- Email: [email protected]
Rebound depolarization in single units of the ventral cochlear nucleus: A contribution to grouping by common onset?
Neuroscience 154 (2008) 139 –146
REBOUND DEPOLARIZATION IN SINGLE UNITS OF THE VENTRAL COCHLEAR NUCLEUS: A CONTRIBUTION TO GROUPING BY COMMON ONSET? S. BLEECK,a N. J. INGHAM,b J. L. VERHEYc AND I. M. WINTERd*
the suppressive sidebands in its receptive field. The strength of the rebound was positively correlated with the strength of the suppression. These and other results are consistent with the view that low-level mechanisms underlie the psychophysical captor effect. © 2008 IBRO. Published by Elsevier Ltd. All rights reserved.
a
Institute of Sound and Vibration Research, University of Southampton, University Road, Southampton, SO17 1 BJ, UK
b
Wellcome Trust Sanger Institute, Wellcome Trust Genome Campus, Hinxton, CB10 1SA, UK
c
AG Neurosensorik, Institut für Physik, Carl von Ossietzky Universitat Oldenburg, D-26111 Oldenburg, Germany
Key words: wideband suppression, chopper unit, primarylike unit, onset unit, onset asynchrony, vowels.
d
The Centre for the Neural Basis of Hearing, The Physiological Laboratory, Downing Street, Cambridge, CB2 3EG, UK
Despite being continually bombarded with a mixture of sounds originating from a variety of sources the brain is remarkably good at segregating the complex waveform into objects that correspond to separate acoustic events. This segregation involves the sequential grouping of related events across time as well as the simultaneous grouping of acoustic elements across frequency. A key factor influencing simultaneous grouping is common onset time; components that start or stop roughly at the same time are judged as far more likely to have originated from the same source. A component that begins before the others makes a greatly reduced contribution to the timbre of a complex tone or to the phonetic quality of a vowel (e.g. Darwin, 1984). Taking advantage of the fact that it is possible to change the perception of the vowel /I/ to // by manipulating the harmonics around F1, Darwin and Sutherland (1984) suggested that the effect of onset asynchrony was not entirely attributable to neural adaptation. As shown schematically in Fig. 1, increasing the amplitude of one harmonic around F1 (in this case the fourth harmonic) would cause the perception of the vowel to change from /I/ to // (Fig. 1A and B). Extending the increased component before the vowel could, however, greatly reduce this change (Fig. 1C). The role of neural adaptation in this effect was questioned by repeating the experiment but this time using a ‘captor’ tone which was switched on with leading portion of the asynchronous harmonic and off when the vowel started (Fig. 1D). This time the vowel percept did change in a fashion analogous to the effect of an increase in the amplitude of the fourth harmonic. The explanation for this effect was that the captor had grouped with the asynchronous component enabling the remainder of the asynchronous component to be grouped with the remainder of the components. Although the site of action was not specified it was generally assumed that low-level mechanisms were not responsible for this effect. It is possible, however, that a low level explanation could still underlie this phenomenon. Specifically we proposed (Bleeck et al., 2005) that: (i) the reduced effect of
Abstract—Simultaneous grouping by common onset time is believed to be a powerful cue in auditory perception; components that start or stop roughly at the same time are judged as far more likely to have originated from the same source. Here we report a simple experiment designed to simulate a complex psychophysical paradigm first described by Darwin and Sutherland [(1984) Grouping frequency components of vowels. When is a harmonic not a harmonic? Quarterly J of Experimental Psychology: Hum Exp Psychol 36(A):193–208]. It is possible to change the perception of the vowel /I/ to // by manipulating the harmonics around the first formant (F1). Increasing the amplitude of one harmonic around F1 caused the perception of the vowel to change from /I/ to //. Extending the increased component before the vowel could, however, greatly reduce this change. The role of neural adaptation in this effect was questioned by repeating the experiment but this time using a ‘captor’ tone which was switched on with the asynchronous harmonic and off when the vowel started. This time the vowel percept did change in a fashion analogous to the effect of an increase in the amplitude of the fourth harmonic (which is close to F1). This effect was explained by assuming that the captor had grouped with the leading portion of the asynchronous component enabling the remainder of the asynchronous component to be grouped with the remainder of the components. We propose a relatively low-level neuronal explanation for this grouping effect: the captor reduces the neural response to the leading segment of the asynchronous component by activating acrossfrequency suppression, either from the cochlea, or acting via a wideband inhibitor in the ventral cochlear nucleus. The reduction in neural response results in a release from adaptation with the offset of the captor terminating the inhibition, such that the response to the continuation of that component is now enhanced. Using a simplified paradigm we show that both primary-like and chopper units in the ventral cochlear nucleus of the anesthetized guinea pig may show a rebound in excitation when a captor is positioned so as to stimulate *Corresponding author. Tel: ⫹44-1223-333812; fax: ⫹44-1223333840. E-mail address: [email protected] (I. M. Winter). Abbreviations: BF, best frequency; CS, sustained chopper; CT, transient chopper; OC, onset-chopper; PL, primary-like; PN, primary-like with a notch; PSTH, peri-stimulus time histogram; RFT, rise–fall time.
0306-4522/08$32.00⫹0.00 © 2008 IBRO. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.neuroscience.2008.03.020
139
140
S. Bleeck et al. / Neuroscience 154 (2008) 139 –146
Frequency
A)
Vowel /I/
B)
raised harmonic
leading and raised harmonic
raised harmonic
perception: / I/
Leading C) component
/ Ť/
Frequency
with captor
captor
/ I/
E) reference condition
units with different BFs
D)
/ Ť/ F)
time
captor condition
captor in inhibitory frequency band Leading segment
Leading segment
rebound
time
Onset of wideband signal Fig. 1. Schematic illustration of the perceptual paradigm used by Darwin and Sutherland (1984) to demonstrate the importance of onset asynchrony. (A) The harmonic structure of the vowel /i/. (B) Same structure as in A but now with the 4th harmonic raised in amplitude. The perception shifts to //. (C) The onset of the incremented fourth harmonic is now leading the rest of the harmonic complex and the perception of the vowel shifts back toward /i/. (D) An additional tone (captor) is added that begins with the leading 4th harmonic but stops at the onset of the harmonic structure. The captor was initially chosen to be one octave above the leading 4th. The perception of the complex shifted back toward //. The explanation given was that the second tone had grouped with the leading segment of the 4th harmonic leaving the complex to be the same as that shown in (B). (E) cartoon illustrating the possible physiological responses of four units, differing in BF, to the paradigms illustrated by C and (F) the same as in (E) but now for the paradigm shown in (D).
the asynchronous, leading segment, occurs because it has adapted before the other components begin; (ii) a captor that groups with the asynchronous, leading segment, reduces the neural response to that segment by activating across-frequency suppression either from the cochlea or acting via a wideband inhibitor in the ventral cochlear nucleus; (iii) the effect of the reduction in neural response results in a release from adaptation; (iv) the offset of the captor terminates the inhibition, such that the response to the continuation of that component is now enhanced; (v) because the captor offset occurs at the same time as the other components begin, all partials effectively have a common onset time, and so group together. Examining the proposal above, premise (i) is relatively uncontroversial. Premise (ii), however, has important implications. In the Darwin and Sutherland (1984) experiment the captor was harmonically related to the leading segment. Based on our hypotheses above we would predict that the relationship between the captor and leading segment need not be harmonic. This proposal has since been substantiated by Roberts and Holmes (2006, 2007) who have demonstrated that the captor does not have to be harmonically related to the leading segment. Premises (iii)
and (iv) are related in that the suppression produced by the captor will reduce the response to the leading segment and that the offset of the captor will be accompanied by an enhanced response to the continuation of the leading segment. It is this part of the proposal that we examine in this paper. Here we show that for units with wideband suppression the response to a ‘leading segment’ is reduced in the presence of a captor tone and that this suppression is accompanied by an enhancement at the offset of the captor (shown as the blue PSTH in Fig. 1E). Wideband suppression and rebound excitation were observed from both primary-like (PL) and chopper units in the ventral cochlear nucleus and the strength of suppression was positively correlated with the size of the rebound. The final part of the hypothesis (v) remains to be tested and may well occur at a higher level of the auditory system.
