![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
|
|
|
|
|
||
Vol. 17, No. 5, pp. 597-602, March 1, 2003
Department of Genetics, Howard Hughes Medical Institute, Harvard Medical School, Boston, Massachusetts 02115, USA
 |
ABSTRACT |
|---|
 Top
 Abstract
 Introduction
Results and Discussion
Materials and methods
References
|
|---|
Integrins and laminins are important mediators of cell-matrix
interactions in both vertebrates and invertebrates. Here, we show that
germ-band retraction in the Drosophila embryo, during which the
tail end of the embryo retracts to its final posterior position, allows
the investigation of cell spreading and lamellipodia formation in real
time in vivo. We demonstrate that 
1, 2 laminin and PS3
PS
integrin are required for the spreading of a small group of cells of
the amnioserosa epithelium over the tail end of the germ band. We
further implicate a role for this spreading in the process of germ-band
retraction.
| |
Introduction |
|---|
|
Top
Abstract
Introduction
Results and Discussion
Materials and methods
References
|
|---|
Integrins and laminins have been shown to function in signaling,
in stable adhesion of differentiated epithelia to the
basement membrane, and in transient adhesion during cell spreading and migration (Brown et al. 2000
; De Arcangelis and Georges-Labouesse 2000;
Schöck and Perrimon 2002b). Laminins are extracellular matrix molecules consisting of a long chain and shorter and 
chains arranged in a cross-like structure (Martin et al. 1999). One of their
most prominent interaction partners are the integrins, which form
transmembrane heterodimers of and subunits. In
Drosophila there are two , one , and one laminin
genes, and five and two integrin genes (Brown et al. 2000;
Hynes and Zhao 2000; Fig. 1A). Laminins and
integrins have partially overlapping, but not identical phenotypes,
because, even though they bind to each other, they also have various
nonshared interaction partners (Brown et al. 2000; De Arcangelis and
Georges-Labouesse 2000). In Drosophila, several morphogenetic
processes require the presence of integrins, for example the formation
of muscle attachments and wing-blade adhesion (Roote and Zusman 1995;
Brown et al. 2000). Similar to integrin mutants, mutations in
Drosophila laminin are associated with wing-blade adhesion and
muscle-development defects (Henchcliffe et al. 1993; Yarnitzky and Volk
1995; Martin et al. 1999). Further, both 1, 2 laminin and PS
integrin mutants show a transient twisting of the germ band (Roote and
Zusman 1995; Martin et al. 1999).
|
Another process that is affected by the absence of integrins, but which
has not been studied up to now, is germ-band retraction (Leptin et al.
1989; Walsh and Brown 1998). Germ-band retraction occurs in
mid-embryogenesis after germ-band extension and prior to dorsal
closure. It takes ~2 h and involves large-scale epithelial movements,
by which the tail end of the germ band, or embryo proper, moves to its
final posterior position (Martinez Arias 1993). At the end of this
process, the amnioserosa, a squamous epithelium, has spread out and
covers the yolk sac on the dorsal side of the embryo. We have shown
previously that the amnioserosa and the germ band move as one coherent
sheet during retraction. In addition, amnioserosa cells undergo a
dramatic shortening along their dorsoventral axis during retraction
(Schöck and Perrimon 2002a). Most interestingly, retraction is
associated with lamellipodia formation and cell spreading of the
posterior cell row of the amnioserosa, which causes the amnioserosa to
slightly overlap the tail end of the germ band (Schöck and Perrimon
2002a; Fig. 4E, below). The lamellipodia of the amnioserosa spread on
the apical extracellular matrix of the germ band, which is deposited by
the end of embryonic stage 10 (Tepass and Hartenstein 1994).
To explore the mechanisms underlying the observed spreading of the
amnioserosa, we analyzed mutants exhibiting germ-band retraction defects. Here, we show that 1, 2 laminin and PS3PS integrin are required to generate the overlap of the amnioserosa over the tail
end of the germ band, and that they interact genetically during
retraction. Live imaging of embryos completely null for PS integrin
revealed that lamellipodia formation is disrupted and that there is no
cell-matrix adhesion between the amnioserosa and the tail end of the
germ band.
| |
Results and Discussion |
|---|
|
Top
Abstract
Introduction
Results and Discussion
Materials and methods
References
|
|---|
Analysis of maternal-effect U-shaped mutations
To determine whether germ-band retraction is a suitable system to
study cell-matrix interactions and lamellipodia formation, we searched
for genes involved in the cellular functions required for retraction.
