Translational mini-review series on complement factor H: structural and functional correlations for factor H.

The 155-kDa glycoprotein, complement factor H (CFH), is a regulator of complement activation that is abundant in human plasma. Three-dimensional structures of over half the 20 complement control protein (CCP) modules in CFH have been solved in the context of single-, double- and triple-module segments. Proven binding sites for C3b occupy the N and C termini of this elongated molecule and may be brought together by a bend in CFH mediated by its central CCP modules. The C-terminal CCP 20 is key to the ability of the molecule to adhere to polyanionic markers on self-surfaces where CFH acts to regulate amplification of the alternative pathway of complement. The surface patch on CCP 20 that binds to model glycosaminoglycans has been mapped using nuclear magnetic resonance (NMR), as has a second glycosaminoglycan-binding patch on CCP 7. These patches include many of the residue positions at which sequence variations have been linked to three complement-mediated disorders: dense deposit disease, age-related macular degeneration and atypical haemolytic uraemic syndrome. In one plausible model, CCP 20 anchors CFH to self-surfaces via a C3b/polyanion composite binding site, CCP 7 acts as a 'proof-reader' to help discriminate self- from non-self patterns of sulphation, and CCPs 1-4 disrupt C3/C5 convertase formation and stability.

Three-dimensional structure and flexibility of proteins of the RCA family - a progress report.

Members of the regulator of complement activation (RCA) protein family perform a vital role in health and disease. In this report we describe our efforts to solve the structures of human membrane cofactor protein (CD46), the vaccinia virus complement control protein, which mimics mammalian RCA proteins, and human complement receptor type 1 (CD35). These examples illustrate that, despite good progress over the last decade, the regulators of complement, as extracellular multiple domain glycoproteins, still pose formidable problems to structural biologists. Many important questions remain unanswered, in particular with regard to the flexibility of these proteins and the extent to which they undergo conformational rearrangements on engaging their binding partners.

Structure and flexibility of the multiple domain proteins that regulate complement activation.

In this review we summarise more than 10 years of biophysical exploration into the structural biology of the regulators of complement activation (RCA). The five human proteins responsible for regulation of the early events of complement are homologous and are composed largely from building blocks called "complement control protein (CCP) modules". Unlike most multiple domain proteins they do not contain any of the other widely occurring module types. This apparent simplicity of RCA structure, however, is belied by their sophistication of function. In fact, the structures of the individual CCP modules exhibit wide variations on a common theme while the extent and nature of intermodular connections is diverse. Some neighbouring modules within a protein stabilise each other and some co-operate to form specific binding surfaces. The degree of true "modularity" of CCPs is open to debate. The study of RCA proteins clearly illustrates the value of combining complementary structural biology techniques. The results could have implications for folding, evolution, flexibility and structure-function relationships of other molecules in the large, diverse and little understood category of multiple domain proteins.

Solution structure and dynamics of the central CCP module pair of a poxvirus complement control protein.

The complement control protein (CCP) module (also known as SCR, CCP or sushi domain) is prevalent amongst proteins that regulate complement activation. Functional and mutagenesis studies have shown that in most cases two or more neighbouring CCP modules form specific binding sites for other molecules. Hence the orientation in space of a CCP module with respect to its neighbours and the flexibility of the intermodular junction are likely to be critical for function. Vaccinia virus complement control protein (VCP) is a complement regulatory protein composed of four tandemly arranged CCP modules. The solution structure of the carboxy-terminal half of this protein (CCP modules 3 and 4) has been solved previously. The structure of the central portion (modules 2 and 3, VCP approximately 2,3) has now also been solved using NMR spectroscopy at 37 degrees C. In addition, the backbone dynamics of VCP approximately 2,3 have been characterised by analysis of its (15)N relaxation parameters. Module 2 has a typical CCP module structure while module 3 in the context of VCP approximately 2,3 has some modest but significant differences in structure and dynamics to module 3 within the 3,4 pair. Modules 2 and 3 do not share an extensive interface, unlike modules 3 and 4. Only two possible NOEs were identified between the bodies of the modules, but a total of 40 NOEs between the short intermodular linker of VCP approximately 2,3 and the bodies of the two modules determines a preferred, elongated, orientation of the two modules in the calculated structures. The anisotropy of rotational diffusion has been characterised from (15)N relaxation data, and this indicates that the time-averaged structure is more compact than suggested by (1)H-(1)H NOEs. The data are consistent with the presence of many intermodular orientations, some of which are kinked, undergoing interconversion on a 10(-8)-10(-6) second time-scale. A reconstructed representation of modules 2-4 allows visualisation of the spatial arrangement of the 11 substitutions that occur in the more potent complement inhibitor from Variola (small pox) virus.

