natural cytotoxicity that is mediated by natural killer cells (NKs) is important in enabling the innate immune system to cope efficiently with malignancies (17, 30). Epidemiological studies have demonstrated that regular exercise reduces the incidence of mortality associated with most cancers (20, 40). However, patients with malignancies may have a sedentary lifestyle because muscle strength and cardiopulmonary fitness progressively decline, limiting their capacity to perform exercise (40). Acclimatization to systemic hypoxic exposure improves exercise performance by enhancing oxygen delivery and utilization (34, 39). Moreover, systemic hypoxia also increases the recruitment of NKs to the circulation during exercise (12). However, initial increases in NK count and activity during exposure to extreme hypoxia are rapidly followed by NK cytopenia and dysfunction during the recovery phase (11, 22). To the authors' knowledge, an effective antitumor strategy of exercise combined with hypoxia based on promotion of exercise-induced NK effector function and minimization of the risk of immune suppression in the recovery period has not yet been established.

Both the ability to recognize target cells and the production of cytotoxic proteins/cytokines are fundamental to the antitumor cytotoxicity of NKs (13, 27). NKs express an array of activating and inhibitory receptors that are capable of recognizing transformed cells, without the need for immunization or preactivation (13). In natural cytotoxicity, the NKs kill their cellular targets via both granule (perforin/granzyme)- and death receptor (Fas/CD95)-mediated cytotoxic pathways (27). According to earlier studies, acute exercise mobilizes the senescent lymphocytes that enter the bloodstream from peripheral tissues (25, 26) and modulates the antitumor cytotoxicity of NKs by altering cellular granule protein contents (36) in an intensity-dependent manner. However, no clear and comprehensive picture of exercise-mediated NK subset mobilization or natural cytotoxicity of NKs under various hypoxic conditions has become available.

Nasopharyngeal carcinoma (NPC), a common cancer in Southern China, is a highly malignant and commonly metastasized tumor (6, 9). The goal of this study was to clarify the effect of acute hypoxic exercise on the subset mobilization and natural cytotoxicity of NKs. We hypothesize that systemic hypoxia affects redistribution of NK subset populations and anti-NPC cytotoxicity of NKs during exercise, in which the magnitude of response is dependent on the level of hypoxia. To answer the above questions, this investigation compared how various exercise regimens with/without hypoxia affect 1) NK count and subset distribution, 2) cytotoxic protein levels in NKs, 3) capacity of NKs to bind to NPCs, and 4) NK-induced NPC apoptosis.

METHODS

Subjects.

The Ethics Committee of Chang Gung Memorial Hospital reviewed and approved the study protocol (no. 96–1284B). The procedures follow institutional guidelines. Sixteen healthy sedentary men (mean ± SE: age, 24.2 ± 1.2 yr; height, 172.1 ± 1.4 cm; weight, 67.5 ± 3.2 kg; body mass index, 22.8 ± 1.1 kg/m²) participated after they had provided informed consent. All subjects were nonsmokers, did not use medication/vitamins, were infection free and cardiopulmonary risk free, and were recruited from Chang Gung University, Taiwan. Moreover, no subject had engaged in any regular physical activity or mountain climbing for at least 1 yr before the study. Subjects fasted for at least 8 h and were instructed to refrain from exercise for at least 48 h before the study. All subjects arrived at the testing center at 9:00 AM to eliminate any possible diurnal effects.

Experimental protocol and blood collection.

Subjects came to the laboratory on 6 days to follow six distinct experimental protocols in an air-conditioned normobaric hypoxia chamber (Colorado Mountain Room). The chamber temperature was maintained at 22°C ± 0.5°C, and the relative humidity was 60% ± 5%. The chamber had a CO₂ scrubber to eliminate CO₂ from the air (<3,500 ppm) (34, 35, 39). The first protocol comprised 2 min of unloaded pedaling, after which the workload was increased incrementally by 20–30 W every 3 min until exhaustion under 21% O₂ air [high-intensity exercise (HE)] (38). This exercise test determined the maximal oxygen consumption (V̇o_{2 max}) in test subjects, as described (38). Subject V̇o_{2 max} was 42.5 ± 2.3 ml·min⁻¹·kg⁻¹ (mean ± SE) (Table 1). Two weeks after the first test, another five protocols were randomly assigned to subjects in a counterbalanced order. They were separated by intervals of 2 wk to ensure complete recovery between trials. In protocols 2–6, subjects exercised at 50% of predetermined V̇o_{2 max} (ME) on a bicycle ergometer (Corvial 400; Lode) under 12%, 15%, and 21% O₂ air for 30 min and also rested in the sitting position under 12% and 15% O₂ air for 30 min. Arterial O₂ saturation (SaO₂) was measured by finger pulse oximetry (model 9500; Nonin Onyx); blood pressure (BP) and heart rate (HR) were monitored using an automatic blood pressure system (model 412; Quinton) (34). For safety reasons, the test was terminated immediately when the level of O₂ saturation dropped to <70% or the subject complained of discomfort. All subjects were free of acute mountain sickness symptoms during the experimental period. Moreover, all compliance rates for these interventions were 100%.

