Alcohol abuse and smoking alter inflammatory mediator production by pulmonary and systemic immune cells

Abstract

Alcohol use disorders (AUDs) and tobacco smoking are associated with an increased predisposition for community-acquired pneumonia and the acute respiratory distress syndrome. Mechanisms are incompletely established but may include alterations in response to pathogens by immune cells, including alveolar macrophages (AMs) and peripheral blood mononuclear cells (PBMCs). We sought to determine the relationship of AUDs and smoking to expression of IFNγ, IL-1β, IL-6, and TNFα by AMs and PBMCs from human subjects after stimulation with lipopolysaccharide (LPS) or lipoteichoic acid (LTA). AMs and PBMCs from healthy subjects with AUDs and controls, matched on smoking, were cultured with LPS (1 μg/ml) or LTA (5 μg/ml) in the presence and absence of the antioxidant precursor N-acetylcysteine (10 mM). Cytokines were measured in cell culture supernatants. Expression of IFNγ, IL-1β, IL-6, and TNFα in AMs and PBMCs was significantly increased in response to stimulation with LPS and LTA. AUDs were associated with augmented production of proinflammatory cytokines, particularly IFNγ and IL-1β, by AMs and PBMCs in response to LPS. Smoking diminished the impact of AUDs on AM cytokine expression. Expression of basal AM and PBMC Toll-like receptors-2 and -4 was not clearly related to differences in cytokine expression; however, addition of N-acetylcysteine with LPS or LTA led to diminished AM and PBMC cytokine secretion, especially among current smokers. Our findings suggest that AM and PBMC immune cell responses to LPS and LTA are influenced by AUDs and smoking through mechanisms that may include alterations in cellular oxidative stress.

community-acquired pneumonia (CAP) is the leading cause of infection-related death in Western countries and a major reason for intensive care unit admissions internationally (11). Severe CAP in the intensive care unit is linked to increased mortality (23) and is an accepted risk factor for development of the acute respiratory distress syndrome (ARDS) (45). Environmental factors, including the alcohol use disorders (AUDs) alcohol abuse and alcohol dependence, and tobacco smoking, are commonly observed among hospitalized patients and frequently coexist (24, 29). AUDs and smoking have been linked to an increased risk for development of CAP (13, 42, 44) and ARDS (21, 33). Alterations in immune defense are likely in part responsible for these associations. Nevertheless, there is a paucity of human data detailing effects of in vivo alcohol and smoke exposure on the response of immune cells to pathogens; therefore, no specific interventions to mitigate risk for CAP and ARDS in populations with AUDs and smoking have been developed.

Alveolar macrophages (AMs) are responsible for initiation of the pulmonary immune response against invading pathogens via recognition of pathogen-associated molecular patterns (PAMPs), including lipopolysaccharide (LPS) on gram-negative pathogens and lipoteichoic acid (LTA) on gram-positive pathogens. Their actions are mediated in part through Toll-like receptors (TLRs)-2 and -4 on the AM surface. AM activation after PAMP recognition results in elaboration of inflammatory mediators, including proinflammatory cytokines and chemokines, alerting additional components of the immune system to the presence of pathogen invaders and, thereby, promoting host immunity (1, 40, 44).

Although overexuberant pulmonary and systemic inflammation have been linked to worsened disease severity and poorer outcomes in CAP (34) and ARDS (4), the impact of environmental alcohol and cigarette exposure on inflammation in these compartments remains controversial. For example, studies support an association between chronic alcohol exposure and diminished AM activity (6, 7, 41) and, also, tissue macrophage activation (19, 28). Cigarette smoke exposure has been alternatively associated with an anti-inflammatory AM phenotype (39, 50) but, also, with increased inflammatory cytokine expression (10, 17). Moreover, investigations examining the concomitant impact of AUDs and smoking on immune cells, in the lung or systemically, are absent from the literature.

Understanding the alterations in AM and systemic inflammatory responses that are related to AUDs and cigarette smoking could provide insight into the increased predisposition for severe CAP and ARDS among individuals who consume unhealthy amounts of alcohol, many of whom smoke. Alterations in oxidative stress have been demonstrated in AUDs (47) and cigarette smoking (14) and may hamper normal AM activity (18, 31). An understanding of the modulation of oxidative stress in immune cells in vitro with an exogenous antioxidant could provide insight into its efficacy as a therapy in vivo. Therefore, we sought to determine the effects of LPS and LTA exposure on the elaboration of interferon (IFN)-γ, interleukin (IL)-1β, IL-6, and tumor necrosis factor (TNF)-α by freshly isolated AMs from bronchoalveolar lavage (BAL), as well as peripheral blood mononuclear cells (PBMCs), from individuals with AUDs, both smokers and nonsmokers, and healthy controls. We hypothesized that proinflammatory cytokine production by AMs and PBMCs would be more exuberant in subjects with AUDs but that active smoking would dampen cytokine elaboration. We further sought to determine the influence of the antioxidant precursor N-acetylcysteine (NAC) on these responses and differences in TLR expression by AMs and PBMCs that may be associated with AUDs and smoking.

METHODS

Subject screening, recruitment, and enrollment.

Subjects with AUDs were recruited between August 2011 and October 2015 at the Denver Comprehensive Addictions Rehabilitation and Evaluation Services Center, an inpatient detoxification facility affiliated with Denver Health and Hospital Administration (Denver, CO). Control subjects without AUDs were recruited via approved flyers and electronic mail advertisements in the Denver metropolitan area. Investigators submitted the protocols for the study to the Colorado Multiple Institutional Review Board, which approved this study, and all subjects provided written informed consent prior to participation.

Subjects with AUDs were eligible to participate if they met all the following criteria at study entry: 1) an Alcohol Use Disorders Identification Test (AUDIT) score of ≥8 for men or ≥5 for women, 2) alcohol use within the 7 days before enrollment, and 3) ≥21 yr of age. The AUDIT questionnaire is a standardized survey to detect current and previous alcohol abuse that has been validated in a variety of clinical settings (36). AUDIT scores of <8 for men or <5 for women were required for control subjects. Screening of potential controls focused on balancing these individuals with AUD subjects in terms of age and smoking history. A subject was considered to be a smoker if he/she acknowledged current active use of cigarettes, without regard for duration of cigarette use. The ultimate sample sizes of subjects and controls were chosen on the basis of feasibility to recruit comparable numbers of AUD subjects and controls who could be matched on the basis of smoking and on prior investigations by our group.

