Predicted Transmembrane Helices Research Articles

Hiding in plain sight: The discovery of complete genomes of 11 hypothetical spindle-shaped viruses that putatively infect mesophilic ammonia-oxidizing archaea.

The genome of a putative Nitrosopumilaceae virus with a hypothetical spindle-shaped particle morphology was identified in the Yangshan Harbour metavirome from the East China Sea through protein similarity comparison and structure analysis. This discovery was accompanied by a set of 10 geographically dispersed close relatives found in the environmental virus datasets from typical locations of ammonia-oxidizing archaeon distribution. Its host prediction was supported by iPHoP prediction and protein sequence similarity. The structure of the predicted major capsid protein, together with the overall N-glycosylation site, the transmembrane helices prediction, the hydrophilicity profile, and the docking simulation of the major capsid proteins, indicate that these viruses resemble spindle-shaped viruses. It suggests a similarly assembled structure and, consequently, a possibly spindle-shaped morphology of these newly discovered archaeal viruses.

The First Quarter Century of the Dense Alignment Surface Transmembrane Prediction Method.

The dense alignment surface (DAS) transmembrane (TM) prediction method was first published more than 25 years ago. DAS was the one of the earliest tools to discriminate TM proteins from globular ones and to predict the sequence positions of TM helices in proteins with high accuracy from their amino acid sequence alone. The algorithmic improvements that followed in 2002 (DAS-TMfilter) made it one of the best performing tools among those relying on local sequence information for TM prediction. Since then, many more experimental data about membrane proteins (including thousands of 3D structures of membrane proteins) have accumulated but there has been no significant improvement concerning performance in the area of TM helix prediction tools. Here, we report a new implementation of the DAS-TMfilter prediction web server. We reevaluated the performance of the method using a five-times-larger, updated test dataset. We found that the method performs at essentially the same accuracy as the original even without any change to the parametrization of the program despite the much larger dataset. Thus, the approach captures the physico-chemistry of TM helices well, essentially solving this scientific problem.

Enzyme stabilization and thermotolerance function of the intrinsically disordered LEA2 proteins from date palm

In date palm, the LEA2 genes are of abundance with sixty-two members that are nearly all ubiquitous. However, their functions and interactions with potential target molecules are largely unexplored. In this study, five date palm LEA2 genes, PdLEA2.2, PdLEA2.3, PdLEA2.4, PdLEA2.6, and PdLEA2.7 were cloned, sequenced, and three of them, PdLEA2.2, PdLEA2.3, and PdLEA2.4 were functionally characterized for their effects on the thermostability of two distinct enzymes, lactate dehydrogenase (LDH) and β-glucosidase (bglG) in vitro. Overall, PdLEA2.3 and PdLEA2.4 were moderately hydrophilic, PdLEA2.7 was slightly hydrophobic, and PdLEA2.2 and PdLEA2.6 were neither. Sequence and structure prediction indicated the presence of a stretch of hydrophobic residues near the N-terminus that could potentially form a transmembrane helix in PdLEA2.2, PdLEA2.4, PdLEA2.6 and PdLEA2.7. In addition to the transmembrane helix, secondary and tertiary structures prediction showed the presence of a disordered region followed by a stacked β-sheet region in all the PdLEA2 proteins. Moreover, three purified recombinant PdLEA2 proteins were produced in vitro, and their presence in the LDH enzymatic reaction enhanced the activity and reduced the aggregate formation of LDH under the heat stress. In the bglG enzymatic assays, PdLEA2 proteins further displayed their capacity to preserve and stabilize the bglG enzymatic activity.

Coordination of the N-Terminal Heme in the Non-Classical Peroxidase from Escherichia coli.

