(C) PLOS One
This story was originally published by PLOS One and is unaltered.
 . . . . . . . . . .



Conserved signalling functions for Mps1, Mad1 and Mad2 in the Cryptococcus neoformans spindle checkpoint [1]

['Koly Aktar', 'Institute Of Cell Biology', 'School Of Biological Sciences', 'University Of Edinburgh', 'Edinburgh', 'United Kingdom', 'Thomas Davies', 'Ioanna Leontiou', 'Ivan Clark', 'Christos Spanos']

Date: 2024-06 

Cryptococcus neoformans is an opportunistic, human fungal pathogen which undergoes fascinating switches in cell cycle control and ploidy when it encounters stressful environments such as the human lung. Here we carry out a mechanistic analysis of the spindle checkpoint which regulates the metaphase to anaphase transition, focusing on Mps1 kinase and the downstream checkpoint components Mad1 and Mad2. We demonstrate that Cryptococcus mad1Δ or mad2Δ strains are unable to respond to microtubule perturbations, continuing to re-bud and divide, and die as a consequence. Fluorescent tagging of Chromosome 3, using a lacO array and mNeonGreen-lacI fusion protein, demonstrates that mad mutants are unable to maintain sister-chromatid cohesion in the absence of microtubule polymers. Thus, the classic checkpoint functions of the SAC are conserved in Cryptococcus. In interphase, GFP-Mad1 is enriched at the nuclear periphery, and it is recruited to unattached kinetochores in mitosis. Purification of GFP-Mad1 followed by mass spectrometric analysis of associated proteins show that it forms a complex with Mad2 and that it interacts with other checkpoint signalling components (Bub1) and effectors (Cdc20 and APC/C sub-units) in mitosis. We also demonstrate that overexpression of Mps1 kinase is sufficient to arrest Cryptococcus cells in mitosis, and show that this arrest is dependent on both Mad1 and Mad2. We find that a C-terminal fragment of Mad1 is an effective in vitro substrate for Mps1 kinase and map several Mad1 phosphorylation sites. Some sites are highly conserved within the C-terminal Mad1 structure and we demonstrate that mutation of threonine 667 (T667A) leads to loss of checkpoint signalling and abrogation of the GAL-MPS1 arrest. Thus Mps1-dependent phosphorylation of C-terminal Mad1 residues is a critical step in Cryptococcus spindle checkpoint signalling. We conclude that CnMps1 protein kinase, Mad1 and Mad2 proteins have all conserved their important, spindle checkpoint signalling roles helping ensure high fidelity chromosome segregation.

Cryptococcus undergoes fascinating changes during its life and infection cycle. Here we study a control process regulating normal cell divisions, monitoring the movement of DNA between mother and daughter cells. This surveillance system is known as the spindle checkpoint and we demonstrate that its components (Mps1, Mad1 and Mad2), and their mechanism of action, are well conserved. Knowing this, we can now target subtle changes in this pathway for future drug development.

Cryptococcus neoformans is an environmental fungus that kills a large fraction of AIDS patients through meningitis. The World Health Organisation recently published a report on disease-causing fungi and ranked C. neoformans at the top, in the critical priority group of fungal pathogens. This position reflects the public health importance of this organism, the lack of effective drugs for fighting its infection, and the need to prioritise its research and development.

Funding: This work was supported by grants from the Leverhulme Trust (RPG-2018-379 to IL, KGH); the Darwin Trust of Edinburgh (KA); the Wellcome Trust (SRF 202811 to AAJ; Wellcome Centre for Cell Biology core grant 203149; Wellcome iCM programme 218470 to TD; Wellcome Sir Henry Dale Fellowship 208779 to EW; the Wellcome-University of Edinburgh Institutional Strategic Support Fund for IC); and the European Research Council (ERC Advanced Grant CHROMSEG; 101054950, AAJ). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of this manuscript.

Copyright: © 2024 Aktar et al. This is an open access article distributed under the terms of the Creative Commons Attribution License , which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Here we have knocked out the MAD1 and MAD2 genes and compared their phenotypes to that of the mps1Δ strain. We demonstrate that all three components are essential for spindle checkpoint function: in response to anti-microtubule drug treatment, deletion mutants are unable to maintain sister-chromatid cohesion, continue to divide and die. Overexpression of CnMps1 kinase is sufficient to induce metaphase arrest and this is dependent on both Mad1 and Mad2. We identify Mps1 phosphorylation sites in the Mad1 C-terminus and demonstrate that the C-terminal Mad1 T667A substitution abolishes checkpoint signalling and GAL-MPS1 arrest. We conclude that many aspects of spindle checkpoint signalling and Mps1, Mad1 and Mad2 functions are conserved in basidiomycetes.

