Influence of substituents in fluorobenzene derivatives on the cytochrome P450-catalyzed hydroxylation at the adjacent ortho aromatic carbon center.

In a previous study, the in vivo cytochrome P450-catalyzed regioselectivity of aromatic ring hydroxylation for a series of (poly)fluorobenzenes could be quantitatively predicted by the calculated frontier orbital density distribution in the aromatic ring [Rietjens et al. (1993) Biochemistry 32, 4801-4812]. However, the relative small fluorine, its size almost comparable to a hydrogen, is not expected to influence the regioselectivity of aromatic hydroxylation due to steric hindrance. The aim of the present study was to investigate the influence of substituents larger than a hydrogen or fluorine on the possibilities for hydroxylation at the adjacent carbon center. First, the in vivo regioselectivity of aromatic ring hydroxylation of a series of C4-substituted fluorobenzenes was investigated. The results obtained demonstrate that a chlorine and cyano C4 substituent do not hamper hydroxylation at the positions ortho to the C4 carbon atom. For 4-chloro- and 4-cyanofluorobenzene, the observed regioselectivity of aromatic hydroxylation correlated with the regioselectivity predicted on the basis of the frontier orbital density distribution. In contrast, a bromine and iodine substituent affected the hydroxylation at the adjacent ortho carbon centers, reducing it to respectively 40 and 6% of the calculated intrinsic reactivity of the carbon centers. Additional experiments were performed to investigate whether the regioselectivity of the aromatic hydroxylation of the C4-substituted fluorobenzene model compounds was influenced by changes in the cytochrome P450 enzyme pattern. Results obtained demonstrate that for these relatively small substrates the regioselectivity of their hydroxylation was not significantly influenced by several cytochrome P450 inducers. This suggests that the active sites of the cytochromes P450 catalyzing the aromatic hydroxylation do not impose a stereoselective orientation of the aromatic rings with respect to the iron-oxo porphyrin reaction center. Thus, the working hypothesis for additional experiments was that the deviations for the regioselectivity of aromatic hydroxylation observed for 4-bromo- and 4-iodofluorobenzene may be ascribed to steric hindrance by the bromine and iodine substituents hampering the attack of the cytochrome P450 iron-oxo species on the adjacent carbon centers in the benzene derivative. This working hypothesis was further tested by investigating whether useful steric correction factors could be derived from the results obtained with the series C4-substituted fluorobenzenes. These correction factors should make it possible to correct calculated relative reactivities of carbon sites for steric hindrance by substituents positioned ortho with respect to the carbon to be hydroxylated. This will make it possible to better explain and predict the regioselectivities for other chlorine-, bromine-, iodine-, and cyano-containing fluorobenzenes. The in vivo regioselectivity of aromatic ring hydroxylation of a series of five chlorine-, bromine-, iodine-, or cyano-containing fluorobenzenes did not correlate with the noncorrected calculated reactivities (r = 0.49). However, upon correction of the calculated reactivity values by using the steric correction factors, a correlation between the observed and calculated regioselectivity for the substrates of the present study was obtained (r = 0.91). Together these results strongly indicate that for the fluorobenzenes studied the main factors directing the regioselectivity of their aromatic hydroxylation are (i) the nucleophilic chemical reactivity of the site to be hydroxylated and (ii) the steric influence of the substituent ortho with respect to the site of hydroxylation. This latter effect appears to be negligible for a fluorine, chlorine, and cyano substituent but significant for a bromine and iodine substituent.