A wide spectrum of birth defects is caused by deletions of the DiGeorge syndrome chromosomal region at 22q11. Characteristic features include cranio-facial, cardiac and thymic malformations, which are thought to arise from disturbances in the interactions between hindbrain neural crest cells and the endoderm of the pharyngeal pouches. Several genes have been identified in the shortest region of deletion overlap at 22q11, but nothing is known about the expression of these genes in mammalian embryos. We report here the isolation of several murine embryonic cDNAs of the DiGeorge syndrome candidate gene HIRA. We identified several alternatively spliced transcripts. Sequence analysis reveals that Hira bears homology to the p60 subunit of the human Chromatin Assembly Factor I and yeast Hir1p and Hir2p, suggesting that Hira might have some role in chromatin assembly and/or histone regulation. Whole mount in situ hybridization of mouse embryos at various stages of development show that Hira is ubiquitously expressed. However, higher levels of transcripts are detected in the cranial neural folds, frontonasal mass, first two pharyngeal arches, circumpharyngeal neural crest and the limb buds. Since many of the structures affected in DiGeorge syndrome derive from these Hira expressing cell populations we propose that haploinsufficiency of HIRA contributes to at least some of the features of the DiGeorge phenotype.
Hemizygosity of a region of chromosome 22q11 has been associated with a wide variety of congenital malformations that receive several diagnostic labels, the main features being covered by the acronym CATCH22 [Cardiac defect, Abnormal facies, Thymic hypoplasia, Cleft palate, Hypocalcemia, (interstitial) 22q11 deletions] (1 ). The congenital malformation syndromes associated with these deletions include DiGeorge syndrome (DGS), velo-cardio-facial syndrome (VCFS), conotruncal anomaly face syndrome and Opitz syndrome (2 -8 ). For the latter syndrome Robin et al. recently reported linkage with loci on chromosome 22q11 and Xp22 (9 ).
Several well characterized DNA probes can detect CATCH22 deletions (3 ,4 ,10 -15 ). The interstitial deletions are large, spanning 2-3 Mb of DNA. A limited number of CATCH22 patients have been reported with an interstitial deletion smaller than the commonly deleted region (16 -18 ). The current shortest region of deletion overlap (SRDO1) is defined between the centromeric breakpoint in DGS patient G (16 ) and the unbalanced translocation breakpoint in cell line GM00980 [from a VCFS patient with a t(11;22)(q25;q11) translocation (19 )]. Recently, Kurahashi et al. provided evidence for the existence of a second critical region (SRDO2) located in the distal part of the commonly deleted region (17 ). The two SRDOs are mutually exclusive.
At present it is not clear whether the CATCH22 phenotype is the result of haploinsufficiency of one or more genes. The balanced t(2;22)(q14.1;q11.1) translocation breakpoint in DGS patient ADU (20 ) was the major target of positional cloning in various laboratories. These attempts led to the isolation of DGCR3 and DGCR5. DGCR3 is an open reading frame (ORF) of 260 amino acids (aa) that is interrupted by the translocation breakpoint (21 ). However, no cDNAs were isolated and expression studies were only informative under low stringency conditions. This prompted Sutherland et al. (22 ) to screen for adjacent coding sequences in the region. They identified a series of alternatively spliced transcripts from the DGCR5 gene. The DGCR3 sequence is embedded in an intron of DGCR5 and in the same transcriptional orientation. The DGCR5 transcripts do not have an obvious protein encoding potential. They may be functional RNAs. The recent refinement of the SRDO places the ADU breakpoint ~100 kb centromeric to the proximal deletion boundary of SRDO1 (16 ), so the relevance of DGCR3 and DGCR5 for CATCH22 etiology is unclear. It could very well be that the CATCH22 phenotype results from haploinsufficiency of several relevant genes in SRDO1 and that the ADU breakpoint exerts a position effect on such genes.