The non-classical bacterial peroxidase from Escherichia coli, YhjA, is proposed to deal with peroxidative stress in the periplasm when the bacterium is exposed to anoxic environments, defending it from hydrogen peroxide and allowing it to thrive under those conditions. This enzyme has a predicted transmembrane helix and is proposed to receive electrons from the quinol pool in an electron transfer pathway involving two hemes (NT and E) to accomplish the reduction of hydrogen peroxide in the periplasm at the third heme (P). Compared with classical bacterial peroxidases, these enzymes have an additional N-terminal domain binding the NT heme. In the absence of a structure of this protein, several residues (M82, M125 and H134) were mutated to identify the axial ligand of the NT heme. Spectroscopic data demonstrate differences only between the YhjA and YhjA M125A variant. In the YhjA M125A variant, the NT heme is high-spin with a lower reduction potential than in the wild-type. Thermostability was studied by circular dichroism, demonstrating that YhjA M125A is thermodynamically more unstable than YhjA, with a lower TM (43 °C vs. 50 °C). These data also corroborate the structural model of this enzyme. The axial ligand of the NT heme was validated to be M125, and mutation of this residue was proven to affect the spectroscopic, kinetic, and thermodynamic properties of YhjA.

Native structure of mosquito salivary protein uncovers domains relevant to pathogen transmission

Female mosquitoes inject saliva into vertebrate hosts during blood feeding. This process transmits mosquito-borne human pathogens that collectively cause ~1,000,000 deaths/year. Among the most abundant and conserved proteins secreted by female salivary glands is a high-molecular weight protein called salivary gland surface protein 1 (SGS1) that facilitates pathogen transmission, but its mechanism remains elusive. Here, we determine the native structure of SGS1 by the cryoID approach, showing that the 3364 amino-acid protein has a Tc toxin-like Rhs/YD shell, four receptor domains, and a set of C-terminal daisy-chained helices. These helices are partially shielded inside the Rhs/YD shell and poised to transform into predicted transmembrane helices. This transformation, and the numerous receptor domains on the surface of SGS1, are likely key in facilitating sporozoite/arbovirus invasion into the salivary glands and manipulating the host’s immune response.

BoMiProt 2.0: An update of the bovine milk protein database

Milk is a biofluid with various functions, containing carbohydrates, lipids, proteins, vitamins, and minerals. Owing to its importance and availability of vast proteomics information, our research group designed a database for bovine milk proteins (N = 3159) containing the primary and secondary information called BoMiProt. Due to the gaining interest and intensively published literature in the last three years, BoMiProt has been upgraded with newer identified proteins (N = 7459) from peer-reviewed journals, significantly expanding the database from different milk fractions (e.g., whey, fat globule membranes, and exosomes). Additionally, class, architecture, topology, and homology, structural classification of proteins, known and predicted disorder, predicted transmembrane helices, and structures have been included. Each protein entry in the database is thoroughly cross-referenced, including 1392 BoMiProt defined proteins provided with secondary information, such as protein function, biochemical properties, post-translational modifications, significance in milk, domains, fold, AlphaFold predicted models and crystal structures. The proteome data in the database can be retrieved using several search parameters using protein name, accession IDs, and FASTA sequence. Overall, BoMiProt represents an extensive compilation of newer proteins, including structural, functional, and hierarchical information, to help researchers better understand mammary gland pathophysiology, including their potential application in improving the nutritional quality of dairy products.

Exploratory Algorithm of a Multi-epitope-based Subunit Vaccine Candidate Against Cryptosporidium hominis: Reverse Vaccinology-Based Immunoinformatic Approach.

Cryptosporidiosis is the leading protozoan-induced cause of diarrheal illness in children, and it has been linked to childhood mortality, malnutrition, cognitive development, with retardation of growth. Cryptosporidium hominis, the anthroponotically transmitted species within the Cryptosporidium genus, contributes significantly to the global burden of infection, accounting for the majority of clinical cases in numerous nations, as well as its emergence in the last decade is largely due to detections obtained through noteworthy epidemiologic research. Nevertheless, there is no vaccine available, and the only licensed medication, nitazoxanide, has been demonstrated to have efficacy limitations in a number of patient groups recognized to be at high risk of complications. Therefore, current study delineates the computational vaccine design for Cryptosporidium hominis, the notable pathogen for enteric diarrhea. Firstly, a comprehensive literature search was conducted to identify six proteins based on their toxigenicity, allergenicity, antigenicity, and prediction of transmembrane helices to make up a multi-epitope-based subunit vaccine. Following that, antigenic non-toxic HTL epitope, CTL epitope with B cell epitope were predicted from the selected proteins and construct a vaccine candidate with adding an adjuvant and some linkers with immunologically superior epitopes. Afterwards, the constructed vaccine candidates and TLR2 receptor were put into the ClusPro server for molecular dynamic simulation to know the binding stability of the vaccine-TLR2 complex. Following that, Escherichia coli strain K12 was used as a cloning host for the chosen vaccine construct via the JCat server. As a result of the findings, it was resolute that the proposed chimeric peptide vaccine could improve the immune response to Cryptosporidium hominis.

