A Mouse Model for Vascular Cognitive Impairment and Dementia Based on Needle-guided Asymmetric Bilateral Common Carotid Artery Stenosis

Summary

The needle method of asymmetric bilateral common carotid artery stenosis is proposed to create a mouse model for vascular cognitive impairment and dementia. It results in longer-term outcomes compared to previously established models and is compatible with live MRI. Visual representation demonstrating the procedure provides guidance for mastering the surgery.

Abstract

Vascular cognitive impairment and dementia (VCID) results from vascular brain injury. Given VCID's high incidence, which is expected to continue rising as the population ages, it is critical to establish a robust animal model for the disease. This paper presents a novel method of creating a mouse model of VCID that is based on asymmetric bilateral common carotid artery stenosis, which mimics human chronic cerebral hypoperfusion caused by carotid atherosclerosis.

Briefly, common carotid arteries (CCAs) are ligated to different gauge needles (32 G for the right CCA and 34 G for the left CCA) using 7-0 silk sutures followed by immediate needle removal. The remaining suture rings cause persistent blood flow reduction and long-term cognitive impairment associated with white matter injury, microinfarcts, and reactive gliosis, thus closely mimicking the pathogenesis of VCID. Importantly, in this needle model, the clinical representations do not revert with time, providing reliable long-term cognitive impairment. Moreover, the survival rate 24 weeks post surgery was 81.6%, which is higher compared to the other established models of VCID with a similar level of blood flow reduction.

Additional advantages include low material cost and compatibility with MRI to monitor brain injury in live animals since no metal is implanted. The main challenge in employing the needle model of VCID is the requirement for developing advanced surgical skills since mouse CCAs are less than 0.6 mm in diameter and are very fragile. High-quality visual representation of the surgery will thus help researchers to master this technique and advance our understanding of VCID, potentially leading to the development of novel therapeutic modalities to decrease the devastating cognitive decline associated with VCID.

Introduction

Vascular cognitive impairment and dementia (VCID) is the second leading cause of cognitive decline. Despite the undeniable progress that has been made toward understanding VCID pathogenesis and its risk factors, the mechanism of how neurovascular dysfunction contributes to the decline of cognitive ability remains vague. A number of rodent models of various complexity have been established to induce cerebral ischemia to mimic the clinical representations of human VCID¹. Some of these models are based on creating transient cerebral hypoperfusion; however, most of them are generated by inducing chronic cerebral hypoperfusion, the main mechanism that leads to VCID in human patients².

Chronic cerebral hypoperfusion can be introduced using either bilateral carotid artery occlusion (BCAO), which causes severe, but often fatal results, or bilateral carotid artery stenosis (BCAS). BCAS is usually performed using one of two methods: by placing identical microcoils around both CCAs, resulting in symmetric stenosis³; or by implanting an ameroid constrictor and a microcoil around the left and right CCAs, respectively, causing gradual occlusion and ~50% blood flow reduction on the left and right CCAs, correspondingly⁴. The drawbacks of both methods are high mortality rate if stenosis is too severe or if the CCA is occluded and incompatibility with live-animal MRI scan due to the presence of metal in the body. A few genetic mouse models have also been established^1,5,6,7,8. Additional options include cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy mouse models^9,10. Nevertheless, none of the proposed models mimics the full range of ischemic damage that is presented in human patients, and therefore the search for updated VCID models continues.

This paper presents a novel surgical method of inducing asymmetric bilateral common carotid artery stenosis (ABCS) in mice, where CCA stenosis is performed using silk sutures and is controlled by ligating CCAs to needles of various diameters followed by immediate needle removal¹¹. As a result, suture rings of precise diameters are permanently left on the CCAs to ensure chronic stenosis. The benefit of using ABCS over a symmetric method is that moderate hypoperfusion on the right ensures better survival while more pronounced hypoperfusion on the left assures long-term neurological and pathological representations. This needle model has several advantages over the traditional BCAS models¹¹ such as persistent results, lower mortality, ultra-low cost, flexibility, and the possibility to use special analytic approaches.

