The Cerebellar Dentate Nucleus in Autism: Are Neurons and Their Perineuronal Nets Preserved Despite Missing Purkinje Cells?

Background: Human postmortem studies from our laboratory and others have shown differences in Purkinje cell (PC) number, size and gene expression in the autism brain. In fact, PC dysfunction is the most consistent neuropathological finding in autism, with as many as 75% of cases showing reduced numbers. The lateral hemisphere displays the greatest PC decrease reported in autism. However, the dentate nucleus (DN), which normally receives projections from the lateral hemisphere, did not seem to be affected in preliminary studies of total neuron numbers in autism, although very few postmortem cases have been analyzed. The DN harbors the greatest percentage of neurons in the brain surrounded by a perineuronal net (PNN). Genes involved in PNN formation and function have been implicated in autism.

Objectives: Given the known PC reductions in autism, we aimed to determine whether there are further deficits in neuronal numbers within the cerebellar circuitry, as DN neurons are likely to be impacted by PC deficits. PNNs around the DN neurons were also quantified.

Methods: Immunohistochemistry was performed on human postmortem brain tissue from twenty control and twenty autism cases. From each case, five 40µm sections at every sixth interval through the DN were immunostained for anti-HPLN1, a link protein in the PNN, with neutral red as a counterstain for total neuron counts. Tissue was treated for antigen retrieval with tris-buffer (pH 9) for ten minutes in a scientific microwave at 98˚C, followed by twenty minutes of hydrogen peroxide, forty-eight hours of incubation with primary antibody (anti-HPLN1 1:125, R&D systems), two hours with anti-goat secondary antibody (1:600, Vector Laboratories), one and a half hours with avidin-biotin complex (Vector Laboratories) with buffer rinses between each step. Nickel-DAB reaction was followed by 30 minutes of neutral red, dehydration through a series of ethanol solutions and five minutes of xylene before being coverslipped. A Zeiss Microbrightfield Stereoinvestigator system was used to quantify neuronal densities within the DN contour. The density of neurons surrounded by a PNN, neurons without a PNN and total neuron numbers were estimated with the optical fractionator method then divided by the total estimated area using the Cavalieri method within the software.

Results: The density of neurons with PNNs, based on HPLN1 expression, was not different between control (mean= 2,429.39±892.37 neurons/mm³) and autism (mean= 2,141.63±569.80 neurons/mm³). The density of neurons without a PNN also showed no differences in control (mean= 2,896.47±958.23 neurons/mm³) or autism (mean= 2,901.33±995.33 neurons/mm³) and therefore total neuron numbers were similar (control mean= 5,325.86±1,628.85 and autism mean= 5,042.96±1,202.45 neurons/mm³).

Conclusions: Despite reports of significant reductions in the number of PCs in the lateral hemisphere of the cerebellum, similar reductions are not evident in the DN. Thus the targets of PC output within the DN appear to be preserved by the remaining PCs and inferior olive input. Furthermore, the proportion of neurons surrounded by PNNs compared to those without appear to be unaltered. Further work is underway to determine whether activity-dependent components of the PNN, such as aggrecan, are affected while total PNN numbers remain unchanged.