A Model To Estimate the Sources of Tobacco-Specific Nitrosamines in

Article pubs.acs.org/crt

A Model To Estimate the Sources of Tobacco-Specific Nitrosamines in Cigarette Smoke Peter J. Lipowicz*,† and Jeffrey I. Seeman‡ †

Research, Development & Regulatory Affairs, Altria Client Services LLC, 601 East Jackson Street, Richmond, Virginia 23219, United States ‡ SaddlePoint Frontiers, 12001 Bollingbrook Place, Richmond, Virginia 23236, United States ABSTRACT: Tobacco-specific nitrosamines (TSNAs) are one of the most extensively and continually studied classes of compounds found in tobacco and cigarette smoke.1−5 The TSNAs N-nitrosonornicotine (NNN) and 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone (NNK) have been characterized by the US Food and Drug Administration (FDA) as harmful and potentially harmful constituents in tobacco products,6 and cigarette manufacturers report their levels in cigarette tobacco filler and cigarette smoke to the FDA. NNN and NNK are classified by IARC as carcinogenic to humans.7 TSNAs transfer from tobacco to smoke by evaporation driven by heat and the flow of gases down the cigarette rod. Other TSNA sources in smoke include pyrorelease, where room temperature-unextractable TSNAs are released by smoking, and pyrosynthesis, where TSNAs are formed by reactions during smoking. We propose the first model that quantifies these three sources of TSNA in smoke. In our model, evaporative transfer efficiency of a TSNA is equated to the evaporative transfer efficiency of nicotine. Smoke TSNA measured in excess of what is transferred by evaporation is termed “pyrogeneration,” which is the net sum of pyrorelease and pyrosynthesis minus pyrodegredation. This model requires no internal standard, is applicable to commercial cigarettes “as is,” and uses existing analytical methods. This model was applied to archived Philip Morris USA data. For commercial blended cigarettes, NNN pyrogeneration appears to be unimportant, but NNK pyrogeneration contributes roughly 30−70% of NNK in smoke with the greater contribution at lower tobacco NNK levels. This means there is an opportunity to significantly reduce smoke NNK by up to 70% if pyrogeneration can be decreased or eliminated, perhaps by finding a way to grow and cure tobacco with reduced matrix-bound NNK. For burley research cigarettes, pyrogeneration may account for 90% or more of both NNN and NNK in smoke.

■

INTRODUCTION Tobacco-specific nitrosamines (TSNAs) are one of the most extensively and continually studied classes of compounds found in tobacco and cigarette smoke.1−5 The TSNAs N-nitrosonornicotine (NNN) and 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone (NNK) have been characterized by the US Food and Drug Administration (FDA) as harmful and potentially harmful constituents (HPHC) in tobacco products,6 and cigarette manufacturers report their levels in cigarette tobacco filler and cigarette smoke to the FDA. NNN and NNK are classified by IARC as carcinogenic to humans.7 TSNAs form in tobacco during curing by nitrosation of alkaloids in the tobacco leaf and are released into smoke when a cigarette is combusted. A considerable amount of research has been performed over the years to determine the sources of TSNA in smoke, not all of which has resulted in complementary conclusions. There has been no uniformly accepted model that explains and predicts the origin of TSNA in mainstream tobacco smoke. Indeed, there has been little agreement in the literature about the relative importance of the various mechanisms of TSNA occurrence in smoke.8−12 NNK © 2017 American Chemical Society

pyrorelease, which includes release of NNK from a matrixbound moiety, has been recently reported as a major contributor of NNK in smoke from cigarettes having pure blends of either burley or flue-cured tobacco.8 Other researchers10,12,13 had previously concluded that pyrosynthesis of NNK and NNN is either minor or does not occur at all in commercial blended cigarettes. In the January 2017 issue of this journal, Edwards and colleagues at the FDA and Watson and colleagues at the Centers for Disease Control and Prevention (CDC) reported what they importantly characterized as “the first comprehensive assessment of TSNA amounts and smoke yields as a percentage of filler content of all four TSNAs in tobacco filler and smoke, by two smoking regimes, for a broad variety of domestic [USA] cigarette brands.”5 Levels of four TSNAs in tobacco filler and in mainstream smoke (from two machine-smoking regimens) for over 50 cigarettes were reported. Tobacco filler TSNA concentrations were determined using a previously validated Received: February 22, 2017 Published: June 27, 2017 1556

