- Email: [email protected]
Convergence of olfactory and gustatory connections onto the endopiriform nucleus in the rat
Neuroscience 126 (2004) 1033–1041
CONVERGENCE OF OLFACTORY AND GUSTATORY CONNECTIONS ONTO THE ENDOPIRIFORM NUCLEUS IN THE RAT W. FU,a T. SUGAI,a H. YOSHIMURAb AND N. ONODAa*
probably receives temperature signals from the tongue (Kosar et al., 1986; Norgren, 1995). The GI caudal to the middle cerebral artery (MCA) receives general visceral afferents rather than taste signals (Cechetto and Saper, 1987; Saper, 1995). Further, the prominent efferent projections to autonomic structures arise from the agranular division (AI) of the insular cortex (IC), just ventral to the DI (Allen et al., 1991; Saper, 1995). Furthermore, intrinsic connections between the divisions in the IC have been reported (Shi and Cassell, 1998). However, there are many different observations related to communication and differentiation of function between the IC divisions. The primary olfactory cortex is defined as the region that receives direct fiber projections from the olfactory bulb (Price, 1987). The piriform cortex (PC) is the largest structure in the olfactory cortex and further the piriform region includes a nucleus localized in between the PC proper and the caudate-putamen (Haberly, 1998), which is named the “endopiriform nucleus” (EPN). The PC is an important center before delivery of olfactory information to the higher system via the EPN (Haberly, 1998). Tracer studies have shown that the long axons of the EPN neurons form a net of intrinsic connections and project over long distances to the forebrain structures, such as the olfactory cortex, IC, and amygdala (Price, 1987; Behan and Haberly, 1999). Unit recording studies demonstrated that odor stimuli activated single neurons in the PC (Haberly, 1969; Tanabe et al., 1975; Nemitz and Goldberg, 1983; Schoenbaum and Eichenbaum, 1995; Wilson, 1998, 2000), hypothalamus (Scott and Pfaffmann, 1972; Kogure and Onoda, 1983), and orbitofrontal cortex (Tanabe et al., 1975; Yarita et al., 1980; Motokizawa and Ino, 1983; Onoda et al., 1984; Schoenbaum and Eichenbaum, 1995) via the thalamus (Imamura et al., 1984). Single neurons in the orbitofrontal cortex responded to taste and odor stimuli in the monkey (Rolls, 1989). These results suggest that the orbitofrontal cortex is the highest olfactory center, where the olfactory and gustatory information might be integrated. However, taste-sensitive neurons in the orbitofrontal cortex have not been observed in other mammal species. Anterograde tracer injection into the EPN demonstrated labeled axons in the DI and AI (Behan and Haberly, 1999). Further, voltage imaging in slices has also revealed that paroxysmal epileptiform activity can develop in layer VI of the AI in synchrony with that in the EPN (Demir et al., 1998). Thus, the EPN and AI are predicted as the regions where olfactory and gustatory information are centrally integrated. In this study, we present mapping data obtained from electrical and optical recordings to identify these integrative regions in the CNS.
a Department of Physiology, Kanazawa Medical University, Uchinada, Ishikawa 920-0293, Japan b Department of Oral and Maxillofacial Surgery, Kanazawa Medical University, Uchinada, Ishikawa 920-0293, Japan
Abstract—Electrical and optical recordings were made from slice preparations including the piriform and gustatory cortices. Electrical stimulation of the gustatory cortex evoked a characteristic field potential in the endopiriform nucleus. A field potential was induced in the endopiriform nucleus by stimulation of the piriform cortex. Voltage-sensitive dye studies showed that stimulation of the piriform cortex induced signal propagation from the piriform cortex to endopiriform nucleus, whereas stimulation of the gustatory cortex did the same from the gustatory cortex to endopiriform nucleus via the agranular division of the insular cortex. After stimulation of the endopiriform nucleus, optical signals propagated not only to the piriform cortex but also to the gustatory cortex via the agranular division of the insular cortex. The olfactory and gustatory pathways appeared to be reciprocally connected. Unit recordings indicated that olfactory and gustatory activity converged onto a single neuron of the endopiriform nucleus. It is suggested that the cortical integration of olfactory and gustatory information could modulate mechanisms involved in food selection and emotional reactions relating to the chemical senses. © 2004 IBRO. Published by Elsevier Ltd. All rights reserved. Key words: field potential, optical imaging, voltage-sensitive dye, signal propagation, insular cortex.
The primary gustatory cortex (GC) is defined as the cortical region that receives direct fiber projection from the parvicellular part of the ventral posteromedial nucleus in the thalamus, and the terminal labeling area is situated rostrally in the granular division (GI) and caudally in the dysgranular division (DI) of the insular cortex (Nakashima et al., 2000). In fact, unit recordings found gustatory responses of neurons from both the GI and DI (Yamamoto et al., 1985; Ogawa et al., 1992), suggesting the location of the GC in both divisions. The ventral part (Vp) of the GI *Corresponding author. Tel: ⫹81-76-218-8102; fax: ⫹81-76-286-3523. E-mail address: [email protected] (N. Onoda). Abbreviations: ACSF, artificial cerebrospinal fluid; AI, agranular division of the insular cortex; CGRP, calcitonin gene-related peptide; DI, dysgranular division of the insular cortex; EPN, endopiriform nucleus; EPSC, excitatory postsynaptic current; EPSP, excitatory postsynaptic potential; GABAB-R, B-type of receptor of GABA; GC, gustatory cortex; GI, granular division of the insular cortex; IC, insular cortex; MCA, middle cerebral artery; MD, mediodorsal thalamic nucleus; OB, olfactory bulb; PC, piriform cortex; PDH, pulse density histogram; Vp, ventral part.
0306-4522/04$30.00⫹0.00 © 2004 IBRO. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.neuroscience.2004.03.041
1033
1034
W. Fu et al. / Neuroscience 126 (2004) 1033–1041
B
L V
ACAd
A
D M
RhS ec
PC
OB
EPN
ac
LOT
5 3 2 1
MCA
C
R1
GI DI
6
1
1
2 1
D
S1
AI
3
LOT
RhS
DI
2 3
2 3
LOT
6
R4
IV 4
S4 ec L
D
EPN V
M
R2
S2
RhS
PC
AI
PC
5
GI DI AI
PC I II III EPN
1
3
500 µm
2
Fig. 1. Slice preparations and their morphological features. (A) A lateral view of the rat brain. Slice preparations were cut parallel to the MCA as indicated by a dotted line. (B) A frontal section with a cutting angle (approximately 10° off the frontal plane). A ventro-lateral region surrounded by thick dotted lines is expanded in C. (C) A schematic drawing of experimental arrangements. Stimulation site 1 (S1) and recording site 1 (R1) are in layer II of the PC. S2 and R2 are in the EPN. S4 and R4 are in the GC. (D) A micrograph of the frontal section stained with Neutral Red corresponds to the region surrounded by thick dotted lines. This region was a histologically determined camera field in the optical recordings in Fig. 3B and C. This micrograph was reconstructed from resliced sections after dye was deposited in recording and stimulation sites. Spot 1 is in layer II of the PC. Spot 2 is in the EPN. Spot 3 is in layer V/VI of the AI of the IC. Spot 4 is in the DI of the IC, the GC. Closed arrowheads indicate boundaries between divisions of IC. An open arrowhead indicates layer IV of GI. ac, anterior commissure; D, dorsal direction; ec, external capsule; L, lateral direction; I, layer I; II, layer II; III, layer III; IV, layer IV; LOT, lateral olfactory tract; M, medial direction; RhS, rhinal sulcus; V, ventral direction.
EXPERIMENTAL PROCEDURES Animal protocols used in this study complied with all pertinent institutional and Japanese Government regulations, and every attempt was made to minimize the number of killed animals and their suffering.
Materials Young Wistar rats (n⫽25, 100 –150 g) were used. To prepare slices (340 – 400 m thick) including the piriform and gustatory cortices, slices were cut in a plane approximately 10° off the frontal (near parallel to the MCA), from 2.0 mm rostral to the anterior edge of the optic chiasm, or from 0.8 mm rostral to the MCA to 1.2 mm caudal to the MCA (Fig. 1A and B). The slices were perfused continuously (⬎2 h at 30 °C) with oxygenated (95% O2/5% CO2) artificial cerebrospinal fluid (ACSF) containing (in mM): 124 NaCl, 5 KCl, 1.24 KH2PO4, 1.3 MgSO4, 2.4 CaCl2, 26 NaHCO3, and 10 glucose.
Preparation for electrophysiological recording After recovery, a slice was transferred to the recording chamber mounted on the stage of an inverted microscope (IMT-2; Olympus, Japan) and perfused with oxygenated ACSF. Bath temperature was maintained at 26 –28 °C. Mono-tungsten stimulating electrodes were inserted into layer I/II of the PC, layer II/III (corresponding to spot 4 in Fig. 1D) of the GC and/or EPN,
and single square pulses (0.02–1 mA, 80 s) were delivered at 0.05– 0.07 Hz. The stimulus intensity was set to about 5– 8⫻ the threshold, by which the optical signal reached its maximum spatial extent. Glass microelectrodes containing 2% Brilliant Blue in 0.85% NaCl (4 –30 M⍀, DC) were used to record extracellular field potentials, record unit activity, and mark the location of the electrode tip with dye deposit (10 A for 3–5 min). Four to eight sweeps of field potentials were processed for averaging. Patch electrodes for current clamp recordings from EPN neurons contained (in mM): 126 K-gluconate, 4 KCl, 1 CaCl2, 1 MgCl2, 10 HEPES, 10 EGTA, 2 Na2-ATP, 0.3 Na2-GTP (pH⫽7.3 with KOH). For voltage clamp recordings, internal solution was identical except that 126 CsMeSO4, 4 CsCl were substituted for 126 K-gluconate, 4 KCl, respectively. The pH was adjusted to 7.3 with CsOH. The osmolarity of these pipette solutions was 295–299 mOsm. These electrophysiological data, including field and unit responses, were stored and analyzed with the use of pClamp software (Axon Instruments, Foster City, CA, USA).
Preparation for optical recording After recovery, a slice was incubated with a voltage-sensitive dye NK2761 (0.13 mg/ml; Nippon Kankoh-Shikiso Kenkyusho, Okayama, Japan) for 18 min, transferred to the recording chamber, and perfused with ACSF (26 –28 °C). Mono-stimulating electrodes were inserted into the layer I/II of the PC, GC and/or EPN.
W. Fu et al. / Neuroscience 126 (2004) 1033–1041
1035
Fig. 2. Mapping data of field potential responses in three recording sites (R1, R2, and R4) following stimulation of three different sites (S1, S2, and S4) indicated in Fig. 1C. (A) A time-lapse of field potentials in the EPN following stimulation of PC neurons in normal solution. (B) Field potentials were averaged from eight sweeps. PC-evoked EPN responses in normal (a) and Mg2⫹ free solution (b). GC-evoked EPN responses in normal (d) and Mg2⫹ free solution (e). EPN-evoked PC (c) and GC responses (f) in Mg2⫹ free solution. Closed arrowheads under traces indicate the onset of stimulation.
