Multiomics Analysis-Based Biomarkers in Diagnosis of Polycystic Ovary Syndrome

Knowledge graph construction based on granulosa cells transcriptome from polycystic ovary syndrome with normoandrogen and hyperandrogen

PCOS is a widespread disease that primarily caused in-pregnancy in pregnant-age women. Normoandrogen (NA) and Hyperandrogen (HA) PCOS are distinct subtypes of PCOS, while bio-markers and expression patterns for NA PCOS and HA PCOS have not been disclosed. We performed microarray analysis on granusola cells from NA PCOS, HA PCOS and normal tissue from 12 individuals. Afterwards, microarray data were processed and specific genes for NA PCOS and HA PCOS were identified. Further functional analysis selected IL6R and CD274 as new NA PCOS functional markers, and meanwhile selected CASR as new HA PCOS functional marker. IL6R, CD274 and CASR were afterwards experimentally validated on mRNA and protein level. Subsequent causal relationship analysis based on Apriori Rules Algorithm and co-occurrence methods identified classification markers for NA PCOS and HA PCOS. According to classification markers, downloaded transcriptome datasets were merged with our microarray data. Based on merged data, causal knowledge graph was constructed for NA PCOS or HA PCOS and female infertility on NA PCOS and HA PCOS. Gene-drug interaction analysis was then performed and drugs for HA PCOS and NA PCOS were predicted. Our work was among the first to indicate the NA PCOS and HA PCOS functional and classification markers and using markers to construct knowledge graphs and afterwards predict drugs for NA PCOS and HA PCOS based on transcriptome data. Thus, our study possessed biological and clinical value on further understanding the inner mechanism on the difference between NA PCOS and HA PCOS.

Significant alteration of protein profiles in a mouse model of polycystic ovary syndrome

Polycystic ovary syndrome (PCOS) is an endocrine disorder, affecting women of child-bearing age, and the incidence rate is growing and assuming epidemic proportions. The etiology of PCOS remains unknown and there is no cure. Some animal models for PCOS have been established which have enhanced our understanding of the underlying mechanisms, but omics data for revealing PCOS pathogenesis and for drug discovery are still lacking. In the present study, proteomics analysis was used to construct a protein profile of the ovaries in a PCOS mouse model. The result showed a clear difference in protein profile between the PCOS and control group, with 495 upregulated proteins and 404 downregulated proteins in the PCOS group. The GO term and KEGG pathway analyses of differentially expressed proteins mainly showed involvement in lipid metabolism, oxidative stress, and immune response, which are consistent with pathological characteristics of PCOS in terms of abnormal metabolism, endocrine disorders, chronic inflammation and imbalance between oxidant and antioxidant levels. Also, we found that inflammatory responses were activated in the PCOS ovarium, while lipid biosynthetic process peroxisome, and bile secretion were inhibited. In addition, we found some alteration in unexpected pathways, such as glyoxylate and dicarboxylate metabolism, which should be investigated. The present study makes an important contribution to the current lack of PCOS ovarian proteomic data and provides an important reference for research and development of effective drugs and treatments for PCOS.

A novel model based on necroptosis to assess progression for polycystic ovary syndrome and identification of potential therapeutic drugs

Background

Polycystic ovary syndrome (PCOS), a common endocrine and reproductive disorder, lacks precise diagnostic strategies. Necroptosis was found to be crucial in reproductive and endocrine disorders, but its function in PCOS remains unclear. We aimed to identify differentially diagnostic genes for necroptosis (NDDGs), construct a diagnostic model to assess the progression of PCOS and explore the potential therapeutic drugs.

Methods

Gene expression datasets were combined with weighted gene co-expression network analysis (WGCNA) and necroptosis gene sets to screen the differentially expressed genes for PCOS. Least absolute shrinkage and selection operator (LASSO) regression analysis was used to construct a necroptosis-related gene signatures. Independent risk analyses were performed using nomograms. Pathway enrichment of NDDGs was conducted with the GeneMANIA database and gene set enrichment analysis (GSEA). Immune microenvironment analysis was estimated based on ssGSEA algorithm analysis. The Comparative Toxicogenomics Database (CTD) was used to explore potential therapeutic drugs for NDDGs. The expression of NDDGs was validated in GSE84958, mouse model and clinical samples.

