Elevated Cellular Oxidative Stress in Circulating Immune Cells in Otherwise Healthy Young People Who Use Electronic Cigarettes in a Cross‐Sectional Single‐Center Study: Implications for Future Cardiovascular Risk

Abstract

Background

Tobacco cigarettes (TCs) increase oxidative stress and inflammation, both instigators of atherosclerotic cardiac disease. It is unknown if electronic cigarettes (ECs) also increase immune cell oxidative stress. We hypothesized an ordered, “dose‐response” relationship, with tobacco‐product type as “dose” (lowest in nonsmokers, intermediate in EC vapers, and highest in TC smokers), and the “response” being cellular oxidative stress (COS) in immune cell subtypes, in otherwise, healthy young people.

Methods and Results

Using flow cytometry and fluorescent probes, COS was determined in immune cell subtypes in 33 otherwise healthy young people: nonsmokers (n=12), EC vapers (n=12), and TC smokers (n=9). Study groups had similar baseline characteristics, including age, sex, race, and education level. A dose‐response increase in proinflammatory monocytes and lymphocytes, and their COS content among the 3 study groups was found: lowest in nonsmokers, intermediate in EC vapers, and highest in TC smokers. These findings were most striking in CD14_dimCD16⁺ and CD14⁺⁺CD16⁺ proinflammatory monocytes and were reproduced with 2 independent fluorescent probes of COS.

Conclusions

These findings portend the development of premature cardiovascular disease in otherwise healthy young people who chronically vape ECs. On the other hand, that the COS is lower in EC vapers compared with TC smokers warrants additional investigation to determine if switching to ECs may form part of a harm‐reduction strategy.

Registration

URL: https://www.clinicaltrials.gov; Unique identifier: NCT03823885.

Nonstandard Abbreviations and Acronyms

COS

cellular oxidative stress

EC

electronic cigarette

NK

natural killer

TC

tobacco cigarette

Clinical Perspective

What Is New?

• Electronic cigarette (EC) vaping, which has grown to epidemic proportions among young people, is perceived as safer than tobacco cigarette smoking, but it remains unknown if otherwise healthy young EC vapers, like tobacco cigarette smokers, have increased oxidative stress and inflammation compared with nonsmokers.

• A dose‐response increase in proinflammatory monocytes and lymphocytes, and their cellular oxidative stress content was found: lowest in nonsmokers, intermediate in EC vapers, and highest in tobacco cigarette smokers.

What Are the Clinical Implications?

• These findings portend the development of premature cardiovascular disease in otherwise healthy young people who chronically vape ECs.

• On the other hand, that the cellular oxidative stress is lower in EC vapers compared with tobacco cigarette smokers warrants additional investigation to determine if switching to ECs may form part of a harm‐reduction strategy.

Oxidative stress and inflammation are implicated in the pathogenesis of most human diseases, including cardiovascular diseases. ¹ Long‐term exposure to excessive levels of reactive oxygen species (ROS) introduced through environmental exposures or through dysfunctional endogenous enzymatic systems overwhelm antioxidant defense systems, resulting in cellular damage and activation of circulating immune cells. ¹ , ² Activated immune cells, in turn, generate additional ROS, driving oxidation of lipoproteins and further recruitment of monocytes and macrophages, which then enter the vascular wall. Thus, ongoing oxidative stress and inflammation contribute to the initiation and progression of atherosclerotic vascular disease that may present decades later.

Tobacco cigarette (TC) smoking is the most prevalent modifiable risk factor for numerous human diseases, including atherosclerosis, in which oxidative stress and inflammation are known to play a critical role. ² , ³ Over 90% of TC smokers begin smoking in their teens, ⁴ but TC‐related diseases are insidious, presenting only after decades of TC smoking. Each puff of TC smoke contains 10¹⁵ free radicals ⁵ and >7000 different chemicals, ⁶ several of which are known toxicants or even carcinogens. Major prooxidant constituents in TC smoke generate cellular production of ROS when they interact with cellular enzymatic systems. ² Innate and adaptive immune cells, such as myeloid cells (monocytes, macrophages, and dendritic cells), natural killer (NK) cells, and lymphocytes (B and T cells) are activated by TC smoking, ⁷ and are also major sources of systemic oxidative stress. ⁸ Cigarette smoke activates leukocytes to release reactive oxygen and nitrogen species and contributes to development and progression of atherosclerotic cardiovascular disease through several mechanisms, such as secretion of proinflammatory cytokines and increased adherence of monocytes to the endothelium. ² , ³ Although cellular oxidative stress (COS) has been studied in the setting of tobacco smoking and atherosclerosis, there is limited evidence regarding COS among electronic cigarette (EC) vapers.

