Affinity-tuned mesothelin CAR T cells demonstrate enhanced targeting specificity and reduced off-tumor toxicity

Our plain-English summary of Affinity-tuned mesothelin CAR T cells demonstrate enhanced targeting specificity and reduced off-tumor toxicity, by Y. Yang, Y. Vedvyas, Y. Alcaina, et al (JCI Insight, 2024).

The study

• Authors: Y. Yang, Y. Vedvyas, Y. Alcaina, S.J. Trumper, D.S. Babu, I.M. Min, J.M. Tremblay, C.B. Shoemaker, M.M. Jin
• Journal: JCI Insight
• Published: 2024;9(22)
• Cited by: 27 articles, as counted by NIH iCite
• Read it: PubMed · Publisher · Free full text

Abstract

The application of chimeric antigen receptor (CAR) T cell therapy in solid tumors is hindered by life-threatening toxicities resulting from on-target, off-tumor killing of nonmalignant cells that express low levels of the target antigen. Mesothelin (MSLN) has been identified as a target antigen for CAR T cell treatment of mesothelioma, lung, ovarian, and other cancers because of its high expression on tumor cells and limited expression on mesothelial cells. However, fatal off-tumor toxicity of high-affinity MSLN-targeting CAR T cells has been reported in multiple clinical trials. In this study, we constructed CARs using mutant variants of a single-domain nanobody that bind both human and mouse MSLN with a wide range of affinities and examined tumor responses and their toxicities from on-target, off-tumor interactions in mouse models. CAR T cells with low nanomolar affinity (equilibrium dissociation constant, KD) exhibited profound systemic expansion with no apparent infiltration into the tumor. With a gradual reduction of CAR affinity toward the micromolar KD, the expansion of CAR T cells became more restricted to tumors. Our preclinical studies demonstrated that high-affinity MSLN CARs were associated with fatal on-target, off-tumor toxicity and that affinity-tuned CARs rendered T cells more selective for MSLN-high tumors.

In plain English

This was a laboratory and mouse study of chimeric antigen receptor (CAR) T cells directed against mesothelin (MSLN), a protein expressed on mesothelioma and other tumours and on normal mesothelial cells. The authors immunised alpacas with human and mouse MSLN, selected a cross-reactive single-domain antibody (JZQ-B4), and made single alanine substitutions in its binding regions to lower binding strength. The N9A variant had above 100 nM KD for human MSLN, roughly a 500-fold drop in affinity from the parent, and was the only variant whose killing in culture tracked with how much MSLN target cells carried. In mice bearing subcutaneous MSTO-211H/hMSLN mesothelioma tumours, the higher-affinity CAR T cells (JZQ-B4, Y4A, T2A and R3A) failed to control tumour growth and caused weight loss requiring euthanasia within 3 weeks; JZQ-B4, Y4A and T2A cells expanded in lungs, liver and bone marrow with minimal tumour infiltration, while R3A cells showed notably less off-tumour expansion, delayed toxicity and tumour infiltration. N9A CAR T cells produced tumour regression with little off-tumour expansion. The authors concluded that lowering CAR affinity made T cells more selective for MSLN-high tumours and reduced fatal on-target, off-tumour toxicity.

We wrote this plain-English summary from the full article with the help of AI, and a second AI model checked each sentence against the article. It is not the authors' wording; the abstract and the article are their own account.

