Abstract
Diabetes is induced in mice by using streptozotocin (STZ), a compound that has a preferential toxicity toward pancreatic β cells. We evaluated nude male mice from various sources for their sensitivity to a single high dose (160 to 240 mg/kg) of STZ. Diabetes was induced in male mice (age: median, 12 wk; interquartile range, 11 to 14 wk; body weight, about 30 g) from Taconic Farms (TAC), Jackson Laboratories (JAX), and Charles River Laboratories (CRL). Mice were monitored for 30 d for adverse side effects, blood glucose, and insulin requirements. In CRL mice given 240 mg/kg STZ, more than 95% developed diabetes within 4 to 5 d, and loss of body weight was relatively low (mean, 0.4 g). In comparison, both TAC and JAX mice were more sensitive to STZ, as evidenced by faster development of diabetes (even at a lower STZ dose), greater need for insulin after STZ, greater body weight loss (mean: TAC, 3.5 g; JAX, 3.7 g), and greater mortality. We recommend conducting exploratory safety assessments when selecting a nude mouse source, with the aim of limiting morbidity and mortality to less than 10%.
Abbreviations: CRL, Charles River Laboratories; JAX, Jackson Laboratories; STZ, streptozotocin; TAC, Taconic Farms.
Rodent models commonly are used to study immunologic mechanisms and metabolic function in diabetes.19 In our institution, the mouse diabetes model is used to test islet cell preparations for their activity in diabetes reversal. We use congenitally athymic nude mice to avoid any interference of immune rejection on the outcome of results.
Diabetes is induced by streptozotocin (STZ), a glucosamine–nitrosourea compound derived from Streptomyces achromogenes that is used clinically as a chemotherapeutic agent in the treatment of pancreatic β cell carcinoma. STZ damages pancreatic β cells, resulting in hypoinsulinemia and hyperglycemia.10 STZ can induce a diabetic state in 2 ways, depending on the dose. The selectivity for β cells is associated with preferential accumulation of the chemical in β cells after entry through the GLUT2 glucose transporter receptor: chemical structural similarity with glucose allows STZ to bind to this receptor. The mode of action has best been demonstrated in mouse studies. At high doses, typically given singly, STZ targets β cells by its alkylating property corresponding to that of cytotoxic nitrosourea compounds.4 At low doses, generally given in multiple exposures, STZ elicits an immune and inflammatory reaction, presumably related with the release of glutamic acid decarboxylase autoantigens. Under this condition, the destruction of β cells and induction of the hyperglycemic state is associated with inflammatory infiltrates including lymphocytes in the pancreatic islets.12 STZ has well-known adverse side effects, which include hepatotoxicity and nephrotoxicity.4,13-15,17
Nude mice (Fox1nu/Fox1nu homozygotes) are available from a variety of vendors.2,8,18 Although all nude mice ultimately have the same origin (a mutant in a colony at the NIH), each vendor provides a unique subline. The development of sublines occurs as each population of mice becomes generationally distant from a common ancestor. Individual populations will develop identifiable genotype and phenotypic responses based on the accumulation and maintenance of normal random genetic mutations.6,16
We used nude mice from 3 vendors (Charles River Laboratories [CRL], Taconic Farms [TAC], and Jackson Laboratories [JAX]) in establishing the mouse STZ-induced diabetes model. In using mice from these vendors, we observed remarkable differences in their mice's sensitivity to STZ induction of diabetes and in morbidity and mortality after STZ treatment. Differences in sensitivity to STZ between mouse strains have been described anecdotally, but we are unaware reports of differences among mice of essentially the same strain provided by different vendors. Because these differences include adverse events and mortality, the present evaluation has relevance to animal loss and wellbeing and adaptations to currently accepted generalized dose practices.
Materials and Methods
Animals.
