NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

Montmayeur JP, le Coutre J, editors. Fat Detection: Taste, Texture, and Post Ingestive Effects. Boca Raton (FL): CRC Press; 2010.

Cover of Fat Detection

Fat Detection: Taste, Texture, and Post Ingestive Effects.

Show details

Chapter 12Oral and Postoral Determinants of Dietary Fat Appetite

and .

12.1. INTRODUCTION

12.1.1. General Issues in the Study of Fat Appetite

Dietary fat is an important determinant of palatability and energy density in a wide variety of foods. The appetite for fat can be traced to both unlearned attraction to orosensory factors and learned appreciation of high-fat foods based on postoral factors. While these two factors normally operate together, they can be separated experimentally by various techniques. This chapter describes the appetite for dietary fat as a function of oral and postoral effects, and discusses the emerging information about the genetics of fat appetite. Much of the data comes from animal models (rats and mice), but studies of human fat appetite are also noted.

Understanding the basis for fat appetite is important, as it is thought to contribute to the increasing levels of obesity in affluent nations. This stems in part from the higher metabolizable energy content of fat, which is more than twice that of carbohydrate and protein. Foods that are high in fat are thus more energy-dense than lower-fat foods (an exception might be high-fat foods that are also high in water or fiber content). Inexpensive, palatable high-fat foods are widely available. In addition, high-fat food is often found to be less satiating than an equal volume of high-carbohydrate food, which can exacerbate the overconsumption of high-fat foods.

The measurement of appetite for fat can be done by comparing intake of fat to that of other foods presented at the same time (preference) or by absolute amounts consumed, as compared to other foods offered at other times (acceptance). A potential drawback of foods with a mixture of nutrients is that it is difficult to ascribe preference or acceptance to particular components of the food. Accordingly, much of the work that evaluates fat appetite uses isolated nutrients to simplify the analysis. However, foods can be constructed to vary in fat and carbohydrate content, while holding the other nutrients constant, which makes them reasonable alternatives to single-nutrient analysis.

12.1.2. Rodent Models to Study Orosensory and Visceral Determinants of Fat Appetite

Rodent models of feeding behavior have provided extensive information on the origin of fat appetite and the metabolic consequences of eating high-fat diets. Early studies focused on the weight gains and adiposity produced by feeding rats and mice on high-fat composite diets, i.e., a single diet varying in fat content (e.g., Corbit and Stellar 1964; Mickelsen et al. 1955). Later experiments investigated the feeding responses of rodents given access to a pure fat source such as vegetable shortening (e.g., Corwin et al. 1998; Lucas et al. 1989), corn oil, or oil emulsion (Lucas et al. 1989; Takeda et al. 2001) in addition to their low-fat maintenance diet. High-fat and fat option diets typically, but not always, promote overeating, which led to investigations of the determinants of fat-induced overeating.

High-fat diets may increase energy consumption in part because the conversion of dietary fat to body fat reduces its availability for oxidation (Ramirez et al. 1989). That is, foods high in fat may be less satiating than low-fat foods. In addition, high-fat diets may have a palatable flavor that promotes overeating and postoral rewarding effects that, via a conditioning process, further increase the attractiveness of the diet’s flavor. The availability of inbred mouse strains that differ in their avidity for fat as well as the development of knockout (KO) mice missing components of taste transduction systems have provided new data on the orosensory controls of fat intake. The following sections review the evidence that both orosensory (flavor) and postoral (conditioning) factors are important determinants of fat appetite in rodents (see Figure 12.1).

FIGURE 12.1. Oral and postoral determinants of fat appetite.

FIGURE 12.1

Oral and postoral determinants of fat appetite. Both unlearned and learned sources of flavor reward contribute to the appetite for fat. Orosensory detection of fat flavor has olfactory, somatosensory (trigeminal), and taste components. Candidate taste (more...)

12.2. OROSENSORY FACTORS

12.2.1. Fat Preference and Texture

12.2.1.1. Nutritive vs. Nonnutritive Oil

In choice tests, rats typically prefer high-fat to low-fat foods (Hamilton 1964; Rockwood and Bhathena 1990). This preference is based, at least initially, on the flavor (taste, texture, and odor) of the high-fat food because it is apparent in the very first test session. Furthermore, rats prefer foods made oily or greasy with the addition of noncaloric mineral oil or petrolatum over dry, powdered food in choice tests (Carlisle and Stellar 1969; Hamilton 1964). These early findings indicated that the oily/greasy texture of fat was palatable to rats even when separated from the postoral nutritive actions of fat. Hamilton (1964) observed that rats given the choice between a 30% fat (lard) diet and a 30% petrolatum diet initially consumed similar amounts but developed a strong preference for the fat diet over the course of several days. Similar results were observed in rats given the choice of a 25% corn oil diet vs. a 25% mineral oil diet: rats equally preferred the diets for 2 days but then developed an 80% preference for the corn oil diet by day 8 of testing (Carlisle and Stellar 1969). The late developing preference for the fat-rich food over the food containing the non-nutritive fat substitute provided early evidence that the postoral effects of dietary fat influence food preferences in animals via a learning process.

More detailed studies of the orosensory contribution to fat appetite have been conducted using nutritive and nonnutritive oils presented as emulsions or suspensions. Rats as young as 12 days of age displayed comparable ingestive responses to intraoral infusions of 30% corn oil and 30% mineral oil emulsions, which exceeded their response to water infusions (Ackroff et al. 1990). Ingestion by 21-day-old weanling rats was stimulated by corn oil emulsions as low as 6.25% and peak intakes were obtained with a 50% emulsion (Ackerman et al. 1992). Adult rats with no prior experience with oils consumed equivalent amounts of 30% corn oil and 30% mineral oil emulsions during 3 min one-bottle intake tests that minimized postoral nutritive effects (Ackroff et al. 1990). The rats displayed a slight preference for corn oil over mineral oil emulsions in 3 min two-bottle tests, which developed into a small (62%) but significant preference when the test session was extended to 30 min. Food restriction, i.e., maintenance on limited daily chow rations, increased the corn oil emulsion preference to 80%. This strong preference persisted when the animals were no longer food restricted. Other oil-naive adult rats displayed similar deprivation-induced increases in corn oil and mineral oil emulsion intakes during daily 30 min one-bottle tests. These findings indicate that nutritive and nonnutritive oil emulsions are equally acceptable to neonatal and adult rats, that nutritive oil is only slightly preferred by freely fed animals in short-term choice tests, and that deprivation increases the acceptance of both oils. However, when food deprived, animals develop a strong preference for the nutritive oil.

The potent feeding stimulatory effects of nutritive and nonnutritive oil have also been revealed in sham-feeding experiments. With this preparation, the ingested fluid drains out an open gastric fistula which greatly reduces but does not completely eliminate postoral nutritive effects (Mindell et al. 1990). Food-restricted rats with no prior oil experience displayed identical increases in their one-bottle sham-feeding intakes of pure corn oil and mineral oil over successive test sessions. In two-bottle tests, however, the rats preferred the corn oil to the mineral oil, which indicates that in pure form, the nutritive oil has a more attractive flavor than the nonnutritive oil. The sham-feeding response to different corn oil emulsions revealed that a 25% corn oil emulsion stimulates greater intakes than do higher (50% or 100%) and lower (0.78%–12.5%) concentrations. Long-term (24 h/day) real-intake tests have also revealed higher intakes of midrange concentrations of oil emulsions than higher or lower concentrations (Castonguay et al. 1984; Lucas et al. 1989).

The findings that neonatal and adult rats with no prior experience with corn oil or mineral oil display ingestive responses to these oils when first exposed suggest that there is an unlearned attraction to the orosensory properties of oil (Ackerman et al. 1992; Ackroff et al. 1990; Mindell et al. 1990). However, it is possible that the rat’s response to oil flavor is conditioned by the animal’s very early experience with mother’s milk (Ackerman et al. 1992; Ackroff et al. 1990). Newborn rat pups without natural suckling experience ingest more milk (commercial “half and half”) and sweet solutions (saccharin or sucrose) than water from a surrogate nipple (Petrov et al. 2004). A comparison of the ingestive responses of milk-naive and milk-experienced rat pups to oil emulsions would be most informative on the issue of the innate basis of fat appetite.

The ingestive responses of mice to different oils and oil emulsions have been investigated in some detail. In 10 min two-bottle tests, mice (ddY strain) displayed strong preferences for 1% emulsions of corn oil, canola oil, or vegetable oil over the xanthan gum vehicle (Takeda et al. 2000). In another study, ddY mice displayed comparable preferences for 2% nutritive corn oil and 2% nonnutritive sorbitol fatty acid ester oil over xanthan gum vehicle during 10 min tests. They also displayed equal preferences for the nutritive and nonnutritive oils over a range of concentrations (2%–100%) in 10 min tests. In a 24 h choice test with pure oils, however, the mice preferred the nutritive oil, which presumably reflects the differential postoral effects of the two oils (Suzuki et al. 2003). In contrast to these results, BALB/c mice consistently preferred corn oil to mineral oil in 10 min two-bottle tests at corn oil concentrations of 1%–100% (Yoneda et al. 2007). An unusual aspect of this test was that the animals had the choice between different concentrations of corn oil mixed into mineral oil vs. pure mineral oil. This suggests that mice, like rats, prefer nutritive to nonnutritive oils when the oils are in pure form but not necessarily when they are presented as emulsions or suspensions.

Some work has attempted to determine the particular textural quality that rodents use to discriminate oils. Ramirez (1994) showed that rats trained to avoid an oil suspension (triglyceride, silicone, or mineral oil) generalized their avoidance to the other oils but not to a nonoily gum with similar viscosity. These results suggested that a textural property other than viscosity, such as lubricity, was critical for oil detection and preference. Consistent with this interpretation, whereas nondeprived and deprived rats preferred oil emulsions (0.5%–30%) to water in two-bottle tests, they did not prefer the gum vehicle to water (Kimura et al. 2003). Studies of the central neural response to orosensation in the primate orbitofrontal cortex revealed that individual nerve cells respond to the texture of nutritive and nonnutritive oils but not to nonoily stimuli of similar viscosity (Rolls et al. 1999; Verhagen et al. 2003).

12.2.1.2. Strain Differences: Oil vs. Vehicle Tests

It is well established that inbred mouse strains differ in their preference (or aversion) for sapid solutions including salt, bitter, sweet, and umami, so it is not surprising that they also differ in their preference for oil emulsions (Bodnar et al. 2008). These studies have generally used 24 h two-bottle tests comparing emulsion intake vs. water or the emulsion/suspension vehicle. In an early study, SWR/J mice displayed stronger preferences and higher intakes than did AKR/J mice at several corn oil emulsion concentrations during 24 h two-bottle tests (0.005%–50% oil emulsion vs. emulsifier vehicle) (Smith et al. 2001). In another study, C57BL/6ByJ mice had higher intakes and preferences for soybean oil than did 129P3/J mice at low (1%–10%) but not high (30%–100%) concentrations vs. xanthan gum vehicle (Bachmanov et al. 2001). Similar results were obtained with C57BL/6J and 129 (129P3/J, 129X1/SvJ) mice tested with oil emulsions prepared using Intralipid, a stable, commercially available soybean oil emulsion (Sclafani 2007b, see below). These findings were confirmed and extended in a study of 11 inbred mouse strains tested with a wide range of dilute Intralipid emulsions (0.00001%–5%) vs. water (Lewis et al. 2007). Interestingly, the mouse strains displaying stronger oil preference and/or acceptance in these experiments were those that also showed stronger preferences for sucrose and saccharin (Bachmanov et al. 2001; Bodnar et al. 2008; Smith et al. 2001). However, an earlier study comparing macronutrient intake patterns in 13 inbred mouse strains found that variations in fat selection appeared unrelated to sweetener preferences (Smith et al. 2000).

