Emerging studies hint at the roles of autophagy-related proteins in various cellular processes. To understand if autophagy-related proteins influence genome stability, we sought to examine a cohort of 35 autophagy mutants in Saccharomyces cerevisiae. We observe cells lacking Atg11 show poor mitotic stability of minichromosomes. Single-molecule tracking assays and live cell microscopy reveal that Atg11 molecules dynamically localize to the spindle pole bodies (SPBs) in a microtubule (MT)-dependent manner. Loss of Atg11 leads to a delayed cell cycle progression. Such cells accumulate at metaphase at an elevated temperature that is relieved when the spindle assembly checkpoint (SAC) is inactivated. Indeed, atg11∆ cells have stabilized securin levels, that prevent anaphase onset. Ipl1-mediated activation of SAC also confirms that atg11∆ mutants are defective in chromosome biorientation. Atg11 functions in the Kar9-dependent spindle positioning pathway. Stabilized Clb4 levels in atg11∆ cells suggest that Atg11 maintains Kar9 asymmetry by facilitating proper dynamic instability of astral microtubules (aMTs). Loss of Spc72 asymmetry contributes to non-random SPB inheritance in atg11∆ cells. Overall, this study uncovers an essential non-canonical role of Atg11 in the MT-mediated process of chromosome segregation.

Atg11 plays a role in high-fidelity chromosome transmission

To understand the role of autophagy protein(s) in the cell cycle in nutrient-rich conditions, we carried out a phenotypic screen to determine the growth fitness of null mutants of autophagy-related genes (atg) in S. cerevisiae to a MT depolymerizing drug, thiabendazole (TBZ). Among all the 35 atg mutants (S1A Fig), atg11∆ cells displayed maximum sensitivity to TBZ (Fig 1A), supporting an earlier observation [63]. Transgenic wild-type ATG11 complemented the TBZ-induced growth defects in atg11∆ cells (Fig 1A). Furthermore, atg11∆ cells grew as good as wild type at 14 °C but significantly slower at 37 °C (Fig 1B), a condition at which the rate of depolymerization of MTs is more dominant than the rate of polymerization [64]. These results hint towards a role of Atg11 in MT-dependent processes.

Any defects in MT-dependent processes negatively impact the fidelity of chromosome segregation. Indeed, a strain harboring an extra linear chromosome displayed a significantly higher rate of chromosome loss in atg11∆ cells at 30 °C that was further exacerbated at 37 °C (Fig 1C). Similarly, the average mitotic stability of a monocentric circular plasmid was found to be significantly lower in atg11∆ cells than in the wild type but comparable to that of a kinetochore mutant, ctf19∆ (Fig 1D). No other atg mutants tested displayed any decrease in the mitotic plasmid stability. Mutants of key autophagy proteins such as Atg1, essential for both cytoplasm-to-vacuole targeting (Cvt] and autophagy pathways, or Atg17, required during starvation [62], neither displayed any TBZ sensitivity nor any reduction in plasmid stability (Figs 1D and S1A). Using GFP-Atg8 processing assay [65], as previously reported, the atg1∆ cells were defective for autophagy (absence of free GFP), while both atg11∆ and atg17∆ cells have reduced levels of autophagy relative to the wild type (S1B Fig). This suggests that while Atg1 is indispensable for autophagy, it does not play any role in chromosome segregation as shown by our assays. Taken together, we conclude that the loss of Atg11 in S. cerevisiae reduces the fidelity of chromosome segregation leading to an increased frequency of chromosome loss.

Atg11 displays a dynamic association with the SPBs

To understand how Atg11 contributes to MT-dependent processes, we studied the dynamics of Atg11 during the cell cycle progression. We first confirmed the canonical localization of sfGFP-Atg11 onto the vacuolar membrane as punctate signals at all stages of the cell cycle (S2A Fig). In addition, we observed that Atg11 puncta were also at the interface of vacuole and nucleus during the cell cycle (S2B Fig). A protein-protein interaction network analysis of components involved in autophagy and chromosome segregation identified Spc72, one of the components of SPBs, as a key interactor of Atg11 (S3A Fig) [66], raising the possibility that Atg11 could be associated with SPBs during the cell cycle. Live-cell imaging revealed that in 18 out of 24 cases, Atg11 is transiently localized proximal to the dSPB prior to anaphase onset (Fig 2A). Proximal localization was studied by partial overlap between Atg11 (cyan) and Spc42 (magenta), giving a white-colored overlap between the two proteins.

To strengthen our observations of the transient localization of Atg11 to SPBs, we employed single-molecule tracking (SMT) to study the dynamic binding of Atg11 with SPBs and the cytoplasmic Atg11 puncta. For the SMT analysis, we endogenously tagged ATG11 with a HaloTag at its C-terminus for the controlled labeling using HaloTag Ligand (JF646-HTL) (Fig 2B) [67]. The time-lapse movies acquired with slow (200 ms time interval) and fast (15 ms time interval) imaging enabled the quantification of survival time and diffusion coefficient, respectively. We quantified the dwell time of Atg11 over SPBs (Spc42-mCherry) and Atg11 puncta (sfGFP-Atg11) in the cytoplasm from 200 ms time interval movies. The cumulative survival time distribution fits well with the double exponential fit (Fig 2C), suggesting two molecular populations: (a) Atg11 molecules bound with long residence time (specific binding) and (b) bound with short residence time (non-specific binding). We observed specific binding of Atg11 at the cytoplasmic puncta and SPBs (12.3 ± 1.1% and 14.5 ± 2%, respectively, Fig 2C). The mean residence time of Atg11 at SPBs was half as compared to the cytoplasmic puncta (4.4 ± 0.47 s versus 8.6 ± 1.02 s, Fig 2C), suggesting a fast turnover of Atg11 at SPBs compared to the cytoplasmic puncta. We employed Spot-On-based kinetic modeling [68] for robust estimation of the fraction of bound and unbound molecules with their mean diffusion coefficients (Fig 2D–2F). A marginal increase in the bound fraction of Atg11 over SPBs, compared to the cytoplasm (51% versus 46%, respectively) (Fig 2F) was observed. However, the diffusion coefficient of Atg11 over SPBs is three times higher than that of cytoplasm (D = 0.013 µm²/s versus 0.004 µm²/s) (Fig 2F), suggesting more dynamic interactions of Atg11 with SPBs. We also carried out SMT assays in TBZ-treated cells to understand Atg11 dynamics in the absence of MTs. Under MT-depolymerizing conditions, the specific bound fraction of Atg11 to Atg11-cytoplasmic puncta increased by over 2.5-fold and to SPBs by 2-fold (Fig 2G). The residence time of Atg11 reduces marginally in the presence of TBZ at the Atg11-cytoplasmic puncta (8.6 ± 1.02 s to 4.5 ±  2.9 s, Fig 2G), suggesting that the absence of MTs marginally changed its turnover rate. However, the residence time of Atg11 at SPBs increased from 4.4 ± 0.5 s to 41.8 ± 3.9 s upon TBZ treatment (Fig 2G). Based on these observations, we conclude that the dynamic interactions of Atg11 at the SPBs are MT-dependent (Fig 2H).

To support that Atg11 is transiently localized at the SPBs, we utilized a proximity-based bimolecular fluorescence complementation (BiFC) assay with Spc72, a known interactor of Atg11 (S3A and S3B Fig) [69]. A fluorescent punctate signal resulting from the reconstitution of the Venus molecule at the vacuolar periphery, marked by FM4-64, suggested an in vivo interaction between Atg11 and Spc72 (S3B Fig). Co-localization with Spc42 further demonstrated the in vivo interaction between these two proteins at the SPBs (S3C and S3D Fig). The absence of BiFC signals in diploid cells expressing either VN-tagged Spc72 or the VC-tagged Atg11 alone (S3E and S3F Fig), suggested that neither VN nor VC fluoresces on its own. To further probe how Atg11 is associated with SPBs, we first confirmed a previously reported interaction between Atg11 and Spc72 [70] by the yeast two-hybrid (Y2H) assay (S3G and S3H Fig). The Y2H assay with Cnm67, a protein from the outer plaque of the SPB, did not show any interactions with Atg11 (S3I Fig), suggesting the spatial position of Atg11 and Spc72 favors their physical interactions. Taken together, the results from the SMT assays and sub-cellular localization studies, we convincingly demonstrate the existence of a novel and dynamic association of Atg11 at SPBs, in addition to its canonical localization onto the vacuolar membrane crucial for selective autophagy [59].

