Auditory Neuroscience: Balancing Excitation and Inhibition during Development

Abstract

Two recent studies have described how the coupling of excitatory and inhibitory inputs to neurons in the auditory cortex changes during development. This process is driven by experience and, once complete, may limit the plasticity of the cortex in later life.

Human auditory perceptual abilities mature over different timescales depending on which aspect of hearing is considered, in some cases not reaching the levels seen in adults until several years after birth. Consistent with this is the finding that many of the response characteristics of neurons recorded in animals change during development, although the rate at which they mature depends not only on the property in question, but also on the level in the auditory pathway and the species in which this is examined [1,2]. While previous research has focused on the use of extracellular recordings of spiking activity to investigate the developmental emergence of auditory representations in the brain, two new studies [3,4] have provided intriguing insights into this process by using in vivo whole-cell recordings to characterize the maturation of excitatory and inhibitory synaptic inputs to neurons in the primary auditory cortex (A1).

A universal feature of A1 neurons is that they are tuned to particular sound frequencies. This tuning originates in the biomechanical properties of the basilar membrane in the inner ear and is conveyed — via the receptor hair cells — through successive stages of the central auditory pathway. But the properties of A1 neurons are determined not only by their thalamic inputs, but also by the circuitry of the cortex itself and, in particular, by intracortical inhibition [5].

Previous whole-cell voltage-clamp recordings from A1 neurons in adult rats used different holding potentials to measure excitatory and inhibitory synaptic currents. These conductances are largely matched in amplitude and exhibit similar frequency tuning, with inhibition following excitation after a delay of a few milliseconds [6-9]. It is easy to see how inhibitory inputs might sharpen the frequency selectivity of cortical neurons — either by simply reducing the excitatory response across the full span of frequencies so that a more restricted range induces suprathreshold activity [7,8] or, if inhibitory tuning is slightly broader than excitation, by exerting a relatively larger effect on the flanks of the excitatory tuning curve [9]. Because cortical inhibition also appears to contribute to the selectivity of cortical neurons for upward or downward frequency sweeps [6], an important component of many natural sounds, it is of particular interest to understand how the pattern of synaptic inputs emerges during development and whether this process is shaped by experience.

Two groups [3,4] have now used the same approach to investigate the development of excitation and inhibition in A1. Both made whole-cell recordings from rat A1 neurons and observed changes in the balance of excitation and inhibition with age. But that’s where the similarity ends, as one study [3] emphasized developmental changes in the strength of excitatory inputs, whereas the other [4] reported that inhibitory conductances become more sharply tuned with age.

By presenting tones of various frequencies and intensities Sun et al. [3] were able to determine the full synaptic frequency-intensity response area for both excitatory and inhibitory inputs. When recordings were made from rat A1 neurons at postnatal day 12–13 — just after the onset of hearing — their thresholds were, as expected, found to be much higher than those in older animals. This almost certainly reflects the developmental status of the peripheral auditory system, as a similar reduction in threshold with age was observed when the auditory brainstem response was measured.

Perhaps surprisingly, Sun et al. [3] found that the shape and range of the excitatory and inhibitory tuning curves of A1 neurons were closely matched at postnatal day 12–13. Recordings in slightly older rats revealed that the neurons’ thresholds declined rapidly over the next few days and the frequency-intensity response areas of their synaptic inputs expanded. After this stage, however, Sun et al. [3] found that the excitatory tuning curves sharpened over the following weeks of postnatal development, whereas the inhibitory tuning did not change any further (Figure 1A).

Figure 1.

Two different models for the developmental refinement of sound-driven excitatory and inhibitory responses of rat A1 neurons.

(A) The data of Sun et al. [3] suggest that from ~2 weeks after birth, excitatory inputs become more sharply tuned with age, whereas inhibitory tuning is stable. (B) By contrast the data of Dorrn et al. [4] suggest that excitatory tuning remains the same, but that inhibitory tuning improves over this period.

