Oleandrin Induces Immunogenic Cell Death Via the PERK/elF2α/ATF4/CHOP Pathway in Breast Cancer

Background Chemotherapeutic agents have been linked to immunogenic cell death (ICD) induction that is capable of augmenting antitumor immune surveillance. The cardiac glycoside oleandrin, which inhibits Na + /K + -ATPase pump (NKP), has been shown to suppress breast cancer growth via inducing apoptosis. However, the implications of oleandrin in antitumor immune response and potential ICD induction remain unexplored till now, which is investigated in the present study. Methods Calreticulin (CRT) exposure was detected by immunouorescence and ow cytometry, and high mobility group protein B1 (HMGB1) and Adenosine Triphosphate (ATP) secretion was quantied by ELISA and both the intracellular and extracellular expression of Heat shock protein 70/90 (HSP70/90) was detected by western blotting in breast cancer cells treated with oleandrin. Dendritic cells (DCs) were co-cultured with oleandrin-treated breast cancer cells before the expressions of activation markers and cytokines were examined by quantitative real time polymerase chain reaction (qRT-PCR), ELISA, and ow cytometry. Immune activation effects of oleandrin were determined in murine breast cancer model using BALB/C mice. The differential mRNA expression incurred by oleandrin was investigated by mRNA sequencing and subsequently conrmed by qRT-PCR and Western blotting. Results Oleandrin treatment induced CRT exposure on cell surface and the release of HMGB1, HSP70/90 and ATP. The maturation and activation of DCs were increased by co-culturing with oleandrin-treated cancer cells, which subsequently enhanced CD8 + T cell cytotoxicity. In animal models, oleandrin inhibited tumor growth and increased tumor inltrating lymphocytes including DCs and T cells. Mechanistically, oleandrin induced endoplasmic reticulum (ER) stress associated caspase independent immunogenic cell death (ICD) mainly through PERK/elF2α/ATF4/CHOP Activation of IRE1 but ATF6, the other two was also BAP31: CHOP, DDIT3: 3; ATF4: activating transcription factor 4; UPR: unfolded protein response; PERK: protein kinase R–like ER kinase; IRE1: inositol requiring enzyme 1α; ATF6: activating transcription factor 6; eIF2α: eukaryotic translation initiation factor 2α; GADD34, PPP1R15A: growth arrest and DNA damage-inducible transcript 34.

Results Oleandrin treatment induced CRT exposure on cell surface and the release of HMGB1, HSP70/90 and ATP. The maturation and activation of DCs were increased by co-culturing with oleandrin-treated cancer cells, which subsequently enhanced CD8 + T cell cytotoxicity. In animal models, oleandrin inhibited tumor growth and increased tumor in ltrating lymphocytes including DCs and T cells. Mechanistically, oleandrin induced endoplasmic reticulum (ER) stress associated caspase independent immunogenic cell death (ICD) mainly through PERK/elF2α/ATF4/CHOP pathway. Activation of IRE1 pathway but not ATF6, the other two canonical sensors of ER stress, was also observed.
Conclusion Oleandrin triggered ER stress and induced ICD-mediated immune destruction of breast cancer cells. Oleandrin combined with immune checkpoint inhibitors might improve the e cacy of immunotherapy.

