When is a synbio company a gene therapy company?

You can find the consultation here.

At first glance, this may look like a technical regulatory update relevant mainly to advanced therapy developers. In reality, it raises a broader and more consequential question: where will regulators draw the boundaries of “gene therapy” for the next generation of biotechnology?

The arbitrary line between biological and synthetic

The current regulatory framework contains a distinction that is increasingly difficult to justify scientifically.

As formulated today, a GTMP must be a biological medicinal product, meaning that its active substance is derived from a biological starting material.

This has a counterintuitive consequence.

Two nucleic acid constructs may be functionally identical - for example, DNA templates used to produce an mRNA therapeutic - yet only one may fall within the GTMP definition because of how it was made. A version derived from a biological plasmid may be captured, whereas a chemically synthesised equivalent may fall outside the framework.

The MHRA now identifies this distinction as effectively arbitrary in light of modern manufacturing technologies and proposes to remove the requirement entirely.

There is an interesting parallel here with the recent shift in plant gene-editing regulation in the UK, and to some extent, the EU. In both contexts, the regulatory focus is moving away from crude distinctions based solely on how a product or organism was made. But the comparison is not exact: in plant regulation, the emphasis is increasingly on the nature of the resulting genetic change, whereas in the medicinal context, the MHRA’s proposal is more clearly anchored in therapeutic mechanism. The common thread is a move away from origin-based formalism, even if the replacement test is not the same.

From “what it is” to “what it does”

One of the most significant shifts in the consultation is a move away from defining gene therapy based on origin or composition toward defining it based on mechanism of action.

Under the proposed framework, a GTMP may include:

• ‍products containing recombinant or synthetic nucleic acids that undergo transcription or translation; and
• products whose mechanism of action involves sequence-specific genome editing.

In parallel, the consultation proposes removing the requirement that the nucleic acid must directly confer the therapeutic effect. Instead, classification may be driven by the properties of the active substance or its role in the system, even where genetic modification plays a supporting or enabling role. This makes the boundaries of “mechanism of action” especially important.

The proposed focus on “mechanism of action” may also pull a wider class of edited cell therapies into the GTMP category - including products that do not express recombinant nucleic acids at the point of administration. A cell therapy edited, for example, to reduce immunogenicity, improve persistence or remove an inhibitory pathway might previously have been viewed primarily as a somatic cell therapy. Under the new model, however, it could fall within GTMP if that genome edit is treated as part of the product’s relevant mechanism. That makes the undefined boundaries of “mechanism of action” particularly important: the key question may no longer be whether the edit delivers the main therapeutic payload, but whether it is considered to play a sufficiently important enabling role in how the product works.

Two products may ultimately deliver the same therapeutic molecule in a patient. Yet their regulatory classification could differ significantly, depending on the mechanism by which that therapeutic effect is achieved.

This illustrates a central principle of the consultation: regulation is driven by mechanism, not outcome.  A similar contrast may arise where the same gene is silenced by different routes. Silencing achieved through deliberate, sequence-specific genome editing may bring a product within the GTMP framework, whereas silencing achieved through epigenetic modification that does not alter the underlying genetic sequence, or administration of an siRNA, would, under the proposal, fall outside it. Again, the regulatory distinction turns not on the therapeutic objective, but on the mechanism used to achieve it.

Synthetic biology moves into scope

For the engineering biology sector, this shift is particularly significant.

The MHRA highlights existing regulatory precedent showing that products do not need to act directly on human cells to be classified as gene therapies. Genetically modified bacteria, for example, have previously been classified as GTMPs even where the administered genetic material remains within the microbial system.

The proposed definition reinforces and extends this position. It would capture products that:

• act through the expression of recombinant or synthetic nucleic acids, or
• achieve sequence-specific genome editing,

regardless of whether this activity occurs in human cells, donor cells, or other biological systems.

In practice, this means the regulatory analysis is agnostic to the “cellular context”. The “cell” may be human, donor-derived, or microbial. What matters is the role played by engineered nucleic acids, not the organism in which they are expressed.

This has clear implications for:

• engineered live biotherapeutics,
• programmable microbial platforms,
• synthetic genetic circuits, and
• cell-based delivery systems.

The same logic may also be relevant to modalities such as CAR-T, in vivo CAR-T approaches, and certain therapeutic nanostructures. Ex vivo CAR-T products already sit comfortably within the advanced therapy landscape, but the consultation’s mechanism-based framing may be particularly important for newer in vivo approaches in which engineered genetic information is delivered and expressed within the patient to generate the therapeutic cell population. Similar questions could also arise for DNA- or RNA-based nanostructures where the therapeutic effect depends not simply on the presence of nucleic acid material, but on its expression or its role in a sequence-specific editing system.

