Skip to main content
NCBI home page
Scientific Reports logoLink to Scientific Reports
. 2026 Mar 17;16:9566. doi: 10.1038/s41598-025-17342-1

A targeting lentiviral vector for generation of CAR-T cells in vivo

Muhadasi Tuerxunyiming 1,2,#, Jianguo Michael Yin 3,#, Ping Zhu 4, Maoxuan Liu 5, Zheng Fu 6, Qing Zhao 7,8,✉
PMCID: PMC13009191  PMID: 41844687

Abstract

Chimeric antigen receptor (CAR) T cell therapy has demonstrated remarkable therapeutic efficacy in treating cancer and autoimmune diseases. However, current CAR-T cell therapy requires ex vivo T cell engineering, which is both time-consuming and cost-prohibitive, adding complexity to the overall treatment. In this study, using an engineered Sindbis virus envelope, we developed a lentiviral vector system with high specificity for targeting human T cell line and primary T cells, but not targeting other immune cell subsets. Notably, this T cell-specific lentiviral vector does not require additional anti-CD3/CD28 stimulation for primary T cell activation during infection in vitro. Furthermore, the lentiviral vector successfully delivered a CD19-targeting CAR molecule to human primary T cells in vivo. The in vivo generated CD19-CAR-T cells efficiently mediated B cell lymphoma clearance. Overall, our study provides a promising tool for the development of in vivo T cell engineering approaches.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-025-17342-1.

Keywords: Lentiviral vector, Chimeric antigen receptor, In vivo, CAR-T

Subject terms: Genetic transduction, Genetic engineering, Molecular medicine, Haematological cancer, Tumour immunology

Introduction

Chimeric antigen receptor (CAR)-T cell therapy has demonstrated substantial therapeutic efficacy in treating B cell lymphoma and shows significant promise in addressing various autoimmune diseases1–4. Currently, six ex vivo autologous CAR-T cell products have been approved for clinical use in both the United States and China5,6. However, the ex vivo manufacturing process for CAR-T cells is time-consuming, typically requiring three to six weeks for production7,8. Moreover, the ex vivo production of CAR-T cells necessitates specialized manufacturing centers, involving complex equipment and consumables for T cell isolation, activation, expansion, and quality control, which significantly increases costs7,8. Additionally, the treatment protocol requires lymphodepleting chemotherapy prior to CAR-T cell infusion, leading to substantial toxicities7–9. The complexity of this treatment process also poses challenges for both clinical care providers and patients9–11. As a result, the high costs and intricate procedures associated with ex vivo CAR-T therapy severely limit its accessibility for patients who could benefit from this treatment10,11.

Efforts have been made to develop strategies for generating CAR-T cells in vivo. Lipid nanoparticle (LNP)-mediated mRNA delivery for transient expression has been widely utilized in vaccine development12. Recently, CD5-targeted LNPs have been reported to induce CAR-T cells for the treatment of cardiac injury13. CAR-T cell therapies for cancer typically require CAR-T cells to persist in vivo for extended periods—ranging from several months to years—to maintain therapeutic efficacy14,15. Due to the transient nature of mRNA expression via LNPs, LNP-based CAR-T cell generation has been mainly investigated for autoimmune disease treatments, with less focus on cancer applications16.

Lentiviral vector, by contrast, can integrate delivered genes into the host genome, enabling long-term expression of the transgene17. As a result, lentiviral vectors have been explored for in vivo CAR-T cell generation for cancer therapy18–23. CD4-targeted or CD8-targeted lentiviral vectors have been used to generate CD19-CAR-expressing CD4+T cells or CD8+T cells in vivo19–21. In addition, CD3-targeting lentiviral vector have also been investigated for in vivo generation of CAR-T cells, which could be further enhanced by IL-7 or Dasatibib administration21,24. These lentiviral vectors are derived from lentiviral vector pseudotyped with envelope of nipah or measles viruses and require ex vivo activation of PBMCs before engraftment into NSG mice, or additional interleukin-7 (IL-7) administration18–21. Additionally, lentiviral vectors pseudotyped with envelope of cocal virus have been developed for CAR-T cell generation in mouse and non-human primate models, showing promising therapeutic effects in B cell lymphoma22,23. Recent clinical trials have also explored lentiviral vector-based in vivo CAR-T cell therapies for treating B cell lymphoma (NCT06528301, NCT06743503, NCT06539338, NCT06678282, and NCT06890065).

Sindbis virus, a member of the alphavirus family, has a broad host range due to the widespread distribution of its cellular receptor25. Modifications to the Sindbis virus envelope have enhanced its viral titer and targeting specificity, while minimizing non-specific infectivity in the liver and spleen25–27. Lentiviral vectors pseudotyped with engineered Sindbis virus envelope have been developed for in vivo gene delivery25,27. A previous study demonstrated that a lentiviral vector pseudotyped with a mutant Sindbis envelope, coupled with a bispecific antibody that redirects the virus to CD3⁺ human T cells, can mediate T cell infection in vivo in a mouse model28. PBMCs required prior activation before injection into NSG mice, and the infection efficiency remained to be improved28. Therefore, an efficient Sindbis virus envelope-based lentiviral vector capable of targeting human primary T cells for CAR-T cell generation has yet to be developed.

In this study, we evaluated the delivery efficiency and specificity of engineered Sindbis virus envelope-pseudotyped lentiviral vector in combination with human CD3 and CD28 antibodies, demonstrating high specificity and efficacy in infecting human primary T cells both in vitro and in vivo. The successful delivery of a CD19-targeting CAR construct enabled efficient in vivo elimination of B cell lymphoma. These findings introduce a novel approach for in vivo T cell engineering, with potential applications in the generation of CAR-T and T cell receptor (TCR)-T cells for clinical therapeutic use.