Atg11 enables the timely progression of the cell cycle

SPBs nucleate both nuclear and aMTs, critical for kinetochore-MT interactions and spindle orientation along the mother-bud axis, respectively. Having established a non-canonical localization of Atg11 at the SPBs, we, sought to examine the status of both nuclear and aMT regulation in the absence of Atg11. First, we probed cell cycle progression in synchronized cells and studied the dynamics of MTs (GFP-Tub1) and SPBs (Spc42-mCherry) in live cells. We observed a delay in spindle elongation and a concomitant delay in the cell cycle progression in atg11∆ cells (Figs 3A, 3B, S4A, and S4B).

We also measured the time taken by atg11∆ cells to enter a) metaphase (spindle length = 1.5–2 µm) after release from G1 arrest and b) telophase upon anaphase onset (spindle length > 2 µm). We observed a significant delay at both these stages in atg11∆ cells (Figs 3B and S4B). Further analysis revealed that the overnight grown (t = 0) atg11∆ cells displayed a significantly increased proportion of large-budded cells compared to the wild type (S4C Fig). When grown at 37 °C for 6 h, the atg11∆ cells further accumulated an increased proportion of large-budded cells (S4C Fig), with a 2-fold higher proportion of large-budded cells with metaphase spindle of 1.5–2 µm compared to the wild type at 37 °C (Fig 3C), hinting towards a delay in anaphase onset.

Defects in kinetochore-MT attachments lead to the activation of the SAC, an error correction mechanism that delays the metaphase-to-anaphase transition [71]. We, therefore, examined whether the SAC sensed atg11-associated defects of delayed anaphase onset. Deletion of MAD2 in atg11∆ cells alleviated the proportion of large-budded cells with unsegregated nuclei demonstrating that indeed SAC senses atg11∆ -associated defects leading to delayed anaphase onset (Fig 4A). Cells of atg11∆ mad2∆ double mutant grew slower at 37 °C and in the presence of TBZ (Fig 4B). The SAC activation leads to the inhibition of the APC/C-Cdc20 complex, stabilizing securin (Pds1), thus preventing the activation of separase. Inactivation of separase impedes sister chromatid separation and therefore blocks metaphase-to-anaphase transition [72]. Consistent with our observation of SAC-mediated defects, atg11∆ cells displayed stabilized Pds1 levels both upon release from nocodazole treatment and G1 arrest (Figs 4C and S5A). Our results, therefore, provide evidence that atg11∆ cells fail to establish correct kinetochore-MT attachments leading to a delayed anaphase onset.

Does Atg11 help in maintaining the structural integrity of kinetochores or in regulating MT dynamics? The CEN transcriptional read-through assay [73] ruled out the role of Atg11 in the maintenance of the kinetochore ensemble (S5B Fig). The kinetochore and SAC mutants display growth defects in the presence of MT poisons due to a defect in chromosome segregation. An increased chromosome mis-segregation upon loss of Atg11 and the metaphase arrest of atg11∆ cells in the presence of TBZ suggest that the fidelity of chromosome segregation is compromised. To test this, we monitored the segregation dynamics of a marked chromosome in a strain with CEN3 labeled with LacO repeats and LacI tagged with GFP [74]. Most wild-type cells displayed bioriented sister kinetochores, as studied by co-localizing Spc110-mCherry with CEN3-GFP, after recovery following nocodazole treatment (Figs 4D and S5C). However, a significant increase in mono-oriented kinetochores was observed in atg11∆ cells. This was further corroborated by live-cell imaging after nocodazole arrest and release (Fig 4E and 4F). Mono-oriented kinetochores do not generate tension and are sensed by Aurora kinase B^Ipl1, which activates SAC by creating unattached kinetochores [10]. With an increased proportion of cells displaying mono-oriented kinetochores and to determine whether the SAC-dependent delay in the cell cycle in atg11∆ cells is due to Ipl1 activation, we deleted ATG11 in ipl1-321 mutant cells [75]. Cells of atg11∆ ipl1-321 double mutant grew slower both at 30 °C and in the presence of TBZ (Fig 4G). These cells failed to arrest and exhibited nuclear mis-segregation defects (Fig 4H), suggesting that the SAC-dependent delay in atg11∆ is due to Ipl1-mediated improper kinetochore–MT interactions. We also asked if Atg11 mediates this function by localizing to the kinetochores by SMT assays. We carried out SMT assays in a strain co-expressing ATG11-Halo and MTW1-mCherry. Our results show no specific binding of Atg11 at the kinetochores (S5D Fig), suggesting that Atg11 primarily functions at SPBs in mediating MT dynamics during the cell cycle.

Fluorescence recovery after photobleaching (FRAP) of the metaphase spindle established that the kinetochore MTs are dynamic in budding yeast [76]. We examined if there is any perturbation in the dynamics of the metaphase spindle upon loss of ATG11. The wild-type cells exhibited 50% fluorescence recovery of the bleached region after 300 s (n = 12 live-cells) (Fig 5A), while the atg11∆ cells took 338 s (n = 11 live-cells) (Fig 5B). The corrected ratio of FRAP between the bleached and unbleached regions of the spindle after full recovery was approximately 70% for both wild-type and atg11∆ cells. As shown previously, the photobleached cells proceeded through anaphase, suggesting that bleaching of the metaphase spindle did not damage the spindle (Fig 5C). However, we observed the migration of the metaphase spindle to daughter cells after recovery only in atg11∆ cells (Fig 5D). Overall, we observed a marginal delay of 38 s in atg11∆ cells to achieve 50% recovery of the bleached region. In summary, our results reveal that Atg11 plays a critical role in cell cycle progression in conditions that destabilize MTs either by nocodazole treatment or by growing cells at 37 °C. The frequency of erroneous kinetochore-MT attachments increases in the absence of Atg11 under such conditions.

As cells progress from metaphase to anaphase, the serine/threonine phosphatase, Cdc14, is first released from the nucleolus into the nucleus by the Cdc Fourteen Early Anaphase Release (FEAR) network in yeast [77, 78]. Cdc14 release is required to inactivate CDKs during exit from mitosis [79–82]. While securin (Pds1) and nucleolar protein Fob1 negatively regulate the FEAR network, the same network is positively regulated by separase Esp1, the kinetochore/spindle protein Slk19, Spo12, Bns1, and the polo kinase Cdc5 [80,83–86]. Since Pds1 levels are found to be stabilized in the absence of Atg11, we examined whether the atg11∆ cells display any delay in Cdc14 release. We probed for Cdc14 (Cdc14-GFP) dynamics in alpha-factor synchronized G1 cells. Indeed, atg11∆ cells showed a significant delay in Cdc14 release from the nucleolus (S6A and S6B Fig).

