Hypocretin/Orexin Peptides Excite Rat Neuroendocrine Dopamine Neurons through Orexin 2 Receptor-Mediated Activation of a Mixed Cation Current

Hypocretin/Orexin (H/O) neurons of the lateral hypothalamus are compelling modulator candidates for the chronobiology of neuroendocrine output and, as a consequence, hormone release from the anterior pituitary. Here we investigate the effects of H/O peptides upon tuberoinfundibular dopamine (TIDA) neurons – cells which control, via inhibition, the pituitary secretion of prolactin. In whole cell recordings performed in male rat hypothalamic slices, application of H/O-A, as well as H/O-B, excited oscillating TIDA neurons, inducing a reversible depolarising switch from phasic to tonic discharge. The H/O-induced inward current underpinning this effect was post-synaptic (as it endured in the presence of tetrodotoxin), appeared to be carried by a Na+-dependent transient receptor potential-like channel (as it was blocked by 2-APB and was diminished by removal of extracellular Na+), and was a consequence of OX2R receptor activation (as it was blocked by the OX2R receptor antagonist TCS OX2 29, but not the OX1R receptor antagonist SB 334867). Application of the hormone, melatonin, failed to alter TIDA membrane potential or oscillatory activity. This first description of the electrophysiological effects of H/Os upon the TIDA network identifies cellular mechanisms that may contribute to the circadian rhythmicity of prolactin secretion.


Whole-Cell Recordings.
For electrophysiological experiments rats (n = 27) were deeply anaesthetized with sodium pentobarbital and decapitated. The brain was rapidly removed and placed in an ice-cold and oxygenated (95%O 2 /5%CO 2 ) 'slicing' solution containing (in mM) sucrose (214), KCl (2.0), NaH 2 PO 4 (1.2), NaHCO 3 (26), MgSO 4 (1.3), CaCl 2 (2.4), D-glucose (10). The meninges were gently removed, and the brain was blocked and glued to a vibratome (Leica) where 250 μ m thick coronal slices of the hypothalamus containing the Arc were prepared. Slices were immediately transferred to artificial cerebrospinal fluid (aCSF) containing (in mM) NaCl (127), KCl (2.0), NaH 2 PO 4 (1.2), NaHCO 3 (26), MgCl 2 (1.3), CaCl 2 (2.4), D-glucose (10), in a continuously oxygenated holding chamber at 35 °C for a period of 25 min. Subsequently, slices were allowed to recover in 'recording' solution at room temperature for a minimum of 1 h before recording. For whole-cell recordings, slices were transferred to a submerged chamber and placed on an elevated grid that allows perfusion both above and below the slice. An Axioskop 2 FS Plus upright microscope (Carl Zeiss) was used for infrared -differential interference contrast visualization of cells. Recordings were performed at room temperature (22 °C) and slices were continuously perfused with oxygenated 'recording' solution (as above) at a rate of ca. 5 ml/min, unless otherwise described. All pharmacological compounds were bath applied.
Whole cell current-and voltage-clamp recordings were performed with pipettes (3-7 MΩ when filled with intracellular solution) made from borosilicate glass capillaries (World Precision Instruments) pulled on a P-97 Flaming/Brown micropipette puller (Sutter). The intracellular recording solution contained (in mM) K-gluconate (140), KCl (10), HEPES (10), EGTA (1), Na 2 ATP (2), pH 7.3 (with KOH). The concentration used for tetrodotoxin (TTX) was 500 nM. In voltage clamp experiments, unless otherwise stated, the holding potential was − 60 mV. In experiments that required the inhibition of Na + -dependent conductances a 'Zero Na + ' extracellular solution was prepared where NaCl was substituted with an equimolar concentration of TRIS-HCl. Recordings were performed using a Multiclamp 700B amplifier and pClamp9 software (Molecular Devices). Slow and fast capacitative components were automatically compensated for. Access resistance was monitored throughout the experiments, and neurons in which the series resistance was > 25 MΩ or changed > 15% were excluded from the statistics. Liquid junction potential was 16.4 mV and not compensated. The recorded current was sampled at 10 kHz and filtered at 2 kHz unless otherwise stated.
