Muscle Regeneration Agent-Based Model Predicts Enhanced Recovery Outcomes with Altered Cytokine Dynamics

Megan Haase

​

Department of Biomedical Engineering, University of Virginia, Charlottesville, VA, USA�Email: mh2uk@virginia.edu

Twitter: @biomeganics

​

IMAG/MSM Working Group

June 8^th, 2023

​

2

Muscle injuries account for more than 30% of all injuries and are one of the most common complaints to orthopedics

Exercise-induced muscle damage

Skeletal muscle has remarkable regenerative capabilities

Background

Muscle regeneration is essential for everyday life

Valle et al. (2011) Br J Sports Med

Muscle regeneration is a highly coordinated cellular process dependent on various signaling factors

3

Inflammatory phase

Muscle damage

Damaged fiber

Background

Skeletal muscle regeneration requires numerous cell types and microenvironmental signals

Yang & Hu (2018) J Orthop Translat

4

Computational modeling allows us to study complex dynamics and gain new insight into muscle regeneration

Background

5

ABMs allow simulation of intricate interactions between cells

Example model agents

Satellite stem cell

Neutrophil

phagocytosis of damage

apoptosis

Macrophage

phagocytosis of damage and apoptotic cells

Division signal: TNF-α + VEGF - TGF-β

Differentiation signal: 3*IL10 - HGF - TNF - TGF-β

Background

6

Limited diffusion and simplified cell dynamics

Previous regeneration models weren’t built for studying cytokine dynamics

Martin et al. (2015) J Appl Physiol

Focuses on spatial cell dynamics and diffusion

No spatial diffusion or spatial dynamics of inflammatory cells

Virgilio et al. (2018) J Appl Physiol

Background

7

Goal: Build ABM to predict regeneration outcomes with altered cytokine dynamics

1. develop a computational model of muscle regeneration that includes cellular and cytokine spatial dynamics as well as the microvascular environment
2. calibrate the model to capture cell behaviors from published experimental studies
3. validate model outcomes by comparison with multiple published experimental studies
4. conduct in silico experiments to predict how altering cytokine dynamics could impact muscle regeneration

8

Complex cell behaviors and cytokines were modeled spatially

Model Development

9

Spatial characteristics are defined from histology

Model Development

10

Microvascular growth and remodeling plays a key role in muscle regeneration

Umek et al. (2019) Histochem and Cell Bio

Model Development

11

Dynamic ECM properties were integrated to more accurately represent cytokine diffusion with altered collagen density

Filion & Popel (2005) Am J Physiol - Hear Circ Physiol

r_s = cytokine radius

r_f = fiber radius

ϕ = fiber volume fraction

D_∞ = cytokine diffusivity in free solution

D = diffusivity of cytokine in matrix

​

Model Development

12

Cellular behaviors are governed by literature derived rules

Model Development

13

Cellular behaviors are governed by literature derived rules

Model Development

14

15

CaliPro can identify a robust parameter space for stochastic biological models

Joslyn et al. (2020) Cellular and Molecular Bioengineering

Model Calibration

16

Rules that have not been experimentally quantified create unknown parameters

Model Calibration

See supplemental slides for 38 rule references

17

Parameter density estimation and partial rank correlation coefficient to calibrate the model

Model Calibration

18

CSA recovery, SSC and fibroblast count outputs were fit to literature data

Ochoa et al. (2007) Am J Physiol - Regul Integr Comp Physiol

Murphy et al. (2011) Development

Model Calibration

19

Model replicates inflammatory cell count and capillary count per fiber area

Hardy et al. (2016) PLoS One

Wang et al. (2018) J Neuroimmunol

Nguyen et al. (2011) Sci World J

Ochoa et al. (2007) Am J Physiol - Regul Integr Comp Physiol

Model Validation

20

¹Teixeira et al. (2003) Muscle & Nerve

²Raimondo & Mooney (2018) PNAS

³Deng et al. (2012) J Immunol

⁴Hardy et al. (2019) Skelet Muscle

⁵Arsic et al. (2004) Mol Ther

⁶Lu et al. (2011) FASEB J.

⁷Chen et al. (2005) Am J Physiol Cell Physiol.

⁸Liu et al. (2016) Cell Bio Int.

Model Validation

Model inputs altered to simulate various experimental conditions

21

Model Predictions

ABM provides deeper understanding of response to altered angiogenesis during muscle regeneration

22

Model Predictions

Cytokine knockout perturbations revealed crosstalk and temporal interplay between cytokines

23

LHS-PRCC relationships between cytokine parameters and key regeneration metrics

Model Predictions

24

Combined alterations of cytokine dynamics enhance muscle regeneration

Model Predictions

25

Agent-based modeling provides new hypotheses for in vivo experiments

Conclusion

• LHS and PRCC are useful for complex calibration and can be implemented to identify potential therapeutics mechanisms
• Cytokine dynamics play an essential role on regeneration, with single KO conditions leading to cascading impacts
• The ABM allowed for a deeper exploration of the temporal effects of cytokines on regeneration metrics
• In-silico results found that combined alterations of specific cytokine decay and diffusion parameters enhanced regeneration outcomes
• Future work will explore treatment optimization and alterations with various injury types

Acknowledgements

26

M3 Lab Members:

Silvia Blemker, PhD

Xiao Hu, PhD

Emily McCain, PhD

Matthew DiSalvo

Ridhi Sahani

Allison McCrady

Mario Garcia

Jacob Dunn

Undergraduates:

Autumn Routt

Anne Felipe

Brendan Shea

Keerthana Vijayaragavan

Josiah Calhoun

CC3D Collaborators:

James Glazier

T.J. Sego, PhD

Tien Comlekoglu

Alexa Petrucciani

Funding:

Figures created with BioRender.com

27

Thank You!

28

Thank You!