Background
Breast cancer is the most common malignant tumor occurring in women. With the incidence and mortality rate of breast cancer increasing annually, this malignancy has remained as one of the most serious threats to women's health. In 2019, over 2 million cases of breast cancer were reported worldwide, and more than half a million patient death was attributed to breast cancer (1). Treatment options of breast cancer include surgery, radiotherapy, chemotherapy, endocrine therapy, and targeted therapy (2).
However, treatment outcome is still far from satisfactory, especially for triple-negative breast cancers that lack effective treatment targets.
In recent years, immunotherapies represented by immune checkpoint inhibitors have made remarkable achievements in tumor treatment. Phase III clinical trial has showed that paclitaxel combined with atezolizumab, a programmed death-ligand 1 (PD-L1) blocking antibody, has signi cantly prolonged the progression free survival and total survival of participants with metastatic triple-negative breast cancer compared with paclitaxel combined with placebo group (3). The application of atezolizumab has been approved for the treatment of advanced breast cancer. Several studies have also shown that the number and co-stimulation of tumor in ltrating lymphocytes, especially CD4 + T, CD8 + T cells and dendritic cells (DCs), are indicators of curative effects (4). Therefore, immune regulation holds great potential in the treatment of breast cancer and enhancement of tumor lymphocytes in ltration will likely show bene cial effects.
The immune system functions routinely to eliminate dead cells during normal cell turnover, infections and injures (5). Certain chemotherapeutic agents such as anthracyclines, oxaliplatin and paclitaxel can trigger tumor cell death via enhance immune destruction via eliciting the release of damage-associated molecular patterns (DAMPs) to enhance tumor immunogenicity. Immunogenic cell death (ICD) is a cell death process characterized by the upregulation of various DAMPs (6). Calreticulin (CRT), high mobility group box 1 (HMGB1), adenosine triphosphate (ATP) and heat shock protein 70/90 (HSP 70/90) belong to DAMPs. As an "eat me" signal, CRT attracts antigen-presenting cells (APCs) to phagocytize the dead cells. ATP acts as a " nd me" signal, which leads to immune cell in ltration into tumor sites (7-10).
Na + /K + -ATPase pump (NKP) is a transmembrane ion transporter expressed in various cells such as neurons and cardiomyocytes (11). NKP serves as a multifunctional signal transducer that is essential for regulating cell apoptosis, in ammation, adhesion and maintaining cell homeostasis (12). NKP consists of α, β and γ subunits. The α1 and α3 subunits of NKP are frequently overexpressed in various cancers, such as colorectal cancer, glioblastoma, and breast cancer (13,14). Interestingly, cardiac glycosides, such as NKP inhibitors that are used for many years in the treatment of cardiac congestion, were recently showed to have potentials in the treatment of cancer (15). Digoxin, a cardiac glycoside, has been reported to trigger ICD in osteosarcoma cells (16). Retrospective analysis also revealed that digoxin combined with standard chemotherapy prolonged the overall survival of patients with breast cancer, colon cancer, head and neck cancer and hepatocellular carcinoma (17).
In the present study, we report that oleandrin, a monomer compound extracted from the leaves of Nerium oleander that belongs to the cardiac glycoside family (18)(19)(20), suppresses breast cancer cell growth by inducing ICD and leading to increased immune destruction of tumor cells. The effects of oleandrin were linked to ER stress mainly through PERK/elF2α/ATF4/CHOP pathway. Activation of IRE1 pathway but not ATF6, the other two canonical sensors of ER stress, was also observed. Our ndings reveal the potential of oleandrin in cancer treatment via regulating antitumor immune activation.
In vitro cytotoxicity assay 2×10 4 of MDA-MB-231 cells were seeded in 6-well plate and divided into the following 5 groups: cells cocultured with DCs, cells co-cultured with CD8 + T cells, cells pretreated with oleandrin, pretreated cells and co-cultured with DCs, pretreated cells co-cultured with DCs and CD8 + T cells. MDA-MB-231 cells alone were used as control group. After 48 h of co-culture, cells were washed twice with PBS to remove the immune cells and oleandrin. Cells were continued to grow for 14 days. Finally, cells were xed and stained with crystal violet for 15 min. The numbers of colonies were counted. University animal center. 2×10 5 of EMT6 were inoculated into mammary fat pads of BALB/c mice. After 7 days, both long and short diameters of the tumors reached about 5 mm. The mice were randomly divided into 3 groups with 5 mice in each group: PBS as control group, oleandrin treatment with 0.3 mg/kg and 0.6 mg/kg groups. Oleandrin was administered intraperitoneally every day. Tumor volume was measured every day and quanti ed as 0.5 × length × width × width. After 7-day administration, mice were sacri ced, and tumors were weighed. Tumor primary cells and splenocytes were harvested for ow cytometric analysis.
Flow cytometric analysis 1x10 6 of cells were collected and suspended in 100 μl of PBS. Cells were incubated with the following antibodies 5 μl /test of each at room temperature for 30 min, PerCP anti-human HLA-DR, APC anti-human CD11c, FITC anti-human CD80, PE anti-human CD86.

