Light-aversion and cephalic allodynia in an intravenous CGRP model of migraine-like behaviour in male and female rats

Abstract

Background

Photophobia is a common and highly bothersome migraine-associated symptom, where light may cause discomfort and exacerbate headache pain. Calcitonin gene-related peptide (CGRP) may be involved in nociceptive signalling during photophobia. In clinical provocation studies, intravenous CGRP induces both migraine attacks and photophobia. Animal studies attempting to model this have used intraperitoneal or intracranial injections and behavioural tests requiring very high light intensities to provoke light-aversive behaviour, a surrogate for photophobia-like responses. Here, we evaluated the effects of intravenous CGRP on light aversion alongside periorbital sensitivity in naïve rats.

Methods

Male and female Sprague Dawley rats received intravenous injections of 100 µg/kg human αCGRP or saline (n = 6–12/group). After 20 min, light aversion was assessed in an optimized light/dark box (LDB) test. The LDB was optimized by adding bedding material to minimize novelty-related anxiety. Following a washout period, the same rats were re-dosed and tested in either a dim/dark box (DDB) test or an open field test (OFT) to evaluate anxiety in the absence of bright light. Cephalic allodynia, used as an indicator of headache-like pain, was assessed with von Frey measurements.

Results

CGRP-injected rats of both sexes spent significantly less time in the bright compartment of the LDB than controls. This avoidance was absent in the DDB test, and no CGRP-related effect was seen in the OFT, indicating that the effect was light-specific rather than anxiety-based. CGRP also significantly increased periorbital sensitivity, which was not seen following saline injection. Housing conditions and test order influenced outcomes, with female rats seemingly most affected. Grouping animals by treatment and using naïve rats for von Frey appeared to provide the most consistent results.

Conclusions

Intravenous CGRP induced light aversion in both male and female rats, supported by a lack of effect in complementary anxiety tests. The same dose also induced cephalic allodynia, suggestive of headache-like pain. These findings show parallels to clinical provocation studies and highlight the translational potential of this intravenous model for studying migraine-related photophobia-like behaviour. Moreover, the study underscores that housing and test order can strongly influence behavioural outcomes, emphasizing the importance of experimental design in preclinical migraine research.

Graphical Abstract

Supplementary Information

The online version contains supplementary material available at 10.1186/s10194-026-02278-2.

Background

Photophobia, or hypersensitivity to light, is a prominent symptom in headache disorders, particularly migraine, affecting approximately 80–90% of patients [1, 2]. It has been reported as one of the most bothersome associated symptoms, frequently exacerbating the headache, and is included among the diagnostic criteria for migraine in the International Classification of Headache Disorders, which defines it as “hypersensitivity to light, even at normal levels, usually causing avoidance” [1, 3].

Despite its prevalence and clinical importance, the underlying mechanisms of photophobia remain incompletely understood, much like those of migraine pain itself [4]. Several neuroanatomical and neurochemical pathways have been proposed, with particular focus on the trigeminovascular system, where calcitonin gene-related peptide (CGRP), a major pro-nociceptive neurotransmitter and well-established key player in the pathophysiology of migraine, has emerged as a central candidate strongly implicated in the neurobiology of photophobia by both clinical and preclinical studies [5–7]. Anti-CGRP therapy has proven effective in alleviating migraine headache and associated symptoms, including photophobia [8]. Correspondingly, clinical provocation studies have shown that intravenous CGRP infusion can induce migraine attacks and increase light sensitivity in people with migraine [9, 10]. Preclinical studies in mice likewise demonstrate that both intraperitoneal (peripheral) and intracranial (central) administration of CGRP can induce light aversive behaviours, indicative of photophobia-like responses, further supporting a role for CGRP in photophobia [11–13]. These findings are consistent with a predominantly peripheral site of action of CGRP in migraine and photophobia, as peripherally administered CGRP, and to a large extent, its antagonists, do not cross the blood-brain-barrier [14].

Although intravenous CGRP is widely used in human provocation studies, this route has, to our knowledge, not been applied in preclinical migraine models assessing light-aversive behaviour. Furthermore, headache-like pain is rarely evaluated alongside photophobia-like behaviour and has not been examined following intravenous CGRP administration in animal models, further limiting translational insight. To address this, the present study investigated the effects of intravenous CGRP on light-aversive behaviour, serving as a behavioural surrogate for photophobia-like responses, in male and female Sprague Dawley rats. Using behavioural paradigms such as the light/dark box, dim/dark box, and open field test, we sought to disentangle light-induced aversion (photophobia-like behaviour) from anxiety-related behaviours. In parallel, and to better mirror clinical findings of headache-associated photophobia, we evaluated cephalic allodynia – widely used as an indicator of headache-like pain in animal models and a common feature in migraine patients often linked to photophobia – by measuring periorbital mechanical sensitivity following intravenous CGRP using the von Frey test [15, 16].

Methods

Animals

A total of 99 male (300–400 g) and 48 female (220–280 g) Sprague-Dawley rats (Taconic, Ejby, Denmark) were used. Rats were housed in Eurostandard Type VI cages in groups of three to five. The rats were provided ad libitum access to standard chow and water and maintained in a controlled environment at 22 °C ± 2 °C with 55 ± 10° % relative humidity. They were kept on a 12 h light/dark cycle with ambient lighting of approximately 35 lx in the cage. All animals were allowed a minimum of seven days to acclimate before handling and were individually numbered. The rats were randomly assigned to treatment groups separately within each sex using stratified randomization based on body weight, ensuring similar mean weights across groups.

This study complied with the European Communities Council Directive (86/609/ECC) and received approval from the Danish Animal Experimentation Council (Licenses: 2020-15-0201-00708 and 2020-15-0201-00751).

Drugs

Human αCGRP (Bachem AG, Bubendorf, Switzerland, (prod.no. 4013281)) was reconstituted in filtered (0.2 μm) ultrapure sterile water to 1 mM, and further serially diluted to the desired concentration (15, 45, or 100 µg/ml) in sterile PBS. Saline (0.9% NaCl) was used as control. Both CGRP and saline were delivered intravenously (i.v.) via a tail vein at 1 ml/kg, corresponding to CGRP-doses of 15, 45, and 100 µg/kg.

