Renal Histopathology of a Baboon Model with Type-2 Diabetes

Abstract

Naturally occurring type-2 diabetes has been found in a colony of baboons. Ongoing characterization of the baboon colony maintained at the Southwest National Primate Research Center has revealed a significant range of glucose sensitivity with some animals clearly diabetic. Seven baboons, 4 with diabetes and 3 without diabetes underwent histopathological investigation. Three diabetic animals were diagnosed using fasting blood glucose, hemoglobin A1C and intravenous glucose tolerance test, and a fourth one was known to have hyperglycemia. One control baboon and 3 baboons with diabetes had microalbuminuria. On kidney biopsy, diabetic baboons had thickening of the glomerular basement membrane and mesangial matrix expansion compared to controls. Immunohistochemistry showed the diabetic animals had increased mesangial expression of cellular fibronectin ED-A. Two diabetic animals with microalbuminuria had evidence of mesangiolysis with formation of an early nodule. One diabetic animal had a Kimmestiel-Wilson nodule. We conclude that the baboon represents a useful primate model of diabetes and nephropathy that resembles the nephropathy associated with type-2 diabetes in humans.

Introduction

Diabetic nephropathy is one of the major complications of diabetes mellitus and is the leading cause of chronic renal failure in man. The early glomerular morphological features of diabetic nephropathy include thickening of the glomerular basement membrane, and mesangial matrix expansion, which may progress to glomerulosclerosis (Steffes et al., 1989).

Current animal models of type-1 and type-2 diabetes have limitations in that the histological changes in the kidney differ from those in humans. Although more severe renal pathology in these models can be induced by high fat diet, superimposition of obesity or hypertension, there remains a need for a spontaneous model of diabetic nephropathy that exhibits histological changes that resemble more closely those found in humans (Comuzzie et al., 2003; Heffernan et al., 1995; Stout et al., 1986). Baboons are valuable models for the study of complex diseases and physiological processes expressing a pattern of susceptibility and complications very similar to that seen in humans (Rogers and Hixson, 1997). Some primates, including baboons, have been reported to develop spontaneous diabetes and diabetic nephropathy (Rosenblum and Coulston, 1983; Stout et al., 1986; Weber and Greeff, 1973). As in humans, development of diabetes in baboons appears to be associated with increased fat accumulation and age (Chavez et al., 2008; Higgins et al, 2010).

Although the glomerular histology of baboons with type-1 diabetes resembles that found in humans with type-1 diabetes (Heffernan et al. 1996; Stout et al., 1986), there are no studies that have characterized the histological changes of diabetic nephropathy in baboons with type-2 diabetes. In this study we describe the glomerular morphology of baboons with type-2 diabetes. On random screening of a pedigreed baboon colony at the Southwest Regional Primate Research Center in San Antonio, we noted that some animals have developed spontaneous type-2 diabetes. To further characterize this model, we screened approximately 478 baboons for the presence of diabetes based on blood glucose. We performed biopsies on a subset of seven baboons, matched for age, weight and sex, in order to characterize the renal disease and glomerular histopatholology in these animals diagnosed with type-2 diabetes.

Materials and Methods

Animals

The pedigreed baboon colony has between 3,000 and 4,000 baboons at any given time, product of a mixture of two subspecies: Papio hamadryas anubis and Papio hamadryas cynocephalus. All animals in this study are housed in large outdoor enclosures, and are provided ad libitum access to a low fat standard monkey chow diet (Harlan Teklad) and fresh water. This population consists of large 2–4 generation, non-inbred pedigrees ranging in size from 6 to 213 members. The animals involved in the study were used in an appropriate and humane manner, in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals, and the animal protocols were reviewed and approved by the Institutional Animal Care and Use Committee.

We first performed a population assessment of diabetic status by screening for fasting blood glucose in 478 baboons randomly chosen out of the entire colony: 329 females and 149 males, with a mean age of 13.2 years. The baboon colony has a much larger number of females than males, but there was no difference in the incidence of diabetes nor in the mean fasting blood glucose with regards to gender. 48% of the males and 43% of the females had diabetes based on a blood glucose value surpassing 126 mg/dl. The mean fasting blood glucose ± Standard Error was 161.9 ± 4.44 mg/dL for males and 165.3 ± 6.23 mg/dL for females, p=0.6. We then selected at random a sub sample of 7 age- and weight matched female baboons from the group that underwent fasting blood glucose screening, for use in an intensive phenotyping protocol: 3 animals without diabetes that served as controls and 4 animals with spontaneous type-2 diabetes. Six of these animals underwent intravenous glucose tolerance test.

