during the photocycle of bacteriorhodopsin - Europe PMC

1 downloads 0 Views 682KB Size Report
Jan 17, 1985 - Department of Chemistry and Biochemistry, University of California, 405 Hilgard Avenue, Los Angeles, CA 90024. Contributed by M.A. ...
Proc. Natd. Acad. Sci. USA Vol. 82, pp. 3662-3664, June 1985 Biophysics

Importance of bound divalent cations to the tyrosine deprotonation during the photocycle of bacteriorhodopsin (deionized bacteriorhodopsln/acidifled bacteriorhodopsin/blue membrane/tyrosinate formation/trans-is isomerization)

PAUL Dupuis, TIMOTHY C. CORCORAN, AND M. A. EL-SAYED Department of Chemistry and Biochemistry, University of California, 405 Hilgard Avenue, Los Angeles, CA 90024

Contributed by M. A. El-Sayed, January 17, 1985

(M412 formation) (28). The ratio of the amplitudes of the fast and slow M412 components is very sensitive to pH and, at any pH, can be used to calculate a pKa value of about 9.6. This was explained (22) by assuming two different sites for the protonated Schiff base within the protein. In one site, the Schiff base is near an amino acid with a pKa of 9.6-e.g., tyrosine. This gives rise to the slow M412 component. In the second site, the Schiff base is near the conjugate base of the amino acid and this gives rise to the fast M412 component. A model was then proposed to account for the comparable rate of deprotonation ofboth the tyrosine and the slow component of the protonated Schiff base (22). In this model, a positively charged species comes in close proximity to both acids and leads to their deprotonation due to a reduction in their pKa. The positively charged species can either be Ca2' or Mg2± or a positively charged amino acid-e.g., arginine. The fact that the removal of Ca2+ and Mg2+ ions inhibits the proton pumping mechanism but retains the initial intermediates of the photocycle (17) supports this simple-minded model and suggests that the positively charged species is Ca2' or Mg2'. If this is true, then the removal of these two ions should also prevent the deprotonation of tyrosine during the photocycle. For the purple membrane, two absorption changes are observed at 297 nm (19). The first change has a fast rise time and is assigned (19) to changes in the retinal absorption that result from its rapid trans-cis isomerization. The second one is assigned to a tyrosinate ion formation (19) supported by the fact that its amplitude decreases with pH and has a titration curve with a pKa value similar to that of tyrosine (29). In this paper, we study the transient absorption at this wavelength for the deionized and acidified membranes. The transient absorption assigned to the trans-cis isomerization of the retinal chromophore is observed in both the blue and purple bR photocycles, confirming the previous observations that the earlier intermediates in the photocycle are formed in the blue membranes (17). However, the slow rising absorption resulting from tyrosine deprotonation is found to be absent for the blue membranes but can be restored by the addition of Ca2' and Mg2+, suggesting that the deprotonation of tyrosine, like that of the protonated Schiff base, requires the presence of bound Ca2+ or Mg2+.

ABSTRACT The transient absorption changes occurring at 297 nm during the photocycles of the deionized and acidified bacteriorhodopsins (blue membranes) were studied. As opposed to what happens during the photocycle of the purple membrane, for the blue membranes only the fast absorption increase corresponding to trans-cis isomerization of the retinal chromophore is present; the slow rise attributed to the tyrosine deprotonation is not observed; The addition of different salts to the deionized membrane restores the original color and causes a tyrosine deprotonation during the photocycle. This suggests that the presence of cations is required for the deprotonation of tyrosine as it is for the deprotonation of the retinylidene Schiff base. These results are discussed in terms of the recently proposed cation model for the observed deprotonation processes in the photocycle of bacteriorhodopsin.

Bacteriorhodopsin (bR), the only protein in the purple membrane of Halobacterium halobium, contains one molecule of retinal covalently bound to the e-amino group of a lysine residue via a protonated Schiff base (1, 2). When it absorbs light, it undergoes a photochemical cycle (3) during which protons are pumped from the inside to the outside of the cell resulting in a pH gradient across the cell membrane. -This proton gradient drives metabolic processes such as ATP synthesis (4-9). The absorption spectrum of bR shows a maximum at 570 nm. However, when the membrane is deionized on a cationexchange column -(10) or treated with NaCl (10, 11), EDTA (11, 12), or acid (1, 13-16), its Xmax shifts to =600 nm. All of the experiments done so far indicate that the different blue membranes obtained by these methods are indistinguishable from the acidified membrane (11, 12), which does not pump protons (15). The photocycles of both the deionized and acidified bR do not have a blue shifted intermediate corresponding to the M412 transient in the photocycle of native bR (11, 17). Elemental analysis has also shown that while the purple membrane binds approximately 1 Ca2+ and 3 Mg2+ per bR molecule, there is almost no Ca2+ or Mg2+ present in the blue membrane preparations (12). In the purple membrane photocycle, a tyrosine deprotonation occurs on a time scale similar to that for the L550-tO-M412 transformation (18-23). The effect of chemically modified tyrosines on the rate of decay of M412 suggests a coupling between some tyrosines and M412 (24-27). Recently (22), the kinetics of formation of both the tyrosinate ion and M412 were studied simultaneously as a function of pH and temperature.

