Muscle actuator sourcing: C2C12, primary human, iPSC or insect

For a first biohybrid actuator, use C2C12. It is a mouse myoblast line, it is cheap, it is available from any repository, it fuses readily, and the overwhelming majority of published biohybrid work uses it — which means when your construct misbehaves, someone has already written down why.

Move to primary human or iPSC-derived cells when the mouse origin becomes a problem: when the result must be human-relevant, when you need a specific genotype or disease mutation, or when you need force output above what a cell line reliably delivers. Consider insect dorsal vessel tissue only if room-temperature, incubator-free, self-contracting operation is the requirement, and accept that there is no commercial supply.

The four routes compared

Route 1: C2C12, and why it dominates

C2C12 is the default for a reason that has nothing to do with performance. It is immortalised, so it expands indefinitely; it fuses into myotubes reliably under simple serum withdrawal; the media is cheap; and there is a decade of biohybrid literature to compare against.

Its ceiling is real and now well characterised. Chen and colleagues, in benchmarking their own work against the field, state plainly that most C2C12 muscle tissues generate less than 1 mN of active force, with only a few reports exceeding 2 mN, and stresses consistently below 2 mN/mm². Their own result — 4.21 mN twitch and 7.05 mN tetanic, at 5.08 and 8.51 mN/mm² on a mean cross-section of 0.83 ± 0.15 mm² — is presented as the highest reported for C2C12-derived actuators and required a purpose-built self-training maturation platform to achieve.

For comparison at the low end of the same cell source, Raman and colleagues’ optogenetic C2C12 bioactuator generated up to 300 µN (0.56 kPa) of active tension in response to optical stimulation. That 20-fold spread within one cell type is the clearest possible demonstration that cell source is not the dominant variable in force output. Anchor stiffness, maturation protocol, matrix composition and construct geometry all matter more than which myoblast you started with, up to a point.

Choose C2C12 when: you are iterating on device geometry, control strategy or stimulation, and you need many cheap replicates. Which is to say, at the start of almost every project.

Route 2: primary human myoblasts

This is where the force data gets both better and better-characterised, because the biomedical literature has been measuring human engineered muscle for a decade.

Madden and colleagues expanded cells from ten separate human donor sources and formed functional myobundles from all ten, with an average specific force of 2.1 ± 0.9 mN/mm² twitch and 7.0 ± 2.2 mN/mm² tetanic. That “all ten” is the important part: primary human myoblasts work reproducibly, though the standard deviations tell you donor variability is not negligible.

Khodabukus and colleagues then showed what maturation does to those numbers. One week of intermittent electrical stimulation increased myobundle size, sarcomeric protein abundance, calcium transient amplitude by roughly two-fold and tetanic force by roughly three-fold, producing 19.3 ± 0.63 mN/mm² at 1 Hz against 9.1 ± 0.38 mN/mm² unstimulated — described in the paper as the highest specific force reported for engineered human muscle, though still well below native human skeletal muscle at 150–250 mN/mm².

On a commercial casting platform, Smith and colleagues report primary human myoblast engineered muscle tissues producing average twitch forces of 306.5 ± 22.2, 468.8 ± 37.5 and 568.0 ± 103.7 µN at days 7, 10 and 14, with tetanic forces of 740.8 ± 91.3, 1160.5 ± 118.8 and 1387.7 ± 275.7 µN over the same timepoints. Notably, unlike the iPSC-derived tissues in the same study, the primary tissues were still increasing at day 14.

Supply. Several catalogue suppliers list human skeletal myoblasts. For anything at scale or with donor requirements, Cook MyoSite is the significant name: they advertise muscle and other adherent cell production, cryopreservation at any scale, procurement and processing of muscle cells from specific donor populations, custom media formulations, and phase-appropriate GMP manufacturing in a 3,470 sq ft ISO 7 / ISO 5 cleanroom, backed by more than twenty years in muscle-derived cell therapy.

What to specify when ordering: donor age and sex, passage at shipment, purity by CD56 or desmin with a threshold, fusion index if the supplier reports it, and lot reservation if you need consistency across a multi-month programme. Donor-to-donor variation is the largest uncontrolled variable on this route, and the mitigation is buying one large lot rather than several small ones.

Route 3: iPSC-derived myogenic progenitors

The longest route and the only one that gives you genotype control.

