Leptospirosis is a zooanthroponosis, widely distributed throughout the world and aetiologically caused by pathogenic bacteria belonging to the genus, Leptospira (Bharti et al., 2003; Pappas et al., 2008). Pathogenic Leptospira species are invasive and infection results from their ability to colonize and invade the renal tubes of incidental hosts. While the complete mechanism involved in leptospiral pathogenicity is not fully elucidated, several studies on leptospiral virulence and virulence-associated factors indicate the involvement of haemolysins, adhesins, heat shock proteins, flagellins/motility, lipopolysaccharide (LPS), catalase KatE, heat-inducible ClpB chaperone, and several outer membrane proteins (Artiushin et al., 2004; Cullen et al., 2005; Barbosa et al., 2006; Nally et al., 2007; Dong et al., 2008; Lourdault et al., 2011; Andrade and Brown, 2012; Eshghi et al., 2012).

Pathogenicity is multifactorial, requiring integrated mechanisms and pathways to establish an infection. As diverse pathogenic bacteria share common strategies to cause disease and infection, the extent of damage to host tissue is determined by multiple gene or gene products involved in pathways of signal transduction, invasiveness and toxigenesis. Bacteria rely on the ability to sense and respond to environmental cues, including changes in temperature, pH, osmolarity, oxygen availability, and nutrient conditions (Thomas and Wigneshweraraj, 2014). The changing environment, often influenced by climate change, prompts adaptation, and regulatory responses to enhance the survival of the pathogen, and by extension, the ability to cause infection is largely due to the pathogen’s ability and adaptability (Reis et al., 2008; Lau et al., 2010; Batchelor et al., 2012).

The state of physiological imbalance between the natural or exogenous production of and/or exposure to high levels of oxidants and the organism’s ability to counteract their harmful effects induce oxidative stress (Fulda et al., 2010). In biological systems, oxidation reactions involving organic molecules usually generate unstable free radicals including reactive oxygen species (ROS) and reactive nitrogen species (RNS) (Nathan and Shiloh, 2000; Seis, 2014). Unequivocal interplay between oxidant and anti-oxidant countermeasures trigger a myriad of cascades that are likely to induce oxidative stress. For example, in Escherichia coli, the presence of as little as 0.5 μM H₂O₂, regardless of the source, can be cytotoxic (Sobota and Imlay, 2011). Unlike enteric bacteria, saprophytic leptospires lack the two main transcriptional regulators of oxidative stress response in enteric bacteria, OxyR and SoxRS (Anjem and Imlay, 2012). However, Leptospira interrogans possesses four predicted FurR homologs and PerR, a negative peroxide regulator with sensitivity to low H₂O₂ levels (Fillat, 2014), which exhibits similarity to PerR in Bacillus subtilis which controls katA and ahpC expression (Lo et al., 2010). Other leptospiral defense mechanisms against oxidative stress involve but are not limited to peroxiredoxin LinAhpC, catalase KatE in L. interrogans compared to KatG in L. biflexa, heat inducible ClpB chaperone, and glutathione and thiol peroxidases, among others which function in the capacity as metalloproteins (Lourdault et al., 2011; Eshghi et al., 2012).

Adaptation by an organism during serial passage is well established since the generation of live attenuated vaccines (Gamberini et al., 2005). In most instances, gene expression among pathogenic bacteria, including Leptospira is attenuated in highly passaged cultures and associated with loss of or attenuated virulence. Cullen et al. (2002) noted concomitant changes in leptospiral surfaceome, colonial morphology and loss of virulence associated with highly passaged Leptospira. Other studies have reported observations of non-synonymous variant alleles (Lehmann et al., 2015), attenuation of genes involved in invasion (Toma et al., 2014) and plasminogen binding (Vieria et al., 2009). Gene function and virulence may be restored by passage through a host and/ or activation by stimuli.

