Nature Genetics
27, 412 - 416 (2001)
doi:10.1038/86912
The amnionless gene, essential for mouse gastrulation, encodes a visceral-endoderm−specific protein with an extracellular cysteine-rich domainSundeep Kalantry1, 2, 4, Sharon Manning1, 2, 4, Olivia Haub1, Carol Tomihara-Newberger1, 2, Hong-Gee Lee1, Jennifer Fangman3, Christine M. Disteche3, Katia Manova1
& Elizabeth Lacy1, 21 Molecular Biology Program, Memorial Sloan-Kettering Cancer Center, Weill Graduate School of Medical Sciences of Cornell University, New York, New York, USA. 2 Molecular Biology Graduate Program, Weill Graduate School of Medical Sciences of Cornell University, New York, New York, USA. 3 Department of Pathology, Box 357470, University of Washington, Seattle, Washington, USA. 4 These authors contributed equally to this work.
Correspondence should be addressed to Elizabeth Lacy e-lacy@ski.mskcc.orgFate-mapping experiments in the mouse have revealed that the primitive streak can be divided into three functional regions1: the proximal region gives rise to germ cells and the extra-embryonic mesoderm of the yolk sac; the distal region generates cardiac mesoderm and node-derived axial mesendoderm; and the middle streak region produces the paraxial, intermediate and lateral plate mesoderm of the trunk. To gain insight into the mechanisms that mediate the assembly of the primitive streak into these functional regions, we have cloned and functionally identified the gene disrupted in the amnionless (amn) mouse, which has a recessive, embryonic lethal mutation that interferes specifically with the formation and/or specification of the middle primitive streak region during gastrulation2. Here we report that the gene Amn encodes a novel type I transmembrane protein that is expressed exclusively in the extra-embryonic visceral endoderm layer during gastrulation. The extracellular region of the Amn protein contains a cysteine-rich domain with similarity to bone morphogenetic protein (BMP)-binding cysteine-rich domains in chordin, its Drosophila melanogaster homolog (Short gastrulation) and procollagen IIA (ref. 3). Our findings indicate that Amn may direct the production of trunk mesoderm derived from the middle streak by acting in the underlying visceral endoderm to modulate a BMP signaling pathway.
The amn mouse has an insertional mutation generated by the integration of a human CD8 transgene (T81) into distal mouse chromosome 12, downstream from the 3' end of Traf3 (ref. 4). The transgene insert is approximately 200 kb (Fig. 1a) and consists of a co-integrate of T81 DNA and sequences from mouse chromosome 6 (data not shown). To define the nature and extent of any rearrangement of chromosome 12 sequences at the transgene insertion site, we isolated the wild-type gene, Amn, in the form of BAC clones that hybridize to a 3' Traf3 probe. The genomic organization of one of these clones, BAC80, indicated that the transgene insertion site resides in a 54-kb NotI fragment (Fig. 1a). Our previous studies had demonstrated that the transgene insertion does not perturb Traf3 or sequences immediately upstream of the NotI site (N1; Fig. 1a; ref. 4). Thus, we focused on sequences within the 54-kb NotI fragment and within the 6-kb fragment to its right (Fig. 1a). We determined the nucleotide sequence of a total of 42 kb from the wild-type Amn locus, including 10 kb upstream and 26 kb downstream of N1 and 6 kb downstream of N2 (Fig. 1a). We developed primer pairs from these sequences and applied them in PCR assays on heterozygous (Amn+/amn) and homozygous (Amnamn/amn) amn DNAs to determine if any sequences were deleted from the amn allele. These assays identified a deletion of 1,020 to 1,040 bp at the transgene insertion site; all other chromosome 12 sequences are intact (Fig. 1a,b).
