Brain & Development xxx (2015) xxx–xxx www.elsevier.com/locate/braindev
Original article
Circadian-relevant genes are highly polymorphic in autism spectrum disorder patients Zhiliang Yang a,1, Ayumi Matsumoto a,1, Kazuhiro Nakayama b, Eriko F. Jimbo a, Karin Kojima a, Koh-ichi Nagata c, Sadahiko Iwamoto b, Takanori Yamagata a,⇑ a Department of Pediatrics, Jichi Medical University, Shimotsuke, Tochigi, Japan Division of Human Genetics, Center for Molecular Medicine, Jichi Medical University, Shimotsuke, Tochigi, Japan c Department of Molecular Neurobiology, Institute for Developmental Research, Aichi Human Service Center, Kasugai, Aichi, Japan b
Received 2 December 2014; received in revised form 27 March 2015; accepted 10 April 2015
Abstract Background: The genetic background of autism spectrum disorder (ASD) is considered a multi-genetic disorder with high heritability. Autistic children present with a higher prevalence of sleep disorders than has been observed in children with normal development. Some circadian-relevant genes have been associated with ASD (e.g., PER1, PER2, NPAS2, MTNR1A, and MTNR1B). Methods: We analyzed 28 ASD patients (14 with sleep disorders and 14 without) and 23 control subjects of Japanese descent. The coding regions of 18 canonical clock genes and clock-controlled genes were sequenced. Detected mutations were verified by direct sequencing analysis, and additional control individuals were screened. Results: Thirty-six base changes with amino acid changes were detected in 11 genes. Six missense changes were detected only in individuals with ASD with sleep disturbance: p.F498S in TIMELESS, p.S20R in NR1D1, p.R493C in PER3, p.H542R in CLOCK, p.L473S in ARNTL2, and p.A325V in MTNR1B. Six missense changes were detected only in individuals with ASD without sleep disturbance: p.S1241N in PER1, p.A325T in TIMELESS, p.S13T in ARNTL, p.G24E in MTNR1B, p.G24E in PER2, and p.T1177A in PER3. The p.R493C mutation in PER3 was detected in both groups. One missense change, p.P932L in PER2, was detected only in the control group. Mutations in NR1D1, CLOCK, and ARNTL2 were detected only in individuals with ASD with sleep disorder. The prevalence of the mutations detected only single time differed significantly among all ASD patients and controls (p = 0.003). Two kinds of mutations detected only in individuals with ASD with sleep disorder, p.F498S in TIMELESS and p.R366Q in PER3, were considered to affect gene function by three different methods: PolyPhen-2, scale-invariant feature transform (SIFT) prediction, and Mutation Taster (www.mutationtaster.org). The mutations p.S20R in NR1D1, p.H542R in CLOCK, p.L473S in ARNTL2, p.A325T in TIMELESS, p.S13T in ARNTL, and p.G24E in PER2 were diagnosed to negatively affect gene function by more than one of these methods. Conclusion: Mutations in circadian-relevant genes affecting gene function are more frequent in patients with ASD than in controls. Circadian-relevant genes may be involved in the psychopathology of ASD. Ó 2015 The Japanese Society of Child Neurology. Published by Elsevier B.V. All rights reserved.
Keywords: Autism spectrum disorder; Circadian-relevant genes; Next-generation sequencing; Mutation; Single-nucleotide polymorphism; TIMELESS; Period
⇑ Corresponding author at: Department of Pediatrics, Jichi Medical University, 3311-1 Yakushiji, Shimotsuke, Tochigi 329-0498, Japan. Tel.: +81 285 58 7366; fax: +81 285 44 6123. E-mail address: [email protected] (T. Yamagata). 1 Equally contribute.
http://dx.doi.org/10.1016/j.braindev.2015.04.006 0387-7604/Ó 2015 The Japanese Society of Child Neurology. Published by Elsevier B.V. All rights reserved.
Please cite this article in press as: Yang Z et al. Circadian-relevant genes are highly polymorphic in autism spectrum disorder patients. Brain Dev (2015), http://dx.doi.org/10.1016/j.braindev.2015.04.006
2
Z. Yang et al. / Brain & Development xxx (2015) xxx–xxx
1. Introduction The genetic background of autism spectrum disorder (ASD) is highly heterogeneous with respect to copy number variation, single gene mutations, and multifactorial inheritance. Many genes are reportedly associated with ASD, and most of them have roles in synaptic formation or function [1]. Therefore, a primary pathological characteristic of ASD is synaptic dysfunction. Approximately 44–83% of children with ASD experience sleep problems. In contrast, only 10–30% of typically developing infants and children have sleep problems [2]. Autistic children also present with a higher prevalence of sleep disorders than has been observed in children with other developmental disorders, especially intellectually handicapped children [3]. The sleep problems of ASD patients include disturbances of initiation, maintenance, and both. In humans, sleep and wakefulness are regulated by an endogenous circadian clock. The circadian rhythm is established by a system of positive and negative feedback loops of clock genes. The Period genes constitute the main framework of the system. The expression of three Period genes (PER1, 2, 3) and two Cryptochrome genes (CRY1, 2) is activated by a dimer of the proteins CLOCK and ARNTL/BMAL1 [4,5]; PER2 is also able to negatively regulate the ARNTL2/BMAL2-CLOCK complex [6]. CRYs can repress the activity of CLOCK and BMAL1, and thus affect their own expression through negative feedback [7]. In addition to the PER and CRY genes, CLOCK and BMAL1 activate the expression of NR1D1/REV-ERBa, which closes another loop by repressing the expression of ARNTL/BMAL1 [8]. DBP is activated by CLOCK-BMAL1 through Eboxes and inhibited by PER and CRY, and NFIL3/ E4BP4 is a D-box negative regulator and leads to an opposite effect of DBP [9]. Roles in the clock have been proposed for other genes. For example, TIMELESS is essential for resetting the biological clock and interacts directly with the PER proteins [10], and BHLHE40/ Dec1 and BHLHE41/Dec2 are regulators of the mammalian molecular clock [11]. Casein kinase (CK), CSNK1D/CK Id, and CSNK1E/CKIe target PER/ CRY for degradation and regulate their nuclear translocation [12]. Melatonin is synthesized during the night and regulates circadian rhythm through the melatonin receptors MTNR1A and MTNR1B [13], which are expressed in the suprachiasmatic nuclei (SCN), where a master circadian clock is located in mammals. Genetic bases of some familial sleep disorders have been reported [14]. Therefore, genetic analysis of familial sleep and circadian rhythm disorders is important for identifying the specific gene(s) responsible for the regulation of sleep and circadian rhythms in humans.
PER1, NPAS2 [15], MTNR1A, and MTNR1B [16,17] are reportedly associated with ASD. PER2 is one of the candidate genes for 2q37-deletion syndrome, showing ASD and some dysmorphic features [18]. In the Database of Chromosomal Imbalance and Phenotype in Humans (DECIPHER, https://decipher.sanger.ac. uk/) [19], many circadian-related genes, ARNTL, ARNTL2, BHLHE40, CRY2, CSNK1E, DBP, MTNR1A, MTNR1B, NR1D1, PER2, and PER3, were located in the region of copy number variation (CNV) in ASD patients. Sleep is important for normal synaptic development and brain maturation [20]. Bourgeron proposed the idea that the synaptic pathway and the pathway relating to the setting of the clock are altered in ASD. Thus far, there is no direct evidence that the two pathways relate with each other. However, the melatonin system might affect the synaptic pathway through the modulation of CRE-binding protein (CREB) or the GABAergic system [21]. Therefore, it is possible that dysfunction of the circadian genes causes ASD or contributes to the ASD pathophysiology, and further understanding the underlying mechanisms involved with sleep disturbance in children with autism will serve to support a better understanding of the etiology of ASD. In this study, we screened for mutations in the coding regions of circadian-relevant genes in ASD patients with and without sleep disorders, and detected mutations in several genes. 2. Materials and methods 2.1. Samples We screened the circadian-related genes in 14 ASD patients with (Group A), 14 ASD patients without (Group B) sleep disorders, as well as in 23 control individuals. All participants were Japanese. Patients in Group A were 5 men and 9 women (age range 3–28 years), and patients in Group B were 12 men and 2 women (age range 3–19 years). ASD was diagnosed according to the Diagnostic and Statistical Manual of Mental Disorders, 4th Edition; all patients fulfilled the criteria for autism. Regarding sleep disorders, all patients had difficulty initiating sleep, and 6 patients also had difficulty with sleep maintenance. All patients had intellectual disability (ID); their IQs ranged from 14 to 75. Regarding severity of ID, 7 patients were severe, 3 were moderate, and 4 were mild to borderline in Group A; 4 were severe, 4 were moderate, and 6 were mild to borderline in Group B. Eleven patients in Group A and 8 patients in Group B displayed panic and selfinjury. Six patients and five patients, respectively, had epilepsy. Control samples were obtained from healthy individuals after informed consent was obtained.
