Supplementary Table 1. Known lissencephaly genes in order of discovery, and mutation frequency in 216 patients with unexplained lissencephaly
Year / Gene / Mutations / OriginalReferencesAll / ACMG
1993 / LIS1 (PAFAH1B1) / 23 (11%) / 21 (10%) / 1,2
1998 / DCX / 3 (1%) / 3 (1%) / 3
2002 / ARX / 0 / 0 / 4
2000 / RELN / 1 (<1%) / 1 (<1%) / 5
2007 / TUBA1A / 5 (2%) / 4 (2%) / 6
2009 / TUBB2B / 5 (2%) / 3 (1%) / 7
2009 / TUBA8 / 0 / 0 / 8
2009 / VLDLR / 0 / 0 / 9
2010 / TUBB3 / 1 (<1%) / 1 (<1%) / 10
2011 / NDE1 / NT / NT / 11-13
2011 / RNU4ATAC / NT / NT / 14
2012 / TUBB (TUBB5) / 1 (<1%) / 1 (<1%) / 15
2012 / ACTB / 0 / 0 / 16
2012 / ACTG1 / 4 (2%) / 3 (1%) / 16
2012 / DYNC1H1 / 20 (10%) / 14 (7%) / 17
2013 / KIF2A / 0 / 0 / 18
2013 / KIF5C / 0 / 18
2013 / TUBG1 / 5 (2%) / 5 (2%) / 18
2015 / CDK5 / NT / NT / 19
2016 / CRADD / 6 (3%) / 6 (3%) / 20
TOTAL / 74 (34%) / 62 (29%)
None detected / 142 (66%) / 154 (71%)
Supplementary Table 2. Molecular data on patients with disease causing mutations identified by smMIP panel sequencing or whole exome sequencing.
Supplementary Table 3. Clinical data on patients with disease causing mutations identified by smMIP panel sequencing and whole exome sequencing.
Supplementary Table 4. smMIP panel sequencing - regions with low coverage.
SUPPLEMENTARY METHODS
Targeted sequencing
Genomic DNA was extracted from patients’ blood or saliva using different extraction methods. Samples from subjects enrolled during the last 5 years were processed with the Puregene Blood Core Kit (Qiagen, Venlo, Netherlands) with RNase for blood, and the Oragene Saliva Kit (DNA genotek, Ottawa, ON, Canada).
Using an established MIPGEN pipeline,21 we designed a pool of 673 single molecule molecular inversion probes (smMIP) targeting the coding regions of 17 known LIS-associated genes: ACTB, ACTG1, ARX, CRADD, DCX, DYNC1H1, KIF2A, KIF5C, LIS1, RELN, TUBA1A, TUBA8, TUBB, TUBB2B, TUBB3, TUBG1 and VLDLR. For four regions in tubulin genes (TUBA1A, TUBG1, TUBB2B), unique smMIPs could not be designed using default parameters due to high homology. We therefore designed smMIPs without that requirement, expecting to amplify multiple regions (non-unique primers). In total, smMIPs were tiled across 46,789 bp of coding sequence. 100 ng capture reactions were performed in parallel using previously published protocols.21,22 Massively parallel sequencing was done with the Illumina HiSeq and MiSeq platforms.
The smMIP sequencing data was processed with MIPGEN followed by PEAR 0.8.1 ( using default options. This produced high quality single molecule consensus (smc) reads. Variants were called with FreeBayes using the -F 0.1 option to capture low frequency variants, as it retains any variant having two or more non-reference capture events. Variants in non-unique regions were identified using the GATK Haplotype Caller, and FreeBayes on the non-collapsed reads. All variants were annotated using ANNOVAR. Variants were filtered for function against the NHLBI exome variant server, the ExAC server, and 1000 genomes datasets. All variants with greater than 1% frequency in these public databases were excluded. For genes with dominant inheritance (ACTB, ACTG1, DYNC1H1, KIF2A, KIF5C, LIS1, TUBA1A, TUBB, TUBB2B, TUBB3, and TUBG1), we analyzed only variants not present or present in only one ExAC individual.
Copy number variants within smMIP data were detected using ONCOCNV,23 using samples with previously identified causative SNVs used as controls. In the first step we used default parameters of ONCOCNVpipelines that allowed identification of deletions of entire genes (filtered for the loss – copy number value 1 or 1.5 –with the p-value <0.009). While it provides robust whole gene copy number detection from amplicon data, ONCOCNV was not designed to identify one-point copy number change (gain or loss of one exon/amplicon region). However, we used an available functionality to output P-values and outlier statuses for each targeted region and filtered for regions where two consecutive smMIPs had a change. The majority of the calls represented artifacts, since many patients had CNVs of the same amplicon detected. By filtering the obvious false positive calls we identified three unique losses, but only NN were subsequently confirmed by qPCR.
