1

TETRAHYMENA POT2 IS A DEVELOPMENTALLY REGULATED PARALOG OF POT1 THAT LOCALIZES TO 1

CHROMOSOME BREAKAGE SITES BUT NOT TO TELOMERES 2

3

Stacey Cranerta1, Serena Heyseb1*, Benjamin R. Lingera**, Rachel Lescasseb***, Carolyn Pricea# 4

5

aDepartment of Cancer Biology, University of Cincinnati, OH 45267, USA. bDepartment of 6

Molecular Genetics, Microbiology, and Biochemistry, University of Cincinnati, OH 45267, USA. 7

1These authors contributed equally to this work 8

9

10

11

Running Head: Pot2 localizes to chromosome breakage sites 12

13

14

15

#Address correspondence to Carolyn Price. E. Mail; Carolyn.price@uc.edu 16

17

*Current address: The Procter and Gamble Company, Cincinnati, OH 45202, USA 18

**Current address: Department of Chemistry, Indiana Wesleyan University, Marion, IN 46953, 19

USA 20

***Current address: Education Nationale, Académie de Créteil, France 21

22

Accession numbers for new sequences: POT1: KM406495, POT2: KM406494, PAT1: 23

KM406496, TPT1: KM406497 24

25

26

EC Accepts, published online ahead of print on 10 October 2014Eukaryotic Cell doi:10.1128/EC.00204-14Copyright © 2014, American Society for Microbiology. All Rights Reserved.

on May 29, 2021 by guest

http://ec.asm.org/

Dow

nloaded from

2

ABSTRACT 27

Tetrahymena telomeres are protected by a protein complex composed of Pot1, Tpt1, 28

Pat1 and Pat2. Pot1 binds the 3’ overhang and serves multiple roles in telomere maintenance. 29

Here we describe Pot2, a paralog of Pot1, which has evolved a novel function during 30

Tetrahymena sexual reproduction. Pot2 is unnecessary for telomere maintenance during 31

vegetative growth as telomere structure is unaffected by POT2 macronuclear gene disruption. 32

Pot2 is expressed only in mated cells where it accumulates in developing macronuclei around 33

the time of two chromosome processing events: Internal Eliminated Sequence (IES) excision 34

and chromosome breakage. Chromatin immunoprecipitation (ChIP) demonstrated Pot2 35

localization to regions of chromosome breakage but not to telomeres or IESs. Pot2 association 36

with Chromosome Breakage Sites (CBSs) occurs slightly before chromosome breakage. Pot2 37

did not bind CBS or telomeric DNA in vitro suggesting that it is recruited to CBSs by another 38

factor. The telomere proteins Pot1, Pat1 and Tpt1 and the IES binding factor Pdd1 fail to co-39

localize with Pot2. Thus, Pot2 is the first protein found to associate specifically with CBSs. The 40

selective association of Pot2 versus Pdd1 with CBSs or IESs indicates a mechanistic difference 41

between the chromosome processing events at these two sites. Moreover, ChIP revealed that 42

histone marks characteristic of IES processing, H3K9me3 and H3K27me3, are absent from 43

CBSs. Thus, the mechanisms of chromosome breakage and IES excision must be 44

fundamentally different. Our results lead to a model where Pot2 directs chromosome breakage 45

by recruiting telomerase and/or the endonuclease responsible for DNA cleavage to CBSs. 46

47

INTRODUCTION 48

Telomeres are dynamic complexes of protein and nucleic acid that protect the ends of 49

linear eukaryotic chromosomes (1, 2). The telomere proteins prevent the chromosome terminus 50

from being recognized as DNA damage. They also regulate access of telomerase, the enzyme 51

that synthesizes telomeric DNA to maintain telomere length (2, 3). In the ciliate Tetrahymena 52

on May 29, 2021 by guest

http://ec.asm.org/

Dow

nloaded from

3

thermophila, telomeres are bound by a four protein complex composed of Pot1 (formally Pot1a), 53

Tpt1, Pat1, and Pat2 (4, 5). The complex binds to the 3’ single-strand overhang that is present 54

on the G-rich strand of the telomeric DNA (the G-overhang). Pot1 binds the overhang directly, 55

Tpt1 binds to Pot1 and together the two proteins stop the overhang from eliciting a DNA 56

damage response and prevent excessive telomere elongation (5, 6). Pat1 and Pat2 are not 57

needed for telomere protection but they are required for telomerase to maintain telomere length 58

(4, 5). The POT1 gene was originally identified as one of two Tetrahymena homologs (POT1 59

and POT2) of the Oxytricha Telomere End Binding Protein (TEBP) (6). The POT1 and POT2 60

genes lie ~1.3 kb apart in the macronuclear genome suggesting they arose through gene 61

duplication (Fig. 1A). The encoded proteins have 42% sequence identity with each other and 62

∼25% identity to TEBP. Although the role of Pot1 in telomere maintenance and end protection is 63

well established (6), the function of Pot2 has remained unclear. We now address the role of 64

Pot2 and show that it functions during Tetrahymena genome reorganization but that it is not 65

needed for telomere maintenance during normal vegetative cell growth. 66

Tetrahymena have an unusual nuclear organization that is generated through genome 67

reorganization during the sexual stage of the life cycle (7). The cells are bi-nucleated with a 68

somatic macronucleus that is transcriptionally active during vegetative growth and a germline 69

micronucleus that is silent with its chromatin in a heterochromatic state (8). During sexual 70

reproduction, the old parental macronucleus is destroyed and a zygotic copy of the 71

micronucleus gives rise to new micronuclei and macronuclei (7, 9). Formation of the new 72

macronucleus involves a developmental program during which the micronucleus-derived 73

chromosomes are reorganized and matured into a transcriptionally active state. There are two 74

major genomic reorganization events: chromosome breakage with new telomere addition and 75

