Dysregulated Metabolites Serve as Novel Biomarkers for Metabolic Diseases Caused by Vaping and Cigarette Smoking

Metabolites are essential intermediate products in metabolism, and metabolism dysregulation indicates different types of diseases. Previous studies have shown that cigarette smoke dysregulated metabolites; however, limited information is available with electronic cigarette (E-cig) vaping. We hypothesized that E-cig vaping and cigarette smoking altered systemic metabolites, and we propose to understand the specific metabolic signature between E-cig users and cigarette smokers. Plasma from non-smoker controls, cigarette smokers, and e-cig users were collected, and metabolites were identified by UPLC–MS (Ultraperformance liquid chromatography-mass spectrometer). Nicotine degradation was activated by e-cig vaping and cigarette smoking with increased concentrations of cotinine, cotinine N-oxide, (S)-nicotine, and (R)-6-hydroxynicotine. Additionly, we found significant decreased concentrations in metabolites associated with tricarboxylic acid (TCA) cycle pathways in e-cig users verses cigarette smokers, such as: D-glucose, (2R,3S)-2,3-dimethylmalate, (R)-2-hydroxyglutarate, O-phosphoethanolamine, malathion, D-threo-isocitrate, malic acid, and 4-acetamidobutanoic acid. Cigarette smoking significant up-regulated sphingolipid metabolites, such as D-sphingosine, ceramide, N-(octadecanoyl)-sphing-4-enine, N-(9Z-octadecenoyl)-sphing-4-enine, and N-[(13Z)-docosenoyl]sphingosine, verses e-cig vaping. Overall, e-cig vaping dysregulated TCA cycle realted metabolites while cigarette smoking altered sphingolipid metabolites. Both e-cig and cigarette smoke increased nicotinic metabolites. Therefore, specific metabolic signature altered by e-cig vaping and cigarette smoking could serve as potential systemic biomarkers for early cardiopulmonary diseases.


Introduction
E-cig vaping has been increasing rapidly in the United States during recent decades since e-cig is considered a relatively safer alternative to help quit smoking 1 . The e-cig devices deliver aerosolized e-liquid with different concentrations of nicotine. The constituents from e-cig liquid, which are usually propylene glycol (PG) and vegetable glycerin (VG), which are Generally Recognized as Safe (GRAS). Although PG and VG are GRAS, the aerosolized constituents have proven to be toxicants 2

. It has been
known that e-cig delivers more nicotine than cigarette smoke 3,4 . Furthermore, we have shown that e-cig vapor contained various chemical constituents that can affect the downstream metabolism 5 . Cigarette smoke is known to contain thousands of toxic chemicals 6 . The chemicals generated from e-cig or cigarette smoking as xenobiotic chemicals in human organisms could dysregulate metabolomics profiles 7-10 and increase the risk of lung diseases, even lung cancers 11 . Commonly, cotinine is one of the significant metabolites during nicotine degradation, which has been used to identify the smoker or e-cig user 12 . We have shown circulating biomarkers are increased from e-cig users or cigarette smokers, predicting the risk of lung and heart diseases 13,14 . In this study, we have found e-cig vaping to be more associated with bioenergy synthesis (TCA cycle) than cigarette smoking, while cigarette smoking leads more active the sphingolipid pathway.
Bioenergy synthesis, including gluconeogenesis, glycolysis, and TCA cycle, is one of the major metabolic reactions in mitochondrion for generating energy among all the organs/tissues. Previous studies reported that e-cig vaping and cigarette smoking inhibited bioenergy synthesis and induced mitochondrial dysfunction 15,16 .
Mitochondrial metabolism alternation in lungs was followed by cigarette smoke exposure 15,16 ; e-cig exposure induced mitochondrial-oxidative stress and DNA damage 17,18 . Interestingly, a previous study explained that circulated PG would be metabolized into lactic acid in the liver and go through the TCA cycle 19 . However, no study is available to show the bioenergy synthesis-related circulating metabolites in e-cig users and cigarette smokers compared to healthy controls.
Sphingolipids are lipids that contained sphingoid structures and major constituents of plasma membrane 20,21 . Recent studies have shown that sphingolipid metabolites regulate pulmonary inflammatory responses, and they are essential mediators in lung cancer 20,22 . Cigarette smoke-induced accumulation of sphingolipid metabolites in the lungs is mediated with mitophagy, necroptosis, autophagy, and oxidative stress [22][23][24][25] .
Interestingly, previous reports have described dysregulated plasma sphingolipids associated with lung cancer and Chronic Obstructive Pulmonary Disease (COPD) phenotypes 26,27 . In this study, we determined the dysregulation of sphingolipid metabolites in plasma from cigarette smokers or e-cig users.
We collected plasma from healthy controls, e-cig users, and cigarette smokers for metabolites analysis. Our results showed that metabolites related to nicotine degradations are both dysregulated in the plasma from e-cig users and cigarette smokers. TCA cycle-related metabolites showed alternation only in the plasma of e-cig users, while sphingolipid metabolites presented dysregulation only in cigarette smokers' plasma.