The saturation cloning and sequencing of the 22q11 region are at an advanced stage (23 ,24 ). Two transcription maps of SRDO1 have recently been reported (24 ,25 ). Currently at least six transcription units have been identified: HIRA (DGCR1/ TUPLE1) (26 ,27 ), CLH-22 (clathrin heavy chain like, CLTCL/ES3/DGS-K) (24 ,25 ,28 ,29 ), CTP (mitochondrial citrate transporter protein, DGS-J) (24 ,30 ), DGS-I/ES2 (24 ,25 ), DGS-H (24 ) and DGS-G (a serine/threonine kinase) (24 ). Sequences homologous to the 3' UTR of the human Dishevelled 1 gene were mapped just centromeric to the HIRA gene (31 ). It is unclear whether these sequences belong to a gene. Little is known about the expression of the CATCH22 candidate genes in adult and embryonic tissues. It will be important to establish the embryonic expression patterns and functions of the candidate genes.
We examined the evolutionary conservation of HIRA by hybridization of radiolabeled probe hD (from the 4 kb insert of one of our human HIRA cDNA clones) (Fig. 1 B) to a blot containing EcoRI digested genomic DNA from various species. The probe hybridizes to homologous sequences in the DNA of all mentioned organisms (Fig. 2 ). This high level of conservation from man to yeast indicates an essential function for HIRA.
As an initial step to study whether haploinsufficiency of the HIRA gene contributes to the CATCH22 phenotype, we hybridized probe hE (Fig. 1 B) to an E11.5 mouse embryo cDNA library under stringent conditions. This screening yielded seven positive cDNA clones of the mouse homologue of the human HIRA gene (Fig. 1 A). Complete sequencing and assembly of selected clones allowed us to reconstitute a cDNA of 3852 bp (accession number EMBL X99712), containing an ORF of 3045 bp. The 1015 aa predicted protein has seven WD40 repeats (at aa positions 10-44, 68-98, 129-159, 172-202, 234-259, 266-313 and 326-356), two bipartite nuclear localization signals (at aa 267-286 and 626-643), a penta-Q stretch (aa 408-412) and a molecular mass of 111.72 kDa.
Sequence comparison between the human and mouse proteins reveals an overall identity of 95.8% and a similarity of 97.2%. The N-terminal third of the protein, containing the WD40 repeats, is 99.4% conserved (identity and similarity). The difference is in one amino acid in WD40 repeat 3 and one in repeat 6. The C-terminal two-thirds have a relatively higher divergence (identity 93.7%; similarity 96%). Compared to the 1017 aa human HIRA translated sequence, the Hira sequence lacks amino acids 510 and 607 or 608. Our Hira sequence is 99% identical to the recently reported mouse Hira sequence (35 ) and 100% identical to the mouse Tuple1 cDNA clone MF2 (26 ).
In addition to homology with yeast Hir1p and Hir2p, BlastP and BEAUTY database searches with our mouse Hira sequence reveal high homology to the p60 subunit of human Chromatin Assembly Factor I (CAF1A) (34 ) and a Caenorhabditis elegans protein (36 ) which, in the absence of a name, we will henceforward refer to as C.elegans HIRA-like protein (ceHILP). The overall homology with ceHILP is higher than with Hir1p (P = 6.1 × 10-77 versus 7.6 × 10-45). We aligned the murine Hira protein sequence with ceHILP, Hir1p, Hir2p and CAF1A (Fig. 3 ). The overall similarities in this alignment are 39% with ceHILP, 38% with CAF1A, 35% with Hir1p and 31% with Hir2p. It can be readily seen that there are at least three different conserved (functional?) regions: the highest homologies are found in the N-terminal part, region A (aa 1-371) (average similarity 47%); in the middle part, region B (aa 473-658), homology drops (average similarity 18%); C-terminal region C (aa 744-1015) shows again an increased homology (average similarity 35%). Sequence stretches with many gaps in the alignment were excluded from these regions [comparable with the hinge region suggested by Scamps et al. (35 )]. In the absence of experimental evidence for boundaries of functional domains exon-exon boundaries were chosen to provisionally delimit the regions. The C-terminal boundary of region A corresponds with the insertion point for exon 11A. In our alignment the C-termini of Hira, ceHILP and Hir2p almost coincide. CAF1A terminates at the end of region B and Hir1p terminates in region C.
The presence of Hira mRNA in adult mice and embryos was examined by Northern blot analysis of poly(A)+ RNA extracted from embryos at different stages of development and from various adult organs. Probes 12E and mX (Fig. 1 A and B) detect transcripts of 4.4 and 2.3 kb (Fig. 4 A). The 4.4 kb transcript is more abundant in most organs while the 2.3 kb transcript is more abundant in the adult heart.
