The most notable symptom of migraine is severe head pain, and it is hypothesized that this is mediated by sensory neurons innervating the meninges. Here, we present a method to locally apply substances to the dura in a minimally invasive manner while using facial hypersensitivity as an output.
The cranial meninges, comprised of the dura mater, arachnoid, and pia mater, are thought to primarily serve structural functions for the nervous system. For example, they protect the brain from the skull and anchor/organize the vascular and neuronal supply of the cortex. However, the meninges are also implicated in nervous system disorders such as migraine, where the pain experienced during a migraine is attributed to local sterile inflammation and subsequent activation of local nociceptive afferents. Of the layers in the meninges, the dura mater is of particular interest in the pathophysiology of migraines. It is highly vascularized, harbors local nociceptive neurons, and is home to a diverse array of resident cells such as immune cells. Subtle changes in the local meningeal microenvironment may lead to activation and sensitization of dural perivascular nociceptors, thus leading to migraine pain. Studies have sought to address how dural afferents become activated/sensitized by using either in vivo electrophysiology, imaging techniques, or behavioral models, but these commonly require very invasive surgeries. This protocol presents a method for comparatively non-invasive application of compounds on the dura mater in mice and a suitable method for measuring headache-like tactile sensitivity using periorbital von Frey testing following dural stimulation. This method maintains the integrity of the dura and skull and reduces confounding effects from invasive techniques by injecting substances through a 0.65 mm modified cannula at the junction of unfused sagittal and lambdoid sutures. This preclinical model will allow researchers to investigate a wide range of dural stimuli and their role in the pathological progression of migraine, such as nociceptor activation, immune cell activation, vascular changes, and pain behaviors, all while maintaining injury-free conditions to the skull and meninges.
Migraine pain remains a major public health issue worldwide. The World Health Organization ranks it as the sixth-most prevalent disease in the world, afflicting just under 15% of the Earth's population1 and causing a substantial socioeconomic burden on society2,3. Treatment options and their efficacy have been suboptimal and only provide symptomatic relief and do not significantly modify pathophysiological events that underly migraine occurrence4,5. The lack of treatment success is likely due to migraine being a multifactorial disorder whose pathology is poorly understood, leading to a limited number of therapeutic targets. Migraine is also challenging to fully capture in animal models, especially given that migraine diagnosis is made based on verbal communication with patients who describe their experience with migraine hallmarks such as aura, headache, photophobia, and allodynia. Notwithstanding, it is important to note that recent advances in migraine treatments are currently outperforming treatments for many neurological conditions that have been well validated by preclinical models. For instance, monoclonal antibodies and small molecules that target calcitonin gene-related peptide, or its receptor have been very successful in improving the quality of life of migraine sufferers and can potentially transform the clinical management of migraine. While there has been advancement in understanding this disorder, there continues to be much yet to be elucidated.
Based on preclinical animal models and human studies, it is widely accepted that migraine headaches are initiated by aberrant activation of nociceptive fibers within the meninges that signal through the trigeminal and upper cervical dorsal-root ganglia6,7,8,9,10. Despite this theory, many studies still use systemic administration of drugs to understand underlying contributing mechanisms in migraine. While systemic dosing of drugs has substantially strengthened our understanding, these findings do not directly assess whether local actions within the target tissue of interest play a role in migraine. Conversely, several studies have taken an approach to stimulate the dura; however, these experiments require cannula implantation via an invasive craniotomy and extended recovery times11,12. Because of these limitations, we developed a minimally invasive approach to locally stimulate the dura where the lack of a craniotomy eliminates post-surgical recovery and allows for immediate testing in awake animals12,13,14. These injections are performed under light isoflurane anesthesia and administered at the junction of the sagittal and lambdoid sutures in mice.
