- Email: [email protected]
hypocretin-containing neurons in rat
Neuroscience 121 (2003) 269 –275
STIMULANT DOSES OF CAFFEINE INDUCE c-FOS ACTIVATION IN OREXIN/HYPOCRETIN-CONTAINING NEURONS IN RAT J. A. MURPHY, S. DEURVEILHER AND K. SEMBA*
have been suggested for the stimulant effects of caffeine. One is the activation of catecholaminergic systems, in particular the nigrostriatal (Garrett and Griffiths, 1997; Wisor et al., 2001) and mesolimbic dopaminergic systems (Solinas et al., 2002; but Acquas et al., 2002), and dopamine-adenosine interaction involving A2A receptors in the striatum (Fredholm, 1995; Lindskog et al., 2002). A second mechanism is through antagonism at the primarily inhibitory, adenosine A1 receptor (Snyder et al., 1981; Kaplan et al., 1992) present on wake-promoting neurons including basal forebrain neurons (Porkka-Heiskanen et al., 1997; Strecker et al., 2000). The possibility of direct actions at other wake-promoting neurons remains open. Recent evidence indicates that orexin (also known as hypocretin)-containing neurons play a critical role in wakefulness (Kilduff and Peyron, 2000; Willie et al., 2001; Taheri et al., 2002). In the CNS, these neurons are distributed solely in the posterior hypothalamic region surrounding the fornix, and send projections throughout the brain, prominently to monoaminergic and cholinergic wake-related brainstem neurons (Peyron et al., 1998; Nambu et al., 1999). Orexin neurons show an increase in c-Fos immunoreactivity after periods of active wakefulness (Estabrooke et al., 2001; Torterolo et al., 2001, 2003), and the majority of neurons in the perifornical region increase firing during wakefulness and rapid eye movement sleep (Alam et al., 2002). Orexin levels increase in the perifornical area and basal forebrain during these behavioural states (Kiyashchenko et al., 2002). I.c.v. or local administration of orexins promotes wakefulness, whereas impaired orexin transmission is associated with the sleep disorder narcolepsy (see above reviews). Collectively, these findings suggest that orexin-containing neurons play an important role in sleep–wake control. In the present study we investigated whether orexin neurons are activated following the administration of locomotorinducing, low doses of caffeine, using dual immunostaining for c-Fos and orexin B. We previously reported that stimulant doses of caffeine were not effective in inducing significant c-Fos expression in wakefulness-associated cell groups (Bennett and Semba, 1998). However, a more sensitive cFos antibody has since become available, and the recent evidence for the critical role of orexins provided a compelling reason to reinvestigate this issue with respect to orexin neurons.
Department of Anatomy and Neurobiology, Faculty of Medicine, Dalhousie University, Tupper Medical Building, Room 13MN, 5850 College Street, Halifax, Nova Scotia, Canada B3H 1X5
Abstract—Although caffeine is a commonly used CNS stimulant, neuronal mechanisms underlying its stimulatory effect are not fully understood. Orexin (hypocretin)-containing neurons play a critical role in arousal and might be activated by acute administration of caffeine. We examined this possibility by using dual-immunostaining for orexin B and c-Fos protein as a marker for neuronal activation. Rats were administered intraperitoneally with 10, 30 or 75 mg/kg caffeine, or saline. As previously reported, caffeine increased locomotion at 10 and 30 mg/kg, but not at 75 mg/kg. The numbers of orexinimmunoreactive and non-orexin-immunoreactive neurons expressing c-Fos were analysed using three counting boxes within the orexin field in the posterior hypothalamus. Compared with saline, all doses of caffeine increased the number of cells immunoreactive for both orexin and c-Fos. The average magnitude of this increase across doses in orexin neurons differed amongst regions; c-Fos expression increased by 343% in the perifornical area and by 158% in the more medial, dorsomedial nucleus. In the lateral hypothalamic area, c-Fos expression increased by 226% at 10 and 30 mg/kg but no change was seen at 75 mg/kg. In contrast, caffeine significantly increased the number of non-orexin-immunoreactive neurons expressing c-Fos only in the dorsomedial nucleus. These results indicate that systemically administered caffeine preferentially activates orexin neurons over non-orexin neurons in the same field, and that this activation is most pronounced in the perifornical region where orexin neurons are most concentrated. The activation of orexin neurons might play a role in the behavioural activation by caffeine. © 2003 IBRO. Published by Elsevier Ltd. All rights reserved. Key words: c-Fos immunohistochemistry, perifornical area, dorsomedial hypothalamic nucleus, lateral hypothalamic area, adenosine, wakefulness.
Caffeine is the most widely consumed mild stimulant. At low doses it increases neocortical arousal and vigilance in humans, whilst augmenting locomotion in animals; however, at high doses it induces nervousness and irritability in humans and immobility in animals (Nehlig et al., 1992; Sawynok, 1995). Pharmacologically, at low doses caffeine acts mainly as a non-selective antagonist at adenosine A1 and A2A receptors (Fredholm, 1995). Two mechanisms
EXPERIMENTAL PROCEDURES
*Corresponding author. Tel: ⫹1-902-494-2008; fax: ⫹1-902-4941212. E-mail address: [email protected] (K. Semba). Abbreviations: ABC, avidin– biotin– horseradish peroxidase complex; DAB, diaminobenzidine; DMH, dorsomedial hypothalamic nucleus; ir, immunoreactive; LH, lateral hypothalamic area; PeF, perifornical area.
Animals Male Wistar rats (Charles River Canada, St. Constant, PQ, Canada), 270 – 400 g, were pair-housed in cages under a 12-h light/
0306-4522/03$30.00⫹0.00 © 2003 IBRO. Published by Elsevier Ltd. All rights reserved. doi:10.1016/S0306-4522(03)00461-5
269
270
J. A. Murphy et al. / Neuroscience 121 (2003) 269 –275
dark cycle (lights on at 7:00 AM), with food and water available ad libitum. At least 1 week was allowed for acclimatization. The Dalhousie University Committee on Laboratory Animals in accordance with the Canadian Council on Animal Care approved all procedures involving animals. All efforts were made to minimise the number of animals used and their suffering.
Caffeine administration and behavioural observation Animals were weighed and separated into individual cages between 7:30 and 10:00 AM, and were left undisturbed for at least 90 min before injection of caffeine or saline. Caffeine (10, 30, or 75 mg/ml/kg, i.p. in saline; BDH Chemicals, Toronto, Ontario, Canada) or saline was injected between 10:30 AM and 12:55 PM, followed by 5-min behavioural observation immediately and 25, 55 and 85 min postinjection. These doses of caffeine were chosen on the basis of previous studies (Nehlig et al., 1992; Sawynok, 1995) including one from our laboratory (Bennett and Semba, 1998), showing that caffeine increases locomotor activity at 10 and 30 mg/kg, but not at 75 mg/kg. Animals were tested in groups of two or three, each animal with a different drug condition, and injection times and drug conditions were randomly assigned. All injections were made in the room that housed the animals.
Perfusion Ninety minutes after caffeine or saline injection, rats were overdosed with sodium pentobarbital (100 mg/kg, i.p.), and perfused with 0.1 M phosphate-buffered saline (pH 7.4), followed by 4% paraformaldehyde in 0.1 M phosphate buffer (pH 7.4), both at room temperature. Brains were post-fixed in the same fixative overnight at 4 °C, and cryoprotected with 30% sucrose in 0.1 M phosphate buffer.
Immunohistochemistry for c-Fos and orexin Immunohistochemistry was performed on every third coronal 40 m section using a standard avidin– biotin– horseradish peroxydase complex (ABC) method as previously described (Bennett and Semba, 1998), with a rabbit anti-Fos-oncoprotein antibody (1:20,000; catalogue no. PC38; Oncogene Research Products, Cambridge, MA, USA), a biotinylated donkey anti-rabbit IgG antibody (1:1000; Jackson Immunoresearch Laboratories, West Grove, PA, USA), ABC (1:200; Vectastain Elite ABC Kit; Vector Laboratories, Burlingame, CA, USA), and diaminobenzidine (DAB) in the presence of nickel ammonium sulphate to produce a black–purple reaction product. Sections containing the perifornical area (PeF) were then reacted for orexin-B, using an affinitypurified goat anti-orexin-B antibody (1:60,000; Santa Cruz Biotechnologies, Santa Cruz, CA, USA) and an ABC method with DAB to produce a brown reaction product. In a pilot experiment to visualise orexin/hypocretin-containing neurons, we used both antiorexin-A and anti-orexin-B antibodies (Santa Cruz Biotechnologies). Because of no apparent difference in the distribution of orexin-A-labelled neurons and orexin-B-labelled neurons, only anti-orexin-B was used in the present study. In support of these observations, hypocretin-1 was found to be colocalized with hypocretin-2 in the cat hypothalamus (Zhang et al., 2002).
Cell counts Cell counts were performed on a single section that contained the largest number of orexin-immunoreactive (ir) neurons (2.8 – 3.14 mm posterior to bregma; Paxinos and Watson, 1986). The examiner (J.A.M.) was blind to the treatment conditions. Single c-Fos-ir cells, single orexin-ir cells and double c-Fos-ir, orexin-ir cells were mapped using a computer-based image analysis system (Neurolucida 2000; MicroBrightField, Colchester, VA, USA). Three mediolaterally contiguous counting boxes, each measuring
Fig. 1. Section drawing illustrating the distribution of orexin-ir neurons, and placement of three cell counting boxes. Each dot represents one orexin-ir neuron. See Experimental Procedures for details. Abbreviations: 3V, third ventricle; f, fornix; ic, internal capsule; mt, mammillothalamic tract; opt, optic tract; VMH, ventromedial hypothalamic nucleus.