The stimulating parameters were the same as above. Glass microelectrodes containing blue dye were used to monitor field potentials and to deposit the dye.
Optical recording The camera unit of the optical imaging system (Fujix HR-Deltaron 1700; Fujifilm Microdevices, Tokyo) contains a 128⫻128 photodiode array. With the 4⫻ objective, the whole array corresponded to a 2.26⫻2.26 mm2 area of tissue. The light passed through a narrow band interference filter (700⫾30 nm). Sixteen responses to single shocks were averaged to form a run. Neural activity was recorded optically by monitoring absorption changes in transmitted light intensity in each element at a rate of one frame per 0.6 ms. The details have been described elsewhere (Ichikawa et al., 1993; Sugitani et al., 1994; Tanifuji et al., 1994; Obaid and Salzberg, 1996; Sugai et al., 1997).
Histological reconstruction and immunohistochemistry After electrical or optical recordings, appropriate sites in the pathway of signal propagation following stimulation of the PC/GC and/or EPN were confirmed on a display screen, and then each site was marked. A tip of a glass pipette containing dye was superimposed on the mark on the screen after the display was replaced with the real image of the slice preparation, and then the dye was deposited and the site was penetrated by the stimulating electrode. The slice was fixed with 1% paraformaldehyde in 0.1 M phosphate buffer and resectioned (40 m). Sections were processed by a immunohistochemical technique. After elimination of endogenous peroxidase activity with 0.1% H2O2, alternative sections were incubated with an antibody (Chemicon, Temecula, CA, USA) against a B-type receptor of GABA (GABAB-R) at a dilution of 1:10,000 or with an antibody (Biogenesis, Kingston, NH, USA) against calcitonin gene-related peptide (CGRP) at a dilution of 1:4000, and then incubated in avidin– biotin–peroxidase complex solution (Vector, Burlingame, CA, USA). All sections were photographed. The location of the camera field in the slice was also histologically reconstructed.
RESULTS
Fig. 1C, where the ventral region of the slice surrounded by thick dotted lines in Fig. 1B was expanded. Fig. 1D shows an example of a frontal section on which blue dye was deposited. Spot 1 is in layer II of the PC and spot 2 is in the EPN. Mapping studies of field potentials Electrical stimulation of layer II in the PC evoked a field potential with a latency of 35 ms in the central region of the EPN (Fig. 2A). The PC-evoked field responses occurred in initial stimulations, but their latencies became longer gradually (oblique dotted line) in ACSF containing Mg2⫹ (normal solution) and often disappeared after multiple stimulation (Fig. 2A and “a” in Fig. 2B). Disappearance of the PC-evoked field potentials was blocked in ACSF without Mg2⫹ (“b” in Fig. 2B). A field potential with a latency of 66 ms was elicited by a single shock of layer II/III of the GC in the EPN in normal solution (“d” in Fig. 2B). The GC-evoked field responses occurred in initial stimulations, but their latencies became longer gradually in normal solution and then the field responses often disappeared after multiple stimulation (data not shown). Disappearance of the GCevoked field potentials was also blocked in Mg2⫹ free solution (“e” in Fig. 2B). When the stimulation sites in the GC and PC were moved from layer to layer, the onset of the field potentials (latent periods) differed somewhat, but the waveform of individual field potentials was identical. These electrophysiological results suggested that excitation traveled a long distance, and that both the PC and GC are connected with the EPN. Next, electrical stimulation of the EPN provoked characteristic field potentials with their latency of 24 and 114 ms in the PC (“c” in Fig. 2B) and GC (“f” in Fig. 2B), respectively, in Mg2⫹ free solution. These results suggested that the PC and GC are connected reciprocally with the EPN.
Slice preparations and their morphological features
Results from optical imaging studies of signal propagation
Fig. 1B shows the morphological feature of a slice. The experimental arrangement was schematically drawn on
Optical recordings were carried out on brain slices in normal or Mg2⫹ free solution. Field potentials and optical
1036
W. Fu et al. / Neuroscience 126 (2004) 1033–1041
A
RhS
GC
PC
S1
4.8 ms
10.8
16.8 ms
28.8
13.8 ms
16.8
22.8
28.8
142.8
31.8
46.8
58.8
91.8
43.8
57.0
67.8
79.8
109.8
7.8 ms
35.4
43.8
61.8
75.0
142.2
28.8 ms
79.8
127.8
145.8
157.8
193.8
II
R2 EPN
ec
S4
B
PC
GC R2 II AI
EPN
ec
S4
C
PC
GC
II
AI
EPN
ec
R2 PC
D
II
GC
S2
E
EPN
R4
R1 ec
GC AI
ec
PC II
0.2 mV
S2 EPN
Stim 0%
0.075 % (A,B,C) 0.04 % (D, E)
50 ms (A, B, C, D) 100 ms (E)
Fig. 3. A time-lapse optical image of neural excitation in tissue evoked by PC (A), GC (B, C) and EPN stimulation (D, E). (A, C–E) In Mg2⫹ free solution. (B) In normal solution. Numerals on each image are lapse-periods after stimulation. White arrows and arrowheads in images indicate the EPN and layer VI of AI, respectively. The former corresponds to spot 2 in Fig. 1D and the latter corresponds to spot 3. Schematic drawings of individual camera fields are given on the far left. Closed arrowheads indicate individual stimulation sites (S1, S2, and S4). Open arrowheads indicate individual recording sites (R1, R2, and R4) of field potentials on the far right. The left and top are the dorsal and lateral directions, respectively. Colors below E indicate changes in light absorption (in percent). A thin arrow under the lowest trace on the far right marks the onset of stimulation. Abbreviations are the same as in Fig. 1.
signals were recorded concurrently (far right of Fig. 3). PC stimulation (S1) provoked an early optical response, which traveled to other stimulation sites in the PC (image at 4.8 ms) in Mg2⫹ free solution (Fig. 3A). Subsequently, the optical signal was found to propagate 1) horizontally along layer II and vertically toward the deep layers in the PC (image at 10.8 ms after S1), 2) vertically toward the EPN and horizontally along layer III (16.8 ms), and 3) horizontally along the EPN (22.8 ms). The distance that electrically-induced optical signals in the PC and EPN (white arrow) traveled was maximum by 28.8 ms; these signals faded gradually by 142.8 ms, and almost disappeared by 300 ms. The pattern of signal propagation and propagation velocity (51.3⫾10.5 mm/s [mean⫾S.E., n⫽6]) were similar in six slices. The GC shock (S4) provoked an early optical response around stimulation sites in four slices in normal solution (Fig. 3B). The optical signal propagated 1) vertically toward deep layers in the GC (16.8 ms after S4), 2) horizontally within the deep layer and obliquely
toward deep layers in the AI (white arrowhead in 28.8 ms), 3) obliquely toward the EPN (31.8 ms), and finally 4) horizontally along the EPN (white arrow in 46.8 ms). The distance that the optical signals in the EPN (white arrow) traveled was maximum by 58.8 ms; signals in the GC and AI weakened by 91.8 ms and almost disappeared by 120 ms. The pattern of signal propagation and propagation velocity (42.5⫾7.6 mm/s [n⫽4]) were similar in all slices. The GC shock (S4) produced stronger and longer optical responses in 10 slices in Mg2⫹ free solution than in normal solution (Fig. 3C). In particular, optical responses in the deep layers of the AI (white arrowhead in 43.8 ms) and EPN (white arrows in 67.8 and 79.8 ms) were markedly intense. The signal propagation velocity was 33.9⫾ 3.0 mm/s (n⫽10). These results suggested that the PC and GC were connected with the EPN. It appeared that the GC-EPN propagation passed through the AI. EPN stimulation (S2) provoked early optical response (7.8 ms) in four slices in Mg2⫹ free solution (Fig. 3D). Then
W. Fu et al. / Neuroscience 126 (2004) 1033–1041
Dp
AI
Dp
Vp
AI
B
A
GI
Vp
1037
I
C
II
RhS
III
IV 3
AI Dp
DI
4
V VI
500 µm
3
Fig. 4. Immunohistochemical results and histological reconstruction. (A) A microphotograph of a resectioned slice incubated with an antibody against CGRP. Note that CGRP-positive fibers (thick arrows) were dominant in the Vp of the AI. Spot 3 (thin arrows) indicated by white arrowheads in Fig. 3B, C and E was found in the AI. (B) A microphotograph of an adjacent section incubated with an antibody against B-type of receptor of GABAB-R. Spot 3 (thin arrows) was found under the layer containing GABAB-R-positive cells (thick arrows). (C) A microphotograph expanded from the left part of Fig. 1D. A distinct layer of cells in between the thick arrows is the GI, suggesting layer IV. Spot 4 (thin arrow) was located in the region without layer IV and just dorsal to the AI, suggesting layer III of the DI, i.e. the GC. Abbreviations are the same as in Fig. 1. Dp, dorsal part of the AI.
