Sponge localizes to actin caps, metaphase furrows and subapical domains of cellularization furrows
The ELMO-Sponge complex as an unconventional Rap1 GEF provided a reasonable explanation for localization of Canoe and its downstream factors.
As I could already show the functional relevance of ELMO-Sponge I wanted to analyze if ELMO and Sponge also showed a subapical localization during cellularization. To analyze this aspect, I stained fixed wild type embryos for Sponge (Figure 17) as its localization during early embryogenesis was not described yet. Figure 17 shows Sponge staining during different nuclear cycles in sagittal and top views. During interphases of syncytial blastoderm embryos,
Figure 16 ELMO and Sponge are required for subapical restriction of Canoe.
(A-C) Images of fixed (A) wild type, (B) ELMO and (C) sponge embryos stained for Canoe (grey/green), Dlg (grey/red) and DNA (blue). Merged images are shown in right panel with a furrow in higher magnification in insets. (D) Heat maps and averaged values of relative fluorescence intensity along apical-basal axis aligned to the peak value measured for 9 furrows in 3 embryos. Error bars represent SEM. Scale bars 10 µm, insets 2 µm.
Sponge was localized at the actin caps but not at intercap regions (Figure 17, interphase 11).
During mitosis, Sponge localized to the lateral domain of the metaphase furrow and was also found at the apical domain (Figure 17, mitosis 13). Also, at onset of cellularization in interphase 14, before nuclear elongation and with still very short furrows, Sponge was found at the actin caps with some
Figure 17 Sponge localization during early embryonic development.
Images of fixed wild type embryos at indicated stages stained for Sponge (grey/red) and DNA (blue). Sagittal and planar views. Scale bar 10 µm.
enrichment at the rims of the caps, as the top view shows (Figure 17, interphase 14). With start of nuclear elongation, which still resembles early cellularization, Sponge reorganized and was enriched at the newly introduced subapical domain, which is shown in the side view (Figure 17, interphase 14).
The top views of interphase 14 show a more ring like distribution of Sponge compared to its localization in interphase 11, during which the whole actin cap was stained resembling a disc-like localization. From the localization of Sponge, it is possible that it could activate Rap1 locally at the subapical domain at onset of cellularization.
Sponge localization depends on ELMO but not on Rap1, Canoe and Scribble
As described before, Sponge and ELMO build a complex in which ELMO is responsible for membrane binding (Côté et al., 2005). To test if ELMO is necessary for Sponge localization at the membrane in vivo, I stained fixed wild type and ELMO mutant embryos against Sponge and Dlg (Figure 18).
Compared to the subapical Sponge enrichment in wild type embryos during early cellularization (Figure 18A), membrane localization of Sponge was indeed completely absent in ELMO mutant embryos (Figure 18B) showing, that ELMO is needed for membrane targeting of Sponge.
Furthermore, I tested if Sponge localization is affected by its downstream factors Rap1 and Canoe or by the lateral factor Scribble by staining of Sponge in the associated mutant situations (Figure 19). In Rap1 mutant embryos, Sponge enrichment was detected in the subapical region and did not spread into the lateral domain (Figure 19B), although it is possible that Sponge was
Figure 18 Membrane localization of Sponge is mediated by ELMO.
Images of fixed (A) wild type and (B) ELMO embryos in early cellularization stained for Sponge (grey/green), Dlg (grey/red) and DNA (blue). Merged images are shown in red panel with a furrow in higher magnification in inset. Scale bar 10 µm, insets 2 µm.
not well restricted from the apical domain compared to wild type (Figure 19A).
Also, in canoe mutants, Sponge was subapically enriched and did not spread into the lateral domain (Figure 19D). This was also the case for scribble mutants, in which Sponge enrichment was visible at the subapical domain and Sponge did not spread into the lateral domain (Figure 19C). Taken together, the subapical enrichment of Sponge was dependent on its interactor ELMO in vivo, but not on its downstream factors Rap1 and Canoe as well as on the lateral domain protein Scribble.
