The projection from auditory cortex to cochlear nucleus in guinea pigs: an in vivo anatomical and in vitro electrophysiological study
- 123 Downloads
Previous anatomical experiments have demonstrated the existence of a direct, bilateral projection from the auditory cortex (AC) to the cochlear nucleus (CN). However, the precise relationship between the origin of the projection in the AC and the distribution of axon terminals in the CN is not known. Moreover, the influence of this projection on CN principal cells has not been studied before. The aim of the present study was two-fold. First, to extend the anatomical data by tracing anterogradely the distribution of cortical axons in the CN by means of restricted injections of biotinylated dextran amine (BDA) in physiologically characterized sites in the AC. Second, in an in vitro isolated whole brain preparation (IWB), to assess the effect of electrical stimulation of the AC on CN principal cells from which intracellular recordings were derived. BDA injections in the tonotopically organized primary auditory cortex and dorsocaudal auditory field at high and low best frequency (BF) sites resulted in a consistent axonal labeling in the ipsilateral CN of all injected animals. In addition, fewer labeled terminals were observed in the contralateral CN, but only in the animals subjected to injections in low BF region. The axon terminal fields consisting of boutons en passant or terminaux were found in the superficial granule cell layer and, to a smaller extent, in the three CN subdivisions. No axonal labeling was seen in the CN as result of BDA injection in the secondary auditory area (dorsocaudal belt). In the IWB, the effects of ipsilateral AC stimulation were tested in a population of 52 intracellulary recorded and stained CN principal neurons, distributed in the three CN subdivisions. Stimulation of the AC evoked slow late excitatory postsynaptic potentials (EPSPs) in only two cells located in the dorsal CN. The EPSPs were induced in a giant and a pyramidal cell at latencies of 20 ms and 33 ms, respectively, suggesting involvement of polysynaptic circuits. These findings are consistent with anatomical data showing sparse projections from the AC to the CN and indicate a limited modulatory action of the AC on CN principal cells.
KeywordsBiotinylated dextran amine (BDA) Intracellular recording Intracellular staining Corticobulbar axons
The cochlear nucleus (CN) occupies a strategic position in the auditory system. Being the first relay in the auditory pathways, it receives stereotyped inputs from the auditory nerve (AN), modifies this information, and then sends output signals to multiple nuclei of the brainstem and midbrain. In addition, it receives reciprocal, descending auditory inputs which constitute a substantial proportion of connections in the auditory system (Huffman and Henson 1990; Romand and Avan 1997; Rouiller 1997). The sources of the descending auditory projections to the CN have been shown to arise from the contralateral CN (Cant and Gaston 1982; Wenthold 1987; Shore et al. 1992; Schofield and Cant 1996), the superior olivary complex (Adams 1983; Spangler et al. 1987; Brown et al. 1988; Schofield and Cant 1999), the dorsal nucleus of the lateral lemniscus (DLL) (Caicedo and Herbert 1993), and the inferior colliculus (IC) (Van Noort 1969; Carey and Webster 1971; Adams and Warr 1976; Kane and Finn 1977; Andersen et al. 1980; Conlee and Kane 1982; Hashikawa 1983; Faye-Lund 1986, 1988; Coleman and Clerici 1987; Herbert et al. 1991; Malmierca et al. 1996; Schofield 2001). The auditory cortex (AC) is the most remote auditory region giving rise to descending auditory projections to the CN. For a long time, it was believed that the AC might influence the CN only via indirect pathways involving the midbrain. However, recent anatomical experiments (Feliciano et al. 1995; Weedman and Ryugo 1996a, 1996b; Saldaña et al. 1996; Schofield et al. 2001; Doucet et al. 2002) have demonstrated the existence of direct, bilateral projection from the AC to the CN. It has been shown that projections from AC to CN in the rat and the guinea pig arise from layer V neurons of the AC and terminate mainly in the granule cell domain of the CN (Weedman and Ryugo 1996a, 1996b; Schofield et al. 2001). However, the precise relationship between the origin of the projection in the AC and the distribution of axon terminals in the CN remains largely unknown. Although the existence of descending auditory inputs to CN suggests a possible mechanism for cortical feedback control of afferent information at peripheral levels, the functional nature of the influence exerted by direct descending projections from the AC to CN remains unknown.