Odorant inhibition in mosquito olfaction mediated by inverse agonists

The insect repellent methyl salicylate elicits excitatory responses upon interaction with CquiOR32, an odorant receptor (OR) from the southern house mosquito, Culex quinquefasciatus. By contrast, eucalyptol binds to CquiOR32 to generate electrophysiological and behavioral inhibitory responses. In an attempt to identify CquiOR32 variants displaying more robust inhibitory responses for more accurate current-voltage analysis, we sequenced 31 CquiOR32 clones. In the Xenopus oocyte recording system, CquiOR32V2/CquiOrco-expressing oocytes yielded eucalyptol-elicited outward (inhibitory) currents relatively larger than methyl salicylate-generated inward (excitatory) currents. Rescuing experiments showed that two of the amino acid substitutions in CquiOR32V2 located in a predicted transmembrane helix of the receptor are determinants of the outward/inward ratios. These findings, along with co-stimulus assays, suggest that odorant and inhibitor may bind to the same binding pocket. Current-voltage relationships obtained with standard perfusion buffer and those devoid of Na+ or Cl− indicated that both excitatory and inhibitory currents are mediated, at least in part, by cation. We then concluded that eucalyptol is an inverse agonist, which shifts the open ⇔ closed equilibrium of the receptor toward the closed conformation, thus reducing the spontaneous activity. By contrast, the binding of methyl salicylate shifts the equilibrium towards the open conformation and, consequently, leads to an increase in cation influx.

Importance of transmembrane helix 4 of l-alanine exporter AlaE in oligomer formation and substrate export activity in Escherichia coli.

AlaE is the smallest amino acid exporter identified in Escherichia coli. It exports l-alanine using the proton motive force and plays a pivotal role in maintaining intracellular l-alanine homeostasis by acting as a safety valve. However, our understanding of the molecular mechanisms of substrate export by AlaE is still limited because structural information is lacking. Due to its small size (149 amino acid residues), it has been speculated that AlaE functions by forming an oligomer. In this study, we performed chemical cross-linking and pull-down assays and showed that AlaE indeed generates homo-oligomers as a functional unit. Previous random mutagenesis experiments identified three loss-of-function AlaE point mutations in the predicted transmembrane helix 4 (TM4) region, two of which are present in the GxxxG motif. When alanine-scanning mutagenesis was applied to the TM4 region, the AlaE derivatives that had amino acid substitutions around the GxxxG motif showed low l-alanine export activities, indicating that the GxxxG motif in TM4 plays an important role in substrate export. However, these AlaE variants with low activity could still form oligomers. We therefore concluded that AlaE forms homo-oligomers and that the GxxxG motif in the TM4 region plays an essential role in AlaE activity but is not involved in AlaE oligomer formation.

A bipartite chromatophore transit peptide and N-terminal protein processing in the Paulinella chromatophore.