Results

Efficient gene delivery in human T cells

To generate a lentiviral vector specific for human T cells both in vitro and in vivo, we first employed a previously reported engineered Sindbis envelope protein27 pseudotyped in a two-vector lentiviral vector system (hereafter referred to as SLV) (Fig. 1A). The engineered Sindbis envelope protein contains mutations that abolish its original receptor-binding capacity, and is inserted with two Z domains of protein A that specifically bind to the Fc region of IgG27. Results showed that conjugation of the SLV with a human HLA antibody (SLV-HLA), but not IgG (SLV-IgG), enabled efficient infection of 293 T cells (expressing HLA antigen) in the two-vector lentivirus system (Fig. 1B, C, Figure S1A). Of note, we observed that the infection efficiency of SLV-HLA was significantly reduced in normal serum medium (10% FBS) compared to serum-free medium (Figure S1B, C). We also tested the engineered Sindbis envelope protein in a three-vector second-generation lentivirus system (Figure S1D). The results showed efficient infection of 293 T cells when SLV was conjugated with human HLA antibody, but not IgG (Figure S1E). Interestingly, the two-vector lentivirus system yielded a significantly higher viral titer compared to the three-vector system, indicating that this three-vector system requires further optimization for efficient lentiviral vector production (Fig. 1D). Therefore, the two-vector system was used in the subsequent experiments unless otherwise specified in this study.

Fig. 1.

Fig. 1

Open in a new tab

SLV-HLA exhibits potential for infecting human T cells in vitro. (A) Schematic representation of the plasmid vectors used for packaging lentiviral vector pseudotyped with the engineered Sindbis virus envelope (SLV). Sin-G-ZZ represents the mutated Sindbis virus envelope containing two Z domains of protein A, as described in the Methods. CMV, cytomegalovirus promoter; LTR, long terminal repeat. (B) Fluorescence microscopy analysis of SLV-IgG and SLV-HLA infection in 293 T cells. SLV-IgG, SLV conjugated with IgG. SLV-HLA, SLV conjugated with HLA antibody. Scale bar: 50 μm. For panels B-G, lentiviral vector supernatants were pre-incubated with the indicated antibodies (0.25 µg/mL) at 4 °C for 30 min, followed by 100-fold concentration using centrifugal filtration. For viral infection, 1 µL of concentrated lentiviral vector was added to the 100 µL cell culture supernatant in 96-well plate and centrifuged at 800 g for 2 h at 37 °C. After centrifugation, the medium was replaced with fresh culture medium. Data were collected 24 h after infection in B, C and 72 h after infection for D-G. (C) Summary of percentages of GFP-positive cells in 293 T cells from Figure S1A (n = 3, mean ± SD). (D) Lentiviral vector titer in the supernatants of lentiviral vector generated from the two-vector system and three-vector system. (E) Fluorescence microscopy analysis of SLV-IgG and SLV-HLA infection in Jurkat cell. Scale bar: 50 μm. (F) Representative flow cytometry plots showing GFP expression in Jurkat cells following lentiviral vector infection. (G) Summary of percentages of GFP-positive cells in Jurkat cells from F (n = 3, mean ± SD). Technical replicates are shown in C, D and G. Data represent results of at least two independent experiments in B-G. *, P < 0.05. **, P < 0.01. ***, P < 0.001. ****, P < 0.0001.

To evaluate whether the SLV system could mediate efficient gene delivery into human T cells, we first tested it in Jurkat cells, a human CD4+T cell line expressing both HLA and CD3 antigens. Results indicated that SLV-HLA, but not SLV-IgG, efficiently facilitated gene delivery in Jurkat cells (Fig. 1E-G). These results suggest that this engineered Sindbis envelope protein-pseudotyped lentiviral vector may serve as a viable platform for gene delivery in human T cells.

Specific gene delivery in Jurkat cells

CD3ε is an antigen specifically expressed in human T cells29. To generate a T cell-specific lentiviral vector, we investigated whether conjugating the SLV with a human CD3ε antibody could target the virus specifically to human T cells. Results indicated that CD3ε antibody-conjugated SLV (hereafter referred to as SLV-T), but not SLV-IgG, efficiently infected Jurkat cells, as evidenced by fluorescence microscopy (Fig. 2A). Flow cytometry analysis further confirmed the high infection efficiency of SLV-T, but not SLV-IgG, in Jurkat cells (Fig. 2B, C). In contrast, both SLV-T and SLV-IgG failed to efficiently infect Raji cells, a human B cell line that does not express CD3ε (Fig. 2D, E, Figure S2), indicating that SLV-T specifically targets CD3ε-expressing Jurkat cells. These results suggest that SLV-T could potentially be utilized for targeted and efficient gene delivery into human T cells.

Fig. 2.

Fig. 2

Open in a new tab

Specific infection of SLV-T in human T cells in vitro. (A) Fluorescence microscopy analysis of SLV-IgG and SLV-T in Jurkat cells. SLV-T, SLV conjugated with CD3ε antibody. SLV supernatants were pre-incubated with indicated antibodies (0.25 µg/mL) at 4 °C for 30 min, followed by 100-fold concentration using centrifugal filtration. Scale bar: 50 μm. For viral infection, 1 µL of concentrated lentiviral vector was added to the 100 µL cell culture supernatant in 96-well plate and centrifuged at 800 g for 2 h at 37 °C. After centrifugation, the medium was replaced with fresh culture medium. Data were collected 72 h after infection. (B) Representative flow cytometry plots showing GFP expression in Jurkat cells following infection with SLV-IgG or SLV-T. (C) Summary of percentages of GFP-positive cells in Jurkat cells from B (n = 3, mean ± SD). (D) Representative flow cytometry plots showing GFP expression in Raji cells following infection with SLV-IgG or SLV-T. (E) Summary of percentages of GFP-positive cells in Raji cells from D (n = 3, mean ± SD). Technical replicates are shown in C and E. Compiled data from one experiment. Data represent results of two independent experiments. One-way ANOVA with Tukey correction for multiple comparisons. ****, P < 0.0001.