the optical signal propagated 1) horizontally along the EPN (white arrow) and vertically toward layer III in the PC (35.4 ms), 2) vertically toward layers II and III in the PC (43.8 ms), and 3) horizontally along layers II and III in the PC (61.8 ms). The distance that the optical signal in the PC traveled was maximum by 75.0 ms. Optical signals faded gradually (142.2 ms) and almost disappeared at 300 ms after S2. The signal propagation and propagation velocity (40.4⫾5.9 mm/s [n⫽4]) were similar in all slices. After the early optical response to S2 in five slices in Mg2⫹ free solution (Fig. 3E), the optical signal traveled to the GC 1) horizontally toward a deep layer in the AI (white arrowhead in 28.8 ms), 2) vertically toward the surface of the AI (127.8 ms), and 3) horizontally along superficial layers in the GC. The distance that the optical signal traveled in the GC was maximum by 157.8 ms and the signal disappeared 300 ms after S2. The pattern of signal propagation and propagation velocity (19.7⫾2.4 mm/s; n⫽5) was similar in all slices. These results suggested that the PC and GC were connected reciprocally with the EPN. It appeared, thus, that both GC-EPN and EPN-GC propagations passed through the AI. Immunohistochemical results and histological reconstruction After the signal patterns in slices were optically recorded (Fig. 3B, C and E), blue dye was deposited into regions indicated by white arrowheads. CGRP-positive fibers (thick arrows) were observed in the AI (Fig. 4A). Spot 3 was found in the region containing CGRP-positive fibers, which were mainly in the Vp of the AI (Yasui et al., 1989). GABAB-R-positive cells were observed in an adjacent section (Fig. 4B). Spot 3 was found under the layer containing GABAB-R-positive cells, which were located in layer V of
the AI (Jasmin et al., 2003). Thus, spot 3 appeared to be in layer VI of the AI. The left part of Fig. 1D is magnified in Fig. 4C. A distinct layer of cells in between thick arrows can be seen in the GI, suggesting layer IV. Spot 4 was located in the region without layer IV and just dorsal to the AI. Thus, spot four appeared to be in layer III of the DI, which is the GC. A boundary between the AI and DI was determined histochemically by presence or absence of CGRPlabeled fibers. Unit recording study Activity of the EPN units in Mg2⫹ free solution is demonstrated in Fig. 5A and B. PC stimulation provoked a long burst of spikes (“a” in Fig. 5A). A large decrease in amplitude of subsequent spikes was observed, and spike amplitude recovered gradually. Bursting activity can be seen clearly in a pulse density histogram (PDH) and initial spikes appear at 25–30 ms after every PC shock (“d” in Fig. 5A). The same EPN unit responded to GC shocks (“b” in Fig. 5A). Subsequent to an initial spike, a large decrease in amplitude of spikes was also observed. The latency of the initial spike was 38 – 43 ms and the delay of the latency after GC shocks can be seen clearly in the PDH (“e” in Fig. 5A). Unit responses to PC and GC shocks showed similar long bursting patterns. As stimulation of either the PC or GC elicited a long burst of discharges, effects of simultaneous stimulation were less clear (“c” and “f” in Fig. 5A). A burst of spikes was recorded from another EPN unit after PC stimulation (“a” in Fig. 5B). This unit discharged bursting spikes in response to GC shocks (“b” of Fig. 5B). These discharge patterns were similar although this unit fired initially at 28 and 70 ms after PC and GC shocks, respectively. Further, the bursting frequency in Fig. 5B was lower than that shown in Fig. 5A. Fluctuation of the latency of
W. Fu et al. / Neuroscience 126 (2004) 1033–1041
A No. of spikes/bin
a
b
c
C
200 ms
12 10 8 6 4 2 0 0 12 10 8 6 4 2 0 0 12 10 8 6 4 2 0 0 10 8 6 4 2 0
No. of spikes/bin
a
b 200 ms
10 8 6 4 2 0
B
d 20
40
60
80 100
20
40
60
80 100
20
40
60
80 100
e
f
a
Ti me (ms)
c 0 20 40 60 80 100 120140 160
d
No. of spikes/bin
1038
b c
D
200 ms
12 10 8 6 4 2 0
0 10 8 6 4 2 0 12 0 10 8 6 4 2 0 0
d e
20
40
60
80 100
20
40
60
80 100
40
60
80 100
f 20
Ti me (m s)
a b c
50 mV 500 ms
0 20 40 60 80 100 120 140 160
Ti m e (ms)
Fig. 5. Mapping data of unit activity and whole cell patch recordings. (A) Unit activity in the EPN following stimulation of the PC and GC. Bursting spike responses of an EPN neuron to a single PC (a) or GC shock (b). Spike response of the same unit from simultaneous stimulation of the PC and GC (c). PDHs processed from 16 sweeps of unit responses to PC (d), GC (e), and both PC and GC stimulation (f). (B) Spike responses of another EPN unit to a single PC (a) or GC shock (b). PDHs processed from 16 sweeps of PC- (d) and GC-evoked responses (e). Spike response (c) and the PDH (f) from simultaneous stimulation of the PC and GC. (C) Superimposed responses in PC (a) and GC units (b) from eight sweeps following stimulation of the EPN. PDHs processed from 16 sweeps of EPN-evoked responses in PC (c) and GC (d) units. (D) Whole cell patch recording made in a current clamp mode from EPN neurons in Mg2⫹ free or 0.5 mM Mg2⫹ solution. Long-duration depolarizing potentials with superimposed action potential responses in an EPN neuron to a single PC shock (a). A GC-evoked long EPSP with superimposed action potentials from two other EPN neurons (b) and (c). Arrows under traces and PDHs indicate the onset of stimulation.
initial spikes and the difference between latency after PC shock and latency after GC shock can be seen clearly in PDHs (“d” and “e” in Fig. 5B). Simultaneous stimulation of the PC and GC produced a similar response as shown in “a” and “d” of Fig. 5B, but GC-evoked response was not clear (“c” and “f” in Fig. 5B). Stimulation of EPN evoked spike discharges in layer II of the PC (“a” in Fig. 5C) and layer II/III of the GC (“b” in Fig. 5C). Large increases in the impulse rate and the difference of their latencies of initial spikes in the PC and GC can be seen clearly in PDHs (“c” and “d” in Fig. 5C). The average latency of initial spikes was 59.7⫾4.2 ms (mean⫾S.E., n⫽19) for PC neurons and 96.9⫾4.9 ms (n⫽28) for GC neurons. It appeared that discharge patterns of EPN neurons differed largely from those of PC and GC neurons. Of the 30 EPN units we recorded, 25 units responded to both PC and GC shocks. The discharge pattern in
response to PC and GC shocks showed a similar burst in single EPN units, although latencies of individual responses were different. These results suggested that olfactory and gustatory connections converged onto a single EPN neuron. Further, EPN shocks evoked neural activity in the PC and GC. Out of 24, 19 PC units responded to EPN shocks, whereas out of 31, 28 GC units did. Thus, PC-EPN and GC-EPN connections appeared to be reciprocal. Patch clamp recording study Finally, to confirm results obtained from unit responses to PC and GC shocks, electrical recordings were made from EPN neurons in the whole-cell configuration in Mg2⫹ free or 0.5 mM Mg2⫹ solution using the patch clamp technique. The EPN neurons were identified by their size and location
W. Fu et al. / Neuroscience 126 (2004) 1033–1041
in the EPN. PC shocks evoked a long lasting depolarizing potential (⬎1.5 s) with a superimposed burst of action potentials with low amplitude in a current clamp mode (“a” in Fig. 5D). Immediately after an initial action potential, the amplitude of subsequent action potentials was small, but the decrease in amplitude recovered gradually. This depolarizing potential was the longest lasting (half-amplitude duration, 980 ms) among EPN neurons responding to PC shocks. The average reversal potential of the PC-evoked long-lasting depolarizing inward currents was 0.9⫾1.4 mV (mean⫾S.E.) in four other EPN neurons in a voltage clamp mode, suggesting excitatory postsynaptic currents (EPSCs). The average duration of the half amplitude of PCevoked excitatory postsynaptic potentials (EPSPs) was 633⫾136 ms (mean⫾S.E., n⫽10). Next, GC shocks evoked a burst of high frequency action potentials that became superimposed on the ongoing long-duration depolarizing potential (⬎2.5 s) in another EPN neuron in the current clamp mode (“b” in Fig. 5D). Amplitude of the action potentials gradually recovered. This depolarizing potential of this EPN neuron was the longest lasting (half amplitude duration, 1390 ms) of the potentials in all EPN neurons recorded. A GC-evoked depolarizing potential (⬍1 s) was recorded from another EPN neuron (“c” in Fig. 5D) and its half amplitude duration was 325 ms. GCevoked EPSCs were reversed at 1.2⫾1.3 mV (mean⫾S.E., n⫽5), whereas mean half amplitude duration of GC-evoked EPSPs was 444⫾88 ms (mean⫾S.E., n⫽14). EPN neurons fired with low or high frequency following stimulation of the GC, because duration of GCevoked EPSPs varied. These results of extracellular recordings suggested that the long lasting EPSPs were responsible for a large decrease and subsequent gradual increase in unit but not multiunit amplitude following stimulation of the PC or GC.
DISCUSSION In this study, frontal slices were cut as parallel as possible to the MCA and the cutting angles were determined by results obtained from electrical recordings. Further, since the bed nucleus of the anterior commissure, which is used as a landmark in the stereotaxic rat brain (Paxinos and Watson, 1986), was involved in these slices, it was ready to calculate the positioning of the slice structures from the stereotaxic reference points. In some slices, PC but not GC shocks evoked the EPN response, while in others the converse was true. Fiber connections were probably severed. When slices were in ACSF containing Mg2⫹ (normal solution), the field potential in response to either PC or GC shocks disappeared for a short while after stimulation. The disappearance of the responses was blocked by Mg2⫹ free ACSF. Under either Mg2⫹ containing or Mg2⫹ free conditions in this study, following stimulation of layer I/II in the PC or layer II/III in the GC, characteristic field potentials were recorded in the EPN. Effects of Mg2⫹ free solution on field potentials have been investigated in PC slices of guinea-pigs (Richards and Sercombe, 1970) and rats (Hoffman and Haberly, 1989, 1991, 1993). Under Mg2⫹
1039
free conditions, the surface potentials became hyperexcitable (Richards and Sercombe, 1970) and stimulation of association fibers elicited epileptiform responses in the EPN (Hoffman and Haberly, 1989, 1991, 1993). Intracellular recording studies demonstrated that burst response evoked by stimulation of fiber tracts in PC slices in Mg2⫹ free solution consisted of high amplitude, long-duration depolarizing potentials with superimposed action potentials, suggesting an interaction of cells via activated NMDA glutamate receptors (Hoffman and Haberly, 1989, 1991, 1993). Our patch clamp recordings from EPN neurons in Mg2⫹ free solution demonstrated that PC and/or GC shocks evoked long-lasting EPSPs with superimposed action potentials. These results suggest that a reason the field response disappeared in normal solution might be the lack of activation of NMDA glutamate receptors. Optical imaging in the PC region of guinea-pig sagittal slices has demonstrated that optical signals elicited by afferent shocks propagated along layers II and III, then into the EPN, where neural activity moved slowly in the caudal direction along the outer edge of the external capsule (Sugitani et al., 1994). Epileptiform activity evoked by stimulation of association fibers in Mg2⫹ free medium was recorded from the EPN in transverse and longitudinal slices (Demir et al., 1998, 2001). Present optical results in the PC region of rat frontal slices are consistent with previous optical results showing olfactory excitation reached the EPN after stimulation of fiber tracts. A tracer injection study into the olfactory bulb (OB) demonstrated that OB of the mouse had a direct terminal projection to a sector of the IC, and this part of the IC had inputs from the major ascending gustatory and visceral afferent relay structures (Shipley and Geinisman, 1984). However, a direct OB–IC projection has not been found in other mammal species. An electrophysiological study demonstrated that field potentials and unit responses were generated in the prefrontal cortex following electrical stimulation of the OB, PC, mediodorsal thalamic nucleus (MD), and amygdala (Cinelli et al., 1987), suggesting the OB–IC projection is via several synaptic relays. In all species for which data are available, at least two additional routes from OB to IC are suggested: (1) one way is from OB to PC and from PC to IC (Motokizawa and Ino, 1983); (2) the second route is from PC by way of the EPN to the MD and from the MD to the IC (Gerfen and Clavier, 1979; Price, 1987). From a study of anterograde tracer injection in the EPN, labeled axons were found not only in the MD but also in both the DI and AI, and only a few in the GI, and label was present in all layers but concentrated in layer I (Behan and Haberly, 1999). The PC–IC projection by way of the EPN is considered as another route accordingly. In fact, typical extracellular responses associated with abruptonset epileptiform-EPSP, which originated in the EPN, were observed in IC neurons in Mg2⫹ free medium (Hoffman and Haberly, 1993). Voltage imaging in slices has revealed that paroxysmal epileptiform activity can develop in layer VI of the AI in synchrony with that in EPN in low-chloride medium (Demir et al., 1998). Other optical data also identified EPN-AI propagation (Demir et al.,
1040
W. Fu et al. / Neuroscience 126 (2004) 1033–1041
2000). Our optical results demonstrated EPN–GC propagation of signals via the AI and vice versa, although polysynaptic connections were responsible for slow propagation speed of optical signals. The remarkably slow propagation times may be due to bath temperature, smaller diameter axons in the EPN (Behan and Haberly, 1999), inhibitory synaptic action, or wave-like propagation. The AI is associated with visceral sensory and autonomic functions rather than taste (Cechetto and Saper, 1987; Krushel and van der Kooy, 1988; Cechetto and Chen, 1990; Allen et al., 1991; Saper, 1995). An immunohistochemical study revealed that the AI and DI were innervated by CGRP-immunoreactive fibers and the density of CGRPpositive innervation was considerably richer in the ventral AI (Yasui et al., 1989). Our results indicated a blue spot formed over the region where CGRP-positive fibers were abundant (Fig. 4A). Together with our optical results, this blue spot appeared to be evidence of signal propagation to the GC from the EPN via the AI and vice versa. A recent immunohistochemical study has shown, further, that GABAB-Rs are concentrated on pyramidal neurons of layer V in the AI, and GABAB neuron activity could change the pain threshold and then might cause hyperalgesia (Jasmin et al., 2003). In this study, spot 3 was observed under a GABAB-R-positive layer (Fig. 4B). It appeared that spot 3 positioned in layer VI of the AI. From results obtained from unit recordings, furthermore, olfactory and gustatory activity converged onto a single EPN neuron, although the long-duration EPSPs evoked by either stimulus were responsible for the reduced effects of simultaneous stimulation of PC and GC on EPN neurons. The EPN–PC and EPN–GC connections were also observed. Our results led us to conclude reciprocal connections between olfactory and gustatory pathways. Together with the descending connections from the EPN and AI to the amygdala (Price, 1987; Haberly, 1998; Shi and Cassell, 1998; Price, 1995; Nakashima et al., 2000), it is possible that the cortical integration of olfactory, gustatory, visceral and nociceptive information could modulate mechanisms involved in food selection and emotional reactions relating to the chemical and pain senses. Acknowledgements—The authors wish to thank Mr. Adachi and Mr. Muramoto for technical assistance. This work was partly supported by a Grant for Collaborative Research, C2001-01 from Kanazawa Med. Univ.