We screened a collection of zygotic lethal mutations for mutants
associated with maternal-effect U-shaped embryonic phenotypes (Perrimon
et al. 1996; N. Perrimon, unpubl.). U-shaped cuticle phenotypes are
characterized by the dorsal location of abdominal denticle belts 6-8
and are indicative of a failure of retraction (Frank and Rushlow 1996).
One of the mutants we uncovered, 4Y18, shows a U-shaped
phenotype when both the zygotic and the maternal contributions are
removed using the germ-line clone method (Chou and Perrimon 1992; Fig.
1B,C). Mapping experiments revealed that 4Y18 fails to
complement mutations at the wing blister (wb) locus,
which encodes 1, 2 laminin, a polypeptide of the trimeric laminins
(Martin et al. 1999). Sequencing of PCR-amplified 4Y18 genomic
DNA revealed a mutation causing a truncation after 1427 amino acids,
which deletes the trimerization domain of Wb (Fig. 1D). The mutant Wb
protein appears to be unstable, because it was not detected with an
N-terminal anti-Wb antibody in homozygous mutant embryos (data not
shown), suggesting that wb4Y18 is a null allele.
1, 2 laminin mutant shows germ-band retraction defects
Consistent with our finding that the wb4Y18
maternal and zygotic (m/z) mutants exhibit U-shaped cuticle phenotypes,
m/z mutant embryos in the PS integrin myospheroid
(mys) also show a U-shaped phenotype (Leptin et al. 1989; Fig.
5A, below). In addition, time-lapse recordings show that the U-shaped
cuticle defect of wb4Y18 and
mysXG43 m/z mutant embryos results from a failure of
germ-band retraction (data not shown; Fig. 3B, below). At the cellular
level, wb4Y18 m/z mutant embryos exhibit pleiotropic
defects most evident in the amnioserosa. Anti- Spectrin antibody
stainings showed that, similar to mys m/z mutants (see below),
the overlap of the amnioserosa over the tail end of the germ band is
defective, whereas cell-shape changes in the amnioserosa occur normally
(Fig. 1, cf. F and wild type, E). In addition, the amnioserosa
epithelium frequently exhibits multilayering, suggesting polarity
defects (Fig. 1G, arrowheads). Laminin functions in epithelial polarity
in other systems, for example during the development of the mammalian
kidney epithelium, in which laminin is required to provide polarity
cues (O'Brien et al. 2001). We cannot address where exactly 1, 2 laminin is required to mediate germ-band retraction, because it is a
secreted protein. However, the wb4Y18 amnioserosa
defects indicate that cell-matrix interactions are required between
the amnioserosa and the underlying yolk sac and/or between the
overlapping part of the amnioserosa and the apical extracellular matrix
of the germ band.
We also analyzed the only other Drosophila laminin, 3,
5 laminin, encoded by lamininA (lanA; Kusche-Gullberg
et al. 1992; Henchcliffe et al. 1993; Martin et al. 1999). We generated
m/z mutant embryos of the null mutant lanA9-32 to
remove a weak maternal contribution that had not been analyzed previously (Kusche-Gullberg et al. 1992), and observed a range of
defects. In more severe cases, embryos disintegrate before the onset of
retraction, making it impossible to assess the contribution of
lanA to germ-band retraction. However, in milder cases,
lanA m/z embryos exhibit fusion or deletion of segments and
twisting, but no retraction defects (Fig. 1H).
PS3 integrin mutant exhibits germ-band retraction defects and
interacts with 1, 2 laminin
Another mutant (5J38) identified in our screen for
maternal-effect U-shaped phenotypes, was mapped to the scab
(scb) locus, which encodes the PS3 integrin subunit (Stark
et al. 1997; Fig. 2A). PS3 integrin
mutants have dorsal closure defects, and, in addition, PS3 is
expressed in morphogenetically active tissues, such as the salivary
glands and the tracheal pits during their invaginations, as well as in
the amnioserosa (Stark et al. 1997). To assess whether PS3 zygotic
expression, particularly in the amnioserosa, is required for
retraction, we analyzed zygotic scb mutant embryos and found
that they have U-shaped phenotypes as well. The penetrance of these
phenotypes is low, but is enhanced (about threefold) in embryos
transheterozygous for scb and a deficiency that deletes both
PS3 and the closely related PS4 integrin (Fig. 2C), suggesting
functional redundancy of these two closely related subunits. We have
not yet assessed the contribution of the remaining two unrelated PS
integrin subunits, PS1 and PS2.