In an effort to minimize potential confounding related to comorbidity, AUD subjects and controls were ineligible to participate in the study if they met any of the following criteria: 1) prior medical history of liver disease (documented history of cirrhosis, total bilirubin ≥2.0 mg/dl, or albumin <3.0), 2) prior medical history of gastrointestinal bleeding (due to concern for varices), 3) prior medical history of heart disease (history of left ventricular ejection fraction <50%, myocardial infarction, or severe valvular dysfunction), 4) prior medical history of renal disease (end-stage renal disease requiring dialysis or serum creatinine ≥2 mg/dl), 5) prior medical history of lung disease, defined as an abnormal chest radiograph or spirometry (volume in 1st s of exhalation or forced vital capacity <75%), 6) concurrent illicit drug use, defined as a toxicology screen positive for cocaine, opiates, or methamphetamines, 7) prior history of diabetes mellitus, 8) prior history of HIV infection, 9) failure of the patient to provide informed consent, 10) pregnancy, and 11) abnormal nutritional status, defined by the calculated nutritional risk index (NRI) with the subject's albumin, current weight, and usual weight values with the following equation: NRI = 1.519 (albumin in g/l) + (current weight/usual weight) * 100 + 0.417 (2). Patients were considered to have a normal nutritional status if the NRI was ≥90. Potential subjects >55 yr of age were also excluded to minimize concomitant, but asymptomatic, comorbidities.

Clinical protocol.

Eligible subjects with AUDs and controls, group-matched on the basis of current cigarette smoking, were admitted to the University of Colorado Hospital's Clinical and Translational Research Center for bronchoscopy and blood sampling. All bronchoscopy procedures were performed utilizing telemetry monitoring and standard conscious sedation protocols as previously described (22). The bronchoscope was wedged into a subsegment of the right middle lobe or the lingula. Three to four 50-ml aliquots of sterile, room temperature 0.9% saline were sequentially instilled and recovered with gentle aspiration. The first aspirated aliquot was not utilized in experiments for this investigation. All subsequent aliquots were combined and used in experiments as representative of the distal air spaces. BAL samples were transported to the laboratory in sterile 50-ml conical tubes. Whole blood samples were also collected. Informed consent for blood sampling only, without bronchoscopy, was obtained from a subset of AUD subjects and controls.

Laboratory processing.

BAL fluid was immediately centrifuged (900 g, 10 min) after collection to separate cellular and acellular components; cells were then resuspended in RPMI medium (Corning, Corning, NY) with additives as indicated below. PBMCs were isolated from 10 ml of whole blood using a Vacutainer CPT cell preparation tube (Becton Dickinson, Franklin Lakes, NJ), which promotes purification of peripheral white blood cells (WBCs) that contain, on average, 98% mononuclear cells with 2% neutrophil contamination after centrifugation. In commercial tests, within the mononuclear cell fractions isolated, 86% of recovered cells are lymphocytes, while 14% are circulating monocytes, similar to percentages that would be expected in peripheral blood after elimination of neutrophils (3). CPT tubes were centrifuged immediately at 1,800 g for 30 min, supernatant was decanted, and the cells were washed twice in 1× PBS (450 g, 5 min) and finally resuspended in RPMI medium (Corning) with additives as indicated below. Cellular viability was determined via Trypan blue exclusion; cells from all subjects were estimated to exceed 95% viability prior to culture. Cell counts were performed with BAL and PBMCs to determine total cell number. Cytospins of BAL cells were examined after staining to determine differential cell types from ≥200 cells by an observer blinded to the subject history.

After BAL centrifugation, 5 × 10⁵ BAL cells/well were plated in 2 ml of RPMI medium with antibiotics (Corning) in 12-well plates. Cells were cultured for 1 h (37°C, 10% CO₂); then medium was removed and immediately replaced with each of the following conditions per well: RPMI medium only, RPMI medium and 1 μg/ml LPS (from Escherichia coli: 055:B5, Sigma Aldrich, St. Louis, MO), or RPMI medium and 5 μg/ml LTA (from Staphylococcus aureus, InvivoGen, San Diego, CA). Serum was not added to BAL culture medium in an effort to more closely replicate serum-free conditions in lung. For each of these three conditions, in separate wells, supplemental NAC (10 mM; Sigma Aldrich) was also added. Doses of LPS, LTA, and NAC utilized in AM experiments were chosen on the basis of preliminary experiments in our laboratory demonstrating a measurable impact of these compounds on AM cytokine production in vitro. Additionally, doses were comparable to those used by our group and others in prior investigations (6, 15). All wells were cultured for 18 h (37°C, 10% CO₂); then cell culture supernatant and cells were collected separately and stored at −80°C. In a subset of subjects, AMs were cultured for 18 h (37°C, 10% CO₂) with and without reagents, and cytotoxicity was measured using the CytoTox-ONE kit (Promega, Madison, WI) according to the manufacturer's protocol to calculate percent lactate dehydrogenase release as a surrogate for cytotoxicity.

After PBMC isolation, 5 × 10⁵ cells were plated in 2 ml of RPMI medium containing 10% fetal bovine serum (Atlanta Biologicals, Atlanta, GA) and antibiotics in 12-well plates. In addition to RPMI medium, PBMCs were exposed to 1 μg/ml LPS (E. coli: 055:B5, Sigma Aldrich) or 5 μg/ml LTA (S. aureus, InvivoGen), with and without supplemental NAC (10 mM, Sigma Aldrich). Doses of LPS, LTA, and NAC were again chosen after preliminary experiments demonstrated an ability of the compounds at these doses to measurably influence cytokine production in culture by PBMCs. PBMCs were cultured for 18 h (37°C, 5% CO₂), and supernatants and cells were collected separately and stored at −80°C for analyses.

IFNγ, IL-1β, IL-6, and TNFα were assayed in culture supernatants [proinflammatory panel 1 (4-plex) for tissue culture, MesoScale Discovery, Rockville, MD] according to the manufacturer's directions. For all condition types, experimental replicates were performed.

In a subset of subjects and controls, uncultured AMs and PBMCs were lysed using buffer RLT from the RNeasy Mini Kit (Qiagen, Valencia, CA). Additionally, cells collected after 18 h in culture with medium only, in medium with LPS, or in medium with LTA were lysed in a similar way. Lysed cells were homogenized using QIAshredder columns (Qiagen) and stored at −80°C. RNA was extracted using the RNeasy Mini Kit according to the manufacturer's instructions utilizing the on-column DNase digestion with the RNase-free DNase set (Qiagen). After elution, RNA quantity and quality were determined using a NanoDrop spectrophotometer (Thermo Scientific, Wilmington, DE). For all available samples, 1 μg of total RNA was reverse-transcribed utilizing a high-capacity cDNA reverse transcription kit (Applied Biosystems, Foster City, CA). The manufacturer's protocol was followed, and the cDNA was quantified using a NanoDrop spectrophotometer. For the quantitative real time-PCR, 100 ng of cDNA were brought up to 9 μl with nuclease-free water and combined with 1 μl of TaqMan Gene Expression Assay and 10 μl of TaqMan Gene Expression Master Mix (Applied Biosystems). Assays (including a no-template control) were performed in triplicate under standard real-time PCR conditions (50°C for 2 min, 95°C for 10 min, and 40 cycles of 95°C for 15 s followed by 60°C for 1 min) using a real-time PCR system (model 7300, Applied Biosystems) with sequence detection software. Inventoried TaqMan Gene Expression Assays (Applied Biosystems) were used to measure mRNA expression levels for TATA box-binding protein (assay ID Hs00427620_m1), TLR2 (assay ID Hs00152932_m1), and TLR4 (assay ID Hs00152939_m1). TATA box-binding protein was used as the endogenous control.

