GCL pruning of PIP3 establishes the soma-germline boundary

  1. Mariyah Saiduddin
  2. Juhee Pae
  3. Asier M. Vidal
  4. Marty Alani
  5. Ruth Lehmann

Reviewed by

Read the full article See related articles

Discuss this preprint

Start a discussion What are Sciety discussions?

Listed in

Log in to save this article

Abstract

Primordial germ cells (PGCs) are the first cells specified in the Drosophila embryo and serve as precursors to the germline. Their formation requires suppression of somatic fates, a process achieved by excluding the receptor tyrosine kinase Torso from the posterior pole through degradation mediated by the ubiquitin ligase adaptor Germ Cell-Less (GCL). Although Torso is known to antagonize PGC formation, the underlying mechanism has remained unclear. Here, we combine optogenetic Ras activation and Ras effector loop mutants to show that Ras signaling suppresses PGC formation independently of the canonical Raf/MEK/ERK pathway. We identify an unexpected early role for Torso in activating phosphoinositide 3-kinase (PI3K), generating posterior membrane domains enriched in phosphatidylinositol (3,4,5)-trisphosphate (PIP3). Elevated PI3K activity disrupts PGC formation, while reduced PI3K activity leads to ectopic PGCs. We further demonstrate that GCL remodels the posterior pole membrane by suppressing Torso-dependent PI3K activation. Clearing PIP3 enables Myosin II enrichment, thereby constricting the pole bud for PGC formation. Together, our findings reveal how antagonistic Torso and GCL activities establish the soma–germline boundary by regulating cortical lipid organization.

Graphical Abstract

  • Membrane domains with high Torso activity (purple) in the early embryo have a higher PIP3 content.

  • At the posterior pole, GCL-containing germplasm (green) degrades Torso, resulting in a PIP3-low membrane.

  • Clearing PIP3 enables Myosin II pole bud constriction required for PGC formation.

Article activity feed

  1. Review Commons

    Note: This response was posted by the corresponding author to Review Commons. The content has not been altered except for formatting.

    Learn more at Review Commons

    Reply to the reviewers

    Reviewer #1

    Figure 1D: It would be useful to indicate the number of embryos analyzed for these experiments (n = ?).

    Number of embryos now included in figure legend

    Figure 3B: The control condition for gcl⁻/⁻; ras-RNAi is labeled as "EV". This terminology (presumably "empty vector") is not defined in either the text or the figure legend. In addition, the magenta channel for the Ras-G37 condition appears to be flipped horizontally.

    We replaced with “-“ in figure and figure legend

    Page 7: The text states that "Ras-C40 activates the PI3K pathway," whereas the figure depicts Ras-C40 as activating the RalA pathway. This discrepancy could be confusing for the reader and should be corrected.

    The diagram has been corrected

    Figures 4 and 5: To facilitate interpretation, it may be helpful to include a schematic of the PI3K complex indicating the different subunits used in the study, along with information (potentially color-coded) about whether each construct primarily acts as an activator or inhibitor of PI3K function.

    Figure 4E and Figure 5E were added

    Figure 4A and 4B: For clarity and consistency with the text, the panels (and corresponding plots) for dp110-WT and dp110-CAAX could be placed before those for dp110-D954A and dp110-ΔRBD.

    Order of constructs was rearranged

    Figure 5C: The term "p60-TCEp3," which appears to correspond to the germ plasm-targeted p60-WT construct, is not defined in either the figure legend or the main text.

    Clarification was added to the text (p.11, line 225)

    Page 12: The reference "(Fig. S1A, Movie 1)" should be corrected to "(Fig. S2A, Movie 1)."

    Corrected

    Page 13: There is a missing word in the sentence "the biosensor appeared to be enrich to...", which should be corrected to "enriched."

    Corrected

    Figure 7A: Although the data presented are interesting and ultimately support the authors' conclusion that Torso regulates PIP3 levels, the results are somewhat counter-intuitive and may be confusing for readers. The authors might consider moving this panel to the Supplementary Figures. In addition, it could be informative to include PIP3 measurements for gcl⁻/⁻ (and possibly gcl⁺/⁻) pole buds in Figure 7B, as PIP3 appears particularly enriched in these conditions compared to wild type.

