DeepReg: a deep learning hybrid model for predicting transcription factors in eukaryotic and prokaryotic genomes.

Deep learning models (DLMs) have gained importance in predicting, detecting, translating, and classifying a diversity of inputs. In bioinformatics, DLMs have been used to predict protein structures, transcription factor-binding sites, and promoters. In this work, we propose a hybrid model to identify transcription factors (TFs) among prokaryotic and eukaryotic protein sequences, named Deep Regulation (DeepReg) model. Two architectures were used in the DL model: a convolutional neural network (CNN), and a bidirectional long-short-term memory (BiLSTM). DeepReg reached a precision of 0.99, a recall of 0.97, and an F1-score of 0.98. The quality of our predictions, the bias-variance trade-off approach, and the characterization of new TF predictions were evaluated and compared against those produced by DeepTFactor, as well as against experimental data from three model organisms. Predictions based on our DLM tended to exhibit less variance and bias than those from DeepTFactor, thus increasing reliability and decreasing overfitting.

Can Sb(III)-oxidizing prokaryote also oxidize As(III) under aerobic and anaerobic conditions, and vice versa?

Sb(III) and As(III) share similar chemical features and coexist in the environment. However, their oxidase enzymes have completely different sequences and structures. This raises an intriguing question: Could Sb(III)-oxidizing prokaryotes (SOPs) also oxidize As(III), and vice versa? Regarding this issue, previous investigations have yielded unclear, incorrect and even conflicting data. This work aims to address this matter. First, we prepared an enriched population of SOPs that comprises 55 different AnoA genes, lacking AioAB and ArxAB genes. We found that these SOPs can oxidize both Sb(III) and As(III) with comparable capabilities. To further confirm this finding, we isolated three cultivable SOP strains that have AnoA gene, but lack AioAB and ArxAB genes. We observed that they also oxidize both Sb(III) and As(III) under both anaerobic and aerobic conditions. Secondly, we obtained an enriched population of As(III)-oxidizing prokaryotes (AOPs) from As-contaminated soils, which comprises 69 different AioA genes, lacking AnoA gene. We observed that the AOP population has significant As(III)-oxidizing activities, but lack detectable Sb(III)-oxidizing activities under both aerobic and anaerobic conditions. Therefore, we convincingly show that SOPs can oxidize As(III), but AOPs cannot oxidize Sb(III). These findings clarify the previous ambiguities, confusion, errors or contradictions regarding how SOPs and AOPs oxidize each other's substrate.

Intracellular signaling in proto-eukaryotes evolves to alleviate regulatory conflicts of endosymbiosis.

The complex eukaryotic cell resulted from a merger between simpler prokaryotic cells, yet the role of the mitochondrial endosymbiosis with respect to other eukaryotic innovations has remained under dispute. To investigate how the regulatory challenges associated with the endosymbiotic state impacted genome and network evolution during eukaryogenesis, we study a constructive computational model where two simple cells are forced into an obligate endosymbiosis. Across multiple in silico evolutionary replicates, we observe the emergence of different mechanisms for the coordination of host and symbiont cell cycles, stabilizing the endosymbiotic relationship. In most cases, coordination is implicit, without signaling between host and symbiont. Signaling only evolves when there is leakage of regulatory products between host and symbiont. In the fittest evolutionary replicate, the host has taken full control of the symbiont cell cycle through signaling, mimicking the regulatory dominance of the nucleus over the mitochondrion that evolved during eukaryogenesis.

A widespread bacterial protein compartment sequesters and stores elemental sulfur.

Subcellular compartments often serve to store nutrients or sequester labile or toxic compounds. As bacteria mostly do not possess membrane-bound organelles, they often have to rely on protein-based compartments. Encapsulins are one of the most prevalent protein-based compartmentalization strategies found in prokaryotes. Here, we show that desulfurase encapsulins can sequester and store large amounts of crystalline elemental sulfur. We determine the 1.78-angstrom cryo-EM structure of a 24-nanometer desulfurase-loaded encapsulin. Elemental sulfur crystals can be formed inside the encapsulin shell in a desulfurase-dependent manner with l-cysteine as the sulfur donor. Sulfur accumulation can be influenced by the concentration and type of sulfur source in growth medium. The selectively permeable protein shell allows the storage of redox-labile elemental sulfur by excluding cellular reducing agents, while encapsulation substantially improves desulfurase activity and stability. These findings represent an example of a protein compartment able to accumulate and store elemental sulfur.

