Improved Antitumor Efficacy of Chimeric Antigen Receptor T Cells That Secrete Single-Domain Antibody Fragments

Published OnlineFirst February 4, 2020; DOI: 10.1158/2326-6066.CIR-19-0734

CANCER IMMUNOLOGY RESEARCH | RESEARCH ARTICLE

Improved Antitumor Efﬁcacy of Chimeric Antigen Receptor T Cells that Secrete Single-Domain Antibody Fragments Yushu Joy Xie1,2, Michael Dougan3, Jessica R. Ingram4,†, Novalia Pishesha1,2, Tao Fang1, Noor Momin2,5, and Hidde L. Ploegh1

ABSTRACT ◥ Chimeric antigen receptor (CAR) T-cell therapy is effective in the that can endow CAR T cells with desirable properties. The secretion treatment of cancers of hematopoietic origin. In the immunosup- of an anti-CD47 VHH by CAR T cells improves engagement of the pressive solid tumor environment, CAR T cells encounter obstacles innate immune system, enables epitope spreading, and can enhance that compromise their efﬁcacy. We developed a strategy to address the antitumor response. CAR T cells that secrete anti–PD-L1 or these barriers by having CAR T cells secrete single-domain antibody anti–CTLA-4 nanobodies show improved persistence and demon- fragments [variable heavy domain of heavy chain antibodies (VHH) strate the versatility of this approach. Furthermore, local delivery of or nanobodies] that can modify the intratumoral immune landscape secreted anti-CD47 VHH-Fc fusions by CAR T cells at the tumor and thus support CAR T-cell function in immunocompetent ani- site limits their systemic toxicity. CAR T cells can be further mals. VHHs are small in size and able to avoid domain swapping engineered to simultaneously secrete multiple modalities, allowing when multiple nanobodies are expressed simultaneously—features for even greater tailoring of the antitumor immune response.

Introduction cytokines such as IL12, IL15, and IL18 can modify the tumor microenvironment, make it more hospitable for T-cell activity, and thus Chimeric antigen receptor (CAR) T-cell therapy relies on T cells promote an antitumor response (9–11). Other means of manipulating that have been redirected with a receptor designed to recognize an the tumor microenvironment, for example by having CAR T cells antigen of choice. CAR T cells have been used to successfully treat release proteins that affect cell–cell interactions, therefore deserve hematologic cancers (1, 2). In the case of tumors of B-cell origin, consideration as well. CD19-targeted CAR T cells lead to remission in patients refractory to Variable heavy domain of heavy chain antibodies (VHH), also several other lines of therapy (3). CAR T cells have shown less success referred to as nanobodies, are single-domain antibody fragments in the treatment of solid tumors. Obstacles include a paucity of tumor- derived from the variable region of camelid heavy-chain-only anti- specific targets and a high rate of antigen escape (4, 5). A dense bodies (12). They are small, standalone proteins of approximately extracellular matrix, characteristic of many solid tumors, and the 15 kDa that retain binding affinities comparable with full sized presence of inhibitory checkpoint signals both blunt the immune mAbs (13). VHHs are stable, soluble, and can be expressed in excellent response (6–8). Suppressive cell types, such as T-regulatory cells or yields without the need for extensive optimization (14). For use as myeloid-derived suppressor cells, can further inhibit or cause therapeutic agents, their small size and sequence similarity to human exhaustion not only of CAR T cells but also of endogenous T cells immunoglobulin V regions render them less immunogenic than that have infiltrated the tumor (6). This combination of factors murine mAbs, and humanization of VHHs is possible (15). CAR T prevents CAR T cells from recognizing and attacking the tumor. To cells can be engineered so that they express and secrete VHHs with overcome some of these challenges, “armored” CAR T cells that secrete immunomodulatory properties to enhance their antitumor effect. Current immunotherapy approaches aim to improve adaptive immune recognition and killing of tumor cells, but the importance 1Program in Cellular and Molecular Medicine, Boston Children's Hospital, Boston, of engaging the innate immune system for enhanced antigen presen- 2 Massachusetts. Department of Biological Engineering, Massachusetts Institute tation and epitope spreading is increasingly recognized (16, 17). For of Technology, Cambridge, Massachusetts. 3Division of Gastroenterology, example, the CD47 protein delivers a “don't eat me signal” to pha- Massachusetts General Hospital, Boston, Massachusetts. 4Department of Cancer Immunology and Virology, Dana-Farber Cancer Institute, Boston, Massachu- gocytes. Consequently, a blockade of this signal synergizes with the setts. 5Koch Institute for Integrative Cancer Research, Massachusetts Institute of efficacy of mAb therapy for a variety of cancers in preclinical Technology, Cambridge, Massachusetts. models (18–21). Here, we show that engagement of the innate immune Note: Supplementary data for this article are available at Cancer Immunology system through blockade of CD47 by anti-CD47 VHH-secreting CAR Research Online (http://cancerimmunolres.aacrjournals.org/). T cells improves their antitumor effect. – †Deceased. We further demonstrate the modularity of VHH and VHH fusion secreting CAR T cells by developing CAR T cells that secrete VHHs Corresponding Author: Hidde L. Ploegh, Boston Children's Hospital, 1 Blackfan and VHH fusion proteins specific for multiple checkpoints. Such CAR Circle, RB09213, Boston, MA 02115-5713. Phone: 617-919-1613; Fax: 617-432-4775; E-mail: [email protected] T cells show less exhaustion and improved persistence when compared with nonsecreting CAR T cells. Because CAR T cells traffic to and Cancer Immunol Res 2020;8:518–29 persist at sites where their antigen is present (in our case, the tumor doi: 10.1158/2326-6066.CIR-19-0734 microenvironment), this strategy can limit immune-related adverse 2020 American Association for Cancer Research. effects associated with checkpoint blockade (22). Furthermore, we

