The Crystal Structure of the Transthyretin-like Protein from Salmonella dublin, a Prokaryote 5-Hydroxyisourate Hydrolase

J. Mol. Biol. (2006) 359, 1389–1399

doi:10.1016/j.jmb.2006.04.057

The Crystal Structure of the Transthyretin-like Protein from Salmonella dublin, a Prokaryote 5-Hydroxyisourate Hydrolase Sarah C. Hennebry 1 †, Ruby H. P. Law 2,3 †, Samantha J. Richardson 1 2, 3, 4 Ashley M. Buckle 2,3 ⁎ and James C. Whisstock ⁎ 1

The Department of Biochemistry and Molecular Biology, Bio21 Molecular Science and Biotechnology Institute, 30 Flemington Road, The University of Melbourne, Parkville, VIC 3010, Australia 2

The Protein Crystallography Unit, and The Department of Biochemistry and Molecular Biology, School of Biomedical Sciences, Faculty of Medicine, Monash University, Clayton, VIC 3800, Australia 3

Victorian Bioinformatics Consortium, P O Box 53, Monash University, Clayton, VIC 3800, Australia 4ARC Centre for Structural and Functional Microbial Genomics, Monash University, Clayton, VIC 3800, Australia

*Corresponding authors

The mechanism of binding of thyroid hormones by the transport protein transthyretin (TTR) in vertebrates is structurally well characterised. However, a homologous family of transthyretin-like proteins (TLPs) present in bacteria as well as eukaryotes do not bind thyroid hormones, instead they are postulated to perform a role in the purine degradation pathway and function as 5-hydroxyisourate hydrolases. Here we describe the 2.5 Å X-ray crystal structure of the TLP from the Gram-negative bacterium Salmonella dublin, and compare and contrast its structure with vertebrate TTRs. The overall architecture of the homotetramer is conserved and, despite low sequence homology with vertebrate TTRs, structural differences within the monomer are restricted to flexible loop regions. However, sequence variation at the dimer–dimer interface has profound consequences for the ligand binding site and provides a structural rationalisation for the absence of thyroid hormone binding affinity in bacterial TLPs: the deep, negatively charged thyroxine-binding pocket that characterises vertebrate TTR contrasts with a shallow and elongated, positively charged cleft in S. dublin TLP. We have demonstrated that Sdu_TLP is a 5-hydroxyisourate hydrolase. Furthermore, using site-directed mutagenesis, we have identified three conserved residues located in this cleft that are critical to the enzyme activity. Together our data reveal that the active site of Sdu_TLP corresponds to the thyroxine binding site in TTRs. © 2006 Elsevier Ltd. All rights reserved.

Keywords: transthyretin; 5-hydroxyisourate

Introduction Transthyretin (TTR) is a tetrameric protein that is responsible for transport of thyroxine and related molecules. Numerous human mutations of TTR have been characterised that result in protein misfolding, amyloidosis and the development of neurodegenerative diseases (for example, familial † S.C.H. and R.H.P.L. contributed equally to this work. Abbreviations used: 5-HIU, 5-hydroxyisourate; OHCU, 2-oxo-4-hydroxy-4-carboxy-5-ureidoimidazoline; Sdu_TLP, Salmonella dublin TLP. E-mail addresses of the corresponding authors: [email protected]; [email protected]

TLP;

predicting

function

from

structure;

amyloidotic polyneuropathy) and cardiomyopathies.1–3 Transthyretin-like proteins (TLPs) are a family of homologous proteins present in a large range of bacterial, fungal, plant, invertebrate and vertebrate species and are postulated to have arisen as a result of an early gene duplication event.4–6 TLPs characterised to date share significant (∼35%) sequence similarity with TTRs, are able to form tetramers, however are not able to bind thyroid hormones. 4 While the structure, function and dysfunction of vertebrate TTRs is well characterised, the role of the related TLPs remains to be fully understood. Whereas while much structural information is available for vertebrate TTRs, the structure of a representative TLP remains to be determined.

0022-2836/$ - see front matter © 2006 Elsevier Ltd. All rights reserved.

Crystal Structure of Sdu_TLP

1390 Knockout studies performed in Bacillus subtilis together with biochemical investigation of the TLP from this organism suggested that TLPs function in the uricase reaction, the pathway responsible for the oxidation of uric acid to allantoin (Figure 1).7–9 Specifically, a study by Lee et al.8 suggested that TLPs catalyse the hydrolysis of 5-hydroxyisourate (5-HIU) to 2-oxo-4-hydroxy-4-carboxy-5-ureidoimidazoline (OHCU). In addition, these authors show that a related TLP from Escherichia coli also possesses the ability to hydrolyse 5-HIU. Finally, a recent study revealed that murine TLP is also able to function to accelerate 5-HIU hydrolysis, supporting a global role for TLPs in purine metabolism.10 Here we have determined the X-ray crystal structure of the TLP from the Gram-negative bacterium Salmonella dublin (Sdu_TLP). It is demonstrated that, like B. subtilis TLP, Sdu_TLP is able to catalyse the hydrolysis of 5-HIU to OHCU. Furthermore, mutagenesis studies revealed that the active site of Sdu _ TLP is centred upon the region corresponding to the thyroxine binding site in TTRs and that mutation of an absolutely conserved histidine in this region abolishes activity.