The mitotic spindle checkpoint is a key regulator of both mitotic and meiotic divisions, and has been studied in detail in several model systems and human cells [ 14 , 15 ]. This cell cycle checkpoint monitors interactions between kinetochores and spindle microtubules, and if any problems are apparent the checkpoint provides additional time for them to be resolved, by delaying the metaphase to anaphase transition. Its molecular components were identified in budding yeast genetic screens [ 16 – 18 ], and the MAD, BUB and MPS1 genes and their mode of action have been found to be extremely well conserved through eukaryotic evolution [ 19 ]. The spindle checkpoint has not yet been described in Cryptococcus, although both the BUB1 and MPS1 kinases were shown to be relevant to virulence in a genome-wide screen where 129 protein kinases were knocked out [ 20 ].

During human infection Cryptococcal cells are stressed and undergo fascinating morphological transitions. They can form a small morphotype, known as seed cells, which appear to help Cryptococcus enter organs beyond the lung [ 7 ]. They also form polyploid titan cells in the lung [ 8 ]. These large cells (up to 100 micron diameter) are protective: the immune system struggles to clear them because of their size and their protective outer capsule [ 9 ]. Titan cells are polyploid and a recent study identified a specific cyclin, Cln1, as a key regulator of this transition [ 10 ]. Once formed, titan cells continue to divide by budding off small daughter cells, which have a high viability and are haploid [ 8 ]. It has been proposed that this division is likely to be somewhat error-prone, leading to aneuploid daughters. Such aneuploidy would increase genetic diversity in the fungal population and thus be relevant to the generation of drug-resistance in the clinic [ 4 ]. In support of this, aneuploidy (chromosome 1 disomy) has been found to confer fluconazole resistance to Cryptococcus [ 11 , 12 ]. In general, aneuploidy leads to reduced fitness [ 13 ] and cells have evolved a number of surveillance systems, known as cell cycle checkpoints, that keep chromosome mis-segregation to a minimum.

Cryptococcus neoformans is a model basidiomycete. These are very distant from the other major division of fungi, the ascomycetes which include the well characterised model organisms Saccharomyces cerevisiae and Schizosaccharomyces pombe. Important previous studies have described cell division in Cryptococcus neoformans. It has 14 chromosomes [ 5 ] and undergoes a partially open mitosis with the nuclear division taking place in the bud and one set of chromosomes moving back to the mother cell during anaphase [ 6 ].

Cryptococcus neoformans was recently designated a critical priority human fungal pathogen by the WHO [ 1 ], being responsible for the death of around 20% of AIDS patients. Recent estimates have 152,000 annual cases of Cryptococcal meningitis worldwide, mainly in Sub-Saharan Africa, leading to ~112,000 deaths in AIDS patients [ 2 ]. Few drugs are currently available and drug-resistant strains have emerged in the clinic, so new treatments are urgently required [ 3 ]. We are investigating mitotic control of chromosome segregation as a possible future drug target in Cryptococcus neoformans [ 4 ].

Results

Analysis of the Cryptococcus neoformans genome sequence in the FungiDB database identified CNAG_04824 and CNAG_01638 as likely homologues of MAD1 and MAD2.

Their sequences are well conserved, with CnMad1 being predicted by alphafold to be an elongated coiled-coil protein, likely to bind both Mad2 and Bub1 (Fig 1A). To test whether their function is conserved we knocked out both genes, using the amdS Blaster (dominant recyclable marker) approach, developed in Cryptococcus by James Fraser et al [21]. First we replaced the MAD gene with the amdS marker by homologous recombination (S1A Fig). AmdS, encoding the Aspergillus nidulans acetamidase gene, enables acetamide to be used as both a carbon and nitrogen source by cells, and transformants were selected for on acetamide plates. In a second step the amdS marker was allowed to recombine out, via flanking repeat sequences, and its loss was selected for by growth on fluoracetamide. This compound kills any cells still containing the amdS marker as it is metabolised into toxic fluoroacetyl CoA and fluorocitrate, disabling aconitase and inhibiting the citric acid cycle. PCR analysis of genomic DNA was used to confirm targeted integration and recombination (S1B Fig). To confirm the mad1 deletion, we made a polyclonal anti-Mad1 antibody, by expressing amino acids 1–200 of Mad1 fused to 6xHis-MBP (maltose binding protein) in bacteria and injecting the purified MBP-Mad1 protein as antigen into a sheep. The resulting serum was affinity-purified and then used to immunoblot Mad1 proteins from several Cryptococcus strains (Fig 1B) confirming that Mad1p is no longer expressed in the mad1Δ strain.

PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 1. CnMad1 and Mad2 proteins are conserved and mad1 and mad2 mutants are sensitive to anti-microtubule drugs. (A) An alphafold model of Mad1-Mad2 hetero-tetramer indicating conserved domains in CnMad1: the Mad2 interaction site; 2 putative Bub1 binding motifs (RLK) and the structurally conserved C-terminal domain (RWD). (B) The mad1Δ strain is benomyl sensitive and can be complemented with an ectopic GFP-Mad1 construct. Strains were grown on YPD plates containing the indicated concentrations of benomyl for 3 days at 30°C. (C) Anti-Mad1 immunoblot confirms the mad1Δ strain. The three strains indicated were grown to log phase, harvested and whole cell extracts generated. The upper band (~115kD, labelled *) recognised by the anti-Mad1 antibody is a cross-reacting band and is used here as a loading control. (D) Benomyl sensitivity of the mad2Δ and complementation with ectopic GALp-MAD2. Strains were grown on plates (with 2% glucose or 2% galactose) containing the indicated concentrations of benomyl for 3 days at 30°C. (E) Plate reader experiments confirm that mad1, mad2 and mps1 mutants are all temperature sensitive, with populations expanding slower than wild-type. Cells were diluted to OD 600 0.2 and grown with shaking at the indicated temperatures. For viability assay (colony forming units) see S3 Fig. (F) The mad1,mad2 double mutant does not display a synthetic phenotype. mad1, mad2, mad1,mad2 double and mps1 strains were diluted, plated and grown on YPD plates at the indicated temperatures and drug concentrations for 3 days. https://doi.org/10.1371/journal.pgen.1011302.g001

To analyse the loss of function phenotype we plated mad1Δ cells on the anti-microtubule drug benomyl. Fig 1C shows that the mad1Δ strain is sensitive to this drug, but not as sensitive as an mps1Δ strain. We used mps1Δ as a control: this strain has been previously described by the Bahn group, when analysing the kinome of Cryptococcus neoformans [20]. Fig 1C shows that we can rescue the benomyl sensitivity of the mad1Δ strain by complementation with a GFP-Mad1 construct, confirming that we have knocked out the right gene and that the benomyl sensitivity observed is due to loss of Mad1 function. Fig 1B shows that the expression level of this GFP-Mad1 fusion protein is similar to wild-type Mad1 levels and Fig 1C shows that the GFP-Mad1 fusion protein is functional.

Next we knocked out the MAD2 gene using the same approach, and confirmed homologous recombination through PCR analysis of genomic DNA, before and after the amdS marker was lost (S2 Fig). These strains were also found to be benomyl sensitive and could be rescued by expression of an ectopic copy of Myc-MAD2 expressed from the GAL7 promoter in galactose media (Fig 1D).

We next made a mad1, mad2 double mutant, by repeating this experiment and sequentially knocking out MAD2 in the mad1Δ strain. Fig 1F shows that the double mutant (mad1, mad2) was no more sensitive than the single mutants, suggesting that these proteins carry out their function(s) in a concerted fashion. This is consistent with known behaviour of these Mad proteins in other systems where they form a stable, constitutive complex [22–25].

As expected [26], the strain lacking Mps1 protein kinase is significantly more sensitive to the anti-microtubule drugs than the mad mutants (see Fig 1E). Our interpretation is that CnMps1 kinase is likely to have other mitotic functions, in addition to checkpoint signalling, such as error-correction and bi-orientation [27–29]. mps1Δ mutants were also reported to be temperature sensitive [20]. We found a subtle temperature-sensitive phenotype on plates for the mad mutants (Fig 1E), and Fig 1F shows that growth of liquid cultures in a plate reader confirms the phenotype at 37°C for both mad1 and mad2 mutants. This ts phenotype was not as severe as that displayed by the mps1Δ mutant. Thus, as expected of a spindle checkpoint component, the mad1 and mad2 mutants are sensitive to anti-microtubule drugs. However, they are also sensitive to other stresses, such as high temperatures, which to our knowledge has not been reported in other yeasts.