Initially before the experimental tests, and immediately after and 2 h after the experimental tests, blood samples were collected from an antecubital vein using a clean venipuncture (20-gauge needle) under controlled venous stasis at 40 Torr. The first 2 ml of blood was discarded, and the remaining blood was used to measure hematological parameters and blood cell functions. Blood lactate concentration were measured using an i-STAT clinical analyzer (+CG4; i-STAT) (34, 35). Blood NK count was determined by two-color flow cytometry using a Simultest CD3/CD16+CD56 kit (Becton Dickinson) (36).

NPC culture and NK isolation.

Dr. J.-K. Chen (Dept. of Physiology, Chang Gung Univ., Taiwan) kindly provided the human NPC cell line (NPC 076). The cells were cultured in a DMEM/F12 medium (Sigma), supplemented with 10% vol/vol fetal bovine serum (Gibco) and 100 IU/ml penicillin-streptomycin (Sigma) in an atmosphere of 5% CO₂ at 37°C (33, 36).

Thirty-milliliter blood samples were transferred to polypropylene tubes that contained sodium citrate (3.8 g/dl: 1 vol to 9 vol of blood) (Sigma). Peripheral blood mononuclear cells (PBMCs) were separated by density-gradient centrifugation on the Lymphoprep tube (Nycomed) (36). NKs were then isolated from the PBMCs by the MACS-negative immunomagnetic selection method using a commercial NK Isolation Kit II (Miltenyi Biotechnology) (36). T cells, B cells, stem cells, dendritic cells, monocytes, granulocytes, and erythroid cells were indirectly magnetically labeled by using a cocktail of biotin-conjugated antibodies (Miltenyi Biotechnology) and the NK Cell MicroBead Cocktail (Miltenyi Biotechnology). Isolation of highly pure NK cells was achieved by depletion of magnetically labeled cells. The purity of NKs, determined by flow cytometry using anti-CD56 and anti-CD3 monoclonal antibodies, was >90%. The NK number was adjusted by adding DMEM/F12 medium (Sigma) to 2 × 10⁶ cells/ml.

Blood NK phenotypes.

The NK suspensions (2 × 10⁶ cells/ml) were incubated with a saturating concentration (10 μg/ml) of monoclonal anti-human NKGD2 (Santa Cruz Biotechnology), KLRG1 (Santa Cruz Biotechnology), CD57 (eBioscience), CD28 (eBioscience), CD62L (eBioscience), or CD11a (eBioscience) monoclonal antibody that had been conjugated with fluorescein isothiocyanate (FITC), or anti-rabbit IgG (eBioscience) control antibody that had been conjugated with FITC in the dark for 30 min at 4°C. NKs that had been treated with the control antibody (eBioscience) were utilized to correct for background fluorescence. Following fixation with 4% formaldehyde (Sigma) in Hanks' balanced salt solution (HBSS) (Sigma), the mean fluorescence obtained from 5,000 events, representing the NKs, was calculated using a FACScan flow cytometer (Becton Dickinson).

Cellular perforin, granzyme B and interferon-γ staining of NKs.

One milliliter of commercial fixation buffer (eBioscience) was added to 2 × 10⁶ purified NKs in darkness for 20 min at room temperature. Following fixation, the cells were twice washed and permeabilized using a commercial permeabilization washing buffer (eBioscience); they were then incubated in darkness for 30 min at 4°C with a saturation concentration (10 μg/ml) of FITC-conjugated anti-perforin monoclonal antibody (eBioscience), FITC-conjugated anti-granzyme B monoclonal antibody (eBioscience), FITC-conjugated anti-interferon-γ (IFN-γ) monoclonal antibody (eBioscience), or FITC-conjugated anti-IgG control antibody (eBioscience). Subsequently, the fixed and intracellularly labeled cells were resuspended with 1 ml of commercial cell staining buffer (eBioscience). Finally, the mean fluorescent intensity obtained from 5,000 events was measured using FACScan flow cytometry (Becton Dickinson) (36).

