Abstract Background Bacterial pneumonia is a severe respiratory infection that damages the lungs, representing one of the most serious public health problems worldwide due to its morbidity and mortality across all age groups. The challenge is further exacerbated by emergence of multidrug-resistant strains of Streptococcus pneumoniae, Staphylococcus aureus, Klebsiella pneumoniae, and Haemophilus influenzae, making pneumonia treatment increasingly difficult due to the persistent spread of antimicrobial resistance. Consequently, it is crucial to explore novel therapeutic targets to treat multidrug-resistant pneumonia. Results The current study is based on comprehensive analyses of proteo-genomics data from multidrug-resistant leading pneumonia pathogens. Its aim is to identify promising alternative anti-pneumonia drug targets. The analyses have identified 21 druggable targets enriched in pathogen-specific essential metabolic pathways, showing significant druggable potential. Among these, the nitrate reductase-A subunit-α have been prioritized as a new drug target, as there were no literature reports available about its druggability potential. Downstream analyses, including molecular docking, molecular dynamics simulations, virtual screening, and ADME analyses, predicted several potent druggable inhibitors against nitrate reductase-A subunit-α. Notably, aripiprazole and mirabegron, both established drugs from the DrugRep resource, were predicted as lead inhibitors of nitrate reductase-A subunit-α. Conclusion The current study prioritizes several druggable targets against MDR pneumonia, including the nitrate reductase-A subunit-α as a potential alternative target for developing novel anti-pneumonia therapies. Supplementary Information The online version contains supplementary material available at 10.1186/s12866-025-03883-y. Keywords: Pneumonia, Comparative proteomics, Anti-pneumonia therapeutics, Multi-drug resistance, Drug designing and development Background Pneumonia is a severe respiratory infection that damages the lungs. In cases of pneumonia, the air sacs (alveoli) in the lungs become filled with mucus and purulent material, making it difficult to breathe and reducing the ability to pick up oxygen [[38]1]. Pneumonia affects individuals of all ages; however, it poses more serious implications for immune-compromised patients, infants, and the elderly. Approximately 450 million people worldwide are affected by pneumonia, which causes around four million deaths annually [[39]2]. Pneumonia is the world's leading infectious cause of death in children. The disease is prevalent across all regions, but the morbidity rate is higher in southern Asia and sub-Saharan Africa [[40]3]. Survival rates against pneumonia have improved since the introduction of antibiotics and vaccines. However, despite these all advancement, pneumonia remains a life-threatening disease among the elderly, young, and chronically ill populations in developing countries [[41]4]. Common bacterial pneumonia complications include pleural effusion, collapsed lungs, respiratory failure, lung abscess, renal failure, bacteraemia, and sepsis [[42]5]. To date, patients have been treated empirically with broad-spectrum antibiotics, despite the fact that several causal pathogens of the disease have developed critical drug resistance [[43]6]. Multidrug resistance (MDR) pneumonia is a significant public health burden worldwide [[44]7]. Pneumonia causes around 230,000 deaths annually in Europe [[45]8]. According to Hoare and Lim [[46]9], the annual incidence rate of pneumonia among individuals aged 18- 39 years is approximately 6 cases per 1000 people. According to Anevlavis and Bouros [[47]10], community-acquired pneumonia is responsible for the death of 5.6 million Americans annually, making it the sixth leading cause of death. South East Asia has the highest incidence rate of pediatric pneumonia, with 0.36 episodes per child per year under the age of five [[48]11]. The highest rates of pediatric pneumonia is reported from India, China, and Pakistan with 43, 21, and 10 million cases, respectively [[49]12]. A study conducted by Hariri et al. [[50]13] indicates that third-generation cephalosporin, such as amoxicillin, are no longer effective in treating pneumonic infection. Therefore, it is crucial to develop alternative therapeutic strategies to address the leading multidrug-resistant (MDR) pathogens. Identifying key essential proteins shared among the leading MDR pneumonia pathogens could offer promising anti-pneumonia targets for prioritization [[51]14]. Currently, the drugs used for treating pneumonia infections exhibit multiple adverse effects, accompanied by an alarming rise in drug-resistant strains. Therefore, to address the growing problem of drug resistance, it is crucial to identify new therapeutic targets. The availability of extensive data of the whole-genome sequences of MDR pneumonia pathogens provide an opportunity to identify novel drug targets through computational genomic approaches [[52]15]. The search for potential drug targets increasingly relies on comparative genomic approaches [[53]16, [54]17]. These approaches screen the entire genome-wide proteome sets of both the host and pathogen to prioritize pathogen-specific essential proteins with therapeutic potential [[55]18]. These strategies have been implemented to prioritize therapeutic targets against leading pathogens, including Clostridium botulinum [[56]19], Rickettsia [[57]20], Neisseria gonorrhoeae [[58]21], and Shigella dysenteriae [[59]22]. The primary objective of the current study was to utilize annotated proteomics data from pathogens and hosts to identify new and promising broad-spectrum therapeutic targets against MDR pathogens responsible for bacterial pneumonia. A stringent druggability pipeline was employed to identify potential drug targets sharing homology among various