Abstract
Pathogenic species of Leptospira cause leptospirosis, a global zoonotic disease. Leptospira colonize renal tubules of chronically infected maintenance hosts, from where they are shed in urine to the environment and survive in suitable moist conditions. Transmission of disease to new hosts is facilitated by contact with contaminated urine or water sources, because Leptospira can penetrate broken skin or mucosal surfaces of new hosts. Infection of new hosts may be asymptomatic, as with chronically infected maintenance hosts, or may result in an acute disease process in which clinical signs can include fever, jaundice, renal failure, and pulmonary hemorrhage. Those factors that determine if an animal will suffer an acute or a chronic infection are not fully understood but include host animal species, infecting serovar, and infecting dose. During chronic infection, renal colonization and leptospiruria persist despite cellular and humoral responses by the host. Tubulointerstitial nephritis is the most common lesion associated with chronic infection, and this may progress to fibrosis and subsequent renal failure. This review aims to address how Leptospira cause tubulointerstitial nephritis during chronic leptospirosis and to summarize the mechanisms by which Leptospira might evade host immune responses during chronic colonization of the renal tubule.
Introduction
Pathogenic Leptospira spp. infect a wide range of mammals and may cause a range of acute or chronic disease processes. This review focuses on chronic disease, in which a biological equilibrium is observed between particular hosts and certain leptospiral serovars, e.g., Leptospira borgpetersenii serovar Hardjo and Leptospira interrogans serovar Hardjo in cattle, Leptospira interrogans serovar Canicola in dogs, Leptospira interrogans serovar Pomona in pigs, and Leptospira interrogans serovar Copenhageni in rats. 8,15 In these situations, maintenance hosts are typically clinically asymptomatic, and Leptospira succeed in evading the immune response to colonize renal tubules from which they are shed in urine. In contrast, accidental hosts may suffer a range of clinical symptoms of disease. In addition, acute leptospirosis may resolve and progress to chronic infection and persistent leptospiruria. Leptospiruria in maintenance hosts is of high intensity, constant, and long duration compared with accidental hosts, where it is of low intensity, intermittent, and short duration. 9 Mechanisms of immune evasion by Leptospira during renal tubular colonization remain unclear, because leptospires survive and are shed from chronically infected hosts despite a specific host immune response. 31 Evidence of targeted immune responses include the detection of circulating anti-leptospiral immunoglobulin in the serum and urine, 31,32 and varying degrees of lymphocytic and plasmacytic interstitial nephritis. 4,28,31,50,61 In addition, the expression of major histocompatibility class II antigens (MHC II) on tubular epithelial cells in cases of porcine leptospiral nephritis was observed. 42 The molecular mechanisms underlying the constant battle between the maintenance host immune response and Leptospira that colonize renal tubules during chronic disease are poorly understood. Here, we review some of the specific interactions between the host and the pathogen, and put forward some hypotheses on how leptospiruria persists in the face of a specific immune attack.
Dissemination of Leptospira to the kidney
Direct contact of domestic and wild animal species with infected urine, or contaminated water sources, facilitates transmission of leptospirosis, because leptospires can penetrate breaches of the skin or mucosal surfaces, e.g., the conjunctiva. Early studies of mice, rats, and guinea pigs showed that uninjured skin and nasal mucosal membranes are barriers to Leptospira infection. 47 The kidney is a primary target of Leptospira during both acute and chronic infection, where conditions in the renal tubules favor Leptospira survival. 15,58
After experimental infection of the rat (Rattus norvegicus) with L. interrogans serovar Copenhageni, extensive dissemination to almost all tissues is observed during the early stages of infection. 3 This is followed by a clearance, within days, of Leptospira from most tissues, except for the selective survival and proliferation of leptospires in the kidneys, followed by leptospiruria on day 7. 3,31 This clearance from most tissues is likely facilitated by circulating anti-leptospiral immunoglobulin (IgM and IgG), which is detectable in rats by 7 days after infection (Monahan and Nally, unpublished), 31 supporting the argument that the kidney, therefore, is somewhat immune privileged and thus facilitating persistent colonization by leptospires. 3,15 Because Leptospira initially disseminate to all tissues and not specifically the kidney, this does not support the hypothesis that Leptospira specifically targets the kidneys because of tropism. 3 The mechanisms that facilitate this immune privileged state are difficult to ascertain from the literature (Table 1). For instance, renal tubules may be immunologically “favored” for Leptospira colonization, if the anti-leptospiral immunoglobulin present in tubular and bladder urine is unable to kill the bacteria, possibly because of the absence of complement (Tables 1, 2). 1
Future research that will expand knowledge and help explain the pathogenic mechanisms of Leptospira.