DOI: 10.1021/acs.chemrestox.7b00046 Chem. Res. Toxicol. 2017, 30, 1556−1561

Article

Chemical Research in Toxicology

evaporative transfer is a reasonable model for TSNA evaporative transfer.

aqueous ammonium acetate-room temperature solvent extraction-13C-isotope dilution liquid chromatography tandem mass spectrometry method developed by Rickert et al.14 in 2008. One next research challenge that follows this notable FDACDC joint study is to go beyond the implicit assumption that smoke TSNAs are derived solely from room temperature solvent-extractable tobacco TSNAs and determine the levels of these TSNAs in smoke derived by other mechanisms. Thus, reduction of smoke TSNAs is hindered by the incomplete understanding of sources of TSNA in smoke. In 2015, Lang and Vuarnoz8 demonstrated that more drastic extraction methods can quantify what they and others have termed “matrix-bound” TSNAs. Matrix-bound TSNAs are not quantified by room temperature solvent extraction methodologies. Furthermore, there is evidence that TSNAs can be pyrosynthesized from non-TSNA tobacco constituents during the smoking process.12 To date, there is no method to even estimate the amount of smoke TSNAs derived exclusively from solvent-extractable TSNAs in tobacco filler or from other sources such as matrix-bound TSNAs or synthesis from nonTSNA precursors during smoking. In this publication, we report the first model that will quantitatively estimate the fate of tobacco TSNAs and the role of net TSNA synthesis during smoking. We then use this model to quantify these sources of smoke NNN and smoke NNK for a wide range of commercial and research cigarettes. This model is the first step in filling important informational gaps needed in efforts to further reduce smoke TSNAs in cigarette smoke. Sidestream smoke TSNAs were not considered as there is insufficient data to test a sidestream model.

Figure 1. Chemical structures of nicotine, NNN, and NNK.

Additional evidence that TSNA transfer efficiency is well estimated by nicotine transfer efficiency is that other highboiling components of tobacco that do not thermally degrade have transfer efficiencies similar to that of nicotine. Davis et al.16 found that transfer of endogenous nornicotine and certain sterols was the same as nicotine transfer. Values of glycerin transfer17,18 are approximately what would be expected for nicotine transfer. Davis et al.19 measured dotriacontane transfer and found that it was similar to what would be expected for nicotine. There is also some direct evidence of the model’s validity. Stepanov et al.20 measured transfer of NNK using radiolabeled NNK and found values in the range of 7−13%, which are roughly consistent to nicotine transfer in US commercial cigarettes. Fischer et al.12 measured transfer of NNN and NNK added to nonfilter cigarettes and found values of 14 and 15%, respectively, and the authors noted that the transfer was within the range of nicotine transfer (11−16%) expected. An important consequence of our model is that, if the measured smoke yield of a TSNA exceeds that calculated by multiplying the nicotine transfer efficiency by the amount of room temperature-extractable TSNA in the tobacco blend, very strong evidence of net TSNA pyrogeneration has been obtained. This is illustrated in the following calculation from data for a 1R4F reference cigarette (University of Kentucky, Center for Tobacco Reference Products; https://ctrp.uky.edu) smoked under ISO conditions. Nicotine transfer efficiency is calculated as 0.73 mg of smoke nicotine/(0.759 g of filler × 17.8 mg/g of filler nicotine) = 5.4%, and NNK transfer efficiency is 75.6 ng of smoke NNK/(0.759 g of filler × 973 ng/ g of filler NNK) = 10.2%. By the model, the portion of smoke NNK that derives from evaporative transfer of NNK from the filler is 0.054 × 973 × 0.759 = 39.9 ng. The pyrogenerated smoke NNK is 75.6−39.9 = 35.7 ng, which is 47% of the measured smoke NNK. An important advantage of this model is that no internal standard is required for the cigarette smoke analyses, i.e., no 13 C, 14C, or 2H-labeled nicotine or TSNA is needed. Commercial cigarettes can be tested “as is” without the addition of secondary alkaloids or radiolabeled compounds to estimate pyrogeneration as has been necessary in other literature studies.10,20 The analysis also requires no new analytical methods.