Solution structure and dynamics of an open beta-sheet, glycolytic enzyme, monomeric 23.7 kDa phosphoglycerate mutase from Schizosaccharomyces pombe.

The structure and backbone dynamics of a double labelled (15N,13C) monomeric, 23.7 kD phosphoglycerate mutase (PGAM) from Schizosaccharomyces pombe have been investigated in solution using NMR spectroscopy. A set of 3125 NOE-derived distance restraints, 148 restraints representing inferred hydrogen bonds and 149 values of (3)J(HNHalpha) were used in the structure calculation. The mean rmsd from the average structure for all backbone atoms from residues 6-205 in the best 21 calculated structures was 0.59 A. The core of the enzyme includes an open, twisted, six-stranded beta-sheet flanked by four alpha-helices and a short 3(10)-helix. An additional smaller domain contains two short antiparallel beta-strands and a further pair of alpha-helices. The C(alpha) atoms of the S. pombe PGAM may be superimposed on their equivalents in one of the four identical subunits of Saccharomyces cerevisiae PGAM with an rmsd of 1.34 A (0.92 A if only the beta-sheet is considered). Small differences between the two structures are attributable partly to the deletion in the S. pombe sequence of a 25 residue loop involved in stabilising the S. cerevisiae tetramer. Analysis of 15N relaxation parameters indicates that PGAM tumbles isotropically with a rotational correlation time of 8.7 ns and displays a range of dynamic features. Of 178 residues analysed, only 77 could be fitted without invoking terms for fast internal motion or chemical exchange, and out of the remainder, 77 required a chemical exchange term. Significantly, 46 of the slowly exchanging (milli- to microsecond) residues lie in helices, and these account for two-thirds of all analysed helix residues. On the contrary, only one beta-sheet residue required an exchange term. In contrast to other analyses of backbone dynamics reported previously, residues in slow exchange appeared to correlate with architectural features of the enzyme rather than congregating close to ligand binding sites.

RETRACTED: Crystal structure of a complement control protein that regulates both pathways of complement activation and binds heparan sulfate proteoglycans.

This article has been retracted: please see Elsevier Policy on Article Withdrawal (https://www.elsevier.com/about/our-business/policies/article-withdrawal). This article has been retracted at the request of request of the editors. Cell is retracting this paper reporting structures of a poxvirus protein, VCP, that inhibits the complement system. The paper presents a structural model derived from two crystal forms of the protein (PDB: 1G40 and 1G44) that defines an interaction surface implicated in inhibition of complement C3 proteins and visualizes heparin binding sites. We were contacted by the University of Alabama, Birmingham (UAB), the corresponding author's institution, with a report detailing concerns about the veracity of the structures and recommending that the structures be retracted from the Protein Data Bank. We then conducted an assessment with input from experts in the field who found that the structures as presented in the paper were not consistent with available data, including spatial packing and structure (B) factors. These findings were consistent with issues contained in the UAB report. A subsequent investigation by the Department of Health and Human Services Office of Research Integrity (https://www.federalregister.gov/documents/2018/04/16/2018-07782/findings-of-research-misconduct) has concluded that the corresponding author, Krishna H.M. Murthy, engaged in research misconduct and that the structures were falsified and/or fabricated. Given the results of our own assessment and the institutional investigations, the most appropriate course of action is to retract the paper. Co-authors Nick Mullin, Paul N. Barlow, and Craig M. Ogata support this retraction.

DNA recognition by the methyl-CpG binding domain of MeCP2.