 | | Figure 1. The amn locus harbors a small DNA deletion. | | |  | a, Location of primer pairs used to define the deletion at the amn locus and to genotype the wild-type and amn alleles. Top, the locations of primer pairs used to identify the amn allele. Bottom, the locations of primer pairs within the 2.24 kb present in an 18-kb lambda clone spanning the transgene insertion site. Middle, the genomic DNA insert in BAC80, representing the wild-type locus, Amn. b, PCR assays carried out on DNA prepared from E9.5 Amn+/amn and Amnamn/amn embryos. Lane 1, 100-bp ladder; lane 2, the PF-1 + PR-2 primer pair upstream of the transgene insertion site yields a PCR product of 538 bp in both Amn+/amn and Amnamn/amn DNAs; lanes 3−6, primer pairs generate a PCR product in Amn+/amn but not in Amnamn/amn DNA. PF-3 + PRwt, 181 bp (lane 3); PF-4 + PR-3, 618 bp (lane 4); PF-5 + PR-3, 462 bp (lane 5); PF-6 + PR-3, 356 bp (lane 6). Lanes 7, 8, primer pairs generate a PCR product in both Amn+/amn and Amnamn/amn DNAs. PF-7 + PR-3, 335 bp (lane 7); PF-8 + PR-3, 253 bp (lane 8). Lane 9, Amn genotyping assay using PF-2, PRwt and Tg-R-2. The PF-2 + PRwt fragment is 302 bp, and the PF-2 + Tg-R-2 amn fragment is 541 bp. Lane 10, negative control. PF-7 lies immediately adjacent to PF-6; thus the downstream terminus of the deletion must fall within the 20 bp comprising PF-6. The length of the deletion must be within 1,020 to 1,040 bp, as PF-6 extends from bp 1,020−1,040 downstream from transgene insertion site, which itself resides 75 bp downstream of N1.
 Full Figure and legend (61K) |
| |
The minimal rearrangement of chromosome 12 sequences at the transgene insertion site indicated that Amn was contained within BAC80. Consistent with this supposition, a BAC80 transgene supported the development of Amnamn/amn embryos into viable, fertile adult mice (Fig. 2). Thus, this rescue established that Amn and its required regulatory elements reside on the 120 kb included in BAC80.
| | |
We identified three genes on BAC80 (Fig. 3a). Previously, we had excluded one of these, Traf3, as an Amn candidate, as Traf3-null mice die postnatally2,
5. Cdc42bpb is the mouse homolog of the rat gene encoding myotonic dystrophy kinase-related Cdc42-binding kinase- (ref. 6). We positioned Cdc42bpb on BAC80 through a combination of exon-trapping experiments and queries of BAC80 sequences against the non-redundant database. Cdc42bpb is transcribed in the opposite orientation from Traf3 and its 3' end resides 18 kb from the N1 site (Fig. 3a). Consequently, the deletion at the amn locus did not remove any 3' exon sequences from Cdc42bpb. BAC80 is missing the 5' portion of Cdc42bpb, including the promoter, translation start site and first 149 codons. Therefore, the rescue of the amn gastrulation defects by BAC80 also seems to exclude Cdc42bpb as a candidate for Amn.
| | Figure 3. Identification of an Amn candidate that is disrupted and partially deleted by the transgene insertion. | | |  | a, Schematic diagram of the amn locus. An approximately 200-kb transgene fragment is integrated 7.9 kb downstream of Traf3 on distal chromosome 12. BAC80 contains 120 kb of the wild-type Amn locus, including all exons of Traf3, except those encoding the 5' UTR and translation start site. Similarly, BAC80 does not contain the 5' end of Cdc42bpb. The most 5' exon of Cdc42bpb present in BAC80 resides in the terminal 6-kb fragment. It is homologous to base pairs 495−644 of the full-length 5.3-kb rat cDNA and corresponds to aa 150−199. A BLAST search against dbEST using a 6-kb contig assembled from wild-type sequences on either side of the transgene insertion site yielded 19 mouse, 17 rat and 10 human ESTs. These fall into two groups demarcated by the NotI restriction site N1, the 5' ESTs to the left of N1 and the 3' ESTs to the right. This distribution of ESTs reflects a directional cloning strategy in which the 5' and 3' ends of a message are defined, respectively, by EcoRI and NotI sites artificially introduced into the cDNAs during the cloning procedure. Homology of ESTs to genomic sequence is depicted in red. b, Amn occupies 5.3 kb of genomic sequence, contains 12 exons and 11 introns, and encodes an ORF of 458 aa. The transgene inserted into intron 7 and deleted between 1,020 and 1,040 bp. An analysis of the predicted amino acid sequence using the TMAP and PSORT II programs indicated that it represents a type I transmembrane protein with a signal sequence and transmembrane domain between amino acids 1 and 19 and 361 and 381, respectively.
