Fat taste and feeding behavior
The dissection of factors involving fat intake leading to alterations in feeding behavior has been a topic of research for over 60 years. The first finding that oral properties of fat alter food consumption was that addition lard, stimulates food intake in rodents . In general, laboratory rodents prefer HF foods relative to a standard low-fat diet [73-75]. In one study examining the general propensity of laboratory mice to prefer HF foods, it was demonstrated that 10 out of a total 13 mouse strains examined prefered HF foods to low-fat foods . Using two-bottle preference tests, it has also been demonstrated that rodents prefer oil emulsions with varying degrees of fat relative to water or non-nutritive control solutions with similar texture . To minimize post-oral feedback, which has a potent influence on stimulating food intake following previous exposures with nutritive solution, a majority of these experiments have been conducted using brief access to the test preference. In addition to brief access tests, research also demonstrates that animals subjected to the sham feeding paradigm freely drink nutritive corn oil or non-nutritive mineral oil emulsions during one-bottle feeding tests [77, 78]. However, when presented with two bottles in a choice preference test with one bottle containing corn oil and the other mineral oil, animals always prefer corn oil to mineral oil . With the current discovery of fatty-acid receptors on the lingual epithelium, this effect can likely be attributed to the free fatty acids found in corn oil, which activate lingual taste receptor cells and are not present in mineral oil . Finally, demonstrating the ability of fat taste to alter energy consumption, sham feeding fats stimulates increased intake of a standard rat diet following fat exposure . Together, these data exemplify the profound role that oral fat detection can have on influencing feeding behavior.
Mechanisms of oral fat detection
The location of oral fat detection is thought to occur almost solely in the posterior lingual epithelium, specifically, in the taste receptor cells of the circumvallate papillae [81, 82]. Lingual fat sensing involves several receptors expressed on taste receptor cells such as g-protein coupled receptor 40 (GPR40)  and GPR120 [84, 85] as well as the fatty acid translocase CD36  and delayed-rectifying potassium channels . Similarly to sweet taste, fat detection is transduced via gustatory nerves that innervate the lingual epithelium, which transmit signals to higher brain centers ultimately controlling food intake .
Fat activates taste receptor cells via an apical sensor, such as voltage-gated ion channels that contribute to neurotransmitter release on the basolateral portion of the cell . For example, patch-clamp recordings on isolated fungiform taste cells demonstrate that free-fatty acids inhibit delayed rectifying potassium channels (DRKs) . This is specific to poly-unsaturated fatty-acids that are applied extracellularly, denoting these channels may establish oral fat preferences. Various DRKs have been found in the fungiform papillae , but the specific Kv1.5 channel that is found in cardiac tissue and inhibited by fatty acids  is present in rodent lingual epithelium . The proposed mechanism of activation of this channel is via a direct binding between fatty acids and a domain of the Kv1.5, which has been reported in cardiomyocytes . Furthermore, DRKs may contribute to taste modulation by enhancing perceived intensity of taste, which is in agreeance with findings that fatty acids enhances perceived intensity of various tastes .
G-protein coupled receptors on the lingual epithelium may also be responsible for the oral detection of fat. Normally localized on intestinal epithelium enteroendocrine cells, GPRs respond to various lengths of free fatty acids and mediate the release of gut hormones CCK [91, 92] and GLP-1 . Recently, researchers have identified GPR120, which responds to long chain fatty acids, in the sensory fungiform and circumvallate papillae, whereas GPR40, which responds to medium and long-chain fatty acids, was absent from the lingual epithelium . Interestingly, the enteroendocrine cell model, STC-1, expresses GPR120 and activation of this receptor induces cellular depolarization via a PLCβ2, Ca+2-dependent mechanism similar to the activation of taste receptor cells by a taste stimulus [93, 94]. Furthermore, genetic ablation of GPR120 results in reduced fatty-acid preference and loss of fatty acid-induced activation of nerves innervating the lingual epithelium . Together, these data exemplify a role of GPR120 in the detection of oral fats.
