Structural Integrity of Centromeric Chromatin and Faithful Chromosome Segregation Requires Pat1

Abstract

The kinetochore (centromeric DNA and associated protein complex) is essential for faithful chromosome segregation and maintenance of genome stability. Here we report that an evolutionarily conserved protein Pat1 is a structural component of Saccharomyces cerevisiae kinetochore and associates with centromeres in a NDC10-dependent manner. Consistent with a role for Pat1 in kinetochore structure and function, a deletion of PAT1 results in delay in sister chromatid separation, errors in chromosome segregation, and defects in structural integrity of centromeric chromatin. Pat1 is involved in topological regulation of minichromosomes as altered patterns of DNA supercoiling were observed in pat1Δ cells. Studies with pat1 alleles uncovered an evolutionarily conserved region within the central domain of Pat1 that is required for its association with centromeres, sister chromatid separation, and faithful chromosome segregation. Taken together, our data have uncovered a novel role for Pat1 in maintaining the structural integrity of centromeric chromatin to facilitate faithful chromosome segregation and proper kinetochore function.

HIGH-FIDELITY chromosome segregation is important for normal cell growth and development. Errors in chromosome segregation result in aneuploidy, which has been associated with several human diseases and disorders, such as birth defects, developmental problems, and likely cancer (Lengauer et al. 1998). The kinetochore is essential for high-fidelity chromosome segregation in every eukaryotic organism (Burrack and Berman 2012; Maddox et al. 2012). The features of a kinetochore include centromeric (CEN) DNA, specialized protein complexes, and a unique chromatin structure (Bloom and Carbon 1982; Allshire and Karpen 2008; Burrack and Berman 2012; Choy et al. 2012; Maddox et al. 2012). The size and composition of CEN DNA varies among organisms ranging from the ∼125 bp of defined DNA sequences in Saccharomyces cerevisiae (Clarke and Carbon 1980), to 4 kb–10 Mb of DNA composed of species-specific satellite DNA, microsatellite repeats, or retrotransposon-derived sequences in other eukaryotes (Allshire and Karpen 2008; Burrack and Berman 2012; Maddox et al. 2012). Despite the lack of CEN DNA sequence conservation, several kinetochore proteins and multiprotein kinetochore subcompexes are evolutionarily conserved (Kitagawa and Hieter 2001; Meraldi et al. 2006; Burrack and Berman 2012; Maddox et al. 2012). The structural and functional analysis of kinetochores has been an active area of research, and so far about 70 kinetochore proteins have been characterized using molecular genetics and biochemical approaches in S. cerevisiae (Cho et al. 2010). Many S. cerevisiae kinetochore proteins have orthologs in fission yeast, flies, Caenorhabditis elegans, and humans (Hartzog et al. 1996; Kitagawa and Hieter 2001; Meraldi et al. 2006; Przewloka and Glover 2009; Lampert and Westermann 2011).

In addition to kinetochore proteins, structural integrity of CEN chromatin ensures high-fidelity chromosome segregation during the cell cycle (Newlon 1988; Verdaasdonk and Bloom 2011). CEN chromatin is organized into a unique, highly ordered, and topologically distinct structure in chromosomes (Bloom and Carbon 1982; Bloom et al. 1984; Saunders et al. 1990). Several factors such as chromatin-remodeling complexes (RSC complex) (Hsu et al. 2003), as well as post-translational modifications (Boeckmann et al. 2013), and altered dosage of histones regulate CEN chromatin structure for faithful chromosome segregation (Saunders et al. 1990; Pinto and Winston 2000; Au et al. 2008; Choy et al. 2011; Verdaasdonk and Bloom 2011; Yu et al. 2011; Choy et al. 2012; Au et al. 2013). Despite these studies, the molecular mechanisms that ensure the structural integrity of CEN chromatin for high-fidelity chromosome segregation are not fully understood.

In this study, we report that an evolutionarily conserved protein, Pat1, associates with CEN chromatin and regulates kinetochore structure and function to maintain faithful chromosome segregation. A deletion of PAT1 results in altered structure of CEN chromatin, altered patterns of minichromosome supercoiling, defects in sister chromatid separation, and chromosome missegregation. Pat1, initially identified as a protein associated with topoisomerase II (Wang et al. 1996), is involved in the assembly of cytoplasmic foci containing nontranslating mRNAs and decapping factors called P-bodies (Pilkington and Parker 2008), and a disruption of the PAT1 gene causes errors in chromosome transmission fidelity (Wang et al. 1996). We determined that the role of Pat1 in chromosome segregation is independent of its function in P-body assembly and translation repression. We have identified an evolutionarily conserved region (amino acid residues 254–422), which mediates the CEN association of Pat1 and is required for sister chromatid separation and faithful chromosome segregation.

Materials and Methods

Strains, plasmids, and growth conditions

Yeast strains and plasmids used in this work are listed in Table 1. Strains were grown in YPD (1% yeast extract, 2% Bacto-peptone, 2% glucose) or in synthetic medium containing 2% glucose and supplements to allow selection of plasmids.

Table 1. Saccharomyces cerevisiae strains and plasmids.