500 m⫻800 m, were used for quantitative analysis (Fig. 1). These boxes were placed such that the midpoint of the ventral segment of the PeF box and the centre of the fornix were 250 m apart vertically. Thus, these boxes approximately corresponded to the dorsomedial hypothalamic nucleus (DMH), PeF and lateral hypothalamic area (LH), respectively, and together, they covered most of the orexin neurons in a single section. Stereological methodologies were not used because the purpose of cell counting was to obtain a relative measure for c-Fos expression after caffeine.
Statistical analysis Bilateral counts of neurons immunolabeled for c-Fos and/or orexin in each counting box were averaged in each animal. The number of orexin-ir, c-Fos-ir cells was also converted into a percentage of orexin-ir neurons as well as a percentage of the control level. These values were analysed using three-way repeated measures ANOVA, followed by one-way and two-way ANOVA and post hoc Fisher’s pairwise comparisons tests using StatView 5.0 software (SAS Institute, Cary, NC, USA), with P⬍0.05 considered significant.
RESULTS The present results were derived from 25 rats. Seven received an injection of saline, and another seven, five and six received 10, 30 and 75 mg/kg caffeine, respectively. Behavioural responses to caffeine During the first 5-min period after injection, rats in all conditions displayed locomotion, sniffing, rearing and grooming. At 30 min, most saline-injected rats lay as if sleeping, with some animals displaying moderate levels of locomotor activity. At 60 and 90 min, all saline-injected rats appeared to be asleep. In contrast, rats administered with either 10 or 30 mg/kg caffeine exhibited locomotor activity at 30, 60, and 90 min. Rats administered with 75 mg/kg caffeine, however, showed minimal activity at 30 min, and at 60- and 90-min postinjection, they lay mostly immobile although apparently awake. These biphasic behavioural
J. A. Murphy et al. / Neuroscience 121 (2003) 269 –275
271
(mean⫾S.D.⫽222.8⫾39.3, n⫽25). These numbers did not differ between drug conditions (F3,21⬍1, NS). Of the three counting boxes, the PeF contained the highest number of orexin-ir neurons at 100.1⫾23.7 (45% of total population), followed by 74.9⫾16.2 (34%) in the DMH, and 47.8⫾10.9 (21%) in the LH (Fig. 1). Distribution of c-Fos-ir neurons after saline and caffeine injections
Fig. 2. Sections double-stained for c-Fos (black) and orexin-B (grey) in a saline-injected rat (A) and a rat injected with 30 mg/kg caffeine (B). Double black arrows indicate an orexin-ir, c-Fos-ir neuron; single black arrows indicate a non-orexin-ir, c-Fos-ir neuron; and the white arrow indicates an orexin-ir, non-c-Fos-ir neuron (unlabeled nucleus is visible at a different focal plane).
effects of caffeine are consistent with previous reports (Gupta et al., 1971; Nehlig et al., 1992; Bennett and Semba, 1998; Gasior et al., 2000). The numbers of orexin-ir neurons in the orexin field The total number of orexin-ir neurons from the three counting boxes (DMH, PeF, and LH) ranged from 111 to 281
Fig. 2 shows examples of c-Fos immunoreactivity in orexin-ir and non-orexin-ir neurons in the orexin field following saline (A) or 30 mg/kg caffeine injection (B). Purple– black c-Fos immunoreactivity was localised to cellular nuclei, whereas brown orexin immunoreactivity was cytoplasmic. The large size of c-Fos-labelled nuclei of orexinnegative cells suggested that they are neurons (see also Bennett and Semba, 1998). Fig. 3 shows examples of the distribution of c-Fos-ir neurons that are orexin-ir (A, C) or non-orexin-ir (B, D) over the three counting boxes after saline (A, B) or 10 mg/kg caffeine (C, D). As seen in these examples, c-Fos expression after saline and caffeine varied between the cell types and counting boxes. These differences were analysed quantitatively as described below. c-Fos expression in saline conditions Considerable differences in c-Fos expression were observed between the cell types and counting boxes under the saline condition (Fig. 3). There were generally fewer orexin-ir, c-Fos-ir neurons than non-orexin-ir, c-Fos-ir neurons, at a ratio of 1:6. For both types of neurons, more than one-half of c-Fos-ir neurons were concentrated in the DMH, and these numbers were greater than those for the PeF or LH (c-Fos-ir, orexin-ir neurons: mean⫾S.D.⫽ 17.6⫾10.7, 9.7⫾5.4; 3.0⫾1.6, for DMH, PeF and LH, respectively; F2,12⫽10.12, P⬍0.005; P⬍0.05 for DMH ver-
Fig. 3. Examples of distributions of orexin-ir, c-Fos-ir neurons and single labelled c-Fos-ir neurons in the DMH, PeF and LH counting boxes in a saline-administered rat (A, B) and a rat injected with 10 mg/kg caffeine (C, D). Each dot represents a double-labelled orexin-ir, c-Fos-ir cell; each “X” represents a single labelled c-Fos-ir nucleus.
272
J. A. Murphy et al. / Neuroscience 121 (2003) 269 –275
sus PeF/LH; c-Fos-ir, non-orexin-ir neurons: 112.1⫾59.7, 49.8⫾29.9, 34.0⫾24.6, for DMH, PeF and LH, respectively; F2,12⫽15.31, P⫽0.0005; P⬍0.025 for DMH versus PeF/LH). Because of these regional and cell type-specific differences under the saline condition, the effects of caffeine on c-Fos immunoreactivity were analysed as the percentages of the saline condition. Caffeine-induced c-Fos expression Caffeine increased the total number of c-Fos-ir cells in the orexin field compared with saline and regardless of cell type (F3,21⫽4.52, P⬍0.025, for Dose effect; Fig. 4A). However, the dose response of c-Fos expression was significantly different between the two cell types (F3,21⫽4.83, P⬍0.025, for the Cell type⫻Dose interaction). Thus, caffeine at all doses increased the total number of c-Fos-ir, orexin-ir neurons by more than 200% relative to saline values (F3,21⫽5.19, P⬍0.01; all doses, P⬍0.01 versus saline), with no significant difference amongst doses. In contrast to orexin-ir neurons, caffeine slightly increased the number of c-Fos-ir, non-orexin-ir neurons above control values (F3,21⫽3.00, P⫽0.0534 for Dose effect; Fig. 4A). The dose response in c-Fos expression differed between the counting boxes (F6,42⫽2.48, P⬍0.05 for Dose⫻Box interaction; Fig. 4B–D). Thus, the number of orexin-ir, c-Fos-ir neurons was generally greater for the PeF than for the DMH (P⬍0.005) or the LH (P⬍0.005). The number of non-orexin-ir, c-fos-ir neurons did not significantly differ across the three boxes or doses. The DMH. In the DMH, both orexin-ir and nonorexin-ir neurons showed increases in c-Fos expression after caffeine compared with saline (Fig. 4B). These increases were significant at all doses for orexin-ir neurons (by 158%; F3,21⫽4.21, P⬍0.025; P⬍0.05 versus saline), and at 30 and 75 mg/kg doses for non-orexin-ir neurons (by 115%; F3,21⫽4.35, P⬍0.025; P⬍0.05 versus saline). There was no significant difference between the two cell types. The PeF. Caffeine increased the number of orexin-ir neurons expressing c-Fos at all doses, by 343%, compared with saline (F3,21⫽4.69, P⬍0.025; P⬍0.025 versus saline; Fig. 4C). However, non-orexin-ir neurons did not show significant change in c-Fos expression (F3,21⫽1.77, NS). Thus, the difference in percent change in c-Fos expression between the two cell types was significant at 10 and 30 mg/kg doses (F3,21⫽6.49, P⬍0.005, for the Cell type⫻Dose interaction; P⬍0.05 versus saline). The LH. Caffeine increased c-Fos expression in orexin-ir neurons in the LH, by 226% at 10 and 30 mg/kg, Fig. 4. A. Numbers (mean⫹S.E.M.) of orexin-ir, c-Fos-ir cells and non-orexin-ir, c-Fos-ir cells combined for the three counting boxes (A), and separately for the DMH (B), PeF (C) and LH (D). E. Percentage of orexin-ir cells expressing c-Fos within each counting box. Stars, P⬍0.05 versus saline; asterisks, P⬍0.05 versus 75 mg/kg caffeine; horizontal brackets, P⬍0.05 orexin-ir, c-Fos-ir cells versus non-orexinir, c-Fos-ir cells.
