Affinity-matured CD72-targeting Nanobody CAR T-cells Enhance Elimination of Antigen-Low B-cell Malignancies

Abstract

Background: Chimeric antigen receptor (CAR) T-cell therapies are highly efficacious for several different hematologic cancers. However, for most CAR T targets it is observed that low surface antigen density on tumors can significantly reduce therapeutic efficacy. Here, we explore this dynamic in the context of CD72, a surface antigen we recently found as a promising target for refractory B-cell cancers, but for which CD72 low antigen density can lead to therapeutic resistance in preclinical models.

Methods: Primary samples were accessed via institutional review board-approved protocols. Affinity-matured and humanized nanobody clones were previously described in Temple et al.¹ CAR T-cells were generated via lentiviral transduction. In vitro cytotoxicity assays were performed using luciferase-labeled cell lines. In vivo studies were performed using cell line- or patient-derived xenografts implanted in NOD scid gamma (NSG) mice.

Results: We first confirmed ubiquitous CD72 expression across a range of primary B-cell non-Hodgkin lymphomas. We further found that after resistance to CD19-directed therapies, across both B-cell acute lymphoblastic leukemia (B-ALL) models and primary tumor samples, surface CD72 expression was largely preserved while CD22 expression was significantly diminished. Affinity maturation of a nanobody targeting CD72, when incorporated into chimeric antigen receptor (CAR) T-cells, led to more effective elimination in vitro of isogenic models of CD72 low-expressing tumors. These results suggested that nanobody-based CAR T-cells (nanoCARs) may exhibit a similar relationship between binder affinity, antigen expression, and efficacy as previously demonstrated only for scFv-based CAR T-cells. Surprisingly, however, this significantly improved in vitro efficacy only translated to modest in vivo survival benefit. As a parallel strategy to enhance CAR T function, we found that the small molecule bryostatin could also significantly increase CD72 surface antigen density on B-cell malignancy models. Structural modeling and biochemical analysis identified critical residues improving CD72 antigen recognition of our lead affinity-matured nanobody.

Conclusions: Together, these findings support affinity-matured CD72 nanoCARs as a potential immunotherapy product for CD19-refractory B-cell cancers. Our results also suggest that for B-ALL in particular, CD72 may be a preferable second-line immunotherapy target over CD22.

Keywords: Affinity maturation; CD72; Chimeric Antigen Receptor; Immunotherapy; Leukemia; Lymphoma; Nanobody.

Figure 1.. Transcriptome analysis and flow cytometry confirm CD72 is a B-cell specific target ubiquitously present in B-cell malignancies and retained after CD19 CAR T relapse.

A) CD72 transcript expression from microarray profiling in healthy B-cells and B-NHL lymphoma cells. GSE12453: DLBCL (n=11), FL (n=5) and BL (n=5). B-D) Kaplan–Meier plots of overall survival (OS) and relapse-free survival (RFS) in B-cell lymphoma patient cohorts selected by CD72 expression: top 50% expression in red and bottom 50% expression in green. B) GSE10846, DLBCL, treated with R-CHOP, n=233, p=0.0382 by Log-Rank test. C) GSE23967, ABC-DLBCL, treated with R-CHOP, n=20, p=0.0388 by Log-Rank test. D) GSE22762, CLL, n=107, p=0.00124 by Log-Rank test. E) Flow cytometry evaluation of CD72 expression in primary B-NHL lymphoma patient tumor samples from Kanazawa University, Japan (top panel, isotype control was used as negative control) and Hospital Universitario 12 de Octubre, Madrid, Spain (bottom panel, fluorescence minus one was used as negative control). F) Flow cytometry of B-NHL lymphoma tumor cell lines with a broad range of CD72 expression. G) Flow cytometry of two pediatric B-ALL patients profiled for expression of CD19, CD22 and CD72 before and after CD19 CAR T therapy relapse. B-ALL blasts CD10+/CD45 low+ (Patient 1) and CD10+/CD45 low+/− (Patient 2).

Figure 2.. Affinity matured CD72 CAR-T-cells show increased tumor control in vitro and in vivo against B-NHL lymphoma models compared to H24 CAR.