Tether System

Animals were housed in single cages in the veterinary clinic, and placed on a tether system consisting of venous and arterial catheters, a backpack, and a fiberglass box which accommodated catheters, blood pressure transducers and electrical junctions. Two superficial catheters were inserted under general anesthesia: an intravenous catheter in the right internal jugular vein, and an intra-arterial catheter in the right profundal femoris artery. Animals were immobilized with ketamine hydrochloride 10 mg/kg of body weight intramuscularly (IM) and diazepam 0.25 mg/kg intravenously (IV). The catheters were tunneled subcutaneously (SC) to the back, where they exited the skin and connected to a pump inside the backpack. The backpack was attached to a jacket that was comfortably worn by the animals. A flexible shaft connected to a swivel mechanism carried infusion lines that connected to the pump, and cables that connected to a transducer. The tether system allowed for frequent blood sampling, as well as for continuous blood pressure monitoring.

Hemodynamic Parameters

Six animals, three control and three diabetic, had continuous 7-day intra-arterial blood pressure monitoring. Computerized data of daily blood pressure 1-second observations were collected for 23 hours per day, and averaged for each day of observation. Hypertension was defined as systolic blood pressure (SBP) ≥ 140 mmHg, and/or diastolic blood pressure (DBP) 90 mmHg. For our studies lights on occurs at 06:00 hrs, lights off occurs at 18:00 hrs and feeding occurs at 13:00–14:00 hrs. The daytime maximum blood pressure for each day and for each baboon was calculated by averaging all blood pressure observations from 12:00–18:00 hrs and the nighttime minimum blood pressure was calculated by averaging all blood pressure observations from 00:00 –06:00 hrs. The nighttime decrease in blood pressure for each 24-hour period in each baboon was calculated by subtracting the average daytime maximum blood pressure from the average nighttime minimum blood pressure.

Blood and Urine Tests

All the laboratory tests were analyzed in a quality controlled clinical pathology laboratory. Fasting blood glucose was measured using enzymatic reference method with hexokinase; animals with a fasting blood glucose ≥ 126 mg/dl were considered to have diabetes mellitus. After an overnight fast, intravenous glucose tolerance test was performed in 6 baboons: 3 control and 3 with diabetes. Fasting blood glucose was measured at 6, 4, 2 and 0 minutes prior to the glucose load. An intravenous bolus of 50% dextrose (0.5 mg/kg of body weight) was administered over 30 seconds and blood glucose levels were measured at 2, 4, 6, 8, 10, 15, 20, 30, 40, 50, and 60 minutes after the bolus. All blood samples were immediately placed in iced heparinized tubes, centrifuged and plasma was sent for glucose analysis. Glucose disappearance rate was calculated using the period between 5- and 20-minute time points, as previously described (Bodkin and Hansen, 1995).

Serum creatinine was measured by the buffered kinetic Jaffe reaction without deproteination; hemoglobin A-1C using a competitive immunoassay (Micromat II from Bio-Rad); and C-peptide using a competitive immunoassay (Immulite, DPC).

Urine was collected under general anesthesia using a suprapubic needle aspiration. Urine glucose was measured by dipstick. Urine albumin was measured by immunoturbidimetric method. Urine creatinine was measured by buffered kinetic Jaffe reaction without deproteination. Urine albumin excretion was expressed as urine albumin (mg/dl)/creatinine (mg/dl) ratio. These results are based on the analysis of one urine collection, using microalbuminuria thresholds established for humans. Microalbuminuria was defined as urinary albumin/creatinine (Ur A/C) > 0.03.

Renal Biopsy

An open kidney biopsy was performed in the 7 baboons under general anesthesia, using a retroperitoneal approach. Animals were immobilized with ketamine hydrochloride 10 mg/kg of body weight IM and diazepam 0.25 mg/kg IV. An endotracheal tube was inserted and inhalation anesthesia was given with isofluorane 1.5% volume/volume. After the procedure, the animals were allowed to recover on the operating room table. Analgesia was provided with buprenorphine hydrochloride 0.15 mg/kg of body weight SC, twice a day. After the biopsy, amoxicillin 25 mg/kg of body weight was given IV for 7 days post-procedure. Baboons were kept under close observation in the veterinary clinic.

Tissue samples were divided into 4 portions: one was immediately frozen in OCT compound for cryosection, another part was fixed in formalin for routine histological and immunohistochemical microscopy; an additional part was fixed in Michel’s solution for immunohistochemistry after paraffin embedding; and the last portion was fixed in glutaraldehyde/formaldehyde cocktail for electron microscopy as described in detail below. Tissue was analyzed for diabetic glomerular disease characterized by the presence of diffuse and/or nodular increases in mesangial matrix, fibronectin expression, thickening of the glomerular basement membrane, and/or arteriolar hyalinization. In addition, tissue was examined for mesangial immunoglobulin or paraprotein deposits by immunofluorescence microscopy, amyloid deposits by Congo red staining or electron microscopy, and electron dense deposits within the GBM or glomerular capillary subendothelial space. Changes in glomerular size, basement membrane abnormalities, mesangial matrix expansion, and development of fibrosis were assessed by computer-assisted image analysis.

