Exploring the role of differentially expressed metabolic genes and their mechanisms in bone metastatic prostate cancer

Abstract

Background: Approximately 10-20% of patients diagnosed with prostate cancer (PCa) evolve into castration-resistant prostate cancer (CRPC), while nearly 90% of patients with metastatic CRPC (mCRPC) exhibit osseous metastases (BM). These BM are intimately correlated with the stability of the tumour microenvironment.

Purpose: This study aspires to uncover the metabolism-related genes and the underlying mechanisms responsible for bone metastatic prostate cancer (BMPCa).

Methods: Gene Expression Omnibus (GEO) and The Cancer Genome Atlas (TCGA) datasets of PCa and BM were analyzed through R Studio software to identify differentially expressed genes (DEGs). The DEGs underwent functional enrichment via Kyoto Encyclopedia of Genes and Genomes (KEGG) and Gene Ontology (GO), with key factors screened by a random forest utilized to establish a prognostic model for PCa. The study explored the relationship between DEGs and the stability of the immune microenvironment. The action and specificity of CRISP3 in PCa was validated through western blot analysis, CCK-8 assay, scratch assay, and cellular assay.

Results: The screening of GEO and TCGA datasets resulted in the identification of 199 co-differential genes. Three DEGs, including DES, HBB, and SLPI, were selected by random forest classification model and cox regression model. Immuno-infiltration analysis disclosed that a higher infiltration of naïve B cells and resting CD4 memory T cells occurred in the high-expression group of DES, whereas infiltration of resting M1 macrophages and NK cells was greater in the low-expression group of DES. A significant infiltration of neutrophils was observed in the high-expression group of HBB, while greater infiltration of gamma delta T cells and M1 macrophages was noted in the low-expression group of HBB. Resting dendritic cells, CD8 T cells, and resting T regulatory cells (Tregs) infiltrated significantly in the high-expression group of SLPI, while only resting mast cells infiltrated significantly in the low-expression group of SLPI. CRISP3 was established as a critical gene in BMPCa linked to DES expression. Targeting CRISP3, d-glucopyranose may impact tumour prognosis. During the mechanistic experiments, it was established that CRISP3 can advance the proliferation and metastatic potential of PCa by advancing epithelial-to-mesenchymal transition (EMT).

Conclusion: By modulating lipid metabolism and maintaining immunological and microenvironmental balance, DES, HBB, and SLPI suppress prostate cancer cell growth. The presence of DES-associated CRISP3 is a harbinger of unfavorable outcomes in prostate cancer and may escalate tumor proliferation and metastatic capabilities by inducing epithelial-mesenchymal transition.

Keywords: Bone metastasis (BM); Epithelial to mesenchymal transition (EMT); Immuno-microenvironmental homeostasis; Prostate cancer (PCa); scRNA-seq.

Figure 1. Identification of key genes via TCGA and GSE32269 Datasets.

(A) Intersecting the differential genes of the TCGA and GSE32269 datasets, yielding 199 common genes. (B) Bar graphs depicting the Gene Ontology enrichment analysis of the differential genes. (C) Bubble plots portraying the Kyoto Encyclopedia of Genes and Genomes enrichment analysis of the differential genes. (D) Fold plots displaying error rates and the premier genes.

Figure 2. TCGA data-based cox regression analysis.

(A) Identification of 27 genes relevant to prognosis of bone metastatic prostate cancer (BMPCa); (B) selection of Genes with AOC greater than 0.6; (C) Kaplan–Meier Survival Curves for HBB, (D) SLPI, and (E) DES.

Figure 3. An assessment of DES, HBB, and SLPI using TCGA Data.

(A) Comparison of High- and Low-DES gene expressions through heat map analysis; (B) analysis of pathway variations between high- and low-DES expression groups through GSVA; (C) comparison of high- and low-HBB gene expressions through heat map analysis; (D) analysis of pathway enrichments between high- and low-HBB Expression groups using GSVA; (E) comparison of high- and low-SLPI gene expressions through heat map analysis; (F) analysis of pathways linked to high- and low-SLPI expressions.

Figure 4. Analysis of immune infiltration by CIBERSORT.

(A) assessment of the presence of 22 distinct immune cell types in groups with low or high expression of DES, HBB, and SLPI, respectively; (B) correlation between the expression levels of DES, HBB, and SLPI, and the 22 immune cell types; (C) linear correlation between DES and M1 macrophages; (D) linear correlation between DES and naive B cells; (E) linear correlation between HBB and neutrophils; (F) linear correlation between HBB and M1 macrophages; (G) linear correlation between SLPI and T-regulatory cells; and (H) linear correlation between SLPI and M2 macrophages.

Figure 5. Based on single-cell RNA-seq data, four clusters of prostate cancer (PCa) cells were identified with different annotations.

(A) After quality control of 3,000 cells from the tumor cores of three human LNCaP and VCaP samples, 2,857 cells were included in the analysis; (B) the number of genes detected was significantly correlated with the depth of sequencing (Pearson correlation coefficient = 0.93); (C) analysis of variance (ANOVA) plot showing 19,752 corresponding genes in all glioblastoma multiforme cells (GBMs); (D) principle component analysis (PCA) showing clear separation of cells; (E) PCA identified 20 cases of principle component, with estimated P-value of <0.05; (F) the 20 pc were downscaled using the t-distributed random neighborhood embedding (tSNE) algorithm, followed by cell type annotation and classification of four cell clusters; and (G) differential analysis identified 8,025 marker genes, and the top 20 marker genes for each cell cluster are shown in the heat map.

Figure 6. An examination of pivotal genes for bone metastatic prostate cancer (BMPCa).

(A) Intersection analysis of the marker genes from single-cell GSE168733 with differential genes in BMPCa; (B) intersection analysis of 60 commonly differential genes with genes linked to expression of DES, HBB, and SLPI; and (C–F) Survival analysis of CRISP3 (C), SLC4A4 (D), SMS (E), and BANK1 (F) using the TCGA database.

Figure 7. The proliferation and migration of bladder cancer cells elicited by CRISP3 is correlated with epithelial-to-mesenchymal transition (EMT).

(A) qRT-PCR analysis of CRISP3 expression in prostate cancer cell lines; (B) results from the CCK8 assay suggest that the expression of CRISP3 significantly impacted cellular viability; (C) CRISP3 cells underwent transfection with negative control or empty vector or were overexpressed with CRISP3 after being treated with an invasion inhibitor; (D–E) qRT-PCR (D) and Western blot (E) were employed to assess the expression of EMT-related mRNA and proteins.