To elaborate on these advantages, three ligations cause a fragment of CCA stenosis rather than focal point stenosis, leading to persistent hypoperfusion, white matter injury, and cognitive decline in ~90% of mice. The mortality of the needle mice was ~17%, lower than that of Hattori's ameroid restrictor/microcoil model⁴, which has ~30% mortality over 16 weeks based on our experience. Each BCAS model typically costs around $100 due to the expensive microcoils or ameroid restrictors, while the needle model costs only about $1 per mouse. Moreover, the gauge of the needles could be modified depending on the research-specific requirements for the blood flow restriction on either side. In the variation presented in the current paper, the needle model mimics the pathophysiology of severe carotid stenosis, caused by unilateral permanent stenosis without occlusion, which is the most common representation of the disease in clinic¹¹. Further, ameroid restrictors and microcoils that are used in traditional BCAS models are made of metal, which can cause significant artefacts if in vivo MRI is performed, even though the metal is implanted not in the brain but in the chest. It might be difficult to predict how exactly the presence of metal would affect the imaging.

Generally, in vivo MRI that is performed after microcoil implantation is usually simple anatomical imaging, not suitable for quantitative analysis of multiple insults, which is highly desirable for VCID research. In contrast, the needle model presented here uses only silk sutures and is fully compatible with any kind of in vivo MRI. This is significant for two reasons: (1) MRI is extremely sensitive to small brain lesions, microbleeds, or superficial siderosis¹², thus is preferred over the other methods of analysis, such as CT scan (2) in vivo MRI should be preferred over the ex vivo MRI, as VCID research can undoubtedly benefit from tracking the dynamics of lesion progression/healing, especially in response to the proposed novel treatments. Additionally, functional (fMRI) can be performed in the needle model to provide critical insights into the integrity of neurovascular coupling in response to cerebral hypoperfusion. Thus, the possibility of using in vivo MRI opens an avenue for in-depth analysis of the intricate correlation between the size and the location of the lesions and cognitive function, as well as neurovascular coupling, especially in pharmacodynamics studies.

Protocol

All animal protocols were approved by the Institutional Animal Care and Use Committee at the University of Pittsburgh and performed in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals. Sterile techniques must be maintained in all survival surgeries. Twelve-week-old male C57BL/6J mice with body weight 25-30 g were used in the presented experiments.

1. Preparation of the materials and working space

1. Prepare the needle fragments (~4 mm long; 32 G for the right CCA and 34 G for the left CCA). Slightly blunt the sharp end of the needles by carefully tapping the sharp end against a hard surface and carefully separate the needles from the plastic pieces using a needle holder. Precut 7-0 silk sutures into 1-2 cm long pieces.
2. Sterilize surgical instruments, needle fragments, and sutures by autoclaving or any other appropriate method.
3. Disinfect the surgical working space with 70% ethanol immediately before the surgery.