Please cite this article in press as: Yang Z et al. Circadian-relevant genes are highly polymorphic in autism spectrum disorder patients. Brain Dev (2015), http://dx.doi.org/10.1016/j.braindev.2015.04.006
Z. Yang et al. / Brain & Development xxx (2015) xxx–xxx
2.2. Ethical statement Written informed consent was obtained from the parents of the study participants, and from the participants themselves if they were capable of providing consent. The study was permitted by the Bioethics Committee for Human Gene Analysis of Jichi Medical University. 2.3. Nucleic acid extraction All lymphocyte samples were obtained from patients with informed consent from themselves or their parents. Samples were transfected with EB virus (to establish lymphoblasts) and cultured. Genomic DNA was extracted from lymphoblasts using salting-out methods. For each sample, DNA concentrations were determined using a Nano Drop 2000 (Thermo Fisher Scientific, Waltham, MA, USA). 2.4. The next-generation sequencing analysis Eighteen circadian-relevant genes were selected as candidate genes for analysis (Table 1). Five hundred nanograms of genomic DNA was nebulized for 1 min using 30 psi of pressure, and purified using Agencourt AMPure XP beads (Beckman Coulter, Fullerton, CA, USA). Different multiplex identifiers (MID) (12 different MIDs; Titanium Rapid Library MID adaptor; Roche, South San Francisco, CA, USA) were ligated to the fragmented DNA to enable its use as a library. The adapter-ligated library was amplified through ligation-mediated (LM)-PCR using specific 454 primers. Primers were designed to amplify exons and introns near the genes. Next, 1 lg of the PCR amplification product was hybridized to the custom-designed SeqCap EZ Choice library (Roche NimbleGen, Madison, WI, USA) for 64–72 h at 47 °C in a thermocycler. The captured DNA was bound to Streptavidin M-270 beads (Invitrogen Dynal, Oslo, Norway), was amplified again by LM-PCR using the same specific 454 primers, and was diluted to 1 108 molecules/ll after the fragment sizes were verified with the Aglient High Sensitivity DNA assay. The 12 resulting libraries (each corresponding to one sample) were pooled together in equimolar amounts and combined with capture beads at a ratio of 2 molecules of each DNA library per capture bead. The pooled DNA was then amplified by emulsion PCR (emPCR), following the emPCR amplification manual for the 454 GS Junior Titanium sequencer (Roche). The bead-attached DNAs were denatured, eluted, and quantified with the provided bead counter. We mixed 500,000 enriched DNA beads with Packing Beads. Then, the Pico Titer Plate (PTP) was sequentially loaded with Prelayer Beads, DNA-Packing Beads, Postlayer Beads, and PPiase Beads. Finally, the PTP
3
was mounted in the 454 GS Junior Sequencer, and the program was run in full processing mode for shotgun sequencing (Roche). 2.5. Data analysis The reads were aligned and variants were compared with the reference genome using the 454 integrated software (GS Reference Mapper; Roche). We searched for mutations in the exons of targeted genes. A sample was considered mutated if a mutation was present in at least approximately 50% of confident reads. Genomic reference sequences were ascertained from the NCBI website (http://www.ncbi.nlm.nih.gov/gene/). Each coding polymorphism was compared with the dbSNP database (http://www.ncbi.nlm.nih.gov/snp/) and the corresponding reference (rs) number was assigned to previously identified polymorphisms (Table 2, Supplementary Table 1). The polymorphisms that were detected only in the patients were screened in approximately 100 controls. The hypothesized effects of the mutations on respective protein functions were analyzed using the Polymorphism Phenotyping v2 (PolyPhen-2) prediction tool (http://genetics.bwh. harvard.edu/pph2/), SIFT (http://sift.jcvi.org/), and Mutation Taster (mutation t@sting, http://www. mutationtaster.org/). Statistical significance was analyzed using Fisher’s exact test with IBM SPSS statistics 21 software. 2.6. Direct sequencing analysis Exons in which mutations were detected were verified by direct sequencing analysis. PCR was performed to amplify each exon with a mutation and its neighboring introns. PCR products were purified by passing them through microconcentrating centrifugal filter columns (Millipore, Bedford, MA, USA), and subjected to sequence analysis using the 3730xl DNA Analyzer (Applied Biosystems, Forster City, CA, USA). 2.7. Controls and pedigree analysis The existence of each mutation was confirmed by assessing up to 133 controls, as well as the parents of the patient with the respective mutation. 3. Results We detected 68 (33 in Table 1, 35 in Supplementary Table 2) base changes in the coding regions of 15 genes. Thirty-three base changes with amino acid changes were detected in 11 genes (Table 1), 24 base changes on eight genes were previously reported as single-nucleotide polymorphisms (SNPs), and nine base changes were not reported on the database (Table 1). Detected
Please cite this article in press as: Yang Z et al. Circadian-relevant genes are highly polymorphic in autism spectrum disorder patients. Brain Dev (2015), http://dx.doi.org/10.1016/j.braindev.2015.04.006
4
Gene
Base change
Amino acid change
SNP number
PolyPhen-2 analysis
SIFT analysis
Mutation Taster analysis
Group A (n = 14)
Group B (n = 14)
Control individuals (n = 23)
ARNTL ARNTL2 CSNK1E CLOCK MTNR1A
c.38G > C c.1418T > C c.1484G > A c.2551A > G c.178G > A
p.S13T p.L473S p.S408N p.H542R p.R54 W
– – rs77945315 rs3762836 rs1800885
Tolerated Tolerated Damaging Tolerated Damaging
Disease Disease Disease Disease Disease
causing causing causing causing causing
0 1 2 1 1
1 0 1 0 2
0 0 3 0 1
MTNR1A MTNR1B MTNR1B NR1D1 PER1 PER1 PER2 PER2 PER2 PER2
c.488G > A c.174G > A c.974G > A c.58A > C c.2884G > C c.3722G > A c.2863T > C c.2795C > T c.3257T > C c.3682G > C
p.A157V p.G24E p.A325V p.S20R p.A962P p.S1241 N p.F876L p.P932L p.M1086T p.P1228A
rs1800884 rs8192552 – – rs2585405 – rs78832829 – rs150509214 –
Tolerated Tolerated Tolerated Tolerated Tolerated Tolerated Tolerated Tolerated Tolerated Tolerated
Polymorphism Polymorphism Polymorphism Disease causing Polymorphism Polymorphism Polymorphism Polymorphism Polymorphism Polymorphism
0 0 1 1 4 0 1 0 0 0
1 1 0 0 14 1 1 0 1 1
3 0 0 0 10 0 2 1 1 0
PER2 PER3
c.3968G > A c.1361G > A
p.G1244E p.R366Q
rs934945 rs191016322
Tolerated Damaging
Polymorphism Disease causing
2 1
7 0
8 0
PER3 PER3
c.1477C > T c.2204T > G
p.R493C p.V647G
– rs10462020
Tolerated Tolerated
Polymorphism Polymorphism
1 1
1 0
0 1
PER3 PER3 PER3 PER3 PER3
c.2768T > C c.2854C > G c.3310G > A c.3320C > T c.3326C > T
p.L835P p.P864A p.A1016T p.T1019I p.S1021L
rs228696 rs228697 rs1776342 rs12033719 rs12750400
Tolerated Tolerated Tolerated Tolerated Tolerated
Polymorphism Polymorphism Polymorphism Polymorphism Polymorphism
14 2 8 4 3
14 4 9 5 4
23 4 8 7 4
PER3 PER3 PER3 PER3 TIMELESS
c.3374T > C c.3380T > C c.3737A > G c.3793A > G c.1141G > A
p.M1037T p.L1039S p.H1158R p.T1177A p.A325T
rs2640909 rs77477722 rs10462021 rs35072750 rs113125615
Tolerated Tolerated Tolerated Tolerated Tolerated
Polymorphism Polymorphism Polymorphism Polymorphism Disease causing
2 0 1 0 0
3 1 0 1 1
4 2 1 0 0
TIMELESS TIMELESS
c.1531A > T c.1493T > C
p.I455L p.F498S
rs774027 –
Tolerated Damaging
Polymorphism Disease causing
13 1
14 0
14 0
TIMELESS TIMELESS
c.2660G > A c.3221C > T
p.R831Q p.P1018L
rs774047 rs2291739
Benign Benign Benign Benign Probably damaging Benign Benign Benign Benign Benign Benign Benign Benign Benign Probably damaging Benign Probably damaging Benign Probably damaging Benign Benign Benign Benign Probably damaging Benign Benign Benign Benign Probably damaging Benign Probably damaging Benign Probably damaging
Tolerated Tolerated
Polymorphism Polymorphism
12 9
13 12
18 20
Z. Yang et al. / Brain & Development xxx (2015) xxx–xxx
Please cite this article in press as: Yang Z et al. Circadian-relevant genes are highly polymorphic in autism spectrum disorder patients. Brain Dev (2015), http://dx.doi.org/10.1016/j.braindev.2015.04.006
Table 1 Detected mutations and resulting amino acid changes in ASD patients and controls.
Group
Sample code
Gene
Base change
Amino acid change
Inheritance
Control screening
SNP number
PolyPhen-2 analysis
SIFT analysis
Mutation Taster analysis
Group A
AWS1
TIMELESS
c.1493T > C
p.F498S
Maternal
0/132
None
Damaging
Disease causing
AWS2
NR1D1
c.58A > C
p.S20R
Paternal
0/133
None
Probably damaging Benign
Tolerated
Disease causing
AWS7 AWS10 AWS12
PER3 CLOCK ARNTL2 PER3
c.1477C > T c.2551A > G c.1418T > C c.1361G > A
p.R493C p.H542R p.L473S p.R366Q
N/A Maternal Paternal Paternal
0/92 0/106 0/78 0/98
None rs3762836 None rs191016322
Benign Benign Benign Probably damaging Benign
Tolerated Tolerated Tolerated Damaging
Polymorphism Disease causing Disease causing Disease causing
Tolerated
Polymorphism
Tolerated Tolerated Tolerated
Polymorphism Polymorphism Disease causing
Tolerated Tolerated Tolerated
Disease causing Polymorphism Polymorphism
FHIT (3p21) deletion
Tolerated
Polymorphism
CTNNA3 (10q21) deletion, paternal
Tolerated
Polymorphism
Group B
Control
AWS13
MTNR1B
c.974G > A
p.A325V
Maternal
0/85
None
AWNS1 AWNS3 AWNS4
PER3 PER1 TIMELESS
c.1477C > T c.3722G > A c.1141G > A
p.R493C p.S1241 N p.A325T
Paternal N/A N/A
0/92 0/89 0/103
None None rs113125615
AWNS6 AWNS7 AWNS8
ARNTL MTNR1B PER2
c.38G > C c.174G > A c.3682G > C
p.S13T p.G24E p.P1228A
N/A N/A N/A
0/92 0/118 0/94
None rs8192552 None
AWNS14
PER3
c.3793A > G
p.T1177A
Maternal
0/116
rs35072750
Benign Benign Probably damaging Benign Benign Probably damaging Benign
C15
PER2
c.2795C > T
p.P932L
N/A
1/103
None
Benign
Group A: ASD patients with sleep disorders (AWS); Group B: ASD patients without sleep disorders (AWNS).
