Insights into the Etiology of Mammalian Neural Tube Closure Defects from Developmental, Genetic and Evolutionary Studies.

The human neural tube defects (NTD), anencephaly, spina bifida and craniorachischisis, originate from a failure of the embryonic neural tube to close. Human NTD are relatively common and both complex and heterogeneous in genetic origin, but the genetic variants and developmental mechanisms are largely unknown. Here we review the numerous studies, mainly in mice, of normal neural tube closure, the mechanisms of failure caused by specific gene mutations, and the evolution of the vertebrate cranial neural tube and its genetic processes, seeking insights into the etiology of human NTD. We find evidence of many regions along the anterior⁻posterior axis each differing in some aspect of neural tube closure-morphology, cell behavior, specific genes required-and conclude that the etiology of NTD is likely to be partly specific to the anterior⁻posterior location of the defect and also genetically heterogeneous. We revisit the hypotheses explaining the excess of females among cranial NTD cases in mice and humans and new developments in understanding the role of the folate pathway in NTD. Finally, we demonstrate that evidence from mouse mutants strongly supports the search for digenic or oligogenic etiology in human NTD of all types.

Epigenetic mechanism causes Wnt9b deficiency and nonsyndromic cleft lip and palate in the A/WySn mouse strain.

BACKGROUND: The heritable multifactorial etiology of human nonsyndromic cleft lip with or without cleft palate (CL ± P) is not understood. CL ± P occurs in 15% of neonates in the homozygous A/WySn mouse strain, with a multifactorial genetic etiology, the clf1 and clf2 variant genes. Clf1 acts as a mutant allele of Wnt9b but its coding sequence is normal. An IAP (intracisternal A particle) retrotransposon inserted near the Wnt9b gene is associated with clf1. METHODS: Transcription of noncoding sequence between the IAP and the Wnt9b gene was examined in A/WySn embryos. The levels of Wnt9b transcript and of an "IAP antisense" transcript initiated in the IAP and extending into the noncoding interval were assayed in A/WySn and C57BL/6J whole embryos or heads across embryonic days 8 to 12. Methylation of the 5' LTR of the IAP was examined in E12 A/WySn embryo heads. RESULTS: Mean Wnt9b transcript levels were lower in A/WySn than in C57BL/6J at all ages examined and lower in CL ± P embryos than in their normal littermates. The "IAP antisense" transcript was found in all A/WySn embryos and was highest in CL ± P embryos. The IAP at Wnt9b was generally unmethylated in CL ± P embryos and approximately 50% methylated in normal littermates. CONCLUSION: The clf1 mutation in A/WySn is a "metastable epiallele", in which stochastic deficiency in some individuals of DNA methylation of a retrotransposon uniquely inserted near the Wnt9b gene allows transcriptional activity of the retrotransposon and interference with transcription from Wnt9b. Methylation of metastable epialleles should be investigated in human nonsyndromic CL ± P.

A consideration of the evidence that genetic defects in planar cell polarity contribute to the etiology of human neural tube defects.

A variety of human birth defects originate in failure of closure of the embryonic neural tube. The genetic cause of the most common nonsyndromic defects, spina bifida (SB) or anencephaly, is considered to be combinations of variants at multiple genes. The genes contributing to the etiology of neural tube closure defects (NTDs) are unknown. Mutations in planar cell polarity (PCP) genes in mice cause a variety of defects including the NTD, craniorachischisis, and sometimes SB or exencephaly (EX); they also demonstrate the role of digenic combinations of PCP mutants in NTDs. Recent studies have sought rare predicted-to-be-deleterious alterations (putative mutations) in coding sequence of PCP genes in human cases with various anomalies of the neural tube. This review summarizes the cumulative results of these studies according to a framework based on the embryopathogenesis of NTDs, and considers some of the insights from the approaches used and the limitations. Rare putative mutations in the PCP genes VANGL2, SCRIB, DACT1, and CELSR1 cumulatively contributed to over 20% of cases with craniorachischisis, a rare defect; no contributing variants were found for PRICKLE1 or PTK7. PCP rare putative mutations had a weaker role in myelomeningocele (SB), being found in approximately 6% of cases and cumulated across CELSR1, FUZ, FZD6, PRICKLE1, VANGL1, and VANGL2. These results demonstrate that PCP gene alterations contribute to the etiology of human NTDs. We recommend that future research should explore other types of PCP gene variant such as regulatory mutations and low frequency (1 to 5%) deleterious polymorphisms.

