Tobacco Documents Online

PHILIP MORRIS U. S. A.. I', NiT E R' - O F' F I C' E C' O R R E S P O N D E N C E Ri;chmosrd, Virginia R'ESTRISITED1 D !ISiRIBUTION Tb: Dr. A. Warfield Date: 2'5' April 1989 From: W. P. Hempfling Subject: Unextracted nicotine as potential source of NNK in s2noke:~ a proposal for study The search for the identity or identities of "unextracted nicotine", which, may be a source or sources of'NNK in smoke, comprises the need to answer three questions. The resoliution of those questions is addressed'in the following. A. Is there a correlation between UNiin base webs and NNK in sm©ke? Strategv; Using base webs, the delivery of NNK in smoke of which is known, measure contents of total nicotine, extractable nicotine and unextracted nicotine. Compile a "nicotine balance". Compare values (original dry weight basis) obtained with, values of NNK in smoke. Measure endogenous NNK to estimate transfer into smoke. 1. Total nicotine (TN) Extract base webs with 1 N NaOH, 900 C', 24 hr (.3x water wash) . Measure nicotine content (by GC or HPLC, method to be determined), first extracting combined aqueous portions with dichloromethane (DCM). This treatment sieems to reveal the largest amount of'nicotine of any method now extant. Special attention must be paid to the extent of solubilization of the tobacco fiber, since an indication exists that nicotine remains bound to the small marc: remaining after alkaline digestion. It will be: useful to submit tobacco samples to ARD for estimations of nicotine by the pyran procedure. If conditions of pyrolysis are identified' that liberate significantly greater amounts of nicotine than does hot alkaline digestion, those values should be used as the measure of TN. 2. Extractable nicotine (EN) Extractions are designed to remove soluble nicotine without modifying nicotine-containing compounds that might be present. r~ . ~ a. Extraot base webs (!SW) with neutral buffer (50 mM potassium phosphate, ~ pH 7) in the cold, 24 hr. Wash marc 3x with buffer and combine:. Reserve marc. (~. ~ b. Extract BW with neutral buffier, 90'O C, 24 hr. Wash marc 3x with buffer ~ and combine. Reserve marc. ~ The aqueous extracts of a. and b. above are made 0.5 - 1 N with NaOH and 0 extracted 3x with DCM, and the DCM fraction assayed for nicotine content. ~

C. Extract: BW with methanol/10%: ammonia ,in the: co]Jd, 241 hr. Wash marc 3x with methanol and combine. Reserve marc. Measure nicotine content directly. . Unextracted nicotine (UN) we postulate the existence of two, forms of UN in BW': water-soluble and water-insoluble. The latter form is bound in some manner to the cellulosic tobacco matrix. Both forms yield nicotine in smoke upon combustion.,, and upon prolonged exposure to hot alkali. a . Water-soluble UN ('"Nic-X"'). Mix portions of extracts 2.a. and 2.b. (above) with NaOH to reach 1 N. Heat at 9i0'O' C for 24' hr. Measure nicotine content of' 2.al2 and 2.b. ('DCM extraction,) before andd after heating. The difference is the amount of Nic- X. b. Water-insioluble UN ('"Nic-Y'"'). Mix the (dried) mares from extractions 2. a., 2.b.. and 2'.c. with 1 N NaOH and heat 0100 C) for 24 hr. ESctract aliiqiaots of the alkaline extractt with DCM and measure nicotine content. That quantity is the amount of'Nic- Y. Special attention must be paid to the extent of solubilization of the tobacco fiber, since an indicationlexists that nicotine remains bound to the small marc remaining after alkaline digestion. . Nicotine balance The sum of water-soluble nicotine (2.a. and 2.b.) plus their respective Nic-X and Nic-Y contents should equal TN. This should also be true of the methanol/ammonia-extracted material, unless part or all of Nic-X is soluble in that reagent' (there is some reason to expect solubility in methanol - Ref. 3) . A hot alkaline digestion of a portion of the methanol/ammonia extract should be performed to test the solubility of' Nic-X in that solution. 5. Minor alkaloids The preparation of hot-alkali extracts of base webs for nicotine estimation afford's the opportunity to examine their contents for the presence of' otherwise unextracted forms of the minor alkaloids nornicotine and anatabine. Should this be considiered desirable, sufficient material ought to be available for those assays. B. What is or are the chemical identity or identities of UN? It seems to me highly important that we establish the chemical identities of both, Nic-X and Nic-Y, especially if one or both of these sources of nicotine is correlated with NNK:generation in smoke. Knowledge of their chemical natures will enable us to infer experimentally testable hypotheses about the means of NNK' generation at high temperatures. Moreover, that information may suggest ways of treating tobacco, prior to SCFE, in order to reduce or eliminate the unextracted forms of nicotine.

