Targeting Fatty Acid-Activated Pathways in the Somatosensory System

Understanding the mechanisms that contribute to the sensory recognition ofdietary fat has become increasingly important given the epidemic of obesity which ispartially driven by high dietary fat intake. While the textural properties of dietary fat (e.g.triglycerides) have been well documented to occur via the activation of trigeminalganglionic neurons (TGNs) in the somatosensory system, the ability of fatty acids (FAs)to activate TGNs remained unexplored. To investigate the ability of FAs to activate thesomatosensory system, I have used fura-2 based calcium imaging and patch clamprecording as the functional readout to explore the ability of FAs to activate TGNs fromadult Sprague-Dawley rats. FAs (1-100 μM) elicit robust changes in intracellular calciumconcentration([Ca]i) in approximately two-thirds of TGNs in a concentration-dependent fashion. In general, responses to monounsaturated or saturated FAs occurredin neurons that also responded to polyunsaturated FAs. Moreover, cells exhibit FA-responses that are less dependent of extracellular sodium(Na) but are either dependentor independent of extracellular calcium(Ca) possibly indicating multiple functional celltypes. Store depletion by thapsigargin significantly reduces but does not abolish the FA-induced Ca responses. I also measured FA-induced membrane depolarization in TGNsby patch clamp recording. Linoleic acid (LA) elicits membrane depolarization in TGNswith a time course similar to that seen for the rise in[Ca]i. Inhibition of G protein 34activation with GDP-β-S and inhibition of phospholipase C activity with U73122 blocked77% and 58% of LA-induced depolarization of TGNs, respectively, which suggests theinvolvement of G protein/PLC pathways and is consistent with the activation of upstreamFAs sensitive G protein-coupled receptor (GPCRs) in the FAs signaling pathway inTGNs. IntroductionOne role of the peripheral somatosensory system in the oral cavity is to enable therecognition of the temperature, pungency and the textural properties of foods. Cellularand molecular approaches have yielded information regarding the molecularunderpinning of these processes. Several membrane ion channels have been discovered asthe receptors of capsaicin (from hot chili) (1), menthol (2, 3) and pungent agents (4, 5)such as mustard oil and garlic. These unconventional sensations, along with the moreconventional ones like taste and olfaction, contribute to the ability of enjoying thepalatability of foods while rejecting those compounds that are potentially dangerous.The ability to sense fat either in terms of its texture or taste appears to becorrelated with dietary fat preference and, ultimately, obesity. Human studies suggest thatindividuals with a high body mass index (BMI) have lower sensitivity to oral FAscompared to those with lower BMIs (6). Animal behavioral studies indicate a clearpreference to FAs compare to mineral oil (textural) control (7), consistent with the notionthat FAs have a taste quality. The conditioned taste aversion test showed strong aversionsto LA at the concentration of 100 μM (8). The cellular/molecular mechanisms underlyingthe taste of FAs remain to be elucidated. The textual cues associated with dietary fat (e.g. 35triglycerides) are generally accepted, though the concentrations that produce thissomatosensory experience have not been well established empirically. Recent reportssuggest that the concentrations of free fatty acids (FFAs) in edible oil generally rangebetween 0.01%-0.1% (9). Lingual lipase activity, on the other hand, could generateadditional 0-12 μmol fatty acids-/min per litter (6), which helps increase the FFAs to adetectable level in the oral cavity. Therefore, based upon the concentrations found in fat-containing foods and/or generated by the activity of lingual lipases, FAs have thepotential of being effective sensory stimuli in the oral cavity. Indeed, research hadindicated these concentrations of FAs were able to activate taste receptor cells (TRCs) byinhibiting delayed rectifying potassium channels (DRKs) (10). However, whether FAsare capable of activating the somatosensory system directly like capsaicin and menthol isunknown.The peripheral nervous system (PNS) within oral cavity contains three cranialnerves. Cranial nerves VII and IX (referred to as the facial and glossopharyngeal nerves,respectively) are responsible for the taste perception, while cranial nerve V (referred to astrigeminal nerve) is the one that transmits somatosensory