MC38 murine colon cancer cell line was obtained from Kerafast (USA) and was negative for mycoplasma assays. MC38 cells were cultured in DMEM medium complemented with 10% fetal bovine serum, 100 U/mL penicillin and streptomycin. Cells were incubated in a humidified incubator with 5% CO2 at 37°C.

All experiments and protocols were compliant with the ARRIVE guidelines and were approved by the Animal Ethics Committee of the University Claude Bernard of Lyon (CEEA: DR2021-44v2) and the Ethics Committee of Centre Léon Bérard Comprehensive Cancer Center (Lyon) (2021-SCAR-107). Six-week-old female C57BL/6 mice (n = 120; Janvier Labs, Le Genest-St-Isle, France) placed in a box equipped with 12:12 inverted light cycle at 22 ± 2°C, were distributed in enriched cages with nest and paper (5-6 mice/cage), with ad libitum access to food and water. After reception, mice followed an acclimatization period of one week in the animal house.

We adopted the methodology and treatment described in the study by Grasselly et al. (2018). MC38 cells were injected in 6-week-old C57BL/6 mice. Suspensions of exponentially growing cancer cells diluted in 0.2 mL of PBS were injected subcutaneously into the right flank of mice (5.10⁶ cells for MC38). When tumor volume reached 200 mm³ (around 5 days), mice were randomized in one of the four groups (1) Control (CRTL), (2) Treatment (TRT), (3) Exercise (EXE) and (4) Treatment plus Exercise (TRT/EXE) and the first treatment was administered (D0). Tumor growths were measured by manual measurement length x width (in mm³) using a calliper, every 2 days. To account for a possible loss of body mass, the mice were also weighed 3 times a week. The tumor volume was determined using the formula: V = 4/3 × π × r³. In the case of excessive tumor volume (>1600 mm³) or if weight loss was too great (15 to 20% in a few days compared with the weight at the start of treatment), the animals were excluded and sacrificed.

The combination regimen used for the MC38 colon adenocarcinoma mouse model was composed of capecitabine (Accord), administered per os 5 days/week at a dose of 250mg/kg. Oxaliplatin (Accord 5mg/mL) was injected i.p. once a week at a dose of 5 mg/kg. Anti-PD-1 (RMP1-14, BioXCell) was injected i.p. once a week at a dose of 12.5 mg/kg (20).

In the first set of experiments, mice were trained for 5 days and were sacrificed at 14 days after cancer cell injection. Tumors were dissected and analyzed by flow cytometry (n=20/groups).

In the second set of experimentation, non-treated EXE group (n=8) were trained for 1 week whereas TRT/EXE groups were trained for 16 days. CTRL group (n=8) and EXE group were sacrificed after 12 days of cancer cell injection. TRT (n=12) and TRT/EXE (n=12) groups were sacrificed at 25 days after cancer cell injection. Tumors were dissected and cut in two equal parts at the center; half were frozen in liquid nitrogen and the other half were embedded in OCT prior to being frozen. All samples were stored at −80 °C.

Flow cytometry was used to analyze the tumor immune microenvironment. All tumor tissue samples per group were collected. Cells from the tumors were counted and cell surface markers were stained with the following fluorescently conjugated antibodies: anti-CD45 (30F11, BD Biosciences), anti-CD4 (RM4-5, BD Biosciences), anti-CD8 (53-6.7, Miltenyi Biotec), anti-CD3 (17A2, BD Biosciences), antiCD11b (M1/70, Invitrogen), anti-PD-1 (10F.9G2, BioLegend), anti-CD25 (PC61, Biolegend), anti-CD49b (DX5, BD Biosciences), anti-Granzyme b/Cryofix (GB11, Invitrogen), Viability UV Zombie. Flow cytometry data were acquired on the BD LSRFortessa X20 cytometer and FlowJo software (Ashland, OR, USA) was used for analyses.

