Monkey retrovirus protein could make CAR-T cancer therapy cheaper to manufacture
Korean scientists have identified a novel envelope protein from simian retrovirus 2 that outperforms the industry-standard RD114 protein for manufacturing CAR-T and CAR-NK cell therapies. The discovery, published in Nature Communications, boosted viral titres, gene transduction and tumour control in mice, offering a route to more efficient and affordable cancer immunotherapy production.
Researchers at the Korea Research Institute of Chemical Technology (KRICT) have discovered a viral envelope protein that could ease one of the most persistent bottlenecks in cell and gene therapy manufacturing: the production of viral vectors used to engineer CAR immune cells.
The team, led by Dr Chi Hoon Park with first author Dr Moonjung Jeun, identified a novel envelope protein derived from simian retrovirus type 2 (SRV2) that demonstrated superior performance compared with RD114, the feline endogenous retrovirus-derived envelope that has served as the standard pseudotyping tool for CAR-T and CAR-NK manufacturing for years. The findings were published online in Nature Communications on 23 April 2026.
(from the left) Principal Researcher Chi Hoon Park (Corresponding Author), Postdoctoral Researcher Moonjung Jeun (First Author). © Korea Research Institute of Chemical Technology (KRICT)
A bottleneck in cell therapy manufacturing
CAR immune cell therapies, engineered from a patient’s own T cells or natural killer (NK) cells to recognise and destroy cancer cells, have transformed outcomes for some blood cancers. But their manufacture remains complex and expensive, largely because of the viral vectors required to deliver the therapeutic CAR gene into immune cells. Envelope glycoproteins on these vectors recognise receptors on the surface of immune cells and mediate gene delivery, making them a critical determinant of manufacturing efficiency.
RD114 has been the standard envelope for retroviral CAR-T and CAR-NK production, while VSV-G is more widely used for lentiviral systems. According to the authors, “no studies to date have reported pseudotyping retroviruses or lentiviruses with SRV glycoproteins” prior to this work.
Why SRV2 stands out
The KRICT team turned to SRV2, a type-D retrovirus originally isolated from a rhesus monkey mammary tumour, because both it and RD114 use the same cellular entry receptor, ASCT2, a neutral amino acid transporter abundantly expressed on T cells and NK cells. Sequence analysis showed the receptor-binding domain of SRV2 envelope protein shares meaningful homology with RD114 and another envelope protein, BaEV, hinting that it might support efficient gene delivery into immune cells in the same way. That prediction held up experimentally. SRV2-pseudotyped retrovirus (SRV2 RV) was produced without apparent toxicity to the HEK293T producer cells and achieved substantially higher physical titres than RD114 RV for both GFP-reporter and CAR-encoding constructs. When used to transduce primary human T, NK and B cells, SRV2 RV consistently outperformed RD114 RV on fluorescence intensity, vector copy number and effective multiplicity of infection.
Notably, the advantage was specific to SRV2 among the SRV family. Of eight known SRV serotypes, only SRV2 (with SRV8 showing modest activity) achieved efficient gene transduction into T and NK cells; the researchers describe this as demonstrating, “for the first time, that pseudotyping retroviral vectors with SRV2 represents a potent strategy for gene transduction into immune cells”. Interestingly, the advantage did not extend to lentiviral systems. Lentivirus pseudotyped with SRV2 envelope failed to mediate meaningful transduction into T cells, even after the team engineered several cytoplasmic tail variants intended to improve compatibility with lentiviral protease processing. The authors concluded that SRV2 is compatible with retroviral, but not lentiviral, vector platforms.
Stronger CAR-T and CAR-NK cells, better tumour control
Moving to CAR constructs, T cells transduced with SRV2 RV showed roughly 20–25% higher CAR expression than those made with conventional RD114 RV or VSV-G lentivirus, across CD19, FOLR1 and PD-L1 CAR targets and across three independent blood donors. These SRV2-derived CAR-T cells also killed antigen-positive tumour cells more effectively in culture, while cytokine secretion profiles were broadly comparable to RD114-derived cells, suggesting the functional gain reflects genuinely higher transduction rather than an artefact of altered cytokine output. Similar improvements of around 5-12% in CAR-positive cell yield were seen in NK cells. The clearest evidence came from mouse models. Untreated mice bearing NALM-6 leukaemia xenografts developed tumours within roughly ten days and had all died by day 46. Mice treated with SRV2-based CD19 CAR-T cells showed markedly delayed tumour growth, with three of four animals remaining tumour-free for the full study period, compared with two of four in the RD114-treated group. “This study is significant because we identified a new candidate envelope protein that outperforms RD114, the gene-delivery key that has been widely used worldwide,” said Dr Chi Hoon Park, the study’s corresponding author. KRICT President Seok Min Shin added: “Although CAR therapies offer remarkable anticancer efficacy, their high manufacturing costs remain a major challenge. We hope this new vector platform will contribute to both improved manufacturing yields and enhanced therapeutic outcomes.”
Caveats and next steps
The authors are candid about the study’s limitations. They note that “the present study did not assess the integration-site preference or potential genotoxic risk of SRV2 RV,” an important safety consideration for any new retroviral vector heading towards clinical use, and that the in vivo cohorts were small, meaning larger, adequately powered studies are needed before firmer conclusions on therapeutic benefit can be drawn. Nonetheless, the team has already optimised plasmid ratios and production protocols for SRV2 RV manufacture and plans further work towards large-scale production and eventual commercialisation, positioning SRV2 as a candidate platform for reducing the manufacturing costs that currently limit access to CAR-T and CAR-NK therapies.
Journal reference
Jeun, M., Kim, Y., Lee, H. K., et al. (2026). Discovery of a novel envelope protein derived from simian retrovirus 2 for pseudotyping retroviral vectors used for production of CAR immune cells. Nature Communications, 17, 5643. https://doi.org/10.1038/s41467-026-72024-4