Results Structure of TLP from Salmonella dublin The final model consists of residues 1–28 and 30– 111 (residue 29 is not visible in the electron density), two sulphate molecules and 32 water molecules. It has a Rcryst of 22.6% and a Rfree of 26.1% for all reflections between 47 and 2.5 Å. All residues are in the most favoured and allowed regions of the Ramachandran plot. A summary of statistics is provided in Table 1. The overall structure is consistent with the archetypal TTR fold:11,12 a prealbumin-like fold consisting of an eight-stranded β-sandwich, each sheet adopting a greek key topology, with an additional two-turn α-helix

Table 1. Data collection and refinement statistics Data collection Space group Cell dimensions (Å): a, b, c Resolution (Å) Total number of observations Number of unique observations Multiplicity Data completeness (%) Rmerge (%) a

P6 94.70, 94.70, 57.38 46.98–2.49 51,332 10,308 5.0 99.5 (96.7) 14.1 (2.1) 11.1 (45.7)

Structure refinement Non-hydrogen atoms Protein Solvent Sulphates Rfree (%) b Rcryst (%)

1774 2739 32 2 26.1 22.6

Rms deviations from ideality Bond lengths (Å) Bond angles (°)

0.005 1.19

B factors (Å2) Mean main chain Mean side chain Mean water molecule Mean sulphate molecule r.m.s. deviation bonded Bs

24.5 24.8 38.7 33.7 1.0

Values in parentheses refer to the highest resolution shell. a Agreement between intensities of repeated measurements of the same reflections and can be defined as: Σ(Ih,i–)/Σ Ih,i, where Ih,i are individual values and is the mean value of the intensity of reflection h. b The free R factor was calculated with the 5% of data omitted from the refinement.

between strands E and F (Figure 2(a)). The asymmetric unit consists of two protein molecules, which dimerise via intermolecular hydrogen bonds between H-strands to form an eight-stranded β-sheet (Figure 2(a)). This interaction is found in all TTR crystal structures published to date. Tight noncrystallographic symmetry (NCS) restraints applied during refinement yield identical molecules in the asymmetric unit. A tetrameric ‘‘dimer of dimers’’, a

Figure 1. The uricase pathway responsible for the oxidation of uric acid to allantoin. Adapted from Kahn and Tipton.

Crystal Structure of Sdu_TLP

1391

Monomer A *

F C

D

A

B G

H

H

E

E

G

A

B

C

D

F *

*

Monomer B

(a) Monomer A

Monomer A’

Monomer B

Monomer B’

(b) feature of all other TTR structures, can be generated by applying crystallographic symmetry, and consists of two dimers that associate via non-polar interactions between the loops joining strands G and H with loops joining strands A and B (Figure 2(b)). Gel filtration confirmed that Sdu_TLP adopts a tetrameric form in solution (Figure 3). Structural and sequence comparisons of Sdu_TLP with transthyretins We compared the structure of Sdu_TLP with human (1F41),13 chicken (1TFP),14 rat (1KGI),15

Figure 2. The 2.5 Å crystal structure of Sdu_TLP. (a) Dimeric structure within the asymmetric unit of the crystal, showing the interaction between the β-strands H of each monomer. Secondary structure elements are labelled. The C and N termini are marked with an asterix (*). (b) The homotetrameric structure generated by crystal symmetry, in an orthogonal orientation to (a). Approximate positions of thyroid hormone binding sites at the dimer–dimer interface (shown as a dotted line) are indicated by black circles.

and fish (1SN0)16 TTR. A comparison of the relative orientations and positions of monomers within the tetrameric arrangement of Sdu_TLP, human, rat, fish and chicken TTRs reveals that human, rat and fish TTRs superimpose to within an RMSD of 1.0 Å, whereas chicken TTR and Sdu_TLP monomers show larger shifts (up to ∼4 Å) in the relative orientations of each monomer within the tetramer (Figure 4(a)). Further, the total contacting surface area between molecules within the tetramer (5275 Å) is 6–14% smaller compared to fish (5625 Å), human (5810 Å), chicken (5828 Å), and rat (6133 Å).

