Sage Journals: Discover world-class research

Abstract

Background

Opening of K_ATP channels by systemic levcromakalim treatment triggers attacks in migraine patients and hypersensitivity to von Frey stimulation in a mouse model. Blocking of these channels is effective in several preclinical migraine models. It is unknown in what tissue and cell type K_ATP-induced migraine attacks are initiated and which K_ATP channel subtype is targeted.

Methods

In mouse models, we administered levcromakalim intracerebroventricularly, intraperitoneally and intraplantarily and compared the nociceptive responses by von Frey and hotplate tests. Mice with a conditional loss-of-function mutation in the smooth muscle K_ATP channel subunit Kir6.1 were given levcromakalim and GTN and examined with von Frey filaments. Arteries were tested for their ability to dilate ex vivo. mRNA expression, western blotting and immunohistochemical stainings were made to identify relevant target tissue for migraine induced by K_ATP channel opening.

Results

Systemic administration of levcromakalim induced hypersensitivity but central and local administration provided antinociception respectively no effect. The Kir6.1 smooth muscle knockout mouse was protected from both GTN and levcromakalim induced hypersensitivity, and their arteries had impaired dilatory response to the latter. mRNA and protein expression studies showed that trigeminal ganglia did not have significant K_ATP channel expression of any subtype, whereas brain arteries and dura mater primarily expressed the Kir6.1 + SUR2B subtype.

Conclusion

Hypersensitivity provoked by GTN and levcromakalim in mice is dependent on functional smooth muscle K_ATP channels of extracerebral origin. These results suggest a vascular contribution to hypersensitivity induced by migraine triggers.

Keywords

Arterial dilation blood-brain barrier Kir6.1 migraine nociception

Introduction

The mechanisms of migraine are incompletely understood. A recent human experimental study showed that intravenous infusion of levcromakalim provoked migraine attacks in all subjects tested (1) and similarly it induced tactile hypersensitivity in a mouse model (2,3). Inhibition of ATP-sensitive potassium (K_ATP) channels rescued hypersensitivity in both the glyceryl trinitrate (GTN) mouse model of migraine and in a congenital rat model of migraine (2). This suggests that K_ATP channels could be an interesting novel target for migraine drugs. K_ATP channels are, however, ubiquitously present and blocking all K_ATP channels would probably have unwanted side effects. We have previously suggested the Kir6.1-SUR2B channel subtype as a possible drug target (2).

In the present study, we focused on the effect of K_ATP channel opening in the brain versus outside the brain and identification of relevant target tissue. Importantly, both anti- and pronociceptive transmission has been reported depending on the site of action (2,3–6). We used three approaches all including in vivo mouse models: (i) systemic and central activation of K_ATP channels; (ii) provocation of hypersensitivity by clinically validated migraine-inducing substances GTN and levcromakalim; and (iii) conditional knockout of K_ATP channel subunit Kir6.1 in smooth muscle. Our hypothesis was that activation of K_ATP channels outside the blood-brain barrier induce hypersensitivity, while activating K_ATP channels in the brain would have the opposite effect. We show that K_ATP channels in vascular smooth muscle outside the brain are responsible for the sensitisation by systemic administration of GTN and levcromakalim.

Materials and methods

Experimental animals

Experiments were performed under license number 2017-15-0201-01358 from the Danish Animal Experiments Inspectorate and in accordance with the European Community guide for the care and use of animals (2010/63/UE). A total of 125 mice were used to complete the study. Subjects were adult male and female mice 8-23 weeks of age on C57Bl/6J background: Wild-type mice (N = 88) were C57Bl/6JBomTac (Taconic, Ejby, Denmark), intracerebroventricular- (ICV) cannulated mice (N = 20) were C57Bl/6J (Charles River, France) and the smooth muscle Kir6.1 conditioned knockouts (N = 17) B6-Kcnj8^tm1(Cre)/Geno mice (7) were donated from Professor Andrew Tinker (Barts and the London School of Medicine and Dentistry, England). Knockouts were termed ‘F/F cre’ and the floxed wild types not expressing the smooth muscle promotor termed ‘F/F’. Mice were group-housed in a climate-controlled room under a 12 h light/dark cycle, lights on at 7 AM. Food and water were freely available. ICV cannulated and Kir6.1 knockout males were single-housed to protect the ICV probe and due to aggressive behaviour, respectively.

Mouse models of migraine

Repeated systemic dosing of levcromakalim or GTN induces a progressive hypersensitivity to tactile stimulation both within trigeminal and somatic pain circuits. The GTN model is highly validated in terms of responses to anti-migraine drugs (8 –11), whereas the levcromakalim model is novel and not yet as validated as the former (2,3). Importantly, the models are directly mirrored in the migraine clinic where both GTN and levcromakalim are known triggers of migraine or headache and do not cause any other pains (1,12). Levcromakalim (Tocris, Bio-Techne, Abingdon, UK) 1 mg/kg or GTN (Cambrex via Hospital pharmacy) 10 mg/kg, or their respective vehicles 2% DMSO and 12.2% ethanol in saline, were injected IP (10 ml/kg volume) every other day for a total of 2 or 6 times depending on the experiment. With the applied doses and statistical power, GTN normally induces tactile sensitivity already after a single injection, whereas levcromakalim often needs to be injected twice before a significant hypersensitivity is present. Both models also develop basal hypersensitivity which is measured every test day prior to injections (2,3,8,10).

Intracerebroventricular administration of levcromakalim

Mice were delivered surgically prepared with ICV cannulas (C315GS-4/Spc, 2.3 mm length, Plastics One, Bilaney, Düsseldorf, Germany) placed at the following stereotactic coordinates from bregma: anterior-posterior: −0.22 mm, lateral: 1 mm and dorso-ventral: −2.3 mm. Location of the cannula was confirmed by post-mortem injection of Evans blue (Figure 1(g)). Under isoflurane anaesthesia, levcromakalim 10 µg/mouse or vehicle (5% DMSO in PBS) was infused over five minutes in 5 µl volume. The vehicle was different from the systemic vehicle to avoid precipitation in the small volume. The injector was kept in place for another two minutes to avoid back flow. ICV infusions were repeated twice separated by one day. In total three out of twenty mice were excluded from this experiment. One vehicle- and one levcromakalim-treated mouse were completely excluded due to occlusion of guide cannula and unsatisfactory post-mortem staining. One in the levcromakalim group was excluded on day 3 due to significant backflow upon infusion. The dose was selected based on available literature showing analgesic effect of 3–100 µg/mouse (6,13,14), and equals a dose of 0.4 mg/kg for a 25 g mouse, which were considered a large dose given directly in the brain ventricle.

Figure 1.

Systemic, but not central and local administration of levcromakalim induces cutaneous hypersensitivity. (a)-(f) Hotplate test (day 1 and 3 at 20 min and 135 min) and von Frey test (day 1 and 3 at 120 min) after levcromakalim (LEV) or vehicle (VEH) administration either systemically (IP; 1 mg/kg) or centrally in the lateral ventricle (ICV; 10 µg/mouse). Hotplate data are expressed asContinued.

