Anti-CD21 Chimeric Antigen Receptor T cells for the Treatment of T Cell Acute Lymphoblastic Leukemia

doi:10.21203/rs.3.rs-2227661/v1

Download PDF

Article

Anti-CD21 Chimeric Antigen Receptor T cells for the Treatment of T Cell Acute Lymphoblastic Leukemia

https://doi.org/10.21203/rs.3.rs-2227661/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Relapsed/refractory (R/r) T cell acute lymphoblastic leukemia (T-ALL) has a dismal prognosis, with an unmet need for effective novel therapies. The successes seen in chimeric antigen receptor (CAR)-T cell therapy for B-ALL have yet to be fully translated to T-ALL. Most strategies have targeted pan-T antigens (CD7, CD5) but these may be limited by T cell aplasia and fratricide, requiring elimination of CAR-T antigen expression during manufacture and salvage hematopoietic stem cell transplantation (HSCT).

Here, we describe CD21 as a novel immunotherapeutic strategy for the treatment of T-ALL. CD21 is largely confined to malignant T cells with expression in 57% of diagnostic T-ALL but only on a minor fraction of mature T cells (10%). While anti-CD21 CAR-T targeting membrane distal epitopes were ineffective, CAR-T cells utilising a novel Fab-CAR architecture and binding to membrane proximal epitopes showed no fratricide and were potent against low antigen density cell line and patient-derived xenograft models of T-ALL in vitro and in vivo. Further, we showed that CD21 expression in T-ALL can be upregulated by inhibition of the PI3K/ axis.

CD21 is a novel target for CAR-T cell therapy in T-ALL, avoiding the fratricide and T cell aplasia seen with many T-ALL CAR-T strategies.

Biological sciences/Cancer/Haematological cancer/Leukaemia/Acute lymphocytic leukaemia

Biological sciences/Immunology/Immunotherapy

Chimeric antigen receptor (CAR) T cell therapy

CD21

T-ALL

antigen density manipulation

CD21 is expressed in T-ALL but is only expressed on a minority of normal T cells
Anti-CD21 CAR-T cells targeting membrane-proximal epitopes are highly effective against models of T-ALL in vitro and in vivo

T cell acute lymphoblastic leukemia (T-ALL) is an aggressive malignancy of T cell precursors, accounting for ~ 15% of ALL diagnoses in children and 25% in adults.(1, 2) Outcomes have historically been poor, but advances in treatment have brought them in line with B-ALL.(3, 4) However, the prognosis for relapsed or refractory (r/r) disease is dismal, with only 6.5% long-term survival in adults and < 25% in children.(2, 5)

CAR-T cell therapy is useful in r/r B-ALL, inducing sustained responses in a proportion of patients; early studies suggest CAR-T cell therapies may also be useful in T-ALL. However, CAR-T cell targeting of T-ALL is more complex: analogous approaches to CD19 CAR-T cell therapy, where T cell lineage antigens (e.g. CD5, CD7)(6–8) are targeted carry risks of fratricide and T cell aplasia, requiring additional engineering and possible rescue by HSCT. Targeting antigens only expressed on T-ALL blasts, without expression on normal T cells (e.g. TRBC1, CD1a, CCR9)(9–11), is a more desirable strategy. However, this latter approach is limited since only a proportion of T-ALL cases express target-selective antigens. Hence, identification of additional targets is required to make CAR-T cell therapies as broadly available as possible.

CD21 (CR2, complement receptor 2) is a type 1 transmembrane glycoprotein, expressed on mature B lymphocytes and follicular dendritic cells (FDCs), where it binds complement degradation fragments within the B cell co-receptor complex alongside CD19, CD81 and Leu-13.(12)(13) Importantly, it has no known expression on non-immune tissues.

Here, we explore CD21 as a novel CAR-T target for T-ALL and demonstrate that CD21 is expressed on a high proportion of T-ALL cases but only 10% of normal T cells. CD21 is a complex CAR-T target due to its large (15/16 sushi repeat), flexible and highly glycosylated extracellular domain. Despite this, we show that anti-CD21 CAR-T cells targeting membrane-proximal epitopes using a Fab (fragment of antigen binding) antigen binding domain show robust in vitro and in vivo anti-tumour activity against multiple T-ALL disease models. Further, we show that efficacy can be enhanced through PI3K-mediated pharmacological upregulation of CD21 antigen density.

Cell lines and maintenance

The HEK-293T cell line was cultured in IMDM (Lonza) with 10% FBS (HyClone, GE) and 2mM GlutaMAX (Invitrogen). All others were cultured in RPMI (Lonza) with 10% FBS and 2mM GlutaMAX. All were routinely tested for mycoplasma and were obtained from the Deutsche Sammlung von Mikroorganismen und Zellkulturen (DSMZ).

Flow cytometry

UCL Research Ethics Committee granted ethical approval. Whole blood was obtained from consenting healthy donors or leucocyte cones from NHS Blood and Transplant. Peripheral blood mononuclear cells (PBMCs) were isolated using Ficoll Paque (GE Healthcare) and SepMate™-50 tubes (STEMCELL technologies) following manufacturer’s protocol.

T-ALL samples were obtained from UCL/UCLH Biobank, Great Ormond Street Biobank, Blood Cancer UK Leukaemia CellBank and the Biobank of Ospedale Papa Giovanni XXIII, Bergamo, Italy.

Flow cytometry was performed on a BD LSR Fortessa instrument (BD Biosciences), a CytoFLEX (Beckman Coulter) or NovoCyte (Agilent) 96-well flow cytometer. CD21 antigen density was determined using Quantibrite™ PE beads (BD) as per the manufacturer’s instructions. See supplementary methods for a list of antibodies used.

