Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2017 Dec 20.
Published in final edited form as: Cell Rep. 2016 Dec 20;17(12):3115–3124. doi: 10.1016/j.celrep.2016.11.067

Beyond Epilepsy and Autism: Disruption of GABRB3 Causes Ocular Hypopigmentation

Ryan J Delahanty 1,2, Yanfeng Zhang 1, Terry Jo Bichell 3, Wangzhen Shen 3, Kelienne Verdier 3, Robert L Macdonald 3,4,5,6, Lili Xu 3, Kelli Boyd 7, Janice Williams 8, Jing-Qiong Kang 3,6,9,*
PMCID: PMC5240804  NIHMSID: NIHMS836970  PMID: 28009282

Abstract

Reduced ocular pigmentation is common in Angelman syndrome (AS) and Prader-Willi syndrome (PWS) and long thought to be caused by OCA2 deletion. GABRB3 is located in the 15q11-13 region flanked by UBE3A, GABRA5, GABRG3 and OCA2. Mutations in GABRB3 have been frequently associated with epilepsy and autism, consistent with its role in neurodevelopment. We report here a robust phenotype in the mouse in which deletion of Gabrb3 alone causes nearly complete loss of retinal pigmentation due to atrophied melanosomes, as evidenced by electron microscopy. Using exome and RNA sequencing, we confirmed that only the Gabrb3 gene was disrupted while the Oca2 gene was intact. However, mRNA abundance of Oca2 and other genes adjacent to Gabrb3 is substantially reduced in Gabrb3−/− mice, suggesting complex transcriptional regulation in this region. These results suggest that impairment in GABRB3 downregulates OCA2 and indirectly causes ocular hypopigmentation and visual defects in AS and PWS.

Keywords: GABAA receptors, GABRB3, OCA2, deletion, epilepsy, autism, eye, hypopigmentation

Graphical abstract

graphic file with name nihms836970f5.jpg

Introduction

It is well documented that reduced ocular pigmentation and other visual defects are common symptoms of Angelman syndrome (AS) and Prader-Willi syndrome (PWS) (Fryburg et al., 1991; Michieletto et al., 2011). However, the regulatory mechanisms that control gene expression in the 15q11-q13 chromosome region which includes UBE3A, GABRB3, GABRA5, GABRG3 and OCA2, and the alterations in genes involved in AS and PWS, are complex and are not well understood (Saitoh et al., 1994). The exact function of the encoded OCA2 melanosomal transmembrane protein is unknown, but it is believed that this protein may transport molecules in and out of melanosomes where melanin is produced. The failure to transport melanin by OCA2 results in hypopigmentation. The hypopigmentation noted in AS and PWS, as well as that of oculocutaneous albinism type II, is thought to be caused simply by deletion or mutation of OCA2, formerly called the P gene (Rinchik et al., 1993), but inconsistent data exist. Hypopigmentation is present when OCA2 is included in the deleted region with its flanking genes, but not with impairment in OCA2 alone (Spritz et al., 1997), suggesting some unknown mechanisms underlying the phenotype of hypopigmentation in AS and PWS. Additionally, maternal deletion of UBE3A and mutations in UBE3A alone cause AS, thus further complicating the pathogenesis of ocular pigmentation (Michieletto et al., 2011).

GABRB3 is located in the middle of the AS and PWS deletion region 15q11–13, flanked by UBE3a, GABRA5, GABRG3 and OCA2. GABRB3 is an important neurodevelopmental gene and is abundantly expressed in the early mammalian brain. Its encoded β3 subunit is important for GABAA receptor assembly, trafficking and brain development, including stem cell proliferation (Kang and Barnes, 2013; Andang et al., 2008). Genetic abnormalities in GABRB3 have been associated with multiple brain disorders including AS, PWS (Fridman and Koiffmann, 2000; Woodage et al., 1994), childhood absence epilepsy, autism and cognitive disabilities (Tanaka et al., 2012b; Delahanty et al., 2011; DeLorey and Olsen, 1999), but the association of GABRB3 with defects in organs other than brain has not been reported.

We report here a very robust phenotype in the mouse in which deletion of Gabrb3 alone via disruption of its promoter and exons 1–3 caused almost complete loss of eye pigmentation. Eyes from wild-type mice in the C57BL/6J background appeared to be black while those from heterozygous Gabrb3+/− littermates appeared to be brown, and eyes from homozygous Gabrb3−/− littermates were pale yellow. With electronic microscopy, we demonstrated that melanosomes in the mutants were atrophied, and almost no mature melanosomes were detected in the retinal pigment cells in the homozygous Gabrb3−/− mouse eyes, while mature melanosomes were reduced in the heterozygous Gabrb3−/− mouse eyes. We confirmed with exome sequencing and transcriptome analysis RNA sequencing (RNA-Seq) that Gabrb3 was the only gene that was disrupted in this model while Oca2 gene structure was intact. Disruption of the Gabrb3 promoter with siRNAs also down-regulated the transcripts of Oca2 in mouse neuronal cultures. This thus suggests a mechanism in which deletion of Gabrb3, but not Oca2, could also result in the failure of melanin transport by downregulating Oca2 transcription, thus leading to hypopigmentation. This extends the biological role of GABRB3 beyond epilepsy and autism and provides critical insights into understanding the visual defects in AS and PWS.

Results

Deletion of Gabrb3 is associated with ocular hypopigmentation in both humans and mice

Deletion positive patients with Angelman syndrome often exhibit reduced retinal pigmentation, as illustrated by images of the eye of a patient with a Class II maternal deletion in chromosome 15q11-13 including GABRB3 and OCA2, compared to his unaffected parents and sibling (Figure 1A). We made an unexpected finding in Gabrb3 knockout mice, that heterozygous Gabrb3+/− mice had reduced retinal pigmentation while all the homozygous Gabrb3−/− mice had no visible eye pigmentation (Figure 1B) when bred in multiple different mouse genetic backgrounds including the darkly pigmented C57BL/6J.