Intraplantar administration of levcromakalim

Levcromakalim 2.5 µg/mouse or vehicle (2% DMSO in saline) was injected subcutaneously between the foot pads of the right hind paw in 20 µl volume using a 0.3 ml 31 G insulin syringe. One person held the mouse by the scruff of the neck while another person performed the injection by restraining the foot of the mouse by one hand and injecting with the other. Tactile sensitivity was measured on both right and left hind paws of the individual mouse in order to control for sensitisation caused by absorption of levcromakalim to the systemic circulation. Hypertonic saline was injected as positive control. The intraplantar injection was given twice with one day in between. The dose was selected based on the finding that 0.1 mg/kg IP did not induce hypersensitivity until the fourth repeated injection (day 7 of the mouse model protocol) (2). Therefore, this dose injected IPL should not sensitise the mice if distributed to the systemic circulation. Contrarily, 1:10 of the active systemic dose of 1 mg/kg IP is a very high dose applied locally.

Cutaneous sensitivity to tactile stimulation

As a surrogate measure of pain/pronociception (10,15,16), tactile sensitivity to stimulation with von Frey monofilaments (Ugo Basile, Gemonio VA, Italy) within periorbital or hind paw nociceptive circuits was measured using the up-down paradigm (17,18) as previously described in detail (10). Hind paw testing was performed with the mouse placed in clear plexiglas chambers on a mesh floor. Mice were placed in the chambers for acclimatisation 30-45 minutes prior to testing (10). For periorbital testing, the mouse was placed in an espresso-sized paper cup (19) secured to a table edge with the open end of the cup 2 cm over the edge. Here, the mice would sit facing out of the cup allowing filament access to the periorbital area. Mice were habituated to the cups for 1 hr the day before testing and 1 hr just before testing. Generally, the sensitivity of the cephalic area is higher than the hind paw and the effect window smaller (10,20). Calculation of 50% withdrawal thresholds was done using the free online calculator at https://bioapps.shinyapps.io/von_frey_app/ with application of inter-filament steps (21). Tactile sensitivity was measured at baseline and 20 min and/or 120 min after administration of levcromakalim and 120 min after GTN by a blinded experimenter.

Cutaneous sensitivity to heat stimulation

The von Frey test of tactile sensitivity is not always capable of detecting an antinociceptive/analgesic effect of substances due to passive lifting of the mouse by the heavier filaments (22). Therefore, we applied the hot plate test for the detection of possible antinociceptive effects of levcromakalim. Mice were baseline tested and at 20 and 135 min after levcromakalim administration on a 52°C hot plate (IITC LifeScience Inc, Woodland Hills, California) and latency to robust withdrawal or licking of hind paws was recorded. Testing was performed by an experimenter blinded to treatment.

Ex vivo vasoactivity

Kir6.1 smooth muscle conditional knockout mice and floxed controls were anaesthetized with gas (70% CO₂ and 30% O₂) and decapitated following 14-15 days of washout since day 11 of the GTN in vivo experiment in which all received GTN. The brain was carefully released from the skull and immersed in cold oxygenated Na⁺Krebs buffer (in mM: 119 NaCl, 4.6 KCl, 1.5 CaCl₂, 1.2 MgCl₂, 1.2 NaH₂PO₄, 15 NaHCO₃, 5.5 glucose and 0.03 EDTA; pH 7.4). The basilar artery (BA) was isolated and divided into two segments (1–2 mm). BA segments were mounted on 25 µm diameter wires in a Mulvany-Halpern wire myograph (Danish Myo Technology, Hinnerup, Denmark) with Na⁺Krebs buffer (37°C). Following 15 min of equilibration, the vessels were stretched to reach a pretension of 1–2 mN/mm. Initially, K⁺Krebs (similar to Na⁺Krebs, except 55.4 mM NaCl is exchanged for KCl, making it 60 mM K⁺) was added to examine the contractile potential of vessels. Vasorelaxation was induced by cumulative exposure to levcromakalim (0.01–30 µM) after U46619-induced (0.3 µM) precontraction. As a positive control, each vessel was exposed to 100 µM papaverine (Sigma-Aldrich, St. Louis, Missouri, US) at the end to induce maximal vasorelaxation. Myograph responses were collected in LabChart™ (ADInstruments, Colorado Springs, Colorado, US). Accumulated vasodilation, % was calculated as $100 - (100 * \frac{x - \max dilation}{100 % precontraction - \max dilation})$ where x was the tension of the blood vessel after stimulation. 100% precontraction and max dilation were constants for individual blood vessels.

Desorption Electrospray Ionization - Mass Spectrometry Imaging (DESI-MSI)

Wild-type mice were given glibenclamide 1 mg/kg IP and euthanized 120 min later by cervical dislocation and the body frozen at −80 °C. The snout, up to the eyes of the frozen mouse head was removed using a hacksaw, and the head was mounted on a cryo-microtome specimen plate using a 5% carboxymethyl cellulose gel as adhesive. Coronal cryo-sections of the head (including skull, brain, jaws etc.) were cut on a Leica CM3050S cryo-microtome (Leica Microsystems, Wetzlar, Germany) at a thickness of 30 µm, thaw-mounted on regular glass slides (VWR, Radnor, Pennsylvania, US) and stored at −80°C until analysis. Prior to DESI-MS analysis, the sample slides were thawed in a vacuum desiccator for 10 min.

The samples were imaged by DESI-MSI on a Thermo LTQ XL ion trap mass spectrometer (Thermo Scientific, Waltham, Massachusetts, US) equipped with a custom built DESI-MSI ion source, as previously described (23). The spray solvent consisted of methanol and water (95:5) delivered at a flow rate of 5 µL/min and a nebulizer gas pressure of 8 bar. Imaging was performed with 250 µm pixel size in negative ion mode using the displaced dual mode imaging method (24) with targeted drug imaging in MS/MS mode and lipid analysis in full scan mode (scan range m/z 450–900). Glibenclamide was detected based on the m/z 492→367 transition. After successful imaging, the imaged sections were H&E stained as described in detail elsewhere (25), and optical images were made on an Olympus BH-2 microscope (Olympus Optical Co., Hamburg, Germany). Raw data from the mass spectrometer was converted to imzML files (26), and images were generated in MSI Reader 1.01 (27). Co-registration of the MS images and the optical images was based on full-scan images of lipids, and the images were overlaid using Corel PaintShopPro 2018 (Corel Corporation, Ottawa, Canada).

mRNA expression studies

Wild-type mice were anaesthetized with pentobarbital and transcardially perfused with 25-30 ml ice-cold PBS buffer. The brain stem (representing TNC), TG and heart were taken out, while the cerebral blood vessels and dura mater were carefully dissected under a microscope. Samples were placed in RNase-free Eppendorf tubes and immediately stored at −80 $° C$ . Total RNA was purified using the NucleoSpin miRNA for small and large RNA (Macherey-Nagel, Dueren, Germany). 300 µl buffer ML and beads were added to the tissue and it was disrupted and homogenized by using the FastPrep-24 5G Instrument (MP Biomedicals, Irvine, California, US) at 6m/s 3 times 20 s with a 20 s break. Total RNA was purified according to manufacturer’s recommendations. After the purification step, yield and purity of the RNA was assessed by spectrophotometrically measuring the 260 nm/280 nm ratio. Cerebral blood vessels and dura samples were pooled from three mice to get more RNA. Additional twelve individual cerebral blood vessel samples were included from our in-house tissue library.cDNA was synthesized from 500 ng purified RNA using the iScript cDNA synthesis kit (BioRad, Hercules, California, US) according to manufacturer’s recommendations. Each mixture was prepared on ice. A negative control was created by not adding reverse transcriptase to the RNA samples in order to control for environmental and genomic DNA contribution. The samples were incubated in a thermal cycler using the following protocol: 5 min at 25°C, 20 min at 46°C, 5 min at 85°C, hold at 4°C and then the cDNA samples were stored at −20°C until further use.