ECs are the most rapidly rising tobacco product used in the United States today. EC aerosol, generated from heating, without combustion, solvents, flavors, and usually nicotine, contains significantly lower levels of toxicants compared with TC smoke. ⁹ Because of the long lag time for disease presentation, the health risks of ECs relative to TCs are unknown, yet ECs have been promoted as a smoking cessation, harm reduction, strategy. Alarmingly, largely because of the perceptions that ECs are safe, EC vaping has reached epidemic levels in never‐smoking middle and high school students, with 30% of high school seniors (typically 17–18 years old) reporting EC vaping in the previous month. ¹⁰

Although an urgent public health issue, the health risks associated with EC vaping, especially relative to TC smoking, remain unknown. The purpose of the current study was to pair sensitive flow cytometry with fluorescent probes to quantify the relative immune cell‐type populations and their intracellular content of ROS in otherwise healthy young EC vapers compared with TC smokers, and nonsmokers. We hypothesized a continuum of oxidative stress and immune cell activation, essentially a “dose‐response” relationship, with the “dose” defined as tobacco‐product type (lowest in the nonsmokers, intermediate in the long‐term EC vapers and highest in the long‐term TC smokers), and the “response” defined as measures of immune cell subtypes and their COS.

Methods

The data that support the findings of this study are available from the corresponding author on reasonable request. H.R.M. and T.K. had full access to all the data in the study and take responsibility for their integrity and the data analysis.

Materials

Flow cytometry reagents, including flow cytometry staining buffers and antibodies, were purchased from Biolegend. CellROX Green (catalog No. C10444) and CellROX Deep Red (catalog No. C10442) were obtained from Thermo Scientific.

Study Population

Healthy male and female volunteers between the ages of 21 and 45 years were eligible for enrollment if they were long‐term (≥1 year) (1) TC smokers, (2) EC vapers (no dual users), or (3) nonsmokers. Former TC smokers were eligible if >1 year had elapsed since quitting. End‐tidal CO, elevated >10 ppm in smokers, was measured in EC vapers and nonsmokers to confirm none were surreptitiously smoking TCs. All participants were required to meet the following criteria: (1) nonobese (≤30 kg/m² body mass index); (2) no known health problems; (3) alcoholic intake ≤2 drinks per day and no regular illicit drug use, including marijuana, determined through screening questionnaire and urine toxicology testing; (4) no prescription medications (oral contraceptives allowed); and (5) not exposed to second‐hand smoke, or using licensed nicotine replacement therapies. The experimental protocol was approved by the Institutional Review Board at the University of California, Los Angeles, and written, informed consent was obtained from each participant.

Experimental Protocol

After abstaining from caffeine, tobacco product use, and exercise for at least 12 hours, fasting participants reported to the UCLA Clinical Translational Research Center at the same time of day, ≈8 am. Blood was drawn by trained medical assistants and prepared for flow cytometry and measurement of cotinine levels.

Flow Cytometry

Freshly isolated whole blood was immediately processed for flow cytometric determination of cellular ROS. COS was determined by the use of the CellROX Green Reagent, a measure of total (cytoplasmic and nuclear) cellular ROS, ¹¹ , ¹² , ¹³ and the use of the CellROX Deep Red Reagent, a measure of cytoplasmic cellular ROS. ¹⁴ , ¹⁵ , ¹⁶ The efficiency of CellROX Green to determine COS has previously been validated in several cells, including sperm, epithelial and melanoma cells, neurons, bacteria, and immune cells, such as macrophages. ¹⁷ The efficiency of CellROX Deep Red to assess COS has previously been validated in several cells, including sperm, endothelial and epithelial cells, hepatocytes, neurons, cardiomyocytes, and immune cells. ¹⁵ The CellROX Deep Red has been previously used to detect the ex vivo impact of cigarette smoke on cellular ROS by flow cytometry in spermatocytes. ¹⁶

See Data S1 for detailed methods.