Tumor samples were lysed in buffer pH 7.4 (10 mM Tris Base, 0.5M sucrose, 50 mM NaCl, 5 mM EDTA, 30 mM Sodium pyrophosphate, 1% NP40, 0.25% sodium deoxycholate, 5µl/ml of inhibitors of proteases cocktail, 50mM NaF and 100µM sodium orthovanadate) and centrifuged at 12 000 g for 12 min (4°C). Protein concentration was determined using Lowry protein assay (Lowry et al., 1951). Proteins (50 μg) were separated by SDS-PolyAcrylamide Gel Electrophoresis (PAGE) and transferred onto nitrocellulose membranes (Bio-Rad, Hercules, CA, USA). The membranes were blocked with 5% BSA or non-fat dried milk in TBS-Tween (0.05%) and were incubated overnight at 4°C with the appropriate primary antibodies ( Table 1 ). Then, the membranes were washed three times with TBS-Tween (0.05%) and were incubated with secondary antibodies for 1 hour at room temperature. Immunoreactive bands were visualized with Odyssey Infrared Imaging System (LI-COR Biosciences, Lincoln, NE, USA) and protein loading was normalized to HSC70 expression.

Total RNA was extracted from frozen tumor tissue samples using TRIzol^® reagent according to the manufacturer’s protocol (Invitrogen, Carlsbad, CA, USA). RNA amounts and purity were determined on a microplate reader (Varioskan Lux, ThermoScientific, Waltham, MA, USA) with a µDrop^® plate (ThermoScientific, Waltham, MA, USA) and RNA integrity was controlled on 1.2% agarose gel using the FlashGel electrophoresis system (Lonza, Rockland, ME, USA). Reverse transcription reaction was carried out on 1 μg of total RNA using the iScript Supermix (Bio-Rad, Hercules, CA, USA). QPCR was performed on a CFX-96 Real Time System (Bio-Rad, Hercules, CA, USA) using Sybrgreen method (SsoAdvanced Universal Sybr^® Green Supermix, Bio-Rad, Hercules, CA, USA). The expression of target genes was normalized to reference genes (HPRT1, RPL4 and RPL19) and the relative expression was calculated using the ΔΔCt method ( Table 2 ).

All statistical analyses were performed in Prism9 (GraphPad software, San Diego, CA, USA). All variables were tested for normality using the Shapiro-Wilk test, which tests for normal distribution of the data. The tumor volume data were analyzed using a two-way (Exercise and Treatment) repeated measures ANOVA for experiment 1 and using repeated measures ANOVA (group effect: TRT vs. TRT/EXE) for experiment 2, followed by a Fisher’s Least Significant Difference (LSD) test. Flow Cytometry data (experiment 1) were analyzed using a two-way (Exercise and Treatment) ANOVA, followed by a Fisher’s LSD test. Western Blot and RT-qPCR (experiment 2), independent t-test or Mann-Whitney test was utilized to compare TRT vs. TRT/EXE because distribution was not normal. The comparisons were considered statistically significant for p < 0.05. Results were expressed as mean ± SEM (Standard Error of the Mean).

To investigate the impact of acute exercise prior to the administration of a combination of immune checkpoint inhibitors (anti-PD1) and platinum-based chemotherapy, we initially monitored tumor growth over a 7-day period ( Figure 1A ). At Day 0 (D0), the measured tumor volumes were similar among all groups ( Figure 1B ). By Day 7 (D7), exercise alone did not exhibit any effect on tumor growth when compared to the control group (EXE vs. CTRL). In contrast, both the control and exercise groups (CTRL and EXE) showed significantly larger tumors at D7 ( Figure 1B ). However, it is worth noting that a trend toward a more substantial reduction in tumor volume was observed in the EXE/TRT group compared to the TRT group. Specifically, on the final day of Experiment 1 (D7), the mean tumor volume in the TRT group measured 463.5 ± 134.5 mm3, while the EXE/TRT group had tumors measuring 185.5 ± 62.2 mm3 (p=0.0747, Figure 1B ). Intriguingly, between D4 and D7, tumor volumes remained stable in the TRT group but decreased in the EXE/TRT group, with a significant reduction in tumor volume observed between these two time points (p=0.0213, Figure 1B ).

We characterized the tumor immune infiltrate profiles by flow cytometry for each group at D7. This characterization encompassed total immune infiltrate CD45+ cells and the total count of CD3+ T cells, including effector cells like CD4+ T cells and CD8+ cytotoxic T cells. The following results are presented as the mean percentages of cells within each treatment group, normalized to the total live cells count.

