Microgliosis: a double-edged sword in the control of food intake

Maintaining energy balance is essential for survival and health. This physiological function is controlled by the brain, which adapts food intake to energy needs. Indeed, the brain constantly receives a multitude of biological signals that are derived from digested foods or that originate from the gastrointestinal tract, energy stores (liver and adipose tissues) and other metabolically active organs (muscles). These signals, which include circulating nutrients, hormones and neuronal inputs from the periphery, collectively provide information on the overall energy status of the body. In the brain, several neuronal populations can specifically detect these signals. Nutrient-sensing neurons are found in discrete brain areas and are highly enriched in the hypothalamus. In turn, specialized brain circuits coordinate homeostatic responses acting mainly on appetite, peripheral metabolism, activity and arousal. Accumulating evidence shows that hypothalamic microglial cells located at the vicinity of these circuits can influence the brain control of energy balance. However, microglial cells could have opposite effects on energy balance, that is homeostatic or detrimental, and the conditions for this shift are not totally understood yet. One hypothesis relies on the extent of microglial activation, and nutritional lipids can considerably change it.

Evidence for Constitutive Microbiota-Dependent Short-Term Control of Food Intake in Mice: Is There a Link with Inflammation, Oxidative Stress, Endotoxemia, and GLP-1?

Aims: Although prebiotics, probiotics, and fecal transplantation can alter the sensation of hunger and/or feeding behavior, the role of the constitutive gut microbiota in the short-term regulation of food intake during normal physiology is still unclear. Results: An antibiotic-induced microbiota depletion study was designed to compare feeding behavior in conventional and microbiota-depleted mice. Tissues were sampled to characterize the time profile of microbiota-derived signals in mice during consumption of either standard or high-fat food for 1 h. Pharmacological and genetic tools were used to evaluate the contribution of postprandial endotoxemia and inflammatory responses in the short-term regulation of food intake. We observed constitutive microbial and macronutrient-dependent control of food intake at the time scale of a meal; that is, within 1 h of food introduction. Specifically, microbiota depletion increased food intake, and the microbiota-derived anorectic effect became significant during the consumption of high-fat but not standard food. This anorectic effect correlated with a specific postprandial microbial metabolic signature, and did not require postprandial endotoxemia or an NOD-, LRR-, and Pyrin domain-containing protein 3-inflammasome-mediated inflammatory response. Innovation and Conclusion: These findings show that the gut microbiota controls host appetite at the time scale of a meal under normal physiology. Interestingly, a microbiota-derived anorectic effect develops specifically with a high-fat meal, indicating that gut microbiota activity is involved in the satietogenic properties of foods. Antioxid. Redox Signal. 37, 349-369.

The hepatocyte insulin receptor is required to program the liver clock and rhythmic gene expression

Liver physiology is circadian and sensitive to feeding and insulin. Food intake regulates insulin secretion and is a dominant signal for the liver clock. However, how much insulin contributes to the effect of feeding on the liver clock and rhythmic gene expression remains to be investigated. Insulin action partly depends on changes in insulin receptor (IR)-dependent gene expression. Here, we use hepatocyte-restricted gene deletion of IR to evaluate its role in the regulation and oscillation of gene expression as well as in the programming of the circadian clock in the adult mouse liver. We find that, in the absence of IR, the rhythmicity of core-clock gene expression is altered in response to day-restricted feeding. This change in core-clock gene expression is associated with defective reprogramming of liver gene expression. Our data show that an intact hepatocyte insulin receptor is required to program the liver clock and associated rhythmic gene expression.

Postprandial Hyperglycemia Stimulates Neuroglial Plasticity in Hypothalamic POMC Neurons after a Balanced Meal

Mechanistic studies in rodents evidenced synaptic remodeling in neuronal circuits that control food intake. However, the physiological relevance of this process is not well defined. Here, we show that the firing activity of anorexigenic POMC neurons located in the hypothalamus is increased after a standard meal. Postprandial hyperactivity of POMC neurons relies on synaptic plasticity that engages pre-synaptic mechanisms, which does not involve structural remodeling of synapses but retraction of glial coverage. These functional and morphological neuroglial changes are triggered by postprandial hyperglycemia. Chemogenetically induced glial retraction on POMC neurons is sufficient to increase POMC activity and modify meal patterns. These findings indicate that synaptic plasticity within the melanocortin system happens at the timescale of meals and likely contributes to short-term control of food intake. Interestingly, these effects are lost with a high-fat meal, suggesting that neuroglial plasticity of POMC neurons is involved in the satietogenic properties of foods.

