Introduction

Leptospirosis, a zoonosis caused by bacterial species belonging to the genus Leptospira, phylum Spirochaetota, is one of the neglected tropical diseases posing global health concerns. Leptospires have been classified into pathogenic, intermediate, and saprophytic groups, and more recently into clades P1, P2, S1, and S2, in which P1/pathogenic clade accommodates important pathogenic species related to leptospirosis1,2,3. Pathogenic Leptospira strains are highly virulent and can infect humans4 as well as domesticated animals such as dog5, horse6, swine7 and cattle8. Leptospira has been detected in wild animals such as rats4,9, capybara10, water vole11, hamster10 and tiger12. Reservoir animals’ urine contributes to the transmission of leptospirosis, and both urban areas and open fields carry subclinical Leptospira10,12. Species distinction of causative agents is crucial in understanding the epidemiology of leptospirosis. Leptospira interrogans RGAT (= ATCC 43642 T)13 and L. kirschneri 3522CT14 are two species of Leptospira commonly infecting humans4, and share a close phylogenetic neighborhood. These two species can be distinguished from each other as they share 91.3% and 42.3% Orthologous average nucleotide identity (OrthoANI) and digital DNA–DNA hybridization (dDDH), which were well below the threshold (95‒96.0% and 70%) recommended for the species delineation15.

The genomic analysis provided evidence for potential horizontal gene transfer in Leptospira species16. Leptospira harbors genes for virulence include ligA, lolC, lolD, lipA, fliG, and fliF; and antimicrobial resistance such as ermA, _tetA and mdtA17. Resistance to sulfonamides, neomycin, actidione, polymyxin, nalidixic acid, vancomycin and rifampicin has been reported in Leptospira18. The biofilm formers require a higher minimum inhibitory concentration (MIC) of penicillin, fluoroquinolone, and tetracycline as compared to its planktonic forms5. L. interrogans forms a biofilm in the renal tubules of Rattus norvegicus9. Biofilm formation enhances antimicrobial resistance19,20 and drives chronic colonization.

A Leptospira-infected tiger displayed diffuse tan foci on the liver parenchyma and jaundice12. Leptospirosis was attributed to abortion in swine7 and horse17. In humans, renal insufficiency, hepatic dysfunction, and pulmonary failure are the pathological complications associated with leptospirosis21. In addition to their ability to evade or hijack the immune system, the spiral shape and endoflagellar motility have been regarded as key traits that contribute to the success of leptopires as pathogens22. Our understanding of the pathogenesis of Leptospira is still limited despite the discovery of a causative spirochaetal agent > 100 years ago23. In particular, the biology and virulence factors of leptospires still await further exploration for strategic therapeutic intervention with minimal antibiotic usage.

Earlier studies have emphasized the environmental persistence of L. interrogans and the associated risks to human health24. Leptospira species without clearly established virulence features can inhabit soil and surface waters2. Once excreted via animal urine, leptospires require moisture or a water body to survive and stay infectious25. We hypothesized that the sewage water originating from a wildlife enclosure could serve as a reservoir for the pathogenic L. interrogans. We tested the hypothesis by isolating a pure culture from the sewage and carrying out a detailed phenotypic, genotypic and biochemical characterization. We further trace a compound that can retard the growth of L. interrogans when provided in aqueous and volatile form.

Results

Morphology of KeTo

KeTo appeared Gram-staining-negative, thin and filamentous organisms under 100X oil immersion objective, whereas staining with acridine orange showed bright green slender organisms against a dark background, arranged in clusters and singles with a typical hooked-end appearance of the genus (Fig. 1a‒b). Transmission electron microscopy showed the spiral elongated structure of the cells of KeTo in aggregates (Fig. 1c‒d).

Fig. 1
figure 1

Morphological analysis of Leptopira interrogans KeTo. Light (a), fluorescence (b) and transmission electron microscopic images (c, d) of KeTo. Cells were grown in enriched liquid EMJH for 5 days at 30 °C under darkness at shake cultures (120 rpm). Asterisks and arrows represent aggregates and single cells of KeTo, respectively. Scale bar in (a) and (b), 100 µm. Scale bar in (c) and (d), 500 nm and 200 nm, respectively.

