This protocol describes repetitive hypoxic preconditioning, or brief exposures to systemic hypoxia that reduce infarct volumes and blood-brain barrier disruption following transient middle cerebral artery occlusion in mice. It also details dual quantification of infarct volume and blood-brain barrier disruption after stroke to assess the efficacy of neurovascular protection.
Experimental animal models of stroke are invaluable tools for understanding stroke pathology and developing more effective treatment strategies. A 2 week protocol for repetitive hypoxic preconditioning (RHP) induces long-term protection against central nervous system (CNS) injury in a mouse model of focal ischemic stroke. RHP consists of 9 stochastic exposures to hypoxia that vary in both duration (2 or 4 hr) and intensity (8% and 11% O2). RHP reduces infarct volumes, blood-brain barrier (BBB) disruption, and the post-stroke inflammatory response for weeks following the last exposure to hypoxia, suggesting a long-term induction of an endogenous CNS-protective phenotype. The methodology for the dual quantification of infarct volume and BBB disruption is effective in assessing neurovascular protection in mice with RHP or other putative neuroprotectants. Adult male Swiss Webster mice were preconditioned by RHP or duration-equivalent exposures to 21% O2 (i.e. room air). A 60 min transient middle cerebral artery occlusion (tMCAo) was induced 2 weeks following the last hypoxic exposure. Both the occlusion and reperfusion were confirmed by transcranial laser Doppler flowmetry. Twenty-two hr after reperfusion, Evans Blue (EB) was intravenously administered through a tail vein injection. 2 hr later, animals were sacrificed by isoflurane overdose and brain sections were stained with 2,3,5- triphenyltetrazolium chloride (TTC). Infarcts volumes were then quantified. Next, EB was extracted from the tissue over 48 hr to determine BBB disruption after tMCAo. In summary, RHP is a simple protocol that can be replicated, with minimal cost, to induce long-term endogenous neurovascular protection from stroke injury in mice, with the translational potential for other CNS-based and systemic pro-inflammatory disease states.
As the leading cause of adult disability and the fourth leading cause of death, stroke is one of the most debilitating disease states facing the adult population of the United States.1 Animal models of stroke allow for experimental investigation of new methods of reducing ischemic injury and improving post-stroke recovery. One novel avenue for such translational research is preconditioning. Preconditioning is the intentional use of a non-damaging stimulus to reduce damage from a subsequent, and more severe, injury.2 Hypoxic preconditioning has been shown to produce pleiotropic changes in the brain that provide protection against stroke in both in vivo and in vitro studies.3 However, a single exposure to hypoxia only offers short-term neuroprotection, inducing less than 72 hr of tolerance against ischemia in adult mice.4 Even after four weeks of 14 hr daily exposures to hypobaric hypoxia, Lin et al. found that neuroprotection was only sustained for one week.5 Repetitive hypoxic preconditioning (RHP) is characterized by stochastic variations in frequency, duration, and intensity of hypoxic exposures. In contrast to a single preconditioning challenge, RHP induces a cerebroprotective phenotype that lasts up to eight weeks in mice.6 RHP reduced infarct volumes, blood-brain barrier (BBB) disruption, vascular inflammation, and leukocyte diapedesis for weeks after the final hypoxic exposure. RHP specifically reduced inflammation in the ischemic brain by reducing T cell, monocyte, and macrophage populations, while maintaining B cell populations in the ischemic hemisphere.7 In fact, RHP induced an immunosuppressive phenotype in mice prior to any CNS injury, including stroke. RHP-treated B cells isolated from RHP-treated healthy mice exhibited a unique anti-inflammatory phenotype, with a downregulation of both antigen presentation and antibody production. The overall reduction in pro-inflammatory adaptive immune mechanisms makes RHP an excellent methodology to induce endogenous immunosuppression for not only CNS-specific inflammatory diseases, but also systemic injury or disease models that include a pro-inflammatory pathology.