Results
Global metabolic profiling of plasma from healthy controls, e-cig users, and cigarette smokers analyzed by UPLC-MS.
We performed global metabolites profiling based on negative and positive ion modes to identify dysregulated metabolites in plasma from cigarette smokers and e-cig users through UPLC-MS (Figure 1. A&B). During the profiling, a total of 1018 metabolite features were detected in negative ion mode, and 7244 metabolite features were detected in positive ion mode. To determine the significance of metabolomics profiling, we have applied multivariate statistical analysis via the PCA model (Figure 1.

C&D).
In the negative ion mode UPLC-MS measurement, the absolute value of metabolites between the control and cigarette smoking groups are majorly overlapped, while metabolites in the e-cig group show significantly different metabolites distribution ( Figure 1C). Interestingly, we have found an overlapped metabolites distribution in control and e-cig users' plasma from positive ion mode, and cigarette smokers showed a significant difference in dysregulated metabolites ( Figure 1D).
We also screened and identified the dysregulated metabolic pathways in cigarette smoke and e-cig groups ( Table 1). Metabolic pathways, including nicotine degradation III, serotonin degradation, and gluconeogenesis, were altered in both cigarette smoke and e-cig groups. Interestingly, the TCA cycle, D-galactose degradation, and UDP-N-acetyl-D-galactosamine biosynthesis II were found to have dysregulation in the e-cig group, while nicotine degradation IV was altered in the cigarette smoke group.

Nicotine degradation-related metabolites increased in both plasma from cigarette smokers and e-cig users.
Nicotine degradation is commonly seen after cigarette smoking and e-cig (with nicotine) vaping 28 . As expected, we have shown increased metabolites related to nicotine degradation in plasma from both e-cig users and cigarette smokers (    Other dysregulated metabolites in plasma from e-cig users or cigarette smokers. In addition to the TCA cycle or sphingolipid metabolites, we also identified other significantly dysregulated metabolites (Figures 5-6). Among the metabolites significantly dysregulated in e-cig user's plasma, we observed increased jasmonic acid in e-cig users compared to cigarette smokers and healthy controls (  We also detected significantly dysregulated metabolites from cigarette smokers' plasma compared to e-cig users and healthy controls (Figure 6). We found significantly increased metabolites such as glycolic acid, 6-hydroxy-2-naphthoic acid, 2-beta-Dglucosyle anthranilate, and budesonide, as well as significantly downregulated metabolites such as L-(-)-methionine, 2-methylthiazolidine, 4-(stearoylamino)butanoic acid, and 3-methylsulfolene (Figure 6).

