- Email: [email protected]
Serotonergic dorsal raphe neurons from obese zucker rats are hyperexcitable
Neuroscience 120 (2003) 627– 634
SEROTONERGIC DORSAL RAPHE NEURONS FROM OBESE ZUCKER RATS ARE HYPEREXCITABLE P. OHLIGER-FRERKING,a*1 B. A. HORWITZa,b AND J. M. HOROWITZa,b
The ventromedial hypothalamus (VMH) has long been known to be an important mediator in the feeding pathways of humans (Blundell and Hill, 1987), cats (reviewed by Anokhin and Shuleikina, 1977), rats (reviewed by Je´quier and Tappy, 1999) and mice (Nonogaki et al., 1998). A variety of neurotransmitters affect activity of cells in the VMH. Among these is serotonin (5-HT) which modulates both the pattern of eating and the sensation of hunger/ satiety in humans (Blundell and Hill, 1987). In rats, 5-HT has been shown to be released at the VMH at the beginning of a meal, in anticipation of a meal, and during feeding (Schwartz et al., 1990; Orosco and Nicolaidis, 1992; Meguid et al., 2000). In addition, a large number of studies have shown that microinjection of 5-HT into the hypothalamic region reduces food intake (reviewed by Leibowitz and Alexander, 1998). Together, these studies indicate a modulatory role for 5-HT in the regulation of feeding behavior. The 5-HT released in the VMH during feeding derives from nerve terminals rather than from other potential sources such as glia, platelets, or mast cells (Auerbach et al., 1989), and the largest population of 5-HT-containing neurons in the brain that project to the VMH are found in the midbrain dorsal raphe nucleus (DRN; Jacobs et al., 1984). DRN activity increases during oral– buccal movements, such as occur during feeding (Fornal et al., 1996). Furthermore, electrical stimulation of the DRN and feeding behavior both result in increased 5-HT at the VMH (De Fanti et al., 2000; Orosco and Nicolaidis, 1992; Meguid et al., 2000). These studies provide further support for the view that DRN-to-VMH signals play a modulatory role in the neural control of food intake. Genetically obese Zucker rats as young as 16 days are hyperphagic (McLaughlin and Baile, 1981) and also present a number of other phenotypic changes related to obesity including decreased brown fat thermogenesis (Moore et al., 1985), increased white adipocyte size (Boulange et al., 1979), and hyperinsulinemia (Blonz et al., 1985). These rats lack a normally functioning leptin receptor (Chua et al., 1996) and, in addition, exhibit a number of abnormalities in the brain 5-HT system. Several studies of basal 5-HT levels or 5-HT turnover in the hypothalamus of young (ⱕ12 weeks) Zucker rats have been performed, but little consensus has been reached. Orosco et al. (1986) have suggested that resting 5-HT levels in the medial hypothalamus of lean and obese Zucker rats, revealed by chromatography, are not different. De Fanti et al. (2000), using microdialysis, also found that basal 5-HT levels were not different. However, Routh et al. (1994), in studies using tissue punches, found decreased 5-HT turnover in the
a Physiology Graduate Group, University of California, Davis, CA 95616, USA b
Section of Neurobiology, Physiology, and Behavior, University of California, Davis, CA 95616, USA
Abstract—Release of serotonin (5-HT) from dorsal raphe nucleus (DRN) neurons projecting to the ventromedial hypothalamus (VMH) has a modulatory effect on the neural pathway involved in feeding, hunger, and satiety. The obese Zucker rat, an animal model of genetic obesity, exhibits differences in serotonin signaling as well as a mutated leptin receptor. To evaluate possible mechanisms underlying this difference in serotonin signaling, we have compared electrophysiological responses of DRN neurons from 14- to 25-day-old male lean (Fa/Fa) and obese (fa/fa) Zucker rats using the whole-cell patch clamp technique on cells in brain slices from these animals. We found that the resting properties of these neurons are not different, but the DRN neurons from obese rats are hyperexcitable in response to current injection. This hyperexcitability is not accompanied by an increase in the depolarization caused by current injection or by changes in the threshold for spiking. However, the hyperexcitability is accompanied by reduction in the size and time course of the afterhyperpolarization (AHP) following an action potential. DRN neurons of obese rats recover from the AHP faster due to a smaller amplitude AHP and a faster time constant () of decay of the AHP. These deficits are not due to changes in the spike waveform, as the spike amplitude and duration do not differ between lean and obese animals. In summary, we provide evidence that serotonergic DRN neurons from obese Zucker rats are intrinsically hyperexcitable compared with those from lean rats. These results suggest a potential mechanism for the reported increase in 5-HT release at the VMH of obese rats during feeding, and provide the first direct evidence of changes in the intrinsic activity of serotonergic neurons, which are crucial regulators of feeding behavior, in a genetic model of obesity. © 2003 IBRO. Published by Elsevier Science Ltd. All rights reserved. Key words: afterhyperpolarization, hyperexcitability, serotonin, dorsal raphe nucleus, obesity, satiety. 1
Present address: Neurological Sciences Institute, Oregon Health and Science University, 505 Northwest 185th Avenue, Beaverton, OR 97006, USA. *Correspondence to: P. Ohliger-Frerking, Neurological Sciences Institute, Oregon Health and Science University, 505 Northwest 185th Avenue, Beaverton, OR 97006, USA. Tel: ⫹1-503-418-2661; fax: ⫹1-503-418-2501. E-mail address: [email protected] (P. Ohliger-Frerking). Abbreviations: ACSF, artificial cerebrospinal fluid; AHP, afterhyperpolarization; DRN, dorsal raphe nucleus; PCR, polymerase chain reaction; PE, phenylephrine; TTX, tetrodotoxin; VMH, ventromedial hypothalamus; 5-HT, serotonin; , time constant of decay.
0306-4522/03$30.00⫹0.00 © 2003 IBRO. Published by Elsevier Science Ltd. All rights reserved. doi:10.1016/S0306-4522(03)00360-9
627
628
P. Ohliger-Frerking et al. / Neuroscience 120 (2003) 627– 634
VMH of obese rats under resting conditions (1994). The reasons for this discrepancy are unclear, although it should be noted that microdialysis studies have poor spatial resolution and may have measured 5-HT levels outside of the VMH. Studies focusing on feeding-induced changes in hypothalamic 5-HT are in better agreement, consistently showing increased feeding-induced 5-HT release in the medial hypothalamus of obese versus lean Zucker rats (Lemierre et al., 1998; De Fanti et al., 2001). These results suggest that activity of serotonergic neurons in the DRN of obese Zucker rats may differ from those of lean rats, and, in fact, we have found previously that DRN neurons from obese rats exhibit stronger responses to exogenous adrenergic agonists (Ohliger-Frerking et al., 2002). In this study, we have investigated the electrophysiological properties of serotonergic DRN neurons in brain slices from lean (Fa/Fa) and obese (fa/fa) Zucker rats to determine if they differ in their intrinsic excitability.
EXPERIMENTAL PROCEDURES All experimental protocols in this work were reviewed and approved by the University of California Davis Animal Care and Use Committee, to minimize animal use and suffering, in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals (NIH 85-23, revised 1985).
Slice preparation Male Zucker rats were obtained from the colony maintained by the UC Davis Clinical Nutrition Research Unit. The pups were decapitated and their brains rapidly removed and placed in cold (4 °C) artificial cerebral spinal fluid (ACSF). All animals used were between 14 and 25 days old, an age range over which the (fa/fa) rats are not yet visibly obese; this was done so that any differences in cellular activity would be more likely to lie upstream of the onset of obesity, rather than be a consequence of obesity. ACSF containing (in mM) 125 NaCl, 2.5 KCl, 1 MgCl2, 1.25 NaH2PO4, 25 NaHCO3, 10 dextrose and 2 CaCl2, was aerated with 95% O2/5% CO2 at all times, and had a pH of 7.3 and an osmolarity of approximately 300 mOsm. The brain was trimmed and 250 m thick coronal midbrain slices were prepared using a vibratome filled with cold (4 °C) ACSF. Slices containing DRN were visually identified by the presence of the cerebral aqueduct and the medial longitudinal fasciculi. Three DRN-containing slices were typically obtained from each brain. Slices were transferred to a holding chamber filled with warm (35 °C) ACSF where they were allowed to recover for a minimum of 1 h prior to recording.
