Sage Journals: Discover world-class research

Abstract

Introduction

Carbon monoxide was previously considered to just be a toxic gas. A wealth of recent information has, however, shown that it is also an important endogenously produced signalling molecule involved in multiple biological processes. Endogenously produced carbon monoxide may thus play an important role in nociceptive processing and in regulation of cerebral arterial tone.

Discussion

Carbon monoxide-induced headache shares many characteristics with migraine and other headaches. The mechanisms whereby carbon monoxide causes headache may include hypoxia, nitric oxide signalling and activation of cyclic guanosine monophosphate pathways. Here, we review the literature about carbon monoxide-induced headache and its possible mechanisms.

Conclusion

We suggest, for the first time, that carbon monoxide may play an important role in the mechanisms of migraine and other headaches.

Keywords

Carbon monoxide haeme oxygenase headache pain processing cerebral vessels

Introduction

Carbon monoxide (CO) is well known for its neurotoxic effects, but recent studies suggest that CO is also a pain-modulating neurotransmitter. The International Classification of Headache Disorders (ICHD-III beta) (1) defines carbon monoxide-induced headache as a bilateral headache, with quality and intensity that is related to the severity of CO intoxication (Table 1). According to ICHD-III beta (1), it develops within 12 hours of exposure and resolves within 72 hours after elimination of CO. CO headache shares many characteristics with migraine. It is usually frontal, moderate to severe, aggravated by physical activity, and nausea and vomiting are also common symptoms (1 –4). Hypoxia was previously thought to be the sole mechanism behind CO-induced headache (5,6), but CO headache may also appear after exposure to non-toxic doses (2,6,7). Interestingly, CO is now known to be endogenously produced (8) and recent data suggest that CO is involved in pain transmission via cyclic guanosine monophosphate (cGMP) pathways (9 –11). Furthermore, CO interacts with nitric oxide (NO), a crucial molecule in migraine (12), dilates cerebral vessels (13 –15) and may cause production of free radicals (16,17). These data suggest that CO may be importantly implicated in the mechanisms of migraine and other headaches. A thorough review of CO and its headache-generating mechanisms, therefore, seems timely.

Table 1.

The diagnostic criteria for carbon monoxide-induced headache.

(a) Bilateral headache fulfilling criterion (c)

(b) Exposure to carbon monoxide has occurred

1. headache has developed within 12 hours of exposure to carbon monoxide

2. headache intensity varies with the severity of carbon monoxide intoxication

3. headache has resolved within 72 hours of elimination of carbon monoxide

(d) Not better accounted for by another ICHD-3 diagnosis

Comments: Typical symptoms at carboxyhaemoglobin levels of

10%–20% mild headache without gastrointestinal or neurological symptoms

20%–30% moderate pulsating headache and irritability

30%–40% severe headache with nausea, vomiting and blurred vision

Adapted from the third edition of the International Classification of Headache Disorders (ICHD-III beta) (1).

Biological chemistry and body stores

CO binds to haeme with an affinity 200–250 times higher than oxygen (18) and increases the affinity of haeme groups to oxygen (Figure 1) (19). Thus, CO both decreases the oxygen-carrying capacity and impairs oxygen release to the tissues. Hence, during exogenous CO exposure carboxyhaemoglobin (COHb) levels increase and oxygen-demanding tissues will eventually become hypoxic (20).

Figure 1.

CO sources and body stores. CO in the body originates from exogenous uptake by the lungs from ambient air and from endogenous production by the cells in the body. The majority of CO (80%) binds to haemoglobin as carboxyhaemoglobin (COHb). The minority of CO (15%) binds to extravascular haeme-proteins. CO: carbon monoxide; COHb: carboxyhaemoglobin; NADPH: reduced nicotinamide adenine dinucleotide phosphate; NADP+: oxidised nicotinamide adenine dinucleotide phosphate.

The hypoxic effect of CO was previously thought to be the only explanation for the symptoms of CO intoxication (5,6). The dogma was challenged by Goldbaum (21) in a canine model, where blood was either replaced by blood containing 80% COHb or CO was injected intraperitoneally. Following both of these exposures the animals showed no symptoms of intoxication. By contrast, animals that inhaled 13% CO, reaching average COHb levels of 64%, died within an hour. The authors explained this difference by the CO effect on extravascular haeme-proteins (21) since 15% CO in the body is bound to extravascular haeme-proteins (Figure 1) (8). They argued that tissue uptake of CO is relatively high in the heart and brain because of low oxygen tension in these rapidly metabolising tissues and a high CO tension in the pulmonary veins (21). This is supported by a significant CO tension in the pulmonary capillary blood (22) and a relatively slow saturation of haemoglobin with CO in erythrocytes (21).

The cellular uptake of CO is highly dependent on the partial pressure of oxygen because CO and oxygen compete for the same iron and copper binding sites. Thus, at low oxygen tension CO binds cellular haeme proteins noncompetitively and less CO is needed to produce an effect (8). Therefore, the clinical effects are probably due to a combination of COHb formation and the direct effects of CO at the cellular level.

CO as a pain-modulating neurotransmitter

CO is synthesised endogenously by oxidative cleavage of haeme by the enzyme haeme oxygenase (HO) (Figure 1) (11). There are two major HO isoforms: HO-1 and HO-2 (8,11). They facilitate the same synthesis, but their distribution, regulation and function differ. HO-1 is poorly expressed in the nervous system, but can experimentally be induced by hypoxia, heat shock and endotoxins and seems likely to antagonise oxidative stress (11,23,24). In relation to pain, the most interesting enzyme is HO-2, which is widely distributed and continuously expressed in cortical regions, hippocampus, basal ganglia, hypothalamus, cerebellum, brain stem and the spinal cord (Figure 2) (9,25,26). HO-2 is also expressed in cells in the trigeminal, sphenopalatine, superior cervical and dorsal root ganglia (Figure 2) (9,23,27), which are important nociceptive structures of the peripheral nervous system (28). In human autopsies, HO-2 is expressed in 40% of the cell bodies in the superior cervical ganglia and in 60% of the cell bodies in the trigeminal ganglia. Forty per cent of the trigeminal cells also co-expressed calcitonin gene-related peptide (CGRP) (9), a well-known migraine trigger (29). HO-induced cleavage of haeme results in CO, biliverdin and free iron (Figure 1). Since biliverdin and free iron are not known to have catalysing abilities, HO-activity is assumed to be proportional to the amount of CO in the process (11) and the important functions of HO-2 in nociception are likely exerted by CO as a neurotransmitter.

