Did α-Synuclein and Glucocerebrosidase Coevolve? Implications for Parkinson's Disease

Abstract

Mutations in the GBA1 gene are associated with increased risk of Parkinson's disease, and the protein produced by the gene, glucocerebrosidase, interacts with α-synuclein, the protein at the center of the disease etiology. One possibility is that the mutations disrupt a beneficial interaction between the proteins, and a beneficial interaction would imply that the proteins have coevolved. To explore this possibility, a correlated mutation analysis has been performed for all 72 vertebrate species where complete sequences of α-synuclein and glucocerebrosidase are known. The most highly correlated pair of residue variations is α-synuclein A53T and glucocerebrosidase G115E. Intriguingly, the A53T mutation is a Parkinson's disease risk factor in humans, suggesting the pathology associated with this mutation and interaction with glucocerebrosidase might be connected. Correlations with β-synuclein are also evaluated. To assess the impact of lowered species number on accuracy, intra and inter-chain correlations are also calculated for hemoglobin, using mutual information Z-value and direct coupling analyses.

Fig 1. Synuclein evolutionary tree and sequence features.

A) The types of synuclein proteins found in different branches of vertebrates are shown. Some ray-finned fish have a fourth synuclein that resembles α-syn in the NAC region. B) The sequence features of α-syn, β-syn and γ-syn are diagrammed. The boxes represent the imperfect amphipathic repeats, and the dashed line for β-syn indicates the gap in the NAC region, which is indicated by the line above α-syn. C) Table of sequence identities (%) comparing synucleins in tetrapods (human and coelacanth), ray-finned fish (medaka), cartilaginous fish (ghostshark) and jawless fish (lamprey).

Fig 2. DCA and MI Z-value correlated mutation analyses of hemoglobin.

A) Intra-chain and B) inter-chain DI values (y-axis) for the 314 species hemoglobin MSA, ranked largest to smallest on the x-axis, between all non-invariant residues (light red dots), with those corresponding to contacts shown as orange circles. The inset shows an expanded view of the top 100 (intra-chain) or top 200 (inter-chain) results. The percentages of correctly predicted contacts for the top 10, 100, and 1000 results are shown. Because of the large number of contacts, histograms are shown below, with the number of contacts binned every 500 DI values (intra-chain) or 1000 DI values (inter-chain). C) Intra-chain and D) inter-chain MI Z-values for the 314 species hemoglobin MSA, shown similarly to A) and B). E) Intra-chain and F) inter-chain DI values for the 72 species hemoglobin MSA, corresponding most closely with species for which complete α-syn and GCase sequences are known, shown similarly to A) and B). G) Intra-chain and H) inter-chain MI Z-values for the 72 species hemoglobin MSA, shown similarly to A) and B). Note that for the 72 species hemoglobin MSA, there are fewer non-invariant residues, and thus fewer potential intra- and inter-chain pairs along the x-axis.

Fig 3. Top ten MI Z-value ranked α-syn, GCase correlated residue pairs

The pairs of α-syn and GCase residues for 72 vertebrate species with the most highly correlated mutations, as determined by the Z-value analysis, are displayed, with the Z-values and MI values shown. Also shown is the #15 ranked correlated pair, since it, along with the #1 and #8 ranked pairs, has an α-syn residue corresponding to a PD-associated mutation. The species are organized on the class, sub-class, or infra-class levels to highlight phylogenetic aspects of the residue variations. The α-syn residues for the cartilaginous fish, C. milii, are also shown, though since its complete GCase sequence is not known, C. milii was not included in the analysis.

Fig 4. GCase residues for the top ranked correlated pairs and conserved surface regions

A) A ribbon structure of human GCase (pdb 1OGS) with the residues from the top correlated pairs shown. B) The surface corresponding to invariant GCase residues is shown in green, with the surfaces of selected correlated pair residues in red and labeled. A large patch of invariant surface lies between V78 and G115.

Fig 5. Sequence and structure of α-syn showing top correlated residues and regions known to interact with GCase

A) The α-syn residues of the top ranked correlated pairs with GCase are indicated. Invariant residues are capitalized, negatively charged residues in red, positively charged in blue, hydrophobic in green, all others gray. The seven imperfect repeat regions are indicated by the dashed boxes and the NAC region is underlined (purple). In solution only residues 118–137 (dashed blue line) interact with GCase. In the presence of lipid vesicles, fluorescent labels at residues 57, 100, and 136 (green asterisks) showed interaction with GCase, indicating a much larger region of interaction than in solution. B) The α-syn structure with residues of the top ranked correlated pairs indicated. The residue coloring is the same as for the α-syn sequence. The NAC region is indicated by the purple oval, and the C-terminal region that interacts with GCase in solution is indicated by the dashed blue oval. The green asterisks indicate the locations at residues 57, 100, and 136 where fluorescent labels showed interaction with GCase in the presence of lipid vesicles. The structure shown is micelle bound α-syn (pdb 1XQ8).

Fig 6. Top MI Z-value ranked α-syn, β-syn and β-syn, GCase correlated residue pairs.

Residue pairs of β-syn with α-syn and GCase residues for 55 vertebrate species with the most highly correlated mutations, as determined by the Z-value analysis, are displayed, with the Z-values and MI values shown. For β-syn, sequences for E. caballus and C. ferus were not found, so the closely related E. ferus przewalskii and C. bactrianus were used.