REFERENCES Allen GV, Saper CB, Hurley KM, Cechetto DF (1991) Organization of visceral and limbic connections in the insular cortex of the rat. J Comp Neurol 311:1–16. Behan M, Haberly LB (1999) Intrinsic and efferent connections of the endopiriform nucleus in rat. J Comp Neurol 408:532–548. Cechetto DF, Chen SJ (1990) Subcortical sites mediating sympathetic responses from insular cortex in rats. Am J Physiol 258:R245– R255. Cechetto DF, Saper CB (1987) Evidence for a viscerotopic sensory representation in the cortex and thalamus in the rat. J Comp Neurol 262:27–45. Cinelli AR, Ferreyra-Moyano H, Barragan E (1987) Reciprocal func-
tional connections of the olfactory bulbs and other olfactory related areas with the prefrontal cortex. Brain Res Bull 19:651–661. Demir R, Haberly LB, Jackson MB (1998) Voltage imaging of epileptiform activity in slices from rat piriform cortex: onset and propagation. J Neurophysiol 80:2727–2742. Demir R, Haberly LB, Jackson MB (2000) Characteristics of plateau activity during the latent period prior to epileptiform discharges in slices from rat piriform cortex. J Neurophysiol 83:1088 –1098. Demir R, Haberly LB, Jackson MB (2001) Epileptiform discharges with in-vivo-like features in slices of rat piriform cortex with longitudinal association fibers. J Neurophysiol 86:2445–2460. Gerfen CR, Clavier RM (1979) Neural inputs to the prefrontal agranular insular cortex in the rat: horseradish peroxidase study. Brain Res Bull 4:347–353. Haberly LB (1969) Single unit responses to odor in the prepiriform cortex. Brain Res 12:481–484. Haberly LB (1998) Olfactory cortex. In: The synaptic organization of the brain (Shepherd GM, ed), pp 377–416. Oxford: Oxford University Press. Hoffman WH, Haberly LB (1989) Bursting induces persistent all-ornone EPSPs by an NMDA dependent process in piriform cortex. J Neurosci 9:206 –215. Hoffman WH, Haberly LB (1991) Bursting-induced epileptiform EPSPs in slices of piriform cortex are generated by deep cells. J Neurosci 11:2021–2031. Hoffman WH, Haberly LB (1993) Role of synaptic excitation in the generation of bursting-induced epileptiform potentials in the endopiriform nucleus and piriform cortex. J Neurophysiol 70:2550 – 2561. Ichikawa M, Iijima T, Matsumoto G (1993) Real-time optical recording of neuronal activities in the brain. In: Brain mechanisms of perception and memory (Ono T, Squire LR, Raichle ME, Perrett DI, Fukuda M, eds), pp 638 –648. New York: Oxford University Press. Imamura K, Onoda N, Takagi SF (1984) Odor response characteristics of thalamic mediodorsal nucleus neurons in the rabbit. Jpn J Physiol 34:55–73. Jasmin L, Rabkin SD, Granato A, Boudah A, Ohara PT (2003) Analgesia and hyperalgesia from GABA-mediated modulation of the cerebral cortex. Nature 424:316 –320. Kogure S, Onoda N (1983) Response characteristics of lateral hypothalamic neurons to odors in unanesthetized rabbits. J Neurophysiol 50:609 –617. Kosar E, Grill HJ, Norgren R (1986) Gustatory cortex in the rat. I: Physiological properties and cytoarchitecture. Brain Res 379:329 – 341. Krushel LA, van der Kooy D (1988) Visceral cortex: integration of the mucosal senses with limbic information in the rat agranular insular cortex. J Comp Neurol 270:39 –54. Motokizawa F, Ino Y (1983) A non-thalamic olfactory pathway to the orbital gyrus in the cat. Brain Res Bull 10:83–88. Nakashima M, Uemura M, Yasui K, Ozaki HS, Tabata S, Taen A (2000) An anterograde and retrograde tract-tracing study on the projections from the thalamic gustatory area in the rat: distribution of neurons projecting to the insular cortex and amygdaloid complex. Neurosci Res 36:297–309. Nemitz JW, Goldberg SJ (1983) Neuronal responses of rat pyriform cortex to odor stimulation: an extracellular and intracellular study. J Neurophysiol 49:188 –203. Norgren R (1995) Gustatory cortex. In: The rat nervous system (Paxinos G, ed), pp 751–771. Australia: Academic Press. Obaid AL, Salzberg BM (1996) Micromolar 4-aminopyridine enhances invasion of a vertebrate neurosecretory terminal arborization: optical recording of action potential propagation using an ultrafast photodiode-MOSFET camera and a photodiode array. J Gen Physiol 107:353–368. Ogawa H, Hasegawa K, Murayama N (1992) Difference in taste quality coding between two cortical taste areas, granular and dysgranular insular areas in rats. Exp Brain Res 91:415–424.
W. Fu et al. / Neuroscience 126 (2004) 1033–1041 Onoda N, Imamura K, Obata E, Iino M (1984) Response selectivity of neocortical neurons to specific odors in the rabbit. J Neurophysiol 52:638 –652. Paxinos G, Watson C (1986) The rat brain in stereotaxic coordinates. San Diego: Academic Press. Price JL (1987) The central olfactory and accessory olfactory systems. In: Neurobiology of taste and smell (Finger TE, Silver WL, eds), pp 179 –203. New York: John Wiley and Sons. Price JL (1995) Thalamus. In: The rat nervous system (Paxinos G, ed), pp 629 –648. Australia: Academic Press. Richards CD, Sercombe R (1970) Calcium, magnesium and the electrical activity of guinea-pig olfactory cortex in vitro. J Physiol 211: 571–584. Rolls ET (1989) Information processing in the taste system of primates. J Exp Biol 146:141–164. Saper CB (1995) Central autonomic system. In: The rat nervous system (Paxinos G, ed), pp 107–135. Australia: Academic Press. Schoenbaum G, Eichenbaum H (1995) Information coding in the rodent prefrontal cortex. I: Single-neuron activity in orbitofrontal cortex compared with that in pyriform cortex. J Neurophysiol 74:733– 750. Scott JW, Pfaffmann C (1972) Characteristics of responses of lateral hypothalamic neurons to stimulation of the olfactory system. Brain Res 48:251–264. Shi CJ, Cassell MD (1998) Cortical, thalamic, and amygdaloid connections of the anterior and posterior insular cortices. J Comp Neurol 399:440 –468. Shipley MT, Geinisman Y (1984) Anatomical evidence for conver-
1041
gence of olfactory, gustatory, and visceral afferent pathways in mouse cerebral cortex. Brain Res Bull 12:221–226. Sugai T, Sugitani M, Onoda N (1997) Subdivisions of the guinea-pig accessory olfactory bulb revealed by the combined method with immunohistochemistry, electrophysiological and optical recordings. Neuroscience 79:871–885. Sugitani M, Sugai T, Tanifuji M, Onoda N (1994) Signal propagation from piriform cortex to the endopiriform nucleus in vitro revealed by optical imaging. Neurosci Lett 171:175–178. Tanabe T, Iino M, Takagi SF (1975) Discrimination of odors in olfactory bulb, piriform-amygdaloid areas and orbitofrontal cortex of the monkey. J Neurophysiol 38:1284 –1296. Tanifuji M, Sugiyama T, Murase K (1994) Horizontal propagation of excitation in rat visual cortical slices revealed by optical imaging. Science 266:1057–1059. Wilson DA (1998) Habituation of odor responses in the rat anterior piriform cortex. J Neurophysiol 79:1425–1440. Wilson DA (2000) Comparison of odor receptive field plasticity in the rat olfactory bulb and anterior piriform cortex. J Neurophysiol 84: 3036 –3042. Yamamoto T, Yuyama N, Kato T, Kawamura Y (1985) Gustatory responses of cortical neurons in rats. II: Information processing of taste quality. J Neurophysiol 53:1356 –1369. Yarita H, Iino M, Tanabe T, Kogure S, Takagi SF (1980) A transthalamic olfactory pathway to the orbitofrontal cortex in the monkey. J Neurophysiol 43:69 –85. Yasui Y, Saper CB, Cechetto DF (1989) Calcitonin gene-related peptide immunoreactivity in the visceral sensory cortex, thalamus, and related pathways in the rat. J Comp Neurol 290:487–501.
