| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Department of Neuroscience (N.U., S.P.K.), University of Florida McKnight Brain Institute, College of Medicine, Gainesville, Florida 32610-0244 Department of Physiology and Functional Genomics (M.G.D., P.S.K.), Gainesville, Florida 32610-0274; and Division of Diabetes, Digestive, and Kidney Diseases (N.U., A.I.), Department of Clinical Molecular Medicine, Kobe University Graduate School of Medicine, Kobe 650-0017, Japan
Address all correspondence and requests for reprints to: Satya P. Kalra, Ph.D., Department of Neuroscience, McKnight Brain Institute, P.O. Box 100244, Gainesville, Florida 32610-0244. E-mail: skalra{at}mbi.ufl.edu.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
, free fatty acids and insulin, concomitant with normoglycemia; and elevated ghrelin levels for extended period. Body weight and plasma levels of leptin and metabolic variables were suppressed to a lesser extent in rAAV-lep wt mice without decreasing food intake. The sustained high leptin transgene expression decreased only the dark-phase phagia in both genotypes, but wt mice escaped from leptin restraint during the lights-on phase, resulting in normal overall food intake. Leptin administration rapidly decreased plasma gastric ghrelin and adipocyte adiponectin but not TNF levels, thereby demonstrating a peripheral restraining action of leptin on the secretion of hormones of varied origins. Whereas ghrelin administration readily stimulated feeding in controls, it was completely ineffective in rAAV-lep-treated wt mice. Thus, leptin expressed locally in the hypothalamus counteracted the central orexigenic effects of peripheral ghrelin. Cumulatively, these results identify newer central and peripheral modulatory influences of leptin on hormonal signals of disparate origin implicated in weight homeostasis and metabolic disorders. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
, and pancreatic insulin have been implicated also in the pathogenesis of hyperglycemia and hyperinsulinemia, the concomitants of type 2 diabetes (7, 8, 9). Furthermore, anorexigenic leptin is a hormone with pleiotropic effects that include the pancreatic insulin and blood glucose homeostasis (10, 11). Similarly, orexigenic ghrelin has the potential to influence several similar metabolic variables (12, 13, 14). However, the relationship of the adipocyte hormones, adiponectin and TNF, with leptin or ghrelin, two of the primary afferent signals in integration of weight homeostasis, is not known. Similarly, the neural and hormonal factors that regulate ghrelin, adiponectin, and TNF remain to be determined. Therefore, to begin to understand the interrelationship among these metabolic and afferent signals, the first objective of this study was to evaluate the effects of experimentally induced shifts in leptin, insulin, and ghrelin levels on adiponectin and TNF levels. Recently we reported that enhanced leptin transgene expression evoked by an injection of recombinant adeno-associated virus encoding the leptin gene (rAAV-lep) into various hypothalamic sites, including the arcuate nucleus (ARC)-paraventricular nucleus (PVN) axis, severely depleted adipose tissue and depressed weight gain by increasing energy expenditure with or without decreasing energy intake in rats (15, 16, 17, 18). Consistent with the loss of adipose tissue, blood leptin levels were reduced by more than 80% and free fatty acids (FFAs), triglycerides, and insulin levels were drastically suppressed. Surprisingly, ghrelin secretion was more than 2-fold higher but was not accompanied by the expected increase in food intake in these hypoleptinemic rats. This reciprocity of reduced leptin and enhanced ghrelin blood levels simultaneously with reduced adiposity due to enhanced central leptin action led us to hypothesize that normally leptin restrains the orexigenic and adiposity-promoting effects of ghrelin in two ways: it restrains ghrelin’s orexigenic effects by a central action and ghrelin secretion by a peripheral action. Therefore, the second objective of these investigations was to examine the long-term central and acute peripheral effects of leptin on ghrelin levels and adipocyte hormones in three experimental paradigms with varied ranges of ghrelin secretion: wild-type (wt) mice with normal ghrelin levels, leptin-deficient ob/ob mice displaying relatively high ghrelin (19), and hypoleptinemic ob/ob and wt mice displaying markedly elevated ghrelin secretion evoked by an intracerebroventricular (icv) injection of rAAV-lep vector (20). Because ob/ob mice lack leptin, this model allows us to identify the peripheral and central effects of leptin on ghrelin and other hormones.
