Structural characterization of a β-hydroxyacid dehydrogenase from Geobacter sulfurreducens and Geobacter metallireducens with succinic semialdehyde reductase activity.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54

BIOCHI4453_proof ■ 29 May 2014 ■ 1/9

Biochimie xxx (2014) 1e9

Contents lists available at ScienceDirect

Biochimie journal homepage: www.elsevier.com/locate/biochi

Research paper

Structural characterization of a b-hydroxyacid dehydrogenase from Geobacter sulfurreducens and Geobacter metallireducens with succinic semialdehyde reductase activity Q8

Yanfeng Zhang a, b, Yi Zheng b, Ling Qin b, Shihua Wang a, *, Garry W. Buchko c, *, R. Michael Garavito b, * a b c

College of Life Sciences, Fujian Agriculture and Forestry University, Fuzhou, Fujian, China Department of Biochemistry and Molecular Biology, Michigan State University, East Lansing, MI 48824-1319, USA Biological Sciences Division, Department of Biochemistry and Structural Biology, Pacific Northwest National Laboratory, Richland, WA 99352, USA

a r t i c l e i n f o

a b s t r a c t

Article history: Received 24 March 2014 Accepted 9 May 2014 Available online xxx

Beta-hydroxyacid dehydrogenase (b-HAD) genes have been identified in all sequenced genomes of eukaryotes and prokaryotes. Their gene products catalyze the NADþ- or NADPþ-dependent oxidation of various b-hydroxy acid substrates into their corresponding semialdehyde. In many fungal and bacterial genomes, multiple b-HAD genes are observed leading to the hypothesis that these gene products may have unique, uncharacterized metabolic roles specific to their species. The genomes of Geobacter sulfurreducens and Geobacter metallireducens each contain two potential b-HAD genes. The protein sequences of one pair of these genes, Gs-bHAD (Q74DE4) and Gm-bHAD (Q39R98), have 65% sequence identity and 77% sequence similarity with each other. Both proteins are observed to reduce succinic semialdehyde, a 4-carbon substrate instead of the typical b-HAD 3-carbon substrate, to g-hydroxybutyric acid. To further explore the structural and functional characteristics of these two b-HADs with a less frequently observed substrate specificity, crystal structures for Gs-bHAD and Gm-bHAD in complex with NADPþ were determined to a resolution of 1.89 Å and 2.07 Å, respectively. The structures of both proteins are similar, composed of 14 a-helices and nine b-strands organized into two domains. Domain 1 (1e165) adopts a typical Rossmann fold composed of two a/b units: a six-strand parallel b-sheet surrounded by six a-helices (a1ea6) followed by a mixed three-strand b-sheet surrounded by two a-helices (a7 and a8). Domain 2 (166e287) is composed of a bundle of seven a-helices (a9ea14). Four functional regions conserved in all b-HADs are spatially located near each other, with a buried molecule of NADPþ, at the interdomain cleft. Comparison of these Geobacter structures to a closely related b-HAD from Arabidopsis thaliana in the apo-NADPþ and apo-substrate bound state suggests that NADPþ binding effects a rigid body rotation between Domain 1 and 2. Bound near the Substrate-Binding and Catalysis Regions in two of the eight protomers in the asymmetric unit of Gm-bHAD is a glycerol molecule that may mimic features of bound biological substrates. © 2014 Published by Elsevier Masson SAS.

Keywords: Succinic semialdehyde reductase Structural biology GABA shunt Bio-inspired catalysis Protein X-ray crystallography

1. Introduction Beta-hydroxyacid dehydrogenases (b-HAD) are a superfamily of ubiquitous enzymes found in both prokaryotic and eukaryotic

Q1

Abbreviations: b-HAD, b-hydroxyacid dehydrogenases; Gs-bHAD, Geobacter sulfurreducens; Gm-bHAD, Geobacter metallireducens; SSA, succinic semialdehyde; GHB, g-hydroxybutyric acid; GABA, g-aminobutyric acid. * Corresponding authors. E-mail addresses: [email protected] (S. Wang), [email protected] (G.W. Buchko), [email protected] (R.M. Garavito).