EXPERIMENTAL PROCEDURES The results in this paper come from the responses of single units recorded from the left ventral cochlear nucleus of anesthetized guinea pigs (Cavia porcellus) weighing between 365 and 500 g. Details of the neural recording and surgical procedures may be found in Bleeck et al. (2006). The experiments performed in this
# off spikes
study have conformed to international guidelines on the ethical use of animals and have been carried out under the terms and conditions of the project license issued by the United Kingdom Home Office to the fourth author. Every effort was made to minimize the number of animals used and their suffering. All stimuli were digitally synthesized in real-time with a PC equipped with a DIGI 9636 PCI card that was connected optically to an AD/DA converter (ADI-8 DS, RME Audio Products, Germany). The AD/DA converter was used for digital-to-analog conversion of the stimuli as well as for analog-to-digital conversion of the amplified (⫻1000) neural activity. The sample rate was 96 kHz. The AD/DA converter was driven using ASIO (Audio Streaming Input Output) and SDK (Software Developer Kit) from Steinberg (Lloyd, 2002). Following digital-to-analog conversion, the stimuli were equalized (phonic graphic equalizer, model EQ 3600; Apple Sound) to compensate for the speaker and coupler frequency response and fed into a power amplifier (Rotel RB971) and a programmable end attenuator (0 –75 dB in 5 dB steps, custom build) before being presented over a speaker (Radio Shack 30 –1777 tweeter assembled by Mike Ravicz, MIT, Cambridge, MA, USA) mounted in a coupler designed for the ear of a guinea pig. The stimuli were monitored acoustically using a condenser microphone (4134; Bruel & Kjaer, Naerum, Denmark) attached to a calibrated 1-mm diameter probe tube that was inserted into the speculum close to the eardrum. Upon isolation of a unit, its best frequency (BF) and threshold were determined using audiovisual criteria. Spontaneous activity was measured over a 10-s period. Single units were classified based on the shape of their peri-stimulus time histograms (PSTHs) according to conventional criteria (e.g. Young et al., 1988; Blackburn and Sachs, 1989). The PSTHs were generated from the spike times collected in response to 250 sweeps of a 50-ms tone burst 20 dB above the units BF threshold. Rise–fall time (RFT) of the tones was 1 ms (Cos2 gate) and the starting phase was randomized. Repetition rate was 4 Hz. For all units with BFs greater than 0.3 kHz PSTHs were classified as primary-like (PL), primary-like with a notch (PN), sustained chopper (CS) and transient chopper (CT). To examine the effects of onset asynchrony we simplified the paradigm described in the introduction (see Fig. 1E and 1F) by using just two signals; a suprathreshold ‘probe’ tone at the units BF and a second, ‘captor’ tone positioned in a suppressive sideband. In some cases captor tones were used at a variety of frequency/level combinations. We ran two different stimulus conditions: (i) a reference condition consisting of a 200 ms, 20 dB suprathreshold BF tone burst, (ii) a captor condition consisting of a 100-ms tone gated simultaneously with the probe tone onset. To determine the effect of the captor tone the strength of the rebound depolarization was compared against the strength of suppression. As illustrated in Fig. 2 for both the reference and captor conditions the number of spikes between 60 and 100 ms post-stimulus onset was counted (red-shaded area). The strength of the suppression was then quantified as the ratio of spike counts between reference and captor conditions. To estimate the strength of the rebound a similar ratio was calculated between 100 and 110 ms (blueshaded area). A receptive field was estimated for all units by measuring the response to single tones (50 ms duration; 5 ms RFT) of various frequency/intensity combinations. Frequency was varied from three octaves below BF to two octaves above BF in 0.1 octave steps. Sound level was varied in 5 dB steps from approximately 10 dB below BF threshold to 80 dB above BF threshold. When spontaneous discharge rate was low the receptive field was also measured in the presence of a low level (⬃5–10 dB suprathreshold) tone burst at BF to produce a surrogate spontaneous rate. This ‘tickle-tone’ paradigm could reveal the presence of suppressive sidebands. For 19 units we presented a suprathreshold, 50 ms probe tone at the unit’s BF gated simultaneously with a 25 ms
# of spikes
S. Bleeck et al. / Neuroscience 154 (2008) 139 –146
141
reference condition
captor condition suppression
rebound
captor probe 60 ms
100ms 110ms
Fig. 2. The calculation of the strength of suppression and rebound. The top figure (A) shows the response to a 200 ms tone at BF (reference condition) while the lower figure (B) illustrates the response to the same reference tone but in the presence of a ‘captor’ tone positioned in a suppressive sideband of the unit (captor condition). The strength of suppression was defined as the ratio of spikes in the reference condition to those in the captor condition measured over the region indicated by the red-shaded area. A similar measure was used to define the strength of the rebound effect. The time window is indicated by the blue-shaded area.
captor tone. The captor tone was positioned at a variety of frequency/intensity combinations and the effect of suppression and rebound strength was calculated.
RESULTS The results are based upon the recordings from 49 units in the ventral cochlear nucleus which demonstrated the presence of suppressive sidebands. Examples of the response from single units to the reference and captor conditions are shown in Fig. 3. For each case in the reference condition the PSTH shows clear adaptation. For the majority of cases the level of the reference tone was set at or close to the saturated rate for that unit. In the captor condition the presence of the captor clearly causes a reduction in the response to the reference condition alone followed by an enhancement of the response (effectively a second onset) at the captor offset. The captor level was always set at, or above, the level of the reference tone. An increase of at least 50% in the response to the BF tone at the termination of the captor was found in 23 units. Table 1 shows a summary of the effects of suppression and the amount of response enhancement as a function of unit type.
142
S. Bleeck et al. / Neuroscience 154 (2008) 139 –146
transient Chopper
sustained chopper
reference condition
reference condition
captor condition
captor condition
Primary like
Primary with notch
reference condition
reference condition
captor condition
captor condition
Time [ms]
Time [ms]
Fig. 3. Examples of the responses of four single unit types to the reference (top panel) and captor (bottom panel) conditions. All examples were chosen to illustrate a substantial effect in the captor condition. The CS unit had a BF of 9.2 kHz and a captor frequency of 3 kHz. The CT unit had a BF of 13.8 and a captor tone frequency of 6.6 kHz. The PN unit had a BF of 10.4 kHz and a captor frequency of 5.2 kHz. The PL unit had a BF of 7.15k Hz and a captor frequency of 8.84 kHz. Bin-width was 2 ms.
