Anti-BCMA chimeric antigen receptors with fully human heavy-chain-only antigen recognition domains

Abstract

Chimeric antigen receptor (CAR)-expressing T cells targeting B-cell maturation antigen (BCMA) have activity against multiple myeloma, but improvements in anti-BCMA CARs are needed. We demonstrated recipient anti-CAR T-cell responses against a murine single-chain variable fragment (scFv) used clinically in anti-BCMA CARs. To bypass potential anti-CAR immunogenicity and to reduce CAR binding domain size, here we designed CARs with antigen-recognition domains consisting of only a fully human heavy-chain variable domain without a light-chain domain. A CAR designated FHVH33-CD8BBZ contains a fully human heavy-chain variable domain (FHVH) plus 4-1BB and CD3ζ domains. T cells expressing FHVH33-CD8BBZ exhibit similar cytokine release, degranulation, and mouse tumor eradication as a CAR that is identical except for substitution of a scFv for FHVH33. Inclusion of 4-1BB is critical for reducing activation-induced cell death and promoting survival of T cells expressing FHVH33-containing CARs. Our results indicate that heavy-chain-only anti-BCMA CARs are suitable for evaluation in a clinical trial.

Fig. 1. A murine anti-BCMA CAR can be immunogenic.

a 11D5-3-NS, a truncated nonsignaling CAR containing only the murine 11D5-3 scFv, hinge, and transmembrane regions was designed. Irradiated, autologous 11D5-3-NS-transduced T cells were used to stimulate PBMC in culture. PBMC were from a patient who received 11D5-3-CD828Z CAR T cells on a clinical trial. Seven days later, the PBMC were stimulated again with 11D5-3-NS-transduced T cells. Seven days after the second stimulation, the PBMC were cultured overnight with autologous T cells that were either untransduced, transduced with the human NGFR gene, or transduced with the 11D5-3-NS gene. Culture supernatants were assayed for IFNγ by ELISA. 11D5-3-NS-specific release of IFNγ was found. b PBMC collected after CAR T-cell infusion to a different patient than in a were stimulated with autologous 11D5-3-CD828Z CAR⁺ T cells as in a. Peptide reactivity was assessed by culturing the stimulated PBMC for 6 h with autologous dendritic cells pulsed with 15-mer peptides of a peptide library covering all possible 15 mers of the 11D5-3 scFv. Specific IFNγ production by T cells was found in an ICCS assay against peptide pool 5 and peptide 59 from pool 5. c A diagram of the FHVH33-CD8BBZ CAR with the fully human heavy-chain binding domain FHVH33, hinge and transmembrane domains from human CD8α, a human 4-1BB domain, and a human CD3ζ domain. d 11D5-3-CD8BBZ has a murine scFv binding domain. Otherwise, 11D5-3-CD8BBZ has an identical sequence as FHVH33-CD8BBZ. e Except for substitution of CD28 for 4-1BB, FHVH33-CD828Z is identical to FHVH33-CD8BBZ.

Fig. 2. Murine scFv 11D5-3 versus fully human heavy-chain domain FHVH33.

a Flow cytometry results of T cells from the same donor transduced with 11D5-3-CD8BBZ or FHVH33-CD8BBZ or left untransduced are shown; cells were stained with BCMA-Fc-PE. b, d–h show mean ± s.e.m. 11D5-3 means 11D5-3-CD8BBZ. FHVH33 means FHVH33-CD8BBZ. All comparisons are two-tailed, paired t-tests. P < 0.05 was considered statistically significant. N.S. is not statistically significant. b Median fluorescence intensity of CD3⁺, BCMA-Fc-PE⁺ cells expressing 11D5-3-CD8BBZ or FHVH33-CD8BBZ is shown (n = 7, P = N.S.). c Relative affinity of 11D5-3-CD8BBZ versus FHVH33-CD8BBZ was determined by staining CAR-expressing T cells with decreasing concentrations of BCMA-Fc-PE and performing flow cytometry. Y-axis is percent maximum specific binding (% of Bmax). d Relative K_D values were determined by nonlinear regression from binding curves of 11D5-3-CD8BBZ-expressing T cells and FHVH33-CD8BBZ-expressing T cells. K_D values were calculated based on the concentration of BCMA-Fc-PE yielding half-maximal binding (n = 7; P = N.S). T cells were transduced with either 11D5-3-CD8BBZ or FHVH33-CD8BBZ and stimulated in culture for 4 h; degranulation of CD4⁺ and CD8⁺ T cells was assessed by CD107a expression. T cells were stimulated with either BCMA⁺ C17-BCMA⁻K562 cells (e and f, n = 4) or BCMA⁺ RPMI8226 cells (g and h, n = 5). CD4⁺ or CD8⁺ %CD107a⁺ events are from flow cytometry plots gated on live CD3⁺ cells. Background degranulation with BCMA-negative NGFR-K562 cell stimulation was subtracted from degranulation with BCMA⁺ cell stimulation. Results were normalized for CAR expression. i T cells expressing 11D5-3-CD8BBZ, FHVH33-CD8BBZ, or negative-control CAR SP6-CD828Z were tested in a 4-h cytotoxicity assay (one of similar experiments). Points represent mean cytotoxicity of replicate wells ±s.e.m.