Internal Eliminated Sequence (IES) excision. Chromosome breakage is the process whereby 76

the five large micronuclear chromosomes are broken into ∼200 smaller units and telomeric 77

on May 29, 2021 by guest

http://ec.asm.org/

Dow

nloaded from

4

repeats are added de novo to the broken chromosome ends (10-12). IES excision involves 78

removal of nearly one third of the micronuclear genome by excising segments of DNA and 79

ligating the broken ends back together (13, 14). Subsequent DNA amplification generates a 80

copy number of ∼45 for all chromosomes except the rDNA which is amplified to ∼9000 copies 81

(7, 15). 82

Chromosome breakage occurs at a well conserved 15 base-pair consensus sequence 83

that contains a 10-bp invariant core (16, 17). About 200 of these loci, termed chromosome 84

breakage sites (CBSs), exist in the micronuclear genome (11, 18). After breakage, telomeres 85

are added to the newly formed ends by the enzyme telomerase (10, 12). It is not yet known 86

whether telomere addition occurs concurrently with, or shortly after, breakage and the proteins 87

involved in DNA cleavage and telomerase recruitment remain to be identified. While the rDNA is 88

also processed by this mechanism, after cleavage two rDNA molecules are ligated to form a 89

palindrome with telomeres on each end (15, 19). 90

Unlike chromosome breakage, there is no consensus sequence for IES excision. 91

Instead, the parental macronucleus is used as a template for directing removal of sequences in 92

the developing macronucleus by an RNAi-like mechanism (8, 14). Early in conjugation, the 93

micronucleus is transcribed into long non-coding RNA which is then processed into small RNAs, 94

termed scan RNAs (scnRNAs), by a Dicer-like protein Dcl1 (20-22). The scnRNAs then “scan” 95

the genome of the parental macronucleus and if sequence homologous to a scnRNA is found, 96

the RNA is degraded (23). The RNAs that were not degraded contain sequences that are 97

absent from the parental macronucleus and thus represent sequences to be targeted for 98

excision from the developing macronucleus. The remaining scnRNAs are transported to the 99

developing macronucleus where they target homologous sequences for excision by mediating 100

histone methylation to generate tri-methylated H3K9 and H3K27 (H3K9Me3 and H3K27Me3) (8, 101

24-26). The H3K9Me3 and H3K27Me3 marks are recognized by the HP1-like protein Pdd1 and 102

by Tpb2 a PiggyBac transposase-like protein (27, 28). Pdd1 is needed for assembly of the 103

on May 29, 2021 by guest

http://ec.asm.org/

Dow

nloaded from

5

marked IESs into heterochromatin bodies while Tpb2 carries out the actual DNA cleavage (29, 104

30). Pdd1 also has poorly understood roles early in conjugation when it localizes to both the 105

parental macro- and micronuclei (28, 29). 106

Interestingly, gene disruptions in essentially all components of the IES excision pathway 107

examined to date prevent not only IES excision but also chromosome breakage (21-23, 26-28, 108

31, 32). PDD1 is a notable exception in that somatic knockout of early expression leaves 109

chromosome breakage unaffected although cells are unable to complete IES excision (28). At 110

first sight, the finding that most IES components are required for chromosome breakage 111

suggests that the two DNA processing pathways utilize essentially the same RNAi based 112

mechanism. However, the requirement for early expression of Pdd1 in order to complete IES 113

excision but not chromosome breakage raises the possibility that chromosome breakage 114

proceeds by a mechanistically different process that can be indirectly affected by the status of 115

the IES excision pathway. For example, disruption of IES excision might trigger a developmental 116

checkpoint which prevents the cell from proceeding with the chromosome breakage program. 117

Here, we show that the telomere protein paralog Pot2 is developmentally regulated with 118

expression coinciding with chromosome breakage and telomere addition. Moreover, Pot2 119

localizes to sites of chromosome breakage but not to telomeres or IESs. Thus, Pot2 is the first 120

protein to be specifically associated with chromosome breakage. We also show that CBSs lack 121

H3K9 and H3K27 methylation. This result indicates that IES excision and chromosome 122

breakage must proceed via different mechanisms. 123

124

MATERIALS AND METHODS 125

Growth, mating, and transformation of Tetrahymena. Cells were grown in 1X SPP or 1.5X 126

PPYS media with 1X antibiotic/antimycotic as described previously (33). To obtain growth 127

curves, cells were maintained in log phase growth (1-2x105 cells/mL). Cell lines with disruption 128

of the macronuclear POT2 gene were generated by using biolistic transformation to introduce a 129

on May 29, 2021 by guest

http://ec.asm.org/

Dow

nloaded from

6

gene replacement construct containing the Neo3 drug selection cassette into the POT2 gene 130

locus of Cu428 and B2086 cells (Sup Fig. 1A). The cells were selected with increasing 131

concentrations of paromomycin until all copies of the macronuclear POT2 gene were replaced. 132

Cells were checked at intervals to verify that they retained the full gene replacement. For 133

mating, wild type Cu428 and B2086 or POT2 knockout cells were starved in 10 mM Tris buffer 134

pH 7.5 for 18-24 hours, 1x105 cells/mL of each mating type were then mixed and aliquots were 135

taken at indicated times. The percent of cells in each stage was assessed by fluorescent 136

microscopy after staining with 0.01% acridine orange. The POT1 and PDD1 macronuclear 137

knockout cell lines were as previously described (6, 28). The PDD1 micronuclear knockout cell 138

line was provided by Douglas Chalker (Washington University, Missouri). 139

140

Telomere analysis. Genomic DNA from wild type or POT2 macronuclear knockout cells was 141

digested with HindIII and analyzed by Southern blot using a sub-telomeric probe specific to the 142

rDNA (Table 1) as described previously (34). Analysis of the G-overhang length was performed 143

as described (6). 144

RT-PCR. RNA was purified using the Qiagen RNAeasy kit then treated with DNase. Reverse 145

transcription was performed using random hexamers. The cDNA was then diluted and used as 146

a template for PCR. 147

148

Generation of Pot2 antibody. Pot2 antibody was made by immunizing rabbits with denatured, 149

full-length Pot2 expressed in E. coli. Antibody was purified on a column made by coupling 150

purified Pot2 to NHS-activated Sepharose 4 Fast Flow (GE). Antibody was eluted with Pierce 151

gentle elution buffer before dialysis into TBS and addition of 0.02% NaAzide and 10% glycerol. 152

