- Email: [email protected]
Absence of the tight junctional protein AF-6 disrupts epithelial cell–cell junctions and cell polarity during mouse development
880
Research Paper
Absence of the tight junctional protein AF-6 disrupts epithelial cell–cell junctions and cell polarity during mouse development Alexander B. Zhadanov*†, D. William Provance Jr.*, C.A. Speer‡, J. Douglas Coffin*§, Dee Goss*, J.A. Blixt‡, Cheryl M. Reichert* and John A. Mercer* Background: The establishment, maintenance and rearrangement of junctions between epithelial cells are extremely important in many developmental, physiological and pathological processes. AF-6 is a putative Ras effector; it is also a component of tight and adherens junctions, and has been shown to bind both Ras and the tight-junction protein ZO-1. In the mouse, AF-6 is encoded by the Af6 gene. As cell–cell junctions are important in morphogenesis, we generated a null mutation in the murine Af6 locus to test the hypothesis that lack of AF-6 function would cause epithelial abnormalities.
Addresses: *McLaughlin Research Institute, 1520 23rd Street South, Great Falls, Montana 594054900, USA. ‡Department of Veterinary Molecular Biology, Montana State University, Bozeman, Montana 59717-3610, USA. Present address: §Department of Pharmaceutical Sciences, University of Montana, Missoula, Montana 59812, USA. †Deceased.
Results: Although cell–cell junctions are thought to be important in early embryogenesis, homozygous mutant embryos were morphologically indistinguishable from wild-type embryos through 6.5 days post coitum (dpc) and were able to establish all three germ layers. The earliest morphological abnormalities were observed in the embryonic ectoderm of mutant embryos at 7.5 dpc. The length of the most apical cell–cell junctions was reduced, and basolateral surfaces of those cells were separated by multiple gaps. Cells of the embryonic ectoderm were less polarized as assessed by histological criteria and lateral localization of an apical marker. Mutant embryos died by 10 dpc, probably as a result of placental failure.
Correspondence: John A. Mercer E-mail: [email protected] Received: 8 February 1999 Revised: 5 July 1999 Accepted: 5 July 1999 Published: 5 August 1999 Current Biology 1999, 9:880–888 http://biomednet.com/elecref/0960982200900880 © Elsevier Science Ltd ISSN 0960-9822
Conclusions: AF-6 is a critical regulator of cell–cell junctions during mouse development. The loss of neuroepithelial polarity in mutants is consistent with a loss of efficacy of the cell–cell junctions that have a critical role in establishing apical/basolateral asymmetry.
Background The dynamic modulation of cell–cell contacts is a critical step in many developmental processes, including the establishment of epithelial cell polarity and complex movements of sheets of cells [1]. Proteins defined by the PDZ sequence motif — a protein interaction motif originally identified as repeated regions of homology between the synaptic protein PSD-95, the product of the Drosophila dlg tumor suppressor gene and the epithelial tight-junction protein ZO-1 — have been shown to function at many different sites of cell–cell interaction, including tight junctions, adherens junctions and synapses. It has been shown that PDZ domains participate in cell–cell junction organization, intracellular signaling and receptor clustering by mediating protein–protein interactions [2]. The human AF6 gene was first identified as a fusion partner with the ALL-1 locus in acute myeloid leukemias with a t(6;11)(q27;q23) translocation [3]. It was also isolated in a yeast two-hybrid screen with H-Ras [4]. The AF6 locus (Af6 in mouse) encodes a modular protein (AF-6 or afadin) that contains a PDZ domain [5]; two motifs that interact with Ras and the tight-junction protein ZO-1
in vitro [4,6,7]; regions with homology to myosin-V-like and kinesin-like cargo-binding domains [8]; and an actinbinding domain present only in a long isoform produced by alternative splicing [9]. From studies of the Drosophila homologs of AF-6, ZO-1 and Ras, we can predict multiple and complex roles for the mammalian proteins in cell–cell interactions during development. The canoe (cno) gene is a Drosophila homolog of Af6 [10,8]. Genetic manipulation of the activity of Cno modulated the effects of both constitutively active and dominantnegative forms of Ras1 in cone cell differentiation [11]. The tamou (tam) locus of Drosophila encodes a protein homologous to ZO-1 [12]. Flies doubly homozygous for the viable hypomorphic alleles pydtam and cnomis1 die as embryos, suggesting a genetic interaction between tam and cno [13]. Of the known Ras effectors [14], AF-6 is the most likely candidate for directly mediating the effects of Ras (or other small GTPases) on cell–cell junctions. The AF-6 protein colocalizes with ZO-1 at tight junctions in epithelial cells [15] and with cadherin at adherens junctions in both rat liver and intestinal epithelium [9]. Overexpression
Research Paper Neuroectodermal abnormalities in Af6 mutant embryos Zhadanov et al.
of activated Ras protein in Rat1 fibroblasts causes a perturbation of cell–cell contacts, with a decreased accumulation of ZO-1 and AF-6 at the cell surface [15]. Together with in vitro binding data, these data suggest that the interaction of AF-6 with ZO-1 is required for the assembly and dynamic modulation of cell–cell junctions, and that this interaction is inhibited by activated Ras. Here, we present evidence that AF-6 function is required for maintenance of cell–cell junctions and cell
881
polarity in the neuroepithelium of embryos at 7.5 days post coitum (dpc).
Results Loss of AF-6 function leads to embryonic lethality by 10.5 dpc
AF-6 protein was detected in blastocysts and throughout development in wild-type embryos. The protein was widely distributed in 8.75 and 9.5 dpc embryos, with strongest signals in the tail bud, epithelial cells of branchial
Figure 1
8.75 dpc +/+
m
–7.9 kb
wt
–6.1 kb
–
–
+/
+/
#53
+/
+
–
#17
+/
– +/
#4
+/ + +/ +
– +/ + +/ +
m wt
–15.1 kb –13.3 kb
–
+
+/
+
+/
+
+/
–
+/
+
+/
+/
–
+/
–
(e) +/
9.5 dpc –/–
(d)
#53
+/
#17
+/ + +/ +
(c)
9.5 dpc +/+
+/ + +/ +
(a)
ES kbp –7.9 kb –6.1 kb
m wt Intercross F2
H
E H
6.1 kb
13.5 kb
M
Backcross N2 M
+/ + +/ – +/ – +/ – +/ + +/ + +/ +
E
(b)
+/ – +/ – +/ – +/ + +/ – +/ – +/ +
(f)
S E E
A
EH 3′
1 3 neo
tk N X Targeting construct
E E
m
650 bp
wt
353 bp
AN (g) M
5′ probe A H E X Targeted Af6 allele
C C
Parents
3′ probe B neo
E E E
H
–/ – +/ – +/ – +/ – +/ + +/ – +/ + +/ + +/ – +/ + +/ – +/ + +/ – –/ –
H E X 5′ Wild-type Af6 locus
15.3 kb
1
2 3 7.9 kb
A
EH E H
m
650 bp
wt
353 bp
Current Biology
AF-6 expression, targeting vector, homologous recombination and germline transmission. (a) AF-6 protein expression at 8.75 and 9.5 dpc in wild-type (+/+) and homozygous mutant (–/–) embryos. High levels of AF-6 were detected in the tail bud (black arrow), epithelial cells of branchial arches (black arrowhead) and neural tube (white arrow). (b) Targeting strategy. Maps of the wild-type Af6 locus (top), the targeting vector (middle) and the disrupted Af6 allele (bottom) are shown. Black boxes, Af6 exons; black bars, external 5′ and 3′ probes used for Southern blot analysis; neo, neomycin resistance selectable marker (with the direction of transcription indicated as an arrow pointing in the 5′ to 3′ direction); tk, thymidine kinase gene (with the direction of transcription indicated as an arrow pointing in the 5′ to 3′ direction); grey boxes, promoter derived from the gene for phosphoglycerate kinase
(PGK); arrows (1–3), PCR primers for genotyping; A, ApaI; E, EcoRI; H, HindIII; N, NotI; S, StuI; X, XhoI. EcoRI and HindIII fragments (with sizes in kb) diagnostic of the wild-type and targeted alleles are shown, respectively, above the map of the wild-type locus and below the map of the targeted allele. (c,d) Southern blot analysis showing homologous recombination at the Af6 locus in ES cell clones 4, 17, and 53 using (c) the 3′ probe B on EcoRI-digested DNA, and (d) the 5′ probe A on HindIII-digested DNA. (e) Southern blot analysis demonstrating germline transmission of the targeted Af6 allele in the progeny of a chimera for ES cell clone 17 using the 3′ probe B on EcoRI-digested genomic DNA. (f,g) PCR genotyping of 9.5 dpc embryos. Lanes labelled C in (g) denote no DNA control samples. In (c–g), diagnostic fragments for the wild type (wt) and the mutant (m) are indicated.
882
Current Biology, Vol 9 No 16
Table 1
we used PCR to genotype embryos from intercrosses of heterozygotes (Table 1). Partially resorbed embryos were observed at 10.5 and 11.5 dpc and were found on genotyping to be homozygotes (Figure 1f,g). At earlier times, homozygous mutants were present at the expected Mendelian proportion (Table 1).
PCR genotyping of progeny derived from intercrosses between Af6/+ heterozygotes. Age of progeny
+/+
+/–
–/–
Liveborn
41
73
0
13.5 dpc
3
6
0
11.5 dpc
2
5
2 (resorptions)
10.5 dpc
5
9
4 (resorptions)
9.5 dpc
13
31
13
8.5 dpc
11
16
9
7.5 dpc
12
26
11
To confirm the absence of AF-6 protein in mutants, mutant and wild-type embryos were analyzed by wholeembryo immunohistochemistry (Figure 1a) and western blot analysis (data not shown). No AF-6 protein was detected in mutant homozygous by either method. The earliest abnormality in mutant embryos is a hyperplastic embryonic ectoderm
To characterize the abnormalities of Af6 homozygous mutant embryos, we analyzed litters from intercrosses of heterozygous mice at stages between 5.5 and 9.5 dpc. The mutant phenotype was identical in mice derived from both targeted ES cell lines, and so results from both lines have been pooled in this study. All embryos were genotyped by PCR.
arches, neural tube, and mid/hindbrain neuroepithelium at 9.5 dpc (Figure 1a). We generated a null mutation at the Af6 locus by homologous recombination in embryonic stem (ES) cells by inserting the PGK–neo cassette into exon 4 of the Af6 locus (Figure 1b). Two independently derived ES cell clones (Figure 1c,d) generated chimeras that transmitted the mutant allele in the germ line (Figure 1e). Mice heterozygous for the Af6 mutation appeared normal and were fertile.