A) Amino acid sequences of anti-CD72 nanobody used in CAR T designs, including parental NbD4 (originally described in ref.), humanized H24, affinity matured NbD4.1, NbD4.3, NbD4.7, and NbD4.13 (originally described in ref.). Complementarity determining regions (CDRs) highlighted. B) In vitro 48-hour luciferase-based cytotoxicity assay of CD72 CAR Ts versus lymphoma cell lines Namalwa and Toledo. Data normalized to untransduced T-cells. n=3 technical replicates, performed at noted E:T ratios. C-D) Kinetics of CD72 CAR-T cytotoxicity and proliferation versus mCherry-labeled JeKo-1 MCL cell line by Incucyte live cell imaging. Data normalized to time zero of each well. n=3 technical replicates. Performed at 1:3 Effector:Tumor (E:T) ratio. E-F) Luciferase-labeled JeKo-1 (1e6 tumor cells) implanted intravenously in NSG mice (n=5 per group) and treated with 3.5e6 CAR T-cells 7 days after. Bioluminescence images (E) and quantified average radiance (F) plot of JeKo-1 tumor in mice treated with Empty CAR (negative control), CD19 CAR, H24, NbD4.7 or NbD4.13 CAR. G) Quantification of blood CAR T expansion in murine peripheral blood over time. CAR Ts quantified based on human CD3+/GFP+ cells. Statistical analysis was performed by two-way ANOVA with Tukey’s multiple comparisons test. H) Kaplan–Meier curve of overall survival in JeKo-1 study. Statistical analysis was performed by log-rank (Mantel-Cox) test. (*p<0.05, **p<0.01). I) Murine study with 1e6 mantle cell lymphoma PDX DFBL44685 cells injected into mice intravenously and subsequently intravenously implanted with 5e6 CAR Ts 10 days after. Murine tumor burden assessed by spleen ultrasound at 7 day intervals. J) CAR T expansion in murine blood on day 51 after PDX tumor injection. K) Survival analysis of high tumor burden stress test with 2e6 PDX cells implanted intravenously and treated with CAR-Ts 14 days after. Kaplan–Meier curve of overall survival. Statistical analysis was performed by log-rank (Mantel-Cox) test. L) Spleen volume, M) tumor burden in blood and N) CAR T percentage in blood at 6 weeks after tumor injection. Statistical analysis in I, J, L-N) was performed by Student’s t-test. (ns p>0.05, *p<0.05, **p<0.01, ***p≤0.001, ****p≤0.0001). Independent T-cell donors were used for studies in (B), (C-D), (F-H), and (I-N).

Figure 3.. Affinity matured CAR-Ts have superior effector function to H24 in vitro versus Jeko1-CD72 low model.

A) JeKo-1 tumor cells were isolated from mice implanted with WT tumor and relapsed after H24 CAR therapy (from our prior study; previously shown to have decreased CD72 expression at relapse). These mCherry-labeled JeKo-1 CD72lo^r cells were co-cultured with CAR-Ts at 1:3 E:T in Incucyte live cell analyzer. The bar graph represents cytotoxicity at 36 hours of co-culture. Data analyzed by unpaired t-test. B) Flow cytometry plots showing CD72 expression in JeKo-1 WT, JeKo-1 CD72 KO and engineered JeKo-1 CD72-lo^oe model. C-D) In vitro 24-hour luciferase-based cytotoxicity assay of Empty, CD19, H24, NbD4.13 CAR Ts versus JeKo-1 WT (C) and JeKo-1 CD72 lo^oe (D) models. Normalized to tumor cells, n=3 technical replicates. E) In vitro 24-hour luciferase-based cytotoxicity assay of Empty, CD19, H24, NbD4.13 and NbD4.13-H24 CAR-Ts versus JeKo-1 CD72lo^oe model. Normalized to untransduced T-cells, n=3 technical replicates. Statistical analysis in D) and E) was performed by two-way ANOVA with Tukey’s multiple comparisons test (**p<0.01, ***p≤0.001, ****p≤0.0001). Performed with CAR T-cells generated from an independent donor from (C-D). F) Cytokine profiling of CD72 CAR T supernatant by Luminex assay (see Methods) after 24hr exposure to JeKo-1 WT and JeKo1-lo^oe tumor models at 1:1 E:T. Statistical analysis of data is performed by ordinary one-way ANOVA (ns p>0.05, *p<0.05, **p<0.01, ***p ≤ 0.001, ****p ≤ 0.0001). G) Cytotoxicity kinetics of affinity matured (NbD4.13), affinity matured+humanized (NbD4.13-H24), and humanized (H24) CD72 CAR-Ts with repetitive stimulation with JeKo-1 CD72 lo^oe tumor (mCherry-labeled), 2 and 5 exposures to tumor. Data obtained from Incucyte live cell analyzer. Data normalized to time zero of each well. n=3 technical replicates. Performed with CAR T-cells generated from an independent donor from (C-F).