The amoeba Paulinella chromatophora contains photosynthetic organelles, termed chromatophores, which evolved independently from plastids in plants and algae. At least one-third of the chromatophore proteome consists of nucleus-encoded (NE) proteins that are imported across the chromatophore double envelope membranes. Chromatophore-targeted proteins exceeding 250 amino acids (aa) carry a conserved N-terminal extension presumably involved in protein targeting, termed the chromatophore transit peptide (crTP). Short imported proteins do not carry discernable targeting signals. To explore whether the import of proteins is accompanied by their N-terminal processing, here we identified N-termini of 208 chromatophore-localized proteins by a mass spectrometry-based approach. Our study revealed extensive N-terminal acetylation and proteolytic processing in both NE and chromatophore-encoded (CE) fractions of the chromatophore proteome. Mature N-termini of 37 crTP-carrying proteins were identified, of which 30 were cleaved in a common processing region. Surprisingly, only the N-terminal ∼50 aa (part 1) become cleaved upon import. This part contains a conserved adaptor protein-1 complex-binding motif known to mediate protein sorting at the trans-Golgi network followed by a predicted transmembrane helix, implying that part 1 anchors the protein co-translationally in the endoplasmic reticulum and mediates trafficking to the chromatophore via the Golgi. The C-terminal part 2 contains conserved secondary structural elements, remains attached to the mature proteins, and might mediate translocation across the chromatophore inner membrane. Short imported proteins remain largely unprocessed. Finally, this work illuminates N-terminal processing of proteins encoded in an evolutionary-early-stage organelle and suggests host-derived posttranslationally acting factors involved in regulation of the CE chromatophore proteome.

Quantitative proteome analysis reveals changes of membrane transport proteins in Sedum plumbizincicola under cadmium stress

Sedum plumbizincicola is an herbaceous species tolerant of excessive cadmium accumulation in above-ground tissues. The implications of membrane proteins, especially integrative membrane proteins, in Cd detoxification of plants have received attention in recent years, but a comprehensive profiling of Cd-responsive membrane proteins from Cd hyperaccumulator plants is lacking. In this study, the membrane proteins of root, stem, and leaf tissues of S. plumbizincicola seedlings treated with Cd solution for 0, 1 or 4 days were analyzed by Tandem Mass Tag (TMT) labeling-based proteome quantification (Data are available via ProteomeXchange with identifier PXD025302). Total 3353 proteins with predicted transmembrane helices were identified and quantified in at least one tissue group. 1667 proteins were defined as DAPs (differentially abundant proteins) using fold change >1.5 with p-values <0.05. The number of DAPs involved in metabolism, transport protein, and signal transduction was significantly increased after exposure to Cd, suggesting that the synthesis and decomposition of organic compounds and the transport of ions were actively involved in the Cd tolerance process. The number of up-regulated transport proteins increased significantly from 1-day exposure to 4-day exposure, from 5 to 112, 16 to 42, 18 to 44, in root, stem, and leaf, respectively. Total 352 Cd-regulated transport proteins were identified, including ABC transporters, ion transport proteins, aquaporins, proton pumps, and organic transport proteins. Heterologous expression of SpABCB28, SpMTP5, SpNRAMP5, and SpHMA2 in yeast and subcellular localization showed the Cd-specific transport activity. The results will enhance our understanding of the molecular mechanism of Cd hypertolerance and hyperaccumulation in S. plumbizincicola and will be benefit for future genetic engineering in phytoremediation.

In silico characterization and structural modeling of bacterial metalloprotease of family M4

BackgroundThe M4 family of metalloproteases is comprised of a large number of zinc-containing metalloproteases. A large number of these enzymes are important virulence factors of pathogenic bacteria and therefore potential drug targets. Whereas some enzymes have potential for biotechnological applications, the M4 family of metalloproteases is known almost exclusively from bacteria. The aim of the study was to identify the structure and properties of M4 metalloprotease proteins. ResultsA total of 31 protein sequences of M4 metalloprotease retrieved from UniProt representing different species of bacteria have been characterized for various physiochemical properties. They were thermostable, hydrophillic protein of a molecular mass ranging from 38 to 66 KDa. Correlation on the basis of both enzymes and respective genes has also been studied by phylogenetic tree. B. cereus M4 metalloprotease (PDB ID: 1NPC) was selected as a representative species for secondary and tertiary structures among the M4 metalloprotease proteins. The secondary structure displaying 11 helices (H1-H11) is involved in 15 helix-helix interactions, while 4 β-sheet motifs composed of 15 β-strands in PDBsum. Possible disulfide bridges were absent in most of the cases. The tertiary structure of B. cereus M4 metalloprotease was validated by QMEAN4 and SAVES server (Ramachandran plot, verify 3D, and ERRAT) which proved the stability, reliability, and consistency of the tertiary structure of the protein. Functional analysis was done in terms of membrane protein topology, disease-causing region prediction, proteolytic cleavage sites prediction, and network generation. Transmembrane helix prediction showed absence of transmembrane helix in protein. Protein-protein interaction networks demonstrated that bacillolysin of B. cereus interacted with ten other proteins in a high confidence score. Five disorder regions were identified. Active sites analysis showed the zinc-binding residues—His-143, His-147, and Glu-167, with Glu-144 acting as the catalytic residues. ConclusionMoreover, this theoretical overview will help researchers to get a details idea about the protein structure and it may also help to design enzymes with desirable characteristics for exploiting them at industrial level or potential drug targets.

