The inheritance of glomerulosclerosis in mice is controlled by multiple ...

Nephrol Dial Transplant (1998) 13: 3074–3078

Nephrology Dialysis Transplantation

Original Article

The inheritance of glomerulosclerosis in mice is controlled by multiple quantitative trait loci Oliver Lenz1,3, Feng Zheng1, Jose Vilar2, Sophie Doublier1,*, Enrico Lupia1,*, Susanne Schwedler1,*, Liliane J. Striker1,* and Gary E. Striker3 1Renal Cell Biology Section, Metabolic Diseases Branch, NIDDK, NIH, Bethesda, MD, USA and 2INSERM U319, Paris, France, 3Ivax Research Institute, Miami, FL, USA

Abstract Background. Glomerulosclerosis, the common terminal event in chronic glomerular diseases such as diabetic nephropathy or IgA nephropathy, leads to end-stage renal disease. The considerable variation in both the risk of developing glomerulosclerosis and the rate of progression in individual patients suggest a role for genetic factors which have not been identified so far. In this study we sought to examine the mode of inheritance of glomerulosclerosis in mice. Methods. F1 animals of a mating between glomerulosclerosis-prone ROP-Os/+ male and non-sclerotic C3H female mice were backcrossed to the ROP strain. We took advantage of the radiation-induced mutation oligosyndactylism (Os) to identify glomerulosclerosis at the age of 3 months. Kidneys were perfused in situ with PBS/Formalin 10%. The extent of glomerulosclerotic lesions was evaluated on PAS stained paraffin sections using computer-aided morphometry. Results. F1 mice did not show any glomerulosclerosis. In the backcross offspring, we found a wide distribution of glomerular lesions between individual animals, ranging from normal to very severe. We calculated that at least 8–10 loci determine the severity of glomerulosclerosis in mice. Conclusions. Our data show that glomerulosclerosis is inherited in a recessive fashion involving at least 8–10 loci. Key words: genetics; glomerulosclerosis; mice; oligosyndactyly

progression is largely under genetic control. The considerable individual variation in the course of renal disease progression, has been studied extensively in diabetes mellitus, but is still poorly understood [1]. Ethnic background plays a role. For instance, African-Americans develop HIV-related glomerulosclerosis, whereas this form of rapidly progressive glomerulosclerosis is virtually non-existent in Caucasians [2]. Similarly, Pima Indians with non-insulin-dependent diabetes (NIDDM ) have a much higher incidence of diabetic nephropathy than Northern Europeans with NIDDM [3,4]. Although these epidemiological observations pinpoint the role of genetic factors, the mode of inheritance of glomerulosclerosis in man is still largely unknown. We have previously shown that ROP mice are prone to glomerulosclerosis. The disease process can be accelerated and uncovered by the introduction of the Os mutation [5]. This mutation leads to an inborn reduction in nephron number, glomerular hypertrophy, oligosyndactyly, and it is lethal in homozygous animals [6–9]. Since only sclerosis-prone mice carrying this mutation develop glomerulosclerosis [10], Os may be used as a tool to uncover susceptibility to glomerulosclerosis. In this study we demonstrate the inheritance of glomerulosclerosis by crossing sclerosis prone ROP mice with non-sclerotic C3H mice. We show that the susceptibility to develop glomerulosclerosis is under the control of multiple recessive modifiers. At least five loci, which may now be determined by linkage analysis, seem to be involved in the determination of the individual risk to develop glomerulosclerosis.

Introduction

Subjects and methods

Multiple lines of evidence suggest that the risk for a given individual to develop glomerulosclerosis and its rate of

Animals

Correspondence and offprint requests to: Gary E. Striker MD, Director, Ivax Research Institute, 4400 Biscayne Boulevard, Miami FL 33137, USA. *Current address: Renal Cell Biology Laboratory, Division of Nephrology, Department of Medicine, University of Miami School of Medicine, PO Box 016960 (R-126), Miami, FL 33101, USA.

Mice were obtained from The Jackson Laboratory, Bar Harbor, Maine. As progenitor strains we chose female C3H/HeJ (stock # 000659) and male ROP/Le-Os Es1b/+ Es1a (stock # JR0566). Male and female (C3H×ROP)-Os/+ F1 animals were backcrossed to the ROP strain to create 35 backcross offspring (BC ). Except for C3H mice, only 1

© 1998 European Renal Association–European Dialysis and Transplant Association

Inheritance of glomerulosclerosis in mice

animals carrying the Os mutation were used for morphometric analysis.

Histology All mice were sacrificed at the age of 90 days. The aorta was punctured distal and clamped proximal of the renal arteries. After broadly opening the vena cava, both kidneys were perfused in situ with 6 ml PBS and 3 ml 10% formalin, fixed in Carnoy’s solution for at least 12 h, and embedded in paraffin. Four-micrometre sections were stained with PAS. In addition, kidney and body weights were recorded.