Routine Histology

Three-micron-thick tissue sections were mounted on slides and studied by light microscopy with hematoxylin-eosin (H&E), Masson’s trichrome and periodic acid Schiff-silver methenamine (PAS/Jones) using routine staining techniques.

Immunohistochemistry

Immunofluorescence and immunoperoxidase were performed using methods previously described (Barnes et al., 1999). Direct immunofluorescence was used to examine glomerular localization of endogenous IgG, IgA, IgM, C3, C1q, fibrinogen and albumin. Frozen tissue sections 4-μm-thick were mounted on slides and allowed to air-dry for 10 minutes. The slides were fixed in 95% ethanol for 30 minutes, and rinsed twice in TBS for 5 minute intervals. The mounted sections were incubated for 1 hour with FITC-conjugated polyclonal rabbit anti-human IgG, IgA, C3, C1q, fibrinogen, and albumin (DAKO, Carpinteria, CA), diluted 1:10 in 1% BSA in TBS. Excess antiserum was washed away by rinsing with three changes of TBS at 10 minute intervals. The slides were mounted with an aqueous mounting medium and examined using an Olympus AX70 Research microscope (Melville, NY) equipped for epifluorescence and filter modules optimized for fluorescein, and Cy3.

Immunoperoxidase histochemistry was used to examine the degree of glomerular mesangial matrix expansion assessed by the expression of laminin and the ED-A domain of alternatively spliced cellular fibronectin, a subtype of fibronectin that is expressed exclusively in embryonic tissue and in adult tissue in the presence of disease. After air drying, 6 μm sections were blocked with donkey IgG, followed by biotin/avidin kit for endogenous biotin_(Vector Laboratories), and with 0.6% hydrogen peroxide in methanol for endogenous peroxidase. The sections were incubated for 30 minutes with mouse monoclonal anti-fibronectin (IST9) antibody (Harlan, Indianapolis, IN) or mouse monoclonal anti-laminin antibody (Chemicon, Temecula, CA), 10 μg/ml) in PBS containing 0.1 % bovine serum albumin (BSA). Control sections were incubated with mouse IgG, 10 μg/ml in PBS, (Chemicon, Temecula, CA), in place of the primary antibody. All sections were incubated for 30 minutes with biotinylated donkey anti-mouse IgG (Chemicon, Temecula, CA) diluted 1:150 as the secondary antibody. Immunoperoxidase detection was performed using the avidin-biotin complex (ABC) method (Vector Laboratories, Burlingame, CA) using 3, 3′-diaminobenzidine tetra hydrochloride (DAB, Zymed Laboratories, San Francisco, CA)as substrate.

Buffered Congo Red for Amyloid

Six-μm-thick sections of paraffin-embedded kidney tissue were deparaffinized in xylene, hydrated and rinsed in distilled water, and stained in 0.5% buffered Congo Red solution in Sorenson-Walbum Buffer, pH 10.0, for 20 minutes. After incubation, slides were differentiated in 70% ethanol, dehydrated quickly in 95% and absolute ethanol, rinsed in a mixture of absolute ethanol and xylene, cleared in several changes of xylene and mounted in Permount.

Image Analysis

Image analysis was performed by 2 researchers who were blinded with regards to the control or diabetes status of the animals. Kidney sections stained with Jones silver and fibronectin were used to quantitate mesangial matrix expansion. Digital images of 20 randomly selected glomeruli in each experiment were taken using a 20× objective. The extent of silver stain or DAB reaction product was measured in the images using the segmentation tool of Image-Pro Plus (Media Cybernetics, Silverspring, MD). That portion of the image showing staining product was extracted by selecting a lower and upper range of gray scale within the limits of background and the highest intensity of staining. The image data were then masked and pseudo-colored for measurement of total staining area relative to the area of interest, set to encompass the glomerular capillary tuft. Renal disease parameters that were measured included overall glomerular tuft area, area of silver staining and the accumulation of extra-cellular matrix fibronectin. All image measurements were calibrated to a stage micrometer.

Electron Microscopy

Electron microscopy of the glomerular capillaries was performed with a Philips EM208S electron microscope, using a 14,000× magnification with internal image calibration.

