Turning anecdotal irradiation-induced anti-cancer immune responses into reproducible in situ cancer vaccines via disulfiram/copper-mediated enhanced immunogenic cell death of breast cancer cells

Irradiation (IR) induces immunogenic cell death (ICD) in tumors, but it rarely leads to the abscopal effect (AE). However, combining IR with immune checkpoint inhibitors has shown anecdotal success in inducing AEs. In this study, we aimed to enhance the IR-induced immune response and generate reproducible AEs using the anti-alcoholism drug disulfiram (DSF) and copper complex (DSF/Cu) via induction of tumor ICD. We measured ICD in vitro and in vivo. In mouse tumor models, DSF/Cu was injected intratumorally followed by localized tumor IR, creating an in situ cancer vaccine. We determined the anti-cancer response by primary tumor rejection and assessed systemic immune responses by tumor rechallenge and the occurrence of AEs, i.e., spontaneous lung metastasis. Additionally, we analyzed immune cell subsets and quantified proinflammatory and immunosuppressive chemokines/cytokines in the tumor microenvironment (TME) and blood of the vaccinated mice. Immune cell depletion was investigated for its effects on the vaccine-induced anti-cancer response. The results showed that DSF/Cu and IR induced more potent ICD under hypoxia than normoxia in vitro. Low-dose intratumoral injection of DSF/Cu and IR demonstrated strong anti-primary and -rechallenged tumor effects and robust AEs in mouse models. These vaccinations also increased CD8 + and CD4 + cell numbers while decreasing Tregs and myeloid-derived suppressor cells in the 4T1 model, and increased CD8+, DC, and decreased Treg cell numbers in the MCa-M3C model. Depleting both CD8 + and CD4 + cells abolished the vaccine’s anticancer response. Moreover, vaccinated tumor-bearing mice exhibited increased TNFα levels and reduced levels of immunosuppressive chemokines/cytokines. In conclusion, our novel approach generated an anti-cancer immune response, resulting in a lack of or low tumor incidence post-rechallenge and robust AEs, i.e., the absence of or decreased spontaneous lung metastasis in tumor-bearing mice. This approach is readily translatable to clinical settings and may increase IR-induced AEs in cancer patients.


BACKGROUND
The abscopal effect (AE) denotes the ability of a localized therapy, such as irradiation (IR), to initiate a systemic antitumor response against non-irradiated metastatic cancer far outside the primary treatment area.Evidence suggests that the AE is mediated by systemic antitumor immune responses [1,2].Although AEs following IR have been found in many cancer types [3], IR alone rarely induces AEs.Even when IR was combined with immunotherapy (e.g., anti-CTLA-4 or anti-PD1 antibodies), AEs were observed in few patients; 47 cases were reported in 6 years from 2012 to 2018.[4] AEs in patients with cancer receiving IR with or without immunotherapy have been reported anecdotally [5,6].IR induces immunogenic cell death (ICD) in various cancer cell types [7].ICD is a phenomenon mediated by immunostimulatory signals from apoptotic cells.It is inducible by chemotherapeutics such as anthracyclines, oxaliplatin, bortezomib, radiotherapy, and photodynamic therapy [8][9][10], leading to effective antitumor immunity [11][12][13][14][15].
The combination of IR and DSF/Cu induced more potent ICD in human breast and pancreatic cancer cells than either method alone.This IR-and DSF/Cu-induced ICD was determined and measured using ICD parameters, including apoptosis, CRT and HSP90 cell surface expression, and the release of HMGB1 and ATP [26].Based on these ndings, we tested whether IR and DSF/Cu-mediated enhanced ICD of breast cancer (BC) cells can act as an in situ cancer vaccine in preclinical mouse BC models to turn anecdotal irradiation-induced anticancer immune responses into a reproducible AE in cancer-bearing hosts.

Cell lines
Mouse 4T1 and human UACC812 BC cell lines were purchased from the American Type Culture Collection (ATCC) and the human MDA-MB-231 BC cell line was acquired from the Duke Comprehensive Cancer Center Cell Culture Facility.Human MDA-MB-231 and UACC812 cells were cultured in Dulbecco's modi ed Eagle's medium (DMEM; Corning, Corning, NY, USA) supplemented with 10% heat-inactivated Gemini Foundation fetal bovine serum (FBS; Gemini Bioproducts, LLC, West Sacramento, CA, USA).The mouse HER2/neu + MCa-M3C cell line, developed at Massachusetts General Hospital, was cultured in DMEM supplemented with 15% heat-inactivated FBS [27].Mouse triple-negative 4T1 cells [28] were cultured in RPMI 1640 medium (Corning) supplemented with 10% FBS.All the cells were cultured at 37°C in a 5% CO 2 atmosphere.

