The kinetochore provides the end-on attachment of central core complexes, such as Ndc80, to the spindle microtubules (MTs) for a balanced chromosome segregation. The K-acetyltransferase Gcn5, involved in acetylation dependent processes, is also engaged in the control of the cell cycle progression. Here we show that in budding yeast the deletion of the KAT GCN5 gene leads to a complete growth recovery of a strain missing the central kinetochore component Ndc80. Accordingly, we obtained full deletion of the essential, highly conserved, Ndc80 KT subunit in the gcn5Δ strain. We also demonstrated that the deletion of Gim3, a subunit of the tubulin chaperone, abrogates the recovery of cell growth in the ndc80-1-gcn5Δ strain suggesting an involvement of the kinetochore-MT attachment in this process. Our observations suggest the notion that KAT Gcn5 exerts a regulatory role in the interaction of the central kinetochore Ndc80 complex to the spindle microtubules exhorting us to dig more into the mechanistic of this process.

Mitotic stability requires a successful attachment of the kinetochore (KT) to the mitotic spindle^[1][2] achieved also through the activity of the central KMN network^[3][4]. In S.cerevisiae, mutations in KT and KMN subunits are lethal due to kinetochore aberrations preventing the accomplishment of mitosis. The central KT-hub is the Ndc80 complex composed by Ndc80, Spc24, Spc25 and Nuf2 proteins, conserved from fungi to humans and implicated in several essential functions like binding to microtubules and, recruitment of the mitotic checkpoint^[5][6]. Depletion of Ndc80 and Nuf2 results in mislocalization of several kinetochore subunits^[7]. Ndc80c is a 4 unit dumbbell shaped coiled coil complex with 2 positively charged head domains (Ndc80/Nuf2) on one end, Ndc80 binds the negative microtubule surface^[8], on the opposite site, C-terminal domains of Spc24/Spc25 bind to the kinetochore^[9]. It has been proposed that Nuf2 is not directly involved in the binding with microtubules suggesting a substantiallly different contribution compared to Ndc80^[10]. Ndc80 also shows an internal loop that mediates the recruitment of Dam1^[11], a component of the ring embracing the mitotic spindle^[12] and a toe CH domain that mediates the microtubule binding at high affinity^[13]. Finally, Ndc80 is essential to recruit the SAC proteins Mad1 and Mad2 to unattached kinetochores while KMN-Spc105 recruits Bub1/Bub3 and Mad1/Mad2 necessary to delay cell division^[14].

Here we assay the role of K-acetyltransferase Gcn5 on kinetochore function and its interaction with mitotic spindle. Unexpected results highlighted the relevance of Gcn5 in this phenomenon and show that in absence of Gcn5 cells can override mitotic arrest due to an aberrant kinetochore and recover growth.

Ndc80 central kinetochore component is a tetrameric elongated complex composed of Ndc80, Nuf2, Spc24 and Spc25 proteins (Fig. 1A) that connects the kinetochore to the spindle microtubules. Almost all the KT subunits are essential for viability in budding yeast, indeed the corresponding TS mutants^[15][16] are lethal at restrictive temperature (34°C). The involvement of Gcn5 in mitotic progression^[17] is due to its many roles, among others the acetylation of regulatory proteins^[18] and the expressions of several cell cycle-related genes^[19] was demonstrated. In an attempt to analyze additional effects of KAT Gcn5 on the kinetochore functions, we carried out a screening of S.cerevisiae TS mutants in kinetochore subunits where we deleted K-acetyltransferase Gcn5.

