Abstract
A series of five 3-nitroflavones were tested for their ability to inhibit the formation of colon aberrant crypt foci (ACF) induced by a s.c. injection of azoxymethane (C2H6N2O) in rats. Our aim was to relate the electron-donating effects of the 3-nitroflavones as characterized by their Hammett substitution constants with their efficacy in inhibiting ACF. In a first assay (initiation, protocol A) the 3-nitroflavone as well as the 4′-substituted nitro-, methoxy-, fluoro-, and hydroxy-3-nitroflavones were continuously present in the diet. In a second assay (postinitiation, protocol B) they were given for a period of 4 weeks after the last azoxymethane injection. The different substituents of the 3-nitroflavones at the 4′-position spanned a spectrum of Hammett constants (ςp+), going from +0.79 for the electron-withdrawing group, NO2, to −0.92 for the electron-donating group, OH. For both protocols the percentages of inhibition plotted versus the Hammett substitution constants showed a linear correlation, the most efficacious ACF inhibition being produced by the molecules with the most electron-donating substituents. Moreover, the nitroflavones were not only chemoprotective during initiation of the ACF, but also therapeutic in the postinitiation progression assay. The above correlations may be of predictive value in the search for new chemoprotective agents. The overall molecular mechanism of the inhibition of ACF by the 3-nitroflavones under study appears to involve redox reactions.
INTRODUCTION
A number of natural and synthetic compounds have been identified as potential agents against colon tumor development. Several epidemiological studies and clinical trials have demonstrated that treatment with NSAIDs4 caused regression of pre-existing colonic adenomas. The most active chemoprotective agents against colonic precancerous or cancerous lesions investigated in animal models include NSAIDs and retinoids (1), organoselenium compounds (2, 3), and flavonoids (4, 5). The presence of ACF in the colonic mucosa of patients with colon cancer, the expression of mutations in APC and ras, and overexpression of cyclooxygenase-2 and inducible nitric oxide synthase in the aberrant crypts suggest that ACF are precursor lesions from which adenomas and carcinomas develop in the colon. Also, many animal model studies have shown a positive correlation between ACF formation, and colon tumor incidence and multiplicity.
In the present work, five systematically substituted 3-nitroflavones were tested for their ability to inhibit the formation of colon ACF. Our aim was to verify whether the efficacy of chemoprotection against ACF formation and progression is related to the electronic effects of substituents of the chemoprotective agents as characterized by their Hammett substitution constants. The choice of the 3-nitroflavone as well as four substituted 3-nitroflavones (4′-hydroxy, 4′-methoxy, 4′-fluoro, and 4′-nitro) was guided by the fact that the different substituents at position 4′ spanned a spectrum of Hammett constants (ςp+) going from +0.79 for the electron-withdrawing group, NO2, to −0.92 for the electron-donating group, OH. According to our hypothesis, the order of efficacy in inhibiting ACF formation would be related to the nature of the 4′ substituent with OH > OCH3 > F > H > NO2, because this order is related to the decreasing ability of these different substituents to donate electrons, as characterized by their substituent Hammett constants.
MATERIALS AND METHODS
Chemopreventive Agents and Carcinogen.
The 3-nitroflavone and four analogues were synthesized on a multigram scale, using a procedure described previously (6, 7). The four analogues were substituted on the 4′-position of the B ring: 4′-OH, 4′-OCH3, 4′-F, and 4′-NO2 (Fig. 1).
The carcinogen AOM, purchased from Ash Stevens Inc. (Detroit, MI), was injected at a dose level of 15 mg/kg body weight. The negative control animals were injected with saline in place of AOM. The rats were sacrificed by CO2 asphyxiation and the colon removed for measuring ACF. This study was approved by the Institutional Animal Care and Use Committee at the American Health Foundation.
ACF Assay.
