Human Reproduction Update Advance Access originally published online on August 27, 2004
Human Reproduction Update 2004 10(6):487-496; doi:10.1093/humupd/dmh039
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Syncytin: the major regulator of trophoblast fusion? Recent developments and hypotheses on its action
Department of Anatomy II, University Hospital Aachen, Wendlingweg 2, D-52057 Aachen, Germany
1 To whom correspondence should be addressed. Email: apotgens{at}ukaachen.de
 |
Abstract |
|---|
TOP
Abstract
IntroductionSyncytin is a membrane protein derived from the envelope gene of an endogenous retrovirus of the HERV-W family. The gene appears to be almost exclusively expressed in placenta; the protein was found in particular in syncytiotrophoblast. After transfection into various cell types it has proven to be a very fusogenic protein, inducing the formation of syncytia. Therefore, the question rises as to whether syncytin is responsible for the fusion process of villous cytotrophoblast into syncytiotrophoblast in vivo. If so, how is this fusion process regulated if syncytin is found all over the syncytiotrophoblast? Can this process be regulated through local or temporal changes in syncytin expression, or is syncytin merely one factor in a cascade of events leading to fusion limited at some other level? This review will try to summarize the published data on the regulation of fusion in trophoblast models as well as on the localization and regulation of syncytin expression and of its presumed receptors. Assuming that syncytin is the key factor inducing trophoblast fusion, a number of models will be presented by which syncytin and/or its receptors might regulate this process. In some of the hypotheses proposed, local coexpression of syncytin and receptor, leading to blocking of one factor by the other, is of functional relevance.
Key words: cell–cell fusion / HERV-W / retrovirus receptor type D / syncytin / trophoblast
|
Introduction |
|---|
Syncytial fusion of cytotrophoblast with syncytiotrophoblast is an important process in implantation and placentation. Syncytiotrophoblast is a multinuclear tissue forming the outer surface of the fetal part of the placenta and the barrier with maternal blood. This syncytium is responsible for many of the functions performed by the placenta, such as transport of oxygen, nutrients and waste products, hormone production and immune tolerance (for review see Benirschke and Kaufmann, 2000
). Nuclei in the syncytiotrophoblast were shown to be non-proliferative (Richart, 1961; Moe, 1971). By contrast, villous cytotrophoblast cells comprise a proliferative population of stem cells necessary for trophoblast growth and regeneration (Kosanke et al., 1998). The trophoblast fusion rate was estimated to be severalfold higher than that needed only for growth (Huppertz et al., 1998). The syncytiotrophoblast itself has a high turnover rate, in that syncytial nuclei become apoptotic and are shed in so-called syncytial sprouts, corresponding to giant apoptotic bodies, into maternal blood (Ikle, 1961; Yasuda et al., 1995; Huppertz et al., 1998). Trophoblast fusion must also compensate for this loss of material. In addition, various studies have indicated that in the syncytiotrophoblast very little uridine is incorporated as compared to cytotrophoblast (Kaufmann et al., 1983; Huppertz et al., 1999, 2003), suggesting that transcription is low in syncytiotrophoblast, and that much of the RNA necessary for its massive protein production is likely to be transferred from cytotrophoblast cells into the syncytium through fusion. This does not imply, however, that any mRNA species present in the syncytiotrophoblast is necessarily transcribed in the villous cytotrophoblast.
It is not difficult to imagine that trophoblast fusion must be tightly controlled. If fusion would occur too frequently, the regenerative pool of villous cytotrophoblast cells would soon be exhausted, whereas too little fusion would lead to rarefaction of the syncytiotrophoblast layer and to functional deficits. However, it is difficult to say at what level and by which mechanism this fusion process is limited, and whether fusion is initiated by cytotrophoblast cells or by syncytiotrophoblast. Differentiation of cytotrophoblast cells as a prerequisite for fusion would guarantee that proliferative cytotrophoblast cells are locked out from the process so that the regenerative pool is not exhausted. Studies performed in recent years suggest that fusion is preceded by differentiation of cytotrophoblast cells, accompanied by early stages of apoptosis (Huppertz et al., 1999; Black et al., 2004). In this view, fusion would be driven by cytotrophoblast cells reaching a certain stage of differentiation.
On the other hand, ultrastructural and enzyme histochemical studies have demonstrated distinct regions within the syncytiotrophoblast showing signs of ageing or degeneration such as loss of polyribosomes, degranulation of endoplasmic reticulum, condensation of chromatin and loss of enzymes of energy metabolism (Kaufmann, 1972; Kaufmann and Stark, 1972; Martin and Spicer, 1973). The data in these old descriptive papers suggested that such ageing areas of the syncytiotrophoblast were preferably involved in fusion. In this view, fusion would be driven by aged areas of syncytiotrophoblast that need fusion with an underlying cytotrophoblast cell for replenishment.