Strategv; Isolate the water-soluble forrki of UN (Nic-X) and determine its chemical structure (mass spectrometry).. Subject the water-insoluble matrix containing Nic-Y to treatment with enzymes that digest cellulose and protein in an attempt to render the Nic-Y moiety or moieties sufficiently water- soluble to employ separation methods that are to be worked out during Nic-X isolation. Isolate the water-solubilized form of Nic-Y and determine its chemical structure (further enzymatic: or cheznical degradation, masss spectrometry). 1. Isolation of Nic-X The raw material from which Nic-X will be isolated will be CEL. The CEL~ (bright, burley or oriental) chosen for work-up will be identified by measuring the content of Nic-X as already described (3.b., above), selecting, that CEL with the highest Nic-X content. Nicotine will first be extracted by rendering the solution alkaline ('ca. 01.5 N Naf9H)~ and mixing repeatedly with DCM, (room temperature) for brief periods of time. It is possible that Nic-X may be:solubZe.or partially soluble in DCM, and this must be initially determined by performing a hot alkali digestion of the DCM extract (;after removing DCM); andiof the DCM-extracted CEL. If Nic-X is found to be soluble in DCM, the: organic extract can be used as a, starting point for further (simpler) purification. We ought to consider further solvent extraction schemes if'this~be true. If Nic-X proves not to be soluble in DCM, the aqueous fractionifrom which nicotine has been removed will serve as starting material. T will assume: the latter to be the caBe, because it will present the:greater difficulty in isolation. The presence of Nic-X in various fractions arisingi during, purification will be determinedi as in A.3.b., above. Isolation, ought to: be.achieved by (in sequenceY the following steps, preceded by high-speed centrifugation in the cold and micropore filtration to~remove particulate matter. a,. Gell permeation chromatography in the cold using BioGel P-4 or similar matrix allowingiresolution below 10 kD (acetate buffer, pH 5). b. Cation-exchange:chromatography (strong resin) in the cold at pH 5 (acetate: buffer), using either an increasing,gradient of acetic acid (01 to 1 M) or NaOH (0 to 1 M) ; c. Fbliliowing neutralization of the combined fractions containing Niic-X,. HPLC reversed-phase: separation (C'l e) with methanoli/water or acetonitrile/water, or other gradients (to be determined)1, at suitable pH values. UV absorption by the nicotine moiety can be used to monitor the progress of chromatography. Should Nic-X be soluble in DCM or methanol, the reversed-phase HPLC'' separation will be applied directly to solvent-free fractions, and fully orgianic or polar org,anic/water gradients may be employed (to be determined by consultation).. The most hig,hly purified Nic-X fractions will be submitted for examination by mass spectrometry. In addition, and foLlowing hot alkaline digestion of a portion of the purified material, reversed-phase HPLC will be used to separate free nicotine from other components thereby liberated, which will themselves be submitted for MS analysis. It will be useful to subject Nic-X to treatment with P-glucosidase (see C. below). -3-