Whole-cell recording In preparation for each experiment, single DRN-containing slices were transferred to a recording chamber perfused with warm (35 °C) ACSF at 2 ml/min. Borosilicate glass microelectrodes (2.5– 6 M⍀) were filled with recording solution containing (in mM) 140 K-gluconate, 5 NaCl, 1 MgCl2, 0.3 EGTA, 10 HEPES, 3 MgATP and 0.2 NaGTP. The osmolarity of the internal solution was approximately 300 mOsm and the pH was adjusted to 7.4. Single cells in the DRN were visualized in the slice using a Nikon IR-DIC microscope (Technical Instruments, Burlingame, CA, USA) and an Olympus camera (Scientific Instrument Co., Sunnyvale, CA, USA). Whole-cell patch clamp recordings were made from each cell using the axoclamp 1C patch clamp amplifier in current clamp mode. Whole-cell voltages were filtered at 2 kHz and digitized at 5 kHz. Voltage records were stored directly on a computer and then analyzed off-line using Pclamp 6 software. Cell
capacitance, series resistance, and input resistance were monitored with a brief voltage step of 5 mV in voltage clamp mode. Series resistances were generally between 5 and 20 M⍀, and input resistances were generally between 400 and 800 M⍀. Resting membrane potential was measured either directly, if the cell was silent at rest, or as the voltage during occasional plateaus between spikes, if the cell was spontaneously active. 5-Hydroxytryptamine (5-HT) and tetrodotoxin (TTX) were purchased from Sigma (St. Louis, MO, USA). Both were dissolved in ACSF to the given concentration and bath applied via a drip system at 2 ml/min when assessing their effects on the cells. Serotonergic dorsal raphe neurons can be separated from non-serotonergic cells in the DRN by a number of electrophysiological criteria. 1) Serotonergic dorsal raphe neurons have autoreceptors that hyperpolarize the cell in response to exogenous 5-HT. 2) Serotonergic dorsal raphe neurons show pacemaker activity, either at rest or, for those cells that are silent at rest, in response to current injection. 3) Spikes from serotonergic DRN neurons are followed by a pronounced afterhyperpolarization (AHP). 4) Serotonergic dorsal raphe neurons have slower action potentials (averaging slightly less than 2 ms) than do non-serotonergic neurons (averaging around 1 ms), resulting from a ‘shoulder’ during the tail phase of the spike. In this study, serotonergic dorsal raphe neurons were identified using all four criteria. The cell was classified as serotonergic if 1) bath application of 20 M 5-HT caused a sustainable hyperpolarization of at least 5 mV that returned toward baseline upon washout with ACSF (Fig. 1A); 2) the cell showed pacemaker activity at rest, or in response to current injection (Fig. 1B1); 3) spikes in the cell were followed by an AHP of at least five mV; and 4) spikes in the cell had a duration, measured at the base, of greater than 1 ms (Fig. 1B2). Only cells that met all of these criteria were used in the study (n⫽71 cells).
Genotyping The Zucker rats in this study were too young to be visually identifiable as lean or obese. It was, therefore, necessary to genotype each animal after the experiment to determine the group to which it belonged. To facilitate this, following decapitation the spleens were removed and stored at ⫺70 °C until analysis. To determine the genotype of each animal, a small amount of the spleen was homogenized in digestion buffer (20 mM Tris, 200 mM EDTA, 3% SDS, 0.1 mg/ml RNAse A, 0.1 mg/ml proteinase K, pH 8), and DNA was isolated. Samples were extracted with phenol/ chloroform, and the aqueous portion was extracted a second time with chloroform. Following addition of ethanol, DNA was precipitated from the aqueous layer, rinsed with 70% ethanol, and vacuum dried. After drying, DNA was resuspended in 200 l of 10 mM Tris buffer solution containing 1 mM EDTA to a final concentration of 0.25 mg/ml. Genotypes were assessed via the method of Chua et al., (1996). Briefly, polymerase chain reaction (PCR) amplification of DNA was performed in a thermocycler with the following protocol: 36 amplification cycles (92 °C, 30 s; 54 °C, 60 s; 68 °C, 30 s) in 4 mM NTPs, 750 nM primers (5'-GTTTGCGTATGGAAGTCACAG-3' and 5'-ACCAGCAGAGATGTATCCGAG-3'), 1⫻ PCR buffer and 2 U Taq polymerase. Following amplification, 16 ml of PCR product was digested with 5 U MspI for 90 min at 37 °C, then fractionated in a 2% agarose gel containing ethidium bromide. Gels were trans-illuminated with UV light and photographed for documentation. An uncut 143-bp product was obtained from wildtype alleles (Fa) but two smaller products, one of 106 bp and one of 37 bp, were produced by the mutant (fa) allele. Following visual analysis of the gels, animals were identified as lean (Fa/Fa; n⫽22), heterozygotes (Fa/fa; n⫽14), or obese (fa/fa; n⫽23). Cells from heterozygotes (n⫽23) were found to be statistically indistinguishable from those from lean animals for all variables studied, consistent with the phenotypic similarity between lean and heterozygous animals (heterozygous data not shown).
P. Ohliger-Frerking et al. / Neuroscience 120 (2003) 627– 634
629
Fig. 1. Serotonergic DRN neurons were identified by several criteria. (A) Serotonergic DRN neurons hyperpolarize in response to 5-HT. 5-HT (20 M) was bath applied during the time indicated by the bar. The representative neuron shown hyperpolarized and then returned to baseline voltage following washout of 5-HT with normal ACSF. (B1) Serotonergic DRN neurons show pacemaker activity. In the representative cell shown, pacemaker activity was present at rest; in other cells, a modest (10 – 60 pA) current injection was required to elicit pacemaker activity. Note that this cell also had a pronounced AHP following each spike. (B2) Serotonergic DRN neurons have broad action potentials. A spike from the cell in B1 is shown on an expanded time scale to illustrate the long duration of the action potential. Scale bar⫽4 mV/4 min in (A), 39 mV/820 ms in (B1), and 39 mV/3.1 ms in (B2).
Data analysis Data were analyzed using the Student’s t-test (Figs. 5, 6, and 7), a two way repeated measure ANOVA (Figs. 3 and 4), or the Pearson Product Moment test (Fig. 2) as appropriate. Post hoc analysis of the ANOVA was done with the Tukey test. The data in Fig. 2 were not normally distributed and, therefore, are presented as median (IQR). All other results are presented as the mean⫾S.E.M. Statistical significance was determined at Pⱕ0.05.
RESULTS Resting properties The resting properties of serotonergic DRN neurons from lean and obese Zucker rats were examined first. Using the whole-cell patch clamp technique in voltage clamp mode, we injected a 5 mV voltage pulse and measured the input resistance of each cell to determine if plasma membranes of DRN neurons from obese Zucker rats were more or less leaky than those of lean Zucker rats. This was not the case, as indicated by the lack of genotype difference in the input resistance of the cells (689⫾51 M⍀, n⫽24 lean; 698⫾49 M⍀, n⫽18 obese). Membrane capacitance, which is directly pro-
Fig. 2. Spontaneous firing rates of serotonergic DRN neurons from lean and obese Zucker rats do not differ. (A) Representative voltage traces from spontaneously active serotonergic DRN neurons of lean (top) and obese (bottom) Zucker rats. Average spontaneous firing frequency was not different in cells from lean and obese rats (B). The averages shown in (B) include all cells, even those that were silent at rest.
portional to the surface area of the cell, also did not differ in cells from lean and obese animals (25.4⫾1.3 pF, n⫽24 lean; 25.1⫾1.2 pF, n⫽20 obese). Next we measured resting membrane potential using the wholecell patch clamp technique in current clamp configuration to determine if the cells from obese rats were more or less depolarized at rest compared with cells from lean animals. We found no genotype difference in the resting membrane potential of the cells (⫺52.6⫾1.7 mV, n⫽23 lean; ⫺53.1⫾1.3 mV, n⫽21 obese). Serotonergic neurons from the DRN have pacemaker activity and fire spontaneously in vivo as well as in the slice preparation, although some of the neurons in the slice preparation are silent, presumably due to the loss of excitatory adrenergic afferents (Mosko and Jacobs, 1974; Aghajanian and Vandermaelen, 1982; Van-
630
P. Ohliger-Frerking et al. / Neuroscience 120 (2003) 627– 634
Fig. 3. Serotonergic DRN neurons of obese rats are hyperexcitable in response to current injection. (A) Representative voltage traces from a serotonergic DRN neuron of a lean Zucker rat. The middle third of each trace shows the response of the cell to current injection. The cell was injected with 20, 60, and 100 pA of current (top to bottom), and each subsequently larger injection provoked more rapid firing. Representative voltage traces from a serotonergic DRN neuron from an obese rat are shown in (B). Current injections larger than 20 pA evoked a significantly faster firing frequency in the cells from obese rats compared with those from lean rats. The average responses of serotonergic DRN neurons from lean (closed circles) and obese rats (open circles) to current injection are indicated in (C). Subsequently larger current injections produced an increase in average firing rate in cells from both lean and obese animals, but cells from obese rats respond with a significantly faster average firing frequency at each level of current injection over 20 pA.