1392

Crystal Structure of Sdu_TLP

Figure 3. Purification of Sdu_TLP and determination of native molecular mass. The molecular mass of native wildtype Sdu_TLP proteins determined by gel filtration was compared with denatured molecular mass determined by SDSPAGE. The results revealed that TLPs are tetramers. Inset (i) shows the calibration of the gel filtration column. Standards used were albumin (66 kDa), ovalbumin (44 kDa), chymotrypsinogen (25 kDa) and cytochrome c (13 kDa). Inset (ii): 15% SDS-PAGE of TLPs following purification by ion-exchange and gel filtration chromatography. Lane 1, molecular mass marker; lane 2, 20 μg wild-type Sdu_TLP; lane 3, 20 μg H6A TLP; lane 4, 20 μg H95A TLP; lane 5, 10 μg Y108F TLP. Proteins are stained with Coomassie blue. Native molecular masses obtained from mutants H6A, H95A and Y108F were the same as that of the wild-type Sdu_TLP (data not shown).

Assuming an approximate empirical relationship between oligomeric buried surface area and oligomer stability,17 the smaller oligomeric surface contact area in Sdu_TLP is consistent with reports that the E. coli TLP tetramer is less stable than its vertebrate counterparts.18 Structural superposition of monomers from each species reveals that the overall fold is conserved. Generally, the structural deviation between Sdu_TLP and vertebrate TTRs is of the same order of magnitude to that within the set of vertebrate TTRs (Table 2). Exceptions and regions of plastic deformation (Figure 4(b)) include the D strand and DE loop (residues 47–57) and FG loop (residues 86–90). It is noteworthy that amongst TTRs, chicken TTR shows the largest structural deviations, particularly in the αhelix and the loop connecting this helix to the F strand. This has been attributed to the presence of a sulphate ion that binds to the FG loop and C terminus.16 A sequence alignment of Sdu_TLP with vertebrate TTRs of known structure reveals sequence identities of 29–36% (Table 3 and Figure 4(c)). Residues that are absolutely conserved in the TLP (but not the TTR) family cluster in the vicinity of the dimer–dimer interface and the region corre-

sponding to the thyroxine binding site in TTRs (Figure 4(d)). It is generally well understood that critical functional residues in an enzyme's active site are strongly selected for during evolution, and are therefore more likely to be conserved in comparison to non-functional residues. The clustering of conserved residues at the putative active site of Sdu_TLP indicates that these residues have undergone some selective pressure to maintain enzymatic function (Figure 4(c) and (d)). In contrast, TLP-conserved residues that are also conserved with the vertebrate TTRs are more diffusely located in the molecule, consistent with these amino acids playing structural rather than functional roles. Given this pattern of sequence conservation, together with the functional role of the dimer–dimer interface in ligand binding in TTRs, we postulated that the two pockets formed by the dimer–dimer interface in Sdu_TLP represent two active sites and contained residues crucial for 5-HIU hydrolysis. Nature of the putative active site of Sdu_TLP Examination of the dimer–dimer interface of Sdu_TLP reveals that the conserved residues His6,

Crystal Structure of Sdu_TLP

1393

His95 and Tyr108 localise to this region. Comparison with human TTR reveals that Tyr108 in particular has a profound effect on the size and shape of the putative ligand-binding site (Figures 5 and 6). In TTRs the amino acid corresponding to residue 108 is a conserved Thr119 and substitution to a conserved Tyr in TLPs dramatically reduces the depth of this pocket.

The conserved Leu99 in Sdu_TLP also adopts a conformation that plugs the base of the pocket in comparison to Leu110 in human TTR (Figure 6(a)). Thus whilst the TTR surface is deep and funnel-shaped, the TLP surface is more groove-like (Figure 6). Furthermore, the surface charge of the putative ligand-binding site of Sdu_TLP differs considerably from the TTR

(a) FG loop DE loop

H95 Y108

(b) Figure 4 (legend on next page)

H6

R44

Crystal Structure of Sdu_TLP

1394

(c) S dublin Fish Human Chicken Rat

A

E S dublin Fish Human Chicken Rat

B

C

D

* * MILSVHILDQQTGKPAPGVEVVLEQKK-DNGWTQLNTGHTDQDGRIKALWP APTPTDKHGGSDTRCPLMVKILDAVKGTPAGSVALKVSQKTADGGWTQIATGVTDATGEIHNLIT GPTGTGESKCPLMVKVLDAVRGSPAINVAVHVFRKAADDTWEPFASGKTSESGELHGLTT APLVSHGSVDSKCPLMVKVLDAVRGSPAANVAVKVFKKAADGTWQDFATGKTTEFGEIHELTT GPGGAGESKCPLMVKVLDAVRGSPAVDVAVKVFKKTADGSWEPFASGKTAESGELHGLTT 1 10 20 30 40 50 α

F

G

H

* * * EKAAAPGDYRVIFKTGQYFESKKLDTFFPEIPVEFHISKT-NEHYHVPLLLSQYGYSTYRGS EQQFPAGVYRVEFDTKAYWTNQGSTPFHEVAEVVFDAHPEGHRHYTLALLLSPFSYTTTAVVSSVRE EEEFVEGIYKVEIDTKSYWKALGISPFHEHAEVVFTANDSGPRRYTIAALLSPYSYSTTAVVTNPKE EEQFVEGVYRVEFDTSSYWKGLGLSPFHEYADVVFTANDSGHRHYTIAALLSPFSYSTTAVVSDPQE DEKFTEGVYRVELDTKSYWKALGISPFHEYAEVVFTANDSGHRHYTIAALLSPYSYSTTAVVSNPQN 60 70 80 90 100 110

