Tobacco Documents Online

CONFIDENTIAL ON THE SELECTIVE REDUCTION OF TOBACCO SPECIFIC N-NITROSAMINES FROM CIGARETTE SMOKE A RESEARCH PROPOSAL November 3, 1980 Dietrich Hoffmann, Ph.D. American Health Foundation Valhalla, New York 10595 Copy 3

TABLE OF CONTENTS SUMMARY A. AIM OF STUDY B. PRELIMINARY STUDIES C. METHODS 1. New Analytical Method for TSNA 2. NNK in Cigarette Smoke a. Transfer Rate b. Pyrosynthesis GC-Analysis of Tobacco Alkaoids 3. Selective Reduction of TSNA from Cigarette Smoke D. EXPERIMENTAL CIGARETTES E. TIME SCHEDULE - STAFF NEEDS 7 F. REFERENCES G. PHF - PUBLICATIONS ON N-NITROSAMINES ADDENDUM: Reprints of AHF Publications on Subject ESTIMATED BUDGET r 8

S [JMMARY Cigarette smoke contains tobacco specific N-nitrosamines (0.8 - 5 ug/cigarette). These TSNA are formed exclusively from the nicotiana alkaloids during tobacco processing and smoking. Three (3) major individual TSNA have been identified so far, namely N'-nitrosonornicotine (NNN), 4-(N-methyi-N-nitrosamino)-1-(3-pyridyl)-1-butanone (NNK) and N'-nitrosoanatabine (NAT). This research proposal is designed toward the selective reduction of TSNA in cigarette smoke and includes three parts: 1) the development of a rapid analytical GC-TEA method for TSNA, 2) studies on the transfer rate into smoke, the pyrosynthesis of NNK f rom nicotine and the inhibition of this pyrosynthesis, and 3) the selective reduction of TSNA in cigarette smoke by filter tips with and without perforation. Since TSNA are biologically highly active smoke constituents, their reduction in cigarette smoke is an essential facet of the development of cigarettes with significantly lowered biological activity. -

ON THE SELECTIVE REDUCTION OF TOBACCO SPECIFIC N-NITROSAMINES FROM CIGARETTE SMOKE A RESEARCH PROPOSAL A. AIM OF STUDY In the experimental setting, the tobacco specific N-nitrosamines (TSNA) represent one group of biologically important tobacco constituents (1,2). These agents are formed from nicotiana alkaloids by N-nitrosation during tobacco processing (aging, curing and fermentation) and during tobacco smoking. So far, three TSNA were identified. These are N'-nitrosonornicotine (NNN), 4-(N-methyl-N-nitros- amino)-1-(3-pyrid.yl)-1-butanone (NNK), and N'-nitrosoanata- bine (NAT). Nornicotine Nicotine Anatabine

G:- -17 The total concentration of the three TSNA in tobacco _ = ranges from 0.8 - 80 ppm and in the smoke of one cigarette between 0.4 - 5 ug (3,4). In addition, it appears that tobacco and its smoke also contain traces of N'-nitrosoana- basine basine (NAB). It is the aim of this study: O 1, to develop an improved method for TSNA analysis which is amenable to in depth study on these tobacco consti- tuents, -- - 2. to determine the transfer rate into smoke and the pyro- synthesis of NNK in cigarettes made entirely of flue- cured tobacco, Burley tobaco and of a U.S. cigarette tobacco blend, and _ _ 3. to explore the methods which lead'to a selective reduc- tion of TSNA in cigarette smoke. B. PRELIMINARY STUDIES Earlier reports have been reviewed during the last 2 years (1,2) and recent papers have delineated the metabolic activation of NNN (5), and NNK (6). Recently, we also pre- sented preliminary data which indicate the possibility to reduce TSNA selctively from cigarette smoke by filtration (5). The same study also suggests that filter perforation exerts a major influence on the selective removal of TSNA from the smoke by filter tips.