Blood pressure measurement

The effect of i.v. administered CGRP on blood pressure was assessed in male rats (n = 10) using an invasive femoral artery blood pressure device. The rats were placed under general inhalation anaesthesia (isoflurane 2–5%, air 0.9 L/min, and O₂ 0.35 L/min) and femoral artery access and catheterization were carried out as previously described [17]. In short, an incision was made in the inguinal area and the femoral artery was exposed by blunt dissection. Once isolating and ligating the artery a small cut was made and a rat femoral artery catheter (Instech Laboratories, PA, USA (art.no. C10PU-RFA1310)) was inserted. Before insertion, the catheter was filled with heparinized saline (10 IU/ml) and once placed in the artery, the catheter was flushed with sterile saline (heparin free) every 1–2 min to prevent clotting. Blood pressure was measured using a blood pressure device (MLT0670 disposable BP transducer, ADinstruments, Dunedin, New Zealand) attached to the intraarterial catheter, and data was acquired using PowerLab 8/30 and LabChart software version 6.0.30 (ADinstruments, Dunedin, New Zealand). Mean arterial pressure (MAP) was recorded for approximately 5 min prior to CGRP administration (15 µg/kg (n = 3), 45 µg/kg (n = 3), or 100 µg/kg (n = 4)), and continuously for 20–30 min following the injection. Average MAP was calculated for four defined time periods: from start of recording and until CGRP administration (baseline MAP); 0–2 min post-administration; 5–10 min post-administration; and the final 5 min before recording stopped (about 20–25 min post-administration). For each of the four time periods average MAP was calculated by averaging the 5-second mean values obtained from consecutive 5-second bins. Short bursts of MAP fluctuations during catheter flushing were not included in the average. The animals were euthanized after the last measurement by decapitation while still under general anaesthesia.

Fundus imaging

Fundus imaging was performed to assess the time course of systemic vasodilation following a single i.v. injection of CGRP. Male rats (n = 3) were anesthetized (isoflurane), and the eyes were treated with a local anaesthetic (Minims^® Oxybuprocain 0.4%, Bausch&Lomb, Vaughan, ON, Canada) and mydriasis was induced using tropicamide eye drops (10 mg/ml, Tropicavet, VAPP, Carnaxide, Portugal), and finally a lubricant was applied to both eyes. Pictures of the fundus were taken using a fundus camera for rodents (Phoenix micron i.v., Micron Discover software v.2.2, Phoenix research laboratories Inc, Pleasantin, CA, USA) before i.v. administration of 100 µg/kg CGRP and then every 30 s for 10 min, again at 15 min and finally at 60 min following the injection.

Photosensitivity assessment

In order to assess potential light aversive behaviour following CGRP injection, a new set of rats were evaluated in a light/dark box (LDB) test combined with a low-light anxiety test; the dim/dark box (DDB) test or the open field test (OFT). Behaviour in these tests was always recorded for 10 min using a camera mounted above the arena and tracking software (ANY-maze video tracking system v.7.48, Stoelting Co., Wood Dale, IL, USA) was used to quantify time spent in each zone/compartment, distance travelled, and time spent resting (immobile for minimum 5 s). Rats were tested in a random cage-wise fashion, the experimenter was aware of treatment allocation but had no influence on the automated tracking of behaviour.

Light/dark box test

The LDB setup consisted of two identical Plexiglas boxes, each box measuring 50 × 50 cm with a height of 34.5 cm. The light box, open at the top, was illuminated using Philips Hue light strips, overhead lighting, and a floor lamp to create a bright environment. The dark box featured opaque black walls and a lid, creating a pitch-black environment. Infrared light was used for video recording in the dark box. A 10 × 10 cm opening connected the two boxes, allowing the rats to freely move between them. Each trial began with the rat placed at the centre of the light box, oriented away from the opening to the dark box. The desired lux was set according to the test and measured in the centre of the open box using a Mini Light meter (UT383, Uni-Trend Technology, Guangdong, China).

Dim/dark box test

The DDB test was performed in the LDB setup but with very low light intensity (2 lx) in the open compartment.

Open field test

The OFT arena measured 1 × 1 m with 34.5 cm high anti-reflective walls to prevent escape. The area was divided into zones: the outer zone, defined as a 25 cm border along the inside walls, and the centre zone, the inner square measuring 50 × 50 cm. To minimize light aversive behaviour in the centre zone the room was dimly lit at 10 lx. Each trial began with the rat placed in the centre of the OFT arena.

Electronical von Frey

Mechanical sensitivity of the periorbital and plantar regions was assessed as previously described using a handheld electronic von Frey (EvF) anaesthesiometer (IITC, Life Science Inc, Los Angeles, CA, USA), that utilizes a pressure-sensitive filament to quantify withdrawal thresholds [18–20]. A blunt EvF filament was applied with gradually increasing pressure up to 400 g or until a clear withdrawal response was observed, at which point the applied force was recorded. Periorbital pressure was recorded at three periorbital sites and one control neck site, the first periorbital site being tested twice (Fig. 1). Withdrawal threshold was calculated as the average pressure of the three periorbital areas, only including the second measurement at the first site. The rat was placed in the lap of the person performing the test, gently held in a kitchen towel with only the head exposed. For plantar assessment, four distinct areas between the footpads of the right hind paw were tested, the first site being tested twice, first and last (Fig. 1). Plantar withdrawal threshold was calculated as the average pressure recorded at the four sites, only including the second measurement at the first site. The plantar test was performed right after the periorbital. Here the rat was again gently held in the towel but this time with the right foot exposed and the body and head covered by the fabric. A habituation von Frey session was performed a day or two before baseline measurements to familiarize the animals to the procedure. Ambient light intensity was approximately 100 lx during the tests. To maintain blinding, both investigators conducting the tests wore coloured glasses to obscure visible signs of compound allocation (erythema on ears and paws following i.v. CGRP).

Fig. 1.

Electronical von Frey test. Pressure points for periorbital (right) and plantar (left) von Frey measurements. For periorbital measurements the filament was applied to three periorbital areas, the first tested twice (measurements 1 and 4), and finally a control point in the neck (measurement 5) was tested. For calculation of average pressure threshold measurements 2 through 4 were used. For plantar measurements four points between the footpads were tested, the first (point 1) was measured again in the end (point 5). For calculation of average pressure threshold measurements 2 through 5 were used. Created with http://www.BioRender.com

Study design for behavioural assessment

Optimizing the light/dark box test (Pilot study)

To refine the LDB into a more light-focused assay, we sought to reduced novelty-related anxiety and minimize overall stress by adding a thin, evenly distributed layer of home-cage bedding material across the floors of both compartments, aiming for naïve rats to spend more time in the light compartment. This condition was then compared with the standard empty-arena setup in male SD rats (n = 46) to assess its effect on time spent in the light compartment. The bedding used (dust-free wood granulate bedding and paper shreds nesting material) came from the rat’s own home cage. It has been demonstrated that increasing light intensity reduces the exploratory drive in the LDB test [21], therefore the effect of bedding was evaluated at different light intensities; 300 lx (n = 4 with bedding, n = 4 no bedding), 1900 lx (n = 16 with bedding, n = 16 no bedding), and 4700 lx (n = 6 with bedding). 4700 lx was the maximum light intensity achievable with our setup. With 4700 lx in the open box the light intensity in the centre of the dark covered box was measured at 3 lx, and < 3 lx in the darker corners.

