PD-1 immune checkpoint blockade enhances FLOT chemotherapy toxicity in oesophageal adenocarcinoma cells via immune-independent mechanisms

Chemotherapy upregulates immune checkpoint (IC) expression on the surface of tumour cells and IC-intrinsic signalling confers a survival advantage against chemotherapy in several cancer-types including oesophageal adenocarcinoma (OAC). However, the signalling pathways mediating chemotherapy-induced IC upregulation and the mechanisms employed by ICs to protect OAC cells against chemotherapy remain unknown. Longitudinal proling revealed that FLOT-induced IC upregulation on OE33 OAC cells was sustained for up to 3 weeks post-treatment, returning to baseline upon complete tumour cell recovery. Pro-survival MEK signalling mediated FLOT-induced upregulation of PD-L1, TIM-3, LAG-3 and A2aR on OAC cells promoting a more immune-resistant phenotype. Single agent PD-1, PD-L1 and A2aR blockade decreased OAC cell viability, proliferation and mediated apoptosis. Mechanistic insights demonstrated that blockade of the PD-1 axis decreased stem-like marker ALDH and expression of DNA repair genes. Importantly, combining single agent PD-1, PD-L1 and A2aR blockade with FLOT enhanced cytotoxicity in OAC cells. These ndings reveal novel mechanistic insights into the immune-independent functions of IC-intrinsic signalling in OAC cells with important clinical implications for boosting the ecacy of the rst-line FLOT chemotherapy regimen in OAC in combination with ICB to not only boost anti-tumour immunity but also to suppress IC-mediated promotion of key hallmarks of cancer that drive tumour progression. 5-year overall survival rates as low as 15-40% depending on tumour stage at clinical presentation 4 . Improvements in the ecacy of rst-line chemotherapy regimens were achieved through combining immune checkpoint blockade (ICB) with chemotherapy as depicted in the recent ndings from the phase III Checkmate 649 trial, which demonstrated the synergy between nivolumab and rst-line chemotherapy (FOLFOX and XELOX) in previously untreated OGJ patients (n=1,581), in which a signicant improvement in overall survival in patients with a PD-L1 combined positive score of 5 or greater was observed (14.4 months (nivolumab plus chemotherapy arm) vs. 11.1 months (chemotherapy arm)) 5 . Furthermore, the nivolumab plus chemotherapy arm also reduced the risk of death by 29% (HR, 0.71; 98.4% CI, 0.59-0.86; p<0.0001) 5 . The ndings from this trial highlight the potential synergy that can be exploited between chemotherapy and ICB in the rst-line setting. Rational explanations for the improvement in overall survival are likely attributed to the ICB-mediated reinvigoration of anti-tumour immune responses. Moreover, mounting evidence in the literature suggest that the FOLFOX/XELOX chemotherapy regimens may possess immunostimulatory properties with the


