Systems analysis reveals neuregulin-1 control of cardiomyocyte size and shape mediated by distinct PI3K and p38 pathways

By integrating high-content morphological profiling, phospho-protein arrays, and systems modeling, this study reveals that neuregulin-1 uniquely drives cardiomyocyte elongation and area expansion through distinct PI3K and p38 signaling pathways, offering new mechanistic insights for selectively targeting maladaptive cardiac remodeling.

The Gist

Imagine your heart is a bustling city made of millions of tiny workers called cardiomyocytes (heart muscle cells). When the city faces stress—like high blood pressure or a heart attack—these workers have to work harder. To cope, they get bigger. This is called hypertrophy.

But here's the twist: getting bigger isn't just about growing a bigger belly. Sometimes, the workers get wider and rounder (like a balloon), and sometimes they get longer and thinner (like a stretched rubber band). The paper you're asking about is a detective story that figures out why some stress makes the cells stretch out while others make them puff up, and how the cell's internal "control panel" decides which shape to take.

Here is the breakdown of their discovery, using some everyday analogies:

1. The Mystery: Different Stresses, Different Shapes

The researchers knew that different stressors (like Angiotensin II, Endothelin-1, and Neuregulin-1) made heart cells grow, but they didn't know exactly how the cells decided their shape.

• The Analogy: Imagine you have three different construction foremen (the stress signals). Foreman A tells the workers to build a wide, squat warehouse. Foreman B tells them to build a long, narrow tunnel. The researchers wanted to know: What are the blueprints inside the workers that tell them which shape to build?

2. The Investigation: Taking a "Snapshot" of the Cell's Brain

To solve this, the team used a high-tech camera to take "snapshots" of the proteins inside the cells (the cell's internal machinery) at different times after the stress started. They used a method called Reverse-Phase Protein Arrays, which is like taking a photo of every single switch and dial on the control panel to see which ones were turned on or off.

They also used a super-smart computer program (called PLSR, or Partial Least Squares Regression) to act as a translator. This program looked at the "switches" that were flipped and tried to guess: "If these switches are on, what shape will the cell become?"

3. The Big Discovery: The "Stretch" vs. The "Puff"

The computer analysis revealed a fascinating split in the cell's control system. They focused heavily on a signal called Neuregulin-1 (Nrg1), which is known to be a "good guy" signal often associated with healthy exercise.

They found that Nrg1 does two things at once:

1. It makes the cell bigger (more area).
2. It makes the cell longer (elongation).

But here is the magic: The cell uses two different internal pathways to do these two things. Think of the cell's control panel as having two distinct teams of workers:

• Team PI3K (The "Stretch & Grow" Crew): This team is responsible for both making the cell bigger AND making it stretch out longer. If you turn off this team, the cell stops growing in both size and length.
• Team p38 (The "Puff" Crew): This team is only responsible for making the cell bigger (wider). It has nothing to do with stretching the cell out. If you turn off this team, the cell stops getting wider, but it can still stretch out if Team PI3K is working.

The Metaphor:
Imagine a piece of dough.

• Team PI3K is the baker who kneads the dough and pulls it into a long loaf.
• Team p38 is the baker who just puffs the dough up like a bun.
• When Neuregulin-1 arrives, it tells both bakers to work. The result? A long, thick loaf of bread.
• If you stop the "Puff" baker (p38), you still get a long loaf, just a bit thinner.
• If you stop the "Stretch" baker (PI3K), you get no loaf at all.

4. The Proof: Testing the Theory

To prove they were right, the researchers played "what-if" games. They used drugs to temporarily "turn off" specific teams (PI3K or p38) while the stress signal was active.

• Result: When they turned off PI3K, the cells stopped getting longer and stopped getting bigger.
• Result: When they turned off p38, the cells stopped getting wider, but they still got longer!

This confirmed that the cell has a "decoupling" mechanism. It can control its length and its width independently using different internal pathways.

