Skip to main content
PMC home page
The Journal of Physiology logoLink to The Journal of Physiology
. 2001 Jan 15;530(Pt 2):287–293. doi: 10.1111/j.1469-7793.2001.0287l.x

Optical recordings of taste responses from fungiform papillae of mouse in situ

Yoshitaka Ohtubo *, Toshiyuki Suemitsu *, Satoshi Shiobara *, Takafumi Matsumoto *, Takashi Kumazawa †, Kiyonori Yoshii *
PMCID: PMC2278412  PMID: 11208976

Abstract

  1. Single taste buds in mouse fungiform papillae consist of ∼50 elongated cells (TBCs), where fewer than three TBCs have synaptic contacts with taste nerves. We investigated whether the non-innervated TBCs were chemosensitive using a voltage-sensitive dye, tetramethylrhodamine methyl ester (TMRM), under in situ optical recording conditions.

  2. Prior to the optical recordings, we investigated the magnitude and polarity of receptor potentials under in situ whole-cell clamp conditions. In response to 10 mM HCl, several TBCs were depolarized by ∼25 mV and elicited action potentials, while other TBCs were hyperpolarized by ∼12 mV. The TBCs eliciting hyperpolarizing receptor potentials also generated action potentials on electrical stimulation.

  3. A mixture of 100 mM NaCl, 10 mM HCl and 500 mM sucrose depolarized six TBCs and hyperpolarized another three TBCs out of 13 identified TBCs in a taste bud viewed by optical section. In an optical section of another taste bud, 1 M NaCl depolarized five TBCs and hyperpolarized another two TBCs out of 11 identified TBCs.

  4. The number of chemosensitive TBCs was much larger than the number of innervated TBCs in a taste bud, indicating the existence of chemosensitivity in non-innervated TBCs. There was a tendency for TBCs eliciting the same polarity of receptor potential to occur together in taste buds. We discuss the role of non-innervated TBCs in taste information processing.

Single taste buds comprise many cells, but only a few elongated taste bud cells (TBCs) are in synaptic contact with taste nerves in mouse fungiform papillae (Kinnamon et al. 1993). A quantitative study showed that the number was less than three (Seta & Toyoshima, 1995). Non-innervated TBCs have been considered to have supportive functions. However, we may conclude from studies showing a substantial number of TBCs responding to taste substances that, like innervated TBCs, non-innervated TBCs are chemosensitive.

The taste responses of TBCs have been investigated electrophysiologically as reviewed by Herness & Gilbertson (1999) and Lindemann (1996). Although these recordings promoted a better understanding of taste receptor mechanisms, they were inappropriate for counting all of the chemosensitive TBCs in a taste bud. Optical recordings are much more suitable because they examine the taste responses of all TBCs in a taste bud at once. Several scientists have already investigated taste responses optically (Akabas et al. 1988; Bernhardt et al. 1996; Hayashi et al. 1996). However, they focused on taste transduction mechanisms rather than on the present issues.

We developed an in situ whole-cell patch clamp technique with a peeled mouse lingual epithelium, which was as effective as the tongue epithelium in situ in protecting basolateral membranes from solutions applied to receptor membranes (Furue & Yoshii, 1997a, 1998). In this study, we first investigated the receptor potential of TBCs under in situ patch clamp conditions. Then we counted the number of responding TBCs by applying an optical recording method, where we chose taste-stimulating solutions (a mixture stimulus and 1 m NaCl solution) to maximize the number of TBCs responding.

The in situ patch clamp recordings showed that taste substances depolarized several TBCs and hyperpolarized other TBCs. The optical recordings showed that at least six TBCs were depolarized and at least another three TBCs were hyperpolarized in the same taste bud in response to the mixture of taste substances. These results indicate that non-innervated TBCs are chemosensitive. We discuss the possibility that non-innervated TBCs modify the output of innervated TBCs.

METHODS

Peeled lingual epithelia

We prepared peeled lingual epithelia as described previously (Furue & Yoshii, 1997a, 1998) in accordance with Guiding Principles for the Care and Use of Animals in the Field of Physiological Sciences approved by the Council of the Physiological Society of Japan. In brief, we subcutaneously injected a collagenase solution into tongues obtained from mice that had been anaesthetized with ether and then decapitated. Epithelia were peeled from the tongues with forceps ∼5 min after collagenase injection, and mounted on a recording platform (Fig. 1), where the peeled epithelium protected the basolateral membrane as an effective barrier against deionized water and taste solutions such as high concentrations of HCl or NaCl for more than 60 min (Furue & Yoshii, 1997a, 1998).

