| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
,1
* Physiologie de la Nutrition, UMR INSERM U 866, Ecole Nationale Supérieure de Biologie Appliquée à la Nutrition et à l’Alimentation, Université de Bourgogne, Dijon, France;
Chimioréception, Centre Européen des Sciences du Goût, UMR 5170 Centre National de la Recherche Scientifique/1214 Institut National de la Recherche Agronomique (INRA)/Université de Bourgogne, Dijon, France;
Physiologie de la Nutrition et du Comportement Alimentaire, Institut National Agronomique Paris Grignon (INAPG), UMR 914 INRA/INAPG, Paris, France; and
Unité Propre de Recherche de l’Enseignement Supérieur (UPRES) 4183 Lipides et Signalisation Cellulaire, Université de Bourgogne, Dijon, France
2Correspondence: Physiologie de la Nutrition, UMR INSERM U866, ENSBANA—Université de Bourgogne, 1, Esplanade Erasme F-21000, Dijon, France. E-mail: pbesnard{at}u-bourgogne.fr
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: dietary lipid perception • gustation • feeding behavior • obesity risk
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, 2)
. Humans display similar patterns of feeding behavior, as obese patients have been shown to have a greater preference for fatty foods than lean ones (3
, 4)
. Reasons for this difference in the feeding behavior remain unclear. Until recently, it has been thought that oral lipid detection takes place only through somesthesic and olfactory cues (5)
. This restrictive view has been challenged by recent observations in laboratory rodents. Indeed, mice and rats continue to display a significant preference for fat, even if olfactory, textural, and postingestive influences are simultaneously excluded (6
, 7)
. Dietary lipids consist mainly of triglycerides, but long-chain fatty acids (LCFAs) seem to be the chemical cue involved in this orosensory perception of lipids. Rats subjected to a two-bottle preference test display a lower appetence for triglycerides and medium-chain fatty acids than for LCFAs (1)
. Moreover, pharmacological inhibition of lingual lipase, which efficiently hydrolyzes triglycerides in rodents, leads to a dramatic decrease in spontaneous fat preference in a free choice situation (8)
.
We have recently shown in mice that the integral membrane lipid-binding protein CD36 plays a major role in this orosensory perception of LCFAs in the mouse (9)
. Indeed, CD36 gene inactivation abolishes the spontaneous preference for LCFA-enriched solutions or solid diets observed in wild-type mice. This effect is lipid specific, as the preference for sweet taste and aversion to bitterness reported in controls were also found in CD36-null mice. Moreover, CD36 gene disruption abolished the short-term increase in the flux and protein content of pancreato-biliary secretions triggered by the direct deposition of LCFAs on the tongue. Distribution of CD36 in lingual mucosa seems to be highly suitable for the perception of LCFAs and the resulting generation of physiological effects. CD36 is found specifically on the apical side of some of the taste receptor cells of gustatory papillae. It is especially strongly expressed in the circumvallate papillae (9)
. Lingual lipase is released directly into the clefts of this papillae in the mouse (8)
. Consequently, a large proportion of the CD36-positive taste bud cells are likely to be found in a microenvironment enriched in LCFAs. The efficiency of this lipid perception system is enhanced by the high binding affinity of CD36 for LCFAs (10)
.