Full Figure and legend (27K) |
| |
We identified the third gene contained within BAC80 through a BLAST search of dbEST using 6 kb of sequence flanking the transgene insertion site (2 kb to the left and 4 kb to the right of N1; Fig. 3a). This search yielded multiple mouse, as well as human and rat, ESTs that were assembled into a contig with a consensus sequence of 1,566 bp, including an ORF of 1,374 bp. The ORF encodes a predicted type I transmembrane protein of 458 amino acids. Based on an alignment of the EST contig with genomic DNA sequence, this Amn candidate consists of 12 exons and 11 introns spread over 5.3 kb (Fig. 3a,b). The T81 transgene fragment inserted into intron 7, physically disrupting the gene and deleting the remaining 3' exons and introns, except the 3' UTR and the terminal 6−12 codons. As this gene is entirely encoded on the rescuing BAC, it became the prime candidate for Amn.
We next generated a line of mice carrying a targeted null allele of the Amn candidate (Amntm1Ehl, hereafter Amngfp). In this allele, the genes encoding enhanced green fluorescent protein (eGFP) and Neo replace the entire protein-coding region of the Amn candidate (Fig. 4a). We dissected embryos at embryonic day (E) 7.5 and E8.5 from 3 different matings: Amn+/amn intercrosses; crosses between Amn+/amn and Amn+/gfp mice; and Amn+/gfp intercrosses. The resulting Amnamn/amn, Amngfp/amn and Amngfp/gfp embryos have nearly identical phenotypes (Fig. 4d). The lack of complementation between the Amngfp and Amnamn alleles in the compound heterozygotes demonstrates that the gene identified at N1 is Amn. The comparable phenotypes of the Amnamn/amn and Amngfp/gfp embryos also rule out any other candidate gene that might reside in the 14−15 kb between the 3' ends of Amn and Cdc42bpb. In agreement with this conclusion is the absence of detectable ESTs or exons from this region (data not shown). Although the Amnamn allele contains an intact promoter region (Fig. 3a) and is transcriptionally active in amn mutants (data not shown), it must nevertheless act as a null allele because the Amnamn/amn, Amngfp/amn and Amngfp/gfp embryos have equivalent developmental defects.
| | |
In situ hybridization analysis of wild-type blastocysts and early postimplantation stage embryos revealed that Amn is first transcribed in the primitive endoderm at E4.5 (Fig. 5a−d). We found Amn expression specifically in the visceral endoderm at all subsequent stages of gastrulation examined (E5.5−E8.5; Fig. 5e−h, and data not shown). As shown by the immunohistochemistry assays (Fig. 5i,j), the Amn protein is localized to the apical surface of the visceral endoderm cells at E6.0 and E7.5. A similar distribution of Amn is detected at E5.5; in particular, Amn is expressed throughout the visceral endoderm and is not localized to either the posterior or anterior side of the embryo (data not shown). These expression data agree with our previous finding that the gastrulation defects of amn mice are not rescued in chimeras generated by injection of wild-type ES cells into Amnamn/amn blastocysts; thus Amn must act in an extra-embryonic tissue2. The expression pattern indicates that Amn functions exclusively in the visceral endoderm to direct epiblast growth and assembly of middle primitive streak.
| | Figure 5. Amn is expressed exclusively in primitive endoderm and visceral endoderm during early postimplantation stages. | | |  | a−h, Sections of wild-type embryos hybridized to a full-length Amn cDNA probe prepared from the composite EST contig. i,j, Immunohistochemical staining using an antisera raised against the N-terminal portion of Amn (aa 138−204). a,c,e,g,h, Dark-field photographs. b,d,f,i,j, Bright-field photographs. a,b, Amn is not transcribed in E3.5 blastocysts. Arrow in (a) identifies the blastocyst that was transferred to an oviduct for tissue processing. c,d, Amn is transcribed specifically in the primitive endoderm of E4.5 postimplantation embryos. e,f,g,h, Amn is transcribed in the visceral endoderm derivative of the primitive endoderm at E5.5 (e), E6.5 (g) and E7.5 (h). Amn protein localizes to the apical surface of the visceral endoderm in pre- (i) and post- (j) gastrula stage embryos. i, Sagittal section of a pre-gastrula E6.0 embryo. j, Sagittal section of an E7.5 embryo.