Hypothesis: the female excess in cranial neural tube defects reflects an epigenetic drag of the inactivating X chromosome on the molecular mechanisms of neural fold elevation.

Females have long been known to be in excess among cranial neural tube defect (NTD) cases. Up to two thirds of human anencephalics and mouse exencephalics from various genetic causes are female, but the cause of this female excess is unknown. It appears not to be attributable to gonadal hormones, developmental delay in females, or preferential death of affected males. Recent studies of the Trp53 mouse mutant showed that exencephaly susceptibility depends on the presence of two X chromosomes, not the absence of the Y. Over a decade ago, we hypothesized that the relevant difference between female and male mammalian embryos at the time of cranial neural tube closure is the fact that females methylate most of the DNA in the large inactive X chromosome after every cell division, reducing the methylation available for other needs in female cells. Recently, the Whitelaw laboratory identified several proteins in mice (Momme D genes) involved in epigenetic silencing and methylation and shared in the silencing of transgenes, retrotransposons, and the inactive-X, and suggested that the inactive-X acts as a "sink" for epigenetic silencing proteins. The "inactive-X sink" hypothesis can be used to suggest expected changes in sex ratio in cranial NTDs in response to various genetic or environmental alterations. We recommend that observation of sex ratio become a standard component of all NTD studies. We suggest that the female excess among cranial NTDs is an epigenetic phenomenon whose molecular investigation will produce insight into the mechanisms underlying NTDs.

Strain-specific modifier genes of Cecr2-associated exencephaly in mice: genetic analysis and identification of differentially expressed candidate genes.

Although neural tube defects (NTDs) are common in humans, little is known about their multifactorial genetic causes. While most mouse models involve NTDs caused by a single mutated gene, we have previously described a multigenic system involving susceptibility to NTDs. In mice with a mutation in Cecr2, the cranial NTD exencephaly shows strain-specific differences in penetrance, with 74% penetrance in BALB/cCrl and 0% penetrance in FVB/N. Whole genome linkage analysis showed that a region of chromosome 19 was partially responsible for this difference in penetrance. We now reveal by genetic analysis of three subinterval congenic lines that the chromosome 19 region contains more than one modifier gene. Analysis of embryos showed that although a Cecr2 mutation causes wider neural tubes in both strains, FVB/N embryos overcome this abnormality and close. A microarray analysis comparing neurulating female embryos from both strains identified differentially expressed genes within the chromosome 19 region, including Arhgap19, which is expressed at a lower level in BALB/cCrl due to a stop codon specific to that substrain. Modifier genes in this region are of particular interest because a large portion of this region is syntenic to human chromosome 10q25, the site of a human susceptibility locus.

The clf2 gene has an epigenetic role in the multifactorial etiology of cleft lip and palate in the A/WySn mouse strain.

BACKGROUND: The A/WySn mouse strain with 15 to 20% penetrance of cleft lip and palate (CLP) is an animal model for human multifactorial CLP. The CLP is due to two unlinked genes that interact epistatically, Wnt9b(clf1) and clf2, plus a maternal effect. The Wnt9b(clf1) mutation is an IAP transposon insertion. The clf2 gene, with unknown function, was located in a 13.6 Mb region of chromosome 13 containing 145 genes. METHODS: To reduce the clf2 candidate region, 1146 mice segregating for A/WySn and C57BL/6J alleles at clf2 were screened for recombinants by simple sequence-length polymorphism haplotypes; recombinants' testcross progeny were typed for CLP and simple-sequence length polymorphisms. To identify the function of clf2, the effect of clf2 genotype on risk of CLP was tested in Wnt9b(null/null) knockouts and in compound mutants (Wnt9b(clf1/null) ), and the methylation of the IAP at Wnt9b was assayed in the Wnt9b(clf1/null) mutants by combined bisulfite restriction analysis. RESULTS: The location of clf2 was redefined to 3.0 Mb between Cntnap3 and AK029746 containing 48 genes, of which 30 are Zfp genes. The clf2 genotype had no detectable effect on Wnt9b(null/null) embryos, but strongly affected risk of CLP and methylation of the IAP in Wnt9b(clf1/null) embryos. CLP was associated with low levels of methylation of the IAP. CONCLUSIONS: The clf2 gene is the first identified polymorphism that affects the epigenetic methylation and silencing of IAP retrotransposons. This CLP model raises the question of whether parallel epigenetic factors are involved in risk and environmental sensitivity of human CLP.

