Fission yeast Pcp1 links polo kinase-mediated mitotic ... - BioMedSearch

The EMBO Journal Review Process File - EMBO-2009-71857

Manuscript EMBO-2009-71857

Fission yeast Pcp1 links polo kinase-mediated mitotic entry to γ-tubulin-dependent spindle formation Chii Shyang Fong, Masamitsu Sato, Takashi Toda Corresponding author: Takashi Toda, Cancer Research UK, London Res. Inst.

Review timeline:

Submission date: Editorial Decision: Revision received: Editorial Decision: Revision received: Accepted:

09 July 2009 30 July 2009 30 September 2009 15 October 2009 16 October 2009 16 October 2009

Transaction Report: (Note: With the exception of the correction of typographical or spelling errors that could be a source of ambiguity, letters and reports are not edited. The original formatting of letters and referee reports may not be reflected in this compilation.)

1st Editorial Decision

30 July 2009

Thank you for submitting your manuscript for consideration by The EMBO Journal. It has now been seen by three reviewers, whose comments are attached below. As you will see, all referees in principle consider your findings interesting and potentially important. They do nevertheless all raise a number of substantive points that would have to be satisfactorily addressed before publication may be warranted. I realize that the list of specific criticisms is quite extensive, and would therefore prefer not to go through it in detail in this letter. Given the overall positive assessment of the referees in principle, I would nevertheless be inclined to offer you the possibility to respond to the referees' concerns in a revised version of the manuscript. Thus, should you feel confident that you might be able to adequately address the various points and to improve the manuscript through additional experimental evidence, we should be happy to consider a revised manuscript for publication. Please be reminded that it is EMBO Journal policy to allow a single round of revision only, and that it is therefore essential that you diligently answer to all the points raised at this stage if you wish the manuscript ultimately to be accepted. In any case, please do not hesitate to get back to us should you need feedback on any issue regarding your revision. Thank you for the opportunity to consider your work for publication. I look forward to your revision. Yours sincerely,

© European Molecular Biology Organization

1


Editor The EMBO Journal _____ REFEREE REPORTS: Referee #1 (Remarks to the Author): In this paper the authors describe the identification of two novel fission yeast pcp1 mutant alleles. Using a number of elegant cell biological, molecular, EM, and biochemical techniques they report each pcp1 allele has distinct effects upon the recruitment of the gamma-tubulin complex and the polo like kinase, Plo1, to the SPB. This manuscript is of interest to a general cell biological audience and in my opinion is suitable for publication in EMBO J. - ONCE each of the points below have been addressed. Throughout the paper there is variation in the time which cells are held at the restrictive temperature (some expts for 2 hrs, some for 4 hrs). Need to have uniform timings throughout the paper (show both 2 and 4 hr data points in supp data - if necessary). Intro Pg3 para1 Change "...SPB is involved in such multiple pathways" to "..SPB is involved in each distinct pathway." Pg4 ln4 Change "places" to "foci". Pg4 ln9 Change "...except that it is..." to "...except to identify that it is...". Pg4 ln11 Change "...roles in mitotic progression..." to "...roles during mitotic progression...". Pg4 ln14 Change "...have regulatory subunits that act as molecules responsible for targeting kinase catalytic subunits to specific cellular locations. Instead it is often the case that certain components localising to individual subcellular structures bind polo kinase, thereby loading it to these places (Elia et al, 2003)." To "...have regulatory subunits that act as molecules responsible for targeting kinase catalytic subunits to specific cellular locations. Instead it is Polo kinase has been shown to localize by associating with proteins already recruited to a distinct cellular position (Elia et al, 2003)." Pg4 last ln "Despite this distinct cell-cycle dependent Plo1 localisation, its spatial regulation remains elusive." I would disagree with this statement (See studies from Hagan lab for example which have gone some way to establish role Fin1 and Cut12 have in regulating Plo1 recruitment to the SPB). Please change text accordingly. Results Supplementary Figure S2: Need immunofluorescence images at 4 hr, so corresponds with biochemical data. Should discuss (in a sentence) significance of ability to localise but not function. Pg6 ln4