The force data is competitive. In the Smith study, iPSC-derived engineered muscle tissues under optimised conditions increased from 385.8 ± 51.1 µN twitch at day 7 to 1156.3 ± 129.1 µN by day 10, with force plateauing thereafter at 943.1 ± 162.1 µN at day 14. Tetanic forces reached 2498.4 ± 58.3 µN at day 10. A second iPSC line in the same study reached lower tetanic outputs — 607.9, 1693.7 and 995.3 µN at days 7, 10 and 14 — which is a useful reminder that line-to-line variation within iPSC is as real as donor-to-donor variation within primary.

The genuine advantage shows up in disease modelling. The same study compared dystrophin-null iPSC-derived tissues against healthy controls at day 10: normal tissues produced 1319.9 ± 68.2 µN twitch and 1654.1 ± 97.6 µN tetanic, while dystrophic tissues produced 414.4 ± 21.9 and 267.2 ± 20.2 µN. That kind of isogenic comparison is not available on any other route.

Timeline is the cost. Curi Bio publishes 8–16 weeks for CRISPR line engineering, 4–6 weeks for iPSC expansion and banking, and 4–8 weeks for directed differentiation to skeletal myoblasts. If you start from a patient line and need a construct, that is a multi-month programme before any device work begins.

Choose iPSC when: the genotype is the point, you need isogenic controls, or you need a renewable supply of a specific human genetic background. Do not choose it because it sounds more advanced — for pure force output, primary human myoblasts reach comparable numbers faster.

Route 4: insect dorsal vessel tissue

The outlier, and worth understanding even if you never use it.

Akiyama and colleagues built an autonomously moving microrobot powered by insect dorsal vessel tissue, reporting a whole-tissue contracting force calculated at 20 µN — up from 4.7 µN in their earlier work — driving autonomous locomotion at 3.5 × 10⁻² µm/s. The forces are small. The operating envelope is what makes it interesting: room temperature, no CO₂ incubator, no external stimulation, spontaneous rhythmic contraction sustained over long periods.

For an untethered device that must function outside a tissue culture facility, that combination has no equivalent in the mammalian routes. Every mammalian construct in this field needs 37 °C, CO₂ buffering, medium exchange and, for most, an electrical or optical stimulus.

The supply problem is total. There is no commercial source. Every published use involves an insect colony and a dissection. Practically, this route requires either an existing entomology capability or a collaboration with a group that has one, and the sourcing question becomes a collaboration question rather than a procurement question.

Cross-cutting: what actually determines your force

Four levers, roughly in order of how much they move the number:

1. Chronic electrical stimulation. The single largest documented effect. One week of stimulation roughly tripled tetanic force in human myobundles and produced the field’s highest reported specific force. If your construct is underperforming and you are not pacing it, that is the first thing to change.

2. Anchor stiffness. Published Mantarray data show day-10 twitch force falling from 1156.3 ± 129.1 µN at 1.92 N/m post stiffness to 624.0 ± 154.6 µN at 0.96 N/m — a near-two-fold change from device mechanics alone. Two softer stiffnesses (0.16 and 0.48 N/m) bent so far under passive tension that no active contraction could be measured at all.

3. Media formulation. In the same study, maintenance in 2% horse serum throughout produced greater force development than switching to 15% fetal bovine serum after day 7, or than a serum-free knockout serum replacement formulation.

4. Cell source. Real, but smaller than the above three within the mammalian routes. A 20-fold spread exists within C2C12 alone depending on how the tissue was matured.

The practical implication for sourcing: do not spend three months acquiring a premium cell source to fix a force problem caused by an unpaced construct on soft posts in the wrong medium.

What to buy alongside the cells

None of these routes produces an actuator on its own. A working construct also needs:

• A casting device with anchors. eNUVIO’s OMEGA-MP states compatibility with all three mammalian routes — iPSC-derived myogenic precursors, primary biopsy cells and immortalised myogenic lines — at USD $440.00 for four devices and eight experiments. See 3D muscle culture devices.
• ECM hydrogel. Typically fibrin or collagen with a basement membrane extract. Lot variability here is a real source of tissue-to-tissue variation and is worth controlling by reserving lots.
• Growth and differentiation media, matched to the cell source.
• Stimulation hardware, for maturation and for actuation.
• A force readout — optical post tracking or a dedicated instrument.
• Four to six weeks of skilled culture labour per iteration.

How we can help

We source cells against a written specification — species, donor characteristics, passage, purity marker and threshold, lot reservation — and put the quotes side by side on a cost-per-construct basis rather than a cost-per-vial basis. Where the honest answer is “start with C2C12 and a $440 device”, that is what we will tell you, because a premium cell source does not fix a device problem. If your requirement is stated as a force, read the force specification reference first so the spec is comparable to published work.