Cell fusion during development

Abstract

Most readers of this review originated from a sperm-egg fusion event. Cell fusion is a process that is crucial at many intersections later during development. However, we do not know which molecules (fusogens) fuse the membranes of gametes to form zygotes, myoblasts to form myotubes in muscles, macrophages to form osteoclasts in bones, or cytotrophoblasts to form syncytiotrophoblasts in placentas. There are five gold standards that can be applied for the identification of genuine fusogens. Based on these criteria, a numerical score can be used to assess the likelihood of protein fusogenicity. We compare distinct families of candidate developmental, viral and intracellular fusogens and analyze current models of membrane fusion.

Full text

Cell fusion during development
Meital Oren-Suissa and Benjamin Podbilewicz
Department of Biology, Technion-Israel Institute of Technology, Haifa, 32000 Israel
Most readers of this review originated from a sperm–egg
fusion event. Cell fusion is a process that is crucial at
many intersections later during development. However,
we do not know which molecules (fusogens) fuse the
membranes of gametes to form zygotes, myoblasts to
form myotubes in muscles, macrophages to form osteo-
clasts in bones, or cytotrophoblasts to form syncytio-
trophoblasts in placentas. There are five gold standards
that can be applied for the identification of genuine
fusogens. Based on these criteria, a numerical score
can be used to assess the likelihood of protein fusogeni-
city. We compare distinct families of candidate develop-
mental, viral and intracellular fusogens and analyze
current models of membrane fusion.
Introduction
Whether a plant, protist or primate, sexual reproduction
begins with the simple process of membrane fusion, forming
the first cellular unit of the organism by mixing two cells
with different genetic and cytoplasmic contents [1]
(Figure 1). During early embryonic development, cell fusion
is involved in muscle formation [2], and continuous devel-
opment depends upon correct fusion events. For example, in
mammals, macrophage fusion forms the bone-resorbing
osteoclasts [1], and trophoblast cells {see Glossary} fuse to
generate the placentalsyncytiotrophoblast layer that serves
as a barrier between maternal and fetal environments [3].In
nematodes, fused epithelial cells form barriers and con-
strain cell migration, by forming the skin, vulva, uterus,
hymen and glands of the worm (Figures 2)[4]. In some
sponges, cells (blastomeres) fuse early in development, and
somatic cell fusions have been detected in fungi, leeches,
insects and many other organisms (Figure 3)[5–8].
Most of our knowledge on the mechanism of cell fusion
comes from intracellular vesicle fusion and viral–host cell
fusion [9]. Cell fusion processes generally involve bringing
two lipid bilayers into close proximity, followed by the
formation of a fusion pore and cytoplasmic mixing, to form
a syncytium (Figure 4). The molecules that take part in the
fusion reaction can be divided into three groups according
to their involvement in the fusion stages. The first group is
composed of molecules that function before fusion, such as
adhesion molecules. The second group includes the
mediators of fusion that directly rearrange the lipid
bilayers and lead to the formation of fusion pores. These
proteins are defined as fusogens and are the focus here.
The third group consists of molecules that lead to extension
of the fusion pores, and to complete disassembly of
membranes.
Developmental cell fusion is still poorly understood, and
although many proteins required for various cell–cell
fusion events have been identified over the past decades,
only a few have been thoroughly characterized and defined
as fusogens. Defining whether a protein is a true fusogen
has proven to be a difficult task. A primary obstacle is how
to distinguish between proteins which function in fusion
rather than adhesion. In sperm–egg fusion, for example,
impairment of either fusion or adhesion will lead to a
reduction in fertility and, as a consequence, to a small
brood size. Recently, the generation of knockout mice has
proved useful in identifying proteins directly involved in
sperm–egg fusion, and refuted several proteins previously
identified as fusogens [10].
Here, we suggest the use of five defined criteria for the
identification of genuine fusogens (Box 1). Although some
proteins which meet these criteria have been identified for
viral and intracellular membrane fusion, for cell–cell fusion
only a few exist that meet most criteria (Table 1). We discuss
current candidate fusogens for various cell fusion events in
different organisms, and evaluate how well the best candi-
dates meet the five criteria for bona fide fusogens.
Gamete fusion during mating and fertilization
Sexual reproduction usually occurs through a sequence
of fast occurring events that include gamete attraction
Review TRENDS in Cell Biology Vol.17 No.11
Glossary
Acrosome reaction: The binding of the sperm to the zona pellucida triggers the
fusion of the acrosomal vesicle found at the sperm head with the sperm cell
membrane. The fusion results in the release of the acrosomal proteolytic
enzymes that lyze the zona pellucida, enabling the sperm to reach the egg
membrane.
ADAM proteins: a family of membrane proteins with a disintegrin and
metalloprotease domain. The ADAMs are involved in diverse processes such
as development and cell–cell interactions.
Cytotrophoblast: the inner layer of trophoblast cells.
FF family: a novel family of nematode-specific type-1 membrane proteins,
necessary and sufficient for cell fusion. These proteins contain a conserved
signature of cysteines at their ectodomain.
HERV: the human endogenous retroviral protein family.
HERV-W: a new family of HERVs, that is not replication competent.
IgSF proteins: the immunoglobulin superfamily is a large group of proteins
that share structural similarities, all possessing the immunoglobulin domain.
The family includes cell surface and soluble proteins involved in adhesion,
recognition and binding of cells.
SPE-9: defective spermatogenesis 9 protein.
FER-1: spermatogenesis or fertilization defective 1 protein.
Syncytiotrophoblast: multinucleated layer formed by the fusion of trophoblast
cells.
Trophoblast cells: epithelial cells that form the outer layer surrounding the
blastocyst and attach the embryo to the uterus wall. The trophoblasts provide
nourishment from the mother and develop into a large part of the placenta.
Utse: an ‘H’-shaped syncytium (hymen) that connects the uterus to the vagina
and attaches to the lateral epidermal seam cells in C. elegans.
Zona pellucida: a thick glycoprotein outer layer surrounding the mammalian
oocyte. The sperm binds to the zona pellucida, and this binding triggers the
acrosomal reaction, which enables the sperm to penetrate.
Corresponding author: Podbilewicz, B. (podbilew@tx.technion.ac.il).
Available online 5 November 2007.
www.sciencedirect.com 0962-8924/$ – see front matter ß2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.tcb.2007.09.004

and migration, recognition, attachment and adhesion,
followed by fusion. Gamete fusion is a unique heterotypic
fusion process that merges two genetically different cells
into one. Studying gamete fusion requires detailed
analysis of membrane dynamics, using light and electron
microscopy. However, owing to the dynamic nature of
these events, it is still difficult to differentiate between
gamete fusion failure and defects in the processes that
precede fusion.