Specific gene delivery in human primary T cells by SLV-T in vitro

Next, we assessed whether SLV-T could specifically infect human primary T cells derived from human peripheral blood mononuclear cells (PBMCs). We first screened donors whose PBMCs exhibited higher infection efficiency using SLV-T. The results revealed that infection efficiency varied between different donors, with a few showing obvious infection by SLV-T (Figure S3A). PBMCs from donors exhibited obvious infection efficiency were used for subsequent studies, unless otherwise noted. CD3 and CD28 antibodies for activation of TCR signaling are commonly used to facilitate lentiviral transduction and T cell proliferation during the generation of CAR-T cells in vitro5. Results showed that, following CD3 and CD28 antibody stimulation, both SLV-T and vesicular stomatitis virus glycoprotein (VSV-G)-pseudotyped lentiviral vector (VSV-G-LV), but not SLV-IgG, could infect primary T cells from PBMCs (Figure S3B-D). Of note, VSV-G-LV showed better infection in primary T cell compared to SLV-T, following CD3 and CD28 antibody stimulation (Figure S3B-D). However, in the absence of CD3 and CD28 antibody stimulation before lentiviral vector infection, VSV-G-LV exhibited minimal infection of primary T cells from PBMCs (Fig. 3A, B). In contrast, SLV-T, but not VSV-G-LV or SLV-IgG, was able to infect primary T cells even without additional CD3 and CD28 stimulation before virus infection (Fig. 3A, B). The infection of SLV-T was observed specifically in T cells, including both CD4+T cells and CD8+T cells, but not in other immune cells, such as B cells and natural killer (NK) cells (Fig. 3C, Figure S3D). These results indicate that SLV-T could mediate gene delivery specifically into primary T cells without the requirement for additional CD3 and CD28 antibody stimulation.

Fig. 3.

Fig. 3

Open in a new tab

SLV-T exhibits potential for infecting human PBMCs in vitro. (A) Representative flow cytometry plots showing GFP expression in PBMCs following infection with SLV-IgG, SLV-T (0.25 µg/mL), or with lentiviral vector pseudotyped with vesicular stomatitis virus glycoprotein (VSV-G-LV). PBMCs were cultured in 24-well plates (1 × 106/well) without addition of anti-CD3/CD28 beads before lentiviral vector infection. Data were collected 96 h after infection. (B) Summary of percentages of GFP-positive cells in PBMCs from A (n = 3, mean ± SD). One-way ANOVA with Tukey correction for multiple comparisons. **, P < 0.01. ***, P < 0.001. (C) Summary of percentages of GFP-positive cells in each immune subset from Figure S3C (n = 3, mean ± SD). Two-tailed unpaired Student’s t-test; **, P < 0.01. ***, P < 0.001. Technical replicates are shown in B and C. Compiled data from one experiment. Data represent results of at least two independent experiments using PBMCs from different donors.

Enhancing infection efficiency of SLV-T in human primary T cells in vitro

Given that the gene delivery efficiency of SLV-T in human primary T cells was relatively low (Fig. 3A, B), and that additional TCR activation with CD3 and CD28 antibody treatment before virus infection slightly enhanced SLV-T infection in primary T cells (Fig. 3A, B; Figure S3A, B), we investigated whether co-conjugation of CD3 and CD28 antibodies to SLV (hereafter referred to as enSLV-T) could further improve the infection efficiency in primary T cells compared to SLV-T (conjugation of CD3 antibody alone). The results demonstrated that enSLV-T significantly enhanced gene delivery efficiency in primary T cells compared to SLV-T (Fig. 4A, B). The infection efficiency of enSLV-T was increased in both CD4+T cells and CD8+T cells (Fig. 4C, Figure S4). The non-specific infection of enSLV-T in other immune cells, such as B cells and NK cells, remained minimal (Fig. 4C, Figure S4A). Of note, there was a trend of increased vector copy number of lentiviral vector in enSLV-T-generated CAR-T cells compared to in VSV-G-LV-generated CAR-T cells (Figure S4B). These results suggest that enSLV-T enhances specific gene delivery to human primary T cells in vitro.

Fig. 4.

Fig. 4

Open in a new tab

EnSLV-T shows improved infection efficiency in human PBMCs in vitro. (A) Representative flow cytometry plots showing GFP expression in PBMCs following infection with SLV-T or enSLV-T. Antibodies were used at a concentration of 0.25 µg/mL for conjugation. PBMCs were cultured in 24-well plates (1 × 106/well) without addition of anti-CD3/CD28 beads before lentiviral vector infection. Data were collected 96 h after infection. (B) Summary of percentages of GFP-positive cells in PBMCs from A (n = 3, mean ± SD). (C) Summary of percentages of GFP-positive cells in each immune subset from Figure S4 (n = 3, mean ± SD). Technical replicates are shown in B and C. Compiled data from one experiment. Data represent results of at least two independent experiments using PBMCs from different donors. Two-tailed unpaired Student’s t-test; *, P < 0.05.

In vivo generation of CD19-CAR-T cells by enSLV-T

We then assessed the delivery efficiency and specificity of enSLV-T in a Raji-bearing mouse model, which is commonly used for B cell lymphoma studies. A previously reported CD19-targeting CAR construct5along with a GFP reporter, was packaged in enSLV-T (hereafter referred to as CD19-enSLV-T) (Fig. 5A). NOD.Cg-Prkdcscid Il2rgtm1Vst/Vst (NPG) mice engrafted with human PBMCs and luciferase-expressing Raji cells were used as recipients (Fig. 5B). The recipients were injected with either CD19-enSLV-T or in vitro-generated CD19-CAR-T cells (Fig. 5B, Figure S5A). Notably, there was no significant difference in the in vitro cytotoxicity between CD19-CAR-T cells generated by CD19-enSLV-T and those generated by VSV-G-LV (Figure S5B). CD19-CAR-T cells in the PMBCs of mice were monitored dynamically by flow cytometry analysis (Fig. 5B, C), and bioluminescence imaging was used to assess the growth of Raji cells in the recipients.

Fig. 5.

