Turning the Immune System’s Brake Into Medicine: RegCell’s Road From Nobel Science to Autoimmune Cell Therapy

The liver can be injured in silence. In autoimmune hepatitis, an immune system built to eliminate infection turns on the body’s own liver cells. Some patients arrive with fatigue or jaundice, others only with abnormal blood tests. Untreated inflammation can progress to fibrosis, cirrhosis and liver failure.

Today’s treatment weakens the attack with corticosteroids and immunosuppressants. It saves lives, but patients may manage infection, bone and metabolic side effects and years of therapy. Most relapse after treatment is stopped. The unmet need is not simply a stronger way to disable immunity. It is a way to reteach the immune system what it must not attack.

RegCell, born from research in Kyoto and Osaka and now headquartered in Emeryville, California, wants to conduct that lesson with living cells. It proposes to collect disease-relevant CD4-positive T cells from a patient and reverse their role while preserving what they recognize. The sword is not discarded. It returns to the same place carrying a shield.

What RegCell is trying to make

The company’s core product concept is a stable and functional induced regulatory T cell, abbreviated S/F-iTreg. Conventional CD4-positive cells from the patient—ideally cells that recognize the drivers of the disease—are processed outside the body until they express FOXP3 and acquire the regulatory program and epigenetic state associated with a stable Treg identity.

The objective is antigen-specific immune tolerance. Rather than broadly dampening immune responses, the infused cells should become active where a disease antigen is presented and suppress T- and B-cell responses to that target. In principle, defense against infection and vaccine responses could remain intact. Showing that selectivity in patients will be the platform’s most important test.

An immune system is not born with a perfect enemy list

T cells use T-cell receptors, or TCRs, to recognize small peptide fragments displayed by antigen-presenting cells. The enormous diversity of those receptors prepares us for pathogens evolution has not seen. Random diversity also generates receptors that can recognize the self.

The thymus eliminates many dangerous cells, a process called central tolerance. It cannot catch every one. A second system must restrain self-reactive cells that reach the body’s periphery and end inflammation after an infection. Regulatory T cells are central to that peripheral tolerance. Immunity has an accelerator and a brake that knows where to press.

The suppressor T-cell idea was once abandoned

During the 1970s, researchers debated “suppressor T cells” that might restrain immune reactions. No unique marker identified the proposed cells, and weakly reproducible experiments and incorrect molecular explanations damaged the field. By the 1980s, many immunologists treated the entire idea as discredited and emphasized deletion of self-reactive cells in the thymus.

That history made Shimon Sakaguchi’s work unusually difficult. Arguing for a suppressive cell sounded like reviving a failed theory. The field required more than a concept. It needed surface markers that other laboratories could use and experiments in which removing a defined population caused autoimmunity and replacing it prevented disease.

1979: the paradox of the three-day-old mouse

At the Aichi Cancer Center Research Institute, Sakaguchi studied an experiment in which the thymus was removed from newborn mice. The expected result was immune deficiency. But when surgery occurred three days after birth, immunity ran out of control and attacked organs including ovaries, thyroid and stomach.

Transferring a subset of CD4-positive T cells from healthy mice prevented the disease. The immune system appeared to contain cells that actively restrained other cells, not merely a thymus that deleted dangerous clones. Sakaguchi began pursuing those cells in 1979 and spent more than a decade confronting technical ambiguity and professional skepticism.

1995: CD4 and CD25 reveal the security guards

In 1995 Sakaguchi and colleagues showed that removing T cells carrying both CD4 and CD25 caused autoimmunity, while restoring the population prevented it. The result made it possible for other researchers to isolate and test the same cells. Regulatory T cells returned to immunology as a defined, functional population.

Doubt did not vanish at publication. Replication by American immunologist Ethan Shevach and others helped change opinion. Science rarely turns because an authority declares victory. It turns when a claim becomes an operation that skeptics can repeat.

2001–2003: FOXP3 joins two stories

Mary Brunkow and Fred Ramsdell showed in 2001 that mutations in Foxp3 caused severe autoimmunity in scurfy mice and in boys with the human IPEX syndrome. In 2003 Sakaguchi and other researchers tied FOXP3 to the development and function of the Tregs he had identified. The 2025 Nobel Prize in Physiology or Medicine honored all three for discoveries concerning peripheral immune tolerance.

