Tuning shifts of the auditory system by corticocortical and corticofugal projections and conditioning

Abstract

The central auditory system consists of the lemniscal and nonlemniscal systems. The thalamic lemniscal and nonlemniscal auditory nuclei are different from each other in response properties and neural connectivities. The cortical auditory areas receiving the projections from these thalamic nuclei interact with each other through corticocortical projections and project down to the subcortical auditory nuclei. This corticofugal (descending) system forms multiple feedback loops with the ascending system. The corticocortical and corticofugal projections modulate auditory signal processing and play an essential role in the plasticity of the auditory system. Focal electric stimulation - comparable to repetitive tonal stimulation - of the lemniscal system evokes three major types of changes in the physiological properties, such as the tuning to specific values of acoustic parameters of cortical and subcortical auditory neurons through different combinations of facilitation and inhibition. For such changes, a neuromodulator, acetylcholine, plays an essential role. Electric stimulation of the nonlemniscal system evokes changes in the lemniscal system that is different from those evoked by the lemniscal stimulation. Auditory signals ascending from the lemniscal and nonlemniscal thalamic nuclei to the cortical auditory areas appear to be selected or adjusted by a "differential" gating mechanism. Conditioning for associative learning and pseudo-conditioning for nonassociative learning respectively elicit tone-specific and nonspecific plastic changes. The lemniscal, corticofugal and cholinergic systems are involved in eliciting the former, but not the latter. The current article reviews the recent progress in the research of corticocortical and corticofugal modulations of the auditory system and its plasticity elicited by conditioning and pseudo-conditioning.

Fig. 1. Tuning shifts, the auditory pathway and auditory cortex

Focal electric stimulation (ES) of the lemniscal auditory system evokes facilitation, inhibition and tuning shifts in AI and subcortical auditory nuclei. There are two types of tuning shifts of “tuning-unmatched” neurons: centripetal (a) and centrifugal (b). (a) and (b) The unbroken and broken triangles represent the tuning curves in the control and shifted conditions, respectively. “Tuning-matched” neurons do not show tuning shifts, but facilitation of their auditory responses. Centripetal and centrifugal tuning shifts both have been found in the frequency and time domains. (c) The dorsolateral view of the brain of the mustached bat, Pteronotus parnellii parnellii (P.p.p.). The arrows indicate the ascending and descending (corticofugal) systems. (d) A neurophysiological map of the auditory cortex (AC) of the mustached bat. The numbers and lines in the anterior (AIa) and posterior (AIp) divisions and the Doppler-shifted constant frequency (DSCF) area in AI indicate iso-best-frequency lines. The CF/CF area sensitive to combinations of constant-frequency (CF) signals consists of two subdivisions that contain a Doppler-shift (velocity) axis. The frequency modulation-frequency modulation (*FF), dorsal fringe (DF) and ventral fringe (VF) areas are sensitive to combinations of frequency-modulated (FM) signals. Each area consists of three subdivisions. These areas contain an echo-delay (range) axis. CBL, cerebellum; CER, cerebrum; CN, cochlear nucleus; DIF, dorsal intrafossa area; DM, dorsomedial area; DP, dorsoposterior area; IC, inferior colliculus; MGB, medial geniculate body; NLL, nucleus of the lateral lemniscus; P.p.r., Pteronotus parnellii rubiginosus which is larger than P.p.p.; SOC, superior olivary complex; VA, ventroanterior area; VM, ventromedial area; VP, ventroposterior area (Suga and Ma 2003). Focal electric stimulation of the AIp, DF or VF area evokes centripetal tuning shifts, whereas that of the DSCF or FF area evokes centrifugal tuning shifts. [*The FF area had been called the FM-FM area because it consists of FM-FM combination-sensitive neurons. Both the DF and VF areas, subsequently found, also consist of FM-FM neurons. So, the FM-FM area is now called the FF area (Tang and Suga 2008).]