These findings suggest that lingual CD36 is probably either a lipid sensor or involved in a lipid sensor cascade, facilitating both the ingestion and digestion/absorption of fatty foods in the mouse, although this remains to be definitively demonstrated. The CD36-mediated perception of dietary lipids may be due purely to textural (i.e., trigeminal) cues. However, as the addition to control solution of agents mimicking lipid texture does not seem to abolish the attraction for lipids in two-bottle preference tests (6
, 7)
, involvement of the gustatory pathway appears to be a likely alternative. We test this hypothesis by carrying out biochemical, physiological, and behavioral experiments in mice to determine whether 1) LCFAs specifically affected the free intracellular calcium levels ([Ca2+]i) in taste bud cells, a parameter known to be involved in the generation of gustatory signals in the afferent gustatory fibers (11)
, 2) gustatory nerves (i.e., chorda tympani and glossopharyngeal nerves) were involved in the spontaneous fat preference and cephalic phase of digestive secretions triggered by oral lipid stimulation, and 3) LCFA deposition on the tongue led to activation of the neurons of the nucleus of the solitary tract (NST), the first gustatory relay in the brain. We used pharmacological inhibition of CD36 and CD36-null mice to confirm the crucial role played by lingual CD36 in this gustatory cascade.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
) were individually housed in a controlled environment (constant temperature and humidity, darkness from 7:00 p.m. to 7:00 a.m.) and fed a standard laboratory chow ad libitum (A04; Scientific Animal Food & Engineering, Augy, France). The strategy used to invalidate the CD36 gene eliminated a portion of introns 2 and 3 and all of exon 3. No regulatory element has ever been identified in these intronic sequences (12)
. The nearest neighboring genes to CD36 on chromosome 5 are 
0.05 Mb away. Furthermore, no other gene or expressed sequence tag has been shown to overlap the CD36 gene. In a first set of experiments, we investigated the role of gustatory nerves both in LCFA preference and conditioned taste aversion, by bilaterally sectioning the chorda tympani (CT) and/or glossopharyngeal (GL) nerve in anesthetized (40 mg/kg i.p. pentobarbital sodium) mice. The GL nerves were sectioned after gentle dissection of the sternohyoid, omohyoid, and digastric muscles. The CT was transected under the pterygoid muscle before its junction with the lingual nerve. For controls (sham-operated, SO), nerves were visualized but were left undisturbed.
In a second set of experiments, we assessed the impact of bilateral section of the gustatory nerves (GLX or GLX+CTX) on the cephalic phase of digestive secretions triggered by oral administration of linoleic acid (Sigma-Aldrich, St. Louis, MO, USA). Ten days after surgery, mice were fasted overnight, anesthetized (40 mg/kg i.p. pentobarbital sodium), esophageal ligation (to prevent the ingestion of linoleic acid) and pancreato-biliary catheterization (to collect digestive juice) were carried out. As soon as the pancreato-biliary flux had become stable, linoleic acid was delivered directly on the tongue. Digestive juice was collected every 5 min for 30 min. The efficiency of pancreato-biliary derivation was checked by infusing 0.5 ml of 0.1 N HCl in saline into the duodenal lumen at the end of experiment to trigger the secretin-mediated stimulation of digestive secretions.
In a third set of experiments, the effect of oral lipid stimulation of the NST was explored by using the Fos method (13)
. Linoleic acid deposition (70 µl) was gently placed on the tongue of concious CD36+/+ or CD36–/– mice with a small paintbrush. Water alone, or 0.3% xanthan gum (Sigma-Aldrich), which mimics lipid texture, was placed on the tongue of control animals in the same manner. On the basis of the dynamics of Fos protein synthesis in the central nervous system (13)
, tissue samples were taken and fixed 90 min after oral stimulation. After an intraperitoneal injection of pentobarbital sodium (40 mg/kg), the thoracic cavity was opened and mice were perfused intracardially with 50 ml of ice-cold heparin in saline followed by 50 ml of ice-cold 4% paraformaldehyde in 0.1 M phosphate buffer pH 7.4. The entire brain was removed and fixed by incubation in 4% paraformaldehyde for 2 h. Samples were cryoprotected by incubation in 15% sucrose in 0.1 M phosphate buffer for 2 h, followed by overnight treatment with 30% sucrose in phosphate buffer. Brain samples were frozen in isopentane maintained at –40°C and were then stored at –80°C.
Immunomagnetic isolation of CD36-positive taste bud cells
Mouse circumvallate papillae were isolated as described previously (9)
. Taste bud cells were then dissociated by incubation for 20 min in RPMI 1640 medium supplemented with 2 mM EDTA, 1.2 mg/ml elastase, 0.6 mg/ml collagenase (type I), and 0.6 mg/ml trypsin inhibitor at 37°C. Taste bud cells were recovered by centrifugation [600 g, 10 min, at room temperature (RT)], and incubated for 2 h with a phycoerythrin-coupled anti-CD36 antibody (Santa Cruz Biotechnology, Santa Cruz, CA, USA). The cells were washed in PBS and centrifuged (600 g, 10 min, RT), then incubated with antiphycoerythrin IgG-coated microbeads. The cell suspension was loaded onto a Miltenyi wire-mesh separation column. Magnetically labeled cells (CD36-positive cells) were retained in the column by a magnetic field, whereas unlabeled cells (CD36-negative cells) passed through. Labeled cells were released and collected by abolishing the magnetic field.