No morphological abnormalities were observed before 6.5 dpc (data not shown). By 7.5 dpc, homozygous mutant embryos were morphologically distinguishable from normal embryos (Figure 2a). Mutant conceptuses showed a reduction in the size of the embryonic portion relative to the extraembryonic portion; posterior body elongation was retarded and neural folds were not observed (Figure 2a). In wild-type embryos, the neural plate was clearly defined anteriorly, with developing neural folds (Figure 2a, NF).
No liveborn homozygous mutants were found among the progeny of matings between heterozygotes (Figure 1e). To determine when Af6 homozygous null embryos died, Figure 2 (a)
(b)
EX
EX
+/+
(c)
–/–
D EM NF
EM
EX
+/+
NF
+/+
–/–
7.5 dpc (e)
(d)
(f)
C
C C
AL
A
EX
EX
EM
–/– +/+ EM 8.5 dpc (g) C AL
9.5 dpc Y (i)
(h)
E
N NE
A
A
S E
PS +/+
A
E
E
E PS –/–
S +/+
–/–
PS +/+
–/– Current Biology
Morphology of wild-type (+/+) and Af6 mutant (–/–) embryos at 7.5, 8.5 and 9.5 dpc. (a) Embryos dissected from the decidua at 7.5 dpc. Normal embryos exhibit AP polarity; the developing neural folds mark the anterior aspect of embryos. (b) Dissected wild-type embryo and a partially dissected mutant embryo at 8.5 dpc. (c) Wild-type and a typical comma-shaped mutant embryo at 9.5 dpc. (d–i) Hematoxylin–eosin-stained sections of wild-type and mutant embryos at (d,e) 7.5 dpc, (f,g) 8.5 dpc and (h,i) 9.5 dpc. EM, embryonic portion of the embryo; EX, extraembryonic portion of the embryo; A, amnion; AL, allantois; C, chorion; D, decidua; E, embryonic ectoderm; N, neural tube; NE, neuroepithelium; NF, neural fold; PS, primitive streak; S, somites; Y, yolk sac. The arrowheads in (e) denote accumulations of (upper) mesoderm and (lower) ectoderm. The bars represent 100 µm in (a,b,c,g,h,i); 50 µm in (d,e); and 200 µm in (f). All embryos shown were genotyped by (a–c) PCR or (d–i) immunohistochemistry.
Research Paper Neuroectodermal abnormalities in Af6 mutant embryos Zhadanov et al.
883
Figure 3 Whole-mount in situ hybridization analysis of T (Brachyury) and Hesx1 expression. (a) T expression at the posterior aspect of the embryo in primitive streak mesodermal cells (arrows) in both wild-type and mutant embryos at 7.5 dpc. (b) Normal T expression (white arrows) at 8.5 dpc in wild-type embryos, and aberrant (but asymmetric) expression (white arrowhead) in mutants at the same stage. (c) Hesx1 expression in anterior visceral endoderm (arrows) at 7.0 dpc marks the anterior aspect of both wild-type and mutant embryos. The bars represent 100 µm.
(a)
(b)
(c) +/+
+/+ T (Brachyury)
+/+
+/+
–/–
Sectioning of 7.5 dpc embryos revealed that the mutants formed all three germ layers, as mesoderm had clearly passed through the primitive streak region (Figure 2d,e). The embryonic ectoderm, however, which at the midline will become the neuroepithelium of the neural plate, was more stratified than columnar (Figure 2e, lower arrowhead; Figure 4). The neuroepithelium was present in mutants, but it was poorly developed at the anterior and posterior of the embryo (Figure 2d,e). The allantois, which forms from a projection of mesodermal cells from the posterior margin of the embryonic ectoderm, was present in normal embryos at this stage. Mutant embryos lacked an allantois; in its place, we observed a posterior accumulation of mesodermal cells in the mutant embryo (Figure 2e, upper arrowhead). By 8.0–8.5 dpc, the presumptive neuroectodermal layer in mutant embryos became thick and disorganized centrally, while the anterior and posterior margins of the layer were thin. Mesodermal structures such as notochord and somites were not observed in 8.5 dpc mutant embryos (Figure 2b,f,g), and expression of a notochord marker, HNF-3β [16,17], was not detected (data not shown).
Hesx1
–/–
T (Brachyury) –/–
+/+ –/–
–/– Current Biology
formation. The T gene (also known as Brachyury) is transiently expressed from the onset of gastrulation in cells ingressing into the primitive streak and in nascent mesoderm [18]. Mutant embryos expressed T transcripts at 7.5 dpc, suggesting the presence of a primitive streak (Figure 3a) despite the abnormal morphology of the mutant mesoderm. By the next day of development, however, expression of T was confined to small patches of cells, possibly corresponding to axial mesoderm (Figure 3b, white arrowhead). T expression was nearly absent by 9.5 dpc in mutant embryos (data not shown). The expression of the homeobox gene Hesx1 (also known as Rpx) marks the anterior aspect of the embryo [19]. Mutant embryos expressed Hesx1 in anterior visceral endoderm at 7.0 dpc, indicating that AP polarity was established despite morphogenetic abnormalities (Figure 3c). By the next day, Hesx1 transcripts are normally localized to rostral neuroectoderm [19]; Hesx1 expression, however, was not detected in 8.5 dpc mutant embryos (data not shown), consistent with the absence of a head fold in mutant embryos. The earliest known anterior marker, Hex Figure 4
Mutant embryos were comma-shaped at 9.5 dpc and lacked recognizable normal structures (Figure 2c,h,i). Histological examination showed that the 9.5 dpc mutant embryos consisted predominantly of loose mesenchyme (Figure 2i). The death of mutant embryos by 10 dpc is likely to be the result of failure to establish any connection between the embryonic and maternal circulatory systems (data not shown), but the developmental abnormalities are so extensive that many alternative explanations are possible.
(a) +/+
(b) –/–
E
M
VE
Mutant embryos asymmetrically express early markers for anteroposterior axis establishment
Although histology provided evidence of morphogenetic failure, it did not indicate whether early differentiation events were arrested as well. Therefore, we examined early markers associated with anteroposterior (AP) axis
Current Biology
Semi-thin (1 µm) sections from 7.5 dpc embryos fixed and embedded for electron microscopy. Sections from (a) wild-type and (b) mutant embryos were stained in toluidine blue. E, embryonic ectoderm; M, mesoderm; VE, visceral endoderm.
884
Current Biology, Vol 9 No 16
[20], also was expressed asymmetrically in mutant embryos (data not shown). Abnormal morphology of apical cell–cell junctions and basolateral intercellular gaps in embryonic ectoderm
From the gross phenotype, we hypothesized that the cell–cell junctions in mutant embryonic ectoderm might be weakened or absent, and therefore unable to maintain the integrity of the ectoderm during growth and movement of the epithelial sheet. To examine cell–cell junction structure in the absence of AF-6, embryos were fixed and embedded for electron microscopy. Initial 1 µm sections examined by light microscopy (Figure 4) showed that the mutant embryonic ectoderm was more stratified and less columnar, and the mesoderm was less dense than the corresponding germ layers in wild-type embryos. No significant differences were observed between wild-type and mutant visceral endoderm at this level of resolution. We then examined all three germ layers of 7.5 dpc embryos by transmission electron microscopy (Figure 5). The most striking difference between wild-type and mutant embryonic neuroectoderm was the presence of multiple gaps between ectodermal cells of the lumenal layer in the mutant embryo (Figure 5d, arrows). In addition, structures resembling microvilli were present in the intercellular gaps (Figure 5d, arrowheads). Higher magnification views of apical junctions between neuroectodermal cells are shown in Figure 5b,e; the wild-type neuroectoderm had classical tight and adherens junctions
(Figure 5b), but junctions of mutant neuroectoderm had consistently shorter electron-dense regions that could not be classified as tight or adherens junctions. These electron-dense regions were frequently displaced basally (Figure 5e). The differences between mutant and wildtype embryonic endoderm (Figure 5c,f) were less striking than those seen in ectoderm, correlating with the absence of endodermal abnormalities at the light microscopic level in the mutant embryos. AF-6 colocalizes with both ZO-1 and E-cadherin in the neuroectoderm of 7.5 dpc embryos
Yamamoto et al. [15] observed that AF-6 colocalizes with ZO-1, and Mandai et al. [9] observed that AF-6 colocalizes with E-cadherin, showing more overlap with E-cadherin than with ZO-1. Initially, we used laser confocal microscopy to determine whether these associations were present in wild-type embryos at 7.5 dpc. We observed extensive overlap in the subcellular distribution of AF-6 with both ZO-1 (Figure 6a) and E-cadherin (Figure 6b) in embryonic ectoderm and visceral endoderm. These overlaps were present in blastocysts as well (data not shown). No detectable changes in the subcellular localization of ZO-1 and E-cadherin were found in Af6 mutant embryos
Because it was difficult to classify the abnormal cell–cell junctions in mutant neuroectoderm, and as AF-6 colocalized with components of both tight and adherens junctions, we examined the distribution of molecular markers for tight and adherens junctions in 7.5 dpc mutant
Figure 5
(a)
+/+
(b)
(c)
(d)
–/–
(e)
(f)
Transmission electron micrographs of (a,b,d,e) embryonic ectoderm and (c,f) visceral endoderm of (a–c) wild-type and (d–f) homozygous mutant embryos at 7.5 dpc. The apical surface is on the right in (a,d), and at the top in (b,c,e,f). (c,d) Extended cellular processes (arrowheads in panel d) are found at the lateral surface of the mutant cells. (d) Multiple gaps (arrows) are present between cells of the mutant; gaps at the center of (d) are enlarged in the inset. The magnification is 3,600 × in (a,d); 45,000 × in (b,e); and 56,000 × in (c,f).
