Figure 1.
Diacylglycerol localises to the NE, Golgi and ER in mammalian cells.
HeLa (A) and COS-7 (B) cells were transfected with EGFP-PKCεC1aC1b (DAG probe-green), fixed and imaged by confocal microscopy. DAG was localised at the NE (yellow arrows), ER (green arrow) and Golgi (white arrows). (C–D) Calreticulin (ER marker) was detected by indirect immunofluorescence (red). Apart from a minor detection of the Golgi, the signal at the ER and NE (insets) was absent in cells transfected with the DAG non-binding mutant (C1b W264G) (D). HeLa (E-H) and COS-7 (I) cells were followed through mitosis by live confocal microscopy. EGFP-PKCεC1aC1b in HeLa (F) and COS-7 (I) presented similar distributions as DiOC6 (E), GFP-POM121 (G), and ER tracker (H). ER tubules (green arrows) and NE reformation (yellow arrows) were observed. To label chromatin, cells were incubated with Hoechst 333432 or transfected with mCherry-H2B. Scale bars: 10 μm.
Figure 2.
Acute depletion of DAG results in an incomplete NE.
(A) Diagram of the rapalogue dimerisation device. After rapalogue (R) treatment, RFP-Flag-FRB-DGKεK (DGKε) dimerises with EGFP-2FKBP-LBRΔTM2-8 (LBR) and is recruited to LBR in the ER and NE. (B) Confocal images of live interphase HeLa cells transfected with LBR (green) and DGKε (red) show the recruitment of DGKε (inset) to the NE and ER, 45 min after addition of 500 nM rapalogue. (C) HeLa cells transfected with EGFP-2FKBP-LBRΔTM2-8 (LBR) only and treated with rapalogue showed a normal NE reformation (yellow arrows) between late telophase (left panel) and cytokinesis (right panel), similarly to what was observed in (D) LBR and DGKε co-expressing HeLa cells in the absence of rapalogue. In HeLa (E) and COS-7 (F) cells treated with rapalogue NE reformation was impaired. Images representative of n = 10 experiments. (G) When DGKε was replaced by its inactive mutant (D434N) the NE formation was normal (yellow arrows). Images representative of n = 6 experiments. Scale bars: 10 μm.
Figure 3.
Acute depletion of PtdIns(4,5)P2 results in an incomplete NE: DAG is not observed upon malformation of the NE.
(A) Diagram of the rapalogue dimerisation device. After rapalogue (R) treatment, RFP-Flag-FRB-SKIP (SKIP) dimerises with EGFP-2FKBP-LBRΔTM2-8 (LBR) and is recruited to LBR in the ER and NE. (B) Confocal images of live interphase HeLa cells transfected with LBR (green) and SKIP (red) show the recruitment of SKIP to the NE and ER, 45min after addition of 500 nM rapalogue. (C) LBR localisation during mitosis in LBR and SKIP co-expressing HeLa cells in the absence of rapalogue shows complete NE reformation (yellow arrows) between telophase (left panel) and cytokinesis (right panel). (D) In HeLa cells treated with rapalogue NE reformation was impaired. (E–F) When SKIP was replaced by its inactive mutant (D310G) the NE formation was normal (yellow arrows), in the absence (E) or presence (F) of rapalogue. (G-H) EGFP-C1a-C1b (DAG probe-green) localisation during mitosis in dark (EGFP) LBR and SKIP co-expressing cells. In the absence of rapalogue (G), NE formation was normal (yellow arrows) and ER tubules were visible (green arrow), contrary to what was observed in the presence of rapalogue (H). Images representative of n = 3 experiments. Scale bars: 10 μm.
Table 1.
Quantification of NE phenotype in single cell experiments of HeLa cells co-expressing LBR and DGKε.
Figure 4.
NE reformation is disrupted in a dose-dependent manner in DAG-depleted mitotic cells.
