西亚试剂优势供应上万种化学试剂产品，欢迎各位新老客户咨询、选购！
中国
联系我们
联系方式：400-990-3999 / 邮箱：sales@xiyashiji.com
西亚试剂 —— 品质可靠，值得信赖
VANGL2 Mutations in Human Cranial Neural-Tube Defects
Yun-Ping Lei, B.S. Ting Zhang, Ph.D. Hong Li, M.D. Bai-Lin Wu, M.Med., Ph.D. Li Jin, Ph.D. Hong-Yan Wang, Ph.D.
Mutations in more than 200 genes are known to cause neural-tube defects in mice; less is known about the genetic cause of neural-tube defects in humans.1 Kibar and colleagues2 hypothesized that human neural-tube defects are caused by mutations in VANGL1 and VANGL2, genes that affect planar cell polarity and cause neural-tube defects in mice. They identified mutations in VANGL1 but not in VANGL2 in humans.2 We hypothesized that mutations in VANGL2 are lethal to the fetus, and therefore we sequenced VANGL2 in 163 stillborn or miscarried Han Chinese fetuses with neural-tube defects (Table 1 in the Supplementary Appendix, available with the full text of this letter at NEJM.org) and 508 apparently unrelated healthy Han Chinese infants. We obtained written informed consent from the parents and collected and analyzed samples with the approval of the institutional review board of Fudan University.
We identified three novel missense mutations in VANGL2. All were heterozygous in fetuses with a cranial neural-tube defect: S84F (737CT), R353C (1543CT), and F437S (1796TC). R353C was detected in a male fetus at 21 weeks' gestation. This fetus had anencephaly with occipital and cervical spina bifida. F437S was detected in a male fetus at 24 weeks' gestation with anencephaly, and S84F was detected in a female fetus at 22 weeks' gestation with holoprosencephaly. All three mutations affect conserved residues in VANGL2 proteins across species (see the Figure in the Supplementary Appendix) and were absent in controls. The prevalence of other variants was similar among cases and controls (Table 2 in the Supplementary Appendix).
R353 and F437 are located in the cytoplasmic domain, adjacent to the carboxy-terminal PDZ-binding domain. The mutations R353C and F437S are predicted to affect protein structure, and both affect residues that are highly conserved across species (see the Figure in the Supplementary Appendix). Similarly positioned mutations (D255E and S464N) of Vangl2 in mice have been shown to affect Vangl2 function, and they are predicted to disrupt interactions with the cytoplasmic protein, disheveled (Dvl).3,4 S84F predicts the substitution of a serine residue at position 84 (which is highly conserved across species) with a phenylalanine residue (see the Figure in the Supplementary Appendix). Its association with holoprosencephaly is uncertain.
Using a yeast two-hybrid system, we tested the ability of VANGL2 mutants (carrying either the R353C or the F437S mutation) to bind Dvl. All constructs were stably expressed at similar levels (Figure 1A). F437S completely abrogated interaction with Dvl, whereas R353C diminished but did not abolish this interaction (Figure 1B, 1C, and 1D). In contrast, and serving as a positive control, was the interaction between nonmutant VANGL2 and Dvl.
Because we identified VANGL2 mutations in miscarried fetuses with severe cranial neural-tube defects, we surmise that their lethal effect during in utero development precludes their presence in living persons with less severe defects. Our results provide support for studies that emphasize the role of planar-cell-polarity genes in neural-tube closure, although craniorachischisis, not anencephaly, is the invariable in mice that are homozygously deficient in Vangl2.5