J. A. Murphy et al. / Neuroscience 121 (2003) 269 –275
but not at 75 mg/kg (F3,21⫽6.03, P⬍0.005; P⬍0.01 versus saline; Fig. 4D). Unlike orexin-ir neurons, caffeine nonsignificantly increased the number of non-orexin-ir, c-Fos-ir neurons above control values (F3,21⫽1.95, NS). Thus, orexin-ir neurons showed a greater overall percent increase than did non-orexin-ir in the LH (F3,21⫽6.78, P⬍0.005, for the Cell type⫻Dose interaction; Fig. 4D). The number of orexin-ir cells expressing c-Fos was also analysed as a percentage of orexin-ir neurons (Fig. 4E). Caffeine increased this percentage in the total orexin field from a mean of 16% in the saline condition to 41– 46% at the three doses (F3,21⫽5.42, P⬍0.01 for Dose effect; P⬍0.0001 versus saline). This increase was significant at all doses in the PeF and DMH (F3,21⫽5.16 and 4.09, respectively, P⬍0.025; P⬍0.025 versus saline), and at 10 and 30, but not 75 mg/kg, in the LH (F3,21⫽7.47, P⫽0.001; P⬍0.005 versus saline). Overall, the DMH showed the highest percentage, at 59% (all P⬍0.0001, DMH versus PeF/LH), followed by the PeF at 41%, and the LH at 17% (P⬍0.0001, PeF versus LH). Compared at specific dose conditions, the DMH showed a higher percentage than the PeF at 0, 30 and 75 mg/kg (P⬍0.005), and the LH at all doses (P⬍0.025); the PeF showed a higher percentage than the LH at all doses (P⬍0.05).
DISCUSSION We found that caffeine increased c-Fos expression in orexin neurons by approximately 200% at each of the three doses used (10, 30 and 75 mg/kg, i.p.) compared with control levels with saline. These neurons accounted for 40% of orexin neurons. In contrast, non-orexin neurons intermixed with orexin neurons responded to caffeine with a non-significant increase in c-Fos expression. c-Fos expression showed regional differences within the orexin field. Technical considerations The c-Fos expression in the present study was generally greater than in our previous study that also investigated caffeine-induced c-Fos immunoreactivity but without neurochemical identification (Bennett and Semba, 1998). The timing of perfusion after caffeine injection is slightly different, 90 min in the present study versus 120 min in the previous study. However, both are within the range of peak c-Fos immunoreactivity after stimulation (Dragunow and Faull, 1989; Morgan and Curran, 1989). Thus, the main reason for the discrepancy is probably the use of a more sensitive c-Fos antibody in the present study. A maximum of 60% of orexin neurons were activated in any subregion of the orexin field at any dose used. This percentage may be underestimated due to limited penetration of antibodies, which might favour detection of the cytoplasmic orexin, as opposed to the nuclear c-Fos. However, we believe that such bias is unlikely because only large neurons were counted. Furthermore, up to 78% of orexin neurons were shown to express c-Fos in actively moving cats (Torterolo et al., 2001). Thus, 60% is probably not far from the true proportion of orexin neurons that
273
express c-Fos to caffeine. The partial activation of orexin neurons might be due to the functional heterogeneity of orexin neurons (see below) and/or the possibility that some orexin neurons express c-Fos more readily. Mechanisms of caffeine-induced activation of orexin neurons Both direct and indirect mechanisms might mediate the c-Fos expression of orexin neurons after low doses of caffeine. First, c-Fos activation can be a direct result of caffeine’s antagonism of endogenous adenosine at adenosine A1 receptors present on orexin neurons. The primary action of caffeine at low, psychostimulant doses is thought to be antagonism at A1 and A2A receptors (Fredholm, 1995). Of these, A2A receptors are virtually absent from the posterior hypothalamus (Rosin et al., 1998), whereas A1 receptors, which are primarily inhibitory, are present in the area where orexinergic neurons are concentrated (Reppert et al., 1991; Rivkees et al., 1995). It was recently reported that 30% of orexin neurons were immunoreactive for A1 receptors (Thakkar et al., 2002). Thus, caffeine could activate orexin neurons by blockade of postsynaptic A1 receptors. Orexin neurons receive glutamatergic input (Li et al., 2002). Thus, caffeine could also activate orexin neurons by presynaptically disinhibiting the glutamatergic input through a well-documented presynaptic mechanism involving A1 receptor activation (Dunwiddie, 1985; Greene and Haas, 1991). Alternatively, the c-Fos expression of orexin neurons at the low caffeine doses could be secondary to caffeine’s actions elsewhere, for example, the nigrostriatal (Garrett and Griffiths, 1997; Wisor et al., 2001; Lindskog et al., 2002) and mesolimbic dopaminergic systems (Solinas et al., 2002; but Acquas et al., 2002). According to this scenario, caffeine’s primary action at these extrahypothalamic sites is responsible for the behavioural activation, which secondarily activates orexin neurons. These direct and indirect effects are not mutually exclusive, and the c-Fos expression of orexin neurons could result from combined effects. The mechanisms for the c-Fos expression at the highest dose (75 mg/kg) of caffeine are less clear, in part due to caffeine’s additional action as a phosphodiesterase inhibitor (Choi et al., 1988). Phosphodiesterase inhibitors produced behavioural depression (Choi et al., 1988), which is consistent with the lack of behavioural activation at 75 mg/kg caffeine. Similar to lower dose effects, both direct and indirect effects are likely to be involved. Activation of orexin neurons and locomotor activity Behavioural activation was accompanied by the activation of orexin neurons, as measured by c-Fos expression, in all subregions of the orexin field. However, the highest dose of caffeine, which induced behavioural immobility, was also associated with c-Fos expression in orexin neurons in the DMH and PeF, but not in the LH. These observations suggest that orexin neurons in all subregions that are activated by low doses of caffeine might be involved in behavioural activation, although the possibility remains
274
J. A. Murphy et al. / Neuroscience 121 (2003) 269 –275
that caffeine acts elsewhere (see above) to induce behavioural activation which in turn activates orexin neurons. The biphasic behavioural response is most closely paralleled by the inverted U-shaped pattern of c-Fos expression of orexin neurons in the LH. The selective activation of orexin neurons during caffeine-induced behavioural activation is consistent with the recent evidence for close association of orexin neuron activity with motor activity. Orexin neurons selectively expressed c-Fos during wake periods with somatomotor activity in cat (Torterolo et al., 2003). Orexin levels in the cerebrospinal fluid were higher during wake states with motor activity than quiet wake states (Kiyashchenko et al., 2002; Wu et al., 2002). Wake-related neurons, which accounted for 37% of neurons recorded in the rat PeF, increased firing rate during muscle activity across all sleep– wake states (Alam et al., 2002). These muscle activityrelated orexin neurons might enhance locomotion via previously described descending projections (Peyron et al., 1998; Nambu et al., 1999) to midbrain dopaminergic neurons and/or somatic motor neurons. The former possibility is supported by the blockade of orexin-induced increase in locomotion by dopamine receptor antagonists (Nakamura et al., 2000). Two explanations may be offered for the c-Fos expression in the DMH and PeF after the immobility-inducing dose of caffeine. First, the downstream pathway of orexin neurons activated at high doses of caffeine might be inhibited by mechanisms that are initiated only at the highest dose of caffeine. This could result in behavioural immobility despite the neuronal activation. An alternative explanation is that high doses of caffeine have peripheral cardiovascular effects, which can secondarily activate neurons involved in autonomic regulation (Bennett and Semba, 1998). The posterior hypothalamus contains neurons involved in cardiovascular regulation (Allen and Cechetto, 1992), and the possibility exists that the orexin neurons in the DMH and PeF that express c-Fos after 75 mg/kg caffeine might not be the same neurons as those that expressed c-Fos after the lower doses of caffeine, and may be involved in the cardiovascular response to peripheral caffeine. Regional heterogeneity in the responsiveness of orexin neurons to caffeine, as well as the partial participation in the response, is not surprising in light of their suggested roles not only in wakefulness, but also in locomotion, feeding, and neuroendocrine and autonomic regulation (Kilduff and Peyron, 2000; Willie et al., 2001; Taheri et al., 2002). Thus, low doses of caffeine might activate only those orexin neurons involved in wakefulness and locomotion. Comparisons with other stimulants Caffeine appears to induce similar levels of c-Fos expression in orexin neurons as modafinil, a drug of unknown mechanism of action used clinically to treat excessive daytime sleepiness (Chemelli et al., 1999; Scammell et al., 2000). Modafinil also only slightly increases c-Fos expression in non-orexin neurons within the PeF. However, modafinil increases waking dose-dependently (Scammell
et al., 2000), and without increasing locomotion (Edgar and Seidel, 1997). Whether or not the activation of orexin neurons is necessary for the wake-enhancing effect of modafinil has not been determined. The amount of c-Fos expression after caffeine seems to be less than c-Fos expression reported after antipsychotic stimulant methamphetamine (0.5 mg/kg, i.p.; Estabrooke et al., 2001) or amphetamine (1.5 mg/kg, i.p.; Fadel et al., 2002). The c-Fos expression induced by amphetamine was also limited to orexin neurons located medial to the fornix, but clozapine, an antipsychotic drug with weight loss side effect, increased c-Fos expression in orexin neurons in the lateral LH/PeF (Fadel et al., 2002).
CONCLUSIONS The present results indicate that the acute administration of caffeine preferentially activates orexin neurons over non-orexin neurons in the orexin field, and that this activation is most pronounced in the PeF where orexin neurons are most concentrated. The activation of orexin neurons might play a role in behavioural activation by caffeine. Acknowledgements—This work was supported by the Canadian Institutes of Health Research (MOP14451). S.D. was a recipient of Fondation Singer-Polignac and Nova Scotia Health Research Foundation Fellowships. We thank Douglas Rasmusson for helpful comments and Joan Burns for technical assistance.