The methyl-CpG binding domain (MBD) of the transcriptional repressor MeCP2 has been proposed to recognize a single symmetrically methylated CpG base pair via hydrophobic patches on an otherwise positively charged DNA binding surface. We have tested this binding model by analysis of mutant derivatives of the MeCP2 MBD in electrophoretic mobility shift assays complemented by NMR structural analysis. Exposed arginine side chains on the binding face, in particular Arg-111, were found to be critical for binding. Arg-111 was found to interact with the conserved aspartate side chain Asp-121, which is proposed to orientate the arginine side chain to allow specific contacts with the DNA. The conformational flexibility of the disordered B-C loop region, which forms part of the binding face, was also shown to be important. In contrast, mutation of the exposed hydrophobic side chains had a less severe effect on DNA binding. This suggests that the Arg-111 side chain may contribute to sequence-specific recognition of the CpG site rather than simply making nonspecific contacts with the phosphate backbone. The majority of missense mutations within the MBD found in the human genetic disorder Rett syndrome were shown or predicted to affect folding of the domain rather than the DNA recognition event directly.

Characterization of active-site mutants of Schizosaccharomyces pombe phosphoglycerate mutase. Elucidation of the roles of amino acids involved in substrate binding and catalysis.

The roles of a number of amino acids present at the active site of the monomeric phosphoglycerate mutase from the fission yeast Schizosaccharomyces pombe have been explored by site-directed mutagenesis. The amino acids examined could be divided broadly into those presumed from previous related structural studies to be important in the catalytic process (R14, S62 and E93) and those thought to be important in substrate binding (R94, R120 and R121). Most of these residues have not previously been studied by site-directed mutagenesis. All the mutants except R14 were expressed in an engineered null strain of Saccharomyces cerevisiae (S150-gpm:HIS) in good yield. The R14Q mutant was expressed in good yield in the transformed AH22 strain of S. cerevisiae. The S62A mutant was markedly unstable, preventing purification. The various mutants were purified to homogeneity and characterized in terms of kinetic parameters, CD and fluorescence spectra, stability towards denaturation by guanidinium chloride, and stability of phosphorylated enzyme intermediate. In addition, the binding of substrate (3-phosphoglycerate) to wild-type, E93D and R120,121Q enzymes was measured by isothermal titration calorimetry. The results provide evidence for the proposed roles of each of these amino acids in the catalytic cycle and in substrate binding, and will support the current investigation of the structure and dynamics of the enzyme using multidimensional NMR techniques.

Dissecting sites important for complement regulatory activity in membrane cofactor protein (MCP; CD46).

Membrane cofactor protein (MCP; CD46), a widely distributed regulator of complement activation, is a cofactor for the factor I-mediated degradation of C3b and C4b deposited on host cells. MCP possesses four extracellular, contiguous complement control protein modules (CCPs) important for this inhibitory activity. The goal of the present study was to delineate functional sites within these modules. We employed multiple approaches including mutagenesis, epitope mapping, and comparisons to primate MCP to make the following observations. First, functional sites were located to each of the four CCPs. Second, some residues were important for both C3b and C4b interactions while others were specific for one or the other. Third, while a reduction in ligand binding was invariably accompanied by a parallel reduction in cofactor activity (CA), other mutants lost or had reduced CA but retained ligand binding. Fourth, two C4b-regulatory domains overlapped measles virus interactive regions, indicating that the hemagglutinin docks to a site important for complement inhibition. Fifth, several MCP regulatory areas corresponded to functionally critical, homologous positions in other CCP-bearing C3b/C4b-binding proteins. Based on these data and the recently derived crystal structure of repeats one and two, computer modeling was employed to predict MCP structure and examine active sites.

Conserved surface-exposed K/R-X-K/R motifs and net positive charge on poxvirus complement control proteins serve as putative heparin binding sites and contribute to inhibition of molecular interactions with human endothelial cells: a novel mechanism for evasion of host defense.