CD36-positive cells were resuspended in a fresh RPMI medium supplemented with 10% fetal calf serum, 200 U/ml penicillin and 0.2 mg/ml streptomycin, seeded in a Biocoat Poly-D-lysine-coated dishes, and cultured for 24 h. They were then stained with trypan blue to assess their viability before [Ca2+]i determination.
[Ca2+]i determination
CD36-positive and CD36-negative taste bud cells were loaded by incubation with Fura-2/AM (1 µM) in loading buffer (110 mM NaCl, 5.4 mM KCl, 25 mM NaHCO3, 0.8 mM MgCl2, 0.4 mM KH2PO4, 20 mM sodium HEPES, 0.33 mM Na2HPO4, 1.2 mM CaCl2; pH 7.4) for 60 min at 37°C. Fluorescence intensity was determined with a spectrofluorometer at 340 nm, 380 nm (excitation filters), and 510 nm (emission filters). The fatty acids (20 µM) tested were added during recording. Sulfo-N-succinimidyl oleate ester [SSO, 400 µM in 0.5
DMSO (14)
], which inhibits binding to CD36, was added 20 min before the fatty acids. [Ca2+]i was calculated, as described previously (15)
.
Real-time RT-PCR
Total RNA was isolated from isolated taste bud cells on RNeasy minicolumns (Qiagen, Valencia, CA, USA) and treated with DNase using the RNase-free DNase Set (Qiagen). We generated first-strand cDNA from 0.2 µg total RNA by reverse transcription, according to the kit manufacturer’s instructions (Omniscript Reverse Transcription, Qiagen). CD36 mRNA levels in taste bud cells were determined by real-time RT-PCR (iCycler IQ, Bio-Rad, Hercules, CA, USA). We checked that the cell preparations tested were indeed of gustatory origin by also assaying the G-protein 
-gustducin, which is found in the taste bud cells. RNA levels were normalized with respect to 36B4 mRNA levels. We mixed the cDNA (corresponding to 10 ng total RNA), 300 nM of the forward and reverse primers and 100 nM of the probe with the Taqman PCR mastermix (qPCR MasterMix No Rox, Eurogentec, Herstal, Belgium). Primer/probe sets were designed by Eurogentec. The oligonucleotide sequences of the primers and probes were as follows (PubMed accession numbers are given in brackets): CD36 (NM 007643): forward 5'-GGCCAAGCTATTGCGACATG-3', probe 5'-CACAGACGCAGCCTCCTTTCCACCT-3', reverse 5'-CCGAACACAGCGTAGATAGAC-3'; -gustducin (XM 144196): forward 5'-ACACATTGCAGTCCATCCTAGC-3', probe 5'-TGAAGTTGTTCTTGGTCCTCTCGGCTCC-3', reverse 5'-ATCACCATCTTCTAGTGTATTTGCC-3'; 36B4 (NM 007475): forward 5'-GCCACCTGGAGAACAACCC-3', probe 5'-AGGTCCTCCTTGGTGAACACGAAGCC-3', reverse 5'-GCCAACAGCATATCCCGAATC-3'. Each real-time PCR cycle consisted of a denaturation at 95°C for 15 s followed by annealing at 60°C for 30 s. The comparative 
Ct method was used for quantification (16)
.
Behavioral experiments
Two-bottle preference test
The effects of selective bilateral section of gustatory nerves (CT and/or GL) on the spontaneous preference for linoleic acid-enriched solutions were investigated by means of the two-bottle preference test. Sham-operated animals were used as controls. Ten days after surgery, individually caged mice were allowed to choose between 2% purified linoleic acid emulsified in 0.3% xanthan gum in water or water with vehicle alone (0.3% xanthan gum) over a period of 48 h. The water intake of mice was restricted for 3 h before the experiment, and the position of bottles (on the right or the left) was changed daily to avoid the development of preference for the bottle on a particular side. Two weeks later, these mice were subjected to the same protocol for 0.5 h. The specificity of the LCFA effect and surgical efficiency were checked by carrying out additional short-term (0.5 h) and long-term (48 h) behavioral tests with a sweet or a bitter molecule (4% sucrose and 0.1 mM denatonium, respectively). Mice were allowed a resting period of 1 wk between two successive tests.
