Evidence for Two Geminate Rebinding States Following Laser

May 21, 2005 - Evidence for Two Geminate Rebinding States Following Laser Photolysis of R State Hemoglobin Encapsulated in Wet Silica Gels ... The ana...
0 downloads 8 Views 90KB Size
11411

2005, 109, 11411-11413 Published on Web 05/21/2005

Evidence for Two Geminate Rebinding States Following Laser Photolysis of R State Hemoglobin Encapsulated in Wet Silica Gels Silvia Sottini, Stefania Abbruzzetti, and Cristiano Viappiani* Dipartimento di Fisica, UniVersita` degli Studi di Parma, Parco Area delle Scienze 7/A, 43100 Parma, Italy

Stefano Bettati Dipartimento di Sanita` Pubblica, UniVersita` degli Studi di Parma, Via Volturno 39, 43100 Parma, Italy

Luca Ronda and Andrea Mozzarelli Dipartimento di Biochimica e Biologia Molecolare, UniVersita` degli Studi di Parma, Parco Area delle Scienze 23/A, 43100 Parma, Italy ReceiVed: March 18, 2005; In Final Form: May 9, 2005

In this letter we report the first experimental evidence for CO rebinding to human hemoglobin from multiple geminate states. The analysis of the rebinding kinetics using a maximum entropy method allowed the identification of two distinct rebinding states within the protein matrix, which become populated under conditions of increased viscosity in a silica gel at high glycerol concentration. Our findings suggest the presence of at least two distinct docking sites for the photolyzed ligand. Assuming a minimal four-state model, we estimate the microscopic rates and the activation energies for the elementary processes.

Photolysis of CO complexes of hemoglobin (Hb) has been traditionally used for the characterization of the ligand binding kinetics. The fraction of photodissociated CO molecules which react with the heme and do not reach the solvent phase gives rise to the geminate recombination. This process has been treated as a first-order rate process, characterized by multiple exponential decays,1-4 reflecting structural relaxation of the protein.4,5 We report here the first kinetic evidence for multiple geminate rebinding states after photolysis of HbCO. Migration of the photodissociated ligand within the protein matrix of myoglobin (Mb) revealed the internal protein structure and its fluctuations on the time scale extending from picoseconds to microseconds. Ligand rebinding kinetics for different Mb mutants and under Xe pressures6,7 suggested that the ligand may migrate to a hydrophobic cavity, called Xe1 cavity, on the proximal side, possibly via the Xe4 cavity on the distal side. Low temperature8 and time-resolved9-11 X-ray crystallographic studies, have located the dissociated CO in these docking sites by monitoring ligand rebinding and migration and the associated structural changes within the protein. Molecular dynamics simulations mapping the ligand escape pathways also proposed the involvement of the Xe cavities.12 In the case of Hb, a detailed understanding of the internal pathways for CO migration is not yet available because of the limited structural data. It was observed that the photodissociated CO molecules are located only in the distal pocket at 25 K,13 but the low temperature likely prevented migration of the ligand, as previously observed for Mb.8 Therefore, kinetic data taken under conditions where exit to the solvent phase is strongly inhibited are extremely * Corresponding author. E-mail: [email protected]

10.1021/jp0514224 CCC: $30.25

Figure 1. Rebinding kinetics (left) and lifetime distributions (right) for wet R state HbCO gels at T ) 10 °C in the absence (red circles and red line) and presence of 80% glycerol (black circles and black line). Blue lines in the left panel are the fit with MEM. The bathing solution contains 0.1 M Hepes, pH 7.

relevant to gain understanding on the internal ligand dynamics. To this goal we have encapsulated HbCO in wet nanoporous silica gels and carried out nanosecond laser flash photolysis experiments as a function of glycerol concentration and temperature (Supporting Information). The constraints imposed by the gel matrix simplify the rebinding kinetics by preventing the quaternary relaxation.14 Furthermore, the increased geminate yield for CO rebinding to R state Hb inside silica gels in the presence of glycerol15 is expected to probe with higher sensitivity the geminate state. To retrieve unbiased kinetic information, and detect multiple geminate states, we have used a maximum entropy method (MEM)16,17 to obtain a distribution of lifetimes associated with the CO rebinding kinetics. The geminate rebinding kinetics to R state HbCO gels (Figure 1, left panel) becomes clearly biphasic at low temperature and high glycerol concentration. The additional constraints imposed on the protein by glycerol decrease the fraction of CO molecules escaping to © 2005 American Chemical Society