Figure 1. Deletion of Gabrb3 including the promoter region is associated with ocular hypopigmentation in both humans and mice.

Figure 1

Open in a new tab

(A). Representative eye images of the normal parents and sibling of a child with Angelman syndrome (AS) caused by a Class II maternal deletion in chromosome 15 containing the GABRB3 and OCA2 genes. The eye color of the AS child as indicated by blue arrow was lighter than that of both parents and the sibling. (B). Images of wild-type (wt) Gabrb3+/+, heterozygous (het) Gabrb3+/−, and homozygous (hom) Gabrb3−/− knockout (KO) mice at post-natal day 0 (P0). (C). Images of dissected eyeballs from P0 wt, het and hom Gabrb3 KO mice. (D). A close up image of the P0 wt, het and hom Gabrb3 KO mouse eyes. There was no obvious pigment in the hom eyes as indicated by blue arrows. (E). A schematic illustration of GABRB3 knockout construct, ie. a part of the Gabrb3 promoter and exon 1–3 were replaced by neomycin phosphotransferase (PGKneo) in Gabrb3 KO mice. (F). Representative blots for protein expression of α5 subunit (∼50 kDa), β3 subunit (∼50 kDa) and UBE3a (∼90 kDa) in cerebellar lysates of P0 wt, het and hom Gabrb3 KO mice. The specific band in each gel was highlighted by blue box. Gapdh (∼37 kDa) used as loading control. (G). Quantification of relative α5, β3 and UBE3a protein integrated density values (IDVs) in Gabrb3 KO mice (n = 4). (H). Fontana Masson (FM) staining indicating that P0 hom Gabrb3 KO mice had no visible melanin in retinal pigment epithelial (RPE) cells, a layer which was distinct in wt, as indicated by the blue arrow. (I). Histogram showing P0 het and hom Gabrb3 KO mice had reduced melanin intensity in retinal pigment epithelial cells from FM staining compared to P0 wt littermates (n = 6).

Eyeballs from post-natal day zero (P0) wild-type Gabrb3+/+ C57BL/6J mice appeared to be black while those from heterozygous Gabrb3+/− littermates appeared to be brown, and eyes from homozygous Gabrb3−/− littermates were pale yellow or white (Figure 1C). Consistently, the eyes in the wild-type and the heterozygous mice were black or brown while the eyes in the homozygous pups appeared to be pink (Figure 1D). As previously described (Homanics et al., 1997), a part of the promoter and exon 1 and 3 of Gabrb3 were replaced with the selection marker neomycin phosphotransferase (PGKneo) in the original construction of the Gabrb3 knockout mouse model (Figure 1E).

Total protein expression of the β3 subunit encoded by Gabrb3 in heterozygotes was reduced (0.47 ± 0.06 for het vs 1= for wt, n = 5) and no β3 subunit protein was detected in homozygotes. Total protein expression of the α5 subunit encoded by Gabra5 in heterozygotes and homozygotes was also reduced (0.56 ± 0.08 for het and 0.16 ± 0.05 for hom vs 1= for wt, n = 5). Because mutation in UBE3A alone is also associated with AS, and unsilencing the dormant allele of UBE3A has been proposed as a therapeutic strategy for AS (Huang et al., 2012), we sought to determine the protein expression of Ube3a in the Gabrb3 knockout mice and found that Ube3a protein level was unaltered in both heterozygous and homozygous Gabrb3 knockout mice (Figure 1F, G) (1.02 ± 0.01 for het, 0.99 ± 0.03 for hom vs 1 for wt, n = 5).

Retinal pigment epithelial (RPE) cells are the major melanin producing cells in the eye. We thus assayed the melanin in RPE cells. The proportion of homozygous mice without visible retinal pigment by both hematoxylin and eosin (H&E) staining and by visual observation was nearly 100% (n = 22 from 11 litters and 5 different dams), while about 9% of the heterozygous mice had no visible eye pigmentation (n= 35, 3/35). We then determined the melanin in the retinal pigment cell layer with Fontana Masson staining, which specifically stains melanin (Figure 1H). The average pigment intensity value in the heterozygous mice was 68% of the wild-type mice while the value in the homozygous pups was only 3% of the wild-type mice (Figure 1I). Since mutations in UBE3A have also been associated with AS and the Ube3a maternal deletion mouse model has been proposed to be the model for AS (Jiang et al., 1998) we did parallel experiments in Ube3a knockout mice. The full knockout Ube3a−/− mice in C57BL/6J background appeared to have normal eye pigmentation. The melanin intensity was similar between the wild-type and the Ube3a−/− mice in both H&E staining and Fontana Masson staining (Figure S1).

Melanosomes were atrophied and the mature melanosomes were reduced in the mutant Gabrb3 knockout mice