Minced tissue obtained from kidney cortex was fixed overnight in cold 4% para-formaldehyde, 1% glutaraldehyde in phosphate buffer then embedded in epon 812 resin by routine laboratory methods. 0.5 μm plastic sections were cut and stained with toluidine blue for identification of representative areas and subsequent ultramicrotomy. Ultrathin sections (70–90 nm) were stained with uranyl acetate and lead citrate. Digital images were taken of capillary walls in which the profiles of glomerular epithelium, basement membrane and endothelium were sectioned at a cross section perpendicular to the capillary axis to avoid tangential cuts. Captured images of two to three glomeruli were examined per animal and analysis of GBM thickness was performed in seven random capillaries of each specimen using the linear measurement tool of Image-Pro Plus. In each calibrated image, nine to fourteen evenly spaced linear measurements of GBM thickness spanning from epithelial to endothelial cell membranes were taken. Results are expressed in nanometer lengths.

Statistical Analysis

Glomerular tuft area, basement membrane thickness, glomerular area of silver staining, and area of mesangial matrix accumulation (fibronectin ED-A staining) per glomerulus were calculated and expressed as means for each baboon. Group means were presented with the SD as the index of dispersion. Because of limited sample size, the results were expressed with the medians (interquartile ranges) and mean ± SD, appropriately. Statistical probability for each measure was examined by Student’s t test. P-values were two-sided and those < 0.05 were indicated statistically significant. All statistical analyses were performed with Statistical Analysis System (SAS) version 9.3.

Results

In the preliminary fasting blood glucose screening of 478 baboons, 44 % had glucose level ≥ 126 mg/dl (Figure 1). The following are the results of the sub-set of 7 baboons for more detailed analysis.

Figure 1.

Distribution of fasting blood glucose in 478 adult baboons, on a standard monkey chow. 210 animals had fasting blood glucose ≥ 126 mg/dl (dark bars).

Hemodynamic Parameters (Table 1)

Table 1.

Hemodynamic Data

Animals	Non-Diabetic Baboon			Diabetic Baboon
	1	2	3	1	2	3
Age (years)	20	19	18	24	20	19
Weight (kg)	14	20	14	15	14	22
SBP	143 ± 9	123 ± 11	119 ± 9	139 ± 13	118 ± 8	146 ± 14
DBP	83 ± 7	73 ± 8	77 ± 7	85 ± 7	74 ± 6	98 ± 12
MAP	107 ± 7	93 ± 8	95 ± 7	106 ± 8	92 ± 6	117 ± 11

SBP: systolic blood pressure. DBP: diastolic blood pressure. MAP: mean arterial pressure.

SBP, DBP and MAP are stated as mean ± SD.

All hemodynamic data between Non-Diabetic and Diabetic Animals had p-values >0.2.

Two animals had hypertension (HTN): one animal with diabetes had mean SBP (± SD) 146 ± 14/mean DBP (± SD) 98 ± 12 and one control animal had mean SBP (± SD) 143 ± 9/mean DBP (± SD) 83 ± 7. Overall, mean blood pressure was slightly higher in diabetic vs controls (105.6±7.1 vs 98.7±4.4, but did not reach statistical significance (P=0.45). Baboons exhibit a circadian variation in blood pressure as in man and other animal species. In our baboons, this daily blood pressure variation typically exhibits a sustained maximum increase in blood pressure during the 12:00–18:00 hrs time period and a sustained maximum decrease in blood pressure during the 00:00–06:00 hrs time period. The remaining portions of a typical 24 hr period represent times of progressive increases in blood pressure (06:00–12:00 hrs) and progressive decreases in blood pressure (00:00–06:00 hrs). For our studies, lights on occurs at 06:00 hrs, feeding occurs at 13:00–14:00 hrs and lights off occur at 18:00 hrs. These events most likely contribute to much of the daily variation in blood pressure in our baboons. The 3 control baboons exhibited a nighttime fall of systolic blood pressure of 7.4%, 7.7% and 7%, respectively. The control baboon with hypertension had a fall in systolic blood pressure similar to that seen in the 2 other control baboons. The 3 tethered diabetic baboons exhibited a nighttime fall of systolic blood pressure of 12%, 11.4% and 0.5%, respectively. The diabetic baboon with hypertension did not have a fall of systolic blood pressure, compared to the normotensive baboons with diabetes. Diabetic animals exhibited lower heart rate variability compared to controls, but the difference did not reach statistical significance. Blood and Urine Tests (Figure 2, Table 2)

Figure 2.

Results of intravenous glucose tolerance test in a sample of 6 adult baboons. Circles: control animals. Diamonds: diabetic animals. Control animals had normal fasting blood glucose and normal glucose disappearance rates 20 minutes after intravenous glucose bolus. This is contrasted with results for diabetic animals that exhibited fasting hyperglycemia, followed by higher glycemic peaks, slow glucose disappearance rates, and persistent hyperglycemia with a failure to return to baseline glycemic levels even at 60 minutes after intravenous glucose load.

