Data were obtained from a population-based case–control study conducted by the California Birth Defects Monitoring Program (CBDMP). The CBDMP is an active, population-based surveillance system for collecting information on infants and fetuses with congenital malformations, which has been described elsewhere [19]. Included for study were 192 singleton infants with spina bifida (cases) and 190 non-malformed infants (controls). Cases were randomly selected from all live born cases and a random sample of non-malformed control infants ascertained by the CBDMP corresponding to birth years 1994–1998. The case and control infants were linked to their newborn bloodspot. All samples were obtained with approval from the State of California Health and Welfare Agency Committee for the Protection of Human Subjects.

DNA for genotyping for this study derived from anonymous newborn bloodspots. Bloodspots are collected on all newborns in California for genetic testing purposes by the State of California. The State retains the residual, unused, portion of the bloodspot and makes these bloodspots available to approved researchers. The approval process includes detailed review by the State of California Committee for the Protection of Human Subjects (the primary IRB).

The original collection of bloodspots for newborn testing includes an information form but it is not an official informed consent form. The purpose of the form is to disseminate information to the parents as to what occurs with their babies’ bloodspots and provides them with the instructions so that they can opt out and request destruction by writing to the State of California. Therefore this process is similar to the newborn genetic screening tests - "informed dissent". For the use of anonymous bloodspots for research, an "opt out" policy is applied. In other words, parents are given written materials which explain that if they do not want their child’s specimen used in research studies, they can write to the State and the State will destroy the sample. Thus, no bloodspots were used in this research project for anyone whose parents had "opted out".

Genomic DNA was extracted from newborn screening bloodspots using the Puregene DNA Extraction Kit (Qiagen, Valencia, CA) and amplified using the GenomiPhi Kit (GE Healthcare). Coding exons and flanking exon-intron regions of the human SCRIB gene (NM_015356) were amplified by polymerase chain reactions (PCR) from the whole genome amplification (WGA) product. Primer sequences are available upon request. The PCR products were sequenced using the Prism Bigdye Terminator Kit (v3) on an ABI 3730XL DNA analyzer (Life Technologies, Carlsbad, CA). Both case and control samples were sequenced with either a specific forward or reverse primer. Sequencing results were analyzed using the Mutation Surveyor software V 4.0.7 (Softgenetics, Stage College, PA). The detected mutations were subsequently confirmed by a second round of whole genome amplification, PCR and sequencing analysis.

Mutations were annotated according to the HGVS nomenclature (http://www.hgvs.org/mutnomen/). Nucleotide numbering reflects cDNA numbering with +1 corresponding to the A of the ATG translation initiation codon 1 in the reference sequence. A variant was designated as novel if it was not found in dbSNP Build 136 or in 1000 genome data (www.1000genome.org) or the Exome Variant Server (NHLBI GO Exome Sequencing Project (ESP), Seattle, WA (URL: http://evs.gs.washington.edu/EVS/) [November 2012 accessed]). Rare mutations were defined as having an allele frequency of less than 1%. The potential pathogenic effect of the missense mutations on protein function was predicted using two online programs: PolyPhen (Polymorphism Phenotyping) (http://genetics.bwh.harvard.edu/pph/) and SIFT (http://sift.jcvi.org/). CLUSTAL W program with built in Mega software (V 5.1) (http://www.megasoftware.net/) was used to analyze the conservation of SCRIB amino acids that were changed by the mutations. Residues were considered to be highly conserved if there was no variation in amino acid properties observed across the compared seven orthologous proteins, included five mammalian orthologus plus zebrafish and Drosophila.

Human SCRIB full-length cDNA was cloned into pEGFP-C1 plasmid (GFP-SCRIB) and was graciously provided for our use by Dr. Patrick Humbert (Peter MacCallum Cancer Centre, Melbourne, Australia). GFP-SCRIB I285K mutant construct was acquired from Dr. Philip Stanier (UCL Institute of Child Health, London, UK). Human influenza hemagglutinin (HA) tagged VANLG2 (HA-VANGL2) plasmid was obtained from Dr. Hongyan Wang (Fudan University, Shanghai, China). SCRIB missense changes were introduced into GFP-SCRIB by GeneArt® Site-Directed Mutagenesis System (Life Technologies, Carlsbad, CA).

MDCK II cell line was chosen for use in SCRIB mutation subcellular localization studies due to its epithelial characteristic and being commonly used in epithelial cell surface polarity studies. MDCK II cells were purchased from Sigma-Aldrich and cultured according to the manufacturer’s protocols. One day before transfection, cells were seeded in 4 chamber 35 mm glass bottom dishes (4x10⁴/chamber) (In Vitro Scientific, Sunnyvale, CA). Transfection was performed using Lipofectamine 2000 regent (Life Technologies, Carlsbad, CA) according to the manufacturer’s manual. Forty-eight hours later, cells were washed twice with PBS and incubated 5 minutes with Hochest 3342 (1µg/ml) (Invitrogen) and CellMask™ Plasma Membrane Stains (Life Technologies, Carlsbad, CA) (2µg/ml) solution. The cells were then washed 3 times with PBS and fixed in 4% PFA (paraformaldehyde in phosphate-buffered saline) for 10 minutes at 37^°C, followed by 3 more PBS washes. Cells were examined and photographed by a laser scanning confocal microscope (LSM710, Leica). The transfection efficiency of GFP-SCRIB wild type and its mutant constructs was measured and calculated on the Operetta High Content Screening System (PerkinElmer).

