ABSTRACT
The layers of the epithelial syncytium, i.e., syncytiotrophoblasts, differentiate from chorionic trophoblasts via cell fusion and separate maternal and fetal circulations in hemochorial placentas. L-type amino acid transporter 1 (LAT1) and its covalently linked ancillary subunit 4F2hc are colocalized on both maternal and fetal surfaces of syncytiotrophoblasts, implying their roles in amino acid transfer through the placental barrier. In this study, LAT1 knockout, in addition, revealed a novel role of LAT1 in syncytiotrophoblast development. LAT1 at midgestation was selectively expressed in trophoblastic lineages in the placenta, exclusively as a LAT1-4F2hc heterodimer. In LAT1 homozygous knockout mice, chorionic trophoblasts remained largely mononucleated, and the layers of syncytiotrophoblasts were almost completely absent. The amount of 4F2hc protein, which possesses a fusogenic function in trophoblastic cells, as well as in virus-infected cells, was drastically reduced by LAT1 knockout, with less affecting the mRNA level. Knockdown of LAT1 in trophoblastic BeWo cells also reduced 4F2hc protein and suppressed forskolin-induced cell fusion. These results demonstrate a novel fundamental role of LAT1 to support the protein expression of 4F2hc via a chaperone-like function in chorionic trophoblasts and to promote syncytiotrophoblast formation by contributing to cell fusion in the developing placenta.
INTRODUCTION
The placenta serves as a site of maternofetal exchange of various substances, as well as an immunological barrier and an endocrine organ. Although the overall tissue organization varies among species, the importance of the epithelial syncytium, i.e., syncytiotrophoblasts, that separates maternal and fetal circulations is common to hemochorial placentas in which the maternal blood directly contacts with the fetal chorion. In the labyrinth of mouse placentas, two syncytiotrophoblast layers start to emerge at midgestation by cell fusion of chorionic trophoblasts (1, 2). Two retrovirus-derived murine proteins, syncytins, have been identified to mediate syncytialization of trophoblastic cells in the developing placenta (3, 4). Although the in vivo relevance has not been well addressed, several other proteins, including the single-membrane-spanning type II membrane glycoprotein 4F2hc/CD98hc/FRP-1, have been also suggested to be involved in the cell fusion of trophoblastic cells (5–12). 4F2hc is the heavy chain subunit of 4F2 antigen (CD98) and is covalently linked to a nonglycosylated light chain. The suppression of 4F2hc expression in human trophoblastic BeWo cells reduces forskolin-induced cell fusion, suggesting its significant implication in the syncytiotrophoblast development (8, 13). Independently, 4F2hc has also been characterized as a fusion regulatory protein (FRP-1) relevant to viral infection. The formation of multinucleated giant cells is one of the histopathological features in some types of viral infection. Because some monoclonal antibodies against 4F2hc enhance or suppress the cell fusion induced by different types of viruses, 4F2hc has been regarded as one of the key regulatory molecules for the virus-induced cell fusion (14, 15). Similar effects of anti-4F2hc antibodies on the cell fusion have also been reported in the differentiation of osteoclasts (14, 15). Therefore, the fusogenic function of 4F2hc seems not only to be limited to pathological cell fusion but also to be involved in physiological cell fusion.
As a light chain of 4F2 antigen, we have identified a multiple-membrane-spanning protein, L-type amino acid transporter 1 (LAT1; SLC7A5), which forms heterodimeric complex with 4F2hc via an extracellular disulfide bond and exhibits the function of transporter for large neutral amino acids (16–18). Following the discovery of LAT1, five other members of solute carrier family 7 (SLC7 family) have also been shown to associate with 4F2hc as light chains to form distinct heterodimeric amino acid transporters (19). The heterodimerization is essential for their plasma membrane targeting and functional expression (19). It is now widely accepted that LAT1 is upregulated in various types of cancers to support tumor growth by providing cancer cells with amino acids (20, 21). In contrast, LAT1 protein expression has been reported only in limited normal tissues/organs such as the blood-brain barrier and placenta (22–25). In the human term placenta, LAT1 has been colocalized with 4F2hc at two polarized plasma membrane domains of the syncytiotrophoblast, i.e., the maternally facing apical membrane and the fetally facing basal membrane (22, 25, 26), implying the importance of LAT1-4F2hc heterodimer in the maternofetal transfer of amino acids. Based on studies using primary human trophoblasts, it has been proposed that abrogated cell surface localization of LAT1 may be implicated in intrauterine growth restriction (IUGR) (27, 28), in which placental amino acid transport is impaired (29, 30). To further understand the physiological relevance of LAT1 in vivo, we constructed a LAT1 knockout mouse strain. Unexpectedly, we have found severe defects in the syncytiotrophoblast formation in developing placenta in LAT1 knockout mouse. Because we found that LAT1 suppression decreases 4F2hc expression in the labyrinth and impairs the cell fusion of trophoblastic cells, we propose a novel aspect of placental development that LAT1 contributes to trophoblastic cell fusion by supporting membrane presentation of 4F2hc.
RESULTS
Homozygous deletion of the Lat1 gene results in embryonic lethality at midgestation.Targeted disruption of the Lat1 gene was performed as summarized in Fig. 1A to C (see also Materials and Methods). Heterozygous Lat1 gene knockout (Lat1+/−) mice were phenotypically indistinguishable from wild-type littermates. In contrast, no homozygous knockout (Lat1−/−) offspring were obtained by intercrossing Lat1+/− mice, indicating that systemic homozygous Lat1 gene knockout causes embryonic lethality, as previously reported (31). To determine the timing of lethality in utero, embryos at different developmental stages were analyzed by PCR genotyping. Although three genotypes were obtained at the approximate Mendelian ratio from embryonic day 8.5 (E8.5) to E10.5, no living Lat1−/− embryos were found at E11.5 (Table 1). As shown in Fig. 1D, Lat1−/− embryos exhibited no obvious gross abnormalities at E8.5. Shortly after embryonic turning, the mutant embryos exhibited slightly smaller heads, as well as shorter trunks and tails, which became more obvious at E10.5 (Fig. 1D). All of the dead Lat1−/− embryos at E11.5 were severely degenerated. These results demonstrate that LAT1 is essential for embryogenesis in mice at midgestation.
Targeted disruption of the mouse Lat1 gene. (A) Schematic illustration of the wild-type allele, the targeting construct, and the targeted allele. The entire first exon (Exon1) of the Lat1 gene containing the initiation codon and a part of first intron were replaced with a neomycin resistance cassette that is driven by the RNA polymerase II promoter (polII-neor). The cassette was flanked by 2.3 and 7.5 kb of genomic sequences (gray bars). A diphtheria toxin A subunit gene (DTA) was introduced to the targeting construct as a negative selection marker to enrich the site-directed recombinant clones. The EcoRI restriction sites and two probes used for Southern blot analysis (black bars) and three primers for genotyping PCR analysis (arrowheads) are indicated. (B) Southern blot analysis of wild-type TT2 mouse embryonic stem cells (Lat1+/+) and the recombinant clone used to establish the knockout strain (Lat1+/−). The Ext probe in the external position of the targeting construct yielded 6.0- and 7.7-kb bands for the wild-type and targeted alleles, respectively (left panel). The Neo probe designed against the neomycin resistance cassette yielded a band at 7.7 kb only for the targeted allele (right panel). (C) Genotyping PCR analysis of wild-type (Lat1+/+), heterozygous knockout (Lat1+/−), and homozygous knockout (Lat1−/−) embryos at E10.5 using genomic DNA extracted from the yolk sac. Bands at 1.0 and 1.5 kb were amplified from the wild-type and targeted alleles, respectively. (D) Gross morphology of Lat1+/+ and Lat1−/− embryos at midgestation. Scale bars, 1 mm.