Table 2.

Serum and Urine Data

Clinical chemistry	Non-Diabetic Baboon			Diabetic Baboon
	1	2	3	1	2	3	4*
Serum:
Glucose (mg/dl)	87.0	92.0	96.0	185.0	216.0	188.0	79.0
Creatinine (mg/dl)	0.8	0.7	0.7	0.6	0.6	0.7	0.6
Hgb A1C (%)	5.7	5.0	5.9	8.4	7.0	8.9	11.0
C-Peptide (ng/ml)	4.2	4.7	2.2	4.9	1.6	3.3	1.1
Urine:
Creatinine (mg/dl)	157.0	18.0	118.0	66.0	113.0	125.0	24.0
Albumin (mg/dl)	8.7	< 0.8	2.8	1.7	4.3	11.9	3.1
Albumin/Creatinine	0.055	< 0.03	0.024	0.026	0.038	0.095	0.129

^*Animal being treated with insulin.

The 3 control animals had mean fasting blood glucose of 91.7 mg/dl and normal glucose disappearance rates by 20 minutes post-glucose bolus on IGTT, representing normal glucose tolerance and insulin sensitivity. This is contrasted by the IGTT results for the 3 diabetic animals that exhibited fasting hyperglycemia, with mean fasting blood glucose 196 mg/dl (P<0.0005), followed by higher glycemic peaks, slow glucose disappearance rates, and a failure to return to baseline glycemic values even at 60 minutes after intravenous glucose load, confirming the diagnosis of diabetes determined by a fasting blood glucose 126 mg/dl. Also, serum levels of hemoglobin A1C in the diabetic baboons were above >6% generally considered diagnostic for humans (Diabetic = 8.1±0.57 vs. controls 5.5 ±0.27, P= 0.015). The remainder of clinical chemistries did not show statistical significance between the groups. The fourth diabetic animal had advanced type-2 diabetes and was provided insulin treatment (15 IU of NPH insulin SC once a day). All animals had normal serum creatinine and hemoglobin C-peptide.

The 4 diabetic animals had glycosuria by dipstick, compared to control animals. Microalbuminuria, defined as urinary albumin/creatinine (Ur A/C) > 0.03 was found in one control animal and three of the four diabetic animals, where mean ratios were 0.42±0.34 in diabetic baboons) vs. 0.036±0.0003 for controls. However, because of the variability, the values did not reach significance (P=0.28).

Histology

All diabetic animals displayed variable degrees of renal disease. Diabetic baboons had larger glomeruli, increased glomerular basement membrane thickness, and matrix expansion, relative to controls (Table 3, Figures 3–5). Also, there was increased expression of laminin in the glomerular capillary wall of diabetic animals relative to controls (Figure 4C, D). Image analysis of glomerular size revealed a 37% increase in glomerular tuft area averaging approximately 26,000 μm² in control vs 36,000 μm² in diabetic baboons when the animal receiving insulin was excluded (p-value = 0.04, Table 3, Figure 3,). Increased glomerular size was associated with an expanded matrix assessed by image analysis of PAS Jones (Figure 3) and fibronectin (Figure 4) staining, with increases of 50% and 45% in staining area respectively, however these values did not reach statistical significance). Glomerular basement membrane thickness measured in electron micrographs nearly doubled in diabetic animals showing an mean of 327 μm in controls vs 564 μm in diabetic animals (p= 0.029, Figure 5, Table 3). Focal areas of mesangiolysis and early nodule development were observed in two diabetic animals with microalbuminuria (Figure 3). One diabetic animal had a glomerulus with an area of marked increase in mesangial matrix consistent with beginning stages of glomerulosclerosis, arteriolar hyalinosis, and a capsular drop (Figures 5 and 7). Immunofluorescence microscopy in the diabetic animals showed 3+ staining for IgG and albumin in the capillary wall, Bowman’s capsule and tubular basement membrane and 2+ staining for IgM in both glomerular mesangium and capillary wall (Figure 6). Immunofluorescence staining for IgA, C3, and C1q was negative in all glomeruli examined. In the control animals, immunoflurescence showed mild staining of the capillary wall for IgG, but it was negative for IgA, IgM, C3 and C1q. There were no mesangial deposits by EM in either group was negative in all 7 animals, ruling out immune-mediated glomerular disease. Similarly, Congo red staining was negative in all animals indicating an absence of glomerular alterations attributable to amyloid deposition. There was increased thickening of the tubular basement membrane in diabetic baboons, compared to control baboons. There was no significant interstitial pathology in either group.