|
To determine whether 1, 2 laminin interacts with integrins in vivo,
we tested for genetic interactions with a hypomorphic viable PS
integrin allele (mysnj42) that shows a wing blister
phenotype (Walsh and Brown 1998). Wing blisters arise due to a failure
of cell-matrix interactions required for the apposition of two
epithelia that form the wing blade (Walsh and Brown 1998). Consistent
with the similarity of the mys and wb embryonic
phenotypes, removal of one copy of wb in a hemizygous PS
integrin mutant background results in a significant increase in wing
blisters (Fig. 2B), whereas removal of one copy of lanA does
not cause an increase (Henchcliffe et al. 1993). We next used the low
penetrance of germ-band retraction defects in PS3 integrin mutants
to test for genetic interactions with 1, 2 laminin mutants during
retraction. We analyzed zygotic wb scb double mutants and
zygotic scb mutants with only one copy of wb. In both
cases, we observed a significant increase in the frequency of U-shaped
cuticle defects (Fig. 2C). These results show that PS3 integrin and
1, 2 laminin interact in vivo during germ-band retraction.
PS3PS integrin-containing S2 cells spread on 1,
2 laminin
We next asked whether PS3PS integrin is able to mediate
cell-matrix interactions with 1, 2 laminin, using a cell-spreading assay in tissue culture. Schneider's line 2 (S2) Drosophila
cells are particularly suitable for such assays because, even though they express PS integrin, they do not spread due to the lack of
expression of PS integrins (Gotwals et al. 1994). The PS integrin
subunit heterodimerizes with PS1, PS2, and PS3 integrin (Gotwals et al. 1994; Stark et al. 1997), and it was shown previously that PS2PS integrin interacts with 1, 2 laminin in a
cell-spreading assay (Graner et al. 1998). We therefore transfected S2
cells with PS1, the splice variant PS2m8, or PS3 and plated
them on plates coated with a N-terminal 1, 2 laminin fragment
containing the RGD motif required for interaction with PS2PS to
compare their adhesion behavior (Fig. 2D-G). Similar to PS2PS,
PS3PS mediates spreading on the N-terminal 1, 2 laminin
fragment, whereas PS1PS does not (Fig. 2D). To assess the
contribution of the RGD motif in the observed spreading behavior, we
tested spreading on a short fibronectin analog peptide containing the
RGD motif (GRGDSPK). Only PS2PS exhibits spreading on this
peptide, indicating that the contribution of the RGD motif to spreading
of PS3PS is smaller than for PS2PS or is absent (Fig. 2D).
Finally, all three integrin heterodimers may also bind to and spread on
laminin domains not included in the 1, 2 laminin fragment used for
this spreading assay, as PS1PS integrin-containing cells had been shown previously to spread on full-length 3, 5 laminin (Gotwals et
al. 1994). Altogether, our results demonstrate that PS3PS integrin mediates spreading on 1, 2 laminin, that these integrin and
laminin subunits are required for germ-band retraction, and that they
interact genetically.
PS integrin mutant retracts the germ band only partially
To address more precisely the role of integrins during germ-band
retraction, we used live imaging of PS integrin (mys)
mutant embryos. We focused our analyses on mys mutants,
because integrins must heterodimerize with integrins in order
to function, and mys m/z mutants show the highest penetrance
of U-shaped cuticle phenotypes. mysXG43 m/z embryos
are completely null for PS integrin (Leptin et al. 1989), and 65%
of the mutants display a severe U-shaped cuticle phenotype (Fig. 5E,
below). To outline all cells in time-lapse recordings, we marked
embryos with ubiDE-cadherinGFP. In paternally rescued maternal
mys mutant embryos, which were genotyped by the additional GFP
expression in the amnioserosa, retraction proceeds normally as in
wild-type embryos (Fig. 3A). In contrast,
in mys m/z mutant embryos, retraction progresses only for a
short time, then stops prematurely, which results in an average of two
to three abdominal segments remaining on the dorsal side (Fig. 3B). Most cell-shape changes are unaffected; in particular, amnioserosa cells still shorten along their dorsoventral axis (Fig. 3B; data not
shown).
