Diaphanous is required for cytokinesis in Drosophila and shares domains of similarity with the products of the limb deformity gene.
ABSTRACT We show that the Drosophila gene diaphanous is required for cytokinesis. Males homozygous for the dia1 mutation are sterile due to a defect in cytokinesis in the germline. Females trans-heterozygous for dia1 and a deficiency are sterile and lay eggs with defective eggshells; failure of cytokinesis is observed in the follicle cell layer. Null alleles are lethal. Death occurs at the onset of pupation due to the absence of imaginal discs. Mitotic figures in larval neuroblasts were found to be polyploid, apparently due to a defect in cytokinesis. The predicted 123 x 10(3) M(r) protein contains two domains shared by the formin proteins, encoded by the limb deformity gene in the mouse. These formin homology domains, which we have termed FH1 and FH2, are also found in Bni1p, the product of a Saccharomyces cerevisiae gene required for normal cytokinesis in diploid yeast cells.
- SourceAvailable from: Jörg Grosshans[Show abstract] [Hide abstract]
ABSTRACT: Members of the Diaphanous (Dia) protein family are key regulators of fundamental actin driven cellular processes, which are conserved from yeast to humans. Researchers have uncovered diverse physiological roles in cell morphology, cell motility, cell polarity, and cell division, which are involved in shaping cells into tissues and organs. The identification of numerous binding partners led to substantial progress in our understanding of the differential functions of Dia proteins. Genetic approaches and new microscopy techniques allow important new insights into their localization, activity, and molecular principles of regulation.Communicative & integrative biology 11/2013; 6(6):e27634.
- [Show abstract] [Hide abstract]
ABSTRACT: During muscle development, myosin and actin containing filaments assemble into the highly organized sarcomeric structure critical for muscle function. Although sarcomerogenesis clearly involves the de novo formation of actin filaments, this process remained poorly understood. Here we show that mouse and Drosophila members of the DAAM formin family are sarcomere-associated actin assembly factors enriched at the Z-disc and M-band. Analysis of dDAAM mutants revealed a pivotal role in myofibrillogenesis of larval somatic muscles, indirect flight muscles and the heart. We found that loss of dDAAM function results in multiple defects in sarcomere development including thin and thick filament disorganization, Z-disc and M-band formation, and a near complete absence of the myofibrillar lattice. Collectively, our data suggest that dDAAM is required for the initial assembly of thin filaments, and subsequently it promotes filament elongation by assembling short actin polymers that anneal to the pointed end of the growing filaments, and by antagonizing the capping protein Tropomodulin.PLoS Genetics 02/2014; 10(2):e1004166. · 8.17 Impact Factor
- [Show abstract] [Hide abstract]
ABSTRACT: Actin-based protrusions are important for signaling and migration during development and homeostasis. Defining how different tissues in vivo craft diverse protrusive behaviors using the same genomic toolkit of actin regulators is a current challenge. The actin elongation factors Diaphanous and Enabled both promote barbed-end actin polymerization, and can stimulate filopodia in cultured cells. However, redundancy in mammals and Diaphanous' role in cytokinesis limited analysis of whether and how they regulate protrusions during development. We use two tissues driving Drosophila dorsal closure, migratory leading-edge (LE) and non-migratory amnioserosal (AS) cells, as models to define how cells shape distinct protrusions during morphogenesis. We found that non-migratory AS cells produce filopodia that are morphologically and dynamically distinct from those of LE cells. We hypothesized that differing Enabled and/or Diaphanous activity drive these differences. Combining gain- and loss-of-function with quantitative approaches revealed Diaphanous and Enabled each regulate filopodial behavior in vivo and defined a quantitative "fingerprint", the protrusive profile, which our data suggest is characteristic of each actin regulator. Our data suggest LE protrusiveness is primarily Enabled-driven, while Diaphanous plays the primary role in the AS, and reveal each has roles in dorsal closure, but its robustness ensures timely completion in their absence.Molecular Biology of the Cell 08/2014; · 4.55 Impact Factor
To complete mitosis successfully, a cell must carry out two
separate processes: karyokinesis, or chromosome segregation,
and cytokinesis, or the division of the entire cell. In animal
cells, cytokinesis is believed to be mediated by the contractile
ring, a transient cytoskeletal structure located midway between
the spindle poles at the cell cortex. Actin and myosin, which
are concentrated at the cleavage furrow in a variety of cells,
are believed to generate the force leading to contraction of the
ring and, ultimately, the separation into two daughter cells
(Satterwhite and Pollard, 1992).