Discussion
E-cig vaping has rapidly increased since it has been presumed as a safe alternative to cigarette smoke, evoking public concerns about the health risks of e-cig vaping 29 . Our previous studies have proven that both acute and chronic e-cig exposure can induce pulmonary inflammation and oxidative stress 30,31 . Many studies have shown about cigarette smoking-induced metabolic disease with dysregulated metabolites 9, 32 ; however, limited studies have elucidated the effects of e-cig on metabolic disorders which in turn to identify promising metabolite biomarkers related to potential diseases 32,33 . In this study, we have successfully identified dysregulated metabolites from the plasma of e-cig users and cigarette smokers related to nicotine degradation, TCA cycle, and sphingolipid metabolism, as well as some other metabolites introduced by e-cig aerosol and cigarette smoke.
Nicotine degradation pathways were the most commonly activated metabolic responses in cigarette smokers and e-cig users (nicotine contained e-cig vaping). When nicotine from cigarette and e-cig aerosol was inhaled into the human body, a number of metabolites are metabolized from nicotine 28 . The most important and commonly used metabolite to identify nicotine degradation is cotinine, which will be converted from 70%-80% of nicotine introduced into the human body 28 . The other cotinine-associated metabolites identified from our study, including cotinine N-oxide and trans-3hydroxycotinine. Around 35%~42% of the total cotinine will be transformed to cotinine N-oxide and trans-3-hydroxycotinine 28 . Nicotine-related metabolites, such as nornicotine and 6-hydroxynicotine, will be converted from nicotine 28 . About 10% of the nicotine will not be metabolized, and we have detected it as (S)-nicotine, and 28 . From our and other previous studies cotinine has been used as a biomarker to identify nicotine degradation, which is the commonly activated metabolism after smoking and nicotine vaping 30,34 . Furthermore, nicotine, cotinine, cotinine N-oxide, and trans-3hydroxycotinine are considered to be primary metabolites in total nicotine equivalent (TNE), which have been used as standards to validate nicotine intake 35 . Other metabolites such as nor-nicotine and 6-hydroxynicotine are less concentrated (<2%) and lower in abundance compared to TNE metabolites 28 . Hence, they are not considered as regular biomarkers for the characterization of nicotine inhalation 28,35,36 .
A previous study has identified that nor-nicotine preserves a longer half-life compared to either nicotine or cotinine 37 , and nor-nicotine was highly relevant to TNE in smokers' urine compared to health control 38 . Consistent with these data, our results confirm that although nor-nicotine or 6-hydroxynicotine are low abundance in body fluids, they are still sufficient to serve as biomarkers to identify smoking status as well as an indicator for nicotine degradation pathway activation.
The TCA cycle is a series of biochemical conversions with the generation of bioenergy, which usually occurs in mitochondrion with the products from glycolysis. A previous study has shown that either PG or PG/VG inhibited the glucose metabolism and ATP generation in airway epithelium 39 . The aerosolized PG/VG inhaled into lungs were unlikely deposited and accumulated in the bloodstream since the half-life for PG is ~4h; PG will be converted to lactic acid via alcohol dehydrogenase in the liver and then merged in the TCA cycle 19 . Our previous studies described that e-cig exposure is capable of inducing oxidative stress in the mitochondrion and dysregulation of mitochondrial complexes in lung fibroblasts 40 . Furthermore, e-cig exposure causes an increased amount of damaged mitochondrial DNA in plasma, as well as increases the risk of cardiovascular diseases 41 . In this study, we showed that most of the TCA cyclerelated metabolites are downregulated in e-cig users while there were no changes in the cigarette smokers compared to the healthy control. This is the first study to report that a series of metabolites associated with the TCA cycle are altered in e-cig users since former studies are focused on nicotine-related metabolites identified from e-cig users.
Surprisingly, we did not find significant difference between the cigarette smokers and the healthy control about the TCA cycle metabolites in plasma. It is well-known that cigarette smoke inhibits mitochondrial respiratory function and dysregulates TCA cycle 15 . The dysregulated TCA cycle-related metabolites identified from the e-cig group provide information that vaping might associate with synthetic bioenergy metabolism.
Therefore, a larger sample size is needed for the future study.
Sphingolipid metabolites are associated with lung inflammation, emphysema, and COPD 22,27,42 . Among all the known sphingolipid metabolites, sphingosine-1phosphate (S1P) and ceramide are well-studied 20 . Increased ceramide levels found in the elastase-induced mouse emphysema model and ceramides inhibitors were capable of attenuating elastase caused airspace enlargement 42,43 . We found that the cigarette smoke group showed significantly higher plasma levels of ceramide and sphingosine compared to e-cig users and the healthy control group. Since chronic cigarette smoking is shown to cause COPD/emphysema, our results are indirectly in agreement with previous studies [42][43][44] . Additionally, ceramide accumulation and the disproportion of sphingolipids were identified from the lungs of COPD/emphysema patients and smokers 45 . We have observed increased sphingosine as well, which can be converted from S1P, which is one of the downstream products of ceramide. Both ceramide and S1P were involved in the pathogenesis of various lung diseases 22 metabolic rates and weight loss, which have also been showed in cigarette smokers 48,49 . The dysregulated metabolites from cigarette smokers or e-cig users are all capable of serving as promising biomarkers for various diseases.
In conclusion, various dysregulated metabolites were identified from e-cig users or cigarette smokers when compared to healthy controls/non-smokers. Dysregulated metabolites from both e-cig users and cigarette smokers were correlated with nicotine degradation, which has been shown previously. Dysregulated metabolites related to the TCA cycle were found only in e-cig users, and altered sphingolipid metabolites were shown only in cigarette smokers; specific dysregulated metabolites identified in different groups preserve the potential as novel biomarkers for vaping and smoking associated with metabolic diseases. Further biochemical measurements of altered metabolites are required to confirm our findings in a larger cohort.