2. Performing the surgery

1. Weigh the mouse and induce anesthesia by placing the mouse for 2-3 min in a chamber ventilated with 3% isoflurane in a mixture of 25% O₂ and 72% N₂O.
2. Place the mouse (ventral side up) on a heating pad to maintain a constant body temperature (37 °C) and secure a face mask for ventilation with 1% isoflurane in a mixture of 25% O₂ and 74% N₂O. Secure the mouse's limbs with adhesive tape.
3. Shave the fur on the neck with an electric shaver. Clean the fine hair with adhesive tape.
4. Sterilize the surgical site with Betadine Solution (10% iodine). Deiodize the skin with 70% ethanol. Ensure that the level of anesthesia is adequate by a lack of response to a firm toe pinch (pedal reflex).
5. Make a vertical midline incision along the trachea and separate the bilateral thyroid glands with micro forceps. Pull the skin and tissue away with sterile small skin retractors.
6. Under the microscope, use angled tweezers to carefully expose and blunt dissect one of the CCAs from the vagus nerve and the sheath. Use sterile water or phosphate-buffered saline (PBS) to wet the incision site if the CCA tends to stick to the forceps. Place a small plastic syringe under the neck to support posture if required.
   NOTE: Take extra care not to rupture the CCA.
7. Thread three precut silk suture fragments (size 7-0) under the CCA (1 mm apart) using angled tweezers.
8. Draft a very loose double knot around the CCA on one of the suture fragments.
   NOTE: Wetting the surgical site and the sutures with sterile water or PBS might help with drafting a knot.
9. Place a fragment of needle (32 G for the right CCA and 34 G for the left CCA) parallel to the CCA, inside the loose knot (Figure 1A). Carefully tighten the primary knot around both the needle and the CCA until no blood flow is observed and secure it with a secondary knot. Immediately pull out the needle to restore partial blood flow and trim the ends of the suture.
   NOTE: The surgeon must be very careful trimming the suture ends after the knots are completed. Cutting too close to the knot might result in loosening the knot. Additionally, it is critical to observe the CCA after removing the needle to make sure that the blood flow is still enabled downstream of the knot. If the CCA looks very pale after needle removal, it means that the CCA is overcompressed.
10. Repeat steps 2.8-2.9 for the 2^nd and 3^rd suture threads on the same CCA, approximately 1 mm apart from each other (Figure 1B).
11. Repeat steps 2.6-2.10 with the opposite CCA using a 32 G needle fragment.
12. Carefully examine whether both CCAs are efficiently ligated: confirm that all three suture bands are securely knotted in place but are not so tight as to completely block the blood flow (check that the CCAs are not pale downstream of the knots). Check that the three suture bands on each CCA are approximately 1 mm apart from each other.
13. Close the skin using sterile silk sutures and treat the closed incision site with a 2-3-mm-thick layer of EMLA cream.
14. Immediately after the surgery, inject 100-150 µL of ketoprofen (1 mg/mL stock; 5 mg/kg body weight) intraperitoneally to relieve postoperative pain. Repeat this injection 24 h and 48 h after the surgery.
15. Place the mouse on a homeothermic blanket at 37 °C for continuous monitoring for 2 h prior to returning the animal to the animal facility.
    NOTE: For sham procedure, steps 2.7-2.12 should be omitted.

3. Validation of the model

NOTE: Brain injury was analyzed with in vivo MRI and further verified with Luxol fast blue (LFB) staining and behavioral tests.