Y108

H6 H95

(d) Figure 4. Sequence and structural comparisons Sdu_TLP with vertebrate transthyretins. (a) Stereo diagram showing a superposition of tetramers of Sdu_TLP (magenta) with human (1F41, cyan),13 rat (1KGI, yellow),15 chicken (1TFP, orange)14 and fish (1SNO, green)16 TTRs. Tetramers were superposed using the A chain only. (b) Stereo diagram showing a superposition of monomers of Sdu_TLP (magenta) with human (1F41, cyan), rat (1KGI, yellow), chicken (1TFP, orange) and fish (1SNO, green) TTRs. Bound thyroxine ligand in the human structure is shown as sticks. Regions of plastic deformation are labelled. Active site residues are shown as red sticks. (c) Sequence alignment. Sdu_TLP residues conserved within the TLP family are highlighted in green. Residues that are conserved also in fish, human, chicken and rat, are highlighted in orange. Active site residues are indicated by an asterisk. Residues mutated in this study are indicated by a red asterisk. Secondary structure elements are indicated. Residue numbering is for Sdu_TLP. (d) Cartoon representation, in stereo, of the Sdu_TLP tetramer. Sdu_TLP residues conserved within the TLP family are shown as green sticks. Residues that are conserved also in fish, human, chicken and rat are shown as orange sticks. Active site residues that are mutated in this study are labelled. The dimer–dimer interface is indicated by a dotted line.

counterpart: residues Arg44, Arg46 and Arg109 in particular produce an electrostatically positive groove. In contrast the thyroxine-binding site of vertebrate TTRs is negatively charged (Figure 6(c) and (d)). We were interested to investigate the role of TLP conserved residues in and around the putative active site of Sdu_TLP in 5-HIU hydrolysis. We chose residues that could logically perform a chemical role in catalysis and thus targeted His6, His95 and Tyr108 for site-directed mutagenesis. These residues were

mutated to Ala, Ala and Phe, respectively, and their role in TLP function investigated. Sdu_TLP is able to catalyse 5-HIU hydrolysis and mutations in the active site abolish this activity We confirmed that recombinant Sdu_TLP possessed similar 5-HIU hydrolysis activity to that observed in B. subtilis, E. coli and mouse TLPs.8,10 We calculated the rate of 5-HIU hydrolysis using the published

Crystal Structure of Sdu_TLP

1395

Table 2. RMSD (Å) between vertebrate TTRs and Sdu_TLP Structure

Fish

Human

Chicken

Rat

S. dublin

1.04

1.59 1.22

1.04 0.34 1.22

1.36 1.40 1.88 1.44

Fish Human Chicken Rat Net RMSD = 1.31

Superposition was performed using MUSTANG. Net RMSD was calculated for 110 residues in each structure, and is defined as by Konagurthu et al.30

extinction coefficient of 5-HIU at 312 nm (8500 M−1 cm−1)7 and demonstrated that the Sdu_TLP possesses 5-HIUase activity (Figure 7). One milligram of wildtype TLP catalysed the hydrolysis of 10 μmol of 5-HIU in 1 min (Figure 7). The mutants revealed that all three residues were important for Sdu_TLP function. The H95A and Y108F mutations reduced 5-HIU hydrolysis activity by 90%. The H6A mutation completely abolished activity (Figure 7). Using gel filtration we confirmed that all mutations formed tetramers indistinguishable from wild-type (data not shown). These data demonstrate that the active site of Sdu_TLP localises to the dimer–dimer interface and, in addition reveal that key conserved residues, in particular His6 participate in 5-HIU hydrolysis.

structure of Sdu_TLP (Figure 4(d)). Interestingly, several highly conserved residues in the TLP family (His6, His95 and Tyr108) including the characteristic C-terminal YRGS motif, localized to the dimer– dimer interface corresponding to the thyroxinebinding site in vertebrate TTRs. Examination of this region in TLP immediately revealed a structural basis for the inability of TLP to bind thyroxine in an analogous fashion to TTRs. Specifically, in TLPs the corresponding cavity is positively rather than negatively charged, and is less deep, primarily as a result of substitution of Thr to Tyr at residue 108 (Sdu_TLP numbering) and a change in conformation of Leu99. However, the presence of highly conserved residues led us to test the hypothesis that this cleft may represent the active site of Sdu_TLP and we therefore investigated this region using site-directed mutagenesis of plausible catalytic residues (based upon physiochemical characteristics). Our data reveal that His6, at the top of the cavity is absolutely