Yeast mating and membrane fusion
Mating in the yeast Saccharomyces cerevisiae involves
secretion of pheromones by a- and a-mating type haploid
cells. Pheromone detection activates a signaling cascade
that leads to the activation of gene transcription, cell cycle
arrest and cell polarization. The mating partners adhere,
the cell wall is degraded and fusion occurs [11]. A reverse
genetic screen designed to identify uncharacterized phero-
mone-regulated membrane proteins [12] uncovered a can-
didate fusogen, Prm1, a multispanning transmembrane
protein, which localizes to the site of membrane fusion
after pheromone induction (Figure 1a). In prm1 mutants,
the cell wall is degraded but the frequency of fusion be-
tween the apposing membranes is decreased. However,
Prm1 is not essential for fusion because a mutation in
both mating partners results only in 50% fusion failure,
Figure 1. Gamete fusion in various organisms. (a) Prm1 (red) is localized to the site of cell fusion. In prm1 mutants, the cell wall degrades but fusion does not occur
[1,12,13].(b) Sperm–egg fusion in C. elegans requires the expression of EGG-1 and EGG-2 in the egg, and SPE-9, SPE-38 and SPE-42 in the sperm [18,21,22,24].Itis
unknown whether sperm proteins interact with oocyte EGG-1 and EGG-2, and the ligands for SPE-9, SPE-38 and SPE-42 have not yet been identified. (c) In mice, CD9 is
expressed in the microvillar region of the egg [39]. Izumo becomes detectable after the acrosomal reaction on the sperm surface [35]. It is not clear whether Izumo and CD9
interact and whether this interaction is required for the fusion process.
Figure 2. Epithelial cell fusion in humans and worms. (a) In the formation of the human placenta, secondary trophoblast fusion involves fusion of cytotrophoblasts with the
syncytiotrophoblast layer. Syncytins are the best fusogenic candidates identified for these cell fusion events, and several receptors have been identified [3,46].(b) EFF-1 is
necessary and sufficient for homotypic epithelial cell fusion in C. elegans [15,56]. In the embryo, most of the dorsal cells and some of the ventral cells (green) fuse to form
the hypodermal syncytia [4,87]. Modified, with permission, from Ref. [29].(c) In C. elegans, fusion between the AC and the utse occurs only after they both express AFF-1.
Other epithelial fusions are also mediated by AFF-1 in C. elegans [16]. Modified, with permission, from Ref. [16].
Figure 3. Mesodermal cell fusions in flies and mammals. (a) Myoblast fusion in Drosophila begins with recognition and attachment of the founder cell and fusion-
competent cells. Sns is expressed on the surface of the FCM, and Duf/kirre and Rst/IrreC are expressed on the surface of the founder cell. The first round of fusion leads to
the formation of a precursor myotube, and additional rounds of fusion between FCMs and nascent myotubes form a mature myotube [1].(b) Mononuclear osteoclasts fuse
to form bone-resorbing, multinucleate osteoclasts in mammals. DC-STAMP is required for the fusion process, and fusion can occur only if DC-STAMP is expressed in one of
the fusing cells, suggesting interaction through an unknown ligand [23].
538 Review TRENDS in Cell Biology Vol.17 No.11
www.sciencedirect.com

and when missing from one mating cell, the fusion levels
decrease only slightly [12].
An additional screen for mutants that enhance the
prm1-dependent fusion-failure phenotype identified Kex-
2[13].kex2 mutations enhance the prm1 phenotype but do
not completely block fusion. Its molecular function as a
Golgi protease suggests the existence of a proteolitically
activated protein involved in the cell fusion machinery. For
the above reasons, it is not clear if Prm1 is the yeast
fusogen, or whether Kex-2 is responsible for the processing
of an unidentified fusogen. Recently, the activated phero-
mone receptors were shown to have a novel and surprising
late role in yeast mating, apparently at the stage of cell
wall digestion and membrane juxtaposition before mem-
brane fusion [14].
Fertilization in Caenorhabditis elegans
In recent years, the nematode C. elegans has proven to
be a useful model for studying the process of
fertilization. Sperm–egg fusion in C. elegans is indepen-
dent of known fusogens that function in somatic fusion
[i.e epithelial fusion failure 1 (EFF-1) and anchor cell
fusion failure 1 (AFF-1) [15,16];seelater].C. elegans
sperm lacks a flagellum and an acrosome, and uses a
single pseudopod for amoeboid motility. Nonetheless, it
carries out all essential reproduction steps common to
most spermatozoa: migration to the site of fertilization,
recognition, adhesion and gamete fusion. Moreover,
because C. elegans hermaphrodites are self-fertile, iso-
lation of mutants affected only in spermatogenesis is
relatively simple because such mutant hermaphrodites
Figure 4. Simplified comparison between proposed models for viral, intracellular and developmental membrane fusion mechanisms. (a) The Env proteins of retroviruses
(e.g. HIV) and probably Syncytin exist in the native conformation before being triggered by binding to a receptor (not shown) that induces a dramatic conformational
change. This conformational change involves the formation of coiled coils and insertion of amphipathic fusion peptides (red) into the target membranes (only two
fusogenic trimeric proteins are shown). Formation of hairpins and six-helix bundles induce tightening of the membranes followed by membrane merger [88–90].(b) Native
heterotrimeric tSNAREs (blue, green and violet) bind to the vSNAREs (red). Activation by coiled-coil formation assembles the complexes in a prefusion conformation that
involves bending and tightening of the membranes. Zippering of the complexes is believed to induce hemifusion (not shown), followed by complete membrane merger
(postfusion) [84].(c) Homotypic model for EFF-1-mediated fusion based on heterotypic hairpins shown in (a). EFF-1 proteins (only two monomers are shown) undergo
conformational changes and insertion of putative fusion peptides or loops (red) into the apposing membrane. Hypothetical refolding and hairpin formation results in
membrane tightening followed by hemifusion (not shown) and complete merger of the membranes. (d) Model for homotypic binding between FF proteins based on the
zippering mechanism shown in (b). Homotypic interactions activate EFF-1 proteins in both cell membranes. Assembly of complexes bends the membranes (only two trans-
homodimers are shown). Zippering of the complexes induces fusion through hemifusion (not shown). In addition to the four models shown here, a hybrid mechanism
involving homotypic zippering of hairpins (c+d) or other novel mechanisms are conceivable for FF and other fusogens [58].
Review TRENDS in Cell Biology Vol.17 No.11 539
www.sciencedirect.com

produce viable offspring when mated with a wild-type
male [17].
Genetic screens have identified several genes required
for fertilization. The best characterized C. elegans
candidate fusogen so far is SPE-9 (see Glossary), a
sperm-specific single transmembrane protein containing
ten epidermal growth factor (EGF)-like repeats, with sim-
ilarity to ligands of the Notch receptor [18] (Figure 1b). spe-
9mutant hermaphrodites lay unfertilized oocytes, whereas
males are completely sterile. spe-9 mutant sperm show
normal morphology and motility, and are able to compete
with hermaphrodite sperm. After sperm activation, SPE-9
localizes to the pseudopod [18,19]. Although spe-9 is
required specifically for fertilization, it is unclear whether
it is required for the membrane fusion step or for earlier
stages such as recognition and tight adhesion, as might be
interpreted from function of other EGF-motif-containing
proteins that are known to mediate adhesive or ligand–
receptor interactions. However, no receptor for SPE-9 on
the surface of oocytes has yet been identified [20].
Additional sperm proteins required for fertilization are
SPE-38 and SPE-42 (Figure 1b). Similarly to spe-9
mutants, spe-38 and spe-42 mutant sperm are fully motile
and able to migrate to the site of fertilization but are
incapable of fertilizing oocytes [21,22]. SPE-38 and SPE-
42 proteins share structural features with proteins in other
phyla that have been linked to membrane fusion. SPE-38 is
predicted to encode a novel four-pass integral membrane
protein with some structural similarities to the yeast Prm1
[12] and [21]. However, SPE-38 is not a member of the
tetraspanin family, which has been implicated in the
process of mammalian sperm–egg fusion (see later).
