Class III obesity is associated with chronic inflammation and a variety of changes in immune function. Yet surprisingly little was known about the status of neutrophils that represent the first line of immune defense. The aim of this study was to assess key functions of neutrophils from class III obese patients, namely phagocytosis, superoxide production, chemotaxis, and response to endotoxin challenge, and compare their responses with lean controls. Thirty obese patients (BMI 48.8 ± 6.6 kg/m2) with comorbidities such as diabetes, hyperlipidemia, high blood pressure, etc. and nine lean (BMI between 20 and 25) subjects were enrolled in the study. Neutrophils from class III obese patients phagocytosed Escherichia coli (E. coli) at similar rates and with adequate numbers of bacteria taken up per cell compared with cells from lean subjects. Neutrophil production of superoxide, which is key to rapid killing of pathogens, showed modest diminution in the class III obese, which increased among patients with BMI >50. Chemotactic activity of neutrophils from class III obese patients was not altered. However, neutrophils from obese subjects showed an increased response to low-dose endotoxin, with concomitant reduced apoptosis and extension of their half-life compared with lean subjects, which suggests possible hyperresponsiveness of these neutrophils. Overall, neutrophil activity was not significantly altered by age, gender, diabetic status, or hyperlipidemia. Collectively, these results suggest that class III obese patients, even with comorbidities, have normal or nearly normal phagocytic, chemotactic, and superoxide generating capacity.
Class III obesity (morbid) is associated with chronic low-grade inflammation which is marked by increased levels of inflammatory factors in sera (1,2). Such inflammation not only plays a role in the development of several comorbidities in obese patients, such as heart disease and diabetes (3), but also is an indicator of immune dysregulation. Indeed, acute phase proteins are elevated in severely obese patients (4) and activated macrophages have been found in the adipose tissue of obese patients (5). In addition, circulating monocytes in class III obese have been shown to be polarized toward a pro-inflammatory phenotype (6). At the cellular level, impaired dendritic cell and CD8+ T cell function has been observed during influenza infection of diet-induced obese mice (7). Similarly, deficient memory T cell responses (8) and bone marrow macrophage impairment in nitric oxide and cytokine production were observed in diet-induced obese mice (9). In humans, reduced natural killer cell activity has been observed in class III obese patients (10).
Such observed changes in immune defense during obesity can impair responses to infection. For example, diet-induced obese mice are more susceptible to influenza infection (11) and have increased susceptibility to bacterial sepsis (12) and periodontal disease (13). In humans, obese subjects have been shown to have impaired vaccine responses, increased susceptibility to surgical site infection, increased risk of infection after trauma, and increased risk of periodontitis (14,15,16,17). Recent studies have shown that obese subjects are more susceptible to H1N1 influenza infection (18). Thus, immune dysregulation appears to be prevalent in obesity.
Neutrophils are the most prevalent immune cell in blood and represent 50–70% of white cells. These cells are key to the initial response to infections utilizing phagocytosis of pathogens and the secretion of reactive oxygen species, proteases, and other antimicrobial compounds to quickly destroy bacteria, protozoa, fungal invaders, etc. Neutrophils have a short half-life, 8–10 h in humans, which can be extended during times of stress or inflammation (19). Given their importance, we know surprisingly little about the impact of obesity on key functions of the neutrophil. Because of the alterations in immune function just discussed, in particular the potential increased susceptibility of obese subjects to bacterial infection, it was important to ascertain if the chronic inflammatory state of class III obesity and general immune dysfunction also disrupted key neutrophil functions. It was also important to determine if comorbidities associated with obesity as well as anthropomorphic attributes such as gender or age contributed to altered neutrophil function. For these reasons, the function of neutrophils from class III obese subjects, with comorbidities such as diabetes and hyperlipidemia, were compared with neutrophils from lean subjects.
Methods and Procedures
Class III obese subjects (N = 30; average BMI 48.8 ± 6.6 kg/m2) between the ages of 29 and 64 years, scheduled for laparoscopic Roux-en-y gastric bypass, were included in the study after obtaining informed consent. The study was done with approval from Michigan State University institutional review board. A group of nine lean donors (BMI between 20 and 25) between the ages of 22 and 50 were enrolled as controls.
