Author's response

Susanne M.A. Lens shared about A motor-based approach to induce chromosome-specific mis-segregations in human cells

What kind of experiments do you propose to evaluate the cell’s immediate response to acquiring the new aneuploidy?

By immunofluorescence, we will initially check a number of the known aneuploidy (stress) responses, such as DNA damage and lysosomal/proteosome responses. Alterations, if detected, can in principle, be directly correlated with aneuploidy state of single cells reported by the number of (dCas9) GFP foci. By single cell RNA-seq, performed immediately after a round of targeted mis-segregation we hope to reveal, in an unbiased manner, altered signalling pathways in daughter cells with the targeted aneuploidy. For this we preferentially want to single out and compare the daughter pairs that have gained and lost the chromosomal arm.

Did you ever try to continue culturing the cells? Do they ever retain the extra piece of DNA (e.g. through translocations/rearrangements with other chromosomes)?

Not yet, but for follow-up studies we plan to sort out the cells with a targeted mis-segregation (as for RNA-seq). It will be very interesting to see what happens to the chromosomal arms that are likely missing a centromere, such as 9q and 1p. We plan to do this both for p53wt (the ones we used in the manuscript) and p53-/- RPE1-dCas9-Kin14VIb cells.

Can you speculate about why the ‘out> 1-1’ subpopulation, specifically, has a delay in anaphase onset?

We can indeed only speculate. It might be that in this subpopulation the TetO chromosome was more often first transported by the kinesin towards one of the spindle poles, i.e. before its KTs were captured by MTs. For instance, because at NEB the TetO chromosome was in close proximity of the pole. This could make it more challenging for the KTs of that chromosome to attach to MTs and to bi-orient, causing a mitotic delay. However, once attached to MTs, the kMT pulling forces will counteract the poleward transport force by the kinesins at the telomere. Depending on the number of kinesin molecules accumulating on the TetO locus, this kMT pulling force could lead to detachment of the kinesins from the microtubule, resulting in a 1-1 segregation of the locus. As said, pure speculation, to confirm this we would need to image the cells at short time intervals and with markers for KT attachment, but this severely limits the number of cells that we can follow per experiment. Also, we feel that the population of cells with a 2-0 distribution of the locus is the population of interest.

Sarah E. McClelland shared about Inducing Specific Chromosome Mis-Segregation in Human Cells

What kind of experiments do you propose to evaluate the cell’s immediate response to acquiring the new aneuploidy?

We started this project (many years ago!) with the major aim to characterise the immediate signalling responses activated by changes in chromosome number. These have not yet been fully dissected since the majority of aneuploid model systems have been analysed for aneuploidy responses long after the initial change in chromosome number or structure. We reason that having a cellular population that had all (or nearly all) very recently undergone the same aneuploidy event would represent a powerful tool to undertake population level ‘omics assays to systematically assess in an unbiased manner, the ‘acute’ aneuploidy state. One obvious reason for this is to understand the potential negative impacts of aneuploidy on cells, and here also to investigate how the acute aneuploidy response might vary between different chromosomes, and gains or losses, whole, or partial chromosomes. We have also become very interested in the topic of gene dosage compensation given lots of interesting discussions around this recently at conferences. In particular, we wonder how and when dosage compensation (if it is happening) occurs after the initial aneuploidy event, and how this could vary between monosomic and trisomic events.

Additionally, using longer term cell outgrowth experiments, we plan to address whether each newly acquired aneuploidy is maintained or lost after a few generations, and the dependence on functional p53 for this.  Particularly important in the context of cancer evolution will be to understand whether, for a specific chromosome, gains or losses are maintained at the same rate, an important and yet not fully answered question.

We have also long wondered how cancer cells are able to continually reshuffle their genomes, constantly suffering new chromosomal alterations and yet continuing proliferation. Our assumption is that initial aneuploidy responses that would usually arrest cell proliferation have been adapted to in cancer. Ultimately, we aim to discover the pathways acutely activated by aneuploidy, and thus provide a starting point to discover mechanisms by which cancer cells evade these responses, and provide routes to capitalise on this for cancer therapy.