Determination of Plasma Cotinine Levels

The assay for plasma cotinine, using the method of chromatography/mass spectrometry, was run by the commercial laboratory, Quest Laboratories (Quest Diagnostics Incorporated, Madison, NJ), with a limit of quantitation of 2 ng/mL and a reference range in smokers of 16 to 145 ng/mL.

Statistical Analysis

We hypothesized an ordered, dose‐response relationship of oxidative stress across the 3 study groups: lowest in nonsmokers, intermediate in long‐term EC vapers, and highest in long‐term TC smokers. We considered the “dose” to be the type of tobacco product used, and the “response” to be the immune cell subtype and its COS. To test this hypothesis, the ordered trend (F) test across the 3 ordered groups (nonsmokers, EC vapers, TC smokers) was computed under an ANOVA model. ¹⁸ Means±SEM are reported. If the overall trend P value or the overall ANOVA P value was ≤0.05, then the pairwise post hoc t test P values are reported between 2 groups (Fisher least significant difference criterion). The ordered trend test was considered statistically significant when P≤0.05. For continuous outcomes, examination of normal quantile plots and the Shapiro‐Wilks statistic confirmed that the distributions followed the normal distribution. Overall and pairwise P values for comparing categorical covariates (sex, race, and education) across the 3 study groups were computed using the Fisher exact test.

Sample Size Calculation

Our primary outcomes are COS in proinflammatory monocytes, given their role in cardiovascular disease. ¹⁹ Given absence of data on monocyte frequencies or COS in immune cells in EC vapers, and based on data on frequencies of proinflammatory monocytes in otherwise healthy people without clinical disease, ²⁰ a sample size of 9 participants per group (nonsmokers, EC vapers, and TC smokers) was sufficient to permit detection of a Δ of 2.9% with 80% power and 2‐sided α=0.05. A total of 9 to 12 participants were included in each study group. This study, largely exploratory, is not powered to detect effect sizes with adjustments for multiple comparisons. ²¹ , ²² This is an interim report of our study registered at ClinicalTrials.gov (NCT03823885), which is a short‐term exposure, crossover study.

RESULTS

Baseline Characteristics

A total of 33 participants, including 12 nonsmokers (age, 24.3±2.2 years; 5 women), 12 long‐term EC vapers (age, 24.1±4.3 years; 4 women), and 9 long‐term TC smokers (age, 24.9±4.1 years; 5 women), participated in the study. Baseline characteristics of the 3 groups are shown in the Table. There were no differences among the groups in any variable, including age, sex, race, body mass index, or education level. All smokers and vapers used their tobacco product daily. Ten EC vapers reported using a “pod” device (eg, JUUL), and one each used a “mod” or a “cigalike” device; all EC vapers used flavored, nicotine‐containing liquid. Plasma cotinine levels were not significantly different in TC smokers and EC vapers (58 versus 85 ng/mL, respectively; P=0.34), consistent with similar, and relatively light, smoking burden.

Table 1.

Baseline Characteristics

Characteristic	Nonsmokers	EC Vapers	TC Smokers	P Value
	(n=12)	(n=12)	(n=9)
Age, y	24.3±2.15	24.1±4.34	24.9±4.08	0.54
Sex (men/women)	7/5	8/4	4/5	0.61
Race				0.65
White	4	6	2
Asian	4	5	3
Black	2	0	1
Hispanic	2	1	1
Unknown	0	0	2
BMI, kg/m2	24±3.66	22.6±2.89	23.0±3.47	0.37
Plasma cotinine, ng/mL	0	85.0±126.2	58.0±39.5*
Highest level education				1.0
<High school	0	0	0
≥College	12	12	9

Values are given as number or mean±SD. BMI indicates body mass index; EC, electronic cigarette; and TC, tobacco cigarette.

^* P=0.34, EC vapers vs TC smokers.

