Re: 운동은 NK cell의 조절을 통해 항암효과

[문형철]

beyond reason

Natural killer (NK) cells are the most responsive immune cells to exercise, displaying an acute mobilization to the circulation during physical exertion. Recently, exercise-dependent mobilization of NK cells was found to play a central role in exercise-mediated protection against cancer. 

Here, we review the link between exercise and NK cell function, focusing on circulating exercise factors and additional effects, including vascularization, hypoxia, and body temperature in mediating the effects on NK cell functionality. Exercise-dependent mobilization and activation of NK cells provides a mechanistic explanation for the protective effect of exercise on cancer, and we propose that exercise represents a potential strategy as adjuvant therapy in cancer, by improving NK cell recruitment and infiltration in solid tumors. 

Exercise is Beneficial for Cancer Patients–A Role for NK Cells
Circulating immune cells were first shown to be regulated by exercise in 1893 by Schulte, who described how lymphocytes were recruited to the blood stream during physical exertion [1]. Through such immune cell recruitment, exercise may directly stimulate the immune system to combat disease. Within the past 30 years, focus on getting patients across a number of diseases out of bed and into active rehabilitation early in the disease trajectory has emerged [2]. In cancer, the benefits of engaging patients in an active lifestyle despite their disease are becoming increasingly evident. At present, more than 100 exercise intervention studies in cancer patients have reported favorable effects on both patient-reported outcomes (Box 1) and physical functioning (see Glossary) when exercise is performed during or after antineoplastic therapy [3]. In addition, accumulating evidence suggests that exercise has a direct effect on tumor growth. Early evidence from observational studies has shown that physical activity reduces the risk of disease recurrence in colorectal, prostate, and breast cancer patients [4–6] (Box 2). Moreover, a 2-year training intervention study has recently shown that high intensity endurance training can increase PSA doubling time, a surrogate marker for delayed tumor progression in prostate cancer patients [7].

Accordingly, several preclinical studies have shown that voluntary wheel running can inhibit tumor growth across a range of experimental animal tumor models, including both genetic and trans- plantable tumor models [8]. In line with this, our laboratory has recently shown that voluntary wheel running in mice could induce an exercise-dependent increase in intratumoral immune cell infiltration in various genetic and transplantable murine tumor models [9]. Intratumoral infiltrates included natural killer (NK) cells, and exercise-mediated induction of intratumoral NK cells contributed to a 50–60% reduction in tumor growth. Further mechanistic analyses demonstrated that tumor control could be achieved through an epinephrine-dependent mobilization of NK cells, together with subsequent IL-6-induced redistribution and activation of NK cells. These recent findings have linked exercise, epinephrine (‘fight or flight’ response), and IL-6 to NK cell mobilization and redistribution, and ultimately, to control of tumor growth in mice [9]. In light of this, we review the current knowledge on exercise-mediated regulation of NK cells, seeking to improve our under- standing of the role of exercise and exercise factors in the mobilization and activation of NK cells in tumor control. The review is mainly focused on human findings, yet, to provide additional insight into relevant mechanisms, animal studies are also discussed. The ability to further our insight into the mechanistic pathways regulating NK cell antitumoral activity and cytotoxicity are timely, as they may facilitate additional approaches–including physiology-based ones–to improving potential cancer treatments.

Box 1. Exercise and Depression in Cancer Patients

Many patients diagnosed with cancer suffer from depression, yet exercise may comprise an effective intervention. Several Cochrane systematic reviews have documented beneficial effects of exercise on patient-related outcomes, including depression, in patients with cancer both during and after completed anticancer therapy [96,97]. While depression may arise from multifactorial causes, kynurenine-mediated tryptophan degradation has been proposed as a mechanism, whereby stress induces depression. Ruas and colleagues recently showed that exercise training could decrease kynurenine levels through increased kynurenine metabolism in the muscles, proposing a mechanistic link between exercise training and reduced depression [98].

Box 2. Exercise and Cancer Prevention

Over the past 10 years, epidemiological studies have demonstrated that exercise and physical activity are associated with markedly reduced cancer-related, and overall mortality, across the ‘big’ cancer diagnoses, that is, breast, colon, and prostate cancers [5,99,100]. The Centers for Disease Control and Prevention (CDC) has summarized the scientific evidence stating that: ‘physical active people have a lower risk of developing colon or breast cancer than people who are not active’; and ‘physical active people may have a lower risk of developing endometrial and lung cancer, although the scientific evidence is not final yet’ (http://www.cdc.gov/physicalactivity/basics/pa-health/index.htm#ReduceCancer). 

Basic NK Cell Biology

NK cells were originally described as a subset of lymphocytes with a ‘natural killing’ ability towards cancer cells without previous priming [10], classifying them within the innate immune system of fast responders. Recent classification groups NK cells within the emerging population of innate lymphoid cells (ILCs) [11]. NK cells develop in the bone marrow from CD34+ hemato- poietic precursor cells and are subsequently distributed widely throughout the body including the bone marrow (BM), lymph nodes (LNs), spleen, peripheral blood, lung, and liver (Figure 1) [12]. 

In healthy adult individuals, NK cells constitute 5–15% of all circulating lymphocytes [13]. Upon activation, the main function of NK cells is to kill infected (e.g., virus) or transformed (malignant) cells, and, to trigger the adaptive immune response through cytokine release. For further characterization of NK cell phenotype and function, see Box 3.

Box 3. NK Cell Phenotype and Function

NK cells are characterized by a combination of surface molecules, neither being independently specific for the NK lineage of cells. Mature NK cells are often identified as CD3neg lymphocytes with a differential expression‎ of CD16 and CD56. The expression‎ level of the latter two has been ascribed to different NK cell maturation stages and functions. CD56bright NK cells, which are primarily found in secondary lymphoid compartments, that is, LNs and tonsils, are considered to be the least mature. These cells have low-to-none CD16 expression‎, and are potent cytokine producers with poor lytic function. At the other end of the maturation spectrum, CD56dim CD16+ NK cells comprise approximately 90% of NK cells in the blood and spleen. CD56dim CD16+ NK cells are highly cytotoxic and produce only few, if any, cytokines [101]. In addition to this classical understanding, accumulating evidence indicates a more complex range of phenotypic and functional subtypes of mature NK cells, which are formed as a result of epigenetic modifications and microenvironmental stimuli [102].

