A full-length single nuclei transcriptomic atlas of human skeletal muscle insulin resistance

By integrating full-length single-nucleus transcriptomics with gold-standard metabolic clamps, this study redefines skeletal muscle insulin resistance as a nucleus-resolved multicellular process, identifying specific myonuclear populations and molecular regulators like ZIP14 and EGF signaling that drive metabolic dysfunction.

The Gist

Imagine your body's skeletal muscles (the ones that let you walk, run, and lift things) as a massive, bustling city. For a long time, scientists thought this city was mostly made up of one type of building: the "muscle fiber factories" that burn sugar for energy. They believed that if the city got sick (specifically, if it stopped listening to the "sugar manager" called insulin), it was because the factories themselves were broken.

But this new study is like sending a high-tech drone fleet with super-cameras into that city. Instead of just looking at the whole neighborhood from a distance (which is what older studies did), they zoomed in on every single nucleus (the "control room" inside every cell) to see exactly what was going on inside.

Here is the story of what they found, explained simply:

1. The City Has Hidden Neighborhoods

Scientists used to think all muscle cells were just "Type 1" (slow, endurance) or "Type 2" (fast, power). But this study discovered that the "control rooms" (nuclei) inside the muscle fibers are actually very diverse. They found three distinct types of neighborhoods that we didn't know about before:

• The "EGF" Neighborhood: These control rooms are like a chaotic construction site. They are full of signals that cause inflammation and confusion.
• The "MYH7B" Neighborhood: These are the "Zen Masters." They are calm, efficient, and great at managing energy.
• The "CKM" Neighborhood: These are the hardworking energy plants.

The Big Discovery: In healthy people, the city has a good balance of Zen Masters and Energy Plants. But in people with Type 2 Diabetes (where the body can't manage sugar well), the city is overrun by the chaotic EGF neighborhoods, and the Zen Masters are scarce. The more chaotic neighborhoods you have, the worse your sugar control becomes.

2. The "Gold Standard" vs. The "Guessing Game"

To figure out who was healthy and who was sick, the researchers used two different tools:

• The Guessing Game (HOMA2-IR): This is like looking at a car's dashboard while it's parked. It uses fasting blood sugar and insulin to guess how the engine runs. It's easy, but it's often wrong because it doesn't tell you what's happening while the car is driving.
• The Gold Standard (The Clamp): This is like putting the car on a dynamometer (a test track) and driving it at full speed while measuring exactly how much fuel it burns. This study used this "Gold Standard" method.

Why it matters: When they used the "Guessing Game," the results were messy and confusing. But when they used the "Gold Standard," they saw a crystal-clear picture: specific control rooms (nuclei) were directly linked to how well the muscle burned sugar. It turns out, you can't understand the city's traffic just by looking at the parked cars; you have to watch the traffic while it's moving.

3. The Zinc Hero (ZIP14)

The study found a specific "hero" molecule called ZIP14 (a zinc transporter).

• In the Liver: ZIP14 is usually a villain; it helps the liver ignore insulin.
• In the Muscle: ZIP14 is a hero. It acts like a traffic cop, opening the gates to let sugar into the muscle cells.

The researchers tested this by removing ZIP14 from mice. Without this "traffic cop," the muscle cells couldn't take in sugar, even when insulin told them to. This suggests that boosting ZIP14 could be a new way to help people with diabetes manage their blood sugar.

4. The "Bad Neighbor" Effect (EGF Signaling)

The study also found that the chaotic EGF neighborhoods don't just sit there; they talk to the immune cells (the city's police force).

• When the EGF neighborhoods signal the immune cells, it creates a "riot." The immune cells get angry and start a fire (inflammation).
• This fire stops the muscle from breaking down "Branch-Chain Amino Acids" (BCAAs), which are like high-grade fuel. When you can't burn this fuel, it builds up in the blood, which is a classic sign of diabetes.
• Essentially, the EGF neighborhoods are starting a riot that clogs the fuel lines.

The Takeaway

This paper changes the story of Type 2 Diabetes. It's not just that "muscles are broken." It's that the balance of different control rooms inside the muscle has shifted.

• The Problem: Too many chaotic "EGF" neighborhoods and not enough "Zen Master" neighborhoods.
• The Cause: A lack of the "Zinc Hero" (ZIP14) and a riot caused by bad signaling between muscle and immune cells.
• The Hope: By targeting these specific neighborhoods (like boosting ZIP14 or calming the EGF signals), we might be able to fix the city's traffic and help people with diabetes manage their sugar much better.

