Molecular stratification of the human fetal vaginal epithelium by spatial transcriptome analysis

Abstract

The human vaginal epithelium is a crucial component of numerous reproductive processes and serves as a vital protective barrier against pathogenic invasion. Despite its significance, a comprehensive exploration of its molecular profiles, including molecular expression and distribution across its multiple layers, has not been performed. In this study, we perform a spatial transcriptomic analysis within the vaginal wall of human fetuses to fill this knowledge gap. We successfully categorize the vaginal epithelium into four distinct zones based on transcriptomic profiles and anatomical features. This approach reveals unique transcriptomic signatures within these regions, allowing us to identify differentially expressed genes and uncover novel markers for distinct regions of the vaginal epithelium. Additionally, our findings highlight the varied expressions of keratin ( KRT) genes across different zones of the vaginal epithelium, with a gradual shift in expression patterns observed from the basal layer to the surface/superficial layer. This suggests a potential differentiation trajectory of the human vaginal epithelium, shedding light on the dynamic nature of this tissue. Furthermore, abundant biological processes are found to be enriched in the basal zone by KEGG pathway analysis, indicating an active state of the basal zone cells. Subsequently, the expressions of latent stem cell markers in the basal zone are identified. In summary, our research provides a crucial understanding of human vaginal epithelial cells and the complex mechanisms of the vaginal mucosa, with potential applications in vaginal reconstruction and drug delivery, making this atlas a valuable tool for future research in women's health and reproductive medicine.

Keywords: keratin; spatial transcriptomics; vaginal epithelium.

Figure 1

Spatial transcriptome analysis of the human fetal vagina (A) Spatial transcriptomic analysis workflow of the human fetal inferior/lower vagina. (B) H&E-stained cross-sections of vaginas from fetuses at 22 weeks of age near the orifice or introitus (left: V22W-5; right: V22W-6). (C) Spatial transcriptomics results mapped onto the two vaginal sections (left: V22W-5; right: V22W-6). (D) UMAP clustering of cell-covered spots from two tissue sections, resulting in 9 unique clusters; spots within the same cluster were color-coded identically. (E) Cluster abundance analysis of V22W-5 and V22W-6 based on spot numbers.

Figure 2

Spatially variable features of the human fetal vagina revealed by differential gene expression analysis (A) Heatmap presenting the top 10 DEGs per cluster, sorted by log 2(fold change). For clusters with fewer than 10 genes, all DEGs are displayed. (B) Variability in the number of DEGs across different clusters.

Figure 3

Zone-specific marker expression in the human fetal vaginal epithelium (A) Left: violin plot depicting the spatial distribution of DEGs, including LCE3E, FLG, SPINK7, C15orf48, PI3, and IGSF9. Right: spatial mapping further illustrates the specific zones where these genes are predominantly expressed. (B) IHC staining of LCE3E (red triangle arrow) and FLG (red triangle arrow) in fetus vagina. (C) IHC staining of SPINK7 and C15orf48 (red triangle arrow) in the adult vagina. (D) IHC staining of PI3 in the adult vagina. Notably, there are PI3-negative cells in the basal layer. (E) IHC staining of IGSF9 in the adult vagina. Scale bars are provided in the panels, with units in micrometers (μm). Notably, SPINK7, C15orf48, PI3, and IGSF9 data were cross-validated using the Human Protein Atlas (HPA) for additional confirmation (see Methods for details). Scale bars are shown in μm.

Figure 4

Cluster assignment using cell type-specific markers and constraints from histological characteristics (A) Heatmap displaying the relative expression levels of canonical markers in each cluster. (B) Spatial distribution features of markers (IVL, KRT5, CDH1, TGM1, and VIM) in the human fetal vagina illustrated through spatial mapping and violin plots. (C) Verification of markers expressed in vaginal epithelia by immunofluorescence (KRT5) and immunohistochemical staining (IVL, CDH1, TGM1, and VIM) in the fetal vagina. (D) IHC staining (left, middle) and spatial mapping of CD207 in the basal layer (red triangular arrow) in the fetal vagina. (E) IHC staining of PGR and ESR1 in the human fetal vagina. Scale bars are shown in μm.

Figure 5

Keratin gene analysis reveals heterogeneity of the vaginal epithelium (A) Heatmap analysis illustrating the zone-specific dominant keratin family genes across the vaginal epithelium. (B) Left: spatial mapping of KRT13 across the vaginal epithelium. Right: IHC staining of KRT13 in fetal vagina. (C) Spatial mapping and IHC staining of KRT14 in fetal vagina. (D) Spatial mapping of KRT78 and IHC staining of KRT78 in fetal vagina. (E) Spatial mapping of KRT17 and IHC staining of KRT17 in fetal vagina. KRT17 positivity (red arrow); KRT17 negativity (green arrow). (F) Spatial mapping of KRT6A, KRT6B, and KRT6C. Scale bars are shown in μm.

Figure 6

KEGG analysis reveals diverse pathways associated with various epithelial zones Bubble plot presenting all the KEGG pathways to elucidate the pathways active within each region of the vaginal epithelium. The colors represent the ‒log 10(q-value), and the bubble size indicates the number of genes enriched within each pathway.

Figure 7

Potential progenitor epithelial cells in the vagina (A) IHC staining for proliferation marker (MKI67) and stem cell-related markers (NGFR, AXIN2, COL17A1, and SOX2) in the active basal layer of the fetal vaginal epithelium. (B) Immunofluorescence (IF) staining of MKI67 and SOX2 in the fetal vagina. (C) IF staining of MKI67 and TP63 in the fetal vagina. Scale bars are shown in μm. (D) Spatial mapping of MKI67, NGFR, AXIN2, SOX2 and COL17A1.