Welfare complications in the male BN/Crl and RjHan:SD rat streptozotocin-induced diabetes models

Abstract

German

Spanish

French

The streptozotocin (STZ)-induced hyperglycaemic rat model is widely used in diabetes research, particularly for investigating diabetic retinopathy; however, animal welfare concerns are often underreported. This study evaluated welfare outcomes in two commonly used rat lines, Brown Norway (BN/Crl) and Sprague Dawley (RjHan:SD). Data were collected from a series of diabetic retinopathy studies, in which a total of 183 BN/Crl and 76 RjHan:SD male rats received 40–65 mg/kg STZ via intraperitoneal injection and were monitored for 5–12 weeks. Welfare parameters, including body weight development and urologic complications (notably paraphimosis), were recorded and compared. Both lines achieved hyperglycaemia (≥16 mmol/l) within three weeks. However, BN/Crl rats exhibited a high incidence (82.5%) and severity of paraphimosis, along with marked weight loss, resulting in 13.7% of the animals reaching humane endpoint criteria for euthanasia. Weight loss positively correlated with STZ doses, with the highest dose (65 mg/kg) leading to 17.2% humane endpoint rate. In contrast, RjHan:SD rats exhibited significantly fewer urologic complications and maintained better weight gain, with none reaching humane endpoint. Our findings suggest that, while BN/Crl rats may offer advantages for ocular research owing to their pigmented eyes, their susceptibility to severe welfare issues raises concerns regarding their routine use. Furthermore, standardised supportive treatments, such as insulin supplementation, are worth considering for the model. This study highlights that careful selection of model animals, disease induction protocols and supportive treatments can optimise research outcomes and avoid loss of experimental animals, while adhering to the 3Rs.

Keywords

Laboratory animal welfare ethics and welfare refinement animal model organisms and models in vivo endocrinology physiology

Introduction

Streptozotocin (STZ)-induced hyperglycaemic rodent models are among the most widely used preclinical research tools for studying various diabetes mellitus-related conditions, particularly diabetic retinopathy.^1,2 STZ, a natural antibiotic produced by Streptomyces achromogenes,³ selectively binds to GLUT2 transporters found on the β cells of the islets of Langerhans in the mammalian pancreas.^3,4 This binding triggers DNA alkylation and cross-linking, leading to destruction of insulin-producing β cells and chronic hyperglycaemia.⁵ Since oestrogen protects against STZ-induced β-cell apoptosis,^6,7 male animals are typically preferred in this model.

In diabetic rats, chronic hyperglycaemia leads to various health and welfare concerns. These include gastrointestinal problems caused by disrupted gut motility and enlargement of intestinal components.^3,8,9 These issues lead to diarrhoea and constipation, as well as difficulty maintaining body weight.^3,10–12 Additionally, hyperglycaemia induces polyuria⁸ and associated urologic complications, which in male rats can lead to paraphimosis.^3,11 The STZ injection is also linked to acute mortality, and the risk positively correlates with the age of the animals at the time of induction.^8,12

Over time, STZ-induced hyperglycaemia leads to ophthalmic phenotype resembling diabetic retinopathy. This includes vascular changes such as neovascularization and vascular leakage,^13–18 as well as neuronal pathology characterised by apoptosis and functional deficits in multiple retinal cell types. These neuronal changes are seen, for example, in amacrine cells, retinal ganglion cells and bipolar cells.^{13–15,17,18} In addition to diabetic retinopathy, other ophthalmic complications, such as diabetic cataracts,⁸ develop. As diabetic retinopathy typically emerges after long exposure to hyperglycaemia, it is often necessary to maintain hyperglycaemic animals over extended periods.¹ The STZ model offers a relatively fast onset of hyperglycaemia in rats, typically within a few days following STZ injection, leading to significant diabetic retinopathy symptoms within just three months. This accelerated progression of diabetes and diabetic retinopathy symptoms enables researchers to shorten the follow-up period for experimental animals, aligning with the 3Rs framework outlined in Directive 2010/63/EU.

As the model causes moderate harm to the animals and cannot be replaced by non-animal research methods, it is important to continuously refine the methods. However, any refinement strategies must be carefully validated to avoid interfering with the diabetic retinopathy model development. One approach involves selecting the most suitable animals, as different rodent strains and stocks exhibit significant physiological and behavioural variation,^10,12,16 influencing phenotype development and the severity of side effects. In this manuscript, we use the term line to collectively refer to both strains and stocks. We propose that selecting an optimal line can serve as the primary refinement strategy in the diabetic retinopathy model. This study aimed to evaluate the welfare concerns associated with STZ-induced diabetes in male Sprague Dawley (RjHan:SD, Janvier Labs) and Brown Norway (BN/Crl, Strain#091, Charles River Laboratories) rats.

Materials and methods

Animals

Data were retrospectively collected from five independent diabetic retinopathy experiments performed at a private research organisation specialising in preclinical ophthalmic research during 2020–2022. A total of 96 outbred male Sprague Dawley rats (RjHan:SD, Janvier Labs, Genest-Saint-Isle, France) and 220 inbred male Brown Norway rats (BN/Crl, Charles River Laboratories, Sulzfeld, Germany) were used. Because of the retrospective nature of this study, a priori power analysis was not performed. At induction, animals were 6–10 weeks old and maintained under Specific Pathogen-Free conditions in accordance with FELASA recommendations.