Therefore, the aim of the present study was two-fold. The first aim was to extend the available anatomical data, by establishing the distribution of the corticobulbar axons in the CN. The second objective was to assess the effects of activating the descending projections from the AC on CN neurons. Two types of experiments were conducted in the guinea pig to address these questions. First, the distribution of corticobulbar axons in the CN was established neuroanatomically, based on restricted injections of the anterograde tracer biotinylated dextran amine (BDA) in physiologically characterized sites in the AC. Second, an in vitro isolated whole brain (IWB) preparation was used to perform intracellular recording and staining of single neurons in the CN in order to test the presence and the physiological properties of synaptic responses to electrical stimulation of the AC. Since, in the IWB preparation, it was impossible to record intracellularly from small neurons, the effects of AC stimulation could be assessed only in CN principal cells.
Materials and methods
In vivo experiments
BDA injections into the AC were performed in five adult guinea pigs (body weight 530–850 g). Under general anesthesia (diazepam 8 mg/kg, followed 15 min later by pentobarbital 20 mg/kg; see Evans 1979), a craniotomy (about 5×8 mm) was made in the temporal bone. Small punctures were made in the dura under microscopy control to insert low impedance glass pipettes (<3 MΩ) that allowed multiunit recordings (3–5 neurons) at a depth of 500–1000 µm. The frequency tuning at each recording site was tested between 20 and 80 dB SPL using a previously described sound delivery system (Manunta and Edeline 1997, 1999). Briefly, ascending sequences of pure tones (frequency 0.1–35 kHz, 100 ms duration, 1 s intertone interval) were delivered to the contralateral ear. Online displays of the peri-stimulus time histograms allowed determination of the frequency tuning at each intensity. A regular progression of the best frequency (BF) from low to high when moving from rostral to caudal was interpreted as a location in the primary auditory cortex (AI), whereas a regular progression of the BF from high to low was assumed to be typical of the dorsocaudal field (DC) (Redies et al. 1989; Wallace et al. 2000). Locations in secondary auditory fields (DCB dorsocaudal belt, VCB ventrocaudal belt; Wallace et al. 2000) were characterized by (1) long-latency responses, (2) very broad tuning, and (3) a lack of tonotopy. At a previously characterized locus, a pipette (tip 25–50 µm) filled with BDA (10%; Molecular Probes, Eugene, OR, USA) was lowered 1500 µm below the pial surface and the tracer was iontophoretically ejected for 45 min (8 µA positive current, 7 s "on/off"). After a survival time of 7–10 days, the animals were deeply anesthetized (pentobarbital, 120 mg/kg) and perfused with a 4% paraformaldehyde (PFA) solution in 0.1 M phosphate buffer (PB, pH 7.4). After 24 h postfixation, the brain was placed in a solution of 30% sucrose in 0.1 M PB until it sank. Serial coronal sections (40 µm) were prepared with a freezing microtome and collected in 0.05 M Tris (pH 7.6). BDA labeling was vizualized using previously described protocols (see for details Bajo et al. 1999). Individual sections were examined under brightfield illumination (×4, ×10, ×20, and ×40), and reconstruction of the corticobulbar axons in the CN was made using a drawing tube.
