Figure 1.
Enzymes and metabolites involved in the GABA shunt.
Enzymes catalysing reactions are numbered, (1) glutamate decarboxylase (GAD), (2) γ-aminobutyric transaminase (GABA-T), (3) succinic semialdehyde dehydrogenase (SSADH) and (4) succinic semialdehyde reductase (SSR). The blue highlighted boxes show enzymes that are found in the gab operon of E. coli. The thick black line indicates the pathway is blocked, SSA is then converted to γ-hydroxybutyrate (GHB) by succinic semialdehyde reductase (SSR), pathway coloured in red, which is described in SSDAH deficiency. Figure adapted from Blasi et al. [9]
Figure 2.
Crystal structure of E. coli SSADH.
a) Two cartoon representations of E. coli SSADH monomer (rotated by 180°) with NADP+ bound (orange) comprises of the catalytic domain (blue and light blue) with catalytic loop (red), the cofactor binding domain (green and yellow, where yellow illustrates the Rossmann fold) and the oligomerisation domain (magenta); b) A cartoon representation of the E. coli SSADH dimer with NADP+ bound (orange), it can be seen that the 3-stranded oligomerisation domain β- sheet (dark green) of the green monomer is extending the 7-stranded catalytic domain β- sheet (dark blue) of the blue monomer to form a 10-stranded β- sheet. c) Two surface representation models of the SSADH tetramer (rotated by 180°) showing the dimer of dimer formation between the blue and red monomer and the green and light blue monomers. NADP+ (orange) can be seen on the same face of the dimer, the substrate binding pocket has also been labelled.
Figure 3.
A single molecule superposition of E. coli SSADH and human SSADH.
A) Cα trace of monomer A of E. coli SSADH (green) superposed with the human SSADH molecule (PDB ID: 2w8r [26]: yellow: r.m.s.d. = 0.712 over 473 residues), with the NADP+ moiety (orange) from E. coli SSADH. Two structurally variable regions have been highlighted with dashed lines and labelled B–C. Figures B–C show Cα traces of all four E. coli SSADH monomers A–D (green) and 5 human SSADH monomers (open loop, PDB ID: 2w8o, 2w8p, 2w8q, 2w8r yellow; closed loop, PDB ID: 2w8n magenta)[26] superposed onto each other, only one NADP+ molecule (orange) from monomer A of E. coli SSADH is shown. B) Shows the region surrounding the 3 amino acid insertion (261RKN263) in human SSADH, which clashes with the 2'phosphate of NADP+. C) The loop motif connecting s2D and s3D in the catalytic domain, residues A379–G388 in E. coli SSADH and M432–G441 in human SSADH (r.m.s.d. of 3.4 Å over 10 residues). The E. coli SSADH loop is conserved throughout the ALDH family and is stabilised by 7 hydrogen bonds. The novel loop in human SSADH is stabilised by only 3 hydrogen bonds, furthermore this same loop in the reduced wild type human SSADH (PDB ID: 2w8o)[26] is highly flexible and could not be determined using X-ray crystallography.
Figure 4.
Comparisons between human (PDB ID: 2w8r)[26] and E.coli (monomer A) SSADH cofactor binding and SSA binding pockets, visualised using electrostatic surface representations (red represents negatively charged surfaces and blue represents positively charged surfaces).
A–B) both human (A: NAD+ in yellow) and E. coli (B: NADP+ in orange) SSADH have a two pocket NAD(P)+ binding site per molecule, the first (mostly blue), positively charged and close to the surface, accommodates the adenosine moiety (and the 2'phosphate in E. coli SSADH). The second binding pocket, deep in the active site, houses the nicotinamide ribose moiety (absent in human SSADH). The smaller human SSADH cofactor binding pocket has a large positive protrusion, which closes the bottom of the pocket, while the larger E. coli SSADH cofactor binding pocket can clearly accommodate the 2'phosphate of the NADP+. C–D) shows the positively charged SSA binding pocket of both human (C) and E. coli (D) SSADH highlighted by a white dashed line. The human SSA binding pocket is larger than the E. coli SSA binding pocket.
Figure 5.
SSADH substrate binding and the active site.