Chemical reagents and antibodies
Tetraethylthiuram disul de (disul ram, DSF) and copper chloride were purchased from Sigma-Aldrich (St. Louis, MO, USA).DSF and CuCl 2 were dissolved in DMSO and Milli-Q water, respectively.A stock solution of DSF (10 mM) was aliquoted and stored at 20°C for up to 1 year and freshly diluted with cell culture medium (in vitro assays) or PBS (in vivo assays) prior to use.

Cell proliferation and viability assays
Tumor cells (MDA-MB-231, UACC812, and 4T1) were seeded in 96-well plates at a concentration of 5 × 10 3 cells/well in 100 µL culture medium and incubated at 37°C and 5% CO 2 for 24 h.Next, culture medium containing DSF/Cu at the indicated concentrations was added.Following incubation at 37°C and 5% CO 2 for 24 h, the proliferative/viable cells were evaluated using an MTT assay (Sigma-Aldrich).
The IC 50 values for each cell line were calculated using GraphPad Prism 8.All experiments were performed in triplicate.

Apoptotic cell analysis
We seeded cells (3 × 10 5 cells/well) in 6-well plates (Corning) and treated with DSF/Cu at the indicated doses and time points.Apoptosis cells were detected by Annexin V/7-AAD Apoptosis Detection Kit (640922, BioLegend, UK).The percentage of apoptotic cells was determined as described [26].

Detection of calreticulin on the cell surface
We detected translocation of calreticulin to the cell surface by immuno uorescence staining and ow cytometry, as described [26].Determination of intracellular ATP, which re ects the level of extracellular ATP Six-well plates were seeded with 3 × 10 5 cells/well and treated with or without DSF/Cu, as described [26].

Irradiation
In vitro IR was performed on cells seeded in 6-well plates (3 × 10 5 cells/well in 2 mL culture medium) at 8, 12 Gy.The X-RAD 320 Biological Irradiator (Precision X-ray, Inc., North Branford, CT) was used for IR experiments.
In situ tumor vaccination Six-week-old female BALB/c and FVB mice were purchased from the Massachusetts General Hospital COX7 animal facility.To establish tumors, a single cell suspension of mouse 4T1 cells (3.5 × 10 5 per mouse) in RPMI1640 serum-free medium was subcutaneously (s.c.) injected into the hind legs of mice (day 0).When the tumors were palpable, we initiated a cycle of in situ tumor vaccination consisting of 3site intratumoral injections of DSF/Cu (at the indicated doses in 100 µL PBS on days 4 and 6) and a single dose of IR (12 Gy) delivered locally to each mouse tumor (on day 5).The same vaccination cycle was repeated on day 9. On day 9, the mice were rechallenged with 5 × 10 5 4T1 cells injected s.c.into the other side of the hind leg.The FVB mice were inoculated orthotopically with 1 × 10 6 MCa-M3C cells/per mouse in the left mammary fat pad and treated with DSF/Cu (1.5 µM/1 µM) and IR (12 Gy for cycles 1, 2, and 8 Gy for cycle 3).We measured tumor growth daily with a caliper and calculated the volume according to the following formula: V = 1/2 (longer diameter long × shorter diameter 2 ).At time of sacri ce, the tumors, lungs, and spleens were collected for analysis of immune cell subtypes, cytokines and cellular level metastasis.All animal studies were approved by the Institutional Animal Care and Use Committee.

Mouse primary tumor and lung metastasis sample preparations for histologic and ow analyses
For hematoxylin and eosin (H&E) staining, entire lung tissues (4T1 muse model) were collected from each mouse at sacri ce and formalin-xed and para n-embedded (FFPE).For ow cytometry analysis or detection of lung metastatic cancer cells using cell culture techniques, each tumor or lung (50% tissue of each lung, MCa-M3C mouse model), and spleen was collected at sacri ce.Primary tumors or lungs were minced into 3 × 3 mm pieces and digested with Collagenase IV (LS004188) (1 mg/mL PBS) (Worthington Biochemical Corp.) at 37°C for 1 h.Spleens were mechanically homogenized using a ne metal mesh net and then the red blood cells in the splenocyte suspension were lysed with ammonium-chloride-potassium (ACK) lysing.The digested tumor or splenocyte suspension were ltered through a 40 µM cell strainer to obtain single-cell suspensions for ow analysis.

