- Email: [email protected]
Modulation of tumor promotion in mouse skin by the food additive citral (3,7-dimethyl-2,6-octadienal)
Cancer Letters, 56 (1991) 25-28 Elsevier Scientific Publishers Ireland
25 Ltd.
Modulation of tumor promotion in mouse skin by the food additive citral (3,7-dimethyl,6-octadienal) M. J. Connor Dioision of Dermatology,
Department
of Medicine,
UCLA School
of Medicine,
Los Angeles,
CA 90024 (CJ.S.A.J
(Received 11 September 1990) (Revision received 18 October 1990) (Accepted 19 October 1990)
Sammary Citral inhibits the formation of retinoic acid from retinol in mouse epidermis. Since skincarcinogenesis is sensitioe to retinoid status, and retinoic acid may be the actioe form of vitamin A in the epidermis, citral was tested for its ability to modulate tumor promotion in a two-stage skin-carcinogenesis study in hairless mice. The dorsal skins of female skh/hrl mice were initiated with 0.1 pmol dimethylbenzanthracene, and tumors were promoted by twice-weekly application of 10 nmol of tetradecanoylphorbol-13-acetate (TPA) for 20 weeks. Prior to each TPA application groups were dosed with 0, 1 pmol, or 10 pmol citral. Citral had a dosedependent inhibitory effect on tumor-production in the TPA promoted groups. At 10 weeks of promotion the percentage of mice with tumors were 88%, 72% and 60%, for the 0,l and 10 pmol citral treated groups, and the numbers (mean f S. D.) of tumors per affected animal were 7.3 f 6.6, 3.9 A 4.2, and3.7 f 3.5, respectively. At 15 weeks of promotion the tumor incidence was 96%, 96% and 84%, respectiuely, and the number of tumors per ajfected animal were 9.5 f 6.8, 7.2 f 4.6 and
Correspondence to: M.J. Connor, Division of Dermatology, Department of Medicine, UCLA School of Medicine, Los Angeles, CA 90024 U.S.A.
0304-3835/91/$03.50 Published
0 1991 Elsevier Scientific Publishers
and Printed in Ireland
4.5 f in the fewer at 20
3.3, respectiuely. Again the affected mice high dose citral group had significantly tumors. When the study was terminated weeks ofpromotion all mice had at least
one tumor, but the number of tumors per afjetted mouse were lower in the citral treated groups.
These findings concur with the proposal that there is a retinoid requirement for skin tumor
promotion, have
and establishes
potential
uses
that anti-retinoids
as modulators
of car-
cinogenesis. Keywords: citral; tumor-promotion; A; retinoic acid; carcinogenesis
vitamin
Introduction The well established murine two-stage skincartinogenesis model has been used extensively to study the relationships between retinoids and cancer. Recently, results of a complete study of the impact of vitamin A deprivation on skin carcinogenesis in this system were published [5]. Because these authors observed that skin tumor promotion was inhibited in severely vitamin Adeficient mice and that tumors appeared when the mice were re-fed retinyl palmitate or the provitamin beta-carotene, they suggested that vitamin A is required for skin tumor-promotion. Prior studies of the effects of exogenous retinoids Ireland Ltd.