C. METHODS 1. New Analytical Method for TSNA So far, we employed an analytical method for TSNA in tobacco and in smoke which utilizes HPLC-separation followed by detection with a thermal energy analyzer (TEA; 3). How- ever, this method is too time consuming for an in-depth study on the selective filtration of TSNA. We plan, therefore, to utilize GC-TEA in the final step of the analysis. This requires lining with glass of-the injection port and the entire GC up to the pyrolysis unit of the TEA. The GC-column will be directly connected with the pyrolysis tube. With this set up, we avoid thermal degradation of some of the TSNA, an occurrence which so far could not be avoided in the case of N-nitrosamines with relatively low volatility. We will also explore the suitable GC-packings for the separation of NNN,, NNK, NAT, and NAB (Z- and E-isomers are not separated by GC at elevated temperatures). For quantitative work, we will employ NNN-2'-14C (8), as an internal standard. - 2.- NNK in Cigarette Smoke: Transfer Rate and Pyrosynthesis a. Transfer Rate: One hundred cigarettes will be spiked with NNK-carbonyl- 4C, recently synthesized by us (9,10), and will be smoked under standard conditions (11) to 23 mm butt length. For the actual analysis, we will employ NNK-3H as internal standard (10). In order to assure that the 0-activity of the endfraction originates only from the intact and transfered NNN-carbonyl-14C, the HPLC separation must distinguish between NNN-and the Z- and E-isomers of NNK (3). Therefore, the isolated NNK-containing fraction will be rechromatographed by reverse phase HPLC. The dual counting of the (3-activity of the isolated NNK-14C and of the internal standard NNK-3H will permit their calculation of the absolute amount of transfered NNK within +10% (10). The transfer rate of NNK will be determined for a cigarette made entirely from flue-cured tobacco, Burley tobacco, and from a U.S. cigarette tobacco blend (experimental cigarettes 1, 2, and 3). " b. Pyrosynthesis for NNK: One hundred cigarettes without filter tip will be spiked with nicotine-methyl-14C (12) and will be smoked under standard conditions (11) to a 23 mm-butt length. The analysis will be similar to that out- lined under "a. Transfer Rate" with NNK-3H as internal standard. The analysis will again be completed for a cigar- ette made entirely of flue-cured tobacco, Burley tobacco, and from a U.S. cigarette tobacco blend (experimental cigarettes 1, 2, and 3) -3- 1

The data from 2a and 2b will give us the best possible information currently attainable for the formation in cigar- ette smoke of the biologically highly active NNK. We will also explore the possibility of inhibiting the pyrosynthesis of-NNK by addition of ascorbic acid (5 mg/cig- arette) to the 62 mm tobacco column of those experimental cigarettes which do not have filter tips. We will spike each cigarette with only 5 mg ascorbic acid, another time with both ascorbic acid (5 mg) and with nicotine-methyl-14C. Whereas the volatile N-nitros amines appear to be formed pri- marily by pyrolysis of protein and NO in the burning cone, it is our working hypothesis that the TSNA are pyrosynthesized from the alkaloids and NO in the hot zone of a burning cigar- ette, thus, ascorbic acid may serve as a scavenger of the NO. GC-Analysis of Tobacco Alkaloids: The study of the for- mation of TSNA requires detailed information as to the pres- ence and concentration of the individual nicotiana alkaloids in the tobacco to be smoked. Although application of the recently developed HPLC-method for the alkaloids (13) will yield the needed information, Fie believe that a capillary GC-MS method would result in better resolution of the indi- vidual tobacco alkaloids. Thus, we propose to develop such a method. 3. Selective Reduction of TSNA from Cigarette Smoke 1. Conventional Filter Tips -" A number of studies have demonstrated that modification of the tobacco and of the f ilter of a ciqarette can effec- " tivel y reduce "tar" nicotine and carbon monoxide in the -- ~ `-~- smoke. filter Such modifications include the use of very active materials, filter tip perforation and the use of high- • =~ lv norous paper wrappers. However the presently popular U, S f'1t ci arettes d 1' k th t 1 a t 0 8 I ~ +~I' r iv i e g e er smo e wi a e s . ug ~- -- _ TSNA (3,4). To our knowledge, there have been only limited p;~F:~"' efforts towards the reduction of TSNA in cigarette smoke. We ropose several experiments which should produce the neces- p sary base line data for the reduction of TSNA in commercial F+ A cigarettes. (This study should, be completed with 85 mm long filter cigarettes having a U.S. tobacco blend, a 20 mm filter tip, a cigarette wrapper of medium porosity, and cellulose acetate filter material with about 6 denier fibers, except otherwise stated ) ..i