Whole exome sequencing
We also performed whole exome sequencing (WES) in one trio (one affected child and both parents) and one foursome (two affected siblings and their unaffected parents). Five other patients were studied with both WES and our targeted smMIP panel. Library preparation, exome enrichment, and WES were done at the University of Washington Center for Mendelian Genomics, or Broad Institute Genomic Services with enrichment kits used on a routine basis at the corresponding center (NimbleGen v2 or Agilent SureSelect). Captured libraries were sequenced on a HiSeq 2000 (Illumina, San Diego, CA, USA).
Sequence reads were aligned to the human genome (hg19) using BWA software or the CLC Biomedical Genomics workbench. Downstream processing was done with the Genome Analysis Toolkit, SAMtools, and Picard Tools. Single nucleotide variants and indels were subsequently called by GATK Unified Genotyper24and a variant quality score of ≥10 and were annotated using SeattleSeq SNP annotation and ANNOVAR (see URLs). Variants were then filtered using standard hard-filtering parameters 24. Specifically, only variants with a quality score of ≥30, sequencing depth of ≥10, quality/depth ratio of ≥5, length of homopolymer run of ≤5.0 and allelic balance of ≤0.80 were considered for downstream analysis.
We analyzed variants affecting coding regions and essential splice sites and excluded all variants with frequencies higher than 1% in multiple genome databases including the Single Nucleotide Polymorphism Database (dbSNP), 1000 Genomes, and the Exome Aggregation Consortium (ExAC).
SUPPLEMENTARY RESULTS
Regions with low coverage
The average collapsed coverage (number of single molecule consensus sequences or “smc-reads” across all samples was 21,22 but coverage was not uniform. The genomic coordinates of regions with average collapsed coverage <5 smc-reads are shown in Supplementary Table 4. ARX was excluded from analysis because of poor enrichment and uniformly low coverage. ACTG1, ACTB, TUBB3, TUBB2B and TUBA8 had low coverage with 20% of the coding sequence covered with 5 or fewer smc-reads. Low coverage correlates with high (>65%) or low (<30%) GC content of the target sequence, and with low in silico prediction scores (SVR score < 1.4).21 These regions represent diagnostic gaps that may have decreased the mutation yield in our cohort. One mutation in TUBB2B was found during clinical testing that was missed during our initial variant calling due to low coverage. However, direct Sanger sequencing of poorly covered regions in ACTB and ACTG1 in patients with frontal predominant pachygyria did not reveal mutations missed by smMIPs.
SUPPLEMENTARY DISCUSSION
Reelinopathies: VLDLR
Biallelic mutations of VLDLR have been reported in the “disequilibrium syndrome” in 40 subjects from 14 families.9,25-31When provided, brain imaging shows the same rare subtype of LIS seen with RELN mutations except that they appear on average less severe. Mutations of VLDLR were a very rare cause of LIS in our cohort, seen only in two previously reported children.30 Both had anterior predominant thin LIS with severe cerebellar hypoplasia and loss of the foliar pattern, the same pattern seen with RELN mutations. Too few mutations of this gene have been reported to draw further conclusions.
Forebrain transcription factors
Severe mutations of ARX including truncations and missense mutations in the homeodomain cause X-linked lissencephaly with abnormal genitalia (XLAG) in males, sometimes with severe hydrocephalus as well.4,32-34 About half of females carrying the same mutations have a less severe phenotype consisting of agenesis of the corpus callosum and epilepsy.35 Less severe mutations including a few homeodomain mutations cause a wide range of phenotypes including agenesis of the corpus callosum, infantile spasms and other types of early life epilepsy, dyskinesia and intellectual disability in males, while females are normal.36-41 The relatively high frequency of XLAG in our cohort most likely reflects very active recruitment for several years. The true frequency is likely to be closer to 1% or less.