Introduction
Oesophageal adenocarcinoma (OAC) is the predominant subtype of oesophageal cancer in Western countries 1 . Moreover, OAC is an exemplar model of an obesity-driven cancer and as such its incidence is rapidly increasing in parallel with the rising level of obesity 2 . Response rates to the standard of care FLOT , oxaliplatin and docetaxel) chemotherapy regimen remain low with a complete pathologic response rate of 16.6% 3 and subsequent clinical outcomes are dismal with a median overall survival rate of 50 months 3 and 5-year overall survival rates as low as 15-40% depending on tumour stage at clinical presentation 4 . Improvements in the e cacy of rst-line chemotherapy regimens were achieved through combining immune checkpoint blockade (ICB) with chemotherapy as depicted in the recent ndings from the phase III Checkmate 649 trial, which demonstrated the synergy between nivolumab and rst-line chemotherapy (FOLFOX and XELOX) in previously untreated OGJ patients (n=1,581), in which a signi cant improvement in overall survival in patients with a PD-L1 combined positive score of 5 or greater was observed (14.4 months (nivolumab plus chemotherapy arm) vs. 11.1 months (chemotherapy arm)) 5 . Furthermore, the nivolumab plus chemotherapy arm also reduced the risk of death by 29% (HR, 0.71; 98.4% CI, 0.59-0.86; p<0.0001) 5 . The ndings from this trial highlight the potential synergy that can be exploited between chemotherapy and ICB in the rst-line setting. Rational explanations for the improvement in overall survival are likely attributed to the ICB-mediated reinvigoration of anti-tumour immune responses. Moreover, mounting evidence in the literature suggest that the FOLFOX/XELOX chemotherapy regimens may possess immunostimulatory properties with the potential ability to convert a 'cold non-in amed' into a 'hot in amed' tumour microenvironment. 5uorouracil and oxaliplatin chemotherapies which comprise the FLOT regimen, have been shown to induce immunogenic tumour cell death in several cancer types. Therefore, addition of ICB to the FLOT/XELOX regimen likely reinvigorated exhausted anti-tumour immunity and prevented exhaustion of chemotherapy-induced anti-tumour immune responses, producing synergistic ampli cation of antitumour immunity, which translated into durable clinical responses in gastroesophageal cancer patients.
Oxaliplatin induced immunogenic cell death in colorectal cancer 6 and lung carcinoma 7,8 , and 5-FU induced immunogenic cell death in colon carcinoma 9 and gastric cancer 10 . Immunogenic cell death is a particular modality of cell death that can be triggered by selected anticancer chemotherapeutics 11 .
Tumour cells undergoing immunogenic cell death are characterised by the release or exposure of damage-associated molecular patterns (DAMPs) that stimulate the attraction, activation and maturation of dendritic cells and eventually the antigen-speci c priming of cytotoxic T lymphocytes 12 . This can induce an adaptive anticancer immune response that targets residual cancer cells with the same antigenic pro le 13 .
Aside from ICB-mediated inhibition of immune evasion, another possible explanation of the bene cial outcomes with ICB may be attributed to their effect on immune-independent hallmarks of cancer.
Abundant evidence in the literature has highlighted that use of ICB to inhibit PD-L1 14 , A2aR 15 or TIM-3 16 tumour cell-intrinsic signalling can suppress tumour cell invasion and migration and subsequent metastasis in a range of cancer models of gastric cancer, lung cancer and cervical cancer. Moreover, the PD-1 axis has been implicated in conferring chemo(radio)-resistance through promotion of stem-like characteristic in lung cancer 17 and enhancement of radiation-induced DNA repair in osteosarcoma 18 .
Additionally, PD-L1 signalling induced proliferation of tumour cells in a range of cancer types including hepatocellular carcinoma 19 , melanoma and ovarian cancer 20 .
However, the mechanistic role of PD-1, PD-L1 and A2aR tumour cell-intrinsic signalling in the context of OAC remains to be investigated. Therefore, this study explores how the rst-line FLOT chemotherapy regimen alters IC expression pro les of OAC cells longitudinally. The effect of single agent ICB on the viability of OAC cells is assessed including the ability of ICB to enhance FLOT chemotherapy cytotoxicity in OAC cells. Mechanistic insights are also provided surrounding the regulation of FLOT-induced upregulation of ICs on OAC cells and the effect of IC signalling on OAC cell proliferation, DNA repair and expression of cancer stem-like properties with and without FLOT treatment. Collectively, the ndings of this study will offer important insights for the design of synergistic ICB-chemotherapy combinations for OAC patients.

Methods
Culture of cell lines OE33 and SK-GT-4 oesophageal adenocarcinoma cells were purchased from European Collection of Cell Cultures. OE33 cells and SK-GT-4 cells were grown in RPMI 1640 medium with 2 mM L-glutamine (ThermoFisher Scienti c, Ireland) and supplemented with 1% (v/v) penicillin-streptomycin (50 U/ml penicillin 100µg/ml streptomycin (P/S)) and 10% (v/v) foetal bovine serum (FBS) (ThermoFisher Scienti c, Ireland) and detached using trypsin-EDTA solution (Gibco). All cell lines were maintained in a humidi ed chamber at 37°C 5% CO 2 and were tested regularly to ensure mycoplasma negativity. The OE33 cell line was established from a poorly differentiated stage IIA adenocarcinoma of the lower oesophagus (Barrett's metaplasia) of a 73-year old female patient. Whereas, the SK-GT-4 cell line originated from a well differentiated oesophageal adenocarcinoma arising in Barrett's epithelium from an 89-year-old Caucasian male. Detection of Ki67 staining by ow cytometry Cells were collected by trypsinization, xed in ice-cold 70% ethanol (Merck, Darmstadt, Germany) and incubated at room temperature for 30 minutes. The xative was decanted after centrifugation at 1,300 RPM for 3 mins and cells were resuspended in 100 µl of FACs buffer and stained with 1 µl of Ki67-AF647 (BioLegend). OAC cells were washed and resuspended in FACs buffer and acquired using BD FACs CANTO II (BD Biosciences) using Diva software and analysed using FlowJo v10 software (TreeStar Inc.).
Cell viability CCK-8 assay A CCK-8 assay (Sigma, USA) was used to assess the effect of ICB with and without FLOT chemotherapy regimen on the proliferation rate of OE33 cells and SK-GT-4 cells. 5 × 10 3 OAC cells were adhered in a 96 well plate at 37°C, 5% CO2 overnight. The media was removed, and cells were cultured for 48h in complete RPMI in the absence or presence of nivolumab (10 µg/ml), atezolizumab (10 µg/ml), A2aR antagonist (3 µM, Bio-techne, USA) with and without the FLOT chemotherapy regimen (IC 50 dose). 5 µl of CCK-8 solution was added to each well, followed by a 1.5h incubation in the dark at 37°C, 5% CO2. The optical density at 450 nm and 650 nm (reference wavelength) was measured using the Versa Max microplate reader (Molecular Devices, Sunnyvale, CA, USA) to determine a viable cell number. All of the data were analysed from three independent experiments.