Identification of a novel tetracycline resistance gene, tet(63), located on a multiresistance plasmid from Staphylococcus aureus.

To identify and characterize a novel tetracycline resistance gene on a multiresistance plasmid from Staphylococcus aureus SA01 of chicken origin. MICs were determined by broth microdilution according to CLSI recommendations. The whole genome sequence of S. aureus SA01 was determined via Illumina HiSeq and Oxford Nanopore platforms followed by a hybrid assembly. The new tet gene was cloned and expressed in S. aureus. The functionality of the corresponding protein as an efflux pump was tested by efflux pump inhibition assays. A novel tetracycline resistance gene, tet(63), was identified on a plasmid in S. aureus SA01. The cloned tet(63) gene was functionally expressed in S. aureus and shown to confer resistance to tetracycline and doxycycline, and a slightly elevated MIC of minocycline. The tet(63) gene encodes a 459 amino acid efflux protein of the major facilitator superfamily that consists of 14 predicted transmembrane helices. The results of efflux pump inhibitor assays confirmed the function of Tet(63) as an efflux protein. The deduced amino acid sequence of the Tet(63) protein exhibited 73.0% identity to the tetracycline efflux protein Tet(K). The plasmid pSA01-tet, on which tet(63) was located, had a size of 25664 bp and also carried the resistance genes aadD, aacA-aphD and erm(C). A novel tetracycline resistance gene, tet(63), was identified in S. aureus. Its location on a multiresistance plasmid might support the co-selection of tet(63) under the selective pressure imposed by the use of macrolides, lincosamides and aminoglycosides.

MoonProt 3.0: an update of the moonlighting proteins database.

MoonProt 3.0 (http://moonlightingproteins.org) is an updated open-access database storing expert-curated annotations for moonlighting proteins. Moonlighting proteins have two or more physiologically relevant distinct biochemical or biophysical functions performed by a single polypeptide chain. Here, we describe an expansion in the database since our previous report in the Database Issue of Nucleic Acids Research in 2018. For this release, the number of proteins annotated has been expanded to over 500 proteins and dozens of protein annotations have been updated with additional information, including more structures in the Protein Data Bank, compared with version 2.0. The new entries include more examples from humans, plants and archaea, more proteins involved in disease and proteins with different combinations of functions. More kinds of information about the proteins and the species in which they have multiple functions has been added, including CATH and SCOP classification of structure, known and predicted disorder, predicted transmembrane helices, type of organism, relationship of the protein to disease, and relationship of organism to cause of disease.

A predicted transmembrane region in plant diacylglycerol acyltransferase 2 regulates specificity toward very-long-chain acyl-CoAs

