INTRODUCTION

Some atypical antipsychotic and other drugs effective in schizophrenia, depression, obsessive–compulsive disorder, and neurodegenerative disease are considered to act at serotonin (5-hydroxytryptamine) 5-HT2 receptors (Barnes and Sharp, 1999; Breier, 1995; Ramasubbu et al, 2000; Rauser et al, 2001; Reynolds, 2001; Thase, 2002). In brain, 5-HT2 receptors can be coupled via G-proteins to phospholipase C (PLC) activation, generating inositol 1,4,5-trisphosphate (IP3) and diacylglycerol as second messengers (Conn and Sanders-Bush, 1986b; Edagawa et al, 2000), or to phospholipase A2 (PLA2) activation, releasing arachidonic acid (AA) from phospholipids (Axelrod, 1995; Berg et al, 1998; Felder et al, 1990; Tournois et al, 1998). Both AA and its eicosanoid metabolites are important second messengers (Shimizu and Wolfe, 1990).

A 5-HT2 syndrome has been described in rats following administration of the 5-HT2A/2C receptor agonist (±)2,5-dimethoxy-4-iodophenyl-2-aminopropane (DOI) (Glennon, 1986; Johnson et al, 1998; Pranzatelli, 1990; Wettstein et al, 1999). The syndrome is characterized by head and body shakes, ear scratching, skin jerks, and forepaw tapping. It is maximal in response to 3.0 mg/kg intraperitoneal (i.p.) DOI, and 2.5 mg/kg i.p. DOI has been used widely in behavioral and biochemical studies of the syndrome. Additionally, 2.5 mg/kg i.p. DOI in rats markedly stimulates the release of corticotropin (ACTH), corticosterone, oxytocin, renin, and prolactin, and activates hypothalamic corticotropin-releasing factor and oxytocin-expressing neurons (Van de Kar et al, 2001). DOI also induces hyperthermia in rats (Mazzola-Pomietto et al, 1997).

Despite the marked behavioral and neuroendocrine effects of 2.5 mg/kg DOI, the regional cerebral metabolic rate for glucose (rCMRglc), a marker of neuronal activity measured with intravenous [14C]2-deoxy-D-glucose, was minimally affected in unanesthetized rats given this dose of DOI (Freo et al, 1991). Of 75 brain regions examined using quantitative autoradiography, this dose of DOI reduced rCMRglc significantly in layer IV of the pyriform cortex, the ventral CA3 region of the hippocampus, the cortical nucleus of the amygdala, and the olfactory tubercle. The reductions were ascribed to inhibition by DOI of neuronal spike activity (Ashby et al, 1990; Bloom, 1985; Cooper et al, 1996), to which rCMRglc is said to be coupled (Sokoloff, 1999). In another study, adrenalectomy or pretreatment with metyrapone (an inhibitor of 11-β-hydroxylase, a rate-limiting enzyme in corticosterone syntheses) abolished rCMRglc declines in the dorsal CA1, CA2 and CA3 regions of the hippocampus in response to 10 mg/kg i.p. DOI, suggesting to the authors that hippocampal activity can be modulated by the hypothalamic–pituitary–adrenal axis (Freo et al, 1992).

It is not evident why 2.5 mg/kg i.p. DOI produces the marked behavioral activation of the 5-HT2 syndrome, while at the same time causing decrements in rCMRglc. We thought that this discrepancy might be clarified if we could examine postsynaptic signal transduction in vivo, secondary to 5-HT2 receptor occupancy by DOI. As noted above, such signaling can occur through activation of PLC or PLA2. No method currently exists to image brain PLC activation in vivo, whereas PLA2 activation can be imaged by using quantitative autoradiography to measure incorporation into brain of intravenously injected, radiolabeled AA (DeGeorge et al, 1991; Hayakawa et al, 2001; Nariai et al, 1991; Rapoport, 2001; Robinson et al, 1992). We thought that we would use this latter method. Tracer incorporation in response to an appropriate agonist reflects PLA2-mediated hydrolysis of unlabeled AA from the stereospecifically numbered (sn)-2 position of synaptic brain phospholipids (DeGeorge et al, 1991; Fonlupt et al, 1994; Grange et al, 1998; Jones et al, 1996), independent of changes in regional cerebral blood flow (rCBF) (Chang et al, 1997; DeGeorge et al, 1991; Robinson et al, 1992; Robinson and Rapoport, 1986; Yamazaki et al, 1994). Receptors coupled to PLA2 via membrane G-proteins include cholinergic muscarinic M1 and M3 receptors, dopaminergic D2 receptors, and serotonergic 5-HT2 receptors (Axelrod, 1995; Bayon et al, 1997; Cooper et al, 1996; DeGeorge et al, 1991; Felder et al, 1990; Hayakawa et al, 2001; Vial and Piomelli, 1995). PLA2 can also be activated when Ca2+ enters cells by glutamate acting at N-methyl-D-aspartate (NMDA) receptors or by acetylcholine acting at nicotinic receptors (Brooks et al, 1989; Cooper et al, 1996; Vijayaraghavan et al, 1995).