Full Figure and legend (58K) |
| |
Queries of the predicted Amn amino acid sequence against the non-redundant database did not detect any related proteins. But a stretch of 70 amino acids in the extracellular region of Amn has similarity to cysteine-rich (CR) modules present in a small group of proteins known to function as BMP inhibitors: chordin, Short gastrulation (SOG) and procollagen IIA (Fig. 6a). These CR regions mediate the biological activity of chordin, SOG and procollagen IIA by binding to BMPs and blocking their access to signaling receptors3,
7. Recent reports have described additional proteins with similar CR modules, including CRIM1 (ref. 8), Xenopus laevis Kielin9, Drosophila Crossveinless-2 (Cv-2; ref. 10), and Drosophila and X. laevis Twisted-gastrulation (TSG and Tsg, respectively; refs. 11,12). Whereas Kielin and the Caenorhabditis elegans homolog of CRIM1 exert anti-BMP activity3,
9, Cv-2 acts genetically to enhance BMP/DPP signaling10. Depending on the experimental or genetic context, Tsg and TSG are capable of either inhibiting or increasing BMP activity11,
12. A BMP binding site maps to the amino-terminal region of Tsg, which contains its single CR module12. Therefore, it is notable that the Amn CR module has greater identity and similarity to the X. laevis chordin CR2 than does the Tsg CR module (Fig. 6a).
| | Figure 6. Amn encodes a cysteine-rich (CR) region with homology to CR2 in chordin. | | |  | a, Alignment of the CR module in Amn (aa 184−253) with the second of four CRs in X. laevis (x) and mouse (m) chordin and with the single CR in mouse procollagen IIA and X. laevis Twisted gastrulation (xTsg). Dark shading, conserved residues; light shading, similar residues. The 70-aa CR of Amn contains 9 cysteines that align with the C-terminal 9 of 10 cysteines in the chordin CR. The seven cysteines in the xTsg CR align with the C-terminal seven cysteines in chordin. % identity/% similarity, Amn CR to xChordin CR2, 29/39; Amn CR to chordin CR2, 29/41; xTsg CR to xChordin CR2, 27/32; xTsg CR to Amn CR, 38/44. The Amn CR aligns with chordin CR2 over a stretch of 49 aa, whereas the xTsg CR aligns over only 34 aa to chordin CR2 and Amn CR. b, A human and putative Drosophila homolog of Amn. % identity/% similarity between the predicted amino acid sequences for human and mouse AMN and for Drosophila and mouse Amn is shown for five different segments of the protein: N terminal to the CR; CR; extracellular region C terminal to the CR; TM, transmembrane domain; Cytopl., cytoplasmic domain. The sizes of human AMN, mouse Amn and Drosophila Amn are, respectively, 454, 458 and 505 aa. The cytoplasmic domain of Drosophila Amn is nearly double the size of that in mouse and human.
Full Figure and legend (49K) |
| |
We have identified a human homolog of Amn on chromosome 14q32. Consistent with the synteny between this region and distal mouse chromosome 12, human AMN resides between TRAF3 and CDC42BPB, in the same transcriptional orientation as TRAF3. Queries of the fly genome have also identified a putative Drosophila homolog of Amn (ref. 13). An alignment of the predicted amino acid sequences of the Drosophila, human and mouse genes indicates that the N-terminal portion, together with the CR module, is the most conserved segment of Amn (Fig. 6b) and thus that it is likely to be critical for Amn function. Therefore, future studies will seek to establish whether Amn does, in fact, act as a modulator of BMP signaling.