Immunofluorescence and image analysis. Male 6-10 weeks old Sprague Dawley rats (n = 5), were deeply anesthetized and perfused with a fixative containing 4% paraformaldehyde and 0.2% picric acid according to Zamboni and deMartino 31 . Fourteen μ m thick sections spanning the entire rostro-caudal length of the Arc were processed for immunofluorescence using the tyramide signal amplification (TSA) Plus protocol (Perkin Elmer) as previously described 32 . Briefly, sections were incubated with primary anti-tyrosine hydroxylase (TH) immunoglobulin (1:1,000; mouse monoclonal; MAB318; Millipore) combined with anti-H/O antiserum (1:10,000; raised in rabbit; gift from Drs. E. Mignot and K. Eriksson) at 4 °C for 16 hours. Sections were then pre-incubated with TNB blocking reagent as supplied with TSA Plus kit (Perkin Elmer) for 30 min, and incubated for 60 min with horseradish-peroxidase-conjugated swine anti-rabbit immunoglobulin (1:500 in TNB buffer; Dako) mixed with donkey anti-mouse antiserum conjugated with Alexa-594 (Life Technologies). The sections were next incubated for 10 min with tyramide-conjugated fluorescein (1:500) and subsequently mounted with anti-fading agent (2.5% DABCO, Sigma) diluted in glycerol. Three tissue sections per animal, spaced at regular intervals throughout the Arc were investigated with confocal microscopy using an Olympus FV1000 microscope (Tokyo, Japan). For each brain section and hemisphere, high magnification confocal stacks were acquired from the dm Arc. Image stacks were processed and analyzed with BitPlane Imaris software. Close appositions were assigned manually by investigating the stack's 3D projections, and confirmed by investigating single optical sections in the xy-, xz-and yz-planes.
Statistical Analysis and Reagents. Data analysis was performed with OriginPro8.5 (OriginLab) and Clampfit 9 (Molecular Devices) software. Statistical significance was set at P < 0.05 and determined using the stated statistical test (*P < 0.05, **P < 0.01, ***P < 0.001, ns = non-significant). To determine the H/O-A induced current in the presence of TTX frequency distribution histograms were plotted for ten Sec of control, H/O-A and wash traces, with Gaussian fits performed on the individual and pooled distributions. The mean difference between the Gaussian peaks for control and H/O-A was used as the value for the H/O-A induced current.
All reagents were purchased from Sigma Pharmaceuticals with the exception of TTX which was purchased from Alomone Labs, H/O-A and -B which were purchased from Bachem and SB 334867, Ala 11 D-Leu 15 -orexin B and TCS OX2 29 which were purchased from Tocris Bioscience.

TIDA neurons are innervated by H/O-immunoreactive fibers.
The H/O neurons of the lateral hypothalamus project widely throughout the brain, innervating numerous structures -including the Arc 17,33,34 . It remains unknown, however, to what extent this innervation includes the dmArc and TIDA neurons. Double-label immunofluorescence staining of H/O and tyrosine hydroxylase (TH) -the rate-limiting enzyme in monoamine biosynthesis -was performed to address this issue (Fig. 1). Hypocretin/Orexin-immunoreactive (-ir) fibers were observed in all three sectors of the Arc (the dorsomedial, ventromedial and ventrolateral 35 ). Only scattered axon terminals were seen in the median eminence and then exclusively in the internal layer, separate from the TH-ir terminals in the external layer. TH-ir neurons clustered in the dmArc were observed to form close appositions on both cell somata and proximal dendrites. Such contacts could be seen throughout the rostrocaudal extension of the dmArc. Individual appositions were verified in confocal microscope images in x-, y-, and z-planes to ensure that they did not simply represent superimpositions upon TH-ir elements in the coronal plane. Thus, there appears to exist an anatomical substrate for direct H/O-TIDA interactions. We next explored this possibility using whole-cell patch clamp recordings.
