Inducible expression of kallikrein in renal tubular cells protects mice against spontaneous lupus nephritis



This article is corrected by:

  1. Errata: Error in Author Name in Article by Shao et al (Arthritis Rheum, March 2013) Volume 65, Issue 4, 1106, Article first published online: 28 March 2013



To ascertain whether engineered expression of kallikreins within the kidneys, using an inducible Cre/loxP system, can ameliorate murine lupus nephritis.


In mice with a lupus-prone genetic background, we engineered the expression of tamoxifen-inducible Cre recombinase under the control of a kidney-specific promoter whose activation initiates murine kallikrein-1 expression within the kidneys. These transgenic mice were injected with either tamoxifen or vehicle at age 2 months and then were monitored for 8 months for kallikrein expression and disease.


Elevated expression of kallikrein was detected in the kidney and urine of tamoxifen-injected mice but not in controls. At age 10 months, all vehicle-injected mice developed severe lupus nephritis, as evidenced by increased proteinuria (mean ± SD 13.43 ± 5.65 mg/24 hours), increased blood urea nitrogen (BUN) and serum creatinine levels (39.86 ± 13.45 mg/dl and 15.23 ± 6.89 mg/dl, respectively), and severe renal pathology. In contrast, the tamoxifen-injected mice showed significantly reduced proteinuria (6.6 ± 4.12 mg/24 hours), decreased BUN and serum creatinine levels (15.71 ± 8.17 mg/dl and 6.64 ± 3.39 mg/dl, respectively), and milder renal pathology. Tamoxifen-induced up-regulation of renal kallikrein expression increased nitric oxide production and dampened renal superoxide production and inflammatory cell infiltration, alluding to some of the pathways through which kallikreins may be operating within the kidneys.


Local expression of kallikreins within the kidney has the capacity to dampen lupus nephritis, possibly by modulating inflammation and oxidative stress.

Systemic lupus erythematosus (SLE) is a systemic autoimmune disease characterized by autoantibody production and immune complex (IC) deposition at multiple sites (1–6). Immune-mediated nephritis caused by deposition of pathogenic autoantibodies and ICs in the glomeruli contributes to mortality and morbidity in this disease. Studies of spontaneous lupus nephritis in murine models and experimental anti–glomerular basement membrane (anti-GBM) disease have provided valuable insights into the underlying mechanisms of human lupus nephritis (7–12). The experimentally induced model of anti-GBM disease has proven to be a particularly useful tool for studying the susceptibility of end organs to immune-mediated damage (7–10).

Previous studies by our group revealed that NZW mice exhibit increased susceptibility to experimental anti-GBM–induced glomerulonephritis (EAG) compared with C57BL/6 (B6) mice (10). Several lupus susceptibility loci derived from the NZW/NZM2410 strain have already been bred onto the normal B6 background as congenic intervals (13–21). One such lupus susceptibility interval is Sle3, on chromosome 7 (17, 20, 21). Unlike other congenic strains tested, B6.Sle3 mice (bearing the NZW/NZM2410-derived “z” allele of Sle3 on the nonautoimmune B6 background) exhibit increased susceptibility to EAG compared with B6 mice (8, 21, 22). Further fine mapping using recombinant subcongenics bearing progressively narrowed subintervals of Sle3 revealed the susceptibility genes to be located between D7mit157 and D7mit158 on chromosome 7 (21), an interval harboring the kallikrein (klk) cluster of genes. Microarray and real-time polymerase chain reaction (PCR) studies indicated that the B6.Sle3 congenic as well as several EAG-susceptible strains (such as 129/SvJ, NZW, and DBA/1) had significantly reduced renal expression of kallikreins compared with B6 and BALB/c control strains, following anti-GBM challenge. Furthermore, there were also indications that particular klk alleles may be associated with lupus nephritis in patients with SLE (22).

The above-mentioned studies suggested that kallikreins may be renoprotective in immune-mediated nephritis. The renoprotective effect of kallikrein was further underscored by 2 functional experiments. Delivering the kallikrein-1 (Klk1) gene via adenoviral vector into B6.Sle3 mice (EAG-sensitive strain) attenuated the severity of anti-GBM–induced nephritis (23), while blocking the kallikrein function in BALB/c mice (EAG-resistant strain) using a kinin B2 receptor antagonist (HOE140) rendered these mice sensitive to anti-GBM–induced nephritis (22, 24). However, the studies did not address the question of whether kallikreins could also modulate the renal injury that arises during the course of spontaneous lupus nephritis. Therefore, we established a novel transgenic mouse strain with inducible, kidney-specific kallikrein expression on the background of B6.Sle1.Sle3 of double-congenic mice, using the Cre-ERT2/LoxP system. Importantly, renal-specific expression of Klk1 significantly ameliorated lupus nephritis. The protective effect of Klk1 against lupus nephritis may be related to suppression of oxidative stress and reduction of inflammation.


Transgenic mice.

CAG-lacZflx/−-Klk1-EGFP–transgenic mice.