Full size image

Molecular identification of KeTo as a new strain of L. interrogans

KeTo shared high 16S rRNA gene sequence similarity with L. interrogans RGAT (100%) followed by L. kirschneri 3522CT (99.9%) (Table S1). In addition, it showed > 99.1% 16S rRNA gene sequence similarities with 11 authentic species of Leptospira and established distinct and tight (92% bootstrap support) phyletic lineage in the cluster that accommodated both L. interrogans RGAT and L. kirschnerii 3522CT (Fig. 2a). KeTo genome containing 4,726,352 bp spanned 169 contigs having 35% GC content, 5193 coding sequences and 39 RNAs. Genome sequence analysis at TYGS revealed discrete phyletic lineage established by KeTo associated with L. interrogans RGAT 99% bootstrap support) (Fig. 2b). OrthoANI analysis showed highest genomic relatedness of KeTo to L. interrogans RGAT (= ATCC 43642T, 99.3%), followed by L. kirschneri 3522CT (91.3%) and other related species (80.8‒74.3%) (Fig. 3a). A comparative genomic map of KeTo and other related Leptospira species is shown in Fig. 3b. Based on these data, KeTo is proposed to be a new strain of L. interrogans.

Fig. 2.
figure 2

16S rRNA gene based and genome based phylogenetic analysis of Leptospira interrogans KeTo. Phylogenetic tree generated for KeTo based on manual submission of 16S rRNA gene (a) and genome data (b). Trees were generated using TYGS.

Full size image
Fig. 3
figure 3

Genomic analysis of Leptospira interrogans KeTo. Heatmap showing the genomic relatedness of KeTo with L. interrogans RGAT, L. kirschneri 3522CT and other closely related type strains of Leptospira species based on OrthoANI (a). Comparative genome map showing genomic features of KeTo and other related species (b). Genes involved in the metabolism of ketone body and ethyl acetoacetate (green fonts), and conferring virulence (red fonts) are annotated; details are summarized in Table S3 and S4. OrthoANI plots and Circular map were generated using OAT software53 and CGveiw55, respectively.

Full size image

Multidrug resistance and preferential utilization of a major ketone body in KeTo

Biolog GNIII assay was performed on KeTo to understand its nutritional requirement and drug resistance pattern. KeTo oxidized diverse organic acids available as sole carbons at GNIII (Fig. S1). Interestingly, oxidation of acetoacetate was highest as compared to other carbon sources and the oxidation response declined with time. The strain was resistant to vancomycin, rifamycin, aztreonam, lincomycin, nalidixic acid, minocycline and troleandomycin (Fig. S2). Table S2 summarizes the genes contributing to the drug resistance in KeTo. Taken together, KeTo is a multidrug-resistant strain having the potential ability for preferential utilization of acetoacetate, a major ketone body of the mammalian system.

Genomic potential for ketone body utilization in KeTo and other Leptospira

KeTo genome was mined for the genetic markers of ketone body utilization and relevant genes are annotated in Fig. 3b and listed in Table S3. A gene encoding acetoacetyl CoA synthetase (AACS), succinyl-CoA:3 oxoacid CoA transferase (SCOT) and acetoacetate decarboxylase were found in the genome (Fig. 4; Table S3). In addition, a gene encoding acetoacetate decarboxylase (ADC) that transforms acetoacetate to acetone was found; however, the genome lacked a gene for β-hydroxybutyrate dehydrogenase that catalyzes the reversible conversion of acetoacetate to β-hydroxybutyrate. KeTo genome additionally harbored genes for esterase, alcohol dehydrogenase and aldehyde dehydrogenase that catalyze EA catabolism, substantiating EA utilization in vitro. Interestingly, genes dedicated to ketone body utilization and EA metabolism were also found in the closely related type strains of the species of Leptospira. All the important genes listed Table S3 are found in all the species except for the ADH gene (fig|6666666.1062727.peg.2469), while high-identity match is only found in L. mayottensis 200901116T. Interestingly L. mayottensis 200901116T contains a region of high similarity to KeTo that is apparent in the map (Fig. S3). This finding implies the occurrence of horizontal gene transfer between KeTo and L. mayottensis 200901116T, in which latter was originated from human blood26.

Fig. 4
figure 4

Pathways involved in ketone body and ethyl acetoacetate metabolism in Leptospira interrogans KeTo. AACS, acetoacetyl-CoA synthetase; ADC, acetoacetate decarboxylase; ADH, alcohol dehydrogenase; ALDH, aldehyde dehydrogenase; SCOT, succinyl-CoA:3-oxoacid-CoA transferase.