organisms that catalyze the NADþ- or NADPþ-dependent oxidation of b-hydroxy acid substrates into their corresponding semialdehyde [1,2]. Characterized enzymes in this superfamily include 6phosphogluconate dehydrogenase [3], 3-hydroxyisobutyrate dehydrogenase [4], tartronate-semialdehyde reductase [5], and Lserine dehydrogenase [6], enzymes that play important roles in their respective metabolic pathways. A common feature of this list of b-HADs, with the exception of 6-phosphogluconate dehydrogenase, is a preference for a 3-carbon substrate core. However, these proteins are somewhat promiscuous and may also catalysis oxidation of substrates with a 2- and 4-carbon substrate core as

http://dx.doi.org/10.1016/j.biochi.2014.05.002 0300-9084/© 2014 Published by Elsevier Masson SAS.

Please cite this article in press as: Y. Zhang, et al., Structural characterization of a b-hydroxyacid dehydrogenase from Geobacter sulfurreducens and Geobacter metallireducens with succinic semialdehyde reductase activity, Biochimie (2014), http://dx.doi.org/10.1016/j.biochi.2014.05.002

55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

BIOCHI4453_proof ■ 29 May 2014 ■ 2/9

2

Y. Zhang et al. / Biochimie xxx (2014) 1e9

Fig. 1. A) Consensus sequence of the four conserved regions generally observed in the primary amino acid sequence of all b-hydroxyacid dehydrogenases [1]. Residues labeled with an X are usually hydrophobic. B) ClustalW2 multiple amino acid sequence alignment of Gs-bHAD (Q74DE4), Gm-bHAD (Q39R98) and characterized b-hydroxyacid dehydrogenases from Pseudomonas aeruginosa (Pa-bHAD, Q91516), Arabidopsis thaliana (At-bHAD, Q9LSV0), and Homo sapiens (Hs-bHAD, P31937) illustrated with ESPript [45]. Red shaded regions highlight invariant residues and red residues highlight conserved residues in the alignment. The consensus sequences of the four conserved regions observed in most b-HADs are shown below the alignment: Dinucleotide Cofactor-Binding Region (blue, * ¼ D or R), Substrate-Binding Region (cyan), Catalysis Region (green, # ¼ N or Q), and Cofactor-Binding Region #2 (gray).


66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

BIOCHI4453_proof ■ 29 May 2014 ■ 3/9


Q2

well. Indeed, b-HADs from Arabidopsis thaliana and Haemophilus influenzae have been identified that preferentially oxidize glyoxylate (2-carbon) [7] and D-b-hydroxybutyrate (4-carbon) [8], respectively. Because of this substrate diversity and the observation that multiple homologs of b-HADs are often present in the genome sequences of fungi and bacterial species (often with low overall sequence similarity), it has been suggested that some b-HAD superfamily members may have unique metabolic roles specific to their species [6,8]. The b-HADs are typically around 300 amino acids in length and are identified in sequenced genomic DNA by the presence of four conserved regions [1] as shown in Fig. 1A. The most conserved region is the N-terminal Dinucleotide Cofactor-Binding Region composed of the consensus sequence -GXXGXGXMGXXXAX NXXXGXXXXXXD/R- (where X is usually a hydrophobic residue). In the majority of family members, when the C-terminal amino acid of the Dinucleotide Cofactor-Binding Region is an aspartic acid residue the protein is specific for NADþ [9]. The next conserved region is perhaps the best signature of a b-HAD, the Substrate-Binding Region composed of the consensus sequence -DAPVSGGXXXAXXG-. Within this consensus sequence the -SGG- triad is especially conserved with the side chain of the serine residue observed to make contact with the carbonyl group of the substrate in the 6phosphogluconate dehydrogenase co-crystal structure with 6phosphogluconate [3]. It remains to be determined if a similar contact between this serine residue and the carbonyl group of a substrate is a feature common to all b-HADs. The third highly conserved region is the Catalysis Region composed of the consensus sequence -GXXGXGXXXKXXXN/Q-. Mutagenesis studies [9] and co-crystal structures [3] suggest the lysine residue is crucial for catalysis. The fourth region is more conserved in prokaryotes than eukaryotes with a consensus sequence of -KDLGXAXD-. While it is spatially near the Dinucleotide Cofactor-Binding Region in crystal structures and has been postulated to play a role in cofactor binding [1], the functional role of this conserved region is still unclear. The genomes of Geobacter sulfurreducens and Geobacter metallireducens each contain an open reading frame for two proteins containing the four regions described in Fig. 1A. To further explore the structural and functional characteristics of b-HADs, we cloned, expressed, purified, and crystallized one of the two bHADs proteins from G. sulfurreducens (Gs-bHAD) and G. metallireducens (Gm-bHAD), Q74DE4 and Q39R98 (UniProtKB), respectively [10]. Both proteins contain 287 residues and have 65% sequence identity and 77% sequence similarity with each other and reduce the 4-carbon substrate succinic semialdehyde (SSA) in the presence of nicotinamide adenine dinucleotide phosphate (NADPþ) [10]. Here we report the crystal structures of Gs-bHAD and Gm-bHAD in complex with NADPþ determined to a resolution of 1.89 Å and 2.07 Å, respectively, and describe their structural features in relation to the four conserved consensus sequences (Fig. 1A) characteristic of b-hydroxyacid dehydrogenases. 2. Materials and methods 2.1. Structure determination and refinement Details of the methods used to prepare clones of the Gs-bHAD and Gm-bHAD genes, express and purify the recombinant protein, and grow diffraction quality crystals have been described [10]. Native X-ray diffraction data were collected with a Mar CCD 300 detector on the ID-21-LS-CAT section G beamline at the Advanced Photon Source at Argonne National Laboratory. Diffraction data were processed using DENZO and the integrated intensities were