Figure 2.

Haeme oxygenase 2 (HO-2) distribution in the central and peripheral nervous system. Animal studies reported an abundant presence of HO-2 mRNA and protein throughout the nervous system investigated by in situ hybridization, polymerase chain reaction and immunohistochemistry (10,15,23 –27). In addition, HO-2 was found in the nodose, jugular and otic ganglia (23,27). Only one study investigated human autopsies in the peripheral ganglia and found HO-2 in the trigeminal and superior cervical ganglia (9). HO-2: haeme oxygenase 2; CNS: central nervous system; PNS: peripheral nervous system.

In a rat model of neuropathic pain Gordh et al. (30) showed an ipsilateral upregulation of HO-2 expression in the grey matter of the spinal cord at the level of the ligated spinal nerves (Table 2). In a similar model Li and Clark (31) showed an increase in total HO activity in the spinal cord segment ipsilateral to the ligated nerve root or paw incision, using a non-specific HO assay. In the same study, subcutaneous administration of blood-brain barrier (BBB)-permeable HO-inhibitors reduced pain, whereas non-BBB-permeable HO-inhibitors had no analgesic effect (31). In a formalin-induced nociception mouse model, BBB-permeable HO-inhibitors also dose-dependently reduced hyperalgesia, whereas a BBB-non-permeable drug had no effect (32). The observed increase in spinal cord HO activity during pain and the analgesic effect of BBB-permeable HO-inhibitors implies that HO exerts its effects on central nociceptive signalling rather than at the peripheral nervous system. The pro-nociceptive effect of HO-2 has also been explored in HO-2-null mutant mice in different pain models. Thus, HO-2-null mutant mice show reduced pain-related behaviour compared to wild-type mice in glutamate (33), incisional and inflammatory pain models (32,34). An interesting case of HO-1 deficiency in a 6-year-old boy with severe growth retardation has been reported, but so far no human data on the possible relationship between HO deficiency and headache or cases of HO-2 deficiency have been reported (35). Data from the above animal studies are summarised in Table 2.

Table 2.

Role of carbon monoxide in nociception in animal models of pain.

Study description		Major findings
Animal, pain model	Methods	HO in the spinal cord	Non-selective HO inhibitors	HO-2 null animals compared to wild type
Rat, neuropathic	• Spinal nerve ligation (30,31) • Partial sciatic nerve ligation (34)	HO-2 expression ↑(30) Total HO activity ↑(31)	• BBB permeable: pain behaviours ↓(31) • BBB impermeable: no effect (31)	• Pain behaviours ↓(34) • Brush-induced spinal cord Fos expression ↓(34)
Rat, incisional	• Paw incision (31)	Total HO activity ↑	• BBB permeable: pain behaviours ↓ • BBB impermeable: no effect	Not performed
Mouse, inflammatory	• Paw formalin injection (32,36,37) • Freund’s adjuvant injection (34)	Not measured	• BBB permeable: pain behaviours ↓(32) and formalin-induced spinal cord Fos expression ↓(36) • BBB impermeable: no effect (32)	• Pain behaviours ↓(32,34,37) • Nociception-induced expression of nociception-related genes ↓(34,36,37)
Mouse: glutamate induced	• Intrathecal injection of glutamate, NMDA and AMPA (33)	Not measured	• BBB permeable: pain behaviours ↓and glutamate-, NMDA- and AMPA-induced cGMP accumulation in spinal cord blocked	• Pain behaviours ↓ • No cGMP response to glutamate

Arrow up: increased; arrow down: decreased. HO-2: heme oxygenase 2; BBB: blood-brain barrier; NMDA: N-methyl-D-aspartate; AMPA: alpha-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid; cGMP: cyclic guanosine monophosphate.

Some studies have, however, suggested that endogenous CO may have antinociceptive effects in the spinal cord and in the peripheral nervous system. Treatment with a non-specific HO-inhibitor stimulated nociceptor hypersensitivity in a mechanical hypersensitivity rat model (38) and a formalin model (39). This effect was countered by CO administration and blocked by an inhibitor of soluble guanylate cyclase (sGC) (38). Interestingly, HO-1 exerts antinociceptive effects and HO-1 inducers have been suggested as new analgesics (40). Thus, increased knowledge of the different properties of HO-1 and HO-2 is important to understand the complex role of CO in nociception. It should be noted that all the above-mentioned experiments on CO have been conducted in non-headache models of pain, and research on the pro- or antinociceptive role of CO in headache models is needed.

Pain- and headache-inducing mechanisms of CO

Activation of the cGMP pathway is likely an important part of migraine and other headache pathophysiology (12). cGMP is converted from guanosine triphosphate (GTP) by sGC and can in turn activate cGMP-dependent ion channels, phosphodiesterases and protein kinases and thereby affect numerous intracellular functions (12). There is evidence to suggest a correlation between HO activation, sGC and cGMP in the processing of pain. Thus, in a mouse model, the normal cGMP increase after intrathecal glutamate injection is blocked by HO-inhibitors (33) and the increase is not found in HO-null mutant mice (Table 2) (33). sGC and HO-2 have similar locations in the central nervous system and administration of an HO-inhibitor depletes cGMP content in olfactory neurons (10). CO has high affinity to haeme in sGC and binding stimulates the activity of sGC (10,41). CO can increase cGMP production in spinal cord slices (42) and may regulate glutamate release (Figure 3). CO increases glutamate release in vivo in rat cortex and hippocampus (43). In synaptic preparations from rat cortex depolarisation-dependent release of glutamate could be inhibited by a non-specific HO-inhibitor (44).