dermaelen and Aghajanian, 1983; Crunelli et al., 1983). We have previously found that DRN neurons from obese rats fire more in the presence of the ␣1 adrenergic agonist, phenylephrine (PE), than do those from lean rats (Ohliger-Frerking et al., 2002). However, because PE depolarizes serotonergic DRN neurons from obese rats more than it depolarizes those from lean rats (Ohliger-Frerking et al., 2002), it remains unclear if any of this increased firing in cells from obese rats is due to a change in the intrinsic spontaneous firing rate. Therefore, in the absence of adrenergic agonists¸ we examined the spontaneous firing rate of serotonergic DRN neurons in brain slices from lean and obese Zucker rats in current clamp mode (Fig. 2A). We found that approximately half of the cells recorded in lean animals were spontaneously active (56%; n⫽25), and a similar proportion were spontaneously active in obese animals (48%; n⫽21; P⬎0.5). The spontaneous firing rates of serotonergic DRN neurons were not different for lean and obese animals, when all cells were considered (Fig. 2B), or when only those cells that were spontaneously active were considered (not shown). The spontaneous firing rate of all cells from obese animals was (median [IQR]) 0 (0 –1.33), and for all cells from lean animals, was 0.33 (0 –1.5). Cell excitability Although the resting properties of serotonergic DRN neurons from lean and obese animals were not different, it remained to be seen if they responded similarly to
depolarization. To test if serotonergic DRN neurons from obese rats are differentially excitable, a current pulse of 1 s duration incrementing in steps from 20 to 100 pA was injected into cells of lean and obese rats every 3 s, and changes in voltage and firing rate evoked by depolarization were recorded. Fig. 3A shows representative voltage traces evoked by 20, 60 and 100 pA current injection into a DRN neuron in a brain slice from a lean Zucker rat. As expected, as the magnitude of the current injection increased incrementally from 20 to 100 pA, firing rate also increased. In DRN cells from lean Zucker rats, the average evoked firing frequencies at current injection levels of 20, 40, 60, 80, and 100 pA were 1.2⫾0.3 Hz, 2.8⫾0.5 Hz, 4.3⫾0.6 Hz, 6.1⫾0.6 Hz, and 7.9⫾0.6 Hz, respectively (Fig. 3C, closed circles; n⫽25). Fig. 3B shows representative samples of the voltage and firing response of a single DRN neuron from an obese Zucker rat evoked by current injections of 20, 60, and 100 pA. As in cells from the lean rats, subsequently larger current injection steps provoked a significantly larger increase in firing rate of cells from obese rats at every increase in current injection level (Fig. 3C, open circles; n⫽21). In cells from obese rats, 20, 40, 60, 80, and 100 pA current injections evoked average firing rates of 2.2⫾0.4 Hz, 4.7⫾0.5 Hz, 6.9⫾0.7 Hz, 9.2⫾0.8 Hz, and 11.0⫾1.0 Hz, respectively. Evoked firing rates of cells from obese rats were significantly higher than those of the lean cells at all current steps greater than 20 pA, indicating that serotonergic DRN neurons of obese rats are hyperexcitable.
P. Ohliger-Frerking et al. / Neuroscience 120 (2003) 627– 634
631
Fig. 4. Voltage changes in response to current injection were not different in serotonergic DRN neurons from lean and obese Zucker rats. Representative voltage traces from cells of lean (A) and obese (B) rats show the responses of these cells to 20, 60, and 100 pA current injections (top to bottom) in ACSF to which the sodium channel blocker TTX was added to abolish spiking. The average current-evoked change in membrane potential of serotonergic DRN neurons of lean (closed circles) and obese (open circles) rats is shown in (C). There was no difference in current-evoked depolarization in cells from lean and obese rats at any current injection level.
Depolarization in TTX One explanation for this hyperexcitability is that cells from the obese animals respond with greater depolarization following a given current injection. This was not qualitatively apparent from the data and was not easily evaluated quantitatively because spiking following current injection made it difficult to clearly measure the depolarization. Therefore we repeated the previous experiment with the sodium channel blocker TTX added to the perfusate to abolish spiking. Bath-applied TTX (1–2 M) abolished action potentials and allowed a more accurate measurement of cell depolarization following current injection. We found no difference in the magnitude of membrane depolarization of cells from lean and obese rats at any current injection level in TTX (Fig. 4; n⫽11 lean, n⫽14 obese), suggesting that the hyperexcitability of serotonergic DRN neurons from obese rats is not due to increased depolarization when current is injected into the cell. Threshold potential for spike initiation We next examined the possibility that hyperexcitability of serotonergic DRN neurons from obese rats reflected a lower threshold for firing an action potential; i.e. less current injection or depolarization would be required to bring the cells to threshold potential. To test this hypothesis, we measured the threshold of action potentials for each cell (that is, the membrane voltage at the onset of the rapid depolarizing spike due to voltage-gated Na⫹ channel activation; Fig. 5A, arrow). These measurements showed no genotype difference in threshold potentials (Fig. 5B, ⫺30.5⫾0.9 mV, n⫽21 lean; ⫺32.5⫾0.9 mV, n⫽21 obese).
This information together with the TTX data suggests that hyperexcitability of serotonergic DRN neurons from obese animals is not due to a change in the coupling between depolarization and spiking. The possibility remained, however, that the hyperexcitability might be due to a difference in how the cells recover from the spike. Afterhyperpolarization analysis Spikes in serotonergic DRN neurons are followed by a pronounced AHP that is mediated by the opening of Ca2⫹-gated K⫹ channels (Crunelli et al., 1983; Aghajanian, 1985). We reasoned that a change in the AHP that caused the cells to recover more rapidly could be responsible for their hyperexcitability. To test this hypothesis, we measured the amplitudes of the AHPs in response to single isolated spikes in cells from lean and obese rats. The average AHP of cells from obese rats was significantly smaller than that of cells from lean rats. Representative AHPs from cells of lean and obese rats are superimposed in Fig. 6A. The AHP amplitude for cells from lean rats averaged ⫺17.9⫾1.3 mV while that for cells from obese rats averaged ⫺14.2⫾0.8 mV (Fig. 6B; n⫽25 lean, n⫽21 obese). On closer comparison of the AHPs from lean and obese rats, a difference in time course of the decay was also apparent. To quantitate this effect, we fit the decay of each AHP with a single exponential. The measured time constant of decay () was, on average, significantly faster in cells from obese rats than in cells from lean animals (216⫾29 ms, lean; n⫽25; 130⫾15 ms, obese; n⫽21; Fig. 6C).
632
P. Ohliger-Frerking et al. / Neuroscience 120 (2003) 627– 634
Fig. 5. Threshold potentials of serotonergic DRN neurons from lean and obese Zucker rats do not differ. Representative voltage traces of cells from lean (A1) and obese (A2) rats are shown. Threshold potential was determined to be the voltage at which the rapid depolarization due to voltage-gated Na⫹ channel activation began (arrow). There were no significant differences in the mean threshold potential of serotonergic DRN neurons from lean versus obese pups (B).
Spike analysis The observation that the AHP in serotonergic DRN neurons is smaller and faster in cells from obese animals could indicate a deficit in the cellular mechanisms directly responsible for the AHP, including calcium influx, calcium buffering, and calcium-sensitive potassium channel activation. However, an alternative is that the spike in obese animals is shorter or smaller, so that less calcium current is activated. To distinguish between these possibilities, we examined the properties of spikes recorded in serotonergic DRN neurons from lean and obese animals. We found that spikes were not qualitatively different in cells from lean (n⫽25) and obese (n⫽21) animals (Fig. 7A). There were also no significant differences in the spike amplitude (85.8⫾2.09 mV, lean; n⫽25; 88.3⫾1.55 mV, obese; n⫽21; Fig. 7B), duration (1.92⫾0.09 ms, lean; n⫽25; 1.86⫾0.07 ms, obese; n⫽21; Fig. 7C), half-width or rise time (data not shown), revealed by a quantitative analysis of DRN neurons from lean and obese animals.
DISCUSSION In this study, we conducted an electrophysiological comparison of serotonergic DRN neurons in lean and obese animals. We found no differences in the resting properties or the spontaneous firing rates of these neurons. In fact, the resting firing rates of DRN neurons from lean and obese rats were comparable to those recorded by others in lean rats both in brain slice preparations and in vivo
Fig. 6. The AHP serotonergic DRN neurons of obese rats is smaller and faster than that of lean rats. (A) Representative voltage traces of serotonergic DRN neurons from lean and obese rats superimposed to allow a more direct visual comparison of the AHP amplitudes of these cells. The average amplitude of the AHP of cells from obese rats is significantly smaller than that of lean rats (B). The average time constant () of decay of the AHP was significantly faster for cells from obese rats compared with that of their lean counterparts (C).
(Mosko and Jacobs, 1974; Vandesmaelen and Aghajanian, 1983). When subjected to current injection, however, DRN cells from obese Zucker rats responded with a significantly higher firing rate for currents greater than 20 pA. This hyperexcitability was not due to a difference in coupling between current injection and depolarization, as evidenced by the fact that current injection in the presence of TTX elicited the same voltage change in cells from lean and obese animals. However, the AHP following an action potential in cells from obese rats was significantly smaller and faster, allowing a more rapid return to resting potential. This results in hyperexcitability of these cells when they fire repetitively at frequencies at which firing rate is limited by the recovery from the AHP. Because the spike waveform is not resolveably different between lean and obese animals, the deficit in the AHP
P. Ohliger-Frerking et al. / Neuroscience 120 (2003) 627– 634
633
Fig. 7. The spike waveform of serotonergic DRN neurons is not significantly different for lean and obese rats. (A) Representative spikes are shown from serotonergic DRN neurons from lean and obese rats. (B) Spike amplitude was not significantly different for cells from lean and obese rats. (C) Spike duration, measured at the base of the action potential, was not different for cells from lean and obese rats.
in obese animals is likely to be due to a change in one of the underlying mechanisms directly involved in the generation of the AHP. The AHP in DRN neurons is mediated by a Ca2⫹-activated K⫹ channel (Crunelli et al., 1983; Aghajanian, 1985). Briefly, Ca2⫹ flows through voltage-gated Ca2⫹ channels into the cell where it is subject to buffering. The remaining free Ca2⫹ is available to bind to the Ca2⫹activated K⫹ channel and activate it. The faster, smaller AHP in DRN neurons from obese rats could, therefore, be due to reduced Ca2⫹ influx, increased Ca2⫹ buffering, and/or a decreased number, affinity, or conductance of Ca2⫹-activated K⫹ channels. A comprehensive study of the ion fluxes underlying the AHP in lean and obese rats could distinguish among these alternatives. Our results show that, when excited, serotonergic DRN neurons from obese rats fire faster than do those of lean rats, reflecting, at least in part, the likelihood that during excitation the AHP becomes rate-limiting. Feeding activity has been shown to excite DRN neurons (Fornal et al., 1996), and obese Zucker rats show greater 5-HT release than do lean rats in response to feeding (Lemierre et al., 1998; De Fanti et al., 2001). The latter may be the result of the hyperexcitability of serotonergic neurons in the DRN that we have described, although the extent to which this hyperexcitability controls serotonergic DRN firing and subsequent 5-HT release in vivo remains to be established. Moreover, the functional significance of the increased 5-HT release is unclear, because increased 5-HT at the VMH causes hypophagia rather than the hyperphagia present in obese Zucker rats.