3

66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 3. Results and discussion 90 91 3.1. Structure of Gs-bHAD and Gm-bHAD 92 93 Despite crystallizing in distinctly different space groups (see 94 Table 1), the crystallographic asymmetric unit for each protein 95 Table 1 96 X-ray data collection and structure refinement statistics for Gs-bHAD and Gm-bHAD. 97 Q7 Values in parentheses are for the highest resolution shell. 98 Parameters Gs-bHAD Gm-bHAD 99 100 PDB entry 3PDU 3PEF Data collection 101 X-ray source APS: 21-ID-G APS: 21-ID-G 102 Detector Mar CCD 300 Mar CCD 300 103 Space group P21212 P1 104 a (Å) 99.61 75.97 105 b (Å) 147.49 79.14 c (Å) 182.47 95.47 106 a( ) 90 82.15 107 b ( ) 90 88.80 108 g ( ) 90 87.66 109 Resolution (Å) 28.81e1.89 43.16e2.07 (1.97e1.89) (2.15e2.07) 110 Wavelength (Å) 0.9793 0.9793 111 Multiplicity 6.9 (5.3) 2.5 (2.3) 112 I/s(I) 19.2 (3.2) 9.6 (4.4) 113 Completeness (%) 92.3 (82.7) 95.2 (89.0) a 114 Rmerge (%) 8.3 (42.7) 8.1 (20.4) Refinement 115 Rwork (%)b 0.151 0.157 116 c Rfree (%) 0.186 0.209 117 RMSD in bond length (Å) 0.007 0.007 118 RMSD in bond angle ( ) 1.070 1.166 Average B-factor (Å2) 119 Protein 27.72 39.05 120 Solvent 38.07 38.96 121 Molprobity analysis 122 Clashscore 4.88 15.08 123 Ramachandran favored (%) 97.98 97.27 Ramachandran outlier (%) 0.00 0.09 124 Residues with bad bonds (%) 0.00 0.04 125 Residues with bad angles (%) 0.00 0.18 126 P P a Rmerge ¼ jIi j/ , where Ii is the observed intensity and is the 127 average intensity over symmetry equivalent measurements. P P 128 b Rwork ¼ jjFobsj jFcalcjj/ jFobsj. c 129 Rfree is the same as Rwork but for 5% of all reflections that were not used in 130 crystallographic refinement.