Perhaps the most intriguing feature of Amn is that it acts in the visceral endoderm to affect cell behaviors in the adjacent epiblast. The apical location of Amn places it on the opposite side of the visceral endoderm from the epiblast; therefore Amn is unlikely to interact directly with epiblast cells. We propose instead that Amn modulates a BMP signaling pathway within visceral endoderm to direct the production of a set of gene products that do interact directly with epiblast cells and influence their behavior. Unlike chordin and Sog, Amn is transmembrane-bound. This indicates that Amn might modulate BMP receptor function by serving as an accessory or co-receptor that, through its CR domain, facilitates or hinders BMP binding. Notably, a type I BMP receptor, Alk2 (ActRIA, encoded by Acvr1), is also located on the apical surface of the visceral endoderm during early postimplantation development14. Moreover, similar to amn mice, developmental arrest of Acvr1-deficient embryos at the mid-streak stage can be rescued by the provision of wild-type visceral endoderm in chimeras14,
15. Finally, the apical location of both Amn and Alk2 indicates that they may potentially interact not only with ligands expressed by visceral endoderm, such as BMP2 (ref. 16), but also with ligands expressed in the surrounding maternal tissue, for example BMP7 and BMP2 (refs. 16,17). Genetic and biochemical experiments are under way to explore these models of Amn function.
Methods
amn mice.
We generated the transgenic line carrying the Amnamn allele on the inbred C57BL/6J background4 and then backcrossed it to and maintained it on the 129S3/SvImJ background (The Jackson Laboratory, stock number JR002448; ref. 2).
Characterization of transgene-insertion locus.
We isolated BAC80 from the Genome Systems 129/SvJ library using a unique 1.1 XbaI probe defined previously in Traf3 (ref. 4). We determined the nucleotide sequence of 42 kb of the wild-type amn locus using 3 different clones: an 18-kb  -clone spanning N1 (Fig. 1) that we had isolated from a 129/SvJ -FIX II library (Stratagene); a 26-kb fragment of a P1 clone (RIII library, Genome Systems) containing sequences to the right of N1; and a 6-kb subclone of BAC80 including sequences to the right of N2. We determined nucleotide sequence using the BioResource Center/DNA Sequencing Facility of Cornell University. The sequence and location of primers used in the PCR assays are as follows: PF-1, 5'−GACTCGGACCCAGCTTGTAA−3', bp 9,757−9,776; PR-2, 5'−GAAAGCAGTCAGGTCCTCGT−3', bp 10,294−10,275; PF-2, 5'−CCCTACTGCCTCAAGGGACAATG−3', bp 10,578−10,599; PF-3, 5'−CTGTCCCAGGAGCTATCGTG−3', bp 10,699−10,718; PRwt, 5'−TTGGTCCGGACTCTGTCATC−3', bp 10,879−10,860; PF-4, 5'−AGGTGGAGAAGGCATGAAGA−3', bp 11,382−11,401; PF-5, 5'−GAGCCCTCTGACCTTGTGAC−3', bp 11,540−11,559; PF-6, 5'−CACTCTTTGCTGGGGAGGCAG−3', bp 11,646−11,666; PF-7, 5'−AGGCAGAGGCCTGACCTGC−3', bp 11,667−11,685; PF-8, 5'−CTGGTGGCAAGGACAGACTT−3', bp 11,749−11,768; PR-3, 5'−ATCCTGGAGGCTGAGACAGA−3', bp 12,001−11,982; Tg-R-2, 5'−GATTTCGCATTTGGAGGATG−3', bp 492−473 bp in T81a.
BAC transgene rescue.
