Combining protein sequence,structure, and dynamics: A novel approach for functional evolution analysis of PAS domain superfamily

PAS domains are widespread in archaea, bacteria, and eukaryota, and play important roles in various functions. In this study, we aim to explore functional evolutionary relationship among proteins in the PAS domain superfamily in view of the sequence‐structure‐dynamics‐function relationship. We collected protein sequences and crystal structure data from RCSB Protein Data Bank of the PAS domain superfamily belonging to three biological functions (nucleotide binding, photoreceptor activity, and transferase activity). Protein sequences were aligned and then used to select sequence‐conserved residues and build phylogenetic tree. Three‐dimensional structure alignment was also applied to obtain structure‐conserved residues. The protein dynamics were analyzed using elastic network model (ENM) and validated by molecular dynamics (MD) simulation. The result showed that the proteins with same function could be grouped by sequence similarity, and proteins in different functional groups displayed statistically significant difference in their vibrational patterns. Interestingly, in all three functional groups, conserved amino acid residues identified by sequence and structure conservation analysis generally have a lower fluctuation than other residues. In addition, the fluctuation of conserved residues in each biological function group was strongly correlated with the corresponding biological function. This research suggested a direct connection in which the protein sequences were related to various functions through structural dynamics. This is a new attempt to delineate functional evolution of proteins using the integrated information of sequence, structure, and dynamics.     

Expansion of the APC superfamily of secondary carriers

The amino acid‐polyamine‐organoCation (APC) superfamily is the second largest superfamily of secondary carriers currently known. In this study, we establish homology between previously recognized APC superfamily members and proteins of seven new families. These families include the PAAP (Putative Amino Acid Permease), LIVCS (Branched Chain Amino Acid:Cation Symporter), NRAMP (Natural Resistance‐Associated Macrophage Protein), CstA (Carbon starvation A protein), KUP (K⁺ Uptake Permease), BenE (Benzoate:H⁺ Virginia Symporter), and AE (Anion Exchanger). The topology of the well‐characterized human Anion Exchanger 1 (AE1) conforms to a UraA‐like topology of 14 TMSs (12 α‐helical TMSs and 2 mixed coil/helical TMSs). All functionally characterized members of the APC superfamily use cation symport for substrate accumulation except for some members of the AE family which frequently use anion:anion exchange. We show how the different topologies fit into the framework of the common LeuT‐like fold, defined earlier (Proteins. 2014 Feb;82(2):336‐46), and determine that some of the new members contain previously undocumented topological variations. All new entries contain the two 5 or 7 TMS APC superfamily repeat units, sometimes with extra TMSs at the ends, the variations being greatest within the CstA family. New, functionally characterized members transport amino acids, peptides, and inorganic anions or cations. Except for anions, these are typical substrates of established APC superfamily members. Active site TMSs are rich in glycyl residues in variable but conserved constellations. This work expands the APC superfamily and our understanding of its topological variations. Proteins 2014; 82:2797–2811. © 2014 Wiley Periodicals, Inc.     

Grp1‐associated scaffold protein (GRASP) is a regulator of the ADP ribosylation factor 6 (Arf6)‐dependent membrane trafficking pathway

GRASP interacts with Grp1 (g eneral r eceptor for p hosphoinositides 1; cytohesin 3), which catalyses nucleotide exchange on and activation of Arf6 (ADP‐ribosylation factor‐6). Arf6 is a low‐molecular‐mass GTPase that regulates key aspects of endocytic recycling pathways. Overexpressed GRASP accumulated in the juxtanuclear ERC (endocytic recycling compartment). GRASP co‐localized with a constitutively inactive mutant of Arf6 in the ERC such that it was reversed by expression of wild‐type Grp1. Co‐expression of GRASP and Grp1 promoted membrane ruffling, a cellular hallmark of Arf6 activation. GRASP accumulation in ERC was found to block recycling of the MHC‐I (major histocompatibility complex‐I), which is trafficked by the Arf6‐dependent pathway. In contrast, overexpression of GRASP had no effect on the recycling of transferrin receptors, which are trafficked by a clathrin‐dependent pathway. The findings suggest that GRASP regulates the non‐clathrin/Arf6‐dependent, plasma membrane recycling and signalling pathways.     