Fig. 5

Open in a new tab

Generation of CD19CAR-T cells by CD19-enSLV-T in vivo. (A) Schematic representation of the CD19-targeting CAR-expressing construct. T2A, self-cleaving peptide. (B) Experimental design of the CD19-CAR-T cell or lentiviral vector treatment against Raji cells in NPG mice homogenized with human PBMCs. PBMCs were i.v. injected into NPG mice (1 × 107/mouse) seven days before i.v. injection of CD19-CAR-T cells (5 × 106/mouse). Raji cells were intraperitoneally (i.p.) injected into NPG mice five days before injection of CD19-CAR-T cells. Lentiviral vectors were injected three times with an interval of 2 days through i.v. injection (2 × 107 TU/mice). (C) Representative flow cytometry plots showing CD3 and CAR(GFP) expression in total PBMCs of blood from mice shown in B. CD19-enSLV-T, enSLV-T containing the CD19-targeting CAR construct shown in A. Data were collected at day15 after CAR-T cell injection or the third-time lentiviral vector injection. For C-E, data are from one independent experiment. Data represent results of two independent experiments using PBMCs from different donors. (D) Summary of percentages of CD3+GFP+ cells in PBMCs from C (n = 3 for CD19-CAR-T cell group, n = 4 for the rest, mean ± SD). One-way ANOVA with Tukey correction for multiple comparisons; *, P < 0.05. (E) Representative bioluminescence images of NPG mice xenografted with Raji cells as designed in B (n = 4 for each group, mean ± SD). Representative figures from one independent experiment. (F) Kaplan-Meier survival of Raji-bearing NPG mice treated with CD19-CAR-T cells or indicated lentiviral vector (n = 8 for each group) (log-rank Mantel-Cox test with Bonferroni’s correction for multiple comparisons; ****, P < 0.0001). Compiled data from two independent experiments. Data represent results of at least two independent experiments.

As expected, CD19-enSLV-T specifically targeted T cells in the PBMCs, with barely detectable infection in non-T cells (Fig. 5C, D, Figure S5C). The expression levels of co-inhibitory molecules did not show significant difference between CD19-CAR-T cells from mice injected with CD19-enSLV-T and those from mice receiving in vitro-generated CD19-CAR-T cells by VSV-G-LV (Figure S5D, E). Interestingly, a slight reduction in the proportion of central memory-like T cells was observed in CD19-CAR-T cells from mice injected with CD19-enSLV-T compared to those receiving in vitro-generated CD19-CAR-T cells (Figure S5F, G). Both groups of mice receiving either CD19-CAR-T cells or CD19-enSLV-T effectively controlled tumor growth in vivo (Fig. 5E, Figure S5H-J). Notably, the proportion of CAR-expressing T cells was significantly lower in the CD19-enSLV-T-injected mice compared to the CD19-CAR-T cell-injected mice (Fig. 5C, D). The survival was significantly improved in both the CD19-CAR-T cell-injected and CD19-enSLV-T-injected groups of mice compared to the IgG-SLV-injected group of mice (Fig. 5F). No significant changes in body weight were observed between the groups of mice receiving CD19-CAR-T cells or CD19-enSLV-T (Figure S5K). These results indicate that CD19-enT-SLV specifically targets T cells and successfully generates therapeutic CD19-CAR-T cells in vivo.

In summary, our study provides a promising approach for the in vivo engineering of human T cells, which may substantially reduce the cost, time, and complexity associated with clinical ex-vivo T cell-based therapies.

Discussion

Ex vivo CAR-T therapy has demonstrated significant clinical benefits in the treatment of various cancers, particularly B cell lymphoma1. However, the complexity and high costs associated with ex vivo CAR-T therapy have significantly limited its widespread application in patients who could benefit from this treatment. In this study, we developed a T-cell-specific lentiviral vector system for the generation of CAR-T cells both in vitro and in vivo. This lentiviral vector does not require additional CD3/CD28 activation for infection and exhibits negligible infection of non-T cells. Importantly, in vivo-generated CD19-CAR-T cells were capable of efficiently eliminating B cell lymphoma in a Raji cell-derived xenograft mouse model, underscoring the clinical potential of this approach. Intriguingly, CAR-T cells generated through antibody-conjugated lentiviral transduction showed therapeutic efficacy comparable to adoptively transferred CAR-T cells, despite their lower transduction efficiency. This phenomenon may be attributed to the use of a third-generation CD19-CAR construct, which incorporates signal transduction domains of both human CD28 and 4-1BB, thereby enabling potent tumor clearance with fewer effector cells.

Our study provides a potentially valuable tool for broad T cell engineering in vivo. The targeted lentiviral vector could also be adapted for the delivery of other CAR molecules, such as GPC-CAR, HER2-CAR, and CLDN18.2-CAR, for targeting solid tumors. In addition to CAR-T cell therapy, other T cell-based immunotherapies, such as TCR-T cell therapy, have been developed, which require the delivery of tumor antigen-specific TCR molecules to T cells30. Therefore, the T-cell-targeting lentiviral vector system developed in our study may also serve as a platform for the generation of TCR-T cells in vivo. Furthermore, this lentiviral vector could be applied for in vivo gene editing in T cells.

While the CD3 antibody enables the lentiviral vector to specifically recognize and bind to human T cells, the combination of CD3 and CD28 antibodies on the lentiviral vector mimics the CD3/CD28 activation signal typically used during ex vivo CAR-T cell generation. CD28 is a costimulatory molecule that plays a key role in T cell activation31. In addition to CD28, several other costimulatory molecules, such as 4-1BB, OX40, ICOS, GITR, and CD27, contribute to T cell activation32. Some of these molecules can have both stimulatory and inhibitory effects depending on the context32. Previous studies have shown that incorporating CD58 and CD80 T cell costimulatory ligands, along with anti-CD3 scFv, into the cocal envelope-based lentiviral vector enhances both T cell activation and costimulation, thus improving T cell infection efficiency23. Our findings demonstrate that the inclusion of CD3 and CD28 antibodies in S-LV significantly increases T cell infection efficiency compared to CD3 alone, suggesting that the appropriate incorporation of costimulatory molecules into lentiviral vectors can enhance primary T cell infection. Notably, the infection efficiency varied between PBMCs from different donors, which may be due to differences in IgG-producing capacity of B cells in PBMCs from various donors. Therefore, it would be advisable to screen PBMC infection efficiency prior to conducting in vivo gene delivery or potential clinical applications.