2


"....which coincided with mitotic entry (light red columns in Figure 1B, C). This viability drop paralleled the emergence of lethal chromosome missegregation (blue lines)." Only refers to Figure 1 B NOT C. Chromosome missegregation is not necessarily lethal. Figure 1 C show cells with post anaphase array not mitotic spindle. Should show mitotic cells. If author keeps current Fig1C (in addition to requested images), should describe it in the text. Figure 1D Spindle is seen to associate with both SPBs (microtubules must be interdigitating as SPBs migrating apart around the nuclear membrane) in pcp1-18 mutant. I do not understand how this can occur if one of the two SPBs does not embed into the nuclear membrane (EM data or figure 2A). Pg6 ln6 Finding is consistent with recent study of Tallada et al. Sentence to discuss. Figure 3 (& S3) Should also show images of Alp14 & Alp16 localisation in a pcp1-15 and pcp1-18 mutant with 2 separated SPBs (with appropriate colabelled SPB marker as control). Pg8 ln7 Change "..address the significance of -TuC localisation for Pcp1 function,..." to "..examine how Pcp1 function affects -TuC localisation,..." Figure 4 B Show wt control in same axes. Would be useful if authors could differentiate between Alp4 on 1 SPB or 2 SPBs. Figure 5 (with 3 & S3). Why does the gamma tubulin complex only affect the pcp1-15 and not the pcp1-18 allele if in the latter one of the SPBs does not enter into the nuclear membrane. Is the gamma tubulin complex on each SPB in cells in which these organelles are separating SPBs? Pg9 ln14 "..Pcp1 is required for -TuC recruitment to the mitotic SPB, but also suggest that -TuC recruitment during interphase is mediated by a Pcp1- and Mto1-independent echanism" The authors need to expand on how this occurs when the SPB is outside the nuclear membrane in these cells. Figure 6 A: Appears like there is an extra accumulation of fragmented SPB (or Cut11 anyway) in the pcp1-18 mutant. Explain/ D: Should mix cells and observe simultaneously to compare intensity of Plo1 signal in pcp1-15 and pcp1-18 cells. One would assume that Plo1 is only recruiting to the "old" or "maternal" SPB in pcp1-18 mutant. Show cells with 2 distinct SPBs and appropriate markers to examine this. Pg11 & Figure 6F If the stf1.1 mutant bypasses requirement for Cdc25, enhances Plo1 recruitment to the SPB and its overexpression suppresses pcp1-18, logic would dictate it may also bypass the requirement for Pcp1 (/ alleviate the effect of the pcp1-18) for Plo1 SPB recruitment. Therefore authors should address whether Plo1 is on one / two SPBs in stf1.1 pcp1.18 cells? Referee #2 (Remarks to the Author): Summary:


3


Fong et al. provide evidence that the fission yeast pericentrin-like Pcp1 regulates multiple functions of the SPB. It does this through recruitment of two important proteins to the SPB, the gamma tubulin complex and polo kinase (plo1). A separation of function for Pcp1 is shown through the characterization of two alleles (pcp1-15 and pcp1-18). Both mutants result in abnormal spindles (specifically, monopolar spindles emanating from only the mother SPB); however the molecular mechanism behind this defect appears to be different in the two alleles. Pcp1-15 fails to recruit gamma-tubulin to the mitotic SPB and pcp1-18 fails to recruit Plo1 (polo kinase). Pcp1-18 also shows a defect in SPB insertion into the nuclear envelope during mitosis. This defect can be rescued by overexpressing NPC proteins, such as cut11 (Ndc1), cut12, and pom152. This study provides new insight into how the SPB interacts with several pathways and further characterizes a relatively newly identified protein. Interesting points: ï Includes interesting further characterization of Pcp1 - a fission yeast SPB component found by the Davis lab in 2002. ï Shows two very distinct functions of the Pcp1 proteins by isolating 2 different alleles - also demonstrates how two pathways can lead to defects in MT nucleation from the daughter SPB ï Pcp1-18 mutant shows that Pcp1 has a role in the reorganization of the nuclear envelope in mitosis for the insertion of the SPB into nuclear envelope. This is supported by an interesting genetic linkage to the Cut11/Cut12 pathway. ï Pcp1-15 mutant demonstrates the role of Pcp1 in gamma-tubulin recruitment is important for mitotic spindle nucleation Weak points to address: ï Claims in Introduction that the roles for Pcp1 have not been characterized for cell division and MT formation - but Rajagopalan S. et al (2004, Curr. Biol. 14:69-74) does characterize its role in the spindle orientation checkpoint. Furthermore, there is important pertinent information on the Cdc5 and Spc110 genes in budding yeast that should be presented in the paper, particularly in the Discussion for comparison to the new data presented in this manuscript. ï The claim that pcp1-18 and pcp1-15 identified different functions of the protein would be well supported by intragenic complementation. Do these alleles show intragenic complementation? ï A similar genetics question. Do the authors find synthetic lethality in double mutant strains containing pcp1-18 or pcp1-15 and a mutant allele of an interacting gene. Such interactions would support the suppression data and may support the separation-of-function argument. ï Supplementary figure 1 should be moved to the figures in the paper or and the specific mutations made in the 2 alleles should be discussed. There is plenty of structure/function work on Spc110 in budding yeast to make the comparison worthwhile. ï Figure 1 - when characterizing mutant effects on viability, the authors do not show or discuss any effects on cell cycle progress. Are there any delays in G1 or other parts of the cell cycle? ï Figure 1 - movies: It looks like pcp1-15 nucleates MTs from one pole and then the second pole later on, is this correct? If so, the phenotype is more complex than reported. ï Figure3 - Can low levels of gamma tubulin complex protein explain why the mother can still nucleate MTs even without Alp localization? ï Figure 4 - The phenotype seems less prominent than Figure 3 - Alp4 decreases by less than half in the Fig. 4 experiment. Is there an explanation for this result? ï Figure 7 - The IPs experiments are weak because there is no control showing that the Pcp1 IP is in anyway specific for Plo1 and gamma-tubulin. It seems possible that the IP is pulling down most of the SPB. If so, that could be a great reagent for this group, but not very useful in this context. ï Figure 7 - The model suggests an unusual localization for Plo1 in the SPB. Is there any data to support this localization? Minor points: ï The first sentence of the Introduction is awkward. ï First section of results - when discussing pcp1 mutants, it refers to S2. It should refer to S1. ï On pages 6 & 7, the description of the monopolar spindles would be more clear if some of the use of "spindle" was replaced with "spindle microtubules." ï Figure 1 - put %s in figures 1C and 1D on the figure of what portion of cells behave this way, also Figure2 put %s on figure ï Figure 2 - How are they sure they are in the same mitotic phase in 3 cells? Also, looks like an elongation defect of the existing MT.


