Accurate and Fast Prediction of Intrinsically Disordered Protein by Multiple Protein Language Models and Ensemble Learning.

Intrinsically disordered proteins (IDPs) play a vital role in various biological processes and have attracted increasing attention in the past few decades. Predicting IDPs from the primary structures of proteins offers a rapid and facile means of protein analysis without necessitating crystal structures. In particular, machine learning methods have demonstrated their potential in this field. Recently, protein language models (PLMs) are emerging as a promising approach to extracting essential information from protein sequences and have been employed in protein modeling to utilize their advantages of precision and efficiency. In this article, we developed a novel IDP prediction method named IDP-ELM to predict the intrinsically disordered regions (IDRs) as well as their functions including disordered flexible linkers and disordered protein binding. This method utilizes high-dimensional representations extracted from several state-of-the-art PLMs and predicts IDRs by ensemble learning based on bidirectional recurrent neural networks. The performance of the method was evaluated on two independent test data sets from CAID (critical assessment of protein intrinsic disorder prediction) and CAID2, indicating notable improvements in terms of area under the receiver operating characteristic (AUC), Matthew's correlation coefficient (MCC), and F1 score. Moreover, IDP-ELM requires solely protein sequences as inputs and does not entail a time-consuming process of protein profile generation, which is a prerequisite for most existing state-of-the-art methods, enabling an accurate, fast, and convenient tool for proteome-level analysis. The corresponding reproducible source code and model weights are available at https://github.com/xu-shi-jie/idp-elm.

Evolutionary Engineering of a Cp*Rh(III) Complex-Linked Artificial Metalloenzyme with a Chimeric ß-Barrel Protein Scaffold.

Evolutionary engineering of our previously reported Cp*Rh(III)-linked artificial metalloenzyme was performed based on a DNA recombination strategy to improve its catalytic activity toward C(sp2)-H bond functionalization. Improved scaffold design was achieved with α-helical cap domains of fatty acid binding protein (FABP) embedded within the ß-barrel structure of nitrobindin (NB) as a chimeric protein scaffold for the artificial metalloenzyme. After optimization of the amino acid sequence by directed evolution methodology, an engineered variant, designated NBHLH1(Y119A/G149P) with enhanced performance and enhanced stability was obtained. Additional rounds of metalloenzyme evolution provided a Cp*Rh(III)-linked NBHLH1(Y119A/G149P) variant with a >35-fold increase in catalytic efficiency (kcat/KM) for cycloaddition of oxime and alkyne. Kinetic studies and MD simulations revealed that aromatic amino acid residues in the confined active-site form a hydrophobic core which binds to aromatic substrates adjacent to the Cp*Rh(III) complex. The metalloenzyme engineering process based on this DNA recombination strategy will serve as a powerful method for extensive optimization of the active-sites of artificial metalloenzymes.

One-Step Preparation of Fe/N/C Single-Atom Catalysts Containing Fe-N4 Sites from an Iron Complex Precursor with 5,6,7,8-Tetraphenyl-1,12-Diazatriphenylene Ligands.

Fe/N/C single-atom catalysts containing Fe-Nx sites prepared by pyrolysis are promising cathode materials for fuel cells and metal-air batteries due to their high oxygen reduction reaction (ORR) activities. We have developed iron complexes containing N2- or N3-chelating coordination structures with preorganized aromatic rings in a 1,12-diazatriphenylene framework tethering bromo substituents as precursors to precisely construct Fe-N4 sites in an Fe/N/C catalyst. One-step pyrolysis of the iron complex with carbon black forms atomically dispersed Fe-N4 sites without iron aggregates. X-ray absorption spectroscopy (XAS) and electrochemical measurements revealed that the iron complex with N3-coordination is more effectively converted to Fe-N4 sites catalyzing ORR with a TOF value of 0.21 e site-1 s-1 at 0.8 V vs. RHE. This indicates that the formation of Fe-N4 sites is controlled by precise tuning of the chemical structure of the iron complex precursor.

Directed Evolution of a Cp*RhIII -Linked Biohybrid Catalyst Based on a Screening Platform with Affinity Purification.

