First published online November 26, 2007
doi: 10.1242/10.1242/dev.014027
Development 134, 4469-4478 (2007)
Published by The Company of Biologists 2007
The transmembrane protein Perdido interacts with Grip and integrins to mediate myotube projection and attachment in the Drosophila embryo
Beatriz Estrada1,2,*,
,
Stephen S. Gisselbrecht1,* and
Alan M. Michelson1,3,
1 Division of Genetics, Department of Medicine, Brigham and Women's Hospital and
Harvard Medical School, Boston, MA 02115, USA.
2 Centro Andaluz de Biología del Desarrollo (CABD), Universidad Pablo de
Olavide-CSIC, Carretera de Utrera Km. 1, 41013 Sevilla, Spain.
3 National Heart, Lung and Blood Institute, National Institutes of Health,
Bethesda, MD 20892, USA.
View larger version (80K):
[in this window]
[in a new window]
|
Fig. 1. perd is expressed in a subset of muscle founder cells (FCs) and
is required for proper muscle development. (A) Double-labeling of a
stage 12 wild-type embryo for perd RNA (blue) and Kruppel (Kr; brown)
protein (which marks a subset of FCs) shows coexpression in some (arrows) but
not all (arrowheads) Kr-positive FCs. (B) A high-magnification view of
a stage 15 embryo shows that perd RNA is present in the muscle
(arrows) and not in the tendon cell (bracket). (C-F) In contrast to
injection of inactive control lacZ dsRNA (C,D), injection of
double-stranded RNA corresponding to a portion of the perd gene (E,F)
into embryos expressing tau-GFP in the somatic musculature causes specific
muscles to develop with a rounded morphology and incorrect attachments
(arrows), while other muscle groups are unaffected (arrowhead). (G)
perd encodes a single-pass transmembrane protein of 2355 amino acids
with two laminin G domains near its amino terminus and a carboxyl-terminal
class II PDZ-binding motif (NQYWV). Asterisks indicate the positions of
nonsense mutations in the EMS-induced alleles H2-5 (1), F1-3 (2), F2-5 (3) and
187(C2) (4).  indicates the position of a four-nucleotide deletion in
H1-4, resulting in a frameshift and early termination. (H,I)
Immunostaining for myosin heavy chain shows that genetic loss of perd
function causes the same muscle phenotype as RNAi, in which ventral
longitudinal and segment border muscles have a rounded or teardrop shape
(arrows), whereas lateral transverse muscles are normal (arrowhead).
|
|
View larger version (86K):
[in this window]
[in a new window]
|
Fig. 2. perd mutant muscles fail to form correct attachments. In
wild-type embryos expressing β-galactosidase in muscle VL1 (with the 5053
Gal4 line) (A-F), this muscle sends out blunt-ended processes
anteriorly during stage 13 (arrowheads in D), which continue to extend during
stage 14 (E) and form mature-appearing connections by stage 16 (F). In a
perd mutant (G-I), at stage 13 myotube projections appear
thinner and less well-spread but are correctly oriented (arrowheads in G),
then fail to extend through stage 14 (H). By stage 16, muscles are rounded up
and new projections extend in inappropriate directions (arrow, I).
|
|
View larger version (62K):
[in this window]
[in a new window]
|
Fig. 3. perd and Grip have similar phenotypes and are
coexpressed. (A-C) The terminal muscle phenotype in Grip
mutant embryos (C) appears similar to, although less severe than, the
perd phenotype (B). (D-F) The time course of development of
the Grip phenotype is very similar to that of perd (compare
with Fig. 2). (G-I)
Simultaneous detection of perd RNA (G) and Grip RNA (H) by
fluorescence in situ hybridization shows that many cells coexpress both genes
(arrowheads). (J,K) Quantification of muscle VL1 phenotypes in
perd (J) and Grip (K) mutant embryos at three different
stages of development. Each mutation causes a majority of muscles to appear
abnormal in the earliest phases of process extension, but most muscles recover
and achieve an appropriate morphology at later stages in the Grip
mutant, whereas the perd mutant phenotype is increasingly severe. In
addition, a small number of muscles with misdirected projections can be
observed at stage 15 in both genotypes (yellow).