Structure,function and evolution of the hemerythrin‐like domain superfamily

Hemerythrin‐like proteins have generally been studied for their ability to reversibly bind oxygen through their binuclear nonheme iron centers. However, in recent years, it has become increasingly evident that some members of the hemerythrin‐like superfamily also participate in many other biological processes. For instance, the binuclear nonheme iron site of YtfE, a hemerythrin‐like protein involved in the repair of iron centers in Escherichia coli, catalyzes the reduction of nitric oxide to nitrous oxide, and the human F‐box/LRR‐repeat protein 5, which contains a hemerythrin‐like domain, is involved in intracellular iron homeostasis. Furthermore, structural data on hemerythrin‐like domains from two proteins of unknown function, PF0695 from Pyrococcus furiosus and NMB1532 from Neisseria meningitidis, show that the cation‐binding sites, typical of hemerythrin, can be absent or be occupied by metal ions other than iron. To systematically investigate this functional and structural diversity of the hemerythrin‐like superfamily, we have collected hemerythrin‐like sequences from a database comprising fully sequenced proteomes and generated a cluster map based on their all‐against‐all pairwise sequence similarity. Our results show that the hemerythrin‐like superfamily comprises a large number of protein families which can be classified into three broad groups on the basis of their cation‐coordinating residues: (a) signal‐transduction and oxygen‐carrier hemerythrins (H‐HxxxE‐HxxxH‐HxxxxD); (b) hemerythrin‐like (H‐HxxxE‐H‐HxxxE); and, (c) metazoan F‐box proteins (H‐HExxE‐H‐HxxxE). Interestingly, all but two hemerythrin‐like families exhibit internal sequence and structural symmetry, suggesting that a duplication event may have led to the origin of the hemerythrin domain.     

Geniposide inhibits glucolipotoxicity and cooperates with Txnip knockdown to potentiate cell adaption to endoplasmic reticulum stress in pancreatic beta cells

Thioredoxin‐interacting protein (Txnip), a negative regulator of thioredoxin, has become an attractive therapeutic target to alleviate metabolic diseases. Our previous data demonstrated that geniposide improved glucose‐stimulated insulin secretion by accelerating Txnip degradation and prevented the early‐stage apoptosis of pancreatic β cells induced by palmitate, but the underlying mechanisms are still unclear. The objective of this study is to identify the role of Txnip in geniposide preventing the apoptosis of pancreatic β cells induced by high glucose and palmitate (HG/PA). The results revealed that geniposide attenuated HG/PA‐induced cell apoptosis and the expression of Bax and caspase‐3, while increasing mitochondrial membrane potential and the anti‐apoptotic protein levels of heme‐oxygenase‐1 (HO‐1) and Bcl‐2 in INS‐1 rat pancreatic β cells. Knockdown of the Txnip gene raised the levels of anti‐apoptotic proteins HO‐1 and Bcl‐2 and geniposide potentiated the effect of Txnip when the INS‐1 cells were challenged by HG/PA. Furthermore, geniposide enhanced the adoptive unfolded protein response by increasing the phosphorylation of PERK/eIF2α and IRE1α in HG/PA‐treated INS‐1 cells. The results together suggest that geniposide might be useful to antagonize glucolipotoxicity and Txnip might be a pleiotropic cellular factor in pancreatic β cells.     