The nomenclature used to describe the strains below is based on the recommendations of the International Committee on Standardized Genetic Nomenclature for Mice.11 This standardized nomenclature reflects both the genotype and the original source of the strain from a distinct supplier. SPF male congenitally athymic nude mice with a median age of 12 wk (interquartile range, 11 to 14 wk) and weighing approximately 30 g were obtained from 3 commercial suppliers (CRL, TAC, and JAX); the following background information was published on the vendors’ respective websites.2,8,18
CRL mice are Crl:NU-Foxn1nu mice from Charles River Laboratories (Wilmington, MA). These animals originated from NIH and were originally thought to be BALB/c congenics. It was later determined that they were not inbred; therefore they were maintained as outbred and were not associated with any stock or strain.2 TAC mice are CrTac:NCr-Foxn1nu animals from Taconic Farms (Germantown, NY). These mice have both BALB/c inbred and NIH(S) outbred stock in their genetic background. The outbred background originated from an accidental cross between the BALB/c inbred nude and NIH(S) outbred nude mice. The company received the NCr nude spontaneous mutant model from the National Cancer Institute in 1993 after several years of random breeding.18 JAX mice are NU/J animals from Jackson Laboratories (Bar Harbor, ME). The company imported the nude mutation from the NIH on an outbred stock in 1975. As of 2008, the strain has been inbred for at least 100 generations. The strain is on a BALB/c background.8
The mice were kept under SPF conditions by using cages equipped with filter tops and absorbent bedding (7092 Teklad Corn Cob Bedding, Harlan Laboratories, Madison, WI) in a temperature-controlled environment (22 to 25 °C) on a 12:12-h light:dark photoperiod. Mice were housed in groups of 2 to 4 per cage. Water was provided as libitum, and animals were fed irradiated rodent diet (Diet 2919 Teklad Global 19% Protein Rodent Diet, Harlan Laboratories). This diet was selected preferentially over standard chow because of its slightly higher energy density compared with that of conventional mouse diet (3.3 kcal/g and 2.9 kcal/g, respectively). This higher energy level is advantageous in support of the diabetic state. Studies using these mice for assessment of islet cell preparations were approved by the University of Minnesota Institutional Animal Care and Use Committee, conducted in compliance with the Animal Welfare Act,1 and adhered to principles stated in the Guide for Care and Use of Laboratory Animals.7
Diabetes induction and animal monitoring.
The mice were injected intraperitoneally with a single high dose of 160 to 240 mg/kg STZ. We used a pharmaceutical-grade formulation of STZ (Zanosar Teva Pharmaceuticals, Irvine, CA) to avoid impurities that may have harmful biologic activities. STZ is administered at a higher dose level in nude mice than in immunocompetent mice because the destruction of β cells in nude mice must result from direct drug action with no assistance from the immune-mediated response.10 We followed the dose-range recommendations of the Clinical Islet Transplant (CIT) consortium regarding procedures for the mouse bioassay in testing human islet cell preparations to be used in clinical trials (150 to 240 mg/kg).3 Historically we used a dose of 240 mg/kg to achieve consistently a diabetic state with limited morbidity and mortality. However, when mice were obtained from alternate suppliers, unacceptable toxicity was observed, and dose corrections within the range recommended by the consortium were attempted immediately. The dose adjustments were made based on the severity of adverse events and in a prospective and stepwise fashion: first reducing to 200 mg/kg, then 180 mg/kg, and finally 160 mg/kg. The average dose and range of each cohort is presented in Table 1. The mice subsequently were hydrated with 1.0 mL normal physiologic saline administered subcutaneously. Blood was obtained by lancet prick in the tail. Body weight or blood glucose concentration (or both) was monitored once before and daily after STZ injection until a diabetic state was confirmed by the glucose dehydrogenase method (AlphaTRAK, Abbot Laboratories, Chicago, IL). Mice with a glucose concentration exceeding 300 mg/dL were considered diabetic. Insulin therapy (glargine; Lantus, Sanofi-Aventis US, Bridgewater, NJ) was initiated at a dose of 0.5 U every other day after 3 consecutive glucose measurements exceeding 300 mg/dL. The dose and frequency of insulin administration (range, 0.1 to 1.2 U/kg daily) was increased or decreased in response to a combination of measured glucose and body weight. For example, when the blood glucose was greater than 600 mg/dL in combination with a weight loss greater than 3% with respect to the previous value, the insulin dose was increased. In cases of dehydration or progressive weight loss, the frequency of dosing was increased.
Table 1.