Vaccinia virus complement control protein (VCP) has been shown to possess the ability to inhibit both classical and alternative complement pathway activation. The newly found ability of this protein to bind to heparin has been shown in previous studies to result in uptake by mast cells, possibly promoting tissue persistence. It has also been shown to reduce chemotactic migration of leukocytes by blocking chemokine binding. In addition, this study shows that VCP-through its ability to bind to glycosaminoglycans (heparin-like molecules) on the surface of human endothelial cells-is able to block antibody binding to surface major histocompatibility complex class I molecules. Since heparin binding is critical for many functions of this protein, we have attempted to characterize the molecular basis for this interaction. Segments of this protein, generated by genetic engineering of the DNA encoding VCP into the Pichia pastoris expression system, were used to localize the regions with heparin binding activity. These regions were then analyzed to more specifically define their properties for binding. It was found that the number of putative binding sites (K/R-X-K/R), the overall positive charge, and the percentage of positively charged amino acids within the protein were responsible for this interaction.

Structural changes accompanying pH-induced dissociation of the beta-lactoglobulin dimer.

We have used NMR spectroscopy to determine the three-dimensional (3D) structure, and to characterize the backbone dynamics, of a recombinant version of bovine beta-lactoglobulin (variant A) at pH 2. 6, where the protein is a monomer. The structure of this low-pH form of beta-lactoglobulin is very similar to that of a subunit within the dimer at pH 6.2. The root-mean-square deviation from the pH 6.2 (crystal) structure, calculated for backbone atoms of residues 6-160, is approximately 1.3 A. Differences arise from the orientation, with respect to the calyx, of the A-B and C-D loops, and of the flanking three-turn alpha-helix. The hydrophobic cavity within the calyx is retained at low pH. The E-F loop (residues 85-90), which moves to occlude the opening of the cavity over the pH range 7.2-6.2, is in the "closed" position at pH 2.6, and the side chain of Glu89 is buried. We also carried out measurements of (15)N T(1)s and T(2)s and (1)H-(15)N heteronuclear NOEs at pH 2.6 and 37 degrees C. Although the residues of the E-F loop (residues 86-89) have the highest crystallographic B-factors, the conformation of this loop is reasonably well defined by the NMR data, and its backbone is not especially mobile on the pico- to nanosecond time scale. Several residues (Ser21, Lys60, Ala67, Leu87, and Glu112) exhibit large ratios of T(1) to T(2), consistent with conformational exchange on a micro- to millisecond time scale. The positions of these residues in the 3D structure of beta-lactoglobulin are consistent with a role in modulating access to the hydrophobic cavity.

3D HCCH(3)-TOCSY for resonance assignment of methyl-containing side chains in (13)C-labeled proteins.

Two 3D experiments, (H)CCH(3)-TOCSY and H(C)CH(3)-TOCSY, are proposed for resonance assignment of methyl-containing amino acid side chains. After the initial proton-carbon INEPT step, during which either carbon or proton chemical shift labeling is achieved (t(1)), the magnetization is spread along the amino acid side chains by a carbon spin lock. The chemical shifts of methyl carbons are labeled (t(2)) during the following constant time interval. Finally the magnetization is transferred, in a reversed INEPT step, to methyl protons for detection (t(3)). The proposed experiments are characterized by high digital resolution in the methyl carbon dimension (t(2max) = 28.6 ms), optimum sensitivity due to the use of proton decoupling during the long constant time interval, and an optional removal of CH(2), or CH(2) and CH, resonances from the F(2)F(3) planes. The building blocks used in these experiments can be implemented in a range of heteronuclear experiments focusing on methyl resonances in proteins. The techniques are illustrated using a (15)N, (13)C-labeled E93D mutant of Schizosacharomyces pombe phosphoglycerate mutase (23.7 kDa).

Simultaneous CT-13C and VT-15N chemical shift labelling: application to 3D NOESY-CH3NH and 3D 13C,15N HSQC-NOESY-CH3NH.