Full size image

EA suppressed cell density and metabolic activity in KeTo

The impacts of EA and incubation time on KeTo were evaluated. The optical cell density and metabolic activity in KeTo declined significantly (p < 0.0001) with time as evidenced through the t-test analysis (Fig. S4a and b). KeTo further showed a drastic decline in cell density and metabolism when incubated with ≥ 2.5 mM EA. A significant (p < 0.1) decline in residual EA was also seen when KeTo was exposed to ≥ 2.5 mM EA for 60 h (Fig. S4c). These data highlighted a dose-dependent decline in the optical cell density and metabolic activity of KeTo despite its ability to utilize EA as a sole carbon.

The time course of cell density, Alamar blue dye reduction (viability) and media acidity/alkalization responses of KeTo in basal EMJH supplemented with EA (Bea_KeTo) as a sole carbon source were tested and the results differentiating the treatments are shown in Fig. 5a‒f. Quantification of EA using SNP-based colorimetric assay showed declined residual EA in the presence of KeTo suggesting EA degradation (Fig. 6a‒b). Classical purple ring formation during Rothera’s test, a gold standard test for the ketone body, substantiated this notion as the thickness of the ring decreased suggesting acetoacetate consumption (Fig. 6c).

Fig. 5
figure 5

Impact of ethyl acetoacetate on Leptospira interrogans KeTo. The influence of ethyl acetoacetate on optical cell density (OD) (a), metabolic activity (b), and media acidity (c) and alkalinity (d) are shown. Portion of microplates showing OD, Alamar blue (AB) dye reduction and chromogenic variations in phenol red (PR) treatment at 0h (e) and 96 h (f) are shown (top to bottom). Error bar, mean (n = 4) ± SD. B, Basal EMJH without enrichment; B_KeTo, B inoculated with KeTo; Bea, B containing ethyl acetoacetate; Bea_KeTo, B inoculated with KeTo treated with EA.

Full size image
Fig. 6
figure 6

Colorimetric determination of ethyl acetoacetate utilization in Leptospira interrogans KeTo. Time course assay showing residual EA as determined using in-house SNP assay. Error bar, mean (n = 4) ± SD (a), a corresponding section of microplate showing differential chromogenic reaction (b). Rothera’s test showing declined concentration of EA in Bea inoculated with L. interrogans KeTo (c). Bea, B containing ethyl acetoacetate; Bea_KeTo, Bea inoculated with KeTo.

Full size image

Biofilm forming ability of KeTo and the impact of EA

The impacts of nutrient availability and EA input were tested on planktonic cell and biofilm formation abilities of KeTo. Planktonic cells were significantly (p < 0.0001) high while biofilm formation was low (p < 0.0001) under nutrient enrichment as compared to their nutrient-limited counterparts (Fig. 7a‒b). EA input significantly (p < 0.0001) hampered the planktonic cells and biofilm formation, particularly under nutrient-limited conditions. Thus, KeTo can tune metabolism and colonization as per nutrient availability and EA exerts antibiofilm activity on KeTo under oligothrophic conditions.

Fig. 7
figure 7

Impact of nutrient enrichment and ethyl acetoacetate input on the planktonic cells and biofilm formation in Leptospira interrogans KeTo. Formation of planktonic cells of KeTo (bar with slant lines) in the absence of enrichment (light blue background) and the presence of enrichment (light green background) (a); Formation of biofilm of Keto (bar with slant lines) in the absence of enrichment (light blue background) and the presence of enrichment (light green background) (b); inset shows portion of microplates of corresponding treatment. The biofilm formed on microplate wells under various treatments was fixed with 99% methanol, stained with 0.1% crystal violet. The excess stain was removed by aqueous wash and dried. The dye was solubilized using 33% acetic acid and read at 570 nm. Error bar, mean (n = 6) ± SD. ****p < 0.0001; ns, non-significant. Asterisks placed over the horizontal line represents level of significance based on the t-test; *p < 0.05. 48 h old data. B, Basal EMJH without enrichment; B_KeTo, B inoculated with KeTo; Bea, B containing ethyl acetoacetate; Bea_KeTo, Bea inoculated with KeTo; E, Enriched EMJH medium; E_KeTo, E inoculated with KeTo; Eea, E containing ethyl acetoacetate; Eea_KeTo, Eea inoculated with KeTo.

