Research Letter: Is the 129S1/SvImJ Mouse Strain More Suitable to Study Anticoagulant-Related Nephropathy Than the C57BL/6 Strain?

Abstract

Background:

We have previously demonstrated that excessive anticoagulation with warfarin or dabigatran may result in acute kidney injury with red blood cell (RBC) tubular casts in some patients with chronic kidney disease, and this condition was named anticoagulant-related nephropathy (ARN). 5/6 nephrectomy (5/6NE) rats treated with warfarin or dabigatran reproduce the main pathologic features of human ARN. We had reported that 5/6NE C57BL/6 mice only partially develop ARN with increased serum creatinine and hematuria but no RBC tubular casts in the kidney.

Objectives:

The aim of this study was to investigate whether ARN can develop in 5/6NE 129S1/SvImJ mice.

Methods:

5/6NE was performed in 129S1/SvImJ mice. Three weeks after 5/6NE, mice were treated with warfarin (1.0 and 1.5 mg/kg/day) or vehicle for 7 days. Serum creatinine, hematuria, and prothrombin time (PT) were monitored daily. Renal morphology was evaluated at the end of the studies.

Results:

Treatment with warfarin resulted in PT elevation 2 to 3 folds from baseline (1.0 mg/kg/day warfarin) and 4 to 5 folds from baseline (1.5 mg/kg/day warfarin) by day 7. Serum creatinine and hematuria elevated by day 7 in a dose-dependent manner. Histologically, 2 of 8 (25%) 5/6NE mice had RBCs in the tubules, and there was acute tubular epithelial cell injury in all warfarin-treated 5/6NE 129S1/SvImJ mice.

Conclusions:

Our findings suggest that 129S1/SvImJ mouse strain is a more suitable murine model to study ARN than C57BL/6 mouse strain.

Introduction

We had previously reported that excessive anticoagulation with warfarin may result in acute kidney injury (AKI) in some patients with chronic kidney disease (CKD). ¹ Later, clinical case reports and experimental data showed that such AKI may be associated not only with warfarin but also with other anticoagulants.^2,3 This AKI associated with anticoagulation was named as anticoagulant-related nephropathy (ARN). We demonstrated that when 5/6 nephrectomy (5/6NE) rats were treated with excessive doses of warfarin or dabigatran, these animals developed ARN. Morphologically, kidneys had glomerular hemorrhage, red blood cell (RBC) tubular casts, and acute tubular necrosis.^{3 -5} Unfortunately, genetic manipulations in rats are not easy to perform; therefore, a murine model of ARN would provide better tools to study the pathogenesis of ARN. We had reported that 5/6NE C57BL/6 mice only partially develop ARN. Although they show an increase in serum creatinine and hematuria when treated with warfarin or dabigatran, there were no RBC tubular casts in the kidney. ⁶ It has been reported that 5/6NE 129S1/SvImJ mice are more susceptible to the development of CKD than 5/6NE C57BL/6 mice.^7,8 Here, we test the hypothesis whether ARN develops in 5/6NE 129S1/SvImJ mice.

Methods

129S1/SvImJ mice were obtained from the Jackson laboratory (Bar Harbor, Maine). 5/6NE was performed in 3- to 4-month-old (20-25 g) male mice as we previously described. ⁶ Animals were anesthetized with isoflurane/oxygen (1:5), and a middle laparotomy was performed. After the right nephrectomy, two-thirds of the left kidney was surgically resected. The animals were kept on the standard rodent diet with free access to water.

Animals were treated with daily warfarin (1.5 mg/kg/day, n = 8; 1.0 mg/kg/day, n = 8) or vehicle (control group, water, n = 5) via oral gavage for a total of 7 days starting 3 weeks after 5/6NE. Blood (from the submandibular vein, 100 µL/time, using a 23-gauge needle) and urine (spontaneous urine) samples were collected daily. Urine and blood samples were analyzed immediately. Animals were sacrificed at day 7 of the treatment. The histology of the kidney was evaluated on 2- to 3-µm sections of paraffin-embedded tissue stained with hematoxylin and eosin (H&E) by a renal pathologist (S.V.B.) who was blinded to the treatment group.

