Type 2 Diabetes Mellitus in the Northern Tree Shrew (Tupaia belangeri)

Abstract

Within a 2-year period, 5 adult, single-housed, northern tree shrews (Tupaia belangeri) presented for a range of clinical signs including lethargy, hindlimb lameness, decreased visual acuity, polyuria, and polydipsia. Diagnostic findings, including point-of-care blood glucose and ketone measurements, urine dipstick analysis, insulin, fructosamine, and glycosylated hemoglobin levels, and gross and histopathologic lesions, were consistent with a diagnosis of type 2 diabetes mellitus. Although spontaneous diabetes mellitus has been reported in other members of the order Scandentia, to our knowledge, this represents the first published case series describing naturally occurring type 2 diabetes mellitus in the northern tree shrew.

Introduction

Tree shrews (Tupaia spp. and related genera) are small mammals native to Southeast Asia that have been gaining attention as animal models due to their close evolutionary relationship to primates and physiologic similarities to humans.^1–3 In particular, tree shrews have been studied for their potential as models for hepatitis B and C virus infections, neurodegenerative disease, and diabetes.^2,4,5 Their endocrine physiology has some primate-like features (eg, cortisol being the primary stress hormone rather than corticosterone as in rodents), and they exhibit some unique metabolic traits such as naturally high HDL cholesterol levels.^2,6–8

Diabetes mellitus is a common metabolic disease presentation in many species, with a human prevalence ∼11%, and type 2 constituting 95% of these cases.⁹ The cause of this disease process stems from a combination of a decreased peripheral response to insulin and a decreased insulin secretion by β cells.¹⁰ In veterinary medicine, type 2 diabetes has been reported in many species, including cats, pigs, and nonhuman primates (NHPs).^11–13 Both obesity and advanced age are risk factors associated with the onset of insulin resistance leading to type 2 diabetes.^10,11,13 Treatment in humans first and foremost involves lifestyle changes such as diet reform, with first-line pharmacological interventions aimed at controlling hyperglycemia, including biguanides, SGLT2 inhibitors, and GLP-1 receptor agonists.^14–16

Although experimental induction of diabetic phenotypes has been described in tree shrews, reports of spontaneous diabetes mellitus are rare.⁶ A single historical report documents spontaneous diabetes mellitus in the Mindanao tree shrew (Urogale everetti); however, spontaneously occurring diabetes mellitus has not previously been reported in Tupaia belangeri, although in one case report it has been suggested as a potential diagnosis.^6,17,18

Case Report

Over the course of 1 year, 4 mature adult northern tree shrews (5-7 years of age) presented for a range of clinical signs including lethargy, hindlimb lameness, decreased visual acuity, polyuria, and polydipsia. Two male (shrews S and R) and 2 female (shrews M and T) adult intact shrews comprised the cohort. All animals presented were participants in visual acuity task-based behavior work-for-juice reward. Polyuria and polydipsia presented in 3 of the 4 individuals (S, M, and T) whereas decreased visual acuity, hindlimb lameness, lethargy, unkempt coat, and acute weight gain each occurred in one of the shrews (S). General colony weight ranged from 160 to 260 g; although absolute weight did not correlate with the onset of type 2 diabetes, an acute (within a 1- to 2-month period) weight gain in adults (weights ranging from 200 to 240 g) did present with other concurrent clinical signs (S). Around 1.5 years following the presentation of the cohort of 4 affected animals, a fifth mature adult (5.5-year-old) female shrew (H) presented with acute onset of polydipsia.

