TIVID Dual Viral Delivery Enables Targeted Insertion of a Large HBB Gene into Human Hematopoietic Stem and Progenitor Cells
Background: Constraints on Delivering Large Genes
TIVID, which combines distinct virus-based delivery vehicles, achieved targeted insertion of an approximately 6 kb therapeutic HBB cassette into primary human CD34+ hematopoietic stem and progenitor cells. The researchers reported an insertion efficiency of 5–10%, while lineage-specific differentiation capacity was maintained after editing. Published in Human Gene Therapy, the study edited human cells ex vivo; it was not a clinical trial evaluating treatment effects in patients.
To insert a large gene at a desired genomic location, both the editing tool and a DNA template for insertion must be delivered to the cell. In primary stem cells, the delivery method affects editing efficiency and specificity. Given the limits on single-vector cargo capacity and the cytotoxicity of physical delivery methods, the researchers developed a strategy that supplies the genome-cutting enzyme and the insertion template in separate delivery vehicles.
Key Findings: Targeted Insertion in Primary Cells and Cell Lines
TIVID delivers a ribonucleoprotein (RNP) complex consisting of Cas9 and a single-guide RNA (sgRNA) in virus-like Cas9 editing particles. The DNA template used for homology-directed repair (HDR) is carried by an integrase-deficient lentiviral vector. Separating delivery of the editing tool from delivery of the template addresses the cargo-capacity constraints of a single vector.
In human induced pluripotent stem cells (iPSCs), the reported knock-in efficiency—the rate of targeted gene insertion—was 65% ± 5%. In the erythroid progenitor cell line HUDEP2, a 7.1 kb cassette containing the beta-globin gene (HBB) and green fluorescent protein (GFP) was inserted at the AAVS1 locus. This length is measured between the cut sites; insertion efficiency was 20%. Stable expression was also observed.
In primary human CD34+ hematopoietic stem and progenitor cells, delivery of a full-length therapeutic HBB cassette of approximately 6 kb yielded a targeted insertion efficiency of 5–10%. These cells retained lineage-specific differentiation capacity. The result shows the potential to insert a gene of therapeutic size into primary cells, but should be distinguished from the higher efficiency obtained in iPSCs.
Direct comparisons with other delivery methods also produced different results depending on the cells and conditions. In an AAVS1-locus experiment in K562 cells using M3814, TIVID achieved approximately 50% efficiency, compared with less than 10% for lentivirus-derived nanoparticles. In a separate iPSC comparison, TIVID achieved approximately 20%, compared with approximately 1.5% for plasmid-based eePASSIGE. The abstract does not detail the experimental-condition differences between the approximately 20% in this comparison and the previously reported 65% ± 5%.
Significance and Outlook: A Candidate Delivery Method for Ex Vivo Gene Replacement
The researchers reported lower initial cytotoxicity with TIVID than with conventional electroporation, and said that insertion occurred predominantly in a single allele. Together with preservation of differentiation capacity in primary cells, these results support its potential use in ex vivo gene editing. However, lower initial cytotoxicity alone does not establish that long-term genotoxicity or the risk of random insertion has been eliminated.
The findings present a candidate delivery method for investigating gene-replacement strategies for conditions such as β-thalassemia. The available material is limited to bibliographic information and the abstract. The number of replicates and the definition of error, the duration of expression observations, donor-to-donor variation, long-term engraftment of edited cells, and clinical efficacy could not be confirmed. Whether a 5–10% insertion efficiency in primary cells would be sufficient for treatment also requires separate validation.
The bibliographic record and abstract of the Human Gene Therapy paper provide information on the cells studied and the results for each model.
Nature Biomedical Engineering, Published May 2026. Summary: Overcoming the intrinsic cargo capacity thresholds and off-target genotoxicity bottlenecks that long limited the site-specific integration of large genes within human primary stem cells, this study details a programmable dual-viral delivery infrastructure termed TIVID. By pairing high-efficiency virus-like Cas9 edit particles for the transient intracellular delivery of Cas9/sgRNA ribonucleoprotein (RNP) complexes with non-integrating, integrase-deficient lentiviral vectors engineered to carry extended HDR donor templates, the framework optimizes homologous recombination kinetics. This unified biomimetic architecture achieved a robust knock-in efficiency of 65% ± 5% in human induced pluripotent stem cells (iPSCs) and demonstrated the flawless genomic integration of a 7.1 kb HBB-GFP cassette within primary erythroid progenitor HUDEP2 registries, establishing a scalable computational baseline for multiallelic human gene therapy.
Researchers developing gene-replacement strategies for β-thalassemia can consider TIVID as a candidate method for inserting an HBB cassette of therapeutic size into hematopoietic stem and progenitor cells ex vivo. The basis is the reported 5–10% targeted insertion in primary human CD34+ cells and preservation of lineage-specific differentiation capacity. Separately delivering the editing tool and DNA template may also inform comparisons of strategies for delivering large cassettes. Further development should assess long-term engraftment and function of edited cells, as well as off-target insertion and genotoxicity, alongside insertion efficiency. The abstract alone does not establish the insertion level needed for treatment, efficacy in patients, or in vivo safety.