Yost Chapter 15: The Central Auditory Nervous System pp. 227-251. (19 March, 2004)

 

Yost provides a very schematic diagram of the central auditory system in Figure 15.1: while schematic and terribly incomplete, it is also very complex! And as a first step, it provides a framework for the complications, some of which are given later on. Note that the first staging point into the brain is the cochlear nucleus (there is not one but two major divisions, one having two parts, that are very different in their cellular structure and their connections: namely, the dorsal and the ventral cochlear nuclei, with the ventral further subdivided into the anterior and the posterior: but you can see how Yost has structured those divisions in his connections of the auditory nerve). All of these nuclei include the incoming afferent auditory pathways; but they may also have downstream afferents from other nuclei, and certainly interneurons as well, which are responsible for processing the afferent information before it is passed on to other nuclei. In addition there are substantial inputs from other senses, certainly for the dorsal cochlear nucleus, which permits integration across senses and possibly learning (neuroplasticity).

 

From the ventral cochlear nucleus, which is the simpler part of the CN anatomically, a ventral pathway swings across the midline going directly up to the inferior colliculus; an intermediate pathway goes to one of the nuclei of the trapezoid body, and hence to the superior olive; and the third goes directly to the superior olive: all of these are contralateral connections. Other connections of the ventral cochlear nucleus go to the superior olive on the same side, actually to the same divisions of the superior olive: one division of the superior olive (the lateral nucleus) receives an excitatory connection from the ipsilateral ear, and an inhibitory connection from the same part of the ventral cochlear nucleus of the other ear, its sign reversed (+ to -) in the trapezoid body, considered a part of the superior olive. Another part of the superior olive (the medial nucleus) receives an excitatory connection from both sides. This part of the brain gets the first representation of input from both ears. The lateral nucleus mostly receives high frequency stimuli, the medial nucleus mostly receives low frequency stimuli: can you then imagine what these two binaural nuclei do? Which might be large in the elephant? Which in the mouse? [there are at least 8 more of these nuclei, but not much is known about the others. You might try to find out, if you are looking for a paper topic].

 

Note that the dorsal cochlear nucleus also connects to the superior olive and also to the inferior colliculus, and note too that the superior olivary complex connects to the inferior colliculus on both sides. Figure 15.1 has these several connections going through the lateral lemniscus, which is a fiber tract: but there are also ventral and dorsal nuclei within the lateral lemniscus, which connect up to the inferior colliculus on both sides (note the Commissure of Probst). The inferior colliculus is not just one structure but three and perhaps four: a central nucleus, mostly providing upstream information, which processes auditory information and sends it on to the medial geniculate (the thalamic nucleus of the auditory system, which is not one nucleus but certainly three) and then on to auditory cortex (which is not one place but possibly four or more). At each place “side” pathways go off to non-auditory structures -- for example, for the acoustic startle reflex from the cochlear nucleus to the reticulospinal tract in the brain stem; for directed eye movements, from the inferior colliculus to the surrounding midbrain and to the superior colliculus. And other pathways provide input into the auditory pathways from other senses (for example, the dorsal cochlear nucleus as described above) and from down stream connections of higher processing centers (such as the downstream pathway from the superior olivary complex to the cochlear nucleus, a collateral of the path to the haircells in the cochlea).

 

Figure 15.2 is even more simple but interesting in its showing that at the superior olive and beyond binaural information is present in each site. Figure 15.3 is also simple and also interesting because it shows downstream influences. A safe rule of thumb is that each upstream pathway is accompanied by a downstream pathway. Then Figure 15.4 shows some real cells in the cochlear nucleus and in the inferior colliculus. We look at these more seriously in class to see what can made of structure/function relationships.

 

The next section Yost calls “strategies for studying the central auditory system.” The first “principle of organization” is its topographic organization: beginning with the basilar membrane and continuing through to the cortex each part of the system has a"dimension of pitch," with similar frequencies together organized by increasing frequency of the stimulus. Typically these are studied by using a micro-electrode to record cellular activity, and finding the BF of a cell; then moving the electrode in one spatial dimension, finding the BF again, and so forth. Along some dimensions it may be that a best frequency stays much the same (going vertically down into a “barrel” of the cortex, for example), while another spatial dimension reveals a systematic shift in BF as one proceeds (from anterior to posterior in the cortex, dorsal to ventral in the cochlear nucleus, etc.). Yost makes the interesting point that other dimensions of organization seem to be difficult to find (though the barn owl has a spatial location for sound in the superior colliculus and the inferior colliculus), but this may be because the right kind of research has not been done as yet. At least in the inferior colliculus it is thought that other information (binaural/monaural?; temporal acuity? SAM?) may be orthogonal to the vertical frequency dimension.