Genotypes of offspring from intercrossed Lat1+/− mice
LAT1 depletion causes abnormalities in developing placental labyrinth.Since Lat1−/− embryos start to show abnormalities at midgestation when the placenta starts to be formed, we examined Lat1−/− placentas. As illustrated in Fig. 2A, the allantois attaches and fuses with the chorionic plate at E8.5. The allantoic mesodermal villi then start to invade the chorionic plate to form primary fetal blood vessels. The blood vessel network increases its complexity by further invasion, elongation, and branching, thereby increasing the contact surface area with maternal blood sinus. Simultaneously, at around E9.5 and E10.5, chorionic trophoblasts surrounding the endothelial cells start to fuse and differentiate into the two layers of syncytiotrophoblasts, which separate the maternal and fetal circulations (Fig. 2B). There were no gross abnormalities in Lat1−/− placentas at E9.5 (Fig. 2C). The allantois normally contacted and fused with the chorionic plate. Invasion of mesodermal villi from the fetal side into the chorionic plate also occurred normally. The ectoplacental cavity occluded similar to the wild-type mice, so that the chorionic plate was placed close to the ectoplacental cone (Fig. 2C). At E10.5, the primitive architecture of the labyrinth started to form in wild-type placentas (Fig. 2D). Fetal blood vessels and maternal sinusoids were intertwined with each other. The syncytiotrophoblast-like membranous structures separating the fetal vessels and maternal sinusoids were found in well-vascularized regions of the labyrinth. In contrast, the chorionic plate in the Lat1−/− placenta was obviously undifferentiated, i.e., chorionic trophoblasts remained tightly packed, and only a few less-branched fetal blood vessels and maternal sinusoids were formed (Fig. 2D). The impaired fetal blood vessel formation in the Lat1−/− placenta was also confirmed by immunostaining of an endothelial marker CD34 (Fig. 2F). Consistent with the gross abnormality of labyrinth, when a fluorescent dye, rhodamine 123, was injected into a pregnant female at E10.5, a Lat1−/− embryo exhibited obviously lower fluorescence compared to Lat1+/+ and Lat1+/− littermates, presumably due to its reduced placental permeation of the dye (Fig. 2G). Despite such severe developmental defects in the labyrinth, the trophoblast giant cell layer and junctional zone seemed unaffected in the Lat1−/− placenta (Fig. 2D). At E11.5, fetal capillaries were almost completely absent in the labyrinth of the Lat1−/− placenta (Fig. 2E). Although the number of maternal sinusoids was increased in the Lat1−/− placenta, the enlarged sinusoids were often filled with coagulations of maternal blood. These results indicate that LAT1 is essential for the development of the labyrinth.
Developmental abnormalities in Lat1−/− placentas. (A) Tissue organizations of the developing mouse placenta at midgestation. Abbreviations for the placental tissues used in the microscopic images throughout this study are indicated below. (B) Architecture of labyrinth demonstrating maternal and fetal circulation separated by two syncytiotrophoblast layers. Fetal red blood cells were shown with blue-colored nuclei. (C to E) Hematoxylin-eosin-stained sagittal sections of placentas from wild-type (Lat1+/+) and homozygous knockout (Lat1−/−) littermates. The upper and lower images in panels C and E, as well as the top and middle images in panel D, are low and high magnifications, respectively. The bottom images in panel D are further enlarged images of the labyrinth. Insets show mesodermal villi invading into the chorionic plate. (C) At E9.5, the tissue organization of Lat1−/− placentas appeared to be normal. The allantois (“Al”) contacted and fused with the chorionic plate (“CP”). The ectoplacental cavity between chorionic plate and ectoplacental cone (“EPC”) was occluded. Invasion of mesodermal villi (arrowheads) into the chorionic plate also occurred normally. (D) At E10.5, the development of labyrinth (“La”) in Lat1−/− placentas was impaired. In the Lat1+/+ placenta, fetal blood vessels containing nucleated erythrocytes and maternal sinusoids contacting enucleated mature erythrocytes were intertwined with each other and separated by the membranous syncytiotrophoblasts (arrows). The labyrinth of mutant placenta at E10.5 was thin and less vascularized. The chorionic plate (“CP”) remained tightly packed. Syncytiotrophoblast-like structures were not developed, whereas the trophoblast giant cell layer (“TGC”) and junctional zone (“JZ”) were well organized. (E) At E11.5, the tissue organization of the labyrinth in the Lat1−/− placenta was severely impaired. Fetal blood vessels were almost completely absent, except at the initial invasion sites of mesodermal villi (arrowheads). Coagulations of maternal blood were often observed in enlarged sinusoidal cavities (asterisks). (F) Immunofluorescence staining of CD34 in Lat1+/+ and Lat1−/− placentas from littermates at E10.5. The nuclei were stained with DAPI. (G) Rhodamine 123 permeation assay on E10.5 littermates. Rhodamine 123 fluorescence was compared among Lat1+/+, Lat1+/−, and Lat1−/− embryos. De, decidua. Paraffin sections were used for panels C to F. Scale bars: 500 μm for the upper rows of panels C to E and panel F, 100 μm for the lower rows of panels C and E and the middle row of panel D, 20 mm for bottom row of panel D, and 1 mm for panel G.
LAT1 is expressed in trophoblastic lineages of the chorionic plate and ectoplacental cone in the developing placenta.We examined the expression of LAT1 and its localization in the developing placenta. Western blot analysis of the E10.5 placenta showed that the expression of LAT1 protein is already detectable at this developmental stage (Fig. 3A). Compared to the wild-type placenta, the expression of LAT1 was reduced to ∼50% in the Lat1+/− placenta, whereas it was completely absent in the Lat1−/− placenta, confirming that the targeted disruption of Lat1 gene resulted in a null allele (Fig. 3A). At the mRNA level, LAT1 was most abundantly expressed in the chorionic plate in an E9.5 placenta (Fig. 3B). Lower but significant expression was detected in a limited region of the ectoplacental cone adjacent to the chorionic plate. LAT1 expression was maintained in the corresponding regions in an E10.5 placenta (Fig. 3C), in which the chorionic plate and ectoplacental cone differentiated into the labyrinth and junctional zone, respectively. In contrast, the trophoblast giant cell layer lacked the expression of LAT1 mRNA.
Expression of LAT1 in developing placentas. (A) Western blot analysis of Lat1+/+, Lat1+/−, and Lat1−/− placentas (E10.5) from littermates using an anti-LAT1 antibody. (B and C) In situ hybridization to reveal LAT1 mRNA expression. The upper and lower images in panels B and C show low and high magnifications, respectively. (B) LAT1 mRNA was most abundantly expressed in the chorionic plate (“CP”) at E9.5. Lower but significant expression was also detected in a region of the ectoplacental cone (“EPC”) adjacent to the chorionic plate at E9.5. (C) LAT1 mRNA was persistently expressed in the corresponding regions in the E10.5 placenta, i.e., the labyrinth (“La”) and junctional zone (“JZ”), but was absent in the trophoblast giant cell (“TGC”). Dashed lines indicate the boundaries between placental tissues. Al, allantois; De, decidua. Paraffin sections were used for panels B and C. Scale bars: 500 μm for upper images in panels B and C and 250 μm for lower images in panels B and C.