Pot1, Pat1 and Tpt1 antibodies were as previously described (5, 6). Pdd1 antibody was from 153

on May 29, 2021 by guest

http://ec.asm.org/

Dow

nloaded from

7

Abcam (ab5339) and H3K27Me3 and H3K9Me3 antibodies were from Millipore (07-449 and 07-154

442). 155

156

Immunolocalization. Fixation was performed as previously described (35). 3.0 mL of mating 157

cells were fixed with 10 μl of fixative (2:1, saturated mercuric chloride: 95% EtOH) for 5 min. 158

Cells were pelleted, resuspended in 100% MeOH, dropped on slides and air dried. Slides were 159

blocked with PBT buffer (3% BSA, 0.1% Tween 20, 60 mM PIPES, 25 mM HEPES, 10 mM 160

EGTA, and 2 mM MgCl2, pH 6.9) for 1 hr and then incubated with a 1:100 dilution of purified 161

Pot2 antibody in PBT (0.35 mg/ml) followed by 1:500 dilution of Cy2-conjugated goat anti-rabbit 162

secondary antibody. Counter staining was with 0.1 μg/ml DAPI for 10 min. 163

164

11 kb PCR. The 11 kb PCR assay was performed using one primer complementary to the 165

telomeric repeats (F3) and one internal primer that hybridized within the 17S RNA coding 166

sequence (R3c) (Table 1). The PCR was performed using 5Prime PCR extender® kit, the 167

conditions were as follows: 10 pmol R3c, 10 pmol F3, 160 ng of genomic DNA, 2 μl 10X PCR 168

extender Buffer, 0.4 µl 10 mM dNTP's, and 0.3 μl enzyme mix. 94°C 2 min, 40 cycles of 94°C 169

20 sec, 50°C 20 sec, 72°C 2 min, then 72°C 5 min. 170

171

Chromatin Immunoprecipitation (ChIP). Cells were washed with 10 mM Tris pH 7.5, fixed 172

with formaldehyde, the DNA sheared by sonication and the soluble fraction prepared as 173

previously described (6). 50 μL of soluble chromatin, at 5x106 cells/mL, was used per 174

immunoprecipitation in 1 ml of lysis buffer (150 mM NaCl, 25 mM Tris, pH 7.5, 5 mM EDTA, pH 175

8.0, 1% Triton X-100, 0.1% SDS, 0.5% NaDoc, and 1X protease inhibitor cocktail). Samples 176

were incubated with antibody overnight with rotation at 4°C, Protein A beads were added for ≥2 177

hours and the beads were processed as described previously (4). Briefly, a sample of input 178

chromatin was collected for DNA isolation. DNA was purified by boiling the beads with 50 μL 179

on May 29, 2021 by guest

http://ec.asm.org/

Dow

nloaded from

8

10% Chelex slurry in water (Chelex® 100 from BioRad) (36). The samples were treated with 180

Proteinase K (100 μg/mL) for 30 min at 55°C then boiled again for 10 minutes. The supernatant 181

was collected, the beads were re-extracted with 50 μL ddH20, and supernatants pooled. The 182

supernatant was used directly as a template for real-time quantitative PCR. The PCR was 183

performed using SYBR® Advantage® qPCR Premix from Clontech Laboratories, Inc. The 184

conditions were as follows; 95°C for 2 min, followed by 40 cycles of 95°C for 15 sec, 55°C for 20 185

sec, 72°C for 30 sec . The primers are shown in Table 1. 186

187

Protein isolation and DNA binding analysis. Full length TAP-tagged Pot1 was purified from 188

insect cells using IgG beads and the tag was removed by TEV cleavage. MBP-tagged Pot2 was 189

purified from insect cells on amylose resin. MBP-tagged Pot1 and Pot2 N-terminal domains (a.a. 190

1-287 for Pot1 and a.a. 1-270 for Pot2) were expressed in E. coli and purified on amylose resin. 191

The truncation site was based on modeling of structural homology to the Oxytricha TEBP DNA-192

binding domain using Phyre2 and included both predicted OB folds. For DNA binding studies, 193

oligonucleotides (Table 1) were 5’ end labeled with 32P γ-ATP and duplexes were formed by 194

heating and slow cooling. The 15 μl binding reactions contained 25-200 ng of purified protein 195

and 4 fmol of labeled oligonucleotide in 40 mM Tris pH 7.5, 125 mM NaCl, 5% glycerol, 1 mM 196

DTT, and 0.25% NP-40. After 30 min incubation at room temperature, binding reactions were 197

separated in 5% non-denaturing acrylamide gels made with 1X TBE and 2.5% glycerol. Gels 198

were run with 0.5X TBE at 130V and analyzed by Phosphor Imaging. 199

200

RESULTS 201

POT2 is unnecessary for macronuclear telomere maintenance 202

To gain more insight into the function of Pot2, we disrupted the macronuclear POT2 203

gene by replacing the endogenous gene locus with a neomycin resistance cassette (Fig. S1A). 204

