GentiBio.

A patient just got dosed. That is the whole point.

In a clinic somewhere in 2026, a person recently diagnosed with type 1 diabetes received a single infusion of their own cells - re-engineered in a lab to do one specific job. The trial is called POLARIS. The therapy is GNTI-122. And for GentiBio, a clinical-stage biotech in Cambridge, Massachusetts, that infusion is the moment a six-year bet stopped being a slide deck and became medicine flowing into a vein.

GentiBio is roughly 45 people working on one stubborn idea: that the immune system can be taught to stand down. Not suppressed with a blunt instrument, not bypassed - taught. The company builds engineered regulatory T cells, or "EngTregs," and aims them at diseases where the body has turned on itself. It has raised about $177 million, signed a partnership with Bristol Myers Squibb worth up to $1.9 billion in potential milestones, and collected an FDA Fast Track designation along the way. For a company most people have never heard of, that is a surprising amount of weather.

Your immune system is brilliant. Until it isn't.

Autoimmune disease is, at heart, an act of mistaken identity. The same immune system that fights off infections decides that the pancreas, or the gut lining, or some other piece of you, is the enemy. In type 1 diabetes, immune cells destroy the insulin-producing cells in the pancreas. The standard answer for a century has been to replace what was lost - inject insulin, forever - rather than to stop the destruction.

The body actually ships with a fix. Regulatory T cells, or Tregs, are the cells whose entire job is to keep the rest of the immune system in line. They are the brakes. The trouble, as GentiBio's CEO has put it, is that Tregs are "rare and heterogeneous" - too scarce to harvest in useful numbers, too inconsistent to behave like a reliable drug. You cannot bottle a brake pedal you can barely find.

So the field had a tantalizing target and no practical way to use it. That gap - between the brake that biology already invented and a brake you could actually manufacture - is the entire reason GentiBio exists.

Don't hunt for Tregs. Build them.

The wager that started GentiBio was deceptively simple: if natural Tregs are too rare and too unreliable, stop hunting for them and engineer them instead. Take abundant, ordinary cells and edit them - including the FOXP3 gene, the master switch that makes a Treg a Treg - until they behave the way you want, at the scale you need. The science came out of a 2020 paper in Science Translational Medicine describing exactly that gene-editing strategy.

The team assembled to chase it spans two continents. Adel Nada, the co-founder and CEO, came from senior roles at Intellia Therapeutics and Casebia - a gene-editing pedigree. The scientific co-founders are a who's who of Treg biology: David Rawlings of Seattle Children's Research Institute, Jane Buckner of the Benaroya Research Institute, and Gideon Gross of the MIGAL Galilee Research Institute in Israel. Rawlings and Buckner led the research team behind that founding paper.

The investors agreed the bet was worth funding. GentiBio launched in 2020 with $20 million in seed money from OrbiMed, the Novartis Venture Fund, and RA Capital. A year later it raised a $157 million Series A led by Matrix Capital Management - which the company called the largest Series A in the Treg therapeutic space to date. The smart immunology money, it turned out, liked the idea of building brakes.

A guard posted at the scene of the crime

The platform is modular by design: take cells from abundant autologous (the patient's own) or allogeneic (donor) sources, engineer in the right genes and the right targeting, and produce Tregs that are stable, disease-specific, and tunable. The promise is to fix the two things that made natural Tregs unusable - their scarcity and their inconsistency - in one manufacturing process.

GNTI-122, the lead program, shows what that looks like in practice. It is a single-dose, autologous, antigen-specific engineered Treg therapy for type 1 diabetes. The cells are designed to migrate to the pancreas and stay there, quietly suppressing the local immune attack so the body can keep making its own insulin. The goal is not to manage the disease. It is to reduce or eliminate the need for lifelong insulin injections in people caught early.

The money, the milestones, the regulator

Conviction is cheap; validation is not. GentiBio's comes in three currencies. First, capital: $177 million from investors who specialize in seeing biology early, including OrbiMed, Novartis Venture Fund, RA Capital, Matrix Capital, and Avidity Partners. Second, a strategic partner: Bristol Myers Squibb staked an undisclosed upfront payment plus up to $1.9 billion in potential milestones on GentiBio's Tregs for inflammatory bowel disease - the kind of check large pharma does not write casually.

Third, and most important, the regulator and the clinic. The FDA accepted GNTI-122's IND in 2025, then awarded it Fast Track Designation - a status reserved for therapies addressing serious unmet needs. And in 2026, POLARIS dosed its first participant. Each step is the kind of thing that cannot be bought, only earned.

Treat the cause, not the symptom

GentiBio's stated aim is to "functionally cure" autoimmune, alloimmune, autoinflammatory, and allergic diseases - a deliberately bigger word than "manage." The company's own phrasing is plainer: it wants to help patients live their lives, better. That is a low-key way to describe trying to end a century of daily insulin injections.

The ambition is not modest, even if the company is. Where conventional treatment fights an overactive immune system with broad suppression - turning down the volume on everything and accepting the infection risk that comes with it - GentiBio is trying to fix the specific wiring fault. A Treg aimed at one disease leaves the rest of the immune system intact. The brakes work without unplugging the engine.

If the brakes hold

None of this is settled. A Phase 1 trial proves safety and gathers early signals, not cures; engineered cell therapies are hard to manufacture and harder to make affordable; and the 2023 layoffs were a reminder that even good science has to survive a bad market. The honest version of GentiBio's story is that the company has done everything right up to the part that actually matters - whether the cells work in people. That answer is being written now, one POLARIS participant at a time.

But the size of the prize is what keeps the smart money and a major pharma at the table. If engineered Tregs can be aimed at one disease, they can in principle be aimed at many - the same platform, retargeted. Type 1 diabetes is the proof of concept. Inflammatory bowel disease is next in line through Bristol Myers Squibb. Behind them sits a list of conditions where the immune system has simply made a mistake and never corrected it.

Return to that clinic in 2026. A single infusion of re-engineered cells, designed to travel to the pancreas and stand guard. If it holds, the patient who received it may need fewer injections, or none. That is the whole bet, reduced to one quiet moment: not a cure announced from a podium, but a brake pedal, built from scratch, finally pressed.