of Npytet/tet skin. At 35 weeks of age, chronic NPY overexpression induces an increase in macrophages in the dermis of Npytet/tet skin, but this increase was only statistically significant at a less stringent p-value of 0.1 (Figure 2C,D).
We previously reported no overt qualitative changes in the abundance of CD4+ and CD8+ T cells in Npytet/tet skin [29], yet using the pan T cell marker and quantitative assessment we find that T cells are significantly more abundant in Npytet/tet skin compared to Npy+/+ littermates at 22 and 32 weeks old. This is attributable to T cell accumulation in the dermis and a trend for their accumulation in the adipose tissue (Figure 3A,B). This pattern persists in 35-week-old Npytet/tet skin (Figure 3C,D). To further characterize the T cell pool, we assessed for the abundance of regulatory T cells (Tregs) via FOXP3 expression. In 22-week-old skin, Tregs are significantly more abundant in Npytet/tet skin, which is attributed to their significant accumulation in the dermis (Figure 4A,B). This persists in 35-week-old skin, along with a trend for their accumulation in the adipose tissue (Figure 4C,D). Tregs are known to express FOXP3 as well as CD4 [35]. Although we previously observed no qualitative differences in CD4+ T cell abundance [29], the improved quality of the paraffin-embedded tissues used in the current study compared to the frozen tissue used in our previous study revealed quantitative differences in T cells expressing the FOXP3 marker for Tregs. These findings suggest that chronic NPY overexpression in the skin can induce infiltration or proliferation of macrophages and T cells, specifically Tregs, resulting in a chronically inflamed environment within the skin.
NPY has also been shown to induce mast cell activation [36]. Accordingly, we assessed for NPY-mediated changes in mast cell numbers within 22-week-old Npy+/+ and Npytet/tet skin and found that mast cells are similarly abundant in the skin of both genotypes (Supplementary Figure 5). Pathological phenotypes, by H&E staining and immunolabeling, are also observed in heterozygous Npytet/+ mice and these phenotypes are generally similar to Npytet/tet mice or are intermediate between those observed in Npy+/+ and Npytet/tet mice (Supplementary Figures 1-4).
3. Transcriptional changes in 22-week-old Npytet/tet skin further supports inflammation as a major response to chronic NPY overexpression and a contributing factor to NPY-mediated skin pathologies
To identify transcriptional changes in the skin that might elaborate the mechanisms by which NPY induces skin pathology and inflammation, we harvested whole skin from 22-week-old Npy+/+ and Npytet/tet mice to be assessed by RNA sequencing (RNAseq) (Figure 5A). Npy upregulation observed by RNAseq was independently validated in skin from a different anatomical region of the same mice via qPCR (Figure 5B). From this RNAseq data we made several comparisons; all Npytet/tet versus all Npy+/+ samples, Npytet/tet versus Npy+/+ samples based on sex, an intra-genotype evaluation of Npytet/tet samples (Figure 5C). When comparing Npytet/tet animals to Npy+/+ littermates, a short list of 32 differentially expressed genes (DEGs; -1 > log2FC < 1, padj < 0.05) was obtained. Six of these DEGs were upregulated. When comparing gene signatures specific to a certain cell type and biological process were most apparent in the downregulated genes. Twenty-six of 32 DEGs were downregulated— 11 of these DEGs were related to melanogenesis and pigmentation, such as Dct, Mc1r, Pmel, and Tyr, and 4 of these DEGs are usually expressed in anagen-stage mouse hair follicle keratinocytes, including Dlx2, Dlx3, Msx2, Mycn (Figure 5D) [37,38].
By principal component analysis (PCA), we noticed that our RNAseq samples clustered somewhat by sex rather than by genotype (Supplementary Figure 6). Thus, repeating the differential expression analysis while only comparing Npytet/tet to Npy+/+ animals of the same sex revealed some of the same DEGs as above along with additional DEGs not apparent when the samples of both sexes were evaluated together (see overlaps in Figure 5D). Interestingly, comparing male mice only (Npytet/tet males (n=2) versus Npy+/+ males (n=3)) showed an upregulation of Retnla, a gene which participates in the negative regulation of Th2 responses and induction of fibrosis and is upregulated in response to IL-4 and IL-13 in macrophages [39-41]. Npytet/tet males also downregulated Skint3 and Skint9. These genes are expressed by epidermal keratinocytes to signal dendritic epithelial T cells to promote proper skin wound healing and are downregulated during skin aging [42]. On the other hand, comparing Npytet/tet and Npy+/+ females only [Npytet/tet females (n=3) versus Npy+/+ females (n=2)], we discovered an abundance of new DEGs only significant in this sex (67 total). This included 9 upregulated genes, 4 of which are involved in adipogenesis and lipogenesis, Sult1e1, Thrsp, Fasn, Agpat2 [43-45], and one Pla2g2d, that is known as a pro-resolving lipid mediator that works to terminate inflammation [46]. The remainder of the DEGs downregulated by Npytet/tet females are enriched in genes expressed by various layers of anagen-stage hair follicles, including Cryba4, Edn2, Fbp1, Fgf5, Foxe1, Krt28, Krt31, Krt33a, Krt35, Krt71, Krtap11-1, Krtap15, Padi3, S100a3, Sct, Scl39a8, and Tchh. It is unclear, however, whether this downregulation of hair follicle-related genes is reflective of an NPY-related pathology or differences in the hair stage of Npytet/tet and Npy+/+ females at harvest as only a subset of these animals was evaluated for histopathology.