Table 3.

GBM thickness, glomerular tuft area, glomerular area of silver staining and glomerular area of fibronectin staining in control and type-2 diabetic baboons.

Glomerular Morphometric	Non-Diabetic Baboon n=3	Diabetic Baboon n=4	p-value
GBM thickness (nm)	269 (267, 445) 327 ± 102	530 (443, 600) 521 ± 103 564 ± 70*	0.056 0.029*
Tuft Area (μm2)	27,046 (21,146, 30,207) 26,133 ± 4,598	34784 (29,276, 37,850) 33,563 ±5,510 35,967 ± 3,297*	0.118 0.039*
Silver Staining (μm2)	4,675 (4,137, 6,045) 4,952 ± 984	6,973 (4,883, 9,965) 7,424 ± 3,824 7,730 ± 4,623*	0.334 0.366*
Fibronectin Staining (μm2)	10,382 (6,936, 11,342) 9,553 ± 2,316	12,005 (9,193, 15,713) 12,453 ± 4,448 14,058 ± 3,772*	0.357 0.153*

The values are stated medians (interquartile range) and mean ± standard deviation.

^*= N=6, excluding animal treated with insulin.

Figure 3.

Glomerular PAS/Jones staining in control (A) and diabetic (B) baboons. Diabetic baboons tended to have larger glomeruli than control animals and features of diabetic nephropathy characterized by mesangiolysis with early nodular formation affecting one third of the diabetic glomerulus (arrow). The amount of silver staining of glomerular matrix of the diabetic animal is also increased relative to control (see Table 3). Bar = 100 μm.

Figure 5.

Electron micrographs of capillary wall cross-sections in control (A) and diabetic (B) baboons. Thickness of the glomerular basement membrane (GBM) was measured perpendicular to the capillary wall at sites depicted by lines. There was significant increase of the GBM thickness in the diabetic baboons relative to controls. Bar = 500 nm. Other features of diabetic nephropathy observed by electron microscopy include marked mesangial matrix expansion (C) and a capsular drop (D, arrow). Bar = 2 μm.

Figure 4.

Immunoperoxidase detection of fibronectin ED-A (A, B) and laminin (C, D) in glomeruli from control (A, C) and diabetic (B, D) baboons. There is increased expression of fibronectin in the mesangium of the diabetic baboon (B) relative to control (A). Laminin expression is increased in the glomerular capillary walls of the diabetic baboon relative to control (C, D). Bar = 100 μm.

Figure 7.

H & E staining of a glomerulus from a 23 year -old female diabetic baboon weighing 18.9 kg. Arteriolar hyalinosis(arrow ) and capsular drop (arrowhead) are illustrated. Bar = 100 μm.

Figure 6.

Glomerular immunofluorescence study from a diabetic baboon. (A) IgG, (B) IgM, (C) albumin, and (D) C3. There was 3+ staining for IgG and albumin, and 2+ staining for IgM. Immunofluorescence for C3 was negative. Bar = 100 μm.

Discussion

In this communication, we have identified spontaneous type-2 diabetes in baboons that is associated with morphological changes in the kidney that resemble diabetic nephropathy in man. Kidney biopsy showed thickening of the glomerular and tubular basement membranes and expansion of mesangial matrix with increased deposition of fibronectin ED-A and laminin. In addition, mesangiolysis and glomerulosclerosis associated with nodular lesions typical of human diabetic nephropathy were observed in this baboon population

Old world primates have been reported to develop spontaneous type-2 DM (Bodkin, 2001). Because of their close phylogenetic relationship to humans, baboons not only share many biological, anatomical, biochemical and physiological characteristics, but also have many of the same genes influencing relevant phenotypes operating on a similar genetic background (Howard, 1982; Rogers and Hixson, 1997). The type-2 diabetic baboon could be a useful primate model of human diabetes and its complications. Indeed, baboons in the colony at Southwest Regional Primate Research Center show a wide range of insulin sensitivity and there is a strong inverse correlation between indexes of adiposity and insulin sensitivity (Chavez et al., 2008). Development of obesity and insulin resistance in baboons in this population could likely occur by a change from the wild to captivity (Chavez et al., 2008). Importantly, young adult male baboons on a high sugar high fat diet develop increased body fat and triglyceride concentrations, altered adipokine concentrations, and evidence of altered glucose metabolism such as elevated hemoglobin A1c levels (Higgins et al., 2010), demonstrating a clinically-relevant animal model of diet-induced metabolic dysregulation. Baboons have a long life span that would allow for longitudinal studies of diabetes mellitus and its complications.