Rats were pair-housed in individually ventilated cages (NexGen Rat 900, floor area 900 cm², Allentown LLC, Allentown, USA) with aspen bedding (Populus tremula, Tapvei Estonia OÜ, Paekna, Estonia). Environmental enrichment included nesting material, gnawing blocks (P. tremula, Tapvei Estonia OÜ) and polycarbonate red tubes (Datesand Limited, Bredbury, UK). Vivarium conditions were maintained at a temperature of 22°C ± 1°C, relative humidity of 50% ± 10%, and a 12-h light/dark cycle (lights on from 07:00 h to 19:00 h). Animals had ad libitum access to feed (Rat/Mouse maintenance V1534-000 or V1535-000, ssniff Spezialdiäten GmbH, Soest, Germany) and tap water (Kuopion Vesi Oy, Kuopio, Finland).

Rats were randomised into treatment groups at the cage level, ensuring that STZ-induced and healthy control animals were pair-housed separately to minimise intra-cage competition. All animals underwent a minimum of one week acclimatisation period before the experiments commenced.

Induction of hyperglycaemia and blood glucose monitoring

A total of 76 RjHan:SD rats and 183 BN/Crl rats were induced with 1–3 intraperitoneal (i.p.) injections of STZ in 10 mmol/l sodium citrate buffer (pH 4.5) at doses of 40, 45 or 65 mg/kg (Table 1). Control animals were either naïve or sham-induced, with the latter receiving only buffer via i.p. injection. Animals were fasted for 4 h (08:00–12:00 h) prior to induction and provided with 10% sucrose water to drink for two days post-induction. If an animal appeared apathetic one day after induction, 5% glucose solution (1 ml/rat) was administered subcutaneously (s.c.).

Table 1.

Summary of the experiments where BN/Crl and RjHan:SD rats were induced with streptozotocin (STZ).

Experiment ID	Rat line	Follow-up period (weeks)	Total number of animals	Number of STZ- induced animals	Number of naïve/sham animals	STZ dose (manufacturer^a)	Number of anaesthesia time points during follow-up	Age at induction (weeks)	Opportunistic bacteria status	Feed ssniff cat#
1. BN40	BN/Crl	12	33	27	6	40 mg/kg (CC)	4	9	Staphylococcus aureus+ Pseudomonas aeruginosa+ Proteus mirabilis+	V1535-000
2. BN45	BN/Crl	12	65	58	7	45 mg/kg (SA)	5	7–9	S. aureus+ P. aeruginosa+ P. mirabilis+	V1534-000
3. BN65	BN/Crl	5–9	122	98	24	65 mg/kg (SA)	5	9–10	S. aureus+ Klebsiella oxytoca+ P. aeruginosa+	V1535-000
4. SD45A	RjHan:SD	6–12	54	38	16	45 mg/kg (SA)	0–4	6–7	Negative	V1535-000
5. SD45B	RjHan:SD	12	42	38	4	45 mg/kg (SA)	0–1	6–7	Negative	V1535-000

Experiment IDs are coded based on the rat line and the STZ dose. The animals arriving from a barrier positive for opportunistic bacteria are indicated.

Manufacturers of the STZ: SA = Sigma-Aldrich, Saint Louis, USA (cat# S0130), CC = Cayman Chemicals, Ann Arbor, USA (cat#13104).

cat#: catalogue number.

Blood glucose levels were measured 2–4 days post-induction, and animals with blood glucose levels ≥16 mmol/l were considered diabetic. Those with lower blood glucose levels underwent re-induction. Blood glucose was subsequently monitored 1–2 times per week using either Accu-Chek Aviva (Hoffmann-La Roche, Basel, Switzerland) or AlphaTRAK blood glucose meters (Zoetis US LLC, Parsippany, USA), from femoral vein blood samples, between 08:00 h and 12:00 h.

Anaesthesia and ophthalmic phenotype follow-up

Anaesthesia was administered via s.c. injection of medetomidine hydrochloride (0.3 mg/kg) and ketamine hydrochloride (40 mg/kg) mixture (Cepetor Vet 1 mg/ml, CP-Pharma Handelsgesellschaft mbH, Burgdorf, Germany; Ketaminol Vet 50 mg/ml, Intervet International B.v., Boxmeer, Netherlands) and reversed with atipamezole (1.0 mg/kg s.c., Revertor Vet 5 mg/ml, CP-Pharma Handelsgesellschaft mbH, Burgdorf, Germany). Some animals received intravitreal injections for evaluating potential diabetic retinopathy treatments; however, no systemic treatment for diabetes was administered. Animals were anaesthetised for the treatments, or for ophthalmic imaging to monitor diabetic retinopathy development, 0–5 times per experiment.

Welfare monitoring and humane endpoints

Animals were monitored daily for 5–12 weeks post-induction or until a humane endpoint (HEP) was reached. The daily welfare assessments included activity levels, hydration status and paraphimosis scoring (Supplemental File 1). Body weight was measured at least weekly with more frequent monitoring for animals experiencing significant weight loss. Research staff were blinded to group allocation; however, blinding was partially compromised due to observable welfare concerns and reduced growth in STZ-induced animals.

If an animal developed paraphimosis, treatment involved lubricating the prolapsed penis with local analgesic gel (2% Xylocain, AstraZeneca UK Limited, Cambridge, UK) and gently repositioning it into the prepuce. The prevalence and severity of penile prolapse were scored as follows: 0: Healthy, no paraphimosis; 1: Mild, temporary paraphimosis; 2: Moderate, penis out of the prepuce without swelling; 3: Marked, penis completely exposed, swollen, red or purple in colour; 4: Marked paraphimosis with urine retention.

