cns hormonal and nutritional regulation of glucose … · 2017. 12. 19. · title: cns hormonal and...
TRANSCRIPT
CNS HORMONAL AND NUTRITIONAL REGULATION OF
GLUCOSE AND ENERGY HOMEOSTASIS
by
Mona Anna Abraham
A dissertation submitted in conformity with the requirements for the degree of Doctor of Philosophy
Department of Physiology University of Toronto
© Copyright by Mona Anna Abraham (2017)
ii
Title: CNS Hormonal and Nutritional regulation of Glucose and Energy homeostasis Name: Mona Anna Abraham
Degree: Doctor of Philosophy
Department: Physiology, University of Toronto
Year of Convocation: 2017
GENERAL ABSTRACT
The mediobasal hypothalamus (MBH) and the dorsal vagal complex (DVC) regions contain a
leaky blood-brain barrier and therefore act as critical sites of action for circulating hormones and
nutrients on glucose and energy homeostasis. For example, hormones such as insulin, leptin and
glucagon and nutrients such as lipids, glucose and amino acids act in the MBH and DVC to
regulate hepatic glucose production, glucose tolerance, food intake and body weight, but the
underlying CNS sensory mechanisms in rodents and humans remain elusive. The current
dissertation unveils novel mechanisms for glucagon action in the MBH and glycine sensing in
the DVC in rats and mice that maintain glucose and energy homeostasis. In Study 1, we
discovered that KATP channel is necessary for MBH glucagon action to exert glucose control. In
Study 2, we found that pharmacological and molecular manipulation of glycine transporter 1 in
the DVC enhances glycine sensing and potently regulates glucose and energy homeostasis. In
conclusion, I have discovered novel mechanisms for glucagon and glycine sensing in the brain
that regulate glucose and energy homeostasis, and have unveiled CNS therapeutic targets for
diabetes and obesity.
iii
Acknowledgments
Completing my doctoral work at the University of Toronto has been one of the most rewarding
and challenging experiences of my life. I’m extremely thankful to all the intellectual mentors and
colleagues who have guided and supported me through the last 5 years.
To my PhD supervisor, Dr. Tony Lam. I have greatly benefitted from his work ethic, keen
scientific insight, and ability to simplify complex questions. Sincerely, I appreciate everything he
has done for me.
To my committee: Drs. Adria Giacca, Richard Bazinet, Daniel Drucker. They have been
extremely generous with their time and always provided helpful suggestions, valuable advice,
and support.
To my current and past colleagues in the lab. Every result in this dissertation was accomplished
with the assistance of fellow lab-mates. Dr. Jessica Yue, Dr. Beatrice Filippi, Mary, Dr. Frank
Duca, Paige and I worked together on several different projects, and without their efforts my
doctoral work would have been far more difficult.
I gratefully acknowledge the support of the Banting and Best Diabetes Centre Fellowship and
Canadian Institutes for Health Research for their financial support to make this work possible.
I’m extremely thankful for my grandparents and parents who instilled in me a love of learning,
and lastly, I’m grateful most of all to God, who opens doors of destiny.
iv
Table of Contents
Acknowledgments .......................................................................................................................... iii
Table of Contents ........................................................................................................................... iv
List of Tables ................................................................................................................................ vii
List of Figures .............................................................................................................................. viii
List of Publications that contributed to this Thesis ..........................................................................x
Chapter 1 Introduction .....................................................................................................................1
1.1 Diabetes and Obesity ............................................................................................................1
1.2 The role of the hypothalamus in the regulation of glucose and energy homeostasis ............3
1.1.1 Hormonal action .......................................................................................................4
1.1.2 Nutrient sensing .....................................................................................................13
1.2 The role of the dorsal vagal complex in the regulation of glucose and energy homeostasis ........................................................................................................................18
1.2.1 Hormonal action .....................................................................................................19
1.2.2 Nutrient sensing .....................................................................................................21
1.3 The role of glial cells in the regulation of glucose and energy homeostasis. ......................25
1. 4 Summary of Introduction/Rationale of Study 1 and Study 2 .............................................26
Chapter 2 ........................................................................................................................................28
Study 1 ...........................................................................................................................................28
2.1 Abstract ...............................................................................................................................29
2.2 Introduction .........................................................................................................................30
2.3 Materials and Methods ........................................................................................................32
2.3.1 Animal Preparation ...................................................................................................32
2.3.2 Adenovirus injection .................................................................................................32
2.3.3 Pancreatic-euglycemic clamp ...................................................................................32
2.3.4 Western blotting for phosphorylated ACC ...............................................................33
v
2.3.5 PKC-δ activity assay .................................................................................................34
2.3.6 Biochemical Analysis ...............................................................................................35
2.3.7 Statistical Analysis ....................................................................................................35
2.4 Results .................................................................................................................................35
2.4.1 Role of MBH AMPK in glucagon action .................................................................35
2.4.2 Role of MBH PKC-δ in glucagon action ..................................................................37
2.4.3 Role of MBH KATP channels in glucagon action ......................................................37
2.5 Discussion ...........................................................................................................................38
Chapter 3 ........................................................................................................................................45
Study 2 ...........................................................................................................................................45
3.1 Abstract ...............................................................................................................................46
3.2 Introduction .........................................................................................................................47
3.3 Materials and Methods ........................................................................................................49
3.3.1 Animal preparation and surgical procedures ............................................................49
3.3.2 Intravenous glucose tolerance test ............................................................................50
3.3.3 DVC treatments ........................................................................................................51
3.3.4 Pancreatic basal insulin euglycemic clamp in rats ....................................................52
3.3.5 Hepatic branch vagotomy in rats ..............................................................................53
3.3.6 Microdialysis .............................................................................................................54
3.3.7 DVC virus injection ..................................................................................................54
3.3.8 Brain tissue sampling in rats .....................................................................................55
3.3.9 Microdialysate sample glycine analysis ....................................................................55
3.3.10 Acute (3-d) and chronic (28-d) high-fat feeding in rats ..........................................56
3.3.11 Intravenous ALX infusion clamps ..........................................................................56
3.3.12 Induction of experimental type 2 diabetes ..............................................................57
3.3.13 Fasting-refeeding experiments ................................................................................57
vi
3.3.14 Pancreatic basal insulin euglycemic clamp in mice ................................................57
3.3.15 Western blot analyses .............................................................................................58
3.3.16 Biochemical analysis ..............................................................................................59
3.3.17 Calculations and statistics .......................................................................................59
3.4 Results .................................................................................................................................60
3.4.1 Gluco-regulation by DVC GlyT1 inhibition in healthy rodents ...............................60
3.4.2 Anti-diabetic effect of DVC GlyT1 inhibition ..........................................................63
3.4.3 Metabolic benefits of DVC GlyT1 inhibition in obesity ..........................................65
3.4.4 DVC GlyT1 inhibition regulates energy balance ......................................................66
3.5 Discussion ...........................................................................................................................68
Chapter 4 Summary, Discussion and Future Directions ................................................................86
4.1 Summary .............................................................................................................................86
4.2 Discussion ...........................................................................................................................89
4.3 Limitations and Future directions .......................................................................................98
References ....................................................................................................................................103
vii
List of Tables
Table 2. 1 Plasma insulin, glucagon and glucose concentrations during basal and clamp
conditions. Data are means ±SEM. ............................................................................................... 44
viii
List of Figures
Figure 2. 1 Schematic representation of the working hypothesis. ................................................ 40
Figure 2. 2 Role of MBH AMPK in glucagon action. .................................................................. 41
Figure 2. 3 Role of MBH PKC- δ in glucagon action. .................................................................. 42
Figure 2. 4 Role of MBH KATP channels in glucagon action. ....................................................... 43
Figure 3. 1 Schematic representation of the working hypothesis. ................................................ 72
Figure 3. 2 Chemical inhibition of DVC GlyT1 regulates glucose homeostasis in healthy rats. . 73
Figure 3. 3 Molecular inhibition of DVC GlyT1 regulates glucose homeostasis in healthy rats. 74
Figure 3. 4 DVC and iv infusion of ALX regulates glucose homeostasis in 3d-HFD rats. .......... 75
Figure 3. 5 Inhibition of DVC GlyT1 regulates glucose homeostasis in diabetic rats. ................ 76
Figure 3. 6 Chemical inhibition of DVC GlyT1 regulates glucose homeostasis in obese rats. .... 77
Figure 3. 7 Molecular inhibition of DVC GlyT1 regulates metabolic homeostasis in obese rats. 78
Figure 3. 8 Chemical and molecular inhibition of DVC GlyT1 regulate energy balance. ........... 79
Supplementary Figure 3. 1 Metabolic effects of chemical inhibition of DVC GlyT1 in healthy
rats. ................................................................................................................................................ 80
Supplementary Figure 3. 2 Metabolic effects of chemical inhibition of GlyT1 in the 4th ventricle
of mice and in the DVC of hepatic vagotomized rats. .................................................................. 81
Supplementary Figure 3. 3 Brain regions included in the GlyT1 protein analysis. ...................... 82
ix
Supplementary Figure 3. 4 Metabolic effects of molecular inhibition of DVC GlyT1 in healthy
rats. ................................................................................................................................................ 83
Supplementary Figure 3. 5 Metabolic effects of DVC and iv infusion of ALX in 3d-HFD rats. 84
Supplementary Figure 3. 6 Metabolic effects of chemical and molecular inhibition of DVC
GlyT1 in obese rats. ...................................................................................................................... 85
Figure 4. 1 Summary of Study 1 and Study 2. .............................................................................. 88
x
List of Publications that contributed to this Thesis
Study 1:
Abraham, M.A., Yue, J.T., LaPierre, M.P., Rutter, G.A., Light, P.E., Filippi, B.M., and Lam, T.K. (2014). Hypothalamic glucagon signals through the KATP channels to regulate glucose production. Mol Metab 3, 202-208.
Study 2:
Abraham, M.A.*, Yue, J.T.*, Bauer, P.V., LaPierre, M.P., Wang, P., Duca, F.A., Filippi, B.M., Chan, O., and Lam, T.K. (2016). Inhibition of glycine transporter-1 in the dorsal vagal complex improves metabolic homeostasis in diabetes and obesity. Nat Commun 7, 13501. *Equal contribution
Review Papers:
Abraham, M.A., and Lam, T.K. (2016). Glucagon action in the brain. Diabetologia 59, 1367-1371.
Abraham, M.A., Filippi, B.M., Kang, G.M., Kim, M.S., and Lam, T.K. (2014). Insulin action in the hypothalamus and dorsal vagal complex. Exp. Physiol. 99, 1104-1109.
LaPierre, M.P.*, Abraham, M.A.*, Filippi, B.M., Yue, J.T., and Lam, T.K. (2014). Glucagon and lipid signaling in the hypothalamus. Mamm. Genome 25, 434-441. *Equal contribution
Filippi, B.M.*, Abraham, M.A.*, Yue, J.T., and Lam, T.K. (2013). Insulin and glucagon signaling in the central nervous system. Rev. Endocr. Metab. Disord. 14, 365-375. *Equal contribution
Other studies contributing to the completion of this dissertation:
LaPierre, M.P., Abraham, M.A., Yue, J.T., Filippi, B.M., and Lam, T.K. (2015). Glucagon signalling in the dorsal vagal complex is sufficient and necessary for high-protein feeding to regulate glucose homeostasis in vivo. EMBO Rep 16, 1299-1307.
Filippi, B.M., Bassiri, A., Abraham, M.A., Duca, F.A., Yue, J.T., and Lam, T.K. (2014). Insulin signals through the dorsal vagal complex to regulate energy balance. Diabetes 63, 892-899.
1
Chapter 1 Introduction
1.1 Diabetes and Obesity In 1878, the French physiologist Claude Bernard described that ‘All the vital mechanisms,
however varied they may be, have only one object, that of preserving constant the conditions of
the internal environment which make a free and independent life possible’.1,2 This maintenance
of an internal state defended against changes is particularly true for the regulation of blood
glucose levels and body weight. Low levels of plasma glucose would deplete the brain of its only
energy source, leading to seizures, unconsciousness, and death. On the other hand, sustained
elevation of blood glucose can also be fatal as it causes diabetes and associated complications.
Therefore, it is vital for the body to maintain blood glucose at a fairly narrow range of 4-7
mmol/L. Glucose homeostasis is that process of maintaining blood glucose at a steady-state
level. Similarly, in a healthy body, body weight and body fat is also defended against acute
perturbations under the influence of a tightly regulated homeostatic process called energy
homeostasis. For instance, rats when subjected to caloric restriction display significant weight
loss but when returned to free access of food, quickly rebound regaining their initial body weight
within days3,4. Remarkably, this precise nature of body weight regulation is also true in humans,
both lean and obese. In fact, the literature documents the weight gain rate among obese men
(0.04 kg·BW/year) is slower than men with no history of obesity (0.18 kg·BW /year)5.
However, despite these robust homeostatic systems, Type 2 diabetes and obesity are the
two most challenging public health concerns of the 21st century. Where the global prevalence of
diabetes was 9.3% in 2015, this number is predicted to increase to 12.1 % by 20256. Coincident
with this diabetes epidemic, the prevalence rates of obesity has also been escalating, with about
2
13% of the world’s adult population obese in 20147. People affected by diabetes and obesity are
predisposed to developing cardiovascular diseases, and cancer8, which in turn reduce their life
expectancy and cost heavily in health care expenses. In Canada alone, the economic burden of
diabetes, which was approximated at $12.2 billion in 2010, is projected to increase by another
$4.7 billion by 20209. Meanwhile, the annual economic burden of obesity in Canada ranges from
$ 4.6 billion to $7.1 billion10. Given the survival and financial crisis caused by these diseases and
the expected rise in the number of affected individuals, the need for therapeutic interventions
aimed at combating the diabetes and obesity epidemic is more than ever today.
In this regard, considerable advances have been made scientifically in understanding the
different mechanisms that provide effective feedback to regulate glucose and energy
homeostasis. It is now recognized that the central nervous system (CNS) plays a critical role in
coordinating and integrating various components of glucose and energy regulation. In particular,
hormonal and nutrient signals from the periphery, relaying the body’s energy status, are detected
by the brain and integrated in CNS pathways to maintain constant blood glucose levels and body
weight stability. It follows that Type 2 diabetes and obesity develops as a result of dysregulation
in the ability of the CNS hormone and sensing pathways to appropriately couple the body’s
energy needs with nutrient intake and endogenous nutrient output. As such, delineating the
mechanisms of CNS hormonal signaling and/or nutrient sensing is vital in understanding and
identifying potential molecular targets to therapeutically restore regulation of feeding behaviour
and glucose homeostasis. Till date, tremendous progress has been made to elucidate the
molecular and cellular pathways, primarily within the hypothalamus and hindbrain, comprising
hormonal action and nutrient sensing circuits (which will be reviewed as follows).
The goal of this dissertation is to characterize novel CNS mechanisms of hormonal
signaling and nutrient sensing involved in the control of glucose and energy balance, thereby
3
unveiling potential new therapeutic targets and experimental approaches to improve metabolic
control in diabetes and obesity.
1.2 The role of the hypothalamus in the regulation of glucose and energy homeostasis Of the different anatomical regions in the brain, the hypothalamus in particular is a privileged
CNS site to sense and integrate peripheral signals to regulate metabolic homeostasis. The median
eminence located at the mediobasal hypothalamus (MBH) and adjacent to the arcuate nucleus
(ARC) is a circumventricular organ, which is lined by fenestrated brain endothelium. This
permits circulating molecules to traverse past the blood brain barrier (BBB) and access the
hypothalamic ARC. There are two sets of first order neurons in the ARC, on which peripheral
metabolic hormones such as insulin, leptin, and glucagon and nutrients such as fatty acids and
glucose act: 1) the neurons that produce the anorexigenic (appetite suppressing) neuropeptides,
pro-opiomelanocortin (POMC) and cocaine- and amphetamine-regulated transcript (CART), and
2) the neurons that produce the orexigenic (appetite promoting) neuropeptides, agouti-related
peptide (AgRP) and neuropeptide Y (NPY)11. Projections from these first order neurons to
second order neurons in other hypothalamic areas such as the paraventricular hypothalamus,
lateral hypothalamus, and ventromedial hypothalamus or extrahypothalamic areas including the
nucleus tractus solitarius ultimately influence peripheral metabolism.
4
1.1.1 Hormonal action
Insulin
Insulin is the principal metabolic hormone that controls blood glucose by promoting glucose
uptake into muscle and adipose tissues, and suppressing glucose production by the liver. It is
now well-established that the brain is an important insulin-sensitive organ that has a
physiological role in the regulation of glucose homeostasis as well as energy balance. Pioneering
experiments by Porte and group revealed that intracerebroventricular (icv) injection of insulin
stimulated insulin secretion from the pancreas of dogs12 and reduced body weight and intake of
food in baboons13. Similarly, intranasal administration of insulin also led to decreased plasma
glucose in circulation in dogs14 and rhesus monkeys15. Studies in the past decade have illustrated
that the hypothalamus specifically is an important CNS site of insulin action for metabolic
regulation. Obici et al. demonstrated that insulin receptors (IR) within the hypothalamus have a
physiological role in the regulation of food intake, fat mass and hepatic insulin action. They
showed that rats treated with injection of IR antagonists into the third cerebral ventricle (icv-3) to
block hypothalamic insulin signalling, failed to suppress hepatic glucose production during
hyperinsulinemic clamp studies as well as displayed hyperphagia and increased fat mass16.
Subsequent studies revealed that hypothalamic insulin signals via the insulin receptor substrate-
phosphatidylinositol 3-kinase (IRS-PI3K) pathway17. Hypothalamic overexpression of either
IRS-2 or protein kinase B (PKB, a main downstream signalling molecule of PI3K action
enhanced the ability of peripheral insulin treatment to lower glucose by approximately 2-fold in
rats with uncontrolled diabetes induced by streptozotocin (STZ) 17. Moreover, activation of ATP-
sensitive potassium (KATP channels) is also required for hypothalamic insulin signaling. Insulin
stimulates KATP channel activity in hypothalamic glucose-sensing neurons in lean rats, but not in
obese rats nor in the presence of the KATP channel inhibitor tolbutamide2. Likewise, in the
5
presence of icv-3 infusion of the KATP channel inhibitor glibenclamide, central insulin failed to
lower glucose production and regulate peripheral glucose fluxes18. Indeed, the hepatic vagus
nerve transmit the brain-liver communication of central insulin action and, at the level of the
liver, the hepatic interleukin-6 (IL-6)/STAT3 signaling mediates the effect of brain insulin to
lower hepatic glucose production19. A recent study has proposed inhibition of the α7-nicotinic
acetylcholine receptor (nAchR) contributes to the effect of central insulin to activate hepatic IL-
6/STAT3 and regulate gluconeogenic responses20. The exact mechanism underlying this brain-
liver axis of insulin action remains unclear and merits investigation.
Regarding the neuronal circuitry, AgRP neurons play a major role in insulin’s ability to
lower glucose production21,22. In specifically targeted AgRP-neuron insulin receptor (IR)
knockout mice, insulin failed to suppress glucose production during hyperinsulinemic clamps21.
On the other hand, in L1 mice (mice that have ~90 % reduction of IR levels in the ARC and
which have hyperinsulinemia and impaired regulation of glucose production23), the restoration of
insulin signaling in AgRP neurons was sufficient for central insulin to lower glucose
production22.
Importantly, overnutrition appears to affect the ability of central insulin to lower food
intake and glucose production. Obese rats develop insulin resistance in the brain which prevents
an insulin-dependent decrease of food intake24. High-fat diet (HFD) feeding induces the
expression of several pro-inflammatory cytokines and inflammatory responsive proteins in the
hypothalamus; this together with the accumulation of pro-inflammatory lipids increases local
inflammation in the hypothalamus and impairs the anorectic effect of insulin25,26 inducing
hyperphagia and consequently, body weight gain. Recent evidence suggested that switching to a
low-fat diet reversed hypothalamic insulin resistance caused by diet-induced obesity and restored
the anorectic effect of hypothalamic insulin27. Similarly, HFD feeding for 3 days (3d HFD) also
6
blunts the glucose production-lowering effect of hypothalamic insulin28. This is associated with
the activation of hypothalamic S6 kinase (S6K), which phosphorylates insulin receptor substrate
(IRS) adaptor proteins to create a negative feedback that shuts off insulin signaling28. Further,
overfeeding-induced endoplasmic reticulum (ER) stress could also prevent the ability of insulin
to lower glucose production, since icv-3 treatment with the ER stress- inducer, thapsigargin,
prevented the ability of central insulin to lower blood glucose levels29. Taken together, these
studies associate hypothalamic insulin resistance to diet-induced dysregulation of glucose and
energy balance.
Leptin
Like insulin, leptin is another hormone that has high relevance in diabetes and obesity. Leptin
circulates in the plasma in proportion to body fat stores, enters the brain and interacts with its
receptors to regulate food intake, body weight, and glucose homeostasis30. It is now evident that
the metabolic effects of leptin are likely mediated through the brain. In both rats31 and mice32, icv
administered leptin show similar metabolic effects as intravenously (iv) administered leptin. For
instance, low-dose icv leptin recapitulates the effects of iv leptin in reversing the insulin
resistance and diabetic phenotype of leptin deficient, ob/ob33 and lipodystrophic mice34.
Similarly, icv leptin normalizes hyperglycemia in STZ-induced Type 1 diabetic rats whereas iv
leptin at the same dose is unable to do so35,36. Notably, the hypothalamus per se can regulate
leptin-glucose control as evident by Coppari et al.’ and Morton et al.’s studies where selective
restoration of leptin receptors in the hypothalamus normalizes blood glucose levels of the obese,
hyperinsulinemic and severely diabetic leptin receptor (LepRb)-null mice37 and rats38,
respectively. Consistently, injection of leptin specifically into the ARC leads to reduced food
intake and body weight gain in rats39 and icv administration of leptin in ob/ob mice attenuates
7
obesity37. The STAT3 pathway plays a critical role in central leptin’s regulation of energy
homeostasis while there is some controversy over its role in glucose regulation. Mice with a
whole-body disruption of LepRb à STAT3 signaling (s/s mice) are hyperphagic and obese with
reduced energy expenditure, but less hyperglycemic compared to LepRb deficient, db/db
mice40,41. In contrast, mice with neural-specific ablation of STAT3, in addition to being
hyperphagic and obese, show impaired glucose tolerance and hyperinsulinemia to the same
extent as db/db mice42. Mice with specific deletion of STAT3 from AgRP neurons display
obesity, increased NPY expression, but did not differ in glucose levels compared to control
mice43. STAT3 deletion from POMC neurons increases adiposity, but the effect is milder than
that observed for the AgRP-specific knockout, suggesting a greater role for STAT3 in leptin’s
regulation of energy homeostasis in AgRP neurons than in POMC cells44. It should be noted,
evidence also suggests most of the anorectic actions of leptin is dependent on the inhibition of
GABA neurons outside of the ARC causing a reduced reduced inhibitory tone to POMC
neurons45. Importantly, STAT3 activation is a functional determinant for hypothalamic leptin’s
regulation of glucose homeostasis, as pharmacological and molecular inhibition of STAT3 in the
third ventricle abolishes the ability of MBH leptin to lower glucose production in rodents fed a
high fat diet46. The extracellular signal–regulated kinase (ERK), a member of the mitogen-
activated protein kinase (MAPK) family, is another downstream pathway of the leptin receptor.
The ERK pathway appears to modulate brain leptin’s control on energy and glucose balance.
Chemical blockade of hypothalamic ERK1/2 negated the food intake- and weight-reducing
effects of icv leptin47, while genetic blockade of ERK signaling in POMC neurons negated the
glucose lowering effects of iv leptin with only a modest effect on food intake48. In addition to
STAT3 and ERK, the LepRb receptor also signals through the PI3K pathway. Pharmacological
blockade of PI3K activity in the hypothalamus abolishes the ability of leptin to lower food
8
intake49. Interestingly, chemical blockade of PI3K prevents insulin-stimulated suppression of
glucose production otherwise seen upon restoration of LepRb in Koletsky fak/fak rats38. The PI3K
pathway is also implicated in leptin-induced depolarization of POMC neurons50, whose role in
regulating normoglycemia is also dependent on PI3K signaling51. Notably, like insulin, leptin
also activates KATP channels to hyperpolarize hypothalamic neurons obtained from lean rats, but
not in obese rats52.
Furthermore, AMP-activated kinase (AMPK) has also been identified as an important
player in leptin's hypothalamic intracellular signaling cascade. Exogenous administration of
leptin inhibits AMPK in the ARC, while constitutive activity of AMPK in the ARC abolishes
leptin's effect on anorexia, thereby indicating that inhibition of hypothalamic AMPK is necessary
for leptin's regulation on energy homeostasis53. Subsequent studies have demonstrated that
activation of acetyl-CoA carboxylase (ACC), the downstream target of AMPK inhibition, further
mediates hypothalamic leptin’s control on food intake and body weight54. Consistently, increased
levels of malonly-CoA, a direct product of ACC activation and an important mediator in the
lipid-sensing pathway, are also required for hypothalamic leptin-induced reduction in food
intake55.
Importantly, hypothalamic leptin resistance occurs in an obese state since icv leptin fails
to induce anorexia in diet-induced obese (DIO) rats56, contributing to the notion that
hypothalamic dysregulation of hormonal action leads to obesity. Surprisingly, the effect of icv
leptin to decrease glucose production is intact in both 3 d HFD-fed as well as STZ diabetic
rats35,57. It is possible, that these differences in CNS leptin resistance could be due to specific
signaling pathways activated by leptin. Where activation of hypothalamic STAT3 is required for
MBH leptin’s effect on glucose production46, HFD impaired the anorectic effects of leptin
despite intact STAT3 signaling58.
9
Ghrelin
The peripherally derived-orexigenic hormone, ghrelin, is another critical regulator of metabolism
modulated through hypothalamic pathways. Ghrelin regulates feeding, and glucose homeostasis
via ghrelin receptors abundantly expressed in the ARC AgRP/NPY neurons 59,60. Indeed, direct
ghrelin administration into the ARC increases food intake, induces NPY and AgRP mRNA
expression in the ARC and its orexigenic effect on food intake was negated in the presence of icv
administration of anti-NPY or anti-AgRP antibodies 61. Further, icv ghrelin infusion has also
been reported to improve glucose tolerance in pair-fed mice, compared to icv ghrelin ad libitum
mice and icv vehicle treated mice 62. While the molecular players for the glucoregulatory action
of icv ghrelin remain largely unknown, the key pathway for the orexigenic effect of ghrelin
includes AMPK since icv ghrelin increases AMPK phosphorylation in the hypothalamus63 and
molecular inhibition of AMPK markedly blunted the feeding effect of ghrelin64. Consistently,
ghrelin-induced AMPK phosphorylation decreased ACC levels, accompanied by reduced
formation of malonyl-CoA levels and enhanced CPT-1 activity, which in turn leads to
mitochondrial fatty oxidation64. This hypothalamic fatty oxidation pathway increases reactive
oxygen species (ROS), activating uncoupling protein 2 (UCP2), which mediates sustained firing
of AgRP/NPY neurons, and increased GABAergic inhibitory tone to POMC neurons –
ultimately increasing feeding behaviour65. Whether an AMPK mediated pathway is involved in
the glucose regulating effects of icv ghrelin remains to be identified.
Glucagon
The second principal pancreatic hormone, glucagon, is well known for its catabolic effect on fuel
metabolism. Once released into the circulation, glucagon stimulates hepatic glucose output
through the breakdown of glycogen, inhibition of glycogen synthesis and stimulation of
10
gluconeogenesis66. Glucagon exerts these biological effects by binding to G-protein coupled
glucagon receptors in the liver and activating the cyclic AMP-protein kinase A (cAMP-PKA)
signaling pathway67,68. In addition to the liver, glucagon receptors (GR) are also expressed in the
brain69,70, and the findings that other key components of the glucagon signaling pathway
including cAMP and PKA are present in the hypothalamus71, and that glucagon can be
transported from blood into the cerebrospinal fluid72 and into the brain to affect hypothalamic
neuronal activity73 suggests that circulating glucagon passes through the BBB and could exert
part of its biological effects through a hypothalamic pathway.
In fact, multiple studies have documented the ability of icv glucagon administration to
modulate peripheral glucose levels in different species of animals74-76. Most intriguing, however,
is the earliest study in dogs where a high dose icv injection of 10 ng of glucagon transiently
produced hypoglycemia followed by hyperglycemia77. The hypoglycemic effect was abolished in
vagotomised dogs, suggesting the involvement of a brain-liver axis in the glucose-lowering
effect of central glucagon; whereas pancreatectomy prevented the hyperglycemic effect,
attributing a pancreatic role to the rise of glucose from icv glucagon injections 77,78. Given the
non-specific administration of icv glucagon and the use of relatively high glucagon dosage in
these experiments, more recent studies have administered much lower doses of glucagon
specifically into the MBH and evaluated whether MBH glucagon action accounts for the
glucose-lowering effect or hyperglycemic effect of icv injections in the early dog studies.
Indeed, direct infusion of glucagon into the MBH actually lowered hepatic glucose
production in a pancreatic basal-insulin euglycemic clamp condition79. This effect of MBH
glucagon required the activation of MBH GRs, PKA signalling, and intact vagal nerves, since
their ablation negated the metabolic effects of MBH glucagon infusion. The inhibitory metabolic
effects of central glucagon were also confirmed in another rodent model; central glucagon, but
11
not saline, administration in C57BL/6 normal mice during a pancreatic clamp increased the
glucose infusion rate required to maintain euglycemia, owing to a suppression in glucose
production79. Furthermore, a genetic knockout model involving mice that lack functional
glucagon receptors (Gcgr−/−) also confirmed the role of central glucagon receptors in mediating
the central glucagon effect since central infusion of glucagon in these mice failed to increase the
glucose infusion rate and lower glucose production compared to Gcgr+/+ control mice. Taken
together, these data illustrate that hypothalamic/central glucagon receptor signaling regulates
glucose production in both rats and mice and is possibly the mechanism responsible for the
initial hypoglycemic effect of central glucagon injections seen in dogs as previously reported77,78.
The next key question was whether the hypothalamic action of glucagon to lower glucose
production is physiologically relevant—does the MBH mediate the effect of circulating glucagon
to regulate glucose production? Despite the classical view of glucagon to be a gluco-stimulatory
hormone, continuous iv infusions of glucagon only transiently stimulate glucose production, with
a subsequent decline back to baseline after approximately 40 min in both dogs80 and humans81,82.