Figure 3.

CO may induce headache and pain through the following mechanisms: 1) hypoxia, 2) impairment of mitochondrial function, 3) modulation of gene expression, 4) activation of cGMP signalling pathways, 5) activation of calcium-gated potassium channels, 6) interactions with nitric oxide. CO: carbon monoxide; sGC: soluble guanylate cyclase; cGMP: cyclic guanosine monophosphate; NOS: nitric oxide synthase; NO: nitric oxide; K_Ca+-channels: calcium-gated-potassium channels; ROS: reactive oxygen species; COHb: carboxyhaemoglobin.

HO-2 may also be a regulator of nociception-related gene expression (Figure 3). In inflammatory and neuropathic pain models HO-2-null mice or mice treated with HO-inhibitors showed reduced Fos expression in dorsal horn neurons after noxious stimuli (34,36). The same was found for c-Fos, c-Jun, Jun-B and nerve growth factor-induced genes (Table 2) (37). Further studies in anatomic compartments related to headache are needed to understand the impact of CO on pain signalling and perception.

CO as a cerebral vasodilator

Paulson et al. (45) showed that exposure to moderate amounts of CO (mean COHb 20%) increased human cerebral blood flow (CBF) by 26%. This was regarded as a hypoxia-induced compensatory response. However, the CBF response to CO in humans is highly variable and not well correlated with COHb (46). Animal in vivo and in vitro studies provide evidence that CO affects the cerebral circulation directly (13,14,47,48). The dilatory effect of CO on cerebral arteries is interesting as sensory input from perivascular nociceptors is important in migraine headache (28) and since cerebral vessels are dilated during spontaneous migraine attacks (49). Ex vivo experiments suggest that CO induces dilation of cerebral and systemic arteries (Table 3) (47,48,50,51). In vivo, topical application of CO dilates pial arterioles in newborn pigs in a dose-dependent manner (Table 3) (13,26).

Table 3.

Role of carbon monoxide in the cerebral circulation.

Study description		Major findings
Type	Methods	Vascular response and CO production	HO-Inhibitors	Other inhibitors
Piglet, in vivo	• CO exposure in closed cranial window (13,26,52,53)	• Arteriolar dilation (13,26,52,53)	• Dilation blocked (26)	• Dilation blocked by inhibitors of nitric oxide synthase (52), K_Ca+-channels (26) and soluble guanylate cyclase (52,53)
Piglet, in vitro	• CO exposure to arteriolar branches of middle cerebral artery (54)	• Arteriolar dilation	Not performed	• Dilation blocked by inhibitors of nitric oxide synthase and K_Ca+-channels
Dog (47), mouse (48), ex vivo	• CO exposure to basilar artery (47) and middle cerebral artery (48)	• Arterial dilation (47,48)	• Arterial constriction (48)	• Dilation blocked by inhibitors of soluble guanylate cyclase (47) and K_Ca+-channels (47)
Piglet, in vivo	• In closed cranial window exposure to glutamate (13), AA, PGE (14) and ADP (55)	• Arteriolar dilation (13,14,55) • CO production ↑(13,14,55)	• Blocked arteriolar dilation (14,55) • Blocked CO ↑(14)	Not performed
Piglet, in vitro	• Glutamate exposure to cerebral arteriole myocytes and astrocytes (15)	• CO production ↑ in astrocytes • activation of K_Ca+-channels in myocytes • In HO-2-null piglets above responses absent	• Activation of K_Ca+-channels absent	Not performed
Piglet, in vitro	• ADP exposure to astrocytes and cerebral microvessels (55) • Glutamate exposure to astrocytes (56)	• CO production ↑ in astrocytes (55,56) and cerebral microvessels (55)	• Blocked CO ↑(56)	Not performed

Arrow up: increased; arrow down: decreased. CO: carbon monoxide; HO-2: haeme oxygenase 2; K_Ca+-channels: calcium-gated-potassium channels; AA: arachidonic acid; PGE: prostaglandin E2; ADP: adenosine diphosphate.

HO-2 is highly expressed in large and small cerebral vessels, astrocytes and neurons (15,24,26). Astrocytes are key contributors to chronic pain mechanisms (57) and play a key role in the ability of neurons to regulate smooth muscle tone in cranial blood vessels in response to metabolic needs. Glutamate (15) and adenosine diphosphate (ADP) (55) increase CO production in isolated astrocytes in piglets and this is abolished in HO-2-null mutant piglets (15). Injury to astrocytes blocks glutamate (13) and ADP (55)-dependent vasodilation and CO production. In artificial closed cranial window models, glutamate- (13), ADP- (55), prostaglandin E2 (PGE₂-) and arachidonic acid (AA) (14)-induced arteriolar dilation was accompanied by increased CO concentration in the cerebrospinal fluid. In these studies dilation of piglet arterioles was blocked by pretreatment with an HO-inhibitor or genetic ablation of HO-2 (14,15,55). This suggests that glutamate, ADP, PGE₂ and AA dilate pial arterioles in piglets through CO and hence CO may be an important diffusible messenger in neurovascular coupling. Abundant evidence suggests that the vasodilatory effects of CO both on cerebral arteries and arterioles are at least partially mediated by activation of calcium-gated-potassium channels (K_Ca+-channels) and the cGMP/protein kinase G-signalling pathway (Figure 3) (15,26,41,47,50 –52,56). Data from the above studies are summarised in Table 3.