Full size image

Genomic potential for virulence and influence of EA on hemolytic activity of KeTo

Genes encoding virulence factors such as hemolysin, sphingomyelinases (n = 4), heme oxygenase and LipL41-expression chaperone Lep were found in KeTo genome (Fig. 3b, Table S4). In line with the detection of hemolysin and sphingomyelinases, KeTo exhibited hemolytic activity on sheep blood agar (Fig. 8a). The growth of KeTo on semisolid EMJH agar without and with 5% (v/v) blood supplement as a function of EA exposure, when compared to respective controls are shown in Fig. 8b‒e. Exposure to EA significantly inhibited the growth of KeTo on the tested semisolid agars as compared to control (Fig. 8f). No marked change in media pH was recorded in KeTo inoculated test and control plates (Fig. 8g‒h). However, AB assay revealed a lack of metabolic activity in KeTo exposed to EA. These data substantiated the inhibitory impact of volatile EA on the growth and metabolism of KeTo.

Fig. 8
figure 8

Hemolytic behaviour (a) and impact of volatile ethyl acetoacetate on growth (b‒f) and metabolism (g) of Leptospira interrogans KeTo. Hemolytic activity on sheep blood agar on 0th (i), 48th (ii), 96th (iii), 120th (iv) and 196th hours (v) of incubation at 30°C. (a); Partition plate assay results showing the effect of EA on growth on EMJH semi-solid media without (b‒c) and with blood supplementation (d‒e); bar diagram shows corresponding quantitative data. Asterisks placed over the horizontal line represent level of significance based on one-way ANOVA (f). Partition plate without EA (g) and with EA exposure (h) loaded with Alamar blue (blue spots) and phenol red (orange spots) to probe cell viability and change in media pH, respectively (representative images). Results in b‒h were obtained after 13 days of incubation at 30°C; E_KeTo, Enriched EMJH inoculated with KeTo; Eea_KeTo, Enriched EMJH inoculated with KeTo exposed to volatile EA; EB_KeTo, 5% blood-supplemented enriched EMJH inoculated with KeTo; EBea_KeTo, 5% blood-supplemented enriched EMJH inoculated with KeTo exposed to volatile EA.

Full size image

EA susceptibility in multiple strains of L. interrogans

Leptospira interrogans serogroup Australis sv. Australis strain Ballico and L. interrogans serogroup Icterohaemorrhagiae sv. Lai Like strain AF6127 were included in addition to KeTo to evaluate possible wide-spread susceptibility to EA in Leptospira. Strains Ballico and AF61 were chosen based on their similar growth kinetics in enriched EMJH (Fig. S5a-b). These two strains were haemolytic as that of KeTo. While the drug resistance pattern of Ballico was in line with KeTo, AF61 was found to be sensitive to troleandomycin and nalidixic acid. All three strains exhibited significant (p < 0.05) reduction in the optical cell density within 48 h when provided with EA (Fig. S5c). SNP assay detected significantly (p < 0.05) lesser amount of residual EA, particularly in Ballico and AF61, while the decline was non-significant in KeTo at tested time point (Fig. 5d-e). Results of Rothera’s test further reflected the declined EA in Leptospira-inoculated samples ((Fig. 5f). Taken together, all strains degraded EA and were found susceptible to this compound.

Discussion

Leptospira interrogans is a complex pathogenic species harboring various strains/serovars/serotypes/pathovars that mutually share 100% 16S rRNA gene sequence similarities but display a remarkable disparity in phenotypic traits or host-specificity. 16S rRNA gene analysis, traditionally used for bacterial species delineation, is not the most effective method for differentiating species within the genus Leptospira2. Since the taxonomic status of KeTo was not resolved through the 16S rRNA gene sequencing, we carried out whole genome sequencing followed by OrhtoANI and dDDH, two commonly used indices to discriminate the genomic species of prokaryotes. KeTo shared high dDDH and OrthoANI values with L. interrogans RGAT (93.4% and 99.3%, respectively), which were well above the threshold (95–96 and 70% for ANI and dDDH, respectively) set for species delineation15. Genome sequences aided the detection of possible horizontal gene transfer between KeTo and L. mayottensis 200901116T, as predicted earlier in the species of Leptospira16.