SPE-42, a predicted seven-pass transmembrane protein,
has a DC-STAMP-like domain. The DC-STAMP protein is
required for osteoclast and macrophage fusion (see later)
[23]. Although their precise function in fertilization, or
whether they interact with one another, is unknown, it
is possible that these proteins function during the fusion
process and that their function has been conserved in
evolution.
The first-discovered egg components required for
fertilization were EGG-1 and EGG-2 type II transmem-
brane, low-density lipoprotein (LDL) receptor repeat-con-
taining proteins [24]. Hermaphrodites that lack these
mutually redundant proteins are completely sterile. A
model in which EGG-1 and EGG-2 function as receptors
for unknown sperm ligands can be deduced from their
specific egg surface expression and the molecular nature
of LDL receptor-related molecules. So far, no interaction
has been established between any egg and sperm com-
ponents in C. elegans, so it will be interesting to see
whether any interaction exists between SPE-38, SPE-42
or SPE-9 and EGG-1 or EGG-2, which seem to be the best
candidate proteins to be directly involved in C. elegans
sperm–egg fusion.
Fertilization in Drosophila
Although Drosophila melanogaster has served as a
powerful model organism for over a century, little is known
about the mechanisms underlying sperm–egg fusion. More
than 30 years ago, Perotti’s [25] classic ultrastructural
studies of fertilization in Drosophila revealed that the
whole sperm enters the egg, and showed that the sperm
membrane around the tail is intact. However, because the
ultrastructure of the sperm membrane at the fusion stage
could not be visualized, this does not serve as sufficient
proof for lack of fusion between sperm and oocyte mem-
branes (M-E. Perotti, personal communication) [25]. Intri-
guingly, several recent papers have claimed that the sperm
enters the egg without membrane fusion, by puncturing a
hole in the oocyte membrane, followed by sperm membrane
breakdown [26,27]. Plasma membrane breakdown and
vesiculation probably occur after membrane fusion events
required to remove and recycle the sperm plasma mem-
brane, as occurs in other cell fusion processes in mammals,
flies and worms (see later) [28,29]. Plasma membrane
breakdown in Drosophila requires the proteins Sneaky
and Misfire [26,27]. Sneaky shares domain similarity with
DC-STAMP and SPE-42 (Table 1), and one of the homologs
of Misfire is C. elegans FER-1, a protein required for the
fusion of a membranous organelle during maturation of
spermatids to motile spermatozoa [30]. Taken together,
there is no proof for the absence of sperm–egg fusion in
Drosophila. The homology between the proteins required
for membrane breakdown in Drosophila and proteins
implicated in fusion in other organisms implies conserved
protein function. The proteins that mediate sperm–egg
fusion in Drosophila remain to be identified.
Sperm–egg fusion in mammals
Fertilization in mammals includes migration and
activation of sperm in the female tract, penetration through
cell layers surrounding the egg, recognition, binding and
induction of the acrosomal reaction, which enables penetra-
tion through the zona pellucida, and, finally, sperm–egg
fusion [31]. Although much attention has been given over
the years to several fusion candidate proteins such as
the ADAM proteins on the sperm (fertilins) and the egg
integrins, knockout mouse experiments showed that
these proteins are not essential for cell fusion [10,32,33].
Recently, the immunoglobulin superfamily (IgSF)
Box 1. The gold standards of a bona fide membrane
fusogen
Several criteria must be met to define a protein properly as a bona
fide fusogen [58]. First, the protein must be essential for membrane
fusion. Second, the fusogen should be expressed at the site and
time of fusion. Third, the molecule should be sufficient to fuse cells
that normally do not fuse in situ. Fourth, expression of the protein in
heterologous cells must lead to fusion of these cells, scored by the
formation of multinucleate cells. Fifth, biochemical reconstitution of
the candidate fusogen(s) into an isolated in vitro system such as
liposomes must fuse them. Exceptions might exist – for example, for
proteins that are fusogens but have not been shown to be essential
for fusion – because there are several redundant proteins. Addi-
tional examples include the formation of protein complexes
required for fusion, or the requirement of fusion cofactors. These
proteins cannot be evaluated under the same criteria, and their
identification as fusogens will be more complicated.
Each candidate fusogen scores two points for fulfilling each
criterion. Proteins that fuse cells in a heterologous system will score
two points for this criterion only if it has been shown that
multinucleation was a result of fusion and not failed cytokinesis
(Table 1).
540 Review TRENDS in Cell Biology Vol.17 No.11
www.sciencedirect.com

Table 1. Examples of candidate fusogens
a
Candidate
fusogen
Family Organism Expression Fusogenic
score
Essential
for fusion
Expressed
at the time
and place
of fusion
Sufficient
in situ
Fuse cells
(heterologous)
In vitro
liposome
fusion
Homotypic/
heterotypic
Refs
Env Class I
enveloped
Viruses
HIV CD4+ cells 10 + + + + + Heterotypic [83,91]
HA Influenza Lung, epithelia 10 + + + + + Heterotypic [9,83]
vSNAREs,
tSNAREs
SNAREs Eukaryotes Intracellular 10 + + + + + Heterotypic [9,84]
Syncytin-1 Syncytins
(retroviruses)
Homo sapiens Cytotrophoblasts and
syncytiotrophoblast
7+ ND + + Heterotypic (?) [44,45,50]
EFF-1 FF C. elegans Epithelia, vulva, pharynx 8 + + + + ND Homotypic [15,56–58]
AFF-1 Epithelia, AC–utse, vulva 8 + + + + ND ND [16]
Examples of plasma membrane proteins required for fusion
Duf/Kirre IgSF D. melanogaster Founder cell 2 + ND Homo/Hetero
b
[66,69,92]
Sns Fusion-competent
myoblasts (FCM)
4++  ND Heterotypic [68,69]
Rst/IrreC Founder cell and FCM 2 + ND Heterotypic [67]
IZUMO M. musculus Sperm or testis 4 + + ND ND ND Heterotypic [35]
CD9 Tetraspanin M. musculus Multiple tissues, oocyte
microvilli
4 + + ND ND ND Heterotypic (?) [36,37]
SPE-38 4TM C. elegans Sperm 4 + + ND ND ND Heterotypic [21]
Prm1 5TM S. cerevisiae Mating cells 3 +NDND NDND [12]
SPE-42 Multispan
DC-STAMP-
like
C. elegans Sperm 2 + ND ND ND ND Heterotypic [22]
DC-STAMP M. musculus Dendritic cells,
macrophages, osteoclasts
3+ND ND ND Heterotypic [23,93]
SPE-9 EGF repeats C. elegans Sperm 4 + + ND ND ND Heterotypic [18,94]
EGG-1;
EGG-2
LDL repeats C. elegans Oocyte 4 + + ND ND ND Heterotypic [24]
a
The fusogenic score (0–10) was calculated using the following scoring system: + (requirementfulfilled) = 2 points, (requirement not fulfilled) = 0 points, ND (not determined) = 0 points and (requirement partially fulfilled) = 1 point.
b
Abbreviations: homo/hetero, homotypic/heterotypic molecular recognition; TM, transmembrane; utse, uterine seam cells;?, contradictory results.