(Accepted 18 March 2004) (Available online 2 June 2004)
CONVERGENCE OF OLFACTORY AND GUSTATORY CONNECTIONS ONTO THE ENDOPIRIFORM NUCLEUS IN THE RAT W. FU,a T. SUGAI,a H. YOSHIMURAb AND N. ONODAa*
probably receives temperature signals from the tongue (Kosar et al., 1986; Norgren, 1995). The GI caudal to the middle cerebral artery (MCA) receives general visceral afferents rather than taste signals (Cechetto and Saper, 1987; Saper, 1995). Further, the prominent efferent projections to autonomic structures arise from the agranular division (AI) of the insular cortex (IC), just ventral to the DI (Allen et al., 1991; Saper, 1995). Furthermore, intrinsic connections between the divisions in the IC have been reported (Shi and Cassell, 1998). However, there are many different observations related to communication and differentiation of function between the IC divisions. The primary olfactory cortex is defined as the region that receives direct fiber projections from the olfactory bulb (Price, 1987). The piriform cortex (PC) is the largest structure in the olfactory cortex and further the piriform region includes a nucleus localized in between the PC proper and the caudate-putamen (Haberly, 1998), which is named the “endopiriform nucleus” (EPN). The PC is an important center before delivery of olfactory information to the higher system via the EPN (Haberly, 1998). Tracer studies have shown that the long axons of the EPN neurons form a net of intrinsic connections and project over long distances to the forebrain structures, such as the olfactory cortex, IC, and amygdala (Price, 1987; Behan and Haberly, 1999). Unit recording studies demonstrated that odor stimuli activated single neurons in the PC (Haberly, 1969; Tanabe et al., 1975; Nemitz and Goldberg, 1983; Schoenbaum and Eichenbaum, 1995; Wilson, 1998, 2000), hypothalamus (Scott and Pfaffmann, 1972; Kogure and Onoda, 1983), and orbitofrontal cortex (Tanabe et al., 1975; Yarita et al., 1980; Motokizawa and Ino, 1983; Onoda et al., 1984; Schoenbaum and Eichenbaum, 1995) via the thalamus (Imamura et al., 1984). Single neurons in the orbitofrontal cortex responded to taste and odor stimuli in the monkey (Rolls, 1989). These results suggest that the orbitofrontal cortex is the highest olfactory center, where the olfactory and gustatory information might be integrated. However, taste-sensitive neurons in the orbitofrontal cortex have not been observed in other mammal species. Anterograde tracer injection into the EPN demonstrated labeled axons in the DI and AI (Behan and Haberly, 1999). Further, voltage imaging in slices has also revealed that paroxysmal epileptiform activity can develop in layer VI of the AI in synchrony with that in the EPN (Demir et al., 1998). Thus, the EPN and AI are predicted as the regions where olfactory and gustatory information are centrally integrated. In this study, we present mapping data obtained from electrical and optical recordings to identify these integrative regions in the CNS.
a Department of Physiology, Kanazawa Medical University, Uchinada, Ishikawa 920-0293, Japan b Department of Oral and Maxillofacial Surgery, Kanazawa Medical University, Uchinada, Ishikawa 920-0293, Japan
Abstract—Electrical and optical recordings were made from slice preparations including the piriform and gustatory cortices. Electrical stimulation of the gustatory cortex evoked a characteristic field potential in the endopiriform nucleus. A field potential was induced in the endopiriform nucleus by stimulation of the piriform cortex. Voltage-sensitive dye studies showed that stimulation of the piriform cortex induced signal propagation from the piriform cortex to endopiriform nucleus, whereas stimulation of the gustatory cortex did the same from the gustatory cortex to endopiriform nucleus via the agranular division of the insular cortex. After stimulation of the endopiriform nucleus, optical signals propagated not only to the piriform cortex but also to the gustatory cortex via the agranular division of the insular cortex. The olfactory and gustatory pathways appeared to be reciprocally connected. Unit recordings indicated that olfactory and gustatory activity converged onto a single neuron of the endopiriform nucleus. It is suggested that the cortical integration of olfactory and gustatory information could modulate mechanisms involved in food selection and emotional reactions relating to the chemical senses. © 2004 IBRO. Published by Elsevier Ltd. All rights reserved. Key words: field potential, optical imaging, voltage-sensitive dye, signal propagation, insular cortex.
The primary gustatory cortex (GC) is defined as the cortical region that receives direct fiber projection from the parvicellular part of the ventral posteromedial nucleus in the thalamus, and the terminal labeling area is situated rostrally in the granular division (GI) and caudally in the dysgranular division (DI) of the insular cortex (Nakashima et al., 2000). In fact, unit recordings found gustatory responses of neurons from both the GI and DI (Yamamoto et al., 1985; Ogawa et al., 1992), suggesting the location of the GC in both divisions. The ventral part (Vp) of the GI *Corresponding author. Tel: ⫹81-76-218-8102; fax: ⫹81-76-286-3523. E-mail address: [email protected] (N. Onoda). Abbreviations: ACSF, artificial cerebrospinal fluid; AI, agranular division of the insular cortex; CGRP, calcitonin gene-related peptide; DI, dysgranular division of the insular cortex; EPN, endopiriform nucleus; EPSC, excitatory postsynaptic current; EPSP, excitatory postsynaptic potential; GABAB-R, B-type of receptor of GABA; GC, gustatory cortex; GI, granular division of the insular cortex; IC, insular cortex; MCA, middle cerebral artery; MD, mediodorsal thalamic nucleus; OB, olfactory bulb; PC, piriform cortex; PDH, pulse density histogram; Vp, ventral part.
0306-4522/04$30.00⫹0.00 © 2004 IBRO. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.neuroscience.2004.03.041
1033
1034
W. Fu et al. / Neuroscience 126 (2004) 1033–1041
B
L V
ACAd
A
D M
RhS ec
PC
OB
EPN
ac
LOT
5 3 2 1
MCA
C
R1
GI DI
6
1
1
2 1
D
S1
AI
3
LOT
RhS
DI
2 3
2 3
LOT
6
R4
IV 4
S4 ec L
D
EPN V
M
R2
S2
RhS
PC
AI
PC
5
GI DI AI
PC I II III EPN
1
3
500 µm
2
Fig. 1. Slice preparations and their morphological features. (A) A lateral view of the rat brain. Slice preparations were cut parallel to the MCA as indicated by a dotted line. (B) A frontal section with a cutting angle (approximately 10° off the frontal plane). A ventro-lateral region surrounded by thick dotted lines is expanded in C. (C) A schematic drawing of experimental arrangements. Stimulation site 1 (S1) and recording site 1 (R1) are in layer II of the PC. S2 and R2 are in the EPN. S4 and R4 are in the GC. (D) A micrograph of the frontal section stained with Neutral Red corresponds to the region surrounded by thick dotted lines. This region was a histologically determined camera field in the optical recordings in Fig. 3B and C. This micrograph was reconstructed from resliced sections after dye was deposited in recording and stimulation sites. Spot 1 is in layer II of the PC. Spot 2 is in the EPN. Spot 3 is in layer V/VI of the AI of the IC. Spot 4 is in the DI of the IC, the GC. Closed arrowheads indicate boundaries between divisions of IC. An open arrowhead indicates layer IV of GI. ac, anterior commissure; D, dorsal direction; ec, external capsule; L, lateral direction; I, layer I; II, layer II; III, layer III; IV, layer IV; LOT, lateral olfactory tract; M, medial direction; RhS, rhinal sulcus; V, ventral direction.
EXPERIMENTAL PROCEDURES Animal protocols used in this study complied with all pertinent institutional and Japanese Government regulations, and every attempt was made to minimize the number of killed animals and their suffering.
Materials Young Wistar rats (n⫽25, 100 –150 g) were used. To prepare slices (340 – 400 m thick) including the piriform and gustatory cortices, slices were cut in a plane approximately 10° off the frontal (near parallel to the MCA), from 2.0 mm rostral to the anterior edge of the optic chiasm, or from 0.8 mm rostral to the MCA to 1.2 mm caudal to the MCA (Fig. 1A and B). The slices were perfused continuously (⬎2 h at 30 °C) with oxygenated (95% O2/5% CO2) artificial cerebrospinal fluid (ACSF) containing (in mM): 124 NaCl, 5 KCl, 1.24 KH2PO4, 1.3 MgSO4, 2.4 CaCl2, 26 NaHCO3, and 10 glucose.
Preparation for electrophysiological recording After recovery, a slice was transferred to the recording chamber mounted on the stage of an inverted microscope (IMT-2; Olympus, Japan) and perfused with oxygenated ACSF. Bath temperature was maintained at 26 –28 °C. Mono-tungsten stimulating electrodes were inserted into layer I/II of the PC, layer II/III (corresponding to spot 4 in Fig. 1D) of the GC and/or EPN,
and single square pulses (0.02–1 mA, 80 s) were delivered at 0.05– 0.07 Hz. The stimulus intensity was set to about 5– 8⫻ the threshold, by which the optical signal reached its maximum spatial extent. Glass microelectrodes containing 2% Brilliant Blue in 0.85% NaCl (4 –30 M⍀, DC) were used to record extracellular field potentials, record unit activity, and mark the location of the electrode tip with dye deposit (10 A for 3–5 min). Four to eight sweeps of field potentials were processed for averaging. Patch electrodes for current clamp recordings from EPN neurons contained (in mM): 126 K-gluconate, 4 KCl, 1 CaCl2, 1 MgCl2, 10 HEPES, 10 EGTA, 2 Na2-ATP, 0.3 Na2-GTP (pH⫽7.3 with KOH). For voltage clamp recordings, internal solution was identical except that 126 CsMeSO4, 4 CsCl were substituted for 126 K-gluconate, 4 KCl, respectively. The pH was adjusted to 7.3 with CsOH. The osmolarity of these pipette solutions was 295–299 mOsm. These electrophysiological data, including field and unit responses, were stored and analyzed with the use of pClamp software (Axon Instruments, Foster City, CA, USA).
Preparation for optical recording After recovery, a slice was incubated with a voltage-sensitive dye NK2761 (0.13 mg/ml; Nippon Kankoh-Shikiso Kenkyusho, Okayama, Japan) for 18 min, transferred to the recording chamber, and perfused with ACSF (26 –28 °C). Mono-stimulating electrodes were inserted into the layer I/II of the PC, GC and/or EPN.
W. Fu et al. / Neuroscience 126 (2004) 1033–1041
1035
Fig. 2. Mapping data of field potential responses in three recording sites (R1, R2, and R4) following stimulation of three different sites (S1, S2, and S4) indicated in Fig. 1C. (A) A time-lapse of field potentials in the EPN following stimulation of PC neurons in normal solution. (B) Field potentials were averaged from eight sweeps. PC-evoked EPN responses in normal (a) and Mg2⫹ free solution (b). GC-evoked EPN responses in normal (d) and Mg2⫹ free solution (e). EPN-evoked PC (c) and GC responses (f) in Mg2⫹ free solution. Closed arrowheads under traces indicate the onset of stimulation.