MDR pneumonia-causing strains. This approach led to the exploration of several promising drug targets against MDR pneumonia, including nitrate reductase-A subunit-α which identified as the top-ranked alternative target based on druggability parameters. Results Retrieval of MDR proteome A total of 144,019 protein sequences of the MDR pneumonia pathogens, including S. aureus (n = 35,143 proteins), S. pneumonia (n = 8,233 proteins), H. influenza (n = 1,822 proteins) and K. pneumonia (n = 9,882 proteins) were obtained from NCBI. Selection of orthologous and non-paralogous proteins Comparative sequence analysis via BLASTp identified 13,588 orthologous, redundant protein sequences among S. aureus, S. pneumonia, H. influenza, and K. pneumonia (Supplementary file 1). Among these, a total of 12,412 paralogous protein sequences were discarded and the remaining 1,176 non-redundant orthologous protein sequences were retained for downstream analyses (Supplementary file 2). Identification of human host and gut microbiota non-homologous proteins Effective drugs candidate proteins need to be non-homologous to human host genome encoding proteins as well as human gut microbiome proteins to reduce the adverse effects of the therapy [[60]16]. A total of 884 human non-homologous pathogen orthologous proteins were identified (Supplementary file 3). Additionally, screening of these proteins against human gut microbiome proteome set identified 479 pathogen orthologous proteins that were non-homologous to human as well as human gut proteome sets (Supplementary file 4). Identification of pathogens essential proteins The pathogen vital proteins are catalogued in pathogen essential protein database (DEG) [[61]23]. Scanning of this repository is important for significant therapeutic target candidates’ identification. The gut microbiota non-homologous pathogens proteins were scanned against DEG database to identify pathogens key proteins essential for cellular survival. The analysis identified 358 essential proteins in the integrated dataset of pneumonia pathogens based on the threshold parameters as mentioned in the Method section (Supplementary file 5). Pathogenic virulence factors and antibiotic resistance genes identification Identification of virulence factors has been suggested as an optimistic approach for identifying therapeutic targets [[62]24]. The BLASTp scanning of VFDB database identified a total of 30 virulence factor homologs in pneumonia pathogen-essential, human non-homolog proteins set acquired from the previous steps (Supplementary file 6). However, the BLASTp-based comparative sequence scanning of these prioritized proteins against the ARG-ANNOT (Antibiotic Resistance Gene-ANNOTation) repository identified no antibiotic-resistant homologs proteins. Subcellular localization prediction The cellular localization analysis is imperative to understand protein's role within a cell. The prioritized drug-candidate proteins were subjected to subcellular localization analysis. This predicted a total of 344 cytoplasmic, 12 periplasmic, 21 inner-membrane, 1 extracellular, and 4 outer-membrane proteins (Fig. [63]1). Among these the cytoplasmic, inner membrane, and periplasmic proteins were prioritized for the downstream analysis for putative drug candidate’s identification. Fig. 1. [64]Fig. 1 [65]Open in a new tab The subcellular localization result of pathogen essential proteins, which are non-homologous to human and human host gut microbiota proteome sets DrugBank database scanning and approved targets identification The prioritized proteins enlisted in the above step were searched against the DrugBank database. Pathogen’s protein sequences showing no homology to existing drug target, catalogued in the DrugBank database, are anticipated as novel or alternative drug targets [[66]25]. Among the previously prioritized 358 selected protein sequences, 4 proteins were found homologous to already approved drug targets and were filtered out (Supplementary file 7). Metabolic pathways analysis The pathogens prioritized targets, i.e. non-homologous to human host, essential for pathogen survival, and non-homologous to DrugBank entries were subjected to pathway enrichment analysis. The KEGG Automated Annotation Server (KAAS) provides useful gene annotation through BLAST comparisons with the manually assembled KEGG GENES catalogue [[67]26]. The KO (KEGG Orthology) task is assigned to each protein in the output result. Out of 354 prioritized targets, 345 were assigned KO numbers and were identified as pathway-dependent proteins, while the remaining 9 proteins, where no KO number was assigned were identified as pathway- independent proteins. Additionally, the metabolic pathways of S. aureus, S. pneumonia, H. influenza, K. pneumonia, and human host were retrieved from KEGG and manually compared to identify pathogen unique and host–pathogen common pathways. This analysis identified 36 proteins that enriched in pathogen-specific metabolic pathways and worthy to be prioritized as therapeutic targets (Table [68]1). Table 1. MDR pneumonia pathogens unique metabolic pathway-dependent proteins identified using KEGG resource Sr.no Protein IDs MDR-pneumonia pathogens orthologous protein KO Number Metabolic pathway CELLO prediction 1 [69]ELP29309 Phosphotransacetylase [70]K00625 Methane metabolism, Methane metabolism Cytoplasmic 2 [71]ELP28599 UDP-N-acetyl muramate dehydrogenase [72]K00075 Peptidoglycan biosynthesis, Peptidoglycan biosynthesis, Peptidoglycan Periplasmic; Inner membrane 3 [73]ELP28614 Preprotein translocase subunit SecA [74]K03070 Quorum sensing, Bacterial secretion system, Bacterial secretion system Cytoplasmic 4 [75]ELP28638 Phosphopyruvate hydratase [76]K01689 Methane metabolism Cytoplasmic 5 [77]ELP28247 Malate dehydrogenase [78]K00027 Two-component system Cytoplasmic 6 [79]ELP27930 UDP-N-acetylglucosamine 1-carboxyvinyltransferase [80]K00790 Peptidoglycan biosynthesis Cytoplasmic 7 [81]ELP27941 UDP-GlcNAc 2-epimerase [82]K01791 O-Antigen nucleotide sugar biosynthesis, Two-component system Cytoplasmic 8 [83]ELP27954 UDP-N-acetylglucosamine 1-carboxyvinyltransferase [84]K00790 Peptidoglycan biosynthesis Cytoplasmic 9 [85]ELP27943 Serine hydroxyl methyltransferase [86]K00600 Cyanoamino acid metabolism, Methane metabolism Cytoplasmic 10 [87]ELP27311 UDP-N-acetylglucosamine 2-epimerase [88]K01791 O-Antigen nucleotide sugar biosynthesis, Two-component system Cytoplasmic 11 [89]ELP27711 Phosphoglyceromutase [90]K01834 Methane metabolism Cytoplasmic 12 [91]ELP27126 Mannose-6-phosphate isomerase [92]K01809 O-Antigen nucleotide sugar biosynthesis Cytoplasmic 13 [93]EMR27157 Isopropylmalate isomerase small subunit [94]K01704 C5-Branched dibasic acid metabolism Cytoplasmic 14 [95]EMR19505 S-adenosylmethionine synthetase [96]K00789 Biosynthesis of various plant secondary metabolites Cytoplasmic 15 [97]EMR19980 serine hydroxyl methyltransferase [98]K00600 Cyanoamino acid metabolism, Methane metabolism Cytoplasmic 16 [99]EMR19785 UDP-N-acetylglucosamine 1-carboxyvinyltransferase [100]K00790 Peptidoglycan biosynthesis Cytoplasmic 17 [101]EMR17033 UDP-N-acetylglucosamine 2-epimerase [102]K01791 O-Antigen nucleotide sugar biosynthesis,, Two-component system Cytoplasmic 18 [103]EMR22476 Nitrate reductase-A subunit alpha [104]K00370 Two-component system Periplasmic; Inner membrane 19 MBU7086541 3-isopropylmalate dehydratase small subunit [105]K01704 C5-Branched dibasic acid metabolism Cytoplasmic 20 MBU7097181 Nitrate reductase subunit beta [106]K00371 Two-component system Periplasmic; Inner membrane 21 MBU7098010 Methionine adenosyl transferase [107]K00789 Biosynthesis of various plant secondary metabolites, Cytoplasmic 22 [108]KKW85303 Preprotein translocase subunit SecA [109]K03070 Quorum sensing, Bacterial secretion system Cytoplasmic 23 [110]KKW85116 UDP-N-acetylglucosamine 1-carboxyvinyltransferase [111]K00790 Peptidoglycan biosynthesis Cytoplasmic 24 [112]KKW84912 Phosphotransacetylase [113]K00625 Methane metabolism Cytoplasmic 25 [114]KKW84834 UDP-N-acetyl enolpyruvoyl glucosamine reductase [115]K00075 Peptidoglycan biosynthesis Cytoplasmic 26 [116]KKW83868 Aspartate kinase [117]K00928 Monobactum biosynthesis, Lysine biosynthesis Inner membrane; Cytoplasmic 27 [118]KKW83911 UDP-N-acetylglucosamine 2-epimerase [119]K01791 O-Antigen nucleotide sugar biosynthesis, Two-component system Cytoplasmic 28 [120]KKW83795 S-adenosylmethionine synthetase [121]K00789 Biosynthesis of various plant secondary metabolites Cytoplasmic 29 [122]TWV02058 succinate–CoA ligase subunit alpha [123]K01902 C5-Branched dibasic acid metabolism Inner membrane; Cytoplasmic 30 [124]TWV02173 UDP-N-acetylglucosamine 1-carboxyvinyltransferase [125]K00790 Peptidoglycan biosynthesis Inner membrane; Cytoplasmic 31 [126]TWV02252 3-isopropylmalate dehydratase small subunit [127]K01704 C5-Branched dibasic acid metabolism Cytoplasmic 32 [128]TWV02259 6-phosphofructokinase [129]K00850 Methane metabolism Inner membrane; Cytoplasmic 33 [130]TWV02309 Phosphopyruvate hydratase [131]K01689 Methane metabolism Cytoplasmic 34 [132]TWV02353 Serine hydroxyl methyltransferase [133]K00600 Cyanoamino acid metabolism, Methane metabolism, Cytoplasmic 35 [134]TWV02469 Acetyl-CoA C-acetyltransferase [135]K00626 Benzoate degradation, Two-component system, Cytoplasmic 36 [136]TWV01756 Aspartate-ammonia ligase [137]K01914 Cyanoamino acid metabolism Cytoplasmic [138]Open in a new tab Structure homologs search and druggability analyses The prioritized protein targets enriched in pathogen unique metabolic pathways were screened against the PDB database to examine their 3-dimensinal (3D) structure availability. The PDB scanning identified that the structural information of these proteins from MDR pneumonia-causing pathogens are not yet reported. Therefore, the structures of these shortlisted proteins were modelled using homology modelling based on their close structural homologous templates, acquired from PDB database. Total 21 proteins of MDR pneumonia-causing pathogens were shortlisted for druggability analyses. The Swiss model server was used to predict the 3D structures of these prioritized proteins, and the ERRAT tool verified the structure models accuracy with a quality factor value of > 85%. The Ramachandran plot additionally verified the validity of the protein 3D structure models by predicting the existence of > 85 to 90% of their residues in the allowed regions of the plot [[139]27]. The capability of a protein pocket to feasibly and strongly bind drug-like compounds is of paramount importance in the characterization of druggable protein targets [[140]28]. Among the shortlisted drug candidates, 3 proteins were predicted to have no druggable binding pockets (Table [141]2), whereas the remaining 18 proteins were found to have potential binding pockets with Pockdrug druggable probability scores of > 0.5 [[142]29]. The shortlisted proteins were further subjected to PPIs analysis using STRING v10.5 database, to prioritize druggable targets based on their function as hub proteins [[143]30]. Significant hub proteins were identified based on a node degree (K) value of ≥ 5, indicating the number of resulting interactions (Fig. [144]2). Among 21 selected druggable targets, 17 were identified as hub proteins (Table [145]2). Table 2. Druggability analyses of unique pathway-dependent essential and human host non-homologous proteins of MDR pneumonia pathogens NCBI accession numbers PDB homolog ID’s Orthologous protein Molecular weight (Dalton) ERRAT quality factor QMEAN > −4 Ramachandran plot (residues in favoured region %) Node degree (K) > 5 (STRING analysis) Pock drug score > 0.5 (Residues in pocket) [146]ELP29309^a 4e4r.1.A Phosphotransacetylase 34,951.74 97.7813 1.44 98.15% 9.09 0.94 (14) [147]ELP28599 1hsk.1.A