Loss of Gcn5 induces cell growth in Ndc80 complex conditionally lethal-mutants

Here we show that TS mutants ndc80-1, spc24-1 spc25-1 and spc105-15 of central kinetochore, involved in the interaction with spindle microtubules, grow normally at 28°C and lose viability at non permissive temperatures (34°C). After deletion of KAT-Gcn5 we surprisingly found that the double mutants ndc80-1-gcn5Δ, spc24-1-gcn5Δ and spc25-1-gcn5Δ and spc105-15-gcn5Δ strains were able to rescue the growth arrest even at 34°C (Fig. 1B). To confirm the direct role of Gcn5 in this observation, we re-induced the expression of GCN5 in those double mutants by introducing the centromeric plasmid pRS316 containing a functional copy of Gcn5 (pGcn5) and empty vector pRS316 (vector) was used as control. According to our previous findings, the re-introduction of a functional wild type Gcn5p abolished the growth recovery also at 34°C, while the empty vector control did not (Fig. 1C). These results provide a direct evidence that Gcn5 acts as a strict regulator of the cell cycle during the kinetochore-MT interaction.

Role of the tubulin chaperon subunit Gim3

The chaperon complex Gim, composed by five subunits (Gim1-5) in yeast is involved in the formation of functional α and γ-tubulins, needed for a correct assembly of MTs^[20]. In order to understand if the dynamics and the stability of spindle MTs can play a role in the cell growth recovery of defective kinetochore mutants in absence of Gcn5 we disrupted the Gim3 prefoldin, component of the tubulin chaperon Gim complex. Loss of Gim3 abolished the growth recovery in ndc80-1-gcn5Δ mutant, as shown by the growth assay of the triple mutant ndc80-1-gim3Δ-gcn5Δ (Fig. 1D). This result demonstrates that the recovery of aberrant Ndc80 kinetochore in absence of Gcn5 is related to kinetochore-MT interaction and the relevance of acetylation on mitotic spindle with stabilising effects caused by Gcn5 dependent ipoacetylation.

Role of SAC mitotic checkpoint

Next, we wanted to test the component of SAC mitotic checkpont^[21] devoted to control the recovery of aberrant or unattached kinetochores by deleting the checkpoint proteins Mad2 and Bub3 in the ndc80-1 mutant. Spot growth assay (Fig. 1E) shows that the absence of Mad2 does not abrogate the growth recovery observed in ndc80-1-gcn5Δ mutant, while the deletion of Bub3 subunit is unable to skip the cell cycle arrest under restrictive temperature conditions. These results suggest that the recovery of growth in absence of Gcn5 in the ndc80-1 mutant is dependent on Mad2 and not Bub3 in the cell. Next, we analyzed cell morphology of the different mutants by optical microscopy, in figure 1F shows the distribution of the different cell types observed: normal G1 and G2 cells, aberrant elongated cells, undivided cell chains and unhealthy cell clumps. Results show that Mad2 deletion increased the number of dividing G1 cells in ndc80-1, while in absence of Bub3 there was a drastic decrease of dividing G1 cells and an enhancement of inviable aberrant or undivided chains indicating defects in cytokinesis. Collectively, the number and type of cell morphologies observed were in full agreement with the growth assay shown above. The recovery obtained both in ndc80-1-mad2Δ and in ndc80-1-gcn5Δ is similar to the triple ndc80-1-mad2Δ-gcn5Δ strain suggesting that, in absence of Gcn5, cells may overcome the Mad2 dependent cell cycle arrest possibly for improved kinetochore-MT interactions.

Production of ndc80Δ-gcn5Δ strain

Based on our collected results showing that the depletion of GCN5 in the conditionally-lethal mutant ndc80-1 is viable, we wanted to push further our assay asking whether we may fully disrupt the essential NDC80 gene in absence of Gcn5 (gcn5Δ) and obtain a viable strain. We therefore transformed gcn5Δ strain with the integrative cassette NDC80::HIS. HIS3 trasformants obtained were analyzed by colony PCR analysis was performed to confirm the correct chromosomal integration of NDC80::HIS cassette and the concomitant presence of the GCN5::KAN.MX in the strain (Fig. 1G). According to our predictions, we obtained the double deleted mutant ndc80Δ-gcn5Δ viable both at permissive and restrictive temperatures (28°C and 34°C). This result reinforces the meaning of our findings thus confirming the role of Gcn5 in the control of the central KT Ndc80 complex and its interaction with spindle microtubules during chromosome segregation.

