Rule-based modeling (RBM) offers a more scalable and modular approach to simulate these complex molecular interactions.

Methods

Experimental: Muscle cells were stimulated with 100nM of insulin → collected sample data on the time points of 0, 2, 5, 10, 20, 60 minutes → measured key proteins involved in insulin signaling pathways.

Computational: RBM was made by using BioNetGen software to generate a network of all possible reactions that could happen in the system

The system was solved using ODEs → calculated how the molecular levels change over time.

Includes 42 reaction rules and 101 parameters encoding interactions among 61 chemical species.

→ Used COPASI to perform a parametric sensitivity analysis to help identify key regulators in the pathway.

This difference in ERK1/2 was linked to the strength of the negative feedback in the model.

Artificially boosting feedback improved simulation accuracy with experimental data.

In contrast, upstream molecules react quickly to enable rapid signal propagation.

The overshooting response, followed by a steady state, enables rapid propagation along one signaling route, preparing molecules for other routes immediately afterward.

Removal of ERK1/2-GRB2/SOS negative feedback (blue lines) resulted in minor or no changes.

Fig. 5. P70S6K-IRS1 negative feedback is essential for a controlled glucose uptake, and when enhanced, reduces the insulin sensitivity and glucose uptake.

Sensitivity analysis revealed that GLUT4 and ERK1/2 are more sensitive to changes without ERK1/2-GRB2/SOS.

Additionally, its modularity permits expansion with new discoveries.

Overall, it is a powerful tool for understanding metabolic diseases and developing targeted therapies.

- Without modifications in the model

- Retrieval of GLUT4 values at every concentration of Insulin after 60 min

Model without p70S6K-IRS1 (red line):

- Model alteration - Reduction of Parameter k15 from 12 to 0

- Retrieval of GLUT4 values at every concentration of Insulin after 60 min

Model with p70S6K-IRS1 enhanced (green line):

- Model alteration - Enhancement of Parameter k15 from 12 to 8 times that value modifying his function:

f15()=8*(k15*(Vmax)*(p70S6K_pT389)^n_p70/(Kd_p70+(p70S6K_pT389)^n_p70))

- Retrieval of GLUT4 values at every concentration of Insulin after 60 minutes

Species: IR_phos_memb, PI3K_variant_p, AMPK_pT172, aSrc, Akt_pS

Behavior: Reach steady state in 2–5 min. Consistent, low variability, and minimal fluctuation across concentrations

Comparison:

• No noticeable visual differences among the three concentrations.

• The response is consistent and fast in all cases.

• Slight variation in magnitude, but the overall profile remains the same.

  
  

Cluster 2: Rapid overshooting (green)

Species: IRS1_SHP2_complex, IRS1_GS_complex, PKC_pT410, Akt_pT, PI345, IRS1_PI3K_complex, IRS1_pY, RasGTP

Behavior: Fast peak followed by a drop to steady state. Overshoot intensity strongly increases with IRS1 concentration.

Comparison:

• Significant increase in response amplitude with rising IRS1:

  • For example, IRS1_pY goes from ~5,000 (6u) to over 600,000 (600u).

  • IRS1_PI3K_complex, PI345, and Akt_pT also show major increases.

• The time to peak remains similar (fast), but the intensity of the overshoot greatly increases with concentration.

Cluster 3: Slow and growing response (yellow)

Species: GLUT4_memb, IR_RasGAP_complex, TSC1_TSC2_pS1387, mTORC1_pS2448, p70S6K_pT389

Behavior: Gradual increase over time. Reaches a higher plateau with increasing IRS1, reflecting delayed but sustained activation.

Comparison:

• The slow growth pattern is maintained at all concentrations.

• GLUT4_memb shows a drastic increase in magnitude:

  • From ~150,000 (6u) to 10 million (600u).

• p70S6K_pT389 and mTORC1_pS2448 also show stronger activation, suggesting IRS1-dependence.

• IR_RasGAP_complex slightly decreases with increasing IRS1, which may suggest an inhibitory or saturation effect.

  
  

Cluster 4: Slow overshooting (purple)

Species: Erk_ppY204_Y187, Mek_pS218_S222, aaRaf, aRaf

Behavior: Delayed peak followed by decrease. Shows marked amplification with higher IRS1 input.

Comparison:

• There’s a considerable increase in peak amplitudes with higher IRS1.

  • Erk_ppY204_Y187 goes from ~1 (6u) to ~20,000 (600u).

• The slow overshoot profile remains, but intensity varies significantly.

• Suggests that the Ras/MAPK pathway is highly sensitive to IRS1, although more slowly.

• The temporal profiles (curve shapes) do not change between concentrations, only the intensity.

• The cluster classification remains valid across all simulations, reinforcing model robustness.

Differences:

• Amplitude of responses increases dramatically with IRS1 concentration, especially in clusters 2, 3, and 4.

• Some species show non-linear responses (e.g., GLUT4_memb and IRS1_pY) to IRS1.

• IR_RasGAP_complex shows a decreasing trend, contrasting with the others.

- Without modifications in the model

- Both positives feedback loops are active:

• p70S6K -> IRS1 via the function f15()

• ERK1/2 -> GRB2/SOS via IRS1-SHP2 complex formation

- Simulation run for 60 min

Model without p70S6K-IRS1 (red line):

- Model alteration - removal of the positive feedback loop from p70S6K to IRS1

- Specifically, removal of the reaction using f15():

IRS1(YXXM,Y896,Y~u,S636~u)  -> IRS1(YXXM,Y896,Y~u,S636~p)  f15() 

- All other components of the model remain unchanged.