|
PS integrin mutant does not adhere to the tail end of the
germ band and does not form lamellipodia
We then characterized the effects of PS integrin mutants on
cell-matrix adhesion between the posterior part of the amnioserosa and
the tail end of the germ band. In mys m/z mutants, the
amnioserosa no longer overlaps the germ band, but instead, crowds in
front of the tail end of the germ band (Fig. 4, cf. B and
A). At a slightly later stage, the
amnioserosa collapses below the tail end of the germ band and cuts in
under the germ band to the level of the hindgut (Fig. 4C). The collapse
of the amnioserosa below the germ band may be a result of the
dorsoventral shortening of the amnioserosa, which normally occurs
during retraction (Schöck and Perrimon 2002a). This phenotype occurs
with the same penetrance as the U-shaped cuticle phenotype. To
determine what causes the amnioserosa to not overlap the tail end of
the germ band, we analyzed time-lapse recordings of dorsal views of
mys mutants marked with
Kr-Gal4, UASsrcEGFP. Paternally rescued maternal
mys mutants show normal lamellipodia formation and adhesion of
the amnioserosa to the tail end of the germ band (Fig. 4E). However,
analysis of mys m/z mutant embryos revealed that lamellipodia
are generally absent from the posterior row of amnioserosa cells (Fig.
4F), and if present, are very small and transient (Fig. 4D). Thus,
cell-matrix adhesion of the amnioserosa to the apical extracellular
matrix of the germ band is disrupted. In conclusion, in both
wb4Y18 and mysXG43 m/z mutants,
the amnioserosa does not overlap the tail end of the germ band as in
wild-type embryos, indicating a requirement for cell-matrix adhesion
between amnioserosa and germ band to maintain the overlap.
|
Expression of PS integrin in the amnioserosa rescues
retraction defects
Finally, to determine whether PS integrin in the amnioserosa is
sufficient to mediate germ-band retraction, we rescued the U-shaped
phenotype of mys m/z mutants by expressing PS integrin in
the amnioserosa with the Gal4/UAS system (Brand and Perrimon 1993).
Expression of UASmys with the Kr-Gal4 line, which
drives expression in the amnioserosa and six segments of the germ band (T3-A5), rescues the U-shaped cuticle phenotype of mys m/z
mutants substantially (Fig. 5B,E). Notably,
expression of UASmys exclusively in the amnioserosa with the
LP1-Gal4 driver rescues the retraction defects of mys
m/z mutants to the same degree (Fig. 5C,E). In contrast, neither the
UASmys construct alone, nor the amnioserosal expression of an
adhesion-deficient PS integrin capable only of signaling
(Martin-Bermudo and Brown 1999; Martin-Bermudo 2000) rescued the
Ushaped phenotype (Fig. 5D,E). These data indicate that PS
integrin expression is required in the amnioserosa, and that the
adhesive properties of PS integrin are most likely required for
retraction.
|
Conclusions
Altogether, we propose that in wild-type embryos, dorsoventral
contraction of amnioserosa cells contributes to the extension of
germ-band epithelial cells along the dorsoventral axis. This dorsoventral extension of the germ-band epithelium may then promote the
retraction of the remaining segments that are still folded back. In the
absence of proper laminin/integrin-mediated adhesion, the overlap of
the amnioserosa over the tail end of the germ band is defective. We
suggest that dorsoventral contraction of the amnioserosa in this mutant
situation results in the collapse of the amnioserosa underneath the
germ band, leading to an incorrect deployment of forces on the germ
band and eventual complete or partial failure of retraction. We further
note that the amnioserosa may also not adhere properly to the
underlying yolk sac in the integrin mutants, which could also
contribute to retraction defects. Our study demonstrates the
involvement of PS3PS integrin and 1, 2 laminin in germ-band
retraction and provides a hypothesis for how retraction proceeds.