Results

Four necroptosis-related signature genes, IL33, TNFSF10, BCL2 and PYGM, were identified to define necroptosis for PCOS. The areas under curve (AUC) of receiver operating characteristic curve (ROC) for training set and validation in diagnostic risk model were 0.940 and 0.788, respectively. Enrichment analysis showed that NDDGs were enriched in immune-related signaling pathways such as B cells, T cells, and natural killer cells. Immune microenvironment analysis revealed that NDDGs were significantly correlated with 13 markedly different immune cells. A nomogram was constructed based on features that would benefit patients clinically. Several compounds, such as resveratrol, tretinoin, quercetin, curcumin, etc., were mined as therapeutic drugs for PCOS. The expression of the NDDGs in the validated set, animal model and clinical samples was consistent with the results of the training sets.

Conclusion

In this study, 4 NDDGs were identified to be highly effective in assessing the progression and prognosis of PCOS and exploring potential targets for PCOS treatment.

Identification of key biomarkers in ischemic stroke: single-cell sequencing and weighted co-expression network analysis

PURPOSE

At present, there is a lack of accurate early diagnostic markers for ischemic stroke.

METHODS

By using dimensionality reduction cluster analysis, differential expression analysis, weighted co-expression network analysis, protein-protein interaction network analysis, cell heterogeneity and key pathogenic genes were identified in ischemic stroke. Immunomicroenvironment analysis was used to explore the immune landscape and immune associations of key genes in ischemic stroke. The analysis platform we use is R software (version 4.0.5). PCR experiments were used to verify the expression of key genes.

RESULTS

Single cell sequencing data in ischemic stroke can be annotated as fibroblast cells, pre-B cell CD34, neutrophils cells, bone marrow (BM), keratinocytes, macrophage, neurons and mesenchymal stem cells (MSC). By the intersection of differential expression analysis and WGCNA analysis, 385 genes were obtained. Gene ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis showed that these genes were highly correlated with multiple functions and pathways. Protein-protein interaction network analysis revealed that MRPS11 and MRPS12 were key genes, both of which were down-regulated in ischemic stroke. The Pseudo-time series analysis found that the expression of MRPS12 decreased gradually with the differentiation of pre-B cell CD34 cells in ischemic stroke, suggesting that the downregulation of MRPS12 expression may play an important role in ischemic stroke. At last, PCR showed that MRPS11 and MRPS12 were significantly down-regulated in peripheral blood of patients with ischemic stroke.

CONCLUSIONS

Our study provides a reference for the study of pathogenesis and key targets of ischemic stroke.

Multi-omics insights and therapeutic implications in polycystic ovary syndrome: a review

Polycystic ovary syndrome (PCOS) is a common gynecological disease that causes adverse effects in women in their reproductive phase. Nonetheless, the molecular mechanisms remain unclear. Over the last decade, sequencing and omics approaches have advanced at an increased pace. Omics initiatives have come to the forefront of biomedical research by presenting the significance of biological functions and processes. Thus, multi-omics profiling has yielded important insights into understanding the biology of PCOS by identifying potential biomarkers and therapeutic targets. Multi-omics platforms provide high-throughput data to leverage the molecular mechanisms and pathways involving genetic alteration, epigenetic regulation, transcriptional regulation, protein interaction, and metabolic alterations in PCOS. The purpose of this review is to outline the prospects of multi-omics technologies in PCOS research by revealing novel biomarkers and therapeutic targets. Finally, we address the knowledge gaps and emerging treatment strategies for the management of PCOS. Future PCOS research in multi-omics at the single-cell level may enhance diagnostic and treatment options.