Oligonucleotides – short strands of chemically synthesized nucleic acids – offer exceptional precision in modulating gene expression. From silencing toxic genes to editing RNA and correcting splicing errors, oligonucleotides can open up new therapeutic strategies that were once out of reach with traditional small molecules and antibodies. In diseases where the genetic culprit is known but difficult to target with small molecules or antibodies, oligonucleotides can slip past traditional barriers and shut down problems at the molecular source.
While oligonucleotides were initially pursued for rare, monogenic disorders, such as Huntington’s disease, momentum has surged across the broader pharma landscape. Advances in chemical modifications, delivery systems, and regulatory understanding are all helping to expand the oligonucleotide field.
But promise alone isn’t enough. To truly unlock the potential of oligonucleotide drugs, developers must tackle critical discovery and development hurdles, such as how to identify optimal RNA targets, design safe and potent sequences, and ensure delivery to the right tissues.
In this interview, Hilary Brooks, VP and Modality Lead for Oligonucleotide Therapeutics at Evotec, shares her unique perspective on oligonucleotide drug discovery.
What drew you into the oligonucleotide space?
I’ve worked in different areas, including metabolic diseases and oncology, but when I look back, oligonucleotides were a common thread. I’ve worked with large European consortia trying to develop cell-penetrating peptides to deliver oligonucleotides, and I’ve worked with Australian biotech companies trying to apply DNA delivery of siRNA therapeutics to oncology. I’ve also worked with French biotechs on oncology drugs.
Moving into oligonucleotides more fully – and into a larger company like Evotec – was a natural next step. In biotech, you're always trying to get the right assays in place and it’s incredibly difficult to do with a small team.
At a big company, all the expertise is already there to help design the best, safest, and most potent drugs. I enjoy having a big Lego set where all I need to do is pick the pieces I need! Maybe it’s rapid-fire mass spectrometry or perhaps it’s deep disease biology expertise. Or maybe it’s something specific like retinal organoids, which take time to grow but can provide an exquisitely tailored response for a particular disease indication.
It also means I can go after targets that haven’t even been fully validated yet – because I can validate them. We’ve got the mechanisms in place – and even before they’re used as therapeutics, oligonucleotides are one of the best tools biologists have for figuring out whether a gene is involved in a particular disease mechanism.
When a partner comes to us with a question – such as how to do something – if I don’t know the answer I can ask around. Sure enough, there is usually someone in the office next door who is an expert! It’s fantastic. We can go after genetic targets, prove their relevance using in vitro and in vivo models, and if the right model doesn’t exist yet, we’ll build it.
What strategies are used in oligonucleotide drug discovery?
I’ve seen many different strategies – including quite a few serendipitous ones! As a partner helping to progress these molecules along the drug discovery pathway, we see people coming in at every possible stage.
Sometimes someone will approach us with a tool compound that they were using to screen small molecules before they realized that the siRNA was working better than any of the hits. In this case, you have a validated target, so the next step is to find or design a compound that can do the job.
This typically begins with an in silico screen, but it’s not just about looking for a complementary RNA sequence; the target you pick must be truly relevant for your patient population. RNA targets are not static. RNA expression varies by age, disease state, stress levels, and even cell type. There are countless variants, so you must ensure your target exists in the relevant patient population. You do not want to spend years developing a drug only to find out, at the eleventh hour, that 20 percent of your intended patients either carry a single nucleotide polymorphism or do not express that RNA isoform at all. That would be a devastating – and completely avoidable – “oops.”
So, start strong with robust in silico analysis. That includes off-target predictions, of course. But even then, your results are only as good as the RNA databases you’re using. Honestly, there’s no substitute for putting your compound in cells and just seeing what it does.
This is where the power of transcriptomics, proteomics, and metabolomics comes in. These tools can tell you what your molecule is actually doing, and help identify and eliminate promiscuous compounds early on. Remember, this is a sequence-based targeting strategy. You can have perfect matches and imperfect matches – and we are still far from predicting what those matches might give with different chemical modifications and their varying placements along the strand. We need more data, which means wet-lab testing, cell assays, and real-world experimentation.
In summary, you need a strong screening cascade where you evaluate safety and potency head-to-head right from the start. Don’t wait until late-stage development to ask whether a compound is safe. We already know where to look and how to assess risk in oligonucleotide drug development. Use that knowledge and build it into your discovery strategy from day one.
What are your thoughts on using AI for oligonucleotide drug development?
You can trust the data that AI is producing – providing you’ve put in data that’s trustworthy in the first place. I don’t think the industry is at this stage yet for oligonucleotide drug development.
We don’t know if AI prediction of potency and safety can be target-agnostic. We see a lot of oligo programs in a target agnostic, chemistry and disease area agnostic fashion but we know that both the type of chemistry used and the sequence – together in combination – affect the parameters of potency and safety. People have their preferred chemistries, and they’ve got their preferred safety assays. If you haven’t benchmarked to the same standards, and you’re not using the same chemistry, then you can’t draw the same conclusions.