AACRJournals.org | 518

VHH-Secreting CAR T Cells

show that, owing to their modest size, multiple VHH and VHH-fu- spinfected twice in 6-well plates at 2,000 g with the virus and polybrene sions can be expressed from the same vector in the same cell, under- mixture for 90 minutes. Fresh media was added after each spinfection, scoring the flexibility of this approach. as described previously (27, 30). After two spinfections, cells were Using CAR T cells as a vehicle for local delivery of VHH and VHH expanded for 2 days in complete RPMI IL2 and used for in vitro and fusion proteins should limit systemic exposure to immune modulators in vivo assays. such as anti-CD47 mAbs that have off-tumor toxicity. Anti-CD47 blocks engagement of the CD47 ligand, SIRP1a, while enhancing Anti-HA immunoprecipitation and immunoblots FcgR-mediated phagocytosis (23). The ubiquitous expression of Anti-HA immunoprecipitations (IP) were performed on culture CD47, especially on red blood cells (RBC), complicates systemic supernatants of transduced cells using the Pierce HA-Tag IP/Co-IP Kit administration of anti-CD47. The anemia that results from such (Thermo Fisher Scientific). HA tags were located at the C-terminus of treatment is sufficiently severe to have led to interruption of several the secreted VHHs. Supernatant (850 mL) was incubated with anti-HA clinical trials because of safety concerns (24, 25). Even so, CD47 agarose beads overnight, and eluates were run on a 12% SDS PAGE gel blockade remains a promising treatment if the toxicity concerns can and transferred to a PDVF membrane (Bio-Rad TurboBlot). Mem- be dealt with (26). We demonstrate that an anti-CD47 VHH, fused to branes were blocked in 5% milk in TBST and incubated with anti-HA- an Fc portion and secreted by CAR T cells, provides a stimulus HRP mAb overnight. Membranes were developed with western ECL for macrophage engulfment. CAR T cells that secrete anti-CD47 Fc substrate (PerkinElmer) and imaged on a Bio-Rad ChemiDoc imager. fusions show superior antitumor efficacy compared with CAR T-cell For immunoblots on cell lysates, cells were lysed with RIPA buffer therapy without such supplementation. Toxicity associated with sys- (25 mmol/L Tris, 150 mmol/L NaCl, 1% NP-40, 0.5% sodium temic exposure to anti-CD47 Fc is thus avoided. deoyxcholate, and 0.1% SDS) and protease inhibitor (complete mini protease inhibitor, Sigma). DNA was removed through benzonase digestion, and lysates were boiled with Laemmli sample buffer and Materials and Methods 2-mercaptoethanol. VHH-CAR retroviral design and construction Vectors used for the generation of CAR T cells were comprised of a In vitro cytotoxicity and activation assays fluorescent protein tracer linked to a P2A sequence followed by a CD8 Cytotoxicity of CAR T cells was measured using various coculture signal sequence and the VHH domain linked to a CD8 hinge and assays. The cytotoxicity of the PD-L1–targeted CAR T cells was shown transmembrane domain. The VHH sequences used are as follows: A4 – by coculture with B16F10 melanoma cells, which express PD-L1. B16 QVQLVESGGGLVEPGGSLRLSCAASGIIFKINDMGWYRQAPGK- cells were plated in complete RPMI (Corning) and recombinant RREWVAASTGGDEAIYRDSVKDRFTISRDAKNSVFLQMNSLKP- murine Il2, and A12 or A12 A4 CAR T cells were added to the EDTAVYYCTAVISTDRDGTEWRRYWGQGTQVTVSS; A12 – QV- B16F10 cells for a 24-hour incubation. T cells were then washed out of QLVESGGGLVQAGGSLRLSCTASGSTFSRNAMAWFRQAPGKER- the plate using PBS, and the number of remaining viable B16F10 cells EFVSGISRTGTNSYDADSVKGRFTISKDNAKNTVTLQMNSLKPE- were measured using a Cell Titer Glo Assay (Promega). Supernatants DTAIYYCALSQTASVATTERLYPYWGQGTQVTVSS; H11– QVQ- of each coculture were collected for IFNg ELISA measurements (BD LQESGGGLAQPGGSLRLSCAASGSTISSVAVGWYRQTPGNQRE- Biosciences). WVATSSTSSTTATYADSVKGRFTISRDNAKNTIYLQMNSLKPED- TAVYYCKTGLTNWGQGTQVTVSS; and B2 – QVQLVETGGGLV- Tumor cell lines QAGGSLRLSCAASGSTFSHNAGGWYRQAPEKQRELVAGISSDG- B16F10 cells obtained from the ATCC and IDEXX cleared for NINYADSVKDRFTISRDNASNTMYLQMNNLKPEDTAVYVCNI- Mycoplasma (2017) were used in the animal tumor models and in vitro RGSYGNTYYSRWGQGTQVTVSS. These were connected to a CD28 studies. Cells were not reauthenticated within the past year. Cells were costimulatory signal and a CD3z activation signal as described pre- maintained in complete RPMI, RPMI1640 supplemented with 10% viously (27, 28). For VHH-secreting CARs, a P2A, IRES, or separate heat-IFS, nonessential amino acids, 1 mmol/L sodium pyruvate, cytomegalovirus (CMV) construct followed the CD3z activation 2 mmol/L L-glutamine, 100 U/mL Penicillin/streptomycin, and domain, and these were linked to an IgK signal sequence to drive 50 mmol/L 2-mercaptoethanol. All cell lines were kept in culture – – secretion of the VHH, which was tagged at the C-terminus with an HA for up to 2 months. B16F10 PD-L1 / lines were generated using tag. The CAR constructs are contained in a murine stem cell virus– CRISPR-Cas9, and single clones were selected and sequenced to based g retroviral vector, the XZ vector, a gift from the Lodish verify gene disruption. MC38 was obtained from Kerafast (2017) laboratory (29). To generate virus, HEK 293T cells were transfected and cleared for Mycoplasma (2017), and cells were grown in using Fugene 6 (Promega) with the XZ CAR vector and a pCL-Eco Dulbecco medium supplemented with 10% heat-IFS, nonessential packaging vector. Cloning was carried out in DH5a E. coli cells. amino acids, 1 mmol/L sodium pyruvate, and 100 U/mL penicillin/ streptomycin. T-cell isolation and CAR transduction and generation To generate CAR T cells, pan T cells were isolated from mouse Protein production, purification, and endotoxin removal splenocytes using a magnetic bead-based–negative selection strategy Recombinant VHHs were produced by periplasmic expression according to the manufacturer's instructions (Thermo Fisher Dyna- in WK6 cells. Recombinant VHH-Fc fusions were produced in beads Untouched Mouse T Cell Kit). T cells were activated overnight Expi293 cells, which were cultured for 4 days after polyethylenimine with anti-mouse CD3 and CD28 (BD Biosciences) in complete PEI transfection. Proteins were purified from supernatants or RPMI [RPMI supplemented with 10% heat-inactivated FBS (IFS), periplasmic fractions on Ni-NTA beads (Qiagen) and eluted with nonessential amino acids, 1 mmol/L sodium pyruvate, 2 mmol/L L- 500 mmol/L imidazole. Eluted proteins were further purified with size glutamine, 100 U/mL Penicillin/streptomycin, and 50 mmol/L exclusion chromatography on a SuperDex 75 column run in PBS. 2-mercaptoethanol] and mIL2 (recombinantly, house-made) media Endotoxin removal was performed by extensive washing of column- and then transduced using the retrovirus and polybrene. T cells were bound materials with endotoxin-free PBS and 0.1% TritonX114.

AACRJournals.org Cancer Immunol Res; 8(4) April 2020 519

Xie et al.

Tumor models for efficacy vectors that encode fluorescently labeled A12 (anti–PD-L1) CARs All animal procedures performed were in accordance with institu- linked to a self-cleaving P2A peptide or IRES sequence and a secreted tional guidelines and approved by the Institutional Animal Care anti-CD47 VHH (A4; refs. 18, 27). We also tested constructs with and Use Committee of Boston Children's Hospital (IACUC protocol monocistronic expression, in which one promoter controls expres- 16-12-3328). To test the efficacy of the A4-secreting PD-L1–targeted sion of the CAR and a second promoter controls expression of a CAR T cells, PD-L1–deficient mice were inoculated with 5 105 cells VHH (Fig. 1B). Success of transduction was gauged by expression of B16F10 wild-type (WT) cells. Two days later, mice were treated with of the fluorescent reporter molecule (Fig. 1C). To verify that the A12 A4-secreting CAR T cells, A12 CAR T cells with daily systemic A4, transduced CARs express anti-CD47 VHH, a flow cytometry–based no CAR T cells with daily systemic A4, or left untreated. A total of three assay was used. A population of both transduced and control CAR T CAR T-cell injections were given intravenously in combination with a cells was stained with a fluorescently labeled anti-CD47 mAb that twice weekly intraperitoneal injection of the anti-TRP1 antibody competes for binding with the A4 anti-CD47 VHH. Populations of TA99, a gift from the D. Wittrup laboratory. Tumor measurements non-A4–secreting CAR T cells were also stained by the anti-CD47 were taken by caliper every 1–2 days. Animals were sacrificed following mAb, because all T cells express CD47. Populations that contain the CAC protocols when the tumor area was >125 mm2, and mice with A4-secreting CAR T cells showed a reduction of anti-CD47 staining ulcerated tumors were removed from the studies. For the MC38 model, in both transduced and untransduced cells. The A4 anti-CD47 3 105 cells were inoculated into PD-L1–deficient mice, and 5 days VHH is thus secreted and can bind to any cell present, regardless later, mice were treated with A12-A4Fc–secreting CAR T cells, A12 of whether that cell itself was transduced (Fig. 1D; Supplementary CAR T cells with a single dose of systemic A4Fc, A12 CAR T cells only, Fig. S1). Expression and secretion of A4 was further confirmed by no CAR T cells with a single dose of systemic A4Fc, or left untreated. A performing an anti-HA immunoblot on CAR T-cell lysates or on an total of three intravenous CAR T-cell injections were given, and tumor anti-HA immunoprecipitate of the culture supernatant of trans- measurements were taken by caliper every 1–2 days. To test the effect ductants (Fig. 1E). The 1B7 pCMV A4 construct is a VHH-based on persistence of the A12-secreting B2-targeted CAR T cells, WT mice CAR T cell, in which the CAR VHH recognizes a Toxoplasma gondii were inoculated with 1 105 cells of B16F10 WT cells. Four days later, (irrelevant) antigen. The 1B7 CAR T cell was likewise engineered to mice were treated intravenously with A12-secreting B2 CAR T cells, B2 secrete the A4 VHH to serve as a control. The A12 (anti–PD-L1) CAR T cells, or left untreated. To test the effect on persistence of the CAR in which a separate CMV promoter drives expression of A4 H11Fc-secreting B2-targeted CAR T cells, WT mice were inoculated showed the most A4 secretion. Because of limitations in transduc- with 1 105 cells of B16F10 WT cells. Four days after, mice were tion capabilities, possibly due to the large construct size of the CMV treated with H11Fc-secreting B2 CAR T cells, B2 CAR T cells with and IRES vectors, we performed all further experiments with the three systemic doses of H11Fc, only three systemic doses of H11Fc, or A12 P2A A4 construct. To confirm that killing by A12 CAR T cells left untreated. In persistence experiments, spleens, lymph nodes, and was not negatively affected by secretion of the A4 VHH, we þ tumors were removed to check for the presence of CAR T cells. Mice performed a killing assay on PD-L1 B16F10 cells. Cell killing was from each treatment group were cohoused to minimize variability due measured with a CellTiter Glo assay, and T-cell activation was to external factors. Experiments were repeated at least twice. measured by production of IFNg (Fig. 1F). Killing of B16F10 cells was comparable for the A12 CAR T cells and A12 P2 AA4 CAR T ELISPOT cells. Enforced secretion of A4 therefore neither impairs CAR T-cell IFNg ELISPOT assays were performed by coculturing with B16F10 activity nor production of IFNg. or B16F10 PD-L1–knockout (KO) cells using a mouse IFNg ELISPOT Kit (BD Biosciences). B16F10 and B16F10 PD-L1–KO tumor cells Delay of syngeneic tumor growth with anti-CD47 VHH-secreting were cultured with IFNg overnight. Splenocytes (1 106) from each CAR T cells treated mouse were then cocultured with 25,000 irradiated tumor cells Having confirmed the generation of A4-secreting A12 (anti–PD- (100 Gy) per well. Plates were developed according to the manufac- L1) CAR T cells, we examined whether secretion of A4 by CAR T cells turer's protocol and imaged with CTL Immunospot S6 analyzer. would improve their antitumor activity (Fig. 2A). PD-L1 KO mice were used as a source of T cells to generate A12 CAR T cells, as we Statistical analysis have observed better persistence of these cells due to a reduction in Statistical analyses were performed using GraphPad Prism. The log chronic CAR signaling in cis, caused by basal PD-L1 expression on WT rank (Mantel Cox) test was used to analyze survival curves. Flow T cells (27). WT CAR T cells generated in the presence of the anti–PD- cytometry data were analyzed using two-tailed, unpaired Student t L1 VHH behave similarly to CAR T cells generated in a PD-L1–KO tests, considering multiple comparison corrections. Statistical signif- background. We have previously demonstrated that chronic signaling icance was determined using the methods described in the Results from WT CAR T cells leads to upregulation of exhaustion markers (27). section or in the figure legends using the Bonferroni correction for B16F10 and MC38 tumors express CD47, as determined by flow comparison between multiple groups. cytometry (Supplementary Fig. S2). C57BL/6 PD-L1–KO mice were inoculated with B16F10 tumors on day 0, and received the treatments depicted in Fig. 2A. All groups also received the melanoma-specific Results anti-TRP1, TA99, to enhance antitumor activity. This allows macro- Anti-CD47 VHH-secreting anti–PD-L1 (A12) CAR T-cell phages and other phagocytic cells to exert an effect on tumor growth generation and function in vitro through FcgR engagement, which is enhanced by A4-VHH–mediated CD47 VHH-secreting anti–PD-L1 (A12) CAR T cells were gener- blockade of the CD47 “don't eat me” signal (Fig. 2A). Tumor sizes ated using the constructs depicted in Fig. 1A (27). We chose to develop were measured by caliper every 1–2 days. Treatment with the A12 CAR T cells that secrete an anti-CD47 VHH without the Fc portion to (anti–PD-L1) CAR T cells in conjunction with systemically delivered limit the toxicities associated with Fc engagement, as CD47 is A4 VHH improved survival when compared with mice treated with A4 expressed across a wide range of normal cells. We generated viral VHH only, or in comparison with mice that received no treatment

