- Email: [email protected]
Excitatory effects of gastric electrical stimulation on gastric distension responsive neurons in ventromedial hypothalamus (VMH) in rats
Neuroscience Research 55 (2006) 451–457 www.elsevier.com/locate/neures
Excitatory effects of gastric electrical stimulation on gastric distension responsive neurons in ventromedial hypothalamus (VMH) in rats Xiangrong Sun a, Ming Tang b, Jing Zhang c, Jiande D.Z. Chen c,d,* a
Department of Pathophysiology, Medical College of Qingdao University, China b Department of Physiology, Medical College of Qingdao University, China c Veteran Research Foundation, VA Medical Center, Oklahoma City, OK, United States d Division of Gastroenterology, University of Texas Medical Branch, Galveston, TX, United States Received 19 January 2006; accepted 2 May 2006
Abstract Introduction: Gastric electrical stimulation (GES) has been used for the treatment of obesity with unclear central mechanisms. The purpose of this study was to investigate the effects of GES on the neuronal activity in the ventromedial hypothalamus (VMH). Methods: Extracellular potentials of single neurons in VMH were recorded in 52 anesthetized rats. Neurons were classified as gastric distensionexcitatory (GD-E) neurons or GD-inhibitory (GD-I) neurons. GES with four sets of parameters was applied for comparison. Results: Eighty two neurons out of 96 (85.41%) in VMH responded to gastric distension (GD). 37.8% were GD-E neurons and 51(62.2%) were GD-I neurons. 55.0%, 17.6%, 77.8%, 14.3% of GD-E neurons were excited by four sets of parameters: GES1 (standard), GES2 (reduced pulse numbers), GES3 (increased pulse width) and GES4 (reduced frequency), respectively. More GD-E neurons were excited by GES3 (P < 0.05 versus GES2 or GES4) and by GES1 (P < 0.02 versus GES2 or GES4). Among the GD-I neurons, 63.6, 37.9, 73.3, and 51.9% neurons were excited by GES1–4, respectively. Conclusion: GES with parameters used for treating obesity excites GD-responsive neurons in VMH. The excitatory effect of GES is related to the strength of stimulation, including pulse frequency and width as well as pulse train on-time. # 2006 Elsevier Ireland Ltd and the Japan Neuroscience Society. All rights reserved. Keywords: Gastric electrical stimulation; Ventromedial hypothalamus; Gastric distension responsive neurons; Rats; Obesity
1. Introduction Obesity is a major public health problem in western societies. This condition affects approximately one-fifth of the U.S. population and is considered to be a contributing factor in 280,000 annual deaths in the U.S. (Mokdad et al., 1999), with an estimated cost of more than $100 billion a year (Martin et al., 1995). Various treatments options are available for obesity, such as diet, exercises, drugs and surgery. However, none of the available therapies is satisfactory and there is an urgent need to develop safe and effective methods to treat patients with morbid obesity.
* Corresponding author at: 1108 the Strand, Room 221, Galveston, TX 77555-0632, United States. Tel.: +1 409 747 3071; fax: +1 409 747 3084. E-mail address: [email protected] (Jiande D.Z. Chen).
Treatment of morbid obesity with gastric electrical stimulation (GES) was first suggested by Cigaina et al. who found decreased food intake and body weight with long-term GES in swine (Cigaina et al., 1996), followed with a number of clinical studies showing the efficacy of GES in reducing weight in obese patients (Miller, 2002; D’Argent, 2002; Cigaina, 2002) with increased satiety and reduced appetite (De Luca et al., 2004). Recently a number of studies have shown peripheral mechanisms of GES therapy for obesity (Chen, 2004). These data suggested that GES induced gastric distention or reduced gastric tone (Zhu and Chen, 2005), inhibited antral contractions (Lei et al., 2005), chronically impaired intestinal gastric slow waves (Ouyang et al., 2003). Except the mediating effect of nucleus tractus solitarri (NTS) was reported (Qin et al., 2005), little is known about the central mechanisms of GES. Food intake is controlled by a highly complex physiological process involving the integrative operation of various brain regions, discrete neuronal pathways, as well as gastrointestinal
0168-0102/$ – see front matter # 2006 Elsevier Ireland Ltd and the Japan Neuroscience Society. All rights reserved. doi:10.1016/j.neures.2006.05.001
452
X. Sun et al. / Neuroscience Research 55 (2006) 451–457
and metabolic factors. Recent investigations (Iversen et al., 2000; Woods and Stricker, 1999) have demonstrated that many of these effects are due to involvement of descending and ascending pathways of the cerebral cortex or the basal forebrain passing through the hypothalamus. During a meal, the distension and secretory activities of the gastrointestinal tract elicited by the ingestion of food may activate the vagal afferent fibers, which comprise the primary neuroanatomical substrate of the neural–gut–brain axis. Several lines of evidence suggest that some meal-associated visceral signals can reach the hypothalamus via the vagal afferent fibers supplying the upper gastrointestinal tract (Jeanningros and Mei, 1980; Zhang et al., 2003; Bary, 2000; Schwartz, 2000), therefore, sensory information from gastrointestinal receptors particularly that transmitted by vagal primary afferents, has been a focus of many investigations due to the recognized importance of this information in the regulation of satiety and gastrointestinal function (Jeanningros and Mei, 1980). It is well established that the VMH in the hypothalamus is closely related to the regulation of feeding behavior and plays an important role in the mediation of satiety. Bilateral VMH lesions consistently caused hyperphagia and obesity, while the opposite effect often seemed to be produced by electrical stimulation of the neurons in the VMH (Raybould, 1992; PlataSalaman, 1998). There is good evidence that implantable electric stimulation of the stomach activates neurons in NTS (Qin et al., 2005) but there is no data regarding roles of the VMH in the central mechanism of GES. It is reasonable to expect that there are gastric distension responsive neurons in the VMH which may faithfully reflect the central representation of gastric mechanoreceptor activation at the level of the hypothalamus and these VMH GD-responsive neuron may involve in regulation of food intake. The present study was undertaken to investigate the effects of implantable gastric electric stimulation with different parameters at the distal stomach on the neuronal activity of gastric-related neurons in the VMH, therefore, to explore the hypothesis that the neurons in the VMH is involved in the central mechanisms of the GES treatment of obesity.
2. Materials and methods 2.1. Animals Adult female or male Wistar rats weighing 250–300 g were used in this study (Institute of Pharmaceutical Research of Qingdao, China). They were housed in a controlled temperature (25 2 8C) and were maintained in individual plastic cages on a 12:12-h light–dark schedule before surgery. Food and water were available ad libitum. The study was approved and all procedures were performed in accordance with institutional guidelines of the Animal Care and Use Committee at Qingdao University.
2.2. Surgery The animals were anesthetized with urethane (1 g/kg, i.p.) and supplemental anesthetics were administrated as needed during the experiment. Anesthesia was confirmed by the absence of paw pinch reflex in response to punching at legs. The heart rate was continuously monitored. The trachea was cannulated to
maintain an open airway. Core temperature was monitored throughout testing via a rectal thermometer and body temperature was regulated to remain at 37 8C through manual adjustments to a heating pad. All surgeries were performed under aseptic conditions in the following order. 2.2.1. Abdominal surgical A laparotomy was made and the stomach was exposed. Gastric contents were removed via a small incision in the fundus wall and the stomach was cleaned with warm isotonic saline. Then a gastric balloon (see below) was gently inserted in the corpus; the wound was ligated around the balloon’s shaft. The balloon catheter was connected to an injector, through which to allow distension of the stomach by infusion of small volumes of saline (3–5 ml, 10–30 s) (Guan et al., 2003). One pair of platinum electrodes (0.3 cm apart used for electric stimulation) was sutured on the serosal surface of gastric antrum close to the lesser curvature. The abdomen was then closed and a small piece of gauze was used as drainage to prevent the accumulation of secretory fluids in the abdomen. 2.2.2. Cranial surgery The rat was placed in a stereotaxic frame (SN-3, Narashige, Tokyo, Japan) and the skull was exposed. A small hole was drilled in the skull to expose the cortex and the dura was cut. Open part of the brain was covered with warm agar (3% in saline) to improve stability for neuronal recording. The stereotaxic coordinates were used in accordance with the atlas of Paxinos and Waston (1998). A one-barrel glass microelectrode filled with 0.5 M sodium acetate and 2% potamine sky blue (tip diameter: 1–3 mm, resistance: 15–20 MV) was advanced in 10 mm steps with the aid of hydraulic micropositioner into the area of the VMH (2.3–2.8 mm posterior to bregma, 0.5–1.0 mm left lateral to the midline and 9.0–10.4 mm below the outer surface of the skull). The microelectrode was used for the extracellular recording of neurons discharge and histological verification of the recording site.