Statistics.

Descriptive statistics related to patient characteristics were calculated and summarized. Primary outcome variables for these investigations included IFNγ, IL-1β, IL-6, and TNFα measured in AM and PBMC cell culture supernatants. To investigate mean differences in supernatant cytokines elaborated by AMs or PBMCs among the four groups (i.e., control nonsmokers, control smokers, AUD nonsmokers, and AUD smokers) in response to culture in 1) medium only, 2) medium with LPS, or 3) medium with LTA, one-way ANOVA was conducted for each of the four outcome variables based on logarithmically transformed data, with post hoc testing performed between groups when P ≤ 0.05 for these comparisons. Nonparametric paired statistical (Wilcoxon's signed-rank) tests were also performed within subject groups to compare changes across conditions (e.g., medium only vs. medium with LPS).

To further investigate alterations in cytokine secretion by AMs in response to medium only and in the presence of LPS or LTA and differences among the groups based on AUD and smoking history, linear mixed modeling was also performed (16, 26), where the two repeated measures (i.e., medium and LPS, or medium and LTA), the group variables AUD (yes or no), current smoking (yes or no), and all interactions were included as fixed effects, with unstructured covariance assumed within subjects. For statistical modeling, outcome variables were logarithmically transformed to achieve normality. Moreover, to examine the effect of NAC on outcome variables, results from linear mixed modeling with three repeated measures from AM culture supernatants, namely, 1) medium only, 2) medium with LPS or LTA, and 3) medium with LPS + NAC or LTA + NAC, along with the group variables AUD (yes or no), current smoking (yes or no), and their interactions was included as fixed effects for each of the AM outcomes.

For RT-PCR data, relative gene expression of TLR2 and TLR4 by AMs and PBMCs was examined across the four groups, with control nonsmokers as the referent group. Additionally, relative gene expression alterations in AMs and PBMCs with LPS or LTA exposure were further examined in each group individually and across groups. The comparative cycle threshold (2^−ΔΔC_T) method was utilized in the calculation of gene expression (38, 38). ANOVA was used to compare mean values across the four groups, while paired t-tests were used to compare relative expression changes within groups.

All statistical analyses were performed using SAS 9.3 (SAS Institute, Cary, NC). A significance level of 0.05 was assumed throughout the study.

RESULTS

Subjects and controls were enrolled in investigations between 2011 and 2015. All subjects enrolled and consented for BAL completed sampling procedures without incident (Table 1). Ages of individuals across groups were similar, and smoking habits did not differ among smokers in the BAL group. Spirometry was comparable between groups, as were baseline laboratory values. There tended to be more men among AUD subjects who smoked. Specific racial differences were noted between the groups. There were more control Caucasian subjects and more Native American AUD subjects. More AUD subjects identified their ethnicity as Hispanic.

Table 1.

Clinical features of subjects and controls with BAL

	Group
	Control nonsmokers (n = 19)	Control smokers (n = 20)	AUD nonsmokers (n = 19)	AUD smokers (n = 31)	P Value
Age, yr	41.5 ± 8.0	40.7 ± 6.6	43.2 ± 6.0	44.8 ± 7.7	0.21
Sex, %men	73.7	60.0	78.9	87.1	0.17
Race, %
Caucasian	89.5	60.0	47.4	74.2	0.02
African-American	15.8	20.0	0	16.1	0.21
Native American	0	0	36.8	9.7	0.0007
Other/mixed race	0	20	15.8	0	0.001
Ethnicity, %Hispanic	15.8	15.0	42.1	22.6	0.20
AUDIT score	2.9 ± 2.1	2.4 ± 2.1	30.0 ± 7.0	29.4 ± 7.5	<0.0001
Age began consuming alcohol, yr	17.8 ± 2.9	17.9 ± 2.7	16.2 ± 3.6	16.2 ± 3.6	0.19
Current smokers, %	0	100	0	100	1.0
Pack-yr smoking	N/A	15.2 ± 13.5	N/A	19.8 ± 16.2	0.30
Spirometry
FEV1, %predicted	102 ± 14	92 ± 10	101 ± 11	96 ± 13	0.05
FVC, %predicted	102 ± 14	94 ± 9	102 ± 12	101 ± 13	0.10
FEV1/FVC	80 ± 6	80 ± 5	79 ± 7	76 ± 8	0.12
BMI, kg/m2	29.1 ± 5.1	27.9 ± 5.6	27.6 ± 5.6	24.9 ± 3.1	0.02
BAL (entire sample)
Total no. of WBCs, ×106	15.3 ± 5.5	32.0 ± 19.9	14.3 ± 6.2	26.3 ± 14.1	<0.0001*
%Alveolar macrophages	85 ± 8	91 ± 3	77 ± 14	89 ± 10	<0.0001†
No. of lymphocytes, ×106	13.1 ± 4.8	29.2 ± 18.5	11.1 ± 5.7	24.2 ± 13.2	<0.0001*
%Lymphocytes	13 ± 7	7 ± 3	21 ± 14	8 ± 5	<0.0001‡
No. of neutrophils, ×106	2.1 ± 1.4	2.4 ± 2.0	3.0 ± 2.4	1.9 ± 1.4	0.25
%Neutrophils	1 ± 2	1 ± 2	2 ± 1	3 ± 6	0.51
%Other	0 ± 1	0 ± 0	0 ± 0	0 ± 0	0.65
Peripheral WBC count, ×103	5.7 ± 1.1	7.4 ± 1.8	6.5 ± 1.6	6.6 ± 1.6	0.02
MCV, fl	87 ± 5	89 ± 6.2	90 ± 5	92 ± 8.1	0.11
Serum
Total bilirubin, mg/dl	0.8 ± 0.3	0.7 ± 0.3	1.0 ± 0.5	0.8 ± 0.3	0.04
Albumin, mg/dl	3.9 ± 0.4	3.8 ± 0.2	4.1 ± 0.3	3.9 ± 0.3	0.02
Creatinine, mg/dl	1.0 ± 0.2	0.9 ± 0.2	0.8 ± 0.1	0.8 ± 0.2	0.004

Values are means ± SD or percentages; n, number of subjects. Across-groups comparisons were performed with 1-way ANOVA for continuous variables (post hoc testing with Tukey's honestly significant difference test where applicable) and Fisher's exact test for categorical variables.