A major concern with using lentiviral vector vectors for generating CAR-T cells, whether ex vivo or in vivo, is the potential for random integration of the delivered genes into the host cell genome, which could potentially disrupt tumor suppressor gene expression and lead to tumorigenesis17,33. However, comprehensive studies have shown that the risk of secondary malignancies following ex vivo CAR-T cell therapy is low and comparable to that seen in patients undergoing stem cell transplantation34–36. The CAR gene is rarely found in secondary primary malignancies after CAR-T therapy, with only one case observed among thousands of patients35,36. No direct evidence has been presented to link secondary primary malignancies to CAR-T therapy.

Recently, several methods for targeted gene integration have been explored37,38. One study reported the use of an engineered R2 retrotransposon system that facilitates RNA-mediated targeted gene integration in mammalian cells38. Given that lentiviral vector delivers genes via RNA, combining targeted gene integration technologies with targeted lentiviral vectors could mitigate concerns regarding random genome integration.

Several studies have also explored the use of cell-specific promoters to drive gene expression in CAR-T cells22,23,39. The MND promoter, a synthetic promoter that drives constitutive expression within the hematopoietic lineage, has been reported to enhance CAR expression in CAR-T cells22. Furthermore, T cell-specific synthetic promoters have been developed for driving CAR expression in T cells (patent: WO2025003526A1)39. The use of T cell-specific promoters in conjugation with T cell-targeting lentiviral vectors could provide an added layer of safety for potential clinical applications.

There are several challenges to the practical clinical use of this gene delivery system. One major issue is that it currently employs a two-plasmid lentiviral vector packaging system, which could raise health concerns. Therefore, developing and optimizing this gene delivery system using a third- or fourth-generation lentiviral vector system would be a more appropriate approach for future applications. Additionally, we observed that the current antibody-conjugated lentiviral vector system demonstrates relative instability, particularly in media containing IgGs. Future improvements may include replacing the ZZ domain with suitable anti-CD3/CD28 nanobodies, which could enhance both stability and clinical applicability.

In conclusion, our study presents an efficient tool for T cell engineering that offers a promising approach to the development of cost-effective and safe T cell therapies for a broad spectrum of cancers.

Methods

Mice

All animal experiments were approved by the Institutional Animal Care and Use Committee (IACUC) of the Suzhou Institute of Systems Medicine, Chinese Academy of Medical Sciences (CAMS-ISM) (ISM-IACUC-0174-R). NOD.Cg-Prkdcscid Il2rgtm1Vst/Vst (NPG) mice, aged 6–8 weeks, were obtained from Beijing Vitalstar Biotechnology. Both male and female mice were used for experiments unless otherwise specified. For Raji cell-derived xenograft model, PBMCs were intravenously (i.v.) administered into NPG mice (1 × 107/mouse) seven days prior to the i.v. injection of CD19-CAR-T cells (5 × 106/mouse) or lentiviral vectors. Raji cells (5 × 105/mouse) were intraperitoneally (i.p.) injected into NPG mice five days before the administration of CD19-CAR-T cells or one day before the first-time injection of lentiviral vectors. Recepient mice with more than 5% T cell engraftment on the day of lentiviral vector infection were selected and randomly assigned to different treatment groups. Lentiviral vector was administered three times via i.v. injection (2 × 107 TU/mice) with a 2-day interval. Following CAR-T cell or lentiviral vector administration, blood samples were collected from the orbital venous plexus of the mice at specified time points. Red blood cells were lysed using ACK lysis buffer (Cat# A1049201, Gibco). The PBMCs were subsequently analyzed by flow cytometry. Mice injected with Raji cells were monitored using in vivo imaging (IVIS Lumina III, PerkinElmer) at designated time points. D-luciferin potassium salt (Cat# 122799, PerkinElmer) was dissolved in PBS to create a working solution at 15 mg/mL. All mice were administered an intraperitoneal injection of luciferin solution (150 mg/kg) 10 min prior to in vivo imaging.

Plasmid construction and lentiviral vector production

Lentiviral vector production was carried out as previously described5. For lentiviral vector production, 293 T packaging cells were seeded in DMEM (Cytiva) complete medium, supplemented with 10% fetal bovine serum (FBS) (Cat# 16-250-078, Gibco™ Fetal Bovine Serum, ultra-low IgG) and 1% Penicillin/Streptomycin, in a 15 cm dish and cultured until the cells reached 80–90% confluence. For VSV-G pseudotyped lentiviral vector, the following plasmids were used: psPAX2 (Cat#12260, Addgene, 15 µg), pMD2.G (Cat#12259, Addgene, 15 µg), and the CMV-CD19-CAR/Ef1a-GFP-expressing plasmid (Cat# LC0001-3, Public Protein/Plasmid Library, 30 µg). The viral supernatant was collected 24-, 48-, and 72-hours post-transfection.

The sequence of the engineered Sindbis virus envelope with the ZZ domain inserted was derived from plasmid 2.2 (Cat#34885, Addgene) and synthesized by GENEWIZ. It was then cloned into the pMD2.G backbone to replace the VSV-G envelope, yielding the plasmid pCMV-Sin-G-ZZ. pNL4-3-GFP was generated from pNL4-3.Luc.R-E- (Cat#V012797, NovoPro) by replacing the luciferase-coding sequence with the GFP-coding sequence. pNL4-3-CD19CAR-GFP was generated from pNL4-3.Luc.R-E- by replacing the luciferase-coding sequence with the CD19CAR-T2A-GFP-coding sequence.