There was no significant difference in the strength of the rebound or the suppression strength between units (ANOVA F(3,45)⫽2.47, P⬎0.05 and F(3,45)⫽1.8). From these data we also observed a correlation (r2⫽0.25, P⬍0.05) between the strength of suppression and the strength of the rebound response. Such a result might not be surprising as we deliberately sought to position the captor tone in the measured suppressive sidebands of the single unit. No systematic attempt was made to place the captor tone in the region that gave the strongest suppression. Therefore, to investigate if the strength of the rebound activity could be more strongly correlated with the strength of the suppression, we varied the frequency and the amplitude of the captor tone systematically through the receptive field in 19 single units that we could record from for long enough. Fig. 4 shows an example of one such experiment. The response map was measured using the described ‘tickle-tone’ paradigm to reveal the otherwise invisible suppression. This neuron’s response map shows an area of strong inhibition above and below its BF. The probe tone was fixed at approximately 20 dB above the neuron’s threshold at BF (star in Fig. 4B). The captor tone was systematically changed over a range of 80 dB, two octaves below and one octave above BF, in five steps per octave (100 repeats). The green dots indicate the positions
of the captor tone. Example PSTHs are shown in the figure (the corresponding frequency and level are indicated by the green dots surrounded by black circles and highlighted by the arrows). At low sound levels the captor has no effect, the PSTH is the same as for the probe alone condition (not shown). When the captor is positioned in the suppressive sidebands (middle left and right), the response to the first half of the tone is smaller, and a rebound Table 1. The relationship between unit type, suppression strength and size of the rebound following captor termination Unit type
N
50% Rebound effect
Suppression strength (%)
Rebound strength (%)
CT CS PL PN Total
18 18 4 9 49
11 (61.1%) 5 (27.8%) 2 (50.0%) 5 (55.6%) 23 (46.9%)
38.3%⫾31.9 57.9%⫾21.6 44.0%⫾25.3 57.3%⫾33.3 49.4%⫾29.0
84.2%⫾76.7 31.5%⫾25.1 63.6%⫾39.8 66.5%⫾70.5 59.9%⫾61.4
There was no significant difference in the strength of the rebound or the suppression strength between units (ANOVA F(3,45)⫽2.47, P⬎0.05 and F(3,45)⫽1.8). From these data we also observed a correlation (r2⫽0.25, P⬍0.05) between the strength of suppression and the strength of the rebound response. Such a result might not be surprising as we deliberately sought to position the captor tone in the measured suppressive sidebands of the single unit. No systematic attempt was made to place the captor tone in the region that gave the strongest suppression.
S. Bleeck et al. / Neuroscience 154 (2008) 139 –146
143
A.
Tickle tone
B. 0 10 20 30 40 50
0 10 20 30 40 50
0 10 20 30 40 50
0 10 20 30 40 50
Probe tone
C. 0 10 20 30 40 50
0 10 20 30 40 50
Fig. 4. (Caption overleaf).
144
S. Bleeck et al. / Neuroscience 154 (2008) 139 –146
is visible at the beginning of the second half. Positioning the captor in the region of strongest suppression there is no activity in the first half of the response and a clear rebound is visible in the second. When the captor tone is in the receptive field of the unit it is now excitatory and therefore acts as an amplifier of the response to the probe tone (top left). The unit shows a very strong response to the first half and instead of a rebound effect, a temporary suppression is visible at the beginning of the second half. From these data there appears to be a correlation between strength of suppression and the size of the rebound. This correlation is shown in Fig. 4C, and for this unit is reasonably strong (r2⫽0.79; P⬍0.001; one tailed). There was no difference in the strength of suppression for captor placement above or below BF. Analysis of all 19 units measured with the same paradigm reveals an average correlation between suppression strength and rebound strength of r2⫽0.41⫾0.22. This positive correlation was significant for 13/19 units.
DISCUSSION We have shown that neurons in the VCN show an enhanced response to a pure tone at BF at the termination of a captor tone which is positioned in a suppressive sideband. The strength of the suppression is correlated with the strength of the rebound activity. The experiments with systematic variation of the captor tone throughout the whole receptive field (and beyond) never revealed a frequency/level combination of the captor where we observed a rebound activity without prior suppression. Consistent with previous studies examining two-tone suppression at the level of the VCN, suppression was observed for PL, PN, CT and CS units and was observed most frequently in CT units and least in PL units (e.g. Blackburn and Sachs, 1989; Winter and Palmer, 1990; Spirou and Young, 1991; Rhode and Greenberg, 1994). The origin of the suppression and the rebound, as observed in single units in the VCN, is unclear. Two-tone suppression is already observed in the mechanical responses of the basilar membrane (e.g. Ruggero et al., 1992; Rhode and Cooper, 1993) and a rebound effect can be observed in the responses of the auditory nerve (e.g. Kiang et al., 1965; Sachs and Kiang, 1968; Schmeidt, 1982; Delgutte, 1990). This phenomenon can be quite pronounced; under some conditions the second tone (the suppressor) completely shuts down the neural activity evoked by the first tone (the probe). Psychophysically this could correspond to the suppressor rendering the probe inaudible: the suppressor is said to mask the probe. The
effects of two-tone suppression are clearly confusable with the effects of neural inhibition, however, for the functioning of the model suggested in this paper and described by Roberts and Holmes (2006, 2007) and Holmes and Roberts (2006), it is not essential if the origin of the suppression is neural or mechanical. It has been suggested that the bandwidth of two-tone suppression at the level of the cochlea and auditory nerve is insufficient to explain the captor effect but the bandwidth of lateral suppression in the cochlear nucleus can be considerably broader (Rhode and Greenberg, 1994). In addition the effects of captor tones presented to the contralateral ear (see below) cannot be explained at the level of the auditory nerve. It is for these reasons that we suggest that the responses of units in the VCN may be important. A neural processing scheme based on the responses of two unit types from the VCN that might plausibly account for the captor effect observed psychophysically, first proposed by Bleeck et al. (2005) and subsequently elaborated by Roberts and Holmes (2006), is shown in Fig. 5. We suggest onset-chopper (OC) units act as wideband inhibitors (e.g. Winter and Palmer, 1995). The bandwidth of the observed suppression regions is compatible with the typical bandwidth of OC units (Winter and Palmer 1995; Jiang et al., 1996). We suggest CT units receive both narrowband excitatory input from the cochlea and wideband inhibitory input from OC units, and these connections are assumed to be between cells that share a similar BF (Ferragamo et al., 1998; Pressnitzer et al., 2001). In this scheme the excitatory response of CT cells to the leading probe tone will be reduced by inhibitory input from OC units with the same BF and the addition of a remote captor tone will further reduce the response of the CT units to the leading portion as the captor also falls within the wide receptive field of the OC units. The offset of the captor, as the vowel begins, will produce a transient rebound in the excitatory response through release from inhibition. We postulate that this rebound effect is due to cell intrinsic rebound depolarization. In brain slices of the deep cerebellar nucleus Aizenman and Linden (1999) found that a rebound depolarization was elicited after the offset of a hyperpolarizing stimulus leading to a Na⫹ spike burst in neurons. This mechanism could also be responsible for the rebound effect described here. As discussed in the introduction our suggestion that the effect of the captor is mediated by wideband suppression is at odds with the original suggestion that the captor and leading segment grouped on the basis of common harmonicity. There is no evidence that wideband suppres-
Fig. 4. (A) Receptive field from a CT unit measured with a ‘tickle’ tone to reveal suppressive sidebands (outlined by the blue lines). The excitatory area is bounded by the red line. The BF was approximately 8 kHz and threshold was 91 dB attenuation. (B) An outline of the excitatory and inhibitory response areas of the same response area shown in (A). The green dots indicate the frequency/intensity position of a captor tone which was presented simultaneously with a probe tone (position indicated by the red star). The resulting post-stimulus time histograms for six captor positions are shown to the side of the receptive field. The red line beneath the PSTHs indicates the duration of the captor tone; the blue line indicates the duration of the probe. Note that when the captor was positioned in the suppressive sidebands there was a reduction in the response to the probe tone and also a rebound in the response at the captor offset. (C) The correlation between suppression strength and rebound strength for the single unit shown in Fig. 5. The values when the captor tone was positioned below BF are indicated by open circles and when positioned above are indicated by crosses.
