Mice

Adapting the protocol of Wu et al. (2010), 14-month-old female PAPP-A KO and wild-type (WT) mice were injected with streptozotocin i.p (50 μg/g per injection) freshly dissolved in 137 mM NaCitrate, pH 4.5, every 2–3 days for a total of 5–7 injections in order to induce stable hyperglycemia. Mice were placed in clean cages and fasted for four hours prior to each injection. Body weight and blood glucose levels were taken prior to the initial streptozotocin injection and weekly for four months following, also after a four hour fast. Mice were considered diabetic when their blood glucose level was sustained at ≥250 mg/dL. There was less than 10% mortality, and dead or euthanized animals were excluded from further study. Mice were not treated with insulin at any time during the study. After four months of stable diabetes, body and kidney weights were recorded, blood was collected, and right and left kidneys were prepared for fixation. All procedures were approved by Mayo Clinic’s Institutional Animal Care and Use Committee.

Renal histology

The right kidney was cut in cross section and the left kidney was cut longitudinally, rinsed in PBS and fixed overnight in Bouin’s solution. Fixed tissues were flushed for 10 minutes with tap water and stored at 4C in PBS until shipping to the Animal Histology Core at Mayo Clinic, Scottsdale, AZ for sectioning and staining with hematoxylin-eosin. Kidney ections were qualitatively assessed by an American College of Veterinary Pathologists board-certified veterinary pathologist (R.J.M). A blinded histopathology report was based on standard descriptions and terminology (see legend to Table 2), and given one of four subjective grades of minimal, mild, moderate and marked. Sections of kidney were also stained with periodic acid-Shiff (PAS) to measure mesangial area within each glomerulus using Adobe PhotoShop software (Adobe Systems, Inc.). At least 300 glomeruli from three mice were assessed per group.

Table 2

Diabetic kidney histology

Immunohistochemistry

De-paraffinized sections of mouse kidney were stained for macrophages using F4/80 as primary antibody. Slides were rinsed with Tris-buffered saline and 0.01% Tween-20 for 3 minutes, treated with Peroxidase Blocking Reagent (Dako, Carpinteria, CA) for 5 minutes, rinsed with Wash Buffer (Dako), and treated with Rodent Block M, (Biocare Medical, Concord, CA) for 30 minutes. Slides were then incubated with primary antibody (F4/80 CI:A3-1, rat monoclonal, Serotec) for 60 minutes at room temperature at a concentration of 1:200 [diluted in Background Reducing Diluent (Dako)], followed by incubation with secondary antibody (Rat on Mouse Probe, Biocare, prediluted) for 15 minutes at room temperature. Slides were rinsed 3 times in Wash Buffer, incubated with tertiary reagent (Rat on Mouse HRP, Biocare) for 15 min at room temperature, rinsed in Wash Buffer, and then stained for visualization with diaminobenzidine (Biocare) for 5 minutes. Counterstaining with hematoxylin, followed by dehydration in increasing concentrations of ethyl alcohol and xylene was performed prior to cover slipping.

Blood chemistries

Blood urea nitrogen (BUN) measurements were performed using a Urea Nitrogen assay kit purchased from Pointe Scientific Inc. (Canton, MI). Serum IGF-I was measured using a mouse IGF-I ELISA kit generously provided by ImmunoDiagnostic Systems (Fountain Hills, AZ)

General characteristics of diabetic and non-diabetic mice

There was a significant difference in body weight between WT and PAPP-A KO mice, as reported previously (Conover et al. 2004), but there was no significant difference in body weight in the diabetic mice (WT and PAPP-A KO) when compared to their respective non-diabetic controls (Table 1). There was a significant difference in the kidney weight/body weight ratio between diabetic WT and PAPP-A KO mice. However, this was not caused by the diabetic phenotype, since a similar difference was seen in nondiabetic mice. There was no significant difference in the harvest blood glucose level between WT and PAPP-A KO mice, although levels were significantly elevated in the diabetic mice compared to non-diabetic mice as expected from this model. There was no significant difference in blood urea nitrogen levels between diabetic WT and PAPP-A KO mice, although slightly elevated in diabetic compared to non-diabetic mice. These characteristics suggest changes in kidney function with four months of hyperglycemia, although urine was not collected for measurement of creatinine or albumin. Furthermore, serum IGF-I levels were not significantly different between diabetic WT and PAPP-A KO mice or between diabetic and non-diabetic WT mice, confirming previous characterization of PAPP-A KO mice as a model of autocrine/paracrine versus endocrine regulation of IGF action (Conover et al. 2004).

Table 1

General characteristics of diabetic and non-diabetic WT and PAPP-A KO mice

Renal histology

Diabetes induced by streptozotocin in aged mice results in glomerular, tubulointerstitial, and vascular kidney lesions with many features of human diabetic nephropathy (Wu et al. 2010, Fioretto and Mauer 2007). These are morphologically characterized by thickening of the Bowman’s capsule, glomerular hypertrophy with expansion of the extracellular matrix surrounding mesangial cells, and thickening of the basement membrane underlying proximal tubular epithelial cells.

WT diabetic mice developed thickening of Bowman’s capsule, tubular changes, and interstitial infiltrates (Fig. 2 and Table 2). The glomeruli were enlarged in 80% and the basement membrane of Bowman’s capsules was thickened in 100% of WT diabetic kidneys. Tubular dilation was noted in 80% of the kidneys and mononuclear infiltrates noted in 90%. Forty percent of the kidneys had additional abnormalities and all WT diabetic kidneys had two or more abnormalities. In contrast, diabetic PAPP-A KO kidneys showed little or no changes in glomerular size, Bowman’s capsule, and tubular dilation. However mononuclear cell infiltration was not different between diabetic WT and PAPP-A KO mice. To distinguish age-related versus hyperglycemia-induced effects on renal histology, we also examined 12 age-matched non-diabetic WT kidneys and saw little or no thickening of Bowman’s membrane or increase in glomerular size, although mononuclear cell infiltration and tubular dilation was similar to what was seen in diabetic WT kidneys (Table 2).

