Tobacco Documents Online

PRIVILEGED AND CCi;F,DEidTIAl, REDUCTION OF CARCINOGENICITY ON MOUSE SKIN OF CIGARETTE SMOKE CONDENSATE WITH PALLADIUM CATALYST §Life Sciences Division, Arthur D. Little, Inc., Cambridge, Mass. 02140; and -fiResearch Department Liggett Group, Inc., Durham, North Carolina 27702 ABSTRACT The carcinogenicity of cigarette smoke condensate (CSC) toward mouse skin has been dramatically reduced by 77-100% through a com- bination of palladium metal catalyst with blends high in Burley tobacco content or with lower Burley tobacco blends supplemented by the addition of compensating amounts of a nitrate salt. This lowering of carcinogenicity is associated with decreases in the level of polycyclic aromatic hydrocarbons (PAH) in the smoke con- densate. Through the use-of filters designed to complement this additive combination, the other measured components of the smoke stream are either lower than or equivalent to those of conventional best sell- ing US cigarette-brands._ This is confirmed by lowered in vitro ciliatoxicity and cytotoxicity responses as well as in vivo sub- acute inhalation studies with rats.-

REDUCTION OF CARCINOGENICITY ON MOUSE SKIN OF CIGARETTE SMOKE CONDENSATE WITH PALLADIUM CATALYST §Life Sciences Division, Arthur D. Little, Inc., Cambridge, Mass., 02140;_tResearch Department., Liggett Group, Inc., Durham, North Carolina 27702 In 1953 Wynder, Graham and Croninger reported that cigarette smoke condensate (CSC) was carcinogenic for the skin of the CAF1 , mouse. Epidemiological evidence had already appeared associating lung cancer in man with cigarette smoking (Doll, 1954; Wynder, 1950). Experiments were soon initiated in a number of laboratories, in- cluding our own, using the mouse skin assay to attempt to identify active carcinogens in the-smoke condensate and to study processes or treatments to reduce this potentially harmful biological acti- vity. A short-term mouse skin sebaceous gland assay (Suntzeff, 1957; Bock,_ 1958) was used in conjunction with the long-term mouse skin painting experiments to follow the concentrating of the active materials. In agreement with the results of a number of other investiga- tors (Wynder, 1957; Bock,1962; Lee, 1977; Severson, 1976), we found that the bulk of this activity was present in the fraction containing polycyclic aromatic hydrocarbons (PAH). Our findings were in essen- I tial agreement with the reports of these other investigators, namely,

that a major part of the mouse skin tumor-initiating activity is associated with a very complex mixture of unsubstituted-and alkyl- substituted PAH compounds. Further separation of this fraction has shown that significant sebaceous gland suppression (SGS) activity is largely restricted to 4- and 5-ring PAH rather than to those with fewer or more rings. These separations also showed that the SGS activity is mostly associated with alkyl-substituted PAH rather than with the unsubstituted compounds. This was verified, with respect to mouse skin carcinogenicity, by preparing the PAH fraction from pyrolysis of cigarettes at 1000°C to provide a source of an array of unsubstituted FAH like those found in_cigarette smoke. The unsubstituted PAH did not-increase the carcinogenicity when added to CSC in amounts-equivalent to their presence in unfortified smoke. The PAH from CSC,-containing both unsubstituted and alkyl-substituted PAH, enhanced the=-carcinogenicity of either whole CSC or the heptane- soluble CSC in proportion to the rate of its addition. Since the SGS activity of the PAH fraction of CSC has been shown to be distributed=over many different classes and-numerous compounds within each of these classes, it was concluded that efforts to re- duce the carcinogenicity of the smoke must be aimed -at reduction of the PAH compounds in general, with particular importance attached to the reduction of simple alkyl-substituted PAH of 4 or 5 fused rings. We therefore developed a routine procedure for_concentrating the most active PAH compounds in one fraction so that--we might use this to measure the effectiveness of tobacco treatments. The product of this fractionation amounted to about 0:2$ of the total dry con- densate (40-60 pg/cigt) and gave an ED50 for SGS activity of-about

- 3 - 200 pg [one-fourth as active as benzo(a)pyrene (BaP ED50-- 50 ug)]. , i d-A"« j By addition-of isotopically labeled BaP to the smoke condensate prior to carrying out the routine preparation of the-PAH fraction, it was possible to show that 81% of this material was recovered in the end fraction. The PAH concentrate, prepared in this fashion was charac- terizied by applying HPLC and GLC techniques to give further separa- tion into the-expected array of substituted and unsubstituted PAH (Severson, 1976)-demonstrating that a highly concentrated mixture of these materials had been obtained. Initially the yield of the active PAH fraction from various smoke or pyrolysis condensates was assessed by means of-the four-day, mouse skin SGS test. As a con- venience, it was later found possible to use infrared absorbance as a measure of the-amount of active PAH since this was shown to correlate well with the SGS test. Since the manufacture of cigarettes requires inconveniently large amounts o-f both tobacco and additives, it was desirable to design a technique for preliminary screening using smaller samples. An appa-ratus wasdesigried to-approximate the conditions present in a burning cigarette, although this cannot be-don-e in any precise way. An important feature of this technique is that the heat required for the pyrolysis is provided by combustion of the tobacco at the peri- phery. Control of the rate of pyrolysis was exercised by adjusting ~ _L ," - ~ the rate of flow of nitrogen gas through the tobacco mass and into ~ GJ the cold traps where the condensate was collected. The reactor ~ N ~ conditions were adjusted to obtain a PAH fraction similar in compo- `O sition to that found in cigarette smoke. The most promising addi- tives, i.e., those which gave significant reductions in the pyroly-

