Inducing stroke in aged, hypertensive, diabetic rats

Abstract

Animal models of ischemic stroke often neglect comorbidities common in patients. This study shows the feasibility of inducing stroke by 2 h of thread occlusion of the middle cerebral artery in aged (56 week old) spontaneously hypertensive rats (SHRs) with both acute (2 weeks) and chronic (36 weeks) diabetes. After modifying the streptozotocin dosing regimen to ensure that old SHRs survived the induction of diabetes, few died after induction of stroke. Induction of stroke is feasible in rats with multiple comorbidities. Inclusion of such comorbid animals may improve translation from the research laboratory to the clinic.

Introduction

Preclinical testing of candidate stroke therapies rarely considers the risk factors common in diseases of the elderly. Hypertension, present in >50% of stroke patients, increases stroke risk by 20 to 30% for a 10 mm Hg increase in arterial blood pressure (Alberts and Atkinson, 2004). Diabetes or acute hyperglycemia is evident in 25 and 40% of stroke patients, respectively, (Kaarisalo et al, 2005; Williams et al, 2002) and is associated with poor outcome (Poppe et al, 2009). Hypertension is an additional comorbidity in 55% of diabetic patients (Kaarisalo et al, 2005). Importantly, better diabetic control and blood pressure reduction both reduce stroke risk (Gaede et al, 2003). Age itself is another powerful risk factor for stroke (Frost et al, 2007), which is only rarely considered in stroke modeling (Rosen et al, 2005).

In this study, we sought to establish the feasibility of modeling stroke in aged hypertensive diabetic rats.

Materials and methods

Animals

All procedures were carried out in accordance with institutional and national guidelines. A total of 36 male spontaneously hypertensive rats (SHRs) (12 weeks) were randomized to nondiabetic, acute and chronic diabetic groups. A cohort of 12 age-matched nonhypertensive, nondiabetic Wistar-Kyoto (WKY) rats were used as controls. Although blinded induction of ischemia was part of the experimental design, reduced body weights in diabetic animals made this impossible to maintain. Infarct analysis was blinded by recoding tissue upon collection.

Induction of Diabetes

Chronic diabetes (Bell and Hye, 1983) was induced at 16 weeks and maintained for 36 weeks. After overnight fasting, streptozotocin (STZ; 50 mg/kg, dissolved in 0.1 mol/L sodium citrate buffer, pH=4.5 (Sigma, Castle Hill, NSW, Australia)) was injected through the tail vein. Acute diabetes was induced at 54 weeks and maintained for 2 weeks by daily tail vein injection of 10 mg/kg STZ until consistent hyperglycemia was achieved. Once diabetes was confirmed (blood glucose >15 mmol/L 2 days after injection) diabetic animals (acute and chronic) were administered 3-Units of human Protaphane insulin (NovoNordisk, Baulkham Hills, NSW, Australia) subcutaneously daily to maintain health. Nondiabetic controls (WKY and SHR) received tail vein injection of 0.1 mol/L sodium citrate buffer at 16 weeks of age.

Induction of Stroke

Middle cerebral artery occlusion (MCAo) was performed at 56 weeks of age in all animals. Anesthesia was induced (5%) and maintained (2%) by spontaneous breathing of isofluorane (50:50 O₂:air) through a nose cone and maintained until the occluding thread had been inserted and resistance felt. All incisions were closed using a silk suture, and animals were allowed to recover from anesthesia during the 2-h occlusion period. This allowed for neurologic deficit to be assessed just before reperfusion. During surgery, heart rate and pO₂ were monitored and rectal temperature was regulated to 37.4°C. Stroke was induced by 2-h transient MCAo using the thread occlusion method proposed by Longa et al (1989) as modified by Spratt et al (2006). A silicon-coated suture (coating diameter and length 0.35 and 5 mm, respectively) was maneuvered through the external and internal carotid arteries to block the MCA. After 2 h, the animal was reanesthetized and the thread was withdrawn into the external carotid stump, sealed in place and the surplus trimmed.

Assessments

Neurologic deficit was assessed at 2 h (immediately before reperfusion) and at 24 h after MCAo using a modification of the procedure described by Petullo et al (1999). Twenty-four hours after MCAo, rats were killed by isoflurane overdose (achieved by saturating the air in a ‘knock-down box' using undiluted isoflurane soaked in cotton wool) and perfused with 4% paraformaldehyde before processing for hematoxylin and eosin staining to delineate infarction at eight coronal planes. The infarct volume was corrected for edema (infarct volume × contralateral hemisphere/ipsilateral hemisphere volume) and expressed as a proportion of contralateral hemispheric volume to correct for differences in brain size resulting from age or diabetes. Additional measures of body weight, systolic blood pressure (by tail-cuff method), blood glucose, and rectal temperature were made shortly before surgery for stroke induction.

Data Analysis

All data are presented as mean±s.d. Analysis of variance with Tukey's post hoc error correction was used to determine the statistical significance of differences between groups.