Despite increasing knowledge about mitosis and the cell
cycle, many aspects of cytokinesis remain poorly understood.
For example, it is not clear how actin and myosin are recruited
to the contractile ring, nor how the contractile ring is attached
to the cell membrane. The position of the contractile ring is
believed to be determined by a signal originating at the mitotic
spindle (Rappaport, 1986), but the nature of this signal is not
known. Lastly, few of the structural and regulatory proteins
involved in cytokinesis have been identified.
The isolation of mutations that disrupt cytokinesis in
organisms such as yeast or Drosophila promises to be a useful
approach for identifying genes required for cytokinesis.
Indeed, mutations identified through genetic screens in
Drosophila (Karess et al., 1991), or through a ‘reverse genetic’
approach in S. cerevisiae(Watts et al., 1987) and Dictyostelium
(DeLozanne and Spudich, 1987), have demonstrated that
myosin function is required for cytokinesis. More recently, the
peanut locus has been shown to be required for cytokinesis in
Drosophila and to encode a homolog of Cdc12 and related
proteins required for cytokinesis in yeast (Neufeld and Rubin,
The diaphanous (dia) locus was identified in a P element
screen for recessive male-sterile mutations (Castrillon et al.,
1993). Here we present evidence that dia is generally required
for cytokinesis. A failure in cytokinesis during spermatogene-
sis results in multinucleate spermatids. A defect in oogenesis
in females is associated with a failure of cytokinesis in the
somatically derived follicle cells, which surround each devel-
oping egg chamber. We further show that null dia alleles result
in early pupal lethality and that the lethal phenotype is consis-
tent with a defect in cytokinesis. Lastly, we have defined two
evolutionarily conserved domains that are common to the
protein products of diaphanous, the limb deformity locus of
mouse and chicken, and BNI1, a yeast gene identified on the
basis of its interaction with CDC12.
MATERIALS AND METHODS
Drosophila markers and manipulations
All crosses were performed at 25°C on yeasted cornmeal molasses
agar. Genetic markers and balancers are described and referenced by
Lindsley and Zimm (1992). Lethal dia mutations were balanced with
In(2LR)Gla, Bc. RNA from germlineless flies was prepared from the
progeny of females homozygous for the tud1mutation.
Remobilization of P element to generate new alleles
The dia1chromosome was brought together with a transposase source
by crossing dia1/CyO flies to flies carrying the P[ry+, ∆2-3] trans-
posase source on the third chromosome (Robertson et al., 1987). In
the next generation, the P[ry+, ∆2-3] chromosome was crossed out,
and ry−derivatives of the dia1chromosome were selected to establish
Of the 226 rosy−lines generated, 17 were homozygous lethal. The
results of several genetic tests argue that these events are lethal alleles
Development 120, 3367-3377 (1994)
Printed in Great Britain © The Company of Biologists Limited 1994
We show that the Drosophila gene diaphanous is required
for cytokinesis. Males homozygous for the dia1mutation
are sterile due to a defect in cytokinesis in the germline.
Females trans-heterozygous for dia1and a deficiency are
sterile and lay eggs with defective eggshells; failure of
cytokinesis is observed in the follicle cell layer. Null alleles
are lethal. Death occurs at the onset of pupation due to the
absence of imaginal discs. Mitotic figures in larval neuro-
blasts were found to be polyploid, apparently due to a defect
in cytokinesis. The predicted 123×103 Mr protein contains
two domains shared by the formin proteins, encoded by the
limb deformity gene in the mouse. These formin homology
domains, which we have termed FH1 and FH2, are also
found in Bni1p, the product of a Saccharomyces cerevisiae
gene required for normal cytokinesis in diploid yeast cells.