a. Nonfilter Cigarettes with Paper Wrapper of Different Porosity: Cigarette 4. Low porosity 3. __ Medium porosity 5. High porosity b. Cigarette with Cellulose Acetate Filter Tip: Cigarette 6. Cellulose secondary acetate (=1.6 denier fiber) 7. Cellulose_secondary acetate (°'3.2 denier fiber) 8. Cellulose tertiary acetate (=3.2 denier fiber) c. Filter Filter: Cigarettes with Different Plasticizers in Cigarette 6. Control cigarette .9. 10. 11. as 6, but with =6% glyceryl triacetate as 6, but with =6$ polyethylene glycol as 6, but with '='6$ polyethylene glycol tereph,thalate d. Filter Cigarettes with Perforation in Middle of 20 mm Filter Tip : - Cigarette 6. Control cigarette (no perforation) 12. Perforation resulti.ng in =10~ air dilution 13. Perforation resulting in 20-25% air 14. dilution Perforation resulting in =30-35$ air 15. dilution Perforation resulting in =45-50% air dilution We will smoke all experimetal nonfilter cigarettes under standard conditions to 23 mm butt length (1-5), and in the case of the filter cigarettes, to a butt length of 20 mm plus overwrap, plus 3mm. The smoke of each cigarette will be ana- lyzed for "tar" (FTC), nicotine,CO and tobacco specific N-ni- trosamines (TSNA). In the "d-series" we will also separately determine the TSNA in the filter part which is between tobac- co column and filter perforation (upstream filter partk and in the filter part which is between filter perforation and filter end (downstream filter part).

This study should clearly answer questions in regard to the factors that play a role in the formation of TSNA and their selective reduction in the smoke. We consider this a research program of major importance towards the development of cigarettes with highly reduced biological activity. D. EXPERIMENTAL CIGARETTES As outlined_in "C. Methods", this study would require 15 different experimental cigarettes, as follows: , Cigarette 1. 2. 3. 4. .5. 6. 7. 8. 9. 10. Flue-cured tobacco only - nonfilter Burley tobacco only - nonfilter U.S. cigarette tobacco blend - nonfilter As 3, but with low porosity paper - non- filter As 3, but with high porosity paper - non- filter U.S. cigaette tobacco blend - cellulose secondary acetate filter, =1.6 denier filter (CA) As 6, but with =3.2 denier fiber filter As 6, but with =3.2 denier tertiary acetate As 6, but filter with =6$ glyceryl triace- tate _ col As 6, but filter with =6$ polyethylene gly- col terephthalate -12. As 6, but with perforation resulting in =10$ air dilution - 13. As 6, but with perforation resulting in 14. 15. 20-25% air dilution As 6, but with perforation 30-35% air dilution As 6, but with perforation 45-50% air dilution resulting in resulting All cigarettes should be 85 mm long. In the case in f ilter cigarettes, the filter column would be 20 mm long (cigarettes 6-15). In the case of the filter cigarettes with perforation (cigarettes 12-15), the single line of holes would be in the middle of the cellulose acetate filter tip. Cigarettes 3-15 would contain the same U.S. tobacco blend, cigarettes 1-3 and 6-15 would have the same cigarette wrapper paper and filter cigarettes 6-8 and 12-15 the same concentra- tion of the same plasticizer in the filter. The different tobaccos of experimental cigarettes 1 and 2 and the rest will also be analyzed for individual alkaloids. -6- a*: As 6, but filter with =6% polyethylene gly- _--~-~--~

E. TIME SCHEDULE - STAFF NEEDS The proposed research plan can be completed within 6 months after receiving the experimental cigarettes, or a portion of them, with the following staff : 1. ~Senior Analytical Chemist 35% 2. Research Chemist - 100% 3. Research Chemist 100% 4. Research Chemist 100% 5. Research Chemist 100% 6. Smoke Technician 50% Supervisor Project 1 and Method for Tobacco Alkaloids Project 2a and 2b Project 3 For all projects Special Chemicals needed: Nicotine-methyl-14C NNK-random tritiated