Consistent with these observations, immunofluorescence microscopy of an E10.5 placenta revealed that LAT1 protein was localized in the labyrinth and the junctional zone (Fig. 4A). Particularly high expression of LAT1 was often detected in the region adjacent to the CD34-positive fetal blood vessels in the labyrinth. Although some immunoreactive signals were observed in the perinuclear regions of trophoblast giant cells, these were most likely to be nonspecific because similar signals were detected in a Lat1−/− placenta. As shown under a higher magnification in Fig. 4B, LAT1 was localized at the plasma membrane of each mononucleated chorionic trophoblast, which had not differentiated into the syncytiotrophoblasts. LAT1 was not detectable at the plasma membrane of sinusoidal giant trophoblasts (Fig. 4B), a subtype of giant trophoblasts in maternal sinusoids (2). The plasma membrane of trophoblasts in the junctional zone was also positive for LAT1 staining (Fig. 4B). In well-vascularized regions of the labyrinth, in addition to the plasma membrane of remaining mononucleated chorionic trophoblasts, LAT1 was often detected as two parallel membranous structures located between the maternal sinusoids and fetal capillaries (Fig. 4C). LAT1 expression tended to be higher in the vicinity of CD34-positive fetal capillaries in developing syncytiotrophoblasts at E10.5 (Fig. 4C and D). Connexin 26-positive gap junctions (32, 33) were located between the two lines of LAT1-positive membranous structures (Fig. 4D). Therefore, LAT1 is localized to the apical surface of developing syncytiotrophoblast layer I (SynT-I) and the basal surface of syncytiotrophoblast layer II (SynT-II), which face the maternal sinusoids and fetal capillaries, respectively (see Fig. 2B). Such localization of LAT1 in syncytiotrophoblasts was further supported by the observation in a mature placenta, showing that LAT1 and its heavy chain subunit 4F2hc were detected in distinctive two parallel lines that were accompanied by an intervening line of connexin 26-positive gap junctions (Fig. 5).
Localization of LAT1 in midgestational placentas. (A) Immunofluorescence microscopy of LAT1 and CD34 in Lat1+/+ (upper panels) and Lat1−/− placentas (lower panels) from littermates at E10.5. High expression of LAT1 was often observed around the CD34-positive fetal capillaries (arrowheads). Trophoblast giant cells (“TGC”) harbored nonspecific perinuclear signals in LAT1 staining in both genotypes (asterisks). The Lat1−/− labyrinth (“La”) and junctional zone (“JZ”) were mostly negative for CD34 staining due to poor fetal blood vessel development. (B) Magnified images of LAT1 immunofluorescence microcopy in the nonsyncytialized region of the Lat1+/+ labyrinth (left panels) and junctional zone (right panels). Cells filling the greater part of the observed fields were mononucleated chorionic trophoblasts (left panels) and trophoblasts in the junctional zone (right panels), respectively, which express LAT1 at the plasma membrane. Sinusoidal giant trophoblasts lacking LAT1 expression are indicated (open triangles). (C) Magnified images of LAT1 and CD34 immunofluorescence microcopy in syncytialized regions of the Lat1+/+ labyrinth. In addition to the plasma membrane of the remaining mononucleated chorionic trophoblasts, LAT1 was often detected in two parallel membranous structures (arrowheads) between maternal sinusoids (asterisks) and CD34-positive fetal capillaries. Sinusoidal giant trophoblasts lacking LAT1 expression in maternal sinusoids are indicated (triangles). Most of the remaining mononucleated cells that express LAT1 at the plasma membrane were undifferentiated chorionic trophoblasts. (D) Enlarged images of triple staining of LAT1, CD34, and connexin 26 (CX26) in the area indicated with a dashed square in panel C. The intense signals of connexin 26 staining in the maternal sinusoids most likely arose from a cross-reaction of the secondary antibody with mouse IgG in the maternal sinusoids (asterisks). Al, allantois; CP, chorionic plate; De, decidua. Frozen sections were used for panels A to D. Scale bars: 50 μm.
Localization of LAT1 and 4F2hc in Lat1+/+ mature placental labyrinth at E17.5. (A) Immunofluorescence microscopy of LAT1 and 4F2hc in the Lat1+/+ labyrinth at E17.5. LAT1 and 4F2hc were colocalized in two parallel lines in labyrinth. (B) Double staining of 4F2hc and CD34 indicating that 4F2hc-expressing double-lined membranous structures, thereby expressing LAT1 based on panel A were located adjacent to CD34-positive fetal capillaries. (C) Double staining of 4F2hc and connexin 26. Connexin 26-positive gap junctions were located between the two lines of 4F2hc-expressing two parallel lines, which indicates that 4F2hc, as well as LAT1, was expressed at the apical surface of syncytiotrophoblast layer I and the basal surface of syncytiotrophoblast layer II (see Fig. 2B). Frozen sections were used for panels A to C. Magnified views are shown in each inset. Scale bars: 50 μm (10 μm for the inset images).
The expression of heavy chain subunit 4F2hc in chorionic trophoblasts highly depends on the presence of LAT1.Because LAT1 and 4F2hc are colocalized in the labyrinth of the mature placenta (Fig. 5A), we examined whether LAT1 is also associated with 4F2hc in the midgestational placenta and whether the LAT1 knockout affects the expression of 4F2hc. In a Western blot analysis of an Lat1+/+ placenta under nonreducing conditions, LAT1 and 4F2hc were detected at identical molecular sizes of 130 to 150 kDa, whereas, under reducing conditions, the bands corresponding to LAT1 and 4F2hc shifted to ∼37 kDa and 70 to 90 kDa, respectively (Fig. 6A). This suggests that LAT1 exists exclusively as a heterodimer associated with 4F2hc in the placenta at this developmental stage. The localization of 4F2hc was quite similar to that of LAT1 in the E10.5 placenta (Fig. 6B). 4F2hc immunostaining was positive in the plasma membranes of trophoblasts in the labyrinth and junctional zone but not in the trophoblast giant cell layer (Fig. 6B). Consistent with the expression of LAT1, 4F2hc was often detected as two parallel membranous structures at the boundary between the maternal sinusoids and fetal blood vessels, suggesting its colocalization with LAT1 on the apical surface of SynT-I and the basal surface of SynT-II. Similar to LAT1, the expression of 4F2hc tended to be higher on the basal surface of SynT-II than that on the apical surface of SynT-I at E10.5.
LAT1-dependent expression of 4F2hc in developing placentas. (A) Western blot analysis of a Lat1+/+ placenta was performed using an anti-LAT1 antibody (left panel) and an anti-4F2hc antibody (right panel) under reducing [(+)] and nonreducing [(−)] conditions. The bands corresponding to LAT1-4F2hc heterodimer at ∼140 kDa are indicated by arrows. The bands of LAT1 (∼37 kDa) and 4F2hc (70 to 90 kDa) monomers are indicated by black and white arrowheads, respectively. (B) Immunofluorescence microscopy of 4F2hc and CD34 in Lat1+/+ and Lat1−/− placentas from littermates at E10.5. The localization of 4F2hc in a Lat1+/+ placenta (left panel) was quite similar to that of LAT1. The region indicated by the dashed rectangle is enlarged and shown in the lower panels. The two parallel membranous 4F2hc-positive structures located between maternal sinusoids and CD34-positive fetal capillaries (arrowheads) are shown in the lower panels. Lat1−/− labyrinth (“La”) almost completely lacked 4F2hc expression. JZ, junctional zone; TGC, trophoblast giant cell layer. Paraffin sections were used. Scale bars: 50 μm. (C) Western blot analysis of whole lysates of Lat1+/+ and Lat1−/− placentas from littermates at E9.5 to E11.5 using an anti-LAT1 antibody and an anti-4F2hc antibody under reducing conditions. β-Actin was detected as a loading control. (D) Quantification of LAT1 and 4F2hc mRNAs by real-time PCR. Placental RNA was isolated from whole littermates at E9.5, E10.5, and E11.5 and analyzed. Data were normalized with respect to the expression of GAPDH mRNA and are reported as a percentage of the averaged values for Lat1+/+ littermates. PCR analyses were performed in quadruplicate (n = 4) and are reported as means ± the standard deviations (SD).