Strain	Genotype	Reference
A. Saccharomyces cerevisiae strains
YPH1015	MATa ura3-52 lys2-801 ade2-101 trp1Δ63 his3Δ200 leu2Δ1 CFIII (CEN3L.YPH278) HIS3 SUP11	Spencer et al. (1990)
JK421	MATa ura3-1 leu2-3,112 his3-11 trp1-1 can1-100 ade2-1 ndc10-1	Goh and Kilmartin (1993)
JBY577	MATa leu2,3-112::LacO-LEU2 his3-11, 15::LacI-GFP-HIS3 trp1-1 ura3-1 ade2-1 can1-100 CEN4-LacO-LEU2	Warsi et al. (2008)
JG595	MATa ura3-1 leu2,3-112 his3-1 trp1-1 ade2-1 can1-100 Δbar1 CSE4-12Myc::URA3 SCM3-3Flag::kanMX	(Camahort et al. 2007)
BY4741	MATa ura3Δ0 leu2Δ0 his3Δ1 met15Δ0	Open Biosystems
YMB6398	MATa ura3-52 lys2-801 ade2-101 trp1Δ63 his3Δ200 leu2Δ1 CSE4-13Myc::LEU2	Au et al. (2008)
YMB7963	MATa ura3-1 leu2,3-112 his3-1 trp1-1 ade2-1 can1-100 Δbar1 CSE4-12Myc::URA3 SCM3-3Flag::kanMX PAT1-3HA::TRP1	This study
YMB8307	MATa ura3-1 leu2-3,112 his3-11 trp1-1 can1-100 ade2-1 ndc10-1 PAT1-3HA::TRP1	This study
YMB8345	MATa ura3-52 lys2-801 ade2-101 trp1Δ63 his3Δ200 leu2Δ1 pat1Δ::kanMX CFIII (CEN3L.YPH278) HIS3 SUP11	This study
YMB8402	MATa ura3Δ0 leu2Δ0 his3Δ1 met15Δ0 pat1Δ::kanMX	This study
YMB8421	MATa leu2,3-112::LacO-LEU2 his3-11, 15::LacI-GFP-HIS3 trp1-1 ura3-1 ade2-1 can1-100 CEN4-LacO-LEU2 pat1Δ::kanMX	This study
YMB8422	MATa ura3-52 lys2-801 ade2-101 trp1Δ63 his3Δ200 leu2Δ1 CSE4-13Myc::LEU2 pat1Δ::URA3	This study
YMB8660	MATa ura3-52 lys2-801 ade2-101 trp1Δ63 his3Δ200 leu2Δ1 CEN-URA3 CFIII (CEN3L.YPH278) HIS3 SUP11	This study
YMB8678	MATa ura3-52 lys2-801 ade2-101 trp1Δ63 his3Δ200 leu2Δ1 edc3Δ::kanMX CFIII (CEN3L.YPH278) HIS3 SUP11	This study
YMB8706	MATa ura3-52 lys2-801 ade2-101 trp1Δ63 his3Δ200 leu2Δ1 pat1Δ::kanMX CEN-URA3 CFIII (CEN3L.YPH278) HIS3 SUP11	This study
YMB8680	MATa ura3-52 lys2-801 ade2-101 trp1Δ63 his3Δ200 leu2Δ1 dhh1Δ::kanMX CFIII (CEN3L.YPH278) HIS3 SUP11	This study
YMB8728	MATa ura3-52 lys2-801 ade2-101 trp1Δ63 his3Δ200 leu2Δ1 TOP2-18Myc-URA3 CFIII (CEN3L.YPH278) HIS3 SUP11	This study
YMB8729	MATa ura3-52 lys2-801 ade2-101 trp1Δ63 his3Δ200 leu2Δ1 pat1Δ::kanMX TOP2-18Myc-URA3 CFIII (CEN3L.YPH278) HIS3 SUP11	This study
YMB8900	MATa leu2,3-112::LacO-LEU2 his3-11, 15::LacI-GFP-HIS3 trp1-1 ura3-1 ade2-1 can1-100 CEN4-LacO-LEU2 mad2Δ::kanMX	This study
YMB8901	MATa leu2,3-112::LacO-LEU2 his3-11, 15::LacI-GFP-HIS3 trp1-1 ura3-1 ade2-1 can1-100 CEN4-LacO-LEU2 pat1Δ::URA3 mad2Δ::kanMX	This study
B. List of plasmids
Plasmid	Description	Reference
pRS416	CEN URA3	Sikorski and Hieter (1989)
pRP1490 (PAT1)	CEN URA3 PAT1-Flag	Pilkington and Parker (2008)
pRP1491 (Δ254-422)	CEN URA3 PAT1:Δ254-422-Flag	Pilkington and Parker (2008)
pRP1492 (Δ422-697)	CEN URA3 PAT1:Δ422-697-Flag	Pilkington and Parker (2008)
pTW009	CEN URA3 TOP2-18Myc	Warsi et al. (2008)

Chromosome fragment loss assay

A colony color assay developed previously (Spencer et al. 1990) was used to measure the loss of a nonessential reporter chromosome fragment (CF). Strains were grown to logarithmic phase in synthetic media to maintain the CF; cultures were then diluted and plated on synthetic medium with limiting adenine at 30°. Loss of the CF results in red sectors in an otherwise white colony. The chromosome loss frequency was calculated by determining the percentage of colonies that were at least half red, indicating loss of the CF at the first cell division. At least 3000 colonies were counted from three individual transformants for each strain (Mishra et al. 2011).

Chromatin immunoprecipitation experiments

Chromatin immunoprecipitation (ChIP) experiments were performed with three independent biological replicates as described (Mishra et al. 2011). Antibodies used to capture protein–DNA complexes were α-FLAG (A2220, Sigma), α-HA (A2095, Sigma), α-Myc (A7470, Sigma), or α-GST (Z-5, sc-459, Santa Cruz Biotech). Quantitative real time ChIP–PCR was performed using Fast SYBR Green Master Mix in 7500 Fast Real Time PCR system (Applied Biosystems) as previously described (Mishra et al. 2011). The enrichment values were calculated using the ΔΔCt method (Livak and Schmittgen 2001) and are presented as percentage input. ChIP–PCR was performed using Hotstar-Taq polymerase Master Mix (Qiagen) (Mishra et al. 2011). Primer sequences are listed in Table 2.