The vascular effect of CO may also be NO dependent. Thus, nitric oxide synthase (NOS) inhibitors block CO-induced vasodilation in piglets (Table 3) (53,54). However, CO may also have vasoconstrictor effects due to interference with NO generation in the brain. Ishikawa et al. (58) showed that suppression of endogenous CO production by an HO-inhibitor induced dilatation of pial arterioles in rats and increased regional formation of NO, which was blocked by CO (58). This vasoconstrictor effect may be related to the duration of exposure since exposure to CO in doses of ≥2 × 10^–7M for two hours induced initial dilation but also delayed contraction of piglet pial arterioles. When NO concentration was held constant the initial dilation was sustained, suggesting that the vasoconstrictor effect may be caused by a long-term inhibition of NO (59).

Interactions with NO

NO is a signalling molecule importantly involved in primary headaches and especially in migraine. In contrast to CO, NO has been extensively studied in animal and human headache models (12). Chronic tension-type headache (60), migraine (12) and cluster headache (61) can all be induced experimentally by glyceryl trinitrate (GTN), a pro-drug for NO (62). Furthermore, NOS-inhibitors may be effective as migraine treatment (63). Through the NO-cGMP cascade NO affects nociceptive signalling and dilates cerebral arteries (12). Like NO, CO is an endogenously produced gaseous neurotransmitter in the peripheral and central nervous system. Both can be generated from the vascular wall and both bind and activate sGC thus elevating intracellular cGMP (Figure 3). CO and NO bind competitively to the haeme group of sGC but CO with much less affinity (11). Interestingly, there are areas in the brain with high expression of HO-2, sGC and cGMP without any NOS expression, indicating that it is CO that in these areas increases cGMP (10,11).

NO and CO have other shared functions as they both bind to haemoglobin and inhibit the oxygen transport system (11). They also inhibit mitochondrial oxidative phosphorylation by competitive binding to cytochrome c oxidase (CCO). NO can bind both oxidised and reduced CCO. CO binds only reduced CCO (64). The two gases also affect each other’s synthesis. NO increases CO production by upregulating HO-1 gene expression in smooth muscle cells and endothelial cells (65,66). NO is synthesised from L-arginine by NOS, which is a cytochrome P-450 type haemoprotein to which CO can also bind and affect its activity (Figure 3) (67). The latter varies between cells and diseases. Thus, CO increases inducible NOS (iNOS) in hepatitis (68) and, in pulmonary hypertension, endothelial NOS (eNOS) is increased by CO whereas iNOS stays normal (69). The effect of CO on NO synthesis also seems dose dependent. In renal resistance arteries low doses of CO increase NO release and high doses of CO inhibit eNOS activity and decrease NO release (70). Very high CO doses inhibit NOS isolated from the rat cerebellum (67). However, in rat brain NO production increased nine-fold after severe CO intoxication (20 minutes in 3000 parts per million (ppm)) (71). Another important mechanism may be the competitive use of the electron donor nicotinamide adenine dinucleotide phosphate (NADPH) and the co-substrate molecular oxygen in the formation of NO and CO (11).

Thus, NO and CO seem to display intricate autoregulatory mechanisms that can both enhance and decrease the effects of each other. The possible implication of these mechanisms in relation to headache pathophysiology is unclear.

Oxidative stress

Oxidative stress has been proposed to play a role in migraine pathophysiology (72). CCO inhibition by CO may be an important mechanism causing oxidative stress and cellular damage. CCO is an important enzyme in the mitochondrial oxidative metabolism and inhibition of CCO may lead to increased production of reactive oxygen species and eventually cellular damage by oxidation of essential enzymes, lipids and mitochondrial DNA (Figure 3) (73,74). In vitro and in vivo rat experiments have shown significant oxidative stress in the brain after exposure to CO in toxic doses, indicated by lipid peroxidation (75), increased hydrogen peroxide production, mitochondrial glutathione depletion and non-enzymatic hydroxylation of salicylate (17,41,76). It has been hypothesised that the acute and delayed neuronal injury in specific regions of the brain caused by CO intoxication is caused by this oxidative stress (17,41,76). However, these experimental effects were conducted with very high CO concentrations seen only in very severe intoxication (COHb 50%–70%) (17,43,74,76). However, one in vitro study showed that experimentally increased endogenous CO production in human embryonic kidney cells can inhibit cellular respiration and CCO activity moderately (16). Exposure to CO in low concentrations stimulates release of NO and production of the strong oxidant peroxynitrite from bovine pulmonary artery endothelial cells (77). This reactive oxygen species production at low-dose CO and its importance in headache need to be explored.

CO-induced headache

CO is colourless, odourless and non-irritating. Patients may therefore be exposed to CO without noticing. CO induces headache, nausea, weakness, palpitations, dizziness and fatigue (1 –3,78). The diagnosis may be overlooked in clinical practise (7,79) and COHb measurement may show normal values due to time delay from exposure to COHb measurement. Fortunately this is rare because the half-life of COHb is relatively long (240–360 minutes in normal air) (8). The symptoms caused by CO exposure are correlated to the COHb levels according to the diagnostic criteria of ICHD III beta (1). Thus, 10%–20% concentration of COHb is associated with mild headaches, 20%–30% COHb is associated with moderate pulsating headache and irritability, and 30–40% COHb is associated with severe headache accompanied by nausea, vomiting and blurred vision (1). Chief Surgeon Sayers and his two colleagues performed 16 self-exposures to CO in a gas chamber (80). Headache data from 50% of the self-exposures are questionable, as they were performed after strenuous exercise, high temperature, high humidity or with pre-existing headache due to sinus infection. The remaining experiments showed a clear trend toward 10%–20% COHb causing tightness across the forehead, which on exertion increased to headache, while 20%–25% COHb caused headache which increased on exertion (80).

