Curi Bio, a Seattle-based tissue engineering firm, and Battelle Memorial Institute have announced a strategic partnership to commercialise a three-dimensional human neuromuscular junction (NMJ) assay for botulinum neurotoxin (BoNT) potency testing, positioning the platform as a GLP-ready alternative to the mouse lethality bioassay across pharmaceutical and medical countermeasure development pipelines.
The announcement arrives at a moment of genuine regulatory momentum. The FDA published its NAM validation roadmap in April 2025, outlining a stepwise strategy to reduce and ultimately replace animal-based preclinical testing, and followed that with a further draft guidance in early 2026 clarifying the validation framework for non-animal models submitted in support of drug applications. The NIH simultaneously committed USD 150 million to human-relevant research alternatives. That policy architecture now gives platforms like Curi Bio’s Mantarray ecosystem a clearer filing path than they had even two years ago, which changes the commercial calculus for a partnership of this kind.
What the mouse lethality bioassay’s persistence reveals about the size of the opportunity
The mouse lethality bioassay has been the regulatory standard for BoNT potency testing for decades. Its persistence is not a reflection of scientific confidence but of regulatory inertia and the lack of broadly validated alternatives at sufficient scale. BoNT products require batch-by-batch potency release testing, a requirement that generates significant and ongoing use of animal models across both therapeutic and medical countermeasure pipelines. Industry observers tracking the field note that even companies that have internally developed cell-based alternatives have in many cases continued running the mouse bioassay in parallel, either to satisfy reference standard requirements or to maintain comparability data for existing market authorisations. The systemic barriers are well documented: regulatory frameworks in multiple jurisdictions still list the LD50 test as the default, and companies face the burden of demonstrating equivalence to displace it.
This is the context in which the Curi Bio-Battelle deal matters. The platform is not simply a research-grade NMJ model — it is being positioned as a commercial-scale, GLP-deployable assay that can generate data directly comparable to what regulators have historically accepted from animal studies. That distinction separates it from the majority of academic and early-stage microphysiological system work, which has demonstrated biological plausibility without completing the regulatory translation step.
How the Mantarray platform’s functional readout differs from earlier cell-based potency assays
Cell-based potency assays for BoNT are not new. The most widely referenced precedent is the SiMa cell-based assay developed for onabotulinumtoxinA, which received FDA, Health Canada, and European approval for BOTOX lot-release testing by measuring SNAP25 cleavage via ELISA. That approach, while validated and accepted, captures a biochemical surrogate for toxin activity rather than a direct functional measurement of neuromuscular transmission. Curi Bio’s NMJ model takes a different approach: it generates real-time contractile force data from three-dimensional co-cultures of iPSC-derived motor neurons and skeletal muscle tissue, with NMJ-specific inhibition confirmed by optical stimulation that selectively activates the neuronal component.
The distinction has practical implications for potency testing fidelity. A functional assay measuring loss of contractile output at the NMJ is mechanistically closer to the physiological process BoNT disrupts in vivo than an ELISA surrogate of substrate cleavage. Published data from Curi Bio’s laboratory shows the platform reproduces expected dose-dependent responses to both BoNT/A and BoNT/B presynaptically, and to postsynaptic antagonists including tubocurarine, with greater than 95 percent tissue formation success rates across a 24-well format. Whether that mechanistic alignment translates into superior regulatory predictivity relative to validated ELISA-based approaches is a question that validation studies conducted under GLP conditions will need to answer — and it is precisely this validation work that the Battelle partnership is intended to enable.
Why Battelle’s GLP infrastructure is the critical constraint this partnership addresses
The history of microphysiological systems is populated with technically capable platforms that have failed to achieve regulatory adoption, not because the science was inadequate but because the operational infrastructure for GLP-compliant deployment did not exist or was prohibitively expensive to construct. Curi Bio’s Mantarray ecosystem has demonstrated strong performance at the research level and has attracted partnerships, including earlier MOUs with Genetox and DreamCIS for a BoNT potency filing targeting US FDA and Korean regulatory approval. The Battelle relationship addresses a different and arguably more consequential gap: the institutional capacity to transfer an assay into a GLP environment, validate it against regulatory-standard acceptance criteria, and deploy it at the scale required for routine lot-release testing across multiple sponsors.
Battelle brings specialised biosafety infrastructure, regulatory affairs expertise, and established relationships across both commercial pharmaceutical accounts and government medical countermeasure programmes. The latter is significant. BoNT is a Category A bioterrorism agent, and the US government maintains stockpiled medical countermeasures including both antitoxins and therapeutic BoNT formulations for potential emergency use. Lot-release potency testing for those stockpiled countermeasures is a government-funded, recurring requirement — one that creates a defined commercial demand independent of the cosmetic and therapeutic BoNT market. Regulatory watchers suggest the inclusion of MCM pipelines in the partnership scope reflects a deliberate effort to address both commercial and governmental revenue streams simultaneously.
