Mal3 is a multi-copy suppressor of the sensitivity to microtubule-depolymerizing drugs and chromosome mis-segregation in a fission yeast pka1 mutant

The cAMP-dependent protein kinase Pka1 is known as a regulator of glycogenesis, transition into meiosis, chronological aging, and stress responses in the fission yeast, Schizosaccharomyces pombe. We demonstrated here that Pka1 is responsible for normal growth in the presence of the microtubule-destabilization drug TBZ and proper chromosome segregation. The deletion of the pka1 gene resulted in the TBZ-sensitive phenotype and chromosome mis-segregation. We isolated the mal3 gene as a multi-copy suppressor of the TBZ-sensitive phenotype in the pka1Δ strains. Overexpression of the CH domain (1–143) or the high-affinity microtubule binding mutant (1–143 Q89R) of Mal3 rescued the TBZ-sensitive phenotype in the pka1Δ and mal3Δ strains, while the EB1 domain (135–308) and the mutants defective in microtubule binding (1–143 Q89E) failed to do so in the same strains. Chromosome mis-segregation caused by TBZ in the pka1Δ or mal3Δ strains was suppressed by the overexpression of the Mal3 CH domain (1–143), Mal3 CH domain with the coiled-coil domain (1–197), or full-length Mal3. Overexpression of EB1 orthologs from Saccharomyces cerevisiae, Arabidopsis thaliana, Mus musculus, or Homo sapiens suppressed the TBZ-sensitive phenotype in the pka1Δ strains, indicating their conserved functions. These findings suggest that Pka1 and the microtubule binding of the Mal3 CH domain play a role in the maintenance of proper chromosome segregation.


Introduction
Microtubules are the major cytoskeletal fibers in eukaryotic cells, which maintain the structure of cells and provide platforms for substance transport. Microtubules grow by the polymerization of the tubulin dimer, which consists of the α-tubulin and ß-tubulin, and shrink by depolymerization. In mitosis, microtubules dramatically change their structure and regulate chromosome segregation. Microtubule associated proteins (MAPs), such as XMAP215, EB1, CLIP170, and CLASP1, regulate microtubule function and stability [1,2]
To construct pGBKT7-Tip1, the oligonucleotide primers Tip1-BF and Tip1-SSR were used to amplify a 1,386 bp fragment containing the complete tip1 protein coding sequence from S. pombe genomic DNA. The amplified tip1 gene fragment was digested with BamHI and SalI and ligated into the corresponding sites of pGBKT7 to generate the plasmid pGBKT7-Tip1. To construct pGAD424-Tip1, the plasmid pGBKT7-Tip1 was digested with BamHI and SalI and ligated into the corresponding sites of pGAD424.
The wild-type, pka1Δ, or mal3Δ cells were transformed with the pREP3X-derived plasmids (LEU2 marker) and selected onto EMMU (EMM containing uracil but lacking leucine) plates which also contained 15 μM thiamine to repress the gene expression under the nmt1 promoter. To test the growth following mal3 overexpression, transformants were grown on EMMU containing 15 μM thiamine for 2 days at 30˚C, followed by transferred onto EMMU without thiamine to induce the mal3 gene expression under the nmt1 promoter, and incubated for 1 day at 30˚C. The cells were then spotted on EMMU plates containing TBZ with or without 15 μM thiamine, and incubated for 3 to 5 days at 30˚C.
When the expression of Mal3 was induced from the nmt41 promoter, cells were first grown in EMM containing 15 μM thiamine, to the mid-log phase (~5 x 10 6 cells/mL), at 30˚C. The cells were washed three times with water and resuspended in EMM lacking thiamine. The cells were incubated for 48 h at 30˚C.

Mini-chromosomal loss assay
This assay was performed as previously described [36]. Two independent Ade + isolates of the wild type, pka1Δ, and mal3Δ strains, were analyzed for the stability of the mini-chromosome16 (Ch16) containing the ade6-M210 mutation, on adenine-limited EMMU plates, in the presence or absence of 7.5 μg/mL TBZ. The host ade6-M210 cells are rendered Ade + by allelic complementation. Transformed cells were plated on adenine-limited EMMU plates in the presence or absence of 7.5 μg/mL TBZ and incubated at 30˚C for 3 to 5 days and then at 4˚C for 1 to 2 overnight periods to allow deepening of the red color of the Adecolonies. The frequency of chromosome loss was determined by counting the total colonies and the red colonies.

