In silico transcriptomic analysis reveals shared molecular signatures and immune-associated pathways between Hashimotos thyroiditis and type 2 diabetes with exploratory drug repurposing

This study utilizes in silico transcriptomic analysis to identify shared immune-associated molecular signatures and key genes between Hashimoto's thyroiditis and type 2 diabetes, subsequently prioritizing three candidate drugs (gliquidone, oleanolic acid, and glipizide) for potential repurposing to address the comorbidity.

The Gist

Imagine your body is a bustling city with two very important neighborhoods: the Thyroid District (which controls your energy and metabolism) and the Sugar Control Zone (which manages your blood sugar).

In this study, researchers noticed that sometimes, when the Thyroid District gets attacked by the city's own security forces (an autoimmune disease called Hashimoto's Thyroiditis), the Sugar Control Zone also starts having trouble (a condition called Type 2 Diabetes). It's like a power outage in one part of the city causing a traffic jam in another. Doctors often have to treat these two problems separately, but the researchers wondered: Are these two neighborhoods actually connected by the same broken pipes?

Here is the story of how they found out, explained simply:

1. The Great Data Detective Work

The researchers didn't go into a lab with test tubes. Instead, they went into the "digital library" of the human body. They grabbed two huge piles of genetic blueprints (transcriptomic data):

• Pile A: From people with Hashimoto's.
• Pile B: From people with Type 2 Diabetes.

They used a computer program to scan these blueprints, looking for genes that were acting up in both piles. Think of it like comparing two lists of "troublemakers" in a school. They found 59 genes that were causing trouble in both the Thyroid District and the Sugar Control Zone.

2. Finding the "Big Bosses" (The Key Genes)

Out of those 59 troublemakers, the researchers used a network map (like a social media graph) to see who was talking to whom. They realized that five specific genes were the "Big Bosses" or the Key Generals leading the chaos in both diseases.

These five generals are named: CDC42, CD74, FOS, RAC2, and YWHAB.

• The Metaphor: Imagine these five genes are the conductors of an orchestra. In a healthy body, they keep the music (immune response and metabolism) playing smoothly. In these sick patients, these conductors are screaming the wrong notes, causing the immune system to attack the thyroid and messing up how the body handles sugar at the same time.

3. The "Why" Behind the Chaos

The researchers asked, "What are these generals actually doing?"

• They found that these genes are heavily involved in immune system activity. It turns out the body's security forces are confused and overactive in both conditions.
• They also found that these genes are controlled by specific "managers" (transcription factors and microRNAs) who are giving the wrong orders.
• The Analogy: It's like a factory where the foreman (the gene) is telling the workers to build too many walls (inflammation) instead of fixing the machines (metabolism).

4. The "Drug Repurposing" Treasure Hunt

This is the most exciting part. The researchers asked: "Do we already have drugs in our medicine cabinet that can fix these specific generals?"

Instead of inventing new drugs from scratch (which takes years and costs billions), they looked for existing drugs that might accidentally fit the "locks" of these five genes. They ran a massive computer simulation, like trying thousands of keys in a lock until one clicks.

They found three promising keys:

1. Gliquidone: A drug already used for diabetes.
2. Oleanolic Acid: A natural compound found in olive leaves.
3. Glipizide: Another common diabetes drug.

The Magic: These drugs are already approved for diabetes, but the computer simulation showed they might also calm down the angry immune system in Hashimoto's. It's like finding out that a tool you use to fix a leaky faucet can also tighten a loose screw in your car engine.

5. The "Stress Test" (Simulation)

Before celebrating, they put these three drugs through a rigorous virtual stress test:

• Will the body absorb them? Yes.
• Are they toxic? No, they seem safe.
• Do they stay stuck to the target? They ran a 100-second "movie" (molecular dynamics simulation) of the drug hitting the gene. The drugs held on tight and didn't fall off, proving they are stable candidates.

The Bottom Line

This study is a hypothesis generator. It's not a cure yet, and doctors can't prescribe these drugs for Hashimoto's based on this alone.

Think of it this way: The researchers just handed the medical community a very strong map. They said, "Look, these two diseases share the same broken wiring. Here are five specific wires causing the problem, and here are three tools we already have that might fix them."

Now, real scientists need to go into the lab (the "wet lab") to test if these drugs actually work in living cells and animals. If they do, we might soon see a future where a single treatment plan helps patients manage both their thyroid and their blood sugar, making life much easier for millions of people.

Technical Summary

1. Problem Statement

Hashimoto's Thyroiditis (HT), an autoimmune disorder causing hypothyroidism, frequently co-occurs with Type 2 Diabetes (T2D). Managing patients with both conditions is clinically complex due to overlapping genetic, metabolic, and immune dysregulations, as well as potential therapeutic conflicts. While epidemiological links are established, the molecular mechanisms underlying this comorbidity remain incompletely characterized. There is a lack of systematic identification of shared gene expression signatures and a need for computational strategies to identify candidate drugs that could target these shared pathways for dual management.