It is well known that ghrelin stimulates appetite and promotes adiposity by modulating hypothalamic appetite regulating pathways, primarily the orexigenic neuropeptide Y (NPY) network in the ARC-PVN axis (4, 12, 21). Leptin has also been shown to exert a modulatory influence on energy homeostasis through the NPY network in the ARC-PVN axis (2, 3). The third objective of the current investigation was to see whether increased leptin expression in the hypothalamus would affect the orexigenic action of peripheral ghrelin.
Rodents display a circadian feeding pattern, consuming relatively larger amounts during the dark phase of the daily light-dark cycle (2, 22, 23, 24). Blood leptin levels also show a daily rhythm characterized by increased secretion concomitant with the dark-phase phagia (2, 24, 25). However, the significance of these high leptin levels on energy intake is poorly defined. The fourth objective of this study, therefore, was to determine the effects of sustained increased leptin expression in the hypothalamus on the daily feeding pattern in ob/ob and wt mice.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Effects of icv rAAV-lep on BW, FI, and circulating metabolic hormones in ob/ob mice
The aim of this experiment of was 2-fold: first, to assess the effects of selective leptin expression in the hypothalamus on efflux of hormones produced by adipocytes, pancreas, and stomach and then examine the acute impact of peripheral leptin on these hormones.
Two weeks after adaptation, ob/ob and wt mice were anesthetized with sodium pentobarbital (60 mg/kg, ip), and blood samples were collected from the orbital-sinus (80 µl, wk 0) using heparin-coated hematocrit tubes between 0900 and 1000 h. Plasma was stored at –20 C until analysis of metabolic hormones. Thereafter, each of the two groups of experimental animals was divided into two subgroups: one subgroup of each genotype was injected icv with the nonimmunogenic, nonpathogenic rAAV encoding the green fluorescent protein (rAAV-GFP) gene, and the second subgroup received rAAV-lep. Because ob/ob mice are highly sensitive to leptin feedback, compared with wt mice (26, 27, 28, 29, 30), and based on our pilot study, ob/ob mice (n = 5) received less rAAV-lep (8.6 x 107 infectious particles in 1.5 µl) than the wt controls (2.7 x 109 infectious particles in 1.0 µl, n = 6). Control ob/ob (n = 5) and wt (n = 6) mice received equivalent amount of rAAV-GFP. For icv injection, anesthetized mice were placed in a David Kopf stereotaxic apparatus with mouse adapter. The stereotaxic coordinates for third cerebroventricle injections were 0.3 mm posterior to bregma, 0.0 mm lateral to midline, and 4.2 mm below the dura.
The acute effects of exogenous leptin on hormone levels were examined between d 35 and 46 after injection when BW reduction had stabilized. Food was withdrawn at 0700 h. Three hours later, the ob/ob and wt mice received ip saline alone (control) on d 35, recombinant 1 µg mouse leptin in saline (R&D Systems Inc., Minneapolis, MN) on d 40 and 10 µg leptin on d 46 after icv injection. The selection of leptin doses was based on previous studies demonstrating the efficacy of peripheral leptin on BW and FI (27, 28, 30). Blood samples were collected from the tail vein (0 min) and at 60 min after saline or leptin injections, and plasma samples were stored at –20 C until analysis of metabolic hormones.
To assess the long-term efficacy of rAAV-lep on metabolic hormones, mice were anesthetized with sodium pentobarbital for withdrawal of a blood sample by intracardiac puncture and then killed between 0900 and 1000 h on d 52 after icv treatment. Brain was dissected out, and the hypothalamus was excised and stored frozen at –80 C for analysis of leptin mRNA by RT-PCR. Plasma was stored for analysis of metabolic hormones.
Because rAAV-lep injection decreased food intake in ob/ob, but not wt mice, an additional group of ob/ob mice was pair fed (PF) the amount of food consumed by rAAV-lep-treated ob/ob mice for 52 d, and blood samples were processed as described above.
