Impaired hypothalamic Fto expression in response to fasting and glucose in obese mice.

OBJECTIVE: Recent genome-wide association studies have identified a strong association between obesity and common variants in the fat mass and obesity associated (FTO) gene. FTO has been detected in the hypothalamus, but little is known about its regulation in that particular brain structure. The present study addressed the hypothesis that hypothalamic FTO expression is regulated by nutrients, specifically by glucose, and that its regulation by nutrients is impaired in obesity. RESEARCH DESIGN AND METHODS: The effect of intraperitoneal (i.p.) or intracerebroventricular (i.c.v.) administration of glucose on hypothalamic Fto mRNA levels was examined in fasted mice. Additionally, the effect of glucose on Fto mRNA levels was also investigated ex vivo using mouse hypothalamic explants. Lastly, the effect of i.p. glucose injection on hypothalamic Fto immunoreactivity and food intake was compared between lean wild-type and obese ob/ob mice. RESULTS: In wild-type mice, fasting reduced both Fto mRNA levels and the number of Fto-immunoreactive cells in the hypothalamus, whereas i.p. glucose treatment reversed this effect of fasting. Furthermore, i.c.v. glucose treatment also increased hypothalamic Fto mRNA levels in fasted mice. Incubation of hypothalamic explants at high glucose concentration increased Fto mRNA levels. In ob/ob mice, both fasting and i.p. glucose treatment failed to alter the number of Fto-immunoreactive cells in the hypothalamus. Glucose-induced feeding suppression was abolished in ob/ob mice. CONCLUSION: Reduction in hypothalamic Fto expression after fasting likely arises at least partly from reduced circulating glucose levels and/or reduced central action of glucose. Obesity is associated with impairments in glucose-mediated regulation of hypothalamic Fto expression and anorexia. Hypothalamic Fto-expressing neurons may have a role in the regulation of metabolism by monitoring metabolic states of the body.

Lobotomy of genes: use of RNA interference in neuroscience.

Galen of Pergamon studied nerve function by shearing nerves in various species including monkeys, dogs, bulls and even elephants (humans being off limits to researchers; Sartan, 1954). An analogous strategy to determine gene function by ablating gene expression has recently been developed. RNA interference (RNAi) is a cellular response to double-stranded RNA (dsRNA) apparently as a defense against viral or transposon activity (Denli and Hannon, 2003; Dykxhoorn et al., 2003; Plasterk, 2002; Zamore, 2002). By activating this ancient defense mechanism through the introduction of artificial dsRNA, it is now possible to inhibit expression of almost any gene in almost any cell type, among them neuronal cells. In mammalian cells the active RNAi species must be short, approximately 21 nucleotide RNAs; these 21-bp species are called short interfering RNA (siRNA; Fig 1).

Daily nasal inoculation with the insulin gene ameliorates diabetes in mice.

The present study examined the feasibility of liposome-mediated gene transfer via nasal administration, for treating insulin-dependent diabetes mellitus. The rat insulin gene was packed under control of the CMV promoter, complexed with DC-chol/DOPE-based liposomes and administered daily via the nasal route in mice made severely diabetic by streptozocin. Sustained expression of the insulin gene was achieved and insulinopenia, ketonuria and death were prevented. Hyperglycemia and body weight reduction were significantly suppressed without evidence of hypoglycemia throughout the experimental period. RT-PCR and FISH analysis indicated that insulin was produced in the alveolar epithelial cells of the lung. Liposome-mediated in vivo gene transfer via nasal administration may provide an efficacious route for delivery of hormonal and other gene products into the blood stream.

Brain glucose-sensing mechanisms: ubiquitous silencing by aglycemia vs. hypothalamic neuroendocrine responses.