For the production of SLV, the following packaging plasmids were used for the two-vector lentivirus system: pCMV-Sin-G-ZZ (37.5 µg), and pNL4-3-GFP (37.5 µg) or pLN4-3-CD19CAR-GFP (37.5 µg); the following packaging plasmids were used for the three-vector lentivirus system: pCMV8.2dVPR (Cat#8455, Addgene, 15 µg), pCMV-Sin-G-ZZ (15 µg), and either the pLenti-CMV-luc2CP-GFP-Puro plasmid (Cat#80045-4c, Public Protein/Plasmid Library, 30 µg) or the CMV-CD19-CAR/Ef1a-GFP-expressing plasmid (Cat# LC0001-3, Public Protein/Plasmid Library, 30 µg). Polyethylenimine (PEI) (Polysciences, Catalog# 23966-1) was used for the plasmid transfection as previously described5. One hour prior to transfection, the culture medium of 293 T cells was washed with PBS (Cytiva) twice and replaced with DMEM without FBS. The viral supernatant was harvested 24 h after transfection and filtered through a 0.45 μm filter, pre-incubated with the indicated antibodies (0.25 µg/mL) at 4 °C for 25–30 min, and then concentrated 100-fold using a centrifugal filter unit (Cat# UFC910024, Millipore) at 4,000 g for 30 min at 4 °C, or concentrated and purified for about 100-fold by ultracentrifugation in sucrose gradient and resuspended in PBS at roughly 1/100th of the original volume (e.g., a total of 0.2 mL PBS used to resuspend a total of 20 mL of lentiviral supernatant). Lentiviral vector titer was measured using the Lentivirus Titer p24 ELISA Kit according to the manufacturer’s instructions (Cat#L01019, GenScript) (30–50 ng/mL of p24 corresponds to about 10⁶ TU/mL of the lentivirus vectors). The following antibodies were used: anti-human HLA (Cat# 327002 and Cat# 335902, BioLegend), mouse IgG2a (Cat# 400202, BioLegend), anti-human CD3 (Cat# 317302, BioLegend), and anti-human CD28 (Cat# 302902 and Cat# 377204, BioLegend). For viral infection in cell lines, cells were washed with PBS twice and replaced with cell culture medium without FBS before infection. 1 µL of the concentrated lentiviral vector was added to 100 µL of cell culture supernatant in a 96-well plate and centrifuged at 800 g for 2 h at 37 °C, unless otherwise noted. After centrifugation, the medium was replaced with fresh culture medium. Flowcytometry and/or fluorescence microscopy were used for analysis the GFP expression in cells. To measure the relative copy number of lentiviral vector in CAR-T cells, the following primers were used: GFP primers (Forward: 5‘-CTTCCTGCACGCCATCAACAACG, Reverse: 5′-GATGATCTTGTCGGTGAAGATCACG) and human ACTB primers (Forward: 5′-CTGGAACGGTGAAGGTGACA, Reverse: 5′-AAGGGACTTCCTGTAACAACGCA). CAR-T cells were sorted by GFP positive and then genome DNA was extracted (DNeasy Blood and Tissue Kit (Qiagen)). FastStart™ universal SYBRR Green mix (ROX) and Bio-Rad iQ5 Real Time PCR were used for the Quantitative PCR (qPCR).

Ex-vivo CAR-T production

Ficoll-Paque (Cat# 25710, Dongfang Huahui Co. Ltd., Beijing) was used to isolate PBMCs as previously described40. The PBMCs were subjected to T cell enrichment using the T Cell Isolation Kit (Cat# 17951, STEMCELL Technologies) according to the manufacturer’s instructions. The isolated PBMCs or enriched T cells were activated using Dynabeads™ Human T-Activator CD3/CD28 (Cat# 11132D, Gibco; T cells: beads ratio = 1:1) for 24 h. PBMCs without activation or activated PBMCs and T cells washed twice with PBS (Cytiva) and resuspended in X-VIVO™ 15 medium (Catalog #02-053Q, Lonza) without FBS prior to lentiviral transduction. For SLV infection in PBMCs, multiplicity of infection (MOI) equal 50 was used. SLVs were infected twice with an interval of 24 h. During infection, concentrated lentiviral vector was added to the cells in serum-free medium, and the mixture was centrifuged at 800 g for 2 h at 37 °C. Two hours post-transduction, the lentiviral vector-containing medium was replaced with X-VIVO™15 medium supplemented with IL-2 (Peprotech, Cat# 200-02, 100 U/mL).

Flow cytometry analysis

ACK lysis buffer (Cat# A1049201, Gibco) was used to lyse red blood cells from whole blood according to the manufacturer’s instructions. For surface marker staining, cells were first stained with the LIVE/DEAD™ Fixable Aqua Dead Cell Stain Kit (Cat# L34966, Invitrogen), followed by Fc receptor blocking (Cat# 422301, BioLegend). The cells were then stained with the indicated antibodies (1:200 final dilution unless otherwise specified) for 30 min at 4 °C. The following antibodies were used: anti-human CD3ε (Cat# 317336 and Cat# 317302, BioLegend), anti-human CD19 (Cat# 392504 and Cat# 302218, BioLegend), anti-human CD20 (Cat# 302322, BioLegend), anti-human CD4 (Cat# 47-0048-42 and Cat# 48-0049-42, eBioscience), anti-human CD8α (Cat# 25-0088-42, eBioscience), anti-human CD16 (Cat# 48-0168-42, Invitrogen), anti-human CD56 (Cat# 362542, BioLegend). Data collection was performed using a Cytek Aurora spectral flow cytometer, and FACS data analysis was conducted using FlowJo software (version 10.6). In vitro killing assay was performed as previously described5. Briefly, CAR-T cells (effector) were were sorting purified by GFP positive with BD and co-cultured with Raji cells stably expressing firefly luciferase (target) at indicated ratios in a 96-well cell culture plate for 48 h, cancer cell lysis was detected using the Steady-Glo® Luciferase assay system (Cat#E2520, Promega). Bioluminescence was measured 48 h after co-culture.

Statistical analyses

Statistical analyses were performed using a two-tailed unpaired Student’s t-test with GraphPad Prism 6.01 software, unless otherwise specified. For survival curve analysis, the Log-rank (Mantel-Cox) test with Bonferroni’s correction for multiple comparisons was applied, also using GraphPad Prism 6.01. Data are presented as mean ± SD, unless otherwise indicated. Statistical significance was defined as follows: *, P < 0.05. **, P < 0.01. ***, P < 0.001. ****, P < 0.0001.

