**3. Animal Models of Migraine**

Animal models have been developed with the aim to comprehend migraine disorder, even though all of them show shortcomings. To be considered reliable, animal models should display a similar etiology and phenotype to human migraine. Although this disorder is considered complex with a variable phenotype, there is currently no animal model able to replicate all its features [52]. In this paragraph, we consider murine models developed to study the migraine disorder. In Table 3, we summarized the current animal models of migraine in which rats or mice were used.

Currently, animal models focus on activation of nerves in TCC. Nociceptors are located in the terminal structures of meningeal and trigeminovascular afferents deriving from the ophthalmic division of the trigeminal nerve that innervate intracranial structures sensitive to pain, such as the dura mater and meningeal vasculature, large cerebral arteries and the paranasal sinuses. Headaches similar to migraine can be caused by stimulation of nerves that innervate these structures [53,54]. Animal models based on chemical provocations that use different vasodilating agents are probably the most investigated in preclinical research. The administration of a mix of inflammatory mediators, named "inflammatory soup", e.g., using a mixture composed by prostaglandin (PGE2), histamine, 5-HT and bradykinin, has been used to stimulate meningeal and trigeminovascular nociceptors [55]. In this model, the inflammatory soup can be administered by injection using a micro-catheter placed into the cisterna magna through the atlanto occipital membrane of animals that received anesthetics. This injection of chemicals induces the activation of the primary sensory fibers supplying the meninges. In addition, topical application on the dura mater of rats is also used causing a reversible cephalic mechanical sensitivity [56–58]. Potential limitations of this technique are linked to the use of chemicals that can compromise the functionality of the blood-brain-barrier (BBB), resulting in activation of central sites directly rather than synaptically by the activation of meningeal afferent fibers [52].

Another model widely used, based on nitroglycerin (NTG), nitric oxide donor glycerol trinitrate, intravenously or intraperitoneally administered, has been studied as method to provoke migraine-like pain [52]. In rodents, hyperalgesia provoked by NTG has been used to develop a model to study sensory hypersensitivity associated with migraine [59,60]. The infusion of NTG in mice caused thermal and mechanical allodynia, symptom reversed by sumatriptan [59]. Moreover, in a transgenic mouse model of familial migraine that was studied, animals that expressed a human migraine gene, casein kinase 1δ, in which a major sensitivity to hyperalgesia induced by NTG, compared to controls, was observed [61]. In addition, NTG-administered in mice was able to induce aversion to light and increased meningeal blood flow [60,62]. This mouse model was used to develop a test able to model the progression of the disorder from an acute to a chronic condition by means intermittent injection of NTG. This modality of treatment evoked acute hyperalgesia and developed a progressive basal hypersensitivity to mechanical stimulation [63]. Both acute hyperalgesia and hypersensitivity were blocked by topiramate and propranolol, whereas hyperalgesia was inhibited by sumatriptan, suggesting that this model may be considered in the screening of novel therapies in migraine treatment [64]. The use of NTG in rodents may effectively model migraine-like symptoms [65], even though doses, type of administrations and the determination of the time of observation need to be carefully monitored when animal models based on NTG are adopted for the study of the TGVS [52].

Medication overuse headache (MOH) is a condition that in predisposed subjects affected by migraine or tension-type headache typically occurs. In these conditions, an overusing of drugs such as triptans and opioids, e.g., for more than 10 days per month over 3 months, may transform the headache from an episodic to a chronic state. In rodents, MOH may be used as a model in which repeated administration of these drugs induced a long-lasting state of latent sensitization [66]. It has been reported that TEV-48125, a humanized CGRP antibody, inhibited cutaneous allodynia both induced by bright light stress and NO donor in a MOH model, in which rodents were prior treated with sumatriptan or morphine, thus highlighting the importance of this model in discovering migraine medications [67].