Current Biology
Research Paper Neuroectodermal abnormalities in Af6 mutant embryos Zhadanov et al.
Figure 6
885
Figure 7 +/+
(b)
M
–/–
(a)
NE
(b)
ZO-1
(a)
NE
NE M
NE
M 50 µm
M
VE
(c)
(d)
E-cadherin
NE 20 µm M Current Biology
NE
M
VE Colocalization of AF-6 (red) with (a) ZO-1 (green) and (b) E-cadherin (green) in neuroectoderm (NE) of wild-type 7.5 dpc embryos. M, mesoderm. (e)
embryos by confocal microscopy. An antibody against AF-6 was used to identify mutant embryos (data not shown). As the results of Yamamoto et al. [15] suggested that activated Ras reduces cell–cell contacts in Rat1 cells by removing AF-6 and ZO-1 from the cell membrane, we hypothesized that ZO-1 localization would be changed in mutant embryonic ectoderm. Surprisingly, ZO-1 localization was not altered in the mutant embryos (Figure 7a,b), indicating that ZO-1 does not depend on AF-6 for its submembrane localization in this tissue. Localization of the adherens junction marker E-cadherin was also not significantly different in the neuroepithelia of mutant embryos, compared with the wild type (Figure 7c,d). The localization of β-catenin was similar to that of E-cadherin (data not shown). Both ZO-1 and E-cadherin colocalized with AF-6 in wild-type blastocysts, but the localization of either protein was not changed in mutant blastocysts (data not shown). In our electron micrographs (Figure 5), it appeared that microvilli were present in the basolateral gaps between mutant neuroepithelial cells. To confirm this identification and the reduction in polarity with a molecular marker, we performed immunofluorescent localization using a monoclonal antibody to prominin, an apical marker found in microvilli of neuroepithelium and many other epithelia [21]. The change in prominin localization was dramatic (Figure 7e,f). As shown in Figure 7e, in the wild-type neuroectoderm, prominin (green) had minimal overlap with E-cadherin (red); in mutant embryos, however, there was significant basolateral colocalization of prominin and
NE
M VE
E-cadherin (red)/prominin (green)
VE (f)
NE
M VE Current Biology
ZO-1 and E-cadherin localization is normal in 7.5 dpc Af6 mutant embryos, but the mutant neuroepithelium has lost its polarity. (a,b) Localization of ZO-1 in (a) wild-type and (b) Af6 mutant embryos. (c,d) Localization of E-cadherin in (c) wild-type and (d) Af6 mutant embryos. (e,f) Localization of E-cadherin (red) and prominin (green) in (e) wild-type and (f) Af6 mutant embryos. NE, neuroectoderm; M, mesoderm; VE, visceral endoderm.
E-cadherin (Figure 7f, yellow). Three-dimensional representations of Figure 7e,f are published with this article on the internet (see Supplementary material). These data demonstrate that the polarity of neuroectoderm is dramatically reduced in the absence of AF-6 function.
Discussion We propose that AF-6 is necessary for the dynamic regulation of tight and/or adherens junctions. More importantly, our data indicate that AF-6 function is only necessary in specific cell types that do not correspond entirely with the cell types in which AF-6 is expressed. Not unexpectedly, the regulation of tight and adherens junction during postimplantation development is far more complicated
886
Current Biology, Vol 9 No 16
than has been suggested by studies of cultured cells and preimplantation embryos. The Af6 mutant phenotype is interesting in several respects. First, the surprising lack of abnormalities in preimplantation development suggests that AF-6 is not an essential structural component of either tight or adherens junctions, which (along with desmosomes) are prominent in early development [22]. Desmosomes do not appear to be relevant to the initial abnormalities, as we did not identify desmosomes in several hundred wild-type and mutant embryonic cells (data not shown). All of our data are consistent with the hypothesis that AF-6 has a regulatory, not a constitutive or structural, function, although these roles are not mutually exclusive and functional redundancy is an alternative explanation. Second, the loss of columnar epithelial morphology in mutant neuroectoderm resembles hyperplasia, consistent with a role for AF-6 as a Ras effector [4] that could participate in the epithelial–mesenchymal transition [23] in carcinomas. This predicts that transformed Af6 homozygous mutant cells may be more invasive or metastatic than wildtype cells. Electron microscopy (Figure 5) showed that mutant neuroectoderm had electron-dense cell–cell contacts that were not identifiable as tight or adherens junctions. The large intercellular gaps observed are consistent with either reduced strength of cell–cell junctions or alterations in transepithelial transport, which also might be predicted from the loss of polarity that we observed. Similar intercellular gaps have been observed in Xenopus embryos following injection of dominant-negative N-cadherin [24] and E-cadherin [25]. It is important to note that the gaps are present along the entire lateral surfaces of the cells, not just apically. This is surprising given the normal apical location of AF-6 (Figure 6). A detailed immunoelectron microscopic characterization of cell–cell junctions in blastocysts, embryoid bodies and homozygous mutant ES cells is now in progress. In a developmental context, the tissue specificity of the Af6 mutant phenotype is consistent with results from other investigators, as wild-type neuroepithelium has been shown to reorganize and lose polarity between 8 and 9 dpc [26]. During this time, expression of occludin is decreasing and expression of ZO-1 is increasing in chick embryos [27]. These results can be integrated by hypothesizing that the epithelial character of these cells is beginning to be lost at 7.0 dpc, and that AF-6 may represent a pivotal mechanism for modulating cell–cell junctions downstream of Ras or other small GTPases during this period. Any roles of AF-6 in later morphogenetic events are likely to be masked by the severe abnormalities caused by the initial disruption at 7.5 dpc. There is some controversy over the identification of AF-6 as a component of tight junctions [15] or (as afadin) as a
component of adherens junctions [9]. Our demonstration of colocalization of both ZO-1 and E-cadherin with AF-6 in wild-type neuroectoderm suggests that there may be no conflict between these two results. All three analyses were performed under different conditions and therefore any conclusions derived from them are clearly not mutually exclusive. Moreover, the abnormalities we observed in neuroepithelial cell–cell junctions defy description in classical terms and do not help to distinguish between a role of AF-6 in tight versus adherens junction regulation. At this point, our data are consistent with both hypotheses. Our failure to detect any change in localization of ZO-1 in the mutant embryos was not expected, however, given the AF-6–ZO-1 interactions demonstrated in vitro by Yamamoto et al. [15]. Despite the apparent discrepancy between the effects they observed upon overexpression of Ras and the effects we observed upon removal of AF-6, both suggest a regulatory role for AF-6; hypothesizing a more structural function for AF-6 in directing the localization of ZO-1 would predict mislocalization of ZO-1 in the mutant. The striking loss of neuroepithelial polarity in the Af6 mutant embryos is consistent with data from other systems [28], which suggest a role for cell–cell junctions in directing polarity. Although we observed lateral localization of both apical structures and an apical marker, the basolateral localization of E-cadherin itself was not perceptibly changed in the mutant embryos. Whether this loss in polarity is a downstream effect of the alterations in cell–cell junctions or directly downstream of AF-6 cannot be determined. In summary, these data suggest that the regulation of cell–cell junctions during embryogenesis is tissue-specific, and that AF-6 is an important part of this mechanism in at least one important embryonic epithelium. More importantly, these data indicate that this regulation may be complex and indirect, as the mechanism of cell–cell junction disruption was not illuminated at the level of light microscopy by immunolocalization of the basic markers of tight and adherens junctions.
Materials and methods Generation of Af6 mutant mice A 412 bp cDNA probe was obtained by reverse-transcriptase PCR from mouse brain cDNA using oligonucleotide primers corresponding to the amino-acid sequences EKFRPDM and PETSFTR, which are conserved between human AF6 and Canoe. This cDNA probe was mapped (using the Jackson Laboratory BSS mapping panel) to proximal mouse chromosome 17, which corresponds to human 6q27. The probe was used to screen a 129/SvJ genomic bacterial artificial chromosome library (Genome Systems). A 6.2 kb XhoI–ApaI Af6 genomic fragment was subcloned into pGEM11 (Figure 1b). The neo expression cassette from pPNT [29] was subcloned into a StuI site. This Af6–neo fragment was inserted into pPNT. The linearized targeting vector was electroporated into R1 ES cells [30], which were subjected to double selection with G418 and gancyclovir. Four independent homologous recombination events occurred among 75 doubly resistant colonies, as assessed by Southern blot hybridization with both 5′
Research Paper Neuroectodermal abnormalities in Af6 mutant embryos Zhadanov et al.
(Figure 1d) and 3′ (Figure 1e) flanking probes. All four clones were injected into host NIH Black Swiss (Tac:N:NIHS-BCfBR) × C57BL/6J F1 blastocysts, and two of the four clones generated chimeras that transmitted the mutation to their offspring (Figure 1e). The mutant phenotype was identical in mice derived from both original targeted ES cell lines; results from both lines have been pooled in this study. Sequences of PCR primers used for genotyping were: #1, 5′-GAAGGAGGCATTGAGACAAGC-3′; #2, 5′-AATGGGCTGACCGCTTCCTCGTG-3′; and #3, 5′-CAAAGGATCAGCTGAAGAGAGAC-3′.
Western blot analysis
Mouse embryos and adult brain were solubilized and 10 µg protein per lane was separated by SDS–PAGE and transferred to nitrocellulose. The separated proteins were probed with rabbit polyclonal antibody against AF-6 [15], a gift from Kozo Kaibuchi.
Histology, in situ hybridization and immunohistochemistry Embryos were fixed for 2 h in 4% paraformaldehyde and embedded in paraffin wax. Serial 5 µm sections from representative blocks were collected on Superfrost Plus slides (Fisher Scientific) and stained with hematoxylin/eosin using standard procedures. Whole-mount in situ hybridization analysis using digoxygenin-labeled probes was performed as described by Wilkinson et al. [31], using Hesx1 [19], Hex [20], HNF-3β [16] and T [31]. Stained embryos were cleared by incubation in a mixture of benzyl benzoate and benzyl alcohol (2:1) after dehydration in methanol. A minimum of six wild-type and six mutant embryos were used for each probe. Immunofluorescent localization was performed with rabbit polyclonal antibodies against ZO-1 (Zymed) and AF-6 [15]; mouse monoclonal antibodies against AF-6 (Transduction Laboratories) and E-cadherin (Transduction Laboratories); and rat monoclonal antibodies against ZO-1 (Chemicon) and prominin [21]. Embryo sections were examined with a BioRad MicroRadiance laser confocal unit attached to a Nikon Optiphot microscope.
