Exploration of the cytochrome c oxidase pathway puzzle and examination of the origin of elusive mutational effects

Abstract

Gaining detailed understanding of the energetics of the proton-pumping process in cytochrome c oxidase (CcO) is a problem of great current interest. Despite promising mechanistic proposals, so far, a physically consistent model that would reproduce all the relevant barriers needed to create a working pump has not been presented. In addition, there are major problems in elucidating the origin of key mutational effects and in understanding the nature of the apparent pK(a) values associated with the pH dependencies of specific proton transfer (PT) reactions in CcO. This work takes a key step in resolving the above problems, by considering mutations, such as the Asn139Asp replacement, that blocks proton pumping without affecting PT to the catalytic site. We first introduce a formulation that makes it possible to relate the apparent pK(a) of Glu286 to different conformational states of this residue. We then use the new formulation along with the calculated pK(a) values of Glu286 at these different conformations to reproduce the experimentally observed apparent pK(a) of the residue. Next, we take the X-ray structures of the native and Asn139Asp mutant of the Paracoccus denitrificans CcO (N131D in this system) and reproduce for the first time the change in the primary PT pathways (and other key features) based on simulations that start with the observed structural changes. We also consider the competition between proton transport to the catalytic site and the pump site, as a function of the bulk pH, as well as the H/D isotope effect, and use this information to explore the relative height of the two barriers. The paper emphasizes the crucial role of energy-based considerations that include the PT process, and the delicate control of PT in CcO.

Figure 1

The key proton pathways in CcO at the end of the D pathway. The notations used in this figure are taken from ref.. Here B⁻ designates the Fe-bound OH⁻ and the W_i (i=1,2,3,4) designate the water molecules that may be involved in the PT process. In the reaction step that is investigated here, there is PT through the Dpathway, via Glu286 to B⁻ and to a transient acceptor site for pumped protons (the hemea₃ D-propionate, Prd), respectively. Proton pumping is impaired in the N139D mutant CcO.

Figure 2

(a) A schematic depiction of the reaction steps in CcO. The numbers indicate the order of reactions. Due to space limitation, we do not draw each step separately and thus the charges correspond to the first step in each figure. Here, we designate a reduced and oxidized heme (squares) by a “+” and no charge, respectively. We do not exclude a possibility that the proton moves from Prd to another site after step 3. The labels are explained in the legend to Fig. 1. a and a3 are hemes a and a₃, respectively.

Figure 3

A schematic diagram for the states that are involved in the determination pK_app in mutants where the transfer to Prd is blocked. Here, PT designates proton transfer to B⁻. E₁H denotes the protonated E286 in the downward configuration and E₂H denotes the upward configuration (see Fig. 7 for example structure). pK_a1 and pK_a2 are the corresponding pK_a values in the two different configurations. K₁ and K₂ denote the equilibrium constants of rotation of the protonated and deprotonated E286, respectively.

Figure 4

Regions of allowed pK_a1 and pK_a2 for feasible ranges of K₁ and experimentally known pK_app, based on Eq. 6.

Figure 5

The energetics of the states depicted in Fig. 3 based on a representative set of computed values of K₁, K₂ and pK_app as presented in Table 4. The computed values are for the native CcO: K₁=0.2, K₂=3.0×10⁻⁵, pK_a1 = 9.4, pK_a2 = 13.2, pK_app=9.5 and for the N139D mutant CcO:K₁=1.0, K₂=2.0×10⁻⁵, pK_a1=11.5, pK_a2=16.2, pK_app = 11.8. Here, the solid line corresponds to the native CcO with pK_app=9.5 and the dashed line corresponds to the N139D mutant CcO with pK_app=11.8. Note that the free energy of theE⁻ species here does not include the free energy of the protons in the bulk solution at the given pH.

Figure 6

A tentative description of the energetics of the first few reaction steps in the native and the N139D CcOs. A more quantitative description is given in Fig. 9 (where the state notations are given in a different and more general way). Here E, P, B⁻ denote the E286, Prd, binuclear center, respectively. A “-” sign and “H” indicates a deprotonated or protonated site, respectively. a₃⁺ and a₃ are the oxidized and reduced hemea₃, respectively.