Inhibition of MAPK‐mediated ACE expression by compound C66 prevents STZ‐induced diabetic nephropathy

A range of in vitro, experimental and clinical intervention studies have implicated an important role for hyperglycaemia‐induced activation of the renin‐angiotensin system (RAS) in the development and progression of diabetic nephropathy (DN). Blockade of RAS by angiotensin converting enzyme (ACE) inhibitors is an effective strategy in treating diabetic kidney diseases. However, few studies demonstrate the mechanism by which hyperglycaemia up‐regulates the expression of ACE gene. Our previous studies have identified a novel curcumin analogue, (2E,6E)‐2,6‐bis(2‐(trifluoromethyl)benzylidene)cyclohexanone (C66), which could inhibit the high glucose (HG)‐induced phosphorylation of mitogen‐activated protein kinases in mouse macrophages. In this study, we found that the renal protection of C66 in diabetic mice was associated with mitogen‐activated protein kinase (MAPK) inactivation and ACE/angiotensin II (Ang II) down‐regulation. Generally, MAPKs have been considered as a downstream signalling of Ang II and a mediator for Ang II‐induced pathophysiological actions. However, using C66 and specific inhibitors as small molecule probes, in vitro experiments demonstrate that the MAPK signalling pathway regulates ACE expression under HG stimulation, which contributes to renal Ang II activation and the development of DN. This study indicates that C66 is a potential candidate of DN therapeutic agents, and more importantly, that reduction in ACE expression by MAPKs inhibition seems to be an alternative strategy for the treatment of DN.     

Prediction of pKa and redox properties in the thioredoxin superfamily

Electrostatic interactions play important roles in diverse biological phenomena controlling the function of many proteins. Polar molecules can be studied with the FDPB method solving the Poisson-Boltzmann equation on a finite difference grid. A method for the prediction of pK(a)s and redox potentials in the thioredoxin superfamily is introduced. The results are compared with experimental pK(a) data where available, and predictions are made for members lacking such data. Studying CxxC motif variation in the context of different background structures permits analysis of contributions to cysteine DeltapK(a)s. The motif itself and the overall framework regulate pK(a) variation. The reported method includes generation of multiple side-chain rotamers for the CxxC motif and is an effective predictive tool for functional pK(a) variation across the superfamily. Redox potential follows the trend in cysteine pK(a) variation, but some residual discrepancy indicates that a pH-independent factor plays a role in determining redox potentials for at least some members of the superfamily. A possible molecular basis for this feature is discussed.     

Adrenomedullin disulfide bond mimetics uncover structural requirements for AM1 receptor activation

Adrenomedullin (ADM) is a vasoactive peptide hormone of 52 amino acids and belongs to the calcitonin peptide superfamily. Its vasodilative effects are mediated by the interaction with the calcitonin receptor‐like receptor (CLR), a class B G protein‐coupled receptor (GPCR), associated with the receptor activity modifying protein 2 (RAMP2) and functionally described as AM‐1 receptor (AM₁R). A disulfide‐bonded ring structure consisting of six amino acids between Cys¹⁶ and Cys²¹ has been shown to be a key motif for receptor activation. However, the specific structural requirements remain to be elucidated. To investigate the influence of ring size and position of additional functional groups that replace the native disulfide bond, we generated ADM analogs containing thioether, thioacetal, alkane, and lactam bonds between amino acids 16 and 21 by Fmoc/t‐Bu solid phase peptide synthesis. Activity studies of the ADM disulfide bond mimetics (DSBM) revealed a strong impact of structural parameters. Interestingly, an increased ring size was tolerated but the activity of lactam‐based mimetics depended on its position within the bridging structure. Furthermore, we found the thioacetal as well as the thioether‐based mimetics to be well accepted with full AM₁R activity. While a reduced selectivity over the calcitonin gene‐related peptide receptor (CGRPR) was observed for the thioethers, the thioacetal was able to retain a wild–type‐like selectivity profile. The carbon analog in contrast displayed weak antagonistic properties. These results provide insight into the structural requirements for AM₁R activation as well as new possibilities for the development of metabolically stabilized analogs for therapeutic applications of ADM.     