EXPERIMENTAL The purple membrane was purified from the ET1-001 strain of Halobacterium halobium as in ref. 22. Membrane suspensions were deionized by passage through a cation-exchange column as described (10) and the acidified membrane was obtained by adding HCO to a bR suspension until pH 2.0 was reached; the bR concentration was kept constant at 16 1AM. Salts from stock solutions of 4 mM CaCl2, 4 mM MgCl2, and 0.4 M NaCl in double-deionized water were added to 3 ml of

The results have shown two formation rates for M412, with the slower one dominating under physiological conditions and with a rate slightly faster than that for the tyrosine deprotonation. This eliminated the possibility that the tyrosinate formation is a prerequisite for the Schiff base deprotonation The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. §1734 solely to indicate this fact.

Abbreviation: bR, bacteriorhodopsin.

3662

Biophysics: Dupuis et al.

Proc. Natl. Acad. Sci. USA 82 (1985)

the deionized membrane suspension with a micropipet. All blue membrane suspensions were kept in plastic containers and contact with glass was avoided to prevent contamination of the solutions. The absorption spectrum of each sample was recorded on a Hewlett-Packard 8451A diode array spectrophotometer with a spectral resolution of 2 nm. The experimental setup used to monitor the transient UV absorption changes is the same as the one described in refs. 22 and 23. The monitoring wavelength was 297 nm and the photolysis of the samples was accomplished with a focused 6-ns, 590-nm beam obtained from a dye laser pumped by the frequency doubled output of a Nd:YAG laser (Quanta Ray, Mountain View, CA), at a repetition rate of 10 Hz. The dye used was kiton red and photolysis powers were kept at 50 mW for all samples.

RESULTS AND DISCUSSION For native bR, upon photolysis, a biphasic absorbance change is monitored at 297 nm (see Fig. 1). The faster component corresponds to trans-cis isomerization of the

cE o~~~~ X Cu

I Wce

Cu_ c

0 0 .0J

0

200

400

600

800

1000

Time, ,us FIG. 1. The transient absorption increase monitored at 297 nm during the photocycle of bR and its modified preparations. The rapid rise observed for all samples is due to the trans-cis isomerization of the retinal chromophore while the slow rising absorption present in curves 3-7 results from a tyrosine deprotonation. The different formation curves are for acidified bR (curve 1), deionized bR (curve 2), deionized bR + 13 AtM, 27 ,M, and 78 ,uM CaCl2 (curves 3, 4, and 5, respectively), deionized bR + 78 ,AM MgCI2 (curve 6), and native bR (curve 7). The bR concentration was kept constant at 16 ,uM and all suspensions were prepared in double-deionized water. The curves have been drawn so that the amplitude of the fast component is constant for each one. The arrow marks the time at which the photolysis laser pulse is fired.