The murine Klk1 gene coding region was amplified using primers (forward 5′-AGCGTCGACTCCTGTTACCATGAGG-3′ and reverse 5′-AGCGTCGACAATGTGATACTCAGTC-3′) and inserted into the multiple cloning site (Sal I site) in the CAG-lacZflx/−-EGFP vector system (a kind gift from Dr. Tomokazu Fukuda [25]). In this construct, the expression of Klk1 is inhibited by the presence of lacZ, which is flanked by loxP sites (floxed). In the absence of active Cre, the ubiquitous CAG promoter drives the expression of lacZ, thus providing a detectable marker of cells expressing the unmodified construct (25, 26). The recombinant construct that contains the murine Klk1 coding region in front of IRES-EGFP was digested using Afl II and Spe I to remove it from the vector. This DNA fragment was extracted from the gel (QIAquick Gel Extraction Kit; Qiagen), and diluted to 2 μg/ml, using 1 mM Tris HCl (pH 8.0) and 0.1 mM EDTA. The purified transgene DNA was submitted to the Transgenic Core Facility of University of Texas–Southwestern Medical Center for pronuclei microinjection into 0.5-day-old B6 mouse embryos (The Jackson Laboratory). Microinjected embryos were cultured in KSOM (Sigma) for 1 day, and embryos that reached the 2-cell stage were transferred into the oviducts of pseudopregnant female mice. Transgenic founders were identified by PCR analysis of tail biopsy specimens using primers for the Klk1 and green fluorescent protein (GFP) genes (forward 5′-GTTCATCTGCACCACCGGC-3′ and reverse 5′-TTGTGCCCCAGGATGTTGC-3′). Transgenic founders were crossed with C57BL/6J mice to generate permanent lines. The offspring were genotyped by PCR detection of the inserted lacZ and murine Klk1 gene fragments and X-Gal staining of multiple tissues. One line that highly expressed the transgene in the kidney was selected for further study.

Ksp/CreERT2-transgenic mice.

This transgenic line has been established previously, and its unique tissue-specific properties have been confirmed (26, 27). In this transgenic mouse, the expression of CreERT2 fusion protein is under the control of the kidney-specific Ksp-cadherin (cadherin 16) gene promoter, which directs gene expression exclusively in kidney tubular epithelial cells. The CreERT2 protein contains Cre recombinase fused in-frame to a modified human estrogen receptor and is normally retained in the cytosol in an inactive state. Upon binding to 4-hydroxytamoxifen, the fusion protein translocates to the nucleus where it mediates Cre/loxP recombination (28). By crossing CreERT2-transgenic mice with mice containing essential exon(s) of a gene of interest flanked by loxP sites, gene targeting can be temporally regulated. All mice were bred on the B6 background.

Breeding of Ksp/CreERT2 mice with CAG-lacZflx/−-Klk1-EGFP mice.

The Ksp/CreERT2–transgenic and CAG-lacZflx/−-Klk1-EGFP–transgenic strains were crossed to each other to generate a Ksp/CreERT2/CAG-lacZflx/−-Klk1-GFP double-transgenic line. The offspring bearing both Ksp/CreERT2 and CAG-lacZflx/−-Klk1-GFP transgenes were identified by genotyping, using multiple PCR primer sets including Cre (forward 5′-AGGTTCGTGCACTCATGGA-3′ and reverse 5′-TCGACCAGTTTAGTTACCC-3′) and Klk1-GFP (forward 5′-GTTCATCTGCACCACCGGC-3′ and reverse 5′-TTGTGCCCCAGGATGTTGC-3′). The litters carrying both transgenes were crossed to each other to generate a constant line of double-transgenic mice.

Creation of Ksp/CreERT2/CAG-lacZflx/−-Klk1-GFP/Sle1.Sle3 mice.

The Ksp/CreERT2/CAG-lacZflx/−-Klk1-GFP mice were then crossed with B6.Sle1.Sle3-congenic mice to generate transgenic mice carrying homozygous Sle1 and Sle3 lupus-susceptibility intervals, all on the B6 background (20, 21). This bicongenic strain of mouse was generated previously and has been a reliable tool for studying spontaneous lupus nephritis (13–21). After 2–3 rounds of backcrossing with the B6.Sle1.Sle3 mice, we successfully generated a recombinant strain, Ksp/CreERT2/CAG-lacZflx/−-Klk1/Sle1.Sle3. Genotyping with primer sets for Sle1 (for primer pair 1, forward 5′-GTGTCTGCCTTTGCACCTTT-3′ and reverse 5′-CTGCTGTCTTTCCATCCACA-3′; for primer pair 2, forward 5′-CCATAAGCCTCCTGTTTCCC-3′ and reverse 5′-AAAATGAACTCAGCGGGTTG-3′) and Sle3 (for primer pair 1, forward 5′-CTTCATCTGAGCCTGGGAAG-3′ and reverse 5′-ACTGTAGACCCATGTTCTGATTAGG-3′; for primer pair 2, forward 5′-TGAATTCACACACATGTGCG-3′ and reverse 5′-TGAATGCAGATTCCTTCATCC-3′) distinguished the mice that were homozygous for Sle1 and Sle3. All experiments involving animals were performed under the auspices of the University of Texas–Southwestern Institutional Animal Care and Research Advisory Committee.