The neuropeptide 26RFa in the human gut and pancreas: potential involvement in glucose homeostasis

OBJECTIVE

Recent studies performed in mice revealed that the neuropeptide 26RFa regulates glucose homeostasis by acting as an incretin and by increasing insulin sensitivity. However, in humans, an association between 26RFa and the regulation of glucose homeostasis is poorly documented. In this study, we have thus investigated in detail the distribution of 26RFa and its receptor, GPR103, in the gut and the pancreas, and determined the response of this peptidergic system to an oral glucose challenge in obese patients.

DESIGN AND METHODS

Distribution of 26RFa and GPR103 was examined by immunohistochemistry using gut and pancreas tissue sections. Circulating 26RFa was determined using a specific radioimmunoassay in plasma samples collected during an oral glucose tolerance test.

RESULTS

26RFa and GPR103 are present all along the gut but are more abundant in the stomach and duodenum. In the stomach, the peptide and its receptor are highly expressed in the gastric glands, whereas in the duodenum, ileum and colon they are present in the enterocytes and the goblet cells. In the pancreatic islets, the 26RFa/GPR103 system is mostly present in the β cells. During an oral glucose tolerance test, plasma 26RFa profile is different between obese patients and healthy volunteers, and we found strong positive correlations between 26RFa blood levels and the BMI, and with various parameters of insulin secretion and insulin resistance.

CONCLUSION

The present data suggest an involvement of the 26RFa/GPR103 peptidergic system in the control of human glucose homeostasis.

Neuropeptide 26RFa (QRFP) is a key regulator of glucose homeostasis and its activity is markedly altered in obese/hyperglycemic mice

Recent studies have shown that the hypothalamic neuropeptide 26RFa regulates glucose homeostasis by acting as an incretin and increasing insulin sensitivity. In this study, we further characterized the role of the 26RFa/GPR103 peptidergic system in the global regulation of glucose homeostasis using a 26RFa receptor antagonist and also assessed whether a dysfunction of the 26RFa/GPR103 system occurs in obese hyperglycemic mice. First, we demonstrate that administration of the GPR103 antagonist reduces the global glucose-induced incretin effect and insulin sensitivity whereas, conversely, administration of exogenous 26RFa attenuates glucose-induced hyperglycemia. Using a mouse model of high-fat diet-induced obesity and hyperglycemia, we found a loss of the antihyperglcemic effect and insulinotropic activity of 26RFa, accompanied with a marked reduction of its insulin-sensitive effect. Interestingly, this resistance to 26RFa is associated with a downregulation of the 26RFa receptor in the pancreatic islets, and insulin target tissues. Finally, we observed that the production and release kinetics of 26RFa after an oral glucose challenge is profoundly altered in the high-fat mice. Altogether, the present findings support the view that 26RFa is a key regulator of glucose homeostasis whose activity is markedly altered under obese/hyperglycemic conditions.

Lack of Hypothalamus Polysialylation Inducibility Correlates With Maladaptive Eating Behaviors and Predisposition to Obesity

High variability exists in individual susceptibility to develop overweight in an obesogenic environment and the biological underpinnings of this heterogeneity are poorly understood. In this brief report, we show in mice that the vulnerability to diet-induced obesity is associated with low level of polysialic acid-neural cell adhesion molecule (PSA-NCAM), a factor of neural plasticity, in the hypothalamus. As we previously shown that reduction of hypothalamic PSA-NCAM is sufficient to alter energy homeostasis and promote fat storage under hypercaloric pressure, inter-individual variability in hypothalamic PSA-NCAM might account for the vulnerability to diet-induced obesity. These data support the concept that reduced plasticity in brain circuits that control appetite, metabolism and body weight confers risk for eating disorders and obesity.