In the present study, we injected tritiated AA ([3H]AA) intravenously in unanesthetized rats and used quantitative autoradiography to determine regional brain incorporation coefficients k* of the tracer in response to 2.5 mg/kg i.p. DOI. The racemic DOI commonly is used to study effects of in vivo 5-HT2A/2C receptor activation. Both stereoisomers bind with equivalent affinities to 5-HT2A/2C receptors, although (−)DOI is twice as potent as (+)DOI in inducing head twitches in mice (Glennon, 1986,1987; PDSP Drug Database, 2000; Pranzatelli, 1990; Roth et al, 2000).

We also quantified k* for [3H]AA in response to chronically administered mianserin, an atypical tetracyclic antidepressant that has been used as a 5-HT2-receptor antagonist in many animal studies, although having some adrenergic α2-antagonist activity as well (Anji et al, 2000; Ashby et al, 1990; Blackshear and Sanders-Bush, 1982; Dijcks et al, 1991; Hoyer et al, 1995; PDSP, 2000; Pranzatelli, 1990; Rocha et al, 1994; Roth and Ciaranello, 1991; Roth et al, 2000; Sanders-Bush et al, 1987; Schreiber et al, 1995). Finally, we measured k* in response to 2.5 mg/kg i.p. DOI 24 h after mianserin administration (Arvidsson et al, 1986; Berendsen and Broekkamp, 1991; Sanders-Bush et al, 1987), by which time mianserin is known to be largely washed out from the brain (Dijcks et al, 1991; Sanders-Bush et al, 1987).

An abstract of part of this work has been published (Qu et al, 2001).

MATERIALS AND METHODS

Chemicals

Radiolabeled [5,6,8,9,11,12,14,15-3H]AA ([3H]AA) at a specific activity of 200 Ci/mmol was purchased from Moravek Biochemicals (Brea, CA). Radiochemical purity by thin-layer chromatography always exceeded 96%. Mianserin and DOI were purchased from Sigma-Research Biochemicals International (Natick, MA). Pentobarbital sodium was purchased from Richmond Veterinary Supply Co. (Richmond, VA).

Animals

Male Fischer-344 rats (Charles River Laboratories, Wilmington, MA), 12 weeks old and weighing 290–320 g, were housed under standard laboratory conditions under a 12-h light/12-h dark cycle, with ready access to standard laboratory chow and water. The experimental protocol was approved by the National Institute of Child Health and Human Development Animal Care and Use Committee and conformed to the Guide for the Care and Use of Laboratory Animals (National Institute of Health Publication 86-23).

Arterial and Venous Catheterization

Rats were placed in four experimental groups of 10 animals each: (1) controls; (2) rats given 2.5 mg/kg i.p. DOI acutely; (3) rats administered 10 mg/kg i.p. mianserin daily for 14 days, then not given mianserin for 24 h; (4) rats administered 10 mg/kg i.p. mianserin daily for 14 days, then not given mianserin for 24 h, and then given 2.5 mg/kg i.p. DOI.