Real-time PCR was performed using the QuantStudio™ Flex System instrument (ThermoFisher Scientific, Waltham, Massachusetts, US) and SYBR Green I dye chemistry (SensiFAST™ SYBR® NO-ROX kit [Bioline]). Each primer pair (Supplementary Table 1) was verified in the heart as a positive control for the different K_ATP channel subtypes using a 1:10 dilution series of the cDNA. For the final experiments, the cDNA was diluted 1:5 and two endogenous reference genes (Hprt1 and Aip) were used as controls. The RT-negative control and a water blank (no cDNA added) were analysed in all experiments to validate the results. The run was as follows: 2 min at 95°C to activate the polymerase followed by the 2-step cycling PCR Stage with a denaturation step for 15 s at 95°C and an annealing/extension step for 30 s at 60°C. Lastly, the run had a melt curve stage with 15 s at 95°C, 1 min at 60°C and then 15 s at 95°C in order to study the specificity of the amplified products. The amplicons were verified by gel electrophoresis. The data was quantified by using the C_t method by the following equation: $2^{- C_{t} (Target gene) - C_{t} (Reference gene)}$ .

Bioinformatics, RNA-sequencing in human, mouse and rat

RNA-sequencing data of the trigeminal ganglion (TG) was investigated in human, mouse and rat in order to support our protein and RNA expression data of the K_ATP subunits. Human TG samples (n=16) were downloaded from the Sequence Read Archive (SRA) database (PRJNA384203) (28). Mouse samples (n=15) were downloaded from three projects within the SRA database (PRJNA396069 (29), PRJNA390912 (30), PRJNA205891 (31)). Rat samples (n=12) were present in-house and previously published (32). Quality of samples was assessed with FastQC (33), leading to the removal of one mouse sample. Quality of the reads was assessed with Next Generation Sequencing Quality Control Generator (NGSQC) (34) and low-quality reads were filtered out using default settings. Quantification was performed with kallisto (35), using the reference genomes GRCh38.p12 (human), Rnor6.0 (rat) and GRCm38.p6 (mouse). Human-rat and human-mouse orthologs were retrieved using the R-package biomaRt (36); in case of multiple ortholog matches, the mean gene count of the orthologs was used. Blood was removed from the rat prior to sequencing. As this was not done for human and mouse samples genes encoding hemoglobin subunits were removed from the human and mouse data. Gene counts were imported using the R-package tximport (37) and normalized into transcripts per million.

Protein expression studies

Wild-type mice were anaesthetized with gas (70% CO₂ and 30% O₂) and decapitated. Trigeminal ganglion (TG), brain stem (representing trigeminal nucleus caudalis [TNC]), dura mater and cerebral blood vessels were removed quickly and stored at -80°C. Dura and cerebral blood vessel tissues were pooled from three mice. The frozen tissues were first grinded on dry ice using a mortar and a pestle and further homogenized by sonicating the samples in RIPA lysis buffer (Hospital Pharmacy) containing a cocktail of protease inhibitors (PhosStop tablet (Roche, Basel, Switzerland) and complete mini inhibitor tablet (Roche). The samples were incubated on ice for 30 min and centrifuged at 14,000 g for 20 min at 4°C, the supernatants were collected and stored at −80°C until further processing. Protein concentrations were determined using the DC protein assay kit II (BioRad).

Protein samples were diluted to 1 mg/ml in 5× loading buffer (Life Technologies, Carlsbad, California, US) heated at 70°C for 10 min and loaded on a 4−12% NuPAGE™ Gel (Invitrogen). The gel was run at 180 V for 30 min in running buffer (20 mL 20× 3-(N-morpholino) propanesulfonic acid (MOPS) and 380 ml milli-Q water). After size separation, the proteins were transferred to PVDF membrane using the iBlot 2 dry blotting system (Life Technologies). The membranes were then blocked in 5% non-fat dry milk (Merck, Søborg, Denmark) in tris-buffered saline (TBS) with 1% Tween-20 (TBS-T) at room temperature for 1 h while being tilted back and forth. The membrane was incubated with primary rabbit polyclonal anti-Kir6.1 (Alamone labs, Jerusalem, Israel APC-105) diluted 1:250 in TBS-T containing 5% non-fat dry milk at 4°C overnight. Next day, the membranes were washed for 3 × x5 min in TBS-T and probed with goat anti-rabbit (DAKO, Tilst, Denmark P044701-2) 1:2000 at room temperature for 1 h while being tilted. The membranes were washed again as described above. Lastly, membranes were developed with Amersham™ ECL™ Prime kit (GE Healthcare) and visualized with ImageQuant LAS 4000 mini (GE Healthcare, Chicago, Illinois, US). In order to re-assess the protein amount loaded in each lane, the membranes were stripped for the bound antibodies using Restore™ PULS Western Blot Stripping Buffer (ThermoFisher) and the membranes were re-probed with rabbit monoclonal anti-GAPDH (Cell signalling 5174S) 1:1000 as described above. To further validate the products observed using the anti-Kir6.1 antibody, the antibody was preincubated with a negative control antigen (Alomone labs) (1 µg:1 µg) in 5% non-fat dry milk in TBS-T. In parallel, in a second tube anti-Kir6.1 antibody was diluted in 5% non-fat dry milk in TBS-T as described above. Both tubes were incubated for 1 h at room temperature. The membrane was cut in half and then added to the tube with or without the negative control antigen.

Immunofluorescence

Wild-type, smooth muscle Kir6.1 F/F, and Kir6.1 F/F cre mice were deeply anaesthetized with sodium pentobarbital IP and transcardially perfused with PBS and subsequently 25 ml 4% PFA (Hospital Pharmacy). Anterior-, middle- and posterior cerebral arteries were dissected in situ within brain parenchyma and stored in PBS. Serial sections 3–5 μm thick were cut in transverse planes and placed on silanized slides and stored at 4°C. For immunofluorescence microscopy, sections were deparaffinized and rehydrated in xylene followed by a series of graded alcohol in accordance with established procedures, treated with TEG buffer before blocking with Dako REAL™ Antibody Diluent (Agilent, Glostrup, Denmark) and left in primary antibodies (rabbit polyclonal anti-Kir6.1 Alamone labs APC-105 1:200 and mouse monoclonal anti-alpha smooth muscle actin Sigma-Aldrich A2547 1:500) overnight at 4 °C in a humidified chamber. After PBS washing steps, cells were incubated in corresponding Alexa Flour® secondary antibodies 1:600 (Invitrogen, Waltham, Massachusetts, US A-10037 and A-2106) for 45 min at room temperature. Cell nuclei were labelled by DAPI staining. Coverslips were mounted with either Dako fluorescent mounting medium or with PBS containing 90% glycerol, and 2% N-propyl-gallate and subjected to epifluorescence microscopy. Images were captured on a fully motorized Olympus BX63 upright microscope with an Olympus DP72 color, 12.8-megapixel, 4.140 × 3.096-resolution camera and with a fully motorized and automated Olympus IX83 Inverted microscope with a Hamamatsu ORCA-Flash 4.0 camera (C11440-22CU). The software used was Olympus CellSens dimension, which was able to do deconvolution on captured z stacks, and images were processed for publication using Image J and Adobe Photoshop CS6.