S. Bleeck et al. / Neuroscience 154 (2008) 139 –146
145
Fig. 5. Schematic of the proposed neural circuit in the ventral cochlear nucleus that underlies the captor effect as observed in this study. The input to the units in the ventral cochlear nucleus is represented by two PL PSTHs positioned at the captor and probe tones. A wideband inhibitor is represented by an OC PSTH pattern. The output unit is represented by the CT unit. Above the PSTHs are outlines of the receptive fields of each unit. The auditory nerve fiber input is relatively narrowband while the OC unit is broadband. The CT unit which receives inhibitory input from the OC unit as well as excitatory input from the auditory nerve is characterized by inhibitory sidebands. Note that the BF of the OC and CT unit is at the probe tone frequency.
sion is greatest when there is an octave relationship between the probe and suppressor. This idea was tested, using psychophysical methods in humans, by Roberts and Holmes (2006) who varied the frequency of a synchronous-onset captor between 900 and 2250 Hz when the probe tone was at 500 Hz. They found that captor efficacy was dependent on frequency proximity rather than harmonic relationship and had an upper cutoff around 1500 Hz. Captor efficacy was also found to be dependent upon captor sound level; the lower the captor level the less effect it has (Roberts et al., 2007). Replacing the captor tone with a noise-band captor does not reduce captor efficacy. These results are all compatible with the idea of wideband suppression playing a role in segregation by onset asynchrony, however, other observations are more problematic for the wideband suppression hypothesis. For instance, a captor tone may end as much as 80 ms before onset of the vowel and still be just as effective (Holmes and Roberts, 2006). The rebound in excitation in our data does not persist for as long as 80 ms but it could be that the general level of the response in the captor channel is still augmented relative to the non-captor condition. Of course, other stages of processing may also be involved. Roberts and Holmes (2007) have reported that the captor is just as effective when presented to the contralateral ear as when played to the ipsilateral ear. This result rules out the possibility of two-tone suppression at the level of the cochlea
but may be compatible with what is known about binaural responsiveness in the mammalian cochlear nucleus (e.g. Joris and Smith, 1998; Shore et al., 2003; Ingham et al., 2006).
CONCLUSION The findings are consistent with an account of the psychophysical captor effect in terms of the proposed VCN scheme and are in general agreement with our original proposal: a captor tone reduces the neural response to the asynchronous component and creates a rebound effect when placed in a suppressive sideband. The long decay time for the captor effect after captor offset is, however, inconsistent with our current knowledge of VCN single unit responses but it is clear that the existence of wideband suppression at the level of the VCN could play a role in processes previously believed to occur at higher levels of the auditory pathway. Auditory grouping mechanisms have been divided into (i) primitive grouping, which involves the use of general properties of sound to determine the source of acoustic events and (ii) schema-based grouping, which is generally learned and therefore dependent on the listener’s specific experience (Bregman, 1990). Neural adaptation and wideband suppression are, we suggest, two important mechanisms for primitive grouping at the level of the cochlea and cochlear nucleus.
146
S. Bleeck et al. / Neuroscience 154 (2008) 139 –146
Acknowledgments—Supported by the Biotechnology and Biological Sciences Research Council (BBSRC) UK. We thank Professor Brian Roberts for helpful comments on the manuscript.
REFERENCES Aizenman CD, Linden DJ (1999) Regulation of the rebound depolarization and spontaneous firing patterns of deep nuclear neurons in slices of rat cerebellum. J Neurophysiol 82:1697–1709. Blackburn CC, Sachs MB (1989) Classification of unit types in the anteroventral cochlear nucleus: PST histograms and regularity analysis. J Neurophysiol 62:1303–1329. Bleeck S, Ingham NJ, Verhey JL, Winter IM (2005) Wideband suppression in the cochlear nucleus: a role in grouping by common onset? Abstracts of the 28th midwinter meeting of the Association for Research in Otolaryngology. Vol. 28:236. Bleeck S, Sayles M, Ingham NJ, Winter IM (2006) The time course of recovery from suppression and facilitation from single units in the mammalian cochlear nucleus. Hear Res 212:176 –184. Bregman AS (1990) Auditory scene analysis. Cambridge: MIT Press. Darwin CJ (1984) Perceiving vowels in the presence of another sound: constraints on formant perception. J Acoust Soc Am 76:1636 – 1647. Darwin CJ, Sutherland NS (1984) Grouping frequency components of vowels. When is a harmonic not a harmonic? Q J Exp Psychol Hum Exp Psychol 36(A):193–208. Delgutte B (1990) Two-tone rate suppression in auditory-nerve fibers: dependence on suppressor frequency and level. Hear Res 49: 225–246. Ferragamo MJ, Golding NL, Oertel D (1998) Synaptic inputs to stellate cells in the ventral cochlear nucleus. J Neurophysiol 79:51– 63. Holmes SD, Roberts B (2006) Inihibitory influences on asynchrony as a cue for auditory segregation. J Exp Psych Hum Percept Perform 32:1231–1242. Ingham NJ, Bleeck S, Winter IM (2006) Contralateral inhibitory and excitatory frequency response maps in the mammalian cochlear nucleus. Eur J Neurosci 24:2515–2529. Jiang D, Palmer AR, Winter IM (1996) Frequency extent of two-tone facilitation in onset units in the ventral cochlear nucleus. J Neurophysiol 75:380 –395. Joris PX, Smith PH (1998) Temporal and binaural properties in dorsal cochlear nucleus and its output tract. J Neurosci 18:10157–10170.
Kiang NYS, Watanabe T, Thomas EC, Clark LF (1965) Discharge patterns of single fibers in the cat’s auditory nerve. MIT Research Monograph number 35. Cambridge, MA: MIT Press. Lloyd MI (2002) A fast inexpensive stimulus presentation and data aquisition system for auditory neuroscience. Int J Audiol 41:263. Pressnitzer D, Meddis R, Delahaye R, Winter IM (2001) Physiological correlates of comodulation masking release in the mammalian ventral cochlear nucleus. J Neurosci 21:6377– 6386. Rhode WS, Cooper NP (1993) Two-tone suppression and distortion production on the basilar membrane in the hook region of cat and guinea pig cochleae. Hear Res 66:31– 45. Rhode WS, Greenberg S (1994) Lateral suppression and inhibition in the cochlear nucleus of the cat. J Neurophysiol 71:493–514. Roberts B, Holmes SD (2006) Asynchrony and the grouping of vowel components: captor tones revisited. J Acoust Soc Am 119: 2905–2918. Roberts B, Holmes SD (2007) Contralateral influences of wideband inhibition on the effect of onset asynchrony as a cue for auditory grouping. J Acoust Soc Am 121:3655–3665. Roberts B, Holmes SD, Bleeck S, Winter IM (2007) Wideband inhibition modulates the effect of onset asynchrony as a grouping cue. In: Hearing: from sensory processing to perception (Kollmeier B et al., eds), pp 333–341. Berlin: Springer. Ruggero MA, Robles L, Rich NC (1992) Two-tone suppression in the basilar membrane of the cochlea: mechanical basis of auditory nerve rate suppression. J Neurophysiol 68:1087–1099. Sachs MB, Kiang NYS (1968) Two-tone inhibition in auditory-nerve fibers. J Acoust Soc Am 43:1120 –1128. Schmeidt RA (1982) Boundaries of two-tone rate suppression of cochlear nerve activity. Hear Res 7:335–351. Shore SE, Sumner CJ, Bledsoe SC, Lu J (2003) Effects of contralateral sound stimulation on unit activity of ventral cochlear nucleus neurons. Exp Brain Res 153:427– 435. Spirou GA, Young ED (1991) Organization of dorsal cochlear nucleus type-IV unit response maps and their relationship to activation by band-limited noise. J Neurophysiol 66:1750 –1768. Winter IM, Palmer AR (1990) Responses of single units in the anteroventral cochlear nucleus of the guinea pig. Hear Res 44:161–178. Winter IM, Palmer AR (1995) Level dependence of cochlear nucleus onset unit responses and facilitation by second tones or broadband noise. J Neurophysiol 73:141–159. Young ED, Robert JM, Shofner WP (1988) Regularity and latency of units in ventral cochlear nucleus: implications for unit classification and generation of response properties. J Neurophysiol 60:1–29.