Emerging hypotheses to explain how Leptospira evade the renal immune response.
A study that examined experimental infection of swine with serovar Pomona showed that leptospires entered the kidney hematogenously and migrated through the endothelium to the interstitium during an “interstitial phase,” which was characterized by edema, vasculitis, and leptospiremia. 10 This was followed by a “tubular phase,” during which leptospires associated with lumina of the proximal convoluted tubule. 10 Similar results were detected in experimentally infected mice in which Leptospira were shown to migrate from the capillary lumina to the interstitial tissue of the kidney within the first 4 days of infection. 26 By day 10, Leptospira were observed between the epithelial cells of the proximal convoluted tubules, with colonization of the renal tubule lumina by day 14. However, despite this early migration and establishment of leptospires, morphologic alterations in the kidney and specific lymphocyte infiltration were not observed in the rat before 21 days after infection. 37,50 This suggests that Leptospira colonize renal tubules in rats without any initial significant damage to the kidney or stimulation of a local cellular immune response (Table 2).
For leptospires to disseminate within the host, leptospiral invasion is likely mediated by enzyme secretion, which allows for degradation of host cell membranes. 38,43 Several genes of L. interrogans were found to encode for hemolysins, which may lead to hemorrhage and localized ischemia, possibly contributing to renal tubular necrosis in some cases of acute leptospirosis. 39 Genes encoding for sphingomyelinases, which initiate endothelial cell damage (increased permeability, membrane aggregation, and fusion 16 ) are also expressed by pathogenic and not by saprophytic leptospires and may be significant during dissemination and pathogenesis. 40
Tubulointerstitial nephritis and leptospiral immune evasion in the kidney
Kidney injury is an early manifestation of acute leptospirosis, occurring within days of infection, in comparison with chronic infection in which kidney damage occurs later, after weeks or months of renal colonization. Tubulointerstitial nephritis is a primary lesion during acute renal injury in leptospirosis, with characteristic lesions, such as interstitial edema and lymphocytic infiltrates. 58 Renal lesions associated with chronic leptospirosis were documented in a range of carrier animals, including dogs, rats, pigs, cattle, and raccoons. 4,19,32,50,61
Our previous work in the rat showed that, after 160 days of experimental infection with L. interrogans serovar Copenhageni, a renal inflammatory response is evident in the form of multifocal, mild-to-severe interstitial nephritis. 31 Experimental infection of rats shows that interstitial nephritis does not occur until after 1 month or more of infection, despite renal carriage being established 7 days after infection. 31,50 Interstitial nephritis was the only lesion attributable to leptospiral infection. 50 The inflammatory infiltrate is lymphocyte rich, with discrete aggregations of macrophages, primarily located in the interstitium, in the region of proximal and distal tubules. 31 Similar immune lesions were also observed in studies of trapped wild rats, with 56% of culture positive wild rat kidneys showing distinct foci of nephritis. 50 However, in the same study, similar inflammatory infiltrates were observed in 53% of wild rat kidneys, which did not culture positive for Leptospira. It would have been of interest to know, for example, by immunohistochemistry, if lesions in the kidneys of wild rats were directly associated with Leptospira colonization. Little information is available about the effect of the wide array of environmental, chemical, biological, and nutritional factors that may influence disease progression and the host renal immune response during chronic infection in nature.