Directed evolution of Cp*RhIII -linked nitrobindin (NB), a biohybrid catalyst, was performed based on an in vitro screening approach. A key aspect of this effort was the establishment of a high-throughput screening (HTS) platform that involves an affinity purification step employing a starch-agarose resin for a maltose binding protein (MBP) tag. The HTS platform enables efficient preparation of the purified MBP-tagged biohybrid catalysts in a 96-well format and eliminates background influence of the host E. coli cells. Three rounds of directed evolution and screening of more than 4000 clones yielded a Cp*RhIII -linked NB(T98H/L100K/K127E) variant with a 4.9-fold enhanced activity for the cycloaddition of acetophenone oximes with alkynes. It is confirmed that this HTS platform for directed evolution provides an efficient strategy for generating highly active biohybrid catalysts incorporating a synthetic metal cofactor.

Triazolecarbaldehyde Reagents for One-Step N-Terminal Protein Modification.

Site-specific modification of peptides and proteins is a key aspect of protein engineering. We developed a method for modification of the N terminus of proteins using 1H-1,2,3-triazole-4-carbaldehyde (TA4C) derivatives, which can be prepared in one step. The N-terminal specific labeling of bioactive peptides and proteins with the TA4C derivatives proceeds under mild reaction conditions in excellent conversion (angiotensin I: 92 %, ribonuclease A: 90 %). This method enables site-specific conjugation of various functional molecules such as fluorophores, biotin, and polyethylene glycol attached to the triazole ring to the N terminus. Furthermore, a functional molecule modified with a primary amine moiety can be directly converted into a TA4C derivative through a Dimroth rearrangement reaction with 1-(4-nitrophenyl)-1H-1,2,3-triazole-4-carbaldehyde. This method can be used to obtain N-terminal-modified proteins via only two steps: 1) convenient preparation of a TA4C derivative with a functional group and 2) modification of the N terminus of the protein with the TA4C derivative.

Hemoproteins Reconstituted with Artificial Metal Complexes as Biohybrid Catalysts.

In nature, heme cofactor-containing proteins participate not only in electron transfer and O2 storage and transport but also in biosynthesis and degradation. The simplest and representative cofactor, heme b, is bound within the heme pocket via noncovalent interaction in many hemoproteins, suggesting that the cofactor is removable from the protein, leaving a unique cavity. Since the cavity functions as a coordination sphere for heme, it is of particular interest to investigate replacement of native heme with an artificial metal complex, because the substituted metal complex will be stabilized in the heme pocket while providing alternative chemical properties. Thus, cofactor substitution has great potential for engineering of hemoproteins with alternative functions. For these studies, myoglobin has been a focus of our investigations, because it is a well-known oxygen storage hemoprotein. However, the heme pocket of myoglobin has been only arranged for stabilizing the heme-bound dioxygen, so the structure is not suitable for activation of small molecules such as H2O2 and O2 as well as for binding an external substrate. Thus, the conversion of myoglobin to an enzyme-like biocatalyst has presented significant challenges. The results of our investigations have provided useful information for chemists and biologists. Our own efforts to develop functionalized myoglobin have focused on the incorporation of a chemically modified cofactor into apomyoglobin in order to (1) construct an artificial substrate-binding site near the heme pocket, (2) increase cofactor reactivity, or (3) promote a new reaction that has never before been catalyzed by a native heme enzyme. In pursuing these objectives, we first found that myoglobin reconstituted with heme having a chemically modified heme-propionate side chain at the exit of the heme pocket has peroxidase activity with respect to oxidation of phenol derivatives. Our recent investigations have succeeded in enhancing oxidation and oxygenation activities of myoglobin as well as promoting new reactions by reconstitution of myoglobin with new porphyrinoid metal complexes. Incorporation of suitable metal porphyrinoids into the heme pocket has produced artificial enzymes capable of efficiently generating reactive high valent metal-oxo and metallocarbene intermediates to achieve the catalytic hydroxylation of C(sp3)-H bonds and cyclopropanation of olefin molecules, respectively. In other efforts, we have focused on nitrobindin, an NO-binding hemoprotein, because aponitrobindin includes a ß-barrel cavity, which provides a robust structure highly similar to that of the native holoprotein. It was expected that the aponitrobindin would be suitable for development as a protein scaffold for a metal complex. Recently, it was confirmed that several organometallic complexes can bind to this scaffold and function as catalysts promoting hydrogen evolution or C-C bond formation. The hydrophobic ß-barrel structure plays a significant role in substrate binding as well as controlling the stereoselectivity of the reactions. Furthermore, these catalytic activities and stereoselectivities are remarkably improved by mutation-dependent modifications of the cavity structure for the artificial cofactor. This Account demonstrates how apoproteins of hemoproteins can provide useful protein scaffolds for metal complexes. Further development of these concepts will provide a useful strategy for generation of robust and useful artificial metalloenzymes.