Effects of icv rAAV-lep on the daily pattern of FI and on ghrelin-induced appetite stimulation
Groups of ob/ob (n = 6) and wt (n = 6) mice were treated icv with either rAAV-GFP (control) or rAAV-lep as in experiment 1. On d 35 after icv injection, light- and dark-phase FI was assessed. Weighed amounts of food were placed in feeders at 1800 h before lights-off, and the amount consumed was analyzed the following morning after lights-on (0700 h). At this time, freshly weighed amount of food was supplied and the amount consumed during the lights-on phase until 1800 h was recorded. To determine the effects of ghrelin on FI, 2 d later, these groups of mice were injected ip with saline on d 37, 0.4 µg/g recombinant rat ghrelin (Peptide International Inc., Louisville, KY) at 0900 h on d 43 and 1.2 µg/g ghrelin on d 49 after icv injections. Doses of ghrelin for injections were selected on the basis of previous reports (19, 31) and a pilot study in age-matched untreated mice (n = 6). Amount of food consumed after 1 h after injection was recorded.
Construction and packaging of rAAV vectors
Rat leptin cDNA derived from pCR-leptin (a gift from Dr. Roger H. Unger, Southwestern Medical School, Dallas, TX) was subcloned into rAAV vector plasmid pAAV ß-GEnh (15, 16, 26). The vectors were packaged, purified, concentrated, and titered in the Vector Core Laboratory at the University of Florida as described earlier and used in our previous studies (15, 16, 17, 18). The control virus, rAAV-GFP, was similarly constructed to encode the GFP gene.
Analysis of leptin mRNA expression in the hypothalamus
Total RNA was extracted from hypothalami using an RNA isolation kit (Qiagen, Inc., Valencia, CA). First-strand cDNA was obtained using a RNA PCR kit (reverse transcription system; Promega, Madison, WI). Primers common for mouse and rat leptin were designed to amplify a 310-bp region: sense, 5'-TGACACCAAAACCCTCATCA; antisense, 5'-ATCCAGGCTCTCTGGCTTCT. Cyclophilin, used as endogenous control, was generated as a 199-bp fragment with the primers: sense, 5'-ATGTGGTACGGAAGGTGGAG; antisense, 5'-TGGCTACCTTCGTCTGTGTG. The PCR products, generated as described earlier (15, 16, 17, 18), were sequenced and independently verified to be completely homologous with mouse and rat leptin and mouse cyclophilin.
Analyses of metabolic hormones
Plasma leptin was measured by a sensitive rat/mouse leptin RIA kit (Linco Research, St. Charles, MO), insulin by mouse insulin ELISA kits (ALPCO Diagnostics, Windham, NH), total ghrelin by ghrelin RIA kits (Phoenix Pharmaceuticals, Belmont, CA), adiponectin by mouse adiponectin RIA kits (Linco Research), TNF by CytElisa mouse kit (ALPCO Diagnostics), and FFAs by NEFA C kit (Wako Chemicals Inc., Richmond, VA). Blood glucose levels were measured with a blood glucose meter (Glucometer Elite XL; Bayer, Elkhart, IN).
Statistical analyses
BW and FI were analyzed using two-way repeated-measures ANOVA with time and treatment as variables. Comparisons of metabolic hormones within the treatment group were analyzed by paired t test and between two groups by Student’s t tests and three groups by one-way ANOVA followed by Bonferroni’s multiple comparison post hoc test. Daily FI pattern and the effect of ghrelin injection on FI were analyzed by Student’s t test. Significance was set at P < 0.05 for all analyses.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
)., inset). In contrast, leptin mRNA was detectable in the hypothalamus of control wt mice injected with rAAV-GFP, and expression increased further in response to rAAV-lep injection at 52 d after injection (Fig. 1B, inset).
|
)., A and B). However, BW suppression at d 52, compared with the respective rAAV-GFP control groups, was greater in ob/ob mice (56.2%, F2,119 = 299.1) than in wt mice (19.3%, F1, 95 = 80.7). Similarly, the effects on FI were different in the two genotypes. As reported earlier in wt mice and rats in response to either chronic leptin administration or rAAV-lep injection (28, 29, 33), the rAAV-lep dose used in this study did not influence 24-h FI in wt mice (Fig. 1B, F1, 95 = 1.85). However, presumably due to increased sensitivity to leptin as reported previously (26, 27, 28, 29, 30), even the lower-dose rAAV-lep used in ob/ob mice suppressed FI by 48% during the 52-d treatment (F1, 79 = 473.6, Fig. 1A). Pair feeding of ob/ob mice produced a different pattern in the BW response (Fig. 1A). After a significant decrease, PF mice steadily regained weight to the d 0 level. A similar response of progressive return of BW in PF rodents has been reported by others and us (2, 15, 16, 17, 33), presumably due to a compensatory decrease in energy expenditure.