Results

Mortality

Mortality during follow-up to 56 weeks of age was low. Three animals died due to unknown reasons late in the experiment but before diabetic or surgical intervention. All rats survived induction of diabetes at 16 weeks but 1 died during the chronic treatment to 56 weeks. When acute diabetes was induced in old age with a single bolus of STZ, the first two animals died. After dividing the 50 mg/kg bolus into 10 mg/kg doses administered daily through the tail vein until stable hyperglycemia ensued, no further deaths were encountered. All nondiabetic control animals (12/12 WKY and 11/11 SHR) survived for 24 h after MCAo. In the SHR-acute diabetic and SHR-chronic diabetic groups, 7 of 8 and 8 of 11 survived MCAo, respectively. Six brains (four WKY control, one SHR-acute diabetic and one SHR-chronic diabetic) were irreparably damaged during perfusion and fixation and could not be used for infarct volume determination (Table 1a).

Table 1. Experimental mortality, body weight, physiological variables and infarct volume.

(a) Experimental mortality
Group	Initial number	Incidental deatha	Age at diabetes induction	STZ diabetes	Post-STZ loss	tMCAo (56 weeks)	Poststroke death	Technical lossb	Infarct analysis
WKY nondiabetic	12	0	16 weeks	Citrate buffer	—	12	0	4	8
SHR-nondiabetic	12	1	16 weeks	Citrate buffer	—	11	0	0	11
SHR-acute diabetic	12	2	54 weeks	10	2	8	1	1	6
SHR-chronic diabetic	12	0	16 weeks	12	1	11	3	1	7

(b) Body weight, infarct volume, and physiologic variables at/after MCAo
Measurement	Time	WKY Nondiabetic (56 weeks) n=8	SHR Nondiabetic (56 weeks) n=11	SHR acute diabetic (56 weeks) n=6	SHR-chronic diabetic (56 weeks) n=7
Body weight (g)	<30 mins before anesthesia	516±37	495±21	446±23 **	366±25 **
Contralateral hemisphere volume (mm3)	24 h after tMCAo	438±32	398±28	433±9	376±33
Edema-corrected infarct volume (mm3)	24 h after tMCAo	81±22	139.3±22*	129.8±42	175.0±51
Plasma glucose (mmol/L)	<30 mins before anesthesia	8±2	8±2	29±5**	34±10**
Blood pressure (mm Hg)	<30 mins before anesthesia	124±9*	202±6	212±7**	213±5**
Rectal temperature
(°C)	<30 mins before anesthesia	35.9±0.8*	37.5±0.9	36.2±1**	35.6±0.8**
	24 h after tMCAo	35.2±0.8	36.7±1.1	33.8±1.8**	35.5±2.1

MCAo, middle cerebral artery occlusion; SHR, spontaneously hypertensive rat; tMCAo, transient middle cerebral artery occlusion; WKY, Wistar-Kyoto.

All values reported as mean±1 s.d.

^*P<0.05, WKY control versus SHR nondiabetic controls, ^**P<0.05, SHR nondiabetic controls versus SHR diabetic.

^aIncidental losses included two SHRs that died due to unknown reasons before STZ treatment and one nondiabetic SHR that died after induction of diabetes but before stroke.

^bTechnical problems with tissue processing made tissue unsuitable for infarct analysis.

Presurgical Impact of Diabetes in Hypertensive Rats

Both cohorts of diabetic animals maintained plasma glucose concentrations approximately fourfold higher than did control animals (P<0.05) (Table 1b). In the SHR-chronic diabetic group, plasma glucose was relatively stable for the duration of the experiment. This was accompanied by an initial marked reduction in body weight, which led to differences in body weight compared with nondiabetic SHR for the duration of the experiment (Figure 1C). Body weight was also reduced immediately before stroke in the SHR-acute diabetic group (P<0.05). Both cohorts of diabetic animals sustained a modest (∼10 mg Hg) increase in blood pressure and a decrease in rectal temperature when compared with the SHR-nondiabetic group (P<0.05) (Table 1b). Before and 24 h after stroke surgery, the diabetic cohorts had reduced rectal temperatures than did SHR controls (Table 1b).

Figure 1.

Infarct size, location, and physiologic variables in aged WKY and SHR animals with acute and chronic diabetes. (A) Total infarct volume as a proportion of contralateral hemisphere. Each data point represents an individual animal. (B) Regional involvement in infarction showing contribution of cortical, striatal, and additional subcortical structures. (C) Weight and plasma glucose concentration profile over the chronic diabetic period in SHR. Solid line=weight, dashed line=plasma glucose. circle=WKY nondiabetic; triangle=SHR non-diabetic; diamond=SHR acute diabetic; square=SHR chronic diabetic. (D) Neurologic deficit at 2 and 24 h after MCAo. Scale measures deficit in forelimb flexion, torso twisting, lateral push and mobility. A higher score indicates greater deficit. Solid bars represent deficit at 2 h after MCAo; open bars represent deficit at 24 h after MCAo. All data are presented as mean±s.d. ^*P<0.05 when compared with appropriate nondiabetic WKY control or for specific comparisons indicated.