RHP reduces both infarct volume and BBB disruption following a transient middle cerebral artery occlusion (tMCAo). Animal models of stroke, such as the commonly used tMCAo, dramatically improve the understanding of the pathophysiology of stroke, as well as the design of more effective neurotherapeutics. First developed by Koizumi et al., in 1986,8 the tMCAo procedure is a widely used method of inducing stroke in rodents and one of the preferred methods for investigating inflammation following reperfusion. As the methods for tMCAo evolve, the more recent use of silicone-coated filaments further reduce the risk of subarachnoid hemorrhage compared to other models9,10 and improve reliability, though unfortunately tMCAo often produces a wide variation in infarct volumes.11-13 Most of these studies delineate infarction regions in coronal brain sections by staining with 2,3,5- triphenyltetrazolium chloride (TTC), considered a gold standard for infarct quantification because it is a simple and inexpensive way to produce vivid, replicable results. TTC serves as a substrate of dehydrogenases present in mitochondria. When brain slices are exposed to the TTC solution, TTC is selectively taken into living cells where its non-soluble reduction product, formazan, precipitates to a deep red color in viable mitochondria. Because of mitochondrial dysfunction in the ischemic tissue, this tissue remains white, allowing for differentiation of damaged and healthy tissue.14
RHP also reduces BBB disruption in the ischemic hemisphere.6 Therefore, the dual quantification of BBB integrity within the same brains as TTC-based infarct volume determinations15 would provide useful information about the full efficacy of endogenous protection, and potential causal relations between BBB disruption and infarction in untreated and treated animals. The influx of peripheral blood through a disrupted BBB, secondary to stroke, increases leukocyte populations, pro-inflammatory cytokines, oxidative stress, vasogenic edema, and hemorrhagic transformation in the ischemic hemisphere, ultimately increasing the rates of infection and mortality in patients with ischemic stroke.16,17 A common method of measuring BBB disruption in animal models is through quantification of Evans blue (EB) dye leak into the brain.15,18-21 EB selectively binds to serum albumin, a globular protein (MW=65 kDa) that does not cross the BBB in uninjured animals.22 Following ischemic stroke, EB infiltrates the brain, and fluoresces at 620 nm, allowing for measurement of optical density within the perfused injured parenchyma.22 The optical density is directly proportional to the permeability of the BBB when EB has been washed out of the post-mortem cortical vasculature by transcardiac perfusion. With the immediate processing of TTC-stained brains in animals with EB administration, both the infarct volume and BBB disruption can be effectively quantified. It should be noted, however, that neuronal injury and BBB disruption are not concomitant processes in the post-stroke brain,23,24 so the selection of time of sacrifice is an important consideration.
The protocol that follows details the RHP method, the tMCAo method for inducing a temporary arterial occlusion that models middle cerebral artery occlusions in human patients, and the dual histological methods for determining neural and vascular stroke injury endpoints. TTC measures cell death and cumulative tissue damage, allowing for the quantification of an overall infarct volume, while EB provides for the hemispheric quantification of BBB damage.
NOTE: This protocol was approved by the Institutional Animal Care and Use Committee (IACUC) at UT Southwestern Medical Center which abides by the National Institutes for Health (NIH) policy for experimental animal usage.
1. Repetitive Hypoxic Preconditioning
2. Transient Middle Cerebral Artery Occlusion (tMCAo)
3. Evans Blue (EB) Injection
4. 2,3,5- Triphenyltetrazolium Chloride (TTC) Staining
5. Infarct Volume Quantification
6. Blood-brain Barrier (BBB) Integrity Quantification
This study included 25 male Swiss Webster mice that were 10 weeks of age at the start of randomization into RHP (n = 10) or 21% O2 (n = 15) groups. Two weeks after the final RHP exposure, surgical procedures were performed, with groups blinded and counterbalanced between days. Following tMCAo, 1 mouse died during post-operative recovery and 1 mouse was excluded from the study because it did not meet the reperfusion CBF criterion. Both excluded mice were from the 21% O2 group. In accordance with ARRIVE guidelines,28 Table 2 shows surgical parameters. Indirect infarct volumes, hemispheric swelling (i.e., edema), and Evans blue (EB) extravasations are shown in Figure 2. All data were analyzed with unpaired t-tests (mean, standard deviation shown), and outliers detected with a False Discovery Rate less than 1% (Graphpad Prism), which removed 4 outliers (2 from 21%, 2 from RHP) from the EB data set.
