is a significant reservoir host of leptospirosis. It is also an important experimental animal model making it ideal to study pathogenic mechanisms of chronic leptospirosis using a relevant host of persistent carriage and infection. Recent studies have confirmed that Leptospira
regulate gene and protein expression during acute and chronic disease 
and whilst several studies have examined the acutely infected host response to infection, there has been limited work to explore the molecular basis of the chronically infected host response that facilitates persistent carriage.
Experimentally infected Rattus norvegicus
excrete large numbers of leptospires and differential protein expression is evident in the urine of experimentally infected rats compared to non-infected controls 
. In order to normalize for the presence of leptospires in urine from infected rats, urine from non-infected controls was spiked with in vitro cultivated Leptospira
. Therefore, results identify not only host derived proteins that are differentially expressed during chronic leptospirosis, but also those proteins of Leptospira
that are differentially expressed in response to host signals encountered during renal colonization.
The expression of several host derived proteins was diminished in infected rat urine samples relative to non-infected controls and included membrane metalloendopeptidase (Mme), malate dehydrogenase 1 (Mdh1), napsin A aspartic peptidase (Napsin A) and aspartocylase. Membrane metalloendopeptidase (also known as neprilysin), a type-II membrane anchored enzyme, has roles in posttranslational modification, protein turnover, and as a chaperone. Napsin A aspartic peptidase is a kidney-derived aspartic protease-like protein expressed in kidney, lung and spleen, and is excreted as a functional protease in urine 
. Whilst the significance of this reduced expression during chronic leptospirosis is not yet clear, both Mme and Napsin A are reported to be expressed in renal tubules, and decreased levels of expression are indicative of renal tubule injury 
. Alternatively, Leptospira
might directly affect the expression of these host proteins as a protective mechanism to minimize host-initiated proteolytic degradation of leptospiral proteins. Primary injury of renal tubules is regarded as the hallmark of the kidney in human patients suffering leptospirosis 
, and both Mme and Napsin A are conserved in humans and dogs.
In contrast, the expression of host derived proteins identified as components of immunoglobulin G and A, vacuolar H+ATPase B2 and kidney aminopeptidase, were identified in protein spots detected in increased amounts in urine of chronically infected rats. Vacuolar H+
-ATPases mediate the ATP-dependent transport of protons, are expressed in the plasma membrane in the kidney and contribute to proximal tubular bicarbonate reabsorption. The importance in final urinary acidification along the collecting system is highlighted by monogenic defects in two subunits of the vacuolar H+
-ATPase in patients with distal renal tubular acidosis 
. Aminopeptidases are proteolytic enzymes that remove L-amino acids sequentially from the amino termini of polypeptide chains. Both vacuolar H+ATPase and kidney aminopeptidase have been identified in the urinary proteome of rats 
Since experimentally infected rats can persistently excrete leptospires for months 
, samples taken over a three-week period were selected for DIGE to eliminate day-to-day variability in urine samples. However, continued evidence of differential gene expression in infected kidney tissue compared to non-infected controls was provided by qRT-PCR at the end-point of the experiment. In contrast to gene expression levels for igk
which were significantly upregulated, decreased level of mme
gene transcripts were detected which is in general agreement with proteomic data and indicative of differential protein expression. However, it will be important to ascertain at exactly what time post infection differential proteomic and gene expression changes occur, and whether such changes are linked to the appearance of pathology.
Our results indicate that increased amounts of Loa22, a surface exposed putative lipoprotein, is expressed by leptospires excreted in urine from chronically infected rats compared to in vitro cultures. Similarly, increased amounts of multiple isoforms of GroEL are detected. Loa22 is the first genetically defined virulence factor in pathogenic Leptospira
species since mutation of the gene encoding loa22
results in attenuation of virulence 
. In addition, expression of Loa22 is upregulated during acute disease relative to other outer membrane proteins 
. Increased expression of GroEL can be induced by modifying growth temperature of cultures from 30 to 37°C overnight, but interestingly there is no corresponding increase in gene transcript 
. LipL32 is an outer membrane lipoprotein which is expressed during acute and chronic disease, the function of which remains to be elucidated 
. Paradoxically, LipL32 is a major outer membrane protein specific to pathogenic species of Leptospira
, yet mutations in this gene do not confer any change in phenotype including virulence 
. Although LipL32 is detected in significant amounts in leptospires excreted in urine, differential expression was not detected for those isoforms excised for identification. Decreased amounts of flagellin were detected in leptospires excreted in urine; this may reflect diminished motility during renal tubule colonization, and that motility increases when leptospires are in the external environment. A similar diminution in levels of flagellin was observed when leptospires were cultured at 37°C from 30°C in order to emulate host conditions encountered during infection 
To further verify that the increased immunoglobulin content in urine of infected rats was in response to infection, immunoblots were performed using in vitro cultivated leptospires. Immunoblots were probed directly with urine from infected rats compared to urine from non-infected controls for the presence of IgG. Urinary IgG from infected rats was specific for leptospires and reacted with several protein antigens. In previous studies, it has been shown that sera from chronically infected rats reacts with LipL32 as expressed by both leptospires excreted in urine and the in vitro cultivated leptospires with which rats were experimentally infected, but relatively few other antigens 
. This suggests that antigen expression was down-regulated by leptospires in renal tubules to avoid host antibody responses; it will be interesting to determine the specificity of urinary IgG compared to serum IgG, particularly for antigens expressed by leptospires excreted in urine. During bovine leptospirosis, there is a correlation between the rise in urinary agglutinating antibody levels and a reduction in the detection of viable leptospires present in urine 
. Plasma cells associated with tubulointerstitial nephritis in dogs have been shown to produce anti-leptospiral antibody locally 
, but no anti-kidney antibody had been detected in any of the renal eluates, suggesting that antibody production is directed specifically against Leptospira
and not against renal antigens 
. The identification of protein antigens reactive with urinary IgG can also provide for the development of novel diagnostics to detect reservoir hosts of infection, including humans 
Quantitative RT-PCR suggests increased expression of IgA in addition to IgG. Although immunoblots confirmed detection of IgG specific for leptospires, a corresponding detection of IgA specific for leptospires was not detected in urine (data not shown). This may be due to the different time point of sampling or alternatively, indicative of local production of IgA which is not excreted in urine. Our analysis at this time is limited to pellets of urine from infected rats which contain excreted leptospires as well as host derived proteins. However, it is likely that analysis of the supernatant will provide for the identification of additional host derived proteins. It will be of interest to further investigate the supernatant for biomarkers of chronic leptospirosis and monitor relative expression of these in the presence of urine over the course of months, as occurs in naturally infected hosts suffering chronic disease. By definition of the experimental design, differentially expressed proteins serve as biomarkers to identify chronically infected rats compared to non-infected controls, but such biomarkers will need to be validated in hosts with alternative kidney pathologies e.g. chronic kidney disease. Finally, it will be important in future studies to determine what components of Leptospira induce increased/decreased expression of each of these host derived proteins and whether inhibiting such changes reduces colonization.
In conclusion, results demonstrate the use of the rat model of chronic leptospirosis to identify differentially expressed proteins in urine derived from both host and pathogen. Differentially expressed host derived proteins include known markers of kidney function and immune response. Differential expression was validated at the level of gene transcription and in the case of immunoglobulin G, further validated through to the production of antibody which was specific for leptospires. Differentially expressed pathogen derived proteins include the known virulence factor Loa22 and the stress response protein GroEL.