Western blotting
Cells were collected and lysed using radio immunoprecipitation assay (RIPA) buffer. Cell culture supernant was were centrifuged for 30 min with 10,000 xg and protein concentration was measured using the protein concentration kit (2772T, Thermo scienti c, Shanghai, China). Protein was separated by 10% sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to PVDF membrane. Primary antibodies were diluted and incubated at 4°C overnight. The membranes were then washed and incubated with secondary antibodies at room temperature for 1 h. Chemiluminescence was performed using Supersignal West Pico plus (Thermo Fisher Scienti c, Inc.) and detected by BIO-RAD

Statistical analysis
Statistical analysis was performed using the Statistical Package for the Social Sciences (SPSS) software version 21 (IBM Corp., Armonk, NY, USA). The results were represented as mean ± standard deviation. The comparison between the two groups was conducted by t-test or one-way ANOVA. p<0.05 was considered statistically signi cant (*, p<0.05, **, p<0.01).

Oleandrin induces ICD in breast cancer cells in vitro
Our previous study has established the IC50 concentrations of oleandrin against MCF7 and MDA-MB-231 cells, which were 14.5 nM and 24.62 nM, respectively (21). Therefore, the treatment concentrations of MCF7 and MDA-MB-231 cells were used at 15 nM and 25 nM in the following experiments. CRT exposure on cell surface, a marker for ICD, is an 'eat me' signal to stimulate DC maturation and immune activation. To investigate whether oleandrin induces ICD in breast cancer, MCF7 and MDA-MB-231 cells were treated with oleandrin for 10 h before immuno uorescence assays to stain CRT (Fig. 1A). Compare to the control, oleandrin treatment increased CRT expression on cell surface. Which was further con rmed by ow cytometry analyses. The proportion of PI -/CRT + subpopulation was compared between the control and oleandrin treatment group (Fig. 1B). Consistent with immuno uorescence results, oleandrin treatment increased cell surface CRT expression in both MCF7 and MDA-MB-231 cells. The release of other DAMPs, including HMGB1, ATP and HSP70/90, was also compared between the control and oleandrin-treated groups. The secreted HMGB1 was detected after 12 h and increased signi cantly at 24 h and 48 h in MCF7 cells following oleandrin treatment. No signi cant release of HMGB1 was observed in MDA-MB-231 cells at 12 h but increased signi cantly at 24 h and 48 h (Fig. 2A). Distinct from the results of HMGB1 release, ATP secretion was observed 4 h after oleandrin treatment in both MCF7 and MDA-MB-231 cells and reached the peaks at 12 h (Fig. 2B). Moreover, both the intracellular and extracellular expressions of HSP70/90 were detected, as presented in Fig. 2C, intracellular expressions of HSP70/90 were not affected by oleandrin in MCF7 and MDA-MB-231 cells. However, the extracellular expression of HSP70 and HSP90 were detected in both cells at 48 h after oleandrin treatment. These data indicated that oleandrin treatment triggered ICD in breast cancer cells.

Oleandrin induces ICD-associated immune activation in vitro
Tumor cell ICD is capable of inducing antigen presenting function of DCs. As a result, a series of immune responses are activated, including cytokine secretion and T cell activation (22). To eliminate the uncertainty caused by different human leukocyte antigen (HLA), we detected the HLA subtypes of MCF7, MDA-MB-231 cells and two volunteers. As shown in supplementary Fig. 1A and B, HLA subtype of MDA-MB-231 was HLA-A2 that matched that of the volunteer. To investigate whether oleandrin treatment induces DC maturation and activation, we cultured DCs isolated from PBMC and then co-cultured with MDA-MB-231 cells with or without oleandrin treatment. The morphologic features of DCs were observed, which showed that the adherent cells were round or oval on day 1 and extended branched projections on day 5 (sFig. 1C). Flow cytometry data con rmed that 99.37±0.15 % of cells were CD11c positive that indicated mature DCs on day 5 (Fig. 3B) Cell surface expressions of CD80, CD86, and HLA-DR were detected by ow cytometry and represented by Δ mean uorescence intensity (ΔMFI). Cells stained with isotype control were considered as control (Fig.  3B). IL-2, IL-10, and IFN-γ expressions in the co-culture supernatant were further con rmed by ELISA (Fig.  3C). Consistent with the qRT-PCR results, DCs co-cultured with oleandrin-treated MDA-MB-231 cells showed signi cantly enhanced levels of CD86, IL-2 and IFN-γ but decreased IL-10 expression. Taken together, these results suggest that co-culture with oleandrin-treated breast cancer cells signi cantly enhanced DC maturation and activation.