520 Cancer Immunol Res; 8(4) April 2020 CANCER IMMUNOLOGY RESEARCH

VHH-Secreting CAR T Cells

A B A12 P2A A4 (2.4 kb) VHH GFP VHH CD28 CD3zeta P2A VHH HA

Signal CD8 hinge and TM Signal sequence P2A sequence A12 IRES A4 (3.0 kb) GFP VHH CD28 CD3zeta IRES VHH HA CAR P2A Signal CD8 hinge and TM Signal sequence sequence A12 pCMV A4 (3.1 kb) GFP VHH CD28 CD3zeta pCMV VHH HA CAR T cell Signal CD8 hinge and TM Signal sequence P2A sequence

C A12 A12 P2A A4 A12 IRES A4 A12 pCMV A4 + + + GFP CAR cells GFP CAR cells + GFP CAR cells + 38.0 31.9 GFP CAR cells 40.9 6.81 A12 = anti–PD-L1 A4 = anti-CD47

CAR CAR CAR CAR

A12 pCMV A4 D A4 A12 P2A A4 A12 IRES A4 CAR A4-secreting A12 CAR T cell CD47 A12 CAR + excess soluble A4

A4 A12 CAR A4

Anti-CD47 mAb Anti-CD47 mAb Anti-CD47 mAb Anti-CD47 mAb 1. 2. 3. 4. 5. 6. E 1. 2. 3. 4.

1. Untransduced 1. Ladder 2. A12 2. A12 3. A12 IRES A4 3. A12 P2AA4 4. A12 pCM VA4 4. (+) control – 5. 1B7 pCMV A4 HA Ub 10 ng 6. HA-Ubiquin [+ control]

F B16 killing - 24 hours IFNγ production - 24 hours

100 20,000

80 15,000 None A12 (anti–PD-L1) 60 10,000 A12 P2A A4

(pg/mL) (anti-CD47–secreting γ

% Killing 40 5,000 anti–PD-L1 CAR) IFN 20 0

0 −5,000 1:2E:T 1:1E:T 5:1E:T 10:1E:T 1:2E:T 1:1E:T 5:1E:T 10:1E:T

Figure 1. Anti-CD47 VHH-secreting A12 CAR T cells can be generated and function in vitro. A, Schematic of VHH-secreting CAR T cell. B, Design of various VHH-secreting CAR constructs. Constructs consist of fluorescent protein tracers and a CAR separated by a P2A sequence. The secreted VHH is encoded either multi-cistronically, with a P2A sequence or IRES sequence, or with a separate CMV promoter. TM, transmembrane. C, Transduction of constructs traced by a fluorescent reporter protein. D, Flow cytometry–based assay to detect expression of A4 VHH. Populations analyzed included transduced and untransduced cells and were stained with a fluorescent anti-CD47 mAb that competes for binding with A4. A4-secreting CAR T cells showed blockade of anti-CD47 binding. E, Anti-HA immunoblot on lysate of A12 P2A A4 CAR T cells shows production of A4 VHH (left). Anti-HA IP on supernatant of A12 A4-secreting CAR T cells shows secretion of A4 (right). F, B16F10 killing is measured with a CellTiter Glo–based assay (left). CAR T-cell activation is measured by IFNg ELISA (right). E:T, effector-to-target ratio.

AACRJournals.org Cancer Immunol Res; 8(4) April 2020 521

Xie et al.

A (Host background/tumor type/CAR T-cell background) Day 0 Day 2 Day 6 Day 12 Day ? Four groups: all TA 99 treated Sample size PD-L1KO /B16WT/None n =4 KO WT n KO PD-L1 /B16 /None + sA4 =4 PD-L1 mouse B16 CAR CAR CAR Measure PD-L1KO / B16WT /A12PD-L1KO+sA4 n =4 injecon (1× 107) (1 ×× 107) (9 106) survival PD-L1KO / B16WT /A12P2AA4PD-L1KO n =7 (5× 105) Daily injections of A4 VHH for sA4 groups

B Survival Tumor area average Tumor area individual 150 150 2

2 None 100 None + sA4 100 100 A12 CAR + sA4 *** A12 P2A A4 ** 50 50 50 A12 = anti–PD-L1 A4 = anti-CD47 Tumor area mm *** Tumor area mm Percent survival 0 0 0 0 20406080 020406080 020406080 Days postinoculation Days postinoculation Days postinoculation C D Survival Day 0 Day 2 Day 7 Day 14 Day ? 100 None A12 ** A12 A4 *

B16 CAR CAR CAR Measure 50 A12 = anti–PD-L1 WT mouse 6 6 6 injecon (5 × 10 ) (4 × 10 ) (4 × 10 ) survival A4 = anti-CD47 (5 × 105)

Percent survival 0 010203040 Days postinoculation Anti-CD47 Anti-HA E F (Host background/tumor type/CAR T-cell background) A4-secreting B2 CAR (B2P2AA4) Four groups: all TA99 treated Sample size WT B2 CAR + excess WT/B16 /None n =5 soluble A4 WT/B16WT/B2 n =5 B2 CAR WT/B16WT/B2P2AA4 n =8

Survival G H 100 None Day 0 Day 4 Day 9 Day 18 Day ? B2 ** B2P2AA4 * 50 B2 = anti-EIIIB WT mouse B16 CAR CAR CAR Measure A4 = anti-CD47 injecon (8.5 × 106) (8.5 × 106) (9× 106) survival

Percent survival 0 (1 × 105) 0 10203040 Days postinoculation

Figure 2. Treatment with anti-CD47 VHH-secreting CARs further delays tumor growth in syngeneic, immunocompetent tumor models. A, Schematic of in vivo experiment to determine efficacy of A12 CAR T cells that secrete A4 VHH. B, Kaplan–Meier curves (left), average tumor growth curves (middle), and individual tumor growth curves (right) of mice treated with no CARs, daily systemic doses of A4 VHH, A12 CAR T cells together with systemic A4 VHH, or with A12 CAR T cells that secrete A4 VHH (P ¼ 0.0001; A12 þ sA4/A12P2AA4, P ¼ 0.002; None þ sA4/A12P2AA4, P ¼ 0.0008; None/A12P2AA4, P ¼ 0.0006). All groups received the anti-TA99. Experiments were repeated twice. C, Schematic of the in vivo experiment to test the efficacy of A12P2AA4 CAR T cells in WT mice. D, Kaplan–Meier curves show significant improvement in survival with A12P2AA4 CAR T-cell treatment (None/A12P2AA4, P ¼ 0.0011; A12/A12P2AA4, P ¼ 0.043). E, B2 CAR T cells that secrete A4. Anti-CD47 staining and anti-HA staining show that A4 VHH is secreted. Experiments were repeated twice. F, In vivo experimental groups to test the efficacy of B2 A4 CAR T-cell treatment. G, Schematic of the in vivo experiment for treatment with B2P2AA4 CAR T cells. H, Kaplan–Meier curves and individual tumor growth curves (B2/B2P2AA4, P ¼ 0.0168; None/B2P2AA4, P ¼ 0.0002). Experiments were repeated twice. , P ≤ 0.05; , P ≤ 0.01; , P ≤ 0.001; , P ≤ 0.0001.