Detection of γH2AX by ow cytometry
After treatment, cells were collected by trypsinization, xed in ice-cold 70% ethanol (Merck, Darmstadt, Germany), and incubated at room temperature for 30 minutes. The xative was decanted after centrifugation at 1,300 RPM for 3 mins with 2 ml of PBS containing 2% FBS. Cells were resuspended in 100 µl γH2AX staining (1:100 dilution of antibody) solution [Triton X100 (0.1%), FBS (2%)] for 2h at room temperature. OAC cells were resuspended in FACs buffer and acquired using BD FACs CANTO II (BD Biosciences) using Diva software and analysed using FlowJo v10 software (TreeStar Inc.).

RNA isolation and quanti cation
Cells were seeded at a density of 3×10 6 cells in a T75 ask in 11 ml of cRPMI and allowed to adhere overnight. Following drug treatment RNA was isolated from cell lines using the TRI Reagent® method. The RNA pellet was re-suspended in 30 µl RNAase free molecular grade H 2 0 and stored at −80°C. RNA quanti cation was determined spectrophotometrically, using a Nanodrop 1000 spectrophotometer (version 3.1.0, Nanodrop technologies, DE, USA).

cDNA synthesis
For cell line samples, total RNA (1 µg total RNA) was reverse transcribed to cDNA using the manufacturer's instructions for the cDNA Reverse Transcription Kit from eBioscienes (4368814). In brief, to anneal the primers to the RNA, a master mix containing RNaseOUT 25X dNTP Mix 10 mm, (10 mM, prepared as a 1:1:1:1 ratio of dATP, dGTP, dTTP and dCTP), 10x rt random primers, Bioscript reverse transcriptase (200units/µl) and 10X Bioscript Reaction Buffer in RNase-free water was added to each sample and the samples were incubated for 10 mins at 25°C, for 120 mins at 37°C then for 5 mins at 85°C and held at 4°C. The resulting cDNA was stored at -20°C.
Quantitative real time PCR and analysis qPCR was performed using TaqMan primer probes (MLH1, SMUG1, PARP1, MMS19 (Roche)) and a Quant Studio 5 real-time thermal cycler (Thermo Fisher Scienti c). 18S (eBioscienes) was used as an endogenous control for data normalization. Data analysis was performed using Thermo sher Scienti c Connect qPCR application software.

Aldehyde dehydrogenase (ALDH) assay
Aldehyde dehydrogenase (ALDH) enzyme activity was assessed using the Alde uor® assay (Stem Cell Technologies), according to the manufacturer's instructions. Brie y, cells were trypsinised and resuspended at a density of 1 x 10 6 cells/mL in Alde uor® assay buffer containing ALDH substrate (bodipy-aminoacetaldehyde) (5 µL/mL). Immediately following this, half of the resuspended cells were added to a tube containing the ALDH inhibitor diethylaminobenzaldehyde (DEAB) to provide a negative control. Cells were acquired using BD FACs CANTO II (BD Biosciences) using Diva software and analysed using FlowJo v10 software (TreeStar Inc.).

Statistical analysis
Data were analysed using GraphPad Prism (GraphPad Prism, San Diego, CA, USA) and was expressed as mean ± SEM. Statistical differences between two treatments in a particular cell line were analysed using a paired parametric Student's t-test. Statistical signi cance was determined as p≤0.05.

