Volume 584, Issue 8, Pages 1455-1462 (16 April 2010)

5 of 34

ABSTRACT

FULL TEXT

FULL-TEXT PDF (533 KB)

Crystal structure of the HIV-1 integrase core domain in complex with sucrose reveals details of an allosteric inhibitory binding site

Edited by Hans Eklund

Received 8 January 2010; received in revised form 5 March 2010; accepted 9 March 2010. published online 12 March 2010.

Abstract

HIV integrase (IN) is an essential enzyme in HIV replication and an important target for drug design. IN has been shown to interact with a number of cellular and viral proteins during the integration process. Disruption of these important interactions could provide a mechanism for allosteric inhibition of IN. We present the highest resolution crystal structure of the IN core domain to date. We also present a crystal structure of the IN core domain in complex with sucrose which is bound at the dimer interface in a region that has previously been reported to bind integrase inhibitors.

Structured summary

MINT-7713125: IN (uniprotkb:P04585) and IN (uniprotkb:P04585) bind (MI:0407) by X-ray crystallography (MI:0114)

Keywords: Anti-HIV drug, Ligand binding, HIV-1 integrase, X-ray crystallography

Article Outline

• Abstract

• 1. Introduction

• 2. Materials and methods

• 2.1. Mutagenesis

• 2.2. IN expression and purification for crystallization

• 2.3. Crystallization

• 2.4. Data collection, structure determination and refinement

• 3. Results and discussion

• 3.1. Expression, purification, crystallization and data collection

• 3.2. Structure of IN

• 3.3. Structure of IN

• 4. Conclusions

• Acknowledgment

• References

• Copyright

1. Introduction

HIV integrase (IN) is one of the three enzymes encoded in the HIV genome. IN mediates the retroviral process of integration–insertion of the proviral DNA into that of the host cell [1]. IN acts minimally as a pair of homodimers [2] and catalyzes two reactions: 3′-processing, where two bases are removed from the blunt-ended proviral DNA, and strand transfer, where each end of the proviral DNA is covalently bonded to the host DNA [3]. The essential nature of integration in HIV replication makes IN a prime target for anti-HIV drug discovery.

After a long period of development, the first IN inhibitor raltegravir [4] is being successfully applied in the clinic [5], [6]. Raltegravir is believed to act through interaction with the IN active site and a coordinated metal ion, located in the core domain, and the bound viral DNA [7], [8]. A common problem with drugs targeting HIV is the rapid development of mutations in the HIV genome which diminish their efficacy. Already such mutations have arisen for IN that confer resistance to raltegravir. Moreover many of these mutations show cross resistance profiles to elvitegravir, a second IN inhibitor in phase III clinical trials [8]. Taken together this suggests that modulation of IN at sites other than the active site would be desirable as this would lessen the possibility of cross resistance. It is becoming increasingly clear that during the process of viral replication IN also makes functional interactions with a number of viral and host proteins. Targeting these interactions raises the prospect of additional approaches to the treatment of HIV. This proposition for IN has been discussed by a number of authors [9], [10], [11], [12], [13], [14] and the approach of developing drugs that do not target the active site of an enzyme has already proven successful for drugs targeting HIV reverse transcriptase [15].

The crystal structures of truncated IN proteins (INCORE [16], [17], [18], [19], [20], [21], [22], [23], [24], [25], [26], INCORE+C [19], [27], [28], INCORE+N [29], [30]) have been determined but no structure of the full length enzyme or IN with viral DNA is available. NMR structures of the isolated N-terminal [31], [32], [33] or C-terminal [34], [35] IN domains have also been reported. Only a small number of ligands have been observed by crystallography to bind to IN and unfortunately, in all cases, the observed ligand orientations are perturbed by crystal contacts with either a neighboring protomer or a crystallographically related ligand molecule. The best known IN complex contains the diketo-acid inhibitor, 1-(5-chloroindol-3-yl)-3-hydroxy-3-(2H-tetrazol-5-yl)-propenone (5CITEP), which is present in the IN active site [22], although the bound orientation is strongly perturbed by interaction with a second ligand molecule. Two tetraphenyl arsonium compounds have also been reported to interact with the lens epithelium-derived growth factor (LEDGF) binding site, although these ligands also make contact with adjacent protomers in the crystal lattice [24]. The naphthalenedisulfonic acid Y3, an inhibitor of strand transfer and 3′-processing has been observed to bind to the closely related avian sarcoma virus IN (ASV IN) near residues 145–152 (equivalent to residues 140–147 in HIV-1 IN) which form a flexible surface loop adjacent to the active site [26]. In the ASV IN-Y3 complex two symmetry-related ligands bind in close proximity to each other and may not represent the biologically bound orientation of the molecule. The absence of unperturbed ligand complexes is a significant impediment to the development of drugs that target this enzyme.

The best described IN–protein interaction for IN is with the human cellular protein, LEDGF [25], [29], [36], [37], [38], [39], [40], [41] which has been characterized biochemically and by X-ray crystallography. LEDGF binds to IN with the principal contact being a pocket at the IN dimer interface. The IN binding domain (IBD) of LEDGF has been crystallized bound to the INCORE [25] and more recently to both a two domain fragment of HIV-2 INCORE+N [29] and maedi-visna virus (MVV) INCORE+N [42] where the IBD domain makes additional contacts with the N-terminal domain of IN. A variety of other cellular binding partners for IN have been identified, including transportin 3 (TNP03) [43], embryonic ectoderm development (EED) protein [44], [45], SNF 5 [46], [47], [48] and the C-terminal domain of nucleoporin 153 [49]. IN has also been shown to directly interact with HIV reverse transcriptase [50], [51], [52] and more specifically Wilkinson et al. [50] mapped the residues involved in this interaction to the C-terminal domain of IN using a series of heteronuclear single quantum coherence (HSQC) NMR experiments. At present, apart from LEDGF and reverse transcriptase, the nature of the IN–protein interactions have not been characterized in detail.

In spite of this, several sites on IN have been identified as potential mediators of protein–protein interactions including the LEDGF binding site [53], the C-terminal domain [50] and a putative nuclear localization sequence between residues 161–173 [54]. One potential allosteric binding site is a pocket formed at the dimer interface near amino acid residue K173. This pocket has been identified as a site of inhibitor binding by several research groups [12], [13], [14]. Shkriabai et al. [14] used affinity acetylation and mass spectrometry to identify this pocket as the binding site of a chicoric acid analogue and more recently showed that the compound modulates dynamic interactions between IN subunits in a dose dependent manner [13]. Du et al. [12] identified a bis-1-pyrrolidineacetamide compound that competitively inhibited IN binding to viral DNA. Molecular docking was used to propose a binding orientation, which was corroborated by site-directed mutation of residues in the proposed binding pocket.