AUDIT, alcohol use disorders (AUDs) identification test; BAL, bronchoalveolar lavage; BMI, body mass index; FEV₁, forced expiratory volume in 1st s of exhalation; FVC, forced vital capacity; MCV, mean corpuscular volume; N/A, not applicable; WBC, white blood cell.

^*P ≤ 0.03, smokers (AUD and control) vs. nonsmokers (AUD and control).

^†P = 0.0004, AUD smoker vs. AUD nonsmoker.

^‡P ≤ 0.02, AUD nonsmoker vs. other groups.

Alveolar macrophage culture experiments.

Viability of freshly collected BAL cells was 95 ± 3%. The numbers of BAL WBCs and relative percentages of AMs and lymphocytes significantly varied across the four subgroups (P < 0.0001). Current cigarette smoking was associated with higher numbers of BAL WBCs and higher absolute numbers of AMs (Table 1) in subjects with AUDs or in controls. AUD nonsmokers had a higher percentage of lymphocytes, although absolute numbers of lymphocytes were not associated with AUDs or smoking (P = 0.47).

After 18 h of culture in medium only, secretion of IFNγ, IL-1β, IL-6, and TNFα by AMs was relatively low and did not differ by AUD or smoking history; addition of 2.5–10 mM NAC similarly did not alter basal cytokine expression. However, addition of LPS to AMs in culture was associated with increased secretion of all four cytokines by all four subject types (P < 0.0001; Fig. 1). IFNγ secreted by LPS-stimulated AMs was significantly different across the four subject types (P = 0.03; Fig. 1A). Post hoc analyses indicated that IFNγ secreted by LPS-stimulated AMs from AUD nonsmokers was significantly elevated compared with the other three subject groups (P ≤ 0.03). In separate mixed-modeling analysis of these data, IFNγ production by LPS-stimulated AMs was positively associated with AUDs (P = 0.03), while smoking had a negative impact on IFNγ secretion (P = 0.03). In this analysis, no significant interactions between AUDs and smoking were observed on IFNγ expression (P = 0.36). Addition of 10 mM NAC with LPS was associated with less IFNγ expression by AMs for all subject groups (P < 0.04), except AUD nonsmokers. IL-1β secreted by LPS-stimulated AMs also significantly varied between the four subject groups (P = 0.03; Fig. 1B). Post hoc analyses of the data revealed that LPS-stimulated AMs from AUD nonsmokers expressed quantitatively more IL-1β than AMs from either control group (P ≤ 0.03). Mixed-modeling analysis indicated a significant positive effect of AUDs on IL-1β expression within these groups (P = 0.04) but no effect of smoking (P = 0.18). Similar to observations with IFNγ, NAC added with LPS was associated with less IL-1β secretion (P < 0.001 for each group except AUD nonsmokers). Post-LPS AM IL-6 production tended to vary across groups (P = 0.054, Fig. 1C), with IL-6 secreted by AMs from AUD nonsmokers exceeding that in AUD smokers (P ≤ 0.03). AMs cultured in NAC and LPS was associated with significantly less IL-6 secretion in all groups (P < 0.001) than did AMs cultured in LPS. The quantity of TNFα secreted by LPS-stimulated AMs did not vary across subject groups (P = 0.23; Fig. 1D). NAC added with LPS was associated with less TNFα secretion (P < 0.001), except in AUD nonsmokers. Additional mixed-modeling analyses indicated that current smoking was associated with more robust decreases in cytokine response to LPS with concomitant NAC (P ≤ 0.01).

Fig. 1.

Freshly obtained alveolar macrophages (AMs) from control subjects and subjects with alcohol use disorders (AUDs), both smokers and nonsmokers, were cultured for 18 h in medium only, in medium with 1 μg/ml lipopolysaccharide (LPS), or in medium with 1 μg/ml LPS and 10 mM N-acetylcysteine (NAC). Interferon-γ (IFNγ), interleukin (IL)-1β, IL-6, and tumor necrosis factor (TNF)-α were measured in supernatants. Addition of LPS was associated with increased production of all cytokines by all 4 subgroups (P < 0.0001; media + LPS). Addition of NAC with LPS was generally associated with less AM cytokine expression (media + LPS + NAC). A: IFNγ secreted by LPS-stimulated AMs significantly differed across the 4 groups; IFNγ values were more elevated among AUD nonsmokers than control smokers and nonsmokers, as well as AUD smokers. Addition of NAC was associated with less IFNγ expression for all groups, except AUD nonsmokers. B: IL-1β secreted by LPS-stimulated AMs significantly differed between groups and between AUD nonsmokers and control smokers and nonsmokers. Addition of NAC was associated with less IL-1β expression for all groups except AUD nonsmokers. C: IL-6 secretion by LPS-stimulated AMs tended to differ across groups (P = 0.054) and was significantly different between AUD smokers and nonsmokers. Addition of NAC was associated with less IL-6 production by all 4 groups. D: TNFα secretion by LPS-stimulated AMs did not differ across groups (P = 0.23). Addition of NAC was associated with less TNFα expression for all 4 groups, except AUD nonsmokers. Subject numbers were as follows: 12 control nonsmokers, 12 control smokers, 12 AUD nonsmokers, and 24 AUD smokers. Bars indicate medians; error bars indicate median absolute deviation. *P ≤ 0.03, between-group differences after LPS. **P < 0.04 and ***P < 0.001, media + LPS vs. media + LPS + NAC within groups.

Analogous to LPS experiments, stimulation of AMs by LTA was associated with significantly increased secretion of all four cytokines, regardless of subject type (P < 0.0001; Fig. 2). Absolute values of IFNγ after LTA stimulation did not differ between the four groups (P = 0.26; Fig. 2A). Addition of 10 mM NAC with LTA was associated with diminished IFNγ expression, toward values expressed by unstimulated AMs, for all groups (P ≤ 0.001), except AUD nonsmokers (P = 0.08). Similarly, absolute values of IL-1β, IL-6, and TNFα secretion after LTA stimulation were not statistically different across the groups (P > 0.05 for all comparisons; Fig. 2, B–D). In post hoc comparisons, IL-1β secretion by LTA-stimulated AMs differed only between AUD nonsmokers and smokers (P = 0.009). Addition of NAC to AMs with LTA was associated with less IL-1β secretion (P ≤ 0.006) than LTA alone for all groups except the control nonsmokers (P = 0.054; Fig. 2B). Moreover, addition of NAC with LTA was associated with less IL-6 (P ≤ 0.0002; Fig. 2C) and TNFα (P ≤ 0.0005; Fig. 2D) secretion by all four groups than addition of LTA only, regardless of AUD or current cigarette smoking. In mixed-modeling analyses, current smoking was associated with diminished AM secretion of IFNγ (P < 0.04), IL-6 (P < 0.02), and TNFα (P < 0.01) in response to concurrent NAC and LTA administration; however, AUDs did not impact release of cytokines by AMs in response to concurrent NAC and LTA.