Southern blot analysis confirmed full gene replacement, indicating that POT2 is not essential 205

on May 29, 2021 by guest

http://ec.asm.org/

Dow

nloaded from

9

(Fig. S1B). The knockout cells exhibited normal cytology (data not shown) and the growth rate 206

was unaffected (Fig. 1B). Thus, unlike Pot1 which is essential, Pot2 is not needed for growth of 207

vegetative cells. 208

Since Pot1 is required for several aspects of macronuclear telomere maintenance, we 209

asked if Pot2 also has telomeric functions. To determine the effect of Pot2 loss on telomere 210

length, we used Southern-blot analysis to compare the size of the rDNA telomeric restriction 211

fragments from wild type and POT2 knockout cells. Unlike Pot1 depletion which causes rapid 212

telomere elongation (6), loss of Pot2 had no effect on telomere length (Fig. 1C). As Pot1 213

depletion also causes an increase in G-overhang length and a change in the 3’ terminal 214

nucleotide, we next asked whether Pot2 loss affects G-overhang structure. This was achieved 215

using a previously described oligonucleotide ligation and primer extension procedure (Fig. 1D) 216

(33). An adaptor complex composed of a unique sequence DNA duplex with a 5 nt extension of 217

C-strand telomeric repeat was hybridized and then ligated to the G-overhangs on telomeric DNA 218

isolated from wild type or POT2 knockout cells. The oligonucleotide with the C-strand extension 219

(the guide oligo) was then primer extended with T4 DNA polymerase to the junction with the 220

telomere duplex DNA and the reaction products were visualized by gel electrophoresis. In initial 221

experiments, we monitored ligation of a series of adaptor complexes harboring guide oligos with 222

different permutations of the C-strand telomeric repeat. Only guide oligos with a 3’CCCA 223

extension allowed ligation of the adaptor complex to the telomere (data not shown), indicating 224

that the overhang terminated with the sequence 5’-G4T and hence was unchanged by loss of 225

Pot2 (33). When we performed the primer extension reaction and visualized the reaction 226

products, we found that the pattern of products obtained with DNA from wild type and POT2 227

knockout cells was essentially the same. Most products corresponded to overhangs of 14 or 20 228

nucleotides, indicating that overhang length was unaffected by Pot2 loss (Fig. 1E). As expected, 229

DNA from POT1 knockout cells gave rise to longer primer extension products indicating G-230

on May 29, 2021 by guest

http://ec.asm.org/

Dow

nloaded from

10

overhang elongation (6, 33). Overall our results indicate that Pot2 is not needed for length 231

regulation or correct G-overhang processing at macronuclear telomeres. 232

233

234

Pot2 is expressed during macronuclear development at the time of new telomere 235

synthesis 236

Since Pot2 is not needed for telomere maintenance in vegetative cells, a likely 237

alternative role would be in new telomere synthesis or one of the other processing events that 238

take place when cells mate (Fig. 2A). To determine if Pot2 expression is developmentally 239

regulated, we isolated RNA from conjugating cells and used RT-PCR to examine transcript 240

abundance for both POT2 and POT1 throughout the mating process. Wild type cells of opposite 241

mating types were starved for 24 hours, mixed to initiate mating and RNA was isolated at two 242

hour intervals until 18 hours post mixing. RNA was isolated again at 24 hours, then the mated 243

cells were fed so they could resume vegetative growth and RNA was isolated two hours later at 244

26 hours. RNA was also collected from vegetatively growing and starved cells. The RT-PCR 245

analysis revealed that POT2 is expressed during macronuclear development and the 246

expression pattern is quite different from that of POT1 (Fig. 2B). 247

POT1 mRNA was detected in actively dividing vegetative cells but not in starved cells. It 248

was then up-regulated ∼2 hours after initiating mating and remained abundant until 6 hours. 249

This timing of expression spans the pre-zygotic meiotic and mitotic divisions. POT1 mRNA 250

abundance then diminished as conjugation proceeded until 14 hours when a slight up-regulation 251

occurred to coincide with DNA amplification from 4N to 8N after anlagen II (31, 37, 38). POT1 252

mRNA also accumulated after re-feeding which coincides with the final rounds of DNA 253

amplification in the new macronucleus (37, 38). 254

In contrast to POT1, the POT2 mRNA was undetectable or of very low abundance in 255

vegetative cells and during early time points after the initiation of mating (Fig. 2B). The transcript 256

on May 29, 2021 by guest

http://ec.asm.org/

Dow

nloaded from

11

was first detected at the10 hour time point which corresponded to the anlagen II stage (Fig. 2A) 257

when cells start the process of macronuclear development. Transcript levels remained elevated 258

until 14 hours by which time most cells were separated. This pattern of expression overlaps that 259

seen for proteins involved in late stages of IES excision (28, 39, 40). Since, chromosome 260

breakage is also thought to occur around this time, our observations suggested that Pot2 may 261

have a role in either IES excision or chromosome breakage. An additional increase in POT2 262

expression was observed at the 24 hour time point. The reason for the increase is unclear as 263

chromosome processing and conjugation events should be complete by this time (Fig 2A). 264

To determine if the timing of POT2 transcription coincides with the onset of chromosome 265

breakage, we used a PCR reaction to detect the 11kb rDNA (Fig S2A and Fig. 2C) (32). The 266

11kb molecule is a byproduct of rDNA processing that is only seen in conjugating cells following 267

chromosome breakage and new telomere addition. It is an excellent marker for these events 268

because the product gradually disappears from progeny cells after 11 population doublings 269

meaning that it is absent from the parental cells used for mating (41). When we performed the 270

PCR assay with DNA isolated from the same batches of mated cells used for the transcript 271

analysis (Fig. 2B), the first time point that the 11kb rDNA product could be detected was 12 272

hours after initiating mating (Fig. 2C). At this time point, the cells were leaving the anlagen II 273

stage of development and the pairs were beginning to separate. More 11 kb rDNA product was 274

detected at 14 hours by which time most cells had separated (Fig. 2A) and it rose further by 24 275

hours. The initial appearance of the 11 kb rDNA product at 12-14 hours indicates that 276

chromosome breakage and new telomere synthesis occur around the time of cell separation 277

and that POT2 transcription starts earlier. 278

To further explore the link between Pot2 expression and chromosome processing, we 279

examined the distribution of Pot2 protein in mated cells. Antibodies to Pot2 were raised for this 280

purpose and used to perform immunolocalization studies with mated and unmated cells (Fig. 281