■

MODEL TSNAs transfer from tobacco to cigarette smoke by evaporation driven by heat and the flow of gases down the cigarette rod. In our model, we first estimate the TSNAs in smoke from any chosen cigarette if they only arose by evaporative transfer of room temperature-extractable TSNAs in the tobacco blend. Any TSNA in smoke measured in excess of that amount is termed “pyrogeneration,” which is the sum of pyrorelease and pyrosynthesis minus pyrodegredation. The key assumption of this analysis is that, for any cigarette, the evaporative transfer efficiency of a TSNA from tobacco to smoke is the same as the evaporative transfer efficiency of nicotine from tobacco to smoke for the same cigarette under the same smoking conditions. Transfer efficiency, as used herein, means the ratio of the amount of the analyte in smoke divided by the amount of the analyte measured in the filler, i.e., the tobacco blend. Its value is dependent on the cigarette design and the method of machine smoking used. Nicotine transfer efficiency ranges from approximately 1% in very low tar cigarettes smoked under International Organization of Standardization (ISO) machine smoking conditions to approximately 30% in unfiltered cigarettes smoked under Health Canada (HC) conditions. Nicotine evaporatively transfers into mainstream smoke with minimal (∼3%) thermal or oxidative degradation.15 The estimated ambient vapor pressures of NNN and NNK are 0.07 and 0.01 Pa (National Library of Medicine Toxnet; https://toxnet.nlm.nih.gov), which are lower than nicotine at 5 Pa but still sufficiently volatile to be released by puff temperatures and gas flows down the cigarette rod during puffing. The similarity of the structures of nicotine and the TSNAs, as shown in Figure 1, is further evidence that nicotine

■

METHODS

We now describe the application of the model to existing data sets to estimate TSNA pyrogeneration. We analyzed four sets of archived TSNA data from Philip Morris USA (PM USA). Set 1 contains data from 38 commercial blended PM USA and Philip Morris International (PMI) cigarettes sampled in 2000 and 2001 and smoked under ISO, HC, and Massachusetts Department of Public Health (MDPH) conditions. All of the cigarettes were filtered and were a variety of 1557

DOI: 10.1021/acs.chemrestox.7b00046 Chem. Res. Toxicol. 2017, 30, 1556−1561

Article

Chemical Research in Toxicology cigarette designs and ISO tar levels. The data from set 1 were reported by Counts et al.21 Set 2 contains data from 143 samples of PM USA commercial blended cigarettes sampled in 2011 and 2012 and measured under ISO and HC conditions. Four of the cigarettes were unfiltered, and the rest were a variety of cigarette designs and ISO tar levels. Set 3 contains data from eight samples from burley tobacco grown in 2002 and tested at FTC smoking conditions. All the cigarettes were king-size full-flavor filtered cigarettes of the same design. The burley filler used varied by storage conditions. Set 4 contains data from 48 burley tobacco samples grown in 2004 and tested under FTC conditions. All the cigarettes were king-size fullflavor filtered cigarettes of the same design, which was very similar to the Set 3 cigarette design. The burley filler varied by growing location, seed variety, amount and method of application of fertilizer, and curing conditions. The analytical methods used for Set 1 were published,21 and all the data were measured using comparable validated and standardized methods. Methods and laboratories were consistent across each set of cigarettes, but several laboratories were used overall. We recognize that the use of data collected in different studies from different laboratories over time has the potential to increase the variability of the results. Nevertheless, this procedure allows for use of a large amount of data to test the model. Indeed, that using tobacco and smoke data obtained from different laboratories in this new TSNA model results in consistent conclusions suggests that the new model is robust and that the data and the conclusions are generalizable and reliable.

