Investigating the structural changes in amino acids conformations interacting with a toxic effector molecule within the Hcp1 tail/tube complexes of the type VI secretion system using artificial intelligence and deep learning platform

doi:10.21203/rs.3.rs-5231960/v1

Download PDF

Research Article

Investigating the structural changes in amino acids conformations interacting with a toxic effector molecule within the Hcp1 tail/tube complexes of the type VI secretion system using artificial intelligence and deep learning platform

https://doi.org/10.21203/rs.3.rs-5231960/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Background: The primary objective of this study is to understand how a putative toxic effector of the type VI secretion system (T6SS) in Acinetobacter baumannii triggers the contraction of the Hcp1 nanotube through the application of an Artificial Intelligence (AI) and Deep Learning (DL) framework. Furthermore, the virtual assessment of components associated with this secretory system was also investigated.

Methods: The amino acid sequences of T6SS components were retrieved from the GenBank database. AI software such as AlphaFold2, and neural DL software like Rosetta-Fold were used to generate the 3D structures of T6SS protein components. The Rosetta Packer (DLP) program was employed to identify the side chains of the amino acids involved in binding to the effector. A backbone-dependent rotamer library for amino acids side chains was developed based on the Dunbrack rotag package.

Results: Through the machine learning AI system, it has been found that each specific effector molecule binds specifically toa particular set of amino acids (in this case; Lys, Phe, Arg, and His) within the Hcp1 monohexameric ring by H-bound. This interaction induces a rotameric shift in the dihedral angles (Φ/Ψ) of the aforementioned amino acids' side chains launching the contraction of the Hcp1 tail/tube complex and injection effector molecule to the prey cell. Furthermore, TssB/C, TssM, and ClpV ATPase T6SS components are essential for the propulsion of the effector molecule.

Conclusion: This study reveals the mechanism by which the effector molecule induces changes in the conformational of the amino acid side chain from the apo to the hollo state along the Hcp1 tail, resulting in the contraction and propulsion of the substrate into the target cell.

Acinetobacter baumannii

Type VI secretion system

Rotameric shift

Dihedral angel

Phylogenetic tree

AI system.

Acinetobacter baumannii, a hospital-acquired pathogen

A. baumannii is a Gram-negative, opportunistic pathogen commonly found in soil, water, hospital sinks, bed sheets of patients, health care settings, human body skin, and hospital settings [1]. This organism is often associated with nosocomial infections with high mortality and morbidity rates in intensive care units (ICUs) responsible for approximately 2%–10% of all Gram-negative hospital-acquired infections worldwide [1, 2]. The isolates are often collected from the sputum, central nervous system, burn and wound, blood, and urine of patients with prolonged ICU stays [3, 4]. Interestingly, A. baumannii exhibits high resilience to tough environmental conditions and survives on medical devices for extended periods [5]. This durability is attributed to its ability to form biofilm and secrete toxic compounds that can lead to the death of adjacent cells by direct contact [6, 7]. A. baumannii also exhibits high genomic plasticity, adapting to various environmental conditions such as the presence of antibiotics, heavy metal ions, disinfectants, and elevated temperatures [8, 9]. As part of the ESKAPE organisms, A. baumannii poses a global threat to human health and presents a therapeutic challenge due to the emerging, increasing, and spread of antibiotic-resistance genes in Multidrug-resistant A. baumannii (MDR) strains which have been reported worldwide, leading to the high mortality rate in critically ill patients in ICUs [10]. In addition, the emergence and prevalence of carbapenem-resistant (CRAB) MDRAB strains attributed to a significant increase in mortality rates compared to non-CARB-resistant strains in the ICU [11,12]. This organism causes several hospital infections such as ventilator-associated pneumonia (VAP), bloodstream infections (BSI), wound and skin infections, soft tissue, and urinary tract infections in patients with immunocompromised systems or those with underlying conditions [13-15]. Besides, CRAB isolates have been designated as the top priority pathogen requiring urgent management for treatment by the World Health Organization (WHO) [16]. It is now upgraded to a number 1 public health threat by the CDC in the USA (https://www.cdc.gov/drugresistance/pdf/threats-report/2019-ar-threats-report-508.pdf). However, CARB isolates are now problematic across Asia and the Americas, except in Japan and Canada. Moreover, Oceania, Western Europe, the Nordic region, and part of central Europe have the lowest rate (<10%), however, in areas surrounding the Mediterranean, including southern Europe, the Middle East, the Indian subcontinent, and North Africa, A. baumannii clinical isolates that are resistant to carbapenems raised to 90% [17, 18]. Factors contributing to the spread of CARB isolates indicate poor antimicrobial stewardship programs and high administration in management of hospital hygiene and load of antibiotics in hospital settings [19]. Furthermore, the emergence and spread of extensive drug-resistant (XDR) and pan-drug-resistant (PDR) strains of A. baumannii in ICUs worldwide, poses a significant threat as many isolates do not respond to any antibiotic treatment [20]. In light of these challenges, which require a detailed understanding of this organism’s pathobiology, comprehensive strategies for antibiotic administration must be developed to reduce the propensity of infection by MDR strains of this microorganism.

