Auditory processing in zebrafish and its relevance to autism

Autism is clinically defined by criteria that include differences in social behavior, communication, and language, as well as restrictive and repetitive patterns of behavior. In 2013, the DSM-5 includes differences in sensory responsivity as part of the autistic profile, including hyper-sensitivity to sound. Auditory processing differences have previously been described, and include increased sensitivity, reduced filtering, increased connectivity, and a reduction in auditory short- and long-term adaptation from the same repetitive auditory stimulus. Neurologically, some of these auditory processes have been studied in humans, but technological limitations and ethical considerations prevent the detailed dissections of the underlying neural mechanisms.

For this reason, we look to animal models to help build a more in-depth picture of whole-brain and regional processes, in this case, larval zebrafish models of autism together with selective plane illumination microscopy (SPIM) to investigate whole brain and regional processes. To date, we have a partial and inconsistent knowledge base regarding the auditory processing capabilities larval zebrafish. To determine a baseline for larval zebrafish hearing, a novel method for delivering water-borne sounds was used with a diverse assembly of acoustic stimuli and SPIM imaging. Using a conditional thresholding analysis, results reveal responses to frequencies ranging from 100Hz-4kHz, with frequency discrimination from 100Hz-2.5kHz. Frequency-selective neurons are present, with topographic organization of frequencies in multiple auditory regions, suggesting developmental progress towards eventual tonotopic organization in adult fish. Using functional clustering, the analysis identified categories of neurons that are selective for pure-tone stimuli, white noise complex sounds, sounds with a sharp onset, and sounds involving a gradual crescendo. These results suggest a more nuanced auditory system than has previously been described in larval fish

With a baseline for larval auditory processing established, the next two studies in this thesis examined various auditory processes in the zebrafish models of autism. In these studies, the fragile X fmr1-/- model is used. Fragile X is the most common monogenetic form of autism, and individuals with autism and fragile X are often hypersensitive to sound. The first of the two studies explored differences in auditory sensitivity between the fmr1-/- model and its wildtype (WT) siblings by presenting acoustic stimuli at a wide range of volumes. Using functional analysis, auditory responses were more plentiful overall, particularly in the hindbrain. The primary auditory region, the octavolateral nuclei (ON) had more auditory responsive neurons in the fmr1-/- fish compared to their WT siblings. Three regions all downstream from the primary auditory region responded more strongly to auditory stimuli – the thalamus, torus semicircularis, and tegmentum, with a 2-fold (6 dB) shift in sensitivity. Results suggest that the fmr1-/- fish are more sensitive to sound, and more auditory information is transmitted throughout the auditory pathway, suggesting less filtering of information in the fmr1-/- fish. A graph theory analysis showed an increase in functional connectivity between regions at all volumes in the fmr1-/- fish.

The final experiment set out to investigate auditory adaptation in larval zebrafish and how this differed in fmr1-/- fish. Using whole brain and regional clustering approaches, we found adaptation rates varied between brain regions, with the fastest adaptation rates evident in the fore- and midbrain regions and slower adaptation in the hindbrain. Whole brain clustering revealed the fmr-/- fish show a higher magnitude of response to the first stimuli, demonstrating auditory sensitivity, and reduced filtering downstream as seen in the previous experiment. Additionally, the fmr-/- fish were shown to have a reduced recovery to the re-encountered stimulus after a rest period. WT fish show almost full restoration of response after a rest period and then show sensitization.

Using regional clustering, all ten regions were found to have at least one cluster that was similar across regions, with a low magnitude of response, as well as auditory sensitivity in the fmr1-/- fish and sensitization in the WT after a rest period. Five regions had a second, more responsive cluster that was distinct from the first cluster found in all regions, these being the thalamus, tegmentum, torus semicircularis, ON, and remaining hindbrain. This additional cluster had a higher magnitude of response than the first, with a denser spatial organization. The midbrain regions had a higher magnitude of response than the primary auditory region (ON). These results suggest that these midbrain regions, specifically the torus semicircularis and the tegmentum could play a pivotal role in adaptation and affect regions downstream in the fmr1-/- larval zebrafish.

Results from this thesis firstly demonstrate larval zebrafish have a more sophisticated auditory processing system than previously described and are a good model to investigate auditory processing. In the fmr1 model, results include increased sensitivity, reduced filtering, increased regional connectivity and an auditory adaptation phenotype that includes reduced recovery in fmr1-/- fish. These results can help direct and inform research in other animal models, and in human models with regards to auditory processing and autism.