Incorporation of a Cp*Rh(III)-dithiophosphate Cofactor with Latent Activity into a Protein Scaffold Generates a Biohybrid Catalyst Promoting C(sp2)-H Bond Functionalization.

A Cp*Rh(III)-dithiophosphate cofactor with "latent" catalytic activity was developed to construct an artificial metalloenzyme representing a new type of biohybrid catalyst which is capable of promoting C(sp2)-H bond functionalization within the ß-barrel structure of nitrobindin (NB). To covalently conjugate the Cp*Rh(III) cofactor into a specific position of the hydrophobic cavity of NB via a maleimide-Cys linkage, strong chelation of the dithiophosphate ligand is employed to protect the rhodium metal center against attack by nucleophilic amino acid residues in the protein. It is found that subsequent addition of the Ag+ ion induces dissociation of the dithiophosphate ligands, thereby activating the catalytic activity of the Cp*Rh(III) cofactor. The resulting "active" biohybrid catalyst promotes cycloaddition of acetophenone oxime with diphenylacetylene via C(sp2)-H bond activation. This catalytic activity is enhanced 2.3-fold with the introduction of two glutamate residues (A100E/L125E) adjacent to the Cp*Rh(III) cofactor. The Cp*Rh(III) cofactor with switchable activity from a "latent" form to an "active" form provides a new strategy for generating biohybrid catalysts incorporating a variety of highly reactive transition metal complexes specifically within its protein scaffolds.

Site-Specific Modification of Proteins through N-Terminal Azide Labeling and a Chelation-Assisted CuAAC Reaction.

Site-specific modification of peptides and proteins is an important method for introducing an artificial function to the protein surface. Recently, we found that new bioconjugation reagents, 6-(azidomethyl)-2-pyridinecarbaldehyde (6AMPC) derivatives, allow specific N-terminal modification and enhance the reaction rate of the subsequent bioconjugation in a chelation-assisted CuAAC reaction. The N-terminal specific azide-labeling of bioactive peptides and proteins occurs under mild reaction conditions with 6AMPC derivatives (angiotensin I: 90%, ribonuclease A: 90%). Kinetic analysis of the CuAAC reaction with azide-labeled proteins reveals that the ligation is promoted in the presence of a copper-chelating pyridine moiety. Importantly, the introduction of an electron-donating methoxy group to the pyridine moiety further accelerates the CuAAC ligation. We demonstrate that this method enables site-specific conjugation of various functional molecules such as fluorophores, biotin, and polyethylene glycol.

A Pyrene-Linked Cavity within a ß-Barrel Protein Promotes an Asymmetric Diels-Alder Reaction.

A unique π-expanded reaction cavity tethering a polycyclic moiety which provides a platform for substrate binding was constructed within the robust ß-barrel structure of nitrobindin (NB). NB variants with cavities of different sizes and shapes are coupled with N-(1-pyrenyl)maleimide (Pyr) to prepare a series of NB-Pyr conjugates. The orientation of the pyrene moiety is fixed within the cavity by the coupling reaction. The fluorescent quenching analysis of NB-Pyr indicates that azachalcone (aza), which is a dienophile for a Diels-Alder (DA) reaction, is efficiently incorporated within the pyrene-linked reaction cavity by the aromatic interaction. The DA reaction between aza and cyclopentadiene proceeds within the reaction cavity of NB-Pyr in the presence of CuII ion in high yield and high enantio- and regioselectivity.