Type VI Secretion System in A. baumannii

Many Gram-negative bacteria can thrive in competitive polymicrobial communities when faced with extreme environmental conditions in the soil, aquatic environments, and healthcare settings [21]. They implement this by injecting toxic effector molecules into neighboring cells, leading to cell death. This behavior is especially prominent among soil microbiota, which compete for vital micronutrients, cationic ions, and growth factors essential for survival in nutritionally deprived ecosystems. As a result, the bacteria have developed various secretory mechanisms to overcome such difficult conditions. One key mechanism is the type VI secretion system (T6SS), first identified in soil-resident bacteria [22]. The T6SS is crucial for the survival of bacterial host cells in unfavorable conditions, allowing them to eliminate adjacent cells through direct contact [23]. This macromolecular nanomachine acts as a syringe and needle to inject toxic effector proteins into neighboring prokaryotic and eukaryotic cells while protecting itself and sister cells with complex cognate immunity genes, typically located upstream or downstream of the corresponding T6SS effector [24]. Four different subtypes of T6SSs have been described in the literature (T6SSi-iv), with variations in the number of conserved components, and low homology among subtypes proteins [25]. It is estimated that 60% of proteobacteria acquired this system by horizontal gene transfer through mobile genetic elements. On the other hand, T6SS structural clusters are present in over 25 % of Gram-negative bacteria, varying in number from one to six different genetic clusters per organism [26]. Many pathogens use the T6SS to escape immune cells, eliminate commensal bacteria, and facilitate infection [27]. T6SS cluster plasticity suggests evolutionarily divergent systems were acquired horizontally by genetic elements like plasmids, transposons, and integrons [28]. Five distinct phylogenetic groups containing T6SS genes have evolved. Despite this, the main component of the T6SS system, particularly the core genes that encode T6SS structural components, remains highly conserved across various bacterial species, indicating a fundamental role in their functionality [29]. Additionally, the system confers advantages for invading eukaryotic cells and functions as a virulence factor, since certain toxic effectors can kill eukaryotic cells by inducing either apoptosis or necrosis, in both cases targeting cells releases cytokines like TNF-α, interleukins, and interferon-ɣ all which cause inflammatory response [30]. For instance, Vibrio cholerae V5 delivers a toxic effector that interferes with the host cytoskeleton of intestinal cells, similarly, Folkesson et al., [31] reported that in the case of Salmonella enterica, a genomic island [Salmonella centrisome island (SCI)] related to the IcmF gene cluster is associated with eukaryotic cell invasion of the epithelial cells and cause death. The T6SS machinery can transport effector proteins in two ways to targeted cells, i) either by fusion with structural components or ii) by non-covalent interaction with one of the core T6SS protein components in both cases, the effector's molecule is associated with the Hcp/VgrG/PAAR structure complex essential for the activity of T6SS (injectosome) [32]. A. baumannii can acquire antibiotic-resistant genes, secretion systems, and various enzymes, including phospholipases, DNase, and peptidoglycan-hydrolyzing amidase genes from other pathogenic bacteria through gene transfer phenomenon [33]. Analysis of the architecture and building block of the T6SS in this bacterium revealed that the core components comprise 13 essential genes that are clustered together occupying about 25% of its total genomic content [34]. Some of the T6SS baseplate components have also been identified as structural homologues of T4 bacteriophage proteins. The similarities between the Hcp, VgrG, and TssE proteins and three subunits associated with the bacteriophage tails (bacteriophage T4 gp19, gp27–gp5, and gp25, respectively) indicate an evolutionary transfer gene of this component to bacteria [35]. Subsequent, evaluation of function using in silico study and Cryo-electron microscopy tomography method indicates that assembly of the T6SS baseplate follows a pathway similar to that of the T4 phage, in contrast, the membrane proteins originate from the bacterium itself [36, 37]. For example, protein-protein interaction studies and protease susceptibility assays indicated that TssM undergoes an ATP binding-induced conformational change and that subsequent ATP hydrolysis is crucial for recruiting Hcp to interact with the periplasmic domain of the TssM-interacting protein TssL (an IcmH/DotU family protein) into a ternary complex and mediating Hcp1 secretion [38]. Normally, the inner Hcp tube assembles onto the base of VgrG and extends into the cytoplasm. A few extended Hcps might carry diverse toxins in their C-termini to serve as T6SS effectors [39]. Furthermore, T6SS-active clinical strains were found to survive better in the presence of human serum and were more frequently detected in patients with catheter-related bloodstream infection, hematopoietic stem cell transplantation, and immunosuppressive agent therapy [33]. Higher hcp expression was found to increase the invasiveness of A. baumannii under respiratory infection and could be triggered by the acid environment [33]. In Acinetobacter baylyi, this structure has a syringe-like shape and is often used by bacteria to deliver toxic substances into the other cells. The effectors may constitute extensions of any of the secreted components Hcp, VgrG, or PAAR, or bind non-covalently to these in terms of cargo’ effectors [40]. The process is a one-step mechanism where bacterial cytoplasmic substrates are conveyed directly into a target cell or to the extracellular space. The operational mechanism of the T6SS macromolecule is divided into three stages: 1) assembly of components to form the active T6SS complex (syringe and needle structure), 2) delivery phase, where the effector molecule triggers the contraction of Hcp1 with the help of VgrG1(gp27) contractile sheath and TssB/C, resulting in the propelling of the effector molecule through the Hcp1 nanotube towards the targeted cell, (the energy for this function obtains from hydrolysis of ATP molecule by ClpV-ATPase cleavage, and 3) disassembly phase, during which the toxin effector is released into target cell led to cell death through various mechanisms [41]. The sequence analysis of the VgrG, Hcp, and PAAR proteins by Boyer et al., [42] confirmed that some C-terminal regions possess domains for binding effector molecules showing similarities with adhesins or proteins with enzymatic functions. In A. baumannii, the T6SS system is negatively regulated by the acquisition and loss of plasmid DNA containing the regulatory genes characterized by three conserved regions: a locus encoding the type IV secretion system (T4SS) conjugative machinery; a region encoding two TetR transcriptional regulators for example, conjugative plasmid pAB3 produce proteins within the H-NS family control the expression of Hcp1 nanotube [43].

Rotameric shift and conformational change in protein structure

A critical aspect of forecasting protein structure and its interactions with various compounds, including toxic agents or peptide ligands, is the evaluation of side chain flexibility [44]. A rotameric shift is characterized by a shift in the dihedral angles (ϕ, ψ) of an amino acid's side chain, either clockwise or counterclockwise, leading to a conformational change in that amino acid configuration. Each amino acid possesses a unique set of rotamers, and a comprehensive collection of these rotamers for all amino acids can be organized into libraries [45]. The rotamericity of an amino acid is fundamentally influenced by the diverse environments encountered in actual protein structures. The elements such as the torsional angles of the backbone residue and the protein's secondary structure play a role in altering the dihedral angles of the amino acid side chain [46]. It has been proposed that torsional angles can either directly influence the dihedral rotation of the backbone (φ between Cα and N, ψ between Cα and Cω, and between C and N) or indirectly impact the side chains of amino acids (χ1 between Cα and Cβ, χ2 between Cβ and Cγ) [47]. Rotamers can be classified as backbone-independent, secondary-structure-dependent, or backbone-dependent, with distinctions made based on dihedral angles and conformations. Backbone-dependent predictions result in rotameric shifts in both the backbone (χ1/χ2, and Cγ) and the dihedral angles (ϕ, ψ) of amino acids, whereas, backbone-independent rotamer libraries reflect the frequencies and average dihedral angles for all side chains in proteins, irrespective of each residues conformation [48]. Research indicates that backbone-dependent rotamer libraries offer considerable advantages over their backbone-independent counterparts, particularly when utilized as an energy term, as they enhance the efficiency of side-chain packing algorithms employed in protein structure prediction [49]. Moreover, the condition for matching side chain conformations must be the same as that used for the evaluation next to the last rotamer library where a correct position must lie within 40° in each χ dihedral [50]. Usually, a rotational angle is in the apo form or Holo form (bound state). The rotamer frequencies are primarily due to steric repulsions between backbone atoms Cγ whose position depends on φ and ψ side chain heavy atoms. Moreover, the φ and ψ angles of amino acids can be determined by the Ramachandran plot [50].

