Maternal age-dependent APC/C-mediated decrease in securin causes premature sister chromatid separation in meiosis II

Sister chromatid attachment during meiosis II (MII) is maintained by securin-mediated inhibition of separase. In maternal ageing, oocytes show increased inter-sister kinetochore distance and premature sister chromatid separation (PSCS), suggesting aberrant separase activity. Here, we find that MII oocytes from aged mice have less securin than oocytes from young mice and that this reduction is mediated by increased destruction by the anaphase promoting complex/cyclosome (APC/C) during meiosis I (MI) exit. Inhibition of the spindle assembly checkpoint (SAC) kinase, Mps1, during MI exit in young oocytes replicates this phenotype. Further, over-expression of securin or Mps1 protects against the age-related increase in inter-sister kinetochore distance and PSCS. These findings show that maternal ageing compromises the oocyte SAC–APC/C axis leading to a decrease in securin that ultimately causes sister chromatid cohesion loss. Manipulating this axis and/or increasing securin may provide novel therapeutic approaches to alleviating the risk of oocyte aneuploidy in maternal ageing.

Loss of Securin protein in MII oocytes: Securin levels in MII oocytes are very low, therefore western blots may be close to detection level, and small variations in protein levels may look much more important. The western blot in MII oocytes ( Figure 2) was only done twice and should repeated a third time to be sure that this is not due to fluctuations in protein detection. Alternatively, are background bands visible on the western blot at around the same size as Securin, which show that there is indeed no difference in blotting efficiency at that molecular size?
As far as I know the inhibitor AZ3146 has not been used in mouse oocytes to inhibit Mps1. The authors should show that the inhibitor indeed overrides the checkpoint in meiosis I, to be sure that the drug is functional to inhibit Mps1 in oocytes. Alternatively the authors should cite references where this drug has been used in oocytes.
In the discussion the authors state that Securin does not re-accumulate after meiosis I and that low levels of Securin remaining in meiosis I oocytes are important for Separase inhibition in meiosis II. As far as I know it has not been shown that remaining Securin levels in meiosis II are indeed "left over" from meiosis I and not newly translated as oocytes enter meiosis II. Even though levels are low they may have been newly synthesized.
Minor point: Figure 2, numbering is not correct in the text (page 7) Reviewer #3 (Remarks to the Author) This manuscript investigates one mechanistic aspect of premature sister chromatid separation in meiosis II of mouse oocytes. The major findings support the conclusion that old oocytes have lower levels of securin protein in MII, which is at least in part causative of premature sister chromatid separation. Further data indicate that the kinetics of securin degradation during anaphase I dictate the level of securin in MII and that the Mps1 kinase partly controls the kinetics.
The conclusions are interesting and this is an important topic. Mostly, the data support the conclusions but there are some issues that should be addressed before it would be appropriate to support publication. Main concerns: (1) Abstract. "This increased APC/C activity... can be normalized by...inhibiting Mps1." This is strangely worded. It is not increased securin degradation in the old eggs that is corrected by MPS1 inhibition -degradation actually gets faster in the old eggs with the inhibitor.
(2) General point. Through the manuscript they claim the data show APC/C activity is higher in the old eggs during anaphase of MI. This is not shown, just assumed based on more securin degradation. Clearly the data are consistent with the likely possibility that APC/C activity is different in the old eggs, but this is not directly shown. It remains possible that another aspect of securin degradation is affected, for example a securin post-translational modification, securin localization, proteasome regulation. Since other possible factors have not been ruled out, and there is no direct assessment of APC/C activity, the wording needs to be changed throughout to align with the data -i.e. they show that more securin is degraded; mechanism not resolved at this time. Comparing the old mice in 3a,c with or without the MPS1 inhibitor, there appear to be important differences in securin degradation kinetics: (i) increased lag (delayed initiation of degradation) with inhibitor; (ii) faster rate of degradation with inhibitor (the slopes of the curves are quite different in a and c). It is not clear how this translates to equal degradation rates reported in panels b,d. How can this be explained? It seems to me that Mps1 is still restraining the degradation of securin in the old oocytes; it just seems to be having a greater impact in the young oocytes.

Reviewer #1
The present manuscript by Nabti and colleagues aims at clarifying why oocytes from older mice missegregate sister chromatids at higher rates than younger mice. The authors found a decrease in Securin levels in oocytes from older mice, which is likely due to increased APC activity and thereby a failure to properly maintain Separase inactive during the meiosis II arrest. The authors propose that this leads to increased interkinetochore distances and precocious loss of centromeric cohesion leading to the observed missegregations.
Overall, the study is well executed and can be published once the following points have been addressed: The phrase that "APC/C activity can be normalized by Mps1 inhibition (abstract and page 5)" is misleading, because not old oocytes are somehow rescued to behave like young oocytes, but young oocytes show the same defects as old oocytes, upon Mps1 inhibition. I think this should be formulated differently. (for example: "SAC inhibition in young oocytes results in increased APC/C activity, such as observed in old oocytes") Response: Agreed, this is unclear. We have re-formulated the sentence as requested.
Loss of Securin protein in MII oocytes: Securin levels in MII oocytes are very low, therefore western blots may be close to detection level, and small variations in protein levels may look much more important. The western blot in MII oocytes (Figure 2) was only done twice and should repeated a third time to be sure that this is not due to fluctuations in protein detection. Alternatively, are background bands visible on the western blot at around the same size as Securin, which show that there is indeed no difference in blotting efficiency at that molecular size?
Response: There is a misunderstanding here. The Western blots in MII oocytes were repeated 3 times. The only blot that was done twice was the one using MI oocytes, which shows no change with age. We have now performed an additional replicate and the new data is included in the revised Figure 2. Unfortunately there are no background bands that can be used as controls for blotting efficiency. We achieve this through reblotting for actin and normalizing the securin bands against the actin controls.
As far as I know the inhibitor AZ3146 has not been used in mouse oocytes to inhibit Mps1. The authors should show that the inhibitor indeed overrides the checkpoint in meiosis I, to be sure that the drug is functional to inhibit Mps1 in oocytes. Alternatively the authors should cite references where this drug has been used in oocytes. Figures 3 and 4)).

