Pulling wire to avoid misinsertion of filament and vascular damage improves ischemic stroke models' success rate and repeatability

Ethical approval

The animals in this experiment were sacrificed according to the 2006 Guideline of the Chinese Ministry of Science and for the Care and Use of Laboratory Animals. Our study passed the ethical review requirements of Yunnan University of Traditional Chinese Medicine, ethical review no: R-062021002.

Animals

Seventy-five specific pathogen free (SPF) male SD rats, 8–10 weeks old, weighing 250–280g, were provided by Hunan Slack Jingda Biological Company. The breeding and modeling of experimental animals in this study were carried out in the SPF level experimental operation room of the university [License number: SYXK (Yunnan) K2017-0005]. The use of experimental animals followed the 3R principle and all experimental protocols were approved by the Experimental Animal Welfare Ethics Committee of Yunnan University of Traditional Chinese Medicine. Before the operation, four to five animals were housed in each cage, the rats were maintained under a 12h/12h light/dark cycle and could obtain food and water without restriction. The rats were randomly divided into¹ no-surgery group (n=15),² sham group (n=15),³ Zea longa group (n=15), and⁴ optimal group (n=15). At the end of the experiment, the animals were sacrificed by spinal cord dislocation after an intraperitoneal injection of sodium pentobarbital. All efforts were made to ameliorate any suffering of the animals.

Reagents

Preoperative and postoperative care

After seven days of adaptive feeding, the rats were diet fasted for six hours before surgery. Each rat was kept in a separate cage after surgery and the wound was disinfected with iodophor twice a day to prevent infection. The food was softened with cow's milk before the rats were fed.

Modeling method

The operation was performed under general anesthesia (2% isoflurane in 30% O₂ and 70% N₂O). Heating pads and rectal thermometers were used to detect and maintain the body temperature of the rats between 36.5–37.0°C. The MCAO standard procedure Zea longa approach for the surgery was as follows: After 75% ethanol skin disinfection, a 2 cm long median incision was made in the neck, and the right side was dissected CCA, external carotid artery (ECA), internal carotid artery (ICA), occipital artery (OA), superior thyroid artery (STA), and PPA. The distal ends of OA, STA, PPA, and ECA were ligated, and CCA and ICA were temporarily clamped. Insert the filament from the ECA into the CCA and finally insert the ICA to block the main MCA and ICA. The filament was pulled out after one hour and the muscle skin was sutured. The wire lifting method of this study is specifically as follows: After skin disinfection with 75% ethanol, a 2 cm long median incision was made in the neck, and the right CCA, ECA, ICA, and carotid artery are dissected. The proximal end of the common artery was ligated to block the blood supply, the ICA was temporarily clamped. In addition, pulling thread a (Figures 1 and 2) was tied around the ECA above 2–3 mm of the bifurcation of the ICA and ECA and a 45° incision was made on the CCA below 5–6 mm of the bifurcation of the ICA and ECA. The filament was placed into the ICA through the incision access. Then the ECA arteriole clamp was loosened, the pulling thread b was tied on the CCA, and the pulling thread a, b was lifted left with the index finger and middle finger (raise the CCA by about angle 20°–30° between the ICA and PPA to avoid the filament passing into the PPA). The filament was moved slowly forward in the ICA till a hindrance was felt indicating that the filament had approached the front end of the MCA. After ischemia of one hour, the filament was removed then the muscle skin was sutured, and the wound was disinfected. The filament was not inserted in the sham group, but the other operations were the same as those in the Zea longa group. The no-surgery group did not do have any intervention. Please see underlying data (Video file).¹⁴

Figure 1. Schematic diagram of the main arteries of the head and neck of the rat during the modelling of the different groups middle cerebral artery occlusion (MCAO) and the position of the filament.

A. In the Zea longa method. B. The optimization method of this study.

Figure 2. Schematic diagrams of the embolization paths of different methods.

A. The anatomy of the rat neck vessels: common carotid artery (CCA), external carotid artery (ECA), internal carotid artery (ICA) and pterygoid artery (PPA). D and E are the threaded paths, in which the angle between CCA and PPA is changed by pulling (a and b) two threads to the left, and CCA and ICA tend to form a straight line at the same level, and the head of the thread is easily inserted into ICA within a short period of time (a is the pulling thread, b is the pulling thread, c is the fixed thread, and d is the head of the thread).

