Role of diabetes in lung injury from acute exposure to electronic cigarette, heated tobacco product, and combustible cigarette aerosols

Patients with diabetes are more vulnerable to the detrimental respiratory effects of combustible cigarette smoke (CS) when compared to the general population. Electronic cigarettes (ECIG) and heated tobacco products (HTP) are marketed as less harmful alternatives to CS. In this study, we compared the effects of acute ECIG, HTP and CS exposure on the lungs of type II diabetes versus non-diabetic mice in an animal model Type II Diabetic (Diab) and Non-Diabetic (Non-Diab) mice were divided into Control, ECIG, HTP and CS groups. Animals were exposed for 6 hrs./day to either air, ECIG, HTP or CS for seven days. Lung injury was determined by a) histopathology, b) wet to dry ratio, c) albumin concentration in bronchoalveolar lavage uid, d) expression of TNF-α, IL-6, and IL-1 β, e) reactive oxygen species production (ROS), and f) assessment of cellular apoptosis. and in diabetic mice exposed to of for upregulated in Diab + HTP and Diab + ECIG but in Non-Diab + HTP. + less Non-Diab + CS and weakly noted in ECIG + leak observed Quantitative Polymerase Chain Reaction (q-PCR) was utilized to assess inammatory mediators’ transcriptional levels. RNA was extracted from the lung using the TRIzol method (Invitrogen, Carlsbad, CA, USA). Briey, 1 ml of TRIzol reagent was used per 50–100 mg of a tissue sample, followed by chloroform extraction. RNA samples were precipitated and stored at -80 o C. RNA was quantied using a 260/280 nm absorbance ratio method. Total RNA (5 µg) was reverse-transcribed into rst strand cDNA. Real time-PCR was performed using the iCycler (Bio-Rad Laboratories, CA, USA) with SYBR Green. Specic primers (Tib-Molbiol, Berlin, Germany) were used to assess the expression of the inammatory mediators in these tissues Fw IL-6:


Introduction
Combustible cigarette smoke (CS) contains toxicants that are harmful to human health and that result in organ damage and early death [1]. Chronic obstructive pulmonary disease (COPD), for example, is associated with CS and exposure and was projected to become the third most common cause of death by 2020 [2,3]. Electronic cigarettes (ECIG) and heated tobacco products (HTP) are marketed as less harmful or not harmful alternatives to combustible cigarettes, and sometimes as effective tools for smoking cessation [4]. ECIGs are battery-powered devices that heat and vaporize a liquid solution that usually contains nicotine, propylene glycol, glycerin, and other ingredients [5]. To attract more users, ECIG manufacturers are now adding a wide range of avors to the liquid solutions [6]. Some ECIG users are advocating for ECIG exemption from indoor-smoking policies [7]. Despite the inconclusive data on their long-term health effects and population-wide impact on tobacco use patterns, ECIGs continue to penetrate the global markets with increased sales and consumption [8,9].
Another category of tobacco products is known as heat-not-burn or heated tobacco products (HTP).
These products involve electrically heating a tobacco ller similar to that found in combustible cigarettes.
HTPs produced by Phillip Morris International are marketed under the IQOS brand, which has been cleared for marketing as a Modi ed Risk Tobacco Product by the US Food and Drug Administration [10].
IQOS uses a battery-powered heating blade to reportedly heat the ller to a temperature of 350 o C, producing an aerosol that mimics that of a combustible cigarette, but with reduced toxicant content [11].
The adverse effects of HTP use are not thoroughly tested and scienti cally established.
Type II diabetes is a major global health problem and it is recognized that cardiovascular and respiratory complications are augmented in cigarette smokers [12,13]. Smoking is associated with signi cantly higher rates of mortality and morbidity in diabetic than in non-diabetic populations [14]. To the best of our knowledge, the effects of ECIG or HTP on subjects with comorbid medical conditions like diabetes have not been examined. In this study, we studied whether ECIG or HTP exposure exaggerates the respiratory effects in diabetics when compared to non-diabetics in an animal model. In addition, the study compared the effects of ECIG and HTP to the detrimental effects of CS exposure in diabetic and non-diabetic mice.