ELMO shows a ring-like localization at onset of cellularization marking the position of the newly forming subapical domain
To follow the dynamics of ELMO-Sponge in vivo, I made use of a transgenic line expressing GFP-tagged ELMO under the control of its own promoter. The transgene was made by Dr. Zhiyi Lv. Furthermore, I used CherrySlam as a marker for the basal domain. Imaging of living embryos expressing ELMO-GFP and CherrySlam was done from the top view with z-stacks with a step size of 0.5 µm and an interval of 1 min during end of mitosis 13 and beginning of interphase 14 and onset of cellularization (Figure 20). Timepoint zero was
Figure 19 Genetic control of subapical Sponge.
Images of fixed embryos in early cellularization stained for Sponge (grey/green), Dlg (grey/red) and DNA (blue). Merged images are shown in right panel, inserts show zoom in of one furrow. Genotypes (A) wild type, embryos from germline clones for (B) Rap1, (C) scribble, (D) canoe. Scale bar 10 µm, insets 2 µm.
defined by the formation of new cellularization furrows. As shown in Figure 20, ELMO-GFP localized to metaphase furrows in metaphase 13.
Figure 20 ELMO-GFP and CherrySlam dynamics during mitosis 13 and interphase 14.
Images of a time lapse recording of an embryo expressing Elmo-GFP (green) and CherrySlam (red). Time from left to right, apical basal position from up to down. Scale bar 10 µm.
With onset of cellularization, ELMO-GFP localized to caps and showed an enrichment at the rims of the cap to form a ring-like structure, that was stable at least during the first 8 minutes of cellularization when furrows elongated, shown by basal CherrySlam signal, as indicated. Figure 21A shows ELMO-GFP signal from an actin cap during an earlier interphase, that had no enrichment at the rim of the caps, forming a more disc-like localization.
To analyze the dynamics of ELMO-GFP in more detail I selected significant z-positions along the apical-basal axis and timepoints at onset of cellularization, which are depicted in Figure 21C, with a schematic overview of the sagittal view in Figure 21B. During interphases in syncytial blastoderm, ELMO-GFP localized to the whole cap domain in a disc like pattern (Figure 21A). At onset of interphase 14 (Figure 21C, 1 min), ELMO-GFP got enriched at the rims of the actin caps at a z-position of around 0.5-1.5 µm, as also described before for Sponge in fixed embryos. The new furrow was formed at the position where two rings came together as indicated by a yellow arrowhead. Similar to the
Figure 21 Dynamics of ELMO-GFP during early cellularization.
(A) Image from a time laps recording of an embryo expressing ELMO-GFP shows ELMO-GFP localization at the cap during interphase 13. (B) Scheme for furrow formation and invagination in early cellularization. Subapical and basal domains are marked in green and red, respectively. Axial (apical-basal) axis with approximate scale is indicated. (C) Images from time lapse recordings including axial stacks of embryo expressing ELMO-GFP (grey/green) and CherrySlam (grey/red) during mitosis 13 and early interphase 14. Axial position is indicated. Yellow arrowhead points to position of “new” furrows. (D) Relative fluorescence intensity of ELMO-GFP (green) and CherrySlam (red) at “new” furrows measured along the apical-basal axis at indicated times. Error bars represent SEM. Scale bar 10 µm.
quantifications I showed for CanoeYFP and ScribbleGFP, I measured the fluorescence intensity of ELMO-GFP and CherrySlam along the apical-basal axis of three new furrows in one embryo, normalized the data to their peaks and plotted the averages of the relative fluorescence intensity against the apical-basal position (Figure 21D). After 2 minutes, when the new furrow was still a shallow groove, revealed by the missing peak of CherrySlam intensity (dashed red line), ELMO-GFP already showed a subapical enrichment which was diminished in basal direction (dashed green line). After the formation and invagination of the new furrow at 6 min, indicated by a peak of CherrySlam intensity at around 4.5-5.5 µm (red line), ELMO-GFP intensity was still subapically enriched and decreased in direction of the basal domain (green line).
Taken together, I could show that the unconventional GEF ELMO-Sponge is a likely candidate to activate Rap1 locally to position Canoe and its downstream factors. The change of ELMO-Sponge localization from a more disc-like distribution during blastoderm interphases to the ring-like enrichment during onset of cellularization may provide an explanation for the switch from three cortical domains in metaphase and two cortical domains in interphase during syncytial blastoderm to four cortical domains in cellularization by the insertion of the new subapical domain.