CD21-scFv generation

Genetic rat vaccination for CD21 was performed by Aldevron and anti-CD21 scFvs identified through biopanning then screening of an immune phage display both manually and using PacBio next generation sequencing (Genewiz), as previously described(14)

Surface plasmon resonance

Kinetic analysis was performed on a Biacore T200 (GE Healthcare) using recombinant human CD21 protein covalently coupled to a CM5 sensor chip according to manufacturer’s recommendations. Serial dilutions of scFvs with a murine IgG2a Fc were captured and injected over the flow cells at a rate of 30ul/minute and equilibrium binding analysis was performed. Kinetic rate constants were obtained by curve fitting according to a Langmuir 1:1 binding model. Data analysis was performed using Biacore Insight Evaluation software v3.0 (GE Healthcare).

Differential scanning fluorimetry

A Prometheus NT.48. (NanoTemper) was used. The emission of fluorescence radiation was measured at wavelengths 330nm and 350nm across a temperature range from 20°C to 95°C at a rate of increase of 1°C min^− 1 and used to determine the protein melting temperatures.

Generation of CAR-T cells

CAR constructs were expressed in the SFG vector and retroviral supernatant generated through transient transfection of HEK293T cells. Donor PBMCs were transduced using a standardised protocol as previously described.(11)

Generation of engineered cell lines

Raji and SUPT-1 CD21 knockout (KO) cell lines created using CRISPR/Cas9 engineering were donated by Autolus Ltd. Retroviral vectors containing CD21 SCRs 1–5, 6–10 and 11–15 and a BFP marker were used to transduce SUPT-1 and Raji CD21 KO cell lines. Cells were bulk-sorted for CD21 and BFP double positive populations, then single-cell cloned.

In vitro cytotoxicity and cytokine release assays

Target cells were stained with CellTrace™ Violet (CTV) (Thermofisher), plated at 25–50,000 per well with three technical repeats. CAR-T cell to target cell ratios of 1:1, 1:2, 1:4, 1:8 and 1:16 were used. Controls were non-transduced (NT) cells, a non-binding aCD21 CAR (Ph2) and aCD19 FMC63 CAR. At 48hrs, enzyme-linked immunosorbent assay (Biolegend) for IFN-γ and IL-2 was performed on cell supernatant as per manufacturer’s instructions. 72hr co-culture of primary T-ALL cells were performed using allogeneic donor T cells. Residual lymphoblasts were identified using human CD45, murine CD45, CD3 and CTV staining.

In vitro proliferation assays

CAR-T cells labelled with CTV were co-cultured with irradiated (30Gy) target cells in a 1:2 E:T ratio for 7 days. CTV MFI on days 0 and 7 determined CTV dilution and fold-expansion from baseline. CAR-T cell counts were established by staining with aCD34 antibody for the RQR8 CAR marker gene.

In vivo models of T-ALL

Work was performed under a UK Home Office-approved project license and was approved by the UCL Biological Services Ethical Review Committee.

PDX1 was derived from a 10-year old boy with biallelic CDKN2A-deleted T-ALL. PDX2 was derived from a 1-year old girl with ATM-deleted T-ALL. Both were developed at the Institute of Child Health by Professor Owen Williams.

6–8 week old female NSG mice (Charles River) were used. Mice were monitored behaviourally and with regular weighing. Mice were culled over behavioural concerns or > 15% weight loss at investigator discretion or at pre-specified time points. Cell suspensions were injected via tail vein. 50ul tail vein bleeds were taken at set timepoints. Bone marrow and splenic tissue was collected at necroscopy for certain experiments. Tumour cells were identified as human CD45 + CD3-. Bioluminescence (BLI) imaging was performed using the IVIS system (PerkinElmer). General anaesthesia was induced using inhaled isoflurane. Following induction, intraperitoneal injection (IP) of luciferin (200 µl) was undertaken.

Pharmacological manipulation of CD21 antigen density

Drug concentrations used: L685 (1uM), GDC0941 (1uM), BEZ235 (1uM), AZD5363 (5uM), rapamycin (20nM), copanlisib (100nM). Target cells at 5x10⁵/ml were incubated with drugs in 24-well or 6-well T cell coated plates for 24, 48 or 72 hours. T lymphoblasts were maintained in culture using the HS5 stromal cell line, donated by Professor Kwee Yong. Blasts were identified as CD7 + BFP–. Drugs were washed off three times using sterile PBS. In vivo, 14mg/kg copanlisib or DMSO/PBS control was injected intraperitoneally.

Statistical analysis

Statistical analysis was performed using GraphPad Prism v9 for Windows (GraphPad Software, San Diego, California USA). Student’s t test or 2-way ANOVA were used for samples with normally distributed variables. Survival curves were generated using the Kaplan-Meier method. P < 0.05 showed statistical significance. Data is summarised as mean +/- standard deviation.

DATA SHARING STATEMENT

For original data, please contact [email protected].

CD21 is expressed in T-ALL and mature B cell malignancies

70% (11/16) of T-ALL cell lines expressed CD21 (mean copy number (ABC) 2569 copies/cell) (Fig. 1A). 33/58 (57%) of diagnostic T-ALL samples expressed CD21 (Fig. 1B), with a mean ABC of 1262 (range 408–2875) (Fig. 1C). Expression varied by maturational stage with highest seen in cortical (80%), pre-T (72%), and mature T-ALL (67%) and lesser expression in ETP-ALL (25%) and Pro T-ALL (17%) (Fig. 1D). 45% (5/11) of relapsed samples were CD21+ (mean ABC 1279). 6/30 (20%) primary refractory samples expressed CD21 (mean ABC 1478) (Figs. 1E-F).