Based on the HSQC scheme, we have designed a 2D heterocorrelated experiment which combines constant time (CT) 13C and variable time (VT) 15N chemical shift labelling. Although applicable to all carbons, this mode is particularly suitable for simultaneous recording of methyl-carbon and nitrogen chemical shifts at high digital resolution. The methyl carbon magnetisation is in the transverse plane during the whole CT period (1/J(CC) = 28.6 ms). The magnetisation originating from NH protons is initially stored in the 2HzNz state, then prior to the VT chemical shift labelling period is converted into 2HzNy coherence. The VT -15N mode eliminates the effect of 1J(N,CO) and 1,2J(N,CA) coupling constants without the need for band-selective carbon pulses. An optional editing procedure is incorporated which eliminates signals from CH2 groups, thus removing any potential overlap with the CH3 signals. The CT-13CH3,VT-15N HSQC building block is used to construct two 3D experiments: 3D NOESY-CH3NH and 3D 13C,15N HSQC-NOESY-CH3NH. Combined use of these experiments yields proton and heteronuclear chemical shifts for moieties experiencing NOEs with CH3 and NH protons. These NOE interactions are resolved as a consequence of the high digital resolution in the carbon and nitrogen chemical shifts of CH3 and NH groups, respectively. The techniques are illustrated using a double labelled sample of the CH domain from calponin.

Central modules of the vaccinia virus complement control protein are not in extensive contact.

The 28.6 kDa vaccinia virus complement control protein (VCP) is an inhibitor of the complement system and has therapeutic potential. It is composed of four domains or modules and is a homologue of complement receptor 1 (CR1) and other mammalian regulators of complement activation. A key aspect to structure-function relationships in these proteins is the extent of intramolecular module-module interactions, since these dictate the overall shape and flexibility of the molecules. A protein fragment (VCP approximately 2,3) encompassing modules 2 and 3 of VCP was over-expressed in Pichia pastoris. Ultracentrifugation showed that VCP approximately 2,3 is highly asymmetric with an axial ratio of 5.3:1, which is consistent with an end-to-end arrangement of the two modules. NMR spectroscopy, differential scanning calorimetry, CD and intrinsic tryptophan fluorescence were used to monitor unfolding of VCP approximately 2,3. Experiments performed over a range of temperatures and concentrations of guanidinium chloride revealed that module 2 unfolds under milder conditions than, and independently of, module 3. Unfolding of module 2 is not associated with extensive changes in amide (15)N and (1)H chemical shifts of module 3, implying that the modules do not form an extensive intermodular interface. Results obtained in this work for VCP approximately 2,3 are compared with those obtained in a study of CR1 modules 15-17 [Kirkitadze, Krych, Uhrin, Dryden, Smith, Cooper, Wang, Hauhart, Atkinson and Barlow (1999) Biochemistry 38, 7019-7031].

Co-operativity between modules within a C3b-binding site of complement receptor type 1.

Complement receptor type 1 (CR1) has 30 modules in its extracellular portion. An understanding of structure-function relationships within CR1 is being assembled gradually from studies of overlapping protein fragments. A CR1 fragment corresponding to modules 16 and 17 was expressed recombinantly as a non-glycosylated protein and its stability and unfolding characteristics studied using biophysical techniques. The results were compared with data collected previously on a CR1 fragment encompassing modules 15, 16 and 17 which together constitute a C3b-binding site (Kirkitadze, M.D., Krych, M., Uhrin, D. , Dryden, D.T.F., Smith, B.O., Wang, X., Hauhart, R., Atkinson, J.P. and Barlow, P.N. (1999) Biochemistry 38, 7019-7031). Modules within CR1 were found to co-operate during unfolding. The folding, stability and flexibility of this protein is therefore likely to be a complex function, and not just the sum, of contributions from individual modules.

The solution structure of the domain from MeCP2 that binds to methylated DNA.

MeCP2 is an abundant mammalian protein that binds methylated CpG (mCpG) sequences within double-stranded DNA, represses transcription by recruiting histone deacetylases, and is essential for embryonic development. It is one of a family of proteins which mediate the biological consequences of DNA methylation. These proteins each possess a sequence motif of about 70 residues which, in MeCP2, form a domain necessary and sufficient for binding to mCpG. The solution structure of the mCpG-binding domain (MBD) from MeCP2 has been solved and the DNA-binding surface of the domain mapped using NMR spectroscopy. Residues 95-162 of MeCP2 adopt a novel fold forming a wedge-shaped structure. An N-terminal four-stranded antiparallel beta-sheet forms one face of the wedge, while the other face is formed mainly by a C-terminal helical region. The thin end of the wedge is extended by a long loop between beta-strands B and C containing many basic residues. The B-C loop together with residues in strands B, C and D, and at the N terminus of the alpha-helix, appears to form an interface with methylated DNA. Unstructured residues at the NH2 terminus of the domain are also involved in formation of the complex. The presence of numerous arginine and lysine side-chains on the DNA-binding surface of MBD is consistent with the requirement for the mCpG site to be flanked by non-specific sequences of base-pairs. The absence of symmetry in the domain implies that recognition does not exploit the symmetry of the binding site. A conserved hydrophobic pocket containing the side-chains of Tyr123 and Ile125 on the positively charged beta-sheet face is a candidate for the region of contact with the methyl-groups of the modified cytosine residues.

