Developmental sensory experience balances cortical excitation and inhibition

Abstract

Early in life, neural circuits are highly susceptible to outside influences. The organization of the primary auditory cortex (A1) in particular is governed by acoustic experience during the critical period, an epoch near the beginning of postnatal development throughout which cortical synapses and networks are especially plastic. This neonatal sensitivity to the pattern of sensory inputs is believed to be essential for constructing stable and adequately adapted representations of the auditory world and for the acquisition of language skills by children. One important principle of synaptic organization in mature brains is the balance between excitation and inhibition, which controls receptive field structure and spatiotemporal flow of neural activity, but it is unknown how and when this excitatory-inhibitory balance is initially established and calibrated. Here we use whole-cell recording to determine the processes underlying the development of synaptic receptive fields in rat A1. We find that, immediately after the onset of hearing, sensory-evoked excitatory and inhibitory responses are equally strong, although inhibition is less stimulus-selective and mismatched with excitation. However, during the third week of postnatal development, excitation and inhibition become highly correlated. Patterned sensory stimulation drives coordinated synaptic changes across receptive fields, rapidly improves excitatory-inhibitory coupling and prevents further exposure-induced modifications. Thus, the pace of cortical synaptic receptive field development is set by progressive, experience-dependent refinement of intracortical inhibition.

Figure 1

Refinement of excitatory-inhibitory balance during AI critical period. a, Imbalanced synaptic frequency tuning at P14. Top, frequency tuning of excitation (filled) and inhibition (open). Bottom, excitation and inhibition were uncorrelated (linear correlation coefficient r: −0.01, p>0.8). b, Balanced tone-evoked excitation and inhibition in adults. Top, frequency tuning. Bottom, excitation and inhibition were correlated (r: 0.87, p<0.001). c, Increase of excitatory-inhibitory balance during AI critical period. Circles, individual recordings. Squares, averages. d, Summary of developmental changes to excitatory-inhibitory balance. Top, excitatory-inhibitory correlation in young and adults (P12-P21, r: 0.37±0.06, n=43; adults, r: 0.71±0.05, n=31, p<10⁻⁴ compared to P12-P21, Student’s two-tailed t-test). **, p<0.01. Bottom, difference in excitatory and inhibitory best frequencies in young and adults (P12-P21, best frequency difference: 1.4±0.2 octaves, n=43; adults, 0.2±0.1 octaves, n=25, p<10⁻⁶). Error bars, s.e.m.

Figure 2

Delayed maturation of inhibitory frequency tuning. a, Excitatory frequency tuning was sharper than inhibitory tuning at P14. Lines, Gaussian fits (σ_Exc: 4.0, σ_Inh: 9.6). Same recording as in Figure 1a. b, Excitatory and inhibitory tuning were both sharp in adulthood (σ_Exc: 2.7, σ_Inh: 2.5). Same recording as in Figure 1b. c, Excitatory frequency tuning sharpened before inhibition (P12-P15, σ_Exc: 5.4±1.0, σ_Inh: 9.4±1.6, n=15, p<0.02).Circles, excitation (filled) and inhibition (open) for each cell. Squares, averages. d, Summary of developmental changes to tuning. Top, tuning sharpness in young (P12-P21, σ_Exc: 4.9±0.4, σ_Inh: 7.7±0.8, n=43, p<0.0004) and adults (σ_Exc: 4.4±0.4, σ_Inh: 4.5±0.6, n=31, p>0.8). Bottom, excitation-to-inhibition ratio (E:I ratio) was unchanged during AI critical period. E:I ratios were similar between young (P12-P21, E:I ratio: 1.3±0.2, n=43) and adults (E:I ratio: 1.2±0.2, n=31, p>0.6). Error bars, s.e.m.