26
in the murine skin carcinogenesis and related model systems have yielded equivocal results. In some cases retinoids applied to the skin during the promoting stage have inhibited tumorpromotion [ll]. However, in other similar studies the opposite response has occurred, and retinoids have enhanced promotion or acted as tumor promoters [6--81. Since retinoic acid has been reported to promote [6,7] and to inhibit promotion [lo] in both CD1 and SENCAR mice; and orally administered retinoids have been reported to enhance [8] and to inhibit [9] promotion in studies reported by the same investigators, differences in mouse strain, the chemical structure of the retinoid used, route of administration of the retinoid (oral vs. topical), or the laboratory conducting the study appear to be unlikely reasons for these discrepancies. Retinoids also exhibit apparently paradoxical behavior in predictive assays for tumor promoters, such as their ability to inhibit the induction of and to induce murine epidermal omithine decarboxylase activity [3]. In this case inhibition of ornithine decarboxylase induction depends upon the direct presence of the retinoid. Retinoid induced ornithine decarboxylase activity is only detected at later times, when the epidermal retinoid concentration has diminished, and is probably associated with retinoid induced epiderma1 hyperplasia [2]. Citral (3,7-dimethyl-2,6-octadienal) is widely employed as a flavor and fragrance additive by the soap, cosmetic and perfume industries. Locally applied citral inhibited retinoic acid formation from retinol in mouse epidermis in vivo 141, and acted as a specific retinol antagonist in two bioassays in mouse epidermis, having no effect on the activity of retinoic acid in the same assays [ 11. Retinoic acid formation may therefore be required for vitamin A activity in the skin. Since skin-carcinogenesis is sensitive to retinoid status, and retinoic acid may be the active metabolite of vitamin A in the epidermis, we tested the ability of citral to modulate tumor promotion in a 2 stage skin-carcinogenesis study in hairless mice. Citral inhibited tumor promotion in a dose dependent manner in this system, suggesting that inhibitors of retinoic acid synthesis
may have potential agents. Materials
as anti-tumor-promoting
and Methods
Mice and carcinogenesis protocol Female &h/h1 mice were obtained from the Skin and Cancer Hospital (Temple University, Philadelphia, PA). They were housed five/cage in the UCLA Division of Laboratory Medicine, on a 12-h dark/light cycle and were fed Wayne Rodent Blox and tap water ad libitum. The mice were initiated by the application of a single dose of 0.1 pmol dimethylbenzanthracene (DMBA) in 0.1 ml acetone to their dorsal skins, when they were 7-8 weeks of age. The promotion phase was begun 1 week later, with twice weekly applications of the tumor promoter TPA (10 nmol in 0.1 ml acetone) to the dorsal skins. Citral solutions in acetone were made up freshly just before use, and were applied to the dorsal skins 1 h prior to TPA. The initiated mice were allocated to one of five treatment groups, each containing 25 mice. Group 1 received vehicle and TPA. Group 2 received 1 pmol citral and TPA. Group 3 received 10 prnol citral and TPA. Group 4 received 10 pmol citral and vehicle. Group 5 were initiated only. During the promotion phase of the study the mice were weighed weekly. They were carefully examined once a week for the presence of skin tumors, and the number of tumors present on each affected mouse was recorded. Skin tumors were defined as lesions with a diameter greater than 1 mm that were present for at least two consecutive observations. Analysis of data Differences in the percentage of mice in each group with tumors, the number of tumors per treatment group, and the mean number of tumors per affected mouse for the promoted groups treated with 0, 1 pmol, or 10 pmol citral throughout the promotion phase were assessed by analysis of variance. Differences in tumor incidence at specified times were determined using the Fisher Exact test, and in tumor load and size by Student’s t-test.
Histology Representative mice from each group were killed at the completion of the study, and skin tumors and punch biopsies of skin were recovered, fixed in neutral buffered formalin, and processed for histological evaluation.
10
8 8 I
Results
t a
Skin tumors did not occur on the mice that were initiated only or on initiated mice that were treated with citral only. This confirms that the dose of DMBA used was a sub-carcinogenic initiating dose, and establishes that citral is not a skin-tumor promoter. No animals in any of the treatment groups died during the study. Chronic treatment with citral was well tolerated by the mice, and no gross or histological changes attributable to citral were observed in the skin. All the groups that were promoted with TPA developed tumors. Histological evaluation of representative tumors indicated that the tumors were all of the pedunculated papilloma type, and the tumors varied in size from 2-35 mm in diameter. Figures 1 and 2 show the percentage of mice with tumors and the number of tumors r
100
60
t
20
00 0
a
5
10
20
Weeks of Promotion
Fig. 1. Effect of citral on the percentage of mice with tumors during promotion with TPA. Groups were treated with 0 (-0 o-_), 1 pmol (-0 l-), or 10 pmol (-V V -) citral.
6
Weeks of Promotion
Fig. 2. Effect of citral on the number of tumors per mouse during promotion with TPA. Groups were treated with 0 (-0 o-), 1 pmol (-0 l-), or 10 pmol (- v v -) citral.
per affected mouse respectively, over the 20-week promotion period for the TPA promoted groups dosed with 0, 1 or 10 prnol citral. Analysis of variance of the number of tumors occurring over the 20-week period revealed that the differences between the three groups promoted with TPA and receiving 0, 1, and 10 prnol citral were significant for the number of tumors per group (F ratio = 63.4, P c 0.01) and the number of tumors per affected animal (F ratio = 48.6, P < 0.01). Although the dose of TPA used was sufficient to promote tumors in all animals by 20 weeks, tumor appearance was retarded in the citral treated groups compared to controls (see Fig. 2). At 10 weeks of promotion the percentages of mice with tumors (incidence) were 88%) 72% and 60%, for the 0, 1 and 10 pmol citral treated groups respectively, and the incidences in the 1 pmol and 10 pmol citral treated groups were significantly lower (P < 0.05, Fisher Exact test). At 15 weeks of promotion the percentage of mice with tumors were 96 % ,96 % and 84 % , respectively. Again, the affected mice in the high dose citral group had significantly fewer tumors (P < 0.05,
28
Fisher Exact test). By 20 weeks of promotion all the mice in the TPA promoted groups had at least one tumor, the mean number of tumors per mouse being 8.3 f 5.5, 6.6 f 4.4, and 4.4 f 3.1 for the 0, 1 and 10 pmol citral treated groups, respectively.