 

A second “principle of organization” is the concept of excitation and inhibition. Yost makes the point that the central auditory nervous system (the CANS) does its work by a complex aggregations of excitation and inhibition: by adding or subtracting. Thus in the lateral nucleus of the superior olive EI (excitation/inhibition) cells are excited by ipsilateral stimulation and inhibited by contralateral stimulation. Of course we know the next part, that E or I cells have different neurotransmitters. Yost mentions acetylcholine, GABA, and glycine, but there are many more (how many: 50 counting subtypes? even more?).

 

A third principle of organization in the type of response one might find in single fibers/single neurons. For example, in the auditory nerve spontaneous firing and driven rates may follow the phase locked period of the cell regardless of its input. However, in the CANS the spontaneous rates depend on lots of things, such as the state of attention, Also, in the auditory nerve the response typically is related in a simple fashion to stimulus level (save for saturation). But in the central nervous system a cell may fire only to a particular intensity — probably not because of the characteristic of that cell alone, but because of its inhibitory sidebands in the circuits that drive that cell. Central cells thus tend to have a small dynamic range, smaller than the auditory nerve even. Also the auditory nerve may phase lock to 3 - 4 kHz or so; but in the CANS it may phase lock, if at all, to only a few hundred hertz (save for the bushy cells in the VCN and the principle cells of the MNTB, which match the auditory nerve input). The patterns of firing of central neurons may be very different: some show “on responses”, others “off responses”, others give a little burst on onset, then small bursts if the stimulus is continued, “chopper cells” and so forth (see Figure 15.5). A final way to analyze a unit is by its latency to fire, which might mean that if two cells in the same nuclear complex fire with different latencies, that one is part of a circuit with more cells/synapses in its pathway (though this is not the only explanation). Yost presents an interesting comparison from the famous paper by Kiang (1965) showing resonant phase locking to a click in an auditory nerve fiber, but just a single pulse in a cochlear nucleus neuron (Figure 15.6). What happened to the rest of the information in the input? It went off to some other neurons that connect up with different circuits and provide different sensory experiences.
              

Yost then looks at neural circuits. He describes in Figure 15.7 a very familiar type of circuit that is common in other sensory systems, and common in audition as well, just not in the cochlea. It is the circuit that yields lateral inhibition. It works (in this case, in his schematic) by having adjacent cells in a bank of afferents mutually inhibit (presynaptically) their connections to their targets. Imagine a tone is presented, that activates a band of these cells. Just at the edge of the band, the first excited cell inhibits the adjacent non-excited cell, and drives it below zero; while in turn, that cell just off the edge has no inhibition to inhibit the first excited cell, which, however, inhibits the adjacent excited cells on the other side. This little (hypothetical) neural circuit will result in an edge enhancement, and sharpen the frequency discrimination (see page 237). There are indeed lots of cells that inhibit nearby cells in the auditory system, and further, there are even “projection” inhibitory GABA neurons that inhibit at long distances. I have read that these exist only in the auditory system (lots of the commissural cells are GABA or glycine projection neurons).

 

Yost next introduces the concept of the “evoked potential” as a way of examining auditory activity in large groups of neurons that may by inaccessible to single unit recording, or can be done only under anesthesia (which changes how cells respond). Yost describes some “event related evoked potentials” recorded from the scalp, with the additional use of signal averaging to draw a small but regular potential from a large but irregular background. Evoked potentials have been used for about 40 years, but they are not typically used nowadays for “science as hypothesis testing,” but instead for clinical testing. Here the brain stem evoked response (ABR) is now the best known and most useful (see Figure 15.8), and it is possible to describe changes in the speed of neural processing by comparing the latencies of different peaks in the ABR (Figure 15.9, and 15.10).