The parcellation of the CN and the corresponding nomenclature used in the present study were derived from the original description of the CN in the guinea pig by Hackney et al. (1990) in Nissl material. Two main divisions of the CN are clearly distinguishable. Caudally, the dorsal cochlear nucleus (DCN) has been subdivided in four layers. Rostrally, the ventral cochlear nucleus (VCN) contains two distinct regions, the anteroventral (AVCN) and posteroventral (PVCN) subdivisions separated by the cochlear nerve root. AVCN comprises mainly the spherical cell area ("sca"), whereas PVCN includes mainly the globular and octopus cell areas ("gca" and "oca", respectively). In addition, several territories formed by small neurons surround the CN. They consist of the superficial granular layer ("sgrl") covering the lateral and dorsal surface of VCN, the granular lamina ("lam") separating the DCN and PVCN, and the small cell cap region ("cap") located medially between the dorsal edge of PVCN and the sgrl. For simplification, lam and sgrl were grouped in the present report and referred to as the granule cell domain (GCD).
In vitro experiments
The physiological experiments were performed on guinea pigs of both genders, weighing 200–600 g. The methods used to isolate the brain and maintain it in vitro were similar to those described previously (Babalian et al. 1997, 1999, 2002a). Briefly, the animals were deeply anesthetized with pentobarbital (150–300 mg/kg i.p.) and perfused through the ascending aorta with a cold (10–13°C) Ringer's solution for 5–6 min. Then the animal was decapitated and the skull bones covering the brain were removed. The exposed brain was transected rostrally at the level of olfactory bulbs and caudally at the cervical segments C1–C2. The isolation of the brain from the skull was completed by gradual cutting of cranial nerves and arteries on the ventral surface of the brain. The brain was transferred to the incubation/recording chamber filled with saline at 13°C, and secured in the ventral side up position. The brain was cannulated through one of the vertebral arteries with a metallic tube (external diameter 0.4 mm) that was connected to the perfusion system. Following elimination of all major leaks from the brain's arterial system by ligating the arteries severed during the dissection (second vertebral artery, carotids, hypophyseal and labyrinthine arteries), the temperature of the chamber and perfusate was gradually increased to 29°C. The perfusion rate was progressively increased during rewarming to reach 4.5–5 ml/min. The perfusion solution had the following composition (in mM): NaCl 126; NaHCO3 26; MgSO4 1.3; KH2PO4 1.2; KCl 3; CaCl2 2.4; glucose 15; dextran 1.5% (MW 70000, Macrodex 70; Braun Medical AG, Sempach, Switzerland). The solution was continuously bubbled with a mixture of 95% O2 and 5% CO2.
The auditory nerves and the ipsilateral auditory cortex were stimulated using bipolar metallic (stainless steel or tungsten) electrodes with interpolar distance of 0.5 and 3–4 mm for AN and AC stimulation, respectively. The electrodes were inserted in the stump of the AN and 1–1.5 mm below the surface of the AC, at a location where a field potential was present in response to AN stimulation. Intracellular recordings from CN neurons were made by micropipettes filled with a solution of 1–2% neurobiotin in 2 M potassium acetate (resistance 120–200 MΩ). At the end of the recording session, neurons were stained by passing positive currents of 0.5–1.5 nA for 5–30 min in a 200 ms on/off duty cycle. The recorded signals were displayed and stored using a computer-based data acquisition system (PowerLab; ADInstruments, Castle Hill, Australia). After each successful experiment, the brain was fixed with a solution of 4% PFA in 0.1 M PB delivered through the perfusion system for 30–45 min. The brain was then postfixed overnight and placed in 30% sucrose solution until further processing. Injected cells were visualized on coronal brain sections (100 μm) using standard ABC histochemistry with some modifications (Wan et al. 1992).
In vivo experiments
Properties of biotinylated dextran amine injection sites and presence of labeled axons in the cochlear nucleus (CN). AI Primary auditory cortex, DC dorsocaudal auditory field, DCB dorsocaudal belt
Location of injection
Best frequency (kHz)
Presence of labeled axon terminals
Concerning the trajectory of the BDA-labeled axons, these were seen to travel from AC in the most dorsal part of cerebral peduncle and reach the lateral lemniscus in its rostral corner. From the lateral lemniscus, AC fibers traveled in the dorsal acoustic stria and entered the DCN. In the DCN, they ran in layers III and IV, from dorsal to ventral and from medial to lateral. More rostrally, the majority of axons were seen in the GCD and small cell cap surrounding the CN, whereas few of them reached the PVCN. Very few BDA-labeled axons traveled in the trapezoid body.
