Effects of Feeding-Related Peptides on Neuronal Oscillation in the Ventromedial Hypothalamus

The ventromedial hypothalamus (VMH) plays an important role in feeding behavior, obesity, and thermoregulation. The VMH contains glucose-sensing neurons, the firing of which depends on the level of extracellular glucose and which are involved in maintaining the blood glucose level via the sympathetic nervous system. The VMH also expresses various receptors of the peptides related to feeding. However, it is not well-understood whether the action of feeding-related peptides mediates the activity of glucose-sensing neurons in the VMH. In the present study, we examined the effects of feeding-related peptides on the burst-generating property of the VMH. Superfusion with insulin, pituitary adenylate cyclase-activating polypeptide, corticotropin-releasing factor, and orexin increased the frequency of the VMH oscillation. In contrast, superfusion with leptin, cholecystokinin, cocaine- and amphetamine-regulated transcript, galanin, ghrelin, and neuropeptide Y decreased the frequency of the oscillation. Our findings indicated that the frequency changes of VMH oscillation in response to the application of feeding-related peptides showed a tendency similar to changes of sympathetic nerve activity in response to the application of these substances to the brain.


Introduction
The ventromedial hypothalamus (VMH) plays an important role in feeding behavior, obesity, and thermoregulation [1]. The VMH contains glucose-sensing neurons, the firing of which depends on the level of extracellular glucose [2]. The information on changes in glucose levels detected by glucose-sensing neurons in the VMH is transmitted to the sympathetic nervous system in order to maintain the blood glucose levels [3]. The VMH also expresses various receptors of the peptides related to feeding [4]: insulin [5], neuropeptide Y (NPY) [6], orexin [7], galanin [8], ghrelin [9], cocaine-and amphetamine-regulated transcript (CART) [10], cholecystokinin (CCK)-A [11], corticotropin-releasing factor (CRF) [12] and pituitary adenylate cyclase-activating polypeptide (PACAP) [13]. These neuropeptides are divided largely into two groups based on their effects on feeding behavior: anorexigenic peptides (insulin, PACAP, CRF, leptin, CCK, and CART) [14] and orexigenic peptides (orexin, galanin, ghrelin, and NPY) [1,15,16]. However, it is not clear whether these peptides work via receptors in the VMH. In addition, it has been suggested that the feeding-related peptides affect sympathetic nerve activity (SNA) [1], whereas it is not well understood whether the action of feeding-related peptides mediates the activity of glucose-sensing neurons in the VMH.

Drugs
Orexin, insulin, CRF, PACAP, galanin, ghrelin, NPY, CCK, and CART were supplied from the Peptide Institute, Inc. (Osaka, Japan). Leptin was supplied from Sigma Aldrich (Tokyo, Japan). The concentration of each drug was referred from previous papers [20][21][22][23][24][25][26][27][28][29]. We also tested the effects of the peptides at 10-100 nM in preliminary experiments and determined the minimum concentration that caused clear and reversible effects. All drugs were dissolved in standard ACSF and applied to the preparation for 10-15 min by superfusion. Field potential recordings were attempted from the bilateral VMH region (see below), and one preparation was tested for less than three different conditions, when the VMH activity recovered after 20-30 min washout. Thus, a total of 97 field potential recordings were examined.

Electrophysiological Measurements of Neuronal Activity within the VMH
For the extracellular recordings by the field potential technique, glass electrodes (50-150 µm diameter) were placed on the bilateral VMH region of the preparation (Figure 1) [17]. Neuronal activities were recorded by an AC amplifier (MEG-5200, Nihon Kohden, Tokyo, Japan) through a 0.5-Hz low-cut filter and stored on hard-disc memory through a PowerLab system (ADInstruments, Castle Hill, Australia) with a 4 kHz sampling rate. The field potentials were rectified and integrated with a 0.  The dashed lines denote the VMH. This preparation was briefly stained by 0.05% methylene blue. The field potential (fp) was recorded with a glass pipette set within the VMH. f: fornix.

Statistics
All of the data are presented as means ± SD. Differences in the VMH response to pre and post superfusion with all peptides and glucose were examined by paired t-test with the use of GraphPad InStat (version 3.01, GraphPad Software Inc., La Jolla, CA, USA). p-values <0.05 were considered to indicate statistical significance.