An update to the list of mouse mutants with neural tube closure defects and advances toward a complete genetic perspective of neural tube closure.

The number of mouse mutants and strains with neural tube defects (NTDs) now exceeds 240, including 205 representing specific genes, 30 for unidentified genes, and 9 multifactorial strains. These mutants identify genes needed for embryonic neural tube closure. Reports of 50 new NTD mutants since our 2007 review (Harris and Juriloff, 2007) were considered in relation to the previously reviewed mutants to obtain new insights into mechanisms of NTD etiology. In addition to null mutations, some are hypomorphs or conditional mutants. Some mutations do not cause NTDs on their own, but do so in digenic, trigenic, and oligogenic combinations, an etiology that likely parallels the nature of genetic etiology of human NTDs. Mutants that have only exencephaly are fourfold more frequent than those that have spina bifida aperta with or without exencephaly. Many diverse cellular functions and biochemical pathways are involved; the NTD mutants draw new attention to chromatin modification (epigenetics), the protease-activated receptor cascade, and the ciliopathies. Few mutants directly involve folate metabolism. Prevention of NTDs by maternal folate supplementation has been tested in 13 mutants and reduces NTD frequency in six diverse mutants. Inositol reduces spina bifida aperta frequency in the curly tail mutant, and three new mutants involve inositol metabolism. The many NTD mutants are the foundation for a future complete genetic understanding of the processes of neural fold elevation and fusion along mechanistically distinct cranial-caudal segments of the neural tube, and they point to several candidate processes for study in human NTD etiology.

Accelerated embryonic development associated with increased risk of neural tube defects induced by maternal diet in offspring of SELH/Bc mice.

BACKGROUND: The SELH/Bc mouse strain has a high risk of the NTD, exencephaly, caused by multifactorial genetics. All SELH/Bc embryos have delayed elevation of neural folds; some never elevate (future exencephalics). Maternal diets affect SELH/Bc exencephaly rates: 25-35% on Purina Diet 5015 versus 5-10% on Purina Diet 5001. We hypothesized that in SELH/Bc, the diets affect maternal blood glucose and embryonic developmental rate. METHODS: We compared mice fed the two diets. On GD 9.4 we tested maternal blood glucose and examined embryos for developmental age (somite count) and cranial neural fold morphology. We observed GD 14 exencephaly rates. RESULTS: Diet 5015 caused fivefold more exencephaly (40 vs. 7% on GD 14), significantly higher mean maternal blood glucose in replicate experiments (6.3 vs. 5.5, p < .05; 6.3 vs. 5.3 mmol/L, p < .05), and significantly higher mean litter somite count on GD 9.4 (18.4 vs. 15.0, p < .05; 16.7 vs. 14.4 somites, p < .05). Among midrange embryos (15-16 somites), embryos from Diet 5015 were significantly shifted to earlier stages of midbrain fold morphology and had significantly more distance between the tips of the folds (p < .05). CONCLUSIONS: In SELH/Bc mice, the 5015 diet causes higher maternal blood glucose, a faster overall embryonic developmental rate during neural tube closure, and delayed midbrain fold elevation relative to overall development. This pattern suggests that maternal dietary effects that modestly increase embryonic growth rate may exacerbate a lack of coordination between genetically delayed neural folds and normally developing underlying tissues, increasing risk of NTD.

Mouse genetic models of cleft lip with or without cleft palate.

Nonsyndromic cleft lip and palate (CLP) is among the most common human birth defects. Transmission patterns suggest that the causes are "multifactorial" combinations of genetic and nongenetic factors, mostly distinct from those causing cleft secondary palate (CP). The major etiological factors are largely unknown, and the embryological mechanisms are not well understood. In contrast to CP or neural tube defects (NTD), CLP is uncommon in mouse mutants. Fourteen known mutants or strains express CLP, often as part of a severe syndrome, whereas nonsyndromic CLP is found in two conditional mutants and in two multifactorial models based on a hypomorphic variant with an epigenetic factor. This pattern suggests that human nonsyndromic CLP is likely caused by regulatory and hypomorphic gene variants, and may also involve epigenetics. The developmental pathogenic mechanism varies among mutants and includes deficiencies of growth of the medial, lateral or maxillary facial prominences, defects in the fusion process itself, and shifted midline position of the medial prominences. Several CLP mutants also have NTD, suggesting potential genetic overlap of the traits in humans. The mutants may reflect two interacting sets of genetic signaling pathways: Bmp4, Bmpr1a, Sp8, and Wnt9b may be in one set, and Tcfap2a and Sox11 may be in another. Combining the results of chromosomal linkage studies of unidentified human CLP genes with insights from the mouse models, the following previously unexamined genes are identified as strong candidate genes for causative roles in human nonsyndromic CLP: BMP4, BMPR1B, TFAP2A, SOX4, WNT9B, WNT3, and SP8.