Independently melting modules and highly structured intermodular junctions within complement receptor type 1.

A segment of complement receptor type 1 (CR1) corresponding to modules 15-17 was overexpressed as a functionally active recombinant protein with N-glycosylation sites ablated by mutagenesis (referred to as CR1 approximately 15-17(-)). A protein consisting of modules 15 and 16 and another corresponding to module 16 were also overexpressed. Comparison of heteronuclear nuclear magnetic resonance (NMR) spectra for the single, double, and triple module fragments indicated that module 16 makes more extensive contacts with module 15 than with module 17. A combination of NMR, differential scanning calorimetry, circular dichroism, and tryptophan-derived fluorescence indicated a complex unfolding pathway for CR1 approximately 15-17(-). As temperature or denaturant concentration was increased, the 16-17 junction appeared to melt first, followed by the 15-16 junction, and module 17 itself; finally, modules 15 and 16 became denatured. Modules 15 and 16 adopted an intermediate state prior to total denaturation. These results are compared with a previously published study [Clark, N. S., Dodd, I, Mossakowska, D. E., Smith, R. A. G., and Gore, M. G. (1996) Protein Eng. 9, 877-884] on a fragment consisting of the N-terminal three CR1 modules which appeared to melt as a single unit.

Total chemical synthesis and high-resolution crystal structure of the potent anti-HIV protein AOP-RANTES.

BACKGROUND: RANTES is a CC-type chemokine protein that acts as a chemoattractant for several kinds of leukocytes, playing an important pro-inflammatory role. Entry of human immunodeficiency virus-1 (HIV-1) into cells depends on the chemokine receptor CCR5. RANTES binds CCR5 and inhibits HIV-1 entry into peripheral blood cells. Interaction with chemokine receptors involves a distinct set of residues at the amino terminus of RANTES. This finding was utilized in the development of a chemically modified aminooxypentane derivative of RANTES, AOP-RANTES, that was originally produced from the recombinant protein using semisynthetic methods. RESULTS: AOP-RANTES has been produced by a novel total chemical synthesis that provides efficient, direct access to large amounts of this anti-HIV protein analog. The crystal structure of chemically synthesized AOP-RANTES has been solved and refined at 1.6 A resolution. The protein is a dimer, with the amino-terminal pentane oxime moiety clearly defined. CONCLUSIONS: Total chemical synthesis of AOP-RANTES provides a convenient method of producing the multi-milligram quantities of this protein needed to investigate the molecular basis of receptor binding and antiviral activity. This work provides the first truly high-resolution structure of a RANTES protein, although the structure of RANTES was known from previous nuclear magnetic resonance (NMR) determinations.

The herpes simplex virus triplex protein, VP23, exists as a molten globule.

Two proteins, VP19C (50,260 Da) and VP23 (34,268 Da), make up the triplexes which connect adjacent hexons and pentons in the herpes simplex virus type 1 capsid. VP23 was expressed in Escherichia coli and purified to homogeneity by Ni-agarose affinity chromatography. In vitro capsid assembly experiments demonstrated that the purified protein was functionally active. Its physical status was examined by differential scanning calorimetry, ultracentrifugation, size exclusion chromatography, circular dichroism, fluorescence spectroscopy, and 8-anilino-1-naphthalene sulfonate binding studies. These studies established that the bacterially expressed VP23 exhibits properties consistent with its being in a partially folded, molten globule state. We propose that the molten globule represents a functionally relevant intermediate which is necessary to allow VP23 to undergo interaction with VP19C in the process of capsid assembly.