Results
The average frequency in the control was 0.068 ± 0.016 Hz (n = 97). Superfusion with the anorexigenic neuropeptides PACAP (50 nM) and CRF (50 nM) increased the frequency of the VMH oscillation by 11.9% ± 10.3% (n = 6, p < 0.05) and 43.1% ± 24.5% (n = 12, p < 0.01), respectively ( Figure  2A,B). , and after 20-min washout (wash). The frequency of the oscillation was increased by the application of each of the peptides. Note that the application of PACAP or CRF also increased background tonic activity. FP: raw field potential; ʃFP: integrated field potential; wash: washout. The dashed lines denote the VMH. This preparation was briefly stained by 0.05% methylene blue. The field potential (fp) was recorded with a glass pipette set within the VMH. f: fornix.

Statistics
All of the data are presented as means ± SD. Differences in the VMH response to pre and post superfusion with all peptides and glucose were examined by paired t-test with the use of GraphPad InStat (version 3.01, GraphPad Software Inc., La Jolla, CA, USA). p-values < 0.05 were considered to indicate statistical significance.

Statistics
All of the data are presented as means ± SD. Differences in the VMH response to pre and post superfusion with all peptides and glucose were examined by paired t-test with the use of GraphPad InStat (version 3.01, GraphPad Software Inc., La Jolla, CA, USA). p-values <0.05 were considered to indicate statistical significance.