scaled using SCALEPACK from the HKL-2000 program package [11]. The structure for the Gs-bHAD crystal was solved by molecular replacement using AMoRe [12] from the CCP4 program suite [13] with the crystal structure of glyoxylate reductase 1 from Arabidopsis (AtGLYR1, PDB entry 3DOJ) [7]; as the search model after alignment modification with CHAINSAW [14]. Missing side chains were manually rebuilt into the electron density maps using Coot [15] followed by numerous iterative rounds of restrained refinements using REFMAC [16] and PHENIX (v1.5_2) [17]. Improvements in structure quality during the iterative process were monitored by decreases in Rwork and Rfree. The structure for the GmbHAD crystal was solved in a similar manner except the final coordinates for the Gs-bHAD structure was used as the molecular replacement search model. The final Gs-bHAD model yielded a Rwork of 0.151 and a Rfree of 0.186 while the final Gm-bHAD model yielded a Rwork of 0.157 and a Rfree of 0.209. In the later stages of refinement the stereochemistry of the final structures was assessed by MOLPROBITY [18] and any conflicts addressed. MOLPROBITY showed good geometry and fitness for the final Gs-bHAD and GmbHAD models (Table 1). These coordinates and structure factors were deposited in the RCSB Protein Data Bank (PDB) with accession numbers 3PDU and 3PEF, respectively.


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

BIOCHI4453_proof ■ 29 May 2014 ■ 4/9

4


contained eight protein monomers, each with a single bound molecule of NADPþ. Eleven glycerol and 1809 water molecules were also modeled into the asymmetric unit of Gs-bHAD, while seven glycerol, five polyethylene glycol, one 1,2-ethanediol and 1099 water molecules were modeled into the asymmetric unit of GmbHAD. Missing, or uninterpretable electron density for the main chain occurred only for a few residues at the N- and C-termini of a few chains (Gs-bHAD: M1 chain B, C, F, and H, A287 chain G; GmbHAD: M1 chain F and H, R287 chain D, E, and F). There are no significant conformational differences in the structures between the eight monomers within the asymmetric unit for either protein; the a-carbons for all monomers superimpose on each other with an RMSD ranging from 0.2 to 0.4 Å for Gs-bHAD and from 0.2 to 0.5 Å for Gm-bHAD (UCSF-Chimera) [19]. Fig. 2A is a superposition of a single protein monomer in the asymmetric unit of Gs-bHAD and Gm-bHAD that illustrates the structures of both proteins are very similar. Between the two illustrated structures, the RMSD of the backbone a-carbons is 0.56 Å and of all non-proton atoms is 0.83 Å [20]. The elements of secondary structure observed for each protein unit in Gs-bHAD and Gm-bHAD are superimposed on the secondary structure alignment in Fig. 1B and show the minor differences between the two proteins are small variations in the length of a few elements of secondary structure. Both proteins are essentially composed of 14 a-helices and nine b-strands organized into two domains (Fig. 2B). Domain 1 (1e165) is a typical Rossmann fold [21] composed of two a/b units: a six-strand parallel b-sheet (b3:b2:b1:b4:b5:b6) surrounded by six a-helices (a1ea6) followed by a mixed b-sheet ([b7:Yb8:Yb9) surrounded by two a-helices (a7 and a8). Domain 2 (166e287) is composed of a bundle of a-helices (a9ea14) with the long a-helix, a9, connecting the two domains. Within the asymmetric unit of the Gs-bHAD and Gm-bHAD crystals, the protein monomers are clearly arranged as tetramers. Analysis of the inter-subunit interactions with the quaternary prediction program PISA [22] supports this interpretation; the most stable biological unit for each protein is a tetramer created by a dimer of dimers as shown in Fig. 3 for Gs-bHAD. The adoption of such a quaternary structure results in 19,618 Å2 and 19,550 Å2 of buried surface area for Gs-bHAD and Gm-bHAD, respectively.

