Fig 1.
Phylogeny and evolution of host- and non-host adapted Leptospira lineages.
(a) Distribution of virulence-associated genes and sources of isolation across Leptospira species. The phylogenetic tree of Leptospira is based on soft-core genes (present in at least 95% of the genomes). The subclade P1, formerly referred to as the “pathogens” lineage, can be separated into two distinct groups: P1+ and P1-. P1+ consists of species associated with severe infections and diverged after a specific node of evolution (filled circle), while P1- comprises species that have not been isolated from patients and are considered as “low-virulent pathogens”. The species used in this study are indicated in the phylogenetic tree in bold typeface. Species in the P1- group and P2 subclade isolated from animals are indicated by a red rectangle according to previous studies (L. alstonii (frogs) [19], L. tipperaryensis (shrew) [20], L. licerasiae (humans and rats) [21–23], L. venezuelensis (rodents, cattle and humans) [15], and L. fainei (pigs and wild boars) [24,25]). The distribution of virulence-associated genes (S2 Table) within the genus Leptospira is also shown using a heat map representation. (b) Schematic representation of the Leptospira genealogy. Evolutionary model and reconstruction of ancestral phenotypes in the genus Leptospira by PastML analysis using all maximum likelihood methods [26]. Branches are annotated with bars representing the sum of gene gain (blue bar) and loss (red bar). S, P2, P1- and P1+ clades and groups are indicated by spheres (whose size corresponds to the number of species) while most-recent common ancestors are indicated by dashed spheres. The dotted circles represent the most-recent common ancestors of each Leptospira group (S, P2, P1- and P1+), and the color indicates the most likely phenotype of that ancestor. (c) Virulence of representative Leptospira species in the hamster model. The virulence of Leptospira was assessed by infecting hamsters (n = 4) with 108 leptospires by the intraperitoneal route. After 2 and 4 days of infection, the burden was assessed in kidney (red symbols) and liver (blue symbols) by quantitative PCR. Data are means ± SD; the absence of values indicates that Leptospira DNA was not detected. (d-f) Survival of representative Leptospira species in water. Leptospires were incubated at RT in filter-sterilized spring water. At 21 days, leptospires were harvested, labelled with DAPI and analyzed by confocal microscopy (d) (scale bar: 10μm). The roundness of DAPI-positive leptospires was performed using Icy software (e) (n = 100 leptospires). The survival of Leptospira in filter-sterilized spring water after 21 days was determined by CFU (f). S: L. biflexa; P2: L. licerasiae, L. fluminis; P1- group: L. adleri, L. gomenensis, L. tipperyarensis, L. yasudae; P1+ subgroup: L. interrogans, L. noguchii, L. weilii, L. santarosai, L. mayottensis.
Table 1.
List of isolates used in this study.
Fig 2.
Only P1+ species escape the human complement system and have reduced interaction with human macrophages.
(a) Survival of Leptospira upon exposure to human serum. Each Leptospira species was incubated in 20% of normal or heat-inactivated human serum for 2 hr. Living bacteria were enumerated by CFU (counted in triplicate). The survival was compared to the inactivated-human serum. Unpaired two-tailed Student’s t test was used. ***p<0.0001. (b) MAC deposition in Leptospira was detected by indirect immunofluorescence. CFSE-stained Leptospira species were incubated with human serum for 30 min, fixed and then incubated with an anti-MAC antibody (C5b9). Indirect immunofluorescence was quantified by flow cytometry. The percentage was calculated by comparing the number of positive MAC-Leptospira to the number of negative MAC-Leptospira. Unpaired two-tailed Student’s t test was used. ***p<0.0001. (c) Bacterial adhesion and entry into human macrophages. To assess bacterial internalization and adhesive bacteria, infected THP-1 macrophages at 2 hr post-infection (PI) were washed with PBS and lysed directly before (adhesive bacteria) and after gentamicin treatment (bacterial internalization). Bacteria were enumerated by CFU (counted in triplicate). (d-e) Infection of human macrophages with CFSE-stained Leptospira. After 2 hr pi, macrophages were labelled with LysoTracker (Red). The fluorescence was analyzed by confocal microscopy. DAPI (blue) was used to visualize nuclei, CFSE (green) was used to visualize leptospires (scale bar: 20 μm). Quantification of CFSE positive macrophages (infected macrophages) compared to CFSE negative macrophages (uninfected macrophages) was performed using Icy software. Unpaired two-tailed Student’s t test was used. *p<0.01. (f) Human macrophage response to Leptospira infection. After 6 hr pi, IL-1β, TNF-α and IL-6 cytokines release from supernatant were measured by ELISA. Unpaired two-tailed Student’s t test was used. **p<0.001. (g) Representative fluorescence microscopy images of macrophages uninfected or infected with leptospires for 6 hr. Macrophages were labeled with antibody against NF-κB (red). DAPI (white) was used to visualize nuclei, respectively (scale bar: 20 μm). The NF-κB translocation to the nucleus can be visualized by the increase of NF-κB fluorescence intensity (red here) into the nucleus (white). (h) Ratio between nuclear and cytosolic NF-κB fluorescence intensity (n > 100 cells per condition, two-way ANOVA test; ***p<0,01). LPS: Escherichia coli LPS. (i) Heatmap showing relative expression of several genes regulated by NF-κB after 6hr pi for Leptospira infected macrophages. Expression of genes was analyzed and normalized using gapdh gene. Hierarchical clustering was performed using Ward’s method. Data are the mean ± SD (panels a-d, f, h, and i) or representative (panels e and g) of three independent biological replicates. S: L. biflexa; P2: L. licerasiae, L. fluminis; P1- subgroup: L. adleri, L. gomenensis, L. tipperyarensis, L. yasudae; P1+ subgroup: L. interrogans, L. noguchii, L. weilii, L. santarosai, L. mayottensis.