The stimulating parameters were the same as above. Glass microelectrodes containing blue dye were used to monitor field potentials and to deposit the dye.
Optical recording The camera unit of the optical imaging system (Fujix HR-Deltaron 1700; Fujifilm Microdevices, Tokyo) contains a 128⫻128 photodiode array. With the 4⫻ objective, the whole array corresponded to a 2.26⫻2.26 mm2 area of tissue. The light passed through a narrow band interference filter (700⫾30 nm). Sixteen responses to single shocks were averaged to form a run. Neural activity was recorded optically by monitoring absorption changes in transmitted light intensity in each element at a rate of one frame per 0.6 ms. The details have been described elsewhere (Ichikawa et al., 1993; Sugitani et al., 1994; Tanifuji et al., 1994; Obaid and Salzberg, 1996; Sugai et al., 1997).
Histological reconstruction and immunohistochemistry After electrical or optical recordings, appropriate sites in the pathway of signal propagation following stimulation of the PC/GC and/or EPN were confirmed on a display screen, and then each site was marked. A tip of a glass pipette containing dye was superimposed on the mark on the screen after the display was replaced with the real image of the slice preparation, and then the dye was deposited and the site was penetrated by the stimulating electrode. The slice was fixed with 1% paraformaldehyde in 0.1 M phosphate buffer and resectioned (40 m). Sections were processed by a immunohistochemical technique. After elimination of endogenous peroxidase activity with 0.1% H2O2, alternative sections were incubated with an antibody (Chemicon, Temecula, CA, USA) against a B-type receptor of GABA (GABAB-R) at a dilution of 1:10,000 or with an antibody (Biogenesis, Kingston, NH, USA) against calcitonin gene-related peptide (CGRP) at a dilution of 1:4000, and then incubated in avidin– biotin–peroxidase complex solution (Vector, Burlingame, CA, USA). All sections were photographed. The location of the camera field in the slice was also histologically reconstructed.
RESULTS
Fig. 1C, where the ventral region of the slice surrounded by thick dotted lines in Fig. 1B was expanded. Fig. 1D shows an example of a frontal section on which blue dye was deposited. Spot 1 is in layer II of the PC and spot 2 is in the EPN. Mapping studies of field potentials Electrical stimulation of layer II in the PC evoked a field potential with a latency of 35 ms in the central region of the EPN (Fig. 2A). The PC-evoked field responses occurred in initial stimulations, but their latencies became longer gradually (oblique dotted line) in ACSF containing Mg2⫹ (normal solution) and often disappeared after multiple stimulation (Fig. 2A and “a” in Fig. 2B). Disappearance of the PC-evoked field potentials was blocked in ACSF without Mg2⫹ (“b” in Fig. 2B). A field potential with a latency of 66 ms was elicited by a single shock of layer II/III of the GC in the EPN in normal solution (“d” in Fig. 2B). The GC-evoked field responses occurred in initial stimulations, but their latencies became longer gradually in normal solution and then the field responses often disappeared after multiple stimulation (data not shown). Disappearance of the GCevoked field potentials was also blocked in Mg2⫹ free solution (“e” in Fig. 2B). When the stimulation sites in the GC and PC were moved from layer to layer, the onset of the field potentials (latent periods) differed somewhat, but the waveform of individual field potentials was identical. These electrophysiological results suggested that excitation traveled a long distance, and that both the PC and GC are connected with the EPN. Next, electrical stimulation of the EPN provoked characteristic field potentials with their latency of 24 and 114 ms in the PC (“c” in Fig. 2B) and GC (“f” in Fig. 2B), respectively, in Mg2⫹ free solution. These results suggested that the PC and GC are connected reciprocally with the EPN.
Slice preparations and their morphological features
Results from optical imaging studies of signal propagation
Fig. 1B shows the morphological feature of a slice. The experimental arrangement was schematically drawn on
Optical recordings were carried out on brain slices in normal or Mg2⫹ free solution. Field potentials and optical
1036
W. Fu et al. / Neuroscience 126 (2004) 1033–1041
A
RhS
GC
PC
S1
4.8 ms
10.8
16.8 ms
28.8
13.8 ms
16.8
22.8
28.8
142.8
31.8
46.8
58.8
91.8
43.8
57.0
67.8
79.8
109.8
7.8 ms
35.4
43.8
61.8
75.0
142.2
28.8 ms
79.8
127.8
145.8
157.8
193.8
II
R2 EPN
ec
S4
B
PC
GC R2 II AI
EPN
ec
S4
C
PC
GC
II
AI
EPN
ec
R2 PC
D
II
GC
S2
E
EPN
R4
R1 ec
GC AI
ec
PC II
0.2 mV
S2 EPN
Stim 0%
0.075 % (A,B,C) 0.04 % (D, E)
50 ms (A, B, C, D) 100 ms (E)
Fig. 3. A time-lapse optical image of neural excitation in tissue evoked by PC (A), GC (B, C) and EPN stimulation (D, E). (A, C–E) In Mg2⫹ free solution. (B) In normal solution. Numerals on each image are lapse-periods after stimulation. White arrows and arrowheads in images indicate the EPN and layer VI of AI, respectively. The former corresponds to spot 2 in Fig. 1D and the latter corresponds to spot 3. Schematic drawings of individual camera fields are given on the far left. Closed arrowheads indicate individual stimulation sites (S1, S2, and S4). Open arrowheads indicate individual recording sites (R1, R2, and R4) of field potentials on the far right. The left and top are the dorsal and lateral directions, respectively. Colors below E indicate changes in light absorption (in percent). A thin arrow under the lowest trace on the far right marks the onset of stimulation. Abbreviations are the same as in Fig. 1.
signals were recorded concurrently (far right of Fig. 3). PC stimulation (S1) provoked an early optical response, which traveled to other stimulation sites in the PC (image at 4.8 ms) in Mg2⫹ free solution (Fig. 3A). Subsequently, the optical signal was found to propagate 1) horizontally along layer II and vertically toward the deep layers in the PC (image at 10.8 ms after S1), 2) vertically toward the EPN and horizontally along layer III (16.8 ms), and 3) horizontally along the EPN (22.8 ms). The distance that electrically-induced optical signals in the PC and EPN (white arrow) traveled was maximum by 28.8 ms; these signals faded gradually by 142.8 ms, and almost disappeared by 300 ms. The pattern of signal propagation and propagation velocity (51.3⫾10.5 mm/s [mean⫾S.E., n⫽6]) were similar in six slices. The GC shock (S4) provoked an early optical response around stimulation sites in four slices in normal solution (Fig. 3B). The optical signal propagated 1) vertically toward deep layers in the GC (16.8 ms after S4), 2) horizontally within the deep layer and obliquely
toward deep layers in the AI (white arrowhead in 28.8 ms), 3) obliquely toward the EPN (31.8 ms), and finally 4) horizontally along the EPN (white arrow in 46.8 ms). The distance that the optical signals in the EPN (white arrow) traveled was maximum by 58.8 ms; signals in the GC and AI weakened by 91.8 ms and almost disappeared by 120 ms. The pattern of signal propagation and propagation velocity (42.5⫾7.6 mm/s [n⫽4]) were similar in all slices. The GC shock (S4) produced stronger and longer optical responses in 10 slices in Mg2⫹ free solution than in normal solution (Fig. 3C). In particular, optical responses in the deep layers of the AI (white arrowhead in 43.8 ms) and EPN (white arrows in 67.8 and 79.8 ms) were markedly intense. The signal propagation velocity was 33.9⫾ 3.0 mm/s (n⫽10). These results suggested that the PC and GC were connected with the EPN. It appeared that the GC-EPN propagation passed through the AI. EPN stimulation (S2) provoked early optical response (7.8 ms) in four slices in Mg2⫹ free solution (Fig. 3D). Then
W. Fu et al. / Neuroscience 126 (2004) 1033–1041
Dp
AI
Dp
Vp
AI
B
A
GI
Vp
1037
I
C
II
RhS
III
IV 3
AI Dp
DI
4
V VI
500 µm
3
Fig. 4. Immunohistochemical results and histological reconstruction. (A) A microphotograph of a resectioned slice incubated with an antibody against CGRP. Note that CGRP-positive fibers (thick arrows) were dominant in the Vp of the AI. Spot 3 (thin arrows) indicated by white arrowheads in Fig. 3B, C and E was found in the AI. (B) A microphotograph of an adjacent section incubated with an antibody against B-type of receptor of GABAB-R. Spot 3 (thin arrows) was found under the layer containing GABAB-R-positive cells (thick arrows). (C) A microphotograph expanded from the left part of Fig. 1D. A distinct layer of cells in between the thick arrows is the GI, suggesting layer IV. Spot 4 (thin arrow) was located in the region without layer IV and just dorsal to the AI, suggesting layer III of the DI, i.e. the GC. Abbreviations are the same as in Fig. 1. Dp, dorsal part of the AI.
