Anticancer effects of exercise: Insights from single-cell analysis

Highlights

• •Physical exercise can exert antitumorigenic effects; however, the molecular mechanisms are still poorly understood.
• •Single-cell analysis may help to characterize the molecular mechanisms underlying the effects of exercise on anticancer immune function as well as on the complex tumor microenvironment.
• •Recent research using single-cell analysis provides preliminary insights into the molecular mechanisms behind an improved antitumor immunity in response to exercise. Particularly, there is evidence for a “reprogramming” of several immune effectors towards a higher antitumoral toxicity.

A recent special topic in the Journal of Sport and Health Science reported the health benefits conferred by traditional and innovative m-health exercise and multimodal programs with respect to several types of cancer.¹ A possible mechanism behind this protective effect could be enhanced cancer immunosurveillance.^2,3 A single bout of exercise, such as 20–60 min of running or cycling, has been demonstrated to elicit a robust mobilization of immune effectors with a high cytotoxic capacity—notably, natural killer (NK) and CD8⁺ T cells—into the bloodstream, with the repetition over time of these bouts (“regular” exercise) eventually resulting in higher immune infiltrates into tumors and a delayed cancer growth.³ There is, however, controversy as to which immune effectors preferentially infiltrate tumors. “Non-natural” scenarios (e.g., the use of athymic mice lacking CD8⁺ T cells) are occasionally employed to artificially create isolated conditions and examine the impact of exercise exclusively on a particular cell subset (e.g., NK cells). However, it is crucial to acknowledge that in actual anticancer responses, all immune components collaborate synergistically, and thus “non-natural” scenarios may not fully reflect the intricate interplay necessary for an effective immune response against cancer.³

In recent decades, “omics” technologies have allowed researchers to study how exercise impacts the expression of thousands of genes in different immune cells.⁴ For instance, tailored regular exercise for several weeks can induce remarkable improvements in the specific function of NK cells (as assessed by their ability to kill cultured tumor cells) together with significant changes at the intracellular proteome level.⁵ Yet “traditional” gene sequencing approaches do not allow for the differentiation of the biological responses of each specific cell subset residing in the tumor microenvironment—that is, a heterogeneous and continuously evolving collection of infiltrating and resident host cells. To this effect, state-of-the-art single-cell RNA sequencing (scRNA-seq) analysis is currently employed for the examination of individual cells within biological samples, which provides researchers with the capability to investigate cellular heterogeneity and understand the distinctive characteristics of each cell and different subtype within a given cell family.⁶ Notably, although CD8⁺ T-cell infiltrates are generally crucial to establish an efficient antitumor immunity, not all CD8⁺ T-cell subpopulations living in the tumor microenvironment contribute equally to prognosis, disease progression, or response to immunotherapy. Seven different CD8⁺ T-cell subpopulations were recently identified in melanoma, with higher and lower infiltrates of “exhausted” and cytotoxic subsets, respectively, found in later-stage tumors.⁷ Additionally, 3 genes (PMEL, TYRP1, and EDNRB) overexpressed in the exhausted subpopulations were associated with poor prognosis.⁷ It is conceivable that the information provided by scRNA-seq could also help to characterize the molecular mechanisms underlying the effects of physical exercise on anticancer immune function as well as on the complex tumor microenvironment, as recently illustrated by 2 relevant studies^8,9 (Fig. 1).

Fig. 1.

Main findings from single-cell analysis. (A) Acute and regular exercise improve cancer immunosurveillance; (B) Findings from Savage et al.⁸; (C) Findings from Batatinha et al.⁹ This figure was created using  BioRender (https://biorender.com/). APC = antigen presenting cells; IFN = interferon; MHC = major histocompatibility complex; OXPHOS = oxidative phosphorylation; PBMCs = peripheral blood mononuclear cells; scRNA-seq = single-cell RNA sequencing; TAM = tumor-associated macrophages.