Immune Cell Subtypes

To assess the impact of long‐term smoking on immune cells, we first determined the frequency of immune cell subtypes among smoking groups (Figure ). Gating strategies for viability dye and antibody staining are shown in Figure S1. Neutrophils, CD14dimCD16⁺ monocytes, and NK, T, and B cells were found in the lowest proportion in the nonsmokers, intermediate in the EC vapers, and in the greatest proportion in TC smokers and were lower in nonsmokers compared with TC smokers (Figure 1A through 1J).

Figure 1. Frequency of immune cell types among smoker groups.

Flow cytometry was used to determine the percentage of different immune cell types in CD45⁺ immune cells (A through J). The compared groups were nonsmokers (NSs; white), electronic cigarette vapers (EC vapers; light gray), and tobacco cigarette smokers (TC smokers; dark gray). Summary of data (% cellular marker⁺ of parent population) is shown for CD45⁺CD15⁺CD16⁺CD14⁻hi‐SSC neutrophils (A), CD45⁺CD14⁺⁺CD16⁻ classic monocytes (B), CD45⁺CD14⁺⁺CD16⁺ intermediate monocytes (C), CD45⁺CD14dimCD16⁺ nonclassic (patrolling or CD14⁺CD16⁺⁺) monocytes (D), CD45⁺CD14⁺CD16⁺ total proinflammatory monocytes (intermediate and nonclassic) (E), CD45⁺CD3⁺ T cells (F), CD45⁺CD3⁺CD4⁺ T cells (G), CD45⁺CD3⁺CD8⁺ T cells (H), CD45⁺CD3⁻CD56⁺CD16⁺ natural killer (NK) cells (I), and CD45⁺CD19⁺ B cells (J). Data represent box‐and‐whisker boxes that display the minimum, mean, and maximum (n=9–12 participants per group). The ANOVA statistical test was used to compare 3 groups, and the t test was used to compare 2 groups. The trend P analysis tested the continuum of the difference in measures among groups in an ordered direction (NSs→EC vapers→TC smokers). *P<0.05, **P<0.01, ***P<0.001.

COS in CD45⁺ Immune Cells

Given the lack of data on the impact of EC vaping on COS, we then determined the relative impact of long‐term TC smoking or EC vaping on COS, as measured by flow cytometry using the fluorescent probes CellROX Green, a measure of total (cytoplasmic and nuclear) cellular ROS, and CellROX Deep Red, a measure of cytoplasmic cellular ROS. There was a dose‐response relationship among the 3 study groups for the percentage of CD45⁺ immune cells that were positive for total (Figure 2A and 2B) and cytoplasmic (Figure 2C and 2D) ROS (lowest in nonsmokers, intermediate in EC vapers, and greatest in TC smokers). In addition, the mean fluorescence intensity of total (Figure 2E and 2F) and cytoplasmic (Figure 2G and 2H) ROS in CD45⁺ immune cells also demonstrated this same, consistent dose‐response relationship. Between‐group comparisons consistently showed significantly greater COS in TC smokers compared with nonsmokers (Figure 2A through 2H). Cytoplasmic ROS was greater in TC smokers compared with EC vapers as well (Figure 2C and 2D).

Figure 2. Cellular oxidative stress in CD45⁺ immune cells among smoker groups.

Flow cytometry was used to determine total (nuclear and cytoplasmic) and cytoplasmic reactive oxygen species. The compared groups were nonsmokers (NSs; white), electronic cigarette vapers (EC vapers; light gray), and tobacco cigarette smokers (TC smokers; dark gray). Representative data of percentage of immune (CD45⁺) cells that had positive staining for CELLROX Green among compared groups are shown (A). Summary of data for (A) is shown (B). Representative data of percentage of CD45⁺ cells that had positive staining for CELLROX Deep Red among compared groups are shown (C). Summary of data for C is shown (D). Representative data of CellROX Green change in mean fluorescence intensity (∆MFI) in CD45⁺ cells are shown (E). Fluorescence intensity of a positive cell population was compared with a negative cell population (fluorescence minus one negative control for staining) (∆MFI). Summary of data for E is shown (F). Representative data of CellROX Deep Red ∆MFI in CD45⁺ cells is shown (G). Summary of data for G is shown (H). Data represent box‐and‐whisker boxes that display the minimum, mean, and maximum (n=9–12 participants per group). The ANOVA statistical test was used to compare 3 groups, and the t test was used to compare 2 groups. The trend P analysis tested the continuum of the difference in measures among groups in an ordered direction (NSs→EC vapers→TC smokers). *P<0.05, **P<0.01. SSC indicates side scatter.