Results
The average frequency in the control was 0.068 ± 0.016 Hz (n = 97). Superfusion with the anorexigenic neuropeptides PACAP (50 nM) and CRF (50 nM) increased the frequency of the VMH oscillation by 11.9% ± 10.3% (n = 6, p < 0.05) and 43.1% ± 24.5% (n = 12, p < 0.01), respectively ( Figure  2A,B). , and after 20-min washout (wash). The frequency of the oscillation was increased by the application of each of the peptides. Note that the application of PACAP or CRF also increased background tonic activity. FP: raw field potential; ʃFP: integrated field potential; wash: washout. a and b correspond to records at "a" and "b" in (A,B), and after 20-min washout (wash). The frequency of the oscillation was increased by the application of each of the peptides. Note that the application of PACAP or CRF also increased background tonic activity. FP: raw field potential; FP: integrated field potential; wash: washout.
Although the oscillations on the right and left sides of the VMH were not synchronized, the two sides showed no differences in the changes of frequency in response to the neuropeptides. Application of CRF that facilitated the VMH oscillation induced an increase in background tonic activity, as shown in Figure 2B. A similar increase in background tonic activity was also observed with the application of orexin (trace b in Figure 5) or PACAP (trace b in Figure 2A), which facilitated the VMH oscillation. Interestingly, some peptides (i.e., CCK, CART, and ghrelin) that depressed the VMH oscillation also induced background tonic activity in 10-15% of the preparations (trace b in Figure 4B as an example with CCK). correspond to records at "a" and "b" in (A), and after 20-min washout (wash). The frequency of the oscillation was increased by the application of insulin. FP: raw field potential; FP: integrated field potential; wash: washout.
Although the oscillations on the right and left sides of the VMH were not synchronized, the two sides showed no differences in the changes of frequency in response to the neuropeptides. Application of CRF that facilitated the VMH oscillation induced an increase in background tonic activity, as shown in Figure 2B. A similar increase in background tonic activity was also observed with the application of orexin (trace b in Figure 5) or PACAP (trace b in Figure 2A), which facilitated the VMH oscillation. Interestingly, some peptides (i.e., CCK, CART, and ghrelin) that depressed the VMH oscillation also induced background tonic activity in 10-15% of the preparations (trace b in Figure 4B as an example with CCK).      oscillation was increased by the application of orexin. Note that its application increased background tonic activity.
Thus, there are reciprocal interactions between some feeding-related peptides and the sympathetic nervous system, as previously suggested [1]. Our findings indicated that the frequency changes of the VMH oscillation in response to the application of feeding-related peptides and glucose [17] showed tendencies similar to those of SNA in response to the application of these substances to the brain, except for leptin, CART, and CCK. These findings are consistent with our previous report in which we showed that the VMH oscillation might be involved in frequency modulation of the SNA. It is not clear whether CCK induces excitation or inhibition of the sympathetic nervous system. The systemic application (intravenously [38] or intra-arterially [14]) of
Thus, there are reciprocal interactions between some feeding-related peptides and the sympathetic nervous system, as previously suggested [1]. Our findings indicated that the frequency changes of the VMH oscillation in response to the application of feeding-related peptides and glucose [17] showed tendencies similar to those of SNA in response to the application of these substances to the brain, except for leptin, CART, and CCK. These findings are consistent with our previous report in which we showed that the VMH oscillation might be involved in frequency modulation of the SNA. It is not clear whether CCK induces excitation or inhibition of the sympathetic nervous system. The systemic application (intravenously [38] or intra-arterially [14]) of CCK caused sympathoinhibition, whereas the intracerebroventricular application of CCK resulted in sympathoexcitation [34]. CCK reduced the frequency of the VMH oscillation in the present study, whereas it enhanced the background tonic activity in the VMH. Similar effects were also observed in some preparations when CART was applied. An increase of the background tonic activity might correlate with the excitatory effect of CCK and CART on SNA [33,34]. Previous electrophysiological studies reported that VMH neurons were either depolarized or hyperpolarized by leptin. This difference in response depended on the subpopulation of the VMH [39,40]. In the present study, the frequency of the VMH oscillation decreased by superfusion with leptin. Therefore, the subpopulation of the VMH neurons that is hyperpolarized by leptin might be involved in the generation of VMH oscillation.
Energy expenditure by sympathoexcitation induces the loss of body weight. It has been reported that the activation of steroidogenic factor 1 (SF1) neurons induces an increase in the blood glucose level by gluconeogenesis via sympathoexcitation [41]. Hypothalamic orexin stimulates feeding-associated glucose utilization in skeletal muscle via the sympathetic nervous system [7]. In addition, immunohistochemical analysis revealed that the mouse VMH contains many cells positive for SF1 [7]. PACAP neurons within the VMH are thought to be SF1 neurons, which affect the sympathetic nervous system via leptin [13]. Insulin reduced the firing frequency of the SF1 neurons within the VMH, which might contribute to obesity development [42], although our present results indicated that insulin caused no significant (or slight excitatory) effects on the VMH oscillation, suggesting presence of subpopulations with different properties. Because SF1 neurons in the VMH project to cardiorespiratory centers in the medulla [43], we presume that one of the cellular characteristics of the VMH oscillator might be SF1 neurons, although histological identification remains a subject for future study.
Stimulation of the VMH induced excitation of the renal [44] and interscapular brown adipose tissue (iBAT) [45] SNA. The VMH stimulation also activated hepatic gluconeogenesis via the sympathetic nervous system [46]. These findings suggest that the VMH and sympathetic nervous system have an important role in the counter-regulatory response (CRR) against hypoglycemia [12]. Furthermore, glucose-inhibited neurons in the VMH may play a key role in the CRR [47]. The activation of type 1 CRF receptors in the VMH induced CRR [12]. SF1 neurons of the VMH are the specific target of lateral parabrachial nucleus (LPBN) CCK glucoregulatory neurons. This discrete CCK (LPBN) and SF1 (VMH) neurocircuit is both necessary and sufficient for induction of the CRR [48,49]. Thus, these previous studies suggested that some feeding-related peptides affect the CRR via the VMH. In addition, the stimulation of SF1 neurons within the VMH induced hyperglycemia [50].
The pattern of the VMH oscillation in response to the application of glucose or neuropeptides is consistent with the hypothesis that these substances act on the sympathetic nervous system involved in the CRR via VMH oscillation. However, the secretion of ghrelin with hunger does not stimulate the sympathetic nervous system [36], and our results indicated that ghrelin inhibited VMH oscillation. Therefore, the inhibition of the VMH activity by some neuropeptides such as ghrelin may exert their effect via iBAT-SNA rather than by the CRR.

Conclusions
VMH oscillation showed sensitivity to various feeding-related peptides and glucose. Changes in the frequency of the VMH oscillation may affect energy expenditure and the CRR against hypoglycemia via SNA.

Conflicts of Interest:
The authors declare no conflict of interest.