Animals experiencing dehydration and/or excessive weight loss were hydrated with Ringer-Lactate solution (s.c., 2–4 ml/rat, Animalcare Limited, York, UK). Open skin lesions or ulcerations were treated with disinfecting Vetericyn+ VF Wound & Skin Care Hydrogel (Innovacyn Inc., Rialto, USA). At the experimental endpoint or upon meeting HEP criteria, animals were humanely euthanised using an anaesthetic overdose followed by decapitation. The criteria for HEP are detailed in Table 2.¹⁹

Table 2.

Criteria for humane end points (HEPs) for streptozotocin-induced rats.

HEP	Description of criteria	Action
1. Marked paraphimosis	Marked, score 4, prolapse, where penis is completely out from the preputium, swollen, red or purple in colour, and urine retention	Immediate euthanasia
2. Marked body weight (BW) loss	25% BW loss from baseline. Animal is active and shows no other signs of discomfort^a	BW was followed for maximum five days^b, after which the animal was euthanised if the weight did not recover
3. General welfare reduction	Reduced welfare state including two or more of the following symptoms: dehydration, inactivity, 20% weight loss, and/or paraphimosis	Immediate euthanasia

The 25% body weight loss was exclusively accepted by the Project Authorisation Board of Finland to follow five days before euthanising the animals in this disease model, as the hyperglycaemia causes reduction in the nutrient absorption and thus fluctuations in the body weight. Defining too strict body weight-related HEPs could lead to unnecessary loss of experimental animals, as suggested by, for example, Talbot et al. (2020).¹⁹

During the five-day follow-up, feeding and drinking behaviours were confirmed by visual observations, and the animals received daily subcutaneous injections of balanced electrolyte solution (Ringer-Lactate).

Data collection

The following parameters were assessed to evaluate the diabetic phenotype and associated health complications: 1. Diabetic resilience: induction success rate, induction-related mortality, anaesthesia-related mortality, and the rate of HEPs; 2. Paraphimosis: prevalence and severity; 3. Body weight; and 4. Blood glucose.

Statistics

Data analysis and plotting were performed using GraphPad Prism software (Version 10, Dotmatics, Boston, USA). A p-value <0.05 was considered statistically significant. For statistical analysis, the data from experiments SD45A and SD45B were combined into a single dataset (SD45) owing to the same induction protocol, and the origin and induction age of the rats. Control animals from each line were grouped as BN Control (n = 37) and SD Control (n = 20).

Body weight was monitored for 5–12 weeks, 1–7 times per week, following induction of hyperglycaemia, and weekly averages from weeks 0–9 were used for the analysis. Body weight data were normalised to baseline, presented as mean ± standard deviation (SD). The data were analysed using a restricted maximum likelihood (REML) mixed effects model, and each treatment group was compared with the line untreated control with Šídák’s or Dunnett’s multiple comparisons tests.

Blood glucose was measured 1–2 times per week, and weekly averages were used for the analysis. In studies BN45 and BN40, the blood glucose meter used had a maximum limit of 33.3 mmol/l (Accu-Chek Aviva). In studies BN65 and SD45, the maximum limit was 41.7 mmol/l (AlphaTRAK). Statistical analyses were not performed between groups measured with different blood glucose meters. Residual plots were assessed for normality and homogeneity of variances. Data are presented as mean ± SD throughout.

Diabetes resilience and paraphimosis were analysed using either Fisher’s exact test or chi-squared test for categorical variables. HEP rates were compared between STZ-induced groups using Kaplan–Meier survival analysis with a log-rank Mantel–Cox test.

Results

Diabetes resilience

Two (2.0%) BN/Crl rats receiving the highest dose (BN65) died from the STZ induction, but no animals died from the lower doses (BN40 or BN45). The success rate for hyperglycaemia induction was slightly but not significantly higher in the BN/Crl strain (95.6%) compared with the RjHan:SD stock (90.7%) (Fisher’s exact test: p = 0.063, Table 3). While the dose of STZ and the manufacturer varied across studies in the BN/Crl group, these factors did not appear to affect the induction success rate. Anaesthesia-associated mortality was slightly higher in RjHan:SD rats compared with BN/Crl (4.0% vs. 0.5%). No correlations between mortality rates and number of anaesthesia sessions were observed.

Table 3.

The diabetes resilience of BN/Crl and RjHan:SD rats after streptozotocin (STZ) induction, assessed as the number of animals lost before experimental end point due to unsuccessful induction, death or from reaching a humane endpoint (HEP).