The in vivo fatty acid method has been described elsewhere (DeGeorge et al, 1991; Hayakawa et al, 2001). Briefly, rats in each of the four groups were anesthetized with halothane (1–3% v/v in O2). PE 50 polyethylene catheters (Clay Adams, Lincolnshire, IL) filled with heparinized saline (100 IU/ml) were surgically implanted into a femoral artery and vein, after which the incision site was infiltrated with a local anesthetic (lidocaine) and closed with wound clips. The rats were wrapped loosely in a fast-setting plaster cast, secured to a wooden block with the upper body free, and allowed to recover from anesthesia in a temperature-controlled and sound-dampened box for 4 h. Body temperature was kept at 36–37°C by means of a rectal thermometer and a feedback heating device.

Drug Administration and Tracer Infusion

After the rat recovered from anesthesia for 4 h, 125 μl arterial blood was withdrawn to measure pH, pO2, and pCO2. Rats (8–10 per group) were administered either saline (control) or 2.5 mg/kg i.p. DOI. After 20 min, 1.75 mCi/kg [3H]AA in 2 ml of 5 mM HEPES buffer, pH 7.4, containing 50 mg/ml fatty-acid-free bovine serum, was infused through the venous cannula with an infusion pump (Harvard Instrument Co., Holliston, MA) at a rate of 400 μl/min for 5 min.Timed 125-μl arterial blood samples were collected from the beginning of infusion to 20 min, when the rats were killed with 65 mg i.v. sodium pentobarbital. Brains were removed and frozen in 2-methylbutane at −50°C for subsequent autoradiography. Plasma was separated from arterial blood by centrifugation, and lipids were extracted using the method of Folch et al (1957). Radioactivity in the organic fraction was measured by liquid scintillation spectroscopy.

Autoradiography and Calculations

Frozen brains were sectioned on a cryostat at −20°C. Sets of three adjacent 20-μm sections were collected and mounted on glass coverslips at 140-μm coronal intervals and dried. The sections were exposed together with [3H]methylmethacrylate autoradiographic standards (Amersham, Arlington Heights, IL) to [3H]hyperfilm (Amersham) for 15–18 weeks and then developed following the manufacturer's instructions. One of the three adjacent sections was collected and stained with cresyl violet to identify brain regions with reference to a rat-brain atlas (Paxinos and Watson, 1987).

Regional brain radioactivity was measured in sextuplicate by quantitative densitometry using the public domain image analysis program NIH Image (version 1.62) created by Wayne Rasband (National Institutes of Health, Bethesda, MD), installed on a Macintosh computer (Apple Computer, Cupertino, CA). Regional brain incorporation coefficients k* were calculated as

where k* is in units of ml/(s g); c brain * (20 min) is the brain radioactivity at 20 min in nCi/g, c plasma * is the plasma fatty acid radioactivity in nCi/ml, and t is time after onset of [3H]AA infusion.

Data were compared using Prism software for the Macintosh (Abacus Concepts, Berkeley, CA) and are reported as means ±SEM. A one-way ANOVA and Dunnett's (Dunnett, 1964) multiple comparison test were used to evaluate statistical significance between experimental and control means; p<0.05 was taken as statistically significant.

RESULTS

Table 1 summarizes mean physiological parameters in unanesthetized control rats and in rats treated chronically with mianserin. These values are similar to published values.

Table 1 Physiological Parameters of Rats after Surgery
Full size table

As illustrated in Figure 1, coronal autoradiographs showed widespread increments in k* (brain radioactivity divided by integrated plasma radioactivity; Equation (1)) for [3H]AA after 2.5 mg/kg i.p. DOI, compared with k* from control rats. The largest increments were in motor and somatosensory cortical areas.

Figure 1
figure 1

[3H]arachidonate incorporation coefficients k* in coronal sections from brain of (a) control rat and (b) rat given DOI (2.5 mg/kg i.p.); k* is color-coded.

Full size image

Mean regional [3H]AA incorporation coefficients (k*) in saline-treated control rats are presented in the first data column of Table 2 The values are comparable to previously published control values (DeGeorge et al, 1991; Hayakawa et al, 2001). Notable is the 6- to 10-fold greater k* at the choroid plexus than in the brain parenchyma.

Table 2 Regional [3H]Arachidonate Incorporation Coefficients k* (ml/(s g) brain×104) in Rat Brain under Different Conditions
Full size table

Compared with controls, 2.5 mg/kg i.p. DOI produced widespread and statistically significant increments in k* for [3H]AA, of the order of 60%, in many brain regions (second data column of Table 2), but particularly in the neocortex.

