CEP44 is required for maintaining centriole duplication and spindle integrity

Reviewer #1 (Public Review):

Summary:
Zhang et al. describe novel roles for the centriolar protein CEP44, namely that it is required for centriole engagement (and thus inhibition of centriole reduplication) and that it promotes microtubule stability. While a function of CEP44 in centriole engagement is somehow convincingly shown, the data do not support a role for CEP44 in microtubule stabilization.

Strengths:
The finding that centriole engagement relies on CEP44 is novel and of great interest to the centriole field. Interestingly, the authors correlate reduced CEP44 expression levels with the occurrence of breast carcinoma, which makes this study also very interesting for a broad audience.

Weaknesses:
The paper has important findings, but unfortunately, the main claims are only partially supported.

1. The role of CEP44 in microtubule stability is not clear from the presented data:
   - Fig. 7A and S6 A, there is no visible difference in microtubule density/intensity between the different groups of cells. In Fig. 7C, the CEP44 S324A spindle looks even brighter than the WT spindle. The authors need to indicate how many cells were analyzed. This information is actually lacking in all the experiments.

2. Several figure parts are not properly labelled.

3. Several of the experiments (WBs) likely miss proper controls: How did the authors detect proteins that run at very similar sizes: 55 kDa (alpha-tubulin), 44 kDa (Cep44), and 57 kDa (Cep57 and Cep57L)? The loading control needs to be detected in the same lane as the protein of interest. Did the authors strip and reprobe membranes? If so, this needs to be indicated and included in the methods section.

4. It is not clear how such a low CEP44-FLAG expression (Fig. 5A) can rescue a CEP44 KO.

Reviewer #2 (Public Review):

Zhang and Wei, et al. investigated the role of a centrosomal protein, CEP44, in regulating centrosomes and spindle integrity, with a focus on processes that may be dysregulated in breast cancer. The authors found that a breast cancer cell line, MDA-MB-436, lacks CEP44 protein and has amplified centrioles. CEP44 expression is reduced in samples from breast cancer patients. By super-resolution microscopy, the authors localize CEP44 to the proximal inner lumen of centrioles, as has also been previously shown by another group (Atorino et al 2020). Next, the authors investigate the role of CEP44 in centrosome regulation. They found that loss of CEP44 in HeLa cells results in extra puncta of CEP97 or Centrin-3, while ectopic overexpression of CEP44 in MDA-MB-436 cells reduces the number of CEP97 foci. Only one of the excess puncta in a CEP44-depleted HeLa cell recruits CEP164 or ODF2, indicating that extra foci were not the result of cytokinesis failure. In G1, most (~80%) of CEP44-depleted cells have 2 centrin foci, while in G2, a small population (~20%) have more than 4 centrin foci, and gamma-tubulin is recruited in foci in G2. The authors were able to observe centriole disengagement and amplification using live cell imaging. The authors propose that CEP44 acts in regulating centriole engagement by recruiting CEP57 and CEP57L1 to centrioles. The authors made CEP44 knockout cell lines using CRISPR and found that loss of CEP44 results in multipolar spindles, correlated with an increase in centriole amplification. Finally, the authors investigate the role of CEP44 at the mitotic spindle. The authors find that CEP44 localizes to spindles and is phosphorylated by Aurora A at G2/M on Ser324. Phosphorylation of CEP44 is required for its proper distribution between centrosomes and the spindle and microtubule stability within both spindles and interphase microtubules. Together, these studies shed light on the roles of CEP44 within centrosomes and spindles and will be of interest to cell biologists and cancer biologists studying cell division and centrosomes.

The conclusions of this paper are only partially supported. The analyses could be improved to address the concerns about the major conclusions.

Reviewer #3 (Public Review):

Summary:
The manuscript by Zhang et al. analyzes the function of the centrosomal protein CEP44 in centriole duplication and in the formation of the mitotic spindle. The first part addresses the role of CEP44 at centrioles. Using mostly RNAi-mediated depletion in cell lines and in some cases KO cells, the authors find increased centriole numbers in depleted cells and, based on quantification of centrioles stained with various centriole markers as well as live imaging, conclude that this is due to premature centriole disengagement and overduplication. The second part, which is largely independent of the first, focuses on the role of CEP44 in the mitotic spindle. The authors find that CEP44 is phosphorylated in mitosis in an Aurora A-dependent manner and identify the phosphorylation site, which controls CEP44 spindle localization and functions in maintaining spindle integrity.

Strength:
The manuscript makes the interesting observation that reduced expression of CEP44 is observed in breast cancer and correlated with poor survival in patients.
The analysis of mitotic phosphorylation including the identification of the modified site and its role in spindle recruitment is interesting and useful.

Weakness:
The authors seem to largely ignore previously published work that contrasts with the findings presented in the current study. The previous work found a role of CEP44 in centriole formation and centrosome conversion and observed reduced centriole numbers in depleted cells, whereas the current study claims the opposite, a role in centriole engagement that leads to overduplication and increased centriole number in depleted cells. However, the supporting evidence is not strong enough, especially in light of the previous work. Considering that CEP44 depletion also disrupts mitosis, which could affect centriole numbers by failed segregation/division, a more careful analysis in synchronized cultures would be needed. Also, cell cycle analysis would be required to rule out cell cycle effects in CEP44-depleted cells, which could also explain altered centriole numbers. Moreover, the quality of the imaging is often not sufficient to support the claims.
The second part is largely disconnected from the first and reads as if it was a separate study. There is no attempt to integrate both parts. For example, the second part seems to largely focus on normal bipolar spindles, even though the first part reveals multipolarity as a phenotype after CEP44 knockdown. It remains unclear if the spindle defects are due to centriole defects, defective spindle microtubule stability/organization, or both, and whether the centriole-localized or spindle-localized CEP44 is involved.

Another weak aspect is that neither for RNAi nor for KO cells the authors show that CEP44 is depleted at centrioles and to what extent. This is only shown in cell extract.