Statistical analyses

Group sizes for in vivo experiments were decided without the use of a priori sample size calculations but based on our in-house experience with the mouse models. 10-12 mice per group have previously provided adequate power to detect intermediate effects of antagonists at high levels of significance in experimental designs with four test groups and five multiple comparisons between groups (2,3,10,38). Here, group sizes were 8-10, as fewer comparisons were generally made. Actual group sizes are given in figure legends. Randomisation of animals to experimental groups was achieved by draw but counterbalanced with respect to baseline measurements and home cage in order to avoid confounding by these two factors. von Frey data are presented and analysed as mean ± SEM of square root transformed (SQRT) 50% withdrawal thresholds expressed in grams. This data transformation improves the Gaussian distribution to allow parametric testing. All behavioural data were analysed with a repeated-measure mixed model with Sidak’s post hoc comparison between groups. In the case of cephalic von Frey test, an unpaired t-test was applied as only one timepoint and two groups were tested. Vasoactivity measures were analysed by repeated-measure mixed model and Sidak’s post hoc comparison between groups. All data were analysed using GraphPad Prism 8 (Graph Pad Software Inc., CA, USA). In figure legends, the overall treatment effects are indicated with two-tailed p-values and F-statistics are reported with P < 0.05 considered statistically significant. In the figures, level of significance is indicated for Sidak’s post hoc comparisons as follows: *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Results

Systemic but not central or local K_ATP channel activation induces cutaneous hypersensitivity

We compared systemic (intraperitoneal, IP) and central (intracerebroventricular, ICV) administration of K_ATP channel agonist levcromakalim. There was no effect on the hot plate 20 min after systemic administration (Figure 1(a)). Repeated systemic administration induced tactile hypersensitivity 120 min after injection as measured by von Frey filaments. The 50% withdrawal thresholds were significantly lower 120 min after the second injection of levcromakalim, P = 0.0003 (Figure 1(c)). Similarly, hypersensitivity to heat also developed after the second injection on day 3 with latency to withdrawal decreasing by 4.9 s ± 2.1 s, P = 0.05 (Figure 1(b)).

When levcromakalim was administered directly in the lateral ventricle of the brain, we observed an acute analgesic effect on the hotplate with latency to withdrawal increasing by 4.5 s ± 1.7 s, P = 0.029 (Figure 1(d)). The effect size was smaller and not significant following the second injection on day 3 and no difference was observed at 135 min after administration (Figure 1(e)). There was no difference in tactile sensitivity between vehicle and levcromakalim groups following ICV administration (Fig 1(f)). Correct ICV cannula placement was confirmed by post-mortem injection of dye in all mice (Figure 1(g)). To summarize, levcromakalim IP resulted in delayed hypersensitivity responses at 120 min, whereas levcromakalim ICV provided acute analgesia after 20 min with no delayed effects.

We have previously shown that the K_ATP channel blocker glibenclamide inhibits both GTN- and levcromakalim-induced cutaneous tactile hypersensitivity after systemic IP administration (2). To investigate the site of action of glibenclamide in relation to the blood-brain barrier, we performed DESI-MS visualisation of head frontal sections of animals exposed to an analgesic dose of glibenclamide. Qualitative assessment revealed no substantial presence of glibenclamide inside the blood-brain barrier 120 min after dosing (Figure 1(h)), which is also supported by previous studies (39).

The observed hypersensitivity 120 min after IP administration of levcromakalim could be due to local pronociceptive actions on peripheral nociceptors in the paws. To test this possibility, we injected levcromakalim or vehicle subcutaneously between the foot pads (intraplantarily, IPL). Administration was done on the right hind paw using the contralateral left paw as non-injected control in order to test for the effect of systemic absorption. We observed no effect on tactile sensitivity after levcromakalim IPL in ipsilateral paws at 20 min neither on day 1 nor on day 3, P = 0.41 and 0.98. Similarly, there were no differences between the groups 120 min after injection, P = 0.65 and 0.55 at day 1 and day 3, respectively (Figure 1(i)). Moreover, the contralateral non-injected hind paw was not sensitised (Supplementary Figure 1), showing that the intraplantar levcromakalim doses was not absorbed to the systemic circulation in amounts able to induce hypersensitivity. Hypertonic saline was injected as positive control and induced a prominent hypersensitivity in the ipsilateral paw. Thus, the effect of systemic levcromakalim was not a result of direct action of levcromakalim on peripheral nociceptors in the hind paws.

K_ATP channel subunit distribution in migraine relevant tissues

To identify possible target tissues for levcromakalim-induced hypersensitivity associated with migraine pain, we investigated the mRNA and protein expression of the K_ATP channel subunits Kir6.1, Kir6.2, SUR1, SUR2A and SUR2B in the trigeminal ganglion (TG), trigeminal nucleus caudalis (TNC), dura mater and cerebral blood vessels of the mouse. We found a low or absent mRNA expression of subunits Kir6.1, Kir6.2 and SURs in TG (Figure 2(a)). This was also found in publicly available RNAseq data of TG from human, rat and mouse, where transcripts per million were well below 10 (interpreted as low or absent expression) for all subunits (Table 1). mRNA expression of Kir6.2 and SUR1 was found in TNC. In contrast, the blood vessels and dura mater had dominant mRNA expression of Kir6.1 and SUR2B (Figure 2(a)). Western blotting verified that the protein expression of Kir6.1 was only in dura mater and blood vessels and not in TNC and TG (Figure 2(b)). Immunofluorescent staining showed that the Kir6.1 protein localisation was primarily within the vascular smooth muscle in cerebral blood vessels (Fig 2(c)).

Figure 2.

K_ATP channel subunit expression profile of tissues relevant to migraine. (a) mRNA expression of K_ATP channel subunits of mouse cerebral blood vessels (CBV) n = 14, dura mater n = 3, trigeminal ganglion (TG) n = 9, and trigeminal nucleus caudalis (TNC) n = 9 expressed relative to Hprt1. (b) Protein expression of the Kir6.1 subunit in TNC, TG, dura and CBV (see Figure S2 and S3 for uncut gels and blocking experiment). (C) Representative immunofluorescent staining of Kir6.1 and α-smooth muscle actin (α-SMA) in middle cerebral artery of wild-type mouse, n = 3.

Table 1.

mRNAseq data showing expression of K_ATP subunits and house keeping genes (HPRT1 and AIP) in human, mouse and rat trigeminal ganglion, expressed as transcripts per million. Values under 10 is regarded as none or very limited expression.