These results therefore indicate that the balance between excitatory and inhibitory tuning actually breaks down slightly during postnatal development, with mature A1 neurons having frequency-intensity response areas in which inhibition is broader than excitation. This is consistent with previous findings obtained in adult rats by this group [9], and suggests that inhibitory effects at the flanks of the receptive fields do indeed contribute to the frequency selectivity of A1 neurons. Moreover, because Sun et al. [3] found that the total range of sound frequencies over which A1 neurons can be excited or inhibited by sound did not change after about postnatal day 16, they argued that postnatal refinement involves a frequency-dependent adjustment in the strength of excitatory inputs, rather than a loss of those inputs.

The other study to investigate the development of synaptic inputs to rat A1 neurons also found that excitatory and inhibitory responses are present and equally strong just after hearing onset. But in contrast to the findings of Sun et al. [3], Dorrn et al. [4] reported that excitatory and inhibitory responses tended to peak at different frequencies. Indeed, the lack of correlation between excitation and inhibition suggested that they were far from co-tuned at this stage of development.

Dorrn et al. [4] also recorded from older rats and found that the correlation between the excitatory and inhibitory inputs improved during the third and fourth weeks of postnatal development. In their hands, this was not due to a change in excitatory frequency tuning, which resembled that seen in adults as early as postnatal day 15. Rather, they found that A1 inhibition was initially less tuned than excitation, and that adult levels of inhibitory frequency selectivity were reached only toward the end of the first postnatal month (Figure 1B).

The difference between the results from these two studies is quite striking. The explanation for this may lie in some key methodological differences in the way they were conducted. For example, as previously mentioned, Sun et al. [3] recorded synaptic currents over a range of sound frequencies and intensities; however, Dorrn et al. [4] compared excitatory and inhibitory responses at a single fixed intensity, which is potentially significant when thresholds change so dramatically with age. There were also differences in the cortical location of the neurons that were sampled, with Sun et al. [3] focusing on layer 4, while Dorrn et al. [4] recorded over a larger region, spanning layers 3–6.

Role of experience in excitation–inhibition coupling

It is well established that auditory experience has a profound influence on the postnatal development of the central auditory pathway [1,2]. The maturation of the frequency tuning of A1 neurons is no exception to this, with passive repeated exposure to particular sounds inducing a reorganization of the cortical frequency map [10,11]. This plasticity is developmentally regulated, with as little as three days’ exposure to pulsed single-tone stimuli during early postnatal life causing an expansion of the region of A1 tuned to that frequency [12].

This obviously leads to the question of what effect auditory experience has on the synaptic properties of A1 neurons. Dorrn et al. [4] went on to address this by exposing young rats to repetitive single-tone stimuli for periods that varied in duration from a few minutes to a few days. They found that as little as 3–5 minutes of tonal stimulation altered both excitatory and inhibitory synaptic responses across a range of frequencies, causing their tuning profiles to become more similar. This plasticity lasted a few hours, and was not seen in adult rats.

Thus, brief periods of sensory experience appear to temporarily accelerate the balancing of excitatory and inhibitory synaptic inputs that these authors described during postnatal development. Early exposure to single-tone stimuli for 1–3 days again enhanced excitatory-inhibitory correlation, but this time the changes lasted for weeks. Once balanced excitation and inhibition had been achieved, additional episodes of tonal stimulation had no further effect on synaptic tuning, implying that the matching of these inputs may be the limiting factor in determining whether the receptive fields of cortical neurons can change in response to an altered acoustic environment.

While these findings suggest that inhibitory inputs are particularly susceptible to experience, it remains the case that the conflicting conclusions of the two studies highlighted here have to be reconciled. Nevertheless, it seems highly likely that in vivo whole-cell recordings will be able to provide very valuable insights into the maturation of cortical sensitivity to other, more complex stimulus features and how this is shaped by experience. Just as important, however, for establishing a link with developmental hearing disorders in humans is the application of behavioral studies in young animals [13].