Yeast two-hybrid system
The yeast Matchmaker Two-Hybrid System 3 (Clontech) was used to examine the in vivo selfinteraction of the truncated Mal3 proteins. For the assays, the plasmids of two fusion constructs into pGBKT7 (a Gal4 DNA-binding domain vector for Gal4-BD fusion) and pGAD424 (a Gal4-activating domain vector for Gal4-AD fusion) were then co-introduced into AH109 yeast cells using the lithium acetate method. Transformed yeast cells were grown on the synthetic dextrose medium (SC) lacking leucine and tryptophan (SC-LW) plate, for 3 days at 30˚C. The specific protein-protein interaction was determined by the growth of yeast cells on SC medium lacking leucine, tryptophan, and histidine (SC-LWH).

Fluorescence microscopy of GFP fusion protein
S. pombe cells were grown in EMMU liquid medium containing 15 μM thiamine, washed three times with water, transferred to the EMMU liquid medium lacking thiamine, and incubated for 24 h at 30˚C. The GFP-tagged Mal3 and the mRFP-tagged Sad1 proteins in living cells were visualized and imaged using a BX51 microscope (Olympus) with a DP70 digital camera (Olympus) or BZ-X700 microscope (Keyence). 1:1000) and rabbit polyclonal anti-PSTAIRE (Cdc2; diluted 1:1000) antibodies were purchased from Santa Cruz Biotechnology. Horseradish peroxidase-conjugated anti-mouse IgG (Santa Cruz Biotechnology) or anti-rabbit IgG antibody (Promega) was used as the secondary antibody.

Data and statistical analyses
Experiments were performed three times and the average values and standard deviations (SD) were calculated. Data from the control and the overexpression of Mal3 were compared using the two-sample t-test. P-values < 0.05 were considered statistically significant. All statistical analyses were performed using EZR version 3.5.1 (Saitama Medical Center, Jichi Medical University, Saitama, Japan) [38], which is a graphical user interface for R (The R Foundation for Statistical Computing, Vienna, Austria).

Reproducibility
All experiments were conducted at least twice to confirm the reproducibility of the results.

Multi-copy Mal3 suppresses the TBZ-and MBC-sensitive phenotype of the pka1Δ strains
To gain further the insight into the novel functions of Pka1 in S. pombe, the sensitivity of the wild type, cgs1Δ, and pka1Δ strains toward various drugs were tested. While the wild type and cgs1Δ strains grew on YES containing 0.1 μM Latrunculin A (LatA), 15 μg/mL thiabendazole (TBZ), 5 μg/mL carbendazim (MBC), or 1 M KCl, the pka1Δ strain did not on 0.1 μM LatA or 1 M KCl containing media as previously shown [20,22,39,40]. We found the pka1Δ strain also exhibited growth retardation by 15 μg/mL TBZ and 5 μg/mL MBC (Fig 1A), which are microtubule-destabilization drugs. We next analyzed the role of Cyr1 and its dependency of Pka1 activity on the growth retardation phenotype by TBZ. We found that the cyr1Δ strain exhibited the TBZ-sensitive phenotype at the concentration of 18 μg/mL TBZ, as similarly observed in the pka1Δ strain, and deletion of cgs1 reversed its phenotype ( Fig 1B). Deletion of cyr1 results in the inactivation of Pka1 and concomitant deletion of cgs1 results in the constitutive activation of Pka1. These findings indicate that the TBZ-sensitive phenotype in the cyr1Δ strain is dependent on the Pka1 activity.
Because this phenotype of the pka1Δ strains was novel, we first analyzed mitotic microtubule of the pka1Δ strain. Both the wild type and pka1Δ strains showed normal mitotic microtubules structure in the presence of TBZ similar to its absence (S1 Fig). Because mitotic microtubules of the pka1Δ strain was indistinguishable with the wild type strain, we next conducted a screening for a multi-copy suppressor to identify the possible target of Pka1. The pka1Δ strain was transformed with pREP3X-S. pombe cDNA library, which is under the control of thiamine-repressible nmt1 promoter, and screened on EMMU containing 15 μg/mL TBZ. As a result, we obtained 15 candidates including Mal3, which is a microtubule plus-end EB1 family protein. Next, we examined whether Mal3 is a specific multi-copy suppressor for the TBZ-sensitive phenotype of the pka1Δ strain. Toward this, the pka1Δ strain was transformed with pREP3X, pREP81-pka1, or pREP3X-mal3 and the transformed cells were spotted onto EMMU plates in the presence or absence of 15 μg/mL TBZ. Overexpression of Mal3 clearly rescued the TBZ-sensitive phenotype of the pka1Δ strains, although it did not rescue the growth on 1 M KCl or 0.1 μM LatA (Fig 1C). These results indicate that Mal3 is a specific multi-copy suppressor for TBZ sensitivity in the pka1Δ strain. These results indicate that the function of Mal3 was not depressed in the pka1Δ strains. Because the mal3Δ strain exhibited the TBZ-sensitive phenotype similar to the pka1Δ strains [11,41], we analyzed the genetic interaction between Pka1 and Mal3 by growing the cgs1Δ mal3Δ and pka1Δ mal3Δ double mutants on YES containing TBZ. As shown in Fig 1D, the wild type and cgs1Δ strains were not sensitive to TBZ, but the cgs1Δ mal3Δ double mutant as well as the mal3Δ strains exhibited sensitivity to 15 μg/mL TBZ. The TBZ sensitivity of the pka1Δ mal3Δ double mutant, however, was enhanced as it did not grow on a lower concentration of TBZ (10 μg/mL TBZ) compared with the pka1Δ or mal3Δ strains. The cgs1Δ strain also exhibited the TBZ-tolerance phenotype compared to the wild type strain (Fig 1D). The result, that the deletion of pka1 enhances the TBZ sensitivity of the mal3Δ strain, suggests that Pka1 has other targets in addition to Mal3.