2.3. Procedure 2.3.1. Extracelluar recording Once the microelectrode was advanced into the area of the VMH, the extracellular action potentials of single unit was recorded. The recorded signals were amplified using MEZ8201 amplifier (Nihon Kohden, Tokyo, Japan) and displayed on a oscilloscope (VC-II, Nihon Kohden, Tokyo, Japan). Electrical signals from the amplifier were input into SUMP-PC bioelectric signal processing system and all data were stored in a computer for subsequent analysis. 2.3.2. Identification of GD-responsive neurons After a steady baseline firing rate was obtained for at least 5 min, the test for gastric distension was performed by inflating the gastric balloon with 3–5 ml 37 8C water at a rate of 0.5 ml/s and maintained for 10–30 s. A neuron was identified as GD-responsive neurons if its mean firing rate changed by at least 20% from the mean basal firing level. The GD-responsive neurons were further classified into GD-excitatory (GD-E) neurons and GD-inhibitory (GD-I) neurons according to the spontaneous discharge increased or decreased with GD. 2.3.3. Gastric electrical stimulation study For each GD-responsive neuron, gastric electrical stimulation with four different sets of parameters was applied randomly for 1 min: GES-1 (pulse train of standard parameters—pulse amplitude: 6 mA, pulse width: 0.3 ms, pulse frequency: 40 Hz, train on-time of 2 s and off-time of 3 s); GES-2 (same as GES-1 but reduced on-time to 0.1 s); GES-3 (same as GES-1 but increased pulse width to 3 ms); GES4 (same as GES-1 but reduced pulse frequency to 20 Hz). GES-1 is the set of parameters which are most commonly used in the clinical application of GES for obesity (Miller, 2002; D’Argent, 2002; Cigaina, 2002; De Luca et al., 2004). The other sets of parameters were used as control to study the roles of train on-time, pulse width and pulse frequency involved in GES. The reason we chosen to use the parameters (set #1) identical to the humans was that the frequency of gastric myoelectrical activity in rats is close to that in humans and the impedance between the pair of stimulation electrodes placed in the gastric serosal is similar among different species (rats, dogs and humans). In addition, a number of previous studies performed in our lab revealed similar effective parameters for GES among different species (Chen,
X. Sun et al. / Neuroscience Research 55 (2006) 451–457
453
Table 1 Effects of GES with various parameters on activity of VMH neurons responding to GD Responsive to GD
GES-1
GES-2
GES-3
GES-4
R
NR
R
NR
R
NR
R
NR
GD-E GD-I
11E, 4I 21E, 9I
5 3
3E, 5I 11E, 10I
9 8
14E, 3I 22E, 7I
1 1
2E, 6I 14E, 7I
6 6
Total(R/tested)
45*/53 (84%)
29/46 (63%)
46*,#/48 (96%)
29/41 (70%) *
GES: gastric electrical stimulation, GD: gastric distension (3–5 ml, 10–30 s); R: response to GES; NR: no response to GES. P < 0.05 compared with population of neurons responding to GES2; #P < 0.01 compared to population of neurons responding to GES4. 2004; Zhu and Chen, 2005; Lei et al., 2005; Ouyang et al., 2003; Qin et al., 2005).
2.4. Histology At the end of each experiment, pontamine sky blue was ejected from the recording microelectrode to the site recorded by applying cathodal current (10 mA, 20 min). Animals were transcardially perfused with isotonic saline 300 ml and 10% formalin 300 ml. The brain was then removed and stored in fixative solution at 4 8C for 24 h. Using a freezing micotome, the brain tissue was cut in 20 mm sections to visualize the location of the pontamine sky blue spot. If the recording sites were out of the VMH, the data were excluded from the analysis.
2.5. Statistical analysis The responsive patterns of VMH neurons to the gastric distension and GES were determined by comparing the discharge frequency of recorded cells in responsive window of post-stimulation with the basal firing rates of the cells in control window of pre-stimulation. The raw tracing of neuronal responses to gastric stimuli was processed by software to eliminate GES artifacts. All data are expressed as mean S.E. Paired Student’s t-test was used to study the difference between any paired data. Chi-square analysis was applied to investigate the difference in the neuronal response patterns to four sets of parameters. Comparison of data were considered statistically significant if P < 0.05.
were affected by GES3 (46/48) than GES2 (29/46) (P < 0.001) or GES4 (29/41) (P = 0.003). Besides GES3, GES1 (45/53) had a more significant effect on GD-responsive neurons than GES2 (P = 0.023) but there was no significant difference between GES1 and GES3 (P > 0.05) (Table 1, Fig. 1). Among the GD-E neurons, 55.0, 17.6, 77.8 and 14.3% of tested neurons showed excitatory responses to GES1–4, respectively. More GD-E neurons were excited by GES3 (P < 0.001 versus GES2; P = 0.022 versus GES4) and GES1 (P = 0.002 versus GES2; P = 0.016 versus GES4). Excitatory response of GD-E neurons to GES3 (increased by 343.6 89.2%) were significantly greater than to GES1 (increased by 97.4 33.7%) (P < 0.05). But there was no significant difference in the average responsive duration (Tables 1 and 2; Figs. 2 and 3). Among the GD-I neurons, 63.6, 37.9, 73.3 and 51.8% were excited by GES1–4, respectively. There was no significant difference in the excitatory effect among GES1, 3 and 4; however, GES2 was noted to be the least effective (compared with GES1, P = 0.043 and compared with GES3, P = 0.006) (Tables 1 and 2; Figs. 4
3. Results 3.1. Responses to gastric distension A total of 96 neurons in the VMH were recorded in this study. The spontaneous unit discharges exhibited three different patterns: phasic, continuous and single. Among 92 recorded neurons, 82 units responded to gastric distension (82/96, 85.4%). Thirty-one (31/82, 37.8%) were GD-E neurons and 51 (51/82, 62.2%) were GD-I neurons. The firing frequency of GD-E neurons was (1.51 0.33) Hz before gastric distension and increased to (2.84 0.62) Hz during gastric distension (P < 0.01) and the firing frequency of GD-I neurons firing frequency decreased from (1.20 0.20) Hz at baseline to (0.61 0.10) Hz during gastric distension. The response latency time was between 0 and 20 s. The average duration of excitatory effect was (86.7 15.6) s and the inhibitory effect was (50.1 9.8) s. 3.2. Effects of GES on GD-responsive neurons The percentage of GD-responsive neurons responded to GES1–4 were 84, 63, 96 and 70%, respectively. More neurons
Fig. 1. Effects of GES with different parameters on GD-responsive neurons. The data were presented as percentage of neurons responding or not responding to GES. *P < 0.05 vs. GES2; #P < 0.01 vs. GES2; §P < 0.01 vs. GES4. R: responding to GES; NR: no responding to GES.
Fig. 2. Excitatory effects of GES with different parameters on GD-E neurons. The data were presented as percentage of neurons excited or not excited by GES. *P < 0.05 vs. GES4; #P < 0.01 vs. GES2. E: excited by GES; I + NR: inhibited and no responded.