The FEAR pathway regulates multiple anaphase events, including timely exit from mitosis, midzone spindle stabilization, segregation of ribosomal DNA, and nuclear positioning [87]. Separase and Slk19 regulate the stabilization of fragile spindle midzone during anaphase by targeting the Ipl1-Sli15-Bir1 complex to the spindle midzone in early anaphase, which in turn facilitates the recruitment of the spindle-stabilizing protein Slk19 [88–90]. The absence of either separase or Slk19 leads to the collapse of anaphase spindle. We observed MT buckling in atg11∆ cells at anaphase, although in a small proportion of cells (5% of the live-cell) (Fig 6A). The FEAR network is also vital for nuclear positioning and is required to activate pulling forces at aMTs in the mother cell during anaphase. Loss-of-function of separase results in the movement of mitotic spindles into daughter cells called the ‘daughterly phenotype’ [91]. atg11∆ cells displayed this phenotype in 14% of the live cells analyzed (n = 29 live-cell) (Fig 6B), along with anaphase spindle elongation in the mother cell (Fig 6B, 10% of the live cells). Kar9, a MAP, that regulates nuclear positioning at metaphase, is known to be dephosphorylated in early anaphase and is suggested to be a target of Cdc14 [28, 29]. We, therefore, examined whether Atg11 influences the Kar9- or dynein-dependent pathway of spindle positioning by examining the effect of low temperature, known to enhance the binucleated cell phenotype [43]. The dyn1∆ , kip2∆ , and dyn1∆ kip2∆ double mutant strains exhibited a significant increase in the proportion of binucleated cells as expected (Fig 6C) [92]. Remarkably, atg11∆ dyn1∆ double mutant cells had a significantly increased proportion of binucleated cells. These cells also displayed mitotic exit with mispositioned spindles, marked by the accumulation of multinucleated and anucleate cells (Fig 6C). The kar9∆ and atg11∆ kar9∆ mutants had more binucleated cells than those of the wild-type (Fig 6C). However, the proportion of binucleated cells of the atg11∆ kar9 double mutant was similar to kar9∆ (Fig 6C), suggesting that Atg11 contributes to spindle positioning in a Kar9-dependent manner. Since Kar9 regulates spindle positioning at metaphase, we, therefore, examined whether the inactivation of Atg11 perturbed the spindle alignment and migration of metaphase spindles to the bud-neck. We measured the distance between the metaphase spindle and the bud-neck along with the angle between the spindle axis and the mother-bud axis. The atg11∆ cells displayed a 2-fold increase in the proportion of cells having a metaphase spindle at the central region in the mother cell (Fig 6D), as well as a significantly higher angle of alignment of the metaphase spindle than the wild-type (Fig 6E).

Kar9 is asymmetrically localized at the dSPB, which in turn is required for correct spindle positioning in the wild-type cells. Therefore, we examined if the defects in spindle positioning upon loss of Atg11 are due to alterations in the asymmetric localization of Kar9 at the dSPB. To address this, we analyzed the distribution of Kar9, tagged with GFP, in metaphase cells co-expressing Spc72-mCherry that marks the SPB. While 16% of wild-type cells displayed weak asymmetric Kar9 localization, atg11∆ cells displayed a 2-fold increase in the proportion of cells with weak asymmetric Kar9 localization (Fig 6F). This loss in Kar9 asymmetry in atg11∆ cells explains the 2-fold increase in the proportion of cells having a metaphase spindle at the central region in the mother cell. Taking these observations together, our results suggest that Atg11 facilitates timely Cdc14 release from the nucleolus during early anaphase upon timely degradation of securin, which is necessary for spindle stabilization and Kar9-dependent spindle positioning and alignment in budding yeast.

Atg11 facilitates dynamic instability of aMTs via timely degradation of Clb4

Dynamic instability of MTs also facilitates the interaction of MTs with the bud cortex, crucial for spindle orientation and positioning in budding yeast. Various factors like motor proteins, MT-binding proteins, and MT-nucleation proteins regulate MT dynamics and proper positioning of the mitotic spindle [92–94]. We, therefore, sought to understand if Atg11 additionally contributes to the dynamic instability of aMTs in budding yeast by being localized at the SPBs. We studied aMT (GFP-Tub1) dynamics in metaphase cells of wild-type and atg11∆ cells (Figs 7A and 7B). The atg11∆ cells displayed a significantly increased catastrophe frequency (Fig 7C), while the rescue frequencies were comparable to those of the wild-type (Fig 7C). A closer examination revealed a significant increase in the number of catastrophe events, decreased aMT length at catastrophe and lower time spent in polymerization before catastrophe in atg11∆ cells (S7A, S7B and S7C Fig). The atg11∆ cells also showed a significant increase in the number of rescue events (S7D Fig). However, there was no difference observed in the minimum length of aMTs and the time spent for depolymerization before the rescue onset between wild-type and atg11∆ cells (S7E and S7F Fig). Remarkably, loss of Atg11 lowered both polymerization and depolymerization rates of aMTs in atg11∆ cells (Fig 7C). Consequently, the length of aMTs in atg11∆ cells was also significantly shorter at the metaphase and anaphase stages of the cell cycle (S7G–S7I Fig). These results revealed that Atg11 facilitates dynamic instability of aMTs and prevents catastrophe frequency in budding yeast.

The aMT dynamics are tightly regulated in S. cerevisiae, switching from an unstable to a stable state at anaphase onset. This transition requires the degradation of the mitotic cyclin Clb4 by the APC/C^Cdc20 complex via the proteasomal degradation pathway, which is crucial for the establishment of proper spindle positioning [95]. The unstable aMTs in atg11∆ cells led us to investigate if the Clb4 levels were altered. The wild-type cells showed cycling levels of Clb4, being highest at metaphase and levels dropped as cells entered anaphase both at 37 and 30 °C (Figs 7D and S8A). Consistent with our observation of unstable aMTs in atg11∆ cells, we observed stabilized Clb4 levels with loss of Atg11 (Figs 7D and S8A). Taken together, these results suggest that atg11∆ cells fail to facilitate the timely degradation of Clb4 crucial for aMT stability at anaphase onset in budding yeast.

Both asymmetric localization of Kar9 at the dSPB and proper integrity of aMTs ensure the asymmetric inheritance of the old SPB (dSPB) into the daughter bud [28,36]. With a random Kar9 distribution during metaphase and shorter aMTs due to loss of Atg11, we examined whether atg11∆ cells display any defects in SPB inheritance. To test this possibility, we tagged Spc42 (SPB marker) with mCherry. The slow folding kinetics of mCherry and ordered assembly of SPBs allow age discrimination with the new SPB (mSPB) being significantly dimmer than the dSPB [51]. Deletion of KAR9, known to display random SPB inheritance [36], led to a randomized SPB inheritance relative to the wild-type (Fig 8A). Strikingly, atg11∆ cells showed a 4-fold increase in SPB inheritance defects (Fig 8A). Notably, autophagy mutant cells lacking a key autophagy protein Atg1 segregated SPBs like in the wild-type cells (Fig 8A), hinting that SPB inheritance is autophagy-independent in budding yeast. A previous report on the localization interdependency of Spc72 and Kar9 onto the SPBs, placed Spc72 upstream of Kar9 [96], while a recent study demonstrates Cdc5 as a molecular timer facilitating timely recruitment of both Spc72 and Kar9 onto the SPBs [97]. Based on the SPB inheritance defects, we asked whether Atg11 contributes to the asymmetric localization of Spc72 and Kar9 at the SPBs. Strikingly, the loss of ATG11 randomized the SPB localization of Spc72 (Fig 8B and 8C). Taken together, we conclude that being localized at the SPBs, Atg11 prevents catastrophe frequency of aMTs, crucial for proper aMT length and asymmetric SPB inheritance in budding yeast.

Fig 9. SPB-associated Atg11 ensures high-fidelity chromosome segregation in budding yeast.

This study reveals two distinct pools of Atg11 in S. cerevisiae. A lower proportion of Atg11 which is less dynamic canonically localizes at the pre-autophagosomal structures (PAS) onto the vacuolar membrane [59], necessary for selective autophagy and dependent on actin [104,105]. While the more dynamic Atg11 localizes at the SPBs in a higher proportion which is critical for high-fidelity chromosome segregation. The asymmetric Kar9 localization together with the dynamic instability of MTs is crucial for pre-anaphase spindle positioning and alignment, and chromosome biorientation. Before anaphase onset, aMTs are unstable and cyclin Clb4 levels are high. The mono-oriented kinetochores have defective tension, leading to the activation of Ipl1. Ipl1 carries out the disassembly of defective kinetochore-MT attachments via phosphorylation and activates SAC. SAC delays anaphase onset until biorientation is achieved. Upon biorientation, SAC is silenced, Pds1 gets degraded, separase is activated, and Cdc14 is released from the nucleolus via the FEAR pathway leading to proper sister chromatid separation [134]. APC-C/Cdc20 also degrades Clb4 upon anaphase onset crucial for the stability of aMTs. Atg11 ensures asymmetric localization of both Spc72 and Kar9 at the dSPB, critical for non-random SPB inheritance.

https://doi.org/10.1371/journal.pbio.3003069.g009

Yeast strains, plasmids, and primers

All the strains used in this study are specified in S1 Table. Primer sequences and the plasmids used for strain construction are specified in S2 and S3 Tables, respectively.