In initial urinalysis, using a VET-10 urine dipstick (Jorgensen Laboratories, Loveland, CO) and refractometer, one of the individuals (S) presented with a decreased urine specific gravity of 1.010 (the baseline for the colony was established at 1.020-1.030 using the urine of young animals without polyuria/polydipsia). Moderate ketones were detected in the same individual (4 mmol/L) and were absent in the other 3. All 4 shrews (S, M, T, and R) had urine glucose values ranging from 1,000 to 2,000+ mg/dL on the dipstick analysis of pan-collected urine. After initial evaluation, venipuncture (via the saphenous vein with manual restraint) was performed and initial blood glucose values ranged from 319 to >650 mg/dL (Table 1). These values were severely elevated compared with published ranges (100-150 mg/dL)⁶ and unpublished ranges (115-175 mg/dL; Max Planck Florida Institute for Neuroscience). Insulin levels were obtained around the time of diagnosis for one of the females, 1.08 ng/mL (T), which falls within a normal range of 0.3-1.1 ng/mL (Antech Diagnostics, Fountain Valley, CA).⁶ The presence of normoinsulinemia in the face of marked hyperglycemia supported a diagnosis of early stage type 2 diabetes mellitus in T. Glycosylated hemoglobin (HbA1c) was determined for one individual in the initial cohort (M) and presented >14% with a reference range of 5%-8% (Antech Diagnostics, Fountain Valley, CA).¹⁹ Serial blood glucose measurements in H taken with the point-of-care Contour glucometer (Ascensia Diabetes Care, Basel, Switzerland) in the days prior to euthanasia indicated blood glucose ranging from 342 to 412 mg/dL, with consistent moderate glucosuria on dipstick analysis (1,000-2,000+ mg/dL). Insulin level at euthanasia was 0.4 ng/mL, which was considered within normal limits. Blood glucose on chemistry analysis was 588 mg/dL (Antech Diagnostics, Fountain Valley, CA). The glycosylated hemoglobin was 7.2%, which is on the high end of a normal reference range.⁶ A fructosamine level obtained at euthanasia had a value of 345 µmol/L (Antech Diagnostics, Fountain Valley, CA). Normal fructosamine levels for tree shrews are currently not established in the literature, but a working reference range has been published for rhesus macaques (157-230 µmol/L).^20,21 Unlike HbA1c, which reflects 1-4 months of glucose history, fructosamine reflects the average blood glucose over the previous 2-4 weeks, indicating a much shorter timeline of identification of type 2 diabetes onset for shrew H compared with shrew M.²²

Table 1.

Blood Glucose Values (mg/dL) per Individual at Time of Diagnosis and Throughout Treatment as It Applies to Each

Event	S	M	R	T	H
Time of diagnosis	>650	477	319	470	405
Metformin 5 mg/kg	>650	342	N/A	N/A	N/A
Metformin 10 mg/kg	350-450	321	250-350	N/A	N/A
Euthanasia	747	321	446	470	588

Two male (S and R) and 3 female (M, T, and H) mature adult intact shrews comprised the cohort. All values obtained after treatment were sampled at least 3 wk after initiation of that treatment.

The most significant histopathologic findings across individuals were within the hepatobiliary system. Microscopic examination of liver sections revealed severe nodular hepatic lipidosis and moderate glycogen vacuolation of frequent hepatocytes and the biliary ductular epithelium (Figure 1). Microscopic examination of pancreas sections revealed glycogen vacuolation of the ductular epithelium and islet cells and islet lipidosis (Figure 2). Finally, microscopic evaluation of kidney slides revealed glomerulosclerosis, mesangial proliferation, and hyaline thickening of vascular basement membranes as well as glycogen nephrosis (Figure 3). Due to experimental needs of the investigator, histopathologic findings of ocular changes could not be assessed.

Figure 1.

Liver. Multifocally randomly throughout the parenchyma are discrete nodules of hepatic lipidosis, which compress the surrounding parenchyma. Affected hepatocytes are markedly expanded by discrete microvesicular to macrovesicular cytoplasmic vacuolation. Remaining hepatocytes are moderately expanded by feathery cytoplasmic vacuolation (glycogen). Frequently within portal tracts, there is mild diffuse glycogen-type vacuolation of the biliary ductular epithelium. Hematoxylin and eosin; original magnification ×40.

Figure 2.