In this cohort, RHP reduced infarct volumes by almost half (33.9 ± 23.0 mm3) compared to 21% O2-treated mice (62.6 ± 42.6 mm3), but this difference was non-significant (p = 0.10). However, an a priori Power analysis predicted a required n = 10 per group, which we did not achieve with the RHP-treated mice after outliers were removed. This should be considered in designing future experiments using RHP. There was no effect of RHP on hemispheric swelling, a metric derived from the TTC volume analysis that could be equated with edema.6 RHP-treated mice exhibited a trend (p = 0.05) in the reduction of BBB disruption within the ischemic hemisphere normalized to the contralateral hemisphere. Figure 2D shows a range of values for EB leak for both treatment groups.
Figure 1: Custom-designed RHP chambers. Upper panel depicts the custom-built flowmeters for individual monitoring of air flow in up to four chambers. Air impermeable tubing is shown leaving the flowmeter and attaching to the inlet port of the 15 L induction chamber in the lower panel. Outlet port remains open during RHP to allow for air circulation.
Figure 2: RHP reduces infarct volumes and BBB disruption. RHP (A) reduces infarct volumes by 46% compared to control mice, but has (B) no affect on hemispheric swelling derived from the TTC data. (C) In the same animals, RHP reduces BBB disruption (p = 0.05) as defined by Evans Blue leak normalized to the contralateral hemisphere. (D) Representative TTC stains from both RHP-treated and 21% O2-exposed control mice with corresponding infarct volumes and Evans Blue leak shown below each sample. Values shown are mean +/- standard deviation.
A single exposure to systemic hypoxia (i.e., 2 hr of 11% O2) in mice “transiently” protects the brain from tMCAo,29 meaning the epigenetic response to the hypoxic preconditioning challenge is short-lasting, and the baseline phenotype is restored within days. Repetitive presentations of the hypoxic preconditioning stimulus dramatically extend the duration of the neuroprotective phenotype.6 Many studies have shown that the frequency, magnitude, and duration of the repetitive stimulus train are critical determinants of this response. For example, simply repeating the same intensity and duration of hypoxia (2 hr of 11% oxygen) 3 times per week (M-W-F) over 2 weeks did not extend the therapeutic window for neurovascular protection from tMCAo,6 but was sufficient to induce long-term ischemic tolerance in the retina.30 Interestingly, although 5 exposures to systemic hypoxia (5 hr of 8% O2) spaced 6 days apart protected against tMCAo induced 3 days after the last hypoxic exposure, neuroprotection was lost if the 5 hypoxic exposures were placed 3 days apart.31 The reason for this difference remains unclear, but may have been due to the duration of the severe hypoxia relative to this protocol and/or insufficient recovery time between hypoxic challenges of this severity.