Syncytin appears to play a key role in mediating syncytial fusion of villous trophoblast (Frendo et al., 2003b). Moreover, an amino acid transporter (ATB0, also known as ASCT2) was identified as a receptor for syncytin (Blond et al., 2000). However, at present it is an open question whether or how syncytin and its receptor can regulate (drive and limit at the same time) the fusion process on their own. In this review we will summarize what has been published on the function and expression of syncytin so far, as well as on other molecules involved in trophoblast fusion. Learning more about the spatial and temporal expression of syncytin, its receptors, and other fusion-relevant molecules might help in understanding the regulation of syncytial fusion. In the final part of this paper we will speculate about theoretical models on how the interplay between syncytin and its receptor(s) might determine the fusion of cytotrophoblast with syncytiotrophoblast.
|
The discovery of syncytin and its receptors |
|---|
The name syncytin was introduced by (Mi et al., 2000
) after finding that the envelope gene of an endogenous retrovirus of the HERV-W family (Blond et al., 1999) encoded a functional protein expressed in syncytiotrophoblast and that, upon transfection into COS cells, it induced the formation of multinuclear syncytia. It was proposed that syncytin is a key factor in regulating trophoblast fusion. In the same year Blond et al. (2000) published a paper supporting this message, but which in addition identified the type D mammalian retrovirus receptor (RDR, also known as a neutral amino acid transporter system: ATB0, ASCT2, or SLC1A5) (Kekuda et al., 1996; Rasko et al., 1999; Tailor et al., 1999) as a receptor for the HERV-W envelope. A functional receptor was present on pig, monkey and human cells, but lacking on cells of quail, rat, mouse, hamster or cat origin, such that fusion of syncytin-expressing cells with the latter cells did not occur. In contrast, Mi et al. (2000) described fusion of syncytin-expressing COS cells with hamster (CHO) cells, and between insect (Sf9) cells transfected with syncytin. The latter experiments were not described in detail, making it difficult to judge whether methodological differences can explain the observed discrepancy. Recently, a potential alternative receptor for syncytin has been described, which is an amino acid transporter system homologous to RDR/ATB0/ASCT2/SLC1A5, called ASCT1 (or SLC1A4) (Lavillette et al., 1998; Marin et al., 2000, 2003). To prevent confusion with these alternative names, we will in the following text refer to the two amino acid transporters and presumed syncytin receptors as ASCT2 and ASCT1.
A second fusogenic envelope protein derived from an endogenous retrovirus has recently been described (Blaise et al., 2003). The envelope protein of a member of the HERV-FRD family was identified after screening many retroviral sequences present in the human genome for having an intact open reading frame encoding the envelope protein. Out of 16 envelope-encoding sequences, only two were found to cause cell–cell fusion: the envelope of HERV-W (syncytin) and the envelope of HERV-FRD. The latter protein was then named syncytin 2. Syncytin 2 appeared to utilize a different receptor than syncytin. The mRNA encoding syncytin 2 was specifically found in placenta tissue.
Syncytin is able to function as an actual retroviral envelope protein: although it could not be packaged into retroviral particles derived from Moloney murine leukaemia virus (MLV; Blond et al., 2000), it could be packaged into recombinant retroviral particles derived from HIV-1 (An et al., 2001). These viral pseudotypes were able to infect and transfer genetic material to human cells not expressing the HIV receptor CD4.
Frendo et al. (2003b) showed that syncytin expression was up-regulated after stimulating the fusion of primary cytotrophoblast isolates into syncytia by a cAMP analogue. Furthermore, they showed that antisense inhibition of syncytin in these primary cell cultures actually inhibited fusion, demonstrating an active role of syncytin in driving the fusion of trophoblast. By doing so, syncytin appears to have an important role in the morphogenesis of the human placenta. In an evolutionary sense, the envelope gene of this HERV-W virus seems to have adopted an important function in human reproduction. It has thus prevented its own functional extinction, which was the fate of most other human endogenous retroviral genes that no longer encode proteins (Kjellman et al., 1995; Blond et al., 1999; Hughes and Coffin, 2001; Ling et al., 2002).