Mouse mutants with neural tube closure defects and their role in understanding human neural tube defects.

BACKGROUND: The number of mouse mutants and strains with neural tube closure defects (NTDs) now exceeds 190, including 155 involving known genes, 33 with unidentified genes, and eight "multifactorial" strains. METHODS: The emerging patterns of mouse NTDs are considered in relation to the unknown genetics of the common human NTDs, anencephaly, and spina bifida aperta. RESULTS: Of the 150 mouse mutants that survive past midgestation, 20% have risk of either exencephaly and spina bifida aperta or both, parallel to the majority of human NTDs, whereas 70% have only exencephaly, 5% have only spina bifida, and 5% have craniorachischisis. The primary defect in most mouse NTDs is failure of neural fold elevation. Most null mutations (>90%) produce syndromes of multiple affected structures with high penetrance in homozygotes, whereas the "multifactorial" strains and several null-mutant heterozygotes and mutants with partial gene function (hypomorphs) have low-penetrance nonsyndromic NTDs, like the majority of human NTDs. The normal functions of the mutated genes are diverse, with clusters in pathways of actin function, apoptosis, and chromatin methylation and structure. The female excess observed in human anencephaly is found in all mouse exencephaly mutants for which gender has been studied. Maternal agents, including folate, methionine, inositol, or alternative commercial diets, have specific preventative effects in eight mutants and strains. CONCLUSIONS: If the human homologs of the mouse NTD mutants contribute to risk of common human NTDs, it seems likely to be in multifactorial combinations of hypomorphs and low-penetrance heterozygotes, as exemplified by mouse digenic mutants and the oligogenic SELH/Bc strain.

Wnt9b is the mutated gene involved in multifactorial nonsyndromic cleft lip with or without cleft palate in A/WySn mice, as confirmed by a genetic complementation test.

BACKGROUND: Nonsyndromic cleft lip (CL) with or without cleft palate (CLP) is a common human birth defect with complex genetic etiology. One of the unidentified genes maps to chromosome 17q21. A mouse strain, A/WySn, has CLP with complex genetic etiology that models the human defect, and 1 of its causative genes, clf1, maps to a region homologous to human 17q21. Extensive studies of the candidate region pointed to a novel insertion of an IAP transposon 3' from the gene Wnt9b as the clf1 mutation. Independently a recessive knockout mutation of Wnt9b (Wnt9b-) was reported to cause a lethal syndrome that includes some CLP. METHODS: A standard genetic test of allelism between clf1 and the Wnt9b- mutation was done. A total of 83 F1 embryos at gestation day 14 (GD 14) from Wnt9b-/+ males crossed with A/WySn females, and 79 BC1 GD 14 embryos from F1 Wnt9b-/clf1 males back-crossed to A/WySn females were observed for CL. Embryo genotypes at clf1 and Wnt9b were obtained from DNA markers. Genotypes for a second unlinked modifier locus from A/WySn, clf2, were similarly obtained. RESULTS: The compound mutant embryos (Wnt9b-/clf1) had high frequencies of CL: 27% in the F1 and 63% in the BC1. The clf2 modifier gene was found to have 3 alleles segregating in this study and to strongly influence the penetrance of CL in the compound mutant. CONCLUSIONS: The noncomplementation of clf1 and Wnt9b- confirms that clf1 is a mutation of the Wnt9b gene. The homologous human WNT9B gene and 3' conserved noncoding region should be examined for a role in human nonsyndromic CLP.

Maternal diet alters exencephaly frequency in SELH/Bc strain mouse embryos.