Review TRENDS in Cell Biology Vol.17 No.11 541
www.sciencedirect.com

membrane glycoprotein Izumo was identified using a mouse
monoclonal antibody that specifically inhibits sperm–egg
fusion [34] (Figure 1c). Izumo is sperm specific, and is
detectable on the sperm surface only after sperm have
undergone the acrosomal reaction [35]. Whereas Izumo-
deficient female mice show normal fertility, Izumo-deficient
male mice are completely sterile, despite normal mating
behavior and normal sperm migration and motility.
Accumulation of sperm in the perivitelline space suggests
that Izumo is required for fusion but not for adhesion or zona
penetration. Izumo() sperm injected directly into wild-type
eggs produced fertilized eggs and normally developed
embryos, ruling out an Izumo-induced defect in later de-
velopment. The human Izumo homolog is also detectable
specifically in human sperm. When adding polyclonal anti-
bodies against the human Izumo to a xeno-species fusion
system containing zona-free hamster eggs and human
sperm, no fusion occurs [35]. Thus, it seems that Izumo
represents the best sperm–egg candidate fusogen identified
so far, being the first mammalian sperm proteinessential for
fusion (Table 1). However, the question of whether Izumo is
the sperm–egg fusogen requires further investigation.
On the egg plasma membrane, CD9 is a tetraspanin
protein family member that has been shown to be essential
for fertilization (Figure 1c). Three different laboratories
have generated CD9 knockout mice, which show severely
reduced fertility (up to 98%) [36–38]. CD9 contains four
transmembrane domains and two extracellular loops of
unequal size. Many possibilities exist for the involvement
of CD9 in sperm–egg fusion. CD9 localizes to the micro-
villar region of the oocyte [39], the area to which the sperm
binds and later fuses with. In CD9 knockout mice, micro-
villi morphology is impaired and their curvature altered
[39]. In addition, it was shown that CD9 associates with the
IgSF proteins EWI-2 and EWI-F, which bind to actin
filaments in microvilli [31]. This association links CD9
to the actin cytoskeleton network, suggesting a possible
function in remodeling of surface curvature, to enable
membrane fusion. Although known to associate with IgSF
molecules, no connection between CD9 and Izumo has yet
been shown (Figure 1c). Recent work has shown that
membrane fragments containing CD9 from the oocyte
plasma membrane are transferred to the fertilizing sperm
found in the perivitelline space [40]. This intriguing
phenomenon might be important for reorganizing mem-
brane domains on the sperm necessary for the fusion
process.
Another link between the tetraspanin CD9 and the
fusion machinery comes from in vitro assays in which
knocking down expression of CD9 and an additional tetra-
spanin, CD81, resulted in enhanced syncytia formation
and viral entry induced by HIV-1 envelope proteins [41],
and, in another case, knockdown enhanced fusion of mono-
nuclear phagocytes [42]. Thus, CD9 might be involved in
various aspects of the fusion process, with pro- or anti-
fusogenic activities in different experimental systems.
Early development
Epithelial fusion
1. Trophoblast fusion in placenta formation Trophoblast
fusion in the mammalian placenta can be divided into
two modes. In the first, neighboring trophoblasts of the
blastocyst undergo cell–cell fusion to form the syncytiotro-
phoblast (Figure 2a). The second occurs when cytotropho-
blasts fuse with the syncytiotrophoblast for expansion and
maintenance of the syncytiotrophoblast, which serves as a
barrier between maternal and fetal blood vessels [3]. The
best characterized candidate fusogen for trophoblast fusion
in humans is Syncytin-1, encoded by the HERV-W retro-
viral element, and belonging to the class I viral fusogens, as
HIV [43] (Table 1). Syncytin-1 transfection into various cell
lines results in the formation of multinucleate cells [44,45],
demonstrating the fusogenic properties of Syncytin-1. Con-
flicting results have been obtained regarding the cellular
localization of Syncytin-1 but the majority of expression
seems to be in the syncytiotrophoblast and the cytotropho-
blasts [46].
Several receptors for Syncytin-1 have been identified,
including the D mammalian retrovirus receptor RDR [44]
and ASCT1 [47–49] (Figure 2a). Rat sarcoma cells, which
do not fuse when expressing Syncytin-1, showed elevated
fusion activity when cotransfected with the RDR receptor,
suggesting a heterotypic mechanism of fusion for Syncytin-
1[44]. Although Syncyin-1 inhibition in primary cultures
by antisense oligonucleotides [50] and by anti-Syncytin
antiserum in human trophoblastic cell line leads to a
decrease in cell fusion, it is unclear whether Syncytin-1
is essential for trophoblast fusion in situ.
Another, less characterized retroviral gene expressed
in cytotrophoblastic cells is Syncytin-2, encoded by the
retroviral env gene HERV-FRD [51,52]. Transient trans-
fection of cell lines with Syncytin-2 also leads to ectopic
fusion [51], suggesting a fusogenic role for Syncytin-2.
Syncytin-1 and -2 are found only in highly evolved
primates [51], leading to the question of how placenta–
trophoblast fusion occurs in other species, such as rodents.
Recently, two murine viral envelope genes have been
discovered, encoding Syncytin-A and -B [53]. Although
evolutionarily distinct from the genes encoding Syncy-
tin-1 and -2, both proteins display fusogenic activity fol-
lowing expression in transfected cells. No receptor has yet
been identified for Syncytin-A and -B [53]. As occurs in the
case of Syncytin-1, inhibition of Syncytin-A by antisense
oligonucleotides and anti-Syncytin-A antiserum in cul-
ture leads to a decrease in cell fusion [54].Structural
and functional studies of Syncytin-A show that the heptad
repeats region can form a stable a-helical complex, typical
of class I viral fusion proteins [55].
It seems that at least twice during evolution, retroviral
genes have adapted independently to perform apparently
similar functions in trophoblast fusion. The mysterious
existence of two independent Syncytins, in two different
species, poses an intriguing question: is there a specific
requirement for two proteins, each complementing the
fusogenic activity of the other, or is the dual existence a
fail-safe mechanism for such an important process? In
summary, Syncytins in rodents and primates are related
to retroviral glycoproteins from the class I viral fusogens
and are sufficient to fuse heterologous cells in culture. To
determine whether mouse Syncytins are also essential for
trophoblast fusion, detailed analyses of knockout mice of
both Syncytins is required (Table 1).
542 Review TRENDS in Cell Biology Vol.17 No.11
www.sciencedirect.com

2. C. elegans skin formation In C. elegans, epithelial cells
start forming the syncytial hypodermis 340 min after the
first cleavage [4]. The hypodermis is the outer layer of
the body, which establishes the body shape, secretes the
cuticle and stores nutrients ([4] and the Wormatlas website,
http://www.wormatlas.org). Many cells in the hypodermis
form large syncytia; for example, 23 hypodermal cells
fuse during embryonic development and 116 additional
cells fuse to this giant syncytium during larval
development (Figure 2b). Genetic screens for fusion
failure resulted in the identification of EFF-1, a type I
membrane protein essential for epithelial cell fusion [15].
In eff-1 mutants, most of the fusion events do not occur,
affecting the elongation and motility of larvae and adults.