|
Alternative functions of syncytin |
|---|
In addition to its fusogenic nature, two other possible functions of the syncytin protein have been postulated. First, a potential immunosuppressive region has been identified in the amino acid sequence of syncytin (Blond et al., 1999
). This may play a role in suppressing the maternal immune response against the fetus. Second, expression of syncytin has been shown to cause cellular resistance against infection by spleen necrosis virus (Ponferrada et al., 2003) and is likely to prevent infection by other retroviruses of the same family. It is thus feasible that expression of syncytin by the syncytiotrophoblast prevents vertical transmission of a number of retroviruses through the placenta. |
Distribution of syncytin expression in human placenta |
|---|
Syncytin mRNA was detected by Northern blot in placenta, and to a much lesser extent also in testis (Mi et al., 2000
). By in situ hybridization on term and preterm placental sections, the mRNA encoding syncytin was found only in syncytiotrophoblast (Mi et al., 2000). In sections from pregnancies complicated by pre-eclampsia, the mRNA signal was much weaker (Lee et al., 2001). No explicit comments were made in these papers as to whether at least in some cytotrophoblast cells syncytin mRNA was detected. For localizing the syncytin protein, various antibodies have been used. A rabbit polyclonal antibody raised against peptides from various parts of the syncytin protein (Mi et al., 2000) was used to localize syncytin by immunohistochemistry. In normal placentae of gestational age 21–40 weeks, the authors found staining at the basal syncytiotrophoblast membrane, but in some cases also in the apical syncytiotrophoblast membrane. In placentae from pregnancies complicated by pre-eclempsia, syncytin was usually detected in the apical syncytiotrophoblast membrane (Lee et al., 2001). By using the same antibody on first trimester placenta specimens, syncytin appeared to be expressed by extravillous trophoblast cells as well as by villous trophoblast (Muir et al., 2003).
A monoclonal antibody (6A2B2) described by (Blond et al., 2000) was raised against a recombinant protein comprising a large part of the syncytin protein (residues 68–446). On sections of a placenta of 13 weeks gestation, the authors showed most staining in the syncytiotrophoblast, but there were also areas of positive staining in villous cytotrophoblast. Another study using the same monoclonal demonstrated anti-syncytin reactivity at the apical syncytiotrophoblast membrane in both first trimester and term placental sections (Frendo et al., 2003b). By using this monoclonal in western blot and immunohistochemistry, Smallwood et al. (2003) demonstrated that syncytin expression was highest in first trimester placenta compared to term placenta. These authors furthermore showed that syncytin staining was strong in syncytiotrophoblast and in extravillous trophoblast cells, whereas staining in villous cytotrophoblast was faint.
Using a polyclonal guinea-pig antibody raised against a recombinant protein comprising syncytin residues 21–215 (Yu et al., 2002) we found uniform staining in the syncytiotrophoblast of first trimester placental paraffin sections. In the villous cytotrophoblast layer, few cells were stained (Figure 1).
|
Varying specificities of the existing antibodies, and/or varying staining methods or tissue sources used may have caused these discrepancies. Knowing the exact distribution of syncytin within the trophoblast is, however, of crucial importance for understanding how syncytin might regulate trophoblast fusion (see below). To date, however, there is little consensus in the literature on this issue. However, the major site of syncytin expression appears to be the syncytiotrophoblast, while it was also detected in cytotrophoblast. This would imply that syncytin expression occurs mainly after fusion. This may seem to be a paradoxical finding if one presumes that syncytin initiates the fusion process although it is synthesized mainly after fusion.
|
Regulation of syncytin expression and of cell–cell fusion in trophoblast model systems |
|---|
In theory, expression of syncytin may be regulated at various levels: transcription, alternative splicing, stability of the messenger, efficiency of translation, post-translational modifications, intracellular targeting, and stability of the protein. The transcription rate of the syncytin gene, being an endogenous retrovirus, is regulated by the promoter region in and close to the 5' long terminal repeat region (5'LTR) of the provirus. Due to alternative splicing, several mRNA can be produced, and the ratio of different splice forms found in placenta varied throughout pregnancy (Blond et al., 1999
; Smallwood et al., 2003). Most likely only a 3.1 kb transcript is translated into syncytin protein. In testis, only an 8.0 kb splice variant was expressed (Mi et al., 2000). To manifest its fusogenic activity, the syncytin preprotein must be cleaved by furin to yield a transmembrane (TM) unit and an extracellular surface (SU) subunit (Blond et al., 1999).
Yu et al. (2002) demonstrated that the placenta-specific transcription factor GCMa interacts with two binding sites upstream of the HERV-W 5'LTR, thereby stimulating transcription of the syncytin mRNA. Overexpression of GCMa in trophoblast-derived cell lines BeWo and JEG3 stimulated syncytin mRNA and protein expression as well as cell–cell fusion. Overexpression of GCMa in HeLa cells did not lead to expression of a syncytin transcript, suggesting that the effect of GCMa on syncytin expression is lineage-dependent. Lee et al. (2003) demonstrated that the 3' terminus of the U3 region in the HERV-W 5'LTR, containing a TATA box, is associated with cell-specific expression of syncytin.