Metabolic hormones (Figs. 2 and 3)
Adipocyte hormones: leptin.
As expected, in rAAV-GFP-treated ob/ob mice, plasma leptin in peripheral plasma was undetectable (data not shown). Moreover, leptin in plasma was also undetectable, despite leptin expression in the hypothalamus after rAAV-lep injection, thereby indicating little transport of centrally produced leptin to the periphery (data not shown). On the other hand, rAAV-lep in wt mice reduced plasma leptin levels by 91% from initial values and 92% from rAAV-GFP group at 52 d after injection, reflecting marked diminution of adipose tissue (Fig. 3).
|
|
Adipocyte hormones: adiponectin.
Adiponectin levels were also suppressed by rAAV-lep treatment in ob/ob mice (P < 0.05, Fig. 2). In addition, another type of adiponectin response was observed in ob/ob mice (Fig. 2). At d 0, adiponectin levels were not different between rAAV-lep and control groups, but they decreased significantly at d 52 in rAAV-GFP (53.9%), presumably due to increased fat deposition observed earlier (9, 34, 35, 36), and rAAV-lep groups (60.1%). Additionally, hypothalamic leptin transgene expression suppressed adiponectin further by 26% from that found in the rAAV-GFP control at d 52 (P < 0.05). In PF ob/ob mice, with little change in BW at d 52, adiponectin levels increased slightly (13.8%, P < 0.05, Fig. 2). In wt mice, rAAV-lep treatment significantly decreased plasma adiponectin (P < 0.05, Fig. 3) from the initial d 0 value; at d 52 levels were 27% lower than in rAAV-GFP group (P < 0.05).
Adipocyte hormones: TNF.
rAAV-lep also decreased TNF levels in ob/ob mice in contrast to little impact in control rAAV-GFP or PF groups of mice (Fig. 2). In wt mice, rAAV-lep failed to affect circulating TNF levels, despite a loss of weight and adipose tissue (Fig. 3).
Blood insulin and glucose levels (Figs. 2 and 3).
Leptin expression in the hypothalamus of ob/ob mice attenuated hyperinsulinemia by 90%. Because plasma leptin levels were undetectable in these mice, it implied an action of leptin confined to central targets in inhibiting insulin secretion from pancreatic ß-cells (Fig. 2). Similar reduction in plasma insulin levels (80%) was also observed in response to rAAV-lep treatment in wt mice (Fig. 3). Despite the suppressed insulin levels, blood glucose levels were significantly suppressed (P < 0.05 vs. respective rAAV-GFP groups) to normoglycemic range in ob/ob mice and wt mice (Figs. 2 and 3). Pair feeding of ob/ob mice also reduced circulating insulin and glucose levels (P < 0.05), but the reduction was markedly smaller, compared with that after rAAV-lep treatment (Fig. 2, P < 0.05).
Gastric hormone ghrelin (Figs. 2 and 3).
In contrast to the suppressive effects on adipocyte and pancreatic hormones, rAAV-lep treatment increased ghrelin levels in both ob/ob (Fig. 2) and wt mice (Fig. 3). The increase was much greater in ob/ob mice than in wt mice because ghrelin levels rose by more than 30-fold in ob/ob mice (Fig. 2), compared with the almost 3-fold increase in wt mice (Fig. 3).