Interestingly, Mighiu et al. demonstrated blocking MBH glucagon signaling prolonged the
gluco-stimulatory effect of iv glucagon, thereby ascribing the short-lived elevation of glucose
production and glycemia induced by circulating glucagon to, and illustrates the physiological
relevance of, MBH glucagon action.
On the contrary, in diabetic83 and obese84 conditions chronic hyperglucagonemia is
associated with increased glucose production and blood glucose levels. For instance, in diabetic
rodents, antagonism of glucagon action by iv injection of either glucagon receptor antagonist
THG85 or monoclonal glucagon antibody86 alleviates hyperglycemia, suggesting that glucagon
leads to a sustained increase in blood glucose levels in diabetes. Therefore, given that glucagon’s
gluco-stimulatory effect is evanescent in normal physiology whereas it is sustained in diabetes, it
12
is likely that the absence of glucagon’s transient effect in diabetes and obesity may be attributed
to hypothalamic glucagon resistance. In fact, 3 d HFD feeding has been shown to disrupt MBH
glucagon action to lower glucose production79. This indicates that hypothalamic glucagon
resistance indeed exists in a pathological state. However, this resistance could be reversed by
direct activation of hypothalamic PKA, which implies, at this preliminary stage, that the
hypothalamic glucagon signaling defect lies upstream of PKA. Investigating what lies
downstream of PKA in MBH glucagon signaling thus becomes of scientific interest, especially
from a therapeutic standpoint to identify targets that could potentially enhance or restore
hypothalamic glucagon action in diabetes and/or obesity. Interestingly, PKA decreases AMPK
activation in adipocytes87 and hypothalamic cell lines88. The aim of Study 1 in this dissertation is
to further uncover the molecular pathways that lie downstream of MBH glucagon-PKA signaling
in the regulation of glucose homeostasis.
Notably, the effects of MBH glucagon are not just limited to glucose regulation. The
earliest evidence that hypothalamic glucagon action regulates food intake comes from Inokouchi
et al.’s study which showed that icv-3 injection of glucagon significantly suppressed feeding in
rats more potently than the effect observed with peripheral administration89. Similar anorexigenic
effects of central glucagon were also demonstrated in chicks90 and sheep91. Notably, one study
showed that central glucagon administration did not alter food intake in baboons13; however, the
glucagon dose used in this study was 10 times lower than that what was used to achieve feeding
inhibition in rats89. Although iv infusion of glucagon reduces meal size and suppresses appetite
in humans92, whether this is due to an effect of glucagon in the hypothalamus is not clear.
In summary, peripherally derived hormones such as insulin, leptin and glucagon have
been documented to act on their respective receptors in the hypothalamus to regulate glucose and
energy homeostasis via seemingly distinct signaling pathways, but perhaps also converging at a
13
common downstream effector (s). Notably, the activation of KATP channels presents as a likely
candidate, as not only is it necessary for the action of MBH insulin and leptin, but also in MBH
nutrient sensing mechanisms which be reviewed as follows.
1.1.2 Nutrient sensing
In addition to integrating hormonal signals, the hypothalamus is also a sensor of nutritional
inputs including those of fatty acids, glucose and amino acid leucine to modulate glucose
homeostasis and energy balance.
Fatty acids
Although fatty acids per se are not an energy source for the brain, lipid sensing mechanisms have
emerged as important indicators of nutrient availability to the CNS to regulate glucose and
energy homeostasis. In fact, icv-3 infusion of the long chain fatty acid (LCFA), oleic acid during
basal insulin clamping leads to a significant inhibition in the rate of hepatic glucose production,
and during an unclamped postprandial setting, effectively lowers plasma glucose levels93. The
hypothalamus can also sense circulating fatty acids to regulate glucose homeostasis94.
Circulating fatty acids can passively diffuse across the BBB95. At the same time, facilitated
transport of lipids can also occur via plasma membrane fatty acid transport proteins and fatty
acid translocases/CD36 96-98. Upon entry into cells, LCFAs are subsequently esterified to LCFA-
CoA by acyl-CoA synthetase (ACS) which prevents diffusion out of neurons. In fact, the
accumulation of intracellular LCFA-CoA is necessary to activate hypothalamic lipid sensing
pathway and lower glucose production as direct inhibition of MBH ACS negates the ability of
circulating LCFA to restrain glucose production94. Intracellular LCFA-CoA levels are also
dependent on carnitine palmitoyltransferase-1 (CPT1) activity, which regulates the uptake of
14
LCFA-CoA into mitochondria for β-oxidation. In line with this, genetic inhibition of
hypothalamic CPT1 expression increased LCFA-CoA levels and suppressed glucose
production99. Therefore, factors regulating CPT1 activity are also key components of the lipid
sensing pathway. One such factor is malonyl-CoA which inhibits CPT1 to prevent lipid entry
into the mitochondria100. Thus, malonyl-CoA, which is a product of glucose metabolism and
therefore, abundant during nutrient availability, promotes the accumulation of LCFA-CoA to
lower glucose production and circumvent circulating nutrient excess101. AMPK is another
important negative regulator of hypothalamic lipid sensing. As described earlier, AMPK inhibits
ACC activity consequently preventing the formation of malonyl-CoA from acetyl-CoA and this
abolishes the glucose-lowering actions of MBH lipid sensing mechanisms102.
As with MBH insulin and leptin infusions, hypothalamic KATP channels are also activated
by hypothalamic lipid sensing, and are necessary for hypothalamic LCFA to suppress glucose
production103. Consistent with this, blockade of hypothalamic KATP channels by both
pharmacological and genetic approaches abolished the glucose production-lowering effect
achieved by physiological increases in circulating LCFAs94. The activation of KATP channels
subsequently signals through the hepatic vagus nerve to communicate with the liver as hepatic
vagotomy negates the hypothalamic lipid sensing mechanism to inhibit the hepatic expression of
gluconeogenic enzymes and glucose production104. Further, hypothalamic protein kinase C
(PKC) was later shown to be an important mediator in the hypothalamic lipid-KATP signaling to
regulate glucose homeostasis. While direct hypothalamic infusion of the PKC activator, OAG
lowered glucose production, this effect was eliminated by co-infusion with rottlerin, an inhibitor
specifically targeting the PKC-δ isoform as well as by blocking hypothalamic KATP channel105.
Lastly, inhibition of hypothalamic PKC negated the effect of hypothalamic lipids to lower
glucose production105.
15
Hypothalamic fatty acid sensing also modulates energy balance. Rats that received a
bolus of icv-3 oleic acid 1 hour prior to the dark cycle showed a significant reduction in food
intake- an effect that lasted for 2 days103. Further, the reduced expression of hypothalamic NPY
mRNA levels observed in rats treated with icv-3 oleic acid vs vehicle-bolus, likely represents the
underlying mechanism behind the anorectic effect of icv oleic acid.
Similar to how various hypothalamic hormonal signaling mechanisms are disrupted in
models of high fat feeding, 3 d HFD feeding disrupts the ability of icv-3 oleic acid infusion to
lower glucose production93. Interestingly, genetic inhibition of hypothalamic CPT1 expression
by icv infusion of the ribozyme CPT1A-Ribo restored LCFA-CoA levels and rescued the gluco-
regulatory effect of hypothalamic lipid sensing in the 3d HFD fed rats106. Moreover, direct
activation of MBH PKC-δ was also able to restore the suppression of glucose production during
3d HFD105. These data are indicative that the HFD-induced defect in lipid sensing lies at the
level of CPT1 and LCFA-CoA accumulation, while the downstream signaling pathway of
LCFA-CoA remains intact after a 3d HFD. Notably, icv oleic acid also failed to suppress food
intake in rats fed 3d HFD103. This proves compelling evidence that hypothalamic lipid sensing
can become defective and is of therapeutic relevance in diabetes and obesity.
Because of the involvement of 1) AMPK and KATP channels in the hypothalamic action of
leptin and/or insulin 2) AMPK, PKC- δ, and KATP channels in hypothalamic lipid sensing
mechanism, and 3) that PKA inhibits AMPK, we tested in study 1, the role these respective
intracellular signaling effectors in MBH glucagon action.
16
Glucose and lactate
Interestingly, the hypothalamic lipid sensing axis illustrated as above is positioned to mediate
hypothalamic glucose sensing. In addition to being the primary fuel for the brain, glucose
sensing in the hypothalamus plays an important role in the physiological control of blood glucose
and feeding regulation. It was Lam et al. who first provided evidence that during a pancreatic
clamp, a direct infusion of glucose or lactate, a byproduct of glucose metabolism, into the MBH
leads to an inhibition of hepatic glucose production, which, like MBH insulin and leptin action,
also requires activation of hypothalamic KATP channels107. Notably, when hypothalamic lipid
sensing axis is blocked via activation of MBH AMPK, it abolished the effect of MBH glucose
and lactate sensing to lower glucose production102. It is not surprising, increase in hypothalamic
glucose flux lowers food intake, and this anorectic effect is also mediated by inhibiting AMPK
and activating ACC activity, which in turn elevates hypothalamic malonyl–CoA108. Taken
together, AMPK-mediated lipid sensing appears to be a critical intermediary mechanism by
which hypothalamic glucose sensing works to regulate glucose production and feeding behavior.
The literature documents that defective hypothalamic glucose sensing maybe an
important pathogenic component of diabetes and obesity. This is particularly true for MBH
glucose sensing in uncontrolled diabetes. Chari et al. showed in rats with STZ-induced
uncontrolled diabetes, icv-3 administration of glucose was unable to suppress the rate of glucose
production, and they demonstrated this defect was due to a glucotoxic insult induced by
sustained hyperglycemia109. Intriguingly, MBH lactate administration, using the same dose as for
normal rats, retained its ability to inhibit glucose production in the STZ-induced diabetic rats, as
well as in experimentally induced hypoinsulinemic rats, and in rats fed a 3d HFD110. Thus, it is
likely that defective mechanisms upstream of lactate metabolism are at play in hindering
effective hypothalamic glucose sensing.
17
Amino acid, leucine
Amino acid sensing in the hypothalamus also elicit potent metabolic influence on peripheral
glucose and energy regulation. Leucine is one of the three branched-chain amino acids that have
been documented to represent a physiological signal of hypothalamic amino acid availability.111
Dietary leucine escapes first-pass metabolism resulting in a rapid increase in plasma leucine
levels during postprandial conditions, and leucine reaches the brain sooner than other amino
acids112. A recent study shows that direct infusion of leucine in the MBH lowers blood glucose
through suppression of glucose production113. This gluco-regulatory effect requires the
metabolism of leucine to acetyl-CoA and the subsequent conversion to malonyl-CoA as well as
the activation of hypothalamic KATP channels. Similarly, leucine sensing also contributes to the
hypothalamus modeulated lowering of food intake and body weight, and this effect requires
leucine -induced activation of the hypothalamic mammalian target of rapamycin (mTOR)
signaling114. Studies investigating the downstream neural targets in hypothalamic leucine sensing
revealed that in addition to hypothalamic POMC neurons, MBH leucine also activates brainstem
DVC neurons to lower food intake111. This finding identified for the first time an MBH-DVC
neurocircuit engaged by MBH leucine to exert homeostatic control on energy homeostasis. As
with fatty acids and glucose, impaired hypothalamic leucine sensing may also have a role in
disease development. While molecular disruption of leucine sensing in the MBH resulted in
hyperglycemia in rats fed a high-protein diet113, 3d HFD feeding also attenuated the ability of
central leucine to modulate glucose metabolism115.
18
1.2 The role of the dorsal vagal complex in the regulation of glucose and energy homeostasis
Although the historical emphasis on the CNS regulation of glucose and energy homeostasis has
been on the MBH, the DVC, in the brain stem, is another key brain site of hormonal action and
nutrient sensing that can regulate metabolic homeostasis. Like the MBH, the DVC contains a
circumventricular organ called the area postrema. In addition, the DVC contains the nucleus of
the solitary tract (NTS) and the dorsal motor nucleus of the vagus (DMX) with synaptic
connections to vagal afferent and efferent fibres, respectively. These features enable humoral and
neural signals to reach the DVC in order to modulate various autonomic functions including
glucose and energy regulation. As described previously, the DVC is a critical site for the
integration of lipid, glucose and leucine sensing inputs from the hypothalamus to affect glucose
and feeding control. This is because neuronal projections from the hypothalamus reach the
hindbrain forming a hypothalamic-DVC neuronal axis to exert homeostatic feedback on glucose
and feeding control. In this regard, studies have identified the N-Methyl-d-aspartate (NMDA)
receptors in the DVC to act as an important switchboard in integrating hypothalamic signals to
maintain glucose and energy homeostasis. Previously, neurotransmission by the DVC NMDA
receptors was shown to play an important role in relaying signals generated by intestinal nutrient
and hormonal signalling mechanisms to lower glucose production and feeding behavior116-120.
However, it was Lam et al. who directly tested and confirmed that pharmacological and
molecular blockade of DVC NMDA receptors indeed also negates the ability of hypothalamic
nutrient sensing mechanisms activated by glucose (or more specifically lactate) metabolism or
hypothalamic lipid sensing to lower glucose production121.
19
In addition to the role of the DVC as a relay centre for hypothalamic-sensed signals,
multiple studies have also confirmed that the DVC is equally a direct sensor for different
hormones and nutrients to regulate metabolic homeostasis.
1.2.1 Hormonal action
Insulin
The earliest evidence that the DVC is insulin responsive came from Filippi et al122 who showed
that direct insulin infusion into the DVC (targeting the NTS) activated insulin receptor-mediated
signaling events in a dose-dependent manner. Indeed, DVC insulin signaling lowers glucose
production in rodents, while inhibition of DVC insulin receptors negated the glucose production-
lowering effect. Secondly, this study also addressed which of the DVC insulin receptor-mediated
signaling pathways regulate glucose production. When insulin was infused into the DVC at the
same dose as that which activated PI3K when administered into the MBH28,123, insulin failed to
activate DVC PI3K. Instead, DVC insulin activated Erk1/2, while inhibition of DVC Erk1/2
signaling by a chemical approach, as well as a molecular approach abolished the ability of
insulin to lower glucose production. In direct contrast, inhibition of DVC PI3K signaling did not
alter the ability of insulin to control glucose levels. Furthermore, directly activating DVC Erk1/2
lowered glucose production, strengthening the gluco-regulatory role of DVC Erk1/2 signaling.
Thus, insulin activates an Erk1/2-dependent signaling pathway in the DVC to lower glucose
production, as opposed to the MBH, where insulin lowers glucose production via PI3K signaling.
Finally, activation of KATP channels in the DVC was also found necessary for insulin–Erk1/2
signaling to inhibit glucose production.
Indeed, direct administration of insulin into the DVC also affected feeding behavior.
DVC insulin infusion lowered food intake as early as 90 min compared with DVC saline
20
infusions124. The mechanism underlying DVC insulin-induced satiety involves the activation of
DVC Erk1/2, because molecular and chemical inhibition of Erk1/2 signaling negated the ability
of DVC insulin infusion to reduce feeding. In contrast, inhibition of the PI3K–Akt pathway did
not affect the action of insulin in the DVC, thereby suggesting that DVC insulin activates an
Erk1/2-dependent and a PI3K–Akt-independent pathway to regulate feeding.
High-fat feeding disrupts the action of insulin in the DVC and dysregulates feeding
control and glucose production in association with the inability of insulin to activate Erk1/2 in
the DVC122,124. However, future studies are warranted to characterize the underlying mechanisms
of insulin resistance in the DVC that concurrently dysregulate feeding and glucose control.
Leptin
As for the effects of insulin, the DVC is also a target for the inhibitory effect of leptin on food
intake. Rats treated with an injection of leptin into the 4th ventrice (icv-4), presumably targeting
the DVC, showed reduced food intake at 2, 4, and 24 h after injection and displayed significant
reduction in body weight gain125. Further, increased expression of the leptin transgene in the
DVC completely curbed the time-related gain in body weight in rats despite only a transient
suppression of food intake126.
Glucagon
In light of the fact that glucagon receptors are abundantly expressed in the brainstem70,127, and
peripherally administered glucagon alters neuronal activity in the brainstem128, it was plausible
that, analogous to the MBH, glucagon could initiate a signalling cascade in the DVC to regulate
peripheral metabolism. While it remains to be investigated whether glucagon acts in the DVC to
regulate food intake, a recent study has revealed that glucagon action in the DVC can impact
glucose homeostasis, and has postulated a novel physiological role of DVC glucagon action in
21
postprandial conditions129. During pancreatic basal insulin clamp conditions, glucagon infusion
into the DVC reduced hepatic glucose production in healthy rodents. Using complementary
chemical and molecular loss-of-function approaches, blocking activation of the glucagon
receptor, PKA, ERK 1/2 or KATP channels in the DVC reversed the glucose production-lowering
effect of DVC glucagon infusion, thereby implicating each of these as the downstream signalling
mediators in the DVC glucagon signalling pathway. The physiological relevance of DVC
glucagon action in regulating glucose homeostasis was investigated in the context of the
postprandial period following high-protein meals. During fasting–refeeding experiments, high-
protein feeding acutely decreased plasma glucose levels compared with a low-protein diet; but
when DVC glucagon receptor signalling was blocked, the ability of high-vs low-protein feeding
to suppress the rise in plasma glucose was negated.
1.2.2 Nutrient sensing
Glucose
It has long been shown that there are glucose responding neurons in the DVC130. Specifically,
these neurons employ glucokinase and KATP channels to alter their action potential firing in
response to varying glucose concentrations131. Microinjections of glucose into the NTS or DMV
sites of the DVC lead to inhibition of gastric motility and increased intragastric pressure132.
Surprisingly, very little work has been done to investigate the functional effects of direct glucose
administration in the DVC neurons on the control of glucose and energy regulation, other than
the studies that have identified that glucose sensing neurons in the DVC participate in the
regulation of feeding and glycemic responses elicited by hypoglycemia133. For instance,
microinjection of glucose antimetabolites into various hindbrain sites has been shown to induce
at least 1.5 g more in food intake or at least more than 25 mg/dl in hyperglycemic response
22
compared to vehicle-injected rats133. Further insights into the physiological potential of direct
DVC glucose sensing as a signal of energy availability remain unexplored.
Leucine
Similar to the MBH, the DVC possesses amino acid nutrient sensing capabilities modulating
metabolic homeostasis. Direct NTS administration of leucine, using a postprandial physiological
dose, significantly reduced meal size, 24 hour food intake and 24 hour body weight change, and
this effect is mediated by activation of the mTOR/S6K1 pathway134. The same study also
identified catecholaminergic and POMC neurons as the primary leucine- sensing neurons in the
DVC. Moreover, DVC leucine administration also elicits activation of DVC Erk1/2 signaling
pathway, which is also necessary for the anorectic effect of DVC leucine. Lastly, it was also
observed that co-administration of subthreshold doses of CCK admininstered ip or DVC
melanotan II (an analogue of the melanocortin peptide hormone) or DVC leptin enhanced the
acute anorectic effects of DVC leucine, thereby highlighting the integrating role DVC leucine
sensing plays with gut-derived signals, forebrain-descending melanocortinergic signals and
adiposity signals of energy availability to control energy homeostasis134.
As with DVC insulin action, HFD feeding leads to an impairment in DVC leucine
sensing in the regulation of feeding135. HFD negates the anorectic effects of DVC leucine
administration. This was accompanied by an increase in the baseline activity of the S6K1 and
impaired leucine-induced activation of this pathway in the DVC of HFD-fed mice, further
implicating impaired DVC mTOR sensing during HFD. Notably, the synergistic suppressive
effect of ip CCK and DVC leucine was also blunted in the DIO mice, thereby indicating that
HFD also impairs the gut brain integration in the DVC in the control of meal size.
23
Glycine
Accumulating evidence shows that sensing of the amino acid glycine in the DVC has a
physiological role in the regulation of peripheral homeostasis. Data show there is a high
concentration of glycine terminals in the DVC and that glycine injections into the DVC elicit
changes in heart rate and blood pressure136. Glycine binds to the GluN1 subunit of the NMDA
receptor and acts a co-agonist along with glutamate to potentiate NMDA receptor activation137.
In line with this, direct administration of glycine in the DVC activates NMDA receptors to lower
hepatic glucose production, which was abolished in the presence of NMDA receptor antagonism
and hepatic vagotomy138. Subsequent studies have revealed that glycine sensing via DVC
NMDA receptors is sufficient to also lower hepatic-triglyceride secretion139. These data are in
agreement with other studies demonstrating that binding of glycine to NMDA receptor serves to
potentiate NMDA receptor-mediated neurotransmission 140-142.
Given that CNS hormonal and nutritional sensing mechanisms are impaired in models of
obesity and/or diabetes, the question arises whether the effectiveness of DVC glycine to lower
glucose production is intact in settings of insulin resistance, uncontrolled diabetes and DIO. We
are encouraged by the findings that DVC glycine infusion, at the same dose infused in normal
rodents, is still able to lower hepatic lipid production in an acute diet-induced insulin resistance
model139 and that other central nutrient sensing mechanisms, such as those of central lactate are
preserved in an early-onset model of STZ-diabetes110. Further, no studies till date have directly
tested whether DVC glycine sensing can regulate energy homeostasis. However, given that DVC
NMDA receptors play a role in the control of food intake (as described previously) elicited by
hormones such as cholecystokinin143, and that glycine binding acts as a modulator to NMDA
transmission in the DVC, it is plausible that glycine triggers a sensing mechanism via DVC
NMDA receptors to reduce appetite.
24
However, glycine as a CNS-active drug suffers from the limitation of having a poor
pharmacokinetic profile144. For instance, following oral administration of glycine, although
elevated glycine levels in the cerebral cortex, there was a concomitant increase in the rate of
glycine uptake and rapid conversion of glycine into serine in brain tissue, thereby limiting brain
exposure to extracellular glycine145. Indeed, studies from the schizophrenia field have reported
the use of glycine administration as an approach to activate NMDA receptor-mediated
transmission in schizophrenic patients. Schizophrenia displays reduced NMDA receptor
function, and therefore increasing NMDA receptor function via pharmacological manipulation is
integral to the treatment for schizophrenia. While there are some encouraging studies reporting
that large doses of glycine can improve the negative symptoms in schizophrenic patients146,147,
there are studies that have failed to confirm the efficacy of glycine as a therapeutic agent148. It is
not clear whether the conflicting evidence is due to low CNS exposure to extracellular glycine.
In fact, the level of extracellular glycine in the CNS is primarily dependent on regulation
by the Na+/Cl−-dependent glycine transporters, GlyT1 and GlyT2149. In particular, GlyT1, which
is expressed on glial cells, is the primary transporter of glycine mediating the uptake of glycine
into cells near NMDA receptors. Thus, blockade of GlyT1 could increase synaptic glycine levels,
thereby potentiating activation of NMDA receptors. Indeed, GlyT1 inhibitors have been tested in
rodents and preliminary clinical studies and have proved to be beneficial in the treatment of
schizophrenia150-153.
Given the unfavourable properties of glycine administration, the question arises as to
whether GlyT1 inhibition in the DVC would be a favourable approach to trigger glycine sensing
in the brainstem and more importantly, whether there is a novel therapeutic potential for DVC
GlyT1 inhibition, in the treatment for diabetes and obesity to lower glucose levels and body
weight via the activation of NMDA receptors.
25
1.3 The role of glial cells in the regulation of glucose and energy homeostasis. When glial cells were first discovered in the 1800s, they were viewed as merely “cellular glue”
for the brain, holding neurons together. However, in recent years, there has been accumulating
evidence pointing that far from being passive, glial cells play a critical role in the normal
functioning of the brain including synaptic plasticity, development, neurotransmission and
metabolism154. In fact, astrocytes, a type of glial cells, are emerging as important regulators of
nutrient and energy sensing mechanisms in the CNS, and they do so by expressing specific
hormonal receptors and nutrient transporters. For instance, in the hypothalamus, astrocytic
insulin receptors are indispensible for proper glucose and insulin entry into the brain, in turn
contributing to CNS regulation of systemic glucose and energy homeostasis155. Similarly, leptin
signaling in hypothalamic astrocytes has also been reported to play an important role in the CNS
control of feeding156. At the same time, studies also show that glial specific glucose transporter,
GLUT1 is vital for hypothalamic glucose sensing to regulate peripheral glucose levels109, and
that lipoprotein lipase (LPL) in astrocytes controls lipid uptake in the hypothalamus for central
regulation of body weight and glucose metabolism157. Notably, in addition to transporting
circulating nutrients and expressing hormonal receptors, glial cells can also contribute to
systemic metabolic control via uptake of neurotransmitters from the synaptic cleft. For instance,
glial cells exclusively expressing glutamate transporter (GLT)-1, mediates glutamate uptake into
astrocytes, which is critical for regulating synaptic transmission by this excitatory amino acid.
Further, glutamate uptake into astrocytes has also been reported to increase glycolysis and lactate
production, in turn modulating nutrient availability for neurons107,158. Therefore, CNS regulation
of nutrient sensing and hormonal signals is, atleast in part, directed by glial cells. However, what
remains to be shown and will be addressed in this dissertation, is whether the glial specific
26
glycine transporter, GlyT1 like GLT1, can couple CNS nutrient mechanisms to control systemic
metabolic homeostasis.
1. 4 Summary of Introduction/Rationale of Study 1 and Study 2 Type 2 diabetes and obesity, the two largest public health concerns of today, are progressive
metabolic disorders of glucose and energy homeostasis. Over the last two decades, significant
progress has been made in support of the role the brain plays in regulating peripheral glucose and
energy homeostasis. The two key regions of the CNS: the MBH and the DVC can receive and
integrate information from hormones and nutrients to subsequently direct changes in hepatic
glucose production and feeding behaviour, and they do so by using distinct and sometimes
common receptors and intracellular signalling pathways. AMPK and KATP are two such crucial
intracellular signaling pathways responsible for the metabolic effects of MBH insulin, leptin,
glucose, fatty acids and amino acid leucine to regulate whole-body energy homeostasis and
glucose control. More recently, a novel gluco-regulatory role of MBH glucagon was discovered
to lower glucose production via the PKA pathway. Whether MBH glucagon action is
downstream mediated by AMPK and KATP signaling remains to be investigated. The focus of
Study 1 was to evaluate whether AMPK-mediated lipid sensing and KATP channels are
necessary for MBH glucagon for glucose regulation.
Moreover, the smallest amino acid glycine triggers neurotransmission in the DVC to
regulate metabolic homeostasis including glucose and lipid metabolism using NMDA receptors.
Importantly, DVC glycine sensing can normalize the hypersecretion of lipids induced by 3d
HFD. These findings highlight the therapeutic potential of glycine sensing to lower blood lipids
in individuals with obesity and diabetes. Could there be a novel therapeutic potential for DVC
glycine manipulation in lowering blood glucose and body weight in obese and diabetic
27
individuals? Because glycine has poor pharmacokinetics, and because glycine levels in the
CNS are enhanced by GlyT1 inhibition, in Study 2, we investigated whether DVC GlyT1
inhibition could sufficiently trigger glycine sensing to improve glucose and energy
homeostasis in diabetes and obesity.
28
Chapter 2 Study 1
Hypothalamic Glucagon Signals through the
KATP Channels to Regulate Glucose Production
Modified from:
Abraham, M.A., Yue, J.T., LaPierre, M.P., Rutter, G.A., Light, P.E., Filippi, B.M., and Lam,
T.K. (2014). Hypothalamic glucagon signals through the KATP channels to regulate glucose
production. Molecular Metabolism 3, 202-208.
Permission to reproduce portions of the above manuscript has been obtained from the copyright
owner: Elsevier Limited
29
2.1 Abstract Background and Aims: Gluco-regulatory hormones such as insulin, leptin and GLP-1 signal in
the mediobasal hypothalamus (MBH) to lower hepatic glucose production glucose production.
MBH glucagon action also inhibits glucose production, but the downstream MBH signaling
mediators remain largely unknown. In parallel, a lipid-sensing pathway involving MBH AMPK -
> malonyl-CoA -> CPT-1 -> LCFA-CoA -> PKC-δ leading to the activation of MBH KATP
channels has been documented to lower glucose production. Given that glucagon signals through
the cAMP-PKA pathway in MBH to lower glucose production, and PKA inhibits AMPK in
hypothalamic cell lines, a possibility arises that glucagon-PKA action in the MBH would inhibit
AMPK, elevates LCFA-CoA levels to activate PKC-δ, and activates KATP channels to lower
glucose production. Methods: Using molecular and chemical approaches, we inhibited the MBH
lipid-sensing pathway in normal male Sprague-Dawley rats via (i) activation of MBH AMPK or
inhibition of MBH PKC-δ, and (ii) MBH KATP channels in the presence of MBH glucagon
stimulation and evaluated the changes in glucose kinetics during the pancreatic (basal insulin)
euglycemic clamps. Results: We found that neither molecular and chemical activation of MBH
AMPK nor inhibition of PKC-δ negated the gluco-regulatory effect of MBH glucagon. In
contrast, molecular and chemical inhibition of MBH KATP channels negated MBH glucagon’s
effect to lower glucose production. Conclusion: Our data collectively indicates that MBH
glucagon signals through a lipid-sensing independent but KATP channel-dependent pathway to
regulate glucose production.
30
2.2 Introduction A role of glucagon action in the MBH was documented, in contrast to the hormone’s hepatic
stimulatory effect, to lower glucose production79. This glucose production-lowering effect
required the activation of the MBH glucagon receptor-cAMP-PKA signaling pathway. In an
experimental model of high-fat feeding, hypothalamic glucagon resistance disrupts the control on
glucose production. However, direct activation of MBH PKA bypasses this resistance to lower
glucose production79. Since MBH glucagon resistance lies upstream of PKA in response to a
high-fat diet, the potential downstream targets of PKA in MBH glucagon action warrants
investigation.
The activation of cAMP-PKA pathway has been documented to inhibit AMPK in
hypothalamic cell lines88 and adipocytes87. These findings are of interest as direct inhibition of
MBH AMPK is sufficient to lower glucose production102, while activating MBH AMPK negates
glucose sensing to inhibit glucose production102. It is believed that activation of MBH AMPK
negates the ability of glucose flux to increase malonyl-CoA levels and relieves the inhibition on
CPT-1, leading to a reduction of cytosolic LCFA-CoA levels101,118. An accumulation of MBH
LCFA-CoA levels is necessary to activate MBH PKC-δ105 and the KATP channels94,105 to lower
glucose production. Given that MBH PKA signaling is necessary for glucagon to inhibit glucose
production79 and that PKA inhibits AMPK in vitro as discussed above87,88, we here tested the
hypothesis that MBH lipid-sensing pathway involving AMPK -> LCFA-CoA -> PKC-δ and the
subsequent activation of the KATP channels are necessary for MBH glucagon to lower glucose
production (Figure 2.1).