In vitro experiments
The two cells had similar physiological properties, which are illustrated in Fig. 5C for the giant cell. Both cells responded to stimulation of the ipsilateral AN with a sequence of excitatory (EPSPs) and inhibitory postsynaptic potentials (IPSPs), and exhibited long latency EPSPs to stimulation of AC. Stimulation of the AC evoked an EPSP of 2.9 mV at a latency of 20 ms in the giant cell (Fig. 5C). For the pyramidal cell, the EPSP had an amplitude of 7 mV and a latency of 33 ms (not shown). These long latencies suggest that responses of CN neurons to AC stimulation were induced polysynaptically.
In order to ensure stimulation of adequate sites in the AC, field potentials induced by stimulation of auditory nerve were systematically recorded through the stimulating electrode in the AC. Figure 5D illustrates an example of responses in the AC to stimulation of the ipsilateral or contralateral AN in the same animal. Stimulation of auditory nerves evoked a monophasic field potential in the AC. The latency of the field potentials was in the range of 10–15 ms and their amplitudes varied from 0.1 to 0.5 mV across the experiments. These data confirm that the location of the electrical stimulations was in the AC.
The original contribution of the present study is to extend previous tracing data on the projection of the AC to the CN, by introducing physiological characterization of the AC injection sites, and an investigation of corticobulbar effects in the IWB preparation. The anterograde transport of BDA was used to localize the axon terminal fields formed by the direct projection from the AC to the CN. The corticobulbar axons terminate mainly in the GCD of the ipsilateral CN and, to a smaller extent, in the DCN and the PVCN. This result is consistent with previous retrograde and anterograde tracing studies in the rat and guinea pig showing direct projections from the AC to the CN (Weedman and Ryugo 1996a, 1996b; Schofield et al. 2001). We found some axon terminals in the contralateral CN but they were much less numerous than in the ipsilateral CN. This observation is also in accordance with a recent anterograde tracing study in the guinea pig demonstrating bilateral projections from the AC to the CN, with predominance of the ipsilateral ones (Schofield et al. 2001). A new finding of the present study is that contralateral AC projections were not observed in the two animals subjected to BDA injection in the high BF regions of AI or DC. This result is not due to a particularly small injection site (see Fig. 1B,D) and/or poor labeling (see Fig. 3). Nevertheless, the observed difference between projections originating from high versus low BF regions is based on a small number of animals (n=4) and needs further confirmation.