Transmission electron microscopy Whole embryos were fixed by immersion in Karnovsky’s fixative at 4°C overnight and treated in 1% OsO4 in 0.1 M sodium cacodylate buffer for 1 h. During dehydration, embryos were stained in uranyl acetate for 2 h. Semi-thin and ultra-thin sections were stained with toluidine blue and lead citrate, respectively; the former were examined by light microscopy (Figure 4) and the latter were examined on a JEOL 100C × electron microscope (Figure 5). At least two embryos of each genotype were used in each experiment.
Supplementary material Supplementary material including three-dimensional representations of Figure 7e,f is available at http://current-biology.com/supmat/supmatin.htm.
Acknowledgements The authors dedicate this manuscript to the memory of Alexander (Sasha) Zhadanov, who died suddenly and unexpectedly on 12 April 1999 of pulmonary emboli, and Georgean Wald, who died on 19 April 1999 after a long battle with breast cancer. Sasha and Georgean were fine friends and colleagues, and we will miss them dearly. We thank J. Rossant and R. Beddington for probes; W. Huttner for the anti-prominin antibody; O. Sundin, L. Van Aelst and K. Kaibuchi for both antibodies and helpful criticism; K. Michaels, C. Silan and D. Corcoran for technical assistance; and G. Carlson, W. Crain and J. Bermingham for reviewing the manuscript. A.B.Z. was a postdoctoral fellow of the Northwest Region of the American Heart Association and J.A.M. is an Established Investigator of the American Heart Association.
References 1. Gumbiner BM: Cell adhesion: the molecular basis of tissue architecture and morphogenesis. Cell 1996, 84:345-357. 2. Ponting CP, Phillips C, Davies KE, Blake DJ: PDZ domains: targeting signalling molecules to sub-membranous sites. Bioessays 1997, 19:469-479.
887
3. Prasad R, Gu Y, Alder H, Nakamura T, Canaani O, Saito H, et al.: Cloning of the ALL-1 fusion partner, the AF-6 gene, involved in acute myeloid leukemias with the t(6;11) chromosome translocation. Cancer Res 1993, 53:5624-5628. 4. Van Aelst L, White MA, Wigler MH: Ras partners. Cold Spring Harb Symp Quant Biol 1994, 59:181-186. 5. Ponting C, Phillips C: DHR domains in syntrophins, neuronal NO synthases and other intracellular proteins. Trends Biochem Sci 1995, 20:102-103. 6. Kuriyama M, Harada N, Kuroda S, Yamamoto T, Nakafuku M, Iwamatsu A, et al.: Identification of AF-6 and canoe as putative targets for Ras. J Biol Chem 1996, 271:607-610. 7. Ponting CP, Benjamin DR: A novel family of Ras-binding domains. Trends Biochem Sci 1996, 21:422-425. 8. Ponting CP: AF-6/cno: neither a kinesin nor a myosin, but a bit of both. Trends Biochem Sci 1995, 20:265-266. 9. Mandai K, Nakanishi H, Satoh A, Obaishi H, Wada M, Nishioka H, et al.: A novel actin filament-binding protein with one PDZ domain localized at cadherin-based cell-to-cell adherens junction. J Cell Biol 1997, 139:517-528. 10. Miyamoto H, Nihonmatsu I, Kondo S, Ueda R, Togashi S, Hirata K, et al.: Canoe encodes a novel protein containing a GLGF/DHR motif and functions with Notch and scabrous in common developmental pathways in Drosophila. Genes Dev 1995, 9:612-615. 11. Matsuo T, Takahashi K, Kondo S, Kaibuchi K, Yamamoto D: Regulation of cone cell formation by Canoe and Ras in the developing Drosophila eye. Development 1997, 124:2671-2680. 12. Takahisa M, Togashi S, Suzuki T, Kobayashi M, Murayama A, Kondo K, et al.: The Drosophila tamou gene, a component of the activating pathway of extramacrochaetae expression, encodes a protein homologous to mammalian cell-cell junction-associated protein ZO-1. Genes Dev 1996, 10:1783-1795. 13. Takahashi K, Matsuo T, Katsube T, Ueda R, Yamamoto D: Direct binding between two PDZ domain proteins Canoe and ZO-1 and their roles in regulation of the jun N-terminal kinase pathway in Drosophila morphogenesis. Mech Dev 1998, 78:97-111. 14. Sternberg PW, Alberola-Ila J: Conspiracy theory: RAS and RAF do not act alone. Cell 1998, 95:447-450. 15. Yamamoto T, Harada N, Kano K, Taya S, Canaani E, Matsuura Y: The Ras target AF-6 interacts with ZO-1 and serves as a peripheral component of tight junctions in epithelial cells. J Cell Biol 1997, 139:785-795. 16. Ang SL, Rossant J: HNF-3 beta is essential for node and notochord formation in mouse development. Cell 1994, 78:561-574. 17. Sasaki H, Hogan BL: HNF-3 beta as a regulator of floor plate development. Cell 1994, 76:103-115. 18. Kispert A, Herrmann BG: Immunohistochemical analysis of the Brachyury protein in wild-type and mutant mouse embryos. Dev Biol 1994, 161:179-193. 19. Thomas P, Beddington R: Anterior primitive endoderm may be responsible for patterning the anterior neural plate in the mouse embryo. Curr Biol 1996, 6:1487-1496. 20. Thomas PQ, Brown A, Beddington RS: Hex: a homeobox gene revealing peri-implantation asymmetry in the mouse embryo and an early transient marker of endothelial cell precursors. Development 1998, 125:85-94. 21. Weigmann A, Corbeil D, Hellwig A, Huttner WB: Prominin, a novel microvilli-specific polytopic membrane protein of the apical surface of epithelial cells, is targeted to plasmalemmal protrusions of non-epithelial cells. Proc Natl Acad Sci USA 1997, 94:12425-12430. 22. Collins JE, Fleming TP: Epithelial differentiation in the mouse preimplantation embryo: making adhesive cell contacts for the first time. Trends Biochem Sci 1995, 20:307-312. 23. Birchmeier W, Behrens J, Weidner KM, Hulsken J, Birchmeier C: Epithelial differentiation and the control of metastasis in carcinomas. Curr Top Microbiol Immunol 1996, 213:117-135. 24. Kintner C: Regulation of embryonic cell adhesion by the cadherin cytoplasmic domain. Cell 1992, 69:225-236. 25. Levine E, Lee CH, Kintner C, Gumbiner BM: Selective disruption of E-cadherin function in early Xenopus embryos by a dominant negative mutant. Development 1994, 120:901-909. 26. Aaku-Saraste E, Oback B, Hellwig A, Huttner WB: Neuroepithelial cells downregulate their plasma membrane polarity prior to neural tube closure and neurogenesis. Mech Dev 1997, 69:71-81.
888
Current Biology, Vol 9 No 16
27. Aaku-Saraste E, Hellwig A, Huttner WB: Loss of occludin and functional tight junctions, but not ZO-1, during neural tube closure—remodeling of the neuroepithelium prior to neurogenesis. Dev Biol 1996, 180:664-679. 28. Wang AZ, Ojakian GK, Nelson WJ: Steps in the morphogenesis of a polarized epithelium. I. Uncoupling the roles of cell-cell and cell-substratum contact in establishing plasma membrane polarity in multicellular epithelial (MDCK) cysts. J Cell Sci 1990, 95:137-151. 29. Tybulewicz V, Crawford C, Jackson P, Bronson R, Mulligan R: Neonatal lethality and lymphopenia in mice with a homozygous disruption of the c-abl proto-oncogene. Cell 1991, 65:1153-1163. 30. Nagy A, Rossant J, Nagy R, Abramow-Newerly W, Roder JC: Derivation of completely cell culture-derived mice from earlypassage embryonic stem cells. Proc Natl Acad Sci USA 1993, 90:8424-8428. 31. Wilkinson DG, Bhatt S, Herrmann BG: Expression pattern of the mouse T gene and its role in mesoderm formation. Nature 1990, 343:657-659.
Because Current Biology operates a ‘Continuous Publication System’ for Research Papers, this paper has been published on the internet before being printed. The paper can be accessed from http://biomednet.com/cbiology/cub — for further information, see the explanation on the contents page.
S1
Supplementary material Absence of the tight junctional protein AF-6 disrupts epithelial cell–cell junctions and cell polarity during mouse development
Alexander B. Zhadanov, D. William Provance Jr., C.A. Speer, J. Douglas Coffin, Dee Goss, J.A. Blixt, Cheryl M. Reichert and John A. Mercer Current Biology 5 August 1999, 9:880–888 Figure S1 (a)
ve
e
ps/m (b)
ve
e
ps/m Current Biology
Semithin (1 µm) sections from 7.5 dpc embryos fixed and embedded for electron microscopy. Sections from (a) wild-type and (b) mutant embryos were stained with toluidine blue: e, embryonic ectoderm; ps/m, primitive streak/mesoderm; ve, visceral endoderm. The extraembryonic portions of the embryos are at the right in this figure and in Figures S2 and S3.
Figure S2 A stereo pair of a larger thickness of the wildtype embryo shown in Figure 7e, demonstrating the planar apical staining for prominin. Localization of E-cadherin is shown in red and prominin in green. Twenty-four sections in a Z-series were stacked and projected. The data from this figure are shown as a rotating stereo movie at http://www.montana.edu/wwwmri/wt.mov.
S2
Supplementary material
Figure S3 A stereo pair of a larger thickness of the Af6 mutant embryo shown in Figure 7f, showing localization of E-cadherin (red) and prominin (green). Much of the prominin staining is clearly basolateral, although in some regions it remains apical. Twenty-eight sections in a Z-series were stacked and projected. The data from this figure are shown as a rotating stereo movie at http://www.montana.edu/wwwmri/mut.mov.