Strain differences in sweetener preference are due in part to allelic variations in the Tas1r3 gene that codes for the T1R3 sweet taste receptor. Glendinning et al. (2008) explicitly examined the relationship between the avidity for sugar and oil by comparing the initial (5 s) licking response to sucrose and Intralipid in inbred strains with the sensitive form of the T1R3 receptor (C57BL/6J, FVB/NJ, SWR/J, SM/J) and the subsensitive form (129P3/J, BALB/cJ, DBA/2J, AKR/J). Consistent with prior work, the sweet sensitive strains licked more vigorously than did the sweet sub-sensitive strains for low sucrose concentrations (≤ 0.3 M, ~10%). There was, however, no clear relationship between sweet taste sensitivity and the licking response to the oil emulsions (1%–20%). Nevertheless, there were significant strain differences in the initial oil licking as well as in 24 h oil preferences (Intralipid vs. water). Licking rates and 24 h oil preferences were highly correlated for 1% oil but not at the higher concentrations.

Taken together, these data indicate that responsiveness to orosensory cues primarily determines the intake of dilute oil emulsions but that postoral mechanisms influence the intake of more concentrated oil emulsions. This is not unique to fat consumption because sweet taste sensitivity does not fully account for the intake of concentrated sugar solutions (Glendinning et al. 2005; Sclafani 2006). Further evidence that sweet taste sensitivity does not account for oil consumption is indicated by the finding that mice lacking the T1R3 sweet receptor, which are indifferent to dilute sugar solutions, do not differ from wild-type (WT) mice in their intake and preference for 1% Intralipid (S. Zukerman and A. Sclafani, unpublished findings).

The source of the mouse strain differences in the orosensory response to oils is not known, but it may be related in part to differences in oral somatosensation. In the Intralipid study mentioned above (Sclafani 2007b), C57BL/6J (B6) and 129 (129P3/J, 129X1/SvJ) mice were first tested with the nonnutritive oil Sefa Soyate, a sucrose polyester oil used in the preparation of olestra. In 24 h oil vs. vehicle tests, nonnutritive oil intake and preference were greater in B6 mice than 129 mice at 0.313%–2.5% concentrations. Similar strain differences were obtained with low (0.313%–5%) but not high (10%–20%) Intralipid concentrations. However, when retested for their Intralipid preferences, the B6 and 129 mice showed robust (>90%) preferences for all oil concentrations, although B6 mice still consumed more oil at low concentrations. A second test with the nonnutritive oil revealed increased preference and acceptance in the B6 and 129X1/SvJ mice. The nonnutritive oil preference and intake were much less than that for the nutritive oil. We attributed this experiential effect to a conditioned increase in the animals’ evaluation of oil flavor due to the postoral nutritive effects of Intralipid. As discussed below, odor or taste stimuli may mediate the increased preference for nutritive over nonnutritive oils.

12.2.2. Fat Detection and Odor

Although pure, processed oils have virtually no odor, oil and other fats develop odors as they interact with oxygen, enzymes, and chemicals in food (Mela and Marshall 1992). This led to the conjecture that animals could distinguish among nutritive oils, and discriminate them from nonnutritive oils, on the basis of odor. Larue (1978) investigated the role of olfaction in fat detection using a conditioned aversion paradigm. Rats were poisoned after they consumed a 20% fat chow (butter, margarine, lard) or chow containing a nonnutritive fat substitute (20% vaseline). In rats with intact olfaction, aversions generalized from one nutritive fat to others but not to vaseline. Rats made anosmic with zinc sulfate treatment showed smaller aversions when trained with vaseline and little difference from intact rats when trained with the nutritive fats. These results were taken as evidence against the idea that a fatty odor is importantly involved in fat appetite.

Other studies investigated the role of olfaction in the preference for oil emulsions. Rats made anosmic by removal of the olfactory bulb lost their preference for 0.5% corn oil or triolein oil emulsions but not for 1% oil emulsions (Ramirez 1993). In contrast, mice with zinc-induced anosmia, unlike control mice, did not prefer 1% and 3% corn oil emulsions, although they did prefer 5% and 10% oil emulsions (Takeda et al. 2001). Anosmic mice, unlike control mice, also failed to prefer triolein oil (0.1%–2%) or oleate (0.1%–0.2%) to xanthan gum vehicle (Fukuwatari et al. 2003). In evaluating these results, however, it should be noted that anosmia in rats also reduced the preference for dilute sucrose, starch, and Polycose (maltodextrin) solutions (Ramirez 1993) while anosmia in mice blocked the preference for a concentrated sucrose solution (0.5 M) (Uebayashi et al. 2001). Thus, experimentally induced anosmia produces a generalized reduction in responsiveness to nutrient flavors.

Distinguishing nutritive from nonnutritive fat may be aided by olfaction when the fat is embedded in a mixed food. After mice given corn oil and mineral oil adulterated chows acquired a preference for the corn oil version, the olfactory nerve was sectioned in one group. These mice did not distinguish between the nutritive and nonnutritive chows until nerve regrowth occurred, suggesting that mice may rely on odor cues to detect nutritive oil in a mixed food (Kinney and Antill 1996). Postoral conditioning studies discussed below also demonstrate that animals can use odor cues to learn to prefer nutritive oils.

12.2.3. Fatty Acid Taste

Fat had long been considered to be “tasteless” (Mela 1992), but there is accumulating evidence for the existence of a fat taste mediated by a gustatory response to fatty acids (Abumrad 2005; Fukuwatari et al. 1997; Gilbertson et al. 1997; Laugerette et al. 2005). In rodents, the taste of fat depends on the presence of long-chain free fatty acids. In 5 min two-bottle tests, rats preferred 1% linolenic acid, linoleic acid, and oleic acid, in that order, to xanthan gum vehicle but did not prefer caprylic acid, a medium-chain fatty acid (Tsuruta et al. 1999). The attraction to triglycerides appears to be mediated, in part, by fatty acids released in the mouth by the action of salivary lipase (Kawai and Fushiki 2003). Rats preferred a 2% emulsion of purified triolein to a triolein emulsion containing the lipase inhibitor orlistat during a 5 min two-bottle choice. In contrast to this finding, rats equally preferred a 2% corn oil emulsion and a corn oil + orlistat emulsion. This indicates that attractiveness of natural food oils does not depend upon lipolysis in the mouth. Also, long-chain free fatty acids are attractive to rodents at dilute concentrations but they lose their appeal at higher concentrations. In one study, BALB/c mice showed a monotonic increase in their 60 s licking response to corn oil as concentration increased from 1% to 100%, whereas their peak response to linoleic acid occurred at the 1% concentration (Yoneda et al. 2007). In separate 24 h intake tests, C57BL6/J mice consumed substantially more of 2% soybean oil than 2% linoleic acid when both were suspended in a xanthan gum vehicle (19.3 vs. 8.0 g/day) (Sclafani et al. 2007a). Free fatty acids may have irritant effects in the mouth that reduce their palatability at higher concentrations (see Chale-Rush et al. 2007).

Evidence that free fatty acids are detected by the gustatory system is reviewed in the Chapters 3, 4 and 5 by Gilbertson, Pittman, and Contreras. The following section reviews work from our laboratory investigating fat preferences in genetically modified mice missing different taste signaling elements.

12.2.3.1. Fatty Acid Receptors in Taste Buds

In 1997, Fukuwatari et al. reported the localization of the fatty acid binding protein CD36 in taste buds and they speculated that it participates in the oral detection of fat. Laugerette et al. (2005) subsequently confirmed the localization of CD36 in the taste buds of the rodent circumvallate papillae where it was colocalized in taste cells with alpha-gustducin. Further work demonstrated that linoleic acid increases intracellular calcium and neurotransmitter release in CD36-positive taste cells and activates gustatory neurons in the nucleus of the solitary tract (El-Yassimi et al. 2008; Gaillard et al. 2008). Evidence for CD36 involvement in fat preference was indicated by the failure of CD36 KO mice, unlike WT mice, to prefer 2% linoleic acid to xanthan gum vehicle in 0.5 or 24 h two-bottle tests (Laugerette et al. 2005). Linoleic acid preference was also blocked by gustatory nerve transection in WT mice (Gaillard et al. 2008). CD36 is also implicated in the cephalic phase digestive response to fat (Hiraoka et al. 2003). Lingual stimulation with linoleic acid promoted pancreatobiliary secretion in WT mice but not in CD36 KO mice (Laugerette et al. 2005).

We further investigated the role of CD36 in fat appetite by comparing the 24 h preferences of CD36 KO and C57BL/6J WT mice for nutritive soybean oil and non-nutritive Sefa Soyate oil as well as linoleic acid (Sclafani et al. 2007a). In an initial oil vs. vehicle test, CD36 KO mice showed no preference for 0.313%–2.5% Sefa Soyate oil, whereas WT mice preferred Sefa Soyate at the 2.5% concentration. In a second test, CD36 KO mice showed only weak preferences for 0.313%–1.25% soybean oil but displayed a significant preference for 2.5% oil. In contrast, WT mice strongly preferred even the lowest concentration. In a choice between 0.625% soybean and Soyate oils, the B6 mice strongly preferred the nutritive oil whereas the KO mice were indifferent to the oils. However, after experience with the 2.5% soybean oil, the CD36 KO displayed a significant, albeit attenuated, preference for 0.625% soybean oil over Soyate oil. The oil-experienced CD36 KO mice also strongly preferred 0.313%–20% Intralipid although they consumed less oil than the WT mice at 2.5%–20% concentrations.

In confirmation of prior findings (Laugerette et al. 2005), naive CD36 KO mice were indifferent to 1%–2% linoleic acid emulsions preferred by WT mice (Sclafani et al. 2007a). However, after they were tested with, and developed a preference for soybean oil emulsions, the experienced CD36 KO mice preferred a range of linoleic acid emulsions (0.025%–2%) over the xanthan gum vehicle and did not differ from WT mice in their fatty acid preference or intake. Thus, despite their initial deficit in fat and fatty acid preference, the KO mice developed significant preferences for soybean oil and linoleic acid even at low concentrations. This late-developing preference was probably due to a postoral conditioning effect. As discussed below, CD36 KO as well as WT mice learn to prefer flavors paired with intragastric (IG) infusions of soybean oil. The failure of CD36 KO mice to prefer the nonnutritive Soyate oil was surprising, as nonnutritive oil preference is assumed to be mediated by oily texture rather than taste. Conceivably, Soyate oil may have a fat-like taste to WT mice because it consists of fatty acids attached to a sucrose core as well as containing a small amount of unbound free fatty acids. However, in a follow-up experiment, naive CD36 KO mice displayed weaker preferences for dilute (0.313%–2.5%) mineral oil emulsions compared to WT mice and the reason for this deficit is not clear.

