Ko-Hsin Chin

The structure and inhibition of a GGDEF diguanylate cyclase complexed with (c-di-GMP)2 at the active site

Cyclic diguanosine monophosphate (c-di-GMP) is a key signalling molecule involved in regulating many important biological functions in bacteria. The synthesis of c-di-GMP is catalyzed by the GGDEF-domain-containing diguanylate cyclase (DGC), the activity of which is regulated by the binding of product at the allosteric inhibitory (I) site. However, a significant number of GGDEF domains lack the RxxD motif characteristic of the allosteric I site. Here, the structure of XCC4471(GGDEF), the GGDEF domain of a DGC from Xanthomonas campestris, in complex with c-di-GMP has been solved. Unexpectedly, the structure of the complex revealed a GGDEF-domain dimer cross-linked by two molecules of c-di-GMP at the strongly conserved active sites. In the complex (c-di-GMP)(2) adopts a novel partially intercalated form, with the peripheral guanine bases bound to the guanine-binding pockets and the two central bases stacked upon each other. Alteration of the residues involved in specific binding to c-di-GMP led to dramatically reduced K(d) values between XCC4471(GGDEF) and c-di-GMP. In addition, these key residues are strongly conserved among the many thousands of GGDEF-domain sequences identified to date. These results indicate a new product-bound form for GGDEF-domain-containing proteins obtained via (c-di-GMP)(2) binding at the active site. This novel XCC4471(GGDEF)-c-di-GMP complex structure may serve as a general model for the design of lead compounds to block the DGC activity of GGDEF-domain-containing proteins in X. campestris or other microorganisms that contain multiple GGDEF-domain proteins.

A cyclic GMP-dependent signalling pathway regulates bacterial phytopathogenesis

Cyclic guanosine 3′,5′‐monophosphate (cyclic GMP) is a second messenger whose role in bacterial signalling is poorly understood. A genetic screen in the plant pathogen Xanthomonas campestris (Xcc) identified that XC_0250, which encodes a protein with a class III nucleotidyl cyclase domain, is required for cyclic GMP synthesis. Purified XC_0250 was active in cyclic GMP synthesis in vitro. The linked gene XC_0249 encodes a protein with a cyclic mononucleotide‐binding (cNMP) domain and a GGDEF diguanylate cyclase domain. The activity of XC_0249 in cyclic di‐GMP synthesis was enhanced by addition of cyclic GMP. The isolated cNMP domain of XC_0249 bound cyclic GMP and a structure–function analysis, directed by determination of the crystal structure of the holo‐complex, demonstrated the site of cyclic GMP binding that modulates cyclic di‐GMP synthesis. Mutation of either XC_0250 or XC_0249 led to a reduced virulence to plants and reduced biofilm formation in vitro. These findings describe a regulatory pathway in which cyclic GMP regulates virulence and biofilm formation through interaction with a novel effector that directly links cyclic GMP and cyclic di‐GMP signalling.

Nucleotide binding by the widespread high-affinity cyclic di-GMP receptor MshEN domain

C-di-GMP is a bacterial second messenger regulating various cellular functions. Many bacteria contain c-di-GMP-metabolizing enzymes but lack known c-di-GMP receptors. Recently, two MshE-type ATPases associated with bacterial type II secretion system and type IV pilus formation were shown to specifically bind c-di-GMP. Here we report crystal structure of the MshE N-terminal domain (MshEN1-145) from Vibrio cholerae in complex with c-di-GMP at a 1.37 Å resolution. This structure reveals a unique c-di-GMP-binding mode, featuring a tandem array of two highly conserved binding motifs, each comprising a 24-residue sequence RLGxx(L/V/I)(L/V/I)xxG(L/V/I)(L/V/I)xxxxLxxxLxxQ that binds half of the c-di-GMP molecule, primarily through hydrophobic interactions. Mutating these highly conserved residues markedly reduces c-di-GMP binding and biofilm formation by V. cholerae. This c-di-GMP-binding motif is present in diverse bacterial proteins exhibiting binding affinities ranging from 0.5 μM to as low as 14 nM. The MshEN domain contains the longest nucleotide-binding motif reported to date.