Fig. 2.

Freshly obtained AMs from control subjects and subjects with AUDs, both smokers and nonsmokers, were cultured for 18 h in medium only, in medium with 5 μg/ml lipoteichoic acid (LTA), or in medium with 5 μg/ml LTA and 10 mM NAC. IFNγ, IL-1β, IL-6, and TNFα were measured in supernatants. Addition of LTA was associated with significantly increased secretion of all 4 cytokines by AMs across all groups (P < 0.0001). Addition of NAC with LTA was associated with a decrease in cytokine expression toward quantities expressed by unstimulated AMs. A: IFNγ secreted by LTA-stimulated AMs did not differ across groups (P = 0.26). Addition of NAC with LTA was associated with less IFNγ expression (P ≤ 0.001), except in AUD nonsmokers (P = 0.08). B: IL-1β secretion by LTA-stimulated AMs did not differ across groups (P = 0.20) but was significantly greater in AUD smokers than nonsmokers (P = 0.009, by post hoc analysis). Addition of NAC with LTA was associated with less IL-1β secretion, except in control nonsmokers (P = 0.054). C and D: IL-6 and TNFα secretion by LTA-stimulated AMs was not significantly different across groups (P = 0.53 and P = 0.57, respectively). However, addition of NAC with LTA to AM culture was associated with less IL-6 (P ≤ 0.003) and TNFα secretion (P ≤ 0.0005) for all groups. Subject numbers were as follows: 19 control nonsmokers, 20 control smokers, 19 AUD nonsmokers, and 31 AUD smokers. Bars indicate medians; error bars indicate median absolute deviation. *P ≤ 0.01, between-groups comparisons. **P ≤ 0.001, within-group comparisons.

Cytotoxicity of cultured AMs was assessed at the 18-h time point. As a group, cells cocultured with 10 mM NAC had improved survival (34 ± 10% vs. 23 ± 12%, P = 0.02). Diminished cytotoxicity with NAC was more evident in smokers (P = 0.03) and in subjects without AUDs (P = 0.06). Addition of LPS or LTA was not associated with differences in cytotoxicity at the 18-h time point, regardless of AUD or smoking history.

PBMC culture experiments.

PBMCs were collected from 40 subjects: 9 control nonsmokers, 10 control smokers, 10 AUD nonsmokers, and 11 AUD smokers (Table 2). Age and sex differences between groups were similar. Race and Hispanic ethnicity were also similar between groups, although there tended to be more Caucasians among controls. Among smokers, smoking habits did not differ. The total number of PBMCs per unit of blood varied across groups (P = 0.04) but was statistically different only between the AUD nonsmokers and control smokers (P = 0.04). Moreover, neither the subjects' peripheral WBCs nor differentials (data not shown) were significantly different.

Table 2.

Clinical features of subjects and controls with PBMCs utilized in the investigations

	Group
	Control nonsmokers (n = 9)	Control smokers (n = 10)	AUD nonsmokers (n = 10)	AUD smokers (n = 11)	P Value
Age, yr	39.8 ± 8.6	39.3 ± 4.6	45.7 ± 6.9	4.8 ± 8.3	0.26
Sex, %men	88.8	77.8	88.9	90.9	0.92
Race, %
Caucasian	100	50	50	82	0.07
African-American	0	33	11	9	0.07
Native American	0	0	33	9	0.11
Other	0	17	6	0	0.71
Ethnicity, %Hispanic	11	10	40	18	0.40
AUDIT score	3.0 ± 1.6	3.0 ± 2.4	29.7 ± 8.4	31.4 ± 7.8	<0.0001
Current smokers, %	0	100	0	100	1.0
Pack-yr smoking	N/A	15.2 ± 13.5	N/A	19.8 ± 16.2	0.81
BMI, kg/m2	26.7 ± 4.3	29.5 ± 4.2	27.5 ± 5.5	23.7 ± 2.5	0.20
Peripheral WBCs, ×103	5.6 ± 0.6	6.8 ± 1.9	6.6 ± 1.6	7.4 ± 1.8	0.40
MCV, fl	89.5	83 ± 7	92 ± 3	97 ± 7	0.02
Total PBMCs, ×106/10 ml whole blood	11.3 ± 5.4	15.6 ± 4.3	9.0 ± 3.2	10.1 ± 6.1	0.04*

Values are means ± SD or percentages. Peripheral WBC, MCV, and BMI values are from 19 subjects only. Across-groups comparisons were performed with ANOVA for continuous variables (post hoc Tukey's honestly significant difference test where applicable) or Fisher's exact test for categorical variables.

PBMC, peripheral blood mononuclear cell.

^*P = 0.04, AUD nonsmoker vs. control smoker.

As observed with AMs, for all analyses, basal cytokine secretion by PBMCs in medium did not differ substantially; moreover, 2.5–10 mM NAC did not appreciably influence basal cytokine secretion. Addition of LPS significantly increased secretion of all four cytokines by all four subject groups compared with the medium-only conditions (P ≤ 0.001; Fig. 3). Examination of cytokine secretion by LPS-stimulated PBMCs revealed significant between-group differences only for IL-1β secretion (P = 0.03; Fig. 3B) that appeared to be driven by differences between AUD nonsmokers and control smokers (P = 0.04). Addition of NAC along with LPS was associated with significantly diminished secretion of IFNγ, IL-1β, IL-6, and TNFα by all subject types, with the exception of IFNγ by control nonsmokers, which was not influenced by the presence of NAC with LPS (Fig. 3A).

Fig. 3.