mad1Δ and mad2Δ mutants are checkpoint defective To test whether these mutants have spindle checkpoint defects, we carried out additional assays that analyse how individual cells respond to anti-microtubule drug treatment. More specifically, we asked whether they can arrest in mitosis and maintain sister-chromatid cohesion when treated with the microtubule-depolymerising agent nocodazole. Fig 2A and 2B reveal that, unlike wild-type (H99) cells, mad1Δ, mad2Δ, or mps1Δ mutants were unable to arrest as large-budded cells when treated with nocodazole. Wild-type cells display a very robust arrest under these conditions (2μg/ml nocodazole treatment for 90 minutes at 30°C). A subtle mitotic delay, in response to nocodazole treatment, might be missed with such fixed time point analysis, so we also employed microfluidics to analyse in more detail how single cells respond to nocodazole over time. Strains were pre-grown in SC media and then put in the microfluidics device, where single cells were captured (Fig 2C) before imaging. After 5 hours, nocodazole was added (2μg/ml) and imaging continued for a further 6 hours, with images being captured every 2 minutes. Movies were then analysed to see how many individual cells arrested in the nocodazole, and how many continued to re-bud and divide. Fig 2D confirms that the mad1Δ, mad2Δ, and mps1Δ mutants were all unable to delay mitotic progression in response to nocodazole treatment. We conclude that deletion of any one of these genes has likely completely abrogated the checkpoint response. Our results are in agreement with those from the Sanyal lab which independently generated a mad2Δ strain and demonstrated that it was unable to arrest as large-budded cells upon thiabendazole treatment [30]. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 2. Cnmad1Δ and mad2Δ strains fail to checkpoint arrest. (A) mad1, mad2 and mps1 strains fail to maintain a large-budded nocodazole arrest. The strains indicated were grown to log phase and then nocodazole was added to a final concentration of 2.5 μg/ml in YPD media for 3 hours. Scale bar is 10μm. (B) Quantitation of the large-budded cells (bud diameter >4μm) observed at the 3 hour time point during this nocodazole challenge. This experiment was repeated 3 times and 300 cells were counted for each strain per experiment. (C) Images taken from a microfluidics analysis of this behaviour in nocodazole (for H99, mad1, mad2 and mps1 mutants). Cells were pre-grown in synthetic complete media supplemented with 0.2g/ml glucose and 0.05% w/v bovine serum albumin (Sigma) and then injected into the microfluidics device and analysed for 4 hours (+/- 2.5μg/ml nocodazole). Scale bar is 10μm. (D) Quantitation of the microfluidics experiment: movies were analysed manually for re-budding. This experiment was repeated three times and at least 100 cells counted per strain each time. https://doi.org/10.1371/journal.pgen.1011302.g002 The key roles of the SAC are to delay the metaphase to anaphase transition and to protect sister-chromatid cohesion. If sister-chromatids separate prematurely, before bi-orientation of all chromosomes is complete, then sisters will segregate randomly leading to aneuploidy. GFP-marked chromosomes are widely used in model organisms to monitor and quantitate sister-chromatid separation [31,32]. We integrated an array of 240 lac operators (lacO, [33]) into chromosome 3 and expressed a lacI-NeonGreen fusion protein to mark just one of the fourteen Cryptococcus chromosomes (Fig 3A). We knocked out MAD2 in this strain as above (producing KA159) and then complemented that strain with an ectopic copy of MAD2 expressed from its own promoter (producing KA196). Comparison of these strains in the presence of nocodazole demonstrates a sister-chromatid cohesion defect in the mad2Δ strain (Fig 3B). After 90 minutes of nocodazole treatment ~40% of mad2Δ cells have separated sisters, compared to only 2% of wild-type cells in (Fig 3C). It should be noted that in such assays not all sisters will visibly separate, as in the absence of microtubules there are no spindle forces pulling them apart. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 3. The mad2Δ mutant fails to maintain sister-chromatid separation. (A) Schematic representation of the strain with fluorescently-marked chromosome 3. An array of 240 lac operators were integrated at the safe-haven on chromosome 3 and lacI-mNeonGreen expressed. When sister chromatids separate 2 spots are seen. (B) mNeonGreen was imaged in YPD cultures of ‘wild-type’ (mad2Δ, complemented by an ectopic MAD2 construct) and mad2Δ strains, 3 hours after the addition of 2.5μg/ml nocodazole. Scale bar is 10μm. (C) Quantitation of this experiment after three biological repeats. 100 cells were scored in each condition for each experiment. Note, as nocodazole is present there are no microtubule spindles to pull sister chromatids apart in this experiment, so not all sisters appear to separate. Images show relevant magnified cells from 3B. https://doi.org/10.1371/journal.pgen.1011302.g003 Next, we analysed the rate of death of the mad mutants in response to anti-microtubule drug treatment. To quantitate this, we determined the number of colony forming units (CFU) on plates following a time-course of nocodazole treatment for wild-type and sac- mutant cultures. The mad and mps1 mutants all die faster than wild-type cells, likely as a consequence of their first division in nocodazole (S3 Fig). In summary, in response to anti-microtubule drug treatments, the mad mutants fail to arrest, fail to maintain sister-chromatid cohesion, re-bud and die. We conclude that mad1, mad2 and mps1 mutants are all spindle checkpoint defective.

[END]
---
[1] Url: https://journals.plos.org/plosgenetics/article?id=10.1371/journal.pgen.1011302

Published and (C) by PLOS One
Content appears here under this condition or license: Creative Commons - Attribution BY 4.0.

via Magical.Fish Gopher News Feeds:
gopher://magical.fish/1/feeds/news/plosone/