Serum creatinine was measured based on the Jaffe reaction using a creatinine reagent assay (Pointe scientific, Inc., Canton, Michigan) as we previously described.^5,6

Hematuria was evaluated by dipsticks (Siemens Reagent Strips; Siemens, Tarrytown, New York) and expressed in a semiquantitative scale from 0 to 3, where 0 is absent, 1 is mild, 2 is moderate, and 3 is severe.

Prothrombin time (PT) was measured using the Biobase Coagulation Analyzer (model COA01; Genprice Inc., San Jose, California) based on the manufacturer’s protocol as we reported earlier.^5,6

Statistical Analysis

Descriptive statistics were used to analyze differences between experimental groups. Data are present as mean ± standard deviation, unless otherwise specified. Student 2-tailed t test was used to analyze differences between 2 different groups at the same time point; 2-way analysis of variance (ANOVA) was used to analyze differences between experimental groups.

Results

Treatment with 1.0 mg/kg/day warfarin resulted in 12.5% mortality by day 7 of the studies, whereas mice treated with 1.5 mg/kg/day warfarin had 37.5% mortality by day 7.

Treatment with warfarin resulted in a dose-dependent increase in PT (calculated as surrogate international normalized ratio [sINR]). Thus, warfarin 1.0 mg/kg/day (n = 7) resulted in a 2-fold increase in PT by day 3, and the sINR was in the range of 2.0 to 2.5 (Figure 1A), similar to the therapeutic INR range in humans. Warfarin 1.5 mg/kg/day (n = 5) resulted in a rapid increase in PT to 4 times from the baseline, and the sINR was in the range of 3.5 to 4.0 (Figure 1A), emulating super-therapeutic INR in humans. Two-way ANOVA showed significant changes in sINR after treatment (P < .0001). Serum creatinine was increased in 5/6NE mice treated with both 1.0 and 1.5 mg/kg/day warfarin (Figure 1B) in a dose-dependent manner. Thus, in 5/6NE mice treated with 1.0 mg/kg/day warfarin, serum creatinine at day 7 of treatment was 0.77 ± 0.009 mg/mL (control 0.59 ± 0.01 mg/dL, P < .0001), and in 5/6NE mice treated with 1.5 mg/kg/day warfarin, serum creatinine at day 7 was 0.92 ± 0.02 mg/dL (control 0.59 ± 0.01 mg/dl, P < .0001) (Figure 1B). Two-way ANOVA showed significant changes in serum creatinine after treatment (P < .0001)

Figure 1.

Effects of warfarin treatment on kidney function in 129S1/SvImJ mice. (A) Coagulation effects of warfarin treatment in 129S1/SvImJ mice. sINR was calculated as PT increase from control (5/6 NE without treatment) (as described in Methods). (B) Changes in serum creatinine in 129S1/SvImJ mice treated with warfarin. (C) Effects of warfarin treatment on hematuria in 129S1/SvImJ mice. (D) Representative image of red blood cell casts (arrow) in the tubules in a 129S1/SvImJ mouse treated with 1.5 mg/kg/day warfarin. Hematoxylin and eosin (H&E), magnification 200×. (E) Unremarkable renal cortex in a control mouse. H&E, magnification 200×.

Warfarin treatment was associated with an increase in hematuria. While 5/6NE mice were treated with 1.0 mg/kg/day warfarin, hematuria did not significantly increase from control (Figure 1C). Treatment with warfarin 1.5 mg/kg/day resulted in a significant increase in hematuria by day 4, and this elevated hematuria was observed until the end of the study (Figure 1C). Two-way ANOVA showed significant changes in hematuria after treatment (P < .0001)

Histologically, animals treated with warfarin had signs of acute tubular epithelial cell injury. Although the majority of treated animals did not have RBC casts in the kidney (despite increased hematuria), there were 2 mice (25%) treated with 1.5 mg/kg/day warfarin that had RBC casts in the kidney (Figure 1D). The RBC casts were focal and observed in less than 5% of the tubules. The control mice had an unremarkable renal cortex (Figure 1E).

