Abstract
Octreotide is a long-acting somatostatin analogue that has been effectively used to treat migraine. Octreotide poorly penetrates the blood–brain barrier, but has potential central target sites in the trigeminal nucleus caudalis, which is the primary central relay station for trigeminal nociceptive information in the brain. We studied the effect of intracisternally applied octreotide in a model of trigeminovascular stimulation in the unrestrained rat using intracisternal capsaicin infusion to stimulate intracranial trigeminal nerves. Fos expression in the outer layers of the trigeminal nucleus caudalis (TNC I-II) and behavioural analysis were used to measure the effects of octreotide on capsaicin-induced trigeminovascular activation. Increases of head grooming and scratching behaviour are an indication of octreotide-induced trigeminal activation. However, octreotide did not alter the average capsaicin-induced Fos expression in the TNC I-II and capsaicin sensitive behaviours were not modified by octreotide pre-treatment. This argues against a role for central (TNC I-II) somatostatin receptors in the processing of the nociceptive trigeminovascular signals.
Somatostatin (somatotropin release-inhibiting factor, SRIF)) is a hypothalamic pituitary regulatory hormone, which is present in brain, spinal cord (1), gut and pancreas (2). It suppresses growth hormone release and inhibits the release of regulatory peptides of the gastroenteropancreatic endocrine system (2, 3). Octreotide (SMS 201–995) is a long-acting somatostatin analogue, which, due to its inhibitory actions on growth hormone and gastric peptide release, is used in patients suffering from acromegaly, cancer, gastrointestinal diseases and pancreatitis (4). Until now, five different somatostatin receptor genes have been cloned (sst1 to sst5) that fit within previously defined SRIF1 and SRIF2 groups. The SRIF1 group consists of the sst2, sst3 and sst5 receptor with relative high (sst2, sst5) to moderate (sst3) affinity for octreotide and a marked structural similarity. The SRIF2 group consists of the sst1 and sst4 receptor with low affinity for octreotide and high mutual structural similarities (5). The sst2 receptor has been identified in two spliced isoforms, the sst2(a) and sst2(b) receptor that have similar binding properties but may differ in G-protein coupling (6, 7).
Somatostatin and octreotide have effectively been used to relief cluster headache (8) and migraine (9), respectively. The target of action of octreotide for migraine relief is not known but is most likely peripheral as it can poorly penetrate the brain (10, 11). Its mode of action may involve inhibition of release of several vasoactive substances since somatostatin can inhibit trigeminal substance P release (12), a neuropeptide that has been implicated in the pathophysiology of migraine (13–17). Antidromic release of substance P from sensory nerve endings, and the vasodilatation it triggers, is inhibited by somatostatin administration (18). Octreotide has, like many other antimigraine drugs (19–22), effectively been used to antagonize neurogenic plasma protein extravasation in the dura mater after trigeminal ganglion stimulation and intravenous capsaicin administration (23). As octreotide could not reduce substance P-induced plasma protein extravasation, the mode of action of octreotide most likely involves receptors located at the prejunctional trigeminal sensory nerve endings that innervate the dura mater (23).
An extensive complex of somatostatin containing neurones and fibres is present in the rat trigeminal nucleus caudalis, layer I-II (TNC I-II), the primary relay nucleus for trigeminal nociceptive signals (24–26). Somatostatin positive fibres in layer II originate predominantly from primary trigeminal afferents, whereas somatostatin immunoreactivity in layer I most likely originates from interneurones in layer I and II of the TNC (24, 25). Receptor binding sites for somatostatin, based on autoradiography with the (Tyr3) derivative of octreotide, (125I)204–090, accordingly have been reported in the substantia gelatinosa of the trigeminal nuclear complex of rat (27) and human (28). Moderate densities of sst3 receptor mRNA (29, 30) and sst2(b) receptor mRNA (31) have been shown in the spinal trigeminal nucleus. This provides a potential central target for octreotide to modulate orthodromic trigeminal nociception. Centrally acting octreotide-like drugs may show improved analgesic efficacy in migraine due to inhibition of both antidromic and orthodromic trigeminovascular pain processing.