Next we used transmission electron microscopy to identify melanosomes in the RPE cells. Examination of wild-type eyes revealed all four stages of melanosome development with obvious increasing pigmentation in the latter stages. The melanosomes ranged in shapes from ovoid to circular. We found that most of the melanosomes in the RPE cells of wild-type mice were mature, while there were almost no mature melanosomes in the RPE cells in Gabrb3−/− knockout mice. In the RPE cells of heterozygous Gabrb3+/− mice, the intensity of melanosomes and the number of mature melanosomes were reduced (103 ± 12, n = 25 for wt 71 ± 7, n = 55 for het, 18 ± 3, n = 40 for hom). Melanosomes of all different stages were observed in wild-type and heterozygous mice. However, there were more III-IV stage melanosomes (83 ± 4, n = 7 for wt 61 ± 7, n = 8 for het, 0 ± 0, n=8 for hom) and fewer I-II stage melanosomes (17 ± 5, n = 7 for wt 39 ± 7, n = 8 for het, 100 ± 0.10, n = 8 for hom) in the wild-type than in heterozygous and homozygous mice (Figure 2 A, B, C). There were no melanosomes at III-IV stages detected in the homozygous eye. As shown in the enlarged images, samples from both wild-type and heterozygous mice had melanosomes at different stages (Figure S2). Samples from homozygous mice, however, had only stage I immature melanosomes, and some of these melanosomes exhibited abnormal morphology. Contrary to both wild-type and heterozygous eyes, the homozygous eyes were devoid of the darker electron dense mature melanocytes (Figure S2). The ultra-structure was poorly maintained showing many open spaces not evident in wild-type eyes. The loss of tissue integrity was unlikely due to experimental procedures because the finding was consistent from multiple mice (Figure S2). It is unknown if the loss of Gabrb3 or alteration of other genes due to the disruption of Gabrb3 would affect cytoskeleton or microtubule networks which could lead to a compromise in tissue integrity. Nevertheless, this suggests that some unknown mechanisms may underlie the defects in tissue integrity especially in the homozygous mice, but this issue needs further investigation.

Figure 2. Melanosomes in Gabrb3 knockout mice were less mature, reduced in total numbers and more diffusely distributed.

Figure 2

Open in a new tab

(A). Representative images from transmission electron microscopy (TEM) of retinal pigment epithelial (RPE) cells showing melanosomes in P0 wild-type (wt) Gabrb3+/+, heterozygous (het) Gabrb3+/−, and homozygous (hom) Gabrb3−/− knockout (KO) mice. Blue arrows point to the location of melanosomes, while pink roman numerals indicate the stage of each melanosome. (B). Quantification of TEM image intensity (over background) normalized to wt. (C). Percent of total number of melanosomes per field at different stages by genotype. (D). Representative TEM images of RPE cells (2700X) from P0 wt, het and hom Gabrb3 KO eyeballs. (E). Quantification of span of melanosomes in µm by genotype. (F). Representative TEM images of RPE cells (6500X) from P0 wt, het and hom Gabrb3 KO eyeballs (indicated by the blue arrows for hom Gabrb3 sample). (G). Quantification of number of melanosomes per field from TEM images of RPE cells. The P0 het mice had reduced mature melanosomes, compared to wt, while those from hom mice had minimal to no mature melanosomes. In B, C, E and G, *p < 0.05; ** p < 0.01; *** p < 0.001, n = 4 mice for each genotype.

Melanosomes from the Gabrb3 knockout mice were more diffusely distributed and less clustered

The distribution of melanosomes in the RPE cells from wild-type mice were clustered and concentrated apically, while the melanosomes from the heterozygous mice were more diffusely distributed and less clustered. The melanosomes in RPE cells from homozygous mice were distributed more diffusely than in the wild-type and RPE cells from heterozygous mice (Figure 2D). We then measured the span of the melanosome distribution which is the width of the layer of melanosome for each genotype and found that the span was longest in homozygous mice (0.54 ± 0.02 µm, n = 4 for wt 0.78 ± 0.07 µm, n = 4 for het, 1.2 ± 0.14 µm, n = 4 for hom) (Figure 2E). The melanosomes in the wild-type were localized to a single layer within the eye. In the heterozygous eyes, the overall ultra-structural organization was less precise than wild-type and the melanosomes were not confined to a single layer within the eye. In the homozygous eye, the early stage immature melanosomes were not confined to a distinct layer within the eye, but rather extended into the deeper layers. This suggests abnormalities in cellular architecture but needs further elucidation. In the mutant mice, there was a reduction in the total number of melanosomes, including the stage I-IV melanosomes) (29.57 ± 1.6, n = 7 for wt 19.9 ± 1.56, n = 10 for het, 1.71 ± 0.36, n = 7 for hom) (Figure 2 F, G). It is interesting that the distribution pattern of melanosomes is similar between the wild-type and the heterozygous but was very different for the homozygous mice. It is likely that there are some unknown compensatory cellular mechanisms that could attenuate the phenotype in the heterozygous but not in the homozygous, mice. A similar phenomenon has been noticed in the phenotype for another GABAA receptor subunit gene GABRG2. In Gabrg2 knockout mice, the homozygous knockout was lethal while the heterozygous knockout only had mild absence epilepsy and anxiety but a normal survival (Warner et al., 2016; Xia et al., 2016). The lack of pigmentation in the Gabrb3−/− mice is likely to be due to the defects in melanin transport and melanosome maturation instead of melanin synthesis because there was no change with tyrosine hydroxylase staining (Figure S3).

Only Gabrb3 transcripts were absent in Gabrb3 knockout mice but multiple genes in eye morphogenesis and amine binding were upregulated while genes in neuroactive ligand receptors, calcium signaling pathway and steroid synthesis were downregulated in Gabrb3 knockout mice

GABRB3 is flanked by UBE3A, GABRA5, GABRG3 and OCA2 in both human and mouse (Figure 3A). We then performed high-throughput RNA-Seq for the total brain tissues to determine gene expression at the transcriptional level (Figure 3B). Only Gabrb3 was substantially down-regulated in the hom Gabrb3−/− mouse (−11.19 fold) while other flanking genes had minimal downregulation compared with the wild-type (−0.31 for Oca2, −0.08 for Gabrg3, −0.13 for Gabra5 and −0.24 for Ube3a) (Figure 3B). As indicated in the heat map and the gene ontology results (Figure 3C, Figure S3 and Table S1), the relative expression of many genes was altered, especially those involved in eye morphogenesis, amine binding, neuroactive ligand receptor interaction, calcium signaling and neurotransmitter receptor activity and steroid biosynthesis. However, the expression of genes related to pigment synthesis were unchanged, as were GABAA receptor genes in other chromosomes. Unaltered expression of pigment synthesis related genes further validates that the reduced pigmentation is due to defects in melanin transport and melanosome maturation instead of melanin synthesis. However, the RNA-Seq data suggest that the Gabrb3 deletion has far-reaching impact on neurodevelopment and the significance of these extensive gene alterations needs to be further explored (Table S1).

