Pathogenic mutations in the chromokinesin KIF22 disrupt anaphase chromosome segregation
This article has been Reviewed by the following groups
Listed in
- Evaluated articles (ASAPbio crowd review)
Abstract
The chromokinesin KIF22 generates forces that contribute to mitotic chromosome congression and alignment. Mutations in the α2 helix of the motor domain of KIF22 have been identified in patients with abnormal skeletal development, and we report the identification of a patient with a novel mutation in the KIF22 tail. We demonstrate that pathogenic mutations do not result in a loss of KIF22’s functions in early mitosis. Instead, mutations disrupt chromosome segregation in anaphase, resulting in reduced proliferation, abnormal daughter cell nuclear morphology, and, in a subset of cells, cytokinesis failure. This phenotype could be explained by a failure of KIF22 to inactivate in anaphase. Consistent with this model, constitutive activation of the motor via a known site of phosphoregulation in the tail phenocopied the effects of pathogenic mutations. These results suggest that the motor domain α2 helix may be an important site for regulation of KIF22 activity at the metaphase to anaphase transition. In support of this conclusion, mimicking phosphorylation of α2 helix residue T158 also prevents inactivation of KIF22 in anaphase. These findings demonstrate the importance of both the head and tail of the motor in regulating the activity of KIF22 and offer insight into the cellular consequences of preventing KIF22 inactivation and disrupting force balance in anaphase.
Article activity feed
-
-
-
Author response
We thank the reviewers who provided feedback on our manuscript through ASAPbio Crowd Review. We have revised the preprint based on their thoughtful commentary. Specifically, reviewers asked whether in Figure 3 polar ejection force levels depend on the expression level of wild type or mutant KIF22-GFP. We measured KIF22-GFP expression level in these cells and now include plots indicating that polar ejection forces are not correlated with expression levels of KIF22-GFP (Figure 3E and 3F). Our manuscript also includes a plot demonstrating that anaphase recongression is correlated with KIF22-GFP expression level (Supplemental Figure 3A). In addition to adding data to address this question raised by the reviewers, we have revised the text to clarify several points based on reviewer comments, including the efficiency of siRNA …
Author response
We thank the reviewers who provided feedback on our manuscript through ASAPbio Crowd Review. We have revised the preprint based on their thoughtful commentary. Specifically, reviewers asked whether in Figure 3 polar ejection force levels depend on the expression level of wild type or mutant KIF22-GFP. We measured KIF22-GFP expression level in these cells and now include plots indicating that polar ejection forces are not correlated with expression levels of KIF22-GFP (Figure 3E and 3F). Our manuscript also includes a plot demonstrating that anaphase recongression is correlated with KIF22-GFP expression level (Supplemental Figure 3A). In addition to adding data to address this question raised by the reviewers, we have revised the text to clarify several points based on reviewer comments, including the efficiency of siRNA knockdown of KIF22, the absence of identified pathogenic mutations in P148 or R149 in other members of the kinesin superfamily, and the potential role of mutations in altering phosphoregulation of T463. We appreciate the interest of the reviewers in the use of a chondrocyte cell line or mouse model to study how mutations in KIF22 result in tissue-specific pathology in patients. We agree that these questions are important for follow-up studies.
-
-
This review reflects comments and contributions by Julia Grzymkowski, Jake Herman, Yogaspoorthi Subramaniam, and Vladimir Volkov.
Alex Thompson and colleagues characterize the molecular defects of four previously documented and one newly described disease-associated mutation in the human KIF22 chromokinesin. They describe the patient’s history and go on to study the function of this kinesin in mitosis by expressing known pathogenic mutations of KIF22 in HeLa and RPE-1 cells.
After rejecting several hypotheses, the authors conclude that the mutations affect the fidelity of anaphase by an excessive polar ejection force. This work provides compelling evidence that the mutations prevent the normal silencing of this chromosome-associated kinesin motor during the final step of chromosome segregation. This failure to silence KIF22 results in …
This review reflects comments and contributions by Julia Grzymkowski, Jake Herman, Yogaspoorthi Subramaniam, and Vladimir Volkov.
Alex Thompson and colleagues characterize the molecular defects of four previously documented and one newly described disease-associated mutation in the human KIF22 chromokinesin. They describe the patient’s history and go on to study the function of this kinesin in mitosis by expressing known pathogenic mutations of KIF22 in HeLa and RPE-1 cells.
After rejecting several hypotheses, the authors conclude that the mutations affect the fidelity of anaphase by an excessive polar ejection force. This work provides compelling evidence that the mutations prevent the normal silencing of this chromosome-associated kinesin motor during the final step of chromosome segregation. This failure to silence KIF22 results in chromosome segregation failures, abnormal spindle morphology, and proliferation defects.
This study was executed with robust controls and the conclusions presented are clearly represented by the data. A majority of the phenotypes were assayed in multiple cell lines and useful 'negative' results were reported (e.g. mutations do not alter the dynamic nor steady state localization of the protein). The imaging data are beautiful and sufficiently explained, the data is striking. The paper is well written. Altogether the work sheds light on important fundamental and medical questions.