Various external factors have been shown to mediate syncytin expression and cell–cell fusion in trophoblast model systems. The most usual model to study trophoblast fusion is a panel of choriocarcinoma cell lines: BeWo, JAR and JEG3. BeWo and JAR cells were shown to fuse upon stimulation with forskolin (Lyden et al., 1993; Adler et al., 1995), an agent that increases the levels of intracellular cAMP. Forskolin was later shown to increase syncytin mRNA levels in both JEG3, JAR and BeWo cells (Mi et al., 2000; Kudo and Boyd, 2002; Borges et al., 2003). However, levels of syncytin mRNA in JEG3 cells (whether stimulated with forskolin or not) were low compared to those in JAR and BeWo cells (Mi et al., 2000; Borges et al., 2003). Expression of syncytin, therefore, appears to be regulated through a cAMP-dependent mechanism. Using cell line BeWo, Kudo et al. (2003a) and Knerr et al. (2003) have demonstrated that under hypoxia syncytin mRNA levels decreased, and were less inducible by forskolin, compared to ‘normoxia’ (20% oxygen). Hypoxia also decreased fusion of BeWo cells (Kudo et al., 2003a). Down-regulation of syncytin expression by hypoxia may explain the findings of Lee et al. (2001) that syncytin expression was very low in placentae of pre-eclamptic pregnancies. The cases investigated in this study were, however, pathohistologically unclassified so that there is no evidence whether these were combined with intraplacental hypoxia (Todros et al., 1999) or with intraplacental hyperoxia (Kingdom and Kaufmann, 1997).
Another cellular model to study trophoblast fusion is culture of primary isolates of trophoblast cells from first trimester or term placenta. Fusion is generally monitored by measuring disappearance of desmoplakin immunostaining and/or by measuring the release of a product of syncytiotrophoblast: hCG. It has long been known that addition of cAMP analogues stimulates the fusion of primary trophoblast cell isolates (Keryer et al., 1998). Frendo et al. (2003b) showed that syncytin mRNA and protein were present in isolated cytotrophoblast cells, and that stimulation of cell–cell fusion in such a system by cAMP was associated with an increase of their levels. Furthermore, inhibition of syncytin up-regulation by an antisense strategy inhibited fusion, showing that syncytin has a direct influence on syncytial fusion. Also the oxidative state influenced syncytin expression in isolated trophoblast cells: by overexpressing the antioxidant enzyme copper/zinc superoxide dismutase (SOD) in cytotrophoblast cells, syncytin mRNA was down-regulated and fusion was inhibited (Frendo et al., 2001). A preliminary report by (Cariño et al., 2003) indicated that estradiol may play a role in regulating syncytin expression: addition of estradiol in physiological concentrations increased the levels of syncytin mRNA in primary trophoblast cell cultures.
|
Local expression of syncytin receptors and regulation of receptor gene expression |
|---|
Little is known about the expression of ASCT2 (also known as RDR or ATB0) or the alternative syncytin receptor ASCT1 within the trophoblast compartment. To the best of our knowledge, no antibodies exist against ASCT1, and no one has performed in situ hybridization on placental sections to find the site of production of the ASCT1 and ASCT2 mRNA. An antibody against ASCT2 has only recently been described (Green et al., 2004
), but studies on its reactivity in placenta have not been published yet. The only knowledge we have on ASCT2 or ASCT1 expression in placenta comes from functional studies on amino transporters in isolated basal or apical (microvillous) syncytiotrophoblast membranes. System ASC activity (a sodium-dependent transport system of neutral amino acids) is performed by transporters ASCT2 and ASCT1, both of which are considered receptors for syncytin (Kekuda et al., 1996; Lavillette et al., 1998; Rasko et al., 1999; Tailor et al., 1999; Marin et al., 2000, 2003). System ASC activity appears to be present in basal but not in microvillous (apical) syncytiotrophoblast membranes. The messenger encoding ASCT2 is readily expressed in placenta, whereas the ASCT1 mRNA was found to be low in placenta (reviewed in Jansson, 2001 and in Cariappa et al., 2003). To our knowledge, nothing is known about ASCT2 and ASCT1 expression and activities in cytotrophoblast cells.
In a renal cell line NBL-1, ASCT2 activity was shown to be up-regulated upon amino acid starvation (Plakidou-Dymock et al., 1994). ASCT2 mRNA was demonstrated in choriocarcinoma cell lines BeWo, JAR and JEG3. Forskolin, inducing the expression of syncytin in choriocarcinoma cells, had no, or a slightly suppressing, effect on the levels of ASCT2 mRNA in BeWo, JAR and JEG3 cells and on ASCT2 activity in BeWo cells (Kudo and Boyd, 2002; Borges et al., 2003). Hypoxia, which suppressed syncytin expression, also had no or only a slightly suppressing effect on ASCT2 expression and activity in BeWo cells (Knerr et al., 2003a; Kudo et al., 2003a). In JAR cells, epidermal growth factor (EGF) up-regulated ASCT2 mRNA levels and activity (Torres-Zamorano et al., 1997). Interestingly, EGF also increased cell–cell fusion in primary cytotrophoblast isolates (Morrish et al., 1997; Crocker et al., 2001). One can speculate whether the increasing fusion is caused by the up-regulation of syncytin receptor.
|
Additional factors needed for fusion |
|---|
In a recent paper, our group showed that JAR cells expressed the mRNA for ASCT2 and syncytin at equal or higher levels than BeWo cells, but unlike BeWo cells, did not fuse (Borges et al., 2003
). It is possible that either ASCT2 or syncytin mRNA in these cells are not translated into functional protein. Another explanation may be that apart from syncytin and receptor other factors are necessary for fusion to occur, and that JAR cells lack one of these factors. Indeed, there are lines of evidence that a number of other proteins, but also phospholipids, play a role in trophoblast fusion (for review see Pötgens et al., 2002; Huppertz et al., 2002).