2D, Fig. S2B). As expected from the transcript analysis, no Pot2 was observed in unmated cells 282

on May 29, 2021 by guest

http://ec.asm.org/

Dow

nloaded from

12

and it remained absent during the early stages of conjugation. The protein was first detected at 283

the anlagen II stage when it localized to the developing macronucleus. It was not present in the 284

micronuclei or the old macronucleus. One antibody gave some staining of the oral apparatus but 285

this was not observed with a second antibody (Fig. S2C) and hence was non-specific. Thus, the 286

timing of Pot2 expression and localization to the developing macronucleus is consistent with a 287

role in IES excision or chromosome breakage and new telomere synthesis. 288

289

Pot2 localizes to sites of chromosome breakage 290

If Pot2 is involved in IES excision or chromosome breakage, it is likely to associate with 291

IESs or CBSs. To test for this, we performed ChIP with cells harvested at the various stages in 292

macronuclear development (Fig. 3, Fig, S3A). Mated wild type cells were harvested at different 293

time points and portions of the culture were used to monitor progression through macronuclear 294

development and the timing of chromosome breakage (Fig. 3A). The remaining cells were 295

cross-linked with formaldehyde, the chromatin sheared and precipitated with antibody to Pot2. 296

Precipitations were also performed with antibody to Pot1 or Pdd1 as positive controls for 297

telomere or IES association. The precipitated DNA was purified and analyzed by real-time PCR. 298

Primer sets were designed to monitor association with two different CBSs, an IES, the rDNA 299

telomere and an internal rDNA control sequence. As expected, ChIP with Pot1 antibody 300

consistently enriched for the rDNA telomere (Fig. 3C) while Pdd1 antibody enriched for the IES 301

(Fig. 3D). However, the Pot2 antibody enriched for the CBSs but not the telomere or the IES 302

(Fig. 3B). CBS enrichment was apparent by 8 hours, it peaked between 10-12 hours and then 303

declined at 14 hours. When we used the 11 kb PCR assay to monitor the timing of chromosome 304

breakage, we found that this occurred slightly later than Pot2 association with the CBSs (Fig. 305

3A). The 11 kb rDNA was first detected at 10-12 hours and it became much more abundant by 306

14 hours. These results indicate that Pot2 binds directly or indirectly at regions of chromosome 307

breakage prior to CBS cleavage but it does not localize to telomeres or IESs. 308

on May 29, 2021 by guest

http://ec.asm.org/

Dow

nloaded from

13

We attempted to confirm the above ChIP data using HA antibody and cells expressing 309

HA-tagged Pot2. The cells had the endogenous macronuclear POT2 gene replaced with 310

sequence encoding N- or C-terminally tagged Pot2. However when these cells were used for 311

ChIP we obtained inconsistent results (data not shown). Since the timing of Pot2 expression 312

coincides with the start of mRNA production in the developing macronucleus (Fig S3B), the 313

problem may have been that the tagged protein encoded by the parental macronucleus was 314

replaced by wild type Pot2 expressed from the developing macronucleus. A similar 315

phenomenon would explain the lack of phenotype seen after mating POT2 macronuclear 316

knockout cells (data not shown). We attempted to directly test the role of Pot2 in chromosome 317

breakage by generating cell lines with combined micro- and macronuclear gene disruptions but 318

the micronuclear POT2 gene locus was refractory to disruption. 319

320

Pot2 does not bind directly to CBS or telomeric DNA in vitro 321

Since Pot2 localizes to CBSs, we next asked whether it can bind directly to the CBS 322

consensus sequence or to other DNA sequences. MBP-tagged Pot2 was expressed in insect 323

cells using baculovirus, purified on amylose resin (Fig. S4B) and tested for DNA binding 324

specificity using gel shift assays. Purified Pot1 was used as a positive control for telomeric DNA 325

binding (Fig. S4B). As Pot2 is predicted to contain several OB folds (Fig. 1A) and these motifs 326

primarily bind single-stranded nucleic acid (42), we tested for binding both to an oligonucleotide 327

duplex corresponding to the CBS consensus sequence and to the individual oligonucleotides 328

corresponding to the two strands of this sequence. However, no binding was observed (Fig. 329

4A). We also tested for binding to telomeric G-strand DNA but only detected binding by a 330

contaminating protein in the Pot2 preparation (Fig. 4B, Fig. S4D). As expected, Pot1 bound the 331

telomeric G-strand DNA with high affinity. 332

Although an N-terminal tag does not prevent Pot1 from associating with telomeres (6), 333

we were concerned that that the MBP tag might disrupt Pot2 binding. It was not possible to 334

on May 29, 2021 by guest

http://ec.asm.org/

Dow

nloaded from

14

perform binding assays with the untagged full-length Pot2 because the protein became 335

insoluble when the tag was removed. Since the isolated DNA-binding domains of POT family 336

members are generally soluble and bind DNA with high affinity (43, 44), we sought to avoid 337

potential problems due to the MBP tag by expressing C-terminal truncations of Pot1 and Pot2 338

that contained the predicted OB-fold domain. These putative DNA-binding domains were quite 339

soluble and the Pot1 truncation bound telomeric G-strand DNA as expected, indicating correct 340

protein folding However, the Pot2 truncation failed to bind any single or double stranded 341

telomeric or CBS substrate tested (Fig S4A). While we cannot rule a problem due to incorrect 342

folding of the isolated domain, our results suggest that that Pot2 does not bind directly to the 343

CBS consensus sequence. Thus, the association of Pot2 with CBSs may be via an unknown 344

binding partner. Alternatively, binding may only occur in the context of a nucleosome. The 345

inability of Pot2 to bind telomeric DNA fits with the lack of telomere localization seen in the ChIP 346

analysis, again indicating that Pot2 is not a canonical telomere binding protein. 347