Brain Control of Plasma Cholesterol Involves Polysialic Acid Molecules in the Hypothalamus

The polysialic acid (PSA) is a large glycan that is added to cell-surface proteins during their post-translational maturation. In the brain, PSA modulates distances between cells and controls the plasticity of the nervous system. In the hypothalamus, PSA is involved in many aspects of energy balance including food intake, osmoregulation, circadian rhythm, and sleep. In this work, we investigated the role of hypothalamic PSA in the regulation of plasma cholesterol levels and distribution. We report that HFD consumption in mice rapidly increased plasma cholesterol, including VLDL, LDL, and HDL-cholesterol. Although plasma VLDL-cholesterol was normalized within the first week, LDL and HDL were still elevated after 2 weeks upon HFD. Importantly, we found that hypothalamic PSA removal aggravated LDL elevation and reduced HDL levels upon HFD. These results indicate that hypothalamic PSA controls plasma lipoprotein profile by circumventing the rise of LDL-to-HDL cholesterol ratio in plasma during overfeeding. Although mechanisms by which hypothalamic PSA controls plasma cholesterol homeostasis remains to be elucidated, these findings also suggest that low level of hypothalamic PSA might be a risk factor for dyslipidemia and cardiovascular diseases.

Transient Receptor Potential Canonical 3 (TRPC3) Channels Are Required for Hypothalamic Glucose Detection and Energy Homeostasis

The mediobasal hypothalamus (MBH) contains neurons capable of directly detecting metabolic signals such as glucose to control energy homeostasis. Among them, glucose-excited (GE) neurons increase their electrical activity when glucose rises. In view of previous work, we hypothesized that transient receptor potential canonical type 3 (TRPC3) channels are involved in hypothalamic glucose detection and the control of energy homeostasis. To investigate the role of TRPC3, we used constitutive and conditional TRPC3-deficient mouse models. Hypothalamic glucose detection was studied in vivo by measuring food intake and insulin secretion in response to increased brain glucose level. The role of TRPC3 in GE neuron response to glucose was studied by using in vitro calcium imaging on freshly dissociated MBH neurons. We found that whole-body and MBH TRPC3-deficient mice have increased body weight and food intake. The anorectic effect of intracerebroventricular glucose and the insulin secretory response to intracarotid glucose injection are blunted in TRPC3-deficient mice. TRPC3 loss of function or pharmacological inhibition blunts calcium responses to glucose in MBH neurons in vitro. Together, the results demonstrate that TRPC3 channels are required for the response to glucose of MBH GE neurons and the central effect of glucose on insulin secretion and food intake.

Plasticity of the Melanocortin System: Determinants and Possible Consequences on Food Intake

The melanocortin system is one of the most important neuronal pathways involved in the regulation of food intake and is probably the best characterized. Agouti-related peptide (AgRP) and proopiomelanocortin (POMC) expressing neurons located in the arcuate nucleus of the hypothalamus are the key elements of this system. These two neuronal populations are sensitive to circulating molecules and receive many excitatory and inhibitory inputs from various brain areas. According to sensory and metabolic information they integrate, these neurons control different aspects of feeding behavior and orchestrate autonomic responses aimed at maintaining energy homeostasis. Interestingly, composition and abundance of pre-synaptic inputs onto arcuate AgRP and POMC neurons vary in the adult hypothalamus in response to changes in the metabolic state, a phenomenon that can be recapitulated by treatment with hormones, such as leptin or ghrelin. As described in other neuroendrocrine systems, glia might be determinant to shift the synaptic configuration of AgRP and POMC neurons. Here, we discuss the physiological outcome of the synaptic plasticity of the melanocortin system, and more particularly its contribution to the control of energy balance. The discovery of this attribute has changed how we view obesity and related disorders, and opens new perspectives for their management.

Cerebral cell renewal in adult mice controls the onset of obesity

The hypothalamus plays a crucial role in the control of the energy balance and also retains neurogenic potential into adulthood. Recent studies have reported the severe alteration of the cell turn-over in the hypothalamus of obese animals and it has been proposed that a neurogenic deficiency in the hypothalamus could be involved in the development of obesity. To explore this possibility, we examined hypothalamic cell renewal during the homeostatic response to dietary fat in mice, i.e., at the onset of diet-induced obesity. We found that switching to high-fat diet (HFD) accelerated cell renewal in the hypothalamus through a local, rapid and transient increase in cell proliferation, peaking three days after introducing the HFD. Blocking HFD-induced cell proliferation by central delivery of an antimitotic drug prevented the food intake normalization observed after HFD introduction and accelerated the onset of obesity. This result showed that HFD-induced dividing brain cells supported an adaptive anorectic function. In addition, we found that the percentage of newly generated neurons adopting a POMC-phenotype in the arcuate nucleus was increased by HFD. This observation suggested that the maturation of neurons in feeding circuits was nutritionally regulated to adjust future energy intake. Taken together, these results showed that adult cerebral cell renewal was remarkably responsive to nutritional conditions. This constituted a physiological trait required to prevent severe weight gain under HFD. Hence this report highlighted the amazing plasticity of feeding circuits and brought new insights into our understanding of the nutritional regulation of the energy balance.

