RNA's Evolutionary Logic Is Quietly Rewriting The Future Of Therapeutics: A Conversation With Ahmet Berkyurek

A conversation with Ahmet Berkyurek

If evolution is the original systems engineer, Ahmet Berkyurek, Ph.D., cofounder & CEO, Cambridge Medixine Ltd., argues that modern RNA drug development is entering an era where biological constraints are not obstacles but design rules — and where the most transformative innovations may arise from building therapies that behave more like RNA and less like molecules engineered against their nature.

To understand this evolution — both molecular and conceptual — Life Science Connect’s Michael Soloway recently spoke with Berkyurek. His work, alongside contributions from colleagues Virginia Castilla Llorente, Ph.D., and Andreas W. Claas, Ph.D., explores RNA’s extraordinary adaptability.

In their conversation, Berkyurek reflected on the Darwinian logic that underlies RNA biology — variation, selection, and adaptation — and how these ancient principles continue to guide the next generation of RNA therapeutics.

RNA As Evolution’s Most Adaptive System

Advancing RNA: You’ve said before that when it comes to RNA therapeutics, “biology has already solved more problems than we have.” What do you mean by that?

Berkyurek: We often look at RNA through the lens of what we want it to do today — encode a protein, silence a gene, activate an immune pathway, or self-amplify. But none of those capabilities arose because biology was trying to build medicines. RNA is not just a molecule we manipulate; RNA’s properties are the result of billions of years of evolutionary tinkering.

RNA had to:

• fold quickly
• respond to small chemical or physical cues
• interact transiently with proteins
• store information without committing it
• degrade predictably and cleanly
• act catalytically when needed
• sense stress, temperature, energy availability
• move to different locations.

These pressures shaped its chemistry and architecture. When we treat RNA as a blank slate for engineering, we lose insights embedded in those pressures. But when we tune our designs to align with evolutionary logic, RNA becomes more predictable and more powerful.

What I’m saying is that evolution has given RNA a playbook. We’re still learning to read it.

Advancing RNA: Do you think modern RNA technologies — from mRNA vaccines to small activating RNAs — represent distinct “evolutionary branches” in the broader tree of RNA-based medicine?

Berkyurek: Yes, definitely. Each class of RNA technologies represents a different function and property, similar to evolutionary lineages. These can be branched as:

• protein-expression branch
• gene silencing branch
• gene activating branch
• stability branch
• fast onset of action branch
• editing and guiding branch
• translating genetic information (tRNA).

They use different endogenous RNA pathways, have distinct structural adaptations, potentially occupying a different therapeutic niche, and evolve in their own selective pressure based on clinical route, drug delivery, etc.

Reframing ‘Fragility’: Transience As A Therapeutic Feature

Advancing RNA: A lot of therapeutic design tries to overcome RNA’s instability. You argue that instability was evolutionary intent — not an error.

Berkyurek: Exactly. We call RNA “fragile,” but that word is a human projection. RNA’s short half-life is one of its critical biological functions. Without it:

• cells couldn’t respond to environmental change
• differentiation signals wouldn’t be reversible
• immune responses couldn’t be turned off
• stress responses would spiral
• embryonic development would lose temporal precision.

For years, progress in RNA therapeutics has been defined by engineering milestones: cleaner oligo chemistry, tunable nucleoside modifications, smarter LNPs, and more predictable manufacturing. Each advancement has elevated RNA from a fragile biological intermediary to a programmable therapeutic modality. These engineering feats, while necessary, still obscure a deeper truth the industry has not fully absorbed.

RNA binding proteins (RBPs) are really the ones defining the fate of the RNAs in the cells. The fact that we don´t know them that well and they are difficult to produce has led to humans developing other strategies to control the fate of the RNA, including stability. Using RNA binding proteins in combination with RNA, we could shape much better the fate of the therapeutic RNAs, as that is how cells do it.

Biology leverages RNA for processes that need to be precise and finite. That is not fragility. That is kinetic intelligence.