The function of mature NK cells is to kill infected or transformed cells. There are several ways in which NK cells can achieve this: (i) by releasing cytoplasmic granules containing perforin and granzymes that leads to target cell lysis (cytotoxicity); (ii) via death receptor-mediated apoptosis (through Fas–FasL, TNF–TNFR receptor–ligand interactions); (iii) by secreting various effector molecules such as IFN-g, stimulating adaptive immunity; and (iv) through antibody- dependent cell-mediated cytotoxicity (ADCC) via FcRgIII(CD16)-mediated binding to antibody coated target cells (including tumor cells) [103]. 

Exercise and the Control of NK Cell Mobilization

Characterization of exercise-mediated changes in circulating immune cells began in the 1980s, with technological advancements in flow cytometry. Now, a general model of exercise-driven modulation of immune cell distribution in tissues has been proposed, describing how NK cells, T cells, and to a lesser extent, B cells, are mobilized to the circulation during exercise (Figure 2) [14]. This mobilization appears to represent a recruitment of stored immune cells, rather than a generation of new cells [15].

Kinetics and Intensity of NK Cell Mobilization

Exercise-mediated mobilization of NK cells is a very rapid phenomenon. As little as 70 s of stair climbing has been shown to increase the frequency of NK cells in the blood by 6-fold [16]. Subsequently, several studies have shown mobilization of NK cells within minutes with exercise, when performed with an intensity associated with breathlessness, increased heart rate, and elevated plasma epinephrine levels [17–19]. In general, maximal mobilization of NK cells is achieved within 30 min of endurance training, after which continued exercise does not lead to further increases in NK cell numbers [20]. After the initial mobilization, elevated NK cell levels are maintained throughout endurance training for up to 3 h, whereas prolonged exercise exceeding 3 h leads to a decrease in circulating NK cell numbers [20]. In fact, after cessation of exercise at any length, a decline in the number of circulating NK cells is observed, which is believed to reflect a redistribution of the mobilized NK cells to peripheral tissues. The very rapid and uniform response suggests that there is a minimal threshold needed to engage NK cells, and once this stimulus is obtained, the response is all or nothing. In line with this, a study reported that when exercise training intensity was increased in individuals, the subsequent progression in training intensity did not boost the frequency of circulating NK cells further [21].

Both epinephrine and norepinephrine have consistently been shown to drive NK cell mobilization in humans, linking the rapid and well-described mobilization of NK cells to exercise intensity dependent responses in catecholamine concentrations. For instance, early studies have reported that systemic administration of epinephrine can mimic exercise-induced increases in the frequency of circulating CD16+ cells, as well as overall NK cell cytotoxic activities immediately after exercise. Moreover, these studies also observed a reduction in both human NK cell numbers and activity hours after intravenous epinephrine administration [22,23]. The same authors showed that norepinephrine induced a similar mobilization of NK cells, as eval‎uated by an increase in peripheral blood mononuclear cell (PBMC) cytotoxicity after intravenous infusion of norepinephrine in healthy young men [24]. NK cells have the greatest density of b-adrenergic receptors of all lymphocytes [25], and studies using nxxonselective and selective b1- and b2-blockers have indicated that the epinephrine-dependent mobilization of NK cells seen in humans during exercise is dependent on b2-adrenergic signaling [26,27]. 

Age, Gender, and Disease Considerations

Exercise-dependent NK cell mobilization is a general phenomenon, observed across all ages and genders [28,29]. Moreover, NK cell mobilization occurs in lean and obese, as well as trained and inactive individuals [30,31]. While immune function may be somewhat compromised in elderly and inactive people [32], the variation in NK cell mobilization is more likely to reflect a barrier in exercise performance than a functional deficit in NK cell mobilization. To date, the only condition where NK cell mobilization has been reported to be attenuated has been in patients with heart failure on b-blocker medication [33,34]. Yet, heart failure patients not treated with b-blockers respond with a robust NK cell mobilization during exercise, just as healthy individuals do [33,35], highlighting an association between adrenergic signaling and NK cell mobilization.

In cancer patients, only one study has addressed the occurrence of acute mobilization of NK cells. In breast cancer survivors, the induction of NK cells is equivalent to that observed in age- matched control subjects, although the resting levels of NK cells has been shown to be slightly reduced, due to preceding chemotherapeutic therapy [36]. Additionally, a few studies have investigated the effect of exercise training on NK cell cytotoxicity in cancer patients. In breast cancer patients, exercise training was found to increase NK cell cytotoxicity and proliferation through H3 incorporation when tested ex vivo [37,38]. These effects were also seen in stomach cancer patients receiving early rehabilitation while still bedridden from surgery [39]. However, these studies included no eval‎uation on the effect of increased NK cell cytotoxicity on clinical outcomes.

Selective Mobilization of Mature Cytotoxic NK Cells

Studies in healthy individuals suggest that the subtype of NK cells, primarily deployed by exercise, consists of mature cytotoxic NK cells, when compared with less ‘mature’ cytokine- producing NK cells. For example, this has been evident when stratifying NK cells according to CD56 surface expression‎, with CD56dim/CD16+ NK cells displaying greater mobilization than CD56 bright/CD16–cells [20,21], or, when stratifying NK cells based on the combined expression‎ of activating and inhibitory receptors, that is, NKG2A, KIR, and CD57 [40]. Here, a study indicated that the most differentiated (NKG2A–/KIR+) NK cells displayed a greater mobilization than the less mature (NKG2A+/NKG2Chigh) NK cells [41]. One should keep in mind that CD56bright and CD56dim NK cells reside in different compartments, so their differential responses in mobilization frequency by exercise might potentially reflect a selective engagement of different storage beds. CD56dim NK cells are primarily stored within the spleen and vascular bed, both organs being engaged during exercise [18,42]. In addition, selective NK cell mobilization with exercise might be explained, in part, by higher expression‎ of b1 and b2-adrenergic receptors on CD56dim NK cells, suggesting that these cells might be more sensitive to epinephrine increases than CD56bright NK cells during exercise. Finally, one cannot discard the possibility that exercise might induce a maturation of CD56bright into CD56dim cells, as studies eval‎uating NK cell mobilization by exercise have based their analyses on the number of peripheral blood CD56dim cells [34]. Thus, further studies are warranted to fully elucidate the mechanism(s) behind the selective mobilization of cytotoxic CD56dim NK cells with exercise.