454
X. Sun et al. / Neuroscience Research 55 (2006) 451–457
Table 2 The effects of GES1 and GES3 on GD-E and GD-I neurons in VMH Responsive type
GES parameters
Firing frequency before GES (Hz)
Firing frequency after GES (Hz)
Firing varied rate (%)
Latency (s)
Duration (s)
GD-E
GES1 GES3
1.50 0.34 1.36 0.35
2.63 0.83 3.47 0.58
97.44 33.70 343.60 89.21*
0–120 0–70
85.56 14.15 76.47 8.16
GD-I
GES1 GES3
1.05 0.25 1.47 0.41
1.49 0.30 7.58 4.61
112.00 14.67 366.30 87.20*
0–190 0–20#
119.2 14.70 104.1 14.89
*
P < 0.05 vs. GES1; #P < 0.01 vs. GES1.
and 5). In both GD-E and GD-I neurons, the latency of GES3 was markedly shorter than GES1 (Table 2). 4. Discussion GES or implantable gastric stimulation is currently being investigated as an alternative for the treatment of obesity. The present study demonstrated the effect of GES on gastricresponsive afferent neurons in VMH and thereby provided direct evidence of a central neural mechanism that might contribute to the effects of GES on feeding or obesity. Our results showed that parameters 1 and 3 were more effective than parameters 2 and 4 in activating gastric-related neurons in the VMH. This result suggested that the GES used in the previous clinical studies for treating obesity might affect feeding by exciting neurons in the VMH, which is known to have a strong effect in terminating feeding. Compared with parameter 1, parameter 3 showed more intense effect on gastric distension responsive neurons in the VMH. However, it should be noted that the difference in the effect, although significant with these neurons, was to a much lesser degree than the difference in pulse width (10 times) or energy consumption, which is one of major concerns with an implant device. It is clear that the neuronal responses of the VMH to gastric distension observed in the study represent the centrally directed traffic from the activation of vagal gastric mechanoreceptors. Signals from a variety of sensors in the gut that respond to mechanical (distention, contraction) and chemical stimuli transmit to CNS by visceral afferent fibers in vagal and sympathetic nerves. Most of afferent vagal fibers pass through nodose ganglia to terminate on NTS (Schwartz, 2000). Fibers from the esophagus end at its rostrallateral part, from the stomach at caudal-medial part and from intestines at central and rostral parts of NTS (Schwartz, 2000). NTS is adjacent to the dorsal motor nucleus (DMN) of the vagal nerve within the DVC area. Connection by interneurons to the bodies of DMN completes the vago-vagal reflex pathways controlling various functions of the digestive system. Some signals from the gut are transmitted onward toward higher neural centers via ascending tract from the NTS up to hypothalamus including lateral hypothalamic area (LHA), paraventricular nucleus (PVN), arcuate nucleus of hypothalamus (ARC) and VMH to influence higher autonomic centers such as those involved in appetitive behaviors (Schwartz, 2000). Gastric distension mimicked the feeding process during which, as the stomach fills during ingestion, sensory receptors on the mucosal surface and in the gastric wall transmit the signal of feeding to the central nervous
system (Konturek et al., 2004). In our study, we have revealed the existence of gastric distension responsive neurons in the VMH and both excitatory and inhibitory effects of GD were observed. These results were similar to those by Maddison et al. (Maddison and Horrell, 1979). In a previous study by Qin et al., it was showed that about 77% GD-responsive neurons were GD-E neurons and in this study, 38.7% GD-responsive neurons were GD-E neurons and 60.3% neurons are GD-I neurons. This may indicate that NTS received afferent fibers from mechanical receptor in the stomach directly and the VMH as a feeding center in higher level of the brain integrates gastrointestinal signals indirectly by complex interconnections with other brain regions, because of that ascending visceral afferent input from the second and higher order neurons of NTS by way of relay in the pontine parabranchial nucleus (PBN), to midbrain central gray, LHA, PVN of hypothalamus, amygdala and finally to the visceral sensory cortex. Previous studies suggest that during a meal the gastric vagal fibers innervating the upper gastrointestinal tract can be activated by mechanical, chemical and gut-peptide meal-related stimuli and subsequently transmit those negative-feedback visceral signals through NTS to hypothalamus so as to reduce the meal size (Bary, 2000). The present study on the rats showed that 85% VMH neurons responded to gastric distension and most of them were activated by GES, so we presumed that GES might activate the GD responsive neurons in the VMH via afferent vagal fibers (Liu et al., 2004) to inhibit feeding behaviors by regulating gastric motility (Tang et al., 2000) or feeding related peptides secretion, such as serotonin and dopamine (Megued et al., 1996).
Fig. 3. The majority responses of GD-E neurons to GES with different parameters in VMH. GES was applied for 1 min, so the bar for time did not given. (A) Firing frequency increased 10 s after GD was given and recovered soon after GD terminated. The second arrow stands for the termination of GD. (B) Firing frequency increased during GES1 applied and the excitatory response lasted for about 100 s. (C) Firing frequency did not change during GES2 applied and decreased when GES2 terminated. The inhibitory response lasted for about 30 s. (D) Firing frequency increased soon after GES3 began and lasted for about 130 s. (E) Firing frequency increased and the excitatory response lasted for 1 min. Latency was 1 min.
X. Sun et al. / Neuroscience Research 55 (2006) 451–457
455
Fig. 4. Excitatory effects of GES with different parameters on GD-I neurons. The data were presented as percentage of neurons excited or not excited by GES. * P < 0.05 vs. GES1; #P < 0.01 vs. GES3. E: excited by GES; I + NR: inhibited by GES and no responding to GES.
In a separate study, we found that two hours GES1 increased expression of anorexigenic peptide (oxytocin)-containing neurons in the PVN and decreased expression of orexigenic peptide (orexin)-containing neurons in LHA (Tang et al., 2006b). This has extended our understanding of effects and the mechanism of GES on the neuronal and hormonal regulation in the control of food intake. Furthermore, VMH plays an important role in glucoregulation and this may also affect feeding behaviors (Storlien, 1985). A recent study reported (Zhu et al., 2004) that the gastrointestinal inputs conveyed by the gastric vagal afferent fibers could not only reach the VMH but also mostly impinge on those glycemia-sensitive neurons of the VMN and it further demonstrated that the modulation of the gastric vagal afferent inputs on the VMN glycemia-sensitive neurons might play an important role in the short-term feeding regulation. Whether both of characters that are sensitive to gastric distension and to glycemia could be present in a same neuron in VMH and what is the effect of GES on glycemia-sensitive neurons need us to explore in the future. From the present study we have observed that GES with parameter 1 or 3 was more effective on GD-E neurons compared with other parameters. Since there have been no
reports of inhibitory responses in either vagal or splanchnic afferents during gastric distension (John et al., 2001) the GD-E neurons may have direct relationship to this satiety signal. The present study demonstrated that modulation of the GD-E neurons in the VMH might play an important role in the feeding regulation of GES directly and both the pulse frequency and the train on-time of stimulation were related to the effect of GES. As the report of John et al. (2001) indicated that an inhibitory influence of gastric distension on excitatory taste responses is consistent with a role for distension as an inhibitory feedback signal that enhances satiation and diminishes ingestive taste reactivity. Gastric distension has an inhibitory effect on tasteresponsive neurons in PBN and the global activation of the LHA and the amygdala has a more general excitatory effect on taste responding neurons in NTS (Young et al., 2002). Since the functions of LHA and VMH are opposite, we presumed that global activation of VMH had inhibitory effect on this kind of neurons in NTS and regarding the effect of GES on GD-I neuron in VMH, we presumed that GD-I neurons might take part in this taste-responding feedback. Whether the taste responding neurons in NTS could be inhibited by stimulation of GD-I neurons in VMH need further investigate.
456
X. Sun et al. / Neuroscience Research 55 (2006) 451–457
Fig. 5. The majority responses of GD-I neurons to GES with different parameters in VMH. (A) Firing frequency decreased during GD and recovered as soon as GD ended. (B) The neuron was excited by GES1. Latency was 1 min and excitatory response lasted for 50 s. (C) GES2 had no significant effect on the neuron. (D) The neuron was excited by GES3. Latency was 40 s and responsive time lasted for 100 s. (E) Firing frequency increased during GES4 applied and excitatory response lasted for about 80 s.
We believe that the findings of this study are of clinical significance: more effective parameters of GES in activating VMH neurons are expected to yield a higher decrease in food intake. This is because the VMH is a well recognized satiety center and the neurons being investigated in this study are believed to be related to food intake due to the followings: (1) bilateral VMH lesions are known to result in hyperphagia and obesity, whereas electrical stimulation of the VMH is linked to a reduction of food intake (Bruce, 2006). Electrical stimulation of the VMH directly activates neurons, which is similar to neuronal activation induced by gastric electrical stimulation in this study; (2) the neurons being studied for their responses to gastric electrical stimulation were those that responded to gastric distention which mimics food ingestion. Although feeding was not possible in our study and thus a directly correlation could not be established, recent brain-imaging studies have shown a marked increase in activity in the VMH during feeding. The activation of individual neurons with gastric electrical stimulation observed in this study was in agreement with the global increase in neuronal activity noted in brain-imaging studies during feeding (Bruce, 2006; Cigaina et al., 1996) in separate studies, gastric electrical stimulation with similar parameters used in this study was repeatedly shown to reduce food intake in rats and dogs as well
as humans (Ouyang et al., 2003; Yin and Chen, 2005; Yao et al., 2005). A number of nuclei are involved in the regulation of food intake, such as the paraventricular nucleus (PVN) and arcuate nucleus (ARC) in addition to the VMH. In this study, we have only investigated the neuronal responses to GES in the VMH. A similar study was performed to investigate the effect of GES on neuronal activity in the PVN (Tang et al., 2006a). The findings were in agreement with those reported in this study. The effects of GES on neuronal activity in the ARC may be investigated in a future study. In summary, the findings of this study verify the existence of gastric distension responsive neurons in the VMH and the central neuronal mechanisms of GES involving the VMH. The efficacy of GES on gastric-related neurons in the VMH depends on the pulse frequency, train on-time and pulse width of stimulation. Presumably, these data not only reveal the central mechanism of GES treatment for obesity, but also contribute to the optimization of stimulation parameters in clinic remedy for obesity. References Bary, G.A., 2000. Afferent signals regulating food intake. Proc. Nutr. Soc. 59 (3), 373–384.