348

POT2 associates with CBSs without Pat1 or Tpt1 349

At the telomere, Pot1 functions in combination with Tpt1, Pat1 and Pat2. Since Pot1 and 350

Pot2 have significant sequence identity and a similar domain structure, we wished to know 351

whether Pot2 also functions in association with these proteins. Initially we examined the mRNA 352

expression profile of two other components of the telomeric G-overhang binding complex; the 353

Pot1-binding partner Tpt1, and Pat1 which binds to Tpt1. As before, RNA was isolated from 354

mated cells at various time points during conjugation and macronuclear development. RT-PCR 355

analysis indicated that Pat1 and Tpt1 were both expressed in vegetative cells as expected. 356

They were also expressed during macronuclear development but their patterns of up- and 357

down-regulation were less distinct than those of Pot1 or Pot2 (Fig. 5A). Overall the profiles were 358

more similar to that of Pot1 as both Pat1 and Tpt1 were present during the early stages of 359

conjugation and at 26 hours after refeeding. 360

on May 29, 2021 by guest

http://ec.asm.org/

Dow

nloaded from

15

We next used ChIP to examine the association of Tpt1 and Pat1 with CBSs, IESs and 361

telomeres. The ChIP was performed with mated wild type cells and Tpt1 or Pat1 antibody (5). 362

The chromatin was from the same mating experiments shown in Fig. 3. Our analysis revealed 363

that the Tpt1 and Pat1 antibodies both enriched for telomeric DNA but not for CBS or IES DNA 364

(Fig. 5B). In each case, the overall ChIP profile was similar to that observed with antibody to 365

Pot1. We therefore conclude that Pat1 and Tpt1 do not associate with Pot2 at CBSs and that 366

the binding partners of Pot2 are different from those of Pot1. 367

368

CBSs lack H3K27Me3, and H3K9Me3 369

Our finding that Pot2 localizes to CBSs but not IESs indicates that there are likely to be 370

mechanistic differences between the processes of chromosome breakage and IES excision. 371

This conclusion is supported by our observation that Pdd1 localizes to IESs but not CBSs (Fig. 372

3D) and a report that this is also true for the Pdd1 interaction partner Pdd3 (39). To further 373

explore possible mechanistic differences between the two DNA processing pathways, we used 374

ChIP to determine if the heterochromatin marks H3K9Me3 and H3K27Me3 accumulate at or 375

near CBSs prior to chromosome breakage. The generation of these histone modifications is an 376

obligatory step in the scnRNA-mediated IES excision pathway because they are recognized by 377

proteins needed for the DNA cleavage reaction (Pdd1 and Tpb2) (9, 26, 29, 30, 45). Thus, their 378

absence from CBSs would indicate a major difference between the IES and chromosome 379

breakage pathways. To test for histone methylation at CBSs, we performed ChIP (Fig. 6A-B) 380

using antibody to tri-methylated H3K9 or H3K27 and chromatin from the mating time courses 381

described above. Analysis of the precipitated DNA confirmed previous studies indicating that 382

H3K9Me3 and H3K27Me3 accumulate at IESs with H3K27 methylation peaking slightly before 383

H3K9 methylation (26, 46). Interestingly, only low levels of CBS DNA were precipitated by the 384

tri-methylated H3K9 or H3K27 antibody and there was no significant enrichment of CBS DNA 385

relative to the control rDNA sequence. Thus, the nucleosomes at or near CBSs do not 386

on May 29, 2021 by guest

http://ec.asm.org/

Dow

nloaded from

16

accumulate H3K9Me3 and H3K27Me3 marks during the course of macronuclear development. 387

We therefore conclude that, the mechanisms of chromosome breakage and IES excision must 388

be quite different. 389

Given this mechanistic difference, it is interesting that disruption of IES excision can also 390

impair chromosome breakage. The one exception in the literature is the macronuclear knockout 391

of PDD1 which leaves chromosome breakage intact (28). Since the macronuclear knockout 392

disrupts only early Pdd1 expression, we revisited this observation using cells with a full 393

macronuclear and micronuclear gene disruption to prevent both early and late Pdd1 expression. 394

Chromosome breakage was monitored with the 11 kb rDNA assay using DNA isolated from 395

mated cells. Consistent with published data, disruption of early Pdd1 expression had no effect 396

on the timing or the amount of 11 kb PCR product that was generated (Fig. S3C, left panel). 397

However, disruption of early and late Pdd1 expression not only delayed the occurrence of 398

chromosome breakage, but decreased the overall amount of 11 kb PCR product (Fig. S3C, right 399

panel). Thus, it appears that impaired chromosome breakage is a general, but most likely 400

indirect, outcome of impaired IES excision. 401

402

DISCUSSION 403

Here we describe a new member of the POT (Protection of Telomeres) protein family, 404

Tetrahymena Pot2, which functions outside of the established telomere maintenance pathway. 405

Although Pot2 resembles Pot1 in sequence and predicted protein structure, Pot2 is unable to 406

bind telomeric DNA, it does not localize to telomeres and it is not needed for telomere length 407

regulation or other aspects of macronuclear telomere maintenance. Instead, Pot2 is expressed 408

during the sexual stage of the Tetrahymena life cycle when it accumulates in the developing 409

macronucleus slightly before the onset of chromosome breakage and new telomere synthesis. 410

ChIP studies revealed that Pot2 localizes to CBSs, the sequences that mark the site of 411

chromosome breakage, but not to IESs. Thus, Pot2 is the first protein in Tetrahymena to be 412

on May 29, 2021 by guest

http://ec.asm.org/

Dow

nloaded from

17

specifically linked to the process of chromosome breakage and new telomere synthesis. 413