Other findings
SHANK2 c2646C > T p.R835Q maternal
Z. Yang et al. / Brain & Development xxx (2015) xxx–xxx 5
Please cite this article in press as: Yang Z et al. Circadian-relevant genes are highly polymorphic in autism spectrum disorder patients. Brain Dev (2015), http://dx.doi.org/10.1016/j.braindev.2015.04.006
Table 2 Mutations from the analysis of circadian-relevant genes in ASD patients with sleep disorders (Group A), ASD patients without sleep disorders (Group B), and controls.
6
Z. Yang et al. / Brain & Development xxx (2015) xxx–xxx
non-synonymous SNPs number were about two times more in ASD than control group; 86 in Group A (n = 14), 113 in Group B (n = 14) and 135 in control (n = 23). Thirteen SNPs were predicted as damaging at least one program, and detected numbers were 21 in Group A, 22 in Group B and 29 in control. The prevalence of nine SNPs that were not reported previously in database differed significantly between all ASD patients and controls (p = 0.006); five in Group A, four in Group B and one in control. Also, 12 SNPs were appeared only single time among samples, and the prevalence differed significantly between all ASD patients and controls (p = 0.003); five in Group A, six in Group B and one in Control. We focused on the mutations that were not reported as SNPs and not listed in the database previously, as well as SNPs with amino acid changes detected only in ASD participants or controls (Table 2). Seven types of mutations in six circadianrelated genes were detected in six patients in ASD patients with sleep disorders (Group A). In ASD patients without sleep disorders (Group B), seven patients exhibited seven types of mutations in six genes. One mutation, p.R493C on PER3, was detected in both Group A and Group B. Only one mutation was detected in one of 23 control individuals, but not in any participants with ASD. These mutations, which were detected in groups of individuals with ASD, were screened in up to 133 controls and were not detected. This finding strengthened the association of these mutations with ASD. PolyPhen-2 analysis categorized two mutations in Group A, two mutations in Group B, and zero mutations in the control group as “probably damaging”. SIFT analysis suggested that only two mutations in Group A were “damaging”. Mutation Taster analysis indicated seven mutations, five in Group A and two in Group B, as “disease causing”. All three methods indicated that two mutations in Group A affected gene function, and at least one method indicated that five mutations in Group A and three mutations in Group B affected gene function (Table 2). In Group A, detected mutations that were not associated with reported SNPs were p.F498S in TIMELESS, p.S20R in NR1D1, p.R493C in PER3, p.L473S in ARNTL2, and p.A325V in MTNR1B (Table 2). The p.H542R mutation in CLOCK and p.R366Q in PER3 were previously reported as SNPs, but were detected only in ASD patients and not in controls. In Group B, there were four detected mutations that were not associated with reported SNPs: p.R493C in PER3, p.S1241N in PER1, p.S13T in ARNTL1, and p.P1228A in PER2. Detected mutations associated with previously reported SNPs were p.A325T in TIMELESS, p.G24E in MTNR1B, and p.T1177A in PER3. Among the SNPs detected in Group A, p.F498S in TIMELESS and p.R366Q in PER3 were considered to affect gene
function by PolyPhen-2, SIFT and Mutation Taster analyses, and p.S20R in NR1D1, p.H542R in CLOCK, and p.L473S in ARNTL2 were considered “disease causing” by Mutation Taster analysis. In Group B, p.A325T in TIMELESS were considered to affect gene function by PolyPhen-2 and Mutation Taster, and p.S13T in ARNTL and p.P1228A in PER2 were considered to affect gene function by Mutation Taster and PolyPhen-2 analyses, respectively (Table 2). The patient AWS1, a 9-year-old boy with p.F498S in TIMELESS, had ASD phenotype with mild disturbance of sociability, delayed speech development and severely restricted interest, and mild intellectual disability (ID). His sleep pattern was inverted so that he had been falling asleep in the morning at approximately 0500–0700 h since infancy. This mutation was also detected in his mother, who had also experienced sleep disturbance since childhood. The mutation p.S20R in NR1D1 was detected in AWS2, a 23-year-old female patient with typical ASD and severe ID who had trouble with sleep initiation. We had detected the SHANK2 mutation of c.2646C > T (p.R835Q) in this patient during previous autism candidate gene screening. The NR1D1 mutation that was considered disease-causing by Mutation Taster analysis was inherited from her father, and the SHANK2 mutation that was considered likely damaging by PolyPhen-2 analysis was inherited from her mother (Table 2). Her parents did not exhibit any obvious phenotype of ASD or sleep disturbance. We detected p.L473S in ARNTL2 and p.R366Q in PER3 in AWS12, a 4-year-old boy with ASD and mild ID, and both base changes were inherited from his father. This patient experienced strong disturbances of both sleep initiation and sleep maintenance. The p.R366Q mutation in PER3 is thought to negatively affect gene function by three methods. The p.R493C mutation in PER3 was detected in one patient in Group A and one patient in Group B. This SNP was not reported previously, but was considered as benign. The p.P1228A mutation in PER2, which was considered damaging, was detected in one patient in Group B. AWNS14 had inherited a PER3 mutation from her mother, as well as a CTNNA3 deletion from her father that was detected by array comparative genomic hybridization (aCGH). AWNS6 had an ARNTL mutation and also FHIT deletion in aCGH, although inheritance was not defined. In addition to these findings in ASD patients, several base changes that were also detected in control individuals were indicated to affect gene function by at least one method. Among these, p.R54W in MTNR1A was detected in three patients with ASD and one control individual, and was indicated to affect gene function by three methods. The p.S408N mutation in CNSKIE, which two methods considered pathogenic, was detected in three ASD patients and three control participants.
Please cite this article in press as: Yang Z et al. Circadian-relevant genes are highly polymorphic in autism spectrum disorder patients. Brain Dev (2015), http://dx.doi.org/10.1016/j.braindev.2015.04.006
Z. Yang et al. / Brain & Development xxx (2015) xxx–xxx
Two mutations in PER3 and one in TIMELESS that were detected in both ASD patients and control participants were considered probably damaging by PolyPhen-2 (Table 1). 4. Discussion This study is the first to document screening for mutations in circadian-relevant genes in ASD patients with or without sleep disorders. Genes detected in this study were candidate genes for circadian abnormality and ASD. We detected several mutations/SNPs in the ASD patients, and rare variant SNPs are much frequently found in ASD group comparing to the control group. It was suggested that ASD group has unique and different SNPs pool comparing to the control population. In addition, the existence of five different mutations in patients with sleep disorder that were indicated to affect gene function suggests that circadian genes are related to some mechanism of ASD and sleep disorder. In our analysis, SNPs of PER1, PER2 and PER3 were detected more frequently in ASD patients than in controls. Among these, functional analysis of gene mutations led us to believe that p.P1228A in PER2 and p.R366Q in PER3 negatively affect gene function. These results implicate PER proteins, especially PER2 and PER3, in the pathogenesis of ASD; they likely control gene expression through the E-box, which is functional in brain tissue. The biological function of TIMELESS is essential for resetting the biological clock, which interacts directly with the PER proteins [22]. Studies have also revealed roles for TIMELESS in bipolar disorder type 1, schizophrenia, and schizoaffective disorder [23]. We found the mutation p.F498S in the TIMELESS gene in an ASD patient who has been falling sleep in the morning since infancy, and also in his mother, who had also experienced sleep disturbance. The mutation p.F498S was indicated to affect gene function by all three methods. We are currently obtaining results that this mutation affects neuronal function (in preparation), and preparing a functional analysis of TIMELESS using knockout mice. NR1D1/Rev-erba knockout mice, which lack the nuclear receptor Rev-erba (a potent transcriptional repressor and core clock component), displayed marked hyperactivity and impaired response habituation in novel environments [24]. The NR1D1 mutation detected in one of our patients was also detected in her father. During our previous candidate gene screening for ASD, we detected a mutation in SHANK2 in this patient and her mother. SHANK2 is a synaptic scaffold protein and a homologue of SHANK3, and de novo mutations and inherited mutations from unaffected parents reportedly contribute to the pathogenesis of ASD [25]. The p.S20R mutation in NR1D1 was considered diseasecausing only by Mutation Taster and not by other
7
methods. However, our findings also suggest that this mutation affects neuronal function, and we are analyzing the effects of the combination of these two gene mutations on the pathogenesis of the patient’s phenotypes. As another example of two alternating genes, previous aCGH analysis detected CNVs in two patients. AWNS6 with ARNTL mutation had partial deletion of FHIT, and AWNS14 with PER3 mutation had partial deletion of CTNNA3. FIHT was reportedly included in recurrent CNVs in ASD [26]. Compound heterozygote deletion of CTNNA3 was reported in ASD patients [27] and a possible association of hemizygote deletion/mutation with ASD was proposed [27,28]. In ASD, the inability to engage in a flow of either verbal or non-verbal interaction has been described as social timing difficulty. Interactive timing, exact synchronization with certain types of adult communication, and rapidly shifting attention from one mode to another were important for normal social and symbolic development. Biological oscillators are essential for information processing, and faster oscillators are employed for communication. It has been hypothesized that anomalies in clock genes that operate as timing genes in high-frequency oscillator systems, in addition to circadian functions, may underlie the timing deficits of autism [29]. ARNTL/BMAL1 and ARNTL2/BMAL2 interact with CLOCK and induce E-box-dependent transactivation by PER2 [6,30,31]. Studies in ARNTL/BMAL1knockout mice and ARNTL/BMAL1-ARNTL2/ BMAL2 double knockout mice have demonstrated that the ARNTL2/BMAL2 complex is more important to the circadian rhythm than BMAL1 [32]. The results of our study that unreported mutations were detected in ARNTL in ASD patients without sleep disorders and in ARNTL2 in ASD patients with sleep disorders might be a reflection of these findings. In addition to TIMELESS and PER3, associations of ARNTL and ARNTL2 with bipolar disorder have been detected in several reports [23]. These base changes were considered disease-causing by Mutation Taster analysis, and the possible association with ASD should be elucidated further. Melatonin is synthesized during the night and is involved in various physiological functions, including sleep induction, circadian rhythm regulation, and immune response. Melatonin signaling is mainly mediated by the guanine nucleotide binding protein-coupled receptors (GPCRs) MTNR1A and MTNR1B. A previous mutation analysis study for melatonin-related genes in 109 ASD patients detected p.V124I in MTNR1B and base changes in the upstream regulatory regions of MTNR1A and MTNR1B, and suggested that these genes were interesting candidates for ASD [16]. In another study regarding the mutation of melatoninrelated genes, several base changes in MTNR1A and MTNR1B were detected at the same frequency in
Please cite this article in press as: Yang Z et al. Circadian-relevant genes are highly polymorphic in autism spectrum disorder patients. Brain Dev (2015), http://dx.doi.org/10.1016/j.braindev.2015.04.006
8
Z. Yang et al. / Brain & Development xxx (2015) xxx–xxx
participants with ASD and control participants [17]. In both studies, the p.V124I mutation in MTNR1B was detected only in ASD patients. The p.A325V mutation, which was detected in a patient in our study and was benign, was also detected in one Asian individual of the human genome pluriethnic diversity panel (HGDP). The p.R54W mutation in MTNR1A, which was considered to affect gene function by three methods, was detected in three patients with ASD and in one control participant. This finding might reflect the existence of a sleep disorder in a control participant, or it might indicate a polygenic factor underlying ASD. The relationship between these rare variants and ASD should be investigated further. Although our sample size was small, our identification of the rare variant SNPs/mutations with functional alternation in ASD patients at a significantly higher frequency than in controls supports the notion that clock genes may be involved in the pathogenesis of ASD. In particular, we consider that PER2, PER3, and TIMELESS may be associated with ASD and partially related to sleep disturbance because we detected several mutations that are likely to induce damage to gene function. Our results are also supported by previous studies that implicate clock genes in autism [16]. Moreover, the presence of severe ID in a patient with the observed mutation may suggest a broader role for circadian-relevant genes in neurodevelopmental disorders. Some base changes were considered benign, and were inherited from unaffected parents. However, the detection of a combination of mutations on two genes in a single patient suggests that these genes may be working as polygenic factors. Acknowledgments This study was supported by the Program for the Strategic Research Foundation at Private Universities 2011-2015 “Cooperative Basic and Clinical Research on Circadian Medicine” from the Ministry of Education, Culture, Sports, Science and Technology of Japan, and by a Grant-in-Aid from Health Labor Sciences Research Grants from the Ministry of Health, Labor, and Welfare, Japan. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10. 1016/j.braindev.2015.04.006. References [1] Zoghbi HY, Bear MF. Synaptic dysfunction in neurodevelopmental disorders associated with autism and intellectual disabilities. Cold Spring Harb Perspect Biol 2012;4.