Interest in brain glucose-sensing mechanisms is motivated by two distinct neuronal responses to changes in glucose concentrations. One mechanism is global and ubiquitous in response to profound hypoglycemia, whereas the other mechanism is largely confined to specific hypothalamic neurons that respond to changes in glucose concentrations in the physiological range. Although both mechanisms use intracellular metabolism as an indicator of extracellular glucose concentration, the two mechanisms differ in key respects. Global hyperpolarization (inhibition) in response to 0 mM glucose can be reversed by pyruvate, implying that the reduction in ATP levels acting through ATP-dependent potassium (K-ATP) channels is the key metabolic signal for the global silencing in response to 0 mM glucose. In contrast, neuroendocrine hypothalamic responses in glucoresponsive and glucose-sensitive neurons (either excitation or inhibition, respectively) to physiological changes in glucose concentration appear to depend on glucokinase; neuroendocrine responses also depend on K-ATP channels, although the role of ATP itself is less clear. Lactate can substitute for glucose to produce these neuroendocrine effects, but pyruvate cannot, implying that NADH (possibly leading to anaplerotic production of malonyl-CoA) is a key metabolic signal for effects of glucose on glucoresponsive and glucose-sensitive hypothalamic neurons.

Cerulenin mimics effects of leptin on metabolic rate, food intake, and body weight independent of the melanocortin system, but unlike leptin, cerulenin fails to block neuroendocrine effects of fasting.

Cerulenin and a related compound, C75, have recently been reported to reduce food intake and body weight independent of leptin through a mechanism hypothesized, like leptin, to involve hypothalamic nutrition-sensitive neurons. To assess whether these inhibitors act through mechanisms similar to mechanisms engaged by leptin, ob/ob and Ay (agouti) mice, as well as fed and fasted wild-type mice, were treated with cerulenin. Like leptin, cerulenin reduced body weight and food intake and increased metabolic rate in ob/ob mice, and cerulenin produced the same effects in wild-type mice, whereas lithium chloride, at doses that produce conditioned taste aversion, reduced metabolic rate. However, in contrast to leptin, cerulenin did not prevent effects of fasting on plasma corticosterone or hypothalamic levels of neuropeptide Y, agouti-related peptide, pro-opiomelanocortin, or cocaine- and amphetamine-related peptide mRNA. Also, in contrast to leptin, cerulenin was highly effective to reduce body weight in Ay mice, in which obesity is caused by blockade of the melanocortin receptor. These data demonstrate that cerulenin produces metabolic effects similar to effects of leptin, but through mechanisms that are independent of, or down-stream from, both leptin and melanocortin receptors.

Neuroendocrine and pharmacological manipulations to assess how caloric restriction increases life span.

As part of an effort to review current understanding of the mechanisms by which caloric restriction (CR) extends maximum life span, the authors of the present review were requested to develop a list of key issues concerning the potential role of neuroendocrine systems in mediating these effects. It has long been hypothesized that failure of specific neuroendocrine functions during aging leads to key age-related systemic and physiological failures, and more recently it has been postulated that physiological neuroendocrine responses to CR may increase life span. However, although the acute neuroendocrine responses to fasting have been well studied, it is not clear that these responses are necessarily identical to those observed in response to the chronic moderate (30% to 50% reduction) CR that increases maximum life span. Therefore the recommendations of this panel fall into two categories. First, further characterization of neuroendocrine responses to CR over the entire life span is needed. Second, rigorous interventional studies are needed to test the extent to which neuroendocrine responses to CR mediate the effects of CR on life span, or alternatively if CR protects the function of essential neuroendocrine cells whose impairment reduces life span. Complementary studies using rodent models, nonhuman primates, and humans will be essential to assess the generality of elucidated mechanisms, and to determine if such mechanisms might apply to humans.

Adrenalectomy reverses obese phenotype and restores hypothalamic melanocortin tone in leptin-deficient ob/ob mice.

In genetically obese leptin-deficient ob/ob mice, adrenalectomy reverses or attenuates the obese phenotype. Relative to lean controls, ob/ob mice also exhibit decreased hypothalamic proopiomelanocortin (POMC) mRNA and increased hypothalamic agouti-related peptide (AGRP) mRNA and neuropeptide Y (NPY) mRNA. It has been hypothesized that this profile of hypothalamic gene expression contributes to the obese phenotype caused by leptin deficiency. To assess if reversal of obese phenotype by adrenalectomy entails normalization of hypothalamic gene expression, male wild-type and ob/ob mice were adrenalectomized (with saline supplementation) or sham adrenalectomized at 2 months of age. Mice were sacrificed 2 weeks after adrenalectomy, during which time food intake and body weight were monitored daily. After sacrifice, hypothalamic gene expression was assessed by Northern blot analysis as well as in situ hybridization. In wild-type mice, adrenalectomy significantly decreased AGRP mRNA but did not significantly influence POMC or NPY mRNA. In ob/ob mice, adrenalectomy reduced the levels of plasma glucose, serum insulin and corticosterone, and food intake toward or below wild-type levels, and it restored hypothalamic POMC and AGRP mRNA but not NPY mRNA to wild-type levels. These studies suggest that adrenalectomy reverses or attenuates the obese phenotype in ob/ob mice, in part by restoring hypothalamic melanocortin tone toward wild-type levels. These studies also demonstrate that factors other than leptin may play a major role in regulating hypothalamic melanocortin function.