Photocatalytic Properties of TiO2 Composites Immobilized with Gold Nanoparticle Assemblies Using the Streptavidin-Biotin Interaction.

A method using biomolecules to precisely fabricate the morphology of metal nanoparticles immobilized on the surface of a semiconductor using biomolecules is described. A biotin moiety (Biot) is introduced onto the surface of a gold nanoparticle (AuNP) by covalent coupling with α-lipoic acid to assemble AuNPs in the presence of streptavidin (STV). The assembly of Biot-AuNP/STV is immobilized on the surface of TiO2 chemically modified with 1-(3-aminopropyl)silatrane (APS) to provide a positively charged surface. The Au content immobilized on the surface of TiO2 is clearly increased to 9.5 wt % (Au) as a result of the STV-biotin interaction and the electrostatic interaction between negatively charged Biot-AuNPs and the positively charged surface of APS/TiO2. Transmission electron microscopy (TEM) analysis reveals that the composite has an ordered surface geometry in which Biot-AuNPs are spread over the composite surface in two dimensions. The photocatalytic activity toward decomposition of methyl orange dye promoted by this composite is 55%, which is higher than that of the other composites. The Biot-AuNP/STV@APS/TiO2 composite efficiently reduces O2 molecules at Eonset = -0.23 V vs Ag|AgCl, which is more positive than that of other composites (Eonset = -0.40 to -0.32 V). The result suggests that an increased number of AuNPs immobilized in close contact with the TiO2 surface facilitates photoinduced charge transfer. This strategy, which takes advantage of the specific interactions provided by biomolecules and the chemical modification on the surface, has remarkable potential for efficient fabrication of metal nanoparticles on the surface of the semiconductor, which accelerates the reduction of oxygen molecules.

Artificial Diels-Alderase based on the transmembrane protein FhuA.

Copper(I) and copper(II) complexes were covalently linked to an engineered variant of the transmembrane protein Ferric hydroxamate uptake protein component A (FhuA ΔCVF(tev)). Copper(I) was incorporated using an N-heterocyclic carbene (NHC) ligand equipped with a maleimide group on the side arm at the imidazole nitrogen. Copper(II) was attached by coordination to a terpyridyl ligand. The spacer length was varied in the back of the ligand framework. These biohybrid catalysts were shown to be active in the Diels-Alder reaction of a chalcone derivative with cyclopentadiene to preferentially give the endo product.

meso-Dibenzoporphycene has a Large Bathochromic Shift and a Porphycene Framework with an Unusual cis Tautomeric Form.

meso-Monobenzoporphycene (mMBPc) and meso-dibenzoporphycene (mDBPc), in which one or two benzene moieties are fused at ethylene-bridged positions (meso-positions) of porphycene, were prepared in an effort to further delocalize the π-electrons within the porphycene molecule. mMBPc and mDBPc were fully characterized by mass spectrometry, (1)H and (13)C NMR spectroscopy, and X-ray crystallography. The longest-wavelength Q-bands of mMBPc and mDBPc are red-shifted by 92 nm and 418 nm, respectively, compared to that of the unsubstituted porphycene (Pc). Electrochemical measurements indicate that the HOMO is destabilized and the LUMO is stabilized by the fused benzene moieties at the meso positions. Furthermore, both XPS and theoretical studies support the presence of a cis tautomeric form in the ground state of mDBPc, despite the fact that essentially all known porphycene derivatives adopt the trans tautomeric form.

Enzyme-substrate complex structures of CYP154C5 shed light on its mode of highly selective steroid hydroxylation.