Study approval

This study is reported in accordance with Animal Research: Reporting of In Vivo Experiments (ARRIVE) guidelines. All methods were performed in accordance with the relevant guidelines and regulations, as approved by the Ethics Committee of Suzhou Institute of Systems Medicine and Second Xiangya Hospital (ISM-IACUC-0174-R and Ref# 052 (2018)). This study involves PBMCs from healthy donors and was approved by the Institutional Review Board at the Second Xiangya Hospital (Ref# 052 (2018)). All study participants provided informed consent prior to participation. The study was conducted in compliance with the Declaration of Helsinki.

Supplementary Information

Below is the link to the electronic supplementary material.

Supplementary Material 1 (1.3MB, docx)

Acknowledgements

We thank Mr. Qingyuan He from CAMS-ISM animal facility for the help of animal services for several mouse experiments. We would like to thank Dr. Yi Liao from Xiamen University for the help of statistical analyses.

Author contributions

Q.Z. and Z.F. conceived the research idea. Q.Z. and Z.F. designed and supervised this study. M.T., J.Y., P.Z., and M.L. acquired the data. M.T., J.Y., P.Z., M.L. and Q.Z. analyzed the data. M.T., J.Y., Q.Z., Z.F., P.Z., and M.L. prepared the manuscript.

Data availability

The datasets used and/or analysed during the current study available from the corresponding author on reasonable request.

Declarations

Competing interests

The authors declare no competing interests.

Footnotes

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

These authors contributed equally to this work: Muhadasi Tuerxunyiming and Jianguo Michael.

