Natural killer cells in obesity: impaired function and increased susceptibility to the effects of cigarette smoke

Abstract

Background: Obese individuals who smoke have a 14 year reduction in life expectancy. Both obesity and smoking are independently associated with increased risk of malignancy. Natural killer cells (NK) are critical mediators of anti-tumour immunity and are compromised in obese patients and smokers. We examined whether NK cell function was differentially affected by cigarette smoke in obese and lean subjects.

Methodology and principal findings: Clinical data and blood were collected from 40 severely obese subjects (BMI>40 kg/m(2)) and 20 lean healthy subjects. NK cell levels and function were assessed using flow cytometry and cytotoxicity assays. The effect of cigarette smoke on NK cell ability to kill K562 tumour cells was assessed in the presence or absence of the adipokines leptin and adiponectin. NK cell levels were significantly decreased in obese subjects compared to lean controls (7.6 vs 16.6%, p = 0.0008). NK function was also significantly compromised in obese patients (30% +/- 13% vs 42% +/-12%, p = 0.04). Cigarette smoke inhibited NK cell ability to kill tumour cell lines (p<0.0001). NK cells from obese subjects were even more susceptible to the inhibitory effects of smoke compared to lean subjects (33% vs 28%, p = 0.01). Cigarette smoke prevented NK cell activation, as well as perforin and interferon-gamma secretion upon tumour challenge. Adiponectin but not leptin partially reversed the effects of smoke on NK cell function in both obese (p = 0.002) and lean controls (p = 0.01).

Conclusions/significance: Obese subjects have impaired NK cell activity that is more susceptible to the detrimental effects of cigarette smoke compared to lean subjects. This may play a role in the increase of cancer and infection seen in this population. Adiponectin is capable of restoring NK cell activity and may have therapeutic potential for immunity in obese subjects and smokers.

Figure 1. Natural killer cell level and function in obesity.

A. Circulating natural killer cells were significantly reduced in obese subjects (n = 40) compared to lean healthy controls (n = 20), *p = 0.0008. B. NK cytotoxic function was also significantly lower in obese subjects (n = 40) compared to lean subject (n = 10). *p = 0.04 with Mann Whitney U test.

Figure 2. The effects of CSE on NK cytotoxic function.

A. NK cell cytotoxicity was significantly enhanced in the presence of NK activatory cytokine IL-2 (n = 40, ***p = 0.0001). CSE significantly reduced NK cell ability to lyse target cells (n = 40, ***p<0.0001 using Students paired t test). CSE equivalent to 20 cigarettes per day was used in all experiments. Error bars represent standard error of the mean. B. CSE acted in a dose dependent manner. CSE equivalent to 40 cigarettes per day was more detrimental to NK cell cytotoxic function than CSE equivalent to 10 or 20 cigarettes per day (n = 5, correlation of number of cigarettes versus NK cytotoxicity: R² = 0.847).

Figure 3. The effects of CSE on NK cytotoxic function in obese and lean subjects.

A. Cigarette smoke reduced NK cell ability to lyse tumour cells in both obese (***p<0.0001; n = 40) and lean subjects (**p = 0.004; n = 10). B. The percentage decrease in NK cell cytotoxicity was significantly greater in obese subjects (33% reduced cytotoxicity) compared to lean controls (25% reduced cytotoxicity, *p = 0.01). C. Adiponectin significantly enhanced NK cell ability to kill target cells in lean controls, compared to when adiponectin was absent (42% vs 47% target cells killed, n = 10, **p = 0.01). Adiponectin did not significantly affect NK cell function in obese subjects (30% vs 24%, n = 40, p = 0.7). Leptin inhibited NK cytoxicity in lean subjects (n = 10, p = *0.04) but in not obese subjects (n = 40). Error bars represent standard error of the mean.

Figure 4. The effects of adipokines on NK cells in the presence of CSE.

A. CSE significantly inhibited NK cell cytotoxic function. Adiponectin, but not leptin could partially restore NK cell function to normal in both obese and lean subjects (adiponectin + smoke vs smoke alone, obese: **p = 0.002, lean: *p = 0.01). Graphs represent the mean of obese (n = 40) and lean subjects (n = 10). B. Representative histograms of NK cell cytoxicity of one obese subject. Tumour cells were stained with green membrane stain CFSE to distinguish target cells from effector cells. After 4 hours, viability stain 7-AAD was added to measure the proportion of target cells lysed. CFSE⁺ tumour cells were gated and the histograms represent the proportion of dead cells (7-AAD⁺). Percentages illustrate the mean fluorescent intensity (MFI) of 7-AAD.

Figure 5. Cigarette smoke inhibits upregulation of NK cell activation markers.

A. Graph representing expression of perforin marker CD107 and activation marker CD69 on NK cells after stimulation with K562 tumour cells for 4 hours (n = 14). Cigarette smoke (equivalent to 20/day) significantly inhibited the upregulation of both CD107 (*p = 0.01) and CD69 (**p = 0.006). Adiponectin in the presence of smoke could significantly restore CD107 expression to normal (*p = 0.03) but not CD69 (p = 0.3). B. Representive flow cytometry dot plots of one subject illustrating the inhibition of CD107 and CD69 upregulation in the presence of smoke. Using flow cytometry, CD45⁺ lymphocytes were gated and within this gate, NK cells (CD56⁺) were plotted against CD107 or CD69. Cells in the upper right hand quadrant represent the proportion of NK cells expressing CD107 or CD69. C. Cigarette smoke did not significantly alter intracellular production of IFN-γ or IL-10 by NK cells after tumour challenge (n = 5, p = 0.09, p = 0.1 respectively), but when adiponectin was added in addition to smoke, production of IFN-γ was significantly increased (**p = 0.006).