[2] Glickman G. Circadian rhythms and sleep in children with autism. Neurosci Biobehav Rev 2010;34:755–68. [3] Cotton S, Richdale A. Brief report: parental descriptions of sleep problems in children with autism, down syndrome, and Prader– Willi syndrome. Res Dev Disabil 2006;27:151–61. [4] Etchegaray JP, Lee C, Wade PA, Reppert SM. Rhythmic histone acetylation underlies transcription in the mammalian circadian clock. Nature 2003;421:177–82. [5] Travnickova-Bendova Z, Cermakian N, Reppert SM, SassoneCorsi P. Bimodal regulation of mPeriod promoters by CREBdependent signaling and CLOCK/BMAL1 activity. Proc Natl Acad Sci USA 2002;99:7728–33. [6] Sasaki M, Yoshitane H, Du NH, Okano T, Fukada Y. Preferential inhibition of BMAL2-CLOCK activity by PER2 reemphasizes its negative role and a positive role of BMAL2 in the circadian transcription. J Biol Chem 2009;284:25149–59. [7] Sato TK, Yamada RG, Ukai H, Baggs JE, Miraglia LJ, Kobayashi TJ, et al. Feedback repression is required for mammalian circadian clock function. Nat Genet 2006;38:312–9. [8] Preitner N, Damiola F, Lopez-Molina L, Zakany J, Duboule D, Albrecht U, et al. The orphan nuclear receptor REV-ERBalpha controls circadian transcription within the positive limb of the mammalian circadian oscillator. Cell 2002;110:251–60. [9] Yamajuku D, Shibata Y, Kitazawa M, Katakura T, Urata H, Kojima T, et al. Cellular DBP and E4BP4 proteins are critical for determining the period length of the circadian oscillator. FEBS Lett 2011;85:2217–22. [10] Barnes JW, Tischkau SA, Barnes JA, Mitchell JW, Burgoon PW, Hickok JR, et al. Requirement of mammalian Timeless for circadian rhythmicity. Science 2003;302:439–42. [11] Honma S, Kawamoto T, Takagi Y, Fujimoto K, Sato F, Noshiro M, et al. Dec1 and Dec2 are regulators of the mammalian molecular clock. Nature 2002;419:841–4. [12] Eide EJ, Woolf MF, Kang H, Woolf P, Hurst W, Camacho F, et al. Control of mammalian circadian rhythm by CKIe-regulated proteasome-mediated PER2 degradation. Mol Cell Biol 2005;25:2795–807. [13] Simonneaux V, Ribelayga C. Generation of the melatonin endocrine message in mammals: a review of the complex regulation of melatonin synthesis by norepinephrine, peptides, and other pineal transmitters. Pharmacol Rev 2003;55:325–95. [14] Reid KJ, Chang AM, Dubocovich ML, Turek FW, Takahashi JS, Zee PC, et al. Familial advanced sleep phase syndrome. Arch Neurol 2001;58:1089–94. [15] Nicholas B, Rudrasingham V, Nash S, Kirov G, Owen MJ, Wimpory DC. Association of Per1 and Npas2 with autistic disorder: support for the clock genes/social timing hypothesis. Mol Psychiatry 2007;12:581–92. [16] Jonsson L, Ljunggren E, Bremer A, Pedersen C, Lande´n M, Thuresson K, et al. Mutation screening of melatonin-related genes in patients with autism spectrum disorders. BMC Med Genomics 2010;3:10. [17] Chaste P, Clement N, Mercati O, Guillaume JL, Delorme R, Botros HG, et al. Identification of pathway-biased and deleterious melatonin receptor mutants in autism spectrum disorders and in the general population. PLoS One 2010;5:e11495. [18] Leroy C, Landais E, Briault S, David A, Tassy O, Gruchy N, et al. The 2q37-deletion syndrome: an update of the clinical spectrum including overweight, brachydactyly and behavioural features in 14 new patients. Eur J Hum Genet 2013;21:602–12. [19] Firth HV, Richards SM, Bevan AP, Clayton S, Corpas M, Rajan D, et al. DECIPHER: database of chromosomal imbalance and phenotype in humans using ensembl resources. Am J Hum Genet 2009;84:524–33. [20] Peirano PD, Algarı´n CR. Sleep in brain development. Biol Res 2007;40:471–8.
Please cite this article in press as: Yang Z et al. Circadian-relevant genes are highly polymorphic in autism spectrum disorder patients. Brain Dev (2015), http://dx.doi.org/10.1016/j.braindev.2015.04.006
Z. Yang et al. / Brain & Development xxx (2015) xxx–xxx [21] Bourgeron T. The possible interplay of synaptic and clock genes in autism spectrum disorders. Cold Spring Harb Symp Quant Biol 2007;72:645–54. [22] Sangoram AM, Saez L, Antoch MP, Gekakis N, Staknis D, Whiteley A, et al. Mammalian circadian autoregulatory loop: a timeless ortholog and mPer1 interact and negatively regulate CLOCK-BMAL1-induced transcription. Neuron 1998;21: 1101–13. [23] Mansour HA, Wood J, Logue T, Chowdari KV, Dayal M, Kupfer DJ, et al. Association study of eight circadian genes with bipolar I disorder, schizoaffective disorder and schizophrenia. Genes Brain Behav 2006;5:150–7. [24] Jager J, O’Brien WT, Manlove J, Krizman EN, Fang B, GerhartHines Z, et al. Behavioral changes and dopaminergic dysregulation in mice lacking the nuclear receptor Rev-erba. Mol Endocrinol 2014;28:490–8. [25] Berkel S, Marshall CR, Weiss B, Howe J, Roeth R, Moog U, et al. Mutations in the SHANK2 synaptic scaffolding gene in autism spectrum disorder and mental retardation. Nat Genet 2010;42:489–91. [26] Girirajan S, Dennis MY, Baker C, Malig M, Coe BP, Campbell CD, et al. Refinement and discovery of new hotspots of copy-
[27]
[28]
[29] [30]
[31]
[32]
9
number variation associated with autism spectrum disorder. Am J Hum Genet 2013;92:221–37. Bacchelli E, Ceroni F, Pinto D, Lomartire S, Giannandrea M, D’Adamo P, et al. A CTNNA3 compound heterozygous deletion implicates a role for aT-catenin in susceptibility to autism spectrum disorder. J Neurodev Disord 2014;6:17. Butler MG, Rafi SK, Hossain W, Stephan DA, Manzardo AM. Whole exome sequencing in females with autism implicates novel and candidate genes. Int J Mol Sci 2015;16:1312–35. Wimpory D, Nicholas B, Nash S. Social timing, clock genes and autism: a new hypothesis. J Intellect Disabil Res 2002;46:352–8. Zylka MJ, Shearman LP, Weaver DR, Reppert SM. Three period homologs in mammals: differential light responses in the suprachiasmatic circadian clock and oscillating transcripts outside of brain. Neuron 1998;20:1103–10. Shearman LP, Sriram S, Weaver DR, Maywood ES, Chaves I, Zheng B, et al. Interacting molecular loops in the mammalian circadian clock. Science 2000;288:1013–9. Shi S, Hida A, McGuinness OP, Wasserman DH, Yamazaki S, Johnson CH. Circadian clock gene Bmal1 is not essential; functional replacement with its paralog, Bmal2. Curr Biol 2010;20:316–21.
