CAR-T in Autoimmunity: Considerations for Next-Generation Allogeneic CAR-T Design

Editor: Dr. Lumeng Ye

January 30, 2025

In July 2024, Cell published a clinical study of an allogeneic CAR-T therapy named TyU19 (NCT05859997) for treating autoimmune diseases. The study was led by Chinese medical and biotech teams [1]. The therapeutic T cells were engineered with a CAR targeting the CD19 antigen on B cells, coupled with a human leukocyte antigen CD3 knockout. Since these CD19 CAR-T cells were derived from healthy donor T cells, the study highlighted the potential of allogeneic CAR-T therapies for autoimmune conditions.

Less than three months later, the Chinese biotech company BRL Medicine announced IND approval in China by the CDE (Center for Drug Evaluation, NMPA) for its pipeline BRL-203 program for treating systemic lupus erythematosus (SLE). Notably, BRL-203 is the world’s first non-viral gene-edited CAR-T therapy for autoimmune disease and demonstrated excellent safety and efficacy in a prior trial for treating relapsed/refractory non-Hodgkin B cell lymphoma (BRL-201) [2,3].

“The data should be a boon to biotechs currently developing allogeneic CAR-T cell autoimmune therapies of their own,” Fierce Biotech commented in July. Since then, BRL Medicine has completed a Series B+ funding round, while Poseida Therapeutics secured a $1.5 billion acquisition deal with Roche. Clearly, we are witnessing “CAR-Ts sweep into autoimmunity” [4].

Why Non-Viral Gene Editing for Allogeneic CAR-T?

Allogeneic CAR-T therapy engineered via non-viral CRISPR gene editing offers several advantages: precise CAR gene knock-in to control location, copy number, and expression level; enhanced safety and quality through engineered processes; and simultaneous multiple gene knockouts to reduce host immune response triggers. This one-step, non-viral manufacturing process can also shorten turnaround time and lower overall therapy costs.

Large-fragment knock-in using CRISPR double-stranded breaks (DSBs) and homology-directed repair (HDR) is commonly employed in cell therapy development. Balancing editing efficiency and cell viability requires optimizing the format and quality of DNA payloads to achieve high CAR+ cell proportions and maintain cell fitness after editing and manufacturing.

Evaluating DNA Payload Formats for T Cell Engineering

To compare different DNA payloads, we tested GFP insertion at the TRAC exon 1 in primary T cells as part of our CRISPR CAR-T Knock-In Optimization Kit. We also knocked out B2M and PDCD1 genes to eliminate MHC-I and PD-1 surface expression, creating an allogeneic universal CAR-T model. We evaluated eight DNA payload formats encoding GFP with 300 bp homology arms flanking the TRAC DSB site:

• Linear ssDNA (GenExact™ ssDNA, or lssDNA)
• Classic 5’-ssCTS with annealing oligo (ssCTS) (Shy et al. [5])
• 5’-ssCTS with hairpin structure (5'CTS hairpin)
• Circular single-stranded DNA (cssDNA)
• Open-ended linear dsDNA by PCR (open-ends dsDNA)
• Closed-ended linear dsDNA (GenWand™ dsDNA)
• Supercoiled circular dsDNA with minimized backbone (GenCircle™)
• Classic pUC57 plasmid

Figure 1. Illustration of eight different DNA payloads formats used as HDR templates

We purified all DNA payloads using the same protocol to remove enzyme/host proteins, endotoxins, inorganic salts, and solvents, ensuring comparable purity for T cell engineering via electroporation. As shown in Figure 2 below, DNA resuspended in different buffers exhibited variable band patterns during agarose gel electrophoresis due to dimer and aggregate formation.

Figure 2. Agarose gel electrophoresis results of DNA payloads resuspended in electroporation buffer vs. nuclease-free water

To achieve GFP knock-in at the TRAC locus, we electroporated T cells with CRISPR RNP complexes comprising eSpCas9 protein and sgRNA, using payloads resuspended in nuclease-free water. Detailed protocols for CTS annealing, electroporation using the Thermo Fisher Neon™ platform, and DNA payload preparation are available here: CRISPR Knock-in Comprehensive Guide.

Key Findings

• dsDNA templates exhibited lower editing efficiency than ssDNA, likely due to cytotoxicity from cGAS-STING pathway activation during electroporation [6], as well as the short insert length.
• Linear dsDNA showed slightly higher efficiency than circular dsDNA, but the difference was not significant.
• ssDNA, whether linear or circular, achieved comparable editing efficiency.