Human Subjects
Participants in this study have provided information including age, sex, gender, and ethnicity. Detailed information about cigarette smoking, e-cig vaping, and health control allowed us to categorize the condition groups as described previously 50 .

Institutional Review Board (IRB) Statement
This study was conducted at general clinical research center of the University of Rochester Medical Center with IRB approval (RSRB00064337). Participants in this study have provided information including age, sex, gender, and ethnicity. Detailed information about cigarette smoking, e-cig vaping, and health control allowed us to categorize the condition groups as described previously 50 .

Plasma samples collection
Blood samples were centrifuged at 1000 rpm for 5 min at room temperature, and plasma was collected and stored at −80 °C until UPLC-MS analysis. and used to prepare mobile phases and solutions.

Data Processing
Spectral features were extracted from the raw data using Compound Discoverer v2.1 software (Thermo Fisher Scientific, Inc., Waltham, MA) and XCMS software. This procedure included chromatographic alignment, peak picking, peak area integration, and QC-based compound area normalization. Features that eluted with the chromatographic solvent front with retention times <0.5 min in RP data sets and <0.9 min in HILIC data sets were considered unreliable due to potential ion suppression effects (PMID: 12816898). The screening criteria for differential metabolic indicators include p <0.05, fold change >2, or <0.5. Further filtering was carried out by removing features that were not present in 50% of at least one of the plasma sample groups at 10 times the baseline abundance, defined as the peak area of the sample blank run.
Welch's t-test with a Benjamini Hochberg correction was applied to cigarette smoke vs Control, and E-cig user vs control. A further selection of dysregulated metabolic pathways was based on overlap size (>6).
Change fold of metabolite was calculated based on the normalized area from positive or negative mode spectrums. In brief, normalized areas from the control group will be averaged and used as the baseline. The individual normalized area from different samples will be divided by the averaged normalized area from the control group as change folds compared to the baseline.

Statistical analysis
One-way ANOVA and student's t-test were used here to determine the significant difference in the change fold of metabolites among groups through GraphPad Prism Software version 8.0 (La Jolla, CA). Data were presented as mean ± SEM, and p < 0.05 was considered as a statistical difference.

Author Contributions
QW and IR. Conceived and designed the experiments; QW and XJ. Conducted experiments; QW and XJ. Analyzed the data; QW, XJ and IR. Wrote and revised/edited the manuscript.

Conflicts of Interest
The authors have declared that no competing interests exist.