In addition, limited information is known of the active processes involved in clearing Leptospira infection or the immune evasion tactics used by Leptospira to avoid clearance. Anti-leptospiral IgG is detectable in the serum of naturally and experimentally infected rats. 48 During bovine leptospirosis, there is a correlation between the rise in urinary antibody levels and a reduction in the detection of viable leptospires present in urine. 24 Plasma cells associated with tubulointerstitial nephritis in dogs were shown to produce anti-leptospiral antibody locally. 32 No anti-kidney antibody has been detected in any of the renal eluates, which suggests that antibody production is directed specifically against Leptospira and not against renal antigens. 32
Interestingly, early work on rodent carrier models of Leptospira showed that Leptospira derived from renal tubules, i.e., in vivo sources, is less reactive with serum of chronic carriers than in vitro derived Leptospira. 13,14 This is thought to be because of fundamental antigenic differences between the in vitro and in vivo derived organisms. Furthermore, our recent proteomic analysis of Leptospira in rat urine compared with in vitro cultured Leptospira demonstrates that antibody in the serum of chronically infected rats reacts with relatively more antigens of the in vitro derived Leptospira compared with Leptospira derived from rat urine. 31 These results support the hypothesis that differential protein expression and downregulation of specific antigenic proteins by Leptospira during chronic infection facilitates renal tubule colonization and leptospiruria persistence in the presence of a specific host antibody response (Table 2).
Our previous work showed that, although some renal tubules positive for the presence of leptospires are surrounded by inflammatory infiltrate, other renal tubules that contain Leptospira are devoid of any obvious inflammation, perhaps indicating that such tubules are recently colonized or that the bacteria are in some way evading the immune response. 31 Similar findings were observed in a study of infected dogs. 32 Therefore, it has been suggested that leptospires present in tubules not surrounded by cellular infiltrates have a different antigenic profile that facilitates evasion of an immune response and thus enabling leptospiruria to persist (Table 2). 31
Biofilms are colonies of bacteria encapsulated within a protective matrix, which enable bacterial growth in hostile environmental conditions. 18 Genes thought to facilitate alginate biosynthesis that is involved in biofilm formation were found in the genomes of both saprophytic and the pathogenic Leptospira spp., correlating with the fact that both are capable of biofilm formation. 44 Interestingly, these genes are absent in pathogenic L. borgpetersenii, which has reduced environmental survival. 40 As yet, there is no evidence to support the hypothesis that biofilm formation is required for Leptospira renal tubule colonization and immune evasion during chronic infection.
The role of complement may also be important in the context of immune evasion by Leptospira. The complement system is an integral part of the innate immune response, in which various pathways and cascades lead to the elimination of invading microorganisms. 23 As a self-defense mechanism, host cells bind plasma factor H, which promotes deactivation of the complement activation system, thus preventing host cell destruction. 23 The genome of pathogenic Leptospira encodes a group of proteins called Len (leptospiral endostatin-like) proteins, some of which were shown to bind complement regulatory proteins, e.g., plasma factor H. 49 These group of proteins are absent from nonpathogenic Leptospira, which supports the hypothesis that they function in immune evasion. Pathogenic Leptospira spp. can also bind the human complement regulator C4BP, which can provide added resistance against host complement. 5 The leptospiral genome encodes many hypothetical proteins, the functions of which are unknown. Additional, as yet uncharacterized, proteins may also facilitate immune evasion by interfering with the host complement activation system (Table 1).
Genetic predisposition to acute Leptospira infection in humans has been associated with allelic differences that alter the MHC receptor. 25 MHC II products are transmembrane molecules that are required for the presentation of antigens to CD4+ T cells 12 and are normally found on professional antigen-presenting cells, such as dendritic cells, macrophages, and B cells. MHC II may also play a role in the genetic predisposition of animals to infection. MHC II expression was shown on certain epithelial cells, including renal tubular cells, which allowed them to function as nonprofessional antigen-presenting cells. 42,51 In the normal canine kidney, MHC II molecules were detected in interstitial dendritic cells, but, in cases of tubulointerstitial nephritis, MHC II was also expressed on epithelial cells of cortical and medullary tubules. 51 However, in this study, the tubulointerstitial nephritis may have been caused by other factors and was not always specifically caused by leptospirosis. 51 Additional evidence of MHC II involvement in leptospiral evasion was presented in pigs, in which marked variability in MHC II expression and distribution during chronic leptospiral nephritis may reflect the ability to activate different immune response phenotypes. 