(a) G110

H95* Y108* H6* L99

Discussion The function of TTR as a hormone transport protein is well characterised, as are the misfolding events of this protein that lead to amyloidosis and disease. The role of TLPs, however, remains obscure. Recent studies have suggested that rather than playing a transport role, TLPs are enzymes that function in purine metabolism, catalysing the conversion of 5-HIU to OHCU.8,10 In order to further understand the function of TLPs we have determined the 2.5 Å X-ray crystal structure of the TLP from S. dublin. The molecule adopts the typical prealbumin-like fold and is similar to the known structures of vertebrate TTRs, both in terms of the arrangement of molecules in the tetramer and within the monomeric subunit. Despite relatively low sequence identity between Sdu_TLP and vertebrate TTRs, the structural variation is no greater than that within the vertebrate family (Table 2). In order to investigate the catalytic function of TLPs we mapped conserved TLP residues onto the Table 3. Sequence identities (%) within vertebrate TTR family and Sdu_TLP Sequence Fish Human Chicken Rat

Fish

Human

Chicken

Rat

S. dublin

54

62 75

58 83 77

36 29 33 29

Multiple sequence alignment was performed using CLUSTALW.36

R44

(b) E54 K15

T119

T106

T119

Figure 5. A comparison of the ligand-binding cleft at the dimer interface in Sdu_TLP (a) and (b) human TTR. Residues that make up the binding pockets are shown. Hydrogen bonds are shown as broken cyan lines. Thyroxine is shown in stick representation in yellow. Some elements of secondary structure are not shown, for clarity. Residues mutated in this study are indicated with an asterisk. Residues His6, His95, and Y108 of TLP are equivalent (upon structure alignment) to K15, T106, and T119 of TTRs, respectively.

Crystal Structure of Sdu_TLP

1396

L99

H95 L99 Y108 Y108 H95 L99

H6 Y108

(a)

Y108 L99 H6

(b)

R46

R109 R44 Y108 H95 H6 H6 H95 Y108 R44 R46 R109

(c)

(d)

Figure 6. Surface characteristics and interactions at the ligand-binding sites. (a) Molecular surface of human TTR (apo-form) showing superimposed Sdu_TLP residues, at the dimer–dimer interface (dimers are coloured differently). (b) Electron density (contoured at 1σ, where σ is the rmsd of electron density in the unit cell) at the dimer–dimer interface of Sdu_TLP, looking down the non-crystallographic 2-fold axis. Water molecules are represented as red spheres. (c) Electrostatic potential surface of human TTR showing the thyroxine ligand inside the hormone-binding pocket at the dimer–dimer interface. Positively charged residues are in blue; negatively charged residues are in red. (d) Electrostatic potential surface of Sdu_TLP in the equivalent orientation.

required for catalytic activity (Figure 7). In contrast, mutation of His95 and Tyr108 reduced substantially (by 90%), but did not abolish activity. None of the mutations affected the tetrameric assembly of the molecule. These results clearly demonstrate that the active site of Sdu_TLP is centered in the groove at the dimer–dimer interface and that key conserved residues in this region perform a functional rather than structural role. The more open groove or channel in Sdu_TLP, indicated by a smaller dimer– dimer contact surface area compared to TTRs, is consistent with a classic enzyme site topology, allowing access and egress of substrates and products. This is in stark contrast to the deep, secluded thyroxine-binding pocket of TTRs that provide a tight binding, protective environment consistent with a role as hormone-binding protein. The residues that we mutated are 100% conserved across all TLP sequences (from all kingdoms).4 It is therefore likely that TLPs from other organisms are also able to hydrolyse 5-HIU and that this activity represents a general function for this branch of the family. Previously it has been postulated that TTR and TLP arose as a result of an early gene

duplication event.4–6 Indeed, classical TTRs are only found in vertebrates. In contrast, bioinformatic investigation reveals that TLPs are more widely distributed with members in prokaryotes, plants, lower eukaryotes, invertebrates and vertebrates. Thus it is reasonable to suggest that the transthyretin-like fold functioned originally in purine metabolism and that a non-enzymatic transport function arose later in evolution.

Materials and Methods Expression and purification Sdu_TLP was synthesised and purified as described.4 Briefly, the Sdu_TLP gene was cloned into the expression vector pET11a (Novagen), which was then transformed into Rosetta BL21 cells. Protein synthesis was induced with 1 mM IPTG at 30 °C for 4 h. Cells were harvested by centrifugation and then lysed by sonication in 25 mM TrisHCl (pH 7.5) containing 100 μg/ml lysozyme. Soluble protein was isolated by centrifugation of the sonicate at 15,000g for 10 min.