Media and growth conditions

Yeast strains were grown on YPD media (yeast extract 1%, peptone 2%, and dextrose 2%) at 30 °C unless stated otherwise. For the growth assay in the presence of TBZ, overnight grown cultures of wild-type, autophagy mutants, ctf19∆ , mcm16∆ , and ATG11-complemented strains were serially diluted and spotted on YPD containing 50 µg mL⁻¹ and imaged on 2 dpi or 36 h. The analyses for Figs 3C and 7A–7C were done by growing the cells at 37 °C for 6 h.

Interaction network

First, we constructed a protein-protein interaction map by identifying the proteins that have been reported to either genetically or physically interact with each of the Atg proteins. We further extended the network to include the proteins that are known to interact with any of the primary interactor proteins. We gleaned these connections using Perl scripts to query the latest curation of the BioGRID yeast interactomes [66]. To this end, only those protein-protein interactions were retained which were confirmed and corroborated across three or more independent studies.

Construction of yeast strains

The ATG11 gene was deleted from the S. cerevisiae genome by one-step gene replacement [122]. A 40-bp homology region corresponding to the upstream and downstream sequence of ATG11 having HPH marker providing resistance against hygromycin B was amplified from pAG32 using HR34 and HR35 primer pairs. Similarly, ATG11 was deleted in F2351 [51], with Kar9-GFP and Spc72-mCherry tagged, and in SBY322 [ipl1-321, [75]] using primers (HR34 and HR35) carrying 40-bp homology regions from upstream and downstream of the ATG11 gene that were used to amplify the HPH marker sequence from pAG32. The ATG11 gene deletion in the desired transformants was confirmed by PCR using 5′ and 3′ junction primers (HR36 and HR37).

Double deletion mutants were obtained either by transforming the above ATG11 deletion cassette in a single deletion mutant background or by crossing a single deletion mutant in MATa background having the G418 marker with atg11∆ cells in MATα (BY4742) background having the HPH marker. The diploid cells were sporulated and after digestion of the cell wall and vortexing, the cells were plated first on YPD with hygromycin. Growing colonies were patched on G418 and colonies grown on both the selection plates were mated again with both MATa (BY4741) and MATα (BY4742) to differentiate between haploid and diploid cells. Cells showing shmoo formation with only one mating type were taken forward for the assays.

For atg11∆ complementation, a ~ 3.9 kb fragment comprising the upstream and full-length sequence of ATG11 was fused with a ~ 800 bp of HIS3 fragment with its promoter having a 40-bp homology sequence of ATG11 at its 3′ end. The purified PCR product was transformed in the atg11∆ cells and the desired transformants were confirmed using HR116 and HR117 primers.

Construction of a strain expressing GFP-Tub1 and Spc42-mCherry under its native promoter

To study the dynamics of MTs, TUB1 was tagged with GFP using the plasmid, pAFS125 [123]. To visualize SPBs, Spc42, a component of the central plaque of SPB, was tagged with mCherry using the primers HR142 and HR143.

Construction of MAD2 deletion strain

The MAD2 gene was deleted by one-step gene replacement [122], having a 40-bp homology region each from upstream and downstream of the gene using HR52 and HR53 primer pairs. These primers were used to amplify the HPH marker from pAG32. The desired deletion strain was confirmed by PCR using 5′ and 3′ junction primers (HR66 and HR67).

Construction of Pds1-6xHA tagged strain

A PCR fragment containing 6xHA and the HPH selection marker was amplified from pYM16 [124] using the primer pair RA140 and RA141. To obtain Pds1-6xHA, the PCR product was transformed in wild-type BY4741 cells generating the strain ScRA403. Transformants were selected using the hygromycin selection marker. To tag Pds1 in atg11Δ, MATa RA403 was crossed with MHR08 (MATα). Diploids were selected using methionine and lysine prototrophy. Diploids were sporulated and haploid progeny were identified by mating with both MATa and MATα strains. The desired strain (ScRA405) was confirmed for both ATG11 deletion and Pds1 tagging using the primer sets HR36 and HR37 and RA146 and RA147, respectively. Pds1-6xHA tagging both in wild-type and atg11∆ cells were confirmed by western blot analysis.

Construction of strains for Bimolecular Fluorescence Complementation (BiFC) assay

We used N-terminal protein tagging vectors, pFA6a-KanMx6-pGAL1-VN173 and pFA6a-His3Mx6-pGAL1-VC155 for tagging either Atg11 or Spc72 [69]. VN and VC designate N-terminal and C-terminal fragments of Venus, a variant of yellow fluorescent protein, respectively [69]. VN173-Atg11 and VC155-Atg11 were amplified using HR296 and HR297 and HR296 and HR298 primer pairs respectively. Similarly, VN173-Spc72 and VC155-Spc72 were amplified using HR299 and HR300 and HR299 and HR301 primer pairs respectively. VN-tagged fragments with the G418 selection and VC-tagged fragments with the HIS3 selection were transformed into BY4741 and BY4742 backgrounds respectively. The haploid colonies obtained were mated and selected on G418 plates without histidine to obtain diploid cells expressing both fragments of the fluorophore. The control diploid cells expressing either VN-Spc72 or VC-Atg11 co-expressing Spc42-mCherry were generated by mating BY4741 expressing VN-Spc72 with BY4742 or BY4742 expressing VC-Atg11 with BY4741.

Construction of strains for Single Molecule Tracking (SMT) assay

The PDR5 gene, coding for the membrane transporter protein Pdr5 was deleted to allow the retention of the HaloTag ligands (JF646) inside the yeast cell [125]. The PDR5 gene was deleted in both MHR90 MATa strain (sfGFP-Atg11 Spc42-mCherry) and yTB283 MATa strain (sfGFP-Atg11) using a 40-bp homology region each from upstream and downstream of the PDR5 gene using HR333 and HR334 primer pairs. These primers were used to amplify the HPH marker from pAG32. The desired deletion strain was confirmed by PCR using 5′ and 3′ junction primers (HR335 and HR336).

MHR411_a with PDR5 deletion and sfGFP-Atg11 were transformed with Mtw1-mCherry cassette amplified using RA177 and RA178 primer pairs from pAW8-mCherry plasmid. Mtw1 expressing mCherry was confirmed using fluorescence microscopy. For labeling Atg11 using HaloTag Ligand (HTL), the ATG11 gene was fused with -HaloTag using the URA3 selection marker by homologous recombination. The HaloTag cassette was amplified from pTSK561 (Addgene: 190816) using HR329 and HR330 primer pairs and transformed in yTK1501 MATα strain. The tagged strains were confirmed by HR36 and HR37 primer pairs. The diploid cells were selected on CM-ura media and YPD supplemented with hygromycin and geneticin.

Construction of Clb4-9xMyc tagged strain

A PCR fragment comprising 9xMyc and HIS3 selection marker was amplified from pYM19 [124] using the primer pair HR323 and HR324. The cassette was transformed in both wild-type and atg11∆ cells. The transformants were selected on CM-his drop-out media. The transformants expressing Clb4-9xMyc were confirmed by western blot analysis.

Stability of monocentric plasmid

The atg1∆ , atg6∆ , atg11∆ , atg15∆ , atg17∆ , ctf19∆ , and the isogenic wild-type strains were transformed with pRS313 [126], a centromeric plasmid, and transformants were selected on a dropout media lacking histidine (SD-his). The transformants were then grown in a non-selective media, YPD, for 7–10 generations, followed by plating to obtain single colonies. The single colonies were subsequently patched onto SD-his (selective) and YPD (non-selective) media. The mitotic stability was calculated by the number of colonies that were able to grow on selective media (SD-his) per 100 colonies grown on non-selective media (YPD).