(Fig. 2B). However, the A4-secreting A12 CAR T cells provided a and tumor sizes were measured (Fig. 2C). TA99 treatment was given to significant survival benefit when compared with all other conditions. all groups. In these WT, immunocompetent models, even at lower Secretion of (anti-CD47) A4 by CAR T cells further inhibits, but does doses, A12 P2A A4 CAR T cells also show improved antitumor activity not fully control, tumor growth in this aggressive, syngeneic tumor compared with A12 CAR T cells (Fig. 2D). model (Fig. 2B). The antitumor effects of A12 P2A A4 CAR T cells A nanobody, B2 that recognizes a splice variant, EIIIB, of were also evaluated in WT mice. WT C57BL/6 mice were treated with fibronectin detects the tumor neovasculature and tumor stroma A12 CAR T cells or A4-secreting A12 CAR T cells, or left untreated, and was used to generate CAR T cells (27, 28). In similar fashion, we

522 Cancer Immunol Res; 8(4) April 2020 CANCER IMMUNOLOGY RESEARCH

VHH-Secreting CAR T Cells

A B 100 Tumor area individual None + sA4 2 Day 0 Day 2 Day 6 Day 12 Day 21 80 A12 + sA4

60 A12 P2A A4

KO PD-L1 mouse B16 CAR CAR CAR Harvest lymphoid 40 A12 = anti–PD-L1 injecon (1× 107) (1×× 107) (9 106) organs and tumors A4 = anti-CD47 (5× 105) 20

Daily injecons of A4 VHH for sA4 groups Tumor area mm 0 0 5 10 15 20 25 CDDays postinoculation - Spleen - %CD4 CAR Spleen - %CD8+ CAR Spleen - CD11b+ cells - Spleen - CD11c+ cells CD3 2.0 4 40 CD3 1.5 + None + sA4 + None + sA4 1.5 A12 + sA4 30 A12 + sA4 T cells 3 T cells + + A12P2AA4 1.0 A12P2AA4 1.0 2 20 of CD45 of CD45 + CD4 + CD8 + + 0.5 0.5 1 10

0.0 0 0 % CAR % CAR %CD11c %CD11b 0.0 + LN - %CD8 CAR - LN - %CD4+ CAR LN - CD11b+ cells - LN - CD11c+ cells 1.0 3 None + sA4 6 0.5 CD3 CD3 None + sA4 + 0.8 A12 + sA4 + T cells 0.4

T cells A12 + sA4 + A12P2AA4 + 2 0.6 4 A12P2AA4 0.3 CD8

CD4 0.4 of CD45 + of CD45

+ 0.2

1 +

2 + 0.2 0.1

0.0 % CAR 0 % CAR 0 0.0 %CD11b %CD11c

E F B16 B16 PD-L1KO G ELISPOT None + sA4 A12 + sA4 None + sA4 300 A12 P2A A4

200 splenocytes

A12 + sA4 6 100

A12 P2A A4 0 SFU per 10 B16 B16 PD-L1–KO

Figure 3. Secretion of anti-CD47 VHHs by CAR T cells is not detrimental to CAR T-cell survival and increases the presence of innate immune cells. A, Schematic of the experiment to evaluate changes in immune populations mid-treatment. B, Individual tumor growth curves of mice treated with either systemic A4 VHH, A12 CAR T cells together with systemic A4 VHH, or with A12 CAR T cells that secrete A4 VHH. All mice received anti-TA99. C, Flow cytometry was used to detect the presence of A12 CAR T cells or A12 A4-secreting CAR T cells in the spleen and draining lymph nodes (LN). Secretion of A4 by A12 does not affect the persistence of CAR T cells. D, Numbers of CD11bþ and CD11cþ cells in the CD45þCD3– population in the spleen and draining lymph nodes (LN) of treated mice. Treatment with A4-secreting A12 CAR T cells shows a slight, although not statistically signiﬁcant, increase of CD11bþ and CD11cþ cells. E, Tumors isolated from experiment described in D. One of the tumors from mice treated with A4-secreting A12 CAR T cells appeared to lack melanocytes, consistent with epitope spreading to TRP1þ cells. F, ELISPOT assays were performed by coculturing splenocytes with B16F10 cells or with B16F10 PD-L1–KO cells. Activity of splenocytes on B16F10 PD-L1–KO cells indicates that epitope spreading may have occurred, as the CAR T cells recognize PD-L1. G, Quantiﬁcation of ELISPOT data in F. SFU, spot-forming units.

developed EIIIB-targeted CAR T cells that secrete A4 VHH Effect of anti-CD47 VHH secretion by CAR T cells on their (B2P2AA4 CARs) to explore the A4 secretion strategy in a survival and on innate immune cells different model. Flow cytometry shows that these B2-A4 CAR T To determine whether secretion of (anti-CD47) A4 inhibits expan- cells secrete A4 VHH (Fig. 2E). We tested an in vivo tumor model in sion and persistence of the CAR T cells, we monitored immune cell WT mice that compared no treatment with treatment with either B2 populations in the course of treatment (Fig. 3A). Mice were given daily CAR T cells or B2P2AA4 CAR T cells (Fig. 2F). Mice were doses of A4 systemically, received A12 CAR T cells in conjunction with inoculated with B16F10 tumors on day 0, treated with the various daily doses of A4, or received A4-secreting A12 CAR T cells (Fig. 3B). CAR T cells on days 4, 9, and 18, and survival was measured On day 21 after tumor inoculation, lymphoid organs and tumors were (Fig. 2G). Secretion of the A4 VHH by B2 CAR T cells improves harvested and ﬂow cytometry was performed on cells recovered from þ þ survival when compared with treatment with B2 CAR T cells only in lymphoid organs. The numbers of CD8 and CD4 CAR T cells in this immunocompetent mouse model (Fig. 2H). Although the spleen and draining lymph nodes were comparable for mice treated secretion of anti-CD47 VHH by both B2 and A12 CAR T cells with A12 CAR T cells or with the A12 P2A A4 CAR T cells (Fig. 3C). improved outcomes in an aggressive, syngeneic tumor model, Secretion of the (anti-CD47) A4 VHH therefore did not affect per- þ þ tumors eventually relapsed. sistence of the CAR T cells. CD11b and CD11c cells showed a slight,

AACRJournals.org Cancer Immunol Res; 8(4) April 2020 523

Xie et al.

B A 12345 1. Ladder GFP VHH CD28 CD3zeta P2A VHH 2. Positive control – HA Ub 50 ng/mL 3. Positive control – HA Ub 500 ng/mL P2A Signal CD8 hinge and TM Signal sequence sequence 4. B2 (63% transduced) 15 kDa 5. B2 P2A A12 (55% transduced)

C PD-1 CTLA-4 TIM3 LAG3

A12-secreting B2 CAR (B2 P2A A12) D Day 0 Day 4 Day 11 Day 18 B2 CAR

Untransduced T cells WT mouse B16 CAR CAR CAR injection (1.2 ×× 107) (1 107) (1 × 107) B2 = anti-EIIIB (1 × 105) A12 = anti–PD-L1

E Tumor - %CAR CD8 T cells Spleen - %CAR CD8 T cells dLN - %CAR CD8 T cells 30 1.5 **** 3 **** **** None B2 CAR 20 1.0 T cells B2 P2A A12 CAR T cells + T cells + 2 + 10 0.5 CD8 CD8 + CD8 + + 1 0.0 0 %CAR %CAR %CAR -0.5 -10 0 Tumor - %CAR CD4 T cells Spleen - %CAR CD4 T cells dLN - %CAR CD4 T cells 5 * 0.4 **** None 1.5 *** 4 B2 CAR 0.3 T cells B2 P2A A12 CAR + T cells

T cells 3 + 1.0 + 0.2 2 CD4 + CD4 CD4 0.1 + + 1 0.5 0.0 0 %CAR %CAR %CAR -1 -0.1 0.0

F GFP VHH CD28 CD3zeta P2A H11 Fc G 1234 1. Ladder P2A Signal CD8 hinge and TM Signal sequence sequence 50 2. B2 37 3. B2 H11Fc H PD-1 PD-L1 4. HA-Ub 250 ng (positive)

15 H11Fc-secreting B2 CAR (B2 H11Fc) B2 CAR

B2 = anti-EIIIB H11 = anti–CTLA-4

I Spleen - %CAR CD8 T cells dLN - %CAR CD8 T cells Tumor - %CAR CD8 T cells

* 3 15 4 ** None None + sH11FC 3 T cells T cells T cells 2 10

+ B2 + sH11Fc + +

2 B2 H11Fc CD8 CD8 CD8 + + + 1 5 1 %CAR %CAR %CAR 0 0 0 Spleen - %CAR CD4 T cells dLN - %CAR CD4 T cells Tumor - %CAR CD4 T cells * 0.8 * 4 1.0 **** None None + sH11FC 0.8 0.6 3 T cells T cells

