http://orcid.org/0000-0002-2657-7153
Figures
Abstract
Leptospirosis is one of the leading zoonotic infections worldwide. As with other infectious diseases, report of antimicrobial resistance to existing therapeutic arsenal poses challenges in the management of disease. Hence, identification of novel drug targets for the pathogen deems essential. Present study used combined approach of comparative and subtractive genomics to identify putative drug targets. Crucial genes of 16 pathogenic Leptospira strains were filtered and subjected to homology search via target identification tool “TiD”. Thereafter, comparative analysis was performed for non-homologous, essential genes to accomplish the broad-spectrum drug target. Consequently, 37 essential genes were found to be conserved in at least 10 strains of Leptospira. Further, prioritization of resultant set of genes revealed 18 were hubs in protein–protein interaction network. Sixteen putative targets among the hub genes were conserved in all strains of Leptospira. Out of sixteen, fourteen were enzymes while 8 were novel and 4 were involved in virulence mechanism. In addition, genome scale metabolic network reconstruction and choke point analysis revealed cobA (porphyrin and chlorophyll metabolism) and thiL (thiamine metabolism) as chokepoints in their respective metabolic pathways. The proposed hub genes could act as putative broad-spectrum drug targets for Leptospira species, however, these putative targets should be validated to ensure them as real one prior to utilizing them for target based lead discovery.
Citation: Gupta R, Verma R, Pradhan D, Jain AK, Umamaheswari A, Rai CS (2019) An in silico approach towards identification of novel drug targets in pathogenic species of Leptospira. PLoS ONE 14(8): e0221446. https://doi.org/10.1371/journal.pone.0221446
Editor: Manas Ranjan Dikhit, Rajendra Memorial Research Institute of Medical Sciences, INDIA
Received: May 5, 2019; Accepted: August 6, 2019; Published: August 20, 2019
Copyright: © 2019 Gupta et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the manuscript and its Supporting Information files.
Funding: The author(s) received no specific funding for this work.
Competing interests: The authors have declared that no competing interests exist.
Introduction
The globally widespread occurrence of bacterial resistance to present drugs has stressed the necessity to find novel targets [1]. Human leptospirosis is a zoonotic infection caused by pathogenic spirochete Leptospira with the prevalence of about one million cases per year of which 60,000 failed to survive annually [2–4]. It is an endemic, occupational as well as recreational disease in tropical and rural areas [5–7]. Humans, who are exposed frequently to the diseased rodents, pet animals and polluted water, are at high risk of leptospirosis [8]. Clinical symptoms of the disease range from mild fever, vomiting, flu like illness, headache, diarrhea and muscle ache to multi-organ system complications which include kidney, liver, central nervous system and lungs with death rates 5% to 40% [9–10].
Leptospira infection cases have been highly recorded from Indonesia, Thailand, India, Sri Lanka and Maldives. India has been recognized as a major hub for Leptospira spp. to cause leptospirosis since 20th century as it is a confluence of environmental, socio-economic and demographic factors [11]. Climate changes, population size, global warming, natural calamities (cyclone, floods), lack of facilities, poor sanitary infrastructure and clinical suspicion etc. increase the incidence of leptospirosis in coastal and rainy states of India like Gujarat, Kerala, Tamil Nadu, Maharashtra and Andaman-Nicobar islands [8,11–13].Various studies conducted in past few years confirmed higher prevalence of leptospirosis. Most substantial outbreaks have been observed in Chennai floods in 2015, spontaneous eruptions in Gujarat in 2011, flash floods in Mumbai in 2005 and cyclone in Orissa in 1999.
Taylor & Goyal et al., 1931, isolated L. andamans and L. grippotyphosa from diseased patients [14]. In 1960, Dalal P. M. provided the evidence of L. icterohaemorrhagiae antigen in jaundice patients. Similar report was submitted by Joseph K. M. in 1983 who found Leptospira infection in patients of jaundice. Muthusethupathi and Shivkumar observed renal failure in patients of Madras due to Leptospirosis. In 1996, Saravanan & Rajendran isolated L. javanica from urine sample of renal failure, cases of Chennai. Gujarat reported 130 deaths in 2011 within a span of two months due to leptospirosis. Recently, Kochi and Kerala reported 209 cases with 12 deaths. In October 2012, Gujarat reported 16 deaths. These reports highlights the continuous alarming jeopardy the disease presents at this hour.
Presently, treatment of severe leptospirosis is still unclear [15]. Antibiotics (penicillin, cephalosporins, azithromycin, doxycycline) and vaccines were relatively unsuccessful against Leptospira spp. This emphasizes the need of a new drug targets for evolution of competent drug that kill the pathogen [16–18]. In the last two decades, classical research approach was being used to classify protein targets towards the expansion of subunit and recombinant vaccines against leptospirosis. In 2005, whole genome analysis was carried out to classify potential vaccine against L. Interrogans [19]. Amineni et al., 2010 proposed 88 putative drug targets for two serovars of L. interrogans [20]. Lack of data, information as well as methodology, restrict the development of novel potent targets. Previous studies lack the gaps or information that would be useful in revealing of potent target like (1) coverage of all pathogenic strains, (2) metabolic reconstruction and system biology analysis, (3) exploration of hub genes that are analogous for endurance and virulence of pathogen, (4) recognition of choke point enzymes or reactions and (5) common target among the pathogenic strains.
In this study, we have tried to design a protocol to overcome the limitations of previous studies. Our protocol includes complete genome analysis, subtractive approach, comparative approach, and protein-protein interaction analysis that ultimately link to metabolomics. Various bioinformatics tools and cheminformatics techniques represent an attractive source of alternative method for target identification. The drug targets could be used, sequentially, for optimization and new lead recognition through free energy calculation, molecular modeling, molecular dynamics simulations, multiple docking strategies and drug-likeliness determination to establish new antibacterial agent against leptospirosis.