information. Generallyspeaking, noxious stimuli are detected by nociceptors that have small to mediumdiameter perikarya that further correspond to unmyelinated C and lightly myelinated Aδnerve fibers. Innocuous stimuli, such as light touch, are detected by large diameterneurons that correspond to the more heavily myelinated Aα or Aβ fibers (referred to asmechanoreceptors). Based on molecular markers and pharmacological characteristics,there are about 50% of sensory neurons that respond to capsaicin (transient receptor 36potential vanilloid 1(TRPV1) positive) (4), 50% of capsaicin sensitive cells respond tomustard oil (mainly allyl isothiocyanate) (TRPV1 and transient receptor potential ankyrin1 (TRPA1) positive), the other 50% respond to sanshool (2P domain and backgroundpotassium channels (KCNK) positive), which by itself activates 52% of sensory neurons(5). Menthol activates a separate subpopulation (~20%) that does not respond tocapsaicin or mustard oil (transient receptor potential melastatin 8 (TRPM8) positive only)(5). Most of these neurons are small diameter myelinated neurons that are not sensitive tostretch, except for some TRPV1 and sanshool sensitive neurons that are stretch sensitivealso, and fall into the class of non-peptidergic C-fiber nociceptors (11).The primary targets for studies on somatosensation occur in sensory neurons fromthe dorsal root ganglia and trigeminal ganglia. The trigeminal nerve is a mixed nervecontaining both afferent and efferent nerve fibers with broad innervation patterns. TGNsare considered polymodal neurons that consist of a heterogeneous population of cells.How to identify and functionally separate these neurons has been a major confound in theresearch of the physiology of sensory neurons. The most widely used approach is todivide sensory neurons by their sizes, i.e. small and medium neurons (<26 μm diameter)are classified as nociceptors, while larger neurons (>28 μm diameter) are believed torepresent the population of mechanoreceptors (11, 12).Although progress has been made in the cellular and molecular basis for thetransduction of spicy and pungent compounds in somatosensory system, the ability ofTGNs to respond to the components contained in fat and the mechanisms underlying theability of FAs to activate sensory neurons remains elusive. Here, I take the multiple 37approaches to elucidate the molecular and cellular basis of the FA-induced signalingpathway in TGNs. I have shown that FAs activate a subset of TGNs by increasing the[Ca]i and depolarizing the membrane. Moreover, I found that FAs caused the release ofintracellular Ca store in this subset of neurons, a process that contributes to thedevelopment of the FA-induced receptor potential. Preliminary data implicated theprimary receptor for FAs appears to be one or more of the recently identified FA-activated GPCRs. Several FA-sensitive GPCRs were found expressed in the trigeminalganglia, and I provided the first evidences that the G protein/phospholipase C (PLC)pathway is involved in FAs signaling. Thus, I conclude that FAs are effective stimuli forsomatosensory neurons, which contribute to our perception of dietary fat. Materials and MethodsAcute isolation of trigeminal ganglionic neurons. All surgical and experimentalprocedures were reviewed and approved by the Institutional Animal Care and UseCommittee, Utah State University. TGNs were isolated according to methods modifiedfrom Simon and colleagues (13, 14). Four-to 8-week-old male Sprague-Dawley rats weredeeply anaesthetized by 50 mg/kg i.p. injection of sodium pentobarbital. The pairedtrigeminal ganglia were then harvested and diced into small pieces into Hank’s balancedsalt solution (HBSS), followed by a 45-minute incubation with 2.5 mg/ml collagenase(Type XI-S) at 37°C in HBSS also and a 10-minute incubation at 37°C in 100 μg/mlDNase I in culture medium (DMEM/F-12 1:1 medium supplemented with 10% FBS).The neurons were centrifuged and resuspended in culture medium by gentle triturationwith a flame-polished Pasteur pipette. Subsequently, neurons were plated onto poly-D38lysine (0.1 mg/ml)-coated glass coverslips (12-15 mm in diameter) and cultured at 37°Cin a water-saturated atmosphere with 5% CO2. For electrophysiology, neurons with somadiameters ranging between 18 to 25 μm were used within 8 hours after plating oncoverslips. For calcium imaging, neurons were cultured 36-48 hours in culture mediumbefore use. These cell based assays were handled in the room temperature unless stated.Calcium imaging. TGNs were loaded with