In a large published pediatric dataset, CD21 (CR2) RNA expression was detected in 80% of T-ALL cases, across all molecular subtypes.(15) In this and the UKALL14 adult dataset(16) CR2 correlated with the NOTCH1 target Deltex 1, with only limited correlation with CCR9, consistent with NOTCH control of CD21 expression.(11) (Supplementary Fig. 1A-D) Further, we compared CD21 expression to previously described T-ALL selective targets CCR9 and CD1a (Supplementary Fig. 2) by flow cytometry and demonstrated partially overlapping expression with each.

We also evaluated CD21 expression in B cell malignancies. 67% of B lymphoblasts (6/9) were CD21 positive by flow, but expression was variable with low copy number. However, bright and homogenous CD21 expression was seen on 100% of CLL samples (mean ABC 1516, range 915–2491). (Fig. 1G-H)

CD21 is only expressed on a minor fraction of healthy T cells

In normal donor PBMCs, CD21 was highly expressed on B cells, but was not detected on granulocytes, natural killer (NK) cells, NK-T cells and monocytes. CD21 was only detected on few alpha-beta T cells with similar expression in CD4 (mean 10% CD21+) and CD8 cells (mean 11% CD21+, Fig. 1I). T cells expressed CD21 at low surface density (mean CD3 + T cell ABC 316), 10-fold lower than B cells and 4-fold lower than primary T-ALL (Fig. 1J). CD21 expression was highest in gamma-delta and naive T cell subsets (mean 15 and 20% respectively), still at low density (mean naïve T cell ABC 316, mean G-D T cell ABC 335, mean B cell ABC 4250). CD21 did not correlate with markers of exhaustion (Fig. 1J-K) and was not upregulated by activation (Fig. 1L). RNA expression data confirmed limited expression of CR2 on normal tissues with expression limited to lymphoid cells. (Supplementary Fig. 1E-F). On developing thymocytes CR2 was highest on double-negative subsets, before becoming lost in more mature double and single positive subsets.(17) (Supplementary Fig. 1G)

CARs targeting CD21 membrane-distal epitopes are ineffective

Six anti-CD21 scFvs were selected from an immune phage library generated by immunization of rats with full-length CD21 ectodomain.(18) All 6 scFvs had nanomolar binding affinities (Fig. 2A) and bound to full-length CD21 on Raji WT, with 5/6 binding to CD21 membrane-distal epitopes (SCRs 1–5) and none binding to membrane-proximal epitopes. (Fig. 2B).

Second-generation 41BBz CARs derived from these scFvs (Fig. 2C) were not functional in vitro, with no specific interferon-gamma (IFN-γ) (Fig. 1D) or IL-2 (Fig. 1E) secretion when co-cultured with CD21-positive cell lines. Function was not improved when two of these CARs (C43/C48) were tested with different length spacers (Supplementary Fig. 3). To explore the cause of poor efficacy, SUPT-1 cells were engineered to express only CD21 SCRs 1–5 on the cell surface. The same CARs showed significantly improved cytokine release (Figs. 2F-G) which could not be explained by higher antigen expression (Fig. 2H). This suggested that the membrane-distal epitope location and/or epitope accessibility was responsible for poor in vitro CAR function.

CARs targeting CD21 membrane-proximal epitopes show improved in vitro function

A new immune library was generated from rats immunized with truncated CD21 (SCRs11-15). Next-generation sequencing(14) identified 16 scFvs with nanomolar binding affinities (Fig. 3A-B). All but one bound to the 5 most membrane-proximal epitopes on CD21 (SCRs 11–15) (Fig. 3C). In addition, a scFv was derived from the anti-CD21 antibody clone Bu32. This bound to SCR 11–15 with a higher binding affinity than any phage display-derived scFv. (Fig. 3A-C). Notably, binder diversity was limited with 16 scFvs identified belonging to 1 of 9 highly similar heavy chain CDR3 families (Supplementary Fig. 4A-B).

All scFvs were cloned into 41BBz CAR architecture as before (Fig. 3D). Six CARs (NM2, Ph9, Ph10, H1, Ph15 and Ph17) triggered in vitro IFN-G release against a CD21 positive cell line; albeit with high basal release against a CD21-negative cell line (Figs. 3G and J). Only 3 CARs (NM2, Ph9 and Ph10) resulted in CD21-specific cytotoxicity (Figs. 3E-F, I) (Supplementary Fig. 4). Notably, NM2, Ph9 and Ph10 have similar VH CDR3. No CD21-specific IL-2 secretion was observed (Figs. 3H,K).

CD21 Fab-CARs show greater stability, and CD21 specific cytotoxicity and cytokine release

We hypothesized that the increased stability of a Fab-based CAR would improve performance and reduce tonic signalling of these CARs. scFvs recognising membrane-proximal epitopes were formatted into 41BB-Z Fab CARs expressed by a bi-cistronic vector (Figs. 4A-B). Four Fab-CARs demonstrated CD21-specific cytotoxicity and cytokine release (Fig. 4C-E), with NM2 Fab showing greatest potency across multiple donors (Fig. 4F-G).

We explored the reason for this improved function. Fab-CARs had no altered binding affinity compared to their scFv-based counterparts (Fig. 4H) but did show increased thermal stability (Fig. 4I-J), with NM2 Fab particularly stable. Moreover, Fab-CARs, when expressed on HEK293T cells (Fig. 4K) and PBMCS (Fig. 4L) showed lower surface expression compared to their scFv-based CAR equivalents. Thus, reduced surface density and higher stability may have led to reduced basal signalling.