(A–B) HeLa cells labelled with DiOC6 (green) were followed through mitosis by confocal microscopy, fixed at early anaphase (A) or telophase (B) and prepared for high-resolution imaging using CLEM (Fig. S4). The segmentation showed that at early anaphase the NE (red) was incomplete with wide gaps of 4 to 5 μm (purple arrows), at telophase the NE was close to completion with gaps of 50 nm. Segmentation of the ER (blue) at anaphase showed that it was mainly tubular. Note that the NE of the early anaphase cell was segmented around both sets of chromosomes. Dashed white line indicates axis of symmetry. Orange arrow highlights centriole (only one in this section). Serial sections are shown in Movie S4. (C-D) Same experiment with rapalogue-treated HeLa cells expressing LBR (green) and low (C) or high (D) levels of DGKε (red) fixed at early telophase (C) and cytokinesis (D). Dose-dependent effects upon DAG depletion included large gaps in the NE (purple arrows) and aggregation of the ER (blue). The ER phenotype consisted of large multi-lamellar sheets of membrane (insets-green arrows) with minimal NE contact (inset-yellow arrow). Movies of serial sections are shown in Movies S5–6. (E-F) 3D models reconstructed from manually-segmented CLEM serial images of control (E) and DAG-depleted (F) HeLa cells (Movie S7). In control cells, the NE at anaphase (red) was incomplete, while virtually complete at telophase. NE of the early anaphase cell was segmented around both sets of chromosomes. In DAG-depleted cells, the NE was not formed. Centrioles shown in yellow. Scale bars: fluorescence 10 μm; CLEM as indicated on the images.
Figure 5.
1,2- and 1,3-DAG rescue the fragmented NE phenotype.
(A) Confocal images of live HeLa cells 1 min after addition of small unilamellar vesicles (SUVs) containing BODIPY-PtdCho and unsaturated 1,2 DAG (80∶20 mole% respectively). Incorporation of SUVs (green) into interphase and metaphase (white arrows) cells are shown. (B) LBR localisation during mitosis in rapalogue-treated, LBR and DGKε-expressing HeLa cells after addition at metaphase of SUVs containing PtdCho and unsaturated 1,2 DAG (80∶20 mole%). NE reformation (yellow arrows) was rescued. (C) Ultrastructure of the NE (yellow arrow) of the same cell at cytokinesis imaged using CLEM (Fig. S4). LBR localisation in green, DGKε in red. Serial sections are shown in Movie S8. (D) Comparison of 3D models reconstructed from serial images of DAG-depleted (left panel) and DAG-rescued (right panel) cells shows the NE reformed in the presence of 1,2 DAG. (E–G) Similar results as in (A–C) respectively were obtained with SUVs of the non-C1 domain-binding DAG isomer 1,3 DAG. (H) CLEM images of a rapalogue-treated, dark LBR and DGKε-expressing HeLa cell fixed at cytokinesis, after addition of (60∶40 mole %) SUVs with BODIPY-PtdCho and unsaturated 1,3 DAG. Incorporation of the SUVs into cell membranes in green, DGKε in red. EM images show 1,3 DAG completely rescued NE reformation. Images representative of n = 11 experiments. Scale bars: confocal 10 μm, CLEM as indicated on the images.
Figure 6.
Effects of DGK and Synaptojanin1 microinjection on sea urchin embryos and eggs.
(A) Fertilised eggs. The ER was labelled by microinjection of DiIC18 into sea urchin eggs between 10–25 min post fertilisation. Karyomeres in the first cell cycle resolve into individual nuclei between 72 and 96 min post-fertilisation and in the second cycle (4-cell stage) between 116 and 123 min. (B) At 5 μg/ml pipette concentration of DGK, curved stacked sheets of ER formed by 144 min that coalesced into more concentrated “aggregates”. (C) At 100 μg/ml of DGK, either chromosomes condensed but the NE did not break down or an apparently normal metaphase occurred but karyomere fusion (D) was greatly retarded compared to the first cycle controls. (E) Synaptojanin 1 (Syn1) induced a similar phenotype of karyomere resolution delay to DGK as well as formation of curved stacked sheets of ER that coalesced into “aggregates” of sheets (0.17 μg/ml and 1.7 μg/ml shown). Unfertilised haploid eggs (F) were injected with either 50 μg/ml DGK (G) or 1.7 μg/ml Syn1 (H) which rapidly converted the ER in a progressive coarsening of the pattern similar to fertilised eggs already detectable by 13 or 17 minutes post-injection (Movie S9). The upper dark circle in (G) images is the injected oil droplet; lower green circle (arrow) is the zygote nucleus with NE that does not undergo breakdown. (I) Unfertilised eggs were incubated with SUVs containing PtdCho and 1,3 DAG (80∶20 mole%) prior to Syn 1 injection (17 μg/ml) at T0. Their appearance did not change for more than 74 min (Movie S10). All embryos were injected with enzymes ∼40 min post-fertilisation. YOYO®-1 iodide was included to label nucleic acid green and monitor injection. Scale bars: 10 μm.