A Haplolethal Locus Uncovered by Deletions in the Mouse t Complex

Corresponding author: John C. Schimenti, 600 Main St., Bar Harbor, ME 04609., jcs{at}jax.org (E-mail)

Communicating editor: D. M. KINGSLEY

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Proper levels of gene expression are important for normal mammalian development. Typically, altered gene dosage caused by karyotypic abnormalities results in embryonic lethality or birth defects. Segmental aneuploidy can be compatible with life but often results in contiguous gene syndromes. The ability to manipulate the mouse genome allows the systematic exploration of regions that are affected by alterations in gene dosage. To explore the effects of segmental haploidy in the mouse t complex on chromosome 17, radiation-induced deletion complexes centered at the Sod2 and D17Leh94 loci were generated in embryonic stem (ES) cells. A small interval was identified that, when hemizygous, caused specific embryonic lethal phenotypes (exencephaly and edema) in most fetuses. The penetrance of these phenotypes was background dependent. Additionally, evidence for parent-of-origin effects was observed. This genetic approach should be useful for identifying genes that are imprinted or whose dosage is critical for normal embryonic development.

DELETION syndromes, also referred to as segmental aneusomy syndromes or contiguous gene syndromes (CGSs), such as Cri-du-chat [Online Mendelian Inheritance in Man (OMIM) 123450], Smith-Magenis (OMIM 182290), Wolf-Hirschhorn (OMIM 194190), and Williams (OMIM 194050), occur when an individual has a deletion-bearing chromosome in trans to a normal chromosome; the disease is thus a consequence of hemizygosity for one or more genes. CGSs illustrate that diploidy of certain genes is critical for proper development. Presumably, the requirement for diploidy reflects the necessity for precisely controlled levels of a particular gene product.

The molecular bases of CGSs are difficult to study, as the phenotypes are complex and the deletions in individual patients are usually large. Thus, it is unclear whether the various phenotypes are caused by haploinsufficiency of one gene or of multiple genes. Dissection of these syndromes is complicated by the fact that the deletions associated with syndromes such as Prader-Willi, Williams, and DiGeorge have common endpoints, having arisen from recurring unequal recombination events catalyzed by particular duplicated sequences (DUTLY and SCHINZEL 1996 ; URBAN et al. 1996 ; AMOS-LANDGRAF et al. 1999 ). It is possible that CGSs are caused by a "synergistic" effect of multiple deleted genes, since heterozygosity for most null mutations typically has no adverse effects. However, there are cases where heterozygosity for null alleles causes marked phenotypes, but these genes are not always associated with CGSs [e.g., Alagille syndrome has been shown to be caused by a mutation in Jagged1 (LI et al. 1997 ; ODA et al. 1997 )].

Although triploidy and haploidy of chromosomal regions in humans often cause viable or semiviable developmental defects, it is difficult to identify regions of the genome that are haplolethal. One study found that 11% of the genome was never found to be deleted in malformation-associated chromosomal regions, indirectly suggesting that these regions contained gestational haplolethal genes (BREWER et al. 1998 , BREWER et al. 1999 ). One conclusion that can be drawn is that whereas most segmental aneuploidies cause birth defects, just a few regions may be strictly haplolethal. Genes in such regions may be of critical developmental importance in terms of precise requirements for gene product levels. Further complicating the conclusions drawn from human studies is that each cytogenetic variant is a unique case, so issues of penetrance, genetic background, possible semilethality, and expressivity are difficult to distinguish.

Technologies to manipulate the mouse genome afford opportunities to model and genetically dissect human CGSs. Deletions generated in embryonic stem (ES) cells by Cre/loxP-mediated recombination or irradiation have been used to model the Prader-Willi (YANG et al. 1998 ; GABRIEL et al. 1999 ; TSAI et al. 1999 ), DiGeorge (LINDSAY et al. 1999 , LINDSAY et al. 2001 ; MERSCHER et al. 2001 ), and Wolf-Hirschhorn (NAF et al. 2001 ) syndromes and are providing valuable information toward identifying the genes that are responsible for the phenotypes of these diseases. These technologies can also be harnessed to identify imprinted or haplolethal genes and to identify clusters of genes whose coordinate reduction to hemizygosity is essential for phenotypic consequence. Identification of these genes and gene interactions may provide useful clues into those developmental pathways that are the most exquisitely sensitive to levels of particular gene products.