T lymphopenia in genetically obese-diabetic Wistar fatty rats: effects of body weight reduction on T cells.

Patients with long-standing diabetes may have a propensity for infection-related mortality. In this study, lymphocyte subsets, the proliferative response of splenocytes to mitogens, and circulating levels of tumor necrosis factor alpha (TNF-alpha) in genetically obese-diabetic Wistar fatty (fa/fa) rats (WF) were longitudinally compared versus lean (+/?) litters (WL). Moreover, the effects of weight reduction with voglibose treatment on immunity were evaluated (WFV and WLV). Body weight was significantly increased in WF compared with WL. Hyperglycemia and hyperlipidemia developed, respectively, 11 weeks and 5 weeks thereafter throughout the observation periods. Circulating T cells and T-cell subsets of WF were significantly reduced after 22 weeks. There were also significant decreases in CD4+ and CD8+ thymocytes and the proliferative response of splenocytes. Circulating levels of TNF-alpha were significantly increased in WF. Treatment with voglibose resulted in significantly reduced blood glucose, insulin, cholesterol, triglyceride, and body weight in WFV. After weight reduction, circulating T cells and T-cell subsets were increased and TNF-alpha was decreased significantly in WFV. Our results suggest that the number and function of T cells in WF may be reduced, which may be related at least in part to elevated TNF-alpha levels, although the role of the other factors such as glucose, insulin, cholesterol, and triglycerides on T-cell immunity should be further investigated.

Changes in estrogenic regulation of estrogen receptor alpha mRNA and progesterone receptor mRNA in the female rat hypothalamus during aging: an in situ hybridization study.

We examined two molecular responses to estrogen, reduction in estrogen receptor alpha (ER alpha) mRNA and increase in progesterone receptor (PR) mRNA, in the hypothalamus of 3- (young) and 10-month-old (middle-aged) cycling, and 15-month-old (old) acyclic, Fischer 344 female rats. The rats were ovariectomized and then given silastic capsules containing 5% 17beta-estradiol. or empty implants, and killed 4 days after implantation. By means of in situ hybridization, we found that, in young rats, estrogen reduced ER alpha mRNA in both the ventromedial hypothalamus (VMH) and arcuate nucleus (ARC) but not in the preoptic area (POA). In contrast, the effect of estrogen on ER alpha mRNA in the VMH and ARC of middle-aged and old rats was not statistically significant. On the other hand in all regions the induction of PR mRNA by estrogen was at least as strong in middle-aged and old as in young rats. The present study revealed that the induction of PR mRNA by estrogen in the hypothalamus was not impaired with age but ER alpha mRNA in the VMH and ARC was significantly impaired with age, but not in the POA.

Liposome mediated gene transfer into GH3 cells, and rat brain, liver and gut: comparison of different polar or aliphatic domains.

BACKGROUND: Cationic liposomes may be used in gene transfer. However, different liposome configurations have varying efficiency in different tissues. AIMS: To compare multiple lipids during gene transfer into the intestinal mucosa, liver and central nervous system in the adult rat. We evaluate different lipid aliphatic and polar head domains. MATERIALS AND METHODS: Nine cationic or neutral phospholipids, combined with a cationic cholesterol derivative, have been compared to Lipofectin. Transfection was into GH3 cells and the adult rat brain, liver or intestinal mucosa. Results Optimum DNA:lipid ratio was lowest (1:2) in the intestinal mucosa and highest in GH3 cells (1:40). Lipofectin ", was most effective in brain and GH3 cells but had no activity in intestinal mucosa. Saturated cationic lipids transfect differently in GH3 cells and GI mucosa than in liver and brain. However, with saturated neutral phospholipids, GH3 cells, intestinal mucosa and liver transfect similarly. DOTAP the longest unsaturated cationic lipid (18:1) was most effective in the intestine, whereas DMEPC the shortest saturated neutral lipid (14:0) was optimal in the liver. CONCLUSIONS: In this study we propose a rational approach, based on systematic variations of lipids, to optimize liposome mediated gene transfer into the ventricular system of the brain, the liver and gastro-intestinal tract in the adult rat. Additionally, we demonstrate the feasibility of gene transfer into the mucosal cells of the gastro-intestinal tract as well as throughout the ventricular system of the rat brain. This requires liposomes which contain a cationic cholesterol derivative.