42 MHC II expression during leptospiral nephritis in swine may stimulate the local inflammatory reaction and systemic humoral response. 42 Interestingly, no colocalization of MHC II expression and leptospiral antigen is observed within renal tubules, whereas some adjacent tubules without leptospiral colonization show variable degrees of MHC II expression, which suggests a potential defense mechanism by the host against further leptospiral colonization of renal tubules. 42 Alternatively, tubules that are MHC positive but free of bacteria have already been cleared by the immune response. MHC II–positive tubular cells may activate peritubular CD4+ T cells to produce proinflammatory cytokines, which contribute to the immune response against Leptospira. 42
Pathogenesis of tubulointerstitial nephritis
Tubulointerstitial nephritis associated with leptospirosis can be caused by direct damage to host tissue by Leptospira or the presence of leptospiral antigen, initiating a renal immune response. 56 The underlying mechanisms of renal changes may vary, particularly between chronically infected maintenance hosts, which may remain asymptomatic compared with acutely infected, symptomatic, accidental hosts. Interstitial nephritis appears as an early event during acute systemic disease and a late progressive event during the chronic disease process. 46 Differences in the time course of renal lesions are striking when comparing chronic infection of rats to acute infection of guinea pigs; interstitial nephritis is not evident in rats before 21 days of asymptomatic Leptospira infection 37,50 compared with acute infection in guinea pigs, in which the inflammatory infiltrate is composed primarily of neutrophils present in association with tubular epithelial cell necrosis as early as day 4 after infection. 36 In one study of hamster infection, glomerular changes that included mild and focal inflammatory cell infiltration around the glomeruli was noted within 3 hours of inoculation with L. interrogans and became pronounced after 9 hours. 46 Tubulointerstitial nephritis may progress to tubular atrophy and renal fibrosis as seen in chronically infected animals, such as dogs infected with serovar Canicola 28 and rats infected with serovar Icterohaemorrhagiae. 48
Several leptospiral factors have been implicated as toxins or endotoxins, 7,45 producing cytopathic effects on a variety of cells in vitro. 30,55 Early attempts to characterize the toxic agent were irreproducible and inconclusive. 2,15 Leptospira toxins are thought to be closely associated with the leptospiral cell, because whole cells produce a more extensive cytopathic effect than culture supernatants alone. 29,30 Leptospiral-associated direct nephrotoxicity is mediated by outer membrane proteins and lipopolysaccharide. However, this tissue-culture–based observation may not reflect occurrences in the kidney in vivo, because protein secretion or lysed leptospiral fragments that occur in vitro may differ.
Outer membrane components of Leptospira have been shown to be responsible for renal dysfunction. Cultured mouse medullary thick ascending loop of Henle epithelial cells showed increased expression of many genes related to tubular injury and inflammation in the presence of Leptospira outer membrane protein extract. 59 Additional supporting evidence of the pathogenic significance of outer membrane proteins is shown with outer membrane protein extracts of Leptospira santarosai serovar Shermani. When added to cultured mouse proximal tubules, a dose-dependent increase of monocyte chemoattractant protein-1 (MCP-1), RANTES, nitrite and TNF-α is induced within 48 hours. 60 RANTES and MCP-1 are chemokines that are likely to recruit mononuclear leukocytes to the kidney. 60 Antibody raised against the well-characterized outer membrane lipoprotein LipL32 prevented the stimulatory effects of outer membrane protein extract on MCP-1 and iNOS messenger RNA (mRNA) expression, which suggests that LipL32 is a major component of pathogenic Leptospira involved in the mechanism by which pathogenic Leptospira bacteria stimulate the development of tubulointerstitial nephritis. 60 Immunohistochemistry analysis has shown LipL32 to be broadly expressed on the surface of Leptospira present in the proximal tubule and the interstitium, which suggests a function in tubular colonization. 17 It has also been shown to be a dominant reactive antigen in human and animal serologic studies of infection 15 and is not expressed on the surface of saprophytic species of Leptospira. 17 Surprisingly, a recent article that describes a LipL32 knockout mutant showed that LipL32 is not required for either chronic infection of rats or acute infection of guinea pigs. 34 Thus, the function of this protein that is so abundant on the leptospiral outer membrane during disease remains elusive. It would be interesting to see if the newly created LipL32 mutant causes interstitial nephritis during chronic infection in the same way as the wild type. Future genetic manipulations of Leptospira will no doubt help elucidate the pathogenic mechanisms of chronic disease and identify, in vivo, the leptospiral components that cause interstitial nephritis.