Neurological function score

After the rats had been awake for 24 hours, the behavior of the rats was scored according to the Zea longa 5-level 4-point neurological scoring principle. The Zea longa 5-level 4-point system scoring principles: 1. 0 points, indicates no signs of neurological abnormalities. 2. 1 point indicates unable to fully extend the upper limb on the contralateral side of the lesion. 3. 2 points, indicates the rat rotates to the opposite side while walking. 4. 3 points, indicates a fall to the opposite side while walking. 5. 4 points, indicates no spontaneous activity with reduced consciousness. A score of 1–4 points is a successful model; rats that died accidentally during the operation, death within 24 hours of reperfusion, subarachnoid hemorrhage, and 0 points were excluded. Please see underlying data (Data file 2).¹⁴

TTC staining

The first day after the successful model, the rats were sacrificed after spinal cord dislocation under pentobarbital sodium anesthesia to remove the brain. The brain was washed twice with normal saline, and frozen in a refrigerator at -20°C for 10 min. After removal, the brain was placed in the brain groove and cut into five pieces at intervals of 2 mm with a coronal plane slice. The first brain cut was between the optic chiasm and the anterior pole of the brain, the second cut was on the optic chiasm, the third cut was on the pituitary gland, and the fourth place was between the pituitary gland and the posterior pole of the brain. The pieces of brain were incubated in a 0.08% TTC-0.01 mol/L pH 7.26 phosphate buffer solution stored in the dark at 37 °C for 30 min, turning once every 5 min. The brain slices were washed three times with 0.08% TTC-0.01 mol/L pH 7.26 phosphate buffer solution and the solution was then discarded. Then they were put into 40 g/L paraformaldehyde solution for 24 hours. Finally, these brain slice images were captured by a Nikon D3100 camera and ImageJ v1.8.0 analysis software was used to calculate the cerebral infarction volume of the brain slice images (Cerebral infarction volume (%) = total cerebral infarction area/total brain slice area × 100%). Please see underlying data (Data file 1).¹⁴

Nissl staining

On the first day after the successful modeling, and after the rats were sacrificed, the heart was perfused with physiological saline and then perfused with 4% paraformaldehyde 0.1 mol/L phosphate-buffered saline (PBS, pH = 7.4). Paraffin embedding took place after brain extraction. For the paraffin section, the thickness of the clean, glass slide was about 4 μm, and the paraffin section was deparaffinized with water. The sections were placed sequentially into xylene I for 20 min; xylene II for 20 min; absolute ethanol I for 5 min; absolute ethanol II for 5 min; 75% alcohol 5 min, then washed with distilled water. For the Toluidine blue staining: the tissue sections were placed in the dye solution for 5 min, washed with water, and slightly differentiated with 1% glacial acetic acid. The reaction was terminated by washing with distilled water. The degree of differentiation was controlled under a microscope (Olympus BX43, 10X). After washing with distilled water, the sections were placed in an oven (DHG Series, Shanghai Yiheng Biological Company) and dried. Transparent mounting was performed by slicing into clean xylene for 5 min and sealing with neutral gum sealer. The pathological changes of Nissl-stained tissues in the cerebral cortex and hippocampus were observed under a mounting microscope (Olympus BX43, 10X, 40X) Please see underlying data (Data file 3).¹⁴

Recording of modeling, survival time, and weight changes

After the completion of the operation, the survival time and the number and weight of the rats in each group were recorded each day.

Statistical analysis

The experimental data include at least three independent and repeated experimental values, and the values are described by the mean + standard deviation. Student’s-t Test was performed in a two-group comparison, <0.05 was considered statistically significant (SPSS 22.0, SPSS Inc, Chicago, IL, USA)

This method improves the successful efficiency of model

To compare the success rate of each group, neurological function score was adopted to evaluate the function damage of the rats. According to the model success criteria, our result indicated there was no significant difference in the neurological function scores in the sham-operated group and no-surgery group. A total of 45 rats were used in the Zea longa group and our optimal group. Ten models failed and 35 succeeded in the Zea longa method group, with a success rate of 77.77% (Figure 3A); five models failed and 40 succeeded in the optimal group, with a success rate of 88.88% (Figure 3B) and the difference was statistically significant (Figure 3C).

Figure 3. Neurological function scores and success rates of rats in different groups.