When adequate quantities of purified Nic-X are available, they will be tested in cigarettes to determine if'NNK is produced owing to Nic-X addition (matrix to be determine'd). 2. Isolation of water-soluble forms of Nic-Y Prepare hot-water-extracted base webs (A.2.b., above) that are essentially free of EN and Nic-X. Subject these Nic-Y-bealring materials (well-commin'uted), to digestion with enzymes under appropriate conditions (from the literature) of pH, ionic strength, temperature and time. Since treatment with one kind of hydrolase may render the remaining insoluble fraction more susceptible to digestion by one with a different substrate spec:ificity, the enzymes should be applied singly and in succession. Fbr example, digestion with, endoprotease may render the matrix more susceptible to cellulase action, or the reverse. Suitable commerc'ially available enzymes include:: a. cellulase from Trichoderma viride; b. endoprotease (e. g. Pronase from Streptomvices)s' c. phenolase (tyrosinase) from potato;; d. pectinesterase from orange peel; and e. pectinase from Aspercrillus niare;r. Success in liberating water-soluble forms of Nic-Y will be indicated by digesting the supernatant portion (following centrifugation) with hot alkali (A.3.b., above), and showing the presence of nicotine in the digest. Of' course, the possible presence of" nicotine prior to alkaline digestion will also be determined. Isolation of soluble Nic-Y'will proceed as outlined for Nic-X ('B.l.a., b., above), with modifications dictated by the results'of enzyme digestions, which will allow the inference of the types of compounds associated with nicotine in the bound!state. Mass spectrometric analysis will be applied when suitable isolates are available. C. Can nicotine-N-glucoside serve as an esperimntal analog of Nic-X? At lea~st two p'ublications have appeared (1,2) which describe the presence of nicotinic acid-Nrglucoside in tobacco and parsley cells. This compound releases nicotinic acid when treated with A-g,lucosidiase, results reminiscent of C. Kroustalis"s earlier finding (3) that suchh enzymatic treatment of tobacco fractions releases nicotine. We need to be attentive to the possibilit y that a nicotine-N-glycoside may actually be Nic-X, Nic-Y (in a water-insoluble form due to linkage to the cellulosic matrix) , or bothi. According,ly, purified Nic-X fractions will be digested with Q-glucosidase (3) and the products identified. Recent consultation with Dr. A. Wol'if has produced the opinion that such a compound will likel!y possess at least some of'the properties of Nic-X; namely, release of nicotine in hot alkali but not in hot acid. In any event, we should request the preparation of nicotine-N- glucoside, as a possible analog of Nic-X, for smoking studies (NTiKK production). One might also imagine that other potential NNK precursors could exist as giycos'ides (e.g. pseudooxynicotine). I do not mean to aver that Nic-X or Nic-Y i_s a glycoside or glycosides.. ~ 1. Extractions and digestions ought to be designed so that other modes of nicotine immobilization might also be revealed if they in fact exist. ~. -4-

I look forwa~rd to discussions with you and members of Project 6908 to develop a detailediplan of operation with time lines consonant with the importance of'this matter. Refesencew 1. Mizusaki, S., Tanabe, Y., Kisaki, T. and T'aQnaki, E. Metabolism of nicotinic:acidiin tobacco plants. Phytochem., 9, 549 - 554, 1970. 2. Upmeier, B., Thomzik, J. and Barz, W. Nicotinic acid-I?Trglucoside in heterotrophic!parsley cel!l suspension cultures. Phytochem. 27, 3489 - 3493, 1988. 3. Kroustalis, C. Personal communication, 4 April 1989. Copies to,: Pt. Carchman R. Izac J. C'. X. X. B. S. S. Charles Ellis Fenner Ferguson Handy Hassam,L,/ Haut R. N. C. R. C. D. A. Jenkins Jensen Keene Kinser Kroustalis Leyden Lilly R. Morgan K. Shafer D. Watson J. Whidby