Figure 7

The two configurations of E286 as found in the crystal structure of the N139D mutant (PDB code: 3EHB) in P. denitrificans CcO. They are named as follows. Configuration E₁: downward E286 found in the crystal structures of both native and mutant enzymes; Configuration E₂: upward E286 found only in the crystal structure of the mutant enzyme.

Figure 8

The rotated E286 configurations generated by us. Configuration E₃: within 6Å from B⁻; Configuration E₄: within 6Å from Prd.

Figure 9

A schematic diagram for the energetics of the early competing steps in CcO (based on ref. and subsequent estimates) this diagram does not provide definitive conclusive results but rather it is given to clarify the nature of the competing paths. The paths shown allow us to consider the fraction of protons that first go to Prd and then to B⁻. The kinetic diagram for the key pathways is presented in Fig. 10. The figure depicts in boxes the given configurations where all relevant elements (heme a, heme a₃, Prd, E286, B⁻, Cu+2 etc.) are clarified in the first panel. The estimated barriers (in kcal/mol) are written on the arrows that lead between the different states with […]_N and […]_M indicating the values for native and N139D mutant CcOs, respectively, and the estimated free energy of each state is given in a small square at the left-bottom corner of the large square that represents the given state.

Figure 10

The subset of kinetic pathways shown in Fig. 9 that has been used for our simulated kinetics study. Here all species names correspond to the different states as described in Fig. 9. The species I can transfer a proton to a site in the D pathway, which is in equilibrium with the bulk solution. These steps are not included in Fig. 9, but are considered in our kinetic scheme. The relevant rate constants are discussed in the text. The species XV’, XIV’ and X have been clubbed together to signify the states where PT occurs to B⁻. The rate constant values marked in red have been used for the N139D mutant.

Figure 11

a) The barriers for PT from E286 to B⁻ in the native (solid line) and mutant (dashed line) enzymes. b) The barriers for PT from E286 to Prd in the native (solid line) and mutant (dashed line) enzymes. Two alternatives are shown in each panel; with one (red) or two (black) water molecules between the proton donor and acceptor. The transfer from 2^nd water to Prd or B⁻ goes downhill in each case. Since we are interested only in the highest energy point (rate determining step) along the PT path, the last steps are not shown.

Figure 12

Simulated kinetics for the pathways and rate constants given in Fig. 10. The simulations show the concentration of relevant species as a function of time for (a) native: k₁=2×10⁴, k₋₁=10⁵, k₂=10³, k₋₂=10², k₃=2×10⁴, k₋₃=10⁵, k₄=10³ and k₋₄=10 and (b, c) two cases with the N139D mutant: k₁=10⁵, k₋₁=10⁵, k₂=10⁴, k₋₂=10, k₃=10⁴, k₋₃=10⁶, k₄=10³ and k₋₄=10 or k₁=2×10⁴, k₋₁=10⁵, k₂=5×10⁴, k₋₂=10², k₃=2×10⁴, k₋₃=10⁵, k₄=5×10² and k₋₄=10. Note that in the 3^rd case the values of k₁ and k₋₁ are the same as in the native CcO.

Figure 13

Simulated kinetics for the rate of proton transfer to B⁻ as a function of pH for both the native and N139D mutant CcOs. Here the rate has been computed as [BH](t)/t = ([XV ′](t) + [XIV′](t) + [X](t))/t because XV′, XIV′ and X represent the species where PT occurs to B⁻. Since the computed rate is a function of time, we plot the rate obtained at t=10⁻⁴ that roughly corresponds to 100 μs. Here the reaction rate profiles have been shown for both native (black) and mutant (red) with both pathways to B⁻ and Prd are open with rate constants of Fig. 10. The obtained pH profiles emphasize that pK_app can change based on the relative contribution of the pathways.

Figure 14

The P_R → F rate as a function of pH/pD for the wild type and N139D mutant CcO in H₂O and in D₂O. The N139D data in H₂O are from ref.. The wild-type data in D₂O are from ref.. Experimental conditions were as in ref..