5. The Computer Model: The Virtual Heart

Finally, the team built a digital simulation (a logic-based model) of these pathways. They fed the computer the rules they discovered (e.g., "PI3K controls length," "p38 controls width").

• The computer ran the simulation and successfully predicted exactly what the real cells did in the lab.
• This proves that their understanding of the "wiring diagram" is correct. They didn't just guess; they built a working model that explains the biology.

Why Does This Matter?

In heart failure, the heart often gets sick because the cells grow the wrong shape. They might get too wide and stiff, or too long and weak.

• The Takeaway: Because we now know that PI3K and p38 control different parts of the shape, doctors might one day be able to design drugs that target only the bad growth (like the dangerous widening) while leaving the good growth (like the healthy stretching seen in exercise) alone.

In a nutshell: This paper is like finding the specific wiring diagram for a car's engine. They discovered that one button (Neuregulin-1) turns on two different circuits (PI3K and p38). One circuit makes the car go faster (grow longer), and the other makes the car heavier (grow wider). By understanding the wiring, we might be able to fix a broken engine without taking the whole car apart.

Technical Summary

1. Problem Statement

Cardiac hypertrophy is a compensatory response to stress characterized by increased cardiomyocyte size and mass. While the signaling pathways governing cell size (hypertrophy) are relatively well-studied, the mechanisms controlling cell shape (specifically elongation vs. widening) remain poorly understood.

• Pathological Context: Pressure overload (e.g., hypertension) typically causes concentric hypertrophy (increased cross-sectional area), while volume overload (e.g., mitral regurgitation) or endurance exercise causes eccentric hypertrophy (predominant cell elongation).
• Knowledge Gap: It is unclear how interconnected signaling networks differentially regulate cell size versus cell shape, particularly in response to specific hypertrophic ligands like Neuregulin-1 (Nrg1), Angiotensin II (AngII), Endothelin-1 (ET1), and Insulin-like Growth Factor-1 (IGF1).

2. Methodology

The authors employed a multi-modal systems biology approach combining high-throughput experimental profiling with computational modeling:

• Experimental Models:

  • Cell Types: Neonatal Rat Cardiomyocytes (NRCMs) and human induced pluripotent stem cell-derived cardiomyocytes (hiPSC-CMs).
  • Stimuli: Treatment with hypertrophic ligands (Nrg1, AngII, ET1, IGF1) at 1 hour and 48 hours.
  • Perturbations: Pharmacological inhibition of key kinases: MEK (PD0325901), p38 (SB203580), and PI3K (LY294002).

• Data Acquisition:

  • Proteomics: Reverse-Phase Protein Arrays (RPPA) measuring 172 total and phospho-protein probes to capture signaling dynamics.
  • Phenotyping:
    • Live-cell imaging: Tracking GFP-expressing cardiomyocytes to measure area, perimeter, and elongation.
    • Fixed-cell high-content imaging: Immunostaining for $\alpha$-actinin to measure the entire population. A novel metric, Feret Elongation (ratio of Max Feret Diameter to Min Feret Diameter minus 1), was developed to sensitively quantify shape changes independent of scale.
  • Transcriptomics: qPCR analysis of 8–11 hypertrophy-related mRNAs (e.g., BNP, CITED4, CTGF).

• Computational Modeling:

  • Partial Least Squares Regression (PLSR): Used to map the high-dimensional proteomic data (predictor block) to phenotypic outcomes (response block: mRNA and morphology). This identified latent variables linking specific signaling signatures to cell shape/size.
  • Logic-Based Network Modeling: A differential equation model (implemented in Netflux) was constructed to simulate the dynamics of the PI3K/Akt and p38 pathways, incorporating negative feedback via Dual-Specificity Phosphatases (DUSPs).