Temperature and oxidative stress represent two of the main external (host and environmental) stresses which influence virulence and viability of pathogenic Leptospira. Given that L. interrogans serovar Portlandvere and L. borgpetersenii serovar Jules account for more than 60% of cases of human leptospirosis in Jamaica, and with the paucity of information regarding molecular pathogenicity and the drivers of virulence in these species, this study sought to determine the effect of temperature and oxidative stress on virulence associated genes in L. borgpeterseneii Jules and L. interrogans Portlandvere.

The transmission cycle of pathogenic Leptospira species necessitates the ability to respond to environmental changes and is supported by altered gene expression of leptospiral exoproteins predominantly involved in motility, signal transduction and energy-generating functions during infection (Matsui et al., 2012; Eshghi et al., 2015). This study examined the effects of temperature and oxidative stress on virulence associated genes in L. interrogans serovar Portlandvere and L. borgpetersenii serovar Jules. Several studies have demonstrated differential gene transcription in leptospires transitioned from environmental to host simulated conditions involving changes in temperature (Lo et al., 2006, 2009), osmolarity (Matsunaga et al., 2007); serum exposure (Patarakul et al., 2010), and macrophage interactions (Xue et al., 2010).

In this study, the variability in gene expression may possibly be attributed to any individual and/or combination of growth temperature, strain diversity, oxidative conditions or other unknown factors. In most cases, temperature by itself resulted in significant changes in gene expression within the individual strains and between the strains for the five genes analyzed in this study. Not surprisingly, but noteworthy, there was generally higher transcription in Portlandvere compared to Jules (notwithstanding the higher fold in Jules increases relative to 16S rRNA expression). This supports reports of loss of gene function involved in environmental sensing and metabolic transport and utilization in L. borgpetersenii (Picardeau, 2015), possibly due to the presence of more pseudogenes (around 12%) in L. borgpetersenii compared to <4% in L. interrogans, and gene reduction in L. borgpetersenii (Bulach et al., 2006). The approximately 700 kb reduced genome of L. borgpetersenii, being about 16% smaller than L. interrogans, exemplifies a restrictive lifecycle of a host to host mode of transmission compared to the adaptability to either aqueous or mammalian host environs for the larger sized genome of L. interrogans. The lack of expression of ligB or mce using both genomic and expression analysis may be due in part to loss or attenuation of these genes in highly passaged cultures, lack of ORFs and/or regulatory dysfunction. In virulent leptospiral strains, ligB has been shown to be upregulated upon exposure to temperature and osmolarity, with its expression lost in high passaged cultures (Bulach et al., 2006). The eleven and nine genes that were negligibly detected using endpoint RT-PCR in Jules and Portlandvere, respectively, may have possibly been expressed below a detectable threshold; subject to amplification failure, pre- or post-transcriptional modifications in the highly passaged cultures and/or responsive to factors other than temperature, oxidative intensity and duration. Differential gene expression in Jules was observed across conditions of low and high oxidative stress intensities for both short and long durations of oxidative stress. We propose that under oxidative stress, the expressions of loa22, lipL32, and lipL41 in Jules are co-regulated by temperature and oxidative stress, as their expressions were reduced at 37°C, and oxidative conditions of long duration yielded the most abundant transcription when the bacteria were grown at 30°C and upshifted temperatures. For Portlandvere, temperature played a greater role in differential gene expression compared to oxidative stress and duration.

Notwithstanding lower lenA expression in Jules, it was clear that there was co-regulation by temperature and oxidative stress in Portlandvere, evidenced by better yields at upshifted temperature vs 30°C, at 10 mM (high oxidative intensity) vs. 1 mM H₂O₂ (low oxidative intensity), and at 30 min vs 60 min. Interestingly, in the absence of oxidative stress, differential lenA expression in Portlandvere was particularly enhanced at upshifted temperatures and at 37°C compared to 30°C and increased following 30 min oxidative duration, particularly at high oxidation intensity. While the specific function of lenA remains unclear, it is reported to facilitate adherence to host ECM and plasma components to result in degradation of fibrinogen, connective tissue and immunoglobulin (Verma et al., 2010), suggestive of a putative role in leptospiral dissemination and/or evasive strategies. The finding of higher lenA transcripts among upshifted cultures of Portlandvere alludes to the possible involvement at early onset of infection possibly in host recognition to facilitate adhesion, bind plasminogen, overcome host derived ROS, and/or other unknown functions.