Figure 4.. Affinity matured and humanized CAR-Ts have a modest survival benefit versus JeKo-1 low model.

NSG mice were implanted with 0.5e6 JeKo-1 CD72 lo^oe tumor cells through tail vein injection and treated with 5e6 CAR-T-cells 7 days after (n=5 per arm). A) Tumor burden assessed by weekly bioluminescence imaging. B) Quantified BLI data over time. Data analyzed by multiple unpaired t-test. (ns p>0.05, **p<0.01, ****p≤0.0001). C) Peripheral CAR-T expansion on day 9 after CAR-T injection. Statistical analysis by ordinary one-way ANOVA. D) Kaplan–Meier curve of overall survival. Statistical analysis was performed by log-rank (Mantel-Cox) test (ns p>0.05, *p<0.05).

Figure 5.. Bryostatin is an antigen modulator of B-cell associated targets.

A) SC-1 or DOHH-2 cells treated with 0.1% DMSO or increasing concentrations of bryostatin. Surface antigen density of CD72, CD19, CD22 profiled by flow cytometry and normalized to DMSO treated cells. B) SC-1 untreated cells and cells pre-treated with 1 nM of bryostain for 72 hr were then co-cultured with CAR Ts in a 24 hr luciferase-based cytotoxicity assay at 1:1, 1:3 and 1:10 E:T ratios. C) Nalm-6 WT and CD19 therapy relapsed-models (Nalm-6 CD19-R) were treated with 0.1% DMSO or increasing concentrations of bryostatin. Surface antigen density of CD72, CD19, and CD22 profiled by flow cytometry and normalized to DMSO treated cells. Statistical analysis in A) and C) was performed by ordinary one-way ANOVA with Dunnett’s multiple comparisons test (ns p>0.05, *p<0.05, **p<0.01, ***p ≤ 0.001, ****p ≤ 0.0001). D) Pathway analysis of bulk RNA-seq of bryostatin treated SC-1 cells demonstrates 591 upregulated genes and 204 downregulated genes (FDR<0.05). B-cell activation and signaling are among the top pathways represented among upregulated genes.

Figure 6.. Structural modeling and biolayer interferometry studies identify critical residues in the nanobody-CD72 interaction.

A) AlphaFold and HADDOCK modeling predicts NbD4 and NbD4.13 bind a similar epitope on the CD72 monomer: junction of CD72 coiled-coil and the lectin-like globular domain (amino acids 210–250). Only one of the three affinity maturation mutations of NbD4.13 is predicted to improve binding to CD72 (CDR2, Ala52Phe), leading to a favorable interaction of aromatic ring of NbD4.13 (Phe 52) with the proline ring (Pro 222) and adjacent aromatic phenylalanine (Phe 223) of CD72. B) Biolayer interferometry plots demonstrating NbD4, H24, NbD4.13 and NbD4.13 F52A nanobody binding to the biotinylated extracellular domain of CD72 (30 nM). Data are representative of a single experiment performed at multiple concentrations of input nanobody (0,0.625, 1.25, 2.5, 5, 10 μM). The table below illustrates fold-change in affinity (as measured by ratio of K_D’s) for each combination of CD72 and tested in nanobody, all benchmarked to the parental NbD4 clone binding to CD72 WT extracellular domain. Higher values indicate increased affinity compared to NbD4 binding to CD72 WT; lower values indicate decreased affinity.