Lipid rafts enriched in monosialylGb5Cer carrying the stage-specific embryonic antigen-4 epitope are involved in development of mouse preimplantation embryos at cleavage stage

Abstract

Background: Lipid rafts enriched in glycosphingolipids (GSLs), cholesterol and signaling molecules play an essential role not only for signal transduction started by ligand binding, but for intracellular events such as organization of actin, intracellular traffic and cell polarity, but their functions in cleavage division of preimplantation embryos are not well known.

Results: Here we show that monosialylGb5Cer (MSGb5Cer)-enriched raft domains are involved in development during the cleavage stage of mouse preimplantation embryos. MSGb5Cer preferentially localizes at the interfaces between blastomeres in mouse preimplantation embryos. Live-imaging analysis revealed that MSGb5Cer localizes in cleavage furrows during cytokinesis, and that by accumulating at the interfaces, it thickens them. Depletion of cholesterol from the cell membrane with methyl-beta-cyclodextrin (MbCD) reduced the expression of MSGb5Cer and stopped cleavage. Extensive accumulation of MSGb5Cer at the interfaces by cross-linking with anti-MSGb5Cer Mab (6E2) caused F-actin to aggregate at the interfaces and suppressed the localization of E-cadherin at the interfaces, which resulted in the cessation of cleavage. In addition, suppression of actin polymerization with cytochalasin D (CCD) decreased the accumulation of MSGb5Cer at the interfaces. In E-cadherin-targeted embryos, the MSGb5Cer-enriched raft membrane domains accumulated heterotopically.

Conclusions: These results indicate that MSGb5Cer-enriched raft membrane domains participate in cytokinesis in a close cooperation with the cortical actin network and the distribution of E-cadherin.

Figure 1

MSGb5Cer localize at the interfaces between blastomeres in mouse preimplantation embryos. (A) Localization of MSGb5Cer, E-cadherin and GM1 in an unfertilized egg and an embryo at the zygote, 2-cell stage and 8-cell stage. MSGb5Cer (top row), E-cadherin (middle row), and GM1 (bottom row) in a living mouse egg or embryos were stained with Alexa Fluor^® 488-conjugated 6E2, ECCD-2/Alexa Fluor^® 546 Rabbit Anti-rat IgG, and biotinylated-CTX B/Alexa Fluor^® 546 streptavidin, respectively, and examined. Thirty sample unfertilized egg and 45 sample embryos of other stages were used for this experiment represented in image. (B) Distribution of BODIPY^®FL-MSGb5Cer and BODIPY^®FL-GM1 in a 2-cell stage embryo. The distribution of MSGb5Cer and GM1 is shown in the form of optical slice images. BODIPY^®FL-C12 fatty acid was used as a negative control. All scale bars represent 20 μm. Ten sample 2-cell stage embryos were used for this experiment represented in each image. (C) Distribution of BODIPY^®FL-MSGb5Cer in a 2-cell stage embryo cultured in the presence of 6E2. The BODIPY^®FL-MSGb5Cer introduced 2-cell embryo was cultured for 1 hour in M16 medium containing 15 μg/ml of 6E2, and then stained with Alexa Fluor^® 546-conjugated 6E2. Scale bar represent 20 μm. Ten sample 2-cell stage embryos were used for this experiment represented in image.

Figure 2

MSGb5Cer accumulates at the interface during cytokinesis in preimplantation embryos. (A) Frames from a series of confocal time-lapse movies of live cytokinesis-stage cells stained with 6E2 for MSGb5cer (green) and with DAPI for nuclei (red). Frames are shown at 60-minute intervals from 0 to 120 minutes, 15-minute intervals from 120 minutes to 240 minutes, and 60-minute intervals from 240 minutes to 540 minutes. Time is expressed in minutes after the start of culture. The arrow at 150 minutes points to the site of a future furrow in the spherical zygote. Scale bar represents 20 μm. The images shown are representative of nine sample zygotes. (B) Temporal changes in fluorescence intensity at the furrow and the interface region during cytokinesis. The fluorescence intensity in pixel values was measured in boxes of equal size located at the furrow (red square) and the interface (black square) of the frames taken every 15 minutes from a confocal time-lapse movie of the zygote of Figure 2A. The fluorescence intensity values at the furrows (closed circles) and interface (open circles) were plotted against time in minutes. Schematic embryos at the round stage (1), during cytokinesis (2), after cytokinesis (3), and the late 2-cell stage (4) were cultured for 105 minutes, 165 minutes, 240 minutes, and 405 minutes, respectively. This data shown is representative of nine sample zygotes. (C) The fluorescence intensity of MSGb5Cer at the furrow (closed column) and interface (dotted column) in round, cytokinesis, post-cytokinesis and late 2-cell stage embryos. Mean values and s.d. were calculated from the results obtained from the time-lapse movies of 6 embryos shown in Figure 2B.