Dissection of the dimerization modes in the DJ-1 superfamily

The DJ-1 superfamily (DJ-1/ThiJ/PfpI superfamily) is distributed across all three kingdoms of life. These proteins are involved in a highly diverse range of cellular functions, including chaperone and protease activity. DJ-1 proteins usually form dimers or hexamers in vivo and show at least four different binding orientations via distinct interface patches. Abnormal oligomerization of human DJ-1 is related to neurodegenerative disorders including Parkinson’s disease, suggesting important functional roles of quaternary structures. However, the quaternary structures of the DJ-1 superfamily have not been extensively studied. Here, we focus on the diverse oligomerization modes among the DJ-1 superfamily proteins and investigate the functional roles of quaternary structures both computationally and experimentally. The oligomerization modes are classified into 4 types (DJ-1, YhbO, Hsp, and YDR types) depending on the distinct interface patches (I-IV) upon dimerization. A unique, rotated interface via patch I is reported, which may potentially be related to higher order oligomerization. In general, the groups based on sequence similarity are consistent with the quaternary structural classes, but their biochemical functions cannot be directly inferred using sequence information alone. The observed phyletic pattern suggests the dynamic nature of quaternary structures in the course of evolution. The amino acid residues at the interfaces tend to show lower mutation rates than those of non-interfacial surfaces.     

The classic signalling and trans‐signalling of interleukin‐6 are both injurious in podocyte under high glucose exposure

Interleukin‐6 (IL‐6) is a multifunctional cytokine that employs IL‐6 classic and trans‐signalling pathways, and these two signal channels execute different or even opposite effects in certain diseases. As a cardinal event of diabetic kidney disease (DKD), whether the podocyte abnormalities are associated with IL‐6 signalling, especially classic or trans‐signalling respectively, remains unclear. In this study, we identified that the circulatory IL‐6, soluble IL‐6R (sIL‐6R) and soluble glycoprotein 130 (sgp130) levels are elevated in patients with DKD. The expressions of membrane‐bound IL‐6R (mIL‐6R), sIL‐6R and gp130 are enhanced in kidney cortex of diabetic mice accompanying with activated STAT3 by tyrosine 705 residue phosphorylation, while not serine 727. Above data infer both classic signalling and trans‐signalling of IL‐6 are activated during DKD. In cultured podocyte, high glucose (HG) up‐regulates the expression of mIL‐6R and gp130, as well as STAT3 tyrosine 705 phosphorylation, in a time‐dependent manner. Entirely blocking IL‐6 signalling by gp130 shRNA, gp130 or IL‐6 neutralizing antibodies attenuates HG‐induced podocyte injury. Interestingly, either inhibiting IL‐6 classic signalling by mIL‐6R shRNA or suppressing its trans‐signalling using sgp130 protein dramatically alleviates HG‐induced podocyte injury, suggesting both IL‐6 classic signalling and trans‐signalling play a detrimental role in HG‐induced podocyte injury. Additionally, activation of IL‐6 classic or trans‐signalling aggravates podocyte damage in vitro. In summary, our observations demonstrate that the activation of either IL‐6 classic or trans‐signalling advances podocyte harming under hyperglycaemia. Thus, suppressing IL‐6 classic and trans‐signalling simultaneously may be more beneficial in podocyte protection and presents a novel therapeutic target for DKD.     