Conditioned taste aversion
To further investigate the involvement of gustatory nerves in LCFA preference, a bilateral transection of the GL and CT nerves was carried out in mice previously conditioned to avoid linoleic acid. Sham-operated animals were used as controls. Six days before conditioning, access to water was restricted (access to water supplemented with 0.3% xanthan gum from 10:00 to 12:00 a.m. for 3 days and then from 10:00 to 11:00 a.m. for the next 3 days), to train the animals to drink during a short period in the morning. We ensured that the animals did not suffer dehydratation by providing them with free access to drinking water from 3:00 to 3:30 p.m. On the day of conditioning, mice were supplied with 2% linoleic acid-enriched solution from 10:00 to 11:00 a.m., received an intraperitoneal injection (16.5 µl/g body wt) of 0.15 M LiCl (conditioned mice) or 0.15 M NaCl (controls). A two-bottle preference test (1 h) was carried out the next day with 2% linoleic acid-enriched solution vs. vehicle alone (0.3% xanthan gum in water) to assess the extent to which aversion conditioning was successful. Two mice with no clear aversion to linoleic acid were excluded. The conditioned aversion to linoleic acid was strengthened by subjecting the remaining 11 mice to an additive conditioning, as described previously (17
, 18)
. A second set of preference tests was performed to ensure that all mice displayed an aversion to linoleic acid. The specificity of aversion to linoleic acid was assessed using a two-bottle preference test in which 4% sucrose was added in one bottle. Ten days after bilateral gustatory nerves transection, short-term (1 h) two-bottle preference tests were performed in the presence of 2% linoleic acid followed by 4% sucrose. Consecutive tests were separated by a resting period of 1 week.
Immunocytochemistry
CD36-positive cells
CD36-positive cells selected from the mouse circumvallate papillae using the Miltenyi magnetic cell sorting system, as described above, were subjected to immunocytochemical staining. Cells were fixed in 95% ethanol and rehydrated in 0.1 M PBS (pH 7.4). They were blocked by incubation for 20 min at RT in 10% goat serum (Sigma-Aldrich) and 2% Triton X-100 in PBS and then were incubated overnight at 4°C with both a 1:100 dilution of anti-mouse CD36 antibody (19)
and a 1:50 dilution of polyclonal anti-mouse -gustducin antibody raised in rabbit (Santa Cruz Biotechnology). The slides were washed and incubated for 2 h at RT with Cy3-conjugated anti-mouse IgG (for CD36 detection) diluted 1:600 and with FITC-conjugated anti-rabbit IgG (for -gustducin detection) diluted 1:500. Cells were counterstained with Hoechst 33342 (0.1 mg/ml; Sigma-Aldrich). Staining specificity was assessed by carrying out the same procedure but omitting the primary antibodies. Cells were washed twice in PBS, a drop of mounting medium was added to the slides, and cells were then analyzed by confocal microscopy (Leica TCS SP2; Leica Microsystems, Wetzlar, Germany).
Denervation
The extent to which GL nerve transections were successful was determined post mortem by evaluating the number of taste buds in the circumvallate papillae by immunohistochemistry, using -gustducin as a specific marker of taste bud cells (20)
. Circumvallate papillae were embedded in OCT medium (Tissue-Tek; Oxford Instruments, Abingdon, Oxfordshire, UK) and snap-frozen in isopentane chilled with liquid nitrogen. Cryostat sections (14 µm) were allowed to dry in air for 2h at RT, fixed by incubation in 95% ethanol for 5 min and rehydrated by incubation in 0.1 M PBS (pH 7.4) for 10 min. Rehydrated sections were blocked by incubation in 1.5% goat serum and 0.2% Triton X-100 in PBS for 30 min at RT, and were then incubated overnight at 4°C with a 1:200 dilution of a commercial polyclonal anti-mouse -gustducin antibody raised in rabbit. Sections were washed and incubated for 2h at RT with labeled secondary antibodies (1:600 dilution of Alexa 488-conjugated anti-rabbit IgG; Santa Cruz Biotechnology). Staining specificity was assessed by carrying out the same procedure, but omitting the primary antibodies.