REFERENCES Acquas E, Tanda G, Di Chiara G (2002) Differential effects of caffeine on dopamine and acetylcholine transmission in brain areas of drugnaı¨ve and caffeine-pretreated rats. Neuropsychopharmacology 27: 182–193. Alam MN, Gong H, Alam T, Jaganath R, McGinty D, Szymusiak R (2002) Sleep-waking discharge patterns of neurons recorded in the rat perifornical lateral hypothalamic area. J Physiol 538:619 –631. Allen GV, Cechetto DF (1992) Functional and anatomical organization of cardiovascular pressor and depressor sites in the lateral hypothalamic area: I. Descending projections. J Comp Neurol 315:313– 332. Bennett HJ, Semba K (1998) Immunohistochemical localization of caffeine-induced Fos protein expression in the rat brain. J Comp Neurol 401:89 –108. Chemelli RM, Willie JT, Sinton CM, Elmquist JK, Scammell T, Lee C, Richardson JA, Williams SC, Xiong Y, Kisanuki Y, Fitch TE, Saper CB, Yanagisawa M (1999) Narcolepsy in orexin knockout mice: molecular genetics of sleep regulation. Cell 98:437–451. Choi OH, Shamim MT, Padgett WL, Daly JW (1988) Caffeine and theophylline analogues: correlation of behavioral effects with activity as adenosine receptor antagonists and as phosphodiesterase inhibitors. Life Sci 43:387–398. Dragunow M, Faull R (1989) The use of c-fos as a metabolic marker in neuronal pathway tracing. J Neurosci Methods 29:261–265. Dunwiddie TV (1985) The physiological role of adenosine in the central nervous system. Int Rev Neurobiol 27:63–139. Edgar DM, Seidel WF (1997) Modafinil induces wakefulness without intensifying motor activity or subsequent rebound hypersomnolence in the rat. J Pharmacol Exp Ther 283:757–769. Estabrooke IV, McCarthy MT, Ko E, Chou TC, Chemelli RM, Yanagisawa M, Saper CB, Scammell TE (2001) Fos expression in orexin neurons varies with behavioral state. J Neurosci 21:1656 –1662. Fadel J, Bubser M, Deutch AY (2002) Differential activation of orexin
J. A. Murphy et al. / Neuroscience 121 (2003) 269 –275 neurons by antipsychotic drugs associated with weight gain. J Neurosci 22:6742–6746. Fredholm BB (1995) Adenosine, adenosine receptors and the actions of caffeine. Pharmacol Toxicol 76:93–101. Garrett BE, Griffiths RR (1997) The role of dopamine in the behavioral effects of caffeine in animals and humans. Pharmacol Biochem Behav 57:533–541. Gasior M, Jaszyna M, Peters J, Goldberg SR (2000) Changes in the ambulatory activity and discriminative stimulus effects of psychostimulant drugs in rats chronically exposed to caffeine: effect of caffeine dose. J Pharmacol Exp Ther 295:1101–1111. Greene RW, Haas HL (1991) The electrophysiology of adenosine in the mammalian central nervous system. Prog Neurobiol 36:329 – 341. Gupta BD, Dandiya PC, Gupta ML (1971) A psycho-pharmacological analysis of behaviour in rats. Jpn J Pharmacol 21:293–298. Kaplan GB, Greenblatt DJ, Kent MA, Cotreau MM, Arcelin G, Shader RI (1992) Caffeine-induced behavioral stimulation is dose-dependent and associated with A1 adenosine receptor occupancy. Neuropsychopharmacology 6:145–153. Kilduff TS, Peyron C (2000) The hypocretin/orexin ligand-receptor system: implications for sleep and sleep disorders. Trends Neurosci 23:359 –365. Kiyashchenko LI, Mileykovskiy BY, Maidment N, Lam HA, Wu MF, John J, Peever J, Siegel JM (2002) Release of hypocretin (orexin) during waking and sleep states. J Neurosci 22:5282–5286. Li Y, Gao XB, Sakurai T, van den Pol AN (2002) Hypocretin/orexin excites hypocretin neurons via a local glutamate neuron-A potential mechanism for orchestrating the hypothalamic arousal system. Neuron 36:1169 –1181. Lindskog M, Svenningsson P, Pozzi L, Kim Y, Fienberg AA, Bibb JA, Fredholm BB, Nairn AC, Greengard P, Fisone G (2002) Involvement of DARPP-32 phosphorylation in the stimulant action of caffeine. Nature 418:774 –778. Morgan JI, Curran T (1989) Stimulus-transcription coupling in neurons: role of cellular immediate-early genes. Trends Neurosci 12:459 – 462. Nakamura T, Uramura K, Nambu T, Yada T, Goto K, Yanagisawa M, Sakurai T (2000) Orexin-induced hyperlocomotion and stereotypy are mediated by the dopaminergic system. Brain Res 873:181–187. Nambu T, Sakurai T, Mizukami K, Hosoya Y, Yanagisawa M, Goto K (1999) Distribution of orexin neurons in the adult rat brain. Brain Res 827:243–260. Nehlig A, Daval J-C, Debry G (1992) Caffeine and the central nervous system: mechanisms of action, biochemical, metabolic and psychostimulant effects. Brain Res Rev 17:139 –170. Paxinos G, Watson C (1986) The rat brain in stereotaxic coordinates, 2nd ed. Orlando, FL: Academic Press. Peyron C, Tighe DK, van den Pol AN, de Lecea L, Heller HC, Sutcliffe JG, Kilduff TS (1998) Neurons containing hypocretin (orexin) project to multiple neuronal systems. J Neurosci 18:9996 –10015. Porkka-Heiskanen T, Strecker RE, Thakkar M, Bjorkum AA, Greene RW, McCarley RW (1997) Adenosine: a mediator of the sleepinducing effects of prolonged wakefulness. Science 276:1265– 1268.
275
Reppert SM, Weaver DR, Stehle JH, Rivkees SA (1991) Molecular cloning and characterization of a rat A1-adenosine receptor that is widely expressed in brain and spinal cord. Mol Endocrinol 5:1037– 1048. Rivkees SA, Price SL, Zhou FC (1995) Immunohistochemical detection of A1 adenosine receptors in rat brain with emphasis on localization in the hippocampal formation, cerebral cortex, cerebellum, and basal ganglia. Brain Res 677:193–203. Rosin DL, Robeva A, Woodard RL, Guyenet P, Linden J (1998) Immunohistochemical localization of adenosine A2A receptors in the rat central nervous system. J Comp Neurol 401:163–186. Sawynok J (1995) Pharmacological rationale for the clinical use of caffeine. Drugs 49:37–50. Scammell TE, Estabrooke IV, McCarthy MT, Chemelli RM, Yanagisawa M, Miller MS, Saper CB (2000) Hypothalamic arousal regions are activated during modafinil-induced wakefulness. J Neurosci 20:8620 –8628. Snyder SH, Katims JJ, Annau Z, Bruns RF, Daly JW (1981) Adenosine receptors and behavioral actions of methylxanthines. Proc Natl Acad Sci USA 78:3260 –3264. Solinas M, Ferre S, You ZB, Karcz-Kubicha M, Popoli P, Goldberg SR (2002) Caffeine induces dopamine and glutamate release in the shell of the nucleus accumbens. J Neurosci 22:6321–6324. Strecker RE, Morairty S, Thakkar MM, Porkka-Heiskanen T, Basheer R, Dauphin LJ, Rainnie DG, Portas CM, Greene RW, McCarley RW (2000) Adenosinergic modulation of basal forebrain and preoptic/ anterior hypothalamic neuronal activity in the control of behavioral state. Behav Brain Res 115:183–204. Taheri S, Zeitzer JM, Mignot E (2002) The role of hypocretins (orexins) in sleep regulation and narcolepsy. Annu Rev Neurosci 25:283– 313. Thakkar MM, Winston S, McCarley RW (2002) Orexin neurons of the hypothalamus express adenosine A1 receptors. Brain Res 944: 190 –194. Torterolo P, Yamuy J, Sampogna S, Morales FR, Chase MH (2001) Hypothalamic neurons that contain hypocretin (orexin) express cfos during active wakefulness and carbachol-induced active sleep. Sleep Res Online 4:25–32. Torterolo P, Yamuy J, Sampogna S, Morales FR, Chase MH (2003) Hypocretinergic neurons are primarily involved in activation of the somatomotor system. Sleep 26:25–28. Willie JT, Chemelli RM, Sinton CM, Yanagisawa M (2001) To eat or to sleep? Orexin in the regulation of feeding and wakefulness. Annu Rev Neurosci 24:429 –458. Wisor JP, Nishino S, Sora I, Uhl GH, Mignot E, Edgar DM (2001) Dopaminergic role in stimulant-induced wakefulness. J Neurosci 21:1787–1794. Wu M-F, John J, Maidment N, Lam HA, Siegel JM (2002) Hypocretin release in normal and narcoleptic dogs after food and sleep deprivation, eating, and movement. Am J Physiol Integr Comp Physiol 283:R1079 –R1086. Zhang JH, Sampogna S, Morales FR, Chase MH (2002) Co-localization of hypocretin-1 and hypocretin-2 in the cat hypothalamus and brainstem. Peptides 23:1479 –1483.