Evolutionary divergence of the APETALA1 and CAULIFLOWER proteins

Abstract APETALA1 (AP1) and CAULIFLOWER (CAL) are a pair of paralogous genes that were generated through the pre‐Brassicaceae whole‐genome duplication event. AP1 and CAL have both partially redundant and unique functions. Previous studies have shown that the K and C regions of their proteins are essential for the functional divergence. However, which differences in these regions are the major contributors and how the differences were accumulated remain unknown. In the present study, we compared the sequences of the two proteins and identified five gaps and 55 amino acid replacements between them. Investigation of genomic sequences further indicated that the differences in the proteins were caused by non‐synonymous substitutions and changes in exon–intron structures. Reconstruction of three‐dimensional structures revealed that the sequence divergence of AP1 and CAL has resulted in differences between the two in terms of the number, length, position and orientation of α‐helices, especially in the K and C regions. Comparisons of sequences and three‐dimensional structures of ancestral proteins with AP1 and CAL suggest that the ancestral AP1 protein experienced fewer changes, whereas the ancestral CAL protein accumulated more changes shortly after gene duplication, relative to their common ancestor. Thereafter, AP1‐like proteins experienced few mutations, whereas CAL‐like proteins were not conserved until the diversification of the Brassicaceae lineage I. This indicates that AP1‐ and CAL‐like proteins evolved asymmetrically after gene duplication. These findings provide new insights into the functional divergence of AP1 and CAL genes.     

Solution structure of the soluble domain of the NfeD protein YuaF from <Emphasis Type="Italic">Bacillus subtilis</Emphasis>

The transmembrane protein YuaF from B. subtilis is a member of the NfeD-like clan with a potential role in maintaining membrane integrity during conditions of cellular stress. nfeD-genes are primarily found in highly conserved operon structures together with the gene of another membrane protein belonging to the SPFH superfamily, in this case YuaG. This strongly suggests a functional if not physical interaction between YuaF and YuaG. Secondary structure predictions of NfeD proteins that accompany SPFH proteins all indicate a high content of beta-sheets in the C-terminal domains indicating a conserved core structure despite very low homology at the level of primary structure. Here we report the high-resolution solution structure of YuaF's soluble C-terminal domain derived from NMR data (sYuaF, residues 97-174 of full-length YuaF). Full backbone and side chain assignments of sYuaF were obtained from triple-resonance spectra. The structure was determined from distance restraints derived from 3D NOESY spectra collected at 600 MHz and 800 MHz, together with phi, psi, and chi(1) torsion angle restraints based on the analysis of (1)H(N), (15)N, (1)H(alpha), (13)C(alpha), (13)CO, and (13)C(beta) chemical shifts, and HNHA, HNHB and HACAHB-COSY spectra. Structures were calculated using CYANA 2.0 and refined in AMBER 8. sYuaF is composed of an extended N-terminal alpha-helix and a beta-barrel formed by five beta-strands. This beta-sheet core structure is well known from the diverse class of OB-fold proteins and can also be found in the distantly related NfeD protein Ph0471 from the archaeon P. horikoshii. Despite significant differences of their amino acid sequences the structural homology of these proteins suggests a conserved function of SPFH-associated NfeD proteins.     

Functional Clustering of Immunoglobulin Superfamily Proteins with Protein–Protein Interaction Information Calibrated Hidden Markov Model Sequence Profiles

Secreted and cell-surface-localized members of the immunoglobulin superfamily (IgSF) play central roles in regulating adaptive and innate immune responses and are prime targets for the development of protein-based therapeutics. An essential activity of the ectodomains of these proteins is the specific recognition of cognate ligands, which are often other members of the IgSF. In this work, we provide functional insight for this important class of proteins through the development of a clustering algorithm that groups together extracellular domains of the IgSF with similar binding preferences. Information from hidden Markov model-based sequence profiles and domain architecture is calibrated against manually curated protein interaction data to define functional families of IgSF proteins. The method is able to assign 82% of the 477 extracellular IgSF protein to a functional family, while the rest are either single proteins with unique function or proteins that could not be assigned with the current technology. The functional clustering of IgSF proteins generates hypotheses regarding the identification of new cognate receptor–ligand pairs and reduces the pool of possible interacting partners to a manageable level for experimental validation.     