After 14 days of mianserin administration, and allowing 24 h for mianserin to be washed out from the brain (Dijcks et al, 1991; Sanders-Bush et al, 1987), there was no significant difference in mean k* for [3H]AA in any brain region compared with the respective k* in control animals (third data column of Table 2). Furthermore, when DOI was administered after 2 weeks of mianserin after allowing for washout (fourth data column of Table 2), no statistically significant difference in mean k* was found in any brain region or in the choroid plexus, compared with the respective control mean. Thus, chronic mianserin completely blocked all DOI-induced increments in [3H]AA incorporation.

DISCUSSION

The 5-HT2A/2C receptor agonist DOI, at a dose of 2.5 mg/kg i.p., caused widespread and large (as high as 60%) increments in k* for [3H]AA in brains of unanesthetized adult rats. These increments are consistent with the reported marked behavioral (5-HT2 syndrome) and neuroendocrine responses provoked by this dose (Johnson et al, 1998; Pranzatelli, 1990; Van de Kar et al, 2001; Wettstein et al, 1999). The increments in k* could be completely blocked by chronic pretreatment with mianserin, a 5-HT2 receptor agonist that has been reported to block the 5-HT2 syndrome and the hyperthermia produced by DOI (Berendsen and Broekkamp, 1991; Mazzola-Pomietto et al, 1997).

The interpretation that k* for [3H]AA reflects regional PLA2 activation derives from experimental observations that k* is independent of rCBF, that incorporation of labeled AA from blood into brain phospholipids is very rapid, and that k* reflects brain PLA2 but not PLC activity (Rapoport, 2001; Rapoport et al, 2001; Robinson et al, 1992; Washizaki et al, 1991). That k* is independent of rCBF is evident from several observations. As shown in Table 2, k* for [3H]AA was markedly elevated in response to 2.5 mg/kg DOI (Table 2), despite evidence that rCMRglc, to which rCBF is coupled (Reivich, 1974), declines or does not change with this dose (Freo et al, 1991,1992). Likewise, administration to rats of arecoline, a cholinergic agonist that acts at muscarinic M1 receptors coupled to PLA2, increased rCMRglc and rCBF (as well as k*) for labeled AA, without affecting k* for labeled palmitic acid (DeGeorge et al, 1991; Jones et al, 1996; Maiese et al, 1994; Soncrant et al, 1985). Thus, fatty acid uptake by the brain is not increased by increased rCBF per se. Finally, values for k* for both labeled palmitate and arachidonate were shown to be unaffected by two-fold increments in rCBF induced by hypercapnia in rats and monkeys (Chang et al, 1997; Yamazaki et al, 1994).

The independence of k* from rCBF arises because circulating plasma albumin, to which fatty acid is highly bound but from which it can rapidly dissociate (Svenson et al, 1974), acts as an ‘infinite source’ of intravascular tracer for entry into brain (Robinson et al, 1992; Robinson and Rapoport, 1986; Washizaki et al, 1991). As blood passes through the brain, unesterified unbound labeled fatty acid is rapidly extracted and replaced by fatty acid released from albumin. About 5% of a plasma fatty acid is stripped from albumin as blood passes through the brain (Pardridge and Mietus, 1980).

Within 2 min after entering rat brain following its intravenous injection, 90% of radiolabeled AA has been incorporated into ‘stable’ brain lipids, largely into the sn-2 position of phospholipids. The remainder, found in the aqueous fraction, represents metabolites arising from comparatively slow β-oxidation (Osmundsen and Hovik, 1988). The rate of disappearance of labeled AA from brain phospholipids is only 10% per hour (DeGeorge et al, 1989; Rapoport, 2001; Rapoport et al, 2001; Washizaki et al, 1994), which means that we can image tracer incorporation at 20 min without worrying about loss from the phospholipids. Finally, inhibiting brain PLA2 activity in vivo by drug produces a proportional reduction in k* for [3H]AA (Grange et al, 1998).