There is one conceptual question which may be relevant to address to strengthen the interpretations. Mutant KIF22 constructs are introduced via two inducible pathways in the background of siRNA depletion of the endogenous KIF22, and the resulting expression levels of the mutants exceed the endogenous KIF22 by 2-3 fold, as judged by immunofluorescence. The polar ejection forces are then estimated by treating cells with monastrol and quantifying the distance between the spindle monopole and the chromosome ring around it. However, even expression of wild-type KIF22 led to elevated polar ejection force (Fig 3D), raising a question about whether overexpression of KIF22, rather than mutations, produces this phenotype. Is it possible to modulate the expression levels to address this question? At least, can the levels of KIF22-GFP in individual cells be correlated to the pole-chromosome distance in this experiment? The same question applies to the description of chromosome recongression phenotypes.
General comments:
- How complete is siRNA depletion of endogenous KIF22? Can a western blot be provided to test this?
- In the FRAP experiments, which regions were photobleached at different cell cycle stages?
Introduction
The introduction could feature more about the molecular aspect of KIF22 mutations and shorten the paragraph(s) about the disease pathologies.
‘These mutations occur in adjacent residues P148 and R149 in the α2 helix of the KIF22 motor domain (Figure 1B). P148 and R149 are conserved in kinesin-10 family members across species (Figure 1C) and in many human members of the kinesin superfamily (Figure 1D).’ - In addition to the mention in Figure 1A, it would be useful to describe the specific four KIF22 mutations in the text. Have mutations to the paralogous Pro and Arg residues in other kinesin proteins been correlated with pathologies?
‘However, KIF22 knockout in mice did not affect skeletal development’ - Often, a complete KO of a gene is relatively less/non-detrimental than a dominant mutant ascribing to its (im)proper functions. Given that, comparison with mutant mice may be more relevant than comparisons with KIF22 KO mice.
Results
‘(RPE-1) cell lines expressing wild type and mutant KIF22-GFP to assess any differences between the consequences of expressing mutant KIF22 in aneuploid cancer-derived cells (HeLa-Kyoto) and genomically stable somatic cells.’ - Based on the introduction, mutations in the KIF22 motor domain affect skeletal development severely and selectively. Epithelial cells such as HeLa or RPE may not fully represent the outcomes of KIF22 mutations, would usage of a chondrocyte cell line be relevant?
‘KIF22-GFP with pathogenic mutations demonstrated the same localization pattern throughout the cell cycle as wild type motor’ - Over-expression of KIFF22-GFP is several folds higher compared to the nascent content of KIFF22 in cells and the localization pattern remains unaltered. What does this imply with respect to force generated by excessive motor loading on microtubules and associated chromosomes?
‘Relative polar ejection forces were compared by measuring the distance from the spindle pole to the maximum DAPI signal (Figure 3A). Expression of mutant motor did not reduce polar ejection forces (Figure 3B and 3C).’ - Figure 3A is incredibly helpful and clear for interpreting the data. In Figure 3B, it is difficult to see overlap between GFP and DAPI when using green and blue, since the DAPI is already shown to the left, it may be more helpful to show only Centrin and GFP. In Figure 3B, is GFP signal similar between different mutants?
‘Together, the localization of mutant KIF22 and the ability of mutant KIF22 to generate polar ejection forces indicate that pathogenic mutations P148L, P148S, R149L, R149Q, and V475G do not result in a loss of KIF22 function during early mitosis.’ - Quantifying rescue phenotypes in terms of either force generated by them (activity) or protein content may be relevant to support this conclusion.
‘This phenotype was dominant and occurred in the presence of endogenous KIF22’ - In the introduction KIF22 mutants are described as dominant because they cause disease phenotypes as heterozygotes. Then the anaphase recongression and nuclear morphology phenotypes are described as dominant because they are observable without siRNA treatment. Recommend not applying the same term to both conditions because they are not equivalent. The outstanding characterization in Supplement 1 shows that ectopic KIF22 is expressed in excess of endogenous KIF22. Ectopic copies may also silence or outcompete endogenous KIF22 because ectopic KIF22 shows no decrease in mitotic localization when endogenous KIF22 is depleted (S1C and S1E). This is not a comment on the data but rather how they are discussed.
‘Additionally, the distance between chromosome masses at the time of cleavage furrow ingression was reduced in cells expressing KIF22-GFP R149Q or V475G, suggesting that the position of the chromosome masses may be physically obstructing cytokinesis’ - Does this mean that karyokinesis failed too, as the chromosome masses fail to sufficiently segregate?
‘was imaged in HeLa-Kyoto cells expressing fluorescent markers for the poles (pericentrin-RFP) and centromeres (CENPB-mCh) (Figure 5A)’ - It is difficult to see the centromere staining, can it be false colored to a different color than the poles?
‘Since expression of KIF22-GFP T463A does not cause anaphase recongression (Figure 8E), the level of compaction of the segregating chromosome masses was explored as a possible explanation for this modest increase in the percentage of cells with nuclear morphology defects.’ - How does KIF22A protein conformation influence availability/accessibility of T463 to phosphorylation by CDK1/cyclin B? head-tail auto-inhibition, does it occur due to masking of the T463 site?
-