Using Lab-Grown Human Muscle Models for Preclinical Drug Evaluation
Lab-grown human muscle models replicate key in vivo drug responses, enabling efficient and accurate preclinical testing.
Following the US Food and Drug Administration (FDA) announcement on the phasing out of animal testing, the drug development industry has welcomed a wave of new approach methodologies (NAMs).
From organs-on-a-chip to in silico screening, NAMs offer a human-centric approach to replace or complement animal testing at various stages of the preclinical pipeline. The short-term advantages—increased efficiency and reduced cost—can scale into much larger long-term benefits, with industry experts anticipating improvements in drug development success driven by greater predictive power.
However, to realize this goal, researchers must establish the validity and reliability of their models, ensuring they have the scientific capability to accurately predict human outcomes while also meeting expectations for regulatory readiness.
Prof. Mark Lewis, co-founder and chief executive officer of Myomaker Bio, understands these challenges all too well. Drawing on decades of academic experience in regenerative and exercise medicine at Loughborough University, Lewis shared his insights into the challenges of producing human muscle models and his broader aim of accelerating and improving drug development success with Myomaker Bio’s human muscle models.
Could you outline the development journey behind your range of muscle models, and the main challenges you faced in bringing them to this stage?
Mark Lewis, PhD (ML):
Human tissues such as skin, muscle, and bone are made up of cells, so the first challenge in recreating tissue outside the body is obtaining those cells and producing enough of them to build a viable model. The next challenge is recreating the structural environment that those cells need to function properly. For example, muscles attach to bones via tendons, so we had to design anchor points to replicate that structure.
In the early days, we improvised. Our first anchoring units were made from orthodontic wire, which is biocompatible, and from mesh sourced through a haberdasher that happened to have the right weight properties. It was something of a cottage industry—we were literally assembling these systems by hand. But that approach wasn’t very reproducible and required significant operator skills to achieve consistent results. The emergence of 3D printing changed that dramatically. It allowed us to produce identical components at low cost and reliably, as long as we ensured the materials were biocompatible.
We also had to determine how best to measure muscle contraction. Force transducers proved to be the most effective solution, and as they became more widely available, the process became much easier. Another challenge was sourcing human motor neurons, but advances in stem cell technology have since made them more accessible, allowing us to incorporate them into our models.
IH:
To what extent do the engineered muscles predict human in vivo drug responses?
ML:
We believe our muscle models behave in ways that closely reflect normal human muscle, and we’ve established several validation points that support this (Table 1). Ultimately, predicting drug responses is exactly what our company is focused on—it’s the core reason we founded MyoMaker Bio.
Table 1: Myomaker Bio organ responses to pharmaceutical agents.
| Pharmaceutical agent | Response of Myomaker Bio organ |
| Barium chloride (BaCl₂) | BaCl₂ induced injury and an in vivo-like inflammatory response while preserving a regenerative mononuclear cell population. |
| Leucine | Leucine activated downstream effectors of mTORC1 in a dose-dependent manner, promoting protein synthesis and supporting muscle growth. |
| d-tubocurarine | d-tubocurarine inhibited nicotinic acetylcholine receptors, suppressing spontaneous contractions. |
| Insulin | Prolonged insulin exposure attenuated insulin-stimulated glucose uptake, insulin signaling, and induced compensatory changes in glucose transporter expression. |
| Resolvin E1 (RvE1) | RvE1 attenuated inflammation and mitigated inflammation-associated reductions in muscle function. |
| Testosterone | Testosterone promoted myotube hypertrophy and enhanced differentiation and fusion-impaired myoblasts. |
IH:
What are the current constraints or limitations of your models, from biological modeling and measurements to manufacturing?
ML:
At present, we primarily work with what I would describe as “normal” muscle tissue. For example, we might take a biopsy from a healthy individual in their twenties with no underlying muscle disease. Participants complete a detailed questionnaire beforehand so we can confirm their background.
The next step is extending this approach to disease contexts. Can we build models using muscle samples from people with muscular dystrophy, for instance? I believe we can.
Muscular dystrophy
Muscular dystrophy is a group of genetic disorders characterized by progressive muscle degeneration. Mutations in genes responsible for maintaining healthy muscle structure and function mean that muscle fibers are more susceptible to damage and are not sufficiently repaired.
The further challenge is modeling more complex conditions, such as motor neuron disease, where multiple biological systems are involved. Recreating those disease environments in a meaningful way will be key to deepening our understanding of these conditions.
Motor neuron disease
Motor neuron disease is a group of neurological disorders characterized by progressive degeneration of motor neurons—nerve cells that transmit signals from the brain and spinal cord to the skeletal muscles for voluntary muscle movement. As neurons deteriorate, muscles weaken and waste, ultimately impacting a range of functions including movement, speech, swallowing, and breathing. The cause of motor neuron disease is not fully elucidated, but it is thought to be multifactorial, involving genetics and environmental factors.
IH:
Do you anticipate that models like these could ever replace animal testing?
ML:
Yes, I do, but it won’t happen overnight. Moving directly from animal testing to no animal testing isn’t realistic because animal studies are deeply embedded in both legislation and pharmaceutical development processes.
The framework that guides this transition is the three Rs: refinement, reduction, and replacement.
Models like ours can already contribute significantly to refinement by helping researchers better understand biological processes and ask more targeted questions. That, in turn, allows them to reduce the number of animals used in experiments. Over time, replacement might become possible, but there are still many regulatory and scientific hurdles to address before that happens.
What we are already seeing is refinement in practice. Where a scientist might have previously carried out three procedures in animals, two of those may now be conducted using an in vitro model instead.
IH:
What role do you see lab-grown human muscle models playing in transforming drug development timelines and success rates?
ML:
Even with recent advances, drug development remains slow, expensive, and heavily reliant on animal models that do not always predict human outcomes accurately. Our human muscle platforms are designed to help bridge that gap by providing more human-relevant data earlier in the process.
By improving the predictive power of preclinical testing, these models have the potential to make drug development faster, more efficient, and ultimately more successful.
Our ambition is that this approach will significantly shorten development timelines and increase the likelihood that new drugs will succeed in the clinic.
Add Technology Networks as a preferred Google source to see more of our trusted coverage.