COS in Specific Immune Cell Types

We then determined the impact of smoking exposures on COS among immune cell types (Figures 3, 4, 5). Group comparisons between TC smokers and EC vapers showed that there were no differences in ROS in neutrophils (Figure 3A through 3D). The proportion of B cells that had detectable total ROS (Figure 3I) and the proportion of NK (Figure 3G), B (Figure 3K), and total CD3⁺, CD4⁺, and CD8⁺ T cells (Figure 4C, 4G, and 4K) that had detectable cytoplasmic ROS were greater in TC smokers compared with EC vapers. Similar data were seen for the mean content for cytoplasmic ROS in NK cells (Figure 3H) and for the mean content for total (Figures 4J and 5J) and cytoplasmic (Figures 4L and 5H, 5L, 5P) ROS in CD8⁺ T cells (Figure 4J and 4L) and proinflammatory monocytes (Figure 5H, 5J, 5L, and 5P). There were no differences in total ROS (Figure 3E and 3F) or the mean content for total (Figures 3J and 4B, 4F) and cytoplasmic (Figure 3L) ROS in NK (Figure 3E and 3F) and B cells (Figure 3L) in TC smokers compared with EC vapers.

Figure 3. Cellular oxidative stress in neutrophils, natural killer (NK) cells, and B cells among smoker groups.

Flow cytometry was used to determine total (nuclear and cytoplasmic) and cytoplasmic reactive oxygen species. The compared groups were nonsmokers (NSs; white), electronic cigarette vapers (EC vapers; light gray), and tobacco cigarette smokers (TC smokers; dark gray). Summary data of percentage of immune cells that had positive staining for CELLROX Green (A, E, I) and CELLROX Deep Red (C, G, K) and for change in mean fluorescence intensity (∆MFI) CellROX Green (B, F, J) and ∆MFI CellROX Deep Red in cells (D, H, L) among compared groups are shown for CD45⁺CD15⁺CD16⁺CD14⁻hi‐SSC neutrophils (A through D), CD45⁺CD3⁻CD56⁺CD16⁺ NK cells (E through H), CD45⁺CD19⁺ B cells (I through L). Data represent box‐and‐whisker boxes that display the minimum, mean, and maximum (n=9–12 participants per group). The ANOVA statistical test was used to compare 3 groups, and the t test was used to compare 2 groups. The trend P analysis tested the continuum of the difference in measures among groups in an ordered direction (NSs→EC vapers→TC smokers). *P<0.05, **P<0.01, ***P<0.001.

Figure 4. Cellular oxidative stress in T cell subsets among smoker groups.

Flow cytometry was used to determine total (nuclear and cytoplasmic) and cytoplasmic reactive oxygen species. The compared groups were nonsmokers (NSs; white), electronic cigarette vapers (EC vapers; light gray), and tobacco cigarette smokers (TC smokers; dark gray). Summary data of percentage of immune cells that had positive staining for CELLROX Green (A, E, and I) and CELLROX Deep Red (C, G, and K) and for change in mean fluorescence intensity (∆MFI) CellROX Green (B, F, and J) and ∆MFI CellROX Deep Red in cells (D, H, and L) among compared groups are shown for CD45⁺CD3⁺ T cells (A through D), CD45⁺CD3⁺CD4⁺ T cells (E through H), and CD45⁺CD3⁺CD8⁺ T cells (I through L). Data represent box‐and‐whisker boxes that display the minimum, mean, and maximum (n=9–12 participants per group). The ANOVA statistical test was used to compare 3 groups, and the t test was used to compare 2 groups. The trend P analysis tested the continuum of the difference in measures among groups in an ordered direction (NSs→EC vapers→TC smokers). *P<0.05, ***P<0.001.

Figure 5. Cellular oxidative stress in monocyte subsets among smoker groups.