Hypocretin/orexin A and B excite oscillating TIDA neurons replacing phasic discharge with tonic firing via a post synaptic mechanism. Earlier work has demonstrated that in the rat, TIDA neurons can be reliably identified in vitro on the basis of their electrophysiological properties 2,25,26 . These features include a robust membrane potential oscillation that is synchronised between neurons, such that TIDA cells alternate rhythmically between depolarised UP states crowned by action potentials and hyperpolarised DOWN states during which discharge is absent ( Fig. 2A).
Bath application (90-120 seconds) of H/O-A (200 nM) consistently resulted in the depolarisation of TIDA neurons and the replacement of phasic discharge with tonic firing ( Fig. 2A; n = 12/12; 100%), transient effects readily reversible upon wash of drug from the recording chamber. To evaluate the possibility that these effects are mediated by direct, postsynaptic actions of H/O-A on TIDA neurons, action potential-dependent pre-synaptic influences were blocked by bath application of the voltage-gated Na + channel antagonist TTX (500 nM), a compound that also abolishes the TIDA oscillation, likely by its inhibitory effect on persistent Na + currents 25    In a proportion (n = 3/9; 33%) of TIDA neurons application of H/O-A, in addition to the mixed cationic conductance, also generated a slow onset, TTX-insensitive oscillation (Fig. 4A). This rhythmical activity was characterised by a frequency of 0.3 ± 0.06 Hz (n = 3) and an amplitude of 6.2 ± 1.1pA; (n = 3), and manifested itself following the peak H/O-A induced response. (This H/O-generated, TTX-insensitive oscillation is significantly faster and of lower amplitude than the default, TTX-sensitive TIDA oscillation recorded at the same temperature (frequency: 0.05 ± 0.002 Hz; amplitude: 22.7 ± 2.9pA; n = 9)). The fast oscillation appeared to be mediated by an L-type Ca + current as it was abolished by the application of the L-type antagonist, nimodipine (20 μ M; Fig. 4B; n = 3/3).

Discussion
In this study we investigated the effects of H/O peptides upon the electrical activity of TIDA neurons. We found that both H/O-A and -B excite TIDA neurons by inducing a depolarising switch from phasic discharge to tonic firing, a response that was frequently associated with a TTX-insensitive L-type Ca 2+ channel-dependent oscillation. While earlier work has shown that centrally administered H/O can potently lower plasma Prl levels [22][23][24] , and has suggested that this may be relayed in part (but not fully) through dopaminergic actions 23 , the site of action and mechanism for this neuroendocrine modulation has remained elusive. As neuroendocrine dopamine exerts a powerful inhibitory influence on pituitary lactotrophs (see ref. Neuronal excitation as a consequence of H/O receptor activation has been shown to occur in many neuronal populations and to be underpinned by an assortment of electrophysiological mechanisms. For example, H/Os cause excitation via closure of a K + conductance in thalamic neurons 46,47 , sublayer 6b cortical neurons 48 and layer 2-3 pyramidal cells 36 . In histaminergic neurons of the tuberomammillary nucleus 49 and both pro-opiomelanocortin 50 and GABA-ergic neurons of the Arc 51 H/O causes excitation via the activation of a Na + /Ca 2+ exchanger. In TIDA neurons the H/O effect is here shown to be underpinned by the activation of a mixed cation current, a mechanism shared with H/O-induced excitation in serotonin neurons of the dorsal raphe 52,53 , area postrema neurons 54 and cholinergic neurons of the laterodorsal tegmentum 55,56 . In the latter example the H/O response islike the TIDA effect -associated with the augmentation of L-type Ca 2+ currents. Importantly, in addition to H/O, other Prl-inhibiting factors -such as TRH 25 (when acting centrally) and oxytocin 27 , as well as Prl itself 26 -induce excitation via the activation of a mixed cation current. It therefore appears that the modulation of these currents constitutes a key regulatory axis, or point of convergence, for factors controlling the output of the TIDA network.