Key words: cytokinesis, mitosis, spermatogenesis, oogenesis, P
diaphanous is required for cytokinesis in Drosophila and shares domains of
similarity with the products of the limb deformity gene
Diego H. Castrillon and Steven A. Wasserman
Department of Biochemistry, University of Texas Southwestern Medical Center, Dallas, Texas 75235-9038, USA
of the dia locus: (1) the cytologically visible deficiency Df(2L)TW84,
which deletes the entire dia region, fails to complement any of these
lethal mutations; (2) complementation tests indicate that these 17
lethal events are due to mutations in the same locus; (3) the lethal
events fail to complement the original dia1 allele and (4) a genomic
fragment containing a single intact transcription unit can fully rescue
the male-sterile and lethal phenotypes (discussed in more detail
Nucleic acid manipulations and analysis
Standard protocols were performed as described (Sambrook et al.,
1989). For plasmid rescue, genomic DNA prepared from adults
homozygous for dia1was digested to completion with XbaI and NheI,
self-ligated and transformed into competent E. coli cells following a
standard protocol (Ashburner, 1989) with selection for kanamycin
resistance. The 140 bp fragment from one of the plasmid rescue clones
was gel-purified, labelled and used to screen a Drosophila genomic
library (EMBL3 vector) provided by J. Tamkun. The 3.6 kb dia cDNA
was obtained by screening a 12-24 hour embryonic library (Brown
and Kafatos, 1988). fod cDNAs were cloned from an adult testis
library provided by T. Hazelrigg.
RNA amplification (RT-PCR) was performed as described in
Kawasaki (1990). Synthetic oligonucleotides were synthesized to the
5′ end of the 3.6 kb cDNA and to the 140 bp fragment flanking the P
element insertion site (see Fig. 7). The synthetic oligonucleotide com-
plimentary to the genomic 140 bp fragment is 5′-ATGGTACCCAA-
GAAAAAGTGTTCGGAGGG-3′. The first 8 bases were added to
create an Asp718I/KpnI site to facilitate subcloning. The sequence of
the oligonucleotide complementary to the 5′ end of the cDNA is 5′-
CTTCCGGGGATCCAGGGATTG-3′. This oligonucleotide was
designed around a naturally occurring BamHI site. 20 picomoles of
each oligonucleotide were used per PCR reaction.
To avoid the complication of sequence polymorphisms that are
common in non-isogenized Drosophila stocks, RT-PCR was
performed on RNA from adult flies carrying Df(2L)TW84, which
deletes the entire dia locus. 0.3 µg of total RNA from
Df(2L)TW84/CyO adults or 0.3 µg of poly(A)+RNA from 8-16 hour
embryos (Oregon-R) were used for cDNA synthesis preceding PCR.
75 picomoles of random hexanucleotides were used to prime cDNA
synthesis by M-MLV reverse transcriptase. Amplification was carried
out with 2.5 units of Taq polymerase in 35 cycles on a Perkin-Elmer-
Cetus DNA thermal cycler using the following parameters: (1) raise
temperature to 95°C, then hold 15 seconds; (2) lower to 55°C, then
hold 30 seconds; (3) raise to 72°C, then hold 30 seconds. A total of
four polymorphisms were identified among the Df(2L)TW84/CyO
PCR products sequenced. Sequencing of an additional embryonic RT-
PCR clone revealed no significant differences.
The 3.6 kb dia cDNA and various PCR products were subcloned
into M13 mp18 and mp19 for sequencing by the dideoxy chain ter-
mination method. Sequences were assembled using Assemblylign
(IBI/Kodak) and analyzed using the BLASTP sequence comparison
program (Altschul et al., 1990). The BLOSUM62 matrix was used for
Predictions of coiled-coil structure from protein sequence were
performed using the algorithm of Lupas et al. (1991). The mean score
for globular proteins and coiled-coil sequences calculated by this
method are 0.77 and 1.63, respectively.
Germline transformation and phenotypic rescue
pDC4 was constructed by ligating the 11 kb SalI fragment from
λ4138R to the w+P element vector pCaSpeR 4 (Pirrotta, 1988)
digested with XhoI. P element-mediated germline transformation into
a w1118background was performed as described by Spradling (1986),
except that a coinjected plasmid carrying P[ry+, ∆2-3] (Robertson et
al., 1987) served as the transposase source. Genetic crosses were
carried out to generate mutant flies carrying one copy of the P[w+,
pDC4] third chromosome. The second chromosomes of all dia alleles
described in this paper are marked with cn, which allowed us to
confirm that the rescued flies were mutant for dia.