Considering the trends observed in the reduction of tumor growth during Experiment 1, Experiment 2 extended the exercise period in the treated groups to evaluate the kinetics of tumor growth with and without pre-treatment exercise ( Figure 2A ). Similar to Experiment 1, tumor volume was regularly assessed throughout Experiment 2 to evaluate the tumor growth. Data from the CTRL and EXE groups compared to TRT and TRT/EXE confirm Experiment 1, with similar tumor growth between D0 and D7 and the same treatment efficacy ( Supplementary Data ). In the TRT/EXE group, tumor volume tended to increase less compared to the TRT group, with a significant difference observed on D2 (p=0.0245, Figure 2B ). Additionally, both groups showed a return to volumes close to those at day 0 by day 16, with this effect occurring by day 7 for the TRT group and by day 4 for the TRT/EXE group.

To evaluate the impact of the exercise pre-treatment administration on cell death, we measured apoptosis by immunoblotting of the cleaved caspase-3 (cCASP3). Apoptotic effector cCASP3 protein amount remained unchanged with TRT/EXE compared to TRT ( Figure 3A ). The activation of AKT through its phosphorylation on MC38 tumor cells was evaluated, to assess its promotion of proliferative and survival pathways. The pAKT/AKT ratio showed no change in the group of animals running prior to immune-chemotherapy administration ( Figure 3B ).

Angiopoietin-1 (ANGPT1) and Angiopoietin 2 (ANGPT2) involved in promoting vascular hyperpermeability and weakening of the vascular barrier were not different between TRT and TRT/EXE ( Figure 3C ). We observed that the ANGPT2:1 ratio tended to be lower in the TRT/EXE group than in the TRT group ( Figure 3D ). Il2rβ, a subunit of IL-2 receptor essential for regulation of immune response and expressed by NK cells, was not affected by the combination of exercise-treatment (vs treatment alone). Similar results were observed for Tumor Necrosis Factor alpha (TNF-α) and Interleukine-6 (IL-6) within tumor, 2 cytokine genes coding for pro-inflammatory. Genic expression of CHOP (C/EBP homologous protein-10), a key transcription factor for initiation of apoptotic program under extreme Endoplasmic Reticulum (ER) stress conditions was not different between TRT and TRT/EXE ( Figure 3C ).

The significantly slower kinetics of tumor growth in the early stage of MC38 cell development displayed in both experiments by the group receiving exercise in addition to immuno-chemotherapy compared to the group receiving immuno-chemotherapy only, suggests that exercise may enhance the effects of immuno-chemotherapy. However, our second experiment did not support the hypothesis of sustained exercise-induced effects on tumor growth in addition to immunotherapy beyond D7, as the tumor volumes in both groups returned to their initial sizes (D0) on D16. This observation could be the result of the particularly strong effect of this specific immune-chemotherapy combination on MC38 tumor growth, as previously demonstrated, that may overshadow the exercise effect (20). MC38 is a syngeneic colorectal cancer model which has been extensively used in the context of immune checkpoint inhibitor therapy as it has shown robust responses to anti PD1 and anti PDL1 antibodies. Although the subcutaneous (SC) and orthotopic localizations have been shown to differ in terms of tumor immune microenvironment, the SC tumors remain largely used, in particular for the description of resistance mechanisms to immune checkpoint inhibitors (21, 22).