Table 2. List of PCR primers used in this study.

Locus	Forward (5′–3′)	Reverse (5′–3′)	Reference
CEN1	CTCGATTTGCATAAGTGTGCC	GTGCTTAAGAGTTCTGTACCAC	Choy et al. (2011)
CEN3	GATCAGCGCCAAACAATATGG	AACTTCCACCAGTAAACGTTTC	Choy et al. (2011)
CEN5	AAGAACTATGAATCTGTAAATGACTGATTCAAT	CTTGCACTAAACAAGACTTTATACTACGTTTAG	Choy et al. (2011)
HSC82	AGGGTAAATTGGCTTATCTGATTATA	TCCAAATACACAATCCAGCATATC	Mizuguchi et al. (2007)
HML	CACAGCGGTTTCAAAAAAGCTG	GGATTTTATTTAAAAATCGAGAGG	Choy et al. (2011)
CEN3–L1	ATATTGTTTGGCGCTGATCGCC	TTGATGAACTTTTCAAAGATGAC	Choy et al. (2011)
CEN3–L2	TCATCTTTGAAAAGTTCATCAAGG	GATAACAAAGCATGGTATGGCG	Choy et al. (2011)
CEN3–L3	GCCATACCATGCTTTGTTATCGTC	TATTATGCTCCCCTGGATTTTATGCG	Choy et al. (2011)
CEN3–L4	AGCATTAGAGCCACTGTCATTTC	ATAATTAAGATACGAATGTGTTCGTTG	This study
CEN3–R1	TTTACTGGTGGAAGTTTTGCTCA	GTCAACGAGTCCTCTCTGGCTA	Choy et al. (2011)
CEN3–R2	GAGAGGACTCGTTGACGTAGAA	GAATATGATAATGGTTACACCAGTAGG	Choy et al. (2011)
CEN3–R3	TGTAACCATTATCATATTCATGAC	GATTTAATGCACGTTATGTTTCG	Choy et al. (2011)
CEN3–R4	ACTGACAGCACCATTAATCAATCA	TATCCTCAGTAGAGGGCAAAGTT	This study
CEN3–DraI	TTGATGAACTTTTCAAAGATGAC	GTCAACGAGTCCTCTCTGGCTA	Choy et al. (2011)
ADP1–DraI	ATCCAAATGTGCTCAAGATAGTAGC	CACCAAACAACATTTACTAGCAGTG	This study

Western blotting

Protein samples for Western blotting were prepared using the TCA procedure (Mishra et al. 2011) and quantified with the Bio-Rad DC protein assay (Bio-Rad, Hercules, CA). Equal amounts of protein for each sample were size separated on polyacrylamide gels and transferred to nitrocellulose membrane. Primary antibodies used were α-HA (clone 12CA5, Roche), α-Flag (F-3165, Sigma), and α-Tub2 (Mishra et al. 2011). Secondary antibodies used were HRP-conjugated sheep α-mouse IgG (NA931V, Amersham), and HRP-conjugated sheep α-rabbit IgG (NA934V, Amersham).

Sister chromatid separation assay

Strains were grown in YPD or in selective medium to logarithmic phase at 30°, treated with 3 µM α-factor (T-6901, Sigma) for 2 hr to synchronize cells in G1, washed, and released into pheromone-free YPD medium. α-factor was readded after the completion of mitosis (80 min after release) to arrest the cells in the subsequent G1 phase of the cell cycle. Samples were taken at 20-min time points after release from G1 for microscopic analysis. Cells were fixed as described (Warsi et al. 2008) and LacI–GFP foci were examined using a Zeiss Axioskop 2 microscope (Zeiss, Thornwood, NY).

Minichromosome supercoiling assay

Wild-type and pat1Δ strains carrying a topologically closed circular minichromosome (pRS416) were grown in YPD at 25° to midlog phase, synchronized in G1 with α-factor, and released into pheromone free and nocodazole (20 μg/mL) containing medium at 35° for 3 hr. Total DNA was purified as described (Warsi et al. 2008). DNA was fractionated by electrophoresis on 0.6% agarose gels containing 3 μg/mL chloroquine in both the running buffer and the gel for supercoiling analysis. DNA was transferred to a nylon membrane and hybridized with a radiolabeled probe (1549-bp ScaI/PvuII fragment located adjacent to CEN6 from pRS416). Cells for fluorescence-activated cell sorting (FACS) analysis were fixed in 70% ethanol, treated with RNase A, Proteinase K, washed with 0.2 M Tris-buffer, stained with propidium iodide, and assayed using a FACSort flow cytometer and Cell Quest software (BD Biosciences, Bedford, MA) (Mishra et al. 2011). Cells were examined using a Zeiss Axioskop 2 microscope (Zeiss) to determine the cell-cycle stages following the cell morphology and nuclear position as described previously (Calvert and Lannigan 2010).

Preparation of yeast nuclei and DraI accessibility assay

Yeast nuclei were prepared from wild-type and pat1Δ strains grown in YPD at 30° to midlog phase as described previously (Saunders et al. 1990) and digested with DraI (0, 20, and 100 U/mL of nuclei) for 30 min at 37°. DNA was extracted from the chromatin as described previously (Saunders et al. 1990). For Southern hybridization, DNA was cleaved with HindIII, separated on 1.4% agarose gels, transferred to nylon membranes, and hybridized with a 900-bp radiolabeled DNA fragment flanking CEN3 (Chr3: 114945–115844). Quantitative real-time PCR (qPCR) was performed to determine the susceptibility of chromatin to DraI digestion (100 U/mL) using primers flanking DraI sites of CEN3 and ADP1. Values were normalized to an untreated control (no DraI) to determine the fraction of DNA digested by the DraI (100 U/mL) enzyme.