In normotensive humans, the average nighttime blood pressure decreases by 10–20% (dippers) and this blood pressure variation pattern is usually also preserved in patients with hypertension (Marinakis et al., 2003). A variety of diurnal blood pressure patterns occur in patients in whom the nocturnal fall of blood pressure may be > 20% (extreme dippers), < 10% (non-dippers), or even reversed (reverse dippers) (Kario et al., 1996). There is evidence that non-dipper patients experience increased frequency of cardiovascular disease (Kuwajima et al., 1992), cerebrovascular disease (Kario et al., 1996; Kario et al., 2003), microalbuminuria (Lurbe et al., 2002) and progression of renal disease (Farmer et al., 1998). Elderly patients with extreme dipper and reverse dipper patterns of nocturnal blood pressure experience increased frequency of cerebrovascular disease than patients with appropriate nocturnal blood pressure dipping (Kario et al., 2003). Reverse dipper hypertensive patients experience worsening microalbuminuria compared with non-dippers and dippers (Marinakis et al., 2003). Moreover, normoalbuminuric patients with type-1 diabetes who have an increase in systolic blood pressure during sleep are more susceptible to developing microalbuminuria than patients with a normal dipping pattern (Lurbe et al., 2002). In our experience, baboons in which we measure blood pressure while on tether normally have a nighttime decrease of 5–12 mm Hg. We attribute this smaller decrease in blood pressure to the observation that our tethered baboons do not sleep in a recumbent position; i.e. they usually sleep sitting upright wedged into a corner of the tether cage. Thus, there probably is not a central shift in blood volume in our tethered baboons as there is in recumbence. Nevertheless, the diabetic baboon with the highest blood pressures did not have nighttime decrease in blood pressures compared to the other 5 tethered baboons. There are no data available about nighttime blood pressure in baboons sleeping in recumbent position and the pattern of nocturnal blood pressure dipping, extreme dipping, non-dipping or reverse dipping in the baboon.

In our study, diabetic baboons showed glycosuria, fasting hyperglycemia and abnormally elevated hemoglobin A1C levels, consistent with diabetes in spite of adequate insulin production as evidenced by normal C-peptide levels. Moreover, the 3 diabetic baboons that underwent IGTT were asymptomatic, and exhibited fasting hyperglycemia and higher glycemic peaks, followed by slow glucose disappearance rates, and a failure to return to baseline glycemic values even at 60 minutes after intravenous glucose load. One of the 4 diabetic baboons developed hyperglycemia and polyuria prior to the fasting glucose screening, and was started on insulin treatment instead of oral hypoglycemic agents because of preference by the veterinary staff that are more familiar with the use of insulin. Nevertheless, this animal showed normal C-peptide levels in spite of insulin treatment, suggesting that it has type-2 diabetes.

There is no information in the literature describing renal lesions in baboons with type-2 diabetes. A previous study of pancreatectomized type-1 diabetic baboons by Stout et al showed that diabetic baboons exhibited morphological lesions that are similar to those of mild to moderate diabetic glomerulosclerosis in humans (Stout et al., 1986). In that study, renal biopsies were performed at baseline, at 0.5 to 6 years and at 8 to 12 years of diabetes mellitus. Insulin dosage was adjusted to maintain all diabetic animals in poor metabolic control, with average daily glycosuria > 2 g. Glomerular basement membrane thickness was found to increase progressively with duration of diabetes (Stout et al., 1986). Baboons with diabetes had mesangial expansion due to increased volume of mesangial matrix, relative to controls. There was considerable inter-animal variability in the rate of progression of diabetic nephropathy secondary to type-1 diabetes in these animals, and hyalinized arterioles and early nodules were rarely seen (Stout et al., 1986 ). In this current study, baboons with type-2 diabetes showed classical characteristics of type-2 diabetes in humans. Pathological alterations included glomerular hypertrophy, increased glomerular and tubular basement membrane thickness, and mesangial matrix expansion with increased deposition of laminin and alternatively spliced cellular fibronectin ED-A, compared to controls. As in human diabetic nephropathy (Morita et al., 1998; Saito et al., 1988; Stout et al., 1993), mesangiolysis and early nodular lesions were found in two of the baboons with diabetes. In addition, one baboon showed a capsular drop and nodular glomerulosclerosis by EM, a feature seldom seen in conditions other than diabetic nephropathy (Barrie et al., 1952). In humans, mesangiolysis begins by focal and segmental disintegration of the mesangium, resulting in cystic or aneurismal dilatation of the involved capillary tufts. The dilated tufts are then filled with lysed mesangial matrix in a reticular or fibrillar arrangement, followed by a concentric arrangement and the ultimate formation of diabetic nodules (Saito et al., 1988; Stout et al., 1993 ).