Triacylglycerols are the main constituent of seed oil. The specific fatty acid composition of this oil is strongly impacted by the substrate specificities of acyltransferases involved in lipid synthesis, such as the integral membrane enzyme diacylglycerol acyltransferase (DGAT). Two forms of DGAT, DGAT1 and DGAT2, are thought to contribute to the formation of seed oil, and previous characterizations of various DGAT2 enzymes indicate that these often are associated with the incorporation of unusual fatty acids. However, the basis of DGAT2's acyl-donor specificity is not known because of the inherent challenges of predicting structural features of integral membrane enzymes. The recent characterization of DGAT2 enzymes from Brassica napus reveals that DGAT2 enzymes with similar amino acid sequences exhibit starkly contrasting acyl-donor specificities. Here we have designed and biochemically tested a range of chimeric enzymes, substituting parts of these B. napus DGAT2 enzymes with each other, allowing us to pinpoint a region that dramatically affects the specificity toward 22:1-CoA. It may thus be possible to redesign the acyl-donor specificity of DGAT2 enzymes, potentially altering the fatty acid composition of seed oil. Further, the characterization of a DGAT2 chimera between Arabidopsis and B. napus demonstrates that the specificity regulated by this region is transferrable across species. The identified region contains two predicted transmembrane helices that appear to reoccur in a wide range of plant DGAT2 orthologues, suggesting that it is a general feature of plant DGAT2 enzymes.

A Lineage-Specific Paralog of Oma1 Evolved into a Gene Family from Which a Suppressor of Male Sterility-Inducing Mitochondria Emerged in Plants.

Cytoplasmic male sterility (MS) in plants is caused by MS-inducing mitochondria, which have emerged frequently during plant evolution. Nuclear restorer-of-fertility (Rf)genes can suppress their cognate MS-inducing mitochondria. Whereas many Rfs encode a class of RNA-binding protein, the sugar beet (Caryophyllales) Rf encodes a protein resembling Oma1, which is involved in the quality control of mitochondria. In this study, we investigated the molecular evolution of Oma1 homologs in plants. We analyzed 37 plant genomes and concluded that a single copy is the ancestral state in Caryophyllales. Among the sugar beet Oma1 homologs, the orthologous copy is located in a syntenic region that is preserved in Arabidopsis thaliana. The sugar beet Rf is a complex locus consisting of a small Oma1 homolog family (RF-Oma1 family) unique to sugar beet. The gene arrangement in the vicinity of the locus is seen in some but not all Caryophyllalean plants and is absent from Ar. thaliana. This suggests a segmental duplication rather than a whole-genome duplication as the mechanism of RF-Oma1 evolution. Of thirty-seven positively selected codons in RF-Oma1, twenty-six of these sites are located in predicted transmembrane helices. Phylogenetic network analysis indicated that homologous recombination among the RF-Oma1 members played an important role to generate protein activity related to suppression. Together, our data illustrate how an evolutionarily young Rf has emerged from a lineage-specific paralog. Interestingly, several evolutionary features are shared with the RNA-binding protein type Rfs. Hence, the evolution of the sugar beet Rf is representative of Rf evolution in general.

Functional responses between PMP3 small membrane proteins and membrane potential.

The Plasma Membrane Proteolipid 3 (PMP3, UPF0057 family in Uniprot) family consists of abundant small hydrophobic polypeptides with two predicted transmembrane helices. Plant homologues were upregulated in response to drought/salt-stresses and yeast deletion mutants exhibited conditional growth defects. We report here abundant expression of Group I PMP3 homologues (PMP3(i)hs) during normal vegetative growth in both prokaryotic and eukaryotic cells, at a level comparable to housekeeping genes, implicating the regular cellular functions. Expression of eukaryotic PMP3(i)hs was dramatically upregulated in response to membrane potential (Vm) variability (Vmvar ), whereas PMP3(i)hs deletion-knockdown led to Vm changes with conditional growth defects. Bacterial PMP3(i)h yqaE deletion led to a shift of salt sensitivity; Vmvar alternations with exogenous K+ addition downregulated prokaryotic PMP3(i)hs, suggesting [K+ ]-Vmvar axis being a significant feedback element in prokaryotic ionic homeostasis. Remarkably, the eukaryotic homologues functionally suppressed the conditional growth defects in bacterial deletion mutant, demonstrating the conserved cross-kingdom membrane functions by PMP3(i)hs. These data demonstrated a direct reciprocal relationship between PMP3(i)hs expression and Vm differentials in both prokaryotic and eukaryotic cells. Cumulative with PMP3(i)hs ubiquitous abundance, their lipid-binding selectivity and membrane protein colocalization, we propose [PMP3(i)hs]-Vmvar axis as a key element in membrane homeostasis.