The placental expression of the mRNA encoding another fusogenic envelope protein syncytin 2 (HERV-FRD; Blaise et al., 2003) indicates that there may be alternative mechanisms for regulating trophoblast fusion. Nothing is known yet on its local distribution within the placenta. It is also not clear which receptor is used by syncytin 2, and whether this receptor is present in trophoblast.
Some members of a family of proteins called ADAM (a disintegrin and a metalloproteinase domain) contain putative hydrophobic fusion peptides just like retroviral envelope proteins. ADAM 1 and 2 (fertilin 
and ß) and ADAM 12 (meltrin ) have actually been shown to be involved in cell–cell fusion. Fertilins and ß were shown to be crucial for sperm–oocyte fusion in the mouse (Cho et al., 1997, 1998; Evans et al., 1998). The fertilin gene in the human is, however, dysfunctional and therefore fertilin cannot be relevant for any type of cell–cell fusion in the human (Jury et al., 1997, 1998). Meltrin a triggers fusion of myoblasts into skeletal muscle fibres (Yagami-Hiromasa et al., 1995; Gilpin et al., 1998; Galliano et al., 2000) and also the formation of osteoclasts (Abe et al., 1999). Meltrin mRNA was also detected in the human placenta (Gilpin et al., 1998), but as yet it is not clear whether trophoblast or another cell type (e.g. blood mononuclear cells) was responsible for this expression.
CD98 is a subunit of a family of amino acid transporter molecules. Apart from its transporter function it has been proposed to play a role in cellular differentiation, activation, adhesion, as well as fusion. It has been shown to play a modulating role in cell fusion of monocytes/macrophages into osteoclasts, as well as in virus-mediated cell fusion (reviewed in Devés and Boyd, 2000). Kudo et al. (2003b) have recently shown that reduction of CD98 expression through an antisense oligonucleotide inhibited fusion of BeWo cells. The mechanism by which CD98 influences cell–cell fusion is not yet clear (Devés and Boyd, 2000).
Gap junctional communication was shown to be essential to trophoblast differentiation and fusion (Cronier et al., 2003). Connexin 43 (Cx43) is a protein involved in trophoblastic gap junctional communication. By antisense strategies in primary trophoblast cultures it was shown that Cx43 is directly involved in trophoblast fusion. While fusion was inhibited in antisense-treated cells (Frendo et al., 2003a), the mRNA encoding syncytin was also decreased (Malassiné et al., 2003). Connexin 43 therefore appears to have a direct or indirect effect on syncytin expression as well.
Adhesion molecules cadherin-11 and E-cadherin appear to play opposing roles in mediating trophoblast fusion (Getsios and MacCalman, 2003). Whether their mode of action also involves changes in syncytin expression has not yet been studied.
The differentiation of villous trophoblast cells is accompanied by early steps of the apoptosis cascade (Huppertz et al., 1999). These steps also include the caspase-mediated inactivation of flippases (translocases), leading to a flip of phosphatidylserines from the inner to the outer leaflet of the plasmalemma (PS flip). This PS flip was shown to be a prerequisite for fusion of choriocarcinoma cell lines (Lyden et al., 1993; Adler et al., 1995). A monoclonal anti-phosphatidylserine antibody was able to inhibit fusion in this model. The central role of the early molecular machinery of the apoptosis cascade for the process of syncytial fusion is further underlined by the fact that antisense and peptide inhibition of caspase 8, critical for inactivation of flippases, inhibited trophoblast fusion in vitro (Black et al., 2004).
The factors mentioned in this section, and other, perhaps still unknown, fusion-relevant factors may influence fusion at various levels. The roles of some of these factors in fusion have just begun to be investigated. Some of them may influence expression of other fusion-relevant genes, such as the gene encoding syncytin. Other factors may be necessary as co-factors for syncytin-induced fusion, required at the same location at the same time as syncytin or its receptor. Finally, some of the fusion-relevant factors mentioned above may be part of fusion-inducing machineries operating independently of, or alternatively to, syncytin-induced fusion. The expression of syncytin 2 in placenta (Blaise et al., 2003) may represent a redundant mechanism for initiating trophoblast fusion. At present, however, the way by which all these fusion-relevant molecules and mechanisms cooperate to regulate trophoblast fusion is unclear.