A possible explanation for this paradoxical increase is that the hyperexcitability of DRN neurons and the increased 5-HT release is a compensatory mechanism to offset other deficits in feeding regulation caused by the reduction of leptin receptor signaling. Leptin administration reportedly enhances 5-HT turnover (Calapai et al., 1999), and it has been suggested that this increase in 5-HT turnover is due to decreased 5-HT reuptake (Calapai et al., 1999; Charnay et al., 2000). Consistent with these results, genetically obese mice (ob/ob) lacking leptin show decreased effectiveness of centrally administered 5-HT (Currie, 1993). It seems plausible that this reduced effectiveness could be offset by hyperexcitability of the serotonergic DRN neurons. While the physiological role of this hyperexcitability remains obscure, it complements the previously described increase in sensitivity that these cells show to PE (OhligerFrerking et al., 2002). The mechanism underlying the increased sensitivity to PE in obese rats is clearly different than that responsible for the hyperexcitability that we have described here, i.e. PE evokes a greater depolarization in cells from obese rats (Ohliger-Frerking et al., 2002), while the hyperexcitability we observed here involves a change in the size and time course of the AHP. The two mechanisms, therefore, could be expected to act synergistically in vivo to make the cells extremely sensitive to adrenergic input. In conclusion, serotonergic DRN neurons of obese Zucker rats are hyperexcitable compared with those of lean Zucker rats when activated, but not at rest. This study provides the first direct evidence for changes in the intrin-
634
P. Ohliger-Frerking et al. / Neuroscience 120 (2003) 627– 634
sic activity of serotonergic neurons in a genetic model for obesity. These results suggest a mechanism for the previously described increase in 5-HT release at the VMH of obese Zucker rats during feeding and they support previous studies suggesting a compensatory role for serotonergic signaling in response to perturbations in leptinergic signaling. Acknowledgements—This work was supported by NIH grant DK32907 (B.H.), with the rats provided by the University of California, Davis, Clinical Nutrition Research Unit (NIH DK35747).
REFERENCES Aghajanian GK (1985) Modulation of a transient outward current in serotonergic neurones by alpha 1-adrenoceptors. Nature 315:501– 503. Aghajanian GK, Vandermaelen CP (1982) Intracellular recordings from serotonergic dorsal raphe neurons: pacemaker potentials and the effect of LSD. Brain Res 238:463–469. Anokhin PK, Shuleikina KV (1977) System organization of alimentary behavior in the newborn and the developing cat. Dev Psychobiol 10:385–419. Auerbach SB, Minzenberg MJ, Wilkinson LO (1989) Extracellular serotonin and 5-hydroxyindoleacetic acid in hypothalamus of the unanesthetized rat measured by in vivo dialysis coupled to highperformance liquid chromatography with electrochemical detection: dialysate serotonin reflects neuronal release. Brain Res 499:281– 290. Blonz ER, Stern JS, Curry DL (1985) Dynamics of pancreatic insulin release in young Zucker rats: a heterozygote effect. Am J Physiol 248:E188 –E193. Blundell JE, Hill AJ (1987) Serotoninergic modulation of the pattern of eating and the profile of hunger-satiety in humans. Int J Obesity 11 Suppl 3141–155. Boulange A, Planche E, de Gasquet P (1979) Onset of genetic obesity in the absence of hyperphagia during the first week of life in the Zucker rat (fa/fa). J Lipid Res 20:857–864. Calapai G, Corica F, Corsonello A, Sautebin L, Di Rosa M, Campo GM, Buemi M, Mauro VN, Caputi AP (1999) Leptin increases serotonin turnover by inhibition of brain nitric oxide synthesis. J Clin Invest 104:975–982. Charnay Y, Cusin I, Vallet PG, Muzzin P, Rohner-Jeanrenaud F, Bouras C (2000) Intracerebroventricular infusion of leptin decreases serotonin transporter binding sites in the frontal cortex of the rat. Neurosci Lett 283:89 –92. Chua SC Jr, Chung WK, Wu-Peng XS, Zhang Y, Liu SM, Tartaglia L, Leibel RL (1996) Phenotypes of mouse diabetes and rat fatty due to mutations in the OB (leptin) receptor. Science 271:994 –996. Crunelli V, Forda S, Brooks PA, Wilson KC, Wise JC, Kelly JS (1983) Passive membrane properties of neurones in the dorsal raphe and periaqueductal gray recorded in vitro. Neurosci Lett 40:263–268. Currie PJ (1993) Differential effects of NE, CLON, and 5-HT on feeding
and macronutrient selection in genetically obese (ob/ob) and lean mice. Brain Res Bull 32:133–142. De Fanti BA, Gavel DA, Hamilton JS, Horwitz BA (2000) Extracellular hypothalamic serotonin levels after dorsal raphe nuclei stimulation of lean (Fa/Fa) and obese (fa/fa) Zucker rats. Brain Res 869:6 –14. De Fanti BA, Hamilton JS, Horwitz BA (2001) Meal-induced changes in extracellular 5-HT in medial hypothalamus of lean (Fa/Fa) and obese (fa/fa) Zucker rats. Brain Res 902:164 –170. Fornal CA, Metzler CW, Marrosu F, Ribiero-do-Valle LE, Jacobs BL (1996) A subgroup of dorsal raphe serotonergic neurons in the cat is strongly activated during oral-buccal movements. Brain Res 716: 123–133. Jacobs BL, Gannon PJ, Azmitia EC (1984) Atlas of serotonergic cell bodies in the cat brainstem: an immunocytochemical analysis. Brain Res Bull 13:1–31. Je´quier E, Tappy L (1999) Regulation of body weight in humans. Physiol Rev 79:451–480. Leibowitz SF, Alexander JT (1998) Hypothalamic serotonin in control of eating behavior, meal size, and body weight. Biol Psychiatry 44:851–864. Lemierre S, Rouch C, Nicolaidis S, Orosco M (1998) Combined effect of obesity and aging on feeding-induced monoamine release in the rostromedial hypothalamus of the Zucker rat. Int J Obesity Relat Metab Disord 22:993–999. McLaughlin CL, Baile CA (1981) Ontogeny of feeding behavior in the Zucker obese rat. Physiol Behav 26:607–612. Meguid MM, Fetissov SO, Blaha V, Yang ZJ (2000) Dopamine and serotonin VMN release is related to feeding status in obese and lean Zucker rats. Neuroreport 11:2069 –2072. Moore BJ, Horwitz BA, Stern JS (1985) Brown fat thermogenesis and its role in the development of obesity. Brain Res Bull 14:577–583. Mosko SS, Jacobs BL (1974) Midbrain raphe neurons: spontaneous activity and response to light. Physiol Behav 13:589 –593. Nonogaki K, Strack AM, Dallman MF, Tecott LH (1998) Leptin-independent hyperphagia and type 2 diabetes in mice with a mutated serotonin 5-HT2C receptor gene. Nat Med 4:1152–1156. Ohliger-Frerking P, Horowitz JM, Horwitz BA (2002) Adrenergic excitation of serotonergic dorsal raphe neurons is enhanced in genetically obese rats. Neurosci Lett 332:107–110. Orosco M, Nicolaidis S (1992) Spontaneous feeding-related monoaminergic changes in the rostromedial hypothalamus revealed by microdialysis. Physiol Behav 52:1015–1019. Orosco M, Trouvin JH, Cohen Y, Jacquot C (1986) Ontogeny of brain monoamines in lean and obese female Zucker rats. Physiol Behav 36:853–856. Routh VH, Stern JS, Horwitz BA (1994) Serotonergic activity is depressed in the ventromedial hypothalamic nucleus of 12-day-old obese Zucker rats. Am J Physiol 267:R712–R719. Schwartz DH, Hernandez L, Hoebel BG (1990) Serotonin release in lateral and medial hypothalamus during feeding and its anticipation. Brain Res Bull 25:797–802. Vandermaelen CP, Aghajanian GK (1983) Electrophysiological and pharmacological characterization of serotonergic dorsal raphe neurons recorded extracellularly and intracellularly in rat brain slices. Brain Res 289:109 –119.