Effects of tumor necrosis factor alpha on leptin secretion and gene expression: relationship to changes of glucose metabolism in isolated rat adipocytes.

OBJECTIVE: Our objective was to determine the effects of prolonged exposure to tumor necrosis factor-alpha (TNF-alpha) on leptin secretion from and leptin (OB) gene expression in isolated adipocytes. Because glucose uptake and the metabolism of glucose beyond lactate are important determinants of leptin production in adipocytes, we examined the effects of TNF-alpha on glucose uptake and lactate production and their relationship to leptin secretion. DESIGN AND METHODS: Isolated rat adipocytes were anchored in a defined matrix of basement membrane components and cultured with media containing 5 mM glucose, 0.16 nM insulin and several concentrations of TNF-alpha. Leptin secretion, steady-state levels of leptin mRNA levels, glucose uptake, and lactate production were assessed over 96 h. RESULTS: TNF-alpha at concentrations of 0.024, 0.24, 2.4 and 24 ng/ml did not affect leptin secretion over 24 h. TNF-alpha at concentrations of 0.24 to 24 ng/ml significantly inhibited leptin secretion over 96 h by 19-60%. TNF-alpha at concentrations of 0.024 to 24 ng/ml significantly decreased steady-state levels of leptin mRNA after 96 h by 32-95%. In addition, TNF-alpha at concentrations of 2.4 and 24 ng/ml significantly increased glucose uptake and lactate production over 96 h by 30-57%. TNF-alpha at a concentration of 0.024 ng/ml did not affect leptin secretion, glucose uptake or lactate production. Overall, for the TNF-alpha concentrations tested, leptin secretion was inversely related to the percent of glucose carbon released as lactate; however, TNF-alpha did not induce a proportional increase of lactate production from glucose. CONCLUSION: Short-term (24 h) exposure of isolated adipocytes to TNF-alpha does not affect leptin secretion. Prolonged exposure to TNF-alpha produces a concentration-dependent inhibition of leptin secretion and gene expression. This suggests that the acute effect of TNF-alpha to increase circulating leptin levels in vivo may be indirect. TNF-alpha at higher concentrations increases glucose uptake, but does not increase the conversion of glucose to lactate. Therefore, TNF-alpha appears to induce a dissociation between adipocyte glucose metabolism and leptin production.

Fasting regulates hypothalamic neuropeptide Y, agouti-related peptide, and proopiomelanocortin in diabetic mice independent of changes in leptin or insulin.

Fasting increases hypothalamic neuropeptide Y (NPY) and agouti-related peptide (AGRP) messenger RNA (mRNA) and reduces hypothalamic POMC mRNA, and is also characterized by a reduction in plasma leptin, insulin, and glucose, each of which has been implicated in the regulation of hypothalamic gene expression. To further evaluate the roles of leptin, insulin, and glucose in mediating effects of fasting, we examined hypothalamic gene expression in nondiabetic and streptozotocin (STZ)-induced diabetic mice both under ad lib fed and 48-h fasted conditions. In both diabetic and nondiabetic mice, fasting stimulated hypothalamic NPY and AGRP mRNA and inhibited hypothalamic POMC mRNA and adipose leptin mRNA. However, in diabetic mice fasting had no effect on plasma leptin and insulin while decreasing plasma glucose, whereas in nondiabetic mice fasting decreased plasma leptin, insulin, and glucose. Furthermore, in nondiabetic fasted mice, NPY and AGRP mRNA were higher, and POMC mRNA and plasma glucose were lower, than in diabetic ad lib fed mice, even though insulin and leptin were similar in these two groups. These data are consistent with the hypothesis that although leptin and insulin regulate hypothalamic gene expression, glucose or other factors may have independent effects on hypothalamic and adipose gene expression under conditions of low insulin and leptin.