Fig 3.
Acquisition of specific P1+ genes involved in host-associated lifestyle.
(a) Genes gained in P1+ species. Annotation of the 64 P1+-specific genes is indicated. (b) Positive selection analysis of the P1+-specific genes. Significative positive selection was determined using the PoSeiDon pipeline. Significant genes (p-value <0.05, dashed line) and rate of non-synonymous (dN) and synonymous (dS) in the alignment of orthologous sequences are indicated. (c-d) Heterologous expression of putative virulence factors in P1- strains affects the virulence characteristics. The virulence of L. adleri (c) and L. yasudae (d) P1- strains was assessed by infecting hamsters (n = 4) by intraperitoneal route with 108 leptospires. After 1 day of infection, leptospiral load in kidney and liver was assessed by quantitative PCR. pLIMLP_09380, pLIMLP_11655 and pLIMLP_03665 correspond to constructs that enable the heterologous expression of L. interrogans genes in L. adleri and L. yasudae. Unpaired two-tailed Student’s t test was used. *p< 0.01, **p<0.001, ***p<0.0001, ns: not significant. (e) Silencing of putative virulence factors in the pathogen L. interrogans affects the virulence characteristics. Survival of hamsters (n = 4) infected intraperitoneally with 106 Leptospira for each construct. L.i. pdcas9-empty, L.i. pdcas9-LIMLP_09380, L.i. pdcas9-LIMLP_11655 and L.i. pdcas9-LIMLP_03665 correspond to knock-down L. interrogans mutant strains of the corresponding genes using dcas9. Statistical significance in comparison with L.i pdcas9-empty was determined by a Log rank Mantel Cox test (**p<0.0021). (f) Heterologous expression of LIMLP_09380 in P1- strains affects the survival in human serum. P1- species (L. adleri and L. yasudae) producing or not LIMLP_09380 were incubated in 20% of human serum or inactivated-human serum for 2 hr; L. interrogans WT (L.i) is shown here as a reference for P1+ species. After incubation, the bacteria were enumerated by CFU (counted in triplicate). The percentage of surviving bacteria was calculated using the inactivated-human serum as normalization. Unpaired two-tailed Student’s t test was used. **p<0.001, ***p<0.0001. (g-h) Heterologous expression of LIMLP_11655 in P1- strains affects the innate immune response of macrophages. Ratio between nuclear and cytosolic NF-κB fluorescence intensity (n > 100 cells per condition, two-way ANOVA test; ****p<0,001; ns: not significant) in the different Leptospira strains (g). Heatmap showing relative expression of several genes regulated by NF-κB after 6hr pi for Leptospira infected cells. Expression of genes were analyzed and normalized using gapdh gene. Hierarchical clustering procedure of Leptospira genus was performed using Ward’s method (h).
Fig 4.
Heatmap representation of the main features of representative Leptospira species described in this study.
With the exception of L. licerasiae [21–23], only P1+ species are responsible for infections in humans. Hierarchical clustering was performed using Ward’s method.
Fig 5.
Schematic representation of the evolutionary transition from environmental-adapted Leptospira species (P1- group) to host-adapted Leptospira species (P1+ group).
Host-associated factors found in both P1- and P1+ species (S2 Fig) are indicated in grey. Factors found exclusively in P1+ and P1- are indicated in red and green, respectively. Specific lipoproteins, Leucine-rich repeat (LRR)-encoding proteins, sphingomyelinase-like proteins, virulence-modifying (VM) proteins and uncharacterized proteins are prominent among P1+ isolates. The different factors contributing to host adaptation of P1+ species are represented, including a collagenase (encoded by LIMLP_03665), a hypothetical protein (encoded by LIMLP_09380) and a VM protein (encoded by LIMLP_11655). LIMLP_09380 participates in evasion from complement-mediated killing and the VM proteins is involved in prevention of host inflammatory response. In addition, several factors are contributing to cell binding and ECM (extracellular matrix) degradation in P1+ species. The lipopolysaccharides (LPS) of P1+ species, which have a higher complexity than those of other Leptospira species [7,34] may differentially interfere with the host and confer resistance to immune surveillance. The reduced in vitro microbial and metabolic activities of P1+ species in comparison to P1- species might be also important for adaptation to the host. The catalase activity of P1+ species is higher than in P1- species, allowing them to better tolerate H2O2 as encountered inside a host.