Chromosome loss assay

The chromosome loss assay was performed in a strain harboring an extra non-essential chromosome that carries SUP11 that suppresses ade2 mutation in the assay strain, YMH58a as previously described [127]. Cells retaining this extra chromosome grow as white-colored colonies, complementing the ade2 mutation. Loss of extra chromosome leads to the red-pigmented colony, displaying ade2 mutant phenotype, implying loss of SUP11. A single white colony of wild-type, atg11∆ , and ctf19∆ cells were grown in non-selective media (YPD) for 5–7 generations, both at 30 and 37 °C. The cells were plated on YPD. The number of colonies showing chromosome loss at the first division was scored for each condition in three independent biological experiments.

In vivo assay for kinetochore integrity: CEN transcriptional read-through assay

The CEN transcription read-through assay was done as previously described [73]. Briefly, the reporter plasmid, pAKD06+CEN4 contains a β-galactosidase reporter gene with the selectable marker URA3 was transformed in wild-type, ctf19∆ and atg11∆ cells and selected on SD-ura media. The single colony was inoculated in SD-ura media, containing 2% galactose and 0.3% raffinose and was grown to 1 OD₆₀₀. The cell pellet was washed with water and resuspended in 1 mL of Z-buffer (60 mM Na₂HPO₄, 40 mM NaH₂PO₄.2H₂O, 10 mM KCl, 5 mM β-ME). 0.1 mL of this was taken for the determination of OD₆₁₀. 0.1 mL of Z-buffer was added to the remaining 0.9 mL of cells. The cells were then permeabilized by adding 50 µL of 0.1% SDS and 100 µL of chloroform (incubated at 30 °C for 15 min). 0.2 mL of 4 mg mL⁻¹ ONPG (ortho-nitrophenyl-β-galactoside) was added to the cell suspension and incubated at 30 °C till yellow color developed. The reaction was stopped by adding 0.5 mL of 1 M Na₂CO₃. Cells were spun down at 10,000g and the clear supernatant was transferred to a fresh tube. The optical density of this solution was measured at 420 and 550 nm. The enzyme activity was normalized with respect to cell density. Units of β galactosidase were measured as:

Live-cell imaging

For tethering the cells to the glass-bottom dishes for live-cell imaging, the dishes were treated with 6% Concanavalin A (Catalogue no. C2010, Sigma-Aldrich) for 10 min. Wild-type and atg11∆ cells carrying GFP-Tub1 (spindle) and Spc42-mCherry (SPBs), or Cdc14-GFP and Tub1-mCherry, or sfGFP-Atg11 and Spc42-mCherry were grown overnight in CM medium, re-inoculated in CM at 0.2 OD₆₀₀ and grown for 3 h. The cells were first arrested at G1 using 3.0 µg mL⁻¹ α-factor at 30 °C (Merck, Cat. No. T6901). After 90 min, 3.0 µg mL⁻¹ α-factor was again added and incubated for another 90 min. Cells were washed thrice with CM medium and released into Concanavalin A-treated glass bottom dishes and incubated at 30 °C or 37 °C for 10 min. Live-cell imaging was performed on an inverted confocal microscope (ZEISS, LSM880) equipped with a temperature-controlled chamber (Pecon incubator, XL multiSL), a Plan Apochromat 63× NA oil 1.4 objective and GaAsP photodetectors. For time-lapse microscopy, images were captured at a 4 min interval with 5% intensity exposure with 0.5 µm Z-steps using 488 and 587 nm for GFP and mCherry as excitation wavelength. All the images displayed were maximum intensity projections of images (except for Fig 2A, which is represented as a single plane image) for each time using ImageJ.

Microscopic image acquisition and processing

Wild-type and deletion mutants carrying either GFP-Tub1 (spindle) or Spc42-mCherry (SPB marker) were grown overnight in YPD medium and re-inoculated in YPD at 0.2 OD₆₀₀ for 3 h at 30 °C. The cells were pelleted at 3,000g and mounted on a glass slide having a 2% agarose pad supplemented with 2% dextrose. The pad was sealed with a coverslip. The cells were imaged using an inverted microscope (ZEISS, Axio Observer.Z1/7), a Plan-Apochromat 100×/1.40 oil DIC M27. The filters used were GFP/FITC 488, mCherry 561 for excitation, GFP/FITC 500/550 bandpass or GFP 495/500 and mCherry 565/650 or mCherry 580/750 bandpass for emission. Z-stacks at an interval of 0.5 µm were taken and maximum intensity projection for each image using ImageJ was projected.

Biorientation assay

For the CEN3 biorientation assay, overnight grown wild-type and atg11∆ cells were reinoculated to 0.2 OD₆₀₀ and grown for 3 h. The cells were arrested at metaphase in the presence of 15 µg mL⁻¹ of nocodazole for 2 h. The cells were then released in YPD in the presence of 50 µ M bathocuproinedisulfonic acid disodium salt (BCS) at 30 or 37 °C for 1 h. Cells with two CEN3-GFP puncta between the two SPBs (Spc110-mCherry) were considered bioriented, while those with one CEN3-GFP punctum closer to either of the SPBs or away from the polarity axis were considered mono-oriented. The large-budded cells with a budding index of >0.6 were used for the analysis.

For the CEN3 biorientation assay using live-cell, the overnight grown wild-type and atg11∆ cells were re-inoculated to 0.2 OD₆₀₀ and grown for 4 h in CM in the presence of 50 µ M BCS and grown at 30 °C. The cells were arrested at metaphase in the presence of 15 µg mL⁻¹ of nocodazole for 2 h. The cells were then released in CM in the presence of 50 µ M BCS and live-cell was carried out at room temperature. The images were captured at 512×512 frame size with 5 min intervals.

Western blot for Pds1 dynamics after nocodazole arrest

Overnight grown cells were diluted to 0.1 OD₆₀₀ for wild-type and 0.15 for atg11∆ cells in 25 mL complete media. The cultures were then grown at 30 °C until 0.3–0.4 OD₆₀₀. The cells were first arrested at G1 using 2.5 µg mL⁻¹ α-factor grown at 25 °C (Cat. No. T6901, Merck). After 90 min, 3.0 µg mL⁻¹ α-factor was again added and incubated for another 90 min. Cells were washed with water thrice and released into YPD media containing 15 µg mL⁻¹ nocodazole (Cat. No. M1404, Merck) for 90 min. Metaphase-arrested cells were washed thrice in water and released in 25 mL YPD. The cultures were grown at 30 °C and 1 OD₆₀₀ equivalent culture was collected and pelleted every 20 min. The cell pellets were lysed by TCA precipitation and resuspended in 50 µ L lysis buffer (0.1N NaOH, 1% SDS and 5× protein loading dye). Pds1 degradation dynamics were analyzed using a western blot. Pds1 levels were monitored using rat anti-HA antibody (Cat. No. 11867423001, Roche) and peroxidase-conjugated rabbit anti-rat secondary antibody (Cat. No. HPO14, Genei) at 1:3,000 and 1:10,000 dilution respectively. GAPDH levels were analyzed using mouse anti-GAPDH antibody (Cat. No. MA5-15738, Invitrogen) and HRP conjugated goat anti-mouse secondary antibody (Cat. No. ab97023, Abcam) at 1:5,000 and 1:10,000 dilution respectively. The blots were developed using a Chemiluminescence substrate (BioRad) and imaged using the ChemiDoc imaging system (BioRad). Pds1 levels were normalized with their respective loading control at every time point for both wild-type and atg11∆ cells.