Research Paper
Absence of the tight junctional protein AF-6 disrupts epithelial cell–cell junctions and cell polarity during mouse development Alexander B. Zhadanov*†, D. William Provance Jr.*, C.A. Speer‡, J. Douglas Coffin*§, Dee Goss*, J.A. Blixt‡, Cheryl M. Reichert* and John A. Mercer* Background: The establishment, maintenance and rearrangement of junctions between epithelial cells are extremely important in many developmental, physiological and pathological processes. AF-6 is a putative Ras effector; it is also a component of tight and adherens junctions, and has been shown to bind both Ras and the tight-junction protein ZO-1. In the mouse, AF-6 is encoded by the Af6 gene. As cell–cell junctions are important in morphogenesis, we generated a null mutation in the murine Af6 locus to test the hypothesis that lack of AF-6 function would cause epithelial abnormalities.
Addresses: *McLaughlin Research Institute, 1520 23rd Street South, Great Falls, Montana 594054900, USA. ‡Department of Veterinary Molecular Biology, Montana State University, Bozeman, Montana 59717-3610, USA. Present address: §Department of Pharmaceutical Sciences, University of Montana, Missoula, Montana 59812, USA. †Deceased.
Results: Although cell–cell junctions are thought to be important in early embryogenesis, homozygous mutant embryos were morphologically indistinguishable from wild-type embryos through 6.5 days post coitum (dpc) and were able to establish all three germ layers. The earliest morphological abnormalities were observed in the embryonic ectoderm of mutant embryos at 7.5 dpc. The length of the most apical cell–cell junctions was reduced, and basolateral surfaces of those cells were separated by multiple gaps. Cells of the embryonic ectoderm were less polarized as assessed by histological criteria and lateral localization of an apical marker. Mutant embryos died by 10 dpc, probably as a result of placental failure.
Correspondence: John A. Mercer E-mail: [email protected] Received: 8 February 1999 Revised: 5 July 1999 Accepted: 5 July 1999 Published: 5 August 1999 Current Biology 1999, 9:880–888 http://biomednet.com/elecref/0960982200900880 © Elsevier Science Ltd ISSN 0960-9822
Conclusions: AF-6 is a critical regulator of cell–cell junctions during mouse development. The loss of neuroepithelial polarity in mutants is consistent with a loss of efficacy of the cell–cell junctions that have a critical role in establishing apical/basolateral asymmetry.
Background The dynamic modulation of cell–cell contacts is a critical step in many developmental processes, including the establishment of epithelial cell polarity and complex movements of sheets of cells [1]. Proteins defined by the PDZ sequence motif — a protein interaction motif originally identified as repeated regions of homology between the synaptic protein PSD-95, the product of the Drosophila dlg tumor suppressor gene and the epithelial tight-junction protein ZO-1 — have been shown to function at many different sites of cell–cell interaction, including tight junctions, adherens junctions and synapses. It has been shown that PDZ domains participate in cell–cell junction organization, intracellular signaling and receptor clustering by mediating protein–protein interactions [2]. The human AF6 gene was first identified as a fusion partner with the ALL-1 locus in acute myeloid leukemias with a t(6;11)(q27;q23) translocation [3]. It was also isolated in a yeast two-hybrid screen with H-Ras [4]. The AF6 locus (Af6 in mouse) encodes a modular protein (AF-6 or afadin) that contains a PDZ domain [5]; two motifs that interact with Ras and the tight-junction protein ZO-1
in vitro [4,6,7]; regions with homology to myosin-V-like and kinesin-like cargo-binding domains [8]; and an actinbinding domain present only in a long isoform produced by alternative splicing [9]. From studies of the Drosophila homologs of AF-6, ZO-1 and Ras, we can predict multiple and complex roles for the mammalian proteins in cell–cell interactions during development. The canoe (cno) gene is a Drosophila homolog of Af6 [10,8]. Genetic manipulation of the activity of Cno modulated the effects of both constitutively active and dominantnegative forms of Ras1 in cone cell differentiation [11]. The tamou (tam) locus of Drosophila encodes a protein homologous to ZO-1 [12]. Flies doubly homozygous for the viable hypomorphic alleles pydtam and cnomis1 die as embryos, suggesting a genetic interaction between tam and cno [13]. Of the known Ras effectors [14], AF-6 is the most likely candidate for directly mediating the effects of Ras (or other small GTPases) on cell–cell junctions. The AF-6 protein colocalizes with ZO-1 at tight junctions in epithelial cells [15] and with cadherin at adherens junctions in both rat liver and intestinal epithelium [9]. Overexpression
Research Paper Neuroectodermal abnormalities in Af6 mutant embryos Zhadanov et al.
of activated Ras protein in Rat1 fibroblasts causes a perturbation of cell–cell contacts, with a decreased accumulation of ZO-1 and AF-6 at the cell surface [15]. Together with in vitro binding data, these data suggest that the interaction of AF-6 with ZO-1 is required for the assembly and dynamic modulation of cell–cell junctions, and that this interaction is inhibited by activated Ras. Here, we present evidence that AF-6 function is required for maintenance of cell–cell junctions and cell
881
polarity in the neuroepithelium of embryos at 7.5 days post coitum (dpc).
Results Loss of AF-6 function leads to embryonic lethality by 10.5 dpc
AF-6 protein was detected in blastocysts and throughout development in wild-type embryos. The protein was widely distributed in 8.75 and 9.5 dpc embryos, with strongest signals in the tail bud, epithelial cells of branchial
Figure 1
8.75 dpc +/+
m
–7.9 kb
wt
–6.1 kb
–
–
+/
+/
#53
+/
+
–
#17
+/
– +/
#4
+/ + +/ +
– +/ + +/ +
m wt
–15.1 kb –13.3 kb
–
+
+/
+
+/
+
+/
–
+/
+
+/
+/
–
+/
–
(e) +/
9.5 dpc –/–
(d)
#53
+/
#17
+/ + +/ +
(c)
9.5 dpc +/+
+/ + +/ +
(a)
ES kbp –7.9 kb –6.1 kb
m wt Intercross F2
H
E H
6.1 kb
13.5 kb
M
Backcross N2 M
+/ + +/ – +/ – +/ – +/ + +/ + +/ +
E
(b)
+/ – +/ – +/ – +/ + +/ – +/ – +/ +
(f)
S E E
A
EH 3′
1 3 neo
tk N X Targeting construct
E E
m
650 bp
wt
353 bp
AN (g) M
5′ probe A H E X Targeted Af6 allele
C C
Parents
3′ probe B neo
E E E
H
–/ – +/ – +/ – +/ – +/ + +/ – +/ + +/ + +/ – +/ + +/ – +/ + +/ – –/ –
H E X 5′ Wild-type Af6 locus
15.3 kb
1
2 3 7.9 kb
A
EH E H
m
650 bp
wt
353 bp
Current Biology
AF-6 expression, targeting vector, homologous recombination and germline transmission. (a) AF-6 protein expression at 8.75 and 9.5 dpc in wild-type (+/+) and homozygous mutant (–/–) embryos. High levels of AF-6 were detected in the tail bud (black arrow), epithelial cells of branchial arches (black arrowhead) and neural tube (white arrow). (b) Targeting strategy. Maps of the wild-type Af6 locus (top), the targeting vector (middle) and the disrupted Af6 allele (bottom) are shown. Black boxes, Af6 exons; black bars, external 5′ and 3′ probes used for Southern blot analysis; neo, neomycin resistance selectable marker (with the direction of transcription indicated as an arrow pointing in the 5′ to 3′ direction); tk, thymidine kinase gene (with the direction of transcription indicated as an arrow pointing in the 5′ to 3′ direction); grey boxes, promoter derived from the gene for phosphoglycerate kinase
(PGK); arrows (1–3), PCR primers for genotyping; A, ApaI; E, EcoRI; H, HindIII; N, NotI; S, StuI; X, XhoI. EcoRI and HindIII fragments (with sizes in kb) diagnostic of the wild-type and targeted alleles are shown, respectively, above the map of the wild-type locus and below the map of the targeted allele. (c,d) Southern blot analysis showing homologous recombination at the Af6 locus in ES cell clones 4, 17, and 53 using (c) the 3′ probe B on EcoRI-digested DNA, and (d) the 5′ probe A on HindIII-digested DNA. (e) Southern blot analysis demonstrating germline transmission of the targeted Af6 allele in the progeny of a chimera for ES cell clone 17 using the 3′ probe B on EcoRI-digested genomic DNA. (f,g) PCR genotyping of 9.5 dpc embryos. Lanes labelled C in (g) denote no DNA control samples. In (c–g), diagnostic fragments for the wild type (wt) and the mutant (m) are indicated.
882
Current Biology, Vol 9 No 16
Table 1
we used PCR to genotype embryos from intercrosses of heterozygotes (Table 1). Partially resorbed embryos were observed at 10.5 and 11.5 dpc and were found on genotyping to be homozygotes (Figure 1f,g). At earlier times, homozygous mutants were present at the expected Mendelian proportion (Table 1).
PCR genotyping of progeny derived from intercrosses between Af6/+ heterozygotes. Age of progeny
+/+
+/–
–/–
Liveborn
41
73
0
13.5 dpc
3
6
0
11.5 dpc
2
5
2 (resorptions)
10.5 dpc
5
9
4 (resorptions)
9.5 dpc
13
31
13
8.5 dpc
11
16
9
7.5 dpc
12
26
11
To confirm the absence of AF-6 protein in mutants, mutant and wild-type embryos were analyzed by wholeembryo immunohistochemistry (Figure 1a) and western blot analysis (data not shown). No AF-6 protein was detected in mutant homozygous by either method. The earliest abnormality in mutant embryos is a hyperplastic embryonic ectoderm
To characterize the abnormalities of Af6 homozygous mutant embryos, we analyzed litters from intercrosses of heterozygous mice at stages between 5.5 and 9.5 dpc. The mutant phenotype was identical in mice derived from both targeted ES cell lines, and so results from both lines have been pooled in this study. All embryos were genotyped by PCR.
arches, neural tube, and mid/hindbrain neuroepithelium at 9.5 dpc (Figure 1a). We generated a null mutation at the Af6 locus by homologous recombination in embryonic stem (ES) cells by inserting the PGK–neo cassette into exon 4 of the Af6 locus (Figure 1b). Two independently derived ES cell clones (Figure 1c,d) generated chimeras that transmitted the mutant allele in the germ line (Figure 1e). Mice heterozygous for the Af6 mutation appeared normal and were fertile.