Figure 3. NNK transfer efficiency vs nicotine transfer efficiency.

The contributions of pyrogeneration to the NNN and NNK smoke yields are shown in Figures 4 and 5, respectively. For

■

RESULTS The results for all four data sets for NNN are shown in Figure 2. The nicotine transfer efficiency varies from approximately 1

Figure 4. Contribution of NNN pyrogeneration to NNN in smoke.

Figure 2. NNN transfer efficiency vs nicotine transfer efficiency.

to 30% because of the variety of cigarettes and smoking conditions used. For the commercial blended cigarettes, the NNN transfer efficiency is mostly scattered around the 1:1 line indicating that the observed NNN transfer efficiencies are approximately equal to nicotine transfer efficiencies. Thus, there is no pyrogeneration of NNN for these cigarettes. By contrast, the NNN transfer efficiencies observed for the all burley research cigarettes are all significantly above the 1:1 line. For one cigarette, the NNN transfer efficiency is nearly 100%. There is nearly as much NNN in the collected smoke as there is in all the tobacco filler in the cigarette. This is unambiguous evidence of pyrogeneration of NNN because the majority of the mass of smoke goes to the sidestream during static burn. The results for all four data sets for NNK are shown in Figure 3. All the data are above the 1:1 line, but the burley research cigarette transfer efficiencies are much higher than those for commercial blended cigarettes. Approximately half of the burley data shows NNK transfer greater than 100% with a maximum value near 200%.

Figure 5. Contribution of NNK pyrogeneration to NNK in smoke.

NNN in the smoke of commercial cigarettes, there is extensive scatter around 0% contribution, i.e., around the 1:1 line. This scatter reflects measurement variability, which is amplified because four measurements are used to derive the value: NNN in smoke, NNN in filler, nicotine in smoke, and nicotine in filler. The scatter is greater when the percent pyrogeneration is lower because percent pyrogeneration is a value calculated from the difference of two values, each with variability. When the difference is small, the resulting scatter becomes large. Irrespective of the data scatter, it is clear that pyrogeneration, if it exists at all, is negligible for NNN for commercial blended 1558

DOI: 10.1021/acs.chemrestox.7b00046 Chem. Res. Toxicol. 2017, 30, 1556−1561

Article

Chemical Research in Toxicology cigarettes. However, for burley research cigarettes, NNN pyrogeneration is extensive. The contribution is as high as 90% at low filler NNN levels and trends downward to approximately 50% as filler NNN increases. NNK pyrogeneration in commercial cigarettes ranges from 70% for low NNK filler values to 30% at higher NNK filler values. NNK pyrogeneration in burley research cigarettes ranges from 95% at low NNK filler to 70% at higher NNK filler values. NNN in smoke from pyrogeneration is correlated to NNK in smoke from pyrogeneration for all the data sets. The correlation coefficients are 0.44, 0.46, 0.96, and 0.89 for sets 1−4, respectively. The correlation is thus very strong for the burley research cigarettes.

(conventionally measured NNK concentration in cigarette filler + predicted matrix‐bound NNK concentration in cigarette filler) × nicotine transfer efficiency × cigarette filler weight = NNK in smoke

(1) 8

For example, for sample TL001 in the Lang and Vuarnoz data, eq 1 becomes (145 ng of NNK/g + predicted matrix-bound NNK ng/g) × 0.221 × 0.675 g = 95 ng of NNK. The predicted matrix-bound NNK is 493 ng/g, and the measured value is 498 ng/g. The results of this calculation applied to all the data are plotted in Figure 6. With the exception of one data point, the

■

DISCUSSION We now compare the pyrogeneration results using our new TSNA model to the pyrogeneration calculated from the data of published studies in Table 1. The recent study by Edwards et Table 1. Comparison of Study Results with Published Studies mean contribution to smoke from pyrogeneration data set 1 1 1 2 2 3 4 Lang and Vuarnoz8 Edwards et al.5 Edwards et al.5

tobacco filler

smoking conditions

samples (N)