In this work, we describe the characterization of a reported binding site in the HIV-1 IN core domain containing a bound molecule of sucrose and present a crystal structure of the HIV-1 IN core domain to a higher resolution than previously published. The sucrose binding site represents a potential target for the development of allosteric inhibitors of HIV replication.

2. Materials and methods

2.1. Mutagenesis

A truncated HIV-1 IN50-212 construct of similar sequence to the NL43 strain was cloned into the Nde1 and Xho1 restriction sites of pET28b+ (Novagen). This construct expressed in the IN50-212 domain containing an N-terminal 6-histidine (H6) tag followed by a thrombin cleavage site. The solubilizing mutations C56S, W131D, F139D and F185H [19] were introduced into the IN50-212 sequence and this clone, H6-IN50-212, was used for protein expression and purification. These point mutations significantly improve solubility whilst retaining integrase activity in vitro ([19] and unpublished results). In addition HIV-1NL43 containing the F185H [55] or C56S [56] mutations in IN exhibited similar replication profiles to wild type HIV-1NL43 in cells.

2.2. IN expression and purification for crystallization

Escherichia coli containing the H6-IN50-212 construct was cultured in Terrific Broth media containing 30μgml−1 kanamycin. When OD600 reached 0.6 expression was induced using 0.5mM IPTG overnight at room temperature. The cells were harvested by centrifugation for 20min at 6000×g and pellets were resuspended in 50% Bugbuster (Novagen) and 50% Buffer A (25mM HEPES pH 8.0, 0.5M sodium chloride, 5mM imidazole, 5mM dithiothreitol (DTT), 10% (v/v) glycerol). The cells were incubated on ice for 45min before debris was removed by centrifugation at 30000×g for 30min. The supernatant was filtered through a 0.45μm Millipore filter and loaded onto a nickel IMAC sepharose column (GE Healthcare) column. H6-IN50-212 was eluted using an imidazole gradient against a solution containing 25mM HEPES pH 8.0, 0.5M sodium chloride, 1M imidazole, 10% (v/v) glycerol and 1mM DTT. Fractions containing H6-IN50-212 were pooled and dialyzed overnight against 25mM HEPES pH 8.0, 0.5M sodium chloride, 5mM imidazole, 10% (v/v) glycerol and 1mM DTT. The N-terminal His6 tag was removed by adding calcium chloride to 2mM followed by the addition of thrombin (Sigma) and incubation on ice. After 90% of the protein was cleaved (monitored by SDS gel electrophoresis) the protein sample was passed over a benzamidine-sepharose column (GE Healthcare). The flow through was collected and then passed over another Ni2+ IMAC column and the flow through containing IN50-212 lacking the N-terminal His6 tag was collected. This protein was buffer exchanged into 25mM HEPES pH 6.5, 0.5M sodium chloride and 5mM DTT and concentrated to 7–10mgml−1 using Vivaspin centrifugal concentrators (Sartorius) with a 5000 MWCO and then used for crystallization.

2.3. Crystallization

Briefly, 2μl of 8mgml−1 IN50-212 protein solution was mixed with an equal volume of reservoir solution (1.8M ammonium sulfate, 0.15M sodium citrate pH 4.6 and 5mM cadmium chloride). Crystallization was performed by the hanging drop vapor diffusion method [57] at 293K. Crystals appeared overnight and grew to typical dimensions of approximately 0.3mm×0.2mm×0.1mm.

2.4. Data collection, structure determination and refinement

Prior to freezing in liquid nitrogen, crystals were soaked in a solution containing mother liquor and either 20% (v/v) glycerol (INGOL) or 25% (w/v) sucrose (INSUC) for several minutes. In the case of INGOL, cadmium chloride was omitted from the cryoprotectant solution. Diffraction data for the IN50-212 crystals were collected at 100K on the 14 BMC beamline or 23 ID-D beamline at the Advanced Photon Source (Chicago, USA). Data were indexed using MOSFLM [58] and scaling was performed using SCALA [59] in the CCP4 interface [60]. Initial phases were determined using AMoRe [61] with the unliganded HIV-1 core domain (PDB ID: 1EXQ [19]) as the search model. Model building and refinement were completed using Coot [62], Refmac5 [60] and Phenix [63]. The INGOL structure was refined using anisotropic temperature factors. The final statistics of refinement are summarized in Table 1. The models were analyzed with the program PROCHECK [64] which showed that their stereochemical quality was similar or better than expected for structures refined at similar resolution. An omit map was calculated by performing a round of simulated annealing in the absence of the sucrose molecules using Phenix [63] and then superimposing the resultant Fo–Fc map over the final INSUC model. Structure alignments were performed using the align command in PyMOL [65]. Each protomer of IN was superimposed over chain A of INGOL using either Cα atoms or all atoms. The coordinates of INGOL and INSUC have been deposited in the Protein Data Bank [66] with accession numbers 3L3U and 3L3V.

Table 1.

Summary of X-ray data processing and refinement statistics.


Data collection	INGOL	INSUC
Wavelength (Å)	0.90	1.00
Temperature (K)	100	100
Maximum resolution (Å)	1.4	2.0
Space group	P212121	P212121
Unit-cell (Å)	a=59.6, b=62.4, c=81.1	a=60.1, b=60.7, c=81.5
No. of measured reflections	754279 (115695)	149441 (21613)
No. of unique reflections	60131 (8679)	20793 (2979)
Completeness (%)	99.9 (100.0)	100.0 (100.0)
Rmerge (%)a	11.2 (43.9)	9.0 (41.0)
Mean I/σ(I)	15.2 (6.6)	12.7 (4.4)
Multiplicity	12.5 (13.3)	7.2 (7.3)
Wilson 〈B〉 (Å2)	11.1	24.2

Refinement
Maximum resolution (Å)	1.4 (1.48–1.4)	2.0 (2.1–2.0)
No. reflections used	56999	20741
No. reflections used for Rfree	3064	1064
Rfree (%)	23.6	24.6
Rfactor (%)	20.3	19.8
No. protein atoms	2492	2332
No. sucrose atoms	–	46
No. sulfate ions	3	2
No. cadmium ions	–	4
No. water molecules	335	84
R.M.S.D. bond angles (°)	2.1	1.1
R.M.S.D. bond lengths (Å)	0.02	0.01

Average B-factors (Å2)
Protein atoms	10.3	35.3
Sucrose conformation 1/2	–	36.4/32.3
Water	23.1	36.2

Ramachandran distribution
Most favoured (%)	99.3	98.6
Allowed (%)	0.7	1.4

Values in parentheses are for the highest resolution shell.