Oleandrin suppresses tumor growth in mouse model
To investigate the anti-tumor effects of oleandrin in vivo, BALB/c mice were used to implant EMT-6 cells into mammary fat pads. 7 days after implantation, tumor-bearing mice were divided into three groups and treated with oleandrin at 0.3 mg/kg and 0. 6mg/kg in peritoneal. Mice treated with vehicle were used as control (Fig. 4A). Compared with the control group, tumor growth was inhibited in the oleandrin treatment groups 1 day after administration. After 7 days of continuous administration, the average tumor size of 0.3 mg/kg treatment group was unchanged compared with day 0, while the average tumor size of 0.6 mg/kg treatment group was even smaller than that at day 0 (Fig. 4B, *p<0.05, **p<0.01). The average tumor weight of 0.6 mg/kg treatment group was 1.58 times lower than that of 0.3 mg/kg treatment group and was 2.66 times lower than that of the control group (Fig. 4 C and D, **p<0.01).

Oleandrin activates antitumor immune responses in mice
To investigate the alterations in the immune microenvironment after oleandrin treatment, splenocytes and tumor primary cells were collected and stained with T cell and DC markers. As presented in Fig. 5A-D, compared with the control group, oleandrin treatment groups showed increased populations of both CD4 + and CD8 + T cell in spleen in a dose-dependent manner ( Fig. 5A and B, **p<0.01). CD11b + /CD11c + or CD11b + /CD11ccells were not changed ( Fig. 5C and D).
Furthermore, tumor in ltrating lymphocytes (TILs) were analyzed by staining with CD45. Compared with the control group, oleandrin treatment increased the portion of CD45 + cells in a dose-dependent manner (Fig. 5E). Moreover, DC (CD45 + /CD11c + ) were also increased by oleandrin treatment (Fig. 5E and F, **p<0.01). Compared with CD45 + cells and CD11c + cells, the proportion of tumor in ltrating T cells were much less. Therefore, the absolute numbers of CD4 + and CD8 + T cells were analyzed. As shown in Fig.   5G and H, oleandrin treatment increased the numbers of both CD4 + and CD8 + T cells (**p<0.01). These data indicated that oleandrin treatment increased DC and T cell in ltration in the tumor sites.
Oleandrin-induced CRT exposure is independent of caspase activity.
Translocation of CRT from endoplasmic reticulum (ER) to cell surface is an indicator for ICD. Previous studies have revealed that chemotherapeutic agents such as anthracyclines and oxaliplatin can induce ICD in tumor cells via caspase-8 dependent pathway. Caspase-8 hydrolyzes the ER anchor protein B cell receptor-associated protein 31 (BAP31), which further transfers CRT to cell surface through Golgi apparatus (24,25). This procedure is essential, as blocking caspase 8 inhibits CRT exposure. Our previous study has shown that oleandrin induced breast cancer cell apoptosis. To explore whether oleandrin-induced ICD is dependent on caspase, MCF7 and MDA-MB-231 cells were pretreated with Z-VAD-FMK, a pan-caspase inhibitor, for 10 h before treated with oleandrin. Cell apoptosis and CRT exposure were detected by ow cytometry. As shown in Fig. 6A, Z-VAD-FMK signi cantly inhibited breast cancer cell apoptosis. However, CRT exposure was not affected (Fig. 6B). These data indicated that oleandrin-induced CRT exposure was likely not associated with caspase.

Oleandrin induces ICD via ER stress
To further explore the mechanism of oleandrin-induced ICD in breast cancer cells, MCF7, T47D and MDA-MB-231 cells were treated with oleandrin and differential mRNA expressions were analyzed by RNA sequencing. A total of 121 common signi cantly different genes in the three different pairs of cells were identi ed in the Venn's diagrams (Fig.7A). The most signi cantly changed genes were categorized into the ATF2 pathway (Fig.7B). The identi ed genes involved in ATF2 pathway were illustrated using a heatmap. As shown in Fig.7C, activating transcription factor 3 (ATF3), DNA damage inducible transcript 3 (DDIT3), and activating transcription factor 4 (ATF4) were reported as the downstream genes of the ER stress pathway. Their altered expressions were further validated by qRT-PCR (Fig. 7D). The mRNA expression levels of these three genes were confrmed to be up-regulated by oleandrin treatment. Consistent with qRT-PCR results, the protein expression levels of ATF3, ATF4 and CHOP (DDIT3) were also increased accordingly (Fig. 7D). These data indicated that oleandrin-induced ICD might be associated with ER stress.