NNN

NNK

blended blended blended blended blended burley burley burley

ISO HC MDPH ISO HC FTC FTC HC

38 38 38 143 143 8 48 50

11% −12% −3% 11% 3% 75% 71% no data

52% 41% 48% 61% 57% 84% 84% 73%

various various

ISO HC

52 52

−6% −18%

55% 49%

Figure 6. Predicted matrix NNK vs measured matrix-bound NNK.

agreement is very good with R2 = 0.93 for the 1:1 line. Thus, our model predicts the Lang and Vuarnoz8 data. This data is consistent only with pyrorelease and not pyrosynthesis, but other studies must be conducted to confirm that pyrorelease is the dominant mechanism of pyrogeneration. Our results show that NNN is not pyrogenerated in commercial blended cigarettes even though such cigarettes contain burley tobacco. Why does the burley not contribute some pyrogenerated NNN in the smoke? This is a valid question raised by our findings. There are other tobacco varieties in the blended cigarette that would dilute any effect of burley, but the dilution is not so great as to obviate the expected effect. Additionally, aged burley as would be used in commercial cigarettes was not included in our data sets. The testing was done on freshly cured burley because the original data was collected to test agronomic means to lower burley TSNAs. We did find two samples in archived data for aged burley. In 2003, there was a test of a sample of cigarettes containing the 1999 and 2000 burley crops. For these two samples, the pyrogeneration contributions for NNN were only 39 and 38% respectively, which are lower than any of the data for the unaged burley shown in Figure 5. This observation suggests that pyrogeneration of NNN is reduced over the time scale (years) relevant to burley aging. We hypothesize that room temperature-unextractable matrixbound NNN slowly becomes room temperature-extractable as tobacco ages or that NNN precursors are slowly converted to room temperature-extractable filler NNN over time. These hypotheses would also explain why pyrogeneration is lower when room temperature-extractable NNN levels are high in the tobacco filler. A long-term experiment tracking both smoke and room temperature-extractable filler TSNAs over time using our model to quantify pyrogeneration would be a test of these

al.5 tested US commercial cigarettes including an all flue-cured cigarette and two research blended cigarettes. Their measurements of mean transfer efficiency across all the cigarettes studies under ISO smoking conditions were 7, 6.6, and 15.6% for nicotine, NNN, and NNK, respectively. The corresponding values under HC conditions were 18, 15.2, and 35%. Our model was applied using these mean transfer efficiencies to determine the contribution of TSNA pyrogeneration. The results were similar to those of sets 1 and 2 for NNK, but the NNN pyrogeneration was slightly more negative. The study by Lang and Vuarnoz8 used research cigarettes of which only the burley results are summarized here. They measured matrixbound NNK using an exceptionally vigorous extraction that included autoclaving the sample in water at 130 °C and 275 kPa for multiple hours. They found significantly more matrixbound NNK than NNK measured by conventional methods. Their result for the mean contribution of NNK pyrogenerated was comparable to the results for burley research cigarettes in our study. One of their samples showed 177% NNK transfer to smoke, which is similar to the maximum value shown in Figure 3. To further test our model, we applied our model to their data to determine if it could predict the amount of matrix-bound NNK for each cigarette tested. Equation 1 was solved to determine predicted matrix-bound NNK. 1559

DOI: 10.1021/acs.chemrestox.7b00046 Chem. Res. Toxicol. 2017, 30, 1556−1561

Article

Chemical Research in Toxicology

tine; TSNA, tobacco-specific nitrosamine; PMI, Philip Morris International; PM USA, Philip Morris USA

hypotheses and a further demonstration of the value of the new model. Indeed, such data sets may already exist, and this model can be applied to them retrospectively much as we have done with our data sets. Finally, the high degree of correlation between NNN and NNK pyrogeneration for burley research cigarettes suggests that there may be a common mechanism for NNN and NNK pyrogeneration.