Relative to the accessible surface area of each monomer, tetramer formation buries ~39% of each protein's surface, a value that strongly favors self-association [22]. The inter-subunit interactions involve Domain 2 of each protomer and are mediated by both hydrophobic contacts and hydrogen bond formation [23]. Such interactions along almost the entire length of one face of two antiparallel helices, a9, are largely responsible for the formation of one set of dimers (Fig. 3B) (along with minor contributions from a10:a9 and a13:a13 contacts). In turn, these two dimers are held together by interactions along part of one face of anti-parallel helices a12 with a12 and a13 with a13 (Fig. 3A). The adoption of such tetramer complexes has been observed for other b-HADs [6,7,23] and was consistent with the molecular weight observed for GsbHAD and Gm-bHAD in solution by size exclusion chromatography [10]. While the structures for Gs-bHAD and Gm-bHAD may be expected to be similar to each other due to the 65% sequence identity and 77% sequence similarity of the two proteins in the same bacterial genera, their overall structures are also generally similar to the structure determined for other b-HADs. This is evident from a DALI search [24] of the Protein Data Bank for structures similar to Gs-bHAD and Gm-bHAD that identifies many structures of b-HADs with Z-scores greater than 20 and backbone a-carbons RMSDs 3.0 Å despite sequence identities as low at 17% (data not shown). Fig. 1B is a ClustalW2 alignment of the amino acid sequences containing two of these DALI hits, b-HADs from the infectious prokaryote Pseudomonas aeruginosa (Pa-bHAD, 3OBB) [6] and the plant A. thaliana (At-bHAD, 3DOJ) [7], together with human b-HAD (Hs-bHAD, no structure in PDB) and Gs-bHAD and Gm-bHAD. The identity and similarity of the aligned sequences, shown in Supplementary Table 1, range from 28 to 65% and from 45 to 77%, respectively. As noted previously [1], there is significant sequence identity and similarity in the consensus sequences of the four signature conserved regions (shaded blue, cyan, green, and gray in Fig. 1B) in the aligned prokaryotic and eukaryotic b-HADs. Dispersed rather evenly between these four conserved signature regions are many other highly conserved regions that are likely responsible for the adoption of the proper folding of the protein and assembly into tetramer quaternary units.

Fig. 2. A) Superposition of a single protein chain in the crystallographic asymmetric unit of Gs-bHAD (chain A, wheat) and Gm-bHAD (chain B, marine) shown as a cartoon representation using the program Superpose [20]. B) Single protein chain in the crystallographic asymmetric unit of Gs-bHAD (chain A, wheat) shown as a cartoon representation. The consensus sequence of the four conserved regions observed in most b-HADs is colored on the structure: Dinucleotide Cofactor-Binding Region (blue), Substrate-Binding Region (cyan), Catalysis Region (green), and Cofactor-Binding Region #2 (black). The bound NADPþ is shown as a spherical representation colored red and the two domains (1e165 and 166e287) are labeled.


66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

BIOCHI4453_proof ■ 29 May 2014 ■ 5/9


5

Fig. 3. A) Cartoon representation of the tetramer in the crystallographic asymmetric unit of Gs-bHAD with the largest buried surface area. Each protomer chain is colored differently with the NADPþ bound to each protomer shown as a red spherical representation. The secondary structural elements of Domain 2 are labeled. B) The orientation in (A) rotated 90 towards the viewer about the x-axis.

3.2. Dinucleotide cofactor-binding site All of the four conserved regions in the primary amino acid sequences of b-HADs necessary for enzyme activity (Fig. 1A) are spatially located near each other at the interdomain cleft in both Gs-bHAD and Gm-bHAD as illustrated in Fig. 2. In the crystal structures of both Gs-bHAD and Gm-bHAD, one molecule of NADPþ is also observed buried deep in this interdomain cleft (Fig. 2B). A close-up of the orientation of the NADPþ within the cleft is shown in Fig. 4A with details of the molecular interactions between NADPþ and each protein summarized schematically in Fig. 5. Aside from two side chain residues, K36 and H236, that interact with NADPþ in Gm-bHAD but are too distant to interact in Gs-bHAD, the same set of (mostly) conserved side chains and main chain groups are responsible for holding NADPþ in place. The adenine moiety of NADPþ, which is located in Domain 1, not only makes hydrophobic contact with the side chain of R32 that sits overtop of the base, but also the C6 amine group forms a hydrogen bond with the side chain oxygen of E73 (Fig. 4A). The 20 phosphate group of the ribose moiety attached to adenine

interacts, perhaps electrostatically, with the side chains of N31 and R32. For Gm-bHAD the side chain of K36 is also close enough to form a salt bridge with the 20 -phosphate group of NADPþ and further stabilize the complex. In the conserved Dinucleotide Cofactor-Binding consensus sequence (Fig. 1A), a negatively charged aspartate residue at the C-terminal end usually signals that the b-HAD prefers NADþ binding while a positively charged arginine residue usually signals that the b-HAD prefers NADPþ binding [9]. In the Gs-bHAD and Gm-bHAD amino acid sequences this position is occupied by neither an aspartate nor arginine residue, but instead, by an asparagine residue. However, one residue outside the conserved Dinucleotide Cofactor-Binding consensus sequence there is an arginine residue. In the crystal structures of Gs-bHAD and Gm-bHAD the side chain of this arginine residue is in a position to interact with the phosphate group of NADPþ. If this arginine residue is responsible for the observed preference of Gs-bHAD and Gm-bHAD for NADPþ over NADþ [10] then it may be necessary to modify the Dinucleotide CofactorBinding consensus sequence of b-HADs slightly to accommodate an extra residue.