For therapeutics, this means the goal isn’t always to maximize stability; it’s to match the kinetics of the desired biological effect. In some cases, longer-lived RNA is valuable; in others, precise degradation is the therapeutic mechanism.

If we embrace degradability where appropriate, we stop fighting RNA and start collaborating with it.

saRNA And Replicons As Evolution-Inspired Machines

Advancing RNA: Self-amplifying RNA (saRNA) has surged in interest. From your perspective, what is the deeper evolutionary significance of these systems?

Berkyurek: saRNA platforms echo primordial strategies for efficiency. Early life operated under severe energy constraints. How do you achieve large functional output from minimal starting material? You replicate.

But replication was never random. It was:

• gated
• context-dependent
• resource-sensitive
• error-prone but tunable.

Today’s replicon systems capture only a fraction of that sophistication.

The next innovation wave will require understanding:

• how amplification interacts with tissue-specific stress pathways
• how local metabolites influence replicase activity
• how replication fidelity shapes antigen expression
• how the cell’s innate sensors can gate, accelerate, or restrict amplification.

In other words, we must understand saRNA not just as a high-expression tool but as an eco-systemic interaction between RNA, host, and environment. Evolution already wrote the script — we’re now learning the staging directions.

RNA Is Tissue-Aware — And We Need To Design Like It

Advancing RNA: The way RNA folds, interacts, and self-regulates almost mimics natural selection at a molecular level. How might that intrinsic adaptability inspire new approaches in RNA drug design?

Berkyurek: I believe the inherent adaptability of RNA as a molecule opens the door to designing dynamic, rather than static, RNA medicines. A prime example is mRNA molecules that can function across different tissues, guided by RNA sensors embedded in smart RNA technologies. The next generation of RNA therapeutics may incorporate secondary structures optimized for stability in specific tissues while degrading more rapidly elsewhere. Probably, different secondary structures promote interaction with different RBPs, which in the end triggers expression, degradation, or movement to a different location within the cells. Over the next decade, we may see tissue-specific mRNA delivery systems precisely tuned with tailored secondary structures.

Advancing RNA: You’ve used the phrase “RNA is tissue-aware,” which is unusual language in drug development. What do you mean?

Berkyurek: We often behave as though RNA functions uniformly across biological environments. That’s not true. A piece of RNA entering a hepatocyte and a piece entering a dendritic cell experience fundamentally different worlds:

• Ionic environment
• Temperature microgradients
• RNA-binding proteins
• RNA decay machinery
• Innate immune sensors
• ATP availability
• Crowding and viscosity

These change folding kinetics, immunogenicity, and translation efficiency.

So, when I say RNA is “tissue-aware,” I mean that RNA’s behavior is determined by local environmental forces. If we design therapeutics without acknowledging those forces, we design against the grain of biology.

The future is tissue-optimized RNA. Not one formulation, but many — each reflecting the environment it must navigate.

AI As A Selective Pressure, Not Just A Design Tool

Advancing RNA: AI is everywhere in drug development. But your framing — AI as a selective pressure — is novel. Can you elaborate?

Berkyurek: Biological evolution works through variation and selection. What AI provides is the ability to rapidly create variation — millions of hypothetical RNA sequences, structures, folding pathways — and then evaluate them under constraints.

But the most important part is the constraint. If we let AI freely optimize, it generates molecules that function well in silico but not in cells. If we impose evolutionary-like constraints — energy cost, environmental noise, folding intermediates, error tolerance —the designs become biologically coherent.

In other words, AI becomes a simulated evolutionary ecosystem.

We are not just designing RNA — we are evolving it in a computational environment that mimics selective pressures biology once imposed.

That is a paradigm shift.

Revisiting RNA’s Role In Inheritance And Memory

Advancing RNA: There’s growing work on transgenerational RNA inheritance. For many developers, this feels fringe. Should it matter to therapeutic design?