Freshly obtained peripheral blood mononuclear cells (PBMCs) from control subjects and subjects with AUDs, both smokers and nonsmokers, were cultured in medium only, in medium with 1 μg/ml LPS, or in medium with 1 μg/ml LPS and 10 mM NAC for 18 h (37°C, 10% CO₂). IFNγ, IL-1β, IL-6, and TNFα were measured in supernatants. Unstimulated PBMC cytokine secretion did not differ between groups (P > 0.05; media). Stimulation of PBMCs with LPS was associated with increased cytokine secretion (P ≤ 0.004; media + LPS), while addition of NAC with LPS was generally associated with diminished cytokine secretion (media + LPS + NAC). A: IFNγ secretion by LPS-stimulated PBMCs did not vary across groups (P = 0.40). Addition of NAC to LPS-stimulated PBMCs was associated with diminished IFNγ expression (P ≤ 0.03) among all subjects, except control nonsmokers. B: IL-1β expression by LPS-stimulated PBMCs varied across groups (P = 0.03), with significant differences between control smokers and AUD nonsmokers (P = 0.04). Addition of NAC to PBMC culture with LPS was associated with less IL-1β expression among all subjects (P ≤ 0.02). C: between-groups differences in LPS-stimulated PBMCs of IL-6 were not statistically significant. Addition of NAC to PBMC culture with LPS was associated with less IL-6 expression (P ≤ 0.001) for all subject groups. D: LPS-stimulated PBMC secretion of TNFα did not vary across groups (P = 0.14). All subjects secreted less TNFα after NAC exposure (P ≤ 0.001). Bars indicate medians; error bars indicate median absolute deviation. Subject numbers were as follows: 9 control nonsmokers, 10 control smokers, 10 AUD nonsmokers, and 11 AUD smokers. *P ≤ 0.04; **P ≤ 0.01.

Stimulation of PBMCs by LTA significantly increased cytokine secretion in AUD and non-AUD groups (P ≤ 0.01; Fig. 4), with the exception of IFNγ expression by nonsmokers (both controls and those with AUDs), where expression was minimally influenced by LTA. No significant across-groups differences in cytokine release after LTA stimulation of PBMCs were observed for any cytokine. When 10 mM NAC was added to PBMCs with LTA, significantly diminished expression of IL-1β, IL-6, and TNFα was observed (P ≤ 0.05) across all four subgroups (Fig. 4, B–D). However, IFNγ secretion was not significantly impacted by addition of NAC with LTA (Fig. 4A).

Fig. 4.

Freshly obtained PBMCs from control subjects and subjects with AUDs, both smokers and nonsmokers, were cultured in medium only, in medium with 5 μg/ml LTA, or in medium with 5 μg/ml LTA and 10 mM NAC for 18 h (37°C, 10% CO₂). IFNγ, IL-1β, IL-6, and TNFα were measured in supernatants. Stimulation of PBMCs by LTA was generally associated with increased secretion of all 4 cytokines (media + LTA), while addition of LTA was associated with diminished cytokine production (media + LTA + NAC). A: addition of LTA was associated with significant increases in IFNγ expression by AMs within control smokers and AUD smokers only (P ≤ 0.05), but across groups, expression of IFNγ after LTA did not differ (P = 0.22). Addition of NAC to PBMC culture with LTA was associated with diminished IFNγ expression, which did not achieve statistical significance. B: LTA-stimulated PBMCs from all groups secreted more IL-1β (P = 0.001), but no significant across-group differences were observed. Addition of NAC to PBMCs with LTA was associated with significantly diminished IL-1β secretion (P ≤ 0.05) across all groups. C: LTA-stimulated PBMCs from all groups secreted more IL-6 (P ≤ 0.004), but across-group differences were not significant. Addition of NAC with LTA was associated with decreased secretion of IL-6 in all groups (P ≤ 0.05). D: TNFα secretion in LTA-stimulated PBMCs was significantly elevated regardless of subject group (P ≤ 0.001), although between-group TNFα secretion did not vary (P = 0.47). Addition of NAC with LTA was associated with decreased secretion of TNFα by AMs (P ≤ 0.05) in all subject groups. Bars indicate medians; error bars indicate median absolute deviation. Subject numbers were as follows: 9 control nonsmokers, 10 control smokers, 10 AUD nonsmokers, and 11 AUD smokers. *P ≤ 0.05.

Magnitude of response between AMs and PBMCs, within subject, by AUDs.

Since collection of PBMCs is less invasive than collection of AMs, we sought to determine if relative cytokine secretion by PBMCs and AMs collected from the same individual, at the same time point, would be comparable and if AUDs would impact this relationship. In nonparametric analyses comparing AM expression with PBMC expression of cytokines within a given individual, only the response of IFNγ secretion to LPS stimulation was determined to be related (Spearman's ρ = 0.61, P = 0.008). Subsequent analyses in the 19 individuals with paired samples indicated that cytokine secretion in response to 1 μg/ml LPS differed substantially between AMs and PBMCs. Across the entire group, PBMCs exposed to 1 μg/ml LPS elaborated quantitatively greater amounts of IFNγ, IL-1β, and IL-6 than did paired AMs from the same individual (P ≤ 0.0003 for each comparison). In contrast, quantitatively more TNFα was secreted by LPS-exposed PBMCs than AMs (P = 0.03). In mixed-modeling analyses, AUDs appeared to further influence PBMC-AM differences in secretion of IFNγ (P = 0.03) and IL-1β (P = 0.002) in response to LPS (Fig. 5, A and B). Within the group of 19 subjects, 5 μg/ml LTA was associated with less PBMC secretion of IFNγ, IL-1β, IL-6, and TNFα than corresponding AMs from the same subject stimulated with 5 μg/ml LTA (P ≤ 0.03 for each comparison). In mixed-modeling analyses, AUDs also influenced PBMC-AM differences in secretion of IFNγ (P = 0.03), IL-1β (P = 0.006), and IL-6 (P = 0.03) in response to LTA (Fig. 5, A–C).

Fig. 5.

Cytokine expression in PBMCs and AMs collected at the same time point in AUD (n = 10) and control (n = 9) subjects (50% of subjects in each group were smokers). In linear mixed-modeling analyses, AUDs appeared to influence expression of IFNγ in response to LPS and LTA (A), IL-1β in response to LPS and LTA (B), and IL-6 in response to LTA (C), but not TNFα in response to LPS or LTA (D). ■ and ▲, median differences between PBMC and AM expression; whiskers, intraquartile ranges.

TLR expression by AMs.

Significant differences in TLR2 (P < 0.0001) expression by uncultured, unstimulated AMs were observed across the four subgroups (Fig. 6A). In post hoc analyses, AM TLR2 expression was diminished in control smokers compared with control nonsmokers (P = 0.003) and AUD nonsmokers (P < 0.0001). AUDs were associated with TLR2 expression comparable to control nonsmokers, although TLR2 expression tended to be less in AUD smokers than AUD nonsmokers (P = 0.06). Differences in TLR4 expression by uncultured, unstimulated AMs were also observed between the four subgroups (P = 0.003; Fig. 6B). In post hoc analyses, TLR4 expression was substantially lower in AMs from control smokers than control nonsmokers (P = 0.007). mRNA expression by AMs before and after stimulation with LPS or LTA was also determined. The average expression changes in TLR2 and TLR4 did not vary across subgroups (Table 3), whether cells had been obtained from control smokers or AUD smokers, and no additive or synergistic effects referable to AUDs or smoking on gene expression response to these PAMPs were observed.

Fig. 6.