Target	Human	Mouse	Rat
SUR1 (ABCC8)	4.65 (3.07)	5.56 (4.52)	3.09 (0.86)
SUR2 (ABCC9)	4.91 (1.77)	0.21 (0.53)	3.19 (1.12)
Kir6.1 (KCNJ8)	5.19 (1.90)	0.30 (0.77)	1.72 (0.67)
Kir6.2 (KCNJ11)	2.05 (1.14)	1.15 (1.34)	3.81 (0.92)
HPRT1	52.56 (22.34)	187.18 (192.45)	103.89 (13.90)
AIP	56.15 (10.18)	143.77 (152.46)	58.96 (2.83)

The Kir6.1 subunit in smooth muscle is essential for cutaneous hypersensitivity after migraine triggers

To understand the role of K_ATP channels in vascular smooth muscle, we repeatedly administrated levcromakalim to Kir6.1 smooth muscle conditional knockout mice (sm22cre + Kir6.1 F/F) (7). The conditional knockout mice showed a significantly reduced response to systemic levcromakalim administration, as compared to appropriate controls (Kir6.1 F/F). On hind paw measurements, there was an overall significant effect of treatment group, F(3,33) = 29.3, P < 0.0001 and the post hoc test showed significance of the Kir6.1 knockout on every test day, P < 0.0001 to 0.01. The cephalic sensitivity was 0.21 SQRT(g) ± 0.04 in the controls vs 0.43 SQRT(g) ± 0.09 in the Kir6.1 knockout, P = 0.03 (Figure 3(b)). A similar protective effect of Kir6.1 smooth muscle conditional knockout was observed after repeated stimulation with GTN, F(3,33) = 109.9, P < 0.0001 and the post hoc comparison again revealed significance of the Kir6.1 deletion on every test day, P < 0.0001 (Figure 3(c)) and on day 11 applying the cephalic von Frey readout P = 0.011 (Figure 3(d)). The effect of Kir6.1 deletion was also evident on basal sensitivity for both levcromakalim and GTN provocations (Supplementary Figure 5).

Figure 3.

Mice with conditional deletion of Kir6.1 in smooth muscle cells are protected from hypersensitivity induced by levcromakalim and GTN. (a) Kir6.1 smooth muscle conditional knockout mice (F/F cre) were protected from levcromakalim- (LEV) induced hypersensitivity at hind paw level (days 1-9), mixed-effects analysis (F_3,33 = 29.3, P < 0.0001), Sidak’s post hoc comparison between all groups: ^#[WT LEV; n = 10 vs WT VEH; n = 10], *[F/F LEV; n = 11 vs F/F cre LEV; n = 7] and (b) in the periorbital area $(day 11), unpaired t-test (t = 2.48, df = 14, P = 0.03). One mouse refused facial testing and was not included. (c) Following 18 days of washout, the protocol was repeated with GTN as inducer of hypersensitivity. Hind paw level (days 1-9), mixed-effects analysis (F_3,33 = 109.9, P < 0.0001), Sidak’s post hoc comparison between all groups: ^#[WT GTN; n = 10 vs WT VEH; n = 10], *[F/F GTN; n = 10 vs F/F cre GTN; n = 7] and (d) in the periorbital area $(day 11), unpaired t-test (t = 2.9, df = 15, P = 0.01). Raw data are available in Supplementary Figure 4 and basal sensitivity data in Supplementary Figure 5.

Arteries from the Kir6.1 smooth muscle conditional knockout mice showed a marked decrease in Kir6.1 expression in the smooth muscle cells compared to the controls (Figure 4(a)). Moreover, arteries had an approximate 50% reduction in their response to levcromakalim as compared to the floxed controls, overall effect of treatment F(1,4) = 8.0, P = 0.047 and levcromakalim 10⁻⁵ M: P = 0.002 and 3 × 10⁻⁵ M: P = 0.001. The K_ATP channel independent vasodilator papaverine were used as positive control, and dilated knockout and control arteries to the same degree (Figure 4(b)–(c)).

Figure 4.

Visual and functional lack of Kir6.1 in the Kir6.1 conditional knockout mouse. (a) Representative immunofluorescent staining for Kir6.1 and α-smooth muscle actin (α-SMA) of middle cerebral arteries from floxed control (Kir6.1 F/F) and Kir6.1 smooth muscle conditional knockout (F/F cre) mice, n = 3. (b) Representative traces of ex vivo dilatory responses of basilar artery segments from Kir6.1 F/F and F/F cre mice stimulated with increasing concentrations of levcromakalim after precontraction with U46619. K_ATP-independent dilation was achieved at last with papaverine as a functional control. (c) The accumulated vasodilatory response to levcromakalim was reduced in the Kir6.1 F/F cre mice, whereas the response to the positive control and maximal dilation remained intact, mixed-effects analysis and Sidak’s post hoc comparison between groups at each stimulation (F_1,4 = 8.0, P = 0.047).

Discussion

Several lines of evidence from this study suggest that K_ATP channels in vascular smooth muscle cells of extracerebral blood vessels play a key role in the development of cutaneous tactile hypersensitivity mediated by migraine-provoking compounds: (i) levcromakalim induced hypersensitivity only by systemic (not central or local) administration; (ii) only blood vessels and dura mater, not trigeminal ganglion and trigeminal nucleus caudalis had expression of K_ATP channel subunit Kir6.1; and (iii) mice with a conditional deletion of K_ATP channel subunit Kir6.1 in smooth muscle cells had decreased arterial dilatory response to levcromakalim and did not develop tactile hypersensitivity as controls following provocations with levcromakalim and GTN.

K_ATP channel effects in- and outside the brain

Comparing acute and delayed nociceptive responses to levcromakalim injected centrally and systemically we found important differences. Following ICV administration, levcromakalim had no acute (20 min post injection) effect on tactile sensitivity, whereas it provided analgesia on the hotplate. Acute analgesic effect of centrally delivered K_ATP channel openers has been described previously and is be believed to be part of the mechanism of action of opioids (4 –6). Activation of opioid receptors stimulates neuronal NO synthase to produce NO which in turn activates a K_ATP channel-dependent potassium current causing hyperpolarisation and decreased neurotransmitter release (6,40,41). The role of K_ATP channels in this cascade has led to the suggestion that K_ATP channel openers could be used as analgesics and in the prevention of opioid tolerance (4,40,41). This effect is acute and short lasting (10-30 min) and central versus peripheral effects are not discussed. One may speculate whether levcromakalim is metabolised within the brain or may be transported out of the CNS (42,43) and perhaps exert systemic effects.

In our data, there were no delayed (120 min post injection) effects after ICV dosing neither anti- nor pronociceptive. On the von Frey measure following ICV dosing we did see a progressive lowering of thresholds that was not significantly different from vehicle, but as there was also a drop in the vehicle group there is a possibility that the vehicle (5% DMSO) may be masking the effect of levcromakalim. This effect was not observed on the hotplate though, and the applied dose was likely insufficient to induce hypersensitivity via the systemic circulation after 2 injections even in case it was transported out of the CNS (2). Contrarily, repeated systemic administration induced hypersensitivity to both heat and tactile stimulation, as also described previously (2,3). Furthermore, the systemic effect was not due to local effects on hind paw K_ATP channels or nociceptors. Similarly, in humans levcromakalim only provokes migraine following systemic administration and not after local subcutaneous injection (1,44). Moreover, in humans levcromakalim dilates extracerebral, but not cerebral arteries upon systemic administration (45) indicating that levcromakalim does not cross the blood-brain barrier. Together with our data, this suggests that the hypersensitivity mediating effects are outside the blood-brain barrier. Of migraine relevant tissue outside the blood-brain barrier, only dura mater had a substantial expression of K_ATP channel subunits. The trigeminal ganglion had very limited expression of all subunits, which was further confirmed by RNAseq data from human, rat and mouse.