Previously we reported that the onset of macroscopic hair graying in the Npytet/tet mice initiates at 25-27 weeks [29], yet this timeframe is variable as graying was visible in a subset of this cohort of mice as early as 22 weeks (Figure 5A). Interested in interrogating transcriptomic signatures that might explain the accelerated onset of graying in the “gray” subset of Npytet/tet mice, we reanalyzed our RNAseq data from the 22-week-old skin to compare the “gray” and “non-gray” Npytet/tet mice. Interestingly, the gray subset is comprised of male Npytet/tet mice, while the non-gray subset contains female Npytet/tet mice. By PCA analysis, Npytet/tet animals clustered independently based on both graying and sex with 71% variance across principal component 1 (PC1) (Figure 6A). Npy+/+ animals also cluster independently across PC1 based on sex, yet with only 57% variance. This suggests that the additional variance seen between Npytet/tet animals across PC1 may be explained by DEGs that are specific to graying. It is important to note that although the male Npytet/tet mice of this particular 22-week-old cohort exhibited hair graying before their female counterparts, this male-first presentation of the phenotype is not consistent across multiple litters (data not shown).
To identify the genes that contribute to graying onset within the gray cohort of Npytet/tet mice, we first removed any DEGs that could be considered sex-associated and may be confounding. First, differential expression analysis between the sexes was performed for each genotype and a list containing both upregulated and downregulated genes for each contrast were generated (Npytet/tet males versus Npytet/tet females = 918 DEGS; and Npy+/+ males versus Npy+/+ females = 303 DEGs). Next, we overlapped these two DEG lists and removed the 138 genes which were present in both lists and which we deemed to be common, sex-associated genes (Figure 6B). The remaining 780 DEGs within the Npytet/tet group were evaluated via gene set enrichment analysis (GSEA) to investigate the transcriptomic differences between gray (i.e., male) and non-gray (i.e., female) Npytet/tet skin.
GSEA to identify the biological processes upregulated in non-gray Npytet/tet skin indicates that immune system development is the second-most enriched process. Specifically, the processes of lymphocyte activation, interleukin-4 (IL-4) production, and cell proliferation are among the most enriched and match expectations based on known NPY signaling. Additionally, processes such as thermogenesis and neurogenesis are enriched in non-gray Npytet/tet skin (Supplementary File). Conversely, GSEA shows that epidermal cell differentiation is the third-most enriched process in gray Npytet/tet skin. Additionally, myeloid leukocyte (i.e., macrophage) activation and response to cytokines are enriched in the skin of gray Npytet/tet mice (Figure 6C). Despite the gray Npytet/tet mice clearly exhibiting a more progressed graying phenotype, the expression of pigment genes is not consistently up- or down-regulated in either subset of the Npytet/tet group. Interestingly, genes related to T cells and B cells are more enriched in the non-gray Npytet/tet skin, while genes related to macrophages are enriched in the gray Npytet/tet skin (Figure 6D). This observation is further supported by GSEA for the cell types that are enriched in the skin of each subset of the Npytet/tet genotype. The transcriptome of non-gray Npytet/tet skin shows high enrichment for cells of neural origin (Supplementary File), as well as enrichment for fibroblasts and immune cells, more specifically B cells (Figure 6E). This contrasts with gray Npytet/tet skin, in which dendritic cells, proliferating macrophages, and proliferating basal cells are within the 10 most enriched cell types (Figure 6E). These changes in transcriptomic signatures from non-gray to gray may reflect a progression in cellular pathology that ultimately drives the more severe hypopigmentation observed in 35-week-old mice.
GSEA to identify human phenotypes with similar gene enrichment profiles indicates that abnormal function in the human gastrointestinal tract and personality abnormalities, including depressivity, are among the only 5 human phenotypes with similar gene enrichment as the non-gray Npytet/tet skin, the former of which could be due to the upregulation of immune cell activation in non-gray skin (Figure 6F). The same GSEA shows that gray Npytet/tet skin gene enrichment is similar to that of multiple hair abnormalities, with alopecia being the most similar human phenotype (Supplementary File). Notably, gray Npytet/tet skin is highly enriched for abnormal hair growth, thickened skin, and hyperkeratosis (Figure 6F), which match our histopathological findings (Figure 1).