Humans also exhibit arteriolar hyalinosis of the afferent and efferent arterioles. None of the baboons biopsied in our study had interstitial fibrosis or arteriolar hyalinosis. However, it should be emphasized that the entire baboon colony has been maintained on a relatively reno-protective diet, low in fat and carbohydrates and rich in nuts and vegetables. Other risk factors that contribute to the development and progression of renal lesions in type-2 diabetes, such as severe hypertension, smoking or hyperlipidemia were not present. Nevertheless, this is a very small sample and these types of lesions could develop in these animals at a later stage of disease progression or occur in other animals with type-2 diabetes in this colony. For example, the kidney obtained from an additional 23 year-old female baboon with type-2 diabetes at necropsy showed arteriolar hyalinosis (Figure 7). Although this animal was not part of the functional study group, it was included to further illustrate morphological features of diabetic nephropathy in type-2 diabetic baboons. There is evidence for variability of renal histological findings in humans with type-2 diabetes. A study in 34 patients with type-2 diabetes and microalbuminuria showed there is heterogeneity of renal lesions, with 29.4% of patients having normal renal histology, 29.4% having typical histological changes of diabetic nephropathy, and 41.2% having atypical patterns of injury with more pronounced tubulo-interstitial or vascular lesions (Fioretto et al., 1996).

Our study has some limitations due to the small sample size, unknown pattern of microalbuminuria in baboons and unknown duration of diabetes in these animals. Fasting blood glucose measurements were not performed in this baboon colony prior to 1999 and the natural course of type-2 diabetes in this species is unknown. Although all diabetic animals exhibited various degrees of renal pathology, the mesangial matrix increase and nodular formation are not very prominent. Two of the diabetic baboons showed mesangiolysis, but we do not have an adequate sample size to assess for the pattern of mesangiolysis and nodular formation in this model. It is unknown whether nephropathy intrinsically progresses more slowly in this species. The practical utility of the model may be limited due to extended timeframe required for animals to develop diabetic nephropathy and the expense. Nevertheless, the changes described so far show many similarities to human renal disease which make the type-2 diabetic baboon an attractive animal model of human diabetic nephropathy, deserving further characterization. For example, a recent fasting blood glucose screening in the colony has identified animals as young as 8 years old with diabetes. Therefore, we believe it is important that the natural history of type-2 diabetes and diabetic nephropathy be further characterized in this colony.

One control baboon and 3 diabetic baboons showed microalbuminuria. In humans with type-1 diabetes, microalbuminuria is a useful marker for early diagnosis of diabetic nephropathy, as well as for monitoring its natural course and effectiveness of treatment (Warram et al., 1996). Although in type-1 diabetes persistent microalbuminuria defines the state of incipient diabetic nephropathy, some patients with type-1 diabetes and microalbuminuria have normal renal morphology (Fioretto et al., 1994). In contrast, in some patients who will develop type-2 diabetes, microalbuminuria may precede or even predict the development of type-2 diabetes (Schmitz, 2000). The course of renal function is heterogeneous among microalbuminuric and proteinuric Caucasian patients with type-2 diabetes (Fioretto et al., 1995; Fioretto et al., 1996; Fioretto et al., 1998; Gambara et al., 1993; and Nosadini et al., 2000). Patients with increased volume of mesangial fraction and GBM thickness at baseline are more likely to rapidly lose renal function, have worse metabolic control and are less responsive to tight blood pressure control (Fioretto et al., 1995; Nosadini et al., 2000). Hyperglycemia may cause different patterns of renal injury in patients with type-2 diabetes, who tend to be older, compared with patients with type-1 diabetes, who tend to be younger (Fioretto et al., 1997; Ritz et al., 1997). The fact that the morphological changes in baboons resemble those of humans may allow using serial kidney biopsies to follow the progression of the disease within a reasonable timeframe, rather than progression of microalbuminuria to proteinuria. Therefore, it is important to identify a non-human primate model of type-2 diabetes that develops spontaneous diabetic nephropathy lesions that are similar to those occurring in humans. This primate model may be useful to study the course of nephropathy in type-2 diabetes, identify pathogenetic mechanisms and assess efficacy of newer therapeutic interventions.

Abbreviations

ABC

avidin-biotin complex

BSA

bovine serum qwalbumin

DAB

3′,3′-diaminobenzidine tetra hydrochloride

DBP

diastolic blood pressure

DM

diabetes mellitus

ED-A

fibronectin extradomain-A

GBM

Glomerular basement membrane

H & E

Hematoxylin and eosin

IGTT

intravenous glucose tolerance test

PAS

periodic acid Schiff

PBS

phosphate-buffered saline

SBP

systolic blood pressure

Ur A/C

urinary albumin/creatinine