A discrepancy between COHb and clinical symptoms has, however, been observed in patients with CO poisoning (4,7,78,81). A systematic prospective study of headache characteristics and COHb levels of 100 CO-poisoned patients (mean COHb of 21.3%) showed no correlation between COHb and peak intensity of pain (4). Interestingly, the authors reported that 41% of patients developed throbbing headache but whether these patients experienced migraine-like features such as unilateral location, aggravation by physical activity, phonophobia, photophobia, nausea or vomiting is unknown (4). It has previously been demonstrated that substances may provoke mild headache in healthy controls, but migraine in migraineurs (82). A similar study by Hampson and Dunn with data from 1323 CO-poisoned patients suggested that headache may not be a mandatory symptom of CO poisoning (81). However, the authors did not report whether patients had personal or family history of migraine, and it should also be noticed that none of the known migraine-provoking substances induce migraine in 100% of patients (82). Furthermore, CO-poisoned patients are investigated during decreasing COHb levels from an unknown peak concentration. Data from accidental CO exposure may also be misleading because of simultaneous exposure to other possible headache-inducing/-relieving compounds from motor vehicles, forklift, furnaces, etc.

Headache can appear at COHb levels below 10% (2,4,6,79). CO levels in non-smokers are 1%–3%, and in smokers 3%–8% but even levels up to 15% are seen in heavy smokers without headache or other symptoms of intoxication. Increased tolerability with chronic exposure has been suggested (83). Smokers do not seem to be more tolerant to severe acute exposure but they recover faster (78). Interestingly, smoking may be associated with an increased risk of migraine with aura but not migraine without aura (84). Whether this may be related to CO is unknown.

The severity of CO-induced symptoms may correlate better with duration of exposure than with COHb concentration (78). Stewart et al. (2) exposed 18 healthy subjects to CO at different concentrations and durations. In one experiment, a low CO concentration and long duration was tested. The authors reported no headache during inhalation of 25, 50 (end COHb 5.9%) and 100 (end COHb 12%) ppm CO for eight hours and no headache at 50 ppm for 24 hours (end COHb 7.9%). However, three subjects were exposed to 200 ppm CO (end COHb 16%) for four hours and developed a mild headache in the last hour. Two subjects were exposed to 500 ppm CO on three occasions for 1.8–2.3 hours (maximum COHb 26%) and repeatedly developed mild frontal headache after approximately one hour, when COHb was about 15%. The headaches were aggravated by minimal physical activity. The headaches remained mild during the first post-exposure hour. Then they intensified into severe occipitofrontal headaches accompanied by nausea and reached an intensity peak 3.5 hours after exposure. Treatment with hyperbaric oxygen immediately following the exposure ameliorated the headaches (2). Finally, two subjects were exposed to rising CO concentration over two hours, until 1000 ppm was reached and sustained for 30 minutes. Both subjects had mild frontal headache after two hours and developed severe post-exposure headache lasting up to 12 hours (2). None of the headaches could be ameliorated by acetylsalicylic acid (2), which is effective in migraine and episodic tension-type headache (85). Thus, cyclooxygenase 1 and 2 activity (which is inhibited by acetylsalicylic acid) does not seem to be involved in CO-induced headache. Stewart et al. (3) investigated experimental exposure to high concentrations of CO (range: 1000 ppm for 10 minutes to 35.600 ppm for 45 seconds) in six healthy subjects. Two subjects reported mild frontal headache after a few minutes of exposure to 10,000 and 30,000 ppm CO, reaching a COHb of 9.1% and 11.6%, respectively. The headaches disappeared after 20 minutes of 100% oxygen. The long-term effect of CO on headache is not well examined, but chronic post-intoxication headache has been reported among 30 patients one to 24 months after CO intoxication (86). A number of studies using CO exposure have been conducted without serious adverse events (87 –89). However, more human studies are needed to explore the importance of CO in headache pathophysiology.

Conclusion and future perspectives

CO is not only a toxic pollutant but also an important endogenously produced volatile signalling molecule involved in many biological mechanisms. The headache-inducing mechanisms of CO are not understood. Hypoxia may be one but CO also has many similarities and interacts with NO, which is an important molecule in primary headaches. Like NO, CO can activate the cGMP pathway, cause vasodilation and possibly induce oxidative stress. Whether CO normally plays a role in the mechanisms of migraine, tension headache and cluster headache is unknown. However, growing evidence implicates CO in nociceptive processing in the peripheral and central nervous system.

Article highlights

Carbon monoxide (CO) is endogenously produced by the enzyme haeme oxygenase (HO).

HO is present in important nociceptive structures of the nervous system.

CO-induced headache shares many characteristics with migraine and other headaches.

The mechanisms whereby CO causes headache may include hypoxia, nitric oxide signalling, activation of cyclic guanosine monophosphate pathways, cerebral vasodilation and production of free radicals.

CO may play an important role in the mechanisms of migraine and other headaches.

Abbreviations

CO: carbon monoxide; ICHD III beta: International Classification of Headache Disorders; cGMP: cyclic guanosine monophosphate; NO: nitric oxide; COHb: carboxyhaemoglobin; HO: haeme oxygenase; CGRP: calcitonin gene-related peptide; BBB: blood-brain barrier; sGC: soluble guanylate cyclase; GTP: guanosine triphosphate; CBF: cerebral blood flow; ADP: adenosine diphosphate; PGE₂: prostaglandin E2; AA: arachidonic acid; K_Ca+-channels: calcium-gated-potassium channels; NOS: nitric oxide synthase; GTN: glyceryl trinitrate; CCO: cytochrome c oxidase; iNOS: inducible NOS; eNOS: endothelial NOS; ppm: parts per million; NADPH: nicotinamide adenine dinucleotide phosphate; NMDA: N-methyl-D-aspartate; AMPA: alpha-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid; NADP+: oxidised nicotinamide adenine dinucleotide phosphate; neuronal PAS2: neuronal PAS domain-containing protein 2; CNS: central nervous system; PNS: peripheral nervous system.