X. Sun et al. / Neuroscience Research 55 (2006) 451–457 Bruce, M.K., 2006. The rise, fall, and resurrection of the ventromedial hypothalamus in the regulation of feeding behavior and body weight. Physiol. Behav. 87 (2), 221–244. Chen, J.D.Z., 2004. Mechanisms of an implantable gastric stimulator for obesity. Obes. Surg. 14 (Suppl. 1), 28–32. Cigaina, V., 2002. Gastric pacing as therapy for morbid obesity: preliminary results. Obes. Surg. 12 (Suppl. 1), 12–16. Cigaina, V.V., Saggioro, A., Rigo, V.V., Pinato, G., Ischai, S., 1996. Long-term effects of gastric pacing to reduce feed intake in swine. Obes. Surg. 6 (3), 250–253. D’Argent, J., 2002. Gastric electrical stimulation as therapy of morbid obesity: preliminary results from the French study. Obes. Surg. 12 (Suppl. 1), 21S– 25S. De Luca, M., Segato, G., Busetto, L., Favretti, F., Aigner, F., Weiss, H., de Gheldere, C., Gaggiotti, G., Himpens, J., Limao, J., Scheyer, M., Toppino, M., Zurmeyer, E.L., Bottani, G., Penthaler, H., 2004. Progress in implantable gastric stimulation: summary of results of the European multi-center study. [Clinical Trial. Journal Article. Multicenter Study] Obes. Surg. Suppl.1, S33–S39. Guan, Y., Tang, M., Jiang, Z.Y., Peeters, T.L., 2003. Excitatory effects of motilin in the hippocampus on gastric motility in rats. Brain Res. 984 (1/2), 33–41. Iversen, S., Iversen, L., Saper, C.B., 2000. The hypothalamus integrates autonomic and endocrine functions with behavior. In: Kandel, E.R., Schwartz, J.H., Jessell, T.M. (Eds.), Principles of Neural Science. 4th ed. McGraw-Hill, New York, pp. 974–981. Jeanningros, R., Mei, N., 1980. Vagal and splanchnic effects at the level of the ventromedian nucleus of the hypothalamus (VMH) in the cat. Brain Res. 185 (2), 239–251. John, P.B., Susan, P.T., Joseph, B.T., 2001. Integration of gastric distension and gustatory responses in the parabrachial nucleus. Am. J. Physiol. Regul. Integr. Comp. Physiol. 281 (5), R1581–R1593. Konturek, S.J., Konturek, J.W., Pawlik, T., Brzozowki, T., 2004. Brain–gut axis and its role in the control of food intake. J. Pharm. Pharmacol. 55 (1 Pt 2), 137–154. Lei, Y., Xing, J., Chen, J.D.Z., 2005. Effects and mechanisms of implantable gastric stimulation on proximal gastric distention in conscious dogs. Obes. Surg. 15 (4), 528–533. Liu, J., Qiao, X., Chen, J.D., 2004. Vagal afferent is involved in shortpulse gastric electrical stimulation in rats. Dig. Dis. Sci. 49 (5), 729– 737. Maddison, S., Horrell, R.I., 1979. Hypothalamic unit responses to alimentary perfusions in the anesthetized rat. Brain Res. Bull. 4 (2), 259–266. Martin, L.F., Hunter, S.M., Lauve, R.M., O’Leary, J.P., 1995. Severe obesity: expensive to society, frustrating to treat, but important to confront. South Med. J. 88 (9), 895–902. Megued, M.M., Yang, Z.J., Gleason, J.R., 1996. The gut–brain brain–gut axis in anorexia: toward an understanding of food intake regulation. Nutrition 12 (Suppl. 1), S57–S62. Miller, K.A., 2002. Implantable electrical gastric stimulation to treat morbid obesity in the human: operative technique. Obes. Surg. 12 (Suppl. 1), 17S– 20S.
457
Mokdad, A.H., Serdula, M.K., Dieta, W.H., Bowman, B.A., Marks, J.S., Koplan, J.P., 1999. The spread of the obesity epidemic in the United States, 1991–1998. JAMA 282 (16), 1519–1522. Ouyang, H., Yin, J.Y., Chen, J.D.Z., 2003. Inhibitory effects of chronic gastric electrical stimulation on food intake and weight and their possible mechanisms. Dig. Dis. Sci. 48 (4), 698–705. Paxinos, G., Waston, C., 1998. The Rat Brain in Stereotaxic Coordinates, 4th ed. Academic Press, USA. Plata-Salaman, C.R., 1998. Hypothalamus and the control of feeding: fifteen decades of direct association. Nutrition 14 (1), 67–70. Qin, C., Sun, Y., Chen, J.D.Z., Foreman, R., 2005. Gastric electrical stimulation modulates neuronal activity in nucleus tractus solitarii in rats. Auton. Neurosci. 119 (1), 1–8. Raybould, H.E., 1992. Vagal afferent innervation and the regulation of gastric motor function. In: Ritter, S., Ritter, R.C., Barnes, C.D. (Eds.), Neurooanatomy and Physiology of Abdomeninal Vagal Afferents. CRC Press, Boca Raton, FL, pp. 193–219. Schwartz, G.J., 2000. The role of gastrointestinal vagal afferents in the control of food intake: current prospects. Nutrition 16 (10), 866–873. Storlien, L.H., 1985. The role of the ventromedial hypothalamic area in periprandial glucoregulation. Life Sci. 36 (6), 505–514. Tang, M., Zhang, H.Y., Jiang, Z.Y., Xu, L., Peeters, T.L., 2000. Effect of central administration of motilin on the activity of gastric-related neurons in brain stem and gastric motility of rats. Sheng Li Xue Bao 52 (5), 416–420. Tang, M., Zhang, J., Chen, J.D.Z., 2006a. Central mechanism of gastric electrical stimulation involving neurons in paraventricular nucleus of hypothalamus in rats. Obes. Surg. 16 (3), 344–352. Tang, M., Zhang, J., Xu, L., Chen, J.D.Z., 2006b. Implantable gastric stimulation (IGS) alters expression of oxytocin- and orexin-containing neurons in the hypothalamus of rats. Obes. Surg. 16 (6), 762–769. Woods, S.C., Stricker, E.M., 1999. Food intake and metabolism: gastric distension and calories provide satiety signals. In: Zigmond, M.J., Bloom, F.E., Landis, S.C., Roberts, J.L., Squire, L.R. (Eds.), Fundamental Neuroscience. Academic Press, New York, pp. 1091–1098. Yao, S., Ke, M., Wang, Z., Xu, D., Zhang, Y., Chen, J.D., 2005. Retrograde gastric electrical stimulation reduces food intake and delays gastric emptying in humans: a therapeutic potential for obesity? Dig. Dis. Sci. 50 (9), 1569–1575. Yin, J.Y., Chen, J.D.Z., 2005. Retrograde gastric electrical stimulation reduces food intake and weight in obese rats. Obes. Res. 13 (9), 1580–1587. Young, K.C., Cheng, S.L., David, V.S., 2002. Taste responses of neurons of the hamster solitary nucleus are enhanced by lateral hypothalamic simulation. Neurophysiology 87 (4), 1981–1992. Zhang, Y.P., Ma, C., Wen, Y.Q., Wang, J.J., 2003. Convergence of gastric vagal and cerebellar fastigial nuclear inputs on glycemia-sensitive neurons of lateral hypothalamic area in the rat. Neurosci. Res. 45 (1), 9–16. Zhu, H.B., Chen, J.D.Z., 2005. Implantable gastric stimulation inhibits gastric motility via sympathetic pathway in dogs. Obes. Surg. 15 (1), 95–100. Zhu, J.N., Zhang, Y.P., Song, Y.N., Wang, J.J., 2004. Cerebellar interpositus nuclear and gastric vagal afferent inputs reach and converge onto glycemiasensitive neurons of the ventromedial hypothalamic nucleus in rats. Neurosci. Res. 48 (4), 405–417.