Furthermore, only a few rodent studies combined exercise with immuno-chemotherapy in various clinical cancer settings. Several murine studies have investigated the combination of immunotherapy with exercise in different types of cancer including breast cancer, melanoma liver and lung carcinoma with divergent results (23, 24). Overall, the tumor models (i.e. syngeneic transplanted mouse models, drug-induced or transgenic models) and exercise modalities (i.e. pre-injection vs post-injection exercise, programmed vs voluntary exercise, intensity) were heterogeneous. In their Patient-Derived Xenograft (PDX) model of non-small cell lung cancer, Martin Ruiz et al. found that combining immunotherapy (specifically nivolumab, a PD-1 inhibitor) with programmed exercise post-injection (aerobic and resistance training, 8 weeks) did not exhibit any additional effect on the inhibition of tumor growth compared to immunotherapy alone (25). In addition, the final tumor volume resulting from the nivolumab + exercise intervention tended to be larger than that observed in the other study groups. The authors attribute this outcome to a ‘paradoxical’ or ‘unconventional’ response associated with the immune cell recruitment and the intratumoral inflammatory environment triggered by these cells. Bay et al. showed an additive effect of a combination of 2 weeks of voluntary wheel running with anti-PD-1 treatment on tumor growth in a B16 melanoma model (26). Gomes-Santos et al. investigated how physical activity, either in conjunction with anti-PD-1 treatment alone or when combined with anti-CTLA-4 treatment, influenced tumor growth in three murine breast cancer models (14). The authors observed that ICI treatment alone had minimal impact on orthotopic breast cancers, whereas combining ICI with programmed treadmill running led to a delay in tumor growth and a reduction in the final tumor volume. Wennerberg et al., found in mice with established 4T1 triple negative mammary carcinoma that combining anti-PD1 and radiotherapy with programmed treadmill exercise significantly slowed tumor growth compared to anti-PD1 and radiotherapy alone (27).

In the present study, the groups without immuno-chemotherapy treatment (CTRL and EXE) exhibited larger tumor size than the treated groups from D4 up to D16, without difference between the two groups. These results confirmed the successful tumor cell implementation and the efficiency of the immuno-chemotherapy treatment previously reported (20). Moreover, the lack of a difference in tumor volume and intratumoral T-cell infiltrate between the control and the exercise only groups suggests that the observed beneficial effect of combined exercise-immuno-chemotherapy might be explained by a potentiation effect of exercise rather than a proper effect of exercise itself.

Similar to Grasselly et al. (2020), we reported here that tumor volume reached its peak volume at D7 after treatment initiation and decreased afterwards, in both the TRT and TRT/EXE groups. At D7, we observed a significant increase in total CD45+ leukocyte, CD3+ T lymphocyte and CD8+ T cytotoxic cells infiltration in the combined exercise-immuno-chemotherapy group tumors compared to immune-chemotherapy alone, suggesting that that one week of exercise before immuno-chemotherapy administration is sufficient to increase total leukocyte and T lymphocyte tumor infiltration included the CD8+ cytotoxic T cells. As previously mentioned, exercise is known to have a profound influence on the human immune system (28). Currently, it remains unclear whether the anticancer effects of physical activity are mediated by the acute mobilization and redistribution of cytotoxic effector cells in response to acute exercise sessions or through an adaptive training effect that enhances tissue-associated T cells. In our study, we demonstrated that acute exercise leads to an increase in intratumoral immune cell infiltration, which may explain our observations of reduced tumor growth after 7 days of acute exercise prior to immuno-chemotherapy administration.

To date, several studies in different murine models have observed that exercise without anti-tumor treatment increases intratumoral immune infiltrates after several weeks of treadmill or voluntary wheel running training (12, 27, 29, 30). The results by Wang et al. suggested that a minimum of exercise intensity, comparable to the present protocol, is needed to observe beneficial effects on intratumoral immune cell infiltration (31). Of note, we did not observe an effect of pre-treatment exercise on the infiltration of CD4+ and CD8+ lymphocyte subpopulations. A study showed that training an ApcMin/+ murine model of intestinal cancer on a treadmill for 12 weeks reduced tumor burden and increased CD8+ expression in the tumor (29). Martín-Ruiz et al. (2020) observed that neutrophil tumor infiltration trended to be higher in the exercise group in combination with nivolumab compared to the nivolumab alone group (25).