Other LIS causes
Our data shows that a very high proportion of LIS is genetic. Indeed, we are aware of no reliable data suggesting a non-genetic etiology for any form of LIS. Several studies from the 1970’s suggested intrauterine hypoxia or perfusion failure based on pathological analysis, and another reported LIS with prenatal cytomegalovirus (CMV) infection.42-44But the abnormal cortical histology proposed for perfusion failure is the same as seen in classic (genetic) LIS, while the histology linked to CMV infection would be better classified as polymicrogyria, which has been reported with CMV many times.45-47 We have seen no compelling evidence to support any extrinsic, non-genetic cause of LIS based on experience in ~1400 children over 30 years. However, polymicrogyria, hemimegalencephaly, focal cortical dysplasia type 2, and some other malformations of cortical development can be mistaken for LIS on low-resolution brain-imaging studies or by less experienced interpreters.
Previous studies
Our work represents the first systematic study of a nearly complete set of known causal genes in a large well-studied LIS cohort, finding mutations in 68 of 216 (31%) unexplained and partially pretested LIS patients (Supplementary table 4) by targeted smMIP sequencing. Two prior studies have tested a smaller number of patients and genes. The first reported mutations of KIF5C, KIF2A, DYNC1H1 or TUBG1 in 8 of 162 (5%) patients with LIS or PMG (including tubulinopathies) who previously tested negative for mutations in LIS1, DCX and three tubulin genes.18 The second study reported mutations in 8 of 47 (17%) patients with LIS (primarily pachygyria) on a panel of 12 LIS genes (similar to our panel but excluding four low frequency genes), and in 9 of 30 (30%) patients with SBH on a panel of 6 LIS genes including DCX and LIS1.48 No information was provided regarding prior testing in this cohort. The data provided was insufficient for us to determine why the mutation detection rate in these two studies was so much lower than our rate
Mosaicism
One study cited above emphasized the occurrence of mosaic mutations in LIS phenotypes, finding mosaicism in 6 of 9 (67%) individuals with SBH but only 1 of 8 (13%) with LIS.48 The discovery of mosaicism in patients with SBH is not surprising, as prior reports have shown that most mutations of DCX in males with SBH and most mutations of LIS1 in SBH are mosaic.49-54 In our historical cohort, 10 of 23 (43%) patients with posterior-predominant SBH had mutations of LIS1, which were shown to be mosaic using old technology in seven, germline in two and not determined in one. In our smMIP sequencing cohort, 2 of 15 (13%) patients with posterior SBH had LIS1 mutations, both mosaic. We conclude that mosaicism is common in SBH (except for females with diffuse or frontal SBH where germline DCX mutations are most common), but not in other forms of LIS. However, both symptomatic mosaicism and germline mosaicism (in parents) do occur.
REFERENCES FOR SUPPLEMENT
1.Reiner O, Carrozzo R, Shen Y, et al. Isolation of a Miller-Dieker lissencephaly gene containing G protein beta-subunit-like repeats. Nature. 1993;364:717-721.
2.Chong SS, Pack SD, Roschke AV, et al. A revision of the lissencephaly and Miller-Dieker syndrome critical regions in chromosome 17p13.3. Hum Mol Genet. 1997;6:147-155.
3.Gleeson JG, Allen KM, Fox JW, et al. Doublecortin, a brain-specific gene mutated in human X-linked lissencephaly and double cortex syndrome, encodes a putative signaling protein. Cell. 1998;92:63-72.
4.Kitamura K, Yanazawa M, Sugiyama N, et al. Mutation of ARX causes abnormal development of forebrain and testes in mice and X-linked lissencephaly with abnormal genitalia in humans. Nat Genet. 2002;32:359-369.
5.Hong SE, Shugart YY, Huang DT, et al. Autosomal recessive lissencephaly with cerebellar hypoplasia is associated with human RELN mutations. Nat Genet. 2000;26:93-96.
6.Keays DA, Tian G, Poirier K, et al. Mutations in alpha-tubulin cause abnormal neuronal migration in mice and lissencephaly in humans. Cell. 2007;128:45-57.
7.Jaglin XH, Poirier K, Saillour Y, et al. Mutations in the beta-tubulin gene TUBB2B result in asymmetrical polymicrogyria. Nat Genet. 2009;41:746-752.
8.Abdollahi MR, Morrison E, Sirey T, et al. Mutation of the variant alpha-tubulin TUBA8 results in polymicrogyria with optic nerve hypoplasia. Am J Hum Genet. 2009;85:737-744.
9.Boycott KM, Flavelle S, Bureau A, et al. Homozygous deletion of the very low density lipoprotein receptor gene causes autosomal recessive cerebellar hypoplasia with cerebral gyral simplification. Am J Hum Genet. 2005;77:477-483.