To develop a comprehensive set of resources for functional analysis of the mouse t complex on chromosome 17, a region containing numerous genes required for development and gametogenesis (SILVER 1985 ), sets of deletion complexes are being generated. This laboratory previously reported the development of an ES cell irradiation technology that was used to induce a collection of deletions centered at the centromeric end of the t complex, at a locus called D17Aus9 (YOU et al. 1997A ). Here, the creation of additional deletion complexes centered at the Sod2 and D17Leh94 is reported. Examination of animals harboring nested breakpoints resulted in the identification of a small region that, in the hemizygous state, causes severe exencephaly or edema in animals. Additionally, females carrying the deletion are fertile but do not transmit it to their offspring, suggesting the presence of an imprinted gene.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Inducing deletions in ES cells:
The generic targeting scheme of YOU et al. 1997B was used to build targeting vectors for the D17Leh94 and Sod2 loci. Genomic DNA fragments from these loci were isolated from a 129/Sv lambda Fix II library (Stratagene, La Jolla, CA). For D17Leh94, a 10-kb SacI fragment was first subcloned into the SacI site of pSKBam and then a cassette containing the neomycin resistance (neo) and Herpes simplex virus thymidine kinase (tk) genes from the pBAN/TK cassette was ligated into a unique BamHI site in the 10-kb insert. The resulting plasmid was linearized with NotI and electroporated into v6.4 ES cells, which are of the genotype (129S4SvJae x C57BL/6J)F₁ (YOU et al. 1998 ). DNA from clones that were resistant to 300 µg/µl of geneticin (Life Technologies) were digested with ScaI and probed with a 1.58-kb SacI fragment, located 1.5 kb downstream of the tk/neo integration site, to confirm proper targeting. For purposes of determining the exact location of the D17Leh94 locus relative to other sequences in the Celera mouse genome assembly, the sequence of the probe fragment used to screen the genomic library, originally called Tu94 (ROHME et al. 1984 ), was determined and submitted to GenBank (accession no. AF408196).

To target the Sod2 locus, an 8-kb SacI fragment from a Sod2-positive -phage clone was subcloned into pSKBam, the tk/neo cassette (described above) was ligated into a unique BamHI site in the mouse fragment, and the resulting vector was linearized with ClaI and electroporated into v6.4 cells. DNA from geneticin-resistant clones was digested with EcoRI and probed with a PCR-generated 850-bp fragment from Sod2 exon 4.

The methodology for generating and selecting ES cell lines containing deletions of the targeted insertions has been described (YOU et al. 1997A , YOU et al. 1997B ).

Identification of deletion-bearing ES cell clones and mapping of deletion breakpoints:
We identified in a stepwise fashion 1-2'-deoxy-2'-fluoro-ß-D-arabinofuranosyl-5-iodouracil (FIAU)-resistant clones containing deletions of desired sizes. First, clones that arose due to transcriptional inactivation of the tk gene rather than deletion were identified by PCR (using primers for the tk gene) or Southern blot analysis (to determine presence or absence of the targeted allele). Second, clones that lost the neo/tk cassette by mitotic recombination or huge deletions were identified by PCR analysis of a simple sequence length polymorphism (SSLP) marker (D17Mit93) that is polymorphic between the 129/SvJae ("129") and C57BL/6J ("B6") parental chromosomes and is located distal to the t complex at centimorgan position 45. Remaining clones were typed for a number of polymorphic SSLP markers within the t complex to determine the extent of each deletion.

Deletion breakpoints were further narrowed by identifying new B6 vs. 129 microsatellite polymorphisms from the Celera database of mouse DNA sequence. These markers, shown in Fig 1, and their corresponding primer sequences are listed below:

• D17Jcs1 GAAGTGAGAGTCGCATGGTG TGCTTCACACACTTCCCAAA

  
  

  View larger version (29K): In this window In a new window Download PPT slide   Figure 1. Map of Sod2 and D17Leh94 deletions. The proximal region of chromosome 17 is depicted as a horizontal line, with the centromere (circle) on the left. Map positions (in megabases, Mb) were deduced as described in MATERIALS AND METHODS. Microsatellite loci shown above the chromosome are abbreviated by exchanging the prefix "M" for "D17Mit." Those below the chromosomes, which were created for this study, are abbreviated by exchanging the prefix "J" for "D17Jcs." Markers below the chromosome correspond to microsatellite repeats developed in this report (see MATERIALS AND METHODS). Deletions are indicated as horizontal rectangles and are color coded to the locus at which they were induced (red, D17Leh94; blue, Sod2; black, D17Aus9). The amount of DNA known to be absent in each deletion is spanned by the rectangles. The thin lines extending from the ends of the rectangles indicate the regions in which the deletion breakpoints reside. The critical region containing the haplolethal locus is contained within the shaded area.

• D17Jcs8 TACCACCAGGACAGGGCTAT CGATGACGAAGAAGGTGATG

• D17Jcs11 CCCCTAGGCCCTTATGGTAA TGCTAGCCACATGGACATTC

• D17Jcs35 TGAGGAAGTACTTGCATACACCA CAGGAGCCTCTAGCCAAGTG

• D17Jcs37 TGCCAGTAACATGAGATGCAG GGCTTATTTTCATTTCCATGC

• D17Jcs80 GTTTGATCCCTGGAACTTGC GCAAAACGGTCAGATGTTTG

Multiple cell lines that carried deletions within the t complex (as judged by Southern and tk PCR data) were isolated but did not remove any flanking Mit markers. For example, 8 of 11 cell lines in the first Sod2 irradiation experiment deleted tk but not any Mit markers; none of these cell lines were used to make chimeras. For both targeted cell lines, there was a high background of FIAU-resistant clones that did not appear to actually carry a deletion. For example, of 288 colonies picked from a deletion experiment using the SodC1 targeted cell line, only 47 (16%) of the cell lines were genotyped as having a deletion of tk. However, the background was almost completely eliminated by isolating subclones that were free of tk nonexpressing cells. A larger set of deletions derived from this subclone will be described elsewhere.