Based on various studies in humans, a consistent picture of a rapid and general mobilization of NK cells during an acute bout of exercise is evident (Figure 2). As such, considering the reliance on b-adrenergic signaling for exercise-dependent NK cell mobilization, the performed exercise must be of an intensity that is associated with increased catecholamine levels to elicit a mobilization response.

Exercise Factors Involved in NK Cell Activation and Function

During an acute bout of exercise, additional exercise factors beyond catecholamines are induced. These include muscle-derived myokines, which are peptides derived from muscle fibers and released into the circulation during exercise [43]. A large number of these myokines are cytokines, and their release during exercise is believed to direct energy substrate fluxes to working muscles, coordinating the adaptive metabolic response in muscle following exercise cessation [44]. Yet, several of the proposed exercise-induced myokines, are primarily recog- nized for their role in controlling immune cells, including IL-15, IL-7, and IL-6 [45]. These immune cell-modulating cytokines provide a conceptual basis for understanding how contracting muscles communicate with circulating immune cells during exercise, establishing a muscle-to- immune cell crosstalk axis that links physical activity to immune cell regulation.

IL-15

IL-15 is highly expressed in muscle tissue, and IL-15 protein expression‎ has been shown to increase in human muscles following 12 weeks of endurance training [46]. IL-15 is also acutely regulated by exercise, as muscular IL15 mRNA expression‎ increases within 6 h of training [47]. It is currently unclear whether IL-15 is actually released from contracting muscles during exercise or whether it accumulates within muscles. Studies using primary murine myotubes or C2C12 muscle cells in vitro suggest that IL-15 is secreted in complex with the IL-15 receptor / (IL- 15R/) [48]. Within muscles, IL-15 functions both as an anabolic factor, leading to the accumu- lation of myosin heavy chain proteins, and as a metabolic regulator, defining a myogenic phenotype characteristic of fast-switch muscles [49].

IL-15 activates cytotoxic immune cells, such as NK and T cells through trans-presentation of the IL-15/IL-15R/ complex on dendritic cells [50]. Muscle cells also express the IL-15/IL-15R/ complex, and in primary murine myotube cell culture systems, muscle-derived IL-15/IL-15R/ can activate murine cytotoxic T cells [48], indicating that trans-presentation can also occur from muscles. IL-15 is recognized as the main regulator of NK cells, based on mutational loss-of- function studies in both humans and mice [51,52]. Yet, the response to IL-15 may be concen- tration-dependent, with high IL-15 concentrations inducing NK cells to proliferate, while low IL- 15 concentrations, leads to NK cell maturation [53]. Recent studies have demonstrated that NK cells and other cytotoxic immune cells increase their cellular metabolism upon activation; this includes increased rates of glycolysis and oxidative phosphorylation [53,54]. Upon NK cell activation, upregulation of nutrient transporters, augmented substrate uptake, and increased cellular size ensue, among other things [55]. Mechanistic studies in mice have shown that IL-15 may drive immune cell activation and metabolism through mTORC1-dependent signaling, leading to a cytotoxic profile characterized by interferon (IFN)-g production and granzyme B expression‎ [53,54]. By contrast, eval‎uation of human peripheral NK cells have shown that CD56bright NK cells are more metabolically active than CD56dim NK cells, and while the combination of IL-12 and IL-15 might increase the metabolic activity of these human NK cells, the dependence on mTORC1 signaling in induction of the metabolic response has not been verified [56].

The great dependency on IL-15 for NK cell proliferation and maturation has been pursued within NK cell therapy. Recently, a first-in-human study with administration of recombinant IL-15 in melanoma patients showed that IL-15 administration markedly increased the number of circu- lating NK and gd T cells [57]. Subsequently, IL-15 administration resulted in clinical benefits with clearance of lung metastases in two patients [57]. This study highlights the clinical potential of IL- 15 in NK cell therapy, and several clinical trials are underway pursuing the therapeutic effect of IL-15 or the IL-15/IL-15R/ complex in cancer patients [58].

IL-7

IL-7 has also been identified as a myokine. Primary human myotubes have been shown to release IL-7 to the supernatant in cell culture systems, and men undergoing an 11 week strength training program showed increased IL-7 expression‎ in major muscle groups [59,60]. In primary human myocytes, recombinant IL-7 has been shown to stimulate satellite cell proliferation and reduce myotube differentiation by moderating the expression‎ of terminal myogenic markers such as myosin heavy chain 2 and myogenin [60].

IL-7 is a member of the gC family of cytokines, which is generally recognized for their role in T and B cell development. Premature and immature NK cells express high levels of the IL-7 receptor / [61], yet the number or functions of NK cells in IL-7 knockout mice are unaffected [62,63]. Even though IL-7 may have little overall importance for NK cell physiology, there is a small population of NK cells residing in the thymus that have been shown to depend on IL-7, as IL-7R knockout mice lack this NK cell population [63]. These NK cells are characterized as being cytokine- producing rather than cytotoxic [62,64,65].