The purpose of this study is to exploit the mechanism of contraction of the Hcp1 tail/tube complexes upon binding to a specific toxic effector molecule in the A. baumannii by employing machine learning techniques such as AI and Deep learning systems along with bioinformatics databases. The study also aims to analyze the structures, functions, and domain analysis attributes of A. baumannii T6SS, as well as to examine the phylogenetic tree relationships, and to classify Hcp and VrgG1 of T6SS components.

Analysis of the structure and function of T6SS in A. baumannii

The amino acid sequence accession numbers AOX94828.1, WP_000556919.1, WP_000653195.1, WLC49609.1, AKR15105.1, AKR15106.1, PQI80388.1, BDI26490.1, and AKR15119.1 were initially retrieved from the GenBank (NBI) database and used as query sequences to identify the constituents of T6SS proteins. Subsequently, the sequences were confirmed for the prediction of the molecular dynamic structure of each T6SS component by homology modeling using platforms like the SWISS-MODEL (https://swissmodel.expasy.org/), I-TASSER (https://zhanggroup.org/I-TASSER/), and Deep learning software (https://geekflare.com/deep-learning-software).

A diffusion modeling program was instigated to generate protein structures based on Rosetta-Fold (https://www.rosettacommons.org/software). For analysis of domain structures, Rosetta Packer was employed (https://www.rosettacommons.org/software). The system uses a unique three-track neural network and integrates three types of information: the sequential patterns in proteins, the interplay charges between amino acids, and the probable 3D configurations. Rosetta-Packer software was also used for the detection of any change in the side chain of amino acids and conformational isomerism. Quality controls for the accuracy of predicted models were assessed using Z-Score, RMSD index, B-factor, melting point, MolProbity, and Ramachandran plot. The visualization of T6SS components' structures was carried out using PyMOL through the Schrödinger platform (https://www.schrodinger.com/products/pymol). The AlphaFold2 (AF2) model quality and reliability were evaluated based on parameters such as predicted local distance difference test (pLDDT) and predicted aligned error (PAE) values [51]. The per-residue pLDDT score implies the indicated quality of a protein, ranging from 0 to 100, with higher scores indicating higher quality of the predictions [52]. Residues with pLDDT scores above 90 are considered reliable, 70–90 scores are less reliable, and below 50 scores indicate low quality. PAE measures the average distance between the predicted and true residue positions after aligning both structures using a variant of the Kabsch algorithm [53]. The PAE is useful for assessing confidence between domains or chains. The mechanism of the AF2 model system was divided into three main parts: model input, Evoformer, and structure module [54]. In this module, AF2 takes the inputs from the first module and passes them through a deep learning module (called Evoformer). The Evoformer estimates the distance between residues and processed multi-sequence alignment representing faster and fewer than simpler pairformer modules. Structure Module estimates a 3D structure based on the last updated output representations from the Evoformer. The Evoformer gives more flexibility to the neural network, allowing it to dynamically learn interactions between non-neighboring nodes in the AI system. In this manner, the AI system can learn what relations are relevant and which can be ignored to focus on computing the graph of how each residue relates to another [52].

Bioinformatics analysis of T6SS in A. baumannii

The TOPCONS web server (https://topcons.net/) was utilized for consensus prediction of membrane protein topology (https://dtu.biolib.com/DeepTMHMM). This server has successfully differentiated between globular and membrane proteins. The phylogenetic tree relationship between each component of the type VI secretion system was investigated using the BioAlign.IO package BioPython 1.75 documentation (https://biopython.org/docs/1.75/api/Bio.AlignIO.html). The functional analysis of protein was also conducted by classifying them into superfamily, family, and predicting domains with the help of the EMBL-EBI port InterPro/Pfam database (https://www.ebi.ac.uk/interpro/). Additionally, the hierarchical relationship among the proteins involved in the T6SS functions was confirmed using the Class Architecture Topology Homologous Superfamily (CATH) database (https://www.cathdb.info/). The CATH provides an evolutionary classification of proteins and domains based on their sequence similarity versus structural diversity among superfamilies. The molecular structure of the entire T6SS system in A. baumannii was predicted by machine learning system.

Rotameric shift and stereoisomerism

The important part of this study was to determine how the toxic effector propels from the inner membrane to the outer membrane injectosome (VgrG1 + PAAR+ Hcp1). To answer this question, the open-source software package rotag was used to create a rotameric library using https://www.github.com/agrybauskas/rotag site. Here, to confirm the rotameric shifts of dihedral angles of the staggered amino acids within the Hcp1 monohexameric ring through binding to the specific effector molecule, various strategies have been employed, including the use of rotamer libraries, minimizing free energy (ΔG), and utilizing the GROMACS software (http://www.gromacs.org/) to create an index file with preferred dihedral angles.

The rotameric libraries are used to predict, build, design, and solve new protein structures. Here, predicted amino acid dihedral angle (Φ/ψ) side chains and backbone (χ1/ χ 2, and Cγ) were assessed for the possibility of rotation and changes in amino acid angles, and protein conformation. The effect of backbone conformation on side chain rotamer frequencies is primarily due to steric repulsions between backbone atoms Cγ whose position depends on φ and ψ side chain heavy atoms [55]. One of the greatest advantages of the rotag is its ability to deal with nonstandard amino acids. There are only a handful of structures with selenocysteine and pyrrolysine in the PDB which precludes the collection of extensive statistics. This is, however, not a problem for rotag since its method does not require pre-calculated averaged dihedral angle datasets. In this study, the Ramachandran plot was also used to check the φ, ψ rotameric interactions of the amino acids side-chain. The current rotameric library and residue data are available on the Dunbrack website (http://dunbrack.fccc.edu/bbdep/). Details of the rotamer library can be downloaded from http://www.dynameomics.org site.