The sole carbon utilization pattern of L. interrogans sv. Manilae earlier deduced through Biolog GNIII complemented specific phenotypic traits with the genomic potential28. KeTo was found to oxidize diverse organic acids, amino acids and sugars, and exhibited resistance to multiple antibiotics on Biolog GNIII. The oxidation of acetoacetate served as primary evidence for the possible ketone body utilization in KeTo. This is the first report on resistance exhibited by ketone body-oxidizing Leptospira against diverse glycopeptide, monobactam, quinolone, tetracycline and macrolide antibiotics. Antibiotics of the class tetracycline (doxycycline), penicillin/aminopenicillin (amoxicillin, ampicillin and penicillin), cephalosporin (ceftriaxone and cefotaxime) and macrolide (erythromycin) have been used to treat leptospirosis. Thirty two genes have been linked to antimicrobial resistance in Leptospira, consisting of 20 essential genes being present in the majority of strains16. Antibiotics can be pumped out of cells by certain efflux pump systems, such as those in sv. Pomona, which makes the antibiotics ineffective. Antibiotic-inactivating enzymes that stop antibiotics from working. Genetic mechanisms like efflux pump and antibiotic-inactivating enzymes can also play a pivotal role in conferring resistance. Detection of diverse drug resistant genes in KeTo (Table S2) and drug resistant pattern in vitro were alarming and required urgent countermeasures.

Our pangenomic analysis suggested co-occurrence of pathogenic traits and ketone body metabolic ability within the genus Leptospira. The quantitative variations detected in the residual EA in KeTo, Ballico and AF61 cultures suggest differential degradation of EA in the pathogenic strains of L. interrogans. However, further studies are warranted to validate active ketone body degradation and susceptibility to EA in other species in general and the saprophytic species in particular. Mammals can transform energy stored in ketone bodies to high-energy phosphates to replenish the energetic deficit elicited by starvation/reduced carbohydrate intake29. Ketone bodies are the major metabolic fuels for the respiration of extrahepatic tissues such as the brain, kidney and heart30,31,32 and it appears that KeTo, Ballico and AF61 are most likely to exploit the ketone body metabolic pathway for instant energy. The detection of SCOT catalyzing the reversible transfer of CoA from succinyl-CoA to acetoacetate as the first step of ketone body utilization33, and AACS converting acetoacetate to acetoacetyl-CoA that provides acetyl units for lipogenesis34 substantiated the ketone body utilization in Leptospira species.

Secretion of sphingomyelinase is one of the virulence factors reported in Leptospira21,35. Genes coding for hemolysin and four distinct sphingomyelinases including the most virulent sph221,27 were identified in KeTo. The hemolytic behavior of KeTo was validated through blood agar in vitro suggesting one of its possible pathogenic attributes. Biofilm formation has been reported in pathogenic and saprophytic Leptospira thriving in various biotic and abiotic environments19. Earlier reports identify antibiofilm and/or antimicrobial activities of EA against some pathogenic bacterial strains36,37. Since external (extracellular DNA) and internal factors (c-GMP based signaling) govern biofilm24,38, the possible influence of EA on these features in Leptospira needs further exploration.

There was a clear indication for degradation of EA and associated susceptibility in KeTo, Ballico and AF61 under oligotrophic conditions as evident through SNP and Rothera’s assays. Since EA did not contribute to cell biomass despite undergoing degradation, this compound is most likely a “lethal attractant” for Leptospira. EA is a thermostable food preservative39 and is less toxic to animals40. EA has an LC0 (maximum tolerable concentration) of 0.96‒2.28 mM against fishes and EC0 (the concentration at which CO2 evolution was not suppressed) of 16 mM against bacteria. The maximum dose of EA tested in our study was ~ 8 mM, in which strain-dependent variations in terms of the susceptibility was encountered. Further in vivo studies are warranted to define LC0 of EA on mammals and validate the possible application of EA as a therapeutic agent for leptospirosis.

In conclusion, the metabolic fingerprinting of L. interrogans KeTo, a novel multidrug-resistant hemolytic strain originating from a wildlife-associated sewage sample, revealed its potential ability to thrive in the presence of diverse organic acids and ketone bodies. Comparative genomics showed the widespread co-occurrence of virulence factors and ketone body utilization in Leptospira. KeTo could be a potential pathogen as its genome possesses genes encoding a hemolysin and four distinct sphingomyelinases, sharing 99‒100% amino acid sequence similarities with that of L. interrogans RGAT. The planktonic cell and biofilm-forming ability of KeTo in nutrient-rich and nutrient-limited conditions reflect dual life strategy of the organism to withstand/strive in the blood stream and urinary tract, respectively. Our data based on KeTo, Ballico and AF61 may pave the way for the further exploration of molecular basis underpinning EA-driven growth inhibition for the therapeutics of leptospirosis thereby aiding in healthcare.