Results

Pat1 regulates chromosome segregation and associates with CEN DNA

A previous study reported that a disruption of the PAT1 gene by insertion of LEU2 within the coding region results in errors in chromosome segregation in mitosis and meiosis (Wang et al. 1996). We deleted the entire PAT1 gene in a wild-type strain containing a reporter CF and examined the frequency of CF loss using a colony color assay in which the loss of CF results in red sectors in an otherwise white colony (Spencer et al. 1990). Colonies that were at least half red reflecting the loss of CF in the first cell division were counted (Figure 1A, see inset). The frequency of CF loss in pat1Δ strains is about 20-fold higher than that of the wild-type strain (Figure 1A). The CF loss phenotype of pat1Δ strains is complemented by a plasmid expressing PAT1 from its native promoter (Figure 1A). The frequency of CF loss observed in pat1Δ strain is similar to that reported for deletion of kinetochore genes (Kastenmayer et al. 2005; Ma et al. 2012).

Figure 1.

Pat1 regulates chromosome segregation and associates with CEN DNA. (A) Deletion of PAT1 results in a chromosome-loss phenotype. Frequency of chromosome loss in wild ype (WT; YPH1015), pat1Δ (YMB8345), and pat1Δ strain transformed with CEN–PAT1 (pRP1490) was determined as described in Materials and Methods. Representative image (see inset) showing the loss of the nonessential chromosome fragment in the first cell division (≥half red sector) is shown. At least 3000 colonies were counted and average from three independent transformants ±SE is shown. (*) P-value <0.05, Student’s t-test. (B) Pat1 is enriched at the core CEN3. ChIP was performed for Pat1–3HA in wild-type strain (YMB7963) grown in YPD at 30°. Enrichment of Pat1 to the region surrounding CEN3 was determined by qPCR and is shown as % input. Average from three independent biological replicates ±SE is shown. The non-CEN locus HML was used as a control. (C) Pat1 associates with CEN DNA in a NDC10-dependent manner. ChIP experiments were performed for Pat1–3HA in wild-type (WT, YMB7963) and ndc10-1 (YMB8307) strains grown in YPD at permissive temperature (25°) and after a 6-hr shift to nonpermissive temperature (37°). Association of Pat1 at CEN1 is examined by PCR. IN (input), IP (DNA from ChIP with α-HA antibodies), and M (DNA from ChIP with α-GST antibodies). (D) Pat1 protein levels are not altered in ndc10-1 strain as revealed by Western blotting with α-HA antibodies. Tub2 used a loading control. (E) Enrichment of Pat1 to CEN (CEN1, CEN3, CEN5), and non-CEN (HML, HSC82) DNA in wild-type (WT, YMB7963) and ndc10-1 (YMB8307) strains was determined by qPCR and is shown as % input. Average from three independent biological replicates ±SE is shown. There was no significant enrichment of Pat1 at non-CEN regions (HML, HSC82) relative to that observed at the CENs. (*) P-value <0.05, (**) P-value <0.01, ns, significantly not different at 5% level (P-value >0.05, Student’s t-test).

The chromosome-loss phenotype observed in pat1Δ strain (Figure 1A) prompted us to examine whether Pat1 associates with CEN DNA. Previous studies have shown that Pat1 localizes to the nucleus (Teixeira and Parker 2007) and associates with the chromatin fraction in S. cerevisiae (Wyers et al. 2000). However, the genomic regions of Pat1 association are yet to be identified. We constructed a strain that expressed HA-tagged PAT1 from its native promoter and determined that the epitope tagging does not affect the function of PAT1 as this strain does not exhibit a growth defect (Supporting Information, Figure S1). ChIP experiments revealed that Pat1 associates with CEN1, CEN3, and CEN5 in wild-type cells (Figure 1, B, C, and E). ChIP–qPCR results showed that Pat1 associates with about a 1.5-kb DNA region around CEN3, exhibiting relatively higher enrichment at the central core region of CEN3 (Figure 1B). As expected, no significant enrichment was detected at the non-CEN HML locus (Figure 1B). In a control experiment using an untagged strain, no significant enrichment was observed by ChIP–qPCR at either the CEN or non-CEN HML regions (Figure S2).

Previous studies have shown that the inner kinetochore protein Ndc10 is required for the assembly of functional kinetochores in budding yeast (Goh and Kilmartin 1993). Several kinetochore proteins fail to associate with centromeres in the ndc10-1 mutant at the nonpermissive temperature (37°) (Goh and Kilmartin 1993; Janke et al. 2001; Hajra et al. 2006; Camahort et al. 2007; Winey and Bloom 2012). Hence, we examined whether CEN association of Pat1 is dependent on NDC10. ChIP experiments were done to examine the CEN association of Pat1 in wild-type and ndc10-1 strains grown at the permissive temperature (25°) and after shift to the nonpermissive temperature (37°). Western blot analysis showed that the expression of Pat1 is similar in wild-type and ndc10-1 strains grown at 25° and 37° (Figure 1D). ChIP–qPCR showed that Pat1 levels at CEN1, CEN3, and CEN5 are reduced (∼6–10 fold), specifically in ndc10-1 mutants at the nonpermissive temperature (Figure 1, C and E). There was no significant enrichment of Pat1 at non-CEN regions (HML and HSC82) relative to that observed at the CENs (Figure 1E). Based on these results we conclude that Pat1 associates with centromeres and this requires NDC10 gene.