The substrate specificity switch FlhB assembles onto the export gate to regulate type three secretion

Protein secretion through type-three secretion systems (T3SS) is critical for motility and virulence of many bacteria. Proteins are transported through an export gate containing three proteins (FliPQR in flagella, SctRST in virulence systems). A fourth essential T3SS protein (FlhB/SctU) functions to “switch” secretion substrate specificity once the growing hook/needle reach their determined length. Here, we present the cryo-electron microscopy structure of an export gate containing the switch protein from a Vibrio flagellar system at 3.2 Å resolution. The structure reveals that FlhB/SctU extends the helical export gate with its four predicted transmembrane helices wrapped around FliPQR/SctRST. The unusual topology of the FlhB/SctU helices creates a loop wrapped around the bottom of the closed export gate. Structure-informed mutagenesis suggests that this loop is critical in gating secretion and we propose that a series of conformational changes in the T3SS trigger opening of the gate through interactions between FlhB/SctU and FliPQR/SctRST.

Topology Prediction Improvement of α-helical Transmembrane Proteins Through Helix-tail Modeling and Multiscale Deep Learning Fusion

Transmembrane proteins (TMPs) play important roles in many biological processes, such as cell recognition and communication. Their structures are crucial for revealing complex functions but are hard to obtain. A variety of computational algorithms have been proposed to fill the gap by predicting structures from primary sequences. In this study, we mainly focus on α-helical TMP and develop a multiscale deep learning pipeline, MemBrain 3.0, to improve topology prediction. This new protocol includes two submodules. The first module is transmembrane helix (TMH) prediction, which features the capability of accurately predicting TMH with the tail part through the incorporation of tail modeling. The prediction engine contains a multiscale deep learning model and a dynamic threshold strategy. The deep learning model is comprised of a small-scale residue-based residual neural network and a large-scale entire-sequence-based residual neural network. Dynamic threshold strategy is designed to binarize the raw prediction scores and solve the under-split problem. The second module is orientation prediction, which consists of a support vector machine (SVM) classifier and a new Max-Min assignment (MMA) strategy. One typical merit of MemBrain 3.0 is the decision mode composed of the dynamic threshold strategy and the MMA strategy, which makes it more effective for hard TMHs, such as half-TMH, back-to-back TMH, and long-TMH. Systematic experiments have demonstrated the efficacy of the new model, which is available at: www.csbio.sjtu.edu.cn/bioinf/MemBrain/.

Crystal structure of heliorhodopsin.

Heliorhodopsins (HeRs) are a family of rhodopsins that was recently discovered using functional metagenomics1. They are widely present in bacteria, archaea, algae and algal viruses2,3. Although HeRs have seven predicted transmembrane helices and an all-trans retinal chromophore as in the type-1 (microbial) rhodopsin, they display less than 15% sequence identity with type-1 and type-2 (animal) rhodopsins. HeRs also exhibit the reverse orientation in the membrane compared with the other rhodopsins. Owing to the lack of structural information, little is known about the overall fold and the photoactivation mechanism of HeRs. Here we present the 2.4-Å-resolution structure of HeR from an uncultured Thermoplasmatales archaeon SG8-52-1 (GenBank sequence ID LSSD01000000). Structural and biophysical analyses reveal the similarities and differences between HeRs and type-1 microbial rhodopsins. The overall fold of HeR is similar to that of bacteriorhodopsin. A linear hydrophobic pocket in HeR accommodates a retinal configuration and isomerization as in the type-1 rhodopsin, although most of the residues constituting the pocket are divergent. Hydrophobic residues fill the space in the extracellular half of HeR, preventing the permeation of protons and ions. The structure reveals an unexpected lateral fenestration above the β-ionone ring of the retinal chromophore, which has a critical role in capturing retinal from environment sources. Our study increases the understanding of the functions of HeRs, and the structural similarity and diversity among the microbial rhodopsins.