11412 J. Phys. Chem. B, Vol. 109, No. 23, 2005

Letters on the discovery of hydrophobic cavities in a number of different globins,20,21 suggest that such cavities might be present also in Hb. The dependence of the geminate lifetime distribution on glycerol concentration (Figure 2A) and temperature (Figure 2B) exhibits a progressive split in two bands. Figure 2 shows that the second, slower band can be easily identified only at glycerol concentrations above 60% and temperatures below 30 °C. It is therefore difficult to extract viscosity dependence for the rate constants. In the absence of structural data, we can only provide a kinetic interpretation of our results. We have followed a scheme proposed for Mb22 to describe the rebinding kinetics with a simplified branched four-state model (Supporting Information). In this model we assumed that, after photolysis, the ligand can rebind to the heme from fast rebinding sites, likely present within the distal pocket, or from slower rebinding sites, located farther in the protein matrix, and, possibly, on the proximal side. Escape of the ligand to solvent occurs mainly from the distal pocket, as observed for Mb.22,23 The differential equations corresponding to Scheme 1 can be solved analytically SCHEME 1: Simplified Four-State Kinetic Model

Figure 2. (A) Contour plots for the lifetime distribution of CO rebinding kinetics to R state HbCO gels as a function of glycerol concentration. T ) 15 °C. Two separate bands appear in the geminate phase at glycerol concentrations above ≈60%. (B) Contour plots for the lifetime distribution of CO rebinding kinetics to R state HbCO gels in the presence of 80% glycerol as a function of temperature.

the solvent. This fraction accounts for less than 10% of the photodissociated ligands at 10 °C and 80% glycerol. The recovered lifetime distribution, g(log(τ)), in the presence of glycerol (Figure 1, right panel) shows the appearance of two peaks (centered at log(τ) ) -7.7 and -6.4) for the geminate phase. The corresponding distribution in the absence of glycerol evidences a single, broad band at log(τ) ) -7.1. Bimolecular rebinding from solution phase occurs at much longer times, with peaks at log(τ) ) -3.7 and -2.7, in the absence and presence of glycerol, respectively. As expected, the bimolecular rebinding rate scales with viscosity (Supporting Information). The difference in the log(τ) values for the geminate rates in glycerolcontaining samples is likely too large (more than 1 order of magnitude) to be attributed to rebinding to R and β chains.3,18 The single peak in the bimolecular rebinding gives additional support to this argument, showing that the reaction of the ligand with the two different chains occurs with undistinguishable rates. Since the observed distribution becomes bimodal at glycerol concentrations above ≈40%, the splitting seems to be related to the internal protein dynamics rather than static heterogeneity. Bimodal lifetime distributions in the geminate phase were previously reported for MbCO solutions with 79% glycerol, and were interpreted as arising from binding from the Xe1 and Xe4 sites.7 It seems reasonable to interpret the two bands in the geminate phase distribution for HbCO as the first evidence of CO rebinding from different sites inside the protein, in analogy with the observed behavior in Mb. The increased difficulty encountered by CO in the escape to the solvent phase, due to the high glycerol concentration, allows the ligand to migrate also to docking sites which would not be populated otherwise. This interpretation is supported by recent data on HbO2 photolysis, showing evidence for two rebinding states inside Hb subunits,19 and by the strict analogy with MbCO kinetic data in the gel (unpublished results). Moreover, recent reports

and, under the hypothesis of preequilibrium for the rebinding from solution, microscopic rates can be obtained. From the linear Arrhenius plots in the temperature range 10-50 °C (Supporting Information) the activation energies for the microscopic rate constants were obtained (Table 1). The rates kCA and kCB are characterized by negligible activation barriers, regardless of glycerol content. On the contrary, the activation energy for rate kBC is 5.5 kcal/mol, a value similar to the barrier that CO must overcome to escape to solvent (rate kCS). TABLE 1: Activation Energy and Rate Constants at 15 °C in the Absence and Presence of 80% Glycerol CfA No glycerol k (106 s-1) Ea (kcal/mol) 80% glycerol k (106 s-1) Ea (kcal/mol)

5.0 0 14 0

BfC

CfB

CfS

SfC

-

-

5.5 6.3

11.1 M-1 15.2

3 5.5

17 0

5 5.2

1.7 M-1 10.8

Although the barrier for process CfA is unchanged by the addition of glycerol, the intrinsic probability for returning to state A and, mostly importantly, to state B are dramatically increased, resulting in much higher corresponding rates (Table 1). This explains the high efficiency with which state B is populated in Hb, even in the presence of a decreased barrier for process CfS. Acknowledgment. The authors thank MIUR (FIRB and COFIN2004), CNR (Functional Genomics project) and INFM for financial support. P.J. Steinbach is kindly acknowledged for the use of MemExp. Supporting Information Available: Experimental methods and data analysis. Equations relating observed with microscopic rate constants for the kinetic model in Scheme 1. Temperature and viscosity dependence of microscopic rate constants. This material is available free of charge via the Internet at http:// pubs.acs.org.