Overexpression of the CH domain of Mal3 suppresses the TBZ-sensitive phenotype of the pka1Δ strain
Mal3 has three domains namely, the Calponin homology (CH) domain at the N-terminus (3-103), a coiled-coil domain at the middle region (167-194), and an end binding (EB1) domain at the C-terminus (197-241) [12]. To identify which domain is important for the suppression of the TBZ sensitivity of the pka1Δ strains, we constructed plasmids containing specific parts of Mal3 as shown in Fig 2A. We first analyzed the interaction of the Mal3 fragments by a yeast two-hybrid system. Mal3 , Mal3 (1-218), Mal3 (1-241), and Mal3 (135-308) interacted each other as well as with the full-length Mal3 (1-308) ( Fig 2B). However, Mal3 (1-143) did not interact with itself as in Fig 2B and also as previously described [42].
Next, we analyzed the suppression of the TBZ sensitivity of the pka1Δ strains by the individual domains. As shown in Fig 2C, the overexpression of Mal3 (1-143) and Mal3  rescued the TBZ-sensitive phenotype of the pka1Δ strains, while Mal3 (1-218) showed no effect. Mal3 (135-308) exhibited a growth defect in the pka1Δ strains at 12 μg/mL TBZ, and Mal3 (1-241) exhibited growth inhibition even in the absence of TBZ (Fig 2C). Overexpression of the Mal3 CH domain is sufficient for the suppression of the TBZ sensitivity of the pka1Δ strains and the coiled-coil domain has a negative effect on growth (Fig 2C). We then analyzed the localization of Mal3 (full-length, 1-308) and Mal3 (1-143) by GFP fusion expressed under the nmt41 promoter in the pka1Δ sad1-mRFP strain, to visualize the end points of the microtubules. As shown in Fig 2D, the full-length of Mal3-GFP clearly localized to the microtubules in interphase and mitosis. We did not observe any localization of Mal3 (1-143)-GFP in interphase, but detected localization on microtubules during prophase to anaphase in mitosis ( Fig  2D, arrowheads). These findings indicate that the microtubular localization of the CH domain dilutions were prepared as described in Fig 1A and spotted on YES in the presence or absence of 18 μg/mL TBZ. The plates were incubated for 3 or 5 days at 30˚C (YES for 3 days and YES+18μg/mL TBZ for 5 days). (C) The wild type (PR109) and pka1Δ (YMP36) strains harboring pREP3X (vector), pREP81-pka1, or pREP3X-mal3 were cultured for 1 day on EMMU (EMM with uracil) containing 15 μM thiamine at 30˚C to repress its expression of the mal3 gene by the nmt1 promoter. Cells were washed three times with water, transferred onto EMMU lacking thiamine and incubated for 24 h at 30˚C to induce mal3 by the nmt1 promoter. Culture dilutions were prepared as described above and spotted onto EMMU in the presence or absence of 15 μg/mL TBZ, 0.1 μM LatA, or 1 M KCl. All plates were incubated for 5 days at 30˚C. (D) Wild type (PR109), cgs1Δ (YMP40), pka1Δ (YMP36), mal3Δ (TTP1), cgs1Δ mal3Δ (TTP5), and pka1Δ mal3Δ (TTP7) S. pombe strains were grown on YES, harvested, and resuspended in water at 10 7 cells/mL. Culture dilutions were prepared as described in Fig 1A and