Fig. 3. T cells expressing either 11D5-3-CD8BBZ or FHVH33-CD8BBZ specifically recognized BCMA and primary multiple myeloma cells.

a T cells from the same donor were transduced with either 11D5-3-CD8BBZ or FHVH33-CD8BBZ or were left untransduced. T cells were cultured overnight with the indicated target cells, and an IFNγ ELISA was performed on the culture supernatant. Bars represent the means of duplicate wells. Bars representing cultures including untransduced T cells and cultures with BCMA-negative target cells are not visible because the values are too small. Similar results were obtained with cells from four different donors. BCMA-K562, RPMI8226, and MM.1 S are BCMA⁺; other target cells are BCMA⁻ (b–d) RPMI8226 target cells were cultured overnight with either 11D5-3-CD8BBZ (11D5-3) T cells or FHVH33-CD8BBZ (FHVH33) T cells. Next, an ELISA assay was performed to measure IFNγ, TNF, or IL-2 in the culture supernatant. The amount of nonspecific cytokine release in response to the BCMA-negative cell line NGFR-K562 in parallel cultures was subtracted from the cytokine release in response to RPMI8226 to obtain the RPMI8226-specific IFNγ release. Results are normalized for CAR expression. Mean  ±  s.e.m. are shown; n = 4. P = N.S. (not statistically significant) by paired two-tailed t-tests for IFNγ and IL-2; P = 0.018 for TNF. e Patient bone marrow cells that consisted of 56% BCMA⁺ multiple myeloma cells were cultured for 4 h with autologous T cells that were untransduced (UT) or transduced with the SP6-CD828Z negative-control CAR, 11D5-3-CD8BBZ, or FHVH33-CD8BBZ. Cells were stained for CD107a to detect degranulation of CD4⁺ T cells (top row) or CD8⁺ T cells (bottom row). Plots are gated on CD3⁺ live lymphocytes.

Fig. 4. FHVH33 CARs with either a CD28 or 4-1BB costimulatory domain have similar in vitro function.

a T cells from ten donors were transduced with either FHVH33-CD828Z (28) or FHVH33-CD8BBZ (BB). Four or 5 days after transduction, T cells were stained with anti-CD3 and BCMA-Fc-PE to detect CAR-expressing T cells. The mean ± s.e.m. of the percentages of T cells expressing each CAR is shown. All statistical comparisons in this figure were by paired two-tailed t-test. P < 0.05 was considered statistically significant. b T cells expressing either FHVH33-CD828Z or FHVH33-CD8BBZ or left untransduced were cultured overnight with the target cells indicated, and IFNγ was measured in the culture supernatant by ELISA. BCMA-K562 and RPMI8226 were BCMA⁺; the other cell lines were BCMA⁻. Bars represent the means of duplicate wells. The bars representing cultures including untransduced T cells and cultures with BCMA-negative target cells are not visible because the values are too small. Seven experiments with similar results were conducted. c T cells expressing either FHVH33-CD828Z, FHVH33-CD8BBZ, or the control CAR SP6-CD828Z were tested in a 4-h cytotoxicity assay. Two experiments with similar results were conducted. Symbols represent means of duplicate wells  ±s.e.m. Degranulation of CD4⁺ d or CD8⁺ e T cells was measured by detecting CD107a upregulation with flow cytometry (n = 6). The percentages of T cells with CD107a upregulation after a 4-h culture with RPMI8226 cells minus background CD107a upregulation after culture with NGFR-K562 cells are shown. Release of f IFNγ, g TNF, and h IL-2 after overnight culture with RPMI8226 cells minus background cytokine production after culture with NGFR-K562 cells is shown (n = 7 for IFNγ and TNF, n = 5 for IL-2). For d–h, the mean ± s.e.m. are shown; results are normalized for CAR expression; N.S, not statistically significant. P < 0.05 was considered statistically significant.