Animals:
All animal procedures were conducted according to the National Institutes of Health (NIH) Guide for the Care and Use of Laboratory Animals and were approved by the Institutional Animal Care and Use Committee at the American University of Beirut. Four-month-old non-obese type II female diabetic FVB-Tg (Ckm-1GF1R*K1003R)1Dlr/J mice weighing 22 to 25 grams (Diab) and age matching (Non-Diab) FVB-Tg adult female mice were kept in a temperature-controlled room and on a 12/12 hrs dark/light cycle and had standard chow and water access except when mice were placed in the exposure apparatus. Female mice were selected because of their established vulnerability to CS when compared to males. The ECIG, HTP, CS or exposure apparatus (ONARES, CH Technologies, USA) consists of a smoke generator, mixing/conditioning chamber and a twelve "nose only" rodent exposure carousel. One port was dedicated for sampling analysis, and the remaining 11 ports were used for animal exposure. Non-Diabetic mice were each divided into four groups: Control Non-Diab, Non-Diab + ECIG, Non-Diab + HTP and Non-Diab + CS.
Similarly, diabetic mice were divided into: Control Diab, Diab + ECIG, Diab + HTP and Diab + CS. Before placement in the exposure carousel, animals were acclimated to retainers for one week. As described previously [15], mice were then positioned in retainers and placed into the compartments of the carousel and received a continuous ow of air or air mixed with ECIG or HTP or CS aerosols. Mice were exposed for seven consecutive days, twice daily (am session and pm session). Each session lasted for 3 hrs.
At the conclusion of the experiment, animals were anesthetized, and the trachea was cannulated with polyethylene tubing. Animals were exsanguinated by severing the aorta. The diaphragm was dissected to allow free lung expansion. The left lung was clipped, and the right lung was then lavaged three times by slowly instilling 0.5 ml of PBS (Ca ++ and Mg ++ free, 37 °C) and then gradually aspirating the lavaged uid. The lower lobe of the left lung was excised for pulmonary water content evaluation. The upper lobe of the left lung was xed in formalin for pathology examination and TUNEL assay. The remaining right lung lobes were individually frozen in liquid nitrogen for RNA extraction.
Aerosol generation and exposure protocol ECIG aerosol was generated using pre-lled V4L CoolCart (3.5 Ohm, 18 mg/mL labeled nicotine concentration) cartomizer cartridges, connected to an automatically activated 4.2 V Vapor Titan Soft Touch battery. The cartridges and batteries were purchased from an internet vendor (www.vapor4life.com) and passed a quality control test before their use. To ensure a steady aerosol generation during the 3-hour exposure sessions, cartridges and batteries were replaced every 30 min. ECIG puff parameters were set at 4 s puff duration, 1.2 L/min ow rate, and 14 s inter-puff interval, as in Talih et al. [16].
HTP aerosols were generated utilizing the 'I quit smoking system' (IQOS). The tobacco sticks, holders and the batteries were purchased from the internet vendor (www.IQOS.COM). The holder was fully charged before its use and the tobacco sticks were replaced every 6 min. The puff duration was 2 sec and the time in-between the puffs was 58 sec (i.e. 1 puff for 2 seconds every 1 minute). The puff depth was 35 ml. CS was generated from 3R4F cigarettes (University of Kentucky, Lexington, KY) with 0.9 mg TPM, 9.4 mg tar, and 0.726 mg nicotine per cigarette. The machine was set at one puff every minute with a duration of two seconds per puff and a volume of 35 ml per puff as well. With the current setup, cotinine levels were sampled from different groups. CS animals displayed levels that ranged from (40 to 55 ng/ml) as compared to ECIG and HTP groups where levels were consistently > 100 ng/ml. Cotinine levels signi ed adequate exposure to HTP and ECIG when compared to CS. With the current utilized protocol, nondiabetic animals would be exposed to levels that will result is measurable lung injury (17). The differential effects on Diabetic mice can then be ascertained.

Lung histology:
The upper lobe of the right lung was xed in 10% buffered formalin, embedded in para n, sectioned and stained with hematoxylin and eosin (H&E). The degree of lung injury was determined based on the following histological features: alveolar interstitial wall edema, congestion, degree of in ammatory cell in ltration, and intra-alveolar edema.

Wet-to-dry lung weight
The left lower lobe was weighed and then placed in a 95 °C oven to dry for two days. The dried tissue was weighed, and the wet-to-dry ratio (W/D) was then calculated.