Phylogenetic relation of the T6SS components in A. baumannii

Furthermore, the phylogenetic tree analysis of individual T6SS components was conducted using the BioPython 1.58 Align package (https://biopython.org). Here, the AlignIO and AlignInfo modules were imported. The AlignIO module provides functions to read and write sequence alignments in various formats, while the AlignInfo module offers classes to summarize and analyze alignment information. It is important to note the AlignIO read always the alignment file in Clustal format.

Components of the Type VI secretion system in A. baumannii

Inner membrane subcomplexes

The first initial subcomplexes of the T6SS in the A. baumannii are the inner membrane base plate complexes comprising the proteins TssK, TssF, TssG, and TssE. Figure 1A, B, C, and D illustrate stereochemical predictions of the inner membrane baseplate complexes of A. baumannii T6SS by artificial intelligence. Structurally, TssK is identified as a key component of the inner membrane structure (Fig. 1A), it is a trimeric cytoplasmic protein that shares structural features with receptor-binding proteins found in T4 phage. This protein is integral to the binding of membrane complex proteins TssL and TssM. Analyses conducted through AF2 indicate that the three distinct domains of TssK facilitate its interaction with other T6SS systems (Fig. 1A). It is a central component of the type VI secretory system in A. baumannii situated in the base plate region of inner membrane complexes composed of three domains, the N-terminal region of TssK comprises several short β-strands inter-connected by minor loops, playing a crucial role in the oligomerization of Hcp1. The second domain encompasses an α-helical core formed by four elongated transmembrane helices (α1, α2, α3, and α4) that are arranged in the form of a helix-turn-helix configuration, linking the baseplate to outer membrane constituents. The third domain is characterized by the presence of six β-sheets (β1 through β6) that engage with the inner membrane, and contribute to the stability of the baseplate complex (Fig. 1A). The cartoon structure of the second protein, TssF indicates that it is a trimeric protein (Fig. 1B) composed of three domains; one situated at the C-terminal consisting of four β-strands containing 460–587 amino acid residues attached to one α-helix via a hydrophobic core. The second domain contains a central alpha-helical core made up of 45–85 amino acid residues that help in the correct conformation of the protein structure, and the third domain consists of three β-sheets (90–457 residues) attached to the inner membrane assembly. The TssF is structurally homologous to phage T4 receptor-binding protein IcmH (Fig. 1B). The third protein, TssG (Fig. 1C) is an essential component of the baseplate complexes. The C-terminal region of this protein acts as an adaptor that interacts with both TssF and TssK proteins. It adopts a folded chain similar to the TssF protein although these two proteins do not share detectable similarity. In A. baumannii, two extended loops were detected protruding from opposite ends of TssG (loop1 and loop 2) anchoring baseplate components to the cytoplasmic membrane and aiding in the proper folding of TssG (Fig. 1C). The analysis of the 3D structure of the fourth protein TssE by the AF2 revealed a small single-domain protein comprising four β-sheets at the C-terminal region. This domain facilitated the assembly of other baseplate components and conferred a binding site for the TssB protein (Fig. 1D).

The membrane sub-complex

The baseplate is attached to the outer membrane via a complex assembled of three proteins namely TssM, TssL, and TssJ, however, a later protein binds to the cell membrane lipoprotein. They contain a peptidoglycan-binding domain [56]. Researchers by Cryo-electron tomography along with fluorescence microscopy discovered that TssM and TssL have strong homology with the phage proteins IcmF, and IcmH (DotU), respectively [57].

TssM is a multi-component protein containing three conserved domains that carry different functions. The first domain of TssM contains 8 β-sheets at the C-terminal connected via two hairpins mainly β1 and β2 loops (Fig. 2A). Previous works have indirectly demonstrated that the C-terminal domain of TssM could be at least transiently exposed to the extracellular environment, where it functions as a prey cell detector [58] and facilitates interaction with inner membrane protein by long transmembrane helix at the periplasmic region near C-terminal (Fig. 2A). These long transmembrane α-helices were also detected by TOPCON software and function to stabilize the helices themselves. The third domain is anchored to the inner membrane components by several transmembrane helices (Fig. 2A). Likewise, the protein TssL domain-1 is bound to the outer membrane complex via a single periplasmic α-helix at the C-terminal region, while the second domain consists of 9 α-helices (32–196 amino acids) anchored to the cytoplasmic membrane via the N-terminal region (Fig. 2B). It has been shown that two amino acid residues of TssL in A. baumannii, Asp98, and Glu99, are strongly conserved among T6SS-encoding strains, and remarkably impact the dynamics, expression, and performance of this protein [59]. Finally, the TssJ protein structure illustrated in Fig. 2C is composed of 8 α-helices in the N-terminal run side by side connected by a long α-helix at the C-terminal region to the lipoprotein of the bacterial cytoplasmic membrane of the targeted cell.

Tail tube/sheath sub-complex

The T6SS in A. baumannii contain a third sub-complex containing a tail tube/sheath composed of a TssB-C/ Hcp1 nanotube (TssD) complex, with three VgrG1 (gp5) (or TssI) protein on top, and sometimes a single PAAR protein [46]. The TssB (Fig. 3A) consists of a central domain with a mix of α-helices and β-sheets, predicted with high confidence (PLDD accuracy > 90%) by AlphaFold. This domain plays a crucial role in the contraction of HcP1 tail phage-like and VgrG1 (gp27). The TSSB/C complex wraps around the Hcp1 nanotube, contributing to the stability of the complex. Hcp1, VgrG, and PAAR proteins (part of the injectosome) bind to an antibacterial effector and are then injected into a target cell (Fig. 3B). The structures of Hcp1 and VgrG1 have the C-terminal domain assembled to effector form a needle-like spike. The schematic overview of a monohexameric ring of Hcp1 C (Fig. 3C) also demonstrates that inside of the Hcp1 tube, several β-sheet arranged side by side with several loops connected to a few α-helices. Nevertheless, the PAAR protein situated at the tip of the VrgG1 (Fig. 3B) connected mostly with Hcp1 contraction helps inject the effector into the prey cells.