UDP-N-acetylmuramate dehydrogenase 33,797.43 97.9381 1.39 98.01% 8.18 No pocket [148]ELP28614 1tf2.1.A Preprotein translocase subunit SecA 95,960.07 95.7784 −1.92 94.51% 10 1 (14) [149]ELP28638 5boe.1.A Phosphopyruvate hydratase 47,116.95 97.164 0.48 96.52% 8.91 No pocket [150]ELP27930 3sg1.1.A UDP-N-acetylglucosamine 1-carboxyvinyltransferase 44,995.72 90.8642 −1.57 95.70% 7.64 0.75 (11) [151]ELP27941 5enz.1.A UDP-GlcNAc 2-epimerase 42,412.71 94.7945 0.43 98.12% 4.55 0.98 (8) [152]ELP27943^a 1yjz.1.A Serine hydroxymethyltransferase 45,172.29 93.9623 0.19 95.79% 7.45 0.99 (18) [153]EMR19505^a 7lo2.1.A S-adenosylmethionine synthetase 41,907.57 93.2065 0.67 97.36% 6 0.97 (19) [154]EMR19980^a 1dfo.1.A S-adenosylmethionine synthetase 45,439.81 98.2843 1.21 96.62% 7.45 0.98 (15) [155]EMR19785 4r7u.1.A UDP-N-acetylglucosamine 1-carboxyvinyltransferase 44,585.26 96.7081 −0.74 96.15% 6.91 0.99 (16) [156]EMR17033^a 1f6d.1.A UDP-N-acetylglucosamine 2-epimerase 41,820.13 96.1905 0.09 96.52% 6.91 0.99 (14) [157]EMR22476^a 1q16.1.A Nitrate reductase-A subunit-α 140,618.8 91.4425 −1.62 94.77% 8.18 0.79 (23) MBU7097181 3egw.1.B Nitrate reductase-A subunit-β 59,372.49 92.8261 −2.59 93.80% 8.55 0.86 (8) [158]KKW85116^a 5wi5.1.A UDP-N-acetylglucosamine 1-carboxyvinyltransferase 45,867.89 91.9315 −0.21 97.62% 5.82 0.79 (28) [159]KKW83911^a 4fkz.1.A UDP-N-acetylglucosamine 2-epimerase 40,923.94 95.7447 0.32 96.94% 6 1 (12) [160]TWV02058^a 1jkj.1.A Succinate–CoA ligase subunit alpha 30,275.09 95.203 1.46 97.19% 7.82 0.99 (10) [161]TWV02173 4r7u.1.A UDP-N-acetylglucosamine 1-carboxyvinyltransferase 45,268.97 96.1776 −0.95 95.93% 7.27 No pocket [162]TWV02309^a 6bfz.1.A Phosphopyruvate hydratase 46,173.38 93.4679 0.87 96.31% 7.82 0.45 (15) [163]TWV02353 1dfo.1.A Serine hydroxymethyltransferase 45,976.28 98.0512 0.67 97.13% 4 0.97 (9) [164]TWV02469 5f0v.1.A Acetyl-CoA C-acetyltransferase 40,801.15 91.5559 0.49 96.86% 4.91 0.99 (25) [165]TWV01756 11as.1.A Aspartate–ammonia ligase 37,433.8 90.2208 −1.6 92.92% 4.25 0.98 (7) [166]Open in a new tab ^a top-ten druggable protein targets enriched in pathogen-specific metabolic pathways Fig. 2. [167]Fig. 2 [168]Open in a new tab Protein–protein interaction network of top-prioritized druggable protein targets predicted by STRING tool. The query protein is indicated by the red colour. A Phosphotransacetylase B UDP-N-acetylmuramate dehydrogenase C Preprotein translocase subunit SecA D Phosphopyruvate hydratase E. Nitrate reductase-A subunit-α F. Nitrate reductase-A subunit-β Virtual screening Among the shortlisted drug target candidate proteins, nitrate reductase-A subunit-α was prioritized for virtual screening based on druggability analysis, capability to anchor small drug-like molecules, and availability of comparatively more accurate homolog 3D structure. Based on DrugRep scores, top ten -hit compounds were shortlisted against nitrate reductase-A subunit-α by screening the approved drugs library, experimental-stage chemical drugs, and traditional Chinese medicine repositories of the database. The top hits were subjected to molecular docking analysis against nitrate reductase-A subunit-α target (Fig. S1a-e, S2a-e). The docking analysis predicted that all the top-screened compounds exhibited substantial molecular interactions with the receptor protein (Table [169]3). The approved drugs Mirbegron and Aripiprazole have been recognized as lead compounds targeting the nitrate reductase-A subunit-α, and there is potential to repurpose these drugs against MDR-pneumonia. The 3D structure model of nitrate reductase-A subunit-α and the biochemical interactions of the top-ranked drug-like compound against its significant binding pocket are illustrated in Fig. [170]3. Table 3. The top ten hit compounds predicted to inhibit nitrate reductase A-subunit-α were identified during virtual screening of DrugRep resource. The CB-Dock2 score for each compound was determined using the CB-Dock2 server S.no Pubchem CIDs Compound name Library DrugRep score CB-Dock2 score 1 21,081,761 1-(4-fluorophenyl)-N-[3-fluoro-4-(1H-pyrrolo[2,3-b]pyridin-4-yloxy)phen yl]−2-oxo-1,2-dihydropyridine-3-carboxamide Experimental drug −11 −11.4 2 90,376,256 Mirabegron Approved drug −10.3 −9.5 3 5,289,566 5-(5-(4-(5-hydro-4-methyl-2-oxazolyl)phenoxy)pentyl)−3-methyl isoxazole Experimental drug −9.7 −10.2 4 14,777,879 Tricoumaroyl spermidine Traditional Chinese medicine −9.6 −9.1 5 1786 5-(5-(4-(4,5-dihydro-2-oxazoly)phenoxy)pentyl)−3-methyl osoxazole Experimental drug −9.5 −9.9 6 445,650 4-(4-STYRYL-PHENYLCARBAMOYL)-BUTYRIC ACID Experimental drug −9.4 −10.1 7 6,470,990 3-fluoro-N-[3-(1H-tetrazol-5-yl)phenyl]benzamide Experimental drug −9.4 −10.5 8 60,795 Aripiprazole Approved drug −9.3 −9.4 9 446,665 [(4-{4-[4-(Difluoro-Phosphono-Methyl)-Phenyl]-Butyl}-Phenyl)-Difluoro-M ethyl]-Phosphonic Acid Experimental drug −9.3 −9.9 10 1788 5-(5-(6-CHLORO-4-(4,5-DIHYDRO-2-OXAZOLYL)PHENOXY)PENTYL)−3-METHYL ISOXAZOLE Experimental drug −9.3 −9.6 [171]Open in a new tab Fig. 3. [172]Fig. 3 [173]Open in a new tab A The 3D structure of nitrate reductase-A subunit-α, modelled via the Swiss Model resource, followed by model accuracy validation via Ramachandran plot analysis (Table [174]2) (B) Molecular binding interactions of the top-hit docked compound (pubchem CID: 21,081,761) within the binding site of nitrate reductase-A subunit-α ADME analysis of the shortlisted compounds The ADME molecular properties of drug-like small molecules are essential for the effective design, synthesis, and clinical use of drugs. Absorption, distribution, metabolism, and excretion are the preliminary important pharmacokinetics parameters. A lead compound needs to fulfil the ADME criteria to be an effective drug [[175]31]. The pharmacokinetic properties of the top ten hit compounds were calculated using chemical descriptors based on physiological properties and chemical