|
Are syncytin and its receptors the rate-limiting factors in trophoblast fusion? |
|---|
Taking everything into account, it is tempting to speculate that syncytin and its receptors are merely part of a cascade of events leading to fusion. The immunohistochemical findings of syncytin being present in the syncytiotrophoblast, associated with the functional presence of ASC activity in the basal syncytiotrophoblast membrane, poses the question how trophoblast fusion is limited. If syncytiotrophoblast would just fuse with any underlying cytotrophoblast cell, the pool of regenerative trophoblast cells would soon be exhausted. Is the process leading to fusion limited by a lack of syncytin or receptor at places where fusion should not occur, or is it limited at some other level: by the absence of one of the other factors mentioned above, or by some unknown mechanism? In fact, we still have the impression that syncytin and its receptors can be the most important players in this process, and that the interplay between these two molecules can be rate-limiting to the fusion process. In the remaining part of this review we would like to speculate about a number of models on how syncytin and its receptors might regulate fusion of the trophoblast. These models assume that syncytin and its receptors are the only actors in regulating fusion, for reasons of simplicity only. By discussing these models, we do not want to imply that other molecules or mechanisms are not important. These models are largely theoretical and speculative, but are based upon knowledge about the interaction between retroviral envelope proteins and their receptors.
|
Interaction of retroviral envelope proteins and their receptors—receptor interference |
|---|
Infection of retrovirus into target cells is initiated by fusion of the retroviral envelope with the target cell plasmalemma. The retroviral envelope protein interacts with a cellular receptor, after which an interaction of the envelope protein with the target cell membrane's phospholipid bilayer associated with conformational changes lead to membrane fusion. This process has been studied and described in much detail for some viruses including HIV (reviewed recently by Colman and Lawrence, 2003
). Infection of cells by many retroviruses, including HIV, leads to expression of the retroviral envelope protein on the cell surface, and to syncytialization of the infected cells with other cells expressing the viral receptor (Weiss and White, 1993; Higuchi et al., 2000). The process of cell–cell fusion induced by a retroviral envelope appears to be very similar to retrovirus–cell fusion. In fact, mechanisms and kinetics of retrovirus-cell fusion are often studied in cellular fusion models (Lineberger et al., 2002).
Retroviruses have been divided into various groups based on their receptor usage, even before many of the receptor proteins themselves were known. This grouping of retroviruses was based on observations that cells infected by a certain retrovirus could not be superinfected again by the same retrovirus or by a group of other retroviruses. These groups were called receptor interference groups as it became clear that the viruses within one such group utilized the same receptor. After expression of the retroviral envelope protein by the infected cell, the receptor is blocked and no longer available for infection with a second retrovirus of the same group. Receptor interference was also tested through syncytium formation assays (Sommerfelt and Weiss, 1990). As a consequence, for retrovirus–cell fusion (and cell–cell fusion) to occur, the target cell must not only express the appropriate receptor, but this receptor must also be available for interacting with the viral envelope protein, i.e. it should not be blocked by endogenously produced envelope protein.
Syncytin, being the retroviral envelope protein of HERV-W, very likely regulates membrane fusion in a way similar to other retroviral envelope proteins. Receptors for syncytin have been described (ASCT2 and ASCT1). ASCT2 and ASCT1 are receptors for a large group of retroviruses including spleen necrosis virus (Blond et al., 2000; Marin et al., 2000, 2003). Thus, HERV-W belongs to the same interference group as all the retroviruses using the same receptors. As mentioned above, expression of syncytin by cells normally permissive to spleen necrosis virus induced resistance to infection with this virus (Ponferrada et al., 2003). Syncytin, therefore, has the ability to block its own receptor if coexpressed with its receptor. Consequently, the mere presence of syncytin in a certain cell or syncytial membrane does not necessarily mean that syncytin is available to interact with a receptor molecule in an opposing cell membrane. If syncytin is coexpressed with receptor, either syncytin or the receptor may be blocked, dependent on the expression levels of both proteins. Hence, the balance between local syncytin and receptor expression levels may determine which of both molecules is available.
|
Hypothetical models for regulating trophoblast fusion |
|---|
As explained above, there must be a mechanism limiting syncytiotrophoblast fusion such that it occurs only between certain cytotrophoblast cells and/or certain areas of syncytiotrophoblast, at a certain rate only. One way of regulating this process could be through changing local availabilities of syncytin and its receptors.