With molecular and chemical approaches, we inhibited the (i) MBH lipid-sensing
pathway via activation of MBH AMPK or inhibition of MBH PKC-δ, and (ii) MBH KATP
31
channels in the presence of MBH glucagon stimulation and evaluated the changes in the rate of
glucose production and glucose uptake in normal rats.
32
2.3 Materials and Methods 2.3.1 Animal Preparation
Adult male Sprague Dawley rats aged 8 weeks (260-280g) from Charles River Laboratories
(Montreal, Quebec, Canada) were studied. Rats underwent stereotaxic implantation with a 26-
gauge stainless steel bilateral guide catheter (C235G, Plastics One Inc. Virgina, USA) placed
into the MBH using the coordinates 3.1 mm posterior to bregma, 0.4 mm lateral of midline and
9.6 mm below skull surface as described79. After six days of recovery, vascular catheters were
inserted into the internal jugular vein and carotid artery for infusion and blood sampling79,122. All
experiments in rats complied with the rules of the Institutional Animal Care and Use committee
of the University Health Network.
2.3.2 Adenovirus injection
Immediately following brain surgery, a group of rats received 3µl of adenovirus containing the
constitutively active (CA) form of AMPK (Ad-CA AMPK α1312 [T172D] (3.83x1010pfu/ml)102; or
the dominant negative (DN) form of PKC-δ or LacZ (4x108pfu/ml; gift from Dr. J Soh,
Biomedical Research Centre for Signal Transduction, Incheon, Korea)159 ; or the DN Kir6.2
AAA (3.1x1010pfu/ml) or green fluorescence protein (GFP) (3.0x1010 pfu/ml)105 through each
side of the MBH catheters, as described102,105,109.
2.3.3 Pancreatic-euglycemic clamp
Four days following vascular catheterization, conscious and unrestrained rats with at least 90%
recovery in their food intake and body weight were used in clamp studies. All rats were limited
to 15g of food the night before the clamp to ensure comparable nutritional status79,122. At the start
of the experiment (t=0), a primed continuous infusion of [3-3H]-glucose (Perkin-Elmer; 40µCi
33
bolus, 0.4µCi min-1) was initiated and maintained until the end of the experiment (t=210) to
assess glucose kinetics using tracer dilution methodology. From t=90-210, a pancreatic basal
insulin clamp was performed during which somatostatin (3µg kg-1 body weight min-1) and
insulin (1.1mU kg-1 body weight min-1) were continuously infused to replace insulin to basal
levels, along with a variable infusion of 25% glucose to maintain euglycemia. At 10-min
intervals, plasma samples were taken for determination of [3-3H]-glucose concentrations, as well
as plasma insulin and glucagon concentrations. At the end of the experiment, rats were
anesthetized and injected with 3ul bromophenol blue through each side of the MBH catheter to
verify the correct placement of the catheter. The MBH wedges were then collected, frozen in
liquid nitrogen and stored at −80°C for subsequent analysis.
Treatments administered into the MBH at a rate of 0.006µl/min included: Saline (t=90-210);
5pg/µl glucagon (t=90-210); 25mmol/l AMPK activator AICAR (t=0-90) with 25mmol/l
AICAR+saline (t=90-210) or 25mmol/l AICAR+5pg/µl glucagon (t=90-210); 60µmol/l PKC-δ
inhibitor rottlerin (t=0-90) with 60µmol/l rottlerin+saline or 60µmol/l rottlerin+5pg/µl glucagon
(t=90-210); 100µmol/l KATP channel inhibitor glibenclamide (t=0-90) with 100µmol/l
glibenclamide+saline or 100µmol/l glibenclamide+5pg/µl glucagon (t=90-210 min).
2.3.4 Western blotting for phosphorylated ACC
MBH wedges were homogenized in a lysis buffer constituting 50mM Tris-HCl (pH 7.5), 1mM
EGTA, 1mM EDTA, 1% (w/v) Nonidet P40, 1mM sodium orthovanadate, 50mM sodium
fluoride, 5mM sodium pyrophosphate, 0.27M sucrose, 1µM Dithiotritolo (DTT), and protease
inhibitor cocktail (Roche). The Pierce 660nm protein assay (Thermo Scientific) was used to
measure protein concentrations of the homogenized tissues. Protein lysates (20ug) were
subjected to SDS-PAGE and transferred onto nitrocellulose membranes (Amersham). The
34
membranes were first incubated for 1h at room temperature with blocking solution (5% BSA in
Tris-buffered saline and 0.2% Tween-20) and then overnight at 4°C with the primary antibody
(indicated below) diluted 1000-fold. Protein expression was detected using a horseradish
peroxidase (HRP)-linked rabbit-specific secondary antibody (diluted 1/4,000 in blocking
solution) and an enhanced chemoluminescence commercial kit (Pierce). The phosphorylation of
ACC was quantified by densitometry with the Quantity One software and normalized for the
total protein (ACC or β-tubulin). Primary antibodies include: anti-phospho ACC, anti-total ACC,
anti-β-tubulin (Cell signaling Technology).
2.3.5 PKC-δ activity assay
PKC-δ was immunoprecipitated from MBH wedges obtained after performing 10-minute MBH
infusions of saline, glucagon or OAG in rats. To do this, 3 MBH wedges from each treatment
group were pooled together to yield an n =1. MBH tissues were homogenized as described
above, following which 500ug of tissue lysate was incubated overnight with 8µg of PKC-δ
polyclonal antibody (sc-213; Santa Cruz Biotechnology) on a rotating wheel and then incubated
with 25µl of protein A/G sepharose beads for 2h at 4°C. The beads were then centrifuged at
8000rpm for 1min. After removal of the supernatant, the beads were then washed (2x with 1ml
lysis buffer (as above) with 0.5M NaCl, 1x with 1ml lysis buffer (as above) with 0.15M NaCl
and 2x with 1ml buffer A containing 50mM Tris-HCl pH 7.5, 0.1mM EGTA and 1µM DTT).
With the supernatant removed, 20µl of buffer A was further added giving a final sample volume
of 25µl. We then proceeded with the Biotrak protein Kinase C (PKC) enzyme assay system (GE
Healthcare). Additionally, to normalize for the amount of PKC-δ immunoprecipitated in each
sample, PKC-δ protein was separated from the beads using Laemmli sample buffer and subjected
35
to SDS-PAGE and quantified (as described above). Results were then presented as
pmol/min/protein.
2.3.6 Biochemical Analysis
Plasma glucose levels were determined by the glucose oxidase method (Glucose analyzer GM9;
Analox Instruments, Lunenberg, MA). Radioimmunoassays (Linco Research, St. Charles, MO)
were used to determine plasma insulin and glucagon concentrations.
2.3.7 Statistical Analysis
Unpaired Student’s t tests and ANOVA as appropriate. Significance was accepted as P<0.05.
2.4 Results 2.4.1 Role of MBH AMPK in glucagon action
We activated MBH AMPK in the presence of MBH glucagon by infusing the AMPK activator
AICAR and examined whether it would negate the ability of MBH glucagon to lower glucose
production. This AICAR dose negated the ability of hypothalamic glucose infusion to lower
glucose production102. Firstly, consistent with previous finding79, MBH glucagon infusion led to
a significant increase in the glucose infusion rate (Figure 2.2A) to achieve euglycemia during
the pancreatic basal insulin clamp and significantly lowered glucose production (Figures 2.2 B,
C) compared to MBH saline treatment independent of changes in glucose uptake and plasma
levels of insulin, glucagon and glucose (Figure 2.2 D, Table 2.1). Interestingly, glucagon co-
infused with AICAR into the MBH still led to a significant increase in glucose infusion rate and
decrease in glucose production. Notably, rats that received MBH AICAR infusion alone showed
no significant difference in the glucose infusion rate and glucose production compared with the
MBH saline group, thereby showing that AICAR per se has minimal effect on glucose kinetics in
36
these experimental conditions. These data indicate that chemical activation of hypothalamic
AMPK does not reverse the ability of MBH glucagon to lower glucose production.
Alternatively, we activated hypothalamic AMPK by injecting an adenovirus expressing
the CA form of AMPK, which has been previously shown to negate the metabolic effects of
hypothalamic nutrient-sensing mechanisms102. We examined whether MBH glucagon inhibition
on glucose production is then blocked in rats expressing MBH CA AMPK. Firstly, infusion of
MBH glucagon with prior Ad-GFP (control) injection significantly increased the glucose
infusion rate required to maintain euglycemia compared to MBH saline infusions (Figure 2.2E),
owing to a reduction in glucose production (Figures 2.2F, G) as opposed to a difference in
glucose uptake (Figure 2.2H). Similarly, in rats injected with MBH CA AMPK, MBH glucagon
also led to an increase in the glucose infusion rate and suppression of glucose production, while
glucose uptake was similar in all groups (Figures 2.2E-H). Similar to our chemical approach
data, these results show that molecular activation of hypothalamic AMPK does not negate the
ability of MBH glucagon to lower glucose production.
As added confirmation, we evaluated AMPK activity (i.e., protein content ratio of pACC
/ total ACC) in MBH wedges obtained from rats that received MBH saline or glucagon
treatments during clamps. Acetyl CoA carboxylase (ACC) is phosphorylated by AMPK; thus, a
lower ratio of phospho (P)-ACC/ total ACC is indicative of a lower degree of AMPK activation.
MBH glucagon infusion showed a similar degree of AMPK activation (0.6±0.1) as with MBH
saline infusion (0.5±0.1) (Figure 2.2I) (in contrast to our original hypothesis where MBH
glucagon action was postulated to inhibit AMPK). These results suggest that the MBH glucagon
infusion does not alter MBH AMPK activity in these experimental conditions and hence changes
in MBH AMPK activity is not necessary for glucagon to lower glucose production.
37
2.4.2 Role of MBH PKC-δ in glucagon action
We next investigated whether inhibition of MBH PKC-δ will negate the glucose production-
regulating effect of MBH glucagon. MBH PKC-δ was inhibited by infusing PKC-δ inhibitor
rottlerin directly into the MBH at a dose previously shown to block MBH lipid or PKC activator
(OAG) infusion to lower glucose production and MBH OAG to activate PKC-δ105. We found
that MBH rottlerin did not attenuate the ability of MBH glucagon to increase the exogenous
glucose infusion rate (Figure 2.3A) and lower glucose production (Figures 2.3B, C) during the
clamp, while glucose uptake remained unchanged (Figure 2.3D). Alternatively, MBH glucagon
was also equally potent to increase the glucose infusion rate (Figure 2.3E) and lower glucose
production (Figures 2.3 F, G) in rats injected with MBH DN PKC-δ as compared with LacZ
injected rats, while glucose uptake remained similar between groups (Figure 2.3H). Of note,
injection of DN PKC-δ has been previously shown to reduce PKC-δ activity and negate the
metabolic effects of OAG in the duodenum159 . Collectively, these data suggest that both
chemical and molecular inhibition of hypothalamic PKC-δ do not block the ability of MBH
glucagon to lower glucose production. Consistent with these observations, no significant
difference in MBH PKC-δ activity between glucagon- and saline- treated MBH tissues was
detected, whereas MBH OAG (positive control) infusion markedly stimulated PKC-δ activity in
the same experimental conditions (Figure 2.3I).
2.4.3 Role of MBH KATP channels in glucagon action
We inhibited MBH KATP channels using glibenclamide in the presence of MBH glucagon to
examine whether the glucose production-lowering effect of MBH glucagon is abrogated. MBH
glibenclamide has been reported to negate MBH lipid105, OAG105 but insulin18 as well to lower
glucose production. Interestingly, co-infusion with glibenclamide attenuated the ability of MBH
38
glucagon to increase the glucose infusion rate (Figure 2.4A) and lower glucose production
(Figures 2.4B, C), without any changes in glucose uptake (Figure 2.4D).
The Kir6.2-SUR1 KATP channels (which are blocked by glibenclamide) are expressed in
both the pancreatic β-cells and neurons18. The adenovirus expressing the DN form of Kir6.2
expresses an AAA mutant subunit of Kir6.2 that co-assembles with endogenous Kir6.2 and
prevents the KATP channels from conducting potassium current160. We here directly injected the
adenovirus DN Kir6.2 AAA into the MBH as described105 and tested the metabolic effect of
MBH glucagon. Consistent with our chemical approach data, MBH glucagon also failed to
increase the exogenous glucose infusion rate (Figure 2.4E) and lower glucose production
(Figures 2.4F, G) in rats expressing MBH DN Kir 6.2 AAA unlike GFP (control virus)-injected
rats, independent of changes in glucose uptake (Figure 2.4H). Together, these chemical and
molecular loss-of-function experiments indicate that MBH KATP channel is necessary for MBH
glucagon to lower glucose production.
2.5 Discussion We presently demonstrate that activation of MBH AMPK or inhibition of MBH PKC-δ did not
negate the glucose production-lowering effect of MBH glucagon. Activation of MBH AMPK
increases protein ratio of pACC / total ACC, inhibits ACC and prevents the formation of
malonyl-CoA (endogenous inhibitor of CPT-1) and LCFA-CoA54,101, while inhibiting MBH
PKC-δ negates LCFA-CoA to lower glucose production94. Since blocking the beginning or the
end of this malonyl-CoA -> CPT-1 -> LCFA-CoA -> PKC-δ, lipid sensing, pathway did not alter
the gluco-regulatory effect of MBH glucagon, MBH glucagon is demonstrated to signal through
a lipid-sensing independent pathway to lower glucose production.
39
Like glucagon, leptin and insulin action in the MBH lowers glucose production18,46,161.
However, in contrast to the inability of MBH glucagon infusion to alter MBH pACC / total ACC
(i.e., AMPK activity) as currently reported, hypothalamic leptin and insulin administration lower
MBH AMPK activity, the protein ratio of pACC / total ACC and activate ACC activity53,54.
These findings raise the possibility that unlike glucagon, leptin and insulin in the MBH signal
through a lipid-sensing dependent pathway to lower glucose production. This working
hypothesis, however, remains to be investigated.
Interestingly, activation of the MBH KATP channels is necessary for MBH glucagon to
lower glucose production. Hypothalamic KATP channels, thus, become the common integrator of
hormonal and nutrient sensing to regulate glucose production as activation of hypothalamic KATP
channels is sufficient18 and necessary for insulin18, GLP-1162 and lipids94 as well to lower glucose
production in rodents. Of note, activating KATP channels in the whole brain of humans163 or the
dorsal vagal complex in rodents122 lowers glucose production, highlighting the gluco-regulatory
role of the KATP channels is not limited to the ones that are expressed in the hypothalamus.
Given that MBH glucagon receptor-PKA signaling79 and the activation of the MBH KATP
channels are required for glucagon to lower glucose production, and that PKA directly
phosphorylates and activates the Kir6.2/SUR1 subunits of the KATP channels164, our data
collectively indicate that glucagon action in the hypothalamus signals via a lipid-sensing
independent but KATP channel dependent pathway to regulate glucose production in vivo.
40
Figure 2. 1 Schematic representation of the working hypothesis.
MBH glucagon may signal through the MBH lipid sensing- KATP channel pathway to lower
glucose production.
41
Figure 2. 2 Role of MBH AMPK in glucagon action.
(A) Glucose infusion rate, (B) glucose production, (C) glucose production suppression expressed
as the percentage decrease from basal period (60-90 min) to the clamp period (180-210 min) and
(D) glucose uptake obtained during the clamps that received MBH saline (n=5), glucagon (n=5),
AICAR+saline (n=5) or AICAR+glucagon (n=7). (E) Glucose infusion rate, (F) glucose
production, (G) glucose production suppression expressed as the percentage decrease from basal
period (60-90 min) to the clamp period (180-210 min) and (H) glucose uptake obtained during
the clamps that received MBH GFP+saline (n=5), GFP+glucagon (n=5), CA AMPK+saline
(n=5) or CA AMPK+glucagon (n=6). (I): Phosphorylation of ACC. Shown above is the
representative western blot of pACC in saline (n=5) and glucagon (n=5) treated MBH wedges
normalized to total ACC and B-tubulin. Shown below is the quantification of pACC normalized
to total ACC. Data are shown as means+SE. *P<0.05.
42
Figure 2. 3 Role of MBH PKC- δ in glucagon action.
(A) Glucose infusion rate, (B) glucose production, (C) glucose production suppression expressed
as the percentage decrease from basal period (60-90 min) to the clamp period (180-210 min) and
(D) glucose uptake obtained during the clamps that received MBH Saline (n=5); Glucagon
(n=5); Rot+saline (n=5); Rot+glucagon (n=5). Rot = Rottlerin. (E) Glucose infusion rate, (F)
glucose production, (G) glucose production suppression expressed as the percentage decrease
from basal period (60-90 min) to the clamp period (180-210 min) and (H) glucose uptake
obtained during the clamps that received MBH LacZ+saline (n=5); Lacz+glucagon (n=5); DN
PKC-δ+saline (n=5); DN PKC-δ+glucagon (n=5). I: PKC-δ activity in MBH wedges. Shown is a
representative quantification from three samples in each treatment group. Data are shown as
means+SE. *P<0.05.
43
Figure 2. 4 Role of MBH KATP channels in glucagon action.
(A) Glucose infusion rate, (B) glucose production, (C) glucose production suppression expressed
as the percentage decrease from basal period (60-90 min) to the clamp period (180-210 min) and
(D) glucose uptake obtained during the clamps that received Saline (n=5); Glucagon (n=5);
Gli+saline (n=5); Gli+glucagon (n=4). Gli = Glibenclamide. (E) Glucose infusion rate, (F)
glucose production, (G) glucose production suppression expressed as the percentage decrease
from basal period (60-90 min) to the clamp period (180-210 min) and (H) glucose uptake
obtained during the clamps that received MBH GFP+saline (n=5); GFP+glucagon (n=5); DN
Kir6.2+saline (n=5); DN Kir6.2+glucagon (n=5). Values are shown as means+SE. *P<0.05
44
Table 2. 1 Plasma insulin, glucagon and glucose concentrations during basal and clamp
conditions. Data are means ±SEM.
45
Chapter 3 Study 2
Inhibition of Glycine Transporter-1 in the
Dorsal Vagal Complex improves Metabolic
Homeostasis in Diabetes and Obesity
Modified from:
Abraham, M.A.*, Yue, J.T.*, Bauer, P.V., LaPierre, M.P., Wang, P., Duca, F.A., Filippi, B.M.,
Chan, O., and Lam, T.K. (2016). Inhibition of glycine transporter-1 in the dorsal vagal complex
improves metabolic homeostasis in diabetes and obesity. Nat Commun 7, 13501. *Equal
contribution
Permission to reproduce portions of the above manuscript has been obtained from the copyright
owner: Elsevier Limited
46
3.1 Abstract Background and Aims: The DVC located in the brainstem is a crucial site in sensing nutrients
and hormones to regulate peripheral glucose and energy homeostasis. Direct administration of
amino acid glycine into the DVC is sufficient to lower glucose production in healthy rats via
activation of NMDA receptors in the DVC. Further, NMDA receptors in the DVC can also relay
gut-hormonal dependent signals to regulate feeding. What remains to be shown is whether DVC
glycine- NMDA receptor axis regulates glucose production in diabetic and obese rodents, and
whether activating NMDA receptors with direct glycine infusion could potentially also regulate
feeding. However, given that glycine has poor pharmacokinetics in vivo, we here sought to
examine whether selectively increasing extracellular glycine levels in the DVC by inhibiting
DVC GlyT1 has a therapeutic potential in diabetes and obesity. Methods: We administered a
GlyT1 inhibitor and/or a molecular GlyT1 knockdown in the DVC of healthy, diabetic and obese
male Sprague-Dawley rats and evaluated changes in glucose regulation during ivGTT and basal-
insulin pancreatic clamps. Loss-of-function approaches targeting the NMDA receptors as well as
the vagus nerve were also utilized to delineate the mechanistic pathway relaying DVC GlyT1
inhibition and glucose regulation. Finally, to determine the effect of DVC GlyT1 inhibition on
food intake and body weight, we also performed a non-clamp fasting-refeeding protocol in
healthy rats. Results: We found that administration of a glycine transporter 1 (GlyT1) inhibitor,
or molecular GlyT1 knockdown, in the DVC suppresses glucose production, increases glucose
tolerance, and reduces food intake and body weight gain in healthy, obese, and diabetic rats.
Conclusions: These findings provide proof of concept that GlyT1 inhibition in the brain
improves glucose and energy homeostasis. We propose that GlyT1 inhibitors have potential as a
treatment of both obesity and diabetes.
47
3.2 Introduction Obesity and diabetes have become a worldwide epidemic. Over 2.1 billion people worldwide are
overweight or obese8 and approximately 422 million are afflicted with diabetes165. Given the
combined economic burden of treating both these diseases and their complications, the
development of safe and effective therapeutic strategies is decidedly crucial. The dysregulation
of glucose and energy homeostasis in diabetes and obesity are caused in part by the aberrant
elevation of hepatic glucose production and energy intake166,167, pathologies which arise from the
collective failure of multiple homeostatic systems involving the liver, pancreas, adipose tissue,
brain, and gastrointestinal tract167-171. As such, the development of pharmacological approaches
to restore the impaired mechanisms within these systems is crucial to restore metabolic
homeostasis in diabetes and obesity. Neural circuits of the central nervous system (CNS) emerge
as a potential target for clinical intervention. The recently FDA-approved anti-obesity drug,
lorcaserin, activates hypothalamic 5-HT2C receptors to reduce food intake172. Moreover, the anti-
diabetic glucagon-like-peptide 1 receptor agonist drug, liraglutide, which is clinically
demonstrated to improve glycemia and reduce body weight173, requires activation of neuronal
glucagon-like-peptide 1 receptor to exert its anorectic effects174. The central actions of other
hormones such as insulin further demonstrate the potential of CNS-based therapies, as intranasal
insulin delivery in humans reduces food intake175 and glucose production176 and improves
whole-body insulin sensitivity177.
CNS nutrient sensing mechanisms also reduce food intake and body weight178 and lower
glucose levels in healthy107 and diabetic110 rodents. Specifically, hypothalamic nutrient sensing
activates a forebrain-hindbrain neuronal axis involving N-methyl-ᴅ-aspartate (NMDA) receptors
in the dorsal vagal complex (DVC) to suppress glucose production121, while these same DVC
NMDA receptors are required for intestinal sensing of nutrients, as well as metformin and
48
resveratrol, to lower glucose production and food intake5,116,119,179,180. Furthermore, directly
targeting the DVC has metabolic benefits, as direct administration of glycine, an obligatory co-
agonist of the NMDA receptor, into the DVC of healthy rats lowers glucose production via
NMDA receptor activation138. Therefore, manipulating glycine levels in the DVC could present a
therapeutic target for the treatment of obesity and diabetes. However, administration of glycine
per se is not suitable as a therapy due to its poor pharmacokinetics in vivo.
On the other hand, regulating glycine concentration by manipulating glycine transporters
(GlyT) has demonstrated clinical feasibility. Since glycine uptake into cells is regulated by
glycine transporters, of which GlyT1 is the primary regulator of glycine levels in the vicinity of
NMDA receptors181, GlyT1 inhibition increases extracellular glycine levels to potentiate the
activation of NMDA receptors142. Modulation of NMDA receptor neurotransmission is currently
used as a therapy for schizophrenia, a disease that displays reduced NMDA receptor function. In
fact, clinical trials have shown that NMDA receptor augmentation via GlyT1 inhibitors improve
symptoms of schizophrenia182,183. However, no studies to date have investigated the therapeutic
potential of GlyT1 inhibition for the treatment of diabetes and obesity. Here, we examined
whether GlyT1 inhibition regulates glucose and energy homeostasis in healthy, obese, and
diabetic rodents (Figure 3.1). We demonstrate that direct inhibition of GlyT1 in the DVC
confers metabolic benefits including improved glucose tolerance, lowered glucose production,
reduced feeding, and lowered body weight gain in diabetic and obese rodents. We also report
that systemic infusion of GlyT1 inhibitor recapitulates the metabolic effects of DVC GlyT1
inhibition. Thus, inhibiting GlyT1 in the brain represents a potential novel therapeutic strategy to
lower plasma glucose levels and body weight in diabetes and obesity.
49
3.3 Materials and Methods 3.3.1 Animal preparation and surgical procedures
Male Sprague-Dawley rats (Charles River Laboratories, Saint-Constant, QC, Canada) weighing
280–300 g (9-week old) were used. For chronic 28-d feeding studies (see below), a separate set
of rats initially weighing 200-220 g and fed with regular chow or a high-fat diet were used. Rats
individually housed, were subjected to a standard light–dark (0700 light, 1900 dark) cycle, and
had ad libitum access to drinking water and standard regular chow or a 10% lard-enriched chow
(high-fat diet, HFD) where indicated (see below). Rats were anesthetized during surgeries
(ketamine, 60 mg/kg; xylazine, 8 mg/kg). Bilateral, 26-gauge, stainless steel guide cannulae
(Plastics One Inc, Roanoke, VA, USA) were stereotaxically implanted into the DVC via
coordinates targeting the nucleus of the solitary tract within the DVC (NTS, 0 mm on the
occipital crest, 0.4 mm lateral to the midline, 7.9 mm below the cranial surface; Supplementary
Figure 3.1)139. Eight days following DVC surgery, indwelling catheters were surgically
implanted in the left carotid artery and right jugular vein for blood sampling and infusions,
respectively184. Post-surgical body weight and food intake were monitored daily. Rats attained a
minimum of 90% of their pre-vascular surgery body weight before undergoing experimentation 5
days following vascular surgery. Rats that did not fully recover were excluded from the study.
Rats were randomly allocated into groups prior to experiments but no blinding was done.
In parallel, microdialysis studies were performed on male Sprague-Dawley rats (Charles
River, Raleigh, NC) which started with a body weight of ~280-300g and were individually
housed in the Yale Animal Resources Center in temperature (22-23oC) and humidity controlled
rooms. The animals had free access to rat chow (Harlan Teklad, Indianapolis, IN, USA) and
water. Upon arrival at the Yale Animal Resources Center, the animals were acclimatized to
50
handling and a 12-hour light cycle (lights on between 0700h and 1900h) for one week before
experimental manipulation. Principles of laboratory animal care were followed, and experimental
protocols were approved by the Institutional Animal Care & Use Committee at Yale University.
The rats were anesthetized with isoflurane and the heads were positioned into a stereotaxic frame
(David Kopf Instruments, Tujunga, CA). A single stainless steel guide cannula for
microinjection and microdialysis (Eicom Corporation, Japan) was implanted intracranially using
the following stereotaxic co-ordinates from Paxinos and Watson (0mm on the occipital crest,
5mm medial-lateral, and 7.4mm ventral at an angle of 35° for microdialysis). This targeted the
1mm microdialysis probe (Eicom Corporation, Japan) to the DVC within the NTS.
Male C57BL/6 mice (Jackson Laboratory, Bar Harbor, ME) at 18 weeks of age were
housed in a standard light-dark cycle with ad libitium access to drinking water and standard
chow. Mice were anesthetized during stereotaxic and vascular surgeries (Avertin, 0.6 mg/g). A
unilateral, 33-gauge, stainless steel guide cannula (Plastics One Inc) was stereotaxically
implanted into the 4th ventricle (ICV-4; 6.0 mm posterior to Bregma, 4.0 mm below the cranial
surface)79,122. One week following ICV-4 surgery, an indwelling catheter was surgically
implanted in the right jugular vein79,122. Post-surgical body weight and food intake were
monitored daily. Mice attained a minimum of 90% of their pre-vascular surgery body weight
before undergoing pancreatic clamp 3-5 days following vascular surgery.
3.3.2 Intravenous glucose tolerance test
Experiments were performed in overnight- (16 to 18 h) fasted male Sprague-Dawley rats 5 days
after vascular catheterization. Basal blood samples were obtained in conscious, unrestrained rats
immediately before the start of DVC infusions (0.33 µl/h, CMA 400 syringe pump, CMA
Microdialysis, Inc., North Chelmsford, MA) of 0.9% saline or ALX (ALX 5407, Tocris
51
Bioscience, 40 nM), which were commenced at t = -240 min and maintained until the end of the
experiment at t = 60 min to ensure that rats received the same duration of DVC treatment as
clamp experiments (see below). After t = 0 min blood samples were obtained, an intravenous
bolus of glucose (20% glucose, 0.25 g/kg) was injected and flushed with saline. Injections were
administered via the jugular vein catheter, and blood was sampled from the carotid artery
catheter to measure plasma glucose and insulin levels for 60 min following glucose injection as
described79.
3.3.3 DVC treatments
(i) 0.9% saline
(ii) MK801 (NMDA receptor antagonist, 0.06 ng/min, dissolved in saline)
(iii) 7CKNA (7-chlorokynurenic acid, non-competitive antagonist of the glycine binding
site of the NMDA receptor, 30 µM, dissolved in saline),
(iv) ALX (ALX 5407, selective inhibitor of GlyT1, Tocris Bioscience, 40 nM, dissolved
in saline)
(v) glycine (10 µM, dissolved in saline).
Using the same DVC infusion protocol as the current study, glycine at 10 µM was validated to
elevate DVC glycine levels by ~1.2 fold and lower glucose production138 and secretion of
triglyceride rich very low density lipoproteins (VLDL-TG)139 in regular chow-fed healthy rats,
while MK-801 at 0.06 ng/min and 30 µM of 7CKNA blocked the effects of DVC glycine
infusion to lower glucose production138, and VLDL-TG secretion139. Thus, we have chosen to
use these same dose and concentrations for the inhibitors in this study examining the effect of
DVC GlyT1 inhibition (likely mediating glycine sensing). In fact, a total amount of ~20 ng of
MK-801 was delivered into the DVC over the course of 330 minutes in the current studies,
52
which is comparable to the 50 ng of MK-801 delivered into the NTS (or DVC) that regulated
feeding behavior185. Similarly, our concentration of 30 µM 7CKNA was also well within the
range reported by other studies that have indicated that 10-50 uM of 7CKNA reduces 80-90% of
glycine binding to rat cerebral cortex synaptic plasma membrane186, while 30 uM of 7CKNA
inhibits NMDA-induced transmitter release from rat hippocampal slices187. More importantly,
DVC infusion of neither MK-801 at 0.06 ng/min nor 30 µM 7CKNA per se resulted in increased
glucose production but only blocked the effect of DVC glycine infusion to lower glucose
production138. Thus, any concern for the non-specific effects of these inhibitors in regulating
glucose homeostasis can be safely excluded. The concentration of 40 nM ALX was chosen based
on the IC50 of ALX for GlyT1 (4 nM)188 and factoring into a dilution factor when chemical
inhibitors are infused into the DVC.