Effects of rAAV-lep on the daily FI pattern (Fig. 4)
Two types of effects of rAAV-lep on daily energy consumption in ob/ob and wt mice were obtained. First, increased hypothalamic leptin expression did not affect the daily rhythm in ingestive behavior; the dark-phase phagia was observed in both genotypes. Second, the cumulative 24-h FI was suppressed in the rAAV-lep-treated ob/ob mice due to a 33% decrease in eating during the dark-phase alone (P < 0.05, Fig. 4A). On the other hand, whereas significant dark-phase reduction (19%) in FI was also elicited in wt mice by rAAV-lep (P < 0.05, Fig. 4B), these mice ate significantly more during the light phase, thereby raising the cumulative 24-h consumption to the range of rAAV-GFP control mice (Fig. 4B).
|
)), both doses of leptin were highly and equally effective in reducing the elevated ghrelin levels in rAAV-lep-injected ob/ob and wt mice. Ghrelin was decreased by 75% in ob/ob and 50% in wt mice at 60 min (P < 0.05, Fig. 5, A and B). Leptin injection also concurrently suppressed plasma adiponectin levels without affecting TNF levels in rAAV-lep-treated ob/ob and wt mice. Whereas ip leptin was ineffective in rAAV-GFP ob/ob controls, the two doses of leptin were equally effective in decreasing adiponectin in wt rAAV-GFP-treated mice (P < 0.05). In rAAV-lep-treated ob/ob mice, both doses of leptin suppressed adiponectin, but only the high leptin dose decreased adiponectin levels in rAAV-lep wt mice (P < 0.05, Fig. 5B).
|
)
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
A selective increase in leptin transgene expression in the hypothalamus, in the complete absence of leptin in the periphery, suppressed weight gain for an extended period. This further strengthens the emerging notion (39) that the energy homeostatic effector pathways in the hypothalamus are exclusive targets for this effect of leptin. Whereas leptin mRNA was detectable in the hypothalamus of wt mice as in the rat hypothalamus (15, 17), it was undetectable in the hypothalamus of ob/ob mice. Leptin expression locally in the hypothalamus after rAAV-lep treatment reinstated weight homeostasis in ob/ob and wt mice. Additionally, evidence that transgene expression after icv injection is largely confined to hypothalamic sites (16, 40) and a similar efficacy on weight suppression manifests after increased leptin transgene expression in response to microinjection of rAAV-lep in discrete hypothalamic sites (17) further endorses the capability of locally produced leptin to control weight homeostasis for extended periods (15, 16, 41, 42). In an ongoing study (Kalra, S., N. Ueno, S. Boghossian, and P. Kalra, unpublished observations), similar long-term control on weight gain is seen for over 15 months.
A greater than 80% reduction in plasma leptin and decreased FFA levels seen in wt mice and a similar decrease in FFA in ob/ob mice implied that weight reduction was due to fat depletion and not loss of muscle mass, as demonstrated previously by direct measurement of fat and muscle mass after rAAV-lep or leptin therapy in rats and mice (15, 16, 27, 33). Furthermore, whereas the low dose of rAAV-lep employed in this study failed to change the 24-h food intake in wt mice, in accord with similar findings seen after either short- or long-term administration of leptin in the periphery or after central infusion in mice (28, 29, 33) and central rAAV-lep injection in rats (15, 16), even a lower rAAV-lep dose was highly effective in ob/ob mice. Because increased central leptin transgene expression consistently increased nonshivering thermogenic energy expenditure as indicated by up-regulation of uncoupling protein-1 mRNA expression in brown adipose tissue of wt rats (15, 16, 17, 18) and ob/ob mice (our unpublished observations), the BW reduction observed in wt mice is likely an outcome of increased energy expenditure. Likewise, the relatively greater loss of BW in ob/ob mice is evidently a result of enhanced energy expenditure in combination with reduced energy intake caused by the leptin hypersensitivity in these mice (26, 27, 28).
There are several additional new findings of these investigations. First, the sustained increase in leptin expression in the hypothalamus failed to affect the occurrence of the circadian clock-driven dark-phase phagia consisting of increased feeding episodes of high amplitude and frequency (2, 22, 23, 24). However, despite ad libitum availability of food, dark-phase intake was reduced in both ob/ob and wt mice. Consequently, we propose that increased leptin expression selectively in the hypothalamus reduces only the dark-phase consumption. Presumably, an attenuated neurogenic appetitive urge underlies the voluntary reduction in FI. Enhanced NPY release in the hypothalamic PVN alone has been attributed to initiation of the dark-phase phagia (2, 43, 44, 45). The observations that NPY rapidly stimulates feeding, activation of PVN NPY receptors with central NPY infusion dose-dependently stimulated episodic pattern of feeding (43, 44), NPY immunoneutralization suppressed dark-phase intake (46), and increased leptin expression with rAAV-lep injection-attenuated NPY mRNA expression in the ARC are in accord with reduced NPYergic signaling in the ARC-PVN axis as one of the underlying factors in decreasing intake (16, 17, 44, 45, 46, 47). Consequently, we infer that attenuated NPY release in the PVN of both wt and ob/ob mice during the dark phase alone is responsible for the selective reduction in FI during this period.