Meanwhile, the schematic representation of the whole T6SS system predicted by AI and deep learning in this study as well as the base plate along the hcp1 tail is depicted in Fig. 4A and Fig. 4B.

Most effectors like Tle1AH have activity against mammalian cells or lysozyme-like effectors (TseP), and type VI secretion NADase effector family 2 (Tnf2) molecules against bacterial cells [60]. Noreen et al. reported that an Arg-to-Ala mutation at position 30 within the extended Hcp1 resulted in a significant decrease in cytotoxicity, suggesting mutations of Hcp1 shut down the functioning of all T6SS components [61]. This protein encompasses both side chain and backbone, along with ATP hydrolysis. To find the stereochemical model of monohexameric Hcp1 with high confidence (PLDD > 90%), we used the AF2 platform in the PyMOL for the graphic analysis of Hcp1 (Fig. 5A). The ring structure exhibited a tight β-barrel connected to a short-turn helix. The β-barrel domain was (~ 12 Å diameter) formed by eight strands of β sheets (β1–β8). Furthermore, Fig. 5B shows a ball and stick model of the Hcp1 ring. The accuracy of folding the alpha helices and beta sheets was estimated by angstrom error plot using the Rosetta-Fold program (Fig. 5C). The amino acids alignment of the Hcp1 of the query sequence with 5 closely related A. baumannii strains in the UniProtKB database is shown in Fig. 5D.

The results showed ≥ 99.0% identities in the sequences. This indicated that strains of A. baumannii Hcp1 originated from a common ancestor during evolution, though there are high variations in sequences among superfamilies. The cartoon structure generated monohexameric rings based on the AI system is displayed in Fig. 6A. The local quality of the Hcp1 protein using the Expasy Swiss platform (Fig. 6B) was used to check the accuracy of domain chain α, β, and αβ mixed strands as well as amino acid angles, Ramachandran plot, MolProbity = 1.10, clash score = 0.71, Ramachandran favor = 94.51, cβ-factor = 0.37, and Z-score = 2.37 were also carried out in this study to confirm shift in the angle of the amino acids (Fig. 6C).

Importance of this study

The molecular mechanism of propelling the effector molecule from the inner membrane to the outer membrane and finally to the prey cell is shown in Fig. 7A and Fig. 7B. The amino acids bind to the effector molecule, and the H-bond formation is predicted in Fig. 7A. Analysis of the amino acids' rotational angle both for the backbone (e.g., φ between Cα and N, ψ between Cα and Cω between C and N) and the side chain (e.g., χ1 between Cα and Cβ, χ2 between Cβ and Cγ) indicate the effector molecule induced rotameric shift in four amino acids within Hcp1 ring. The estimated changes detected by software from the rotamer library were as follows [for Lys (11–13Ǻ), His (18Ǻ), Arg (9–8.3Ǻ), and Phe (9–12Ǻ)], respectively (Fig. 7B). The change in conformation of the above amino acids by toxic effector causes the contraction of the Hcp1 tail/tube complex with the help of phage like VgrG1(gp27) contractile sheath, ClpV ATPase and TSSM and transport it to the head of T6SS (injectosome) where, VgrG1, Hcp1, and PAAR proteins have effector binding domains. The contracted sheath is then detached and disassembled by ClpV ATPase and injecting the Hcp1, VgrG1, and PAAR proteins into the recipient cell, this process is recycled into a new extended T6SS apparatus at either the original or a newly formed base plate complex.

Classification of Hcp and VgrG protein families

InterPro/Pfam software provides functional analysis of protein prediction by classifying them into superfamily, family, and predicting domains. The graphic classification summary of the Hcp1 in A. baumannii is shown in Fig. 8 to predict functional domains and relative classification of Hcp1, and VgrG1. The entry matches representative three domains include, T6SS_Hcp1 (Pfam PF05638), VI_effect_T6SS (TIGR03344), and HIS-I_assembly_Hcp1 (IPR053165), and one homologous superfamily Hcp1_like_SF (IPR053165) (Fig. 8).

Furthermore, using the amino acids percent homology matrix and CATH database, it was confirmed that, a strong evolutionary relationship existed among the Hcp1 family in the A. baumannii isolates as shown in Fig. 9A. The protein matrix homology assay confirmed close similarity among the Hcp family in A. baumannii with identity > 90% (Fig. 9B).

This was supported by phylogenetic tree analysis of Hcp as showing > 90% identity among the Hcp1 amino acid sequence of different isolates of this bacterium (Fig. 10).

In the case of VgrG1, the blast parameters including identity, score, and E-value are shown in Fig. 11A. To check the percent identity among amino acid sequences of VgrG1, 5 closely related strains of A. baumannii were selected, the alignment was performed using the UniProtKB database and the results are demonstrated in Fig. 11B. The results indicate 99% identity among the VgrG1 sequence in 5 closely related strains of A. baumannii. In addition, to determine the error of the predicted distance (PAE) for each pair of residues in both axes indicate the position of the individual amino acids. The uncertainty in the distance of two amino acids is shown in Fig. 11C. In the present study, no errors were observed in the alignment, particularly in the DUF2345 subunit.

Meanwhile, to estimate the number of domains and amino acid contents of VgrG1, the InterPro/Pfam software (Fig. 12) was employed. The protein consists of three main domains: i) VgrG (COG3501), which forms the tip exhibited needle-like structure and has effector binding activity, an uncharacterized conserved protein implicated phage assembly with AF2 PLLD (≥ 90), ii) Phage GDP, this family includes phage late control gene D protein (GDP) and related bacterial sequences, iii) The DUF 2345 superfamily is a member of the Rhs element (Vgr_Rhs_VI) and plays an important role in the injection of the VgrG1 needle-like tips into prey cells (Fig. 12).