Multiple other engineering approaches did not convey greater benefit. Using a high-affinity binder (Bu32 – KD 2.46E-10) (Supplementary Fig. 5A-B) or a CD28 spacer/ TMD did not result in greater in vitro efficacy compared to NM2 Fab-CAR (Supplementary Fig. 5C-H). A dual anti-CD21 CAR co-targeting proximal and distal epitopes simultaneously failed to improve in vitro function (Supplementary Fig. 4I-J).

NM2 CD21 Fab-CAR is effective against CD21 low density T-ALL cell lines and PDX

NM2 Fab-CAR was further evaluated in vitro. NM2 Fab-CAR-T cells proliferated in a 7-day proliferation assay (Fig. 5a-b) with no difference in exhaustion (Fig. 5c) or differentiation (Fig. 5d) compared to control aCD19 CAR-T cells. NM2 Fab-CAR-T cells were CD21-negative (Fig. 5e) and expanded post-transduction comparably to CD19 CAR (Fig. 5f) indicating limited fratricide of CD21-positive T cells.

NM2 Fab-CAR showed CD21-specific cytotoxicity and IFN-γ secretion against 3 CD21-low T-ALL cell lines: Jurkat (ABC-618), MOLT-4 (ABC-1196), P12 (ABC-601) (Fig. 5g).

Further, when co-cultured with 2 CD21-dim (474 and 889 CD21 ABC respectively) patient-derived xenograft (PDX) models of T-ALL in a 72hr cytotoxicity assay at a 1:1 E:T ratio, NM2 Fab CAR specifically eliminated both T-ALL lymphoblast populations (Fig. 5h).

CD21 CARs exhibit anti-leukemic activity in vivo

NM2 Fab CAR was next evaluated in a Jurkat cell line xenograft model of T-ALL. 3x10⁶ Jurkat cells were followed by 1x10⁶ CAR21 or NT cells on day 6 (Fig. 6A). CAR21-treated mice had significantly better tumour control by BLI (Fig. 6B) and tumour burden in bone marrow at necroscopy (Fig. 6C). CAR21 mice had statistically improved survival (median 41days) over NT mice (median 22 days) (Fig. 6D).

We next used two clinically relevant PDX models of T-ALL expressing CD21 at low density (404 and 596 ABC). In PDX1, 1x10⁶ T-ALL lymphoblasts were followed by 4x10⁶ CAR21 or CAR19 cells on day 20 and mice monitored with bi-weekly bleeding (Fig. 6E). CAR19 mice rapidly progressed and died by day 41. CAR21 exerted statistically improved tumour control with complete disease eradication in the blood by day 57 (Fig. 6F) leading to improved survival for CAR21 mice (Fig. 6G). There were late relapses in CAR21 mice, with disease in spleens of 3/4 at necroscopy (Fig. 6H) which remained CD21-positive (Fig. 6I) and was associated with loss of detectable CAR-T cells in this immunodeficient model. Tumour control at this stage was in only 1 mouse with detectable T cells (Fig. 6H).

In PDX2, 0.8x10⁶ CAR21 or CAR19 cells were used (Fig. 6J). CAR19 mice experienced rapid disease progression. CAR21 again exerted tumour control in blood with eradication achieved by day 48 (Fig. 6K). CAR21 mice had a significant survival advantage over CAR19 mice and were all alive and well at experiment end (Fig. 6L). As previously, most mice had no detectable CAR-T cells by day 40 (Fig. 6M).

PI3K/mTOR inhibition increases CD21 expression in T-ALL and enhances CAR-T function.

PI3K inhibition of T-ALL cell lines has been shown to activate the NOTCH-MYC pathway and upregulate CD21(19). We hypothesized that a pharmaceutical approach to increasing CD21 antigen density may be possible. We first explored effect of PI3K/mTOR/AKT and NOTCH pathway inhibition on CD21 expression. Across six T-ALL cell lines, incubation with four PI3K/mTOR/AKT pathway inhibitors (AZD5363 [Akt], GDC0941 [pan PI3K], BEZ235 [pan PI3K + mTOR] and rapamycin [mTOR]) upregulated CD21 expression (Fig. 7A, B) at least 2-fold. Conversely, the gamma-secretase inhibitor L685 downregulated CD21 expression. CD21 modulation persisted up to 48hrs after drug removal (Fig. 7C). CD21 modulation did not occur in a CD21- negative cell lines or normal B and T cells. (Fig. 7D). Notably, at least 2-fold increases in CD21 were seen with the clinically approved dual PI3Kαδ inhibitor copanlisib in 5 T-ALL cell lines (Fig. 7E), similar to rapamycin (Fig. 7F). The impact of copanlisib was specific to CD21, as (excepting CCR9 on SUPT-1) there was no change in CCR9 or CD7 expression across 4 T-ALL cell lines (Fig. 7G). This effect could be reproduced in a CD21-low PDX T-ALL sample (Fig. 7H) and in vivo with xenografted MOLT-4 cells. (Fig. 7I)

We next explored whether PI3K-mediated CD21 upregulation would improve in vitro CAR21 functionality. In MOLT-4, an increase in CD21, through 48hr pre-incubation with either rapamycin or copanlisib, significantly increased IFN-γ and IL-2 production (Fig. 7J. Such increases were replicated in 2 further T-ALL cell lines, CUTLL-1 and RPMI 8402 (Figs. 7k-l). However, when copanlisib was added directly to the co-culture rather than via pre-incubation, there was a significant attenuation of IFN-gamma and IL-2 production suggesting an expected direct inhibitory effect on CAR-T effector functions (Fig. 7K-L).

Patients with r/r T-ALL face a poor prognosis with OS of < 10% in adults and < 25% in children.(2, 5, 20) Salvage chemotherapy to re-induce remission followed by HSCT is the only current curative approach, but this is largely unsuccessful and limited to younger patients without co-morbidities(21, 22). In contrast, several immunotherapies have been developed for r/r B-ALL, including CD19 CAR-T cell therapy which can induce durable remissions.(23) This success has motivated development of CAR-T cell approaches in r/r T-ALL, where there is a clear unmet need.

