TL;DR
This study performed a meta-analysis of gene expression data from lean and obese individuals with Polycystic Ovary Syndrome (PCOS) to uncover distinct molecular pathways and associated comorbidities. The findings suggest that while many genes are downregulated in PCOS, the specific affected pathways and the density of gene-disease networks differ between lean and obese phenotypes, indicating potentially distinct pathophysiological mechanisms and varying risks for comorbidities.
What Problem Does This Paper Address?
Polycystic Ovary Syndrome (PCOS) is a common endocrine disorder affecting women of reproductive age, often leading to various comorbidities like metabolic syndrome, insulin resistance, and an increased risk of certain cancers. While obesity is a known risk factor for PCOS, a significant portion of affected women are lean, and the underlying mechanisms may differ between these phenotypes. Previous research has often focused on obese PCOS or PCOS in general, leaving a gap in our understanding of the specific molecular differences between lean and obese PCOS. This paper aims to address this by systematically comparing gene expression profiles, identifying differentially regulated pathways, and analyzing associated comorbidities in lean versus obese PCOS patients.
Methods
The researchers employed a meta-analysis approach to integrate data from multiple studies. They collected gene expression data from eight relevant datasets in the Gene Expression Omnibus (GEO) database, filtering out irrelevant studies. For each dataset, they performed microarray data pre-processing, including background correction and normalization, followed by differential gene expression analysis using the limma package to identify differentially expressed genes (DEGs). These DEGs were then compared across datasets to find commonly dysregulated genes. Pathway enrichment analysis was conducted using the Gene Set Enrichment Analysis (GSEA) method to identify biological pathways affected in lean and obese PCOS. Finally, they used the KEGG disease database to map DEGs to disease categories and construct gene-disease networks to assess comorbidity risks, particularly comparing lean and obese PCOS.
Datasets
The study utilized publicly available gene expression data from eight human microarray datasets obtained from the Gene Expression Omnibus (GEO) database. These datasets were selected based on specific criteria and included samples from both lean (BMI ≤ 23) and obese (BMI > 23) individuals with PCOS, as well as control groups. The specific GEO accession numbers are: GSE98421, GSE5850, GSE10946, GSE6798, GSE98595, GSE48301, GSE5090, and GSE43264.
Key Insights and Ideas
- Downregulation Dominates: A striking observation is that the majority of differentially expressed genes in PCOS, across both lean and obese phenotypes and various tissue types, are downregulated. This suggests a general suppression of gene activity rather than widespread activation in the disease state.
- Distinct Etiologies: The lack of commonly dysregulated genes between lean and obese PCOS, along with differences in enriched pathways, strongly suggests that the underlying molecular mechanisms and genetic contributors may differ between these two phenotypes. This challenges a one-size-fits-all understanding of PCOS.
- Tissue Specificity and Shared Pathways: While specific genes might be unique to certain tissues (e.g., endometrium and ovary showing the most DEGs), there's evidence of shared dysregulated genes and pathways between ovarian and endometrial tissues, pointing to common pathophysiological processes in these key organs.
- Comorbidity Risk Varies: The gene-disease network for obese PCOS is significantly denser than that for lean PCOS. This indicates a higher potential for multiple comorbidities in obese individuals with PCOS, aligning with clinical observations.
- Cellular Homeostasis Disruption: Key pathways related to cellular homeostasis, such as inflammation, immune response, insulin signaling, and steroidogenesis, are consistently enriched in both lean and obese PCOS, highlighting fundamental cellular processes affected by the syndrome.
Key Results
- Out of 5014 statistically significant DEGs identified across eight GEO datasets, 4891 (unique 4101) were downregulated, and only 123 (unique 96) were upregulated.
- Endometrial tissues showed the highest number of DEGs, followed by ovarian tissues.
- Six genes (GPX7, SERPINI1, TMEM256, SVIP, MAT2A, and SRGN) were commonly dysregulated across ovarian, endometrial, and adipose tissues.
- Pathway analysis revealed 86 common pathway gene sets enriched in both lean and obese PCOS, with 1031 unique to lean PCOS and 159 unique to obese PCOS.
- Biological pathways were clustered into four major themes: Cell-motility and immune response, FAK-related, ERBB1 and PDGFRB signaling, and Mitochondrial gene expression. Notably, the first three were downregulated in lean PCOS and upregulated in obese PCOS, while mitochondrial gene expression showed the opposite trend.
- The gene-disease network for obese PCOS was denser, suggesting a higher risk of comorbidities compared to lean PCOS.
Limitations and Future Work
The authors acknowledge several limitations. Firstly, the study's conclusions are dependent on the availability and quality of existing gene expression data in public repositories, and the consistency of phenotype annotations across these datasets. They note that a larger number of tissue and BMI-matched samples would have strengthened the analysis. Secondly, the comorbidity analysis is limited by the availability of comprehensive and up-to-date gene-disease association databases. They highlight that some documented gene-PCOS associations (e.g., CYP1B1) are not present in the KEGG database. Finally, the study focused solely on coding genes and did not include RNASeq, lncRNAs, or cell line data, which could have provided additional insights.
Future work should focus on validating these findings through experimental studies, further investigating the specific roles of identified DEGs and pathways in both lean and obese PCOS, and developing diagnostic and therapeutic applications based on these discoveries. The distinct molecular profiles suggest the need for phenotype-specific treatment strategies.
Summary updated