Crystal structure of YwpF from Staphylococcus aureus reveals its architecture comprised of a β‐barrel core domain resembling type VI secretion system proteins and a two‐helix pair

The ywpF gene (SAV2097) of the Staphylococcus aureus strain Mu50 encodes the YwpF protein, which may play a role in antibiotic resistance. Here, we report the first crystal structure of the YwpF superfamily from S. aureus at 2.5‐Å resolution. The YwpF structure consists of two regions: an N‐terminal core β‐barrel domain that shows structural similarity to type VI secretion system (T6SS) proteins (e.g., Hcp1, Hcp3, and EvpC) and a C‐terminal two‐helix pair. Although the monomer structure of S. aureus YwpF resembles those of T6SS proteins, the dimer/tetramer model of S. aureus YwpF is distinct from the functionally important hexameric ring of T6SS proteins. We therefore suggest that the S. aureus YwpF may have a different function compared to T6SS proteins. Proteins 2015; 83:781–788. © 2015 Wiley Periodicals, Inc.     

Large-scale prediction of function shift in protein families with a focus on enzymatic function

Protein function shift can be predicted from sequence comparisons, either using positive selection signals or evolutionary rate estimation. None of the methods have been validated on large datasets, however. Here we investigate existing and novel methods for protein function shift prediction, and benchmark the accuracy against a large dataset of proteins with known enzymatic functions. Function change was predicted between subfamilies by identifying two kinds of sites in a multiple sequence alignment: Conservation-Shifting Sites (CSS), which are conserved in two subfamilies using two different amino acid types, and Rate-Shifting Sites (RSS), which have different evolutionary rates in two subfamilies. CSS were predicted by a new entropy-based method, and RSS using the Rate-Shift program. In principle, the more CSS and RSS between two subfamilies, the more likely a function shift between them. A test dataset was built by extracting subfamilies from Pfam with different EC numbers that belong to the same domain family. Subfamilies were generated automatically using a phylogenetic tree-based program, BETE. The dataset comprised 997 subfamily pairs with four or more members per subfamily. We observed a significant increase in CSS and RSS for subfamily comparisons with different EC numbers compared to cases with same EC numbers. The discrimination was better using RSS than CSS, and was more pronounced for larger families. Combining RSS and CSS by discriminant analysis improved classification accuracy to 71%. The method was applied to the Pfam database and the results are available at http://FunShift.cgb.ki.se. A closer examination of some superfamily comparisons showed that single EC numbers sometimes embody distinct functional classes. Hence, the measured accuracy of function shift is underestimated.     

细胞膜蛋白与细胞骨架蛋白相互作用研究进展

细胞膜蛋白与胞浆骨架蛋白的相互作用对于维持细胞正常形态 ,细胞粘附与信号传导有重要作用。含有 4 .1 JEF结构域的蛋白 4 .1超家族与含有PDZ结构域的MAGUK蛋白家族能结合多种膜蛋白胞内区与胞浆蛋白 ,在膜蛋白与胞浆蛋白之间建立联系 ,对于细胞、细胞 -细胞间连接的正常结构与功能的维持有着重要作用。     

Rg1 protects iron‐induced neurotoxicity through antioxidant and iron regulatory proteins in 6‐OHDA‐treated MES23.5 cells

Ginsenoside‐Rg1 is one of the pharmacologically active components isolated from ginseng. It was reported that Rg1 protected dopamine (DA) neurons in 6‐hydroxydopamine (6‐OHDA)‐induced Parkinson's disease (PD) models in vivo and in vitro. Our previous study also demonstrated that iron accumulation was involved in the toxicity of 6‐OHDA. However, whether Rg1 could protect DA neurons against 6‐OHDA toxicity by modulating iron accumulation and iron‐induced oxidative stress is not clear. Therefore, the present study was carried out to elucidate this effect in 6‐OHDA‐treated MES23.5 cells and the possible mechanisms were also conducted. Findings showed Rg1 restored iron‐induced decrease in mitochondrial transmembrane potential in MES23.5 cells, and increased ferrous iron influx was found in 6‐OHDA‐treated cells. Rg1 pretreatment could decrease this iron influx by inhibiting 6‐OHDA‐induced up‐regulation of an iron importer protein divalent metal transporter 1 with iron responsive element (DMT1 + IRE). Furthermore, findings also showed that the effect of Rg1 on DMT1 + IRE expression was due to its inhibition of iron regulatory proteins (IRPs) by its antioxidant effect. These results suggested that the neuroprotective effect of Rg1 against iron toxicity in 6‐OHDA‐treated cells was to decrease the cellular iron accumulation and attenuate the improper up‐regulation of DMT1 + IRE via IRE/IRP system. This provides new insight to understand the pharmacological effects of Rg1 on iron‐induced degeneration of DA neurons. J. Cell. Biochem. 111: 1537–1545, 2010. © 2010 Wiley‐Liss, Inc.     