The expression of Hira in the brain and the heart and its expression during organogenesis, raised the possibility that Hira plays a role in the development of the brain and the pharyngeal region of the head. Therefore we determined the spatial distribution of Hira mRNAduring mammalian embryogenesis. Whole mount in situ hybridization was performed on mouse embryos of 8.5, 9.5 and 10.5 days of gestation, a time-window during which the most critical events in organogenesis are taking place. For these experiments we used single stranded RNA probe mA (Fig. 1 B), that only detects the 4.4 kb transcript.
At E8.5 base levels of Hira transcripts are present throughout the embryo, including the heart (Fig. 6 A and B). Higher levels of transcripts are detected in the cranial neural folds (Fig. 6 A). At E9.5 Hira transcripts are also widely present (Fig. 6 D). Higher levels of expression are seen in pharyngeal arches 1 and 2. We do not detect Hira transcripts in the heart anymore at this stage. Sections of gelatin embedded whole mount embryos reveal Hira transcripts in the mesenchyme of the pharyngeal arches (not shown). At E10.5 the expression pattern of Hira is similar to that at E9.5 (Fig. 6 E). The highest levels of transcripts are in the anterior part (including the maxillary component) of arch 1 and the posterior part of arch 2, the frontonasal mass and the limb buds. Just posterior to arches 1 and 2 a circumscribed spot with a high level of Hira transcripts is present. This region corresponds well to the region where hindbrain neural crest cells reside before entering the developing gut and the third and fourth pharyngeal arch. In chicken embryos this region is known as the circumpharyngeal neural crest (37 ).
Two out of the seven Hira cDNAs (clones 12B and 12E) contain non-canonical sequences (Fig. 1 A). To investigate whether these cDNAs represent valid transcripts, we partially characterized the genomic organization of the murine Hira gene. The intron-exon boundaries we determined coincide with those in the human as recently published by Lorain et al. (38 ). Below we will use their numbering of the exons.
Clone 12B contains a 203 bp insert between exons 3 and 4, which encode part of the first and second WD40 domain respectively (sequence in Fig. 7 A). Genomically, the insert is continuous with exon 3 but is separated from exon 4 by an intron. Therefore, clone 12B must have arisen from the use of an alternative splice donor site. The ORF of clone 12B is interrupted by a frameshift, both in cDNA and genomic clones. The next potential translation initiation site is the methionine codon at nucleotide position 510, with a Kozak score of 0.67, or the methionine codon at nucleotide position 546 with a Kozak score of 0.78.
We identified three regions in Hira through alignment with ceHILP, Hir1p, Hir2p and CAF1A. In addition, mouse/human HIRA alignments and the intron/exon structure of the ceHILP gene further support the presence of these distinct regions within Hira: mouse/human HIRA comparison showed higher homology in the N-terminal third (region A) than in the remainder of the protein; ceHILP exon-exon boundaries are spatially remarkably close to those of Hira. The major part of regions B and C are coded for by large single exons in ceHILP (39 ,40 ), suggesting that these regions represent single functional domains.
Region A contains WD40 domains that are present in a wide range of different proteins (41 ) and are involved in protein-protein interactions (42 ). In the case of yeast Tup1p, after which Hira was originally named, WD40 domains 1 and 2 interact directly with the homeodomain protein [alpha]2p (43 ).
The homology in regions B and C points to conserved domains although the functions have yet to be demonstrated. A clue to their functions might again come from Tup1p. Recently, Edmonson et al. reported that Tup1p binds histones H3 and H4 through a domain outside the WD40 region (44 ). Strikingly, the CAF complex also binds histones H3 and (acetylated) H4 and is thought to chaperone these core histones to the newly replicated DNA (34 ). In view of the homology of Hira with CAF1A it might well be that Hira performs an analogous function. Moreover, CAF1A function might be sensitive to gene dosage. CAF1A has been mapped to the Down syndrome critical region at 21q22 (45 ). The presence of three copies of CAF1A in Down syndrome individuals might contribute to the phenotype. CAF1A works in a complex with at least one other subunit (p150). Changes in gene-dosage might perturb the stoichiometry of the complex, hampering its formation. Hira, like CAF1A and Hir1p and Hir2p, is expected to be functional in a complex. Decreased availability of Hira due to haploinsufficiency of its gene may well have an effect on the formation of such a complex.