4


Referee #3 (Remarks to the Author): This manuscript describes distinct functions of the SPB-localized protein Pcp1 in mitotic spindle assembly through characterization of the phenotypes of two novel temperature sensitive alleles of pcp1 called pcp1-15 and pcp1-18. This study represents a very nice use of genetics to tease apart different roles of a multifunctional protein. Furthermore, this work provides an entry point for future studies of how the SPB/centrosome regulates multiple pathways. Unfortunately, many of the key experiments lack proper quantification of the phenotypes described, making it difficult to evaluate the strength of the conclusions. In addition, characterization of the how the multicopy and genetic suppressors rescue the pcp1-18 mutant allele would strengthen the manuscript. Specific Points: 1) Page 6 - Although it is stated that 80% of pcp1-15 and pcp1-18 mutants have monopolar spindles, the number of cells counted needs to be stated. This is needed throughout the manuscript. In addition, from the images shown in Figure 1D, it looks like both mutants form bipolar spindles that collapse. This phenotype seems to be ignored, and merits some discussion of how this fits with the models for what is wrong in each mutant allele. 2) Quantification is needed for the phenotypes described in Figure 2A, and the number of cells counted for the graph in 2B needs to be shown. 3) The number of cells counted for the graph in 3B needs to be indicated. Quantification is needed for the phenotypes described in Figure 3C. 4) A major conclusion of the paper is that SPB insertion into the nuclear envelope is defective in the pcp1-18 mutant. However, this conclusion appears to be based on 2 electron micrographs, one of which is not very convincing. The SPB only inserts into the nuclear envelope during mitosis, so it is crucial to show that cells are in mitosis when looking for a failure in SPB insertion into the nuclear envelope. The image in figure 5E is pretty convincing, but the one in figure 5D is not. The authors state that the cell in figure 5D is in mitosis because of the nuclear envelope protrusion, which they say is caused by the ends of a monopolar spindle pushing on the nuclear envelope, which is completely possible. However, it is also possible that the cell is actually in interphase, and the protrusion is caused by forces generated by cytoplasmic microtubules attached to the nuclear envelope. If the cell is in mitosis, then microtubules from the monopolar spindle should be apparent in the protrusion. In any case, it is important to make clear whether it was concluded that the pcp118 mutant has an SPB insertion defect is based on just these 2 images, and if not, do the authors have a sense of how frequent the defect occurs. 5) It would be interesting to know more about how the multicopy (cut11, pom152, cut12, kms2, tcg1) and genetic (cut12-s11, wee1-50) suppressors of pcp1-18 rescue the mutant phenotype. For example do the suppressors enhance recruitment of Plo1 to the SPB, or do they rescue the defect in SPB insertion into the nuclear envelope. Similarly, since pcp1-18 has defects in both Plo1 recruitment to the SPB and in SPB insertion into the nuclear envelope, do plo1 mutants have defects in SPB insertion? 6) There are a number of places in the discussion that should be worded a bit more precisely. For example, at the top of page 14, it is stated that Pcp1 activates Plo1 kinase at the SPB, but no evidence for this is presented in the paper. At the bottom of page 14 it is stated that Pcp1 is required for Plo1 recruitment to the SPB. Given the data presented, it would be more accurate to state that Pcp1 plays a role in recruitment of Plo1 to the SPB. On page 15, it is stated that SPB insertion into the nuclear envelope is under the control of Plo1. No evidence for this was presented, therefore, as stated earlier, it is important to show whether plo1 mutants have defects in SPB insertion into the nuclear envelope.


5


1st Revision - authors' response

30 September 2009

Thank you very much for giving us an opportunity of resubmitting our work. We took all the points raised by the three referees seriously and revised the manuscript accordingly, including a number of new experimentations. Please find attached the revised manuscript entitled "Fission yeast Pcp1 links polo kinase-mediated mitotic entry to -tubulin-dependent spindle formation" by Fong et al. (EMBOJ-2009-71857). Our revised manuscript comprises the text, 7 figures and 1 table (42,888 characters including spaces but without references) and Supplementary Information (a single pdf file that contains supplementary methods, one table and 7 figures, and 3 movies). Point-to-point responses to each referee’s comments are as follows. Referee #1 (Remarks to the Author): 1-1. Pg3 para1 "...SPB is involved in such multiple pathways" to "..SPB is involved in each distinct pathway." Corrected. 1-2. Pg4 ln4 Change "places" to "foci". Corrected. 1-3. Pg4 ln9 Change "...except that it is..." to "...except to identify that it is...". Corrected. 1-4. Pg4 ln11 Change "...roles in mitotic progression..." to "...roles during mitotic progression...". Corrected. 1-5. Pg4 ln14 Change "...have regulatory subunits that act as molecules responsible for targeting kinase catalytic subunits to specific cellular locations. Instead it is often the case that certain components localising to individual subcellular structures bind polo kinase, thereby loading it to these places (Elia et al, 2003)." To "...have regulatory subunits that act as molecules responsible for targeting kinase catalytic subunits to specific cellular locations. Instead it is Polo kinase has been shown to localize by associating with proteins already recruited to a distinct cellular position (Elia et al, 2003)." Corrected. 1-6. Pg4 last ln "Despite this distinct cell-cycle dependent Plo1 localisation, its spatial regulation remains elusive." I would disagree with this statement (See studies from Hagan lab for example which have gone some way to establish role Fin1 and Cut12 have in regulating Plo1 recruitment to the SPB). Please change text accordingly.