Model I
In a simple model, cytotrophoblast cells ready for fusion express only syncytin or receptor, while the basal syncytiotrophoblast membrane provides the other factor required for fusion. Fusion would be limited by the absence of one or the other factor in one of the membranes. Initiation of fusion would occur through up-regulation of the lacking factor in the appropriate membrane. A simplified model is shown in Figure 2-I, in which fusion is shown as a process taking place between two different cells.
|
Model II
A more complicated situation arises if at least one of the fusion partners expresses both the envelope protein and the receptor. In this model, fusion is prevented if both fusion partners overexpress the same factor (Figure 2-II). Initiation of fusion can be achieved by changing the balance between ligand and receptor by down-regulating or up-regulating one or both. Figure 2-II shows three ways of achieving this. In addition to the models shown in Figure 2, even more complicated situations are conceivable: in the non-fusogenic state, one cell type may express both factors, while the other cell type expresses none of them; or: both of the cell types coexpress both factors in the non-fusogenic state.
|
Likely models in the light of recent findings |
|---|
In light of the published data on the distribution of syncytin and receptor, not all of the above theoretical models have the same likelihood. ASC activity, caused by one or both of the receptors for syncytin, was found in basal syncytiotrophoblast membranes (Jansson, 2001
; Cariappa et al., 2003). Nothing is known about local differences in expression along the basal syncytiotrophoblast membrane nor about their expression in cytotrophoblast cells. Syncytin appears to be expressed mainly by syncytiotrophoblast, but there is discrepancy about its localization within the syncytiotrophoblast (see above). Apical expression is not likely to play a direct role in the fusion between syncytium and cytotrophoblast cells. In all studies published, there is no obvious local variation in the expression level of syncytin along the syncytiotrophoblast. Data on syncytin expression in villous cytotrophoblast are also unclear: syncytin staining in villous cytotrophoblast was found to be faint (Smallwood et al., 2003), localized to some areas (Blond et al., 2000) or to some cells only (our findings, Figure 1). Given the fact that both ASC activity and syncytin immunoreactivity have been found in syncytiotrophoblast, it is not unlikely that both proteins are actually coexpressed in the basal syncytiotrophoblast membrane, so that a situation as shown in Figure 2, Model II would arise.
Model I
Assuming that cytotrophoblast or syncytium can only express ligand or receptor at a time, the situation presented in Figure 3, example I is most likely: The basal syncytiotrophoblast membrane expresses receptor, whereas syncytin is expressed in mature cytotrophoblast cells only. The initiative to fuse is taken by the cytotrophoblast cell. This model would not fit with the findings of Lee et al. (2001), who demonstrated syncytin at the basal syncytiotrophoblast membrane in normal placentas. However, it would tally with the study of Frendo et al. (2003b), who found only apical syncytin expression in the syncytiotrophoblast, with Blond et al. (2000) and with our findings (Figure 1) showing syncytin staining in some cytotrophoblast cells. Other possibilities within model I are less likely: syncytin expressed ubiquitously in cytotrophoblast cells is in contrast with immunohistochemical findings, and expression of receptor only in cytotrophoblast cells but not or rarely in the syncytium is in contrast with the functional presence of the amino acid transporter system ASC in the basal syncytiotrophoblast membrane.
|
Model II
Assuming that syncytin and receptor can be expressed simultaneously by cytotrophoblast or, more likely, by syncytium opens more possibilities. The possibilities shown in Figure 3 are the most likely ones in light of published data. The non-fusogenic state is characterized either by overexpression of syncytin by both fusion partners (examples IIA) or by overexpression of receptor by both partners (examples IIB). Examples IIAc and IIBc demonstrate how cytotrophoblast cells can take the initiative to fuse by up-regulating a certain protein. Examples IIAb and IIBb demonstrate how the syncytiotrophoblast can take the initiative to fuse by locally down-regulating a particular protein. Up-regulating a certain protein in the syncytiotrophoblast (corresponding to Model IIa in Figure 2) is another possibility to drive fusion, which is not shown in Figure 3 for reasons of clarity.
Example IIAb displays a paradoxical situation: in the basal syncytiotrophoblast membrane, syncytin is required to inhibit fusion by blocking the availability of receptor, while local down-regulation of syncytin would initiate fusion. Although this may seem an interesting hypothesis, the proposition that syncytin should always be present in cytotrophoblast cells is not in accordance with immunohistochemical findings in the literature.
At this stage of knowledge, we cannot rule out any of the models discussed above because the information on the distribution of syncytin and its receptors is quite sparse and in part even contradictory. However, examples IIB, as shown in Figure 3, have some very interesting and attractive features. Syncytin would be ubiquitously expressed in the basal syncytiotrophoblast membrane, in line with the findings of Lee et al. (2001). Receptors would be present in most of the basal syncytiotrophoblast membrane, which is also in line with literature data, and which makes sense in light of the amino acid transport function of this membrane. Initiation of fusion may take place either through up-regulation of syncytin in differentiating cytotrophoblast cells (Figure 3, example IIBc), through local down-regulation of receptor in the syncytiotrophoblast (Figure 3, example IIBb), or through local up-regulation of syncytin in the syncytiotrophoblast (not shown in Figure 3; it would correspond to Model IIa in Figure 2).
The first possibility (IIBc) is supported by immunohistochemical findings (Blond et al., 2000; this paper, Figure 1) showing syncytin staining in some cytotrophoblast cells.