3663

retinal chromophore, which can be observed at this wavelength (21), and the slower component is caused by the formation of a tyrosinate ion (19, 22, 23). As seen in Fig. 1, for both the deionized and acidified membranes, only the fast component is present. The absence of the slow transient absorption at 297 nm in the deionized and acidified samples could either suggest the inhibition of the tyrosine deprotonation or a large shift in the tyrosinate absorption. We have repeated these experiments at 313 nm and have not found any new absorption changes during the photocycle of the modified samples. Thus, the shift in the absorption spectrum of the tyrosinate ion, if it occurs, would have to be towards the protein absorption region. We have observed that the protein absorption does not significantly shift in these modified samples. This, together with the small cation concentrations (-0.1 mM) required to restore the tyrosinate signal at 297 nm, argues against large shifts in the absorption spectrum of the tyrosinate ion. This then suggests that the absence of absorption at 297 nm in the deionized and acidified samples is a result of the lack of tyrosinate formation and thus the result of the inhibition of the tyrosine deprotonation. There are two additional conclusions that can be made. First, the very similar behavior of the two blue membranes seems to corroborate the assumption that both species are identical (1012). Second, the presence of a fast absorbance change at 297 nm indicates that the retinal chromophore still undergoes a trans-cis isomerization during the photocycle of the blue membranes. There have been previous studies on the acidified (15, 17) and deionized (11) membranes that have shown that the equivalent of the K and L intermediates are present in their photocycles, with formation rates identical to those of native bR. These facts, coupled with our observation that the chromophore still isomerizes, indicate that the trans-cis isomerization is not sufficient for the proton pump mechanism to be active. Our observations also indicate that the formation of M412 is inhibited during the photocycle of bR when Ca2' and Mg2+ are not present in the membrane, in agreement with previous observations (11, 17). It was noted in refs. 10 and 12 that once bR was deionized, the addition of Ca2+, Mg2+, Na+, etc., restored the original purple membrane. We added different concentrations of these ions to the deionized membrane and, in each case, we have studied the transient absorbance at 297 nm. We wanted to verify that the Ca- and Mg-restored samples behave identically. We also used one Na+ concentration for the same purpose. It was observed (10) that when the purple color has been restored by addition of a monovalent cation, subsequent dilution regenerates the blue membrane. The results are given in Fig. 1 and in Table 1. Table 1 shows that the results obtained for the samples prepared with Ca2+ or with Mg2+ are indeed very similar. The rate constants obtained for the tyrosinate ion formation are comparable for both sets of solutions. One can also observe that as the visible absorption maximum of the restored bR shifts from 604 nm to 570 nm, the kinetic data become closer and closer to those for native bR. We also observe that it takes more Na+ than Ca2+ or Mg2+ to restore the purple color (10), but the Na-restored bR is found not to differ from Caand Mg-restored bR. As shown in Table 1, the observed rate constant for the tyrosinate ion formation changes slightly with the concentration of the salt used. A possible explanation for this is the following. A close observation of the curve for the completely deionized bR in Fig. 1 (curve 2) shows the presence of a slow decaying component following the fast rising one. Thus, at any salt concentration, the temporal behavior of the slow component is a combination of this slow decay and of the rising amplitude of the tyrosinate ion formed. As the tyrosinate amplitude increases with the salt concentration, its

3664

Biophysics: Dupuis et al.

Table 1. The salt effect on the rate constant and amplitude of the transient tyrosinate formation measured at 297 nm during the photocycle of the blue membranes k x 104 s-1 Sample Amplitudet Ak,* 570 0.94 ± 0.04 0.52 ± 0.02 Native bR 0 600 Acidified bR 0 604 Deionized bR + 13 AuM CaC12 0.44 ± 0.04 0.11 ± 0.01 594 + 27 ,uM CaC12 0.32 ± 0.03 578 0.92 ± 0.05 + 53 ,M CaCl2 0.42 ± 0.02 572 0.96 ± 0.03 + 78 AuM CaCl2 0.48 ± 0.02 1.10 ± 0.05 572 + 13 ,1M MgCl2 0.14 ± 0.02 0.48 ± 0.05 594 + 27 ,uM MgCl2 0.24 ± 0.01 0.75 ± 0.02 582 + 53 1LM MgCl2 0.44 ± 0.04 572 1.10 ± 0.05 + 78 AM MgCl2 1.10 ± 0.05 0.46 ± 0.04 572 + 2.7 mM NaCl 576 0.77 ± 0.06 0.19 ± 0.02 bR concentration = 16 AM. *Measured in a separate experiment with a Hewlett-Packard 8451A diode array spectrophotometer. tThese values correspond to the fraction of the absorption change at 297 nm attributed to the tyrosinate ion signal. The rest of the signal is assigned to trans-cis isomerization of the retinal chromophore.

formation rate constant also "appears" to increase until it reaches the value observed in native bR. The results shown in Table 1 indicate that whatever salt is used to regenerate the purple color, the tyrosinate ion formation rate observed during the photocycle is identical for each sample having a given visible absorption maximum or

relative tyrosinate amplitude. The above results thus show that just as the cations present in the membrane are needed to produce the M412 intermediate (11), they are also needed to produce the tyrosinate ion. This gives an additional support for the mutual coupling of the processes involved in their formation and might give some credibility to the recently proposed model for the deprotonation of the protonated Schiff base and tyrosine (22). This model proposes that a positively charged species is required for the reduction of the pKa of both acids and thus for their deprotonation at a comparable rate. It would be tempting to suggest that this positively charged species is a Ca2+ or a Mg2+ ion bound to the protein. The fact that either ion restores the tyrosine deprotonation might indicate that the site for these ions near the Schiff base and tyrosine can accommodate either Mg2+ or Ca2+. On the other hand, it is also possible that either ion is required for inducing protein conformation changes (10, 16, 30) that bring different positively charged species (e.g., arginine) near the tyrosine and the Schiff base sites during the photocycle. In any case, as was discussed previously (22), the pKa values for both the protonated Schiff base and tyrosine are reduced on a comparable time scale that could be determined by the time required for the transient protein conformation changes to bring together the responsible cation and the deprotonating acids during the photocycle. The attractiveness of this cation model stems from the fact that a cation can reduce the pKa of both acids. The pKa value of the positively charged protonated Schiff base could be reduced on account of the repulsive interaction with the responsible cation. The decrease in the pKa value of tyrosine could result from the attractive interaction between the responsible cation and the negatively charged tyrosinate ion produced upon deprotonation.