Microtubule binding ability of Mal3 is required for TBZ suppression and EB1 proteins from other eukaryotes suppress the TBZ sensitivity of the pka1Δ strains
It has been previously reported that the Mal3 (Q89R) mutant strongly binds to the microtubules, and the Mal3 (Q89E) mutant does not [12,13]. We used the Mal3 mutants (Q89R and Q89E) to identify whether microtubule binding of Mal3 is required for the suppression TBZ sensitivity of the pka1Δ strains on EMM. As shown in Fig 3A and Table 2, overexpression of the full-length Mal3 (Q89R) mutant caused growth inhibition in the pka1Δ strains in the absence of thiamine (control panel) and the full-length Mal3 (Q89E) weakly rescued the growth of the pka1Δ strains on 15 μg/mL TBZ. To further clarify the suppression by the Mal3 mutants, we next constructed the plasmids expressing the mutants of the CH domain: pREP3X-Mal3 (1-143 Q89E) and pREP3X-Mal3 (1-143 Q89R). Mal3 (1-143 Q89R) resulted in the rescue of the pka1Δ strains on EMM in the presence of 15 μg/mL TBZ or 5 μg/mL MBC, while Mal3 (1-143 Q89E) showed no effect on the pka1Δ strains on the same media ( Fig 3A and Table 2). These results indicate that the microtubule binding of Mal3 CH domain is important for the suppression of sensitivity to the microtubule-destabilizing drugs, in the pka1Δ strains.
Mal3 localizes to the microtubule plus end as a plus-end tracking protein (+TIP), thereby controlling cell morphology and microtubule dynamics in fission yeast. Therefore, we next examined whether the other +TIP proteins such as Tip1 (CLIP-170), Tea1 (cell end marker), and Tea2 (kinesin), exhibit the suppression of the TBZ-sensitive phenotype of the pka1Δ strains. As a result, the overexpression of Tip1, Tea1, or Tea2 did not rescue the growth defect of the pka1Δ strains on EMM containing 15 μg/mL TBZ, although the overexpression of Tea2 caused growth defect under normal condition (S3 Fig and Table 2). We also examined the effect of the other microtubule binding proteins Alp7 (TACC protein) and Alp14 (TOG/ XMAP215). As a result, Alp7 or Alp14 did not rescue the TBZ-sensitive phenotype of the pka1Δ strain ( S3 Fig and Table 2). These results indicate that the CH domain of Mal3 has a specific role in the suppression of sensitivity to the microtubule-destabilizing drugs of the pka1Δ strains.