The deficits displayed by CD36 KO mice in initial preference tests with soybean oil and linoleic acid support the view that this fatty acid binding protein is involved in fat taste. Nevertheless, our findings that with experience CD36 KO mice were indistinguishable from WT mice in their preference for dilute soybean oil or linoleic acid emulsions demonstrates that this protein is not essential for the detection or preference for fat. It may be that the mice learn to prefer oils by associating their odor or texture (e.g., lubricity) with their postoral nutritive effects. However, it is also possible that rodents have other fatty acid sensors that function independently of CD36. The role of delayed rectifying potassium channels in taste cells in fatty acid detection is discussed in the Chapter 3. The G protein–coupled receptor GPR120, which is thought to be a free fatty acid receptor in the gut (e.g., Hirasawa et al. 2005), was recently identified in the circumvallate papillae where it may serve as a fatty acid taste receptor (Matsumura et al. 2007).

12.2.3.2. Gustducin Signaling

Gustducin is a G protein found in taste cells that mediates the neural and behavioral response to sweet, umami, and bitter tastants (Nelson et al. 2002; Ruiz-Avila et al. 2001). In particular, gustducin KO mice display reduced preferences for sucrose, glutamate, and quinine solutions. Gustducin is colocalized with CD36 in circumvallate taste cells, which suggests a role for the G protein in fatty acid taste (Laugerette et al. 2005). We therefore compared the preference for nutritive and nonnutritive oils of gustducin KO mice with that of C57BL/6J WT mice (Sclafani et al. 2007b). In 24 h two-bottle tests, gustducin KO mice were similar to WT mice in their preference for Sefa Soyate oil (0.313%–2.5%) and the two genotypes showed identical preferences and intakes for soybean oil over a wide range of concentrations (0.039%–20%). Thus, despite its presence in CD36-containing taste cells, these data indicate that gustducin does not have a critical role in fat preference.

12.2.3.3. Trpm5 Channel Signaling

Trpm5 (transient receptor potential family) is a Ca2+ activated cation channel mediating the chemosensory transduction cascade in gustatory sensory cells. Like gustducin, it is important in mediating sweet, umami, and bitter tastes. To determine if Trpm5 is involved in fat appetite, we compared the oil preferences of Trpm5 KO and WT mice (Sclafani et al. 2007b). In initial tests with 0.313%–2.5% oil emulsions, Trpm5 KO mice failed to prefer either soybean oil or Sefa Soyate oil to vehicle. In a subsequent test with soybean oil in the form of Intralipid, Trpm5 KO showed weak preferences for 0.313%–1.25% oil but developed a significant preference for 2.5% and higher concentrations. Nevertheless, they underconsumed 2.5%–20% Intralipid compared to WT mice. Given that the Trpm5 KO mice developed a strong (95%) preference for concentrated Intralipid emulsions, they were retested with dilute emulsions. In this final test series, the Trpm5 KO mice displayed preferences even at the lowest concentration (0.039%); the only difference from WT behavior was lower intake of Intralipid at the highest concentrations in the series (0.625%–2.5%). The similarities to CD36 KO mice (initial indifference to soybean and Sefa Soyate oil, and persistent lower acceptance even after preferences had been acquired) (Sclafani et al. 2007a) suggest that Trpm5 is part of the CD36 fatty acid signaling pathway.

In a subsequent experiment (A. Sclafani, unpublished) that directly compared the oil preferences of CD36 KO, Trpm5 KO, and WT mice, the Trpm5 KO mice displayed the most profound deficit. In an initial test with 0.313%–2.5% mineral oil, Trpm5 KO mice were completely indifferent to the nonnutritive oil whereas CD36 KO mice displayed an attenuated preference compared to WT mice. Unlike CD36 KO and WT mice, the Trpm5 KO mice were indifferent to 0.313%–2.5% soybean oil. Although Trpm5 KO intakes were lower, all strains displayed preferences for 5%–20% oil. Finally, whereas CD36 KO and WT mice preferred 0.625% soybean oil to mineral oil and Sefa Soyate oil, the Trpm5 KO mice did not discriminate between the nutritive and nonnutritive oils. These data indicate that Trpm5 is more essential for the detection of and preference for dilute oil emulsions than is CD36, which is consistent with the possible existence of other fatty acid taste receptors.

12.2.3.4. Purinergic Receptor Signaling

Taste receptor cell communication with other taste cells and gustatory nerves appears quite complex (Roper 2007). Recent evidence indicates that taste receptor cells use ATP as a neurotransmitter to communicate with gustatory nerves via P2X2 and P2X3 heteromeric receptors. P2X2/P2X3 double KO (P2X Dbl KO) mice show no gustatory nerve response to bitter, sour, salty, or umami stimuli. They also showed greatly reduced two-bottle preference/aversion responses to sweeteners, glutamate, and bitter substances (Finger et al. 2005). Given that P2X Dbl KO mice appear to have a near-total ageusia, we investigated their preference response to dietary fat (Sclafani 2007a). In 24 h two-bottle tests with 0.313%–20% soybean oil emulsions vs. vehicle, P2X Dbl KO mice, unlike WT controls, failed to prefer the intermediate 0.625%–5% concentrations. They displayed a marginal preference for 10% oil but a strong preference (93%) for 20% soybean oil comparable to that of the WT mice (94%). When retested with the same oil concentrations, the P2X Dbl KO mice, like WT controls, displayed robust preferences (>90%) for all concentrations, and the two genotypes consumed comparable amounts of oil.

Thus, the soybean oil preference deficit of the P2X Dbl KO mice was slightly greater than that observed with Trpm5 KO mice and more profound than that of CD36 KO mice in our other studies. However, like the Trpm5 KO and CD36 KO mice, the P2X Dbl KO mice displayed near total preferences for even dilute soybean oil emulsions after experience with concentrated oil. The KO mice presumably associated the nontaste flavor components (odor, texture) of the oil emulsions with their postoral nutrient reinforcing actions.

Together with previously reported data (Laugerette et al. 2005), the findings of our KO mouse studies implicate CD36 in the preference for free fatty acids and triglycerides but revealed an even more important role for Trpm5 and P2X2/P2X3 taste signaling elements. In all cases, however, the initial oil preference deficit displayed by these KO mice disappeared after they had an experience with the concentrated nutritive oil emulsions. This indicates that the postoral nutritive effects of fat can overcome orosensory deficits in fat detection. As discussed next, the potent preference conditioning actions of postoral nutrients are documented by the studies involving IG nutrient infusions.

12.3. POSTORAL FACTORS

As discussed in Section 12.2.1, a learned component to fat appetite was first suggested by early reports of rats developing a preference with repeated testing for a high-fat food over a food containing a nonnutritive fat substitute (Carlisle and Stellar 1969; Hamilton 1964). Similar preference shifts were observed in rats and mice tested with nutritive and nonnutritive oils (Ackroff et al. 1990; Suzuki et al. 2003). In addition, mice show significant increases in their preferences for dilute oil emulsions after consuming concentrated nutritive oil emulsions (Sclafani 2007a,b; Sclafani et al. 2007a,b). Such findings provide indirect evidence that postoral effects of fat produced learned changes in the attractiveness of high-fat foods. There are other possible explanations, however. For example, intake of a high-fat food may have digestive and metabolic effects that allow animals to increase their consumption of fat-rich foods and thereby express a preexisting fat flavor preference (e.g., Reed et al. 1991). The postoral conditioning actions of fat are now well established in numerous learning experiments using flavor or place preference paradigms. This section reviews the literature of fat-conditioned flavor preference; a discussion of place preference conditioning by fat is provided in Chapter 10.

12.3.1. Oral Conditioning Studies

Preference conditioning by dietary fat has been formally investigated using a Pavlovian conditioning procedure in which an arbitrary flavor (the conditioned stimulus or CS+, e.g., grape) is associated with a fat source and a different flavor (the CS−, e.g., cherry) is paired with a nonnutritive alternative (e.g., water or a saccharin solution). In an early application of this paradigm, Mehiel and Bolles (1988) trained food-restricted rats to drink, on alternate days, a CS+ flavored corn oil emulsion (3%) and CS− flavored saccharin (0.25%) solution. In a two-bottle choice test with both flavors presented in water, the rats consumed more of the CS+ flavor solution than the CS− flavor solution. Mehiel concluded that the CS+ preference was reinforced by the postoral caloric effects of the fat rather than its flavor, because a separate test indicated that the corn oil emulsion was less preferred than the saccharin solution. In a subsequent study, we obtained a fat-conditioned preference in food-restricted rats trained with a CS+ flavor added to a 7.1% corn oil emulsion (Elizalde and Sclafani 1990). However, we also observed a preference in rats trained with a flavor added to a nonnutritive 7.1% mineral oil emulsion, although the preference was less pronounced than that displayed by the corn oil trained animals (72% vs. 89%). Our results indicated that the mineral oil had a palatable flavor that, even in the absence of postoral nutritive feedback, was sufficient to condition a flavor preference. The process by which a palatable flavor can produce a preference for a neutral flavor is known as “flavor–flavor” conditioning.

In a second experiment, new groups of rats had a CS+ flavor paired with the consumption of corn oil or mineral oil emulsions, but in this case, a delayed conditioning procedure was used. That is, the rats consumed a CS+ flavored saccharin solution for 10 min followed, after a 10 min delay, by the consumption of an oil emulsion without added flavor. (Consumption of a CS− saccharin solution was not followed by another solution.) With this procedure, only rats trained with the corn oil emulsion acquired a preference for the oil-associated flavor (80%, compared to 44% for the mineral oil group) (Elizalde and Sclafani 1990). The effectiveness of nutritive, but not nonnutritive oil to condition a flavor preference over a delay is consistent with earlier results obtained with nutritive (glucose) and nonnutritive (saccharin) sweeteners (Holman 1975). In that study, only rats trained with a CS+ flavor paired with delayed intake of a glucose solution acquired a preference for the CS+ flavor. Other rats trained with a CS+ flavor paired with the delayed intake of a nonnutritive saccharin solution failed to acquire a preference for the CS+ flavor. Flavor-nutrient learning over a delay presumably evolved to tolerate the natural delay that occurs between orosensory stimulation and postoral effects of food.