A cartoon representation of the E. coli SSADH (monomer A: green) substrate (SSA) binding pocket superposed onto human SSADH C340A mutant with SSA bound (PDB ID: 2w8q [26]: red) and human SSADH containing the catalytic cysteine (PDB ID: 2w8o [26]: yellow). The key SSA binding residues from 2w8q have their interactions with SSA shown as a red dashed line. Superposition of catalytic residues are shown: the catalytic cysteine and the general base as sticks (labeled according the E. coli SSADH) and NADP+ (orange) from E. coli SSADH. It can be seen that the equivalent SSA binding residues from 2w8o and E. coli SSADH (R164, R283 and S445) are in a very similar location and orientation to those of 2w8q. The catalytic cysteine of 2w8o is oriented toward the NADP+ moiety while in E. coli C288 is oriented toward the substrate (SSA). Also two conformations of the general base (E254) can be seen in E. coli SSADH, with (a) being in the hydride conformation and (b) the hydrolysis conformation.
Figure 6.
NADP+ binding of E. coli SSADH.
Stereo view of the active site showing the NADP+ moiety (yellow), SSADH residues (green) involved in binding NADP+, water molecules can be seen as red spheres and all bonds are depicted with a black dashed line. The 2F0–Fc omit electron density of the NADP+ moiety contoured at 1σ is also shown (light blue mesh). Interactions of monomer A and NADP+ can be seen, specifically both AN1 and AN6 of the adenine moiety (labelled adn) interacts with Q239(Oε1), Q243(Oε1) and N217(Oδ1) via water molecules. Adjacent to the adenine moiety, both AO2 and AO3 of the ribose (labelled rb2) hydrogen bonds with T153(O) and K179(NZ, O) via a single water molecule. 3AOP of the 2'-phosphate interacts with K179(NZ). AO2 of the pyrophosphate interacts with S233(N, OG) and the NO1 hydrogen bonds directly with W155(Nε1). Both NO2 and NO3 of the adjacent ribose moiety (labelled rb1) hydrogen bonds with K338(NZ), while NO2 also interacts with E385(Oε1). NN7 of the nicotinamide (labelled nt) moiety interacts with N156(Nδ2) and the catalytic C288(N) via the single water molecule. While NO7 interacts directly with G232(O) and L255(O), as well as with L255(N) and E254(Oε1) via the same water molecule. Up to 13 SSADH residues make 24 van der Waals or hydrogen bonds interactions with NADP+ per monomer, 16 of which are mediated by water (Table S2). Notably, all of the residues involved directly NADP+ binding in E. coli SSADH [48] (Table S2) are also conserved in human SSADH.
Figure 7.
Different binding states of NAD(P)+ in the ALDH family.
Four different conformations of NAD(P)+ are shown as sticks, hydride conformation (PDB ID: 1bpw[31]: yellow), hydrolysis conformation (E. coli SSADH: green), out conformation (PDB ID: 2ilu [29]: magenta) and flexible, where the nicotinamide ribose moiety is unable to be resolved using X-ray crystallography (PDB ID: 2w8r [26]: blue). The general base (E254) and the catalytic cysteine (C288: both orange), which are conserved in human and E. coli SSADH and the whole ALDH family, have been labelled to define the active site.
Figure 8.
Human point mutations mapped onto the E. coli SSADH structure.
A cartoon representation of E. coli SSADH monomer A (cyan, NADP+ orange) showing the 17 point mutations (magenta spheres) that map to the mature human protein. A small region of monomer B catalytic domain (dark grey) and monomer C oligomerisation domain (light grey) have been included to illustrate the proximity of point mutations with regard to dimer and tetramer interfaces. The mutations occur in all three domains, 4 in the oligomerisation domain (G176R, H180Y, P182L and G533R), 6 in the cofactor binding domain (C93F, C223Y, T233M, A237S, N255S and G268E) and 7 in the catalytic domain (N335K, N372S, P382L/Q, V406I, G409D and V487E).
Table 1.
Mutations causing nonsense mutations of SSADH from SSADH deficient patients.
Table 2.
Analysis of point mutations found in human SSADH enzymes as a result of missense mutations.
Table 3.
Data collection and refinement statistics*.