The extent to which bilateral CT sectioning had been successful was checked by counting fungiform papillae stained with methylene blue (21)
. The lingual epithelium was separated from the connective tissue by dissociation with elastase and dispase (2 mg/ml of each enzyme mixed in tyrode buffer (in mM): 140 NaCl, 5 KCl, 10 HEPES, 1 CaCl2, 10 glucose, 1 MgCl2, 10 sodium pyruvate; pH 7.4), stained with 0.5% methylene blue, and rinsed with distilled water. Taste pores appeared as dark blue dots on a background of lightly stained fungiform papillae.
Fos immunostaining
Transverse 40 µm sections corresponding to the NST were cut with a cryostat at –22°C and collected in 0.1 M phosphate buffer, pH 7.4. Alternate sections were mounted on Superfrost Plus glass slide (Milian SA, Geneva, Switzerland) for immunostaining. Sections were incubated in 2% bovine serum albumin and 0.5% Triton X-100 in 0.1 M PBS for 1 h; then they were incubated with the primary antiserum containing the polyclonal pAb5 antibody raised against Fos protein (Oncogene Science, Cambridge, MA, USA) diluted 1:10,000, for 24 h. Sections were washed three times in 0.1% skim milk in 0.01 M phosphate buffer for 10 min each and were then incubated with biotinylated goat anti-rabbit antiserum diluted 1:200 for 3 h, washed twice in 0.01 M PBS for 10 min each, incubated in a saline sodium bicarbonate solution (8.4 g/l NaHCO3, 9 g/l NaCl; pH 8.2) for 10 min and, finally, with avidin-FITC complex diluted 1:500 in saline bicarbonate solution for 30 min. Following a 10-min washout period in saline bicarbonate solution and two 10-min baths in 0.01 M PBS, sections were mounted in Vectashield (Vector Laboratories, Burlingame, CA, USA). Sections were examined under a widefield fluorescence microscope (Zeiss Axio Imager 1). Photomicrographs were analyzed with imaging software (Image J, National Institutes of Health, Bethesda, MD, USA). Fos-positive neurons were counted independently twice in blind conditions (genotype and stimulus conditions). The mean number of positive neurons by section was determined for each area of interest, as determined according to a stereotaxic atlas (22)
.
Statistics
The results are expressed as means ± SE. The significance of differences between groups was evaluated with SigmaStat (Systat Software, Erkrath, Germany). We checked that the data for each group were normally distributed and that variances were equal and then carried out ANOVA or two-tailed Student’s t tests. The test used in each experiment is indicated in the corresponding legend. Significant differences are labeled as follows: *P < 0.05; **P < 0.01; ***P < 0.001.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. This event is known to be mediated by changes in [Ca2+]i in taste receptor cells (11)
. We investigated whether fatty acids could induce this response, by isolating CD36-positive taste bud cells from mouse circumvallate papillae by affinity purification using magnetic beads and anti-CD36 antibody. We checked that this original method was both selective and efficient, by real-time RT-PCR and immunocytochemical analysis of CD36 expression in taste bud cells before and after immunomagnetic selection. Only 11% of the 1.3 x 106 cells (86% viability) isolated from 100 circumvallate papillae (one per mouse) were CD36 positive. CD36 mRNA levels were increased by a factor of seven following selection (Fig. 1
A). This change reflected the increase in the number of CD36-positive cells (Fig. 1B
). Consistent with our previous findings (9)
, CD36 appeared to be coexpressed with -gustducin, demonstrating that the cells selected on the basis of their CD36 expression were taste receptor cells (Fig. 1A, B
). As expected, no CD36 mRNA was detected in unselected (CD36-negative) cells (Fig. 1A
).
|
After 24 h of culture, CD36-positive taste bud cells were used to investigate the possible induction by fatty acids of changes in [Ca2+]i, using a fluorescent probe. CD36-negative taste bud cells were used as controls. Consistent with the binding affinity of CD36 for LCFAs (10)
, saturated and unsaturated LCFAs triggered a rapid and robust increase in the [Ca2+]i in CD36-positive cells (palmitic acid or PA > linoleic acid or LA > docosahexaenoic acid or DHA), whereas the medium-chain fatty acid, capric acid, induced only a small change in [Ca2+]i levels (Fig. 2
A). By contrast, LCFA-induced calcium responses remained low in CD36-negative cells (Fig. 2A
). The origin of differences in [Ca2+]i responsiveness found between PA/LA and DHA remains to be determined.
|
In CD36-positive cells, the addition of a specific CD36 binding inhibitor, sulfo-N-succinimidyl oleic acid ester (SSO) (14)
, to the culture medium fully abolished the linoleic acid-mediated rise in [Ca2+]i (Fig. 2B
). All together, these data provide the first demonstration that LCFAs trigger an increase in [Ca2+]i levels in the taste bud cells, and this event is CD36-dependent.