Isolation of blood neutrophils
Blood was collected in heparinized tubes (BD, Franklin Lakes, NJ) after an overnight fast. To isolate neutrophils, blood was allowed to sediment in 1% Dextran (Sigma-Aldrich, St Louis, MO) in 0.9% NaCl for 45 min at room temperature. Supernatant was collected, washed, and loaded onto a two-layer (60/70%) Percoll gradient in 0.9% NaCl. After centrifugation at 300g for 6 min, neutrophils were collected from the interface between the 60% and 70% layers. Neutrophils were 98% of all nucleated cells in the preparation as determined by flow cytometry and histological examination (19).
Phagocytic activity of neutrophils was determined flow cytometrically using the pHrodo E. coli Bioparticles Phagocytosis kit (Invitrogen, Carlsbad, CA). The engulfed bacteria fluoresce when in the low pH environment of the acidified phagocytic compartment. Blood (100 µl) was mixed with varying amounts of E. coli bioparticles and incubated for 15 min at 37 °C. A duplicate set of each reaction was incubated at 4 °C as phagocytosis-negative control for associated but not ingested bacteria. Samples without E. coli constituted a second set of negative controls. The percent of cells (as determined by flow cytometry gating on neutrophils via forward and side scatter; Supplementary Figure S1a online) with ingested E. coli was determined flow cytometrically by comparing phagocytosis reactions at 37 °C with reactions at 4 °C (Supplementary Figure S1b online). The relative amount of E. coli ingested by neutrophils was determined by dividing mean fluorescence intensity of phagocytosis reactions with that of cells incubated in the absence of E. coli. Representative images of phagocytosis reactions at 4 °C and 37 °C are shown in Supplementary Figure S1c,d online.
Superoxide production by neutrophils was determined by measurement of the reduction of ferricytochrome c (20) via a microplate assay. Neutrophils (1 × 106 cells/ml) were mixed with cytochrome c (50 µmol/l; Sigma-Aldrich) and various concentrations of phorbol myristate acetate (PMA; Sigma-Aldrich; 0 µmol/l, 0.01 µmol/l, 0.1 µmol/l, 1 µmol/l) in a 96-well plate and incubated in a microplate reader at 37 °C for 30 min, with absorbance at 550 nm measured every 2 min. Some samples were also incubated with superoxide dismutase (400 U/ml) to adjust for non-superoxide–mediated reduction of cytochrome c. Superoxide production was measured as total nmoles produced after 30 min (per 105 cells) as well as the rate of production over 30 min (nmoles/min) during the linear phase of the reaction (see Supplementary Figure S2 online).
The ability of neutrophils to migrate toward a chemotactic agent was monitored using the CytoSelect Cell Migration Assay (Cell Biolabs, San Diego, CA), a microplate assay which measured migration of cells from an upper chamber through a 3-µm pore membrane into a lower chamber containing varying concentrations of recombinant human chemokine IL-8 (CXCL8) (Peprotech, Rocky Hill, NJ; 0.1 nmol/l, 1 nmol/l, or 10 nmol/l). Neutrophils (4.5 × 105) were placed in upper chambers and microplates incubated at 37 °C, 5% CO2 for 1 h. Migration of cells to lower chamber was determined after addition of fluorescent CyQuant GR dye to each well and measurement of resultant fluorescence at 480 nm/520 nm with a fluorescence plate reader, according to the manufacturer's recommendations. Fluorescence was normalized to wells containing vehicle (no IL-8) to account for random migration of cells.