Quantitative PCR analysis of Toll-like receptor (TLR) mRNA expression in unstimulated AMs and PBMCs. Values are presented as fold change [arbitrary units (au)] in TLR expression relative to nonsmoking controls (means ± SD). A: AM TLR2 expression differed across the 4 subject groups and was significantly different between smokers/nonsmokers with and without AUDs (P < 0.003, by post hoc analyses). B: AM TLR4 expression varied across the 4 subject groups (P = 0.003) and was substantially diminished in control smokers compared with control nonsmokers (P = 0.007, by post hoc analysis). C: PBMC TLR2 expression varied significantly across the 4 groups (P = 0.003) and was significantly lower in control nonsmokers than the other 3 groups (P < 0.05, by post hoc analysis) D: PBMC TLR4 expression did not vary by subgroup. For AM TLR experiments, n = 10 for each subject group; for PBMC TLR experiments, n = 4 for each subject group. *P ≤ 0.003; **P ≤ 0.05.

Table 3.

Change in mRNA expression of TLR2 and TLR4 by AMs after 18 h of LPS or LTA stimulation

	Group
	Control nonsmokers	Control smokers	AUD nonsmokers	AUD smokers	P Values (between groups)
		TLR2
LPS
Pre	−1.2	−1.3	−1.5	−1.4	0.71
Post	−1.5 to −0.9	−1.7 to −0.9	−2.0 to −1.0	−2.0 to −0.9
LTA
Pre	−1.1	−1.1	−2.1	−1.8	0.30
Post	−1.4 to −0.8	−1.4 to −0.8	−3.3 to −1.0	−2.7 to −0.8
			TLR4
LPS
Pre	−1.0	−0.9	−0.9	−1.2	0.09
Post	−1.3 to −0.7	−1.1 to −0.7	−1.3 to −0.6	−1.5 to −0.9
LTA
Pre	−1.1	−1.1	−1.5	−1.5	0.13
Post	−1.5 to −0.8	−1.3 to −0.9	−2.0 to −0.9	−1.9 to −1.2

Values represent relative change in gene expression.

LTA, lipoteichoic acid; TLR, Toll-like receptor. P values are shown for comparisons across subgroups by ANOVA.

TLR expression by PBMCs.

In uncultured, unstimulated PBMCs, TLR2 expression varied across the four subgroups of subjects (P = 0.003; Fig. 6C). In post hoc comparisons, TLR2 expression was significantly less in control nonsmokers than in control smokers, AUD nonsmokers, or AUD smokers (P < 0.05). No significant differences in TLR4 expression across the four subgroups were observed (P = 0.18; Fig. 6D). Stimulation of PBMCs by LPS or LTA was associated with a relative increase in TLR2 that was not different across subgroups (Table 4). In contrast, stimulation with LPS or LTA was associated with a more substantial decrease in TLR4 expression across the four groups. The magnitude of change did not differ across groups.

Table 4.

Change in mRNA expression of TLR2 and TLR4 by PBMCs after 18 h of LPS or LTA stimulation

	Group
	Control nonsmokers	Control smokers	AUD nonsmokers	AUD smokers	P Values (between groups)
	TLR2
LPS
Pre	1.9	1.5	1.9	2.1	0.85
Post	0.4 to 3.3	0.7 to 2.4	0.2 to 3.5	0.8 to 3.3
LTA
Pre	1.4	1.1	1.1	1.6	0.40
Post	0.6 to 2.2	0.7 to 1.6	0.6 to 1.7	0.7 to 2.5
	TLR4
LPS
Pre	−3.9	−1.2	−2.6	−2.1	0.44
Post	−12.5 to −2.3	−11.1 to 1.7	−7.7 to −1.1	−11.9 to 1.0
LTA
Pre	−3.5	1.3	−1.9	−2.0	0.59
Post	−5.2 to −2.5	−1.2 to 3.8	−3.7 to 1.3	−14.3 to 1.1

Values represent relative change in gene expression. P values are shown for comparisons across groups by ANOVA.

DISCUSSION

Patients with AUDs who acquire pneumonia frequently suffer more complications, including ARDS, and often fare more poorly than individuals without harmful drinking habits. The additional impact of cigarette smoking, a common habit in patients with AUDs, on these outcomes is unknown but may further influence outcomes in pneumonia. Our data highlight alterations in immune cells related to AUDs and current smoking that may be relevant to pulmonary infections. Our work suggests that chronic alcohol consumption, particularly in the absence of concomitant cigarette smoking, is associated with more robust IFNγ and IL-1β secretion by human AMs after exposure to PAMPs, particularly LPS. IFNγ and IL-1β cytokine secretion by PBMCs in response to LPS was affected in parallel with IFNγ and IL-1β secretion in response to LPS in AMs. Current smoking with AUDs appeared to dampen IFNγ, IL-1β, and IL-6 response by AMs to PAMPs, as might be predicted on the basis of prior studies (39, 50), but had less impact on PBMC response to PAMPs. Similar values of AM proinflammatory cytokine secretion in smokers, with or without AUDs, suggest that the influence of smoking on AM proinflammatory cytokine production trumps alterations related to AUDs. Our observations extend previous in vitro chronic alcohol studies (19, 28) to a clinically relevant in vivo exposure system. Additionally, determination of combined effects of in vivo alcohol and cigarette smoke exposure on monocyte cytokine release is completely novel, complementing prior investigations reporting a dampened response of AMs to LPS in the context of smoking (12).

Our observations suggest that AUDs and smoking have the ability to influence proinflammatory cytokine secretion by immune cells, in the lung and systemically, which may have a downstream impact on development and recovery from pneumonia. Our data suggest that chronic alcohol exposure, in particular, may induce an activated, M1-like phenotype in AMs and PBMCs, consistent with prior reports (19, 28, 43). It has been noted that human AMs (in contrast to non-AM mononuclear cells) can be “primed” by LPS exposure in vivo to increase inflammatory mediator production when stimulated by subsequent doses of LPS and LTA (20). Importantly, in the setting of chronic alcohol consumption, intermittent systemic LPS exposure related to defects in the intestinal barrier have been reported (5, 35) and could have led to activation of immune cells in our AUD subjects. Although the enhanced levels of proinflammatory cytokines released in the setting of AUDs might be expected to predict better outcomes in response to pneumonia-causing pathogens, this scenario is not borne out clinically in the setting of AUDs, where ARDS subsequent to an episode of pneumonia is common (32, 33, 44). Possible explanations for this discrepancy may be that the upregulation of inflammatory mediator production by immune cells cannot be sustained in the setting of frank infection or may, perhaps, enhance downregulation of key receptors or other pathways necessary to fully eradicate pathogens or facilitate lung repair. Alternatively, prolonged or enhanced secretion of proinflammatory mediators may contribute to collateral damage by causing overexuberant neutrophil chemotaxis to the lung or promote abnormalities in alveolar-capillary barrier permeability (8, 9), further enhancing the severity of lung injury. Animal data supporting enhanced neutrophilia after LPS exposure in an in vivo alcohol exposure model have been reported (6).