4 sate active PAH content, were subsequently incorporated into ci-ga- - (.[~c k~il.r4 rettes. The smoke condensate was then evaluated to confirm the PAH reduction. U-timetely, the most-promising additives were evaluated by determining the degree of carcinogenicity of the CSC toward mouse skin as compared-to-condensate from untreated cigarettes.- - MATERIALS AND METHODS Preparation of Tobacco Samples Containing Additives. Solid, water-insoluble additives were ground finer than 1001mesh and added / to a casing mi-xture consisting of glycerin, propylene glycol, invert sugar, corn syrup, flavorant materials, and water. Water-soluble additives were added to the premixed casing solution in a minimum amount of water. The thoroughly-mixed dispersions of casing and additive were sprayed onto uncut Bright, Burley, Maryland, and Turkish (BBMT) tobacco blends at levels equivalent to 14% casing and 1-10% additives. Reconstituted tobacco was added in some cases. The treated tobacco was cut in a laboratory guillotine cutter at 32 cuts per inch and allowed to equilibrate at 60% RH and 68°F for a minimum of 48 hours. Preparation of Cigarettes Containing Additives. Soluble addi- tives were dissolved in water and added to a casing mixture of gly- cerin, propylene glycol, invert sugar, corn syrup, flavorant materials, and water. Insoluble additives were added to the casing mixture and pebble-milled for a minimum of 12 hours. The continuously stirred mixtures of casing and additive were applied to blends of BBMT tobaccos in a casing cylinder. Reconstituted tobacco was then added to the blend. The blend-was cut,and during the drying operation,cut tobacco stems and expanded tobacco were added. Cigarettes were manufactured on factory production equipment.

- S - Severai of the additives were-incorporated in reconstituted tobacco sheet to maximize catalytic availability. To prepare the treated reconstituted tobacco, the additives were dry-mixed with a blend of ground BBMT tobacco strips, a-cellulose,_and sodium carboxy- methylcellulose. A solution of water, propylene glycol, and glycerine was added to the dry blend and the combination wet-mixed until the appropriate consistency was achieved. The material was then rolled into sheets on factory production equipment,-cut into 4 inch squares and finally shredded using rotary cutters at a rate of 32 cuts per inch. The resulting product was used to manufacture cigarettes. Tobacco Pyrolysis Procedure. Approximately 160 g of cut tobacco, equilibrated at 60% RH and 74°F, was packed to a density in the range of commercial cigarettes in t he apparatus shown in Figure 1. The tobacco was lit at the surface of the screen cylinder and the rate of combustion was controlled by pulling a vacuum through an exit tube located at the axis of the cylinder while simultaneously bleeding N2 into the tobacco mass through the two end screens. The flow conditions were selected to give a PAH composition similar to that observed for CSC. The condensate was collected in two glass traps, cooled in liquid air, connected in tandem at each end of the exit tube. The traps were washed with distilled acetone and the yield of "dry" con- densate determined by evaporation of an aliquot. Standard Procedure for Preparation of a PAH Concentrate From Cigarette Smoke Condensate. The smoke condensate (ca 30 g) was extracted from aqueous 84% methanol (w/w) with three portions of hexane (1, 0.5, and 0.5 v/v). Distribution coefficients in this ~

- 6 - system indicate that more than 96% of the PAH should be recovered in the hexane extracts. After washing with 0.1 volume of water, the hexane extract was ~ evaporated to 150 ml and extracted three times with 300-m1 portions of N, N-dimethylformamide (DMF) in water (9:1 v/v). The 900 ml of 9:1 DMF:H20 was back extracted with two equal portions of hexane. To the DMF solution was added 300 ml hexane and then 900 ml H20 :t, to increments of about 100 ml,with shaking after each addition. - - ~ The hexane layer was removed, and the 45% DMF was extracted twice more with 300-m1 portions of hexane. The combined 2700-m1 hexane extract was concentrated to 500 ml and washed with two 50-m1 por- tions of water. Recoveries greater than 88% would be expected in this DMF ext-raction procedure for materials with distribution coeffi- cients less than 1.87 in the hexane/90% DMF system. Active compo- nents of CSC by the SGS test had distribution coefficients between 0.15 and 1.87 in this system. Chromatography with hexane of the solids extracted from 45% DMF was accomplished on 40 times its weight of Fisher acid-washed alumina (80-200 mesh). The adsorbent was activated to a picene retention of 15-20% (i.e.,15-20% removal of picene by 1 gram of alumina in 30 minutes from 20 ml of cyclohexane solution containing 0.8 pg of picene per ml). The materials eluted between 0 and 6.0 ml ~ 3. of hexane per gram of alumina were collected to contain the PAH con- centrate. The PAH concentrate was next chromatographed on 200 times its weight of silica gel (Davison 923, 100-200 mesh). The activity of the silica gel was adjusted by addition of water to cause retention I! 'Ii I