The gustatory nerves convey the fatty acid signal
Taste receptor cells from fungiform papillae and, to a lesser extent, from the anterior part of the foliate papillae establish synaptic contacts with the chorda tympani (CT) nerve, whereas the posterior parts of the foliate and circumvallate papillae are innervated by the glossopharyngeal (GL) nerve. To determine whether the gustatory nerves were involved in the LCFA-mediated fat preference and early phase of digestive secretion, the impact of a bilateral transection of the CT and/or GL nerves was explored in wild-type mice. Sectioning of the gustatory nerves is known to have a profound effect on the structure and number of gustatory papillae (23
, 24)
. We evaluated post mortem the efficacy of GL nerve transections (GLX) by determining the number of taste buds in circumvallate papillae by immunohistochemical detection of -gustducin, whereas the effect of CT transection (CTX) on the number of fungiform papillae was determined by methylene blue staining (21)
. According to Tadeka et al. (25)
, only "ghost" buds were found 10 days after bilateral GLX (Fig. 3
A). Consistent with the innervation pattern of the lingual papillae, no additive effect was found in double-denervated mice (GLX+CTX). CTX resulted in a 36% decrease in the number of fungiforms as compared to sham-operated controls, which is in good agreement with the 38% decrease reported by Guargliardo and Hill (26)
(Fig. 3A
). In the mouse, the presence of stained fungiforms 10 days following CTX is consistent with a complete denervation. Indeed, fungiform papillae degenerate incompletely following CTX, unlike circumvallate taste buds (26)
. Moreover, methylene blue method is known to have a small false-positive rate (21)
. For each set of experiments, mice that did not display a significant decrease in the number of vallate taste bud (GLX) and fungiforms papillae (CTX), as compared to SO controls were excluded from the analysis.
|
As expected, a loss of preference for sweet tastes and aversion of bitter tastes was found in denervated mice (Fig. 3B
). In short-term (0.5 h) tests, preference for the lipid-enriched solution, which was decreased by sectioning of the GL nerves, was abolished in double-denervated mice (Fig. 3B
). Bilateral transections of the GL were also associated with a significant decrease in the flux of pancreatobiliary juice, this decrease being especially large in double-denervated mice (Fig. 3C
). These data provide the first demonstration that fat preference and the lipid-mediated increase in digestive secretions require intact GL and CT nerves in the mouse. GL afferent fibers may play a dominant role in transduction of the fatty acid signal. Indeed, GL injury fully abolished preference for the linoleic acid-enriched solution in long-term (48 h) two-bottle preference tests (Fig. 3B
). CD36 expression, which is particularly strong in posterior tongue papillae (9)
, may account for this result.
Conditioned taste aversion tests have shown that rats were able to detect and avoid very small quantities of oleic acid or linoleic acid (<100 µM), providing new evidence to suggest that LCFAs are the chemical components responsible for the orosensory perception of dietary fat (27)
. We investigated the involvement of gustatory nerves in the orosensory perception of LCFAs, by carrying out a bilateral transection of GL and CT nerves in mice conditioned to avoid linoleic acid. Sham-operated animals were used as controls. Like rats (27)
, mice can be conditioned to avoid a solution containing an LCFA (Fig. 4
). This aversive behavior is specific since the sweet preference was not affected in conditioned mice. Sectioning of the gustatory nerves fully suppressed the avoidance of linoleic acid in conditioned mice (Fig. 4)
. These data provide additional evidence that LCFAs are perceived as sapid molecules and are consistent with an involvement of gustatory nerves in the orosensory perception of fatty acids in the mouse.