NK-NPC binding.

The NK suspension was incubated with a saturating concentration of monoclonal anti-human CD56 antibody conjugated with phycoerythrin (CD56-PE) (Becton Dickinson) in darkness for 30 min at 4°C, and then washed twice with HBSS (Sigma). Then 0.5 ml of CD56-PE-labeled NK suspension (2 × 10⁶ cells/ml) was added to 0.5 ml of NPC (1 × 10⁵ cells/ml) suspension (20 NKs:1 NPC) for 4 h with shaking at 100 rpm at 37°C in 5% CO₂. Thereafter, the cell mixtures were transferred immediately into polypropylene tubes that contained 4% formaldehyde (Sigma) in HBSS (Sigma) to terminate the cell-cell interactions. The mean fluorescence from 5,000 events, representing the CD56-PE-labeled NKs bound to NPCs, was then determined using a FACScan flow cytometer (Becton Dickinson) (36). Briefly, the NPCs were gated separately from the NKs on the basis of forward/sideward scatter; then, the number of PE-stained events observed in the NPCs gate was expressed as a percentage of defined NK-NPC bindings (36).

NK-induced CD95 expression on NPC.

Mixtures of NKs (2 × 10⁶ cells/ml) and NPCs (1 × 10⁵ cells/ml) in DMEM/F12 medium (Sigma) were incubated with shaking at 100 rpm at 37°C in 5% CO₂ for 4 h, and then centrifuged at 1,600 g at 4°C for 10 min. Pellets were subsequently collected in the lower layer. The pellets were incubated with a saturating concentration of monoclonal anti-human CD95 antibody that had been conjugated with FITC (eBioscience) in darkness for 30 min at 4°C, and then washed twice with HBSS (Sigma). The NPCs were gated separately from the NKs based on forward/sideward scattering using a FACScan flow cytometer (Becton Dickinson), and the FITC-stained events in the NPCs gate were then expressed as a percentage of defined CD95 expression on NPCs.

NK-induced phosphatidylserine exposure on NPC.

The phosphatidylserine (PS) exposure on NPC was detected by FACScan flow cytometry, using a commercial annexin V Cyt5 kit (BD Bioscience Pharmingen) (36). This assay is based on changes in the cell surface exposure of PS during the early stages of apoptosis and the selective affinity of annexin V for the phospholipids (31). Mixtures of NKs (2 × 10⁶ cells/ml) and NPCs (1 × 10⁵ cells/ml) in DMEM/F12 medium (Sigma) were incubated with shaking at 100 rpm at 37°C in 5% CO₂ for 4 h. The NPCs were gated separately from the NKs based on forward/sideward scattering using a FACScan flow cytometer (Becton Dickinson), and the number of apoptotic fluorescein-stained events in the NPCs gate was then expressed as a percentage of the number of defined apoptotic NPCs (36).

Statistical analysis.

Data are expressed as means ± SE. The statistical software package StatView IV was employed to analyze data. Experimental results were analyzed by six (sessions) × three (time sample points) repeated-measures ANOVA and Tukey's multiple range test to compare hematological parameters, NK counts, phenotypes and cytotoxic-related protein contents, NK-NPC binding and NK-induced CD95 expression and PS exposure on NPC initially before, immediately after, and 2 h after various experimental tests. The criterion for significance was P < 0.05.

RESULTS

Cardiopulmonary functions.

Both 12% and 15% O₂ ME induced more minute ventilation than 21% O₂ ME did (P < 0.05), while the V̇o₂ results did not differ among 12%-21% O₂ ME tests (Table 1). Although neither 12% nor 15% O₂ exposure at rest influenced blood lactate concentration, both 12% and 15% O₂ ME generated higher levels of lactate in blood than 21% O₂ ME (Table 1, P < 0.05). The decline in the degree of SaO₂ by hypoxia was related to the decrease in O₂ concentration in air, while ME reduced the level of SaO₂ to a greater extent by exposure to 12% or 15% O₂ compared with 21% O₂ (Table 1, P < 0.01). The 12% O₂, but not 15% O₂, exposure at rest increased HR, while both of these hypoxic exposures promoted the increase in HR by ME (Table 1, P < 0.05). Although either 12% or 15% O₂ exposure at rest did not alter systolic or diastolic BP, 12% O₂ ME had a higher level of systolic BP than 21% O₂ ME (Table 1, P < 0.05). However, these changes of cardiopulmonary and hematological parameters induced by normoxic or hypoxic exercise tests returned to the pretest state after the 2 h period of recovery.