In short: Muscle health isn't just about the size of the muscle; it's about the specific mix of "control rooms" inside it.

Technical Summary

1. Problem Statement

Skeletal muscle (SkM) insulin resistance is a primary driver of Type 2 Diabetes (T2D). However, the molecular mechanisms underlying this resistance remain incompletely understood due to two major limitations in previous research:

• Bulk Transcriptomics: Prior studies relied on whole-muscle homogenates, which obscure cell-type-specific signals. SkM is a complex tissue containing myofibers (syncytia with multiple nuclei), endothelial cells, immune cells, and progenitor cells, all of which contribute differently to metabolic health.
• Technological Constraints: Existing single-nucleus RNA sequencing (snRNA-seq) studies in SkM have largely used 3' or 5' end-capture technologies, resulting in lower gene coverage and an inability to detect full-length transcripts or alternative splicing events critical for defining cell states.
• Phenotyping Limitations: Many studies rely on surrogate markers (e.g., HOMA-IR) rather than gold-standard measures of insulin sensitivity, potentially missing specific molecular signatures related to actual glucose disposal.

2. Methodology

The authors integrated high-resolution full-length snRNA-seq with gold-standard metabolic phenotyping to create a nucleus-resolved atlas of human SkM insulin resistance.

• Cohort & Phenotyping:

  • Subjects: 18 older adults (>60 years) with obesity, split into T2D (n=9) and Control (n=9) groups. Groups were matched for age, BMI, and aerobic capacity.
  • Gold-Standard Metabolic Assessment: Participants underwent Hyperinsulinemic-Euglycemic (HE) clamps with stable isotope glucose tracers ([6,6-2H2] glucose). This allowed for the direct quantification of insulin-stimulated glucose disposal rates ($R_d$) and the calculation of insulin sensitivity ($R_d/I$). Non-oxidative glucose disposal (NOGD) and carbohydrate oxidation were also measured via indirect calorimetry.
  • Sample Collection: Vastus lateralis muscle biopsies were taken after an overnight fast.

• Single-Nucleus Transcriptomics:

  • Technology: The study utilized the ICELL8 system with full-length snRNA-seq (SMART-Seq Pro). Unlike 3' biased methods, this approach amplifies both 3' and 5' ends, providing full-length transcript coverage.
  • Scale: 25,303 nuclei were profiled, detecting an average of 4,962 genes per nucleus.
  • Clustering: Nuclei were clustered into 8 distinct populations: Myonuclei (84% of total), Muscle Satellite Cells (MuSC), Fibro-adipogenic Progenitors (FAP), Immune cells, Endothelial Cells (EC), and Smooth Muscle Cells/Pericytes (SMC).

• Validation & Mechanistic Studies:

  • Spatial Validation: RNAscope (fluorescent in situ hybridization) was used to validate novel myonuclear populations in tissue sections.
  • Deconvolution: Published bulk RNA-seq datasets (CALERIE, Trim et al., Mahmassani et al., GTEx) were deconvoluted to validate findings across independent cohorts regarding age, sex, and BMI.
  • Functional Assays:
    • In vitro: Primary human myotubes were treated with fatty acids to induce insulin resistance; gene expression was correlated with glucose oxidation/glycogen synthesis.
    • In vivo: Glucose uptake was measured in differentiated myotubes from Wild-Type (WT) vs. ZIP14 Knockout (KO) mice.

3. Key Contributions

• First Full-Length snRNA-seq Atlas of Human SkM: This is the first study to apply full-length snRNA-seq to human skeletal muscle, revealing transcriptional heterogeneity previously undetectable with 3' biased technologies.
• Discovery of Novel Myonuclear Populations: The study identified distinct myonuclear subtypes beyond traditional myosin heavy chain (MyHC) fiber typing, specifically EGF+ and MYH7B+ myonuclei.
• Nucleus-Specific Molecular Signatures: The authors demonstrated that molecular drivers of insulin resistance are highly specific to nucleus types and are often obscured when analyzing bulk tissue or using surrogate metabolic indices (HOMA-IR).
• Mechanistic Insights: Identification of ZIP14 as a positive regulator of insulin-stimulated glucose uptake in muscle (contrasting its role in the liver) and the link between EGF signaling and impaired branched-chain amino acid (BCAA) catabolism.