In this report we have shown that the Hira gene is expressed from as early as day E7 (gastrulation stage) through adulthood. Whole mount in situ hybridization shows that expression is ubiquitous, but that higher levels of transcripts are present in the cranial neural folds, subregions of pharyngeal arches 1 and 2, the circumpharyngeal neural crest and the limb buds. It is of interest that Hira transcripts are present in neural folds, where neural crest cells will be generated. In chicken embryos the zinc finger gene Slug is also expressed in the cranial folds. Treatment of embryos with Slug antisense oligos showed that Slug is necessary for neural crest cell migration (46 ). In mice, the homeobox gene Msx-3 shows highly restricted expression that partially overlaps with Hira expression (47 ).
Most of the mesenchymal cells in the pharyngeal arches derive from hindbrain neural crest cells. It thus seems likely that the Hira expressing cells in pharyngeal arches 1 and 2 represent hindbrain neural crest cells. Special attention needs to be drawn to the regional differences within the arch mesenchyme. Certain subpopulations of cells either in the anterior part (arch 1) or in the posterior part (arch 2) seem to exist. In a recent study it was shown that Bmp-7 has an overlapping distribution pattern (48 ). It will be interesting to study whether the various gene products (Slug, Msx-3, Bmp-7 and Hira) function in the same developmental pathway. Another group of Hira expressing cells that is evident consists of neural crest cells in the circumpharyngeal region. In birds, it has been demonstrated that these neural crest cells are crucial for proper development of the outflow tract of the heart. In mice, these cells express the proto-oncogene Ret, and part of this population of hindbrain neural crest cells is on its way to the developing gut or to the superior cervical ganglion (49 ).
In general, the elevated levels of Hira expression overlap with the expression domains of the cellular retinoic acid binding protein gene Crabp-1. Crabp-1 is also expressed in the frontonasal mass, branchial arches 1 and 2 and the neural folds. Exposure of vertebrate embryos to retinoic acid leads to phenocopy of the CATCH22 phenotype (50 ). The overall Hira expression pattern seems to be conserved across species (50 ) .
Full-size Hira transcripts are present in the developing heart at E8.5, where they are no longer detectable at E9.5 and E10.5. The probe we used for the in situ experiments does not detect the short transcript, which is predominant in the adult heart. It is possible that between E8.5 and E9.5 the embryonic heart switches from the full-size to the 2.3 kb transcript. The developing heart arises through a complex series of morphogenetic interactions. In the mouse embryo the heart field is induced at E7.5. The straight heart tube is subsequently specified in an antero-posterior sequence (E9.5) to form the various regions and chambers of the looped (E8.5) and mature heart (E15) (51 ). The expression in the heart at E8.5 suggests that Hira might play a role in cardiac looping. This is of interest in the context of CATCH22, as the heart is one of the affected organs.
We find that two Hira mRNA species (2.3 and 4.4 kb) are present in embryos and adult organs. Both embryonically and at the adult stage (averaged over the individual organs) the 4.4 kb transcript is expressed at higher levels than the 2.3 kb transcript, without dramatic variations in the ratio between the two. Between adult organs the ratio of the expression of the 2.3 and 4.4 kb transcripts varies. The smaller 2.3 kb transcript is the predominant mRNA in the adult heart. Northern blot analysis reveals that this transcript lacks the sequences of the 3' end of canonical Hira. This mRNA thus encodes a protein that consists of mainly WD40 domains. The 2.3 kb transcript could represent the transcript coding for the truncated 12E or MF2 coded polypeptides. The 4.4 kb transcript must represent the full length Hira transcript. Scamps et al. mention only one full length transcript in fetal mouse RNA (36 ). Most likely, they used 3' probes.