Pat1 is required for the structural integrity of centromeric chromatin

To determine the role of Pat1 in kinetochore structure, we examined the susceptibility of CEN chromatin (CEN3) to digestion by the restriction enzyme DraI in wild-type and pat1Δ strains. Three closely spaced DraI recognition sequences are within the CDE-II region of CEN3 (see schematic, Figure 2A), which are protected from nuclease digestion due to the kinetochore protein complex (Saunders et al. 1990; Meluh et al. 1998; Crotti and Basrai 2004). Altered accessibility to DraI has been observed in kinetochore mutants and this approach has been used extensively to examine the structural integrity of CEN chromatin (Saunders et al. 1990; Meluh et al. 1998; Mythreye and Bloom 2003; Yu et al. 2011). Yeast nuclei treated with different concentrations of DraI were analyzed by Southern hybridization using a 900-bp DNA probe flanking CEN3 (see schematic Figure 2A). We observed increased DraI accessibility of CEN3 chromatin as evident from the higher signal intensity of the smaller (cut) DraI–HindIII fragments in the pat1Δ compared to the wild-type strain (Figure 2B). FACS and nuclear morphology analyses showed no major differences in the cell-cycle distribution between wild-type (WT) and pat1Δ strains (Figure S3A). We validated and quantified the increased DraI sensitivity by quantitative PCR using primers flanking CEN3 or a non-CEN region (ADP1). The CEN chromatin in pat1Δ strain was more susceptible to DraI digestion (approximately ninefold) than that observed for chromatin from a WT strain (Figure 2C). To determine if the increased nuclease sensitivity in pat1Δ strain was centromere specific, we examined the DraI sensitivity of the ADP1 locus, which also contains three closely spaced DraI sites (see schematic, Figure 2C). The ADP1–chromatin (histone H3–chromatin) in pat1Δ and wild-type strains exhibited low (∼1.0–1.5%) and similar levels of accessibility by DraI (Figure 2C). These results suggest that Pat1 is required for the structural integrity of CEN chromatin.

Figure 2.

Pat1 is required for the structural integrity of CEN chromatin. (A) Schematic of restriction enzyme sites in the three elements (CDEI, II, and III) within CEN3. Indicated are the three DraI recognition sites within CDEII of CEN3. The 900-bp probe used for Southern blotting is indicated by a shaded bar. HindIII restriction sites and resulting HindIII–HindIII and DraI–HindIII are shown. (B) CEN chromatin shows increased nuclease sensitivity in pat1Δ strains. Southern blot analysis of nuclei prepared from logarithmic growing cultures of wild-type (WT, YMB6398) and pat1Δ (YMB8422) strains in YPD at 30° were incubated with 0, 20, and 100 U/mL DraI at 37° for 30 min. DNA prepared from DraI-treated samples were digested with HindIII and Southern blots were hybridized with the probe as indicated in A. Signal intensities were quantified using Image J (Schneider et al. 2012), and the percentage of CEN3 chromatin cleaved by DraI (% cut) represents the ratio of signal intensity value of cleaved DNA (cut) over signal intensities from DraI-resistant DNA (uncut). (C) qPCR analysis revealed increased nuclease sensitivity of CEN chromatin in pat1Δ strains. DraI (100 U/mL) treated samples were assayed by qPCR using primers flanking the DraI sites in CEN3 and a non-CEN region encoding ADP1 (marked as arrows on schematic). Graph shows the susceptibility of CEN3 and ADP1 chromatin to the DraI digestion as measured by fraction of DNA cut by 100 U/mL DraI (% cut) over the untreated control (no DraI). Average from three independent biological replicates ±SE is shown. (*) P-value <0.05, and ns, significantly not different at 5% level (P-value >0.05, Student’s t-test).

pat1Δ strain exhibits defects in sister chromatid separation and alteration in superhelicity of minichromosomes

The chromosome-loss phenotype observed in pat1Δ strains may be due to defective kinetochore function and thus, we examined if pat1Δ strains show defects in sister chromatid separation by measuring the separation of sister chromatids labeled with a GFP reporter (LacI–GFP) inserted 13 kb from CEN4 (Warsi et al. 2008). Wild-type and pat1Δ strains were synchronized in G1 with α-factor, and budding index and sister chromatid separation was examined at 20-min intervals after release from the G1. In the wild-type strain a maximum number of large budded cells was observed at 100 min post G1 release, which correlates with the frequency of sister chromatid separation (∼85%; Figure 3). However, in the pat1Δ strain, the highest numbers of large budded cells and sister chromatid separation were observed at 140 min post G1 release (Figure 3). As expected, only 2–5% of G1 synchronized cells exhibited separated sister chromatids (Figure 3). To examine whether delay in sister chromatid separation of pat1Δ strain is due to activation of the spindle assembly checkpoint (SAC), we deleted MAD2, a component of the SAC, in the pat1Δ strain, and measured the budding index and frequency of sister chromatid separation as described above. The delay in sister chromatid separation observed in pat1Δ was suppressed by deletion of MAD2 (Figure 3). Similar to the wild-type strain, the highest numbers of large budded cells and sister chromatid separation were observed at 100 min post-G1 release in pat1Δmad2Δ strain (Figure 3). These results indicate that PAT1 is required for sister chromatid separation and that MAD2 contributes to the delay in sister chromatid separation in pat1Δ strains.

Figure 3.