BACKGROUND: The SELH/Bc mouse inbred strain, with a high frequency of nonsyndromic, genetically-multifactorial exencephaly, is a model for human cranial neural tube defects (NTDs). Maternal diet affects risk of human NTDs. METHODS: Exencephaly frequencies in SELH/Bc embryos were compared in 8 studies in which dams were fed alternative commercial Purina diets (5015 and 5001) or semisynthetic diets, and in several studies in which maternal diet was supplemented with a specific nutrient, either in drinking water or food before and during pregnancy, or by intraperitoneal injection on E7 and/or E8. RESULTS: The exencephaly frequency in SELH/Bc embryos was 2- to 8-fold higher when the dams were fed Purina 5015 (averaging 23% exencephaly) or a semisynthetic diet modeled on Purina 5015 (averaging 28%) or NIH-31 standard diet (23%), compared with Purina 5001 (averaging 7%). The exencephaly frequency remained high (41%) on a semisynthetic diet modeled on Purina 5001. The exencephaly frequency was not reduced significantly by maternal supplementation with folic acid, nor with each of zinc, methionine, niacin, brewers' yeast, riboflavin, vitamin B12, or inositol. Nor was it reduced by maternal diets with supplemental methyl donors and cofactors or with reduced fat. CONCLUSIONS: The frequency of exencephaly in SELH/Bc embryos is strongly influenced by a specific unidentified aspect of the commercial ration Purina 5001 that prevents 55-85% of exencephaly in SELH/Bc embryos, when directly compared with an alternative commercial ration Purina 5015 or its semisynthetic mimic. This strong maternal diet effect on NTD frequency may point to novel nutritional approaches to prevention of human NTDs.

Investigations of the genomic region that contains the clf1 mutation, a causal gene in multifactorial cleft lip and palate in mice.

BACKGROUND: Human nonsyndromic cleft lip and palate, CL(P), is genetically complex, with one contributing gene on chromosome 17q. A potentially homologous gene, clf1 on distal chromosome 11, is part of the digenic cause of the 10-30% CL(P) in the A/WySn mouse strain. Here we report our progress toward identifying the clf1 mutation. METHODS: Transcription from all of the known and predicted genes in the 1.5-Mb candidate region was examined in A/WySn and control (AXB-4/Pgn) ED10-11 embryo heads. The marker haplotype for 28 inbred strains across the clf1 region was obtained. The entire transcripts of Wnt9b and Wnt3 in A/WySn were sequenced. Using long PCR, the genomic region from Wnt3 throughWnt9b was screened in A/WySn for an inserted retrotransposon. RESULTS: Gosr2, Wnt9b, Wnt3, Nsf, Arf2, Crhr1, Mapt, Cdc27, Myl4, Itgb3, chr11_20.152, chr11_20.154, chr11_20.155, and chr11_20.156 are expressed in ED10-11 heads. None is absent or detectably reduced in A/WySn. The ancestral pre-clf1 mutation haplotype was found in CBA/J mice. By a test-cross, CBA/J was confirmed to lack the clf1 mutation. Three single-nucleotide variants in A/WySn (vs. C57BL/6J) were found in each of the 3' untranslated regions (3'UTRs) of Wnt3 and of Wnt9b, respectively; their presence in CBA/J shows that none are the clf1 mutation. An inserted intracisternal A particle (IAP) retrotransposon located 6.6 kb from the 3' end of Wnt9b was found in A/WySn and in all clf1 strains tested. This IAP is absent in C57BL/6J and CBA/J. CONCLUSIONS: The clf1 mutation is a genomic alteration present in A/WySn and absent in the ancestral chromosomal segment in CBA/J. The IAP retrotransposon insertion near Wnt9b in A/WySn fits this criterion; we predict that interference with Wnt9b function by this IAP is the clf1 mutation.

The open-eyelid mutation, lidgap-Gates, is an eight-exon deletion in the mouse Map3k1 gene.

The BALB/cGa mouse strain and its descendants, now called the SELH/Bc strain, have produced two waves of high frequency of spontaneous heritable mutations. One of these, the recessive lidgap-Gates (lg(Ga)) mutation, causes the same open-eyelids-at-birth phenotype as the gene knockout mutations of Map3k1 and co-maps to distal Chr 13. The lg(Ga) mutation is demonstrated to be a 27.5-kb deletion of exons 2-9 in the Map3k1 gene, the first spontaneous mutant allele described at this locus. The lg(Ga) mutation is consistent with a pattern suggesting that the waves of mutation in BALB/cGa and its descendants tend to be large deletions or ETn insertions, whose elevated rate of occurrence is due to an unknown mechanism.

A digenic cause of cleft lip in A-strain mice and definition of candidate genes for the two loci.

