Macrophages and Galectin 3 Control Bacterial Burden in Acute and Subacute Murine Leptospirosis That Determines Chronic Kidney Fibrosis

Abstract

Previous studies have suggested that macrophages may contribute to acute Leptospira dissemination, as well as having a major role in kidney fibrosis. Our aim was to characterize the role of macrophages and galectin 3 (Gal-3) on the survival, clinical course, bacterial burden, interstitial nephritis, and chronic kidney fibrosis in Leptospira interrogans serovar Copenhageni (LIC)-induced experimental murine leptospirosis. C57BL/6J mice depleted of macrophages by liposome-encapsulated clodronate treatment and infected with LIC presented a higher bacterial burden, had reduced subacute nephritis and enhanced chronic kidney fibrosis relative to untreated, infected mice. Moreover, LIC infection in mice whose Gal-3 was disrupted (Lgals3^-/-) had a higher bacterial burden and enhanced subacute nephritis and chronic kidney fibrosis when compared to C57BL/6J wild-type mice. Chronic fibrosis did not correlate with higher transcription levels of TGF-β1 or IL-13 in the kidneys. Kidney fibrosis was found in chronically infected rats as well as in wild infected rats. On the other hand, human fibroblast cultures exhibited enhanced differentiation to myofibroblasts after treatment with LIC. Our results demonstrate that macrophages and Gal-3 play a critical role in controlling the LIC burden but has a minor role in subsequent fibrosis. Instead, kidney fibrosis was better correlated with bacterial burden. Taken together, our results do not support a role for macrophages to disseminate leptospires during acute infection, nor in chronic kidney fibrosis.

Keywords: Leptospira; fibrosis; galectin 3; macrophages; pathogenesis.

Figure 1

Depletion of macrophages resulted in a higher bacterial burden, reduced subacute nephritis, and unaltered renal function. (A) Treatment scheme for intraperitoneal depletion and infection in Mock, Mock + LipClod, LIC and LIC + LipClod groups of C57BL/6J mice. (B) Bacteremia levels at 1 day post-infection (dpi) were significantly increased in LIC-infected mice depleted of macrophages. (C,D) At 14 dpi, kidneys sections of all groups of mice were immunostained for LipL32 (the major Leptospira interrogans antigen) and also bacterial load in LIC + LipClod group resulted in higher levels compared to undepleted but LIC-infected mice. (E–G) Representative haematoxylin-eosin staining of kidney sections. An inflammation score was established by a pathologist and CCL-2 mRNA was quantified by qRT-PCR as a molecular marker of inflammation; β-actin was used as a housekeeping control. Significantly diminished levels of inflammation were detected in LIC + LipClod group. (H,I) Normal urea and creatinine levels in serum samples from all experimental groups were detected and assayed as a measure of renal functions at 14 dpi. Serum samples from LIC-infected gerbils were used as a positive control (+). In all cases, data represent assays of two independent experiments of groups of 6 animals, each. Bars indicate 50 microns; AU, arbitrary units; ^nsP > 0.05; ^*P < 0.05; ^**P < 0.01; ^***P < 0.001; ^****P < 0.0001.

Figure 2

Depletion of macrophages resulted in an increased chronic renal bacterial burden and similar chronic interstitial nephritis. (A) Treatment scheme for intraperitoneal depletion and infection in Mock, Mock + LipClod, LIC and LIC + LipClod groups of C57BL/6J mice. (B) Percentage of body weight change in all groups of mice during the 45 days post-infection (dpi) showing that the LIC + LipClod group had significantly reduced weight gain. Immunostaining with an antibody against F4/80+ cells (macrophages) (C) confirming macrophage depletion or with anti-LipL32 (the major Leptospira interrogans antigen) (D) in kidney sections of all experimental groups of mice at 45 dpi. Depletion of macrophages in LIC-infected mice resulted in more presence of leptospiral antigen. (E) Higher bacterial load in kidneys of LIC + LipClod group was detected compared to undepleted LIC group. β-actin was used as a housekeeping control. (F–H) Representative haematoxylin-eosin staining of kidney sections. An inflammation score was established by a pathologist and CCL-2 mRNA was quantified by qRT-PCR as a molecular marker of inflammation; β-actin was used as a housekeeping control. Similar chronic inflammation was observed among experimental groups. (I,J) Normal urea and creatinine levels in serum samples from all experimental groups were detected and assayed as a measure of renal functions at 45 dpi. In all cases, data represent assays of two independent experiments of groups of 6 animals, each. Bars indicate 50 microns; AU, arbitrary units; ^nsP > 0.05; ^*P < 0.05; ^**P < 0.01.

Figure 3

Depletion of macrophages resulted in an increased fibrosis. (A,B) Picrosirius red was used for collagen staining of kidney sections for all groups of mice at 45 dpi and digital quantification of fibrosis revealed significantly increased collagen deposition in the LIC + LipClod group of mice at 45 dpi. (C–E) α-smooth muscle actin protein immunostaining in all groups of mice and representative western blot image of α-smooth muscle actin protein in kidney samples from LIC and LIC + LipClod groups of mice at 45 dpi; β-actin was used as a housekeeping control. Data represent assays of two independent experiments of groups of 6 animals, each. For digital quantification, the samples derive from the same experiment and the gels/blots were processed in parallel. AU, arbitrary units; ^*P < 0.05; ^****P < 0.0001.