    We agree that at first the findings in the early embryos were confusing, but we prefer including them in the main figure to demonstrate changes in PIP 3 distributions in torso mutants. We are now providing a possible explanation for these findings (p13 line 270-). The differences are quite clear in the older embryos and measurements shown in 7B-D. Pole bud measurements for gcl-/- and gcl+/- are shown in figure 6 E-G.

    Reviewer #2

    Fig. legends to 1C and 1D are swapped.

    Corrected

    Why is csw not necessary for PGC formation? It acts upstream of Ras. This is not discussed.

    We now highlight this point in the text (and refer to studies on the sevenless kinase, which suggested a similar position of Csw parallel or downstream of Ras (page 6 line 107-).

    Fig 3C. Consider changing the order of the ras-variants used: S35, G37, C40 instead of S35, C40, G37.

    We changed the schematic in Figure 3C that should make the order of Ras variants more intuitive.

    Fig 4A, B: Consider changing the order of the panels. Control, dp110-wt, dp110-CAAX, dp110-D954A, dp110-deltaRBD.

    Order of constructs was rearranged

    Fig S4 is mentioned in the text before S2 and S3. Consider changing the suppl. figure order.

    Order of supplementary figures was rearranged

    Page 12: Fig S1 A does not show PIP2 dynamics. Movie 1 is not available to this reviewer. The authors most likely refer to fig. S2.

    Movie 1 was uploaded and figure calls were corrected

    Page 13, 1st para: Why do the authors use glc heterozygous embryos to look at PIP3 and PIP2? Particularly so when they report later in the MS that glc+/- behave differently to wt controls in terms of PIP3 levels (Fig. 7C). By looking at gcl+/+, they might find that now PIP2 levels are different in gcl mutant embryos or that the differences between PIP3 levels in +/+ and -/- are larger than compared with +/-.

    Since gcl+/- embryos form the same number of PGCs as WT but show a statistically significant increase in PI3K activity when comparing membrane to cytoplasm staining intensity, we favor using gcl+/- embryos, as these embryos may represent a more sensitive test for PIP2 and PIP3 levels.

    Pages 15 and 16: revise figure calls in the text.

    Figure calls were revised

    M+M: How were gcl+/- and gcl-/- embryos identified?

    Since all genetic manipulations in this alter the maternal contribution to the embryo, we us the term ‘mutant’ embryos referring to the maternal genotype (indicated on page 3 line 33 and more clearly stated in material and methods and reagent table). Embryos derived from mother of a specific maternal genotype are all identical, thus we can easily distinguish between embryos derived from homozygous mutant mothers (gcl-/-) or heterozygous mutant mothers (gcl-/+) In the reagents table we include the precise genotype description. “CyO” refers to the balancer chromosome commonly used to identify heterozygotes on the second chromosome. Flies with the CyO balancer have curly wings.

    Reviewer #3

    Figure 1B: The authors describe that embryos with OptoSos still form buds which protruded from the cortex, but PGCs largely fail to cellularize (described in pg. 5). I'm not sure what they meant by "fail to cellularize" as this is not obvious to me when looking at the figure. The authors should describe how they know it's cellularized in the controls and not in the OptoSos or change the wording to "suggesting a failure to cellularize".

    We used the word ‘protruded’ to describe our live observations. PGCs were quantified in fixed embryos, immunostained with anti-Vasa antibody to count Vasa positive cells (Fig 1C and D. We observe a lack of Vasa-positive PGCs, only in the light-activated OptoSos condition.

    Fig. 1B, lines 4-5: at what stage are these embryos? Cycle 9? Cycle 14? Both?

    Nuclear cycles of embryos for each panel are noted on the left side of each panel

    Fig. 4A: add dp110-CAAX results to Results section

    dp110-CAAX results are included in the Results section (p.9. line 177)

    Figure 5C: The hyper-clustered phenotype they describe is hard to visualize in this figure (described in pg. 11). The authors should describe what is meant by "hyper-clustered".

    We agree and re-worded the description of this observation to be clearer, page 11, line 226-.