The missing part: the Archaeoglobus fulgidus Argonaute forms a functional heterodimer with an N-L1-L2 domain protein.

Argonaute (Ago) proteins are present in all three domains of life (bacteria, archaea and eukaryotes). They use small (15-30 nucleotides) oligonucleotide guides to bind complementary nucleic acid targets and are responsible for gene expression regulation, mobile genome element silencing, and defence against viruses or plasmids. According to their domain organization, Agos are divided into long and short Agos. Long Agos found in prokaryotes (long-A and long-B pAgos) and eukaryotes (eAgos) comprise four major functional domains (N, PAZ, MID and PIWI) and two structural linker domains L1 and L2. The majority (∼60%) of pAgos are short pAgos, containing only the MID and inactive PIWI domains. Here we focus on the prokaryotic Argonaute AfAgo from Archaeoglobus fulgidus DSM4304. Although phylogenetically classified as a long-B pAgo, AfAgo contains only MID and catalytically inactive PIWI domains, akin to short pAgos. We show that AfAgo forms a heterodimeric complex with a protein encoded upstream in the same operon, which is a structural equivalent of the N-L1-L2 domains of long pAgos. This complex, structurally equivalent to a long PAZ-less pAgo, outperforms standalone AfAgo in guide RNA-mediated target DNA binding. Our findings provide a missing piece to one of the first and the most studied pAgos.

Membrane fusion and fission during eukaryogenesis.

All eukaryotes can be traced back to a single shared ancestral lineage that emerged from interactions between different prokaryotic cells. Current models of eukaryogenesis describe various selective forces and evolutionary mechanisms that contributed to the formation of eukaryotic cells. Central to this process were significant changes in cellular structure, resulting in the configuration of a new cell type characterized by internal membrane compartments. Additionally, eukaryogenesis results in a life cycle that relies on cell-cell fusion. We discuss the potential roles of proteins involved in remodeling cellular membranes, highlighting two critical stages in the evolution of eukaryotes: the internalization of symbiotic partners and a scenario wherein the emergence of sexual reproduction is linked to a polyploid ancestor generated by cell-cell fusion.

Unraveling the plasticity of translation initiation in prokaryotes: Beyond the invariant Shine-Dalgarno sequence.

Translation initiation in prokaryotes is mainly defined, although not exclusively, by the interaction between the anti-Shine-Dalgarno sequence (antiSD), located at the 3'-terminus of the 16S ribosomal RNA, and a complementary sequence, the ribosome binding site, or Shine-Dalgarno (SD), located upstream of the start codon in prokaryotic mRNAs. The antiSD has a conserved 5'-CCUCC-3' core, but inter-species variations have been found regarding the participation of flanking bases in binding. These variations have been described for certain bacteria and, to a lesser extent, for some archaea. To further analyze these variations, we conducted binding-energy prediction analyses on over 6,400 genomic sequences from both domains. We identified 15 groups of antiSD variants that could be associated with the organisms' phylogenetic origin. Additionally, our findings revealed that certain organisms exhibit variations in the core itself. Importantly, an unaltered core is not necessarily required for the interaction between the 3'-terminus of the rRNA and the region preceding the AUG of the mRNA. In our study, we classified organisms into four distinct categories: i) those possessing a conserved core and demonstrating binding; ii) those with a conserved core but lacking evidence of binding; iii) those exhibiting binding in the absence of a conserved core; and iv) those lacking both a conserved core and evidence of binding. Our results demonstrate the flexibility of organisms in evolving different sequences involved in translation initiation beyond the traditional Shine-Dalgarno sequence. These findings are discussed in terms of the evolution of translation initiation in prokaryotic organisms.

The nuclease-associated short prokaryotic Argonaute system nonspecifically degrades DNA upon activation by target recognition.

Prokaryotic Argonautes (pAgos) play a vital role in host defense by utilizing short nucleic acid guides to recognize and target complementary nucleic acids. Despite being the majority of pAgos, short pAgos have only recently received attention. Short pAgos are often associated with proteins containing an APAZ domain and a nuclease domain including DUF4365, SMEK, or HNH domain. In contrast to long pAgos that specifically cleave the target DNA, our study demonstrates that the short pAgo from Thermocrispum municipal, along with its associated DUF4365-APAZ protein, forms a heterodimeric complex. Upon RNA-guided target DNA recognition, this complex is activated to nonspecifically cleave DNA. Additionally, we found that the TmuRE-Ago complex shows a preference for 5'-OH guide RNA, specifically requires a uridine nucleotide at the 5' end of the guide RNA, and is sensitive to single-nucleotide mismatches between the guide RNA and target DNA. Based on its catalytic properties, our study has established a novel nucleic acid detection method and demonstrated its feasibility. This study not only expands our understanding of the defense mechanism employed by short pAgo systems but also suggests their potential applications in nucleic acid detection.