(Accepted 29 April 2003)
SEROTONERGIC DORSAL RAPHE NEURONS FROM OBESE ZUCKER RATS ARE HYPEREXCITABLE P. OHLIGER-FRERKING,a*1 B. A. HORWITZa,b AND J. M. HOROWITZa,b
The ventromedial hypothalamus (VMH) has long been known to be an important mediator in the feeding pathways of humans (Blundell and Hill, 1987), cats (reviewed by Anokhin and Shuleikina, 1977), rats (reviewed by Je´quier and Tappy, 1999) and mice (Nonogaki et al., 1998). A variety of neurotransmitters affect activity of cells in the VMH. Among these is serotonin (5-HT) which modulates both the pattern of eating and the sensation of hunger/ satiety in humans (Blundell and Hill, 1987). In rats, 5-HT has been shown to be released at the VMH at the beginning of a meal, in anticipation of a meal, and during feeding (Schwartz et al., 1990; Orosco and Nicolaidis, 1992; Meguid et al., 2000). In addition, a large number of studies have shown that microinjection of 5-HT into the hypothalamic region reduces food intake (reviewed by Leibowitz and Alexander, 1998). Together, these studies indicate a modulatory role for 5-HT in the regulation of feeding behavior. The 5-HT released in the VMH during feeding derives from nerve terminals rather than from other potential sources such as glia, platelets, or mast cells (Auerbach et al., 1989), and the largest population of 5-HT-containing neurons in the brain that project to the VMH are found in the midbrain dorsal raphe nucleus (DRN; Jacobs et al., 1984). DRN activity increases during oral– buccal movements, such as occur during feeding (Fornal et al., 1996). Furthermore, electrical stimulation of the DRN and feeding behavior both result in increased 5-HT at the VMH (De Fanti et al., 2000; Orosco and Nicolaidis, 1992; Meguid et al., 2000). These studies provide further support for the view that DRN-to-VMH signals play a modulatory role in the neural control of food intake. Genetically obese Zucker rats as young as 16 days are hyperphagic (McLaughlin and Baile, 1981) and also present a number of other phenotypic changes related to obesity including decreased brown fat thermogenesis (Moore et al., 1985), increased white adipocyte size (Boulange et al., 1979), and hyperinsulinemia (Blonz et al., 1985). These rats lack a normally functioning leptin receptor (Chua et al., 1996) and, in addition, exhibit a number of abnormalities in the brain 5-HT system. Several studies of basal 5-HT levels or 5-HT turnover in the hypothalamus of young (ⱕ12 weeks) Zucker rats have been performed, but little consensus has been reached. Orosco et al. (1986) have suggested that resting 5-HT levels in the medial hypothalamus of lean and obese Zucker rats, revealed by chromatography, are not different. De Fanti et al. (2000), using microdialysis, also found that basal 5-HT levels were not different. However, Routh et al. (1994), in studies using tissue punches, found decreased 5-HT turnover in the
a Physiology Graduate Group, University of California, Davis, CA 95616, USA b
Section of Neurobiology, Physiology, and Behavior, University of California, Davis, CA 95616, USA
Abstract—Release of serotonin (5-HT) from dorsal raphe nucleus (DRN) neurons projecting to the ventromedial hypothalamus (VMH) has a modulatory effect on the neural pathway involved in feeding, hunger, and satiety. The obese Zucker rat, an animal model of genetic obesity, exhibits differences in serotonin signaling as well as a mutated leptin receptor. To evaluate possible mechanisms underlying this difference in serotonin signaling, we have compared electrophysiological responses of DRN neurons from 14- to 25-day-old male lean (Fa/Fa) and obese (fa/fa) Zucker rats using the whole-cell patch clamp technique on cells in brain slices from these animals. We found that the resting properties of these neurons are not different, but the DRN neurons from obese rats are hyperexcitable in response to current injection. This hyperexcitability is not accompanied by an increase in the depolarization caused by current injection or by changes in the threshold for spiking. However, the hyperexcitability is accompanied by reduction in the size and time course of the afterhyperpolarization (AHP) following an action potential. DRN neurons of obese rats recover from the AHP faster due to a smaller amplitude AHP and a faster time constant () of decay of the AHP. These deficits are not due to changes in the spike waveform, as the spike amplitude and duration do not differ between lean and obese animals. In summary, we provide evidence that serotonergic DRN neurons from obese Zucker rats are intrinsically hyperexcitable compared with those from lean rats. These results suggest a potential mechanism for the reported increase in 5-HT release at the VMH of obese rats during feeding, and provide the first direct evidence of changes in the intrinsic activity of serotonergic neurons, which are crucial regulators of feeding behavior, in a genetic model of obesity. © 2003 IBRO. Published by Elsevier Science Ltd. All rights reserved. Key words: afterhyperpolarization, hyperexcitability, serotonin, dorsal raphe nucleus, obesity, satiety. 1
Present address: Neurological Sciences Institute, Oregon Health and Science University, 505 Northwest 185th Avenue, Beaverton, OR 97006, USA. *Correspondence to: P. Ohliger-Frerking, Neurological Sciences Institute, Oregon Health and Science University, 505 Northwest 185th Avenue, Beaverton, OR 97006, USA. Tel: ⫹1-503-418-2661; fax: ⫹1-503-418-2501. E-mail address: [email protected] (P. Ohliger-Frerking). Abbreviations: ACSF, artificial cerebrospinal fluid; AHP, afterhyperpolarization; DRN, dorsal raphe nucleus; PCR, polymerase chain reaction; PE, phenylephrine; TTX, tetrodotoxin; VMH, ventromedial hypothalamus; 5-HT, serotonin; , time constant of decay.
0306-4522/03$30.00⫹0.00 © 2003 IBRO. Published by Elsevier Science Ltd. All rights reserved. doi:10.1016/S0306-4522(03)00360-9
627
628
P. Ohliger-Frerking et al. / Neuroscience 120 (2003) 627– 634
VMH of obese rats under resting conditions (1994). The reasons for this discrepancy are unclear, although it should be noted that microdialysis studies have poor spatial resolution and may have measured 5-HT levels outside of the VMH. Studies focusing on feeding-induced changes in hypothalamic 5-HT are in better agreement, consistently showing increased feeding-induced 5-HT release in the medial hypothalamus of obese versus lean Zucker rats (Lemierre et al., 1998; De Fanti et al., 2001). These results suggest that activity of serotonergic neurons in the DRN of obese Zucker rats may differ from those of lean rats, and, in fact, we have found previously that DRN neurons from obese rats exhibit stronger responses to exogenous adrenergic agonists (Ohliger-Frerking et al., 2002). In this study, we have investigated the electrophysiological properties of serotonergic DRN neurons in brain slices from lean (Fa/Fa) and obese (fa/fa) Zucker rats to determine if they differ in their intrinsic excitability.
EXPERIMENTAL PROCEDURES All experimental protocols in this work were reviewed and approved by the University of California Davis Animal Care and Use Committee, to minimize animal use and suffering, in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals (NIH 85-23, revised 1985).
Slice preparation Male Zucker rats were obtained from the colony maintained by the UC Davis Clinical Nutrition Research Unit. The pups were decapitated and their brains rapidly removed and placed in cold (4 °C) artificial cerebral spinal fluid (ACSF). All animals used were between 14 and 25 days old, an age range over which the (fa/fa) rats are not yet visibly obese; this was done so that any differences in cellular activity would be more likely to lie upstream of the onset of obesity, rather than be a consequence of obesity. ACSF containing (in mM) 125 NaCl, 2.5 KCl, 1 MgCl2, 1.25 NaH2PO4, 25 NaHCO3, 10 dextrose and 2 CaCl2, was aerated with 95% O2/5% CO2 at all times, and had a pH of 7.3 and an osmolarity of approximately 300 mOsm. The brain was trimmed and 250 m thick coronal midbrain slices were prepared using a vibratome filled with cold (4 °C) ACSF. Slices containing DRN were visually identified by the presence of the cerebral aqueduct and the medial longitudinal fasciculi. Three DRN-containing slices were typically obtained from each brain. Slices were transferred to a holding chamber filled with warm (35 °C) ACSF where they were allowed to recover for a minimum of 1 h prior to recording.