In short, titration of the repetitive hypoxic challenge across the domains of frequency, magnitude, and duration, seems critical to the development of tolerance while not inducing injury, not to mention tissue-specific and perhaps species-specific effects. For example, 3 of 9 of the RHP exposures involve exposure to 8% O2 for 4 hr, though 6 hr of 8% O2 induces hippocampal neuronal death.32 Therefore, there must be a strict adherence to the RHP protocol with regard to duration and severity of hypoxic exposures to induce endogenous protection. Although the potential effect of circadian rhythm on preconditioning-induced protection needs further exploration,33 performing RHP exposures at the same time of day for a particular cohort reduces the risk of any potential confounding effect. In terms of dose-response and efficacy, the most robust RHP-mediated protection from focal stroke occurred when the tMCAo was induced 2 weeks following the final hypoxic exposure,3 at a time when RHP-treated healthy mice exhibit an immunosuppressive phenotype (in the absence of ischemic injury).7 This 2 week time point may not coincide with the period of maximal protection against other acute CNS injuries, or in other organ systems. Both the RHP protocol, and the temporal features of the epigenetic response, will undoubtedly need optimization for each new translational use.
One of the greatest limitations of the tMCAo procedure is the heterogeneous distribution of infarct volumes, as shown by this representative data set. The collateral circulation provided by the Circle of Willis, leptomeningeal anastomoses, and dorsal collateral junctions can lead to inconsistent infarct volumes.34 Variations between the Circle of Willis in different strains of mice, as well as the presence and patency of posterior communicating arteries,25,35 can further reduce consistency of infarct volumes within groups. Robust sample sizes and replication are often necessary to produce conclusive results and should be included in the design of the experiment after powering accordingly. Other methods of stroke induction, particularly distal focal ischemic methods such as occlusion of the primary MCA branch via a burr hole in the lateral skull, or photothrombosis of distal branches of the MCA, often produce less variation in stroke volume. Photothrombotic strokes are limited, however, as they produce simultaneous extracellular and intracellular edema, a phenomenon not seen in ischemic strokes in humans.36 Conversely, tMCAos are more directly applicable to the clinical setting as over 60% of all strokes in humans occur due to obstruction of the middle cerebral artery.37 Finally, the anesthesia used during tMCAo, isoflurane, has been shown to induce neuroprotection.38 In order to account for this neuroprotection, isoflurane levels should be consistent between surgeries and between treated and untreated experimental groups, and maintained at <3% to reduce its potential neuroprotective effects.39
Time is also critical in the TTC staining process. Although there is great variation in the stroke literature about the time in which TTC staining can be performed after ischemia, varying from 4 hr to 7 days post-stroke,18,40 TTC staining should occur at 24 to 48 hr after stroke for the most consistent and clearly delineated infarct volumes. TTC infarct volumes have been found to stabilize 24 hr after stroke but after 48 hr, an influx of macrophages into the ischemic brain makes defining the infarct volume difficult.14,40 Furthermore, the staining of brain tissue with TTC should immediately follow sacrifice. Delaying staining will decrease the quality of the stain because of increased mitochondrial death resulting from animal death, not cerebral infarction. A similar increase in mitochondrial death will occur if the brain is not kept moist during the insertion of the blades. While this protocol clearly illustrates the benefit of infarct analysis with TTC staining, other immunohistological stains can be used to quantify the infarct volume following tMCAo. Cresyl violet or fluoro-jade staining have less rigid time requirements for staining after sacrifice, but these classic stains require very thin sections obtained by cryostat or microtome, and thus require more time for integrative summation and analysis.14 Furthermore, these stains cannot be used in conjunction with EB quantification of BBB disruption, as developed by others,15 eliminating the possibility of direct comparison between infarct volume and BBB integrity in a given brain. It should be noted that RHP-treated animals typically exhibited pink as opposed to pure white volumes.6 In order to avoid washout of infarct in RHP-treated mice, we use a shorter staining period for TTC that results in more pink healthy tissue as opposed to dark red. Evans Blue processing may also darken the infarct area so pure white infarcts are unlikely with the dual quantification staining presented in this protocol.