Pat1 is required for sister chromatid separation. Wild-type (WT, JBY577), pat1Δ (YMB8421), mad2Δ (YMB8900), and pat1Δmad2Δ (YMB8901) strains marked with LacI–GFP (CEN4–proximal) synchronized in G1 with α-factor, released into pheromone-free medium, and sampled at 20-min time intervals were examined for cell morphology (top) and sister chromatid separation in cells as indicated schematically (bottom). At least 100 cells were examined for every time point. Average ±SE derived from three independent biological replicates is shown.

The interaction of Pat1 with topoisomerase II (Wang et al. 1996), together with the observation showing altered structure of CEN chromatin in a pat1Δ strain, prompted us to examine whether Pat1 affects the topological structure of minichromosomes. The electrophoretic mobility of DNA topoisomers on agarose gels is mainly based on their writhe; however, separation of topoisomers on chloroquine containing agarose gels reduces the twist of DNA. As the linking number in a circular minichromosome is fixed, the reduction in twist is compensated for by an increase in writhe (Prunell 1998; Furuyama and Henikoff 2009; Shivaraju et al. 2012). Negatively supercoiled DNA is relaxed by chloroquine due to addition of positive writhe; hence it migrates more slowly on an agarose gel, whereas positively supercoiled DNA is tightened by chloroquine and runs more quickly on a gel (Prunell 1998; Furuyama and Henikoff 2009; Shivaraju et al. 2012). We assayed the topological behavior of a minichromosome containing CEN6 from wild-type and pat1Δ strains using chloroquine gels. Deletion of PAT1 results in alteration in topological structure of minichromosomes as topoisomers from the minichromosome in pat1Δ strain migrate more slowly on chloroquine containing agarose gel compared to those observed in the wild-type strain (Figure 4A). We next examined the superhelicity of minichromosome in strains treated with nocodazole, which depolymerizes microtubules and releases spindle tension. The supercoiling pattern of minichromosomes in wild-type and pat1Δ strains treated with nocodazole was similar to that observed for untreated cells (Figure 4A). These results suggest that the altered pattern of minichromosome supercoiling in pat1Δ strain is not due to tension imposed by the mitotic spindle. The differences in topological distribution of minichromosomes observed between wild-type and pat1Δ strains is not a consequence of biased cell-cycle distribution (Figure S3B).

Figure 4.

Pat1 affects the topological state of minichromosomes. (A) Deletion of PAT1 causes alteration in topological structure of minichromosomes. Southern blot analysis of total DNA from wild-type (BY4741) and pat1Δ (YMB8402) strains with topologically closed circular minichromosome (pRS416). Total DNA was resolved on agarose gels with and without chloroquine (3 μg/mL), blotted to a nylon membrane, and hybridized with radiolabeled unique DNA fragment derived from minichromosome. Plasmid DNA from Escherichea coli was used as control. Migration pattern of topoisomers are marked with a bracket. (B) CEN association of Top2 is not affected in pat1Δ strains. ChIP of Top2–Myc at CEN (CEN1, CEN3) in wild-type (WT, YMB8728), and pat1Δ (YMB8729) strains grown at 30°. Untagged strains (YMB8660 and YMB8706) were used as a control. Enrichment was determined by qPCR and is shown as % input. Average from three independent biological replicates ±SE is shown. ns, significantly not different at 5% level (P-value >0.05, Student’s t-test).

Topoisomerase II (Top2) associates with CEN chromatin in budding yeast and its activity is required to maintain genome stability (Cline and Hanawalt 2006; Warsi et al. 2008; Baldwin et al. 2009). Based on the physical interaction between Pat1 and Top2 (Wang et al. 1996), we examined if alteration in minichromosome topology and defects in kinetochore function observed in pat1Δ strain are due to alteration in levels of centromere-associated Top2 in these strains. ChIP results showed that the level of CEN-associated Top2 was not affected in pat1Δ (15.75% of input at CEN1, and 11.84% at CEN3) compared with the levels observed in a wild-type strain (12.29% at CEN1 and 9.59% at CEN3) (P-values = >0.05) (Figure 4B). No significant enrichment was detected by ChIP–qPCR at CEN1 and CEN3 using an untagged control strain (Figure 4B).

Role of Pat1 in chromosome segregation is independent of its function in P-body assembly and translation repression

Previous studies have defined domains of Pat1 involved in P-body function in budding yeast (Pilkington and Parker 2008). To define a domain required for the chromosome segregation function of Pat1, we assayed two different alleles of pat1 (Pilkington and Parker 2008) (Figure 5A) for a CF loss phenotype. Western blotting confirmed the expression of the pat1 alleles (Figure 5B). Chromosome loss assays showed that pat1–Δ254–422, which does not show any defects in P-body assembly and translation repression (Pilkington and Parker 2008), exhibits a CF loss phenotype (CF loss rate = 1.92 ± 0.27, mean ± SE) similar to that observed for pat1Δ (1.57 ± 0.22) strains (Figure 5C), whereas pat1–Δ422–697 that exhibits defective P-body assembly and translation repression (Pilkington and Parker 2008) did not show a significant CF loss phenotype (Figure 5C). Wild-type strain expressing pat1 alleles does not result in increased CF loss (Figure S4). Based on these results, we denote the Pat1 domain from amino acids 254–422 as the chromosome transmission fidelity (CTF) region.

Figure 5.