3. Results and discussion

3.1. Expression, purification, crystallization and data collection

The core domain of HIV-1 IN containing solubilizing mutations C56S, W131D, F139D and F185H (IN50-212) was expressed and purified using methods similar to those described by Chen et al. [19] Crystals were readily obtained at 295K using ammonium sulfate as a precipitant. Crystals were cryoprotected using either a 20% (v/v) glycerol (INGOL) or a 25% (w/v) sucrose solution (INSUC) prior to freezing in liquid nitrogen. Both INGOL and INSUC crystallized in the P212121 space group. INGOL diffracted beyond 1.4Å resolution on the 14BM-C beamline and INSUC diffracted beyond 2.0Å resolution on the 23ID-D beamline at the Advanced Photon Source (APS, Chicago, USA). The structures were solved by molecular replacement and the final models were refined to Rfactor and Rfree values of 20.3% and 23.6% respectively for INGOL and 19.8% and 24.6% respectively for INSUC. Data collection, refinement and Ramachandran statistics are shown in Table 1. For both structures, the asymmetric unit contains two copies of the IN core domain arranged as a dimer which is accepted as the minimal biologically active unit. Each INCORE comprises a central five-strand beta sheet flanked by six helices with an RNaseH-like fold and is consistent with other reported crystal structures of this domain [17], [19], [20], [21], [22], [23], [24].

3.2. Structure of INGOL

The higher resolution of the INGOL structure (1.4Å) compared with other published structures (1.6–2.6Å) provides new insights into the IN core domain. Twenty seven side-chains were modeled with two or more alternate conformations (L63, I84, E85, L101, L102, R107, V113, N144, E157, I182, I191, R199, V201, I203, and I204 in chain A and L63, L74, V77, K103, K111, V113, S119, K127, K136, S153, I161 and V165 in chain B) and a 2–3 fold increase in the number of crystallographic water molecules were observed (∼160/protomer compared to ∼50–100/protomer for other published IN structures). Of the alternate conformations the hydrophobic residues (L63, L74, V77, I84, L101, L102, V113, I161, I182, I191, V201, I203, and I204) are mostly buried and involved in Van der Waals contacts. R107, E85 and K103 are located at the dimer interface and are involved in contacts between protomers. K111, K127, K136 and N144 are solvent exposed and make contacts with a neighboring IN molecule in the crystal lattice. The side-chain of E157 is either directed towards the solvent or makes a hydrogen bond with the H183 side-chain. A comparison with published IN crystal structures showed that the overall fold of the core domain was conserved (R.M.S.D. of 0.6Å2 (range 0.4–1.0Å2) using ∼113 Cα atoms or average R.M.S.D. of 0.8Å2 (range 0.6–1.0 Å2) using all atoms (∼820 atoms)).

A flexible loop (residues 140–149) flanking the catalytic D, D, E motif is traceable in chain A (Fig. 1). This loop is typically disordered or contacting residues from a neighboring protomer in the crystal lattice in other published IN structures. Contacts between this loop and neighboring IN molecules in the crystal lattice were inspected. Only residues 140–144 were found to be involved in crystal contacts. Residues in this loop are highly conserved and are important for IN function [67]. In chain A of the INGOL structure, the mobile loop is in an open conformation and folds back on the protein. Interestingly neither Q148 nor E152 are located in the active site region, instead these important residues are both pointing away from this region and towards the solvent. This conformation may represent the orientation of IN prior to viral DNA binding. In contrast residues 141–148 could not be modeled into the electron density for chain B. Despite several attempts to get suitable diffracting crystals in the presence of Mg or Mn ions, a metal ion coordinating to D64 and D116 was absent from the structure; rather D64 is orientated to form a hydrogen bond with N155 which in turn causes a small distortion in the β2 strand. D116 interacts with the backbone nitrogen of G118 and several crystallographic water molecules.

View Large Image
Download to PowerPoint

Fig. 1. Crystal structure of INGOL solved to 1.4Å. (A) INGOL shown in cartoon with protomers colored green and blue respectively and residues 140–149 in chain A colored orange. Selected residues in the active site and the flanking loop are shown in stick. (B) A Connolly surface [77] applied over residues in INGOL with the exception of residues 140–149 which are shown in stick. Crystallographic waters are shown as spheres. Hydrogen bonds between the loop, crystallographic waters and protein are shown as dashed lines.

3.3. Structure of INSUC

For the INSUC structure, a sucrose molecule from the cryoprotectant was identified at the interface between the two IN monomers comprising residues E87, V88, I89, P90, E96, Y99, F100, K103 and K173 of chain A and the same residues in chain B (Fig. 2A). The sucrose binding pocket is flanked by the two LEDGF binding pockets in the IN dimer and is adjacent to the active site D, D, E residues (Fig. 2B). Due to the symmetry of the binding site, sucrose is found in two overlapping, alternate orientations which were each modeled to bind at 50% occupancy based on difference maps. An omit map calculated by completing a round of simulated annealing in the absence of the sucrose molecules was used to confirm the final model (Fig. 2C). Each of the sucrose conformations makes a network of hydrogen bonds with the protein directly, through interactions with the side-chain atoms of residues E96, K103 and K173 and the backbone of V88 from both monomers, and indirectly through the crystallographically observed water molecules; HOH 11, 25, 29, 62, 66, 82 and 84 (shown in Fig. 2D, E and described in Table 2). Water molecules HOH 11 and HOH 25, and HOH 62 and HOH 82 occupy equivalent positions in each monomer whereas the other water molecules in the binding site (HOH 29, HOH 66 and HOH 84) do not. These latter water molecules have high B-factors. There was weak evidence for symmetry equivalent waters for HOH 29, HOH 66 and HOH 84 at low sigma levels in the 2Fo–Fc map; however, after several test model building and refinement cycles electron density around the waters did not improve and B-factors remained high. For this reason these water molecules were not included in the final model even at lower levels of occupancy (<50%). HOH 66 and HOH 84 were modeled to have 50% occupancy which reduced the B-factors of these molecules from ∼50Å2 to a level more consistent with the surrounding environment (∼30Å2). From a comparison of this pocket to other published IN crystal structures, we found that the sucrose binding site typically contains a network of water molecules (18 in the INGOL structure) which are displaced upon sucrose binding, with the exception of HOH 11 and HOH 25, and to a lesser extent HOH 62 and HOH 82 which tend to be conserved (data not shown). These four water molecules are stabilized through multiple hydrogen bonds with protein residues in the pocket in the absence of sucrose but also have a significant role in sucrose binding.