+ B2 + sH11Fc + T cells

+ 0.6 0.4 2 B2 H11Fc CD4 CD4 + + CD4 0.4 + 0.2 1 0.2 %CAR %CAR %CAR 0.0 0.0 0

J GFP VHH CD28 CD3zeta P2A H11 Fc P2A A4

P2A CD8 leader CD8 hinge and TM Signal sequence Signal sequence

KLAnti-CD47 Anti-HA 1. 2. 3. 1. Ladder A12H11FcA4 CAR 2. A12H11FcA4 A12 CAR + soluble A4 (10:1) 50 kDa (~18% transduced) 3. A12 A12 CAR (~25% transduced)

CD47 HA

524 Cancer Immunol Res; 8(4) April 2020 CANCER IMMUNOLOGY RESEARCH

VHH-Secreting CAR T Cells

although not statistically significant, increase in mice that received the significantly impacted (Fig. 4E; Supplementary Fig. S3B). In similar A12 P2A A4 CAR T cells, consistent with greater involvement of the fashion, we generated B2 CAR T cells that secrete an anti–CTLA-4 innate immune system (Fig. 3D). For each group of mice, tumors were VHH-Fc fusion (B2-H11Fc; Fig. 4F). Flow cytometry–based compe- excised on day 21. One of the tumors extracted from the A12 P2A A4 tition assays confirmed that the B2-H11Fc CAR T cells indeed secrete treatment group showed a loss of melanin, suggesting that epitope H11Fc (Supplementary Fig. S3C), as verified also by immunoblotting spreading had occurred (Fig. 3E). To further examine this possibility, (Fig. 4G). Staining of B2-H11Fc CAR T cells for exhaustion markers we performed an ELISPOT assay and cocultured splenocytes from shows less PD-L1 and PD-1 than on B2 CAR T cells (Fig. 4H). Analysis each mouse with either B16F10 WT cells or B16F10 PD-L1–KO cells of the lymphoid organs and tumors of treated tumor-bearing mice (Fig. 3F): the latter lack PD-L1, the target molecule recognized by the showed that the B2-H11Fc–secreting CAR T cells show better per- þ þ A12 CAR. If epitopes other than PD-L1 were recognized, splenocytes sistence of both CD4 and CD8 CAR T cells (Fig. 4I). Administra- should also respond to the B16F10 PD-L1–KO cells. The ELISPOT tion of the B2-H11Fc–secreting CAR T cells did not significantly results for splenocytes obtained from mice that received the A4- improve survival, compared with treatment with a combination of B2 secreting A12 CARs, as well as from mice that received the A12 CARs CAR T cells and systemically delivered H11Fc, but was significantly together with systemic administration of A4 VHH showed reactivity better than no treatment (Supplementary Fig. S3D). Although, the data against both the B16F10 and B16F10 PD-L1–KO lines (Fig. 3F and G). establish that these CAR T cells secrete anti–CTLA-4 (H11Fc), the This shows that one or more antigens other than PD-L1 are recognized amount released may not have been sufficient to provide an additional by T cells in the CAR T-cell recipients, consistent with epitope survival benefit. spreading. We further demonstrated the modularity of this approach by generating CAR T cells that secrete multiple VHHs or VHH fusions. Modular and multi-cistronic expression and secretion of VHHs As a proof of concept, we generated A12 CAR T cells that secrete both and VHH fusions by CAR T cells an anti–CTLA-4 VHH-Fc fusion and an anti-CD47 VHH, encoded by An advantage of using VHHs as the secreted output of CAR T cells a single vector (Fig. 4J), as validated biochemically and by cytofluori- lies in their stability and solubility. With these properties, VHH- metry. Anti-CD47 flow cytometry assays show that A4 VHH is indeed secreting CARs can be made modular. We have generated other secreted (Fig. 4K). An anti-HA IP on the supernatant of cultures VHH-secreting CAR T cells to improve their persistence through confirmed that H11-Fc is secreted as well (Fig. 4L). We did not checkpoint blockade. We have produced EIIIB-specific CAR T cells evaluate the antitumor activity of such double-secreting CAR T cells. that secrete an anti–PD-L1 VHH (A12) with a design similar to that of the A4-secreting CAR T cells (Fig. 4A). To confirm secretion of the Localized delivery of A4-Fc by CAR T cells limits toxicity and A12 VHH, we performed a competition assay similar to that described delays tumor growth for the A4-secreting CAR T cells (Supplementary Fig. S3A). T cells Because the A12 (anti–PD-L1) CAR T cells localize to and can be transduced with the A12-secreting B2 CAR are not stained by the retained in the tumor environment due to the presence of PD-L1, the PD-L1 mAb, showing that secreted A12 blocks recognition by anti– cognate antigen recognized by A12, they can produce and serve as a PD-L1. Immunoblotting confirmed release of the A12 VHH (Fig. 4B). delivery vehicle for biologics. An advantage of such localized delivery is Transduced cell populations, which are a mixture of transduced T the possible avoidance of side effects caused by therapeutics applied cells and unmodified bystander T cells, were stained for the exhaustion systemically. Secretion of an anti-CD47 VHH by CAR T cells was markers PD-1, CTLA-4, TIM3, and LAG3. Populations that contain beneficial in the B16F10 animal model, but CD47 blockade is more the (anti–PD-L1) A12-secreting CAR T cells show less staining for effective if there is a signal to drive phagocytosis (18). In the B16F10 these markers compared with CAR T cells that do not secrete the A12 model, inclusion of the TA99 antibody, which targets TRP1, provides VHH (Fig. 4C). To determine whether these in vitro results correlate this signal. However, in the MC38 model and for several other tumor with improved persistence in vivo, C57BL/6 mice were inoculated with models, such enhancing antibodies are not currently available. There- B16F10 tumors and treated with either B2 CAR T cells, A12-secreting fore, we generated PD-L1–targeted CAR T cells that secrete an A4-Fc B2 CAR T cells, or left untreated (Fig. 4D). The spleens, draining fusion to block CD47 while at the same time providing an activating lymph nodes, and tumors of each mouse were harvested after the IgG2a Fc portion to enhance phagocytosis of anti-CD47–decorated tumor reached >125 mm2, and CAR T cells were enumerated. The targets (Fig. 5A). Anti-human CD47 mAbs have shown promise in A12-secreting B2 CAR T cells showed improved persistence and preclinical xenograft models, but in syngeneic settings, the widespread expansion compared with the B2 CAR T cells, but survival was not expression of CD47 on RBCs in particular has caused complications

Figure 4. Secretion of VHHs and VHH-fusions by CAR T cells is modular, and multi-cistronic constructs can be produced. A, Vector design of A12 VHH-secreting B2 CAR T cells. B, Anti-HA IP on supernatant from the culture of B2 CAR T cells or B2 A12-secreting CAR T cells. TM, transmembrane. C, Flow cytometry for common exhaustion markers in populations of untransduced cells and cells transduced with B2 CARs or B2 A12-secreting CAR T cells. D, Experimental schematic to determine how A12 secretion affects persistence of B2 CAR T cells. E, Flow cytometry to determine the presence of CD8þ and CD4þ CAR T cells in the spleen, tumor-draining lymph nodes, and tumors (spleen CD4, P ¼ 0.0001; spleen CD8, P < 0.0001; draining lymph node CD4, P < 0.0001; draining lymph node CD8, P < 0.0001; tumor CD4, P ¼ 0.0042; tumor CD8, P < 0.0001). dLN, draining lymph node. F, Vector design of anti–CTLA-4 H11Fc-secreting B2 CAR T cells. Experiments were repeated twice. TM, transmembrane. G, Anti-HA IP on supernatant from cells transduced with either B2 CAR or B2 H11Fc-secreting CAR. Samples are probed with an anti-HA-HRP– labeled antibody. H, Staining for common exhaustion markers on populations of cells transduced with either B2 CAR or B2 H11Fc-secreting CARs. I, Following a similar experimental setup as in D, we analyzed the persistence of B2 and H11Fc-secreting B2 CAR T cells in the spleen, draining lymph nodes (dLN), and tumors. The B2 H11Fc CAR T cells showed improved survival over B2 CAR T cells or B2 CARs given systemic doses of H11Fc (spleen CD4, P < 0.0001; spleen CD8, P ¼ 0.0058; dLN CD4, P ¼ 0.0053; dLN CD8, P ¼ 0.0254; tumor CD4, P ¼ 0.0088; tumor CD8, P ¼ 0.0019). Experiments were repeated twice. J, Vector design of a multi-cistronic construct to generate CAR T cells with a VHH-Fc fusion and an additional VHH on a single plasmid. TM, transmembrane. K, Staining with anti-CD47 shows inhibition of binding. Staining with an anti-HA mAb shows production of HA-tagged VHH bound to the T-cell surface, indicating that A4 VHH is being produced. L, Anti-HA IP on supernatant from cells transduced with an H11Fc- and A4-secreting CAR T cell shows that H11Fc is secreted. , P ≤ 0.05; , P ≤ 0.01; , P ≤ 0.001; , P ≤ 0.0001.