The subcellular localization of GFP expression was classified to three different categories: (1) GFP present only at membrane; (2) GFP distributed at the membrane and in the cytoplasm; and (3) GFP exclusively in the cytoplasm. For each transfection, at least 50 cells were counted and each construct transfection was performed three times. Quantitative data between wild type and mutants were evaluated by Chi-square analysis.

The HEK293T cell line was chosen in immunoprecipitation assay because of its high transfectability. In this study, the HEK293T cell line was kindly provided by Dr. Edward Marcotte (Department of Chemistry & Biochemistry, University of Texas at Austin) and was cultured in DMEM (Life Technologies, Carlsbad, CA) supplemented with 10% FBS and 2mM glutamine. One day before transfection, cells were seeded into 6-well plates (6X10⁵cells/well). GFP-SCRIB (2µg) or its relative mutant plasmids were co-transfected with HA-VANGL2 (2µg). Twenty-four hour post-transfection, cells were washed twice with ice-cold PBS and lysed with 200µl 1% NP-40 lysate buffer containing 1X cocktail protease inhibitor (Roche). We removed 30µl lysate as inputs, examined the input by anti-GFP or anti-HA immunoblotting. The remaining cell lysate was used for GFP-SCRIB and HA-VANGL2 co-immunoprecipitation. The co-immunoprecipation was performed using a Pierce HA Tag IP/Co–IP Kit (Thermo Fisher Scientific Inc.), according to the instructions. The co-precipitates were run on SDS-PAGE followed by western blot detection, immunoblotting with anti-GFP antibody for GFP-SCRIB detection and anti-HA antibody for HA-VANGL2 detection.

We detected six SCRIB missense mutations among 192 spina bifida infants. They were: c.1096 G > A (p.A366T); c.1655 C>T (p.T552M); c.3128 C>T (p.P1043L); c.3943G>A (p.A1315T); c.3995C>T (p.P1332L) and c.4559T>G (P.L1520R) (Table 1). None of these mutations was found in the dbSNP database or 1000 genome data, or in the 190 infant controls. p.P1332L was found in the ESP database, with a minor allele frequency of 0.007 among European American (EA) populations. Among of these six mutations, p.A366T was localized to the LRR domain, p.1043L was within the third PDZ domain, while the other four variants did not locate in these two classical domains of SCRIB (Figure 1A, Table 1). Five of the six mutations were predicted-to-be-deleterious by PolyPhen, while three of them were predicted-to-be-deleterious by SIFT. PolyPhen and SIFT each has its own advantages and disadvantages; therefore, combining the results from both programs was thought to increase the prediction sensitivity [20]. Using this strategy, five mutations were determined as “predicted-to-be-deleterious” (Table 1). During amino acid conservation analysis, three (p.P1043L, p.P1332L and p.L1520R) mutations located at highly conserved nucleotides, and the others (p.A366T, p.T552M and p.A1315T) involved less conserved nucleotides (Figure 1B). p.P1043L was identified in a case with an open defect in the thoracic area, others were found among cases with defects in the lumbosacral area. We also detected six novel synonymous mutations in case infants, but not control infants (Table 2). Eight common SNPs were identified in both cases and controls, and there were no significant differences between the groups (Table 3).

Previous studies demonstrated that SCRIB must localize to the cell membrane in order to function normally [13,17]. The SCRIB mutation (p.I285K) which caused craniorachischisis in mice adversely affected SCRIB membrane localization [17]. We compared the SCRIB membrane localization of the six case specific missense mutations with the wild type localization serving as a positive control, and the mouse mutation p.I285K serving as a negative control. Three of the six mutations(p.P1043L, p.P1332L and p.L1520R), all of which change a highly conservative amino acid (Figure 1B), significantly increased the proportion of cells with cytosolic localization, like the mouse pathogenic line-90 mutant (p.I285K), while the other three mutations (p.A366T, p.T552M and p.A1315T) maintained the same subcellular localization pattern as wild type (Figure 2). No significant difference in transfect efficiency was detected between GFP-SCRIB wild type and its mutant constructs (Supplementary Figures S1, S2, and S3 in File S1).

VANGL2 is a core PCP gene that when mutated, causes craniorachischis in mice [5,6] as well as different kinds of NTDs in humans [21,22]. Of the six mutations identified, one of them (p.P1043L) was located in the third PDZ domain (PDZ3) of SCRIB, which could affect the physical interaction with PZD binding motif in VANGL2. We also performed co-immunoprecipitation of SCRIB and its mutants with VANGL2. We found that none of them lost the physical interaction between SCRIB and VANGL2 (Figure 3).