Berkyurek: Transgenerational RNA inheritance is real in multiple organisms, and evidence in mammals is emerging. Whether or not RNA therapeutics ever approach that boundary, the underlying insight is important: RNA can be a carrier of persistent biological information, not just transient signals.

For drug developers, this means:

• some RNA-triggered effects may last longer than expected
• RNA may alter chromatin states indirectly
• RNA can impact stress responses across cell generations
• small RNAs can shape long-term regulatory programs.

The goal isn’t to make medicines hereditary. The goal is to understand RNA’s depths, so our therapies behave predictably, ethically, and safely.

At Cambridge Medixine, we draw on a similar evolutionary perspective to develop what we believe to be the first non-viral self-amplifying RNA platform for therapeutic applications and beyond. Guided by the idea that aspects of ancient RNA biology may still be reflected in certain DNA-based enzymes, we explored primordial self-amplifying DNA systems, examined their enzyme architectures, and identified conserved RNA-associated motifs alongside their DNA-functional domains.

This Is Not A Constraint-Free Molecule — And That’s Good

Advancing RNA: You frequently say the biggest mistake we can make is trying to “turn RNA into DNA or protein.” Why?

Berkyurek: DNA is for storage. Protein is for stable function. RNA evolved to do something else entirely.

Its properties — conditional, transient, malleable — are not limitations; they are what RNA is. When we try to force RNA to behave like a protein drug or a persistent gene therapy, we run into problems.

But if instead we ask: What can only RNA do?

We unlock:

• logic-gated therapeutics
• spatially aware therapies
• tunable bursts of expression
• therapies with natural clearance
• programmable cell-state manipulation
• multilevel regulatory RNA circuits.

These aren’t compromises. They are advantages unique to RNA.

The more we lean into RNA’s identity, the more innovative — and safer — our therapeutics become.

Toward An Evolution-Informed Era Of RNA Therapeutics

Advancing RNA: What does RNA therapeutics look like if the field fully embraces evolutionary logic?

Berkyurek: I’d offer three themes:

1. RNA becomes a dynamic therapeutic, not a payload

We stop viewing RNA as a passive messenger and start using it as a biological event — one that activates, adapts, and extinguishes with precision.

2. Therapy design becomes tissue-centric, not molecule-centric

We will design RNAs, LNPs, and expression kinetics specific to the tissue environment, mirroring how evolution specialized RNA behavior across cell types.

3. AI-driven evolution becomes a design engine

Instead of simulating folding alone, AI will simulate selection pressures. We will evolve RNA for ecological fit, not just chemical function.

Conclusion: Listening To What RNA Has Been Trying To Tell Us

Berkyurek says the next decade of RNA therapeutics will not just be defined by new chemistries or delivery systems. It may be defined by a philosophical shift — one that recognizes RNA’s inherent intelligence, its adaptability, its environmental sensitivity, and its evolutionary roots.

In Berkyurek’s words: “Our greatest innovations may come not from overpowering RNA’s nature, but from embracing it. If we let RNA show us what it was built to do, we’ll build therapies we can’t even imagine yet. Evolution is the best engineer we’ve ever had. We just need to learn how to listen.”

About The Expert

Ahmet Berkyurek, PhD, is co-founder and Chief Executive Officer (CEO) of Cambridge Medixine (CamMed). He completed M.Sc. and Ph.D. degrees at Osaka University with a research fellowship from the Japanese Government, and a B.Sc. degree from Istanbul Technical University (ITU). Berkyurek continued his scientific career at the University of Cambridge as a research fellow with a Marie-Curie Fellowship and worked in the biotech industry at the intersection of biochemistry, cell biology, RNA biology, oligonucleotide and vaccine/RNA therapeutics. Berkyurek was a finalist for Researcher of the Year 2023 by Cambridge Independent Science and Technology Awards for his discoveries in RNA therapeutics to prevent SARS-CoV-2 infections and received the CEO of the Year 2025 Award for the UK from the Global Excellence Network for his contributions to mRNA therapeutics research.