Materials and methods
Pathogenic strains of Leptospira were subjected to a broad-spectrum anti-leptospiral target identification strategy that involves subtractive and comparative genomics approach along with protein-protein interaction network analysis. The complete protocol of the target identification is shown in Fig 1.
Classification of Leptospira strains
Up to 2016, global character was lacking in classification of Leptospira. Fouts et al (2016) and Caimi et al (2017) have established the classification of Leptospira by wide range inter-species genomic comparison of all known infectious and non-infectious species of the bacterium [21–22]. In the present study, we have followed the systems established by Fouts and Caimi to classify various species of Leptospira as pathogenic, intermediate and saprophytic to humans. List of classified Leptospira strains was prepared and target mining using subtractive genomics, comparative genomics along with protein-protein interaction network analysis was restricted to pathogenic strains only.
Mining of potential drug targets
Whole genome of pathogenic strains was retrieved and processed using target identification software called TiD developed by our team [23]. TiD is a standalone program. It consists of modules for paralog analysis, non-homolog analysis and essentiality analysis that exclude duplicates proteins, homolog of human and its gut flora proteins from the essential protein dataset [24–26]. Target prioritization module annotates the screened essential protein dataset on druggability and virulence which is essential for pathogen survival as well as pathogenicity [27].
Target identification and prioritization parameters were adapted from our previously published paper [23]. The standard protocol for target identification was used to remove paralogs from pathogenic leptospiral proteome and, then, define the list of essential genes that are homologs to entries present in Database of Essential Genes (DEG) with e-value ≤ 10−10, bit score ≥ 100. Further, the proteins homology was checked with human and gut microbes. The resulting dataset was compiled as putative drug targets and annotated for virulence, drug likeliness and mapped with UniProt identifiers (http://www.uniprot.org/).
Metabolic pathway analysis
The mapped dataset of putative targets was prioritized at the KAAS (KEGG Automatic Annotation Server) to know the involvement of these proteins in metabolic pathways. Functional annotation of these targets was also acquired through BLAST comparisons against KEGG database [28]. KAAS server has programmed procedure that depends on sequence similarities and bi-directional best hit data to assign K numbers that permits recreation of KEGG pathways.
Subcellular localization
Subcellular localization along with the biological significance of inimitable drug targets was analyzed in CELLO v.2.5. It is a multi-class support vector machine sorting method. This identifies targets as membrane proteins, cellular proteins or surface protein [29]. Results obtained were further analyzed in subcellular localization prediction tool pSORTb v.3.0. It is first standalone software that predicts the location of proteins for all prokaryotes including archaea and bacteria with atypical membrane/cell wall topologies with high accuracy [30].
Genome scale metabolic network reconstruction
The metabolic network reconstruction of L. borgpetersenii serovar Hardjo-bovis str. L550 and L. interrogans serovar Copenhageni str. Fiocruz L1-130 were performed using ‘PathoLogic’ plugins of pathway tools [31]. GenBank file format of each chromosome of organism were used as input in ‘Build’ section that involved ‘Replicon Editor’, ‘Trial Parse’, ‘Automated Build’ and Refine. Pathway hole filler was used to fill the gaps of reconstructed network and then, save the database for choke point analysis.
Choke-point analysis
"Chokepoint responses" are the reactions which exclusively catalyze a particular substrate or preferentially deliver a particular product [32–33]. To distinguish potential drug targets, we searched choke points reactions using Pathway tool. Choke point analysis includes all reactions that raise several pathways while excludes those reactions catalyzed by more than one enzyme and found in human. Result of choke point analysis was compared with putative targets. Results of Pathway tools were also validated with BioCyc webserver [34]. It is an online server that comprised of 13075 pathway genome databases and provides unified data on the genomes and metabolic pathways of thousands of sequenced organisms which are built via metabolic network reconstruction software(s).
Sequence-structure relationship
A multiple sequence alignment was created for cobA and thiL sequences of Leptospira strains using clustalX and UniProt. Structurally conserved elements were identified and formatted with ESPript 3.0 to obtain the conserved and semi conserved residues of both the chokepoints [35]. The secondary structure of cobA and thiL were determined by using PsiPred server. Further, a three dimensional structure model of a target sequences were built by Modeller 9.17 [36]. Energy minimization of best modelled structure having least DOPE score was performed using YASARA server [37]. Validation of energy minimized structures was performed through SAVES server ((http://services.mbi.ucla.edu/SAVES/), QMEAN and ProSA [38–39]. The catalytic pocket and active site residues of validated model were defined using CASTp and PyMOL-2.3.2 [40–41].
Protein-protein network analysis
Protein-protein interaction data is available for two strains of Leptospira in string database. Therefore, network analysis was restricted to these two strains. PPI network of putative drug targets of both strains were constructed in STRING app of Cytoscape v3.7.1 with confidence score ≥0.4 [42]. Network interaction data was figured through network analyzer module [43]. Putative drug targets with interacting partners’ ≥30 in Leptospira proteome were subjected to MCODE plugin for the functional module detection [44]. The cutoff parameters used for molecular complex detection were degree cutoff = 2, node score cutoff = 0.2, k-core = 2, and maximum depth = 100. The highest ranked module was chosen for gene ontology and functionality. Gene enrichment analysis of interacting nodes was also carried in STRING app. The results of two strains were checked for conservancy among the pathogenic strains using multiple sequence alignment tool ClustalX2 [45].