The aim of this study was to evaluate the effects of octreotide in a conscious rat model of intracranial trigeminovascular nociception. Because octreotide can poorly penetrate the blood–brain barrier and the octapeptide analogues of somatostatin (32), somatostatin, and the selective agonists DC 32–87 (sst2) and BIM-23056 (sst3) (33, 34), showed modulation of physiological effects mediated by regions in the brainstem after intracisternal infusion, we choose this route of drug administration.
The larger blood vessels of the brain and the meninges are innervated by sensory nerves of the trigeminal system, together forming the trigeminovascular system. Several animal models of trigeminovascular activation have shown analgesic effects of anti-migraine drugs (19–23, 35–37), including the newly developed triptans with a central site of action (38, 39). A previously described animal model of intracranial trigeminovascular stimulation, based on the intracisternal infusion of the irritant capsaicin in conscious rats, was used since it enables the analysis of behaviour combined with assessment of Fos immunoreactivity (Fos-ir) in the trigeminal nucleus caudalis (40).
Methods
The experiments were performed according to the ethical guidelines for investigations of experimental pain in conscious animals (41) and were approved by the local committee of animal bio-ethics, University Groningen (FDC 2198). Male Wistar rats weighing 250–325 g were used. They were individually housed in a light/dark cycle with lights on from 08.00 till 20.00 h. Food and water were provided ad libitum.
Cisterna magna (CM) cannulation
The cannula was made from a stainless steel needle (0.6 × 25 mm 23 G × 1′′; Braun, Melsungen, Germany) of which 6.5 mm was inserted into the brain. Surgery was performed under semi-sterile conditions. Sodium pentobarbital was used as anaesthetic (60 mg/kg, i.p.). Rats were placed in a stereotaxic apparatus with incisor bar at −7 mm from the horizontal plane. Two holes were drilled into the caudal corners of the interparietal skull and two screws (diameter 1.0 mm) were driven 1.5 mm into the skull. A hole of 1.2 mm was drilled at the mid-line of the external occipital crest through which the CM cannula was carefully placed behind the cerebellum, into the CM. Correct placement was confirmed by withdrawal of cerebrospinal fluid. The cannula was fixed to the skull with dental cement (Kemdemt, Purton, Swindon, UK) and closed by insertion of a metal wire (of the same length as the cannula) within a polyethylene cap to seal the cannula off.
Drugs
Capsaicin (3.05 g) was dissolved in 1 ml saline-ethanol-Tween 80 (8:1:1)
(vehicle-stock) and sonicated for 5 min (capsaicin-stock). The capsaicin-stock
and vehicle-stock solutions were further diluted 1:40 to yield the capsaicin 250
n
Experimental procedures
Octreotide or saline was injected intracisternally through the cannula in a volume of 10 μl, 10 min prior to infusion of capsaicin or vehicle. During the injection we attached a silica tube with internal diameter of 75 μm to the microinjector to reduce the internal diameter of the CM cannula.
For the infusion of 100 μl capsaicin or vehicle, the rats were placed in an observation cage (30 × 30 × 30 cm). This amount was infused in 2 min by a microinjector pump (CMA100, Carnegie Medicin, Stockholm, Sweden). The behaviour of the rats was recorded on videotape from 5 min before to 10 min after the infusion of capsaicin or vehicle.