Pancreas. There is multifocal moderate nodular hyperplasia of islet cell clusters, including ductuloinsular complexes (nesidioblastosis). Islet cells are frequently expanded by marked cytoplasmic vacuolation (glycogen and/or lipid). Hematoxylin and eosin; original magnification ×40.

Figure 3.

Kidney. Diffusely, glomeruli have one or more of the following changes: segmental to global expansion of the mesangial matrix by hyalinized, eosinophilic, fibrillar material with decreased cellularity (glomerulosclerosis), which occasionally replaces entire glomerular tuft; moderate to marked thickening of the basement membrane of Bowman’s capsule and glomerular capillary basement membranes; and periglomerular fibrosis (diabetic nephropathy). The interstitium and renal pelvis are multifocally infiltrated by moderate numbers of lymphocytes and fewer plasma cells. Tubule epithelial cells are multifocally expanded by vacuolation (glycogen nephrosis). Hematoxylin and eosin; original magnification ×40.

Once this cohort of affected animals was identified, a combination of pharmaceutical, dietary, and environmental interventions was pursued in S, M, and R. Initially, metformin was dosed at 5 mg/kg by mouth twice per day via injection into superworms (Zophobas morio) that were then offered to the tree shrews for consumption. This dosing did not modulate blood glucose levels within a 2-week timeline, and thus based on tolerance by the shrews and published dosing in marmosets, a therapeutic dose of 10 mg/kg was chosen.²³ Metformin, combined with dietary modification, ameliorated clinical signs observed in individual animals and decreased both the blood glucose and degree of glucosuria (2,000+ mg/dL to 500 mg/dL) for 3-8 months depending on the individual, at which point hyperglycemia became unmanageable with oral medical management, and humane euthanasia (animals were sedated with a combination of 40 mg/kg ketamine, 0.15-0.2 mg/kg dexmedetomidine, and 1-2 mg/kg butorphanol) was elected (Table 1). Dexmedetomidine is known to raise blood glucose levels, particularly in diabetic patients, which likely contributed to the increased hyperglycemia noted at the time of euthanasia compared with Contour glucometer–derived values.^24,25 Response to metformin varied between individuals and may have been influenced, in part, by variable voluntary consumption of the medicated superworms, resulting in inconsistent oral drug delivery. Body weight did not vary significantly (>10 g) after the initiation of any pharmacological intervention.

Insulin therapy was not pursued due to several factors, including preserved endogenous insulin production at the time of diagnosis, lack of species-specific insulin injection or implant dosing data for tree shrews, difficulty of daily restraint for injection, scientific objectives of the laboratory, and the elevated risk of hypoglycemia in small-bodied exotic mammals. Metformin was selected due to its favorable safety profile, absence of hypoglycemia when used as a monotherapy, ability to be dosed orally, and primary mechanism of action targeting insulin resistance through suppression of hepatic gluconeogenesis and enhancement of peripheral insulin sensitivity.