To date, studies of integrin function during lamellipodia formation and
cell spreading have been mostly limited to tissue culture. A wealth of
information has been obtained on the role of these molecules in the
information flow from the extracellular matrix to the cytoskeleton
during migration (Geiger et al. 2001). The process of germ-band
retraction in the Drosophila embryo now provides an in vivo
model system in which the function of the molecules identified in
tissue culture experiments can be systematically tested. Uniquely, this
system allows the study of lamellipodia formation and cell spreading in
embryos mutant for candidate genes in real time and in vivo. Finally,
the characterization of additional mutations associated with germ-band
retraction defects will allow the identification of novel players
involved in cell-matrix adhesion.
| |
Materials and methods |
|---|
|
Top
Abstract
Introduction
Results and Discussion
Materials and methods
References
|
|---|
Fly strains
The following fly strains were used: Kr-Gal4 40°,
c381-Gal4, LP1-Gal4, UASmys,
UASTorsoDPScyt,
UASsrcEGFP, ubiDE-cadherinGFP, scb5J38
FRTG13, scb2, mysXG43
FRT101, mysnj42, lanA9-32
FRT79DF, and wb4Y18 FRT40A (Leptin et al. 1989;
Martin-Bermudo and Brown 1996, 1999; Oda and Tsukita 1999, 2000; Deng
and Ruohola-Baker 2000; Herranz and Morata 2001). We made the following
recombinants using standard genetic techniques: UASsrcEGFP Kr-Gal4
40°, UASTorsoDPScyt
UASmCD8GFP, UASmys UASsrcEGFP, and wb4Y18
scb2.
Time-lapse recording
Time-lapse recordings were done as described (Schöck and
Perrimon 2002a). To analyze embryos, we used
ubiDE-cadherinGFP, which is ubiquitously expressed and
localizes to adherens junctions, thereby outlining cells (Oda and
Tsukita 2001), and UASsrcEGFP, which localizes to both
cytoplasm and plasma membrane due to the Src myristilation anchor
attached to EGFP. mys germ-line clone (GLC) embryos were
analyzed for lamellipodia by crossing FM7c, Kr-Gal4
UASGFP; UASsrcEGFP Kr-Gal4 40° or FM7c,
Kr-Gal4 UASGFP; ubiDE-cadherinGFP males to
mys GLC females. This allowed us to distinguish paternally
rescued embryos expressing nuclear GFP from m/z embryos.
Genetic studies
From a GLC screen of more than 6000 zygotic lethal mutations
(Perrimon et al. 1996; N. Perrimon, unpubl.), a number of mutants were
isolated that exhibit U-shaped embryonic phenotypes. GLCs were produced
in females of the genotype y w hsFLP; mutant
FRT/ovoD1 FRT using the ovoD1 system
and mated to mutant or wild-type males for determination of paternal
rescue as described (Chou and Perrimon 1992).
wb4Y18 and scb5J38 were mapped using deficiencies and alleles of the respective mutants. 4Y18 was mapped to a small region (35A) containing two loci that mutate to lethality; one of them, wb, did not complement 4Y18. As expected, the smallest deficiency Df(2L)el80f1 did not paternally rescue the 4Y18 GLC mutant. The wb4Y18 maternal effect U-shaped phenotype is partially penetrant and fully paternally rescueable. Twenty-two percent of the embryos from wb4Y18 GLC females crossed to wb4Y18/+ males show a U-shaped phenotype (A7 and A8 on dorsal side), whereas only 2% show this phenotype when wb4Y18 GLC females were crossed to wild-type males; 5J38 was mapped similarly to 51D-52A, and failed to complement mutants at the scb locus.
For the rescue experiments, we crossed mys GLC females to
Kr-Gal4 UASsrcEGFP/CyO; UASmys/TM3, actGFP,
Ser males or to UASmys UASsrcEGFP/+
LP1-Gal4/+ and
UASTorsoDPScyt UASmCD8GFP/+;
c381-Gal4/+ males, respectively. We sorted aged embryos for
the presence of Kr-Gal4 UASsrcEGFP and the absence of TM3,
actGFP or for the presence of UASmys UASsrcEGFP and
UASTorsoDPScyt UASmCD8GFP,
respectively. We then aged these embryos further to allow paternally
rescued mys GLC embryos to hatch.