Fig. 4. A) Close-up view of the position of the bound NADPþ within the interdomain cleft of chain A in Gs-bHAD. The NADPþ atoms are colored as follows: carbon ¼ yellow, oxygen ¼ red, nitrogen ¼ blue, phosphorus ¼ purple. Residues whose side chains make contact with NADPþ are labeled. B) Close-up view of the glycerol molecule modeled into the Substrate-Binding Region and Catalysis Region of the A protomer of Gm-bHAD. The NADPþ is shown in a red spherical representation and the glycerol atom is identified by a black arrow with the atoms colored as follows: carbon ¼ yellow, oxygen ¼ red. Residues whose side chain and/or main chain atoms make contact with glycerol are labeled. In both figures the consensus sequences of the conserved regions are colored as described in Fig. 2.


66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

BIOCHI4453_proof ■ 29 May 2014 ■ 6/9

6


Fig. 5. Schematic representation of hydrophobic contacts (red arc with radiating spokes) and potential hydrogen bonds (green dashed lines) between NADPþ (purple bonds) and either Gs-bHAD (chain A) or Gm-bHAD (chain E) (orange bonds). Residues that make contact with NADPþ via hydrophilic side chain interactions are underlined. The NADPþ atoms are colored as follows: carbon ¼ black, oxygen ¼ red, nitrogen ¼ blue, phosphorus ¼ purple. The image was generated with the program LIGPLOT [46].

The pyrophosphate moiety of NADPþ lies closer to the center of the interdomain cleft and is coordinated by interactions with the main chain amido groups of I11 and M12. For Gm-bHAD this moiety is further coordinated by interaction with the side chain of H236 that lies just outside the consensus sequence of Cofactor-Binding Region #2 (Fig. 1B). The nicotinamide moiety of NADPþ showed planar electron density consistent with an oxidized state (data not shown) with parts of it physically extended into Domain 2. The side chain 3-amino group of the highly conserved K239 residue in the consensus sequence of Cofactor-Binding Region #2 interacts with the 20 - and 30 -hydroxyl groups of the ribose. The side chains of three other highly conserved residues, M12, V121, and F232, form a hydrophobic pocket around the nicotinamide ring. The amide functional group on the nicotinamide ring interacts with the main chain keto group of residue 231 and the side chain hydroxyl group of T124 (Gs-bHAD) or S124 (Gm-bHAD) with the amide group rotated 180 in each protein to accommodate the interactions. The best crystals for both Gs-bHAD and Gm-bHAD were obtained in the presence of 1 mM NADPþ (~2:1 NADPþ:protein molar ratio) during crystallization [10]. No crystals could be obtained for Gm-bHAD under any screened condition in the absence of NADPþ or in the presence of NADþ. On the other hand, crystals for Gs-bHAD were readily grown under a number of conditions in the absence of NADPþ, however, they diffracted poorly (7e8 Å resolution). These results are consist with enzyme activity assays showing that both proteins are capable of reducing SSA in the presence of NADPþ and poorly in the presence of NADþ [10]. Large conformational changes have been observed for some bHADs, such as Escherichia coli 6-phosphogluconate dehydrogenase, upon cofactor binding. While suitable apo-crystals of Gs-bHAD and Gm-bHAD were not obtained, the amino acid sequence of At-bHAD aligns well (no insertions or deletions) with the sequences of both Geobacter enzymes (Fig. 1B). The structure of At-bHAD is for the apo-enzyme [7] and overall its Ca backbone atoms superimpose on the structures of Gs-bHAD and Gm-bHAD with an RMSD of 0.89 and 0.88 Å, respectively (http://fatcat.burnham.org). On the other hand, if only the Ca backbone atoms of each Domain are superimposed, 0.1e0.3 Å lower RMSD values are obtained indicating a difference in the relative orientation of the two Domains. This is illustrated for Domain 2 of Gs-bHAD and At-bHAD (3OBB) in Fig. 6. The clear