FOXP3 is more than a name badge. Yet transient FOXP3 expression alone does not necessarily make a durable regulatory cell. A cell exposed to inflammation needs a deeper regulatory memory—an epigenetic state that can survive division. That insight leads directly to RegCell’s platform.

The same cell can help autoimmunity and protect a tumor

In autoimmune disease, allergy, transplant rejection and graft-versus-host disease, researchers generally want more Treg activity. Tumors can exploit the same biology by recruiting Tregs that shield cancer from immune attack. Cancer treatment may therefore seek to remove the brake.

RegCell’s early history reflects both sides. In 2018 AMED selected a company project intended to weaken Tregs and combine that effect with molecular inhibition against solid tumors. The present company is focused on tolerance, but the founding science generated therapeutic strategies pointing in opposite directions.

2016: knowledge leaves the laboratory and becomes a company

RegCell Co., Ltd. was established in Kyoto on January 27, 2016 to translate Sakaguchi’s work. The early company included regulatory-cell medicine and a regenerative killer-T-cell effort for cancer. In 2019 the latter was separated into Rebirthel, sharpening RegCell’s focus on Treg biology.

A startup is not merely another paper. It must unite patents, financing, manufacturing, quality, toxicology, regulation, reimbursement and supply. Sakaguchi established that Tregs exist and matter. The company must turn variable living cells from individual patients into a product with a defined identity, dose and release specification.

Why autoimmune medicine is difficult to move beyond broad suppression

Corticosteroids, antimetabolites, biologic antibodies and JAK inhibitors have transformed many autoimmune diseases. Broad pathways are useful when no single cause can be named. The same breadth can affect infection control, protective antibodies, vaccine responses, bone, metabolism and reproduction, often requiring long treatment and management of relapse.

Antigen-specific therapy promises to repair a faulty circuit without darkening the whole immune system. The difficulty is that circuits vary among people and across stages of disease. Immune responses may spread from one epitope to others. Precision becomes a weakness if the chosen target is incomplete.

The first stated indication is autoimmune hepatitis

RegCell’s February 2026 manufacturing announcement and June BIO International Convention profile identify autoimmune hepatitis as the first indication. The cause of this immune-mediated inflammatory liver disease remains incompletely understood. It can affect any age, sex or ethnicity and range from silent disease to acute liver failure.

Standard immunosuppression is lifesaving, but treatment may be prolonged and most patients relapse when medication is withdrawn. A disease with multiple or uncertain antigens is a meaningful test of RegCell’s claim that the patient’s own TCR repertoire can capture relevant targets. It also demands rigorous endpoints: ALT, AST, IgG and steroid reduction are useful, but histology, relapse and long-term safety matter too.

The central idea: keep the target memory of the harmful cell

CAR-Tregs receive a synthetic receptor through genetic engineering and are directed toward a known target. RegCell takes another route. It retains the native TCR of a disease-driving effector or memory CD4-positive cell while changing the cell’s role from attack to suppression. Target recognition remains; cellular occupation changes.

If successful, one manufacturing backbone could address diseases with multiple or unidentified antigens. But selecting the disease-driving cells—where to collect them, by which markers and at what purity—is fundamental. The advertised ability to work without knowing every antigen creates a corresponding burden: prove that the starting population contained the right immune memories.

FOXP3 is not enough: teaching a cell to remember its role

Work from Sakaguchi’s laboratory in 2012 showed that FOXP3 expression and Treg-specific DNA hypomethylation are independent and complementary parts of regulatory-cell development. FOXP3 activates the program; an epigenetic state at regions including TSDR/CNS2 helps lock identity through cell division.

An induced Treg that reverts to an inflammatory phenotype could fail or theoretically aggravate disease. The “S/F” in S/F-iTreg stands for stable and functional. The name states the product ambition. How stable the cells remain inside a human inflammatory environment, where they travel and how long they persist cannot be established outside clinical trials.