3.3.4 Pancreatic basal insulin euglycemic clamp in rats
Experiments were performed in male Sprague-Dawley rats fasted for ~4-6 hrs before clamp
experiments to ensure comparable post-absorptive nutritional status. Basal blood samples were
obtained in conscious, unrestrained rats immediately before the start of DVC infusions (0.33
µl/h) of the following infusates: (i) saline, (ii) MK801, (iii) 7CKNA, (iii) ALX, (iv) ALX +
MK801, (v) ALX + 7CKNA, (vi) glycine. Infusions of MK801 or 7CKNA, when used, or saline
as a control, were commenced at t = -90 min; infusions of ALX ± MK801 or 7CKNA were
commenced at t = -60 min, and infusions of glycine ± 7CKNA were commenced at t = 0 min and
maintained for the duration of the experiment (Supplementary Figure 3.1 A). ALX, an
inhibitor of the GlyT1 transporter, inhibits the binding of glycine to its cellular transporter and
elevates extracellular levels of glycine189. ALX infusion was initiated earlier to allow for
extracellular levels of glycine in the DVC to accumulate. Clamp methodology was performed as
53
follows79. A primed, continuous infusion (PHD2000 syringe pump, Harvard Apparatus, Saint
Laurent, QC) of [3-3H]-glucose (Perkin Elmer; 40 µCi bolus + 0.4 µCi infusion) was
commenced at t = 0 min and maintained until the end of the clamp experiment at t = 240 min to
measure glucose kinetics using tracer-dilution methodology. Glucose turnover was calculated
using steady-state formulae, in which the rate of appearance of glucose of glucose is calculated
using [3-3H]-glucose. The total rate of appearance of endogenous glucose production is
equivalent to the rate of glucose utilization during the basal period (t = 60-90 min). The
pancreatic basal insulin-euglycemic clamp was initiated at t = 90 min with the primed continuous
infusion of insulin (1.2 mU/kg/min, somatostatin (SST, 3 µg/kg/min), and a variable infusion of
25% glucose to maintain glycemia at a similar level to the basal period and was maintained until
t = 240 min. Plasma samples were obtained every 10 min for determination of [3-3H]-glucose
specific activity and glucose levels. Wedges containing the DVC, the left and right portions of
spinal trigeminal tr. (sp5), Spinal 5nu caudal part (Sp5C), Spinal 5nu, interpolar (Sp5I), and the
pyramidal tr. (py) (Supplementary Figure 3.3:i-v, see Brain tissue sampling section below)
were collected immediately after the experiments, frozen in liquid nitrogen, and stored at -80°C
for analysis.
3.3.5 Hepatic branch vagotomy in rats
A separate set of male Sprague-Dawley rats underwent hepatic branch vagotomy138 on the same
day as vascular catheterization surgeries. The hepatic branch of the ventral subdiaphragmatic
vagal trunk was transected, and the omentum between the liver and the esophagus was severed
such that any tissue connections between the liver and the esophagus were removed. The neural
communication between the central nervous system and the liver was disrupted upon transection
of the hepatic vagal nerve, and to a much lesser degree, the innervations to the gut were also
54
disrupted. Sham-operated rats underwent similar procedure except for transection of the vagus.
After surgical recovery, rats underwent clamp experiments as described above.
3.3.6 Microdialysis
One week after surgery, the male Sprague-Dawley rats were fasted for 4-6 hrs before
experiments to ensure comparable post-absorptive nutritional status. On the day of the study, the
microdialysis-microinjection probe was inserted through the guide cannula and the animals were
allowed to recover for 2.5hrs prior to collection of the baseline sample (Supplementary Figure
3.2B). Artificial extracellular fluid was perfused through the probe at a rate of 0.5ul/min
throughout the study. Following the recovery period, we collected a baseline sample over the
course of 2 hours prior to the start of ALX (40nM in saline, Tocris Bioscience) infusion. The
ALX was infused into the DVC (via the microinjection needle) at a rate of 0.33ul/h for a total
duration of 300 minutes. Microdialysate samples were collected at 60, 180 and 300 minutes
following the start of ALX infusion. Control animals were infused with saline and sampled under
similar conditions. At the end of the study, the animals were euthanized with an overdose of
sodium pentobarbital.
3.3.7 DVC virus injection
Immediately after stereotaxic surgery while anesthetized, 3 µl of adenovirus or lentivirus were
injected over 30 s in each of the DVC cannulae with microsyringes. An adenovirus expressing
shRNA to the GluN1 subunit of the NMDA receptor (Ad- GluN1 shRNA, 4.0 x 1011 pfu/ml), or
a mismatch sequence as a control (Ad-MM, 4.0 x 1011 pfu/ml), was injected into the DVC for
one set of experiments using the same protocol that we have validated138. This adenoviral GluN1
shRNA knockdown procedure decreases GluN1 protein levels specifically in the region DVC138.
55
In separate sets of experiments, a lentivirus expressing shRNA to GlyT1 (LV-GlyT1 shRNA, 1.0
x 106 infectious units) (sc-270432-V, Santa Cruz Biotechnology, Inc., Dallas, TX), or a mismatch
sequence as a control (LV-MM, 1.0 x 106 infectious units) (sc-108080, Santa Cruz
Biotechnology), was injected in the DVC. Eight days after DVC cannulation and virus injection,
vascular catheterization was performed as described above in rats that would undergo clamp
experiments. Thirteen days after DVC cannulation, virus-injected rats underwent clamp or
feeding experiments as described above.
3.3.8 Brain tissue sampling in rats
At the end of the experiments, rats were injected with 3 µl bromophenol blue through each side
of the bilateral DVC catheter to verify the correct placement of the catheter. Once the whole
brain is harvested from the anesthetized rat via decapitation, the cerebelleum is lifted to expose
the caudal part of the brain (Supplementary Figure 3.3:i-v). Only those data for rats that
showed injection of dye within the vagal triangle (Supplementary Figure 3.3: ii-v) were
included. A spatula was used to extract the section of vagal triangle overlaying the DVC (Blue;
Supplementary Figure 3.3:ii-v). Additionally, sections of tissues were also dissected out from
the left (Yellow) and right (Purple) lateral regions of the caudal brain containing sp5 (spinal
trigeminal tr.), Sp5C (spinal 5nu, caudal part), Sp5I (spinal 5nu, interpolar), as well from the
lower region containing py (pyramidal tr) of the distal caudal brain (Green; Supplementary
Figure 3.3:ii-v) to validate the specificity of the lentiviral injections into the DVC.
3.3.9 Microdialysate sample glycine analysis
Microdialysate samples collected at baseline, 180 and 300 minutes were analyzed using a
fluorometric assay kit (Glycine assay kit (Fluorometric), #K589-100, Biovision Incorporated,
Milpitas, CA) according to the manufacturer’s directions. A sample volume of 50ul was used in
56
the assay and since glycine content was low, each microdialysate sample was spiked with
0.3nmol of the glycine standard to bring the values within a more reliable reading range of the
assay and the calculations were adjusted accordingly. Since glycine levels were relatively stable
during ALX infusion, the concentrations obtained from the last two microdialysate samples were
averaged together and compared to baseline levels. Similar calculations were performed for the
control animals.
3.3.10 Acute (3-d) and chronic (28-d) high-fat feeding in rats
Two separate sets of male Sprague-Dawley rats were fed a palatable, 10% lard-enriched high-fat
diet (HFD, TestDiet #571R, Purina Mills, Richmond, IN), either for 3-d (acute, 3-d HFD) or 28-
d (chronic, 28-d HFD) prior to clamp experiments. The composition of the HFD (3.9 kcal/g)
differs from regular chow (3.1 kcal/g): fat content (34 vs 18%); protein (22 vs 33%) and
carbohydrate (44 vs 49%) content. A 28-d regular chow-fed cohort of rats was used in parallel to
the 28-d HFD group. Both cohorts of HFD rats underwent the pancreatic clamp experiments as
described above. Rats did not overeat were excluded from the studies.
3.3.11 Intravenous ALX infusion clamps
In a separate group of 3-d HFD fed male Sprague-Dawley rats, the pancreatic clamp experiments
were performed as described above, with the exception that a continuous i.v. ALX (4.1
µg/kg/min, dissolved in 6% DMSO infused at 20 µl/min) or i.v. 6% DMSO (20 µl/min) as
vehicle was initiated at t = -90 min. At t = 0 min, a primed continuous infusion of 3[H3]- glucose
was commenced and maintained until the end of the experiment, t = 210 min. At t = 90 min, the
pancreatic basal insulin clamp was initiated with the primed continuous infusion of insulin (1.2
mU/kg/min), somatostatin (SST, 3 µg/kg/min), and a variable infusion of 25% glucose to achieve
euglycemia was administered until t = 210 min. Intravenous (i.v.) ALX was constantly infused at
57
4.1 ug/kg/min to achieve a total amount of 1.23 mg/kg ALX delivered into the blood in 300 min.
This choice of dose is based on the fact that i.v. ALX injected at 1-2 mg/kg has been documented
to potentiate NMDA-evoked firing in PFC neurons of rats in vivo190.
3.3.12 Induction of experimental type 2 diabetes
Six days after DVC surgery, a separate set of male Sprague-Dawley rats was given an
intraperitoneal injection of nicotinamide (Nic, 170 mg/kg) followed by an intraperitoneal low-
dose injection of streptozotocin (STZ, 65 mg/kg) 15 min later and fed with a HFD for 7 days as
described179,180,191, prior to intravenous glucose tolerance tests as described above. Rats that did
not present with fed hyperglycemia (e.g. > 9 mM) were excluded from the study.
3.3.13 Fasting-refeeding experiments
Separate groups of male Sprague-Dawley rats were subjected to a 22-h fast (food removed at 7
pm) prior to undergoing the refeeding experiment. DVC injections (0.04 µl/min for 5 min using
CMA syringe pumps) of 0.9% saline, ALX (40 nM), or glycine (10 µM) were given at t = -60
min (ALX) or t = -10 min (glycine) (Fig 7a,e). To prevent backflow of the injected volume,
injection cannulae were left in guide cannulae for an additional 5 min with the pump off, and
dummy cannulae are re-inserted and secured with dust caps. Regular chow was returned to cages
at 5 pm, t = 0 min. Food intake was measured every 30 min for the first 4 h of the refeeding
experiment, and every 1 h until t = 360 min. Food intake and body weight were measured again
20 h (day 1) and 44 h (day 2) after rats were refed.
3.3.14 Pancreatic basal insulin euglycemic clamp in mice
Experiments were performed in male C57BL/6 mice fasted for ~4-6 hrs before clamp
experiments to ensure comparable post-absorptive nutritional status. Basal blood samples were
58
obtained in conscious, unrestrained mice immediately before the start of ICV-4 infusions (1.02
µl/h) of saline or ALX (40 nM), which were commenced at t = -90 min and maintained for the
duration of the experiment. Clamp methodology in mice was performed as follows 79. A primed,
continuous infusion of [3-3H]-glucose (1 µCi bolus + 0.1 µCi infusion) was commenced at t = 0
min and maintained until the end of the clamp experiment at t = 180 min to measure glucose
kinetics. The basal period was defined as t = 50-60 min. The pancreatic basal insulin-
euglycemic clamp was initiated at t = 60 min with a primed continuous infusion of insulin (1.4
mU/kg/min, SST (8.3 µg/kg/min), and a variable infusion of 10% glucose to maintain glycemia
at a similar level to the basal period and was maintained until t = 180 min. Plasma samples were
obtained every 10 min for determination of [3-3H]-glucose specific activity and glucose levels.
3.3.15 Western blot analyses
GlyT1 protein levels were measured in purified plasma membrane fractions of brain tissue
wedges from rats that received LV GlyT1 shRNA or LV MM injections. Brain tissue wedges
were collected 13 d following lentivirus injection and immediately frozen in liquid nitrogen and
stored at -80°C until analysis. Purified plasma membrane protein fractions were isolated using a
commercial kit suitable for mammalian tissues (Plasma Membrane Protein Extraction Kit
#K268-50, BioVision Incorporated, Milpitas, CA)184. Purified plasma membrane fraction protein
concentrations were measured using a BCA Protein Assay kit (#K812-1000, BioVision
Incorporated, Milpitas, CA), and 6 µg of protein was subjected to electrophoresis on 8%
polyacrylamide gels and transferred to nitrocellulose membranes. Membranes were incubated
with blocking solution (5% BSA in Tris-buffered saline containing 0.2% Tween-20 (TBS-T)) for
1h at room temperature and overnight at 4°C in primary antibody solutions diluted 1/1000 in 5%
BSA in TBS-T of GlyT1 (ab113823 rabbit, Abcam, Cambridge, MA), or insulin receptor (IR) β
59
(L55B10 mouse, #3020, Cell Signaling Technology, Danvers, MA) after 10 min shaking in
antibody stripping buffer (Gene Bio-Application Ltd, Yavne, Israel) and re-blocked as above.
Protein expression was detected using an HRP-linked secondary antibody (rabbit and mouse,
respectively, diluted 1/4000 in blocking solution) and an enhanced chemoluminescence reagent
(Pierce ECL Western Blotting Substrate, Thermo Scientific, Rockford, IL). Immunoblots were
detected using a MicroChemi 4.2 chemiluminescent imaging system and quantified with
GelQuant image analysis software (DNR Bio-Imaging Systems, Jerusalem, Israel). Plasma
membrane GlyT1 protein levels were normalized to the plasma membrane protein levels of IR.
3.3.16 Biochemical analysis
Plasma glucose concentrations were measured by the glucose oxidase method (Glucose Analyzer
GM9, Analox Instruments, Lunenburg, MA). Plasma insulin levels were determined by
radioimmunoassay (Millipore Canada Ltd, Etobicoke, ON).
3.3.17 Calculations and statistics
The sample size for each group was chosen based on study feasibility and prior knowledge of
statistical power form previously published experiments. For pancreatic clamp experiments in
rats, measurements during t = 60-90 min were averaged for the basal period, and t = 210-240
min, and for the intravenous ALX infusion clamps, t=180-210 min were averaged for the clamp
period. In mice, measurements during t = 50-60 min were averaged for the basal period, and t =
160-180 min were averaged for the clamp period. Integration of the area under the curve (AUC)
was calculated with GraphPad Prism 6 software (LaJolla, CA). Unpaired Student’s t-tests were
performed in the statistical analysis of two groups. Where comparisons were made across more
than two groups, ANOVA was performed, and if significant, was followed by Dunnett’s or
Tukey’s post-hoc tests when appropriate. Measurements that were taken repeatedly over time
60
were compared using repeated measures ANOVA; if the time and treatment interaction between
groups was found to be significant, Sidak’s multiple comparisons test or t-tests were used to
determine the statistical significance at specific time points between groups. Differences in the
overall effects of HFD diet on body weight are indicated where significance was found following
repeated measures ANOVA. P value <0.05 was considered statistically significant.
3.4 Results 3.4.1 Gluco-regulation by DVC GlyT1 inhibition in healthy rodents
To first assess a gluco-regulatory function of GlyT1 inhibition in physiological conditions, we
infused the GlyT1 inhibitor, ALX188,192, into the DVC of conscious, unrestrained healthy rats and
monitored plasma glucose levels during an intravenous glucose tolerance test (ivGTT)
(Supplementary Figure 3.1A). DVC ALX infusion for 5 hrs improved glucose tolerance
(Figure 3.2A) independent of a rise in plasma insulin levels (Figure 3.2B) compared to DVC
saline infusions. To begin delineating the mechanism by which DVC GlyT1 inhibition improves
glucose tolerance independent of changes in insulin, we tested whether DVC GlyT1 inhibition
regulates glucose production or uptake during the pancreatic basal insulin euglycemic clamps in
both rats and mice (Supplementary Figure 3.1A), since DVC glycine infusion potentiates
NMDA receptors to inhibit hepatic glucose production138.
Infusion of ALX into the DVC of rats increases the requirement for exogenous glucose
infusion to maintain euglycemia (Figure 3.2C) and lowers the rate of glucose production
(Figure 3.2D) compared to infusions with saline, independent of differences in glucose uptake,
plasma glucose, plasma insulin or body weight (Supplementary Figure 3.1C-F). We also
performed pancreatic clamps in healthy mice that underwent stereotaxic and vascular
cannulation and demonstrated that ICV-4th ventricle ALX infusion correspondingly increases
61
glucose infusion rates and suppresses glucose production without affecting glucose uptake or
plasma glucose levels (Supplementary Figure 3.2A-D).
Whereas inhibition of NMDA receptors with DVC infusion of NMDA receptor blocker
MK801 alone has no effect on glucose metabolism in rats, the ability of DVC ALX infusion to
increase glucose infusion rates and suppress glucose production is abolished with co-infusion of
MK801 (Figure 3.2C, D), without altering glucose uptake or plasma glucose (Supplementary
Figure 3.1C,D). The classical NMDA receptors, which are comprised of two glycine-binding
GluN1 subunits and two glutamate-binding GluN2 subunits, require co-agonism of their subunit
binding sites for activation137,141,142,193. Since GluN1 is an obligatory subunit193 and full agonism
at the glycine site is necessary for full NMDA receptor activation194, we tested the gluco-
regulatory ability of DVC ALX infusion when GluN1 is inactivated. Similar to that which was
observed with NMDA receptor inhibition, specific chemical antagonism of the GluN1 subunit of
NMDA receptors with 7-chlorokynurenic acid (7CKNA) into the DVC nullifies the ALX-
induced increase of the requirement for exogenous glucose and suppression of glucose
production without affecting glucose uptake or plasma glucose levels (Figure 3.2C,D,
Supplementary Figure 3.1C,D). Selective genetic inhibition of DVC GluN1 subunits with
injection of an adenoviral vector expressing GluN1 shRNA (Ad-GluN1 shRNA) likewise
reverses the ability of ALX infusion to increase glucose infusion rates and lower glucose
production compared to adenovirus injected mismatch sequence (Ad-MM) controls (Figure
3.2C,D, Supplementary Figure 3.1C,D).
To test whether hepatic vagal innervation mediates the gluco-regulatory effects of DVC
GlyT1 inhibition, we examined the effect of DVC ALX in rats with hepatic vagotomy vs. sham
surgery. While hepatic vagotomy or sham surgery per se do not affect glucose kinetics, the
62
higher glucose infusion rate and lower glucose production observed in sham rats receiving DVC
ALX compared to DVC saline are negated in hepatic vagotomized rats, without any difference in
glucose uptake or plasma glucose (Figure 3.2E,F, Supplementary Figure 3.2E,F).
We next performed microdialysis to examine the effect ALX infused into the DVC would
have on the extracellular levels of glycine within the DVC in healthy rats in vivo
(Supplementary Figure 3.1B). When ALX vs. saline is infused into the DVC at a comparable
duration and dosage as the IVGTT and clamp infusion studies in healthy rats (Supplementary
Figure 3.1A), ALX results in a ~2.5-fold increase in extracellular glycine levels in the DVC
(Figure 3.2G). Taken together, DVC GlyT1 inhibition via ALX infusion increases glucose
tolerance and elevates extracellular glycine levels in the DVC to potentiate NMDA receptors and
activate a brain-liver axis to lower glucose production in healthy rodents in vivo.
We alternatively tested the gluco-regulatory role of hindbrain GlyT1 inhibition via the
targeted molecular knockdown of GlyT1 within the DVC. We first confirmed that lentiviral
injection of GlyT1 shRNA (LV-GlyT1 shRNA) into the DVC selectively reduces the expression
of both the 70- and 90-kDa isoforms of GlyT1 in plasma membrane fractions of only the DVC
tissue compared to lentiviral injection of mismatch sequence (LV-MM), but not in the 2 adjacent
left and right lateral regions of the DVC containing the Spinal trigeminal track (sp5), Spinal 5nu
caudal part (Sp5C) and Spinal 5nu interpolar (Sp5I), and the region inferior to the DVC
containing the pyramidal tract (py) of the same rats (Figure 3.3A-D, Supplementary Figure 3.3
i-v). The dominant band at the molecular weight of 70-75 kDa corresponds to GlyT1a and b
isoforms and the weaker band at 90-100 kDa corresponds to GlyT1c isoform found in the rat
brain as described195. The 90-100 kDa band is not detected in the sp5, Sp5C, Sp5I (right) and py
regions (Figure 3.3C,D). The immunoblot also reveals a strong band at 55 kDa (Figure 3.3A),
63
which is consistent with the occurrence of partially glycosylated form of GlyT1 in the 55-60 kDa
range as indicated196-198. However, the 55 kDa GlyT1 band in the DVC of LV- GlyT1 shRNA vs.
- MM injected rats is not significantly different as compared to the effect on 70 kDa and 90 kDa
bands (Figure 3.3A). Nonetheless, the specific metabolic role of various forms of GlyT1 in the
brain warrants future investigation.
A 13-day chronic inhibition of GlyT1 robustly increases glucose infusion rates (Figure
3.3E) and diminishes rates of glucose production (Figure 3.3F) during the clamps, independent
of changes in glucose uptake and plasma glucose (Supplementary Figure 3.4A,B). The DVC
LV-GlyT1 shRNA and MM injected regular-chow fed rats received vascular surgery (for the
clamp studies) on day 8 post-DVC viral injection and the clamp studies were conducted on day
13 (Supplementary Figure 3.1A). The body weights of these viral injected rats remain
comparable on the morning of the clamps at which point rats were also fasted for 4-6 hrs
(Supplementary Figure 3.4C). Importantly, the gluco-suppressive effect of this chronic
molecular GlyT1 inhibition is also mediated through the activation of DVC NMDA receptors
since DVC infusion with MK801 abolishes the effect of LV-GlyT1 shRNA to increase glucose
infusion rates and lower glucose production, unaffected by differences in glucose uptake,
glycemia, or body weight (Figure 3.3E,F, Supplementary Figure 3.4A-C). These molecular
loss-of-function studies strengthen the role of DVC GlyT1 inhibition in elevating extracellular
glycine levels and activating NMDA receptors to lower glucose production in healthy rodents.
3.4.2 Anti-diabetic effect of DVC GlyT1 inhibition
We next sought to ascertain a therapeutic relevance for the glucose-lowering capacity of DVC
GlyT1 inhibition first in 3-day high fat diet (3-d HFD) fed rats (Figure 3.4A). Rats placed on a
3-d HFD were first confirmed to be hyperphagic (cumulative food intake: 258 ± 10 vs 178 ± 11
64
kcal, P<0.01 3-d HFD (n=26) vs 3-d RC (n=11), t-tests) and hyperinsulinemic {3-d HFD (1.7 ±
0.2, n=5) vs. 3d RC rats (0.9 ± 0.1, n=8), P<0.05, t-tests}, consistent with the fact that 3d HFD
rats were validated in parallel under hyperinsulinemic-euglycemic clamp conditions in our
research facility to exhibit hepatic insulin resistance179. We here evaluated whether antagonism
of DVC GlyT1 modulates glucose homeostasis in these 3-d HFD rats to the same extent as direct
DVC glycine infusion during the pancreatic (basal insulin)-euglycemic clamp conditions, given
that DVC GlyT1 inhibition increases extracellular DVC glycine levels (Figure 3.2G). Indeed,
DVC GlyT1 inhibition with ALX increases the requirement of glucose (Figure 3.4B) and
suppresses the rate of glucose production (Figure 3.4C) independent of alterations in glucose
uptake (Supplementary Figure 3.5A) and plasma glucose levels (Supplementary Figure 3.5B)
during the pancreatic clamp in HFD rats to the same extent as DVC glycine infusion (Figure
3.4B,C). Further, this glucose production-lowering effect of DVC ALX or glycine in 3-d HFD
rats requires the activation of the NMDA receptor GluN1 subunits as co-infusion of 7CKNA
with ALX or glycine abates the glucose-suppressive ability of ALX and glycine (Figure
3.4B,C).
We next examined whether systemic administration of GlyT1 inhibitor ALX recapitulates
the glucose production-lowering effect of DVC ALX infused-dependent GlyT1 inhibition in 3-
day HFD-fed rats. Strikingly, constant intravenous (i.v.) infusion of ALX for 5 hrs leads to a
higher glucose infusion rate (Figure 3.4D) and lower glucose production (Figure 3.4E)
compared to i.v. 6% DMSO vehicle infusion during the pancreatic clamps, and these metabolic
changes occur independent of glucose uptake (Supplementary Figure 3.5C) and changes in
plasma glucose levels (Supplementary Figure 3.5D).
65
Next, we evaluated the effects of DVC ALX infusion in a rat model of type 2 diabetes
(Figure 3.4A) that is considered a better representation of humans with type 2 diabetes191. Rats
were injected with nicotinamide (Nic) and low-dose streptozotocin (STZ) to prevent beta-cell
compensation for HFD-induced insulin resistance, and maintained on a HFD for 7 days (Figure
3.5A). We have confirmed these 7-d STZ/Nic/HFD rats have fasting hyperglycemia (Figure
3.5B) and validated in parallel in our research facility to exhibit elevated hepatic glucose
production179,180. In addition, diabetic 7-d STZ/Nic/HFD + DVC saline-infused rats are glucose
intolerant as they have markedly elevated total glucose excursions during ivGTT compared to
their non-diabetic regular chow fed DVC saline-infused counterparts (Figure 3.5C).
Interestingly, ALX infusion into the DVC markedly lowers total glucose excursions in diabetic
rats compared to DVC saline infusion (Figure 3.5C). Thus, these experiments indicate a gluco-
regulatory therapeutic potential for DVC GlyT1 inhibition in high-fat fed or diabetic rodents.
3.4.3 Metabolic benefits of DVC GlyT1 inhibition in obesity
We next assessed whether DVC GlyT1 inhibition improves glucose metabolism in 28-d
HFD-induced obese rats (Figure 3.6A). Rats fed a HFD for 28 days were first confirmed to be
obese (Figure 3.6B) and hyperinsulinemic {28d HFD rats (2.5 ± 0.2, n=10) vs. 28d RC rats (1.9
± 0.2, n=9), P<0.05, t-test}, consistent with the fact that this obese model was validated in
parallel under hyperinsulinemic-euglycemic clamp conditions in our research facility to exhibit
hepatic and peripheral insulin resistance179. Importantly, in both 28-d regular chow and HFD
cohorts, we here report that acute inhibition of DVC GlyT1 with ALX infusion into the DVC
increases glucose infusion rates (Figure 3.6C) and lowers glucose production (Figure 3.6D)
independent of changes in glucose uptake (Supplementary Figure 3.6A) and plasma glucose
levels (Supplementary Figure 3.6B) during the pancreatic (basal insulin)-euglycemic clamp
66
conditions. Notably, the glucose production-lowering effect of acute DVC GlyT1 antagonism is
evident in spite of the weight gain incurred by chronic high-fat feeding (body weight on the
morning of clamp experiments: 419 ± 7 vs 390 ± 10 g, P<0.05 28-d HFD vs 28-d RC, t-test).
Given that acute inhibition of DVC GlyT1 improves glucose homeostasis in short-term
(Figure 3.4C) and long-term high-fat fed rats, we postulated that chronic inhibition of GlyT1 in
the DVC might confer a gluco-regulatory benefit during 28 days of HFD-induced obesity. We
tested this hypothesis by subjecting 28-d HFD-fed rats to targeted knockdown of GlyT1 in the
DVC (via DVC LV-GlyT1 shRNA injection on Day 16 after HFD; Figure 3.7A) to determine
whether this chronic (from Day 16-Day 29; Figure 3.7A) intervention modulates glucose
homeostasis. Indeed, chronic genetic inhibition of GlyT1 in the DVC robustly increases the
glucose infusion rate (Figure 3.7B) and suppresses glucose production (Figure 3.7C) as
compared with MM controls. This glucose-lowering effect occurs independent of changes in
glucose uptake (Supplementary Figure 3.6C) and plasma glucose levels (Supplementary
Figure 3.6D). Surprisingly, body weights on the morning of clamp experiments are markedly
lower in 28-d HFD-rats with chronic DVC GlyT1 inhibition (Figure 3.7D). In fact, this lowering
of body weight in 28-d HFD-induced obese rats is evident by 4 d post-viral (LV-GlyT1 shRNA
vs. LV-MM) injection (Figure 3.7E). However, it is unlikely that the gluco-regulatory
improvement results from a decrease in body weight since chronic DVC GlyT1 inhibition lowers
glucose production in healthy rats without affecting body weight on the morning of the clamp
studies (Figure 3.3F, Supplementary Figure 3.4C).
3.4.4 DVC GlyT1 inhibition regulates energy balance
Given that acute DVC GlyT1 inhibition lowers glucose production in 28-d HFD obese rats
(Figure 3.6D), it was important to next investigate whether the local elevation of glycine in the
67
DVC associated with DVC GlyT1 inhibition can also regulate energy balance. First, we tested
the direct effect of DVC glycine sensing on appetite and body weight regulation in healthy rats
that received DVC surgery 13 days prior (Figure 3.8A). Following a 22-h fast, injection of
glycine into the DVC begins to lower food intake compared to saline infused controls by ~120
min post-injection and refeeding (Figure 3.8A,B), an effect that becomes significant by 180 min
and persisted for 1 day after refeeding (Figure 3.8C). DVC glycine has no significant effect on 1
or 2 day post-refeeding percent body weight gain (Figure 3.8D). Secondly, we assessed whether
chemical inhibition of GlyT1 in the DVC could recapitulate these glycine-induced satiation
effects. Indeed, DVC ALX injection reduces food intake by 60 min after refeeding (or 120 min
post-ALX injection) (Figure 3.8E,F) with the effect still present 1 day post-refeeding (Figure
3.8G). DVC ALX also reduces the percent body weight gain after 1 and 2 day but not 3 day of
refeeding (Figure 3.8H). Finally, we evaluated whether molecular inhibition of GlyT1 in the
DVC could regulate energy balance. LV-GlyT1 shRNA or LV-MM was injected into the DVC
of healthy rats to knockdown DVC GlyT1, resulting in reduced body weight of LV-GlyT1
shRNA rats 4 day post-viral injection compared to LV-MM (Figure 3.8I), similar to the effect
observed in obese rats (Figure 3.7E). The viral-injected regular chow-fed rats were then
subjected to a 22-h fast in the evening of day 4, and genetic knockdown of DVC GlyT1 lowers
cumulative food intake as early as 120 min following refeeding (Figure 3.8J) and up to 1 day
post-refeeding (Figure 3.8K). In parallel, LV-GlyT1 shRNA vs LV-MM injection lowers the
percent body weight gain following 1 and 2 day but not 3 day post-refeeding (Figure 3.8L).
