Further characterization of autoantibodies to GABAergic neurons in the central nervous system produced by a subset of children with autism - PubMed (original) (raw)
Further characterization of autoantibodies to GABAergic neurons in the central nervous system produced by a subset of children with autism
Sharifia Wills et al. Mol Autism. 2011.
Abstract
Background: Autism is a neurodevelopmental disorder characterized by impairments in social interaction and deficits in verbal and nonverbal communication, together with the presence of repetitive behaviors or a limited repertoire of activities and interests. The causes of autism are currently unclear. In a previous study, we determined that 21% of children with autism have plasma autoantibodies that are immunoreactive with a population of neurons in the cerebellum that appear to be Golgi cells, which are GABAergic interneurons.
Methods: We have extended this analysis by examining plasma immunoreactivity in the remainder of the brain. To determine cell specificity, double-labeling studies that included one of the calcium-binding proteins that are commonly colocalized in GABAergic neurons (calbindin, parvalbumin or calretinin) were also carried out to determine which GABAergic neurons are immunoreactive. Coronal sections through the rostrocaudal extent of the macaque monkey brain were reacted with plasma from each of seven individuals with autism who had previously demonstrated positive Golgi cell staining, as well as six negative controls. In addition, brain sections from adult male mice were similarly examined.
Results: In each case, specific staining was observed for neurons that had the morphological appearance of interneurons. By double-labeling sections with plasma and with antibodies directed against γ-aminobutyric acid (GABA), we determined that all autoantibody-positive neurons were GABAergic. However, not all GABAergic neurons were autoantibody-positive. Calbindin was colabeled in several of the autoantibody-labeled cells, while parvalbumin colabeling was less frequently observed. Autoantibody-positive cells rarely expressed calretinin. Sections from the mouse brain processed similarly to the primate sections also demonstrated immunoreactivity to interneurons distributed throughout the neocortex and many subcortical regions. Some cell populations stained in the primate (such as the Golgi neurons in the cerebellum) were not as robustly immunoreactive in the mouse brain.
Conclusions: These results suggest that the earlier report of autoantibody immunoreactivity to specific cells in the cerebellum extend to other regions of the brain. Further, these findings confirm the autoantibody-targeted cells to be a subpopulation of GABAergic interneurons. The potential impact of these autoantibodies on GABAergic disruption with respect to the etiology of autism is discussed herein.
Figures
Figure 1
Photomicrographs illustrating coronal sections through the macaque monkey cerebellum. (A) Nissl-stained section illustrating the three major layers of the cerebellum: the granule cell layer (GCL), the Purkinje cell layer (PCL) and the molecular layer (ML). The locations of several large Purkinje cells (asterisks) are shown. (B) Immunoreactivity resulting from exposure of adjacent section to plasma from a representative child with autism (age 4 years). Golgi neurons (black arrows) are clearly labeled and lie adjacent to or below the Purkinje cell layer. Ghost images of Purkinje cells are marked with asterisks. Lighter, though consistent, labeling of basket cells (white arrows) was also observed with this plasma. Only plasma from subjects that provided this level of staining of the Golgi cells in the cerebellum was used for analysis of other brain regions. Calibration bar, 100 μm.
Figure 2
Photomicrographs illustrating coronal sections through primary visual cortex (V1) of the macaque monkey brain. (A) Nissl-stained section showing the lamination pattern (cortical layers I to VI) of neurons in this region. (B) Staining of V1 with plasma from one child with autism (age 6 years). Note that in this panel and Figure 2E, the highest numbers of labeled neurons are located in the superficial layers (I to III). (C) Section through area V1 demonstrating in situ hybridization with a probe to glutamic acid decarboxylase 67 (GAD67). The brown reaction product illustrates neurons that are GABAergic. (D) Very similar representation of GABAergic neurons identified immunohistochemically with a monoclonal antibody to γ-aminobutyric acid (GABA). (E) Staining very similar to that shown in Figure 2B of plasma from another child with autism (age 5 years). (F) Section through area V1 that was reacted with plasma from a typically developing child (age 5 years). While there is light, nonspecific background staining that resembles the distribution of Nissl-stained cell bodies, there is no specific labeling of GABAergic neurons. Calibration bar, 100 μm.
Figure 3
Fluorescence confocal photomicrographs of double-labeling with plasma from a representative child with autism (age 5 years) and an antibody directed at GABA. (A) GABA immunoreactivity in area V1 (green). (B) Immunoreactivity of V1 with plasma from a child with autism (red). (C) Merged images from Figures 3A and 3B showing several double-labeled cells (yellow). The vessels have been labeled (V) to aid in orientation with Figure 3D. (D) Schematic of the same region shown in Figures 3A to 3C, with cells coexpressing GABA and plasma immunoreactivity, depicted as yellow circles, GABAergic cells without plasma immunoreactivity, depicted as green hexagons, and a single GABA-negative, plasma-positive cell, depicted as a red star. The vessels have been labeled (V) to aid in orientation with Figure 3C. Calibration bar, 100 μm.