|
The neuronal activation of NST triggered by an oral stimulation with linoleic acid is CD36-dependent
The NST in the brain stem is the first synaptic relay between the orosensory perception of tastants and the central nervous system. Immunohistochemical detection of Fos, the protein product of the immediate early gene c-fos, has been successfully used to identify populations of neurons activated by various tastants in the NST (28)
. Topographic analyses of the NST of this type has revealed that the rostral-gustatory and caudal-gustatory areas of the solitary tract (rgNST, cgNST) corresponded roughly to first-order gustatory subnuclei receiving synaptic inputs from the CT and GL nerves, respectively. The distal part of the NST, innervated principally by visceral afferences of the vagus nerve, may be considered as a nongustatory area (ngNST, (29)
) (Fig. 5
A). We studied Fos-immunoreactivity to explore whether the deposition of linoleic acid on the tongue activated the gustatory nuclei of the NST. Water or a xanthan gum solution to mimic the texture of fat was applied to the tongue of control mice. Linoleic acid elicited a robust Fos-like immunostaining in the gustatory nuclei of the NST (Fig. 5B, C
). This neuronal activation was not due to intraoral somatosensory effects (i.e., texture and/or mechanical stimuli) because it was very weak in mice treated with xanthan gum solution or water alone. As previously reported for sweet and bitter tastes (30)
, weak neuronal activation was also observed in the visceral NST. A combination of gustatory and visceral activations by LCFAs might account for this result. Indeed, the activation of NST neurons by linoleic acid may be due to the stimulation of taste buds in the upper esophagus and pharynx, which is known to be under the control of vagus nerve and/or visceral signals mediated by lipid-dependent reflexes. To examine whether linoleic acid-mediated activation of gustatory NST required the presence of lingual CD36, a second set of experiments has been performed using CD36-null mice. The deposition of linoleic acid on the tongue led to a significant increase in Fos-like immunoreactivity in both the rostral and caudal gustatory NST in wild-type animals (Fig. 6
). This event was clearly mediated by CD36, as it was not observed in CD36-null mice.
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Saturated and unsaturated LCFAs selectively trigger a rapid and huge increase in [Ca2+]i in CD36-positive taste bud cells isolated from circumvallate papillae. This change is known to induce the release of neurotransmitters, which then activate fibers in gustatory nerves, resulting in taste signals (11)
. This cellular event appears to be CD36-dependent, as the use of SSO to inhibit the binding of LCFAs to CD36 totally abolished the LCFA-induced calcium response. The small change in [Ca2+]i found after the addition of capric acid to the culture medium is consistent with this assumption. Indeed, CD36 is known to display preferential binding to LCFAs (10)
. The signaling cascade triggered by binding of LCFAs to CD36 remains unknown in taste bud cells. -gustducin is systematically found in CD36-positive taste bud cells, but this G-protein is unlikely to be involved, because -gustducin-null mice and wild-type mice exhibit similar lipid preference (31)
. In contrast, protein tyrosine kinases (PTKs) from the Src kinase family may be involved in this regulation. Indeed, the PTK-like Fyn and Lyn proteins are known to be implicated in [Ca2+]i control within cells (32)
. These PTKs have also been shown to interact with the cytosolic C-terminal tail of CD36 (33)
.
Selective sectioning of the CT and/or GL nerves greatly decreased both the spontaneous preference for LCFAs and the LCFA-induced cephalic phase of digestion. Lipid signals therefore appear to be transmitted by a nerve route known to be involved in the transfer of gustatory information to the brain. GL is a mixed nerve that also carries some somesthetic information (34)
. It was therefore possible that texture, rather than taste, was involved. We therefore sectioned the GL and CT bilaterally and investigated the effect of this intervention on LCFA detection in mice previously conditioned to avoid linoleic acid. As reported in rats (35)
, we found that mice could easily be conditioned to detect linoleic acid specifically and to avoid this compound. This avoidance behavior was totally abolished by sectioning of the gustatory nerves, preventing them from distinguishing between linoleic acid and the control solution containing xanthan gum used as a texturing agent. These data suggest that the textural cues carried by the GL nerve do not play a major role in the orosensory perception of LCFAs. Moreover, in contrast to what has been reported for fat (35)
, reports of textural perception of LCFAs by the trigeminal nerve seems to be also anecdotal, as the sectioning of the gustatory nerves was sufficient to abolish the preference for LCFAs.