IL-6

IL-6 is the best described myokine with multiple functions in muscle-to-organ crosstalk and metabolic regulation. Plasma IL-6 increases exponentially during exercise proportionally with the length and intensity of the performed exercise and the amount of muscle tissue engaged [66]. Muscle fibers have been firmly established as the source of exercise-induced IL-6 [44]. Integra- tive physiological investigations in humans have shown that exercise-induced IL-6 leads to increased intramuscular glucose uptake and translocation of the glucose transporter (Glut4) to the sarcolemma; activation of AMPK (50 AMP-activated protein kinase) signaling and downstream alterations in cellular metabolism; increased hepatic glucose production and output; induction of lipolysis and fatty acid oxidation in adipose tissue; and constitution of an anti-inflammatory response with inhibition of lipopolysaccharide (LPS)-induced tumor necro- sis factor / (TNF-/) expression‎ and production of anti-inflammatory cytokines, that is, IL-1ra and IL-10 (as reviewed in Pedersen and Febbraio [44]).

Although NK cells express the IL-6 receptor complex, comprising the IL-6 receptor / and the ubiquitously expressed gp130 subunit [67], IL-6 has not traditionally been linked to NK cell development or activation. Studies in the early 1990s suggested that IL-6 might alter the expression‎ of cell adhesion molecules on human NK cells and increase integrin-mediated adhesion to fibronectin- or laminin-coated plates [68]. Later, IL-6 was implicated in endothelial transmigration of both monocytes and cytotoxic T cells. In these studies, IL-6 was shown to induce changes in cell adhesion molecules and facilitate chemotaxis as eval‎uated by immuno- fluorescent staining, transwell migration (Boyden chamber), and intravital imaging [69,70]. In contrast to a role in promoting immune cell infiltration in tumors, extensive evidence has revealed that elevated IL-6 levels can also be associated with protumorigenic responses stimulating survival, proliferation, and angiogenesis, depending on the context [71]. As exercise leads to a transient increase in IL-6, one might speculate that this increase differs from the chronic elevated IL-6 levels that have been observed in patients with low-grade inflammation and that have been associated with tumor-promoting properties. Indeed, our group has recently shown that blockade of IL-6 signaling during exercise abolishes the exercise-dependent reduction in melanoma tumor growth in mice and eliminated exercise-induced intratumoral NK cell infiltration [9]. By contrast, daily injection of physiological doses of IL-6 could not mimic the effect of exercise on tumor reduction or NK cell infiltration alone [9].

Taken together, muscle-derived myokines may play a role in an exercise-dependent regulation and activation of NK cells, eliciting a muscle-to-immune cell crosstalk axis during exercise (Figure 3, Key Figure). Yet, further studies are warranted to fully elucidate the mechanistic link between exercise-induced myokines and NK cell regulation.

Figure 3. During exercise, a rapid and general mobilization of NK cells to the circulation is observed, and this exercise- dependent mobilization is induced by b-adrenergic signaling and catecholamines. Mobilized NK cells are affected by muscle-derived myokines, exercise-dependent hyperthermia, as well as intratumoral vascularization and perfusion, sub- sequently inducing the regulation, redistribution, and activation of mobilized NK cells.

Additional Exercise Effects

In addition to the release of systemic exercise factors, exercise controls blood perfusion, oxygen consumption, and body temperature, and these physiological changes may also affect NK cell function, perhaps even directly.

Tumor Perfusion, Vascularization, and Resolution of Intratumoral Hypoxia

Direct control of blood circulation by the parasympathetic nervous system during exercise ensures blood perfusion to vital organs and contracting muscles. Studies have shown that endurance training is associated with increased angiogenesis and vascularization in orthotropic mammary and prostate tumors from exercising mice and that such increased vascularization is correlated with enhanced blood perfusion and reduced tumor growth [72–76]. In addition to these long-term training studies in mice, McCullough et al. have shown that perfusion of orthotropic prostate tumors increases immediately during treadmill running in tumor-bearing rats (as evidenced by in vivo injection of Hoechst-33342 stained tumor cells prior to animal culling) [74]. There are several benefits to improved intratumoral vascularization and blood perfusion; first, these increase the accessibility of circulating immune cells and antineoplastic medication to tumors; second, increased blood perfusion can limit intratumoral hypoxia. Using human cell lines in vitro as well as in vivo renal cell and breast cancer mouse models, hypoxia has been shown to inhibit NK cell-mediated lysis, through a mechanism that might involve auto- phagic degradation of granzyme B [76–78]. Thus, resolution of intratumoral hypoxia could indirectly augment the cytotoxicity of tumor-infiltrating NK cells. Furthermore, NK cells have been shown to be recruited to sites of inflammation through a multifaceted chemotactic response, which depends on intratumoral expression‎ of a range of chemokines and NK cell receptor ligands [79]. Yet, in a hypoxic tumor microenvironment, NK cell activating ligands are shed from tumor cells, providing a mechanism where a hypoxic state can facilitate tumoral escape from immune recognition [80–82]. In line with the exercise-dependent resolution of intratumoral hypoxia, tumors from exercising mice have been shown to display marked upregulation of chemokines (CCL3, CXCL1, CX3CL1, Chemerin) and NK cell activating receptor ligands (Mult1, H60a, Clr-b) [9]. While increased tumor perfusion has been previously regarded as unfavorable, the benefits of increasing access of immune cells and antineoplastic drug delivery, in addition to resolving hypoxia-related intratumoral stress may, in the right setting, outweigh the disadvantage of providing more nutrient and oxygen to the tumors through perfusion.

Exercise and Increased Body Temperature

During a bout of exercise, the core temperature of both humans and mice rises [83]. Using transplantable and genetic tumor mouse models, such increases in core body temperature have been observed in systemic thermal therapy (STT), or in induced hyperthermia, and have demonstrated delayed tumor growth through a mechanism involving increased immune cell infiltration [69,84–86]. In a xenograft breast cancer model, the hyperthermia-induced delay in tumor growth was attributed to increased infiltration of NK cells, mediating tumor cell apoptosis at the tumor site [84]. The same study showed that hyperthermia increased the diameter of blood vessels within the tumor and hypothesized that the increased infiltration of immune cells might be correlated to a greater accessibility of immune cells inside the tumor. This hypothesis is supported by similar studies dissecting the effects of thermal therapy on immune cell trafficking [69,85]. Hyperthermia has been shown to modify the tumor vascu- lature by inducing IL-6 trans-signaling in tumor blood vessels, upregulating endothelial adhesion molecules, and making them more permissible to CD8+ T cell trafficking into tumors [85]. This study, however, did not assess the trafficking of NK cells into tumors. Nevertheless, in light of previous studies [84], including the recent report of exercise- mediated mobilization and IL-6-dependent increase of tumor-infiltrating NK cells on tumor control [9], one might speculate the occurrence of a similar trafficking mechanism during exercise. Furthermore, in vitro studies suggest that colon cancer target cells are more readily killed by NK cells due to hyperthermia-induced upregulation of the NK cell activating receptor ligand MICA [86].