Generation of mice bearing deletions:
ES cell clones containing deletions of interest were injected into blastocysts derived from B6, or in some cases FVB, females. If any progeny sired by the chimeras had agouti fur, all animals in the litter were genotyped (regardless of coat color) using at least two SSLP markers from chromosome 17 to determine if they carried the deletion (note that because the F₁ hybrid ES cells were heterozygous for the B6 genome, progeny from a chimera mated to B6 could have black fur, but still inherit the ES cell genome). Ideally, markers on the centromeric and telomeric flanks of the deletion were used; otherwise a marker from within the deletion was informative, depending on the nature of the cross. All but one of the Sod2 deletion lines encompassed the T locus, so these animals could be tracked by the short-tailed Brachyury phenotype. Potential rare double recombination events (rather than deletions) could be detected by crossing males harboring presumed deletions of the 129 chromosome to strains B6 and C3H and testing progeny with markers within the deletion that are polymorphic among all three strains.

Husbandry and timed matings:
All mice were reared at The Jackson Laboratory. For timed matings, the morning that vaginal plugs were detected was considered embryonic gestation day 0.5 (E0.5).

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Generation of two new deletion complexes on mouse chromosome 17:
Two panels of deletions with focal points within the central t complex were generated by the technique of ES cell irradiation (YOU et al. 1997A ). The deletions are centered about Sod2 (a superoxide dismutase) and D17Leh94 (an anonymous DNA marker; see Fig 1). Targeting vectors generated from these loci were used to integrate, by homologous recombination, a Herpes simplex virus tk gene into the cognate locus of (129S4SvJae x C57BL/6J)F₁ hybrid ES cells. Germline-competent, targeted clones from each locus were exposed to ionizing radiation, and potential deletion-bearing clones were selected by growth in FIAU, which kills tk-expressing cells. DNAs isolated from FIAU-resistant clones were used as templates for PCR amplification using polymorphic microsatellite markers distributed throughout the t complex. The targeted insertions in both the Sod2 and D17Leh94 lines occurred in the 129S4SvJae (129) alleles, so those containing interstitial deletions were identified by loss of 129 SSLP alleles (see MATERIALS AND METHODS).

Deletion-bearing ES cell lines were injected into C57BL/6J (B6) blastocysts (in one case, FVB blastocysts were used) to generate chimeric mice. The sizes of deletions contained in these clones are depicted in Fig 1. Initially, the chimeras were bred to B6 females, and the presence of agouti offspring indicated germline transmission of the 129 agouti allele from ES cell-derived gametes.

Seven ES cell lines bearing deletions induced at the Sod2 locus were used to make chimeras (Fig 1). The minimal sizes of these deletions ranged from 8 Mb (Sod2^df6J) to 14 Mb (Sod2^df4J) and extended collectively from a point centromeric to D17Mit164, the most proximal marker available in the t complex, to an interval 1–2 Mb distal to D17Leh94 (15–16 Mb). All but one of the deletions (Sod2^df6J) remove the brachyury (T) locus. Since animals that are heterozygous for a null mutation (or deletion) of T have a short or kinked tail, this serves as a convenient visible marker for tracking the deletions. Each of the deletion-bearing ES cell lines was capable of generating germline chimeras, except for Sod2B10.

Strain-dependent failure to transmit certain deletions:
Of the six deletion ES cell lines that did produce germline chimeras, five exhibited a marked failure to transmit the deletion-bearing chromosome when the chimeras were mated to B6 females (Table 1). For the Sod2^df3J line, six male chimeras were derived that produced agouti offspring in crosses with B6 females (and thus were deemed to be germline chimeras), but none of 112 offspring exhibited brachyury offspring, indicating that none inherited the deletion. The absence of deletion transmission was verified by genotyping all progeny. Similar results were obtained from germline chimeras produced from the Sod2^df1J, Sod2^df2J, Sod2^df4J, and Sod2^df6J deletion-bearing ES cell lines. In all, the germline chimeras derived from these five cell lines sired 326 progeny in matings to B6 females, but no deletion-bearing offspring were observed (Table 1). In contrast, the Sod2^df5J deletion was transmitted by germline chimeras to 63 of 171 progeny (36%) when mated with B6 females. Because the chimeras were derived from 129 x B6 ES cells injected into B6 blastocysts, some of the progeny could have been derived from the B6 host. Maximally, 50% of progeny would be agouti if there were 100% transmission of ES-derived sperm from these chimeras. As indicated in Table 1, the proportion of agouti offspring from chimeras with the various deletions ranged from 11 to 52%, indicating that most progeny, with the exception of Sod2^df3J, were derived from the injected ES cells. Notably, the same chimeras of lines Sod2^df1J, Sod2^df2J, and Sod2^df6J that failed to pass the deletion in crosses to B6 mates were capable of transmitting the deletion when mated to mice of other strains (Table 1 and see below).