Morphometry We chose a morphometric approach for the assessment of glomerulosclerosis. With this technique we could achieve unbiased and reproducible results, independent of which investigator was scoring the slides. For each animal digital images of twenty randomly chosen consecutive glomeruli were captured with a one chip colour CCD camera (Bunton Instruments, Rockville, MD) mounted on an Olympus light microscope (Bunton Instruments). The glomerular tuft was traced and the enclosed area was copied to create a new image. Both the total glomerular area (including nuclei and spaces within capillary loops) and mesangial area (excluding nuclei) were obtained using Adobe Photoshop (Adobe Systems, San Jose, CA) and NIH Image [11] (Figure 1). For each glomerulus, the ratio of mesangial area to total glomerular area was calculated. Ratios were averaged for each slide. The result (‘percent mesangial area’) was used to determine the severity of glomerulosclerosis in each animal. Glomerular size was determined by measuring 50 consecutive, randomly chosen glomeruli for each slide as described [10].

Statistical analysis The ratios mesangial area to total glomerular area (in percent) of all mice were placed into classes between zero and 40%. Classes were chosen to be in 2% increments and plotted against the relative number of animals in each class. Variances were compared using the one-way ANOVA and

Fig. 1. Morphometric assessment of glomerulosclerosis. For each animal, digital images of 20 randomly chosen glomeruli were captured (A). For each glomerulus, we traced total glomerular area (B) and mesangial extracellular matrix (C ) using Adobe Photoshop@. For each we calculated the number of pixels using NIH Image and computed the ratio mesangial area to total glomerular area in percent. The average mesangial area of 20 glomeruli was used as the score for the respective animal.

3075

Tukey’s multiple comparison test. Means were compared using the Student’s t-test, where appropriate with Welch’s correction for unequal variances. The number of modifiers was estimated using the formula derived by Wright [12] for an intercross analysis and known as ‘Wright’s polygene estimate’: (m −m )2 P2 F1 4(s2 −s2 ) BC1 E where n is an approximation of the minimum number of loci, m and m are the mean percent mesangial area in F1 BC1 the diseased parent strain (P2) and F hybrids (F1) respect1 ively, and s2 is the variance of the diseased parent strain. P2 We estimated s2 , which represents the variance due to E environmental influences, from the averages of the P1, P2, and F1 based on weights of 15152. n=

Results ROP-Os/+ mice exhibited severe sclerotic lesions at the age of 3 months. In contrast, C3H/HeJ mice did not have glomerulosclerosis at the age of 3 months or even 15 months ( Figure 2). We measured the severity of glomerulosclerosis in 10 animals from each of the two progenitor strains. We found that the distribution means of C3H and ROP-Os/+ mice were significantly different (P0.05). This finding suggested the absence of dominant alleles in ROP mice. To expose recessive modifiers, which cannot be identified in the F generation, we performed a backcross to the ROP 1 strain. The BC generation [(C3H×ROP)×ROP]1 Os/+ showed sclerotic lesions ranging from very mild to severe (Figure 3). A summary of all results is shown in Table 1. Using raw data, and estimating s2 from E C3H, ROP-Os/+ and F1 animals using weights of

Fig. 2. Glomerular histology in ROP-Os/+ and C3H mice. Representative micrographs of PAS-stained paraffin sections showing glomeruli of 3-month-old ROP-Os/+ mice, 3-month-old C3H mice, and 15-month-old C3H mice. Whereas ROP-Os/+ mice develop severe glomerulosclerosis by the age of 3 months (A), C3H do not show sclerotic lesions at 3 months (B) or 15 months (C ).

3076

O. Lenz et al.

Fig. 3. Glomerular histology in F and backcross progeny. 1 Representative micrographs of PAS-stained paraffin sections showing glomeruli of (ROP×C3H )Fi-Os/+ mice (A) and the backcross progeny (B,C ). Whereas F mice do not show sclerotic lesions, BC 1 1 animals display glomerulosclerosis of a wide array of severity, ranging from normal (B) to severe (C ).

15152, the formula of Wright yields 8.3 (n in Table 1). 1 It can be noted, however, that the variance seems to increase with the mean in the parent and F1 animals. Thus, we recalculated the number of loci using the data after a logarithmic transformation. The result is slightly higher with n =9.6 [12]. 2 We compared the ratio kidney weight to body weight, reflecting the number of nephrons, and the glomerular size in 10 animals of both the F and the 1 BC . Both F and BC animals carrying the Os muta1 1 1 tion showed a significant reduction in the ratio kidney weight to body weight (P