From discovery to dosing: Trace Neuroscience's UNC13A bet is now in the clinic
By Joy N. Ismail, PhD
When Trace Neuroscience announced on June 22 that first patients had been dosed in the LAUNCH ALS trial, it felt like the right moment to revisit our original podcast conversation with co-founder and CEO Eric Green. Eric joined us in December 2024, when Trace had just launched with its Series A and the clinic was still a future goal. The distance between that conversation and this milestone is worth revisiting now that TRCN-1023, their antisense oligonucleotide (ASO) targeting UNC13A in ALS, has reached the clinic.
A target born from two threads of genetics
UNC13A has a peculiar history. It first turned up in Sydney Brenner's lab in the 1960s, in mutagenesis screens in C. elegans looking for worms that moved poorly. Decades of biochemistry then established UNC13A as an essential component of the presynaptic vesicle release machinery. Loss of UNC13A prevents vesicles from docking, leading to paralysis in worms and in mammals.
In 2009, UNC13A emerged as one of the first two GWAS signals in ALS, and alongside C9orf72, it remains one of the most robust hits in the disease to this day. For over a decade, the underlying mechanism remained unclear.
That changed in 2020, when Eric's team at Maze Therapeutics, working with Aaron Gittler's group at Stanford (and in parallel with Pietro Ferrata's team at UCL), cracked open the mechanism. The key was TDP-43, a nucleic acid binding protein that in healthy neurons cycles between nucleus and cytoplasm. In 97% of ALS cases, TDP-43 gets trapped in the cytoplasm. When that happens, it can no longer bind the UNC13A mRNA, and without that binding, a cryptic exon gets spliced in. That cryptic exon triggers nonsense-mediated decay, and the UNC13A protein drops. The genetic variants associated with ALS risk and faster progression sit directly in that cryptic exon region, right where TDP-43 binds, and the risk alleles specifically interfere with that binding, making the splicing defect worse.
In one elegant mechanism, the GWAS signal, the rare Mendelian biology, and the dominant pathological feature of ALS (TDP-43 mislocalization) all snapped together.
Why a low odds ratio isn't a small opportunity
One of the most useful moments in our conversation with Eric came when Patrick pushed back on the genetics: the UNC13A common variants carry an odds ratio well below 1.3. That sounds modest, but Eric drew the right analogy. Common variants in HMG-CoA reductase, the target of statins, also have small effect sizes on LDL. Statins became one of the most impactful drug classes in history. The size of the genetic signal tells you something about the gene's population-level influence on phenotype under normal conditions; it says nothing about the ceiling of a pharmacologic intervention.
In the case of UNC13A, the risk variants modestly exacerbate the cryptic exon inclusion. An ASO that completely suppresses that inclusion should, in principle, have far larger effects than the variants alone predict, in the same way that saturating HMG-CoA reductase with a small molecule moves LDL far beyond what the GWAS effect size would imply. Rare, severely damaging mutations in UNC13A cause early-onset neurodevelopmental and neuromuscular disease in humans, extending the allelic series and reinforcing just how much biology is riding on this protein.
The therapeutic approach
TRCN-1023 is a single-stranded ASO administered intrathecally. The design is conceptually analogous to nusinersen in spinal muscular atrophy (SMA). Whereas that drug blocked a cryptic splice site to restore full-length SMN protein, TRCN-1023 blocks the pathological cryptic exon in UNC13A to prevent its inclusion and restore protein expression. The precedent from SMA is meaningful. Splice-switching ASOs can work, they can be durable with infrequent dosing, and intrathecal delivery can achieve CNS exposure that systemic routes can't.
Preclinical data reported at the time of company launch showed effective cryptic exon suppression in rodent and non-human primate models, with an emerging safety profile that supported moving into humans.
The clinical program
The global program runs on two tracks. FUNCTION ALS is a Phase 1/2 randomized, double-blind, placebo-controlled trial authorized in the UK and Netherlands, expected to enroll approximately 30 participants across North America and Europe. Participants receive TRCN-1023 or placebo with 24 weeks of follow-up, with biomarker analyses and digital assessments built in.
LAUNCH ALS, the IIT being run in partnership with Tenacia Biopharmaceutical at Beijing Tiantan Hospital under principal investigator Yilong Wang, has already dosed its first patients. With approximately 25 participants, it gives Trace a faster read on early safety and PK while generating the kind of global dataset that modern regulatory submissions increasingly demand.
Why now
When Eric founded Trace in 2024, the question of timing was central to the thesis. The mechanistic insight was there. The modality, ASOs, had been validated in SMA and Huntington's. The preclinical package was in hand. What Maze couldn't provide was a fully dedicated team with neuroscience-specific clinical development expertise and a laser focus on this one disease. That's what Trace was built to be.
Eighteen months on, first patients dosed. For a disease with no curative treatment, no approved therapy that meaningfully alters disease course in the 90% without a targetable monogenic cause, and a median survival of two to five years from diagnosis, that is not a small thing.
Listen to the full episode below.