View Large Image
Download to PowerPoint

Fig. 2. Sucrose bound to the HIV-1 IN core domain. (A) Location of the sucrose conformers at the IN dimer interface. IN monomers are colored grey and yellow. (B) Relative location of the sucrose binding pocket to the D, D, E active site residues (red) and residues involved in binding LEDGF (orange). Arrows indicate the channels between the LEDGF and sucrose binding pockets. (C) Stereo diagram of Fo–Fc omit map at 3σ around the two alternate sucrose orientations (green). (D and E) Stereo diagram of the hydrogen bonding network between alternate sucrose conformations and IN (blue or grey dashed lines). Residues within 5Å of the sucrose are shown in stick. Crystallographic water molecules involved in the interactions are shown as spheres.

Table 2.

Direct and solvent-bridged hydrogen bonding interactions between sucrose oxygen atoms and the protein binding site.


Sucrose atom	Sucrose alternative conformation 1		Sucrose alternative conformation 2
Sucrose atom	H-bond partner	Water mediated H-bond	H-bond partner	Water mediated H-bond
O6′	O V88 (B) (2.8Å)		O V88 (A) (2.8Å)
	HOH 82 (2.6Å)	OE1 E87 (B) (2.4Å)	HOH 62 (2.7Å)	OE1 E87 (A) (2.4Å)
		OH Y99 (A) (2.4Å)		OH Y99 (B) (2.5Å)
		N V88 (B) (3.2Å)		N V88 (A) (3.0Å)
		O V88 (B) (3.6Å)		O V88 (A) (3.4Å)
O4′	OE1 E96 (A) (3.5Å)		OE1 E96 (B) (3.5Å)
	OE2 E96 (A) (3.4Å)		OE2 E96 (B) (3.8Å)
	NZ K173 (B) (3.7Å)		NZ K173 (A) (2.9Å)
	HOH 66 (2.9Å)	NZ K173 (B) (2.8Å)	HOH 84 (3.2Å)	OE2 E96 (B) (2.9Å)
		OE2 E96 (A) (3.1Å)
O3′	OE1 E96 (A) (3.1Å)		OE1 E96 (B) (3.09Å)
	HOH 29 (2.5Å)
O1′	NZ K103 (A) (2.5Å)		NZ K103 (B) (2.5Å)
	HOH 11 (3.3Å)	OE1 E96 (A) (2.6Å)	HOH 25 (3.5Å)	OE1 E96 (B) (2.7Å)
		O E96 (A) (2.7Å)		O E96 (B) (2.8Å)
O1	HOH 29 (3.5Å)	O3′ SUC (2.5Å)	HOH 29 (3.5Å)	O3′ SUC (2.5Å)
		O2 SUC (2.9Å)		O2 SUC (2.9Å)
O2	O V88 (B) (3.7Å)		O V88 (B) (3.5Å)
	NZ K173 (A) (2.9Å)
	HOH 29 (2.9Å)	O3′ SUC (2.5Å)
		O1 SUC (3.4Å)
O3	NZ K173 (A) (2.6Å)
	HOH 84 (2.7Å)	OE2 E96 (B) (2.8)
O4	OE1 E96 (B) (2.8Å)		OE1 E96 (A) (2.6Å)
	OE2 E96 (B) (3.5Å)		OE2 E96 (A) (3.3Å)
			HOH 29 (2.9Å)
O5		NZ K103 (B) (3.5Å)		NZ K103 (A) (3.3Å)
O6	NZ K103 (B) (3.3Å)		NZ K103 (A) (3.2Å)
	HOH 25 (2.7Å)	OE1 E96 (B) (2.7Å)	HOH 11 (2.7Å)	OE1 E96 (A) (2.6Å)
		O E96 (B) (2.8Å)		O E96 (A) (2.7Å)

Interactions between the sucrose and IN were dominated by hydrogen bonding with 18 hydrogen bonds between the protein and crystallographic water molecules and Van der Waals contacts with Y99, K173 and E87. The average B-factors of atoms in the sucrose conformers (36 and 32Å2 respectively) were consistent with surrounding residues (32Å2). The position of residues in the sucrose binding pocket was compared to INGOL and with the exception of the solvent exposed K173 side-chains which move to form hydrogen bonds with the sucrose molecule there were no other significant changes observed in residue positions.

Of the residues that make up the sucrose binding pocket, E87, V88 and I89 are invariant and P90, E96, Y99, F100, K103 and K173 share greater then 98% sequence similarity in samples isolated from patients who are naive to IN inhibitors [67], [68] but are more variable in a sequence alignment of the more distantly related lentiviral INs (data not shown). Comparison of simian immunodeficiency virus (SIV [27] and HIV-2 IN [29] crystal structures with HIV-1 IN show that the amino acids and orientation of the residues in the site are conserved with the exception of E96 which is a glutamine, Y99 which is a leucine and K173 which is a glutamate in the HIV-2 and SIV structures. The site has a similar shape but residues are less conserved in the Rous sarcoma virus (RSV) [28] and Maedi-visna virus (MVV) [42] IN structures. The equivalent sucrose binding pocket in the RSV and MVV IN structures both contain more polar residues.

Attempts to determine an IC50 for integration in an assay similar to that described by Ovenden et al. [69] showed that sucrose exhibited no inhibition activity at the maximum tested concentration of 10mM. Sucrose also failed to show affinity for IN50-212 in saturation transfer difference (STD) and HSQC NMR experiments at a concentration of up to 100mM (data not shown). The concentration of sucrose in the cryoprotectant drop was ∼700mM. We propose that sucrose is a specific, low affinity ligand with an affinity greater than 100mM. The lack of inhibition and affinity was surprising considering the large number of hydrogen bonds that sucrose makes with IN.