Please cite this article in press as: Yang Z et al. Circadian-relevant genes are highly polymorphic in autism spectrum disorder patients. Brain Dev (2015), http://dx.doi.org/10.1016/j.braindev.2015.04.006
Original article
Circadian-relevant genes are highly polymorphic in autism spectrum disorder patients Zhiliang Yang a,1, Ayumi Matsumoto a,1, Kazuhiro Nakayama b, Eriko F. Jimbo a, Karin Kojima a, Koh-ichi Nagata c, Sadahiko Iwamoto b, Takanori Yamagata a,⇑ a Department of Pediatrics, Jichi Medical University, Shimotsuke, Tochigi, Japan Division of Human Genetics, Center for Molecular Medicine, Jichi Medical University, Shimotsuke, Tochigi, Japan c Department of Molecular Neurobiology, Institute for Developmental Research, Aichi Human Service Center, Kasugai, Aichi, Japan b
Received 2 December 2014; received in revised form 27 March 2015; accepted 10 April 2015
Abstract Background: The genetic background of autism spectrum disorder (ASD) is considered a multi-genetic disorder with high heritability. Autistic children present with a higher prevalence of sleep disorders than has been observed in children with normal development. Some circadian-relevant genes have been associated with ASD (e.g., PER1, PER2, NPAS2, MTNR1A, and MTNR1B). Methods: We analyzed 28 ASD patients (14 with sleep disorders and 14 without) and 23 control subjects of Japanese descent. The coding regions of 18 canonical clock genes and clock-controlled genes were sequenced. Detected mutations were verified by direct sequencing analysis, and additional control individuals were screened. Results: Thirty-six base changes with amino acid changes were detected in 11 genes. Six missense changes were detected only in individuals with ASD with sleep disturbance: p.F498S in TIMELESS, p.S20R in NR1D1, p.R493C in PER3, p.H542R in CLOCK, p.L473S in ARNTL2, and p.A325V in MTNR1B. Six missense changes were detected only in individuals with ASD without sleep disturbance: p.S1241N in PER1, p.A325T in TIMELESS, p.S13T in ARNTL, p.G24E in MTNR1B, p.G24E in PER2, and p.T1177A in PER3. The p.R493C mutation in PER3 was detected in both groups. One missense change, p.P932L in PER2, was detected only in the control group. Mutations in NR1D1, CLOCK, and ARNTL2 were detected only in individuals with ASD with sleep disorder. The prevalence of the mutations detected only single time differed significantly among all ASD patients and controls (p = 0.003). Two kinds of mutations detected only in individuals with ASD with sleep disorder, p.F498S in TIMELESS and p.R366Q in PER3, were considered to affect gene function by three different methods: PolyPhen-2, scale-invariant feature transform (SIFT) prediction, and Mutation Taster (www.mutationtaster.org). The mutations p.S20R in NR1D1, p.H542R in CLOCK, p.L473S in ARNTL2, p.A325T in TIMELESS, p.S13T in ARNTL, and p.G24E in PER2 were diagnosed to negatively affect gene function by more than one of these methods. Conclusion: Mutations in circadian-relevant genes affecting gene function are more frequent in patients with ASD than in controls. Circadian-relevant genes may be involved in the psychopathology of ASD. Ó 2015 The Japanese Society of Child Neurology. Published by Elsevier B.V. All rights reserved.
Keywords: Autism spectrum disorder; Circadian-relevant genes; Next-generation sequencing; Mutation; Single-nucleotide polymorphism; TIMELESS; Period
⇑ Corresponding author at: Department of Pediatrics, Jichi Medical University, 3311-1 Yakushiji, Shimotsuke, Tochigi 329-0498, Japan. Tel.: +81 285 58 7366; fax: +81 285 44 6123. E-mail address: [email protected] (T. Yamagata). 1 Equally contribute.
http://dx.doi.org/10.1016/j.braindev.2015.04.006 0387-7604/Ó 2015 The Japanese Society of Child Neurology. Published by Elsevier B.V. All rights reserved.
Please cite this article in press as: Yang Z et al. Circadian-relevant genes are highly polymorphic in autism spectrum disorder patients. Brain Dev (2015), http://dx.doi.org/10.1016/j.braindev.2015.04.006
2
Z. Yang et al. / Brain & Development xxx (2015) xxx–xxx
1. Introduction The genetic background of autism spectrum disorder (ASD) is highly heterogeneous with respect to copy number variation, single gene mutations, and multifactorial inheritance. Many genes are reportedly associated with ASD, and most of them have roles in synaptic formation or function [1]. Therefore, a primary pathological characteristic of ASD is synaptic dysfunction. Approximately 44–83% of children with ASD experience sleep problems. In contrast, only 10–30% of typically developing infants and children have sleep problems [2]. Autistic children also present with a higher prevalence of sleep disorders than has been observed in children with other developmental disorders, especially intellectually handicapped children [3]. The sleep problems of ASD patients include disturbances of initiation, maintenance, and both. In humans, sleep and wakefulness are regulated by an endogenous circadian clock. The circadian rhythm is established by a system of positive and negative feedback loops of clock genes. The Period genes constitute the main framework of the system. The expression of three Period genes (PER1, 2, 3) and two Cryptochrome genes (CRY1, 2) is activated by a dimer of the proteins CLOCK and ARNTL/BMAL1 [4,5]; PER2 is also able to negatively regulate the ARNTL2/BMAL2-CLOCK complex [6]. CRYs can repress the activity of CLOCK and BMAL1, and thus affect their own expression through negative feedback [7]. In addition to the PER and CRY genes, CLOCK and BMAL1 activate the expression of NR1D1/REV-ERBa, which closes another loop by repressing the expression of ARNTL/BMAL1 [8]. DBP is activated by CLOCK-BMAL1 through Eboxes and inhibited by PER and CRY, and NFIL3/ E4BP4 is a D-box negative regulator and leads to an opposite effect of DBP [9]. Roles in the clock have been proposed for other genes. For example, TIMELESS is essential for resetting the biological clock and interacts directly with the PER proteins [10], and BHLHE40/ Dec1 and BHLHE41/Dec2 are regulators of the mammalian molecular clock [11]. Casein kinase (CK), CSNK1D/CK Id, and CSNK1E/CKIe target PER/ CRY for degradation and regulate their nuclear translocation [12]. Melatonin is synthesized during the night and regulates circadian rhythm through the melatonin receptors MTNR1A and MTNR1B [13], which are expressed in the suprachiasmatic nuclei (SCN), where a master circadian clock is located in mammals. Genetic bases of some familial sleep disorders have been reported [14]. Therefore, genetic analysis of familial sleep and circadian rhythm disorders is important for identifying the specific gene(s) responsible for the regulation of sleep and circadian rhythms in humans.
PER1, NPAS2 [15], MTNR1A, and MTNR1B [16,17] are reportedly associated with ASD. PER2 is one of the candidate genes for 2q37-deletion syndrome, showing ASD and some dysmorphic features [18]. In the Database of Chromosomal Imbalance and Phenotype in Humans (DECIPHER, https://decipher.sanger.ac. uk/) [19], many circadian-related genes, ARNTL, ARNTL2, BHLHE40, CRY2, CSNK1E, DBP, MTNR1A, MTNR1B, NR1D1, PER2, and PER3, were located in the region of copy number variation (CNV) in ASD patients. Sleep is important for normal synaptic development and brain maturation [20]. Bourgeron proposed the idea that the synaptic pathway and the pathway relating to the setting of the clock are altered in ASD. Thus far, there is no direct evidence that the two pathways relate with each other. However, the melatonin system might affect the synaptic pathway through the modulation of CRE-binding protein (CREB) or the GABAergic system [21]. Therefore, it is possible that dysfunction of the circadian genes causes ASD or contributes to the ASD pathophysiology, and further understanding the underlying mechanisms involved with sleep disturbance in children with autism will serve to support a better understanding of the etiology of ASD. In this study, we screened for mutations in the coding regions of circadian-relevant genes in ASD patients with and without sleep disorders, and detected mutations in several genes. 2. Materials and methods 2.1. Samples We screened the circadian-related genes in 14 ASD patients with (Group A), 14 ASD patients without (Group B) sleep disorders, as well as in 23 control individuals. All participants were Japanese. Patients in Group A were 5 men and 9 women (age range 3–28 years), and patients in Group B were 12 men and 2 women (age range 3–19 years). ASD was diagnosed according to the Diagnostic and Statistical Manual of Mental Disorders, 4th Edition; all patients fulfilled the criteria for autism. Regarding sleep disorders, all patients had difficulty initiating sleep, and 6 patients also had difficulty with sleep maintenance. All patients had intellectual disability (ID); their IQs ranged from 14 to 75. Regarding severity of ID, 7 patients were severe, 3 were moderate, and 4 were mild to borderline in Group A; 4 were severe, 4 were moderate, and 6 were mild to borderline in Group B. Eleven patients in Group A and 8 patients in Group B displayed panic and selfinjury. Six patients and five patients, respectively, had epilepsy. Control samples were obtained from healthy individuals after informed consent was obtained.
