Single Gene Knockout Strips Lab-Grown Heart Cells of Immune Visibility

Scientists in Japan have created lab-grown heart cells that largely hide from the immune system by switching off a single gene, a result that could bring the field closer to off-the-shelf cardiac cell therapy for heart failure. In a study published in iScience, researchers led by Shugo Tohyama of Fujita Health University used the CRISPR-Cas9 gene-editing tool to disrupt TAPBP, a gene that encodes tapasin, a molecular chaperone essential for displaying HLA class I proteins on the cell surface. The resulting cardiomyocytes, grown from human induced pluripotent stem cells, showed dramatically reduced HLA class I expression and evaded attack by cytotoxic T cells in laboratory tests, without provoking the natural killer cell response that has undermined previous attempts at hypoimmunogenic cell lines.

The clinical backdrop explains the urgency. Heart failure affects more than 64 million people worldwide, a figure that nearly doubled between 1990 and 2017 and is expected to keep climbing. For severe cases resistant to drugs and devices, heart transplantation remains the only radical treatment, yet donor hearts are scarce. In Japan, only about 100 heart transplants are performed each year, and patients wait on average more than five years. Transplanting cardiomyocytes derived from induced pluripotent stem cells, or hiPSCs, has long been proposed as a regenerative alternative, and preclinical studies in non-human primates as well as early clinical trials have reported encouraging outcomes.

But immunological rejection remains the central obstacle. Autologous approaches, in which a patient’s own cells are reprogrammed and redifferentiated, are limited by long manufacturing lead times and the demanding quality control required under good manufacturing practice. Allogeneic approaches, using validated donor cell lines produced at scale, promise off-the-shelf delivery but carry the risk of graft rejection. HLA matching, the strategy that underpins organ transplantation, only partially helps because HLA genes are extraordinarily polymorphic; even a large homozygous hiPSC bank covers only a fraction of any population. Patients receiving mismatched cells must therefore take immunosuppressants, trading the risk of graft loss against the risk of infection.

The immunology of rejection centers on HLA class I molecules, which display peptide fragments on the surface of nearly every nucleated cell. Cytotoxic T lymphocytes constantly scan these displays, and mismatched HLA class I triggers rapid killing of the graft. Removing HLA class I entirely, for example by disrupting beta-2 microglobulin, does spare cells from T cell attack, but it creates a new problem: natural killer cells interpret the absence of HLA class I as a missing-self signal, a hallmark of virus-infected or malignant cells, and destroy them. Previous workarounds have required elaborate multi-gene editing, such as knocking out HLA-A and HLA-B while preserving HLA-C, or additionally overexpressing PD-L1, HLA-G, and CD47. Each added modification multiplies off-target risks and complicates quality control for clinical manufacturing.

The Japanese team took a different route by targeting the antigen-processing machinery rather than the HLA genes themselves. Tapasin, a 48-kilodalton transmembrane glycoprotein encoded by TAPBP, bridges HLA class I molecules to the TAP transporter inside the endoplasmic reticulum. It stabilizes unstable HLA class I complexes carrying low-affinity peptides and catalyzes their exchange for high-affinity ones, a keystone function in peptide loading. Mice lacking tapasin show reduced HLA class I expression and impaired cytotoxic T cell responses. The researchers hypothesized that disrupting TAPBP in human stem cells would lower HLA class I enough to escape T cell detection while leaving enough inhibitory signal, or enough other protective features, to avoid provoking natural killer cells.

To test this, they electroporated a plasmid encoding TAPBP guide RNA and Cas9 into an HLA haplotype-homozygous hiPSC line provided by Kyoto University’s Center for iPS Cell Research and Application. Among 56 colonies screened by flow cytometry, they selected the clone with the lowest HLA class I expression and confirmed loss of tapasin protein by western blotting. Sanger sequencing revealed a single nucleotide insertion in the target sequence. The edited cells retained their stemness, expressing the pluripotency markers NANOG and OCT4, showed a normal karyotype, and expanded at rates equivalent to unedited controls. When stimulated with interferon-gamma, a inflammatory signal known to upregulate HLA class I, the knockout cells still responded in a dose-dependent manner, but their expression plateaued at a lower level than controls, and HLA class II remained low throughout.