An analogous approach to CD19 CAR-T cell therapy is targeting of a pan-T cell antigen. Of these approaches, CD7 is the most investigated. (7, 8, 24–26) For instance, a recent phase 1 study found 95% bone marrow MRD-negative complete responses (CR) by D28.(24) CAR-T cell targeting of pan-T cell antigens is expected to cause CAR-T cell fratricide and T cell aplasia. Surprisingly, early CD7 CAR data showed expansion of naturally occurring CD7-negative T cells, although overall T cell levels remained low with opportunistic infections and viral reactivation described.(7, 24) With short follow-up, and most undergoing HSCT, the long-term consequences of CD7 T cell depletion remain unknown. Furthermore, relapses due to CD7 loss have been reported(7, 24)

An alternative approach is targeting of selective antigens which are expressed by T-ALL blasts, but not by normal T cells. A few such antigens have been described: CD1a is a selective target, although only expressed by cortical T-ALL which tend to have a good prognosis.(10) Anti-TRBC1/2 CAR-T cells are more relevant to mature T cell malignancies as only 1/3rd of T-ALL cases express surface TCR(9). We recently proposed CCR9 as a CAR target for T-ALL; although not expressed in all T-ALL cases, CCR9 was enriched in r/r and ETP T-ALL and a clinical trial of anti-CCR9 CAR-T cells is planned(11). Given that selective T-ALL antigens do not give complete coverage of all cases, and that escape through antigen loss has been reported, characterization of more T-ALL target antigens is worthwhile.

CD21 expression has long been reported in mature B cell malignancies such as CLL, DLBCL and follicular NHL.(27–29) Reports of expression on T-ALL cell lines and T-ALL date back several decades(30, 31), and more recently, the Tokyo Children’s Cancer Study Group described CD21 expression in cortical T-ALL(32). Here, we extend these studies and show CD21 expression in T-ALL across a range of subtypes – highest in pre-T (72%), cortical (80%), and mature T-ALL (67%) with lesser expression in ETP-ALL.

CD21 is also expressed on normal cells including T cells, B cells and follicular dendritic cells: B cell aplasia is an expected consequence of CD21 directed CAR-T therapy but is generally well-tolerated.(33) Additionally, depletion of FDCs could result in impaired response to infection. We found that CD21 was only expressed at low density on a fraction of mature T cells (10%), particularly in naïve subsets. This is in keeping with recent literature describing greatest CD21 expression in recent thymic emigrants.(34) Accordingly, anti-CD21 CAR-T cells were not affected by fratricide and significant T cell aplasia is unlikely to be consequent of this strategy. Functional consequences of CD21 + T cell depletion are difficult to predict since the physiological role of CD21 + T cells is unknown. Naïve T cell depletion may theoretically lead to restricted TCR repertoire and consequent impaired immunity, particularly in older adults with thymic involution. In mice the CR2 gene encodes for both CD21 and CD35. CR2 knockout mice exhibit impaired IgM and IgG responses but no T cell deficit has been described. Notably, previous antibody based CD21-directed therapies tested against B cell malignancies in early studies resulted in minimal toxicity.(35, 36)

Of the described selective target antigens, CCR9 is the most widely expressed being found in 73% of T-ALL cases with particular enrichment in r/r disease. CD21 and CCR9 expression in T-ALL at the RNA level shows only limited overlap between the two antigens (Supplementary Fig. 1) and flow cytometry of 50 T-ALL samples identified multiple cases of both CCR9 + CD21- and CCR9-CD21 + disease, identifying an additional 14% of cases that were CD21-positive but CCR9-negative. aCD21 CAR-T cell therapy may then be useful in many cases of CCR9-negative T-ALL. Additionally we found CD21 to be a more useful target than CD1a with only 20% of cases in our series CD1a positive (Supplementary Fig. 2).

CD21 is a challenging CAR target. Its long, flexible highly glycosylated structure could be expected to inhibit CAR-T activation.(37, 38). At 146kDa, it is significantly larger than established targets like CD19 and even CD22.(39). As with CD22,(40) anti-CD21 CARs targeting membrane-distal epitopes were not functional and targeting of membrane-proximal epitopes was required to trigger CAR activation. Despite this, CARs had limited function in vitro. Further, similar to experience targeting CD22, a very high affinity CD21 CAR did not enhance in vitro potency.(40, 41) Fab-based CARs have been previously described (42) and their increased stability has been exploited to improve CAR performance.(43) With this structure, we were able to produce CARs capable of CD21-specific lysis and cytokine release in vitro. We found that converting proximal targeting CARs to a Fab format reduced tonic signalling and improved performance, resulting in a final aCD21 CAR candidate effective in clinically relevant in vivo models.

CD21 is expressed at relatively low density in T-ALL (median density 1262 copies/cell) compared to > 10,000 CD19 copies in B-ALL.(44) Multiple groups have published on correlations between antigen density and CAR-T potency.(44) CD21 expression in T-ALL could be upregulated through incubation with PI3K/mTOR inhibitors, and this led to improved CD21-specific cytokine release by aCD21 CAR-T cells against multiple T-ALL lines. We suggest that PI3K/mTOR inhibitors could be used as a priming therapy prior to aCD21 CAR-T infusion. This work adds to the literature on pharmacological upregulation of CD22, CD38 and BCMA antigen density to enhance therapeutic efficacy of CAR-T cells in patients with heterogenous or dim antigen expression and to attempt to reduce the risk of antigen negative escape.(45–48)

In conclusion, we have proposed CD21 as a tumour-selective antigen for immunotherapeutic targeting of T-ALL, avoiding risks of fratricide and T cell aplasia. aCD21 Fab-CAR-T cells were effective in in vitro and in vivo models of T-ALL. This work adds to the growing literature describing CAR-T cell strategies for T-ALL - important for a heterogenous disease likely to require multi-antigen targeting for optimal outcomes. We also provide further evidence that pharmacological upregulation of antigen density can enhance CARs targeting low antigen density targets, and our experience provides lessons regarding engineering required when attempting to target complex antigens.