In recent decades, several gene mutations have been correlated with some forms of severe and rare migraine by the use of an approach based on classic linkage analysis. This analysis contributed to sustain the hypothesis that an inherited trait is implicated in this disorder. A translational step offered by transgenic mouse technologies, in supporting the understanding about the pathophysiology of migraine and the evaluation of new therapeutic targets, is represented by behavioral characterization of genetic models of migraine [68]. The familial hemiplegic migraine (FHM) is a rare monogenic migraine, in which affected subjects show a severe hemiplegic aura accompanied by weakness perceived on one side of the body. Specifically, three gene mutations were identified to date, CACNA1A (FHM1), ATP1A2 (FHM2), and SCNA1A (FHM3), encoding for subunits of ion channels and transporters that show to have a role in neurotransmission [69]. These mutations, respectively encode for subunits of voltage-gated calcium channels, sodium-potassium ATPases, and voltage-gated sodium channels [70]. In this frame, genetic models of migraine have been created by inserting human mutated genes in the mouse genome obtaining knock-in (KI) mouse models, respectively, two of FHM1 [71,72], and one of FHM2 [73]. The humanized FHM1 KI mouse models contain gain of function missense mutations (R192Q or S218L) in the CACNA1A gene, one of the most investigated genes whose product is the pore-forming α1A subunit of Cav2.1 channels (P/Q type) (55). S218L mice showed a phenotype similar to a severe clinical phenotype of patients showing the same mutation (FHM1 S218L) [72]. In R192Q mice, decreased CGRP-immunoreactivity was observed when compared to controls, in cells of TG and in the superficial laminae of the TCC, suggesting that alteration in the expression of CGRP is induced by FHM-1 CACNA1A mutation [74]. These models showed decreased neuronal response to nociceptive activation of the TGVS in comparison with controls [75], suggesting that they show different reactions to common nociceptive signals observed in migraine. In addition, enhanced susceptibility in showing spontaneous pain behaviors correlated to nociceptive headache and photophobia was highlighted [76,77]. Only one KI mouse model for FHM2 expressed the loss of function W887R missense mutation in ATP1A2 [73], whereas no FHM3 KI mouse model has been developed yet.

In a study based on a rodent model, during the process of evaluating baseline periorbital von Frey thresholds, a male rat affected by spontaneous episodic trigeminal allodynia was discovered [78]. This characteristic was noticed by episodic alteration of periorbital pain threshold. The mating demonstrated that this trait is inheritable in both sexes and with a diversity of phenotypes. Some animals showed a similarity to chronic migraineurs, with thresholds lower than normal for more than two weeks per month. Other animals showed a similarity to episodic migraine, manifesting periods of normal thresholds and periods of lower thresholds. Chemicals such as sumatriptan, ketorolac and DHE were tested in order to validate this model, temporarily reversing the pain thresholds. Furthermore, the treatment with valproic acid for a period of one month blocked spontaneous changes in trigeminal allodynia. After the discontinuation of the treatments, the animals returned to the initial baseline. This study might be considered a unique model of spontaneous allodynia with phenotypes similar to migraine, providing a possible predictive model for drug development and for the investigation of the pathophysiology of spontaneous episodic trigeminal pain disorders [78]. Research based on these genetic models offers an advantage due to the fact that animals are manipulated in order to have predefined phenotypes similar as possible to migraine features. Nonetheless, the disadvantages might regard the common polygenic forms of this disorder in the general population, in which these specific mutations might not be relevant.

Migraine aura and cortical spreading depression (CSD) are transient neurological deficits highlighted in around 30% of migraineurs [79]. CSD is a phenomenon described as a slow wave of depolarization of neuronal and glial cells in the cortex that can be induced experimentally. In rodent models, injections of KCl showed to initiate this phenomenon in the cortex. CSD visualization is operated measuring the electrical activity of cortical neurons by means implanted microelectrodes or changes in cerebral blood flow through laser doppler flowmetry [52]. Furthermore, CSD can be also induced by electrical or mechanical stimulation of the cortex [80]. Progresses in the study of CSD mechanisms have focused to elucidate whether it plays a part in provoking the trigeminovascular activation, resulting in migraine triggering. In rodents, studies based on imaging showed that CSD induces vasodilation of meningeal blood vessels [81], and enhanced neuronal activation at the level of trigemino nuclear complex and in higher cerebral areas of the trigeminal pain pathway [82]. Considering that migraine frequently occurs without aura, the mechanisms of CSD require further studies to be better clarified and for drug discovering in migraine aura treatment.


#### **Table 3.** Current animal models of migraine.