Demographic variables and the effect of STZ on metabolic function in nude mice from 3 sources
| CRL (n = 60) | TAC (n = 23) | JAX (n = 58) | |
| Age at induction (d) | 89 ± 17 | 90 ± 12 | 85 ± 12 |
| Body weight before STZ(g) | 31.3 ± 2.3 | 30.2 ± 1.4 | 29.6 ± 0.8c |
| Average body weight after STZ (g)a | 30.3 ± 2.2 | 27.1 ± 1.8c | 26.3 ± 1.3c |
| Body weight loss (g)a | 0.37 ± 2.3 | 3.5 ± 3.4c | 3.7 ± 2.3c |
| STZ dose (mg/kg) | 240 ± 0 | 223 ± 26 | 181 ± 5c,d |
| STZ dose range (mg/kg) | 240 | 160–240 | 180–200 |
| Days to blood glucose > 300 (mg/dL) | 2.3 ± 0.8 | 1.6 ± 0.7b | 1.0 ± 0.1c,d |
| Average insulin requirement (U/d)a | 0.32 ± 0.12 | 0.70 ± 0.33c | 0.49 ± 0.07c |
| Random blood glucose before STZ (mg/dL) | 157 ± 25 | 144 ± 29 | 134 ± 21c |
| Average random blood glucose after STZ (mg/dL)a | 502 ± 63 | 503 ± 73 | 554 ± 44b,d |
Data are presented as arithmetic mean ± 1 SD.
Averages calculated in animals developing diabetes and surviving at least 7 d after injection (CRL, n = 55; TAC, n = 11; JAX, n = 56).
P < 0.01 compared with CRL
P < 0.001 compared with CRL
P < 0.001 compared with TAC
Mice were inspected daily for signs of pain or distress, including changes in respiration, appetite, urine output, excessive thirst, dehydration, activity (for example, lethargy or hyperactivity), weight loss exceeding 10% of the initial value, unkempt appearance, abnormal posture, and twitching or trembling. The measurement of body weight and observation of body condition (for example, thin, normal, overweight) acted as a surrogate marker of appetite. The measurement of skin turgor and observation of cage bedding for urine output acted as a surrogate marker of thirst. These complications were documented, and mice received routine medical management appropriate to presenting symptoms (for example, warming pads for hypothermia, warm physiologic saline for dehydration, dextrose or insulin for metabolic correction, analgesics for management of pain or distress). Mice that manifested complications that did not respond quickly to medical treatment were euthanized promptly with an overdose of carbon dioxide. The follow-up period was kept at 30 d after STZ injection.
Mice with a successful course after diabetes induction subsequently were allocated to islet transplantation studies.
Statistics.
The data for various demographic and biochemical parameters are expressed as mean ± 1 SD and were compared by using one-way ANOVA followed by a Bonferroni–Dunn test for multiple comparisons. The nonparametric Wilcoxon–Mann–Whitney test was used to compare median values. Proportions were compared by using the Fisher exact test. Kaplan–Meier plots were evaluated by using the log-rank test. All analyses were run by using Prism and Instat software (GraphPad Software, San Diego, CA). Values were considered statistically significant when P < 0.05.
Results
Nude mice from all 3 sources were similar in regard to age, body weight, and glucose level before STZ injection (Table 1). Statistical analysis showed significance (P < 0.05) for the lower body weight and lower glucose level in JAX mice when compared with CRL mice, but the actual values were considered within the expected normal range according to supplier growth curves.
A single high dose (160 to 240 mg/kg) of STZ effectively induced diabetes: 96.5% of mice were diabetic by day 5 after STZ administration (Figure 1 A).
Figure 1.
(A) Diabetes induction and (B) survival after diabetes induction in male nude mice from Charles River Laboratories (CRL), Jackson Laboratories (JAX), and Taconic Farms (TAC). (A) The development of diabetes (glucose greater than or equal to 300 mg/dL) is presented as Kaplan–Meier estimates plotted over time after STZ infusion. 100% of JAX and TAC mice rapidly developed diabetes by day 2 after STZ injection, whereas CRL mice needed 5 d to achieve a diabetic state in 92% of mice (P < 0.001). (B) Death or euthanasia is presented as Kaplan–Meier estimates plotted over time after STZ infusion. CRL mice demonstrated significantly better survival (P < 0.001), compared with JAX and TAC mice.
The time to reach the diabetic state and the average glucose level after STZ differed significantly between the 3 groups of mice (Table 1, Figure 1 A). In all cases, JAX mice developed diabetes within 1 d after treatment, whereas achieving this state took (on average) 1.6 d in TAC mice and 2.3 d in CRL mice. The average glucose level after STZ was highest in JAX mice (554 mg/dL) and lowest in CRL mice (502 mg/dL). The dose of insulin was on average highest in the TAC group (0.70 ± 0.33 U daily) and lowest in the CRL group (0.32 ± 0.12 U daily), P < 0.001. The level of hyperglycemia was not related to the insulin dose (Figure 2),
Figure 2.