AACRJournals.org Cancer Immunol Res; 8(4) April 2020 525

Xie et al.

A B A12 A12-A4Fc VHH-Fc fusion CAR T cell

GFP VHH CD28 CD3zeta P2A VHH Fc

P2A Signal CD8 hinge and TM Signal sequence sequence CAR T cells CAR T cells 27.7 35.5 C Anti-CD47 Anti-IgG2a Anti-HA

A4Fc-secreting A12 CAR (A12-A4Fc) A12 CAR + excess A12 = anti–PD-L1 soluble A4 A4 = anti-CD47 A12 CAR

CD47 IgG2a HA 12345 D E CellTiter Glo 24 hours IFNg ELISA 24 hours 100 60,000 A12 1. Ladder 80 A12-A4Fc 2. A12 50 kDa 40,000

3. A12-A4Fc ng 60 li l 4. HA-Ub 25 ng i K 40 5. HA-Ub 250 ng % 20,000 IFN g (pg/mL) 20 15 kDa 0 0 1:1 E:T 5:1 E:T 10:1 E:T 1:1 E:T 5:1 E:T 10:1 E:T

F (Host background/tumor type /CAR T cell) Bleed Five groups: Sample size Day 0 Day 2 Day 6 Day 12 Day 15 PD-L1KO/B16WT /None n =4 PD-L1KO/B16WT /A12 PD-L1KO n =4 KO B16 CAR PD-L1KO/B16WT /None + sA4Fc n =3 PD-L1 mouse CAR CAR Measure injection (1× 107) (9 ×× 106) (9 106) PD-L1KO/B16WT /A12 PD-L1KO+sA4Fc n =4 (5× 105) + A4Fc injection PD-L1KO/B16WT /A12-A4FcPD-L1KO n =6 GH% Starting weight RBC count I HGB level 120 15 **** 20 **** 110 None 15 10 A12 CAR 100 cells/ m L 6 10 None + sA4Fc 90 5 A12 + sA4Fc 80 **** 5 A12-A4Fc HGB level g/dL RBC *10 % Starting weight 70 0 0 051015 Day post Tx JK(Host background/tumor type/CAR T cell) %A4Fc-bound circulating RBC Three groups: Sample size KO 150 PD-L1 /MC38/None n =5 None A12 = anti–PD-L1 PD-L1KO/MC38/A12 PD-L1KO n =5 A4 = anti-CD47 A12 CAR 100 PD-L1KO/MC38/A12-A4FcPD- L1KO n =6 None + sA4Fc A12 + sA4Fc Day 0 Day 2 Day 6 Day 12 Day 22 Day? 50 A12-A4Fc KO

% RBC bound by A4Fc 0 PD-L1 MC38 CAR CAR CAR CAR Measure mouse (3 × 105) (1× 107) (9 × 106) (6 ×× 106) (6 106) survival Survival N

100 None + + - ++ -

+ Tumor - CD11b CD11c cells Tumor - CD11c CD11b A12 80 8 None

** ** CD3 A12-A4Fc * + A12 60 6 A12-A4Fc

50 of CD45 + 4 40 of CD45 ++ CD11c + 20 2

Percent survival 0 0102030

Days postinoculation 0 %CD11c 0 %CD11b M Tumor area individual Tumor - CD45+CD3+CD69+ cells Tumor - CD45+CD3-CD69+ cells 200 2 None 100 - 80 ** * None 150 A12 CD3 90 + A12 A12-A4Fc 60 A12-A4Fc

100 T cells 80 + 40

70 of CD45 50 +

Tumor area mm 20 %CD69 60 0 0102030 50 %CD69 0 Days postinoculation

526 Cancer Immunol Res; 8(4) April 2020 CANCER IMMUNOLOGY RESEARCH

VHH-Secreting CAR T Cells

and prevented further exploration (24). By delivering anti-CD47 with calipers. A concern was the possibility that binding of the A4-Fc to VHH-Fc fusions locally to the tumor microenvironment, we might the CAR T cells themselves could lead to early elimination and thus avoid such side effects. Anti–PD-L1 CARs that secrete A4Fc fusions compromise their therapeutic efficacy. However, upon analysis of the were generated using a single vector containing a fluorescent reporter of lymphoid organs and tumors at the endpoint of the MC38 experi- protein, a CAR linked to a self-cleaving peptide sequence, and a ments, the A12-A4Fc CAR T cells persisted, although to a slightly secreted VHH-IgG2a Fc fusion with an HA tag (Fig. 5A). This lesser, not statistically significant, degree than the A12 CAR T cells construct was transduced into T cells and adequately expressed (Supplementary Fig. S4). Phagocytosis likely requires a balance of (Fig. 5B). To determine whether the transduced cells secrete (anti- signals, and blockade of CD47 alone may not be sufficient to ensure CD47) A4Fc, we performed flow cytometry–based assays. Staining elimination (31, 32). We saw an improvement in survival of the MC38- with a fluorescently labeled anti-CD47 that competes for binding with inoculated mice upon treatment with the A12-A4Fc CAR T cells in A4 shows that the A4Fc-secreting CAR T cells block binding of the comparison with no treatment or treatment with A12 CARs (Fig. 5L fluorescent anti-CD47 (Fig. 5C). The extent of CD47 blockade is and M). Within the tumors in the A12-A4Fc–treated group, the þ þ similar to that of cells incubated in the presence of excess soluble A4 number of CD11b CD11c cells showed an increase, one of the VHH. The population of A4Fc-secreting CARs binds both an anti- indicators of the presence of M1-like macrophages (Fig. 5N; þ IgG2a and an anti-HA. The A4-Fc fusion is thus secreted and can bind refs. 33, 34). We also saw slightly more CD11c cells, as well as þ þ to CD47 on the T-cell surface (Fig. 5C). Because these cells comprise significantly more CD69 T cells and CD45 cells in the groups treated both transduced and untransduced cells, secreted A4Fc can bind to with A12-A4Fc CAR T cells (Fig. 5N). These observations suggest that bystander cells in amounts sufficient to block all available CD47 local secretion of the A4Fc generates a more inflammatory environ- molecules. This may increase tumor killing efficacy. We further ment within the tumor, improving the activation of immune cells. confirmed secretion of A4Fc by immunoblot (Fig. 5D) and detected the presence of a polypeptide of approximately 50 kDa, consistent with the predicted size of the A4-Fc fusion. Discussion To determine whether secretion of A4Fc affects killing by A12 CAR CAR T-cell therapies have shown success as single-agent therapeu- T cells, we performed a coculture assay with B16F10 cells. Killing of tics in hematologic cancers, but an immunosuppressive environment B16F10 and production of IFNg by A12-A4Fc and A12 CAR T cells poses a challenge for their use in the treatment of solid tumors. was comparable (Fig. 5E). To address possible toxicity of anti-CD47 Antibody-based approaches, alone or in combination with other treatment, immunocompetent mice received either systemic treatment modalities can establish a favorable antitumor environment. (anti-CD47) A4Fc (Supplementary Fig. S3), (anti–PD-L1) A12 CAR However, each additional component may also cause side effects and T cells, (anti–PD-L1) A12 CAR T cells together with systemic (anti- complicate production and implementation. Because CAR T cells are CD47) A4Fc, (anti-CD47) A4Fc-secreting (anti–PD-L1) A12 CAR T living therapies that can be programmed to execute different functions, cells, or no treatment at all. We measured body weight to gauge the we developed CAR T cells that can secrete immune-modulating VHHs toxicity of each treatment (Fig. 5F). Systemic delivery of A4Fc or A12 and VHH fusions to improve their therapeutic efficacy in immuno- CAR T cells together with systemic A4Fc led to rapid weight loss competent animals. (Fig. 5G). However, treatment with the A4Fc-secreting A12 CAR T VHHs are single-domain antibodies that can perform a wide range cell showed no significant weight loss, comparable with untreated of functions depending on their targets. Because of their small size and mice. Mice from each group were bled 3 days after treatment, and a stability, they can be produced at high yields and are readily secreted in complete blood count was done. Both groups that received systemic functional form. Because VHHs are derived from immunoglobulin A4Fc injections showed a drop in RBC numbers and hemoglobin heavy chain–only antibodies, they can traverse the secretory pathway levels, but the group with CAR T-cell–mediated delivery of A4Fc did without the need for additional partner proteins. Where simultaneous not (Fig. 5H and I). A large fraction, around 80%, of circulating RBCs production of multiple single-chain variable fragments (scFv) in a were decorated with systemically delivered A4Fc, whereas circulating single cell can compromise their assembly and secretion through RBCs from mice treated with the A4Fc-secreting CAR T cells showed domain swaps between the VH and VL portions of the different scFvs, no such binding (Fig. 5J). To determine whether A4Fc-secreting VHHs have no such drawback (35). This facilitates the generation of A12 CAR T cells provide an antitumor benefit, we examined the CAR T cells that make use of more than a single VHH. CAR T cells MC38 tumor model (Fig. 5K). We inoculated mice with MC38 tumors serve as ideal vehicles for VHH delivery through local secretion, so that and treated the recipients with either A12 CAR T cells, A12-A4Fc CAR VHHs target and accumulate in the tumor microenvironment. This T cells, or left mice untreated. We measured tumor sizes every 1–2 days strategy not only limits potential adverse effects due to off-tumor–