Western blot for Clb4 dynamics after nocodazole arrest

Overnight grown cells were diluted to 0.2 OD₆₀₀ for wild-type and 0.25 for atg11∆ cells in 25 mL complete media. The cultures were then grown at 30 °C until 0.4–0.5 OD₆₀₀. The cells were first arrested at G1 using 2.5 µg mL⁻¹ α-factor grown at 25 °C (Cat. No. T6901, Merck). After 90 min, 3.0 µg mL⁻¹ α-factor was again added and incubated for another 90 min. Cells were washed with complete media thrice and released into YPD media containing 15 µg mL⁻¹ nocodazole (Cat. No. M1404, Merck) for 90 min. Metaphase-arrested cells were washed twice in water and once with YPD, and released in 50 mL YPD. The cultures were grown at 30 or 37 °C and 1 OD₆₀₀ equivalent culture was collected and pelleted every 20 min. The cell pellets were lysed by TCA precipitation and resuspended in 80 µ L lysis buffer (0.1N NaOH, 1% SDS and 5× protein loading dye). Clb4 degradation dynamics were analyzed using a western blot. Clb4 levels were monitored using mouse anti-Myc antibody (Cat. No. 11867423001, Roche) and peroxidase-conjugated rabbit anti-mouse secondary antibody (Cat. No. ab97023, Abcam) at 1:5,000 and 1:10,000 dilution respectively. GAPDH levels were analyzed using mouse anti-GAPDH antibody (Cat. No. MA5-15738, Invitrogen) and HRP conjugated goat anti-mouse secondary antibody (Cat. No. ab97023, Abcam) at 1:5,000 and 1:10,000 dilution respectively. The blots were developed using a Chemiluminescence substrate (BioRad) and imaged using the ChemiDoc imaging system (BioRad). Clb4 levels were normalized with their respective loading control at every time point for both wild-type and atg11∆ cells.

GFP-Atg8 processing assay

The atg1∆ , atg11∆ , atg17∆ , and the isogenic wild-type strains were transformed with GFP-Atg8 pRS423, a 2 µm plasmid, and selected on a dropout media lacking histidine (SD-his). The overnight growing cells in SD-his media were reinoculated to 0.25 OD₆₀₀ in SD-his till the OD reached 1.0. The cells were pelleted and washed once with a nitrogen starvation medium (yeast nitrogen base +  2% dextrose). 1 OD cell/mL cells were transferred to a pre-warmed nitrogen starvation medium (30 and 37 °C). 1 mL cells corresponding to 1 OD were aliquoted at 0, 2, 4, and 6 h post incubation in a nitrogen starvation medium. The cell pellets were lysed by TCA precipitation and resuspended in 50 µ L lysis buffer (0.1N NaOH, 1% SDS, and 5× protein loading dye). GFP-Atg8 processing assay was analyzed using a western blot. GFP-Atg8 and free GFP levels were monitored using mouse anti-GFP antibody (Roche, Cat. No. 11814460001) and peroxidase-conjugated rabbit anti-mouse secondary antibody (Cat. No. ab97023, Abcam) at 1:6,000 and 1:10,000 dilution, respectively. The blots were developed using a Chemiluminescence substrate (BioRad) and imaged using the ChemiDoc imaging system (BioRad).

Bimolecular Fluorescence Complementation (BiFC) assay

The overnight growing cells co-expressing Spc72 fused to the N-terminal half of Venus and Atg11 tagged to the C-terminal half of the same fluorophore were reinoculated to 0.2 OD₆₀₀ in YEP containing 2% galactose (YPG) for 3 h. To label the vacuolar membrane, FM4-64 was added at a concentration of 7.5 μM for 30 min. The cells were washed twice with 1× PBS and further incubated for 1 h in YPG before microscopy.

Staining protocols

FM4-64 staining was performed to visualize the vacuolar membrane. Briefly, FM4-64 (Cat. No. T3166, Thermo Fischer Scientific) was added at a concentration of 7.5 μM for 30 min. The cells were washed twice with 1× PBS and further incubated for 1 h in YPG before microscopy.

Nuclei were visualized by diamidino-2-phenylindole (DAPI) (Cat. No. 62248, Thermo Fischer Scientific) staining. Briefly, the cells were fixed in 3.7% formaldehyde for 10 min. The cells were washed and treated with 0.05% Triton X-100 for 5 min. The cells were resuspended in 1× PBS and stained with 5 ng mL⁻¹ DAPI for 10 min and imaged.

Vacuoles were stained with 7-amino-4-chloromethyl-coumarin (CMAC) (ThermoFisher Scientific, Cat. No. C2110; CellTracker Blue CMAC dye) dissolved in DMSO. Cells were incubated with 10 µM dye at room temperature for 90 min, washed with 1× PBS thrice, and were imaged.

Yeast two-hybrid assays

The full-length ATG11 (~3,534 bp), SPC72 (~1,869 bp), and CNM67 (~1,746 bp) were PCR amplified and cloned in frame with the Gal4 activation/DNA binding domain (AD or BD domain) in pGADC1 and pGBDC1 respectively at BamHI and SalI sites using oligos listed in S2 Table. Yeast-two hybrid analyses were performed in the strain PJ69-4A as described previously [128]. Briefly, bait and prey plasmids were co-transformed into the S. cerevisiae strain PJ69-4A. Positive interactions were scored by the appearance of white-colored colonies on a synthetic defined medium containing 6 µg mL⁻¹ adenine and/ or by the growth in the presence of 2 mM 3-Amino-1,2,4-triazole (3-AT) (Cat. No. A8056, Sigma-Aldrich).

aMT dynamics

The overnight grown wild-type and atg11∆ cells carrying GFP-Tub1 were reinoculated to 0.2 OD₆₀₀ and were grown for 3 h in CM-ura medium. The cells were first arrested at G1 using 3.0 µg mL⁻¹ α-factor grown at 25 °C (Merck, Cat. No. T6901). After 90 min, 3.0 µg mL⁻¹ α-factor was again added and incubated for another 90 min. Cells were washed thrice with prewarmed CM-ura medium at 37 °C and released in the same medium at 37 °C till the cells reached metaphase (1.5–2.0 µm). For tethering the cells to the glass-bottom dishes for live-cell imaging, the dishes were treated with 6% Concanavalin A (Cat. No. C2010, Sigma-Aldrich) for 10 min. Cells were resuspended in CM-ura medium and were allowed to adhere for 10 min at 37 °C. The unadhered cells were washed and the adhered cells were resuspended in 3 mL of CM-ura and proceeded for live-cell imaging at 37 °C. The images were acquired every 15 s at 1024×1024 frame size for a total of 25 cycles. At each time point, images were taken at 0.4 µm apart, covering a total depth of 4.5–5.5 µm. Images were processed by a deconvolution process automatically performed by the Zeiss software. Image analysis was done using Fiji. As described in [95,129], various events were defined as follows: (a) catastrophe frequency—the total number of polymerization-to-depolymerization transitions divided by the total time of all growth events; (b) rescue frequency—the total number of depolymerization-to-polymerization transitions divided by the total time of all shrinkage events; (c) polymerization rate—divide the net change in length by the change in time for each growth event; (d) depolymerization rate—divide the net change in length by the change in time for each shrinkage event. These events were defined only for the bud-directed aMTs and not from aMTs emanating from dSPB.

Single Molecule Tracking (SMT) assay

The diploid cells tagged with one copy of Atg11 with sfGFP and the other copy tagged with the HaloTag and co-expressing either Spc42-mCherry or Mtw1-mCherry were treated with 40 nM JF646-HTL for 30 min [67]. For carrying out SMT upon TBZ treatment, the cells were first arrested at metaphase in the presence of 200 µg mL⁻¹ TBZ for 90 min and then were treated with 40 nM JF646-HTL for 30 min. The cells were then washed thrice for 15 min with either CSM +  TBZ or CSM media to remove the unbound JF646-HTL. The cells were then imaged for ~ 1 h using a Leica DMi8 infinity TIRF inverted fluorescence microscope equipped with a Photometric Prime95B sCMOS camera, 100× 1.47 NA TIRF objective lens, 638 nm 150 mW laser module. Time-lapse movies of single molecules were acquired with a 15 ms time interval (total time points collected: 1,000; exposure time: continuous acquisition with 15 ms camera processing time) and a 200 ms time interval (total time points collected: 200; exposure time: 50 ms).