10.Poirier K, Saillour Y, Bahi-Buisson N, et al. Mutations in the neuronal ss-tubulin subunit TUBB3 result in malformation of cortical development and neuronal migration defects. Hum Mol Genet. 2010;19:4462-4473.
11.Alkuraya FS, Cai X, Emery C, et al. Human mutations in NDE1 cause extreme microcephaly with lissencephaly [corrected]. Am J Hum Genet. 2011;88:536-547.
12.Bakircioglu M, Carvalho OP, Khurshid M, et al. The essential role of centrosomal NDE1 in human cerebral cortex neurogenesis. Am J Hum Genet. 2011;88:523-535.
13.Paciorkowski AR, Keppler-Noreuil K, Robinson L, et al. Deletion 16p13.11 uncovers NDE1 mutations on the non-deleted homolog and extends the spectrum of severe microcephaly to include fetal brain disruption. Am J Med Genet A. 2013;161A:1523-1530.
14.Abdel-Salam GM, Abdel-Hamid MS, Hassan NA, et al. Further delineation of the clinical spectrum in RNU4ATAC related microcephalic osteodysplastic primordial dwarfism type I. Am J Med Genet A. 2013;161A:1875-1881.
15.Breuss M, Heng JI, Poirier K, et al. Mutations in the beta-tubulin gene TUBB5 cause microcephaly with structural brain abnormalities. Cell reports. 2012;2:1554-1562.
16.Riviere JB, van Bon BW, Hoischen A, et al. De novo mutations in the actin genes ACTB and ACTG1 cause Baraitser-Winter syndrome. Nat Genet. 2012;44:440-444, S441-442.
17.Willemsen MH, Vissers LE, Willemsen MA, et al. Mutations in DYNC1H1 cause severe intellectual disability with neuronal migration defects. J Med Genet. 2012;49:179-183.
18.Poirier K, Lebrun N, Broix L, et al. Mutations in TUBG1, DYNC1H1, KIF5C and KIF2A cause malformations of cortical development and microcephaly. Nat Genet. 2013;45:639-647.
19.Magen D, Ofir A, Berger L, et al. Autosomal recessive lissencephaly with cerebellar hypoplasia is associated with a loss-of-function mutation in CDK5. Hum Genet. 2015;134:305-314.
20.Di Donato N, Jean YY, Maga AM, et al. Mutations in CRADD Result in Reduced Caspase-2-Mediated Neuronal Apoptosis and Cause Megalencephaly with a Rare Lissencephaly Variant. Am J Hum Genet. 2016;99:1117-1129.
21.Boyle EA, O'Roak BJ, Martin BK, Kumar A, Shendure J. MIPgen: optimized modeling and design of molecular inversion probes for targeted resequencing. Bioinformatics. 2014;30:2670-2672.
22.Hiatt JB, Pritchard CC, Salipante SJ, O'Roak BJ, Shendure J. Single molecule molecular inversion probes for targeted, high-accuracy detection of low-frequency variation. Genome research. 2013;23:843-854.
23.Boeva V, Popova T, Lienard M, et al. Multi-factor data normalization enables the detection of copy number aberrations in amplicon sequencing data. Bioinformatics. 2014;30:3443-3450.
24.DePristo MA, Banks E, Poplin R, et al. A framework for variation discovery and genotyping using next-generation DNA sequencing data. Nature genetics. 2011;43:491-498.
25.Ali BR, Silhavy JL, Gleeson MJ, Gleeson JG, Al-Gazali L. A missense founder mutation in VLDLR is associated with Dysequilibrium Syndrome without quadrupedal locomotion. BMC Med Genet. 2012;13:80.
26.Kruer MC, Jepperson TN, Weimer JM, et al. Mutations in VLDLR associated with ataxia with secondary vitamin E deficiency. Mov Disord. 2013;28:1904-1905.
27.Moheb LA, Tzschach A, Garshasbi M, et al. Identification of a nonsense mutation in the very low-density lipoprotein receptor gene (VLDLR) in an Iranian family with dysequilibrium syndrome. Eur J Hum Genet. 2008;16:270-273.
28.Ozcelik T, Akarsu N, Uz E, et al. Mutations in the very low-density lipoprotein receptor VLDLR cause cerebellar hypoplasia and quadrupedal locomotion in humans. Proc Natl Acad Sci U S A. 2008;105:4232-4236.
29.Sonmez FM, Gleeson JG, Celep F, Kul S. The very low density lipoprotein receptor-associated pontocerebellar hypoplasia and dysmorphic features in three Turkish patients. J Child Neurol. 2013;28:379-383.