Regarding dietary modification, tree shrews are both insectivorous and frugivorous in the wild.^26–28 The initial diet consisted of Purina cat chow (Nestle Purina PetCare, St. Louis, MO), Mazuri insectivore chow (Mazuri Exotic Animal Nutrition, Shoreview, MN), hard-boiled eggs, mealworms, superworms, bananas, berries, apples, and apple juice as a reward for behavior task training as administered by the laboratory. This diet was transitioned to Purina cat chow diabetic management (Nestle Purina PetCare, St. Louis, MO), substituting higher glycemic index fruits such as apples and bananas with lower index fruits such as blueberries and grapes, and adding increased protein to the diet.²⁹ To this end, the ratio of in-shell peanuts, hard-boiled eggs, mealworms, and superworms were increased, along with the introduction of more vegetables such as carrots, cucumbers, avocado, sweet potatoes, and snap peas. An attempt was also made to transition rewards from juice to either diluted juice or juices that contained alternative sweeteners such as stevia or monk fruit. Unfortunately, due to the long history (>1 year) of juice reward administration (up to 5 d/wk), the shrews were unwilling to transition to performing their behavior tasks at full ability unless fully concentrated juice was offered. This dietary change appeared to be effective for the colony, in that only 1 other tree shrew (H) out of 13 was identified in the subsequent 2-year period as being diabetic, and this individual had a history of obesity and increased mammary development since transfer from its previous institution. The base chow was changed in recent months from Purina cat chow diabetic management (Nestle Purina PetCare, St. Louis, MO) to 340 g of LabDiet 5L14 (high-density ferret diet; Nestle Purina PetCare, St. Louis, MO), which has not resulted in changes to the average monthly weights for the adult animals in our colony. Additional enrichment rotates seasonally (eg, 1 baby carrot, 3 snap peas, and 2 whole peanuts or 10 mealworms or ¼ hardboiled egg, 3 blueberries, and ⅛ avocado). Other institutions with ongoing tree shrew research have reported success at managing the weight in their colony animals (particularly breeders) with Laboratory Feline Diet 5003 (Nestle Purina PetCare, St. Louis, MO), Mazuri insectivore extruded pellets (Mazuri Exotic Animal Nutrition, Shoreview, MN), and other permutations of lower-fat, high-protein diets.

The environmental factors that were changed in the colony after the initial cohort of diabetic animals appeared included water quantitation to monitor for polydipsia, and obtaining monthly individual weights to identify animals with a propensity for obesity. Weekly, the animal care technicians fill each animal’s water bottle. Daily, the animal care technicians tracked water intake by marking the current water level with a piece of tape and assessing the amount consumed in a 24-hour period. They were instructed to contact veterinary staff if an animal began to drink 50% more than its conspecifics. This methodology was used to identify the final affected animal (H) before the appearance of any clinical signs other than polydipsia. Routine screening of overweight animals for point-of-care blood glucose values, as well as urine screening with dipstick analysis, are relatively easy interventions that require only a single drop of blood and a urine catch pan, all of which are minimally invasive and provide valuable feedback regarding overall colony health.

Housing and care took place on IACUC-approved protocols at a facility accredited by AAALAC International, adhering to the guidelines outlined in the 8th edition of the Guide for the Care and Use of Laboratory Animals. This was done in accordance with the Public Health Service Policy on the Humane Care and Use of Laboratory Animals and the Animal Welfare Act and its amendments. Veterinary care was provided following best practices specific to the species and condition.

Discussion

The northern tree shrew (Tupaia belangeri) serves as a unique small animal model for a multitude of fields of study. These include hepatitis, neoplasia, social stress, depression, aging, learning behaviors, vision, and toxicology.³⁰ Their small size, short reproductive cycle, high brain/body ratio, and close affinity to primates at a lower cost of maintenance make them a desirable model.³¹ Although this is not the first report of spontaneous type 2 diabetes in the order Scandentia, as previous reports in the Mindanao tree shrew exist, to our knowledge, this is the first confirmed report in the northern tree shrew.¹⁷

Insulin therapy is frequently the mainstay treatment for overt clinical diabetes mellitus in many veterinary species, but it requires repeated parenteral dosing and careful monitoring to avoid iatrogenic hypoglycemia, which can be life-threatening. Current veterinary consensus emphasizes selecting a treatment plan that achieves acceptable clinical control while prioritizing avoidance of hypoglycemia and recognizes that the intensity of monitoring and feasibility of repeated dosing are practical constraints that can influence case management.³² In the present case series, insulin therapy was not selected as a first-line therapeutic due to a lack of species-specific dosing and monitoring data for insulin administration as well as welfare and feasibility concerns associated with frequent injections and intensive glucose monitoring in a small-bodied, wild, easily stressed species. For these reasons, oral antihyperglycemic medications were pursued.