Please cite this article in press as: Yang Z et al. Circadian-relevant genes are highly polymorphic in autism spectrum disorder patients. Brain Dev (2015), http://dx.doi.org/10.1016/j.braindev.2015.04.006
Z. Yang et al. / Brain & Development xxx (2015) xxx–xxx
2.2. Ethical statement Written informed consent was obtained from the parents of the study participants, and from the participants themselves if they were capable of providing consent. The study was permitted by the Bioethics Committee for Human Gene Analysis of Jichi Medical University. 2.3. Nucleic acid extraction All lymphocyte samples were obtained from patients with informed consent from themselves or their parents. Samples were transfected with EB virus (to establish lymphoblasts) and cultured. Genomic DNA was extracted from lymphoblasts using salting-out methods. For each sample, DNA concentrations were determined using a Nano Drop 2000 (Thermo Fisher Scientific, Waltham, MA, USA). 2.4. The next-generation sequencing analysis Eighteen circadian-relevant genes were selected as candidate genes for analysis (Table 1). Five hundred nanograms of genomic DNA was nebulized for 1 min using 30 psi of pressure, and purified using Agencourt AMPure XP beads (Beckman Coulter, Fullerton, CA, USA). Different multiplex identifiers (MID) (12 different MIDs; Titanium Rapid Library MID adaptor; Roche, South San Francisco, CA, USA) were ligated to the fragmented DNA to enable its use as a library. The adapter-ligated library was amplified through ligation-mediated (LM)-PCR using specific 454 primers. Primers were designed to amplify exons and introns near the genes. Next, 1 lg of the PCR amplification product was hybridized to the custom-designed SeqCap EZ Choice library (Roche NimbleGen, Madison, WI, USA) for 64–72 h at 47 °C in a thermocycler. The captured DNA was bound to Streptavidin M-270 beads (Invitrogen Dynal, Oslo, Norway), was amplified again by LM-PCR using the same specific 454 primers, and was diluted to 1 108 molecules/ll after the fragment sizes were verified with the Aglient High Sensitivity DNA assay. The 12 resulting libraries (each corresponding to one sample) were pooled together in equimolar amounts and combined with capture beads at a ratio of 2 molecules of each DNA library per capture bead. The pooled DNA was then amplified by emulsion PCR (emPCR), following the emPCR amplification manual for the 454 GS Junior Titanium sequencer (Roche). The bead-attached DNAs were denatured, eluted, and quantified with the provided bead counter. We mixed 500,000 enriched DNA beads with Packing Beads. Then, the Pico Titer Plate (PTP) was sequentially loaded with Prelayer Beads, DNA-Packing Beads, Postlayer Beads, and PPiase Beads. Finally, the PTP
3
was mounted in the 454 GS Junior Sequencer, and the program was run in full processing mode for shotgun sequencing (Roche). 2.5. Data analysis The reads were aligned and variants were compared with the reference genome using the 454 integrated software (GS Reference Mapper; Roche). We searched for mutations in the exons of targeted genes. A sample was considered mutated if a mutation was present in at least approximately 50% of confident reads. Genomic reference sequences were ascertained from the NCBI website (http://www.ncbi.nlm.nih.gov/gene/). Each coding polymorphism was compared with the dbSNP database (http://www.ncbi.nlm.nih.gov/snp/) and the corresponding reference (rs) number was assigned to previously identified polymorphisms (Table 2, Supplementary Table 1). The polymorphisms that were detected only in the patients were screened in approximately 100 controls. The hypothesized effects of the mutations on respective protein functions were analyzed using the Polymorphism Phenotyping v2 (PolyPhen-2) prediction tool (http://genetics.bwh. harvard.edu/pph2/), SIFT (http://sift.jcvi.org/), and Mutation Taster (mutation t@sting, http://www. mutationtaster.org/). Statistical significance was analyzed using Fisher’s exact test with IBM SPSS statistics 21 software. 2.6. Direct sequencing analysis Exons in which mutations were detected were verified by direct sequencing analysis. PCR was performed to amplify each exon with a mutation and its neighboring introns. PCR products were purified by passing them through microconcentrating centrifugal filter columns (Millipore, Bedford, MA, USA), and subjected to sequence analysis using the 3730xl DNA Analyzer (Applied Biosystems, Forster City, CA, USA). 2.7. Controls and pedigree analysis The existence of each mutation was confirmed by assessing up to 133 controls, as well as the parents of the patient with the respective mutation. 3. Results We detected 68 (33 in Table 1, 35 in Supplementary Table 2) base changes in the coding regions of 15 genes. Thirty-three base changes with amino acid changes were detected in 11 genes (Table 1), 24 base changes on eight genes were previously reported as single-nucleotide polymorphisms (SNPs), and nine base changes were not reported on the database (Table 1). Detected
Please cite this article in press as: Yang Z et al. Circadian-relevant genes are highly polymorphic in autism spectrum disorder patients. Brain Dev (2015), http://dx.doi.org/10.1016/j.braindev.2015.04.006
4
Gene
Base change
Amino acid change
SNP number
PolyPhen-2 analysis
SIFT analysis
Mutation Taster analysis
Group A (n = 14)
Group B (n = 14)
Control individuals (n = 23)
ARNTL ARNTL2 CSNK1E CLOCK MTNR1A
c.38G > C c.1418T > C c.1484G > A c.2551A > G c.178G > A
p.S13T p.L473S p.S408N p.H542R p.R54 W
– – rs77945315 rs3762836 rs1800885
Tolerated Tolerated Damaging Tolerated Damaging
Disease Disease Disease Disease Disease
causing causing causing causing causing
0 1 2 1 1
1 0 1 0 2
0 0 3 0 1
MTNR1A MTNR1B MTNR1B NR1D1 PER1 PER1 PER2 PER2 PER2 PER2
c.488G > A c.174G > A c.974G > A c.58A > C c.2884G > C c.3722G > A c.2863T > C c.2795C > T c.3257T > C c.3682G > C
p.A157V p.G24E p.A325V p.S20R p.A962P p.S1241 N p.F876L p.P932L p.M1086T p.P1228A
rs1800884 rs8192552 – – rs2585405 – rs78832829 – rs150509214 –
Tolerated Tolerated Tolerated Tolerated Tolerated Tolerated Tolerated Tolerated Tolerated Tolerated
Polymorphism Polymorphism Polymorphism Disease causing Polymorphism Polymorphism Polymorphism Polymorphism Polymorphism Polymorphism
0 0 1 1 4 0 1 0 0 0
1 1 0 0 14 1 1 0 1 1
3 0 0 0 10 0 2 1 1 0
PER2 PER3
c.3968G > A c.1361G > A
p.G1244E p.R366Q
rs934945 rs191016322
Tolerated Damaging
Polymorphism Disease causing
2 1
7 0
8 0
PER3 PER3
c.1477C > T c.2204T > G
p.R493C p.V647G
– rs10462020
Tolerated Tolerated
Polymorphism Polymorphism
1 1
1 0
0 1
PER3 PER3 PER3 PER3 PER3
c.2768T > C c.2854C > G c.3310G > A c.3320C > T c.3326C > T
p.L835P p.P864A p.A1016T p.T1019I p.S1021L
rs228696 rs228697 rs1776342 rs12033719 rs12750400
Tolerated Tolerated Tolerated Tolerated Tolerated
Polymorphism Polymorphism Polymorphism Polymorphism Polymorphism
14 2 8 4 3
14 4 9 5 4
23 4 8 7 4
PER3 PER3 PER3 PER3 TIMELESS
c.3374T > C c.3380T > C c.3737A > G c.3793A > G c.1141G > A
p.M1037T p.L1039S p.H1158R p.T1177A p.A325T
rs2640909 rs77477722 rs10462021 rs35072750 rs113125615
Tolerated Tolerated Tolerated Tolerated Tolerated
Polymorphism Polymorphism Polymorphism Polymorphism Disease causing
2 0 1 0 0
3 1 0 1 1
4 2 1 0 0
TIMELESS TIMELESS
c.1531A > T c.1493T > C
p.I455L p.F498S
rs774027 –
Tolerated Damaging
Polymorphism Disease causing
13 1
14 0
14 0
TIMELESS TIMELESS
c.2660G > A c.3221C > T
p.R831Q p.P1018L
rs774047 rs2291739
Benign Benign Benign Benign Probably damaging Benign Benign Benign Benign Benign Benign Benign Benign Benign Probably damaging Benign Probably damaging Benign Probably damaging Benign Benign Benign Benign Probably damaging Benign Benign Benign Benign Probably damaging Benign Probably damaging Benign Probably damaging
Tolerated Tolerated
Polymorphism Polymorphism
12 9
13 12
18 20
Z. Yang et al. / Brain & Development xxx (2015) xxx–xxx
Please cite this article in press as: Yang Z et al. Circadian-relevant genes are highly polymorphic in autism spectrum disorder patients. Brain Dev (2015), http://dx.doi.org/10.1016/j.braindev.2015.04.006
Table 1 Detected mutations and resulting amino acid changes in ASD patients and controls.
Group
Sample code
Gene
Base change
Amino acid change
Inheritance
Control screening
SNP number
PolyPhen-2 analysis
SIFT analysis
Mutation Taster analysis
Group A
AWS1
TIMELESS
c.1493T > C
p.F498S
Maternal
0/132
None
Damaging
Disease causing
AWS2
NR1D1
c.58A > C
p.S20R
Paternal
0/133
None
Probably damaging Benign
Tolerated
Disease causing
AWS7 AWS10 AWS12
PER3 CLOCK ARNTL2 PER3
c.1477C > T c.2551A > G c.1418T > C c.1361G > A
p.R493C p.H542R p.L473S p.R366Q
N/A Maternal Paternal Paternal
0/92 0/106 0/78 0/98
None rs3762836 None rs191016322
Benign Benign Benign Probably damaging Benign
Tolerated Tolerated Tolerated Damaging
Polymorphism Disease causing Disease causing Disease causing
Tolerated
Polymorphism
Tolerated Tolerated Tolerated
Polymorphism Polymorphism Disease causing
Tolerated Tolerated Tolerated
Disease causing Polymorphism Polymorphism
FHIT (3p21) deletion
Tolerated
Polymorphism
CTNNA3 (10q21) deletion, paternal
Tolerated
Polymorphism
Group B
Control
AWS13
MTNR1B
c.974G > A
p.A325V
Maternal
0/85
None
AWNS1 AWNS3 AWNS4
PER3 PER1 TIMELESS
c.1477C > T c.3722G > A c.1141G > A
p.R493C p.S1241 N p.A325T
Paternal N/A N/A
0/92 0/89 0/103
None None rs113125615
AWNS6 AWNS7 AWNS8
ARNTL MTNR1B PER2
c.38G > C c.174G > A c.3682G > C
p.S13T p.G24E p.P1228A
N/A N/A N/A
0/92 0/118 0/94
None rs8192552 None
AWNS14
PER3
c.3793A > G
p.T1177A
Maternal
0/116
rs35072750
Benign Benign Probably damaging Benign Benign Probably damaging Benign
C15
PER2
c.2795C > T
p.P932L
N/A
1/103
None
Benign
Group A: ASD patients with sleep disorders (AWS); Group B: ASD patients without sleep disorders (AWNS).