Statistical analysis
Unless otherwise noted, data are presented as the mean ± SEM.We used one-way ANOVA, two-way ANOVA, studentized-range test, and Student's t-test to compare groups and paired samples.All statistical analyses were performed using GraphPad Prism 8. Differences were considered statistically signi cant when the p-value was < 0.05.

DSF/Cu induced ICD of mouse and human BC cells in a dose-and time-dependent manner
Exposure to DSF (0.01-2.5 µM) and a xed concentration of CuCl 2 (1 µM) for 24 h in vitro in mouse 4T1 and human MDA-MB-231 and UACC812 BC cells indicated that the half-maximal inhibitory concentration (IC 50 ) value of DSF was 0.268 µM for 4T1, 0.534 µM for MDA-MB-231, and 0.482 µM for UACC812 cells (Fig. 1A).When 4T1 cells were observed by optical microscopy, the tumor cells started to appear round when treated with 0.025-0.25 µM DSF and CuCl 2 (1 µM) and became prominent at concentrations ≥ 0.15 µM, with more oating cells (Fig. 1B).Flow cytometry analysis revealed that the increase in apoptotic cells was DSF/Cu -dose-dependent and time-dependent (Fig. 1C, D).The molecular characteristics of ICD were assessed in DSF/Cu-treated cells by determining their cell-surface expression or release of highly immunostimulatory DAMPs, including calreticulin (CRT), released ATP levels, and extracellular HMGB1.DSF/Cu induced CRT cell surface expression in a dose-dependent manner in both dead (7AAD + ) and dying (7AAD − ) 4T1 and MBA-MD-231 cells (Fig. 2A-C).DSF/Cu decreased intracellular ATP levels, re ecting increased extracellular release of ATP[26] in a dose-dependent manner for all three cell lines (Fig. 2D).DSF/Cu increased the release of extracellular HMGB1 in a dose-dependent manner in all three cell lines (Fig. 2E).
Next, we tested whether similar results could be obtained using different doses of DSF/Cu (0.2 µM/1 µM) and IR (12 Gy).As expected, under hypoxia, such doses of DSF/Cu and IR tended to induce more potent ICD of 4T1 cells than under normoxia (40.60 ± 9.25% vs. 25.90 ± 11.48%, p = 0.15) (Fig. 3F).4T1 cells treated under hypoxic conditions produced less TGF-β1, a critical marker of inhibitory immune cell activation (Fig. 3G).These results provide a rm rationale to test the hypothesis that a combination of DSF/Cu and IR may convert tumors into an in situ vaccine through the induction of strong ICD of tumor cells and reverse immunosuppressive TME in vivo.