Results and discussion
The current study is an adoption of advanced subtractive and comparative genomic approach and further augmented to protein-protein interaction network and metabolomics analysis. The unique and essential proteins are significant for Leptospira growth, survival and pathogenicity. A search for the hub proteins and choke enzymes in the unique and essential pathways is therefore considered as a promising approach to deal with the challenging leptospirosis infection.
Classification of Leptospira strain
In silico and in vitro target identification research to cure leptospirosis, till date, have been more focused towards most pathogenic L. interrogans serovars named Lai and Copenhagani [20,46–47]. The present study tried to cover all pathogenic strains of Leptospira which are responsible for causing infectious disease. We grouped Leptospira strains based on their pathogenicity. Out of 27 retrieved strains from NCBI FTP server, 16 are observed to be pathogenic in nature, 5 are intermediate and 6 are saprophytic in nature (Table 1). Sixteen pathogenic strains of Leptospira were selected for putative drug target mining [21–22,48].
Sequence retrieval and putative drug targets mining
Complete amino acid sequence (*.faa) of sixteen serovars were successfully retrieved from NCBI-FTP server and subjected to TiD software. Nascimento et al., 2004 and Ren et al., 2003 reported the genomic size of L. interrogans i.e. to be 4.33-Mb chromosome I and 350-kb chromosome [49–50]. In the present study, L. weilii serovar Topaz str. LT2116 proteome consists of most elevated number of proteins while L. borgpetersenii serovar Hardjo-bovis str. L550 has smallest proteome of 2945 proteins. About 13 pathogenic Leptospira strains have larger proteome than L. interrogans. Size of each proteome is mentioned in Fig 2A and 2B. Amineni et al., 2010 mentioned 158 essential genes of Lai and 218 genes of Copenhagni as human non homolog [20]. In our study, 736 proteins of L. kmetyi serovar Malaysia str. Bejo-Iso9 followed by 732 proteins of L. alstonii are identified as essential proteins. L. borgpetersenii serovar Hardjo-bovis str. L550 has lowest number of essential proteins (Fig 2A and 2B). L. kmetyi serovar Malaysia str. Bejo-Iso9 has uppermost whereas L. santarosai serovar Shermani str. LT 821 has minimum number of human non-homolog as well as gut flora non homolog protein. All strains of Leptospira have more than 36 novel drug targets and 31 virulent proteins. Results of drug target mining of each pathogenic strain are reported in Fig 2A and 2B.
The plot of detailed protein products of L. alexandari, L. alstonii, L. borgpetersenii, L. interrogans, L. kirschneri, L. kmetyi, L. mayottensis, L. noguchii is represented in Fig 2A. The plot of detailed protein products of L. sp. B5-022, L. sp. Fiocruz LV3954, L. sp. Fiocruz LV4135, L. sp. P2653, L. sp. ZV016, L. sp. serovar Kenya str. Sh9, L. santarosai serovar Shermani str. LT 821, L. weilii serovar Topaz str. LT2116 is represented in Fig 2B. The resulted drug targets in each step represented using bar graph.
Metabolic pathway analysis
Metabolic pathway analysis of the host-pathogen highlighted sixteen pathogen specific pathways such as biosynthesis of antibiotics, biosynthesis of secondary metabolites, degradation of aromatic compounds, microbial metabolism in diverse environments, sulfur metabolism, cysteine and methionine metabolism, seleno compound metabolism, lysine degradation, beta-alanine metabolism, glutathione metabolism, folate biosynthesis, lipopolysaccharide biosynthesis, ubiquinone and other terpenoid-quinone biosynthesis, phenylpropanoid biosynthesis, bacterial chemotaxis and flagellar assembly. These pathways were observed to have 34 communal drug targets (Fig 3, Table 2). In addition, common drug targets were also identified from 8 pathways which are unique to the survival of 15 strains of Leptospira. The findings of our study are in same line with the Amineni et al., 2010 and Anisetty et al. 2005 [20,51]. Among these common drug targets three are non enzyme and 31 are enzymes (Table 2).
Sixteen pathways are unique to all strains of Leptospira to which 34 common genes were retrieved as putative drug targets. Red color denotes presence while Green color denotes absence of pathways from which common genes were retrieved for further analysis.
Subcellular localization
Localization of 34 common drug targets in current study exposed that 29 drug targets are cytoplasmic and 5 are transmembrane proteins (Table 2). The membrane proteins have the capability to act as useful vaccine drug targets against all strains of Leptospira. However, these transmembrane proteins are involved in lipid polysaccharide synthesis (kdtA), ubiquinone-terpenoid-quinone biosynthesis (ubiA), bacterial chemotaxis (fliN, cheA, mcp)as well as part of flagellar assembly (fliN) and two component system (mcp, cheA). These play an essential role in the survival and pathogenesis of Leptospira. Location of these drug targets is required in future to design drug or vaccine accordingly. Subcellular localization information of a drug target should complement with the pharmacological properties of lead molecules targeted to it, therefore, it is an important aspect in rational drug design.
Genome scale metabolic network reconstruction
Genome scale metabolic network of L. borgpetersenii serovar Hardjo-bovis str. L550 and L. interrogans serovar Copenhageni str. Fiocruz L1-130 were constructed successfully using Pathway Tools Software. Model of L. interrogans serovar Copenhageni str. Fiocruz L1-130 comprised 673 enzymes, 988 enzymatic reactions, 6 transport reactions, 133 pathways, 957 compounds, 37 tRNAs, 6 transporter and 3684 polypeptides (S1 SBML). Likewise, model of L. borgpetersenii serovar Hardjo-bovis str. L550 constructed with 141 pathways, 613 enzymes, 1099 enzymatic reactions, 5 transport reactions, 3290 polypeptides, 12 transporters and 1075 metabolites (S2 SBML). From the set of 34 drug target, 20 and 23 were shaped in reconstructed model of L. borgpetersenii serovar Hardjo-bovis str. L550 and L. interrogans serovar Copenhageni str. Fiocruz L1-130 respectively. These reconstructed models were used for choke point analysis.