Induction of Klk1 expression in Ksp/CreERT2/CAG-lacZflx/−-Klk1/Sle1.Sle3 mice with tamoxifen.

To induce Klk1 expression, 14 female Ksp/CreERT2/CAG-lacZflx/−-Klk1/Sle1.Sle3 mice were selected, allowed to reach age 2 months, and randomly divided into 2 groups (the tamoxifen group and the vehicle control group). Tamoxifen or 4-hydroxytamoxifen (Sigma), which produced equivalent results, was dissolved in 95% ethanol and then diluted 10-fold in sunflower oil to a final concentration of 10 mg/ml. Mice in the tamoxifen group received tamoxifen (4 mg/40 gm body weight) daily by intraperitoneal injection for 5 consecutive days. Control mice received an equal volume of vehicle alone (9.5% ethanol in sunflower oil). The mice were monitored and then killed 8 months after administration of the first injection.

Blood and urine collection.

Blood was collected from anesthetized mice through the retroorbital sinus at 5 different time points: on day 0 and then at bimonthly intervals for 8 months. Blood samples were centrifuged at 1,000g for 20 minutes, and serum samples were collected and stored at −20°C. Twenty-four–hour urine samples were collected at the 5 time points, using metabolic cages. Mice received only water in order to eliminate contamination of food particles in the urine samples. After collection, the urine samples were centrifuged at 1,000g and used for further analysis.

Evaluation of renal disease.

Antinuclear autoantibody (ANA) production, 24-hour proteinuria, blood urea nitrogen (BUN), and serum and urine creatinine were measured to monitor the progression of spontaneous lupus nephritis. ANAs were measured using an ANA enzyme-linked immunosorbent assay kit (Alpha Diagnostic). The urinary protein concentration was determined using a Coomassie Plus Assay Kit (Thermo Scientific Pierce). BUN was measured with a BUN detection kit (Sigma-Aldrich). Serum and urine creatinine levels were measured using a commercial kit (BioAssay System).

Kallikrein, nitrate/nitrite (NOx), and superoxide assays.

The chromogenic substrate H-D-Val-Leu-Arg-pNA (S-2266; DiaPharma) was used to assay urine kallikrein levels. Briefly, 50 μl of urine was added to 50 μl of assay buffer (0.2M Tris HCl, pH 8.2, containing 300 μg/liter soybean trypsin inhibitor and 375 μg/liter EDTA) and incubated at 37°C for 30 minutes. Then, 50 μl of S-2266 was added and incubated at 37°C for 3 hours, and the absorbance was read at 405 nm. The activity of kallikrein per liter of urine or the amount excreted over 24 hours was then calculated. Urine and renal NOx levels were measured with a fluorometric assay, using a QuantiChrom Nitric Oxide Assay Kit (Medibena). Briefly, renal tissue specimens were collected at the end of the experiment and homogenized in lysis buffer (10 mM Tris, pH 7.4, 100 mM NaCl, 1 mM EDTA, 20 mM Na4P2O7, 2 mM Na3VO4, and 1% Triton X-100) containing 1:100 protease inhibitor cocktail (Sigma). Nitrate reduction was assayed by applying activated cadmium granules in the tissue samples for 15 minutes; then, the oxidized NOx level was deduced by recording the optical density at 540 nm. Superoxide levels in renal extracts were quantified by a spectrophotometric assay based on rapid reduction of ferricytochrome c to ferrocytochrome c, according to a modified protocol (29). Reduction of ferricytochrome c independent of superoxide was corrected for by deducting the activity not inhibited by superoxide dismutase. Myeloperoxidase activity was determined as previously described (30).

Western blot analysis.

Western blot analysis was performed using tissue extracts from mouse kidneys. Renal cortex was homogenized in lysis buffer (25 mM Tris HCl, pH 7.4, 1% Triton X-100, 0.1% sodium dodecyl sulfate, 2 mM EDTA) containing 1:100 protease inhibitor cocktail (Sigma-Aldrich) and centrifuged at 4°C for 30 minutes. Protein concentrations were determined using the Lowry method. The homogenates were subjected to Western blot analysis using a custom-generated rabbit anti-mouse Klk1 antibody (1:1,000; Abgent) and chemiluminescence detection (Thermo Scientific Pierce). Anti-GAPDH antibody (Sigma) was used as a loading control.

Immunohistochemical staining.