Thus, not only does exercise increase the vascularization and perfusion of tumors, facilitating the access of immune cell subsets to the tumor site, but exercise-induced hyperthermia may also make the tumor vasculature more permissible to immune transmigration into the tumor (Figure 3).

Clinical Perspectives

The clinical importance of mobilizing NK cells and increasing intratumoral immune cell infiltration is well established. In a longitudinal study in healthy volunteers (n = 3625), it was demonstrated that individuals harboring NK cells with low natural cytotoxicity presented a significantly higher incidence of cancer after 11 years [87]. Moreover, the degree of immune cell infiltration is a strong predictor of clinical outcome across most cancers [88,89]. The link between immune cell infiltration in tumors to prognosis and treatment response prompted Angell and Galon to propose an ‘Immunoscore’, which quantifies the immune context, as a clinical tool to link immune cell infiltration to treatment response [90].

In addition to the direct role of NK cells in tumor cell killing, NK cells may also activate cells of the adaptive and innate immune responses, including Th1 T cells and M1 macrophages through secretion of IFN-g [91,92]. Currently, clinical attention has been placed on generating an inflammatory intratumoral environment through the use of immune checkpoint blockade ther- apy, among other approaches [93]. The mechanistic insight into the immunological role of exercise indicates that exercise might be able to enhance the delivery of such therapeutic intratumoral adaptations, by increasing immune cell infiltration, and triggering alterations in the intratumoral environment that lead to increased vascularization and tumor perfusion (Box 4).

Box 4. Clinician's Corner

International review boards recommend cancer patients to engage in exercise training [104].
At present, more than 1000 exercise intervention trials in cancer patients are registered at Clinicaltrials.gov.

Yet, the research field of exercise oncology generally lacks a mechanistic understanding of how exercise directly affects tumor biology and growth.

Detailed insight into the mechanistic effects of exercise on tumor biology and growth will provide a strong rationale for prescribing effective exercise interventions to cancer patients.

Evidence reviewed here suggests that exercise at high intensity, but short duration, can mobilize cytotoxic NK cells, and this may regulate tumor growth. 

Concluding Remarks

Exercise plays an increasing role as supportive therapy for cancer patients, aiming to improve patients’ quality of life, cancer-related fatigue, and physical functioning. In addition, emerging evidence suggests that exercise may also have direct antioncogenic effects on disease progression [7,9]. To pursue exercise as an ‘anticancer medicine’, detailed insight into the underlying mecha- nistic effects is warranted. Scientific research is currently endeavoring on this, and some of the first studies have indicated that NK cells play a pivotal role linking exercise to cancer control [8,9]. With this insight, exercise training in cancer patients may potentially move from a ‘one fits all’ approach to a prescription framework, based on extensive physiological knowledge of the effects of different modes of exercise [94,95]. In light of the effects of exercise-induced catecholamines and exercise factors on the control of NK cell mobilization and function, as well as on tumor accessibility through enhanced tumor vascularization, perfusion, and transmigration, a regimen of high intensity exercise endurance training might be prescribed to specifically stimulate this response.

In conclusion, although many questions remain (see Outstanding Questions), we propose that exercise may represent a promising complementary anticancer therapeutic approach based on the effect on NK cell mobilization and activation, as well as on changes to blood perfusion and body core temperature. These effects may reinforce immune cell distribution and transmigration into tumors, facilitating tumor lysis. Consequently, this represents an area of investigation that should be pursued and that may yield promising (and more cost-effective) therapeutic antitu- moral strategies.

Acknowledgments

This work was supported by grants from the Danish Cancer Society, the Region of Copenhagen Research funds, and TrygFonden. The Centre for Physical Activity Research (CFAS) is further supported by a grant from TrygFonden. During the study period, the Centre of Inflammation and Metabolism (CIM) was supported by a grant from the Danish National Research Foundation (DNRF55).  

References

1. Tipton, C. (2014) History of Exercise Physiology, Human Kinetics 26. ISBN-13: 9780736083690

2. Pedersen, B.K. and Saltin, B. (2015) Exercise as medicine– evidence for prescribing exercise as therapy in 26 different 27. chronic diseases. Scand. J. Med. Sci. Sports 25, 1–72

3. Ballard-Barbash, R. et al. (2012) Physical activity, biomarkers,
   and disease outcomes in cancer survivors: a systematic review. 28. J. Natl. Cancer Inst. 104, 815–840

4. Irwin, M.L. et al. (2008) Influence of pre- and postdiagnosis physical activity on mortality in breast cancer survivors: the 29. health, eating, activity, and lifestyle study. J. Clin. Oncol. 26, 3958–3964

5. Meyerhardt, J.A. et al. (2006) Physical activity and survival after 30. colorectal cancer diagnosis. J. Clin. Oncol. 24, 3527–3534

6. Friedenreich, C.M. et al. (2016) Physical activity and survival after 31. prostate cancer. Eur. Urol. Published online January 7, 2016. http://dx.doi.org/10.1016/j.eururo.2015.12.032

7. Hvid, T. et al. (2016) Effect of a 2-year home-based endurance training intervention on physiological function and PSA doubling 32. time in prostate cancer patients. Cancer Cause Control 27, 165–
   174