the optical signal propagated 1) horizontally along the EPN (white arrow) and vertically toward layer III in the PC (35.4 ms), 2) vertically toward layers II and III in the PC (43.8 ms), and 3) horizontally along layers II and III in the PC (61.8 ms). The distance that the optical signal in the PC traveled was maximum by 75.0 ms. Optical signals faded gradually (142.2 ms) and almost disappeared at 300 ms after S2. The signal propagation and propagation velocity (40.4⫾5.9 mm/s [n⫽4]) were similar in all slices. After the early optical response to S2 in five slices in Mg2⫹ free solution (Fig. 3E), the optical signal traveled to the GC 1) horizontally toward a deep layer in the AI (white arrowhead in 28.8 ms), 2) vertically toward the surface of the AI (127.8 ms), and 3) horizontally along superficial layers in the GC. The distance that the optical signal traveled in the GC was maximum by 157.8 ms and the signal disappeared 300 ms after S2. The pattern of signal propagation and propagation velocity (19.7⫾2.4 mm/s; n⫽5) was similar in all slices. These results suggested that the PC and GC were connected reciprocally with the EPN. It appeared, thus, that both GC-EPN and EPN-GC propagations passed through the AI. Immunohistochemical results and histological reconstruction After the signal patterns in slices were optically recorded (Fig. 3B, C and E), blue dye was deposited into regions indicated by white arrowheads. CGRP-positive fibers (thick arrows) were observed in the AI (Fig. 4A). Spot 3 was found in the region containing CGRP-positive fibers, which were mainly in the Vp of the AI (Yasui et al., 1989). GABAB-R-positive cells were observed in an adjacent section (Fig. 4B). Spot 3 was found under the layer containing GABAB-R-positive cells, which were located in layer V of
the AI (Jasmin et al., 2003). Thus, spot 3 appeared to be in layer VI of the AI. The left part of Fig. 1D is magnified in Fig. 4C. A distinct layer of cells in between thick arrows can be seen in the GI, suggesting layer IV. Spot 4 was located in the region without layer IV and just dorsal to the AI. Thus, spot four appeared to be in layer III of the DI, which is the GC. A boundary between the AI and DI was determined histochemically by presence or absence of CGRPlabeled fibers. Unit recording study Activity of the EPN units in Mg2⫹ free solution is demonstrated in Fig. 5A and B. PC stimulation provoked a long burst of spikes (“a” in Fig. 5A). A large decrease in amplitude of subsequent spikes was observed, and spike amplitude recovered gradually. Bursting activity can be seen clearly in a pulse density histogram (PDH) and initial spikes appear at 25–30 ms after every PC shock (“d” in Fig. 5A). The same EPN unit responded to GC shocks (“b” in Fig. 5A). Subsequent to an initial spike, a large decrease in amplitude of spikes was also observed. The latency of the initial spike was 38 – 43 ms and the delay of the latency after GC shocks can be seen clearly in the PDH (“e” in Fig. 5A). Unit responses to PC and GC shocks showed similar long bursting patterns. As stimulation of either the PC or GC elicited a long burst of discharges, effects of simultaneous stimulation were less clear (“c” and “f” in Fig. 5A). A burst of spikes was recorded from another EPN unit after PC stimulation (“a” in Fig. 5B). This unit discharged bursting spikes in response to GC shocks (“b” of Fig. 5B). These discharge patterns were similar although this unit fired initially at 28 and 70 ms after PC and GC shocks, respectively. Further, the bursting frequency in Fig. 5B was lower than that shown in Fig. 5A. Fluctuation of the latency of
W. Fu et al. / Neuroscience 126 (2004) 1033–1041
A No. of spikes/bin
a
b
c
C
200 ms
12 10 8 6 4 2 0 0 12 10 8 6 4 2 0 0 12 10 8 6 4 2 0 0 10 8 6 4 2 0
No. of spikes/bin
a
b 200 ms
10 8 6 4 2 0
B
d 20
40
60
80 100
20
40
60
80 100
20
40
60
80 100
e
f
a
Ti me (ms)
c 0 20 40 60 80 100 120140 160
d
No. of spikes/bin
1038
b c
D
200 ms
12 10 8 6 4 2 0
0 10 8 6 4 2 0 12 0 10 8 6 4 2 0 0
d e
20
40
60
80 100
20
40
60
80 100
40
60
80 100
f 20
Ti me (m s)
a b c
50 mV 500 ms
0 20 40 60 80 100 120 140 160
Ti m e (ms)
Fig. 5. Mapping data of unit activity and whole cell patch recordings. (A) Unit activity in the EPN following stimulation of the PC and GC. Bursting spike responses of an EPN neuron to a single PC (a) or GC shock (b). Spike response of the same unit from simultaneous stimulation of the PC and GC (c). PDHs processed from 16 sweeps of unit responses to PC (d), GC (e), and both PC and GC stimulation (f). (B) Spike responses of another EPN unit to a single PC (a) or GC shock (b). PDHs processed from 16 sweeps of PC- (d) and GC-evoked responses (e). Spike response (c) and the PDH (f) from simultaneous stimulation of the PC and GC. (C) Superimposed responses in PC (a) and GC units (b) from eight sweeps following stimulation of the EPN. PDHs processed from 16 sweeps of EPN-evoked responses in PC (c) and GC (d) units. (D) Whole cell patch recording made in a current clamp mode from EPN neurons in Mg2⫹ free or 0.5 mM Mg2⫹ solution. Long-duration depolarizing potentials with superimposed action potential responses in an EPN neuron to a single PC shock (a). A GC-evoked long EPSP with superimposed action potentials from two other EPN neurons (b) and (c). Arrows under traces and PDHs indicate the onset of stimulation.
initial spikes and the difference between latency after PC shock and latency after GC shock can be seen clearly in PDHs (“d” and “e” in Fig. 5B). Simultaneous stimulation of the PC and GC produced a similar response as shown in “a” and “d” of Fig. 5B, but GC-evoked response was not clear (“c” and “f” in Fig. 5B). Stimulation of EPN evoked spike discharges in layer II of the PC (“a” in Fig. 5C) and layer II/III of the GC (“b” in Fig. 5C). Large increases in the impulse rate and the difference of their latencies of initial spikes in the PC and GC can be seen clearly in PDHs (“c” and “d” in Fig. 5C). The average latency of initial spikes was 59.7⫾4.2 ms (mean⫾S.E., n⫽19) for PC neurons and 96.9⫾4.9 ms (n⫽28) for GC neurons. It appeared that discharge patterns of EPN neurons differed largely from those of PC and GC neurons. Of the 30 EPN units we recorded, 25 units responded to both PC and GC shocks. The discharge pattern in
response to PC and GC shocks showed a similar burst in single EPN units, although latencies of individual responses were different. These results suggested that olfactory and gustatory connections converged onto a single EPN neuron. Further, EPN shocks evoked neural activity in the PC and GC. Out of 24, 19 PC units responded to EPN shocks, whereas out of 31, 28 GC units did. Thus, PC-EPN and GC-EPN connections appeared to be reciprocal. Patch clamp recording study Finally, to confirm results obtained from unit responses to PC and GC shocks, electrical recordings were made from EPN neurons in the whole-cell configuration in Mg2⫹ free or 0.5 mM Mg2⫹ solution using the patch clamp technique. The EPN neurons were identified by their size and location
W. Fu et al. / Neuroscience 126 (2004) 1033–1041
in the EPN. PC shocks evoked a long lasting depolarizing potential (⬎1.5 s) with a superimposed burst of action potentials with low amplitude in a current clamp mode (“a” in Fig. 5D). Immediately after an initial action potential, the amplitude of subsequent action potentials was small, but the decrease in amplitude recovered gradually. This depolarizing potential was the longest lasting (half-amplitude duration, 980 ms) among EPN neurons responding to PC shocks. The average reversal potential of the PC-evoked long-lasting depolarizing inward currents was 0.9⫾1.4 mV (mean⫾S.E.) in four other EPN neurons in a voltage clamp mode, suggesting excitatory postsynaptic currents (EPSCs). The average duration of the half amplitude of PCevoked excitatory postsynaptic potentials (EPSPs) was 633⫾136 ms (mean⫾S.E., n⫽10). Next, GC shocks evoked a burst of high frequency action potentials that became superimposed on the ongoing long-duration depolarizing potential (⬎2.5 s) in another EPN neuron in the current clamp mode (“b” in Fig. 5D). Amplitude of the action potentials gradually recovered. This depolarizing potential of this EPN neuron was the longest lasting (half amplitude duration, 1390 ms) of the potentials in all EPN neurons recorded. A GC-evoked depolarizing potential (⬍1 s) was recorded from another EPN neuron (“c” in Fig. 5D) and its half amplitude duration was 325 ms. GCevoked EPSCs were reversed at 1.2⫾1.3 mV (mean⫾S.E., n⫽5), whereas mean half amplitude duration of GC-evoked EPSPs was 444⫾88 ms (mean⫾S.E., n⫽14). EPN neurons fired with low or high frequency following stimulation of the GC, because duration of GCevoked EPSPs varied. These results of extracellular recordings suggested that the long lasting EPSPs were responsible for a large decrease and subsequent gradual increase in unit but not multiunit amplitude following stimulation of the PC or GC.
DISCUSSION In this study, frontal slices were cut as parallel as possible to the MCA and the cutting angles were determined by results obtained from electrical recordings. Further, since the bed nucleus of the anterior commissure, which is used as a landmark in the stereotaxic rat brain (Paxinos and Watson, 1986), was involved in these slices, it was ready to calculate the positioning of the slice structures from the stereotaxic reference points. In some slices, PC but not GC shocks evoked the EPN response, while in others the converse was true. Fiber connections were probably severed. When slices were in ACSF containing Mg2⫹ (normal solution), the field potential in response to either PC or GC shocks disappeared for a short while after stimulation. The disappearance of the responses was blocked by Mg2⫹ free ACSF. Under either Mg2⫹ containing or Mg2⫹ free conditions in this study, following stimulation of layer I/II in the PC or layer II/III in the GC, characteristic field potentials were recorded in the EPN. Effects of Mg2⫹ free solution on field potentials have been investigated in PC slices of guinea-pigs (Richards and Sercombe, 1970) and rats (Hoffman and Haberly, 1989, 1991, 1993). Under Mg2⫹
1039
free conditions, the surface potentials became hyperexcitable (Richards and Sercombe, 1970) and stimulation of association fibers elicited epileptiform responses in the EPN (Hoffman and Haberly, 1989, 1991, 1993). Intracellular recording studies demonstrated that burst response evoked by stimulation of fiber tracts in PC slices in Mg2⫹ free solution consisted of high amplitude, long-duration depolarizing potentials with superimposed action potentials, suggesting an interaction of cells via activated NMDA glutamate receptors (Hoffman and Haberly, 1989, 1991, 1993). Our patch clamp recordings from EPN neurons in Mg2⫹ free solution demonstrated that PC and/or GC shocks evoked long-lasting EPSPs with superimposed action potentials. These results suggest that a reason the field response disappeared in normal solution might be the lack of activation of NMDA glutamate receptors. Optical imaging in the PC region of guinea-pig sagittal slices has demonstrated that optical signals elicited by afferent shocks propagated along layers II and III, then into the EPN, where neural activity moved slowly in the caudal direction along the outer edge of the external capsule (Sugitani et al., 1994). Epileptiform activity evoked by stimulation of association fibers in Mg2⫹ free medium was recorded from the EPN in transverse and longitudinal slices (Demir et al., 1998, 2001). Present optical results in the PC region of rat frontal slices are consistent with previous optical results showing olfactory excitation reached the EPN after stimulation of fiber tracts. A tracer injection study into the olfactory bulb (OB) demonstrated that OB of the mouse had a direct terminal projection to a sector of the IC, and this part of the IC had inputs from the major ascending gustatory and visceral afferent relay structures (Shipley and Geinisman, 1984). However, a direct OB–IC projection has not been found in other mammal species. An electrophysiological study demonstrated that field potentials and unit responses were generated in the prefrontal cortex following electrical stimulation of the OB, PC, mediodorsal thalamic nucleus (MD), and amygdala (Cinelli et al., 1987), suggesting the OB–IC projection is via several synaptic relays. In all species for which data are available, at least two additional routes from OB to IC are suggested: (1) one way is from OB to PC and from PC to IC (Motokizawa and Ino, 1983); (2) the second route is from PC by way of the EPN to the MD and from the MD to the IC (Gerfen and Clavier, 1979; Price, 1987). From a study of anterograde tracer injection in the EPN, labeled axons were found not only in the MD but also in both the DI and AI, and only a few in the GI, and label was present in all layers but concentrated in layer I (Behan and Haberly, 1999). The PC–IC projection by way of the EPN is considered as another route accordingly. In fact, typical extracellular responses associated with abruptonset epileptiform-EPSP, which originated in the EPN, were observed in IC neurons in Mg2⫹ free medium (Hoffman and Haberly, 1993). Voltage imaging in slices has revealed that paroxysmal epileptiform activity can develop in layer VI of the AI in synchrony with that in EPN in low-chloride medium (Demir et al., 1998). Other optical data also identified EPN-AI propagation (Demir et al.,
1040
W. Fu et al. / Neuroscience 126 (2004) 1033–1041
2000). Our optical results demonstrated EPN–GC propagation of signals via the AI and vice versa, although polysynaptic connections were responsible for slow propagation speed of optical signals. The remarkably slow propagation times may be due to bath temperature, smaller diameter axons in the EPN (Behan and Haberly, 1999), inhibitory synaptic action, or wave-like propagation. The AI is associated with visceral sensory and autonomic functions rather than taste (Cechetto and Saper, 1987; Krushel and van der Kooy, 1988; Cechetto and Chen, 1990; Allen et al., 1991; Saper, 1995). An immunohistochemical study revealed that the AI and DI were innervated by CGRP-immunoreactive fibers and the density of CGRPpositive innervation was considerably richer in the ventral AI (Yasui et al., 1989). Our results indicated a blue spot formed over the region where CGRP-positive fibers were abundant (Fig. 4A). Together with our optical results, this blue spot appeared to be evidence of signal propagation to the GC from the EPN via the AI and vice versa. A recent immunohistochemical study has shown, further, that GABAB-Rs are concentrated on pyramidal neurons of layer V in the AI, and GABAB neuron activity could change the pain threshold and then might cause hyperalgesia (Jasmin et al., 2003). In this study, spot 3 was observed under a GABAB-R-positive layer (Fig. 4B). It appeared that spot 3 positioned in layer VI of the AI. From results obtained from unit recordings, furthermore, olfactory and gustatory activity converged onto a single EPN neuron, although the long-duration EPSPs evoked by either stimulus were responsible for the reduced effects of simultaneous stimulation of PC and GC on EPN neurons. The EPN–PC and EPN–GC connections were also observed. Our results led us to conclude reciprocal connections between olfactory and gustatory pathways. Together with the descending connections from the EPN and AI to the amygdala (Price, 1987; Haberly, 1998; Shi and Cassell, 1998; Price, 1995; Nakashima et al., 2000), it is possible that the cortical integration of olfactory, gustatory, visceral and nociceptive information could modulate mechanisms involved in food selection and emotional reactions relating to the chemical and pain senses. Acknowledgements—The authors wish to thank Mr. Adachi and Mr. Muramoto for technical assistance. This work was partly supported by a Grant for Collaborative Research, C2001-01 from Kanazawa Med. Univ.