We hypothesized that exercise combined with anti-PD1 could decrease checkpoint molecule PD-1 expression by CD8+ cells. Our results showed that exercise alone decreases CD8+ cells expressing PD-1+ compared to the control group. However, we found no additive effect on the decrease in CD8+ PD-1+ cell expression when exercise was combined with anti-PD1. Some of the previously mentioned preclinical studies have explored the effects of exercise training on levels of the checkpoint molecule PD-1 expression linked to T cell exhaustion, however results are heterogenous. Wennerberg et al., showed that exercise training, when combined with radiotherapy and anti-PD1 treatment following the injection of 4T1 triple negative breast tumors, reduced the proportion of splenic PD1+CD8+ T cells, which was not achieved by anti-cancer therapy alone, and resulted in reduced tumor growth (27). In their study, Bay et al. showed that voluntary wheel running led to enhanced expression levels of the checkpoint molecule PD-1 and its two ligands PD-L1 and PD-L2 in their B16 melanoma mouse model, without effect on tumor growth with the combination of exercise and anti-PD-1 therapy (26).

Overall, these results suggest that the effects of physical exercise on PD1+ T cells and the efficacy of anti-PD1 immunotherapy are inconclusive. Further research is needed to determine whether the anticancer effects of physical activity are mediated by PD1-dependent or independent pathways.

Other mechanisms might explain the observed effects of exercise on reducing tumor growth, such as the improvement of tumor vascularization and perfusion, which can lead to a more favorable metabolic profile within the tumor, including the mitigation of hypoxia (8–10, 32–34). Angiopoietin-1 (Angpt1) is an activator of tyrosine kinase receptor TEK expressed mainly on endothelial cells. TEK activation and phosphorylation promote vascular maturity and endothelial cell survival. While Angpt1 counteracts hyperpermeability, Angiopoietin 2 (Angpt2) is upregulated in human cancer and its activation weakens the vascular barrier. Increased ANGPT2/ANGPT1 ratio at the mRNA level has been reported to correlate with neo‐angiogenesis and poor prognosis in many cancer types (35). Unfortunately, we did not find any effect of exercise on the expression of these genes alone, but we have observed a trend towards an increase in the ANGPT2/ANGPT1 ratio. A recent review attributed the difficulty to come to a conclusion on the effect of exercise on vascular remodeling is because of the methodological heterogeneity among preclinical studies (36).

Moreover, McCullough et al. suggested that exercise induces a transient increase in tumor perfusion, potentially leading to enhanced immune infiltration and chemotherapy infiltration within the tumor (10). In the present study, we were not able to explore this hypothesis. Future studies exploring the effect of acute exercise just before or during anti-cancer treatment should plan to measure transient tumor perfusion during exercise.

Furthermore, in the present study, the combination of exercise with immuno-chemotherapy had no effect on apoptosis at D16, whereas we expected an increase in cell death markers such as Caspase 3. One study found a potentiation of apoptosis when exercise was combined with radiotherapy compared with radiotherapy alone in a prostate cancer model (34, 37). In a patient-derived xenograft (PDX) non-small-cell lung cancer model, Martín-Ruiz et al. also found that exercise in combination with PD-1 blockade reduced tumor growth, exhibiting diminished tumor cell proliferation and increase of tumor necrosis. The lack of exercise effect (TRT vs TRT/EXE) on tumor markers of apoptosis, vascular integrity and inflammation is consistent with the absence of a difference in tumor volume at D16.

We did not find any difference in the pAKT/AKT ratio in the tumor, a tumor survival pathway, when exercise was combined with immuno-chemotherapy compared to treatment alone at D16. These results of intratumoral molecular analysis at D16 are consistent with the lack of difference in tumoral volumes observed at that time between the TRT and TRT/EXE groups.

The present study has several limitations. Firstly, the MC38 subcutaneous murine model of colorectal cancer that was used in this study exhibited aggressive tumor growth (22). Most untreated animals reached ethical endpoint criteria, namely a tumor size > 1600 mm3 and tumor necrosis, within 14 days of subcutaneous tumor implantation. Additionally, tumor volumes in the treated groups reached their maximum after 7 days of treatment and exercise. Moreover, it is possible that the impact of the capecitabine+oxaliplatin+anti-PD1 combination was already substantial after a few days, making it difficult to observe additional benefits of exercise when administered alone. For future studies, it would be relevant to investigate the combination of ICI with exercise in less aggressive models of colorectal cancer characterized by slower tumor growth and lower treatment doses. Such models would better mimic the clinical conditions observed in humans.