Delay conditioning is a useful method to separate the oral vs. postoral effects of nutrients and is not unique to fat. We have conditioned similar flavor preferences based on fat, carbohydrate, and protein sources using this procedure (Pérez et al. 1995). An alternate way to separate oral from postoral effects of nutrients is pharmacological inhibition of digestive enzymes. We trained rats to eat two different flavored fat diets (vegetable shortening supplemented with vitamins, minerals, and fiber) during alternate 30 min/day meals (Ackroff and Sclafani 1996). One flavored diet (e.g., grape) contained the lipase inhibitor orlistat and the other flavored diet (e.g., cherry) was drug free (plain diet). In subsequent two-choice tests, the animals consumed substantially more (84%) of flavored plain diet than flavored orlistat diet whether or not the latter diet contained orlistat during the choice test. We attributed the learned flavor preference to the positive postoral effects of the digested fat in the plain diet. Alternately, the animals may have learned an aversion to the flavored orlistat diet due to direct actions of drug or of the undigested fat in the gut. This seemed unlikely, however, because the animals consumed comparable amounts of the two diets during initial training. Another possibility is that the orlistat diet was less tasty than the plain diet because orlistat prevented lingual lipase from producing free fatty acids in the mouth. However, other animals given a two-choice test between unflavored plain and orlistat fat diets consumed comparable amounts during the first two 30 min tests. Thus, as in the case with corn oil (Kawai and Fushiki 2003), inhibiting lingual lipase does not appear to have an immediate effect on the palatability of vegetable shortening.

12.3.2. Gastric Conditioning Studies

A more direct approach to studying postoral conditioning by dietary fat is to pair the intake of a CS+ flavor solution with IG infusion of nutritive oils. This eliminates the possibility that the palatable flavor of the oil contributed to the learning of the CS+ preference. In the first study to report flavor preferences conditioned by IG fat infusions (Lucas and Sclafani 1989), we trained food-restricted rats with a CS+ flavored saccharin solution paired with IG infusions of 7.1% corn oil emulsion, and a CS− flavored saccharin solution paired with IG water during daily 10 min one-bottle sessions; IG infusion volumes were fixed at 7 mL. In a subsequent 10 min choice test conducted without IG infusions, the rats showed a significant, but modest (61%) CS+ preference. Other animals were given 24 h/day access to food and, on alternate days, CS+ and CS− flavored saccharin solutions paired with IG corn oil and water infusions, respectively. In this case, the amount of infused IG was automatically matched to the oral intake of the CS solution, i.e., whenever the animal licked the CS+ or CS− solution, a computer turned on the oil or water infusion pump, respectively. In a subsequent two-bottle choice test during which both CS solutions remained paired with their appropriate IG infusions (reinforced test), the animals displayed a 76% CS+ preference. The rats continued to prefer the CS+ flavor during four additional days of testing during which intake of both solutions was paired with IG water infusions (extinction test). The latter finding demonstrates that, once learned, fat-conditioned flavor preferences are quite persistent even in the absence of continued reinforcement.

We have replicated the effectiveness of IG fat infusions to condition flavor preferences in many experiments that investigated various aspects of the conditioning process. Although significant, the 61% CS+ preference obtained in our original corn oil short-term conditioning experiment was weaker than the preferences (>85%) typically obtained in studies involving IG carbohydrate infusions (Drucker and Sclafani 1997; Lucas and Sclafani 1996a; Lucas et al. 1998b; Sclafani and Lucas 1996). We attempted to improve fat conditioning by using IG infusions of “preingested” fat, that is, a corn oil emulsion that was ingested by a donor rat and then collected from its stomach. The preingested oil contained salivary and gastric secretions thought to facilitate nutrient conditioning (Molina et al. 1977; Puerto et al. 1976). The pre-ingested fat emulsion was no more effective than a fresh emulsion in conditioning a flavor preference (Lucas and Sclafani 1989). We also determined if adding a small amount of oil to the CS solution would enhance IG conditioning by activating fatty taste-induced cephalic reflexes that facilitate fat digestion (see Laugerette et al. 2005). Training animals with 2% corn oil added to the CS solutions did not enhance the conditioning response to IG oil infusions (Lucas and Sclafani 1989). Perhaps, the addition of a free fatty acid (e.g., linoleic) rather than a triglyceride would be more effective in light of the recent findings of Laugerette et al. (2005).

One manipulation that did improve flavor conditioning was feeding animals a high-fat chow diet rather than the standard low-fat chow used in our laboratory. It is well documented that diets high in fat enhance lipase secretion, fat digestion, and absorption as well as voluntary fat consumption (see Reed and Friedman 1990; Reed et al. 1991). We therefore compared flavor conditioning in rats maintained on a low-fat (LFC, 12% fat) standard chow diet with rats fed a high-fat (HFC, 48% fat) chow-oil diet (Lucas and Sclafani, 1996b). The food-restricted rats were trained in 30 min daily sessions with CS+ and CS− solutions paired with matched IG infusions of 7.1% corn oil and water, respectively. By the end of training, the HFC and LFC groups displayed CS+ preferences of 90% and 62%. Interestingly, the rats in the HFC group retained their strong CS+ preference after being switched to the LFC diet. This indicates that a high-fat diet enhances the acquisition of a fat-based preference and is not necessary for its continued expression.

We also determined if fats with different fatty acid compositions (chain length and saturation) differ in their ability to condition flavor preferences (Ackroff et al. 2005). The food-restricted rats were trained (30 min/day) with three flavored solutions: two CS+ flavors were paired with IG infusions of two different fat sources (7.1% emulsion) and a third flavor (CS−) was paired with IG water infusion. The results revealed that corn oil conditioned a stronger preference than did a medium-chain triglyceride mixture (84% vs. 65%). In contrast, emulsions of corn oil, beef tallow, vegetable shortening, or safflower oil produced similar CS+ preference relative to the CS− flavor. In direct comparisons, however, the rats preferred the corn oil-paired CS+ flavor to the CS+ flavors paired with beef tallow and vegetable oil. The flavors paired with corn oil and safflower oil were equally preferred. Taken together, the results demonstrate that a variety of fat sources can condition flavor preferences, but fats with high polyunsaturated content and/or lower saturated fat content are the most effective. Another notable aspect of this study is that all the oils, except for the medium-chain triglyceride oil, produced relatively strong flavor preferences (~85%) even though the animals were fed a low-fat chow diet. This would appear inconsistent with the maintenance diet study cited above. It may be that training animals with two different oils within an experiment facilitates learning in animals on a low-fat diet. Another potential factor is that our ability to prepare stable emulsions has improved over the years as new techniques have been adopted which may enhance the digestion of the infused oil emulsions (Ledeboer et al. 1999).

Mice, like rats, are sensitive to the postoral reinforcing actions of dietary fat. It is possible, therefore, that the different fat preferences observed in inbred mouse strains may be due, in part, to differences in the postoral nutritive effects of fat. We investigated this possibility in C57B6/6J (B6) and 129 mice, which displayed strong and weak preferences, respectively, for dilute nutritive oil emulsions (Bachmanov et al. 2001; Glendinning et al. 2008; Lewis et al. 2007; Sclafani 2007b). The mice were fitted with a chronic gastric catheter and trained to drink flavored CS+ and CS− saccharin solutions paired with IG infusions of 5% Intralipid and water, respectively. The animals had ad libitum access to low-fat chow during the 23 h/day training sessions. In our initial experiment, the CS+ solutions were sweetened with saccharin and the B6 mice acquired a stronger preference (96%) for the CS+ flavor than did the 129 mice (80%). However, the 129 mice drank only half as much of the CS solutions as did the B6 mice during training, which we attributed to their less avid response to the saccharin-sweetened solutions.

In a second experiment, therefore, we matched the attractiveness of the CS solutions for the two strains by adding more sweetener in the solutions offered to the 129 mice. This equated the CS solution training intakes of the two strains and resulted in equally strong CS+ preferences in the 129 mice (98%) and B6 mice (93%). Furthermore, the 129 mice overconsumed the oil-paired CS+ solution relative to the water-paired CS− solution to the same degree as did the B6 mice. This contrasts with the finding that 129 mice consume about half as much of a comparably dilute Intralipid solution (Sclafani 2007b). Taken together, these findings indicate that the B6 and 129 mice are equally responsive to the postoral reinforcing actions of Intralipid and 129 mice underconsume the oil emulsion because of a reduced orosensory response to the emulsion. Whether other inbred strains differ in their oral or postoral response to fat remains to be determined, and the IG conditioning procedure provides a powerful tool to dissect out the origin of the strain differences.

12.3.2.1. Fat vs. Carbohydrate Conditioning

The overconsumption of high-fat foods is often attributed to their palatable flavor, but given the IG conditioning results reviewed above, it is also possible that postoral factors contribute to the palatability of such foods. Yet, in early experiments the flavor conditioning effects produced by IG fat infusions were weaker than those obtained in comparable studies using IG carbohydrate infusions (Elizalde and Sclafani 1990; Lucas and Sclafani 1989, 1996b; Ramirez 1997). To directly compare the postoral reinforcing effects of fat and carbohydrate, we used a conditioning paradigm in which rats were trained with one flavor (the CS+F) paired with IG fat, a second flavor (the CS+C) paired with IG carbohydrate and a third flavor (the CS−) paired with IG water infusions. This allows two kinds of comparisons: the relative strength of preferences for each CS+ over the CS−, and the preference between the CS+F vs. CS+C flavors.

In one experiment (Lucas and Sclafani 1999a), we trained food-restricted rats 30 min/day, with intake of the CS+F solution paired with IG 7.1% corn oil and intake of the CS+C solution paired with IG 16% Polycose (a maltodextrin isocaloric to corn oil at this concentration). Half of the rats were fed a low-fat chow ration (LFC; 12% fat) and the other half was fed a high-fat ration (HFC; 48% fat energy) throughout the experiment. Overall, in the posttraining choice tests, the rats consumed more CS+C and CS+F solution than CS− solution. The preferences, relative to the CS−, were slightly greater for CS+C than CS+F, and in the HFC rats than the LFC rats but the differences were not significant (HFC: CS+C vs. CS+F, 83% and 76%; LFC: 73% vs. 68%). In the critical CS+C vs. CS+F choice test, however, CS+C was significantly preferred by both the HFC (68%) and LFC (70%) groups. We used a similar training procedure in another experiment in which rats were fed restricted rations of low-fat chow or a varying menu of palatable foods including high-fat items (e.g., cookies, cheese, meat). Both diet groups significantly preferred the CS+C to the CS+F (76%–78%). Other investigators reported a similar preference for CS+C over CS+F in chow-fed rats (Tracy et al. 2004).

To determine if food restriction and/or short-training sessions influenced the preference for a carbohydrate-paired flavor over a fat-paired flavor, we studied flavor conditioning in nondeprived rats trained 22 h/day. The animals, which had ad libitum access to either a LFC or HFC diet, were trained with CS solutions paired with IG infusions of fat (7.1% corn oil), carbohydrate (16% sucrose), or water (Ackroff and Sclafani 2007). The chow diet did not affect preferences for the CS+F over CS− (86% LFC vs. 82% HFC), but the CS+C vs. CS− preference was stronger in the LFC group (86% vs. 77%). In the choice test with the two nutrient-paired CS+ solutions, the rats in both diet groups preferred the CS+C to the CS+F (60% or 72%). Thus, the results of this series of experiments were quite consistent: irrespective of maintenance diet, food deprivation state, or session length, rats develop stronger preferences for the CS+C than the CS+F. This indicates that the postoral reinforcing actions of fat are less pronounced than those of carbohydrate at the concentrations tested.