A critical question was whether the gene edit would compromise the cells’ ability to become functional heart muscle. Using a differentiation protocol involving WNT modulation followed by metabolic selection in glucose- and glutamine-depleted, lactate-supplemented medium, the team generated cardiomyocytes from both control and TAPBP-knockout lines. Purity, measured by cardiac troponin T positivity, exceeded 95 percent by day 17 in both groups. Immunostaining showed comparable sarcomere organization and alignment, quantitative PCR revealed similar expression ratios of ventricular and atrial isoforms of myosin light chain, troponin, and myosin heavy chain genes, and mitochondrial distribution patterns were indistinguishable. Calcium transient recordings captured with a functional drug screening system showed no differences in beat rate, contraction amplitude, or relaxation kinetics, and both cell types increased their beating rate in a dose-dependent response to the beta-adrenergic agonist isoproterenol.

The immunological tests then delivered the study’s central findings. Cytotoxic T lymphocytes were expanded from three healthy donors whose HLA haplotypes mismatched the cell line. In ELISpot assays measuring interferon-gamma production, T cells exposed to knockout cardiomyocytes produced far fewer spots than those exposed to controls, and the gap persisted even after the target cells were pre-treated with interferon-gamma to maximize HLA expression. Degranulation assays, which detect surface CD107a as a marker of perforin and granzyme release, told the same story: with interferon-gamma pre-treatment, the percentage of degranulating T cells fell from 9.24 percent to 1.92 percent against cells from donor A, from 29.6 to 15.5 percent for donor B, and from 27.2 to 10.1 percent for donor C. In calcein-based killing assays, cytotoxicity against knockout cells was significantly reduced for two of the three donors without pre-treatment and remained significantly lower after pre-treatment for two donors. Co-culture with T cells also raised the beating rate of control cardiomyocytes more than that of knockout cells, suggesting the edited cells experienced less immune stress.

The natural killer cell results were equally important. Because reduced HLA class I can trigger missing-self killing, the team co-cultured natural killer cells from three donors with both cell types and measured interferon-gamma secretion, CD107a degranulation, and cytotoxicity, using the K562 leukemia line as a positive control. Across all assays, responses to control and knockout cardiomyocytes were equivalent and minimal, whether or not the cells had been treated with interferon-gamma. The explanation emerged when the researchers examined activating ligands: unlike K562 cells, both control and knockout cardiomyocytes lacked expression of ULBP and MICA/MICB, the stress ligands that natural killer cells rely on. With no activating signal to outweigh the lost inhibitory signal, the edited cells escaped innate attack despite their near-invisible HLA profile.

The authors are careful about limitations. Their experiments tested immune responses under interferon-gamma-induced inflammation in vitro, whereas the failing heart in patients contains a far more complex mix of cytokines and immune mediators. No in vivo therapeutic effect was evaluated, partly because species-mediated rejection complicates animal models, and humanized mouse platforms still incompletely reconstitute human immune function. Even so, the appeal of the strategy lies in its simplicity: a single gene disruption, rather than a cascade of edits, yields cardiomyocytes that dodge the two major arms of cellular rejection. If future work in more physiologically relevant models confirms these findings, TAPBP-knockout hiPSC-derived cardiomyocytes could become a universal donor cell source, easing the donor shortage that constrains heart transplantation today.

Subject of Research: Generation of hypoimmunogenic cardiomyocytes from human induced pluripotent stem cells via TAPBP knockout

Article Title: Generation of hypoimmunogenic cardiomyocytes from human induced pluripotent stem cells with TAPBP knockout

Article References: Yamamoto, Y., Ohno, R., Shoda, K., Suzuki, T., Momoi, K., Fujii, S., Kenmochi, Y., Yoshikawa, T., Shimomura, M., Tsukamoto, N., Ohnuki, K., Tani, H., Soma, Y., Fukuda, K., Nakatsura, T., & Tohyama, S. (2026). Generation of hypoimmunogenic cardiomyocytes from human induced pluripotent stem cells with TAPBP knockout. iScience, 29(11), Article 117933. https://doi.org/10.1016/j.isci.2026.117933

Image Credits: AI Generated

DOI: 10.1016/j.isci.2026.117933

Keywords: hiPSC, cardiomyocytes, TAPBP, tapasin, HLA class I, CRISPR-Cas9, hypoimmunogenic, heart failure, cytotoxic T lymphocytes, natural killer cells, cell therapy, transplantation