Acid–Base Catalysis and Crystal Structures of a Least Evolved Ancestral GFP-like Protein Undergoing Green-to-Red Photoconversion

Abstract

In green-to-red photoconvertible fluorescent proteins, a three-ring chromophore is generated by the light-activated incorporation of a histidine residue into the conjugated π-system. We have determined the pH-rate profile and high- and low-pH X-ray structures of a least evolved ancestor (LEA) protein constructed in the laboratory based on statistical sequence analysis. LEA incorporates the minimal number of substitutions necessary and sufficient for facile color conversion and exhibits a maximal photoconversion quantum yield of 0.0015 at pH 6.1. The rate measurements provide a bell-shaped curve, indicating that the reaction is controlled by the two apparent pKa values, 4.5 ± 0.2 and 7.5 ± 0.2, flanking the chromophore pKa of 6.3 ± 0.1. These data demonstrate that the photoconversion rate of LEA is not proportional to the A-form of the GFP-like chromophore, as previously reported for Kaede-type proteins. We propose that the observed proton dissociation constants arise from the internal quadrupolar charge network consisting of Glu222, His203, Glu148, and Arg69. Increased active site flexibility may facilitate twisting of the chromophore upon photoexcitation, thereby disrupting the charge network and activating the Glu222 carboxylate for the abstraction of a proton from a carbon acid. Subsequently, the proton may be delivered to the Phe64 carbonyl by a hydrogen-bonded network involving Gln42 or by means of His65 side chain rotations promoted by protein breathing motions. A structural comparison of LEA with the nonphotoconvertible LEA-Q42A variant supports a role for Gln42 either in catalysis or in the coplanar preorganization of the green chromophore with the His65 imidazole ring.

Crystal structure of the fluorescent protein fromDendronephthyasp. in both green and photoconverted red forms

“…Such deviation is presumably owing to the increased flexibility of the chromophore phenolic ring caused by weakening/disruption of the hydrogen bond to the residue at position 142. These data are in line with the previously proposed mechanism, in which the PC is triggered by excitedstate chromophore motions promoted by a decreased density of molecular packing in the immediate chromophore environment (Kim et al, 2013). However, such an increase in the chromophore mobility often decreases its quantum yield, drastically impacting the usability of the corresponding mutants (Mandal et al, 2004)…”

Photoconvertible Fluorescent Proteins and the Role of Dynamics in Protein Evolution

“…In subsequent work, we demonstrated that the LEA quantum yield of photoconversion is 1.5 × 10 −3 [ 25 ], a value that is similar to other well-characterized pcFPs. Other experiments demonstrated that among the 13 sites, the substitutions T69A and Y116N and the deletion of Tyr217 were particularly important in promoting red color [ 22 ].…”

Reconstructing Ancient Proteins to Understand the Causes of Structure and Function

“…Crystal structures of the ancestral protein with and without the red-shifting substitutions showed that these residues have virtually no effect on the overall geometry of the structure (46). But molecular dynamics simulations of these proteins suggested that the key substitutions allow a transient intermediate conformational step to be occupied that makes incorporation of the imidazole into the red chromophore possible; specifically, the chromophore must undergo a light-activated internal twist that organizes the surrounding functional groups properly relative to the histidine (45, 46). The derived residues at the 11 other key sites make the backbone around the chromophore flexible and enable this conformation to be accommodated (46).…”

“…The derived residues at the 11 other key sites make the backbone around the chromophore flexible and enable this conformation to be accommodated (46). In the ancestral state, the backbone is more rigid and would clash with the twisted intermediate chromophore, so the imidazole cannot be incorporated, even if the histidine alone were present (45, 46). …”