Insights into the Alkyl Peroxide Reduction Pathway of Xanthomonas campestris Bacterioferritin Comigratory Protein from the Trapped Intermediate―Ligand Complex Structures

Considerable insights into the oxidoreduction activity of the Xanthomonas campestris bacterioferritin comigratory protein (XcBCP) have been obtained from trapped intermediate/ligand complex structures determined by X-ray crystallography. Multiple sequence alignment and enzyme assay indicate that XcBCP belongs to a subfamily of atypical 2-Cys peroxiredoxins (Prxs), containing a strictly conserved peroxidatic cysteine (C(P)48) and an unconserved resolving cysteine (C(R)84). Crystals at different states, i.e. Free_SH state, Intra_SS state, and Inter_SS state, were obtained by screening the XcBCP proteins from a double C48S/C84S mutant, a wild type, and a C48A mutant, respectively. A formate or an alkyl analog with two water molecules that mimic an alkyl peroxide substrate was found close to the active site of the Free_SH or Inter_SS state, respectively. Their global structures were found to contain a novel substrate-binding pocket capable of accommodating an alkyl chain of no less than 16 carbons. In addition, in the Intra_SS or Inter_SS state, substantial local unfolding or complete unfolding of the C(R)-helix was detected, with the C(P)-helix remaining essentially unchanged. This is in contrast to the earlier observation that the C(P)-helix exhibits local unfolding during disulfide bond formation in typical 2-Cys Prxs. These rich experimental data have enabled us to propose a pathway by which XcBCP carries out its oxidoreduction activity through the alternate opening and closing of the substrate entry channel and the disulfide-bond pocket.

XC1028 from Xanthomonas campestris adopts a PilZ domain-like structure without a c-di-GMP switch.

The crystal structure of XC1028 from Xanthomonas campestris has been determined to a resolution of 2.15 Å using the multiple anomalous dispersion approach. It bears significant sequence identity and similarity values of 64.10% and 70.09%, respectively, with PA2960, a protein indispensable for type IV pilus‐mediated twitching motility, after which the PilZ motif was first named. However, both XC1028 and PA2960 lack detectable c‐di‐GMP binding capability. Although XC1028 adopts a structure comprising a five‐stranded β‐barrel core similar to other canonical PilZ domains with robust c‐di‐GMP binding ability, considerable differences are observed in the N‐terminal motif; XC1028 assumes a compact five‐stranded β‐barrel without an extra long N‐terminal motif, whereas other canonical PilZ domains contain a long N‐terminal sequence embedded with an essential “c‐di‐GMP switch” motif. In addition, a β‐strand (β1) in the N‐terminal motif, running in exactly opposite polarity to that of XC1028, is found inserted into the parallel β3/β1′ strands, forming a completely antiparallel β4↓β3↑β1↓β1′↑ sheet in the canonical PilZ domains. Such dramatic structural differences at the N‐terminus may account for the diminished c‐di‐GMP binding capability of XC1028, and suggest that interactions with additional proteins are necessary to bind c‐di‐GMP for type IV fimbriae assembly. Proteins 2009.

Structural polymorphism of c‐di‐GMP bound to an EAL domain and in complex with a type II PilZ‐domain protein

Cyclic di-GMP (c-di-GMP) is a novel secondary-messenger molecule that is involved in regulating a plethora of important bacterial activities through binding to an unprecedented array of effectors. Proteins with a canonical PilZ domain that bind c-di-GMP play crucial roles in regulating flagellum-based motility. In contrast, noncanonical type II PilZ domains that do not effectively bind c-di-GMP regulate twitching motility, which is dependent on type IV pili (T4P). Recent data indicate that T4P biogenesis is initiated via the interaction of a noncanonical type II PilZ protein with the GGDEF/EAL-domain protein FimX and the pilus motor protein PilB at high c-di-GMP concentrations. However, the molecular details of such interactions remain to be elucidated. In this manuscript, the first hetero-complex crystal structure between a type II PilZ protein and the EAL domain of the FimX protein (FimX(EAL)) from Xanthomonas campestris pv. campestris (Xcc) in the presence of c-di-GMP is reported. This work reveals two novel conformations of monomeric c-di-GMP in the XccFimX(EAL)-c-di-GMP and XccFimX(EAL)-c-di-GMP-XccPilZ complexes, as well as a unique interaction mode of a type II PilZ domain with FimX(EAL). These findings indicate that c-di-GMP is sufficiently flexible to adjust its conformation to match the corresponding recognition motifs of different cognate effectors. Together, these results represent a first step towards an understanding of how T4P biogenesis is controlled by c-di-GMP at the molecular level and also of the ability of c-di-GMP to bind to a wide variety of effectors.