What the FDA’s NAMs validation framework requires from platforms seeking regulatory acceptance
The FDA’s 2026 draft guidance on NAM validation establishes four core validation principles: clear definition of the intended regulatory context of use; demonstration of human biological relevance; reproducibility and transferability of the method; and performance benchmarking against accepted reference standards. For a platform like Curi Bio‘s NMJ assay, the context-of-use requirement is the most immediately operationally demanding. Regulators will want to know whether the assay is being proposed as a standalone replacement for the mouse bioassay, a complement to existing testing packages, or a primary method for specific product categories. The answer shapes the evidence package required.
Industry observers familiar with NAM validation timelines note that the path from a scientifically robust platform to a regulatory-accepted lot-release method typically involves multi-year studies, interlaboratory reproducibility exercises, and comparability datasets generated across multiple product matrices. The Curi Bio-Battelle partnership appears structured to generate exactly that evidence base, but the timeline for regulatory acceptance remains genuinely uncertain. FDA guidance documents create permissive conditions for NAM submissions; they do not accelerate the validation work required to meet those conditions. Sponsors considering whether to incorporate the NMJ assay into development or lot-release programmes will need clarity on how far along that validation pathway the platform has progressed before making package-defining decisions.
Where scalability and manufacturing standardisation create residual risk for the platform
The Mantarray platform’s commercial design incorporates features intended to reduce operator-dependent variability: automated contractility measurement across 24-well plates, direct-from-thaw neurosphere generation, and standardised iPSC-derived cell inputs. These design choices address one of the longstanding criticisms of microphysiological systems as lot-release platforms — that the complexity of living tissue models introduces biological variability incompatible with the precision required for quality control assays. The greater than 95 percent tissue formation success rate reported in Curi Bio’s published data is encouraging on this metric.
However, published data also acknowledges that the platform’s NMJ formation is incomplete and asymmetric in terms of pre- and postsynaptic marker co-localisation, which is described as indicative of an immature NMJ. For BoNT potency testing purposes, what matters is whether the functional response to the toxin is sufficiently reproducible and sensitive relative to the reference standard — maturity of synaptic morphology is secondary to assay performance. That said, regulators reviewing a replacement assay for a decades-old standard will scrutinise biological maturity as part of their human relevance assessment, particularly for a product category where subtle potency differences carry direct patient safety implications.
Cell supply chain is a further variable. Curi Bio has an existing arrangement with Cook MyoSite for GMP-compliant iPSC expansion and differentiation, which supports lot-release readiness for the cell inputs. Maintaining consistent cell line performance across large-scale, multi-lot validation exercises is a challenge well recognised in the field — and one that any regulatory submission will need to address through documented comparability protocols.
What clinicians, regulators, and countermeasure developers will be watching next
For commercial BoNT manufacturers, the near-term question is whether the Curi Bio-Battelle platform can achieve regulatory acceptance for lot-release testing in any major jurisdiction within a credible timeframe. A first regulatory acceptance — even for a single product in a single market — would shift the competitive calculus significantly, because it would establish a precedent that other sponsors could reference in their own submissions. The Genetox-DreamCIS collaboration targeting BOTAONE approval through the US and Korean FDAs represents one potential pathway to that precedent, though its status and timeline have not been publicly updated since the 2024 MOU announcement.
For government countermeasure programmes, the value proposition is distinct: reducing animal burden in ongoing stockpile lot-release testing while potentially generating superior functional data on MCM potency. Regulatory watchers note that government agencies operating under animal welfare mandates have an institutional incentive to accelerate acceptance of validated alternatives that does not apply to the same degree in commercial pharmaceutical procurement decisions.
The broader implication of the Curi Bio-Battelle partnership is that the industrialisation of microphysiological systems for regulatory use is moving from theoretical to operational. The gap between a validated research platform and a GLP-deployed assay has historically been where academic-origin NAM technologies stall. By pairing a technology developer with an organisation that has the regulatory infrastructure to close that gap, the partnership represents a structurally different approach to commercialisation than the field has typically seen. Whether the platform delivers on that commercial logic will depend on what the GLP validation data shows — and on whether the regulatory agencies reviewing the first submissions are prepared to move as quickly as the policy framework now implies they should.
Curi Bio and Battelle commercialise human NMJ assay as GLP-ready alternative to mouse bioassay for botulinum toxin testing added by Pallavi Madhiraju on
View all posts by Pallavi Madhiraju →