30.Boycott KM, Bonnemann C, Herz J, et al. Mutations in VLDLR as a cause for autosomal recessive cerebellar ataxia with mental retardation (dysequilibrium syndrome). J Child Neurol. 2009;24:1310-1315.
31.Kolb LE, Arlier Z, Yalcinkaya C, et al. Novel VLDLR microdeletion identified in two Turkish siblings with pachygyria and pontocerebellar atrophy. Neurogenetics. 2010;11:319-325.
32.Bettella E, Di Rosa G, Polli R, et al. Early-onset epileptic encephalopathy in a girl carrying a truncating mutation of the ARX gene: rethinking the ARX phenotype in females. Clin Genet. 2013;84:82-85.
33.Kato M, Das S, Petras K, et al. Mutations of ARX are associated with striking pleiotropy and consistent genotype-phenotype correlation. Hum Mutat. 2004;23:147-159.
34.Uyanik G, Aigner L, Martin P, et al. ARX mutations in X-linked lissencephaly with abnormal genitalia. Neurology. 2003;61:232-235.
35.Marsh E, Fulp C, Gomez E, et al. Targeted loss of Arx results in a developmental epilepsy mouse model and recapitulates the human phenotype in heterozygous females. Brain. 2009;132:1563-1576.
36.Bienvenu T, Poirier K, Friocourt G, et al. ARX, a novel Prd-class-homeobox gene highly expressed in the telencephalon, is mutated in X-linked mental retardation. Hum Mol Genet. 2002;11:981-991.
37.Conti V, Marini C, Gana S, Sudi J, Dobyns WB, Guerrini R. Corpus callosum agenesis, severe mental retardation, epilepsy, and dyskinetic quadriparesis due to a novel mutation in the homeodomain of ARX. Am J Med Genet A. 2011;155A:892-897.
38.Guerrini R, Moro F, Kato M, et al. Expansion of the first PolyA tract of ARX causes infantile spasms and status dystonicus. Neurology. 2007;69:427-433.
39.Kato M, Saitoh S, Kamei A, et al. A longer polyalanine expansion mutation in the ARX gene causes early infantile epileptic encephalopathy with suppression-burst pattern (Ohtahara syndrome). Am J Hum Genet. 2007;81:361-366.
40.Mirzaa GM, Paciorkowski AR, Marsh ED, et al. CDKL5 and ARX mutations in males with early-onset epilepsy. Pediatr Neurol. 2013;48:367-377.
41.Stromme P, Mangelsdorf ME, Shaw MA, et al. Mutations in the human ortholog of Aristaless cause X-linked mental retardation and epilepsy. Nat Genet. 2002;30:441-445.
42.Stewart RM, Richman DP, Caviness VS, Jr. Lissencephaly and Pachygyria: an architectonic and topographical analysis. Acta Neuropathol. 1975;31:1-12.
43.Jellinger K, Rett A. Agyria-pachygyria (lissencephaly syndrome). Neuropadiatrie. 1976;7:66-91.
44.Norman MG, Roberts M, Sirois J, Tremblay LJ. Lissencephaly. Can J Neurol Sci. 1976;3:39-46.
45.Becroft DM. Prenatal cytomegalovirus infection: epidemiology, pathology and pathogenesis. Perspect Pediatr Pathol. 1981;6:203-241.
46.de Vries LS, Gunardi H, Barth PG, Bok LA, Verboon-Maciolek MA, Groenendaal F. The spectrum of cranial ultrasound and magnetic resonance imaging abnormalities in congenital cytomegalovirus infection. Neuropediatrics. 2004;35:113-119.
47.Manara R, Balao L, Baracchini C, Drigo P, D'Elia R, Ruga EM. Brain magnetic resonance findings in symptomatic congenital cytomegalovirus infection. Pediatr Radiol. 2011;41:962-970.
48.Jamuar SS, Lam AT, Kircher M, et al. Somatic Mutations in Cerebral Cortical Malformations. N Engl J Med. 2014;371:733-743.
49.Quelin C, Saillour Y, Souville I, et al. Mosaic DCX deletion causes subcortical band heterotopia in males. Neurogenetics. 2012;13:367-373.
50.Poolos NP, Das S, Clark GD, et al. Males with epilepsy, complete subcortical band heterotopia, and somatic mosaicism for DCX. Neurology. 2002;58:1559-1562.
51.Gleeson JG, Luo RF, Grant PE, et al. Genetic and neuroradiological heterogeneity of double cortex syndrome. Ann Neurol. 2000;47:265-269.