■

(1) Hecht, S. S., and Szabo, E. (2014) Fifty years of tobacco carcinogenesis research: From mechanisms to early detection and prevention of lung cancer. Cancer Prev. Res. 7, 1−8. (2) Yershova, K., Yuan, J.-M., Wang, R., Valentin, L., Watson, C., Gao, Y.-T., Hecht, S. S., and Stepanov, I. (2016) Tobacco-specific nnitrosamines and polycyclic aromatic hydrocarbons in cigarettes smoked by the participants of the shanghai cohort study. Int. J. Cancer 139, 1261−1269. (3) Zarth, A. T., Upadhyaya, P., Yang, J., and Hecht, S. S. (2016) DNA adduct formation from metabolic 5′-hydroxylation of the tobacco-specific carcinogen n′-nitrosonornicotine in human enzyme systems and in rats. Chem. Res. Toxicol. 29, 380−389. (4) Carlson, E. S., Upadhyaya, P., and Hecht, S. S. (2016) Evaluation of nitrosamide formation in the cytochrome p450-mediated metabolism of tobacco-specific nitrosamines. Chem. Res. Toxicol. 29, 2194−2205. (5) Edwards, S. H., Rossiter, L. M., Taylor, K. M., Holman, M. R., Zhang, L., Ding, Y. S., and Watson, C. H. (2017) Tobacco-specific nitrosamines in the tobacco and mainstream smoke of U.S. Commercial cigarettes. Chem. Res. Toxicol. 30, 540. (6) Food and Drug Administration (2012) Harmful and potentially harmful constituents in tobacco products and tobacco smoke; established list. Federal Regist 77, 20034−20037. (7) IARC (2007) IARC monographs on the evaluation of carcinogenic risks to humans. Volume 89. Smokeless tobacco and some tobacco-specific nitrosamines. http://monographs.iarc.fr/ENG/ Monographs/vol89/mono89.pdf, accessed on December 7, 2016. (8) Lang, G., and Vuarnoz, A. (2015) Matrix-bound 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone in tobacco: Quantification and evidence for an origin from lignin-incorporated alkaloids. J. Nat. Prod. 78, 85−92. (9) Piadé, J. J., Wajrock, S., Jaccard, G., and Janeke, G. (2013) Formation of mainstream cigarette smoke constituents prioritized by the world health organization − yield patterns observed in market surveys, clustering and inverse correlations. Food Chem. Toxicol. 55, 329−347. (10) Moldoveanu, S., and Borgerding, M. (2008) Formation of tobacco specific nitrosamines in mainstream cigarette smoke; part 1, FTC smoking. Beitr. Tabak. Int. 23, 19−31. (11) Baker, R. R. (2006) Smoke generation inside a burning cigarette: Modifying combustion to develop cigarettes that may be less hazardous to health. Prog. Energy Combust. Sci. 32, 373−385. (12) Fischer, S., Spiegelhalder, B., Eisenbarth, J., and Preussmann, R. (1990) Investigations on the origin of tobacco-specific nitrosamines in mainstream smoke of cigarettes. Carcinogenesis 11, 723−730. (13) Spiegelhalder, B., and Bartsch, H. (1996) Tobacco-specific nitrosamines. Eur. J. Cancer Prev. 5, 33−38. (14) Wu, J., Joza, P., Sharifi, M., Rickert, W. S., and Lauterbach, J. H. (2008) Quantitative method for the analysis of tobacco-specific nitrosamines in cigarette tobacco and mainstream cigarette smoke by use of isotope dilution liquid chromatography tandem mass spectrometry. Anal. Chem. 80, 1341−1345. (15) Yu, J., Taylor, L. T., Aref, S., Bodnar, J. A., and Borgerding, M. F. (2006) Influence of puffing parameters and filter vent blocking condition on nicotine fate in a burning cigarette. Part 1. Full flavor cigarettes. Beitr. Tabak. Int. 22, 185−195. (16) Davis, D. L., Bush, L. P., Grunwald, C., and Cornelius, P. L. (1977) Transfer of exogenously applied and endogenous alkaloids and sterols from tobacco to its smoke condensate. J. Agric. Food Chem. 25, 752−756. (17) Liu, C. (2004) Glycerol transfer in mainstream cigarette smoke. Beitr. Tabakforsch. Int. 21, 111−116. (18) Rodgman, A. (2002) Some studies on the effects of additives on cigarette mainstream smoke properties. Ii. Flavorants. Beitr. Tabakforsch. Int. 20, 83−103.