Figure 1. In situ optical and patch clamp recording scheme (A) from a taste bud in a peeled lingual preparation (B).

Figure 1

Open in a new tab

A, continuous arrow, excitation light path; dashed arrow, emission light path. B, the electrode was removed for optical recordings and a photograph of the taste bud was taken with a cooled CCD camera (C4742-95-12, Hamamatsu Photonics K.K.). Scale bar, 20 μm.

Optical recordings

We monitored the membrane potential of TBCs with a voltage-sensitive dye, tetramethylrhodamine methyl ester (TMRM). The basolateral membranes of TBCs were stained with 1 μm TMRM dissolved in a physiological saline solution for 30 min on the recording platform. While staining basolateral membranes, their receptor membranes facing down onto the recording platform were immersed in the control solution (physiological saline). Since the peeled epithelium separated the solution applied on the basolateral membrane side from that on the receptor membrane side by preserving tight junctions between cells, only the basolateral membranes were stained.

After TMRM had been washed from extracellular spaces with the control solution, the epithelium on the recording platform was placed under an upright microscope (BX50, Olympus, Tokyo, Japan) with a x60 water-immersion objective in a dark room (Fig. 1A). The platform chamber (receptor membrane side) was irrigated with adapting deionized water and stimulating solutions. The basolateral membranes were irrigated with the control solution during experiments. The epithelium was epi-illuminated by a 100 W halogen lamp. A dichroic mirror unit (U-MWIG, Olympus) reflected wavelengths between 520 and 550 nm, exciting TMRM (maximum excitation at 548 nm), and transmitted fluorescence through a high-pass filter (565 nm) to a SIT camera (C2400-08, Hamamatsu Photonics K.K., Hamamatsu, Japan).

Sequences of video frames taken at a rate of one frame per 60-120 s were digitized (640 pixels x 480 pixels x 8 bit) to decimal values in the range 0-255 with a video capture board (LG-3, Scion Corporation, Frederick, MD, USA). In order to achieve high-resolution recordings of fluorescence intensitie (FI), we biased against the FI before the onset of stimulation, where the lowest FI (biased level) in a pixel yielded a pixel value of ∼30 and the maximal FI in a pixel was ∼200, though these pixel values differed from epithelium to epithelium. Although such biasing prevented us from obtaining relative changes in fluorescence intensity and a quantitative understanding of the magnitude of the taste responses, our goal in this study was to investigate taste responses qualitatively. The sequential images thus obtained were processed with NIH Image (version 1.6.2). The FI of TMRM was increased on depolarization and decreased on hyperpolarization, which was confirmed with cultured hippocampal neurons and liposomes (data not shown).

The vibration or drift of a peeled lingual epithelium distorts optical recordings. We eliminated vibration by lining epithelia with a nylon mesh taken from stockings and by keeping the hydraulic pressure on them constant (Furue & Yoshii, 1998). The recording platform drifted by ∼0.9 μm for 10 min. Since the maximum duration between taking video frames was 120 s, the amount of drift between the frames was 0.18 μm on average. This value was close to the resolution limit of the microscopes and hence we disregarded the drift.

Scaling of fluorescence intensity

Pixel values differed among pixels relating to the same TBCs (Fig. 3A). Since the horizontal, optical slice of single TBCs must remain at equipotential, the different FI showed inhomogeneous staining with TMRM. In order to neutralize the effects of different staining on taste responses, we averaged the FI in pixels contained in the same TBCs.

Figure 3. Optical responses to the mixture stimulus.