Blood NK counts and phenotypes.

Although both 12% and 15% O₂ exposures at rest did not change total NK count in blood, the blood NK count increased in response to normoxic (21% O₂ ME and HE) and hypoxic (12% and 15% O₂ ME) exercise tests (Fig. 1, P < 0.05). However, initial increases in total blood NK count immediately after 21% O₂ HE and 12% O₂ ME were rapidly followed by NK cytopenia after the 2-h postexercise recovery period (Fig. 1, P < 0.05).

No significant changes in counts and percentages of the NK subsets with CD28⁺ (Table 2), CD57⁺ (Table 2), NKG2D⁺ (Table 3), KLRG1⁺ (Table 3), CD62L⁺ (Table 4), and CD11a⁺ (Table 4) in blood occurred immediately and 2 h after both 12% and 15% O₂ exposure at rest. In blood NK subset counts, although counts of the NKs with CD28⁺, CD57⁺, NKG2D⁺, KLRG1⁺, CD62L⁺, and CD11a⁺ were increased immediately after all normoxic (21% O₂ ME and HE) and hypoxic (12% and 15% O₂ ME) exercise tests, these NK subset counts were decreased below pretest level 2 h after 21% O₂ HE test (Tables 2–4, P < 0.05) (Fig. 2A). In blood NK subset distribution, 21% O₂ HE and 12% O₂ ME reduced the percentage of NKs with CD28⁺ (Table 2, P < 0.05), whereas the two normoxic/hypoxic exercise tests increased the percentages of NKs with CD57⁺ (Table 2, P < 0.05) and KLRG1⁺ (Table 3, P < 0.05). Additionally, the percentage of blood CD62L⁺ NKs (Table 4, P < 0.05) was decreased, and the percentage of blood CD11a⁺ NKs (Table 4, P < 0.05) was increased immediately following 12% and 15% O₂ ME, but not 21% O₂ ME. Furthermore, the blood NK subsets redistributed by normoxic and hypoxic exercise returned to the pretest state after the 2-h postexercise recovery period (Tables 2–4).

Perforin, granzyme B, and IFN-γ levels.

Both cellular perforin (Fig. 3A) and granzyme B (Fig. 3B) levels in blood NKs were increased immediately after 21% O₂ HE and the 12% and 15% O₂ ME (P < 0.05) (Fig. 2B), whereas cellular IFN-γ level (Fig. 4A) in blood NKs and IFN-γ concentration in NK-NPC coincubation (Fig. 4B) were elevated immediately or 2 h after the three tests (P < 0.05). However, neither 21% O₂ ME nor the 12% or 15% O₂ exposure at rest changed cellular perforin (Fig. 3A), granzyme B (Fig. 3B), or IFN-γ (Fig. 4A) level in blood NKs, as well as IFN-γ concentration in NK-NPC coincubation (Fig. 4B).

NK-NPC binding and NK-induced NPC CD95 expression and death.

The percentages of NPCs that were bound to NKs were increased immediately after 21% O₂ HE or the 12% and 15% O₂ ME (Fig. 5, P < 0.05), while NK-induced CD95 expression on NPC was enhanced immediately and 2 h after each of the three tests (Fig. 6, P < 0.05) (Fig. 2C). Furthermore, both 12% and 15% O₂ ME promoted NK-induced NPC apoptosis (NPCs stained with annexin V alone) (Fig. 7A, P < 0.05), whereas 21% O₂ HE facilitated NK-induced NPC apoptosis and necrosis (the NPCs stained simultaneously with annexin V and propidium iodide) (Fig. 7B, P < 0.05) (Fig. 2D). However, no significant change in NK-NPC binding, NK-induced CD95 expression, or apoptosis/necrosis of NPC occurred after either 21% O₂ ME or the 12%-15% O₂ exposure at rest (Figs. 5–7).