Other findings
SHANK2 c2646C > T p.R835Q maternal
Z. Yang et al. / Brain & Development xxx (2015) xxx–xxx 5
Please cite this article in press as: Yang Z et al. Circadian-relevant genes are highly polymorphic in autism spectrum disorder patients. Brain Dev (2015), http://dx.doi.org/10.1016/j.braindev.2015.04.006
Table 2 Mutations from the analysis of circadian-relevant genes in ASD patients with sleep disorders (Group A), ASD patients without sleep disorders (Group B), and controls.
6
Z. Yang et al. / Brain & Development xxx (2015) xxx–xxx
non-synonymous SNPs number were about two times more in ASD than control group; 86 in Group A (n = 14), 113 in Group B (n = 14) and 135 in control (n = 23). Thirteen SNPs were predicted as damaging at least one program, and detected numbers were 21 in Group A, 22 in Group B and 29 in control. The prevalence of nine SNPs that were not reported previously in database differed significantly between all ASD patients and controls (p = 0.006); five in Group A, four in Group B and one in control. Also, 12 SNPs were appeared only single time among samples, and the prevalence differed significantly between all ASD patients and controls (p = 0.003); five in Group A, six in Group B and one in Control. We focused on the mutations that were not reported as SNPs and not listed in the database previously, as well as SNPs with amino acid changes detected only in ASD participants or controls (Table 2). Seven types of mutations in six circadianrelated genes were detected in six patients in ASD patients with sleep disorders (Group A). In ASD patients without sleep disorders (Group B), seven patients exhibited seven types of mutations in six genes. One mutation, p.R493C on PER3, was detected in both Group A and Group B. Only one mutation was detected in one of 23 control individuals, but not in any participants with ASD. These mutations, which were detected in groups of individuals with ASD, were screened in up to 133 controls and were not detected. This finding strengthened the association of these mutations with ASD. PolyPhen-2 analysis categorized two mutations in Group A, two mutations in Group B, and zero mutations in the control group as “probably damaging”. SIFT analysis suggested that only two mutations in Group A were “damaging”. Mutation Taster analysis indicated seven mutations, five in Group A and two in Group B, as “disease causing”. All three methods indicated that two mutations in Group A affected gene function, and at least one method indicated that five mutations in Group A and three mutations in Group B affected gene function (Table 2). In Group A, detected mutations that were not associated with reported SNPs were p.F498S in TIMELESS, p.S20R in NR1D1, p.R493C in PER3, p.L473S in ARNTL2, and p.A325V in MTNR1B (Table 2). The p.H542R mutation in CLOCK and p.R366Q in PER3 were previously reported as SNPs, but were detected only in ASD patients and not in controls. In Group B, there were four detected mutations that were not associated with reported SNPs: p.R493C in PER3, p.S1241N in PER1, p.S13T in ARNTL1, and p.P1228A in PER2. Detected mutations associated with previously reported SNPs were p.A325T in TIMELESS, p.G24E in MTNR1B, and p.T1177A in PER3. Among the SNPs detected in Group A, p.F498S in TIMELESS and p.R366Q in PER3 were considered to affect gene
function by PolyPhen-2, SIFT and Mutation Taster analyses, and p.S20R in NR1D1, p.H542R in CLOCK, and p.L473S in ARNTL2 were considered “disease causing” by Mutation Taster analysis. In Group B, p.A325T in TIMELESS were considered to affect gene function by PolyPhen-2 and Mutation Taster, and p.S13T in ARNTL and p.P1228A in PER2 were considered to affect gene function by Mutation Taster and PolyPhen-2 analyses, respectively (Table 2). The patient AWS1, a 9-year-old boy with p.F498S in TIMELESS, had ASD phenotype with mild disturbance of sociability, delayed speech development and severely restricted interest, and mild intellectual disability (ID). His sleep pattern was inverted so that he had been falling asleep in the morning at approximately 0500–0700 h since infancy. This mutation was also detected in his mother, who had also experienced sleep disturbance since childhood. The mutation p.S20R in NR1D1 was detected in AWS2, a 23-year-old female patient with typical ASD and severe ID who had trouble with sleep initiation. We had detected the SHANK2 mutation of c.2646C > T (p.R835Q) in this patient during previous autism candidate gene screening. The NR1D1 mutation that was considered disease-causing by Mutation Taster analysis was inherited from her father, and the SHANK2 mutation that was considered likely damaging by PolyPhen-2 analysis was inherited from her mother (Table 2). Her parents did not exhibit any obvious phenotype of ASD or sleep disturbance. We detected p.L473S in ARNTL2 and p.R366Q in PER3 in AWS12, a 4-year-old boy with ASD and mild ID, and both base changes were inherited from his father. This patient experienced strong disturbances of both sleep initiation and sleep maintenance. The p.R366Q mutation in PER3 is thought to negatively affect gene function by three methods. The p.R493C mutation in PER3 was detected in one patient in Group A and one patient in Group B. This SNP was not reported previously, but was considered as benign. The p.P1228A mutation in PER2, which was considered damaging, was detected in one patient in Group B. AWNS14 had inherited a PER3 mutation from her mother, as well as a CTNNA3 deletion from her father that was detected by array comparative genomic hybridization (aCGH). AWNS6 had an ARNTL mutation and also FHIT deletion in aCGH, although inheritance was not defined. In addition to these findings in ASD patients, several base changes that were also detected in control individuals were indicated to affect gene function by at least one method. Among these, p.R54W in MTNR1A was detected in three patients with ASD and one control individual, and was indicated to affect gene function by three methods. The p.S408N mutation in CNSKIE, which two methods considered pathogenic, was detected in three ASD patients and three control participants.
Please cite this article in press as: Yang Z et al. Circadian-relevant genes are highly polymorphic in autism spectrum disorder patients. Brain Dev (2015), http://dx.doi.org/10.1016/j.braindev.2015.04.006
Z. Yang et al. / Brain & Development xxx (2015) xxx–xxx
Two mutations in PER3 and one in TIMELESS that were detected in both ASD patients and control participants were considered probably damaging by PolyPhen-2 (Table 1). 4. Discussion This study is the first to document screening for mutations in circadian-relevant genes in ASD patients with or without sleep disorders. Genes detected in this study were candidate genes for circadian abnormality and ASD. We detected several mutations/SNPs in the ASD patients, and rare variant SNPs are much frequently found in ASD group comparing to the control group. It was suggested that ASD group has unique and different SNPs pool comparing to the control population. In addition, the existence of five different mutations in patients with sleep disorder that were indicated to affect gene function suggests that circadian genes are related to some mechanism of ASD and sleep disorder. In our analysis, SNPs of PER1, PER2 and PER3 were detected more frequently in ASD patients than in controls. Among these, functional analysis of gene mutations led us to believe that p.P1228A in PER2 and p.R366Q in PER3 negatively affect gene function. These results implicate PER proteins, especially PER2 and PER3, in the pathogenesis of ASD; they likely control gene expression through the E-box, which is functional in brain tissue. The biological function of TIMELESS is essential for resetting the biological clock, which interacts directly with the PER proteins [22]. Studies have also revealed roles for TIMELESS in bipolar disorder type 1, schizophrenia, and schizoaffective disorder [23]. We found the mutation p.F498S in the TIMELESS gene in an ASD patient who has been falling sleep in the morning since infancy, and also in his mother, who had also experienced sleep disturbance. The mutation p.F498S was indicated to affect gene function by all three methods. We are currently obtaining results that this mutation affects neuronal function (in preparation), and preparing a functional analysis of TIMELESS using knockout mice. NR1D1/Rev-erba knockout mice, which lack the nuclear receptor Rev-erba (a potent transcriptional repressor and core clock component), displayed marked hyperactivity and impaired response habituation in novel environments [24]. The NR1D1 mutation detected in one of our patients was also detected in her father. During our previous candidate gene screening for ASD, we detected a mutation in SHANK2 in this patient and her mother. SHANK2 is a synaptic scaffold protein and a homologue of SHANK3, and de novo mutations and inherited mutations from unaffected parents reportedly contribute to the pathogenesis of ASD [25]. The p.S20R mutation in NR1D1 was considered diseasecausing only by Mutation Taster and not by other
7
methods. However, our findings also suggest that this mutation affects neuronal function, and we are analyzing the effects of the combination of these two gene mutations on the pathogenesis of the patient’s phenotypes. As another example of two alternating genes, previous aCGH analysis detected CNVs in two patients. AWNS6 with ARNTL mutation had partial deletion of FHIT, and AWNS14 with PER3 mutation had partial deletion of CTNNA3. FIHT was reportedly included in recurrent CNVs in ASD [26]. Compound heterozygote deletion of CTNNA3 was reported in ASD patients [27] and a possible association of hemizygote deletion/mutation with ASD was proposed [27,28]. In ASD, the inability to engage in a flow of either verbal or non-verbal interaction has been described as social timing difficulty. Interactive timing, exact synchronization with certain types of adult communication, and rapidly shifting attention from one mode to another were important for normal social and symbolic development. Biological oscillators are essential for information processing, and faster oscillators are employed for communication. It has been hypothesized that anomalies in clock genes that operate as timing genes in high-frequency oscillator systems, in addition to circadian functions, may underlie the timing deficits of autism [29]. ARNTL/BMAL1 and ARNTL2/BMAL2 interact with CLOCK and induce E-box-dependent transactivation by PER2 [6,30,31]. Studies in ARNTL/BMAL1knockout mice and ARNTL/BMAL1-ARNTL2/ BMAL2 double knockout mice have demonstrated that the ARNTL2/BMAL2 complex is more important to the circadian rhythm than BMAL1 [32]. The results of our study that unreported mutations were detected in ARNTL in ASD patients without sleep disorders and in ARNTL2 in ASD patients with sleep disorders might be a reflection of these findings. In addition to TIMELESS and PER3, associations of ARNTL and ARNTL2 with bipolar disorder have been detected in several reports [23]. These base changes were considered disease-causing by Mutation Taster analysis, and the possible association with ASD should be elucidated further. Melatonin is synthesized during the night and is involved in various physiological functions, including sleep induction, circadian rhythm regulation, and immune response. Melatonin signaling is mainly mediated by the guanine nucleotide binding protein-coupled receptors (GPCRs) MTNR1A and MTNR1B. A previous mutation analysis study for melatonin-related genes in 109 ASD patients detected p.V124I in MTNR1B and base changes in the upstream regulatory regions of MTNR1A and MTNR1B, and suggested that these genes were interesting candidates for ASD [16]. In another study regarding the mutation of melatoninrelated genes, several base changes in MTNR1A and MTNR1B were detected at the same frequency in
Please cite this article in press as: Yang Z et al. Circadian-relevant genes are highly polymorphic in autism spectrum disorder patients. Brain Dev (2015), http://dx.doi.org/10.1016/j.braindev.2015.04.006
8
Z. Yang et al. / Brain & Development xxx (2015) xxx–xxx
participants with ASD and control participants [17]. In both studies, the p.V124I mutation in MTNR1B was detected only in ASD patients. The p.A325V mutation, which was detected in a patient in our study and was benign, was also detected in one Asian individual of the human genome pluriethnic diversity panel (HGDP). The p.R54W mutation in MTNR1A, which was considered to affect gene function by three methods, was detected in three patients with ASD and in one control participant. This finding might reflect the existence of a sleep disorder in a control participant, or it might indicate a polygenic factor underlying ASD. The relationship between these rare variants and ASD should be investigated further. Although our sample size was small, our identification of the rare variant SNPs/mutations with functional alternation in ASD patients at a significantly higher frequency than in controls supports the notion that clock genes may be involved in the pathogenesis of ASD. In particular, we consider that PER2, PER3, and TIMELESS may be associated with ASD and partially related to sleep disturbance because we detected several mutations that are likely to induce damage to gene function. Our results are also supported by previous studies that implicate clock genes in autism [16]. Moreover, the presence of severe ID in a patient with the observed mutation may suggest a broader role for circadian-relevant genes in neurodevelopmental disorders. Some base changes were considered benign, and were inherited from unaffected parents. However, the detection of a combination of mutations on two genes in a single patient suggests that these genes may be working as polygenic factors. Acknowledgments This study was supported by the Program for the Strategic Research Foundation at Private Universities 2011-2015 “Cooperative Basic and Clinical Research on Circadian Medicine” from the Ministry of Education, Culture, Sports, Science and Technology of Japan, and by a Grant-in-Aid from Health Labor Sciences Research Grants from the Ministry of Health, Labor, and Welfare, Japan. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10. 1016/j.braindev.2015.04.006. References [1] Zoghbi HY, Bear MF. Synaptic dysfunction in neurodevelopmental disorders associated with autism and intellectual disabilities. Cold Spring Harb Perspect Biol 2012;4.