Figure 3

Open in a new tab

A, sequence of fluorescence images (upper panel) taken every 120 s and corresponding subtracted frames (lower panel) indicated by the numbers below the panels. Note that the spontaneous decrease in FI was not compensated in the images in this figure and in Fig. 4. Scales under the images show the range of digitized FI, 0 to 255 (upper scale) and -84 to +84 (lower scale). D.W., deionized water. B, SIT camera image of a taste bud (upper photograph) and a subtracted frame (‘3 - 2’ of the taste bud shown in A) overlaid with the layout of identified TBCs (lower photograph). Scale bars in A and B, 20 μm. C, relative optical responses of numbered TBCs in these series. Here and in Fig. 4, TBC numbers correspond to trace numbers and filled circles show the optical responses with taste stimulation. Horizontal scale bar, 5 min; vertical scale bar, 100% (see Methods).

The FI spontaneously decreased, probably due to bleaching or to dissociation of TMRM. We estimated the magnitude of the spontaneous decrease by fitting exponential curves to the average FI in identified TBCs obtained from baseline video frames, with sequential frames before and after taste stimulation. The difference between the average FI with taste substances and the FI estimated from the fitted curve yielded the response FI.

Then we normalized the response FI in each TBC relative to the maximal response FI in a TBC found in the same sequential frames. In these experiments, the normalized response FI thus obtained was used as the relative optical response of TBCs.

Patch clamp recordings

The receptor potentials of TBCs in peeled epithelia were investigated with an in situ patch clamp technique (Furue & Yoshii, 1997a, 1998). We placed a recording glass pipette filled with a KCl electrode solution on the basolateral membrane of a TBC under visual guidance to perform whole-cell patch clamp recording (Fig. 1A). The ground electrode was placed outside the recording platform, and was always immersed in the control solution flowing from the basolateral membrane. We also investigated taste responses under a cell-attached condition that is similar to in situ patch clamp recording except that the patch membrane is not ruptured.

We irrigated receptor membranes and basolateral membranes as for the optical recordings. Taste responses thus recorded under current clamp conditions (#3900A, Dagan Corporation, Minneapolis, MN, USA) were digitized with pCLAMP 7/Digidata 1200 (Axon Instruments).

Solutions

All solutions were prepared with deionized water. Control solution (mm): 150 NaCl, 5 KCl, 2 CaCl2, 0.5 MgCl2, 10 glucose and 5 Hepes-NaOH, pH 7.4. KCl electrode solution (mm): 120 KCl, 5 MgCl2, 10 EGTA, 5 Na2ATP, 0.3 Na3GTP and 10 Hepes-NaOH, pH 7.2. Collagenase solution: 0.4% collagenase (Sigma type IA) dissolved in control solution.

RESULTS

Patch clamp recordings

Prior to beginning optical recordings, we investigated the magnitude and polarity of TBC responses under in situ patch clamp conditions. In response to 500 mm NaCl, TBCs elicited depolarizing receptor potentials of ∼40 mV with action potentials on the rising slope of depolarization (Fig. 2A). It is likely that the depolarizing receptor potentials activated voltage-gated channels such as Na+ and K+ channels, which gave rise to action potentials. The long lasting depolarization then inactivated Na+ channels, stopping the firing of action potentials. Two TBCs out of 34 TBCs examined elicited depolarizing responses to 500 mm NaCl.

Figure 2. In situ patch clamp recordings of taste responses obtained from different TBCs.

Figure 2

Open in a new tab

A-C, receptor potentials obtained with the in situ whole-cell clamp technique. The two traces in C were obtained from the same TBC. D, action currents recorded in the in situ cell-attached clamp condition. Arrowheads in A and B indicate the action potentials shown on a magnified time scale on the right. Action potentials in C were elicited by depolarizing current (lower right-hand trace). Bars above the traces represent the application duration of the indicated stimulus solutions. Dashed lines in A-C represent the 0 mV level. Holding currents, -40 pA in A, -6 pA in B, and -20 pA (right) and -30 pA (left) in C. Holding potential, 0 mV in D. Horizontal scale bars, 50 s in A-C and 20 s in D. Vertical scale bars, 25 mV in A-C and 50 pA in D. Horizontal scale bars for magnified action potentials and action current, 20 ms in A and B, 5 ms in C, and 2 ms in D.