				Number of animals lost before experimental end point
Experiment ID	Number of control animals	Number of STZ-induced animals	Successful induction	Unsuccessful induction	Mortality, total	Mortality, anaesthesia	Mortality, STZ	HEP (from successfully induced)
1. BN40	6	27	26 (96.3%)	1 (3.7%)	1 (3.7%)	1 (3.7%)	0 (0.0%)	2 (7.7%)
2. BN45	7	58	56 (96.6%)	2 (3.4%)	0 (0.0%)	0 (0.0%)	0 (0.0%)	6 (10.7%)
3. BN65	24	98	93 (94.9%)	3 (3.1%)	2 (2.0%)	0 (0.0%)	2 (2.0%)	16 (17.2%)
BN total	37	183	175 (95.6%)	6 (3.3%)	3 (1.6%)	1 (0.5%)	2 (1.1%)	24 (13.7%)
4. SD45A	16	38	34 (89.5%)	4 (10.5%)	2 (5.3%)	2 (5.3%)^a	0 (0.0%)	0 (0.0%)
5. SD45B	4	38	35 (92.1%)	3 (7.9%)	1 (2.6%)	1 (2.6%)	0 (0.0%)	0 (0.0%)
SD total (SD45)	20	76	69 (90.7%)	7 (9.2%)	3 (4.0%)	3 (4.0%)	0 (0.0%)	0 (0.0%)

One SD45A rat died under anaesthesia at baseline imaging (not STZ-induced).

A total of 13.7% of BN/Crl rats were euthanised owing to reaching HEP before the experimental endpoint. In contrast, no RjHan:SD rats reached HEP. The primary causes of HEPs in BN/Crl rats included severe (score 4) paraphimosis leading to urine retention (n = 4), extreme (>25%) body weight loss (n = 11), and overall welfare deterioration, including symptoms such as hypothermia, dehydration, inactivity, >20% weight loss, and/or score 3 paraphimosis (n = 9). The HEP rate in BN/Crl rats increased with the STZ dose: 7.7% (2/26) for BN40, 10.7% (6/56) for BN45 and 17.2% (16/93) for BN65. Kaplan–Meier survival analysis revealed a significant negative effect of STZ dose on survival (log-rank Mantel–Cox test: p = 0.006, Figure 1).

Figure 1.

Survival of BN/Crl rats exposed to different doses of streptozotocin (STZ), assessed from humane endpoint rates. STZ dose at induction significantly affected the survival (log-rank Mantel–Cox test: p = 0.006). A dose-dependent effect was observed (log-rank test for trend: p < 0.001). Animals were followed for nine (BN65) or 12 weeks (BN40, BN45).

Paraphimosis

Paraphimosis prevalence was significantly higher (chi-squared test: p < 0.001, Table 4) in the BN/Crl (82.5%, 146/177) compared with the RjHan:SD rats (7.2%, 5/69). In BN/Crl rats, the prevalence of paraphimosis varied with STZ dose: 73.0% in BN40, 85.7% in BN45 and 83.1% in BN65, although a dose-dependent trend could not be substantiated. Furthermore, the severity of paraphimosis was greater in BN/Crl rats, with more than half (51.5%) of affected animals showing a score of 3, and 26.6% a score of 2. In contrast, RjHan:SD rats with paraphimosis predominantly had a score of 1 (4.3%) or 2 (2.9%). Only successfully induced rats were included into the comparison, as no paraphimosis cases were observed in healthy controls or unsuccessfully induced rats in either line.

Table 4.

The occurrence and severity of paraphimosis in the streptozotocin-induced rats, presented as numbers per group (n) and as fractions of row total (prevalence, %).

		Paraphimosis prevalence
		No	Ye s	Severity
		0	1–4	1	2	3	4	Total
Group
BN40	n	7	19	2	13	2	2	26
	%	(26.9)	(73.0)	(7.7)	(50.0)	(7.7)	(7.7)	(100)
BN45	n	8	48	2	19	25	2	56
	%	(14.3)	(85.7)	(3.6)	(33.9)	(44.6)	(3.6)	(100)
BN65	n	16	79	0	15	64	0	95
	%	(16.8)	(83.1)	(0.0)	(15.8)	(67.4)	(0.0)	(100)
BN total	n	31***	146***	4	47	91	4	177
	%	(17.5)	(82.5)	(2.3)	(26.6)	(51.5)	(2.3)	(100)
SD45	n	64***	5***	3	2	0	0	69
	%	(92.8)	(7.2)	(4.3)	(2.9)	(0.0)	(0.0)	(100)
Total	n total	96	151	7	49	91	4	247

The prevalence of paraphimosis was significantly higher in the BN/Crl rats compared with the RjHan:SD rats. Chi-squared test: $χ_{1}^{2}$ = 119.8, ***p < 0.001.

Body weight

At baseline, the BN/Crl rats (n = 225) weighed 217.1 ± 40.1 g, and the RjHan:SD rats (n = 96) weighed 318.3 ± 17.6 g. Because of age differences at induction, baseline weights for BN groups varied: 226.0 ± 42.0 g (BN Control, 7–10 weeks of age), 205.8 ± 15.5 g (BN40, nine weeks), 167.1 ± 15.5 g (BN45, 7–9 weeks) and 246.4 ± 18.9 g (BN65, 9–10 weeks). Consequently, the body weight of each animal was normalised to baseline. Rats in the healthy control groups of both lines continued to grow as expected throughout the study (Figure 2(a) and (b)).

Figure 2.