As mentioned above, in our physiological experiments using the IWB preparation, the effects produced by AC stimulation could be tested only in CN principal cells. The AC-induced responses observed in two DCN cells had very long latencies suggesting their polysynaptic origin. Thus it can be concluded that AC projections to the three CN subdivisions and GCD most likely do not establish direct contacts with somata or dendrites of CN principal cells. Moreover, even the polysynaptic effects were observed in a very low percentage of recorded CN principal neurons (4%). To some extent, this is consistent with the relatively sparse corticobulbar projection observed in AVCN, PVCN and DCN as a result of BDA injection in the AC. In contrast, a much stronger projection was seen in the GCD and small cell cap. The GCD consists of a continuous sheet of small cells that surrounds the magnocellular VCN and penetrates into the DCN as layer II (Mugnaini et al. 1980). These small cells project to some CN principal cells and other CN cells, which in turn can contact CN principal cells. Therefore, one may wonder why so few polysynaptic responses induced by AC stimulation were observed in the present study. One possible explanation could be that the stimulation of AC and/or position of stimulating electrodes in the AC were not optimal. This argument can be ruled out for at least two reasons. First, the position of the stimulating electrodes in the AC was confirmed by recording, through the same electrodes, of field potentials in response to AN stimulation. Second, a large interpolar distance of the stimulating electrodes and high stimulation currents suggest that a large area of AC and, consequently, a large number of projecting AC cells was activated in our experiments. The observation of rare AC-induced effects in the CN principal cells is probably not due to the impaired GCD and other polysynaptic circuits in the IWB preparation. Indeed, in the same set of experiments, a large proportion of CN principal cells responded to stimulation of other sources of inputs to the CN (e.g., dorsal column nuclei, and trigeminal nerve; see Babalian et al. 2001) known also to operate mainly through the GCD (Itoh et al. 1987; Wright and Ryugo 1996; Shore et al. 2000; Haenggeli et al. 2002; Ryugo et al. 2003). This comparison under the same experimental conditions thus supports the interpretation that the AC projection exerts indeed a limited influence on CN principal cells as compared to the other two mentioned sources of non-auditory inputs. The reasons for that are not known. One may, however, speculate that the density of projections and/or the reliability of synaptic transmission in the GCD and other polysynaptic circuits are lower for the projections originating from the AC. For example, AC stimulation might produce predominantly subthreshold modulation of the membrane potential in small interneurons, which would not result in any visible effects on CN principal cells. Possible failure of AC inputs to generate action potential in small relay cells would thus account for the rare observation of responses at the level of CN principal cells. This explanation somewhat contradicts our experiments using train stimulation of AC intended to produce temporal summation of synaptic responses and discharges in relay neurons between AC and CN principal cells. Yet it is possible that parameters of train stimulation were not optimal in our study. Thus, influences of AC on granule cells and other CN interneurons remain to be elucidated.
The polysynaptic nature of the observed excitatory responses in two DCN cells suggests an alternative possibility that these effects were due to activation of AC projections to the IC and/or superior olivary complex, which in turn project to the CN. The existence of these pathways is well documented (Herbert et al. 1991; Feliciano and Potashner 1995; Saldaña et al. 1996; Malmierca et al. 1996; Games and Winer 1998; Schofield and Cant 1999; Budinger et al. 2000; Schofield 2001; Coomes and Schofield 2001). However, this possibility seems to be unlikely for the following reasons. Our previous results in the IWB preparation indicate that electrical stimulation of IC and various auditory and non-auditory brainstem inputs to the CN produce predominantly inhibitory effects in the vast majority of CN cells (Babalian et al. 2002b; Jacomme et al. 2002). In contrast, the two responses observed in the present study were excitatory. Moreover, stimulation of the IC and brainstem sources of inputs to CN induces synaptic responses in CN cells, usually at relatively short latencies (below 10 ms). Even if we take into account propagation and synaptic delay time related to the additional activation of AC projections to IC and/or brainstem, the resulting values will be hardly compatible with response latencies in the present study (20 and 33 ms). Therefore, we think that the observed long-latency responses are mediated through the AC projections to CN and slow signal processing compatible with the granule cell circuits or other interneuronal circuits in the CN. The fact that AC stimulation effects were observed only in DCN cells may indicate that the AC preferentially influence, though weakly, output activity of the DCN. Such a conclusion would be consistent with our observation of higher density of AC projections in the DCN relative to other CN subdivisions.
In conclusion, the anatomical data show sparse projections from the AC to the three CN subdivisions and more dense projections to the GCD. Physiological results in the IWB preparation indicate the existence of a limited modulatory action of the AC on CN principal cells. Thus, in the guinea pig, the auditory cortex might modulate the CN activity essentially at the level of granule cells and CN interneurons.
This work was supported by Swiss National Science Foundation grants No. 31-55836.98 and 31-66731.01, and the National Center of Competence in Research (NCCR) "Neural plasticity and repair".