Oleandrin induces ER stress in breast cancer
In mammalian cells, various cytotoxic stimuli can cause unfolded protein responses (UPRs), which function as an adaptive cellular program to sustain protein homeostasis to protect cells (26). Such imbalance incurred by various stimuli can cause ER stress. ER stress is initiated by three ER transmembrane sensors: protein kinase R-like ER kinase (PERK), inositol requiring enzyme 1α (IRE1) and activating transcription factor 6 (ATF6). Three sensors can further activate downstream signaling pathways. The expression of total PERK, its substrate eukaryotic translation initiation factor 2α (eIF2α) and ATF6 were not affected by oleandrin treatment (Fig. 8A and sFig. 1F). However, phosphorylation levels of PERK and eIF2α (S52) were enhanced 6 h after oleandrin treatment. In addition, the activation of ATF4, and the expressions of ATF4-dependent target protein (growth-arrest-and DNA-damage-induced transcript 34 (GADD34, PPP1R15A) and CHOP) were increased as well (Fig.8A). Previous studies have demonstrated that GADD34 associates with the broadly acting serine/threonine protein phosphatase 1 (PP1) to dephosphorylate eIF2α (27). This may be a feedback mechanism for cell homeostasis. Moreover, the expression of IRE1 and phosphorylation of IRE1 (S724) were signi cantly increased, and subsequently enhanced the downstream XBP1 expression (Fig. 8B).
As shown in Fig. 9, oleandrin, as a Na + /K + ATPase inhibitor, increases the concentration of Na + and activated Na + /Ca 2+ ion exchange channel on cell surface, which causes the in ux of Ca 2+ . Loss of cellular homeostasis and disruption of Ca 2+ leads to activation of ER stress associated pathways including PERK-elF2α-ATF4 and IRE1-XBP1. ER stress enhances the releasement of ATP and HMGB1.
Moreover, ER stress induces CRT exposure to the cell surface. The release of these DAMPs signals eventually leads to the enhancement of immune response.