Catalytically inactive long prokaryotic Argonaute systems employ distinct effectors to confer immunity via abortive infection.

Argonaute proteins (Agos) bind short nucleic acids as guides and are directed by them to recognize target complementary nucleic acids. Diverse prokaryotic Agos (pAgos) play potential functions in microbial defense. The functions and mechanisms of a group of full-length yet catalytically inactive pAgos, long-B pAgos, remain unclear. Here, we show that most long-B pAgos are functionally connected with distinct associated proteins, including nucleases, Sir2-domain-containing proteins and trans-membrane proteins, respectively. The long-B pAgo-nuclease system (BPAN) is activated by guide RNA-directed target DNA recognition and performs collateral DNA degradation in vitro. In vivo, the system mediates genomic DNA degradation after sensing invading plasmid, which kills the infected cells and results in the depletion of the invader from the cell population. Together, the BPAN system provides immunoprotection via abortive infection. Our data also suggest that the defense strategy is employed by other long-B pAgos equipped with distinct associated proteins.

Functionally comparable but evolutionarily distinct nucleotide-targeting effectors help identify conserved paradigms across diverse immune systems.

While nucleic acid-targeting effectors are known to be central to biological conflicts and anti-selfish element immunity, recent findings have revealed immune effectors that target their building blocks and the cellular energy currency-free nucleotides. Through comparative genomics and sequence-structure analysis, we identified several distinct effector domains, which we named Calcineurin-CE, HD-CE, and PRTase-CE. These domains, along with specific versions of the ParB and MazG domains, are widely present in diverse prokaryotic immune systems and are predicted to degrade nucleotides by targeting phosphate or glycosidic linkages. Our findings unveil multiple potential immune systems associated with at least 17 different functional themes featuring these effectors. Some of these systems sense modified DNA/nucleotides from phages or operate downstream of novel enzymes generating signaling nucleotides. We also uncovered a class of systems utilizing HSP90- and HSP70-related modules as analogs of STAND and GTPase domains that are coupled to these nucleotide-targeting- or proteolysis-induced complex-forming effectors. While widespread in bacteria, only a limited subset of nucleotide-targeting effectors was integrated into eukaryotic immune systems, suggesting barriers to interoperability across subcellular contexts. This work establishes nucleotide-degrading effectors as an emerging immune paradigm and traces their origins back to homologous domains in housekeeping systems.

History of Maturation of Prokaryotic Molybdoenzymes-A Personal View.

In prokaryotes, the role of Mo/W enzymes in physiology and bioenergetics is widely recognized. It is worth noting that the most diverse family of Mo/W enzymes is exclusive to prokaryotes, with the probable existence of several of them from the earliest forms of life on Earth. The structural organization of these enzymes, which often include additional redox centers, is as diverse as ever, as is their cellular localization. The most notable observation is the involvement of dedicated chaperones assisting with the assembly and acquisition of the metal centers, including Mo/W-bisPGD, one of the largest organic cofactors in nature. This review seeks to provide a new understanding and a unified model of Mo/W enzyme maturation.

Prokaryotic Argonaute Proteins: A New Frontier in Point-of-Care Viral Diagnostics.

The recent pandemic of SARS-CoV-2 has underscored the critical need for rapid and precise viral detection technologies. Point-of-care (POC) technologies, which offer immediate and accurate testing at or near the site of patient care, have become a cornerstone of modern medicine. Prokaryotic Argonaute proteins (pAgo), proficient in recognizing target RNA or DNA with complementary sequences, have emerged as potential game-changers. pAgo present several advantages over the currently popular CRISPR/Cas systems-based POC diagnostics, including the absence of a PAM sequence requirement, the use of shorter nucleic acid molecules as guides, and a smaller protein size. This review provides a comprehensive overview of pAgo protein detection platforms and critically assesses their potential in the field of viral POC diagnostics. The objective is to catalyze further research and innovation in pAgo nucleic acid detection and diagnostics, ultimately facilitating the creation of enhanced diagnostic tools for clinic viral infections in POC settings.

Type II bacterial toxin-antitoxins: hypotheses, facts, and the newfound plethora of the PezAT system.