Further research demonstrated that eff-1 is also sufficient to
induce embryonic and postembryonic cell fusions in C.
elegans [56,57]. Moreover, expression in heterologous
tissue culture cells results in ectopic fusion and the
formation of multinucleate syncytia [58]. Unlike viral
fusion and Syncytin-mediated fusion, which promote
fusion through a heterotypic mechanism, it was
demonstrated that eff-1 is required for fusion in both
fusing cells, both in vivo and in tissue culture [58].In
addition, green fluorescent protein-tagged EFF-1 has
been shown to accumulate at the contact zone between
fusing cells, before fusion [57]. Meeting four of the five
gold standards for proteins involved in membrane fusion,
EFF-1 is most likely to be a genuine fusogen (Table 1;Box 1).
Late development
Anchor cell fusion in C. elegans
In addition to the hypodermis, fusion was observed in the
formation of the vulva, excretory gland, male tail, anchor
cell (AC), uterus and pharynx [4,59–61]. In the postem-
bryonic stages, eff-1 is required for fusion of additional
epithelial cells, such as the hypodermis and the seam cells,
most of the vulval cells and cells of the tail [15]. However,
eff-1 is not required for fusion of the AC to the uterine seam
cell (utse) or hymen [16,56].InC. elegans, the AC is
responsible for establishing the physical connection be-
tween the uterus and the vulva. Serving as an ‘organizing
center’, the AC induces cell fate change, and differentiation
of the vulva precursor cells and uterine cells [59,62]. Uter-
ine–vulval continuity is further established by fusion of
eight uterine p–cell daughters, generating the utse syncy-
tium, and fusion of the AC to this syncytium leads to the
formation of a thin hymen layer between the vulva and the
uterus [59,60] (Figure 2c). The first laid egg ruptures this
hymen, establishing a direct vulval–uterine connection. A
genetic screen for egg-laying defects identified AFF-1, a
type I transmembrane protein [16]. AFF-1 is essential for
AC fusion, for fusion of the vulval rings vulA and vulD, and
for fusion of the hypodermal seam cells into a continuous
row, forming two extended seam cell syncytia on the lateral
sides of the animal. Similarly to EFF-1, AFF-1 expression
in transfected Sf9 cells is sufficient to induce multinucleate
cell formation (Table 1). AFF-1 and EFF-1 are homologs
with moderate sequence conservation but they share a
remarkable putative structural conservation, in that all
16 cysteines and 11 out of 22 proline residues are conserved
in their ectodomains. Together, EFF-1 and AFF-1 represent
the founding members of a family of developmental
fusogens, mediating cell–cell fusion [16]. As for the Syncy-
tins, it is not clear why two fusogens are required to
perform similar fusion events. It will be interesting to see
whether the constitution of the FF family will lead to the
discovery of additional family members that fuse cells in
different organisms.
Continuous development, tissue remodeling and
repair
Myoblast fusion
During myogenesis in Drosophila, mononucleate myoblasts
fuse to form multinucleate muscle fibers (Figure 3a). This
process involves ‘founder cell’ myoblasts, and ‘fusion com-
petent myoblasts’ (FCM). The founder cell functions as the
organizer of muscle formation, attracting the FCMs, which
migrate toward it and, after recognition, fuse with the
founder cell. Embryonic myoblast fusion starts with the
founder cell fusing with one to two FCMs, to generate a
muscle precursor. In the subsequent steps, FCMs fuse with
the precursors, to generate multinucleate myotubes [63].
Extensive reviews of Drosophila myoblast fusion are pro-
vided by Chen et al. [1] and Horsley and Pavlath [2].
Using electron microscopy, Doberstein et al. [28]
identified an intermediate step in myoblast fusion of paired
vesicles that align on apposing membranes before fusion. It
has been proposed that the myoblast fusogen resides
within those vesicles, and localizes to the plasma mem-
brane before membrane fusion [28,64]. Recently, Estrada
et al. [65] identified Singles Bar, a candidate protein
required for fusion of the paired vesicles to the plasma
membrane.
Several IgSF proteins have been found to be required for
myoblast fusion in Drosophila: Dumbfounded (Duf), also
called Kin of Irregular chiasm C (Kirre) and Roughest
(Rst), also called Irregular chiasm C (IrreC), in the founder
cell, and Sticks and stones (Sns) in the FCM [66–68].
Although single mutations of Duf and Rst show that they
are not essential for myoblast fusion, when both proteins
are mutated, severe myoblast fusion defects are observed
[67], suggesting redundant function in myoblast formation.
Sns was shown to be essential for cell fusion [68], and to
interact with Duf in vitro [69]. However, expression of Sns,
Duf and Rst in S2 culture cells mediates adhesion and
aggregation but not fusion, suggesting that Duf, Rst and
Sns function by mediating recognition, attraction or tight
binding rather than cell fusion [69]. To date, no candidate
proteins have been found to be both necessary and suffi-
cient for myoblast fusion (Table 1).
Recently, more light has been shed on the mechanism of
myoblast fusion in Drosophila, linking myoblast fusion and
cytoskeleton remodeling [64,70], through D-WIP/Sltr and
WASp–ARP2/3 association and F-actin polymerization. D-
WIP/Sltr is recruited to the sites of myoblast fusion and
regulates the actin polymerization machinery.
Similarly to Drosophila, mammalian muscle
development is also a multistage process, in which mono-
nucleate myoblasts fuse with one another to form nascent
myotubes. Multinucleate myotubes undergo additional
rounds of fusion, leading to an increase in size and matu-
ration [2]. As for the sperm–egg fusion, many proteins have
Review TRENDS in Cell Biology Vol.17 No.11 543
www.sciencedirect.com

been implicated in myoblast fusion, including members of
the ADAM family, integrins and various adhesion mol-
ecules. None of these proteins, however, have been shown
to be essential for myoblast fusion, and knockout of most of
them in mice revealed no apparent musculature defects
[71–75]. Recently, Kirrel, a homolog of Drosophila Duf/
Kirre and Rst/IrreC, has been shown to be required for
myoblast fusion in Zebrafish [76]. In morpholino-injected
embryos directed against Kirrel, a large number of unfused
mononucleate monocytes were detected. These findings
reveal an evolutionary conservation in the regulation of
muscle fusion. Nonetheless, the myoblast fusogen remains
to be identified.
Macrophage fusion: multinucleation of osteoclasts
Macrophages fuse and differentiate to form multinucleate
osteoclasts (bone-resorbing cells) in bones or giant cells in
different tissues (Figure 3b). Both types of syncytia have
resorption abilities, such as resorption of the mineralized
matrix of the bone or of invading pathogens, and thus
have an important role in bone remodeling and immune
defense [1]. Several proteins have been identified that have
a role in macrophage fusion. Macrophage fusion receptor
(MFR) and its ligand CD47 are both IgSF cell-surface
proteins. Whereas MFR expression rises before fusion,
the levels of CD47 remain constant [1,77]. An extracellular
peptide form of CD47 binds to macrophages, associates
with MFR and prevents multinucleation [78], and the
extracellular domain of MFR prevents fusion of macro-
phages in vitro [77]. However, no further evidence exists to
support a direct involvement in the fusion machinery.