Immunohistochemical staining was performed on cryosectioned mouse kidney sections. The following antibodies and dilutions were used: murine Klk1 (1:1,000; Abgent), GFP (1:1,000; Invitrogen), Cre (1:1,000; Molecular Probes), kidney injury molecule 1 (KIM-1) (1:1,000; R&D Systems), and ionized calcium–binding adapter molecule 1 (IBA-1) (1:800; Abcam). Secondary antibodies were conjugated to Alexa Fluor 488 or Alexa Fluor 594 (1:200; Molecular Probes) or horseradish peroxidase. Tissue sections were incubated with the primary antibodies at 4°C overnight or at room temperature for 2 hours after blocking of nonspecific binding with blocking solution (10% goat serum, 0.1% bovine serum albumin in phosphate buffered albumin). The tissue sections were incubated with the appropriate secondary antibodies for 1 hour at room temperature. The slides were mounted with Vectashield and examined using a Leica fluorescence microscope. Images were then acquired using a digital camera.

Morphologic and histologic analyses.

Mice were perfused through the left ventricle with 4% paraformaldehyde. The kidneys were removed and sectioned with a cryostat. Sections were stained with periodic acid–Schiff (PAS) or hematoxylin and eosin (H&E). These sections were examined by 2 investigators in a blinded manner, for any evidence of pathologic changes in the glomeruli, tubules, or interstitial areas. The glomeruli were screened for evidence of hypertrophy, proliferative changes, crescent formation, hyaline deposits, fibrosis/sclerosis, and basement membrane thickening.

The severity of glomerulonephritis was graded on a 0–4 scale, where 0 = normal, 1 = mild increase in mesangial cellularity and matrix, 2 = moderate increase in mesangial cellularity and matrix with thickening of the glomerular basement membrane (GBM), 3 = focal endocapillary hypercellularity with obliteration of capillary lumina and a substantial increase in the thickness and irregularity of the GBM, and 4 = diffuse endocapillary hypercellularity, segmental necrosis, crescents, and hyalinized end-stage glomeruli. Similarly, the severity of tubulointerstitial nephritis was graded on a 0–4-point scale, based on the extent of tubular atrophy, inflammatory infiltrates, and interstitial fibrosis, as detailed previously (31). The numbers of infiltrating polymorphonuclear leukocytes, including neutrophils, were directly enumerated based on their typical polymorphonuclear morphology, by examining 50 glomeruli per kidney section per mouse. The extent of monocyte/macrophage infiltration was measured by immunohistochemical staining with anti–IBA-1 antibody (1:500) and Alexa Fluor 488–labeled secondary antibody (1:200). The average percentage of positive cells per total number of cells per high-power field in 10 different fields for each section from each mouse was calculated.

Quantitative real-time PCR.

Renal cortex Klk1 gene expression differences were validated by quantitative real-time PCR using validated TaqMan assays (Applied Biosystems). Total RNA was extracted from kidney using TRIzol (Invitrogen). Complementary DNA (cDNA) was transcribed using a High-Capacity cDNA Archive Kit (Applied Biosystems). Quantitative real-time PCRs were carried out on an Applied Biosystems 7300 Real-Time PCR System, using 10 μl of 2× TaqMan Universal PCR Master Mix and 1 μl of 20× TaqMan Gene Expression Master Mix for murine Klk1 (Mm00834006_g1). Transcription of eukaryotic 18S ribosomal RNA (Hs99999901_s1) was used as an internal control.

Multiplex cytokine assays for detection of inflammatory cytokines.

A Cytokine Mouse Magnetic 20-Plex Panel (Life Technologies) was used for cytokine detection, using a MAGPIX System (Luminex). Kidney proteins were extracted from transgenic mice treated with tamoxifen or vehicle and analyzed for expression of cytokines, including interferon-α, interleukin-1 (IL-1), IL-6, IL-10, IL-12, monocyte chemotactic protein 1, transforming growth factor β1, and tumor necrosis factor α (TNFα). The assays were performed according to the manufacturers' specifications. Standard curves were generated by using reference concentrations supplied by the manufacturer. Each assay was performed in triplicate, and ≥3 mice were included in each group.

Statistical analysis.

Data are presented as the mean ± SD. Differences between the means were calculated using Student's t-test. Analysis of variance was used for multiple comparisons. P values less than 0.05 were considered significant.


Generation of Ksp/CreERT2/CAG-lacZflx/−-Klk1/Sle1.Sle3–congenic/transgenic mice.

The role of Klk1 in spontaneous lupus nephritis was studied using kidney-specific inducible gene targeting. The strategies for generating the transgenic strains are illustrated in Figure 1A. Three congenic/transgenic lines were involved in this study. First, we generated a transgenic line, CAG-lacZflx/−-Klk1-GFP. The transgenic founders were identified by PCR analysis of the DNA fragment encompassing both the LacZ and Klk1 genes (Figure 1B). The transgenic founders were then crossed with the second transgenic line, Ksp/CreERT2 (26, 27), to generate the double-transgenic strain, Ksp/CreERT2/CAG-lacZflx/−-Klk1-GFP, which harbors both the Cre and Klk1 transgenes (Figure 1B). The double-transgenic strain was bred onto B6.Sle1.Sle3 mice, which is a double-congenic lupus-prone strain developed previously (11–17). After several rounds of breeding, Ksp/CreERT2/CAG-lacZflx/−/Klk1/Sle1.Sle3 mice, bearing homozygous Sle1 and Sle3 intervals on the B6 background, were generated.