(Accepted 10 March 2008) (Available online 20 March 2008)
REBOUND DEPOLARIZATION IN SINGLE UNITS OF THE VENTRAL COCHLEAR NUCLEUS: A CONTRIBUTION TO GROUPING BY COMMON ONSET? S. BLEECK,a N. J. INGHAM,b J. L. VERHEYc AND I. M. WINTERd*
the suppressive sidebands in its receptive field. The strength of the rebound was positively correlated with the strength of the suppression. These and other results are consistent with the view that low-level mechanisms underlie the psychophysical captor effect. © 2008 IBRO. Published by Elsevier Ltd. All rights reserved.
a
Institute of Sound and Vibration Research, University of Southampton, University Road, Southampton, SO17 1 BJ, UK
b
Wellcome Trust Sanger Institute, Wellcome Trust Genome Campus, Hinxton, CB10 1SA, UK
c
AG Neurosensorik, Institut für Physik, Carl von Ossietzky Universitat Oldenburg, D-26111 Oldenburg, Germany
Key words: wideband suppression, chopper unit, primarylike unit, onset unit, onset asynchrony, vowels.
d
The Centre for the Neural Basis of Hearing, The Physiological Laboratory, Downing Street, Cambridge, CB2 3EG, UK
Despite being continually bombarded with a mixture of sounds originating from a variety of sources the brain is remarkably good at segregating the complex waveform into objects that correspond to separate acoustic events. This segregation involves the sequential grouping of related events across time as well as the simultaneous grouping of acoustic elements across frequency. A key factor influencing simultaneous grouping is common onset time; components that start or stop roughly at the same time are judged as far more likely to have originated from the same source. A component that begins before the others makes a greatly reduced contribution to the timbre of a complex tone or to the phonetic quality of a vowel (e.g. Darwin, 1984). Taking advantage of the fact that it is possible to change the perception of the vowel /I/ to // by manipulating the harmonics around F1, Darwin and Sutherland (1984) suggested that the effect of onset asynchrony was not entirely attributable to neural adaptation. As shown schematically in Fig. 1, increasing the amplitude of one harmonic around F1 (in this case the fourth harmonic) would cause the perception of the vowel to change from /I/ to // (Fig. 1A and B). Extending the increased component before the vowel could, however, greatly reduce this change (Fig. 1C). The role of neural adaptation in this effect was questioned by repeating the experiment but this time using a ‘captor’ tone which was switched on with leading portion of the asynchronous harmonic and off when the vowel started (Fig. 1D). This time the vowel percept did change in a fashion analogous to the effect of an increase in the amplitude of the fourth harmonic. The explanation for this effect was that the captor had grouped with the asynchronous component enabling the remainder of the asynchronous component to be grouped with the remainder of the components. Although the site of action was not specified it was generally assumed that low-level mechanisms were not responsible for this effect. It is possible, however, that a low level explanation could still underlie this phenomenon. Specifically we proposed (Bleeck et al., 2005) that: (i) the reduced effect of
Abstract—Simultaneous grouping by common onset time is believed to be a powerful cue in auditory perception; components that start or stop roughly at the same time are judged as far more likely to have originated from the same source. Here we report a simple experiment designed to simulate a complex psychophysical paradigm first described by Darwin and Sutherland [(1984) Grouping frequency components of vowels. When is a harmonic not a harmonic? Quarterly J of Experimental Psychology: Hum Exp Psychol 36(A):193–208]. It is possible to change the perception of the vowel /I/ to // by manipulating the harmonics around the first formant (F1). Increasing the amplitude of one harmonic around F1 caused the perception of the vowel to change from /I/ to //. Extending the increased component before the vowel could, however, greatly reduce this change. The role of neural adaptation in this effect was questioned by repeating the experiment but this time using a ‘captor’ tone which was switched on with the asynchronous harmonic and off when the vowel started. This time the vowel percept did change in a fashion analogous to the effect of an increase in the amplitude of the fourth harmonic (which is close to F1). This effect was explained by assuming that the captor had grouped with the leading portion of the asynchronous component enabling the remainder of the asynchronous component to be grouped with the remainder of the components. We propose a relatively low-level neuronal explanation for this grouping effect: the captor reduces the neural response to the leading segment of the asynchronous component by activating acrossfrequency suppression, either from the cochlea, or acting via a wideband inhibitor in the ventral cochlear nucleus. The reduction in neural response results in a release from adaptation with the offset of the captor terminating the inhibition, such that the response to the continuation of that component is now enhanced. Using a simplified paradigm we show that both primary-like and chopper units in the ventral cochlear nucleus of the anesthetized guinea pig may show a rebound in excitation when a captor is positioned so as to stimulate *Corresponding author. Tel: ⫹44-1223-333812; fax: ⫹44-1223333840. E-mail address: [email protected] (I. M. Winter). Abbreviations: BF, best frequency; CS, sustained chopper; CT, transient chopper; OC, onset-chopper; PL, primary-like; PN, primary-like with a notch; PSTH, peri-stimulus time histogram; RFT, rise–fall time.
0306-4522/08$32.00⫹0.00 © 2008 IBRO. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.neuroscience.2008.03.020
139
140
S. Bleeck et al. / Neuroscience 154 (2008) 139 –146
Frequency
A)
Vowel /I/
B)
raised harmonic
leading and raised harmonic
raised harmonic
perception: / I/
Leading C) component
/ Ť/
Frequency
with captor
captor
/ I/
E) reference condition
units with different BFs
D)
/ Ť/ F)
time
captor condition
captor in inhibitory frequency band Leading segment
Leading segment
rebound
time
Onset of wideband signal Fig. 1. Schematic illustration of the perceptual paradigm used by Darwin and Sutherland (1984) to demonstrate the importance of onset asynchrony. (A) The harmonic structure of the vowel /i/. (B) Same structure as in A but now with the 4th harmonic raised in amplitude. The perception shifts to //. (C) The onset of the incremented fourth harmonic is now leading the rest of the harmonic complex and the perception of the vowel shifts back toward /i/. (D) An additional tone (captor) is added that begins with the leading 4th harmonic but stops at the onset of the harmonic structure. The captor was initially chosen to be one octave above the leading 4th. The perception of the complex shifted back toward //. The explanation given was that the second tone had grouped with the leading segment of the 4th harmonic leaving the complex to be the same as that shown in (B). (E) cartoon illustrating the possible physiological responses of four units, differing in BF, to the paradigms illustrated by C and (F) the same as in (E) but now for the paradigm shown in (D).
the asynchronous, leading segment, occurs because it has adapted before the other components begin; (ii) a captor that groups with the asynchronous, leading segment, reduces the neural response to that segment by activating across-frequency suppression either from the cochlea or acting via a wideband inhibitor in the ventral cochlear nucleus; (iii) the effect of the reduction in neural response results in a release from adaptation; (iv) the offset of the captor terminates the inhibition, such that the response to the continuation of that component is now enhanced; (v) because the captor offset occurs at the same time as the other components begin, all partials effectively have a common onset time, and so group together. Examining the proposal above, premise (i) is relatively uncontroversial. Premise (ii), however, has important implications. In the Darwin and Sutherland (1984) experiment the captor was harmonically related to the leading segment. Based on our hypotheses above we would predict that the relationship between the captor and leading segment need not be harmonic. This proposal has since been substantiated by Roberts and Holmes (2006, 2007) who have demonstrated that the captor does not have to be harmonically related to the leading segment. Premises (iii)
and (iv) are related in that the suppression produced by the captor will reduce the response to the leading segment and that the offset of the captor will be accompanied by an enhanced response to the continuation of the leading segment. It is this part of the proposal that we examine in this paper. Here we show that for units with wideband suppression the response to a ‘leading segment’ is reduced in the presence of a captor tone and that this suppression is accompanied by an enhancement at the offset of the captor (shown as the blue PSTH in Fig. 1E). Wideband suppression and rebound excitation were observed from both primary-like (PL) and chopper units in the ventral cochlear nucleus and the strength of suppression was positively correlated with the size of the rebound. The final part of the hypothesis (v) remains to be tested and may well occur at a higher level of the auditory system.