In addition to DRKs and GPRs, the fatty acid translocase, CD36, may play the most pivotal role in oral fat detection as suggested by research over the past 10 years [81, 82]. Originally discovered as the main mechanism of fatty-acid uptake in adipocytes , CD36 is also significantly expressed in the lingual epithelium, specifically in the circumvallate papillae [81, 82]. Located on the apical membrane of taste cells  and conservatively expressed across multiple species [82, 96], CD36 is positioned to bind extracellular fatty acids in the oral cavity. Furthermore, taste receptor cells that express CD36 are located in close proximity to a lipid-rich environment near the Von Ebner’s glands that secrete lingual lipase  (Figure 4). The extracellular structure of CD36, with a large hydrophobic pocket, also illustrates its role in binding fatty acids . Indeed, CD36 reversibly binds fatty acids, specifically in the nanomolar range [99, 100]. Experiments using CD36 KO mice demonstrate the significant physiological role of this receptor in fat taste [82, 101]. The integral function of CD36 in fat detection was established by Laugerette et al, who demonstrated that deletion of CD36 in mice results in abolished spontaneous fat preference and an absence of oral fatty-acid induced neural activation . Together, it is hypothesized that GPR120 and CD36 may mediate fat taste; however, due to its relative infancy in relation to sweet taste, further clarification is needed to examine the exact mechanisms of fat taste.
Figure 4: Localization of CD36 on the apical portion of the circumvallate papillae TRCs. Taste buds containing TRCs expressing CD36 are juxtaposed to the Von Ebner’s gland that secretes lingual lipase, which hydrolyzes triglycerides into free fatty acids in the mouth .
Experiments examining fatty acid-induced activation of TRCs have demonstrated the great dependence of these cells on CD36 in mediating this response. Isolated taste receptor cells expressing CD36 display large increases in cellular activation in response to fatty acid application . This is specifically dependent upon CD36 as blockade of fatty acid binding inhibits increases in intracellular calcium . The second messenger system which mediates CD36-induced activation of taste receptor cells may be PLCβ2 as taste cells express inositol-triphosphate  and CD36 and PLCβ2 are co-localized in other cells of the mammalian system . Together, these findings demonstrate that the intracellular signaling mechanisms for sweet and fat taste may be extremely similar. For example, increases in intracellular calcium activate TRPM5 channels that induce an influx of Na+ that leads to cell depolarization . Taste receptor cells expressing CD36 indeed depolarize in response to fatty acids making a case for this protein as a proposed mechanism in oral fat detection . Furthermore, the finding that TRPM5 KO mice display no preference for oils implicates this protein in mediating CD36-dependent fatty-acid signaling and taste cell depolarization . While it is hypothesized that CD36 is localized on Type II TRCs due similar intracellular signaling markers as sweet taste, this has yet to be examined. Rapid influxes in intracellular calcium also serve to induce neurotransmitter via cellular depolarization, which is observed in CD36-positive TRCs . Some of the proposed neurotransmitters responsible for taste cell signaling to afferent nerve fibers include acetylcholine, norepinephrine, serotonin, and glutamate. Specifically, serotonin and noradrenaline may be the primary neurotransmitters released by CD36-positive taste receptor cells as the transcript for two vital enzymes in the production of these neurotransmitters is present in CD36-positive cells . As such, serotonin and noradrenaline are both released from fatty-acid activated taste receptor cells; however whether this is directly by CD36 expressing cells or indirectly via other TRCs is unknown . Thus, despite the wealth of data illustrating the role for CD36 mediating the oral detection of fats, intriguing questions such as the exact pathways in TRCs that are responsible for fat taste transduction as well as the TRC type that expresses CD36 remain unanswered.
Taste transduction, central signaling, and reward
Regardless of the taste detected, TRCs directly or indirectly release various neurotransmitters [107-112] or peptide hormones [113-117] that activate the CT and GP nerve fibers. The specific mechanism is not completely understood; however, this may involve direct TRC to nerve connections, in which Type II cells generate ATP that interacts via connexins to nerve fibers [65, 66]. A second mechanism is that indirect signaling from TRC Type II cells that sense tastants, release ATP, which then stimulates Type III TRCs to release neurotransmitters that activate afferent nerves. Support of this latter hypothesis comes from the finding that Type III cells are the only TRCs that express voltage gated calcium channels and neurotransmitters, such as serotonin [108, 109]. Additionally, the release of ATP from Type II cells can bind to P2X receptors , which are ion channels that open in response to extracellular ATP and expressed by Type III TRCs (Figure 5). While this intriguing and yet unestablished transduction mechanism is currently under investigation, the afferent CT and GP fibers indeed receive input from taste bud complexes and signal upstream to the nucleus of the solitary tract (NTS) of the caudal brainstem.
Figure 5: Proposed mechanism of cell-to-cell communication for taste transduction. Type II TRCs express the receptors for a variety of tastants, most importantly, T1R2+3. Type II cells relay taste information either directly via cell-afferent nerve connections or indirectly via Type III cell-nerve synapses which repond to extracellular ATP released by Type II cells. Afferent nerves then relay information to the hindbrain for taste coding .