1. Imaging of cerebral blood perfusion.
   1. Anesthetize the mouse using 1-1.5% isoflurane. Sterilize the surgical site with Betadine Solution (10% iodine). Then, deiodize the skin with 70% ethanol.
   2. Secure the skull of the animal in a stereotactic frame. Make a midline saggital incision in the scalp from frontal to occipital bone to expose the skull and clean the skull surface with sterile saline.
   3. Place a charged-coupled device camera 10 cm above the skull using a two-dimensional laser speckle system. Put a probe holder over the craniectomy site and secure it firmly.
   4. Take blood perfusion images 5 min before surgery and immediately after needle release or 7, 14, 21, 28, 35, and 42 days after surgery.
   5. Close the skin using sterile silk sutures and treat the closed incision site with a 2-3-mm-thick layer of ELMA cream.
2. In vivo MRI
   1. Anesthetize the mice using 1-1.5% isoflurane, with respiration continuously monitored and temperature maintained at 37 °C with warm air during image acquisition.
   2. Perform in vivo MRI using a 9.4T scanner, an 86 mm Tx coil and 4-channel mouse brain receiver array, running the associated software. Following positioning and pilot scans, acquire T2-weighted images (T2WI) using a Rapid Acquisition with Relaxation Enhancement (RARE) sequence, with the following parameters: Echo Time/Repetition Time (TE/TR) = 40/4,000 ms, averages = 8,256 × 256 matrix, 16 slices with a 0.5 mm slice thickness, a RARE factor = 4, and a field of view (FOV) of 20 x 20 mm.
   3. Collect Diffusion Tensor Imaging (DTI) data using using an echo planar imaging (EPI)-DTI imaging sequence using the same geometry and parameters as the T2WI with the following exceptions: TR/TE = 2,300/22 ms, acquisition matrix = 128 x 128, 2 segments, 5 A0 images, and 30 noncolinear diffusion images, Δ/δ = 10/3 ms, and a b-value = 1,000 s/mm².
   4. Analyze DTI data using DSI Studio software  (http://dsistudio.labsolver.org/), looking for differences in diffusion scalar parameters (Fractional Anisotropy (FA), Mean Diffusivity (MD), Axial Diffusivity (AD) and Radial Diffusivity (RD)). Draw regions of interest (ROIs) for the corpus callosum (CC), external capsule (EC), internal capsule (IC), Fimbria, anterior commissure (AC), cingulum (Cing), hippocampus (hippo), Cortex (C), and striatum (Str) from both hemispheres.
3. Luxol fast blue (LFB) staining
   1. Prepare brain sections: fix the brain with 4% PFA for 24 h and then immerse the brain in 30% sucrose until the brain sinks. Embed the brain in OCT compound on dry ice. Cut 20 µm-thick coronal brain slice on a sliding microtome. Store brain slices in storing solution (30% glycerol/30% ethylene glycol in PBS).
   2. Immerse brain sections in LFB (0.1% alcohol solution), keep at 56 °C overnight, and wash with distilled water.
   3. Incubate brain sections in 0.05% lithium carbonate and dehydrate through graded alcohols.
   4. Stain the sections with 0.5% cresyl violet for 5 min, differentiated with 70% ethanol.
   5. Mount the stained brain sections with mounting medium.
4. Cognitive function assessment with a modified Morris water maze test
   1. Fill a circular tank with water (25 ^oC, depth 33 cm) and submerge a square ~10 x 10 cm² plexiglass platform 1.2 cm below the water level, 31 cm from the north, east, south, or west edge of the pool.
      NOTE: This platform should remain in the same place for the whole duration of testing.
   2. Turn on a ceiling-mounted video camera. Place a mouse in the tank (do not drop-release at water level), facing the wall and starting at either northeast, southeast, southwest, or northwest. Allow the mouse to swim for a maximum of 90 s to find a submerged platform.
   3. Allow the mouse to remain on the submerged platform for 30 s if the platform is found. If the platform is not found, place the mouse there for 30 s.
   4. Let the mouse rest for 5 min and repeat the swimming trial (steps 3.4.2 - 3.4.3) 2x.
   5. Dry the mouse with a piece of cloth and return it back to its cage under a heating bulb.
   6. Repeat three swimming trials every day for 5 sequential days.
   7. Using video footage, calculate latency to escape (how long does it take to find the platform in each trial).
   8. On the last day of testing, remove the platform and record the time that the mouse spends in the quadrant, where the platform was previously located.

Discussion

Several methods of VCID induction using asymmetric CCA stenosis have been described, and all of them share an important and critical surgical step-isolation of the CCA from the vagus nerve and the sheath and exposure of the CCA to make it accessible for stenosis. While we provide good-quality visual guidance on the surgical exposure of CCAs prior to ligation, we would also like to direct researchers to watch additional videos on CCA isolation that are available online in the context of other mouse and rat surgeries^15,16,17,18. Extra care should be taken not to compress or rupture the CCA as it is the main artery that supplies the brain with oxygenated blood.

Another critical step of the surgery that needs to be mastered is drafting a loose knot and further tightening it around the CCA with the needle followed by needle removal. We highly recommend practicing this step with just the needle alone before attempting to perform it in an anesthetized mouse. This will allow for mastering delicate forceps movements and perfecting manipulations with the sutures without the risk of hurting the mouse. The knot needs to be secured well without falling apart during needle removal, tight enough to hold the needle but loose enough to enable sliding the needle away from the knotted suture using forceps.