Taken together, we provide evidence that DVC GlyT1 inhibition and subsequent glycine
elevation triggers a sensing mechanism in the DVC to lower feeding and body weight in rats.
68
3.5 Discussion We have shown that targeted inhibition of DVC GlyT1 through either administration of a GlyT1
inhibitor or a chronic molecular knockdown improves glucose homeostasis and lowers body
weight gain in diabetic and obese rodents.
The effect of DVC GlyT1 inhibition on glucose production regulation requires a hepatic
vagal-dependent communication between the brain and the liver. Although the neurocircuitry
involved in food intake and body weight regulation by DVC GlytT1 inhibition (or glycine
sensing) remains unclear, the underlying neuronal relay is likely different than glucose
production regulation (since glucose production is altered by DVC GlyT1 inhibition independent
of changes in food intake and body weight (Figure 3.2D, Figure 3.3F, Figure 3.6D) as well as
blood pressure and heart rate regulation (since DVC injection of glycine or glutamate induce
changes of blood pressure and heart rate at a much faster rate199,200 than changes in feeding
induced by DVC glycine injection (Figure 3.8B)). It would be important to follow-up on the
potential long term control of food intake and body weight regulation via repeated injections of
glycine or ALX, particularly knowing that a knock-down of DVC GlyT1 for 13 days exerts an
anti-obesity effect.
Although the individual cells in the DVC involved in the metabolic control of DVC
GlyT1 inhibition remain to be identified, the potentiation and activation of the GluN1/GluN2-
containing NMDA receptor in the DVC is necessary for the gluco-regulatory effect of DVC
GlyT1 inhibition (Figure 3.2-3.4). Given that the NMDA receptors are expressed in the plasma
membrane and are necessary for the metabolic effect of DVC GlyT1 inhibition, DVC ALX
infusion increases extracellular glycine levels within the DVC as assessed by microdialysis
(Figure 3.2G), and that DVC glycine infusion (like GlyT1 inhibition) potentiates DVC NMDA
receptors to lower glucose production in healthy138 and 3-d HFD (Figure 3.4C) rats, glycine is
69
proposed to be the endogenous agonist that mediate the metabolic control of DVC GlyT1
inhibition. D-serine, like glycine, is also a co-agonist of the NMDA receptors. However, it is
unlikely that D-serine is the endogenous agonist that mediates the effects of GlyT1 inhibition as
injection of GlyT1 inhibitor elevates extracellular glycine but not serine and glutamate levels in
the brain of rats189. Future studies are necessary to dissect the specific role of glycine vs. serine
per se as well as in the presence of GlyT1 inhibition in regulating glucose and energy
homeostasis.
Although ketamine (a partial NMDA receptor antagonist) was used to anaesthetize the
animals for brain and vascular surgeries, any potential confounding effects of ketamine on the
gluco-regulatory studies should be absent by the time we carry out the infusion experiments as
body weight and food intake of the rodents have fully recovered. In addition, MK-801 inhibits
the GluN1/GluN2 but not the GluN1/GluN3 NMDA receptors201. Given that in our current study,
DVC MK-801 fully reverses the ability of both DVC ALX infusion and DVC LV-GlyT1 shRNA
viral injection to inhibit glucose production (Figures 3.2, 3.3), it is likely that activation of the
GluN1/GluN2 receptors, and not GluN1/GluN3, is essential for the metabolic effects of DVC
GlyT1 inhibition and glycine sensing. Consistent with this hypothesis, bi-directional changes of
NMDA receptors in the DVC via DVC infusion of NMDA or NMDA receptor antagonist AP5
alter glucose production138, while strychnine-sensitive glycine receptors do not appear to mediate
DVC glycine sensing to regulate glucose production138, altogether strengthening the claim that
GluN1/GluN2-containing NMDA receptors mediate the glucose-lowering effect of DVC GlyT1
inhibition. Nonetheless, a role for DVC GluN1/GluN3 NMDA receptor in glucose regulation
remains to be directly assessed.
DVC GlyT1 inhibition improves glucose tolerance independent of a rise in plasma insulin
levels and lowers glucose production when insulin levels are maintained at basal during the
70
pancreatic clamps. Thus, it is tempting to speculate that DVC GlyT1 inhibition may improve
glucose homeostasis in type 1 diabetic insulin-deficient conditions, particularly knowing that
leptin action in the brain and the gut, as well as nutrient sensing in the gut, have been
documented to improve glucose homeostasis in insulin-deficient type 1 diabetic rodents119,202-204.
This working hypothesis warrants future investigation. On the other hand, it would also be of
future interest to assess whether DVC GlyT1 inhibition reverses insulin resistance in type 2
diabetic and obese rodents using the hyperinsulinemic-euglycemic clamp technique to achieve
insulin-stimulated conditions.
The finding that systemic administration of ALX can recapitulate the glucose production-
lowering effect of GlyT1 inhibition in the DVC during the pancreatic (basal insulin)-euglycemic
clamp settings further substantiates the potential therapeutic relevance of GlyT1 inhibitors in
diabetes and obesity. However, given that NMDA receptors are also expressed in the islets and
alter glucose-stimulated insulin secretion205, future studies are warranted to investigate the short
and long-term metabolic benefits of ALX administration in non-clamp conditions.
Our current set of findings serve as proof of concept for potential of GlyT1 inhibition as a
singular therapeutic target for the concurrent treatment of both diabetes and obesity, in addition
to its current use in the treatment of schizophrenia. Interestingly, patients with schizophrenia
have over four times the risk for abdominal obesity and twice the risk for diabetes compared to
general population controls206, highlighting the possibility that common pathologies may
contribute the development of these diseases.
Among several GlyT1 inhibitors that have undergone clinical trials, bitopertin has seen
the most success by advancing to phase III trials for the treatment of schizophrenia182,183. ALX,
on the other hand, has demonstrated relatively poorer tolerance in vivo207. However, although
ALX never entered clinical trials, it is extensively used as a pharmacological tool for the study of
71
glycine transporter function207. Interestingly, DVC administration of ALX in the present study
increases DVC extracellular glycine levels which mimics a comparable effect of a low oral dose
of bitopertin on CSF glycine levels in rats208. These two drugs may therefore trigger a similar
degree of NMDA receptor-mediated neurotransmissions to elicit comparable metabolic effects.
Given that several GlyT1 inhibitors have successfully demonstrated safety and efficacy in
humans and that systemic ALX infusion recapitulates the ability of ALX infusion into the DVC
to lower glucose production in HFD rats, we propose that GlyT1 inhibitors be considered as
pharmacological agents for the restoration of glucose and energy homeostasis in obesity and
diabetes.
72
Figure 3. 1 Schematic representation of the working hypothesis.
Glycine transporter-1 (GlyT1) facilitates the cellular uptake of glycine in the dorsal vagal
complex (DVC). Chemical (via DVC ALX infusion) or genetic (via DVC lentiviral injection of
GlyT1 shRNA) inhibition of GlyT1 increases extracellular glycine levels in the DVC, which
potentiates the activation of DVC N-methyl-D-aspartate (NMDA) receptors to regulate glucose
production and glucose tolerance, and food intake and body weight gain. MK-801, NMDA
receptor ion channel blocker. 7-chlorokynurenic acid, 7CKNA-antagonist to the GluN1 subunit
of NMDA receptors.
73
Figure 3. 2 Chemical inhibition of DVC GlyT1 regulates glucose homeostasis in healthy
rats.
(A) Plasma glucose levels (inset: integrated area under the curve (AUC)) and (B) plasma insulin
levels during ivGTT with DVC infusion of ALX (n=8, black squares) or saline (n=7, white
squares). †P<0.04, ††P<0.0008 determined by Sidak’s multiple comparisons test following
repeated measures ANOVA. *P<0.05 determined by t-test. (C) Glucose infusion rates and (D)
glucose production during clamps with DVC infusion of saline (n=11), ALX (n=9), MK801
(n=9), ALX+MK801 (n=5), ALX+7CKNA (n=5), Ad-MM+ALX (n=5), or Ad-GluN1
shRNA+ALX (n=5). (C *P<0.002 vs saline, MK801, ALX+MK801, and ALX+7CKNA
determined by ANOVA and Dunnett’s post hoc test; †P<0.002 vs Ad-GluN1 shRNA+ALX
determined by t-test; D: *P<0.02 vs saline, MK801, ALX+MK801, and ALX+7CKNA
determined by ANOVA and Dunnett’s post hoc test; †P<0.0008 vs Ad-GluN1 shRNA+ALX
determined by t-test.) (E) Glucose infusion rates and (F) glucose production during clamps with
DVC ALX infusion in vagotomized (n=7) or sham-operated (n=5) rats or DVC saline infusion in
vagotomized (n=7) or sham-operated rats (n=5) rats. (E,F*P<0.01 compared to all other groups
determined by ANOVA and Dunnet’s post hoc test.). (G) Extracellular glycine levels within the
DVC following DVC infusion of ALX (n=7) or saline (n=7) in microdialysis studies. *P<0.03 vs
saline determined by t-test. Data are shown as the mean + SEM.
74
Figure 3. 3 Molecular inhibition of DVC GlyT1 regulates glucose homeostasis in healthy
rats.
(A) Representative Western blots and protein levels of plasma membrane GlyT1 (55, 70 and 90
kDa isoforms) normalized to insulin receptor (IR) in DVC wedges of rats 13 day-post DVC
lentiviral (LV) injection of GlyT1 shRNA (black bars, n=14) or a mismatch sequence (MM;
white bars, n=11) as a control. *P<0.01, **P<0.001 determined by t-test. (B-D) Representative
Western blots and protein levels of plasma membrane GlyT1 (70 and/or 90 kDa isoforms)
normalized to IR in sp5, Sp5C, Sp5I (L), sp5, Sp5C, Sp5I (R) and py wedges of rats 13 day-post
DVC LV injection of GlyT1 shRNA (black bars, n=5) or MM (white bars, n=5). (E) Glucose
infusion rates and (F) glucose production during clamps in rats injected with LV-MM (n=7), LV-
GlyT1 shRNA (n=7), or LV-GlyT1 shRNA with DVC MK801 infusion (n=6). (E: *P<0.006; F:
*P<0.003 vs LV-MM control and LV-GlyT1 shRNA+MK801 determined by ANOVA and
Dunnett’s post hoc test.). Data are shown as the mean + SEM.
75
Figure 3. 4 DVC and iv infusion of ALX regulates glucose homeostasis in 3d-HFD rats.
(A) Experimental protocol for panels B-C. (B) Glucose infusion rates and (C) glucose
production during clamps with DVC infusion of saline (n=5), glycine (n=6), ALX (n=5),
glycine+7CKNA (n=5), and ALX+7CKNA (n=5). (B: *P<0.0003 vs saline and
Glycine+7CKNA; †P<0.001 vs saline and ALX+7CKNA; determined by ANOVA and
Dunnett’s post hoc test; C: *P<0.006 vs saline and Glycine+7CKNA; †P<0.002 vs saline and
ALX+7CKNA; determined by ANOVA and Dunnett’s post hoc test.) (D) Glucose infusion rates
and (E) glucose production during clamps with iv infusion of 6% DMSO (n=7) or ALX (n=7) in
3-d HFD rats. (D,E: *P<0.001 vs iv DMSO determined by t-test.)
76
Figure 3. 5 Inhibition of DVC GlyT1 regulates glucose homeostasis in diabetic rats.
(A) Experimental protocol for panels B-C. (B) Plasma levels of glucose in overnight-fasted 7d
STZ/Nic/HFD diabetic rats (black bars, n=17) compared to non-diabetic, regular chow-fed
counterparts (white bars, n=13); *P<0.01 determined by t-test. (C) Plasma glucose levels (inset:
integrated area under the curve (AUC)) during ivGTT with DVC infusion of ALX (n=9, grey
triangles) or saline (n=8, black triangles) in 7d STZ/Nic/HFD rats or DVC saline in regular chow
rats (n=7, white squares). †P<0.05, ††P<0.01, ††† P<0.001 vs DVC saline + regular chow rats; ‡
P<0.05, ‡‡ P<0.01 vs DVC ALX + 7d STZ/Nic/HFD rats determined by ANOVA and Dunnett’s
post hoc test; AUC: *P<0.05, **P<0.01 vs. DVC saline + 7d STZ/Nic/HFD rats determined by
ANOVA and Dunnet’s post-hoc test. Data are shown as the mean + SEM.
77
Figure 3. 6 Chemical inhibition of DVC GlyT1 regulates glucose homeostasis in obese rats.
(A) Experimental protocol for panels B-D (B) Body weight gain in rats that were fed with HFD
(white circles, n=18) or regular chow (RC, white squares, n=6). Inflections of the body weight
curves at d 16 and d 24 represent DVC cannulation and vascular catheterization surgery days,
respectively. *P<0.02 main effect of diet, F(1,22)=6.964 determined by repeated measures
ANOVA. (C) Glucose infusion rates and (D) glucose production during clamps in 28d RC-fed
rats with DVC infusion of saline (n=5) or ALX (n=5) and in 28d HFD-fed rats with DVC
infusion of saline (n=7) or ALX (n=7) (C,D: *P<0.01 vs. the respective DVC saline determined
by ANOVA and Tukey’s post-hoc test).
78
Figure 3. 7 Molecular inhibition of DVC GlyT1 regulates metabolic homeostasis in obese
rats.
(A) Experimental protocol for panels B-E. (B) Glucose infusion rates and (C) glucose
production during clamps in 28d HFD-fed rats with DVC lentivirus (LV) injection of GlyT1
shRNA (n=10) or a mismatch sequence (MM, n=9) as a control. (B, C: *P<0.001 vs 28d-
HFD+MM determined by t-test.) (D) Body weights on the morning of clamp experiments in 28d
HFD-fed rats with DVC LV MM or GlyT1 shRNA. *P<0.04 determined by t-test. (E) Percent
body weight change on days 4 and 5 following DVC injection of LV-MM (white circles, n=9) or
GlyT1 shRNA (black circles, n=9) in 28d HFD-fed rats and of MM fed with regular chow (white
squares, n=5). *P<0.05 compared to all other groups determined by ANOVA and Dunnett’s post
hoc test. Data are shown as the mean + SEM.
79
Figure 3. 8 Chemical and molecular inhibition of DVC GlyT1 regulate energy balance.
(A) Experimental protocol for feeding experiments in rats that received DVC injection of
glycine (black squares, n=11) or saline (white squares, n=11). (B) Cumulative food intake during
the feeding experiment. (C) Daily food intake on day 1 and day 2 after food was returned during
the feeding experiment. (D) Percent body weight gain on day 1 or day 2 after the feeding
experiment. (E) Experimental protocol for feeding experiments in rats that received DVC
injection of ALX (black squares, n=8) or saline (white squares, n=8). (F) Cumulative food intake
during the feeding experiment. (G) Daily food intake on day 1 and day 2 after food was returned
during the feeding experiment. (H) Percent body weight gain on day 1, day 2 or day 3 after the
feeding experiment. (I) Percent body weight change on day 4 following DVC injection of LV-
MM (n=8) or GlyT1 shRNA (n=10) in regular chow-fed rats. (J) Cumulative food intake during
the feeding experiment in rats injected with LV-MM (n=6) or GlyT1 shRNA (n=7). (K) Daily
food intake on day 1 and day 2 after food was returned during the feeding experiment. (L)
Percent body weight gain on day 1, day 2 or day 3 after the feeding experiment. †P<0.05,
††P<0.01, †††P<0.001 determined by t-test at each time point, ‡P<0.05 determined by t-test at
each time. Data are shown as the mean + SEM.
80
Supplementary Figure 3. 1 Metabolic effects of chemical inhibition of DVC GlyT1 in
healthy rats.
(A) Experimental protocols for experiments shown in Figure 2. In intravenous glucose tolerance
tests (ivGTT), infusions of saline or ALX into the DVC commenced at t=-240min and were
maintained for the duration of the experiment. In clamps experiments, pre-infusions of saline,
MK801, or 7CKNA into the DVC commenced at t=-90min. Infusions of saline, MK801, ALX,
ALX+MK801, or ALX+7CKNA into the DVC commenced at t=-60min and were maintained for
the duration of the clamps. (B) Experimental protocol for microdialysis studies shown in Fig. 1i.
(C) Glucose uptake, (D) basal and clamp plasma glucose levels during clamps with DVC
infusion of saline (n=11), ALX (n=9), MK801 (n=9), ALX+MK801 (n=5), ALX+7CKNA (n=5),
Ad-MM+ALX (n=5), or Ad-GluN1 shRNA+ALX (n=5), (E) basal and clamp plasma insulin
levels during DVC infusion of saline (n=6) or ALX (n=6), and (F) body weights on the morning
of clamp experiments before DVC infusion of saline (n=11) or ALX (n=9). Data are shown as
the mean + SEM.
81
Supplementary Figure 3. 2 Metabolic effects of chemical inhibition of GlyT1 in the 4th
ventricle of mice and in the DVC of hepatic vagotomized rats.
(A) Glucose infusion rates, (B) rates of clamp glucose production, (C) glucose uptake, and (D)
clamp plasma glucose levels during clamp experiments with ICV-4th ventricle infusion of ALX
(n=5) or saline (n=6) in C57BL/6 mice; *P<0.02 vs saline determined by t-test. (E) Glucose
uptake, and (F) basal and clamp plasma glucose levels during clamp experiments with DVC
infusion of ALX in vagotomized (n=7) or sham-operated (n=5) rats or DVC infusion of saline in
vagotomized (n=7) or sham-operated rats (n=5) rats. Data are shown as the mean + SEM.
82
Supplementary Figure 3. 3 Brain regions included in the GlyT1 protein analysis.
This include the DVC region (blue), the left (L) lateral region (yellow) containing spinal
trigeminal tr. (sp5), spinal 5nu, caudal part (Sp5C), spinal 5nu, interpolar (Sp5I), the right (R)
lateral region (purple) containing sp5, Sp5C, Sp5I and the bottom region (green) containing
pyramidal tr. (py). i: Saggital image representing the rat brain. ii: Vagal triangle overlaying the
DVC located in the caudal part of the brain. iii-v: three coronal images representing the
proximal, medial and distal regions of the caudal brain indicating all the regions included in the
analysis of the GlyT1 protein.
83
Supplementary Figure 3. 4 Metabolic effects of molecular inhibition of DVC GlyT1 in
healthy rats.
(A) Glucose uptake, (B) basal and clamp plasma glucose levels during clamps, and (C) body
weights on the morning of clamp experiments in rats injected with DVC LV-MM (n=7), LV-
GlyT1 shRNA (n=7), or LV-GlyT1 shRNA with DVC MK801 infusion (n=6). Data are shown
as the mean + SEM.
84
Supplementary Figure 3. 5 Metabolic effects of DVC and iv infusion of ALX in 3d-HFD
rats.
(A) Glucose uptake and (B) basal and clamp plasma glucose levels during clamps with DVC
infusion of saline (n=5), glycine (n=6), ALX (n=5), glycine+7CKNA (n=5), or ALX+7CKNA
(n=5) in 3-d HFD rats, (C) Glucose uptake and (D) basal and clamp plasma glucose levels during
clamps with iv infusion of 6% DMSO (n=7) or ALX (n=7) in 3-d HFD rats. Data are shown as
the mean + SEM.
85
Supplementary Figure 3. 6 Metabolic effects of chemical and molecular inhibition of DVC
GlyT1 in obese rats.
(A) Glucose uptake and (B) basal and clamp plasma glucose levels during clamps in 28d RC-fed
rats with DVC infusion of saline (n=5) or ALX (n=5) and in 28d HFD-fed rats with DVC
infusion of saline (n=7) or ALX (n=7), (C) Glucose uptake and (D) basal and clamp plasma
glucose levels during clamps in 28d HFD-fed rats injected with DVC LV-GlyT1 shRNA (n=10)
or LV-MM (n=9). Data are shown as the mean + SEM.
86
Chapter 4 Summary, Discussion and Future
Directions 4.1 Summary The ability of the CNS to detect and integrate peripheral humoral signals is an important
requirement for the regulation of glucose and energy homeostasis in a mammalian organism. The
current set of studies were aimed at further characterizing novel molecular mechanisms of CNS
hormone and nutrient sensing that control glucose and energy homeostasis.
Previously, our lab demonstrated that glucagon action in the MBH, via glucagon
receptor-PKA signaling pathway leads to an inhibition of hepatic glucose production, which is
disrupted in rats fed a high-fat diet for three days. This indicates hypothalamic glucagon
resistance exists in a pathological state. However, this resistance could be reversed by direct
activation of hypothalamic PKA, which implies, that the hypothalamic glucagon signaling defect
lies upstream of PKA. Investigating what lies downstream of PKA in MBH glucagon signaling
thus became of therapeutic interest, to identify targets that could potentially enhance or restore
hypothalamic glucagon action in diabetes and/or obesity. In hypothalamic cell lines, PKA has
been demonstrated to inhibit AMPK88 while MBH AMPK-mediated lipid sensing mechanisms
has been shown to lower glucose production102. It was unknown whether lipid sensing was a
downstream mechanism of MBH glucagon action to regulate glucose homeostasis. In Study 1,
we have shown that MBH glucagon does not signal through the lipid sensing axis involving
AMPK and PKC-δ rather activates KATP channels that lie downstream of MBH glucagon-PKA
signaling to lower glucose production (Figure 4.1).
87
Similar to the MBH, the DVC is anatomically poised to detect hormones and nutrients to
regulate metabolic homeostasis. Direct administration of glycine into the DVC activates NMDA
receptors to lower glucose production in healthy rodents138. What remains to be shown is
whether DVC glycine-NMDA receptor axis works in diabetic and obese rodents to regulate
glucose homeostasis, as well as control energy balance. Given the poor pharmacokinetics of
glycine, Study 2 aimed to address whether inhibiting DVC GlyT1, the main regulator of glycine
levels for NMDA receptors, would sufficiently trigger endogenous glycine sensing to regulate
glucose and energy homeostasis. The major findings of these studies were that DVC GlyT1
inhibition elevates extracellular glycine availability to activate NMDA receptors in the DVC,
leading to improved glucose tolerance, lowered glucose production, reduced body weight gain
and food intake in healthy, diabetic and obese rodents (Figure 4.1). We have also reported that
intravenous infusion of GlyT1 inhibitors resulted in similar metabolic effects to those of DVC
GlyT1 inhibition. The significance of these major findings as well as potential avenues of further
research are discussed below.
88
Figure 4. 1 Summary of Study 1 and Study 2.
Summarized representation illustrating that glucagon action in the MBH exerts an AMPK-
independent but a KATP channel- dependent pathway to exert glucose control, and that GlyT1
inhibition in the DVC sufficiently triggers glycine sensing in the DVC to modulate energy and
glucose homeostasis via activation of NMDA receptors.
89
4.2 Discussion Ion channels in the CNS regulation of glucose and energy homeostasis.
The findings of this dissertation highlight the importance ion channels play in the CNS pathways
regulating glucose balance and food intake. KATP channels are one of the most ubiquitously
expressed ion channels in the brain209, where opening (or activation) of KATP channels induces
efflux of K+ ions and hyperpolarization of the cell membrane leading to suppressed neuronal
activity and excitability. In study 1, the loss of MBH glucagon’s effect to lower glucose
production by MBH injection of the dominant negative Kir6.2 mutant virus as well as by MBH
infusion of the pharmacological inhibitor glibenclamide is indicative that KATP channels mediate
the glucose regulating effects of MBH glucagon action. More specifically, our data suggests that
glucagon induces KATP channel-dependent hyperpolarization in MBH neurons to lower glucose
production. Consistent with this idea, our lab previously documented the co-localization of MBH
glucagon receptors with AgRP neurons79, which makes it conceivable that MBH glucagon
infusion could be causing hyperpolarization and inhibited firing of the orixegenic AgRP neurons
(mediated via KATP channel activation) to lower glucose production. Indeed, activated KATP
channels have been previously reported to mediate the hyperpolarization and inhibitory effects of
insulin on AgRP neurons, which in turn leads to suppressed glucose production21,210. Further,
Study 1 also corroborates those reports that showed pharmacologically depolarizing
hypothalamic KATP channels (via application of KATP channel inhibitors into the MBH) blocks
the glucose production lowering effects that are induced by MBH insulin, glucose, fatty acids or
leucine. Taken together, hypothalamic KATP channels represent a critical and common CNS ion
channel through which different types of hormones and nutrients regulate glucose homeostasis.
Although the role of KATP channels in the regulation of energy homeostasis remains elusive, it
has been shown that activation of hypothalamic KATP channels contributes to age-dependent
90
obesity. For instance, in old mice, there is an increase of KATP channel activity, which causes
silencing of hypothalamic POMC neurons to reduce their leptin-induced release of the
anorexigenic α-MSH, displaying a hyperphagic and obese phenotype211. Investigation of KATP
channel activity within specific hypothalamic neurons (AgRP vs POMC) merits future
investigation for our thorough understanding of the role these channels play for energy
homeostasis.
Study 2 demonstrated that NMDA receptors, which are ligand-gated ion (Ca2+) channels,
in the DVC or more specifically the NTS, play a critical role in mediating the metabolic effects
of GlyT1 inhibition. Earlier studies have shown that NMDA receptor signaling relay ascending
signals from the gut to regulate glucose and energy homeostasis116,117,120, and that direct
activation of NTS NMDA receptors is sufficient138 and necessary121 for hypothalamic nutrient
sensing to lower hepatic glucose production. Recent studies further place emphasis on the role
NMDA receptor activation in the hindbrain play in the reduction of food intake and body
weight212. Upon co-activation by glutamate and glycine, NMDA receptors elicit membrane
depolarization of glutamatergic neurons and Ca2+-dependent signaling cascades via increased
conductance of Na+ and Ca2+ ions. Depolarization of the glutamatergic neurons in the NTS in
turn leads to excitation of DMV neurons, whose axons form the efferent limb of the vagus
nerve213. Consistently, in Study 2, we did show that DVC ALX treatments completely failed to
lower glucose production in hepatic-vagotomozied rats, confirming that glucose-regulation by
DVC GlyT1 inhibition indeed involves a specific neuronal relay via the vagus nerve to the liver.
Interestingly, the gluco-regulatory effects of MBH glucagon also required an intact
hepatic vagal signaling79. This together with the fact that NMDA receptors in the DVC can
integrate information from hypothalamic nutrient signals such as MBH lactate to regulate
91
glucose homeostasis, it opens up the question as to whether MBH glucagon action (from Study1)
are also downstream mediated by NMDA receptors in the DVC. While future experiments are
warranted to test this hypothesis, there is some evidence showing that activation of KATP channel
and its neuronal hyperpolarization influence in fact play a protective role against over-
stimulation of NMDA receptors (or glutamate excitotoxicity), typically known to cause neuronal
cell death214. Interestingly, it is also suggested that NMDA receptor stimulation can lead to KATP
channel activation. For instance, NMDA receptor activation stimulates nitric oxide (NO)
production by Ca2+ -dependent activation of nitric oxide synthase (NOS)215-217, while NO-
mediated activation of cGMP-dependent protein kinase in turn can activate KATP channel
opening218. Whether this coupling of NMDA receptors to KATP channels or vice versa play a role
in CNS regulation of metabolic homeostasis remain to be investigated.
Indeed, NMDA receptors are also present in the hypothalamus but interestingly these
hypothalamic channels are shown to influence energy homeostasis in the opposite direction as in
the DVC. For instance, in the LH, it is the injection of NMDA receptor antagonists that lowers
feeding219 while intrahypothalamic injection of glutamate analogs, that are specific to NMDA
receptors, increases food intake220. Further, deletion of GluN1 or GluN2 subunits from
hypothalamic AgRP neurons cause lowering of food intake and body weight221,222. Nonetheless,
glutamatergic action and NMDA receptors in both the hypothalamus and the brainstem are
important for energy balance control. Of note, there is little evidence on the role of
hypothalamic NMDA receptors in regulating glucose homeostasis.
92
The relevance of MBH glucagon action and DVC GlyT1 inhibition in health and disease
Indeed, the prime objective of this dissertation was to unveil novel molecular mechanisms in the
CNS that would serve as therapeutic targets to lower blood glucose, food intake and body weight
gain in diabetes and obesity. As our previous study has demonstrated that hypothalamic glucagon
resistance in the context of high fat feeding is manifested by the inability of glucagon receptor
signaling to activate PKA79, the findings of Study 1 indicate that activating hypothalamic KATP
channels may be therapeutically advantageous. In fact, activation of hypothalamic KATP channels
by oral administration of the KATP channel activator, diazoxide has already been implicated in
lowering glucose production in humans163. This study performed the same euglycemic –
somatostatin clamping technique as in our dissertation work, fixing the gluco-regulatory
hormones in circulation and showed that in healthy individuals, oral diazoxide treatment led to a
30% suppression of glucose production. Additional studies in healthy rats confirmed that
diazoxide’s inhibitory effects on glucose production are negated in the presence of the KATP
channel blocker glibenclamide, suggesting these effects are likely mediated by KATP channels in
the brain163. Furthermore, intranasal administration of insulin at doses that increases the insulin
concentration in the cerebrospinal fluid (CSF) led to a suppression of glucose production during
pancreatic clamps in humans, likely through activation of CNS KATP channels176.
Additionally, our study contributes to the evolving association between increased
glucagon action and a metabolically healthier phenotype- a theme that has been garnering
scientific attention in the recent years. Multiple studies have reported that activation of glucagon
receptors in conjunction with other G protein-coupled receptors is metabolically advantageous in
diabetes and obesity. For example, a triagonist aimed at simultaneous activation of glucagon,
GLP-1 and glucose-dependent insulinotropic (GIP) receptors improved metabolic and glycemic
profiles in obese and diabetic rodents223. In addition, dual activation of glucagon and GLP-1
93
receptors normalized glucose tolerance and reduced food intake in mice with diet-induced
obesity224,225. These pre-clinical studies have attributed the beneficial effects of glucagon
primarily to increased energy expenditure and decreased food intake, and its hyperglycemic
effects to be countered by the actions of GLP-1 and/or GIP. Whether these polyagonists reach
the brain and activate central glucagon signalling to improve the diabetogenic effects in these
polyagonist therapeutic strategies remain to be investigated. In the same light, a recent
investigation reports that reduced glucagon suppression 2 hours after an intravenous glucose
challenge is associated with a healthier metabolic phenotype including lower BMI, higher insulin
sensitivity and reduced risk of impaired glucose tolerance226. These findings together with our
study reshape our understanding of glucagon’s physiological role in health and disease.