Figure 3

MSGb5Cer is overlapped by cholesterol at the interface in a 2-cell embryo. A two-cell stage embryo was stained with 6E2 for MSGb5cer (green) and then filipin III for cholesterol (blue) as described in Materials and Methods. Scale bar: 20 μm. Ten sample 2-cell stage embryos were used for this experiment represented in images.

Figure 4

Pretreatment with MbCD inhibits cytokinesis and development of mouse preimplantation embryos. (A) Effect of MbCD on the expression of MSGb5Cer (green), GM1 (red) and E-cadherin (red) on 8-cell embryos. The left column and the right column of each panel represent a 2D image and the overlay of a 2D image and a differential interference contrast (DIC) micrograph, respectively. Thirty 8-cell embryos were used for the experiment represented in the images. (B) Representative DIC images of embryos cultured for 0, 24 and 48 hours in M16 medium after preincubation with 0-5 mM MbCD. Scale bars: 100 μm. Twenty 8-cell embryos were used for the experiment represented each images. (C) The percentages of living embryos cultured for 12, 24 and 48 hours after preincubation with 0-5 mM MbCD. Twenty 8-cell embryos were cultured for each concentration of MbCD.

Figure 5

Extensive cross-linking of MSGb5Cer with 6E2 inhibits cytokinesis and development of mouse preimplantation embryos. (A) Deposition of BODIPY^®FL-MSGb5Cer at the interfaces of 2-cell stage embryos cultured in the presence of 6E2. Embryos labeled with BODIPY^®FL-MSGb5Cer (green) were cultured for 1 hour in M16 medium containing 100 μg/ml of 6E2 or 15B2 antibodies. A fluorescence 2D image of BODIPY^®FL-MSGb5Cer and the overlay of a 2D image and DIC micrograph of cultured embryos are shown. These images shown are representative of twelve 2-cell embryos. (B) Effect of cross-linking of MSGb5Cer with 6E2 on cytokinesis and the development of mouse preimplantation embryos. Two-cell embryos were cultured for 60 hours in M16 medium containing 100 μg/ml of 15B2 (red squares), ECCD-2 (blue diamonds), and 6E2 (green triangles), and the living embryos were counted and their DIC images were taken at the times indicated. Arrows indicate compacted embryos. Twenty of 2-cell stage embryos were cultured for each experiment groups. (C) Localization of F-actin and E-cadherin in MSGb5Cer-cross-linked 2-cell embryos. Two-cell embryos treated for 2 hours with 6E2 (left column) or 15B2 (right column) were fixed and stained with Alexa Fluor^® 633 phalloidin and anti E-cadherin/Alexa Fluor^® 546-conjugated rabbit anti-rat IgG. These images shown are representative of twelve 2-cell embryos. (D) Inhibitory effect of cross-linking of MSGb5Cer with 6E2 on the localization of E-cadherin at interfaces. Embryos were immunostained with ECCD-2/Alexa Fluor^® 546-conjugated Rabbit Anti-rat IgG (red) in the presence of various concentrations of 6E2. The upper row and the bottom row of each panel represent a 2D image and the overlay of a 2D image and DIC micrograph, respectively. These images shown are representative of thirty morula embryos.

Figure 6

Disruption of actin filaments suppresses localization of MSGb5Cer at the interface. Two-cell embryos pretreated with 0, 0.2, and 2 μg/ml of CCD were immunostained with Alexa Fluor^® 488-conjugated 6E2 (green). The 3D images were constructed by stacking optical slice images of the area enclosed by the square. Scale bar: 20 μm (2D), 10 μm (3D). These images shown are representative of thirty 2-cell stage embryos.

Figure 7

Absence of E-cadherin causes heterotopic localization of MSGb5Cer on the blastomere surface. Embryos expressing E-cadherin (Genotype; Floxed/+) and lacking E-cadherin (Genotype; Floxed^del/+) were dually immunostained with ECCD-2/Alexa Fluor^® 546-conjugated Rabbit Anti-rat IgG and Alexa Fluor^® 488-conjugated 6E2. 2D images of E-cadherin (red), MSGb5Cer (green) and nuclei (blue) are merged in the far right column. We used thirty maternal knockout embryos and seventeen control embryos in this experiment. MsGb5Cer represents heterotopic localization in all maternal knockout embryos.