No morphological abnormalities were observed before 6.5 dpc (data not shown). By 7.5 dpc, homozygous mutant embryos were morphologically distinguishable from normal embryos (Figure 2a). Mutant conceptuses showed a reduction in the size of the embryonic portion relative to the extraembryonic portion; posterior body elongation was retarded and neural folds were not observed (Figure 2a). In wild-type embryos, the neural plate was clearly defined anteriorly, with developing neural folds (Figure 2a, NF).
No liveborn homozygous mutants were found among the progeny of matings between heterozygotes (Figure 1e). To determine when Af6 homozygous null embryos died, Figure 2 (a)
(b)
EX
EX
+/+
(c)
–/–
D EM NF
EM
EX
+/+
NF
+/+
–/–
7.5 dpc (e)
(d)
(f)
C
C C
AL
A
EX
EX
EM
–/– +/+ EM 8.5 dpc (g) C AL
9.5 dpc Y (i)
(h)
E
N NE
A
A
S E
PS +/+
A
E
E
E PS –/–
S +/+
–/–
PS +/+
–/– Current Biology
Morphology of wild-type (+/+) and Af6 mutant (–/–) embryos at 7.5, 8.5 and 9.5 dpc. (a) Embryos dissected from the decidua at 7.5 dpc. Normal embryos exhibit AP polarity; the developing neural folds mark the anterior aspect of embryos. (b) Dissected wild-type embryo and a partially dissected mutant embryo at 8.5 dpc. (c) Wild-type and a typical comma-shaped mutant embryo at 9.5 dpc. (d–i) Hematoxylin–eosin-stained sections of wild-type and mutant embryos at (d,e) 7.5 dpc, (f,g) 8.5 dpc and (h,i) 9.5 dpc. EM, embryonic portion of the embryo; EX, extraembryonic portion of the embryo; A, amnion; AL, allantois; C, chorion; D, decidua; E, embryonic ectoderm; N, neural tube; NE, neuroepithelium; NF, neural fold; PS, primitive streak; S, somites; Y, yolk sac. The arrowheads in (e) denote accumulations of (upper) mesoderm and (lower) ectoderm. The bars represent 100 µm in (a,b,c,g,h,i); 50 µm in (d,e); and 200 µm in (f). All embryos shown were genotyped by (a–c) PCR or (d–i) immunohistochemistry.
Research Paper Neuroectodermal abnormalities in Af6 mutant embryos Zhadanov et al.
883
Figure 3 Whole-mount in situ hybridization analysis of T (Brachyury) and Hesx1 expression. (a) T expression at the posterior aspect of the embryo in primitive streak mesodermal cells (arrows) in both wild-type and mutant embryos at 7.5 dpc. (b) Normal T expression (white arrows) at 8.5 dpc in wild-type embryos, and aberrant (but asymmetric) expression (white arrowhead) in mutants at the same stage. (c) Hesx1 expression in anterior visceral endoderm (arrows) at 7.0 dpc marks the anterior aspect of both wild-type and mutant embryos. The bars represent 100 µm.
(a)
(b)
(c) +/+
+/+ T (Brachyury)
+/+
+/+
–/–
Sectioning of 7.5 dpc embryos revealed that the mutants formed all three germ layers, as mesoderm had clearly passed through the primitive streak region (Figure 2d,e). The embryonic ectoderm, however, which at the midline will become the neuroepithelium of the neural plate, was more stratified than columnar (Figure 2e, lower arrowhead; Figure 4). The neuroepithelium was present in mutants, but it was poorly developed at the anterior and posterior of the embryo (Figure 2d,e). The allantois, which forms from a projection of mesodermal cells from the posterior margin of the embryonic ectoderm, was present in normal embryos at this stage. Mutant embryos lacked an allantois; in its place, we observed a posterior accumulation of mesodermal cells in the mutant embryo (Figure 2e, upper arrowhead). By 8.0–8.5 dpc, the presumptive neuroectodermal layer in mutant embryos became thick and disorganized centrally, while the anterior and posterior margins of the layer were thin. Mesodermal structures such as notochord and somites were not observed in 8.5 dpc mutant embryos (Figure 2b,f,g), and expression of a notochord marker, HNF-3β [16,17], was not detected (data not shown).
Hesx1
–/–
T (Brachyury) –/–
+/+ –/–
–/– Current Biology
formation. The T gene (also known as Brachyury) is transiently expressed from the onset of gastrulation in cells ingressing into the primitive streak and in nascent mesoderm [18]. Mutant embryos expressed T transcripts at 7.5 dpc, suggesting the presence of a primitive streak (Figure 3a) despite the abnormal morphology of the mutant mesoderm. By the next day of development, however, expression of T was confined to small patches of cells, possibly corresponding to axial mesoderm (Figure 3b, white arrowhead). T expression was nearly absent by 9.5 dpc in mutant embryos (data not shown). The expression of the homeobox gene Hesx1 (also known as Rpx) marks the anterior aspect of the embryo [19]. Mutant embryos expressed Hesx1 in anterior visceral endoderm at 7.0 dpc, indicating that AP polarity was established despite morphogenetic abnormalities (Figure 3c). By the next day, Hesx1 transcripts are normally localized to rostral neuroectoderm [19]; Hesx1 expression, however, was not detected in 8.5 dpc mutant embryos (data not shown), consistent with the absence of a head fold in mutant embryos. The earliest known anterior marker, Hex Figure 4
Mutant embryos were comma-shaped at 9.5 dpc and lacked recognizable normal structures (Figure 2c,h,i). Histological examination showed that the 9.5 dpc mutant embryos consisted predominantly of loose mesenchyme (Figure 2i). The death of mutant embryos by 10 dpc is likely to be the result of failure to establish any connection between the embryonic and maternal circulatory systems (data not shown), but the developmental abnormalities are so extensive that many alternative explanations are possible.
(a) +/+
(b) –/–
E
M
VE
Mutant embryos asymmetrically express early markers for anteroposterior axis establishment
Although histology provided evidence of morphogenetic failure, it did not indicate whether early differentiation events were arrested as well. Therefore, we examined early markers associated with anteroposterior (AP) axis
Current Biology
Semi-thin (1 µm) sections from 7.5 dpc embryos fixed and embedded for electron microscopy. Sections from (a) wild-type and (b) mutant embryos were stained in toluidine blue. E, embryonic ectoderm; M, mesoderm; VE, visceral endoderm.
884
Current Biology, Vol 9 No 16
[20], also was expressed asymmetrically in mutant embryos (data not shown). Abnormal morphology of apical cell–cell junctions and basolateral intercellular gaps in embryonic ectoderm
From the gross phenotype, we hypothesized that the cell–cell junctions in mutant embryonic ectoderm might be weakened or absent, and therefore unable to maintain the integrity of the ectoderm during growth and movement of the epithelial sheet. To examine cell–cell junction structure in the absence of AF-6, embryos were fixed and embedded for electron microscopy. Initial 1 µm sections examined by light microscopy (Figure 4) showed that the mutant embryonic ectoderm was more stratified and less columnar, and the mesoderm was less dense than the corresponding germ layers in wild-type embryos. No significant differences were observed between wild-type and mutant visceral endoderm at this level of resolution. We then examined all three germ layers of 7.5 dpc embryos by transmission electron microscopy (Figure 5). The most striking difference between wild-type and mutant embryonic neuroectoderm was the presence of multiple gaps between ectodermal cells of the lumenal layer in the mutant embryo (Figure 5d, arrows). In addition, structures resembling microvilli were present in the intercellular gaps (Figure 5d, arrowheads). Higher magnification views of apical junctions between neuroectodermal cells are shown in Figure 5b,e; the wild-type neuroectoderm had classical tight and adherens junctions
(Figure 5b), but junctions of mutant neuroectoderm had consistently shorter electron-dense regions that could not be classified as tight or adherens junctions. These electron-dense regions were frequently displaced basally (Figure 5e). The differences between mutant and wildtype embryonic endoderm (Figure 5c,f) were less striking than those seen in ectoderm, correlating with the absence of endodermal abnormalities at the light microscopic level in the mutant embryos. AF-6 colocalizes with both ZO-1 and E-cadherin in the neuroectoderm of 7.5 dpc embryos
Yamamoto et al. [15] observed that AF-6 colocalizes with ZO-1, and Mandai et al. [9] observed that AF-6 colocalizes with E-cadherin, showing more overlap with E-cadherin than with ZO-1. Initially, we used laser confocal microscopy to determine whether these associations were present in wild-type embryos at 7.5 dpc. We observed extensive overlap in the subcellular distribution of AF-6 with both ZO-1 (Figure 6a) and E-cadherin (Figure 6b) in embryonic ectoderm and visceral endoderm. These overlaps were present in blastocysts as well (data not shown). No detectable changes in the subcellular localization of ZO-1 and E-cadherin were found in Af6 mutant embryos
Because it was difficult to classify the abnormal cell–cell junctions in mutant neuroectoderm, and as AF-6 colocalized with components of both tight and adherens junctions, we examined the distribution of molecular markers for tight and adherens junctions in 7.5 dpc mutant
Figure 5
(a)
+/+
(b)
(c)
(d)
–/–
(e)
(f)
Transmission electron micrographs of (a,b,d,e) embryonic ectoderm and (c,f) visceral endoderm of (a–c) wild-type and (d–f) homozygous mutant embryos at 7.5 dpc. The apical surface is on the right in (a,d), and at the top in (b,c,e,f). (c,d) Extended cellular processes (arrowheads in panel d) are found at the lateral surface of the mutant cells. (d) Multiple gaps (arrows) are present between cells of the mutant; gaps at the center of (d) are enlarged in the inset. The magnification is 3,600 × in (a,d); 45,000 × in (b,e); and 56,000 × in (c,f).