Discussion

Anticoagulant-related nephropathy is a well-recognized entity clinically and pathologically. Numerous case reports describe AKI in patients on different anticoagulants and with different coagulopathies. ² The pathogenesis of this condition is not well understood; it has been demonstrated that oxidative stress plays a role in the development of acute tubular necrosis but not glomerular hemorrhage. ⁹ Although an animal model that mimics this condition was developed in rats,^{3 -5} development of a murine model will provide better tools to study the role of each individual gene or protein using knockout mice. C57BL/6 mice are widely used to create different knockout mice. In prior studies, we demonstrated that ARN only partially develops in C57BL/6 mice, when there are AKI and hematuria but no RBC tubular casts. ⁶

It has been reported that different mouse strains react differently to kidney injury. For instance, 129Sv mice have markedly increased proteinuria after 5/6NE compared with C57BL/6 mice.^7,8 Morphologically, glomerulosclerosis, interstitial fibrosis and tubular atrophy, perivascular fibrosis, and vascular thickening are more prominent in 129S1/SvImJ mice compared with C57BL/6 mice 12 weeks after 5/6NE.^7,8 In our study, morphology in 5/6NE 129S1/SvImJ mice was similar to that in 5/6NE C57BL/6 mice 3 to 4 weeks after 5/6NE, which corresponds with earlier observations. ⁷

Our warfarin treatment increased PT (measured as sINR) to the levels that are similar to human therapeutic protocols. Thus, treatment with 1.0 mg/kg warfarin resulted in a 2- to 3-fold increase in PT compared with vehicle-treated 129S1/SvImJ mice, which is similar to the therapeutic range in humans. When 1.5 mg/kg/day warfarin was used, the PT increased 3 to 4 folds, similar to the INR levels in patients who developed ARN. ¹ Of note, the warfarin dose that required PT increase to the level in 129S1/SvImJ mice was lower than the warfarin dose in C57BL/6 mice, where 2.5 and 5.0 mg/kg/day of warfarin had to be used to achieve similar increase in PT. ⁶ This could be explained by different sensitivity of these 2 mouse strains to warfarin. Although there are no literature data on how warfarin works in these mouse strains, differences in coagulation parameters, such as time to first clot formation, between these mouse strains have been described. ¹⁰

Our data suggest that glomerular hematuria develops in warfarin-treated mice regardless of the strain. Similar to what has been demonstrated in a rat model, serum creatinine also increased in both mouse strains, indicating injury to the tubulointerstitium, probably via oxidative stress. ⁹ Interestingly, glomerular hematuria in rats leads to tubular RBC casts in 5/6NE animals, but in sham-operated rats and in 5/6NE C57BL/6 mice treated with anticoagulants, tubular RBC casts did not develop.^{4 -6} The mechanisms of this phenomenon are not clear, but they are probably related to the tubules (either proximal or distal segment) rather than to the glomerular filtration barrier injury because hematuria was observed in all experimental animals treated with different anticoagulants.^{3 -6} In contrast, in 25% of 5/6NE 129S1/SvImJ 5/6 mice there were tubular RBC casts, similar to ARN in humans and rats.^1,2,4

Our data indicate that 5/6NE 129S1/SvImJ mice can reproduce all aspects of human ARN, including AKI, hematuria, and RBC tubular casts (in about 25% of mice). A limitation of our studies is that only warfarin was tested, but based on the previous data on humans and animals, all classes of anticoagulants result in similar morphologic changes in the kidney regardless of the anticoagulants used.^2,3,5,6

Conclusions

We demonstrated that the development of ARN differs among mouse strains. Whereas 5/6NE C57BL/6 mice treated with anticoagulants did not develop RBC casts in the kidney, such RBC casts were seen in a subset of 5/6NE 129S1/SvImJ mice treated with warfarin. Because RBC casts are seen not in all but only in about 25% of 5/6NE 129S1/SvImJ mice, the 129S1/SvImJ mice are better but not the ideal model to study ARN. Therefore, further investigations on other mouse strains are warranted.