Somewhat similar to the taste buds, in which taste receptive fields are reasonably broad, NTS neurons can respond to a broad number of stimuli . Despite this, specific populations of NTS neurons are more responsive to specific taste stimuli [120-122]. This has been shown by various electrophysiological methods, such as single neuron recordings, which has led to the identification of “tuned” neurons. For example, neurons that are highly activated by sweet stimuli evoke a specific neural activation pattern and can be differentiated from neurons responding to other tastes such as umami, sour, or bitter . Furthermore, for each taste, different stimuli evoke specific neural firing patterns. For example, a nonnutritive sweetener, such as saccharin, can be differentiated from a nutritive sweet stimuli, such as sucrose based on neural firing . Using single neuron recordings, researchers have demonstrated that T1R3 is vital in the activation of sweet responsive NTS neurons, providing evidence of this receptor in the central processing of sweet stimuli . Upstream, neurons from the NTS project to the pontine Parabrachial Nucleus (PBN). From the PBN, neurons project in two directions, the first is the gustatory cortex (GC), which is responsible for encoding taste, mechanical, and visceral stimuli. The second pathway consists of PBN neurons projecting to the limbic system, ultimately terminating in the ventral striatum, involving reward function. The gustatory cortex is best viewed as the brain region that assigns hedonic value to taste stimuli. For example, stimuli with similar hedonic values activate similar regions of the GC and differing hedonic values activate more distinct regions [126-128]. Additionally, the GC is also associated with conditioned taste aversion (CTA), in which an animal avoids a normally preferred taste stimuli (such as saccharin) paired with an aversive stimuli (such as lithium chloride (LiCl)) [129, 130]. In this manner, the GC exhibits relative plasticity as saccharin will typically activate specific neuron subsets of the GC, but after pairing with LiCl, a different neural activation patterning occurs . Furthermore, due to the plasticity of this brain region, this effect is fully reversible. Although these pieces of evidence may demonstrate the functional role of the GC in determining taste, an overwhelming amount of evidence places a role of the PBN limbic projection to the ventral striatum in hedonics and controlling reward function of palatable stimuli.
Central reward pathways upstream of the NTS and PBN are strong contributors to both sweet and fat taste. Behavioral evidence of this comes from the finding of reinforcement studies demonstrating the rewarding values of sweet and fat stimuli. For example, reinforcement designs are used in which an animal must work more to obtain the same reward (palatable food). When given access to sucrose or fat, in the absence of post-oral feedback, rats work harder for continued access to these stimuli, demonstrating the behaviorally rewarding value of sweet or fat taste [132, 133]. As previously mentioned, neurons of the PBN extend to the ventral striatum . Using microdialysis, immunohistochemistry, and brain lesioning, several collective studies demonstrated that the PBN-striatal projections are responsible for the stimulation of sweet solution intake and subsequent sucrose-induced influxes of dopamine (DA) in the Nucleus of Accumbens (NAcc), a nuclei in the ventral striatum involved in reward function . While the NAcc consists of two parts, a core and shell, the latter of the two is vital in tracking reward. Thus, in addition to the view that the GC tracks hedonic value, release of DA from the striatum, which contains the NAcc, is associated with this perception as well [136, 137]. For example, influxes of DA in the NAcc are a direct function of oral reward and the motivational state to obtain a reward, such as sweet stimuli [138, 139]. The first initial evidence linking sweet taste to reward function was the finding that drinking saccharin led to increases in NAcc DA, and is associated with learned behavioral processes . As well as saccharin, the consumption of sucrose leads to increases of DA in the NAcc . However, this latter finding could be influenced by the nutritive value of sucrose reaching the intestine. Thus, through a variety of controlled experiments using fixed and non-fixed volumes of low and high concentrations of sucrose during sham-feeding, it was demonstrated that sweet taste alone is sufficient for increasing NAcc DA . In addition to sweet taste, sham feeding of nutritive oils has also demonstrated fat taste leading to increases in DA in the NAcc relative to water . Therefore, although detection of taste arises from the tongue involving apical membrane bound receptors, afferent signals relayed from TRCs indirectly provide input to central brain areas, such as the NAcc, which signal sweet and fat taste as rewarding, and lead to stimulation of intake for these palatable stimuli.