Perfusion and immunohistochemistry
Rats were perfused 2 h after the infusion of capsaicin or vehicle. Prior to the
transcardial perfusion rats were deeply anaesthetized with sodium pentobarbital
(120 mg/kg i.p.) and perfused with 0.9% saline for 1 min, followed by 4%
paraformaldehyde (PF) in 0.1
Free floating sections were immunocytochemically stained for the proto-oncogene protein c-fos according to a standard protocol previously described (40, 42). In brief, after pre-incubation in normal sera and pre-treatment with 0.3% H2O2, the primary antibody, rabbit anti-Fos (1:10000; AB5, Oncogene Science Inc, Cambridge, UK) was applied overnight at room temperature. After thorough washing, biotinylated goat-α-rabbit IgG (1:200, Pierce, Rockford) was applied for 2 h. Subsequently, the sections were washed and then incubated with the avidine-biotine-peroxidase complex (Vector Labs, Burlingame) for 2 h at room temperature. The Ni-enhanced 3,3′-diaminobenzidine tetrahydrochloride reaction was used to visualize the presence of peroxidase. Intermittent washing was performed with KPBS, antibodies were dissolved in KPBS with 0.5% triton X-100. All staining procedures were performed with gentle agitation. Sections were mounted, dehydrated and cover-slipped with DEPEX.
Quantification
Fos expression
Fos-ir cells in layers I and II of the TNC (TNC I-II) were counted at −1, −2, −3, −4, −5 and −6 mm caudal from obex by an observer blinded from experimental procedures. Sections from −0.5 to −1.5 mm were averaged to obtain the count for the −1 mm level and so on. The mean of the total TNC I-II was calculated by averaging the Fos expression at the six levels.
Behaviour
The behaviour shown before, during and after the infusion was analysed with Observer ® software (Noldus Information Technology, Wageningen, The Netherlands). The first 2 min immediately after infusion were not analysed because rats were uncoupled from the microinjector device in that period. Behaviours scored were exploring behaviour (exploring and rearing), head grooming and head scratching and immobilization (all forms other than resting). Other behaviours shown include body grooming, eating and drinking, and resting.
Statistical analysis
The following groups were assembled: saline + vehicle (control,
n = 3), octreotide + vehicle
(Oct, n = 4), saline + capsaicin
(Caps, n = 5) and octreotide +
capsaicin (Oct + Caps, n = 7).
Effects of octreotide independent of capsaicin were tested by comparing the
Oct and the control group. Effects of octreotide on capsaicin-induced
numbers of Fos positive cells in the TNC I-II and behaviour were tested by
comparing the Oct + Caps group to the Caps group. The
Student's t-test was used to test significant
differences if data showed normal distribution, in all other cases the
Mann–Whitney rank sum test was employed. P
< 0.05 was considered significant. Fos data are expressed as mean
number of positive cells ±
Results
Inclusion criteria
As previously described (40), only rats with successful infusions exhibiting dural EB staining patterns ventral from the hindbrain, around the brainstem and the upper levels of the cervical spinal cord were included. Twenty-five rats were used, of which six were excluded because of aberrant EB staining (n = 2) or obstruction of the intracisternal cannula.
Behaviour
Prior to noxious trigeminovascular stimulation (Fig. 1a)
Effects of intracisternal octreotide administration on the time spent
on exploring, immobilization, and head grooming/scratching behaviour
(mean ±
Behavioural data obtained from either octreotide or saline treated rats in the 5 min prior to the infusion of capsaicin or vehicle were pooled regardless of the subsequent type of infusion. Octreotide caused a significant decrease of exploring behaviour (saline: 210 ± 12, octreotide: 141 ± 20) and a significant increase of head grooming/scratching (saline: 34 ± 9, octreotide: 117 ± 22). Changes of immobilization behaviour (saline: 20 ± 8, octreotide: 10 ± 5) or body grooming/scratching (saline: 13 ± 5, octreotide: 8 ± 3) were not observed.
During noxious trigeminovascular stimulation (Fig. 1b)
Administration of octreotide into the cisterna magna did not cause changes of exploring, head grooming and head scratching behaviour during infusion of vehicle (exploring: control: 106 ± 3, Oct: 105 ± 6; head grooming/scratching: control: 10 ± 2, Oct: 8 ± 7). Moreover, it could not prevent the capsaicin-induced reduction of exploring behaviour (Caps: 27 ± 7, Oct + Caps: 32 ± 4) nor the induction of head grooming and head scratching (Caps: 70 ± 9, Oct + Caps: 65 ± 8). Neither capsaicin, octreotide, nor the combination of these two compounds, initiated alterations of immobilization behaviour.