Multiple antihyperglycemic drug classes are used in human type 2 diabetes mellitus, some of which have been evaluated in NHPs as translational intermediaries, but translation to the northern tree shrews is limited by the near absence of species-specific pharmacokinetic dosing and safety data. Thiazolidinediones, particularly pioglitazone, improve insulin sensitivity and glycemic indices in obese, insulin-resistant rhesus macaques, yet are associated with fluid retention/edema and weight gain.^33,34 Clinical consensus statements also highlight edema and congestive heart failure risk in susceptible patients receiving thiazolidinediones.³⁵ Sodium-glucose cotransporter 2 (SGLT2) inhibitors have clear renal pharmacodynamics in NHPs, and in rhesus macaques they can induce characteristic metabolic shifts, including increased ketone body levels.³⁶ The disposition of bexagliflozin has been characterized in cynomolgus macaques and humans, but in people it increases urinary glucose excretion and urine volume and carries risks of volume depletion, genitourinary infection, and ketoacidosis.^37–39 In the present cases, marked baseline glucosuria with polyuria/polydipsia and intermittent ketonuria were already present, and adding an SGLT2 inhibitor was judged likely to exacerbate osmotic diuresis and potentially increase risk of ketotic complications.^36–38,40 Other classes of drugs likewise have limited published experience across NHP species and essentially none in tree shrews, and they may introduce practical constraints (injectable administration) or adverse-effect profiles (notably gastrointestinal intolerance) that complicate management, and thus metformin was selected as the initial oral antihyperglycemic.⁴¹

Metformin was selected because it directly targets a core mechanism of type 2 diabetes mellitus, hepatic glucose production, and improves insulin sensitivity without stimulating insulin secretion directly.⁴² Contemporary known mechanisms and clinical experience support that the principal effect is mediated largely through suppression of hepatic gluconeogenesis, and importantly it rarely causes hypoglycemia when used as a monotherapy.⁴² The most common adverse effects are gastrointestinal (eg, reduced appetite, soft stool/diarrhea) and may limit dose escalation, which we did not experience as a limiting factor.

In our colony, successful treatment was defined as the amelioration of clinical signs. We were able to manage our shrews using metformin at 10 mg/kg as well as dietary changes for 3-8 months before acute declines were noted, necessitating humane euthanasia. After the presentation of the initial four shrews, a screening plan through periodic urine dipstick measurements allowed us to more proactively diagnose the fifth tree shrew, after which blood glucose diagnostics confirmed the case as diabetes. This screening plan is important for any colony of aging tree shrews, as the fifth tree shrew developed clinical signs of diabetes more than a year after the colony had been transitioned to a modified diet and was not receiving juice rewards.

Behavioral management in laboratory settings often relies on positive reinforcement with food rewards. Reinforced choice can become a meaningful contributor to total caloric and simple-sugar exposure when rewards are delivered frequently over long periods.⁴³ In captive macaques, obesity is common and has important research and clinical implications, including associations with insulin resistance and type 2 diabetes, underscoring that ‘extra’ calories from supplements and treats are not metabolically neutral.⁴⁴ Importantly, nonhuman metabolic research demonstrates biologic plausibility for a reward system rich in simple sugars leading to the development of insulin resistance and metabolic syndrome progressing to type 2 diabetes.⁴⁵ This concern is also relevant for tree shrews specifically, as insulin resistance has been seen with higher sugar intake.⁵ Taken together, these data support the interpretation that the original husbandry strategy (including a juice-based, frequent, reinforcement system) may have contributed to the disease onset or acceleration in the present cases. As a practical mitigation strategy, future reinforcement plans for tree shrews should prioritize low levels of added sugar, lower glycemic rewards, portion control, and explicit accounting of reward calories within daily rations. At least one group has reported success training with products such as Ensure Clear (Abbott Laboratories, Abbott Park, IL). Monthly weight checks and fluid intake quantitation via tape markers on the water bottles have served as a reliable colony-wide screening method for technicians to recognize and report obesity and polydipsia, allowing for earlier identification of potentially affected individuals.

Conflict of Interest

The authors have no conflicts of interest to declare.

Funding

This work was supported by the University of Pennsylvania Office of the Vice Provost for Research.