Figure 5. Localized delivery of A4-Fc by CAR T cells limits toxicity and delays tumor growth in immunocompetent animals. A, Schematic of vector design for A4Fc-secreting CAR T cells. TM, transmembrane. B, Transduction of A12-A4Fc–secreting CAR T cells can be traced with a fluorescent reporter. Both the A12 and A12-A4Fc were transduced to a similar extent. C, Secretion of A4Fc detected by flow cytometry. Staining with an anti-CD47 mAb shows that the A4Fc-secreting CAR T cells block the mAb. Staining with anti-IgG2a and anti-HA shows that the secreted molecule is bound to the T-cell surface. D, An anti-HA IP on the supernatant of cultured cells shows that A4Fc is secreted into the supernatant. E, CellTiter Glo measuring B16F10 killing (left) and IFNg ELISA from coculture with B16F10 cells (right). E:T, effector-to- target ratio. F, Design of the experiment to examine safety of A4Fc-secreting CARs in vivo. G, Weight loss curves of untreated mice or mice treated with a single dose of systemic A4Fc, A12 CAR T cells, A12 CAR T cells together with systemic A4Fc, or A12 CAR T cells that secrete A4Fc (P < 0.0001). Tx, treatment. RBC count (H), hemoglobin (HGB) content (P < 0.0001; I), and percentage of RBCs bound by A4Fc (J) of mice in each treatment group 3 days after injection of first treatment. K, Design of experiment to examine efficacy of A4Fc-secreting CAR T cells in an MC38 tumor model. Kaplan–Meier curves (L) and tumor growth curves (M) of mice treated with A12 CAR T cells, treated with A12-A4Fc–secreting CAR T cells, or left untreated (A12/A12-A4Fc, P ¼ 0.0308; None/A12-A4Fc, P ¼ 0.0007). Experiments were repeated twice. N, Flow cytometry analysis of CD11bþCD11cþ cells, CD69þCD45þCD3–, and CD69þCD45þCD3þ T cells within each treatment group (CD11bþCD11cþ, P ¼ 0.0052; CD11c high, P ¼ 0.115; CD69þCD45þCD3þ, P ¼ 0.0139; CD69þCD45þCD3–, P ¼ 0.0025). , P ≤ 0.05; , P ≤ 0.01; , P ≤ 0.001; , P ≤ 0.0001.

AACRJournals.org Cancer Immunol Res; 8(4) April 2020 527

Xie et al.

targeting caused by systemic exposure, but also produces a higher local CAR T cells. A single retroviral vector can encode for the secretion of concentration of VHH, which should be helpful where complete multiple VHHs and VHH-fusions from the same CAR T cell. blockade of a target in the tumor microenvironment is required. In conclusion, we have shown that secretion of VHHs by CAR T Blockade of CD47 can increase phagocytosis of antibody-decorated cells can improve their antitumor activity in immunocompetent hosts targets by macrophages and other myeloid cells (21). However, the with syngeneic tumor models. Treatment with an anti-CD47 VHH- ubiquitous expression of CD47 on self-cells causes off-tumor side secreting CAR T cell enhanced survival over treatment with system- effects that are dose-limiting for clinical application of anti-CD47 ically delivered anti-CD47 VHH along with a CAR T cell. Localized treatment. Dose-escalation studies have shown that treatment-induced secretion of VHHs and VHH-Fc fusions by CAR T cells results in an anemia is of particular concern and prevents attaining the requisite improved safety profile when compared with systemic administration dose of anti-CD47 needed for increased tumor killing in some of these antibody derivatives and can contribute to improved persis- cases (36). We show it is possible to confine the presence of the agent tence of CAR T cells. Personalized, multi-modal treatments tailored mostly to the tumor environment and thus prevent potential systemic toward the characteristics of each tumor indication thus become toxicities by CAR T-cell–mediated local delivery of anti-CD47 VHHs possible. or anti-CD47 VHH-Fc fusions. These studies are performed in immunocompetent hosts, which should more accurately represent systemic Disclosure of Potential Conflicts of Interest toxicities, if any, than the frequently used immunocompromised M. Dougan is a consultant for Tillots Pharma and Partner Therapeutics, is a models. Because phagocytosis by macrophages depends on a balance scientific advisory board member for Neoleukin Therapeutics, and reports receiving a fl between stimulatory and inhibitory signals, delivery of the anti-CD47 commercial research grant from Novartis. No potential con icts of interest were disclosed by the other authors. VHH without an Fc portion requires an additional cue for phagocy- – tosis (31, 32). In the case of the B16F10 melanoma model, the TRP-1 Authors’ Contributions specific TA99 mAb can provide that phagocytic stimulus (37). How- Conception and design: Y.J. Xie, M. Dougan, J.R. Ingram, H.L. Ploegh ever, in less well-characterized models, or in models where such Development of methodology: Y.J. Xie, T. Fang, H.L. Ploegh antibodies are not readily available, it would be beneficial to have a Acquisition of data (provided animals, acquired and managed patients, provided molecule that can simultaneously provide a stimulatory signal and facilities, etc.): Y.J. Xie, N. Pishesha block inhibitory interactions. We therefore generated CAR T cells that Analysis and interpretation of data (e.g., statistical analysis, biostatistics, secrete anti-CD47 VHH-Fc fusions. Treatment with A4Fc-secreting computational analysis): Y.J. Xie, M. Dougan þ Writing, review, and/or revision of the manuscript: Y.J. Xie, M. Dougan, A12 CAR T cells shows an accumulation of CD11c macrophages H.L. Ploegh within the tumor, suggesting an accumulation of proinflammatory þ þ Administrative, technical, or material support (i.e., reporting or organizing data, M1-type macrophages (33, 34). An increase in CD69 CD45 cells constructing databases): Y.J. Xie, N. Momin also indicates enhanced leukocyte activation with inflammatory activ- Study supervision: Y.J. Xie ity. Secretion of an anti-CD47 VHH or VHH-Fc fusion by CAR T cells Other (made essential contributions to the generation and characterization of the thus inhibits tumor growth, while at the same time avoiding the nanobodies and the conceptualization of the VHH-based CAR T cells that were used in this study): J.R. Ingram toxicities associated with systemic anti-CD47 treatment. However, we believe that engineering CAR T cells to secrete the anti-CD47 VHH, Acknowledgments paired with administration of a separate Fc engagement molecule is This research was supported by the Lustgarten Foundation Grant 80939 (to H.L. preferable. This solution avoids the potential risk of increased clear- Ploegh) and the Melanoma Research Alliance (to H.L. Ploegh). Y.J. Xie was supported ance of VHH-Fc–secreting T cells, as CD47 is present on the T-cell by a National Science Foundation Graduate Research Fellowship. M. Dougan was surface. supported by the NIH Mentored Clinical Scientist Development Award Different tumor indications have diverse immune characteristics, 1K08DK114563-01 and the American Gastroenterological Association Research with variable immune infiltration, immunosuppressive capabilities, Scholars Award. N. Pishesha was supported by the Harvard Junior Fellowship. We fi thank Andrew Woodham for technical assistance with the ELISPOT assay. We thank and in ltrating cell types. An ideal CAR T-cell therapy should be Laura Kummer for her technical assistance in amplifying viral DNA. We thank designed to tackle particular aspects of each tumor, with modifications Richard Hynes for helpful discussion in work regarding the B2 CAR T cell. We are that should be easy to engineer. To demonstrate the flexibility of VHH- grateful to Steve Kolifrath for his assistance in animal care. We thank Nicholas secreting CAR T cells, we generated several additional VHH-secreting McCaul for his assistance in revisional experiments. We thank the flow cytometry core CAR T cells based on a standard vector design. In each case, we find at Boston Children's Hospital (BCH) Karp. We are grateful to the Lodish laboratory that VHHs are secreted effectively. For both the PD-L1 VHH- and for providing the XZ vector. We thank Dane Wittrup for providing the TA99. CTLA-4 VHH-Fc fusion–secreting CAR T cells, cells appear less The costs of publication of this article were defrayed in part by the payment of page exhausted in the course of their generation. In vivo, this translates to charges. This article must therefore be hereby marked advertisement in accordance improved persistence and survival of the CAR T cells. Tumor-bearing with 18 U.S.C. Section 1734 solely to indicate this fact. mice treated with the A12-secreting B2 CAR T cells and the H11Fc- secreting B2 CAR T cells show more CAR T cells in the spleen, draining Received September 25, 2019; revised November 13, 2019; accepted January 29, lymph nodes, and tumors when compared with treatment with B2 2020; published first February 4, 2020.