In considering the region(s) of proximal 17 that might be responsible for the failure to transmit the deletion to offspring of matings to B6, an examination of the deletion breakpoints implicates a locus or loci distal to D17Mit131. The region proximal to D17Mit156 seems not to be responsible, since Sod2^df5J, which does transmit the deletion to B6, extends much more centromerically than does Sod2^df6J, which does not transmit to offspring by B6. At the present level of resolution, the locations of the Sod2^df2J-5J proximal breakpoints are indistinguishable. Most notably, however, all of the deletions that were not transmissible to B6 extend more distally than Sod2^df5J. These combined data implicate a locus or loci between D17Mit131 and D17Jcs8 (Fig 1). The Celera mouse genome sequence assembly spans much of this region. Using this sequence (which exists in multiple large contigs with a few gaps between and within them) in conjunction with the mouse bacterial artificial chromosome (BAC) fingerprint database (Vancouver Genome Sequencing Center) and the The Institute for Genomic Research BAC end sequence database, the interval was determined to span 4000 kb.

Two deletion-bearing ES cell lines generated at the D17Leh94 locus, D17Leh94^df4J and D17Leh94^df10J, were injected into blastocysts to generate germline chimeric mice. The distal breakpoints are currently indistinguishable, lying between D17Jcs37 and D17Jcs1. D17Leh94^df4J extends further proximally to the interval between Sod2 and D17Mit131.

In matings of D17Leh94^df10J/+ mice to B6, the deletion was transmitted to 43.5% of progeny (27/62). In contrast, three germline male chimeras derived from the D17Leh94^df4J clone transmitted the deletion to only 2 of 122 progeny in matings with B6 females. Of the two N2 animals that did carry the deletion, one was a female that died prior to reaching reproductive maturity; the remaining male was mated to a C3H/HeOuJ female. Interestingly, this animal transmitted the deletion to roughly half its progeny (J. FOREJT, personal communication). Seven additional germline chimeras were then generated and mated to both B6 and C3H/HeOuJ. Whereas only 4/211 (1.9%) progeny inherited the deletion in matings to B6 females, the transmission rate rose to 39% (30 out of 76 animals typed) in matings of these same chimeras to C3H/HeOuJ females. These data are indicative of a strain-dependent effect on viability of deletion-bearing embryos or in fertilizing ability of deletion-bearing sperm.

Since the D17Leh94^df4J deletion overlaps with all five Sod2 deletions that failed to transmit in matings to B6 (all delete D17Mit213 and D17Jcs35), the possibility was considered that the phenomenon was due to haploinsufficiency of a common region. If so, it might be expected that the Sod2 deletions would exhibit the same strain effects on transmission. To test this, the Sod2^df1J, Sod2^df2J, and Sod2^df6J germline chimeras that failed to transmit the deletion to any progeny in matings to B6 were crossed to C3H/HeOuJ females. Numerous deletion-bearing offspring were obtained, as determined by DNA typing (Table 1). Deletion transmission rates for Sod2^df1J, Sod2^df2J, and Sod2^df6J were 16% (13 of 80), 13% (20 of 149), and 68% (11 of 16), respectively. Thus, the phenotype of complete haplolethality caused by matings to B6 females could be partially rescued by using C3H/HeOuJ dams.

To test whether the elevated transmission to C3H/HeOuJ or the low transmission to B6 was unique to either strain, D17Leh94^df4J/+ males were mated with females of various other strains. The results are shown in Table 2. For each strain, inheritance of the deletion was observed. Transmission in crosses to AKR/J, C57BLKS/J, DBA/2J, B6D2F1/J, and B6C3F1/J was between 2 and 10% after examination of an average of 50 animals per cross. Transmission of the D17Leh94^df4J in crosses to C3D2F1/J, C3FeB6F1/J, and C3HeB/FeJ was higher, with a 29% transmission rate in C3FeB6F1/J (based on examination of 107 progeny) and a 20% rate for C3D2F1/J and C3HeB/FeJ (30 and 55 animals examined, respectively). The data suggest that, overall, transmission of the deletion chromosome to live offspring is significantly impaired, irrespective of genetic background. However, the B6 background is especially incompatible with transmission, and in most cases, this incompatibility is nearly absolute.