Kir6.1 subunit in smooth muscle is important for induced hypersensitivity

Since 1940 when Ray and Wolff described that distention of cranial arteries evoked pain (46), it has been discussed whether vasodilation associated with migraine is the cause of pain or a proxy for central mechanisms (47 –50). In human migraine models, using the same provoking substances as applied here and others like CGRP, dilation of middle meningeal and superficial temporal arteries have been clearly demonstrated (45,51 –53). Moreover, human middle meningeal arteries have previously been shown to express Kir6.1 and SUR2B subunits (54). These K_ATP channels are activated downstream from GTN, CGRP, PACAP and cilostazol (55) that are all triggers of headache and migraine in patients (1,56 –58). Experiments in rodent models suggest the presence of a NO-TRPA1-CGRP-K_ATP signalling pathway: The effect of GTN is completely inhibited by both chemical inhibition and genetic deletion of TRPA1 (3,59,60), while cilostazol (increases cAMP mimicking CGRP receptor activation) and levcromakalim bypasses TRPA1. K_ATP channel inhibition blocks both GTN, cilostazol and levcromakalim induced hypersensitivity indicating that these channels are placed downstream in the proposed signalling pathway (3). Interestingly, CGRP antagonism is also effective in both GTN (10,61) and the cilostazol and levcromakalim models of provoked migraine making the hierarchy of CGRP and K_ATP in the signalling pathway less clear (3).

A 1:1 relationship between vasodilation and pain has not been found in provoked (62) or in spontaneous migraine (63). Time delay from arterial dilation to pain perception and fluctuations in the threshold to pain perception likely contribute to the lack of a direct association (64,65). A critical question is how a normal physiological phenomenon, arterial dilation mediated by K_ATP channels, results in hypersensitivity in animals and is perceived as painful in migraine patients? Likely candidates for future work on the translation of vasodilation to trigeminal pain processing could be the stretch-sensing TRPM8 channel that has been associated with migraine in a large genome-wide association study (66) together with other TRP channels (67), the ASIC- channels (68) and Piezo channels (69). Alternatively, potassium released through vascular smooth muscle K_ATP channels into the neurovascular microenvironment could mediate a direct effect on trigeminal nerve endings surrounding the blood vessels.

Strengths and limitations

The translational limitations of mouse cutaneous hypersensitivity to migraine pain are to be noted. The administered provokers of hypersensitivity in the mouse model are known triggers of migraine in humans and do not cause other pains (12,44). The GTN mouse model is validated in terms of response to specific anti-migraine drugs, so several lines of evidence indicate its relevance to human migraine (8 –11). Activation of headache associated nociceptors in the trigeminovascular system followed by sensitization of second-order neurons in the TNC mediate cephalic hypersensitivity in migraine. Further, whole-body cutaneous hypersensitivity seen in some migraine patients is mediated through activation of neurons in the thalamus (70,71). Together, this suggests that both cephalic and whole-body hypersensitivity can be used as an indication of nociceptive activation of the trigeminovascular system. We used hind paw testing as primary readout in this study as cephalic sensitivity is less robust towards repeated testing, has a narrow effect window and is more variable thereby requiring much larger group sizes to perform this kind of study. This was further discussed by Christensen et al. (3). Cephalic von Frey test was performed in the Kir6.1 knockout and controls only as it was not feasible to perform this test simultaneously in the full cohort of mice. Cephalic hypersensitivity following GTN (10,20,60) and levcromakalim (3) has previously been shown in wild-type mice.

Our study was limited by the very small size of mouse meningeal arteries which are not suitable for direct in vivo or ex vivo investigations. Studies were performed on the larger intracerebral basilar artery that has the same K_ATP channel subunit expression pattern as in dura mater. A major difference between the meningeal arteries and the basilar artery is the blood-brain barrier. However, in the ex vivo myograph experimental setup the test compounds bypass the blood-brain barrier. Previous studies in rat and pig have shown effect of K_ATP channel openers and blockers on the middle meningeal artery (72 –74). These effects were similar to the effect in rat basilar arteries (74,75). Taken together, these data suggest that the results showed here are representative of the mouse middle meningeal artery in the dura mater.

Conclusion

The evidence that K_ATP channels are possible targets for a new class of migraine drugs has been slowly but steadily building up in recent years. Here, we narrowed down the target to subunit Kir6.1 or its associated regulatory SUR2B subunit. This evidence is important as a unique subunit combination is found in different tissues. Targeted inhibition of the K_ATP channel isoform Kir6.1 in the vasculature is possible without disrupting K_ATP channels of other configurations. Furthermore, according to the present data, a selective isoform antagonist should not cross the blood-brain barrier.

Key findings

K_ATP channel opener levcromakalim triggers hypersensitivity in mouse models only after systemic – not central or local – administration.

Smooth muscle K_ATP channels mediate GTN- and levcromakalim- induced hypersensitivity.

Supplemental Material

sj-pdf-1-cep-10.1177_03331024211053570 - Supplemental material for Smooth muscle ATP-sensitive potassium channels mediate migraine-relevant hypersensitivity in mouse models

Supplemental material, sj-pdf-1-cep-10.1177_03331024211053570 for Smooth muscle ATP-sensitive potassium channels mediate migraine-relevant hypersensitivity in mouse models by Sarah L Christensen, Rikke H Rasmussen, Sanne La Cour, Charlotte Ernstsen, Thomas F Hansen, Lisette JA Kogelman, Sabrina P Lauritzen, Gintare Guzaite, Bjarne Styrishave, Christian Janfelt, Søren T Christensen, Qadeer Aziz, Andrew Tinker, Inger Jansen-Olesen, Jes Olesen and David M Kristensen in Cephalalgia

Footnotes

Author contributions

SLC and DMK designed and interpreted the experiments with advice from JO. SLC, DMK and CE performed in vivo experiments. RHR and IJ-O did myograph experiments, TFH and LJAK did the bioinformatics, SLC and SPL did qPCR and western blotting, GG, BS and CJ performed DESI-MSI, STC did the immunofluorescence, and QA and AT bred and developed smooth muscle conditional knockout mice. SLC, DMK and JO wrote the manuscript with contributions from all other authors.

Declaration of conflicting interests

The authors declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: JO owns stocks in the biotech start-up company Cephagenix. All other authors have declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was supported by Candys Foundation and Læge Sofus Carl Emil Friis og Hustru Olga Doris Friis' Legat.

Data availability

Most raw data are provided as supplementary material. Additional data will be provided upon reasonable request.

ORCID iDs

Sarah L Christensen

Rikke H Rasmussen

Charlotte Ernstsen

Lisette JA Kogelman

Søren T Christensen

Qadeer Aziz

Inger Jansen-Olesen

References

Al-Karagholi

MA-M

Hansen

Guo

, et al. Opening of ATP-sensitive potassium channels causes migraine attacks: a new target for the treatment of migraine. Brain 2019; 142: 2644–2654.

Christensen

Munro

Petersen

, et al. ATP sensitive potassium (KATP) channel inhibition: A promising new drug target for migraine. Cephalalgia 2020; 40: 650–664.

Christensen

Rasmussen

Ernstsen

, et al. CGRP-dependent signalling pathways involved in mouse models of GTN- cilostazol- and levcromakalim-induced migraine. Cephalalgia Epub ahead of print 18 August 2021. DOI: 10.1177/03331024211038884.

Khanna

Malhotra

Mehta

, et al. Interaction of morphine and potassium channel openers on experimental models of pain in mice. Fundam Clin Pharmacol 2011; 25: 479–484.

Lohmann

Welch

SP.

Antisenses to opioid receptors attenuate ATP-gated K+ channel opener-induced antinociception. Eur J Pharmacol 1999; 384: 147–152.

Narita

Suzuki

Misawa

, et al. Role of central ATP-sensitive potassium channels in the analgesic effect and spinal noradrenaline turnover-enhancing effect of intracerebroventricularly injected morphine in mice. Brain Res 1992; 596: 209–214.