Authors’ contributions

JO and MA conceived and designed (including search strategies) the review. NA and MKH conducted the literature search and wrote the first and subsequent drafts of the manuscript. HWS contributed to analysis and interpretation of reviewed data. HWS, MA and JO participated in critical revision and writing of the article. All authors have seen and approved the final version.

Footnotes

Acknowledgements

The authors thank Dr Jakob Hakon Christensen for illustrative work with the figures.

Funding

This work was supported by the Capital Region of Denmark, the Foundation for Health Research (grant number A4620); the Novo Nordisk Foundation (grant number A14333); and the Danish Headache Foundation.

Conflict of interest

None declared.

References

Headache Classification Committee of the International Headache Society (IHS). The International Classification of Headache Disorders, 3rd edition (beta version). Cephalalgia 2013; 33: 629–808.

Stewart

Peterson

Baretta

. Experimental human exposure to carbon monoxide. Arch Env Heal 1970; 21: 154–164.

Stewart

Fisher

Baretta

. Experimental human exposure to high concentrations of carbon monoxide. Arch Env Heal 1973; 26: 1–7.

Hampson

. Characteristics of headache associated with acute carbon monoxide poisoning. Headache 2002; 42: 220–223.

Haldane

. Medicolegal contributions of historical interest. The action of carbonic oxide on man. Forensic Sci 1972; 1: 451–483.

Stewart

. The effect of carbon monoxide on humans. J Occup Med 1975; 18: 304–309.

Heckerling

. Occult carbon monoxide poisoning: A cause of winter headache. Am J Emerg Med 1987; 5: 201–204.

Piantadosi

. Biological chemistry of carbon monoxide. Antioxid Redox Signal 2002; 4: 259–270.

Uddman

Tajti

Sundler

. The presence of heme-oxygenase and biliverdin reductase in human cranial ganglia indicates a role for carbon monoxide in neural transmission. Neuro Endocrinol Lett 2004; 25: 423–428.

10.

Verma

Hirsch

Glatt

. Carbon monoxide: A putative neural messenger. Science 1993; 259: 381–384.

11.

Maines

. The heme oxygenase system: A regulator of second messenger gases. Annu Rev Pharmacol Toxicol 1997; 37: 517–554.

12.

Olesen

. The role of nitric oxide (NO) in migraine, tension-type headache and cluster headache. Pharmacol Ther 2008; 120: 157–171.

13.

Leffler

Parfenova

Fedinec

. Contributions of astrocytes and CO to pial arteriolar dilation to glutamate in newborn pigs. Am J Physiol Heart Circ Physiol 2006; 291: H2897–H2904.

14.

Kanu

Leffler

. Arachidonic acid- and prostaglandin E2-induced cerebral vasodilation is mediated by carbon monoxide, independent of reactive oxygen species in piglets. Am J Physiol Heart Circ Physiol 2011; 301: H2482–H2487.

15.

Umstot

. Astrocyte-derived CO is a diffusible messenger that mediates glutamate-induced cerebral arteriolar dilation by activating smooth muscle Cell KCa channels. Cir Res 2008; 102: 234–241.

16.

D’Amico

Lam

Hagen

. Inhibition of cellular respiration by endogenously produced carbon monoxide. J Cell Sci 2006; 119: 2291–2298.

17.

Piantadosi

Zhang

Demchenko

. Production of hydroxyl radical in the hippocampus after CO hypoxia or hypoxic hypoxia in the rat. Free Radic Biol Med 1997; 22: 725–732.

18.

Rodkey

O’Neal

Collison

. Relative affinity of hemoglobin S and hemoglobin A for carbon monoxide and oxygen. Clin Chem 1974; 20: 83–84.

19.

Roughton

FJW

Darling

. The effect of carbon monoxide on the oxyhemoglobin dissociation curve. Am J Physiol 1944; 141: 17–31.

20.

Okada

Tyuma

. Effect of carbon monoxide on equilibrium between oxygen and hemoglobin. Am J Physiol 1976; 230: 471–475.

21.

Goldbaum

. Mechanism of the toxic action of carbon monoxide. Ann Clin Lab Sci 1976; 6: 372–376.

22.

Linderholm

. On the significance of CO tension in pulmonary capillary blood for determination of pulmonary diffusing capacity with the steady state CO method. Acta Med Scand 1957; 156: 413–427.

23.

Magnusson

Ekström

Elmér

. Heme oxygenase-1, heme oxygenase-2 and biliverdin reductase in peripheral ganglia from rat, expression and plasticity. Neuroscience 2000; 95: 821–829.

24.

Scapagnini

D’Agata

Calabrese

. Gene expression profiles of heme oxygenase isoforms in the rat brain. Brain Res 2002; 954: 51–59.

25.

Ewing

Maines

. In situ hybridization and immunohistochemical localization of heme oxygenase-2 mRNA and protein in normal rat brain: Differential distribution of isozyme 1 and 2. Mol Cell Neurosci 1992; 3: 559–570.

26.

Leffler

Nasjletti

. Carbon monoxide and cerebral microvascular tone in newborn pigs. Am J Physiol 1999; 276: H1641–H1646.

27.

Vollerthun

Höhler

Kummer

. Heme oxygenase-2 in primary afferent neurons of the guinea-pig. Histochem Cell Biol 1996; 105: 453–458.

28.

Olesen

Burstein

Ashina

. Origin of pain in migraine: Evidence for peripheral sensitisation. Lancet Neurol 2009; 8: 679–690.