Overexpression of the Mal3 CH domain suppresses the growth defect and chromosome mis-segregation in the presence of TBZ in the mal3Δ strains
It has been previously reported that mal3Δ results in the TBZ-sensitive phenotype and chromosome mis-segregation [11]. We next analyzed whether the deletion of the CH domain causes the TBZ-sensitive phenotype in the mal3Δ strains. Toward this, we used the plasmids expressing the fragmented and mutated Mal3. The overexpression of Mal3 (1-143), Mal3 , Mal3 (1-218), and Mal3 (1-308 full-length) rescued the growth of the mal3Δ strains on EMM in the presence of 15 μg/mL TBZ, whereas the overexpression of Mal3 (135-308) exhibited growth retardation on 12 μg/mL TBZ (Fig 5A and Table 4). The overexpression of Mal3 (1-241) exhibited growth inhibition even in the absence of TBZ as observed in the pka1Δ strain ( Fig 5A and Table 4). Full-length Mal3 (Q89E) and Mal3 (Q89R) exhibited growth suppression on EMM containing TBZ and growth inhibition on EMM in the mal3Δ strain, respectively. Because these results might be due to the effect of the EB1 domain, we next analyzed using the fragmented mutants of Mal3: Mal3 (1-143 Q89R) and Mal3 (1-143 Q89E).
Overexpression of Mal3 (1-143 Q89R), which is an enhanced microtubule binding mutant, rescued the growth defect of the mal3Δ strains, similar to that observed in the pka1Δ strains, on EMM containing 15 μg/mL TBZ (Fig 5A and Table 4) or 10 μg/mL MBC ( Table 4). The mutant defective in microtubule binding, Mal3 (1-143 Q89E), did not restore the sensitivity to 15 μg/mL TBZ (Fig 5A and Table 4) or 5 μg/mL MBC in the mal3Δ strains, similar to that observed in the pka1Δ strains (Table 4), indicating that the microtubule binding of Mal3 CH domain is required for the growth in the presence of microtubule-destabilizing drug. Next, we analyzed whether the CH domain is functional for proper chromosome segregation in the mal3Δ strains, by the mini-chromosome loss assay. As results, the overexpression of the Mal3 CH domain  or the Mal3 full-length (1-308) rescued the mini-chromosome loss in the mal3Δ strains, similar to that observed in the pka1Δ strains ( Fig 5B). Whereas, the overexpression of the Mal3 EB1 domain (135-308) inhibited the growth in the presence of TBZ (data not shown). These results indicate that the CH domain of Mal3 plays a role of maintenance in the normal segregation of chromosome.
Red colonies and sector colonies were counted. Experiments were performed three times: averages with S.D. are shown. Asterisks ( � ) indicate P-value < 0.05 for comparison with the pka1Δ strain harboring pREP3X (vector). (B) Wild type (PR109) and pka1Δ (YMP36) strains were cultured in YES. Cells were incubated with 10 μg/mL TBZ for 24 h at 18˚C. Cells were stained with DAPI and observed by fluorescent microscopy. Arrowhead and asterisk show the cut phenotype and mis-segregation, respectively. Scale bar: 10 μm. (C) Abnormal mitosis including the cut phenotype, in which cytokinesis takes place over the unseparated chromosomes and mis-segregation was determined by DAPI staining. About 50 cells were analyzed in individual strains and experiments were performed three times; averages with S.D. are shown. Double asterisks ( �� ) indicate P-value < 0.01 for comparison with the pka1Δ strain harboring pREP3X (vector).
https://doi.org/10.1371/journal.pone.0214803.g004  1-143 Q89R), or pREP3X-mal3 (1-143 Q89E) were cultured as described in Fig 1C. Culture dilutions were prepared as described in Fig  Next, we analyzed the suppression upon overexpression of other eukaryotic EB1s in the mal3Δ strains. Overexpression of S. cerevisiae Bim1p (ScEB1), human EB1 (HsEB1), mouse EB1 (MmEB1), and A. thaliana EB1a (AtEB1a) rescued the growth defect in the mal3Δ strains in the presence of 12 μg/mL TBZ but A. thaliana EB1b (AtEB1b) and EB1c (AtEB1c) failed to do so, as observed in the pka1Δ strains (Fig 5C). These results indicate that the EB1 from yeasts to higher eukaryotes have conserved role in the restoration of growth defect in the presence of microtubule-destabilizing drugs.

Plasmid
No drug +15 μg/mL TBZ +5 μg/mL MBC +10 μg/mL MBC (1-143 Q89R) The TBZ sensitivity and the MBC sensitivity were analyzed by spotting assay and streak assay, respectively. https://doi.org/10.1371/journal.pone.0214803.t004 Mal3 in S. pombe pka1Δ transformants were cultured on EMMU in the absence of thiamine, transferred on EMMU in the presence or absence of TBZ, and incubated at 30˚C. As a result, overexpression of Mal3 (1-308) or Mal3 (1-143) did not alter their growth even with the presence of a high concentration of TBZ (60 μg/mL), although Mal3 (135-308)-overexpression caused inhibition of cell growth (Fig 6A). This result suggests that the pka1Δ and mal3Δ strains were rescued by the TBZ-tolerance of the Mal3 CH domain ). Finally, we tested the possibility that the Mal3 CH domain would affect the CLIP170 ortholog, Tip1. To do this, we analyze the interaction between the Mal3 CH domain and Tip1 by the yeast two-hybrid system. As a result, Mal3 (1-143) did not interact with Tip1 although Mal3 (1-308: full-length) interacted with Tip1 ( Fig 6B). This result suggests that the suppression of the Mal3 CH domain was not mediated by Tip1.