Blood glucose values (top row) and body weight (bottom row) in response to insulin (glargine) treatment (U/kg daily) after diabetes induction by using STZ in male nude mice from Charles River Laboratories (CRL), Jackson Laboratories (JAX), and Taconic Farms (TAC). Data are presented as mean ± 1 SD. Dashed lines indicate average blood glucose value after STZ (top row) and average baseline weight (bottom row).
Adverse effects of STZ injection included weight loss, respiratory distress, rapid glycemic shifts resulting in life-threatening hypoglycemia, and a generalized poor body condition (Table 2) that manifested with higher frequency in JAX and TAC mice (P < 0.001). The STZ dose was decreased in the cohorts with these complications (JAX, TAC). In JAX and TAC mice, the dose reductions significantly (P < 0.0001) increased the number of days that mice survived after STZ administration. Median survival was 23 d (interquartile range, 15 to 30 d) in mice given less than 200 mg/kg STZ, whereas it was 7 d (interquartile range, 2 to 14 d) in mice receiving 200 mg/kg or more. However, the time to development of diabetes, average insulin requirement after STZ, random glucose level after STZ, and overall complication rates were not affected by dose reduction. This effect might be related to the rather high dose range that was used relative to published recommendations,3 so that within this dose range, only a relationship with severe toxicity at doses exceeding 200 mg/kg became evident. Overall, CRL mice received the highest dose (240 mg/kg) and experienced the lowest rate of complications. Even at lower average doses (223 mg/kg in TAC mice and 181 mg/kg in JAX mice), TAC and JAX mice experienced significantly (P < 0.05) more weight loss despite receiving more insulin (Table 1, Figure 2). Using a 10% weight loss as a threshold value for animal health status, the incidence of weight loss exceeding this threshold was lowest in CRL mice (5%), higher in TAC mice (35%), and highest in JAX mice (55%).
Table 2.
Complications (%) in animals from various sources during a 30-d follow-up after diabetes induction
| CRL | TAC | JAX | |
| (n = 60) | (n = 23) | (n = 58) | |
| Overall complicationa rate | 8 | 83b | 71b |
| By category | |||
| Respiratory distress | 0 | 35b,c | 2 |
| Life-threatening hypoglycemia | 2 | 26d | 64b,e |
| Generalized poor body | 8 | 44b | 45b |
| condition | |||
| At least 10% loss in average | 5 | 35d | 55b |
| body weight after STZ | |||
Complications were defined as an adverse effect from STZ treatment resulting in death or euthanasia.
P < 0.001 compared with CRL
P < 0.001 compared with JAX
P < 0.01 compared with CRL
P < 0.05 compared with TAC
The proportions of mice that survived with no complications were 17% for TAC mice, 13% for JAX mice, and 92% for CRL mice (Table 2). Survival is illustrated in Figure 1 B: 50% survival after STZ treatment was 7 d for TAC mice, 25 d for JAX mice, and greater than 30 d for CRL mice. The occurrence of complications, either separately or in combination, was highest in JAX mice and lowest in CRL mice. Severe complications that either caused death or required euthanasia occurred in 83% of TAC mice, 71% of JAX mice, and 8% of CRL mice.
Discussion
At our institution, we use STZ-induced diabetes in congenitally athymic nude mice to evaluate the in vivo potency of islet cell preparations. Diabetes induction is accomplished by using a single high STZ dose, to achieve optimal toxicity to pancreatic β cells without the possibility of remaining endogenous insulin-synthesizing capacity that would confound the results. Adhering to the dose range criteria for centers participating in Clinical Islet Transplant (CIT) consortium studies (namely, 150 to 240 mg/kg), we used the 240-mg/kg dose initially and adapted the dose level to 160, 180, 200, or 240 mg/kg on the basis of the severity of adverse events.3 Because the approach to dose reduction was essentially a pragmatic one in response to observed adverse events (see Materials and Methods), the present study was not powered to assess the specific influence of specific dosages on achievement of diabetic state. However, the doses used in each of the cohorts were sufficient to induce a diabetic state (Figure 1 A). In using the model, we were confronted with variable but sometimes severe and frequent adverse side effects in our mice, resulting in a small window between effective diabetes induction and complications of STZ treatment. This difficulty was apparent even when we used pharmaceutical-grade STZ (Streptozocin, Zanosar), for which chemical impurities and lot variability are limited. The present comparative evaluation of 3 sources of nude mice shows remarkable variation within this STZ dose range for effectiveness and complication rate.