Albumin level
The concentration of albumin in the BALF was determined by an immune-turbidimetric assay as described before (16). Agglutination, caused by antigen/antibody complexes, was measured turbidimetrically at the clinical chemistry laboratory of the American University of Beirut Medical Center using a Hitachi 912 Autoanalyser (Roche Diagnostics, Basel, Switzerland).
Transcription expression pro le of in ammatory mediators IL-1, IL-6, and TNF-α Quantitative Polymerase Chain Reaction (q-PCR) was utilized to assess in ammatory mediators' transcriptional levels. RNA was extracted from the lung using the TRIzol method (Invitrogen, Carlsbad, CA, USA). Brie y, 1 ml of TRIzol reagent was used per 50-100 mg of a tissue sample, followed by chloroform extraction. RNA samples were precipitated and stored at -80 o C. RNA was quanti ed using a 260/280 nm absorbance ratio method. Total RNA (5 µg) was reverse-transcribed into rst strand cDNA. Real time-PCR was performed using the iCycler (Bio-Rad Laboratories, Hercules, CA, USA) with SYBR Green. Speci c primers (Tib-Molbiol, Berlin, Germany) were used to assess the expression of the in ammatory mediators in these tissues (IL-1β: Fw CACCTCTCAAGCAGAGCACAG, Rw GGGTTCCATGGTGAAGTCAAC; IL-6: Fw TCCTACCCCAACTTCCAATGCTC, Rw TTGGATGGTCTTGGTCCTTAGCC; TNF-α: Fw AATGGGCTCCCTCTCATCAGTTC, Rw TCTGCTTGGTGGTTTGCTACGAC). PCR products and their corresponding melting temperatures were analyzed using the iQ5 Optical System Software" (Bio-Rad Laboratories). Correction for loading was achieved by subtracting for local background and normalizing against the cDNA levels of the GAPDH housekeeping gene (GAPDH: Fw GTATTGGGCGCCTGGTCACC, Rw CGCTCCTGGAAGATGGTGATGG).

Assessment of Reactive Oxygen Species (ROS) production:
Dihydroethidium (DHE) (Invitrogen, Molecular Probes, USA) (10 µmol/L dissolved in DMSO) was applied to lung sections and incubated in a light-protected humidi ed chamber at 37 0 C for 15 min. Fluorescent images of ethidium-stained tissue were scanned for a signal with a scanning confocal microscope (LSM 710, Zeiss, Germany). Ethidium bromide was excited at 488 nm, and uorescence was detected at 560 nm long-pass lter. Mean uorescence intensity of the digitalized image was measured with Image software [National Institutes of Health (NIH), Bethesda, MD] for quanti cation as previously described [16].

Assessment of apoptosis -TUNEL assay
The terminal deoxynucleotidyl transferase-mediated dUTP nick-end labeling (TUNEL) assay was used to monitor the extent of DNA fragmentation. Fluorescein-conjugated dUTP incorporated in nucleotide polymers was detected and analyzed using uorescence microscopy (Zeiss LSM 710, Germany). Positive and negative controls were used to verify the speci city of the TUNEL assay. TUNEL-positive nuclei were distinguished from the TUNEL-negative nuclei by counterstaining with Hoechst 33258.

Results
One animal in the Diab + CS expired on day six of CS exposure. All other animals in the different groups (Diab and Non-Diab) survived the experiment. CS animals displayed limited decrease in body weight that was statistically non-signi cant.
Lung histology H&E staining of lung sections from Non-Diab + ECIG essentially displayed "normal" alveolar structure with isolated foci of in ltration of in ammatory cells (Figs. 1A and 1B). Diab + ECIG demonstrated worsening in ltration of in ammatory cells (macrophages and lymphocytes) around the bronchioles and into the lung parenchyma. Edematous and thickened alveolar walls were also noted ( Fig. 1C and 1D).
Non-Diab + HTP lungs revealed patches of alveolar wall edema and capillary congestion (Figs. 1E and 1F). These ndings were signi cantly worse in Diab + HTP when compared to Non-Diab + HTP. Non-Diab + CS and Diab + CS lungs demonstrated signi cant acute lung injury in both groups. Diab + CS lung injury was signi cantly worse when compared to Non-Diab + CS with worsening edematous and thickened alveolar walls, in addition to signi cant in ltration by in ammatory cells (Fig. 1).

W/D Ratio and Albumin leak
There was signi cant (p < 0.001) increase in the W/D ratio, when compared to Control, for animals exposed to CS (Non-Diab and Diab). HTP and ECIG for both diabetic and Non-Diab animals demonstrated no any signi cant increase in W/D ratio when compared to Control. As for albumin leak into the BALF, no signi cant increase in albumin leak was noted in ECIG, but both HTP animals (Non-Diab and Diab) demonstrated signi cant (p < 0.001) albumin leak into the BALF when compared to Control In ammatory Mediators TNF-α Exposure to ECIG or HTP did not result in an increase in TNF-α in the Non-Diabetic group. On the other hand, Diab + ECIG and Diab + HTP demonstrated a signi cant increase in TNF-α expression when compared to Non-Diab groups and Control ( Fig. 2A). CS exposure resulted in a signi cant increase TNF-α in expression in Non-Diabetic and diabetic groups as well. The intensity of TNF-α expression was signi cantly higher in Diab group when compared to Non-Diab.