The phylogenetic tree analysis

The dendrogram analyses of VgrG1, TssM, and TssG showed high sequence similarity versus structural identity (≤ 90%), among the main components of T6SS, respectively (Fig. 13A, 13B, 13C). The results indicate that each protein of T6SS has an identical ancestor as they have a close evolutionary relationship. This hypothesis was further proved by identity in the amino acid sequence residues with other strains of A. baumannii. Similarly, Fig. 13A shows the phylogenetic tree for Hcp1, Fig. 13B illustrates the phylogenetic tree for VgrG1, Fig. 13C indicates the phylogenetic tree for TssM, and finally, Fig. 13D shows the phylogenetic tree for protein TSSG. The depicted graph shows a high degree of sequence identity among A. baumannii strains. However, in the case of TssJ, either it was absent from the complex or had very diverse amino acid sequences.

This study involved machine learning and artificial intelligence programs to uncover the mechanism of Hcp1 nanotube contraction and the transport of effector molecules to the targeted cell using A. baumannii as the host cell. Artificial intelligence demonstrated significant success in predicting the structure and function of this phenomenon. T6SS allows bacteria to compete with other members of the microbiome by killing them or inhibiting their growth. A key finding of this study was the discovery of the mechanism responsible for the movement of effector molecules in Hcp1 through sheath contraction using AI and DL simulation programs. The binding of effector molecules to specific amino acids in the 40Å hexagonal monolayers of the Hcp1 nanotube was found to induce rotational shifts in the backbone structure as well as the dihedral angles of the side chain amino acids. This shift induces conformational changes and initiates contraction of Hcp1 by using the VgrG1 contractile sheath, resulting in the propulsion of the effector molecule from the inner membrane to the tip of the T6SS syringe using the ATPases ClpV and TssM. This work also highlights the potential of machine learning in drug design and highlights a promising lead with targeted activity against challenging Gram-negative pathogens. Furthermore, it demonstrates the potential of AI to accelerate the development of uniquely active drugs of new protein sequences and structures, predict the effects of genetic variants, and provide new insights into the mechanisms of transport and injection of toxic agents into nearby host cells.

Ethics approval and consent to participate: Not applicable.

Consent for publication: Not applicable.

Availability of data and materials: The data used and/or analyzed during the current study are available from the corresponding author upon reasonable request.

Competing interests: Not applicable.

Funding: No fund was used for this research.