Figure 1.

Generation and characterization of the congenic/transgenic mice. A, Cloning and breeding strategy used to generate the target mice. Two transgenic strains, Ksp/CreERT2 and CAG-lacZflx/−-Klk1-GFP, both on the C57BL/6 (B6) background, were crossed with each other to generate a double-transgenic strain, Ksp/CreERT2/CAG-lacZflx/−-Klk1-GFP. This strain was then crossed with B6.Sle1.Sle3 double-congenic mice to generate the final target strain, Ksp/CreERT2/CAG-lacZflx/−-Klk1-GFP/Sle1.Sle3, which carries the transgenes CreERT2, which is driven by the Ksp promoter, and lacZflx/−-Klk1-GFP, which is driven by the CAG promoter, on the B6.Sle1.Sle3 genetic background. In effect, when induced with tamoxifen, this dual vector system will initiate high-level Klk1 expression in renal tubular cells. B, Genotyping analysis of mouse genomic DNA (from tail) by polymerase chain reaction amplification of Cre and lacZ-Klk1 gene fragments. C–H, Detection of LacZ expression by X-Gal staining in different tissue specimens from transgenic mice (C–F) and control mice (G and H). C, Kidney cortex. D, Kidney medulla. E, Liver. F, Spleen. G and H, Kidney and liver, respectively, from B6 mice as control. Images are representative of samples from 5 mice each. IRES = internal ribosome entry site; EGFP = enhanced green fluorescent protein; pA = polyA.

High-level lacZ expression was detected in the kidneys of the transgenic mice, including the proximal tubule segments, loop of Henle, collecting ducts, and glomeruli, by X-Gal staining (Figures 1C and D). LacZ signals were also detected in other tissues, including the liver and spleen (Figures 1E and F), although the expression levels were lower. No lacZ signal was identified in the kidney and liver of the nontransgenic B6 controls (Figures 1G and H).

Inducible expression of Klk1 in Ksp/CreERT2/CAG-lacZflx/−-Klk1 mice.

The Ksp/CreERT2/CAG-lacZflx/−-Klk1 mice were administered tamoxifen or vehicle (sunflower oil) and killed 2 weeks later. Quantitative PCR showed that Klk1 expression was significantly increased in the kidneys of the tamoxifen-injected mice compared with the kidneys of the vehicle-treated group (P < 0.001) (Figure 2A). The elevated expression of Klk1 in the kidneys of tamoxifen-injected mice was also confirmed by Western blotting (Figure 2B). No Klk1 expression was detected in the other tissues (Figures 2A and B). Immunohistochemical staining showed that the expression of Klk1 and GFP protein was colocalized in the renal tubular cells (Figures 2C and D).

Figure 2.

Tamoxifen-induced expression of the kallikrein-1 gene (Klk1) within the kidney. A, Klk1 gene expression in different tissues after treatment with tamoxifen or vehicle, as determined by polymerase chain reaction. Bars show the mean ± SD of 5 mice per treatment group for each organ type. B, Klk1 protein expression in different tissues after treatment with tamoxifen or vehicle, as determined by Western blotting. Blots are representative of 5 mice per group. C and D, Immunohistochemical staining showing that Klk1 and GFP protein are colocalized in renal tubular cells of transgenic mice after tamoxifen treatment. Images are representative of 5 mice per group. RQ = relative quantity (see Figure 1 for other definitions).

Renal-specific activation of Klk1 in mice with spontaneous lupus nephritis.

Female Ksp/CreERT2/CAG-lacZflx/−-Klk1/Sle1.Sle3 mice were allowed to reach age 2 months and then were randomly divided into 2 groups. One group of mice was administered tamoxifen intraperitoneally for 5 consecutive days to induce Cre/loxP recombination, while the control group received only vehicle sham (sunflower oil). All mice were observed for an additional 8 months before being killed. Figures 3A and B show the renal expression of Cre, Klk1, and GFP in both groups, as determined by Western blotting. Cre expression was detected in all mice independent of tamoxifen treatment, although the expression level in each mouse was variable (Figure 3A). Increased expression of Klk1 and GFP protein was observed in all mice receiving tamoxifen (Figure 3A). Quantitative analysis showed that the expression of Klk1 and GFP was significantly increased in tamoxifen-treated mice compared with vehicle-treated mice (P < 0.05) (Figure 3B). Immunohistochemical staining showed that Klk1 signals were localized in the renal tubular cells of tamoxifen-treated mice (Figure 3C) but not vehicle-treated mice (Figure 3D).

Figure 3.