    Figure 7: When comparing Fig. 7A and 7B torsoHH/WK images, we can see that in Fig. 7A that PIP3 pattern changes such that PIP3 is now at the most posterior end where PGC will eventually form (compared to control that has low PIP3 in this region), but then in Fig. 7B they are looking at the buds and they say PIP3 levels decrease, which does not correspond to Fig. 7A. Are these simply different stages and PIP3 levels change over time (looking at Fig. 7C, PIP3 does not seem to change a lot over time)?

    The figure legend now states more clearly that embryos were of different ages. We also explain in the text the apparent discrepancy in the patterns before and during budding (page13 line 266). The time points in figure 7C span nuclear cycle 10, not earlier (page14 line 274). By measuring membrane to cytoplasmic distribution, a more accurate comparison is possible at this stage.

    p. 5, line 5: "Optosos" is written "OptoSos" elsewhere (suggest using OptoSos throughout)

    Corrected

    Is it possible that inhibition of myosin II recruitment is due to conversion of PIP2 -> PIP3, thus loss of PIP2, or is it that myosin is specifically recruited to regions where PIP2 is high? This seems like a point that should be added to the discussion.

    This point is now discussed on page 20, line 403

    p. 5, line 6: suggest adding a comma after "Ras" for clarity

    Corrected

    p. 5, last line: the genotype is "w^1118" (with ^ indicating a superscript), not "w^-1118", and is italicized (this should be corrected throughout)

    Corrected

    p. 6, line 2: replace "cellularizing" with "cellularization"

    Corrected

    p. 6, lines 11-13: Where is it shown that knockdown of csw, dsor1 and rolled did not restore PGC formation? The data are not present in Fig. 2C (could include in supp fig?)

    We added these data as Supplementary figure 1

    p. 7, line 1: replace "interfere" with "interferes"

    Corrected

    p. 7, last three lines: what is stated here, "Ras-G37 [activates] both the RalA and the PI3K pathways, and Ras-C40 activates the PI3K pathway" is not consistent with what is diagrammed in Fig. 3C, where Ras-C40 is indicated as activating RalA (please correct either the text or the diagram)

    We apologize and corrected the figure

    p. 11, lines 1-2: the Pi3K21B gene and transcript should be italicized (note that Pi3K21B is the official gene name on FlyBase)

    Gene name was italicized

    p. 11, lines 6-10: it might be helpful to explain how the p60 construct was overexpressed (current lines 9-10) before describing the results (current lines 7-8)

    Clarification on p60 construct was added to p.11, line 215-

    p. 12, paragraph 2, line 2: the PIP2 biosensor should be written as "PLCgamma[PH]:mCherry" throughout, not "PLCy[PH]:mCherry"; this should be changed in the figures as well as the text (Symbol font can be used to turn "g" into lower-case "gamma", both in Word and in Illustrator)

    Gamma symbol was added

    It would also be helpful to show the overlap of the PIP2 and PIP3 signals in control vs. gcl mutants at different stages so the relative distribution and intensity of the signals can be better appreciated (consider adding this as a supplementary figure).

    Our data show that PIP2 is not affected by lack of GCL (Fig 6 B-D). We thus do not think that simultaneous imaging of PIP2 and PIP3 in gcl-/- would add to our conclusions. Furthermore, these experiments would require a significant time investment to generate the respective genotypes. Thus, we agree with the reviewer that this is experiment is beyond the scope of the paper.

    p. 12, paragraph 2, line 3: it does not appear that the two PIP markers were used "simultaneously" in Fig. 6A; however, this is evident from Fig. S2 and Movie 1 (consider placing callouts to these earlier in the paragraph or moving the description of simultaneous expression and observation of the two markers later in the paragraph to avoid confusion)

    We did simultaneously image PIP2 and PIP3 sensors and have added this as Movie 1 and also in supplementary Figure S4, which are now clearly referred to in the text.

    p. 12, paragraph 2, line 7: replace "Fig. S1A" with "Fig. S2" (this was confusing)

    Figure call was updated

    p. 16: change "Fig. 7G-I" to "Fig. 8G-I"

    Figure call was updated

    p. 20, Deming reference: there appears to be a stray asterisk in the title

    Asterisk was removed from reference

    Fig. 1D: need to explain that the colors in the graph indicate the numbers of PGCs formed (this could also be added as a label across the top of the graph); in addition, the number of embryos examined for each genotype should be included in the legend

    We added a label at the top of the graph and ‘n’ were added to figure legend

    Fig. 2B: spell out where csw, dsor1 and rolled data are shown; also, "n" is not defined; was this the number of embryos per genotype?