To try and understand why no binding affinity was observed we compared the conformation of sucrose in our structure to that of the solid state sucrose structure [70] and to other protein–sucrose complexes in the Protein Databank [66] to see if it was in a high energy state. In the solid state structure [70], the glucosyl and fructosyl rings of the sucrose adopt a conformation where the bulkier polar groups are positioned equatorially which is consistent with our results. Two internal hydrogen bonds between the fructosyl-6-OH (donor) and glucosyl-2-OH (acceptor) and the fructosyl-6-OH and glucosyl-O were observed that were not present in our sucrose conformation however these are replaced by water-mediated hydrogen bonds in an aqueous environment [71], [72] which is consistent with our results. Other polar groups in the solid state structure adopt a conformation to maximize interactions with neighboring molecules in the crystal lattice. Examples of protein crystal structures containing sucrose included amylosucrase (PDB ID: 1MW1, [73]), DNA polymerase I (PDB ID: 3EYZ, unpublished results), beta lactamase (PDB ID: 1LOG, [74]) and the HIV-1 envelope protein GP120 (PDB ID: 2NXZ, [75]). In each case the conformation of the sucrose rings was consistent with our structure and that of the solid state structure. In addition the torsion angle between the two sugar rings was consistent in all of the structures.

In the protein–sucrose complexes the sucrose binding site tended to be hydrophilic and extensive hydrogen bonding networks were observed between the sucrose and protein residues. In general, the sucrose rings did not interact directly with protein residues with the exception of the orange carotenoid protein (PDB ID: 1M98 [76]). These sites tended to contain a large number of water molecules which helped to further stabilize binding though water-mediated hydrogen bonds reminiscent of what was observed for INSUC. Hydrophobic interactions were minimal. In summary there was no evidence that the conformation of sucrose observed in our structure was in a high energy state. Rather the reason for the lack of observed binding affinity is most likely due to the high energy required to desolvate the sucrose polar groups, i.e. desolvation of the hydroxyl groups will incur an enthalpic penalty that is not compensated for by the formation of favorable bonds with the protein.

We next asked the question what is the importance of this pocket and could it potentially be a site for the development of allosteric IN inhibitors. The sucrose binding pocket is located about 10Å from the LEDGF binding site. It has been observed that the IN mutant L172A/K173A fails to interact with LEDGF even though the LEDGF-IN crystals did not show contact beyond H171 [25], suggesting some interaction between the pockets. Applying a Connolly surface [77] over the protein reveals a channel between the sucrose and LEDGF binding pockets (Fig. 2B, arrows). This channel is positioned over Y99 and is flanked by residues E96, Q95, K173 and H171. It could be exploited for the development of inhibitors targeting LEDGF and provide a method for allosteric inhibition of IN.

The sucrose binding pocket has been previously identified by two research groups as a binding site for inhibitors of IN. Du et al. [12] identified the symmetrical compound 1-pyrrolidineacetamide, N,N′-(methylene-di-4,1-phenylene) bis-1-pyrrolidineacetamide as an IN ligand using surface plasmon resonance (SPR). The compound was found to competitively inhibit DNA binding to IN50-212 and full length IN (residues 1–288) and to inhibit HIV-1 IIIB replication with an EC50 of 41μM in C8166 cells. The binding site of the inhibitor was located using mutagenesis and found to involve residues K103, K173 and T174. Shkriabai et al. [14] also identified this region as the binding site of a largely symmetrical bis-caffeoyl derivative, methyl N, O-bis(3,4-diacetoxycinnamoyl)-d-serinate. This compound exhibited an IC50 of ∼3μM in both an in vitro 3′-processing and strand transfer IN assay [78]. Specifically they used mass spectrometry to identify the transfer of an acetyl group from the compound to K173. The current structure of INSUC provides a description of protein–ligand interactions in a pocket that has been identified as a binding site for IN inhibitors. These details provide a means to design new inhibitors that are capable of exploiting some of these interactions and therefore build a lead with greater affinity.

4. Conclusions

We present two crystal structures of the IN core domain; the first, INGOL, is refined to a higher resolution than previously described and shows a unique orientation of the functionally important flexible loop flanking the active site. The second, INSUC, has a sucrose molecule bound in a symmetrical binding site at the dimer interface comprising residues E87, V88, I89, P90, E96, Y99, F100, K103 and K173 from both monomer chains and characterizes a previously reported inhibitory binding pocket. This site is one of a number clefts on the surface of IN and may represent a functional binding pocket for protein–protein recognition. Although the affinity of the sucrose is low, other inhibitors of HIV replication have been identified that bind in this region with higher affinity. The large number of interactions between sucrose and IN could be used to develop compounds with better binding affinity.

Acknowledgements

We thank Nick Vandegraaff for testing sucrose in the IN assay and Susanne Feil and Mike Gorman for support and advice. We also thank the BioCARS and GM/CA staff for their help at the Advanced Photon Source. This work was supported by the Australian Research Council Linkage project (LP0775192) and the Australian Synchrotron Research Program, which was funded by the Commonwealth of Australia under the Major National Research Facilities Program. Use of the Advanced Photon Source was supported by the U.S. DOE, Basic Energy Sciences and Office of Energy Research. M.W.P. is an ARC Federation Fellow and National Health and Medical Research Council Honorary Fellow.

References

[1]. [1]Brown PO. Integration. In: Coffin JM, Hughes SH, Varmus HE editor. Retroviruses. Plainview, NY: Cold Spring harbour Press; 1997;p. 161–201.

[2]. [2]Guiot E, et al. Relationship between the oligomeric status of HIV-1 integrase on DNA and enzymatic activity. J. Biol. Chem. 2006;281:22707–22719. MEDLINE | CrossRef

[3]. [3]Vandegraaff N, Engelman A. Molecular mechanisms of HIV integration and therapeutic intervention. Expert Rev. Mol. Med. 2007;9:1–19. MEDLINE

[4]. [4]Summa V, et al. Discovery of raltegravir, a potent, selective orally bioavailable HIV-integrase inhibitor for the treatment of HIV-AIDS infection. J. Med. Chem. 2008;51:5843–5855. CrossRef

[5]. [5]Chirch LM, Morrison S, Steigbigel RT. Treatment of HIV infection with raltegravir. Expert Opin. Pharmacother. 2009;10:1203–1211. CrossRef

[6]. [6]Imaz A, Del Saz SV, Ribas MA, Curran A, Caballero E, Falco V, et al. Raltegravir, Etravirine, and Ritonavir-Boosted Darunavir: A Safe and Successful Rescue Regimen for Multidrug-Resistant HIV-1 Infection. J Acquir Immune Defic Syndr. 2009;52:382–386. CrossRef

[7]. [7]Garvey EP, Schwartz B, Gartland MJ, Lang S, Halsey W, Sathe G, et al. Potent inhibitors of HIV-1 integrase display a two-step, slow-binding inhibition mechanism which is absent in a drug-resistant T66I/M154I mutant. Biochemistry. 2009;48:1644–1653.