DISCUSSION

This investigation is the first to demonstrate clearly that systemic hypoxia affects mobilization of the NK subset populations and anti-NPC cytotoxicity of NKs during exercise. Although normoxic HE, but not ME, increased NK count and perforin/granzyme B/IFN-γ levels, capacity of NKs to bind to NPCs, as well as NK-induced CD95 expression and apoptosis of NPC, an extensive mobilization of the replicative senescent/inhibitory NKs into the peripheral blood compartment occurred under the HE regimen. Either 15% or 12% O₂ exposure at rest did not alter the NK subset distribution in blood or the NK-induced NPC apoptotic responses. However, both 15% O₂ and 12% O₂ ME increased cytotoxic protein levels in blood NKs, NK-NPC binding, and anti-NPC cytotoxicity of NKs, whereas only 12% O₂ ME mobilized higher percentages of the NKs with CD57⁺/CD28⁻/KLRG1⁺ into the bloodstream. Several previous studies have indicated suppressed peripheral NK immunity in patients with NPC (18, 29, 41). Therefore, acute 12% O₂ ME may be an effective antitumor strategy of exercise combined with hypoxia in the patients with NPC by promoting NK effector function and minimizing mobilization of replicative senescent/inhibitory NKs.

The balance between activating and inhibitory signals that are transmitted by cell surface receptors regulates NK immune responses (13). The NKG2D is capable of controlling NK activation via interactions with ligands expressed on the surface of transformed cells (19). Conversely, the killer cell lectin-like receptor G1 (KLRG1) ligates with cadherins, which are ubiquitously expressed in various tissues, inhibiting NK expansion and function (10, 23). Furthermore, the KLRG1 (32) and CD57 (3) have been identified as markers of replicative senescence on human lymphocytes, whereas the KLRG1 is coupled with a reduced number of lymphocytes expressing cell surface antigen CD28 (1), which is an important costimulatory molecule for activation and proliferation of naive lymphocytes. According to various investigations, HE (>75% V̇o_{2 max}) mobilizes the KLRG1⁺, CD57⁺, and CD28⁻ T-lymphocytes that enter the bloodstream from peripheral tissues (25, 26), whereas the numbers and phenotypes of blood lymphocytes and NKs remain unchanged or are only slightly modified in response to ME (i.e., 50–74% V̇o_{2 max}) (21, 37). This work further demonstrated that normoxic HE mobilizes higher percentages of the replicative senescent (CD57⁺ and CD28⁻)/inhibitory (KLRG1⁺) NKs into the peripheral blood compartment. Another novel finding of this study is that the ME with 12% O₂ also substantially increased percentages of the replicative senescent/inhibitory NKs in blood, whereas percentages of these NK subsets in blood remained unchanged in response to both 21% and 15% O₂ ME. Senescent NKs are functionally compromised, as they have shortened telomeres that can no longer enter the cell division cycle and are associated with the age-related dysfunction of the immune system (16). Consequently, the results in this investigation suggest that the release of replicative senescent/inhibitory NKs induced by 12% O₂ ME may reduce the capacity of blood NKs for proliferation and function. Accordingly, the intervening hypoxic dosage importantly determines the extent of exercise-mediated mobilization of replicative senescent/inhibitory NKs.

The cytotoxic functions of NKs are critical in enabling the immune system to cope efficiently with malignancy (17). Recently, the authors' results obtained from the flow cytometric assay demonstrated that under normoxic conditions, HE increased perforin/granzyme B contents in NKs, with enhanced NK-induced caspase-3 activation and NPC apoptosis, whereas ME did not significantly change the NK count, granule protein content, or lytic activity (36). Notably, the experimental results herein reveal that both 12% and 15% O₂ ME regimens considerably facilitated the cascade of NK effector functions, elevating perforin, granzyme B, and IFN-γ levels, increasing NK-NPC binding, and promoting NK-induced CD95 expression and PS exposure of NPC. The granule proteins, such as perforin and granzyme B, trigger cell-death mechanisms by operating directly or indirectly via the activation of “effector” caspase-3 (28); and the IFN-γ engages the expression of CD95 on target cells, leading to classical caspase 8-dependent apoptosis (24). Therefore, the two hypoxic ME regimens may promote granule- and death receptor-mediated cytotoxic pathways by increasing these granule protein and cytokine levels in blood NK cells. In this study, changes in levels of cytotoxic proteins of the NKs occurred rapidly, in ∼30 min. Additionally, an earlier study found a decrease in perforin mRNA level in blood NKs following strenuous exercise (15). Taken together, these results suggest that the increased cytotoxic protein levels in blood NKs on hypoxic ME are not caused by the de novo synthesis of new protein, but rather involve numerical redistribution in which NKs with high cytotoxic protein expressions are recruited to the bloodstream from the marginating pool in blood vessels, lymph nodes, spleen, and intestines (21). In human NK subsets, CD56^bright NKs are predominantly immunoregulatory cells that generate high levels of cytokines in response to monokine stimulation, whereas CD56^dim NKs are basically cytotoxic cells with an abundance of granule proteins (8). Although this investigation has not further determined the distributions of NK surface-expressed CD56^bright and CD56^dim in blood, both 12% O₂ and 15% O₂ ME regimens can be reasonable to recruit the two phenotypes of NKs to the circulation, because NKs with high granule proteins and cytokine levels are mobilized.