Body weight change (%) and blood glucose (mmol/l) from streptozotocin (STZ)-induced BN/Crl and RjHan:SD rats. Data are presented as mean ± SD. In panels (c) and (d), the dashed lines mark the hyperglycaemia threshold of 16 mmol/l, and the dotted lines mark the upper limits of the blood glucose meters used for each group, 33.3 and 41.7 mmol/l. (a) Body weight of BN/Crl rats was significantly reduced in all STZ-induced groups (restricted maximum likelihood (REML) mixed-effects model, all: Time, Treatment, and Time × Treatment, p < 0.001). Pairwise comparisons revealed significant differences between BN Control and all STZ-induced groups from week 4 onwards (Dunnett’s multiple comparisons test: BN Control vs. BN40, p < 0.001; BN Control vs. BN45, p < 0.01; BN Control vs. BN65, p < 0.001, three comparisons in 10 families, n = 26–94 rats/group per time point). (b). Body weight of RjHan:SD rats was significantly reduced in the STZ-induced group (REML mixed-effects model, all: Time, Treatment, Time × Treatment, p < 0.001). Pairwise comparisons showed significant difference from week 1 onwards (Šídák’s multiple comparisons test: SD Control vs. SD45, p < 0.001, n = 8–69 rats/group per time point). (c) Blood glucose of all BN/Crl rats reached the 16.0 mmol/l hyperglycaemia threshold within three weeks of induction. No statistical analysis was performed owing to different upper limits in the blood glucose meters. Data are presented from 19–93 rats/group per time point. (d) Blood glucose of all RjHan:SD rats reached the 16.0 mmol/l hyperglycaemia threshold within two weeks of induction. Significant differences between STZ-induced and control animals were observed starting one week post-induction (REML mixed-effects model with Šídák’s multiple comparisons test: SD Control vs. SD45, p < 0.001, 10 comparisons, n = 8–69 rats/group per time point).

In all STZ-induced BN/Crl rats, growth was significantly reduced compared with healthy controls (REML mixed-effects model, all: Time, Treatment, and Time × Treatment, p < 0.001, Figure 2(a) and (b)). Rats induced with 40 mg/kg or 45 mg/kg of STZ showed only minimal weight gain until week 5 post-hyperglycaemia onset, after which the body weight began to decrease. By week 9, the body weight in these groups was almost the same as at baseline, with 2.7% weight loss in BN40 rats, and a minimal 2.4% weight gain in BN45 rats. In contrast, rats induced with the highest STZ dose, 65 mg/kg, had lost on average 15.3% of their baseline body weight by week 9. The weekly body weight across all BN/Crl STZ-induced groups was significantly lower than in healthy controls from week 4 onwards (Dunnett’s multiple comparisons test: p < 0.01, Figure 2(a)).

Similarly, the growth of STZ-induced RjHan:SD rats was significantly reduced compared with the healthy controls (REML mixed-effects model, all: Time, Treatment, and Time × Treatment, p < 0.001, Figure 2(b)). The difference between STZ-induced and healthy control RjHan:SD rats became statistically significant starting one week post-induction (Šídák's multiple comparisons test: SD Control vs. SD45, p < 0.001). Despite the reduced growth, the SD45 group showed steady weight gain of 34.7% by week 9.

Blood glucose

In all experiments, the STZ-induced groups reached the 16.0 mmol/l blood glucose threshold within three weeks of induction (Figure 1). At week 9, the blood glucose averages for the BN/Crl rat groups were 30.8 mmol/l ± 2.8 mmol/l (n = 25) for BN40, 32.8 mmol/l ± 1.1 mmol/l (n = 52) for BN45, and 35.9 mmol/l ± 4.4 mmol/l (n = 64) for BN65, while the control group had a blood glucose of 5.3 mmol/l ± 0.6 mmol/l (n = 31, Figure 2(c)). No statistical analysis was performed between BN groups owing to different blood glucose measurement methods. For the RjHan:SD rat groups, the week 9 blood glucose averages were 30.9 mmol/l ±6.5 mmol/l (n = 19) for SD45 and 6.4 mmol/l ±0.8 mmol/l (n = 8) for SD Control (Figure 2(d)). The difference between STZ-induced and healthy control RjHan:SD rats was statistically significant starting one week post-induction (REML mixed-effects model with Šídák's multiple comparisons test: SD Control vs. SD45, p < 0.001).

Discussion

Despite its widespread use, the STZ-induced diabetic rat model is poorly characterised in terms of animal welfare. In this study we assessed the health complications in two rat lines, Brown Norway (BN/Crl) and Sprague Dawley (RjHan:SD), widely used in diabetic retinopathy research. Hyperglycaemia (blood glucose ≥16.0 mmol/l) successfully developed in both lines within three weeks, aligning with previous studies,^10,18 and led to major diabetic retinopathy symptoms, such as retinal vascular leakage and retinal functional impairment, within the follow-up period (data not presented here). In addition, in both lines, diabetic cataracts developed similarly. However, distinct line-specific welfare outcomes were observed, with BN/Crl rats experiencing significantly more complications.

Induction-related mortality was observed in BN/Crl rats at the 65 mg/kg STZ dose, whereas no mortality occurred at lower doses. This suggests that the induction ages and STZ doses used in these experiments were appropriate, and that the post-induction supportive care (e.g. sucrose water, glucose injections) was effective. Since RjHan:SD rats were tested only at 45 mg/kg, the lower mortality may reflect either line-specific resilience or the lower STZ dose.

Mortality associated with the medetomidine–ketamine anaesthesia protocol was pronounced especially in RjHan:SD rats. This highlights the need to refine anaesthetic regimens to minimise unnecessary loss of experimental animals in this model. Issues with ketamine-α2-agonists producing inadequate anaesthesia depth in STZ-treated rats has previously been documented.¹⁸ We observed a similar phenomenon, where hyperglycaemic rats required higher anaesthetic doses, increasing the risk of mortality. In healthy rodents, α2-agonists are known to cause transient hyperglycaemia, but this effect is minimal in chronically hyperglycaemic animals.¹⁸ The combination is also associated with prolonged recovery.²⁰ A promising alternative with fewer side effects is isoflurane, although in these experiments inhalation anaesthesia was impractical owing to combining multiple in vivo imaging methods at one timepoint. Future studies should explore whether alternative injectable anaesthesia combinations produce a different outcome in diabetic rodents.