Intervention study (Part 1, 2, and 3)

To assess the effects of i.v. CGRP on light-aversive behaviour and cephalic allodynia, a series of behavioural tests were performed on male and female SD rats (Fig. 2). The rats were randomly assigned to i.v. injection with either saline (control) or 100 µg/kg CGRP. Group sizes were based on prior experience with similar behavioural and von Frey assays. In the first part of the study (Part 1), the rats were housed in a randomized manner, mixing both administration groups in each cage. In the second and third part (Parts 2 and 3) new rats were used, and they were housed according to compound allocation. The purpose of housing by administration group in Part 2 and 3 was to minimize emotional contagion or social analgesia, which could otherwise lead rats to mirror the distress or discomfort of cage mates or modulate their pain behaviour, a phenomenon that might be particularly relevant in females [22–26]. Rats of both study setups had the same home cage during the entire study period.

Fig. 2.

Study design for assessing light aversive behaviour and mechanical sensitivity following CGRP provocation. Male and female Sprague Dawley rats were randomized into two equally sized treatment groups, saline (control) and CGRP (100 µg/kg). Part (1) Three behavioural tests were performed with minimum 1 week wash-out period between each test. The rats were randomized to a new or the same compound prior to each new test. No rat received more than two CGRP injections. Twenty-four male rats were tested first in the LDB (saline: n = 12, CGRP: n = 12), then in the EvF (saline: n = 12, CGRP: n = 12), and finally in either the OFT (saline: n = 6, CGRP: n = 6) or the DDB test (saline: n = 6, CGRP: n = 6). Twenty female rats were tested first in the LDB (saline: n = 10, CGRP: n = 10), secondly in the OFT (saline: n = 10, CGRP: n = 10), and finally in the EvF test (saline: n = 10, CGRP: n = 10). The rats were randomly housed, with both injection-groups represented in each cage (mixed cages). Part (2) In the second study the rats were housed based on compound allocation (matched cages). Twelve female rats were tested in the LDB (saline: n = 6, CGRP: n = 6) and next in the DDB test where each rats switched group (saline: n = 6, CGRP: n = 6). Part (3) In the final study the rats were again housed according to injection allocation (matched cages). Sixteen male rats (saline: n = 8, CGRP: n = 8) and 16 female rats (saline: n = 8, CGRP: n = 8) were tested in the EvF only. CGRP; calcitoningene-related peptide, DDB; dim/dark box, EvF; electronical von Frey, i.v; intravenous, LDB; Light/dark box, OFT; open field test. Created with http://www.BioRender.com

In Part 1, female and male rats were first assessed in the LDB test, then in either the EvF test followed by an anxiety test (OFT or DDB), or in the OFT followed by EvF measurements. In Part 2, female rats were assessed in the LDB test followed by a DDB test. In Part 3 both male and female rats were assessed in the EvF test only, with no previous injection or testing performed (Fig. 2).

Following injection, the rats were allowed to rest in a completely dark room for 20 min before being exposed to a behavioural test. After testing each rat, the rat was returned to its home cage, and the testing arena was cleaned with disinfectant wipes and dried off before the next rat was tested. When animals were reused for further tests, they were always allowed at least one week of rest and wash-out before the next test. They were either re-randomized to the next injection (Part 1) or just switched to the opposite administration group (Part 2). Based on the pilot study result the LDB and DDB tests were performed with home cage bedding material in both compartments. No bedding was used in the OFT. Light intensity in the open compartment of the LDB was set to 4700 lx.

Mechanical sensitivity measurements were obtained before injection (baseline) and again 20 min after i.v. administration of saline or 100 µg/kg CGRP. In Part 1 EvF was performed as the second or third behavioural test. In Part 3 EvF was performed as the first and only test, on rats previously naïve to CGRP as well as i.v. injections.

Statistical analysis

Statistical analyses and graphical illustrations were performed using GraphPad Prism (version 10.4.1, GraphPad Software, Boston, MA, USA). Data were analysed using 2way ANOVA and Sidak’s multiple comparisons test for normally distributed data and Friedman’s test with Dunn’s multiple comparisons test for non-parametric data. When comparing two time-points within groups, paired t-tests were used. A p-value < 0.05 was considered statistically significant. All reported statistics are presented in Suppl. Table S1.

Results

CGRP caused a marked but short-lived MAP reduction and long-lasting peripheral vasodilation

CGRP has been shown to affect blood pressure even at very low doses, causing a drop in arterial pressure as a result of systemic vasodilation following i.v. administration [27, 28]. There is a risk that a critical drop in blood pressure following i.v. delivery would change the behaviour of the rat, reducing locomotor and exploratory behaviour in a subsequent behavioural test. Before evaluating light sensitivity, we therefore investigated the time-course effect of different doses of CGRP on arterial blood pressure and potential behavioural changes.

Baseline MAP for the ten rats used was 91.8 ± 13.5 mmHg. Intravenous CGRP resulted in an immediate and significant drop in MAP compared to pre-injection baseline levels for all doses tested (15 µg/kg; p = 0.013, 45 µg/kg; p = 0.013, 100 µg/kg; p = 0.003) (Fig. 3A, B). However, the marked decrease in MAP was short lived, and 5–10 min post injection the blood pressure had increased and was no longer significantly different from baseline. At the end of the observation period (20–25 min after injection) it had reached a relatively stable level at about 75–90% of baseline pressure. While the drastic pressure drop was compensated for rather quickly, aligning with a previous study [29], peripheral vasodilation lasted for at least 60 min, as assessed by observation of cutaneous flushing (red ears and front paws), indicating a continued effect from CGRP. Direct vasoactive effects of i.v. CGRP assessed by live fundus imaging revealed vasodilation of ciliary vessels evident a few minutes after i.v. injection and lasting for at least 15 min (Fig. 4). At the 1 h time-point a clear ciliary vasodilation was no longer visible. Observing rats injected with CGRP while awake did not reveal any signs of adverse side effects such as lethargy, uncoordinated movements, distress, or increased tendance to seek shelter or hide compared to their cage mates. The only clinical feature observed following the injection was redness of ears and paws which lasted for the entire observation period of 60 min, similar to what others have reported in mice [30] and parallelling observations in human provocation studies [31]. These combined findings indicate that the cardiovascular effects of CGRP were quickly compensated, and that the sustained peripheral vasodilation was modest and not associated with clinical adverse effects. Based on these findings, the highest dose tested, 100 µg/kg CGRP, was used for behavioural tests, with a 20 min recovery period applied before testing.

Fig. 3.