Deletion-bearing progeny from matings to B6 die from severe exencephaly or edema:
To determine why few deletion-bearing progeny from crosses with B6 were observed, the possibility that deletions were causing embryonic lethality was investigated. Timed matings were conducted between B6 or C3H females and male chimeras carrying the Sod2^df1J, Sod2^df2J, and Sod2^df5J deletions. Pregnant females were dissected at day E12.5, 13.5, or 14.5. The Sod2 deletions all remove T, providing a convenient marker (short tails) for deletion-containing embryos. Additionally, all embryos/animals were typed with SSLP markers to unambiguously assign their genotypes.

Dissections of B6 females plugged by either Sod2^df1J or Sod2^df2J chimeras revealed that the deletion-bearing progeny from these crosses suffer from severe exencephaly or edema (Fig 2). For example, of 10 embryos carrying the Sod2^df2J deletion, 7 had exencephaly or edema; of four Sod2^df1J deletion-bearing animals, all had exencephaly or edema. Animals with edema were likely resorbed in utero prior to parturition. However, when timed pregnancies were allowed to proceed to birth, some exencephalic, short-tailed liveborn offspring were observed (not shown). Generally, the dams immediately ate such newborns. None of the fetuses produced in matings between B6 females and Sod2^df5J chimeras demonstrated exencephaly or edema (none of seven fetuses carrying the deletion), consistent with the normal transmission of this deletion to weaned offspring as described earlier.

View larger version (73K): In this window In a new window Download PPT slide   Figure 2. Haplolethal phenotype of embryos. (A) A normal late-gestation fetus. (B) Exencephalic Sod2df2J/+ fetus. (C) Edematous phenotype of Sod2df2J/+ fetus.

When Sod2^df1J and Sod2^df2J chimeras were mated with C3H females, deletion-bearing fetuses were more likely to be normal (brachyury notwithstanding) than with matings to B6 (see Table 3). Combining the results, whereas 11/14 embryos produced from matings to B6 were abnormal, only 4/12 were affected in matings to C3H (P = 0.05). Animals carrying either of two different deletions centered at the D17Aus9 locus in the proximal t complex were used as controls for possible generalized defects associated with deletions (Fig 1). In matings to B6, 90% (9 of 10) of the embryos carrying D17Aus9^df26J or D17Aus9^df27J were normal, as were all 3 embryos produced in matings to C3H. The fact that one exencephalic embryo was observed might be related to an observation that the T^hp deletion, which extends more distally than D17Aus9^df26J and D17Aus9^df27J, has been found to cause exencephalic embryos in the C3H background (ROGERS et al. 1997 ). Anyway, since D17Aus9^df26J and D17Aus9^df27J have be rendered congenic on the C57BL/6J background (data not shown), the underlying basis for the rare exencephalic embryo is clearly different from the B6-dependent phenomenon observed with the haplolethal deletions presented here.

Females cannot transmit the D17Leh94^df4J deletion:
To test whether there were sex differences in transmission rates of the D17Leh94 deletions, deletion-bearing females having >50% C3H genetic background were backcrossed to B6 males, and the progeny were genotyped to identify the proportion inheriting the deletion. The Sod2 deletions were not tested, since they remove the imprinted T maternal effect (Tme) locus near Sod2, thus preventing transmission from females. The smaller deletion, D17Leh94^df10J, was transmitted equally well by males and females (data not shown). However, females bearing the D17Leh94^df4J deletion at the N2 or N3 generation into C3H failed to transmit the deletion to their offspring in matings to B6 males: of 66 progeny from a total of 16 litters, none carried the deletion. This is in contrast to a similar set of matings in which D17Leh94^df4J/+ male siblings at the N2 or N3 generation into C3H exhibited a deletion transmission rate of 20% (28/142 progeny from 27 litters). When D17Leh94^df4J/+ females (N2 or N3 generation into C3H) were crossed to C3H males, no deletion-containing progeny were obtained (14 animals in a total of 4 litters). Additionally, D17Leh94^df4J/+ females with a mixed 129/C3H/B6 background were also unable to transmit the deletion, as judged by examination of the 23 offspring from 4 different litters. Finally, one deletion-bearing D17Leh94^df4J/+ female was mated with a CAST/Ei male; this outcross often acted to boost deletion transmission rates of other deletions (data not shown). However, only a single litter of five animals was obtained from this cross, and none carried the deletion. It is worth noting that deletion-bearing females were difficult to obtain, and it was not uncommon for them to die prior to reaching reproductive maturity (data not shown). In summary, transmission of D17Leh94^df4J from a female was never observed (0/108 offspring), raising the possibility that a paternally imprinted locus lies within the deletion region.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Here we describe the generation of two panels of deletions made within the central t complex on mouse chromosome 17. Analysis of mice bearing these deletions indicated that much of the proximal t complex does not contain haplolethal genes, as evidenced by the health and high transmission rates of the D17Leh94^df10J and Sod2^df5J deletions, which collectively span all of the proximal t complex (15 Mb) except for the interval between their respective proximal and distal breakpoints. However, a relatively small region could be defined that, when hemizygous, caused severe exencephaly or edema when the C57BL/6J contribution to genetic background was >50%.