Figure 3. Only Gabrb3 transcripts were absent in RNA-Seq but disruption of Gabrb3 promotor by siRNAs down-regulated Oca2 gene expression in mouse neurons.

Figure 3

Open in a new tab

(A) Schematic illustration of Gabrb3 and flanking genes in mouse chromosome 7. Magenta arrows indicate the transcription direction of each gene. (B). Fold-change of transcripts from genes flanking Gabrb3 from RNA-sequencing (RNA-Seq) in P0 homozygous (hom) Gabrb3−/− knockout (KO) mice when normalized to wild-type (wt) Gabrb3+/+ values. (C). The heatmap demonstrates that expression of many genes was altered in the P0 hom Gabrb3 mice compared with the wt and heterozygous (het) mice as reflected by color key. (D). Schematic illustration of Gabrb3 and flanking Oca2 in mouse chromosome 7. (E, F). The relative mRNA abundance of Gabrb3 (E) or Oca2 (F) in two week old mouse cortical neurons treated with or without siRNAs for 4 days measured by qRT-PCR. Con indicates untreated control neurons, mock indicates neurons treated with scrambled siRNAs. Promoter signifies neurons treated with siRNAs targeting the Gabrb3 promoter region while exon 4 signifies those targeting Gabrb3 exon 4. (G). Neuronal cultures were transfected with HA tagged human GABRB3 subunit (GABRB3HA) cDNAs for 7–8 days. Relative abundance of Oca2 and Gabrb3 transcripts in neurons from Gabrb3 knockout pups of each genotype with or without over-expression of GABRB3HA cDNA. In E, F and G, *p < 0.05; ** p < 0.01; *** p < 0.001 vs wt.

Silencing Gabrb3 against its promoter but not the middle coding region down-regulated Oca2 transcripts

What is the genetic basis for a disruption of Gabrb3 causing hypopigmentation? We hypothesized that it was due to the disruption of Gabrb3 that indirectly down-regulated Oca2 but not the impairment of β3 subunit itself. We then designed siRNAs targeting the promoter region and coding region of Gabrb3 to determine the effect on transcription of Gabrb3 and Oca2 in mouse neuronal culture (Figure S3, Figure 3D). We found that disruption of either the promoter region or of exon 4 by siRNAs down-regulated the expression of Gabrb3 (0.11 ± 0.07 for promoter, 0.24 ± 0.18 for exon 4 vs 0.97 ± 0.15 for scramble, n = 4 cultures) (Figure 3E). However, only disruption of the promoter region, but not the coding region, of exon 4 (Figure 3F) specifically downregulated the transcripts of Oca2 (0.25 ± 0.16 for promoter, 0.79 ± 0.23 for exon 4 vs 1.06 ± 0.09 for scramble, n = 4 cultures).

We then determined if an increase of β3 subunit protein and GABAA receptor function would up-regulate Oca2 expression. We thus over-expressed human GABRB3 cDNA with an HA tag, which prevents it from being detected by the probes against mouse Gabrb3 transcripts because of the insertion of HA tag in the probed region. We demonstrated that increasing the β3 subunit protein by over-expression of the GABRB3HA subunit cDNAs in neurons did not upregulate the expression of Oca2 (0.68 ± 0.18 for het and 0.27 ± 0.06 for hom without β3HA subunit protein vs 0.58 ± 0.12 for het and 0.24 ± 0.08 for hom with β3HA subunit protein) or Gabrb3 transcripts (0.46 ± 0.12 for het 0.04 ± 0.01 for hom without β3HA subunit protein vs 0.49 ± 0.08 for het 0.05 ± 0.01 for hom with β3HA subunit protein) in the neuronal cultures (Figure 3G). This thus suggests that the downregulation of Oca2 in Gabrb3 knockout mice was not due to loss of β3HA subunit protein and the function of GABRB3 subunit containing receptor per se. It is instead likely that some unknown transcriptional mechanisms involving the Gabrb3 promoter region which modulates the expression of the downstream gene like Oca2. It has been reported that GABRB3 has a strong promoter (Kirkness and Fraser, 1993). It is possible that the GABARB3 promoter or its associated binding proteins may regulate the transcription of adjacent genes. However, this requires further elucidation.

Though Gabrb3 is the only deleted gene, transcripts of the downstream genes such as Gabra5, Gabrg3 and Oca2 were downregulated

Because the chromosome 15q11-13 region is unstable, and genomic abnormalities like deletion or duplication are frequently identified in this region, we performed mouse whole exome sequencing to make sure that this mouse line had no mutations or other genetic abnormalities other than the Gabrb3 deletion. Our data verified that only the first two exons and part of exon 3 of Gabrb3 were absent, while the downstream exons of Gabrb3 and those of the flanking genes, including the 11 exons of Oca2, were intact (Figure 4 A, B). This is consistent with the original construction of the Gabrb3 knockout mouse, in which the 2.8 kb of the Gabrb3 promoter and exon 1–3 regions were replaced (Homanics et al., 1997). This thus indicated that the Oca2 gene itself was unaltered in the Gabrb3 knockout mouse and confirms that only the Gabrb3 gene region was interrupted in this mouse model.

Figure 4. Deletion of Gabrb3 including the promoter region downregulated expression of Oca2in the eye as well as in the brain.