8. Pedersen, L. et al. (2015) Effects of exercise on tumor physiology and metabolism. Cancer J. 21, 111–116

Murray, D.R. et al. (1992) Sympathetic and immune interactions during dynamic exercise. Mediation via a beta 2-adrenergic- dependent mechanism. Circulation 86, 203–213
Starkie, R.L. et al. (2001) Effect of adrenergic blockade on lymphocyte cytokine production at rest and during exercise. Am. J. Physiol. Cell Physiol. 281, C1233–C1240

Ogawa, K. et al. (2005) A single bout of exercise influences natural killer cells in elderly women, especially those who are habitually active. J. Strength Cond. Res. 19, 45–50

McFarlin, B.K. et al. (2005) Chronic resistance exercise training improves natural killer cell activity in older women. J. Gerontol. A Biol. Sci. Med. Sci. 60, 1315–1318

Nieman, D.C. et al. (1999) Influence of obesity on immune func- tion. J. Am. Diet. Assoc. 99, 294–299
Campbell, P.T. et al. (2008) Effect of exercise on in vitro immune function: a 12-month randomized, controlled trial among postmenopausal women. J. Appl. Physiol. 104, 1648–1655

Simpson, R.J. et al. (2012) Exercise and the aging immune system. Ageing Res. Rev. 11, 404–420

9. Pedersen, L. et al. (2016) Voluntary running suppresses tumor growth through epinephrine- and IL-6-dependent NK cell mobi- 34. lization and redistribution. Cell Metab. 23, 554–562

10. Kiessling, R. et al. (1975) “Natural” killer cells in the mouse. I. Cytotoxic cells with specificity for mouse Moloney leukemia cells. Specificity and distribution according to genotype. Eur. J. Immu- 35. nol. 5, 112–117

11. Artis, D. and Spits, H. (2015) The biology of innate lymphoid cells. Nature 517, 293–301

12. Sun, J.C. and Lanier, L. (2011) NK cell development, homeosta- 36. sis and function: parallels with CD8+ T cells. Nat. Rev. Immunol.
    11, 645–657

13. Robertson, M.J. and Ritz, J. (1990) Biology and clinical relevance 37. of human natural killer cells. Blood 76, 2421–2438

14. Nielsen, H.B. et al. (1996) Lymphocyte, NK and LAK cell responses to maximal exercise. Int. J. Sports Med. 17, 60–65 38.

15. Walsh, N.P. et al. (2011) Position statement. Part one: immune function and exercise. Exerc. Immunol. Rev. 17, 6–63

16. Millard, A.L. et al. (2013) Brief exercise increases peripheral blood NK cell counts without immediate functional changes, but impairs their responses to ex vivo stimulation. Front. Immunol. 4, 125

17. Schwindt, C.D. et al. (2007) Do circulating leucocytes and lym- phocyte subtypes increase in response to brief exercise in chil- dren with and without asthma? Br. J. Sports Med. 41, 34–40

18. Nielsen, H.B. et al. (1996) Lymphocytes and NK cell activity during repeated bouts of maximal exercise. Am. J. Physiol. 271, R222–R227

19. Gabriel, H. et al. (1992) Mobilization of circulating leucocyte and lymphocyte subpopulations during and after short, anaerobic exercise. Eur. J. Appl. Physiol. Occup. Physiol. 65, 164–170

20. Timmons, B.W. and Cieslak, T. (2008) Human natural killer cell subsets and acute exercise: a brief review. Exerc. Immunol. Rev. 14, 8–23

21. Shephard, R.J. (2003) Adhesion molecules, catecholamines and leucocyte redistribution during and following exercise. Sports Med. 33, 261–284

22. Dimitrov, S. et al. (2010) Selective mobilization of cytotoxic leu- kocytes by epinephrine. J. Immunol. 184, 503–511

23. Kappel, M. et al. (1991) Evidence that the effect of physical exercise on NK cell activity is mediated by epinephrine. J. Appl. Physiol. 70, 2530–2534

24. Kappel, M. et al. (1998) Effects of elevated plasma noradrenaline concentration on the immune system in humans. Eur. J. Appl. Physiol. Occup. Physiol. 79, 93–98

25. Landmann, R. (1992) Beta-adrenergic receptors in human leu- kocyte subpopulations. Eur. J. Clin. Invest. 22, 30–36

Bigley, A.B. et al. (2012) NK-cells have an impaired response to acute exercise and a lower expression‎ of the inhibitory receptors KLRG1 and CD158a in humans with latent cytomegalovirus infection. Brain Behav. Immun. 26, 177–186

Maisel, A.S. et al. (1990) Adrenergic control of circulating lym- phocyte subpopulations. Effects of congestive heart failure, dynamic exercise, and terbutaline treatment. J. Clin. Invest. 85, 462–467

Evans, E.S. et al. (2015) Impact of acute intermittent exercise on natural killer cells in breast cancer survivors. Integr. Cancer Ther. 14, 436–445
Fairey, A.S. et al. (2005) Randomized controlled trial of exercise and blood immune function in postmenopausal breast cancer survivors. J. Appl. Physiol. 98, 1534–1540

Nieman, D.C. et al. (1995) Moderate exercise training and natural killer cell cytotoxic activity in breast cancer patients. Int. J. Sports Med. 16, 334–337

33. Maisel, A.S. et al. (1990) Beta-adrenergic receptors in lympho- cyte subsets after exercise. Alterations in normal individuals and patients with congestive heart failure. Circulation 82, 2003–2010

39. Na, Y.M. et al. (2000) Exercise therapy effect on natural killer cell cytotoxic activity in stomach cancer patients after curative sur- gery. Arch. Phys. Med. Rehabil. 81, 777–779

40. Beziat, V. et al. (2010) NK cell terminal differentiation: correlated stepwise decrease of NKG2A and acquisition of KIRs. PLoS ONE 5, e11966

41. Bigley, A.B. et al. (2014) Acute exercise preferentially redeploys NK-cells with a highly-differentiated phenotype and augments cytotoxicity against lymphoma and multiple myeloma target cells. Brain Behav. Immun. 39, 160–171

