High temporal acuity of auditory processing underlies perception of speech and other rapidly varying sounds. A common measure of auditory temporal acuity in humans is the threshold for detection of brief gaps in noise, which is typically on the order of a few milliseconds. Longer gap-detection thresholds, as observed in patients with developmental disorders, are usually considered evidence for ''sluggish'' auditory processing. Here we show, in an animal model of developmental disorder, that deficits in auditory brain sensitivity to brief gaps in noise do not necessarily imply a general loss of central auditory temporal acuity, and may arise instead from specific impairments in neural coding of sound offsets.

We recorded extracellularly from neurons in three different subdivisions of the auditory thalamus, and also in the inferior colliculus of the auditory midbrain, in the BXSB/MpJ-Yaa mouse model of developmental disorder and gap-detection deficits. Thalamic recordings revealed a stimulus-specific, subdivision-specific deficit in neural sensitivity to brief gaps in noise in experimental animals relative to controls. Neural sensitivity to brief gaps in noise was reduced in lemniscal and non-lemniscal (but not polysensory) subdivisions of the medial geniculate body of the auditory thalamus. However, thalamic responses to other rapidly varying stimuli, such as rapid click trains, were unaffected. Moreover, no deficit in neural sensitivity to brief gaps in noise was observed in the inferior colliculus. These findings indicate that the deficits in thalamic sensitivity to brief gaps in noise do not arise from a general loss of central auditory temporal acuity --- and likely originate within the auditory forebrain.

Further analysis of thalamic recordings revealed that deficits in neural sensitivity to brief gaps in noise in experimental animals reflect a specific abnormality in neural coding of sound offsets, which entirely spares neural coding of sound onsets. To explain these results, we developed a simple phenomenological model of temporal processing in the auditory forebrain, incorporating intensity gain control and dissociable sound-onset-sensitive and sound-offset-sensitive coding mechanisms. We fit the parameters of the model using data-derived measures of thalamic population responses to noise onsets and offsets, and then tested the model using simulated gap-in-noise stimuli, rapid click trains, and clicks following noise. Two versions of the model differing only in the value of a single parameter --- the weighting of the sound-offset-sensitive mechanism --- successfully reproduced both similarities and differences between thalamic population responses recorded in experimental and control animals.

Our findings indicate that deficits in auditory temporal acuity observed in developmental disorders might arise from specific impairments in neural coding of sound offsets. Moreover, the results suggest that while time-scales for integration in the auditory forebrain are typically on the order of tens, hundreds or thousands of milliseconds, temporal processing at millisecond resolution can be achieved through combination of sound-onset-sensitive and sound-offset-sensitive coding mechanisms.