It was striking that the intensity of 4F2hc staining was drastically decreased in the Lat1−/− placenta (Fig. 6B). The labyrinth in the Lat1−/− placenta almost completely lacked the signals of 4F2hc. Only a residual 4F2hc staining was detectable in the junctional zone. This was further supported by Western blot analysis of placentas between E9.5 and E11.5 (Fig. 6C). In the Lat1+/+ placenta, 4F2hc, as well as LAT1, gradually increased from E9.5 to E11.5, whereas in the Lat1−/− placenta, 4F2hc protein was considerably reduced to ∼15% at E9.5 and ∼20% at E10.5 compared to the Lat1+/+ placenta. These results suggest that the expression of 4F2hc in chorionic trophoblasts heavily depends on the coexpression of its light chain subunit LAT1. It is important to emphasize that, in Lat1+/− and Lat1−/− placentas, the amounts of 4F2hc mRNA at E9.5, E10.5, and E11.5 were maintained at a level comparable (mostly higher than 80% and not less than 70%) to that in Lat1+/+ placentas (Fig. 6D), suggesting that the decrease in 4F2hc protein amount in Lat1−/− placentas is not primarily due to the reduced transcription of 4F2hc gene.
Knockdown of LAT1 suppresses 4F2hc expression and forskolin-induced syncytialization of human trophoblastic BeWo cells.LAT1 and 4F2hc endogenously expressed in human trophoblastic BeWo cells colocalized at the plasma membrane, suggesting the presence of LAT1-4F2hc heterodimer (Fig. 7A). Because the suppression of 4F2hc expression in BeWo cells was reported to reduce forskolin-induced cell fusion (8, 13), we performed an in vitro cell fusion assay in combination with small interfering RNA (siRNA)-mediated LAT1 knockdown in BeWo cells to examine whether the suppression of LAT1 expression affected cell fusion in the development of the syncytiotrophoblast. Clones of BeWo cells individually expressing either AcGFP or mCherry were cocultured and stimulated by forskolin to induce cell fusion. The fused cells were then visualized as those positive for both AcGFP and mCherry fluorescence (Fig. 7B). After transfection of three different siRNAs against LAT1, the LAT1 protein was decreased to 10 to 15% compared to mock-treated cells (Fig. 7C). As shown in Fig. 7C, LAT1 knockdown caused a considerable decrease in the amount of 4F2hc (50 to 60% compared to untreated cells). Although this decrease in 4F2hc protein was similar to what we observed in Lat1−/− placentas (Fig. 6C), we found that the 4F2hc mRNA was also significantly decreased by knocking down LAT1 in BeWo cells (Fig. 7D, 40 to 50% compared to untreated cells) in contrast to the sustained 4F2hc mRNA level in Lat1−/− placentas (Fig. 6D). When the cell fusion was induced by forskolin, the syncytium formation was significantly decreased 50 to 60% by LAT1 knockdown, which further supported the essential roles of LAT1 in the expression of 4F2hc and so in trophoblastic cell fusion (Fig. 7E).
Suppression of 4F2hc expression and forskolin-induced syncytialization of human trophoblastic BeWo cells by LAT1 knockdown. (A) Immunofluorescence staining of endogenous LAT1 and 4F2hc in BeWo cells. Magnified views of the area indicated by dashed squares are shown in the insets. LAT1 and 4F2hc were colocalized on the plasma membrane of BeWo cells. (B) In vitro cell fusion assay by using two BeWo cell clones stably expressing either mCherry or AcGFP. The cells were cocultured and treated with either 10 μM forskolin (upper panels) or DMSO (0.01%, solvent for forskolin; lower panels) for 48 h. Syncytialized cells were visualized as those positive for both fluorescent proteins (asterisks). Scale bars: 100 μm. (C and D) Knockdown of LAT1 by siRNA and its effect on the expression of 4F2hc at protein and mRNA level. (C) BeWo cells were transfected with either negative control siRNA (NC siRNA#1 and siRNA#2) or LAT1-targeting siRNA (LAT1 siRNA#1 to siRNA#3). At 72 h after transfection, the cells were collected and subjected to Western blot analysis under reducing conditions using anti-LAT1 antibody or anti-4F2hc antibody. Anti-Na+/K+ ATPase α1 antibody was used for a loading control. (D) Under the condition used for panel C, mRNA expression of LAT1 and 4F2hc was analyzed by real-time PCR. Data were normalized with respect to the expression of GAPDH mRNA and are reported as the percentage of the averaged values of untreated cells (Mock). PCR analyses were performed with quadruplicate (n = 4) and are shown as means ± the SD. (E) Effect of LAT1 knockdown on forskolin-induced cell fusion in an in vitro cell fusion assay. At 24 h after siRNA transfection, the cells were treated with 10 μM forskolin or DMSO (the solvent for forskolin) for 48 h. Cell fusion was quantified by calculating the percentage of nuclei located in cells positive for both AcGFP and mCherry fluorescence. The data shown are means ± the standard errors of the mean (n = 7). Asterisks indicate a statistically significant difference according to the unpaired Student t test (P < 0.05).
Depletion of LAT1 has no apparent influence on cell proliferation or apoptosis of trophoblasts in the placenta.Because the LAT1-4F2hc heterodimer preferentially transports large neutral amino acids, including essential amino acids (17, 18), we performed Ki-67 staining and bromodeoxyuridine (BrdU) incorporation assay (Fig. 8) to examine whether LAT1 depletion affects cell proliferation of trophoblasts and consequently impairs syncytiotrophoblast development. The results from both cell proliferation assays indicated no obvious differences between Lat1+/+ and Lat1−/− placentas. For either genotype, most cells in the chorionic plate and ectoplacental cone at E9.5 were positive for both proliferation markers (Fig. 8A and B). Similarly, in Lat1−/− placentas at E10.5, despite the impaired formation of labyrinth and the marked reduction in its thickness, both proliferation markers were detected in most of the undifferentiated chorionic trophoblasts (Fig. 8C and D). As shown in Fig. 8E, Ki-67 labeling indices determined based on Ki-67 staining were not significantly different between Lat1+/+ and Lat1−/− placentas in the ectoplacental cone and chorionic plate at E9.5 and in the junctional zone and labyrinth at E10.5. Furthermore, apoptotic cell death was not enhanced by homozygous Lat1 gene knockout at either E9.5 or E10.5 (Fig. 9A and B). These results suggest that the impaired placental development in the Lat1−/− genotype is not due to an effect on the proliferation or survival of trophoblasts in the chorionic plate and the ectoplacental cone. In addition, the phosphorylation of p70S6K, 4EBP1, and eukaryotic translation initiation factor 2α (eIF2α), which reflects amino acid availability and the state of protein synthesis (34), was not significantly altered in the LAT1 homozygous knockout placenta (Fig. 9D and E). p70S6K and 4EBP1 are the direct substrates of mechanistic target of rapamycin complex 1 (mTORC1), a serine/threonine kinase complex activated by amino acids, especially leucine, arginine, and glutamine, which regulate protein synthesis (34). The translation initiation factor eIF2α is phosphorylated by general control nonderepressible 2 kinase (GCN2) activated by uncharged tRNAs in response to a shortage of essential amino acids (34). The absence of apparent effects of LAT1 knockout on the phosphorylation of these proteins might be, at least in part, due to the compensatory upregulation of autophagy because the lipidated form of the autophagic marker LC3 (35) was slightly increased in the placentas of LAT1 homozygous knockout mice compared to that of wild-type littermates (Fig. 9D and E). These results support the notion that the indispensable developmental function of LAT1 in midgestational placentas is not primarily to support cell proliferation or survival by supplying amino acids for protein synthesis.