CD44 is a hyaluronan receptor adhesion protein involved
in osteoclast fusion. It is also expressed strongly at the
onset of fusion, and antibodies against CD44 were shown to
inhibit osteoclast formation [79], although knockout mice
do not support a direct involvement in the fusion process
[80]. Recently, a seven-transmembrane protein, DC-
STAMP, was found to be required for cell fusion of osteo-
clasts and macrophage giant cells. DC-STAMP knockout
mice do not form multinucleate osteoclasts, and show
increased bone mass (less bone-resorbing activity) and a
mild osteopetrosis phenotype [23]. DC-STAMP was shown
to be necessary in only one of the fusing cells for fusion to
occur. However, DC-STAMP localization to the plasma
membrane during fusion, and the identification of a ligand,
remain to be explored (Figure 3b; Table 1). Despite the
present lack of definitive evidence, such as fusion of heter-
ologous cells, DC-STAMP is emerging as a promising
candidate for the macrophage and osteoclast fusogen.
Concluding remarks and future directions
Cell fusion failure in humans might be associated with
diseases such as myopathies, osteopetrosis, infertility and
preeclampsia. However, the involvement of mutated can-
didate fusogens in the pathogenesis of these diseases is
speculative. The potential therapeutic applications of cell
fusion in tissue repair using stem cells, and in the treat-
ment of some cancers, are controversial [81,82] and require
fusogen identification and mechanistic characterization
before we can understand and modulate these physiologi-
cal and pathological cell fusion events.
Currently, there are two major models by which
genuine fusion proteins (fusogens) function. The first
model is based on studies of viral fusogens belonging to
three structural classes. Fusogens of enveloped viruses
form ‘hairpins’, which ‘fold back’ through diverse confor-
mational changes to bend and eventually fuse the mem-
branes. Figure 4a shows a simplified model for class I and
retroviral fusogens. Fusogen activation involves a confor-
mational change in which coiled coils cause the fusogens
to become erect, resulting in the insertion of the amphi-
pathic fusion peptides or loops into the target membrane.
Folding of the trimeric fusogens results in the formation of
six-helix bundles, which bend the membranes toward
each other, leading to hemifusion (fusion of the outer
leaflets) followed by a complete merger of the membranes.
Viral fusogens from classes II and III form hairpin struc-
tures using different structural intermediates [83].
Figure 4b shows the second model for the merger of
intracellular membranes by the formation of heterotri-
meric tSNAREs. This assembly induces ‘zippering’ – that
is, approach of the membranes followed by hemifusion,
and complete merger of the bilayers without the insertion
of amphipathic peptides [84]. Based on these models
(Figure 4a,b), we suggest three alternative hypothetical
FF homotypic fusion mechanisms. The first model
involves fusion peptide insertion and hairpin formation
(Figure 4c). The second hypothesis comprises zippering of
FF proteins with or without amphipathic fusion peptides
(Figure 4d). A third model (not shown) involves a combi-
nation of the previous two (hairpins–zippers). Finally, it is
conceivable that FF-mediated fusion follows a completely
different mechanism. Dynamic structural analyses of FF
proteins during fusion, combined with the identification of
functional motifs in these fusogens, will help to define the
mechanism of FF-mediated cell fusion. This cell fusion
process might serve as a paradigm for the yet unidentified
developmental fusogens. Because this mechanistic
characterization is complex, it should be implemented
only on strong fusogen candidates, fulfilling five gold
standards (Box 1):
(i) Is it essential for fusion in vivo?
(ii) Is it expressed at the time and place of fusion?
(iii) Is the protein sufficient to fuse cells that normally do
not fuse in situ?
(iv) Can the protein fuse heterologous cells in tissue
culture?
(v) Can the protein fuse liposomes in vitro?
Because developmental cell–cell fusogens remain, with
the exception of Syncytins and FFs, hypothetical and
undiscovered, it is conceivable that the molecular
mediators of cell membrane fusion are not only hidden,
but might also be completely different from the known viral
and intracellular membrane-fusion proteins. Modifiers of
membrane lipid composition and proteolipid channels have
been proposed as alternative mediators of membrane
fusion [85,86]. However, we believe that the evidence for
viral, intracellular and developmental fusogens support
the hypothesis of a unified concept in which viral-like or
SNARE-like proteins fuse biological membranes through
hairpins, zippers, hairpins–zippers or novel fusogenic
mechanisms.
544 Review TRENDS in Cell Biology Vol.17 No.11
www.sciencedirect.com

Acknowledgements
The authors apologize to those colleagues whose work could not be cited
owing to space limitations. We thank D. Lindell, A. Sapir, G. Shemer, O.
Avinoam and four anonymous reviewers for critical reading of the
manuscript. This work was supported by grants from the Israel Science
Foundation.
References
1 Chen, E.H. et al. (2007) Cell-cell fusion. FEBS Lett. 581, 2181–2193
2 Horsley, V. and Pavlath, G.K. (2004) Forming a multinucleated
cell: molecules that regulate myoblast fusion. Cells Tissues Organs
176, 67–78
3 Potgens, A.J. et al. (2002) Mechanisms of syncytial fusion: a review.
Placenta 2, S107–S113
4 Podbilewicz, B. and White, J.G. (1994) Cell fusions in the developing
epithelia of C. elegans.Dev. Biol. 161, 408–424
5 Forest, C.L. (1987) Genetic control of plasma membrane adhesion and
fusion in Chlamydomonas gametes.J. Cell Sci. 88, 613–621
6 Isaksen, D.E. et al. (1999) Inductive regulation of cell fusion in leech.
Development 126, 3381–3390
7 Misamore, M.J. et al. (2003) The Chlamydomonas Fus1 protein is
present on the mating type plus fusion organelle and required for a
critical membrane adhesion event during fusion with minus gametes.
Mol. Biol. Cell 14, 2530–2542
8 Glass, N.L. et al. (2004) Hyphal homing, fusion and mycelial
interconnectedness. Trends Microbiol. 12, 135–141
9 Sollner, T.H. (2004) Intracellular and viral membrane fusion: a uniting
mechanism. Curr. Opin. Cell Biol. 16, 429–435
10 He, Z.Y. et al. (2003) None of the integrins known to be present on the
mouse egg or to be ADAM receptors are essential for sperm-egg binding
and fusion. Dev. Biol. 254, 226–237
11 White, J.M. and Rose, M.D. (2001) Yeast mating: getting close to
membrane merger. Curr. Biol. 11, R16–R20
12 Heiman, M.G. and Walter, P. (2000) Prm1p, a pheromone-regulated
multispanning membrane protein, facilitates plasma membrane fusion
during yeast mating. J. Cell Biol. 151, 719–730
13 Heiman, M.G. et al. (2007) The Golgi-resident protease Kex2 acts in
conjunction with Prm1 to facilitate cell fusion during yeast mating. J.
Cell Biol. 176, 209–222
14 Shi, C. et al. (2007) A role for a complex between activated G protein-
coupled receptors in yeast cellular mating. Proc. Natl. Acad. Sci. U. S.