Degree of apoptosis after endotoxin challenge
The half-life of neutrophils, reflected by changes in the degree of apoptosis after endotoxin challenge, was measured by incubating purified neutrophils (2 × 106/ml) for 18 h in the presence or absence of varying concentrations of lipopolysaccharide (LPS) (21). Briefly, cells were incubated for 18 h in Roswell Park Memorial Institute media (RPMI) with 2% dextran-coated charcoal treated fetal bovine serum (FBS; Atlanta Biologicals, Lawrenceville, GA) with varying concentrations of LPS (from Escherichia coli O127:B8; Sigma-Aldrich). Fetal bovine serum was dextran-charcoal treated to remove endogenous glucocorticoids from fetal bovine serum as glucocorticoids are known to prolong neutrophil half-life. After culture, cells were washed once and incubated with 17 µg/ml Merocyanine 540 (MC540) in a 100 µl volume of Hank's Balanced Salt Solution without Mg++ or Ca++ (Invitrogen) + 4% fetal bovine serum, pH 7.2 for 10 min at room temperature, followed by the addition of 900 µl buffer and storage on ice. MC540 binds to the outer membrane of human cells and shows increased fluorescence in the disordered and loosely packed membranes of apoptotic cells, allowing detection of the proportion of apoptotic cells by flow cytometry (22,23,24,25). Cells were analyzed flow cytometrically, with MC540 excitation at 488 nm and emission detected with a 575 ± 26-nm filter. Cultures of neutrophils resulted in two peaks, MC540low and MC540hi, with MC540hi corresponding to apoptotic neutrophils as previously described (see Supplementary Figure S3 online) (25).
Differences between samples were analyzed by unpaired Student's t tests with two degrees of freedom. s.e.m. is shown where indicated. A P value less than 0.05 was considered significant. Correlations were analyzed using Pearson's test, with P < 0.05 being considered significant.
Thirty class III obese subjects were enrolled in the study. Demographic characteristics of subjects are shown in Table 1. There were 22 females and 8 males with average BMI of 48.8 ± 6.6 kg/m2 and mean age of 46.9 ± 11.0 years. The gender and diabetic proportion of the patients enrolled was in accordance with the national average of individuals undergoing gastric bypass surgery (26,27). Five of 8 males and 7 of 22 females enrolled were diabetic. A majority of patients (18 of 30) were hyperlipidemic, including 6 of 8 males and 12 of 20 females. A group of lean donors consisting of 4 males and 5 females, with ages ranging from 22 to 50 years and BMI from 20 to 25 kg/m2, were used as controls. Lean donors were nonsmokers with no known medical issues such as diabetes, heart disease, or chronic inflammation.
Table 1. Physical characteristics of obese patients enrolled in study
Phagocytosis of E. coli by neutrophils from lean vs. obese donors
Recognition, engulfment, and destruction of pathogens by phagocytosis are critical aspects of neutrophil microbicidal activity. To compare the ability of neutrophils from obese and lean patients to carry out phagocytosis, a flow cytometric assay using fluorescently labeled E. coli was employed to determine the percent of neutrophils containing ingested E. coli as well as the relative number of E. coli particles found inside each cell. Four different concentrations of E. coli were used to check the overall defense capacity of these cells to low and high amounts of bacteria.
Figure 1 shows the results of phagocytosis assays for 30 class III obese patients and 9 lean subjects. As can be seen, neutrophils from both class III obese and lean subjects readily phagocytosed E. coli as over 90% of neutrophils contained ingested bacteria when varying amounts of E. coli were used (Figure 1a). Reducing the amount of E. coli in the phagocytosis reactions resulted in a reduced percentage of neutrophils with ingested E. coli; however, no significant differences in the percent of neutrophils containing ingested E. coli were observed between cells from class III obese and lean subjects.
The percentage of neutrophils with ingested bacteria is only one measurement of phagocytosis activity. The amount or number of engulfed E. coli per cell is also important, being reflective of the capacity of each individual neutrophil to take up pathogens. The relative number or amount of E. coli inside cells was estimated by comparing the fluorescence intensity of phagocytosis reactions to the fluorescence intensity of negative controls. As shown in Figure 1b, increasing amounts of E. coli led to increased fluorescence intensity of the cells, indicating more internalized E. coli per cell. Neutrophils from class III obese patients tended to ingest somewhat fewer numbers of E. coli particles at all concentrations than cells from lean donors, albeit being statistically significant at only one dose of E. coli (Figure 1b). Taken together, the data show that these class III obese subjects were able to adequately phagocytose and engulf E. coli.