The region containing a haplolethal locus or loci can be defined by correlating presence or absence of the phenotype (failure to transmit when chimeras are mated to B6) with the breakpoints of the deletions. Considering the Sod2 deletions alone, this region must lie distal to D17Mit131 (which is removed by Sod2^df5J, a deletion that does not cause the phenotype) and proximal to D17Jcs8, a locus that is not missing in two deficiencies (Sod2^df1J,2J) that do show the phenotype. The interval can be further narrowed by taking into account the fact that D17Leh94^df10J removes D17Jcs80 and D17Jcs8 yet does not exhibit haplolethality. The combined data place the haplolethal locus/loci between D17Mit131 and D17Jcs80, which spans 3 Mb (Fig 1).

In humans, large chromosome deletions commonly cause mental retardation, growth reduction, and dysmorphologies. In mice, large deletions also tend to reduce fitness and can cause certain defects in the heterozygous condition (CATTANACH et al. 1993 ). A recent attempt to correlate human malformations with deletions of particular regions of the genome revealed that a large proportion of the genome (slightly greater than half) is susceptible to visible phenotypes in association with segmental aneuploidy (BREWER et al. 1998 ). For the most part, however, such knowledge is biased in that the information is derived exclusively from viable patients or mice. Information on regions of the genome that are haplolethal is poor, because the probands (or unborn mice) are less observable.

The ability to manipulate the mouse genome in a targeted fashion allows the direct assessment of genes or specific chromosomal regions for haplolethality. Substantial numbers of genes, when mutated, cause heterozygous phenotypes. These were identified by virtue of spontaneous semidominant mutations (such as T and Kit^W) and mutations induced by gene targeting, including those of Twist, Tbx1, Pax1, and Bmp4 (DUNN et al. 1997 ; BOURGEOIS et al. 1998 ; WILM et al. 1998 ; PAINE-SAUNDERS et al. 2000 ; LINDSAY et al. 2001 ). Published reports of haplolethal null alleles identified by gene targeting are very rare and, to our knowledge, are limited to Vegf (CARMELIET et al. 1996 ; FERRARA et al. 1996 ) and Tcof1, the Treacher Collins Franceshetti syndrome 1 homolog (DIXON et al. 2000 ). Deficiency kits in Drosophila cover a vast majority of the genome, although segmental haploidy is lethal in only a few regions; in one case, it has been determined that a single gene (decapentaplegic) is responsible for the observed haplolethality (PRADO et al. 1999 ). This suggests that expression of a relatively small subset of genes is so precisely regulated as to cause severe developmental defects when dosage is reduced by 50%. Such genes probably play crucial roles in either cellular physiology or in patterning of crucial embryonic structures.

The sparseness of haplolethal genes emphasizes the utility of chromosomal deletions for scanning the genome to uncover and localize such genes. Whereas large deletions may cause decreased viability as a nonspecific additive consequence of reduced dosage of many genes, the discrete phenotypes linked to hemizygosity of DNA in the D17Mit131-D17Jcs80 interval, but not in other regions of the t complex examined, suggest that one gene may be responsible. Another possibility is that the phenotype may be due to deletion of multiple, closely linked genes in the critical region that may be related (by duplication) or function in the same pathway. Interestingly, the Celera Discovery System lists several genes in this interval related to the Kruppel C2H2-type zinc-finger family. Members of this large family of transcription factors play numerous developmental roles (BIEKER 2001 ). Some other genes in this region (according to Celera) include four ribosomal protein genes, Thbs2 (thrombospondin 2), and a putative insulin-like growth-factor-binding protein.

In addition to a gene(s) required for correct neural tube development (the hemizygosity of which causes exencephaly), evidence was obtained for the presence of an imprinted gene(s) within the D17Leh94^df4J deletion interval that is not deleted by D17Leh94^df10J, between Sod2 and D17Jcs80. This interval does not incude the imprinted Tme (Igf2r) locus, which is 200 kb proximal to Sod2. Heterozygous females never transmitted D17Leh94^df4J to any live-born offspring, even in cases where the females had >75% of their genome derived from C3H. However, males of the same background were capable of transmitting the deletion (albeit to only 20%, due presumably to the haplolethal gene or genes). This is consistent with the existence of a paternally imprinted (silenced) gene that must be inherited maternally for offspring to be viable. Because of the difficulty in obtaining substantial numbers of fecund deletion-bearing females, timed matings were not performed for the purpose of determining the phenotype of embryos carrying deletions. It is possible that this potential imprinted gene(s) is identical to those that cause the haplolethal phenotype in matings of male chimeras to B6 females. Further studies will be required to determine if this is the case. While the genomic imprinting maps maintained by Beechy, Cattanach and Blake (http://www.mgu.har.mrc.ac.uk/imprinting/imprinting.html) identify an imprinting region on proximal 17, the proximity of the paternally imprinted Tme to this new potentially imprinted gene is too close to have been distinguished.