[8]. [8]Marinello J, Marchand C, Mott BT, Bain A, Thomas CJ, Pommier Y. Comparison of raltegravir and elvitegravir on HIV-1 integrase catalytic reactions and on a series of drug-resistant integrase mutants. Biochemistry. 2008;47:9345–9354.

[9]. [9]Walker MA. New approaches for inhibiting HIV integrase: a journey beyond the active site. Curr. Opin. Investig. Drugs. 2009;10:129–136.

[10]. [10]Al-Mawsawi LQ, Fikkert V, Dayam R, Witvrouw M, Burke T, Borchers CH, et al. Discovery of a small-molecule HIV-1 integrase inhibitor-binding site. Proc. Natl. Acad. Sci. USA. 2006;103:10080–10085. MEDLINE | CrossRef

[11]. [11]Busschots K, De Rijck J, Christ F, Debyser Z. In search of small molecules blocking interactions between HIV proteins and intracellular cofactors. Mol. Biosyst. 2009;5:21–31. CrossRef

[12]. [12]Du L, Zhao YX, Yang LM, Zheng YT, Tang Y, Shen X, et al. Symmetrical 1-pyrrolidineacetamide showing anti-HIV activity through a new binding site on HIV-1 integrase. Acta Pharmacol. Sin. 2008;29:1261–1267. CrossRef

[13]. [13]Kessl JJ, Eidahl JO, Shkriabai N, Zhao Z, McKee CJ, Hess S, et al. An allosteric mechanism for inhibiting HIV-1 integrase with a small molecule. Mol. Pharmacol. 2009;76:824–832. CrossRef

[14]. [14]Shkriabai N, Patil SS, Hess S, Budihas SR, Craigie R, Burke TR, et al. Identification of an inhibitor-binding site to HIV-1 integrase with affinity acetylation and mass spectrometry. Proc. Natl. Acad. Sci. USA. 2004;101:6894–6899. MEDLINE | CrossRef

[15]. [15]de Bethune MP. Non-nucleoside reverse transcriptase inhibitors (NNRTIs), their discovery, development, and use in the treatment of HIV-1 infection: a review of the last 20 years (1989–2009). Antiviral Res. 2010;85:75–90. CrossRef

[16]. [16]Dyda F, Hickman AB, Jenkins TM, Engelman A, Craigie R, Davies DR. Crystal structure of the catalytic domain of HIV-1 integrase – similarity to other polynucleotidyl transferases. Science (Washington, D.C.). 1994;266:1981–1986.

[17]. [17]Bujacz G, Alexandratos J, Qing ZL, Clement-Mella C, Wlodawer A. The catalytic domain of human immunodeficiency virus integrase: ordered active site in the F185H mutant. FEBS Lett. 1996;398:175–178. Abstract | Full-Text PDF (380 KB) | CrossRef

[18]. [18]Bujacz G, Jaskolski M, Alexandratos J, Wlodawer A, Merkel G, Katz RA, et al. The catalytic domain of avian sarcoma virus integrase: conformation of the active-site residues in the presence of divalent cations. Structure. 1996;4:89–96. MEDLINE | CrossRef

[19]. [19]Chen JC, Krucinski J, Miercke LJ, Finer-Moore JS, Tang AH, Leavitt AD, et al. Crystal structure of the HIV-1 integrase catalytic core and C-terminal domains: a model for viral DNA binding. Proc. Natl. Acad. Sci. USA. 2000;97:8233–8238. MEDLINE | CrossRef

[20]. [20]Goldgur Y, Dyda F, Hickman AB, Jenkins TM, Craigie R, Davies DR. Three new structures of the core domain of HIV-1 integrase: an active site that binds magnesium. Proc. Natl. Acad. Sci. USA. 1998;95:9150–9154. MEDLINE | CrossRef

[21]. [21]Maignan S, Guilloteau JP, Zhou-Liu Q, Clement-Mella C, Mikol V. Crystal structures of the catalytic domain of HIV-1 integrase free and complexed with its metal cofactor: high level of similarity of the active site with other viral integrases. J. Mol. Biol. 1998;282:359–368. MEDLINE | CrossRef

[22]. [22]Goldgur Y, et al. Structure of the HIV-1 integrase catalytic domain complexed with an inhibitor: a platform for antiviral drug design. Proc. Natl. Acad. Sci. USA. 1999;96:13040–13043. MEDLINE | CrossRef

[23]. [23]Greenwald J, Le V, Butler SL, Bushman FD, Choe S. The mobility of an HIV-1 integrase active site loop is correlated with catalytic activity. Biochemistry. 1999;38:8892–8898.

[24]. [24]Molteni V, Greenwald J, Rhodes D, Hwang Y, Kwiatkowski W, Bushman FD, et al. Identification of a small-molecule binding site at the dimer interface of the HIV integrase catalytic domain. Acta Crystallogr. Sect. D. 2001;57:536–544.

[25]. [25]Cherepanov P, Ambrosio AL, Rahman S, Ellenberger T, Engelman A. Structural basis for the recognition between HIV-1 integrase and transcriptional coactivator p75. Proc. Natl. Acad. Sci. USA. 2005;102:17308–17313. MEDLINE | CrossRef

[26]. [26]Lubkowski J, et al. Structure of the catalytic domain of avian sarcoma virus integrase with a bound HIV-1 integrase-targeted inhibitor. Proc. Natl. Acad. Sci. USA. 1998;95:4831–4836. MEDLINE | CrossRef

[27]. [27]Chen Z, et al. X-ray structure of simian immunodeficiency virus integrase containing the core and C-terminal domain (residues 50–293) – an initial glance of the viral DNA binding platform. J. Mol. Biol. 2000;296:521–533. MEDLINE | CrossRef

[28]. [28]Yang ZN, Mueser TC, Bushman FD, Hyde CC. Crystal structure of an active two-domain derivative of Rous sarcoma virus integrase. J. Mol. Biol. 2000;296:535–548. MEDLINE | CrossRef