Impact of Stroke in Hypertensive Diabetic Rats

All cohorts showed equivalent neurologic deficits immediately before reperfusion at 2 h after MCAo, indicating that all animals experienced similar degrees of ischemia. At 24 h, these deficits persisted in all SHR cohorts, although with greater variability. Wistar-Kyoto nondiabetic controls trended toward milder deficits (Figure 1D).

Nondiabetic SHRs sustained larger infarcts than did the WKY nondiabetic group (P<0.05). Spontaneously hypertensive rats with chronic diabetes had a larger proportion of damage than did the SHR-nondiabetic group (Figure 1A; P<0.05). Overall, the larger total infarct size in the SHRs was mainly a consequence of greater cortical infarction, although additional subcortical regions were recruited into the infarct in the chronic diabetic cohort. The degree of striatal injury was the same in all animals (Figure 1B). Importantly, all animals developed an infarct, with all SHRs (irrespective of their diabetic status) sustaining cortical damage. Cortical infarct was observed in 75% of WKY nondiabetic animals. More than 60% of all animals in each group had damage to the hypothalamic and preoptic area. Damage to these regions was most prevalent in the chronic diabetic group in which all animals showed infarct.

Discussion

A weakness in preclinical evaluation of candidate stroke therapies has been a reliance on testing in young healthy rats (Sena et al, 2007). In humans, stroke incidence is strongly dependent on age (Frost et al, 2007), and outcome significantly worsens when common comorbidities such as hypertension and diabetes are also present (Alberts and Atkinson, 2004; Kaarisalo et al, 2005; Williams et al, 2002). Animal models rarely consider the impact of comorbidity factors, such as hypertension (Ginsberg and Busto, 1989), diabetes (Kamada et al, 2007), obesity (Vannucci et al, 2001), or nicotine (Wang et al, 1997), which may be part of the reason why neuroprotection has failed to translate from animal studies to a successful clinical trial. We have shown that stroke can be successfully induced by thread occlusion in aged hypertensive rats with acute or chronic diabetes. Although mortality tended to increase with overall disease burden (Table 1a), it was not sufficient to render modeling impractical. Smaller, daily doses of STZ, rather than bolus dosing with STZ, seemed important for the survival in aged hypertensive animals. Future experiments need to examine the consequences of multiple comorbidities on both long-term mortality and morbidity.

Our data are consistent with the published literature regarding hypertensive animals. Spontaneously hypertensive rats experienced larger infarcts (Barone et al, 1992; Spratt et al, 2006) than did their close genetic relatives, the WKY strain (Figure 1A). There were no animals without infarction. Within the WKY nondiabetic control group, 75% of animals had cortical involvement. All SHRs (nondiabetic, acute diabetic, and chronic diabetic) had cortical infarcts. The additional infarction was primarily cortical with an additional subcortical component in the chronic diabetic cohort (Figure 1B). Given weight differences between groups (Table 1b), the dimensions of the occluding thread together with vessel size may need to be carefully examined to ensure equal occlusion of the MCA between groups before we can assume that larger infarcts are attributed to disease rather than to methodological artifact (Spratt et al, 2006).

After acute hyperglycemia, infarct development seems to be accelerated and extended particularly in the cortex (Martin et al, 2006). Similar effects have also been reported after 1 to 4 weeks of exposure to STZ-diabetes in rats subjected to very short (10 mins)- or moderate (1 h)-duration MCAo (Kamada et al, 2007). Moreover, 1 week of STZ-diabetes seemed to have a greater effect than did acute hyperglycemia when the animals were matched for elevation (∼15 mmol/L) in plasma glucose (Kittaka et al, 1996). In this study, 2 weeks of diabetes (plasma glucose 29±5 mmol/L) immediately before 2 h of MCAo did not increase infarct size beyond that seen in hypertensive SHR controls, but chronic diabetes of 36 weeks duration significantly increased the proportion of damage (Figure 1A). A likely explanation for the relatively modest additional effect of diabetes in our experiments is that the combination of severe ischemia (2 h MCAo) in long-term hypertensive animals already produces near maximal lesions. When diabetes is maintained for long period of time for development of diabetic vascular complications in the absence of additional comorbidity, further damage is accrued.

An unexpected observation derived from this proof-of-concept study was the marked variation in rectal temperature detected in the different animal cohorts before and after stroke (Table 1b). Although it is normal to control body temperature during the immediate surgical period (as was the case in this study), these apparently preexisting differences (Table 1b) might be important.

In this study, we have shown that aged comorbid animals can be used successfully for stroke modeling. Using such animals might provide the filter required to improve translation from the research laboratory to the clinic.

Disclosure/conflict of interest

The authors declare no conflict of interest.