cellent technical assistance. This research was supported by a grant (CA 47758) from the National Cancer Institute. References 1
2
Connor, M.J. (1988) Oxidation of retinol to retinoic acid as a requirement for biological activity in the epidermis. Cancer Res., 48, 7038-7040. Connor, M. J., Ashton, A.E. and Lowe, N.J. (1986) A com-
3
parative study of the induction of epidermal hyperplasia by natural and synthetic retinoids. J. Pharmacol. Exp. Ther., 237, 31-35. Connor, M.J. and Lowe, N.J. (1983) The induction of or-
4
nithine decarboxylase activity and DNA synthesis in hairless mouse epidermis by retinoids. Cancer Res., 43, 5174-5177. Connor, M.J. and Smit, M.H. (1987) Terminal group ox-
Discussion We have shown that topically applied citral has an inhibitory effect on tumor promotion in hairless mouse skin. Citral is used as food and cosmetic additive for its lemon flavor and scent, and occurs naturally in high concentrations in certain foodstuffs such as lemon grass. Our interest in testing citral in the two stage carcinogenesis system arose because there is evidence that links retinoid status and skin tumor promotion (see Introduction), and citral antagonized the ability of retinol to inhibit TPA induction of epidermal ornithine decarboxylase activity, a potential marker for tumor-promotion. Our initial expectation was that by inhibiting retinol function, citral might enhance tumor promotion [l]. However, while the study was in progress, definitive evidence appeared showing that skin tumor promotion is depressed in vitamin A deficiency, suggesting a possible requirement for vitamin A at a crucial step in promotion [5]. Our demonstration that citral can modulate the yield of tumors in the murine two stage skin carcinogenesis system concurs with the latter finding that vitamin A deficiency has an inhibitory impact on tumor promotion. While the results of this study show that citral is a naturally occurring anti-promoter, further studies are required to clearly elucidate the mechanism of action of citral in this role. Moreover, these results indicate that inhibitors of retinoic acid synthesis and activity have a potential role as antitumor-promoting agents. Acknowledgements The author thanks Jacob S. Lim for his ex-
5
6
7
8
9
10
11
idation of retinol by mouse epidermis; inhibition in vitro and in ho. Biochem. J., 244, 489-492. De Luca, L.M., Shores, R.L., Spangler, E.F. and Wenk, M. (1989) Inhibition of initiator-promoter-induced skin tumorigenesis in female SENCAR mice fed a vitamin Adeficient diet and reappearance of tumors in mice fed a diet adequate in retinoid or carotene. Cancer 5400-5406. Fischer, S.M., Klein-Szanto, A.J.P., Adams, Slaga, T.J. (1985) The first stage and complete
Res.,
49,
L.M. and promoting
activity of retinoic acid but not the analog Ro 10-9359. Carcinogenesis 6, 575-578. Hennings, H., Wenk, M.L. and Donahoe, R. (1982) Retinoic acid promotion of papilloma formation in mouse skin. Cancer Lett., 16, l-5. McCormick, D.L., Bagg, B.J. and Hultin, T.A. (1987) Comparative activity of dietary or topical exposure to three retinoids in the promotion of skin tumor induction in mice. Cancer Res.. 47, 5989-5993. McCormick, D.L. and Moon, R.C. (1986) Antipromotional activity of dietary N-(4-hydroxyphenylhetinamide in twostage skin tumorigenesis in CD-l and SENCAR mice. Cancer Lett., 31, 133-138. Verma A.K. (1987) Inhibition of both stage I and stage II mouse skin tumor promotion by retinoic acid and the dependence of inhibition of tumor promotion on the duration of retinoic acid treatment. Cancer Res., 47, 5097-5101. Verma, A.K., Shapas, B.G., Rice, H.M. and Boutwell, R. K. (1979) Correlation of the inhibition by retinoids of tumor promoter-induced mouse epidermal ornithine decarboxylase activity and of skin tumor promotion. Cancer Res., 39, 419-425.