structures. To evaluate the efficacy of these compounds, a number of physicochemical properties were calculated, including molecular weight, hydrogen bonding, hydrophobicity, reactivity, bioavailability, molecular stability, aqueous solubility values (logP and logS), skin permeability coefficient (log kp), gastrointestinal tract absorption (GI), and blood–brain barrier (BBB) (Table [176]4). Besides, the Lipinski rule of five determines a suitable drug feature based on the permeability and solubility properties of a compound [[177]32]. All the compounds shortlisted in current study against nitrate reductase A-subunit-α, except C4, were found to have a molecular weight of < 500 Da, which is optimal for orally administered drugs (Table [178]4). Likewise, all the lead compounds were found to follow the Lipinski rule, except for C9. The compounds C1-C3, C5-C8, and C10 exhibited high gastrointestinal absorption, whereas, the C4 and C9 compounds showed low gastrointestinal absorption. Most of the compounds exhibited low penetrability through the blood–brain barrier (BBB), except, C3, C5, C8, and C10 compounds, which are predicted to cross this barrier. However, additional chemical modifications may enhance the drug-like properties of these lead compounds. Compound C1, C3-C7, C9, and C10 were not predicted as p-glycoprotein substrates, whereas the C2 and C8 compounds were identified as p-glycoprotein substrates. Identifying a drug as a P-glycoprotein substrate have consequences for both its pharmacokinetics and therapeutic effectiveness [[179]14]. It is noteworthy that none of these top hits were predicted to exhibit ADME toxicity or mutagenicity and have the potential to be effective inhibitors of nitrate reductase subunit-α. Table 4. The ADME and pharmacokinetics properties, drug-likeness, and medicinal chemistry friendliness of top ten lead compounds predicted as effective inhibitors against nitrate reductase-A subunit-α Compound C1 C2 C3 C4 C5 C6 C7 C8 C9 C10 Molecular weight (g/mol) 458.42 396.51 328.41 583.67 314.38 309.36 283.26 448.39 470.29 348.82 Hydrogen bond acceptors (HBA) 6 4 5 6 5 3 4 3 10 5 Hydrogen bond donors (HBD) 2 4 0 5 0 2 2 1 4 0 Rotatable Bonds (RB) 6 10 8 18 8 8 5 7 9 8 Topological Polar Surface Area (TPSA) (Å^2) 89.01 128.51 56.85 139.20 56.85 66.40 83.56 44.81 134.68 56.85 LogP(cLogP) 4.17 2.43 3.90 4.00 3.63 3.23 2.25 4.21 3.13 4.13 Molar Refractivity 122.34 113.16 97.05 168.61 92.24 92.32 73.79 133.26 102.19 97.25 GI absorption High High High Low High High High High Low High Log S −5.46 Moderately soluble −3.40 Soluble −4.04 Moderately soluble −5.48 Moderately soluble −3.72 Soluble −3.61 Soluble −3.24 Soluble −5.38 Moderately soluble −4.13 Moderately soluble −4.31 Moderately soluble BBB permeant No No Yes No Yes Yes No Yes No Yes P-gp substrate No Yes No No No No No Yes No No log Kp cm/s −6.18 −7.23 −5.65 −6.59 −5.85 −5.94 −6.56 −5.74 −7.29 −5.61 Druglikeness Based on Lipinski rule Yes Yes Yes Yes; 1 violation: MW > 500 Yes Yes Yes Yes Yes Yes Bioavailability Score 0.55 0.55 0.55 0.55 0.55 0.85 0.56 0.55 0.55 0.55 PAINS (alert) 0 0 0 0 0 0 0 0 0 0 Brenk (alert) 0 0 0 1 alert: michael_acceptor_1 0 0 0 0 1 alert: phosphor 0 Synthetic accessibility 2.98 3.52 4.03 3.92 3.47 2.08 1.97 3.15 3.47 3.54 [180]Open in a new tab Molecular dynamics simulations The iMODS tool was used to assess the collective functional mobility of atoms in a macromolecule complex of nitrate reductase-A subunit-α with top 4 druglike compounds, using the normal modes analysis (NMA) method. In order to evaluate the molecular flexibility of a macromolecule in a cellular environment, each normal mode has a frequency that coincides with the relative motion amplitude and deformation vector, establishing the direction of the atomic displacement of the macromolecule [[181]33]. The NMA results for the nitrate reductase-A subunit-α with top 4 druglike compounds docked complexes are shown in Fig. S3-S6. The root mean square deviation (RMSD) of the complex structures were reduced by repeatedly deforming the input structures along the lowest modes while superimposing the two structures both locally and globally to attain potential transitional simulations. The total atomic displacements at each individual atomic site across all the residue modes are a measure of the main-chain deformability of the complex structure. The flexible portions of the protein are shown by the peaks in the complex's deformability graph. Low values indicate the hard parts of the main-chain residues, whereas high values reveal the flexible parts (hinges/linkers) of the chain (Fig. S3A-S6A). The NMA-derived B-factor was used to calculate the amplitudes of the atomic displacements of the molecular complex around the equilibrium conformation for each protein–ligand complex. The resultant B-factor graphs represented the average RMSD values, showing the relationship between the NMA mobility of the protein–ligand complex and the PDB scores for each druglike compound (Fig. S3B-S6B). The eigenvalue corresponding to every normal mode expresses the motion stiffness and is proportional to the energy needed to deform a structure. The eigenvalues were calculated to be approximately 2.599634e-06 for all the complexes, representing significant structural stability of the protein–ligand complexes (Fig. S3C-S6C). Furthermore, each normal mode of the protein–ligand complexes was connected to the variance graph, representing the individual (purple) and cumulative (green) variances, in an inverse relationship with the eigenvalue (Fig. S3D-S6D). The covariance maps of all the complexes showed connectivity between pairs of residues in the individual system. Covariance analysis represented the atomic motions in the dynamic portions of the complex molecules as correlated (shown in red), uncorrelated (shown in white), or anti-correlated (shown in blue) regions in the protein–ligand