fura-2/AM (5 μM; Molecular Probes,Eugene, OR) for 1 hour in Tyrode’s saline solution with 10% pluronic acid at 37°C in thedark. Neurons were then rinsed and placed in culture medium for 30 minutes to allow de-esterification of the acetoxymethyl ester group from fura-2. The coverslips were mountedinto an imaging chamber (RC-25F and RC-26Z, Warner Instruments, Hamden, CT),placed on an inverted microscope (Nikon, Eclipse TS100, Japan) and perfusedcontinuously with Tyrode’s solution. Neurons were illuminated with a 100-watt xenonlamp and excitation wavelengths (340/380 nm) were delivered by a monochromator(Bentham FSM150, Intracellular Imaging Inc., Cincinnati, OH) at a rate of 20 ratios perminute. Fluorescence was measured by a CCD camera (pixelFly, Cooke, MI) coupled toa microscope and controlled by imaging software (IncytIm2, Intracellular Imaging).The ratio of fluorescence (340/380 nm) was directly converted to Ca concentrationsusing a standard curve generated for the imaging system using a fura-2 calcium imagingcalibration kit (Molecular Probes, Eugene, OR). FAs were applied extracellularly with abath perfusion system at a flow rate of 4 ml/min permitting complete exchange of theextracellular solution in less than 20 seconds. 39Electrophysiological recording and analysis. Whole-cell patch clamp recordingwas used to measure membrane potential (current clamp mode) in TGNs with anAxopatch-200B amplifier (Molecular Devices, Sunnyvale, CA). Patch pipettes werefabricated from borosilicate glass on a Flaming-Brown micropipette puller (model P-97;Sutter Instrument, Novato, CA) and subsequently fire-polished on a microforge (modelMF-9, Narishige, Japan) to a resistance of 2-8 MΩ. Commands were delivered and datawere recorded using pCLAMP software (v. 10, Molecular Devices, Sunnyvale, CA)interfaced to an Axopatch-200B amplifier with a DigiData 1322A analog-to-digital board.Data were filtered online at 1 kHz and sampled at 2-4 kHz. The membrane potentials (VM)of TGNs were recorded continuously before, during and after bath application of LAusing the current clamp mode of the amplifier while holding the cells at its zero currentlevel (at rest).Solutions. Standard extracellular saline solution (Tyrode’s) contained (in mM):140 NaCl, 5 KCl, 1 CaCl2, 1 MgCl2, 10 HEPES, 10 glucose, and 10 Na pyruvate; pH7.40 adjusted with NaOH; 310 mOsm. Sodium-free saline (Na free Tyrode’s) contained(in mM): 140 N-methyl-D-glucamine (NMDG), 5 KCl, 1 CaCl2, 1 MgCl2, 10 HEPES, 10glucose; pH 7.40 adjusted with HCl; and 310 mOsm (adjusted with NMDG). Calciumfree saline (Ca free Tyrode’s) contained (in mM): 140 NaCl, 5 KCl, 1 EGTA, 1 MgCl2,10 HEPES, 10 glucose, and 10 Na pyruvate; pH 7.40 adjusted with NaOH; 310 mOsm.The potassium-based intracellular solution was used for measurement of membranepotential contained (in mM): 140 K gluconate, 1 CaCl2, 2 MgCl2, 10 HEPES, 11 EGTA,3 ATP, and 0.5 GTP; pH 7.2 adjusted with KOH; 310 mOsM. Thapsigargin (the SERCA 40inhibitor on endoplasmic reticulum), U-73122 (an inhibitor of PLC) and U-73343 (theinactive analog of U-73122) were purchased from Sigma (St. Louis, MO). GDP-β-S, aninhibitor of G protein activation was obtained from EMD Biosciences (La Jolla, CA).The linoleic acid, oleic acid (OA) and docosahexaenoic acid (DHA) stocks were made in100% ethanol and stored under nitrogen to prevent oxidation. The myristic acid (MA)stock was made in distilled water. The caproic acid was dissolved in Tyrode’s directly.The working solutions were made from stock solutions immediately prior to use.Quantitative real time-PCR (qRT-PCR). For RNA Isolation, the pairedtrigeminal ganglia were either stored at 20°C in RNAlater (Ambion, Austin, TX) orplaced in TRI Reagent (MRC, Cincinnati, OH) for immediate extraction of RNA.Extraction was done according to manufacturer’s protocol. For PCR, first strand cDNAwas synthesized using the iScript cDNA synthesis kit from Bio-Rad (Hercules, CA). Toget a final volume of 20 μl, 50 ng trigeminal RNA was used for the reaction. DNAcontamination was evaluated by setting up a reaction where the reverse transcriptase wasomitted. After first-strand synthesis, 2 μl cDNA was added to a PCR reaction mix. TheHotMaster Taq DNA polymerase kit (5 Prime, Gaithesburg, MD) was used for the PCRreaction (final concentration: 1X reaction buffer, 5.0 mMMg, 200 μM dNTPs, 300-900nM forward and reverse primers, and 1.25 U/μl HotMaster Taq). Primer sequences forGPR40, 43, 120 and CD36 are shown in Table 2.1. For GPR84 detection, a commerciallyavailable TaqMan® Gene Expression Assay (ABI, Foster City CA) was used. 41Statistical analysis. The significant effects of all the treatments were determinedby a two-tailed Student’s t-test (α=0.05) compared with their controls as described in thetext. Data are presented as mean ± SEM, unless otherwise indicated. ResultsLA depolarizes and elicits a rise in[Ca]i in TGNs. In the present study I haveused cell-based assays (ratiometric calcium imaging and patch clamp recording) toexplore the ability of FAs to activate a signaling cascade in isolated rat TGNs. As the firstapproach, I used the prototypical FA stimulus, LA (1-100 μM), to elicit the[Ca]ichanges in isolated TGNs. As shown in Fig. 2.1A, 30 μM LA elicited a robust andreversible increase in[Ca]i in a subset of rat TGNs. The inset in Fig. 2.1 shows theconcentration-response function for the ability of LA to elicit significant[Ca]i risewhere the EC50 for the response is approximately 17.5 μM. In subsequent experimentsexploring the detailed mechanism of this FA-induced signaling cascade, I have focusedon using 30 μM LA, a concentration that produced a significant, but not maximal,response in TGNs. On average, approximately 65% of TGNs (n = 332) responded to LAwith an average increase in[Ca]i of 272.8 ± 9.4 nM. Neurons that showed less than a100 nM change in[Ca]i were counted as non-responders and cells that did not return tonear baseline Ca levels after LA application were excluded from subsequent analyses.To determine if LA was capable of depolarizing rat TGNs, patch clamp recordingwas performed in the whole cell current clamp configuration. I tested 9 neurons for LA-induced depolarization, which have an average resting (zero current) potential of -62.7 ±3.3 mV. LA elicited a large depolarizing response averaging 61.5 ± 6.4 mV in 7 neurons. 422 out of 9 neurons did not respond to the stimulus. Inherent in these LA-inducedresponses were delays of about 2 min for both the imaging and electrophysiologicalassays (Fig. 2.1A and Fig. 2.1B, respectively), though these two parameters were notsimultaneously measured in individual neurons in the present study. This significantdelay in time could not be solely attributed to perfusion delays from the bath applicationsystem because the delay was not seen with activation of TGNs by other pungent stimuli(data not shown). Moreover, of the FAs examined (see Table 2.1), anotherpolyunsaturated FAs DHA showed a similar consistent delay as LA, while themonounsaturated and saturated FAs had a more variable response time, but once initiated,generated a much more rapid[Ca]i rise.Other classes of FAs activate TGNs. To determine the specificity of FAresponses in rat TGNs, I performed a series of experiments using several FAs varying incarbon length, position and number of double bonds (Table 2.2). Fig. 2.2 A-D showsrepresentative Ca responses of various bath applied FAs (30 μM). One of the diet-richmonounsaturated FAs, oleic acid (OA, C18:1) elicited rapid, reversible rise in[Ca]iwhich often showed biphasic characteristics (n=19/83). Another medium chain saturatedFA, myristic acid (MA, C14:0), showed a response type similar to OA (n=39/151). Itconsistently caused a rapid and sometimes biphasic rise in[Ca]i in a subset of TGNs aswell. DHA (C22:6), another polyunsaturated omega-3 FA, elicited the[Ca]i changes ina similar manner with LA. However, caproic acid (C6:0), a short chain saturated FA,failed to elicit a rise in[Ca]i in TGNs (n=72). Additional differences between MA and 43OA responses and those generated by LA (Fig. 2.1A) and DHA were that responses tothe former FAs adapted, while those to LA and DHA did not.To determine the specificity of the FAs responsiveness on individual TGNs, Iapplied several different classes of FAs (LA, OA and MA) to the same individual cells.As stated above, about 65% TGNs responded to LA, which comprised the biggestresponding population to FAs. OA and MA could activate TGNs separately in somecases, however, most neurons responded to OA and MA were also LA-responsive.Therefore, I counted the number of neurons that were LA and OA and/or MA sensitive.As shown in Fig. 2.2E, 23% (n=83) of the LA-sensitive neurons were also OA-sensitive,while 26% (n=151) of LA-sensitive neurons were MA-sensitive. Moreover, a smallproportion of neurons (4%; n=58) responded to all three FAs. LA-induced[Ca]i increases depend to varying degrees on extracellular Cabut not on extracellularNa. Since LA reliably elicited[Ca]i changes and membranedepolarization in rat TGNs, I next conducted a series of imaging experiments todetermine the dependence