The in vivo chemoprotective activity of the five 3-nitroflavones was measured using a rat colon ACF assay, described earlier by Wargovich et al. (1). The timeline of this assay is shown in Fig. 2. In brief, F344 rats, purchased from Charles River Breeding Laboratories (Kingston, NY) were randomized into groups of 10 at 7 weeks of age. All of the rats were fed the AIN-76A diet (Dyets Inc., Bethlehem, PA) for the duration of the experiment. All of the rats were injected at 8 and 9 weeks of age with the carcinogen AOM (15 mg/kg body weight) to induce aberrant colonic crypts. Two protocols were used to determine the effects of the 3-nitroflavones on colon ACF formation (Fig. 2). Protocol A was designed to measure the ability of 3-nitroflavones to block the onset of ACF formation induced by AOM. Protocol B was designed to evaluate the efficacy of 3-nitroflavones to prevent ACF progression after it had been initiated. Under the initiation protocol A, the animals were fed from 7–13 weeks of age with one of the following dose regimen of 3-nitroflavones: 63 ppm, 125 ppm, or no test compound, and sacrificed at the end of 13 weeks of age. Under the postinitiation protocol B, the animals were exposed to carcinogens as in protocol A at week 8 and 9, but were fed the test diets at the same two concentrations between 12 and 16 weeks of age, and sacrificed at the end of 16 weeks of age. In both protocols, the colons were removed and fixed in cold PBS, cut longitudinally, and stained with 0.2% methylene blue (8). The aberrant crypts of the entire colon were scored under a dissecting microscope at ×40 magnification, and the mean number of aberrant crypts per colon determined. The aberrant crypt multiplicities (crypts/colon) at both treatment doses were statistically compared with the carcinogen-only group using the unpaired t test and were based on “Bonferroni’s Correction,” an adjustment of P in multiple comparisons with a single control group. Statistical significance was set at P < 0.05. Correlation coefficients were determined by the least squares method.
RESULTS
The effects of the 3-nitroflavones on AOM-induced ACF formation in rat colon for protocol A and protocol B are reported in Tables 1 and 2, respectively. Two standard concentrations were used throughout for comparative purposes: a low dose of 63 ppm and a high dose of 125 ppm. Neither dose caused measurable weight changes or other toxic effects in these animals over the course of the experiment. A significant inhibition of ACF onset and progression was observed for the electron-donating substituents, 4′-hydroxy, 4′-methoxy, and 4′-fluoro derivatives, the highest inhibition being for the high dose of 125 ppm (∼45 and ∼40% inhibition for protocols A and B, respectively), as compared with the 63 ppm dose (∼33 and 25% inhibition for protocols A and B, respectively). Tables 3 and 4 report the percentage of inhibition of the ACF onset and progression for the protocols A and B, respectively. Fig. 3, A and B, and Fig. 4, A and B, show the corresponding plots for the percentage of inhibition of ACF versus the Hammett constant. For protocol A, where the 3-nitroflavone chemoprotective agent is present at the onset of ACF formation, the correlation coefficients are 0.95 and 0.91 for the 63 and 125 ppm doses, respectively. Under protocol B, where the 3-nitroflavones were administered during the progression phase of the ACFs, the correlation coefficients are 0.88 and 0.89 for the 63 and 125 ppm dose groups, respectively. For protocol A at the 63 ppm dose, the order of efficacy is OH > OCH3 > F > H > NO2 corresponding exactly to the order of the Hammett constants, and the correlation coefficient is the highest as compared with the other three cases. For the other three assays, these orders are slightly different. These differences decrease the correlation coefficients to a minor extent and are attributable to slightly larger experimental errors for the three other assays. We have used the ςp+ constants reported by Hansch and Leo (9). These ςp+ constants, formulated for substituents capable of delocalizing a positive charge, give the best correlations for our data. It can be concluded that the most efficacious derivatives have the best electron-donating substituents.
DISCUSSION
Our hypothesis, that the 3-nitroflavone efficacy in inhibiting ACF formation in vivo would be related to the increasing ability of these different substituents to donate electrons, is confirmed by the high linear correlation between the in vivo inhibition efficacy of the different 3-nitroflavones and their associated Hammett constants, which characterize the ability to donate electrons (9, 10, 11). The regression lines for data of protocols A and B give four equations (correlation coefficients):
Protocol A (63 ppm): % of ACF onset = 15.2 − 18.3 ςp+ (r = 0.95)
Protocol A (125 ppm): % of ACF onset = 25.9 − 20.7 ςp+ (r = 0.91)
Protocol B (63 ppm): % of ACF progression = 13 − 15.3 ςp+ (r = 0.88)
Protocol B (125 ppm): % of ACF progression = 22.0 − 23 ςp+ (r = 0.98)
From these equations it is possible to predict that for better electron-donating substituents such as −N(CH3)2 or −N(CH2CH3)2, which have more negative Hammett constants (−1.70 and −2.07, respectively) the percentage of inhibition might reach 60–70% in the case of the most efficacious 125 ppm regimen.
A comparison between the chemoprotection by the 3-nitroflavones in the initiation protocol A and that observed in the postinitiation protocol B (Tables 3 and 4) shows that the most active molecules are the same in both protocols and correspond to those with the most negative Hammett constants relating directly to the best electron-donating substituents. Moreover, the similar degree of inhibition reached in both protocols A and B underscores the therapeutic effect as well as the protective effect of these agents.