Modern technologies for generating deletions throughout the genome will be valuable for mapping genes that cause aberrant phenotypes when present in only one copy. Such genes may play key roles in development, possibly encoding molecules such as morphogens or transcription factors that may be required at precise levels for particular processes of differentiation, patterning, or cell migration. The work presented here represents the first step in a project to identify such genes within the t complex, which is known to contain several recessive null mutations that disrupt embryonic development. Since the locus or loci required for proper neural tube closure are flanked by two sets of deletions with staggered breakpoints, the next step in this project will be positional cloning of the underlying gene(s) involved in this phenotype. Higher resolution analysis of key deletion breakpoints may substantially narrow the critical region such that BAC rescue would be practical. Since the deletions here are preserved in ES cells, the rescue experiments can be conducted in one step: a deletion-bearing clone can be transfected with a series of BACs, and the chimeras can be evaluated immediately for ability to transmit the deletion in crosses to B6. Otherwise, additional deletions can be selected in ES cells that have breakpoints within the critical region to refine further the locations of the haplolethal and imprinted loci and to determine if they are one and the same and if single or multiple genes underlie the observed phenomena.

	FOOTNOTES

¹ Present address: Department of Biochemistry, University of Wisconsin, Madison, WI 53706.

	ACKNOWLEDGMENTS

The authors thank Tom Gridley and Wayne Frankel for critical comments on the manuscript. This work was supported by a National Institutes of Health (NIH) grant to J.C.S. (HD-24374), a Cancer Center Grant (CA34196) to the Jackson Laboratory, and an NIH fellowship (HD-08441) to V.L.B.

Manuscript received August 24, 2001; Accepted for publication November 19, 2001.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

AMOS-LANDGRAF, J. M., Y. JI, W. GOTTLIEB, T. DEPINET, and A. E. WANDSTRAT et al., 1999  Chromosome breakage in the Prader-Willi and Angelman syndromes involves recombination between large, transcribed repeats at proximal and distal breakpoints. Am. J. Hum. Genet. 65:370-386[Medline].

BIEKER, J., 2001  Kruppel-like factors: three fingers in many pies. J. Biol. Chem. 276:34355-34358[Free Full Text].

BOURGEOIS, P., A. L. BOLCATO-BELLEMIN, J. M. DANSE, A. BLOCH-ZUPAN, and K. YOSHIBA et al., 1998  The variable expressivity and incomplete penetrance of the twist-null heterozygous mouse phenotype resemble those of human Saethre-Chotzen syndrome. Hum. Mol. Genet. 7:945-957[Abstract/Free Full Text].

BREWER, C., S. HOLLOWAY, P. ZAWALNYSKI, A. SCHINZEL, and D. FITZPATRICK, 1998  A chromosomal deletion map of human malformations. Am. J. Hum. Genet. 63:1153-1159[Medline].

BREWER, C., S. HOLLOWAY, P. ZAWALNYSKI, A. SCHINZEL, and D. FITZPATRICK, 1999  A chromosomal duplication map of malformations: regions of suspected haplo- and triplolethality—and tolerance of segmental aneuploidy—in humans. Am. J. Hum. Genet. 64:1702-1708[Medline].

CARMELIET, P., V. FERREIRA, G. BREIER, S. POLLEFEYT, and L. KIECKENS et al., 1996  Abnormal blood vessel development and lethality in embryos lacking a single VEGF allele. Nature 380:435-439[Medline].

CATTANACH, B. M., M. D. BURTENSHAW, C. RASBERRY, and E. P. EVANS, 1993  Large deletions and other gross forms of chromosome imbalance compatible with viability and fertility in the mouse. Nat. Genet. 3:56-61[Medline].

DIXON, J., C. BRAKEBUSCH, R. FASSLER, and M. J. DIXON, 2000  Increased levels of apoptosis in the prefusion neural folds underlie the craniofacial disorder, Treacher Collins syndrome. Hum. Mol. Genet. 9:1473-1480[Abstract/Free Full Text].

DUNN, N. R., G. E. WINNIER, L. K. HARGETT, J. J. SCHRICK, and A. B. FOGO et al., 1997  Haploinsufficient phenotypes in Bmp4 heterozygous null mice and modification by mutations in Gli3 and Alx4. Dev. Biol. 188:235-247[Medline].

DUTLY, F. and A. SCHINZEL, 1996  Unequal interchromosomal rearrangements may result in elastin gene deletions causing the Williams-Beuren syndrome. Hum. Mol. Genet. 5:1893-1898[Abstract/Free Full Text].

FERRARA, N., K. CARVER-MOORE, H. CHEN, M. DOWD, and L. LU et al., 1996  Heterozygous embryonic lethality induced by targeted inactivation of the VEGF gene. Nature 380:439-442[Medline].