We provided BAC80 DNA, diluted into injection buffer (10 mM Tris-HCl, pH 7.5, 0.1 mM EDTA, 100 mM NaCl, 30  M spermine, 70 M spermidine) at a concentration of 1 ng/l, to the MSKCC Transgenic Mouse Core Facility for injection into fertilized mouse eggs recovered from matings between C57BL/6J females and Amn+/amn males. We identified BAC80 transgenic (Tg+) animals by PCR using primers that amplify a 394-bp fragment from the T7 end of the pBeloBACII vector: 5'−TATCCGCTCACAATTCCACA−3' and 5'−TTCATCATGCCGTTTGTGAT−3'. We confirmed the identity of BAC80Tg+ founders by Southern-blot hybridization to SpeI-digested tail DNA, using as probe a 617-bp PCR fragment derived from pBeloBACII: 5'−GCTTAACTATGCGGCCATCAG−3' and 5'−TCAAACATGAGAATTGGTCG−3'. We identified 10 BAC80Tg+ founders of 97 mice screened. We used a Tg+ founder male that was also Amn+/amn in the rescue experiment following the mating scheme diagrammed in Fig. 2. The transgenic line established from this founder is called 129-Tg(BAC80)62Ehl.
FISH analysis.
We labeled the transgene probe containing a 10.4-kb insert with biotin-11-dATP by nick translation (Gibco BRL). We prepared interphase nuclei and metaphase chromosomes from short-term spleen cultures from transgenic mice. After teasing out cells from the spleens, we separated the lymphocytes on a Lympholyte (Accurate Chemical and Scientific) gradient and placed them in RPMI 1640 containing lipopolysaccharide (LPS, Sigma). We collected the cells after 48 h using KCl (0.075 M) as a hypotonic buffer and methanol:acetic acid (3:1, v/v) as fixative. We carried out the hybridization as described18. We distinguished hybridization signals using a detection system from Vector Laboratories. We visualized the hybridization signals by fluorescence using dual band pass filter (Omega). We scored a total of 20−30 nuclei (with 0, 1 or 2 signals) for each of 9 mice. Approximately 70−100% of nuclei scored for each mouse had a consistent number of signals (1 or 2 signals). Nuclei with a number of signals that differed from the expected likely reflect inefficient hybridization or tetraploidy. We confirmed the data obtained in interphase nuclei by analysis of metaphase chromosomes, which were available for eight of the nine mice.
Generation of the Amngfp allele.
Using Pfu Turbo DNA polymerase (Stratagene), we amplified the 3.7-kb and 2.3-kb arms of homology by PCR from 129/SvJ-derived genomic clones spanning Amn. We subcloned the resulting products into the pPCR-Script Amp SK(+) plasmid (Stratagene). The left arm of homology consists of 3.7 kb of sequence directly upstream of the ATG in Amn. We released the 3.7-kb insert by digestion with XhoI and cloned it into the XhoI site of the pEGFPKT1loxneo targeting vector19. We blunt-end cloned a PGK-HSV-TK cassette, excised from pPNT (ref. 20) by EcoRI and HindIII digestion, into the NdeI site upstream of the 3.7-kb arm of homology. We released the 2.3-kb right arm of homology from pPCR-Script by ClaI digestion and subcloned it into the ClaI site directly downstream of MC1neo in pEGFPKT1loxneo. We linearized the targeting vector by SspI digestion and electroporated it into 129/SvJ ES cells (Genome Systems). We recovered 48 clones following growth in G418 (350 g/ml; Gibco BRL) and gancyclovir (2 M; Roche). We verified correct targeting at both the 5' and 3' ends in three clones by Southern-blot and PCR analysis. A fourth clone was properly targeted only at the 5' end. We obtained germline chimeras from one of the three clones using standard procedures. We genotyped subsequent generations of progeny using a three-primer PCR strategy4: PF-1, 5'−GACTCGGACCCAGCTTGTAA−3'; PRwt, 5'−TTGGTCCGGACTCTGTCATC−3'; and Prmut, 5'−GTCCTCCTTGAAGTCGATGC−3' from eGFP.
Generation of Amn antisera.
We PCR-amplified a cDNA fragment corresponding to aa 138−204 of the Amn protein and cloned it in-frame into the GST fusion vector pGEX-4T-1 (Amersham Pharmacia Biotech). We transfected the sequence-verified construct into BL-21 E. coli cells (Stratagene) and expressed and purified the encoded GST fusion protein as described21. We determined protein concentration as described22. We raised polyclonal rabbit antisera against the GST fusion protein using the services of the Pocono Rabbit Farm and Laboratory.
In situ hybridization and immunohistochemistry.