Choke point analysis
Chokepoint enzymes are enzymes that catalyze a reaction which either uniquely consume a substrate or interestingly deliver a precise product. If an enzyme catalyzes at least one chokepoint reaction, it is classified as a promising drug target. Accordingly, we expect the hindrance of a protein that expends an exceptional substrate brings aggregation in the remarkable substrate. This may be conceivably harmful to the cell and restraint the compound that delivers one of a kind item which further leads to the starvation of special product [32–33].
Pathway tool generated the report of choke point reactions for reconstructed model of L. interrogans serovar Copenhageni str. Fiocruz L1-130 and L. borgpetersenii serovar Hardjo-bovis str. L550. The report consists of choke point reactions on the consuming side as well as producing side. Target cobA (cob(I)yrinic acid a) and thiL (thiamine-phosphate kinase) is found as a choke enzyme in both the reconstructed genome scale metabolic model of Leptospiral strains (Table 2). The cobA with an EC: 2.5.1.17 participated in cobalamin biosynthesis pathway. Pathogenic strains of Leptospira can’t grow in the absence of cobalamin that’s why it is an essential component of the Ellinghausen-McCullough-Johnson-Harris (EMJH) semi-synthetic selective medium [49–50]. In contrast, gene thiL found to be important in conversion of thiamine monophosphate to thiamine pyrophosphate which is essential for pathogen survival [52]. Results of pathway tool were also validated with BioCyc server that shows cobA as choke point in most of the Leptospiral strain and thiL in L. mayottensis 200901122 and L. kmetyi serovar Malaysia str. Bejo-Iso9.
Sequence alignment, homology modelling and validation
Multiple sequence alignment was performed for cobA and thiL gene within the pathogenic Leptospira strain (Table 1). The result showed conserved, semi conserved and dispersed amino acid residues. The semi-conserved residues were showen with dot and conserved with red star (S1 and S2 Files). ESPript defined MSA of cobA showed the 10 αhelix (9α + η), 10 β-sheets and 6 turns (TT) while 9 αhelix (8α + η), 12 β-sheets and 7 turns (TT) were found in thiL. In addition, psiPred server validated the ESPript report for both the chokepoints. All these results were in agreement with each other for the secondary structure elements of cobA and thiL. Afterwards, 3D structure of thiL and cobA were modelled using 2YBQ_A (Query Coverage: 88%, Identity: 39.36%) and 3C9R_A (Query Coverage: 88%, Identity: 30.32%) as template respectively (Fig 4). The quality of the modelled structures after energy minimization was evaluated on the UCLA SAVES server shown in Table 3. CastP predicted the largest binding pocket of both modelled structures. Surface view of cobA is showing a deep binding pocket while thiL represented Gln23, Thr24, Asp25, Asp26, Asp39, Asp68, Arg140, Asp207, Thr209 and Asp210 as key residues in their binding pocket (Fig 4).
(A) Largest binding site of modelled structure of thiL (magenta color) showing the active site cleft (grey color). Key residues of binding pocket are highlighted with blue color while rests are shown in a box. (B) Likewise, surface view of binding pocket of cobA (green color) is very deep in comparison with thiL. Binding residues present in largest pocket of modelled protein is presented in orange color.
Prioritization of drug targets
Putative drug targets were subjected for prioritization through network analysis, molecular complex detection and gene enrichment analysis (Fig 1). As mentioned, string app is restricted to generate protein-protein interaction network of L. borgpetersenii serovar Hardjo-bovis str. L550 and L. interrogans serovar Copenhageni str. Fiocruz L1-130. Network of L. borgpetersenii comprised of 38 nodes with 41 edges. Gene lpxD-2, kdsB-2 and lpxK demonstrated the most noteworthy degree followed by kdtA, lpxD-1, lpxB and lpxC with degree 6 (Table 4; Fig 5A). Similarly, L. interrogans Copenhageni L1-130 had 41 nodes with 40 edges. Most astounding degree hub is observed for kdtA and lpxK followed by lpxD, lpxB, lpxC and lpxA (Table 5; Fig 5C). Extension of these two networks, with all proteins of pathogen, stretched the estimation of best genes of the system and predicts with assurance the essential role of these genes for pathogen (Tables 4 and 5). After network extension, degree of cheR increased from 0 to 91, fliN from 0 to 68, cobA from 0 to 35 and metX from 0 to 38 in case of L. borgpetersenii Hardjo-bovis L550 while degree of cheR increased from 4 to 76, fliN from 0 to 66, cob A from 4 to 47 and metX from 2 to 39 in case L. interrogans Copenhageni L1-130.
(A) & (B) present the protein-protein interaction within the selected proteins of L. borgpetersenii serovar Hardjo-bovis str. L550 whereas (C) & (D) denote the interacting partners of L. interrogans serovar Copenhageni str. Fiocruz L1-130. In PPI network, nodes denote protein and interaction between the 2 nodes denotes the edge. Significant MCODE cluster (B) and (D) denotes the genes which are involved in polysaccharide biosynthesis pathway.