References 1. Barrett DM, Singh N, Porter DL, Grupp SA, June CH. Chimeric antigen receptor 4. Newick K, Moon E, Albelda SM. Chimeric antigen receptor T-cell therapy for therapy for cancer. Annu Rev Med 2014;65:333–47. solid tumors. Mol Ther Oncolytics 2016;3:16006. 2. Jena B, Dotti G, Cooper LJN. Redirecting T-cell speciﬁcity by introducing a 5. Maus MV, June CH. Making better chimeric antigen receptors for adoptive tumor-speciﬁc chimeric antigen receptor. Blood 2010;116:1035–44. T-cell therapy. Clin Cancer Res 2016;22:1875–84. 3. Maude SL, Frey N, Shaw PA, Aplenc R, Barrett DM, Bunin NJ, et al. Chimeric 6. Gilham DE, Debets R, Pule M, Hawkins RE, Abken H. CAR-T cells and solid antigen receptor T cells for sustained remissions in leukemia. N Engl J Med 2014; tumors: tuning T cells to challenge an inveterate foe. Trends Mol Med 2012;18: 371:1507–17. 377–84.

528 Cancer Immunol Res; 8(4) April 2020 CANCER IMMUNOLOGY RESEARCH

VHH-Secreting CAR T Cells

7. Whiteside TL. The tumor microenvironment and its role in promoting tumor 23. WillinghamSB,VolkmerJP,GentlesAJ,SahooD,DalerbaP,MitraSS,etal. growth. Oncogene 2008;27:5904–12. The CD47-signal regulatory protein alpha (SIRPa) interaction is a thera- 8. Dougan M, Dranoff G, Dougan SK. Cancer immunotherapy: beyond checkpoint peutic target for human solid tumors. Proc Natl Acad Sci U S A 2012;109: blockade. Annu Rev Cancer Biol 2018;3:55–75. 6662–7. 9. Yeku OO, Brentjens RJ. Armored CAR T-cells: utilizing cytokines and pro- 24. Huang YT, Ma YC, Gao P, Yao Z. Targeting CD47: the achievements and inflammatory ligands to enhance CAR T-cell anti-tumour efficacy. Biochem Soc concerns of current studies on cancer immunotherapy. J Thorac Dis 2017;9: Trans 2016;44:412–8. E168–E74. 10. Yeku OO, Purdon TJ, Koneru M, Spriggs D, Brentjens RJ. Armored CAR T cells 25. Chao MP, Weissman IL, Majeti R. The CD47-SIRP alpha pathway in cancer enhance antitumor efficacy and overcome the tumor microenvironment. Sci Rep immune evasion and potential therapeutic implications. Curr Opin Immunol 2017;7:10541. 2012;24:225–32. 11. Koneru M, Purdon TJ, Spriggs D, Koneru S, Brentjens RJ. IL-12 secreting tumor- 26. Advani R, Flinn I, Popplewell L, Forero A, Bartlett NL, Ghosh N, et al. CD47 targeted chimeric antigen receptor T cells eradicate ovarian tumors in vivo. blockade by Hu5F9-G4 and rituximab in non-Hodgkin's lymphoma. N Engl J Oncoimmunology 2015;4:e994446. Med 2018;379:1711–21. 12. Muyldermans S, Baral TN, Retamozzo VC, De Baetselier P, De Genst E, Kinne J, 27. Xie YJ, Dougan M, Jailkhani N, Ingram J, Fang T, Kummer L, et al. Nanobody- et al. Camelid immunoglobulins and nanobody technology. Vet Immunol based CAR T cells that target the tumor microenvironment inhibit the growth of Immunopathol 2009;128:178–83. solid tumors in immunocompetent mice. Proc Natl Acad Sci U S A 2019;116: 13. Revets H, De Baetselier P, Muyldermans S. Nanobodies as novel agents for cancer 7624–31. therapy. Expert Opin Biol Th 2005;5:111–24. 28. Jailkhani N, Ingram JR, Rashidian M, Rickelt S, Tian C, Mak H, et al. 14. De Meyer T, Muyldermans S, Depicker A. Nanobody-based products as research Noninvasive imaging of tumor progression, metastasis, and fibrosis using a and diagnostic tools. Trends Biotechnol 2014;32:263–70. nanobody targeting the extracellular matrix. Proc Natl Acad Sci U S A 2019; 15. Vincke C, Loris R, Saerens D, Martinez-Rodriguez S, Muyldermans S, 116:14181–90. Conrath K. General strategy to humanize a camelid single-domain antibody 29. Bousquet M, Harris MH, Zhou B, Lodish HF. MicroRNA miR-125b causes and identification of a universal humanized nanobody scaffold. J Biol Chem leukemia. Proc Natl Acad Sci U S A 2010;107:21558–63. 2009;284:3273–84. 30. Berggren WT, Lutz M, Modesto V. General spinfection protocol. In: StemBook 16. Ikeda H, Chamoto K, Tsuji T, Suzuki Y, Wakita D, Takeshima T, et al. The critical [Internet]. Cambridge (MA): Harvard Stem Cell Institute; 2008. role of type-1 innate and acquired immunity in tumor immunotherapy. 31. Gordon S. Phagocytosis: an immunobiologic process. Immunity 2016;44: Cancer Sci 2004;95:697–703. 463–75. 17. Mantovani A, Sica A. Macrophages, innate immunity and cancer: balance, 32. Ravichandran KS. Find-me and eat-me signals in apoptotic cell clearance: tolerance, and diversity. Curr Opin Immunol 2010;22:231–7. progress and conundrums. J Exp Med 2010;207:1807–17. 18. Sockolosky JT, Dougan M, Ingram JR, Ho CC, Kauke MJ, Almo SC, et al. Durable 33. Lumeng CN, DelProposto JB, Westcott DJ, Saltiel AR. Phenotypic switching of antitumor responses to CD47 blockade require adaptive immune stimulation. adipose tissue macrophages with obesity is generated by spatiotemporal differ- Proc Natl Acad Sci U S A 2016;113:E2646–54. ences in macrophage subtypes. Diabetes 2008;57:3239–46. 19. Ingram JR, Blomberg OS, Sockolosky JT, Ali L, Schmidt FI, Pishesha N, et al. 34. Zhu Y, Zhang L, Lu Q, Gao Y, Cai Y, Sui A, et al. Identification of different Localized CD47 blockade enhances immunotherapy for murine melanoma. macrophage subpopulations with distinct activities in a mouse model of oxygen- Proc Natl Acad Sci U S A 2017;114:10184–9. induced retinopathy. Int J Mol Med 2017;40:281–92. 20. Chao MP, Alizadeh AA, Tang C, Myklebust JH, Varghese B, Gill S, et al. Anti- 35. Gil D, Schrum AG. Strategies to stabilize compact folding and minimize CD47 antibody synergizes with rituximab to promote phagocytosis and eradicate aggregation of antibody-based fragments. Adv Biosci Biotechnol 2013;4: non-Hodgkin lymphoma. Cell 2010;142:699–713. 73–84. 21. Majeti R, Chao MP, Alizadeh AA, Pang WW, Jaiswal S, Gibbs KD, et al. CD47 is 36. Tao H, Qian P, Wang F, Yu H, Guo Y. Targeting CD47 enhances the efficacy of an adverse prognostic factor and therapeutic antibody target on human acute anti-PD-1 and CTLA-4 in an esophageal squamous cell cancer preclinical model. myeloid leukemia stem cells. Cell 2009;138:286–99. Oncol Res 2017;25:1579–87. 22. Michot JM, Bigenwald C, Champiat S, Collins M, Carbonnel F, Postel-Vinay S, 37. Bevaart L, Jansen MJ, van Vugt MJ, Verbeek JS, van de Winkel JG, Leusen JH. The et al. Immune-related adverse events with immune checkpoint blockade: a high-affinity IgG receptor, FcgammaRI, plays a central role in antibody therapy comprehensive review. Eur J Cancer 2016;54:139–48. of experimental melanoma. Cancer Res 2006;66:1261–4.

AACRJournals.org Cancer Immunol Res; 8(4) April 2020 529

Improved Antitumor Efficacy of Chimeric Antigen Receptor T Cells that Secrete Single-Domain Antibody Fragments

Yushu Joy Xie, Michael Dougan, Jessica R. Ingram, et al.

Cancer Immunol Res 2020;8:518-529. Published OnlineFirst February 4, 2020.

Updated version Access the most recent version of this article at: doi:10.1158/2326-6066.CIR-19-0734

Supplementary Access the most recent supplemental material at: Material http://cancerimmunolres.aacrjournals.org/content/suppl/2020/02/04/2326-6066.CIR-19-0734.DC1

Cited articles This article cites 36 articles, 13 of which you can access for free at: http://cancerimmunolres.aacrjournals.org/content/8/4/518.full#ref-list-1

Citing articles This article has been cited by 1 HighWire-hosted articles. Access the articles at: http://cancerimmunolres.aacrjournals.org/content/8/4/518.full#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department Subscriptions at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerimmunolres.aacrjournals.org/content/8/4/518. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.