We carried out histological analyses and in situ hybridizations as described2. For the immunohistochemical analysis of Amn expression, we perfused pregnant mice and then fixed the embryos within their decidua in 4% paraformaldehyde. We stained paraffin sections (8 m) using Vectastain Elite ABC kit (Vector Laboratories) as described23. We used the immune and pre-immune sera at 1:4,000 and 1:2,000 dilutions, respectively.
Identification of human AMN.
We screened the human EST database using mouse Amn cDNA sequence. We assembled human ESTs with the highest homology (83−84%) into a contig, which we queried against the non-redundant database. This query identified a human BAC clone that contains an exact match with the human EST contig and that maps to 14q32, adjacent to TRAF3. This human BAC clone also includes sequences from CDC42BPB, which lies distal to the AMN sequences. An alignment of genomic sequences in the BAC clone with the human EST contig and the mouse Amn cDNA mapped the intron-exon structure of human AMN, which we predict to encode a message of 1,896 bp with an ORF of 454 aa.
GenBank accession numbers.
clone, AF320615; P1 subclone containing sequences to the right of N1, AF320616 and AF320617; 6-kb BAC80 subclone, AF320618; mouse EST clones yielding longest possible consensus ORF on joining, AA871896 and AA097423; mouse Amn cDNA, AF320619; human AMN cDNA, AF328788; human BAC containing AMN, AL133455.
Received 22 December 2000; Accepted 15 February 2001
REFERENCES
- Lawson, K.A., Meneses, J.J. & Pedersen, R.A. Clonal analysis of epiblast fate during germ layer formation in the mouse embryo. Development 113, 891−911 (1991). | PubMed | ISI | ChemPort |
- Tomihara-Newberger, C. et al. The amn gene product is required in extraembryonic tissues for the generation of middle primitive streak derivatives. Dev. Biol. 204, 34−54 (1998). | Article | PubMed | ISI | ChemPort |
- Larrain, J. et al. BMP-binding modules in chordin: a model for signalling regulation in the extracellular space. Development 127, 821−830 (2000). | PubMed | ISI | ChemPort |
- Wang, X. et al. A candidate gene for the amnionless gastrulation stage mouse mutation encodes a TRAF-related protein. Dev. Biol. 177, 274−290 (1996). | Article | PubMed | ISI | ChemPort |
- Xu, Y., Cheng, G. & Baltimore, D. Targeted disruption of TRAF3 leads to postnatal lethality and defective T-dependent immune responses. Immunity 5, 407−415 (1996). | Article | PubMed | ISI | ChemPort |
- Leung, T., Chen, X.Q., Tan, I., Manser, E. & Lim, L. Myotonic dystrophy kinase-related Cdc42-binding kinase acts as a Cdc42 effector in promoting cytoskeletal reorganization. Mol. Cell. Biol. 18, 130−140 (1998). | PubMed | ISI | ChemPort |
- De Robertis, E.M., Larrain, J., Oelgeschlager, M. & Wessely, O. The establishment of Spemann's organizer and patterning of the vertebrate embryo. Nature Rev. Genet. 1, 171−181 (2000). | Article | PubMed | ISI | ChemPort |
- Kolle, G., Georgas, K., Holmes, G.P., Little, M.H. & Yamada, T. CRIM1, a novel gene encoding a cysteine-rich repeat protein is developmentally regulated and implicated in vertebrate CNS development and organogenesis. Mech. Dev. 90, 181−193 (2000). | Article | PubMed | ISI | ChemPort |
- Matsui, M., Mizusek, K., Nakatani, J., Nakanishi, S. & Sasai, Y. Xenopus kielin: a dorsalizing factor containing multiple chordin-type repeats secreted from the embryonic midline. Proc. Natl. Acad. Sci. USA 97, 5291−5296 (2000). | Article | PubMed | ChemPort |
- Conley, C.A. et al. Crossveinless 2 contains cysteine-rich domains and is required for high levels of BMP-like activity during the formation of the cross veins in drosophila. Development 127, 3947−3959 (2000). | PubMed | ISI | ChemPort |