In order to compare EB quantification within and between groups, all animals must undergo equivalent circulation times for EB. Others have injected EB immediately following reperfusion and allowed EB to circulate for up to 72 hr,41 or at 4 hr post-tMCAo to circulate for 4 hr,15 as alternative approaches. The amount of EB that crosses the disrupted BBB and enters brain tissue represents the integral accumulation of albumin-bound dye from time of injection to time of sacrifice. Thus, the method described in this paper reveals the status of the BBB at precisely the time between 22 and 24 hr of reperfusion, and may miss earlier or later openings or closings of the barrier. Conversely, the method employed by Wang et al. reveals the status of the BBB for 72 hr following stroke, including all changes in BBB status during that 3 day post-stroke time.42 Neither protocol is better or worse; although there remains a risk of “missing” a transient opening of the BBB by using shorter circulation times, it allows investigators to identify specific temporal features of BBB pathophysiology and to pair it with other methods, as shown here. Others have combined TTC and EB by injecting EB intracardially 1 min after transcardial perfusion.18 However, this method produced inconsistent results and is a far more complicated procedure than the protocol described above.
One of the most promising future directions for RHP is the translational application of this preconditioning protocol to other disease states. Intermittent hypoxic exposures have been found to be protective in pro-inflammatory CNS conditions other than ischemia. In a rat model of Alzheimer’s, 2 weeks of 4 hr daily exposures to hypobaric hypoxia reduced the impairment of memory retention and loss of cortical neurons after intracerebral injections of beta-amyloid.43 Intermittent hypoxia (2 weeks of 15 hr daily exposures to hypobaric hypoxia) was also found to be protective in rats following L-buthionine-[S,R]-sulfoximine (L-BSO) infusion, a model of Parkinson’s which impairs striatal dopaminergic transmission.44 However, these models of Alzheimer’s and Parkinson’s only investigated the short-term effects of hypoxic preconditioning.43,44 Investigating the long-term neuroprotection induced by RHP may be a fruitful future direction for research. Furthermore, the neuroprotection induced by RHP may extend to improved BBB integrity in mouse models of Alzheimer’s, which could be analyzed with the EB.45 The dual TTC and EB analysis would be even more useful in Parkinson’s model as it induces striatal lesions46 and BBB disruption,47 which can be quantified with TTC48 and EB,49 respectively. Overall RHP is a simple, powerful form of preconditioning with great translational potential as it uniquely induces long-term, anti-inflammatory, and neuroprotective effects. The dual quantification of TTC and EB could also be easily applied to other disease states that involve mitochondrial disruption, cell death, and BBB leakage.
The authors have nothing to disclose.
Special thanks to the Gidday lab for their work in developing the RHP protocol, as well as the Neuro-Models Facility (UTSW) for their assistance in the tMCAo surgeries. This work was supported by grants from the American Heart Association (AMS), The Haggerty Center for Brain Injury and Repair (UTSW; AMS), and The Spastic Paralysis Research Foundation of the Illinois-Eastern Iowa District of Kiwanis International (JMG).
Material/ Equipment | Company | Catalog Number | Comments/Description |
Flowmeters, regulators | VetEquip, Inc | Specialty order | Four flowmeters are attached to 6.0 mm flexible PVC tubing which connects to the inlet port on each induction chamber with a plastic female connector. These flowmeters are bolted to a 6.5" x 1" x 1" metal bar. This metal bar is bolted to a MI-246-P pressure gauge with a DISS outlet. This pressure gauge and flowmeter equipment can be attached to each new gas cylinder with a wrench. |