Chemotherapy-induced upregulation of ICs on the surface of OAC cells is maintained up to 3 weeks posttreatment
We have previously demonstrated that rst-line chemotherapy combinations FLOT and CROSS upregulated ICs on the surface of OAC cells following 48h treatment in vitro 21 . Therefore, we sought to investigate how long this chemotherapy-induced upregulation of ICs on the surface of OAC cells is maintained by longitudinally pro ling IC expression on the surface of OE33 cells following 48h treatment with vehicle control or FLOT chemotherapy regimen ( Figure 1A). 48h treatment with FLOT signi cantly increased PD-L1 expression on the surface of OE33 cells compared with the vehicle control at 48h (50.85 ± 1.2 vs. 2.65 ± 0.3%, p<0.0001, Figure 1B). Interestingly, FLOTinduced PD-L1 upregulation on the surface of OE33 cells remained upregulated at 4 days (45.35 ± 1.5 vs. 1.63 ± 0.5%, p<0.0001) and 21 days (8.88 ± 0.3 vs. 0.97 ± 0.1%, p<0.0001, Figure 1B) post-treatment compared with the vehicle control ( Figure 1B). Following subculture of recovered FLOT-treated OE33 cells PD-L1 was signi cantly upregulated compared with the vehicle control (3.15 ± 0.2 vs. 1.15 ± 0.1, p=0.007, Figure 1B).
Although 48h treatment with FLOT did not signi cantly alter CD160 expression on the surface of OE33 cells compared with the vehicle control at 48h, CD160 was signi cantly upregulated 4 days post-FLOT treatment compared with the vehicle control (73.53 ± 5.6 vs. 3.75 ± 0.7, p=0.0008, Figure 1B). However, CD160 expression had returned to baseline 21 days post-FLOT treatment and following subculture of recovered FLOT-treated OE33 cells with no signi cant difference when compared with the vehicle control ( Figure 1B).
Pro-survival MEK signalling upregulates ICs on the surface of OAC cells following chemotherapy treatment Chemotherapy-induced upregulation of ICs on the surface of OAC cells suggests these tumour cells may be employing ICs as an adaptive survival mechanism to overcome genotoxic stress. However, the signalling pathways mediating FLOT-induced IC upregulation remain unknown. Therefore, we sought to investigate if the pro-survival signalling pathway MEK may be regulating the chemotherapy-induced upregulation of ICs. Combining A2aR antagonist with FLOT signi cantly decreased the viability of OE33 cells compared with FLOT treatment alone (71.69 ± 3.1 vs. 72.60 ± 7.0%, p=0.03, Figure 3A). Combining nivolumab with FLOT signi cantly decreased the viability of SK-GT-4 cells compared with FLOT treatment alone (36.01 ± 4.4 vs. 38.72 ± 2.7%, p=0.01, Figure 3A).
Overall, single agent nivolumab and A2aR antagonism signi cantly decreased the viability of OAC cells alone. Interestingly, combining nivolumab or A2aR antagonist with the FLOT regimen signi cantly enhanced the reduction in viability of OAC cells compared with FLOT treatment alone. In addition, combining FLOT chemotherapy with single agent nivolumab, atezolizumab or A2aR antagonism signi cantly enhanced the reduction in viability of OAC cells compared with ICB treatment alone.
Overall, single agent nivolumab, atezolizumab and A2aR antagonism signi cantly decreased the proliferation of OAC cells alone. Interestingly, combining single agent nivolumab, atezolizumab and A2aR antagonism with the FLOT regimen signi cantly decreased the proliferation of OAC cells compared with FLOT treatment alone. Taken together these ndings suggest that inhibition of the PD-1 axis or A2aR axis decreases the survival of OAC cells and when combined with the FLOT regimen synergistically enhance the toxicity of FLOT against OAC cells in vitro.
Given these ndings we next aimed to investigate how blockade of PD-1, PD-L1 or A2aR signalling axes alone and in combination with FLOT might affect OAC cell apoptosis and cell death (Figure 4).  Figure 5A). Overall, similar trends were also observed at 48h and at 72h in which single agent nivolumab, atezolizumab and A2aR antagonism decreased γH2AX expression in OE33 and SK-GT-4 cells compared with the vehicle control ( Figure 5A). Similarly, at 48h and 72h timepoints, combining single agent nivolumab, atezolizumab and A2aR antagonist with the FLOT regimen signi cantly decreased the levels of γH2AX expression in OAC cells compared with FLOT treated cells ( Figure 5A). Furthermore, Tu et al., 23 demonstrated that intracellular PD-L1 acts as an RNA binding protein enhancing the mRNA stability of NBS1 and BRCA1, thus upregulating the expression of DNA repair proteins NBS1 and BRCA1. Therefore, we assessed if ICB might alter the expression of well described DNA repair genes PARP1, SMUG1, MLH1 and MMS19 alone and in combination with FLOT chemotherapy ( Figure 5). Single agent atezolizumab signi cantly reduced the mRNA expression levels of PARP1 and SMUG1 compared with the vehicle control (PARP1: 0.41 ± 0.1 vs. 0.78 ± 0.2%, p=0.005 and SMUG1: 0.45 ± 0.1 vs. 0.83 ± 0.1%, p=0.008) ( Figure 5B). Interestingly, combining A2aR antagonist with the FLOT regimen signi cantly increased the mRNA expression levels of MMS19 compared with FLOT treated cells (1.12 ± 0.1 vs. 0.13 ± 0.06%, p=0.03) ( Figure 5B).
Overall, single agent nivolumab, atezolizumab and A2aR antagonist decreased γH2AX expression in OAC cells and decreased the gene expression of DNA repair genes. Similarly, combining single agent nivolumab, atezolizumab with FLOT chemotherapy decreased γH2AX expression in OAC cells and expression of DNA repair genes. Interestingly, although single agent A2aR antagonist decreased γH2AX expression in OAC cells, an increase in expression of DNA repair genes was observed.