Role of toll-like receptors during leptospiral infection
It has been hypothesized that the innate immune system, particularly the activation of toll-like receptors (TLR) may play a role in pathogenesis of renal disease during Leptospira infection. 56 Individual TLRs recognize distinct pathogen-associated molecular patterns, such as lipopolysaccharide (LPS) and glycolipids, which have been evolutionarily conserved in specific classes of microbes. 11 In response to TLR activation, numerous cytokines and effector molecules are produced. 56
LPS from Leptospira is atypical in that it activates human cells through TLR2 and not the conventional TLR4 pathway. 53 This may be because of the unique structure of its membrane anchor, the leptospiral lipid-A moiety, which is the active toxic component of LPS. 41 In addition, it may be that more than the lipid moiety of leptospiral LPS is required for TLR2 activation. Unusually, the endotoxic activity of leptospiral LPS is considerably less than that of typical gram-negative LPS. 1,15,20,21,27 Increased LPS O antigen content was associated with chronic renal tubular colonization of rats compared with diminished content during acute lethal infection of guinea pigs. 37 O antigen is reported to play a role in adhesion 6 and resistance to complement. 33 The increased amounts of O antigen detected during chronic infection may provide added resistance to the action of complement.
TLRs, mainly TLR2 and TLR4, are highly expressed on renal tubule epithelial cells in vivo, and in vitro studies have indicated that TLR2 mediates the early inflammatory response caused by pathogenic Leptospira in proximal renal tubules. 54,57 Lipoproteins are classically the most potent ligands of TLR2, inducing cytokine secretion and maturation of dendritic cells. 22 Outer membrane proteins from pathogenic Leptospira that contain LipL32 and purified LipL32 increased the expression of TLR2 and stimulated the release of the MCP-1, RANTES, iNOS, and tumor necrosis factor α (TNF-α) in proximal renal tubule cells. 57,60 In contrast, the outer membrane lipoprotein LipL41 was not found to induce gene-expression changes related to the development of tubulointerstitial nephritis. 60 Outer membrane proteins extracted from L. santarosai serovar Shermani did not increase TLR4 expression but caused TLR2 expression to increase by 28%. 57 TLR2 mRNA levels were significantly increased during incubation with recombinant LipL32 but not during incubation with the recombinant outer membrane protein LipL41. 57 Additional in vitro experiments using recombinant LipL32 and anti-LipL32 antibody indicate that LipL32 directly affects TLR2 gene expression in renal proximal tubule cells. 57
Infection of C3H/HeJ mice that have a mutant TLR4 results in an acute lethal infection. Infection of control C3H/HeN mice with an intact TLR4 suffer a sublethal chronic infection and lower leptospiral burden in liver, lung, and kidney. 52 This suggests that TLR4 plays a role in controlling the leptospiral burden during chronic leptospirosis, contrary to previous findings that TLR activation is solely through TLR2. In addition, leptospiral lipid A was shown to be a murine, but not a human, TLR4 agonist. 35 This suggests that differential lipid A recognition by different animal species may play a key role in disease susceptibility and progression. 35
Conclusion
Leptospirosis causes loss of productivity in livestock through abortion, stillbirth, loss of milk production, or failure to thrive. 15 Leptospires colonize host renal tubules, and leptospiruria persists despite active immune processes, such as interstitial nephritis characterized by lymphocyte infiltration, anti-leptospiral immunoglobulin production, MHC II expression, and TLR activation. Differences in disease susceptibility among different animal species may depend on host genetics, as highlighted by differential LPS activation of human and murine TLR. Differences in genetic predisposition among individuals have been shown to occur as a result of MHC variants. Current theories on immune evasion strategies include differential protein expression on the outer membrane of Leptospira 31 and biofilm formation in the renal tubules. 44 In addition, antibody in the renal tubules may not trigger bacterial killing, because of the absence of complement. Much of the annotated leptospiral genome has revealed genes with unknown function. Elucidation of the functions of these genes and their corresponding products during chronic disease would further contribute to revealing the pathogenic mechanisms of chronic leptospirosis, altering the development of therapeutic and control strategies.
Footnotes
Acknowledgements
Our research group is supported by grant number 05/YI2/B696, President of Ireland Young Researcher Award from Science Foundation Ireland, and grant number RSF06 363 from the Research Stimulus Fund from the Department of Agriculture, Fisheries and Food. The authors would like to thank the two anonymous reviewers for their excellent guidance during the preparation of this manuscript.