Figure 2

Renal histology of diabetic WT and PAPP-A KO mice

Expansion of mesangial matrix that is strongly PAS-positive is a hallmark pathological feature of diabetic nephropathy (Steffes 1989). Mesangial area was quantitated in PAS-stained diabetic kidneys and there was a significant shift to larger area in WT compared to PAPP-A KO mice (Fig. 3). Mean mesangial area in WT and PAPP-A KO was 3587 ± 79 μm² and 2518 ± 132 μm², respectively (P = 0.002). This was not just due to the overall larger body size of the WT mice because mesangial area in non-diabetic WT kidney was similar to what was seen in diabetic PAPP-A KO kidney (data not shown).

Figure 3

Mesangial area

Human diabetic nephropathy

It was strategic for the study of PAPP-A in diabetic nephropathy to ascertain where PAPP-A is expressed in the kidney and if there is site-specific regulation in diabetes. Although mouse kidney has been reported to have high levels of PAPP-A mRNA (Conover et al. 2004, Hourvitz et al. 2002), these studies were done in whole kidney and there were/are no antibodies available that recognize murine PAPP-A. Thus, our data are the first to demonstrate specific immunolocalization of PAPP-A to glomeruli. Furthermore, kidney of a patient with confirmed diabetic nephropathy had markedly elevated PAPP-A associated with the glomerulus as well as the surrounding Bowman’s capsule. This was of particular interest since the glomerulus is a major target of renal injury, especially in diabetes (Schlondorff 2008). It is not clear from immunohistochemistry which cells produce PAPP-A because it is a secreted protein that can associate with neighboring cells as well as the cells of origin (Conover et al 2007). In this way, PAPP-A can act as an autocrine/paracrine regulator of IGF action. A clinical study suggested prognostic value of circulating PAPP-A and increased all-cause mortality in type I diabetic patients with nephropathy (Astrup et al. 2007). However, it is not known whether this increase in circulating PAPP-A reflected changes in the kidney.

Mouse diabetic nephropathy

In order to investigate the physiological and pathophysiological role of PAPP-A in the kidney, we used WT and PAPP-A KO mice and induced stable hyperglycemia in these animals with a regimen of low dose streptozotocin injections, following the model developed by Wu et al. (Wu et al. 2010). In the original report, diabetes was induced in mice starting at 17 months, and these mice demonstrated severe nephropathy associated with elevated markers of oxidative stress, endoplasmic reticular stress, and macrophage-derived inflammation at 22 months (Wu et al. 2010). Induction of diabetes in younger mice (4-months-old) did not produce kidney lesions after a similar level and duration of hyperglycemia. We modified this model to start the steptozotocin injections in 14-month-old mice and assessed kidney histopathology after 4 months of stable hyperglycemia (approximately 18-months-old mice). We chose this younger age to minimize potential confounding effects of age-related cancers that become prominent, especially in WT mice, after 18 months (Conover et al. 2010). In our protocol, mice with visual signs of tumor at harvest would be removed from the study and there would be no further analyses; no animals were removed in this cohort due to tumors. Under this modified model, diabetic WT mice developed several features of nephropathy, i.e., thickening of Bowman’s capsule, and mesangial expansion, albeit not to the level of severity seen in the study by Wu et al. (Wu et al. 2010). They also had increased macrophage immunostaining especially around glomeruli. These were significantly reduced in diabetic PAPP-A KO mice. To confirm diabetes-induced versus age-related nephropathy, we also analyzed kidneys of non-diabetic WT and PAPP-A KO mice. There was no evidence of thickened Bowman’s capsule, increased glomerular size, or mesangial expansion. However, tubule dilation and mononuclear cell infiltration occurred to a similar extent. Inflammation and oxidative stress are important contributors to kidney aging. Diabetes accelerates aging phenotype in the kidney because of the additional hypeglycemia-induced stress (Wu et al. 2010). PAPP-A KO mice show enhanced resistance to inflammatory and oxidative stress (Conover 2010), which may explain, in part, the decrease in age-related kidney tubule dilation and relative resistance to the development of diabetic nephropathy

Jakob H. Mikkelsen: Grant, Ansh Labs; Patents, selective exosite inhibition of PAPP-A activity against IGFBP-4.

Claus Oxvig: Grant, Ansh Labs, Danish Research Council; Patents, Selective exosite inhibition of PAPP-A activity against IGFBP-4.

Cheryl A. Conover: Grants, NIH AG028141, HL074871, Ted Nash Long Life Foundation, Ansh Labs; Patents, Marker for Inflammatory Conditions (issued), Insulin-Like Growth Factor Binding Protein-04 Protease (issued), Transgenic mouse lacking PAPP-A Activity (issued), Treatment of Osteoporosis (issued); Royalties, Marker for Inflammatory Conditions.

FUNDING

This work was funded in part by a sponsored research grant from Ansh Labs (to C.A.C.).

The authors acknowledge the technical assistance of Laurie Bale, Jacquelyn Grell, Suban Chakraborty, Sally West and Anthony Croatt, all from the Mayo Clinic, Rochester, Minnesota.