After noxious trigeminovascular stimulation (Fig. 1c)
Octreotide administration followed by a vehicle infusion caused a significant reduction of exploring behaviour (control: 333 ± 14, Oct: 194 ± 51) but head grooming, head scratching (control: 62 ± 24, Oct: 132 ± 71) and immobilization (control: 47 ± 21, Oct: 121 ± 63) were not significantly altered by this treatment. Octreotide pre-treated animals were not significantly different from saline pre-treated animals in their reaction to capsaicin. Time spent on immobilization (Caps: 395 ± 25, Oct + Caps: 313 ± 36), and grooming and scratching of the head (Caps: 79 ± 24, Oct + Caps: 161 ± 35) was not altered by the treatment. Moreover, there was a trend of increased grooming and scratching of the head in the octreotide pre-treated rats.
Fos expression (Fig. 2)
Number of Fos positive cells (mean ±
Intracisternally applied octreotide alone does not significantly alter Fos expression in the TNC I-II, although it was increased at all levels of the TNC (average: control: 72 ± 7, Oct: 145 ± 42). Capsaicin induced a marked increase of Fos expression at every rostro-caudal level of the TNC I-II (average: Caps: 643 ± 50). Octreotide did not significantly alter the average Fos expression in the TNC I-II (Oct + Caps: 610 ± 15) but caused a small, significant decrease of the number of capsaicin-induced Fos positive cells at 6 mm caudal from obex (Caps: 800 ± 44, Oct + Caps: 702 ± 23).
Discussion
Octreotide infusions into the cisterna magna reduced the exploring behaviour and increased the head grooming/scratching behaviour. Intracisternal octreotide pre-treatment did not affect the pain behaviour after intracisternal administration of capsaicin and accordingly did not reduce the average number of Fos positive cells in the TNC I-II. A small but significant reduction of capsaicin-induced Fos expression by octreotide was found in the caudal-most part of the TNC I-II.
That the intracisternal octreotide administration affected the trigeminal system intracranially can be concluded from the head grooming/scratching behaviour that was observed before the infusion of capsaicin. In control animals, the proportion of head grooming to head scratching is 6.5. In octreotide-treated animals, both behaviours are increased but the proportion of head grooming to head scratching is 1.5, showing that especially the head scratching is increased in octreotide-treated animals. In former experiments (40), we especially observed head scratching at higher concentrations of capsaicin, indicating that octreotide in these experiments may act as an irritant. This is confirmed by reports that describe pain at the (subcutaneous) injection site of octreotide in humans (4).
There was no induction of body grooming or scratching in the rats (close to zero in all groups) so the effect seems to be restricted to the trigeminal system, rather than being a centrally mediated general behavioural response. This, however, is not confirmed by the Fos expression in the outer layers of the TNC, the primary relay station for trigeminal nociceptive afferents (15). Although the Fos expression at every level of the TNC I-II of octreotide-treated animals is higher compared to control animals, the difference did not reach statistical significance. That the changes in behaviour caused by octreotide are not well reflected in TNC I-II Fos expression may be caused by the difference in temporal resolution of both parameters. Behaviour is measured acutely, whereas Fos expression is an accumulation of cell-activating events in the hours preceding perfusion. Alternatively, as was shown by Bereiter (43), some corneal stimulation responsive neurones in the TNC do not express Fos although activation of the cells was detected with electrophysiological methods.