A. 104, 5395–5400
15 Mohler, W.A. et al. (2002) The type I membrane protein EFF-1
is essential for developmental cell fusion in C. elegans.Dev. Cell 2,
355–362
16 Sapir, A. et al. (2007) AFF-1, a FOS-1-regulated fusogen, mediates
fusion of the anchor cell in C. elegans.Dev. Cell 12, 683–698
17 L’Hernault, S. (1997) Spermatogenesis. In C. elegans II (Riddle, D.
et al., eds), pp. 271–294, Cold Spring Harbor Laboratory Press
18 Singson, A. et al. (1998) The C. elegans spe-9 gene encodes a sperm
transmembrane protein that contains egf-like repeats and is required
for fertilization. Cell 93, 71–79
19 Zannoni, S. et al. (2003) Dynamic localization of SPE-9 in sperm: a
protein required for sperm-oocyte interactions in Caenorhabditis
elegans.BMC Dev. Biol. 3, 10
20 Singson, A. (2001) Every sperm is sacred: fertilization in
Caenorhabditis elegans.Dev. Biol. 230, 101–109
21 Chatterjee, I. et al. (2005) The Caenorhabditis elegans spe-38 gene
encodes a novel four-pass integral membrane protein required for
sperm function at fertilization. Development 132, 2795–2808
22 Kroft, T.L. et al. (2005) The spe-42 gene is required for sperm-egg
interactions during C. elegans fertilization and encodes a sperm-
specific transmembrane protein. Dev. Biol. 286, 169–181
23 Yagi, M. et al. (2005) DC-STAMP is essential for cell-cell fusion in
osteoclasts and foreign body giant cells. J. Exp. Med. 202, 345–
351
24 Kadandale, P. et al. (2005) The egg surface LDL receptor repeat-
containing proteins EGG-1 and EGG-2 are required for fertilization
in Caenorhabditis elegans.Curr. Biol. 15, 2222–2229
25 Perotti, M-E. (1975) Ultrastructural aspects of fertilization in
Drosophila. In The Functional Anatomy of the Spermatozoan,
Proceedings of the Second International Symposium (Afzelins, B.A.,
ed.), pp. 57–68, Pergamon Press
26 Wilson, K.L. et al. (2006) Sperm plasma membrane breakdown during
Drosophila fertilization requires sneaky, an acrosomal membrane
protein. Development 133, 4871–4879
27 Ohsako, T. et al. (2003) The Drosophila misfire gene has an essential
role in sperm activation during fertilization. Genes Genet. Syst. 78,
253–266
28 Doberstein, S.K. et al. (1997) Genetic analysis of myoblast fusion:
blown fuse is required for progression beyond the prefusion complex.
J. Cell Biol. 136, 1249–1261
29 Shemer, G. and Podbilewicz, B. (2003) The story of cell fusion: big
lessons from little worms. Bioessays 25, 672–682
30 Achanzar, W.E. and Ward, S. (1997) A nematode gene required for
sperm vesicle fusion. J. Cell Sci. 110, 1073–1081
31 Primakoff, P. and Myles, D.G. (2007) Cell-cell membrane fusion during
mammalian fertilization. FEBS Lett. 581, 2174–2180
32 Stein, K.K. et al. (2005) Defects in secretory pathway trafficking during
sperm development in Adam2 knockout mice. Biol. Reprod. 73, 1032–
1038
33 Cho, C. et al. (1998) Fertilization defects in sperm from mice lacking
fertilin beta. Science 281, 1857–1859
34 Okabe, M. et al. (1987) Capacitation-related changes in antigen
distribution on mouse sperm heads and its relation to fertilization
rate in vitro.J. Reprod. Immunol. 11, 91–100
35 Inoue, N. et al. (2005) The immunoglobulin superfamily protein
Izumo is required for sperm to fuse with eggs. Nature 434, 234–
238
36 Le Naour, F. et al. (2000) Severely reduced female fertility in CD9-
deficient mice. Science 287, 319–321
37 Miyado, K. et al. (2000) Requirement of CD9 on the egg plasma
membrane for fertilization. Science 287, 321–324
38 Kaji, K. et al. (2000) The gamete fusion process is defective in eggs of
Cd9-deficient mice. Nat. Genet. 24, 279–282
39 Runge, K.E. et al. (2007) Oocyte CD9 is enriched on the microvillar
membrane and required for normal microvillar shape and distribution.
Dev. Biol. 304, 317–325
40 Barraud-Lange, V. et al. (2007) Transfer of oocyte membrane fragments
to fertilizing spermatozoa. FASEB J. http://www.fasebj.org/10.1096/
fj.06-8035hyp
41 Gordon-Alonso, M. et al. (2006) Tetraspanins CD9 and CD81 modulate
HIV-1-induced membrane fusion. J. Immunol. 177, 5129–5137
42 Takeda, Y. et al. (2003) Tetraspanins CD9 and CD81 function to
prevent the fusion of mononuclear phagocytes. J. Cell Biol. 161,
945–956
43 Blond, J.L. et al. (1999) Molecular characterization and placental
expression of HERV-W, a new human endogenous retrovirus family.
J. Virol. 73, 1175–1185
44 Blond, J.L. et al. (2000) An envelope glycoprotein of the human
endogenous retrovirus HERV-W is expressed in the human placenta
and fuses cells expressing the type D mammalian retrovirus receptor.
J. Virol. 74, 3321–3329
45 Mi, S. et al. (2000) Syncytin is a captive retroviral envelope
protein involved in human placental morphogenesis. Nature 403,
785–789
46 Huppertz, B. et al. (2006) Trophoblast fusion: fusogenic proteins,
syncytins and ADAMs, and other prerequisites for syncytial fusion.
Micron 37, 509–517
47 Lavillette, D. et al. (1998) A proline-rich motif downstream of the
receptor binding domain modulates conformation and fusogenicity of
murine retroviral envelopes. J. Virol. 72, 9955–9965
48 Marin, M. et al. (2000) Sodium-dependent neutral amino acid
transporter type 1 is an auxiliary receptor for baboon endogenous
retrovirus. J. Virol. 74, 8085–8093
49 Marin, M. et al. (2003) N-linked glycosylation and sequence changes in
a critical negative control region of the ASCT1 and ASCT2 neutral
amino acid transporters determine their retroviral receptor functions.
J. Virol. 77, 2936–2945
50 Frendo, J.L. et al. (2003) Direct involvement of HERV-W Env
glycoprotein in human trophoblast cell fusion and differentiation.
Mol. Cell. Biol. 23, 3566–3574
51 Blaise, S. et al. (2003) Genomewide screening for fusogenic human
endogenous retrovirus envelopes identifies Syncytin 2, a gene
conserved on primate evolution. Proc. Natl. Acad. Sci. U. S. A. 100,
13013–13018
Review TRENDS in Cell Biology Vol.17 No.11 545
www.sciencedirect.com

52 Malassine, A. et al. (2007) Expression of the fusogenic HERV-FRD Env
glycoprotein (syncytin 2) in human placenta is restricted to villous
cytotrophoblastic cells. Placenta 28, 185–191
53 Dupressoir, A. et al. (2005) Syncytin-A and syncytin-B, two fusogenic
placenta-specific murine envelope genes of retroviral origin conserved
in Muridae. Proc. Natl. Acad. Sci. U. S. A. 102, 725–730
54 Gong, R. et al. (2007) Syncytin-A mediates the formation of
syncytiotrophoblast involved in mouse placental development. Cell.