Double-labelling of follicle cells
All steps were carried out at room temperature. Ovaries from 5-day-
old females were dissected out and ovarioles were teased apart in
PBS, fixed in PBT (PBT is PBS, 0.1% Tween-20) + 3.7% formalde-
hyde for 30 minutes, and washed 3× 5 minutes in PBT. To minimize
non-specific propidium iodide staining, the ovaries were incubated
overnight in PBS + 10 µg/ml RNAse (shorter incubations may be suf-
ficient), and washed 3× 5 minutes in PBT. The ovaries were then
stained in PBT + 25 µg/ml BODIPY-Concanavalin A (Molecular
Probes) + 0.5 µg/ml propidium iodide (Sigma). Following three 10
minute washes, the ovaries were viewed immediately. Concanavalin
A strongly binds to the epithelial sheath of the ovariole, which is
closely apposed to the follicle cell layer; the follicle cells could be
visualized by focusing just below this layer.
Cytological examination of dividing neuroblasts
Aceto-orcein squashes of the larval CNS were performed as described
by Karess and Glover (1989).
The male-sterile dia1 allele was identified in a P element screen
for mutations affecting spermatogenesis. Although both sper-
matocytes and spermatids are initially present in dia1 testes,
these cells degenerate and are not replenished. By 5 days after
eclosion, most mutant testes are devoid of germinal contents.
Female fertility is not significantly affected, and the viability
of both males and females is normal. The locus maps to
polytene interval 38E (Castrillon et al., 1993).
During spermatogenesis, a spermatogonium (the product of
a stem cell division) undergoes four rounds of mitotic division
to give rise to a cyst of 16 spermatocytes; meiosis then
produces a cyst of 64 haploid spermatids. Wild-type chromo-
somal segregation and cytokinesis result in spermatids that
each contain two major cytological structures of identical size
and shape (Fig. 1A): a pale round nucleus (arrowhead), and an
adjacent dark nebenkern (arrow). The nebenkern results from
the fusion of all the mitochondria in a spermatid (see Fuller,
1993); its size thus serves as a marker for the amount of
cytoplasm inherited by a spermatid.
The contents of testes from 50 newly eclosed dia1males
were examined. dia1testes contain far fewer spermatids than
normal, due to the reduced germinal content. The majority of
spermatids were large and multinucleate (Fig. 1B). These
abnormal spermatids contained either 2, 4 or 8 nuclei, with the
size of the nebenkerne proportional to the number of nuclei.
Of 159 unelongated spermatids identified, 51 (32%) contained
one nucleus (phenotypically normal), 26 (16%) contained two
nuclei, 81 (51%) contained four nuclei, and 1 (1%) contained
The presence of spermatids containing 2 or 4 nuclei within
a common cytoplasm can be explained by a failure in cytoki-
nesis in one or both of the meiotic divisions. Likewise, the rare
spermatid containing 8 nuclei can be explained by a failure of
cytokinesis in three consecutive cell divisions, the first being
the mitotic division preceding meiosis. The nuclei in defective
spermatids are almost always of wild-type size, indicating that
D. H. Castrillon and S. A. Wasserman
3369diaphanous required for cytokinesis
chromosome segregation is normal in spite of the failure of
cytokinesis. In contrast, in mutants that cause nondisjunction
during meiosis, nuclei are of variable size (Gonzalez et al.,
1989; Karess and Glover, 1989).
The finding that the initial failure of cytokinesis can occur
at distinct points along the spermatogenesis pathway suggests
why the germline in dia1testes is eventually depleted. Since
the 5-9 stem cells present in each testis continually divide to
give rise to spermatogonia (Hardy et al., 1979), failure of
cytokinesis during stem cell divisions should result in the
permanent inactivation of these cells.