Fig. 2. Distribution of centripetal and centrifugal BF shifts in the primary auditory cortex (AI) evoked by focal electric stimulation of AI in the Mongolian gerbil, Meriones unguiculatus

Electric stimulation of 1.1-kHz-tuned neurons in AI evokes centripetal (a, circles) and centrifugal (b, triangles) BF shifts of other AI neurons within 1.0 mm. Locations of recorded neurons along the cortical surface are plotted relative to that of the stimulated neurons at the origin of the coordinates. x- and y-axes: directions parallel and orthogonal to the cochleotopic (tonotopic) axis of AI, respectively. “x” in b indicates a neuron that showed no BF shift. Data are pooled from 16 hemispheres of 11 animals. Confidence ellipses are shown for neurons that showed centripetal (a) or centrifugal (b) BF shifts. The amounts of BF shifts were measured in a zone parallel (1) or orthogonal (2 and 3) to the cochleotopic axis (c). The directions and amounts of BF shifts of neurons in the rostro-caudal (1 in c) and dorso-ventral (2 and 3 in c) zones are respectively plotted in d – f as a function of distance along the cortical surface. BF_e: BF of electrically stimulated AI neurons. BF_r: BF of recorded AI neurons. See the inset at the bottom of (c) for symbols (Sakai and Suga 2002).

Fig. 3. “BF-shift-difference” curves obtained from the primary auditory cortex (AI) or the central nucleus of the inferior colliculus (ICc) of four species of mammals

The BF shift changes as a function of the difference in BF between the recorded collicular (ICc_r, dashed curves) or cortical (AI_r, undashed curves) neurons and the electrically stimulated cortical neurons (AI_s). Each BF-shift-difference curve encompasses a scatter plot of BF shifts of many neurons studied (N) such as in Fig. 2d. Note the differences in the curves between species and between different areas of the same species. (a, b, d and e) Centripetal BF shifts, except where indicated by arrows. A prominent centripetal BF shift occurs at ~ 5 kHz higher than the stimulated cortical BF in the big brown bat, Eptesicus fuscus (a) and at ~ 1 kHz higher than that in the Mongolian gerbil, Meriones unguiculatus (d). By contrast, the prominent centripetal BF shift occurs at ~ 10 kHz lower than the stimulated cortical BF in the posterior division of AI (AIp) of the mustached bat, Pteronotus parnellii parnellii (b). In the house mouse, Mus domesticus, prominent centripetal BF shifts occur at ~ 9 kHz higher and lower than the stimulated cortical BF (e). (c) Prominent centrifugal BF shifts occur at ~ 0.5 kHz higher and lower than the stimulated cortical BF in the Doppler-shifted constant frequency (DSCF) area of the mustached bat. The shape of these BF-shift-difference curves might change with the mean BF of stimulated cortical neurons (AI_s). (f) Distance along the cochleotopic axis of AI from the stimulated cortical neurons. (g) The mean and standard deviation of the BFs of stimulated cortical neurons as well as references are shown for each figure, a – e. The characteristics of the electric stimulation (ES_a) were 0.2 ms, 100 nA pulses for a - d and 1 ms, 500 nA pulses for e (Suga and Ma 2003).

Fig. 4. Best frequency (BF) shifts in the specialized auditory system of the mustached bat

