Mitotic R-loops direct Aurora B kinase to maintain centromeric cohesion

Background:

During mitosis, chromosomes undergo several structural changes to prepare for sister chromatids segregation into two daughter nuclei. The two sister chromatids are linked by a structure known as a centromere; in humans, this structure is found between the short arm and long arm of the chromosome and it is here that the mitotic spindle attachments occur which facilitate the separation of the two. Recently, RNA-DNA hybrids (R-loops), which are secondary DNA structured composed of an RNA strand attached to an open double-strand of DNA, have been found at centromeres in human cells undergoing mitosis suggesting they may play a role in regulating this process (1). Why they form and their functionality in the centromere is unclear, though they were shown to activate a kinase, Aurora B (AUKB), which has known roles in the regulation of mitosis (2). AUKB is the kinase member of a complex known as the chromosomal passenger complex (CPC), whose other members include INCENP (inner centromere protein), Borealin and Survivin. Though AUKB activities play a role in resolving mitosis associated centromeric R-loops, it is unknown if the rest of the CPC play roles. Here, the authors investigate the timing of R-loop formation during mitosis and asked what factors may act in response to centromeric R-loops. They show that centromeric R-loops remain through prophase but are resolved before anaphase in a manner dependent upon AUKB activity, the CPC and an RNA interacting factor called RBMX.

 

Key Findings:

 

1. Repetitive regions form R-loops during mitosis

First, the authors asked when and where R-loops form in mitotic cells. They also asked if R-loop formation was affected when the activities of the kinase AUKB were blocked. Using DRIPseq, a modified form of chromatin immunoprecipitation which uses an R-loop specific antibody (S9.6 antibody; 3), the authors stalled cultured human cells in mitosis using a drug called Colcemid then treated the cells with either DMSO (as a control) or one of two AUKB inhibitors (AZD or ZM; 4,5). After collecting and sequencing the DNA from their DRIPseq experiments, they found R-loops form more readily in mitosis at repetitive regions (compared to unsynchronised cells) including alpha satellite regions located at centromeres. By inhibiting the activities of AUKB, they also found R-loops were enrichment at both transcription termination sites and at these alpha satellite regions relative to uninhibited cells. Together, they show repetitive regions accumulate R-loops and AUKB can act to resolve these R-loops.

Figure shows selected data from Moran et al. Fig.3A. shows representative images of control (DMSO) or AUKB inhibitor (AZD or ZM) treated cells. Fig.5D shows an immunoprecipitation experiment showing RBMX can interact with the CPC. Fig. 5E shows representative images of the localisation of RBMX to the centromere. ACA is used to mark centromeric location. Fig.5H shows representative images of a proximity ligation assay (PLA) of Survivin (a CPC member) and RBMX. Fig. 7K Schematic summary of the findings from Moran et al. Figures adapted from Moran et al under a CC BY-NC-ND 4.0 licence.

 

2) R-loops can only be found in centromeres after prophase

Next, the authors asked where R-loops form using immunofluorescence in unsynchronised human cells. To detect R-loops, they used the antibody S9.6. By measuring and quantifying R-loop signal across the chromosomes, they found that R-loop signal was reduced as mitosis progressed, in particular they saw that R-loop formation correlated with chromosome condensation. After prophase, the only R-loops observed were within centromeric regions. Examining the locations of prophase R-loops further via high resolution microscopy, they saw that prophase R-loops formed at sites of heterochromatin (‘silenced’ chromatin) at the nuclear edge. When they correlated R-loop signal with DAPI signal (which stained the chromosomal DNA), the authors found higher R-loops signals were associated with less bright DAPI staining indicating R-loops form when chromosomes condense in early mitosis.

 

 3) AUKB can resolve mitotic R-loops

To ask whether AUKB played a role in centromeric R-loop regulation within mitosis, the authors localised R-loops in unsynchronised cells via microscopy in the presence or absence of AUKB activity. Using two difference AUKB inhibitors, the authors found that inhibiting AUKB activity was associated with higher R-loop levels at chromosomes in mitosis and at their centromeres. They confirmed this using DRIPseq and PCR analysis of these regions. This effect was prominent at alpha satellite repeat regions associated with centromeres (compared with another repeat region). Additionally, depleting R-loops both affected the intensity of AUKB signal by microscopy and the ability of the downstream of AUKB interacting complex (the chromosomal passenger complex; CPC) to drive chromosome movements within the nucleus.

 

4) CPC acts to regulate centromeric R-loops

To ask if the CPC was directly involved in resolving these centromeric R-loops, the authors purified proteins bound to chromatin in mitotic cells by using CPC factors as bait to capture other bound proteins and a chromatography-based approach (known as Multidimensional Protein Identification Technology; MudPIT; 6) to ask about what proteins were present. They found 111 proteins, including other members of the CPC and known CPC interacting factors such as Topo II alpha. Similar proteins to those found here were also identified when proteins associated with R-loops were purified. The factors identified also implicated both AUKB activity and the downstream CPC complex as key mediators in centromeric R-loop regulation, suggesting these factors can on centromeric R-loops during mitosis.