complexes (Fig. S3E-S6E). The correlation matrix was computed by Ichiye and Karplus, 1991 [[182]34] using Eq. 2 and Cα Cartesian coordinates. The elastic network model of the complexes characterized the interactions between the atoms. Each dot in the graph represents a spring that connects the matching two atoms. The colour of the dots indicates their stiffness; deeper grey dots represent stiffer regions, whereas the lighter dots signify more flexible sections (Fig. S3F-S6F). The overall NMA analysis did not show significant difference in performance in terms of stability and flexibility. All the complexes show almost same results. The atomic conformation pattern was almost the same for all the complexes, which exhibited stability after the docking with the drug-like molecules [[183]35]. Discussion The pathogens causing MDR-pneumonia are resistant to several drugs, making the treatment of pneumonia infections more challenging. This leads to an increase in mortality worldwide due to pneumonia infection. It is vital to design and develop alternative therapeutic strategies to combat multi-drug resistance-mediated pneumonia [[184]36]. The current study is based on the integrated proteome data analyses of MDR-pneumonia pathogens, including S. pneumoniae, H. influenzae, S. aureus, and K. pneumonia, to address promising anti-pneumonia therapeutic targets. A subtractive proteomic analysis approach was utilised to identify pathogen essential proteins, i.e. non-homologous to human and gut microbiota proteins. During the PPI analysis, a vast majority of drug candidates were identified as hub proteins and previously such proteins have been reported as suitable drug candidates [[185]37, [186]38]. Based on the centrality rule, if a protein target of pathogen is disrupted, it could potentially be fatal for its survival [[187]39, [188]40]. The druggable proteins targets were further evaluated according to a stringent druggability criteria. Nitrate reductase-A subunit-α was recognised as a potential novel therapeutic target among the drug candidates shortlisted in this study. Nitrate reductase enzyme complex enables the bacteria to use nitrate as an electron acceptor during anaerobic growth [[189]41]. Within the enzyme complex, the nitrate reduction occurs in the nitrate reductase A-subunit-α (narG) that uses a Mo-bisMGD cofactor (bisMGD) for its catalytic activity [[190]42]. Previously, nitrate reductase subunit-β has been reported as a potent drug target against infectious Brucella melitensis strains [[191]43]; however, the nitrate reductase A-subunit-α have not been previously reported as a drug candidate. Therefore, we speculate that nitrate reductase A-subunit-α is a worthy target to be focused against MDR pneumonia. There are currently no commercial drugs available based on the narG inhibition pathway. We here prioritized the nitrate reductase A-subunit-α as feasible target for drug designing and novel inhibitors prediction based on its druggability features, including pocket druggability and the structural validation scores. Furthermore, a comparatively more accurate 3D structure homolog of nitrate reductase A-subunit-α is available from PDB to pursue the analysis of pharmacophore-based virtual screening. The analyses of virtual screening and ADME profiling shortlisted several compounds as potent inhibitors against nitrate reductase A-subunit-α, including Mirabegron and aripiprazole. Aripiprazole is a second-generation antipsychotic medication used to treat schizophrenia, mania associated with bipolar I disorder, irritability associated with an autism spectrum disorder, disjunctive therapy in major depressive disorder, and Tourette syndrome [[192]44]. Likewise, mirabegron is a sympathomimetic class of drug and β3-adrenoreceptor agonist, is an option to antimuscarinics for treating overactive bladder (OAB) [[193]45]. Repurposing of such approved drugs may accelerate the drug discovery process and provide new alternative solutions against the MDR-associated pneumonia. Clinical and experimental follow-up is required to validate these findings. Besides, several experimental drugs, including 5-(5-(4-(5-hydro-4-methyl-2-oxazolyl)phenoxy)pentyl)−3-methyl isoxazole, 5-(5-(4-(4,5-dihydro-2-oxazoly)phenoxy)pentyl)−3-methyl osoxazole, and 5-(5-(6-CHLORO-4-(4,5-DIHYDRO-2-OXAZOLYL)PHENOXY)PENTYL)−3-METHYL ISOXAZOLE were also predicted as highest-ranked molecules among the top ten hit compounds that possibly inhibit nitrate reductase A-subunit-α target. However, some of these lead compounds need additional chemical modification to enhance their drug-like potential. In addition, the protein phosphotransacetylase (PTA) was also predicted as a potential druggable target against MDR-pneumonia This enzyme plays a role in the metabolic processes involving acetate, taurine and hypotaurine, pyruvate, and propanoate [[194]46, [195]47]. PTA inhibition has been reported as a plausible therapeutic strategy to combat S. aureus-mediated infection [[196]48]. However, PTA has not been reported so far as a potential drug target against MDR-pneumonia bacterial pathogens and may therefore be worthy in future studies. Likewise, the protein S-adenosylmethionine synthetase was predicted among short-listed anti-pneumonia druggable proteins. This enzyme plays a role in the biosynthesis of polyamines, i.e., crucial for a variety of cellular processes that influence growth and development [[197]49]. The enzyme UDP-N-acetylglucosamine 2-epimerase was identified as a putative drug target during the current analysis. UDP-N-acetylglucosamine 2-epimerase is responsible for catalysing the reversible epimerization of UDP-N-acetylglucosamine and produces UDP-N-acetylmannosamine (UDP-ManNAc), which is an active donor of ManNAc residues. ManNAc is necessary for a wide variety of bacterial operations, one of which is the formation of antiphagocytic capsular polysaccharides