It is important to link our correlations for ACF inhibition efficacy with similar linear correlations found previously for other redox-related chemoprotection studies. These earlier correlations showed that more easily oxidized agents are more efficacious. The agents studied include: (a) a series of phenols able to inhibit benzo(a)pyrene-induced neoplasia in the forestomach of mice (12); (b) a series of diphenols that induce production of cancer protective enzymes (13); and (c) a series of NSAIDs acting in vitro and in vivo (14, 15, 16). It is important to emphasize that NSAIDs show promise in the chemoprevention of cancer (17, 18, 19).
Such correlations dealing with chemoprotection redox mechanisms may be related to the redox-sensing of gene transcription (20, 21, 22). Moreover, metabolism of xenobiotics is often ruled by redox mechanisms, as shown by the following: (a) by the correlation of the rates of reduction of 27 molecules by purified cytochrome P-450 reductase with the one-electron reduction potentials of these compounds (23); or (b) by the demonstration that cytochrome P-450 mediates covalent binding of substrates to DNA via one-electron oxidation (24).
In conclusion, the correlations reported in the present work should help the search for and identification of more efficient chemoprotectors. Moreover, our present correlations underscore the link between redox mechanisms of ACF inhibition in the colon with the redox correlations observed in: (a) the chemoprotection by phenols against benzo(a)pyrene-induced neoplasia in the forestomach of mice (12); (b) in the activation by diphenols of chemoprotective enzyme synthesis (13); and (c) in the anti-inflammatory process (14, 15, 16). The latter observations suggest that in our case the mechanisms of retardation might involve enhancing of multiple detoxification systems as well as reduction of AOM-induced inflammation. Flavones are known to inhibit the inflammatory process (25, 26). Thus, the electron-donating molecules such as these nitroflavones may slow or block the progression phase (protocol B). For a better understanding of the molecular mechanisms of cancer chemoprotection, physicochemical properties of the many compounds already identified as efficient agents (27, 28) should be ranked versus physicochemical parameters, which include their redox properties. Such ranking would allow us to predict the cancer protective efficacy of certain classes of chemical agents and to approach some biochemical mechanisms of carcinogenesis and its chemoprotection.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Supported by National Cancer Institute Contract N01-CN-65108.
The abbreviations used are: NSAID, non-steroidal anti-inflammatory drug; ACF, aberrant crypt foci; AOM, azoxymethane.
Compound . | Total ACF/rat ± SE . | . | |
---|---|---|---|
. | Concentration of 3-nitroflavones in diet . | . | |
. | 63 ppm . | 125 ppm . | |
Control (no nitroflavone) | 94 ± 8.7 | 94 ± 8.7 | |
3–4′-Dinitroflavone | 96 ± 10 | 88 ± 9.3 | |
3-Nitroflavone | 80 ± 8.2 | 73 ± 7.9 | |
4′-Fluoro-3-nitroflavone | 72 ± 7.3 | 58 ± 6.2a | |
4′-Methoxy-3-nitroflavone | 70 ± 7.5 | 59 ± 6.8a | |
4′-Hydroxy-3-nitroflavone | 63 ± 7.0b | 51 ± 6.1a |
Compound . | Total ACF/rat ± SE . | . | |
---|---|---|---|
. | Concentration of 3-nitroflavones in diet . | . | |
. | 63 ppm . | 125 ppm . | |
Control (no nitroflavone) | 94 ± 8.7 | 94 ± 8.7 | |
3–4′-Dinitroflavone | 96 ± 10 | 88 ± 9.3 | |
3-Nitroflavone | 80 ± 8.2 | 73 ± 7.9 | |
4′-Fluoro-3-nitroflavone | 72 ± 7.3 | 58 ± 6.2a | |
4′-Methoxy-3-nitroflavone | 70 ± 7.5 | 59 ± 6.8a | |
4′-Hydroxy-3-nitroflavone | 63 ± 7.0b | 51 ± 6.1a |
Significant inhibition at P < 0.01–0.001.
Significant inhibition at P < 0.05.