Mean arterial pressure following intravenous CGRP in Sprague Dawley rats. Ten male Sprague Dawley rats were injected i.v. with 15 µg/kg (n = 3), 45 µg/kg (n = 3), or 100 µg/kg (n = 4) CGRP and MAP was measured over time. A). Mean MAP over time for the three dose groups. Mean MAP was calculated in bins of 5 s for the entire measurement period starting about 5 min before injection and until about 25 min after the injection. In the graph the timelines are centred around the dosing time-point so that the injection time-points are aligned, in order to better compare the three groups. The 45 µg/kg group had a relatively lower baseline average and is therefore presented on the right y-axis (same scale as the left y-axis) to further improve visual comparisons. B) Delta MAP for the three groups calculated for three different time-periods after the injection: 0–2 min, 5–10 min, and 20–25 min after injection, shown as mean difference relative to baseline set to 0 mmHg. Percentage average reduction in MAP relative to baseline is also shown for each time-period in the graph. Statistical comparisons of mean MAP at baseline and the three different time-points were made for each dose group using the Friedman test with Dunn’s multiple comparisons test. *; p < 0.05, **; p < 0.01. CGRP; calcitonin gene-related peptide, i.v.; intravenous, MAP; mean arterial pressure

Fig. 4.

Fundus imaging following intravenous CGRP administration. Three male Sprague Dawley rats were injected i.v. with 100 µg/kg CGRP under isoflurane anaesthesia. Live fundus imaging was performed, and pictures were taken before injection (Baseline) and every 30 s for the first 10 min, again at 15 min after injection, and finally at 60 min after injection. Ciliary vessel dilation over time was assessed visually. Representative fundus images from one rat at baseline and at 3 min, 15 min, and 60 min after CGRP administration are shown. Arrows point at ciliary vessel; arrowheads point at branch from ciliary vessel. CGRP; calcitonin gene-related peptide, i.v.; intravenous

Bedding material improved light/dark box behaviour in naïve rats

Photophobia, or aversive sensitivity to light, is a subjective experience that cannot be directly assessed in animal models, much like headache pain. Consequently, there is currently no single validated assay that fully captures photophobia per se in rodents. The LDB test, originally developed as an anxiety test designed to capture the conflict between a rat’s desire to explore and its reluctance to enter a bright, open area [32], has therefore often been used to evaluate light aversion as a surrogate for photophobia-like responses [12, 33]. In this study, the LDB test was applied for this purpose. However, the inherent novelty of the test may in itself be anxiogenic [34], and in our previous experience rats typically spent little time in the light compartment, providing a narrow window to detect behavioural changes following interventions. We therefore aimed to modify the test setup to better suit a light-specific focus by reducing the stress and anxiogenic-effect associated with the open and novel environment, thus enhancing the relative impact of light as the aversive factor. A modified LDB setup was therefore developed in an exploratory LDB pilot study, where the effect on time spent in the light compartment was investigated when adjusting two factors; (1) adding bedding to both compartments compared to the plain, no-bedding, standard setup, and (2) different light-intensities. Adding bedding improved the exploratory drive of naïve rats, spending closer to equal amount of time in both compartments, significantly more than rats in the plain unenriched setup (Fig. 5). The time spent in the light compartment was affected by both bedding condition (F(1,36) = 17.02; p < 0.001) and light intensity (F(1,36) = 10.60; p = 0.003), however, there was no significant interaction between the two factors. After discovering that bedding had a significant effect on time spent in the light compartment at the two first light intensities tested, 300 lx (p = 0.024) and 1900 lx (p < 0.001), we tested our highest intensity, 4700 lx, only with bedding, to examine if the rats came out in the light compartment also at this intensity (Fig. 5). Upon confirming this, the highest light intensity was then used for the CGRP intervention studies.

Fig. 5.

Effect of bedding vs. no bedding on time in light in a light/dark box test. Male Sprague Dawley rats were assessed in the light/dark box comparing two different box settings. In one group bedding from the rat’s own home cage was added to both the light and dark compartment, and in the other group no material was added (standard plain setup). The two conditions were compared at two different light settings, 300 lx (n = 4/group) and 1900 lx (n = 16/group). Finally, a new group of rats were tested at a higher lux, 4700 lx, but only with bedding (n = 6). Time spent in the light compartment is shown. Statistical comparison between groups (bedding vs. plain) was made for the first two light intensities using a 2way ANOVA with Sidak’s multiple comparisons test. *; p < 0.05, ***; p < 0.001

Intravenous CGRP resulted in light aversive behaviour in male and female rats

Following LDB optimisation, we applied the test to assess whether i.v. CGRP would elicit light-aversive behaviour compared to saline controls. In the LDB test of Part 1 (mixed-caged housing), statistical analysis showed a significant effect of sex (F(1,40) = 25.39, p < 0.0001), but not compound, on time spent in the light compartment (Fig. 6.A). Post tests clarified that CGRP-injection only produced significant aversion from the lit compartment in males (CGRP-injected; 148 ± 46 s vs. Saline-injected; 257 ± 74 s, p = 0.002), but not females (CGRP-injected; 347 ± 66 s vs. Saline-injected; 292 ± 61 s, p = 0.218). Suspecting a higher degree of emotional contagion between females in the mixed-cages (CGRP and saline animals co-housed) of Part 1, potentially confounding the results, we performed a new LDB test (Part 2) with females housed according to administration group (matched-cages). Now there was a clear compound effect in the female group (CGRP-injected; 167 ± 104 s vs. Saline-injected; 334 ± 18 s, p = 0.003, unpaired t-test), and when performing the collected analysis with males from Part 1 there was a significant main effect of compound (F(1,32) = 22.49; p < 0.0001), but not sex, on time in light. Subsequent multiple comparisons test showed a significantly shorter amount of time spent in the light compartment following CGRP-injection compared to saline in both male (p = 0.006) and female (p = 0.003) rats (Fig. 6.A), suggesting a light-aversive response to CGRP in both sexes.

Fig. 6.

Behaviour in the LDB test following CGRP administration. Male and female Sprague Dawley rats were injected i.v. with either CGRP or saline control and 20 min later the rats were tested in the LDB test to assess light aversive behaviour. Part 1 (mixed-cages; CGRP- and saline injected rats co-housed): n = 24 male rats and n = 20 female rats. Part 2 (matched-cages; housing according to administration group): n = 12 female rats. Within each group half of the rats were treated with CGRP and the other half with saline. (A) Time spent in the light compartment. (B) Resting time divided by compartments. (C) Examples of track plots, showing how rats moved during the test. The red bar represents the opening between the two compartments. Statistical comparisons between the two administration groups for male and female rats were made using 2way ANOVA with Sidak’s multiple comparisons test. Statistics on resting time shown were made on total resting time. **; p < 0.01. ns; non-significant p ≥ 0.05. CGRP; calcitonin gene-related peptide, i.v.; intravenous, LDB; light/dark box