Materials and methods

Chemicals, reagents and culture media

Ethyl acetoacetate (EA, 98% purity; Merck), Alamar blue (Cell-Quant™ AlamarBlue Cell Viability Reagent; ABP Biosciences), phenol red (PR; Loba Chemie), ammonia (Merck), ammonium sulphate (Loba Chemie) diethylamine (Merck) and sodium nitroprusside (SNP; CDH) were obtained from respective suppliers and used without further purification. Leptospira basal EMJH (Catalogue No. DF0794-17–1) and enrichment EMJH (Catalogue No. DF0795-73–1) were procured from BD Difco. Basal EMJH contained disodium phosphate 1 g/L, monopotassium phosphate 0.3 g/L, sodium chloride 1 g/L, ammonium chloride 0.25 g/L, thiamine 5 mg/L. Enrichment EMJH contained albumin, polysorbate 80 and additional growth factors for Leptospira. Working enriched EMJH was prepared by adding 10% (v/v) enrichment EMJH to basal EMJH.

Bacterial strains, culture conditions and growth kinetics

Strain KeTo was isolated from the sewage sample collected at Pilikula Biological Park, Moodushedde, Mangalore (12°55′41.5ʺN 74°54′01.3ʺE) on 16th October 2022. The sample was diluted tenfold using physiological saline and filtered through a 0.22 µm syringe filter (HiMedia). Subsamples (0.5 ml) were inoculated in 4.5 ml Ellinghausen McCullough Johnson Harris (EMJH) liquid media incorporated with 5-fluorouracil (5-FU, 200 μg/ml) and incubated at 30 °C under shaking culture until visible growth was seen in the form of pellets. The dose of 5-FU was within the threshold (50‒1000 µg/ml) prescribed by the Programme for Prevention and Control of Leptospirosis, National Centre for Disease Control (NCDC, India) to avoid contamination.

Leptospira interrogans serogroup Australis sv. Australis strain Ballico and L. interrogans serogroup Icterohaemorrhagiae sv. Lai Like strain AF61, procured from the National Institute of Epidemiology, Chennai27, were used as reference strains. Seed cultures were prepared in enriched liquid EMJH. An inoculum of 0.5 ml of each of the Leptospira strains was inoculated into 4.5 ml enriched EMJH liquid media supplemented with 5-FU (200 μg/ml) and incubated at 30 °C under darkness and shaking conditions (120 rpm). The cultures were periodically transferred to microplate and cell density was measured at 600 nm to assess the growth kinetics.

The quality of Leptospira culture was assessed by subculturing the strains on nutrient agar and EMJH plates in parallel. In addition, fluorescent staining with acridine orange (0.1 mg/ml) and traditional Gram’s staining were performed frequently to verify the purity of strains. No growth on nutrient agar and the presence of only Gram negative, long, slender, spiral organisms in microscopy suggested a pure culture of Leptospira and ruled out the presence of contaminants. The purity of the isolate was confirmed through 16S rRNA gene sequencing.

Biolog GNIII microplate assay for ketone body oxidation and drug resistance

Cells of all three strains were collected by centrifugation (10,000 rpm, 5 min, 4 °C) and the pellets were washed thrice with 0.9% NaCl and centrifuged (10,000 rpm, 5 min, 4 °C). Washing was necessary to remove residual enrichment media adhered to the cells. The washed cells were resuspended in inoculation fluid A (IF-A; Catalog No. 72401, Biolog), loaded with an OD of 0.54 into the wells (100 μl/well) of GNIII microplate and incubated at 37 °C for 24 to 48 h. The results were read at 590 and 750 nm using a plate reader (FLUOstar Omega microplate reader). OmniLog (absorbance at 590–750 nm) was estimated to understand the oxidation of the ketone body (acetoacetate) and antibiotic-resistant attributes.