12.3.2.2. High-Fat vs. High-Carbohydrate Diet Conditioning

The results obtained with IG fat and carbohydrate infusions would seem to imply that the preference rats display for high-fat over high-carbohydrate diets can not be attributed to the postoral actions of the diets. However, further research revealed that the conditioning actions of fat and carbohydrate depend upon nutrient context and energy density. This research evolved from a seminal study by Warwick and Weingarten (1995) that dissected the oral and postoral determinants of high-fat diet overeating. These investigators developed a pair of isocaloric liquid diets based on evaporated milk supplemented with corn oil and sucrose that were high in fat (HF: 60% of energy) or carbohydrate (HC: 78% of energy), but equal in protein. In an oral feeding experiment, rats fed the HF diet consumed more energy and gained more weight than rats fed the HC diet. In a sham-feeding choice test, which eliminated the postoral effects of the two diets, rats drank significantly more of the HF than HC diet; this demonstrates the greater palatability of the HF diet. In a final IG feeding experiment, rats were given a saccharin solution to drink that was paired with matched IG infusions of either the HF or HC diet as their only nutrient source. The HF rats self-infused more diet and gained more weight than did the HC rats demonstrating that, in the absence of a flavor difference, the postoral actions of the HF diet were sufficient to promote overeating. The investigators hypothesized that the rats overconsumed the saccharin solution paired with the HF diet infusions because the diet was less satiating than the HC diet.

Another possible explanation for the HF-induced overeating, not incompatible with the reduced satiety hypothesis, is that the HF diet has a more reinforcing postoral effect than does the HC diet. This alternative hypothesis would seem inconsistent with the findings summarized above that IG fat infusions condition weaker, not stronger flavor preferences than do IG carbohydrate infusions. However, these data were obtained with pure nutrient infusions and it seemed possible that different results would be obtained with high-fat and high-carbohydrate mixed diets. Therefore, we compared the flavor conditioning produced by IG infusions of the HF and HC diets (Lucas et al. 1998a). Rats were fed chow ad libitum and trained, on different days, with three flavored saccharin solutions (CS+HF, CS+HC, and CS−) paired with IG infusions of HF diet, HC diet, and water, respectively. After 12 training days, flavor preferences were evaluated in a series of two-bottle tests. During training, the rats consumed more of the CS+HF solution than CS+HC solution, and consequently were infused with more HF diet than HC diet. In choice tests with the CS−, the rats strongly preferred both the CS+HF (91%) and CS+HC (84%) flavors. However, in the critical test with the two CS+ solutions, the animals preferred the CS+HF to the CS+HC by 72%.

Because the animals consumed more CS+HF than CS+HC during training, we considered the possibility that their increased familiarity with this flavor was responsible for the CS+HF preference. This was not the case; rats trained with fixed amounts of the CS+HF and CS+HC solutions (30 mL/day) still preferred the CS+HF in a subsequent choice test with the two CS+ solutions. A final experiment determined if the reduced satiating effect of the CS+HF diet was responsible for flavor conditioning effects. Towards this end, rats were trained with an HC diet that was less calorically dense than the HF diet (1.4 vs. 2.1 kcal/mL). This treatment successfully equated the satiating actions of the HC diet and HF diets, as assessed by the similar intakes and meal patterns of CS+HC and CS+HF solutions during training. In the critical CS+HF vs. CS+HC choice test, the rats did not significantly differ in their intake of the two solutions and their CS+HF preference was only 58%.

Subsequent research further explored the training and test parameters that influence the flavor preferences obtained with HF and HC diets. In one study, we examined the effect of deprivation state and session length (Lucas and Sclafani 1999b). Food-restricted rats trained 30 min/day with flavor solutions paired with isocaloric (2.1 kcal/mL) HF and HC diet infusions tended to prefer the HC to the HF paired flavor. Yet, the same rats given 22 h/day training with the same flavors and infusions developed a significant CS+HF preference (72%). However, this CS+HF preference was only expressed in 22 h/day tests; the animals consumed comparable amounts of the CS+HF and CS+HC solutions during the initial 30 min of testing.

A second study further explored the influence of caloric density on the preferences conditioned by HF and HC diets (Ackroff and Sclafani 2006). Ad libitum fed rats were trained 22 h/day with flavored solutions paired with IG infusions of dilute (0.5 kcal/mL) versions of the HF and HC diets. During one-bottle training, the rats consumed similar amounts of the CS+HF and CS+HC flavored solutions paired with these diets. In posttraining choice tests, they strongly preferred the CS+HF (93%) and CS+HC (97%) solutions to the CS− solution, but preferred the CS+HC to CS+HF by 81%. The same animals were next trained with new flavors paired with the calorically dense versions (2.1 kcal/mL) of the diets and they displayed a significant preference for the CS+HF (67%) confirming our original results (Lucas et al. 1998a).

The results of these studies, although complex, were quite consistent. The energy-dense, high-fat liquid diet promoted greater intake than the isocaloric high-carbohydrate diet whether it was consumed by mouth or by IG self-infusion (Ackroff and Sclafani 2006; Lucas and Sclafani 1999b; Lucas et al. 1998a; Warwick and Weingarten 1995). CS flavors paired with infusions of either diet were significantly preferred to a water-paired CS− flavor, with the preference being stronger for the CS+HF solution, and the rats preferred the CS+HF to the CS+HC in 22 h/day choice tests. This preference was not displayed in short (30 min) tests, and food-restricted animals actually preferred the CS+HC to CS+HF. When trained with the energy-dilute versions of the diet, the rats self-infused similar amounts of HF and HC diets, yet, developed a strong preference for the CS+HC over the CS+HF. Finally, when trained with an energy-dense HF diet (2.1 kcal/mL) but diluted HC diet (1.4 kcal/mL), the animals self-infused similar volumes of the two diets (but less HC in calories) and displayed similar preferences for the CS+HF and CS+HC. Thus, the postoral effects of a high-fat mixed diet are not invariably more or less reinforcing than a high-carbohydrate diet. The HF diet was more reinforcing than the HC diet when the diets were energy-dense and available 22 h/day. We attribute this to the reduced satiating action of the energy-dense HF diet relative to the HC diet. As discussed in detail elsewhere (Sclafani and Ackroff 2004), the postoral reinforcing effects of food are separate from their satiating actions which may, in some cases, actually reduce food reinforcement.

12.3.3. Postoral Conditioning Mechanisms

The physiological mechanisms that mediate flavor preference conditioning by dietary fat have been investigated at several levels but remain incompletely understood. One commonly held view is that nutrient reinforcement involves satiety or energy repletion signals. In fact, some investigators hypothesize that the postoral actions of nutrients are reinforcing only in food-restricted animals (Davidson 1998; Harris et al. 2000). Flavor-nutrient learning is often studied in food-restricted animals, but this is not necessary; such learning occurs in animals given unlimited access to food (Ackroff and Sclafani 2006, 2007; Lucas et al. 1998a; Lucas and Sclafani 1989). In one study, we specifically investigated the influence of energy deprivation state on flavor conditioning by fat (Yiin et al. 2005). Rats were fed either unlimited or limited amounts of a high-fat chow diet and trained 30 min/day to drink CS flavored saccharin solutions paired with IG infusions of 7.1% corn oil or water. Both the hungry (food-restricted) and satiated (unrestricted) groups learned preferences for the fat-paired CS+ flavor that did not significantly differ (85% vs. 78%).

It is adaptive for an animal to recognize the postoral nutritive value of a new food even if it is not hungry at the time. However, it is also important for animals with particular nutritional needs to develop preferences for foods that satisfy those needs. Evidence for such nutritional wisdom is provided by the report that diabetic rats with impaired ability to utilize carbohydrates learned to prefer a flavored food associated with the oral intake of a corn oil emulsion over a different flavored food associated with the intake of a sugar solution (Tordoff et al. 1987). Even in the absence of metabolic disorder, preferential intake of high-fat food would still be valuable under many circumstances, allowing the animal to meet its energy requirements more efficiently. The human predilection for energy-dense foods has been interpreted as an evolved response to seasonal food availability (Ulijaszek 2002).

12.3.3.1. Peripheral Sites of Action

The postoral site at which fat acts to condition flavor preferences is not certain. In most conditioning studies, oil emulsions were infused into the stomach, but intraduodenal (ID) infusions of corn oil emulsion are also effective in producing CS+ flavor preferences (Lucas and Sclafani 1996a). This suggests that the stomach is not a critical site of action. Whether ID fat infusions act in the intestines or a postabsorptive site to condition flavor preferences is not known; there are no studies comparing the conditioning effects of intestinal vs. intravenous fat infusions. With respect to satiety, available evidence indicates that fat-induced satiation is preabsorptive: ID, but not intravenous, infusions of Intralipid suppressed sham feeding in rats (Greenberg et al. 1993; Greenberg and Weatherford 1990). As we have noted in Sections 12.3.2.2 and 12.3.3.3, fat satiety effects are distinct from, and may counteract, the reinforcing actions of fat. The satiating effects of fat are further discussed in Chapter 15.

The nutrient specificity of preference learning is apparent in the differential strength of preferences for isocaloric infusions of fats vs. carbohydrates, as reviewed in Section 12.3.2.1. Other indications that rats can discriminate the postoral effects of nutrients were provided by a conditioned taste aversion study (Tracy et al. 2004). Rats were poisoned after IG or ID infusions of carbohydrate (Polycose) or fat (corn oil emulsion) and subsequently showed reduced acceptance of the poisoned nutrient when given the opportunity to consume it orally. These findings were taken as evidence that the intestinal tract can “taste” nutrients in a manner analogous to the lingual gustatory system, but the identity of the intestinal taste receptors was not revealed.

As discussed in Section 12.2.3.1, the fatty acid binding protein CD36 is a putative lingual fat taste receptor. Some investigators also proposed that CD36 functions as a fatty acid detector in the intestinal tract (Chen et al. 2001; Drover et al. 2005; Fukuwatari et al. 1997). We investigated the role of intestinal CD36 in fat-conditioned flavor preferences using CD36 KO and WT mice (Sclafani et al. 2007a). The animals were fitted with chronic IG catheters and trained to associate CS+ and CS− flavored solutions with IG infusions of 5% Intralipid and water, respectively. Although training and testing intakes were lower in KO mice than in WT mice, the animals did not differ in their conditioned preference for the fat-paired CS+ flavor (99% vs. 94%). The calcium channel Trpm5, which appears to be more essential than CD36 in the mediation of oral fat preferences (Section 12.2.3.3) is also found in intestinal cells (Bezençon et al. 2007). In initial oil vs. water two-bottle tests, Trpm5 KO mice were indifferent to nutritive oil emulsions but after extensive experience they developed strong oil preferences (Section 12.2.3.3). These data suggest that the KO mice learned to prefer the oil emulsions based on the postoral reinforcing actions fat. However, the conditioning response of Trpm5 KO mice to IG fat infusions remains to be documented. The role of the GPR120 receptor in the gut in flavor conditioning by IG fat infusions also requires investigation.