Figure 4

Open in a new tab

(A). Whole exome sequencing of eyeball from P0 homozygous (hom) Gabrb3−/− KO mouse determined that only the early exons 1–3 in Gabrb3 were absent while the adjacent genes were intact. (B). Enlarged view of Gabrb3 and adjacent regions by exome sequencing. The green block indicates the deleted region of Gabrb3 in Gabrb3−/− mice. (C–E). Total RNA was extracted from the cerebella, eyeballs or the forebrains of the newborn wild-type (wt), heterozygous (het) and homozygous (hom) Gabrb3−/− littermates. The total RNAs were transcribed to cDNA and quantified by qRT-PCR. The plot demonstrated the Delta CT values of Oca2 expression in wt, het and hom mouse cerebella (C). (D). Relative Oca2 abundance was reduced in the Gabrb3 knockout mouse cerebella with varied cDNA concentrations (ng) when quantified by qRT-PCR. (E). Both Oca2 and Gabrb3 mRNAs were reduced in the P0 old Gabrb3 knockout mouse eyes by qRT-PCR. (F). Relative abundance of Gabrb3 and its flanking genes in the P0 old Gabrb3 knockout mouse forebrains. (G). Relative abundance of the physically unrelated Gabra1, Gabra3 and eye development essential factors Mitf, Pax6 and Wnt3a in mouse forebrains of each genotype. In E and F, *p < 0.05; ** p < 0.01; *** p < 0.001 vs wt.

What is the basis for hypopigmentation in the Gabrb3 knockout mice if the Oca2 gene is intact? RNA-Seq data indicated a minimal down-regulation of the transcripts of adjacent genes, although the Gabrb3 transcripts were absent. We then quantified transcripts of five Gabrb3 flanking genes including Ube3a, Gabra5, Gabrg3 and Oca2 via qRT-PCR in both brain and eye tissues. Expression of Oca2 mRNA was reduced in the mutant Gabrb3 mouse whole brain tissue, cerebellar tissue and eye samples (Figure 4C, D, E). The relative expression profile of Oca2 in eye tissue (Figure 4E) (0.46 ± 0.07 for het, 0.12 ± 0.05 for hom vs 1.0 for wt, n = 3 pairs of mice) was similar to Gabrb3 (0.53 ± 0.015 for het, 0.0 for hom vs 1.0 for wt, n = 3 pairs of mice for eye) and forebrain (Figure 4F) from each genotype. Although Oca2 mRNA abundance is low in forebrain, it is evident that Oca2 transcripts were reduced in the mutant Gabrb3 knockout compared with the wild-type mouse (Figure 4 F). Gabra5 and Gabrg3 mRNA expression was also reduced (0.65 ± 0.07 for het, 0.12 ± 0.05 for hom for Gabra5, and 0.61 ± 0.15 for het and 0.25 ± 0.16 for hom for Gabrg3 vs 1.0 for wt, n = 3 pairs of mice), but expression of Ube3a was not reduced (Figure 4 F).

We also determined the expression levels of GABAA receptor genes that are not located in chromosome 15q11-13 but have close interaction with Gabrb3 during receptor assembly and biogenesis. These include Gabra1 and Gabra3 as well as a few essential genes for neuronal and eye development, including Mitf, Pax6, Wnt3A, to exclude any defect of these essential transcriptional factors that could explain alterations in RPE cell differentiation(Martinez-Morales et al., 2004). The data demonstrated that Gabra1 (0.98 ± 0.12 for het, 1.04 ± 0.11 for hom vs 1 for wt, n = 3 pairs of mice) and Gabra3 (1.08 ± 0.17 for het, 0.99 ± 0.11 for hom vs 1 for wt, n = 3 pairs of mice) gene expression in the Gabrb3 knockout mice was unaltered. The mRNA abundance of Mitf (1.14 ± 0.15 for het, 1.05 ± 7 for hom vs 1 for wt, n = 3 pairs of mice), Pax6 (1.21 ± 0.19 for het, 1.06 ± 0.14 for hom vs 1 for wt, n = 3 pairs of mice) and Wnt3a (1.22 ± 0.21 for het, 1.30 ± 0.20 for hom vs 1.00 for wt, n = 3 pairs of mice) was also unaltered (Figure 4G), suggesting that the hypopigmentation found in Gabrb3 knockout mice was likely due to the down-regulation of Oca2 transcripts. It is not clear why only minimal downregulation of Oca2 and other adjacent genes was observed with RNA-Seq. One possible explanation is that total RNAs for the RNA-Seq assay were from whole brain lysates while the RNAs for qRT-PCR data were acquired from more specific brain regions such as cerebellum, forebrain or eyes.

Discussion

We first identified that a deletion of Gabrb3 alone unexpectedly caused almost complete loss of eye pigment in mice. It has been commonly observed that reduced ocular pigmentation exists in AS and PWS (King et al., 1993). It has long been thought to be caused by OCA2 deletion and not related to GABRB3 deletion. However, there is a report that ocular hypopigmentation in the PWS correlates with P gene deletion but not with haplotype of the hemizygous P allele (Spritz et al., 1997), suggesting genes other than Oca2 in the region may be involved in contributing to hypopigmentation. It is well established that GABRB3 impairment is associated with epilepsy, autism and other neurodevelopmental disorders (Tanaka et al., 2012b; Delahanty et al., 2011). Deletion in chromosome 15q11-13, a region, which includes the GABRB3 gene, is the major cause of AS and PWS (Williams et al., 1995). In addition to seizures, AS and PWS often exhibit clinical features that are characteristic of tyrosine-positive (Type II) OCA albinism (Rinchik et al., 1993). Reduced retinal pigmentation in AS and PWS is associated with multiple visual defects including reduced visual acuity, nystagmus and strabismus (Fryburg et al., 1991). It had previously been concluded that this reduced pigmentation was caused by deletion of OCA2 (also known as P) (Rinchik et al., 1993) because the region of deletion in AS and PWS often includes OCA2. Our findings indicate that deletion of GABRB3 alone could result in a complex eye phenotype in addition to the epilepsy or autism phenotypes resulting from the loss of function of β3 subunit itself, by affecting the expression of adjacent genes, including OCA2.