Importantly, we found that a single ligation of each CCA is not enough to reliably decrease the blood flow and to maintain long-term cerebral hypoperfusion¹¹. One potential explanation might be that a single ligation causes point stenosis, which is likely to cause a local increase in pressure, leading to increased velocity of blood flow to compensate for the blood flow reduction. We recommend performing three ligations ~1 mm apart from each other to create a fragment of stenosis. An additional benefit of using three ligations is that the knots serve as a proofing mechanism in case one of the knots gets loose during additional needle removal. Indeed, a fragment of stenosis caused by three separate ligations of the same CCA increases model consistency, leading to persistent cerebral hypoperfusion in about 90% of mice. Moreover, fragment stenosis mimics CCA stenosis caused by atherosclerosis in human patients more precisely compared to focal point stenosis, thus increasing the clinical relevance of the needle model.

We strongly recommend prompt removal of the needle after finishing the first ligation to make sure CCA blood flow is partially recovered, and only then move on to the second and third ligations sequentially. It is not recommended to finish all three ligations with the needle inside all three knots, as this would significantly increase the time of complete obstruction of blood flow. Typically, we do not recommend more than 1 min of complete CCA occlusion before removing the needle. This recommendation is based on the report that mice showed no signs of any functional impairment after 3 sessions of 60 s occlusion of both CCAs¹⁹. In our model, the surgeon works on one CCA at a time, which is more forgiving than the occlusion of both CCAs simultaneously, but we still recommend following this timeline to exclude any artifacts caused by prolonged occlusion rather than chronic hypoperfusion on either side.

While this needle model allows for adjusting the diameters of the knotted sutures by using needles of various diameters (based on body weight or the specific requirements of blood flow restriction), in our experience, mice had higher survival rates when hypoperfusion in the right hemisphere was moderate rather than severe. On the other hand, persistent severe hypoperfusion in the left hemisphere produced long-term pathological and neurological outcomes. Thus, we recommend performing an asymmetric ligation by using a thicker needle to ligate the right CCA (causing moderate hypoperfusion in the right hemisphere) and a thinner needle to ligate the left CCA (causing severe hypoperfusion in the left hemisphere).

Lastly, researchers should be aware that different mouse strains may produce different outcomes from the ischemic or traumatic insult, mainly due to differences in cerebral vascular anatomy^20,21. Since many studies nowadays require the generation of novel transgenic mice, the background strain must be carefully considered if BCAO surgery is required at any stage of the experimental design. For example, both C57BL/6 and SV129 strains are a common background choice for generating transgenic animals for stroke research²¹. However, it has been well documented that C57BL/6 mice are much more sensitive to ischemia compared to the other strains tested, including SV129 mice^20,21. In fact, there is evidence that the effect of the murine strain can be even more important than the effect of the technique used to induce VCID²¹. Thus, it becomes critically important to keep the mouse backgrounds consistent throughout all experiments that involve inducing ischemic brain injury in rodents. Importantly, researchers can evaluate the efficiency of the surgery outcome in live animals using neurological scoring systems²² with a score of 0.5 as an inclusion criterion. Brain injury can be further confirmed using Iba1 immunostaining which is very sensitive to brain damage even after minor focal insult.

In summary, it is important to remember that VCI is a complex term that unites many clinical representations and causes under the same umbrella. Therefore, researchers should always keep in mind which model should be selected based on the VCI aspects that they would like to study. There can never be a single universal model for all VCI manifestations. CCA stenosis models severely limit blood flow from the major arteries, thus mimicking patients with arteriosclerotic stenosis. The novel needle method of creating asymmetric BCAS in C57BL/6J mice is a reliable method to mimic VCID that offers several advantages over the previously reported methods (particularly high flexibility, low mortality, long-term outcomes, minimal cost, and live MRI monitoring). Due to its advantages over the other models, it can be used to further advance our knowledge of VCID progression as well as serve as a basis for screening potential therapeutic agents to cure or slow the progression of VCID. Similar to other reported methods of BCAS, the needle model requires advanced surgical skills that can be mastered with time, using this visual demonstration as a guide.