Current Biology
Research Paper Neuroectodermal abnormalities in Af6 mutant embryos Zhadanov et al.
Figure 6
885
Figure 7 +/+
(b)
M
–/–
(a)
NE
(b)
ZO-1
(a)
NE
NE M
NE
M 50 µm
M
VE
(c)
(d)
E-cadherin
NE 20 µm M Current Biology
NE
M
VE Colocalization of AF-6 (red) with (a) ZO-1 (green) and (b) E-cadherin (green) in neuroectoderm (NE) of wild-type 7.5 dpc embryos. M, mesoderm. (e)
embryos by confocal microscopy. An antibody against AF-6 was used to identify mutant embryos (data not shown). As the results of Yamamoto et al. [15] suggested that activated Ras reduces cell–cell contacts in Rat1 cells by removing AF-6 and ZO-1 from the cell membrane, we hypothesized that ZO-1 localization would be changed in mutant embryonic ectoderm. Surprisingly, ZO-1 localization was not altered in the mutant embryos (Figure 7a,b), indicating that ZO-1 does not depend on AF-6 for its submembrane localization in this tissue. Localization of the adherens junction marker E-cadherin was also not significantly different in the neuroepithelia of mutant embryos, compared with the wild type (Figure 7c,d). The localization of β-catenin was similar to that of E-cadherin (data not shown). Both ZO-1 and E-cadherin colocalized with AF-6 in wild-type blastocysts, but the localization of either protein was not changed in mutant blastocysts (data not shown). In our electron micrographs (Figure 5), it appeared that microvilli were present in the basolateral gaps between mutant neuroepithelial cells. To confirm this identification and the reduction in polarity with a molecular marker, we performed immunofluorescent localization using a monoclonal antibody to prominin, an apical marker found in microvilli of neuroepithelium and many other epithelia [21]. The change in prominin localization was dramatic (Figure 7e,f). As shown in Figure 7e, in the wild-type neuroectoderm, prominin (green) had minimal overlap with E-cadherin (red); in mutant embryos, however, there was significant basolateral colocalization of prominin and
NE
M VE
E-cadherin (red)/prominin (green)
VE (f)
NE
M VE Current Biology
ZO-1 and E-cadherin localization is normal in 7.5 dpc Af6 mutant embryos, but the mutant neuroepithelium has lost its polarity. (a,b) Localization of ZO-1 in (a) wild-type and (b) Af6 mutant embryos. (c,d) Localization of E-cadherin in (c) wild-type and (d) Af6 mutant embryos. (e,f) Localization of E-cadherin (red) and prominin (green) in (e) wild-type and (f) Af6 mutant embryos. NE, neuroectoderm; M, mesoderm; VE, visceral endoderm.
E-cadherin (Figure 7f, yellow). Three-dimensional representations of Figure 7e,f are published with this article on the internet (see Supplementary material). These data demonstrate that the polarity of neuroectoderm is dramatically reduced in the absence of AF-6 function.
Discussion We propose that AF-6 is necessary for the dynamic regulation of tight and/or adherens junctions. More importantly, our data indicate that AF-6 function is only necessary in specific cell types that do not correspond entirely with the cell types in which AF-6 is expressed. Not unexpectedly, the regulation of tight and adherens junction during postimplantation development is far more complicated
886
Current Biology, Vol 9 No 16
than has been suggested by studies of cultured cells and preimplantation embryos. The Af6 mutant phenotype is interesting in several respects. First, the surprising lack of abnormalities in preimplantation development suggests that AF-6 is not an essential structural component of either tight or adherens junctions, which (along with desmosomes) are prominent in early development [22]. Desmosomes do not appear to be relevant to the initial abnormalities, as we did not identify desmosomes in several hundred wild-type and mutant embryonic cells (data not shown). All of our data are consistent with the hypothesis that AF-6 has a regulatory, not a constitutive or structural, function, although these roles are not mutually exclusive and functional redundancy is an alternative explanation. Second, the loss of columnar epithelial morphology in mutant neuroectoderm resembles hyperplasia, consistent with a role for AF-6 as a Ras effector [4] that could participate in the epithelial–mesenchymal transition [23] in carcinomas. This predicts that transformed Af6 homozygous mutant cells may be more invasive or metastatic than wildtype cells. Electron microscopy (Figure 5) showed that mutant neuroectoderm had electron-dense cell–cell contacts that were not identifiable as tight or adherens junctions. The large intercellular gaps observed are consistent with either reduced strength of cell–cell junctions or alterations in transepithelial transport, which also might be predicted from the loss of polarity that we observed. Similar intercellular gaps have been observed in Xenopus embryos following injection of dominant-negative N-cadherin [24] and E-cadherin [25]. It is important to note that the gaps are present along the entire lateral surfaces of the cells, not just apically. This is surprising given the normal apical location of AF-6 (Figure 6). A detailed immunoelectron microscopic characterization of cell–cell junctions in blastocysts, embryoid bodies and homozygous mutant ES cells is now in progress. In a developmental context, the tissue specificity of the Af6 mutant phenotype is consistent with results from other investigators, as wild-type neuroepithelium has been shown to reorganize and lose polarity between 8 and 9 dpc [26]. During this time, expression of occludin is decreasing and expression of ZO-1 is increasing in chick embryos [27]. These results can be integrated by hypothesizing that the epithelial character of these cells is beginning to be lost at 7.0 dpc, and that AF-6 may represent a pivotal mechanism for modulating cell–cell junctions downstream of Ras or other small GTPases during this period. Any roles of AF-6 in later morphogenetic events are likely to be masked by the severe abnormalities caused by the initial disruption at 7.5 dpc. There is some controversy over the identification of AF-6 as a component of tight junctions [15] or (as afadin) as a
component of adherens junctions [9]. Our demonstration of colocalization of both ZO-1 and E-cadherin with AF-6 in wild-type neuroectoderm suggests that there may be no conflict between these two results. All three analyses were performed under different conditions and therefore any conclusions derived from them are clearly not mutually exclusive. Moreover, the abnormalities we observed in neuroepithelial cell–cell junctions defy description in classical terms and do not help to distinguish between a role of AF-6 in tight versus adherens junction regulation. At this point, our data are consistent with both hypotheses. Our failure to detect any change in localization of ZO-1 in the mutant embryos was not expected, however, given the AF-6–ZO-1 interactions demonstrated in vitro by Yamamoto et al. [15]. Despite the apparent discrepancy between the effects they observed upon overexpression of Ras and the effects we observed upon removal of AF-6, both suggest a regulatory role for AF-6; hypothesizing a more structural function for AF-6 in directing the localization of ZO-1 would predict mislocalization of ZO-1 in the mutant. The striking loss of neuroepithelial polarity in the Af6 mutant embryos is consistent with data from other systems [28], which suggest a role for cell–cell junctions in directing polarity. Although we observed lateral localization of both apical structures and an apical marker, the basolateral localization of E-cadherin itself was not perceptibly changed in the mutant embryos. Whether this loss in polarity is a downstream effect of the alterations in cell–cell junctions or directly downstream of AF-6 cannot be determined. In summary, these data suggest that the regulation of cell–cell junctions during embryogenesis is tissue-specific, and that AF-6 is an important part of this mechanism in at least one important embryonic epithelium. More importantly, these data indicate that this regulation may be complex and indirect, as the mechanism of cell–cell junction disruption was not illuminated at the level of light microscopy by immunolocalization of the basic markers of tight and adherens junctions.
Materials and methods Generation of Af6 mutant mice A 412 bp cDNA probe was obtained by reverse-transcriptase PCR from mouse brain cDNA using oligonucleotide primers corresponding to the amino-acid sequences EKFRPDM and PETSFTR, which are conserved between human AF6 and Canoe. This cDNA probe was mapped (using the Jackson Laboratory BSS mapping panel) to proximal mouse chromosome 17, which corresponds to human 6q27. The probe was used to screen a 129/SvJ genomic bacterial artificial chromosome library (Genome Systems). A 6.2 kb XhoI–ApaI Af6 genomic fragment was subcloned into pGEM11 (Figure 1b). The neo expression cassette from pPNT [29] was subcloned into a StuI site. This Af6–neo fragment was inserted into pPNT. The linearized targeting vector was electroporated into R1 ES cells [30], which were subjected to double selection with G418 and gancyclovir. Four independent homologous recombination events occurred among 75 doubly resistant colonies, as assessed by Southern blot hybridization with both 5′
Research Paper Neuroectodermal abnormalities in Af6 mutant embryos Zhadanov et al.
(Figure 1d) and 3′ (Figure 1e) flanking probes. All four clones were injected into host NIH Black Swiss (Tac:N:NIHS-BCfBR) × C57BL/6J F1 blastocysts, and two of the four clones generated chimeras that transmitted the mutation to their offspring (Figure 1e). The mutant phenotype was identical in mice derived from both original targeted ES cell lines; results from both lines have been pooled in this study. Sequences of PCR primers used for genotyping were: #1, 5′-GAAGGAGGCATTGAGACAAGC-3′; #2, 5′-AATGGGCTGACCGCTTCCTCGTG-3′; and #3, 5′-CAAAGGATCAGCTGAAGAGAGAC-3′.
Western blot analysis
Mouse embryos and adult brain were solubilized and 10 µg protein per lane was separated by SDS–PAGE and transferred to nitrocellulose. The separated proteins were probed with rabbit polyclonal antibody against AF-6 [15], a gift from Kozo Kaibuchi.
Histology, in situ hybridization and immunohistochemistry Embryos were fixed for 2 h in 4% paraformaldehyde and embedded in paraffin wax. Serial 5 µm sections from representative blocks were collected on Superfrost Plus slides (Fisher Scientific) and stained with hematoxylin/eosin using standard procedures. Whole-mount in situ hybridization analysis using digoxygenin-labeled probes was performed as described by Wilkinson et al. [31], using Hesx1 [19], Hex [20], HNF-3β [16] and T [31]. Stained embryos were cleared by incubation in a mixture of benzyl benzoate and benzyl alcohol (2:1) after dehydration in methanol. A minimum of six wild-type and six mutant embryos were used for each probe. Immunofluorescent localization was performed with rabbit polyclonal antibodies against ZO-1 (Zymed) and AF-6 [15]; mouse monoclonal antibodies against AF-6 (Transduction Laboratories) and E-cadherin (Transduction Laboratories); and rat monoclonal antibodies against ZO-1 (Chemicon) and prominin [21]. Embryo sections were examined with a BioRad MicroRadiance laser confocal unit attached to a Nikon Optiphot microscope.