- Adams JC (1983) Cytology of periolivary cells and the organization of their projections in the cat. J Comp Neurol 215:275–289Google Scholar
- Babalian AL, Jacomme AV, Doucet JR, Ryugo DK, Rouiller EM (2001) Auditory and non-auditory inputs to the cochlear nucleus neurons: an in vitro whole brain study. ARO Abstr 24:197Google Scholar
- Babalian AL, Jacomme AV, Ryugo DK, Rouiller EM (2002b) Functional input from the inferior colliculus to cochlear nucleus neurons: an in vitro whole brain study. ARO Abstr 25:8Google Scholar
- Coleman JR, Clerici WJ (1987) Sources of projections to subdivisions of the inferior colliculus in the rat. J Comp Neurol 262:215–226Google Scholar
- Coomes DL, Schofield BR (2001) Cortical projections to the superior olivary complex contact cells that project to the cochlear nucleus in guinea pigs. ARO Abstr 24:45Google Scholar
- Doucet JR, Weedman DL, Ryugo DK (2002) The projections of auditory cortex to the cochlear nucleus. In: Proceedings of International Symposium: Central Auditory Processing. Integration of Auditory and Nonauditory Information, Monte Verità, Switzerland, 12-15 May 2002, Abstract P28Google Scholar
- Faye-Lund H (1988) Inferior colliculus and related descending pathways in rat. Upsala J Med Sci 93:1–17Google Scholar
- Feliciano M, Saldaña E, Mugnaini E (1995) Direct projection from the rat primary auditory neocortex to the nucleus sagulum, paralemniscal regions, superior olivary complex and cochlear nuclei. Aud Neurosci 1:287–308Google Scholar
- Haenggeli CA, Doucet JR, Ryugo DK (2002) Trigeminal projections to the cochlear nucleus in rats. ARO Abstr 25:7Google Scholar
- Herbert H, Aschoff A, Ostwald J (1991) Topography of projections from the auditory cortex to the inferior colliculus in the rat. J Comp Neurol 304:103–122Google Scholar
- Jacomme AV, Rouiller EM, Ryugo DK, Babalian AL (2002) Physiological inputs from the inferior colliculus to the cochlear nucleus in the guinea pig studied in in vitro whole brain preparation. FENS Abstr 1:A184.9Google Scholar
- Mugnaini E, Warr WB, Osen KK (1980) Distribution and light microscopic features of granule cells in the cochlear nuclei of cat, rat, and mouse. J Comp Neurol 581–606Google Scholar
- Romand R, Avan P (1997) Anatomical and functionnal aspects of the cochlear nucleus. In: Ehret G, Romand R (eds) The central auditory sytem. Oxford University Press, New York, pp 97–191Google Scholar
- Rouiller EM (1997) Functionnal organization of the auditory pathways. In: Ehret G, Romand R, (eds) The central auditory system. Oxford University Press, New York, pp 3–96Google Scholar
- Ryugo DK, Haenggeli C-A, Doucet JR (2003) Multimodal inputs to the granule cell domain of the cochlear nucleus. Exp Brain Res DOI: 10.1007/s00221-003-1605-3Google Scholar
- Saldaña E, Feliciano M, Mugnaini E (1996) Distribution of descending projections from primary auditory neocortex to inferior colliculus mimics the topography of intracollicular projections. J Comp Neurol 371:15–40Google Scholar
- Schofield BR, Coomes DL, Schofield R (2001) Projections from the auditory cortex to the cochlear nucleus in guinea pigs. Assoc Res Otolaryngol, Abs. 24:44Google Scholar
- Spangler KM, Cant NB, Henkel CK, Farley GR, Warr WB (1987) Descending projections from the superior olivary complex to the cochlear nucleus of the cat. J Comp Neurol. 259:452–465Google Scholar
- Van Noort J (1969) The structure and connections of the inferior colliculus. Van Gorcum, AssenGoogle Scholar