Toxin-antitoxin (TA) systems are entities found in the prokaryotic genomes, with eight reported types. Type II, the best characterized, is comprised of two genes organized as an operon. Whereas toxins impair growth, the cognate antitoxin neutralizes its activity. TAs appeared to be involved in plasmid maintenance, persistence, virulence, and defence against bacteriophages. Most Type II toxins target the bacterial translational machinery. They seem to be antecessors of Higher Eukaryotes and Prokaryotes Nucleotide-binding (HEPN) RNases, minimal nucleotidyltransferase domains, or CRISPR-Cas systems. A total of four TAs encoded by Streptococcus pneumoniae, RelBE, YefMYoeB, Phd-Doc, and HicAB, belong to HEPN-RNases. The fifth is represented by PezAT/Epsilon-Zeta. PezT/Zeta toxins phosphorylate the peptidoglycan precursors, thereby blocking cell wall synthesis. We explore the body of knowledge (facts) and hypotheses procured for Type II TAs and analyse the data accumulated on the PezAT family. Bioinformatics analyses showed that homologues of PezT/Zeta toxin are abundantly distributed among 14 bacterial phyla mostly in Proteobacteria (48%), Firmicutes (27%), and Actinobacteria (18%), showing the widespread distribution of this TA. The pezAT locus was found to be mainly chromosomally encoded whereas its homologue, the tripartite omega-epsilon-zeta locus, was found mostly on plasmids. We found several orphan pezT/zeta toxins, unaccompanied by a cognate antitoxin.

Towards unraveling the origins of eukaryotic nuclear genome organization.

With 3D genome mapping maturing over the past decade, studies exposed the differences between eukaryotic and prokaryotic genome organization. This raises the question of how the complex eukaryotic genome organization originated. Here, I explore potential pathways to answering this question, guided by our changing understanding of the origins of eukaryotes.

Sensing preferences for prokaryotic solute binding protein families.

Solute binding proteins (SBPs) are of central physiological relevance for prokaryotes. These proteins present substrates to transporters, but they also stimulate different signal transduction receptors. SBPs form a superfamily of at least 33 protein Pfam families. To assess possible links between SBP sequence and the ligand recognized, we have inspected manually all SBP three-dimensional structures deposited in the protein data bank and retrieved 748 prokaryotic structures that have been solved in complex with bound ligand. These structures were classified into 26 SBP Pfam families. The analysis of the ligands recognized revealed that most families possess a preference for a compound class. There were three families each that bind preferentially saccharides and amino acids. In addition, we identified families that bind preferentially purines, quaternary amines, iron and iron-chelating compounds, oxoanions, bivalent metal ions or phosphates. Phylogenetic analyses suggest convergent evolutionary events that lead to families that bind the same ligand. The functional link between chemotaxis and compound uptake is reflected in similarities in the ligands recognized by SBPs and chemoreceptors. Associating Pfam families with ligand profiles will be of help to design experimental strategies aimed at the identification of ligands for uncharacterized SBPs.

The structural basis of hyperpromiscuity in a core combinatorial network of type II toxin-antitoxin and related phage defense systems.

Toxin-antitoxin (TA) systems are a large group of small genetic modules found in prokaryotes and their mobile genetic elements. Type II TAs are encoded as bicistronic (two-gene) operons that encode two proteins: a toxin and a neutralizing antitoxin. Using our tool NetFlax (standing for Network-FlaGs for toxins and antitoxins), we have performed a large-scale bioinformatic analysis of proteinaceous TAs, revealing interconnected clusters constituting a core network of TA-like gene pairs. To understand the structural basis of toxin neutralization by antitoxins, we have predicted the structures of 3,419 complexes with AlphaFold2. Together with mutagenesis and functional assays, our structural predictions provide insights into the neutralizing mechanism of the hyperpromiscuous Panacea antitoxin domain. In antitoxins composed of standalone Panacea, the domain mediates direct toxin neutralization, while in multidomain antitoxins the neutralization is mediated by other domains, such as PAD1, Phd-C, and ZFD. We hypothesize that Panacea acts as a sensor that regulates TA activation. We have experimentally validated 16 NetFlax TA systems and used domain annotations and metabolic labeling assays to predict their potential mechanisms of toxicity (such as membrane disruption, and inhibition of cell division or protein synthesis) as well as biological functions (such as antiphage defense). We have validated the antiphage activity of a RosmerTA system encoded by Gordonia phage Kita, and used fluorescence microscopy to confirm its predicted membrane-depolarizing activity. The interactive version of the NetFlax TA network that includes structural predictions can be accessed at http://netflax.webflags.se/.