In trans to a chromosomal deficiency or a null allele such as
dia2, the dia1allele exhibits an oogenesis phenotype. Although
the viability of such trans-heterozygous adults is normal and
the male germline phenotype is similar to that of dia1males,
female fertility is dramatically decreased. The ovaries of
dia1/dia2females are smaller than wild type. Egg chambers of
all stages are present, the great majority of which contain 15
nurse cells and one oocyte. However, the eggs laid by dia1/dia2
females are shorter than wild-type and have short, fused, or
extra dorsal appendages (Fig. 2). Only 10% of the eggs hatch.
This eggshell phenotype suggested a defect in the somati-
cally derived follicle cells, which surround each developing
egg chamber and secrete the eggshell. Indeed, follicle cells in
mutant ovaries have an abnormal appearance when viewed by
differential interference contrast (DIC) microscopy (Fig. 3B).
Their nuclei vary considerably in size (arrowheads) relative to
wild-type controls (Fig. 3A). Some of the cells appear to
contain two nuclei (arrows).
To visualize the follicle cell layer better, ovaries were
double labelled with a fluorescent Concanavalin A derivative
(to stain plasma membranes) and with propidium iodide (to
stain nuclei) and examined by confocal microscopy. Individ-
ual follicle cells containing two nuclei were frequently
observed in egg chambers from mutant mothers (Fig. 3D,
arrows); such cells were not found in wild-type egg chambers
(Fig. 3C). Therefore, it appears that cytokinesis fails to occur
in some follicle cells. The abnormally large nuclei seen in Fig.
3B are likely due to nuclear fusion following the failure of
Null mutations result in early pupal death and
absence of imaginal discs
mutations, including null mutations, were
generated by imprecise excision of the P element in the dia1
allele. Null alleles such as dia2result in early pupal lethality
and absence of imaginal discs. This phenotype is consistent
with dia being an essential mitotic gene. Due to the presence
of maternal gene products, an embryo with a null mutation in
such a gene can develop into a larva (Gatti and Baker, 1989).
However, since imaginal disc cells divide during the larval
stages, by which time zygotic gene expression is required,
Fig. 1. Cytokinesis defect in
dia1testis. Photographs are
of unfixed testis contents
visualized by phase-contrast
microscopy. Bar represents
10 µm for both A and B.
(A) Part of a 64-cell cyst of
wild-type spermatids. Each
spermatid contains a single
pale nucleus (arrowhead) and
a single dark nebenkern
(arrow). Although this cyst is
intact, spermatid cysts
typically rupture into smaller
groups of cells due to the
absence of a fixation step. (B) Group of six dia1spermatids, each containing four nuclei (arrowheads) associated with a single large nebenkern
(arrow). Inset: single dia1spermatid containing eight nuclei.
Fig. 2. Short egg phenotype. (A) Wild-
type egg; (B) egg laid by dia1/dia2
mother. Eggs from dia1/dia2females are
significantly shorter and somewhat wider
than wild type and have smaller dorsal
appendages. This is an example of a
mildly affected egg.
such larvae will have defective or absent discs and will die at
the onset of pupation.
Homozygotes for less severe alleles such as dia3also die as
early pupae, but third instar larvae contain imaginal discs,
albeit somewhat smaller than normal. Homozygotes for the
weakest lethal allele, dia9, have imaginal discs of normal
appearance but die as late pupae or pharate adults. Very few
(less than 1%) of dia9flies eclose. These flies are sickly and
have a weak ‘rough eye’ phenotype (not shown), consistent
with a mitotic defect that affects a small fraction of cells.
Testes from these surviving adults are almost completely
devoid of germinal content.
Polyploidy in dividing neuroblasts from the larval
The larval central nervous system (brain and ventral ganglion)
is a rich source of dividing cells (neuroblasts) and is the most
suitable tissue for examining mitosis in larvae. Chromosome
morphology and segregation can be examined in aceto-orcein-
stained preparations of the larval CNS (Gatti et al., 1974). In
wild type, a mitotic figure consists of 3 pairs of major chro-
mosomes (Fig. 4A, arrowheads).
In larvae homozygous for a weak lethal allele, dia9, only
a small fraction of neuroblasts are polyploid and the number
of mitotic figures is not affected (Table 1). Homozygotes for
a stronger allele, dia3, exhibit ploidy ranging from 2n, 4n and
8n to extreme hyperploidy (Fig. 4B,C). In addition, signifi-
cantly fewer mitotic figures are present in dia3homozygotes
than in wild type (Table 1). However, the fraction of mitotic
figures in anaphase is the same as in wild type (see Table 1),
whereas mutations that disrupt spindle function result in a
decrease in the number of anaphases (Gatti and Baker, 1989).
