Autophagy-related protein Atg11 is essential for microtubule-mediated chromosome segregation

Abstract

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.

Fig 1. Cells lacking Atg11 show an increased rate of chromosome loss.

(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 25, 50, 75 μg mL⁻¹ thiabendazole (TBZ). Plates were photographed after incubation at 30 °C for 48 h. (B) Overnight grown cells of indicated strains were 10-fold serially diluted and spotted on YPD plates. Plates were photographed after incubation at 30 and 37 °C for 48 h, or 14 °C for 7 days. (C) Bar graph showing fold increase in chromosome loss in atg11∆ and ctf19∆ strains at 30 and 37 °C from three biological replicates. Error bars indicate standard deviation (SD). Statistical analysis was done using Ordinary one-way ANOVA using Tukey’s multiple comparisons tests (*p = 0.0439). (D) Mitotic stability of a monocentric plasmid (pRS313) was determined in wild-type, atg mutants as indicated, as well as in the kinetochore mutant strain ctf19∆ . The assay was done with three independent sets of transformants. Statistical analysis was done using one-way ANOVA using Dunnett’s multiple comparisons tests (****p < 0.0001). The underlying data for panels C and D can be found in S1 Data.

Fig 2. Single-molecule tracking quantifies the binding dynamics of Atg11 molecules at SPBs.

(A) Time-lapse images showing localization dynamics of Atg11 (sfGFP-Atg11) and SPB (Spc42-mCherry) during the cell cycle after G1 arrest followed by release at 30 °C. Montage 1 and Montage 2 represent two different focal planes, displaying dSPB and mSPB. The white arrowheads denote Atg11’s presence proximal to the dSPB. Histogram plots represent the fluorescence intensity profile of sfGFP-Atg11 with Spc42-mCherry at 48, 56, and 88 min (marked with red boundary). Scale bar, 5 μm. (B) Representative image for sfGFP-Atg11 puncta in the cytoplasm, Spc42-mCherry, and Atg11-HaloTag-JF646 imaged over 200 ms intervals. The regions of interest (ROIs) used for tracking Atg11 molecules present in the cytoplasm (cytoplasmic Atg11) or associated with SPBs (SPB-associated) are shown in white and orange circles, respectively. (C) Survival time distribution of Atg11 in the cytoplasm and SPB-associated was quantified from 200 ms time-interval movies. The distribution fits well with the double exponential curve, suggesting two types of bound population: (1) fast fraction and (2) slow fraction. Pie charts represent the percentage of molecules unbound (gray), bound with short residence time (fast fraction, light green), and bound with long residence time (slow fraction, black). The average residence times of fast and slow fractions are presented next to their representative fractions. n =  number of tracks analyzed. (D) Representative image for sfGFP-Atg11 puncta in the cytoplasm, Spc42-mCherry, and Atg11-HaloTag-JF646 imaged over 15 ms intervals. The regions of interest (ROIs) used for tracking Atg11 over cytoplasmic Atg11 and SPBs are shown in white and orange circles, respectively. (E) Spot-On based kinetic modeling for Atg11-HaloTag (obtained from 15 ms time interval movies). The probability density function histogram of single-molecule displacements for Atg11 in cytoplasm and over the SPBs is shown. The dashed line indicates the model derived from the cumulative distribution function (CDF) fitting in Spot-On. (F) Pie charts represent the fraction of bound and unbound molecules with their mean diffusion coefficients (D), obtained from the Spot-On analysis. n =  number of tracks analyzed. (G) Survival time distribution of Atg11 at cytoplasmic Atg11 puncta and SPBs was quantified from 200 ms time-interval movies in the presence of TBZ. The distribution fits well with the double exponential curve, suggesting two types of bound population for both: (1) fast fraction and (2) slow fraction. Pie charts represent the percentage of molecules unbound (gray), bound with short residence time (fast fraction, light green), and bound with long residence time (slow fraction, black). The average residence time of fast and slow fractions is presented next to their representative fractions. n =  number of tracks analyzed. (H) Schematic showing the consequence of TBZ treatment on the dynamics of Atg11 at SPBs using SMT assays. The underlying data for panels C, E, and G can be found in S1 Data.

Fig 3. Cell cycle progression is significantly delayed in atg11∆ cells.