Oligomerization-mediated activation of a short prokaryotic Argonaute.

Although eukaryotic and long prokaryotic Argonaute proteins (pAgos) cleave nucleic acids, some short pAgos lack nuclease activity and hydrolyse NAD(P)+ to induce bacterial cell death1. Here we present a hierarchical activation pathway for SPARTA, a short pAgo consisting of an Argonaute (Ago) protein and TIR-APAZ, an associated protein2. SPARTA progresses through distinct oligomeric forms, including a monomeric apo state, a monomeric RNA-DNA-bound state, two dimeric RNA-DNA-bound states and a tetrameric RNA-DNA-bound active state. These snapshots together identify oligomerization as a mechanistic principle of SPARTA activation. The RNA-DNA-binding channel of apo inactive SPARTA is occupied by an auto-inhibitory motif in TIR-APAZ. After the binding of RNA-DNA, SPARTA transitions from a monomer to a symmetric dimer and then an asymmetric dimer, in which two TIR domains interact through charge and shape complementarity. Next, two dimers assemble into a tetramer with a central TIR cluster responsible for hydrolysing NAD(P)+. In addition, we observe unique features of interactions between SPARTA and RNA-DNA, including competition between the DNA 3' end and the auto-inhibitory motif, interactions between the RNA G2 nucleotide and Ago, and splaying of the RNA-DNA duplex by two loops exclusive to short pAgos. Together, our findings provide a mechanistic basis for the activation of short pAgos, a large section of the Ago superfamily.

Genetically Encoded Crosslinking Enables Identification of Multivalent Ubiquitin-Deubiquitylating Enzyme Interactions.

Ubiquitin (Ub) proteoforms control nearly every aspect of eukaryotic cell biology through their diversity. Inspired by the widely used Ub C-terminal electrophiles (Ub-E), here we report the identification of multivalent binding of Ub with deubiquitylating enzymes (Dubs) using genetic code expansion (GCE) and crosslinking mass spectrometry. While the Ub-Es only gather structural information with the S1 Dub sites, we demonstrate that GCE of Ub with p-benzoyl-L-phenylalanine enables identification of interaction modes beyond the S1 site with a panel of Dubs of both eukaryotic and prokaryotic origin. Collectively, this represents the next generation of Ub-based affinity probes with a unique ability to unravel Ub interaction landscapes beyond what is afforded by cysteine-based chemistries.

Extraordinary long-stem confers resistance of intrinsic terminators to processive antitermination.

Many prokaryotic operons encode a processive antitermination (P-AT) system. Transcription complexes associated with an antitermination factor can bypass multiple transcription termination signals regardless of their sequences. However, to avoid compromising transcriptional regulation of downstream regions, the terminator at the end of the operon needs to be resistant to antitermination. So far, no studies on the mechanism of resistance to antitermination have been reported. The recently discovered conAn P-AT system is composed of two components that are encoded at the start of many conjugation operons on plasmids of Gram-positive bacteria. Here we report the identification of a conAn-resistant terminator, named TerR, in the conjugation operon of the Bacillus subtilis plasmid pLS20, re-defining the end of the conjugation operon. We investigated the various characteristics of TerR and show that its extraordinary long stem is the determining feature for resistance to antitermination. This is the first P-AT resistance mechanism to be reported.

Metabolic compatibility and the rarity of prokaryote endosymbioses.

The evolution of the mitochondria was a significant event that gave rise to the eukaryotic lineage and most large complex life. Central to the origins of the mitochondria was an endosymbiosis between prokaryotes. Yet, despite the potential benefits that can stem from a prokaryotic endosymbiosis, their modern occurrence is exceptionally rare. While many factors may contribute to their rarity, we lack methods for estimating the extent to which they constrain the appearance of a prokaryotic endosymbiosis. Here, we address this knowledge gap by examining the role of metabolic compatibility between a prokaryotic host and endosymbiont. We use genome-scale metabolic flux models from three different collections (AGORA, KBase, and CarveMe) to assess the viability, fitness, and evolvability of potential prokaryotic endosymbioses. We find that while more than half of host-endosymbiont pairings are metabolically viable, the resulting endosymbioses have reduced growth rates compared to their ancestral metabolisms and are unlikely to gain mutations to overcome these fitness differences. In spite of these challenges, we do find that they may be more robust in the face of environmental perturbations at least in comparison with the ancestral host metabolism lineages. Our results provide a critical set of null models and expectations for understanding the forces that shape the structure of prokaryotic life.