Savage et al.⁸ studied the impact of regular exercise on tumors in 2 distinct murine models (YUMMER and B16F10) of melanoma. Male mice (aged 8–12 weeks) performed forced exercise (treadmill running, 45 min/day) for 12–14 consecutive days. In YUMMER tumors, the exercise intervention suppressed tumor growth and hypoxia while increasing microvessel density (among other changes in the tumor vasculature) as well as the number of active CD8⁺ T cells (expressing CD69) and the expression of programmed cell death 1 (Pdcd1) in these cells, thereby suggesting an improved efficacy of immune checkpoint blockade. The scRNA-seq results showed shifts after the exercise program in several T-cell clusters. Thus, the proportions of cytotoxic (Gzmb, Klrd1, Nkg7, Pdcd1, Havcr2, Lag3, Tox, and Cxcr6) and cycling (similar gene expression to the cytotoxic cluster, but including proliferation genes such as Mki67 and Top2a) CD8⁺ T-cell clusters, as well as of IFNg⁺ cytotoxic (i.e., expressing the highest levels of cytotoxic genes, such as Gzmb, Prf1, Ifng, Ccl3 or Ccl4, and of the immune checkpoint genes Pdcd1, Havcr2, Lag3, and Tox) and central memory T-cell (memory (Tcf7, Sell, and Il7r) and recirculation markers (S1pr1 and Klf2)) clusters, were higher in the YUMMER tumors of exercised mice compared to controls. Additionally, exercise increased tumor-associated macrophage (TAM) clusters expressing high levels of major histocompatibility complex class II genes while decreasing the M2-like TAM cluster. Interestingly, exercise upregulated the expression of major histocompatibility complex class II genes (H2-Aa, H2-Eb1, and H2-Ab1), particularly in M2-like TAMs, while also expanding antigen-presenting cell populations. The M2-like TAMs—which, as opposed to the M1 phenotype, can be considered pro-tumorigenic¹⁰—showed the highest and lowest levels of glycolysis and oxidative phosphorylation-related gene expression, respectively, indicating a potentially greater dependency on glycolysis in this cell population. Additionally, regular exercise upregulated and downregulated genes in myeloid cell populations that are involved in oxidative phosphorylation (mt-Nd1, mt-Nd3, and mt-Nd4) and glycolysis (Gapdh, Pkm, Pgam1, Hk2, Pgk1, Eno1, Aldoa, and Ldha), respectively.

In a group of 16 healthy young adults (6 female, mean age =27 years), Batatinha et al.⁹ performed scRNA-seq in peripheral blood mononuclear cells at baseline and during the last 3–5 min of a submaximal exercise (cycling) session lasting 20–30 min. Flow cytometry analyses showed that the most responsive immune effectors to the exercise stimulus were NK and γδ T cells, as well as CD45RA⁺ effector memory CD4⁺ and CD8⁺ T cells, while CD4⁺FoxP3⁺ regulatory T cells remained unchanged. Within the NK cell population, the more cytotoxic subtype (CD56^dim) was preferentially mobilized over CD56^bright subsets. Additionally, exercise mobilized NK cells expressing the activating receptor NKG2D—which is associated with higher cytotoxic activity and NK cell expansion and, therefore, a stronger immune response—over NK cells expressing the inhibitory receptor NKG2A. On the other hand, scRNA-seq showed how the transcriptomes of several cell types (CD8⁺ T cells, effector memory CD8⁺ T cells, CD4⁺ T cells, central memory CD4⁺ T cells, NK cells, mucosal-associated invariant T cells, and monocytes) were altered by exercise. The greatest number of gene set enrichments was found in CD8⁺ T cells, specifically in genes related to cytotoxic activity (GZMB, GAMH, and PRF1), cellular adhesion and migration (ITGB1, ITGB2, and CCL5), IFNγ production (ANXA1), and antitumor activity (HLA-DPA1 and NKG7). These were mostly driven by the effector memory CD8⁺ T-cell population, with smaller contributions from central memory and naïve CD8⁺ T cells. Interestingly, at end-exercise there was a “high” variability in the leading genes enriched in several pathways related to antitumor immune activity across CD4⁺ and CD8⁺ T cells, NK cells, and monocytes. Otherwise, a higher number of the KIR/HLA family genes enriched in NK cell–mediated cytotoxicity pathway was identified at end-exercise in CD8⁺ T cells compared to NK cells. Also, there was a high gene overlap for those pathways with cytotoxicity and anti-tumor immune activity in CD4⁺ and CD8⁺ T cells, NK cells, and monocytes. On the other hand, not only did the exercise bout mobilize lymphocytes that showed an antitumor transcriptomic profile, when human peripheral blood mononuclear cells collected at end-exercise were administered to xenogeneic mice, they enhanced graft-versus-leukemia effects, thereby resulting in lower tumor burden and higher overall survival compared with the control mice receiving “baseline” peripheral blood mononuclear cells.

These studies have provided preliminary insights into the molecular mechanisms underlying improved antitumor immunity in response to acute/regular exercise with evidence of a reprogramming of different immune effectors towards higher antitumoral toxicity. Of note, however, are the current barriers to translating single-cell sequencing approaches into clinical practice: Not only the high cost, but also the intrinsic difficulties of performing bioinformatic analyses and generating reports that can be understood by non-specialists (for more details, see Ref. 11). We hope that single-cell technology will help to guide the design of future studies seeking to shed light on the anticancer effects of exercise and to identify its most effective prescription.

Peer review under responsibility of Shanghai University of Sport.