■

CONCLUSIONS We herein have proposed the first model to predict the sources of smoke TSNAs in cigarette smoke. The assumption made in this model, that the efficiency of TSNA transfer is modeled by the efficiency of nicotine transfer, is supported by literature data. When this model is applied to tobacco and smoke data sets, the results show significant levels of pyrogeneration for some TSNAs and some cigarette types but not in others. These results are consistent across data sets and are comparable to studies in the scientific literature. Using this model, several specific tobacco smoke conclusions can be proposed. For commercial blended cigarettes, NNN pyrogeneration appears to be unimportant, but NNK pyrogeneration contributes roughly 30−70% of NNK in smoke with the greater contribution at lower tobacco NNK levels. This means there is an opportunity to significantly reduce smoke NNK by up to 70% if pyrogeneration can be decreased or eliminated, perhaps by finding a way to grow and cure tobacco with reduced matrix-bound NNK. For burley research cigarettes, pyrogeneration is the dominant source of both NNN and NNK in smoke and may account for 90% or more of both NNN and NNK in smoke. This conclusion suggests that burley tobacco is a likely source of pyrogeneration of TSNAs and that efforts to reduce TSNA pyrogeneration should focus first on the burley component of cigarette tobacco.

■

REFERENCES

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected] or SciencePublications@ altria.com; phone: (804) 335-2321. ORCID

Peter J. Lipowicz: 0000-0003-1100-1079 Jeffrey I. Seeman: 0000-0003-0395-2536 Funding

This research was funded by Altria Client Services LLC and Philip Morris USA. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS We dedicate this paper to the memory of our former colleague and friend Edward B. (Ted) Sanders (November 10, 1940−July 4, 2016). The authors acknowledge the editorial assistance of Eileen Y. Ivasauskas of Accuwrit Inc., and the technical assistance of Lucy Joseph and Thomas Gannon of Altria Client Services LLC. J.I.S.’s contributions to this research were performed while he was a consultant to PM USA in 2005.

■

ABBREVIATIONS FTC, Federal Trade Commission; HC, Health Canada; ISO, International Organization of Standardization; MDPH, Massachusetts Department of Public Health; NNK, 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone; NNN, N-nitrosonornico1560

DOI: 10.1021/acs.chemrestox.7b00046 Chem. Res. Toxicol. 2017, 30, 1556−1561

Article

Chemical Research in Toxicology (19) Davis, B. R., Houseman, T. H., and Roderick, H. R. (1973) Studies of cigarette smoke transfer using radioisotopically labelled tobacco constituents: Part iii: The use of dotriacontane-16,17−14C as a marker for the deposition of cigarette smoke in the respiratory system of experimentai animals. Beitr. Tabakforsch. 7, 148−153. (20) Stepanov, I., Knezevich, A., Zhang, L., Watson, C. H., Hatsukami, D. K., and Hecht, S. S. (2012) Carcinogenic tobaccospecific N-nitrosamines in US cigarettes: Three decades of remarkable neglect by the tobacco industry. Tob. Control 21, 44−48. (21) Counts, M. E., Morton, M. J., Laffoon, S. W., Cox, R. H., and Lipowicz, P. J. (2005) Smoke composition and predicting relationships for international commercial cigarettes smoked with three machinesmoking conditions. Regul. Toxicol. Pharmacol. 41, 185−227.

1561

DOI: 10.1021/acs.chemrestox.7b00046 Chem. Res. Toxicol. 2017, 30, 1556−1561