Blockade of PD-1 axis signalling in OAC cells decreases ALDH stem-like marker
Cancer stem-like cells exist as part of a subpopulation within tumours and are thought to be a major contributor to tumour recurrence. Moreover, our ndings demonstrate that ICB decreases OAC cell proliferation and viability and induced OAC cell apoptosis and cell death in a subpopulation of OAC cells.
Therefore, we aimed to investigate if ICB might be targeting the stem-like compartment within a population of OAC cells so we assessed the effect of ICB on the levels of ALDH stem-like marker in OAC cells alone and in combination with FLOT ( Figure 6).
Similarly, there were trends toward a signi cant decrease in ALDH activity following single agent atezolizumab treatment compared with the vehicle control in SK-GT-4 cells (3.39 ± 1.3 vs. 12.11 ± 4.8%, p=0.08) ( Figure 6). Overall, PD-L1 blockade decreased the expression of ALDH stem-like marker in OAC cells and PD-1 blockade attenuated the FLOT-induced increase in ALDH activity in OAC cells.

Discussion
We and others have previously shown that tumour cells express an array of both IC ligands and IC receptors 21 . Of particular clinical relevance we have previously shown that the FLOT chemotherapy regimen increases the expression of an array of ICs on the surface of OAC cells driving an immuneresistant phenotype within the tumour 21 . Furthermore, tumour cell-expressed ICs have also been implicated as a potential mechanism of chemo(radio)-resistance in a range of cancer types [24][25][26][27]  Furthermore, elements of the PD-1 signalling axis namely the PD-1 receptor and its cognate ligand PD-L1 have been identi ed on the surface of stem-like tumour cells in melanoma 39 and lung cancer 17 . We have previously shown that FLOT upregulates PD-L1 preferentially on the surface of stem-like OAC cells 21 . In addition, PD-1 and PD-L1 tumour cell intrinsic signalling has been shown to promote stem-like characteristics in both melanoma and lung cancer cells 17,39 . Similarly, the ndings of this study demonstrated that blockade of the PD-1 signalling axis decreased ALDH stem-like marker in OAC cells. Collectively, these ndings highlight an immune-independent role for the PD-1 axis signalling cascade in driving a treatment resistant phenotype, as cancer stem-like cells are thought to play a pivotal role in resistance to rst-line chemotherapy regimens and subsequent tumour recurrence 40,41 .
FLOT comprises of a unique cocktail of chemotherapies including 5-FU (an anti-metabolite), oxaliplatin (DNA intercalator) and docetaxel (taxane) with speci c mechanisms of action that target distinct phases of the cell cycle. 5-FU exerts its anticancer effects through inhibition of thymidylate synthase and incorporation of its metabolites into RNA and DNA resulting in damage, generating considerable amounts of cellular genotoxic stress and hindering the normal functioning and homeostasis of cellular processes that require these RNA and DNA 42 . Oxaliplatin, a platinum-based chemotherapy intercalates with cellular DNA forming platinum-DNA adducts, which induce DNA damage and block DNA replication 43 . Docetaxel inhibits microtubular depolymerization, and attenuation the effects of bcl-2 and bcl-xL gene expression which ultimately culminates in G2/M phase cell cycle arrest apoptotic cell death 44 . Although FLOT induced apoptosis within 48h and decreased viability we also observed that FLOT increased the proliferation of OAC cells, however, the observed FLOT-induced proliferation of OAC cells was diminished longitudinally. Although this initial increase in proliferation was surprising the FLOT regimen may collectively hinder DNA damage signalling or repair and subsequent cell cycle arrest initially or may prevent delays through the cell cycle initially. This may subsequently cause the cells to undergo cell cycling at a faster rate increasing their proliferation initially, which then deceases longitudinally as the FLOT-treated cells die off in culture.
There are a number of studies in the literature highlighting that PD-1, PD-L1 and A2aR intrinsic signalling in tumour cells also promote tumour cell proliferation in a range of cancer types including hepatocellular carcinoma 19 , lung, melanoma 20 , ovarian 20 , pancreatic 45 , gastric 15,16,46 and cervical cancer 16 . The ndings of this study further substantiate the immune-independent role of PD-1, PD-L1 and A2aR tumour cell-intrinsic signalling in promoting proliferation of tumour cells in the context of OAC.
In summary, the ndings from this study highlight the novel immune-independent functions of IC tumour cell-intrinsic signalling in OAC cells promoting a range of hallmarks of cancer including promoting tumour cell growth and proliferation, enhancement of a cancer stem-like phenotype and enhancement of DNA repair. Importantly, blockade of the PD-1 signalling axis suppressed tumour cell growth, decreased cancer stem-like marker ALDH and expression of DNA repair genes alone and in combination with the FLOT chemotherapy regimen. Combining PD-L1, PD-1 or A2aR ICB with the FLOT regimen synergistically enhanced chemotherapy cytotoxicity in OAC cells. Overall, this highlights a strong clinical rational for combining ICB with the rst-line chemotherapy regimens to not only reinvigorate anti-tumour immunity and prevent immune exhaustion but to directly enhance the cytotoxicity of FLOT via inhibition of immune-independent hallmarks of cancer mediated by IC-intrinsic signalling in OAC cells.