Discussion
Nerium oleander is widely distributed in subtropical Asia, Mediterranean coast and southwest America. The owers and leaves are commonly used in folk medicine for the treatment of heart failure, leprosy, analgesic, malaria, ringworms, anti-in ammatory and indigestion. (28,29). Oleandrin is a monomer compound extracted from the leaves of Nerium oleander contains potent cardenolides which shows strong heart clinical effect (30,31). Recent studies have revealed that oleandrin has antitumor activities in various tumors such as ovarian cancer, glioma, colon cancer, osteosarcoma, bladder cancer, breast cancer and leukemia. The working mechanism involves the suppression of Akt phosphorylation and the inhibition of mTOR (32)(33)(34)(35)(36). The botanical drug candidates of oleandrin, Anvirzel™ and PBI-05204, have been tested in phase I clinical trials for the treatment of solid tumors. Pbi-05204 was administered under a dose of 0.2255 mg/kg/day, and no side effects above grade 3 occurred, which proved its safety in patients with advanced cancer (37). Our previous study demonstrated that oleandrin had obvious cytotoxic effects on MCF7 (Luminal A subtype), SK-BR-3 (HER2 + subtype) and MDA-MB-231 (TNBC subtype) cells, but no obvious inhibitory effect on MCF10a was observed (21).
DCs are potent antigen presenting cells and play an important role in ICD-mediated immune response. In this study, we show that DC maturation and activation were not affected by oleandrin treatment. However, CD80 expression increased signi cantly on DCs co-cultured with oleandrin-pretreated MDA-MB-231 cells.
CD80 along with CD86 that are markers of active DCs bind to CD28 on T cells and act as a costimulatory signal for T cell activation (38). Moreover, oleandrin treatment decreased IL-10 expression but increased IL-2 and IFN-γ secretion. As an immunosuppressive cytokine, IL-10 inhibits the function of DC cells and weakens the immune activities of CD4 + and CD8 + T cells (39). IL-2 plays an essential role for the proliferation of antigen speci c T cells (40). IFN-γ induces the differentiation of Th1 and NKT cells and is important for anti-tumor immunity (41). These data indicate that oleandrin triggered ICD and activated DC-mediated immune responses. Consistent with our hypothesis, cytotoxic activities of CD8 + T cells were enhanced by oleandrin pretreatment in vitro. In vivo experiments further con rmed that intraperitoneal administration of oleandrin at both 0.3 mg/kg and 0.6 mg/kg inhibited tumor growth. Oleandrin treatment increased the number of tumor in ltrating CD45 + cells including CD11c + DCs, CD4 + T cells and CD8 + T cells. The in ltrations of CD4 + and CD8 + T cells were dose dependent, especially in CD8 + T cells. Furthermore, we analyzed T cells in tumor-bearing mice using splenocyte preparations and found that oleandrin increased the portion of both CD4 + and CD8 + T cells in the spleen. These data suggested oleandrin treatment induced ICD and stimulated DC-mediated immune responses in vivo.
ICD activates anti-tumor immune response mainly by releasing DAMPs. CRT exposure, ATP secretion and HMGB1 release are all indispensable for ICD, meaning that the absence of one single of these ICD hallmarks abolishes its e cacy (6). In the present study, CRT exposure was observed in MCF7 and MDA-MB-231 breast cancer cells 6 hours after oleandrin treatment. Consistent with the previous study, CRT exposure is earlier than cell apoptosis, which indicates CRT exposure was not the consequent event of DNA damage (9). Endoplasmic reticulum, where CRT locates, plays a crucial role in the maintenance of intracellular signal transduction, protein synthesis and calcium homeostasis (42). The RNA sequencing results suggested that oleandrin-induced ER-stress was associated with ICD. Previous studies have shown that certain chemotherapy agents such as anthracyclines and oxaliplatin induce ER stress via activating the caspase-8 signaling pathway that regulates CRT exposure. Caspase-8 hydrolyzes ER binding protein BAP31, which releases CRT from ER to cell surface (24,25). However, in this study, Z-VAD-FMK, a caspase inhibitor, signi cantly inhibited the cell apoptosis, but had no signi cant effect on CRT exposure, indicating that the CRT exposure induced by oleandrin was not dependent on caspase pathway. Western blotting results further con rmed that oleandrin treatment activated ATF3 and CHOP via PERK-elF2α-ATF4 and IRE1-XBP1 pathways but not ATF6. ER stress culminates in the translocation of the CRT to the cell surface, thereby generating an 'eat-me' signal for DCs In the present study, ATP secretion was detectable after 6 h but HMGB1 release was con rmed after 24 h following oleandrin treatment. In the early stage of ER stress, in order to maintain cell homeostasis, cells activate autophagy which facilitates the release of ATP from dying cells (43,44). ATP, as a ' nd-me' signal, recruits DCs to the dying cells. Distinct from ATP secretion, HMGB1 was passively released into the extracellular space at later post-apoptotic time point, allowing HMGB1 to bind Toll-like receptor 4 on DCs and thus stimulate their antigen presentation functions (45). Moreover, previous study demonstrated that CHOP regulates the release of HMGB1 (46). Therefore, we speculated CRT exposure, ATP secretion and HMGB1 release were associated with ER stress (Fig. 9).
Immunotherapies represented by immune checkpoint inhibitors have shown promising therapeutic outcomes. PD-L1 overexpression, tumor mutation burden (TMB) and tumor in ltrating lymphocytes affect the e ciency of immunosuppressive agents (47). The number and function of tumor in ltrating lymphocytes, especially CD4 + T cells, CD8 + T cells and DCs, are key to the effectiveness of the immunotherapy (4). Immune checkpoint inhibitors combining with ICD-inducing agents might be a more effective approach, especially in the treatment of tumors lacking immune cell invasion that are referred as "cold tumor". Recently, it was demonstrated that immune checkpoint inhibitors increased the antitumor response while combined with ICD-inducing chemotherapy agents or radiation therapy (48,49). Several clinical trials that combine checkpoint inhibitors with ICD inducers are ongoing, and the hope is that these combinations will increase the number of patients who can bene t from checkpoint inhibitor therapies. Further study is needed to demonstrate that whether oleandrin combined with immune checkpoint inhibitors will improve the e cacy of immunotherapy and reduce the side effects of chemotherapeutics through dose reduction.

Conclusions
Oleandrin triggered ER stress and induced ICD-mediated immune destruction of breast cancer cells. Oleandrin combined with immune checkpoint inhibitors might improve the e cacy of immunotherapy. The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.