Microscopic analysis of KeTo

Light microscopy was performed for KeTo after Gram staining (HiMedia) of heat-fixed slides using the manufacturer’s instruction and observed under 100X magnification using immersion oil. Simultaneously, heat-fixed smears were flooded with acridine orange dye, allowed to stand in dark condition for 10 min and observed under a fluorescence microscope using an excitation wavelength of 455 nm. Cell morphology was further investigated by transmission electron microscopy (JEOL JEM-1400) after negative staining of cells with 0.2% (w/v) uranyl acetate.

DNA extraction and determination of preliminary identity

Genomic DNA was isolated from the EMJH liquid culture using a QIAamp DNA mini kit according to the manufacturer’s instruction and stored at − 20 °C for molecular assays. The quality and the quantity of DNA were assessed using the combined application of gel electrophoresis and nanodrop spectrophotometry (Colibri, Titertek Berthold). The isolated DNA was used for 16S rRNA gene sequencing41,42. Primers F27 (5'-AGAGTTTGATCMTGGCTCAG-3') and R1492 (5'-TACGGYTACCTTGTTACGACTT-3') were used41,42 for the PCR amplification of 16S rRNA gene with the following set of condition: an initial denaturation at 95 °C for 2 min, and 35 cycles of denaturation at 95 °C for 30 s, annealing at 57 °C for 30 s and extension at 72 °C for 1.5 min followed by a final extension at 72 °C for 7 min. The 16S rRNA gene amplicon was seen as a sharp band at 1500 bp region under a UV transilluminator and subjected to Sanger sequencing. Forward and reverse sequences were trimmed manually, assembled using CAP343 and identified using EzBiocloud44.

Genome sequencing and bioinformatics

The genomic DNA was sequenced using illumina sequencing at Himedia, Mumbai, India. Two hundred and fifty ng of total DNA was used as input for library preparation using QIASeq FX DNA kit (Qiagen) to fragment and obtain adapter-ligated and indexed library as per manufacturer’s instructions. The indexed library was sequenced on an Illumina Miseq using a 300-cycle paired-end chemistry. FASTQC45 was used to assess the quality of raw fastq files. Quality assessment for genome assemblies generated by Spades46 and Megahit47 assemblers was carried out using the Quast48. It was noticed that the assembly generated by the Spades assembler had comparatively larger assembled contigs than megahit and with better N50 value (minimum contig length to suffice 50% of assembled genome sequence). Hence, assembly with Spades assembler was used for further downstream analysis. Genome annotation was carried out using Prokka49. Phylogenomic analysis was performed using GToTree50.

The sequence was uploaded to the Rapid Annotations using Subsystems Technology (RAST51; http://rast.nmpdr.org/rast.cgi) for automated annotation. BLAST (https://blast.ncbi.nlm.nih.gov/Blast.cgi) feature was used to manually confirm gene features of essential bio systems against a non-redundant database of the National Center for Biotechnology Information (NCBI). The genome sequence of KeTo along with 9 other closely related type strains were uploaded to Type (Strain) Genome Server (https://tygs.dsmz.de/) and compared. Phylogenetic trees were reconstructed for the 16S rRNA gene and draft genome using TYGS and LPSN52.

Putative genes encoding the acetoacetate pathway were identified using the BLAST feature of the RAST server51 versus the reference sequences retrieved from UniProt. The best BLAST hit with the highest alignment length percentage and identity match was checked and assigned as the annotation of the predicted gene. OrthoANI was used for intergenomic comparison and similarity by orthologous average nucleotide identity (OrthoANI53) by utilizing genomes of type strains of Leptospira species retrieved from EzBiocloud. A preliminary circular genome view of L. interrogans was obtained from BV-BRC54. For this analysis and genome visualization using CGView55 and Proksee56, nine genomes of closely related type strains of Leptospira that shared the highest pair-wise 16S rRNA gene sequence similarity were used in addition to the KeTo genome.

Culture based analysis for ketone body utilization

The culture grown in enriched EMJH was centrifuged (10,000 rpm, 5 min, 4 °C) and the pellets were washed thrice with basal EMJH media. Cells were grown in basal EMJH without and with EA supplements (B_KeTo and Bea_KeTo, respectively). Cell-free basal EMJH without and with EA (B and Bea) served as background controls. Culture tubes were incubated at 30 °C under shaking (120 rpm) for 96 h. Cell density was measured periodically at 600 nm. Cultures subjected to centrifugation and 100 μl cell-free media were treated with 5 μl phenol red (0.5% w/v) indicator and assessed for acidity and alkalization by reading the plates at 415 and 560 nm, respectively. The culture suspension (100 μl) was treated with 10% Alamar blue (v/v) and the plates were read at 600 nm and 570 nm. The percentage of Alamar blue reduction was quantified according to the manufacturer’s protocol (Cell-Quan AlamarBlue Cell Viability Reagent; ABP Biosciences).