- Yu, K. et al. Processing of the Drosophila Sog protein creates a novel BMP inhibitory activity. Development 127, 2143−2154 (2000). | PubMed | ISI | ChemPort |
- Oelgeschlager, M., Larrain, J., Geissert, D. & De Robertis, E.M. The evolutionarily conserved BMP-binding protein Twisted gastrulation promotes BMP signalling. Nature 405, 757−763 (2000). | Article | PubMed | ISI | ChemPort |
- Kopczynski, C.C. et al. A high throughput screen to identify secreted and transmembrane proteins involved in Drosophila embryogenesis. Proc. Natl. Acad. Sci. USA 95, 9973−9978 (1998). | Article | PubMed | ChemPort |
- Gu, Z. et al. The type I serine/threonine kinase receptor ActRIA (ALK2) is required for gastrulation of the mouse embryo. Development 126, 2551−2561 (1999). | PubMed | ISI | ChemPort |
- Mishina, Y., Crombie, R., Bradley, A. & Behringer, R.R. Multiple roles for activin-like kinase-2 signaling during mouse embryogenesis. Dev. Biol. 213, 314−326 (1999). | Article | PubMed | ISI | ChemPort |
- Coucouvanis, E. & Martin, G.R. BMP signaling plays a role in visceral endoderm differentiation and cavitation in the early mouse embryo. Development 126, 535−546 (1999). | PubMed | ISI | ChemPort |
- Zhang, H. & Bradley, A. Mice deficient for BMP2 are nonviable and have defects in amnion/chorion and cardiac development. Development 122, 2977−2986 (1996). | PubMed | ISI | ChemPort |
- Edelhoff, S. et al. Mapping of two genes encoding members of a distinct subfamily of MAX interacting proteins: MAD to human chromosome 2 and mouse chromosome 6, and MXII to human chromosome 10 and mouse chromosome 19. Oncogene 9, 665−668 (1994). | PubMed | ISI | ChemPort |
- Godwin, A.R., Stadler, H.S., Nakamura, K. & Capecchi, M.R. Detection of targeted GFP-Hox gene fusions during mouse embryogenesis. Proc. Natl. Acad. Sci. USA 95, 13042−13047 (1998). | Article | PubMed | ChemPort |
- Tybulewicz, V.L., Crawford, C.E., Jackson, P.K., Bronson, R.T. & Mulligan, R.C. Neonatal lethality and lymphopenia in mice with a homozygous disruption of the c-abl proto-oncogene. Cell 65, 1153−1163 (1991). | Article | PubMed | ISI | ChemPort |
- Rossi, G., Salminen, A., Rice, L.M., Brunger, A.T. & Brennwald, P. Analysis of a yeast SNARE complex reveals remarkable similarity to the neuronal SNARE complex and a novel function for the C terminus of the SNAP-25 homolog, Sec9. J. Biol. Chem. 272, 16610−16617 (1997). | Article | PubMed | ISI | ChemPort |
- Katz, L., Hanson, P.I., Heuser, J.E. & Brennwald, P. Genetic and morphological analyses reveal a critical interaction between the C-termini of two SNARE proteins and a parallel four helical arrangement for the exocytic SNARE complex. EMBO J. 17, 6200−6209 (1998). | Article | PubMed | ISI | ChemPort |
- Manova, K. et al. c-kit receptor and ligand expression in postnatal development of the mouse cerebellum suggests a function for c-kit in inhibitory interneurons. J. Neurosci. 12, 4663−4676 (1992). | PubMed | ISI | ChemPort |
Acknowledgments
We thank K. Anderson, L. Niswander, P. Wilson, C. Blobel and F. Costantini for discussions and critical reading of the manuscript; J.-H. Dong, J. Ingenito, R. Lester and W. Mark for technical expertise and advice; Y. Zhang and M. Grunwald for technical assistance; J. Choi for his contribution to database analyses; R. Rivi and F. Lupu for discussions and critical reading of the manuscript; and R. Yeung for assistance in the generation of the Amn polyclonal antisera. This work was supported by NIH grant GM58726 (E.L.) and the MSKCC Support Grant CA-08748.
|