The downregulation of genes adjacent to Gabrb3 was unexpected in Gabrb3 knockout mice, although a previous study found that mutations in a regulatory element located within the Herc2 gene inhibited Oca2 gene expression, giving rise to blue and brown eye colors (Eiberg et al., 2008). Here we have unequivocally demonstrated in Gabrb3 knockout mice, that Oca2 mRNA was substantially reduced in both the eyeball and brain tissues, suggesting that an unknown mechanism is downregulating the expression of adjacent genes flanking Gabrb3 at the transcriptional level because Oca2 gene structure was unaltered (Figure 4A and B). This may suggest a shared promoter or some unknown transcriptional mechanisms involving the GABRB3 exon 1–3 region that is critical for Oca2 gene expression. It has also been previously reported that an intron in HERC2 contains the promoter region for Oca2, suggesting complex transcriptional regulation of Oca2 by its adjacent genes (White and Rabago-Smith, 2011). The GABRB3 promoter region is critical for transcriptional factor binding (Tanaka et al., 2012a) and splicing as previously demonstrated (Kirkness and Fraser, 1993). It is possible that the GABRB3 promoter may regulate transcription of adjacent genes via unknown mechanisms.

High throughput RNA sequencing of the Gabrb3−/− knockout mouse demonstrated that only Gabrb3 transcripts were absent, while Oca2 transcripts have very low baseline expression in mouse brain. Given the fact that GABRB3 is expressed in both embryonic stem (ES) cells and neural crest stem (NCS) cells (Andang et al., 2008), which give rise to a diverse cell lineage including melanocytes, it is possible that GABRB3 deletion would cause diverse abnormalities, including cell differentiation from conception on in the embryo. This is also implicated by the previous finding that Gabrb3 deletion causes cleft palate (Homanics et al., 1997). It has been demonstrated that autocrine/paracrine GABA signaling by means of GABAA receptors negatively control ES cell and peripheral NCS cell proliferation (Andang et al., 2008). Additionally, the promoter region of Gabrb3 binds multiple transcriptional factors involved in stem cell differentiation. Future studies focusing on stem cell differentiation and cell lineage tracing in combination with in vivo Gabrb3 knockout will further elucidate the critical role of GABRB3 in brain and eye development and related neurodevelopmental disorders.

In conclusion, this study has unequivocally demonstrated in the mouse that disruption of Gabrb3 alone in the promoter region causes a downregulation of Oca2. This downregulation of Oca2 by interruption of Gabrb3 results in failure of melanin transport and arrested melanosome maturation, leading to ocular hypopigmentation. However, based on this study, the role of GABRB3 in ocular hypopigmentation is unlikely via the function of β3 subunit in GABAA receptors but via regulation of the expression of Oca2, which is the gene that is directly involved in melanin transport. Considering the conventional role of GABRB3 in epilepsy and neurodevelopmental disorders like autism, this finding extends our understanding of GABRB3 beyond those neuropsychiatric disorders.

Methods and Materials

Human eye image

The human eye images from a family of an AS patient in Figure 1A were provided by Dr. Terry Jo Bichell.

Mice

The Gabrb3 knockout mouse line has been previously generated (DeLorey et al., 1998) and was obtained from Jackson Laboratory. Genotyping was performed following the instructions from Jackson Laboratory. Heterozygous x heterozygous breeding was carried out in C57BL/6J/S129 mixed background and C57BL/6J background. The eyeball and total brain lysates from an Ube3a knockout mouse line in C57BL/6J originally derived by Jiang (Jiang et al., 1998) were kindly provided by Dr. Andre Lagrange’s lab in the Department of Neurology, Vanderbilt University Medical Center. Only P0 day old pups were included.

Eye sample preparations and immunohistochemistry

Eye samples were paraffin embedded in a desired orientation and sectioned at 5 µm. Sectioned brain slices were mounted on Superfrost Plus slides for immunohistochemical studies. The H&E, Fontana Masson staining, and immunohistochemistry were performed with standard procedures in the Vanderbilt Translational Pathology Shared Resource Core (Kang et al., 2015).

Total RNA extraction and qRT-PCR

Total RNA from both eyes and whole brain tissue of P0 pups were extracted with Trizol/chloroform. For qRT-PCR, the cDNAs of all samples were prepared with a Reverse Transcription kit (Invitrogen). We performed qRT-PCR using an Applied Biosystems 7900 with the TaqMan Universal Master Mix and with 6-carboxyfluorescein (FAM)-labeled probes (Applied Biosystems) for each specific gene and a VIC-labeled probe (Applied Biosystems) for the endogenous control, β-actin (4352341E). The specific gene expression was normalized to mouse β-actin (Applied Biosystems) (Kang et al., 2009b).

Neuronal cultures, transfections and siRNAs

Neuronal cultures and delivery of siRNAs were modified from our previous study (Kang et al., 2009a). Mice were genotyped immediately after they were born. Mouse cortice were dissected from the brains of postnatal day 0 (P0) old C57BL/6J pups. Brain tissues from pups with the same genotype were combined for cultures. Human GABRB3HA cDNAs were transfected into neurons by calcium precipitation method as previously described (Kang et al., 2010). SiRNAs were prepared according to the manufacturer’s instructions (Thermo Fisher Scientific), and siRNAs (50 nM) were transfected with Oligofectamine or Lipofectamine (Invitrogen) to cells on the 4–5th day in cultures and were harvested 7–8 days later which was around 12 days old in culture.