TAC mice, and to a lesser extent JAX mice, showed a rather small STZ dose window in contrast to that of CRL mice. This difference was already evident in the variable sensitivity to STZ during diabetes induction, which occurred more rapidly and required a lower dose in TAC and particularly JAX mice, which subsequently had the highest blood glucose level after STZ (Table 1). In addition, the requirement for insulin was highest in TAC and JAX mice. This greater sensitivity to STZ was further apparent in the complication rates. After STZ, both TAC and JAX mice showed substantial weight loss, which was often accompanied by death or the need for euthanasia.
In contrast to the results in TAC and JAX animals, diabetes induction by using STZ was easily achieved in CRL mice without emergence of complications. The lower sensitivity to STZ in CRL mice was reflected by a longer average period to achieve a diabetic state and a lower rate of complications. The loss of mice due to complications was 8%, and only 5% of mice passed the threshold of 10% loss in body weight. This threshold is considered relevant, because it might affect the outcome of studies conducted with the mice.
The reason for this difference between mice from the 3 sources remains to be established. Genotypic analysis such as nucleotide polymorphism assays would be required to determine the degree of genetic variation between the CRL mice and those provided by JAX and TAC. The accumulation of genetic variation in discrete populations could underlie small but possibly significant changes in phenotypic responses between these sublines, because both the origin (NIH) and background strain (BALB/c) were the same among the 3 sources.5,6,9,16
Hepatotoxicity and nephrotoxicity are both well-documented effects of STZ.4,13-15 In addition to causing acidosis that can result from renal tubular damage in the kidney, STZ is one of the nitrosourea drugs or toxins known to cause type B lactic acidosis.17 Analysis of blood gases and chemistries is used to confirm and characterize acidosis, but blood volume requirements might unnecessarily compromise mice in an already weakened state. In the current study, adverse events commonly presented as respiratory distress, weight loss, and a generalized poor body condition (Table 2)—all conditions that are consistent with acidosis. This observation demonstrates the need in future studies to establish species appropriate test methods for monitoring acidosis severity and response to treatment strategies. In addition to manifesting these phenomena, JAX and TAC mice required significantly more insulin, as might be expected with acidosis-induced glucose intolerance and insulin resistance. We also noted severe hypoglycemia, which together with lactic acidosis suggests that the liver was injured to the extent that errors of metabolism involving exaggerated glycolysis resulted.
In conclusion, this comparative evaluation of 3 sources of congenitally athymic nude mice showed that CRL animals demonstrate a window between effective diabetes induction by STZ and complication rate that most easily supports generation of the diabetic nude mouse model. This window appears small for TAC and JAX nude mice. Investigations performed in the nude mouse bioassay constitute a necessary component of the quality assessment in administration of manufactured islet cell products to patients. We recommend performing exploratory safety assessments when selecting a nude mice source if the mice are to be used in the STZ-induced diabetes model. In this assessment, the type and frequency of adverse events should be characterized, with the aim of limiting morbidity and mortality to less than 10% in combination with successful induction of a diabetic state in at least 90% of animals. In a well-characterized model, the adverse events associated with the model are defined and subsequently unlikely to be inadvertently attributed to the investigational product. This characterization in combination with selection of an optimal source enhances animal wellbeing and reduces experimental data variability, presenting an opportunity for refinement.
Acknowledgments
We acknowledge with gratitude the excellent care of the animals by Elisabeth Steger, Theresa DuFour, James Munson, and Angela Craig. In particular, we thank Sandra Wagner and the Mouse Genetics Laboratory (Masonic Cancer Center, University of Minnesota) who provided valuable input and insight into supplier differences. Our study was supported by the Schulze Family Foundation, the National Institutes of Health, and the Juvenile Diabetes Research Foundation.