Figure 4

Genetic disruption of Gal-3 significantly increased the acute bacterial burden and acute kidney inflammation. (A) Immunostaining with an antibody against LipL32 (the major Leptospira interrogans antigen) in kidney sections of Mock, Mock + Lgals3^−/−, LIC and LIC + Lgals3^−/− groups of mice at 14 days post-infection (dpi). (B) Absence of galectin-3 in LIC-infected mice resulted in higher bacterial load in kidneys compared to wild-type LIC group. β-actin was used as a housekeeping control. (C–E) Representative haematoxylin-eosin staining of kidney sections. An inflammation score was established by a pathologist and CCL-2 mRNA was quantified by qRT-PCR as a molecular marker of inflammation; β-actin was used as a housekeeping control. Acute kidney inflammation as well as significantly higher CCL-2 levels were detected in LIC + Lgals3^−/− group. (F,G) Normal urea and creatinine levels in serum samples from all experimental groups were detected and assayed as a measure of renal functions at 14 dpi. Serum samples from LIC-infected gerbils were used as a positive control (+). In all cases, data represent assays of two independent experiments of groups of 6 animals, each. Bars indicate 50 microns; AU, arbitrary units; ^nsP > 0.05; ^*P < 0.05; ^***P < 0.001; ^****P < 0.0001.

Figure 5

Genetic disruption of Gal-3 significantly increased the chronic bacterial burden. (A) Immunostaining with an antibody against LipL32 (the major Leptospira interrogans antigen) in kidney sections of Mock, Mock + Lgals3^−/−, LIC and LIC + Lgals3^−/− groups of mice at 45 days post-infection (dpi). (B) Absence of galectin-3 in LIC-infected mice resulted in higher bacterial load in kidneys compared to wild-type LIC group. β-actin was used as a housekeeping control. (C) Immunostaining with an antibody against F4/80+ cells (macrophages) revealed its presence in both infected groups independent of galectin-3 absence. (D–F) Representative haematoxylin-eosin staining of kidney sections. An inflammation score was established by a pathologist and CCL-2 mRNA was quantified by qRT-PCR as a molecular marker of inflammation; β-actin was used as a housekeeping control. Similar chronic kidney inflammation was observed among experimental groups. (G,H) Normal urea and creatinine levels in serum samples from all experimental groups were detected and assayed as a measure of renal functions at 45 dpi. In all cases, data represent assays of two independent experiments of groups of 6 animals, each. Bars indicate 50 microns; AU, arbitrary units; ^nsP > 0.05; ^*P < 0.05.

Figure 6

Genetic disruption of Gal-3 significantly increased the chronic fibrosis with similar chronic inflammation. (A,B) Picrosirius red was used for collagen staining of kidney sections for all groups of mice at 45 dpi and digital quantification of fibrosis revealed significantly increased collagen deposition in the LIC + Lgals3^−/− group of mice at 45 dpi. (C–E) α-smooth muscle actin protein immunostaining in all groups of mice and representative western blot image of α-smooth muscle actin protein in kidney samples from LIC and LIC + Lgals3^−/− groups of mice at 45 dpi; β-actin was used as a housekeeping control. Data represent assays of two independent experiments of groups of 6 animals, each. For digital quantification, the samples derive from the same experiment and the gels/blots were processed in parallel. AU, arbitrary units; ^**P < 0.01; ^****P < 0.0001.

Figure 7

Cytokine mRNA levels in renal interstitial fibrosis triggered by LIC infection showing that fibrosis was not associated with TGF-β1 or IL-13 mRNA levels in the kidney. qRT-PCR analysis of TGF-β1 (A,B), IL-13 (C,D), IFN-γ (E,F), IL-4 (G,H), and IL-10 (I,J) mRNA levels in kidney samples from LIC, LIC + LipClod, and LIC + Lgals3^−/− groups of mice at 45 dpi. β-actin was used as a housekeeping control. Data represent assays of two independent experiments of groups of 6 animals, each; AU, arbitrary units; ^nsP > 0.05; ^*P < 0.05.

Figure 8

Leptospira triggered kidney fibrosis in experimentally infected and wild rats. (A,B) Picrosirius red was used for collagen staining of kidney sections for Mock, and experimentally LIC-infected rats and digital quantification of fibrosis revealed significantly increased collagen deposition in the LIC group at 45 dpi. (C) Wild rats were screened for Leptospira and 3 out of 48 animals were PCR-positive and Picrosirius red was used for collagen staining of kidney sections. (D) Digital quantification of fibrosis revealed significantly increased collagen deposition in the Infected (PCR-positive) compared with the Uninfected (PCR-negative) rats at 45 dpi. ^*P < 0.05; ^****P < 0.0001.

Figure 9

LIC activated human fibroblasts. (A) Immunofluorescent staining of α-smooth muscle actin protein (a specific marker of activated fibroblasts or myofibroblasts) in a primary cell line of cultured human fibroblast treated previously for 48 h with Mock, LIC conditioned-medium, and LIC. Propidium iodide was used for nuclear staining. (B) Representative FACS plots of SMA mean fluorescent intensity (MFI). (C) Representative FACS plots of SMA MFI of a co-culture (Co-) of fibroblast and macrophages infected or not for 48 h with LIC in absence or presence of a non-specific Galectin-3 inhibitor (Gal3i). Bar indicate 50 microns. ^*P < 0.05; ^***P < 0.001.