Influence of Enzyme Residues on Cofactor Spectra

To determine the impact of neighboring residues in the modulation of the transition energies in the cofactor, a perturbation theory analysis of the excitation was performed.

This includes 32 amino acids around the cofactor inZmPDC, corresponding roughly to a 5 Å shell (see Figure 4.6). A few points should be mentioned about the setup, because several ionization sites are present in the pocket. In Figure 4.6a, these residues are shown together with ThDP. Residue Glu473 was considered to be protonated. This decision was based on mutagenesis data (in particular for variant Glu473Gln) and structural analyses of ZmPDC.^[103;104] In a comparison of the CD spectra of the wild-type enzyme and its Glu473Gln mutant, no significant shift was observed in the excitation energies, indicating that the charge of the mutated residue has probably not been changed. Residue His414 was considered to be doubly protonated, while His114 was set to be monoprotonated at the Nδ atom. The only difference in the MM shell between the two tautomers was in the protonation of the Glu50 residue. Because the latter is a prime candidate for the transfer of the proton to the aminopyrimidine ring, this residue was initially considered to be protonated (uncharged) in the AP tautomer but ionized (negatively charged) in the IP calculations. As one will later observe, such an assumption can strongly influence the relative placement of the transition energies. The results are shown in Figure 4.6b.

As one can observe, there is a limited list of residues that strongly modulate the excitation energy for both tautomers. His414 lays directly above the pyrimidine ring and was considered hereby as being doubly protonated (positively charged). This assignment was based on its placement relative to Asp82, which also contributes to a shift in the transitions, but with a different sign. The positive charge in the histidine stabilizes the π-system of the pyrimidine, while this is counteracted by the negative charge in the aspartate. However, the magnitude of the effect of His414 is approximately twice that

Results and Discussion 63

of the shift resulting from Asp82. The Glu50 residue provides the third most prominent feature in the analysis. As previously mentioned, this residue was considered to be the donor of the proton to N1’ for IP formation, leaving it uncharged in the AP case and charged (unprotonated) in the IP tautomer. This choice of setup has a strong impact on

(a)

-0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8

!

/e V

AP-ThDP IP-ThDP

(b)

E473(A) H114(B)

G413(A)

E50(B) D82(A) H414(A)

ThDP

Figure 4.6.: (a) Active site ofZmPDC. Left: IP-ThDP in the complete amino acid shell. Right: Ball-and-stick representation of ThDP and selected enzyme residues. (b) Perturbative residue-by-residue analysis of the impact on the charge-transfer excitations of AP and IP-ThDP in ZmPDC.

the magnitude of the shifts. In the AP case, it has barely any effect on the transition.

In the IP tautomer, it strongly lowers the energy.

The information from this first set of calculations can be summarized as follows. His414 was tentatively taken to be doubly protonated, given the uncertainty in its assignment.

The results show that such a protonation would strongly shift the absorption energies to higher values. The surrounding residues do not completely compensate for the positive charge next to the pyrimidine ring, and the latter will necessarily increase the energy penalty for an electron transfer to the thiazolium ring. For the simulation of ThDP in the whole enzyme, His414 was singly protonated at Nϵ, as this shift was deemed unlikely.

The protonation state of Glu50 is critically determining the position of the cofactor absorption bands. The excitation energies taken from the simple model amount to 3.79 and 4.74 eV (327 and 262 nm, respectively) for the IP and AP tautomers, respectively.

Given that the character of excitation is similar between the two, it is logical that the second shell of residues should not change the relative placement of the two bands. This, however, disagrees with all experimental observations which place the AP transition slightly below that of IP. One possible explanation would be that an assignment of Glu50 as being uncharged is invalid. This is not at all unlikely, considering that the glutamate could simply serve as a proton relay but not necessarily change its protonation when shifting from IP to the AP form. The X-ray structural analysis of ZmPDC revealed that Glu50 directly interacts with two ordered water molecules, and these could be engaged in reversible proton transfers.^[103;104]Taking the same shift as that observed for the IP case, one would obtain an excitation energy of 3.68 eV (337 nm), much closer to the measured value of 3.65–3.87 eV (320–340 nm).^[23;104] Therefore, in the following calculations both protonation states of the Glu50 residue were considered in the AP case in order to understand its effect in more detail. In the IP tautomer this residue was kept to be deprotonated (charged). All other protonation states of neighboring residues maintained the same for both AP and IP, with the exception of His114. The latter is the main candidate for the role of proton acceptor in the tautomerization process. The reason for this assignment was deduced from the analysis of the molecular dynamics simulations, described in the next section.

Results and Discussion 65