Changes in the direction of the BF shifts of a cortical (A) and a collicular (B) DSCF neuron evoked by an antagonist of inhibitory GABA-A receptors, bicuculline methiodide (BMI). A and B: Focal electric electrical stimulation (ES) of cortical DSCF neurons evokes centrifugal BF shifts (a), whereas BMI applied to the cortical DSCF neurons evokes centripetal BF shifts (b). The arrays of PST histograms display frequency-response curves of the recorded neurons. The vertical and horizontal arrows respectively indicate the BFs of cortical DSCF neurons receiving ES or BMI and centrifugal or centripetal BF shifts of the recorded neurons. 1 – 4: The PST histograms recorded before (control) and after ES or BMI applications. The amplitude of tone bursts was set at 10 dB above the minimum threshold of a given neuron. ES: 0.2-ms 100-nA electric pulses delivered at a rate of 5/s for 7.0 min; BMI: 1.0 nl of 5 mM (Xiao and Suga 2002b). The vertical scale bars indicate 50 impulses. (C) Distribution of centripetal and centrifugal BF shifts in the primary auditory cortex (AI), the central nucleus of the inferior colliculus (ICc) and the cochlea evoked by focal cortical electric stimulation (ES). (1) Dorsolateral view of the cerebral cortex. In AI, the Doppler-shifted constant frequency (DSCF) area is sandwiched between the anterior (AIa) and posterior (AIp) divisions of AI. (2) The DSCF area can be divided into dorsal and ventral divisions (DSCFd and DSCFv) in terms of the effect of cortical electric stimulation. (3) The ICc consists of the dorsoposterior (DPD), anterolateral (ALD) and medial (MD) divisions. The DPD can be divided into the dorsal (DPDd) and ventral (DPDv) portions in terms of the effect of cortical electric stimulation. (4) Cochlea where cochlear microphonic responses (CM) were recorded. Electric stimulation (ES) of DSCFd or DSCFv (2, right) evokes the changes in the BFs of DSCF and DPD neurons and CM. Centripetal and centrifugal BF shifts evoked by DSCFd stimulation are expressed by open and filled triangles, respectively, whereas those evoked by DSCFv stimulation are expressed by open and filled circles, respectively (Xiao and Suga 2005).

Fig. 5. Changes in delay tuning evoked by electric stimulation of cortical delay-tuned neurons through the lateral, contralateral, feedforward, feedback and corticofugal projections

(a) The centrifugal best delay (BDe) shift of a VF neuron tuned to a 3.0 ms echo delay evoked by electric stimulation of FF neurons tuned to a 7.0 ms delay through feedforward projection. The delay-response curves of the VF neuron were obtained before (control, open circles), 85 min after (filled circles) and 137 min after (open triangles) the onset of the FF stimulation. (b) and (c) A centripetal BDe shift of a BDe-unmatched FF neuron (b) and sharpening of the tuning of a BDe-matched FF neuron (c) evoked by electric stimulation of VF neurons via the feedback projection. The BDe’s of the recorded FF and stimulated VF neurons were 3.5 and 2.0 ms, respectively in (b), but both were 1.5 ms in (c). The downward arrows indicate the BDe’s of the stimulated neurons. The BDe-shift-difference curves for the BDe shifts evoked by electric stimulation of either the FF (d), DF (e) or VF (f) neurons. (d) “1–5” respectively show the curves for the BDe shifts elicited by the lateral, contralateral, feedforward to DF or VF and corticofugal projections from (or within) the FF area. cFF, contralateral FF. (e) “1–3” respectively show the curves for the BDe shifts elicited by the feedback, lateral and contralateral projections from (or within) the DF area. cDF, contralateral DF. (f) “1” shows the curve for the BDe shifts elicited by the feedback projection from the VF area to the FF area. (g) The block diagram showing the projections evoking the centripetal (arrow) or centrifugal (line with a short bar at its end) BDe shifts. The short dashed lines indicate either centripetal or centrifugal BDe shifts that are speculated (Tang and Suga 2009).

Fig. 6. Changes in tuning curve, best frequency and response magnitude of thalamic and collicular DSCF neurons evoked by focal activation or inactivation of cortical DSCF neurons in the mustached bat, Pteronotus parnellii parnellii