(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 37 °C. Scale bar, 5 μm. (B) Scatter plot displaying time taken for wild-type and atg11∆ cells, after G1 arrest followed by release at 37 °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 ±  SD. Statistical analysis was done using an unpaired t test with Welch’s correction (**p = 0.0057/0.0011, ****p < 0.001). (C) A bar graph representing the proportion of large-budded cells having metaphase (1.5–2 μm) spindle (GFP-Tub1) in wild-type and atg11∆ cells at 37 °C. n > 89, N = 3. Error bars show mean ±  SD. Statistical analysis was done using an unpaired t test with Welch’s correction (*p = 0.0376). The underlying data for panels B and C can be found in S1 Data.

Fig 4. Ipl1-dependent activation of spindle assembly checkpoint delays cell cycle progression in atg11∆ cells.

(A) A bar diagram representing the proportion of large-budded cells with stages of nuclear division in indicated strains when grown at 37 °C for 6 h. Error bars show mean ±  SEM. Cartoons represent nuclear morphology (black) in large-budded cells (gray). n, a minimum number of cells analyzed, N = 4. Statistical analysis was done by two-way ANOVA using Tukey’s multiple comparisons test (**p = 0.0056, ***p = 0.0006). (B) Overnight grown cells of wild-type, single, and double mutants were 10-fold serially diluted and spotted on YPD at 30 and 37 °C and in the presence of TBZ (50 µg mL⁻¹) at 30 °C. Plates were photographed after 48 h of incubation. (C) Schematic showing steps involved in the arrest and release of wild-type and atg11∆ cells to study Pds1 protein dynamics after metaphase arrest. Western blot analysis shows the expression of Pds1-6xHA in wild-type and atg11∆ cells released in YPD at 30 °C after G1 and metaphase arrest using α-factor and nocodazole, respectively. Cells were collected every 20 min to prepare protein samples and to quantify the budding index (top). Protein levels of GAPDH were used as a loading control. Pds1 normalized values are indicated below each lane. The experiments were repeated twice with similar results. (D) 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 37 °C. n represents the minimum number of large-budded cells (budding index of > 0.6) analyzed in three independent biological replicates. Wild-type (n > 101), atg11∆ (n > 106). Error bars show mean ±  SEM. Statistical analysis was done using two-way ANOVA for multiple comparisons (****p < 0.0001). (E) Time-lapse images of bioriented (top, white arrowheads) or mono-oriented (bottom, yellow arrowheads) kinetochores (CEN3-GFP) and SPBs (Spc110-mCherry) during the cell cycle in wild-type and atg11∆ cells after nocodazole arrest followed by release at room temperature. Scale bar, 5 μm. (F) A bar diagram representing the proportion of cells with mono-oriented kinetochores (CEN3-GFP) (dark gray) obtained from live-cell movies. (G) Overnight grown cells of wild-type, single, and double mutants were 10-fold serially diluted and spotted on YPD at 30 and 35 °C and in the presence of TBZ (50 and 75 µg mL⁻¹) at 30 °C. Plates were photographed after 33 h of incubation. (H) A bar diagram representing the proportion of large-budded cells with stages of nuclear division in indicated strains when grown at 35 °C for 6 h. Error bars show mean ±  SEM. Cartoons represent nuclear morphology (black) in large-budded cells (gray). Wild-type (n > 101), atg11∆ (n > 104), ipl1-321 (n > 69), atg11∆ ipl1-321_T1 (n > 101), atg11∆ ipl1-321_T2 (n > 102). n, a minimum number of cells analyzed, N = 3. Statistical analysis was done by two-way ANOVA using Tukey’s multiple comparisons test (**p = 0.0055, ****p < 0.0001). The underlying data for panels A, C–D, F, and H can be found in S1 Data. Uncropped western blots are available in S1 Raw Images.

Fig 5. Fluorescence recovery after photobleaching (FRAP) of the metaphase spindle.

(A, B) Time-lapse images (left) and graphs (right) representing the fluorescence recovery of photobleached metaphase spindles (blue arrowheads) in (A) wild-type and (B) atg11∆ cells. (C, D) Time-lapse images depicting the fluorescence recovery of metaphase spindle (blue arrowheads) after photobleaching and progression to anaphase in wild-type (C), and migration to the daughter bud in atg11∆ cells (D). t = −30 s shows the metaphase spindle before photobleaching, whereas the t = 0 time point represents the bleached lower-half spindle. Scale bar, 2 μm. The graphs represent the quantification of fluorescence intensities of both halves (as shown) of the metaphase spindle with time. The underlying data for panels A and B can be found in S1 Data.

Fig 6. The atg11∆ cells exhibit Kar9-dependent spindle positioning and alignment defects.