The published process: changing cell fate with pharmacology and culture

A 2025 Science Translational Medicine paper stimulated conventional T cells with antigen and IL-2, used CDK8/19 inhibition to induce high FOXP3 expression and withheld CD28 costimulation to promote Treg-specific epigenetic changes. Repeating the cycle with periods of IL-2-only rest converted naïve and effector/memory CD4 cells—including Th1, Th2 and Th17 cells—into S/F-iTregs.

This describes the published scientific platform, not necessarily every detail of the final GMP process for RC-101. A commercial process must specify starting cells, media, residual reagents, culture duration, vessels, freezing, shipment and release tests. Reproducing a paper and releasing a patient batch are separated by a large body of engineering.

The first 2025 paper: inflammatory cells become stable Tregs

Norihisa Mikami, Sakaguchi and colleagues generated induced cells with transcriptional and epigenetic characteristics resembling natural Tregs. The cells remained phenotypically and functionally stable in mice and suppressed models of inflammatory bowel disease and graft-versus-host disease.

The important inversion was that already differentiated effector and memory cells could be converted. The cells most informed about disease antigens might become the therapeutic material. But mouse inflammatory disease is not human autoimmune hepatitis. Dose, persistence, patient variability, concomitant medication and manufacturing failure remain unanswered.

The second 2025 paper: testing selectivity in pemphigus

A Keio University, RIKEN, University of Osaka and RegCell collaboration used pemphigus vulgaris, a disease with a well-defined autoantigen. Pathogenic CD4 cells that recognized the skin adhesion protein desmoglein 3, or Dsg3, were converted into S/F-iTregs recognizing the same target.

In mice, the cells expanded in skin-draining lymph nodes, selectively suppressed Dsg3-specific helper T and B cells and harmful antibodies, and reduced disease without depleting the total B-cell population. Researchers also generated non-gene-edited S/F-iTregs from patient blood and showed suppression in vitro. It is strong mechanistic support, not a result from treating patients.

The distance between “made from human cells” and “worked in humans”

Generating the cells from people with disease shows that the process may tolerate compromised starting material. It does not prove safe manufacturing, migration to the target organ, persistence, prevention of relapse or clinical benefit. Human autoimmunity is more heterogeneous and slower than most mouse models.

A preclinical package must connect pharmacology, biodistribution, toxicity, tumor risk, unintended suppression and manufacturing consistency. Living cells are difficult to titrate after infusion. Persistence can be the source of durable benefit and the source of durable harm.

The advantage of no genetic modification—and a different complexity

RegCell emphasizes that its approach uses neither gene editing nor viral DNA reagents. It may avoid designing a vector for every target, reduce vector-related testing and insertional concerns, and permit a common manufacturing platform across diseases. Those features could support automation and lower cost.

Non-gene-edited does not mean simple or risk-free. Developers must control pharmacologic induction, residual CDK8/19 inhibitor or other materials, lot-to-lot variation in epigenetic state, contaminating cells and long-term lineage stability. Changing cell fate without changing DNA still requires demanding quality assurance.

“High purity” is harder to define than cell count

A Treg product cannot be defined by FOXP3 positivity alone. Activated nonregulatory human cells can express FOXP3 transiently, and FOXP3-positive human cells contain functionally different populations. Identity may require a combination of surface markers, transcription, TSDR/CNS2 methylation, suppression assays, inflammatory cytokines and TCR repertoire.

Release tests must be fast enough to deliver a product, while durable stability cannot be observed in a short test. Surrogate assays must eventually be shown to predict patient outcomes. Connecting identity, purity, potency, safety and dose is the gate between RegCell’s science and a medicine.

RC-101: the schedule reveals development reality

A March 2025 company announcement said RegCell planned to initiate a first-in-human trial that year. At BIO International Convention on June 24, 2026, however, the company profile still classified lead product RC-101 as preclinical and identified early proof of concept from a first-in-human study in the first half of 2028 as the next value-inflection update.

That does not establish that the program failed. Process transfer, analytical methods, GMP materials, stability, toxicology and regulatory consultation routinely move cell-therapy timelines. It does establish that old targets should not be mistaken for current status. A Nobel Prize does not accelerate the clinical clock, and financing does not eliminate biological or manufacturing uncertainty.