Some people call one thing toxic and another not toxic, but it depends on what toxicity assay you’re using and what your threshold is set to. Maybe it’s not hepatotoxic, but it could be immunotoxic. As an industry, we need to standardize and benchmark. This wasn’t done early on. Something would be labeled non-toxic based on how one assay was run, and then it would fall out of the clinic for a completely different toxicity reason. A model cannot be trained on this kind of data. You need a big dataset that has been built consistently.
At Evotec, we are building AI into the in silico design stage and we want to share our success stories – to help make design faster, libraries smaller, and modifications more consistent. I’d love to see other success stories in the industry too.
You work with a lot of projects in various phases. How do you go about optimizing them – especially from a safety perspective?
In terms of the screening cascade, you need to look at potency – but potency in oligonucleotides is a function of many different factors. The same can be said for safety. The chemical modifications you make and whether you’re using a delivery moiety or not, such as a conjugate, all matter.
A screening will tell you if your target has occupancy, binds your oligo, and engages the process you are aiming for – whether that’s splice switching or RNase H-mediated cleavage. But it doesn’t tell you what’s going to happen in vivo – especially when your compound includes modifications or a conjugate. Can it even get into the cell? Does the conjugate affect affinity for transport proteins or receptors that help get it inside? These are key questions.
The goal should be to use the most realistic system as early as possible. With oligos, you tend to go into vivo models earlier than with small molecules, but even before that, I recommend using cell types that are as primary and disease-relevant as possible. Induced pluripotent stem cell (iPSC)-derived models can be helpful for that. If you’re working with antisense oligonucleotides (ASO) and relying on unassisted or gymnotic uptake, those systems allow you to test the multifunctionality of the ASO, including both potency and safety.
The process is complex, but also quite logical once you understand where the pain points are. You start screening in any available cell line, narrow it down, and then move to more complex models. Eventually, you’re testing your complex molecule in a complex cell type and you’re also going into in vivo models quite early.
Biodistribution is critical. If your ASO or siRNA doesn’t reach the tissue where it’s needed, it’s not going to be an effective drug. This is a particular issue in oncology. We know what to target, but the industry can’t yet get oligos to metastatic tumor cells.
I’d love to see oligonucleotides used in oncology. But the reality is: most people do not die from their primary tumor. It’s metastasis that kills – and we can’t reach those cells yet.
What other advice do you have for oligonucleotide development?
My advice is to focus on the low-hanging fruit and do the work early. Doing work later is not only much harder – but will be far more expensive. For example, a lot of early stage projects come to me that are highly promising in terms of proof of concept – but that have completely neglected safety. You shouldn’t wait until drug candidate nomination before checking for off-target effects.
We also need to work together collectively as an industry to solve the formulation and delivery challenges associated with oligonucleotides. How do we get oligonucleotides to go where we need them to go? Unmodified oligonucleotides have a very short half life and tend to accumulate in the liver and kidneys. Fortunately, there’s a lot of interesting work taking place trying to push their biodistribution to different tissue types such as brain or muscle.
Even when the target organ is readily accessible to oligos, such as the liver, there are interesting advances pushing the field forward; take the example of cardiometabolic diseases. At the moment, there are four major players with oligonucleotide drugs all targeting lipoprotein(a) currently in phase II/III. The efficacy of all of them against the target is looking strong, so ultimately it will be the safety and ease of administration that will determine the winner. Some programs are going for monthly injections. Others are looking at once-a-year administration. Patient comfort will likely be at the front and center of the race – and that’s a win for patients.
The industry is also advancing on other fronts. Take vaccines as an example. It’s currently possible to combine five vaccines into one shot. Oligonucleotides could potentially do the same thing. Oligonucleotides are just strings of nucleotides and their interactions with each other are minimal. Providing you are not forming insoluble aggregates, then why not combine targets?
This approach could be especially promising for oncology – once we figure out how to get oligos to tumors. Cancer is a moving target. You can’t go after just one hallmark of cancer with a single drug because the cancer adapts. This is why combination therapy is the standard in oncology now. With oligos, it could be possible to deliver a combination therapy in one tube. You could target multiple hallmarks of cancer at once, such as proliferation, glycolysis, and DNA repair. If we know what the genetic targets are for those processes, we could build oligonucleotides for each, mix them together, and deliver them in one shot.
And what's more, the off-target effects don’t necessarily compound the way they do with small molecules. Each oligo has its own off-target profile – so if you’ve individually worked out the safety of each sequence, combining them doesn’t automatically amplify toxicity the way traditional drugs might.
There’s huge potential here. But for that to happen, we need to solve the delivery challenges with oligonucleotides.
However, I’m positive about the progress being made. There are some great molecules already on the market, but these aren’t the best molecules we’ll ever see. The best ones – the most potent and safest – still haven’t made it to market. They are coming and they will continue to change the game.
Newsletters
Receive the latest pharmaceutical news, personalities, education, and career development – weekly to your inbox.