Because our patient group was diverse, including patients with a number of comorbidities, phagocytosis data were examined with respect to patient age, gender, diabetic status, and hyperlipidemic status to determine if there were any correlations. Patient age, gender, diabetic status, and hyperlipidemia had no significant effect on neutrophil phagocytosis activity (Figures 2 and 3).
Superoxide production by neutrophils from lean vs. obese donors
Microbicidal reactive oxygen species such as superoxide are rapidly released from activated neutrophils and are critical to the rapid killing of pathogens inside and outside the neutrophil. To assess whether severe obesity affected neutrophil reactive oxygen species generation, neutrophil superoxide production from class III obese patients was examined and compared with neutrophils from lean donors. A range of concentrations of PMA (0.01 µmol/l–1 µmol/l) was used to fully assess the functional capacity of neutrophils. Unstimulated cells served as controls for background activity. In these assays, isolated neutrophils were used which had a purity of at least 98% of nucleated cells. As seen in Figure 4a, superoxide production by neutrophils from obese patients at both 0.1 µmol/l and 1 µmol/l PMA stimulation was reduced a significant degree when compared with neutrophils from lean subjects. Lower but adequate amounts of superoxide were also generated after stimulation with 0.01 µmol/l PMA.
The killing of pathogens in a timely manner is also key to defense against infection; thus, the rate of superoxide production (nmoles/min) by neutrophils was also measured. Neutrophils from obese patients showed somewhat lower rates of superoxide production at 0.1 µmol/l and 1 µmol/l PMA compared with cells from lean donors. These differences were not statistically significant. However, a statistically significant (27.9%) reduction in rate of superoxide production by neutrophils from obese patients was observed at 0.01 µmol/l PMA (Figure 4b). Collectively, these results suggest that superoxide production by stimulated neutrophils, both total nmoles produced and the rate of production, are moderately impaired in class III obese patients.
Of note, baseline superoxide production (total nmoles) by unstimulated neutrophils from obese patients was 59.6% greater than cells from lean subjects (P < 0.05, Figure 4a). The rate of superoxide production by unstimulated neutrophils from obese patients, however, was not significantly different from cells from lean subjects. These results suggest that neutrophils from class III obese patients may be partially activated, in keeping with previous studies showing neutrophil activation in obese patients (28) and our current data showing elevated responses to LPS.
Correlation of superoxide production with weight parameters
Neutrophils from patients with BMI >50 made significantly less superoxide at 0.1 and 1 µmol/l PMA stimulation than class III obese patients with BMI <50 (Figure 4c). Importantly, superoxide production by neutrophils from obese patients with BMI <50 was not significantly altered compared with neutrophils from lean subjects. In addition, superoxide production by neutrophils from obese patients, both total nmoles and rate of production (nmoles/min), after stimulation with the highest dose of PMA (1 µmol/l) showed a negative correlation with BMI (total nmoles, r = −0.39083; rate, r = −0.39856, P < 0.05), indicating that superoxide production was further reduced in class III obese patients with higher BMI (Supplementary Figure S4 online). These results indicated that in obese subjects, neutrophil superoxide production may become more impaired at higher BMI, suggesting that the degree of obesity affects neutrophil superoxide production. Of note, the effects of obesity on neutrophil superoxide production were not significantly altered when patient age, gender, diabetic status, and hyperlipidemic status was considered (Figure 5a–d).