Physiol. Biochem. 20, 517–526
55 Peng, X. et al. (2007) Functional characterization of syncytin- A, a newly
murine endogenous virus envelope protein. Implication for its fusion
mechanism. J. Biol. Chem. 282, 381–389
56 Shemer, G. et al. (2004) EFF-1 is sufficient to initiate and execute
tissue-specific cell fusion in C. elegans.Curr. Biol. 14, 1587–1591
57 del Campo, J.J. et al. (2005) Fusogenic activity of EFF-1 is regulated via
dynamic localization in fusing somatic cells of C. elegans.Curr. Biol. 15,
413–423
58 Podbilewicz, B. et al. (2006) The C. elegans developmental fusogen
EFF-1 mediates homotypic fusion in heterologous cells and in vivo.
Dev. Cell 11, 471–481
59 Newman, A.P. et al. (1996) Morphogenesis of the C. elegans
hermaphrodite uterus. Development 122, 3617–3626
60 Sharma-Kishore, R. et al. (1999) Formation of the vulva in C. elegans:a
paradigm for organogenesis. Development 126, 691–699
61 Nguyen, C.Q. et al. (1999) Morphogenesis of the Caenorhabditis elegans
male tail tip. Dev. Biol. 207, 86–106
62 Sherwood, D.R. (2006) Cell invasion through basement membranes: an
anchor of understanding. Trends Cell Biol. 16, 250–256
63 Bate, M. (1990) The embryonic development of larval muscles in
Drosophila.Development 110, 781–804
64 Kim, S. et al. (2007) A critical function for the actin cytoskeleton in
targeted exocytosis of prefusion vesicles during myoblast fusion. Dev.
Cell 12, 571–586
65 Estrada, B. et al. (2007) The MARVEL domain protein, Singles Bar, is
required for progression past the pre-fusion complex stage of myoblast
fusion. Dev. Biol. 307, 328–339
66 Ruiz-Gomez, M. et al. (2000) Drosophila dumbfounded: a myoblast
attractant essential for fusion. Cell 102, 189–198
67 Strunkelnberg, M. et al. (2001) rst and its paralogue kirre act
redundantly during embryonic muscle development in Drosophila.
Development 128, 4429–4439
68 Bour, B.A. et al. (2000) Drosophila SNS, a member of the
immunoglobulin superfamily that is essential for myoblast fusion.
Genes Dev. 14, 1498–1511
69 Galletta, B.J. et al. (2004) SNS: adhesive properties, localization
requirements and ectodomain dependence in S2 cells and embryonic
myoblasts. Mech. Dev. 121, 1455–1468
70 Massarwa, R. et al. (2007) WIP/WASp-based actin-polymerization
machinery is essential for myoblast fusion in Drosophila.Dev. Cell
12, 557–569
71 Yang, J.T. et al. (1996) Genetic analysis of alpha 4 integrin functions in
the development of mouse skeletal muscle. J. Cell Biol. 135, 829–835
72 Kurisaki, T. et al. (2003) Phenotypic analysis of Meltrin
alpha (ADAM12)-deficient mice: involvement of Meltrin alpha in
adipogenesis and myogenesis. Mol. Cell. Biol. 23, 55–61
73 Kurohara, K. et al. (2004) Essential roles of Meltrin beta (ADAM19) in
heart development. Dev. Biol. 267, 14–28
74 Charlton, C.A. et al. (2000) Neural cell adhesion molecule (NCAM) and
myoblast fusion. Dev. Biol. 221, 112–119
75 Hollnagel, A. et al. (2002) The cell adhesion molecule M-cadherin is not
essential for muscle development and regeneration. Mol. Cell. Biol. 22,
4760–4770
76 Srinivas, B.P. et al. (2007) A conserved molecular pathway
mediates myoblast fusion in insects and vertebrates. Nat. Genet. 39,
781–786
77 Saginario, C. et al. (1998) MFR, a putative receptor mediating the
fusion of macrophages. Mol. Cell. Biol. 18, 6213–6223
78 Han, X. et al. (2000) CD47, a ligand for the macrophage fusion receptor,
participates in macrophage multinucleation. J. Biol. Chem. 275,
37984–37992
79 Kania, J.R. et al. (1997) CD44 antibodies inhibit osteoclast formation.
J. Bone Miner. Res. 12, 1155–1164
80 de Vries, T.J. et al. (2005) Effect of CD44 deficiency on in vitro and in
vivo osteoclast formation. J. Cell. Biochem. 94, 954–966
81 Duelli, D. and Lazebnik, Y. (2003) Cell fusion: a hidden enemy? Cancer
Cell 3, 445–448
82 Duelli, D.M. et al. (2007) A virus causes cancer by inducing
massive chromosomal instability through cell fusion. Curr. Biol. 17,
431–437
83 Kielian, M. and Rey, F.A. (2006) Virus membrane-fusion proteins: more
than one way to make a hairpin. Nat. Rev. Microbiol. 4, 67–76
84 Jahn, R. and Scheller, R.H. (2006) SNAREs–engines for membrane
fusion. Nat. Rev. Mol. Cell Biol. 7, 631–643
85 Peters, C. et al. (2001) Trans-complex formation by proteolipid
channels in the terminal phase of membrane fusion. Nature 409,
581–588
86 Zimmerberg, J. and Chernomordik, L.V. (2005) Ne uroscience. Synaptic
membranes bend to the will of a neurotoxin. Science 310, 1626–1627
87 Podbilewicz, B. (2006) Cell fusion. In WormBook (The C. elegans
Research Community, ed.), WormBook, DOI: 10.1895/wormbook.
1.52.1 (http://www.wormbook.org)
88 Korazim, O. et al. (2006) Functional and structural characterization of
HIV-1 gp41 ectodomain regions in phospholipid membranes suggests
that the fusion-active conformation is extended. J. Mol. Biol. 364,
1103–1117
89 Yang, X. et al. (2005) Stoichiometry of envelope glycoprotein trimers in
the entry of human immunodeficiency virus type 1. J. Virol. 79, 12132–
12147
90 Weissenhorn, W. et al. (2007) Virus membrane fusion. FEBS Lett. 581,
2150–2155
91 Gallo, S.A. et al. (2003) The HIV Env-mediated fusion reaction.
Biochim. Biophys. Acta 1614, 36–50
92 Menon, S.D. et al. (2005) A positive feedback loop between
Dumbfounded and Rolling pebbles leads to myotube enlargement in
Drosophila.J. Cell Biol. 169, 909–920
93 Kukita, T. et al. (2004) RANKL-induced DC-STAMP is essential for
osteoclastogenesis. J. Exp. Med. 200, 941–946
94 Putiri, E. et al. (2004) Functional domains and temperature-sensitive
mutations in SPE-9, an EGF repeat-containing protein required for
fertility in Caenorhabditis elegans.Dev. Biol. 272, 448–459
Free journals for developing countries
In 2002, the WHO and six medical journal publishers launched the Health InterNetwork Access to
Research Initiative, which enabled nearly 70 of the world’s poorest countries to gain free or
reduced-cost access to biomedical literature through the internet. Currently more than 70
publishers are participating in the program, providing access to over 2000 journals.
Gro Harlem Brundtland, former director-general for the WHO, said that this initiative was ‘‘perhaps
the biggest step ever taken towards reducing the health information gap between rich and poor
countries’’.
For more information, visit www.who.int/hinari
546 Review TRENDS in Cell Biology Vol.17 No.11
www.sciencedirect.com