Did you ever try to continue culturing the cells? Do they ever retain the extra piece of DNA (e.g. through translocations/rearrangements with other chromosomes)?

Although here we set out to create an event of chromosome mis-segregation and did not yet examine beyond the first cell division, culturing the newly created aneuploid cells will be really interesting to perform in the future. In view of previous works on aneuploidy evolution over time (e.g. Soto et al, Cell reports 2017, Ippolito et al, Dev cell 2021, Lukow et al., Dev cell 2021 and others), we have reasons to believe that the new karyotypes might evolve during time if kept in culture. We hope that our new system will provide a unique model – whereby we can combine information about the initial cellular responses and the longer-term evolution or tolerance of specific aneuploidies to discover the role of both the aneuploidy itself, and the stresses involved in its creation.

The system relies on repetitive loci with multiple binding sites. How flexible is the system, i.e., how many of those sites would one expect to find in the genome (using your lower limit of 1441 binding sites)?

Several studies have used repetitive regions for imaging purposes across the majority of human chromosomes (with dCas9-GFP; e.g. Stanyte et al, JCB 2018). We are interested to resolve more precisely the minimal length of array that can be used to build a functional ectopic kinetochore which will allow us to define a set of targetable chromosomal regions based on the previously used sites, as it is likely that 1441 sites might be substantially higher than required. Moreover, the newly published telomere to telomere human genome assembly (Nurk et al., 2022) has now unlocked previously unknown sequences of many repetitive sequences and segmental duplication, therefore potentially expanding the flexibility of our system. It is also possible to target imageable foci of dCas9 using a tiled sgRNA array (‘CASFISH’; Deng et al, PNAS 2015), so this could be an option for genomic regions that do not harbour endogenous repetitive sequences. Lastly, for applications where genomic editing of cell lines is possible, we are currently developing tools to enable rapid Cas9-based insertion of custom synthetic arrays into specific loci. So – the short answer is we don’t know yet, but anticipate being able to target most genomic regions depending on the application required.

In figure 2E, it looks like the dCas9-EGFP foci for the Chr9-CEN target site also have an increased CENP-T intensity, almost as much as the native centromeres. The KNL-1 intensity is even higher than the intensity on the native centromeres (Figure 2H). Could you speculate about why this is the case ? Is it because it is difficult to distinguish the signal from the nearby native centromere?

You have a sharp eye there! Considering the close distance of the pericentromeric target site with the endogenous centromere, and the large centromeric region of chromosome 9 (more than 40 Mb), we know in our quantification we are unavoidably capturing both the exogenously-targeted CENP-T, and some of the signal of the endogenous centromere due to proximity. Why KNL1 intensity is slightly higher at chromosome 9 than that at the average centromere is interesting, and could reflect that chromosome 9 has a stronger than average centromere (in terms of KNL1 recruitment at least) in these cells.

The single cell sequencing experiments indicate that it might be challenging to find cells of interest within a population in which mis-segregation was induced. How would you pick out the right cells from the population for further analysis (e.g. for downstream omics)?

Within such a mixed population of cells, picking the right cells for further analysis might be indeed challenging. We are approaching this in several ways for future studies; First, we can create single cell clones after the induction of single chromosome mis-segregation. In this way, it will be possible to differentiate the impact of each type of newly generated aneuploidy (for example by karyotyping at specific time points after aneuploidy induction) although here the immediate responses would be lost to analysis. Second, we can enrich the population for those cells that have ectopic kinetochores using FACS-based cell sorting based on the GFP foci. Third, we can exploit the metaphase arrest to isolate mitotic cells before shakeoff and pulsing with Mps1 inhibitors. Lastly, coupled to the synthetic array system mentioned above we can integrate fluorescent markers to allow the isolation (again using FACS for example) of cells that lost the fluorescence marker due to loss of that chromosome, or chromosome region.