Chemotaxis by neutrophils from lean vs. obese donors
Neutrophils must migrate to sites of infection, inflammation, or tissue injury in order to carry out their functions. This process, known as chemotaxis, is an especially critical process for neutrophil antimicrobial function as well as wound healing. In order to test their responsiveness to chemokines, isolated neutrophils from obese and lean patients were analyzed for their ability to migrate toward the chemokine IL-8 in a transwell assay. IL-8 was used in this study because it is a potent neutrophil chemoattractant often associated with inflammation. Migration was measured relative to random migration in wells with no added IL-8. Figure 6a shows a representative image of cells in lower chambers of transwell assays after 1 h incubation. As can be seen, there was significant migration of neutrophils to lower chambers when 10 nmol/l IL-8 was added compared with random migration of cells in the absence of IL-8. Using this assay, the relative migration of neutrophils from obese and lean patients toward varying concentrations of IL-8 was determined. Figure 6b shows that the migration of neutrophils followed a dose-response curve with respect to IL-8. No statistically significant differences were observed between neutrophils from obese and lean donors to a range of concentrations of IL-8. There also was no correlation between degree of chemotaxis and BMI. These results indicate that neutrophils from class III obese subjects have unimpaired chemotactic activity.
Endotoxin response of neutrophils from lean vs. obese donors: half-life extension
Neutrophils have a short half-life of 8–10 h in humans. As obesity has been shown to produce a chronic, low-grade inflammatory state that might alter the life span of circulating neutrophils, the half-life or degree of apoptosis of a resting population of neutrophils from lean and obese subjects was compared. In addition, the half-life of neutrophils from obese and lean subjects was also assessed after cells were treated with a range of doses of endotoxin (LPS). LPS is a component of gram negative bacterial membranes and is a potent activator of neutrophils, resulting in dramatically extended cellular half-life. Isolated neutrophils were used in these experiments and the degree of apoptosis after 18 h in culture was used as a measure of neutrophil half-life.
When neutrophils from obese subjects were cultured in the absence of endotoxin, only a slight, nonsignificant decrease in apoptosis was observed when compared with neutrophils from lean controls (65.3 vs. 69.6%, respectively; Figure 7a), indicating that resting neutrophils from class III obese patients do not have an altered half-life.
Neutrophils are highly responsive to outside stimuli which may cause activation and extension of half-life (19). When neutrophils were challenged with endotoxin, a significant extension of cell survival was noted at 10 ng/ml and 1 ng/ml LPS in neutrophils from both lean and obese subjects (Figure 7a). At the lowest dose of LPS tested (0.1 ng/ml), a dramatic decrease in apoptosis was observed in neutrophils from obese subjects, while neutrophils from lean subjects were not significantly affected. While decreased apoptosis of neutrophils from obese patients was also observed at 1 ng/ml LPS treatment (18.7 vs. 25.1% apoptosis, respectively; P < 0.05), a difference was no longer evident at 10 ng/ml LPS stimulation as neutrophils from both lean and obese subjects underwent similar amounts of apoptosis (Figure 7a). It is known that endotoxin activates and prolongs the half-life of neutrophils (21), but these results indicate that class III obesity causes an enhanced sensitivity of neutrophils to low-dose LPS, suggesting that obesity has altered neutrophil responsiveness to LPS.
Effect of weight parameters on neutrophil survival at low-dose LPS
The effects of weight-related factors in class III obese patients on neutrophil half-life at low-dose LPS were also investigated. As with superoxide production, BMI significantly affected neutrophil half-life in obese patients. The degree of apoptosis of neutrophils from patients with BMI >50 kg/m2 was markedly reduced to 33.9% after culture with 0.1 ng/ml LPS (Figure 7b). This is an interesting and significant extension of neutrophil half-life. Neutrophils from obese patients with BMI between 37.5 and 50 kg/m2 had higher apoptosis rates than patients with a BMI >50 kg/m2, but still remained more protected from apoptosis than neutrophils from lean subjects (Figure 7b). In sum, neutrophils from both groups of class III obese patients (BMI <50, BMI >50) had significantly reduced neutrophil apoptosis, or longer survival, at low-dose LPS compared with cells from lean subjects (Figure 7b). These results indicate that class III obesity results in dysregulation of the neutrophil response to low-dose LPS as measured by increased survival via significant reduction in apoptotic rate. In addition, this dysregulation becomes more severe as BMI of obese patients increases.