Effect of Lat1 knockout on Ki-67 staining and BrdU incorporation in the midgestational placentas. (A and C) Ki-67 staining of Lat1+/+ and Lat1−/− placentas from littermates at E9.5 (A) and E10.5 (C). The small top panels are images of Ki-67 (left panels) and nuclear (right panels) staining. Merged enlarged images are shown in the lower row of panel A and the middle row of panel C. Magnified images of the labyrinth (“La”) at E10.5 are shown in the lower row of panel C. (B and D) BrdU incorporation assay in Lat1+/+ and Lat1−/− placentas from littermates at E9.5 (B) and E10.5 (D). Low- and high-magnification images are shown in the upper and lower rows, respectively. No apparent difference was detected between Lat1+/+ and Lat1−/− placentas. (E) Ki-67 labeling indices determined based on the Ki-67 staining (A and C) in ectoplacental cone (“EPC”) and chorionic plate (“CP”) at E9.5 (n = 3) and in junctional zone (“JZ”) and labyrinth (“La”) at E10.5 (n = 5). The data reported are means ± the SD. No significant difference was observed between Lat1+/+ and Lat1−/− placentas. NS, not significant. Al, allantois; De, decidua; TGC, trophoblast giant cell. Paraffin sections were used for panels A to D. Scale bars: 500 μm for the upper rows of panels B and D and middle row of panel C and 100 μm for lower rows of panels A to D.
Effect of Lat1 knockout on apoptosis, mTORC1 activity, amino acid availability, and autophagy in the midgestational placenta. (A and B) Apoptotic cells were detected by an in situ TUNEL assay in Lat1+/+ and Lat1−/− placentas from littermates at E9.5 (A) and E10.5 (B). No apparent difference was observed between Lat1+/+ and Lat1−/− placentas. (C) Apoptotic cell death in the sections of rat mammary gland as a positive control of the detection of apoptotic cells. (D) Western blot analysis of the lysates of Lat1+/+ (left four lanes) and Lat1−/− (right three lanes) placentas at E10.5 from littermates was performed using the indicated antibodies. For p70S6K, 4EBP1, and eIF2α, both phosphorylated proteins (upper panels) and total proteins (lower panels) were detected. The bands of LC3 corresponding to the cytosolic form and the autophagosome-associated lapidated form are indicated as LC3-I and LC3-II, respectively. (E) Densitometric analysis of the immunoreactive bands in panel D. The relative band intensities of the phosphorylated and total proteins of p70S6K, 4EBP1, and eIF2α, as well as those of LC3-II and LC3-I, are shown. The data shown are means ± the SD (n = 4 for Lat1+/+ and n = 3 for Lat1−/− placentas). The asterisk indicates a statistically significant difference in the unpaired Student t test (P < 0.05). NS, not significant.
DISCUSSION
Cell-cell fusion is a drastic cellular event that occurs under both physiological and pathological conditions (36, 37). It dominates the differentiation of myotubes in skeletal muscle from myoblasts, osteoclasts in bone from monocytic progenitors, and syncytiotrophoblasts in placentas from chorionic trophoblasts. Cell fusion is therefore commonly involved in the processes to construct the complex architecture of some tissues/organs in multicellular organisms. It also plays a role in pathological syncytium formation in viral infection and cancers. However, the molecular mechanisms underlying cell fusion remain largely unknown. The player molecules participating in the process have not yet been identified. Here, we report that LAT1, the light chain subunit of heterodimeric LAT1-4F2hc complex, is involved in the cell fusion of chorionic trophoblastic cells and is critical for syncytiotrophoblast development in the placenta.
LAT1 and 4F2hc colocalize in the apical and basal surfaces of a single syncytiotrophoblast layer in the human term placenta (22, 25, 26). We found that LAT1, together with 4F2hc, is located in the apical surface of SynT-I and the basal surface of SynT-II in mature mouse placentas (Fig. 5). Even though the number of syncytium layers differs between the species, the analogous localizations in humans and mice strongly suggests the conserved function of LAT1-4F2hc heterodimer in amino acid transfer through the mature placenta. The targeted disruption of Lat1 gene in the present study provided further insight into the fundamental roles that LAT1 plays in the development of the placenta. We found that LAT1 knockout severely affected the development of the labyrinth, preserving the junctional zone developed normally, although LAT1 was expressed both in the trophoblastic cells in a chorionic plate that differentiate into syncytiotrophoblasts in the labyrinth and those in the ectoplacental cone that differentiate into spongiotrophoblasts in the junctional zone (Fig. 4). The chorionic plate remained compactly packed even at E10.5 with undifferentiated chorionic trophoblasts in LAT1 homozygous knockout placenta, where the formation of syncytiotrophoblasts was obviously less than that in the wild type. The substantial decrease in the contact surface area between maternal and fetal circulations due to the defect of labyrinth development was confirmed in LAT1 homozygous knockout mice by a rhodamine 123 permeation assay (Fig. 2G).
Because Western blot analysis conducted under reducing and nonreducing conditions suggested that LAT1 in the midgestational mouse placenta is present exclusively in the form of a heterodimer with 4F2hc (Fig. 6A), we studied whether the expression of LAT1 affects that of 4F2hc. We showed that 4F2hc protein was markedly decreased in the Western blot of whole placenta in LAT1 homozygous knockout mice (Fig. 6C). Furthermore, in immunofluorescence microscopy, the expression of 4F2hc was almost completely lost in chorionic trophoblasts of labyrinth in LAT1 homozygous knockout placenta (Fig. 6B), indicating that LAT1 is essential for the expression of 4F2hc in the labyrinth of midgestational placenta. Despite such a drastic reduction in 4F2hc protein, 4F2hc mRNA was maintained at the level comparable (mostly higher than 80% and not less than 70%) to that of the wild-type placenta (Fig. 6D). LAT1 would thus primarily support the expression of 4F2hc at the protein level but not at the transcriptional level in vivo. It was previously reported in a heterologous expression system that b0,+AT, a light chain subunit of heterodimeric amino acid transporters, supported the plasma membrane expression and thereby influenced the total protein amount of its heavy chain subunit rBAT (38). The present study has for the first time shown that such a chaperone-like function of the light chain subunit is also applicable to 4F2hc, the other heavy chain of heterodimeric amino acid transporters, and furthermore provided an in vivo evidence for the importance of light chain subunits to maintain the protein expression of the heavy chain subunits.