In response to 10 mm HCl, several TBCs elicited similar depolarizing receptor potentials of ∼25 mV (Fig. 2b). Action potentials occurred at the beginning of the depolariz ation but disappeared during large depolarizations, due probably to the inactivation of Na+ channels. Other TBCs elicited hyperpolarizing receptor potentials in response to 10 mm HCl (Fig. 2c). TBCs eliciting such hyperpolarizing receptor potentials generated action potentials on depolarization. HCl (10 mm) depolarized four TBCs and hyperpolarized two TBCs out of 29 TBCs examined.

In addition to the above whole-cell recordings, we examined taste responses under a cell-attached condition, which recorded action currents generated by action potentials. The application of a mixture of 200 mm NaCl, 10 mm HCl and 10 mm quinine suppressed spontaneous discharges (Fig. 2D), indicating that the mixture hyperpolarized this TBC. Although depolarizing receptor potentials also suppressed action potential firing by inactivating voltage-gated Na+ channels, the channels would elicit action potentials before inactivation. The absence of an increase in the number of action currents at the onset of stimulation showed that the mixture hyperpolarized this TBC. The patch clamp recordings thus showed that taste substances depolarized several TBCs and hyperpolarized others.

Optical recordings

A mixture of 100 mm NaCl, 10 mm HCl and 500 mm sucrose changed the FI in a peeled lingual epithelium (Fig. 3). We subtracted each preceding frame from the next one, pixel by pixel, and generated net changes in the FI. On such subtracted frames, the FI was increased in some regions (depolarization) and decreased in others (hyperpolarization) in response to the mixture stimulus (Fig. 3A). The FI changes were reversed when the mixture was washed out with deionized water, indicating that the observed changes mainly resulted from taste receptor potentials.

We identified 13 TBCs by their contours in SIT camera images in the taste bud shown in Fig. 3. Of the 13 TBCs, the FI was increased in six and decreased in another three (Fig. 3b). Changes in FI were undetectable in four TBCs. These results showed that the mixture depolarized six TBCs, hyperpolarized three TBCs and had little effect on the membrane potentials of four TBCs. Depolarized TBCs (nos 1, 3 and 4 in Fig. 3b) were arranged close to each other. This TBC group also involved other TBCs though their contours were not clear. Similarly, TBC no. 2 was neighboured by depolarizing TBCs. The hyperpolarized TBCs (nos 7 and 8) were also neighboured by hyper polarizing TBCs. The results indicated a tendency for closely arranged TBCs to elicit a receptor potential of the same polarity.

Not only the mixture stimulus but also 1 m NaCl elicited both depolarized and hyperpolarized responses in TBCs (Fig. 4). Similar to the responses to the mixture stimulus, the FI changed at the onset of stimulation, and these changes reversed at the end of the stimulation. In this epithelium, we identified 11 TBCs by their contours. Again, five were depolarized, another two were hyperpolarized and four showed no response. As for the layout of TBCs responding to the mixture stimulus, depolarized and hyperpolarized TBCs tended to form different groups.

Figure 4. Optical responses to 1 m NaCl.

Figure 4

Open in a new tab

A, subtracted frames for fluorescence images taken every 60 s. B, SIT camera image of the taste bud (left), and the second subtracted image in A overlaid with the layout of identified TBCs (right). Scale bars in A and B, 20 μm. C, relative optical responses of the numbered TBCs in these series. Horizontal scale bar, 5 min; vertical scale bar, 100% (see Methods).

DISCUSSION

In this study, patch clamp recordings showed that depolarizing receptor potentials elicited action potentials (Fig. 2A and B). Although we hyperpolarized TBCs with inward currents to facilitate action potential firing, our previous study showed that taste substances elicited action currents without hyperpolarization under in situ cell-attached recording conditions (Furue & Yoshii, 1997a). The hyperpolarizing receptor potentials (Fig. 2c) may result from an increase in K+ conductance because the magnitude of hyperpolarization was decreased when we manually clamped TBCs close to the equilibrium potential for K+ (approximately -83 mV at basolateral membranes and more negative at receptor membranes).

Optical recordings also showed both depolarizing and hyperpolarizing receptor potentials. At least six TBCs were depolarized and three TBCs were hyperpolarized in response to the mixture stimulus in the same taste bud (Fig. 3). The increases and decreases in FI were also found in pixels not involved in identified TBCs, showing that the number of chemosensitive TBCs was larger than nine. Also, the number should be increased with taste solutions containing a larger number and a higher concentration of taste substances. Even with such a complex mixture for stimulation, the number of responding TBCs might be an underestimate because a TBC may be depolarized by one taste substance but hyperpolarized by another taste substance contained in the mixture.