Effect of obesity on blood neutrophil numbers
Obesity has been associated with increased white blood cell counts in humans, including increased neutrophil numbers (29,30,31). In this study, slightly increased white cell and neutrophil numbers in blood of class III obese patients were observed when compared with lean subjects (WBC: 7,370 ± 271 vs. 6,400 ± 523 cells/µl blood; neutrophils: 4,528 ± 268 vs. 3,662 ± 427 cells/µl blood, respectively), though these data were not statistically significant. Cell numbers were within normal ranges for all patients.
This study details the effects of class III obesity on key neutrophil functions. The most notable finding in this work was that neutrophils from class III obese patients were relatively unimpaired when it came to major functional attributes. Neutrophil chemotaxis was not significantly altered in class III obese vs. lean subjects, and only slight defects were observed in phagocytosis activity. Superoxide production after stimulation with PMA showed some deficits among class III obese subjects; however, even these changes were modest and were not significantly different at all doses of PMA tested. However, some significant defects in superoxide production were observed in patients with a BMI higher than 50. Despite this, the defect observed in patients with BMI >50 was modest and might not represent an immune defense deficit, given the many compensatory mechanisms utilized by the immune system. Surprisingly, diabetic status or hyperlipidemia of class III obese patients did not affect neutrophil superoxide production, phagocytosis of E. coli, or chemotaxis. In addition, patient age and gender were not factors in the observed changes in neutrophil function.
Neutrophil functional deficits in obesity have been observed in previous studies, including impaired neutrophil bactericidal activity in class III obese individuals, though this earlier study used a small number of patients (32). In other studies, increased adherence and altered expression of surface molecules such as adhesion molecule CD62L and the death receptor CD95 (Fas ligand) (32,33,34) have been observed in neutrophils from class III obese subjects, suggesting that neutrophils are altered by the inflammatory state of obesity. Our study only partially agrees with these previous works, as some neutrophil deficits were observed but on the whole it appears that key neutrophil functions are largely intact in class III obese subjects.
An important but understudied aspect of neutrophil biology is neutrophil half-life or survival. Normally on the order of 8–10 h in humans, neutrophil half-life can be extended by stress and/or exposure of cells to inflammatory factors including LPS stimulation (19). It was hypothesized that in the highly altered environment of obesity, and in the dramatic changes created by metabolic syndrome, neutrophil half-life might be altered by interaction with various inflammatory factors (19). There was no difference between half-life of resting, unstimulated neutrophils from obese and lean subjects, however endotoxin challenge (0.1 ng/ml LPS) of neutrophils from class III obese patients significantly extended neutrophil half-life compared with that observed with neutrophils from lean subjects, suggesting heightened sensitivity of neutrophils from class III obese patients to LPS (21). This heightened LPS sensitivity was most significant at the lowest dose tested, but remained unchanged at the highest doses of LPS. The reasons behind the observed increase in neutrophil sensitivity at low doses of LPS may involve molecules involved in LPS signaling, such as CD14 or lipopolysaccharide binding protein (LBP) (35,36). Further investigation into this possibility is ongoing. Moreover, it is not known whether heightened neutrophil sensitivity to low-dose LPS is a result of the chronic inflammation associated with obesity, or a contributor to it. It has been reported that elevated endotoxemia may play a role in obesity and metabolic syndrome in mice (37) and that high-fat diet induces low-grade endotoxemia in humans (38). Further research is warranted to determine the significance of altered neutrophil sensitivity to LPS.
In summary, class III obesity does not appear to significantly alter neutrophil functional capacity. Some deficits in function were noted, however, but these changes were modest, with the most significant impairment observed in patients with the highest BMI (BMI >50) who exhibited a reduced ability to generate superoxide. Neutrophil apoptosis and responsiveness to LPS stimulation was also largely unaltered, except at the lowest dose of LPS tested. Because of numerous compensatory mechanisms within the immune system, it seems unlikely that the modest changes in neutrophil function observed in this study would have a significant impact on immune function in class III obese individuals.