(a) and (b) Shifts in the frequency-tuning curves of two thalamic (MGBv) neurons evoked by an activation (a) or inactivation (b) of cortical neurons: activation by electric stimulation of a 0.2 ms, 100 nA electric pulse delivered at a rate of 5/s for 7 min (ES_a) and inactivation by 90 nl of 1.0% lidocaine (Lid.). The best frequencies (BFs) of the activated or inactivated cortical neurons are indicated by the arrows. The curves were measured before (control, open circles), during (closed circles), and after (recovery; dashed lines) the cortical activation or inactivation. The data points for the recovery are not shown because almost all of them overlapped with those for the control. (c) and (d) The BF shifts of thalamic (c) and collicular (d) neurons evoked by a focal activation (dashed lines) or inactivation (solid lines and filled circles) of cortical neurons. The abscissae represent the differences in BF between the stimulated cortical (AI) and recorded thalamic (MGBv) or collicular (ICc) neurons in the control condition. The abscissae are the same as those in (e) and (f). The BFs of the stimulated cortical neurons were 61.2 kHz on the average. The triangles and circles represent the data obtained from matched and unmatched subcortical neurons, respectively. The regression lines, their slopes ‘a’ and correlation coefficients ‘r’ are shown in the graphs. The BF shift is centrifugal for the cortical activation, but centripetal for the cortical inactivation. (e) and (f). The ordinates represent the percent change in the response magnitude (number of pulses per tone burst) of thalamic (e) and collicular (f) neurons evoked by the cortical activation. The triangles and circles, respectively, represent percent changes in the response magnitude of matched and unmatched subcortical neurons at the BFs of individual neurons in the control condition. To measure response magnitudes, tone bursts were set at the best amplitude of each neuron in the control condition. Changes in BF (c and d) and response magnitude (e and f) both are larger in the MGBv than in the ICc (Zhang and Suga 2000).

Fig. 7. Corticofugal modulation of collicular neurons

(a) and (b) Corticofugal modulation of duration-tuned collicular (ICc) neurons evoked by electrical stimulation of duration-tuned cortical (AI) neurons in the big brown bat, Eptesicus fuscus. The stimulated cortical (AI_s) and recorded collicular (ICc_r) neurons were matched (a) or unmatched (b) in both best frequency (BF) and best duration (BDu). The open and filled arrows indicate the BDu’s of ICc_r and AI_s neurons, respectively. Cortical stimulation sharpened (a) or shifted (b) duration-tuning. (c) and (d) Distributions of the BDu shifts (c) and width changes (d) in duration-tuning. The abcissae represent a BDu difference between the recorded and stimulated neurons. Each triangle in (c) represents a BDu shift. Each open or filled circle in (d) represents sharpening or broadening of a duration-tuning curve, respectively. Crosses mark neurons that showed neither BDu shift nor change in the width of a duration-tuning curve. The extent of change is linearly related to the BDu difference between ICc_r and AI_s neurons. The BDu’s of the stimulated neurons were 5.5 or 5.6 ms on the average (Ma and Suga 2001b). (e) Centripetal minimum-threshold (MT) shifts of BF-matched collicular neurons in the house mouse, Mus domesticus (Yan and Ehret 2002). (f) Centripetal best azimuth (BAZ) shifts of collicular neurons sensitive to the contralateral auditory fields in the big brown bat. Their BAZs shift toward the midline, i.e., toward the BAZs of the stimulated cortical neurons. L, lateral; M, medial (Zhou and Jen 2005). The MT and BAZ shifts are linearly related to the difference in MT and BAZ between the recorded collicular (ICc_r) and stimulated cortical (AI_s) neurons, respectively.

Fig. 8. Conditioning, pseudo-conditioning, frequency-tuning changes and heart rate change