of the LA-induced[Ca]i increases on extracellular Na andCa and the relative contribution of intracellular Ca stores to this process. Todetermine if extracellular Na influx was contributing to[Ca]i increase, I conductedcalcium imaging experiments using a saline solution whereby I substituted for Na withN-methyl-D-glucamine (NMDG). Half of the experiments were conducted usingNa-freesolution first followed by a regular Tyrode’s solution, while the rest were conducted inthe reverse order. As shown in Fig. 2.3A and 2.3B, Na substitution did cause a small butsignificant decrease in the magnitude of the LA-induced rise in[Ca]i. There were no 44obvious order effects in these experiments. Further, to avoid the response amplitudedifferences caused by cell to cell variation, I normalized each cell’s LA-induced[Ca]iincrease in Na free solution with the LA-induced[Ca]i increase in regular Tyrode’ssolution. The resulting percentages were used as the functional output to further compareacross groups. Fig. 2.3C showed the comparison of the normalized percentages betweenregular Tyrode’s (100%) and Na free Tyrode’s (92.0 ± 2.9 %; n=138; P=0.006, pairedStudent’s t-test). The deduction by Na substitution was only 8%, which indicated that itis probably not physiologically relevant.In contrast to extracellular Na substitution, removal of extracellular Ca lead tosignificant reduction in LA-induced[Ca]i increase in over 70% of neurons examined(Fig. 2.3D, 2.3E), which indicated a role for extracellular Ca entry in these responses.However, the dependence of the LA-induced[Ca]i rise upon extracellular Ca washighly variable among cells (Fig. 2.3D-F). Using the same normalization method, thepercentages of LA-induced[Ca]i increase in Ca free Tyrode’s over the LA-induced[Ca]i increase in regular Tyrode’s were used as the functional output. According tothese normalized percentages, I divided TGNs into two classes—the ones that have thenormalized percentages less than 50% (mean percentage 24.1 ± 1.6%; n=114), and theones that have the normalized percentages greater than 50% (mean percentage 72.9 ±2.5%; n=48). These data revealed that about 70% (n=162) of the cells tested in thismanner showed greater than a 50% reduction in the magnitude of the LA-induced[Ca]iincrease upon removal of extracellular Ca. Nevertheless, although performed in the 45same condition, there were 30% (n=162) of neurons that showed little dependence ofextracellularCa. Thapsigargin treatment blocked LA-induced[Ca]i increases in a subset ofTGNs. Data shown in Fig. 2.3 indicated extracellular Ca played important role in LA-induced TGNs responses, but it also became apparent that Ca release from intracellularstores was contributing to the overall changes in[Ca]i during LA application in many ofthe neurons examined. To determine the role of intracellular Ca release, TGNs werepretreated for up to 1.5 hour with thapsigargin (1 μM), a non-competitive SERCA(sarco/endoplasmic reticulum Ca ATPase) inhibitor, which leads to the depletion ofintracellular Ca stores. Most TGNs treated with thapsigargin showed an approximately2.5-3 fold increase in basal/resting Ca levels, which might indicate the activation ofstore operated Ca entry (SOCE) during store depletion (15-17). Application of LAelicited a rapid decrease in basal Ca concentration because of its ability to inhibit SOCE(16) followed by its characteristic slow increase due to LA stimulation (i.e. Fig. 2.1, 2.3).After the depletion of Ca stores, TGNs were still capable of responding to LA, thoughthere appeared to be two classes of neurons again—more than 68% of neurons (n=119)contained predominantly thapsigargin-sensitive stores (TG+), characterized by acomplete inhibition of LA-induced responses inCa-free extracellular solution (Fig.2.4B and 2.4C) (from 193.8 ± 14.6 nM to 4.7 ± 1.1 nM; n=81), while the remainder ofcells (32%; n=119) responded robustly to LA and these responses were not significantlyinhibited by extracellular Ca removal (Fig. 2.4A and 2.4C) and were broadly classifiedas thapsigargin-insensitive (TG-) (from 350.3 ± 28.1 nM to 305.4 ± 30.0 nM; n=38). 