Flow cytometry was used to determine total (nuclear and cytoplasmic) and cytoplasmic reactive oxygen species. The compared groups were nonsmokers (NSs; white), electronic cigarette vapers (EC vapers; light gray), and tobacco cigarette smokers (TC smokers; dark gray). Summary data of percentage of immune cells that had positive staining for CELLROX Green (A, E, I, and M) and CELLROX Deep Red (C, G, K, and O) and for change in mean fluorescence intensity (∆MFI) CellROX Green (B, F, J, and N) and ∆MFI CellROX Deep Red in cells (D, H, L, and P) among compared groups are shown for CD45⁺CD15⁺CD16⁺CD14⁻hi‐SSC neutrophils (A through D), CD45⁺CD3⁻CD56⁺CD16⁺ natural killer cells (E through H), CD45⁺CD19⁺ B cells (I through L), and CD45⁺CD3⁺ T cells (M through P). Data represent box‐and‐whisker boxes that display the minimum, mean, and maximum (n=9–12 participants per group). The ANOVA statistical test was used to compare 3 groups, and the t test was used to compare 2 groups. The trend P analysis tested the continuum of the difference in measures among groups in an ordered direction (NSs→EC vapers→TC smokers). *P<0.05, **P<0.01, ***P<0.001. MNC indicates monocytes.

Group comparisons between TC smokers and nonsmokers showed that the proportion of B cells (Figure 3I and 3K) and proinflammatory monocytes (Figure 5C, 5E, 5G, 5K, 5L, 5M and 5O) that had detectable cellular total (Figures 3I and 5E, 5L, 5M) and cytoplasmic (Figures 3K and 5C, 5G, 5K, 5O) ROS was greater in TC smokers compared with nonsmokers. Similar results were seen for cytoplasmic ROS in NK (Figure 3G), B (Figure 3K), and T cells (Figure 4C), and T cell (Figure 4G and 4K) and monocyte (Figure 5C, 5G, 5K, and 5O) subsets. The mean cellular content for total (Figures 4J and 5F, 5J) and cytoplasmic (Figures 4L and 5H, 5L, 5P) ROS was higher in CD8⁺ T cells (Figure 4J and 4L) and proinflammatory monocytes (Figure 5F, 5H, 5J, 5L, and 5P) in TC smokers compared with nonsmokers. Similar trends (0.05<P<0.10) were observed in neutrophils (Figure 3D), NK cell (Figure 3F), T cell (Figure 4D), and monocyte subsets (Figure 5D and 5N) but were not consistent among independent readouts of COS. There were no other consistent differences in measures of COS in immune cell types between TC smokers and nonsmokers (Figures 3A through 3F, 3H, 3J, 3L, 4A, 4B, 4D, 4F, 4I, and 5A, 5B, 5D, 5N).

Group comparisons between EC vapers and nonsmokers showed that EC vapers had higher proportion of monocyte subsets (Figure 5C, 5G, 5K, and 5O) that had detectable total (Figure 5E, 5I, and 5M) and cytoplasmic (Figure 5C, 5G, 5K, and 5O) ROS compared with nonsmokers. Similar results were seen for cytoplasmic ROS in NK (Figure 3G) and CD4⁺ T cells (Figure 4G) and the mean cellular content for total (Figure 5J) and cytoplasmic (Figure 5H and 5L) ROS in proinflammatory monocytes. There were no differences in other measures of COS in other immune cell types between compared groups (Figures 3A through 3F, 3H through 3L, 4, and 5A, 5B, 5D, 5F).

There was a dose‐response relationship among the 3 study groups for the mean percentage of NK (Figure 3G), B (Figure 3K), and T cells (Figure 4) and monocyte (Figure 5C, 5G, 5K, and 5O) subtypes with cytoplasmic ROS: lowest in the nonsmokers, intermediate in EC vapers, and greatest in TC smokers. The mean percentage of proinflammatory monocytes positive for total ROS (Figure 5E, 5I, and 5M) and the mean cellular content for total (Figure 5F, 5J, and 5N) and cytoplasmic ROS in proinflammatory monocytes (Figure 5H, 5L, and 5P) and T cell subtypes (Figure 4G and 4K) also followed this same pattern. The COS findings in different immune cell subpopulations and whether or not the dose‐response relationship was observed are summarized in Figure 6.

Figure 6. Ordered, “dose‐response” relationship in cellular oxidative stress among immune cell types and smoker groups, with tobacco‐product type as “dose.”.