Molecular code detection produces one significant cluster with MCODE score = 15.733 for L. borgpetersenii Hardjo-bovis L550. It comprised 16 hubs associated through 118 edges. All 16 hubs are similarly connecting with one another with the degree of 13 to 15. Gene kdtA is found to be the seed protein of PPI network (Fig 5B). Gene enrichment analysis displayed lpxD-2, kdsB2, lpxC, lpxB, kdtA, lpxK, lpxD-1 and manA. These are markedly enriched in positive regulation of lipopolysaccharide biosynthesis while most of the genes are involved in metabolic pathways. Similarly, L. interrogans Copenhageni Fiocruz L1-130 module 1 obtained MCODE score = 14.857 with 15 nodes that interconnected with 104 edges. Gene kdsB is the seed of network which is connected to 13 other genes of the network (Fig 5D). Gene enrichment displayed 4 genes fliN, cheR, mcp and aer. These genes are part of two component system and involved in bacterial chemotaxis. The gene ontology analysis indicated that selected essential genes of both the pathogens are important part of metabolic pathways and lipopolysaccharide biosynthesis (Fig 5). Extended network added ribB, ribC, cysl, cysH, cysD, cobA and cysG genes which are responsible for riboflavin biosynthesis, microbial metabolism in diverse environment, sulfur metabolism porhyrin and cholorophyll metabolism (Tables 4 and 5).
Comparative analysis
Comparison between both the networks showed 37 putative targets having the degree ≥30 among which 18 hubs are observed to be common in both pathogens. Among the hub proteins, 16 are also shortlisted from the set of 34 common putative drug targets. Multiple sequence alignment revealed that these hub targets are conserved in at least 10 strains of pathogen. Among these common hub proteins, 8 are novel in at least 10 strains, 5 are virulent while 2 are novel as well as common drug targets (Figs 6 and 7). Hence, these 8 proteins lpxB, lpxK, kdtA, fliN, cobA, metX, thiL and ubiA are proposed as putative drug targets in the present study, as absence of these proteins would affect the survival and pathogenicity of pathogen. However, target proteins lpxB, lpxK, kdtA, cobA, metX and ubiA were previously reported in the study of Amineni et al., 2010 for 2 pathogenic strains of Leptospira through subtractive genomic analysis but fliN and thiL are novel targets which are reported first time in our study [20] (Fig 8).
X axis consists of number of strains and Y axis contains non-homolog gut flora proteins present in Leptospira strains. Right mark denotes the presence of novel target in each strain. Gene product lpxK, lpxB and cysD found to be novel in all strains of pathogen whereas in 16 out of 15 strains, kdtA, fliN, rpoN and queF are seen to be novel.
X axis consists of number of strains and Y axis contains non-homolog gut flora proteins present in Leptospira strains. Right mark denotes the presence of virulent target in each strain. Gene product lpxK and lpxB found to be virulent in all strains of pathogen whereas in 16 out of 15 strains, kdtA and rpoN are seen to be virulent. Protein fliN is important for virulence in 14 strains of Leptospira.
Comparative analysis revealed 37 genes are essential. Among these, 16 are common in all strains of Leptospira from which 8 are found to be novel. From novel targets of Leptospira, 2 are acting as choke points and fliN protein is virulent in nature.
The functional importance of the target candidates and the pathways involved are discussed here. As mentioned, proposed 8 drug targets majorly belong to 7 essential pathogen specific pathways. Amphipathic lipopolysaccharides are outer cell wall surface antigen of Leptospira and actively participate in virulence and antibiotic sensitivity [53–54]. Lipopolysaccharides also act as primary barrier of pathogen and maintain the osmolarity of the gram negative cell. Inhibitor targeting lpxB, lpxK and kdtA proteins of outer membrane would affect the formation of primary barrier and alter the osmotic environment of pathogenic strains. This could be helpful to cure the leptospiral infections successfully. Taylor et al., 2008 and Raetz et al., 2002 also reported the proteins of this pathway as an efficient drug targets in other bacterial pathogens [20,55–56].
One of the central and defining reasons of increasing prevalence rate of Leptospirosis is the capability of the pathogenic strains to switch their flagellar motility for highly effective translocation via viscous substrates and tissues that permits access to far-away host niches [57]. Flagellar motor protein consists of three proteins among which fliN is one of major protein of flagellar motor switch system present in the basal body and interacts with chemotaxis proteins to define the translational and rotational motion of flagella [49,50]. Liao et al., 2009 showed inactivation of fliN gene that affects the rotative motion and migration in liquid and semi-solid medium respectively [58]. Hence, these complex set of genes involved in endoflagellum and its basal rotor system formation play critical role in the motility and virulence to cause Leptospirosis.
Target metX involved in methionine biosynthesis is found to be essential in all strains [59]. It is a proteinogenic as well as a component of S-adenosyl methionine which acts as a main methyl group carrier in cell. In most of the organisms including bacteria, it plays an important role in initiation of translation. Saint-Macary et al., 2015 reported that biosynthesis of methionine is essential for infection of M. oryzae [60]. Target cobA of cobalamin biosynthesis is present in chromosome I and also plays an essential role in the synthesis of vitamin-B12. Previous experimental evidence showed that vitamin B12 lacking growth medium was unable to grow the L. interrogans [50]. Thiamine monophosphate kinase (thiL) is an enzyme that catalyzes the thiamine monophosphate and converts it into thiamine pyrophosphate which is an essential cofactor in all living organism including Leptospira spp. Bian et al., 2011 discovered that to regulate the ABC transporters, riboswitches restricted to interact first with thiamine pyrophosphates due to which pathogen failed to grow in insufficient environment of thiamine pyrophosphate [52].
Thus, the proposed hub genes through the computational approach herein, support to find out broad spectrum drug targets that would be effective against pathogenic strain of Leptospira species. Although, the significance level of these putative targets need to be validated through experimental approach to ensure them as real one and can be the best against pathogenic strains of Leptospira.