(Accepted 9 June 2003)
STIMULANT DOSES OF CAFFEINE INDUCE c-FOS ACTIVATION IN OREXIN/HYPOCRETIN-CONTAINING NEURONS IN RAT J. A. MURPHY, S. DEURVEILHER AND K. SEMBA*
have been suggested for the stimulant effects of caffeine. One is the activation of catecholaminergic systems, in particular the nigrostriatal (Garrett and Griffiths, 1997; Wisor et al., 2001) and mesolimbic dopaminergic systems (Solinas et al., 2002; but Acquas et al., 2002), and dopamine-adenosine interaction involving A2A receptors in the striatum (Fredholm, 1995; Lindskog et al., 2002). A second mechanism is through antagonism at the primarily inhibitory, adenosine A1 receptor (Snyder et al., 1981; Kaplan et al., 1992) present on wake-promoting neurons including basal forebrain neurons (Porkka-Heiskanen et al., 1997; Strecker et al., 2000). The possibility of direct actions at other wake-promoting neurons remains open. Recent evidence indicates that orexin (also known as hypocretin)-containing neurons play a critical role in wakefulness (Kilduff and Peyron, 2000; Willie et al., 2001; Taheri et al., 2002). In the CNS, these neurons are distributed solely in the posterior hypothalamic region surrounding the fornix, and send projections throughout the brain, prominently to monoaminergic and cholinergic wake-related brainstem neurons (Peyron et al., 1998; Nambu et al., 1999). Orexin neurons show an increase in c-Fos immunoreactivity after periods of active wakefulness (Estabrooke et al., 2001; Torterolo et al., 2001, 2003), and the majority of neurons in the perifornical region increase firing during wakefulness and rapid eye movement sleep (Alam et al., 2002). Orexin levels increase in the perifornical area and basal forebrain during these behavioural states (Kiyashchenko et al., 2002). I.c.v. or local administration of orexins promotes wakefulness, whereas impaired orexin transmission is associated with the sleep disorder narcolepsy (see above reviews). Collectively, these findings suggest that orexin-containing neurons play an important role in sleep–wake control. In the present study we investigated whether orexin neurons are activated following the administration of locomotorinducing, low doses of caffeine, using dual immunostaining for c-Fos and orexin B. We previously reported that stimulant doses of caffeine were not effective in inducing significant c-Fos expression in wakefulness-associated cell groups (Bennett and Semba, 1998). However, a more sensitive cFos antibody has since become available, and the recent evidence for the critical role of orexins provided a compelling reason to reinvestigate this issue with respect to orexin neurons.
Department of Anatomy and Neurobiology, Faculty of Medicine, Dalhousie University, Tupper Medical Building, Room 13MN, 5850 College Street, Halifax, Nova Scotia, Canada B3H 1X5
Abstract—Although caffeine is a commonly used CNS stimulant, neuronal mechanisms underlying its stimulatory effect are not fully understood. Orexin (hypocretin)-containing neurons play a critical role in arousal and might be activated by acute administration of caffeine. We examined this possibility by using dual-immunostaining for orexin B and c-Fos protein as a marker for neuronal activation. Rats were administered intraperitoneally with 10, 30 or 75 mg/kg caffeine, or saline. As previously reported, caffeine increased locomotion at 10 and 30 mg/kg, but not at 75 mg/kg. The numbers of orexinimmunoreactive and non-orexin-immunoreactive neurons expressing c-Fos were analysed using three counting boxes within the orexin field in the posterior hypothalamus. Compared with saline, all doses of caffeine increased the number of cells immunoreactive for both orexin and c-Fos. The average magnitude of this increase across doses in orexin neurons differed amongst regions; c-Fos expression increased by 343% in the perifornical area and by 158% in the more medial, dorsomedial nucleus. In the lateral hypothalamic area, c-Fos expression increased by 226% at 10 and 30 mg/kg but no change was seen at 75 mg/kg. In contrast, caffeine significantly increased the number of non-orexin-immunoreactive neurons expressing c-Fos only in the dorsomedial nucleus. These results indicate that systemically administered caffeine preferentially activates orexin neurons over non-orexin neurons in the same field, and that this activation is most pronounced in the perifornical region where orexin neurons are most concentrated. The activation of orexin neurons might play a role in the behavioural activation by caffeine. © 2003 IBRO. Published by Elsevier Ltd. All rights reserved. Key words: c-Fos immunohistochemistry, perifornical area, dorsomedial hypothalamic nucleus, lateral hypothalamic area, adenosine, wakefulness.
Caffeine is the most widely consumed mild stimulant. At low doses it increases neocortical arousal and vigilance in humans, whilst augmenting locomotion in animals; however, at high doses it induces nervousness and irritability in humans and immobility in animals (Nehlig et al., 1992; Sawynok, 1995). Pharmacologically, at low doses caffeine acts mainly as a non-selective antagonist at adenosine A1 and A2A receptors (Fredholm, 1995). Two mechanisms
EXPERIMENTAL PROCEDURES
*Corresponding author. Tel: ⫹1-902-494-2008; fax: ⫹1-902-4941212. E-mail address: [email protected] (K. Semba). Abbreviations: ABC, avidin– biotin– horseradish peroxidase complex; DAB, diaminobenzidine; DMH, dorsomedial hypothalamic nucleus; ir, immunoreactive; LH, lateral hypothalamic area; PeF, perifornical area.
Animals Male Wistar rats (Charles River Canada, St. Constant, PQ, Canada), 270 – 400 g, were pair-housed in cages under a 12-h light/
0306-4522/03$30.00⫹0.00 © 2003 IBRO. Published by Elsevier Ltd. All rights reserved. doi:10.1016/S0306-4522(03)00461-5
269
270
J. A. Murphy et al. / Neuroscience 121 (2003) 269 –275
dark cycle (lights on at 7:00 AM), with food and water available ad libitum. At least 1 week was allowed for acclimatization. The Dalhousie University Committee on Laboratory Animals in accordance with the Canadian Council on Animal Care approved all procedures involving animals. All efforts were made to minimise the number of animals used and their suffering.
Caffeine administration and behavioural observation Animals were weighed and separated into individual cages between 7:30 and 10:00 AM, and were left undisturbed for at least 90 min before injection of caffeine or saline. Caffeine (10, 30, or 75 mg/ml/kg, i.p. in saline; BDH Chemicals, Toronto, Ontario, Canada) or saline was injected between 10:30 AM and 12:55 PM, followed by 5-min behavioural observation immediately and 25, 55 and 85 min postinjection. These doses of caffeine were chosen on the basis of previous studies (Nehlig et al., 1992; Sawynok, 1995) including one from our laboratory (Bennett and Semba, 1998), showing that caffeine increases locomotor activity at 10 and 30 mg/kg, but not at 75 mg/kg. Animals were tested in groups of two or three, each animal with a different drug condition, and injection times and drug conditions were randomly assigned. All injections were made in the room that housed the animals.
Perfusion Ninety minutes after caffeine or saline injection, rats were overdosed with sodium pentobarbital (100 mg/kg, i.p.), and perfused with 0.1 M phosphate-buffered saline (pH 7.4), followed by 4% paraformaldehyde in 0.1 M phosphate buffer (pH 7.4), both at room temperature. Brains were post-fixed in the same fixative overnight at 4 °C, and cryoprotected with 30% sucrose in 0.1 M phosphate buffer.
Immunohistochemistry for c-Fos and orexin Immunohistochemistry was performed on every third coronal 40 m section using a standard avidin– biotin– horseradish peroxydase complex (ABC) method as previously described (Bennett and Semba, 1998), with a rabbit anti-Fos-oncoprotein antibody (1:20,000; catalogue no. PC38; Oncogene Research Products, Cambridge, MA, USA), a biotinylated donkey anti-rabbit IgG antibody (1:1000; Jackson Immunoresearch Laboratories, West Grove, PA, USA), ABC (1:200; Vectastain Elite ABC Kit; Vector Laboratories, Burlingame, CA, USA), and diaminobenzidine (DAB) in the presence of nickel ammonium sulphate to produce a black–purple reaction product. Sections containing the perifornical area (PeF) were then reacted for orexin-B, using an affinitypurified goat anti-orexin-B antibody (1:60,000; Santa Cruz Biotechnologies, Santa Cruz, CA, USA) and an ABC method with DAB to produce a brown reaction product. In a pilot experiment to visualise orexin/hypocretin-containing neurons, we used both antiorexin-A and anti-orexin-B antibodies (Santa Cruz Biotechnologies). Because of no apparent difference in the distribution of orexin-A-labelled neurons and orexin-B-labelled neurons, only anti-orexin-B was used in the present study. In support of these observations, hypocretin-1 was found to be colocalized with hypocretin-2 in the cat hypothalamus (Zhang et al., 2002).
Cell counts Cell counts were performed on a single section that contained the largest number of orexin-immunoreactive (ir) neurons (2.8 – 3.14 mm posterior to bregma; Paxinos and Watson, 1986). The examiner (J.A.M.) was blind to the treatment conditions. Single c-Fos-ir cells, single orexin-ir cells and double c-Fos-ir, orexin-ir cells were mapped using a computer-based image analysis system (Neurolucida 2000; MicroBrightField, Colchester, VA, USA). Three mediolaterally contiguous counting boxes, each measuring
Fig. 1. Section drawing illustrating the distribution of orexin-ir neurons, and placement of three cell counting boxes. Each dot represents one orexin-ir neuron. See Experimental Procedures for details. Abbreviations: 3V, third ventricle; f, fornix; ic, internal capsule; mt, mammillothalamic tract; opt, optic tract; VMH, ventromedial hypothalamic nucleus.