EC indicates electronic cigarette; NK, natural killer; ROS, reactive oxygen species; and TC, tobacco cigarette.

Discussion

To our knowledge, this is the first study to report alterations in the proportion of circulating innate and adaptive immune cells, as well as their COS content, in otherwise healthy young people who are long‐term EC vapers or TC smokers compared with nonsmokers. Overall, we found a marked and consistent dose‐response increase in proinflammatory monocytes and lymphocytes, and their total cellular and cytoplasmic ROS content among the 3 study groups: lowest in the nonsmokers, intermediate in EC vapers, and highest in TC smokers. These findings were most striking in CD14_dimCD16⁺ and intermediate in CD14⁺⁺CD16⁺ proinflammatory monocytes and were reproduced with 2 independent fluorescent probes that determine total (CellROX Green) and cytoplasmic (CellROX Deep Red) cellular ROS.

Oxidative stress plays a major role in inflammation and cellular activation and is a major contributor to atherosclerotic cardiovascular disease. ¹ , ² , ³ The presence of excessive ROS has been termed the “convergent signaling hub” that underlies inflammatory diseases, including smoking‐related atherosclerotic disease. ²³ These findings of increased COS in key innate and adaptive immune cell subtypes portend the future development of premature atherosclerosis in otherwise healthy young people who chronically vape ECs.

TC smoking is a significant independent risk factor for many chronic and lethal diseases in humans. ¹ , ² Given the powerfully addictive nature of nicotine and the low rate of successful smoking cessation, ECs have been proposed as a potential harm‐reduction strategy, with the ultimate goal of reducing morbidity and mortality while satisfying nicotine addiction. ²⁴ ECs may emit fewer toxicants and carcinogens compared with TCs, but our findings confirm that their long‐term use is associated with increased innate and adaptive immunity with increased COS. Although the proportion of immune cell subtypes, and their burden of COS, may be less in long‐term EC vapers compared with TC smokers, it remains unproven and unknown if there is a “safe” level of chronic oxidative stress and inflammation.

Previous attempts to predict the adverse future health effects of ECs have been hampered by methodological limitations, such as relying on in vitro model systems or focusing on short‐term, not long‐term, EC exposure; in addition, most studies have been significantly underpowered. ²⁵ , ²⁶ , ²⁷ , ²⁸ , ²⁹ In one of the few studies of health effects in long‐term EC vapers, we reported an increased susceptibility to, but not actual presence of, chronic oxidative stress, estimated by low‐density lipoprotein oxidizability, compared with healthy nonsmoking controls. ³⁰ Traditional, clinical biomarkers of inflammation, including fibrinogen and CRP (C‐reactive protein), were not elevated. ³⁰ Admittedly, measurements of biomarkers in plasma lack sensitivity to elucidate the effects of ECs on oxidative stress and immune cell activation.

We found that COS was consistently elevated in CD14_dimCD16⁺ and intermediate CD14⁺⁺CD16⁺ proinflammatory monocytes of TC smokers and EC vapers compared with nonsmokers. CD14⁺CD16⁺ monocytes are known contributors to atherosclerotic cardiovascular disease, ³¹ , ³² , ³³ have increased chemotactic properties, and are potent secretors of interleukin‐1, interleukin‐6, and tumor necrosis factor‐α. ³⁴ However, their specific roles in atherosclerosis progression, lesion stability, and clinical events are uncertain. This monocyte subpopulation was also associated with increased vascular superoxide production in vascular dysfunction. ³⁵ Consistent with our data, it has been shown that CD14⁺CD16⁺ monocytes have lower levels of antioxidant genes and increased aerobic respiration and ROS production capacities. ³⁶ Given that oxidative stress is a known instigator of atherosclerosis, ² , ³ it remains to be shown whether increased pro‐oxidant capacity of CD14⁺CD16⁺ monocytes in the setting of EC vaping during lung chemotaxis may contribute to subsequent oxidative stress in arteries, portending the development of premature cardiovascular disease in otherwise healthy young people who chronically vape ECs.