Conclusion
In the post genomic era, drug designing and discovery method is changing earlier established viewpoints. It routinely reorganizes the drug discovery method by incorporating vast data encrypted in our genome. We have performed subtractive genomic and comparative genomics analyses with network analysis of 16 pathogenic strains of Leptospira and identified 8 common drug targets that can be potential targets for drug designing and vaccine development. Moreover, many of the recognized drug targets have been observed to play a key role in the essential metabolic pathways, lipid biosynthesis, flagellar motor protein system and Bacterial chemotaxis. Among these targets, cobA and thiL also found as potent drug targets during genome scale metabolic reconstruction and choke point analysis. An efficient way to develop drugs against these targets would be substantially positive to diminish the threats of serious leptospirosis. However, these targets should be corroborated by further laboratory research for their role in inhibiting the growth and affecting the virulence of pathogens. Genome scale metabolic model of Leptospira would be useful in future for the basis of in silico gene knockout studies. Determination of qualitative tertiary structure and acknowledgment of functionally critical residues of these putative drug targets would be more effective in future for identification of novel leads and its optimization through in silico approaches like protein-ligand docking, free energy calculation and molecular dynamic simulations to design new anti-leptospiral drug against leptospirosis.
Supporting information
S1 SBML. Genome scale metabolic network reconstruction of L.interrogans serovar Copenhageni str. Fiocruz L1-130.
https://doi.org/10.1371/journal.pone.0221446.s001
(RAR)
S2 SBML. Genome scale metabolic network reconstruction of L.borgpetersenii serovar Hardjo-bovis str. L550.
https://doi.org/10.1371/journal.pone.0221446.s002
(RAR)
S1 File. Multiple sequence alignment of thiL.
https://doi.org/10.1371/journal.pone.0221446.s003
(DOCX)
S2 File. Multiple sequence alignment of cobA.
https://doi.org/10.1371/journal.pone.0221446.s004
(DOCX)
Acknowledgments
Authors highly acknowledge Indian Council of Medical Research for supporting facilities to carry out the research work in Biomedical Informatics Centre, National Institute of Pathology, New Delhi vide RFC No. BIC/NTF/1/2013-2014 dated 18/02/2013.
References
- 1. Fair RJ, Tor Y. Antibiotics and Bacterial Resistance in the 21st Century. Perspect Med Chem. 2014 Aug 28;6:25–64.
- 2. Costa F, Hagan JE, Calcagno J, Kane M, Torgerson P, Martinez-Silveira MS, et al. Global Morbidity and Mortality of Leptospirosis: A Systematic Review. PLoS Negl Trop Dis. 2015 Sep 17;9(9):e0003898. pmid:26379143
- 3.
Haake DA, Levett PN. Leptospirosis in Humans. In: Adler B, editor. Leptospira and Leptospirosis [Internet]. Berlin, Heidelberg: Springer Berlin Heidelberg; 2015 [cited 2019 Apr 6]. p. 65–97. (Current Topics in Microbiology and Immunology). Available from: https://doi.org/10.1007/978-3-662-45059-8_5
- 4. Leptospirosis worldwide, 1999. Releve Epidemiol Hebd. 1999 Jul 23;74(29):237–42.
- 5. Pal M, Hadush A. Leptospirosis: An Infectious Emerging Waterborne Zoonosis of Global Significance. Air Water Borne Dis. 2017 Mar 15;6(1):1–4.
- 6. Victoriano AFB, Smythe LD, Gloriani-Barzaga N, Cavinta LL, Kasai T, Limpakarnjanarat K, et al. Leptospirosis in the Asia Pacific region. BMC Infect Dis. 2009 Sep 4;9(1):147.
- 7.
WHO Western Pacific | World Health Organization [Internet]. [cited 2019 May 5]. Available from: https://www.who.int/westernpacific
- 8. Izurieta R, Galwankar S, Clem A. Leptospirosis: The “mysterious” mimic. J Emerg Trauma Shock. 2008 Jan 1;1(1):21. pmid:19561939
- 9. Fernandes LGV, Vieira ML, Kirchgatter K, Alves IJ, Morais ZM de, Vasconcellos SA, et al. OmpL1 Is an Extracellular Matrix- and Plasminogen-Interacting Protein of Leptospira spp. Infect Immun. 2012 Oct 1;80(10):3679–92. pmid:22802342
- 10. Plank R, Dean D. Overview of the epidemiology, microbiology, and pathogenesis of Leptospira spp. in humans. Microbes Infect. 2000 Aug 1;2(10):1265–76. pmid:11008116
- 11. Sethi S, Sharma N, Kakkar N, Taneja J, Chatterjee SS, Banga SS, et al. Increasing Trends of Leptospirosis in Northern India: A Clinico-Epidemiological Study. PLoS Negl Trop Dis [Internet]. 2010 Jan 12 [cited 2019 May 5];4(1). Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2797087/
- 12. Dhanze H, Kumar S, Mane B. Epidemiology of leptospirosis: an Indian perspective. J Foodborne Zoonotic Dis. 2013 Sep 30;1:6–13.
- 13. Sambasiva RR, Naveen G, P B, Agarwal SK. Leptospirosis in India and the rest of the world. Braz J Infect Dis Off Publ Braz Soc Infect Dis. 2003 Jun;7(3):178–93.
- 14. Taylor J, Goyle A. Leptospirosis in Andamans. Indian J Nedical Res. 1931;20:55–6.