Excitatory effects of gastric electrical stimulation on gastric distension responsive neurons in ventromedial hypothalamus (VMH) in rats Xiangrong Sun a, Ming Tang b, Jing Zhang c, Jiande D.Z. Chen c,d,* a
Department of Pathophysiology, Medical College of Qingdao University, China b Department of Physiology, Medical College of Qingdao University, China c Veteran Research Foundation, VA Medical Center, Oklahoma City, OK, United States d Division of Gastroenterology, University of Texas Medical Branch, Galveston, TX, United States Received 19 January 2006; accepted 2 May 2006
Abstract Introduction: Gastric electrical stimulation (GES) has been used for the treatment of obesity with unclear central mechanisms. The purpose of this study was to investigate the effects of GES on the neuronal activity in the ventromedial hypothalamus (VMH). Methods: Extracellular potentials of single neurons in VMH were recorded in 52 anesthetized rats. Neurons were classified as gastric distensionexcitatory (GD-E) neurons or GD-inhibitory (GD-I) neurons. GES with four sets of parameters was applied for comparison. Results: Eighty two neurons out of 96 (85.41%) in VMH responded to gastric distension (GD). 37.8% were GD-E neurons and 51(62.2%) were GD-I neurons. 55.0%, 17.6%, 77.8%, 14.3% of GD-E neurons were excited by four sets of parameters: GES1 (standard), GES2 (reduced pulse numbers), GES3 (increased pulse width) and GES4 (reduced frequency), respectively. More GD-E neurons were excited by GES3 (P < 0.05 versus GES2 or GES4) and by GES1 (P < 0.02 versus GES2 or GES4). Among the GD-I neurons, 63.6, 37.9, 73.3, and 51.9% neurons were excited by GES1–4, respectively. Conclusion: GES with parameters used for treating obesity excites GD-responsive neurons in VMH. The excitatory effect of GES is related to the strength of stimulation, including pulse frequency and width as well as pulse train on-time. # 2006 Elsevier Ireland Ltd and the Japan Neuroscience Society. All rights reserved. Keywords: Gastric electrical stimulation; Ventromedial hypothalamus; Gastric distension responsive neurons; Rats; Obesity
1. Introduction Obesity is a major public health problem in western societies. This condition affects approximately one-fifth of the U.S. population and is considered to be a contributing factor in 280,000 annual deaths in the U.S. (Mokdad et al., 1999), with an estimated cost of more than $100 billion a year (Martin et al., 1995). Various treatments options are available for obesity, such as diet, exercises, drugs and surgery. However, none of the available therapies is satisfactory and there is an urgent need to develop safe and effective methods to treat patients with morbid obesity.
* Corresponding author at: 1108 the Strand, Room 221, Galveston, TX 77555-0632, United States. Tel.: +1 409 747 3071; fax: +1 409 747 3084. E-mail address: [email protected] (Jiande D.Z. Chen).
Treatment of morbid obesity with gastric electrical stimulation (GES) was first suggested by Cigaina et al. who found decreased food intake and body weight with long-term GES in swine (Cigaina et al., 1996), followed with a number of clinical studies showing the efficacy of GES in reducing weight in obese patients (Miller, 2002; D’Argent, 2002; Cigaina, 2002) with increased satiety and reduced appetite (De Luca et al., 2004). Recently a number of studies have shown peripheral mechanisms of GES therapy for obesity (Chen, 2004). These data suggested that GES induced gastric distention or reduced gastric tone (Zhu and Chen, 2005), inhibited antral contractions (Lei et al., 2005), chronically impaired intestinal gastric slow waves (Ouyang et al., 2003). Except the mediating effect of nucleus tractus solitarri (NTS) was reported (Qin et al., 2005), little is known about the central mechanisms of GES. Food intake is controlled by a highly complex physiological process involving the integrative operation of various brain regions, discrete neuronal pathways, as well as gastrointestinal
0168-0102/$ – see front matter # 2006 Elsevier Ireland Ltd and the Japan Neuroscience Society. All rights reserved. doi:10.1016/j.neures.2006.05.001
452
X. Sun et al. / Neuroscience Research 55 (2006) 451–457
and metabolic factors. Recent investigations (Iversen et al., 2000; Woods and Stricker, 1999) have demonstrated that many of these effects are due to involvement of descending and ascending pathways of the cerebral cortex or the basal forebrain passing through the hypothalamus. During a meal, the distension and secretory activities of the gastrointestinal tract elicited by the ingestion of food may activate the vagal afferent fibers, which comprise the primary neuroanatomical substrate of the neural–gut–brain axis. Several lines of evidence suggest that some meal-associated visceral signals can reach the hypothalamus via the vagal afferent fibers supplying the upper gastrointestinal tract (Jeanningros and Mei, 1980; Zhang et al., 2003; Bary, 2000; Schwartz, 2000), therefore, sensory information from gastrointestinal receptors particularly that transmitted by vagal primary afferents, has been a focus of many investigations due to the recognized importance of this information in the regulation of satiety and gastrointestinal function (Jeanningros and Mei, 1980). It is well established that the VMH in the hypothalamus is closely related to the regulation of feeding behavior and plays an important role in the mediation of satiety. Bilateral VMH lesions consistently caused hyperphagia and obesity, while the opposite effect often seemed to be produced by electrical stimulation of the neurons in the VMH (Raybould, 1992; PlataSalaman, 1998). There is good evidence that implantable electric stimulation of the stomach activates neurons in NTS (Qin et al., 2005) but there is no data regarding roles of the VMH in the central mechanism of GES. It is reasonable to expect that there are gastric distension responsive neurons in the VMH which may faithfully reflect the central representation of gastric mechanoreceptor activation at the level of the hypothalamus and these VMH GD-responsive neuron may involve in regulation of food intake. The present study was undertaken to investigate the effects of implantable gastric electric stimulation with different parameters at the distal stomach on the neuronal activity of gastric-related neurons in the VMH, therefore, to explore the hypothesis that the neurons in the VMH is involved in the central mechanisms of the GES treatment of obesity.
2. Materials and methods 2.1. Animals Adult female or male Wistar rats weighing 250–300 g were used in this study (Institute of Pharmaceutical Research of Qingdao, China). They were housed in a controlled temperature (25 2 8C) and were maintained in individual plastic cages on a 12:12-h light–dark schedule before surgery. Food and water were available ad libitum. The study was approved and all procedures were performed in accordance with institutional guidelines of the Animal Care and Use Committee at Qingdao University.
2.2. Surgery The animals were anesthetized with urethane (1 g/kg, i.p.) and supplemental anesthetics were administrated as needed during the experiment. Anesthesia was confirmed by the absence of paw pinch reflex in response to punching at legs. The heart rate was continuously monitored. The trachea was cannulated to
maintain an open airway. Core temperature was monitored throughout testing via a rectal thermometer and body temperature was regulated to remain at 37 8C through manual adjustments to a heating pad. All surgeries were performed under aseptic conditions in the following order. 2.2.1. Abdominal surgical A laparotomy was made and the stomach was exposed. Gastric contents were removed via a small incision in the fundus wall and the stomach was cleaned with warm isotonic saline. Then a gastric balloon (see below) was gently inserted in the corpus; the wound was ligated around the balloon’s shaft. The balloon catheter was connected to an injector, through which to allow distension of the stomach by infusion of small volumes of saline (3–5 ml, 10–30 s) (Guan et al., 2003). One pair of platinum electrodes (0.3 cm apart used for electric stimulation) was sutured on the serosal surface of gastric antrum close to the lesser curvature. The abdomen was then closed and a small piece of gauze was used as drainage to prevent the accumulation of secretory fluids in the abdomen. 2.2.2. Cranial surgery The rat was placed in a stereotaxic frame (SN-3, Narashige, Tokyo, Japan) and the skull was exposed. A small hole was drilled in the skull to expose the cortex and the dura was cut. Open part of the brain was covered with warm agar (3% in saline) to improve stability for neuronal recording. The stereotaxic coordinates were used in accordance with the atlas of Paxinos and Waston (1998). A one-barrel glass microelectrode filled with 0.5 M sodium acetate and 2% potamine sky blue (tip diameter: 1–3 mm, resistance: 15–20 MV) was advanced in 10 mm steps with the aid of hydraulic micropositioner into the area of the VMH (2.3–2.8 mm posterior to bregma, 0.5–1.0 mm left lateral to the midline and 9.0–10.4 mm below the outer surface of the skull). The microelectrode was used for the extracellular recording of neurons discharge and histological verification of the recording site.