Within the same 19 individuals with paired AM and PBMC samples, PBMCs responded more robustly than AMs to a given dose of LPS, particularly in the setting of AUDs, while the converse was true for LTA stimulation. Variability in response to LPS stimulation in different cell and tissue types has previously been noted (37) and may be clinically relevant, given the potential for recruitment of circulating cells to the lung or for extrapulmonary spread of disease in the setting of pneumonia. Overall, our data and reports of others suggest that the response of immune cells to PAMPs may be influenced by the type of PAMP (e.g., LPS or LTA) and the type of immune cell (e.g., AM or PBMC), as well as environmental factors (e.g., alcohol or smoking). In terms of environment, the effect of alcohol on the ability of immune cells to manufacture and release cytokines has been previously explored, and a suppressive effect of alcohol on interactions between TNFα and the TNFα-converting enzyme (TACE) has been reported. TACE typically cleaves pro-TNFα into the 17-kDa TNFα cytokine at the cell surface (30). In the setting of acute alcohol exposure in experiments utilizing human monocytes and PBMCs, protein-protein interactions between TNFα and TACE in the cell membrane were impaired, thereby blocking TNFα secretion in this setting (51). Effects of chronic alcohol exposure, such as in our subjects, on TACE is unclear; however, chronic alcohol exposure has also been reported to sensitize human monocytes to LPS via mechanisms that include decreased signaling of IL-1-associated receptor kinase M, leading ultimately to increased TNFα elaboration (28). Operative mechanisms underlying cytokine release from immune cells in human subjects with relevant in vivo exposures remain to be explored and might be expected to differ by cell of interest, as well as by duration and intensity of exposure. It is important to acknowledge that, in our investigations, PBMC activity did not perfectly mimic AM activity in the same individuals.

One explanation for differences in cytokine release by AMs and PBMCs in response to PAMPs could be alterations in TLR expression related to AUDs or smoking. However, basal TLR2 and TLR4 expression by AMs was diminished in the setting of AUDs and smoking, while PBMC TLR4 expression was not associated with AUDs or smoking in our cohort; only PBMC TLR2 expression appeared to be relatively affected by AUDs or smoking. These data suggest that the PAMPs LPS and LTA might elicit cytokine release from AMs and PBMCs via non-TLR mechanisms. In the present study, the diminished AM and PBMC cytokine production after PAMP stimulation in the presence of NAC supports the alternative possibility that modulation of oxidative stress has a potent effect on cytokine production by these cells. Depletion of the antioxidant glutathione in the setting of chronic alcohol exposure AUDs has been implicated in cellular oxidative stress (8, 47). Mechanisms that contribute to this observation in AMs may include alcohol-induced upregulation of NADPH oxidases (48, 49), as well as upregulation of mitochondrial oxidative stress (27). Cigarette smoking has also been characterized as a contributor to intrapulmonary oxidative stress, by virtue of its association with downstream effects on proinflammatory mediator production by immune cells (46), and to AM accumulation, retention, and activation (25). The ability of NAC to modulate cytokine production in response to PAMPs in AMs and PBMCs has important therapeutic implications that require further exploration to establish the optimal dose and timing of administration to normalize immune cell functions. Certainly, the ability of NAC to beneficially impact AM and PBMC activities apart from cytokine elaboration, such as phagocytosis, deserves additional exploration.

Our work is not without limitations. First, we acknowledge that cellular activity in vitro does not fully represent the in vivo condition. We have attempted to address this limitation by using cells from individuals with well-characterized and robust alcohol and smoking histories, but without overt comorbid conditions, to focus as closely as possible on the impact of these environmental substances on cell function without additional clinical confounders. Nevertheless, the power of our observations is limited by small sample sizes. Although we were able to match subject groups closely in terms of age, sex, and smoking, differences in the racial composition of our cohort may have an unmeasured impact on our data. Our AUD population was more enriched with subjects who were Native Americans or reported Hispanic ethnicity. Moreover, subtle clinical differences, such as those observed with spirometry, and baseline liver function tests, could have also factored into our observations. Nevertheless, investigations we have performed in a vulnerable population utilizing a more-than-minimal risk procedure are comparable to, if not larger than, similar published studies. Additional studies with larger, more homogenous populations would help validate our observations and further clarify the impact of race on our findings.

We further acknowledge that the specific cell types responsible for differences we observed are not established, and it remains possible that certain subtypes of cells contained within BAL or PBMCs are the primary drivers of our data. However, we chose to perform experiments using the same number of freshly collected cells that had been minimally manipulated ex vivo for each individual subject. Moreover, our methods with BAL cells (wash step) and whole blood (CPT collection tubes), as well as data indicating similar numbers and percentages of each cell type across subject groups, provide reassurance that we are primarily assessing AM and mononuclear cell activity. Certainly, additional cell isolation steps could more clearly establish cell types responsible for these responses. We also realize that our results represent data collected at a single time point (18 h). It is likely that AM and PBMC cytokine secretion is dynamic and fluctuates over time. Additionally, our decision to use serum-free medium for AM culture experiments allowed us to more closely mimic the in vivo condition but might have impacted AM viability at the 18-h time point. However, neither viability of AMs at time 0 nor viability at 18 h varied by subject categorization or PAMP stimulation, suggesting that mechanisms other than viability differences affected AM cytokine secretion.

In summary, in otherwise healthy human subjects, AUDs and smoking influenced the response of innate immune cells in lung and circulation in response to PAMPs. Collectively, our findings suggest that AUDs may prime the immune response in the lung and systemically. Our findings also highlight the possibility that smoking may additionally modify the influence of AUDs on immune cells, particularly AMs, and that strategies to normalize oxidative stress may be able to diminish proinflammatory responses to PAMPs. Our observations pave the way to an understanding of the mechanisms whereby AUDs and cigarette smoking are associated with severe CAP and ARDS.

GRANTS

This work was supported by National Institutes of Health Grants R24 AA-019661 and UL1 TR-001082 and the University of Colorado Department of Medicine.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

J.G. and A.M. performed the experiments; J.G., R.G., P.L.S., and E.L.B. interpreted the results of the experiments; J.G., A.M., R.W.V., P.L.S., and E.L.B. edited and revised the manuscript; R.G., P.L.S., and E.L.B. analyzed the data; R.G. and E.L.B. drafted the manuscript; R.W.V. and E.L.B. developed the concept and designed the research; E.L.B. prepared the figures; E.L.B. approved the final version of the manuscript.