6


In response to this point, the following sentence was added. "its spatial regulation remains largely elusive except that the involvement of the SPB component Cut12 and NIMA-related kinase Fin1 has been reported (Grallert & Hagan, 2002; MacIver et al, 2003)." 1-7. Supplementary Figure S2: Need immunofluorescence images at 4 hr, so corresponds with biochemical data. Should discuss (in a sentence) significance of ability to localise but not function. We replaced images (B) with those of 4 h. We have discussed the significance of ability to localise but not function in page 5 ("we wanted to isolate mutant alleles that were specifically defective in distinct aspects of Pcp1 functions, rather than a complete loss-of-function mutant. To this end, ts mutants that exhibited Pcp1 delocalisation from the SPB at the restrictive temperature were first excluded, as these mutants were expected to mimic Pcp1 deletions") as suggested. 1-8. Pg6 ln4 "....which coincided with mitotic entry (light red columns in Figure 1B, C). This viability drop paralleled the emergence of lethal chromosome missegregation (blue lines)." Only refers to Figure 1 B NOT C. Chromosome missegregation is not necessarily lethal. Figure 1 C show cells with post anaphase array not mitotic spindle. Should show mitotic cells. If author keeps current Fig1C (in addition to requested images), should describe it in the text. As suggested, we changed "lethal" to "massive" and explained Figure 1C in the text (page 6, "in Figure 1B and post mitotic cells with missegregated chromosomes are shown in Figure 1C)."). We showed mitotic phenotypes of pcp1 mutants (live images) in Figure 1D. 1-9. Figure 1D Spindle is seen to associate with both SPBs (microtubules must be interdigitating as SPBs migrating apart around the nuclear membrane) in pcp1-18 mutant. I do not understand how this can occur if one of the two SPBs does not embed into the nuclear membrane (EM data or figure 2A). We believe that the apparent emergence of bipolar spindles (eg. 12 and 16 min in pcp118), which this referee pointed out, is in fact the plus end of mono-orientated spindle that emanated from the opposite pole and that this microtubule tip appeared to contact with the other pole. The referee might notice that spindle intensities in the vicinity of this pole were in fact rather dim. If this spindle were bipolar, we should expect a much stronger signal like that seen around the other pole or in equivalent images in wild type cells (eg. please see 0-4 min images in the top three panels). It should be noted that this type of mono-polar spindle phenotypes with two separated SPBs was reported by Iain Hagan’s group (Bridege et al., 1998; Tallada et al. 2009) and one SPB in cut12.1 mutants fails to be inserted exactly like pcp1-18 cells. 1-10. Pg6 ln6 Finding is consistent with recent study of Tallada et al. Sentence to discuss. Work by Tallada et al. (2009) was cited and discussed in detail in the Discussion (page 16). 1-11. Figure 3 (& S3) Should also show images of Alp14 & Alp16 localisation in a pcp1-15 and pcp1-18 mutant with 2 separated SPBs (with appropriate colabelled SPB marker as control). I guess that the referee meant Alp4 and Alp6, not Alp14 and Alp16. We showed new