CYP154C5 from Nocardia farcinica is a bacterial cytochrome P450 monooxygenase active on steroid molecules. The enzyme has recently been shown to exhibit exclusive regioselectivity and stereoselectivity in the conversion of various pregnans and androstans, yielding 16α-hydroxylated steroid products. This makes the enzyme an attractive candidate for industrial application in steroid hormone synthesis. Here, crystal structures of CYP154C5 in complex with four different steroid molecules were solved at resolutions of up to 1.9 Å. These are the first reported P450 structures from the CYP154 family in complex with a substrate. The active site of CYP154C5 forms a flattened hydrophobic channel with two opposing polar regions, perfectly resembling the size and polarity distribution of the steroids and thus resulting in highly specific steroid binding with Kd values in the range 10-100 nM. Key enzyme-substrate interactions were identified that accounted for the exclusive regioselectivity and stereoselectivity of the enzyme. Additionally, comparison of the four CYP154C5-steroid structures revealed distinct structural differences, explaining the observed variations in kinetic data obtained for this P450 with the steroids pregnenolone, dehydroepiandrosterone, progesterone, androstenedione, testosterone and nandrolone. This will facilitate the generation of variants with improved activity or altered selectivity in the future by means of protein engineering.

Crystal structure, exogenous ligand binding, and redox properties of an engineered diiron active site in a bacterial hemerythrin.

A nonheme diiron active site in a 13 kDa hemerythrin-like domain of the bacterial chemotaxis protein DcrH-Hr contains an oxo bridge, two bridging carboxylate groups from Glu and Asp residues, and five terminally ligated His residues. We created a unique diiron coordination sphere containing five His and three Glu/Asp residues by replacing an Ile residue with Glu in DcrH-Hr. Direct coordination of the carboxylate group of E119 to Fe2 of the diiron site in the I119E variant was confirmed by X-ray crystallography. The substituted Glu is adjacent to an exogenous ligand-accessible tunnel. UV-vis absorption spectra indicate that the additional coordination of E119 inhibits the binding of the exogenous ligands azide and phenol to the diiron site. The extent of azide binding to the diiron site increases at pH ≤ 6, which is ascribed to protonation of the carboxylate ligand of E119. The diferrous state (deoxy form) of the engineered diiron site with the extra Glu residue is found to react more slowly than wild type with O2 to yield the diferric state (met form). The additional coordination of E119 to the diiron site also slows the rate of reduction from the met form. All these processes were found to be pH-dependent, which can be attributed to protonation state and coordination status of the E119 carboxylate. These results demonstrate that modifications of the endogenous coordination sphere can produce significant changes in the ligand binding and redox properties in a prototypical nonheme diiron-carboxylate protein active site.

Fibrous supramolecular hemoprotein assemblies connected with synthetic heme dimer and apohemoprotein dimer.

Supramolecular hemoprotein assemblies via heme-heme pocket interaction were prepared by synthetic heme dimers containing a linker with charged amino acids and apohemoprotein disulfide dimers. The mixture of the negatively charged heme dimer and the apomyoglobin dimer provides heterotropic fibrous hemoprotein assemblies, which were characterized by size-exclusion chromatography (SEC) and atomic force microscopy (AFM).

Photocurrent generation from hierarchical zinc-substituted hemoprotein assemblies immobilized on a gold electrode.

All connected: a protein-immobilized electrode comprising hierarchical assemblies of photoactive cytochrome b(562) reconstituted with zinc protoporphyrin IX exhibits remarkably enhanced photocurrent generation relative to an electrode bearing a single zinc-substituted hemoprotein layer. The protein oligomers, which bear a covalently linked protoporphyrin group, assemble by a supramolecular heme/heme pocket interaction.

Chemically programmed supramolecular assembly of hemoprotein and streptavidin with alternating alignment.

Alternating: a cofactor dyad consisting of a heme group (red in picture) and a bis(biotin) unit (blue) was synthesized and shown to specifically bind to both apomyoglobin and streptavidin. In the presence of the dyad, the 1:1 association of a disulfide-bridged myoglobin dimer (green) with streptavidin (gray) afforded a submicrometer-sized fibrous alternating copolymer.

Crystal structure and spectroscopic studies of a stable mixed-valent state of the hemerythrin-like domain of a bacterial chemotaxis protein.