Hypoxic exercise-induced redistribution of NKs may be associated with increased secretions of stress hormones and elevated shear flow during hypoxic exercise (22). In this work, although ME was performed with the same workload during normoxia and hypoxia, the intensity increased (as indicated by elevated HR and blood lactate level) on exposure to hypoxia. Moreover, hypoxic exercise regimens may induce higher plasma catecholamine (epinephrine and norepinephrine) concentrations than did normoxic exercise (13). Since catecholamine induces the selective detachment of NKs from vascular endothelial cells (2), the elevation of catecholamine levels by hypoxic ME may be responsible for the recruitment of NKs from endothelial venules in marginating pools into the bloodstream.

This investigation suggested that NKs that were mobilized by hypoxic ME extravasated the peripheral blood compartment during the 2-h recovery phase. Leukocyte migration from vascular to extravascular sites is a strictly controlled cascade of events, which is initiated by selectin-mediated tethering and rolling interactions of leukocytes on the endothelial surface, followed by integrin-mediated firm adhesion and endothelial transmigration (4). The experimental results in this work indicate that both 15% and 12% O₂ ME increased the percentages of CD62L⁻ and CD11a⁺ NKs in the bloodstream; these NKs, mobilized by hypoxic ME, then rapidly left the peripheral blood compartment during the 2-h recovery phase. These findings suggest that the two hypoxic ME may induce an adhesion cascade between blood NKs and vascular endothelial cells by shedding selectin-CD62L and subsequently expressing β₂-integrin-CD11a on the NK surfaces, further facilitating the extravasation of blood NK cells.

Limitations of the study.

Since NPC occurs typically in the male population of 30–55 yr of age, one limitation of this work is that the subjects were young and healthy. Hence, the clinical relevance and significance of the present findings need to be further assessed in other subpopulations such as middle-aged or elderly subjects, females, and cancer patients. In the methodological limitations, the NK enrichments prepared in this study may include ∼10% non-NK populations, such as CD56 expressing T-cells. Additionally, this investigation had not clarified the distinct effects of normoxic/hypoxic exercise on mobilization of CD56^dim and CD56^bright NK subsets, because of only choosing one-color flow cytometry to analyze the expressions of surface molecules on the total NKs. Recently, Campbell et al. (5) indicated that memory CD8⁺ T lymphocytes that exhibited a high effector and tissue-migrating potential were preferentially mobilized during normoxic high-intensity exercise (85% Watt_max), whereas this exercise also elicited a greater extent of mobilization of CD56^dim NKs than CD56^bright NKs (5). Therefore, the effects of acute hypoxic exercise on mobilization and functions of CD56^dim and CD56^bright NKs need to be further investigated.

In conclusion, systemic hypoxic exposure affects redistribution of the NK subset populations and anti-NPC cytotoxicity of NKs during exercise in a concentration-dependent manner. The 12% O₂ ME mobilizes larger percentages of the replicative senescent/inhibitory NKs into the bloodstream than does the 15% O₂ ME. However, both 12% O₂ and 15% O₂ ME increase cytotoxic protein levels in NKs, capacity of NKs to bind to NPCs, and anti-NPC cytotoxicity of NKs. These experimental findings suggest that 15% O₂ ME is an “effective and modest workload” hypoxic exercise regimen, which provides antitumor cytotoxicity by promoting NK effector functions without mobilizing more percentages of the replicative senescent/inhibitory NKs into the bloodstream.

GRANTS

This study was supported under National Science Council Grants NSC 95-2314-B-182-035-MY3 and 97-2314-B-182-003-MY3.

DISCLOSURES

No conflicts of interest are declared by the authors.