Fig. 6. Cartoon representation of the structural superposition of the Ca backbone atoms of Domain 2 of Gs-bHAD (3PDU) and At-bHAD (3DOJ). The At-bHAD (wheat) is in the open “apo” form with no substrate or cofactor bound while Gs-bHAD (green) is in the closed form with bound NADPþ (stick representation). The black arrow indicates the direction of the rigid body rotation necessary to align Domain 1 of both proteins. The Ca backbone RMSD of the superposition of Domain 1 is 0.76 Å, Domain 2 is 0.61 Å, and the entire structure is 0.89 Å.


66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

BIOCHI4453_proof ■ 29 May 2014 ■ 7/9


conformational difference observed between the two enzymes is manifested by simple rigid body rotation of Domain 1. Interestingly, the domain motion does not seemingly utilize a “hinge” within or at the beginning of helix a9. Perhaps the inability to grow diffraction quality crystals of Gs-bHAD and Gm-bHAD in the absence of the correct cofactor reflects dynamic or conformational differences between the two domains of the apo-protein and the NADPþbound protein.

7

HAD consensus regions, are in position to make van der Waal contacts with the glycerol. The 3-amino group of the side chain of K171, a highly conserved residue in the Catalysis Region (green) that acts as a general base in the acid/base catalytic mechanism of b-HADs [3,9], is positioned near a terminal hydroxyl group of the glycerol (

Carboxydotrophic growth of Geobacter sulfurreducens.

Kinetic and structural characterization for cofactor preference of succinic semialdehyde dehydrogenase from Streptococcus pyogenes.

Direct interspecies electron transfer between Geobacter metallireducens and Methanosarcina barkeri.

Physiological stratification in electricity-producing biofilms of Geobacter sulfurreducens.

Rational engineering of Geobacter sulfurreducens electron transfer components: a foundation for building improved Geobacter-based bioelectrochemical technologies.

Enhanced uranium immobilization and reduction by Geobacter sulfurreducens biofilms.

Succinic semialdehyde reductase Gox1801 from Gluconobacter oxydans in comparison to other succinic semialdehyde-reducing enzymes.

Proteome of Geobacter sulfurreducens in the presence of U(VI).

Succinic semialdehyde dehydrogenase of wheat grain.

Backbone, side chain and heme resonance assignments of the triheme cytochrome PpcD from Geobacter sulfurreducens.

Thermodynamic and kinetic characterization of two methyl-accepting chemotaxis heme sensors from Geobacter sulfurreducens reveals the structural origin of their functional difference.

Cooperative growth of Geobacter sulfurreducens and Clostridium pasteurianum with subsequent metabolic shift in glycerol fermentation.

Constraint-based modeling of carbon fixation and the energetics of electron transfer in Geobacter metallireducens.

Charge transport in films of Geobacter sulfurreducens on graphite electrodes as a function of film thickness.

Direct and quinone-mediated palladium reduction by Geobacter sulfurreducens: mechanisms and modeling.

Biomarkers in a Taurine Trial for Succinic Semialdehyde Dehydrogenase Deficiency.

Identification of genes specifically required for the anaerobic metabolism of benzene in Geobacter metallireducens.

Incidence and Geographic Distribution of Succinic Semialdehyde Dehydrogenase (SSADH) Deficiency.

Deciphering the electric code of Geobacter sulfurreducens in cocultures with Pseudomonas aeruginosa via SWATH-MS proteomics.

Metabolic efficiency of Geobacter sulfurreducens growing on anodes with different redox potentials.

The iron stimulon and fur regulon of Geobacter sulfurreducens and their role in energy metabolism.

Mechanistic stratification in electroactive biofilms of Geobacter sulfurreducens mediated by pilus nanowires.

Bidirectional Photoinduced Electron Transfer in Ruthenium(II)-Tris-bipyridyl-Modified PpcA, a Multi-heme c-Type Cytochrome from Geobacter sulfurreducens.

Natural history of succinic semialdehyde dehydrogenase deficiency through adulthood.