3. Key Contributions

1. Decoupling Size and Shape: The study provides mechanistic evidence that cardiomyocyte size and shape are regulated by distinct, separable signaling cascades, rather than a single unified hypertrophic program.
2. Novel Metric: Introduction of Feret Elongation as a robust, scale-independent metric for quantifying cardiomyocyte shape in high-throughput imaging.
3. Systems Integration: Successful integration of RPPA proteomics, high-content imaging, and machine learning (PLSR) to predict causal regulators of phenotype, followed by experimental validation.
4. Nrg1 Specificity: Identification of Nrg1 as a potent inducer of both cell area and elongation, mediated by specific pathway dynamics.

4. Key Results

A. Ligand-Specific Proteomic Signatures

• RPPA analysis revealed that while early responses (1 hr) shared some overlap, proteomic signatures became highly ligand-specific by 48 hours.
• Nrg1 uniquely activated AKT/mTOR signaling at 48 hours, whereas ET1 strongly activated MAPK pathways early on.
• PLSR Modeling: The model successfully predicted phenotypic outcomes from proteomic data ($R^2$ = 0.99 for cell area, 0.84 for elongation).
  • Cluster B (Elongation): Strongly associated with PI3K/Akt pathway components (p-PI3K, p-Akt, p-GSK3, p-mTOR) and MAPK family members (p-p38, p-MEK).
  • Cluster C (Area): Associated with different protein abundances (e.g., tuberin, bRaf) and p38 signaling.

B. Validation of Causal Regulators

• PI3K: Inhibition of PI3K (LY294002) abolished both Nrg1-induced cell area expansion and elongation. This confirms PI3K is a master regulator for both dimensions.
• p38: Inhibition of p38 (SB203580) significantly reduced Nrg1-induced cell area but had no significant effect on elongation. In fact, p38 inhibition slightly enhanced elongation in some contexts, suggesting p38 acts as a brake on elongation or a specific driver of area.
• MEK: Inhibition suppressed elongation, consistent with the PLSR prediction that MAPK components are linked to shape.

C. Dynamics and Modeling

• Temporal Dynamics: Nrg1 induces a transient activation of p38 (peaking early, declining by 48h due to DUSP-mediated dephosphorylation) and a sustained activation of PI3K/Akt.
• Network Model: The logic-based model, incorporating DUSP-mediated negative feedback on p38, successfully recapitulated the experimental data:
  • Sustained PI3K drives both area and elongation.
  • Transient p38 drives area expansion but is insufficient for elongation.
  • The model predicted that blocking p38 would reduce area without affecting elongation, which was experimentally confirmed.

D. Cross-Species Validation

• The findings were consistent across neonatal rat cardiomyocytes and two lines of human iPSC-derived cardiomyocytes, confirming the robustness of the Nrg1-induced elongation phenotype and its dependence on PI3K.

5. Significance and Implications

• Mechanistic Insight: The study resolves the "size vs. shape" paradox in cardiac hypertrophy. It demonstrates that PI3K is the primary driver of overall growth (both size and shape), while p38 selectively modulates cell area.
• Physiological vs. Pathological: Nrg1-induced elongation is associated with physiological hypertrophy (similar to endurance exercise) and involves beneficial markers (e.g., CITED4 upregulation, CTGF downregulation). In contrast, other ligands like AngII/ET1 drive pathological concentric hypertrophy.
• Therapeutic Potential: By identifying that size and shape can be uncoupled mechanistically, the study suggests new therapeutic strategies. Future drugs could potentially target specific pathways (e.g., modulating p38 vs. PI3K) to prevent maladaptive remodeling (e.g., excessive area expansion leading to heart failure) while preserving adaptive growth or correcting shape defects.
• Methodological Advance: The workflow serves as a blueprint for using systems biology to dissect complex cellular phenotypes, moving beyond single-pathway analysis to network-level understanding.

In summary, this paper establishes that Neuregulin-1 coordinates cardiomyocyte elongation and area expansion through distinct, temporally regulated signaling dynamics, where PI3K is essential for both, and p38 specifically regulates area, providing a refined map for targeting cardiac remodeling.