4. Key Results

A. Novel Myonuclear Populations and Metabolic Health

• Identification: Myonuclei clustered not just by MyHC (Type I, IIa, IIx) but by CKM+, EGF+, and MYH7B+ signatures.
  • CKM+: Associated with oxidative phosphorylation.
  • EGF+: Associated with action potential and contraction genes.
  • MYH7B+: Associated with cation transport. Note: MYH7B protein production is prevented by alternative splicing in muscle; the signal is driven by non-coding transcripts.
• Metabolic Association:
  • The ratio of MYH7B+/EGF+ myonuclei positively correlated with insulin sensitivity ($R_d/I$).
  • Higher proportions of EGF+ myonuclei correlated with lower NOGD (non-oxidative glucose disposal) and worse glycemic control.
  • This ratio was validated in independent datasets: it is higher in males, increases with healthy aging (up to ~60-69 years), and decreases in states of induced insulin resistance (e.g., bed rest).

B. Nucleus-Specific Transcriptional Signatures

• Gold-Standard vs. Surrogate Modeling:
  • Modeling gene expression against $R_d/I$ (gold standard) revealed highly specific signatures. Only 780 of 4,867 genes correlated with $R_d/I$ overlapped across cell types.
  • Modeling against HOMA2-IR (surrogate) showed poor nucleus specificity and failed to capture the distinct pathways linked to actual glucose disposal.
• Specific Pathways:
  • CKM+ Myonuclei: Carbohydrate metabolism genes positively correlated with $R_d/I$.
  • EGF+ Myonuclei: Protein localization to the plasma membrane genes negatively correlated with $R_d/I$.
  • Immune Cells: Myeloid cells showed negative correlations between EGF/EGFR signaling and $R_d/I$.

C. Mechanistic Findings

• ZIP14 (SLC39A14):
  • Expression of the zinc transporter ZIP14 positively correlated with $R_d/I$ in both CKM+ and EGF+ myonuclei.
  • Functional Validation: In ZIP14 KO mice, insulin-stimulated glucose uptake was significantly blunted compared to WT. In human myotubes, ZIP14 expression correlated with glucose oxidation and glycogen synthesis, particularly under insulin-resistant (lipid-induced) conditions.
  • Mechanism: ZIP14 likely suppresses NF-κB inflammatory signaling, thereby preserving IRS-1 tyrosine phosphorylation and insulin sensitivity.
• EGF Signaling & BCAA Catabolism:
  • EGF+ myonuclei proportions and EGF signaling scores negatively correlated with insulin sensitivity.
  • BCAA Link: Higher EGF scores were associated with the downregulation of BCAA catabolism genes (e.g., BCAT2, BCKDHB, PPM1K) specifically in MYH7B+ myonuclei. This suggests EGF signaling contributes to the elevated plasma BCAA levels seen in T2D by impairing muscle BCAA breakdown.
  • Cell-Cell Communication: EGF+ myonuclei communicate with Myeloid immune cells via EGF/EGFR, driving a pro-inflammatory state that correlates with reduced insulin sensitivity.

5. Significance

• Redefining Insulin Resistance: The study shifts the paradigm from viewing SkM insulin resistance as a uniform tissue defect to a multicellular, nucleus-resolved process. Different nuclei within the same fiber contribute distinctively to metabolic dysfunction.
• Therapeutic Targets:
  • ZIP14: Identified as a novel, positive regulator of muscle glucose uptake, suggesting zinc modulation or ZIP14 upregulation could be therapeutic.
  • EGF/EGFR: The link between EGF signaling, inflammation, and impaired BCAA catabolism suggests that EGFR inhibitors (already in clinical use for cancer) could be repurposed to improve metabolic health.
• Methodological Impact: Demonstrates that full-length snRNA-seq combined with gold-standard phenotyping is essential for uncovering molecular mechanisms that are invisible to bulk transcriptomics or 3'-biased single-cell methods.
• Clinical Relevance: Highlights that the MYH7B+/EGF+ ratio could serve as a biomarker for metabolic health and that interventions targeting specific cell-cell communication (e.g., EGF/Myeloid crosstalk) may be more effective than broad anti-inflammatory approaches.