The second possibility (IIBb) is attractive in that the presence of the receptors ASCT2 or ASCT1 could act as a kind of sensor for syncytial well-being. If the theories on the absence of transcription in the syncytiotrophoblast and the resulting syncytial ageing (see the Introduction section of this paper) are true, ageing areas of the syncytiotrophoblast might lose certain proteins of functional relevance. Only fusion of the syncytium with an underlying cytotrophoblast cell could renew its sources. In this example, loss of amino acid transporters ASCT1 or ASCT2 due to syncytial ageing would stop blockage of syncytin by its receptor(s) and thus initiate the fusion event. In this case the amino acid transporters would serve as syncytial sensors of ageing that directly influence the process that leads to syncytial regeneration.
The last possibility (IIBa, not shown) is that local up-regulation of syncytin in the syncytiotrophoblast (to a level so high that syncytin molecules outnumber receptor molecules), is the driving force towards fusion. Local up-regulation of syncytin may involve transport processes of already available syncytin or its messenger to localized areas of the syncytium where fusion is imminent, or alternatively, may involve de novo transcription of the syncytin gene in such areas.
|
Concluding remarks |
|---|
To fully understand the process of syncytial fusion of trophoblast, there is a number of factors that should be addressed in future work. Focusing on the role of syncytin and its receptors in this type of cell–cell fusion, there is first a need to understand better the molecular mechanisms leading to fusion, as well as a need to clarify the spatial distribution of syncytin and its receptors in trophoblast. To understand the molecular interplay between syncytin and receptors, experimental work is needed involving fusion assays with (genetically engineered) cell lines having inducible expression levels of one or both of these proteins. In this way the influence of syncytin or receptor density on the efficiency and kinetics of fusion can be studied. Similarly, the influence of coexpression of both proteins by one cell could be studied using genetically engineered cell lines. Various antibodies used in immunohistochemistry have led to contradictory data on their distribution. Before the distribution of syncytin can be determined unequivocally, there is a need for comparison and validation of existing and novel anti-syncytin antibodies and of immunohistochemical methods. This would require exchange of existing antibodies and of experimental data between laboratories. In studying the distribution of syncytin, attention should be paid to potential local differences in expression level among the syncytiotrophoblast, and between cytotrophoblast cells. To know the distribution of syncytin receptors in trophoblast, immunohistochemical studies using antibodies against ASCT2 and ASCT1 are required.
|
Acknowledgements |
|---|
We thank Dr H.Chen for providing the polyclonal antibody against syncytin and Angela Rüben for performing the immunohistochemistry. This work was supported by grants from the German Research Council (DFG: PO 718/3-1/3-2 to A.P. and P.K.; FR1245/3-3 to P.K.), and of the Rockefeller Foundation: RF96020#76 and RF99021#114 to P.K.).
|
References |
|---|
Abe E, Mocharla H, Yamate T, Taguchi Y and Manolagas SC (1999) Meltrin-alpha, a fusion protein involved in multinucleated giant cell and osteoclast formation. Calcif Tiss Int 64, 508–515.[CrossRef][ISI][Medline]
Adler RR, Ng A-K and Rote NS (1995) Monoclonal antiphosphatidylserine antibody inhibits intercellular fusion of the choriocarcinoma line JAR. Biol Reprod 53, 905–910.[Abstract]
An DS, Xie Y-M and Chen ISY (2001) Envelope gene of the human endogenous retrovirus HERV-W encodes a functional retrovirus envelope. J Virol 75, 3488–3489.
Benirschke K and Kaufmann P (2000) Pathology of the Human Placenta. 4th edn. Springer Verlag, New York.
Black S, Kadyrov M, Kaufmann P, Ugele B, Emans N and Huppertz B (2004) Syncytial fusion of human trophoblast depends on caspase 8. Cell Death Diff 11, 90–98.[CrossRef][ISI][Medline]
Blaise S, de Parseval N, Bénit L and Heidmann T (2003) Genomewide screening for fusogenic human endogenous retrovirus envelopes identifies syncytin 2, a gene conserved on primate evolution. Proc Natl Acad Sci USA 100, 13013–13018.
Blond J-L, Beseme F, Duret L, Bouton O, Bedin F, Perron H, Mandrand B and Mallet F (1999) Molecular characterization and placental expression of HERV-W, a new human endogenous retrovirus family. J Virol 73, 1175–1185.
Blond J-L, Lavillette D, Cheynet V, Bouton O, Oriol G, Chapel-Fernandes S, Mandrand B, Mallet F and Cosset F-L (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.
Borges M, Bose P, Frank H-G, Kaufmann P and Pötgens AJG (2003) A two-colour fluorescence assay for the measurement of syncytial fusion between trophoblast-derived cell lines. Placenta 24, 959–964.[CrossRef][ISI][Medline]
Cariappa R, Heath-Monnig E and Smith CH (2003) Isoforms of amino acid transporters in placental STB: plasma membrane localization and potential role in maternal/fetal transport. Placenta 24, 713–726.[CrossRef][ISI][Medline]