Role of Pat1 in chromosome segregation is independent of its function in P-body assembly and translation repression. (A) Schematic of full-length PAT1 and its alleles. pat1Δ strain transformed with PAT1 or its mutant alleles as described in Table 2 were used for experiments shown in B and C. Results from chromosome loss (C) are summarized in the table. (^#) Results of P-body assembly and translation repression for these alleles were derived from a previous study (Pilkington and Parker 2008). (B) Western blot analysis with α-Flag antibodies showing expression of Pat1–Flag and its alleles (Pat1–Δ254–422–Flag, Pat1–Δ422–697–Flag). (C) The CTF region of Pat1 (amino acid residues 254–422) regulates chromosome segregation. Wild-type (WT, YPH1015) and pat1Δ strain (YMB8345) transformed with vector, PAT1 and its alleles were assayed for chromosome-loss phenotype as described in Materials and Methods. Values sharing the same letter are not significantly different at a 5% level based on analysis of variance (P-value >0.05) (D) Deletion of P-body encoding genes EDC3 and DHH1 does not exhibit an increased chromosome-loss phenotype. Wild-type (WT, YPH1015), edc3Δ (YMB8678), dhh1Δ (YMB8680), and pat1Δ (YMB8345) strains were assayed for the chromosome-loss phenotype as described in Materials and Methods. Average from three independent transformants ±SE is shown. (**) P-value <0.01, Student’s t-test.

We next examined whether deletion of other components of P-body assembly and mRNA decapping lead to a chromosome-loss phenotype. For example, Pat1 interacts with Edc3 and Dhh1 (Coller and Parker 2005; Swisher and Parker 2011), and these are activators of the P-body assembly, translation repression, and mRNA decapping (Coller and Parker 2005; Pilkington and Parker 2008; Swisher and Parker 2011). Our results showed that the frequency of CF loss in edc3Δ and dhh1Δ strains is statistically similar to that of a wild-type strain (Figure 5D). The pat1Δ strain exhibited a CF loss frequency about 20-fold higher than the wild-type strain (Figure 5D). Together, these data show that the role of PAT1 in chromosome segregation is independent of its function in P-body assembly and translation repression.

The CTF region mediates association of Pat1 with centromeres and is required for sister chromatid separation

To determine if the CTF region is necessary for the CEN association of Pat1, ChIP experiments were performed using strain expressing PAT1 or pat1 alleles (Δ254–422, Δ422–697) (Figure 5A). Full-length Pat1 and Pat1–Δ422–697 were shown to associate with CEN DNA (Figure 6A). Consistent with the increased CF loss (Figure 5C), a deletion of the CTF region in pat1–Δ254–422 results in a failure of Pat1 to associate with CEN DNA (Figure 6A). In a control experiment, no significant ChIP–PCR signal was detected at CEN using an untagged strain (Figure 6A). These results show that the CTF region of Pat1 mediates its CEN association and regulates chromosome segregation.

Figure 6.

The CTF region of Pat1 (amino acid residues 254–422) is required for the CEN association and sister chromatid separation. (A) The CTF region mediates the CEN association of Pat1. ChIP experiments showing association of Pat1–Flag or its alleles to CEN1. Lanes: IN (input), IP (DNA from ChIP using α-Flag antibodies), and M (DNA from ChIP using α-GST antibodies). Untagged strain was used as a control. (B) Deletion of the CTF region of Pat1 contributes to delay in sister chromatid separation. Wild-type (WT, JBY577) with vector (pRS416), and pat1Δ (YMB8421) with vector (pRS416), or pat1 alleles: pat1Δ–254–422 (pRP1491), pat1Δ–422–697 (pRP1492), and full-length PAT1 (pRP1490) marked with LacI–GFP (CEN4–proximal) were grown in 1× SC–URA glucose (2%) at 30°, synchronized in G1 with α-factor, and released into pheromone-free medium. Samples were taken at time points (min) after release from G1. Readded α-factor 80 min after release to block the cells in next G1. Cells were fixed and were examined for cell morphology (top) and sister chromatid separation as indicated schematically (bottom). At least 100 cells were examined for every time point. Average ±SE derived from three independent biological replicates is shown. (C) The CTF region of Pat1 contains an evolutionarily conserved sequence motif. The protein sequences of Pat1 CTF region from different fungal species: S. cerevisiae (S. cer), S. kudriavzevii (S. kud), S. arboricola (S. arb), Zygosaccharomyces rouxii (Z. rou), Ashbya gossypii (A. gos), Kluveromyces lactis (K. lac), Candida glabrata (C. gla), C. albicans (C. alb), C. dubliniensis (C. dub), and human were analyzed using the motif-based sequence analysis software MEME (http://meme.nbcr.net). Sequence alignment of the evolutionarily conserved motif is shown. Identical amino acid residues are shown with blue shading, and similar amino acid residues are marked with an asterisk. (D) The consensus sequence of the evolutionarily conserved motif is shown in the Logo format drawn with WebLogo software (Crooks et al. 2004).

We next examined whether a deletion of the CTF region of Pat1 is sufficient for the defects in sister chromatid separation observed in pat1Δ strains. Similar to the results of a pat1Δ strain (vector), a deletion of the CTF region (pat1–Δ254–422) showed defects in sister chromatid separation compared to wild-type or pat1Δ strain with pat1–Δ422–697 or full-length PAT1 (Figure 6B). As expected, only 2–7% of cells had separated sister chromatids in G1 (Figure 6B).

To determine if the CTF region of Pat1 is evolutionarily conserved, we compared Pat1 protein sequences from budding yeast, other fungal species, and humans using BLAST alignment program (http://www.ncbi.nlm.nih.gov). Sequence alignment revealed a unique 30-amino acid sequence motif within the CTF region of Pat1, which is highly conserved (40% identity, 73% similarity) between budding yeast, other fungi, and humans (Figure 6, C and D) despite the differences in the centromere structure in these species (Mishra et al. 2007; Ketel et al. 2009; Verdaasdonk and Bloom 2011; Burrack and Berman 2012; Choy et al. 2012).