Although Pot1 and its binding partners Tpt1 and Pat1 localize to telomeres in mated cells, we 414

did not detect them at CBSs. This means that Pot2 must function at CBSs in conjunction with 415

novel interaction partners and since Pot2 does not appear to bind directly to the CBS consensus 416

sequence, these proteins may direct Pot2 to the CBS. Taken together our results indicate that 417

Pot2 has evolved a role independent of the telomere protection and maintenance functions 418

normally associated with the POT family of proteins. 419

The functional evolution of POT proteins is not without precedence. Mouse has two Pot 420

proteins, Pot1a and Pot1b, with 72% sequence identity. Both participate in telomere end 421

protection but they show only partial functional overlap as the primary role of Pot1a is to inhibit 422

DNA damage signaling while Pot1b prevents telomeric C-strand resection (47, 48). Arabidopsis 423

has three Pot proteins, Pot1a, Pot1b and Pot1c, which exhibit a greater degree of functional 424

divergence. Pot1b and Pot1c both appear to participate in telomere end protection but Pot1a is 425

necessary for telomerase action (49-51). Pot1a does not bind telomeric DNA but instead is a 426

component of the telomerase holoenzyme (49, 52). 427

We attempted to determine the role of Tetrahymena Pot2 during macronuclear 428

development through POT2 gene disruption, but we were only able to disrupt the macronuclear 429

gene and this did not give rise to a phenotype. The lack of phenotype most likely reflected 430

rescuing Pot2 expression from the micronucleus-derived gene during macronuclear 431

development. Nonetheless, our finding that Pot2 localizes to CBSs but apparently does not bind 432

either telomeric DNA or the CBS consensus sequence, suggests several novel functions for this 433

member of the POT protein family. One possibility is that Pot2 interacts with telomerase in a 434

manner akin to Arabidopsis Pot1a to direct addition of telomeric repeats at the site of 435

chromosome breakage (see model in Fig. 6C). Some form of telomerase recruitment factor is 436

likely to be necessary because the broken ends lack sequence complementary to the 437

telomerase RNA template (16, 17) and hence are not good substrates to seed new telomere 438

on May 29, 2021 by guest

http://ec.asm.org/

Dow

nloaded from

18

addition (12).Thus far, we have been unable to detect an interaction between Pot2 and 439

telomerase by precipitating Pot2 and assaying the precipitate for telomerase activity. However 440

this negative result could be explained if the interaction is transient, it occurs only on chromatin 441

or if involves only a small fraction of the telomerase present in mated cells. An alternative, but 442

not mutually exclusive role for Pot2 might be to recruit the machinery responsible for the DNA 443

cleavage reaction. Cleavage at the CBS must be tightly regulated to prevent inappropriate 444

chromosome breakage in the micronucleus and to ensure rapid telomere addition to the newly 445

broken ends in the developing macronucleus. A delay in telomere addition would allow the 446

broken ends to be recognized as DNA damage and resected by nuclease leading to loss of 447

adjacent coding sequence (32, 40). Pot2 does not appear to contain an endonuclease domain 448

so it is unlikely to carry out the DNA cleavage reaction. However, by associating simultaneously 449

with a CBS recognition factor and telomerase, Pot2 would be well positioned to coordinate DNA 450

cleavage with new telomere addition (Fig. 6C). 451

Our finding that Pot2 associates with CBSs but not IESs, while the converse is true for 452

Pdd1, prompted us to look more closely for differences between the chromosome breakage and 453

IES excision pathways. The discovery that CBS lack tri-methlyated H3K9 and H3K27 indicates 454

a fundamental mechanistic difference in the two chromosome processing pathways. 455

Generation of H3K9Me3 and H3K27Me3 marks at IESs is the culmination of the RNAi-based 456

scanning process to delineate IESs from the surrounding macronuclear-destined sequence (8, 457

25, 26, 46). Moreover, these marks are required for the subsequent cleavage reaction because 458

they are recognized by Pdd1 and the transposase-like protein Tpb2 (14, 27, 29). Thus, the lack 459

of H3K9Me3 and H3K27Me3 on CBSs means that chromosome breakage is unlikely to be 460

driven by scnRNA mediated heterochromatin formation. Given that CBSs contain a well 461

conserved consensus sequence, they are likely to be marked for cleavage by a protein that 462

directly recognizes this sequence. As the ultimate fate of IESs and CBSs is quite different (DNA 463

on May 29, 2021 by guest

http://ec.asm.org/

Dow

nloaded from

19

elimination and re-ligation versus cleavage with telomere addition) it is logical that the cell would 464

mark the two regions differently. 465

It remains to be seen whether any components of the IES excision pathway are used in 466

chromosome breakage. While Tpb2 is a candidate to cleave the DNA at CBSs (27), a 467

transposase-like protein may be an inappropriate choice as there is no subsequent ligation 468

reaction. It is still unclear why disruption of IES excision should also prevent or greatly reduce 469

chromosome breakage. However, cells that are unable to carry out IES excision exhibit a 470

disruption in their developmental program as they arrest with two micronuclei and two 471

macronuclei, rather than one micronucleus and two macronuclei, and they are unable to resume 472

growth on refeeding (21-23, 25, 26, 53). Thus, disruption of IES excision may lead to a 473

developmental checkpoint that also prevents chromosome breakage. 474

In summary, our finding that Pot2 localizes to CBSs but not to telomeres reveals a novel 475

function for a telomere protein homolog. While further studies are required to delineate the 476

precise role of Pot2 in chromosome breakage, the current analysis of Pot2 has uncovered a 477

clear functional separation between the two chromosome processing events associated with 478