Obesity and taste
While the varying detection for sweet and fat taste influences short-term intake of these stimuli, the role of oral sensitivity to sweets and its relation to obesity is less clear. In general, obese animals over consume sweet foods, with evidence suggesting this may be mediated by taste functions. Genetic rodent models of obesity display an inability to detect low concentrations of sweet solutions with an accompanying increased consumption of highly concentrated sweet solutions [143, 144]. Thus, obese animals are typically described as having a decreased oral sensitivity to sweet solutions. Decreased oral detection of sweets could contribute to over consumption as more sucrose would be needed to stimulate lingual sweet receptors and release sufficient DA to signify reward . The mechanism responsible for the observed detection of sweet stimuli in obese rodents may be due in part to the fact that the sweet taste heterodimer is co-localized with the leptin receptor, and loss of leptin receptor function leads to diminished oral sensitivity to sweet substances . Thus, the observed leptin resistance in obese rodent models due to increased adiposity denotes that energy status in these models is a major determinant for these behavioral findings. Furthermore, lean or food deprived animals, where normal or enhanced leptin receptor sensitivity is present, exhibit in an increased sensitivity to oral sweet stimuli . Both of these findings, however, could also be attributed to central reward signaling controlling feeding behavior and guiding taste function as obese animals display increased motivational states [145, 148, 149] demonstrating the complexity of taste signaling in energy homeostasis. While obesity clearly is correlated with impaired sweet taste signaling in animal models already obese, studies examining variations in the sweet taste receptor before the onset of obesity have been less promising. For example, the genotype of T1R3 does not predict sucrose- or fructose-induced hyperphagia . However, this does not necessarily rule out altered taste in the pre-obese state as animals prone to obesity display increases in consumption of palatable sweet stimuli before the onset of obesity [145, 150]. Furthermore, in humans, a polymorphism in the T1R2 gene has been shown to correlate with an overweight BMI and increased consumption of carbohydrates . Whether individuals displaying the T1R2 polymorphism exhibit altered sensitivity to sweets, which would lead to increased carbohydrate consumption is unclear from this study alone. In general, data concerning sweet taste sensitivity in obesity is conflicting, and little data has assessed the contribution of taste to human obesity. Initial data demonstrated that obese individuals do not differ in sweet taste sensitivity relative to normal weight individuals [152-155] with formerly obese individuals displaying increases in sweet taste responsiveness . More recent evidence, however, reveals obese individuals maintain a higher affinity for sweet stimuli relative to lean individuals [157-160]. The observed differences in these studies are explained by improvements in the assessment of sweet taste function . Interestingly, in animals, while obesity is associated with alterations in sweet taste function, more recent evidence indicates that taste is vital in influencing short-term intake of sweet solutions and post-oral feedback contributes to long-term intake .
In addition to oral sweet sensitivity, obese animals also display altered oral sensitivity to fats. For example, the Otsuka Long-Evans Tokushima Fatty (OLETF) rat, which is hyperphagic and subsequently becomes obese due its overeating consumes more of high corn oil concentrations when sham feeding than lean controls during the fed state. Furthermore, at a relatively low concentration of corn oil, OLETF rats exhibit increased intake compared to lean animals when food deprived . While this genetic model of obesity clearly displays alteration in oral fat sensitivity leading to increased consumption of fat, high-fat (HF) feeding in rats or mice, which results in obesity, is associated with decreases in lingual expression of CD36 . Thus, it is hypothesized that by expressing less CD36, these animals are unable to detect lingual fats as well as lean or chow-fed controls and increase fat consumption as a response. However, the effect of HF feeding on lingual CD36 expression in the absence of obesity has not been examined. This is extremely important as HF-feeding influences oral fat sensitivity regardless of the obese state as HF-fed non-obese animals display increased acceptance for fats relative to low-fat (LF) fed controls . Nevertheless, together these findings exemplify a decreased ability of obese rodents to detect oral fats, which is associated with increased fat intake. Despite these data, one study in inbred obese rats, which prefer HF foods over LF foods, shows that these rats exhibit increased sensitivity to linoleic acid relative to obese resistant animals during LF-feeding . However this effect is in conflict with previous data demonstrating that the same obese prone animals exhibited decreased inhibition of delayed-rectifying K+ channels, which would result in decreased activation of taste receptor cells compared to the lean controls . Surprisingly, lingual CD36 in this model has yet to be examined, and may play a significant role in these behavioral findings. Unlike sweet taste, obesity in humans is well correlated with increased affinity for fats [156, 167, 168], with the analyses of possible genetic contributions currently under investigation. While initial data from a European population suggested no differences in BMI or oral fat sensitivity in individuals displaying CD36 polymporphisms , more recent data demonstrates that polymorphisms in CD36 are associated with fat perception and BMI [169, 170].