7


Alp4 images that contained two SPBs (Figure 3A). As for Alp6, this figure is shown as supplementary and we kept our original figure (Supplementary Figure S3). 1-12. Pg8 ln7 Change "..address the significance of γ-TuC localisation for Pcp1 function,..." to "..examine how Pcp1 function affects γ-TuC localisation,..." Corrected. 1-13. Figure 4 B Show wt control in same axes. Would be useful if authors could differentiate between Alp4 on 1 SPB or 2 SPBs. As requested, we showed wild type patterns as a control (left). Alp4 signals in pcp1-15 started to disappear upon mitotic entry, when one SPB was observed and remained undetected (see Figure 3A). Therefore in principle no SPB signals were observed in mitotic pcp1-15 cells whether cells contained one or two SPBs. 1-14. Figure 5 (with 3 & S3). Why does the gamma tubulin complex only affect the pcp1-15 and not the pcp1-18 allele if in the latter one of the SPBs does not enter into the nuclear membrane. Is the gamma tubulin complex on each SPB in cells in which these organelles are separating SPBs? This is an interesting point. As the -TuC localises to the intra-nuclear region underneath the cytoplasmic SPB even during interphase, we envision that this is the reason that in the pcp1-18 mutant, the -TuC appears to localise to both SPBs. 1-15. Pg9 ln14 "..Pcp1 is required for γ-TuC recruitment to the mitotic SPB, but also suggest that γ-TuC recruitment during interphase is mediated by a Pcp1- and Mto1-independent echanism" The authors need to expand on how this occurs when the SPB is outside the nuclear membrane in these cells. As suggested, we added the following sentence in the text (page 9), "Given that centromeres are clustered in close proximity to the SPB during interphase (Funabiki et al. 1993), -TuC might be tethered to this location via interacting with some kinetochore components (Appelgren et al., 2003)." 1-16. Figure 6 A: Appears like there is an extra accumulation of fragmented SPB (or Cut11 anyway) in the pcp1-18 mutant. We think this is related to NE herniation that was observed at this site (Figure 6A, marked with arrowhead). This point is mentioned in Figure legend (page 35). D: Should mix cells and observe simultaneously to compare intensity of Plo1 signal in pcp1-15 and pcp1-18 cells. One would assume that Plo1 is only recruiting to the "old" or "maternal" SPB in pcp1-18 mutant. Show cells with 2 distinct SPBs and appropriate markers to examine this. As requested, we performed a mixing experiment and the result is shown in Supplementary Figure S5. It is clear that only pcp1-18 cells lost Plo1-GFP signals. Plo1 recruitment to the SPB is reduced at both mother and daughter SPBs (see the third panel in new Figure 6F). It is not clearly understood at the moment why the daughter SPB is not inserted into the NE, while the mother SPB manages to be inserted, albeit defectively, into the NE. We propose that this asymmetrical phenotype is due to the intrinsic difference between the mother and the daughter SPB, as suggested in recent work by Tallada et al. (2009).


8


1-17. Pg11 & Figure 6F If the stf1.1 mutant bypasses requirement for Cdc25, enhances Plo1 recruitment to the SPB and its overexpression suppresses pcp1-18, logic would dictate it may also bypass the requirement for Pcp1 (/ alleviate the effect of the pcp1-18) for Plo1 SPB recruitment. Therefore authors should address whether Plo1 is on one / two SPBs in stf1.1 pcp1.18 cells? As requested, we examined Plo1 localisation in pcp1.18cut12.s11 (stf1.1). As shown in new Figure 6G and H, Plo1 recruitment to the mitotic SPB is restore in the double mutant. This fact is described in the text, page 12, "Furthermore it was found that Plo1 is recruited to the SPB in pcp1-18cut12.s11 cells (Figure 6H and G)." and discussed in page 17, "although the suppression of Plo1 recruitment to the SPB by cdc12.s11 or multicopy cut11+ implies more complicated scenario such as a positive feedback loop rather than a simple linear pathway.".

Referee #2 (Remarks to the Author): 2-1. Weak points to address: • Claims in Introduction that the roles for Pcp1 have not been characterized for cell division and MT formation - but Rajagopalan S. et al (2004, Curr. Biol. 14:69-74) does characterize its role in the spindle orientation checkpoint. Furthermore, there is important pertinent information on the Cdc5 and Spc110 genes in budding yeast that should be presented in the paper, particularly in the Discussion for comparison to the new data presented in this manuscript. As suggested, we cited the paper by Rajagopalan et al. (page 4). Regarding budding yeast Cdc5 (polo kinase) and Spc110, as far as we are aware, no papers directly linking these two molecules have been published. We might have missed some references. We would be happy to refer to work if the referee lets us know it. • The claim that pcp1-18 and pcp1-15 identified different functions of the protein would be well supported by intragenic complementation. Do these alleles show intragenic complementation? We performed this experiment. Unfortunately no intragenic complementation was observed (see Supplementary Note 1 and Figure S7A). 2-2. • A similar genetics question. Do the authors find synthetic lethality in double mutant strains containing pcp1-18 or pcp1-15 and a mutant allele of an interacting gene. Such interactions would support the suppression data and may support the separation-of-function argument. As requested, we performed genetic analysis between pcp1-15 or -18 and mutants of genes that functionally interact with Pcp1 that are identified in this study. We found that both pcp1-15 and pcp1-18 are synthetically lethal with cut11.1 or cut12.1. Regarding alp4-1891, alp6-719 and plo1-ts19, these mutants are not synthetically lethal with pcp1-15 or pcp1-18, however in all cases, restrictive temperatures are lowered. These results are shown and summarised in Supplementary Note 2 and Figure S7B and C. We envision that the reason that we do not see allele-specificity between pcp1-15 and pcp1-18 is as follows. As either -TuC or Plo1 is essential for viability, double mutants of each single pathway (eg. pcp1-15 and plo1.ts9 or pcp1-18 and alp4-1891) or the same pathway (eg. pcp1-15 and alp4-1891 or pcp1-18 and cut11.1) would show synthetic phenotypes. We therefore believe that these genetic analyses are consistent, rather than inconsistent, with our results. We discuss this point in the Discussion (page 17), "We have addressed whether it would be possible to genetically differentiate distinct functions of Pcp1 using pcp1-15 and pcp1-18 mutants, but so far these two mutants alleles