Whole-cell recording In preparation for each experiment, single DRN-containing slices were transferred to a recording chamber perfused with warm (35 °C) ACSF at 2 ml/min. Borosilicate glass microelectrodes (2.5– 6 M⍀) were filled with recording solution containing (in mM) 140 K-gluconate, 5 NaCl, 1 MgCl2, 0.3 EGTA, 10 HEPES, 3 MgATP and 0.2 NaGTP. The osmolarity of the internal solution was approximately 300 mOsm and the pH was adjusted to 7.4. Single cells in the DRN were visualized in the slice using a Nikon IR-DIC microscope (Technical Instruments, Burlingame, CA, USA) and an Olympus camera (Scientific Instrument Co., Sunnyvale, CA, USA). Whole-cell patch clamp recordings were made from each cell using the axoclamp 1C patch clamp amplifier in current clamp mode. Whole-cell voltages were filtered at 2 kHz and digitized at 5 kHz. Voltage records were stored directly on a computer and then analyzed off-line using Pclamp 6 software. Cell
capacitance, series resistance, and input resistance were monitored with a brief voltage step of 5 mV in voltage clamp mode. Series resistances were generally between 5 and 20 M⍀, and input resistances were generally between 400 and 800 M⍀. Resting membrane potential was measured either directly, if the cell was silent at rest, or as the voltage during occasional plateaus between spikes, if the cell was spontaneously active. 5-Hydroxytryptamine (5-HT) and tetrodotoxin (TTX) were purchased from Sigma (St. Louis, MO, USA). Both were dissolved in ACSF to the given concentration and bath applied via a drip system at 2 ml/min when assessing their effects on the cells. Serotonergic dorsal raphe neurons can be separated from non-serotonergic cells in the DRN by a number of electrophysiological criteria. 1) Serotonergic dorsal raphe neurons have autoreceptors that hyperpolarize the cell in response to exogenous 5-HT. 2) Serotonergic dorsal raphe neurons show pacemaker activity, either at rest or, for those cells that are silent at rest, in response to current injection. 3) Spikes from serotonergic DRN neurons are followed by a pronounced afterhyperpolarization (AHP). 4) Serotonergic dorsal raphe neurons have slower action potentials (averaging slightly less than 2 ms) than do non-serotonergic neurons (averaging around 1 ms), resulting from a ‘shoulder’ during the tail phase of the spike. In this study, serotonergic dorsal raphe neurons were identified using all four criteria. The cell was classified as serotonergic if 1) bath application of 20 M 5-HT caused a sustainable hyperpolarization of at least 5 mV that returned toward baseline upon washout with ACSF (Fig. 1A); 2) the cell showed pacemaker activity at rest, or in response to current injection (Fig. 1B1); 3) spikes in the cell were followed by an AHP of at least five mV; and 4) spikes in the cell had a duration, measured at the base, of greater than 1 ms (Fig. 1B2). Only cells that met all of these criteria were used in the study (n⫽71 cells).
Genotyping The Zucker rats in this study were too young to be visually identifiable as lean or obese. It was, therefore, necessary to genotype each animal after the experiment to determine the group to which it belonged. To facilitate this, following decapitation the spleens were removed and stored at ⫺70 °C until analysis. To determine the genotype of each animal, a small amount of the spleen was homogenized in digestion buffer (20 mM Tris, 200 mM EDTA, 3% SDS, 0.1 mg/ml RNAse A, 0.1 mg/ml proteinase K, pH 8), and DNA was isolated. Samples were extracted with phenol/ chloroform, and the aqueous portion was extracted a second time with chloroform. Following addition of ethanol, DNA was precipitated from the aqueous layer, rinsed with 70% ethanol, and vacuum dried. After drying, DNA was resuspended in 200 l of 10 mM Tris buffer solution containing 1 mM EDTA to a final concentration of 0.25 mg/ml. Genotypes were assessed via the method of Chua et al., (1996). Briefly, polymerase chain reaction (PCR) amplification of DNA was performed in a thermocycler with the following protocol: 36 amplification cycles (92 °C, 30 s; 54 °C, 60 s; 68 °C, 30 s) in 4 mM NTPs, 750 nM primers (5'-GTTTGCGTATGGAAGTCACAG-3' and 5'-ACCAGCAGAGATGTATCCGAG-3'), 1⫻ PCR buffer and 2 U Taq polymerase. Following amplification, 16 ml of PCR product was digested with 5 U MspI for 90 min at 37 °C, then fractionated in a 2% agarose gel containing ethidium bromide. Gels were trans-illuminated with UV light and photographed for documentation. An uncut 143-bp product was obtained from wildtype alleles (Fa) but two smaller products, one of 106 bp and one of 37 bp, were produced by the mutant (fa) allele. Following visual analysis of the gels, animals were identified as lean (Fa/Fa; n⫽22), heterozygotes (Fa/fa; n⫽14), or obese (fa/fa; n⫽23). Cells from heterozygotes (n⫽23) were found to be statistically indistinguishable from those from lean animals for all variables studied, consistent with the phenotypic similarity between lean and heterozygous animals (heterozygous data not shown).
P. Ohliger-Frerking et al. / Neuroscience 120 (2003) 627– 634
629
Fig. 1. Serotonergic DRN neurons were identified by several criteria. (A) Serotonergic DRN neurons hyperpolarize in response to 5-HT. 5-HT (20 M) was bath applied during the time indicated by the bar. The representative neuron shown hyperpolarized and then returned to baseline voltage following washout of 5-HT with normal ACSF. (B1) Serotonergic DRN neurons show pacemaker activity. In the representative cell shown, pacemaker activity was present at rest; in other cells, a modest (10 – 60 pA) current injection was required to elicit pacemaker activity. Note that this cell also had a pronounced AHP following each spike. (B2) Serotonergic DRN neurons have broad action potentials. A spike from the cell in B1 is shown on an expanded time scale to illustrate the long duration of the action potential. Scale bar⫽4 mV/4 min in (A), 39 mV/820 ms in (B1), and 39 mV/3.1 ms in (B2).
Data analysis Data were analyzed using the Student’s t-test (Figs. 5, 6, and 7), a two way repeated measure ANOVA (Figs. 3 and 4), or the Pearson Product Moment test (Fig. 2) as appropriate. Post hoc analysis of the ANOVA was done with the Tukey test. The data in Fig. 2 were not normally distributed and, therefore, are presented as median (IQR). All other results are presented as the mean⫾S.E.M. Statistical significance was determined at Pⱕ0.05.
RESULTS Resting properties The resting properties of serotonergic DRN neurons from lean and obese Zucker rats were examined first. Using the whole-cell patch clamp technique in voltage clamp mode, we injected a 5 mV voltage pulse and measured the input resistance of each cell to determine if plasma membranes of DRN neurons from obese Zucker rats were more or less leaky than those of lean Zucker rats. This was not the case, as indicated by the lack of genotype difference in the input resistance of the cells (689⫾51 M⍀, n⫽24 lean; 698⫾49 M⍀, n⫽18 obese). Membrane capacitance, which is directly pro-
Fig. 2. Spontaneous firing rates of serotonergic DRN neurons from lean and obese Zucker rats do not differ. (A) Representative voltage traces from spontaneously active serotonergic DRN neurons of lean (top) and obese (bottom) Zucker rats. Average spontaneous firing frequency was not different in cells from lean and obese rats (B). The averages shown in (B) include all cells, even those that were silent at rest.
portional to the surface area of the cell, also did not differ in cells from lean and obese animals (25.4⫾1.3 pF, n⫽24 lean; 25.1⫾1.2 pF, n⫽20 obese). Next we measured resting membrane potential using the wholecell patch clamp technique in current clamp configuration to determine if the cells from obese rats were more or less depolarized at rest compared with cells from lean animals. We found no genotype difference in the resting membrane potential of the cells (⫺52.6⫾1.7 mV, n⫽23 lean; ⫺53.1⫾1.3 mV, n⫽21 obese). Serotonergic neurons from the DRN have pacemaker activity and fire spontaneously in vivo as well as in the slice preparation, although some of the neurons in the slice preparation are silent, presumably due to the loss of excitatory adrenergic afferents (Mosko and Jacobs, 1974; Aghajanian and Vandermaelen, 1982; Van-
630
P. Ohliger-Frerking et al. / Neuroscience 120 (2003) 627– 634
Fig. 3. Serotonergic DRN neurons of obese rats are hyperexcitable in response to current injection. (A) Representative voltage traces from a serotonergic DRN neuron of a lean Zucker rat. The middle third of each trace shows the response of the cell to current injection. The cell was injected with 20, 60, and 100 pA of current (top to bottom), and each subsequently larger injection provoked more rapid firing. Representative voltage traces from a serotonergic DRN neuron from an obese rat are shown in (B). Current injections larger than 20 pA evoked a significantly faster firing frequency in the cells from obese rats compared with those from lean rats. The average responses of serotonergic DRN neurons from lean (closed circles) and obese rats (open circles) to current injection are indicated in (C). Subsequently larger current injections produced an increase in average firing rate in cells from both lean and obese animals, but cells from obese rats respond with a significantly faster average firing frequency at each level of current injection over 20 pA.