Hypothalamic glucose sensor: similarities to and differences from pancreatic beta-cell mechanisms.

Glucose-responsive neurons in the ventromedial hypothalamus (VMH) are stimulated when glucose increases from 5 to 20 mmol/l and are thought to play an essential role in regulating metabolism. The present studies examined the role of glucose metabolism in the mechanism by which glucose-responsive neurons sense glucose. The pancreatic, but not hepatic, form of glucokinase was expressed in the VMH, but not cerebral cortex, of adult rats. In brain slice preparations, the transition from 5 to 20 mmol/l glucose stimulated approximately 17% of the neurons (as determined by single-cell extracellular recording) from VMH but none in cortex. In contrast, most cells in both VMH and cortex were silent below 1 mmol/l and active at 5 mmol/l glucose. Glucosamine, 2-deoxyglucose, phloridzin, and iodoacetic acid blocked the activation of glucose-responsive neurons by the transition from 5 to 20 mmol/l glucose. Adding 15 mmol/l mannose, galactose, glyceraldehyde, glycerol, and lactate to 5 mmol/l glucose stimulated glucose-responsive neurons. In contrast, adding 15 mmol/l pyruvate to 5 mmol/l glucose failed to activate glucose-responsive neurons, although pyruvate added to 0 mmol/l glucose permitted neurons to maintain activity. Tolbutamide activated glucose-responsive neurons; however, diazoxide only blocked the effect of glucose in a minority of neurons. These data suggest that glucose-responsive neurons sense glucose through glycolysis using a mechanism similar to the mechanism of pancreatic beta-cells, except that glucose-responsive neurons are stimulated by glycerol and lactate, and diazoxide does not generally block the effect of glucose.

Targeted deletion of the Vgf gene indicates that the encoded secretory peptide precursor plays a novel role in the regulation of energy balance.

To determine the function of VGF, a secreted polypeptide that is synthesized by neurons, is abundant in the hypothalamus, and is regulated in the brain by electrical activity, injury, and the circadian clock, we generated knockout mice lacking Vgf. Homozygous mutants are small, hypermetabolic, hyperactive, and infertile, with markedly reduced leptin levels and fat stores and altered hypothalamic proopiomelanocortin (POMC), neuropeptide Y (NPY), and agouti-related peptide (AGRP) expression. Furthermore, VGF mRNA synthesis is induced in the hypothalamic arcuate nuclei of fasted normal mice. VGF therefore plays a critical role in the regulation of energy homeostasis, suggesting that the study of lean VGF mutant mice may provide insight into wasting disorders and, moreover, that pharmacological antagonism of VGF action(s) might constitute the basis for treatment of obesity.

Hypothalamic agouti-related protein messenger ribonucleic acid is inhibited by leptin and stimulated by fasting.

Agouti-related protein (AGRP) is a homologue of the agouti gene product and, when overexpressed, promotes obesity. Like neuropeptide Y (NPY) messenger RNA (mRNA), AGRP mRNA is produced in the hypothalamic arcuate nucleus and is elevated in leptin-deficient ob/ob and leptin-resistant db/db mice. These data suggest that AGRP mRNA might be affected by leptin and nutritional status in parallel with NPY mRNA. To test this hypothesis, we examined if AGRP mRNA would be, like NPY mRNA, inhibited by leptin injections and stimulated by fasting. AGRP mRNA was elevated in ob/ob mice about 5-fold compared with wild-type controls and was significantly inhibited by leptin treatment, as assessed by Northern blot analysis. In wild-type mice, AGRP mRNA was increased at least 13-fold by a 2-day fast, as assessed both by Northern blot analysis and in situ hybridization. In ad lib fed db/db mice, AGRP mRNA was elevated about 8-fold compared with ad lib fed wild-type controls, and was further increased by fasting in db/db mice. These data suggest that AGRP mRNA and NPY mRNA respond similarly to leptin and fasting.

Resistance to diet-induced obesity is associated with increased proopiomelanocortin mRNA and decreased neuropeptide Y mRNA in the hypothalamus.