Discussion
It has been reported that Pka1 is involved in glucose repression, chronological aging, regulation of transit from mitosis to meiosis, and stress response to KCl and CaCl 2 in S. pombe [14,[20][21][22][23][24][25]. Pka1 also regulates the transition from the G2 phase to M phase, as the overexpression of Pka1 also causes growth defect and mitotic defects leading to cell elongation and DNA content accumulation [43]. On the other hand, PKA regulates the transition of G1/S phase in S. cerevisiae and the progression of G2/M phase in Xenopus laevis [44,45]. These reports suggest that PKA regulates cell cycle progression at several phases. However, its detailed mechanism is mostly unclear. In the M phase, S. pombe Pka1 is involved in the regulation of the anaphase-promoting complex (APC), but its detailed mechanism has not been elucidated [46][47][48][49][50]. In this study, we found that pka1Δ exhibits the TBZ-sensitive phenotype and abnormal chromosome segregation in the presence of TBZ. In general, microtubule-destabilizing drugs inhibit the progression of the M phase in the cell cycle. Spindle assembly checkpoint (SAC) proteins such as Mad1, Mad2, Mad3, Bub1, Bub3, and Mph1 are not essential for the growth under normal condition. However, inactivated SAC causes the TBZ-sensitive phenotype and chromosome missegregation in the presence of microtubule destabilization drug [51]. SAC proteins negatively regulate APC for progression from metaphase to anaphase in the M phase. These results and our observation, that the pka1Δ strains exhibit the TBZ-sensitive phenotype and chromosome mis-segregation, together suggest that Pka1 regulates SAC and/or APC in the M phase.
To know how Pka1 regulates chromosome segregation, we performed multi-copy suppressor screening and isolated Mal3 (MAPs and +TIP protein) based on its ability to suppress TBZ sensitivity in the pka1Δ strains. The CH domain (1-143) of Mal3, but not other domains, is responsible for this suppression. Our results suggest that the deletion of the pka1 gene prevents the attachment of chromosome to the microtubule and the CH domain of Mal3 enhances the binding stability between the microtubule and chromosome. In fact, a microtubule unattached form of Mal3 (1-143 Q89E) failed to restore the sensitive of TBZ in the pka1Δ strains (Fig 3A). The loss of functional Dis1, a TOG/XMAP215 microtubule plus end tracking polymerase, resulted in chromosome mis-segregation [52]. Leucine 841 and proline 844 of Dis1 at C-terminus are responsible for binding to Mal3, while the C-terminus region from 174 to 247 amino acids residue containing a part of coiled-coil domain and EB1 domain of Mal3 are required for binding with Dis1 [41]. Based on their report, we suggest that the function of Dis1 is not required for the suppression of the TBZ-sensitive phenotype in the pka1Δ strain. The CH domain of Mal3 (1-143) did not localize to the microtubules during interphase and only localized to the microtubules during prophase to anaphase. The +TIP proteins, Tea1, Tea2, and Tip1, mostly play a role of in the cell polarity of microtubules in interphase [53]. Tea1, Tea2, and Tip1 bind through Mal3 to the microtubules [53]. This is because the formation of +TIP complex mediated by EB1 domain of Mal3 is required to attach to microtubules in interphase, and the CH domain alone is not able to play this role as observed in the case of Tea1, Tea2, and Tip1 [53,54]. Therefore, the function of mitotic microtubules is important for suppression of the TBZ-sensitive phenotype in the pka1Δ strain. Overexpression of Alp14, which is a microtubule binding protein containing the TOG domain, functions similar to the CH domain of Mal3, but failed to restore the TBZ sensitive phenotype in the pka1Δ, indicating that the CH domain of Mal3 is required for microtubule stability in the pka1Δ strains. The CH domain of Mal3 does not show discernible differences from the TOG domain regarding the Pka1-mediated microtubule stabilization. Beinhauer et al. have shown that the mal3Δ strain exhibited the TBZ-sensitive phenotype and chromosome mis-segregation [11]. These results were also supported in this study (Figs 1D and 5). The CH domain of Mal3 (1-143) is sufficient to suppress these phenotypes, while an unbinding form of Mal3 (1-143 Q89E) failed to reverse the TBZ sensitive