Combination of intratumoral injection of DSF/Cu and localized tumor IR elicited a robust antitumor immune response in immunocompetent mice
Based on the data shown in Fig. 3, we reasoned that localized delivery of both DSF/Cu and IR could induce ICD effectively within the tumor, and that cells that undergo strong ICD should elicit a systemic antitumor immune response, resulting in regression of primary tumor growth and prevention of metastasis.To this end, after an initial dose titration in vivo, we tested the e cacy of low doses of DSF (1.5, 3, and 9 µM), with a xed Cu dose (1 µM) by multiple site-intratumoral injections (i.t.) of 100µL PBS containing DSF/Cu/per tumor and a single dose of localized tumor IR (12 Gy), de ned as in situ cancer vaccines, in syngeneic 4T1-derived tumors growing subcutaneously in the right hind legs of mice (Fig. 4A,   B).In total, 33.3% (5/15) of the mice received the in situ cancer vaccination with DSF/Cu + IR (at all 3 DSF/Cu doses) exhibited complete tumor rejection (Fig. 4C, D), whereas none of the 4T1-bearing mice receiving either vehicle DSF/Cu or IR alone) had complete tumor rejection (Fig. 4C, D).As expected, this approach resulted in an immunological memory response in the same set of mice, indicated by signi cantly reduced tumor formation (20-40%) than that in mice treated with IR only (100%) after 4T1 cell rechallenge (Fig. 4E).We repeated the same in situ vaccination approach for syngeneic MCa-M3Cderived tumors growing orthotopically in the left mammary fat pad of FVB mice.Tumor volumes remained steadily smaller in mice treated with DSF/Cu (1.5 µM/1 µM) and IR (12 Gy) or IR alone than in those treated with DMSO (Fig. 4F).The difference between these two groups was smaller than that in the 4T1 tumor model, perhaps because MCa-M3C is more sensitive to IR-mediated killing than 4T1, which is known to be radioresistant [28,34].Overall, combined treatment of DSF/Cu and IR had a therapeutic effect on multiple types of BC in mice.
In situ cancer vaccines by intratumoral injection of DSF/Cu and IR resulted in profound AEs on lung metastasis.
The antitumor immune response against primary tumors observed in highly metastatic 4T1 and MCa-M3C mouse tumor models led us to investigate whether such an immune response would have AEs, that is, whether tumor growth would be reduced outside the local treatment elds of DSF/Cu and IR [2].The lungs are common metastatic target sites for BC in humans and they were metastatic sites in 100% of specimens from both mouse models used in this study.Therefore, at the time of euthanizing mice (Fig. 4A, B), the lungs were collected.Hematoxylin and eosin (H&E) stained lung sections from each mouse were thoroughly histologically examined for the absence of metastasis, de ned as zero cancer cell (Fig. 5A, B), which was detected in most of the DSF/Cu + IR-treated mice and fewer DS/Cu-treated mice (Fig. 5A, B).The overall metastatic incidence of all DSF/Cu + IR treated groups (20-40%) was lower than that of either the IR (100%) or DSF/Cu alone treated group (20-60%) (Fig. 5B).In addition, primary tumors treated with DSF/Cu + IR showed extensive AEs by complete prevention of lung metastasis in 100% of MCa-M3C tumor bearing mice, as detected through a sensitive lung tissue culture technique (Fig. 5C, D).The AEs may be attributed to the degree of elimination of BCSCs by IR + DSF/Cu-or IRinduced ICD in primary tumors [26] followed by systemic immune responses as elucidated below.
Antitumor e cacy of the in situ cancer vaccine by localized delivery of DSF/Cu and IR was immune effector CD8 + and CD4 + T cell-dependent and modulated suppressive cytokines in the TME and peripheral blood of mice.
Next, we further evaluated that the anti-tumor e cacy of the in situ cancer vaccine is mediated by immune effector cells via depleting CD8+, CD4 + cells, or both (Fig. 7A) in the 4T1 mouse model.Depletion of both CD8 + and CD4 + cells completely abolished the anti-primary and -rechallenged tumor responses elicited by the in situ cancer vaccine, while depletion of either cell type caused only a modest reduction (Fig. 7B, C).Previously, we performed an initial DSF/Cu dose titration in vitro experiment and found that DSF/Cu is toxic to peripheral blood mononuclear cells (PBMC) at a dose of ~ 0.15/1µM, which decreased ~ 50% PBMC in vitro.To con rm only a low dose of DSF/Cu can be used to avoid toxicity to immune cells in TME continuously recruited from blood, we tested DSF/Cu at 27 /1µM, which is 3 times higher than 9 /1µM, i.t.administered instead of lower DSF doses (1.5-9 µM/1µM) in the 4T1 tumor model.Such a high DSF/Cu dose, which is highly likely toxic to immune cells in TME, combined with IR diminished all anti-primary and -rechallenged tumor effects in the treated mice (Fig. 7D, E).However, the exact optimal dose range of i.t.delivery DSF/Cu is yet to be determined.Moreover, i.t.DSF/Cu delivery was more effective than oral low dose DSF delivery, which only produced modest anti-primary and -rechallenged tumor effects (Fig. 7D, E).However, more dose ranges of DSF/Cu with administered routes as well as IR regimen should be further tested for optimization.Lastly, in situ cancer vaccination with DSF/Cu + IR increased proin ammatory cytokine TNF α level in TME and downregulated the expression levels of immunosuppressive chemokines/ cytokines including KC (also known as (C-X-C motif) ligand 1 (CXCL1)), IL-10, MCP-1 (Monocyte chemoattractant protein-1) and TGF-β in both TME and peripheral blood in 4T1 tumor-bearing mice (Fig. 7F, G), indicating its potential to reverse the immunosuppressive TME (cold tumor) into proin ammatory TME (hot tumor).