29.

Lassen

Haderslev

Jacobsen

. CGRP may play a causative role in migraine. Cephalalgia 2002; 22: 54–61.

30.

Gordh

Sharma

Azizi

. Spinal nerve lesion induces upregulation of constitutive isoform of heme oxygenase in the spinal cord. An immunohistochemical investigation in the rat. Amino Acids 2000; 19: 373–381.

31.

Clark

. The role of heme oxygenase in neuropathic and incisional pain. Anesth Analg 2000; 90: 677–682.

32.

Clark

. Heme oxygenase type 2 plays a role in formalin-induced nociception. Pain 2000; 86: 75–80.

33.

Clark

. Spinal cord heme oxygenase participates in glutamate-induced pain-related behaviors. Eur J Pharmacol 2002; 450: 43–48.

34.

Clark

. Heme oxygenase type 2 participates in the development of chronic inflammatory and neuropathic pain. J Pain 2003; 4: 101–107.

35.

Yachie

Niida

Wada

. Oxidative stress causes enhanced endothelial cell injury in human heme oxygenase-1 deficiency. J Clin Invest 1999; 103: 129–135.

36.

Clark

. Heme oxygenase inhibitors reduce formalin-induced Fos expression in mouse spinal cord tissue. Neuroscience 2001; 105: 949–956.

37.

Lighthall

Liang

. Alterations in spinal cord gene expression after hindpaw formalin injection. J Neurosci Res 2004; 78: 533–541.

38.

Steiner

Branco

Cunha

. Role of the haeme oxygenase/carbon monoxide pathway in mechanical nociceptor hypersensitivity. Br J Pharmacol 2001; 132: 1673–1682.

39.

Nascimento

Branco

. Antinociception synergy between the peripheral and spinal sites of the heme oxygenase-carbon monoxide pathway. Braz J Med Biol Res 2009; 42: 141–147.

40.

Hervera

Leánez

Negrete

. Carbon monoxide reduces neuropathic pain and spinal microglial activation by inhibiting nitric oxide synthesis in mice. PLoS One 2012; 7: e43693–e43693.

41.

Morita

Perrella

Lee

. Smooth muscle cell-derived carbon monoxide is a regulator of vascular cGMP. Proc Natl Acad Sci U S A 1995; 92: 1475–1479.

42.

Clark

. Spinal cord nitric oxide synthase and heme oxygenase limit morphine induced analgesia. Brain Res Mol Brain Res 2001; 95: 96–102.

43.

Piantadosi

Zhang

Levin

. Apoptosis and delayed neuronal damage after carbon monoxide poisoning in the rat. Exp Neurol 1997; 147: 103–114.

44.

Shinomura

Nakao

Mori

. Reduction of depolarization-induced glutamate release by heme oxygenase inhibitor: Possible role of carbon monoxide in synaptic transmission. Neurosci Lett 1994; 166: 131–134.

45.

Paulson

Parving

Olesen

. Influence of carbon monoxide and of hemodilution on cerebral blood flow and blood gases in man. J Appl Physiol 1973; 35: 111–116.

46.

Benignus

Petrovick

Newlin-Clapp

. Carboxyhemoglobin and brain blood flow in humans. Neurotoxicol Teratol 1992; 14: 285–290.

47.

Komuro

Borsody

Ono

. The vasorelaxation of cerebral arteries by carbon monoxide. Exp Biol Med 2001; 226: 860–865.

48.

Andresen

Shafi

Durante

. Effects of carbon monoxide and heme oxygenase inhibitors in cerebral vessels of rats and mice. Am J Physiol Heart Circ Physiol 2006; 291: H223–H230.

49.

Amin

Asghar

Hougaard

. 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.

50.

Achouh

Simonet

Fabiani

. Carbon monoxide induces relaxation of human internal thoracic and radial arterial grafts. Interact Cardiovasc Thorac Surg 2008; 7: 959–962.

51.

Wang

. Carbon monoxide-induced vasorelaxation and the underlying mechanisms. Br J Pharmacol 1997; 121: 927–934.

52.

Leffler

Fedinec

Parfenova

. Permissive contributions of NO and prostacyclin in CO-induced cerebrovascular dilation in piglets. Am J Physiol Heart Circ Physiol 2005; 289: H432–H438.

53.

Leffler

Nasjletti

Johnson

. Contributions of prostacyclin and nitric oxide to carbon monoxide-induced cerebrovascular dilation in piglets. Am J Physiol Heart Circ Physiol 2001; 280: H1490–H1495.

54.

Barkoudah

Jaggar

Leffler

. The permissive role of endothelial NO in CO-induced cerebrovascular dilation. Am J Physiol Heart Circ Physiol 2004; 287: H1459–H1465.

55.

Kanu

Leffler

. Roles of glia limitans astrocytes and carbon monoxide in adenosine diphosphate-induced pial arteriolar dilation in newborn pigs. Stroke 2009; 40: 930–935.

56.

Tcheranova

Basuroy

. Glutamate-induced calcium signals stimulate CO production in piglet astrocytes. Am J Physiol Heart Circ Physiol 2011; 301: H428–H433.

57.

Milligan

Watkins

. Pathological and protective roles of glia in chronic pain. Nat Rev Neurosci 2009; 10: 23–36.

58.

Ishikawa

Kajimura

Adachi

. Carbon monoxide from heme oxygenase-2 is a tonic regulator against NO-dependent vasodilatation in the adult rat cerebral microcirculation. Circ Res 2005; 97: e104–e114.

59.

Knecht

Milam

Wilkinson

. Time-dependent action of carbon monoxide on the newborn cerebrovascular circulation. Am J Physiol Heart Circ Physiol 2010; 299: H70–H75.

60.

Ashina

Bendtsen

Jensen

. Nitric oxide-induced headache in patients with chronic tension-type headache. Brain 2000; 123: 1830–1837.