12.3.3.2. Peripheral Lesion Studies

The postoral reinforcing signal generated by fats that condition flavor preferences could be transmitted to the brain via neural or humoral pathways. To date, evidence for a neural route is lacking. In one study, we treated rats with capsaicin, a neurotoxin that produces partial visceral deafferentation, and trained them to associate a CS+ flavor with ID 3.5% corn oil infusions (Lucas and Sclafani 1996a). The capsaicin-treated rats acquired a CS+ preference that was lower, but not significantly so, than that of control animals (72% vs. 87%). Intake of the CS+ solution was suppressed by the ID fat infusions in the controls but not in the capsaicin-treated animals, confirming earlier findings (Yox and Ritter 1988) that capsaicin blocks the satiating action of infused lipids.

We also investigated the potential role of vagal transmission in fat conditioning using a selective sensory vagotomy procedure (Sclafani et al. 2003). The rats were given restricted rations of high-fat chow, and ID infusions of 3.5% corn oil and water were paired with the consumption of CS+ and CS− solutions. Deafferented rats and sham controls displayed the same 69% preference for the fat-paired CS+ solution in the posttraining choice tests. In a satiation test, however, only the sham rats suppressed their intake of a palatable solution in response to ID fat infusions. These findings indicated that the selective vagotomy disrupted fat satiation, but not reinforcement. Other groups of rats were given celiac-superior mesenteric ganglionectomy alone (to remove nonvagal sensory afferents) or ganglionectomy plus vagal deafferentation. The combined treatment appeared to block fat conditioning (57% CS+ preference) but the results were inconclusive because the sham controls displayed a rather weak CS+ preference (66%) which did not improve with further training. Thus, vagal afferents do not appear to mediate fat-conditioned flavor preferences, but the role of nonvagal afferents is uncertain.

The only potential hormonal signal examined thus far in flavor preference learning is cholecystokinin (CCK). An early study by Mehiel and Bolles (1988) proposed that CCK released by nutrients in the gut may signal nutrient reinforcement to the brain. Some support for this idea was provided by our finding that rats learned a preference for a flavored solution paired with a low dose of exogenous CCK (Pérez and Sclafani 1991). However, in a subsequent study, we found that the CCK-A receptor antagonist devazepide did not prevent flavor preference conditioned by ID carbohydrate infusions (Pérez et al. 1998). Preliminary results also indicated that flavor conditioning by ID fat infusions is not blocked by devazepide treatment (C. Pérez, F. Lucas, and A. Sclafani, unpublished results).

In addition to CCK, there are several other intestinal hormones released by fat, including apolipoprotein A-IV, GLP-1, PYY, and enterostatin (Woods 2004). These hormones are all implicated in the satiating response to dietary fat. Their role, if any, in fat reinforcement is not known. The capsaicin and vagotomy findings reviewed above suggest that different physiological processes mediate fat satiation and reinforcement, but this requires further study.

12.3.3.3. Central Studies

The central neural structures involved in flavor-nutrient learning have been investigated in recent studies. Lesions of the lateral hypothalamus (LH) (Touzani and Sclafani 2001) reduced the magnitude of conditioned preferences for CS+ flavors paired with IG carbohydrate (67%, compared to 84% in controls) and fat infusions (63%, compared to 95% in controls) relative to the water-paired CS− flavor. Choice tests between the two CS+ flavors revealed CS+ carbohydrate preferences in the lateral hypothalamic lesion (67%) and control (84%) groups. Thus, LH lesions attenuated but did not eliminate nutrient-based learning. Another study (Touzani and Sclafani 2002) showed that lesions of the area postrema, which is involved in flavor aversion learning, did not alter conditioned preferences for a CS+ flavor paired with 7.1% corn oil infusions (80% AP lesioned, 74% controls).

The amygdala is also implicated in flavor aversion learning; its combination of orosensory and viscerosensory inputs makes it a potentially important site for flavor learning as well. Lesions of the basolateral nucleus reduced but did not eliminate preferences for carbohydrate-paired flavors (78%, vs. 91% in controls) or fat-paired flavors (73%, vs. 89% in controls) (Touzani and Sclafani 2005). Other rats given large lesions that destroyed the central, corticomedial, and basolateral nuclei, failed to learn a carbohydrate-based preference (52%, compared to 95% in controls) or a fat-based preference (54%, compared to 93% in controls). Further work with carbohydrate infusions in animals with large amygdala lesions showed that they could acquire taste-nutrient preferences (bitter-sweet and salty-sweet CS solutions, 74%, vs. 86% in controls) but not flavor-nutrient (sweet-sour odors, saccharin Kool-Aid, nonsignificant 63%, vs. 87% in controls). This indicated that the preference learning deficit was specific to orosensory cues: the large lesions prevented learning involving olfactory cues, but had only minor effects on association of taste cues with postoral nutrient effects.

Some studies have tested the involvement of central circuits only for carbohydrate-conditioned flavor preferences. Lesions of insular cortex, which receives both gustatory and visceral input, did not prevent flavor conditioning with IG carbohydrate infusions (Touzani and Sclafani 2007). Parabrachial nucleus lesions blocked taste but not flavor preference conditioning with IG carbohydrate infusions, suggesting that its importance is primarily in the integration of gustatory cues with postoral information (Sclafani et al. 2001). Pharmacological blockade of dopamine D1 receptors in nucleus accumbens shell and core prevented the acquisition of preference for flavors paired with IG carbohydrate infusions (Touzani and Sclafani 2008). More work is needed to determine whether these and other areas play a role in fat reinforcement of flavor preferences.

12.4. HUMAN STUDIES

The orosensory response of humans to dietary fat has been studied extensively and is reviewed in Chapters 7 and 11. On the other hand, very few studies have been designed to investigate conditioned changes in food preferences based on the postoral actions of dietary fat in humans. Two early studies from Leann Birch’s laboratory have provided evidence that children acquire preferences for the flavors of high-fat yogurt drinks (Johnson et al. 1991; Kern et al. 1993). The investigators trained children as young as 2–3 years of age to indicate their ranking for flavors, using simple cartoon faces (smiling, neutral, frowning) to establish preferred, neutral, and unpreferred flavors. Following an initial assessment, the children repeatedly consumed two flavors of yogurt drink on separate occasions, in a design parallel to our rodent studies. One flavor was high-fat, and a serving provided about twice the energy of the alternate low-fat flavor; the textural characteristics of the two drinks were matched so that the target flavors were the salient cues. Posttraining evaluation of the high-fat flavor was enhanced but the ranking of the low-fat flavor did not change significantly. Children also ate less in test meals following the high-fat yogurt (Johnson et al. 1991); this effect was only apparent after training, suggesting that it was a learned response. When 4–5-year-old children were trained while hungry (overnight fasted), there was enhanced ranking of the high-fat flavor when the children were also hungry at test, but not when they had just eaten a meal (Kern et al. 1993). Retested 2 months later in the hungry state, the children still tended to show enhanced preference for the high-fat flavor. A recent study of college students investigated flavor conditioning using high-fat and low-fat cream cheese spreads (Capaldi and Privitera 2007). Both spreads contained a cue flavor (orange or banana extract) as well as a bitter taste (quinine). The bitter taste was added to reduce the possibility that a conditioned preference for the cue flavor was due to its association with the palatable flavor of the cream cheese spread. Subjects trained with the high-fat spread rated the pleasantness of the cue flavor higher than those trained with the low-fat spread, which the authors took as evidence for flavor-nutrient learning.

12.5. CONCLUSIONS

Studies of fat appetite in animals have demonstrated clearly that there are both oral and postoral sources of fat reward. Recent advances in knowledge of fat detection have begun to reveal the mechanisms for orosensory appetite for fats. Research findings obtained with KO mice missing taste signaling elements (present chapter) and with gustatory nerve transected animals (Chapters 4 and 5) demonstrate the involvement of the gustatory system in fat preference. The olfactory system is implicated in fat appetite by studies of animals with experimentally induced anosmia. The trigeminal system is also importantly involved in the detection of and preference for dietary fat, and new findings are noted in Chapter 3. The postoral influences on fat preference and acceptance are well documented by experiments involving IG fat infusions in rats and mice. However, the transduction of appetitive information about fat in the gut has been more elusive, and thus far little is known about how this information reaches the brain. Further work in animals should fill the gaps in knowledge and contribute to a better understanding of fat appetite.

Both animal and human studies suggest that the postoral consequences of dietary fat, by enhancing preferences for their associated flavors, could contribute to the appetite for high-fat foods. However, it should be noted that the human studies do not establish a clear role for fat per se, since energy density covaried with fat content. Studies comparing the effects of isocaloric high-fat and high-carbohydrate foods on flavor preferences are needed, to determine whether humans find these nutrients differentially rewarding in the absence of energy differences.

ACKNOWLEDGMENT

The authors’ research was supported by National Institute of Diabetes and Digestive and Kidney Diseases Grant DK-31135.