In the present study, the expression of the calcium-mediated LipL32, one of several plasminogen binding leptospiral OMPs (Picardeau, 2015), was influenced by temperature and the duration of oxidative stress. Temperature was an essential factor in lipL32 expression in both Jules and Portlandvere where changes from 30°C to upshift conditions resulted in increased and decreased lipL32 transcription in Portlandvere and Jules, respectively. Synergistic changes in temperature and the duration of oxidative stress may also possibly co-regulate lipL32 in Portlandvere, as long rather than short duration was important in the increased lipL32 expression in Portlandvere at 30°C and upshift conditions. This suggests that lipL32 is responsive to oxidative stress for Portlandvere, which contrasts down-regulation of LipL32 upon macrophage interaction with L. interrogans serovar Lai (Xue et al., 2010) and in in vivo studies using animal models of infection (Matsui et al., 2012). With some 38,000 copies per cell, the immunodominant subsurface lipoprotein LipL32 (Pinne et al., 2012), conserved among pathogenic Leptospira, is both highly antigenic and immunogenic, and has been shown to induce a robust inflammatory response via the NF-κB signaling pathway in cultured human and murine renal cells within 2 h via TLR2 activation (Yang et al., 2006; Fitzgerald et al., 2007). This is suggestive of an early inflammatory response likely leading to detrimental effects observed in tubule-interstitial nephritis observed during leptospirosis.

This observation contrasted with decreased lipL32 expression in Jules at upshift conditions and 37°C, although of lower abundance when compared to Portlandvere and warrants further investigation of possible strain specificity and a putative role for LipL32 in the oxidative stress mediated responses of L. borgpetersenii in mammalian hosts. Decreased lipL32 expression in Jules at 37°C and upshift compared to 30°C, decreased lenA and sph2 among Jules grown at 37°C and upshifted temperatures suggest no putative role during early infection. Because L. borgpetersenii is usually transmitted host-to-host transmission where the normal body temperature is ∼37°C, it is likely that the genes are expressed constitutively in Jules with a non-significant regulation by oxidative stress. In light of this, we postulate that lenA might be involved in an anticipatory adaptive response to ‘low dosage pre-exposure’ which facilitate resistance to the damaging effects of host derived H₂O₂ (Cabiscol et al., 2000). With more genes involved in signal transduction, regulatory and metabolic processes in the larger sized genome of L. interrogans, the role of lipL32 may be relegated to known functions such as outer membrane stabilization, and adhesion of host cells rather than in oxidative stress response.

Further, the results suggest that possible interplay between oxidative stress duration, temperature and strain diversity may have roles in lipL41 expression, as the gene was expressed following both short and long durations of oxidative stress of varying oxidative intensity in Portlandvere at 30°C and preferentially following high oxidative intensity. Conversely, differentially expressed lipL41 displayed temperature sensitivity in Jules, temperature changes from 30°C to upshift were important to yield increased lipL41 transcripts. However, neither changes in oxidative stress intensity nor duration significantly influenced lipL41 transcription in Jules. This decline in lipL41 in Jules as temperature increased from upshift to 37°C, while concurring with similar reports of downregulation of LipL41 upon interaction with macrophage derived cells of L. interrogans at 37°C (Xue et al., 2010), does contrast with other reports which chronicle expression of LipL41 in infection and the urine of rats. On the other hand, the minimal lipL41 expression at upshifted temperatures and the undetectable lipL41 transcripts among L. borgpetersenii Jules grown at 30°C and 37°C concurred with findings of Lo et al. (2009), indicating unaltered lipL41 expression at 30°C and 37°C. This may be suggestive of regulatory factors other than oxidative stress as indicated by Cullen et al. (2004) and Matsui et al. (2012), or lack of co-transcription of the lep chaperone (King et al., 2013).