Differentiation and neural integration of hippocampal neuronal progenitors: signaling pathways sequentially involved

In the context of their potential implication in regenerative strategies, we characterized cell mechanisms underlying the fate of embryonic rat hippocampal H19-7 progenitors in culture upon induction of their differentiation, and tested their capacities to integrate into a neuronal network in vitro. Without addition of growth factors, nearly 100% of cells expressed various neuronal markers, with a progressive rise of the expression of Synapsin I and II, suggesting that cells developed as mature neurons with synaptogenic capacities. Fully differentiated neurons were identified as glutamatergic and expressed the receptor-associated protein PSD-95. Quantification of ATP showed that 60% of cells died within 24 h after differentiation. Cell death was shown to imply Erk1/2-dependent intrinsic mitochondrial apoptosis signaling pathway, with activation of caspase-9 and -3, finally leading to single-strand DNA. Surviving neurons displayed high levels of Akt, phospho-Akt, and antiapoptotic proteins such as Bcl-2 and Bcl-XL, with decreased caspase activation. In the absence of trophic support, the proapoptotic death-associated protein (DAP) kinase was dramatically stimulated by 24 h postdifferentiation, along with increased levels of p38 and phospho-p38, and caspase reactivation. These findings show that different signaling pathways are sequentially triggered by differentiation, and highlight that ultimate cell death would involve p38 and DAP kinase activation. This was supported by the improvement of cell survival at 24-h postdifferentiation when cells were treated by PD169316, a specific inhibitor of p38. Finally, when seeded on rat hippocampal primary cultured neurons, a significant number of differentiated H19-7 cells were able to survive and to develop cell-cell communication.

Gestational vitamin B deficiency leads to homocysteine-associated brain apoptosis and alters neurobehavioral development in rats

Hyperhomocysteinemia has been identified as a risk factor for neurological disorders. To study the influence of early deficiency in nutritional determinants of hyperhomocysteinemia on the developing rat brain, dams were fed a standard diet or a diet lacking methyl groups during gestation and lactation. Homocysteinemia progressively increased in the offspring of the deficient group and at 21 days reached 13.3+/-3.7 micromol/L versus 6.8+/-0.3 micromol/L in controls. Homocysteine accumulated in both neurons and astrocytes of selective brain structures including the hippocampus, the cerebellum, the striatum, and the neurogenic subventricular zone. Most homocysteine-positive cells expressed p53 and displayed fragmented DNA indicative of apoptosis. Righting reflex and negative geotaxis revealed a delay in the onset of integration capacities in the deficient group. Between 19 and 21 days, a poorer success score was recorded in deficient animals in a locomotor coordination test. A switch to normal food after weaning allowed restoration of normal homocysteinemia. Nevertheless, at 80 days of age, the exploratory behavior in the elevated-plus maze and the learning and memory behavior in the eight-arm maze revealed that early vitamin B deprivation is associated with persistent functional disabilities, possibly resulting from the ensuing neurotoxic effects of homocysteine.

Mild, non-lesioning transient hypoxia in the newborn rat induces delayed brain neurogenesis associated with improved memory scores

