Government proposes new gene therapy workarounds

 

By Wendi Strauch Mahoney

A significant challenge for genetic engineering boils down to the logistics of transporting and delivering cell and gene therapies.  For therapeutic, commercial, and large-scale research applications, DNA and RNA are manufactured in specialized facilities.  Whereas many DNA reagents can ship dried at ambient temperatures, RNA-based products and formulated therapeutics frequently require controlled handling and cold-chain logistics.  They are then delivered into cells using viruses, lipid nanoparticles, electroporation, or other methods.

The “move the molecule” model underpins everything from research plasmids to modern mRNA therapeutics.  If successful, this research signals a paradigmatic shift in the manufacturing of DNA and RNA.

DARPA is openly proposing something more radical: a “massless information transfer” using using light as an information channel to drive DNA/RNA synthesis inside cells, rather than physically shipping the sequence itself.  According to the announcement, “technology developed under GO [Generative Optogenetics] will enable unprecedented control over cellular behavior by facilitating genetic programming with single-cell spatial resolution, temporal precision to deliver messages to a cell sequentially, and remote, scalable dissemination of genetic instructions.”  The program seeks to “harness the power of light to direct the synthesis of DNA and RNA directly within living cells.”

DARPA positions GO as a solution to constraints of current de novo (from scratch) nucleic acid synthesis — centralized manufacturing, scaling limits, and the burdens of downstream assembly and delivery.  The public GO page explicitly ties the vision to “resilient supply chains” and transformative applications across medicine, agriculture, and manufacturing.

On December 9, 2025, DARPA’s Biological Technologies Office (BTO) posted the new program page for Generative Optogenetics.  In parallel, DARPA issued a Special Notice (DARPA-SN-26-19) and scheduled an in-person only “Generative Optogenetics Proposers Workshop” for January 7, 2026 to discuss the new technology.  The text pictured below provides a brief explanation from DARPA, comparing current practices with the newer GO technology (DARPA-SN-25-46):

Also pictured below is a diagram explaining how GO may help overcome some of the “traditional chemical and enzymatic methods for the de novo synthesis of DNA and RNA sequences for novel protein production.”  The big question DARPA asks: Is it possible to produce DNA/RNA sequences with “physics-based (light, mechanical, sound, electromagnetic, thermal, etc.) control of cellular processes?”

DARPA’s central concept is a “Nucleic Acid Compiler” (NAC) — a protein complex that could be expressed inside a living cell and would convert an optical input — think a sequence of light pulses — into a biological output, a newly written DNA or RNA sequence that the cell can then transcribe and/or translate using its normal machinery.  Note: DARPA’s public GO page does not spell out NAC details.  However, the public summary of the workshop notice describes it this way.

The old model requires external physical manufacturing (chemistry, purification, shipping, delivery) of genetic media, whereas the GO model seeks to stream the “file” via light and have the cell write it locally.  In sum, the cell becomes the manufacturing environment, enabling the relaying of genetic instructions to living cells without moving the matter that encodes those instructions.  Notably, GO does not make delivery vanish completely, because cells would still need an initial way to express whatever machinery enables the light-to-sequence writing.

By the end of the program, scientists hope to demonstrate in vivo (inside living cells) synthesis of full-length coding sequences between 3 and 6 kilobases (kb) — gene-sized outputs.  A 3–6 kb coding sequence can encode a complete functional protein — often an enzyme, receptor, structural protein, or transcription factor.  In practical terms, it’s enough to change what the cell can do.

Some of the challenges

A full coding sequence can persist in varying ways, with different implications.  If the output is RNA, effects could be transient but powerful.  If the output is DNA that persists (episomal or integrated), effects could be longer lasting, which intensifies requirements for shutoff, containment, and preventing unintended activation.

Related persistence questions have also been studied in the context of COVID-19 mRNA vaccination.  In that setting, investigators have reported detectable vaccine-related SARS-CoV-2 antigen in plasma for days after vaccination in some studies, and reports of spike antigen and vaccine mRNA signals persisting for weeks in draining lymph node germinal centers as part of ongoing immune responses.  These findings do not, by themselves, establish harm or indefinite “persistent production” in the general case, but they illustrate why “shutoff,” localization, and containment matter for any platform intended to generate gene-length payloads in vivo.

Possible uses or benefits of GO

• Instead of ordering DNA, cloning, transforming, and verifying for every variant, a cell that already contains the “write” machinery could be programmed optically to generate specific coding sequences or libraries on demand, a potentially faster solution.
• Optics can be spatially patterned according to region, or even single cells.  A GO platform could allow for different genetic instructions in various micro-locations.
• Could allow for on-demand therapeutics in accessible tissues, like the eye or skin, instructing engineered cells to produce a therapeutic protein or RNA.
• Cells could be instructed to generate antigen sequences locally for rapid re-tasking during outbreaks, consistent with DARPA’s focus on national security–relevant biotechnology.
• Rapid re-tasking of microbial “factories” — a culture might be re-parameterized optically to express different enzymes or pathways without shipping new DNA every time.
• In controlled agriculture settings (greenhouses/vertical farms), where light is already engineered (timing, intensity, wavelength), a GO-like control channel could, in principle, trigger timed expression of stress-response or growth pathways — more plausibly controlled expression than stable, heritable genome rewriting in the near term.

The implications of this new technology are worth debating, even at the research stage.  The upside for medical applications could be great.  It would allow for in situ programming in reachable tissues and would allow therapies to be updated rapidly and delivered with precision.

Risks

The risks are likely not fully understood and, therefore, warrant serious scrutiny.  Chief among them are biosecurity concerns and the potential misuse of a tool that could be repurposed for dual-use ends.  Some experts warn that, if deployed without robust safeguards, remote cellular manipulation could make COVID-era abuses child’s play.

Remote manipulation of cellular biology also raises ethical and security questions that go well beyond a narrowly “benign” model of localized instruction transfer for therapeutic use.  If biological functions can be triggered or altered remotely, how would we reliably detect such activity and defend against it?  What safeguards would prevent translation or transcription errors that yield unintended, off-target proteins?  How would errors be minimized, detected, and if necessary neutralized in vivo?  And how would dose, duration, and spatial specificity be measured and controlled to prevent runaway or cumulative effects?

Any platform that makes genetic programming faster, cheaper, and more scalable increases the importance of authentication, access control, and governance.  DARPA’s own emphasis on preventing unintended activation is a tacit admission that a light-addressable genetic interface introduces new and sobering risks for humankind.  Recent debates over informed consent and risk communication during the pandemic underscore how quickly public trust can erode when people believe material risks were downplayed or inadequately disclosed.

https://www.americanthinker.com/blog/2025/12/government_proposes_new_gene_therapy_workarounds.html