Until now, two HIRA transcripts have only been reported for human fetal liver (26 ,27 ). Though these transcripts were labelled 3.4 and 3.2 kb, and 4.5 and 4.2 kb we assume that these are identical to our 5.0 and 4.5 kb adult transcripts. In the adult heart and pancreas the 5.0 kb transcript is predominant, whereas the 4.5 kb transcript is predominant in brain. Like the 2.3 kb transcript in the mouse, the 5.0 kb is not detected by a Hira probe that excludes WD40 repeat coding regions. One explanation could be that the 5.0 kb transcript represents an RNA that encodes a HIRA like protein with C-terminal WD40 domains. This protein would have an N-terminal extension, which could be similar to the N-terminal extension of ceHILP. This part of ceHILP, which has no equivalent in HIRA, CAF1A, Hir1p or Hir2p, bears very high homology to other C.elegans proteins that are similar to Schizosaccharomyces pombe SDS22 and L.monocytogenes Internalin (52 ,53 ). The yeast SDS22 positively modulates protein phosphatase 1 (54 ). Internalin is required for the internalization of L.monocytogenes by eukaryotic cells (55 ). Alternatively, the 5.0 kb transcript encodes a protein with N-terminal WD40 repeats followed by hitherto unknown sequences not present in canonical HIRA.
The insertion of exon 3A into the Hira mRNA disrupts the canonical Hira open reading frame. An N-terminally truncated protein starting from either of the next two in-frame translation start sites lacks the first WD40 domain. This is comparable with the protein coded for by the MF2 Tuple1 (Hira) cDNA, which lacks half of the first WD40 domain due to a 25 bp insert in the mRNA. In Tup1p the array of WD40 domains appears to influence the binding specificity of the individual domains. In the context of other WD40 domains, a single domain has different binding specificities than in isolation (43 ). The lack of a WD40 domain is therefore very likely to alter the affinity of Hira for other proteins. The insertion exon 11A introduces stop codons, yielding a polypeptide consisting only of WD40 domains. This truncated polypeptide could compete with full length Hira for binding to other proteins. This putative truncated form of Hira resembles CAF1A which also consists mainly of WD40 repeats.
In conclusion, at least four different splice forms of HIRA can be identified in humans and mice. This complex expression pattern of the mammalian HIRA gene is regulated in an organ- and developmental-specific manner and arises from alternative splicing and exon skipping.
HIRA is being analyzed for point mutations and microdeletions by several groups that are looking for a cause for the CATCH22 phenotype in apparently non-deleted CATCH22 patients. The new exons 2A, 3A and 11A should be included in these screens. Further analyses of the human 5.0 kb transcript should help in identifying more HIRA sequences to check for mutations.
In the absence of direct evidence that the CATCH22 phenotype is the result of a single gene defect, it is most likely that the phenotype is the result of haploinsufficiency of several genes with major effect. The spatio-temporal expression pattern of HIRA renders the HIRA gene one of the genes of which haploinsufficiency could lead to part of the CATCH22 phenotype. Since deletion screens in CATCH22 patients have so far not reduced the SRDO to less than 300 kb, it will be necessary to analyze the function of genes with major effect in mice. Therefore, we are analyzing the effect on development of a null mutation of the murine Hira gene.
One has to bear in mind, that it will also be important to identify genes with major effect in SRDO2. Furthermore, position effects cannot be excluded. In campomelic dysplasia and aniridia, SOX9 and PAX6, respectively, can be influenced by translocation breakpoints up to 250 kb distant from the disease causing gene (56 ,57 ). A similar effect could play a role in CATCH22. It is also possible that as a result of the CATCH22 deletion a heterochromatic region (e.g. close to the centromere) can exert its influence onto former distal regions, thereby silencing the genes in these regions. In this model it is even possible that the genes of major effect are located outside the commonly deleted region. Since there appear to be at least two SRDOs, one cannot exclude that both position effects and gene deletions can lead to the CATCH22 phenotype. Considering the fact that there are many phenocopies of the CATCH22 phenotype, deletions or silencing of different genes on 22q11 could lead to similar phenotypes.
Embryos were derived from crosses between FVB/N mice. The morning of vaginal plug was considered 0.5 days post coitum (dpc). The embryos were stages according to the number of developmental day (E8.5-E12.5). The E8.5 embryos were staged under the dissecting microscope by counting somites.