Although neonatal hypoxia can lead to brain damage, mild hypoxic episodes may be beneficial, as illustrated by tolerance induction by preconditioning, a process that might involve neurogenesis. To examine if brief hypoxia in newborn rats could stimulate the generation of neurons, pups were exposed for 5 min to 100% N2. Cell density and apoptosis were monitored in various brain regions and cell proliferation was studied by the incorporation of bromodeoxyuridine. Hypoxia did not result in detectable cell death but promoted cell proliferation in the ensuing three weeks in the subventricular zone and hippocampal dentate gyrus, with increased cell density in hippocampus CA1 pyramidal cells and granular layer of the dentate gyrus. Newly generated cells expressed neuronal markers (NeuroD or neuronal nuclear antigen) and were able to migrate from germinative zones to specific sites, in particular from the subventricular zone to the CA1 layer along the posterior periventricle. Neurogenesis was associated with an early activation of the extracellular regulated kinase 1/2 pathway, and pre-hypoxic administration of U0126, an inhibitor of mitogen-activated protein kinase kinase, impaired hypoxia effect on cell proliferation. Neurobehavioral capacities of hypoxic rats paralleled those of controls, but early exposure to hypoxia was associated with significantly improved memory retrieval scores at 40 days. In conclusion, brief neonatal hypoxia may trigger delayed generation of potentially functional neurons without concomitant cell death. This may constitute an interesting model for studying cell key events involved in the induction of neurogenesis.

Mild neonatal hypoxia exacerbates the effects of vitamin-deficient diet on homocysteine metabolism in rats

Elevated plasma homocysteine has been linked to pregnancy complications and developmental diseases. Whereas hyperhomocysteinemia is frequently observed in populations at risk of malnutrition, hypoxia may alter the remethylation of homocysteine in hepatocytes. We aimed to investigate the combined influences of early deficiency in nutritional determinants of hyperhomocysteinemia and of neonatal hypoxia on homocysteine metabolic pathways in developing rats. Dams were fed a standard diet or a diet deficient in vitamins B12, B2, folate, month, and choline from 1 mo before pregnancy until weaning of the offspring. The pups were divided into four treatment groups corresponding to "no hypoxia/standard diet," "hypoxia (100% N2 for 5 min at postnatal d 1)/standard diet," "no hypoxia/deficiency," and "hypoxia/deficiency," and homocysteine metabolism was analyzed in their liver at postnatal d 21. Hypoxia increased plasma homocysteine in deficient pups (21.2 +/- 1.6 versus 13.3 +/- 1.2 microM, p < 0.05). Whereas mRNA levels of cystathionine beta-synthase remained unaltered, deficiency reduced the enzyme activity (48.7 +/- 2.9 versus 83.6 +/- 6.3 nmol/h/mg, p < 0.01), an effect potentiated by hypoxia (29.4 +/- 4.7 nmol/h/mg, p < 0.05). The decrease in methylene-tetrahydrofolate reductase activity measured in deficient pups was attenuated by hypoxia (p < 0.05), and methionine-adenosyltransferase activity was slightly reduced only in the "hypoxia/deficiency" group (p < 0.05). Finally, hypoxia enhanced the deficiency-induced drop of the S-adenosylmethionine/S-adenosylhomocysteine ratio, which is known to influence DNA methylation and gene expression. In conclusion, neonatal hypoxia may increase homocysteinemia mainly by decreasing homocysteine transsulfuration in developing rats under methyl-deficient regimen. It could therefore potentiate the well-known adverse effects of hyperhomocysteinemia.

Limitation of the in vitro whole blood assay for predicting the COX selectivity of NSAIDs in clinical use

AIMS

To assess if the inhibitory potency of nonsteroidal anti-inflammatory drugs (NSAIDs) on cyclooxygenase (COX) isoenzymes, when given therapeutically in humans, can be predicted from their in vitro concentration-response curves using the whole blood assay.

METHODS

Twenty-four healthy male volunteers aged 20--27 years were recruited. Inhibition of blood COX isoenzymes was determined in vitro before any drug intake and ex vivo after single and repeated intake of either 7.5 mg meloxicam once, 400 mg ibuprofen three times daily or 75 mg diclofenac SR once, taken in a randomized cross-over design. Production of thromboxane B2 (TXB2) during clotting and of prostaglandin E2 (PGE2) during endotoxin exposure served as indicators of platelet COX-1 and monocyte COX-2 activity, respectively. Drugs were determined in plasma by h.p.l.c., with a chiral separation of ibuprofen and free fractions after equilibrium dialysis.