Chronically administered mianserin had no effect on baseline values of k* for [3H]AA, but prevented DOI-initiated increments in k* (Table 2). The 5-HT2 receptor-mediated activation of PLC by DOI, which increases phosphatidylinositol turnover and Ca2+ mobilization by IP3, is also reported to be inhibited by chronic mianserin (Conn and Sanders-Bush, 1986b; Wolf and Schutz, 1997). Inhibition of signaling in both cases is probably due to mianserin-induced neuroplastic changes, rather than to physical blocking of 5-HT2 receptors by mianserin, as the brain mianserin concentration falls to less than 0.1% of its peak concentration within 24 h after i.p. injection (Dijcks et al, 1991; Sanders-Bush et al, 1987). Chronic mianserin is reported not to alter extracellular serotonin levels in rat brain (Kreiss and Lucki, 1995), but is reported to reduce brain densities of 5-HT2A receptors (Berendsen and Broekkamp, 1991; Blackshear and Sanders-Bush, 1982; Essom and Nemeroff, 1996; Frazer et al, 1988; Roth and Ciaranello, 1991) and 5-HT2C receptors (Rocha et al, 1994). Phosphorylation and interaction of the receptors with membrane G-proteins are altered (Hartman and Northup, 1996; Ozawa et al, 1994; Westphal et al, 1995), and both receptor types are functionally hyposensitive (Mazzola-Pomietto et al, 1997). Chronic mianserin, on the other hand, produces a supersensitivity of adrenergic α2 receptors (Pinder, 1985).

Head twitches of the 5-HT2 syndrome appear to be related more to 5-HT2A than to 5-HT2C initiated signaling (Schreiber et al, 1995); thus, a selective 5-HT2C antagonist (SB 200,646A) did not inhibit the twitches. Additionally, dopaminergic D1 antagonists as well as agonists to α1 and α2 adrenoreceptors and to 5-HT1A receptors reduced DOI-induced head twitches, suggesting a role for nonserotonergic mechanisms (Schreiber et al, 1995). A full 5-HT syndrome has been described in humans, with some components perhaps related to the 5-HT2 syndrome in rodents. The clinical syndrome occurs with excess serotonergic therapy and can be exacerbated by coadministration of a monoamine oxidase inhibitor. Its features include an altered mental status, restlessness, myoclonus, hyperreflexia, diaphoresis, shivering, and tremor (Mills, 1997; Sternbach, 1991); it is treated by discontinuing serotonergic therapy.

The robust increments in k* induced by 2.5 mg/kg i.p. DOI are accompanied by a few reductions in rCMRglc (Freo et al, 1991,1992), which are ascribed to reduced neuronal spike activity (Ashby et al, 1990; Bloom, 1985; Cooper et al, 1996; Freo et al, 1991; Sokoloff, 1999). rCMRglc is a weighted average, reflecting energy consumption by many brain processes, and PLA2-initiated AA release and reincorporation consume only a small fraction of net brain adenosine triphosphate (ATP) consumption (Purdon and Rapoport, 1998; Rapoport, 2001). Large changes in [3H]AA incorporation into the brain in contrast to small changes in rCMRglc have also been noted in rats administered the muscarinic agonist arecoline or the dopaminergic D2 agonist quinpirol (Hayakawa et al, 2001; Nariai et al, 1991; Orzi et al, 1988; Wooten and Collins, 1981).

Recall that PLA2 can be activated when a ligand binds to any of a number of receptor subtypes, including 5-HT2 receptors (Axelrod, 1995; Cooper et al, 1996; Felder et al, 1990). 5-HT2 receptors are widely distributed in rat brain (Appel et al, 1990; Morilak et al, 1993; Pazos and Palacios, 1985). High densities are reported in cerebral cortex, olfactory and pyriform cortex, nucleus accumbens, caudate-putamen body and tail, dentate gyrus of hippocampus, and medial mammillary nucleus of hypothalamus. In the neocortex, highest densities are in layer IV. Frontal and motor cortical regions have higher densities than other cortical regions, whereas densities are comparatively low in the caudate-putamen head, globus pallidus, red nuclei, septal nuclei, and most parts of the hippocampus (McKenna et al, 1989; McKenna and Peroutka, 1989), thalamus, hypothalamus, midbrain, brain stem, and spinal cord.