In Study 2, we took advantage of multiple disease models to test the therapeutic potential
of DVC GlyT1 inhibition/glycine sensing in diabetes and obesity. The fasting hyperglycemic
(type 2 diabetic) model used in our dissertation, arguably is the closest rodent model
recapitulating the pathogenesis of type 2 diabetes in humans. Elegantly described by Samuel et
al.191, a low dose of STZ protected by nicotinamide injection induces partial destruction of beta
cells, thereby preventing beta cell compensation for 7 d HFD-induced insulin resistance but still
maintaining basal insulin levels, consequently leading to fasting hyperglycemia secondary to an
elevation of hepatic glucose production179,180. Other rodent models of diabetes including the
Zucker diabetic fatty and Goto-Kakisaki diabetic rats as well as the db/db mice have their
diabetic characteristics confounded by a rise in glucocorticoid levels, which are not represented
in the hormonal profile of type 2 diabetic patients, and therefore are not accurate models of
clinical diabetes. In comparison, the plasma corticosterone levels are not increased in the
STZ/Nic/HFD model of hyperglycemic rats191. Importantly, we showed that 7d STZ/Nic/HFD
94
diabetic rats are glucose intolerant and that infusion of GlyT1 inhibitor into the DVC leads to
improvement of glucose tolerance.
Further, our study indicates that DVC GlyT1 inhibition could also be effective as an early
intervention for insulin resistance and pre-diabetic conditions. While the 3d HFD-model we used
in Study 2 is not a diabetic model, it is a well-established model for hepatic insulin resistance
validated under hyperinsulinemic pancreatic clamp conditions28,227-229. In fact, a recent study
from our own lab testing the insulin sensitizing effects of resveratrol had validated 3d HFD rats
to be insulin resistant in regulating glucose production under hyperinsulinemic clamp conditions,
which could be reversed upon resveratrol infusion into the duodenum179. Whether GlyT1
inhibitors, like resveratrol, sensitizes insulin to regulate glucose production warrants
investigation; but under basal-insulin clamp conditions, our findings conclusively show that
DVC GlyT1 inhibition directly lowers glucose production independent of changes in basal
insulin levels. Further, based on the observation that 3d HFD rats display hyperphagia, a key
driving force of obesity, our data also implicate that DVC GlyT1 inhibition treatments are
effective in regulating glucose production in conditions of early onset diet-induced obesity as
well. However, instead of solely relying on an early onset obesity model (which is not in fact
obese), we also used a 28-d HFD obese model to show that DVC GlyT1 inhibition is effective in
regulating glucose production even at a later stage in diet-induced obesity. Previously, these 28d
HFD obese rats failed to show suppression of glucose production and stimulation of glucose
uptake compared to regular chow-fed rats under insulin-stimulated conditions (i.e.,
hyperinsulinemic clamps)179, thereby indicating hepatic and peripheral insulin resistance.
The therapeutic relevance of GlyT1 inhibition was further corroborated by our data
showing that systemic administration of GlyT1 inhibitors result in similar desirable metabolic
95
effects to those of direct DVC GlyT1 inhibition. However, given that GlyT1 and NMDA
receptors are also present in non-neural tissues such as the pancreas, the possible side effects of
using GlyT1 inhibitors and/ or increasing glycine systemically will have to be assessed
cautiously. A previous study reports that inhibition of NMDA receptor transmission in the islets
enhances glucose-stimulated insulin secretion205. It is possible then, because glycine is a co-
agonist of the NMDA receptor, that systemic infusion of GlyT1 inhibitors and thus, elevated
systemic glycine levels would have the undesirable side effect of enhancing NMDA receptor
transmission in the islets and consequently lead to reduced insulin secretion- an effect especially
detrimental for Type 2 diabetic patients. Interestingly though, multiple studies have documented
circulating glycine levels to be inversely associated with Type 2 diabetes risk230-232. A recent
study has in fact validated that glycine treatment results in increased insulin secretion from intact
human islets, and a disruption in this glycine-insulin action contributes to impaired insulin
secretion in Type 2 diabetes233. Of particular note, the effect of glycine to stimulate insulin
secretion was mediated via activation of the glycine receptors (GlyR) in the pancreatic islets
since antagonism of GlyR with strynchnine-prevented glycine induced insulin secretion.
Whether elevation of endogenous glycine levels in the pancreas during systemic GlyT1
inhibition could potentiate NMDA receptors to alter insulin secretion merits future investigation.
The use of divergent and sometimes common signaling pathways by circulating hormones
and nutrients in the CNS to regulate glucose and energy homeostasis.
Recent studies have highlighted this theme that a commonly derived pathway for CNS
hormonal action could regulate both glucose and energy homeostasis. As described in Chapter 1,
insulin action in the MBH and the DVC regulates both energy and glucose homeostasis234. In the
MBH, it does via a PI3K-dependent pathway to regulate glucose and feeding control, whereas in
96
the DVC, insulin signals through an ERK-dependent pathway to regulate feeding, body weight
and glucose production122. In regards to CNS glucagon action, a recent study by Quiñones et al.
reported that hypothalamic glucagon requires glucagon receptor and activation of downstream
PKA in the hypothalamic ARC to lower food intake, since icv. co-infusion of the glucagon
receptor antagonist des-His1-[Glu9] glucagon amide or the PKA inhibitor H-89 negated the
ability of central glucagon to decrease feeding235. Of note, these were the same chemical
inhibitors employed to confirm the role of glucagon receptor and PKA in mediating MBH
glucagon action on glucose control79, thereby suggesting that any associated molecular players in
the glucagon receptor–PKA branch are likely to be a part of a common pathway for both feeding
and glucose regulation by CNS glucagon. Moreover, Quiñones et al. also reported changes in
AgRP expression associated with the anorectic action of central glucagon, again consistent with
the observation that MBH glucagon receptors co-localised with AgRP neurons, thereby
demonstrating that in brain glucagon action, AgRP neurons mediate both glucose and feeding
regulation.
However, it appears that not the entire signaling pathway converges for the glycaemic
and satiety effects of hypothalamic glucagon action. Contrary to how MBH glucagon signals to
exert glucose control independent of MBH AMPK as shown in Study 1, the suppressive effect
central glucagon exerts on feeding involved inhibition of AMPK and activation of the
downstream target ACC. Specifically, molecular activation of AMPK in the ARC via injection
of a constitutively active AMPK virus blunted the anorectic effects of central glucagon
injections, whereas we showed that activation of MBH AMPK via the same viral approach had
no effect on the glucose production-lowering effect of MBH glucagon. Consistent with this,
there were decreased AMPK and increased ACC in the ARC associated with the satiety effect of
central glucagon injections, whereas we reported that the glucose-lowering effect of MBH
97
glucagon infusions was associated with no differences in pACC/total ACC levels in the MBH. It
is likely these distinct mechanisms by which brain glucagon acts potentially allows for selective
and independent control of glucose and feeding regulation. It is also possible that there are
different populations of glucagon-responsive neurons within the MBH in which glucagon
signaling pathways might be distinct for glucose and energy regulation, and this warrants future
investigation. However, it cannot be overlooked that perhaps the differences in glucagon dose
and administration—a single bolus i.c.v. glucagon injection at a dose of 480 ng for feeding vs 2 h
of constant MBH glucagon infusion with a much lower dose of 3.6 pg for glucose control —
could explain some of the differences in the regulation of molecular targets for feeding and
glucose regulation by brain glucagon. Alternatively, these findings could be due to differences in
the times at which the glucagon-treated tissues were obtained for molecular analysis: changes in
AMPK and ACC, which mediate the anorectic action of hypothalamic glucagon, were measured
1 h after the i.c.v. glucagon single bolus injection vs after 2 h of constant infusion of MBH
glucagon.
Study 2 describes that DVC GlyT1 inhibition plays a role in glucose tolerance and
glucose production in response to elevated extracellular glycine levels and activated NMDA
receptors. Although the involvement of NMDA receptors in the food intake and body weight
regulation by DVC GlyT1 inhibition was not directly shown in our study, we are encouraged by
other reports that show that blockade of NMDA receptors in the DVC enhances feeding236, and
that activation of DVC NMDA receptors is also required for CCK’s vagally mediated
suppression of food intake143,237. Thus, by postulation, DVC NMDA receptors act as the common
downstream mediator for DVC GlyT1 inhibitory control of glucose and energy homeostasis.
Notably, though, other studies have shown the existence of differential mechanisms of CNS
nutrient sensing regulating glucose and energy balance. For instance, activation of the mTOR
98
pathway is required for the inhibitory effect of MBH leucine on food intake111, while the
suppressive effect of MBH leucine on glucose production is independent of mTOR113. Whether
mechanisms of DVC GlyT1 inhibition (and DVC glycine sensing) diverge downstream of
NMDA receptors to allow for selective and autonomous control of feeding and glucose
regulation remain to be investigated.
In the pursuit of new pharmacological targets that would treat diabetes and obesity, in
order to design molecules where the main goal is to curb both hyperglycemia and hyperphagia it
becomes important to identify the point of convergence in these CNS fuel-sensing mechanisms
that control both glucose and energy homeostasis. However, it becomes equally important to
distinguish the point of divergence when the goal is to target one homeostatic regulation but not
the other. For instance, a lean diabetic individual would not require lowering body weight, as
opposed to reducing his glucose levels. In the same line, an obese non-diabetic individual would
only want to improve his energy balance as opposed to glucose regulation. Nonetheless, our
studies along with others’ point to the finding that hormonal action and nutrient sensing in the
CNS act via common as well as distinct signaling mechanisms to mediate glucose and energy
metabolism.
4.3 Limitations and Future directions Lack of a direct electrophysiological assessment confirming that MBH KATP channels were
inhibited by our molecular and pharmacological approaches is a major limitation in Study 1 that
concludes MBH glucagon activates KATP channels to lower glucose production. Given that
blockade of CNS KATP channels leads to membrane depolarization and increased electrical
activity, we acknowledge performing complementary patch-clamp studies to record changes in
membrane potential and firing activity, presumably, in AgRP neurons of MBH slices in response
99
to DN Kir6.2 (vs GFP) or glibenclamide (vs saline) with glucagon treatment would have greatly
strengthened our findings. An important future direction of Study 1 would also be to show
evidence for a direct PKA-mediated phosphorylation of KATP channels in the hypothalamus. It
has been shown that Kir6.2 can be phosphorylated by PKA at S372 and at S1571 for SUR1 in
pancreatic beta cells164. We could transfect hypothalamic cells lines (e.g. GT1-7, known to
express AgRP) with either wild type Kir6.2 or mutant Kir6.2 cDNA (mutated at S372) together
with either wild type SUR1 or mutant SUR1 cDNA (mutated at S1571), and test their ability to
be phosphorylated after PKA stimulation or glucagon treatment using in vitro phosphorylation
assays.
Further, we also did not measure malonyl- and LCFA-CoA levels in Study 1 that
concludes MBH glucagon works through a lipid-sensing independent mechanism. Interestingly
though, studies show activating or inhibiting the AMPK -> malonly-CoA sensing pathway does
not always translate in changes in LCFA-CoA levels in the hypothalamus55. Exogenous leptin
administration has been shown to increase the levels of malonyl-CoA level without subsequently
affecting the LCFA-CoA levels in the ARC54 whereas ghrelin signaling in the hypothalamus is
known to increase LCFA-CoA levels while inhibiting hypothalamic ACC, which reduces the
malonyl-CoA level65. These studies cast doubt on the reliability of malonyl-CoA and LCFA-
CoA levels as readout for lipid sensing activation. Perhaps then, the strength of Study 1 ought to
be the fact that we did not solely focus on blocking the beginning of the lipid sensing pathway,
we also showed neither inhibition of MBH PKC-δ (blocking the end of the lipid sensing
pathway) affected MBH glucagon, thereby ruling out a lipid sensing dependent mechanism for
MBH glucagon’s effect of glucose homeostasis.
The principal findings of Study 1 were generated using a single manipulation of glucose
homeostasis, the pancreatic basal-insulin clamp. Though closer to physiological conditions than
100
the hyperinsulinemic-euglycemic clamp, our findings would be even more convincing in light of
additional methods. Specifically, what is the relevance of this system in the post-prandial state,
where multiple systems are acting on glucose homeostasis in concert? Performing iv GTTs and
mixed-meal tolerance tests would provide further insight and confirmation into the MBH
glucagon signaling axis.
Given the work of polyagonist therapeutic studies targeting glucagon in conjunction with
GLP-1 and GIP receptors in the management of obesity and diabetes223, another important
question is whether concurrent infusions of glucagon, GLP-1 and GIP into the MBH would
result in redundant, additive or synergistic effects in lowering glucose production. This will
begin addressing whether the CNS penetrance of these polyagonists and whether activation of
MBH glucagon signalling plays a role in counteracting the hyperglycemic effects of peripheral
glucagon in these polyagonist therapeutic strategies.
We have identified, for the first time, that DVC GlyT1 inhibition and thus, DVC glycine
sensing plays a critical role in the regulation of energy balance as defined by changes in food
intake and body weight. However, components of energy expenditure (i.e. physical activity and
thermogenesis) are just as important on the energy balance equation. Given that obesity develops
when energy intake exceeds energy expenditure, the goal of any anti-obesity treatment should be
to reduce energy intake, promote energy expenditure, or both. Therefore, to truly repurpose
GlyT1 inhibitors as an effective obesity therapy, it becomes critical to determine the effect of
DVC GlyT1 inhibition on energy expenditure. Interestingly, the DVC as well as the NMDA
receptors in the DVC has directly been implicated in promoting the thermogenic activity of the
brown adipose tissue238,239. For instance, activation of DVC NMDA receptors is shown to
mediate the effect of increasing brown adipose thermogenesis in response to lipid infusion into
the duodenum239. In light of this, we would expect treatment with DVC GlyT1 inhibitors would
101
sufficiently activate NMDA receptors in the DVC to increase energy expenditure. Indeed, to
confirm this, our next step would be to monitor rats treated with DVC ALX (vs. saline) and/or
injected with LV-GlyT1 shRNA (vs. LV-MM) using metabolic chambers, as described
previously240.
It should be noted that most of the viral vectors we used to modulate various molecular
targets in our studies including AMPK activity or GlyT1 expression were under the ubiquitous
promoter, cytomegalovirus (CMV), thereby altering gene expression in non-target cells which is
a limitation in our work. In the future, we could use cell-specific promoters such as the neuron
specific enolase (NSE) or glial fibrillary acidic protein (GFAP) to alter our molecular targets
specifically in neuronal cells versus glial cells, respectively241.
With state-of the-art technologies that allow real-time manipulation of genetically defined
neuronal populations, an important future direction would be to map out the hypothalamic and
neuronal circuitry in the DVC involved in the regulation of glucose and energy homeostasis. As
an example, NTS neurons are known to excite brain regions such as the lateral parabrachial
nucleus (PBN) to modulate feeding behaviour. Recently, distinct population of neurons in the
PBN expressing the neuropeptide calcitonin gene-related protein (CGRP) (CGRPPBN) were
identified to lower feeding242,243. It is likely that lateral PBN neurons or CGRPPBN neurons would
be a downstream target mediating the effects of DVC GlyT1 inhibition to lower feeding and
body weight gain, especially given that DVC GlyT1 inhibition activates NMDA receptors and
that glutamatergic signalling activates CGRPPBN neurons244. Interesting future direction would be
to check for c-fos labeling in CGRPPBN neurons following DVC ALX treatments as well as to
check whether optogenetically inactivating CGRPPBN neurons (using the inhibitory
channelrhodopsin protein construct) would abolish the anorectic effects of DVC GlyT1
inhibition.
102
By contrast, our understanding of the glucoregulatory regulatory neurocircuits is far less
defined. However, a recent study employing optogenetic circuitry mapping approaches, revealed
that activating AgRP → LHA projections as well as activation of AgRP → anterior bed nucleus
of the stria terminalis (aBNST)vl projections impair systemic insulin sensitivity245. Clearly, future
studies are needed to know whether these neuronal circuits underlie MBH glucagon or DVC
GlyT1’s ability to alter peripheral glucose homeostasis.
103
References
1 Bernard, C. Leçons sur les Phenonemes de la Vie Commune aux Animaux et aux Vegetaux. (Bailliere et Fils, 1879).
2 Spanswick, D., Smith, M. A., Mirshamsi, S., Routh, V. H. & Ashford, M. L. Insulin activates ATP-sensitive K+ channels in hypothalamic neurons of lean, but not obese rats. Nat. Neurosci. 3, 757-758, doi:10.1038/77660 (2000).
3 Levin, B. E. & Keesey, R. E. Defense of differing body weight set points in diet-induced obese and resistant rats. Am. J. Physiol. 274, R412-419 (1998).
4 Woods, S. C., Seeley, R. J., Rushing, P. A., D'Alessio, D. & Tso, P. A controlled high-fat diet induces an obese syndrome in rats. J. Nutr. 133, 1081-1087 (2003).
5 Van Wye, G., Dubin, J. A., Blair, S. N. & Dipietro, L. Adult obesity does not predict 6-year weight gain in men: the Aerobics Center Longitudinal Study. Obesity (Silver Spring) 15, 1571-1577, doi:10.1038/oby.2007.186 (2007).
6 Beagley, J., Guariguata, L., Weil, C. & Motala, A. A. Global estimates of undiagnosed diabetes in adults. Diabetes Res. Clin. Pract. 103, 150-160, doi:10.1016/j.diabres.2013.11.001 (2014).
7 Mendis, S., Davis, S. & Norrving, B. Organizational update: the world health organization global status report on noncommunicable diseases 2014; one more landmark step in the combat against stroke and vascular disease. Stroke 46, e121-122, doi:10.1161/STROKEAHA.115.008097 (2015).
8 Ng, M. et al. Global, regional, and national prevalence of overweight and obesity in children and adults during 1980-2013: a systematic analysis for the Global Burden of Disease Study 2013. Lancet 384, 766-781, doi:10.1016/S0140-6736(14)60460-8 (2014).
9 Association, C. D. An economic tsunami: the cose of diabetes in Canada. (2009).
10 Anis, A. H. et al. Obesity and overweight in Canada: an updated cost-of-illness study. Obes. Rev. 11, 31-40, doi:10.1111/j.1467-789X.2009.00579.x (2010).
11 Roh, E., Song do, K. & Kim, M. S. Emerging role of the brain in the homeostatic regulation of energy and glucose metabolism. Exp. Mol. Med. 48, e216, doi:10.1038/emm.2016.4 (2016).
12 Chen, M., Woods, S. C. & Porte, D., Jr. Effect of cerebral intraventricular insulin on pancreatic insulin secretion in the dog. Diabetes 24, 910-914 (1975).
13 Woods, S. C., Lotter, E. C., McKay, L. D. & Porte, D., Jr. Chronic intracerebroventricular infusion of insulin reduces food intake and body weight of baboons. Nature 282, 503-505 (1979).
104
14 Harai, S., Ikenaga, T. & Matsuzawa, T. Nasal absorption of insulin in dogs. Diabetes 27, 296-299 (1978).
15 Anand Kumar, T. C., Sehgal A. & Bajaj, J. S. Intranasal administration of insulin reduces circulating levels of glucose in the Rhesus monkey. Vol. 49 (Current Science, 1980).
16 Obici, S., Feng, Z., Karkanias, G., Baskin, D. G. & Rossetti, L. Decreasing hypothalamic insulin receptors causes hyperphagia and insulin resistance in rats. Nat. Neurosci. 5, 566-572, doi:10.1038/nn861 (2002).
17 Gelling, R. W. et al. Insulin action in the brain contributes to glucose lowering during insulin treatment of diabetes. Cell Metab. 3, 67-73, doi:10.1016/j.cmet.2005.11.013 (2006).
18 Pocai, A. et al. Hypothalamic K(ATP) channels control hepatic glucose production. Nature 434, 1026-1031, doi:10.1038/nature03439 (2005).
19 Inoue, H. et al. Role of hepatic STAT3 in brain-insulin action on hepatic glucose production. Cell Metab. 3, 267-275, doi:10.1016/j.cmet.2006.02.009 (2006).
20 Kimura, K. et al. Central Insulin Action Activates Kupffer Cells by Suppressing Hepatic Vagal Activation via the Nicotinic Alpha 7 Acetylcholine Receptor. Cell Rep 14, 2362-2374, doi:10.1016/j.celrep.2016.02.032 (2016).
21 Konner, A. C. et al. Insulin action in AgRP-expressing neurons is required for suppression of hepatic glucose production. Cell Metab. 5, 438-449, doi:10.1016/j.cmet.2007.05.004 (2007).
22 Lin, H. V. et al. Divergent regulation of energy expenditure and hepatic glucose production by insulin receptor in agouti-related protein and POMC neurons. Diabetes 59, 337-346, doi:10.2337/db09-1303 (2010).
23 Okamoto, H. et al. Transgenic rescue of insulin receptor-deficient mice. J. Clin. Invest. 114, 214-223, doi:10.1172/JCI21645 (2004).
24 Carvalheira, J. B. et al. Selective impairment of insulin signalling in the hypothalamus of obese Zucker rats. Diabetologia 46, 1629-1640, doi:10.1007/s00125-003-1246-x (2003).
25 De Souza, C. T. et al. Consumption of a fat-rich diet activates a proinflammatory response and induces insulin resistance in the hypothalamus. Endocrinology 146, 4192-4199, doi:10.1210/en.2004-1520 (2005).
26 Posey, K. A. et al. Hypothalamic proinflammatory lipid accumulation, inflammation, and insulin resistance in rats fed a high-fat diet. Am. J. Physiol. Endocrinol. Metab. 296, E1003-1012, doi:10.1152/ajpendo.90377.2008 (2009).
105
27 Begg, D. P. et al. Reversal of diet-induced obesity increases insulin transport into cerebrospinal fluid and restores sensitivity to the anorexic action of central insulin in male rats. Endocrinology 154, 1047-1054, doi:10.1210/en.2012-1929 (2013).
28 Ono, H. et al. Activation of hypothalamic S6 kinase mediates diet-induced hepatic insulin resistance in rats. J. Clin. Invest. 118, 2959-2968, doi:10.1172/JCI34277 (2008).
29 Purkayastha, S. et al. Neural dysregulation of peripheral insulin action and blood pressure by brain endoplasmic reticulum stress. Proc. Natl. Acad. Sci. U. S. A. 108, 2939-2944, doi:10.1073/pnas.1006875108 (2011).
30 Meek, T. H. & Morton, G. J. Leptin, diabetes, and the brain. Indian J. Endocrinol. Metab. 16, S534-542, doi:10.4103/2230-8210.105568 (2012).
31 Liu, L. et al. Intracerebroventricular leptin regulates hepatic but not peripheral glucose fluxes. J. Biol. Chem. 273, 31160-31167 (1998).
32 Kamohara, S., Burcelin, R., Halaas, J. L., Friedman, J. M. & Charron, M. J. Acute stimulation of glucose metabolism in mice by leptin treatment. Nature 389, 374-377, doi:10.1038/38717 (1997).
33 Campfield, L. A., Smith, F. J., Guisez, Y., Devos, R. & Burn, P. Recombinant mouse OB protein: evidence for a peripheral signal linking adiposity and central neural networks. Science 269, 546-549 (1995).
34 Asilmaz, E. et al. Site and mechanism of leptin action in a rodent form of congenital lipodystrophy. J. Clin. Invest. 113, 414-424, doi:10.1172/JCI19511 (2004).
35 Lin, C. Y., Higginbotham, D. A., Judd, R. L. & White, B. D. Central leptin increases insulin sensitivity in streptozotocin-induced diabetic rats. Am. J. Physiol. Endocrinol. Metab. 282, E1084-1091, doi:10.1152/ajpendo.00489.2001 (2002).
36 Hidaka, S. et al. Chronic central leptin infusion restores hyperglycemia independent of food intake and insulin level in streptozotocin-induced diabetic rats. FASEB J. 16, 509-518 (2002).
37 Coppari, R. et al. The hypothalamic arcuate nucleus: a key site for mediating leptin's effects on glucose homeostasis and locomotor activity. Cell Metab. 1, 63-72, doi:10.1016/j.cmet.2004.12.004 (2005).
38 Morton, G. J. et al. Leptin regulates insulin sensitivity via phosphatidylinositol-3-OH kinase signaling in mediobasal hypothalamic neurons. Cell Metab. 2, 411-420, doi:10.1016/j.cmet.2005.10.009 (2005).
39 Satoh, N. et al. The arcuate nucleus as a primary site of satiety effect of leptin in rats. Neurosci. Lett. 224, 149-152, doi:10.1016/S0304-3940(97)00163-8 (1997).
40 Bates, S. H. et al. STAT3 signalling is required for leptin regulation of energy balance but not reproduction. Nature 421, 856-859, doi:10.1038/nature01388 (2003).
106
41 Bates, S. H. et al. LRb-STAT3 signaling is required for the neuroendocrine regulation of energy expenditure by leptin. Diabetes 53, 3067-3073 (2004).
42 Gao, Q. et al. Disruption of neural signal transducer and activator of transcription 3 causes obesity, diabetes, infertility, and thermal dysregulation. Proc. Natl. Acad. Sci. U. S. A. 101, 4661-4666, doi:10.1073/pnas.0303992101 (2004).
43 Gong, L. et al. Signal transducer and activator of transcription-3 is required in hypothalamic agouti-related protein/neuropeptide Y neurons for normal energy homeostasis. Endocrinology 149, 3346-3354, doi:10.1210/en.2007-0945 (2008).
44 Xu, A. W., Ste-Marie, L., Kaelin, C. B. & Barsh, G. S. Inactivation of signal transducer and activator of transcription 3 in proopiomelanocortin (Pomc) neurons causes decreased pomc expression, mild obesity, and defects in compensatory refeeding. Endocrinology 148, 72-80, doi:10.1210/en.2006-1119 (2007).
45 Vong, L. et al. Leptin action on GABAergic neurons prevents obesity and reduces inhibitory tone to POMC neurons. Neuron 71, 142-154, doi:10.1016/j.neuron.2011.05.028 (2011).
46 Buettner, C. et al. Critical role of STAT3 in leptin's metabolic actions. Cell Metab. 4, 49-60, doi:10.1016/j.cmet.2006.04.014 (2006).
47 Rahmouni, K., Sigmund, C. D., Haynes, W. G. & Mark, A. L. Hypothalamic ERK mediates the anorectic and thermogenic sympathetic effects of leptin. Diabetes 58, 536-542, doi:10.2337/db08-0822 (2009).
48 do Carmo, J. M. et al. Shp2 signaling in POMC neurons is important for leptin's actions on blood pressure, energy balance, and glucose regulation. Am. J. Physiol. Regul. Integr. Comp. Physiol. 307, R1438-1447, doi:10.1152/ajpregu.00131.2014 (2014).
49 Niswender, K. D. et al. Intracellular signalling. Key enzyme in leptin-induced anorexia. Nature 413, 794-795, doi:10.1038/35101657 (2001).
50 Hill, J. W. et al. Acute effects of leptin require PI3K signaling in hypothalamic proopiomelanocortin neurons in mice. J. Clin. Invest. 118, 1796-1805, doi:10.1172/JCI32964 (2008).
51 Hill, J. W. et al. Phosphatidyl inositol 3-kinase signaling in hypothalamic proopiomelanocortin neurons contributes to the regulation of glucose homeostasis. Endocrinology 150, 4874-4882, doi:10.1210/en.2009-0454 (2009).
52 Spanswick, D., Smith, M. A., Groppi, V. E., Logan, S. D. & Ashford, M. L. Leptin inhibits hypothalamic neurons by activation of ATP-sensitive potassium channels. Nature 390, 521-525, doi:10.1038/37379 (1997).
53 Minokoshi, Y. et al. AMP-kinase regulates food intake by responding to hormonal and nutrient signals in the hypothalamus. Nature 428, 569-574, doi:10.1038/nature02440 (2004).
107
54 Gao, S. et al. Leptin activates hypothalamic acetyl-CoA carboxylase to inhibit food intake. Proc. Natl. Acad. Sci. U. S. A. 104, 17358-17363, doi:10.1073/pnas.0708385104 (2007).
55 Gao, S. et al. Malonyl-CoA mediates leptin hypothalamic control of feeding independent of inhibition of CPT-1a. Am. J. Physiol. Regul. Integr. Comp. Physiol. 301, R209-217, doi:10.1152/ajpregu.00092.2011 (2011).
56 Widdowson, P. S., Upton, R., Buckingham, R., Arch, J. & Williams, G. Inhibition of food response to intracerebroventricular injection of leptin is attenuated in rats with diet-induced obesity. Diabetes 46, 1782-1785 (1997).
57 Pocai, A. et al. Central leptin acutely reverses diet-induced hepatic insulin resistance. Diabetes 54, 3182-3189 (2005).
58 Metlakunta, A. S., Sahu, M. & Sahu, A. Hypothalamic phosphatidylinositol 3-kinase pathway of leptin signaling is impaired during the development of diet-induced obesity in FVB/N mice. Endocrinology 149, 1121-1128, doi:10.1210/en.2007-1307 (2008).
59 Zigman, J. M., Jones, J. E., Lee, C. E., Saper, C. B. & Elmquist, J. K. Expression of ghrelin receptor mRNA in the rat and the mouse brain. J. Comp. Neurol. 494, 528-548, doi:10.1002/cne.20823 (2006).
60 Wang, Q. et al. Arcuate AgRP neurons mediate orexigenic and glucoregulatory actions of ghrelin. Mol Metab 3, 64-72, doi:10.1016/j.molmet.2013.10.001 (2014).
61 Lawrence, C. B., Snape, A. C., Baudoin, F. M. & Luckman, S. M. Acute central ghrelin and GH secretagogues induce feeding and activate brain appetite centers. Endocrinology 143, 155-162, doi:10.1210/endo.143.1.8561 (2002).
62 Stark, R. et al. Acyl ghrelin acts in the brain to control liver function and peripheral glucose homeostasis in male mice. Endocrinology 156, 858-868, doi:10.1210/en.2014-1733 (2015).
63 Kola, B. et al. Cannabinoids and ghrelin have both central and peripheral metabolic and cardiac effects via AMP-activated protein kinase. J. Biol. Chem. 280, 25196-25201, doi:10.1074/jbc.C500175200 (2005).
64 Lopez, M. et al. Hypothalamic fatty acid metabolism mediates the orexigenic action of ghrelin. Cell Metab. 7, 389-399, doi:10.1016/j.cmet.2008.03.006 (2008).