500 m⫻800 m, were used for quantitative analysis (Fig. 1). These boxes were placed such that the midpoint of the ventral segment of the PeF box and the centre of the fornix were 250 m apart vertically. Thus, these boxes approximately corresponded to the dorsomedial hypothalamic nucleus (DMH), PeF and lateral hypothalamic area (LH), respectively, and together, they covered most of the orexin neurons in a single section. Stereological methodologies were not used because the purpose of cell counting was to obtain a relative measure for c-Fos expression after caffeine.
Statistical analysis Bilateral counts of neurons immunolabeled for c-Fos and/or orexin in each counting box were averaged in each animal. The number of orexin-ir, c-Fos-ir cells was also converted into a percentage of orexin-ir neurons as well as a percentage of the control level. These values were analysed using three-way repeated measures ANOVA, followed by one-way and two-way ANOVA and post hoc Fisher’s pairwise comparisons tests using StatView 5.0 software (SAS Institute, Cary, NC, USA), with P⬍0.05 considered significant.
RESULTS The present results were derived from 25 rats. Seven received an injection of saline, and another seven, five and six received 10, 30 and 75 mg/kg caffeine, respectively. Behavioural responses to caffeine During the first 5-min period after injection, rats in all conditions displayed locomotion, sniffing, rearing and grooming. At 30 min, most saline-injected rats lay as if sleeping, with some animals displaying moderate levels of locomotor activity. At 60 and 90 min, all saline-injected rats appeared to be asleep. In contrast, rats administered with either 10 or 30 mg/kg caffeine exhibited locomotor activity at 30, 60, and 90 min. Rats administered with 75 mg/kg caffeine, however, showed minimal activity at 30 min, and at 60- and 90-min postinjection, they lay mostly immobile although apparently awake. These biphasic behavioural
J. A. Murphy et al. / Neuroscience 121 (2003) 269 –275
271
(mean⫾S.D.⫽222.8⫾39.3, n⫽25). These numbers did not differ between drug conditions (F3,21⬍1, NS). Of the three counting boxes, the PeF contained the highest number of orexin-ir neurons at 100.1⫾23.7 (45% of total population), followed by 74.9⫾16.2 (34%) in the DMH, and 47.8⫾10.9 (21%) in the LH (Fig. 1). Distribution of c-Fos-ir neurons after saline and caffeine injections
Fig. 2. Sections double-stained for c-Fos (black) and orexin-B (grey) in a saline-injected rat (A) and a rat injected with 30 mg/kg caffeine (B). Double black arrows indicate an orexin-ir, c-Fos-ir neuron; single black arrows indicate a non-orexin-ir, c-Fos-ir neuron; and the white arrow indicates an orexin-ir, non-c-Fos-ir neuron (unlabeled nucleus is visible at a different focal plane).
effects of caffeine are consistent with previous reports (Gupta et al., 1971; Nehlig et al., 1992; Bennett and Semba, 1998; Gasior et al., 2000). The numbers of orexin-ir neurons in the orexin field The total number of orexin-ir neurons from the three counting boxes (DMH, PeF, and LH) ranged from 111 to 281
Fig. 2 shows examples of c-Fos immunoreactivity in orexin-ir and non-orexin-ir neurons in the orexin field following saline (A) or 30 mg/kg caffeine injection (B). Purple– black c-Fos immunoreactivity was localised to cellular nuclei, whereas brown orexin immunoreactivity was cytoplasmic. The large size of c-Fos-labelled nuclei of orexinnegative cells suggested that they are neurons (see also Bennett and Semba, 1998). Fig. 3 shows examples of the distribution of c-Fos-ir neurons that are orexin-ir (A, C) or non-orexin-ir (B, D) over the three counting boxes after saline (A, B) or 10 mg/kg caffeine (C, D). As seen in these examples, c-Fos expression after saline and caffeine varied between the cell types and counting boxes. These differences were analysed quantitatively as described below. c-Fos expression in saline conditions Considerable differences in c-Fos expression were observed between the cell types and counting boxes under the saline condition (Fig. 3). There were generally fewer orexin-ir, c-Fos-ir neurons than non-orexin-ir, c-Fos-ir neurons, at a ratio of 1:6. For both types of neurons, more than one-half of c-Fos-ir neurons were concentrated in the DMH, and these numbers were greater than those for the PeF or LH (c-Fos-ir, orexin-ir neurons: mean⫾S.D.⫽ 17.6⫾10.7, 9.7⫾5.4; 3.0⫾1.6, for DMH, PeF and LH, respectively; F2,12⫽10.12, P⬍0.005; P⬍0.05 for DMH ver-
Fig. 3. Examples of distributions of orexin-ir, c-Fos-ir neurons and single labelled c-Fos-ir neurons in the DMH, PeF and LH counting boxes in a saline-administered rat (A, B) and a rat injected with 10 mg/kg caffeine (C, D). Each dot represents a double-labelled orexin-ir, c-Fos-ir cell; each “X” represents a single labelled c-Fos-ir nucleus.
272
J. A. Murphy et al. / Neuroscience 121 (2003) 269 –275
sus PeF/LH; c-Fos-ir, non-orexin-ir neurons: 112.1⫾59.7, 49.8⫾29.9, 34.0⫾24.6, for DMH, PeF and LH, respectively; F2,12⫽15.31, P⫽0.0005; P⬍0.025 for DMH versus PeF/LH). Because of these regional and cell type-specific differences under the saline condition, the effects of caffeine on c-Fos immunoreactivity were analysed as the percentages of the saline condition. Caffeine-induced c-Fos expression Caffeine increased the total number of c-Fos-ir cells in the orexin field compared with saline and regardless of cell type (F3,21⫽4.52, P⬍0.025, for Dose effect; Fig. 4A). However, the dose response of c-Fos expression was significantly different between the two cell types (F3,21⫽4.83, P⬍0.025, for the Cell type⫻Dose interaction). Thus, caffeine at all doses increased the total number of c-Fos-ir, orexin-ir neurons by more than 200% relative to saline values (F3,21⫽5.19, P⬍0.01; all doses, P⬍0.01 versus saline), with no significant difference amongst doses. In contrast to orexin-ir neurons, caffeine slightly increased the number of c-Fos-ir, non-orexin-ir neurons above control values (F3,21⫽3.00, P⫽0.0534 for Dose effect; Fig. 4A). The dose response in c-Fos expression differed between the counting boxes (F6,42⫽2.48, P⬍0.05 for Dose⫻Box interaction; Fig. 4B–D). Thus, the number of orexin-ir, c-Fos-ir neurons was generally greater for the PeF than for the DMH (P⬍0.005) or the LH (P⬍0.005). The number of non-orexin-ir, c-fos-ir neurons did not significantly differ across the three boxes or doses. The DMH. In the DMH, both orexin-ir and nonorexin-ir neurons showed increases in c-Fos expression after caffeine compared with saline (Fig. 4B). These increases were significant at all doses for orexin-ir neurons (by 158%; F3,21⫽4.21, P⬍0.025; P⬍0.05 versus saline), and at 30 and 75 mg/kg doses for non-orexin-ir neurons (by 115%; F3,21⫽4.35, P⬍0.025; P⬍0.05 versus saline). There was no significant difference between the two cell types. The PeF. Caffeine increased the number of orexin-ir neurons expressing c-Fos at all doses, by 343%, compared with saline (F3,21⫽4.69, P⬍0.025; P⬍0.025 versus saline; Fig. 4C). However, non-orexin-ir neurons did not show significant change in c-Fos expression (F3,21⫽1.77, NS). Thus, the difference in percent change in c-Fos expression between the two cell types was significant at 10 and 30 mg/kg doses (F3,21⫽6.49, P⬍0.005, for the Cell type⫻Dose interaction; P⬍0.05 versus saline). The LH. Caffeine increased c-Fos expression in orexin-ir neurons in the LH, by 226% at 10 and 30 mg/kg, Fig. 4. A. Numbers (mean⫹S.E.M.) of orexin-ir, c-Fos-ir cells and non-orexin-ir, c-Fos-ir cells combined for the three counting boxes (A), and separately for the DMH (B), PeF (C) and LH (D). E. Percentage of orexin-ir cells expressing c-Fos within each counting box. Stars, P⬍0.05 versus saline; asterisks, P⬍0.05 versus 75 mg/kg caffeine; horizontal brackets, P⬍0.05 orexin-ir, c-Fos-ir cells versus non-orexinir, c-Fos-ir cells.