Con icts of Interest
The authors declare that there is no con ict of interest that could be perceived as prejudicing the impartiality of the research reported. Figure 1 FLOT upregulates ICs on the surface of OAC cells in vitro, an effect which is maintained 3 weeks posttreatment. OE33 cells were treated with vehicle control (veh) or FLOT for 48h (T1), washed twice and allowed to grow for an additional 48h (T2) after which the cells were sub-cultured in new asks and left to recover for 3 weeks (T3). Following complete recovery, the cells were sub-cultured 1 in 2 and screened for IC expression 3 days later (T4). IC ligand and receptor expression was pro led on the surface of OE33 FLOT chemotherapy regimen in the absence and presence of a MEK inhibitor (meki) for 48h and the expression of IC ligands (PD-L1, PD-L2 and CD160) and IC receptors (PD-1, TIGIT, TIM-3, LAG-3 and A2aR) on the surface of OAC cells was determined by ow cytometry. *p<0.05 and **p<0.01, experiments repeated n=3 times, paired parametric t-test.   Single agent nivolumab and atezolizumab decreased the levels of γH2AX and the levels of DNA repair genes in vitro. (A) OE33 cells and SK-GT-4 cells were treated with nivolumab (10 μg/ml), atezolizumab (10 μg/ml) or A2aR antagonist (3 μM) in the absence or presence of FLOT regimen for 24h, 48h and 72h. Expression of γH2aX was determined by intracellular ow cytometry. γH2ax expression is presented as mean uorescence intensity (MFI). Representative histograms showing the levels of γH2AX in OE33 cells for each treatment at 24h, 48h and 72h. (B) OE33 cells were treated with nivolumab (10 μg/ml), atezolizumab (10 μg/ml) or A2aR antagonist (3 μM) in the absence or presence of FLOT regimen for 48h. mRNA expression levels of PARP1, SMUG1, MMS19 and MLH1 were determined by qPCR (n=3 in triplicate). Expression presented as relative quantity of 18S housekeeping gene. *p<0.05 and **p<0.01, paired parametric t-test.

Figure 6
Nivolumab and atezolizumab treatment decrease the percentage of ALDH+ stem-like OAC cells in vitro. OE33 cells and SK-GT-4 cells were treated with nivolumab (10 μg/ml), atezolizumab (10 μg/ml) or A2aR antagonist (3 μM) in the absence or presence of the FLOT regimen for 48h. ALDH activity was determined