dermaelen and Aghajanian, 1983; Crunelli et al., 1983). We have previously found that DRN neurons from obese rats fire more in the presence of the ␣1 adrenergic agonist, phenylephrine (PE), than do those from lean rats (Ohliger-Frerking et al., 2002). However, because PE depolarizes serotonergic DRN neurons from obese rats more than it depolarizes those from lean rats (Ohliger-Frerking et al., 2002), it remains unclear if any of this increased firing in cells from obese rats is due to a change in the intrinsic spontaneous firing rate. Therefore, in the absence of adrenergic agonists¸ we examined the spontaneous firing rate of serotonergic DRN neurons in brain slices from lean and obese Zucker rats in current clamp mode (Fig. 2A). We found that approximately half of the cells recorded in lean animals were spontaneously active (56%; n⫽25), and a similar proportion were spontaneously active in obese animals (48%; n⫽21; P⬎0.5). The spontaneous firing rates of serotonergic DRN neurons were not different for lean and obese animals, when all cells were considered (Fig. 2B), or when only those cells that were spontaneously active were considered (not shown). The spontaneous firing rate of all cells from obese animals was (median [IQR]) 0 (0 –1.33), and for all cells from lean animals, was 0.33 (0 –1.5). Cell excitability Although the resting properties of serotonergic DRN neurons from lean and obese animals were not different, it remained to be seen if they responded similarly to
depolarization. To test if serotonergic DRN neurons from obese rats are differentially excitable, a current pulse of 1 s duration incrementing in steps from 20 to 100 pA was injected into cells of lean and obese rats every 3 s, and changes in voltage and firing rate evoked by depolarization were recorded. Fig. 3A shows representative voltage traces evoked by 20, 60 and 100 pA current injection into a DRN neuron in a brain slice from a lean Zucker rat. As expected, as the magnitude of the current injection increased incrementally from 20 to 100 pA, firing rate also increased. In DRN cells from lean Zucker rats, the average evoked firing frequencies at current injection levels of 20, 40, 60, 80, and 100 pA were 1.2⫾0.3 Hz, 2.8⫾0.5 Hz, 4.3⫾0.6 Hz, 6.1⫾0.6 Hz, and 7.9⫾0.6 Hz, respectively (Fig. 3C, closed circles; n⫽25). Fig. 3B shows representative samples of the voltage and firing response of a single DRN neuron from an obese Zucker rat evoked by current injections of 20, 60, and 100 pA. As in cells from the lean rats, subsequently larger current injection steps provoked a significantly larger increase in firing rate of cells from obese rats at every increase in current injection level (Fig. 3C, open circles; n⫽21). In cells from obese rats, 20, 40, 60, 80, and 100 pA current injections evoked average firing rates of 2.2⫾0.4 Hz, 4.7⫾0.5 Hz, 6.9⫾0.7 Hz, 9.2⫾0.8 Hz, and 11.0⫾1.0 Hz, respectively. Evoked firing rates of cells from obese rats were significantly higher than those of the lean cells at all current steps greater than 20 pA, indicating that serotonergic DRN neurons of obese rats are hyperexcitable.
P. Ohliger-Frerking et al. / Neuroscience 120 (2003) 627– 634
631
Fig. 4. Voltage changes in response to current injection were not different in serotonergic DRN neurons from lean and obese Zucker rats. Representative voltage traces from cells of lean (A) and obese (B) rats show the responses of these cells to 20, 60, and 100 pA current injections (top to bottom) in ACSF to which the sodium channel blocker TTX was added to abolish spiking. The average current-evoked change in membrane potential of serotonergic DRN neurons of lean (closed circles) and obese (open circles) rats is shown in (C). There was no difference in current-evoked depolarization in cells from lean and obese rats at any current injection level.
Depolarization in TTX One explanation for this hyperexcitability is that cells from the obese animals respond with greater depolarization following a given current injection. This was not qualitatively apparent from the data and was not easily evaluated quantitatively because spiking following current injection made it difficult to clearly measure the depolarization. Therefore we repeated the previous experiment with the sodium channel blocker TTX added to the perfusate to abolish spiking. Bath-applied TTX (1–2 M) abolished action potentials and allowed a more accurate measurement of cell depolarization following current injection. We found no difference in the magnitude of membrane depolarization of cells from lean and obese rats at any current injection level in TTX (Fig. 4; n⫽11 lean, n⫽14 obese), suggesting that the hyperexcitability of serotonergic DRN neurons from obese rats is not due to increased depolarization when current is injected into the cell. Threshold potential for spike initiation We next examined the possibility that hyperexcitability of serotonergic DRN neurons from obese rats reflected a lower threshold for firing an action potential; i.e. less current injection or depolarization would be required to bring the cells to threshold potential. To test this hypothesis, we measured the threshold of action potentials for each cell (that is, the membrane voltage at the onset of the rapid depolarizing spike due to voltage-gated Na⫹ channel activation; Fig. 5A, arrow). These measurements showed no genotype difference in threshold potentials (Fig. 5B, ⫺30.5⫾0.9 mV, n⫽21 lean; ⫺32.5⫾0.9 mV, n⫽21 obese).
This information together with the TTX data suggests that hyperexcitability of serotonergic DRN neurons from obese animals is not due to a change in the coupling between depolarization and spiking. The possibility remained, however, that the hyperexcitability might be due to a difference in how the cells recover from the spike. Afterhyperpolarization analysis Spikes in serotonergic DRN neurons are followed by a pronounced AHP that is mediated by the opening of Ca2⫹-gated K⫹ channels (Crunelli et al., 1983; Aghajanian, 1985). We reasoned that a change in the AHP that caused the cells to recover more rapidly could be responsible for their hyperexcitability. To test this hypothesis, we measured the amplitudes of the AHPs in response to single isolated spikes in cells from lean and obese rats. The average AHP of cells from obese rats was significantly smaller than that of cells from lean rats. Representative AHPs from cells of lean and obese rats are superimposed in Fig. 6A. The AHP amplitude for cells from lean rats averaged ⫺17.9⫾1.3 mV while that for cells from obese rats averaged ⫺14.2⫾0.8 mV (Fig. 6B; n⫽25 lean, n⫽21 obese). On closer comparison of the AHPs from lean and obese rats, a difference in time course of the decay was also apparent. To quantitate this effect, we fit the decay of each AHP with a single exponential. The measured time constant of decay () was, on average, significantly faster in cells from obese rats than in cells from lean animals (216⫾29 ms, lean; n⫽25; 130⫾15 ms, obese; n⫽21; Fig. 6C).
632
P. Ohliger-Frerking et al. / Neuroscience 120 (2003) 627– 634
Fig. 5. Threshold potentials of serotonergic DRN neurons from lean and obese Zucker rats do not differ. Representative voltage traces of cells from lean (A1) and obese (A2) rats are shown. Threshold potential was determined to be the voltage at which the rapid depolarization due to voltage-gated Na⫹ channel activation began (arrow). There were no significant differences in the mean threshold potential of serotonergic DRN neurons from lean versus obese pups (B).
Spike analysis The observation that the AHP in serotonergic DRN neurons is smaller and faster in cells from obese animals could indicate a deficit in the cellular mechanisms directly responsible for the AHP, including calcium influx, calcium buffering, and calcium-sensitive potassium channel activation. However, an alternative is that the spike in obese animals is shorter or smaller, so that less calcium current is activated. To distinguish between these possibilities, we examined the properties of spikes recorded in serotonergic DRN neurons from lean and obese animals. We found that spikes were not qualitatively different in cells from lean (n⫽25) and obese (n⫽21) animals (Fig. 7A). There were also no significant differences in the spike amplitude (85.8⫾2.09 mV, lean; n⫽25; 88.3⫾1.55 mV, obese; n⫽21; Fig. 7B), duration (1.92⫾0.09 ms, lean; n⫽25; 1.86⫾0.07 ms, obese; n⫽21; Fig. 7C), half-width or rise time (data not shown), revealed by a quantitative analysis of DRN neurons from lean and obese animals.
DISCUSSION In this study, we conducted an electrophysiological comparison of serotonergic DRN neurons in lean and obese animals. We found no differences in the resting properties or the spontaneous firing rates of these neurons. In fact, the resting firing rates of DRN neurons from lean and obese rats were comparable to those recorded by others in lean rats both in brain slice preparations and in vivo
Fig. 6. The AHP serotonergic DRN neurons of obese rats is smaller and faster than that of lean rats. (A) Representative voltage traces of serotonergic DRN neurons from lean and obese rats superimposed to allow a more direct visual comparison of the AHP amplitudes of these cells. The average amplitude of the AHP of cells from obese rats is significantly smaller than that of lean rats (B). The average time constant () of decay of the AHP was significantly faster for cells from obese rats compared with that of their lean counterparts (C).
(Mosko and Jacobs, 1974; Vandesmaelen and Aghajanian, 1983). When subjected to current injection, however, DRN cells from obese Zucker rats responded with a significantly higher firing rate for currents greater than 20 pA. This hyperexcitability was not due to a difference in coupling between current injection and depolarization, as evidenced by the fact that current injection in the presence of TTX elicited the same voltage change in cells from lean and obese animals. However, the AHP following an action potential in cells from obese rats was significantly smaller and faster, allowing a more rapid return to resting potential. This results in hyperexcitability of these cells when they fire repetitively at frequencies at which firing rate is limited by the recovery from the AHP. Because the spike waveform is not resolveably different between lean and obese animals, the deficit in the AHP
P. Ohliger-Frerking et al. / Neuroscience 120 (2003) 627– 634
633
Fig. 7. The spike waveform of serotonergic DRN neurons is not significantly different for lean and obese rats. (A) Representative spikes are shown from serotonergic DRN neurons from lean and obese rats. (B) Spike amplitude was not significantly different for cells from lean and obese rats. (C) Spike duration, measured at the base of the action potential, was not different for cells from lean and obese rats.
in obese animals is likely to be due to a change in one of the underlying mechanisms directly involved in the generation of the AHP. The AHP in DRN neurons is mediated by a Ca2⫹-activated K⫹ channel (Crunelli et al., 1983; Aghajanian, 1985). Briefly, Ca2⫹ flows through voltage-gated Ca2⫹ channels into the cell where it is subject to buffering. The remaining free Ca2⫹ is available to bind to the Ca2⫹activated K⫹ channel and activate it. The faster, smaller AHP in DRN neurons from obese rats could, therefore, be due to reduced Ca2⫹ influx, increased Ca2⫹ buffering, and/or a decreased number, affinity, or conductance of Ca2⫹-activated K⫹ channels. A comprehensive study of the ion fluxes underlying the AHP in lean and obese rats could distinguish among these alternatives. Our results show that, when excited, serotonergic DRN neurons from obese rats fire faster than do those of lean rats, reflecting, at least in part, the likelihood that during excitation the AHP becomes rate-limiting. Feeding activity has been shown to excite DRN neurons (Fornal et al., 1996), and obese Zucker rats show greater 5-HT release than do lean rats in response to feeding (Lemierre et al., 1998; De Fanti et al., 2001). The latter may be the result of the hyperexcitability of serotonergic neurons in the DRN that we have described, although the extent to which this hyperexcitability controls serotonergic DRN firing and subsequent 5-HT release in vivo remains to be established. Moreover, the functional significance of the increased 5-HT release is unclear, because increased 5-HT at the VMH causes hypophagia rather than the hyperphagia present in obese Zucker rats.