Author contributions

Contribution: N.C.M designed, led, and performed the in vitro and in vivo experiments, analyzed the data, prepared the figures, and wrote the manuscript; A.B performed hybridoma screening, antibody and CAR subcloning, in vitro and in vivo experiments; M.H performed antibody and CAR subcloning, flow cytometry and in vitro experiments; F.N and P.W provided support with phage display; T.K assisted with in vitro/in vivo data collection; animal work was performed on a Home Office license held by L.L; M.F performed Biacore and differential scanning fluorimetry studies; T.M provided immunohistochemistry support for the study; G.G sourced primary adult leukemia samples for the study; S.O performed Biacore and assisted with PacBio NGS data analysis; D.O performed and analysed flow cytometry data on primary samples; M.R.M supported the experiments and provided RNA sequencing data; A.K and M.A.P conceived the study, obtained funding, and wrote the manuscript; P.M.M designed, led and supervised the in vitro and in vivo experiments, and wrote the manuscript.

CONFLICT OF INTEREST DISCLOSURES

P.M.M. and L.L. own stock and received research funding from Autolus Ltd. M.A.P., S.O and M.F. are current or previous employees of and own stock in Autolus Ltd. The remaining authors declare no competing financial interests.

Acknowledgements

This study was supported by research funding from Cancer Research UK (CRUK), the Medical Research Council (MRC), and Great Ormond Street Hospital Charity (GOSHCC). Primary childhood leukemia samples used in this study were provided by Blood Cancer UK Childhood Leukaemia CellBank, and GOSH Biobank. Primary adult leukemia samples used were provided by UCL/UCLH Biobank and by the Papa Giovanni XXIII hospital leukemia biobank.