Several approaches have been developed to evaluate nociceptive behavioral responses in rodents15. Cutaneous allodynia has been reported in approximately 80% of migraine sufferers16,17 and represents a potential translational endpoint for use in rodents. In preclinical models, the application of von Frey filaments to the plantar region of the rodent paw has been used to assess pain behaviors in preclinical migraine models. The primary limitation of this approach is that it does not test the cephalic region. Facial grimace scoring has been used to capture pain behaviors in rodents by analyzing facial expressions after induction of pain stimuli18,19. However, its limitations include only capturing responses to acute stimuli and not chronic orofacial pain conditions. Facial grooming and decreased rearing are also considered outputs of behavioral responses in preclinical models of migraine20,21. Limitations of the former include difficulty in differentiating pain responses from normal routine grooming and other sensations such as itch. In the case of the latter, rearing behaviors typically decrease quickly after the introduction of rodents to novel environments. Although each of these behavioral endpoints is valuable in the understanding of various mechanisms that contribute to pain conditions, there is a critical need for preclinical models of pain disorders such as migraine to include endpoints that specifically capture cephalic hypersensitivity responses. Assessing tactile hypersensitivity of the periorbital skin following dural stimulation is a method that may provide better insight into mechanisms contributing to migraines where sensory symptoms are predominantly cephalic in nature. Here, we describe a method to administer substances onto the mouse dura as a preclinical model of migraine. Following dural application, we also present a detailed method for testing periorbital tactile hypersensitivity using calibrated von Frey filaments applied in the Dixon up-down method.
All procedures were conducted with prior approval of the institutional Animal Care and Use Committee at the University of Texas at Dallas. ICR (CD-1) (30-35 g) and C57/BL6 (25-30 g) mice aged 6-8 weeks were used in this study.
1. Dural infuser
2. Dural injections
3. Periorbital von Frey
4. Testing for baseline withdrawal thresholds
5. Analysis of von Frey results
This injection method is used to administer stimuli onto the dura of mice so that subsequent behavioral testing may occur. The most common behavioral output measured with this model is cutaneous facial hypersensitivity assessed via von Frey12,13,14. Here we show how this model can be used to assess potential sex-specific contributions to migraine pathology (Figure 3).
This procedure has been used to examine the effects of dural prolactin (PRL) on mechanically evoked facial hypersensitivity14 (Figure 3). The results of this study demonstrated that female ICR mice show significantly reduced facial withdrawal thresholds in response to 5 µg of dural prolactin (Figure 3A). A ten-fold lower dose of 0.5 µg of prolactin (PRL) also showed responses similar to high dose of PRL (Figure 3B).
These injections have also been shown to produce spontaneous pain-related behaviors assessed via grimace. Dural 0.5 µg of PRL caused significant grimacing in female mice (Figure 3C), further demonstrating a clear role for dural PRL in female migraine-like behaviors. We performed grimace assays prior to all testing with von Frey filaments.
Figure 1: Dural infuser and injection placement. (A) The injectors/infusers consist of a modified cannula adjusted to the length of ~0.5 mm- 0.65 mm and attached to a needle cemented on a 10 µL gas-tight syringe via tygon tubing. (B) Aerial view of marked injection site location on the head of the mouse. (C) (Left panel) Diagram of the location of the dural injection. Placement of the injection is on the junction of the lambdoid and sagittal sutures at approximately 4.8 mm posterior to bregma. (Middle panel) Post-mortem aerial view of a mouse skull following dural injection of 5 µL of blue injection dye. (Right panel) Separation of the mouse skullcap from the brain. There was no observable leakage of blue injection dye on the brain. Please click here to view a larger version of this figure.
Figure 2: von Frey Testing chambers. (A) von Frey testing chamber composed of 3.5 in x 3.5 in individual acrylic chambers with lids placed on a wire mesh rack. These are connected via columns of 10 chambers organized in 2 rows. (B) Example of mice in their individual cups housed inside the von Frey testing chambers. Please click here to view a larger version of this figure.
Figure 3: Dural application of prolactin induces behavioral responses in mice. Mechanical withdrawal thresholds were assessed following dural application of PRL (5 µg or 0.5 µg) in female mice. (A) Application of 5 µg of PRL (n = 7 PRL, n = 6 vehicle) induced facial hypersensitivity compared to vehicle. (B) Application of 0.5 µg of PRL (n = 5 PRL, n = 4 vehicle) induced long-lasting facial hypersensitivity. (C) Grimace was also assessed in the same mice treated with 0.5 µg of PRL at each time point. These mice exhibited significantly higher grimace scores compared to the mice treated with vehicle. Statistics: Two-way ANOVA followed by Bonferroni multiple comparison post-hoc analysis. Data are represented as means ± SEM. *p < 0.05, ****p < 0.0001. Please click here to view a larger version of this figure.