    We added these data as Supplemental Figure 1

    Fig. 3B: "EV" should be defined in the legend; is this "empty vector"?

    We are using a “-“ to mark controls without transgene

    Fig. 3C: see previous comment re: mistake in the diagram; I believe Ras-C40 was described as activating PI3K, not RalA

    We apologize and corrected the figure

    Fig. 4B, line 2: was the graph plotted from the data in panel (C) or panel (A)? panel (A) seems more likely, because the data in C is plotted in D; please correct the panel callout

    Figure legend was updated to refer to the correct panel

    Fig. 5C: describe "p60-TCEp3" in the legend

    We added germplasm-targeting 3’UTR (TCEp3) to legend and the construct and reference are provided in Material and Methods section

    Figure 6: In Fig. 6E-G, the "brightness" of PIP3 at the membrane corresponds to the images even with different views (posterior and orthogonal) and agrees with the graph.

    However, when looking at Fig. 6B, it looks to me that PIP2 is brighter in gcl+/-, but the opposite is true when looking at Fig. 6D (i.e., PIP2 looks brighter in gcl-/-). The authors might want to comment on this.

    We have updated the figure to better reflect our observations.

    Fig. 6A: define "(fire)" here or in the first figure legend where this is used

    We added an inset for the fire lookup table to clearly define the pseudcolor scheme used in the image

    Figure 8 title: "Actin fluorescence is increased in gcl-/- pole buds",But their graph in Fig. 8B comparing actin in gcl+/- to -/- is not significant

    Thanks for catching our mistake, myosin not actin is changed

    Fig. 8I: replace "Scarlett" with "Scarlet"

    Corrected

    Fig. 8D-F: Although the plots in panel E agree with the images in panel D, it is unclear why those in panel F are not more concordant. In F, myosin appears enriched at the cortex relative to the cytoplasm in gcl-/- mutants, which is hard to reconcile with the data in D-E.

    We have updated the figure to better reflect our observations.

    Fig. S2A: define the three time points shown here, and clarify that these are shown left to right (if this is indeed the case)

    We removed S2A and updated the movie to replace it

    Fig. S4: change "P60" to "p60" in the figure title

    Corrected

    Movie: The movies showing PIP2 and PIP3 in whole embryos are nice, but it would also be helpful to also include merged images of the two channels, so the reader can examine the relative accumulation of the two PIPs over time.

    Merged images panel was added to the movie.

  2. Review Commons

    Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

    Learn more at Review Commons

    Referee #3

    Evidence, reproducibility and clarity

    Summary:

    Although Torso is known to antagonize primordial germ cell (PGC) formation, the underlying mechanisms remain unclear. Canonical Torso signalling typically results in activation of Ras. However, the authors show that Ras-mediated suppression of PGC formation is independent of the Raf/MEK/ERK pathway. Instead, they uncover an unexpected role for Torso in activating phosphoinositide 3-kinase (PI3K) that promotes formation of PIP3 enriched posterior membrane domains. The resulting increase in PI3K activity disrupts PGC formation. Furthermore, they show that by promoting Torso degradation, the ubiquitin ligase adaptor Germ Cell-Less (GCL) primes the posterior membrane with reduced PIP3 to facilitate PGC formation. Lastly, the authors suggest a model where antagonistic relationship between GCL and Torso influences actomyosin contractility that may allow the bud to constrict for proper PGC formation.

    Major comments:

    Figure 1B: The authors describe that embryos with OptoSos still form buds which protruded from the cortex, but PGCs largely fail to cellularize (described in pg. 5). I'm not sure what they meant by "fail to cellularize" as this is not obvious to me when looking at the figure. The authors should describe how they know it's cellularized in the controls and not in the OptoSos or change the wording to "suggesting a failure to cellularize".