A possible explanation for this paradoxical increase is that the hyperexcitability of DRN neurons and the increased 5-HT release is a compensatory mechanism to offset other deficits in feeding regulation caused by the reduction of leptin receptor signaling. Leptin administration reportedly enhances 5-HT turnover (Calapai et al., 1999), and it has been suggested that this increase in 5-HT turnover is due to decreased 5-HT reuptake (Calapai et al., 1999; Charnay et al., 2000). Consistent with these results, genetically obese mice (ob/ob) lacking leptin show decreased effectiveness of centrally administered 5-HT (Currie, 1993). It seems plausible that this reduced effectiveness could be offset by hyperexcitability of the serotonergic DRN neurons. While the physiological role of this hyperexcitability remains obscure, it complements the previously described increase in sensitivity that these cells show to PE (OhligerFrerking et al., 2002). The mechanism underlying the increased sensitivity to PE in obese rats is clearly different than that responsible for the hyperexcitability that we have described here, i.e. PE evokes a greater depolarization in cells from obese rats (Ohliger-Frerking et al., 2002), while the hyperexcitability we observed here involves a change in the size and time course of the AHP. The two mechanisms, therefore, could be expected to act synergistically in vivo to make the cells extremely sensitive to adrenergic input. In conclusion, serotonergic DRN neurons of obese Zucker rats are hyperexcitable compared with those of lean Zucker rats when activated, but not at rest. This study provides the first direct evidence for changes in the intrin-
634
P. Ohliger-Frerking et al. / Neuroscience 120 (2003) 627– 634
sic activity of serotonergic neurons in a genetic model for obesity. These results suggest a mechanism for the previously described increase in 5-HT release at the VMH of obese Zucker rats during feeding and they support previous studies suggesting a compensatory role for serotonergic signaling in response to perturbations in leptinergic signaling. Acknowledgements—This work was supported by NIH grant DK32907 (B.H.), with the rats provided by the University of California, Davis, Clinical Nutrition Research Unit (NIH DK35747).
REFERENCES Aghajanian GK (1985) Modulation of a transient outward current in serotonergic neurones by alpha 1-adrenoceptors. Nature 315:501– 503. Aghajanian GK, Vandermaelen CP (1982) Intracellular recordings from serotonergic dorsal raphe neurons: pacemaker potentials and the effect of LSD. Brain Res 238:463–469. Anokhin PK, Shuleikina KV (1977) System organization of alimentary behavior in the newborn and the developing cat. Dev Psychobiol 10:385–419. Auerbach SB, Minzenberg MJ, Wilkinson LO (1989) Extracellular serotonin and 5-hydroxyindoleacetic acid in hypothalamus of the unanesthetized rat measured by in vivo dialysis coupled to highperformance liquid chromatography with electrochemical detection: dialysate serotonin reflects neuronal release. Brain Res 499:281– 290. Blonz ER, Stern JS, Curry DL (1985) Dynamics of pancreatic insulin release in young Zucker rats: a heterozygote effect. Am J Physiol 248:E188 –E193. Blundell JE, Hill AJ (1987) Serotoninergic modulation of the pattern of eating and the profile of hunger-satiety in humans. Int J Obesity 11 Suppl 3141–155. Boulange A, Planche E, de Gasquet P (1979) Onset of genetic obesity in the absence of hyperphagia during the first week of life in the Zucker rat (fa/fa). J Lipid Res 20:857–864. Calapai G, Corica F, Corsonello A, Sautebin L, Di Rosa M, Campo GM, Buemi M, Mauro VN, Caputi AP (1999) Leptin increases serotonin turnover by inhibition of brain nitric oxide synthesis. J Clin Invest 104:975–982. Charnay Y, Cusin I, Vallet PG, Muzzin P, Rohner-Jeanrenaud F, Bouras C (2000) Intracerebroventricular infusion of leptin decreases serotonin transporter binding sites in the frontal cortex of the rat. Neurosci Lett 283:89 –92. Chua SC Jr, Chung WK, Wu-Peng XS, Zhang Y, Liu SM, Tartaglia L, Leibel RL (1996) Phenotypes of mouse diabetes and rat fatty due to mutations in the OB (leptin) receptor. Science 271:994 –996. Crunelli V, Forda S, Brooks PA, Wilson KC, Wise JC, Kelly JS (1983) Passive membrane properties of neurones in the dorsal raphe and periaqueductal gray recorded in vitro. Neurosci Lett 40:263–268. Currie PJ (1993) Differential effects of NE, CLON, and 5-HT on feeding
and macronutrient selection in genetically obese (ob/ob) and lean mice. Brain Res Bull 32:133–142. De Fanti BA, Gavel DA, Hamilton JS, Horwitz BA (2000) Extracellular hypothalamic serotonin levels after dorsal raphe nuclei stimulation of lean (Fa/Fa) and obese (fa/fa) Zucker rats. Brain Res 869:6 –14. De Fanti BA, Hamilton JS, Horwitz BA (2001) Meal-induced changes in extracellular 5-HT in medial hypothalamus of lean (Fa/Fa) and obese (fa/fa) Zucker rats. Brain Res 902:164 –170. Fornal CA, Metzler CW, Marrosu F, Ribiero-do-Valle LE, Jacobs BL (1996) A subgroup of dorsal raphe serotonergic neurons in the cat is strongly activated during oral-buccal movements. Brain Res 716: 123–133. Jacobs BL, Gannon PJ, Azmitia EC (1984) Atlas of serotonergic cell bodies in the cat brainstem: an immunocytochemical analysis. Brain Res Bull 13:1–31. Je´quier E, Tappy L (1999) Regulation of body weight in humans. Physiol Rev 79:451–480. Leibowitz SF, Alexander JT (1998) Hypothalamic serotonin in control of eating behavior, meal size, and body weight. Biol Psychiatry 44:851–864. Lemierre S, Rouch C, Nicolaidis S, Orosco M (1998) Combined effect of obesity and aging on feeding-induced monoamine release in the rostromedial hypothalamus of the Zucker rat. Int J Obesity Relat Metab Disord 22:993–999. McLaughlin CL, Baile CA (1981) Ontogeny of feeding behavior in the Zucker obese rat. Physiol Behav 26:607–612. Meguid MM, Fetissov SO, Blaha V, Yang ZJ (2000) Dopamine and serotonin VMN release is related to feeding status in obese and lean Zucker rats. Neuroreport 11:2069 –2072. Moore BJ, Horwitz BA, Stern JS (1985) Brown fat thermogenesis and its role in the development of obesity. Brain Res Bull 14:577–583. Mosko SS, Jacobs BL (1974) Midbrain raphe neurons: spontaneous activity and response to light. Physiol Behav 13:589 –593. Nonogaki K, Strack AM, Dallman MF, Tecott LH (1998) Leptin-independent hyperphagia and type 2 diabetes in mice with a mutated serotonin 5-HT2C receptor gene. Nat Med 4:1152–1156. Ohliger-Frerking P, Horowitz JM, Horwitz BA (2002) Adrenergic excitation of serotonergic dorsal raphe neurons is enhanced in genetically obese rats. Neurosci Lett 332:107–110. Orosco M, Nicolaidis S (1992) Spontaneous feeding-related monoaminergic changes in the rostromedial hypothalamus revealed by microdialysis. Physiol Behav 52:1015–1019. Orosco M, Trouvin JH, Cohen Y, Jacquot C (1986) Ontogeny of brain monoamines in lean and obese female Zucker rats. Physiol Behav 36:853–856. Routh VH, Stern JS, Horwitz BA (1994) Serotonergic activity is depressed in the ventromedial hypothalamic nucleus of 12-day-old obese Zucker rats. Am J Physiol 267:R712–R719. Schwartz DH, Hernandez L, Hoebel BG (1990) Serotonin release in lateral and medial hypothalamus during feeding and its anticipation. Brain Res Bull 25:797–802. Vandermaelen CP, Aghajanian GK (1983) Electrophysiological and pharmacological characterization of serotonergic dorsal raphe neurons recorded extracellularly and intracellularly in rat brain slices. Brain Res 289:109 –119.
(Accepted 29 April 2003)