Crystal Structure of Sdu_TLP

1397 5′-GGA TCC TTA GCT CCC GCG AAA GGT TGA ATA ACC ATA C-3′. All mutant Sdu_TLPs were prepared as described for the wild-type protein. Hydrolysis of 5-hydroxyisourate by Sdu_TLP

Figure 7. Hydrolysis of 5-HIU by wild-type Sdu_TLP and by H6A, H95A and Y108F mutant Sdu_TLPs. 0.05 units/ml uricase was equilibrated in 50 mM potassium phosphate buffer (pH 7.8) for 3 min. The reaction was commenced with the addition of 50 μM uric acid. The appearance of 5-HIU was monitored spectrophotometrically at 312 nm. The amount of 5-HIU peaked after approximately 3 min and began spontaneous decomposition if left unperturbed. Addition of 0.04 μM wild-type Sdu_TLP resulted in the rapid hydrolysis of 5-HIU (10 μmol/min per mg TLP). Addition of equimolar amounts of H95A and Y108F mutant Sdu_TLPs resulted in reduced hydrolysis (1.2 μmol/min per mg TLP; activity reduced by 90%). The H6A mutation abolished hydrolytic activity (negligible hydrolysis of 0.1 μmol/min per mg TLP; activity reduced by 99%).

Soluble protein was dialysed against 50 mM Hepes (pH 6.95) overnight, before application onto a HiTrap SP ion exchange column (GE Healthcare). Sdu_TLP was eluted with 50 mM Hepes (pH 6.95) containing 0.5 M NaCl. Fractions were pooled, concentrated in 25 mM Tris-HCl (pH 8.8) containing 150 mM NaCl and 0.002% (w/v) sodium azide and then applied to a Superose-12 10/30 size-exclusion column (GE Healthcare). The Sdu_TLP was purified on the Superose-12 column as a tetramer (Figure 2). All purification steps were performed at 4 °C. The purity of protein preparation was determined by SDS-PAGE followed by Coomassie staining;19 the identity was confirmed by N-terminal sequencing; the molecular mass was determined by mass spectrometry and the concentration was determined by Bradford20 using BSA as standards as described.4 Mutagenesis Amino acid substitutions were introduced into the Sdu_TLP sequence using Quick-change mutagenesis kit (Stratagene). The amino acid substitutions were: H6A, H95A and Y108F. The primers used were: H6A: forward primer 5′-CAT ATG ATT CTC AGC GTA GCG ATT CTC GAT CAG CAA AC-3′; reverse primer 5′-GTT TGC TGA TCG AGA ATC GCT ACG CTG AGA ATC ATA TG-3′. H95A: forward primer 5′-CGA ATG AGC ACT ATG CGG TGC CGC TGT TAT TAA GTC-3′; reverse primer 5′-GAC TTA ATA ACA GCG GCA CCG CAT AGT GCT CAT TCG3′. Y108F: forward primer 5′-GTA TGG TTA TTC AAC CTT TCG CGG GAG CTA AGG ATC C-3′; reverse primer

Enzyme assays were performed using a modified method of Lee et al.8 Briefly, 50 μM uric acid was added to 0.05 units/ml uricase (Sigma) in 50 mM potassium phosphate buffer (pH 7.8). When the amount of 5-HIU had reached a maximum (approximately 3 min) 0.04 μM Sdu_TLP was added. Reactions were performed in 1 ml quartz cuvettes at 22 °C. The amount of 5-HIU was monitored at 312 nm throughout the experiment. The rate of 5-HIU hydrolysis was calculated with the extinction coefficient 8500 M−1 cm−1.7 The experiment was repeated using the Sdu_TLP mutants. Crystallisation Sdu_TLP was concentrated to 5–7 mg/ml; an equimolar amount of allantoin (50 mM in 1 M NaOH) was added and left to incubate at 4 °C for 1 h prior to crystallisation. Initial screening was carried out using a Cartesian Honeybee™ crystallisation robot (Genomic Solutions, USA). Optimization of crystallisation was then performed using the hanging-drop vapour-diffusion method. Hexagonal crystals were obtained in 0.1 M sodium cacodylate (pH 6.0), 0.2 M MgCl2, 20% (w/v) polyethylene glycol 3350 at 22 °C after three to five days. The crystals were flash-frozen in liquid N2 prior to data collection with 15% glycerol as the cryoprotectant. X-ray data collection, structure determination and refinement The crystals diffracted to 2.5 Å resolution and belong to space group P6, with unit cell dimensions of a = 94.70 Å, b = 94.70 Å, c = 57.38 Å, consistent with two molecules per asymmetric unit. The data were merged and processed using MOSFLM and SCALA.21,22 Subsequent crystallographic and structural analysis was performed using the CCP4i interface23 to the CCP4 suite,24 unless stated otherwise. Five percent of the dataset was flagged for calculation of the free R factor (Rfree) with neither a sigma, nor a low-resolution cut-off applied to the data. A summary of statistics is provided in Table 1. The structure was solved using the molecular replacement method and the program PHASER.25 A search model was constructed from an ensemble of the closest four structural homologues (pdb identifier 1GKO, 1IJN, 1SN0, 1SOK)16,26–28 identified using the FFAS server.29 Structures were structurally superimposed using MUSTANG,30 and a ‘‘mixed’’ model consisting of conserved side-chains (all other non alanine/glycine residues truncated at Cγ atom) was created using the SCRWL server.31 Two outstanding solutions having Z-scores of 8.9 and 15.9 were produced, and packed well within the unit cell. Together with the unbiased features in the initial electron density maps, the correctness of the molecular replacement solution was confirmed. Structure refinement and model building proceeded using the best matching model in the ensemble (1SN0). Maximum likelihood refinement using REFMAC,32 incorporating translation, libration, and screw-rotation displacement (TLS) refinement was carried out, employing a