Compound . | Total ACF/rat ± SEM . | . | |
---|---|---|---|
. | Concentration of 3-nitroflavones in diet . | . | |
. | 63 ppm . | 125 ppm . | |
Control (no nitroflavone) | 132 ± 11 | 132 ± 11 | |
3–4′-Dinitroflavone | 130 ± 13 | 131 ± 12 | |
3-Nitroflavone | 126 ± 12 | 112 ± 11 | |
4′-Fluoro-3-nitroflavone | 103 ± 10 | 83 ± 8.8a | |
4′-Methoxy-3-nitroflavone | 96 ± 9.4b | 79 ± 7.1a | |
4′-Hydroxy-3-nitroflavone | 99 ± 10 | 80 ± 7.8a |
Compound . | Total ACF/rat ± SEM . | . | |
---|---|---|---|
. | Concentration of 3-nitroflavones in diet . | . | |
. | 63 ppm . | 125 ppm . | |
Control (no nitroflavone) | 132 ± 11 | 132 ± 11 | |
3–4′-Dinitroflavone | 130 ± 13 | 131 ± 12 | |
3-Nitroflavone | 126 ± 12 | 112 ± 11 | |
4′-Fluoro-3-nitroflavone | 103 ± 10 | 83 ± 8.8a | |
4′-Methoxy-3-nitroflavone | 96 ± 9.4b | 79 ± 7.1a | |
4′-Hydroxy-3-nitroflavone | 99 ± 10 | 80 ± 7.8a |
Significant inhibition at P < 0.01–0.001.
Significant inhibition at P < 0.05.
Compound . | % Inhibition (±SE) . | . | Hammett constant ςp+ . | |
---|---|---|---|---|
. | Concentration in diet . | . | . | |
. | 63 ppm . | 125 ppm . | . | |
3–4′-Dinitroflavone | −2.1 ± 13.3 | 6.4 ± 12.7 | +0.79 | |
3-Nitroflavone | 14.8 ± 12.0 | 22.3 ± 11.8 | 0 | |
4′-Fluoro-3-nitroflavone | 23.4 ± 11.4 | 38.3 ± 10.7 | −0.07 | |
4′-Methoxy-3-nitroflavone | 25.5 ± 11.5 | 37.2 ± 11.0 | −0.78 | |
4′-Hydroxy-3-nitroflavone | 32.9 ± 11.2 | 45.7 ± 10.6 | −0.92 |
Compound . | % Inhibition (±SE) . | . | Hammett constant ςp+ . | |
---|---|---|---|---|
. | Concentration in diet . | . | . | |
. | 63 ppm . | 125 ppm . | . | |
3–4′-Dinitroflavone | −2.1 ± 13.3 | 6.4 ± 12.7 | +0.79 | |
3-Nitroflavone | 14.8 ± 12.0 | 22.3 ± 11.8 | 0 | |
4′-Fluoro-3-nitroflavone | 23.4 ± 11.4 | 38.3 ± 10.7 | −0.07 | |
4′-Methoxy-3-nitroflavone | 25.5 ± 11.5 | 37.2 ± 11.0 | −0.78 | |
4′-Hydroxy-3-nitroflavone | 32.9 ± 11.2 | 45.7 ± 10.6 | −0.92 |
Compound . | % Inhibition (±SE) . | . | Hammett constant ςp+ . | |
---|---|---|---|---|
. | Concentration in diet . | . | . | |
. | 63 ppm . | 125 ppm . | . | |
3–4′-Dinitroflavone | 1.5 ± 16.9 | 0.8 ± 16.3 | +0.79 | |
3-Nitroflavone | 4.6 ± 16.6 | 15.2 ± 15.7 | 0 | |
4′-Fluoro-3-nitroflavone | 21.9 ± 14.9 | 37.1 ± 14.4 | −0.07 | |
4′-Methoxy-3-nitroflavone | 27.3 ± 14.8 | 40.2 ± 13.4 | −0.78 | |
4′-Hydroxy-3-nitroflavone | 25.0 ± 15.4 | 39.4 ± 13.8 | −0.92 |
Compound . | % Inhibition (±SE) . | . | Hammett constant ςp+ . | |
---|---|---|---|---|
. | Concentration in diet . | . | . | |
. | 63 ppm . | 125 ppm . | . | |
3–4′-Dinitroflavone | 1.5 ± 16.9 | 0.8 ± 16.3 | +0.79 | |
3-Nitroflavone | 4.6 ± 16.6 | 15.2 ± 15.7 | 0 | |
4′-Fluoro-3-nitroflavone | 21.9 ± 14.9 | 37.1 ± 14.4 | −0.07 | |
4′-Methoxy-3-nitroflavone | 27.3 ± 14.8 | 40.2 ± 13.4 | −0.78 | |
4′-Hydroxy-3-nitroflavone | 25.0 ± 15.4 | 39.4 ± 13.8 | −0.92 |
Acknowledgments
We thank Dr. Charles Mize for helpful discussions and constructive criticism. We also thank Drs. Jerome Goodman and Gabriel Laury for providing fruitful suggestions, and Jeff Rigotty and the technical staff of the American Health Foundation Research Animal Facility for expert handling of these efficacy studies.