46Interestingly, the magnitude of LA-induced[Ca]i increase in TGneurons (350.3 ± 28.1nM) was significantly greater than TG+ neurons (193.8 ± 14.6 nM ) (Fig. 2.4C, P<0.001),which might be another hint of the differences of thapsigargin sensitivity between the twoclasses of TGNs. Activation of G protein and phospholipase C pathways in LA-induceddepolarization. GPR120 is a long chain polyunsaturated FAs receptor, which has beenshown to play an instrumental role in the FA-induced[Ca]i rise in enteroendocrine cellline STC-1 cells (18-20). GPR40 is another long chain polyunsaturated FAs receptor thatproved to play a role in FA-induced insulin secretion in pancreatic β cells (20, 21). In thisdissertation research, I checked this group of FA-sensitive GPCRs in trigeminal gangliawith quantitative real time PCR (qRT-PCR). As shown in Fig. 2.5F, these GPCRs wereall expressed in trigeminal ganglia, although the expression level differed dramatically. Inorder to confirm the role of G protein in the signaling pathway, I examined the effect ofblocking G protein activation with the reversible blocker GDP-β-S (1 mM) on LA-induced depolarization. In current clamp recording, GDP-β-S was included in theintracellular solution and membrane potential was recorded 2-3 minutes after achievingthe whole cell configuration. LA-induced depolarization was significantly reduced inpresence of GDP-β-S (14.3 ± 4.9 mV; n=4; p<0.001; Fig. 2.5A, 2.5D, 2.5E) compared tocontrol (61.5 ± 6.4 mV; n=7; p<0.001; Fig. 2.5A, 2.5E) TGNs. Next, I determined theinvolvement of PLC in LA signaling in TGNs. Depolarization in response to LA issignificantly reduced in presence of the PLC blocker U73122 (3 μM) (25.8 ± 4.6 mV; 47n=8; Fig. 2.5B, 2.5E) compared to the treatment with the inactive analog U73343 (3 μM)(61.3 ± 2.6 mV; n=7; p<0.005; Fig. 2.5C, 2.5E). DiscussionThe prevailing view of how fats are detected in oral cavity is attributed largely totheir textural properties (22, 23). Inherent in this viewpoint is the assumption that fats ortheir chemical constituents are capable of activating the somatosensory system. However,direct evidence about the ability of fats to stimulate somatosensory system is lacking andthe mechanisms by which they do so remain to be elucidated. Dietary fat, presumably inthe form of triglycerides contains a significant amount of FAs like linoleic and oleic acids(24, 25). The amount of FAs would further increase by hydrolysis during heating (26).Our results demonstrated that FAs could stimulate a[Ca]i increase as well as directlydepolarize TGNs, which provided the first direct evidence that FAs may act assomatosensory stimuli at concentrations found in vivo.I further attempted to unravel the LA-initiated signaling cascade that leads to the[Ca]i rise. Careful studies of Na substitution revealed that LA-induced[Ca]i increasewas reduced about as low as 8% across all the neurons I tested. This result indicated thatthe extracellular Na did involve in the general FA-sensitive signaling, but the smallamount of reduction suggested its role not crucial. Meanwhile, I performed the parallelstudy of substituting for extracellularCa. Interestingly, the Ca dependence ofresponses to LA varied across neurons. Over 65% TGNs showed more than 50%reduction in the magnitude of LA-induced Ca responses, while the rest (~35%)appeared independent of extracellular Ca. Although this variability precludes me from 48making any definite conclusions about the role of extracellular Ca in LA-inducedresponses, it did lead to the interpretation that there may be different classes of neurons inrespect to their ability in responding to FAs. Furthermore, these results indicated thatthere may be differences in the relative involvement of intracellular Ca stores in the LAsignaling pathway. I took the use of the non-competitive SERCA inhibitor thapsigargin inthis study. Interestingly, thapsigargin blocked the LA-induced[Ca]i rise in 68% ofTGNs, which was consistent with the percentage of neurons that were dependent ofextracellularCa. On the other hand, the neurons that were independent of extracellularCa seemed to have the thapsigargin-insensitive stores. The underlying physiologicalfunction differences between these two groups are still unclear.There are several receptor candidates of FAs that have been identified recently.These include, a group of FA-sensitive GPCRs and the FA translocase, CD36 (18, 27-30). Long chain FAs receptor GPR40, for example, has been reported to amplify glucose-stimulated insulin secretion in pancreatic β cells (20, 21). GPR120, another long chainunsaturated FAs receptor, was thought to regulate the GLP-1 secretion in enteroendocrinecell line STC-1 (18). In this dissertation, I first showed the mRNA expression of FA-sensitive GPCRs in trigeminal ganglia. Further, I checked whether G protein/PLCpathway was involved in the signaling cascade. The results from pharmacological studiesusing G protein blocker GDP-β-S and general PLC blocker U73122 yielded