References

  • 1.Cappell, K. M. & Kochenderfer, J. N. Long-term outcomes following CAR T cell therapy: what we know so Far. Nat. Rev. Clin. Oncol.20, 359–371. 10.1038/s41571-023-00754-1 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Schett, G., Mackensen, A. & Mougiakakos, D. CAR T-cell therapy in autoimmune diseases. Lancet402, 2034–2044. 10.1016/S0140-6736(23)01126-1 (2023). [DOI] [PubMed] [Google Scholar]
  • 3.Di Blasi, R. et al. Outcomes of patients with aggressive B-cell lymphoma after failure of anti-CD19 CAR T-cell therapy: a DESCAR-T analysis. Blood140, 2584–2593. 10.1182/blood.2022016945 (2022). [DOI] [PubMed] [Google Scholar]
  • 4.Baker, D. J., Arany, Z., Baur, J. A., Epstein, J. A. & June, C. H. CAR T therapy beyond cancer: the evolution of a living drug. Nature619, 707–715. 10.1038/s41586-023-06243-w (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Fu, Z. et al. IL-2-inducible T cell kinase deficiency sustains chimeric antigen receptor T cell therapy against tumor cells. J. Clin. Invest.13510.1172/JCI178558 (2024). [DOI] [PMC free article] [PubMed]
  • 6.Ye, X. et al. CD19 -targeted CAR T therapy treating hematologic malignancies: hidden danger is the next neighbor to security? Front. Immunol.16, 1490491. 10.3389/fimmu.2025.1490491 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Amini, L. et al. Preparing for CAR T cell therapy: patient selection, bridging therapies and lymphodepletion. Nat. Rev. Clin. Oncol.19, 342–355. 10.1038/s41571-022-00607-3 (2022). [DOI] [PubMed] [Google Scholar]
  • 8.Ayala Ceja, M., Khericha, M., Harris, C. M., Puig-Saus, C. & Chen, Y. Y. CAR-T cell manufacturing: major process parameters and next-generation strategies. J. Exp. Med.22110.1084/jem.20230903 (2024). [DOI] [PMC free article] [PubMed]
  • 9.Brudno, J. N. & Kochenderfer, J. N. Current Understanding and management of CAR T cell-associated toxicities. Nat. Rev. Clin. Oncol.21, 501–521. 10.1038/s41571-024-00903-0 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Strzelec, A. & Helbig, G. Are we ready for personalized CAR-T therapy? Eur. J. Haematol.112, 174–183. 10.1111/ejh.14039 (2024). [DOI] [PubMed] [Google Scholar]
  • 11.Joy, R., Phair, K., O’Hara, R. & Brady, D. Recent advances and current challenges in CAR-T cell therapy. Biotechnol. Lett.46, 115–126. 10.1007/s10529-023-03461-0 (2024). [DOI] [PubMed] [Google Scholar]
  • 12.Hou, X., Zaks, T., Langer, R. & Dong, Y. Lipid nanoparticles for mRNA delivery. Nat. Rev. Mater.6, 1078–1094. 10.1038/s41578-021-00358-0 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Rurik, J. G. et al. CAR T cells produced in vivo to treat cardiac injury. Science375, 91–96. 10.1126/science.abm0594 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Rafiq, S., Hackett, C. S. & Brentjens, R. J. Engineering strategies to overcome the current roadblocks in CAR T cell therapy. Nat. Rev. Clin. Oncol.17, 147–167. 10.1038/s41571-019-0297-y (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Tang, L., Pan, S., Wei, X., Xu, X. & Wei, Q. Arming CAR-T cells with cytokines and more: innovations in the fourth-generation CAR-T development. Mol. Ther.31, 3146–3162. 10.1016/j.ymthe.2023.09.021 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Nie, S., Qin, Y., Ou, L., Chen, X. & Li, L. In situ reprogramming of immune cells using synthetic nanomaterials. Adv. Mater.36, e2310168. 10.1002/adma.202310168 (2024). [DOI] [PubMed] [Google Scholar]
  • 17.Warnock, J. N., Daigre, C. & Al-Rubeai, M. Introduction to viral vectors. Methods Mol. Biol.737, 1–25. 10.1007/978-1-61779-095-9_1 (2011). [DOI] [PubMed] [Google Scholar]
  • 18.Ho, N. et al. In vivo generation of CAR T cells in the presence of human myeloid cells. Mol. Ther. Methods Clin. Dev.26, 144–156. 10.1016/j.omtm.2022.06.004 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Agarwal, S. et al. In vivo generation of CAR T cells selectively in human CD4(+) lymphocytes. Mol. Ther.28, 1783–1794. 10.1016/j.ymthe.2020.05.005 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Frank, A. M. et al. CD8-Specific designed Ankyrin repeat proteins improve selective gene delivery into human and primate T lymphocytes. Hum. Gene Ther.31, 679–691. 10.1089/hum.2019.248 (2020). [DOI] [PubMed] [Google Scholar]
  • 21.Frank, A. M. et al. Combining T-cell-specific activation and in vivo gene delivery through CD3-targeted lentiviral vectors. Blood Adv.4, 5702–5715. 10.1182/bloodadvances.2020002229 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Michels, K. R. et al. Preclinical proof of concept for vivovec, a lentiviral-based platform for in vivo CAR T-cell engineering. J. Immunother Cancer. 1110.1136/jitc-2022-006292 (2023). [DOI] [PMC free article] [PubMed]
  • 23.Nicolai, C. J. et al. In vivo CAR T-cell generation in nonhuman primates using lentiviral vectors displaying a multidomain fusion ligand. Blood144, 977–987. 10.1182/blood.2024024523 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Braun, A. H., Frank, A. M., Ho, N. & Buchholz, C. J. Dasatinib is a potent enhancer for CAR T cell generation by CD3-targeted lentiviral vectors. Mol. Ther. Methods Clin. Dev.28, 90–98. 10.1016/j.omtm.2022.12.002 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Morizono, K. et al. Lentiviral vector retargeting to P-glycoprotein on metastatic melanoma through intravenous injection. Nat. Med.11, 346–352. 10.1038/nm1192 (2005). [DOI] [PubMed] [Google Scholar]
  • 26.Ohno, K., Sawai, K., Iijima, Y., Levin, B. & Meruelo, D. Cell-specific targeting of Sindbis virus vectors displaying IgG-binding domains of protein A. Nat. Biotechnol.15, 763–767. 10.1038/nbt0897-763 (1997). [DOI] [PubMed] [Google Scholar]
  • 27.da Aires, F., Costa, M. J., Corte-Real, S. & Goncalves, J. Cell type-specific targeting with Sindbis pseudotyped lentiviral vectors displaying anti-CCR5 single-chain antibodies. Hum. Gene Ther.16, 223–234. 10.1089/hum.2005.16.223 (2005). [DOI] [PubMed] [Google Scholar]
  • 28.Huckaby, J. T. et al. Bispecific binder redirected lentiviral vector enables in vivo engineering of CAR-T cells. J. Immunother Cancer. 910.1136/jitc-2021-002737 (2021). [DOI] [PMC free article] [PubMed]
  • 29.Wu, W. et al. Multiple Signaling Roles of CD3epsilon and Its Application in CAR-T Cell Therapy. Cell 182, 855–871 e823, (2020). 10.1016/j.cell.2020.07.018 [DOI] [PubMed]
  • 30.Baulu, E., Gardet, C., Chuvin, N. & Depil, S. TCR-engineered T cell therapy in solid tumors: state of the Art and perspectives. Sci. Adv.9, eadf3700. 10.1126/sciadv.adf3700 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Kroczek, R. & Hamelmann, E. T-cell costimulatory molecules: optimal targets for the treatment of allergic airway disease with monoclonal antibodies. J. Allergy Clin. Immunol.116, 906–909. 10.1016/j.jaci.2005.07.005 (2005). [DOI] [PubMed] [Google Scholar]
  • 32.Azuma, M. Co-signal molecules in T-Cell activation: historical overview and perspective. Adv. Exp. Med. Biol.1189, 3–23. 10.1007/978-981-32-9717-3_1 (2019). [DOI] [PubMed] [Google Scholar]
  • 33.Loewen, N. & Poeschla, E. M. Lentiviral vectors. Adv. Biochem. Eng. Biotechnol.99, 169–191. 10.1007/10_007 (2005). [DOI] [PubMed] [Google Scholar]
  • 34.Hamilton, M. P. et al. Risk of second tumors and T-Cell lymphoma after CAR T-Cell therapy. N Engl. J. Med.390, 2047–2060. 10.1056/NEJMoa2401361 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Ghilardi, G. et al. T cell lymphoma and secondary primary malignancy risk after commercial CAR T cell therapy. Nat. Med.30, 984–989. 10.1038/s41591-024-02826-w (2024). [DOI] [PubMed] [Google Scholar]
  • 36.Tix, T. et al. Second primary malignancies after CAR T-Cell therapy: A systematic review and Meta-analysis of 5,517 lymphoma and myeloma patients. Clin. Cancer Res.30, 4690–4700. 10.1158/1078-0432.CCR-24-1798 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Schenkwein, D., Afzal, S., Nousiainen, A., Schmidt, M. & Yla-Herttuala, S. Efficient Nuclease-Directed integration of lentivirus vectors into the human ribosomal DNA locus. Mol. Ther.28, 1858–1875. 10.1016/j.ymthe.2020.05.019 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Chen, Y. et al. All-RNA-mediated targeted gene integration in mammalian cells with rationally engineered R2 retrotransposons. Cell 187, 4674–4689 e4618, (2024). 10.1016/j.cell.2024.06.020 [DOI] [PubMed]
  • 39.Colamartino, A. et al. A new specific promoter allow hematopoietic stem cell immunotherapy approach against acute lymphoblastic leukemia using chimeric antigen receptor. Transplantation Cell. Therapy.25(3), S164. 10.1016/j.bbmt.2018.12.464 (2019). [Google Scholar]
  • 40.Grievink, H. W., Luisman, T., Kluft, C., Moerland, M. & Malone, K. E. Comparison of three isolation techniques for human peripheral blood mononuclear cells: cell recovery and viability, population composition, and cell functionality. Biopreserv Biobank. 14, 410–415. 10.1089/bio.2015.0104 (2016). [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Material 1 (1.3MB, docx)

Data Availability Statement

The datasets used and/or analysed during the current study available from the corresponding author on reasonable request.

Articles from Scientific Reports are provided here courtesy of Nature Publishing Group

RESOURCES