It is possible that anorexigenic hypothalamic effector pathways (1, 2, 3, 4, 5, 6), also shown to be modulated by rAAV-lep treatment (15, 16, 17, 26), may additionally participate in diminishing dark-phase FI. Nevertheless, whereas central leptin transgene expression failed to affect eating bouts during the lights-on phase in ob/ob mice, intriguingly, increased daytime feeding in wt mice normalized cumulative daily intake. This temporal segregation of effects on ingestive behavior endorses the possibility that central inhibitory feedback mechanisms engaged by leptin on a daily basis operate only during the dark phase concomitant with nocturnal leptin hypersecretion (24, 25) and are more sensitive to leptin in ob/ob relative to wt mice (27, 28, 29). Furthermore, the escape from leptin restraint on feeding during the lights-on phase in wt mice provides for the first time an explanation for the puzzling reports from several investigators, demonstrating the ineffectiveness of rAAV-lep as well as central or peripheral leptin administration on a long-term basis to reduce the daily energy consumption in wt rodents (15, 16, 28, 29, 33). As discussed below, this escape is unlikely to be due to increased ghrelin and thus warrants further investigation to determine the causal neural and hormonal factors.
The second new finding of this study relates to the role of leptin on ghrelin. Ghrelin stimulates feeding, and either daily injections or infusion over extended periods promotes adiposity (4, 12, 48). We observed that increased circulating ghrelin, 30-fold in ob/ob and 2-fold in wt mice, failed to elicit overeating. In an ongoing study, we observed a similar ghrelin hypersecretion concomitant with suppressed FI 30 wk after rAAV-lep injection (Kalra, S., N. Ueno, S. Boghossian, and P. Kalra, unpublished observations). That this ineffectiveness of ghrelin is due to a potent restraint exerted by locally produced leptin is indicated by our findings that, whereas exogenous ghrelin readily stimulated feeding in a dose-dependent manner in control mice, it was completely ineffective in mice expressing transgene in the hypothalamus. Thus, we suggest that one of the mechanisms engaged by leptin in integration of hypothalamic control on weight homeostasis is to counteract the central orexigenic effects of ghrelin. Several lines of evidence show that the central counterbalancing interplay between anorexigenic leptin and orexigenic ghrelin is mediated by hypothalamic NPYergic signaling (1, 2, 3, 4, 12). On the basis of the findings of the current study, it is tempting to postulate that leptin acts in two ways to regulate weight homeostasis: one, by directly modulating various components of the central energy and appetite regulating networks, especially at the level of NPY signaling (2, 3, 4, 12, 21, 48), and two, by restraining ghrelin secretion from the stomach as indicated by the rapid suppression of ghrelin levels after leptin injection in ob/ob and wt mice.