REFERENCES

  1. Abumrad NA. CD36 may determine our desire for dietary fats. J Clin Invest. 2005;115:2965–2967. [PMC free article: PMC1265882] [PubMed: 16276408]
  2. Ackerman SH, Albert M, Shindledecker RD, Gayle C, Smith GP. Intake of different concentrations of sucrose and corn oil in preweanling rats. Am J Physiol. 1992;262:R624–627. [PubMed: 1566927]
  3. Ackroff K, Sclafani A. Effects of the lipase inhibitor orlistat on intake and preference for dietary fat in rats. Am J Physiol. 1996;271:R48–54. [PubMed: 8760203]
  4. Ackroff K, Sclafani A. Energy density and macronutrient composition determine flavor preference conditioned by intragastric infusions of mixed diets. Physiol Behav. 2006;89:250–260. [PubMed: 16854441]
  5. Ackroff K, Sclafani A. Dietary fat content and flavor reinforcement by intragastric sucrose and corn oil in rats. Appetite. 2007;49:272.
  6. Ackroff K, Vigorito M, Sclafani A. Fat appetite in rats: The response of infant and adult rats to nutritive and non-nutritive oil emulsions. Appetite. 1990;15:171–188. [PubMed: 2281948]
  7. Ackroff K, Lucas F, Sclafani A. Flavor preference conditioning as a function of fat source. Physiol Behav. 2005;85:448–460. [PubMed: 15990126]
  8. Bachmanov AA, Reed DR, Tordoff MG, Price RA, Beauchamp GK. Nutrient preference and diet-induced adiposity in C57BL/6ByJ and 129P3/J mice. Physiol Behav. 2001;72:603–613. [PMC free article: PMC3341942] [PubMed: 11282146]
  9. Bezençon C, le Coutre J, Damak S. Taste-signaling proteins are coexpressed in solitary intestinal epithelial cells. Chem. Senses. 2007;32:41–49. [PubMed: 17030556]
  10. Bodnar RJ, Lewis SR, Kest B. Section 33: Feeding and drinking, in Section 3: Autonomous and motor behaviors, in Volume I: Behavioral Genetics of the Mouse. In: Crusio WE, Sluyter F, Gerlai RT, editors. Encyclopedia of Behavior Genetics. Cambridge: Cambridge University Press; 2008.
  11. Capaldi ED, Privitera GJ. Flavor-nutrient learning independent of flavor-taste learning with college students. Appetite. 2007;49:712–715. [PubMed: 17868953]
  12. Carlisle HJ, Stellar E. Caloric regulation and food preference in normal, hyperphagic, and aphagic rats. J Comp Physiol Psychol. 1969;69:107–114. [PubMed: 5259698]
  13. Castonguay TW, Burdick SL, Guzman MA, Collier GH, Stern JS. Self-selection and the obese Zucker rat: The effect of dietary fat dilution. Physiol Behav. 1984;33:119–126. [PubMed: 6505047]
  14. Chale-Rush A, Burgess JR, Mattes RD. Multiple routes of chemosensitivity to free fatty acids in humans. Am J Physiol. 2007;292:G1206–1212. [PubMed: 17234892]
  15. Chen M, Yang Y, Braunstein E, Georgeson KE, Harmon CM. Gut expression and regulation of FAT/CD36: possible role in fatty acid transport in rat enterocytes. Am J Physiol. 2001;281:E916–923. [PubMed: 11595646]
  16. Corbit JD, Stellar E. Palatability, food intake, and obesity in normal and hyperphagic rats. J Comp Physiol Psychol. 1964;58:63–67. [PubMed: 14197042]
  17. Corwin RL, Wojnicki FHE, Fisher JO, Dimitriou SG, Rice HB, Young MA. Limited access to a dietary fat option affects ingestive behavior but not body composition in male rats. Physiol Behav. 1998;65:545–553. [PubMed: 9877422]
  18. Davidson TL. Hunger cues as modulatory stimuli. In: Schmajuk NA, Holland PC, editors. Occasion Setting: Associative Learning and Cognition in Animals. Washington, DC: American Psychological Association; 1998. pp. 223–248.
  19. Drover VA, Ajmal M, Nassir F, Davidson NO, Nauli AM, Sahoo D, Tso P, Abumrad NA. CD36 deficiency impairs intestinal lipid secretion and clearance of chylomicrons from the blood. J Clin Invest. 2005;115:1290–1297. [PMC free article: PMC1074677] [PubMed: 15841205]
  20. Drucker DB, Sclafani A. The role of gastric and post-gastric sites in glucose-conditioned flavor preferences in rats. Physiol Behav. 1997;61:351–358. [PubMed: 9035269]
  21. El-Yassimi A, Hchami A, Besnard P, Khan NA. Linoleic acid induces calcium signaling, SRC-kinase phosphorylation and neurotransmitters release in mouse CD36-positive gustatory cells. J Biol Chem. 2008;283:12949–12959. [PubMed: 18321850]
  22. Elizalde G, Sclafani A. Fat appetite in rats: Flavor preferences conditioned by nutritive and non-nutritive oil emulsions. Appetite. 1990;15:189–197. [PubMed: 2281949]
  23. Finger TE, Danilova V, Barrows J, Bartel DL, Vigers AJ, Stone L, Hellekant G, Kinnamon SC. ATP signaling is crucial for communication from taste buds to gustatory nerves. Science. 2005;310:1495–1499. [PubMed: 16322458]
  24. Fukuwatari T, Kawada T, Tsuruta M, Hiraoka T, Iwanaga T, Sugimoto E, Fushiki T. Expression of the putative membrane fatty acid transporter (FAT) in taste buds of the circumvallate papillae in rats. FEBS Lett. 1997;414:461–464. [PubMed: 9315741]
  25. Fukuwatari T, Shibata K, Iguchi K, Saeki T, Iwata A, Tani K, Sugimoto E, Fushiki T. Role of gustation in the recognition of oleate and triolein in anosmic rats. Physiol Behav. 2003;78:579–583. [PubMed: 12782211]
  26. Gaillard D, Laugerette F, Darcel N, El-Yassimi A, Passilly-Degrace P, Hichami A, Akhtar Khan N, Montmayeur JP, Besnard P. The gustatory pathway is involved in CD36-mediated orosensory perception of long-chain fatty acids in the mouse. FASEB J. 2008;22:1458–1468. [PubMed: 18162488]
  27. Gilbertson TA, Fontenot DT, Liu L, Zhang H, Monroe WT. Fatty acid modulation of K+ channels in taste receptor cells: Gustatory cues for dietary fat. Am J Physiol. 1997;272:C1203–1210. [PubMed: 9142845]
  28. Glendinning JI, Chyou S, Lin I, Onishi M, Patel P, Zheng KH. Initial licking responses of mice to sweeteners: Effects of tas1r3 polymorphisms. Chem. Senses. 2005;30:601–614. [PubMed: 16135742]
  29. Glendinning JI, Feld N, Goodman L, Bayor R. Contribution of orosensory stimulation to strain differences in oil intake by mice. Physiol Behav. 2008;95:476–483. [PubMed: 18691606]
  30. Greenberg D, Weatherford SC. Obese and lean Zucker rats differ in preferences for sham-fed corn oil or sucrose. Am J Physiol. 1990;259:R1093–1095. [PubMed: 2260719]
  31. Greenberg D, Smith GP, Gibbs J. Intravenous triglycerides fail to elicit satiety in sham-feeding rats. Am J Physiol. 1993;264:R409–413. [PubMed: 8447497]
  32. Hamilton CL. Rats’ preference for high fat diets. J Comp Physiol Psychol. 1964;58:459–460. [PubMed: 14241068]
  33. Harris JA, Gorissen MC, Bailey GK, Westbrook RF. Motivational state regulates the content of learned flavor preferences. J Exp Psychol Anim Behav Process. 2000;26:15–30. [PubMed: 10650541]
  34. Hiraoka T, Fukuwatari T, Imaizumi M, Fushiki T. Effects of oral stimulation with fats on the cephalic phase of pancreatic enzyme secretion in esophagostomized rats. Physiol Behav. 2003;79:713–717. [PubMed: 12954413]
  35. Hirasawa A, Tsumaya K, Awaji T, Katsuma S, Adachi T, Yamada M, Sugimoto Y, Miyazaki S, Tsujimoto G. Free fatty acids regulate gut incretin glucagon-like peptide-1 secretion through GPR120. Nat Med. 2005;11:90–94. [PubMed: 15619630]
  36. Holman EW. Immediate and delayed reinforcers for flavor preferences in rats. Learn Motiv. 1975;6:91–100.
  37. Johnson SL, McPhee L, Birch LL. Conditioned preferences: Young children prefer flavors associated with high dietary fat. Physiol Behav. 1991;50:1245–1251. [PubMed: 1798782]
  38. Kawai T, Fushiki T. Importance of lipolysis in oral cavity for orosensory detection of fat. Am J Physiol. 2003;285:R447–454. [PubMed: 12702486]
  39. Kern DL, McPhee L, Fisher J, Johnson S, Birch LL. The postingestive consequences of fat condition preferences for flavors associated with high dietary fat. Physiol Behav. 1993;54:71–76. [PubMed: 8327611]
  40. Kimura F, Endo Y, Fujimoto K. Vigorous intake of oil emulsion caused by chronic food deprivation remains after recovery in rats. Physiol Behav. 2003;78:107–115. [PubMed: 12536017]
  41. Kinney NE, Antill RW. Role of olfaction in the formation of preference for high-fat foods in mice. Physiol Behav. 1996;59:475–478. [PubMed: 8700949]
  42. Larue C. Oral cues involved in the rat’s selective intake of fats. Chem Senses Flav. 1978;3:1–6.
  43. Laugerette F, Passilly-Degrace P, Patris B, Niot I, Febbraio M, Montmayeur JP, Besnard P. CD36 involvement in orosensory detection of dietary lipids, spontaneous fat preference, and digestive secretions. J Clin Invest. 2005;115:3177–3184. [PMC free article: PMC1265871] [PubMed: 16276419]
  44. Ledeboer M, Masclee AA, Biemond I, Lamers CB. Differences in cholecystokinin release and gallbladder contraction between emulsified and nonemulsified long-chain triglycerides. J Parenter Enteral Nutr. 1999;23:203–206. [PubMed: 10421388]
  45. Lewis SR, Dym C, Chai C, Singh A, Kest B, Bodnar RJ. Genetic variance contributes to ingestive processes: A survey of eleven inbred mouse strains for fat (Intralipid) intake. Physiol Behav. 2007;90:82–94. [PubMed: 17028044]
  46. Lucas F, Sclafani A. Flavor preferences conditioned by intragastric fat infusions in rats. Physiol Behav. 1989;46:403–412. [PubMed: 2623061]
  47. Lucas F, Sclafani A. Capsaicin attenuates feeding suppression but not reinforcement by intestinal nutrients. Am J Physiol. 1996a;270:R1059–1064. [PubMed: 8928906]
  48. Lucas F, Sclafani A. The composition of the maintenance diet alters flavor-preference conditioning by intragastric fat infusions in rats. Physiol Behav. 1996b;60:1151–1157. [PubMed: 8884946]
  49. Lucas F, Sclafani A. Differential reinforcing and satiating effects of intragastric fat and carbohydrate infusions in rats. Physiol Behav. 1999a;66:381–388. [PubMed: 10357426]
  50. Lucas F, Sclafani A. Flavor preferences conditioned by high-fat versus high-carbohydrate diets vary as a function of session length. Physiol Behav. 1999b;66:389–395. [PubMed: 10357427]
  51. Lucas F, Ackroff K, Sclafani A. Dietary fat-induced hyperphagia in rats as a function of fat type and physical form. Physiol Behav. 1989;46:937–946. [PubMed: 2780878]
  52. Lucas F, Ackroff K, Sclafani A. High-fat diet preference and overeating mediated by postingestive factors in rats. Am J Physiol. 1998a;275:R1511–1522. [PubMed: 9791068]