responsible for the virulence of pathogens [[198]50]. So far, the UDP-N-acetylglucosamine 2-epimerase has not been examined as a potential therapeutic target against MDR-pneumonia. Similarly, the UDP-N-acetylglucosamine 1-carboxyvinyltransferase was identified as a putative drug target. This enzyme is involved in the biosynthesis of peptidoglycan polymer, i.e. essential component of bacterial cell walls. In addition, this enzyme plays a role in the metabolic processes involving both amino sugar and nucleotide sugar. The UDP-N-acetylglucosamine 1-carboxyvinyltransferase has been suggested as a possible therapeutic target against bacteria-mediated infection in several studies [[199]51–[200]54]. The current analysis identified succinate-CoA ligase subunit-α (ADP producing) as a potent drug target. During anaerobic growth, when the oxidative pathway from 2-oxoglutarate is suppressed, the succinate-CoA ligase participates in production of succinyl-CoA [[201]54]. Earlier studies have not reported succinate-CoA ligase subunit-α as a potential pharmacological target against MDR bacterial pneumonia. Similarly, the serine hydroxymethyltransferase (SHMT) have also been shortlisted as a putative drug target in the current analysis. This enzyme plays a role in folate-mediated one-carbon metabolism. The folate-mediated one-carbon metabolism is a fundamental cellular process that transfers one-carbon units to multiple biochemical pathways, including the biosynthesis of purine and thymidine, homeostasis of amino acids, and epigenetic maintenance [[202]55]. This enzyme has been reported as an effective drug target against parasites, viruses, and cancer [[203]56]. Focusing on SHMT as a drug target might be worth in treating MDR bacterial pneumonia due to its essential role in cell proliferation. Additionally, phosphopyruvate hydratase, known as enolase, is predicted as the potential anti-pneumonia target in the current analyses. Enolase plays a key role in carbon metabolism and participates in cell wall formation, RNA turnover, and as a plasminogen receptor. Previously enolase have been reported as a potential target against gram-negative pathogens [[204]57]. The in-silico methods utilised in the current study could provide a means to discover new therapeutic targets and produce broad-spectrum multi-species medications that may be effective in combating pneumonia infections. The druggable proteins prioritized in the current study are necessary for the survival, development, and virulence of multidrug-resistant pneumonia pathogens within the host. These prioritized proteins might be potent candidates to be utilized against MDR-pneumonia infection combat. The 3D structure exploration of druggable proteins prioritized in current study could potentially provide a framework for the development of effective anti-pneumonia drugs. Follow-up validation experiments and future research on the drug targets identified in the current study will be relevant and needed for planning effective strategies to combat MDR pneumonia. Conclusion In the current study, the entire proteome data of multidrug resistant pneumonic pathogens were used to determine new multispecies drug targets against MDR lead pathogens involved in pneumonia. Subtractive and comparative genomic methods were followed to predict pathogens unique pathways essential proteins that are crucial for virulence, survival, and growth of the pathogens. In-silico druggability studies have ranked several new multispecies drug targets against MDR pathogens that have not been focused to treat pneumonia disease. The nitrate reductase-A subunit-α was identified as top-ranked druggable target. Structure-based virtual screening, ADME characteristics, and docking analysis ranked several drug-like compounds against nitrate reductase-A subunit-α, including aripiprazole and mirabegron. These drug-like compounds could be worthy to assess in future for anti-pneumonia potential. Experimental validation through in vitro and in vivo studies is necessary to confirm the findings of this study. Methodology The current study's methodological layout is illustrated in Fig. [205]4. The subsequent analysis parameters are detailed step by step below. Fig. 4. [206]Fig. 4 [207]Open in a new tab The stepwise workflow diagram adopted for identification of novel druggable targets and small inhibitor molecules identification against MDR lead pathogens causing bacterial pneumonia MDR genomes retrieval A total of 36 complete genomes of clinical MDR strains of pneumonia pathogens, including (S. aureus (n = 13), S. pneumoniae (n = 4), H. influenza (n = 1), and K. pneumoniae (n = 18), based on literature information were obtained (Supplementary file S8) from PubMed and NCBI till July 2022 [[208]58]. Orthologous and non-paralogous sequences identification Orthologous protein sequences from different species of pathogen genomes were identified using the BLASTp tool [[209]59] with a threshold cut-off of E-value ≤ 10^–4, bit score ≥ 100, sequence identity ≥ 50%. The resultant orthologous sequences were subjected to CD-hit clustering analysis to remove paralogous sequences with a threshold of 80% sequence identity [[210]60]. The non-paralogous protein sequences were then analysed downstream. Human host non-homologues proteins identification The non-paralogous pathogen proteins were analysed against human proteome using the BLASTp program to identify human non-homologous proteins. The resultant proteins were further analysed using BLASTp against the human gut microbiota proteome data to identify pathogen proteins that were non-homologous to human gut microbiota proteins. A BLASTp threshold of bit score ≤ 100, E-value ≥ 10^–4, and sequence identity ≤ 35% was set during these comparative sequence analyses. Prioritization of pathogens cellular essential proteins An ideal drug target is based on a protein that is indispensable