STZ induction reduced the growth rate in both rat lines, consistent with previous findings.^3,10 After the induction, BN/Crl rats exhibited a complete cessation of growth, with those in the 65 mg/kg STZ dose group being particularly susceptible to body weight loss. Brown Norway rats are naturally small and grow slowly.²¹ Moreover, hyperglycaemia and insulin deficiency are linked to reduced growth rate not only in rats but also in diabetic children,²² because of, for example, altered production and function of growth hormones.²³ Therefore, future research could benefit from rat body condition scoring²⁴ to better distinguish between pathological malnutrition and merely small body size.

One of the most concerning findings was the high prevalence of urologic complications, specifically paraphimosis in the male BN/Crl rats. These issues developed within weeks of STZ induction, and occurred with all STZ doses, suggesting a possible underlying genetic predisposition. The severity and irreversibility of the paraphimosis raises concerns about the suitability of BN/Crl rats for STZ-induced hyperglycaemia model. The aetiology of paraphimosis is not very thoroughly studied in rats. However, in human patients with diabetes, polyuria causes dehydration and glucose accumulation in the preputial area, leading to tissue irritation and chronic inflammation.²⁵ This results in oedema, fibrosis and impaired lymphatic drainage, contributing to the development of paraphimosis. Various urologic complications, such as spontaneous bladder and ureter carcinomas, have been documented in other Brown Norway substrains, such as BN/OrlRj,²⁶ although they have not been linked to the BN/Crl substrain. We hypothesise that inbreeding-associated genetic factors may underlie latent urinary dysfunction in this strain, which is then revealed by the chronic hyperglycaemia-induced polyuria.

One limitation of this study was the health status of the Brown Norway rats, which were sourced from a facility that tested positive for opportunistic pathogens, including Staphylococcus aureus, Pseudomonas aeruginosa, Proteus mirabilis and Klebsiella oxytoca, whereas the Sprague Dawley rats had specific pathogen and opportunistic pathogen free health status. Given that diabetic animals are immunosuppressed and more prone to infections,²⁷ the opportunistic bacteria status may have been a significant variable in this model. Especially P. mirabilis and K. oxytoca have the potential to cause urinary tract infections in rodents,^28,29 and theoretically could enhance the development of paraphimosis. However, the opportunistic bacteria status of the animals included in our data was not confirmed, and the pathogens have not, to our knowledge, been previously associated with paraphimosis. Therefore, future experiments are needed on the aetiology of this complex urogenital complication.

In contrast, Sprague Dawley rats exhibited greater resilience to hyperglycaemia, with fewer adverse effects. Some individuals developed mild and reversible prolapse, but none required euthanasia. Their body weight growth remained relatively stable despite the hyperglycaemia. This resilience suggests that Sprague Dawley rats may be better suited for STZ-induced diabetes research. However, using an outbred line such as Sprague Dawley introduces the potential for increased variability in study outcomes.^30–32 Inbred strains, such as Brown Norway, offer greater genetic uniformity, leading to more consistent results within a specific laboratory or experimental setting. On the other hand, reliance on a single inbred line puts researchers at risk of committing the so-called ‘standardization fallacy’.³³ Excessive standardisation leads to an increased risk that results from in the highly controlled condition may not apply in other contexts, for example, when we attempt to translate them to the human population. Furthermore, the limited genetic diversity of inbred strains makes them more susceptible to strain-specific health issues. One potential solution to these concerns is to replicate studies using multiple inbred strains, rather than relying on a single line.³²

Another key difference between the two rat lines is their ocular characteristics. Brown Norway rats have pigmented eyes more closely resembling human eyes, whereas the albino eyes of Sprague Dawley rats have some distinct properties that should be considered in diabetic retinopathy research. These include different retinal pigment epithelium and photoreceptor structures, along with delayed neural retina development.^34–36 Moreover, symptoms of diabetic retinopathy, including presence of inflammation in the retina, and vascular abnormalities develop slightly differently across lines.^10,16

Given that this comparison was limited to two distinct rat lines, future research should explore alternative strains or stocks that might offer better welfare outcomes in this model. For studies requiring pigmented eyes, it might be valuable to investigate whether other commonly used pigmented rat lines, such as Long Evans rats,^10,37 demonstrate greater diabetes resilience than Brown Norway rats. Moreover, inclusion of female rats in future studies is recommended, since it would account for sex-specific differences in diabetes progression and treatment responses. However, since female rats exhibit a distinct phenotype after STZ induction,^38,39 varying ocular phenotypes and welfare outcomes may be expected. The use of females could provide information on the origin of the hyperglycaemia-driven urologic complications, although the harm and outcome should be evaluated with caution owing to the anatomical differences between the sexes.