Only a few TBCs are in synaptic contact with taste nerves in mouse fungiform papillae (Kinnamon et al. 1993). This number is much lower than that in mouse vallate papillae where 20-30% of TBCs had synapses (Kinnamon et al. 1985). Seta & Toyoshima (1995) counted the number of type III cells in single taste buds in mouse fungiform papillae and found that the maximum number was three, with one or two being more usual. Also they found that only type III cells had synaptic contact with taste nerves. Although the number of chemosensitive TBCs found in the present study was probably underestimated, it was much larger than three. Therefore, we concluded that non-innervated TBCs respond to taste substances.

Such non-innervated but chemosensitive TBCs may fail to contribute to taste perception. However, we consider that non-innervated chemosensitive TBCs modify the output of innervated TBCs as follows. TBCs elicited action potentials in response to taste substances (Fig. 2). In fact, almost all TBCs investigated in mouse fungiform papillae expressed voltage-gated Na+ and K+ channel currents and many of them elicited action potentials on depolarization (Furue & Yoshii, 1997a, b). Depolarized TBCs increase the extracellular K+ concentration by releasing K+ through voltage-gated K+ channels, as found in the nervous systems of various animals (Frankenhaeuser & Hodgkin, 1956; Orkand et al. 1966; Somjen, 1979), which supports the depolarization of innervated TBCs and stimulates neurotransmitter release. In contrast, hyperpolarized TBCs take up K+ and antagonize the above accelerating effect. These opposing effects may code taste information into complex firing patterns of single taste nerves in response to taste substances (Nagai & Ueda, 1981).

Alternatively, non-innervated chemosensitive TBCs together with innervated TBCs may form cell networks by releasing neurotransmitter among themselves, which also could regulate the firing patterns of taste nerves. In these TBC networks, we classify the receptor potentials into first responses (responses to taste substances) and second responses (responses to transmitters released by TBCs).

Electrical synapses among TBCs were found in rat vallate papillae as gap junctions (Akisaka & Oda, 1978). Also, electrical and chemical synapses were found in amphibians (Bigiani & Roper, 1992, 1994). Further investigations may discover such synapses among TBCs in mouse fungiform papillae, which would support our hypothesis on TBC networks. Alternatively, other messenger systems may function in taste buds as in the brain, where neural networks utilize membrane-permeable transmitters such as NO (Bredt & Snyder, 1994) and CO (Verma et al. 1993).