Authors' contributions: Not applicable

Acknowledgments: Not applicable

Howard A, O'Donoghue, M, Feeney A, & Sleator R D. Acinetobacter baumannii: an emerging opportunistic pathogen. Virulence. 2012; 3(3):243–250. https://doi.org/10.4161/viru.19700
Peleg, A. Y., Seifert, H., & Paterson, D. L. Acinetobacter baumannii: emergence of a successful pathogen. Clin Microbiol Rev. 2008; 21(3):538–582. https://doi:org/10.1128/CMR.00058-07
Ayoub Moubareck C, Hammoudi Halat D. Insights into Acinetobacter baumannii: A review of microbiological, virulence, and resistance traits in a threatening nosocomial pathogen. Antibiotics (Basel). 2020;9(3):119. https://doi:10.3390/antibiotics9030119
Antunes LC, Visca P, Towner KJ. Acinetobacter baumannii: evolution of a global pathogen. Pathog Dis. 2014;71(3):292-301.https://doi:10.1111/2049-632X.12125
Dahal U, Paul K, Gupta S. The multifaceted genus Acinetobacter: from infection to bioremediation. J Appl Microbiol. 2023;134(8): lxad145. https://doi:10.1093/jambio/lxad145
Shakibaie MR. Bacterial biofilm and its clinical implications. Ann Microbiol Res. 2018; 2(1):45-50. https://doi:10.36959/958/568
Carruthers MD, Nicholson PA, Tracy EN, Munson RS. Acinetobacter baumannii utilizes a type VI secretion system for bacterial competition. PLoS One. 2013;8:e59388. https://doi.org/10.1371/journal.pone.0059388
Karampatakis T, Tsergouli K, Behzadi P. Pan-genome plasticity and virulence factors: a natural treasure trove for Acinetobacter baumannii. Antibiotics. 2024; 13(3):257. https://doi.org/10.3390/antibiotics13030257
Shakibaie MR, Kapadnis BP, Dhakephalker P, Chopade BA. Removal of silver from photographic wastewater effluent using Acinetobacter baumannii BL54. Can J Microbiol. 1999;45(12):995-1000.
Parvathy V, Sahana V, Helma D, et al. Revisiting ESKAPE Pathogens: virulence, resistance, and combating strategies focusing on quorum sensing. Front Cell Infect Microbiol 2023; 13: 2235-2988. https://doi.org/10.3389/fcimb.2023.1159798
Lynch JP 3rd, Zhanel GG, Clark NM. Infections Due to Acinetobacter baumannii in the ICU: Treatment Options. Semin Respir Crit Care Med. 2017;38(3):311-325. https://doi:10.1055/s-0037-1599225
12. Huang Y, Zhou Q, Wang W, Huang Q, et al. Acinetobacter baumannii ventilator-associated pneumonia: clinical efficacy of combined antimicrobial therapy and in vitro drug sensitivity test results. Front Pharmacol. 2019; 10:92. https://doi: 10.3389/fphar.2019.00092
13. Mohd Sazlly Lim S, Zainal Abidin A, Liew SM, Roberts JA, Sime FB. The global prevalence of multidrug-resistance among Acinetobacter baumannii causing hospital-acquired and ventilator-associated pneumonia and its associated mortality: A systematic review and meta-analysis. J Infect. 2019;79(6):593-600. https://doi.org/10.1016/j.jinf.2019.09.012
Yee LJKG, Abad CLR, Cabaluna ITG, Lim JA. Outcomes of adult patients with multidrug-resistant Acinetobacter baumannii ventilator-associated pneumonia treated with monotherapy versus combination therapy in a tertiary hospital. Diagn Microbiol Infect Dis. 2024;110(3):116481. https://doi.org/10.1016/j.diagmicrobio.2024.116481
15. Russo A, Bassetti M, Ceccarelli G, et al. Bloodstream infections caused by carbapenem-resistant Acinetobacter baumannii: clinical features, therapy and outcome from a multicenter study. J Infect. 2019;79(2):130–138. https://doi:10.1016/j.jinf.2019.05.017
WHO Bacterial Priority Pathogens List, 2024: bacterial pathogens of public health importance to guide research, development, and strategies to prevent and control antimicrobial resistance. Geneva: World Health Organization; 2024. License: CC BY-NC-SA 3.0 IGO. https://iris.who.int/
Ma C, McClean S. Mapping Global Prevalence of Acinetobacter baumannii and Recent Vaccine Development to Tackle It. Vaccines (Basel). 2021;9(6):570. https://doi:10.3390/vaccines9060570
Vincent, J. L., Rello, J., Marshall, J., Silva, E., Anzueto, A., Martin, C. D., et al. International study of the prevalence and outcomes of infection in intensive care units. JAMA. 2009; 302:2323–2329. https://doi:10.1001/jama.2009.1754
Giamarellou H, Galani L, Karavasilis T, Ioannidis K, Karaiskos I. Antimicrobial Stewardship in the Hospital Setting: A Narrative Review. Antibiotics(Basel). 2023; 12(10):1557. https://doi.org/10.3390/antibiotics12101557
Halim J, Carr RA, Fliorent R, Jonnalagadda K, Kurbonnazarova M, Kaur M, Millstein I, Carabetta VJ. Combinations of Antibiotics Effective against Extensively- and Pandrug-Resistant Acinetobacter baumannii Patient Isolates. Microorganisms. 2024; 12(7):1353. https://doi.org/10.3390/microorganisms12071353
Ho BT, Dong TG, Mekalanos JJ. A view to a kill: the bacterial type VI secretion system. Cell Host Microbe. 2014;15(1):9-21.https://doi: 10.1016/j.chom.2013.11.008
Coulthurst S. The Type VI secretion system: a versatile bacterial weapon. Microbiology (Reading). 2019;165(5):503-515.
Gallegos-Monterrosa R, Coulthurst SJ. The ecological impact of a bacterial weapon: microbial interactions and the Type VI secretion system. FEMS Microbiol Rev. 2021;45(6):fuab033. doi:10.1093/femsre/fuab033
Geller AM, Shalom M, Zlotkin D, et al. Identification of type VI secretion system effector-immunity pairs using structural bioinformatics. Mol Syst Biol 2024; 20: 702–718. https://doi.org/10.1038/s44320-024-00035-8
Mougous JD, Cuff ME, Raunser S, et al. A virulence locus of Pseudomonas aeruginosa encodes a protein secretion apparatus. Science. 2006;312(5779):1526-1530. https://doi:10.1126/science.1128393
Allsopp LP, Bernal P. Killing in the name of T6SS structure and effector diversity. J Microbiol. 2023;169(7):001367. https://doi:10.1099/mic.0.001367
Crisan CV, Van Tyne D, Goldberg JB. 2023. The type VI secretion system of the emerging pathogen Stenotrophomonas maltophilia complex has antibacterial properties. mSphere 8:e00584-23 .https://doi.org/10.1128/msphere.00584-23
Navarro-Garcia F, Ruiz-Perez F, Cataldi Á, Larzábal M. Type VI Secretion System in Pathogenic Escherichia coli: Structure, Role in Virulence, and Acquisition. Front Microbiol. 2019; 10:1965. doi:10.3389/fmicb.2019.01965
Alteri CJ, Mobley HLT. The Versatile Type VI Secretion System. Microbiol Spectr. 2016;4(2):10. https://doi:10.1128/microbiolspec.VMBF-0026-2015
Ma AT, Mekalanos JJ. In vivo, actin cross-linking induced by Vibrio cholerae type VI secretion system is associated with intestinal inflammation. Proc Natl Acad Sci USA. 2010;107(9):4365-4370. https://doi:10.1073/pnas.0915156107
Folkesson A, Löfdahl S, Normark S. The Salmonella enterica subspecies I specific centisome 7 genomic island encodes novel protein families in bacteria close to eukaryotic cells. Res Microbiol. 2002;153(8):537-545. https://doi:10.1016/s0923-2508(02)01348-7
32. Howard, SA, Furniss R, Bonini D, Amin H, and Filloux A. The Breadth and Molecular Basis of HCp-driven Type VI Secretion System Effector Delivery. mBio 2021;12:e0026221. https://doi:10.1128/mBio.00262-21