In addition, chromosome morphology is generally normal,
although a small number of mitotic figures contain highly
condensed chromosomes. Homozygotes for a null allele,
dia2, have very few mitotic figures (Table 1), and all are
D. H. Castrillon and S. A. Wasserman
Fig. 3. Cytokinesis defect in oogenesis. Top two panels are of unfixed ovaries flattened under a coverslip and viewed with DIC; bar represents
20 µm. Bottom two panels are of fixed ovaries labelled with BODIPY-Concanavalin A and propidium iodide viewed by confocal microscopy;
bar represents 5 µm. (A) Wild-type follicle cells. Nuclei are of equal size. The large egg chamber occupying the lower two-thirds of the field is
at stage 8. (B) dia1/dia2follicle cells, stage 8 egg chamber. Nuclei are of variable size (arrowheads). Some cells appear to contain more than
one nucleus (arrows). The cytokinesis defect is apparent in earlier egg chambers, but is more easily observed in larger, more mature egg
chambers. (C) Wild-type follicle cells. Each cell contains one nucleus. (D) dia1/dia2follicle cells. Some cells contain two nuclei (arrows). Note
that the nuclei within such a binucleate cell are of equal size.
3371 diaphanous required for cytokinesis
enormously hyperploid, similar to the example shown in Fig.
The morphology of anaphase figures provides direct
evidence that cytokinesis is defective in dia3cells. Cleavage
furrows are sometimes evident in wild-type anaphase figures,
especially in well-isolated cells (Fig. 4D). Cleavage furrows
have not been observed, however, in any dia3anaphases and
the cells appear completely round even when the chromosomes
have finished migrating to opposite poles (Fig. 4E).
Despite the cytokinesis defect, chromosome segregation
appears to be relatively normal in polyploid cells. In bipolar
dia3anaphases, the spindles are well organized and the chro-
mosomes are equally segregated, with no lagging chromo-
somes or other abnormalities. This is true even in anaphases
that are clearly hyperploid (Fig. 4F). In more extremely hyper-
ploid cells, anaphases are typically multipolar. Such multipo-
lar spindles have also been observed in other mutants that
produce hyperploid cells (Gatti and Baker, 1989; Karess et al.,
Taken together, the male-sterile, female-sterile and lethal
phenotypes associated with dia mutations demonstrate that dia
is required for cytokinesis in both the soma and germline and
in mitosis as well as meiosis.
Cloning and characterization of dia genomic region
A 140 bp fragment of DNA extending from the 5′ end of the
Fig. 4. Aceto-orcein stained mitotic chromosomes from larval neuroblasts. Cells in A and D are from dia2/+ larvae; cells in B, C, E, and F are
from dia3/dia3 larvae. A, D, E and F are at same magnification; bar represents 3 µm. For B, bar represents 5 µm, and for C, bar represents 18
µm. (A) dia2/+. Wild-type mitotic figure consists of three pairs of major chromosomes and a pair of small fourth chromosomes which are not
readily apparent in this micrograph. (B) dia3/dia3 tetraploid cell containing 12 major chromosomes. (C) dia3/dia3 hyperploid cell. This mitotic
figure consists of hundreds of chromosomes packed together. Inset: magnification of boxed region, revealing individual chromosome.
(D) dia2/+. Wild-type anaphase figure. Cleavage furrow is evident (arrowheads). (E) dia3/dia3. Cleavage furrow is not evident in this mutant
cell in anaphase. (F) dia3/dia3. Even though anaphase is clearly hyperploid, chromosomes are being segregated equally, indicating a
functioning spindle. Again, cleavage furrow is absent.
Table 1. Quantitation of mitotic defects observed in dia
For each genotype, mitotic figures from five larvae were counted and
averaged; standard deviation is shown in parentheses. Due to the high level of
polyploidy, the exact number of chromosomes in most abnormal mitotic
figures could not be determined. Brain size was comparable for all genotypes.