Statistical analysis

Protein integrated density values (IDVs) were quantified by using the Quantity One or Odyssey fluorescence imaging system (Li-Cor). Images were quantified by ImageJ software. Data were expressed as mean ± S.E.M values and analyzed with Graphpad Prism 5.0 software. Statistical significance of immunoblot, cell imaging and qRT-PCR data was determined by a Student’s unpaired t test or, if appropriate, single value t test. All analyses used an alpha level of 0.05 to determine statistical significance.

Supplementary Material

1
2
3

In Brief.

Hypopigmentation in Angelman syndrome is thought to stem from deletion of the pigment gene OCA2. Delahanty et al now find that disruption of Gabrb3 alone causes profound loss of retinal pigmentation via downregulation of Oca2 transcription. This suggests unique complexities in gene regulation in the 15q11-13 imprinted region.

Highlights.

Gabrb3 and Oca2 are located within the Angelman-Prader-Will imprinted region.

Oca2 is associated with ocular hypopigmentation while Gabrb3 encodes a GABAAR subunit.

Mice with Gabrb3 disruption have reduced eye pigmentation due to melanosome atrophy.

Disruption of the Gabrb3 promoter region alone downregulates transcription of Oca2.

Acknowledgments

This research was supported by research grants from Citizen United for Research in Epilepsy (CURE), Dravet syndrome foundation (DSF), IDEAleague (Dravet organization) and Vanderbilt Clinical and Translation Science Award (CTSA), NIH R01NS 082635 to KJQ. TB was supported by training program T32ES007028.

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Author Contributions: RD and YZ analyzed data. TB collected the human eye images and edited paper. WS collected data. KV helped on colony maintenance and genotyping. RLM edited paper. LX provided advice on eye sample preparation. KB consulted on mouse eye anatomy. JW helped collect data. JQK designed the project, performed experiments; analyzed the data and supervised the project. RD and JQK wrote the manuscript.