References
- 1.Animal Welfare Act as Amended. 2007.7 USC §2131-2159
- 2.Charles River Laboratories. [Internet] 2011. Immunodeficient animal models. [Cited 16 February 2011]. Available at: http://www.criver.com/sitecollectiondocuments/rm_rm_d_immunodeficient_models.pdf
- 3.DAIT, NIAID, NIH [Internet]. 2008. Purified human pancreatic islets, in vivo islets function. Document no. 3104, A04, effective date 07 July 2008. [Cited 16 February 2011]. Available at: http://www.isletstudy.org/CITDocs/3104,%20A04%20In%20Vivo%20Islets%20Function.pdf
- 4.Dufrane D, van Steenberghe M, Guiot Y, Goebbels RM, Saliez A, Gianello P. 2006. Streptozotocin-induced diabetes in large animals (pigs, primates): role of GLUT2 transporter and β-cell plasticity. Transplantation 81:36–45 [DOI] [PubMed] [Google Scholar]
- 5.Gonzalez C, Cuvellier S, Hue-Beauvais C, Lévi-Strauss M. 2003. Genetic control of nonobese diabetic mice susceptibility to high-dose streptozotocin-induced diabetes. Diabetologia 46:1291–1295 [DOI] [PubMed] [Google Scholar]
- 6.Hughes ED, Qu YY, Genik SJ, Lyons RH, Pacheco CD, Lieberman AP, Samuelson LC, Nasonkin IO, Camper SA, Van Keuren ML, Saunders TL. 2007. Genetic variation in C57BL/6 ES cell lines and genetic instability in the Bruce4 C57BL/6 ES cell line. Mamm Genome 18:549–558 [DOI] [PubMed] [Google Scholar]
- 7.Institute for Laboratory Animal Research 1996. Guide for the care and use of laboratory animals. Washington (DC): National Academies Press; [PubMed] [Google Scholar]
- 8.The Jackson Laboratory [Internet]. 2011. JAX mice database. [Cited 16 February 2011]. Available at: http://jaxmice.jax.org/strain/002019.html
- 9.Leiter EH. 1982. Multiple low-dose streptozotocin-induced hyperglycemia and insulitis in C57BL mice: Influence of inbred background, sex, and thymus. Proc Natl Acad Sci USA 79:630–634 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Lenzen S. 2008. The mechanisms of alloxan- and streptozotocin-induced diabetes. Diabetologia 51:216–226 [DOI] [PubMed] [Google Scholar]
- 11.Mouse Genome Informatics [Internet]. 2010. Guidelines for nomenclature of mouse and rat strains. [Cited 16 February 2011]. Available at: http://www.informatics.jax.org/mgihome/nomen/strains.shtml
- 12.Paik SG, Fleischer N, Shin SI. 1980. Insulin-dependent diabetes mellitus induced by subdiabetogenic doses of streptozotocin: obligatory role of cell-mediated autoimmune processes. Proc Natl Acad Sci USA 77:6129–6133 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Palm F, Ortsäter H, Hansell P, Liss P, Carlsson PO. 2004. Differentiating between effects of streptozotocin per se and subsequent hyperglycemia on renal function and metabolism in the streptozotocin-diabetic rat model. Diabetes Metab Res Rev 20:452–459 [DOI] [PubMed] [Google Scholar]
- 14.Schacht RG, Feiner HD, Gallo GR, Lieberman A, Baldwin DS. 1981. Nephrotoxicity of nitrosoureas. Cancer 48:1328–1334 [DOI] [PubMed] [Google Scholar]
- 15.Schein PS, Loftus S. 1968. Streptozotocin: depression of mouse liver pyridine nucleotides. Cancer Res 28:1501–1506 [PubMed] [Google Scholar]
- 16.Simpson EM, Linder C, Sargent E, Davisson M, Mobraaten L, Sharp J. 1997. Genetic variation among 129 substrains and its importance for targeted mutagenesis in mice. Nat Genet 16:19–27 [DOI] [PubMed] [Google Scholar]
- 17.Szerlip HM. 2009. Metabolic acidosis, p 78 : Greenberg A, Cheung AK. Primer on kidney diseases. Philadelphia (PA): Saunders [Google Scholar]
- 18.Taconic Farms [Internet]. 2011. Animal models. [Cited 16 February 2011]. Available at: http://www.taconic.com/wmspage.cfm?parm1=873
- 19.Van Belle TL, Taylor P, von Herrath MG. 2009. Mouse models for type 1 diabetes. Drug Discov Today Dis Models 6:41–45 [DOI] [PMC free article] [PubMed] [Google Scholar]