Maladaptive changes in the local nociceptive system in the dura are considered a key contributor to the headache phase of migraine attacks despite a lack of tissue injury25,26. Here the study presents a method whereby minimally invasive stimulation of the dura can induce facial tactile hypersensitivity. Elucidating the mechanisms and events involved in dural nociceptor activation without causing damage to the cranium and tissues may more accurately reflect migraine mechanisms in a preclinical model.
Craniotomy and cannula implantation have long been used to assess functions and mechanisms that contribute to migraine pain11,12. However, it has been reported that a craniotomy can induce activation of dural mast cells and increase pial vascular permeability in rodents27. Given that mast cell activation in the dura is highly implicated in migraine7,8,28,29, this technique has major caveats that may skew the interpretation. Administering substances through the junction of the sagittal and lambdoidal sutures effectively diminishes the activation of nociceptors mediated by craniotomy-induced mast cell activation. Moreover, non-invasive dural stimulation does not require post-surgical recovery and administration of analgesics which may alter the interpretation of results. Local application of substances onto the dura allows researchers to focus on this specific target tissue, as opposed to systemic administration of drugs where the site of action is not easily determined12,13,14. While systemic administration of substances such as nitroglycerin and calcitonin gene-related peptide trigger experimental attacks in humans that are similar to migraines, they do not allow for assessment of the location of action in rodent models; more targeted tissue-specific models offer an alternate approach.
This technique described here involves injecting a drug or other solution directly onto the dura mater of the meninges through the junction where the sagittal and lambdoidal sutures of the skull meet. For best results, ICR (CD-1) or C57/BL6 mice aged 6-8 weeks should be used for these experiments. Younger mice may be used; however, the use of ICR (CD-1) mice that are older than 8 weeks are not recommended as their skull plate sutures are typically completely fused by this age, making it impossible to inject without damaging the skull. It is also critical to consider the weight/size of each mouse that will undergo this procedure. It is recommended that these injections are performed on animals that have a weight greater than 19 g as the skull is typically very thin at lower weights and may not withstand the pressure applied during the injection. Of importance, there are also likely factors that contribute to the age/weight at which skull plate fusion occurs (e.g., the composition of lab chow used in animal facilities). Therefore, experimenters may need to determine the age/weight range suitable under their own conditions. Different age ranges and animal weights may be required for other mouse strains or genotypes, depending on when the skull plates fuse in those animals, and may also require optimization of the injection itself.
When learning or practicing this technique, it is highly recommended that a level of comfort is obtained with locating the suture junction in euthanized mice. It may be best to first practice with the scalp excised or peeled back in these mice and slowly advance to locating the junction through the skin. Once establishing the precise location, inks and dyes can be injected into the dura to verify location accuracy and depth of the injection. This technique was developed and optimized using ICR (CD-1) mice (30-35 g) and C57/BL6 mice (25-30 g). An infuser length of 0.5-0.6 mm is sufficient to inject a mouse weighing within the range of 25-35 g. However, the length of the infuser may need to be calibrated if injecting mice that significantly differ from the mice used to optimize this technique. For example, a mouse smaller than 25 g would likely result in the use of an infuser that has a length less than 0.5 mm. Upon mastering this technique and when performed in age-appropriate mice, the success rate of this injection can be close to 100%; however, complications with the injection may stem from issues such as breaking the skull due to applying too much force to insert the infuser as well as abnormal bleeding caused by damaging meningeal blood vessels.
Alterations in tactile sensitivity are an important measurement when assessing pain behaviors in rodents. Here we demonstrate the use of periorbital von Frey testing to assess these behaviors in a preclinical migraine model. A major advantage of using this technique in migraine models is that we can assess hypersensitivity of the head, which has more relevance than other non-cranial locations such as paws. The critical step to ensure reproducible results is to make certain that the mice are fully baselined. This will require a well-trained experimenter that can apply von Frey filaments precisely. It is likely that it will take approximately 7 days for an animal to reach baseline. However, it is possible that not every animal will reach the targeted baseline. In our experience, after about 7 days of working with mice, only 60%-70% of animals will reach a baseline of 0.6 g in the periorbital region, but this is dependent on the cohort of animals. This timing should be considered prior to beginning an experiment to ensure sufficient numbers are used to account for dropout and that animals are the proper age post-baseline for using this non-invasive method to stimulate the dura. The steps for determining a baseline are outlined in protocol section 4.
