Comprehensive multi-omics approach reveals detailed insights into diabetes mechanisms

By Dr. Chinta SidharthanReviewed by Lexie CornerSep 20 2024

Advances in high-throughput phenotyping and multi-omics technologies have enabled the measurement of thousands of genomic, transcriptomic, proteomic, and epigenomic readouts. These methods deepen our understanding of cellular molecular interactions and the pathophysiological processes underlying various diseases.

Image Credit: isak55/Shutterstock.com

A recent study published in Nature Communications employed 18 different omics-based technologies to analyze blood, urine, and saliva samples from a multi-ethnic cohort of diabetes patients. This comprehensive approach aimed to construct a molecular roadmap to understand population traits and subtypes of diabetes.

Background

Recent advancements in high-throughput omics and molecular phenotyping technologies have enabled the detailed measurement and analysis of thousands of proteomic, genomic, and transcriptomic components within cells.

While studies with smaller cohorts have focused on dynamic molecular changes in response to factors like viral infections, exercise, or spaceflight, larger cohort studies, such as those conducted by the United Kingdom Biobank, have investigated the molecular mechanisms underlying diseases like diabetes, obesity, schizophrenia, and Alzheimer’s. These studies employ genomics, proteomics, and metabolomics to examine the complex molecular networks associated with these conditions.

Despite these advancements, selecting appropriate platforms and integrating data from multiple omics layers remain significant challenges. Successfully combining diverse molecular phenotyping technologies in a single study could provide comprehensive insights into the molecular interactions that govern both healthy and diseased states.

About the study

In this study, researchers collected saliva, blood, and urine samples from 391 participants of diverse ethnic backgrounds enrolled in the Qatar Metabolomics Study of Diabetes. These samples were analyzed using 18 different molecular phenotyping methods to measure a range of molecular traits, including metabolites, proteins, lipids, and glycosylation patterns.

The genotyping covered over 1.2 million genetic variants, and the transcriptomes of white blood cells from the blood samples were sequenced. Additionally, DNA methylation was analyzed at 450,000 sites. Various advanced techniques were employed to measure these traits, such as nuclear magnetic resonance (NMR), mass spectrometry, and aptamer-based assays. Technologies like N-glycosylation, SOMAscan, and OLINK were used to measure plasma proteins,

For metabolite analysis, both untargeted and targeted methods were utilized. Untargeted methods included ultra-performance liquid chromatography-mass spectrometry and gas chromatography-mass spectrometry, while targeted approaches such as phenylisothiocyanate derivatization and multiple reaction monitoring were applied to analyze specific metabolites in the samples.

A combination of butanol/methanol extraction and quadrupole ion trap mass spectrometry was used to study lipid molecules and lipoprotein subclasses. Transcriptomic profiles were obtained through RNA sequencing and micro-RNA quantification.

To explore the relationships between molecular traits and their association with type 2 diabetes, the researchers used various statistical tools to analyze cross-platform and within-trait partial correlations. Significant single nucleotide polymorphisms linked to the omics data were identified, and network models were employed to examine interactions between molecular traits. Genome-wide and epigenome-wide association studies were also conducted.

Furthermore, the researchers developed an interactive tool called Comics, enabling visualization of the relationships between various molecular traits and exploration of associations with factors such as age, body mass index, and diabetes.

Major findings

The study demonstrated that integrating data from various omics platforms—genomic, proteomic, and metabolomic—can comprehensively map molecular interactions across different biological systems. This holistic approach revealed significant molecular trait interactions and allowed for a more detailed examination of the complex relationships within the body.

The researchers identified five distinct subgroups of diabetes patients, categorized based on disease progression and complications, providing a framework for developing personalized treatment strategies. The study also uncovered previously unreported associations, such as the link between leptin and CXC motif chemokine ligand 5 (CXCL5) and their role in metabolic disorders.

These findings not only confirmed existing knowledge of metabolic disease mechanisms but also opened new avenues for therapeutic research, particularly in understanding the potential roles of CXCL5 and leptin in white adipose tissue remodeling.

Additionally, the team developed an interactive web-based tool called Comics, which offers researchers access to large-scale multi-omics data for testing molecular interactions, promoting the exploration of molecular networks.

Conclusions

This comprehensive multi-omics study established a foundational framework for future large-scale research, leveraging multiple omics platforms to provide a valuable roadmap for exploring molecular interactions, biological processes, and disease mechanisms.

Journal reference

Halama, A., et al. (2024). A roadmap to the molecular human linking multiomics with population traits and diabetes subtypes. Nature Communications. DOI:10.1038/s4146702451134x, https://www.nature.com/articles/s41467-024-51134-x