The experimentally induced reciprocal fluctuations in plasma ghrelin and leptin are akin to those observed in various other physiological paradigms. A decrease in ultradian leptin secretion was contemporaneous with enhanced ultradian ghrelin secretion in fasted rats (49). Recently we observed a diminution of ghrelin episodic discharge coincident with hyperleptinemia characterized by acceleration in ultradian fluctuations in leptin response to consumption of high-fat diet only in obesity-prone rats (50). The preprandial ghrelin increase is seen when serum leptin levels are at nadir, and with food consumption a rise in leptin secretion is followed by receding ghrelin levels (3, 51). The postprandial decreases in ghrelin levels in humans may be concomitant with rises in circulating leptin (12, 51). Conversely, hyperleptinemia in obese rats and humans was negatively correlated with ghrelin levels (12, 52). Thus, our demonstration of rapid suppression of ghrelin by leptin in wt mice displaying low endogenous leptin and marked suppression of ghrelin in ob/ob mice deficient in peripheral leptin lend credence to the proposal that normally leptin exerts a tonic restraint on gastric ghrelin efflux directly, and conversely, ghrelin secretion is enhanced when this restraint is diminished due to reduced leptin titer. A possible central action of peripherally injected leptin to inhibit ghrelin secretion in our paradigm is highly unlikely because marked suppression of ghrelin was observed also in mice overexpressing leptin centrally. Consequently, we envision leptin receptors in the gastrointestinal tract to mediate the leptin restraint on ghrelin secretion (53, 54). Gastric cells can now be added to the growing list of central and peripheral targets for the pleiotropic effects of leptin (10, 11). These new findings warrant future investigations to elucidate the precise mode of action at cellular and molecular levels of leptin, derived either from adipocyte or locally involving paracrine/autocrine action and of sympathetic nervous system in exerting a restraint on ghrelin secretion.
The third new finding to emerge from this study relates to a possible feedback action of leptin on the other adipocyte hormones, adiponectin and TNF, that were also suppressed in response to leptin transgene expression in the hypothalamus. Whereas adiponectin was suppressed in both wt and ob/ob mice, TNF was reduced only in ob/ob mice hypersecreting TNF. The mechanism responsible for this differential TNF response to central leptin gene therapy in the two genotypes remains to be determined. Nevertheless, it is conceivable that simultaneous decreases in leptin, adiponectin, and TNF secretion could be a consequence of enhanced central leptin action either independently or in conjunction with the drastic depletion of fat depots, the source of these hormones (55, 56). The possibility that immune cells may secrete reduced TNF is not ruled out by our study (55, 56).
Furthermore, our observation that leptin injection suppressed plasma adiponectin levels not only provides an explanation for the coincidence of hyperleptinemia and hypoadiponectemia in obesity (34, 35, 36, 57, 58) but also implies a paracrine/autocrine leptin action mediated by the biologically relevant long form of leptin receptor in adipose tissues (11, 59). Inhibition of adiponectin secretion is quite selective because TNF efflux was not influenced by exogenous leptin. Together with the knowledge of reciprocity of circulating concentrations of leptin and ghrelin (3, 49, 50, 51, 52) and the fact that leptin can readily suppress ghrelin secretion, it is tempting to propose that leptin may be an important regulatory signal at peripheral targets that secrete afferent hormonal signals important in hypothalamic sustenance of energy homeostasis.
The functional significance of the inhibitory effects of central and peripheral leptin on adiponectin remains to be ascertained. A decrease in adiponectin levels in parallel with hyperleptinemia and adiposity has been implicated in the development of insulin resistance, hyperglycemia, and type 2 diabetes (34, 35, 36, 37, 38, 60). However, we observed concurrency of central leptin gene therapy-produced hypoleptinemia and suppressed hyperinsulinemia and hyperglycemia with increased insulin sensitivity in wt and ob/ob mice (61). Therefore, the observation of diminished adiponectin levels accompanying peripheral hypoleptinemia in the current study advocates a reassessment of the precise role of adiponectin in the etiology of insulin resistance in type 2 diabetes. Whether it participates independently of hyperleptinemia, which is thought to be a driving force in the progression of insulin resistance and hyperinsulinemia toward the development of diabetes type 2 (8, 11, 36, 37, 38, 59, 60), warrants further investigation.
The fourth new finding of these studies relates to central leptin action on insulin-glucose homeostasis. Marked suppression of plasma insulin in leptin-deficient ob/ob mice expressing leptin solely in the hypothalamus, without any evidence of transport to peripheral circulation, reinforces the inference that it is the central and not the peripheral action of leptin on pancreatic ß-cells, skeletal muscles, and liver (5, 10, 38, 59, 62, 63) that ameliorates hyperinsulinemia and hyperglycemia (15, 16, 17, 18, 42). The alternative possibility that loss of fat tissue alone is responsible for suppression of hyperinsulinemia and hyperglycemia (59, 63) is not supported by the current observations that whereas a rise in leptin after exogenous injection suppressed ghrelin, it failed to decrease plasma insulin (data not shown in results). Previous studies showing that intact ventromedial hypothalamus is required for suppression of hyperinsulinemia and hyperglycemia by leptin are also in line with the existence of a central site of leptin action in insulin and glucose homeostasis (64).