25 Ltd.
Modulation of tumor promotion in mouse skin by the food additive citral (3,7-dimethyl,6-octadienal) M. J. Connor Dioision of Dermatology,
Department
of Medicine,
UCLA School
of Medicine,
Los Angeles,
CA 90024 (CJ.S.A.J
(Received 11 September 1990) (Revision received 18 October 1990) (Accepted 19 October 1990)
Sammary Citral inhibits the formation of retinoic acid from retinol in mouse epidermis. Since skincarcinogenesis is sensitioe to retinoid status, and retinoic acid may be the actioe form of vitamin A in the epidermis, citral was tested for its ability to modulate tumor promotion in a two-stage skin-carcinogenesis study in hairless mice. The dorsal skins of female skh/hrl mice were initiated with 0.1 pmol dimethylbenzanthracene, and tumors were promoted by twice-weekly application of 10 nmol of tetradecanoylphorbol-13-acetate (TPA) for 20 weeks. Prior to each TPA application groups were dosed with 0, 1 pmol, or 10 pmol citral. Citral had a dosedependent inhibitory effect on tumor-production in the TPA promoted groups. At 10 weeks of promotion the percentage of mice with tumors were 88%, 72% and 60%, for the 0,l and 10 pmol citral treated groups, and the numbers (mean f S. D.) of tumors per affected animal were 7.3 f 6.6, 3.9 A 4.2, and3.7 f 3.5, respectively. At 15 weeks of promotion the tumor incidence was 96%, 96% and 84%, respectiuely, and the number of tumors per ajfected animal were 9.5 f 6.8, 7.2 f 4.6 and
Correspondence to: M.J. Connor, Division of Dermatology, Department of Medicine, UCLA School of Medicine, Los Angeles, CA 90024 U.S.A.
0304-3835/91/$03.50 Published
0 1991 Elsevier Scientific Publishers
and Printed in Ireland
4.5 f in the fewer at 20
3.3, respectiuely. Again the affected mice high dose citral group had significantly tumors. When the study was terminated weeks ofpromotion all mice had at least
one tumor, but the number of tumors per afjetted mouse were lower in the citral treated groups.
These findings concur with the proposal that there is a retinoid requirement for skin tumor
promotion, have
and establishes
potential
uses
that anti-retinoids
as modulators
of car-
cinogenesis. Keywords: citral; tumor-promotion; A; retinoic acid; carcinogenesis
vitamin
Introduction The well established murine two-stage skincartinogenesis model has been used extensively to study the relationships between retinoids and cancer. Recently, results of a complete study of the impact of vitamin A deprivation on skin carcinogenesis in this system were published [5]. Because these authors observed that skin tumor promotion was inhibited in severely vitamin Adeficient mice and that tumors appeared when the mice were re-fed retinyl palmitate or the provitamin beta-carotene, they suggested that vitamin A is required for skin tumor-promotion. Prior studies of the effects of exogenous retinoids Ireland Ltd.
26
in the murine skin carcinogenesis and related model systems have yielded equivocal results. In some cases retinoids applied to the skin during the promoting stage have inhibited tumorpromotion [ll]. However, in other similar studies the opposite response has occurred, and retinoids have enhanced promotion or acted as tumor promoters [6--81. Since retinoic acid has been reported to promote [6,7] and to inhibit promotion [lo] in both CD1 and SENCAR mice; and orally administered retinoids have been reported to enhance [8] and to inhibit [9] promotion in studies reported by the same investigators, differences in mouse strain, the chemical structure of the retinoid used, route of administration of the retinoid (oral vs. topical), or the laboratory conducting the study appear to be unlikely reasons for these discrepancies. Retinoids also exhibit apparently paradoxical behavior in predictive assays for tumor promoters, such as their ability to inhibit the induction of and to induce murine epidermal omithine decarboxylase activity [3]. In this case inhibition of ornithine decarboxylase induction depends upon the direct presence of the retinoid. Retinoid induced ornithine decarboxylase activity is only detected at later times, when the epidermal retinoid concentration has diminished, and is probably associated with retinoid induced epiderma1 hyperplasia [2]. Citral (3,7-dimethyl-2,6-octadienal) is widely employed as a flavor and fragrance additive by the soap, cosmetic and perfume industries. Locally applied citral inhibited retinoic acid formation from retinol in mouse epidermis in vivo 141, and acted as a specific retinol antagonist in two bioassays in mouse epidermis, having no effect on the activity of retinoic acid in the same assays [ 11. Retinoic acid formation may therefore be required for vitamin A activity in the skin. Since skin-carcinogenesis is sensitive to retinoid status, and retinoic acid may be the active metabolite of vitamin A in the epidermis, we tested the ability of citral to modulate tumor promotion in a 2 stage skin-carcinogenesis study in hairless mice. Citral inhibited tumor promotion in a dose dependent manner in this system, suggesting that inhibitors of retinoic acid synthesis