Materials

Name	Company	Catalog Number	Comments
1 mL syringe	BD	309659	IP injection of ketoprofen
2.5 mm blunt retractor tips	Kent Scientific Corporation	SURGI-5016-2	2 needed
Andis Ion T-Blade Pet Trimmer	Andis	BTF3	To shave fur
Betadine Solution	Avrio Health L.P.	NDC 67618-150-17
Braided silk suture, 5-0	Teleflex Medical	106-S	For suturing skin
Braided silk suture, 7-0	Teleflex Medical	103-S	For ligating CCAs
Camera	Med Associates, Inc	VID-CAM-MONO-7	For model validation
cover glass	Fisherbrand	12-541-012	For model validation
cresyl violet	Thermo Scientific	C581-25	For model validation
Dispensing Tip, Needle, Flex, 34 Gauge, 1/4" OAL	Jensen Global	JG34-0.25HPX	For the left CCA
DSIStudio			https://dsi-studio.labsolver.org; for DTI data analysis
Dumont Tweezers; Pattern #1	Roboz Surgical Instrument Co	RS-4960	2 needed
Dumont Tweezers; Pattern #5, 45 Degree Angle	Roboz Surgical Instrument Co	RS-5058	2 needed
Ethanol	Fisher Scientific	64-17-5	Dilute to 70%
ethylene glycol	Thermo Scientific	E178-4	For model validation
Eye Needle, Size #4; 3/8 Circle, Cutting Edge, 11 mm Chord Length	Roboz Surgical Instrument Co	RS-7981-4	For suturing skin
E-Z Anesthesia Classic System	E-Z systems	EZ-7000
glycerol	Thermo Scientific	AAA16205AP	For model validation
Homeothermic Monitoring System	Harvard Bioscience	K 022258
Imaging system EVOS FL Auto	Life Technologies	AMAFD2000	For model validation
Isoflurane	Covetrus	11695-6777-2
Ketoprofen	Zoetis	5487
Leica M320 F12 clinic and surgery microscope	Leica	M320 F12
Lidocaine and Prilocaine Cream, USP 2.5%/2.5%	Padagis	NDC 0574-2042-30	Generic to EMLA cream
lithium carbonate	Thermo Scientific	L119-500	For model validation
Luxol fast blue	Thermo Scientific	AC212170250	For model validation
microscope slides	Fisherbrand	12-550-403	For model validation
Microtome	Leica	SM2010R	For model validation
Morris Water Maze with the hidden platform	Maze Engineers		https://maze.conductscience.com/portfolio/morris-water-maze/
M-Prove Portable Balance	Sartorius	AY711	Scales for weiging the mouse
MRI with ParaVision 6.0.1	Bruker	AV3HD 9.4T	For model validation
Multi gas flow meter	Aalborg Instruments	GMR2-010334	for low flow rate gas blending (N2O and O2)
Nano Ultra Fine Pen Needles - 32G 4mm	BD	58320883	For the right CCA
O.C.T.	Thermo Scientific	23-730-571	For model validation
Paraformaldehyde, 4%	Thermo Scientific	J61899.AK	For model validation
PBS	Thermo Scientific	BP399500	For model validation
PeriCam PSI System with Aperiflux probe holder	Perimed Inc	PeriCam PSI HR	For model validation
permount	Thermo Scientific	SP15-100	For model validation
Round Handle Forceps; Micro Suturing With Tying Platform; Curved	Roboz Surgical Instrument Co	RS-5264	To help with cutting and suturing the skin
Small Animal Heat Lamp, 75 Watt	Morganville Scientific	HL0100	For model validation
Small cotton-tipped applicators	Fisher Scientific	23-400-118
Spring Scissors - 8mm Cutting Edge	Fine Science tools	15024-10
Student Halsey Needle Holder	Fine Science tools	91201-13	3 needed: 2 for holding skin retractors and 1 for suturing the skin
Sucrose	Thermo Scientific	A15583.36	For model validation
Universal Camera Ceiling Mount	Med Associates, Inc	ENV-598	For model validation
water bath	VWR	89032-226	For model validation