- 7 of pyrene until the first 10 ml of hexane per gram of adsorbent had been collected. At this point elution of the PAH was achieved with 23.5 ml of 6% benzene in hexane per gram of silica gel. The procedure was designed to retain PAH's with 4-5 fused rings while eliminating as much of the smaller and larger molecules as possible. The fractionation procedure is outlined in Figure 2. The recovery of non-volatiles in the PAH eluate averaged 40-60 pg/cigt. or 0.2% of the total dry smoke condensate. Evaluation of the PAH concentrate was performed either by the SGS Test (Suntzeff, 1957), or by measurement of total infrared absor- bance at 11.9-14.0p or the ratio of absorbance at 3050 cm-1 to that at 2960 cm-1. When the estimation of activity was by SGS on mouse skin, the activity is reported as the ED50. When infrared absor- bance was used to_predict the biological activity, it is reported as the PED50. In order to evaluate the treatment, results were expressed in terms of the number of PED50's per gram of dry conden- sate (DC). Comparison of the IR traces from 0.5 mg of the PAH concentrates obtained from comparable pyrolysate and CSC from the same tobacco blend are shown in Figure 3. Figure 4 presents the relative mass spectral responses for these pyrolysate and CSC PAH concentrates. The dissimilarities between the PAH concentrate composition obtained from the two processes are within the reproducibility limits inherent in the testing program. Further comparison of the PAH concentrates from tobacco pyrolysates or CSC was made by studying the GLC pro- files of gel permeation chromatography fractions obtained from con- trol, treated, and commercial filter cigarettes. Similar patterns ~

- 8 - were found for all three cigarettes, suggesting similar relative abundances of various PAH types. Although the GLC procedure doesn`t resolve a-11 the components, significant differences in the tion of the PAH mixture would be expected to be composi- evident,if present. To further test this result, reverse phase HPLC was employed on ,c(-Bondapak C-18 (Waters Associate), or Partisil ODS-2 (Whatman) Ac columns, using acetonitrile-water as the -mobile-phase, and 254,280, 340, 384, and 430 nm detection. No significant differences were seen in the relative abundance of PAH with four or more rings. Comparison of the levels of small PAH was not feasible since these compounds vary semewhat due to evaporation-in the workup steps. Mass Spectrometric Analysis of PAH Concentrates. The proce- dure used for these measurements was suggested by a publication by Biros (1970) and is essentially similar to that later described by Blumer (1975). Over the period of time during which samples-were analyzed by this technique several dif-ferent instrumental combina- tions were used. The most recent utilized a DuPont 21-490 mass spectrometer in combination with a Finnegan-Incos Model 2300 Data System. - -- - - - An aliquot of an ether solution was evaporated under vacuum in the probe glass capillary (1.5mm ID x 19mm) to leave 10 pg of the PAH concentrate and 1.2 pg of the internal standard,o-tolidine (mol wt 212). The mass spectrometer ionizing voltage was set at 12 e.v. and the data system was programmed to acquire 150 spectra at the rate of 1-spectrum every 6 sec. Scanning was begun, and the /I sample in the solid sampling probe was introduced through the vacuum lock into the ion source. The temperature of the sampling probe ~

l - 9 - was manually increased from 75°C to a final hold temperature of 290°C over a period of 7.5 minutes. After acquisition of the 150 scans, one hundred consecutive spectra were averaged by the data system beginning with the spectrum taken at the time the probe was first introduced into the ion source. The resultant averaged spectrum was normalized on the internal standard peak at-m/e 212. Figure 4 presents this low voltage mole- cular ion summation for several samples of PAH concentrates calcu- lated as the percent of total ion intensity. Evaluation of CSC for Carcinogenicity by Painting on Mouse Skin. Cigarettes were equilibrated at 60% RH and 74°F and smoked to 30mm butts on Liggett & Myers Model 4 smoking machines(Hackney, 1965). Puff-A of 35-m1 volume and 2-sec duration were taken at one minute intervals. The smoke condensate was collected in two traps cooled in liquid air. It was dissolved in acetone, and the acetone removed under reduced pressure at 40-50°C. The final "dry" conden- sate was dissolved in an appropriate volume of acetone to give 50 mg of condensate per 100 mg of solution. For each test group, 50 Swiss Ha/ICR female albino mice were used. Hair was clipped from the dorsal test area prior to each painting using an Oster Co. (Racine, Wis), Model A-2 small animal clipper with a size 40 blade. The animals were 9 weeks of age at the start of the experiment. Application of 100 mg of smoke con- densate solution was accomplished with a Grurrbacher No. 6 camel hair brush. Test materials were applied five days per week for 80 weeks. Analyses for Constituents of Cigarette Smoke-. Cigarettes used for these analyses were equilibrated for at least 24 hours at 60% RH and 74°F prior to smoking for analysis. The nicotine-free Jii