61.

Ekbom

. Nitrolglycerin as a provocative agent in cluster headache. Arch Neurol 1968; 19: 487–493.

62.

Iversen

Olesen

Tfelt-Hansen

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

63.

Lassen

Ashina

Christiansen

. Nitric oxide synthase inhibition in migraine. Lancet 1997; 349: 401–402.

64.

Cooper

Brown

. The inhibition of mitochondrial cytochrome oxidase by the gases carbon monoxide, nitric oxide, hydrogen cyanide and hydrogen sulfide: Chemical mechanism and physiological significance. J Bioenerg Biomembr 2008; 40: 533–539.

65.

Durante

Kroll

Christodoulides

. Nitric oxide induces heme oxygenase-1 gene expression and carbon monoxide production in vascular smooth muscle cells. Circ Res 1997; 80: 557–564.

66.

Polte

Abate

Dennery

. Heme oxygenase-1 is a cGMP-inducible endothelial protein and mediates the cytoprotective action of nitric oxide. Arterioscler Thromb Vasc Biol 2000; 20: 1209–1215.

67.

White

Marletta

. Nitric oxide synthase is a cytochrome P-450 type hemoprotein. Biochemistry 1992; 31: 6627–6631.

68.

Zuckerbraun

Billiar

Otterbein

. Carbon monoxide protects against liver failure through nitric oxide-induced heme oxygenase 1. J Exp Med 2003; 198: 1707–1716.

69.

Zuckerbraun

Chin

Wegiel

. Carbon monoxide reverses established pulmonary hypertension. J Exp Med 2006; 203: 2109–2119.

70.

Thorup

Jones

Gross

. Carbon monoxide induces vasodilation and nitric oxide release but suppresses endothelial NOS. Am J Physiol 1999; 277: F882–F889.

71.

Ischiropoulos

Beers

Ohnishi

. Nitric oxide production and perivascular nitration in brain after carbon monoxide poisoning in the rat. J Clin Invest 1996; 97: 2260–2267.

72.

Bernecker

Ragginer

Fauler

. Oxidative stress is associated with migraine and migraine-related metabolic risk in females. Eur J Neurol 2011; 18: 1233–1239.

73.

Brown

Piantadosi

. In vivo binding of carbon monoxide to cytochrome c oxidase in rat brain. J Appl Physiol (1985) 1990; 68: 604–610.

74.

Brown

Piantadosi

. Recovery of energy metabolism in rat brain after carbon monoxide hypoxia. J Clin Invest 1992; 89: 666–672.

75.

Thom

. Dehydrogenase conversion to oxidase and lipid peroxidation in brain after carbon monoxide poisoning. J Appl Physiol (1985) 1992; 73: 1584–1589.

76.

Zhang

Piantadosi

. Mitochondrial oxidative stress after carbon monoxide hypoxia in the rat brain. J Clin Invest 1992; 90: 1193–1199.

77.

Thom

Ischiropoulos

. Vascular endothelial cells generate peroxynitrite in response to carbon monoxide exposure. Chem Res Toxicol 1997; 10: 1023–1031.

78.

Burney

Nemiroff

. Mass carbon monoxide poisoning: Clinical effects and results of treatment in 184 victims. Ann Emerg Med 1982; 11: 394–399.

79.

Heckerling

Leikin

Maturen

. Predictors of occult carbon monoxide poisoning in patients with headache and dizziness. Ann Intern Med 1987; 107: 174–176.

80.

Sayers

Meriwether

Yant

. Physiological effects of exposure to low concentrations of carbon monoxide. Public Health Rep 1922; 19: 1127–1142.

81.

Hampson

Dunn

. Symptoms of carbon monoxide poisoning do not correlate with the initial carboxyhemoglobin level. Undersea Hyperb Med 2012; 39: 657–665.

82.

Schytz

Schoonman

Ashina

. What have we learnt from triggering migraine? Curr Opin Neurol 2010; 23: 259–265.

83.

Raub

Benignus

. Carbon monoxide and the nervous system. Neurosci Biobehav Rev 2002; 26: 925–940.

84.

Salhofer-Polanyi

Frantal

Brannath

. Prospective analysis of factors related to migraine aura – the PAMINA study. Headache 2012; 52: 1236–1245.

85.

Lampl

Voelker

Steiner

. Aspirin is first-line treatment for migraine and episodic tension-type headache regardless of headache intensity. Headache 2012; 52: 48–56.

86.

Ely

Moorehead

Haponik

. Warehouse workers’ headache: Emergency evaluation and management of 30 patients with carbon monoxide poisoning. Am J Med 1995; 98: 145–155.

87.

González-Alonso

. Erythrocyte and the regulation of human skeletal muscle blood flow and oxygen delivery: Role of circulating ATP. Circ Res 2002; 91: 1046–1055.

88.

González-Alonso

Richardson

Saltin

. Exercising skeletal muscle blood flow in humans responds to reduction in arterial oxyhaemoglobin, but not to altered free oxygen. J Physiol 2001; 530: 331–341.

89.

Hanada

Sander

González-Alonso

. Human skeletal muscle sympathetic nerve activity, heart rate and limb haemodynamics with reduced blood oxygenation and exercise. J Physiol 2003; 551: 635–647.

Carbon monoxide may be an important molecule in migraine and other headaches

Abstract

Introduction

Discussion

Conclusion

Keywords

Introduction

Biological chemistry and body stores

CO as a pain-modulating neurotransmitter

Pain- and headache-inducing mechanisms of CO

CO as a cerebral vasodilator

Interactions with NO

Oxidative stress

CO-induced headache

Conclusion and future perspectives

Article highlights

Abbreviations

Authors’ contributions

Footnotes

Acknowledgements

Funding

Conflict of interest

References