65 Andrews, Z. B. et al. UCP2 mediates ghrelin's action on NPY/AgRP neurons by lowering free radicals. Nature 454, 846-851, doi:10.1038/nature07181 (2008).
66 Jiang, G. & Zhang, B. B. Glucagon and regulation of glucose metabolism. Am. J. Physiol. Endocrinol. Metab. 284, E671-678, doi:10.1152/ajpendo.00492.2002 (2003).
108
67 Beale, E., Andreone, T., Koch, S., Granner, M. & Granner, D. Insulin and glucagon regulate cytosolic phosphoenolpyruvate carboxykinase (GTP) mRNA in rat liver. Diabetes 33, 328-332 (1984).
68 Lok, S. et al. The human glucagon receptor encoding gene: structure, cDNA sequence and chromosomal localization. Gene 140, 203-209 (1994).
69 Dunphy, J. L., Taylor, R. G. & Fuller, P. J. Tissue distribution of rat glucagon receptor and GLP-1 receptor gene expression. Mol. Cell. Endocrinol. 141, 179-186 (1998).
70 Hoosein, N. M. & Gurd, R. S. Identification of glucagon receptors in rat brain. Proc. Natl. Acad. Sci. U. S. A. 81, 4368-4372 (1984).
71 Wetsel, W. C., Eraly, S. A., Whyte, D. B. & Mellon, P. L. Regulation of gonadotropin-releasing hormone by protein kinase-A and -C in immortalized hypothalamic neurons. Endocrinology 132, 2360-2370, doi:10.1210/endo.132.6.8504741 (1993).
72 Graner, J. L. & Abraira, C. Glucagon in the cerebrospinal fluid. N. Engl. J. Med. 312, 994-995 (1985).
73 Inokuchi, A., Oomura, Y., Shimizu, N. & Yamamoto, T. Central action of glucagon in rat hypothalamus. Am. J. Physiol. 250, R120-126 (1986).
74 Marubashi, S. et al. Hyperglycaemic effect of glucagon administered intracerebroventricularly in the rat. Acta Endocrinol. (Copenh.) 108, 6-10 (1985).
75 Honda, K. et al. The mechanism underlying the central glucagon-induced hyperglycemia and anorexia in chicks. Comp. Biochem. Physiol. A Mol. Integr. Physiol. 163, 260-264, doi:10.1016/j.cbpa.2012.08.005 (2012).
76 Amir, S. Central glucagon-induced hyperglycemia is mediated by combined activation of the adrenal medulla and sympathetic nerve endings. Physiol. Behav. 37, 563-566 (1986).
77 Agarwala, G. C. & Bapat, S. K. Effect of centrally administered glucagon on blood glucose levels in dogs. Indian J. Med. Res. 66, 323-330 (1977).
78 Agarwala, G. C., Mishra, R., Jaiswal, G. & Bapat, V. Effect of centrally administered glucagon on liver glycogen & enzymes in anaesthetised dogs. Indian J. Med. Res. 90, 372-378 (1989).
79 Mighiu, P. I. et al. Hypothalamic glucagon signaling inhibits hepatic glucose production. Nat. Med. 19, 766-772, doi:10.1038/nm.3115 (2013).
80 Eigler, N., Sacca, L. & Sherwin, R. S. Synergistic interactions of physiologic increments of glucagon, epinephrine, and cortisol in the dog: a model for stress-induced hyperglycemia. J. Clin. Invest. 63, 114-123, doi:10.1172/JCI109264 (1979).
109
81 Felig, P., Wahren, J. & Hendler, R. Influence of physiologic hyperglucagonemia on basal and insulin-inhibited splanchnic glucose output in normal man. J. Clin. Invest. 58, 761-765, doi:10.1172/JCI108523 (1976).
82 Bomboy, J. D., Jr., Lewis, S. B., Lacy, W. W., Sinclair-Smith, B. C. & Liljenquist, J. E. Transient stimulatory effect of sustained hyperglucagonemia on splanchnic glucose production in normal and diabetic man. Diabetes 26, 177-174 (1977).
83 Muller, W. A., Faloona, G. R., Aguilar-Parada, E. & Unger, R. H. Abnormal alpha-cell function in diabetes. Response to carbohydrate and protein ingestion. N. Engl. J. Med. 283, 109-115, doi:10.1056/NEJM197007162830301 (1970).
84 Reaven, G. M., Chen, Y. D., Golay, A., Swislocki, A. L. & Jaspan, J. B. Documentation of hyperglucagonemia throughout the day in nonobese and obese patients with noninsulin-dependent diabetes mellitus. J. Clin. Endocrinol. Metab. 64, 106-110, doi:10.1210/jcem-64-1-106 (1987).
85 Johnson, D. G., Goebel, C. U., Hruby, V. J., Bregman, M. D. & Trivedi, D. Hyperglycemia of diabetic rats decreased by a glucagon receptor antagonist. Science 215, 1115-1116 (1982).
86 Brand, C. L. et al. Immunoneutralization of endogenous glucagon with monoclonal glucagon antibody normalizes hyperglycaemia in moderately streptozotocin-diabetic rats. Diabetologia 37, 985-993 (1994).
87 Djouder, N. et al. PKA phosphorylates and inactivates AMPKalpha to promote efficient lipolysis. EMBO J. 29, 469-481, doi:10.1038/emboj.2009.339 (2010).
88 Damm, E., Buech, T. R., Gudermann, T. & Breit, A. Melanocortin-induced PKA activation inhibits AMPK activity via ERK-1/2 and LKB-1 in hypothalamic GT1-7 cells. Mol. Endocrinol. 26, 643-654, doi:10.1210/me.2011-1218 (2012).
89 Inokuchi, A., Oomura, Y. & Nishimura, H. Effect of intracerebroventricularly infused glucagon on feeding behavior. Physiol. Behav. 33, 397-400 (1984).
90 Honda, K. et al. Central administration of glucagon suppresses food intake in chicks. Neurosci. Lett. 416, 198-201, doi:10.1016/j.neulet.2007.02.011 (2007).
91 Kurose, Y. et al. Effects of central administration of glucagon on feed intake and endocrine responses in sheep. Anim Sci J 80, 686-690, doi:10.1111/j.1740-0929.2009.00685.x (2009).
92 Geary, N., Kissileff, H. R., Pi-Sunyer, F. X. & Hinton, V. Individual, but not simultaneous, glucagon and cholecystokinin infusions inhibit feeding in men. Am. J. Physiol. 262, R975-980 (1992).
93 Morgan, K., Obici, S. & Rossetti, L. Hypothalamic responses to long-chain fatty acids are nutritionally regulated. J. Biol. Chem. 279, 31139-31148, doi:10.1074/jbc.M400458200 (2004).
110
94 Lam, T. K. et al. Hypothalamic sensing of circulating fatty acids is required for glucose homeostasis. Nat. Med. 11, 320-327, doi:10.1038/nm1201 (2005).
95 Rapoport, S. I. In vivo labeling of brain phospholipids by long-chain fatty acids: relation to turnover and function. Lipids 31 Suppl, S97-101 (1996).
96 Mitchell, R. W., Edmundson, C. L., Miller, D. W. & Hatch, G. M. On the mechanism of oleate transport across human brain microvessel endothelial cells. J. Neurochem. 110, 1049-1057, doi:10.1111/j.1471-4159.2009.06199.x (2009).
97 Le Foll, C., Irani, B. G., Magnan, C., Dunn-Meynell, A. A. & Levin, B. E. Characteristics and mechanisms of hypothalamic neuronal fatty acid sensing. Am. J. Physiol. Regul. Integr. Comp. Physiol. 297, R655-664, doi:10.1152/ajpregu.00223.2009 (2009).
98 Le Foll, C., Dunn-Meynell, A., Musatov, S., Magnan, C. & Levin, B. E. FAT/CD36: a major regulator of neuronal fatty acid sensing and energy homeostasis in rats and mice. Diabetes 62, 2709-2716, doi:10.2337/db12-1689 (2013).
99 Obici, S., Feng, Z., Arduini, A., Conti, R. & Rossetti, L. Inhibition of hypothalamic carnitine palmitoyltransferase-1 decreases food intake and glucose production. Nat. Med. 9, 756-761, doi:10.1038/nm873 (2003).
100 McGarry, J. D. What if Minkowski had been ageusic? An alternative angle on diabetes. Science 258, 766-770 (1992).
101 Yue, J. T. & Lam, T. K. Lipid sensing and insulin resistance in the brain. Cell Metab. 15, 646-655, doi:10.1016/j.cmet.2012.01.013 (2012).
102 Yang, C. S. et al. Hypothalamic AMP-activated protein kinase regulates glucose production. Diabetes 59, 2435-2443, doi:10.2337/db10-0221 (2010).
103 Obici, S. et al. Central administration of oleic acid inhibits glucose production and food intake. Diabetes 51, 271-275 (2002).
104 Pocai, A., Obici, S., Schwartz, G. J. & Rossetti, L. A brain-liver circuit regulates glucose homeostasis. Cell Metab. 1, 53-61, doi:10.1016/j.cmet.2004.11.001 (2005).
105 Ross, R. et al. Hypothalamic protein kinase C regulates glucose production. Diabetes 57, 2061-2065, doi:10.2337/db08-0206 (2008).
106 Pocai, A. et al. Restoration of hypothalamic lipid sensing normalizes energy and glucose homeostasis in overfed rats. J. Clin. Invest. 116, 1081-1091, doi:10.1172/JCI26640 (2006).
107 Lam, T. K., Gutierrez-Juarez, R., Pocai, A. & Rossetti, L. Regulation of blood glucose by hypothalamic pyruvate metabolism. Science 309, 943-947, doi:10.1126/science.1112085 (2005).
111
108 Wolfgang, M. J. et al. Regulation of hypothalamic malonyl-CoA by central glucose and leptin. Proc. Natl. Acad. Sci. U. S. A. 104, 19285-19290, doi:10.1073/pnas.0709778104 (2007).
109 Chari, M. et al. Glucose transporter-1 in the hypothalamic glial cells mediates glucose sensing to regulate glucose production in vivo. Diabetes 60, 1901-1906, doi:10.2337/db11-0120 (2011).
110 Chari, M., Lam, C. K., Wang, P. Y. & Lam, T. K. Activation of central lactate metabolism lowers glucose production in uncontrolled diabetes and diet-induced insulin resistance. Diabetes 57, 836-840, doi:10.2337/db07-1464 (2008).
111 Blouet, C., Jo, Y. H., Li, X. & Schwartz, G. J. Mediobasal hypothalamic leucine sensing regulates food intake through activation of a hypothalamus-brainstem circuit. J. Neurosci. 29, 8302-8311, doi:10.1523/JNEUROSCI.1668-09.2009 (2009).
112 Kanamori, K., Ross, B. D. & Kondrat, R. W. Rate of glutamate synthesis from leucine in rat brain measured in vivo by 15N NMR. J. Neurochem. 70, 1304-1315 (1998).
113 Su, Y. et al. Hypothalamic leucine metabolism regulates liver glucose production. Diabetes 61, 85-93, doi:10.2337/db11-0857 (2012).
114 Cota, D. et al. Hypothalamic mTOR signaling regulates food intake. Science 312, 927-930, doi:10.1126/science.1124147 (2006).
115 Arrieta-Cruz, I., Su, Y. & Gutierrez-Juarez, R. Suppression of Endogenous Glucose Production by Isoleucine and Valine and Impact of Diet Composition. Nutrients 8, 79, doi:10.3390/nu8020079 (2016).
116 Wang, P. Y. et al. Upper intestinal lipids trigger a gut-brain-liver axis to regulate glucose production. Nature 452, 1012-1016, doi:10.1038/nature06852 (2008).
117 Cheung, G. W., Kokorovic, A., Lam, C. K., Chari, M. & Lam, T. K. Intestinal cholecystokinin controls glucose production through a neuronal network. Cell Metab. 10, 99-109, doi:10.1016/j.cmet.2009.07.005 (2009).
118 Lam, T. K. Neuronal regulation of homeostasis by nutrient sensing. Nat. Med. 16, 392-395, doi:10.1038/nm0410-392 (2010).
119 Breen, D. M. et al. Jejunal nutrient sensing is required for duodenal-jejunal bypass surgery to rapidly lower glucose concentrations in uncontrolled diabetes. Nat. Med. 18, 950-955, doi:10.1038/nm.2745 (2012).
120 Cummings, D. E. & Overduin, J. Gastrointestinal regulation of food intake. J. Clin. Invest. 117, 13-23, doi:10.1172/JCI30227 (2007).
121 Lam, C. K., Chari, M., Rutter, G. A. & Lam, T. K. Hypothalamic nutrient sensing activates a forebrain-hindbrain neuronal circuit to regulate glucose production in vivo. Diabetes 60, 107-113, doi:10.2337/db10-0994 (2011).
112
122 Filippi, B. M., Yang, C. S., Tang, C. & Lam, T. K. Insulin activates Erk1/2 signaling in the dorsal vagal complex to inhibit glucose production. Cell Metab. 16, 500-510, doi:10.1016/j.cmet.2012.09.005 (2012).
123 Obici, S., Zhang, B. B., Karkanias, G. & Rossetti, L. Hypothalamic insulin signaling is required for inhibition of glucose production. Nat. Med. 8, 1376-1382, doi:10.1038/nm798 (2002).
124 Filippi, B. M. et al. Insulin signals through the dorsal vagal complex to regulate energy balance. Diabetes 63, 892-899, doi:10.2337/db13-1044 (2014).
125 Grill, H. J. et al. Evidence that the caudal brainstem is a target for the inhibitory effect of leptin on food intake. Endocrinology 143, 239-246, doi:10.1210/endo.143.1.8589 (2002).
126 Boghossian, S., Lecklin, A., Dube, M. G., Kalra, P. S. & Kalra, S. P. Increased leptin expression in the dorsal vagal complex suppresses adiposity without affecting energy intake and metabolic hormones. Obesity (Silver Spring) 14, 1003-1009, doi:10.1038/oby.2006.115 (2006).
127 Hansen, L. H., Abrahamsen, N. & Nishimura, E. Glucagon receptor mRNA distribution in rat tissues. Peptides 16, 1163-1166 (1995).
128 Parker, J. A. et al. Glucagon and GLP-1 inhibit food intake and increase c-fos expression in similar appetite regulating centres in the brainstem and amygdala. Int. J. Obes. (Lond.) 37, 1391-1398, doi:10.1038/ijo.2012.227 (2013).
129 LaPierre, M. P., Abraham, M. A., Yue, J. T., Filippi, B. M. & Lam, T. K. Glucagon signalling in the dorsal vagal complex is sufficient and necessary for high-protein feeding to regulate glucose homeostasis in vivo. EMBO Rep 16, 1299-1307, doi:10.15252/embr.201540492 (2015).
130 Mizuno, Y. & Oomura, Y. Glucose responding neurons in the nucleus tractus solitarius of the rat: in vitro study. Brain Res. 307, 109-116 (1984).
131 Balfour, R. H., Hansen, A. M. & Trapp, S. Neuronal responses to transient hypoglycaemia in the dorsal vagal complex of the rat brainstem. J. Physiol. 570, 469-484, doi:10.1113/jphysiol.2005.098822 (2006).
132 Ferreira, M., Jr. et al. Glucose effects on gastric motility and tone evoked from the rat dorsal vagal complex. J. Physiol. 536, 141-152 (2001).
133 Ritter, S., Dinh, T. T. & Zhang, Y. Localization of hindbrain glucoreceptive sites controlling food intake and blood glucose. Brain Res. 856, 37-47 (2000).
134 Blouet, C. & Schwartz, G. J. Brainstem nutrient sensing in the nucleus of the solitary tract inhibits feeding. Cell Metab. 16, 579-587, doi:10.1016/j.cmet.2012.10.003 (2012).
113
135 Cavanaugh, A. R., Schwartz, G. J. & Blouet, C. High-fat feeding impairs nutrient sensing and gut brain integration in the caudomedial nucleus of the solitary tract in mice. PLoS One 10, e0118888, doi:10.1371/journal.pone.0118888 (2015).
136 Cassell, M. D., Roberts, L. & Talman, W. T. Glycine-containing terminals in the rat dorsal vagal complex. Neuroscience 50, 907-920 (1992).
137 Paoletti, P. & Neyton, J. NMDA receptor subunits: function and pharmacology. Curr. Opin. Pharmacol. 7, 39-47, doi:10.1016/j.coph.2006.08.011 (2007).
138 Lam, C. K. et al. Activation of N-methyl-D-aspartate (NMDA) receptors in the dorsal vagal complex lowers glucose production. J. Biol. Chem. 285, 21913-21921, doi:10.1074/jbc.M109.087338 (2010).
139 Yue, J. T., Mighiu, P. I., Naples, M., Adeli, K. & Lam, T. K. Glycine normalizes hepatic triglyceride-rich VLDL secretion by triggering the CNS in high-fat fed rats. Circ. Res. 110, 1345-1354, doi:10.1161/CIRCRESAHA.112.268276 (2012).
140 Kleckner, N. W. & Dingledine, R. Requirement for glycine in activation of NMDA-receptors expressed in Xenopus oocytes. Science 241, 835-837 (1988).
141 Baptista, V. & Varanda, W. A. Glycine binding site of the synaptic NMDA receptor in subpostremal NTS neurons. J. Neurophysiol. 94, 147-152, doi:10.1152/jn.00927.2004 (2005).
142 Bergeron, R., Meyer, T. M., Coyle, J. T. & Greene, R. W. Modulation of N-methyl-D-aspartate receptor function by glycine transport. Proc. Natl. Acad. Sci. U. S. A. 95, 15730-15734 (1998).
143 Wright, J. et al. Reduction of food intake by cholecystokinin requires activation of hindbrain NMDA-type glutamate receptors. Am. J. Physiol. Regul. Integr. Comp. Physiol. 301, R448-455, doi:10.1152/ajpregu.00026.2011 (2011).
144 Noetzel, M. J., Jones, C. K. & Conn, P. J. Emerging approaches for treatment of schizophrenia: modulation of glutamatergic signaling. Discov. Med. 14, 335-343 (2012).
145 Kawai, N. et al. Pharmacokinetics and cerebral distribution of glycine administered to rats. Amino Acids 42, 2129-2137, doi:10.1007/s00726-011-0950-y (2012).
146 Heresco-Levy, U. et al. Efficacy of high-dose glycine in the treatment of enduring negative symptoms of schizophrenia. Arch. Gen. Psychiatry 56, 29-36 (1999).
147 Heresco-Levy, U., Ermilov, M., Lichtenberg, P., Bar, G. & Javitt, D. C. High-dose glycine added to olanzapine and risperidone for the treatment of schizophrenia. Biol. Psychiatry 55, 165-171 (2004).
148 Buchanan, R. W. et al. The Cognitive and Negative Symptoms in Schizophrenia Trial (CONSIST): the efficacy of glutamatergic agents for negative symptoms and cognitive
114
impairments. Am. J. Psychiatry 164, 1593-1602, doi:10.1176/appi.ajp.2007.06081358 (2007).
149 Betz, H., Gomeza, J., Armsen, W., Scholze, P. & Eulenburg, V. Glycine transporters: essential regulators of synaptic transmission. Biochem. Soc. Trans. 34, 55-58, doi:10.1042/BST0340055 (2006).
150 Black, M. D. et al. Procognitive and antipsychotic efficacy of glycine transport 1 inhibitors (GlyT1) in acute and neurodevelopmental models of schizophrenia: latent inhibition studies in the rat. Psychopharmacology (Berl.) 202, 385-396, doi:10.1007/s00213-008-1289-2 (2009).
151 Pinard, E. et al. Selective GlyT1 inhibitors: discovery of [4-(3-fluoro-5-trifluoromethylpyridin-2-yl)piperazin-1-yl][5-methanesulfonyl-2-(( S)-2,2,2-trifluoro-1-methylethoxy)phenyl]methanone (RG1678), a promising novel medicine to treat schizophrenia. J. Med. Chem. 53, 4603-4614, doi:10.1021/jm100210p (2010).
152 Boulay, D., Bergis, O., Avenet, P. & Griebel, G. The glycine transporter-1 inhibitor SSR103800 displays a selective and specific antipsychotic-like profile in normal and transgenic mice. Neuropsychopharmacology 35, 416-427, doi:10.1038/npp.2009.144 (2010).
153 D'Souza, D. C. et al. Glycine transporter inhibitor attenuates the psychotomimetic effects of ketamine in healthy males: preliminary evidence. Neuropsychopharmacology 37, 1036-1046, doi:10.1038/npp.2011.295 (2012).
154 Garcia-Caceres, C., Fuente-Martin, E., Argente, J. & Chowen, J. A. Emerging role of glial cells in the control of body weight. Mol Metab 1, 37-46, doi:10.1016/j.molmet.2012.07.001 (2012).
155 Garcia-Caceres, C. et al. Astrocytic Insulin Signaling Couples Brain Glucose Uptake with Nutrient Availability. Cell 166, 867-880, doi:10.1016/j.cell.2016.07.028 (2016).
156 Kim, J. G. et al. Leptin signaling in astrocytes regulates hypothalamic neuronal circuits and feeding. Nat. Neurosci. 17, 908-910, doi:10.1038/nn.3725 (2014).
157 Gao, Y. et al. Disruption of Lipid Uptake in Astroglia Exacerbates Diet Induced Obesity. Diabetes, doi:10.2337/db16-1278 (2017).
158 Pellerin, L. & Magistretti, P. J. Glutamate uptake into astrocytes stimulates aerobic glycolysis: a mechanism coupling neuronal activity to glucose utilization. Proc. Natl. Acad. Sci. U. S. A. 91, 10625-10629 (1994).
159 Kokorovic, A. et al. Duodenal mucosal protein kinase C-delta regulates glucose production in rats. Gastroenterology 141, 1720-1727, doi:10.1053/j.gastro.2011.06.042 (2011).
115
160 Baczko, I. et al. Plasma membrane KATP channel-mediated cardioprotection involves posthypoxic reductions in calcium overload and contractile dysfunction: mechanistic insights into cardioplegia. FASEB J. 19, 980-982, doi:10.1096/fj.04-3008fje (2005).
161 German, J. et al. Hypothalamic leptin signaling regulates hepatic insulin sensitivity via a neurocircuit involving the vagus nerve. Endocrinology 150, 4502-4511, doi:10.1210/en.2009-0445 (2009).
162 Sandoval, D. A., Bagnol, D., Woods, S. C., D'Alessio, D. A. & Seeley, R. J. Arcuate glucagon-like peptide 1 receptors regulate glucose homeostasis but not food intake. Diabetes 57, 2046-2054, doi:10.2337/db07-1824 (2008).
163 Kishore, P. et al. Activation of K(ATP) channels suppresses glucose production in humans. J. Clin. Invest. 121, 4916-4920, doi:10.1172/JCI58035 (2011).
164 Beguin, P., Nagashima, K., Nishimura, M., Gonoi, T. & Seino, S. PKA-mediated phosphorylation of the human K(ATP) channel: separate roles of Kir6.2 and SUR1 subunit phosphorylation. EMBO J. 18, 4722-4732, doi:10.1093/emboj/18.17.4722 (1999).
165 Global report on diabetes. World Health Organization. Geneva (2016).
166 Taylor, S. I. Deconstructing type 2 diabetes. Cell 97, 9-12 (1999).
167 Morton, G. J., Cummings, D. E., Baskin, D. G., Barsh, G. S. & Schwartz, M. W. Central nervous system control of food intake and body weight. Nature 443, 289-295, doi:10.1038/nature05026 (2006).
168 Duca, F. A., Bauer, P. V., Hamr, S. C. & Lam, T. K. Glucoregulatory Relevance of Small Intestinal Nutrient Sensing in Physiology, Bariatric Surgery, and Pharmacology. Cell Metab. 22, 367-380, doi:10.1016/j.cmet.2015.07.003 (2015).
169 Schwartz, M. W. et al. Cooperation between brain and islet in glucose homeostasis and diabetes. Nature 503, 59-66, doi:10.1038/nature12709 (2013).
170 Samuel, V. T. & Shulman, G. I. Mechanisms for insulin resistance: common threads and missing links. Cell 148, 852-871, doi:10.1016/j.cell.2012.02.017 (2012).
171 Sun, K., Kusminski, C. M. & Scherer, P. E. Adipose tissue remodeling and obesity. J. Clin. Invest. 121, 2094-2101, doi:10.1172/JCI45887 (2011).
172 Brashier, D. B., Sharma, A. K., Dahiya, N., Singh, S. K. & Khadka, A. Lorcaserin: A novel antiobesity drug. J. Pharmacol. Pharmacother. 5, 175-178, doi:10.4103/0976-500X.130158 (2014).
173 Buse, J. B. et al. Exenatide once weekly versus liraglutide once daily in patients with type 2 diabetes (DURATION-6): a randomised, open-label study. Lancet 381, 117-124, doi:10.1016/S0140-6736(12)61267-7 (2013).
116
174 Sisley, S. et al. Neuronal GLP1R mediates liraglutide's anorectic but not glucose-lowering effect. J. Clin. Invest. 124, 2456-2463, doi:10.1172/JCI72434 (2014).
175 Hallschmid, M., Higgs, S., Thienel, M., Ott, V. & Lehnert, H. Postprandial administration of intranasal insulin intensifies satiety and reduces intake of palatable snacks in women. Diabetes 61, 782-789, doi:10.2337/db11-1390 (2012).
176 Dash, S., Xiao, C., Morgantini, C., Koulajian, K. & Lewis, G. F. Intranasal insulin suppresses endogenous glucose production in humans compared with placebo in the presence of similar venous insulin concentrations. Diabetes 64, 766-774, doi:10.2337/db14-0685 (2015).
177 Heni, M. et al. Central insulin administration improves whole-body insulin sensitivity via hypothalamus and parasympathetic outputs in men. Diabetes 63, 4083-4088, doi:10.2337/db14-0477 (2014).
178 Lam, T. K., Schwartz, G. J. & Rossetti, L. Hypothalamic sensing of fatty acids. Nat. Neurosci. 8, 579-584, doi:10.1038/nn1456 (2005).
179 Cote, C. D. et al. Resveratrol activates duodenal Sirt1 to reverse insulin resistance in rats through a neuronal network. Nat. Med. 21, 498-505, doi:10.1038/nm.3821 (2015).
180 Duca, F. A. et al. Metformin activates a duodenal Ampk-dependent pathway to lower hepatic glucose production in rats. Nat. Med. 21, 506-511, doi:10.1038/nm.3787 (2015).
181 Cubelos, B., Gimenez, C. & Zafra, F. Localization of the GLYT1 glycine transporter at glutamatergic synapses in the rat brain. Cereb. Cortex 15, 448-459, doi:10.1093/cercor/bhh147 (2005).
182 Harvey, R. J. & Yee, B. K. Glycine transporters as novel therapeutic targets in schizophrenia, alcohol dependence and pain. Nat Rev Drug Discov 12, 866-885, doi:10.1038/nrd3893 (2013).
183 Porter, R. A., and Dawson, L.A. GlyT-1 Inhibitors: From Hits to Clinical Candidates. In Small Molecule Therapeutics for Schizophrenia, S. Celanire, and S. Poli, eds., 51-99 (Cham: Springer International Publishing, 2015).
184 Yue, J. T. et al. A fatty acid-dependent hypothalamic-DVC neurocircuitry that regulates hepatic secretion of triglyceride-rich lipoproteins. Nat Commun 6, 5970, doi:10.1038/ncomms6970 (2015).
185 Gillespie, B. R., Burns, G. A. & Ritter, R. C. NMDA channels control meal size via central vagal afferent terminals. Am. J. Physiol. Regul. Integr. Comp. Physiol. 289, R1504-1511, doi:10.1152/ajpregu.00169.2005 (2005).
186 Kemp, J. A. et al. 7-Chlorokynurenic acid is a selective antagonist at the glycine modulatory site of the N-methyl-D-aspartate receptor complex. Proc. Natl. Acad. Sci. U. S. A. 85, 6547-6550 (1988).
117
187 Woodward, J. J. A comparison of the effects of ethanol and the competitive glycine antagonist 7-chlorokynurenic acid on N-methyl-D-aspartic acid-induced neurotransmitter release from rat hippocampal slices. J. Neurochem. 62, 987-991 (1994).
188 Zhang, J. et al. Pharmacological characterization of [(3)H]CHIBA-3007 binding to glycine transporter 1 in the rat brain. PLoS One 6, e21322, doi:10.1371/journal.pone.0021322 (2011).
189 Nagy, K. et al. Alterations in brain extracellular dopamine and glycine levels following combined administration of the glycine transporter type-1 inhibitor Org-24461 and risperidone. Neurochem. Res. 35, 2096-2106, doi:10.1007/s11064-010-0241-0 (2010).
190 Chen, L., Muhlhauser, M. & Yang, C. R. Glycine tranporter-1 blockade potentiates NMDA-mediated responses in rat prefrontal cortical neurons in vitro and in vivo. J. Neurophysiol. 89, 691-703, doi:10.1152/jn.00680.2002 (2003).
191 Samuel, V. T. et al. Fasting hyperglycemia is not associated with increased expression of PEPCK or G6Pc in patients with Type 2 Diabetes. Proc. Natl. Acad. Sci. U. S. A. 106, 12121-12126, doi:10.1073/pnas.0812547106 (2009).
192 Atkinson, B. N. et al. ALX 5407: a potent, selective inhibitor of the hGlyT1 glycine transporter. Mol. Pharmacol. 60, 1414-1420 (2001).
193 Benarroch, E. E. NMDA receptors: recent insights and clinical correlations. Neurology 76, 1750-1757, doi:10.1212/WNL.0b013e31821b7cc9 (2011).
194 Kemp, J. A. & McKernan, R. M. NMDA receptor pathways as drug targets. Nat. Neurosci. 5 Suppl, 1039-1042, doi:10.1038/nn936 (2002).
195 Aroeira, R. I., Sebastiao, A. M. & Valente, C. A. GlyT1 and GlyT2 in brain astrocytes: expression, distribution and function. Brain Struct Funct 219, 817-830, doi:10.1007/s00429-013-0537-3 (2014).
196 Olivares, L., Aragon, C., Gimenez, C. & Zafra, F. The role of N-glycosylation in the targeting and activity of the GLYT1 glycine transporter. J. Biol. Chem. 270, 9437-9442 (1995).
197 Zafra, F., Gomeza, J., Olivares, L., Aragon, C. & Gimenez, C. Regional distribution and developmental variation of the glycine transporters GLYT1 and GLYT2 in the rat CNS. Eur. J. Neurosci. 7, 1342-1352 (1995).