As indicated above, one of the most remarkable phenotypes of the LAT1 homozygous knockout placenta was the defect in labyrinth development. It was previously reported that the knockdown of 4F2hc suppressed the cell fusion of human trophoblastic BeWo cells (8, 13), although the relevance to syncytiotrophoblast development in vivo has not been demonstrated because of the embryonic lethality caused by 4F2hc knockout before the development of placenta (39). Because LAT1 knockout reduced 4F2hc protein in the chorionic trophoblasts of labyrinth (Fig. 6B) and suppressed the formation of syncytiotrophoblasts and remained mononucleated chorionic trophoblasts undifferentiated in chorionic plate/labyrinth (Fig. 2C to E), we examined whether LAT1 and thus LAT1-4F2hc heterodimer are involved in the syncytialization of chorionic trophoblasts. We showed that the knockdown of LAT1 in BeWo cells reduced the expression of 4F2hc, as observed in LAT1 homozygous knockout placenta, and suppressed the forskolin-induced cell fusion (Fig. 7C and E). Taken together, these observations suggest that the defect in syncytiotrophoblast formation in LAT1 homozygous knockout placenta is possibly due to the reduction in 4F2hc protein in chorionic trophoblasts caused by the depletion of LAT1. It is of note that LAT1 knockdown in BeWo cells caused a decrease in 4F2hc mRNA (Fig. 7D) in contrast to the sustained 4F2hc mRNA level observed in the placentas of LAT1 homozygous knockout mice (Fig. 6D). Considering the long half-life (27.2 h) of 4F2hc protein in cultured cells (40), the observed reduction of the 4F2hc mRNA cannot solely explain the decrease of 4F2hc protein in LAT1 knockdown cells. A similar observation was recently reported regarding the genetic disruption of LAT1 in colon adenocarcinoma LS174T cells (41). In that study, the reduction of 4Fh2c protein (∼80% reduction) was much more drastic than that of mRNA (∼40% reduction). It is thus proposed that, in addition to the chaperone-like function of LAT1 at the protein level, other mechanisms which link the reduced LAT1 expression to the reduction of 4F2hc expression at the transcription level would be working in vitro.
4F2hc has been shown to be involved in cell fusion processes not only in trophoblastic cells but also in virus-infected cells and osteoclasts (8, 13–15). We therefore suggest that some common mechanisms involving 4F2hc underlie such physiological and pathological cell fusions. A previous study indicated that galectin-3, a member of lectin family, is a ligand for 4F2hc and binds to the extracellular domain of 4F2hc to promote BeWo cell fusion (6). In that study, based on the observation that lactose, a disaccharide ligand of galectin-3, interfered with the forskolin-induced cell fusion, galectin-3 was suggested to recognize N-linked sugar chains in the extracellular domain of 4F2hc. Since galectin-3 multimerizes through the intermolecular interactions at its N-terminal domain (42), two plasma membrane may be cross-linked by the assembly of LAT1-4F2hc and galectin-3, which would promote the fusion process.
It is also of note that Na+-dependent small neutral amino acid transporters, ASCT1 and ASCT2, have been reported to be engaged in cell fusion by serving as receptors for human syncytin-1 protein (5, 9). Intriguingly, LAT1-4F2hc and ASCT2 are detected in a chemically cross-linkable protein complex in several human cancer cell lines (43). Therefore, although the receptor for mouse syncytins has not yet been identified, it is tempting to propose a fusion machinery corresponding to the large molecular assembly involving LAT1-4F2hc and ASCT2 in the plasma membrane of trophoblastic cells in both mouse and human placentas. Such a molecular assembly could mediate the cell fusion through the interaction of ASCT2 with syncytins in combination with the interaction of LAT1-4F2hc with galectin-3 described above. Since human and mouse syncytins are retroviral in origin and captured into the genomes during the evolution (44), it would be interesting to propose that the molecular complex involving 4F2hc and ASCT1/ASCT2 integrates the cell fusion process mediated by syncytins and galectin-3 in the generation of multinucleated cells by viral infection and in other conditions without involving viral infection. In trophoblastic cells in the placenta, the plasma membrane expression of 4F2hc protein has been suggested to be primarily dependent on LAT1 among six known light chain subunits for 4F2hc (Fig. 6B and C). However, in virus-infected cells and osteoclasts in which 4F2hc is also involved in cell fusion processes, the contribution of LAT1 or other light chains subunits remains to be investigated in the future.
Although LAT1 is an amino acid transporter for large neutral amino acids, including leucine and other essential amino acids, LAT1 knockout did not negatively influence the cell proliferation and survival of trophoblastic cells in the chorionic plate and the ectoplacental cone (Fig. 8 and 9A and B). BrdU incorporation, Ki-67 staining, and a TUNEL (terminal deoxynucleotidyltransferase-mediated dUTP-biotin nick end labeling) apoptosis assay did not support the suppression of proliferation and survival. We also confirmed that the phosphorylation of p70S6K, 4EBP1, and eIF2α, which reflects amino acid availability and the state of protein synthesis (34), was not significantly altered in LAT1 homozygous knockout placentas (Fig. 9D and E), whereas autophagy seems to be slightly upregulated in the placentas of LAT1 homozygous knockout mice compared to wild-type littermates. These results further suggest that the developmental significance of LAT1 in midgestational placenta is not primarily to support cell proliferation or survival by supplying amino acids for protein synthesis.
It is, however, still possible that the amino acid transport activity of LAT1 has a functional impact on the cell fusion process in addition to the proposed chaperone-like function supporting 4F2hc expression. The trafficking of LAT1 to the plasma membrane is positively regulated downstream of mTORC1 (and probably of mTORC2 as well) in primary human trophoblasts (27, 28). Because mTORC1 is activated by amino acids (34), the amino acid uptake mediated by LAT1-4F2hc may increase the cell surface amount of LAT1-4F2hc itself, thereby further promoting the cell fusion of the chorionic trophoblasts. If such a positive-feedback loop exists, LAT1-4F2hc would amplify the focal amino acid availability information and couple it with the cell fusion process, so that the syncytium formation could be induced in close proximity to maternal sinusoids, where the local amino acid concentration is supposed to be high. Consistent with this, the expression of LAT1 and 4F2hc was particularly higher in developing syncytiotrophoblasts than in mononucleated chorionic trophoblasts (Fig. 4C and 6B). Although mTORC1 activity was not apparently decreased in Western blot analyses of whole placental lysates from LAT1 homozygous knockout mice (Fig. 9D), it is still possible that mTORC1 and/or mTORC2 activity is affected locally by LAT1 knockout.
Although the present study provided compelling in vivo evidence that the amount of 4F2hc protein in the plasma membrane of labyrinth of developing placenta is dependent on the expression of LAT1, we observed in BeWo cells that a substantial amount of 4F2hc was still detectable and the forskolin-induced cell fusion remained by 40 to 50% compared to that of untreated cells even though LAT1 expression was markedly decreased by siRNA (Fig. 7C and D). This may be due to the compensative expression of light chain subunits other than LAT1 in BeWo cells that also support the expression of 4F2hc. In the placenta, in addition to LAT1, other members of SLC7 family, LAT2 and y+LAT1, could form heterodimers with 4F2hc as light chain subunits of amino acid transporters (19). The significance of the neutral amino acid transporter LAT2 in the placental function is still unclear since the homozygous Lat2 gene knockout mice show no apparent phenotypes associated with the defects of placental function (45). In contrast, the knockout of y+LAT1, a transporter for both cationic and neutral amino acids, causes IUGR and highly frequent neonatal lethality (16 of 18 neonates died within 24 h after birth) (46). Although the localization of LAT2 and y+LAT1 in the developing placental tissues has not been investigated in detail, the severe impairment of syncytiotrophoblast formation shown in the present study suggests that at least these other light chains of 4F2hc do not substitute for LAT1 in the placental development in vivo.
In summary, we detected a novel fundamental role of the light chain subunit LAT1 to support the protein expression of its heavy chain subunit 4F2hc via a chaperone-like function. We found that LAT1 is essential to form syncytiotrophoblasts, probably by promoting the cell fusion of chorionic trophoblasts in developing placenta.
