Myalgic Encephalomyelitis/Chronic Fatigue Syndrome (ME/CFS) gene expression signatures for the identification of compounds targeting metabolism

Technical Summary: ME/CFS Gene Expression Signatures for Identifying Metabolic Compounds

Problem Statement
Pathological fatigue is a debilitating symptom prevalent in chronic illnesses such as Myalgic Encephalomyelitis/Chronic Fatigue Syndrome (ME/CFS), Multiple Sclerosis, and Long-COVID-19. Unlike physiological fatigue, pathological fatigue persists despite rest and is associated with exercise intolerance and reduced quality of life. Conventional drug discovery pipelines have struggled to identify effective treatments for this symptom. Clinical trials for ME/CFS have largely failed to produce sustained improvements, with only modest effects observed for specific immunomodulatory or psychostimulant drugs in subsets of patients. Furthermore, the molecular basis of ME/CFS remains unclear, lacking validated therapeutic targets, reproducible biomarkers, and preclinical models, creating a "knowledge vacuum" that hinders rational drug development.

Methodology
The authors employed a data-driven, reciprocal disease-drug signature matching approach to identify potential therapeutic compounds.

1. Gene Expression Analysis: The study analyzed two independent, publicly available ME/CFS gene expression datasets (GSE14577 and GSE227375) derived from peripheral blood mononuclear cells (PBMCs). Differentially expressed genes (DEGs) were identified using limma and DESeq2 packages with a False Discovery Rate (FDR) ≤ 0.05 and fold change ≥ 1.25.
2. Signature Identification: The researchers identified a common 46-gene signature (21 up-regulated and 25 down-regulated genes) shared across both datasets.
3. Connectivity Mapping (CMap L1000): This 46-gene signature was queried against the LINCS CMap L1000 database, which contains over one million gene expression profiles induced by genetic and small-molecule perturbagens. The goal was to find compounds with negative connectivity scores (≤ -0.64), indicating an inverse gene expression profile capable of potentially reversing the ME/CFS pathological phenotype.
4. Pathway and Network Analysis: Functional networks were analyzed using STRING, KEGG, and Reactome to understand the biological context of the identified genes and the mechanisms of the candidate compounds.
5. In Vitro Validation: Four candidate compounds targeting metabolism—pioglitazone and rosiglitazone (PPAR agonists), tideglusib (GSK3 inhibitor), and MHY1485 (mTORC1 agonist)—were tested on mouse C2C12 myoblasts and differentiated myotubes. Assays included mitochondrial membrane potential (TMRE), mitochondrial mass (MitoTracker Green), ATP production, Reactive Oxygen Species (ROS) production (MitoSOX), and cell growth (MTT assay).

Key Contributions and Results

• ME/CFS Gene Signature: The study established a robust 46-gene signature common to two independent cohorts. This signature implicated dysregulation in immune function, oxidative stress, apoptosis, and cellular metabolism, including the TCA cycle, oxidative phosphorylation, and autophagy.
• Network Identification: CMap L1000 analysis of the 46-gene signature linked the disease state to 83 genes involved in cytokine signaling, infection, immunity, metabolism, lipid metabolism, and redox regulation.
• Compound Identification: The analysis identified 14 compounds with strong negative connectivity scores. While half were anti-inflammatory/analgesic, the study focused on metabolic regulators. Four compounds were selected for functional testing: pioglitazone, rosiglitazone, tideglusib, and MHY1485.
• Mitochondrial Effects:
  • Pioglitazone: Induced a modest but significant increase in mitochondrial membrane potential and ATP production in differentiated C2C12 myotubes at 48 hours. This occurred without an increase in mitochondrial mass or ROS production, suggesting enhanced metabolic efficiency rather than biogenesis.
  • Tideglusib and MHY1485: Both induced a transient increase in mitochondrial membrane potential and ATP production (at 24 hours) in myotubes, which was lost by 48 hours. Like pioglitazone, these effects were independent of mitochondrial mass and ROS.
  • Rosiglitazone: Did not significantly affect mitochondrial membrane potential or ATP production in the tested conditions.
• Cell Growth: Pioglitazone and rosiglitazone increased the growth of C2C12 myoblasts, whereas tideglusib and MHY1485 showed no significant effect on myoblast growth.

Significance and Claims
The paper claims that its findings provide a framework for exploring therapeutic potential in conditions involving pathological fatigue, such as ME/CFS, by utilizing reciprocal disease-drug signature matching. The authors posit that their approach offers a rational, data-driven alternative to conventional drug discovery, which is often hindered by a lack of mechanistic understanding in complex chronic diseases.

Specifically, the study suggests that:

1. Metabolic Targeting: Impaired mitochondrial metabolism is a plausible contributor to ME/CFS pathology, and compounds targeting metabolic pathways (PPAR, GSK3, mTORC1) may offer therapeutic avenues.
2. Pioglitazone Potential: The ability of pioglitazone to enhance mitochondrial membrane potential and ATP production in muscle cells without increasing ROS suggests it could restore metabolic function in ME/CFS. The authors note that while pioglitazone is approved for Type 2 Diabetes, its potential to improve pathological fatigue in ME/CFS has not been reported, warranting further investigation.
3. Repurposing Strategy: The study demonstrates that drug repurposing based on unbiased gene signature matching can identify novel candidates (like tideglusib and MHY1485) for conditions where traditional target identification is difficult.

The authors maintain a modest tone regarding clinical translation, acknowledging limitations such as the use of a single cell line (C2C12) rather than primary cells from ME/CFS patients, and the lack of data on whether these compounds revert the specific disease signature in patient-derived cells. They conclude that while these findings are promising, further preclinical investigations and clinical trials are necessary to validate the therapeutic efficacy of these compounds for ME/CFS.