21% O2 tank | AirGas | OX USP200 | |
11% O2 tank | AirGas | Specialty order | |
8% O2 tank | AirGas | Specialty order | |
15L induction chambers | VetEquip | 941454 | |
Moor Laber Dopper Flow | Moor Instruments | moorVMS-LDF1-HP | 0.8mm diameter probe |
High Intensity Illuminator | Nikon | NI-150 | |
Zoom Stereo Microscope | NIkon | SMZ800 | Other surgical microscopes may be used. |
Kent Scientific Right Temperature CODA | Kent Scientific Corporation | Discontinued | Recommended replacement is PhysioSuite with RightTemp Temperature Monitoring and Homeothermic Control (Kent Scientific, #PS-RT). |
Hovabator Incubator | Stromberg's | 2362-E | Our model is the 2362N. 2362E is a later model and includes an electronic thermostat. |
V010 Anesthesia system | VetEquip | 901807 | Includes: ten foot high-pressure oxygen hose, frame, flowmeter, oxygen flush assembly, vaporizer, breathing circuit, chamber, nosecones, waste gas evacuation tubing and two VapoGuard filters |
250 mL isoflurane | Butler Schein | NDC-11695 | |
D-6 Vet Trim Animal Cordless Trimmer | Andis | #23905 | Replacement blades are available from Andis (#23995) |
Betadine | Fisher Scientific | 19-898-867 | |
Q-tips | Multiple sellers | Catalog number not available | |
Gauze Pads | Fisher Scientific | 67622 | |
Surgical drape | Fisher Scientific | GM300 | |
Silk Sutures | Look/Div Surgical Specialties | SP115 | |
Nylon Sutures | Look/Div Surgical Specialties | SP185 | |
Durmont #5 forceps (2) | Fine Science Tools | 11251-35 | Angled 45° |
Surgical Scissors | Fine Science Tools | 14028-10 | |
3mm Vannas | Kent Scientific Corporation | INS600177 | Straight blade |
Hartman Hemostats | Fine Scientific Tools | 13002-10 | |
Occluding filaments | Washington University | Specialty order | Filaments are silicone coated at Washington Univeristy and provided to UTSW facilities for a fee. |
Evans Blue | Sigma Aldritch | E2129-10G | |
Filter Paper | Sigma Aldritch | WHA1001150 | 150 mm, circles, Grade 1 |
Weigh Boats | Fisher Scientific | 02-202-101 | 2.5" diameter |
0.9% Sodium Chloride Injection USP | Baxter Pharmaceutics | 2B1321 | |
0.3cc insulin syringe with 29 g needle | Becton Dickinson Labware | 309301 | |
Flat bottom restrainer | Braintree Scientific | FB M | 2.0" diameter |
TTC | Sigma | T8877 | |
10X PBS, pH 7.4 | Fisher Scientific | BP399-20 | |
Water Bath | Multiple sellers | Catalog number not available | Scintillation tubes with TTC may be manually held under running warm water as an alternative to the water bath. |
Styrofoam board | Multiple sellers | Catalog number not available | |
Large Syringe Kit | PumpSystems Inc | P-SYRKIT-LG | |
Perfusion Pump | PumpSystems Inc | NE-300 | |
60 cc syringe | Fisher Scientific | NC9203256 | |
27g winged infusion set | Kawasumi Laboratories, Inc | D3K1-25G 1 | |
20 ml scintillation vial | Fisher Scientific | 50-367-126 | |
Stainless steel spatula | Fisher Scientific | 14-373-25A | |
Alto acrylic 1.0 mm mouse brain, coronal | CellPoint Scientific | Catalog number not available | |
0.21 mm stainless steel blades, 25 pk | CellPoint Scientific | Catalog number not available | Reusable cryostat blades are an inexpensive alternative. |
4% paraformaldehyde | Santa Cruz Biotechnology | SC-281692 | |
Superfrost microscope slides | Fisher Scientific | 12-550-15 | |
HP Scanjet G4050 | Multiple sellers | Catalog number not available | Other commercial scanners are suitable for this step in the protocol. |
ImageJ | National Institute of Health | Catalog number not available | |
Analytical Balance | Mettler Toledo | XSE 205U | |
Precision Compact Oven | Thermo Scientific | PR305225M | |
1.7 mL microcentrifuge tubes (Eppendorfs) | Denville Scientific | C2170 | |
Formamide | Fisher Scientific | BP228-100 | |
96-well plates | Fisher Scientific | 07-200-9 | |
Epoch Microplate Spectrophotometer | BioTek | Catalog number not available |