EXPERIMENTAL PROCEDURES The results in this paper come from the responses of single units recorded from the left ventral cochlear nucleus of anesthetized guinea pigs (Cavia porcellus) weighing between 365 and 500 g. Details of the neural recording and surgical procedures may be found in Bleeck et al. (2006). The experiments performed in this
# off spikes
study have conformed to international guidelines on the ethical use of animals and have been carried out under the terms and conditions of the project license issued by the United Kingdom Home Office to the fourth author. Every effort was made to minimize the number of animals used and their suffering. All stimuli were digitally synthesized in real-time with a PC equipped with a DIGI 9636 PCI card that was connected optically to an AD/DA converter (ADI-8 DS, RME Audio Products, Germany). The AD/DA converter was used for digital-to-analog conversion of the stimuli as well as for analog-to-digital conversion of the amplified (⫻1000) neural activity. The sample rate was 96 kHz. The AD/DA converter was driven using ASIO (Audio Streaming Input Output) and SDK (Software Developer Kit) from Steinberg (Lloyd, 2002). Following digital-to-analog conversion, the stimuli were equalized (phonic graphic equalizer, model EQ 3600; Apple Sound) to compensate for the speaker and coupler frequency response and fed into a power amplifier (Rotel RB971) and a programmable end attenuator (0 –75 dB in 5 dB steps, custom build) before being presented over a speaker (Radio Shack 30 –1777 tweeter assembled by Mike Ravicz, MIT, Cambridge, MA, USA) mounted in a coupler designed for the ear of a guinea pig. The stimuli were monitored acoustically using a condenser microphone (4134; Bruel & Kjaer, Naerum, Denmark) attached to a calibrated 1-mm diameter probe tube that was inserted into the speculum close to the eardrum. Upon isolation of a unit, its best frequency (BF) and threshold were determined using audiovisual criteria. Spontaneous activity was measured over a 10-s period. Single units were classified based on the shape of their peri-stimulus time histograms (PSTHs) according to conventional criteria (e.g. Young et al., 1988; Blackburn and Sachs, 1989). The PSTHs were generated from the spike times collected in response to 250 sweeps of a 50-ms tone burst 20 dB above the units BF threshold. Rise–fall time (RFT) of the tones was 1 ms (Cos2 gate) and the starting phase was randomized. Repetition rate was 4 Hz. For all units with BFs greater than 0.3 kHz PSTHs were classified as primary-like (PL), primary-like with a notch (PN), sustained chopper (CS) and transient chopper (CT). To examine the effects of onset asynchrony we simplified the paradigm described in the introduction (see Fig. 1E and 1F) by using just two signals; a suprathreshold ‘probe’ tone at the units BF and a second, ‘captor’ tone positioned in a suppressive sideband. In some cases captor tones were used at a variety of frequency/level combinations. We ran two different stimulus conditions: (i) a reference condition consisting of a 200 ms, 20 dB suprathreshold BF tone burst, (ii) a captor condition consisting of a 100-ms tone gated simultaneously with the probe tone onset. To determine the effect of the captor tone the strength of the rebound depolarization was compared against the strength of suppression. As illustrated in Fig. 2 for both the reference and captor conditions the number of spikes between 60 and 100 ms post-stimulus onset was counted (red-shaded area). The strength of the suppression was then quantified as the ratio of spike counts between reference and captor conditions. To estimate the strength of the rebound a similar ratio was calculated between 100 and 110 ms (blueshaded area). A receptive field was estimated for all units by measuring the response to single tones (50 ms duration; 5 ms RFT) of various frequency/intensity combinations. Frequency was varied from three octaves below BF to two octaves above BF in 0.1 octave steps. Sound level was varied in 5 dB steps from approximately 10 dB below BF threshold to 80 dB above BF threshold. When spontaneous discharge rate was low the receptive field was also measured in the presence of a low level (⬃5–10 dB suprathreshold) tone burst at BF to produce a surrogate spontaneous rate. This ‘tickle-tone’ paradigm could reveal the presence of suppressive sidebands. For 19 units we presented a suprathreshold, 50 ms probe tone at the unit’s BF gated simultaneously with a 25 ms
# of spikes
S. Bleeck et al. / Neuroscience 154 (2008) 139 –146
141
reference condition
captor condition suppression
rebound
captor probe 60 ms
100ms 110ms
Fig. 2. The calculation of the strength of suppression and rebound. The top figure (A) shows the response to a 200 ms tone at BF (reference condition) while the lower figure (B) illustrates the response to the same reference tone but in the presence of a ‘captor’ tone positioned in a suppressive sideband of the unit (captor condition). The strength of suppression was defined as the ratio of spikes in the reference condition to those in the captor condition measured over the region indicated by the red-shaded area. A similar measure was used to define the strength of the rebound effect. The time window is indicated by the blue-shaded area.
captor tone. The captor tone was positioned at a variety of frequency/intensity combinations and the effect of suppression and rebound strength was calculated.
RESULTS The results are based upon the recordings from 49 units in the ventral cochlear nucleus which demonstrated the presence of suppressive sidebands. Examples of the response from single units to the reference and captor conditions are shown in Fig. 3. For each case in the reference condition the PSTH shows clear adaptation. For the majority of cases the level of the reference tone was set at or close to the saturated rate for that unit. In the captor condition the presence of the captor clearly causes a reduction in the response to the reference condition alone followed by an enhancement of the response (effectively a second onset) at the captor offset. The captor level was always set at, or above, the level of the reference tone. An increase of at least 50% in the response to the BF tone at the termination of the captor was found in 23 units. Table 1 shows a summary of the effects of suppression and the amount of response enhancement as a function of unit type.
142
S. Bleeck et al. / Neuroscience 154 (2008) 139 –146
transient Chopper
sustained chopper
reference condition
reference condition
captor condition
captor condition
Primary like
Primary with notch
reference condition
reference condition
captor condition
captor condition
Time [ms]
Time [ms]
Fig. 3. Examples of the responses of four single unit types to the reference (top panel) and captor (bottom panel) conditions. All examples were chosen to illustrate a substantial effect in the captor condition. The CS unit had a BF of 9.2 kHz and a captor frequency of 3 kHz. The CT unit had a BF of 13.8 and a captor tone frequency of 6.6 kHz. The PN unit had a BF of 10.4 kHz and a captor frequency of 5.2 kHz. The PL unit had a BF of 7.15k Hz and a captor frequency of 8.84 kHz. Bin-width was 2 ms.
