The pK values of the Schiff bases of several bacteriorhodopsin (BR) preparations have been determined by titration. While for the native protein a high pK of 13 has been reported [Druckmann et al. (1982) Biochemistry 21, 4953], we find that a BR reconstituted from retinal and the apoprotein obtained from the retinal-deficient strain JW5 exhibits a low pK value, 8.5. When the retinal chromophore is added to growing JW5 cells leading to in vivo BR formation, this BR shows a high Schiff base pK, >/=10.2. A value of 9.3 was determined when BR was reconstituted from retinal and BO, obtained from bleaching BR with hydroxylamine. A low pK value of 8.1 was found when 13-trifluoro(CF3)-retinal was used as chromophore for in vitro reconstitution [Sheves et al. (1986) Proc. Natl. Acad. Sci. U.S.A. 83, 3262], which is confirmed in this study. When we add CF3-retinal to growing JW5 cells, this low pK shifts to 9.1. Besides wild-type protein, the apoprotein from the mutant D96N (from the chromophore-deficient strain L-07) was also used for in vitro reconstitution with either chromophore, retinal or CF3-retinal. Irrespective of the chromophore used, both mutant BRs exhibit low pK values of their Schiff bases of 8.1. Flash photolysis with respect to the rise and decay of the M-photocycle intermediate of wild-type and D96N-mutated BR carrying retinal and CF3-retinal revealed that in both proteins the incorporation of the trifluororetinal leads to a faster rise of the M-intermediate and to a slower decay. Since the apoprotein from the chromophore-deficient JW5 strain of H. salinarium, despite its lower boyant density, is arranged into trimers (according to CD measurements), we propose that the high pK value of the BR Schiff base is induced by long-distance interactions between BR molecules in the purple membrane patches which control the pK of the chromophore.