2.3. Procedure 2.3.1. Extracelluar recording Once the microelectrode was advanced into the area of the VMH, the extracellular action potentials of single unit was recorded. The recorded signals were amplified using MEZ8201 amplifier (Nihon Kohden, Tokyo, Japan) and displayed on a oscilloscope (VC-II, Nihon Kohden, Tokyo, Japan). Electrical signals from the amplifier were input into SUMP-PC bioelectric signal processing system and all data were stored in a computer for subsequent analysis. 2.3.2. Identification of GD-responsive neurons After a steady baseline firing rate was obtained for at least 5 min, the test for gastric distension was performed by inflating the gastric balloon with 3–5 ml 37 8C water at a rate of 0.5 ml/s and maintained for 10–30 s. A neuron was identified as GD-responsive neurons if its mean firing rate changed by at least 20% from the mean basal firing level. The GD-responsive neurons were further classified into GD-excitatory (GD-E) neurons and GD-inhibitory (GD-I) neurons according to the spontaneous discharge increased or decreased with GD. 2.3.3. Gastric electrical stimulation study For each GD-responsive neuron, gastric electrical stimulation with four different sets of parameters was applied randomly for 1 min: GES-1 (pulse train of standard parameters—pulse amplitude: 6 mA, pulse width: 0.3 ms, pulse frequency: 40 Hz, train on-time of 2 s and off-time of 3 s); GES-2 (same as GES-1 but reduced on-time to 0.1 s); GES-3 (same as GES-1 but increased pulse width to 3 ms); GES4 (same as GES-1 but reduced pulse frequency to 20 Hz). GES-1 is the set of parameters which are most commonly used in the clinical application of GES for obesity (Miller, 2002; D’Argent, 2002; Cigaina, 2002; De Luca et al., 2004). The other sets of parameters were used as control to study the roles of train on-time, pulse width and pulse frequency involved in GES. The reason we chosen to use the parameters (set #1) identical to the humans was that the frequency of gastric myoelectrical activity in rats is close to that in humans and the impedance between the pair of stimulation electrodes placed in the gastric serosal is similar among different species (rats, dogs and humans). In addition, a number of previous studies performed in our lab revealed similar effective parameters for GES among different species (Chen,
X. Sun et al. / Neuroscience Research 55 (2006) 451–457
453
Table 1 Effects of GES with various parameters on activity of VMH neurons responding to GD Responsive to GD
GES-1
GES-2
GES-3
GES-4
R
NR
R
NR
R
NR
R
NR
GD-E GD-I
11E, 4I 21E, 9I
5 3
3E, 5I 11E, 10I
9 8
14E, 3I 22E, 7I
1 1
2E, 6I 14E, 7I
6 6
Total(R/tested)
45*/53 (84%)
29/46 (63%)
46*,#/48 (96%)
29/41 (70%) *
GES: gastric electrical stimulation, GD: gastric distension (3–5 ml, 10–30 s); R: response to GES; NR: no response to GES. P < 0.05 compared with population of neurons responding to GES2; #P < 0.01 compared to population of neurons responding to GES4. 2004; Zhu and Chen, 2005; Lei et al., 2005; Ouyang et al., 2003; Qin et al., 2005).
2.4. Histology At the end of each experiment, pontamine sky blue was ejected from the recording microelectrode to the site recorded by applying cathodal current (10 mA, 20 min). Animals were transcardially perfused with isotonic saline 300 ml and 10% formalin 300 ml. The brain was then removed and stored in fixative solution at 4 8C for 24 h. Using a freezing micotome, the brain tissue was cut in 20 mm sections to visualize the location of the pontamine sky blue spot. If the recording sites were out of the VMH, the data were excluded from the analysis.
2.5. Statistical analysis The responsive patterns of VMH neurons to the gastric distension and GES were determined by comparing the discharge frequency of recorded cells in responsive window of post-stimulation with the basal firing rates of the cells in control window of pre-stimulation. The raw tracing of neuronal responses to gastric stimuli was processed by software to eliminate GES artifacts. All data are expressed as mean S.E. Paired Student’s t-test was used to study the difference between any paired data. Chi-square analysis was applied to investigate the difference in the neuronal response patterns to four sets of parameters. Comparison of data were considered statistically significant if P < 0.05.
were affected by GES3 (46/48) than GES2 (29/46) (P < 0.001) or GES4 (29/41) (P = 0.003). Besides GES3, GES1 (45/53) had a more significant effect on GD-responsive neurons than GES2 (P = 0.023) but there was no significant difference between GES1 and GES3 (P > 0.05) (Table 1, Fig. 1). Among the GD-E neurons, 55.0, 17.6, 77.8 and 14.3% of tested neurons showed excitatory responses to GES1–4, respectively. More GD-E neurons were excited by GES3 (P < 0.001 versus GES2; P = 0.022 versus GES4) and GES1 (P = 0.002 versus GES2; P = 0.016 versus GES4). Excitatory response of GD-E neurons to GES3 (increased by 343.6 89.2%) were significantly greater than to GES1 (increased by 97.4 33.7%) (P < 0.05). But there was no significant difference in the average responsive duration (Tables 1 and 2; Figs. 2 and 3). Among the GD-I neurons, 63.6, 37.9, 73.3 and 51.8% were excited by GES1–4, respectively. There was no significant difference in the excitatory effect among GES1, 3 and 4; however, GES2 was noted to be the least effective (compared with GES1, P = 0.043 and compared with GES3, P = 0.006) (Tables 1 and 2; Figs. 4
3. Results 3.1. Responses to gastric distension A total of 96 neurons in the VMH were recorded in this study. The spontaneous unit discharges exhibited three different patterns: phasic, continuous and single. Among 92 recorded neurons, 82 units responded to gastric distension (82/96, 85.4%). Thirty-one (31/82, 37.8%) were GD-E neurons and 51 (51/82, 62.2%) were GD-I neurons. The firing frequency of GD-E neurons was (1.51 0.33) Hz before gastric distension and increased to (2.84 0.62) Hz during gastric distension (P < 0.01) and the firing frequency of GD-I neurons firing frequency decreased from (1.20 0.20) Hz at baseline to (0.61 0.10) Hz during gastric distension. The response latency time was between 0 and 20 s. The average duration of excitatory effect was (86.7 15.6) s and the inhibitory effect was (50.1 9.8) s. 3.2. Effects of GES on GD-responsive neurons The percentage of GD-responsive neurons responded to GES1–4 were 84, 63, 96 and 70%, respectively. More neurons
Fig. 1. Effects of GES with different parameters on GD-responsive neurons. The data were presented as percentage of neurons responding or not responding to GES. *P < 0.05 vs. GES2; #P < 0.01 vs. GES2; §P < 0.01 vs. GES4. R: responding to GES; NR: no responding to GES.
Fig. 2. Excitatory effects of GES with different parameters on GD-E neurons. The data were presented as percentage of neurons excited or not excited by GES. *P < 0.05 vs. GES4; #P < 0.01 vs. GES2. E: excited by GES; I + NR: inhibited and no responded.
454
X. Sun et al. / Neuroscience Research 55 (2006) 451–457
Table 2 The effects of GES1 and GES3 on GD-E and GD-I neurons in VMH Responsive type
GES parameters
Firing frequency before GES (Hz)
Firing frequency after GES (Hz)
Firing varied rate (%)
Latency (s)
Duration (s)
GD-E
GES1 GES3
1.50 0.34 1.36 0.35
2.63 0.83 3.47 0.58
97.44 33.70 343.60 89.21*
0–120 0–70
85.56 14.15 76.47 8.16
GD-I
GES1 GES3
1.05 0.25 1.47 0.41
1.49 0.30 7.58 4.61
112.00 14.67 366.30 87.20*
0–190 0–20#
119.2 14.70 104.1 14.89
*
P < 0.05 vs. GES1; #P < 0.01 vs. GES1.