Figure 3

Patterned stimulation rapidly enhanced excitation and inhibition during P12- P21. a, Long-term synaptic enhancement at P19. Before patterned 2 kHz stimulation, excitation and inhibition were moderately correlated (r_pre: 0.57); after stimulation, correlation increased (r_post: 0.86). Top, excitation at 2 kHz increased after patterned stimulation (enhancement of 75.2%, p<0.05). Insets, conductances evoked by 2 kHz before (gray) and after (black) repetitive stimulation. Arrow, frequency chosen for patterned stimulation. Scale bars, 1 nS, 40 msec. Bottom, inhibition at 2 kHz increased after repetitive stimulation (enhancement of 138.5%, p<0.05). b, Patterned stimulation did not affect adult AI. Top, excitation was unaltered after 8 kHz patterned stimulation (enhancement of 7.4%, p>0.3). Scale bars, 0.5 nS, 40 msec. Bottom, inhibition at 8 kHz remained unchanged (enhancement of 1.8%, p>0.3). Excitatory-inhibitory correlation was unaffected (r_pre: 0.68, r_post: 0.74). c, Critical period for synaptic modifications induced by patterned stimulation. Circles, changes to excitation (filled) or inhibition (open) for each recording. d, Time course of synaptic modifications to tone presented during patterned stimulation. Top, P12-P21 (excitation: 63.1±11.3%, n=12, p<0.0002; inhibition: 52.9±14.1%, p<0.004). Horizontal bar, patterned stimulation. Bottom, adults (excitation: 5.2±5.3%, n=11, p>0.3; inhibition: 0.2±5.1%, p>0. 9). Error bars, s.e.m.

Figure 4

Patterned stimulation improved excitatory-inhibitory coupling by coordinated synaptic modifications across multiple inputs. a, Synaptic modifications at the presented tone frequency spread to other inputs within one octave (excitation one octave from presented frequency: 21.6±6.7%, n=12, p<0.01; inhibition: 36.0±12.5%, p<0.02), but not 2+ octaves away (p>0.3). **, p<0.01; *, p<0.05. b, After patterned stimulation, responses at original best frequency were reduced (excitation: −34.8±6.4%, n=12, p<0.0003; inhibition: −22.7−6.1%, p<0.004). c, Patterned stimulation increased excitatory-inhibitory correlation in young (Δr: 0.31±0.08, n=12, p<0.004) but not adults (Δr: −0.03 ± 0.09, n=11, p>0.7). d, Nonspecific modifications across multiple inputs were predominant for balancing excitation and inhibition. Considered separately, synaptic modifications only at the presented frequency were less effective (“presented only”, Δr: 0.12±0.09, p>0.2) than changes to all other inputs (“unpresented only”, Δr: 0.32±0.09, p<0.004). Error bars, s.e.m.

Figure 5

/B> Patterned stimulation prevented additional synaptic modifications. a, Synaptic tuning before first episode of patterned stimulation. Initially, excitatory-inhibitory correlation was low (r_pre: 0.27). b, Same cell as in a, but after first period of 4 kHz patterned stimulation. Excitation and inhibition at 4 kHz were enhanced and excitatory-inhibitory correlation increased (excitation: 108.7%, p<0.03; inhibition: 44.4%, p<0.05; r_post1: 0.77). c, Same cell as in a, but after second period of 4 kHz repetitive stimulation. Excitatory-inhibitory strength and balance were unaffected (excitation: 2.1%, p>0.4; inhibition: 6.2%, p>0.3; r_post2: 0.82). d, Summary. Top, conductance changes at presented tone frequency after first (excitation: 61.4±16.7%, n=5, p<0.03; inhibition: 84.8±26.5%, p<0.04) and second (excitation: 6.2±11.4%, n=5, p>0.6; inhibition: −6.2±18.0%, p>0.7) stimulation periods. Bottom, change in excitatory-inhibitory correlation after first (Δr: 0.48±0.10, p<0.01) and second (Δr: 0.02±0.09, p>0.8) stimulation periods. Error bars, s.e.m.