The bacterial chemotaxis protein of Desulfovibrio vulgaris DcrH (DcrH-Hr) functions as an O(2)-sensing protein. This protein has a hemerythrin-like domain that includes a nonheme diiron center analogous to the diiron center of the hemerythrin (Hr) family. Interestingly, the O(2) affinity of DcrH-Hr is 3.3 × 10(6) M(-1), a value 25-fold higher than that of the Pectinaria gouldii Hr. This high affinity arises from the fast association of the O(2) ligand with DcrH-Hr (k(on) = 5.3 × 10(8) M(-1) s(-1)), which is made possible by a hydrophobic tunnel that accelerates the passage of the O(2) ligand to the diiron site. Furthermore, the autoxidation kinetics indicate that the rate of autoxidation of DcrH-Hr is 54-fold higher than that of P. gouldii Hr, indicating that the oxy form of DcrH-Hr is not stable toward autoxidation. More importantly, a mixed-valent state, semimet(R), which was spectroscopically observed in previous Hr studies, was found to be stable for over 1 week and isolable in the case of DcrH-Hr. The high-resolution crystal structures of the semimet(R)- (1.8 Å) and met-DcrH-Hr (1.4 Å) indicate that the semimet(R)- and met-DcrH-Hr species have very similar coordination geometry at the diiron site.

Construction of a whole-cell biohybrid catalyst using a Cp*Rh(III)-dithiophosphate complex as a precursor of a metal cofactor.

A whole-cell biohybrid catalyst where a (pentamethylcyclopentadienyl)rhodium(III) (Cp*Rh(III)) complex was covalently incorporated into the cavity of nitrobindin (NB), a ß-barrel protein, was prepared on an E. coli cell surface to produce isoquinolines via C(sp2)-H bond activation. In this whole-cell biohybrid system, the Cp*Rh(III)-dithiophosphate complex with latent catalytic activity was utilized as a precursor of the metal cofactor. Strong chelation of the dithiophosphate ligands protects the rhodium complex from being deactivated by abundant nucleophiles in cellular environments during conjugation of the cofactor with the protein scaffold. The whole-cell biohybrid catalyst was then activated upon addition of Ag+ ion to dissociate the dithiophosphate ligands and promoted cycloaddition of acetophenone oxime with diphenylacetylene. Furthermore, the activity of the Cp*Rh(III)-linked whole-cell biohybrid catalyst was enhanced 2.1-fold by introducing glutamate residues at positions adjacent to the Cp*Rh(III) cofactor. These results indicate that the use of the Cp*Rh(III)-dithiophosphate complex with switchable activity from a "latent" form to an "active" form provides a new strategy for generating whole-cell biohybrid catalysts.

Thermally Controlled Construction of Fe-Nx Active Sites on the Edge of a Graphene Nanoribbon for an Electrocatalytic Oxygen Reduction Reaction.

Pyrolytically prepared iron and nitrogen codoped carbon (Fe/N/C) catalysts are promising nonprecious metal electrocatalysts for the oxygen reduction reaction (ORR) in fuel cell applications. Fabrication of the Fe/N/C catalysts with Fe-Nx active sites having precise structures is now required. We developed a strategy for thermally controlled construction of the Fe-Nx structure in Fe/N/C catalysts by applying a bottom-up synthetic methodology based on a N-doped graphene nanoribbon (N-GNR). The preorganized aromatic rings within the precursors assist graphitization during generation of the N-GNR structure with iron-coordinating sites. The Fe/N/C catalyst prepared from the N-GNR precursor, iron ion, and the carbon support Vulcan XC-72R provides a high onset potential of 0.88 V (vs reversible hydrogen electrode (RHE)) and promotes efficient four-electron ORR. X-ray absorption fine structure (XAFS) and X-ray photoelectron spectroscopy (XPS) studies reveal that the N-GNR precursor induces the formation of iron-coordinating nitrogen species during pyrolysis. The details of the graphitization process of the precursor were further investigated by analyzing the precursors pyrolyzed at various temperatures using MgO particles as a sacrificial template, with the results indicating that the graphitized structure was obtained at 700 °C. The preorganized N-GNR precursors and its pyrolysis conditions for graphitization are found to be important factors for generation of the Fe-Nx active sites along with the N-GNR structure in high-performance Fe/N/C catalysts for the ORR.