|
|
View larger version (58K):
[in this window]
[in a new window]
|
Fig. 4. perd and Grip interact specifically in a novel RNAi
interaction assay. (A) Injection of Grip dsRNA at high
concentration phenocopies a loss-of-function mutation. (B-D)
Representative embryos injected with perd dsRNA (B) or Grip
dsRNA (C) at concentrations titrated to give minimal effect, and the stronger
effect observed when they are combined (D). (E) Quantification of
pooled data from four independent experiments, comparing the frequency with
which effects are seen from each dsRNA individually or from both together,
along with the distribution predicted if effects were additive (see Materials
and methods for details). n=the number of informative embryos scored
for each condition. The probability of obtaining the observed frequencies from
the additive distribution was <10-12 ( 2 test).
(F,G) In identical experiments, mib2 dsRNA does not
exhibit synergy with perd or Grip dsRNA, despite causing
detachment and rounding of a similar subset of muscles when injected at high
concentration.
|
|
View larger version (92K):
[in this window]
[in a new window]
|
Fig. 5. Perd protein physically interacts with Grip and localizes it to muscle
attachment sites. (A,B) Individual PDZ domains from Grip
were prepared as GST fusion proteins, incubated with epitope-tagged Perd
intracellular domain, and purified by glutathione affinity (PDZ2 was not
successfully assessed in this experiment). Detection of the epitope tag (A)
shows that PerdIC is present only when Grip PDZ7 is present, whereas anti-GST
antibody (B) reveals that fusion protein was present in all lanes.
(C-H) Double-labeling of wild-type embryos for myosin heavy chain and
Grip protein shows that Grip is localized to sites of muscle attachment. In
perd mutant embryos (I-N), Grip protein fails to localize and
is visible over the entire periphery of Grip-expressing muscles. F-H and l-N
are higher magnification images of part of the embryos shown in C-E and I-K,
respectively.
|
|
View larger version (86K):
[in this window]
[in a new window]
|
Fig. 7. Loss of PS1, but not PS2, integrin function phenocopies perd
loss-of-function in muscles. Each row shows the development of a single
embryo injected with dsRNA from the time GFP expression in muscle VL1 first
becomes detectable (t=0, approximately stage 14) through late stages,
when muscles are contracting and embryos are moving. Additional GFP expression
comes from visceral muscles that also express Gal4, internally to the VL1
muscle. Scale bar: 50 µm. (A-D) In control lacZ
dsRNA-injected embryos, muscles are initially unattached, but rapidly achieve
their mature attachment sites (B) and elongated morphology (C). (E-H)
RNAi directed against perd prevents muscles from ever forming proper
attachments. (I-L) Grip dsRNA affects fewer muscles, with a
timecourse similar to that of perd. (M-P) RNAi for the
 PS1 integrin subunit mew, which affects both maternal and
zygotic transcripts, causes a severe phenotype identical to that of
perd. By contrast, in embryos injected with dsRNA for the PS2
subunit if (Q-T), many muscles develop apparently normal
attachments and elongated morphology at stage 16 (arrowheads in S) before
assuming a rounded-up appearance after muscle contraction begins (arrowheads
in T). Removal of both PS integrins by RNAi directed against the common β
subunit mys (U-X) gives the more severe early phenotype.
|
|
View larger version (26K):
[in this window]
[in a new window]
|
Fig. 8. A model for Perd function in formation of muscle attachments.
(A) Muscle-expressed Perd protein forms a complex with Grip, directing
it to sites of tendon contact. We hypothesize that PS1 integrin heterodimers
(the product of the mew and mys genes) expressed on tendon
cells may serve as the ligand for Perd binding and thus mediate target
recognition. See Discussion for details. The additional PDZ domains of Grip
can then recruit other proteins required for the maturation of the
myotendinous junction; the PS2 integrin expressed on the muscle and known to
be required for stable muscle attachment is an attractive candidate, but no
direct interaction with Grip has been demonstrated. (B-D) Elements of
this model have previously been reported (references, bottom right and in
Discussion) in other systems where Perd orthologues are expressed, including
interaction with an integrin of the laminin-binding class (B), interaction
with Grip (C), and acting as an adhesion coreceptor for an integrin in cis
(D).
|
|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati 
Twitter What's this?
© The Company of Biologists Ltd 2007