We also analysed resting time in the LDB following administration (Fig. 6.B). Statistical analysis of Part 1 results (mixed-cage housing) showed a significant effect of both compound (F(1,40) = 5.53, p = 0.024), sex (F(1,40) = 9.82, p = 0.003), and their interaction (F(1,40) = 5.84, p = 0.020) on total resting time. Post tests revealed an effect of compound only in the male group, where CGRP significantly increased resting time compared to saline (p = 0.002). Again, the analysis was also made with females of Part 2 (matched-cage housing) instead of the females of Part 1, and this time there was a main effect of compound (F(1,32) = 8.85, p = 0.006), but not sex, on total resting time. Again, subsequent multiple comparisons test showed a significant effect of CGRP in males, with more time spent immobile following CGRP-injection (CGRP-injected; 122 ± 67 s vs. Saline-injected; 57 ± 36 s, p = 0.007), but no significant effect in females (CGRP-injected; 88 ± 49 s vs. Saline-injected; 48 ± 32 s, p = 0.312). The male resting time was analysed further to assess if there was a compartmental preference for where the rats chose to rest. Main effects analysis showed a significant effect of compound (F(1, 22) = 8.76, p = 0.007), compartment (F(1, 22) = 16.91, p = 0.0005), and their interaction (F(1, 22) = 7.43, p = 0.012) on resting time, indicating that CGRP-injection resulted in increased resting time in the dark and reduced resting time in light when compared to saline. Subsequent multiple comparisons revealed that CGRP-injected male rats spent significantly more time resting in the dark than in the light compartment (dark: 111 ± 77 s; light: 11 ± 18 s, p = 0.0002), whereas no significant compartmental effect was observed for saline-injected controls (p = 0.561).

Half of the male rats and all the female rats of Part 1 (mixed-cage housing) were subsequently tested in an OFT, following a wash-out period and reassignment to a new treatment group. The OFT, an anxiety test frequently used alongside the LDB test when assessing light aversion [11, 12, 21], was included to assess possible anxiety-related behaviour not related to light exposure following CGRP and saline administration.

Analysis of time spent in the centre zone showed no significant effect of either compound or sex on zone preference, with both females and males spending about 2–3% of the total test time in the centre zone (Fig. 7.A). Resting time was also assessed (Fig. 7.B), and although statistical analysis showed no significant difference between saline- and CGRP-injected rats, a significant effect of sex on resting time was seen (F(1, 28) = 13.00, p = 0.001), and subsequent analysis revealed a significant difference between males and females among both those receiving CGRP (Male; 196 ± 80 s, Female; 99 ± 65 s, p = 0.021) and saline (Male; 171 ± 132 s, Female; 65 ± 32 s, p = 0.013), with males resting more regardless of administration group. No compartmental analysis of resting time was performed since all immobility occurred in the outer zone.

Fig. 7.

Behaviour in the OFT following CGRP administration. Male (n = 12) and female (n = 20) Sprague Dawley rats were treated i.v. with either CGRP or saline control and 20 min later the rats were tested in the OFT to assess anxiety. Within each group half of the rats were treated with CGRP and the other half with saline. The male rats had previously been tested in the LDB and the EvF (LDB - EvF - OFT), and the female rats had previously been tested in the LDB (LDB - OFT). The rats were housed with both administration groups represented in each cage (Part 1: mixed-cages). (A) Time spent in the centre zone. (B) Total resting time. (C) Total distance travelled. (D) Examples of track plots, showing how rats moved during the test. Statistical comparisons between the two administration groups for male and female rats were made using 2way ANOVA with Sidak’s multiple comparisons test. ns; non-significant p ≥ 0.05. CGRP; calcitonin gene-related peptide, EvF; electronical von Frey, i.v.; intravenous, LDB; light/dark box, OFT; open field test

As there were no clear compound effects on zone preference or resting time, we also analysed distance travelled in the OFT (Fig. 7.C). Main effect analysis showed a significant effect of both compound (F(1, 28) = 4.30, p = 0.047) and sex (F(1, 28) = 42.55, p < 0.0001) on distance travelled, without a significant interaction between the two factors, indicating that females travelled a longer distance than males overall, and that CGRP reduced locomotor activity in both sexes to a similar extent. Multiple comparisons testing did however not reveal any significant compound effect on distance travelled when looking at males and females separately (male; p = 0.597, female; p = 0.071) (Fig. 7.C), however, a significant effect of sex was seen for both administration groups (CGRP: Male; 30 ± 7 m, Female; 50 ± 10 m, p = 0.0003. Saline: Male; 35 ± 12 m, Female; 60 ± 8 m, p < 0.0001), with female rats travelling a longer distance regardless of compound received.

All rats, male and female, of both injection groups, seemed quite stressed and vocalized during the OFT, indicating a high level of stress or anxiety, making it difficult to discriminate between the two groups.

The other half of the male rats (mixed-cage housing) and the female rats of Part 2 (matched-cage housing) were assessed in the DDB test instead of the OFT following the LDB test, a washout period, and compound reassignment. Just like the OFT, the DDB test was used as a complementary test to the LDB to assess the impact on behaviour of bright light alone, by testing in an otherwise identical low-light environment. An OFT was not performed with Part 2 females as the DDB test was considered a better complement to the LDB test, based on the result on male rats of Part 1. The OFT was deemed less sensitive due to the high levels of anxiety-like behaviour observed in this setup during the first part of the study, as seen by very little time spent in the centre of the arena and vocalisation.

Time spent in the dim compartment was compared between CGRP- and saline-injected rats (Fig. 8.A). Main effect analysis showed no significant effect of either compound or sex on time spent in the dim compartment. Male and female rats in both injection groups spent roughly 40% and 50% of the test session in the dim compartment, respectively, confirming that CGRP did not induce anxiety-like behaviour. Resting time was also analysed for the DDB test and here there was a significant effect of both compound (F(1.20) = 5.14; p = 0.035) and sex (F(1,20) = 7.12; p = 0.015) on total resting time, but no significant interaction between them, indicating that resting time increased with CGRP-injection regardless of sex. Subsequent multiple comparisons test did however not show any significant effect of administered compound on resting time when looking at males and females separately (male; p = 0.133, female; p = 0.383), suggesting that the observed main effects reflect modest but consistent differences across the full dataset rather than strong effects within individual subgroups (Fig. 8.B). A possible compartmental preference during resting was also evaluated, and for male rats there was a significant effect of compartment (F(1, 10) = 10.06, p = 0.010), but not compound, with rats resting more in the dark than in the dim compartment, regardless of administration group. Multiple comparisons analysis did however not show any significant difference in resting time between compartments within either administration group, again suggesting a more general effect. For female rats there was no significant effect of either compartment or administered compound on resting time distribution.

Fig. 8.