Aziz

Thomas

Gomes

, et al. The ATP-sensitive potassium channel subunit, Kir6.1, in vascular smooth muscle plays a major role in blood pressure control. Hypertension 2014; 64: 523–529.

Pradhan

Smith

McGuire

, et al. Characterization of a novel model of chronic migraine. Pain 2014; 155: 269–274.

Bates

Nikai

Brennan

, et al. Sumatriptan alleviates nitroglycerin-induced mechanical and thermal allodynia in mice. Cephalalgia 2010; 30: 170–178.

10.

Christensen

Petersen

Kristensen

, et al. Targeting CGRP via receptor antagonism and antibody neutralisation in two distinct rodent models of migraine-like pain. Cephalalgia 2019; 39: 1827–1837.

11.

Akerman

Romero-Reyes

Karsan

, et al. Therapeutic targeting of nitroglycerin-mediated trigeminovascular neuronal hypersensitivity predicts clinical outcomes of migraine abortives. Pain 2021; 162: 1567–1577.

12.

Ashina

Hansen

Olesen

Pearls and pitfalls in human pharmacological models of migraine: 30 years’ experience. Cephalalgia 2013; 33: 540–553.

13.

Narita

Takamori

Kawashima

, et al. Activation of central ATP-sensitive potassium channels produces the antinociception and spinal noradrenaline turnover-enhancing effect in mice. Psychopharmacol (Berl) 1993; 113: 11–14.

14.

Kamei

Kawashima

Narita

, et al. Reduction in ATP-sensitive potassium channel-mediated antinociception in diabetic mice. Psychopharmacol (Berl) 1994; 113: 318–321.

15.

Perrotta

Serrao

Tassorelli

, et al. Oral nitric-oxide donor glyceryl-trinitrate induces sensitization in spinal cord pain processing in migraineurs: a double-blind, placebo-controlled, cross-over study. Eur J Pain 2011; 15: 482–490.

16.

De Icco

Perrotta

Grillo

, et al. Experimentally induced spinal nociceptive sensitization increases with migraine frequency: a single-blind controlled study. Pain 2020; 161: 429–438.

17.

Chaplan

Bach

Pogrel

, et al. Quantitative assessment of tactile allodynia in the rat paw. J Neurosci Methods 1994; 53: 55–63.

18.

Dixon

WJ.

Efficient Analysis of Experimental Observations. Annu Rev Pharmacol Toxicol 1980; 20: 441–462.

19.

Burgos-Vega

Quigley

Trevisan Dos Santos

, et al. Non-invasive dural stimulation in mice: A novel preclinical model of migraine. Cephalalgia 2019; 39: 123–134.

20.

Ben Aissa

Tipton

Bertels

, et al. Soluble guanylyl cyclase is a critical regulator of migraine-associated pain. Cephalalgia 2018; 38: 1471–1484.

21.

Christensen

Hansen

Storm

, et al. Von Frey testing revisited: Provision of an online algorithm for improved accuracy of 50% thresholds. Eur J Pain 2020; 24: 783–790.

22.

Bradman

MJG

Ferrini

Salio

, et al. Practical mechanical threshold estimation in rodents using von Frey hairs/Semmes-Weinstein monofilaments: Towards a rational method. J Neurosci Methods 2015; 255: 92–103.

23.

Thunig

Hansen

Janfelt

Analysis of secondary plant metabolites by indirect desorption electrospray ionization imaging mass spectrometry. Anal Chem 2011; 83: 3256–3259.

24.

Janfelt

Wellner

Hansen

, et al. Displaced dual-mode imaging with desorption electrospray ionization for simultaneous mass spectrometry imaging in both polarities and with several scan modes. J Mass Spectrom 2013; 48: 361–366.

25.

Janfelt

Wellner

Leger

, et al. Visualization by mass spectrometry of 2‐dimensional changes in rat brain lipids, including N ‐acylphosphatidylethanolamines, during neonatal brain ischemia. FASEB J 2012; 26: 2667–2673.

26.

Schramm

Hester

Klinkert

, et al. ImzML - A common data format for the flexible exchange and processing of mass spectrometry imaging data. J Proteomics 2012; 75: 5106–5110.

27.

Bokhart

Nazari

Garrard

Muddiman

DC.

MSiReader v1.0: Evolving open-source mass spectrometry imaging software for targeted and untargeted analyses. J Am Soc Mass Spectrom 2018; 29: 8–16.

28.

LaPaglia

Sapio

Burbelo

, et al. RNA-Seq investigations of human post-mortem trigeminal ganglia. Cephalalgia 2018; 38: 912–932.

29.

Nguyen

Bonilla

, et al. Diversity amongst trigeminal neurons revealed by high throughput single cell sequencing. PLoS One 2017; 12: e0185543.

30.

Lopes

Denk

McMahon

SB.

The molecular fingerprint of dorsal root and trigeminal ganglion neurons. Front Mol Neurosci 2017; 10.

31.

Gerhold

Pellegrino

Tsunozaki

, et al. The star-nosed mole reveals clues to the molecular basis of mammalian touch. PLoS One 2013; 8: e55001.

32.

Kogelman

LJA

Christensen

Pedersen

, et al. Whole transcriptome expression of trigeminal ganglia compared to dorsal root ganglia in Rattus Norvegicus. Neuroscience 2017; 350: 169–179.

33.

Babraham Bioinformatics - FastQC A Quality Control tool for High Throughput Sequence Data [Internet]. [Cited 8 September 2020]. Available from: https://www.bioinformatics.babraham.ac.uk/projects/fastqc/

34.

Patel

Jain

NGS QC Toolkit: A Toolkit for Quality Control of Next Generation Sequencing Data. Liu Z, editor. PLoS One 2012; 7: e30619.

35.

Bray

Pimentel

Melsted

Pachter

Near-optimal probabilistic RNA-seq quantification. Nat Biotechnol 2016; 34: 525–527.

36.

Durinck

Spellman

Birney

, et al. Mapping identifiers for the integration of genomic datasets with the R/Bioconductor package biomaRt. Nat Protoc 2009; 4: 1184–1191.

37.

Soneson

Love

Robinson

MD.

Differential analyses for RNA-seq: Transcript-level estimates improve gene-level inferences. F1000Res 2015; 4: 1521.

38.

Christensen

Ernstsen

Olesen

, et al. No central action of CGRP antagonising drugs in the GTN mouse model of migraine. Cephalalgia 2020; 40: 924–934.

39.

Lahmann

Kramer

Ashcroft

FM.

Systemic administration of glibenclamide fails to achieve therapeutic levels in the brain and cerebrospinal fluid of rodents. PLoS One 2015; 10: e0134476.

40.

Ocaña

Cendán

Cobos

, et al. Potassium channels and pain: Present realities and future opportunities. Eur J Pharmacol 2004; 500: 203–219.

41.

Tsantoulas

McMahon

SB.

Opening paths to novel analgesics: The role of potassium channels in chronic pain. Trends Neurosci 2014; 37: 146–158.

42.

Strazielle

Ghersi-Egea

JF.

Physiology of blood-brain interfaces in relation to brain disposition of small compounds and macromolecules. Mol Pharm 2013; 10: 1473–1491.

43.

Zhang

Pardridge

WM.

Mediated efflux of IgG molecules from brain to blood across the blood-brain barrier. J Neuroimmunol 2001; 114: 168–172.

44.