resultsconsistent with the interpretation that the G protein/PLC pathway was involved in theFAs (i.e. LA) signaling pathways in TGNs. Nevertheless, direct evidence implicating the 49involvement of specific GPCRs as FAs receptors in somatosensory neurons will requireadditional studies.TGNs responded to a series of FAs, which varied in carbon chain length, and thenumber and position of double bonds. I observed significant delays in FA-evokedphysiological changes in[Ca]i and in membrane depolarization, which was particularlyevident for polyunsaturated FAs. While the nature of this delayed response is not clear,there could be several explanations. First, the LA-induced[Ca]i increase after storedepletion indicates that SOCE might be involved. It was reported that some potentialcomponents in the SOCE pathways might exhibit physiologically relevant, slowresponses (31, 32). Secondly, FA-induced activation of sensory neurons might involveother membrane proteins, like CD36 (30, 33). The exact mechanisms of FAs’ binding toGPCRs and the involvement, if any, with the FAs transporter CD36 is unclear (34). Arate limiting process might exist to slow the physiological response seen in theseexperiments. Thirdly, I tried other FAs which include OA, MA, and DHA. The[Ca]iincreases to OA (long chain monounsaturated FA) and MA (medium chain saturated FA)were faster than LA and DHA, which suggested that they might activate differentreceptors (28, 35). The details of the cellular/molecular explanations of the delay remainto be studied.A significant confounding variable in studying isolated TGNs lies in the fact ofthe mixed population of polymodal neurons. The trigeminal nerve contains both efferentand afferent nerves, not to mention its innervations to the face, nasal and neck areas.Indeed, I observed the variations among TGNs in the extracellular Ca substitution and 50thapsigargin treatment experiments, consistent with there being multiple populations ofcell types that are capable of responding to FAs. However, the underlying physiologicalexplanations of these variations remained unanswered. To this end, however, I willpartially answer this question in Chapter 3, by using pseudorabies virus constructs as a“live-cell tracer,” to identify that subpopulation of neurons that innervates the oral cavity.My studies of FAs activating somatosensory system provide new insights into thedietary fat detection in oral cavity. The somatosensory system apparently recognizes notonly the triglyceride form of fat but also shows strong cellular activation by FAs, theproximate stimulus for fat “taste.” Future studies on identified subpopulations of TGNsmay better refine the transduction pathways including identifying the FA-activated TRPchannels, their relationship with intracellular Ca store, and direct evidence of GPCRsinvolvement in the FA-transduction pathway. References1. Caterina MJ, et al. (1997) The capsaicin receptor: a heat-activated ion channel inthe pain pathway. Nature 389: 816-824.2. McKemy DD, Neuhausser WM, Julius D (2002) Identification of a cold receptorreveals a general role for TRP channels in thermosensation. Nature 416: 52-58.3. Peier AM, et al. (2002) A TRP channel that senses cold stimuli and menthol. Cell108: 705-715.4. Bautista DM, et al. (2005) Pungent products from garlic activate the sensory ionchannel TRPA1. Proc Natl Acad Sci U S A 102: 12248-12252.5. Bautista DM, et al. (2008) Pungent agents from Szechuan peppers excite sensoryneurons by inhibiting two-pore potassium channels. Nat Neurosci 11: 772-779.6. Stewart JE, et al. (2010) Oral sensitivity to fatty acids, food consumption andBMI in human subjects. Br J Nutr: 1-8. 517. Yoneda T, et al. (2009) Assessing palatability of long-chain fatty acids from thelicking behavior of BALB/c mice. Physiol Behav 96: 735-741.8. Pittman DW, et al. (2008) Orosensory detection of fatty acids by obesity-proneand obesity-resistant rats: Strain and sex differences. Chem Senses 33: 449-460.9. Li Y, Garcia-Gonzalez DL, van de Voort F (2008) Determination of free fattyacids in edible oils with the use of a variable filter array IR spectrometer. J AmOil Chem Soc 85: 599-604.10. Gilbertson TA, Fontenot DT, Liu L, Zhang H, Monroe WT (1997) Fatty acidmodulation of K+ channels in taste receptor cells: Gustatory cues for dietary fat.Am J Physiol 272: C1203-1210.11. Bhattacharya MR, et al. (2008) Radial stretch reveals distinct populations ofmechanosensitive mammalian somatosensory neurons. Proc Natl Acad Sci U S A105: 20015-20020.12. Stucky CL, Lewin GR (1999) Isolectin