Reference List

  1. Andang M, Hjerling-Leffler J, Moliner A, Lundgren TK, Castelo-Branco G, Nanou E, Pozas E, Bryja V, Halliez S, Nishimaru H, Wilbertz J, Arenas E, Koltzenburg M, Charnay P, El MA, Ibanez CF, Ernfors P. Histone H2AX-dependent GABA(A) receptor regulation of stem cell proliferation. Nature. 2008;451:460–464. doi: 10.1038/nature06488. [DOI] [PubMed] [Google Scholar]
  2. Delahanty RJ, Kang JQ, Brune CW, Kistner EO, Courchesne E, Cox NJ, Cook EH, Jr, Macdonald RL, Sutcliffe JS. Maternal transmission of a rare GABRB3 signal peptide variant is associated with autism. Mol. Psychiatry. 2011;16:86–96. doi: 10.1038/mp.2009.118. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. DeLorey TM, Handforth A, Anagnostaras SG, Homanics GE, Minassian BA, Asatourian A, Fanselow MS, Delgado-Escueta A, Ellison GD, Olsen RW. Mice lacking the beta3 subunit of the GABAA receptor have the epilepsy phenotype and many of the behavioral characteristics of Angelman syndrome. J. Neurosci. 1998;18:8505–8514. doi: 10.1523/JNEUROSCI.18-20-08505.1998. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. DeLorey TM, Olsen RW. GABA and epileptogenesis: comparing gabrb3 gene-deficient mice with Angelman syndrome in man. Epilepsy Res. 1999;36:123–132. doi: 10.1016/s0920-1211(99)00046-7. [DOI] [PubMed] [Google Scholar]
  5. Eiberg H, Troelsen J, Nielsen M, Mikkelsen A, Mengel-From J, Kjaer KW, Hansen L. Blue eye color in humans may be caused by a perfectly associated founder mutation in a regulatory element located within the HERC2 gene inhibiting OCA2 expression. Hum. Genet. 2008;123:177–187. doi: 10.1007/s00439-007-0460-x. [DOI] [PubMed] [Google Scholar]
  6. Fridman C, Koiffmann CP. Origin of uniparental disomy 15 in patients with Prader-Willi or Angelman syndrome. Am. J. Med. Genet. 2000;94:249–253. doi: 10.1002/1096-8628(20000918)94:3<249::aid-ajmg12>3.0.co;2-x. [DOI] [PubMed] [Google Scholar]
  7. Fryburg JS, Breg WR, Lindgren V. Diagnosis of Angelman syndrome in infants. Am. J. Med. Genet. 1991;38:58–64. doi: 10.1002/ajmg.1320380114. [DOI] [PubMed] [Google Scholar]
  8. Homanics GE, DeLorey TM, Firestone LL, Quinlan JJ, Handforth A, Harrison NL, Krasowski MD, Rick CE, Korpi ER, Makela R, Brilliant MH, Hagiwara N, Ferguson C, Snyder K, Olsen RW. Mice devoid of gamma-aminobutyrate type A receptor beta3 subunit have epilepsy, cleft palate, and hypersensitive behavior. Proc. Natl. Acad. Sci. U. S. A. 1997;94:4143–4148. doi: 10.1073/pnas.94.8.4143. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Huang HS, Allen JA, Mabb AM, King IF, Miriyala J, Taylor-Blake B, Sciaky N, Dutton JW, Jr, Lee HM, Chen X, Jin J, Bridges AS, Zylka MJ, Roth BL, Philpot BD. Topoisomerase inhibitors unsilence the dormant allele of Ube3a in neurons. Nature. 2012;481:185–189. doi: 10.1038/nature10726. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Jiang YH, Armstrong D, Albrecht U, Atkins CM, Noebels JL, Eichele G, Sweatt JD, Beaudet AL. Mutation of the Angelman ubiquitin ligase in mice causes increased cytoplasmic p53 and deficits of contextual learning and long-term potentiation. Neuron. 1998;21:799–811. doi: 10.1016/s0896-6273(00)80596-6. [DOI] [PubMed] [Google Scholar]
  11. Kang JQ, Barnes G. A common susceptibility factor of both autism and epilepsy: functional deficiency of GABA A receptors. J. Autism Dev. Disord. 2013;43:68–79. doi: 10.1007/s10803-012-1543-7. [DOI] [PubMed] [Google Scholar]
  12. Kang JQ, Shen W, Macdonald RL. Two molecular pathways (NMD and ERAD) contribute to a genetic epilepsy associated with the GABA(A) receptor GABRA1 PTC mutation, 975delC, S326fs328X. J. Neurosci. 2009b;29:2833–2844. doi: 10.1523/JNEUROSCI.4512-08.2009. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Kang JQ, Shen W, Zhou C, Xu D, Macdonald RL. The human epilepsy mutation GABRG2(Q390X) causes chronic subunit accumulation and neurodegeneration. Nat. Neurosci. 2015;18:988–996. doi: 10.1038/nn.4024. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. King RA, Wiesner GL, Townsend D, White JG. Hypopigmentation in Angelman syndrome. Am. J. Med. Genet. 1993;46:40–44. doi: 10.1002/ajmg.1320460109. [DOI] [PubMed] [Google Scholar]
  15. Kirkness EF, Fraser CM. A strong promoter element is located between alternative exons of a gene encoding the human gamma-aminobutyric acid-type A receptor beta 3 subunit (GABRB3) J. Biol. Chem. 1993;268:4420–4428. [PubMed] [Google Scholar]
  16. Martinez-Morales JR, Rodrigo I, Bovolenta P. Eye development: a view from the retina pigmented epithelium. Bioessays. 2004;26:766–777. doi: 10.1002/bies.20064. [DOI] [PubMed] [Google Scholar]
  17. Michieletto P, Bonanni P, Pensiero S. Ophthalmic findings in Angelman syndrome. J. AAPOS. 2011;15:158–161. doi: 10.1016/j.jaapos.2010.12.013. [DOI] [PubMed] [Google Scholar]
  18. Rinchik EM, Bultman SJ, Horsthemke B, Lee ST, Strunk KM, Spritz RA, Avidano KM, Jong MT, Nicholls RD. A gene for the mouse pink-eyed dilution locus and for human type II oculocutaneous albinism. Nature. 1993;361:72–76. doi: 10.1038/361072a0. [DOI] [PubMed] [Google Scholar]
  19. Saitoh S, Harada N, Jinno Y, Hashimoto K, Imaizumi K, Kuroki Y, Fukushima Y, Sugimoto T, Renedo M, Wagstaff J. Molecular and clinical study of 61 Angelman syndrome patients. Am. J. Med. Genet. 1994;52:158–163. doi: 10.1002/ajmg.1320520207. [DOI] [PubMed] [Google Scholar]
  20. Spritz RA, Bailin T, Nicholls RD, Lee ST, Park SK, Mascari MJ, Butler MG. Hypopigmentation in the Prader-Willi syndrome correlates with P gene deletion but not with haplotype of the hemizygous P allele. Am. J. Med. Genet. 1997;71:57–62. doi: 10.1002/(sici)1096-8628(19970711)71:1<57::aid-ajmg11>3.0.co;2-u. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Tanaka M, Bailey JN, Bai D, Ishikawa-Brush Y, Delgado-Escueta AV, Olsen RW. Effects on promoter activity of common SNPs in 5’ region of GABRB3 exon 1A. Epilepsia. 2012a;53:1450–1456. doi: 10.1111/j.1528-1167.2012.03572.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Tanaka M, DeLorey TM, Delgado-Escueta A, Olsen RW. GABRB3, Epilepsy, and Neurodevelopment. 2012b [PubMed] [Google Scholar]
  23. Warner TA, Shen W, Huang X, Liu Z, Macdonald RL, Kang JQ. DIfferential molecular and behavioral alterations in mouse models of GABRG2 haploinsufficiency versus dominant negative mutations associated with human epilepsy. Hum. Mol. Genet. 2016 doi: 10.1093/hmg/ddw168. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. White D, Rabago-Smith M. Genotype-phenotype associations and human eye color. J. Hum. Genet. 2011;56:5–7. doi: 10.1038/jhg.2010.126. [DOI] [PubMed] [Google Scholar]
  25. Williams CA, Angelman H, Clayton-Smith J, Driscoll DJ, Hendrickson JE, Knoll JH, Magenis RE, Schinzel A, Wagstaff J, Whidden EM. Angelman syndrome: consensus for diagnostic criteria. Angelman Syndrome Foundation. Am. J. Med. Genet. 1995;56:237–238. doi: 10.1002/ajmg.1320560224. [DOI] [PubMed] [Google Scholar]
  26. Woodage T, Deng ZM, Prasad M, Smart R, Lindeman R, Christian SL, Ledbetter DH, Robson L, Smith A, Trent RJ. A variety of genetic mechanisms are associated with the Prader-Willi syndrome. Am. J. Med. Genet. 1994;54:219–226. doi: 10.1002/ajmg.1320540308. [DOI] [PubMed] [Google Scholar]
  27. Xia G, Pourali P, Warner TA, Zhang CQ, Macdonald L, Kang JQ. Altered GABAA receptor expression in brainstem nuclei and SUDEP in Gabrg2(+/Q390X) mice associated with epileptic encephalopathy. Epilepsy Res. 2016;123:50–54. doi: 10.1016/j.eplepsyres.2016.04.002. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

1
NIHMS836970-supplement-1.docx (14.3KB, docx)
2
NIHMS836970-supplement-2.pdf (1.2MB, pdf)
3
NIHMS836970-supplement-3.xlsx (32.5KB, xlsx)

RESOURCES