The scientists who built a $1.5 billion idea — then walked away to bet on a harder one

In 1957, eight scientists left Bell Labs' parent company to commercialize the transistor. Seven decades later, the pattern is repeating in mRNA medicine.

David Weinberg and Chris Carlson helped build Orbital Therapeutics' RNA platform — the same platform Bristol Myers Squibb acquired for $1.5 billion last October. ParcelBio, founded in 2023 and formally unveiled this week, is the company they started before the acquisition ever happened. That timing is the whole point: their bet on nanoparticle-free mRNA delivery predates the validation event, which means it was a technical conviction, not a follow-on bet.

Their approach is called APEXm. Rather than encapsulating mRNA in lipid nanoparticles — the same delivery vehicle used in Moderna's COVID vaccines and in most in vivo CAR-T programs — ParcelBio uses synthetic oligonucleotides engineered to recruit the cell's own protein stabilization machinery. The ParcelOligos carry modified sequence elements that bind RNA-binding proteins already present in the cell, which protect the mRNA transcript from degradation and physically shuttle it to the ribosome for sustained translation. LNPs solve the durability problem chemically, by wrapping mRNA in a lipid shell. APEXm attempts to solve it biologically, by making the mRNA a better substrate for the cell's native systems.

The pitch is that LNPs, optimized over a decade for vaccine-scale dosing, are the wrong tool for therapeutic protein production. Getting enough CAR-encoding mRNA into enough T cells inside a living body, with enough expression to treat cancer or tolerize autoreactive immune cells, is a different engineering problem. ParcelBio's answer is to sidestep the nanoparticle entirely.

$13.2 million is not a lot of money for a biotech, even a seed round. But the check writers — Breyer Capital, Y Combinator, General Catalyst — are not emerging managers hoping for a lottery ticket. Their backing, at this stage, is a signal that the technical differentiation is real enough to justify a platform bet.

There is an open question about intellectual property. Weinberg and Carlson were inside Orbital's RNA platform development before leaving in 2023. Whether ParcelBio's oligonucleotide approach steps on BMS/Orbital patents depends on how narrowly those patents are drafted — particularly around the specific sequence elements and modifications ParcelBio uses. ParcelBio has not disclosed a freedom-to-operate analysis, and neither BMS nor Orbital has commented on the competitive overlap. For a six-person team betting $13.2 million on a novel platform, an IP challenge would be existential.

The lead programs are in vivo CAR-T for autoimmune disease — specifically B cell depletion — and oncology. In both cases, the target is a cell type that current LNP delivery systems struggle to reach efficiently. Orbital, which BMS acquired, made headlines last July disclosing preclinical data for a circular RNA CAR-T targeting autoimmune disease, betting that circRNA's longer persistence would give it an edge. ParcelBio is taking a structurally different bet: no nanoparticle, no liver accumulation problem, expression driven by sequence engineering rather than delivery vehicle.

There is no shortage of competition. Around 250 bispecific antibody programs are in active development by 180 companies, many targeting the same B cell depletion biology. In vivo CAR-T specifically includes Carisma Therapeutics, Capstan Therapeutics, and Verismo Therapeutics — all of which are building on LNP delivery. What ParcelBio has is a platform claim: if APEXm's mechanism of recruiting native RNA-binding proteins proves superior to lipid encapsulation for therapeutic protein production, the LNP-centric programs currently in their pipelines face becoming outdated infrastructure.

The regulatory path is genuinely unmapped. No oligonucleotide-mRNA hybrid delivery system has FDA precedent. The agency's closest reference points are the siRNA therapeutics that have reached approval — Patisiran for hereditary transthyretin amyloidosis, for instance — but those are designed to silence genes, not encode therapeutic proteins. An oligonucleotide that functions as both delivery vehicle and mRNA sequence element, expressing a CAR receptor in vivo, would require the agency to evaluate something genuinely new. For a six-person team with $13.2 million, that is either an insurmountable moat or a significant liability — and the answer depends on how the agency responds to a novel molecular entity with a plausible safety profile.

Carlson holds a PhD from UCSF where he studied RNA packaging and protection in viruses, then contributed to multiple RNA stabilization patents at Orbital. Weinberg previously led RNA platform development at Orbital and did molecular research at Freenome. Their advisors include Robbie Majzner, a Stanford CAR-T researcher, and Rachel Green, the Johns Hopkins microbiologist known for CRISPR and phage therapy work.

Weinberg is scheduled to present at the 6th mRNA-Based Therapeutics Summit July 21–24 — alongside Moderna, Pfizer, AstraZeneca, and Sanofi. That stage is where the technical details of APEXm will either hold up under expert scrutiny or get picked apart. For now, the story is the bet: two scientists who watched their former employer get bought for $1.5 billion, and decided the real opportunity was something the buyer hadn't thought to build.