DNA Trivalent Binders: DNA Nanotechnology Meets Viral Geometry!

Happy New Year! 🎉 I’m actually writing this on the 31^st , so a year ago (my father’s humour rubbed off on me during the holidays).

Thank you for reading. It has been an amazing 2025! And I couldn’t be more grateful. And we’ve hit a major milestone, right before the year’s end: 400 subscribers! And the best subscribers, by the way.

So, to start the year on the right foot, let’s jump into some science!

Binders work better when they are together: antibodies, peptides, aptamers. Wouldn’t it be great to develop multimeric binders from the start? Today’s paper solves this problem!

DNA Triple Threat

Binders: Biotech’s Workhorses

Binders are a staple in biotech.

They let us investigate cells in microscopy, interrogate protein-protein interactions, and even treat diseases. So, they are super important in basic and translational research.

Antibodies are the classic high-affinity and selective binders, but they come with problems: animal-based production, immune side effects, and delivery difficulties.

To solve the problems of natural binders, scientists turned to synthetic alternatives. Short peptides, AI-designed nanobodies, and my personal favorites, aptamers.

Aptamers: DNA Antibodies

Aptamers are short, single-stranded DNA or RNA molecules that fold into specific 3D shapes. They’re practically the nucleic acid version of antibodies, and they bind to proteins, small molecules, viruses, and more!

Aptamers offer some advantages over antibodies:

• Cheaper to make

• More stability

• Less immunogenic

Plus, chemical modifications for nucleic acids have exploded over the years, resulting in better stability and binding!

How Aptamers are Made

Aptamer production looks different from that of antibodies.

Most aptamers come from SELEX (short for Systematic Evolution of Ligands by Exponential Enrichment):

1. You start with a huge library of random DNA/RNA sequences.

2. Select the ones that bind to your target molecule.

3. Amplify these “winning” sequences (using PCR).

4. Repeat this cycle to enrich for aptamers with high affinity and specificity.

And you got yourself a new DNA/RNA binder!

I’m a fan of aptamers, but they are not without problems. Often, they bind weakly compared to antibodies, and the binding doesn’t always translate into functional activity.

For example, you might find that an aptamer binds to a protein, but it doesn’t actually inhibit its function! And that’s a problem.

Trimeric Proteins: Viruses’ Door to the Cell

But that’s not the only weakness of SELEX.

Lots of clinically important viruses have large trimeric protein complexes on their surfaces, which they use to enter cells. Scientists have proved that assembling binders for these proteins in a target-specific shape improves their functional effect.

For example, you can take a DNA nanostructure and attach multiple monomeric aptamers or antibodies to transform it into a multimeric binding structure.

But simply copying and pasting a binder multiple times doesn’t always work. It requires extensive optimizations, and there is no guarantee of the performance.

A better way would be to evolve binders directly in a multivalent, target-matching geometry, instead of creating monovalent binders and then trying to multimerize them. But it’s impossible!

Or is it?

MEDUSA: Expanding the Directed Evolution Toolbox

Today’s paper explores this seemingly impossible task. With DNA nanotech!

The authors built MEDUSA, a DNA-based trivalent multimerization scaffold that presents chemically-modified aptamers to match the target protein pattern!

In practice, MEDUSA is a small, triangular DNA scaffold where you can hang up to three aptamers.

Instead of selecting monovalent binders, MEDUSA biases aptamer selection towards cooperative binders with the correct triangular arrangement and produces new classes of active binders, inaccessible via classic methods.

But let’s see how it works!

Design Principles: How Does MEDUSA Work?

The authors used MEDUSA to design binders for SARS-CoV-2.

The spike trimers create a triangular pattern on the surface of the particles, binding to cellular proteins and letting them in. Blocking this binding blocks the virus entry!

The key idea is that geometry drives selection. Geometrical simulations show that the binding affinity is the highest when the scaffold matches the target.

Even small deviations of ±2 nm sharply reduce avidity! In addition, shorter DNA linkers allow better geometry matching and more precise engagement.

So, the simulations gave them predictive rules for scaffold size and linker length!

Scaffold Choices and Library Chemistry

Once they had some good idea where to start, they started designing nanostructures.

The team chose a cyclic single-stranded DNA (cssDNA) scaffold with three hybridization sites: easy to make, and nuclease-resistant.

But what did they put on it?

For their library, they didn’t go for plain DNA aptamers. No, they chose FNAPs.

These are chemically-modified nucleic acid polymers produced by ligase-mediated polymerization. The side chain modifications mimic amino acids and increase the binding of these DNA sequences.

And then, in the selection we go!