References

  1. Akabas MH, Dodd J, Al-Awqati Q. A bitter substance induces a rise in intracellular calcium in a subpopulation of rat taste cells. Science. 1988;242:1047–1050. doi: 10.1126/science.3194756. [DOI] [PubMed] [Google Scholar]
  2. Akisaka T, Oda M. Taste buds in the vallate papillae of the rat studied with freeze-fracture preparation. Archivum Histologicum Japonicum. 1978;41:87–98. doi: 10.1679/aohc1950.41.87. [DOI] [PubMed] [Google Scholar]
  3. Bernhardt SJ, Naim M, Zehavi U, Lindemann B. Changes in IP3 and cytosolic Ca2+ in response to sugars and non-sugar sweeteners in transduction of sweet taste in the rat. Journal of Physiology. 1996;490:325–336. doi: 10.1113/jphysiol.1996.sp021147. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Bigiani A, Roper SD. Reduction of electrical coupling between Necturus taste receptor cells, a possible role in acid taste. Neuroscience Letters. 1994;176:212–216. doi: 10.1016/0304-3940(94)90085-x. [DOI] [PubMed] [Google Scholar]
  5. Bredt DS, Snyder SH. Nitric oxide: a physiologic messenger molecule. Annual Review of Biochemistry. 1994;63:175–195. doi: 10.1146/annurev.bi.63.070194.001135. [DOI] [PubMed] [Google Scholar]
  6. Ewald DA, Roper SD. Intercellular signaling in Necturus taste buds: chemical excitation of receptor cells elicits responses in basal cells. Journal of Neurophysiology. 1992;67:1316–1324. doi: 10.1152/jn.1992.67.5.1316. [DOI] [PubMed] [Google Scholar]
  7. Ewald DA, Roper SD. Bidirectional synaptic transmission in Necturus taste buds. Journal of Neuroscience. 1994;14:3791–3804. doi: 10.1523/JNEUROSCI.14-06-03791.1994. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Frankenhaeuser B, Hodgkin AL. The after-effects of impulses in the giant nerve fibres of Loligo. Journal of Physiology. 1956;131:341–376. doi: 10.1113/jphysiol.1956.sp005467. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Furue H, Yoshii K. In situ tight-seal recordings of taste substance-elicited action currents and voltage-gated Ba currents from single taste bud cells in the peeled epithelium of mouse tongue. Brain Resarch. 1997a;776:133–139. doi: 10.1016/s0006-8993(97)00974-8. [DOI] [PubMed] [Google Scholar]
  10. Furue H, Yoshii K. Different distributions of voltage-gated channels in mouse taste buds. Society for Neuroscience Abstracts. 1997b;23:408.10. [Google Scholar]
  11. Furue H, Yoshii K. A method for in-situ tight-seal recordings from single taste bud cells of mice. Journal of Neuroscience Methods. 1998;84:109–114. doi: 10.1016/s0165-0270(98)00104-6. [DOI] [PubMed] [Google Scholar]
  12. Hayashi Y, Zviman MM, Brand JG, Teeter JH, Restrepo D. Measurement of membrane potential and [Ca2+]i in cell ensembles: application to the study of glutamate taste in mice. Biophysical Journal. 1996;71:1057–1070. doi: 10.1016/S0006-3495(96)79306-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Herness MS, Gilbertson TA. Cellular mechanisms of taste transduction. Annual Review of Physiology. 1999;61:873–900. doi: 10.1146/annurev.physiol.61.1.873. [DOI] [PubMed] [Google Scholar]
  14. Kinnamon JC, Henzler DM, Royer SM. HVEM ultrastructural analysis of mouse fungiform taste buds, cell types, and associated synapses. Microscopy Research and Technique. 1993;26:142–156. doi: 10.1002/jemt.1070260207. [DOI] [PubMed] [Google Scholar]
  15. Kinnamon JC, Taylor BJ, Delay RJ, Roper SD. Ultrastructure of mouse vallate taste buds. I. Taste cells and their associated synapses. Journal of Comparative Neurology. 1985;235:48–60. doi: 10.1002/cne.902350105. [DOI] [PubMed] [Google Scholar]
  16. Lindemann B. Taste reception. Physiological Reviews. 1996;76:718–766. doi: 10.1152/physrev.1996.76.3.719. [DOI] [PubMed] [Google Scholar]
  17. Nagai T, Ueda K. Stochastic properties of gustatory impulse discharges in rat chorda tympani fibers. Journal of Neurophysiology. 1981;45:574–592. doi: 10.1152/jn.1981.45.3.574. [DOI] [PubMed] [Google Scholar]
  18. Orkand RK, Nicholls JG, Kuffler SW. Effect of nerve impulses on the membrane potential of glial cells in the central nervous system of amphibia. Journal of Neurophysiology. 1966;29:788–806. doi: 10.1152/jn.1966.29.4.788. [DOI] [PubMed] [Google Scholar]
  19. Seta Y, Toyoshima K. Three-dimensional structure of the gustatory cell in the mouse fungiform taste buds: a computer-assisted reconstruction from serial ultrathin sections. Anatomy and Embryology. 1995;191:83–88. doi: 10.1007/BF00186781. [DOI] [PubMed] [Google Scholar]
  20. Somjen GG. Extracellular potassium in the mammalian central nervous system. Annual Review of Physiology. 1979;41:159–177. doi: 10.1146/annurev.ph.41.030179.001111. [DOI] [PubMed] [Google Scholar]
  21. Verma A, Hirsch DJ, Glatt CE, Ronnett GV, Snyder SH. Carbon monoxide: a putative neural messenger. Science. 1993;259:381–384. doi: 10.1126/science.7678352. [DOI] [PubMed] [Google Scholar]

Articles from The Journal of Physiology are provided here courtesy of The Physiological Society

RESOURCES