- Simulation run for 60 min

Model without ERK1/2 -> GRB2/SOS feedback (blue line):

- Model alteration - removal of the positive feedback loop from ERK1/2 to GRB2/SOS

- Specifically, removal of SHP2-related complexation reactions:

IRS1(YXXM,Y896,Y~p,S636~u) + SHP2(SH2) <-> IRS1(YXXM,Y896!1,Y~p,S636~u).SHP2(SH2!1) k27,k_27

IRS1(YXXM,Y896!1,Y~p,S636~u).SHP2(SH2!1) -> IRS1(YXXM,Y896,Y~p,S636~u) + SHP2(SH2) k41

- All other components of the model remain unchanged

- Simulation run for 60 min

- Excel files with experimental concentration and figure 5 results

- Codes in Python to generate the graphs

However, during the examination of the signaling pathway, it became evident that IRS1 is crucial at multiple stages of the signaling cascade. This led to the hypothesis that altering the initial concentration of IRS1 could significantly influence the model’s behavior. To simplify the analysis and improve computational accessibility, this hypothesis was tested using the existing model framework. This method enabled the assessment of how changes in IRS1 levels impact downstream signaling elements.

This strategy offers a practical avenue for investigating the regulatory role of IRS1 in the ISP pathway, providing valuable insights into its interactions within insulin signaling and aiding in the understanding of the underlying mechanisms involved.

• C1 – Blue: Fast responses reaching steady state within 2–5 minutes.

• C2 – Green: Fast overshooting responses that peak within 2–5 minutes and decay to a steady state by 10–20 minutes.

• C3 – Orange: Slow responses that gradually reach a steady state in 10–20 minutes.

• C4 – Magenta: Slow overshooting responses peaking between 5–10 minutes and returning to a steady state after 30–60 minutes.

IRS1 concentration strongly influences activation magnitude across nearly all clusters — particularly those involved in metabolism, growth, and protein synthesis.

Overshooting clusters (Clusters 2 and 4) show highly amplified peaks with increasing IRS1:

• This suggests positive feedback loops or signal amplification dependent on IRS1 abundance.

• IRS1_pY, PI345, Akt_pT, and RasGTP respond rapidly and intensely to IRS1, reflecting a strong dependency on IRS1 for triggering the PI3K/Akt and Ras/MAPK pathways.

Slow and growing signals (Cluster 3) — like GLUT4_memb, mTORC1_pS2448, and p70S6K_pT389 — show proportional and sustained increases in activity.

• These nodes are crucial for glucose uptake (GLUT4) and protein synthesis/growth (via mTOR/p70S6K).

• More IRS1 = stronger downstream effect = better glucose uptake and anabolic activity.

Adequate IRS1 ensures strong and rapid activation of metabolic pathways:

• Glucose uptake via Akt → GLUT4 translocation.

• Lipid and protein synthesis via mTORC1/p70S6K.

• Cell survival and growth through both PI3K/Akt and Ras/MAPK axes.

  
  

Insulin resistance/ Type 2 Diabetes:

• IRS1 levels are often downregulated, mislocalized, or dephosphorylated (via stress kinases or serine phosphorylation).

• The simulation at low IRS1 (6u) resembles this state:

  • Weak or delayed Akt activation → impaired GLUT4 translocation → hyperglycemia.

  • mTORC1 and p70S6K responses are diminished → impaired protein synthesis and anabolism.

  • Reduced MAPK signaling → compromised growth and repair functions.

Chronic high insulin (hyperinsulinemia):

• Can lead to IRS1 degradation or desensitization.

• Overactivation of mTORC1/p70S6K can trigger feedback inhibition on IRS1 itself.

  • The data suggests IR_RasGAP_complex decreases with IRS1, possibly hinting at such feedback.

IRS1 activates the PI3K/Akt/mTOR and Ras/MAPK pathways.

Impact:

• Promotes cell cycle progression.

• Supports anabolic processes (protein, lipid, nucleotide synthesis).

Relevance: Overactivation of IRS1 is linked to cancer development, especially breast, liver, and prostate cancers.

2. Neuronal Function & Brain Health:

IRS1 is expressed in neurons and mediates insulin’s effects on synaptic plasticity, memory, and cognitive function.

Impact:

• Enhances long-term potentiation (LTP).

• Supports neuronal survival via PI3K/Akt.

Relevance: Impaired IRS1 signaling is seen in Alzheimer’s disease (“Type 3 diabetes”) — insulin resistance in the brain leads to neurodegeneration.

3. Lipid Metabolism & Obesity:

IRS1 signaling influences lipogenesis, lipolysis, and fat storage through Akt and mTOR.

Impact:

• Promotes lipid synthesis in adipocytes.

• Modulates energy balance.

Relevance: IRS1 mutations or defects are associated with lipodystrophy, fatty liver, and obesity.

4. Skeletal Muscle Function:

In muscle cells, IRS1 mediates insulin-stimulated glucose uptake, glycogen synthesis, and protein synthesis.

Impact:

• Supports muscle growth and repair.

• Regulates glucose disposal.

Relevance: IRS1 defects in muscle are a hallmark of insulin resistance and sarcopenia (muscle wasting).

5. Genetic and Epigenetic Regulation:

IRS1 gene expression is modulated by:

• Transcription factors (e.g., FOXO, SREBP).

• MicroRNAs (e.g., miR-126, miR-145).

• Epigenetic marks (DNA methylation).

Relevance:

IRS1 expression is tissue-specific and environmentally regulated (e.g., by diet, exercise).

SNPs in IRS1 are linked to type 2 diabetes risk and metabolic syndrome.