Genome editing becomes unambiguously gene therapy

Another notable proposal is the explicit inclusion of genome editing within the GTMP definition, regardless of the composition of the active substance.

Under the revised framework:

• any product whose mechanism involves deliberate, sequence-specific genome editing may be classified as a GTMP,
• irrespective of whether that mechanism relies on recombinant nucleic acids.

This reflects a forward-looking approach. The consultation explicitly anticipates the development of non-traditional genome editing systems, including those that may be protein-based or otherwise independent of nucleic acid delivery.

Equally important: What is not included

The consultation also draws clear boundaries around what falls outside the GTMP category. Products that modulate gene expression without altering the underlying genetic sequence - such as antisense oligonucleotides, siRNAs, microRNA-based therapies and epigenetic or epitranscriptomic modifiers - are generally not considered gene therapies. Vaccines for infectious diseases are also explicitly excluded, while phage therapies are not expressly addressed.

These distinctions are critical, as they define the boundary between gene therapy, nucleic acid medicines, and other emerging therapeutic approaches.

Notably, phage therapies do not appear to be expressly mentioned in the consultation either as included within or excluded from the GTMP framework. That contrasts with certain modalities, such as microRNA-based therapies, which are specifically identified as falling outside the category.

An engineered bacterium that expresses an antigen illustrates the point. If the product functions as a vaccine against an infectious disease, the consultation suggests it may fall outside the GTMP category by virtue of the vaccine exclusion, even though its mechanism involves engineered genetic expression within a microbial system. If, however, a similar platform were used in a therapeutic setting outside that exclusion, the analysis could point the other way. The scope of the vaccine carve-out may therefore be especially important for microbial synthetic biology platforms.

That distinction matters commercially as well as legally. Products that fall within the GTMP category remain subject to a demanding ATMP framework, whereas infectious disease vaccines - although still highly regulated - generally sit within a more established and operationally familiar medicines pathway.

Why classification matters

Classification is not merely a technical or administrative issue.

The designation of a product as a GTMP determines whether it falls within the Advanced Therapy Medicinal Products (ATMP) framework, which enables a risk-based regulatory approach tailored to complex biological medicines.

At the same time, it brings specific regulatory expectations, including:

• manufacturing standards,
• clinical development requirements, and
• pharmacovigilance obligations.

For developers, whether a product is classified as a GTMP can shape the entire development pathway.

A clear and predictable classification route is often as important as the flexibility it affords.

A question for the engineering biology community

For companies working in engineered microbes, programmable cells, synthetic nucleic acids, and next-generation genome engineering, the practical question is not simply whether a product may one day be classified as a GTMP. It is what should be done now so that, if it does fall within that framework, the business is not starting from scratch.

That means thinking early about classification, comparability and regulatory evidence generation, even at the platform stage. Companies should be asking whether their product’s mechanism could bring it within the GTMP definition, what that would mean for manufacturing strategy and controls, and whether their preclinical package is being designed to support later regulatory interactions. In some cases, it may also be sensible to engage early with the MHRA on borderline classification questions rather than waiting until clinical development is imminent. Importantly, the consultation is not proposing a lighter-touch regime for these products. GTMPs would remain highly regulated, with the same safety, quality and efficacy expectations. The significance of the proposal is different: it would place a wider range of relevant products within an established and well-understood framework, making the likely route more predictable even if it remains demanding.

For companies operating at the edges of the framework, there is also a strong case for clearer and earlier feedback from the regulator on the kinds of features that are likely to become problematic later - not only obvious issues such as the continued use of antibiotic resistance markers, but also questions around vector or chassis design, persistence, biodistribution, off-target activity, control of gene expression, shedding, insertional risk, immunogenicity, and the level of construct and process characterisation that will ultimately be expected. The earlier those signals are available, the easier it will be for companies to make sensible R&D choices before platform design becomes difficult to unwind.

For businesses in this space, readiness for a future GTMP pathway may increasingly need to be built into platform design, manufacturing assumptions and development planning from the outset. The definitions being developed today could shape regulatory strategy, evidence burdens, partnership models and investment decisions for years to come.

For many engineering biology companies, the GTMP question is no longer one for later. It is becoming a design-stage question now.