Response: We have referenced the ability of AZ3146 to override the checkpoint in oocytes as shown by us previously (Nabti et al,. Dual-mode regulation of the APC/C by CDK1 and MAPK controls meiosis I progression and fidelity. J Cell Biol 204, 891-900, 2014 (see
In the discussion the authors state that Securin does not re-accumulate after meiosis I and that low levels of Securin remaining in meiosis I oocytes are important for Separase inhibition in meiosis II. As far as I know it has not been shown that remaining Securin levels in meiosis II are indeed "left over" from meiosis I and not newly translated as oocytes enter meiosis II. Even though levels are low they may have been newly synthesized. eggs. Dev. Biol. 321, 379-386 (2008)).

Response: We have rephrased this sentence. It is correct that we cannot state whether the securin in MII oocytes is 'left over' from MI or is newly synthesised in the MI-MII transition, or both. Our intention in this sentence is not to define whether the securin is newly synthesised, rather we are trying to make the simple point that securin is responsible for separase inhibition at MII as shown previously (Nabti et al,. Securin and not CDK1/cyclin B1 regulates sister chromatid disjunction during meiosis II in mouse
Minor point: Figure 2, numbering is not correct in the text (page 7) Response: Thank you, this has now been corrected.

Reviewer #3
This manuscript investigates one mechanistic aspect of premature sister chromatid separation in meiosis II of mouse oocytes. The major findings support the conclusion that old oocytes have lower levels of securin protein in MII, which is at least in part causative of premature sister chromatid separation. Further data indicate that the kinetics of securin degradation during anaphase I dictate the level of securin in MII and that the Mps1 kinase partly controls the kinetics. The conclusions are interesting and this is an important topic. Mostly, the data support the conclusions but there are some issues that should be addressed before it would be appropriate to support publication.
Main concerns: Abstract. "This increased APC/C activity... can be normalized by...inhibiting Mps1." This is strangely worded. It is not increased securin degradation in the old eggs that is corrected by MPS1 inhibitiondegradation actually gets faster in the old eggs with the inhibitor.
Response: Yes, agreed, we have re-formulated the sentence as requested. We were trying to say that in the presence of MPS1 Inhibitor securin destruction is similar in old and young eggs.
General point. Through the manuscript they claim the data show APC/C activity is higher in the old eggs during anaphase of MI. This is not shown, just assumed based on more securin degradation. Clearly the data are consistent with the likely possibility that APC/C activity is different in the old eggs, but this is not directly shown. It remains possible that another aspect of securin degradation is affected, for example a securin post-translational modification, securin localization, proteasome regulation. Since other possible factors have not been ruled out, and there is no direct assessment of APC/C activity, the wording needs to be changed throughout to align with the data -i.e. they show that more securin is degraded; mechanism not resolved at this time.
Response: Fair point -we have changed the wording as requested. In the discussion we consider any evidence supporting other possibilities as well as the supporting evidence for a role of the APC/C. As noted by the referee the balance of evidence supports the APC/C mediating the effect, particularly because manipulating the SAC generates the predicted response. However, definitive proof of an effect specifically on the APC/C would require a direct measure of the rate of ubiquitination and this is just not possible in the small numbers of mouse oocytes available. I trust this change in emphasis and extra explanation is in line with the reviewer's expectations. Response: Significance has been calculated and added to the figure   Fig 2. The blots are not entirely convincing and the exposures appear to be saturated in some cases. It is important to perform technical replicates of the blots and ensure that the signals are in the linear range, or else use a quantitative method that is completely linear. Three experimental repeats are needed and standard deviation should be shown.  It is important to confirm that the depth of field was sufficient to capture the entire securin signal in each egg/time-point.

Response
Response: Securin-GFP imaging was performed using conventional microscopy and not a confocal. We therefore collect fluorescence from the entire oocyte. Comparing the old mice in 3a,c with or without the MPS1 inhibitor, there appear to be important differences in securin degradation kinetics: (i) increased lag (delayed initiation of degradation) with inhibitor; (ii) faster rate of degradation with inhibitor (the slopes of the curves are quite different in a and c). It is not clear how this translates to equal degradation rates reported in panels b,d. How can this be explained? It seems to me that Mps1 is still restraining the degradation of securin in the old oocytes; it just seems to be having a greater impact in the young oocytes.