(Aa) Paired conditioned (CS) and unconditioned (US) stimuli and the parameter values of the CS (tone bursts) and US (electric leg-shock) used for the experiments on the big brown bat. (Ab) Facilitation of the response and sharpening of the frequency-tuning of BF-matched neurons and BF shifts of BF-unmatched neurons. These changes are tone-specific changes. Here “BF-matched” means that the BF of a recorded neuron is the same as the frequency of the CS. (Ac) The bat shows a decrease in heart rate to the paired CS-US, i.e., conditioning (1). After the conditioning, the conditioned autonomic response (heart rate change) does not occur to 15-kHz (2) and 60-kHz (3) tone bursts, but to the 30-kHz tone bursts (CS) used for the conditioning (Ac by Ji and Suga 2007). When the US is unpaired with the CS by randomizing it (Ba), nonspecific augmentation (sensitization) of BF-matched and -unmatched neurons in the central auditory system is elicited (Bb). The bat shows a heart-rate decrease to the unpaired CS-US, i.e., pseudo-conditioning. After the pseudo-conditioning, it shows a heart rate decrease not only to the 30- kHz tone bursts used for the pseudo-conditioning, but also to 15-kHz and 60-kHz tone bursts (Bc by Ji and Suga 2008) (Suga et al. 2010).

Fig. 9. The collicular (ICc) and cortical (AI) BF shifts elicited by auditory fear conditioning or by electric stimulation of the auditory and/or somatosensory cortices or by long repetitive acoustic stimulation (big brown bat)

(a) Changes in the responses (left two columns) and frequency-response curves (right graph) of a collicular neuron caused by 30-min-long conditioning consisting of 60 pairs of a train of acoustic stimuli (AS_t = CS) and an electric leg-stimulation (ES_l = US). All of the data were obtained with tone bursts fixed at 10 dB above the minimum threshold of the neuron. The CS was 25 kHz, and the BF of the collicular neuron was 29 kHz. The data were obtained before (1, control), immediately after (2), 90 min after (3), and 125 min after the conditioning (4). BF_c and BF_s, BFs in the control and shifted conditions, respectively. BF_s shifted back to BF_c 125 min after the conditioning (Gao and Suga 1998). (b) Time courses of the BF shifts of collicular (1 and 3) and cortical (2 and 4) neurons evoked by electric stimulation of AI, ES_a (1 and 2) or the conditioning (3 and 4). A second conditioning session 3.5 h after the first also evoked collicular (5) and cortical (6) BF shifts. The horizontal bars indicate the electric stimulation or conditioning of 30 min duration. Each curve is the mean of 10 – 15 curves obtained from different neurons (Ma and Suga 2001a; Gao and Suga 2000). (c) and (d) Bilateral inactivation of the somatosensory cortex (SI) with 0.4 μg of muscimol applied to its surface abolishes development of the conditioning-dependent BF shifts of a collicular (c) and a cortical (d) neuron, but does not change their responses and frequency-response curves. Frequency-response curves were obtained before the conditioning (1, control); during SI inactivation (2); immediately after the conditioning under SI inactivation (3); and 75 or 180 min after the conditioning (4). The frequencies of the CS (AS_t) and the electric current of the US (ES_l) are listed in each graph (Gao and Suga 2000). (e) Collicular and cortical BF shifts evoked by a short train of electric stimuli of AI (ES_a) are augmented by electrical stimulation of the somatosensory cortex (ES_s). ES_s was delivered 1.0 s after ES_a, mimicking the conditioning. Curves 1 and 2 respectively represent the time courses of collicular and cortical BF shifts evoked by ES_a alone. Curves 3 and 4 respectively represent the time courses of collicular and cortical BF shifts evoked by ES_a followed by ES_s. These stimuli were delivered over 30 min (horizontal bars). (f) ES_s following a train of acoustic stimuli (AS_t) augments the collicular and cortical BF shifts evoked by AS_t. Curves 1 and 2 respectively represent the time courses of the collicular and cortical BF shifts evoked by AS_t. Curves 3 and 4 respectively represent the time courses of the collicular and cortical BF shifts evoked by AS_t followed by ES_s, mimicking the conditioning. Note that ES_s has a larger and longer augmenting effect on the cortical BF shift than on the collicular BF shift. Means and standard errors (vertical bars) are based on the data obtained from the number of neurons ranging between 12 and 20, as indicated by N (Ma and Suga 2003).

