Vol. 1, /153-1163, October /995 Clinical Cancer Research 1153 Identification of a Class 3 Aldehyde Dehydrogenase in Human Saliva and Increased Levels of This Enzyme, Glutathione S-Transferases, and DT-Diaphorase in the Saliva of Subjects Who Continually Ingest Large Quantities of Coffee or Broccoli’ Lakshmaiah Sreerama, Matthew W. Hedge, and achieve its full therapeutic potential would be selected based Norman E. Sladek2 on whether high or low enzyme activity would be favorable in that regard. Such measurements may also be useful as an Department of Pharmacology, University of Minnesota Medical indicator when exposure to carcinogenic/teratogenic/other- School, Minneapolis, Minnesota 55455 wise toxic environmental/industrial/dietary agents that in- duce these enzymes is suspected. ABSTRACT Human saliva was tested for the presence of cytosobic INTRODUCTION cbass 3 abdehyde dehydrogenase, gbutathione S-transferases Celbuban expression of ALDH-33 and certain other enzymes a, ,i, and IT, and DT-diaphorase, enzymes that are known to (e.g., glutathione S-transfenases and DT-diaphorase) can be catalyze the biotransformation of many xenobiotics, includ- markedly and coordinately increased by both monofunctional ing some that are carcinogens and some that are antineo- (e.g. , phenolic antioxidants such as catechob, hydnoquinone, and pbastic agents. Each of these enzymes was found to be 2,6-di-tert-butyl-4-hydnoxytobuene) and bifunctional (e.g., present in this fluid. Inducers of these enzymes are known to 3-methybcholanthnene and 3,4-benzpymene) inducems4 (1-3). In- be abundantly present in the human diet, especially in cer- duction of glutathione S-tnansfenases and DT-diaphorase is tam vegetables and fruits. Further investigation revealed viewed with particular interest because these enzymes are that the salivary content of these enzymes rapidly, coordi- known to detoxify certain carcinogens (4). Inducers of these nately, and markedly increased upon daily consumption of enzymes are abundantly present in certain components of the relatively large amounts of coffee or broccobi. The enzyme human diet, e.g. , members of the Cruciferae and Liliaceae activities of interest rapidly returned to basal bevels when families of vegetables (5, 6). Certain food additives (e.g., 2,3- these substances were removed from the diet. Given the tert-butyl-4-hydmoxyanisobe) and pharmaceuticals (e.g., olti- important role that cytosolic class 3 abdehyde dehydroge- praz) also act as inducens of these enzymes (reviewed in Refs. nase, the glutathione S-transferases, and DT-diaphorase are 7-9). Vegetables, fruits, and chemicals that induce these en- thought to pbay in determining the carcinogenic potential of zymes prevent experimental cancinogenesis, i.e., they effect a some cancer-producing agents as well as the cytotoxic po- chemopreventive action, and the former is thought to be caus- tentiab of some antineoplastic agents, and assuming that ative, at beast in part, of the batten (reviewed in Refs. 7-9). their salivary levels reflect their tissue bevels, quantification Chemoprevention mediated in this manner is an immenseby of the salivary content of one or more of these enzymes, a attractive idea for obvious reasons, and clinical evaluation of noninvasive and relatively easy undertaking, could be useful oltipnaz in that regard has already been initiated (10-12). De- in: (a) preliminarily assessing the chemopreventive potential sirabbe in such investigations is an estimate of relevant enzyme of various diets and drugs; (b) establishing the optimal dose induction. One way of doing this might be to quantify the and schedule in Phase I clinical trials for any putatively relevant enzyme activities in serum, peripheral blood bympho- chemopreventive diets or drugs of interest; and (c) the ra- cytes, on tissue biopsies (13-15). Limitations of these ap- tional selection and use of chemotherapeutic agents, since several are inactivated, and a few are activated, by these enzymes; alternatively, the antineoplastic agent could be selected first and then a diet that enables the agent to 3 The abbreviations used are: ALDH-3, human cytosolic class 3 alde- hyde dehydrogenase; mIU, milli-International Unit of enzyme activity [nmol NAD(P)H formed/mm in the case of aldehyde dehydrogenase activity, nmol of the conjugate of 1-chloro-2,4-dinitrobenzene and glu- Received 1/20/95; revised 5/15/95; accepted 5/18/95. tathione formed/mm in the case of glutathione S-transferase activity, 1 Supported by USPHS Grant CA 21737, Bristol-Myers Squibb Com- nmol 2,6-dichborophenol-indophenol reduced/mm in the case of DT- pany Grant I00-R220, and Department of Defense Grant DAMD 17- diaphorase activity, and nmol p-nitrophenol formed/mm in the case of 94-J-4057. Descriptions of parts of this investigation have appeared in esterase activity]; PVDF, polyvinylidene difluoride. abstract form (L. Sreerama, M. Hedge, and N. E. Sladek. Proc. Am. 4 Monofunctional inducers are defined herein as agents that induce Assoc. Cancer Res., 35: 84, 1994). ALDH-3, glutathione S-transferases, and DT-diaphorase [NAD(P)H: 2 To whom requests for reprints should be addressed, at Department of quinone oxidoreductase; NQO1] but not cytochrome P450s IA1 and Pharmacology, University of Minnesota, 3-249 Millard Hall, 435 Del- 1A2; bifunctional inducers are defined herein as agents that induce all of aware Street SE., Minneapolis, MN 55455. these enzymes (Ref. 28 and references cited therein). Downloaded from clincancerres.aacrjournals.org on September 26, 2021. © 1995 American Association for Cancer Research. 