BACKGROUND: Nonsyndromic cleft lip with or without cleft palate, CL(P), is a common human birth defect with a complex unknown genetic cause. The mouse model is the "A/-" strains. Our previous studies mapped two loci: clf1 on Chr11 and clf2 on Chr13--with a strong genetic maternal effect on the level of risk. Here we test the hypothesis that CL(P) is digenic and identify candidate genes for clf1 and clf2. METHODS: We observed E14 CL(P) frequencies in backcross (BC1) embryos from a new cross of A/WySn to AXB-4/Pgn and from test crosses of three new "congenic RI" lines. Using new polymorphic markers from genes and our mapping panels of segregants and RI strains, we identified the candidate genes for clf1 and clf2. We sequenced the coding region of Ptch in A/WySn cDNA. RESULTS: Seventy new BC1 CL(P) segregants (4%) were obtained, as predicted. All three new congenic RI lines homozygous for both clf1 and clf2 had A/WySn-level CL(P) frequencies (10-30%) in test crosses. The clf1 region contains 10 known genes (Arf2, Cdc27, Crhr1, Gosr2, Itgb3, Mapt, Myl4, Nsf, Wnt3, and Wnt9b). The clf2 region contains 17 known genes with human orthologs. Both regions contain additional potential genes. No causal mutation in Ptch coding sequence was found. CONCLUSIONS: In A-strain mice, nonsyndromic CL(P) is digenic, suggesting that nonsyndromic human CL(P) may also be digenic. The orthologous human genes are on 17q (clf1) and 9q, 8q and 5p (clf2), and good candidate genes are WNT3 or WNT9B (17q), and PTCH (9q) or MTRR (5p).

Structure and expression of mobile ETnII retroelements and their coding-competent MusD relatives in the mouse.

ETnII elements are mobile members of the repetitive early transposon family of mouse long terminal repeat (LTR) retroelements and have caused a number of mutations by inserting into genes. ETnII sequences lack retroviral genes, but the recent discovery of related MusD retroviral elements with regions similar to gag, pro, and pol suggests that MusD provides the proteins necessary for ETnII transposition in trans. For this study, we analyzed all ETnII elements in the draft sequence of the C57BL/6J genome and classified them into three subtypes (alpha, beta, and gamma) based on structural differences. We then used database searches and quantitative real-time PCR to determine the copy number and expression of ETnII and MusD elements in various mouse strains. In 7.5-day-old embryos of a mouse strain in which two mutations due to ETnII-beta insertions have been identified (SELH/Bc), we detected a three- to sixfold higher level of ETnII-beta and MusD transcripts than in control strains (C57BL/6J and LM/Bc). The increased ETnII transcription level can in part be attributed to a higher number of ETnII-beta elements, but 70% of the MusD transcripts appear to have been derived from one or a few MusD elements that are not detectable in C57BL/6J mice. This element belongs to a young MusD subgroup with intact open reading frames and identical LTRs, suggesting that the overexpressed element(s) in SELH/Bc mice might provide the proteins for the retrotransposition of ETnII and MusD elements. We also show that ETnII is expressed up to 30-fold more than MusD, which could explain why only ETnII, but not MusD, elements have been positively identified as new insertions.

exma: an X-linked insertional mutation that disrupts forebrain and eye development.

Formation of the neural tube plays a primary role in establishing the body plan of the vertebrate embryo. Here we describe the phenotype and physical mapping of a highly penetrant X-linked male-lethal murine mutation, exma (exencephaly, microphthalmia/anophthalmia), that specifically disrupts development of the rostral neural tube and eye. The mutation arose from the random insertion of a transgene into the mouse X Chromosome (Chr). Eighty-three percent of transgenic male embryos display an open, disorganized forebrain and lack optic vesicles. No transgenic males survive beyond birth. Hemizygous females show a variable phenotype, including reduced viability and occasional exencephaly and/or microphthalmia. Altered or reduced expression patterns of Otx2, Pax6, Six3, and Mrx, known markers of early forebrain and eye development, confirmed the highly disorganized structure of the forebrain and lack of eye development in affected exma male embryos. Physical mapping of the transgene by FISH localized a single insertion site to the interval between Dmd and Zfx on the X Chr. A 1-Mb contig of BAC clones was assembled by using sequences flanking the transgene and revealed that the insertion lies close to Pola1 and Arx, a gene encoding a highly conserved homeobox protein known to be expressed in the developing forebrain of the mouse. Data from Southern blots of normal and transgenic DNA demonstrated that a large segment of DNA encompassing Arx and including part of Pola1 was duplicated as a result of the transgene insertion. From the physical mapping results, we propose a model of the gross rearrangements that accompanied transgene integration and discuss its implications for evaluating candidate genes for exma.