B(4)-positive and -negative nociceptorsare functionally distinct. J Neurosci 19: 6497-6505.13. Liu L, Chang GQ, Jiao YQ, Simon SA (1998) Neuronal nicotinic acetylcholinereceptors in rat trigeminal ganglia. Brain Res 809: 238-245.14. Liu L, Simon SA (2000) Capsaicin, acid and heat-evoked currents in rattrigeminal ganglion neurons: relationship to functional VR1 receptors. PhysiolBehav 69: 363-378.15. Hoth M, Penner R (1992) Depletion of intracellular calcium stores activates acalcium current in mast cells. Nature 355: 353-356.16. Yang KT, Chen WP, Chang WL, Su MJ, Tsai KL (2005) Arachidonic acidinhibits capacitative Ca2+ entry and activates non-capacitative Ca2+ entry incultured astrocytes. Biochem Biophys Res Commun 331: 603-613.17. Chang WC, Parekh AB (2004) Close functional coupling between Ca2+ release-activated Ca2+ channels, arachidonic acid release, and leukotriene C4 secretion. JBiol Chem 279: 29994-29999.18. Hirasawa A, et al. (2005) Free fatty acids regulate gut incretin glucagon-likepeptide-1 secretion through GPR120. Nat Med 11: 90-94.19. Tanaka T, et al. (2008) Free fatty acids induce cholecystokinin secretion throughGPR120. Naunyn Schmiedebergs Arch Pharmacol 377: 523-527. 5220. Morgan NG, Dhayal S (2009) G-protein coupled receptors mediating long chainfatty acid signalling in the pancreatic beta-cell. Biochem Pharmacol 78: 1419-1427.21. Itoh Y, et al. (2003) Free fatty acids regulate insulin secretion from pancreaticbeta cells through GPR40. Nature 422: 173-176.22. Mattes RD (2009) Is there a fatty acid taste? Annu Rev Nutr 29: 305-327.23. Rolls ET (2008) The affective and cognitive processing of touch, oral texture, andtemperature in the brain. Neurosci Biobehav Rev 34: 237-245.24. Bertran E, et al. (1999) Determination of olive oil free fatty acid by Fouriertransform infrared spectroscopy. J Am Oil Chem Soc 76: 611-616.25. Wan PJ, Pakarinen DR, Wakelyn PJ (1998) Concerns for the determination offree fatty acid in cotton seed. J. Am. Oil Chem. Soc 75: 1321-1324.26. Orthoefer FT, Gurkin S, Liu L (1996) Deep frying: Chemistry, nutrition, andpractical applications (AOCS Press, Champaign, IL).27. Briscoe CP, et al. (2003) The orphan G protein-coupled receptor GPR40 isactivated by medium and long chain fatty acids. J Biol Chem 278: 11303-11311.28. Brown AJ, et al. (2003) The Orphan G protein-coupled receptors GPR41 andGPR43 are activated by propionate and other short chain carboxylic acids. J BiolChem 278: 11312-11319.29. Ibrahimi A, Abumrad NA (2002) Role of CD36 in membrane transport of long-chain fatty acids. Curr Opin Clin Nutr Metab Care 5: 139-145.30. Laugerette F, et al. (2005) CD36 involvement in orosensory detection of dietarylipids, spontaneous fat preference, and digestive secretions. J Clin Invest 115:3177-3184.31. Kim SJ, et al. (2003) Activation of the TRPC1 cation channel by metabotropicglutamate receptor mGluR1. Nature 426: 285-291.32. Amaral MD, Pozzo-Miller L (2007) TRPC3 channels are necessary for brain-derived neurotrophic factor to activate a nonselective cationic current and toinduce dendritic spine formation. J Neurosci 27: 5179-5189.33. Abumrad NA (2005) CD36 may determine our desire for dietary fats. J ClinInvest 115: 2965-2967. 5334. Su X, Abumrad NA (2009) Cellular fatty acid uptake: a pathway underconstruction. Trends Endocrinol Metab 20: 72-77.35. Wang J, Wu X, Simonavicius N, Tian H, Ling L (2006) Medium-chain fatty acidsas ligands for orphan G protein-coupled receptor GPR84. J Biol Chem 281:34457-34464. 54Table 2.1: The primer sequences for GPR40, 43, 120 and CD36. Target GenBankAccession No. Sense primer/Antisense primer/ProbeCorresponding nucleotidesequence5’-TTCTTTCTGCCCTTGGTTAT-3’575-594GPR40 NM_153304 5’-GCCCTGAGCTTCCGTTTG-3’652-6695’-ACTGCTTTCTGCTATGTGGGCTGCCTC-3’ 596-6225’-GCGGGCATCAGCATAGAA-3’301-318GPR43 NM_001005877 5’-CCCACCTGCTCGGTTGAGTT-3’451-4705’-TGGCTTTCCCGGTGCAGTACAAGCTATC-3’ 332-3595’-CTGCACATTGGATTGGC-3’624-640GPR120 AB2078685’-TCTGGTGGCTCTCGGAGTAT-3’783-8025’-CTCTTCCGCGAGGCTTTCGTGATCTGT-3’ 738-7645’-GGTGTGCTGGACATTGG -3’1000-1016CD36 AF0724115’-CTATGCTCATCTTCGTTAGG -3’1108-11275’-TCAAGCCTTCGATAGGTTCTGAGACATCA-3’ 1077-1105 55Table 2.2: The structural information of fatty acids used in the present study. CompoundsStructuralNomenclature Bond Positions Response in TGNsCaproic acidC6:0no responseMyristic acidC14:0Yes (26%, n=151)Oleic acidC18:1, n-9 Δ9Yes (23%, n=83)Linoleic acidC18:2, n-6 Δ9, 12Yes (65%, n=332)Arachidonic acid C20:4, n-6 Δ5, 8, 11, 14Yes (36%, n=22)DocosahexaenoicacidC22:6, n-3 Δ4, 7, 10, 13, 16, 19 Yes (59%, n=22)