$45.8 million: dividing risk between investors and the public

RegCell announced an $8.5 million seed financing in 2025 co-led by the University of Tokyo Edge Capital Partners and Fast Track Initiative. Celadon Partners, Mitsubishi UFJ Capital, Osaka University Venture Capital and Kyoto University Innovation Capital participated.

The company also announced up to ¥5.6 billion—then expressed as $37.3 million—in nondilutive AMED support. AMED’s official project title is development of antigen-specific immune cell therapy for autoimmune diseases and related conditions, with UTEC as the certified venture-capital partner. A maximum award is not the same as cash already received or corporate valuation; it is conditional public development support tied to eligible work and progress.

What AMED is buying: the bridge, not another discovery

AMED’s pharmaceutical-startup ecosystem program joins investment by certified venture firms with public support across preclinical and clinical development. According to RegCell, the funding supports IND-enabling work, GMP optimization and clinical proof of concept in the United States.

When taxpayers absorb substantial risk, public return also matters. Where will intellectual property and jobs reside? Will patients have access? Will negative findings be shared? Success should be measured not only by opening a U.S. trial but by whether manufacturing and regulatory knowledge remain available to Japan’s next company.

Why the headquarters moved to California

In 2025 RegCell reorganized around a U.S.-based corporation headquartered in Emeryville. It cited access to international talent, capital, partners, regulatory development and commercialization expertise, while saying core scientific work would continue in Japan.

The move displays Japanese strength and weakness together. The discovery, university science, public support and venture investors came from Japan. When the work reached clinical and commercial scale, the company moved closer to the U.S. ecosystem. Complaining about brain drain is insufficient. Without depth in CMC, trial operations, regulatory strategy and biotech management, strong Japanese science will keep making the same journey.

Kincell Bio: manufacturing becomes the second invention

In February 2026 U.S. contract developer and manufacturer Kincell Bio announced that it would perform process and analytical transfer, scale optimization and GMP clinical-material supply for RegCell’s lead Treg program. A laboratory procedure must become a closed, documented and repeatable process before it can enter a trial.

In autologous cell therapy, every patient is a batch. Collection, shipping, selection, culture, testing, freezing or fresh supply, return and infusion form one chain. A failure anywhere can deny treatment. Starting-cell quality varies with age, disease and prior drugs. Manufacturing is not back-office support for the science; it is half the product.

Automation and point-of-care production remain hypotheses

RegCell and Kincell say a nonviral, non-gene-edited process may be suitable for automation, scale and eventual point-of-care access. Avoiding vector production could reduce time and cost. Reusing one CMC backbone across indications could be commercially important.

Hospital production creates facility-to-facility variation, training, equipment, audit and responsibility problems. Central manufacturing offers control but introduces shipping time and geographic access. Relative cost cannot be known until culture duration, failure rate, volume and cryopreservation performance are public.

Safety: how to find a brake pressed in the wrong place

Perfectly localized Tregs could preserve protection against infection. Impure cells, cross-reactive receptors or migration beyond the target tissue might suppress useful immunity. Chronic infection and tumor surveillance require long observation.

The opposite risk is loss of regulatory identity and return to an inflammatory phenotype. Early trials should monitor more than infusion reactions: infections, malignancies, disease worsening, autoantibodies, cytokines, TCR lineages, persistence and off-target immune suppression. A cell designed to remain can create benefits and risks that outlast dosing.

Antigen specificity: the advantage that is hardest to prove

With a target such as Dsg3 in pemphigus, investigators can test whether cells react to that antigen and preserve other immune responses. In autoimmune hepatitis, where targets may be multiple or unknown, retaining native TCRs is attractive—but it is difficult to measure which disease-relevant specificities entered the final batch.

If patients improve, the trial must distinguish antigen-specific tolerance from broader suppression. That requires tracking disease clones, vaccine and infectious-antigen responses, immunoglobulins and cells inside affected tissue. Precision medicine is not a label. It is an experiment that measures selectivity.

The competitors are not only other Treg companies

For patients, the competition is not which company arrives first. It is which approach delivers more benefit than existing therapy with acceptable risk, durable control and a lower total burden. An expensive one-time therapy can create value if it prevents years of medication, hospitalization, relapse and organ damage. If durability is short, individualized manufacturing becomes harder to justify.