Crystal Structure of Sdu_TLP

1398 bulk solvent correction (Babinet model with mask). Throughout refinement, tight NCS-restraints were imposed on the two molecules. All model building and structural validation was carried out using COOT.33 Water molecules were added to the model using ARP/wARP34 when the Rfree reached 30%. Solvent molecules were retained only if they had acceptable hydrogen bonding geometry contacts of 2.5–3.5 Å with protein atoms or with existing solvent, and were in good 2Fo–Fc and Fo–Fc electron density.

7. 8.

9. Structural analysis PYMOL35 was used to produce Figures 2, 4(a), 4(b), 4(d), 5, 6(a), and 6(b). CCP4MG24 was used to calculate electrostatic surfaces and to produce Figure 6(c) and (d). Structures were superimposed using the program MUSTANG.30 Accessible surface areas were calculated using the CCP4 program AREAIMOL.24

10.

11.

Protein Data Bank accession codes The coordinates have been deposited in the RCSB Protein Data Bank and are available under accession code 2GPZ.

12.

13.

Acknowledgements J.C.W. is a NHMRC Senior Research Fellow and Monash University Logan Fellow. This work was supported by the NHMRC, the Australian Research Council and the state government of Victoria (Australia). S.C.H. is Melbourne University Research Scholar. We thank Katya Ruzyla and Tanya A. Bashtannyk-Puhalovich for their technical assistance. We thank the staff at IMCA-CAT (The Advanced Photon Source, Chicago) for technical assistance.

References 1. Benson, M. D., Yazaki, M. & Magy, N. (2002). Laboratory assessment of transthyretin amyloidosis. Clin. Chem. Lab. Med. 40, 1262–1265. 2. Saraiva, M. J. (1995). Transthyretin mutations in health and disease. Hum. Mutat. 5, 191–196. 3. Westermark, P., Engstrom, U., Johnson, K. H., Westermark, G. T. & Betsholtz, C. (1990). Islet amyloid polypeptide: pinpointing amino acid residues linked to amyloid fibril formation. Proc. Natl Acad. Sci. USA, 87, 5036–5040. 4. Hennebry, S. C., Wright, H. M., Likic, V. & Richardson, S. J. The structural and functional evolution of transthyretin and transthyretin-like proteins. Proteins: Struct. Func. Bioinfo. in the press. 5. Prapunpoj, P., Yamauchi, K., Nishiyama, N., Richardson, S. J. & Schreiber, G. (2000). Evolution of structure, ontogeny of gene expression, and function of Xenopus laevis transthyretin. Am. J. Physiol. Regul. Integr. Comp. Physiol. 279, R2026–R2041. 6. Richardson, S. J. (2002). The evolution of transthyretin

14.

15.

16.

17. 18. 19. 20.

21. 22. 23. 24. 25.