MATERIALS AND METHODS
Targeted disruption of the Lat1 locus and genotyping.A gene-targeting construct was designed to replace the entire first exon and a part of the first intron of the Lat1 gene with a neomycin resistance cassette (Fig. 1A) and electroporated into TT2 mouse embryonic stem cells. After G418 selection, targeted clones were screened by PCR and further confirmed by Southern blot analysis of EcoRI-digested genomic DNA using two distinct DNA probes corresponding to external and internal portions of targeting construct (Fig. 1B). Chimeras obtained by injecting positive clones into eight cell-stage ICR mouse embryos were crossed with C57BL/6J mice. The agouti-colored offspring with germ line transmission were further crossed with C57BL/6J mice to establish the strain. Genotyping PCR analysis was performed with KOD FX neo DNA polymerase (Toyobo, Osaka, Japan) against genomic DNA extracted from either tail tips of adults or yolk sacs of embryos (Fig. 1C). The nucleotide sequences of the primers were as follows: P1 (common forward, 5′-TGTCTGTGAGTTTTCCTGGGGACTCCTTTG-3′), P2 (wild-type allele reverse, 5′-GTCTGTTAGAGCACGTGGAGCGTTAAGATC-3′), and P3 (targeted allele reverse, 5′-AGAGGTTACGGCAGTTTGTCTCTCCCCCTT-3′).
Observation of the gross morphology of embryos.Embryos (E8.5 to E11.5) were collected and immediately observed under an SZX-12 bright-field stereomicroscope with a color charge-coupled device (CCD) DP70 camera (Olympus, Tokyo, Japan).
Histological analysis and immunostaining.Placenta was fixed in 4% paraformaldehyde–phosphate buffer (PFA/PB; pH 7.4) at 4°C overnight. Frozen sections (10 μm thick) were prepared after sucrose cryoprotection and embedding in OCT compound (Sakura Finetech, Tokyo, Japan). Paraffin sections (5 μm thick) were prepared after ethanol dehydration, paraffinization, and embedding in TissuePrep T580 (Fisher Scientific, Pittsburgh, PA). Hematoxylin-eosin staining was performed on paraffin sections. For immunofluorescence microscopy, paraffin sections or frozen sections were used after antigen retrieval with target retrieval solution (pH 9.0; Dako, Glostrup, Denmark). Triple staining of LAT1, CD34, and connexin 26 was performed with frozen sections without antigen retrieval. Slides were blocked with Blocking One Histo (Nacalai Tesque, Kyoto, Japan) for 20 min, followed by incubation with the indicated primary antibodies overnight at 4°C: anti-LAT1 (E026, 1:200; TransGenic, Fukuoka, Japan), anti-4F2hc (sc-7094, 1:1,000; Santa Cruz Biotechnology, Santa Cruz, CA), anti-CD34 (sc-18917, 1:500; Santa Cruz Biotechnology), anti-connexin 26 (13-8100, 1:500; Invitrogen, Carlsbad, CA), and anti-Ki-67 (ab-15580, 1:1,000; Abcam, Cambridge, MA). After a washing step in Tris-buffered saline, the slides were incubated with the following fluorescently labeled secondary antibodies for 1 h at room temperature: Alexa Fluor 568-conjugated anti-rabbit IgG (A10042, 1:2,000) or Alexa Fluor 488-conjugated anti-rabbit IgG (A11034, 1:2,000) for LAT1, Alexa Fluor 568-conjugated anti-goat IgG (A10057, 1:2,000) for 4F2hc, Cy3-conjugated anti-rat IgG (112-165-143, 1:2,000) or Alexa Fluor 488-conjugated anti-rat IgG (A10042, 1:2,000) for CD34, Alexa Fluor 488-conjugated anti-rabbit IgG for Ki-67 (A11034, 1:2,000), and Cy5-conjugated anti-mouse IgG (715-175-150, 1:2,000) or Alexa Fluor 488-conjugated anti-mouse IgG (A21202, 1:2,000) for connexin 26. Alexa Fluor dye- and Cy dye-conjugated antibodies were purchased from Molecular Probes (Eugene, OR) and Jackson ImmunoResearch (West Grove, PA), respectively. DAPI (4′,6′-diamidino-2-phenylindole; Invitrogen) was added to the secondary antibody solution at a concentration of 2 μg/ml when indicated. Sections were washed in Tris-buffered saline, mounted, and subjected to image acquisition using a bright-field BZ-9000 fluorescence microscope (Keyence, Osaka, Japan) or the confocal laser scanning microscope systems LSM510 META or LSM710 (Zeiss, Oberkochen, Germany). For the determination of Ki-67 labeling indices, three (for E9.5) and five (for E10.5) different sections for each genotype were analyzed by using the Cell Counter plug in for ImageJ software (National Institutes of Health [NIH], Bethesda, MD).
For the immunostaining of LAT1 and 4F2hc in BeWo cells, the cells were grown on collagen I-coated coverslips and fixed with methanol for 10 min at −20°C. After a blocking step in phosphate-buffered saline (PBS) containing 0.1% Triton X-100 and 1% bovine serum albumin for 10 min, the cells were incubated with anti-LAT1 antibody (KE026, 1:200; TransGenic) and anti-4F2hc antibody (sc-7094, 1:200; Santa Cruz Biotechnology) for 3 h at room temperature. After being washed with PBS, the cells were further incubated with secondary antibodies (Alexa Fluor 568-conjugated anti-rabbit IgG [A10042, 1:2,000; Molecular Probes] and Alexa Fluor 488-conjugated anti-goat IgG [A11055, 1:1,000; Molecular Probes]) for 1 h. The cells on coverslips were then washed with PBS, mounted, and observed with a bright-field BZ-9000 fluorescence microscope.
Rhodamine 123 permeation assay.Rhodamine 123 permeation assay was performed as previously described with minor modifications (47). On E10.5, pregnant female mice were intraperitoneally injected with rhodamine 123 (Wako Chemicals, Osaka, Japan) in PBS (1 mg/kg [body weight]). Embryos were isolated 30 min after injection, briefly fixed in 4% PFA/PB at 4°C, and then observed under an AF6000 stereofluorescence microscope configured with an M205FA and DFC365 FX camera (Leica, Wetzlar, Germany).
In situ hybridization. In situ hybridization was performed as described previously (48). A fragment corresponding to nucleotide positions 1620 to 2132 of mouse LAT1 mRNA (NCBI RefSeq NM_011404.3 ) was amplified by PCR and subcloned into the pCR4-TOPO cloning vector (Invitrogen). After linearization with SpeI digestion, a digoxigenin (DIG)-labeled antisense probe was synthesized in vitro using DIG RNA labeling mix (Roche Diagnostics, Indianapolis, IN) and T7 RNA polymerase (Stratagene, La Jolla, CA). Bright-field images were acquired using a BX63 microscope (Olympus) equipped with a color CCD DP80 camera.
Western blot analysis.Placentas were separated from the decidua, homogenized on ice with a Physcotron NS-310E II (Microtec, Chiba, Japan) in lysis buffer (50 mM Tris-HCl [pH 7.4], 150 mM NaCl, 1% NP-40, 1 mM phenylmethylsulfonyl fluoride, and a protease inhibitor cocktail [complete EDTA-free; Roche Diagnostics]). For the detection of phosphorylated proteins, 20 mM NaF, 1 mM EDTA, 6 mM EGTA, 20 mM β-glycerophosphate, and 1 mM Na3VO4 were added to the lysis buffer. After solubilization on ice for 30 min, followed by centrifugation at 21,000 × g for 10 min at 4°C, the supernatants were collected, mixed with Laemmli buffer, and subjected to SDS-PAGE. For the detection of LAT1, sample boiling before SDS-PAGE was omitted to avoid aggregation. To investigate the heterodimerization of LAT1 and 4F2hc, a crude membrane fraction was prepared from the placentas. The placentas were homogenized on ice in lysis buffer containing 10% glycerol. After centrifugation at 1,000 × g for 5 min at 4°C, the samples were further cleared by centrifugation at 21,000 × g for 10 min at 4°C. The supernatants were ultracentrifuged at 436,000 × g for 1 h at 4°C (Beckman Optima TLX with a TLA100.1 rotor). The obtained crude membrane pellet was then solubilized using lysis buffer containing 1% NP-40 and 1 mM N-ethylmaleimide on ice for 30 min and subjected to SDS-PAGE. Dithiothreitol (DTT) was omitted from the Laemmli buffer for the nonreducing condition.