CAR-T’s immune reset and Treg reeducation

Trials in autoimmune disease are using CD19 CAR-T to eliminate much of the antibody-producing B-cell lineage and allow immunity to rebuild. Early reports include drug-free remissions in refractory patients, but many regimens involve lymphodepletion, infection risk and cytokine-release syndrome.

RegCell’s premise is not destruction and reboot but redirection: turn cells that recognize disease into agents of tolerance. It may prove more selective and less toxic, but it has no human data yet. CAR-T and Treg approaches need not be universal rivals. Disease, severity, antigen biology and patient risk may determine where each belongs.

The 2026 Tregzi approval opens a door—but it is not the same product

On June 30, 2026, the FDA approved Orca Bio’s Tregzi. Used in matched-donor stem-cell transplantation for blood-cancer patients, the product precisely combines stem and progenitor cells, Tregs and conventional T cells to improve survival free of chronic graft-versus-host disease. It was evaluated in a randomized trial of 187 adults.

The decision is an important precedent for regulating, manufacturing and clinically testing a Treg-containing product. It does not validate RegCell. Tregzi is a donor-derived, multicomponent transplant product—not autologous cells reprogrammed for antigen-specific autoimmune tolerance. It weakens the general claim that Tregs cannot become products; it proves nothing about RC-101’s safety or efficacy.

Why the 2026 white paper named RegCell

Japan’s 2026 Science, Technology and Innovation White Paper follows Sakaguchi’s path from research begun in 1979 to discoveries in 1995 and 2003 and the 2025 Nobel Prize. It singles out RegCell as an example of moving knowledge toward social implementation and notes support through national research organizations and pharmaceutical-startup programs.

Inclusion is neither regulatory endorsement nor an independent review. RegCell embodies the policy story of science and business drawing closer. Yet basic science took decades, the company is ten years old, and clinical testing lies ahead. Proximity does not erase distance. The institutions that finance and manufacture the crossing are what the policy must ultimately be judged on.

What the first clinical trial should answer

A first-in-human study will primarily test safety and feasibility. Among patients whose cells are collected, how many yield an in-specification product? How long does production take, what dose can be delivered and do infusion reactions, infection or worsening liver injury occur? Where do the cells travel and how long do they remain?

Preliminary activity can be explored through ALT, AST, IgG, histology, autoantibodies, steroid reduction and time to relapse. Immunosuppressants cannot be withdrawn recklessly, so efficacy interpretation will be slow. Even a dramatic signal in a small uncontrolled cohort must later be separated from natural fluctuation, concomitant drugs and patient selection.

Twelve gates RegCell must pass

Seventy-seven years of immune tolerance

The next news to watch

First will be an IND submission or FDA clearance and a public trial registration defining patients, dose, conditioning, concomitant drugs and manufacturing sites. Second is successful process transfer and GMP production at Kincell. Third is the biological rationale for autoimmune hepatitis and the method used to collect or identify disease-driving cells.

Fourth is how closely RC-101 corresponds to the published S/F-iTreg method. Fifth is whether the expected 2028 signal shows mechanism-specific tolerance, not safety alone. Sixth is how AMED support, intellectual property, pricing and a path for Japanese patients develop.

The Nobel Prize illuminates the beginning; the clinic is another road

RegCell’s story is compelling because discovery and treatment are not joined by a straight line. A three-day-old mouse, a discredited theory, CD4 and CD25, FOXP3, DNA methylation, a university startup, a cancer-business split, a U.S. headquarters, public funding and a manufacturing partner: 46 years of basic research have transformed into another long development process.

The most elegant idea is to preserve the memory of the harmful cell and use it therapeutically. Keep the TCR that recognizes self, then change the attacker into a protector. Success would mark a transition from medicine that weakens immunity to medicine that teaches tolerance again.

Elegant mechanism is not patient benefit. RegCell must move from its preclinical position through manufacturability, safety, specificity, durability, comparative benefit and access—one falsifiable step at a time. The Nobel Prize validates the importance of the biological question, not the success of a product.

That is also the most honest version of the white paper’s journey from knowledge to value. Social implementation is not turning a discovery into advertising. It is carrying the discovery into a clinical test that is allowed to prove it wrong.