42. Benschop, R.J. et al. (1994) The effects of beta-adrenoceptor stimulation on adhesion of human natural killer cells to cultured endothelium. Br. J. Pharmacol. 113, 1311–1316

43. Catoire, M. and Kersten, S. (2015) The search for exercise factors in humans. FASEB J. 29, 1615–1628

44. Pedersen, B.K. and Febbraio, M.A. (2012) Muscles, exercise and obesity: skeletal muscle as a secretory organ. Nat. Rev. Endo- crinol. 8, 457–465

45. Benatti, F.B. and Pedersen, B.K. (2015) Exercise as an anti- inflammatory therapy for rheumatic diseases-myokine regulation. Nat. Rev. Rheumatol. 11, 86–97

46. Rinnov, A. et al. (2014) Endurance training enhances skeletal muscle interleukin-15 in human male subjects. Endocrine 45, 271–278

47. Nielsen, A.R. et al. (2007) Expression‎ of interleukin-15 in human skeletal muscle effect of exercise and muscle fibre type compo- sition. J. Physiol. 584, 305–312

48. Huang, P.L. et al. (2015) Skeletal muscle interleukin 15 pro- motes CD8+ T-cell function and autoimmune myositis. Skelet. Muscle 5, 33 

    49. Pistilli, E.E. et al. (2011) Loss of IL-15 receptor / alters the endurance, fatigability, and metabolic characteristics of mouse fast skeletal muscles. J. Clin. Invest. 121, 3120–3132

    50. Mortier, E. et al. (2008) IL-15R/ chaperones IL-15 to stable dendritic cell membrane complexes that activate NK cells via trans presentation. J. Exp. Med. 205, 1213–1225

    51. Noguchi, M. et al. (1993) Interleukin-2 receptor g chain mutation results in X-linked severe combined immunodeficiency in humans. Cell 73, 147–157

    52. Kennedy, M.K. et al. (2000) Reversible defects in natural killer and memory CD8T cell lineages in interleukin 15-deficient mice. J. Exp. Med. 191, 771–780

    J. Appl. Physiol. Published online April 28, 2016. http://dx.doi. org/10.1152/japplphysiol.00266.2016

    73. Betof, A.S. et al. (2015) Modulation of murine breast tumor vascularity, hypoxia and chemotherapeutic response by exer- cise. J. Natl. Cancer Inst. 107, djv040

    74. McCullough, D.J. et al. (2014) Modulation of blood flow, hypoxia, and vascular function in orthotopic prostate tumors during exer- cise. J. Natl. Cancer Inst 106, dju036

    75. McCullough, D.J. et al. (2013) Effects of exercise training on tumor hypoxia and vascular function in the rodent preclinical orthotopic prostate cancer model. J. Appl. Physiol. 115, 1846–1854

    53. Marcais, A. et al. (2014) The metabolic checkpoint kinase mTOR
    is essential for IL-15 signaling during the development and acti-
    vation of NK cells. Nat. Immunol. 15, 749–757 6832

    54. Donnelly, R.P. et al. (2014) mTORC1-dependent metabolic reprogramming is a prerequisite for NK cell effector function. J. Immunol. 193, 4477–4484

    55. Sinclair, L.V.R.J. et al. (2013) Control of amino-acid transport by antigen receptors coordinates the metabolic reprogramming essential for T cell differentiation. Nat. Immunol. 14, 500–508

    56. Keating, S.E. et al. (2016) Metabolic reprogramming supports IFN- g production by CD56bright NK cells. J. Immunol. 196, 2552–2560

    57. Conlon, K.C. et al. (2015) Redistribution, hyperproliferation, acti- vation of natural killer cells and CD8T cells, and cytokine produc- tion during first-in-human clinical trial of recombinant human interleukin-15 in patients with cancer. J. Clin. Oncol. 33, 74–82

    58. Waldmann, T.A. (2014) Interleukin-15 in the treatment of cancer. Expert Rev. Clin. Immunol. 10, 1689–1701

    59. Norheim, F. et al. (2011) Proteomic identification of secreted proteins from human skeletal muscle cells and expression‎ in response to strength training. Am. J. Physiol. Endocrinol. Metab. 301, E1013–E1021

    60. Haugen, F. et al. (2010) IL-7 is expressed and secreted by human

        skeletal muscle cells. Am. J. Physiol. Cell Physiol. 298, C807–C816

    61. Carotta, S. et al. (2011) Identification of the earliest NK-cell precursor in the mouse BM. Blood 117, 5449–5452

    62. Vosshenrich, C.A. et al. (2005) Roles for common cytokine receptor g-chain-dependent cytokines in the generation, differ- entiation, and maturation of NK cell precursors and peripheral NK cells in vivo. J. Immunol. 174, 1213–1221

    63. He, Y.W. and Malek, T.R. (1996) Interleukin-7 receptor / is essential for the development of gd + T cells, but not natural killer cells. J. Exp. Med. 184, 289–293

    64. Satoh-Takayama, N. et al. (2010) IL-7 and IL-15 independently program the differentiation of intestinal CD3-NKp46+ cell subsets from Id2-dependent precursors. J. Exp. Med. 207, 273–280

    65. Vosshenrich, C.A. et al. (2006) A thymic pathway of mouse natural killer cell development characterized by expression‎ of GATA-3 and CD127. Nat. Immunol. 7, 1217–1224

    66. Keller, C. et al. (2005) Effect of exercise, training, and glycogen availability on IL-6 receptor expression‎ in human skeletal muscle. J. Appl. Physiol. 99, 2075–2079

    67. Ali, T.H. et al. (2014) Enrichment of CD56dimKIR + CD57 + highly cytotoxic NK cells in tumour-infiltrated lymph nodes of melanoma patients. Nat. Commun. 5, 5639

    68. Rabinowich, H. et al. (1993) Response of human NK cells to IL-6 alterations of the cell surface phenotype, adhesion to fibronectin and laminin, and tumor necrosis factor-//b secretion. J. Immu- nol. 150, 4844–4855

    69. Fisher, D.T. et al. (2011) IL-6 trans-signaling licenses mouse and human tumor microvascular gateways for trafficking of cytotoxic T cells. J. Clin. Invest. 121, 3846–3859

    70. Clahsen, T. and Schaper, F. (2008) Interleukin-6 acts in the fashion of a classical chemokine on monocytic cells by inducing integrin activation, cell adhesion, actin polymerization, chemo- taxis, and transmigration. J. Leukoc. Biol. 84, 1521–1529

    71. Dethlefsen, C. et al. (2013) The role of intratumoral and systemic IL-6 in breast cancer. Breast Cancer Res. Treat. 138, 657–664

    72. Garcia, E. et al. (2016) Blood flow responses to mild-intensity exercise in ectopic versus orthotopic prostate tumors; depen- dence upon host-tissue hemodynamics and vascular reactivity.