[29]. [29]Hare S, Shun M, Gupta S, Valkov E, Engelman A, Cherepanov P, et al. A novel co-crystal structure affords the design of gain-of-function lentiviral integrase mutants in the presence of modified PSIP1/LEDGF/p75. PLoS Pathog. 2009;5:e1000259. CrossRef

[30]. [30]Wang JY, Ling H, Yang W, Craigie R. Structure of a two-domain fragment of HIV-1 integrase: implications for domain organization in the intact protein. EMBO J. 2001;20:7333–7343. MEDLINE | CrossRef

[31]. [31]Cai ML, Huang Y, Caffrey M, Zheng RL, Craigie R, Clore GM, et al. Solution structure of the His12→Cys mutant of the N-terminal zinc binding domain of HIV-1 integrase complexed to cadmium. Protein Sci. 1998;7:2669–2674. MEDLINE | CrossRef

[32]. [32]Cai ML, Zheng RL, Caffrey M, Craigie R, Clore GM, Gronenborn AM. Solution structure of the N-terminal zinc binding domain of HIV-1 integrase. Nat. Struct. Biol. 1997;4:839–840. CrossRef

[33]. [33]Eijkelenboom AP, van den Ent FM, Wechselberger R, Plasterk RH, Kaptein R, Boelens R. Refined solution structure of the dimeric N-terminal HHCC domain of HIV- 2 integrase. J. Biomol. NMR. 2000;18:119–128. MEDLINE | CrossRef

[34]. [34]Eijkelenboom A, Sprangers R, Hard K, Lutzke RAP, Plasterk RHA, Boelens R, et al. Refined solution structure of the C-terminal DNA-binding domain of human immunovirus-1 integrase. Proteins Struct. Funct. Genet. 1999;36:556–564.

[35]. [35]Lodi PJ, Ernst JA, Kuszewski J, Hickman AB, Engelman A, Craigie R, et al. Solution structure of the DNA binding domain of HIV-1 integrase. Biochemistry. 1995;34:9826–9833.

[36]. [36]Cherepanov P. LEDGF/p75 interacts with divergent lentiviral integrases and modulates their enzymatic activity in vitro. Nucleic Acids Res. 2007;35:113–124. CrossRef

[37]. [37]Cherepanov P, Devroe E, Silver PA, Engelman A. Identification of an evolutionarily conserved domain in human lens epithelium-derived growth factor/transcriptional co-activator p75 (LEDGF/p75) that binds HIV-1 integrase. J. Biol. Chem. 2004;279:48883–48892. MEDLINE | CrossRef

[38]. [38]Cherepanov P, Maertens G, Proost P, Devreese B, Van Beeumen J, Engelborghs Y, et al. HIV-1 integrase forms stable tetramers and associates with LEDGF/p75 protein in human cells. J. Biol. Chem. 2003;278:372–381. MEDLINE | CrossRef

[39]. [39]Engelman A, Cherepanov P, Finlay B. The lentiviral integrase binding protein LEDGF/p75 and HIV-1 replication. PLoS Pathog. 2008;4:e1000046. CrossRef

[40]. [40]Poeschla EM. Integrase, LEDGF/p75 and HIV replication. Cell Mol. Life Sci. 2008;65:1403–1424. CrossRef

[41]. [41]Michel F, Crucifix C, Granger F, Eiler S, Mouscadet JF, Korolev S, et al. Structural basis for HIV-1 DNA integration in the human genome, role of the LEDGF/P75 cofactor. EMBO J. 2009;28:980–991. CrossRef

[42]. [42]Hare S, Di Nunzio F, Labeja A, Wang J, Engelman A, Cherepanov P. Structural basis for functional tetramerization of lentiviral integrase. PLoS Pathog. 2009;5:e1000515. CrossRef

[43]. [43]Brass AL, Dykxhoorn DM, Benita Y, Yan N, Engelman A, Xavier RJ, et al. Identification of host proteins required for HIV infection through a functional genomic screen. Science. 2008;319:921–926. CrossRef

[44]. [44]Violot S, et al. The human polycomb group EED protein interacts with the integrase of human immunodeficiency virus type 1. J. Virol. 2003;77:12507–12522. CrossRef

[45]. [45]Rakotobe D, Violot S, Hong SS, Gouet P, Boulanger P. Mapping of immunogenic and protein-interacting regions at the surface of the seven-bladed beta-propeller domain of the HIV-1 cellular interactor EED. Virol. J. 2008;5:32. CrossRef

[46]. [46]Morozov A, Yung E, Kalpana GV. Structure–function analysis of integrase interactor 1/hSNF5L1 reveals differential properties of two repeat motifs present in the highly conserved region. Proc. Natl. Acad. Sci. USA. 1998;95:1120–1125. MEDLINE | CrossRef

[47]. [47]Maroun M, et al. Inhibition of early steps of HIV-1 replication by SNF5/Ini1. J. Biol. Chem. 2006;281:22736–22743. MEDLINE | CrossRef

[48]. [48]Kalpana GV, Marmon S, Wang W, Crabtree GR, Goff SP. Binding and stimulation of HIV-1 integrase by a human homolog of yeast transcription factor SNF5. Science. 1994;266:2002–2006. MEDLINE

[49]. [49]Woodward CL, Prakobwanakit S, Mosessian S, Chow SA. Integrase interacts with nucleoporin NUP153 to mediate the nuclear import of human immunodeficiency virus type 1. J. Virol. 2009;83:6522–6533. CrossRef

[50]. [50]Wilkinson TA, Januszyk K, Phillips ML, Tekeste SS, Zhang M, Miller JT, et al. Identifying and characterizing a functional HIV-1 reverse transcriptase-binding site on integrase. J. Biol. Chem. 2009;284:7931–7939. CrossRef

[51]. [51]Hehl EA, Joshi P, Kalpana GV, Prasad VR. Interaction between human immunodeficiency virus type 1 reverse transcriptase and integrase proteins. J. Virol. 2004;78:5056–5067. CrossRef

[52]. [52]Tasara T, Maga G, Hottiger MO, Hubscher U. HIV-1 reverse transcriptase and integrase enzymes physically interact and inhibit each other. FEBS Lett. 2001;507:39–44. Abstract | Full Text | Full-Text PDF (387 KB) | CrossRef

[53]. [53]Shun MC, et al. Identification and characterization of PWWP domain residues critical for LEDGF/p75 chromatin binding and human immunodeficiency virus type 1 infectivity. J. Virol. 2008;82:11555–11567. CrossRef

[54]. [54]Levin A, Armon-Omer A, Rosenbluh J, Melamed-Book N, Graessmann A, Waigmann E, et al. Inhibition of HIV-1 integrase nuclear import and replication by a peptide bearing integrase putative nuclear localization signal. Retrovirology. 2009;6:112. CrossRef

[55]. [55]Engelman A, Liu Y, Chen HM, Farzan M, Dyda F. Structure-based mutagenesis of the catalytic domain of human immunodeficiency virus type 1 integrase. J. Virol. 1997;71:3507–3514.