There was no significant difference in the strength of the rebound or the suppression strength between units (ANOVA F(3,45)⫽2.47, P⬎0.05 and F(3,45)⫽1.8). From these data we also observed a correlation (r2⫽0.25, P⬍0.05) between the strength of suppression and the strength of the rebound response. Such a result might not be surprising as we deliberately sought to position the captor tone in the measured suppressive sidebands of the single unit. No systematic attempt was made to place the captor tone in the region that gave the strongest suppression. Therefore, to investigate if the strength of the rebound activity could be more strongly correlated with the strength of the suppression, we varied the frequency and the amplitude of the captor tone systematically through the receptive field in 19 single units that we could record from for long enough. Fig. 4 shows an example of one such experiment. The response map was measured using the described ‘tickle-tone’ paradigm to reveal the otherwise invisible suppression. This neuron’s response map shows an area of strong inhibition above and below its BF. The probe tone was fixed at approximately 20 dB above the neuron’s threshold at BF (star in Fig. 4B). The captor tone was systematically changed over a range of 80 dB, two octaves below and one octave above BF, in five steps per octave (100 repeats). The green dots indicate the positions
of the captor tone. Example PSTHs are shown in the figure (the corresponding frequency and level are indicated by the green dots surrounded by black circles and highlighted by the arrows). At low sound levels the captor has no effect, the PSTH is the same as for the probe alone condition (not shown). When the captor is positioned in the suppressive sidebands (middle left and right), the response to the first half of the tone is smaller, and a rebound Table 1. The relationship between unit type, suppression strength and size of the rebound following captor termination Unit type
N
50% Rebound effect
Suppression strength (%)
Rebound strength (%)
CT CS PL PN Total
18 18 4 9 49
11 (61.1%) 5 (27.8%) 2 (50.0%) 5 (55.6%) 23 (46.9%)
38.3%⫾31.9 57.9%⫾21.6 44.0%⫾25.3 57.3%⫾33.3 49.4%⫾29.0
84.2%⫾76.7 31.5%⫾25.1 63.6%⫾39.8 66.5%⫾70.5 59.9%⫾61.4
There was no significant difference in the strength of the rebound or the suppression strength between units (ANOVA F(3,45)⫽2.47, P⬎0.05 and F(3,45)⫽1.8). From these data we also observed a correlation (r2⫽0.25, P⬍0.05) between the strength of suppression and the strength of the rebound response. Such a result might not be surprising as we deliberately sought to position the captor tone in the measured suppressive sidebands of the single unit. No systematic attempt was made to place the captor tone in the region that gave the strongest suppression.
S. Bleeck et al. / Neuroscience 154 (2008) 139 –146
143
A.
Tickle tone
B. 0 10 20 30 40 50
0 10 20 30 40 50
0 10 20 30 40 50
0 10 20 30 40 50
Probe tone
C. 0 10 20 30 40 50
0 10 20 30 40 50
Fig. 4. (Caption overleaf).
144
S. Bleeck et al. / Neuroscience 154 (2008) 139 –146
is visible at the beginning of the second half. Positioning the captor in the region of strongest suppression there is no activity in the first half of the response and a clear rebound is visible in the second. When the captor tone is in the receptive field of the unit it is now excitatory and therefore acts as an amplifier of the response to the probe tone (top left). The unit shows a very strong response to the first half and instead of a rebound effect, a temporary suppression is visible at the beginning of the second half. From these data there appears to be a correlation between strength of suppression and the size of the rebound. This correlation is shown in Fig. 4C, and for this unit is reasonably strong (r2⫽0.79; P⬍0.001; one tailed). There was no difference in the strength of suppression for captor placement above or below BF. Analysis of all 19 units measured with the same paradigm reveals an average correlation between suppression strength and rebound strength of r2⫽0.41⫾0.22. This positive correlation was significant for 13/19 units.
DISCUSSION We have shown that neurons in the VCN show an enhanced response to a pure tone at BF at the termination of a captor tone which is positioned in a suppressive sideband. The strength of the suppression is correlated with the strength of the rebound activity. The experiments with systematic variation of the captor tone throughout the whole receptive field (and beyond) never revealed a frequency/level combination of the captor where we observed a rebound activity without prior suppression. Consistent with previous studies examining two-tone suppression at the level of the VCN, suppression was observed for PL, PN, CT and CS units and was observed most frequently in CT units and least in PL units (e.g. Blackburn and Sachs, 1989; Winter and Palmer, 1990; Spirou and Young, 1991; Rhode and Greenberg, 1994). The origin of the suppression and the rebound, as observed in single units in the VCN, is unclear. Two-tone suppression is already observed in the mechanical responses of the basilar membrane (e.g. Ruggero et al., 1992; Rhode and Cooper, 1993) and a rebound effect can be observed in the responses of the auditory nerve (e.g. Kiang et al., 1965; Sachs and Kiang, 1968; Schmeidt, 1982; Delgutte, 1990). This phenomenon can be quite pronounced; under some conditions the second tone (the suppressor) completely shuts down the neural activity evoked by the first tone (the probe). Psychophysically this could correspond to the suppressor rendering the probe inaudible: the suppressor is said to mask the probe. The
effects of two-tone suppression are clearly confusable with the effects of neural inhibition, however, for the functioning of the model suggested in this paper and described by Roberts and Holmes (2006, 2007) and Holmes and Roberts (2006), it is not essential if the origin of the suppression is neural or mechanical. It has been suggested that the bandwidth of two-tone suppression at the level of the cochlea and auditory nerve is insufficient to explain the captor effect but the bandwidth of lateral suppression in the cochlear nucleus can be considerably broader (Rhode and Greenberg, 1994). In addition the effects of captor tones presented to the contralateral ear (see below) cannot be explained at the level of the auditory nerve. It is for these reasons that we suggest that the responses of units in the VCN may be important. A neural processing scheme based on the responses of two unit types from the VCN that might plausibly account for the captor effect observed psychophysically, first proposed by Bleeck et al. (2005) and subsequently elaborated by Roberts and Holmes (2006), is shown in Fig. 5. We suggest onset-chopper (OC) units act as wideband inhibitors (e.g. Winter and Palmer, 1995). The bandwidth of the observed suppression regions is compatible with the typical bandwidth of OC units (Winter and Palmer 1995; Jiang et al., 1996). We suggest CT units receive both narrowband excitatory input from the cochlea and wideband inhibitory input from OC units, and these connections are assumed to be between cells that share a similar BF (Ferragamo et al., 1998; Pressnitzer et al., 2001). In this scheme the excitatory response of CT cells to the leading probe tone will be reduced by inhibitory input from OC units with the same BF and the addition of a remote captor tone will further reduce the response of the CT units to the leading portion as the captor also falls within the wide receptive field of the OC units. The offset of the captor, as the vowel begins, will produce a transient rebound in the excitatory response through release from inhibition. We postulate that this rebound effect is due to cell intrinsic rebound depolarization. In brain slices of the deep cerebellar nucleus Aizenman and Linden (1999) found that a rebound depolarization was elicited after the offset of a hyperpolarizing stimulus leading to a Na⫹ spike burst in neurons. This mechanism could also be responsible for the rebound effect described here. As discussed in the introduction our suggestion that the effect of the captor is mediated by wideband suppression is at odds with the original suggestion that the captor and leading segment grouped on the basis of common harmonicity. There is no evidence that wideband suppres-
Fig. 4. (A) Receptive field from a CT unit measured with a ‘tickle’ tone to reveal suppressive sidebands (outlined by the blue lines). The excitatory area is bounded by the red line. The BF was approximately 8 kHz and threshold was 91 dB attenuation. (B) An outline of the excitatory and inhibitory response areas of the same response area shown in (A). The green dots indicate the frequency/intensity position of a captor tone which was presented simultaneously with a probe tone (position indicated by the red star). The resulting post-stimulus time histograms for six captor positions are shown to the side of the receptive field. The red line beneath the PSTHs indicates the duration of the captor tone; the blue line indicates the duration of the probe. Note that when the captor was positioned in the suppressive sidebands there was a reduction in the response to the probe tone and also a rebound in the response at the captor offset. (C) The correlation between suppression strength and rebound strength for the single unit shown in Fig. 5. The values when the captor tone was positioned below BF are indicated by open circles and when positioned above are indicated by crosses.