What the study did and found

• Design: Preclinical laboratory study: antibody engineering, in vitro cytotoxicity assays, and experiments in tumour-bearing immunodeficient mice
• Who was studied: Human tumour cell lines (NCI-H226, NCI-H226/KO, MSTO-211H, MSTO-211H/hMSLN), Jurkat and primary human T cells from healthy-donor leukopaks, two immunised alpacas, and male 6- to 8-week-old NOD.Cg-Prkdcscid Il2rgtm1Wjl/SzJ mice (4-8 animals per group across 2 independent experiments)
• Exposure: CAR T cells carrying the MSLN-binding nanobody JZQ-B4 or its alanine variants (T2A, R3A, Y4A, N9A), or an amatuximab-derived scFv CAR (AMAT), given as a single intravenous dose of 1 x 10^7 cells per mouse
• Main results:
  • The N9A CAR had above 100 nM KD for human MSLN, about a 500-fold lower affinity than the parental JZQ-B4 CAR
  • In culture, only the low-affinity N9A CAR T cells killed target cells in proportion to how much MSLN they expressed
  • The higher-affinity variants killed MSLN-low MSTO-211H and MSLN-intermediate NCI-H226 cells almost completely within 24 hours
  • In tumour-bearing mice, the higher-affinity VHH CAR T cells (JZQ-B4, Y4A, T2A, R3A) did not control tumour growth and caused severe toxicity requiring euthanasia within 3 weeks, while N9A CAR T cells caused tumour regression and improved survival relative to the higher-affinity variants
  • PET/CT on day 7 showed off-tumour expansion of JZQ-B4, Y4A and T2A CAR T cells in lungs, liver and bone marrow, with tumour homing seen only in the N9A group
  • R3A CAR T cells expanded outside the tumour notably less than the other higher-affinity variants, and tumour sections from R3A- and N9A-treated mice showed tumour-infiltrating T cells, unlike those from JZQ-B4-, Y4A- and T2A-treated mice
  • In lungs on day 8, over 10% of live single cells were CD3+ T cells in mice given the higher-affinity CARs, whereas N9A-treated mice resembled mice given nontransduced T cells
  • Plasma IFN-gamma and perforin on day 6 were substantially lower in N9A-treated mice than in mice given JZQ-B4, Y4A or T2A CAR T cells
  • The high-affinity AMAT scFv CAR, which does not bind mouse MSLN, eliminated tumours without overt toxicity
• The authors' conclusion: The authors concluded that alanine scanning of nanobody binding regions is an efficient way to tune CAR affinity, that high-affinity MSLN CAR T cells caused fatal on-target, off-tumour toxicity in their mouse model, and that optimising affinity reduced toxicity while improving tumour targeting
• Limitations the authors state:
  • Only alanine substitutions in the complementarity-determining regions were tested; screening other amino acid substitutions might give variants with better affinity and antitumour function
  • Only male mice were used, so the relevance of the findings to female mice was not assessed
  • Survival was followed only to 50 days because mice given human T cells eventually die of xenogenicity, so survival comparisons were restricted to the CAR variant cohorts and benefit over controls had to be judged from tumour size
  • The authors noted that preclinical animal studies have limited ability to capture off-tumour adverse events in humans for conventional human-specific scFv CARs

Topics

Medical terms in this study, explained in our Lexicon: Mesothelin, Mesothelioma, Mesothelial Cells, Clinical Trial.

Related studies

• Efficacy and safety of novel multiple-chain DAP-CAR-T cells targeting mesothelin in ovarian cancer and mesothelioma: a single-arm, open-label and first-in-human study — T. Xu, T. Tian, et al. (2024)
• Membrane-bound TRAIL-armoring augments CAR-T cells in mesothelin-positive solid malignancies — A. Volpe, J. Zurita, et al. (2026)
• JNJ-79032421, a Novel Membrane-Restricted Mesothelin-Targeting T Cell-Engaging Bispecific Antibody for the Treatment of Mesothelin-Positive Cancers — K. Smans, T. Nesspor, et al. (2026)
• Mesothelin is a target of chimeric antigen receptor T cells for treating gastric cancer — J. Lv, R. Zhao, et al. (2019)
• Tumor resistance to anti-mesothelin CAR-T cells caused by binding to shed mesothelin is overcome by targeting a juxtamembrane epitope — X.F. Liu, M. Onda, et al. (2024)
• Development of Highly Effective Anti-Mesothelin hYP218 Chimeric Antigen Receptor T Cells With Increased Tumor Infiltration and Persistence for Treating Solid Tumors — S. Tomar, J. Zhang, et al. (2022)

Browse more research in JCI Insight, or return to medical research.