Detection of Klk1 expression in the kidneys of Ksp/CreERT2/CAG-lacZflx/−-Klk1/Sle1.Sle3 mice. A and B, Cre, Klk1, and GFP protein expression in the kidneys of mice treated with tamoxifen or vehicle, as determined by Western blotting (A), and quantitative analysis of the expression of Cre, Klk1, and GFP (B). After normalization to GAPDH, the expression of Klk1 and GFP was significantly increased in the kidneys of tamoxifen-treated mice compared with vehicle-treated mice. Bars show the mean ± SD of 5 mice per group. ∗ = P < 0.05 versus vehicle-treated mice. C and D, Klk1 expression as determined by immunohistochemical staining with anti-Klk1 antibody. Klk1 expression was detected in the tubular cells of tamoxifen-treated mice (C) but not vehicle-treated mice (D). Images are representative of 5 mice per group. RQ = relative quantity (see Figure 1 for other definitions).

Effect of Klk1 activation on renal function of mice with spontaneous lupus nephritis.

After tamoxifen administration, blood and urine specimens from both groups of mice were collected at the 2-, 4-, 6-, and 8-month time points in order to assay proteinuria, BUN, and serum creatinine. As expected, all mice developed some degree of renal damage, as gauged by the level of proteinuria, serum BUN, and creatinine at 6 months of age (4 months after treatment). At the 8-month time point, however, the levels of proteinuria, BUN, and creatinine were significantly reduced in mice treated with tamoxifen compared with vehicle-treated mice (mean ± SD 6.6 ± 4.12 mg/24 hours versus 13.43 ± 5.65 mg/24 hours [P = 0.018], 15.71 ± 8.17 mg/dl versus 39.86 ± 13.45 mg/dl [P = 0.0018], and 6.64 ± 3.39 mg/dl versus 15.23 ± 6.89 mg/dl [P = 0.039], respectively) (Figures 4A–C).

Figure 4.

Effect of Klk1 activation on renal function and spontaneous lupus nephritis. A–C, Levels of proteinuria (A), blood urea nitrogen (BUN) (B), and serum creatinine (C) in Ksp/CreERT2/CAG-lacZflx/−-Klk1/Sle1.Sle3 mice at 2 months, 4 months, 6 months, and 8 months after treatment with tamoxifen or vehicle. Each data point represents a single mouse; horizontal lines show the mean. D–G, Renal injury, as evaluated by histologic examination of tissue sections stained with hematoxylin and eosin (H&E) and periodic acid–Schiff (PAS). Vehicle-treated mice showed significant renal damage in both the cortex and medulla, including crescent formation, intracapillary hypercellularity with capillary lumen obliteration, tubular dilatation, loss of proximal brush border, as well as accumulation of protein casts in the tubules (D and F). In contrast, tamoxifen-treated mice exhibited well-preserved kidney morphology, with reduced renal damage marked by fewer necrotic cells, fewer dilated tubules with less protein casts, and attenuation of brush border loss (E and G). H and I, Glomerulonephritis (GN) and tubulointerstitial nephritis (TIN) scores for renal damage in the tamoxifen-treated group and the vehicle-treated group. Values are the mean ± SD of 5 mice per group. ∗ = P < 0.05 versus tamoxifen-treated mice.

Renal injury was evaluated by histologic examination of tissue sections stained with H&E and PAS (Figures 4D–G). The tamoxifen-treated mice exhibited well-preserved renal morphology (Figures 4E and G). In contrast, the vehicle-treated mice showed significant renal damage in both the renal cortex and medulla, including crescent formation, intracapillary hypercellularity with capillary lumen obliteration, tubular dilatation, loss of the proximal brush border, as well as accumulation of protein casts in the tubules (Figures 4D and F). Activation of Klk1 resulted in reduced renal damage, as marked by fewer necrotic cells, fewer dilated tubules with fewer protein casts, and attenuation of brush border loss (Figures 4E and G). The extent of renal damage was assessed in PAS-stained slides and quantified using the glomerulonephritis score and tubulointerstitial nephritis score, as described previously (8, 9, 30). The mean ± SD glomerulonephritis score in the tamoxifen-treated group was significantly reduced compared with that in the vehicle-treated group (1.6 ± 0.21 versus 3.30 ± 0.51; P < 0.05) (Figure 4H). Similarly, tamoxifen treatment also reduced the severity of tubular injury, as gauged by the tubulointerstitial nephritis score (0.9 ± 0.4 versus 2.8 ± 0.6; P < 0.01) (Figure 4I). These findings indicated that renal-specific activation of Klk1 attenuated spontaneous lupus nephritis in vivo.

Activation of the Klk1 gene in mice with lupus may confer renoprotection by modulating nitric oxide and superoxide formation.