S. Bleeck et al. / Neuroscience 154 (2008) 139 –146
145
Fig. 5. Schematic of the proposed neural circuit in the ventral cochlear nucleus that underlies the captor effect as observed in this study. The input to the units in the ventral cochlear nucleus is represented by two PL PSTHs positioned at the captor and probe tones. A wideband inhibitor is represented by an OC PSTH pattern. The output unit is represented by the CT unit. Above the PSTHs are outlines of the receptive fields of each unit. The auditory nerve fiber input is relatively narrowband while the OC unit is broadband. The CT unit which receives inhibitory input from the OC unit as well as excitatory input from the auditory nerve is characterized by inhibitory sidebands. Note that the BF of the OC and CT unit is at the probe tone frequency.
sion is greatest when there is an octave relationship between the probe and suppressor. This idea was tested, using psychophysical methods in humans, by Roberts and Holmes (2006) who varied the frequency of a synchronous-onset captor between 900 and 2250 Hz when the probe tone was at 500 Hz. They found that captor efficacy was dependent on frequency proximity rather than harmonic relationship and had an upper cutoff around 1500 Hz. Captor efficacy was also found to be dependent upon captor sound level; the lower the captor level the less effect it has (Roberts et al., 2007). Replacing the captor tone with a noise-band captor does not reduce captor efficacy. These results are all compatible with the idea of wideband suppression playing a role in segregation by onset asynchrony, however, other observations are more problematic for the wideband suppression hypothesis. For instance, a captor tone may end as much as 80 ms before onset of the vowel and still be just as effective (Holmes and Roberts, 2006). The rebound in excitation in our data does not persist for as long as 80 ms but it could be that the general level of the response in the captor channel is still augmented relative to the non-captor condition. Of course, other stages of processing may also be involved. Roberts and Holmes (2007) have reported that the captor is just as effective when presented to the contralateral ear as when played to the ipsilateral ear. This result rules out the possibility of two-tone suppression at the level of the cochlea
but may be compatible with what is known about binaural responsiveness in the mammalian cochlear nucleus (e.g. Joris and Smith, 1998; Shore et al., 2003; Ingham et al., 2006).
CONCLUSION The findings are consistent with an account of the psychophysical captor effect in terms of the proposed VCN scheme and are in general agreement with our original proposal: a captor tone reduces the neural response to the asynchronous component and creates a rebound effect when placed in a suppressive sideband. The long decay time for the captor effect after captor offset is, however, inconsistent with our current knowledge of VCN single unit responses but it is clear that the existence of wideband suppression at the level of the VCN could play a role in processes previously believed to occur at higher levels of the auditory pathway. Auditory grouping mechanisms have been divided into (i) primitive grouping, which involves the use of general properties of sound to determine the source of acoustic events and (ii) schema-based grouping, which is generally learned and therefore dependent on the listener’s specific experience (Bregman, 1990). Neural adaptation and wideband suppression are, we suggest, two important mechanisms for primitive grouping at the level of the cochlea and cochlear nucleus.
146
S. Bleeck et al. / Neuroscience 154 (2008) 139 –146
Acknowledgments—Supported by the Biotechnology and Biological Sciences Research Council (BBSRC) UK. We thank Professor Brian Roberts for helpful comments on the manuscript.
REFERENCES Aizenman CD, Linden DJ (1999) Regulation of the rebound depolarization and spontaneous firing patterns of deep nuclear neurons in slices of rat cerebellum. J Neurophysiol 82:1697–1709. Blackburn CC, Sachs MB (1989) Classification of unit types in the anteroventral cochlear nucleus: PST histograms and regularity analysis. J Neurophysiol 62:1303–1329. Bleeck S, Ingham NJ, Verhey JL, Winter IM (2005) Wideband suppression in the cochlear nucleus: a role in grouping by common onset? Abstracts of the 28th midwinter meeting of the Association for Research in Otolaryngology. Vol. 28:236. Bleeck S, Sayles M, Ingham NJ, Winter IM (2006) The time course of recovery from suppression and facilitation from single units in the mammalian cochlear nucleus. Hear Res 212:176 –184. Bregman AS (1990) Auditory scene analysis. Cambridge: MIT Press. Darwin CJ (1984) Perceiving vowels in the presence of another sound: constraints on formant perception. J Acoust Soc Am 76:1636 – 1647. Darwin CJ, Sutherland NS (1984) Grouping frequency components of vowels. When is a harmonic not a harmonic? Q J Exp Psychol Hum Exp Psychol 36(A):193–208. Delgutte B (1990) Two-tone rate suppression in auditory-nerve fibers: dependence on suppressor frequency and level. Hear Res 49: 225–246. Ferragamo MJ, Golding NL, Oertel D (1998) Synaptic inputs to stellate cells in the ventral cochlear nucleus. J Neurophysiol 79:51– 63. Holmes SD, Roberts B (2006) Inihibitory influences on asynchrony as a cue for auditory segregation. J Exp Psych Hum Percept Perform 32:1231–1242. Ingham NJ, Bleeck S, Winter IM (2006) Contralateral inhibitory and excitatory frequency response maps in the mammalian cochlear nucleus. Eur J Neurosci 24:2515–2529. Jiang D, Palmer AR, Winter IM (1996) Frequency extent of two-tone facilitation in onset units in the ventral cochlear nucleus. J Neurophysiol 75:380 –395. Joris PX, Smith PH (1998) Temporal and binaural properties in dorsal cochlear nucleus and its output tract. J Neurosci 18:10157–10170.
Kiang NYS, Watanabe T, Thomas EC, Clark LF (1965) Discharge patterns of single fibers in the cat’s auditory nerve. MIT Research Monograph number 35. Cambridge, MA: MIT Press. Lloyd MI (2002) A fast inexpensive stimulus presentation and data aquisition system for auditory neuroscience. Int J Audiol 41:263. Pressnitzer D, Meddis R, Delahaye R, Winter IM (2001) Physiological correlates of comodulation masking release in the mammalian ventral cochlear nucleus. J Neurosci 21:6377– 6386. Rhode WS, Cooper NP (1993) Two-tone suppression and distortion production on the basilar membrane in the hook region of cat and guinea pig cochleae. Hear Res 66:31– 45. Rhode WS, Greenberg S (1994) Lateral suppression and inhibition in the cochlear nucleus of the cat. J Neurophysiol 71:493–514. Roberts B, Holmes SD (2006) Asynchrony and the grouping of vowel components: captor tones revisited. J Acoust Soc Am 119: 2905–2918. Roberts B, Holmes SD (2007) Contralateral influences of wideband inhibition on the effect of onset asynchrony as a cue for auditory grouping. J Acoust Soc Am 121:3655–3665. Roberts B, Holmes SD, Bleeck S, Winter IM (2007) Wideband inhibition modulates the effect of onset asynchrony as a grouping cue. In: Hearing: from sensory processing to perception (Kollmeier B et al., eds), pp 333–341. Berlin: Springer. Ruggero MA, Robles L, Rich NC (1992) Two-tone suppression in the basilar membrane of the cochlea: mechanical basis of auditory nerve rate suppression. J Neurophysiol 68:1087–1099. Sachs MB, Kiang NYS (1968) Two-tone inhibition in auditory-nerve fibers. J Acoust Soc Am 43:1120 –1128. Schmeidt RA (1982) Boundaries of two-tone rate suppression of cochlear nerve activity. Hear Res 7:335–351. Shore SE, Sumner CJ, Bledsoe SC, Lu J (2003) Effects of contralateral sound stimulation on unit activity of ventral cochlear nucleus neurons. Exp Brain Res 153:427– 435. Spirou GA, Young ED (1991) Organization of dorsal cochlear nucleus type-IV unit response maps and their relationship to activation by band-limited noise. J Neurophysiol 66:1750 –1768. Winter IM, Palmer AR (1990) Responses of single units in the anteroventral cochlear nucleus of the guinea pig. Hear Res 44:161–178. Winter IM, Palmer AR (1995) Level dependence of cochlear nucleus onset unit responses and facilitation by second tones or broadband noise. J Neurophysiol 73:141–159. Young ED, Robert JM, Shofner WP (1988) Regularity and latency of units in ventral cochlear nucleus: implications for unit classification and generation of response properties. J Neurophysiol 60:1–29.
(Accepted 10 March 2008) (Available online 20 March 2008)