synthesis in vertebrate liver, in primitive eukaryotes and in bacteria. Clin. Chem. Lab. Med. 40, 1191–1199. Kahn, K. & Tipton, P. A. (1998). Spectroscopic characterization of intermediates in the urate oxidase reaction. Biochemistry, 37, 11651–11659. Lee, Y., Lee do, H., Kho, C. W., Lee, A. Y., Jang, M., Cho, S. et al. (2005). Transthyretin-related proteins function to facilitate the hydrolysis of 5-hydroxyisourate, the end product of the uricase reaction. FEBS Letters, 579, 4769–4774. Schultz, A. C., Nygaard, P. & Saxild, H. H. (2001). Functional analysis of 14 genes that constitute the purine catabolic pathway in Bacillus subtilis and evidence for a novel regulon controlled by the PucR transcription activator. J. Bacteriol. 183, 3293–3302. Ramazzina, I., Folli, C., Secchi, A., Berni, R. & Percudani, R. (2006). Completing the uric acid degradation pathway through phylogenetic comparison of whole genomes. Nature Chem. Biol. 2, 144–148. Blake, C. C., Geisow, M. J., Oatley, S. J., Rerat, B. & Rerat, C. (1978). Structure of prealbumin: secondary, tertiary and quaternary interactions determined by Fourier refinement at 1.8 Å. J. Mol. Biol. 121, 339–356. Blake, C. C., Geisow, M. J., Swan, I. D., Rerat, C. & Rerat, B. (1974). Structure of human plasma prealbumin at 2–5 Å resolution. A preliminary report on the polypeptide chain conformation, quaternary structure and thyroxine binding. J. Mol. Biol. 88, 1–12. Hornberg, A., Eneqvist, T., Olofsson, A., Lundgren, E. & Sauer-Eriksson, A. E. (2000). A comparative analysis of 23 structures of the amyloidogenic protein transthyretin. J. Mol. Biol. 302, 649–669. Sunde, M., Richardson, S. J., Chang, L., Pettersson, T. M., Schreiber, G. & Blake, C. C. (1996). The crystal structure of transthyretin from chicken. Eur. J. Biochem. 236, 491–499. Muziol, T., Cody, V., Luft, J. R., Pangborn, W. & Wojtczak, A. (2001). Complex of rat transthyretin with tetraiodothyroacetic acid refined at 2.1 and 1.8 Å resolution. Acta Biochim. Pol. 48, 877–884. Eneqvist, T., Lundberg, E., Karlsson, A., Huang, S., Santos, C. R., Power, D. M. & Sauer-Eriksson, A. E. (2004). High resolution crystal structures of piscine transthyretin reveal different binding modes for triiodothyronine and thyroxine. J. Biol. Chem. 279, 26411–26416. Chothia, C. & Janin, J. (1975). Principles of proteinprotein recognition. Nature, 256, 705–708. Eneqvist, T., Lundberg, E., Nilsson, L., Abagyan, R. & Sauer-Eriksson, A. E. (2003). The transthyretin-related protein family. Eur. J. Biochem. 270, 518–532. Laemmli, U. K. & Favre, M. (1973). Maturation of the head of bacteriophage T4. I. DNA packaging events. J. Mol. Biol. 80, 575–599. Bradford, M. M. (1976). A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72, 248–254. Evans, P. (2006). Scaling and assessment of data quality. Acta Crystallog. sect. D, 72–82. Leslie, A. G. W. (1992). Joint CCP4+ ESF-EAMCB. Newsletter on Protein Crystallography, 26. Potterton, E., Briggs, P., Turkenburg, M. & Dodson, E. (2003). A graphical user interface to the CCP4 program suite. Acta Crystallog. sect. D, 59, 1131–1137. (1994). The CCP4 suite: programs for protein crystallography. Acta Crystallog. sect. D, 50, 760–763. McCoy, A. J., Storoni, L. C. & Read, R. J. (2004). Simple

Crystal Structure of Sdu_TLP

26.

27.

28.

29. 30.

algorithm for a maximum-likelihood SAD function. Acta Crystallog. sect. D, 60, 1220–1228. Hornberg, A., Olofsson, A., Eneqvist, T., Lundgren, E. & Sauer-Eriksson, A. E. (2004). The beta-strand D of transthyretin trapped in two discrete conformations. Biochim. Biophys. Acta, 1700, 93–104. Jiang, X., Smith, C. S., Petrassi, H. M., Hammarstrom, P., White, J. T., Sacchettini, J. C. & Kelly, J. W. (2001). An engineered transthyretin monomer that is nonamyloidogenic, unless it is partially denatured. Biochemistry, 40, 11442–11452. Karlsson, A., Olofsson, A., Eneqvist, T. & SauerEriksson, A. E. (2005). Cys114-linked dimers of transthyretin are compatible with amyloid formation. Biochemistry, 44, 13063–13070. Jaroszewski, L., Rychlewski, L., Li, Z., Li, W. & Godzik, A. (2005). FFAS03: a server for profile-profile sequence alignments. Nucl. Acids Res. 33, W284–W288. Konagurthu, A. S., Whisstock, J. C., Stuckey, P. J. & Lesk, A. M. P. S. (2006). MUSTANG: A multiple structural alignment algorithm. Proteins: Struct. Funct. Bioinformat. in the press.

1399 31. Canutescu, A. A., Shelenkov, A. A. & Dunbrack, R. L. (2003). A graph-theory algorithm for rapid protein side-chain prediction. Protein Sci. 12, 2001–2014. 32. Murshudov, G. N., Vagin, A. A. & Dodson, E. J. (1997). Refinement of macromolecular structures by the maximum-likelihood method. Acta Crystallog. sect. D, 53, 240–255. 33. Emsley, P. & Cowtan, K. (2004). Coot: model-building tools for molecular graphics. Acta Crystallog. sect. D, 60, 2126–2132. 34. Morris, R. J., Perrakis, A. & Lamzin, V. S. (2003). ARP/wARP and automatic interpretation of protein electron density maps. Methods Enzymol. 374, 229–244. 35. DeLano, W. L. (2002). The PyMOL UserTs Manual. Delano Scientific, San Carlos, CA, USA. 36. Thompson, J. D., Higgins, D. G. & Gibson, T. J. (1994). CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucl. Acids Res. 22, 4673–4680.

Edited by M. Guss (Received 17 March 2006; received in revised form 21 April 2006; accepted 22 April 2006) Available online 11 May 2006