IL-1β:
Exposure to ECIG resulted in increased IL-1β expression in both Non-Diab and Diab groups when compared to Control. Exposure to HTP, however, resulted in increased IL-1β expression only in the diabetic group (Fig. 2B). Non-Diab + HTP IL-1β expression was similar to Control. CS exposure was associated with a surge IL-1β expression in both groups (Diab and non-Diab). The intensity of IL-1β expression of CS groups was signi cantly higher when compared to ECIG and HTP for Diab and Non-Diab groups respectively.

IL-6
The intensity of IL-6 expression was signi cantly increased in Diab + ECIG when compared to Non-Diab and Control (Fig. 2C). HTP exposure did not result in any signi cant increase in the IL-6 expression for Diabetic and Non-Diabetic groups. As described previously, CS exposure, IL-6 expression was signi cantly increased in both groups when compared to Control, and the expression of the Diab + CS was signi cantly greater than Non-Diab + CS.
Oxidative stress and apoptosis HTP exposure did not result in any signi cant OS in Diab and Non-Diab animals. Diabetic mice exhibited a signi cant increase in ROS production in ECIG and CS groups when compared to Control and Non-Diabetic groups. CS-induced ROS were signi cantly higher in Diab + CS when compared to Non-Diab + CS. In the Non-Diab mice groups, OS was only observed in CS + Non-Diab when compared to Control (Fig. 3).
Similarly, a signi cant increase in the number of TUNEL positive, apoptotic nuclei, indicating cellular death, was detected in Diab + ECIG + and Diab + CS mice when compared to Control. The number of TUNEL-positive apoptotic nuclei was signi cantly higher with CS compared to ECIG exposure (Fig. 4).
HTP exposure did not result in any signi cant increase in apoptosis in Diab + HTP nor Non-Diab + HTP.

Discussion
This study was designed to examine how diabetes mediates lung injury on exposure to ECIG, HTP, or CS aerosols. We found that when exposed to ECIG aerosols the lungs of diabetic mice demonstrated a signi cant increase in albumin leak, a surge in the expression of IL-1β, IL-6 and TNF-α and worsening OS and apoptosis. In addition, histopathological assessment of the lung tissue demonstrated worsening alveolar edema and in ltration by in ammatory cells. Non-Diabetic mice, exposed to ECIG aerosols, demonstrated lower severity of lung injury with ECIG exposure. IL-1β expression was the only tested in ammatory marker that was increased with ECIG exposure, but no signi cant surge was observed in IL-6, ROS production or apoptosis when compared to Control.
As for HTP exposure, there was no detectable increase in in ammation or lung injury in the Non-Diabetic group. Diab + HTP, however, was associated with a signi cant increase in TNF-α, IL-1β and when compared to Control. HTP exposure was, however, associated with signi cant albumin leak into the BALF. When HTP groups (Diab and not Diab) were compared to ECIG and CS groups, we concluded that HTP exposure was associated with the lowest injury to the lungs in this animal model. As for CS exposure, there was an increase in ROS, apoptosis and in ammatory mediators' expression that was more pronounced in the diabetic group when compared to the Non-Diabetic group with similar CS exposure. The above ndings suggest that diabetic mice are more vulnerable when exposed to external irritants such as those present in and ECIG, HTP or CS aerosol.
The association between comorbid health conditions and their augmentation of lung injury and in ammatory response secondary to illnesses and infections was reported in human and animal studies [18,19]. Diabetes, in particular, is known to exacerbate systemic in ammation as evidenced by "priming the lung" with higher serum high mobility group box 1 (HMGB1), higher cytokine levels and enhanced lung injury in a systemic in ammation animal model [20,21]. In addition, diabetes augments oxidant and pro-in ammatory cytokines expression in lung, liver, and heart in a septic animal model (22). The negative impact of diabetes on the lungs is observed in human epidemiological studies as higher fasting insulin and insulin resistance is associated with worsening pulmonary function testing [23,24]

Conclusion
This study highlights the importance of comorbid health conditions in amplifying inhalational lung injury in ECIG, HTP and combustible tobacco products consumers. While limited or no ndings were noted in the Non-Diabetic mice exposed to HTP or ECIG for one week of exposure, signi cant lung injury was noted with ECIG exposure and to a lesser extent with HTP resulting in alveolar injury, increased OS and apoptosis and a signi cant surge in some of the in ammatory markers in diabetic mice. As more novel tobacco products are introduced into the global markets, it is critical to assess their harm and health effects not only on healthy subjects but also to extend the assessment to subjects with additional comorbid health conditions, such as diabetes.     ROS detection in lung tissues. Five μm thickness slides of lung sections for Control, ECIG and CS from Non-Diabetic and diabetic mice were incubated with Dihydroethidium (DHE), and uorescent images of ethidium-stained tissue were analyzed. ROS levels were signi cantly higher in ECIG and CS and exposed diabetic mice when compared to ECIG and CS Non-Diabetic mice exposure respectively.