Transmission electron microscopy Whole embryos were fixed by immersion in Karnovsky’s fixative at 4°C overnight and treated in 1% OsO4 in 0.1 M sodium cacodylate buffer for 1 h. During dehydration, embryos were stained in uranyl acetate for 2 h. Semi-thin and ultra-thin sections were stained with toluidine blue and lead citrate, respectively; the former were examined by light microscopy (Figure 4) and the latter were examined on a JEOL 100C × electron microscope (Figure 5). At least two embryos of each genotype were used in each experiment.
Supplementary material Supplementary material including three-dimensional representations of Figure 7e,f is available at http://current-biology.com/supmat/supmatin.htm.
Acknowledgements The authors dedicate this manuscript to the memory of Alexander (Sasha) Zhadanov, who died suddenly and unexpectedly on 12 April 1999 of pulmonary emboli, and Georgean Wald, who died on 19 April 1999 after a long battle with breast cancer. Sasha and Georgean were fine friends and colleagues, and we will miss them dearly. We thank J. Rossant and R. Beddington for probes; W. Huttner for the anti-prominin antibody; O. Sundin, L. Van Aelst and K. Kaibuchi for both antibodies and helpful criticism; K. Michaels, C. Silan and D. Corcoran for technical assistance; and G. Carlson, W. Crain and J. Bermingham for reviewing the manuscript. A.B.Z. was a postdoctoral fellow of the Northwest Region of the American Heart Association and J.A.M. is an Established Investigator of the American Heart Association.
References 1. Gumbiner BM: Cell adhesion: the molecular basis of tissue architecture and morphogenesis. Cell 1996, 84:345-357. 2. Ponting CP, Phillips C, Davies KE, Blake DJ: PDZ domains: targeting signalling molecules to sub-membranous sites. Bioessays 1997, 19:469-479.
887
3. Prasad R, Gu Y, Alder H, Nakamura T, Canaani O, Saito H, et al.: Cloning of the ALL-1 fusion partner, the AF-6 gene, involved in acute myeloid leukemias with the t(6;11) chromosome translocation. Cancer Res 1993, 53:5624-5628. 4. Van Aelst L, White MA, Wigler MH: Ras partners. Cold Spring Harb Symp Quant Biol 1994, 59:181-186. 5. Ponting C, Phillips C: DHR domains in syntrophins, neuronal NO synthases and other intracellular proteins. Trends Biochem Sci 1995, 20:102-103. 6. Kuriyama M, Harada N, Kuroda S, Yamamoto T, Nakafuku M, Iwamatsu A, et al.: Identification of AF-6 and canoe as putative targets for Ras. J Biol Chem 1996, 271:607-610. 7. Ponting CP, Benjamin DR: A novel family of Ras-binding domains. Trends Biochem Sci 1996, 21:422-425. 8. Ponting CP: AF-6/cno: neither a kinesin nor a myosin, but a bit of both. Trends Biochem Sci 1995, 20:265-266. 9. Mandai K, Nakanishi H, Satoh A, Obaishi H, Wada M, Nishioka H, et al.: A novel actin filament-binding protein with one PDZ domain localized at cadherin-based cell-to-cell adherens junction. J Cell Biol 1997, 139:517-528. 10. Miyamoto H, Nihonmatsu I, Kondo S, Ueda R, Togashi S, Hirata K, et al.: Canoe encodes a novel protein containing a GLGF/DHR motif and functions with Notch and scabrous in common developmental pathways in Drosophila. Genes Dev 1995, 9:612-615. 11. Matsuo T, Takahashi K, Kondo S, Kaibuchi K, Yamamoto D: Regulation of cone cell formation by Canoe and Ras in the developing Drosophila eye. Development 1997, 124:2671-2680. 12. Takahisa M, Togashi S, Suzuki T, Kobayashi M, Murayama A, Kondo K, et al.: The Drosophila tamou gene, a component of the activating pathway of extramacrochaetae expression, encodes a protein homologous to mammalian cell-cell junction-associated protein ZO-1. Genes Dev 1996, 10:1783-1795. 13. Takahashi K, Matsuo T, Katsube T, Ueda R, Yamamoto D: Direct binding between two PDZ domain proteins Canoe and ZO-1 and their roles in regulation of the jun N-terminal kinase pathway in Drosophila morphogenesis. Mech Dev 1998, 78:97-111. 14. Sternberg PW, Alberola-Ila J: Conspiracy theory: RAS and RAF do not act alone. Cell 1998, 95:447-450. 15. Yamamoto T, Harada N, Kano K, Taya S, Canaani E, Matsuura Y: The Ras target AF-6 interacts with ZO-1 and serves as a peripheral component of tight junctions in epithelial cells. J Cell Biol 1997, 139:785-795. 16. Ang SL, Rossant J: HNF-3 beta is essential for node and notochord formation in mouse development. Cell 1994, 78:561-574. 17. Sasaki H, Hogan BL: HNF-3 beta as a regulator of floor plate development. Cell 1994, 76:103-115. 18. Kispert A, Herrmann BG: Immunohistochemical analysis of the Brachyury protein in wild-type and mutant mouse embryos. Dev Biol 1994, 161:179-193. 19. Thomas P, Beddington R: Anterior primitive endoderm may be responsible for patterning the anterior neural plate in the mouse embryo. Curr Biol 1996, 6:1487-1496. 20. Thomas PQ, Brown A, Beddington RS: Hex: a homeobox gene revealing peri-implantation asymmetry in the mouse embryo and an early transient marker of endothelial cell precursors. Development 1998, 125:85-94. 21. Weigmann A, Corbeil D, Hellwig A, Huttner WB: Prominin, a novel microvilli-specific polytopic membrane protein of the apical surface of epithelial cells, is targeted to plasmalemmal protrusions of non-epithelial cells. Proc Natl Acad Sci USA 1997, 94:12425-12430. 22. Collins JE, Fleming TP: Epithelial differentiation in the mouse preimplantation embryo: making adhesive cell contacts for the first time. Trends Biochem Sci 1995, 20:307-312. 23. Birchmeier W, Behrens J, Weidner KM, Hulsken J, Birchmeier C: Epithelial differentiation and the control of metastasis in carcinomas. Curr Top Microbiol Immunol 1996, 213:117-135. 24. Kintner C: Regulation of embryonic cell adhesion by the cadherin cytoplasmic domain. Cell 1992, 69:225-236. 25. Levine E, Lee CH, Kintner C, Gumbiner BM: Selective disruption of E-cadherin function in early Xenopus embryos by a dominant negative mutant. Development 1994, 120:901-909. 26. Aaku-Saraste E, Oback B, Hellwig A, Huttner WB: Neuroepithelial cells downregulate their plasma membrane polarity prior to neural tube closure and neurogenesis. Mech Dev 1997, 69:71-81.
888
Current Biology, Vol 9 No 16
27. Aaku-Saraste E, Hellwig A, Huttner WB: Loss of occludin and functional tight junctions, but not ZO-1, during neural tube closure—remodeling of the neuroepithelium prior to neurogenesis. Dev Biol 1996, 180:664-679. 28. Wang AZ, Ojakian GK, Nelson WJ: Steps in the morphogenesis of a polarized epithelium. I. Uncoupling the roles of cell-cell and cell-substratum contact in establishing plasma membrane polarity in multicellular epithelial (MDCK) cysts. J Cell Sci 1990, 95:137-151. 29. Tybulewicz V, Crawford C, Jackson P, Bronson R, Mulligan R: Neonatal lethality and lymphopenia in mice with a homozygous disruption of the c-abl proto-oncogene. Cell 1991, 65:1153-1163. 30. Nagy A, Rossant J, Nagy R, Abramow-Newerly W, Roder JC: Derivation of completely cell culture-derived mice from earlypassage embryonic stem cells. Proc Natl Acad Sci USA 1993, 90:8424-8428. 31. Wilkinson DG, Bhatt S, Herrmann BG: Expression pattern of the mouse T gene and its role in mesoderm formation. Nature 1990, 343:657-659.
Because Current Biology operates a ‘Continuous Publication System’ for Research Papers, this paper has been published on the internet before being printed. The paper can be accessed from http://biomednet.com/cbiology/cub — for further information, see the explanation on the contents page.
S1
Supplementary material Absence of the tight junctional protein AF-6 disrupts epithelial cell–cell junctions and cell polarity during mouse development
Alexander B. Zhadanov, D. William Provance Jr., C.A. Speer, J. Douglas Coffin, Dee Goss, J.A. Blixt, Cheryl M. Reichert and John A. Mercer Current Biology 5 August 1999, 9:880–888 Figure S1 (a)
ve
e
ps/m (b)
ve
e
ps/m Current Biology
Semithin (1 µm) sections from 7.5 dpc embryos fixed and embedded for electron microscopy. Sections from (a) wild-type and (b) mutant embryos were stained with toluidine blue: e, embryonic ectoderm; ps/m, primitive streak/mesoderm; ve, visceral endoderm. The extraembryonic portions of the embryos are at the right in this figure and in Figures S2 and S3.
Figure S2 A stereo pair of a larger thickness of the wildtype embryo shown in Figure 7e, demonstrating the planar apical staining for prominin. Localization of E-cadherin is shown in red and prominin in green. Twenty-four sections in a Z-series were stacked and projected. The data from this figure are shown as a rotating stereo movie at http://www.montana.edu/wwwmri/wt.mov.
S2
Supplementary material
Figure S3 A stereo pair of a larger thickness of the Af6 mutant embryo shown in Figure 7f, showing localization of E-cadherin (red) and prominin (green). Much of the prominin staining is clearly basolateral, although in some regions it remains apical. Twenty-eight sections in a Z-series were stacked and projected. The data from this figure are shown as a rotating stereo movie at http://www.montana.edu/wwwmri/mut.mov.