Behaviour in the DDB test following CGRP administration. Male (n = 12) and female (n = 12) Sprague Dawley rats were treated i.v. with either CGRP or saline control and 20 min later the rats were tested in the DDB test to assess aversive behaviour. Within each group half of the rats were treated with CGRP and the other half with saline. The male rats had previously been tested in the LDB and the EvF (LDB - EvF - DDB), and the female rats had previously been tested in the LDB (LDB - DDB). The male rats were housed with both administration groups represented in each cage (Part 1: mixed-cages), and the females were housed according to compound received (Part 2: matched-cages). (A) Time spent in the dim compartment. (B) Resting time divided by compartment. (C) Total distance travelled. (D) Examples of track plots, showing how rats moved during the test. The red bar represents the opening between the two compartments. Statistical comparisons between the two administration groups for male and female rats were made using 2way ANOVA with Sidak’s multiple comparisons test. Statistics on resting time shown were made on total resting time. *; p < 0.05, ns; non-significant p ≥ 0.05. CGRP; calcitonin gene-related peptide, DDB; dim/dark box, EvF; electronical von Frey, i.v.; intravenous

Finally, distance travelled was evaluated (Fig. 8.C), and main effects analysis showed a significant effect of compound (F(1, 20) = 7.88, p = 0.011), and subsequent post tests confirmed a significantly shorter distance travelled following CGRP-injection in male rats (CGRP; 21 ± 7 m, saline; 31 ± 8 m, p = 0.025), but not in females (CGRP; 29 ± 6, saline; 33 ± 4 m, p = 0.411), when compared to saline controls.

Loose faeces were seen in some rats during the behavioural tests. This was much more frequently observed in the CGRP-treated groups, and the faeces were notably malodorous, suggesting some gastrointestinal effects of CGRP, similar to what others have reported [11, 30, 35].

Intravenous CGRP caused periorbital cutaneous allodynia in male and female rats

To evaluate mechanical allodynia following i.v. CGRP administration, tactile sensitivity was assessed using the von Frey test, which is considered the golden standard for measuring mechanical thresholds in rodents [36]. Cephalic allodynia is a common symptom during migraine attacks, but other body regions can also be involved (extracephalic), therefore we examined both periorbital and plantar sensitivity [16, 37].

Tactile sensitivity was assessed 20 min after CGRP or saline injection, same time-point as for the behavioural tests, and compared to pre-injection results (baseline). In Part 1 (mixed-cage housing), EvF was the second or third test performed, LDB being the first. For the male rats there was no significant effect of either CGRP or saline on periorbital (CGRP; p = 0.409, Saline; p = 0.322, Fig. 9A) and plantar (CGRP; p = 0.375, Saline; p = 0.571, Fig. 9B) sensitivity. For female rats there was an increased periorbital sensitivity, significantly reduced pressure thresholds, in both administration groups compared to baseline (CGRP; p = 0.020, Saline; p = 0.001, Fig. 9C), but no significant effect of compound on plantar von Frey (CGRP; p = 0.767, Saline; p = 0.157, Fig. 9D).

Fig. 9.

Periorbital and plantar pressure thresholds following intravenous CGRP administration. Male and female Sprague Dawley rats were treated with CGRP or saline i.v. and 20 min later a von Frey test was performed, first a periorbital and then a plantar examination. Baseline withdrawal threshold results were measured two days before the intervention study. Part 1: n = 24 male rats, n = 20 female rats, (mixed-cages, with CGRP and saline rats co-housed). Part 3: n = 16 male rats, n = 16 female rats, (matched-cages, with one injection group per cage). Half of the rats received CGRP and the other half saline. For male rats in Part 1 the EvF test was their second test (LDB - EvF) and for females it was their third behavioural test (LDB - OFT - EvF). In Part 3 the rats were naïve to prior injections and testing before EvF assessment. Results for male rats from study Part 1 and 3 are shown in (A) periorbital threshold results and (B) plantar threshold results. Results for female rats from Part 1 and 3 are shown in (C) periorbital threshold results and (D) plantar threshold results. Statistical comparisons between the two time-points (baseline and 20 min post injection) for each compound group were made using paired t-tests. *; p < 0.05, **; p < 0.01, ns; non-significant p ≥ 0.05. CGRP; calcitonin gene-related peptide, EvF; electronical von Frey, i.v.; intravenous, LDB; light/dark box, OFT; open field test

In the third part of the study the rats were all naïve to injections and testing and housed according to compound allocation. This time there was a significantly reduced periorbital threshold (increased cephalic sensitivity) following CGRP-injection in both male (Baseline; 233 ± 15 g vs. 20 min; 178 ± 39 g, p = 0.011) and female (Baseline; 218 ± 40 g vs. 20 min; 154 ± 26 g, p = 0.016) rats (Fig. 9A, C). Interestingly, saline-injected male rats had the opposite effect on periorbital sensitivity, with a significantly increased threshold (Baseline; 211 ± 22 g vs. 20 min; 238 ± 31 g, p = 0.009) at 20 min compared to baseline (Fig. 9A). This was not seen in saline-injected female rats (p = 0.748, Fig. 9C). Plantar sensitivity was again not affected by compound received in male rats (CGRP; p = 0.156, Saline; p = 0.606, Fig. 9B), however, for female rats the sensitivity was reduced, with a significantly increased threshold for both CGRP (Baseline; 230 ± 25 g vs. 20 min; 258 ± 19 g, p = 0.019) and saline-injected (Baseline; 232 ± 23 g vs. 20 min; 271 ± 23 g, p = 0.004) rats compared to their baseline levels (Fig. 9D). To assess whether the CGRP-induced reduction in periorbital pressure thresholds differed across individual pressure sites, the three periorbital measurement points were also analysed separately. This secondary exploratory analysis revealed a comparable reduction across all three sites, with no evidence of site-specific effects (Suppl. Fig. S1).

Discussion

In this study, intravenously administered CGRP induced light-aversive behaviour and periorbital allodynia in both male and female rats. Our data also highlight the importance of optimal study design, as emotional contagion or social analgesia appears to influence the behavioural responses, particularly in female rats, and stress from previous injections and testing seems to affect von Frey outcomes in both males and females. Interestingly, we also observed that environmental bedding appeared to reduce avoidance behaviour in the light/dark box test, refining the setup to shift the emphasis from general anxiety to light-driven avoidance.

Optimizing the light/dark box test

In order to optimize the LDB test into a more light-specific assay, reducing novelty-related anxiety, we added bedding material to the test setup. This turned out to significantly improve the exploratory behaviour, shown in naïve rats spending significantly more time in the light compartment compared to when using the standard unenriched setup. This was observed at both 1900 lx and 300 lx, and although data at 300 lx should be interpreted with caution due to the small sample size, the overall pattern suggests that familiar olfactory cues or an enriched environment during testing may attenuate anxiety and promote exploratory behaviour in an otherwise novel and potentially anxiogenic environment. Our findings align with those of Mackay-Sim and Laing, who showed that rats prefer a compartment previously occupied by another rat over a clean, empty one, likely due to social odours [38]. When the previous occupant had been stressed, however, this preference was reduced, but the compartment was not avoided, as rats showed equal preference for both compartments. This supports our suggestion that familiar scents may increase confidence in the test setup. Thus, refining the LDB test by adding bedding materials to reduce avoidance may help isolate light-evoked behaviour from novelty-induced anxiety. This optimization may thereby improve the test’s sensitivity when assessing photophobia-like responses in future intervention studies. With control animals spending close to equal amount of time in both compartments, light-specific behavioural changes can be more easily detected. Others have pursued similar optimization by pre-exposing the animals to the test chambers [39]. Habituation to the chambers also introduces familiar scents, likely having a similar effect as bedding material. The pre-exposure protocol is however much more time-consuming. Moreover, overexposure to the test chambers might reduce exploratory drive. Nonetheless, regardless of optimization strategy, a standard anxiety test should always complement the light-aversive test, such as the OFT or DDB, to strengthen the data and further isolate the effect of light itself.