(A) Time-lapse images showing spindle (GFP-Tub1) buckling defects (blue arrowheads) at 37 °C. (B) Time-lapse images showing the mitotic spindle (GFP-Tub1) and SPB (Spc42-mCherry) dynamics during the cell cycle in atg11∆ cells after G1 arrest followed by release at 37 °C. Yellow and white arrowheads mark the movement of the mitotic spindle completely into the daughter cell and the complete elongation of the anaphase spindle in the mother cell, respectively. Scale bar, 5 μm. (C) A bar diagram showing the proportion of cells with properly segregated nuclei (light blue) or with improperly segregated nuclei represented by binucleated or multinucleated cells (dark gray). Error bars show mean ±  standard error of mean (SEM). Statistical analysis was done using two-way ANOVA using Tukey’s multiple comparisons test (****p < 0.0001). Scale bar, 5 μm. (D) The position of the short bipolar spindle in metaphase cells of wild-type (WT) and atg11∆ grown at 37 °C were analyzed. n, a minimum number of cells analyzed, N = 3. Statistical analysis was done by two-way ANOVA using Sidak’s multiple comparison test (***p = 0.0001 and ****p < 0.0001). Scale bar, 5 μm. (E) The angle of alignment along the mother–bud axis in metaphase cells of wild-type (WT) and atg11∆ grown at 37 °C and displaying a short bipolar spindle were analyzed. n, a minimum number of cells analyzed, N = 3. The statistical significance was done using an unpaired t test with Welch’s correction (*p = 0.0446). (F) A bar diagram showing the proportion of metaphase cells carrying Kar9-GFP either on one SPB (strong asymmetry), both the SPBs unequally (weak asymmetry), or both the SPBs equally (symmetry) in the indicated strains. Metaphase cells were identified by measuring the distance between the two SPBs (marked by Spc72-mCherry puncta). Statistical analysis was done using two-way ANOVA using Sidak’s multiple comparisons test (**p = 0.0030/0.0050). Scale bar, 5 μm. The underlying data for panels C–F can be found in S1 Data.

Fig 7. Cells lacking Atg11 have stabilized Clb4 levels leading to altered astral microtubule dynamics.

(A) Time-lapse images (left) and graphs (right) showing the length of astral MTs (aMTs) (GFP-Tub1) in pre-anaphase cells of wild-type cells after G1 arrest and release at 37 °C. (B) Similar images (left) and graphs (right) are shown for atg11∆ cells grown and analyzed under identical conditions as in the wild-type. Scale bar, 5 μm. Black and gray arrowheads represent the catastrophe and rescue events, respectively. (C) Scatter plot displaying catastrophe frequency, rescue frequency, polymerization rate, and depolymerization rate in wild-type and atg11∆ cells as indicated. Error bars show mean ±  standard error of mean (SEM). Statistical analysis was done using an unpaired t test with Welch’s correction (**p = 0.0022/0.0015, *p = 0.0443). (D) 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 37 °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. Experiments were repeated twice with similar dynamics. The underlying data for panels A–D can be found in S1 Data. Uncropped western blots are available in S1 Raw Images.

Fig 8. Spc72-mediated SPB inheritance is significantly altered in atg11∆ cells.

(A) A bar diagram displaying the proportion of cells having proper (dSPB in the daughter cell), reversed (dSPB in the mother cell), or symmetric (SPBs are indistinguishable) SPB inheritance in the indicated strains based on the Spc42-mCherry signal intensity at 37 °C. Right, representative images of above mentioned SPB inheritance are shown. M and D represent the mother and daughter cells, respectively. n, a minimum number of cells analyzed, N = 3. Statistical analysis was done by two-way ANOVA using Dunnett’s multiple comparisons test (****p < 0.0001). Scale bar, 5 μm. (B) Representative images displaying coincidental or opposite distribution of Kar9 (cyan) and Spc72 (magenta) in wild-type (WT) and atg11∆ cells, exhibiting wild-type-like asymmetric and predominant accumulation of both proteins in the daughter bud (D) or reversed such that asymmetric and predominant accumulation of these proteins in the mother (M) cell. Asymmetric localization of Kar9-GFP and Spc72-mCherry is marked by blue and black arrowheads, respectively. Scale bar, 5 μm. (C) Bar diagram representing the proportion of cells displaying inverted Spc72 localization when Kar9 is asymmetrically localized in daughter cells at 37 °C. n, a minimum number of cells analyzed, N = 3. Error bars show mean ±  SD. The statistical significance was done using an unpaired t test with Welch’s correction (**p = 0.0041). The underlying data for panels A and C can be found in S1 Data.

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 [98,99]. 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 [100]. 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.