A mouse embryo 11.5 dpc cDNA library was purchased from Clontech. Plaques were lifted onto nitrocellulose membranes (Millipore HATF filters) and hybridized as recommended by the manufacturer. The cDNA clones were sequenced by primer walking using the Pharmacia T7 sequencing kit. Sequences were assembled, aligned and compared using the GCG package at the CAOS/CAMM server in Nijmegen. Database searches were performed using Blast programs (58 ). Multiple sequence alignments were edited manually with the GeneDoc program and printed with BoxShade.
Complementary (antisense) or non-complementary (sense) RNA probes were synthesized from linearized DNA templates using the DIG-UTP labelling kit (Boehringer Mannheim). In situ hybridization on whole mouse embryos (E8.5-E11.5) was performed according to the protocol of Wilkinson et al. (59 ), with posthybridization washes modified as follows: 50 and 70 min in solution 1 at 70oC; 20 min in solution 1 and 2 (1:1) at room temperature; 2× 20 min in solution 2 at room temperature; 2× 30 min in solution 2 with RNase A at 37oC; 10 min in solution 2 at room temperature; 2× 30 min in solution 3 at 65oC; 2× 30 min in TBST at room temperature. Adsorption of anti-digoxigenin antibodies to embryo powder was omitted. Embryos were either captured on film and subsequently scanned, or captured electronically with an electronic camera. Images were processed using Adobe Photoshop software.
Adult organs or embryos were homogenized in GIT or LiCl urea with a tissue homogenizer. Total RNA was purified by ultra centrifugation on a cesium-chloride cushion or by LiCl precipitation and organic extractions. Poly(A)+ RNA was isolated from total RNA with the PolyATtract mRNA isolation system III (Promega).
A mouse embryo and a human multiple tissue Northern blot, together with a Zoo blot were purchased from Clontech (Palo Alto, CA). Blots were hybridized as per manufacturer's protocol. The mouse multiple tissue Northern blot was produced by running 5 µg poly(A)+ RNA per lane on a denaturing formaldehyde agarose gel, blotting in 20* SSC onto Qiabrene+ nylon membrane (Qiagen), and UV cross-linking in a Stratalinker (Stratagene).
Complementary (antisense) or non-complementary (sense) [[alpha]-32P]UTP labelled RNA probes were synthesized from linearized Hira cDNA templates, gel-eluted, hybridized to total RNA at 50oC and subjected to RNase protection assay according to ref. 60 .
RT-PCR was performed on 5 µg total RNA or 1 µg poly(A)+ RNA (from E8.5, E9.5, E10.5 and E14.5 mouse embryos and adult mouse brain, spleen, liver and lung) with a mix of 300 ng each of oligo-dT and random hexamer primers according to established protocols. The RT reaction product was used for the following PCR amplification: one cycle of 2 min at 94oC, 2 min at 50oC, 2 min at 72oC; 30 cycles of 1 min at 94oC, 1 min at 50oC, 1.5 min at 72oC and one cycle of 10 min at 72oC. Primer sequences: p1, ctgggaaggttgtgatctgg; p2, ggtattacagagtggccctt; p3, tcgcagacgtccaaatggtt; p4, ccaatgtacgtagcccgctt; p5, atactgctgctgtgctgttg; p6, tttgccaaccgttaatgggg; p7, cccagcaacagcatcactaa; p8, atgctgagcttgtctcccta. Positive controls were 1 pg of cDNAs 12B, E and F. Negative controls were omission of cDNA and a sample from an RT reaction without RNA.
We thank Drs Frans Lohman, Vincent Lui and Maarten Mulder for comments on the manuscript. We thank Peter Scambler for his gift of the human C5 cDNA and sharing unpublished results. Prof. Jan C. Molenaar is acknowledged for his continuous support of our work. This study was financially supported by the Erasmus University, the Sophia Foundation for Medical Research, the European Community and the Dutch Heart Foundation.
E, embryonic age in days post coitum; SRDO, smallest region of deletion overlap.
One of the reviewers pointed to the fact that patients and their families often find the use of the term CATCH22 offensive. We propose to use the acronym in scientific communications only. For clinical purposes the various syndromes could be referred to collectively as the `chromosome 22q11 deletion syndrome'.