for pathogen survival. The bacterial essential genes are catalogued in the Database of Essential Genes (DEG) [[211]61]. A shortlisted pathogen protein, non-homologous to human as well as human gut microbiota proteins, were further screened against the DEG [[212]60] following the BLASTp parameters of E-value ≤ 10^–4, bit score ≥ 100, and sequence identity ≥ 50. Identification of virulence and antibiotic-resistant proteins Bacterial pathogens secrete virulence factors that lead to infections in the host. These virulence factors are catalogued in the VFDB database [[213]24]. We scanned the VFDB repository data via comparatively sequence analysis to identify pneumonia-pathogens virulence factor proteins. The pathogenic microbes’ drug-resistant genes are catalogued in the ARG-ANNOT resource [[214]62]. The antibiotic-resistant proteins from compiled proteome set of pneumonia-causing pathogens were identified and excluded, by scanning the ARG-ANNOT database via the BLASTp criteria of E-value ≤ 10^–4[,] bit-score ≥ 100, and sequence identity ≥ 50%. Prediction of proteins subcellular localization The subcellular localization of pathogen essential, human host non-homologous proteins was identified using the CELLO2GO webserver [[215]63]. CELLO2GO classifies proteins based on their cellular localization features, including the outer membrane, periplasmic, inner membrane, cytoplasmic, and extracellular regions. We prioritized the cytoplasmic, periplasmic, and membrane proteins of pneumonia-causing pathogens for subsequent analyses. DrugBank database scanning The DrugBank database [[216]64] was screened against the pneumonia-causing pathogens shortlisted proteins. The step is important to identify whether the prioritized putative druggable target is new or already been catalogued. Homology-based screening of DrugBank database against prioritized anti-pneumonia targets was performed following the BLASTp parameters of sequence identity ≤ 35%, bit score ≤ 100, and E-value ≥ 10^–4. Identification of target enrichment in pathogen-specific metabolic pathways The metabolic pathway enrichment of pathogen shortlisted targets was identified from the KAAS (KEGG Automatic annotation server). The server [[217]26] assigns KO numbers to shortlisted proteins. Prioritized protein enrichment in pathogens specific pathways was checked manually via comparison with human host pathways. The druggable proteins enriched in the pathogens’ unique metabolic pathways were prioritized in downstream analysis to identify suitable therapeutic targets. Druggability analysis Prioritized proteins targets were assessed for druggability potential. The potential binding pockets in the prioritized targets were identified using druggable scores calculated by PockDrug server [[218]28]. Protein–protein interaction (PPI) analysis was carried out using the STRING database [[219]30] with default parameters. Hub proteins were identified based on the value of node degree (K ≥ 5). Knocking out of hub proteins is reported to be lethal for pathogen's survival and hence considered as ideal target. Homology modelling and structure validation The shortlisted pathogen proteins were searched against the PDB database to assess the availability of their 3D structures [[220]65]. In case where the exact 3D structure information is not available in PDB, it was modelled using the Swiss model based on a closely related homologous template through homology modelling approach [[221]66]. ERRAT tool and RAM-PAGE server were used to check the accuracy of 3D modelled structures [[222]67]. Virtual screening The DrugRep server was used for the structure-based virtual screening of the chemical libraries to identify potential inhibitors against a prioritized drug target. DrugRep is a well-organized resource for drug repurposing, containing approved drug libraries, experimental-stage chemical drugs, and traditional Chinese medicines [[223]68]. The server automatically detects possible binding pockets of the receptor, and then performs batch docking of the compounds by employing the AutoDock Vina program. After virtual screening, the binding affinity of the top ten lead compounds with prioritized target was examined using the CB-Dock2 server [[224]69]. Discovery Studio Visualizer v.4.5 was used for protein–ligand visualization and interaction analysis [[225]70]. Drug likeliness and ADME analysis The physiochemical and pharmacokinetic properties, including absorption, distribution, metabolism, and excretion features of the top-ten lead compounds were examined using the SwissADME tool, adhering to the default parameters [[226]71]. ADME furnishes valuable information about a compound's drug-like properties. Molecular Dynamic Simulation The molecular dynamic simulation investigation of the nitrate reductase-A subunit-α with the top 4 drug-like compounds was performed using the iMODS web server [[227]72]. The server offers an enhanced affine-model-based arrow representation of macromolecular complex domain dynamics, specifies possible conformational changes, and finds elastic network models and resolutions utilizing a range of coarse-grained atomic representations. The service computes the structural dynamics of proteins and docked protein complexes with other proteins and ligands using normal mode analysis (NMA). It also includes elastic network data, covariance maps, variance, deformability, eigenvalues, and B-factor (mobility profiles) [[228]73]. Supplementary Information [229]Supplementary Material 1^ (2.5MB, docx) [230]Supplementary Material 2^ (294.2KB, docx) [231]Supplementary Material 3^ (257.4KB, docx) [232]Supplementary Material 4^ (142KB, docx) [233]Supplementary Material 5^ (112KB, docx) [234]Supplementary Material 6^ (20KB, docx) [235]Supplementary Material 7^ (112.7KB, docx) [236]Supplementary Material 8^ (23.1KB, docx) [237]Supplementary Material 9^ (2.7MB, docx) Acknowledgements