Raetz EA, Teachey DT. T-cell acute lymphoblastic leukemia. Hematology Am Soc Hematol Educ Program. 2016 Dec 2;2016(1):580–8.
Marks DI, Paietta EM, Moorman AV, Richards SM, Buck G, DeWald G, et al. T-cell acute lymphoblastic leukemia in adults: clinical features, immunophenotype, cytogenetics, and outcome from the large randomized prospective trial (UKALL XII/ECOG 2993). Blood. 2009 Dec 10;114(25):5136–45.
Dunsmore KP, Winter SS, Devidas M, Wood BL, Esiashvili N, Chen Z, et al. Children’s Oncology Group AALL0434: A Phase III Randomized Clinical Trial Testing Nelarabine in Newly Diagnosed T-Cell Acute Lymphoblastic Leukemia. JCO. 2020 Aug 19;38(28):3282–93.
Rowntree CJ, Kirkwood AA, Clifton-Hadley L, Farah N, Mansour MR, Hussain J, et al. First Analysis of the UKALL14 Randomized Trial to Determine Whether the Addition of Nelarabine to Standard Chemotherapy Improves Event Free Survival in Adults with T-Cell Acute Lymphoblastic Leukaemia (CRUK/09/006). Blood. 2021 Nov 5;138(Supplement 1):366.
Ko RH, Ji L, Barnette P, Bostrom B, Hutchinson R, Raetz E, et al. Outcome of Patients Treated for Relapsed or Refractory Acute Lymphoblastic Leukemia: A Therapeutic Advances in Childhood Leukemia Consortium Study. JCO. 2010 Feb 1;28(4):648–54.
Mamonkin M, Rouce RH, Tashiro H, Brenner MK. A T-cell-directed chimeric antigen receptor for the selective treatment of T-cell malignancies. Blood. 2015 Aug 20;126(8):983–92.
Pan J, Tan Y, Wang G, Deng B, Ling Z, Song W, et al. Donor-Derived CD7 Chimeric Antigen Receptor T Cells for T-Cell Acute Lymphoblastic Leukemia: First-in-Human, Phase I Trial. JCO. 2021 Oct 20;39(30):3340–51.
Gomes-Silva D, Srinivasan M, Sharma S, Lee CM, Wagner DL, Davis TH, et al. CD7-edited T cells expressing a CD7-specific CAR for the therapy of T-cell malignancies. Blood. 2017 20;130(3):285–96.
Maciocia PM, Wawrzyniecka PA, Philip B, Ricciardelli I, Akarca AU, Onuoha SC, et al. Targeting the T cell receptor β-chain constant region for immunotherapy of T cell malignancies. Nature Medicine. 2017 Nov 13;23(12):1416–23.
Sánchez-Martínez D, Baroni ML, Gutierrez-Agüera F, Roca-Ho H, Blanch-Lombarte O, González-García S, et al. Fratricide-resistant CD1a-specific CAR T cells for the treatment of cortical T-cell acute lymphoblastic leukemia. Blood. 2019 May 23;133(21):2291–304.
Maciocia PM, Wawrzyniecka PA, Maciocia NC, Burley A, Karpanasamy T, Devereaux S, et al. Anti-CCR9 Chimeric Antigen Receptor T cells for T Cell Acute Lymphoblastic Leukemia. Blood. 2022 May 4;blood.2021013648.
Liu YJ, Xu J, de Bouteiller O, Parham CL, Grouard G, Djossou O, et al. Follicular Dendritic Cells Specifically Express the Long CR2/CD21 Isoform. J Exp Med. 1997 Jan 6;185(1):165–70.
Mongini PK, Vilensky MA, Highet PF, Inman JK. The affinity threshold for human B cell activation via the antigen receptor complex is reduced upon co-ligation of the antigen receptor with CD21 (CR2). The Journal of Immunology.:11.
Nannini F, Senicar L, Parekh F, Kong KJ, Kinna A, Bughda R, et al. Combining phage display with SMRTbell next-generation sequencing for the rapid discovery of functional scFv fragments. mAbs. 2021 Jan 1;13(1):1864084.
Liu Y, Easton J, Shao Y, Maciaszek J, Wang Z, Wilkinson MR, et al. The genomic landscape of pediatric and young adult T-lineage acute lymphoblastic leukemia. Nature Genetics. 2017 Aug;49(8):1211–8.
Mansour MR, Sulis ML, Duke V, Foroni L, Jenkinson S, Koo K, et al. Prognostic implications of NOTCH1 and FBXW7 mutations in adults with T-cell acute lymphoblastic leukemia treated on the MRC UKALLXII/ECOG E2993 protocol. J Clin Oncol. 2009 Sep 10;27(26):4352–6.
Casero D, Sandoval S, Seet CS, Scholes J, Zhu Y, Ha VL, et al. Long non-coding RNA profiling of human lymphoid progenitor cells reveals transcriptional divergence of B cell and T cell lineages. Nat Immunol. 2015 Dec;16(12):1282–91.
Nannini F, Parekh F, Wawrzyniecka P, Mekkaoui L, Righi M, Dastjerdi FV, et al. A primer set for the rapid isolation of scFv fragments against cell surface antigens from immunised rats. Sci Rep. 2020 Dec;10(1):19168.
Shepherd C, Banerjee L, Cheung CW, Mansour MR, Jenkinson S, Gale RE, et al. PI3K/mTOR inhibition upregulates NOTCH-MYC signalling leading to an impaired cytotoxic response. Leukemia. 2013 Mar;27(3):650–60.
Fielding AK, Richards SM, Chopra R, Lazarus HM, Litzow MR, Buck G, et al. Outcome of 609 adults after relapse of acute lymphoblastic leukemia (ALL); an MRC UKALL12/ECOG 2993 study. Blood. 2007 Feb 1;109(3):944–50.
Raetz EA, Borowitz MJ, Devidas M, Linda SB, Hunger SP, Winick NJ, et al. Reinduction platform for children with first marrow relapse of acute lymphoblastic Leukemia: A Children’s Oncology Group Study[corrected]. J Clin Oncol. 2008 Aug 20;26(24):3971–8.
Samra B, Alotaibi AS, Short NJ, Khoury JD, Ravandi F, Garris R, et al. Outcome of adults with relapsed/refractory T-cell acute lymphoblastic leukemia or lymphoblastic lymphoma. American Journal of Hematology. 2020;95(9):E245–7.
Roddie C, Dias J, O’Reilly MA, Abbasian M, Cadinanos-Garai A, Vispute K, et al. Durable Responses and Low Toxicity After Fast Off-Rate CD19 Chimeric Antigen Receptor-T Therapy in Adults With Relapsed or Refractory B-Cell Acute Lymphoblastic Leukemia. JCO. 2021 Oct 20;39(30):3352–63.
Lu P, Liu Y, Yang J, Zhang X, Yang X, Wang H, et al. Naturally Selected CD7 CAR-T Therapy without Genetic Manipulations for T-ALL/LBL: First-in-human Phase I Clinical Trial. Blood. 2022 May 2;blood.2021014498.
Zhang M, Chen D, Fu X, Meng H, Nan F, Sun Z, et al. Autologous Nanobody-Derived Fratricide-Resistant CD7-CAR T-cell Therapy for Patients with Relapsed and Refractory T-cell Acute Lymphoblastic Leukemia/Lymphoma. Clinical Cancer Research. 2022 May 17;OF1–14.
Png YT, Vinanica N, Kamiya T, Shimasaki N, Coustan-Smith E, Campana D. Blockade of CD7 expression in T cells for effective chimeric antigen receptor targeting of T-cell malignancies. Blood Advances. 2017 Nov 28;1(25):2348–60.
Otsuka M, Yakushijin Y, Hamada M, Hato T, Yasukawa M, Fujita S. Role of CD21 antigen in diffuse large B-cell lymphoma and its clinical significance. Br J Haematol. 2004 Nov;127(4):416–24.
Nichols EM, Jones R, Watson R, Pepper CJ, Fegan C, Marchbank KJ. A CD21 low phenotype, with no evidence of autoantibodies to complement proteins, is consistent with a poor prognosis in CLL. Oncotarget. 2015 Oct 3;6(32):32669–80.
Li D, Xu Q, Hu Y, Wang W, Xie S, Zhao C, et al. Tumor-intrinsic CD21 expression impacts the response of B-cell malignancy cells to CD19-CAR-T cells. Journal of Leukocyte Biology. 2022;112(4):913–8.
Toben HR, Smith RG. T lymphocytes bearing complement receptors in a patient with chronic lymphocytic leukaemia. Clin Exp Immunol. 1977 Feb;27(2):292–302.
Shevach E, Edelson R, Frank M, Lutzner M, Green I. A human leukemia cell with both B and T cell surface receptors. Proc Natl Acad Sci U S A. 1974 Mar;71(3):863–6.
Ohki K, Takahashi H, Fukushima T, Nanmoku T, Kusano S, Mori M, et al. Impact of immunophenotypic characteristics on genetic subgrouping in childhood acute lymphoblastic leukemia: Tokyo Children’s Cancer Study Group (TCCSG) study L04-16. Genes, Chromosomes and Cancer. 2020;59(10):551–61.
Bhoj VG, Arhontoulis D, Wertheim G, Capobianchi J, Callahan CA, Ellebrecht CT, et al. Persistence of long-lived plasma cells and humoral immunity in individuals responding to CD19-directed CAR T-cell therapy. Blood. 2016 21;128(3):360–70.
Pekalski ML, García AR, Ferreira RC, Rainbow DB, Smyth DJ, Mashar M, et al. Neonatal and adult recent thymic emigrants produce IL-8 and express complement receptors CR1 and CR2. JCI Insight. 2017 Aug 17;2(16):e93739.
Benkerrou M, Jais JP, Leblond V, Durandy A, Sutton L, Bordigoni P, et al. Anti-B-cell monoclonal antibody treatment of severe posttransplant B-lymphoproliferative disorder: prognostic factors and long-term outcome. Blood. 1998 Nov 1;92(9):3137–47.
Czuczman MS, Straus DJ, Divgi CR, Graham M, Garin-Chesa P, Finn R, et al. Phase I dose-escalation trial of iodine 131-labeled monoclonal antibody OKB7 in patients with non-Hodgkin’s lymphoma. J Clin Oncol. 1993 Oct;11(10):2021–9.
Davis SJ, van der Merwe PA. The kinetic-segregation model: TCR triggering and beyond. Nat Immunol. 2006 Aug;7(8):803–9.
Xiao Q, Zhang X, Tu L, Cao J, Hinrichs CS, Su X. Size-dependent activation of CAR-T cells. Science Immunology. 2022 Aug 5;7(74):eabl3995.
Hannan JP. The Structure-Function Relationships of Complement Receptor Type 2 (CR2; CD21). Curr Protein Pept Sci. 2016;17(5):463–87.
Haso W, Lee DW, Shah NN, Stetler-Stevenson M, Yuan CM, Pastan IH, et al. Anti-CD22-chimeric antigen receptors targeting B-cell precursor acute lymphoblastic leukemia. Blood. 2013 Feb 14;121(7):1165–74.
Long AH, Haso WM, Orentas RJ. Lessons learned from a highly-active CD22-specific chimeric antigen receptor. OncoImmunology. 2013 Apr;2(4):e23621.
Nolan KF, Yun CO, Akamatsu Y, Murphy JC, Leung SO, Beecham EJ, et al. Bypassing immunization: optimized design of “designer T cells” against carcinoembryonic antigen (CEA)-expressing tumors, and lack of suppression by soluble CEA. Clin Cancer Res. 1999 Dec;5(12):3928–41.
Hotblack A, Kokalaki EK, Palton MJ, Cheung GWK, Williams IP, Manzoor S, et al. Tunable control of CAR T cell activity through tetracycline mediated disruption of protein–protein interaction. Sci Rep. 2021 Nov 9;11(1):21902.
Majzner RG, Rietberg SP, Sotillo E, Dong R, Vachharajani VT, Labanieh L, et al. Tuning the Antigen Density Requirement for CAR T-cell Activity. Cancer Discov. 2020 May;10(5):702–23.
Ramakrishna S, Highfill SL, Walsh Z, Nguyen SM, Lei H, Shern JF, et al. Modulation of Target Antigen Density Improves CAR T-cell Functionality and Persistence. Clin Cancer Res. 2019 Sep 1;25(17):5329–41.
Yang X, Yu Q, Xu H, Zhou J. Upregulation of CD22 by Chidamide promotes CAR T cells functionality. Sci Rep. 2021 Dec;11(1):20637.
Yoshida T, Mihara K, Takei Y, Yanagihara K, Kubo T, Bhattacharyya J, et al. All-trans retinoic acid enhances cytotoxic effect of T cells with an anti-CD38 chimeric antigen receptor in acute myeloid leukemia. Clin Transl Immunology. 2016 Dec 9;5(12):e116.
Pont MJ, Hill T, Cole GO, Abbott JJ, Kelliher J, Salter AI, et al. γ-Secretase inhibition increases efficacy of BCMA-specific chimeric antigen receptor T cells in multiple myeloma. Blood. 2019 Nov 7;134(19):1585–97.