赤散囊菌MAPKs超家族的鉴定及分析

本研究以赤散囊菌Eurotium rubrum全基因组序列为对象,利用HMMER软件构建隐马尔可夫模型（hidden markov models,HMM）结合BLAST的方法鉴定了促分裂原活化蛋白激酶（mitogen-activated protein kinase,MAPK）超家族。通过构建系统发育树对鉴定蛋白进行分析,并利用MEME软件进行了保守性基序的预测及活性位点注释。分析结果表明,赤散囊菌基因组包含了4个MAPK蛋白,分别属于Hog1-type、MpkC-type、Slt2-type和Fus3/Kss1-type类型;3个MAPK kinase（MAPKK）蛋白,分别属于MKK1-type、Pbs2-type和Ste7-type类型;3个MAPK kinase kinase（MAPKKK）蛋白,分别属于BCK1-type、Ste11-type和Ssk22-type类型。保守性基序分析及注释结果表明,MAPKs超家族蛋白都包含了蛋白激酶活性位点“-D[L/I/V]K-”以及保守性的ATP-binding标签序列。MAPK与MAPKK蛋白分别包含了“-TxY-”和“-SD[I/V]WS-”磷酸化位点,且MAPK蛋白还包含一个保守性的common docking基序（CD motif）,而MAPKKK蛋白则包含了一个功能不明的保守性基序,其一致性序列为“-GTPYWMAPEV-”。研究结果为揭示MAPKs信号途径在赤散囊菌中参与调控的生物学过程奠定了基础。     

Evolutionary Trace Annotation of Protein Function in the Structural Proteome

By design, structural genomics (SG) solves many structures that cannot be assigned function based on homology to known proteins. Alternative function annotation methods are therefore needed and this study focuses on function prediction with three-dimensional (3D) templates: small structural motifs built of just a few functionally critical residues. Although experimentally proven functional residues are scarce, we show here that Evolutionary Trace (ET) rankings of residue importance are sufficient to build 3D templates, match them, and then assign Gene Ontology (GO) functions in enzymes and non-enzymes alike. In a high-specificity mode, this Evolutionary Trace Annotation (ETA) method covered half (53%) of the 2384 annotated SG protein controls. Three-quarters (76%) of predictions were both correct and complete. The positive predictive value for all GO depths (all-depth PPV) was 84%, and it rose to 94% over GO depths 1-3 (depth 3 PPV). In a high-sensitivity mode, coverage rose significantly (84%), while accuracy fell moderately: 68% of predictions were both correct and complete, all-depth PPV was 75%, and depth 3 PPV was 86%. These data concur with prior mutational experiments showing that ET rank information identifies key functional determinants in proteins. In practice, ETA predicted functions in 42% of 3461 unannotated SG proteins. In 529 cases—including 280 non-enzymes and 21 for metal ion ligands—the expected accuracy is 84% at any GO depth and 94% down to GO depth 3, while for the remaining 931 the expected accuracies are 60% and 71%, respectively. Thus, local structural comparisons of evolutionarily important residues can help decipher protein functions to known reliability levels and without prior assumption on functional mechanisms. ETA is available at http://mammoth.bcm.tmc.edu/eta.