Li P, Zhang S, Wang J, et al. Uncovering the Secretion Systems of Acinetobacter baumannii: Structures and Functions in Pathogenicity and Antibiotic Resistance. Antibiotics (Basel). 2023;12(2):195. https://doi:10.3390/antibiotics12020195
Repizo GD, Espariz M, Seravalle JL, Salcedo SP. Bioinformatic analysis of the type vi secretion system and its potential toxins in the Acinetobacter Genus. Front Microbiol. 2019; 10:2519. https://doi:10.3389/fmicb.2019.02519
Zoued A, Brunet YR, Durand E, et al. Architecture and assembly of the Type VI secretion system. Biochim Biophys Acta. 2014;1843(8):1664-1673. https://doi:10.1016/j.bbamcr.2014.03.018
Kamyshinsky RA, Chesnokov YM, & Orekhov AS. Cryo-Electron Tomography Studies of Cell Systems. Crystallogr Rep. 2020; 65:744–748. https://doi.org/10.1134/S1063774520050090
Basler M. Type VI secretion system: secretion by a contractile nanomachine. Philos Trans R Soc Lond B Biol Sci. 2015;370(1679):20150021. https://doi:10.1098/rstb.2015.0021
Ma LS, Narberhaus F, Lai EM. IcmF family protein TssM exhibits ATPase activity and energizes type VI secretion. J Biol Chem. 2012;287(19):15610-15621. https://doi:10.1074/jbc.M111.301630
Ma J, Pan Z, Huang J et al. The Hcp proteins fused with diverse extended-toxin domains represent a novel pattern of antibacterial effectors in type VI secretion systems. Virulence. 2017;8(7):1189–1202. https://doi.org/10.1080/21505594.2017.1279374
Lin L, Ringel PD, Vettiger A, Dürr L, Basler M. DNA uptake upon T6SS-dependent prey cell lysis induces sos response and reduces the fitness of Acinetobacter baylyi. Cell Rep. 2019;29(6):1633-1644.e4. https://doi:10.1016/j.celrep.2019.09.083
Lin J-S, Ma L-S, Lai EM. Systematic Dissection of the Agrobacterium Type VI Secretion System Reveals Machinery and Secreted Components for Subcomplex Formation. PLoS ONE 2013; 8(7): e67647. https://doi.org/10.1371/journal.pone.0067647
Boyer F, Fichant G, Berthod J, et al. Dissecting the bacterial type VI secretion system by a genome-wide in silico analysis: what can be learned from available microbial genomic resources? BMC Genomics. 2009;10:104. https://doi.org/10.1186/1471-2164-10-104
43. Di Venanzio G, Moon KH, Weber BS, et al. Multidrug-resistant plasmids repress chromosomally encoded T6SS to enable their dissemination. Proc Natl Acad Sci USA. 2019;116(4):1378-1383. https://doi:10.1073/pnas.1812557116
Dunbrack RL Jr. Rotamer libraries in the 21st century. Curr Opin Struct Biol. 2002;12(4):431-440. https://doi:10.1016/s0959-440x(02)00344-5
Haddad Y, Adam V, Heger Z. Rotamer Dynamics: Analysis of rotamers in molecular dynamics simulations of proteins. Biophys J. 2019;116(11):2062-2072. https://doi:10.1016/j.bpj.2019.04.017
Mortensen JC, Damjanovic J, Miao J, Hui T, Lin YS. A backbone-dependent rotamer library with high (ϕ, ψ) coverage using meta-dynamics simulations. Protein Sci. 2022; 31(12): e4491. https://doi:10.1002/pro.4491
Kabir MWU, Alawad DM, Mishra A, Hoque MT. TAFPred: Torsion Angle Fluctuations Prediction from Protein Sequences. Biology. 2023; 12(7):1020. https://doi.org/10.3390/biology12071020
Mortensen JC, Damjanovic J, Miao J, Hui T, Lin YS. A backbone-dependent rotamer library with high (ϕ, ψ) coverage using metadynamics simulations. Protein Sci. 2022;31(12): e4491. https://doi:10.1002/pro.4491
Gaudreault F, Chartier M, Najmanovich R. Side-chain rotamer changes upon ligand binding: common, crucial, correlate with entropy and rearrange hydrogen bonding. Bioinformatics. 2012;28(18): i423-i430. https://doi:10.1093/bioinformatics/bts395
Chakrabarti P, Pal D. The interrelationships of side-chain and main-chain conformations in proteins. Prog Biophys Mol Biol. 2001;76(1-2):1-102. https://doi:10.1016/s0079-6107(01)00005-0
Guo HB, Perminov A, Bekele S, et al. AlphaFold2 models indicate that protein sequence determines both structure and dynamics. Sci Rep. 2022;12(1):10696. https://doi:10.1038/s41598-022-14382-9
Jumper J, Evans R, Pritzel A, et al. Highly accurate protein structure prediction with AlphaFold. Nature. 2021;596(7873):583-589. https://doi:10.1038/s41586-021-03819-2
Tejera-Nevado P, Serrano E, González-Herrero A, Bermejo R, Rodríguez-González A. Unlocking the power of AI models: exploring protein folding prediction through comparative analysis. Int J Bioinform Res Appl. 2024;21(2): 20230041. https://doi.org/10.1515/jib-2023-0041
Qiu X, Li H, Ver Steeg G, Godzik A. Advances in AI for Protein Structure Prediction: Implications for Cancer Drug Discovery and Development. Biomolecules. 2024; 14(3):339. https://doi.org/10.3390/biom14030339
Bower MJ, Cohen FE, Dunbrack RL Jr. Prediction of protein side-chain rotamers from a backbone-dependent rotamer library: a new homology modeling tool. J Molecular Biol. 1997; 267 (5):1268–82. https://doi:10.1006/jmbi.1997.0926
56. Cascales E, and Cambillau C. Structural biology of type VI secretion systems. Phil. Trans. R. Soc. B3671102–1111. http://doi.org/10.1098/rstb.2011.0209
Nguyen VS, Douzi B, Durand E, Roussel A, Cascales E, Cambillau C. Towards a complete structural deciphering of Type VI secretion system. Curr Opin Struct Biol. 2018; 49:77-84. http://doi:10.1016/j.sbi.2018.01.007
Kandolo O, Cherrak Y, Filella-Merce I, et al. Acinetobacter type VI secretion system comprises a non-canonical membrane complex. PLoS Pathog. 2023;19(9):e1011687. http://doi:10.1371/journal.ppat.1011687
Ruiz FM, Lopez J, Ferrara CG, et al. Structural characterization of TssL from Acinetobacter baumannii: a key component of the Type VI Secretion System. J Bacteriol. 2020;202(17):e00210-20. http://doi:10.1128/JB.00210-20
Liang X, Pei T, Wang Z, Xiong W, Wu L, Xu P, Lin S, Dong TG, 2021. Characterization of lysozyme-like effector sep reveals the dependence of Type VI Secretion System (T6SS) Secretion on Effectors in Aeromonas dhakensis Strain SSU. Appl Environ Microbiol 87: e00435-21. https://doi.org/10.1128/AEM.00435-21
Noreen Z, Jobichen C, Abbasi R, Seetharaman J, Sivaraman J, Bokhari H. Structural basis for the pathogenesis of Campylobacter jejuni Hcp1, a structural and effector protein of the Type VI Secretion System. FEBS J. 2018;285(21):4060-4070. http://doi:10.1111/febs.14650
Taylor NM, Prokhorov NS, Guerrero-Ferreira RC, et al. Structure of the T4 baseplate and its function in triggering sheath contraction. Nature. 2016;533(7603):346-352. http://doi:10.1038/nature17971

No competing interests reported.

Download PDF

Version 1

posted

You are reading this latest preprint version

Investigating the structural changes in amino acids conformations interacting with a toxic effector molecule within the Hcp1 tail/tube complexes of the type VI secretion system using artificial intelligence and deep learning platform

Status:

Version 1

Abstract

Figures

Background

Methods

Results and Discussion

Inner membrane subcomplexes

The membrane sub-complex

Tail tube/sheath sub-complex

Importance of this study

Classification of Hcp and VgrG protein families

The phylogenetic tree analysis

Conclusion

Declarations

References

Additional Declarations

Status:

Version 1