    Figure 5C: The hyper-clustered phenotype they describe is hard to visualize in this figure (described in pg. 11). The authors should describe what is meant by "hyper-clustered".

    Figure 6: In Fig. 6E-G, the "brightness" of PIP3 at the membrane corresponds to the images even with different views (posterior and orthogonal) and agrees with the graph. However, when looking at Fig. 6B, it looks to me that PIP2 is brighter in gcl+/-, but the opposite is true when looking at Fig. 6D (i.e., PIP2 looks brighter in gcl-/-). The authors might want to comment on this.

    It would also be helpful to show the overlap of the PIP2 and PIP3 signals in control vs. gcl mutants at different stages so the relative distribution and intensity of the signals can be better appreciated (consider adding this as a supplementary figure).

    Figure 7: When comparing Fig. 7A and 7B torsoHH/WK images, we can see that in Fig. 7A that PIP3 pattern changes such that PIP3 is now at the most posterior end where PGC will eventually form (compared to control that has low PIP3 in this region), but then in Fig. 7B they are looking at the buds and they say PIP3 levels decrease, which does not correspond to Fig. 7A. Are these simply different stages and PIP3 levels change over time (looking at Fig. 7C, PIP3 does not seem to change a lot over time)?

    Page 15, last paragraph: "If myosin II recruitment is inhibited when PIP3 levels are high" Is it possible that inhibition of myosin II recruitment is due to conversion of PIP2 -> PIP3, thus loss of PIP2, or is it that myosin is specifically recruited to regions where PIP2 is high? This seems like a point that should be added to the discussion.

    Overall, I think their claim that antagonistic activities of GCL and Torso is crucial for PGC formation is well justified. The combination of optogenetic tools with activation and lof mutants is nicely done. Some clarification regarding the PIP3 and PIP2 levels will be helpful to the reader (see my comments above). The myosin claim is less convincing (see my comment on Fig. 8D-F below).

    Minor comments on the text:

    p. 5, line 5: "Optosos" is written "OptoSos" elsewhere (suggest using OptoSos throughout) p. 5, line 6: suggest adding a comma after "Ras" for clarity p. 5, last line: the genotype is "w^1118" (with ^ indicating a superscript), not "w^-1118", and is italicized (this should be corrected throughout) p. 6, line 2: replace "cellularizing" with "cellularization" p. 6, lines 11-13: Where is it shown that knockdown of csw, dsor1 and rolled did not restore PGC formation? The data are not present in Fig. 2C (could include in supp fig?) p. 7, line 1: replace "interfere" with "interferes" p. 7, last three lines: what is stated here, "Ras-G37 [activates] both the RalA and the PI3K pathways, and Ras-C40 activates the PI3K pathway" is not consistent with what is diagrammed in Fig. 3C, where Ras-C40 is indicated as activating RalA (please correct either the text or the diagram) p. 11, lines 1-2: the Pi3K21B gene and transcript should be italicized (note that Pi3K21B is the official gene name on FlyBase) p. 11, lines 6-10: it might be helpful to explain how the p60 construct was overexpressed (current lines 9-10) before describing the results (current lines 7-8) p. 12, paragraph 2, line 2: the PIP2 biosensor should be written as "PLCgamma[PH]:mCherry" throughout, not "PLCy[PH]:mCherry"; this should be changed in the figures as well as the text (Symbol font can be used to turn "g" into lower-case "gamma", both in Word and in Illustrator) p. 12, paragraph 2, line 3: it does not appear that the two PIP markers were used "simultaneously" in Fig. 6A; however, this is evident from Fig. S2 and Movie 1 (consider placing callouts to these earlier in the paragraph or moving the description of simultaneous expression and observation of the two markers later in the paragraph to avoid confusion) p. 12, paragraph 2, line 7: replace "Fig. S1A" with "Fig. S2" (this was confusing) p. 16: change "Fig. 7G-I" to "Fig. 8G-I" p. 20, Deming reference: there appears to be a stray asterisk in the title

    Minor comments on the figures and figure legends:

    Fig. 1B, lines 4-5: at what stage are these embryos? Cycle 9? Cycle 14? Both? Fig. 1C: see previous comment about "w^1118" genotype nomenclature Fig. 1D: need to explain that the colors in the graph indicate the numbers of PGCs formed (this could also be added as a label across the top of the graph); in addition, the number of embryos examined for each genotype should be included in the legend Fig. 2B: spell out where csw, dsor1 and rolled data are shown; also, "n" is not defined; was this the number of embryos per genotype? Fig. 3B: "EV" should be defined in the legend; is this "empty vector"? Fig. 3C: see previous comment re: mistake in the diagram; I believe Ras-C40 was described as activating PI3K, not RalA Fig. 3E: fix "w^1118" as described above Fig. 4A: add dp110-CAAX results to Results section Fig. 4B, line 2: was the graph plotted from the data in panel (C) or panel (A)? panel (A) seems more likely, because the data in C is plotted in D; please correct the panel callout Fig. 5C: describe "p60-TCEp3" in the legend Fig. 6A: define "(fire)" here or in the first figure legend where this is used Figure 8 title: "Actin fluorescence is increased in gcl-/- pole buds",But their graph in Fig. 8B comparing actin in gcl+/- to -/- is not significant Fig. 8D-F: Although the plots in panel E agree with the images in panel D, it is unclear why those in panel F are not more concordant. In F, myosin appears enriched at the cortex relative to the cytoplasm in gcl-/- mutants, which is hard to reconcile with the data in D-E. Fig. 8I: replace "Scarlett" with "Scarlet" Fig. S2A: define the three time points shown here, and clarify that these are shown left to right (if this is indeed the case) Fig. S4: change "P60" to "p60" in the figure title

    Movie: The movies showing PIP2 and PIP3 in whole embryos are nice, but it would also be helpful to also include merged images of the two channels, so the reader can examine the relative accumulation of the two PIPs over time.

    Referees cross-commenting

    I agree enthusiastically with the comments of the other reviewers, who often came to the same conclusion I did about the manuscript and the data, including some of the detailed points about the figures, etc.

    Significance

    General assessment:

    The many strengths of this manuscript include elegant genetic and optogenetic approaches using well-designed transgenes.

    The main weakness is the lack of experiments showing simultaneous live imaging of the PIP2 and PIP3 sensors in gcl-/- and other genetic backgrounds, which would help the reader better envision how regulators of this pathway affect phospholipid distribution at the level of whole embryos and prospective pole cells. Note that because of the time required, I do not insist that they do this.

    Advance:

    Study demonstrates for the first time an unexpected role of Torso in PI3K regulation

    Audience:

    germ cell afficionados, developmental biologists, cell biologists, PI3K researchers

    My field of expertise:

    Drosophila, germ cell development, genetics, cell biology, live imaging, phosphoinositides

  3. Review Commons

    Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

    Learn more at Review Commons

    Referee #2

    Evidence, reproducibility and clarity

    The MS by Saiduddin et al. investigates the connection between the germ cell-less (gcl) germplasm component, the RTK Torso and cortical PI3P levels in the formation of a normal complement of primordial germ cells (PGCs) in the Drosophila embryo. The authors find that GCL regulates Torso levels, which in turn activate the PI3-kinase in a ras-dependent (but raf-independent) manner. It then follows that the realm of action of gcl defines a region at the posterior pole of the embryo where PI3P levels are sufficiently low to allow PGC formation.

    Specific points:

    • Fig. legends to 1C and 1D are swapped.
    • Why is csw not necessary for PGC formation? It acts upstream of Ras. This is not discussed.
    • Fig 3C. Consider changing the order of the ras-variants used: S35, G37, C40 instead of S35, C40, G37.
    • Fig 4A, B: Consider changing the order of the panels. Control, dp110-wt, dp110-CAAX, dp110-D954A, dp110-deltaRBD.
    • Fig S4 is mentioned in the text before S2 and S3. Consider changing the suppl. figure order.
    • Page 12: Fig S1 A does not show PIP2 dynamics. Movie 1 is not available to this reviewer. The authors most likely refer to fig. S2.
    • Page 13, 1st para: Why do the authors use glc heterozygous embryos to look at PIP3 and PIP2? Particularly so when they report later in the MS that glc+/- behave differently to wt controls in terms of PIP3 levels (Fig. 7C). By looking at gcl+/+, they might find that now PIP2 levels are different in gcl mutant embryos or that the differences between PIP3 levels in +/+ and -/- are larger than compared with +/-.
    • Pages 15 and 16: revise figure calls in the text.
    • M+M: How were gcl+/- and gcl-/- embryos identified?