Fig. 5. 4-1BB versus CD28 CARs: proliferation and survival.

a T cells expressing FHVH33-CD828Z (28) or FHVH33-CD8BBZ (BB) were labeled with CFSE and cultured with irradiated BCMA-K562 cells or NGFR-K562 cells. Changes in CAR⁺ T-cell numbers during the 4-day culture are shown (n = 6). All bar graphs in this figure show mean ± s.e.m.; all statistics are paired two-tailed t-tests, P < 0.05 was considered statistically significant. b BCMA-specific proliferation was represented by dividing the CFSE MFI of T cells stimulated with BCMA-K562 by the CFSE MFI of T cells stimulated with NGFR-K562. BCMA-specific CFSE dilution and proliferation were greater with FHVH33-CD828Z T cells (28) than FHVH33-CD8BBZ (BB) T cells (n = 6). c T cells expressing FHVH33-CD828Z (28) or FHVH33-CD8BBZ (BB) were cultured overnight with either BCMA-K562 cells or NGFR-K562 cells and stained with annexin V to detect apoptosis. As a measure of BCMA-specific apoptosis, the %annexin V⁺ CAR⁺ T cells was calculated as the percentage annexin V⁺ CAR⁺ T cells after BCMA-K562 stimulation minus the percentage annexin V⁺ CAR⁺ T cells after NGFR-K562 stimulation. The percentages of annexin V⁺ CAR⁺ cells were higher for FHVH33-CD828Z versus FHVH33-CD8BBZ for CD4⁺ (P = 0.017) and CD8⁺ T cells ⁽P = 0.007); n = 5. d T cells expressing FHVH33-CD828Z (28) or FHVH33-CD8BBZ (BB) were stimulated with BCMA-K562 cells. CAR⁺ T cells were quantified at the beginning of culture and 7 days later. Changes in CAR⁺ T-cell numbers between initiation and day 7 of culture are shown (n = 4). e After the culture described in d, T cells were stained for annexin V. Mean ± s.e.m. %annexin V⁺ CAR⁺ T cells is shown (n = 4). f Mean ± s.e.m. of CD4:CD8 ratios of CFSE-labeled 28 and BB CAR⁺ T cells after the 4-day culture from a are shown (n = 6).

Fig. 6. CAR T-cell studies in mice.

a NSG mice were injected intravenously with MM.1 S cells. After 10 days, 1 × 10⁶ T cells expressing the indicated CARs were injected intravenously. Mice were imaged every 3 days. Two experiments of five mice each were completed with similar results with cells from different donors. b, c RPMI8226 cells were injected intradermally into NSG mice. After palpable tumors were established, mice were injected intravenously with b 2 × 10⁶ or c 1 × 10⁶ T cells expressing the indicated CARs. Left side graphs show the mean tumor volume of five mice/group. The right side graphs show Kaplan–Meier plots of survival of the same mice. For the 2 × 10⁶ CAR T-cell dose, there was a statistically significant difference in survival between the T cells expressing the SP6-CD828Z negative-control CAR and 11D5-3-CD8BBZ, FHVH33-CD828Z, and FHVH33-CD8BBZ (P = 0.003 for all three comparisons). For the 1 × 10⁶ CAR T-cell dose, there was a statistically significant difference in survival between the T cells expressing the SP6-CD828Z negative-control CAR and 11D5-3-CD8BBZ, FHVH33-CD828Z, and FHVH33-CD8BBZ (P = 0.002 for all three comparisons). Two experiments of five mice/group each were completed with T cells from different donors for each experiment with 2 × 10⁶ CAR⁺ T cells/mouse. d CD3⁺CD4⁺CAR⁺ and e CD3⁺CD8⁺CAR⁺ splenocytes were quantitated by flow cytometry 10 days after infusion of FHVH33-CD828Z (28) or FHVH33-CD8BBZ (BB) T cells. The NSG mice had disseminated MM.1 S tumor cells established prior to intravenous CAR T-cell infusion. Comparison by Mann–Whitney test. P = 0.0011 for the CD4 comparison, and P = 0.0030 for the CD8 comparison. Bars represent medians; n = 8 mice per group.