To investigate the possible mechanisms underlying the protective role of klk in lupus nephritis, renal NOx levels and superoxide formation were analyzed. Tamoxifen-injected mice exhibited increased renal NOx levels compared with the vehicle-treated group (mean ± SD 70.68 ± 8.04 μmoles/mg versus 33.38 ± 3.64 μmoles/mg; P = 0.013) (Figure 5A). Immunohistochemical staining with antinitrotyrosine antibody revealed NOx expression in the renal tubular cells of tamoxifen-treated B6.Sle1.Sle3 mice but not in vehicle- or sham-treated mice (Figures 5C–E). In contrast, renal superoxide formation was lower in tamoxifen- treated mice compared with vehicle-treated mice (75.12 ± 6.42 μmoles/minute−1 · mg protein−1 versus 117.7 ± 8.67 μmoles/minute−1 · mg protein−1; P = 0.017) (Figure 5B). These results indicated that Klk1 activation significantly reduced superoxide formation, and that this might contribute to the protective role of klk in lupus nephritis.

Figure 5.

Levels of renal nitrate/nitrite (NOx) and superoxide in congenic/transgenic mice. NOx levels were measured by fluorometric assay. Superoxide production in renal extracts was quantified by spectrophotometric assay based on rapid reduction of ferricytochrome c to ferrocytochrome c, as previously described (29). A and B, Expression of renal NOx (A) and renal superoxide (B) in tamoxifen-treated mice and vehicle-treated mice. Bars show the mean ± SD of 5 mice per group. ∗ = P < 0.05 versus vehicle-treated mice. C–E, Immunohistochemical staining with antinitrotyrosine antibody for the detection of NOx expression in the renal tubular cells of tamoxifen-treated (C), vehicle-treated (D), and sham-treated (E) B6.Sle1.Sle3 mice.

Activation of the Klk1 gene attenuates renal injury by suppressing inflammation in lupus nephritis.

Kidneys from 10-month-old Ksp/CreERT2/CAG-lacZflx/−- Klk1/Sle1.Sle3 mice treated with tamoxifen or vehicle were examined for kidney injury and inflammatory cell infiltration by staining with KIM-1 and IBA-1 antibodies and then were evaluated by light microscopy. In vehicle-treated mice, moderate to marked signal intensities of KIM-1 and IBA-1 were detected in the glomerular and tubular regions, indicating more severe renal damage in lupus nephritis accompanied by infiltration of inflammatory cells (macrophage/monocytes) (Figures 6A and C). In contrast, tamoxifen-treated mice exhibited significantly reduced renal injury and monocyte/macrophage infiltration in the glomeruli and interstitium of the kidneys (Figures 6B and D).

Figure 6.

Evaluation of kidney injury and inflammation in Ksp/CreERT2/CAG-lacZflx/−-Klk1/Sle1.Sle3 mice treated with tamoxifen or vehicle. A–D, Immunohistologic staining of kidney sections using kidney injury molecule 1 (KIM-1) antibodies (A and B) and ionized calcium–binding adapter molecule 1 (IBA-1) antibodies (C and D). Cryosections were first stained with anti–KIM-1 or anti–IBA-1 antibodies and then with Alex Fluor 488–labeled secondary antibodies. Reduced levels of KIM-1 and IBA-1 expression were observed in the renal tubular region of tamoxifen-treated mice compared with vehicle-treated mice. Images are representative of 5 mice per group. E–J, Levels of inflammatory cytokines as measured using a Luminex assay. Interleukin-10 (IL-10) expression (E) was increased, while the expression of IL-12 (F), tumor necrosis factor α (TNFα) (G), IL-1β (H), IL-1α (I), and IL-6 (J) was decreased in the kidneys of tamoxifen-treated mice compared with placebo-treated mice. Bars show the mean ± SD.

In addition, we measured the expression of inflammatory cytokines using a Luminex assay. The expression of IL-12, TNFα, and IL-1β was significantly down-regulated in the tamoxifen-treated mice compared with the vehicle-treated mice (P < 0.05) (Figures 6F–H). The expression of IL-1α and IL-6 was also reduced in tamoxifen-treated mice compared with vehicle-treated mice, although the differences were not statistically significant (Figures 6I and J). However, IL-10 expression was increased significantly in tamoxifen-treated mice compared with vehicle-treated mice (P < 0.05) (Figure 6E). Hence, another renoprotective mechanism through which kallikrein may be operating involves subduing the inflammatory milieu within the affected kidneys.


In the NZM2410 lupus-prone mouse strain, the interplay of 2 loci, Sle1 and Sle3, leads to the emergence of pathogenic autoantibodies and lupus. Importantly, a key genetic contribution from the Sle3 interval is the kallikrein gene (22, 23). The Sle3 interval in strains such as B6.Sle3, NZW, NZM2410, and (NZB × NZW)F1 harbor a Klk allele that encodes reduced kallikrein production (22). Strains bearing this allele (including NZW, 129/Sv, and DBA/1) exhibit increased anti-GBM–induced nephritis, suggesting that kallikrein may be renoprotective. Evidence for this has emerged from 2 different studies. In one study, adenoviral delivery of kallikrein ameliorated anti-GBM–induced nephritis (23). In another study, blocking this axis using bradykinin B2 receptor antagonists resulted in more severe anti-GBM–induced nephritis (22). Given that renal tubular cells constitute a major source of kallikreins (32–35), we reasoned that local elaboration of kallikreins may confer protection against nephritis. The present study examines whether the above-described model also holds true for spontaneous lupus nephritis. Whereas B6.Sle1.Sle3 mice (which produce less kallikrein because of genetic polymorphisms encoded within the Sle3 interval) exhibit signs of lupus nephritis, deliberate up-regulation of kallikreins within the renal tubules significantly dampened renal disease in this model, as evidenced by both clinical and histopathologic readouts.