    77. Tittarelli, A. et al. (2015) The selective degradation of synaptic connexin 43 protein by hypoxia-induced autophagy impairs nat- ural killer cell-mediated tumor cell killing. J. Biol. Chem. 290, 23670–23679

    78. Baginska, J. et al. (2013) Granzyme B degradation by autophagy decreases tumor cell susceptibility to natural killer-mediated lysis under hypoxia. Proc. Natl. Acad. Sci. U.S.A. 110, 17450–17455

    79. Pachynski, R.K. et al. (2012) The chemoattractant chemerin suppresses melanoma by recruiting natural killer cell antitumor defenses. J. Exp. Med. 209, 1427–1435

    80. Schlecker, E. et al. (2014) Metalloprotease-mediated tumor cell shedding of B7-H6, the ligand of the natural killer cell-activating receptor NKp30. Cancer Res. 74, 3429–3440

    81. Sceneay, J. et al. (2012) Primary tumor hypoxia recruits CD11b+/ Ly6Cmed/Ly6G+ immune suppressor cells and compromises NK cell cytotoxicity in the premetastatic niche. Cancer Res. 72, 3906–3911

    82. Deng, W. et al. (2015) Antitumor immunity. A shed NKG2D ligand that promotes natural killer cell activation and tumor rejection. Science 348, 136–139

    83. Evans, S.S. et al. (2015) Fever and the thermal regulation of immunity: the immune system feels the heat. Nat. Rev. Immunol. 15, 335–349

    84. Burd, R. et al. (1998) Tumor cell apoptosis, lymphocyte recruit- ment and tumor vascular changes are induced by low tempera- ture, long duration (fever-like) whole body hyperthermia. J. Cell. Physiol. 177, 137–147

    85. Chen, Q. et al. (2006) Fever-range thermal stress promotes lym- phocyte trafficking across high endothelial venules via an interleukin 6 trans-signaling mechanism. Nat. Immunol. 7, 1299–1308

    86. Ostberg, J.R. et al. (2007) Enhancement of natural killer (NK) cell cytotoxicity by fever-range thermal stress is dependent on NKG2D function and is associated with plasma membrane NKG2D clustering and increased expression‎ of MICA on target cells. J. Leukoc. Biol. 82, 1322–1331

    87. Imai, K. et al. (2000) Natural cytotoxic activity of peripheral-blood lymphocytes and cancer incidence: an 11-year follow-up study of a general population. Lancet 356, 1795–1799

    88. Zhang, L. et al. (2003) Intratumoral T cells, recurrence and survival in epithelial ovarian cancer. N. Engl. J. Med. 348, 203–213

    89. Galon, J. et al. (2006) Type, density, and location of immune cells within human colorectal tumors predict clinical outcome. Science 313, 1960–1964

    90. Angell, H. and Galon, J. (2013) From the immune contexture to the Immunoscore: the role of prognostic and predictive immune markers of cancer. Curr. Opin. Immunol. 25, 261–267

    91. Dungan, L.S. et al. (2014) Innate IFN-g promotes development of experimental autoimmune encephalomyelitis: a role for NK cells and M1 macrophages. Eur. J. Immunol. 44, 2903–2917

    92. Martin-Fontecha, A. et al. (2004) Induced recruitment of NK cells to lymph nodes provides IFN-g for TH1 priming. Nat. Immunol. 5, 1260–1265

    93. Topalian, S.L. et al. (2015) Immune checkpoint blockade: a common denominator approach to cancer therapy. Cancer Cell 27, 450–461

    94. Jones, L.W. (2015) Precision oncology framework for investigation of exercise as treatment for cancer. J. Clin. Oncol. 33, 4134–4137  

95. Sasso, J.P. et al. (2015) A framework for prescription in exercise- oncology research. J. Cachexia Sarcopenia Muscle 6, 115–124

96. Mishra, S.I. et al. (2012) Exercise interventions on health-related quality of life for people with cancer during active treatment. Cochrane Database Syst. Rev. 8, CD008465

97. Mishra, S.I. et al. (2012) Exercise interventions on health-related quality of life for cancer survivors. Cochrane Database Syst. Rev. 8, CD007566

98. Agudelo, L.Z. et al. (2014) Skeletal muscle PGC-1/1 modulates kynurenine metabolism and mediates resilience to stress- induced depression. Cell 159, 33–45

99. Holmes, M.D. et al. (2005) Physical activity and survival after breast cancer diagnosis. JAMA 293, 2479–2486

100. Kenfield, S.A. et al. (2011) Physical activity and survival after prostate cancer diagnosis in the health professionals follow-up study. J. Clin. Oncol. 29, 726–732

101. Bezman, N.A. et al. (2012) Molecular definition of the identity and activation of natural killer cells. Nat. Immunol. 13, 1000–1009

102. Robinette, M.L. et al. (2015) Transcriptional programs define molecular characteristics of innate lymphoid cell classes and subsets. Nat. Immunol. 16, 306–317

103. Warren, H.S. and Smyth, M.J. (1999) NK cells and apoptosis. Immunol. Cell Biol. 77, 64–75

104. Schmitz, K.H. et al. (2010) American College of Sports Medicine roundtable on exercise guidelines for cancer survivors. Med. Sci. Sports Exerc. 42, 1409–1426