198 Barrera, S. P. et al. PKC-Dependent GlyT1 Ubiquitination Occurs Independent of Phosphorylation: Inespecificity in Lysine Selection for Ubiquitination. PLoS One 10, e0138897, doi:10.1371/journal.pone.0138897 (2015).
199 Talman, W. T., Perrone, M. H. & Reis, D. J. Evidence for L-glutamate as the neurotransmitter of baroreceptor afferent nerve fibers. Science 209, 813-815 (1980).
118
200 Talman, W. T. & Robertson, S. C. Glycine, like glutamate, microinjected into the nucleus tractus solitarii of rat decreases arterial pressure and heart rate. Brain Res. 477, 7-13 (1989).
201 Chatterton, J. E. et al. Excitatory glycine receptors containing the NR3 family of NMDA receptor subunits. Nature 415, 793-798, doi:10.1038/nature715 (2002).
202 Fujikawa, T., Chuang, J. C., Sakata, I., Ramadori, G. & Coppari, R. Leptin therapy improves insulin-deficient type 1 diabetes by CNS-dependent mechanisms in mice. Proc. Natl. Acad. Sci. U. S. A. 107, 17391-17396, doi:10.1073/pnas.1008025107 (2010).
203 German, J. P. et al. Leptin activates a novel CNS mechanism for insulin-independent normalization of severe diabetic hyperglycemia. Endocrinology 152, 394-404, doi:10.1210/en.2010-0890 (2011).
204 Rasmussen, B. A. et al. Jejunal leptin-PI3K signaling lowers glucose production. Cell Metab. 19, 155-161, doi:10.1016/j.cmet.2013.11.014 (2014).
205 Marquard, J. et al. Characterization of pancreatic NMDA receptors as possible drug targets for diabetes treatment. Nat. Med. 21, 363-372, doi:10.1038/nm.3822 (2015).
206 Vancampfort, D. et al. A meta-analysis of cardio-metabolic abnormalities in drug naive, first-episode and multi-episode patients with schizophrenia versus general population controls. World Psychiatry 12, 240-250, doi:10.1002/wps.20069 (2013).
207 Harsing, L. G., Jr. et al. Glycine transporter type-1 and its inhibitors. Curr. Med. Chem. 13, 1017-1044 (2006).
208 Alberati, D. et al. Glycine reuptake inhibitor RG1678: a pharmacologic characterization of an investigational agent for the treatment of schizophrenia. Neuropharmacology 62, 1152-1161, doi:10.1016/j.neuropharm.2011.11.008 (2012).
209 Sun, H. S. & Feng, Z. P. Neuroprotective role of ATP-sensitive potassium channels in cerebral ischemia. Acta Pharmacol. Sin. 34, 24-32, doi:10.1038/aps.2012.138 (2013).
210 van den Top, M., Lee, K., Whyment, A. D., Blanks, A. M. & Spanswick, D. Orexigen-sensitive NPY/AgRP pacemaker neurons in the hypothalamic arcuate nucleus. Nat. Neurosci. 7, 493-494, doi:10.1038/nn1226 (2004).
211 Yang, S. B. et al. Rapamycin ameliorates age-dependent obesity associated with increased mTOR signaling in hypothalamic POMC neurons. Neuron 75, 425-436, doi:10.1016/j.neuron.2012.03.043 (2012).
212 Campos, C. A. & Ritter, R. C. NMDA-type glutamate receptors participate in reduction of food intake following hindbrain melanocortin receptor activation. Am. J. Physiol. Regul. Integr. Comp. Physiol. 308, R1-9, doi:10.1152/ajpregu.00388.2014 (2015).
213 Bach, E. C., Halmos, K. C. & Smith, B. N. Enhanced NMDA receptor-mediated modulation of excitatory neurotransmission in the dorsal vagal complex of
119
streptozotocin-treated, chronically hyperglycemic mice. PLoS One 10, e0121022, doi:10.1371/journal.pone.0121022 (2015).
214 Soundarapandian, M. M., Zhong, X., Peng, L., Wu, D. & Lu, Y. Role of K(ATP) channels in protection against neuronal excitatory insults. J. Neurochem. 103, 1721-1729, doi:10.1111/j.1471-4159.2007.04963.x (2007).
215 Hopper, R. A. & Garthwaite, J. Tonic and phasic nitric oxide signals in hippocampal long-term potentiation. J. Neurosci. 26, 11513-11521, doi:10.1523/JNEUROSCI.2259-06.2006 (2006).
216 Sattler, R. et al. Specific coupling of NMDA receptor activation to nitric oxide neurotoxicity by PSD-95 protein. Science 284, 1845-1848 (1999).
217 Snyder, S. H. Nitric oxide: first in a new class of neurotransmitters. Science 257, 494-496 (1992).
218 Shen, K. Z. & Johnson, S. W. Ca2+ influx through NMDA-gated channels activates ATP-sensitive K+ currents through a nitric oxide-cGMP pathway in subthalamic neurons. J. Neurosci. 30, 1882-1893, doi:10.1523/JNEUROSCI.3200-09.2010 (2010).
219 Stanley, B. G., Willett, V. L., 3rd, Donias, H. W., Dee, M. G., 2nd & Duva, M. A. Lateral hypothalamic NMDA receptors and glutamate as physiological mediators of eating and weight control. Am. J. Physiol. 270, R443-449 (1996).
220 Stanley, B. G., Ha, L. H., Spears, L. C. & Dee, M. G., 2nd. Lateral hypothalamic injections of glutamate, kainic acid, D,L-alpha-amino-3-hydroxy-5-methyl-isoxazole propionic acid or N-methyl-D-aspartic acid rapidly elicit intense transient eating in rats. Brain Res. 613, 88-95 (1993).
221 Liu, T. et al. Fasting activation of AgRP neurons requires NMDA receptors and involves spinogenesis and increased excitatory tone. Neuron 73, 511-522, doi:10.1016/j.neuron.2011.11.027 (2012).
222 Uner, A. et al. The role of GluN2A and GluN2B NMDA receptor subunits in AgRP and POMC neurons on body weight and glucose homeostasis. Mol Metab 4, 678-691, doi:10.1016/j.molmet.2015.06.010 (2015).
223 Finan, B. et al. A rationally designed monomeric peptide triagonist corrects obesity and diabetes in rodents. Nat. Med. 21, 27-36, doi:10.1038/nm.3761 (2015).
224 Pocai, A. et al. Glucagon-like peptide 1/glucagon receptor dual agonism reverses obesity in mice. Diabetes 58, 2258-2266, doi:10.2337/db09-0278 (2009).
225 Day, J. W. et al. A new glucagon and GLP-1 co-agonist eliminates obesity in rodents. Nat. Chem. Biol. 5, 749-757, doi:10.1038/nchembio.209 (2009).
226 Wagner, R. et al. Non-Suppressed Glucagon after Glucose Challenge as a Potential Predictor for Glucose Tolerance. Diabetes, doi:10.2337/db16-0354 (2016).
120
227 Kraegen, E. W. et al. Development of muscle insulin resistance after liver insulin resistance in high-fat-fed rats. Diabetes 40, 1397-1403 (1991).
228 Wang, J. et al. Overfeeding rapidly induces leptin and insulin resistance. Diabetes 50, 2786-2791 (2001).
229 Samuel, V. T. et al. Mechanism of hepatic insulin resistance in non-alcoholic fatty liver disease. J. Biol. Chem. 279, 32345-32353, doi:10.1074/jbc.M313478200 (2004).
230 Thalacker-Mercer, A. E. et al. BMI, RQ, diabetes, and sex affect the relationships between amino acids and clamp measures of insulin action in humans. Diabetes 63, 791-800, doi:10.2337/db13-0396 (2014).
231 Wang-Sattler, R. et al. Novel biomarkers for pre-diabetes identified by metabolomics. Mol. Syst. Biol. 8, 615, doi:10.1038/msb.2012.43 (2012).
232 Floegel, A. et al. Identification of serum metabolites associated with risk of type 2 diabetes using a targeted metabolomic approach. Diabetes 62, 639-648, doi:10.2337/db12-0495 (2013).
233 Yan-Do, R. et al. A glycine-insulin autocrine feedback loop enhances insulin secretion from human beta-cells and is impaired in type 2 diabetes. Diabetes, doi:10.2337/db15-1272 (2016).
234 Abraham, M. A. & Lam, T. K. Glucagon action in the brain. Diabetologia 59, 1367-1371, doi:10.1007/s00125-016-3950-3 (2016).
235 Quinones, M. et al. Hypothalamic CaMKKbeta mediates glucagon anorectic effect and its diet-induced resistance. Mol Metab 4, 961-970, doi:10.1016/j.molmet.2015.09.014 (2015).
236 Hung, C. Y., Covasa, M., Ritter, R. C. & Burns, G. A. Hindbrain administration of NMDA receptor antagonist AP-5 increases food intake in the rat. Am. J. Physiol. Regul. Integr. Comp. Physiol. 290, R642-651, doi:10.1152/ajpregu.00641.2005 (2006).
237 Ritter, R. C., Covasa, M. & Matson, C. A. Cholecystokinin: proofs and prospects for involvement in control of food intake and body weight. Neuropeptides 33, 387-399, doi:10.1054/npep.1999.0051 (1999).
238 Skibicka, K. P. & Grill, H. J. Hypothalamic and hindbrain melanocortin receptors contribute to the feeding, thermogenic, and cardiovascular action of melanocortins. Endocrinology 150, 5351-5361, doi:10.1210/en.2009-0804 (2009).
239 Blouet, C. & Schwartz, G. J. Duodenal lipid sensing activates vagal afferents to regulate non-shivering brown fat thermogenesis in rats. PLoS One 7, e51898, doi:10.1371/journal.pone.0051898 (2012).
121
240 Xu, Y. et al. PI3K signaling in the ventromedial hypothalamic nucleus is required for normal energy homeostasis. Cell Metab. 12, 88-95, doi:10.1016/j.cmet.2010.05.002 (2010).
241 Li, M. et al. Optimal promoter usage for lentiviral vector-mediated transduction of cultured central nervous system cells. J. Neurosci. Methods 189, 56-64, doi:10.1016/j.jneumeth.2010.03.019 (2010).
242 Wu, Q., Clark, M. S. & Palmiter, R. D. Deciphering a neuronal circuit that mediates appetite. Nature 483, 594-597, doi:10.1038/nature10899 (2012).
243 Roman, C. W., Derkach, V. A. & Palmiter, R. D. Genetically and functionally defined NTS to PBN brain circuits mediating anorexia. Nat Commun 7, 11905, doi:10.1038/ncomms11905 (2016).
244 Carter, M. E., Soden, M. E., Zweifel, L. S. & Palmiter, R. D. Genetic identification of a neural circuit that suppresses appetite. Nature 503, 111-114, doi:10.1038/nature12596 (2013).
245 Steculorum, S. M. et al. AgRP Neurons Control Systemic Insulin Sensitivity via Myostatin Expression in Brown Adipose Tissue. Cell 165, 125-138, doi:10.1016/j.cell.2016.02.044 (2016).
Malonyl-CoACPT-1
GlucoseProduc5on
Ro7lerin/DNPKC-δ
Glibenclamide/Kir6.2
LCFA-CoA PKC-δ? Sur1-Kir6.2KATPchannel?
AICAR/CAAMPKACC
Acetyl-CoA ?
MediobasalHypothalamusGlucagon
Glucagon Receptor
PKA
AMPK
Figure2.1
-90
-40
0
5
10
15
GFP+salineGFP+glucagonCAAMPK+salineCAAMPK+glucagon
0
2
4
6
8
024681012
0
5
10
0
2
4
6
8
-90
-40
A
B
C
D
E
F£basal
£basal nclampnclamp
H
MBHsaline
MBHglucagon
pACC(265kDa)
TotalACC(265kDa)
β-tubulin(55kDa)
00.10.20.30.40.50.60.7
MBHSaline MBHGlucagon
pAC
C/TotalACC
I* * * *
* * **
* * **
GlucoseIn
fusion
Rate
(mgkg
-1m
in-1)
GlucoseProdu
c5on
(m
gkg-1m
in-1)
GlucoseProdu
c5on
Supp
ression(%
)GlucoseUptake
(mgkg-1
min
-1)
GlucoseIn
fusion
Rate
(mgkg
-1m
in-1)
GlucoseProdu
c5on
(m
gkg-1m
in-1)
GlucoseProdu
c5on
Supp
ression(%
)GlucoseUptake
(mgkg-1
min
-1)
G
0
5
10
SalineGlucagonAICAR+salineAICAR+glucagon
+ – – –– + – –– – + –– – – +
SalineGlucagonAICAR+salineAICAR+glucagon
+ – – –– + – –– – + –– – – +
SalineGlucagonAICAR+salineAICAR+glucagon
+ – – –– + – –– – + –– – – +
SalineGlucagonAICAR+salineAICAR+glucagon
+ – – –– + – –– – + –– – – +
+ – – –– + – –– – + –– – – +
+ – – –– + – –– – + –– – – +
GFP+salineGFP+glucagonCAAMPK+salineCAAMPK+glucagon
GFP+salineGFP+glucagonCAAMPK+salineCAAMPK+glucagon
+ – – –– + – –– – + –– – – +
GFP+salineGFP+glucagonCAAMPK+salineCAAMPK+glucagon
+ – – –– + – –– – + –– – – +
250kDa
250kDa
50kDa
Figure2.2
0
0.2
0.4
0.6
0.8
Saline Glucagon OAG
*
LacZ+salineLacZ+glucagonDNPKCδ+salineDNPKCδ+glucagon
0
5
10
0
5
10
15
-100
-50
0
0
5
10
15
0
5
10
0
5
10
15
-100
-50
0
0
5
10
15
A
B
E
C
D
F
G
H
* * * *
* * * *
**
* *
GlucoseIn
fusion
Rate
(mgkg
-1m
in-1)
GlucoseProdu
c5on
(m
gkg-1m
in-1)
GlucoseProdu
c5on
Supp
ression(%
)GlucoseUptake
(mgkg-1
min
-1)
GlucoseIn
fusion
Rate
(mgkg
-1m
in-1)
GlucoseProdu
c5on
Supp
ression(%
)GlucoseUptake
(mgkg-1
min
-1)
SalineGlucagonRot+salineRot+glucagon
SalineGlucagonRot+salineRot+glucagon
SalineGlucagonRot+salineRot+glucagon
SalineGlucagonRot+salineRot+glucagon
+ – – –– + – –– – + –– – – +
+ – – –– + – –– – + –– – – +
+ – – –– + – –– – + –– – – +
+ – – –– + – –– – + –– – – +
+ – – –– + – –– – + –– – – +
GlucoseProdu
c5on
(m
gkg-1m
in-1)
+ – – –– + – –– – + –– – – +
+ – – –– + – –– – + –– – – +
+ – – –– + – –– – + –– – – +
LacZ+salineLacZ+glucagonDNPKCδ+salineDNPKCδ+glucagon
LacZ+salineLacZ+glucagonDNPKCδ+salineDNPKCδ+glucagon
LacZ+salineLacZ+glucagonDNPKCδ+salineDNPKCδ+glucagon
PKC-δac5v
ity
(pmol/m
in/protein)
I
Figure2.3
-100
-50
0
0
5
10
0
5
10
15
0
5
10
15
0
5
10
0
5
10
15
-100-80-60-40-200
A
B
C
D
E
F£basal
£basalnclamp
nclamp
0
5
10
G
H
**
**
* *
GlucoseInfusio
nRa
te
(mgkg
-1m
in-1)
GlucoseProd
ucMo
n(m
gkg
-1 min
-1)
GlucoseProdu
c5on
Supp
ression(%
)GlucoseUptake
(mgkg
-1m
in-1)
GlucoseInfusio
nRa
te
(mgkg
-1m
in-1)
GlucoseProd
ucMo
n(m
gkg
-1 min
-1)
GlucoseProdu
c5on
Supp
ression(%
)
GlucoseUptake
(mgkg
-1m
in-1)
SalineGlucagonGli+salineGli+glucagon
SalineGlucagonGli+salineGli+glucagon
SalineGlucagonGli+salineGli+glucagon
SalineGlucagonGli+salineGli+glucagon
+ – – –– + – –– – + –– – – +
+ – – –– + – –– – + –– – – +
+ – – –– + – –– – + –– – – +
+ – – –– + – –– – + –– – – +
+ – – –– + – –– – + –– – – +
+ – – –– + – –– – + –– – – +
+ – – –– + – –– – + –– – – +
GFP+salineGFP+glucagonDNKir6.2+salineDNKir6.2+glucagon
GFP+salineGFP+glucagonDNKir6.2+salineDNKir6.2+glucagon
GFP+salineGFP+glucagonDNKir6.2+salineDNKir6.2+glucagon
GFP+salineGFP+glucagonDNKir6.2+salineDNKir6.2+glucagon
+ – – –– + – –– – + –– – – +
Figure2.4
Figure3.1
0
2
4
6
8
10
12
14
0
1
2
3
4
5
6
0
0.001
0.002
0.003
0.004
0.005
Saline ALX0
2
4
6
8
10
12
0
1
2
3
4
5
6
100
120
140
160
180
200
220
240
260
basal 0 2 5 10
15
20
25
30
40
50
60
8100
8600
9100
£Basal nClamp
Glucoseprodu
c5on
(mg/kg/min)
D
*†
Plasmaglucose(m
g/dl)
0
1
2
3
4
basal 0 2 5 10
15
20
25
30
40
50
60
Plasmainsulin
(ng/ml)
DVCsalineDVCALX
DVCsalineDVCALX
Time(min)
AUC(0-60min)
*
††
††
†
Glucoseinfusion
rate(m
g/kg/min)
SalineSHAMALX
HVAG
+ – + –+ + – –– + – +– – + +
£Basal nClamp
Glucoseprodu
c5on
(mg/kg/min)
A B C
Glucoseinfusion
rate(m
g/kg/min)
+ – + – – – –– + – + + + +– – + + – – –– – – – + – –– – – – – + –– – – – – – +
+ – + – – – –– + – + + + +– – + + – – –– – – – + – –– – – – – + –– – – – – – +
E F
SalineALX
MK8017CKNAAd-MM
Ad-GluN1shRNA
SalineALX
MK8017CKNAAd-MM
Ad-GluN1shRNA
+ – + –+ + – –– + – +– – + +
* †
Time(min)
*
G
DVCextracellular
glycineconcen
tra5
on(n
mol/ul) £Baseline
nInfusion*
*
Figure3.2
0
0.4
0.8
1.2
GlyT1(5
5):IR
A B
0
1
2
―IR
70―100― ―GlyT190
―GlyT170
MW(kDa)
0
0.5
1
1.5
GlyT1(9
0):IR
GlyT1(7
0):IR
GlyT1(9
0):IR
DVCLV-GlyT1shRNA
100―
MW(kDa)
70―
100―
―GlyT170
―IR
―GlyT190
GlyT1(7
0):IR
DVCLV-MM
0
0.4
0.8
1.2
**
0
0.4
0.8
1.2
*
DVC sp5,Sp5C,Sp5I(L)
D
70―
―IR100―
―GlyT170
0
0.5
1
1.5
2
C
70― ―GlyT170
100― ―IR
0
0.5
1
1.5
2
sp5,Sp5C,Sp5I(R)MW(kDa)
MW(kDa)
0
1
2
3
4
5
6
7
8
Glucoseinfusion
rate(m
g/kg/min)
Glucoseprodu
c5on
(mg/kg/min)
+ – –– + +– – +
£Basal nClamp*
LV-MM + – –LV-GlyT1shRNA – + +MK801 – – +
E Fpy
0
2
4
6
8
10
12
*
100―
100―
DVCLV-MM + –DVCLV-GlyT1shRNA – +
+ –– +
+ –– +
+ –– +
100―
DVCLV-MM
DVCLV-GlyT1shRNA
DVCLV-GlyT1shRNA
DVCLV-MM
GlyT1(7
0):IR
+ –– +
DVCLV-MM
DVCLV-GlyT1shRNA
GlyT1(7
0):IR
―GlyT155
+ –– +
55―
DVCLV-MM + –DVCLV-GlyT1shRNA – +
Figure3.3
0
2
4
6
8
10
12
0
1
2
3
4
5
6
7
0
2
4
6
8
10
12
14
13
Glucoseinfusion
rate(m
g/kg/min)
Glucoseprodu
c5on
(mg/kg/min)
3dHFDSaline
GlycineALX
7CKNA
+ + + + ++ – – – –– + – + –– – + – +– – – + +
+ + + + ++ – – – –– + – + –– – + – +– – – + +
B C
£Basal nClamp
A
DVCinfusion:saline/glycine/ALX/glycine+7CKNA/ALX+7CKNA
SST(3µgkg-1min-1)
Insulin(1.2mUkg-1min-1)
-90min
Glucose(asneeded)
[3-3H]-glucose(0.4µCimin-1)
0 240
day0
DVCcannula5on Vascularcatheteriza5on Clamp
8
90
DVCinfusion:saline/7CKNA
10 HFD
-60
*
†
* †
Glucoseinfusion
rate(m
g/kg/min)
+ ++ –– +
3dHFDIvDMSO
IvALX
+ ++ –– +
£Basal nClamp
*
*
D
Glucoseprodu
c5on
(mg/kg/min)
E
0
1
2
3
4
5
Figure3.4
‡‡‡
100
120
140
160
180
200
220
240
260
280
300
basal 0 2 5 10
15
20
25
30
40
50
60
‡‡
‡‡
8000
10000
12000
***
0
20
40
60
80
100
120
140
160
A
day0
DVCcannula5on Vascularcatheteriza5on ivGTT
8 136 HFD
STZ/Nicinjec5on
DVCinfusion:saline/ALX
-240min 600
i.v.glucose(0.25g/kg)
Fasted
plasm
aglucose(m
g/dl)
B
Non-diabeMc + –7dSTZ/Nic/HFD – +
*
Plasmaglucose(m
g/dl)
CAUC
(0-60min)
†††
†† ††
‡‡
ࠠ
†††
RC+DVCsaline7-dSTZ/Nic/HFD+DVCsaline7-dSTZ/Nic/HFD+DVCALX
Figure3.5
A
Glucoseinfusion
rate(m
g/kg/min)
Glucoseprodu
c5on
(mg/kg/min)
28dRC28dHFDSalineALX
+ + – –– – + ++ – + –– + – +
+ + – –– – + ++ – + –– + – +
C D
0
1
2
3
4
5
6
0
1
2
3
4
5
6
7
8
9
10
200
250
300
350
400
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28
Bodyweight(g)
B
Time(days)
28dHFD28dRC
£Basal nClamp
Clamp
DVCinfusion:saline/ALX
SST(3µgkg-1min-1)
Insulin(1.2mUkg-1min-1)
Glucose(asneeded)
[3-3H]-glucose(0.4µCimin-1)
day0
DVCcannula5on
Vascularcatheteriza5on
16 29
0 24090
24
RCorHFD
-60min
*
*
*
*
*
Figure3.6
SST(3µgkg-1min-1)
Insulin(1.2mUkg-1min-1)
Glucose(asneeded)
[3-3H]-glucose(0.4µCimin-1)
day0
DVCcannula5on±LV-MM/
LV-GlyT1shRNA
Vascularcatheteriza5on
16
0min 24090
24
HFD
29
A
+ +
+ –
– +
28dHFD + +
DVCLV-MM + –
DVCLV-GlyT1shRNA – +
+ +
+ –
– +
Clamp
0
1
2
3
4
5
6
0
1
2
3
4
5
6
7
8
9
10
Glucoseinfusion
rate(m
g/kg/min)
Glucoseprodu
c5on
(mg/kg/min)
B C£Basal nClamp
E
300
320
340
360
380
400
420
Bodyweight(g)
D
*
*
*
-6
-4
-2
0
2
4
6
%Bod
yweightcha
nge
28dHFD+DVCLV-MM28dHFD+DVCLV-GlyT1shRNA
RC+DVCLV-MM
-6
-4
-2
0
2
4
6
Day4postDVCviralinjecMon
Day5postDVCviralinjecMon
**
Figure3.7
0
2
4
6
8
10
12
14
16
18
Day1 Day2 Day3
20
25
30
35
40
45
50
Day1
15
20
25
30
35
Day1
20
25
30
35
40
Day220
25
30
35
40
45
50
Day10
2
4
6
8
10
12
14
Day1 Day2 Day3
0
5
10
15
20
25
30
0 60 120 180 240 300 360
Cumula5
vefo
odintake(g)
0
5
10
15
20
25
30
0 60 120 180 240 300 360
Cumula5
vefo
odintake(g)
Time(min)
B
F
DVCsalineDVCglycine
DVCsalineDVCALX
† †† ††††
††
† † †††
†††
††††
††††††
Foodreturned
day0
FeedingExperiment
13
DVCinjec5on:saline/ALX
-60min 3600
Foodreturned
day0
DVCcannula5on FeedingExperiment
13
DVCinjec5on:saline/glycine
-10min 3600
A
E
%Bod
yweightg
ain
from
re-fe
eding
C D
G H
Dailyfo
odintake(g)
‡
‡
Dailyfo
odintake(g)
I
-8
-6
-4
-2
0
%Bod
yweightcha
nge DVCLV-MM
DVCLV-GlyT1shRNA
Day4postDVCviralinjec5on
0
5
10
15
20
25
0 60 120 180 240 300 360
Cumula5
vefo
odintake(g)
DVCLV-MMDVCLV-GlyT1shRNA
J
Dailyfo
odintake(g)
Daysagerre-feeding
K
‡
L
Daysagerre-feeding
‡
0
2
4
6
8
10
Day1 Day2
‡‡
%Bod
yweightg
ain
from
re-fe
eding
%Bod
yweightg
ain
from
re-fe
eding
‡
‡
20
25
30
35
Day2
20
25
30
35
40
Day2
††
††
††
†
Figure3.8
Glucose(asneeded)
[3-3H]-glucose(0.4µCimin-1)
SST(3µgkg-1min-1)
Insulin(1.2mUkg-1min-1)
0
2
4
6
8
10
12
60
Glucoseuptake(m
g/kg/min)
SalineALX
MK8017CKNAAd-MM
Ad-GluN1shRNA
+ – – – – – –– + – + + + +– – + + – – –– – – – + – –– – – – – + –– – – – – – +
C
0
20
40
60
80
100
120
140
160
Plasmaglucose(m
g/dl)
+ – – – – – –– + – + + + +– – + + – – –– – – – + – –– – – – – + –– – – – – – +
£Basal nClamp
Plasmainsulin
(ng/ml)
SalineALX
+ –– +
0.0
0.1
0.2
0.3
0.4
Bodyweight(kg)
+ –– +
A
E
DVCinfusion:saline/MK801/ALX/ALX+MK801/ALX
+7CKNA-90min 0 240
day0
DVCcannula5on±Ad-GluN1shRNA/Ad-MM±LV-GlyT1shRNA/LV-MM
Vascularcatheteriza5on
Experiment
8 13
Clamp
ivGTT
90
DVCinfusion:saline/
MK801/7CKNADVCinfusion:saline/ALX-240min 0
i.v.glucose(0.25g/kg)
-60
D F£BasalnClamp
Bday0
DVCcannula5on MicrodialysisExperiment
8
-120 0 300
DVCinfusion:saline/ALXBaseline
-270min
Recovery
Microdialysisinfusion:aECF
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
SupplementaryFigure3.1
0
5
10
15
20
25
0
2
4
6
8
10
12
0
6
12
18
24
30
0
30
60
90
120
150
180Glucoseinfusion
rate(m
g/kg/min)
ICV-4saline + –ICV-4ALX – +
Glucoseprodu
c5on
(mg/kg/min)
+ –– +
Glucoseuptake(m
g/kg/min)
+ –– +
Plasmaglucose(m
g/dl)
+ –– +
*
Glucoseuptake(m
g/kg/min)
Plasmaglucose(m
g/ml)
E£Basal nClamp
*
A B C D
F
SalineSHAMALX
HVAG
+ – + –+ + – –– + – +– – + +
+ – + –+ + – –– + – +– – + +
0
2
4
6
8
10
12
14
0
20
40
60
80
100
120
140
160
SupplementaryFigure3.2
A i
ii
iii
iv
v
SupplementaryFigure3.3
0
2
4
6
8
10
12
0
20
40
60
80
100
120
140
Glucoseuptake(m
g/kg/min)
Plasmaglucose(m
g/dl)
LV-MM + – –LV-GlyT1shRNA – + +
MK801 – – +
+ – –– + +– – +
0.0
0.1
0.2
0.3
0.4
Bodyweight(kg)
+ – –– + +– – +
A B C£BasalnClamp
SupplementaryFigure3.4
0
20
40
60
80
100
120
140
160
0
2
4
6
8
10
12
0
20
40
60
80
100
120
140
160
180
3dHFDSaline
GlycineALX
7CKNA
+ + + + ++ – – – –– + – + –– – + – +– – – + +
AGlucoseuptake(m
g/kg/min)
+ + + + ++ – – – –– + – + –– – + – +– – – + +
B£Basal nClamp
0
2
4
6
8
10
12
14
Glucoseuptake(m
g/kg/min)
D
Plasmaglucose(m
g/dl)
£BasalnClamp
3dHFDIvDMSO
IvALX
+ ++ –– +
+ ++ –– +
C
Plasmaglucose(m
g/dl)
SupplementaryFigure3.5
0
1
2
3
4
5
6
7
8
9
10
A
Glucoseuptake(m
g/kg/min)
B
Plasmaglucose(m
g/dl)
0
1
2
3
4
5
6
7
8
9
10
0
20
40
60
80
100
120
140
160
28dRC28dHFD
SalineALX
+ + – –– – + ++ – + –– + – +
+ + – –– – + ++ – + –– + – +
C
Glucoseuptake(m
g/kg/min)
D
Plasmaglucose(m
g/dl)
0
20
40
60
80
100
120
140
160£Basal nClamp £Basal nClamp
28dHFD + +DVCLV-MM + –
DVCLV-GlyT1shRNA – +
+ ++ –– +
SupplementaryFigure3.6
!MBH
KATP channel!
Glucagon!
GR!
PKA!↓
↓↓
↓
↓
↓
AMPK!Acetyl-CoA!
Malonyl-CoA!↓ACC!
LCFA!
LCFA!LCFA-CoA!
PKC-𝛿!
CPT-1!
Glucagon!
↓
DVC
Glycine!
GlyT1!NMDAr!
Glucose!homeostasis!
Glucose!homeostasis!
Figure4.1