1 154 Induction of Salivary Chemopreventive Enzymes proaches are that each is invasive, the latter two are labor Human stomach mucosa was provided by the Cooperative Hu- intensive, and the enzyme levels in serum are very bow on nib man Tissue Network (Midwest Division, Columbus, OH). Pu- (Ref. 15 and the present study). Another way of doing it would rifled stomach mucosa ALDH-3 and chicken antistomach mu- be to determine various phammacokinetic parameters, e.g., cosa ALDH-3 IgY were prepared as described previously (22). plasma half-life, of an appropriate test substrate before, during, Collection and further processing of saliva was essentially and after exposure to the suspected inducer (16); this approach as described by Takase et al. (23), except that hyaburonidase was is also invasive, not likeby to be as specific, and, in any event, not added to the collected saliva. Briefly, saliva samples (3-5 likely to be even more labor intensive. ml) were collected between 9 and 1 1 am and were placed into It occurred to us that if one or more of the relevant enzymes small beakers kept at 4#{176}C.DT1’ was added (final concentration is present in the saliva, it might be the ideal tissue/fluid to look 5 mM), and the samples were centrifuged at 9000 X g and 4#{176}C at. Thus, we searched the literature for reports documenting the for 15 mm. The supemnatant fractions thus obtained were used as presence of at least one of these enzymes in saliva. We were such when enzyme activities present therein were to be quanti- unable to find such a report. However, we did come across an fled. They were further processed when they were to be sub- investigation in which the presence of an aldehyde dehydroge- mitted to SDS-PAGE, immunoblot analysis, isoebectnic focus- ing, and/on when column chromatographic purification of the nase, apparently different from the known class 1, 2, and 3 abdehyde dehydnogenases and dubbed ALDH-V, in human sa- abdehyde dehydnogenase present therein was to be attempted. In those cases, they were desabted with the aid of PD-10 (Sephadex liva was demonstrated (17, 18). Although most of the reported G-25) columns (isoelectnic focusing of abdehyde dehydnoge- catalytic and physical properties of this enzyme did indeed nase) and/or concentrated in Centnicon-lO concentrators (Ami- appear to be different from those exhibited by ALDH-3, one, con Division, W. R. Grace & Co., Danvens, MA) by low-speed namely, isoebectnic point values, was not. Thus, we initiated centrifugation (isoelectnic focusing or column chromatographic experiments designed to further evaluate the identity of the purification of aldehyde dehydrogenase; SDS-PAGE and salivary aldehyde dehydnogenase and found it to indeed be immunobbot analysis of glutathione S-tnansfenases). Desalted ALDH-3. In the course of these investigations, we ascertained 9000 x g supemnatant fractions (obtained from 1.5 ml saliva) that, in addition to ALDH-3, glutathione S-tmansferases and were subjected to affinity chromatography on reactive blue DT-diaphonase were also present in the saliva and that levels of 2-Sepharose CL-6B as described previously (22) to eliminate each of these enzymes varied widely, but in approximately nonspecific proteins, and the resultant preparation was then direct proportion to each other. A review of the gender, age, concentrated as described above when SDS-PAGE/immunobbot mace, ethnicity, and dietary and other habits of the subjects who analysis for the presence of ALDH-3 was attempted. had contributed salivary samples strongly pointed to a direct Blood (15-20 ml) was collected in heparinized syringes and relationship between enzyme activities and coffee consumption. immediately centrifuged at 2000 X g and 4#{176}Cfor 15 mm. Rou- A subsequent review of the literature revealed that coffee con- tinely, the plasma thus obtained was then assayed for enzyme tains demonstrated inducens, e.g. , catechob, of these enzymes activities. In those cases where SDS-PAGE/immunobbot analysis ( 19). Further experimentation revealed that the continuous con- for the presence of ALDH-3 was to be performed, plasma samples sumption of relatively barge amounts of coffee did indeed result were first desabted and subjected to reactive blue 2-Sephanose in markedly increased salivary ALDH-3, glutathione S-trans- CL-6B affinity chromatography, after which the resultant fraction fenase, and DT-diaphorase levels as did the continuous con- of interest was concentrated, all as described above.