The most intense increments in k* in response to DOI (Figure 1, Table 1) were seen in brain regions having high densities of 5-HT2A compared with 5-HT2C binding sites (Cooper et al, 1996; Li et al, 2001; Pazos and Palacios, 1985). High densities of 5-HT2A binding sites are found in neocortical areas (layer IV), amygdala and midbrain (lateral amygdaloid nucleus, medial amygdaloid nucleus), and CA1 and CA3 regions of the hippocampus, with lesser densities in the caudate-putamen. There are fewer 5-HT2C binding sites in rodent brain; they are found in the hypothalamus, amygdala and hippocampus, but minimally in the neocortex (except for the temporal horn). The correspondence between the brain distributions of PLA2 activation by DOI and of 5-HT2A receptors could be further examined using DOI and altanserin, a selective 5-HT2A blocker (Hoyer et al, 1995; Leysen et al, 1988; PDSP Drug Database, 2000; Roth et al, 2000), or by studying DOI responses in a 5-HT2A-receptor knockout mouse (Lira et al, 2001).

The choroid plexus has more than four-fold higher densities of 5-HT2C binding sites than brain parenchymal regions as well as high levels of 5-HT2C mRNA and 5-HT2A binding sites (Kaufman et al, 1995; Li et al, 2001). The six- to 10-fold greater control value for k* in the choroid plexus than in brain parenchymal regions, the marked increment in this k* in response to DOI, and the ability of chronic mianserin to block this increment suggest that serotonin-related PLA2 signaling plays an important role in choroid plexus function, particularly secretion of cerebrospinal fluid. On the other hand, serotonin is reported to decrease cerebrospinal fluid secretion by increasing phosphorylation of Na+, K+-ATPase by protein kinase C following activation of PLC (Conn and Sanders-Bush, 1986a; Fisone et al, 1995,1998), or through Ca2+-dependent activation of PLA2 (Kaufman et al, 1995). The uptake mechanism of [3H]AA into the choroid plexus may differ from that into brain, as the plexus, unlike the brain parenchyma, has a leaky vasculature that can allow protein-bound [3H]AA to access directly the choroid epithelium (Rapoport, 1976).

The DOI dose chosen in the present study may have been too large to identify PLA2 signaling solely at 5-HT2 receptors, because of downstream activation of dopaminergic D2 and other receptors also coupled to PLA2 (see the Introduction). A smaller dose may help in this regard. Although the affinities of both the (+) and (−) stereoisomers of DOI are reported to be equivalent at 5-HT2 receptors (Glennon, 1986,1987; PDSP Drug Database, 2000; Pranzatelli, 1990; Roth et al, 2000), as far as we know the affinities of the two stereoisomers at other receptors coupled to PLA2 have not been examined. DOI can increase amphetamine-induced dopamine release in the brain (Ichikawa and Meltzer, 1995), brain extracellular dopamine and noradrenaline concentrations (Gobert and Millan, 1999), and dopamine turnover (Gaggi et al, 1997), and DOI will activate local γ-aminobutyric acid (GABA) inputs to serotonergic neurons in the dorsal raphe nucleus (Liu et al, 2000). 5-HT2 receptor activation can also inhibit glutamate release from rat cerebellar mossy fibers (Marcoli et al, 2001) and the release of acetylcholine in the hippo-campus and neocortex (Feuerstein et al, 1996), which may explain DOI's inhibition of rCMRglc (Freo et al, 1991,1992).

In summary, our results suggest that labeled AA can be used to examine in vivo brain PLA2 signaling initiated by serotonergic drugs. Increments in k* for [3H]AA in response to DOI largely correspond to the distribution of 5-HT2A binding sites in the brain, although downstream receptors coupled to PLA2 are probably activated as well. Chronic mianserin, a 5-HT2 agonist known to inhibit the 5-HT2 syndrome, blocks [3H]AA incorporation completely in response to 2.5 mg/kg i.p. DOI. Imaging information gathered using labeled AA is clearly distinct from that using labeled 2-deoxy-D-glucose, and specific to PLA2 activation rather than to general brain functional activity. As a result of this, it might be worthwhile to extend the fatty acid method to examine PLA2 signaling in the human brain by means of positron emission tomography (Chang et al, 1997; Giovacchini et al, 2001).