[56]. [56]Lu R, Limon A, Devroe E, Silver PA, Cherepanov P, Engelman A. Class II integrase mutants with changes in putative nuclear localization signals are primarily blocked at a postnuclear entry step of human immunodeficiency virus type 1 replication. J. Virol. 2004;78:12735–12746. CrossRef

[57]. [57]McPherson A. Preparation and analysis of protein crystals. New York: John Wiley & Sons; 1982;.

[58]. [58]Leslie AGW. Recent changes to the MOSFLM package for processing film and image plate data Joint CCP4+ESF-EAMCB. Newslett. Protein Crystallogr. 1992;26:.

[59]. [59]Evans P. Scaling and assessment of data quality. Acta Crystallogr. D Biol. Crystallogr. 2006;62:72–82. MEDLINE | CrossRef

[60]. [60]Collaborative Computational Project Number 4. (1994). The CCP4 suite: programs for protein crystallography. Acta Crystallogr. D Biol. Crystallogr. 50, 760–763.

[61]. [61]Navaza J. AMoRe: an automated package for molecular replacement. Acta Crystallogr. A. 1994;50:157–163. CrossRef

[62]. [62]Emsley P, Cowtan K. Coot: model-building tools for molecular graphics. Acta Crystallogr. D Biol. Crystallogr. 2004;60:2126–2132. MEDLINE | CrossRef

[63]. [63]Adams PD, et al. PHENIX: building new software for automated crystallographic structure determination. Acta Crystallogr. D Biol. Crystallogr. 2002;58:1948–1954. MEDLINE | CrossRef

[64]. [64]Laskowski RA, Moss DS, Thornton JM. Main-chain bond lengths and bond angles in protein structures. J. Mol. Biol. 1993;231:1049–1067. MEDLINE | CrossRef

[65]. [65]DeLano WL. The PyMOL Molecular Graphics Systemed. Palo Alto, CA, USA: DeLano Scientific; 2002;.

[66]. [66]Berman HM, Westbrook J, Feng Z, Gilliland G, Bhat TN, Weissig H, et al. The Protein Data Bank. Nucleic Acids Res. 2000;28:235–242. MEDLINE | CrossRef

[67]. [67]Wielens J, Crosby IT, Chalmers DK. A three-dimensional model of the human immunodeficiency virus type 1 integration complex. J. Comput. Aided Mol. Des. 2005;19:301–317. CrossRef

[68]. [68]Ceccherini-Silberstein F, Malet I, D’Arrigo R, Antinori A, Marcelin AG, Perno CF. Characterization and structural analysis of HIV-1 integrase conservation. AIDS Rev. 2009;11:17–29.

[69]. [69]Ovenden SP, et al. Globoidnan A: a lignan from Eucalyptus globoidea inhibits HIV integrase. Phytochemistry. 2004;65:3255–3259. MEDLINE | CrossRef

[70]. [70]Brown GM, Levy HA. Sucrose: precise determination of crystal and molecular structure by neutron diffraction. Science. 1963;141:921–923. MEDLINE

[71]. [71]Immel S, Lichtenthaler FW. Molecular modeling of saccharides, 7. The conformation of sucrose in water: a molecular dynamics approach. Liebigs Annalen. 1995;1995:1925–1937.

[72]. [72]Duker JM, Serianni AS. (13C)-substituted sucrose: 13C–1H and 13C–13C spin coupling constants to assess furanose ring and glycosidic bond conformations in aqueous solution. Carbohydr. Res. 1993;249:281–303. MEDLINE | CrossRef

[73]. [73]Skov LK, Mirza O, Sprogoe D, Dar I, Remaud-Simeon M, Albenne C, et al. Oligosaccharide and sucrose complexes of amylosucrase. Structural implications for the polymerase activity. J. Biol. Chem. 2002;277:47741–47747. MEDLINE | CrossRef

[74]. [74]Bourne Y, Rouge P, Cambillau C. X-ray structure of a (alpha-Man(1–3)beta–Man(1–4)GlcNAc)–lectin complex at 2.1-A resolution. The role of water in sugar-lectin interaction. J. Biol. Chem. 1990;265:18161–18165. MEDLINE

[75]. [75]Zhou T, et al. Structural definition of a conserved neutralization epitope on HIV-1 gp120. Nature. 2007;445:732–737. CrossRef

[76]. [76]Kerfeld CA, Sawaya MR, Brahmandam V, Cascio D, Ho KK, Trevithick-Sutton CC, et al. The crystal structure of a cyanobacterial water-soluble carotenoid binding protein. Structure. 2003;11:55–65. MEDLINE | CrossRef

[77]. [77]Connolly ML. Solvent-accessible surfaces of proteins and nucleic acids. Science (Washington, D.C.). 1983;221:709–713.

[78]. [78]Lin Z, et al. Chicoric acid analogues as HIV-1 integrase inhibitors. J. Med. Chem. 1999;42:1401–1414. MEDLINE | CrossRef

a Medicinal Chemistry and Drug Action, Monash Institute of Pharmaceutical Sciences, 381 Royal Parade, Parkville, Victoria 3052, Australia

b Structural Biology Laboratory, St. Vincent’s Institute, 9 Princes Street, Fitzroy, Victoria 3065, Australia

c Avexa Ltd., 576 Swan Street Richmond, Victoria 3121, Australia

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

Corresponding author. Fax: +61 3 99039582.

Corresponding author at: Structural Biology Laboratory, St. Vincent’s Institute, 9 Princes Street, Fitzroy, Victoria 3065, Australia. Fax +61 3 94162676.

PII: S0014-5793(10)00210-3

doi:10.1016/j.febslet.2010.03.016

5 of 34