Figure 4
Fluorescence confocal photomicrographs of double-labeling with plasma from a representative child with autism and antibodies directed toward calcium-binding proteins in area V1. (A to C) Immunoreactivity of (A) the autism (AU)-specific autoantibodies, (B) calbindin and (C) the merged image showing that several cells are colabeled (arrows), but that not all AU immunoglobulin G (IgG)-labeled cells express calbindin (arrowheads). Double-labeling with plasma from (D) a autoantibody-positive child with autism and (E) anti-parvalbumin shows that some AU IgG-labeled cells express parvalbumin (arrows), while most do not (arrowheads). (G to I) AU immunoreactivity (G) rarely occurs in cells that express calretinin (H) with no co-localization evident (I). Arrowheads point to cells that are labeled by the autoantibody, but not by anti-calretinin (I). Calibration bar, 100 μm.
Figure 5
Photomicrographs of coronal sections through the posterior cingulate cortex of the macaque monkey. This level includes areas 29, 30 and 23. (A) Nissl-stained section showing the laminar pattern of posterior cingulate cortex located just dorsal to the corpus callosum (cc). The layers of cortical area 23 are marked I to VI. (B) Immunostaining with plasma from a representative child with autism. As in the primary visual cortex, the most numerous labeled neurons are located in the superficial layers (I to III), although immunopositive neurons are located throughout all layers. This pattern was recapitulated in all cortical areas. Labeled cells were not observed in the cc, although apparent positive cells were observed in the subcortical white matter. Calibration bar, 250 μm.
Figure 6
Photomicrographs of the CA3 region of the hippocampus and the dentate gyrus of the macaque monkey. (A) Nissl-stained section showing the major layers of the CA3 region of the hippocampus. Deep to superficial layers include so, stratum oriens; pcl, pyramidal cell layer; sl, stratum lucidum, sr, stratum radiatum, sl-m, stratum lacunosum-moleculare. (B) Section of the CA3 region of the hippocampus stained immunohistochemically with plasma from a representative AU subject. Numerous labeled neurons are observed in the stratum radiatum (black arrows) and the adjacent stratum lacunosum-moleculare (open arrows). Only background level staining is observed in the pyramidal cell layer, but there are a variety of labeled neurons in stratum oriens (white arrows). (C) Nissl-stained section showing the layers of the dentate gyrus. Layers include pl, polymorphic layer; gcl, granule cell layer; and ml, molecular layer. (D) Section of the dentate gyrus stained immunohistochemically with plasma from a child with autism. Numerous labeled neurons (black arrows) are shown in the polymorphic cell layer, whereas fewer positive cells (open arrows) are shown in the molecular layer. Calibration bars, 100 μm. The bar in B applies to A, and the bar in D applies to C.
Figure 7
Photomicrographs of the dorsal portion of the amygdala of the macaque monkey. (A) Nissl-stained section of the dorsolateral portion of the amygdala. Much of the field is taken up by the central nucleus that can be divided into lateral (CEl) and medial (CEm)divisions. The basal nucleus (B) is located just ventral to the central nucleus. The substantia innominata (SI) is located just dorsal to the central nucleus. (B) Immunohistochemical staining of the amygdala using plasma from a representative child with autism. Labeled neurons are observed throughout all nuclei of the amygdala. In the central nucleus, the lateral nucleus has relatively few labeled neurons, whereas the density, size and dendritic staining of cells is much more prominent in the medial portion of the central nucleus. Calibration bar, 250 μm.
Figure 8
Photomicrographs of the caudate nucleus of the macaque monkey. (A) Nissl-stained section showing the density of striatal neurons in this region. (B) Phase-contrast photomicrograph illustrating small (approximately 10 μm in diameter), stellate neurons stained with plasma from a representative AU subject. (C) Section though the caudate nucleus reacted with a monoclonal antibody to GABA. Immunoreactivity is associated with numerous axonal varicosities as well as a variety of neuronal cell types. The small, stellate cell type observed by plasma immunohistochemical staining is also shown in the GABA preparations (black arrow). Other, larger GABAergic neurons (white arrow) can also be observed. Calibration bar, 10/100 μm.
Figure 9
Cellular profiles are also shown in the subcortical white matter in sections in which plasma stained GABAergic neurons. (A) Phase-contrast photomicrograph of AU-immunoreactive, presumed microglial cells located in the white matter deep within the somatosensory cortex. (B) Similar region from a section reacted with plasma from a representative typically developing child in which no microglial labeling is observed. Calibration bar, 25 μm.
Figure 10
Illustrations of plasma staining of the mouse somatosensory cortex. (A) Nissl-stained section of the mouse somatosensory cortex. Layers I to VI are indicated on the right-hand side of the image. (B) Autoantibody staining of a section adjacent to the Nissl-stained section depicted in A with plasma from a representative AU subject. There are numerous immunoreactive neurons (arrows) scattered throughout all layers of the cortex. Calibration bar in B = 250 μm and applies to A as well.