Fig. 10. Thalamo-cortical modulation

Changes in the frequency-threshold (1) and frequency-response (2) curves of AI neurons (lower graphs) evoked by electric stimulation of a sharply-tuned MGBv neuron (a, upper graphs) or broadly tuned MGBm neurons (b, upper graphs). Electric stimulation of the MGBv neurons evokes a BF shift, i.e., a shift of the frequency-threshold and frequency-response curves of the AI neuron (a, lower graphs), whereas that of the MGBm neurons evokes broadening of those curves of the AI neuron (b, lower graphs). The open and filled circles represent the curves in the control condition and 30 min after the onset of the electric stimulation, respectively. The dashed lines with dots represent the curves obtained 60 min after the onset of the electric stimulation. The frequency-response curves were measured at 30 dB above the minimum threshold of the given neuron (Ma and Suga 2009).

Fig. 11. A working model for tone-specific plasticity (BF shifts) elicited by auditory fear conditioning, paired CS – US

ACh, acetylcholine; AI, primary auditory cortex; CS, conditioned stimulus (tone bursts); ICc and ICx, central nucleus and external cortex of the inferior colliculus; MGBv and MGBm, ventral and medial divisions of the medial geniculate body; NB, nucleus basalis in the forebrain; non-AI, auditory cortex other than AI; PMT, pontomesencephalic tegmentum; TRN, thalamic reticular nucleus; US, unconditioned stimulus (electric leg-shock). The short bar at the end of a line means a projection from inhibitory neurons. The conditioning elicits the cortical BF shift through the neural net in AI, corticofugal feedback loop and ACh from the NB, and also the collicular BF shift through the corticofugal feedback and ACh from the PMT. See the text (revised version of Suga et al. 2000).

Fig. 12. Nonspecific augmentation of four cortical (AI) neurons elicited by pseudo-conditioning (p-cond), unpaired CS-US

The best frequencies (BFs) of the four collicular (ICc) neurons (a - d) were either 29, 24, 26 or 26 kHz (vertical dotted lines). The arrows along the frequency axis indicate the frequencies of the CS_u (CS used for p-cond) which were either 0 (a), 5 (b), 10 (d) or 15 (c) kHz different from the BFs of the recorded neurons. In each column, 1 – 3 show the receptive fields obtained before (control), 30 and 210 min after the onset of p-cond, respectively. P-cond elicits the augmentation of responses, broadening of the receptive field, and decrease in threshold of all these neurons. For simplicity, all these changes are represented by the term “nonspecific augmentation.” The scale bars from dark blue to dark red show low to high spike counts per 10 stimuli. Note a small BF shift toward 19 kHz (i.e., tone-specific change) in b2 (Ji and Suga 2008).

Fig. 13. Changes in the frequency-response curves of single collicular (ICc) neurons elicited by pseudo-conditioning (p-cond)

(a) The BFs of four collicular neurons (1 – 4) were 24, 20, 26 and 22 kHz, respectively. The difference between the BF (vertical dashed line) and CSu (vertical arrow) frequencies is shown at the top of each graph. Note the overall augmentation of the responses 30 min after the onset of p-cond and, in addition, the 1.0 kHz BF shift (oblique arrow) only in 2. (b) The nonspecific augmentation and small BF shift elicited by p-cond were first confirmed in three collicular neurons (1–3, upper graphs). After the recovery from these changes, the drug effects on the changes elicited by the 2^nd p-cond were studied in the same three neurons (1–3, lower graphs). The frequency-response curves of the same single neuron were obtained before (control), 30, 60 and 180 min after the p-cond. (b1) Atropine (Atr) was ipsilaterally applied to the ICc. (b2) Muscimol (Mus) ipsilaterally applied to AI. (b3) Muscimol bilaterally applied to SI. The BFs of these three neurons were 25, 24 or 27 kHz and were 5.0 kHz higher than the CSu frequency. Note that the drugs blocked the BF shift, but not the nonspecific augmentation. All the curves were obtained with tone bursts at 10 dB above the minimum threshold of a given neuron. The keys of the curves are shown in the inset (Ji and Suga 2009).