RESULTS

Intra-subject variation for COX-1 and COX-2 at baseline was at 26 +/- 18% and 18 +/- 13% respectively, and intersubject variation at 39% and 36%, respectively. The ratios of IC50s and, at best, of IC80s revealed diclofenac and meloxicam as selective COX-2 inhibitors and ibuprofen as a preferential COX-1 inhibitor in vitro. However, after oral intake, ibuprofen inhibited ex vivo COX-2 by 80% whereas diclofenac inhibited COX-1 by 70%. Meloxicam inhibited COX-1 from 30 to 55% depending on the repetition of the dose and increase in plasma concentrations. Using in vitro dose--response curves, the in vivo inhibitory potency of diclofenac was estimated adequately from its circulating concentration ([-0.18, 0.21] for COX-1 and [-0.13, -0.03] for COX-2) but this was not the case for ibuprofen on COX-2 ([-0.14, 0.27]) and meloxicam on COX-1 ([0.31, 1.05]). The limited predictability of the system was not improved through considering the unbound fraction of the drugs or the variable chiral inversion of ibuprofen.

CONCLUSIONS

Assessment of COX-2 selectivity based on in vitro studies and pharmacological modelling has a limited clinical relevance. There is a need to investigate COX selectivity at therapeutic plasma concentrations of NSAIDs using the ex vivo whole blood assay.

Stimulation of cyclooxygenase-2-activity by nitric oxide-derived species in rat chondrocyte: lack of contribution to loss of cartilage anabolism

Cross-talk between inducible nitric oxide synthase (NOS II) and cyclooxygenase-2 (COX-2) was investigated in rat chondrocytes. In monolayers, interleukin-1beta (IL-1beta) induced COX-2 and NOS II expression in a dose- and time-dependent manner, to produce high prostaglandin E(2) (PGE(2)) and nitrite (NO(2)(-)) levels in an apparently coordinated fashion. COX-2 mRNA was induced earlier (30 min. versus 4 hr) and less markedly (4-fold versus 12-fold at 24 hr) than NOS II, and was poorly affected by the translational inhibitor cycloheximide (CHX). IL-1beta did not stabilize COX-2 mRNA in contrast to CHX. Indomethacin and NS-398 lacked any effect on NO(2)(-) levels whereas L-NMMA and SMT reduced PGE(2) levels at concentration inhibiting NO(2)(-) production from 50 to 90%, even when added at a time allowing a complete expression of both enzymes (8 hr). Basal COX activity was unaffected by NO donors. The SOD mimetic, CuDips inhibited COX-2 activity by more than 75% whereas catalase did not. Inhibition of COX-2 by CuDips was not sensitive to catalase, consistent with a superoxide-mediated effect. In tridimensional culture, IL-1beta inhibited radiolabelled sodium sulphate incorporation while stimulating COX-2 and NOS II activities. Cartilage injury was corrected by L-NMMA or CuDips but not by NSAIDs, consistent with a peroxynitrite-mediated effect. These results show that in chondrocytes: (i) COX2 and NOS II genes are induced sequentially and distinctly by IL-1beta; (ii) COX-1 and COX-2 activity are affected differently by NO-derived species; (iii) peroxynitrite accounts likely for stimulation of COX-2 activity and inhibition of proteoglycan synthesis induced by IL-1beta.

Cyclo-oxygenase isoenzymes. How recent findings affect thinking about nonsteroidal anti-inflammatory drugs

The discovery of at least 2 cyclo-oxygenase (COX) isoenzymes, referred to as COX-1 and COX-2, has updated our knowledge of nonsteroidal anti-inflammatory drugs (NSAIDs). This has lead investigators to reconsider what can be awaited from this class of drugs. The 2 COX isoenzymes share structural and enzymatic similarities, but are specifically regulated at the molecular level and may be distinguished apart in their functions, although some physiological overlap between them does occur. The major goal in developing selective COX inhibitors is to improve NSAID tolerability. Classic NSAIDs preferentially inhibit COX-1 in vitro, but it appears hazardous to judge their gastrointestinal (GI) safety profile from these data. New compounds with a high selectivity for COX-2, especially those that are non-acidic, may be better tolerated in the GI tract. While these compounds also might have a potential use in various diseases such as colorectal cancer and neurodegenerative diseases of the Alzheimer type, the possible appearance of adverse effects, perhaps renally-related, must be taken into consideration. Finally, well-designed large clinical trials are required to adequately estimate both the promising therapeutic advantages that may be offered by highly selective NSAIDs, and the potential drawbacks that may be inherent with prolonged COX-2 inhibition.