The deposition of linoleic acid on the tongue also triggered the activation in the brain stem of NST areas known to receive CT and GL afferent fibers. This activation appeared to be CD36-dependent, as it was not observed in CD36-null mice subjected to oral stimulation with linoleic acid. The known involvement of the lateral hypothalamus and nucleus accumbens in food intake and reward, respectively, and in the reception of synaptic inputs from the NST (34)
may account for the spontaneous preference for LCFA-enriched food observed in mice. The digestive projections of the NST (34)
may also contribute to a lipid-mediated reflex, controlling pancreato-biliary secretions directly and/or indirectly through the production of intestinal hormones.
Collectively, the data reported herein indicate that CD36 acts as a lipid sensor on the tongue and show that the gustatory pathway is involved in spontaneous fat preference and the induction by LCFAs of the cephalic phase of digestion. They provide additional support for recent studies demonstrating a role of gustation in dietary fat perception. Given that lipids are widespread in foods, these results suggest that "fatty" may be a sixth taste modality in the mouse.
It remains unclear whether oral detection of LCFAs is a common trait in mammals. Rats and mice are similar in several respects: both display a spontaneous preference for lipids (1)
, which is altered by the gustatory nerve transection (35)
, both express CD36 in circumvallate papillae (36)
, and both respond to oral lipid stimulation by a rapid rise in digestive secretion (9)
. This latter effect seems to be CD36 dependent, as it is not affected by the direct application of LCFAs onto a CD36-negative sensory mucosa, such as the soft palate (9)
.
Clinical data have also provided some evidence of a taste component of fat perception in humans. Subjects submitted to a sham feeding with full-fat food displayed a higher postprandial triglyceridemia than subjects sham-fed a fat-free version (37)
. This metabolic change could not be accounted for by textural and olfactory cues (38)
, and it was therefore suggested that an orosensory perception system devoted to lipids might be involved. Consistent with this assumption, healthy subjects seem to be able to detect low quantities of saturated and unsaturated LCFAs in a specific manner (39)
. However, the mechanism underlying an oral fat perception in humans remains unclear. To date, it is not known whether the CD36 gene is expressed in human taste bud cells. Moreover, the presence of an efficient lingual lipase, which is known to play a significant role in lipid preference in rats and mice (8)
, remains far from proven in humans (40
, 41)
. Nevertheless, the small amounts of free fatty acids found in foods might be sufficient to trigger a fat stimulus if there is a lipid sensor displaying a high affinity for LCFAs, such as CD36, in human gustatory papillae (39)
.
Lipids play a number of roles in nutrition. They are the nutrients with the highest energy density and provide essential fatty acids and fat-soluble vitamins. The existence of an orosensory system devoted to lipid perception, by enabling the animal to select and use fatty foods efficiently within the body, might constitute a physiological advantage when food is scarce. Conversely, during periods of plethora, this system might contribute to the risk of obesity. Indeed, mice chronically overeat corn oil and become obese when this lipid source is provided as an optional supplement to standard laboratory chow (42)
. Nevertheless, the effect of lingual CD36 on this feeding behavior remains to be documented. Determining the role of LCFAs in the gustatory system is thus an important challenge, which may give rise to new nutritional and/or pharmacological approaches for decreasing the risk of obesity in the future.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Received for publication June 5, 2007. Accepted for publication November 29, 2007.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  R. D. Mattes Oral Thresholds and Suprathreshold Intensity Ratings for Free Fatty Acids on 3 Tongue Sites in Humans: Implications for Transduction Mechanisms Chem Senses, June 1, 2009; 34(5): 415 - 423. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. L. Silverstein and M. Febbraio CD36, a Scavenger Receptor Involved in Immunity, Metabolism, Angiogenesis, and Behavior Sci. Signal., May 26, 2009; 2(72): re3 - re3. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. D. Mattes Brief oral stimulation, but especially oral fat exposure, elevates serum triglycerides in humans Am J Physiol Gastrointest Liver Physiol, February 1, 2009; 296(2): G365 - G371. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. M. Stratford, K. S. Curtis, and R. J. Contreras Linoleic acid increases chorda tympani nerve responses to and behavioral preferences for monosodium glutamate by male and female rats Am J Physiol Regulatory Integrative Comp Physiol, September 1, 2008; 295(3): R764 - R772. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