Al-Karagholi

MA-M

Ghanizada

Hansen

, et al. Extracranial activation of ATP-sensitive potassium channels induces vasodilation without nociceptive effects. Cephalalgia 2019; 39: 1789–1797.

45.

Al-Karagholi

MA-M

Ghanizada

Hansen

, et al. Levcromakalim, an adenosine triphosphate-sensitive potassium channel opener, dilates extracerebral but not cerebral arteries. Headache 2019; 59: 1468–1480.

46.

Ray

Wolff

HG.

Experimental studies on headache. Arch Surg 1940; 41: 813.

47.

Mason

Russo

AF.

Vascular contributions to migraine: Time to revisit?

Front Cell Neurosci 2018; 12: 233.

48.

Goadsby

Holland

Martins-Oliveira

, et al. Pathophysiology of migraine: A disorder of sensory processing. Physiol Rev 2017; 97: 553–622.

49.

Moskowitz

MA.

Neurogenic inflammation in the pathophysiology and treatment of migraine. Neurology 1993; 43: S16–20.

50.

Goadsby

PJ.

The vascular theory of migraine–a great story wrecked by the facts. Brain. 2009; 132: 6–7.

51.

Asghar

Hansen

Kapijimpanga

, et al. Dilation by CGRP of middle meningeal artery and reversal by sumatriptan in normal volunteers. Neurology 2010; 75: 1520–1526.

52.

Thomsen

LL.

Investigations into the role of nitric oxide and the large intracranial arteries in migraine headache. Cephalalgia 1997; 17: 873–895.

53.

Khan

Amin

Christensen

, et al. Meningeal contribution to migraine pain: a magnetic resonance angiography study. Brain 2019; 142: 93–102.

54.

Ploug

Sørensen

Strøbech

, et al. KATP channels in pig and human intracranial arteries. Eur J Pharmacol 2008; 601: 43–49.

55.

Ashina

Migraine. Ropper AH, editor. N Engl J Med 2020; 383: 1866–1876.

56.

Guo

Olesen

Ashina

Phosphodiesterase 3 inhibitor cilostazol induces migraine-like attacks via cyclic AMP increase. Brain 2014; 137: 2951–2959.

57.

Iversen

Olesen

Tfelt-Hansen

Intravenous nitroglycerin as an experimental model of vascular headache. Basic characteristics. Pain 1989; 38: 17–24.

58.

Schytz

Birk

Wienecke

, et al. PACAP38 induces migraine-like attacks in patients with migraine without aura. Brain 2009; 132: 16–25.

59.

Demartini

Tassorelli

Zanaboni

, et al. The role of the transient receptor potential ankyrin type-1 (TRPA1) channel in migraine pain: evaluation in an animal model. J Headache Pain 2017; 18: 94.

60.

Marone

De Logu

Nassini

, et al. TRPA1/NOX in the soma of trigeminal ganglion neurons mediates migraine-related pain of glyceryl trinitrate in mice. Brain 2018; 141: 2312–2318.

61.

Greco

Mangione

Siani

, et al. Effects of CGRP receptor antagonism in nitroglycerin-induced hyperalgesia. Cephalalgia 2014; 34: 594–604.

62.

Schoonman

van der Grond

Kortmann

, et al. Migraine headache is not associated with cerebral or meningeal vasodilatation–a 3T magnetic resonance angiography study. Brain 2008; 131: 2192–200.

63.

Amin

Asghar

Hougaard

, et al. Magnetic resonance angiography of intracranial and extracranial arteries in patients with spontaneous migraine without aura: a cross-sectional study. Lancet Neurol 2013; 12: 454–461.

64.

Goadsby

PJ.

Pathophysiology of migraine. Ann Indian Acad Neurol 2012; 15: S15–22.

65.

May

Burstein

Hypothalamic regulation of headache and migraine. Cephalalgia 2019; 39: 1710–1719.

66.

Gormley

Anttila

Winsvold

, et al. Meta-analysis of 375,000 individuals identifies 38 susceptibility loci for migraine. Nat Genet 2016; 48: 856–866.

67.

Dussor

Yan

Xie

, et al. Targeting TRP channels for novel migraine therapeutics. ACS Chem Neurosci 2014; 5: 1085–1096.

68.

Karsan

Gonzales

Dussor

Targeted acid-sensing ion channel therapies for migraine. Neurotherapeutics 2018; 15: 402–414.

69.

Della Pietra

Mikhailov

Giniatullin

The emerging role of mechanosensitive piezo channels in migraine pain. Int J Mol Sci 2020; 21: 696.

70.

Burstein

The development of cutaneous allodynia during a migraine attack. Clinical evidence for the sequential recruitment of spinal and supraspinal nociceptive neurons in migraine. Brain 2000; 123: 1703–1709.

71.

Burstein

Jakubowski

Garcia-Nicas

, et al. Thalamic sensitization transforms localized pain into widespread allodynia. Ann Neurol 2010; 68: 81–91.

72.

Syed

Koide

Brayden

, et al. Tonic regulation of middle meningeal artery diameter by ATP-sensitive potassium channels. J Cereb Blood Flow Metab 2019; 39: 670–679.

73.

Gozalov

Jansen-Olesen

Klaerke

, et al. Role of K ATP channels in cephalic vasodilatation induced by calcitonin gene-related peptide, nitric oxide, and transcranial electrical stimulation in the rat. Headache 2008; 48: 1202–1213.

74.

Ploug

Boni

Baun

, et al. K ATP channel expression and pharmacological in vivo and in vitro studies of the K ATP channel blocker PNU-37883A in rat middle meningeal arteries. Br J Pharmacol 2008; 154: 72–81.

75.

Ploug

Edvinsson

Olesen

, et al. Pharmacological and molecular comparison of KATP channels in rat basilar and middle cerebral arteries. Eur J Pharmacol 2006; 553: 254–262.

Smooth muscle ATP-sensitive potassium channels mediate migraine-relevant hypersensitivity in mouse models

Abstract

Background

Methods

Results

Conclusion

Keywords

Introduction

Materials and methods

Experimental animals

Mouse models of migraine

Intracerebroventricular administration of levcromakalim

Intraplantar administration of levcromakalim

Cutaneous sensitivity to tactile stimulation

Cutaneous sensitivity to heat stimulation

Ex vivo vasoactivity

Desorption Electrospray Ionization - Mass Spectrometry Imaging (DESI-MSI)

mRNA expression studies

Bioinformatics, RNA-sequencing in human, mouse and rat

Protein expression studies

Immunofluorescence

Statistical analyses

Results

Systemic but not central or local KATP channel activation induces cutaneous hypersensitivity

KATP channel subunit distribution in migraine relevant tissues

The Kir6.1 subunit in smooth muscle is essential for cutaneous hypersensitivity after migraine triggers

Discussion

KATP channel effects in- and outside the brain

Kir6.1 subunit in smooth muscle is important for induced hypersensitivity

Strengths and limitations

Conclusion

Key findings

Supplemental Material

sj-pdf-1-cep-10.1177_03331024211053570 - Supplemental material for Smooth muscle ATP-sensitive potassium channels mediate migraine-relevant hypersensitivity in mouse models

Footnotes

Author contributions

Declaration of conflicting interests

Funding

Data availability

ORCID iDs

References

Systemic but not central or local K_ATP channel activation induces cutaneous hypersensitivity

K_ATP channel subunit distribution in migraine relevant tissues

K_ATP channel effects in- and outside the brain