 

 5) RBMX maintains centromere conhesion via AUKB and CPC activities

One factor that stood out to the authors as a putative regulator of R-loop centromeric activities was the protein RNA binding motif gene, X chromosome (RMBX). RBMX was found to interact with R-loops but also has roles in cohesin at the centromeres and interacts with the CPC. RBMX is enriched on chromatin at alpha satellite regions at the centromere, where it is recruited by the CPC complex. The authors depleted RBMX by shRNA and examined for R-loop and AUKB locations. They found R-loops increased and AUKB localisation to the centromeres was compromised. Their data suggests RBMX is recruited by R-loops to the centromere, where it can recruit AUKB and subsequently the CPC. In the absence of RBMX, cohesion in the centromere was lost; the effect was associated with the inadequate recruitment of a centromeric cohesin regulator called Sgo1. Sgo1 is a factor which acts to maintain centromeric cohesin, stopping the dissociation of cohesin occurring too early in mitosis (7). Sgo1 can be recruited by the CPC and cohesin requires AUKB activity. Together, these data, combined with prior studies indicated centromeric R-loops can bring in RBMX and AUKB, which in turn recruit CPC, followed by Sgo1 to maintain chromatin cohesion prior to division.

 

In all, the authors have shed light on a novel pathway surrounding the resolution of mitotic R-loops, linking their findings to the activities of a known mitotic regulatory kinase, AUKB, and its associated factors. Their data supports R-loops as key factors for chromosome segregation.

 

What I liked about this preprint:

I really enjoyed reading this preprint study. Mitotic R-loops were shown recently to have a role at the centromere, but their functions within this discreet region were still uncharacterised. Here, the authors took a combined approach to build upon previous studies and data to ask what mitotic R-loops may be doing. Using high resolution microscopy, known inhibitors and next generation sequencing they were able to propose a model and subsequent pathway for the roles of R-loops at the centromeres during mitosis. I really liked the methodical approach the study took to delineate what events were occurring at the centromere and I found the results very compelling and exciting! Their study raises numerous questions and will open up new routes of investigation in the field of R-loop biology.

 

Questions for the Authors:

Q1: What are the consequences of failing to resolve R-loops in the centromeric region?

Q2: Regarding the two AUKB inhibitor compounds; why following AZD treatment do you see an increase in R-loops in interphase DNA and at centromeres whereas treatment with ZM did not cause this affect?

Q3: To deplete R-loops, you overexpress RNAseH1. Is it possible that the changes in AUKB localisation you see after the overexpression of RNAseH1 are not solely due to the reduced R-loop levels failing to recruit AUKB but possibly the increased levels of the RNAseH1 enzyme may be interfering with, or even preventing, AUKB recruitment to the centromere?

Q4: The Kabeche et al. study from 2018 suggests centromeric R-loops may be sensed and signalled by the atypical kinase ATR. You suggest RBMX is recruited by R-loops to the centromere. Do you think ATR stimulates its recruitment or are you proposing RBMX is recruited independently to ATR kinase activity?

Q5: What machinery/pathway do think operates to resolve the centromeric R-loops?

 

References:

1. Kabeche, L., Nguyen, H.D., Buisson, R. and Zou, L. A mitosis specific and R-loop driven ATR pathway promotes faithful chromosome segregation. Science (2019).
2. Willems, E., Deboddeleer, M., Digregorio, M., Lombard, A., Lumapat, P. N. and Rogister, B. The functional diversity of Aurora kinases: a comprehensive review. Cell Division (2018).
3. Ginno, P. A., Lott, P L. Christensen, H. C., Korf, I. and Chedin, F. R-loop formation is a distinctive characteristic of unmethylated human CpG island promoters. Molecular Cell (2012).
4. Yang. J., Ikezoe, T., Nishioka, C., Tasaka, T., Taniguchi, A., Kuwayama, Y., Komatsu, N., Bandobashi, K., Togitani, K., Koeffler, H. P., Taguchi, H. and Yokoyama, A. AZD1152, a novel and selective aurora B kinase inhibitor, induces growth arrest, apoptosis, and sensitization for tubulin depolymerizing agent or topoisomerase II inhibitor in human acute leukemia cells in vitro and in vivo. Blood. (2007).
5. Walsby. E., Walsh. V., Pepper. C., Burnett. A. and Mills. K. Effects of the aurora kinase inhibitors AZD1152-HQPA and ZM447439 on growth arrest and polyploidy in acute myeloid leukemia cell lines and primary blasts. Haematologica (2008)
6. Schirmer, E. C., Yates 3^rd, J. R. and Gerace, L. MudPIT: A powerful proteomics tool for discovery. Discovery Medicine (2003).
7. Zhang, Q and Liu, H. Functioning mechanisms of Shugoshin-1 in centromeric cohesion during mitosis. Essays Biochem. (2020)

Tags: aukb, aurora kinase, centromere, mitosis, rloops

doi: https://doi.org/10.1242/prelights.27322

Read preprint (No Ratings Yet)