DISCUSSION
Metastatic BC is a major contributor to cancer-related mortality, even in patients diagnosed with earlystage disease.Most BC patients are diagnosed at early stages (Stages I, IIA, IIB, and IIIA) [35,36].All these treatments have achieved great local control of the disease.Nonetheless, ~ 30% of patients with earlystage disease eventually develop metastatic BC [37].Thus, prevention of metastatic BC is an urgent and unmet clinical need.
Radiation therapy uses intense energy beams to kill cancer cells or slow their growth by damaging their DNA.Approximately 50% of patients with cancer receive radiation therapy [38].Radiation therapy is the mainstay of treatment for BC at almost every stage, as it is an effective way to reduce the risk of postsurgery recurrence for stage I-III cancers and alleviate the symptoms caused by stage IV metastatic BC [39].Recently, radiation therapy has evolved from a local to a systemic therapy for cancer, owing to its ability to regulate the immune response.Irradiation may augment immune responses against both the target tumor and metastatic sites by modulating antitumor immunity through the release of tumor antigens, tumor DNA, and cytokines into the TME and induction of ICD [26,40].However, RT alone rarely induces strong systemic antitumor immune responses, such as AEs [4].Indeed, when IR is combined with immunotherapy using the anti-CTLA4 monoclonal antibody ipilimumab, AEs have occurred in melanoma and metastatic non-small cell lung cancer [1,41].Nevertheless, these are rare anecdotal clinical events [42].
To meet this unmet clinical challenge, combined irradiation and immunotherapy has gained considerable interest from researchers.Several studies have used irradiation for in situ tumor vaccination, wherein a patient's tumor is used as a source of tumor-speci c antigens to stimulate effective antitumor immune responses.The advantage of this approach is to generate antitumor immune responses using the most immunogenic, T cell-recognizable, and diversi ed tumor antigens, which are termed "private antigens."Private antigens are derived from patient-speci c mutated and differentiated proteins in a tumor.The immunity elicited by in situ tumor vaccination is expected to target all or multiple tumor cancer antigens; thus, this approach may address the challenge of treating heterogeneous cancers [43,44].Morris et al. reported that combining irradiation and an intratumoral IL2-linked tumor-associated antigen-speci c antibody (anti-GD2 Hu14.18K322A or anti-EGFR cetuximab) in mouse models of melanoma, neuroblastoma, and head and neck squamous cell carcinoma eradicated both large tumors and metastases and elicited T cell immune responses against primary tumors that can be further leveraged by anti-CTLA-4 T cell checkpoint blockade to reduce lung metastasis [44,45].Demaria et al. identi ed DNA exonuclease Trex1 as being an upstream regulator of irradiation-induced antitumor immunity.A proper dose of radiation-induced Trex1 induction, which activates the type-I interferon (IFN-I) pathway mediated via cyclic GMP-AMP (cGAMP) synthase (cGAS) and its downstream adaptor stimulator of interferon genes (STING), optimally stimulating antitumor-speci c CD8 + T response.When irradiation (8 Gy × 3) was combined with the immune checkpoint inhibitor (ICI) anti-CTLA-4, complete and durable regression of both the irradiated and non-irradiated mouse cell line TSA-derived tumor was observed [15].Greenberg et al. also demonstrated abscopal responses driven by anti-CTLA4 therapy and vaccination of irradiation-treated mouse B16 mouse cells via the pattern-recognition receptor cGAS-STING axis [46].Notably, all these studies involved ICIs or immune checkpoint blockers (ICBs), which have been revolutionary drugs for many cancers.However, ICIs have serious limitations: i) only a minority patients receive long-term bene ts (i.e., objective responders), ii) their use is limited by their toxicity, because severe immune-related adverse events can be irreversible and sometimes life-threatening, and iii) they may be too expensive for many patients.
Our study also reported IR-based in situ immunization in the absence of ICIs, it offers the following unique characteristics: i) we report an in situ vaccination composed of combined IR with DFS/Cu enhances IR-induced non-antigen/target-dependent ICD in many cancer types, including BC, pancreatic cancer, osteosarcoma, sarcoma, and melanoma[26](Data not shown) ; ii) this in situ vaccination alone induced profound reproducible AEs, measured by clinical resembling spontaneous lung metastasis in two mouse BC models; iii) our earlier data demonstrated that IR and DSF/Cu could induce ICD in differentiated or differentiating breast non-cancer stem cells (BCSCs), which are the root cause of cancer formation, progression, and metastasis [26]; robust immune responses against such a broad spectrum of tumor antigens, including antigen sources derived from BCSCs, may also explain why this approach could achieve robust reproducible AEs.
Our ndings hold great promise for a quick translation of this simple approach using a single fraction of IR (12 Gy) with intratumoral delivery of a low dose of the FDA-approved DSF (1.5 µM) with copper (1 µM) to turn an anecdotal irradiation-induced anticancer immune response into a reproducible AE for BC patients.

Figures
Figures

Figure 7 Anti
Figure 7