GABRIEL, J. M., M. MERCHANT, T. OHTA, Y. JI, and R. G. CALDWELL et al., 1999  A transgene insertion creating a heritable chromosome deletion mouse model of Prader-Willi and Angelman syndromes. Proc. Natl. Acad. Sci. USA 96:9258-9263[Abstract/Free Full Text].

LI, L., I. D. KRANTZ, Y. DENG, A. GENIN, and A. B. BANTA et al., 1997  Alagille syndrome is caused by mutations in human Jagged1, which encodes a ligand for Notch1. Nat. Genet. 16:243-251[Medline].

LINDSAY, E. A., A. BOTTA, V. JURECIC, S. CARATTINI-RIVERA, and Y. C. CHEAH et al., 1999  Congenital heart disease in mice deficient for the DiGeorge syndrome region. Nature 401:379-383[Medline].

LINDSAY, E. A., F. VITELLI, H. SU, M. MORISHIMA, and T. HUYNH et al., 2001  Tbx1 haploinsufficiency in the DiGeorge syndrome region causes aortic arch defects in mice. Nature 410:97-101[Medline].

MERSCHER, S., B. FUNKE, J. A. EPSTEIN, J. HEYER, and A. PUECH et al., 2001  TBX1 is responsible for cardiovascular defects in velo-cardio-facial/DiGeorge syndrome. Cell 104:619-629[Medline].

NAF, D., L. A. WILSON, R. A. BERGSTROM, R. S. SMITH, and N. C. GOODWIN et al., 2001  Mouse models for the Wolf-Hirschhorn deletion syndrome. Hum. Mol. Genet. 10:91-98[Abstract/Free Full Text].

ODA, T., A. G. ELKAHLOUN, B. L. PIKE, K. OKAJIMA, and I. D. KRANTZ et al., 1997  Mutations in the human Jagged1 gene are responsible for Alagille syndrome. Nat. Genet. 16:235-242[Medline].

PAINE-SAUNDERS, S., B. L. VIVIANO, J. ZUPICICH, W. C. SKARNES, and S. SAUNDERS, 2000  glypican-3 controls cellular responses to Bmp4 in limb patterning and skeletal development. Dev. Biol. 225:179-187[Medline].

PRADO, A., I. CANAL, and A. FERRUS, 1999  The haplolethal region at the 16F gene cluster of Drosophila melanogaster: structure and function. Genetics 151:163-175[Abstract/Free Full Text].

ROGERS, I., K. OKANO, and S. VARMUZA, 1997  Paternal transmission of the mouse Thp mutation is lethal in some genetic backgrounds. Dev. Genet. 20:23-28[Medline].

ROHME, D., H. FOX, B. HERRMAN, A.-M. FRISCHAUF, and J. E. EDSTROM et al., 1984  Molecular clones of the mouse t complex derived from microdissected metaphase chromosomes. Cell 36:783-788[Medline].

SILVER, L. M., 1985  Mouse t haplotypes. Annu. Rev. Genet. 19:179-208[Medline].

TSAI, T. F., Y. H. JIANG, J. BRESSLER, D. ARMSTRONG, and A. L. BEAUDET, 1999  Paternal deletion from Snrpn to Ube3a in the mouse causes hypotonia, growth retardation and partial lethality and provides evidence for a gene contributing to Prader-Willi syndrome. Hum. Mol. Genet. 8:1357-1364[Abstract/Free Full Text].

URBAN, Z., C. HELMS, G. FEKETE, K. CSISZAR, and D. BONNET et al., 1996  7q11.23 deletions in Williams syndrome arise as a consequence of unequal meiotic crossover. Am. J. Hum. Genet. 59:958-962[Medline].

WILM, B., E. DAHL, H. PETERS, R. BALLING, and K. IMAI, 1998  Targeted disruption of Pax1 defines its null phenotype and proves haploinsufficiency. Proc. Natl. Acad. Sci. USA 95:8692-8697[Abstract/Free Full Text].

YANG, T., T. E. ADAMSON, J. L. RESNICK, S. LEFF, and R. WEVRICK et al., 1998  A mouse model for Prader-Willi syndrome imprinting-centre mutations. Nat. Genet. 19:25-31[Medline].

YOU, Y., R. BERGSTROM, M. KLEMM, B. LEDERMAN, and H. NELSON et al., 1997a  Chromosomal deletion complexes in mice by radiation of embryonic stem cells. Nat. Genet. 15:285-288[Medline].

YOU, Y., V. L. BROWNING, and J. C. SCHIMENTI, 1997b  Generation of radiation-induced deletion complexes in the mouse genome using embryonic stem cells. Methods 13:409-421[Medline].

YOU, Y., R. BERSGTRAM, M. KLEMM, H. NELSON, and R. JAENISCH et al., 1998  Utility of C57BL/6J x 129/SvJae embryonic stem cells for generating chromosomal deletions: tolerance to gamma radiation and microsatellite polymorphism. Mamm. Genome 9:232-234[Medline].