and 5). In both GD-E and GD-I neurons, the latency of GES3 was markedly shorter than GES1 (Table 2). 4. Discussion GES or implantable gastric stimulation is currently being investigated as an alternative for the treatment of obesity. The present study demonstrated the effect of GES on gastricresponsive afferent neurons in VMH and thereby provided direct evidence of a central neural mechanism that might contribute to the effects of GES on feeding or obesity. Our results showed that parameters 1 and 3 were more effective than parameters 2 and 4 in activating gastric-related neurons in the VMH. This result suggested that the GES used in the previous clinical studies for treating obesity might affect feeding by exciting neurons in the VMH, which is known to have a strong effect in terminating feeding. Compared with parameter 1, parameter 3 showed more intense effect on gastric distension responsive neurons in the VMH. However, it should be noted that the difference in the effect, although significant with these neurons, was to a much lesser degree than the difference in pulse width (10 times) or energy consumption, which is one of major concerns with an implant device. It is clear that the neuronal responses of the VMH to gastric distension observed in the study represent the centrally directed traffic from the activation of vagal gastric mechanoreceptors. Signals from a variety of sensors in the gut that respond to mechanical (distention, contraction) and chemical stimuli transmit to CNS by visceral afferent fibers in vagal and sympathetic nerves. Most of afferent vagal fibers pass through nodose ganglia to terminate on NTS (Schwartz, 2000). Fibers from the esophagus end at its rostrallateral part, from the stomach at caudal-medial part and from intestines at central and rostral parts of NTS (Schwartz, 2000). NTS is adjacent to the dorsal motor nucleus (DMN) of the vagal nerve within the DVC area. Connection by interneurons to the bodies of DMN completes the vago-vagal reflex pathways controlling various functions of the digestive system. Some signals from the gut are transmitted onward toward higher neural centers via ascending tract from the NTS up to hypothalamus including lateral hypothalamic area (LHA), paraventricular nucleus (PVN), arcuate nucleus of hypothalamus (ARC) and VMH to influence higher autonomic centers such as those involved in appetitive behaviors (Schwartz, 2000). Gastric distension mimicked the feeding process during which, as the stomach fills during ingestion, sensory receptors on the mucosal surface and in the gastric wall transmit the signal of feeding to the central nervous
system (Konturek et al., 2004). In our study, we have revealed the existence of gastric distension responsive neurons in the VMH and both excitatory and inhibitory effects of GD were observed. These results were similar to those by Maddison et al. (Maddison and Horrell, 1979). In a previous study by Qin et al., it was showed that about 77% GD-responsive neurons were GD-E neurons and in this study, 38.7% GD-responsive neurons were GD-E neurons and 60.3% neurons are GD-I neurons. This may indicate that NTS received afferent fibers from mechanical receptor in the stomach directly and the VMH as a feeding center in higher level of the brain integrates gastrointestinal signals indirectly by complex interconnections with other brain regions, because of that ascending visceral afferent input from the second and higher order neurons of NTS by way of relay in the pontine parabranchial nucleus (PBN), to midbrain central gray, LHA, PVN of hypothalamus, amygdala and finally to the visceral sensory cortex. Previous studies suggest that during a meal the gastric vagal fibers innervating the upper gastrointestinal tract can be activated by mechanical, chemical and gut-peptide meal-related stimuli and subsequently transmit those negative-feedback visceral signals through NTS to hypothalamus so as to reduce the meal size (Bary, 2000). The present study on the rats showed that 85% VMH neurons responded to gastric distension and most of them were activated by GES, so we presumed that GES might activate the GD responsive neurons in the VMH via afferent vagal fibers (Liu et al., 2004) to inhibit feeding behaviors by regulating gastric motility (Tang et al., 2000) or feeding related peptides secretion, such as serotonin and dopamine (Megued et al., 1996).
Fig. 3. The majority responses of GD-E neurons to GES with different parameters in VMH. GES was applied for 1 min, so the bar for time did not given. (A) Firing frequency increased 10 s after GD was given and recovered soon after GD terminated. The second arrow stands for the termination of GD. (B) Firing frequency increased during GES1 applied and the excitatory response lasted for about 100 s. (C) Firing frequency did not change during GES2 applied and decreased when GES2 terminated. The inhibitory response lasted for about 30 s. (D) Firing frequency increased soon after GES3 began and lasted for about 130 s. (E) Firing frequency increased and the excitatory response lasted for 1 min. Latency was 1 min.
X. Sun et al. / Neuroscience Research 55 (2006) 451–457
455
Fig. 4. Excitatory effects of GES with different parameters on GD-I neurons. The data were presented as percentage of neurons excited or not excited by GES. * P < 0.05 vs. GES1; #P < 0.01 vs. GES3. E: excited by GES; I + NR: inhibited by GES and no responding to GES.
In a separate study, we found that two hours GES1 increased expression of anorexigenic peptide (oxytocin)-containing neurons in the PVN and decreased expression of orexigenic peptide (orexin)-containing neurons in LHA (Tang et al., 2006b). This has extended our understanding of effects and the mechanism of GES on the neuronal and hormonal regulation in the control of food intake. Furthermore, VMH plays an important role in glucoregulation and this may also affect feeding behaviors (Storlien, 1985). A recent study reported (Zhu et al., 2004) that the gastrointestinal inputs conveyed by the gastric vagal afferent fibers could not only reach the VMH but also mostly impinge on those glycemia-sensitive neurons of the VMN and it further demonstrated that the modulation of the gastric vagal afferent inputs on the VMN glycemia-sensitive neurons might play an important role in the short-term feeding regulation. Whether both of characters that are sensitive to gastric distension and to glycemia could be present in a same neuron in VMH and what is the effect of GES on glycemia-sensitive neurons need us to explore in the future. From the present study we have observed that GES with parameter 1 or 3 was more effective on GD-E neurons compared with other parameters. Since there have been no
reports of inhibitory responses in either vagal or splanchnic afferents during gastric distension (John et al., 2001) the GD-E neurons may have direct relationship to this satiety signal. The present study demonstrated that modulation of the GD-E neurons in the VMH might play an important role in the feeding regulation of GES directly and both the pulse frequency and the train on-time of stimulation were related to the effect of GES. As the report of John et al. (2001) indicated that an inhibitory influence of gastric distension on excitatory taste responses is consistent with a role for distension as an inhibitory feedback signal that enhances satiation and diminishes ingestive taste reactivity. Gastric distension has an inhibitory effect on tasteresponsive neurons in PBN and the global activation of the LHA and the amygdala has a more general excitatory effect on taste responding neurons in NTS (Young et al., 2002). Since the functions of LHA and VMH are opposite, we presumed that global activation of VMH had inhibitory effect on this kind of neurons in NTS and regarding the effect of GES on GD-I neuron in VMH, we presumed that GD-I neurons might take part in this taste-responding feedback. Whether the taste responding neurons in NTS could be inhibited by stimulation of GD-I neurons in VMH need further investigate.
456
X. Sun et al. / Neuroscience Research 55 (2006) 451–457
Fig. 5. The majority responses of GD-I neurons to GES with different parameters in VMH. (A) Firing frequency decreased during GD and recovered as soon as GD ended. (B) The neuron was excited by GES1. Latency was 1 min and excitatory response lasted for 50 s. (C) GES2 had no significant effect on the neuron. (D) The neuron was excited by GES3. Latency was 40 s and responsive time lasted for 100 s. (E) Firing frequency increased during GES4 applied and excitatory response lasted for about 80 s.
We believe that the findings of this study are of clinical significance: more effective parameters of GES in activating VMH neurons are expected to yield a higher decrease in food intake. This is because the VMH is a well recognized satiety center and the neurons being investigated in this study are believed to be related to food intake due to the followings: (1) bilateral VMH lesions are known to result in hyperphagia and obesity, whereas electrical stimulation of the VMH is linked to a reduction of food intake (Bruce, 2006). Electrical stimulation of the VMH directly activates neurons, which is similar to neuronal activation induced by gastric electrical stimulation in this study; (2) the neurons being studied for their responses to gastric electrical stimulation were those that responded to gastric distention which mimics food ingestion. Although feeding was not possible in our study and thus a directly correlation could not be established, recent brain-imaging studies have shown a marked increase in activity in the VMH during feeding. The activation of individual neurons with gastric electrical stimulation observed in this study was in agreement with the global increase in neuronal activity noted in brain-imaging studies during feeding (Bruce, 2006; Cigaina et al., 1996) in separate studies, gastric electrical stimulation with similar parameters used in this study was repeatedly shown to reduce food intake in rats and dogs as well
as humans (Ouyang et al., 2003; Yin and Chen, 2005; Yao et al., 2005). A number of nuclei are involved in the regulation of food intake, such as the paraventricular nucleus (PVN) and arcuate nucleus (ARC) in addition to the VMH. In this study, we have only investigated the neuronal responses to GES in the VMH. A similar study was performed to investigate the effect of GES on neuronal activity in the PVN (Tang et al., 2006a). The findings were in agreement with those reported in this study. The effects of GES on neuronal activity in the ARC may be investigated in a future study. In summary, the findings of this study verify the existence of gastric distension responsive neurons in the VMH and the central neuronal mechanisms of GES involving the VMH. The efficacy of GES on gastric-related neurons in the VMH depends on the pulse frequency, train on-time and pulse width of stimulation. Presumably, these data not only reveal the central mechanism of GES treatment for obesity, but also contribute to the optimization of stimulation parameters in clinic remedy for obesity. References Bary, G.A., 2000. Afferent signals regulating food intake. Proc. Nutr. Soc. 59 (3), 373–384.