[2] Glickman G. Circadian rhythms and sleep in children with autism. Neurosci Biobehav Rev 2010;34:755–68. [3] Cotton S, Richdale A. Brief report: parental descriptions of sleep problems in children with autism, down syndrome, and Prader– Willi syndrome. Res Dev Disabil 2006;27:151–61. [4] Etchegaray JP, Lee C, Wade PA, Reppert SM. Rhythmic histone acetylation underlies transcription in the mammalian circadian clock. Nature 2003;421:177–82. [5] Travnickova-Bendova Z, Cermakian N, Reppert SM, SassoneCorsi P. Bimodal regulation of mPeriod promoters by CREBdependent signaling and CLOCK/BMAL1 activity. Proc Natl Acad Sci USA 2002;99:7728–33. [6] Sasaki M, Yoshitane H, Du NH, Okano T, Fukada Y. Preferential inhibition of BMAL2-CLOCK activity by PER2 reemphasizes its negative role and a positive role of BMAL2 in the circadian transcription. J Biol Chem 2009;284:25149–59. [7] Sato TK, Yamada RG, Ukai H, Baggs JE, Miraglia LJ, Kobayashi TJ, et al. Feedback repression is required for mammalian circadian clock function. Nat Genet 2006;38:312–9. [8] Preitner N, Damiola F, Lopez-Molina L, Zakany J, Duboule D, Albrecht U, et al. The orphan nuclear receptor REV-ERBalpha controls circadian transcription within the positive limb of the mammalian circadian oscillator. Cell 2002;110:251–60. [9] Yamajuku D, Shibata Y, Kitazawa M, Katakura T, Urata H, Kojima T, et al. Cellular DBP and E4BP4 proteins are critical for determining the period length of the circadian oscillator. FEBS Lett 2011;85:2217–22. [10] Barnes JW, Tischkau SA, Barnes JA, Mitchell JW, Burgoon PW, Hickok JR, et al. Requirement of mammalian Timeless for circadian rhythmicity. Science 2003;302:439–42. [11] Honma S, Kawamoto T, Takagi Y, Fujimoto K, Sato F, Noshiro M, et al. Dec1 and Dec2 are regulators of the mammalian molecular clock. Nature 2002;419:841–4. [12] Eide EJ, Woolf MF, Kang H, Woolf P, Hurst W, Camacho F, et al. Control of mammalian circadian rhythm by CKIe-regulated proteasome-mediated PER2 degradation. Mol Cell Biol 2005;25:2795–807. [13] Simonneaux V, Ribelayga C. Generation of the melatonin endocrine message in mammals: a review of the complex regulation of melatonin synthesis by norepinephrine, peptides, and other pineal transmitters. Pharmacol Rev 2003;55:325–95. [14] Reid KJ, Chang AM, Dubocovich ML, Turek FW, Takahashi JS, Zee PC, et al. Familial advanced sleep phase syndrome. Arch Neurol 2001;58:1089–94. [15] Nicholas B, Rudrasingham V, Nash S, Kirov G, Owen MJ, Wimpory DC. Association of Per1 and Npas2 with autistic disorder: support for the clock genes/social timing hypothesis. Mol Psychiatry 2007;12:581–92. [16] Jonsson L, Ljunggren E, Bremer A, Pedersen C, Lande´n M, Thuresson K, et al. Mutation screening of melatonin-related genes in patients with autism spectrum disorders. BMC Med Genomics 2010;3:10. [17] Chaste P, Clement N, Mercati O, Guillaume JL, Delorme R, Botros HG, et al. Identification of pathway-biased and deleterious melatonin receptor mutants in autism spectrum disorders and in the general population. PLoS One 2010;5:e11495. [18] Leroy C, Landais E, Briault S, David A, Tassy O, Gruchy N, et al. The 2q37-deletion syndrome: an update of the clinical spectrum including overweight, brachydactyly and behavioural features in 14 new patients. Eur J Hum Genet 2013;21:602–12. [19] Firth HV, Richards SM, Bevan AP, Clayton S, Corpas M, Rajan D, et al. DECIPHER: database of chromosomal imbalance and phenotype in humans using ensembl resources. Am J Hum Genet 2009;84:524–33. [20] Peirano PD, Algarı´n CR. Sleep in brain development. Biol Res 2007;40:471–8.
Please cite this article in press as: Yang Z et al. Circadian-relevant genes are highly polymorphic in autism spectrum disorder patients. Brain Dev (2015), http://dx.doi.org/10.1016/j.braindev.2015.04.006
Z. Yang et al. / Brain & Development xxx (2015) xxx–xxx [21] Bourgeron T. The possible interplay of synaptic and clock genes in autism spectrum disorders. Cold Spring Harb Symp Quant Biol 2007;72:645–54. [22] Sangoram AM, Saez L, Antoch MP, Gekakis N, Staknis D, Whiteley A, et al. Mammalian circadian autoregulatory loop: a timeless ortholog and mPer1 interact and negatively regulate CLOCK-BMAL1-induced transcription. Neuron 1998;21: 1101–13. [23] Mansour HA, Wood J, Logue T, Chowdari KV, Dayal M, Kupfer DJ, et al. Association study of eight circadian genes with bipolar I disorder, schizoaffective disorder and schizophrenia. Genes Brain Behav 2006;5:150–7. [24] Jager J, O’Brien WT, Manlove J, Krizman EN, Fang B, GerhartHines Z, et al. Behavioral changes and dopaminergic dysregulation in mice lacking the nuclear receptor Rev-erba. Mol Endocrinol 2014;28:490–8. [25] Berkel S, Marshall CR, Weiss B, Howe J, Roeth R, Moog U, et al. Mutations in the SHANK2 synaptic scaffolding gene in autism spectrum disorder and mental retardation. Nat Genet 2010;42:489–91. [26] Girirajan S, Dennis MY, Baker C, Malig M, Coe BP, Campbell CD, et al. Refinement and discovery of new hotspots of copy-
[27]
[28]
[29] [30]
[31]
[32]
9
number variation associated with autism spectrum disorder. Am J Hum Genet 2013;92:221–37. Bacchelli E, Ceroni F, Pinto D, Lomartire S, Giannandrea M, D’Adamo P, et al. A CTNNA3 compound heterozygous deletion implicates a role for aT-catenin in susceptibility to autism spectrum disorder. J Neurodev Disord 2014;6:17. Butler MG, Rafi SK, Hossain W, Stephan DA, Manzardo AM. Whole exome sequencing in females with autism implicates novel and candidate genes. Int J Mol Sci 2015;16:1312–35. Wimpory D, Nicholas B, Nash S. Social timing, clock genes and autism: a new hypothesis. J Intellect Disabil Res 2002;46:352–8. Zylka MJ, Shearman LP, Weaver DR, Reppert SM. Three period homologs in mammals: differential light responses in the suprachiasmatic circadian clock and oscillating transcripts outside of brain. Neuron 1998;20:1103–10. Shearman LP, Sriram S, Weaver DR, Maywood ES, Chaves I, Zheng B, et al. Interacting molecular loops in the mammalian circadian clock. Science 2000;288:1013–9. Shi S, Hida A, McGuinness OP, Wasserman DH, Yamazaki S, Johnson CH. Circadian clock gene Bmal1 is not essential; functional replacement with its paralog, Bmal2. Curr Biol 2010;20:316–21.
Please cite this article in press as: Yang Z et al. Circadian-relevant genes are highly polymorphic in autism spectrum disorder patients. Brain Dev (2015), http://dx.doi.org/10.1016/j.braindev.2015.04.006