J. A. Murphy et al. / Neuroscience 121 (2003) 269 –275
but not at 75 mg/kg (F3,21⫽6.03, P⬍0.005; P⬍0.01 versus saline; Fig. 4D). Unlike orexin-ir neurons, caffeine nonsignificantly increased the number of non-orexin-ir, c-Fos-ir neurons above control values (F3,21⫽1.95, NS). Thus, orexin-ir neurons showed a greater overall percent increase than did non-orexin-ir in the LH (F3,21⫽6.78, P⬍0.005, for the Cell type⫻Dose interaction; Fig. 4D). The number of orexin-ir cells expressing c-Fos was also analysed as a percentage of orexin-ir neurons (Fig. 4E). Caffeine increased this percentage in the total orexin field from a mean of 16% in the saline condition to 41– 46% at the three doses (F3,21⫽5.42, P⬍0.01 for Dose effect; P⬍0.0001 versus saline). This increase was significant at all doses in the PeF and DMH (F3,21⫽5.16 and 4.09, respectively, P⬍0.025; P⬍0.025 versus saline), and at 10 and 30, but not 75 mg/kg, in the LH (F3,21⫽7.47, P⫽0.001; P⬍0.005 versus saline). Overall, the DMH showed the highest percentage, at 59% (all P⬍0.0001, DMH versus PeF/LH), followed by the PeF at 41%, and the LH at 17% (P⬍0.0001, PeF versus LH). Compared at specific dose conditions, the DMH showed a higher percentage than the PeF at 0, 30 and 75 mg/kg (P⬍0.005), and the LH at all doses (P⬍0.025); the PeF showed a higher percentage than the LH at all doses (P⬍0.05).
DISCUSSION We found that caffeine increased c-Fos expression in orexin neurons by approximately 200% at each of the three doses used (10, 30 and 75 mg/kg, i.p.) compared with control levels with saline. These neurons accounted for 40% of orexin neurons. In contrast, non-orexin neurons intermixed with orexin neurons responded to caffeine with a non-significant increase in c-Fos expression. c-Fos expression showed regional differences within the orexin field. Technical considerations The c-Fos expression in the present study was generally greater than in our previous study that also investigated caffeine-induced c-Fos immunoreactivity but without neurochemical identification (Bennett and Semba, 1998). The timing of perfusion after caffeine injection is slightly different, 90 min in the present study versus 120 min in the previous study. However, both are within the range of peak c-Fos immunoreactivity after stimulation (Dragunow and Faull, 1989; Morgan and Curran, 1989). Thus, the main reason for the discrepancy is probably the use of a more sensitive c-Fos antibody in the present study. A maximum of 60% of orexin neurons were activated in any subregion of the orexin field at any dose used. This percentage may be underestimated due to limited penetration of antibodies, which might favour detection of the cytoplasmic orexin, as opposed to the nuclear c-Fos. However, we believe that such bias is unlikely because only large neurons were counted. Furthermore, up to 78% of orexin neurons were shown to express c-Fos in actively moving cats (Torterolo et al., 2001). Thus, 60% is probably not far from the true proportion of orexin neurons that
273
express c-Fos to caffeine. The partial activation of orexin neurons might be due to the functional heterogeneity of orexin neurons (see below) and/or the possibility that some orexin neurons express c-Fos more readily. Mechanisms of caffeine-induced activation of orexin neurons Both direct and indirect mechanisms might mediate the c-Fos expression of orexin neurons after low doses of caffeine. First, c-Fos activation can be a direct result of caffeine’s antagonism of endogenous adenosine at adenosine A1 receptors present on orexin neurons. The primary action of caffeine at low, psychostimulant doses is thought to be antagonism at A1 and A2A receptors (Fredholm, 1995). Of these, A2A receptors are virtually absent from the posterior hypothalamus (Rosin et al., 1998), whereas A1 receptors, which are primarily inhibitory, are present in the area where orexinergic neurons are concentrated (Reppert et al., 1991; Rivkees et al., 1995). It was recently reported that 30% of orexin neurons were immunoreactive for A1 receptors (Thakkar et al., 2002). Thus, caffeine could activate orexin neurons by blockade of postsynaptic A1 receptors. Orexin neurons receive glutamatergic input (Li et al., 2002). Thus, caffeine could also activate orexin neurons by presynaptically disinhibiting the glutamatergic input through a well-documented presynaptic mechanism involving A1 receptor activation (Dunwiddie, 1985; Greene and Haas, 1991). Alternatively, the c-Fos expression of orexin neurons at the low caffeine doses could be secondary to caffeine’s actions elsewhere, for example, the nigrostriatal (Garrett and Griffiths, 1997; Wisor et al., 2001; Lindskog et al., 2002) and mesolimbic dopaminergic systems (Solinas et al., 2002; but Acquas et al., 2002). According to this scenario, caffeine’s primary action at these extrahypothalamic sites is responsible for the behavioural activation, which secondarily activates orexin neurons. These direct and indirect effects are not mutually exclusive, and the c-Fos expression of orexin neurons could result from combined effects. The mechanisms for the c-Fos expression at the highest dose (75 mg/kg) of caffeine are less clear, in part due to caffeine’s additional action as a phosphodiesterase inhibitor (Choi et al., 1988). Phosphodiesterase inhibitors produced behavioural depression (Choi et al., 1988), which is consistent with the lack of behavioural activation at 75 mg/kg caffeine. Similar to lower dose effects, both direct and indirect effects are likely to be involved. Activation of orexin neurons and locomotor activity Behavioural activation was accompanied by the activation of orexin neurons, as measured by c-Fos expression, in all subregions of the orexin field. However, the highest dose of caffeine, which induced behavioural immobility, was also associated with c-Fos expression in orexin neurons in the DMH and PeF, but not in the LH. These observations suggest that orexin neurons in all subregions that are activated by low doses of caffeine might be involved in behavioural activation, although the possibility remains
274
J. A. Murphy et al. / Neuroscience 121 (2003) 269 –275
that caffeine acts elsewhere (see above) to induce behavioural activation which in turn activates orexin neurons. The biphasic behavioural response is most closely paralleled by the inverted U-shaped pattern of c-Fos expression of orexin neurons in the LH. The selective activation of orexin neurons during caffeine-induced behavioural activation is consistent with the recent evidence for close association of orexin neuron activity with motor activity. Orexin neurons selectively expressed c-Fos during wake periods with somatomotor activity in cat (Torterolo et al., 2003). Orexin levels in the cerebrospinal fluid were higher during wake states with motor activity than quiet wake states (Kiyashchenko et al., 2002; Wu et al., 2002). Wake-related neurons, which accounted for 37% of neurons recorded in the rat PeF, increased firing rate during muscle activity across all sleep– wake states (Alam et al., 2002). These muscle activityrelated orexin neurons might enhance locomotion via previously described descending projections (Peyron et al., 1998; Nambu et al., 1999) to midbrain dopaminergic neurons and/or somatic motor neurons. The former possibility is supported by the blockade of orexin-induced increase in locomotion by dopamine receptor antagonists (Nakamura et al., 2000). Two explanations may be offered for the c-Fos expression in the DMH and PeF after the immobility-inducing dose of caffeine. First, the downstream pathway of orexin neurons activated at high doses of caffeine might be inhibited by mechanisms that are initiated only at the highest dose of caffeine. This could result in behavioural immobility despite the neuronal activation. An alternative explanation is that high doses of caffeine have peripheral cardiovascular effects, which can secondarily activate neurons involved in autonomic regulation (Bennett and Semba, 1998). The posterior hypothalamus contains neurons involved in cardiovascular regulation (Allen and Cechetto, 1992), and the possibility exists that the orexin neurons in the DMH and PeF that express c-Fos after 75 mg/kg caffeine might not be the same neurons as those that expressed c-Fos after the lower doses of caffeine, and may be involved in the cardiovascular response to peripheral caffeine. Regional heterogeneity in the responsiveness of orexin neurons to caffeine, as well as the partial participation in the response, is not surprising in light of their suggested roles not only in wakefulness, but also in locomotion, feeding, and neuroendocrine and autonomic regulation (Kilduff and Peyron, 2000; Willie et al., 2001; Taheri et al., 2002). Thus, low doses of caffeine might activate only those orexin neurons involved in wakefulness and locomotion. Comparisons with other stimulants Caffeine appears to induce similar levels of c-Fos expression in orexin neurons as modafinil, a drug of unknown mechanism of action used clinically to treat excessive daytime sleepiness (Chemelli et al., 1999; Scammell et al., 2000). Modafinil also only slightly increases c-Fos expression in non-orexin neurons within the PeF. However, modafinil increases waking dose-dependently (Scammell
et al., 2000), and without increasing locomotion (Edgar and Seidel, 1997). Whether or not the activation of orexin neurons is necessary for the wake-enhancing effect of modafinil has not been determined. The amount of c-Fos expression after caffeine seems to be less than c-Fos expression reported after antipsychotic stimulant methamphetamine (0.5 mg/kg, i.p.; Estabrooke et al., 2001) or amphetamine (1.5 mg/kg, i.p.; Fadel et al., 2002). The c-Fos expression induced by amphetamine was also limited to orexin neurons located medial to the fornix, but clozapine, an antipsychotic drug with weight loss side effect, increased c-Fos expression in orexin neurons in the lateral LH/PeF (Fadel et al., 2002).
CONCLUSIONS The present results indicate that the acute administration of caffeine preferentially activates orexin neurons over non-orexin neurons in the orexin field, and that this activation is most pronounced in the PeF where orexin neurons are most concentrated. The activation of orexin neurons might play a role in behavioural activation by caffeine. Acknowledgements—This work was supported by the Canadian Institutes of Health Research (MOP14451). S.D. was a recipient of Fondation Singer-Polignac and Nova Scotia Health Research Foundation Fellowships. We thank Douglas Rasmusson for helpful comments and Joan Burns for technical assistance.