Mechanisms mediating genetic susceptibility to diet-induced obesity have not been completely elucidated. Elevated hypothalamic neuropeptide Y (NPY) and decreased hypothalamic proopiomelanocortin (POMC) are thought to promote the development and maintenance of obesity. To assess the potential role of hypothalamic neuropeptide gene expression in diet-induced obesity, the present study examined effects of a high-fat diet on hypothalamic NPY and POMC mRNA in three strains of mice that differ in susceptibility to develop diet-induced obesity. C57BL/6J, CBA, and A/J mice were fed either normal rodent chow or a high-fat diet for 14 weeks after which hypothalamic gene expression was measured. On the high-fat diet, C57BL/6J mice gained the most weight, whereas A/J mice gained the least weight. On the high-fat diet, NPY mRNA significantly decreased as body weight increased in CBA and A/J mice, but not in C57BL/6J mice. In addition, POMC mRNA significantly increased as body weight increased in A/J mice, but not in CBA and C57BL/6J mice. Since decreased NPY mRNA and increased POMC mRNA would presumably attenuate weight gain, these results suggest that a high-fat diet produces compensatory changes in hypothalamic gene expression in mice resistant to diet-induced obesity but not in mice susceptible to diet-induced obesity.

Changes in neurotensin mRNA by estrogen in the female rat preoptic area during aging: an in situ hybridization histochemistry study.

The present study examined changes in the response of neurotensin mRNA to estrogen during aging at the single cell level. Ten days after ovariectomy, 3-, 10-, and 15-month-old female rats were implanted with estrogen or cholesterol and sacrificed 4 days later. An in situ hybridization study revealed that estrogen significantly increased the number of cells expressing neurotensin mRNA in the preoptic area of all age groups. Furthermore, frequency analysis indicated that estrogen significantly increased the proportion of heavily labeled cells in older rats but not in younger rats. Distributions of the grains/cells between cholesterol- and estrogen-treated rats suggested that older rats were at least as responsive to estrogen as young rats and possibly even more responsive. The result suggests that, at least as reflected by neurotensin mRNA, reproductive senescence in rats is not due to a general decrease in sensitivity to estrogen. Indeed, there is evidence of an increased responsiveness to estrogen with age.

Hyperphagia and weight gain after gold-thioglucose: relation to hypothalamic neuropeptide Y and proopiomelanocortin.

Genetic obesity is associated with increased neuropeptide Y (NPY) messenger RNA (mRNA) and decreased POMC mRNA in the hypothalamus of ob/ob and db/db mice, or impaired sensitivity to alphaMSH (derived from POMC) in the yellow agouti mouse. Acquired obesity can be produced by chemically lesioning the hypothalamus with either monosodium glutamate (MSG) in neonates or gold thioglucose (GTG) in adult mice. The present study examined whether elevated NPY mRNA and/or decreased POMC mRNA in the hypothalamus are associated with obesity due to hypothalamic lesions. GTG injection into adult mice produced a profound obese phenotype, including hyperphagia, increased body weight, and increased leptin mRNA and peptide, in association with reduced hypothalamic NPY mRNA and POMC mRNA. MSG treatment produced virtual elimination of NPY mRNA in the arcuate nucleus and a reduction of hypothalamic POMC mRNA, and led to elevated leptin. MSG pretreatment did not attenuate GTG-induced hyperphagia and obese phenotype. These results do not support a role for NPY-synthesizing neurons in the arcuate nucleus in mediating hypothalamic acquired obesity, but are consistent with the hypothesis that decreased activity of hypothalamic neurons synthesizing POMC play a role in mediating hypothalamic obesity.

Adrenal neuropeptide Y mRNA but not preproenkephalin mRNA induction by stress is impaired by aging in Fischer 344 rats.

Relatively few molecular markers of stress have been studied in aged individuals. Interactions of age and stress on adrenal neuropeptide Y (NPY) and preproenkephalin (ppENK) expression have not been reported. The purpose of these studies was to characterize the adrenal NPY and ppENK responses to stress using a common stressor, physical restraint for 2 h, in Fischer 344 rats at 7, 16 and 23 months of age. Northern blot techniques were used to evaluate induction by stress of adrenal NPY mRNA and adrenal ppENK mRNA. Two humoral responses to stress, serum glucose and corticosterone, were measured to corroborate that a stress response occurred. We observed that the induction by stress of adrenal NPY mRNA is impaired with age but the stress-induced elevation of adrenal ppENK mRNA, blood glucose, and corticosterone show no evidence of age-related impairments.