CGRP-induced light aversion

Following i.v. administration of CGRP we observed a light-avoidant behaviour in our optimized LDB test, evident in both male and female rats. By complementing the test with an anxiety assessment under low-light conditions (the DDB test or the OFT), which did not show a significant compartment or zone preference related to CGRP administration for either males or females, we were able to substantiate that light was the aversive factor in the LDB test. Although this light-avoidant behaviour is not a direct measure of light-induced discomfort or pain, it can be interpreted as a photophobia-like phenotype. These findings align with results from a clinical provocation study where 72% of migraine patients reported photophobia following i.v. administered CGRP [9]. Our results also support previous preclinical findings demonstrating similar light-induced effects from CGRP injection, where both central (intracranially) and peripheral (intraperitoneally [i.p.]) delivery of CGRP resulted in light-avoidant behaviour [11, 12, 39]. Consistent with our observations, Wattiez et al. likewise reported no sex difference in the CGRP-induced response in the LDB test in mice [13]. However, in these studies a light intensity of 25,000–27,000 lx was needed to induce light aversive behaviour in wildtype mice, both when CGRP was delivered peripherally and centrally, using a dose similar to what was used in our study (100 µg/kg). This is in sharp contrast to the 4700 lx used in the present study, suggesting that both the animal species used and the route of CGRP administration (i.v., i.p., or intracranial) influence the degree of light sensitivity and, consequently, the light intensity required to elicit light-aversive behaviour. Additionally, the sensitivity of the behavioural test used to assess this aversion likely affects both the outcome and the light intensity needed to reveal CGRP’s effect. Importantly, bright environments can be inherently aversive, even in untreated animals. Thus, with a more intense light stimulus, a low CGRP concentration may be enough to provoke an aversive response, whereas more CGRP is needed in low light conditions. For example, in RAMP1-overexpressing transgenic mice, which are hypersensitive to CGRP, relatively low light intensities (55–1000 lx) are sufficient to induce light aversion after intracranial CGRP administration [11, 12, 33]. In contrast, wild-type mice receiving the same CGRP dose require a much higher intensity (27000 lx) to exhibit a comparable behavioural response [39]. It is therefore desirable to design test paradigms that can detect CGRP-induced light aversion even under less intense light levels, to better isolate and characterize the pharmacological effect.

Interestingly, female CGRP-treated rats in Part 1 (mixed-cage housing) behaved differently from males, showing no light avoidance in the LDB test. However, when housing according to administration group in Part 2 (matched-cage housing), naive female rats displayed a behaviour more comparable to the males in the LDB test, supporting a CGRP-specific effect. This may reflect different emotional and social environments in cages with mixed- and matched-treatment housing, and possibly a display of social analgesia or buffering in Part 1 females, where the pain or fear response is lowered by the presence of unaffected cage mates [24, 40]. Depending on the study setup, rodents in pain or discomfort can influence, and are influenced by, others in various ways [22–25]. In general, female rodents appear to be more sensitive to the behaviour and social cues of conspecifics and tend to modulate their own behaviour to a greater extent than males [23, 24, 40]. Although the causative mechanisms behind the differences observed in the present study cannot be identified without further measurements, our results highlight the importance of considering housing and study design when evaluating behavioural effects.

Mechanistic considerations for CGRP-induced light aversion

The background for how CGRP induces light-aversion is not clear. Both general discomfort and pain can result in less time spent in the light compartment. There is a difference between the pathways involved in light causing general discomfort and light exacerbating headache pain, and light causing direct pain in the eye [41]. We would expect peripherally administered (i.v.) CGRP to act outside the CNS, likely within the trigeminovascular system, where it may sensitize trigeminal pathways and enhance their responsiveness to converging stimulatory inputs. One group found light sensitive neurons containing CGRP and melanopsin in the mouse and human trigeminal ganglion [42]. They further demonstrated that following trigeminal stimulation by nitroglycerin a light aversive behaviour was induced (1000 lx), both in intact animals and following optic nerve crush destroying the optic nerve but leaving ciliary and trigeminal nerves intact. Light stimuli may thus indirectly activate trigeminal nociceptors via signalling pathways from the eyes, and if the trigeminal neurons are sensitized or altered by CGRP, then the signal to second-order neurons might be more intense, or the threshold may be lower, resulting in pain signalling [41]. It is possible that in our study, CGRP sensitizes light sensitive trigeminal neurons, which then respond to light input differently from controls, resulting in a light aversive behaviour. Another possibility is that exogenous CGRP may stimulate the trigeminovascular system and, through subsequent signalling via second-order neurons, sensitize thalamic circuits. This speculative mechanism could modulate how thalamocortical neurons respond to input from light-sensitive retinal neurons, thereby influencing third order signalling to cortical areas involved in pain perception [43]. These are similar to the mechanisms believed to occur during exacerbation of headache by light, where light changes pain perception in an already sensitized brain. Thalamic sensitization could potentially also result in activation of other pathways not related to pain perception in response to light, but involved in other responses related to the discomfort reported by some patients in response to light, such as anxiety and dizziness [44].

The present behavioural data do not allow differentiation between light-induced pain, light-induced discomfort, or light-exacerbated cephalic hypersensitivity. All of these processes may contribute to photophobia in migraine and could plausibly underlie the observed light-aversive behaviour in our model. Future studies incorporating neural or physiological correlates of light processing may further strengthen the interpretation of photophobia-like behaviours observed in this model.

CGRP-induced discomfort

We did not only look for a light avoidant behaviour, but also at time spent immobile (resting time) and distance travelled. These factors can be indicators of a more general state of pain or discomfort, especially when seen in the low light tests.