Structure of Stenotrophomonas maltophilia FeoA complexed with zinc: a unique prokaryotic SH3-­domain protein that possibly acts as a bacterial ferrous iron-transport activating factor

Iron is vital to the majority of prokaryotes, with ferrous iron believed to be the preferred form for iron uptake owing to its much better solubility. The major route for bacterial ferrous iron uptake is found to be via an Feo (ferrous iron-transport) system comprising the three proteins FeoA, FeoB and FeoC. Although the structure and function of FeoB have received much attention recently, the roles played by FeoA and FeoC have been little investigated to date. Here, the tertiary structure of FeoA from Stenotrophomonas maltophilia (Sm), a vital opportunistic pathogen in immunodepressed hosts, is reported. The crystal structure of SmFeoA has been determined to a resolution of 1.7 A using an Se single-wavelength anomalous dispersion (Se-SAD) approach. Although SmFeoA bears low sequence identity to eukaryotic proteins, its structure is found to adopt a eukaryotic SH3-domain-like fold. It also bears weak similarity to the C-terminal SH3 domain of bacterial DtxR (diphtheria toxin regulator), with some unique characteristics. Intriguingly, SmFeoA is found to adopt a unique dimer cross-linked by two zinc ions and six anions (chloride ions). Since FeoB has been found to contain a G-protein-like domain with low GTPase activity, FeoA may interact with FeoB through the SH3-G-protein domain interaction to act as a ferrous iron-transport activating factor.

A Novel Tetrameric PilZ Domain Structure from Xanthomonads

PilZ domain is one of the key receptors for the newly discovered secondary messenger molecule cyclic di-GMP (c-di-GMP). To date, several monomeric PilZ domain proteins have been identified. Some exhibit strong c-di-GMP binding activity, while others have barely detectable c-di-GMP binding activity and require an accessory protein such as FimX to indirectly respond to the c-di-GMP signal. We now report a novel tetrameric PilZ domain structure of XCC6012 from the plant pathogen Xanthomonas campestris pv. campestris (Xcc). It is one of the four PilZ domain proteins essential for Xcc pathogenicity. Although the monomer adopts a structure similar to those of the PilZ domains with very weak c-di-GMP binding activity, it is nevertheless interrupted in the middle by two extra long helices. Four XCC6012 proteins are thus self-assembled into a tetramer via the extra heptad repeat α3 helices to form a parallel four-stranded coiled-coil, which is further enclosed by two sets of inclined α2 and α4 helices. We further generated a series of XCC6012 variants and measured the unfolding temperatures and oligomeric states in order to investigate the nature of this novel tetramer. Discovery of this new PilZ domain architecture increases the complexity of c-di-GMP-mediated regulation.

Crystal structure of the C-terminal domain of a flagellar hook-capping protein from Xanthomonas campestris

The crystal structure of the C-terminal domain of a hook-capping protein FlgD from the plant pathogen Xanthomonas campestris (Xc) has been determined to a resolution of ca 2.5 A using X-ray crystallography. The monomer of whole FlgD comprises 221 amino acids with a molecular mass of 22.7 kDa, but the flexible N-terminus is cleaved for up to 75 residues during crystallization. The final structure of the C-terminal domain reveals a novel hybrid comprising a tudor-like domain interdigitated with a fibronectin type III domain. The C-terminal domain of XcFlgD forms three types of dimers in the crystal. In agreement with this, analytical ultracentrifugation and gel filtration experiments reveal that they form a stable dimer in solution. From these results, we propose that the Xc flagellar hook cap protein FlgD comprises two individual domains, a flexible N-terminal domain that cannot be detected in the current study and a stable C-terminal domain that forms a stable dimer.

The crystal structure of XC1258 from Xanthomonas campestris: A putative procaryotic Nit protein with an arsenic adduct in the active site

The crystal structure of XC1258 from Xanthomonas campestris: A putative procaryotic Nit protein with an arsenic adduct in the active site Ko-Hsin Chin, Ying-Der Tsai, Nei-Li Chan, Kai-Fa Huang, Andrew H.-J. Wang, and Shan-Ho Chou* 1 National Chung Hsing University Biotechnology Center, National Chung-Hsing University, Taichung, 40227, Taiwan, Republic of China 2 Institute of Biochemistry, National Chung-Hsing University, Taichung, 40227, Taiwan, Republic of China 3 Core Facility for Protein Crystallography, Academia Sinica, Nankang, Taipei, Taiwan, Republic of China 4 Institute of Biological Chemistry, Academia Sinica, Nankang, Taipei, Taiwan, Republic of China