Expression of loa22, the second most abundant leptospiral OMP (Ristow et al., 2007; Zhang et al., 2010), was noteworthy in Portlandvere compared to Jules, particularly when grown at 30°C and even more so following high oxidative intensity. Temperature was an essential factor in loa22 expression in Portlandvere, where changes from 30°C through to upshift to 37°C resulted in decreased loa22 transcription in Portlandvere. Loa22 sensitivity to intensity of oxidative stress may allude to possible co-regulation as higher yields were observed following treatment with 10mM H₂O₂ in Portlandvere. One study reported modest downregulation upon macrophage interaction with serovar Lai (Xue et al., 2010) while another attributed differences between leptospiral strains and macrophages leading to strain-specific interactions (Toma et al., 2011). Similarities between leptospiral OmpA-like, Loa22 and OmpA in E. coli suggest that Loa22 may be osmoregulated, growth rate/phase dependent, and with reduced expression at lower than optimal temperatures (Smith et al., 2007).

Increased lipL32 and loa22 transcription in Portlandvere under conditions of oxidative stress observed during this study were converse to down-regulation in L. interrogans Lai reported by Xue et al. (2010). In fact, in that study, alterations in the outer membrane of Lai upon interaction with macrophages resulted in a highly downregulated clade 1 consisting of major OMPs, lipL41, ompL1, lipL32, lipL48, and ompL47 and a moderately downregulated clade 2 comprising lipL45 and loa22, among other genes.

Differential sph2 transcription was observed in Jules vs Portlandvere with possible co-regulation by temperature in Jules (growth at upshift gave better yields vs growth at 30°C or 37°C), or temperature and oxidative intensity in Portlandvere (high oxidative intensity gave better yields than low intensity). For this gene, the duration of oxidative stress was not considered a significant regulatory factor for Jules. Further, temperature by itself was not an important driver and contrasted with the findings of other thermo-regulatory studies by Qin et al. (2006) and Eshghi et al. (2015) where lowered sph2 expression was observed at 37°C. Other studies have shown upregulation of magnesium-sensitive sph2 in the presence of physiological osmolarity (Matsunaga et al., 2007) and during infection (Narayanavari et al., 2015). SphH in L. borgpetersenii and Sph2 in L. interrogans share >50% structural similarity to Smase C of pathogenic Listeria ivanovii and the beta toxin of Staphylococcus aureus (Narayanavari et al., 2012), which mediate escape from phagocytic vacuoles to release the bacterial cells in the cytosol (Gonzálvez-Zorn et al., 1999). Notwithstanding the significant differential expressions observed in this study, we cannot rule out the possibility of observations due to the differences in experimental temperature, harvesting time, source and complexity of oxidative stress (i.e., simplified H₂O₂ induction versus macrophage interaction).

Differential gene expressions corresponding with temperature changes from 30°C to upshift; 30°C to 37°C and upshift to 37°C and responsiveness to increased intensity and duration of oxidative stress were of keen interest and prompt further investigation of possible role during infection. With an appreciation of the complexity of integrated mechanisms, and genes and gene products involved in the oxidative response of a cell, the in vitro conditions of this study were not meant to simulate physiological conditions of the complex system of oxidative stress, however, the results serve as an important snapshot of selected gene expression in response to temperature and oxidative stress. While it is clear that expression of many virulence genes in highly passaged strains of Leptospira are attenuated or lost, genetic predisposition, changes in growth temperature and/or oxidative intensity and/or duration were factors which acted in isolation or together with other regulatory cues to contribute to the variable gene expression observed in this study. Overall, differential gene expression in serovar Portlandvere was more responsive to temperature and oxidative stress, although relative to 16S rRNA, fold increases in gene expression were associated with Jules.

TF assisted in the design of the study, carried out the expression assays, performed statistical analyses and drafted the manuscript. PB conceived of the study, its design and coordination and edited the manuscript. Both authors read and approved the final manuscript.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.