- 15. Sivaprakasam V, Zochowski WJ, Palmer MF. Clinical spectrum of severe leptospirosis in the UK. JMM Case Rep [Internet]. 2014 [cited 2019 Apr 6];1(4). Available from: https://jmmcr.microbiologyresearch.org/content/journal/jmmcr/10.1099/jmmcr.0.000003
- 16. Wang Z, Jin L, Węgrzyn A. Leptospirosis vaccines. Microb Cell Factories. 2007 Dec 11;6(1):39.
- 17. Levett PN. Leptospirosis. Clin Microbiol Rev. 2001 Apr 1;14(2):296–326. pmid:11292640
- 18. Phimda K, Hoontrakul S, Suttinont C, Chareonwat S, Losuwanaluk K, Chueasuwanchai S, et al. Doxycycline versus Azithromycin for Treatment of Leptospirosis and Scrub Typhus. Antimicrob Agents Chemother. 2007 Sep 1;51(9):3259–63. pmid:17638700
- 19. Gamberini M, Gómez RM, Atzingen MV, Martins EAL, Vasconcellos SA, Romero EC, et al. Whole-genome analysis of Leptospira interrogans to identify potential vaccine candidates against leptospirosis. FEMS Microbiol Lett. 2005 Mar 1;244(2):305–13. pmid:15766783
- 20. Amineni U, Pradhan D, Marisetty H. In silico identification of common putative drug targets in Leptospira interrogans. J Chem Biol. 2010 May 14;3(4):165–73. pmid:21572503
- 21. Fouts DE, Matthias MA, Adhikarla H, Adler B, Amorim-Santos L, Berg DE, et al. What Makes a Bacterial Species Pathogenic?:Comparative Genomic Analysis of the Genus Leptospira. PLoS Negl Trop Dis. 2016 Feb 18;10(2):e0004403. pmid:26890609
- 22. Caimi K, Repetto SA, Varni V, Ruybal P. Leptospira species molecular epidemiology in the genomic era. Infect Genet Evol J Mol Epidemiol Evol Genet Infect Dis. 2017;54:478–85.
- 23. Gupta R, Pradhan D, Jain AK, Rai CS. TiD: Standalone software for mining putative drug targets from bacterial proteome. Genomics. 2017 Jan 1;109(1):51–7. pmid:27856224
- 24. Kanehisa M, Sato Y, Kawashima M, Furumichi M, Tanabe M. KEGG as a reference resource for gene and protein annotation. Nucleic Acids Res. 2016 Jan 4;44(D1):D457–62. pmid:26476454
- 25. Fu L, Niu B, Zhu Z, Wu S, Li W. CD-HIT: accelerated for clustering the next-generation sequencing data. Bioinformatics. 2012 Dec 1;28(23):3150–2. pmid:23060610
- 26. Ye Y-N, Hua Z-G, Huang J, Rao N, Guo F-B. CEG: a database of essential gene clusters. BMC Genomics. 2013 Nov 9;14:769. pmid:24209780
- 27. Hema K , VP I, Pradhan D, Munikumar M, eep S S, et al. Identification of Putative Drug Targets and Vaccine Candidates for Pathogens Causing Atherosclerosis. Biochem Anal Biochem [Internet]. 2015 Apr 17 [cited 2018 Aug 12];4(2). Available from: https://www.omicsonline.org/open-access/identification-of-putative-drug-targets-and-vaccine-candidates-for-pathogens-causing-atherosclerosis-2161-1009-1000175.php?aid=53170
- 28. Moriya Y, Itoh M, Okuda S, Yoshizawa AC, Kanehisa M. KAAS: an automatic genome annotation and pathway reconstruction server. Nucleic Acids Res. 2007 Jul 1;35(suppl_2):W182–5.
- 29. Yu C-S, Chen Y-C, Lu C-H, Hwang J-K. Prediction of protein subcellular localization. Proteins Struct Funct Bioinforma. 2006;64(3):643–51.
- 30. Yu NY, Wagner JR, Laird MR, Melli G, Rey S, Lo R, et al. PSORTb 3.0: improved protein subcellular localization prediction with refined localization subcategories and predictive capabilities for all prokaryotes. Bioinformatics. 2010 Jul 1;26(13):1608–15. pmid:20472543
- 31. Karp PD, Paley S, Romero P. The Pathway Tools software. Bioinforma Oxf Engl. 2002;18 Suppl 1:S225–232.
- 32. Pandey M, Tiwari D, Maurya S, Singh D, Srivastava P. Choke Point Analysis with Subtractive Proteomic Approach for Insilico Identification of Potential Drug Targets in Shigella Dysenteriae. Int J Comput Appl. 2015 Jan 14;109:29–34.
- 33. Yeh I, Hanekamp T, Tsoka S, Karp PD, Altman RB. Computational Analysis of Plasmodium falciparum Metabolism: Organizing Genomic Information to Facilitate Drug Discovery. Genome Res. 2004 May 1;14(5):917–24. pmid:15078855
- 34. Karp PD, Billington R, Caspi R, Fulcher CA, Latendresse M, Kothari A, et al. The BioCyc collection of microbial genomes and metabolic pathways. Brief Bioinform [Internet]. [cited 2019 May 5]; Available from: https://academic.oup.com/bib/advance-article/doi/10.1093/bib/bbx085/4084231
- 35. Robert X, Gouet P. Deciphering key features in protein structures with the new ENDscript server. Nucleic Acids Res. 2014 Jul;42(Web Server issue):W320–324. pmid:24753421
- 36. Webb B, Sali A. Protein structure modeling with MODELLER. Methods Mol Biol Clifton NJ. 2014;1137:1–15.
- 37. Krieger E, Joo K, Lee J, Lee J, Raman S, Thompson J, et al. Improving physical realism, stereochemistry, and side-chain accuracy in homology modeling: Four approaches that performed well in CASP8. Proteins. 2009;77 Suppl 9:114–22.