Western blot analysis was performed as described previously (49). For reprobing, the membranes were stripped of bound antibodies by incubation in 62.5 mM Tris-HCl (pH 6.8) containing 2% SDS and 100 mM β-mercaptoethanol for 45 min at 55°C. The primary antibodies used were as follows: anti-LAT1 (KE026, 1:1,000; TransGenic), anti-4F2hc (sc-31251 at 1:500 [Santa Cruz Biotechnology] for mouse samples and sc-9160 at 1:2,000 for human samples), anti-Na+/K+-ATPase α1 (sc-21712, 1:2,000; Santa Cruz Biotechnology), anti-β-actin (A5441, 1:5,000; Sigma-Aldrich), anti-p70S6K (sc-230, 1:5,000; Santa Cruz Biotechnology), anti-phospho-Thr389-p70S6K (9243, 1:1,000; Cell Signaling Technology, Danvers, MA), anti-eukaryotic translation initiation factor 4E-binding protein 1 (4EBP1; 9452, 1:1,000; Cell Signaling Technology), anti-phospho-Thr37/46-4EBP1 (2855, 1:1,000; Cell Signaling Technology), anti-eukaryotic initiation factor 2α subunit (anti-eIF2α; 5324, 1:100,000; Cell Signaling Technology), anti-phospho-Ser51-eIF2α (3398, 1:1,000; Cell Signaling Technology), and anti-microtubule-associated protein 1 light chain 3 (anti-LC3; 3868, 1:2,000; Cell Signaling Technology). Densitometric analysis was performed using ImageJ software (NIH).
Quantitative real-time PCR.Total RNA was extracted from mouse placentas after removal of the decidua or from BeWo cells using Isogen II (Nippon Gene, Tokyo, Japan) and used for cDNA synthesis (PrimeScript RT master mix; TaKaRa, Shiga, Japan). Quantitative PCR was performed with SYBR Premix Ex Taq Tli RNase H Plus (TaKaRa) using a 7900HT Fast Real-Time PCR system (Applied Biosystems, Foster City, CA). The obtained data were normalized with respect to the expression of glyceraldehyde 3-phosphate dehydrogenase (GAPDH) mRNA. The following primers were used: mouse LAT1 (forward, 5′-CGGGCTGCCTGTCTACTTC-3′; reverse, 5′-CAGAGCACCGTCACAGAGAA-3′), mouse 4F2hc (forward, 5′-ATGGTGCAGCTGGAGTGTG-3′; reverse, 5′-CCCCGTAGCTAAAAACAGGA-3′), mouse GAPDH (forward, 5′-AGCTTGTCATCAACGGGAAG-3′; reverse, 5′-TTTGATGTTAGTGGGGTCTCG-3′), human LAT1 (forward, 5′-CCGTGAACTGCTACAGCGT-3′; reverse, 5′-CTTCCCGATCTGGACGAAGC-3′), human 4F2hc (forward, 5′-GCTTCCAAGAATGCTGAGGTT-3′; reverse, 5′-GCAGTCGGATCAATTGAGGT-3′), and human GAPDH (forward, 5′-GAGTCCACTGGCGTCTTCAC-3′; reverse, 5′-TTCACACCCATGACGAACAT-3′). Specific amplification of the targeted sequences was confirmed by referring to the dissociation curve and agarose gel electrophoresis of the PCR products.
In vitro cell fusion assay using human trophoblastic BeWo cells.Human trophoblastic BeWo cells (catalog no. 9111; JCRB, Osaka, Japan) were maintained in Ham F-12 medium supplemented with 15% fetal calf serum in a humidified incubator with 5% CO2. Cells were transfected with either pAcGFP-N1 or pmCherry-N1 (Clontech, Mountain View, CA) using Lipofectamine 2000 (Invitrogen). The stable transfectants were then selected by G418 (0.8 mg/ml). The obtained two stable cell lines were cocultured in collagen I-coated six-well plates (104 cells each/well) for 24 h and transfected with siRNA using Lipofectamine RNAi MAX (Invitrogen). Silencer Select siRNA for LAT1#1 (catalog no. s15653), LAT1#2 (catalog no. s15654), and LAT1#3 (catalog no. s15655), as well as Silencer Negative Control siRNA#1 and siRNA#2, were purchased from Ambion (Austin, TX). On the next day, either 10 μM forskolin or dimethyl sulfoxide (DMSO [0.01% solvent for forskolin]) was added to the medium, followed by further incubation for 48 h. The cells were fixed in 4% PFA/PB (pH 7.4) for 30 min at room temperature and stained with DAPI (0.2 μg/ml) for 30 min. Fluorescent images were acquired using a bright-field BZ-9000 fluorescence microscope. To quantify the cell fusion, seven randomly selected fields for each well were analyzed to calculate the percentages of nuclei located in cells positive for both AcGFP and mCherry fluorescence. The knockdown efficiency was confirmed by Western blotting of cells collected at 72 h after siRNA transfection.
In vivo BrdU incorporation assay and in situ apoptosis staining.Pregnant female mice were intraperitoneally injected with bromodeoxyuridine (BrdU) labeling reagent (Invitrogen) at 10 μl/g (body weight). The placenta was isolated 2 h after injection, and paraffin sections were prepared as described above. Incorporated BrdU was detected using a BrdU staining kit (Invitrogen). Apoptotic cells were visualized in paraffin sections of the placenta by TUNEL using an in situ apoptosis detection kit (TaKaRa).
Ethics statement.All animal experiments were conducted in compliance with the protocols approved by the Osaka University Medical School Animal Care and Use Committee.
ACKNOWLEDGMENTS
We thank Miyuki Kurauchi, Yuka Miyoshi, Satori Matsumoto, Kaori Nakano, Yoshie Hayamizu, and Yui Kurokawa for technical assistance. We also thank Yuewi Li for helping with preliminary experiments at the initial stage and Masanori Uchikawa (Osaka University, Osaka, Japan) and Naoto Hayasaka (Yamaguchi University, Yamaguchi, Japan) for generous technical advice.
R. Ohgaki, T. Ohmori, S. Hara, S. Nakagomi, and M. Kanai-Azuma performed experiments and analyzed data. R. Ohgaki and Y. Kanai designed the experiments and wrote the manuscript with the participation of M. Kanai-Azuma, K. Kaneda-Nakashima, S. Okuda, and S. Nagamori. Y. Kanai conceived the project.
This study was supported by Grants-in-Aid for Scientific Research on Priority Areas (17081016 to Y. Kanai) and by the Regional Innovation Strategy Support Program to Y. Kanai from the Ministry of Education, Culture, Sports, Science, and Technology of Japan; Grants-in-Aid for Scientific Research (15H04685 and 22659052 to Y. Kanai) from the Japan Society for the Promotion of Science; and the Advanced Research for Medical Products Mining Program of the National Institute of Biomedical Innovation (12-02 to Y. Kanai).
FOOTNOTES
- Received 21 July 2016.
- Returned for modification 25 August 2016.
- Accepted 9 March 2017.
- Accepted manuscript posted online 20 March 2017.
- Copyright © 2017 American Society for Microbiology.