Discussion

The structural dissection of the yeast kinetochore has been an area of active research since the discovery of centromeres in 1980 (Clarke and Carbon 1980), and about 70 proteins have been identified so far (Cho et al. 2010). In this study, we have shown that the evolutionarily conserved protein Pat1 is a component of the budding yeast kinetochore and is required for the structural integrity of CEN chromatin. These conclusions are derived from data which demonstrate that Pat1 (a) associates with centromeres in a NDC10-dependent manner, (b) affects the structural integrity of CEN chromatin, (c) modulates the superhelicity of minichromosomes, and (d) mediates chromosome segregation and sister chromatid separation.

The association of Pat1 with centromeres suggests that Pat1 is a structural component of CEN chromatin. Specificity of association is derived from experiments in which we observed that CEN enrichment of Pat1 is reduced significantly in the kinetochore mutant ndc10-1 at the nonpermissive temperature (37°). Similar results for a requirement of Ndc10 have been reported for CEN association of other kinetochore proteins (Janke et al. 2001; Crotti and Basrai 2004; Hajra et al. 2006; Camahort et al. 2007). Ndc10 is a prerequisite for kinetochore formation in budding yeast and is essential for the proper assembly of CEN chromatin (Goh and Kilmartin 1993; Camahort et al. 2007). The depletion of CEN-bound Pat1 in an ndc10-1 mutant further suggests that a functional kinetochore is required for CEN association of Pat1.

CEN chromatin is composed of a highly ordered and topologically unique structure in chromosomes (Bloom and Carbon 1982; Saunders et al. 1990). Our data reveal that Pat1 is critical for the structural integrity of CEN chromatin. Increased DraI accessibility of CEN chromatin to the levels observed for pat1Δ strains has been used to define a structural role for Cse4, Chl4, Spt4, and core histones (H2A, H2B, and H4) in kinetochore function (Saunders et al. 1990; Meluh et al. 1998; Pinto and Winston 2000; Mythreye and Bloom 2003; Crotti and Basrai 2004; Yu et al. 2011). Consistent with a role for Pat1 in CEN chromatin structure, the pat1Δ strains exhibit phenotypes characteristic of kinetochore mutants such as defects in sister chromatid separation and chromosome-loss phenotypes (Strunnikov et al. 1995; Meluh et al. 1998; Ortiz et al. 1999; Ma et al. 2012). The chromosome-loss phenotype in pat1Δ strain is further supported by a previous study in which a partial deletion of PAT1 coding region showed errors in chromosome transmission fidelity (Wang et al. 1996).

The altered superhelicity of minichromosomes in a pat1Δ strain suggests that Pat1 regulates the topological state of minichromosomes. Although Pat1 physically interacts with Top2 (Wang et al. 1996), the defects in minichromosome topology in a pat1Δ strain are not likely due to the altered levels of centromere-associated Top2. Unlike pat1Δ strains, catalytically inactive top2-4 mutants exhibit no defects in superhelicity of minichromosomes such that no differences in the mobility of topoisomers were observed between top2-4 and wild-type strains (Warsi et al. 2008). In addition, no alteration in CEN superhelicity was observed in topoisomerase mutants (top1, top2, and top3 mutants) in budding yeast (Saunders et al. 1990). Although it is unclear how Pat1 affects the topological state of minichromosomes, data presented in this study begin to provide a functional link between topological structure and chromosome segregation. Chromosome topology is important for kinetochore function because pat1Δ strains, which show altered supercoiling of minichromosome exhibit defects in chromosome segregation. In a recent study, addition of a kinetochore protein Scm3 into in vitro supercoiling assays containing CEN nucleosomes resulted in some changes in the superhelicity of minichromosomes (Shivaraju et al. 2012). Despite these studies, the molecular mechanisms by which a kinetochore protein affects chromosome topology are yet to be determined.

The kinetochore function of Pat1 is separable from its previously described cytoplasmic role in P-body assembly and translational repression (Pilkington and Parker 2008; Swisher and Parker 2011). This conclusion is based on three independent experimental observations. First, deletion of the CTF regions causes a chromosome loss phenotype, but exhibits no observable defects in P-body assembly and translation repression (Pilkington and Parker 2008). Second, the lack of a chromosome-loss phenotype with a deletion of the PAT1 region containing codons 422–697 that mediates P-body assembly and translation repression (Pilkington and Parker 2008). Third, strains deleted for P-body assembly and decapping components (edc3Δ and dhh1Δ) do not exhibit chromosome segregation defects.

Structure–function analysis uncovered a distinct central CTF region (amino acid residues 254–422) of Pat1 for chromosome segregation. A deletion of the CTF region abolishes association of Pat1 with centromeres, leads to chromosome-loss phenotype, and defects in sister chromatid separation. These results could mean that the CTF region directly mediates CEN binding of Pat1. Alternatively, deletion of CTF region may likely alter the protein structure of Pat1 in a way that prevents its CEN association. Our analysis showed that the CTF region of Pat1 is highly conserved and contains a sequence motif that is found in budding yeast, other fungal species, and humans. We interpret these observations to imply that the role of Pat1 in kinetochore function may be evolutionarily conserved. Although the role of the human ortholog of Pat1 (Pat1b) in kinetochore function has not been defined, nuclear localization of Pat1b has been observed in humans (Marnef et al. 2012). Furthermore, mutations and misregulation of Pat1b have been observed in several human cancers (Araujo et al. 2012; Liu et al. 2012; Dulak et al. 2013). Hence, future studies aimed at understanding the role of the human homolog of Pat1 in the structure and function of the kinetochore may help us identify potential therapeutic targets for cancer therapy.