Qualitative and quantitative determination of residual EA

The culture grown in enriched EMJH was centrifuged (10,000 rpm, 5 min, 4 °C) and the pellets were washed thrice with basal EMJH media. Cells were grown in basal EMJH with EA (Bea_KeTo, Bea_Ballico and Bea_AF61) and without EA (B_KeTo, B_Ballico and B_AF61). Cell-free basal EMJH without and with EA (B and Bea) served as background controls. Culture tubes were incubated at 30 °C under shaking (120 rpm). Cell density was measured at 600 nm.

For qualitative estimation of EA, we performed the gold standard assay for the ketone body, Rothera’s test57. For this, 1 gm of ammonium sulphate was taken in a glass test tube for which 1 ml of cell-free culture supernatant was added and mixed well to obtain a saturated solution. For this mixture, 500 µl of 2% (w/v) SNP reagent was added and mixed. 1 ml of ammonia was introduced through the sides of the tubes and waited for 20‒30 min for a purple ring formation at the junction of the two layers.

The residual EA was quantified using a colorimetric method58 with the following modifications. An assay was performed in a microplate for the quantification of EA by using SNP as a chromogen. The 100 μl cell-free supernatant was added to the microtiter wells and treated with 5 µl each of diethyl amine and SNP. The plates were incubated at room temperature for 10 min and the yellow colour formed was read at 400 nm. Residual EA was quantified using a standard curve plotted for the EA.

Planktonic cells and biofilm formation assay

Cells of KeTo were grown in enriched EMJH media till it reached the log phase and centrifuged (10,000 rpm, 5 min, 25 °C). Pellets were washed twice in basal EMJH and centrifuged (10,000 rpm, 5 min, 25 °C) to collect the pellets. The pellets were separately suspended in basal EMJH and enriched EMJH with and without EA. The cell suspension (200 μl/well) was loaded onto a microplate and incubated at 30 °C for 48 h under darkness. Corresponding cell-free media served as negative controls. Biofilm assay was performed as described elsewhere59. Optical density of the planktonic cells was read at 600 nm and the attached cells were fixed with 99% methanol, stained with 0.1% crystal violet. The excess stain was removed by washing with distilled water. After drying, the dye was solubilized using 33% acetic acid. Reading was recorded at 570 nm.

Determination of hemolysis by KeTo, Ballico and AF61

For the qualitative determination of hemolysis, concentrated cell pellets were placed onto 5% sheep blood agar plates (HiMedia). The plates were tightly sealed to avoid the growth of contaminants due to longer incubation periods and was incubated at 30 °C. Hemolysis was monitored daily up to 196 h and the findings were recorded.

Bipartition plate assay

Impacts of EA vapours was tested on the growth of KeTo using bipartition plate assay60 with minor modifications. EMJH semi-solid media was prepared by addition of 0.3% (w/v) agar into the basal media. The enrichment media (10%, v/v) was added aseptically along with 5FU (200 μg/ml). Another set of EMJH semi-solid media was also incorporated with 5% (v/v) blood as Leptospira representatives are known to grow better using lower percentages of agar than on solid media. Both the media were poured separately on one side of the partition Petri plate and allowed to solidify. Five µl of log phase cultures were spotted onto the prepared semi-solid media. 50 µl of EA was placed in a 0.2 ml tube on the adjacent part of the partition. The lids were secured tightly with a sealing tape in order to avoid the test compound from escaping out. The plates were incubated at 30 °C and monitored daily until the 13th day of incubation. The experiment was performed in biological triplicates. Change in media pH and metabolic activity in control and test samples were qualitatively assessed by drop-coating 5 μl each of phenol red (0.5% w/v) indicator and 5 μl Alamar blue directly onto the semi-solid agar.

Statistical analysis

Statistical significance (*p < 0.1, **p < 0.05, ***p < 0.01, ****p < 0.0001) was determined by one-way ANOVA and/or unpaired t-test using GraphPad Prism. Chemical structures were drawn using Marvin (https://chemaxon.com/products/marvin).