may have potential agents. Materials
as anti-tumor-promoting
and Methods
Mice and carcinogenesis protocol Female &h/h1 mice were obtained from the Skin and Cancer Hospital (Temple University, Philadelphia, PA). They were housed five/cage in the UCLA Division of Laboratory Medicine, on a 12-h dark/light cycle and were fed Wayne Rodent Blox and tap water ad libitum. The mice were initiated by the application of a single dose of 0.1 pmol dimethylbenzanthracene (DMBA) in 0.1 ml acetone to their dorsal skins, when they were 7-8 weeks of age. The promotion phase was begun 1 week later, with twice weekly applications of the tumor promoter TPA (10 nmol in 0.1 ml acetone) to the dorsal skins. Citral solutions in acetone were made up freshly just before use, and were applied to the dorsal skins 1 h prior to TPA. The initiated mice were allocated to one of five treatment groups, each containing 25 mice. Group 1 received vehicle and TPA. Group 2 received 1 pmol citral and TPA. Group 3 received 10 prnol citral and TPA. Group 4 received 10 pmol citral and vehicle. Group 5 were initiated only. During the promotion phase of the study the mice were weighed weekly. They were carefully examined once a week for the presence of skin tumors, and the number of tumors present on each affected mouse was recorded. Skin tumors were defined as lesions with a diameter greater than 1 mm that were present for at least two consecutive observations. Analysis of data Differences in the percentage of mice in each group with tumors, the number of tumors per treatment group, and the mean number of tumors per affected mouse for the promoted groups treated with 0, 1 pmol, or 10 pmol citral throughout the promotion phase were assessed by analysis of variance. Differences in tumor incidence at specified times were determined using the Fisher Exact test, and in tumor load and size by Student’s t-test.
Histology Representative mice from each group were killed at the completion of the study, and skin tumors and punch biopsies of skin were recovered, fixed in neutral buffered formalin, and processed for histological evaluation.
10
8 8 I
Results
t a
Skin tumors did not occur on the mice that were initiated only or on initiated mice that were treated with citral only. This confirms that the dose of DMBA used was a sub-carcinogenic initiating dose, and establishes that citral is not a skin-tumor promoter. No animals in any of the treatment groups died during the study. Chronic treatment with citral was well tolerated by the mice, and no gross or histological changes attributable to citral were observed in the skin. All the groups that were promoted with TPA developed tumors. Histological evaluation of representative tumors indicated that the tumors were all of the pedunculated papilloma type, and the tumors varied in size from 2-35 mm in diameter. Figures 1 and 2 show the percentage of mice with tumors and the number of tumors r
100
60
t
20
00 0
a
5
10
20
Weeks of Promotion
Fig. 1. Effect of citral on the percentage of mice with tumors during promotion with TPA. Groups were treated with 0 (-0 o-_), 1 pmol (-0 l-), or 10 pmol (-V V -) citral.
6
Weeks of Promotion
Fig. 2. Effect of citral on the number of tumors per mouse during promotion with TPA. Groups were treated with 0 (-0 o-), 1 pmol (-0 l-), or 10 pmol (- v v -) citral.
per affected mouse respectively, over the 20-week promotion period for the TPA promoted groups dosed with 0, 1 or 10 prnol citral. Analysis of variance of the number of tumors occurring over the 20-week period revealed that the differences between the three groups promoted with TPA and receiving 0, 1, and 10 prnol citral were significant for the number of tumors per group (F ratio = 63.4, P c 0.01) and the number of tumors per affected animal (F ratio = 48.6, P < 0.01). Although the dose of TPA used was sufficient to promote tumors in all animals by 20 weeks, tumor appearance was retarded in the citral treated groups compared to controls (see Fig. 2). At 10 weeks of promotion the percentages of mice with tumors (incidence) were 88%) 72% and 60%, for the 0, 1 and 10 pmol citral treated groups respectively, and the incidences in the 1 pmol and 10 pmol citral treated groups were significantly lower (P < 0.05, Fisher Exact test). At 15 weeks of promotion the percentage of mice with tumors were 96 % ,96 % and 84 % , respectively. Again, the affected mice in the high dose citral group had significantly fewer tumors (P < 0.05,
28
Fisher Exact test). By 20 weeks of promotion all the mice in the TPA promoted groups had at least one tumor, the mean number of tumors per mouse being 8.3 f 5.5, 6.6 f 4.4, and 4.4 f 3.1 for the 0, 1 and 10 pmol citral treated groups, respectively.