9


behave similarly in terms of non-complementation and synthetic lethal interactions with mutations in other related genes (see Supplementary Notes 1 and 2 and Figure S7)." 2-3. • Supplementary figure 1 Supplementary figure 1 should be moved to the figures in the paper or and the specific mutations made in the 2 alleles should be discussed. There is plenty of structure/function work on Spc110 in budding yeast to make the comparison worthwhile. Mutation sites identified in pcp1-15 and pcp1-18 are unfortunately not that useful except that pcp1-15 contains one mutation in the proposed -TuC binding region (NI172T, discussed in page 15). Thus we would like to keep this figure as Supplementary Figure S1. Regarding mutations in budding yeast Spc110, unfortunately the central coiled coil regions in which mutations in pcp1-15 and pcp1-18 are clustered are excluded from mutant analysis (eg. Sunberg and Davis, 1997). Therefore we could not compare directly mutants of Pcp1 and Spc110. Instead we added coiled coil scores as Supplementary Figure 1SB that display how secondary structures are disrupted in Pcp1-15 and Pcp1-18 mutant proteins by point mutations compared to wild type Pcp1. It is true that the temperature-sensitive spc110-221 allele is reported to be rescued by overproduction of Spc98 (Alp6 homologue) (Ngyugen et al., 1997). We accordingly cited this work in the text (page 15), "In budding yeast, the temperature sensitive spc110-221 allele that carries multiple mutations in the N-terminal region was rescued by overproduction of Spc98 (Alp6 homologue, Nguyen et al, 1998), which might be in parallel with our finding described in this study.". 2-4. • Figure 1 When characterizing mutant effects on viability, the authors do not show or discuss any effects on cell cycle progress. Are there any delays in G1 or other parts of the cell cycle? We discussed mitotic delay in both pcp1 mutants (pages 5 and 6), "In addition live image analysis showed that mitotic progression was delayed in both pcp1 mutants (~15 min in wild type vs > 30 min in pcp1 mutants, Figure 1D).". We do not find any other abnormal cell cycle progression except for mitotic stage. Fission yeast has contains only a very short G1 and it is usually cryptic. 2-5. • Figure 1 - movies: It looks like pcp1-15 nucleates MTs from one pole and then the second pole later on, is this correct? If so, the phenotype is more complex than reported. Thank you very much for pointing this out. The images shown are not the best representatives. We now replaced these pictures with another series of live images, in which monopolar spindle phenotypes could be clearly observed. 2-6. • Figure3 Can low levels of gamma tubulin complex protein explain why the mother can still nucleate MTs even without Alp localization? Low levels (but undetectable under our microscopy system) of gamma tubulin could be one of the reasons. Alternatively, it could be due to the intrinsic difference between mother and daughter SPBs that we yet to understand (also suggested by Tallada et al). We discussed this point in the text (the second paragraph in page 8). 2-7. • Figure 4 The phenotype seems less prominent than Figure 3 - Alp4 decreases by


10


less than half in the Fig. 4 experiment. Is there an explanation for this result? In fact quantitative data between Figures 3 and 4 are very similar, if not identical. In Figure 3D, we saw 50% reduction of Alp4-GFP signals, whilst in Figure 4B, the value is also ~50% reduction (at 120 min time point). 2-8. • Figure 7 The IPs experiments are weak because there is no control showing that the Pcp1 IP is in anyway specific for Plo1 and gamma-tubulin. It seems possible that the IP is pulling down most of the SPB. If so, that could be a great reagent for this group, but not very useful in this context. At least in fission yeast the purification of SPBs as a soluble form is notoriously difficult (eg. see Rosenberg et al., Mol. Biol. Cell 2006, 17, 3793-3805). We believe that co-IP shown in this figure is indeed consistent with other genetic results. 2-9. • Figure 7 The model suggests an unusual localization for Plo1 in the SPB. Is there any data to support this localization? We do not understand why the referee said Plo1 localisation is "unusual". It localises to the nuclear side of mitotic SPB in close proximity with Cut12 and Pcp1. Minor points: 2-10. • The first sentence of the Introduction is awkward. Changed to "The centrosome was classically defined as the primary microtubule organising centre (MTOC) of the cell.". 2-11. • First section of results - when discussing pcp1 mutants, it refers to S2. It should refer to S1. We believe the current description is correct. 2-12. • On pages 6 & 7, the description of the monopolar spindles would be more clear if some of the use of "spindle" was replaced with "spindle microtubules." Corrected. 2-13. • Figure 1 Put %s in figures 1C and 1D on the figure of what portion of cells behave this way, also Figure2 put %s on figure % s were Added. 2-14. • Figure 2 How are they sure they are in the same mitotic phase in 3 cells? Also, looks like an elongation defect of the existing MT. Cdc7 localises to both SPBs in early mitosis, but only to the daughter SPB in later stages of mitosis. The fact that only one Cdc7-GFP signal is detected in these cells indicates that all cells are in a later stage of mitosis. Spindle microtubule elongation defect is expected in the mutants, because proper bipolar spindle are not formed in pcp1 mutants. Referee #3 (Remarks to the Author):