Proc. Natl. Acad. Sci. USA 82 (1985) In this model (22), we have proposod that only one cation deprotonation of both the protonated Schiff base and the tyrosine. Since both Ca2W and Mg2+ are found to induce the deprotonation of tyrosine, one should leave the door open for the possibility of having two different cations responsible for the deprotonation of the two acids. In this case, the similarity of the deprotonation rates of both acids could still be the result of a rate-determining protein-retinal conformation change during the photocycle. causes the near-simultaneous

We thank Dr. R. A. Bogomolni who generously gave us our first deionized sample. This work was supported by the Department of Energy, Office of Basic Sciences. P.D. thanks the Natural Sciences and Engineering Research Council of Canada for a postdoctoral fellowship. 1. Oesterhelt, D. & Stoeckenius, W. (1971) Nature (London) New Biol. 233, 149-152. 2. Bridgen, J. & Walter, I. D. (1976) Biochemistry 15, 792-798. 3. Lozier, R., Bogomolni, R. A. & Stoeckenius, W. (1975) Biophys. J. 15, 955-962. 4. Stoeckenius, W., Lozier, R. & Bogomolni, R. A. (1979) Biochim. Biophys. Acta 505, 215-278. 5. Henderson, R. (1977) Annu. Rev. Biophys. Bioeng. 6, 87-109. 6. Honig, B. (1978) Annu. Rev. Phys. Chem. 29, 31-57. 7. Dencher, N. & Wilms, M. (1975) Biophys. Struct. Mech. 1, 259-271. 8. Belliveau, J. W. & Lanyi, J. K. (1977) Arch. Biochem. Biophys. 178, 308-314. 9. Stoeckenius, W. & Bogomolni, R. A. (1982) Annu. Rev. Biochem. 51, 587-616. 10. Kimura, Y., Ikegami, A. & Stoeckenius, W. (1984) Photochem. Photobiol. 40, 641-646. 11. Chang, C.-H., Suh, C.-K., Govindjee, R. & Ebrey, T. (1984) Biophys. J. 45, 210a (abstr.). 12. Chen, J. G., Chang, C.-H., Govindjee, R. & Ebrey, T. (1984) Biophys. J. 45, 214a (abstr.). 13. Moore, T. A., Edgerton, M. E., Parr, G., Greenwood, C. & Perham, R. N. (1978) Biochem. J. 171, 469-476. 14. Fischer, V. & Oesterhelt, D. (1979) Biophys. J. 28, 211-230. 15. Mowery, P. C., Lozier, R. H., Chae, Q., Tseng, Y.-W., Taylor, M. & Stoeckenius, W. (1979) Biochemistry 18, 4100-4107. 16. Muccio, D. D. & Cassim, J. Y. (1979) J. Mol. Biol. 135, 595-609. 17. Kobayashi, T., Ohtani, H., Iwai, J.-i., Ikegami, A. & Uchiki, H. (1983) FEBS Lett. 162, 197-200. 18. Bogomolni, R. A., Stubbs, L. & Lanyi, J. (1978) Biochemistry 17, 1037-1041. 19. Hess, B. & Kuschmitz, D. (1979) FEBS Lett. 100, 334-340. 20. Bogomolni, R. A. (1980) in Bioelectrochemistry, ed. Keysev, H. (Plenum, New York), pp. 83-95. 21. Kuschmitz, D. & Hess, B. (1982) FEBS Lett. 138, 137-140. 22. Hanamoto, J. H., Dupuis, P. & El-Sayed, M. A. (1984) Proc. Nati. Acad. Sci. USA 81, 7083-7087. 23. Dupuis, P. & El-Sayed, M. A. (1985) Can. J. Chem. 63, in press. 24. Konishi, T. & Packer, L. (1978) FEBS Lett. 92, 1-4. 25. Scherrer, P., Packer, L. & Seltzer, S. (1981) Arch. Biochem. Biophys. 212, 589-601. 26. Lemke, H.-D. & Oesterhelt, D. (1981) Eur. J. Biochem. 115,

595-604. 27. Lam, E., Seltzer, S., Katsura, T. & Packer, L. (1983) Arch. Biochem. Biophys. 227, 321-328. 28. Rosenbach, V., Goldberg, R., Gilon, C. & Ottolenghi, M. (1982) Photochem. Photobiol. 36, 197-201. 29. Fukumoto, J. M., Hanamoto, J. H. & El-Sayed, M. A. (1983) Photochem. Photobiol. 39, 75-79. 30. Tsuji, K. & Rosenheck, K. (1979) FEBS Lett. 98, 368-372.