(score >1 means success). A and B. The neurological function scores of Zea longa group and Optimization group; C. The statistics of the number of successful modeling rats by different methods, and the differences are statistically significant (* indicates p <0.05).

The present modeling method increases the repeatability of the model

TTC staining was performed on the brain tissue of rats in each group, and the results showed that the right and left cerebral hemispheres of the sham-operated group were evenly and symmetrically stained in coronal section, with red, gray matter and light red or porcelain white matter, and no pale infarcted areas. The ischemic side of the brain tissue of the Zea longa group and the wire group showed pale white infarcted areas in the cortex and subcortex, including the lateral basal nucleus, parietal and temporal cortex, hippocampus, striatum and other areas. The infarct volume was 18.36 ± 4.93% in the Zea longa group and 23.75 ± 2.96% in the optimization group, and the difference was statistically significant and apparently smaller within-group, indicating that the success rate and stability of the method were better than those of the traditional method (Figure 4).

Figure 4. Results of the volume and morphological changes of brain infarcts in rats of different groups (MCAO).

A. The results of TTC staining of brain tissues in each group, light red is normal brain tissue, white is infarcted tissue, and the infarcted area was measured to compare the statistical differences between groups. B. shows the comparison of the percentage of infarcted volume in each group for statistical difference comparison. **** indicates p <0.0001.

The present experimental method can significantly damage neuronal cells in the hippocampal region

The cells in the hippocampal area of the blank group and sham-operated group were arranged regularly and neatly under light microscopy. The nucleus and nucleolus were clearly visible, with purple-blue granules or block-like structures in the cytoplasm of the nictitating bodies and naturally arranged neural protrusions. In the Zea longa group and the right hippocampal region of the optimization group, the cell arrangement was disorganized: some of the neuronal cells disappeared, the cytosol of the remaining neuronal cells was wrinkled and invaginated, the nerve fibers were abnormally curved, the intracytoplasmic Nissl disappeared and was blurred, the nucleus was solidified and homogenized, and the nucleolus was translocated (Figure 5).

Figure 5. Nissl staining in hippocampal region of rats in different subgroups (MCAO).

A. Microscopic observation of morphology of hippocampal neurons for Nissl staining, the microscopic magnification of the upper and lower parts was 100x and 400x, respectively); B. The number of Nissl vesicles in hippocampal region of different subgroups were counted to compare the statistical differences. (**** indicates p <0.0001).

This experimental method can save modeling time

We recorded the model preparation time of the two different methods and found that the mean modeling time of the Zea longa group was 21.07 ± 6.11 min, while the mean modeling time of the optimization group was 17.22 ± 2.26 min. The mean modeling time of the optimization group was less than that of the Zea longa group, and the modeling time was shorter, and more importantly, the modeling time of the optimization group was more consistent and repeatable. The difference was statistically significant (Figure 6). Please see underlying data (Data file 5).¹⁴

Figure 6. Modeling time of different methods.

Compared to the Zea longa group, the Optimization group showed a statistically significant reduction in the modeling time. (*** indicates p <0.001).

This experimental method causes less hurt to the animals

The rats fasted for six hours before surgery and after surgery each rat was kept in a separate cage and the wound was disinfected with iodine twice a day to prevent infection. The rats’ food was softened with cow's milk and placed in the cage to facilitate feeding. The animals in the blank and sham-operated groups did not die during the experiment. The bodyweight of the animals in both groups increased gradually over time, but the sham-operated group showed a decrease in body weight on the first postoperative day, which was considered to be due to the physical injury to the rats due to the surgical trauma. The Zea longa group showed a significant weight loss in the first five days after surgery and a gradual increase in weight on the seventh day, while the Zea longa group showed a weight loss in the first three days after surgery and a gradual increase in weight on the fifth day. We speculate that the main reason for this trend is that the different modeling methods caused different damage to the rats. The number of survivors in both the Zea longa group and the optimization group decreased, and the number of survivors in the optimization group decreased less than that in the Zea longa group, and the difference was statistically significant (Figure 7). Please see underlying data (Data file 4 and 6).¹⁴

Figure 7. The number of surviving rats and body weight changes after surgery.

A. The statistics of the number of surviving rats at 10 days after surgery, comparing the differences in the number of surviving rats at different time points; B. The statistics of the body weight of rats at 10 days after surgery, comparing the changes in body weight of rats at different time points. (* indicates p <0.05, ** indicates p <0.01, **** indicates p <0.0001).