For dwell time analysis, the particle tracking was performed (for 200 ms time interval movies) using the “TrackRecord” software developed in Matlab (The Matworks) [67,130,131]. The software provides automated features for particle detection (using intensity thresholding), and tracking (using the nearest neighbor algorithm with molecules allowed to move a maximum of 6 pixels from 1 frame to the next, and only tracks that are at least 4 frames or longer are kept). Gaps in the tracks due to photoblinking of the fluorescent dye were closed upto 4 frames. The dwell time was determined by fitting the survival distribution [130] after photobleaching correction. Briefly, the survival histogram was generated from the time that each particle was stationary. In practice, even tightly bound particles move slightly due to chromatin and nuclear motion, and therefore a maximum frame-to-frame displacement of 470 nm (Rmin), and a two-frame displacement of 610 nm (Rmax) (both obtained from the motion of the chromatin-bound histone H3 (Hht1, [132]) have been used to define bound portions of each particle’s track. Because there is a chance that even a fast-diffusing molecule will move less than these thresholds, a further constraint on the minimum number of time points in the bound segment for each particle (Nmin) is used to reduce <1% the contribution of diffusing molecules to the survival histogram. The Nmin value used for 200 ms time-interval movies was 7. The total bound fraction is then calculated as the ratio of bound track segments to the total number of particles.

To extract residence times, the survival distribution, S(t), is fit by least squares to a mixed exponential decay with two rate constants, k_ns = 1/T_ns and k_s =  1/T_s:

where B is the bound fraction, and F_ns is the fraction of particles non-specifically bound. To check for over-fitting, the distribution is also fit to a single-component exponential:

The fits are compared using an F-test to ensure that the two-component model gives a significant improvement over the single-component decay.

For quantifying diffusion parameters (fraction of bound/unbound molecules, diffusion coefficient), particle tracking was performed (from 15 ms time-interval movies) using DiaTrack Version 3.05 (Vallotton and Olivier 2013), with the following settings [67]: remove blur 0.02, remove dim 50–250, maximum jump 6 pixels, where each pixel is 110 nm. This software determines the precise position of single molecules by Gaussian intensity fitting and assembles particle trajectories over multiple frames. The trajectory data exported from Diatrack was further converged into a single.csv file using a custom computational package ‘Sojourner’ (https://rdrr.io/github/sheng-liu/sojourner/). The Spot-On analysis was performed on three frames or longer trajectories using the web interface https://spoton.berkeley.edu/ [68]. The bound fractions and diffusion coefficients were extracted from the cumulative distribution function (CDF) of observed displacements over different time intervals. The cumulative displacement histograms were fitted with a two-state model.

where F₁ and F₂ are bound and free fractions, σ is single molecule localization error, D₁ and D₂ are diffusion coefficients of bound and free fractions, and Z_CORR is the correction factor for fast molecules moving out of axial detection range [68]. The axial detection range for JF646 on our setup is 294 nm. The following settings were used on the Spot-On web interface: bin width 0.01, number of time points 6, jumps to consider 4, use entire trajectories—No, Max jump (µm) 1.5. For model fitting the following parameters were selected: D_bound (µm²/s) min 0.0001 max 0.5, D_free (µm²/s) min 0.5 max 5, F_bound min 0 max 1, Localization error (µm)-Fit from data-Yes min 0.01 max 0.1. Use Z Correction-No, Model Fit CDF, Iterations 3.

Fluorescence recovery after photobleaching (FRAP) of GFP-tubulin in live-cells

The FRAP assay was carried out as described previously [76]. Briefly, the overnight grown wild-type and atg11∆ cells carrying GFP-Tub1 were reinoculated to 0.2 OD₆₀₀ and were grown for 3 h in CM-ura medium. For live-cell imaging, the glass bottom dishes were treated with 6% Concanavalin A (Cat. No. C2010, Sigma-Aldrich) for 10 min. Cells were resuspended in CM-ura medium and were allowed to adhere for 10 min at room temperature. The unadhered cells were washed and the adhered cells were resuspended in 3 mL of CM-ura and proceeded for live-cell imaging at room temperature. The mitotic spindle at the metaphase stage (1.5–2.0 µm) was used for FRAP analysis. Before photobleaching, the images were captured for three-time points at 30 s intervals. One-half of the mitotic spindle was photobleached with a 488 nm laser at 100% intensity with 100 iterations. Images were captured every 30 s at 512×512 frame size at 0.5 µm intervals for a total of 26 cycles. Images were processed by a deconvolution process automatically performed by the Zeiss software. Image analysis was done using Fiji. The images were analyzed as a maximum-intensity projection and quantified as described [76]. A 5-pixelx5-pixel square was placed over the bleached half-spindle, unbleached half-spindle, and a cytoplasmic reference site (for background subtraction) and integrated density was measured for these three sites for every time point. The fluorescence intensity was normalized with the average fluorescence intensity before photobleaching. We calculated the recovery of the bleached region (ratio of bleached to unbleached region after recovery), and the time taken to recover 50% or greater fluorescence intensity of the bleached region after photobleaching. Loss of fluorescence was observed at the unbleached half-spindle, which is in line with the observation that bleached GFP-Tub1 was replaced with the unbleached half-spindle with similar kinetics [76].

Post-acquisition analysis

Spindle length was measured by tracking it using a straight line, and freehand (for bent spindle) tool in ImageJ after maximum intensity projection for each image. Spindle orientation was measured on images that were maximum intensity projected using the angle tool in ImageJ. The angle was calculated by measuring the smaller of the two angles that the spindle forms intersecting the mother-bud axis of the cell. The aMTs were measured after maximum intensity projection of the images using either a straight line or a freehand tool in ImageJ. The budding index was represented as the ratio of the diameter of the daughter cell to the diameter of the mother cell. Throughout the manuscript, large-budded cells with a budding index of >0.6 are taken into consideration for all the analyses [118]. Statistical analyses were done using GraphPad Prism 8.4.0 software. Student t test followed by Unpaired t test for comparing two groups. The data sets represented as bar graphs were analyzed using either One-way ANOVA or two-way ANOVA for multiple groups and have been mentioned in the figure legends where applicable. Scatter plots with data from three independent biological repeats were plotted as SuperPlots [133], displaying individual color-coded data. The statistical analyses were done with three independent biological replicates.

S1 Fig. The atg11∆ cells are hyper-sensitive to the microtubule depolymerizing drug thiabendazole (TBZ).

(A) Overnight grown cells of wild type, null mutants of autophagy-related genes (atg), and two kinetochore mutants, mcm16∆ and ctf19∆ , were 10-fold serially diluted, spotted on YPD and YPD plates containing dimethylformamide (DMF) only or 50 μg mL⁻¹ thiabendazole (TBZ). Plates were photographed after incubation at 30 °C for 36 h. (B) Immunoblot analysis of whole-cell lysate prepared at indicated time points from the wild-type, atg1∆ , atg17∆ , and atg11∆ strains expressing GFP-Atg8 grown at 30 and 37 °C in nitrogen starvation medium and probed with anti-GFP antibodies. GFP-Atg8 (38 kDa) and free GFP (25 kDa) are labeled with black and blue arrowheads, respectively. Uncropped western blots are available in S1 Raw Images.

https://doi.org/10.1371/journal.pbio.3003069.s001

(TIF)

S2 Fig. Cell cycle-dependent localization of Atg11.

(A) Micrographs showing canonical localization of Atg11 onto the vacuolar membrane during mitosis. The sfGFP-Atg11 expressing cells were co-stained with FM4–64 (staining vacuolar membrane, red) and with CMAC (staining the vacuolar lumen, blue). Scale bar, 5 µm. (B) Micrographs representing different focal planes showing localization of Atg11 (sfGFP-Atg11) closer to the nuclear envelope (Nup49-mCherry, white arrowheads), at various cell cycle stages and co-stained with the vacuolar lumen dye CMAC. Scale bar, 6 µm.

https://doi.org/10.1371/journal.pbio.3003069.s002

(TIF)

S3 Fig. Atg11 interacts with Spc72 at the SPBs.