A limitation to von Frey testing is that it can be difficult to distinguish between pain responses and routine grooming/itch. To help distinguish pain from grooming, it is important to notice the length of time this behavior occurs. Usually, a pain response is one swipe following the filament application, while grooming behaviors tend to be prolonged and can last for several seconds to minutes. If the grooming/itch behavior cannot be distinguished from a hypersensitive response, it is best not to record this as a response. Additionally, improper filament placement (e.g., filament slipping) can result in prolonged grooming of the animal, making it difficult to test properly. If this happens, the experimenter should wait until grooming has stopped and the mouse is calm enough to test. Continue from the same filament used prior to the beginning of the grooming behavior. If the mouse continues for very long bouts of time, place the mouse back in the testing chamber for approximately 5 min. Once the 5 min have passed, try testing the mouse again. If this behavior continues with no resolve, the mice have to be removed from the study. Of importance, it is not recommended to shave the fur on the face as it is unclear whether mouse skin retains the same sensitivity after hair is removed, and the process of hair removal (shaving, depilatory creams) may also influence skin sensitivity.
In most situations, it is ideal for administering substances onto the dura no more than 24 h after the mouse has reached baseline. It is recommended that mice are subjected to von Frey filament testing once per hour. If possible, testing every other hour gives enough time for the animals to calm down after testing. Additionally, experiments should be timed as not to interfere with their circadian patterns. Alterations to the circadian rhythm in mice may alter behavioral phenotypes and ultimately result in irreproducible results.
Periorbital von Frey testing can be used in combination with other behavioral assays to strengthen experimental conclusions. The grimace scale relies on spontaneous facial expressions in rodents rather than evoked responses18,19. This method has high accuracy and reliability when assessing and quantifying acute pain behaviors and has been used in many preclinical models of migraine12,30. When using both grimace and periorbital von Frey assays, the experimenter should consider scoring for grimace prior to application of von Frey filaments to the periorbital region of the mouse. This ensures that the grimacing behavior is spontaneous and not evoked by filament application. Hindpaw mechanical hypersensitivity can also be used in conjunction with periorbital von Frey testing. Contrary to grimace scoring, it is best to test facial hypersensitivity prior to assessing hind paw hypersensitivity. Hindpaw testing requires that the mouse is placed back in the chamber without the cup after periorbital von Frey testing is completed.
In conclusion, periorbital von Frey testing and non-invasive dural stimulation in mice add valuable options to the current range of preclinical models of migraine. When performed correctly, this technique presents a refined approach to generating a headache-like phenotype in rodents, as it does not require surgical implantation of a cannula. In rats, cannulas are prone to bacterial infection, can become clogged, may fall off, and require each animal to be single-housed, creating unnecessary stress on the animal. Furthermore, the dural stimulation protocol can easily be modified to use with several drug applications. Periorbital von Frey testing paradigms can also be modified to best fit the experimental specifications. Additionally, periorbital von Frey testing can be used in other orofacial pain disorders. These techniques are an important tool to help further understand the complex underlying mechanisms of migraine pain.
The authors have nothing to disclose.
This study was supported by the National Institutes of Health (NS104200 and NS072204 to GD).
4 oz Hot Paper Cups | Choice Paper Company | 5004W | https://www.webstaurantstore.com/choice-4-oz-white-poly-paper-hot-cup-case/5004W.html |
Absorbent Underpads | Fisherbrand | 14-206-65 | https://www.fishersci.com/shop/products/fisherbrand-absorbent-underpads-8/p-306048 |
C313I/SPC Internal 28 G cannula | P1 Technologies (formerly Plastics One) | 8IC313ISPCXC | I.D. 18 mm, O.D. 35 mm |
Gastight Model 1701 SN Syringes | Hamilton | 80008 | https://www.hamiltoncompany.com/laboratory-products/syringes/80008 |
Ismatec Pump Tubing, 0.19 mm | Cole-Palmer | EW-96460-10 | https://www.coleparmer.com/i/ismatec-pump-tubing-2-stop-tygon-s3-e-lab-0-19-mm-id-12-pk/9646010 |
Stand with chicken wire | Custom | The galvanized steel chicken wire dimensions are 0.25 in. x 19-gauge | |
Testing Rack with individual Chambers | Custom | Each chamber should have a division between each mouse and lids to contain the mouse. The chambers should also be large enough to hold a 4 oz. paper cup. | |
von Frey Filaments | Touch test/Stoelting | 58011 | https://www.stoeltingco.com/touch-test.html |