Yes there is potential conflict of interest.

221101leukemiasupplementary.pdf

Download PDF

Version 1

posted

You are reading this latest preprint version

Anti-CD21 Chimeric Antigen Receptor T cells for the Treatment of T Cell Acute Lymphoblastic Leukemia

Status:

Version 1

Abstract

Figures

Key Points

Introduction

Methods

Cell lines and maintenance

Flow cytometry

CD21-scFv generation

Surface plasmon resonance

Differential scanning fluorimetry

Generation of CAR-T cells

Generation of engineered cell lines

In vitro cytotoxicity and cytokine release assays

In vitro proliferation assays

In vivo models of T-ALL

Pharmacological manipulation of CD21 antigen density

Statistical analysis

DATA SHARING STATEMENT

Results

CD21 is expressed in T-ALL and mature B cell malignancies

CD21 is only expressed on a minor fraction of healthy T cells

CARs targeting CD21 membrane-distal epitopes are ineffective

CARs targeting CD21 membrane-proximal epitopes show improved in vitro function

CD21 Fab-CARs show greater stability, and CD21 specific cytotoxicity and cytokine release

NM2 CD21 Fab-CAR is effective against CD21 low density T-ALL cell lines and PDX

CD21 CARs exhibit anti-leukemic activity in vivo

Discussion

Declarations

Author contributions

CONFLICT OF INTEREST DISCLOSURES

Acknowledgements

References

Additional Declarations

Supplementary Files

Status:

Version 1