The direct quantification of ROS is a valuable and promising biomarker that can reflect the disease process. However, given the short half‐life of these species, their measurement in biological systems is complex. Determination of ROS has several methodological concerns, and global ROS measurements need to be avoided. ³⁷ Identifying individual molecular targets of oxidation‐reduction regulation is needed, and the complexity of COS can be studied only at the single cell level. ¹² Approaches, such as mass spectrometry and spectrophotometric or luminescence methods, have major methodological limitations. ³⁸ Although there is no single method that detects ROS that does not have limitations, the relative differences among different samples may be assessed reasonably and the bias of each method to detect ROS could be overcome by the evaluation of oxidative stress by using >1 criterion. ¹² Flow cytometry is one of the most powerful tools for single‐cell analysis of the immune system. Many fluorescent probes for the detection of reactive species have been developed in the last years, with a different degree of specificity and sensitivity. ¹²

The CellROX Deep Red has been previously used to detect the ex vivo impact of TC smoke on cellular ROS by flow cytometry in spermatocytes. ¹⁶ The use of these fluorochromes for determination of COS in immune cells has previously been validated both in vitro ¹⁷ and in vivo. ³⁹ The CellROX ROS detection reagents are bright and stable ROS sensors that offer significant advantages over existing ROS sensors because they are compatible with labeling in different media and can be used with fixatives. ⁴⁰ This combined use has previously been described in nonimmune cells. ⁴¹ To the best of our knowledge, this study is pioneering in evaluating the efficiency of these probes in detecting ROS production among unique immune cell subsets.

Our study has limitations. Unlike animal studies, participants in human studies are heterogeneous. It is possible, but unlikely, that unmeasured, confounding differences exist among the 3 study groups, besides the obviously different smoking habits, to explain the marked and consistent differences in the proportion of immune cell subtypes and their oxidative stress. However, by any major demographic measure, including age, sex, race, and education level, the 3 study groups were markedly similar (Table). EC vaping is difficult to quantify objectively and then compare with commonly used measures of TC smoking (eg, number of cigarettes per day). Because all of our vapers used ECs with nicotine, plasma cotinine levels were used as an objective, quantifiable measure, common to both EC and TC users, that could be compared between groups to estimate relative tobacco product burden. Our study is a small single‐center study, and not powered to detect effect sizes with adjustment for multiple comparisons. Rather, consistency, direction, and magnitude of the effect in conjunction with the nominal P values were considered to help distinguish true‐ and false‐positive findings. ²¹ , ²² Accordingly, by leveraging the powerful technique of flow cytometry coupled to 2 different sensitive fluorescent probes, we were able to find a consistent dose‐response relationship in COS among the 3 study groups that was repeated in both innate and adaptive immune cells. We acknowledge, however, that confirmation of these findings in additional participants is warranted.

In conclusion, our study is the first to report an increased proportion of proinflammatory monocytes/macrophages, NK cells, and T and B lymphocytes, in otherwise healthy young people who are long‐term EC vapers compared with nonsmokers. This increased proportion of innate and adaptive immune cell subtypes is coupled with the finding that long‐term EC vapers have elevated COS as well. Because low‐grade oxidative stress and inflammation have been identified as the underlying mechanism that instigates and perpetuates atherosclerotic vascular disease that may manifest only decades later, these findings have important future health implications for young people who vape. On the other hand, that the COS is lower in long‐term EC vapers compared with TC smokers is intriguing and warrants additional investigation to determine if switching to ECs may avoid activation of downstream detrimental cellular pathways, supporting their role as part of a harm‐reduction strategy for cardiovascular disease. Future studies delineating the specific cellular pathways impacted in humans who chronically use ECs compared with TCs may provide further insights into their relative health risks, and whether switching to ECs will result in harm reduction.

Sources of Funding

This work was supported by the Tobacco‐Related Disease Research Program (TRDRP) under contract TRDRP 28IR‐0065 (Middlekauff), and by the National Institutes of Health (NIH) National Center for Advancing Translational Science UCLA Clinical and Translational Science Institute grant L1TR001881. This work was also supported in part by NIH grants R01AG059501 and R03AG059462 (Kelesidis). The flow cytometry machine used in the study was purchased through the UCLA Center for AIDS Research (P30AI28697) grant.

Disclosures

None.

Supporting information

Data S1

Figure S1

References 42 , 43

(J Am Heart Assoc. 2020;9:e016983 DOI: 10.1161/JAHA.120.016983.)