cellent technical assistance. This research was supported by a grant (CA 47758) from the National Cancer Institute. References 1
2
Connor, M.J. (1988) Oxidation of retinol to retinoic acid as a requirement for biological activity in the epidermis. Cancer Res., 48, 7038-7040. Connor, M. J., Ashton, A.E. and Lowe, N.J. (1986) A com-
3
parative study of the induction of epidermal hyperplasia by natural and synthetic retinoids. J. Pharmacol. Exp. Ther., 237, 31-35. Connor, M.J. and Lowe, N.J. (1983) The induction of or-
4
nithine decarboxylase activity and DNA synthesis in hairless mouse epidermis by retinoids. Cancer Res., 43, 5174-5177. Connor, M.J. and Smit, M.H. (1987) Terminal group ox-
Discussion We have shown that topically applied citral has an inhibitory effect on tumor promotion in hairless mouse skin. Citral is used as food and cosmetic additive for its lemon flavor and scent, and occurs naturally in high concentrations in certain foodstuffs such as lemon grass. Our interest in testing citral in the two stage carcinogenesis system arose because there is evidence that links retinoid status and skin tumor promotion (see Introduction), and citral antagonized the ability of retinol to inhibit TPA induction of epidermal ornithine decarboxylase activity, a potential marker for tumor-promotion. Our initial expectation was that by inhibiting retinol function, citral might enhance tumor promotion [l]. However, while the study was in progress, definitive evidence appeared showing that skin tumor promotion is depressed in vitamin A deficiency, suggesting a possible requirement for vitamin A at a crucial step in promotion [5]. Our demonstration that citral can modulate the yield of tumors in the murine two stage skin carcinogenesis system concurs with the latter finding that vitamin A deficiency has an inhibitory impact on tumor promotion. While the results of this study show that citral is a naturally occurring anti-promoter, further studies are required to clearly elucidate the mechanism of action of citral in this role. Moreover, these results indicate that inhibitors of retinoic acid synthesis and activity have a potential role as antitumor-promoting agents. Acknowledgements The author thanks Jacob S. Lim for his ex-
5
6
7
8
9
10
11
idation of retinol by mouse epidermis; inhibition in vitro and in ho. Biochem. J., 244, 489-492. De Luca, L.M., Shores, R.L., Spangler, E.F. and Wenk, M. (1989) Inhibition of initiator-promoter-induced skin tumorigenesis in female SENCAR mice fed a vitamin Adeficient diet and reappearance of tumors in mice fed a diet adequate in retinoid or carotene. Cancer 5400-5406. Fischer, S.M., Klein-Szanto, A.J.P., Adams, Slaga, T.J. (1985) The first stage and complete
Res.,
49,
L.M. and promoting
activity of retinoic acid but not the analog Ro 10-9359. Carcinogenesis 6, 575-578. Hennings, H., Wenk, M.L. and Donahoe, R. (1982) Retinoic acid promotion of papilloma formation in mouse skin. Cancer Lett., 16, l-5. McCormick, D.L., Bagg, B.J. and Hultin, T.A. (1987) Comparative activity of dietary or topical exposure to three retinoids in the promotion of skin tumor induction in mice. Cancer Res.. 47, 5989-5993. McCormick, D.L. and Moon, R.C. (1986) Antipromotional activity of dietary N-(4-hydroxyphenylhetinamide in twostage skin tumorigenesis in CD-l and SENCAR mice. Cancer Lett., 31, 133-138. Verma A.K. (1987) Inhibition of both stage I and stage II mouse skin tumor promotion by retinoic acid and the dependence of inhibition of tumor promotion on the duration of retinoic acid treatment. Cancer Res., 47, 5097-5101. Verma, A.K., Shapas, B.G., Rice, H.M. and Boutwell, R. K. (1979) Correlation of the inhibition by retinoids of tumor promoter-induced mouse epidermal ornithine decarboxylase activity and of skin tumor promotion. Cancer Res., 39, 419-425.