Ndc80 in the complex macrostructure of kinetochore is an essential subunit of central layer devoted to the attachment of the KMN to spindle microtubules which is the prerequisite for faithful chromosome segregation. Despite the relevance of Ndc80 in kinetochore the mechanisms governing KT-MT interactions with microtubules are still unclear. Acetylation was associated to microtubule stabilization and deacetylation to their destabilization^[22][23], however its role is still debated. Here we demonstate that the absence of Gcn5 allows growth of cells with aberrant kinetochores in mutants for Ndc80 complex components. We demonstrated that lack of Gcn5 allows growth of mutants in the central Ndc80 complex at non permissive temperature. We also obtained an unbiased demonstration of the role of Gcn5 in this process by producing the full deletion of the essential Ndc80 subunit in the double deleted strain ndc80Δ-gcn5Δ. Moreover, we have shown that the recovery of cell growth is linked to the Gim3 subunit of the tubulin chaperone and is regulated by SAC checkpoint Mad2.

In S.cerevisiae the kinetochore is simplified respect to higher Eukaryotes. It binds to a single microtubule while, in humans, it interacts with a complex bundles of multiple microtubules. However, while yeast shows a simplifyed interaction of kinetochore with a single microtubule it offers the unique opportunity to carry on defined genetic analysis to dissect protein networks and functional interactions. As a proof of concept our results suprisingly outlined the effects induced by Gcn5 on the interaction between the budding yeast kinetochore Ndc80 complex and the mitotic spindle.

Our collected results demonstrate that the absence of Gcn5 allows the interaction between aberrant kinetochores missing an essential central structure and the spindle microtubules. Microtubules are subjected to a number of different post translational modifications determining their stability and dynamics^[24]. Despite these studies the mechanism and the role of this regulation is far to be determined. Our observations set the stage to carry on further characterizations on the mechanisms involved in such process. To this end, the ndc80Δ-gcn5Δ strain we have obtained represents an excellent strain for future studies on the mechanisms involved and the identification of additional regulatory components. We believe that our experimental data may lead to additional new experiments to dissect the role of acetylation in the interaction of central kinetochore with the mitotic spindle^[12].

Yeast strains and growth:

All yeast strains used for this work are S.cerevisiae W303 background and are singularly listed in Table 1. Cells were grown in YEPD rich medium or in SD minimal medium for transformants selection (0.67% yeast nitrogen base and 2% glucose supplemented with required amino acids). All strains were grown at 28°C or 34°C. Gene disruption were performed as previously described^[25] and confirmed by colony PCR for specific integration. 5-fold serial dilutions of strains were spotted and grow on solid medium at indicated temperatures. Morphology analysis were performed by counting average of 500/600 cells in duplicates.

Construction of S.cerevisiae yRB027 strain (ndc80Δ-gcn5Δ) was obtained by integration of NDC80::HIS3 cassette into YPO4 gcn5Δ strain.

Oligonucleotides used for NDC80 gene disruption:

NDC80-S1 5'-CGTTAGGGTGTACTATAATATATTCCATAAGAGATACCCGTTTCTATAAAT-3'

NDC80-S2 5'-GGTGGGGCTGAGCTTTGCTGTAGATTGCTCGGGTATTATATATCATTTATTTTATTAATCGATGAATTCGAGCTCG-3'

NDC80-S3 5'-GGGCTGGTTGAGAGCCCCG-3'

This work was supported by Sapienza Ateneo Project, C.C. by a FIRB (Basic Research Investment Fund) - Cenci Bolognetti International Grant.

We wish to thank Simonetta Piatti and Peter de Wulf laboratories for providing S.cerevisiae kinetochore mutant strains, Giuseppe Pisaneschi for technical assistance.

The authors declare no conflicts of interest.

Not Applicable.

No fraudulence is committed in performing these experiments or during processing of the data. We understand that in the case of fraudulence, the study can be retracted by ScienceMatters.