phenotype of the mal3Δ strains (Fig 5 and Table 4). These results indicate that the microtubule binding ability of Mal3 is important for proper function. It has been reported that the CH domain of Mal3 binds to microtubules in vitro [13], but function of the CH domain has not been clearly understood in vivo. Our result in this study showed that the CH domain of Mal3 co-localizes with the microtubules and retains a role required for the proper segregation of chromosome. It has also been shown that the Mal3 (Q89R) mutant strongly binds to the microtubules and causes hyper phosphorylation at serine and threonine between 144 and 155 amino acids [12]. Overexpression of Mal3 (full-length Q89R) exhibited sever growth defect even under normal growth condition, probably due to excessive binding to microtubule; whereas overexpression of Mal3 (full-length Q89E) showed suppressive effect on the TBZ-sensitive phenotype of the pka1Δ and mal3Δ strains; even the mutant has only weak microtubule binding ability, probably due to the moderate effect of this mutation. Intriguingly, the effect of Q89 mutation was different when a truncated version of Mal3 was expressed. While Mal3 (1-143 Q89R) suppressed the TBZ-sensitive phenotype of the pka1Δ and mal3Δ strains, Mal3 (1-143 Q89E) failed to do so. This suggests that Mal3 (1-143 Q89E) does not have sufficient microtubule binding ability comparing with Mal3 (full-length Q89E) when overexpressed. We also observed that the overexpression of Mal3 (1-241) causes sever growth inhibition compared to full-length Mal3 and other Mal3 fragmented proteins (Figs 2C and 5A, and our unpublished data). These results suggest that the growth inhibition is caused by overexpression of fragments, including both the CH domain and the EB1 domain. The analysis to understand the cause of the severe growth inhibition observed with the truncated C-terminus mutant is currently underway. Genetic analysis on the pka1Δ and mal3Δ mutants suggest that these two genes work in parallel pathway, because the pka1Δ mal3Δ double mutant exhibited much severe sensitivity to TBZ than each single mutant (Fig 1D). Overexpression of Pka1 did not rescue the TBZ-sensitive phenotype of mal3Δ (data not shown), supporting the idea that Mal3 operates downstream of the cAMP/PKA pathway during mitosis. In fact, the gain of functional PKA mutant, the cgs1Δ strain, rescued the TBZ sensitivity in the cyr1Δ strain ( Fig 1B) and exhibited the TBZ-tolerance phenotype compared to the wild type strain (Fig 1D). This result suggests that the PKA activity is important for the response to the microtubule-destabilizing drugs.
Finally, we showed that the functional suppression is also observed in the EB1 orthologs S. cerevisiae Bim1p (ScEB1), human EB1 (HsEB1), mouse EB1 (MmEB1), and A. thaliana EB1a (AtEB1a), in the pka1Δ and mal3Δ strains. EB1 protein has the CH domain at the N-terminus and EB1 domain at the C-terminus. In all these organisms, the EB1 protein binds to the microtubules and plays the role of microtubule stabilization [55]. We identified that the EB1 proteins is responsible for proper chromosome segregation by genetic analysis. Our complementation analysis indicates that the function of Mal3 is mostly conserved from yeasts to high eukaryotes, as the overexpression of AtEB1b and AtEB1c did not suppress the TBZ-sensitivity of the pka1Δ and mal3Δ strains (Fig 3B). Because AtEB1c has a tail region with patches of basic amino acid residues, its function and structure are different from Mal3, budding yeast Bim1p (ScEB1), mouse EB1 (MmEB1), human EB1 (HsEB1), and AtEB1a [35]. Our suppression results suggest that the function is conserved in fission yeast Mal3, budding yeast Bim1p (ScEB1), mouse EB1 (MmEB1), human EB1 (HsEB1), and A. thaliana AtEB1a. However, we have no explanation on why the overexpression of AtEB1b failed to suppress the TBZ-sensitive phenotype in the pka1Δ and mal3Δ strains.
In conclusion, in addition to the known roles of Pka1 in meiosis, gluconeogenesis, chorological aging, and stress response of S. pombe, we propose that Pka1 is also involved in the regulation of chromosome segregation during mitosis. Our findings provide new insights into the novel function of Pka1 in regard to the CH domain of Mal3, in the regulation of microtubule organization and chromosome segregation in S. pombe.