2. Methodology

The study employed a comprehensive in silico pipeline utilizing public transcriptomic data and bioinformatics tools:

• Data Acquisition:
  • HT Dataset: GSE138198 (Discovery) and GSE29315 (Validation).
  • T2D Dataset: GSE29231 (Discovery) and GSE25724 (Validation).
  • All datasets were normalized using the Robust Multi-array Average (RMA) method.
• Differential Expression Analysis (DEG):
  • Used the LIMMA R package.
  • Thresholds: Adjusted p-value < 0.05 and |Log2FC| > 1.
  • Identified 59 Shared Differentially Expressed Genes (sDEGs) (50 upregulated, 9 downregulated) common to both diseases.
• Network Analysis & Key Gene Identification:
  • Constructed a Protein-Protein Interaction (PPI) network using STRING.
  • Prioritized Shared Key Genes (sKGs) using five topological algorithms in Cytoscape (CytoHubba plugin): Betweenness, Closeness, Degree, MCC, and EPC.
• Functional & Regulatory Analysis:
  • Enrichment: Gene Ontology (GO) and KEGG pathway analysis via Enrichr (FDR corrected).
  • Regulatory Networks: Identified Transcription Factors (TFs) using JASPAR and microRNAs (miRNAs) using TarBase.
  • Immune Infiltration: Estimated immune cell proportions using CIBERSORT and correlated them with sKG expression.
• Drug Repurposing Pipeline:
  • Target Selection: sKGs and associated TFs.
  • Docking: Molecular docking of 247 candidate compounds against target proteins using AutoDock Vina.
  • ADMET Profiling: Assessed Absorption, Distribution, Metabolism, Excretion, and Toxicity using ADMETlab 2.0, ProTox, and pkCSM.
  • Stability Validation: 100 ns Molecular Dynamics (MD) simulations using YASARA (AMBER14 force field) to evaluate complex stability (RMSD, RMSF, MM-PBSA).

3. Key Contributions & Results

A. Identification of Shared Molecular Signatures

• sDEGs: Identified 59 genes concordantly dysregulated in both HT and T2D.
• sKGs: Prioritized five core genes: CDC42, CD74, FOS, RAC2, and YWHAB.
• Validation: Independent validation datasets (GSE29315 for HT, GSE25724 for T2D) confirmed that all five sKGs were significantly upregulated in both disease states compared to controls.

B. Functional & Regulatory Insights

• Pathways: Enrichment analysis revealed shared involvement in:
  • Biological Processes: Regulation of actin cytoskeleton, neutrophil chemotaxis, IL-6 production, and oxidative stress response.
  • KEGG Pathways: MAPK signaling, Antigen processing/presentation, IL-17 signaling, and Hippo signaling.
• Regulators:
  • Transcription Factors: FOXC1 and HNF4A were identified as top regulators of the sKGs.
  • MicroRNAs: hsa-miR-221-3p and hsa-miR-29a-3p were predicted as key post-transcriptional regulators.
• Immune Infiltration: Both diseases exhibited similar immune dysregulation patterns, specifically:
  • Increased: Memory B cells, activated memory CD4+ T cells, follicular helper T cells, and M0 macrophages.
  • Decreased: Resting memory CD4+ T cells and M2 macrophages.
  • Strong correlations were found between sKG expression levels and these specific immune cell populations.

C. Druggability & Repurposing Candidates

• Binding Pockets: All sKGs and TFs possessed druggable binding pockets with favorable scores (e.g., CDC42 and CD74 showed high druggability scores >0.8).
• Top Candidates: From 247 screened compounds, three emerged as the most promising based on binding affinity (< -7 kcal/mol), favorable ADMET profiles, and drug-likeness (Lipinski's Rule of 5):
  1. Gliquidone (Sulfonylurea antidiabetic)
  2. Oleanolic acid (Natural triterpenoid)
  3. Glipizide (Sulfonylurea antidiabetic)
• Stability: MD simulations confirmed that the complexes formed by these drugs with their respective targets (e.g., Gliquidone-CDC42) were structurally stable over 100 ns, with low RMSD fluctuations.

4. Significance and Implications

• Mechanistic Insight: The study provides a hypothesis-generating framework linking HT and T2D through a shared axis of immune-inflammatory dysregulation (specifically involving actin cytoskeleton organization, antigen presentation, and cytokine signaling) and metabolic dysfunction.
• Therapeutic Potential: It identifies Gliquidone, Oleanolic acid, and Glipizide as potential repurposing candidates. These drugs, primarily known for T2D management, possess predicted anti-inflammatory properties that could address the autoimmune component of HT in comorbid patients.
  • Gliquidone and Glipizide are noted for potential immunomodulatory effects beyond glucose control.
  • Oleanolic acid is highlighted for its cytoprotective and anti-inflammatory capabilities.
• Clinical Relevance: The findings suggest that targeting shared molecular nodes (sKGs) could offer a more efficient therapeutic strategy for patients with this specific comorbidity, potentially reducing polypharmacy and improving outcomes.

5. Limitations

• Data Source: The analysis relied on independent datasets; no transcriptomic data from patients with both conditions simultaneously was available.
• In Silico Nature: The findings are computational predictions. The identification of sKGs based solely on topological metrics in PPI networks carries a risk of false positives/negatives.
• Validation Needed: The study explicitly states that experimental validation (in vitro cytotoxicity, in vivo animal models, and clinical trials) is required to confirm the therapeutic efficacy and safety of the repurposed drugs.

Conclusion

This research successfully bridges the gap between two distinct endocrine disorders by identifying a shared molecular landscape driven by immune dysregulation. By leveraging a multi-step bioinformatics approach, the authors have prioritized specific genes and repurposed drugs that warrant further experimental investigation to improve the management of the HT-T2D comorbidity.