Seven Months from Mutation to Medicine | Dr. Tamar Grossman of La Jolla Labs

What we talked about.

Sunny Ray talks with Dr. Tamar Grossman of La Jolla Labs about how her team went from a patient's mutation to a dosed medicine in seven months. Grossman traces two decades in RNA therapeutics, from a UCSD postdoc to a decade at ISIS, and explains how La Jolla Labs survived a funding winter by running a fee-for-service model for rare disease programs. She describes an AI-driven design platform that can raise or lower gene activity or correct splicing, and says it found the Spinraza sequence from only 20 candidates. The conversation covers the ALS case, where generous partners and an accelerated N-of-one regulatory path allowed dosing of two patients. Grossman also explains in plain terms how an antisense drug targets only the mutant gene copy, why the company is focusing on neurodegenerative diseases such as ALS, Parkinson's and Alzheimer's, and what she hopes to achieve with pharma partners and investors in the coming months. She closes with advice on funding science creatively and a look at where rare disease medicine is headed.

Dr. Tamar Grossman explains how La Jolla Labs moved from an ALS mutation to dosing patients in seven months using AI-designed RNA drugs.

The questions, and the answers.

What first pulled you toward RNA as the thing worth designing drugs against?

I have spent about two decades in RNA therapeutics. I started as a postdoc at UCSD working on bioinformatics and genetic diseases, then was recruited to ISIS, a pioneer of ASO therapeutics. I kept a visiting professor position for a while, but after a couple of years I knew this was what I wanted to do. I spent ten years there, and some of my drugs are now in the clinic.

Seven months from finding a mutation to dosing two ALS patients. What made that speed possible?

Our partners were incredibly generous. We told them we had a patient dying and asked for help, and they said yes, for free, so no contracts slowed us down. It was also an N of one, which has accelerated regulatory pathways. We ran a single rat study to show safety, got approval, and dosed the patient the next day.

How do you lower the mutant TDP-43 and leave the healthy copy alone, in plain words?

Our drugs are antisense oligonucleotides, about 20 building blocks of chemically modified DNA, delivered directly into the spinal fluid. We design the sequence around the mutation, so it creates a mismatch with the normal allele. Like a zipper, it only fits the mutant copy. We screen a few options to find the ones that best tell mutant from wild type.

Most teams screen hundreds of candidates. Why do you only need about 20?

It comes from our experience and our software and AI tools, which let us visualize the sequence and find the best regions to design against. As a test, Spinraza took over 600 ASOs and ten years to identify. When we put that gene into our software, it found the actual drug from only 20 designs. With our platform it might have taken one year.

How do you decide whether to raise a gene's activity, lower it, or fix splicing?

It all depends on the mutation. We start with the patient's sequencing data and make sure we understand the disease mechanism, such as a toxic gain of function or a splicing defect. That tells us whether to up-regulate, down-regulate or modulate splicing. Then we take multiple shots on goal with different mechanisms and pick the best one to move forward.

What did five years of paying your own way teach you about funding science?

It is not easy, and you have to be creative. Venture capital was not available in the first couple of years, so we built a platform that earned revenue from services. We also licensed part of our software to an institute to improve cash flow. If you stay lean, believe in the technology and stay positive, you can make it happen.

Who should reach out to La Jolla Labs after this episode?

Three groups. Foundations, hospitals and clinicians who want RNA therapeutics for rare disease can use our service. Big pharma and biotech can partner with us on larger indications like Parkinson's and Alzheimer's. And investors, because we plan to meet with them around JPM in January to fund assets that could reach the clinic in under two years.