- 38. Benkert P, Biasini M, Schwede T. Toward the estimation of the absolute quality of individual protein structure models. Bioinforma Oxf Engl. 2011 Feb 1;27(3):343–50.
- 39. Wiederstein M, Sippl MJ. ProSA-web: interactive web service for the recognition of errors in three-dimensional structures of proteins. Nucleic Acids Res. 2007 Jul;35(Web Server issue):W407–410. pmid:17517781
- 40. Tian W, Chen C, Lei X, Zhao J, Liang J. CASTp 3.0: computed atlas of surface topography of proteins. Nucleic Acids Res. 2018 Jul 2;46(W1):W363–7. pmid:29860391
- 41.
DeLano W.L. The PyMOL Molecular Graphics System, San Carlos. 2002, Schrödinger, LLC
- 42. Cline MS, Smoot M, Cerami E, Kuchinsky A, Landys N, Workman C, et al. Integration of biological networks and gene expression data using Cytoscape. Nat Protoc. 2007;2(10):2366–82. pmid:17947979
- 43. Assenov Y, Ramírez F, Schelhorn S-E, Lengauer T, Albrecht M. Computing topological parameters of biological networks. Bioinformatics. 2008 Jan 15;24(2):282–4. pmid:18006545
- 44. Bader GD, Hogue CW. An automated method for finding molecular complexes in large protein interaction networks. BMC Bioinformatics. 2003 Jan 13;4:2. pmid:12525261
- 45. Larkin MA, Blackshields G, Brown NP, Chenna R, McGettigan PA, McWilliam H, et al. Clustal W and Clustal X version 2.0. Bioinformatics. 2007 Nov 1;23(21):2947–8. pmid:17846036
- 46. Chaemchuen S, Rungpragayphan S, Poovorawan Y, Patarakul K. Identification of candidate host proteins that interact with LipL32, the major outer membrane protein of pathogenic Leptospira, by random phage display peptide library. Vet Microbiol. 2011 Nov 21;153(1):178–85.
- 47. La Scola B, Bui LTM, Baranton G, Khamis A, Raoult D. Partial rpoB gene sequencing for identification of Leptospira species. FEMS Microbiol Lett. 2006 Oct 1;263(2):142–7. pmid:16978348
- 48. Bourhy P, Collet L, Brisse S, Picardeau M. Leptospira mayottensis sp. nov., a pathogenic species of the genus Leptospira isolated from humans. Int J Syst Evol Microbiol. 2014;64(12):4061–7.
- 49. Nascimento ALTO, Ko AI, Martins E a. L, Monteiro-Vitorello CB, Ho PL, Haake DA, et al. Comparative genomics of two Leptospira interrogans serovars reveals novel insights into physiology and pathogenesis. J Bacteriol. 2004 Apr;186(7):2164–72. pmid:15028702
- 50. Ren S-X, Fu G, Jiang X-G, Zeng R, Miao Y-G, Xu H, et al. Unique physiological and pathogenic features of Leptospira interrogans revealed by whole-genome sequencing. Nature. 2003 Apr 24;422(6934):888–93. pmid:12712204
- 51. Anishetty S, Pulimi M, Pennathur G. Potential drug targets in Mycobacterium tuberculosis through metabolic pathway analysis. Comput Biol Chem. 2005 Oct;29(5):368–78 pmid:16213791
- 52. Bian J, Shen H, Tu Y, Yu A, Li C. The Riboswitch Regulates a Thiamine Pyrophosphate ABC Transporter of the Oral Spirochete Treponema denticola. J Bacteriol. 2011 Aug 1;193(15):3912–22. pmid:21622748
- 53. Patra KP, Choudhury B, Matthias MM, Baga S, Bandyopadhya K, Vinetz JM. Comparative analysis of lipopolysaccharides of pathogenic and intermediately pathogenic Leptospira species. BMC Microbiol. 2015 Oct 30;15(1):244.
- 54. Bulach DM, Kalambaheti T, de la Peña-Moctezuma A, Adler B. Lipopolysaccharide biosynthesis in Leptospira. J Mol Microbiol Biotechnol. 2000 Oct;2(4):375–80. pmid:11075908
- 55. Taylor PL, Blakely KM, de Leon GP, Walker JR, McArthur F, Evdokimova E, et al. Structure and function of sedoheptulose-7-phosphate isomerase, a critical enzyme for lipopolysaccharide biosynthesis and a target for antibiotic adjuvants. J Biol Chem. 2008 Feb 1;283(5):2835–45. pmid:18056714
- 56. Raetz CRH, Whitfield C. Lipopolysaccharide Endotoxins. Annu Rev Biochem. 2002 Jun 1;71(1):635–700.
- 57. Charon NW, Goldstein SF. Genetics of Motility and Chemotaxis of a Fascinating Group of Bacteria: The Spirochetes. Annu Rev Genet. 2002 Dec 1;36(1):47–73.
- 58. Liao S, Sun A, Ojcius DM, Wu S, Zhao J, Yan J. Inactivation of the fliY gene encoding a flagellar motor switch protein attenuates mobility and virulence of Leptospira interrogansstrain Lai. BMC Microbiol. 2009 Dec 9;9(1):253.
- 59. Ferla MP, Patrick WM. Bacterial methionine biosynthesis. Microbiology. 2014;160(8):1571–84.
- 60. Saint-Macary ME, Barbisan C, Gagey MJ, Frelin O, Beffa R, Lebrun MH, et al. Methionine Biosynthesis is Essential for Infection in the Rice Blast Fungus Magnaporthe oryzae. PLOS ONE. 2015 Apr 9;10(4):e0111108. pmid:25856162