Histochemistry and sequencing
Primary antibody stainings with polyclonal rabbit anti-Wb 25632 (1:100) and monoclonal mouse anti- Spectrin (3
9 from DSHB, 1:10) were performed as described (Pesacreta et al. 1989; Martin et
al. 1999). Fluorescein-conjugated goat anti-rabbit IgG (Vector, 1:400) and AlexaFluor 488-conjugated donkey anti-mouse IgG
(Molecular Probes, 1:500) were used as secondary antibody. Genomic
DNA of wild-type and wb4Y18 homozygous mutant
embryos was PCR-amplified with 16 primer pairs and sequenced by
standard thermal cycle dideoxy sequencing.
Cell-spreading assay
The cell-spreading assay was performed as described (Zavortink et
al. 1993). Briefly, a 340-amino-acid His-tagged RGD-containing fragment
of 1, 2 laminin was purified under denaturing conditions (Graner et
al. 1998). A 3.6-kb XhoI SpeI full-length cDNA
fragment of PS3 was cloned into pMK33 (Gotwals et al. 1994; Stark et
al. 1997). S2 cells were transiently transfected and expression of PS integrins was induced with 500 µM CuSO4 for 24 h.
Cells were allowed to spread for 6 h on coated 96 well tissue culture
plates and then fixed for 30 min in 4% formaldehyde, and at the same time stained for filamentous actin with AlexaFluor 568 phalloidin (Molecular Probes, 1:50). Experiments were done in triplicate with
at least 150 cells counted in each experiment.
| |
Acknowledgments |
|---|
We thank S. Baumgartner, D. Brower, N. Brown, T. Bunch, R. Hynes, G. Morata, H. Oda, H. Ruohola-Baker, E. Spana, and the Bloomington stock center for flies and reagents, and P. Bradley, R. DasGupta, and E. Küster-Schöck for comments on the manuscript. F.S. was an HHMI Fellow and is currently an HFSPO Long-Term Fellow. N.P. is an HHMI investigator.
The publication costs of this article were defrayed in part by payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 USC section 1734 solely to indicate this fact.
| |
Footnotes |
|---|
[Keywords: Cell spreading; Drosophila; germ-band retraction; integrin; laminin]
Received December 17, 2002; revised version accepted January 22, 2003.
1 E-MAIL fschoeck{at}genetics.med.harvard.edu; FAX (617) 432-7688.
2 E-MAIL perrimon{at}rascal.med.harvard.edu; FAX (617) 432-7688.
Corresponding authors.
Article and publication are at http://www.genesdev.org/cgi/doi/10.1101/gad.1068403.
| |
References |
|---|
|
Top
Abstract
Introduction
Results and Discussion
Materials and methods
References
|
|---|
chain required for cell adhesion and migration during embryonic and imaginal development.
J. Cell Biol.
145:
191-201PS cytoplasmic domain is sufficient to regulate gene expression in the Drosophila embryo.
Genes & Dev.
13:
729-739catenin-GFP fusion protein.
Dev. Genes Evol.
209:
218-225[CrossRef][Medline]. integrin subunit associates with PS and functions in tissue morphogenesis and movement during Drosophila development.
Development
124:
4583-4594[Abstract]. subunit.
Cell Adhes. Commun.
1:
251-264[Medline].
CiteULike 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  C. C. F. Homem and M. Peifer Diaphanous regulates myosin and adherens junctions to control cell contractility and protrusive behavior during morphogenesis Development, March 15, 2008; 135(6): 1005 - 1018. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Jani and F. Schock Zasp is required for the assembly of functional integrin adhesion sites J. Cell Biol., December 31, 2007; 179(7): 1583 - 1597. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Gates, J. P. Mahaffey, S. L. Rogers, M. Emerson, E. M. Rogers, S. L. Sottile, D. Van Vactor, F. B. Gertler, and M. Peifer Enabled plays key roles in embryonic epithelial morphogenesis in Drosophila Development, June 1, 2007; 134(11): 2027 - 2039. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Takahashi, F. Takahashi, K. Ui-Tei, T. Kojima, and K. Saigo Requirements of genetic interactions between Src42A, armadillo and shotgun, a gene encoding E-cadherin, for normal development in Drosophila Development, June 1, 2005; 132(11): 2547 - 2559. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Kozlova and C. S. Thummel Essential Roles for Ecdysone Signaling During Drosophila Mid-Embryonic Development Science, September 26, 2003; 301(5641): 1911 - 1914. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
