Precision targeting of pathogenic B cells and autoantibodies in systemic lupus erythematosus using CAR T cells against the IGHV4-34 B cell receptor

Abstract Introduction: Systemic lupus erythematosus (SLE) is a serious autoimmune disease which chiefly involves B-cell dysregulation and activation, hypergammaglobulinemia, and autoantibody production. Studies have demonstrated that treatment with CART19, a CART targeting CD19, a pan B-cell marker, lead to clinical remissions of SLE. However, broadly targeting B cells with CART19 can lead to B-cell aplasia, hypogammaglobulinemia, and thereby leave patients prone to infections. Of note, in SLE, B cells and autoantibodies expressing the immunoglobulin heavy variable 4-34 (IGHV4-34) gene are highly enriched and associated with SLE severity. Therefore, we hypothesize that CART targeting the IGHV4-34 B-cell receptor (BcR) would preferentially deplete pathogenic B cells without immunosuppression. Methods and Results: To better define the significance of IGHV4-34 BcR as a target, we first aimed to deepen our understanding of how IGHV4-34 immunoglobulins (Ig) contribute to the SLE autoantibody repertoire. We depleted IGHV4-34 Ig from serum and then measured the reduction of specific autoantibodies using a bead-based antigen array. We used sera from SLE patients (n=3), depleted 98-100% of IGHV4-34 Ig, and found a >50% reduction in overall autoantibody levels. Furthermore, we saw reductions of SLE-associated autoantibodies (anti-Ro60, anti-SM, anti-Ribo P0, and anti-DNAse1L3) ranging from 10-100% following IGHV4-34 depletion. Higher levels of IGHV4-34 Ig in the serum have been associated with lupus nephritis (LN). Therefore, we assessed the presence of IGHV4-34+ Ig via immunohistochemistry in 11 randomly selected kidney biopsies from LN patients (5 pediatric LN, 6 adult LN) and 4 controls (1 pediatric C3 glomerulonephritis, 1 pediatric post-streptococcal glomerulonephritis, 2 normal pediatric kidneys). All 13/13 glomerulonephritis biopsies were positive for total IgG, reinforcing the role of immune complexes in nephritis. Remarkably, 10/11 LN biopsies but 0/2 glomerulonephritis controls had IGHV4-34+ antibodies in the affected glomeruli. We, therefore, developed anti-IGHV4-34 CART (CART4-34) using the 9G4 rat monoclonal antibody. Given the large size of the BcR antigen, we optimized CAR:IGHV4-34 immune synapse formation and in vitro and in vivo activity using a shorter CAR hinge domain (G4S hinge [5aa]) instead of the CD8 hinge (44aa). We co-cultured B-cell lines (HBL1, Mec1 WT or IGHV4-34+, Jeko1 WT or IGHV4-34+) with CART4-34 at 0.5-0.25 Effector:Target (E:T) ratios for 48-72hr and observed that CART4-34 exhibits potent activity specifically towards IGHV4-34+ B cells while sparing IGHV4-34-negative B cells (p<0.05). We then developed a model to assess whether CART4-34 could target SLE B cells and reduce autoantibody production. We obtained SLE patient-derived B cells, activated them for 48hr by co-culturing with CpG ODN2006, CD40L, IL2, IL10, and IL15 and differentiated them for 24hr into plasmablasts using IL6, IL2, IL10, and IL15. We then performed a 48hr 0.5 E:T co-culture with CART4-34, CART19, or untransduced T cells. We observed that CART4-34 specifically eliminated IGHV4-34+ B-cells (p=0.005), but not other B cells. Similarly, we measured co-culture total IgG and IGHV4-34+ IgG levels via ELISA and determined that while CART19 significantly depleted all IgG levels (p=0.0004), CART4-34 specifically depleted only IGHV4-34 IgG while simultaneously preserving total IgG. Next, we assessed whether soluble IGHV4-34 Ig could inhibit CART4-34 binding to the target. CART4-34 were co-cultured with HBL1 cells at a 1:1 E:T ratio with 0.1-100ug/ml IGHV4-34+ IgG1 antibody for 72hr. We observed nearly 100% inhibition of CART4-34 function at all Ig concentrations. To overcome this, we hypothesized that Ig depletion, as clinically obtained with plasmapheresis, could restore CART function. Indeed, Ig depletion led to a drastic reduction of the inhibitory effect of the IGHV4-34+ media (p<0.0001). Lastly, to confirm CART4-34 specificity and safety, we conducted next-generation sequencing from normal B-cells cocultured with CART4-34 and controls. CART4-34 led to a statistically significant reduction of only the IGHV4-34 genes out of 46 IGHV genes (p<0.01). Conclusions: These results support our hypothesis that IGHV4-34 Ig plays a role in SLE pathogenesis, particularly in LN. Importantly, CART4-34 can be used to generate potent, targeted, and safe therapy for SLE by targeting IGHV4-34+ B cells, while sparing normal B cells and most of the Ig repertoire.