Across the behavioural tests, CGRP administration primarily affected locomotor activity in male rats, with significantly increased resting time in the LDB and reduced distance travelled in the DDB. The increased immobility in the LDB test could be an effect of light-aversion, and compartment preference analysis showed that the male rats preferred resting in the dark compartment when injected with CGRP, whereas no significant compartment preference for resting was seen following saline administration. This reduced locomotor activity among males in response to CGRP in both the LDB and DDB test might reflect a deliberate reduction in physical activity to avoid aggravating headache symptoms, similar to behaviour observed in migraine patients [45]. Light-independent behavioural effects of CGRP have also been demonstrated in male and female mice, where peripherally (i.p.) administered CGRP caused a pain response (assessed using a facial grimace assay) in the absence of light (0 lx), as well as in bright light [30]. The gastrointestinal effects observed following systemic CGRP administration could potentially contribute to altered locomotor activity; however, these effects were not quantified and therefore cannot be directly linked to the behavioural outcomes. In addition, while blood pressure recovered rapidly, more persistent systemic vascular effects of CGRP, as indicated by visibly lasting vasodilation, may also influence behaviour by contributing to general discomfort. Nevertheless, the absence of a consistent, global suppression of locomotor activity suggests that such systemic effects are unlikely to represent the primary driver of the observed behavioural responses.

In the OFT, the most prominent differences were not compound-related but occurred between males and females. Statistical analyses confirmed that female rats spent significantly less time resting and travelled a greater distance than males, reflecting common observations in rats [46–48]. The OFT also appeared to induce high levels of anxiety-like behaviour in our experiment, and the observed sex-related differences in locomotor activity may reflect distinct behavioural responses to pain and stress, with female rats tending toward escape-like behaviour and male rats more often displaying freezing responses [49]. Both female and male rats spent only about 2–3% of their time in the centre zone, regardless of compound received, providing a very narrow window to detect any worsening in anxiety-like behaviour. In contrast, the DDB appeared to provide a wider window of detecting changes in anxiety-like behaviour, as saline-injected animals spent close to half of the test time in the dim compartment, and no apparent anxiety-related behaviour was observed during, or at the end of, the test in this setup. Any anxiogenic behaviour potentially induced by CGRP in the OFT, e.g. due to pain or discomfort, may therefore have been masked by the overall high level of anxiety observed even in the control-animals. Consequently, the lack of significant differences between CGRP-injected and control rats in this test alone should be interpreted with caution. But with the support of the results from the DDB test, it appears unlikely that i.v. CGRP induces acute anxiety, supporting that the aversion from the light-compartment in the LDB test is a true reflection of light-aversion rather than a general anxiety-related response.

Cephalic hypersensitivity

Cutaneous hypersensitivity, a reliable marker for activation and sensitization of the trigeminal system [50, 51], was also evaluated following CGRP delivery. When EvF-testing was performed in rats naïve to prior injections and testing and housed according to injection group (Part 3), CGRP administration reduced periorbital pressure thresholds in both male and female rats, indicating increased cephalic tactile sensitivity, i.e. cephalic allodynia. Together with the observed light-aversive behaviour, these findings support a CGRP-driven sensitization of trigeminal nociceptive pathways, rendering the animals hypersensitive to normally non-noxious mechanical and light stimuli.

Cephalic allodynia was included as an indicator of headache-like pain in the rat. The von Frey assessment at 20 min post-dose falls within the early window (0–60 min) in which both people with migraine and healthy controls can develop headache following i.v. CGRP infusion in clinical studies [10]. However, migraine-like attacks generally occur later [10, 31]. The higher CGRP dose used in the current study (100 µg/kg, corresponding to a human equivalent dose of approximately 16 µg/kg [≈ 1 mg in a 70 kg patient] [52]) compared with clinical provocation trials (30–40 µg/patient) [10, 31] may contribute to the early onset of cephalic allodynia observed here.

In contrast, when this assessment was performed as the second or third behavioural test (Part 1, mixed-cage housing), female rats showed increased periorbital sensitivity irrespective of the compound they received, while no significant effects of either saline or CGRP were observed in males. The absence of a CGRP effect in the mixed-cage, repeatedly tested Part 1 cohort may reflect repeated handling or test-order effects, with earlier behavioural tests potentially elevating overall stress levels in these rats. Such increased stress may have masked CGRP-specific effects, resulting in a generally heightened sensitivity in females and reduced responsiveness in males, consistent with previously reported sex-related differences in stress responses [49]. In contrast to the LDB test, where we suspect a form of social analgesia among the females in the mixed-cages setup, were CGRP-injected rats are “comforted” by unaffected cage mates, the results here suggest that prior exposure to injection and testing may have influenced the controls’ capacity for social buffering. This supports the notion that prior handling, testing, and mixed housing may have confounded the results. Both emotional contagion and stress are known to influence pain sensitivity and may affect study outcome [22, 23, 53, 54]. Regardless of causative mechanism behind the different effects seen, housing rats according to administration group and testing them while naïve, as done in Part 3, appears to have reduced these confounding factors.

No CGRP-specific effect on plantar sensitivity was observed in either sex, similar to patterns observed in mouse studies following systemic (subcutaneous) CGRP administration [55, 56]. Together, these results align with the predominantly cephalic allodynia reported in migraine patients and support a headache-like behavioural response to i.v. CGRP in rats.

Conclusion

This is the first rodent study to use the intravenous route for CGRP provocation and to evaluate light-aversive behaviour alongside cephalic sensitivity within the same experimental framework, aligning more closely with clinical protocols. Combined, our results show that intravenously administered CGRP to male and female rats induces a photophobia-like light-aversive behaviour, signs of general discomfort, and cephalic hypersensitivity, responses consistent with a migraine-like state. These findings suggest activation of similar CGRP-sensitive trigeminal pathways in rats and humans. Furthermore, our optimized light-aversion test effectively detected light aversive behaviours and may serve as a valuable tool for future mechanistic studies of migraine-related photophobia-like responses and other disorders involving light aversion. Finally, our results highlight the importance of an optimized experimental design to reduce confounding stressors and improve the interpretability of behavioural outcomes. In particular, avoiding co-housing of different treatment groups and carefully considering test order effects are essential for robust behavioural study design.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

ANOVA

Analysis of variance

BP

Blood Pressure

CGRP

Calcitonin Gene-Related Peptide

CNS

Central Nervous System

DDB

Dim/Dark Box

EvF

Electronic von Frey

i.p.

Intraperitoneal

i.v.

Intravenous

LDB

Light/Dark Box

MAP

Mean Arterial Pressure

NaCl

Sodium Chloride

OFT

Open Field Test

PBS

Phosphate-Buffered Saline

RAMP1

Receptor activity modifying protein 1

SD

Sprague Dawley

Author contributions

VKA, BSH, KMLN, and EF performed arterial pressure measurements. VKA and BSH performed LDB, DDB, and OFT measurements. VKA, BSH, and KAH performed von Frey measurements. VKA, BSH, KMLN, SH, and KAH analysed and interpreted the data. VKA and BSH prepared the first draft, all authors edited and reviewed the manuscript and approved the final version.

Funding

Open access funding provided by Copenhagen University. K.A. Haanes was supported by a Lundbeck foundation Fellowship (R345-2020-1977). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Data availability

The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

This study complied with the European Communities Council Directive (86/609/ECC) and received approval from the Danish Animal Experimentation Council (Licenses: 2020-15-0201-00708 and 2020-15-0201-00751).

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.