Note:
Our internal data show that circular ssDNA provides better editing efficiency at low input amounts (e.g., 1 μg per million T cells) due to enhanced stability, but causes increased cytotoxicity and reduced editing efficiency at higher inputs. Thus, the amount of cssDNA used was 25% of that of linear ssDNA.

Figure 3. Editing efficiency of GFP insertion at TRAC by CRISPR DSB and HDR, with and without two more genes modified.

Linear ssDNA with a 5’-CTS sequence showed high editing efficiency due to low cytotoxicity, improved nuclear delivery, and enhanced stability from Cas9 RNP protection. However, replacing the CTS partial dsDNA structure with a self-annealed hairpin reduced efficiency, despite comparable Cas9 RNP binding (internal data). This remains a technical puzzle requiring further investigation.

Interestingly, multiple gene editing yielded higher knock-in efficiency than single-gene editing, potentially due to heightened DNA repair responses triggered by simultaneous DSBs. However, this could also increase the risk of chromosomal rearrangements or off-target insertions. We plan to share nanopore sequencing results soon to address these critical safety considerations for allogeneic universal CAR-T engineering.

Stay tuned for updates!

Learn more about our HDR Knock-in Templates here.

Note from the editor: If you have any questions about this blog or other interesting topic suggestions related to gene editing or gene and cell therapy for future posts, please reach out at lumeng.ye@genscript.com.

References

• [1] Wang X, Wu X, Tan B, Zhu L, Zhang Y, Lin L, Xiao Y, Sun A, Wan X, Liu S, Liu Y, Ta N, Zhang H, Song J, Li T, Zhou L, Yin J, Ye L, Lu H, Hong J, Cheng H, Wang P, Li W, Chen J, Zhang J, Luo J, Huang M, Guo L, Pan X, Jin Y, Ye W, Dai L, Zhu J, Sun L, Zheng B, Li D, He Y, Liu M, Wu H, Du B, Xu H. Allogeneic CD19-targeted CAR-T therapy in patients with severe myositis and systemic sclerosis. Cell. 2024 Sep 5;187(18):4890-4904.e9. doi: 10.1016/j.cell.2024.06.027. Epub 2024 Jul 15. PMID: 39013470.
• [2] Zhang J, Hu Y, Yang J, Li W, Zhang M, Wang Q, Zhang L, Wei G, Tian Y, Zhao K, Chen A, Tan B, Cui J, Li D, Li Y, Qi Y, Wang D, Wu Y, Li D, Du B, Liu M, Huang H. Non-viral, specifically targeted CAR-T cells achieve high safety and efficacy in B-NHL. Nature. 2022 Sep;609(7926):369-374. doi: 10.1038/s41586-022-05140-y. Epub 2022 Aug 31. PMID: 36045296; PMCID: PMC9452296.
• [3] Hu Y, Zu C, Zhang M, Wei G, Li W, Fu S, Hong R, Zhou L, Wu W, Cui J, Wang D, Du B, Liu M, Zhang J, Huang H. Safety and efficacy of CRISPR-based non-viral PD1 locus specifically integrated anti-CD19 CAR-T cells in patients with relapsed or refractory Non-Hodgkin's lymphoma: a first-in-human phase I study. EClinicalMedicine. 2023 May 18; 60:102010. doi: 10.1016/j.eclinm.2023.102010. PMID: 37251628; PMCID: PMC10209187.
• [4] Harrison C. CAR-Ts sweep into autoimmunity. Nat Biotechnol. 2024 Jul;42(7):995-997. doi: 10.1038/s41587-024-02321-0. PMID: 38987661.
• [5] Shy BR, Vykunta VS, Ha A, Talbot A, Roth TL, Nguyen DN, Pfeifer WG, Chen YY, Blaeschke F, Shifrut E, Vedova S, Mamedov MR, Chung JJ, Li H, Yu R, Wu D, Wolf J, Martin TG, Castro CE, Ye L, Esensten JH, Eyquem J, Marson A. High-yield genome engineering in primary cells using a hybrid ssDNA repair template and small-molecule cocktails. Nat Biotechnol. 2023 Apr;41(4):521-531. doi: 10.1038/s41587-022-01418-8. Epub 2022 Aug 25. PMID: 36008610; PMCID: PMC10065198.
• [6] An J, Zhang CP, Qiu HY, Zhang HX, Chen QB, Zhang YM, Lei XL, Zhang CX, Yin H, Zhang Y. Enhancement of the viability of T cells electroporated with DNA via osmotic dampening of the DNA-sensing cGAS-STING pathway. Nat Biomed Eng. 2024 Feb;8(2):149-164. doi: 10.1038/s41551-023-01073-7. Epub 2023 Jul 27. PMID: 37500747.

Do you find this article helpful?