11


3-1) Page 6 Although it is stated that 80% of pcp1-15 and pcp1-18 mutants have monopolar spindles, the number of cells counted needs to be stated. This is needed throughout the manuscript. As suggested, the number of samples analysed is now shown (Figures 1B, 1D, 2A, 2B, 3B, 3D, 4B, 6A, 6C, 6D and 6G). In addition, from the images shown in Figure 1D, it looks like both mutants form bipolar spindles that collapse. This phenotype seems to be ignored, and merits some discussion of how this fits with the models for what is wrong in each mutant allele. The same point as 1-9. 3-2) Quantification is needed for the phenotypes described in Figure 2A, and the number of cells counted for the graph in 2B needs to be shown. Added. 3-3) The number of cells counted for the graph in 3B needs to be indicated. Quantification is needed for the phenotypes described in Figure 3C. Added. 3-4) A major conclusion of the paper is that SPB insertion into the nuclear envelope is defective in the pcp1-18 mutant. However, this conclusion appears to be based on 2 electron micrographs, one of which is not very convincing. The SPB only inserts into the nuclear envelope during mitosis, so it is crucial to show that cells are in mitosis when looking for a failure in SPB insertion into the nuclear envelope. The image in figure 5E is pretty convincing, but the one in figure 5D is not. The authors state that the cell in figure 5D is in mitosis because of the nuclear envelope protrusion, which they say is caused by the ends of a monopolar spindle pushing on the nuclear envelope, which is completely possible. However, it is also possible that the cell is actually in interphase, and the protrusion is caused by forces generated by cytoplasmic microtubules attached to the nuclear envelope. If the cell is in mitosis, then microtubules from the monopolar spindle should be apparent in the protrusion. In any case, it is important to make clear whether it was concluded that the pcp1-18 mutant has an SPB insertion defect is based on just these 2 images, and if not, do the authors have a sense of how frequent the defect occurs. Sample number is 5, in which 4 samples showed very clear insertion defects, which is now described in the legend for Figure 5. Regarding nuclear membrane protrusion, it is true that in fission yeast cytoplasmic microtubules physically push the nuclear membrane, by which the nucleus localises in the centre of the cell during interphase. However in no circumstances nor published data, this type of severe NE herniation has been observed during interphase. Instead there are a number of examples in abnormal mitotic cells (eg. Toya et al., 2007, Nat. Cell Biol. 9, 646653). Furthermore the picture shown in Figure 6A (the third raw) clearly showed that cells containing NE herniation are indeed in mitosis, as two Cut11-SPB dots are visible. Visualisation of spindle microtubules in NE herniation by EM is extremely difficult, as the angle of microtubules often changes (eg. see Toya et al. 2007). 3-5) It would be interesting to know more about how the multicopy (cut11, pom152, cut12, kms2, tcg1) and genetic (cut12-s11, wee1-50) suppressors of pcp1-18 rescue the mutant phenotype. For example do the suppressors enhance recruitment of Plo1 to the SPB, or do they rescue the defect in SPB insertion into the nuclear envelope. Similarly, since


12


pcp1-18 has defects in both Plo1 recruitment to the SPB and in SPB insertion into the nuclear envelope, do plo1 mutants have defects in SPB insertion? The similar point as 1-17. As requested, we observed Plo1-GFP signals in pcp1-18 that carries multicopy suppressors cut11+. For cut12, we used cut12.s11. As shown in Figure 6G and H and Supplementary Figure S6, overproduction and activation of Cut11 and Cut12, respectively, restore Plo1 localisation to the SPB. We feel that Kms2, Pom152 and Tcg1 are beyond the scope of this paper. As for wee1-50, this mutant is very small and technically difficult to handle in cell biology. For this reason, we could not do this experiment. Observation of suppression of SPB insertion defect in EM is extremely timeconsuming and laborious. We hope that the referee would appreciate our EM pictures presented in this work. We believe that SPB insertion would occur normally in these suppressed strains, otherwise proper bipolar spindles would not be formed and consequently chromosome segregation and cell division would not take place. Regarding Plo1 mutants and SPB insertion defects, it is known that Plo1 plays myriad roles in mitotic progression other than spindle formation, and therefore phenotypes of plo1 mutants would be expected to be very pleiotropic and their interpretations not to be straightforward. Thus we feel that this is the future issue. 3-6) There are a number of places in the discussion that should be worded a bit more precisely. For example, at the top of page 14, it is stated that Pcp1 activates Plo1 kinase at the SPB, but no evidence for this is presented in the paper. At the bottom of page 14 it is stated that Pcp1 is required for Plo1 recruitment to the SPB. Given the data presented, it would be more accurate to state that Pcp1 plays a role in recruitment of Plo1 to the SPB. On page 15, it is stated that SPB insertion into the nuclear envelope is under the control of Plo1. No evidence for this was presented, therefore, as stated earlier, it is important to show whether plo1 mutants have defects in SPB insertion into the nuclear envelope. Changed as suggested.

2nd Editorial Decision

15 October 2009

Thank you for submitting your revised manuscript. It has now been seen once more by the original referees 2 and 3, and I am pleased to inform you that both of them consider your study significantly improved and would therefore in principle now support publication in The EMBO Journal. There remain, however, a few issues that would still need to be cleared up before we can proceed with acceptance and publication of the study - two minor points raised by referee 2 (see comments below), as well as a more significant one raised by referee 3. As it stands, this reviewer remains unconvinced that the proposed Polo kinase role nuclear envelope disassembly and SPB insertion is sufficiently supported by the presented data, and suggests additional experimentation to demonstrate this. My feeling is that addressing this point, which was not raised in that form during initial review, experimentally is likely beyond the scope of a final revision round here, but I would in this case nevertheless ask you to alternatively ensure that any conclusions/interpretations in this respect are valid or, where necessary, carefully toned down. I would thus like to ask you to incorporate these additional requested changes in a last round of revision. Once we will have received this final version, we should then be able to proceed with the acceptance of your paper. I am looking forward to receiving your final version.