There are alternative rodent models that could be less invasive and produce fewer welfare problems than the STZ-induced diabetic retinopathy. For example, high-fat-diet-induced diabetes or genetically modified diabetic rodent strains (e.g. Ins2Akita, or BKS db/db mice, The Jackson Laboratory, strains #003548 and #000697).¹ However, these alternatives tend to develop the diabetic retinopathy phenotype more slowly, in 4–16 months, which prolongs the experiments.¹ Additionally, non-hyperglycaemic retinopathy models such as oxygen-induced retinopathy can help study vascular pathology, but owing to different aetiology, these models do not reflect the full diabetic retinopathy phenotype.^1,16,40

Finally, implementation of additional refinement approaches, such as insulin supplementation,^9,17,41 is worth considering. It could reduce the overall severity of the disease and lower the HEP rate. However, thorough line-specific validation for insulin dosing is important, as lowering blood glucose levels might slow or completely inhibit the progression of diabetic retinopathy, thereby prolonging the follow-up time necessary for these studies.

Conclusion

This study highlights line-specific differences in the welfare outcomes of STZ-induced hyperglycaemic male rats. While both rat lines, inbred BN/Crl and outbred RjHan:SD, developed hyperglycaemia similarly, BN/Crl rats exhibited severe welfare issues, including painful urological complications and significant weight loss, making them less suitable for diabetes research. In contrast, Sprague Dawley rats showed greater diabetes resilience, with fewer complications and more stable growth, making them the preferable option when pigmentation is not a critical factor.

Future studies should focus on alternative rat lines that could offer better resilience against hyperglycaemia-induced welfare issues while maintaining the scientific validity of diabetes research. Additionally, further optimised refinement strategies, including alternative anaesthesia protocols, and insulin supplementation, could be explored further to reduce suffering and unnecessary loss of animals. The careful selection of animal models and refinement strategies not only contributes to scientific validity but also aligns with the ethical framework of the 3Rs, emphasising the responsibility of researchers to minimise animal suffering.

Footnotes

Acknowledgements

We would like to thank the technical staff and animal caregivers for their invaluable contributions to this study. Special thanks to Pinja Mertano, Anni Kolehmainen, Aino Mering, Piia Pietikäinen and Oliver Mehtovuori for their efforts in maintaining the well-being of the animals throughout the study. We also appreciate the work of Teemu Nevalainen, Ville Jokinen, Anne-Mari Haapaniemi and Maria Vähätupa for their assistance in experimental procedures and data collection.

Data availability

Data are not publicly available.

Declaration of conflicting interests

The authors have no conflicts of interest to declare.

Ethical review

Data were collected from five independent experiments conducted at Experimentica Ltd (Kuopio, Finland) between 2020 and 2022. All procedures complied with EC Directive 2010/63/EU on the protection of animals used for scientific purposes and the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. The protocols were approved and monitored by the Project Authorisation Board of Finland (project licence number ESAVI-9520-2020).

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

ORCID iDs

Anni Tenhunen

Heidi Isohookana

Supplementary material

Supplementary material for this article is available online.

References

Lai

AKW

ACY.

Animal models of diabetic retinopathy: Summary and comparison. J Diabetes Res 2013; 2013: 1–29.

Lelyte

Ahmed

Kaja

, et al. Structure–function relationships in the rodent Streptozotocin-induced model for diabetic retinopathy: A systematic review. J Ocul Pharmacol Ther 2022; 38: 271–286.

Fox

Eastwood

Gentry

, et al. Critical evaluation of the streptozotocin model of painful diabetic neuropathy in the rat. Pain 1999; 81: 307–316.

Wang

Gleichmann

GLUT2 in pancreatic islets: Crucial target molecule in diabetes induced with multiple low doses of streptozotocin in mice. Diabetes 1998; 47: 50–56.

Elsner

Guldbakke

Tiedge

, et al. Relative importance of transport and alkylation for pancreatic beta-cell toxicity of streptozotocin. Diabetologia 2000; 43: 1528–1533.

Paik

S-G

Michelis

Kim

, et al. Induction of insulin-dependent diabetes by streptozotocin inhibition by estrogens and potentiation by androgens. Diabetes 1982; 31: 724–729.

Le May

Chu

, et al. Estrogens protect pancreatic β-cells from apoptosis and prevent insulin-deficient diabetes mellitus in mice. Proc Natl Acad Sci U S A 2006; 103: 9232–9237.

Wang-Fischer

Garyantes

Improving the reliability and utility of streptozotocin-induced rat diabetic model. J Diabetes Res 2018; 2018: 1–14.

Izbéki

Wittman

Rosztóczy

, et al. Immediate insulin treatment prevents gut motility alterations and loss of nitrergic neurons in the ileum and colon of rats with streptozotocin-induced diabetes. Diabetes Res Clin Pract 2008; 80: 192–198.

10.

Kirwin

Kanaly

Linke

, et al. Strain-dependent increases in retinal inflammatory proteins and photoreceptor FGF-2 expression in streptozotocin-induced diabetic rats. Invest Opthalmol Vis Sci 2009; 50: 5396.

11.

Qayoom

Aneesha

Anagha

, et al. Lecithin-based deferoxamine nanoparticles accelerated cutaneous wound healing in diabetic rats. Eur J Pharmacol 2019; 858: 172478.

12.

Deeds

Anderson

Armstrong

, et al. Single dose streptozotocin-induced diabetes: Considerations for study design in islet transplantation models. Lab Anim 2011; 45: 131–140.

13.

Gong

C-Y

Q-W

, et al. Streptozotocin induced diabetic retinopathy in rat and the expression of vascular endothelial growth factor and its receptor. Int J Ophthalmol 2013; 6: 573–577.