REFERENCES Acquas E, Tanda G, Di Chiara G (2002) Differential effects of caffeine on dopamine and acetylcholine transmission in brain areas of drugnaı¨ve and caffeine-pretreated rats. Neuropsychopharmacology 27: 182–193. Alam MN, Gong H, Alam T, Jaganath R, McGinty D, Szymusiak R (2002) Sleep-waking discharge patterns of neurons recorded in the rat perifornical lateral hypothalamic area. J Physiol 538:619 –631. Allen GV, Cechetto DF (1992) Functional and anatomical organization of cardiovascular pressor and depressor sites in the lateral hypothalamic area: I. Descending projections. J Comp Neurol 315:313– 332. Bennett HJ, Semba K (1998) Immunohistochemical localization of caffeine-induced Fos protein expression in the rat brain. J Comp Neurol 401:89 –108. Chemelli RM, Willie JT, Sinton CM, Elmquist JK, Scammell T, Lee C, Richardson JA, Williams SC, Xiong Y, Kisanuki Y, Fitch TE, Saper CB, Yanagisawa M (1999) Narcolepsy in orexin knockout mice: molecular genetics of sleep regulation. Cell 98:437–451. Choi OH, Shamim MT, Padgett WL, Daly JW (1988) Caffeine and theophylline analogues: correlation of behavioral effects with activity as adenosine receptor antagonists and as phosphodiesterase inhibitors. Life Sci 43:387–398. Dragunow M, Faull R (1989) The use of c-fos as a metabolic marker in neuronal pathway tracing. J Neurosci Methods 29:261–265. Dunwiddie TV (1985) The physiological role of adenosine in the central nervous system. Int Rev Neurobiol 27:63–139. Edgar DM, Seidel WF (1997) Modafinil induces wakefulness without intensifying motor activity or subsequent rebound hypersomnolence in the rat. J Pharmacol Exp Ther 283:757–769. Estabrooke IV, McCarthy MT, Ko E, Chou TC, Chemelli RM, Yanagisawa M, Saper CB, Scammell TE (2001) Fos expression in orexin neurons varies with behavioral state. J Neurosci 21:1656 –1662. Fadel J, Bubser M, Deutch AY (2002) Differential activation of orexin
J. A. Murphy et al. / Neuroscience 121 (2003) 269 –275 neurons by antipsychotic drugs associated with weight gain. J Neurosci 22:6742–6746. Fredholm BB (1995) Adenosine, adenosine receptors and the actions of caffeine. Pharmacol Toxicol 76:93–101. Garrett BE, Griffiths RR (1997) The role of dopamine in the behavioral effects of caffeine in animals and humans. Pharmacol Biochem Behav 57:533–541. Gasior M, Jaszyna M, Peters J, Goldberg SR (2000) Changes in the ambulatory activity and discriminative stimulus effects of psychostimulant drugs in rats chronically exposed to caffeine: effect of caffeine dose. J Pharmacol Exp Ther 295:1101–1111. Greene RW, Haas HL (1991) The electrophysiology of adenosine in the mammalian central nervous system. Prog Neurobiol 36:329 – 341. Gupta BD, Dandiya PC, Gupta ML (1971) A psycho-pharmacological analysis of behaviour in rats. Jpn J Pharmacol 21:293–298. Kaplan GB, Greenblatt DJ, Kent MA, Cotreau MM, Arcelin G, Shader RI (1992) Caffeine-induced behavioral stimulation is dose-dependent and associated with A1 adenosine receptor occupancy. Neuropsychopharmacology 6:145–153. Kilduff TS, Peyron C (2000) The hypocretin/orexin ligand-receptor system: implications for sleep and sleep disorders. Trends Neurosci 23:359 –365. Kiyashchenko LI, Mileykovskiy BY, Maidment N, Lam HA, Wu MF, John J, Peever J, Siegel JM (2002) Release of hypocretin (orexin) during waking and sleep states. J Neurosci 22:5282–5286. Li Y, Gao XB, Sakurai T, van den Pol AN (2002) Hypocretin/orexin excites hypocretin neurons via a local glutamate neuron-A potential mechanism for orchestrating the hypothalamic arousal system. Neuron 36:1169 –1181. Lindskog M, Svenningsson P, Pozzi L, Kim Y, Fienberg AA, Bibb JA, Fredholm BB, Nairn AC, Greengard P, Fisone G (2002) Involvement of DARPP-32 phosphorylation in the stimulant action of caffeine. Nature 418:774 –778. Morgan JI, Curran T (1989) Stimulus-transcription coupling in neurons: role of cellular immediate-early genes. Trends Neurosci 12:459 – 462. Nakamura T, Uramura K, Nambu T, Yada T, Goto K, Yanagisawa M, Sakurai T (2000) Orexin-induced hyperlocomotion and stereotypy are mediated by the dopaminergic system. Brain Res 873:181–187. Nambu T, Sakurai T, Mizukami K, Hosoya Y, Yanagisawa M, Goto K (1999) Distribution of orexin neurons in the adult rat brain. Brain Res 827:243–260. Nehlig A, Daval J-C, Debry G (1992) Caffeine and the central nervous system: mechanisms of action, biochemical, metabolic and psychostimulant effects. Brain Res Rev 17:139 –170. Paxinos G, Watson C (1986) The rat brain in stereotaxic coordinates, 2nd ed. Orlando, FL: Academic Press. Peyron C, Tighe DK, van den Pol AN, de Lecea L, Heller HC, Sutcliffe JG, Kilduff TS (1998) Neurons containing hypocretin (orexin) project to multiple neuronal systems. J Neurosci 18:9996 –10015. Porkka-Heiskanen T, Strecker RE, Thakkar M, Bjorkum AA, Greene RW, McCarley RW (1997) Adenosine: a mediator of the sleepinducing effects of prolonged wakefulness. Science 276:1265– 1268.
275
Reppert SM, Weaver DR, Stehle JH, Rivkees SA (1991) Molecular cloning and characterization of a rat A1-adenosine receptor that is widely expressed in brain and spinal cord. Mol Endocrinol 5:1037– 1048. Rivkees SA, Price SL, Zhou FC (1995) Immunohistochemical detection of A1 adenosine receptors in rat brain with emphasis on localization in the hippocampal formation, cerebral cortex, cerebellum, and basal ganglia. Brain Res 677:193–203. Rosin DL, Robeva A, Woodard RL, Guyenet P, Linden J (1998) Immunohistochemical localization of adenosine A2A receptors in the rat central nervous system. J Comp Neurol 401:163–186. Sawynok J (1995) Pharmacological rationale for the clinical use of caffeine. Drugs 49:37–50. Scammell TE, Estabrooke IV, McCarthy MT, Chemelli RM, Yanagisawa M, Miller MS, Saper CB (2000) Hypothalamic arousal regions are activated during modafinil-induced wakefulness. J Neurosci 20:8620 –8628. Snyder SH, Katims JJ, Annau Z, Bruns RF, Daly JW (1981) Adenosine receptors and behavioral actions of methylxanthines. Proc Natl Acad Sci USA 78:3260 –3264. Solinas M, Ferre S, You ZB, Karcz-Kubicha M, Popoli P, Goldberg SR (2002) Caffeine induces dopamine and glutamate release in the shell of the nucleus accumbens. J Neurosci 22:6321–6324. Strecker RE, Morairty S, Thakkar MM, Porkka-Heiskanen T, Basheer R, Dauphin LJ, Rainnie DG, Portas CM, Greene RW, McCarley RW (2000) Adenosinergic modulation of basal forebrain and preoptic/ anterior hypothalamic neuronal activity in the control of behavioral state. Behav Brain Res 115:183–204. Taheri S, Zeitzer JM, Mignot E (2002) The role of hypocretins (orexins) in sleep regulation and narcolepsy. Annu Rev Neurosci 25:283– 313. Thakkar MM, Winston S, McCarley RW (2002) Orexin neurons of the hypothalamus express adenosine A1 receptors. Brain Res 944: 190 –194. Torterolo P, Yamuy J, Sampogna S, Morales FR, Chase MH (2001) Hypothalamic neurons that contain hypocretin (orexin) express cfos during active wakefulness and carbachol-induced active sleep. Sleep Res Online 4:25–32. Torterolo P, Yamuy J, Sampogna S, Morales FR, Chase MH (2003) Hypocretinergic neurons are primarily involved in activation of the somatomotor system. Sleep 26:25–28. Willie JT, Chemelli RM, Sinton CM, Yanagisawa M (2001) To eat or to sleep? Orexin in the regulation of feeding and wakefulness. Annu Rev Neurosci 24:429 –458. Wisor JP, Nishino S, Sora I, Uhl GH, Mignot E, Edgar DM (2001) Dopaminergic role in stimulant-induced wakefulness. J Neurosci 21:1787–1794. Wu M-F, John J, Maidment N, Lam HA, Siegel JM (2002) Hypocretin release in normal and narcoleptic dogs after food and sleep deprivation, eating, and movement. Am J Physiol Integr Comp Physiol 283:R1079 –R1086. Zhang JH, Sampogna S, Morales FR, Chase MH (2002) Co-localization of hypocretin-1 and hypocretin-2 in the cat hypothalamus and brainstem. Peptides 23:1479 –1483.
(Accepted 9 June 2003)