13


Yours sincerely, Editor The EMBO Journal _____ REFEREE REPORTS: Referee #2 (Remarks to the Author): Fong et al. have provided a very good revision of their manuscript reporting evidence that the fission yeast pericentrin-like Pcp1 regulates multiple functions of the SPB. It does this through recruitment of two important proteins to the SPB, the gamma tubulin complex and polo kinase (plo1). This study provides new insight into how the SPB interacts with several pathways and further characterizes a relatively newly identified protein. The authors carefully and thoroughly addressed most of reviewer's comments and the manuscript has benefited from their effort. I am disappointed that they did not attempt to demonstrate the specificity of the co-IP experiments as suggested: ï "Figure 7 - The IPs experiments are weak because there is no control showing that the Pcp1 IP is in anyway specific for Plo1 and gamma-tubulin. " However, this is not a significant enough problem to preclude publication. Nonetheless, there are two minor points that should be corrected. Minor points: ï The first sentence of the second complete paragraph on page 8 beginning with "To substantiate our..." is awkward if not incorrect and should be fixed. ï Despite the authors claim otherwise, there is a little evidence that budding yeast Cdc5 binds Spc110. See figure 4 of Paulson et al. (2007) A Coupled Chemical Genetic and Bioinformatic Approach to Polo-like Kinase Pathway Exploration. Chem. Biol. 14:1261-1272. This paper should be cited, and it supports the case for the Pcp1/Plo1 interaction. Referee #3 (Remarks to the Author): Although the authors have largely satisfied my concerns, one major point still needs to be addressed. The author's results imply, and they show in their model, that Polo kinase is required to promote localized nuclear envelope disassembly and insertion of the SPB into the nuclear envelope. However the authors provide no evidence to support this conclusion. This would be a major finding of potential broad interest beyond the yeast community if they were able to support this conclusion. The obvious way to address this point is to do electron microscopy in Polo mutant cells.

2nd Revision - authors' response

16 October 2009

Thank you very much for your encouraging letter. Please find attached the revised manuscript entitled "Fission yeast Pcp1 links polo kinase-mediated mitotic entry to -tubulin-dependent spindle formation" by Fong et al. (EMBOJ-2009-71857). Upon your and reviews’ requests, we have changed descriptions in the text carefully to fully satisfy outstanding concerns. Our revised manuscript comprises the text, 7 figures and 1 table (42,698 characters including spaces but without references) and Supplementary Information (a single pdf file that contains supplementary methods, one table and 7 figures, and 3 movies). Point-to-point responses to each referee’s comments are as follows.


14


Referee #2 (Remarks to the Author): 1-1. The first sentence of the second complete paragraph on page 8 beginning with "To substantiate our..." is awkward if not incorrect and should be fixed. We agree that this sentence is a bit odd. In response to this point, we have deleted the latter part of this sentence ("and further examine how Pcp1 function affects -TuC localisation for Pcp1 function"). By this correction, we believe that the sentence is more readable. 1-2. Despite the authors claim otherwise, there is a little evidence that budding yeast Cdc5 binds Spc110. See figure 4 of Paulson et al. (2007) A Coupled Chemical Genetic and Bioinformatic Approach to Polo-like Kinase Pathway Exploration. Chem. Biol. 14:1261-1272. Thank you very much for letting us know this reference. We mentioned this paper {page 13, Interestingly budding yeast Spc110 is also reported to interact with both -TuC and Cdc5 (budding yeast polo kinase) (Knop & Schiebel, 1997; 1998; Snead et al, 2007)}.and cited Snead et al. (2007) in the reference. Referee #3 (Remarks to the Author): Although the authors have largely satisfied my concerns, one major point still needs to be addressed. The author's results imply, and they show in their model, that Polo kinase is required to promote localized nuclear envelope disassembly and insertion of the SPB into the nuclear envelope. However the authors provide no evidence to support this conclusion. This would be a major finding of potential broad interest beyond the yeast community if they were able to support this conclusion. The obvious way to address this point is to do electron microscopy in Polo mutant cells. We have rephrased several sentences in the text in order not to give wrong impressions that we insist that polo kinase (Plo1) directly drives nuclear envelope disassembly and insertion of the SPB into the nuclear envelope. Page 14: the Plo1-dependent process is responsible for NE invagination and subsequent SPB insertion. To Pcp1 interacts with Plo1 and is responsible for NE invagination and subsequent SPB insertion. Page 16 We added the following sentence. It would be of interest to examine whether plo1 mutants display similar SPB insertion defects. Page 36: Note that activated Cdc2 in turn activates Plo1 via a positive feedback loop. This drives NE reorganisation and SPB insertion. To Note that activated Cdc2 in turn activates Plo1 via a positive feedback loop. This drives directly or indirectly NE reorganisation and SPB insertion.


15