Hormone secreting cells of the anterior pituitary are regulated primarily by tropic factors, substances secreted by hypothalamic neuroendocrine cells which then reach the pituitary via the portal circulation 57   neuroendocrine system that controls pituitary Prl release, underscores this contention, demonstrating that H/O neurons are an important regulatory node for hypothalamic-pituitary interaction.
The H/O-ergic system is a key element of the neurocircuitry orchestrating biorhythms 58 . Our data provide a compelling cellular and molecular mechanism by which H/O peptides can influence the daily rhythm of Prl secretion, which is lowest during waking when release of H/Os are at their peak 59,60 and their positive influence on TIDA activity and dopamine secretion (which exhibits a similar circadian pattern 14 ) may be at its greatest. Furthermore, our data demonstrate that this influence occurs post-synaptically, with H/Os directly exciting TIDA neurons in a cell-autonomous fashion. This H/O induced excitation of TIDA neurons is of particular interest as previous data has suggested that part of H/O's effect upon Prl secretion may be independent of hypothalamic dopamine, acting via the NPYergic system 23,61 .
The direct, postsynaptic effects of H/O on TIDA neurons shown here may be a physiological correlate of the heavy innervation of the Arc -where TIDA neurons are located 62 -by H/O-immunoreactive fibers 17,33 . Indeed, we demonstrate that H/O-ir terminals form close appositions onto TIDA cell bodies and proximal dendrites, indicative of synaptic contacts. This evidence, in conjunction with our electrophysiological data, suggest that the powerful Prl-releasing actions of centrally administered H/O 22 are mediated through the TIDA system. As H/O fibers are present in the median eminence and the OX1R receptor is abundantly expressed in (as yet unidentified) pituitary cells 63,64 , parallel actions at the level of the pituitary cannot be excluded. Despite this, conclusive evidence for a pituitary effect of H/O on Prl secretion is lacking; in the rat no such actions are seen 65 ; while in the sheep, only modest, seasonally dependent effects have been shown 66 . Notably, in our immunofluorescence stainings H/O-ir terminals in the median eminence were few in number and substantially lower in density compared to those in the dmArc.
In addition to H/O, the pineal hormone melatonin is another chronobiological factor thought to regulate pituitary Prl secretion. Previous research has indicated that melatonin's Prl-releasing effects may be independent of dopamine neurosecretion [67][68][69] , yet whether melatonin directly modulates the electrical properties of TIDA neurons has not been addressed. We found that melatonin application failed to alter either the oscillatory activity or the membrane potential of TIDA neurons. This lack of response further supports the notion that melatonin exerts its influence upon Prl secretion at a site other than the TIDA network, most likely through the pituitary pars tuberalis (see ref. 45).
Beyond the potential implications for our understanding of the physiological nychthemeral release of Prl, our results are interesting in the context of clinical reports indicating that some narcolepsy patients -who show markedly reduced numbers of H/Oergic cells 18,19 -exhibit elevated serum Prl 70,71 . In this scenario the absence of a wakefulness-correlated stimulatory influence (i.e. H/O) on TIDA neurons could contribute to relieving pituitary lactotrophs from their dopaminergic "brake" resulting in the hyperprolactinaemia observed. It should be noted, however, that other investigators have failed to observe changes in circulating Prl in narcoleptic patients [72][73][74] .
While this clinical question remains to be resolved, the present data support a role for the hypothalamic H/Oergic system in the control of neuroendocrine and anterior pituitary function. The excitation of TIDA neurons shown here may be a part of the explanation for the circadian secretion pattern of Prl. Future investigations will determine how H/O input coordinates with other candidate circadian influences on TIDA neurons, including the suprachiasmatic nucleus 13 and the expression of clock genes in the dopamine neurons themselves 14 .