While oral detection of sweet and fat are thought to be largely stimulatory in regards to food intake, post-oral detection of nutrients, involving predominantly gastric and intestinal feedback typically serve to terminate a meal. The temporary distention of the gastric wall observed upon the entrance of ingesta in the stomach may be the most important means of the stomach to regulate food intake. For example, humans with naturally occurring gastric fistulas remain hungry following a meal . As well, rats with man-made esophageal  or gastric  fistulas consume food continuously when food drains from a cannula. However, in the same animals, closing the cannula to allow passage of ingesta into the stomach and intestine rapidly reduces food intake . While these data do not exclude the contribution of intestinal satiation, experiments that prevent gastric using an inflatable pyloric cuff have shown that gastric loads produce volume-related suppression of liquid diet intake [175, 176]. Thus, the distention of the stomach alone is sufficient for the termination of a meal. Furthermore, the finding that this reduction in intake is extraneous to nutritive value as intragastric delivery of non-nutritive loads suppresses food intake similarly to nutritive loads of the same volume  demonstrates the mechanical, rather than chemical nature of this process in inducing satiation.
Table of contents :
1.1 Obesity and healthcare costs
1.2 The digestive system
1.2.1 Oral cavity
18.104.22.168 Sweet taste
22.214.171.124.1 Sweet taste and feeding behavior
126.96.36.199.2 Mechanisms of oral sweet detection
188.8.131.52 Fat taste
184.108.40.206.1 Fat taste and feeding behavior
220.127.116.11.2 Mechanisms of oral fat detection
18.104.22.168 Taste transduction, central signaling, and reward
22.214.171.124 Obesity and taste
126.96.36.199 Gastric distention
188.8.131.52 Endocrine function
184.108.40.206.1 Gastric leptin
220.127.116.11 Intestinal nutrient satiation
18.104.22.168 Intestinal carbohydrates
22.214.171.124.1 Intestinal carbohydrates and feeding behavior
126.96.36.199.2 Mechanisms of intestinal carbohydrate detection
188.8.131.52 Intestinal fats
184.108.40.206.1 Intestinal fats and feeding behavior
220.127.116.11.2 Mechanisms of intestinal fat detection
18.104.22.168 Intestinal nutrients, conditioned preferences, and reward
22.214.171.124 Obesity and intestinal nutrient satiation
126.96.36.199 Vagal afferents and intstinal nutrient satiation
188.8.131.52 Intestinal satiety peptides
184.108.40.206.1 Cholecystokinin (CCK)
220.127.116.11.2 Glucagon-like Peptide-1 (GLP-1)
18.104.22.168.2 Peptide YY (PYY)
22.214.171.124 Hepatic nutrients and feeding behavior
126.96.36.199 Hepatic metabolism and feeding behavior
188.8.131.52 Obesity and hepatic metabolism
1.3.1 Microbiota and energy harvest
1.3.2 Microbiota and intestinal morphology
1.3.3 Microbiota and intestinal endocrine function
1.3.4 Microbiota and inflammation
1.3.5 Microbiota and obesity
1.4 Overall significance and experimental outline
1.4.1 Specific Aim 1
1.4.2 Specific Aim 2
1.4.3 Specific Aim 3
2 Specific Aim 1: How the absence of gut microbiota affects oral and postoral detection of sweet nutrients as well as intestinal sugar transporters
2.2 Up-regulation of intestinal type 1 taste receptor 3 and sodium glucose luminal transporter 1 expression and increased sucrose intake in mice lacking gut microbiota
2.3 Summary of results and conclusions
3 Specific Aim 2: How gut microbiota affects oral and post-oral detection of lipids and associated changes in nutrient-responsive g-protein coupled receptors (GPRs) and intestinal satiety peptides
3.2 Increased oral detection, but decreased intestinal signaling for fats in mice lacking gut microbiota
3.3 Summary of results and conclusions
4 Specific Aim 3: How gut microbiota affects host liver, intestinal, and adipose metabolic parameters in a GF rat model
4.2 Absence of gut microbiota is not protective of fat deposition in the GF F344 rat model
4.3 Specific Aim 3: Summary of results and conclusions
5 General summary
5.1 General results
5.2 General discussion
5.3 General conclusions
6.1.1 Chapter 1
6.1.2 Chapter 3
6.1.3 Chapter 4
6.1.4 Chapter 5
6.2.1 Chapter 4