14.

Naderi

Zahed

Aghajanpour

, et al. Long term features of diabetic retinopathy in streptozotocin-induced diabetic Wistar rats. Exp Eye Res 2019; 184: 213–220.

15.

Park

H-YL

Kim

Park

CK.

Neuronal cell death in the inner retina and the influence of vascular endothelial growth factor inhibition in a diabetic rat model. Am J Pathol 2014; 184: 1752–1762.

16.

Zhang

Sima

, et al. Genetic difference in susceptibility to the blood–retina barrier breakdown in diabetes and oxygen-induced retinopathy. Am J Pathol 2005; 166: 313–321.

17.

Kohzaki

Vingrys

Bui

BV.

Early inner retinal dysfunction in streptozotocin-induced diabetic rats. Invest Opthalmol Vis Sci 2008; 49: 3595.

18.

Connell

Hookham

, et al. Comparisons of α2-adrenergic agents, medetomidine and xylazine, with pentobarbital for anesthesia: Important pitfalls in diabetic and nondiabetic rats. J Ocul Pharmacol Ther 2022; 38: 156–166.

19.

Talbot

Biernot

Bleich

, et al. Defining body-weight reduction as a humane endpoint: A critical appraisal. Lab Anim 2020; 54: 99–110.

20.

Giroux

M-C

Hélie

Burns

, et al. Anesthetic and pathological changes following high doses of ketamine and xylazine in Sprague Dawley rats. Exp Anim 2015; 64: 253–260.

21.

Charles River Laboratories. Brown Norway rat: BN/Crl, https://www.criver.com/products-services/find-model/brown-norway-rat?region=3616 (accessed 28 February 2025).

22.

Mitchell

DM.

Growth in patients with type 1 diabetes. Curr Opin Endocrinol Diabetes Obes 2017; 24: 67–72.

23.

Carlsson

LMS

Clark

Skottner

, et al. Growth hormone and growth in diabetic rats: Effects of insulin and insulin-like growth factor-I infusions. J Endocrinol 1989; 122: 661–670.

24.

Hickman

Swan

Use of a body condition score technique to assess health status in a rat model of polycystic kidney disease. J Am Assoc Lab Anim Sci. 2010; 49: 155–9.

25.

Golbidi

Laher

Bladder dysfunction in diabetes mellitus. Front Pharmacol 2010; 1; 136. DOI: 10.3389/fphar.2010.00136.

26.

Janvier Labs. Brown Norway rat, https://janvier-labs.com/en/fiche_produit/brown-norway_rat/ (accessed 28 February 2025).

27.

Muller

Golshayan

Ehirchiou

, et al. Immunosuppressive effects of streptozotocin-induced diabetes result in absolute lymphopenia and a relative increase of T regulatory cells. Diabetes 2011; 60: 2331–2340.

28.

Armbruster

Smith

Johnson

, et al. The pathogenic potential of proteus mirabilis is enhanced by other uropathogens during polymicrobial urinary tract infection. Infect Immun 2017; 85: e00808–16.

29.

Bleich

Kirsch

Sahly

, et al. Klebsiella oxytoca: Opportunistic infections in laboratory rodents. Lab Anim 2008; 42: 369–375.

30.

Brekke

Steele

Mulley

JF.

Inbred or outbred? Genetic diversity in laboratory rodent colonies. G3 (Bethesda) 2018; 8: 679–686.

31.

Festing

MFW.

Inbred strains should replace outbred stocks in toxicology, safety testing, and drug development. Toxicol Pathol 2010; 38: 681–690.

32.

Tuttle

Philip

Chesler

, et al. Comparing phenotypic variation between inbred and outbred mice. Nat Methods 2018; 15: 994–996.

33.

Voelkl

Würbel

Krzywinski

, et al. The standardization fallacy. Nat Methods 2021; 18: 5–7.

34.

Creel

Inappropriate use of albino animals as models in research. Pharmacol Biochem Behav 1980; 12: 969–977.

35.

Gurdita

Tan

Joos

, et al. Pigmented and albino rats differ in their responses to moderate, acute and reversible intraocular pressure elevation. Doc Ophthalmol 2017; 134: 205–219.

36.

Weidner

[The presence of an albino ERG in the pigmented rat: Genetic implications. (author’s transl)]. J Physiol (Paris) 1981; 77: 813–821.

37.

Aung

Kim

Olson

, et al. Early visual deficits in streptozotocin-induced diabetic long evans rats. Invest Ophthalmol Vis Sci 2013; 54: 1370–1377.

38.

Yamashita

Sugihara

Yamada

, et al. Effect of estrogen on electroretinographic responses in streptozotocin-induced diabetic female rats. Exp Eye Res 2010; 90: 591–597.

39.

Ramsey

Ripps

Qian

An electrophysiological study of retinal function in the diabetic female rat. Invest Opthalmol Vis Sci 2006; 47: 5116.

40.

Vähätupa

Jääskeläinen

Cerrada-Gimenez

, et al. Oxygen-induced retinopathy model for ischemic retinal diseases in rodents. J Vis Exp 2020; 16: e61482.

41.

Pinheiro

de Melo

Andreazzi

AE.

Protocol of insulin therapy for streptozotocin-diabetic rats based on a study of food ingestion and glycemic variation. Scand J Lab Anim Sci 2011; 38: 117–127.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.26 MB