  53. Lucas F, Azzara AV, Sclafani A. Flavor preferences conditioned by intragastric Polycose in rats: More concentrated Polycose is not always more reinforcing. Physiol Behav. 1998b;63:7–14. [PubMed: 9402608]
  54. Matsumura S, Mizushige T, Yoneda T, Iwanaga T, Tsuzuki S, Inoue K, Fushiki T. GPR expression in the rat taste bud relating to fatty acid sensing. Biomed Res. 2007;28:49–55. [PubMed: 17379957]
  55. Mehiel R, Bolles RC. Learned flavor preferences based on calories are independent of initial hedonic value. Anim Learn Behav. 1988;16:383–387.
  56. Mela DJ. DietaryFats: Determinants of Preference, Selection and Consumption. London: Elsevier Applied Science; 1992.
  57. Mela DJ, Marshall RJ. Sensory properties and perceptions of fats. In: Mela DJ, editor. Dietary Fats: Determinants of Preference, Selection and Consumption. London: Elsevier Applied Science; 1992. pp. 43–57.
  58. Mickelsen O, Takahashi S, Craig C. Experimental obesity. I. Production of obesity in rats by feeding high-fat diets. J Nutr. 1955;57:541–554. [PubMed: 13278777]
  59. Mindell S, Smith GP, Greenberg D. Corn oil and mineral oil stimulate sham-feeding in rats. Physiol Behav. 1990;48:283–287. [PubMed: 2255732]
  60. Molina F, Thiel T, Deutsch JA, Puerto A. Comparison between some digestive processes after eating and gastric loading in rats. Pharmacol Biochem Behav. 1977;7:347–350. [PubMed: 928492]
  61. Nelson G, Chandrashekar J, Hoon MA, Feng L, Zhao G, Ryba NJ, Zuker CS. An amino-acid taste receptor. Nature. 2002;416:199–202. [PubMed: 11894099]
  62. Pérez C, Sclafani A. Cholecystokinin conditions flavor preferences in rats. Am J Physiol. 1991;260:R179–185. [PubMed: 1992819]
  63. Pérez C, Lucas F, Sclafani A. Carbohydrate, fat and protein condition similar flavor preferences in rats using an oral-delay procedure. Physiol Behav. 1995;57:549–554. [PubMed: 7786349]
  64. Pérez C, Lucas F, Sclafani A. Devazepide, a CCKA antagonist, attenuates the satiating but not the preference conditioning effects of intestinal carbohydrate infusions in rats. Pharmacol Biochem Behav. 1998;59:451–457. [PubMed: 9476995]
  65. Petrov ES, Nizhnikov ME, Kozlov AP, Varlinskaya EI, Kramskaya TA, Spear NE. Repetitive exposures to a surrogate nipple providing nutritive and non-nutritive fluids: Effects on suckling behavior of the newborn rat. Appetite. 2004;43:185–194. [PubMed: 15458805]
  66. Puerto A, Deutsch JA, Molina F, Roll P. Rapid rewarding effects of intragastric injections. Behav Biol. 1976;18:123–134. [PubMed: 985291]
  67. Ramirez I. Role of olfaction in starch and oil preference. Am J Physiol. 1993;265:R1404–1409. [PubMed: 8285284]
  68. Ramirez I. Chemosensory similarities among oils: Does viscosity play a role? Chem. Senses. 1994;19:155–168. [PubMed: 8055265]
  69. Ramirez I. Stimulation of fluid intake by nutrients: Oil is less effective than carbohydrate. Am J Physiol. 1997;272:R289–293. [PubMed: 9039020]
  70. Ramirez I, Tordoff MG, Friedman MI. Dietary hyperphagia and obesity: What causes them. Physiol Behav. 1989;45:163–168. [PubMed: 2657817]
  71. Reed DR, Friedman MI. Diet composition alters the acceptance of fat by rats. Appetite. 1990;14:219–230. [PubMed: 2164352]
  72. Reed DR, Tordoff MG, Friedman MI. Enhanced acceptance and metabolism of fats by rats fed a high-fat diet. Am J Physiol. 1991;261:R1084–1088. [PubMed: 1951757]
  73. Rockwood GA, Bhathena SJ. High-fat diet preference in developing and adult rats. Physiol Behav. 1990;48:79–82. [PubMed: 2236281]
  74. Rolls ET, Critchley HD, Browning AS, Hernadi I, Lenard L. Responses to the sensory properties of fat of neurons in the primate orbitofrontal cortex. J Neurosci. 1999;19:1532–1540. [PubMed: 9952429]
  75. Roper SD. Signal transduction and information processing in mammalian taste buds. Pflugers Arch. 2007;454:759–776. [PMC free article: PMC3723147] [PubMed: 17468883]
  76. Ruiz-Avila L, Wong GT, Damak S, Margolskee RF. Dominant loss of responsiveness to sweet and bitter compounds caused by a single mutation in alpha-gustducin. Proc Natl Acad Sci USA. 2001;98:8868–8873. [PMC free article: PMC37527] [PubMed: 11447270]
  77. Sclafani A. Oral, post-oral and genetic interactions in sweet appetite. Physiol Behav. 2006;89:525–530. [PMC free article: PMC2364709] [PubMed: 16647093]
  78. Sclafani A. Carbohydrate and fat preference in the ‘tasteless’ P2X2/P2X3 knockout mouse. Appetite. 2007a;49:329.
  79. Sclafani A. Fat and sugar flavor preference and acceptance in C57BL/6J and 129 mice: experience attenuates strain differences. Physiol Behav. 2007b;90:602–611. [PubMed: 17210165]
  80. Sclafani A, Ackroff K. The relationship between food reward and satiation revisited. Physiol Behav. 2004;82:89–95. [PubMed: 15234596]
  81. Sclafani A, Lucas F. Abdominal vagotomy does not block carbohydrate-conditioned flavor preferences in rats. Physiol Behav. 1996;60:447–453. [PubMed: 8840905]
  82. Sclafani A, Azzara A, Touzani K, Grigson PS, Norgren R. Parabrachial nucleus lesions block taste and attenuate flavor preference and aversion conditioning in rats. Behav Neurosci. 2001;115:920–933. [PubMed: 11508731]
  83. Sclafani A, Ackroff K, Schwartz GJ. Selective effects of vagal deafferentation and celiac-superior mesenteric ganglionectomy on the reinforcing and satiating action of intestinal nutrients. Physiol Behav. 2003;78:285–294. [PubMed: 12576127]
  84. Sclafani A, Ackroff K, Abumrad NA. CD36 gene deletion reduces fat preference and intake but not post-oral fat conditioning in mice. Am J Physiol. 2007a;293:R1823–1832. [PubMed: 17804586]
  85. Sclafani A, Zukerman S, Glendinning JI, Margolskee RF. Fat and carbohydrate preferences in mice: The contribution of {α}-gustducin and Trpm5 taste signaling proteins. Am J Physiol. 2007b;293:R1504–1513. [PMC free article: PMC2375390] [PubMed: 17652359]
  86. Smith BK, Andrews PK, West DB. Macronutrient diet selection in thirteen mouse strains. Am J Physiol. 2000;278:R797–805. [PubMed: 10749765]
  87. Smith BK, Volaufova J, West DB. Increased flavor preference and lick activity for sucrose and corn oil in SWR/J vs. AKR/J mice. Am J Physiol. 2001;281:R596–606. [PubMed: 11448865]
  88. Suzuki A, Yamane T, Imaizumi M, Fushiki T. Integration of orosensory and postingestive stimuli for the control of excessive fat intake in mice. Nutrition. 2003;19:36–40. [PubMed: 12507637]
  89. Takeda M, Imaizumi M, Fushiki T. Preference for vegetable oils in the two-bottle choice test in mice. Life Sci. 2000;67:197–204. [PubMed: 10901287]
  90. Takeda M, Sawano S, Imaizumi M, Fushiki T. Preference for corn oil in olfactory-blocked mice in the conditioned place preference test and the two-bottle choice test. Life Sci. 2001;69:847–854. [PubMed: 11487096]
  91. Tordoff MG, Tepper BJ, Friedman MI. Food flavor preferences produced by drinking glucose and oil in normal and diabetic rats: Evidence for conditioning based on fuel oxidation. Physiol Behav. 1987;41:481–487. [PubMed: 3432403]
  92. Touzani K, Sclafani A. Conditioned flavor preference and aversion: role of the lateral hypothalamus. Behav Neurosci. 2001;115:84–93. [PubMed: 11256455]
  93. Touzani K, Sclafani A. Area postrema lesions impair flavor-toxin aversion learning but not flavor-nutrient preference learning. Behav Neurosci. 2002;116:256–266. [PubMed: 11996311]
  94. Touzani K, Sclafani A. Critical role of amygdala in flavor but not taste preference learning in rats. Eur J Neurosci. 2005;22:1767–1774. [PubMed: 16197517]
  95. Touzani K, Sclafani A. Insular cortex lesions fail to block flavor and taste preference learning in rats. Eur J Neurosci. 2007;26:1692–1700. [PMC free article: PMC2361095] [PubMed: 17822433]
  96. Touzani K, Sclafani A. Activation of dopamine D1-like receptors in nucleus accumbens is critical for the acquisition, but not the expression, of nutrient-conditioned flavor preferences in rats. Eur J Neurosci. 2008;27:1525–1533. [PubMed: 18336564]
  97. Tracy AL, Phillips RJ, Chi MM, Powley TL, Davidson TL. The gastrointestinal tract “tastes” nutrients: evidence from the intestinal taste aversion paradigm. Am J Physiol. 2004;287:R1086–1100. [PubMed: 15256365]
  98. Tsuruta M, Kawada T, Fukuwatari T, Fushiki T. The orosensory recognition of long-chain fatty acids in rats. Physiol Behav. 1999;66:285–288. [PubMed: 10336155]
  99. Uebayashi H, Hatanaka T, Kanemura F, Tonosaki K. Acute anosmia in the mouse: Behavioral discrimination among the four basic taste substances. Physiol Behav. 2001;72:291–296. [PubMed: 11274669]
  100. Ulijaszek SJ. Human eating behavior in an evolutionary ecological context. Proc Nutr Soc. 2002;61:517–526. [PubMed: 12691181]
  101. Verhagen JV, Rolls ET, Kasohisa M. Neurons in the primate orbitofrontal cortex respond to fat texture independently of viscosity. J Neurophysiol. 2003;90:1514–1525. [PubMed: 12761278]
  102. Warwick ZS, Weingarten HP. Determinants of high-fat diet hyperphagia: experimental dissection of orosensory and postingestive effects. Am J Physiol. 1995;269:R30–37. [PubMed: 7631900]
  103. Woods SC. Gastrointestinal satiety signals I. An overview of gastrointestinal signals that influence food intake. Am J Physiol. 2004;286:G7–13. [PubMed: 14665437]
  104. Yiin Y-M, Ackroff K, Sclafani A. Flavor preferences conditioned by intragastric nutrient infusions in food restricted and free-feeding rats. Physiol Behav. 2005;84:217–231. [PubMed: 15708774]
  105. Yoneda T, Saitou K, Mizushige T, Matsumura S, Manabe Y, Tsuzuki S, Inoue K, Fushiki T. The palatability of corn oil and linoleic acid to mice as measured by short-term two-bottle choice and licking tests. Physiol Behav. 2007;71:304–309. [PubMed: 17459430]
  106. Yox DP, Ritter RC. Capsaicin attenuates suppression of sham feeding induced by intestinal nutrients. Am J Physiol. 1988;255:R569–574. [PubMed: 3140678]
Copyright © 2010, Taylor & Francis Group, LLC.
Bookshelf ID: NBK53540PMID: 21452471
PubReader format: click here to try

Views

  • PubReader
  • Print View
  • Cite this Page

Other titles in this collection

Related information

  • PMC
    PubMed Central citations
  • PubMed
    Links to pubmed

Related citations in PubMed

See reviews...See all...

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...