X. Sun et al. / Neuroscience Research 55 (2006) 451–457 Bruce, M.K., 2006. The rise, fall, and resurrection of the ventromedial hypothalamus in the regulation of feeding behavior and body weight. Physiol. Behav. 87 (2), 221–244. Chen, J.D.Z., 2004. Mechanisms of an implantable gastric stimulator for obesity. Obes. Surg. 14 (Suppl. 1), 28–32. Cigaina, V., 2002. Gastric pacing as therapy for morbid obesity: preliminary results. Obes. Surg. 12 (Suppl. 1), 12–16. Cigaina, V.V., Saggioro, A., Rigo, V.V., Pinato, G., Ischai, S., 1996. Long-term effects of gastric pacing to reduce feed intake in swine. Obes. Surg. 6 (3), 250–253. D’Argent, J., 2002. Gastric electrical stimulation as therapy of morbid obesity: preliminary results from the French study. Obes. Surg. 12 (Suppl. 1), 21S– 25S. De Luca, M., Segato, G., Busetto, L., Favretti, F., Aigner, F., Weiss, H., de Gheldere, C., Gaggiotti, G., Himpens, J., Limao, J., Scheyer, M., Toppino, M., Zurmeyer, E.L., Bottani, G., Penthaler, H., 2004. Progress in implantable gastric stimulation: summary of results of the European multi-center study. [Clinical Trial. Journal Article. Multicenter Study] Obes. Surg. Suppl.1, S33–S39. Guan, Y., Tang, M., Jiang, Z.Y., Peeters, T.L., 2003. Excitatory effects of motilin in the hippocampus on gastric motility in rats. Brain Res. 984 (1/2), 33–41. Iversen, S., Iversen, L., Saper, C.B., 2000. The hypothalamus integrates autonomic and endocrine functions with behavior. In: Kandel, E.R., Schwartz, J.H., Jessell, T.M. (Eds.), Principles of Neural Science. 4th ed. McGraw-Hill, New York, pp. 974–981. Jeanningros, R., Mei, N., 1980. Vagal and splanchnic effects at the level of the ventromedian nucleus of the hypothalamus (VMH) in the cat. Brain Res. 185 (2), 239–251. John, P.B., Susan, P.T., Joseph, B.T., 2001. Integration of gastric distension and gustatory responses in the parabrachial nucleus. Am. J. Physiol. Regul. Integr. Comp. Physiol. 281 (5), R1581–R1593. Konturek, S.J., Konturek, J.W., Pawlik, T., Brzozowki, T., 2004. Brain–gut axis and its role in the control of food intake. J. Pharm. Pharmacol. 55 (1 Pt 2), 137–154. Lei, Y., Xing, J., Chen, J.D.Z., 2005. Effects and mechanisms of implantable gastric stimulation on proximal gastric distention in conscious dogs. Obes. Surg. 15 (4), 528–533. Liu, J., Qiao, X., Chen, J.D., 2004. Vagal afferent is involved in shortpulse gastric electrical stimulation in rats. Dig. Dis. Sci. 49 (5), 729– 737. Maddison, S., Horrell, R.I., 1979. Hypothalamic unit responses to alimentary perfusions in the anesthetized rat. Brain Res. Bull. 4 (2), 259–266. Martin, L.F., Hunter, S.M., Lauve, R.M., O’Leary, J.P., 1995. Severe obesity: expensive to society, frustrating to treat, but important to confront. South Med. J. 88 (9), 895–902. Megued, M.M., Yang, Z.J., Gleason, J.R., 1996. The gut–brain brain–gut axis in anorexia: toward an understanding of food intake regulation. Nutrition 12 (Suppl. 1), S57–S62. Miller, K.A., 2002. Implantable electrical gastric stimulation to treat morbid obesity in the human: operative technique. Obes. Surg. 12 (Suppl. 1), 17S– 20S.
457
Mokdad, A.H., Serdula, M.K., Dieta, W.H., Bowman, B.A., Marks, J.S., Koplan, J.P., 1999. The spread of the obesity epidemic in the United States, 1991–1998. JAMA 282 (16), 1519–1522. Ouyang, H., Yin, J.Y., Chen, J.D.Z., 2003. Inhibitory effects of chronic gastric electrical stimulation on food intake and weight and their possible mechanisms. Dig. Dis. Sci. 48 (4), 698–705. Paxinos, G., Waston, C., 1998. The Rat Brain in Stereotaxic Coordinates, 4th ed. Academic Press, USA. Plata-Salaman, C.R., 1998. Hypothalamus and the control of feeding: fifteen decades of direct association. Nutrition 14 (1), 67–70. Qin, C., Sun, Y., Chen, J.D.Z., Foreman, R., 2005. Gastric electrical stimulation modulates neuronal activity in nucleus tractus solitarii in rats. Auton. Neurosci. 119 (1), 1–8. Raybould, H.E., 1992. Vagal afferent innervation and the regulation of gastric motor function. In: Ritter, S., Ritter, R.C., Barnes, C.D. (Eds.), Neurooanatomy and Physiology of Abdomeninal Vagal Afferents. CRC Press, Boca Raton, FL, pp. 193–219. Schwartz, G.J., 2000. The role of gastrointestinal vagal afferents in the control of food intake: current prospects. Nutrition 16 (10), 866–873. Storlien, L.H., 1985. The role of the ventromedial hypothalamic area in periprandial glucoregulation. Life Sci. 36 (6), 505–514. Tang, M., Zhang, H.Y., Jiang, Z.Y., Xu, L., Peeters, T.L., 2000. Effect of central administration of motilin on the activity of gastric-related neurons in brain stem and gastric motility of rats. Sheng Li Xue Bao 52 (5), 416–420. Tang, M., Zhang, J., Chen, J.D.Z., 2006a. Central mechanism of gastric electrical stimulation involving neurons in paraventricular nucleus of hypothalamus in rats. Obes. Surg. 16 (3), 344–352. Tang, M., Zhang, J., Xu, L., Chen, J.D.Z., 2006b. Implantable gastric stimulation (IGS) alters expression of oxytocin- and orexin-containing neurons in the hypothalamus of rats. Obes. Surg. 16 (6), 762–769. Woods, S.C., Stricker, E.M., 1999. Food intake and metabolism: gastric distension and calories provide satiety signals. In: Zigmond, M.J., Bloom, F.E., Landis, S.C., Roberts, J.L., Squire, L.R. (Eds.), Fundamental Neuroscience. Academic Press, New York, pp. 1091–1098. Yao, S., Ke, M., Wang, Z., Xu, D., Zhang, Y., Chen, J.D., 2005. Retrograde gastric electrical stimulation reduces food intake and delays gastric emptying in humans: a therapeutic potential for obesity? Dig. Dis. Sci. 50 (9), 1569–1575. Yin, J.Y., Chen, J.D.Z., 2005. Retrograde gastric electrical stimulation reduces food intake and weight in obese rats. Obes. Res. 13 (9), 1580–1587. Young, K.C., Cheng, S.L., David, V.S., 2002. Taste responses of neurons of the hamster solitary nucleus are enhanced by lateral hypothalamic simulation. Neurophysiology 87 (4), 1981–1992. Zhang, Y.P., Ma, C., Wen, Y.Q., Wang, J.J., 2003. Convergence of gastric vagal and cerebellar fastigial nuclear inputs on glycemia-sensitive neurons of lateral hypothalamic area in the rat. Neurosci. Res. 45 (1), 9–16. Zhu, H.B., Chen, J.D.Z., 2005. Implantable gastric stimulation inhibits gastric motility via sympathetic pathway in dogs. Obes. Surg. 15 (1), 95–100. Zhu, J.N., Zhang, Y.P., Song, Y.N., Wang, J.J., 2004. Cerebellar interpositus nuclear and gastric vagal afferent inputs reach and converge onto glycemiasensitive neurons of the ventromedial hypothalamic nucleus in rats. Neurosci. Res. 48 (4), 405–417.