REFERENCES Allen GV, Saper CB, Hurley KM, Cechetto DF (1991) Organization of visceral and limbic connections in the insular cortex of the rat. J Comp Neurol 311:1–16. Behan M, Haberly LB (1999) Intrinsic and efferent connections of the endopiriform nucleus in rat. J Comp Neurol 408:532–548. Cechetto DF, Chen SJ (1990) Subcortical sites mediating sympathetic responses from insular cortex in rats. Am J Physiol 258:R245– R255. Cechetto DF, Saper CB (1987) Evidence for a viscerotopic sensory representation in the cortex and thalamus in the rat. J Comp Neurol 262:27–45. Cinelli AR, Ferreyra-Moyano H, Barragan E (1987) Reciprocal func-
tional connections of the olfactory bulbs and other olfactory related areas with the prefrontal cortex. Brain Res Bull 19:651–661. Demir R, Haberly LB, Jackson MB (1998) Voltage imaging of epileptiform activity in slices from rat piriform cortex: onset and propagation. J Neurophysiol 80:2727–2742. Demir R, Haberly LB, Jackson MB (2000) Characteristics of plateau activity during the latent period prior to epileptiform discharges in slices from rat piriform cortex. J Neurophysiol 83:1088 –1098. Demir R, Haberly LB, Jackson MB (2001) Epileptiform discharges with in-vivo-like features in slices of rat piriform cortex with longitudinal association fibers. J Neurophysiol 86:2445–2460. Gerfen CR, Clavier RM (1979) Neural inputs to the prefrontal agranular insular cortex in the rat: horseradish peroxidase study. Brain Res Bull 4:347–353. Haberly LB (1969) Single unit responses to odor in the prepiriform cortex. Brain Res 12:481–484. Haberly LB (1998) Olfactory cortex. In: The synaptic organization of the brain (Shepherd GM, ed), pp 377–416. Oxford: Oxford University Press. Hoffman WH, Haberly LB (1989) Bursting induces persistent all-ornone EPSPs by an NMDA dependent process in piriform cortex. J Neurosci 9:206 –215. Hoffman WH, Haberly LB (1991) Bursting-induced epileptiform EPSPs in slices of piriform cortex are generated by deep cells. J Neurosci 11:2021–2031. Hoffman WH, Haberly LB (1993) Role of synaptic excitation in the generation of bursting-induced epileptiform potentials in the endopiriform nucleus and piriform cortex. J Neurophysiol 70:2550 – 2561. Ichikawa M, Iijima T, Matsumoto G (1993) Real-time optical recording of neuronal activities in the brain. In: Brain mechanisms of perception and memory (Ono T, Squire LR, Raichle ME, Perrett DI, Fukuda M, eds), pp 638 –648. New York: Oxford University Press. Imamura K, Onoda N, Takagi SF (1984) Odor response characteristics of thalamic mediodorsal nucleus neurons in the rabbit. Jpn J Physiol 34:55–73. Jasmin L, Rabkin SD, Granato A, Boudah A, Ohara PT (2003) Analgesia and hyperalgesia from GABA-mediated modulation of the cerebral cortex. Nature 424:316 –320. Kogure S, Onoda N (1983) Response characteristics of lateral hypothalamic neurons to odors in unanesthetized rabbits. J Neurophysiol 50:609 –617. Kosar E, Grill HJ, Norgren R (1986) Gustatory cortex in the rat. I: Physiological properties and cytoarchitecture. Brain Res 379:329 – 341. Krushel LA, van der Kooy D (1988) Visceral cortex: integration of the mucosal senses with limbic information in the rat agranular insular cortex. J Comp Neurol 270:39 –54. Motokizawa F, Ino Y (1983) A non-thalamic olfactory pathway to the orbital gyrus in the cat. Brain Res Bull 10:83–88. Nakashima M, Uemura M, Yasui K, Ozaki HS, Tabata S, Taen A (2000) An anterograde and retrograde tract-tracing study on the projections from the thalamic gustatory area in the rat: distribution of neurons projecting to the insular cortex and amygdaloid complex. Neurosci Res 36:297–309. Nemitz JW, Goldberg SJ (1983) Neuronal responses of rat pyriform cortex to odor stimulation: an extracellular and intracellular study. J Neurophysiol 49:188 –203. Norgren R (1995) Gustatory cortex. In: The rat nervous system (Paxinos G, ed), pp 751–771. Australia: Academic Press. Obaid AL, Salzberg BM (1996) Micromolar 4-aminopyridine enhances invasion of a vertebrate neurosecretory terminal arborization: optical recording of action potential propagation using an ultrafast photodiode-MOSFET camera and a photodiode array. J Gen Physiol 107:353–368. Ogawa H, Hasegawa K, Murayama N (1992) Difference in taste quality coding between two cortical taste areas, granular and dysgranular insular areas in rats. Exp Brain Res 91:415–424.
W. Fu et al. / Neuroscience 126 (2004) 1033–1041 Onoda N, Imamura K, Obata E, Iino M (1984) Response selectivity of neocortical neurons to specific odors in the rabbit. J Neurophysiol 52:638 –652. Paxinos G, Watson C (1986) The rat brain in stereotaxic coordinates. San Diego: Academic Press. Price JL (1987) The central olfactory and accessory olfactory systems. In: Neurobiology of taste and smell (Finger TE, Silver WL, eds), pp 179 –203. New York: John Wiley and Sons. Price JL (1995) Thalamus. In: The rat nervous system (Paxinos G, ed), pp 629 –648. Australia: Academic Press. Richards CD, Sercombe R (1970) Calcium, magnesium and the electrical activity of guinea-pig olfactory cortex in vitro. J Physiol 211: 571–584. Rolls ET (1989) Information processing in the taste system of primates. J Exp Biol 146:141–164. Saper CB (1995) Central autonomic system. In: The rat nervous system (Paxinos G, ed), pp 107–135. Australia: Academic Press. Schoenbaum G, Eichenbaum H (1995) Information coding in the rodent prefrontal cortex. I: Single-neuron activity in orbitofrontal cortex compared with that in pyriform cortex. J Neurophysiol 74:733– 750. Scott JW, Pfaffmann C (1972) Characteristics of responses of lateral hypothalamic neurons to stimulation of the olfactory system. Brain Res 48:251–264. Shi CJ, Cassell MD (1998) Cortical, thalamic, and amygdaloid connections of the anterior and posterior insular cortices. J Comp Neurol 399:440 –468. Shipley MT, Geinisman Y (1984) Anatomical evidence for conver-
1041
gence of olfactory, gustatory, and visceral afferent pathways in mouse cerebral cortex. Brain Res Bull 12:221–226. Sugai T, Sugitani M, Onoda N (1997) Subdivisions of the guinea-pig accessory olfactory bulb revealed by the combined method with immunohistochemistry, electrophysiological and optical recordings. Neuroscience 79:871–885. Sugitani M, Sugai T, Tanifuji M, Onoda N (1994) Signal propagation from piriform cortex to the endopiriform nucleus in vitro revealed by optical imaging. Neurosci Lett 171:175–178. Tanabe T, Iino M, Takagi SF (1975) Discrimination of odors in olfactory bulb, piriform-amygdaloid areas and orbitofrontal cortex of the monkey. J Neurophysiol 38:1284 –1296. Tanifuji M, Sugiyama T, Murase K (1994) Horizontal propagation of excitation in rat visual cortical slices revealed by optical imaging. Science 266:1057–1059. Wilson DA (1998) Habituation of odor responses in the rat anterior piriform cortex. J Neurophysiol 79:1425–1440. Wilson DA (2000) Comparison of odor receptive field plasticity in the rat olfactory bulb and anterior piriform cortex. J Neurophysiol 84: 3036 –3042. Yamamoto T, Yuyama N, Kato T, Kawamura Y (1985) Gustatory responses of cortical neurons in rats. II: Information processing of taste quality. J Neurophysiol 53:1356 –1369. Yarita H, Iino M, Tanabe T, Kogure S, Takagi SF (1980) A transthalamic olfactory pathway to the orbitofrontal cortex in the monkey. J Neurophysiol 43:69 –85. Yasui Y, Saper CB, Cechetto DF (1989) Calcitonin gene-related peptide immunoreactivity in the visceral sensory cortex, thalamus, and related pathways in the rat. J Comp Neurol 290:487–501.
(Accepted 18 March 2004) (Available online 2 June 2004)