 

He also presents a section on imaging techniques, showing how they may be used to indicate (it is assumed) where auditory processing occurs in human subjects (Fig. 15.11).

 

In the last section Yost describes activity in four brain centers, the cochlear nucleus, the superior olive, the inferior colliculus, and the auditory cortex. Note in Figure 15.12 how the auditory nerve unravels itself in the cochlear nucleus so that each part gets a complete map of the basilar membrane, the dorsal (upper) part of each segment being the high frequency section of the basilar membrane. Figure 15.12 is hard to follow, as it looks like the different turns go to the different 3 sub-nuclei, which they do not. Figure 15.13 is a realistic photomicrograph from Rose et al. (1959) showing how the same frequencies are represented in the dorsal and in the ventral cochlear nucleus (as one follows down an electrode track from top right to bottom left).

 

The cochlear nucleus, like the other central structures, has cells that appear in a great many different shapes, and appear to respond at least to some extent differently. Figure 15.14 is classic in showing how these cells are in different parts of the nucleus, and respond in different ways. The one type that I find easy to remember is the “octopus ON cell” from the posteroventral CN. Because of its cellular characteristics it only responds to the beginning of a stimulus; and because of its long dendrites that run across the auditory nerve, it responds to the onset of a tone in any frequency. Note that the anteroventral portion of the cochlear nucleus has bushy cells which function largely like auditory nerve fibers, and are excitatory (Type I cells) ; or stellate cells, which may be “chopper cells” and have possible inhibitory surrounds (called Type III cells). The dorsal cochlear nucleus has a marvelous collection of cells some of which are the curious Type IV cells, which have tiny bands of excitation, and respond at BF to a very small range of stimulus levels. Unfortunately this mass of data on different types of CN cells has not been systematically exploited for how it fits into their functional role in the cochlear nucleus, though others now besides Sachs and Young, are continuing their work on vowel encoding in the auditory nerve up into the cochlear nucleus.

 

The next section on the superior olivary complex is also rich in electrophysiological data, now in relationship to localization. Note in Figure 15.15 that a cell in the medial superior olivary nucleus does not fire when the stimuli to the left and right ear arrive at the same time, and fires most when the left ear leads by a substantial difference (the rising curve on the right of the figure with the + time units represents left ear leading). However, strangely note that the cells fires more than “zero” if the right ear leads by a small amount. The cell “likes” left side stimuli but it really doesn’t like straight ahead stimuli it seems!

 

The inferior colliculus is now described as having the central nucleus (which is the relatively simple station for activity that will be processed and passed upstream), and the dorsal cortex and paracentral nuclei. Like the SON, the IC does spatial localization (see Figure 15.16). You might note a curious feature of the X-axis, which is that the interaural time differences are artificial: one wouldn’t expect interaural time differences as long as 4 ms unless the subject was an elephant, perhaps. Yin works with cats. The IC cells are EE and EI, binaural and monaural; tonotopically organized; and possibly even show differences with vertical orientation. Cells that lie outside of the central nucleus cells also show complex attentional effects and habituation effects.

 

The last section deals with auditory cortex, which in primates is hidden within the Sylvian fissure. In cats it is more accessible and the several cortical areas shown in Figure 15.17 have been uncovered — eight areas are shown. Yost emphasizes the “non-obligatory” nature of auditory cortex in his describing the anesthetized animals show beautiful tonotopic organization of auditory cortex, while awake animals do not, or it is less apparent. Cells in the awake animal are more apt to prefer complex stimuli such as gliding FM modulated stimuli, perhaps the sounds of conspecifics; and they may respond to onsets but not continued stimuli. Figure 15.18 gives histograms of some of these cell types, not the most complicated ones. Figure 15.19 shows phase locking to AM stimulation.

 

Yost says just a few words about other techniques, apart from electrophysiological recording: lesioning and stimulation studies for example. Surprisingly, many simple auditory discriminations can be maintained in animals that do not have an auditory cortex, though they seem especially deficient in certain tasks that one would think are easy — picking out a gap between two tone pips, for example, or remembering which of two tones came first. Also he doesn’t say anything about auditory learning and the concept of neuroplasticity. We will look at just a little of this research, showing the effects of hearing loss on frequency tuning, and also some data showing how tuning curves can be modified by reinforcing events.