Human Molecular Genetics Pages
©
Introduction
Results
Murine Hira cDNA clones coding for a protein with homology with human Chromatin Assembly Factor 1A
Two Hira transcripts are present from gastrulation stage onward
Hira is ubiquitously expressed, with enhanced expression in pharyngeal arches, neural folds, circumpharyngeal neural crest and limb buds
cDNA clones 12B and 12E represent splice variants with exons 3A and 11A
Discussion
Multiple alignment suggests the presence of three functional regions in Hira
Hira expression in embryos
Splice variants of Hira mRNA
Molecular genetics of CATCH22
Materials And Methods
Embryos
cDNA cloning, sequencing and database searches
Whole mount in situ hybridization
RNA isolation
Southern and Northern analyses
RNase protection assay
RT-PCR detection of alternative Hira mRNA
Acknowledgements
Abbreviations
References
Note Added In Proof
REFERENCES
This page is maintained by OUP admin. Last updated Fri Jan 31 11:45:11 GMT 1997. Part of the OUP Journals World Wide Web service.Copyright Oxford University Press, 1996
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
 |
|
 |
|
  A. Ahmad, H. Kikuchi, Y. Takami, and T. Nakayama Different Roles of N-terminal and C-terminal Halves of HIRA in Transcription Regulation of Cell Cycle-related Genes That Contribute to Control of Vertebrate Cell Growth J. Biol. Chem., September 16, 2005; 280(37): 32090 - 32100. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Blackwell, K. A. Martin, A. Greenall, A. Pidoux, R. C. Allshire, and S. K. Whitehall The Schizosaccharomyces pombe HIRA-Like Protein Hip1 Is Required for the Periodic Expression of Histone Genes and Contributes to the Function of Complex Centromeres Mol. Cell. Biol., May 15, 2004; 24(10): 4309 - 4320. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Roberts, H. F. Sutherland, H. Farmer, W. Kimber, S. Halford, A. Carey, J. M. Brickman, A. Wynshaw-Boris, and P. J. Scambler Targeted Mutagenesis of the Hira Gene Results in Gastrulation Defects and Patterning Abnormalities of Mesoendodermal Derivatives Prior to Early Embryonic Lethality Mol. Cell. Biol., April 1, 2002; 22(7): 2318 - 2328. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. V. AMELSVOORT, E. DALY, D. ROBERTSON, J. SUCKLING, V. NG, H. CRITCHLEY, M. J. OWEN, J. HENRY, K. C. MURPHY, and D. G. M. MURPHY Structural brain abnormalities associated with deletion at chromosome 22q11: Quantitative neuroimaging study of adults with velo-cardio-facial syndrome The British Journal of Psychiatry, May 1, 2001; 178(5): 412 - 419. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Hall, D. M. Nelson, X. Ye, K. Baker, J. A. DeCaprio, S. Seeholzer, M. Lipinski, and P. D. Adams HIRA, the Human Homologue of Yeast Hir1p and Hir2p, Is a Novel Cyclin-cdk2 Substrate Whose Expression Blocks S-Phase Progression Mol. Cell. Biol., March 1, 2001; 21(5): 1854 - 1865. [Abstract] [Full Text]
|
|
|
|
 |
|
 A. Verreault De novo nucleosome assembly: new pieces in an old puzzle Genes & Dev., June 15, 2000; 14(12): 1430 - 1438. [Full Text]
|
|
|
|
 |
|
 W S Kerstjens-Frederikse, R M W Hofstra, A J van Essen, J H C Meijers, and C H C M Buys A Hirschsprung disease locus at 22q11? J. Med. Genet., March 1, 1999; 36(3): 221 - 224. [Abstract] [Full Text]
|
|
|
|
 |
|
 H. Yamagishi, V. Garg, R. Matsuoka, T. Thomas, and D. Srivastava A Molecular Pathway Revealing a Genetic Basis for Human Cardiac and Craniofacial Defects Science, February 19, 1999; 283(5405): 1158 - 1161. [Abstract] [Full Text]
|
|
|
|
|
|
S. Lorain, J.-P. Quivy, F. Monier-Gavelle, C. Scamps, Y. Lécluse, G. Almouzni, and M. Lipinski Core Histones and HIRIP3, a Novel Histone-Binding Protein, Directly Interact with WD Repeat Protein HIRA Mol. Cell. Biol., September 1, 1998; 18(9): 5546 - 5556. [Abstract] [Full Text]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||