(A) Primary (square boxes) and secondary (oval) interactors of Atg11 as reported in the latest curation of BioGRID [66]. (B) Representative images displaying an in vivo BiFC interaction between Atg11 and Spc72 by the reconstitution of Venus fluorescence (VN-Spc72/VC-Atg11) in diploid cells at the vacuolar periphery, marked by FM4−64 staining. Scale bar, 6 μm. (C) Representative images displaying an in vivo BiFC interaction between Atg11 and Spc72 at the SPBs, (Spc42-mCherry), in the diploid cells. The white arrowheads mark the co-localization of the BiFC signals with the dSPB. Scale bar, 6 μm. (D) A bar diagram displaying the proportion of cells displaying co-localization of BiFC signals with either dSPB, mSPB or both the SPBs (n = 85 large-budded cells). (E) Representative fluorescence images expressing the VN-tagged Spc72 in diploid cells co-expressing Spc42-tagged with mCherry. Scale bar, 6 μm. (F) Representative fluorescence images expressing VC-tagged Atg11 in diploid cells co-expressing Spc42-tagged with mCherry. Scale bar, 6 μm. (G) Schematic showing the spatial position of proteins at the outer plaque of the SPB. (H) Yeast-two hybrid (Y2H) assays in S. cerevisiae strains carrying indicated plasmids were 10-fold serially diluted and spotted on SD - trp - leu or SD - trp - leu - his supplemented with 2 mM 3-amino-1,2,4-triazole (3-AT) and grown at 30 °C. AD- activation domain, BD- DNA binding domain. (I) Yeast-two hybrid (Y2H) assays to study interactions between Atg11 and Cnm67, an outer plaque SPB protein (highlighted in bold in the schematic). The underlying data for panel D can be found in S1 Data.

https://doi.org/10.1371/journal.pbio.3003069.s003

(TIF)

S4 Fig. Cells lacking Atg11 display a delayed cell cycle at 30 °C.

(A) Time-lapse images showing dynamics of the mitotic spindle (GFP-Tub1) and SPBs (Spc42-mCherry) during the cell cycle in wild-type (top) and atg11∆ cells (bottom) after G1 arrest followed by release at 30 °C. Scale bar, 5 μm. (B) Scatter plot displaying time taken for wild-type and atg11∆ cells, after G1 arrest followed by release at 30 °C, for completion of the cell cycle (left), to enter into metaphase (middle, 1.5–2 µm mitotic spindle), and anaphase onset till telophase (right, disassembly of MTs). Error bars show mean ±  SEM. Statistical analysis was done using an unpaired t test with Welch’s correction (**p = 0.0045/0.0031, ****p < 0.001). (C) A bar diagram representing the proportion of unbudded, small-budded, and large-budded cells in wild-type (WT) and atg11∆ cells grown in YPD for 0, 2, 4, and 6 h at 37 °C. More than 100 cells were analyzed for each biological replicate and for every time point, N = 3. Error bars show mean ±  SEM. Statistical analysis was done by two-way ANOVA using Tukey’s multiple comparisons test (**p = 0.0068/0.0017, ****p < 0.0001). The underlying data for panels B-C can be found in S1 Data.

https://doi.org/10.1371/journal.pbio.3003069.s004

(TIF)

S5 Fig. Stabilized Pds1 levels indicate defects in kinetochore-MT interactions in atg11∆ cells.

(A) Schematic (left) showing steps involved in synchronizing followed by the release of wild-type and atg11∆ cells to study Pds1 protein dynamics after G1 arrest and release at 37 °C. Western blot analysis (right) shows the expression of Pds1-6xHA in wild-type and atg11∆ cells. Protein levels of GAPDH were used as a loading control. The experiments were repeated twice with similar results. (B) Schematic (left) describing the transcriptional readthrough assay [73]. Briefly, the LacZ (blue) coding sequence is in-frame with the amino-terminal actin ORF (black). CEN DNA (gray), labeled with I, II, and III represent CDE1, CDEII, and CDEIII respectively. Transcription is under the control of the GAL10 promoter and the solid arrow, below the line diagram, represents maximum activity, while the width of dashed arrows represents β-galactosidase activity. A bar diagram (right) representing β-galactosidase levels of the corresponding strains. The measurement was performed in triplicates. Statistical analysis was done by one-way ANOVA using Dunnett’s multiple comparisons test (p < 0.0001). (C) A bar diagram representing the proportion of cells with bioriented (light blue) or mono-oriented kinetochores (CEN3-GFP) (dark gray) in each strain grown at 30 °C. n represents the minimum number of large-budded cells (budding index of > 0.6) analyzed in three independent biological replicates. Error bars show mean ±  SEM. Statistical analysis was done using two-way ANOVA for multiple comparisons (****p < 0.0001). (D) Survival time distribution of Atg11 at kinetochores was quantified from 200 ms time-interval movies. The distribution does not fit well with the double exponential curve, suggesting only a fast fraction. The pie chart represents the percentage of molecules unbound (gray) and bound with short residence time (fast fraction, light green). The average residence time of fast fractions is presented next to their representative fractions. n =  number of tracks analyzed. The underlying data for panels B-D can be found in S1 Data. Uncropped western blots are available in S1 Raw Images.

https://doi.org/10.1371/journal.pbio.3003069.s005

(TIF)

S6 Fig. Cells lacking Atg11 display a delayed Cd14 release

. (A) Time-lapse images showing Cdc14 (Cdc14-GFP) and spindle (Tub1-mCherry) dynamics during the cell cycle in wild-type (top) and atg11∆ (bottom) cells after G1 arrest and release at 37 °C. The yellow arrowheads show Cdc14 release from the nucleolus. Scale bar, 5 μm. (B) The violin plot displays the time duration for wild-type and atg11∆ cells to release Cdc14 from the nucleolus, after G1 arrest and release at 37 °C (top) or 30 °C (bottom). Statistical analysis was done using an unpaired t test with Welch’s correction (*p = 0.0108, **p = 0.0014). The underlying data for panel B can be found in S1 Data.

https://doi.org/10.1371/journal.pbio.3003069.s006

(TIF)

S7 Fig. The atg11∆ cells display an increased number of catastrophe and rescue events, a shorter aMT length at catastrophe and a shorter time span for polymerization

. (A) Bar diagram showing the number of catastrophe events in wild-type and atg11∆ cells. (B) Bar diagram displaying aMT length at catastrophe. (C) Bar diagram representing the time taken for polymerization before the catastrophe event. (D) Bar diagram showing the number of rescue events. (E) Bar diagram displaying aMT length at the rescue. (F) Bar diagram representing the time taken for depolymerization before the rescue event. Error bars show mean ±  SD. Statistical analysis was done using an unpaired t test with Welch’s correction (***p = 0.0003, ****p < 0.0001). (G) Scatter plot displaying aMT length at S-phase (<1 µm mitotic spindle) in wild-type (WT) and atg11∆ cells grown at 37 °C. (H) Scatter plot representing aMT length at metaphase (1.5–2 µm mitotic spindle) in wild-type (WT) and atg11∆ cells grown at 37 °C. (I) Scatter plot showing aMT length at late anaphase (>7 µm mitotic spindle) in wild-type (WT) and atg11∆ cells grown at 37 °C. n, a minimum number of cells analyzed, N = 3. The statistical significance was done using an unpaired t test with Welch’s correction (*p = 0.0194/0.0191). The underlying data for panels A-I can be found in S1 Data.

https://doi.org/10.1371/journal.pbio.3003069.s007

(TIF)

S8 Fig. The atg11∆ cells exhibit stabilized levels of Clb4.

(A) Schematic (left) showing steps involved in synchronizing followed by the release of wild-type and atg11∆ cells to study Clb4 protein dynamics after G1 and metaphase arrest and release at 30 °C. Cells were collected every 20 min to prepare protein samples. Western blot analysis (right) shows the expression of Clb4-9xMyc in wild-type and atg11∆ cells. Protein levels of GAPDH were used as a loading control. Clb4 normalized values are indicated below each lane and the values are plotted as a line graph. The experiments were repeated twice with similar dynamics. The underlying data for panel A can be found in S1 Data. Uncropped western blots are available in S1 Raw Images.

https://doi.org/10.1371/journal.pbio.3003069.s008

(TIF)