These findings resonate well with previous reports that kallikreins may be renoprotective in other models of renal disease (34–45). Kallikreins have been shown to play protective roles against ischemic stroke–induced renal injury in rats by inhibiting apoptosis and inflammation (36–39). Similar protective effects have also been reported in salt-induced hypertensive glomerulosclerosis (40, 41) and gentamicin-induced nephrotoxicity (42, 43). Further evidence of renal protection has been observed in a mouse model of diabetic nephropathy. It has been reported that kallikreins protect against microalbuminuria in experimental type I diabetes mellitus, while a lack of both bradykinin B1 and B2 receptors enhances nephropathy in diabetic mice (44–46).

The mechanisms through which kallikreins may confer renoprotection have not been fully understood. Previous experiments demonstrated that the protection against tissue damage provided by kallikreins could be attributed to pleiotropic effects in inhibiting oxidative stress, apoptosis, inflammation, and fibrosis (47). Several studies have shown that the kallikrein–kinin system can increase nitric oxide (NO) production through stimulating endothelial cell NO synthase activity and can increase cAMP levels in kidney epithelial cells (48–51). In the kidney, proximal tubular epithelial cells are high oxygen–using cells that are vulnerable to oxidative stress. NO and cAMP decrease mitochondrial oxidative metabolism by inhibiting cytochrome c oxidase and activating NADH-ubiquinone oxidoreductase, which in turn suppresses the level of oxidative stress (52–55). The suppression of oxidative stress could also lead to reduced inflammatory cell recruitment and less apoptosis in renal tubular cells. In vitro experiments with cultured renal tubular cells also showed that kallikreins and kinins suppressed H2O2-induced apoptosis and increased cell viability and Akt phosphorylation (39).

In our studies, we observed that kallikreins may be conferring renoprotection through at least 2 distinct pathways: the inhibition of inflammation and redox balancing. Kallikrein induction inhibited the expression of proinflammatory cytokine/chemokines such as IL-1, IL-6, and TNFα, the expression of which has been reported to be elevated in immune-mediated nephritis in both humans and mice (56, 57). In addition, kallikrein treatment in our model also suppressed the production of renal superoxide, with increased NO production. Superoxides (including H2O2) induce reactive oxygen species (ROS) formation, which can have deleterious effects on cells. ROS produced in renal cells have the potential to damage key cellular components including lipids, proteins, and DNA, and the balance of intracellular redox is a key determinant of cell survival, proliferation, differentiation, and apoptosis in nephritis (58). Kallikreins have previously been reported to be able to inhibit H2O2-induced ROS formation via NO production in cultured tubular cells (39).

With respect to the role of NOx in oxidative stress–induced renal damage, reports have been seemingly contradictory. For example, increased NO in lupus glomerulonephritis has been reported to be associated with increased oxidative stress and disease activity (59), as reviewed recently (60, 61). In contrast, NOx has been shown to play a protective role in oxidative-related organ damage in the heart and kidney (62). It is conceivable that these molecules may exert a multitude of effects, possibly being dependent on the specific cell types and molecular contexts. Our data indicated that tamoxifen-induced Klk1 expression is associated with increased NOx production, and this was corroborated by antinitrotyrosine staining (Figures 5C–E). Admittedly, it remains to be formally demonstrated that these resulting changes in the NO pathway are working to subdue disease, as we hypothesize, or are counteracting the protection conferred by kallikreins.

In conclusion, we observed renal kallikreins to be potent modulators of local pathologic reactions in lupus nephritis, a finding that resonates well with the previously reported beneficial roles of renal kallikreins in immune-mediated nephritis (22), diabetic nephropathy (44, 45), and nephritis caused by other triggers (47–51). At the mechanistic level, kallikreins may be modulating renal disease by inhibiting local inflammation, oxidative stress, and apoptosis. These genetic studies suggest that local delivery of kallikreins to the inflamed kidneys may be a viable treatment option for lupus nephritis. Experiments to directly test this hypothesis are therefore warranted.


All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Drs. Mohan and Q-Z Li had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Study conception and design. Shao, Raman, Wakeland, Igarashi, Mohan, Q-Z Li.

Acquisition of data. Shao, Yang, Yan, Y. Li, Du, Raman, Zhang, Q-Z Li.

Analysis and interpretation of data. Shao, Yang, Yan, Y. Li, Du, Mohan, Q-Z Li.


We acknowledge Drs. Jinchun Zhou and Ling Wang from the University of Texas–Southwestern Microarray Core Facility for assistance with the Luminex assays and quantitative PCR experiments.