MEDUSA: Multivalent SELEX

The selection pipeline looks like this:

1. Translate DNA templates into FNAPs.

2. Hybridize 3 FNAPs to the cyclic scaffold to form trivalent MEDUSA.

3. Perform affinity selections against the trimeric spike.

4. Reverse-translate the recovered FNAPs (the ones binding to the target) to DNA templates.

5. Iterate!

So, a multivalent SELEX! The results are FNAP sequences that bind to the target in a trivalent manner.

They also compared 2 selection strategies.

• Monovalent pre-enrichment for several rounds, then multimerize.

• Fully multivalent from round 1 (the real MEDUSA).

Sequencing analysis showed that strategy 2 yielded more diverse outcomes and produced unique sequences that only work when multimerized.

m1 vs m2 vs m11: Comparing Sequences

The authors focused on 3 families of selected sequences: m1, which dominated the monovalent selection, and m2 and m11, which emerged uniquely using MEDUSA multivalent selection.

The team compared the monovalent vs multivalent binding performance:

• m1 shows low-nanomolar monovalent affinity, with no increase when 3 were present on the scaffold

• m2 has ~200-nM monovalent affinity, but gains ~10-fold after trimerization

• m11 had no detectable monovalent binding, but when assembled trivalently, the binding affinity jumped to 24 nM! A pretty good binder.

MEDUSA discovered new binders that require multimerization to work, and you’d never find these sequences with a standard SELEX!

Do They Work? Functional Testing

Now, the real question.

The SARS-CoV-2 spike trimer binds to a cellular protein called ACE2. If this interaction is inhibited, the virus can’t enter cells and can’t cause harm.

The trivalent m2 and m11 successfully inhibit the ACE2-spike interaction, while m1 binds but doesn’t block. The classic problem with SELEX-discovered aptamers!

m2 and m11 were selective, binding only to the wild-type spike, not to mutants, with m11 binding more broadly.

Wrapping It up!

Cool work!

I’m a big fan of aptamers: coming from the DNA side, I find them easy to work with, and I see lots of potential in bringing more of them into the clinic and diagnostics.

Nanometer-scale positioning is the strongest advantage of DNA nanotech, IMO. And MEDUSA uses it perfectly! Plus, the same scaffold can be reused for other viruses.

Another advantage of MEDUSA is that you don’t need to know the structure of your target, just to purify it (maybe you don’t even need it, for an extracellular target). So, it complements nicely other techniques, where you might need structural information.

And the authors even showed that MEDUSA works as a sensor! I think aptamers are especially good for sensing purposes, where their lower cost and robustness can really shine.

So, cool idea! Go here, read the paper, and get inspired!

If you made it this far, thank you! What do you think of aptamer-based therapeutics? How do you think they compare to more standard techniques? Reply and let me know!

More Room:

• Natural RNA Origami, Part II: Some time ago, we covered some natural, cool RNA constructs. Now, this study reveals that some naturally occurring non-coding RNAs can form large, protein-free multimeric assemblies, expanding the known complexity of RNA-only structures. Using cryo-EM, the authors solved high-resolution structures of ROOL and GOLLD RNAs, showing they assemble into hexameric, decameric, and tetradecameric architectures stabilized by conserved RNA tertiary interactions.

• Zebrafish + DNA Origami: A blast from the past! A paper from one of my ex-colleagues in Stockholm. This study uses live zebrafish embryos to track the in vivo biodistribution and clearance of fluorescent DNA origami nanosheets. The authors show that coating the nanostructures with K-PEG alters their tissue distribution by reducing accumulation in the caudal hematopoietic tissue and slowing clearance by scavenger endothelial cells and macrophages. Overall, the work establishes zebrafish as a powerful single-cell–resolution platform for studying how surface modifications control DNA origami behavior in living organisms.

• DNA Graph Problems: I used some cool graph theory to design DNA origami, back in my PhD. This work studies a graph theory problem inspired by DNA self-assembly, aiming to design tile sets that assemble into a target graph using as few bond-edge types as possible. The authors derive an upper bound on this minimum using properties of multipartite subgraphs and vertex degree conditions, and they apply the framework to graphs corresponding to the Platonic solids. I don’t get all of it, but it’s awesome to read!

• Share Plenty of Room with founders or builders

  I help biotech and deep tech companies transform complex technologies into engaging content that builds credibility with investors, partners, and potential hires. Let’s chat!

• Know someone who’d love this?
  Pass it on! Sharing is the easiest way to support the newsletter and spark new ideas in your circle.

• Got a tip, paper, or topic you want me to cover?
  I’d love to hear from you! Just reply to this email or reach out on social.