In summary, by application of central leptin gene therapy paradigm to selectively affect neural mechanisms and circumvent the multiple confounding effects of peripheral leptin in leptin-deficient ob/ob mice, we have shown the following: 1) sustained leptin expression in the hypothalamus does not interfere with the daily feeding pattern but elicits a reduction in dark-phase FI alone; 2) leptin exerts a restraint on the orexigenic effects of ghrelin in two ways, centrally by countering its appetite promoting effects at the level of effector pathways in the hypothalamus and peripherally by attenuating gastric ghrelin secretion; 3) leptin may also exert a tonic restraint on adipocyte adiponectin secretion, a hormone implicated in insulin resistance; and 4) the results reinforce the view that one of the sites of leptin action in modulating pancreatic insulin and glucose homeostasis may reside in the hypothalamus. Cumulatively, the outcome of these studies advances our understanding of the disparate roles of leptin in the interplay of multiple peripheral hormonal signals involved in the development of obesity and attendant metabolic disorders, such as hyperinsulinemia and insulin insensitivity.
|
Footnotes |
|---|
Presented at the 33rd Annual Meeting of the Society for Neuroscience, New Orleans, Louisiana, 2003 (Abstract 283.8, p 81).
Abbreviations: ARC, Arcuate nucleus; BW, body weight; FFA, free fatty acid; FI, food intake; icv, intracerebroventricular; NPY, neuropeptide Y; PF, pair fed; PVN, paraventricular nucleus; rAAV-GFP, rAAV encoding the green fluorescent protein; rAAV-lep, recombinant adeno-associated virus encoding the leptin gene; wt, wild-type.
Received March 1, 2004.
Accepted for publication May 11, 2004.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-cells of humans and rats and stimulates insulin secretion. Diabetes 51:124–129This article has been cited by other articles:
 |
|
 |
|
  R. M. Luque, Z. H. Huang, B. Shah, T. Mazzone, and R. D. Kineman Effects of leptin replacement on hypothalamic-pituitary growth hormone axis function and circulating ghrelin levels in ob/ob mice Am J Physiol Endocrinol Metab, March 1, 2007; 292(3): E891 - E899. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Perez-Tilve, L. Gonzalez-Matias, M. Alvarez-Crespo, R. Leiras, S. Tovar, C. Dieguez, and F. Mallo Exendin-4 Potently Decreases Ghrelin Levels in Fasting Rats Diabetes, January 1, 2007; 56(1): 143 - 151. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. C Gormsen, J. Gjedsted, S. Gjedde, E. T. Vestergaard, J. S Christiansen, J. O. Jorgensen, S. Nielsen, and N. Moller Free fatty acids decrease circulating ghrelin concentrations in humans. Eur. J. Endocrinol., May 1, 2006; 154(5): 667 - 673. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. Rico, M. Del Rio, A. Bravo, A. Ramirez, J. L. Jorcano, M. A. Page, and F. Larcher Targeted Overexpression of Leptin to Keratinocytes in Transgenic Mice Results in Lack of Skin Phenotype but Induction of Early Leptin Resistance Endocrinology, October 1, 2005; 146(10): 4167 - 4176. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Erdmann, F. Lippl, S. Wagenpfeil, and V. Schusdziarra Differential Association of Basal and Postprandial Plasma Ghrelin With Leptin, Insulin, and Type 2 Diabetes Diabetes, May 1, 2005; 54(5): 1371 - 1378. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. P. Kalra, N. Ueno, and P. S. Kalra Stimulation of Appetite by Ghrelin Is Regulated by Leptin Restraint: Peripheral and Central Sites of Action J. Nutr., May 1, 2005; 135(5): 1331 - 1335. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. Lord, S. Ledoux, B. D. Murphy, D. Beaudry, and M. F. Palin Expression of adiponectin and its receptors in swine J Anim Sci, March 1, 2005; 83(3): 565 - 578. [Abstract] [Full Text] [PDF]
|
|
| |||||||