Letters References and Notes (1) Jones, C. M.; Ansari, A.; Henry, E. R.; Christoph, G. W.; Hofrichter, J.; Eaton, W. A. Biochemistry 1992, 31, 6692-6702. (2) Goldbeck, R. A.; Paquette, S. J.; Bjo¨rling, S. C.; Kliger, D. S. Biochemistry 1996, 35, 8628-8639. (3) Esquerra, R. M.; Goldbeck, R. A.; Reaney, S. H.; Batchelder, A. M.; Wen, Y.; Lewis, J. W.; Kliger, D. S. Biophys. J. 2000, 78, 32273239. (4) Henry, E. R.; Jones, C. M.; Hofrichter, J.; Eaton, W. A. Biochemistry 1997, 36, 6511-6528. (5) Lyons, K. B.; Friedman, J. M. In Hemoglobin and oxygen binding; Ho, C., Eaton, W. A., Collman, J. P., Gibson, Q. H., Leigh, J. S., Margolias, E., Moffat, K., Scheidt, W. R., Eds.; Elsevier/North-Holland: Amsterdam, 1982. (6) Scott, E. E.; Gibson, Q. H. Biochemistry 1997, 36, 11909-11917. (7) Tetreau, C.; Blouquit, Y.; Novikov, E.; Quiniou, E.; Lavalette, D. Biophys. J. 2004, 86, 435-447. (8) Ostermann, A.; Washipky, R.; Parak, F. G.; Nienhaus, G. U. Nature 2000, 404, 205-208. (9) Srajer, V.; Ren, Z.; Teng, T. Y.; Schmidt, M.; Ursby, T.; Bourgeois, D.; Pradervand, C.; Schildkamp, W.; Wulff, M.; Moffat, K. Biochemistry 2001, 40, 13802-13815. (10) Bourgeois, D.; Vallone, B.; Schotte, F.; Arcovito, A.; Miele, A. E.; Sciara, G.; Wulff, M.; Anfinrud, P.; Brunori, M. Proc. Natl. Acad. Sci. U.S.A. 2003, 100, 8704-8709.

J. Phys. Chem. B, Vol. 109, No. 23, 2005 11413 (11) Schotte, F.; Lim, M.; Jackson, T. A.; Smirnov, A. V.; Soman, J.; Olson, J. S.; Phillips, G. N., Jr.; Wulff, M.; Anfinrud, P. A. Science 2003, 300, 1944-1947. (12) Bossa, C.; Anselmi, M.; Roccatano, D.; Amadei, A.; Vallone, B.; Brunori, M.; Nola, A. D. Biophys. J. 2004, 3855-3862. (13) Adachi, S.; Park, S. Y.; Tame, J. R. H.; Shiro, Y.; Shibayama, N. Proc. Natl. Acad. Sci. U.S.A. 2003, 100, 7039-7044. (14) Shibayama, N.; Saigo, S. J. Mol. Biol. 1995, 251, 203-209. (15) Sottini, S.; Viappiani, C.; Ronda, L.; Bettati, S.; Mozzarelli, A. J. Phys. Chem. B 2004, 108, 8475-8484. (16) Steinbach, P. J.; Ionescu, R.; Matthews, C. R. Biophys. J. 2002, 82, 2244-2255. (17) Steinbach, P. J. J. Chem. Inf. Comput. Sci. 2002, 42, 1476-1478. (18) Unzai, S.; Eich, R.; Shibayama, N.; Olson, J. S.; Morimoto, H. J. Biol. Chem. 1998, 273, 23150-23159. (19) Lepeshkevich, S. V.; Karpiuk, J.; Sazanovich, I. V.; Dzhagarov, B. M. Biochemistry 2004, 43, 1675-1684. (20) deSanctis, D.; Dewilde, S.; Pesce, A.; Moens, L.; Ascenzi, P.; Hankeln, T.; Burmester, T.; Bolognesi, M. Bioch. Bioph. Res. Com. 2004, 316, 1217-1221. (21) Ascenzi, P.; Bocedi, A.; deSanctis, D.; Pesce, A.; Bolognesi, M.; Marden, M. C.; Dewilde, S.; Moens, L.; Hankeln, T.; Burmester, T. Biochem. Mol. Biol. Educ. 2004, 32, 305-313. (22) Brunori, M.; Gibson, Q. H. EMBO Rep. 2001, 2, 674-679. (23) Agmon, N. Biophys. J. 2004, 87, 1537-1543.