    Significance

    This work reveals a novel, transcription-independent role for Torso in the regulation of cortical lipid compartmentalization and provides a molecular explanation for how Tor activity at the embryo posterior delimits the area where PGCs arise. The experiments are superbly documented, the MS is a pleasure to read and the hypotheses are elegantly tested. While the broader generality of the findings remains uncertain, particularly since the role(s) of gcl in germ cell development do not seem to be evolutionary conserved, the study sheds light on a ras-dependent, transcription-independent function of RTKs in cellularization, a function most likely to be essential also in other contexts.

  4. Review Commons

    Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

    Learn more at Review Commons

    Referee #1

    Evidence, reproducibility and clarity

    This is an outstanding and elegant study that addresses an important question in developmental and germline biology: how the soma-germline boundary is established during embryogenesis. This represents one of the most fundamental cell-fate decisions during organismal development. The authors combine elegant genetics, optogenetics, quantitative live imaging, and lipid biosensors to provide a compelling mechanistic framework linking receptor degradation, lipid signaling, and cytoskeletal dynamics. They show that Torso signaling via PI3K and PIP3 antagonizes primordial germ cell (PGC) formation and promotes somatic cell fate. Furthermore, they demonstrate that GCL-mediated degradation of Torso at the posterior pole creates a PIP3-low membrane domain that permits myosin II recruitment and pole bud constriction, thereby enabling PGC formation. Together, these results clearly demonstrate how the soma-germline boundary is established.

    We have only minor comments on the manuscript, primarily aimed at improving clarity for non-specialist readers:

    1. Figure 1D: It would be useful to indicate the number of embryos analyzed for these experiments (n = ?).
    2. Figure 3B: The control condition for gcl⁻/⁻; ras-RNAi is labeled as "EV". This terminology (presumably "empty vector") is not defined in either the text or the figure legend. In addition, the magenta channel for the Ras-G37 condition appears to be flipped horizontally.
    3. Page 7: The text states that "Ras-C40 activates the PI3K pathway," whereas the figure depicts Ras-C40 as activating the RalA pathway. This discrepancy could be confusing for the reader and should be corrected.
    4. Figures 4 and 5: To facilitate interpretation, it may be helpful to include a schematic of the PI3K complex indicating the different subunits used in the study, along with information (potentially color-coded) about whether each construct primarily acts as an activator or inhibitor of PI3K function.
    5. Figure 4A and 4B: For clarity and consistency with the text, the panels (and corresponding plots) for dp110-WT and dp110-CAAX could be placed before those for dp110-D954A and dp110-ΔRBD.
    6. Figure 5C: The term "p60-TCEp3," which appears to correspond to the germ plasm-targeted p60-WT construct, is not defined in either the figure legend or the main text.
    7. Page 12: The reference "(Fig. S1A, Movie 1)" should be corrected to "(Fig. S2A, Movie 1)."
    8. Page 13: There is a missing word in the sentence "the biosensor appeared to be enrich to...", which should be corrected to "enriched."
    9. Figure 7A: Although the data presented are interesting and ultimately support the authors' conclusion that Torso regulates PIP3 levels, the results are somewhat counter-intuitive and may be confusing for readers. The authors might consider moving this panel to the Supplementary Figures. In addition, it could be informative to include PIP3 measurements for gcl⁻/⁻ (and possibly gcl⁺/⁻) pole buds in Figure 7B, as PIP3 appears particularly enriched in these conditions compared to wild type.

    Significance

    The biological question is highly interesting, the experimental design is very clear, and the data are convincing throughout. The imaging, quantification, and movies are of very high quality and strongly support the authors' conclusions. Overall, this manuscript represents a significant conceptual and technical advance and will be of broad interest to the fields of germline biology, membrane biology, and embryonic morphogenesis.

  5. Version published to 10.64898/2025.12.30.697122 on bioRxiv