Despite the above-mentioned actions of octreotide in the trigeminovascular system, intracisternal octreotide was not effective in reducing the average capsaicin-induced expression of Fos in the TNC I-II. Furthermore, the behaviours sensitive for capsaicin treatment were not affected by octreotide pre-treatment, indicating that secondary or higher order trigeminal processing is not modified by intracisternal octreotide administration. These results are confirmed by Bereiter (43) who observed that the Fos expression in the largest part of the TNC after stimulation of the cornea was not modified by i.c.v octreotide pre-treatment. A reduction of Fos expression by octreotide was only seen in the caudal-most part of the TNC after trigeminal corneal stimulation, which suggested a differential sensitivity of trigeminal neurones for octreotide. Moreover, Bereiter (43) noticed that the threshold dose of intracerebroventricularly administered morphine necessary to reduce the Fos expression was only 0.045 nmol whereas for octreotide it was 1 nmol. In our investigation a small but significant reduction of the Fos expression was found in the caudal-most part of the TNC with an octreotide concentration of 5 nmol. This is well above the threshold dose for octreotide and sufficient for a two-fold but non-significant general increase of the TNC I-II Fos expression compared to controls.
Although we have found a limited effect of intracisternal octreotide administration in the present study that could not be relevant for treatment of trigeminovascular headaches like migraine, migraineurs do benefit from a subcutaneous octreotide treatment (9). Most likely this effect is caused by inhibition of the plasma protein extravasation (PPE) in the dura mater, as was shown after intravenous octreotide administration in guinea pigs or rats (23). The orthodromic conduction of a nociceptive signal in the TNC I-II was not modified by octreotide pre-treatment in the present study, although somatostatin-positive cells and fibres, and somatostatin receptors, have been identified in the TNC (24, 25, 27, 29–31).
A large number of the layer II cells of the TNC show somatostatin immunoreactivity (24–26) and the sst2 receptor is abundant in the TNC I-II (27, 28, 47). Moreover, a small percentage of the trigeminal afferents employ somatostatin as a neuropeptide cotransmitter (24, 25). Trigeminovascular stimulation induces a glutamate release from the primary afferents (45, 46) that is mediated both by NMDA and by AMPA type glutamate receptors in the TNC I-II. Somatostatin modulates glutamate-mediated nociceptive signals (48, 49) and as such, somatostatin analogue's could be an alternative for glutamate antagonists like MK801 (44) that may have significant adverse effects.
Octreotide, a potent long-acting somatostatin agonist, has been shown to inhibit the cellular response to glutamate by interacting with the sst2 type receptor (49). However, recently it has been demonstrated that brain areas with a high density of somatostatin-positive fibres have a low level of membrane associated sst2(a) receptors (47). Here, the sst2(a) receptors are internalized and not recognized by exogenous somatostatin or analogues. Accordingly, receptor binding studies with selective somatostatin agonists have shown binding mainly in the regions with a low density of somatostatin fibres and presumably high numbers of membrane-associated sst2(a) receptors. In the present study, octreotide pre-treatments may have been ineffective because of a reduced availability of the membrane-associated sst2(a) receptors. Irritation caused by the cannula on the head could be a cause of a high endogenous TNC I-II somatostatin release and this may have triggered the internalization of the receptors before the start of the experiment. The effects of a chronic slight irritation are not visible in the Fos expression because Fos is a novelty marker (50). That octreotide did not have access to the sst2 receptors because of internalization or did not reach the threshold value in the TNC I-II after the intracisternal administration could be an alternative explanation. The spinal trigeminal tract may be a diffusion barrier for octreotide, which is known for a poor penetration of the blood–brain barrier (10, 11). Octreotide-induced alterations of the behaviour after the intracisternal administration may have been caused by alternative routes of trigeminal activation. Octreotide binding, for example, has been shown in the nucleus of the solitary tract (27), which is located near the tip of the cannula and involved in orofacial motor control (51).
In conclusion, somatostatin analogues potentially are interesting compounds for reducing the transduction of glutamate-mediated trigeminal nociceptive signals. However, internalization and inactivation of the somatostatin receptors in somatostatin-rich areas like the TNC I-II may significantly reduce the efficacy and use of long-acting somatostatin analogues. More lipophylic compounds are needed to demonstrate the potential analgesic effects of the somatostatin analogues in models of trigeminovascular stimulation and migraine.
Footnotes
Acknowledgements
The authors would like to thank M.B. Spoelstra for histological work and Glaxo-Wellcome, Zeist, The Netherlands, for their generous financial support.