Tetrahymena macronuclear development. The study also provides hints of a broader 479

developmental checkpoint as the explanation for much of the apparent overlap between 480

chromosome breakage and IES excision pathways. Overall, the work begins to unravel the 481

functional evolution of telomere proteins in Tetrahymena and starts to differentiate the process 482

of chromosome breakage from its better understood counterpart IES excision. 483

484

ACKNOWLEDGEMENTS 485

We thank Douglas Chalker for the PDD1 knockout cells, Jacob Naduparambil for 486

assistance with early experiments and members of the Price lab for helpful discussions. This 487

work was supported by National Institutes of Health (NIH) grant RO1 GM088728 to CMP. BRL 488

was supported by NIH grant T32 CA117846 and SC by NIH grant T32 ES007250. 489

on May 29, 2021 by guest

http://ec.asm.org/

Dow

nloaded from

20

490

REFERENCES 491

1. O'Sullivan RJ, Karlseder J. 2010. Telomeres: protecting chromosomes against 492

genome instability. Nat Rev Mol Cell Biol 11:171-181. 493

2. Stewart JA, Chaiken MF, Wang F, Price CM. 2012. Maintaining the end: roles of 494

telomere proteins in end-protection, telomere replication and length regulation. Mutation 495

research 730:12-19. 496

3. Pfeiffer V, Lingner J. 2013. Replication of telomeres and the regulation of telomerase. 497

Cold Spring Harb Perspect Biol 5:a010405. 498

4. Letchoumy Premkumar V, Cranert S, Linger BR, Morin GB, Minium S, Price C. 499

2013. The 3' overhangs at Tetrahymena telomeres are packaged by four proteins: 500

Pot1a, Tpt1, Pat1 and Pat2. Eukaryot Cell. 501

5. Linger BR, Morin GB, Price CM. 2011. The Pot1a-associated proteins Tpt1 and Pat1 502

coordinate telomere protection and length regulation in Tetrahymena. Molecular biology 503

of the cell 22:4161-4170. 504

6. Jacob NK, Lescasse R, Linger BR, Price CM. 2007. Tetrahymena POT1a regulates 505

telomere length and prevents activation of a cell cycle checkpoint. Molecular and cellular 506

biology 27:1592-1601. 507

7. Karrer KM. 2012. Nuclear dualism. Methods Cell Biol 109:29-52. 508

8. Chalker DL, Meyer E, Mochizuki K. 2013. Epigenetics of ciliates. Cold Spring Harb 509

Perspect Biol 5:a017764. 510

9. Chalker DL. 2008. Dynamic nuclear reorganization during genome remodeling of 511

Tetrahymena. Biochim Biophys Acta 1783:2130-2136. 512

10. Fan Q, Yao M. 1996. New telomere formation coupled with site-specific chromosome 513

breakage in Tetrahymena thermophila. Molecular and cellular biology 16:1267-1274. 514

on May 29, 2021 by guest

http://ec.asm.org/

Dow

nloaded from

21

11. Hamilton E, Bruns P, Lin C, Merriam V, Orias E, Vong L, Cassidy-Hanley D. 2005. 515

Genome-wide characterization of tetrahymena thermophila chromosome breakage sites. 516

I. Cloning and identification of functional sites. Genetics 170:1611-1621. 517

12. Yu GL, Blackburn EH. 1991. Developmentally programmed healing of chromosomes by 518

telomerase in Tetrahymena. Cell 67:823-832. 519

13. Chalker DL, Yao MC. 2011. DNA elimination in ciliates: transposon domestication and 520

genome surveillance. Annual review of genetics 45:227-246. 521

14. Mochizuki K. 2012. Developmentally programmed, RNA-directed genome 522

rearrangement in Tetrahymena. Dev Growth Differ 54:108-119. 523

15. Tanaka H, Yao MC. 2009. Palindromic gene amplification--an evolutionarily conserved 524

role for DNA inverted repeats in the genome. Nature reviews. Cancer 9:216-224. 525

16. Yao MC, Zheng K, Yao CH. 1987. A conserved nucleotide sequence at the sites of 526

developmentally regulated chromosomal breakage in Tetrahymena. Cell 48:779-788. 527

17. Yao MC, Yao CH, Monks B. 1990. The controlling sequence for site-specific 528

chromosome breakage in Tetrahymena. Cell 63:763-772. 529

18. Hamilton EP, Williamson S, Dunn S, Merriam V, Lin C, Vong L, Russell-Colantonio 530

J, Orias E. 2006. The highly conserved family of Tetrahymena thermophila chromosome 531

breakage elements contains an invariant 10-base-pair core. Eukaryotic cell 5:771-780. 532

19. Kapler GM. 1993. Developmentally regulated processing and replication of the 533

Tetrahymena rDNA minichromosome. Curr Opin Genet Dev 3:730-735. 534

20. Mochizuki K, Gorovsky MA. 2004. Conjugation-specific small RNAs in Tetrahymena 535

have predicted properties of scan (scn) RNAs involved in genome rearrangement. 536

Genes Dev 18:2068-2073. 537

21. Malone CD, Anderson AM, Motl JA, Rexer CH, Chalker DL. 2005. Germ line 538

transcripts are processed by a Dicer-like protein that is essential for developmentally 539

on May 29, 2021 by guest

http://ec.asm.org/

Dow

nloaded from

22

programmed genome rearrangements of Tetrahymena thermophila. Molecular and 540

cellular biology 25:9151-9164. 541

22. Mochizuki K, Gorovsky MA. 2005. A Dicer-like protein in Tetrahymena has distinct 542

functions in genome rearrangement, chromosome segregation, and meiotic prophase. 543

Genes Dev 19:77-89. 544

23. Mochizuki K, Fine NA, Fujisawa T, Gorovsky MA. 2002. Analysis of a piwi-related 545

gene implicates small RNAs in genome rearrangement in tetrahymena. Cell 110:689-546

699. 547

24. Aronica L, Bednenko J, Noto T, DeSouza LV, Siu KW, Loidl J, Pearlman RE, 548

Gorovsky MA, Mochizuki K. 2008. Study of an RNA helicase implicates small RNA-549

noncoding RNA interactions in programmed DNA elimination in Tetrahymena. Genes 550

Dev 22:2228-2241. 551

25. Liu Y, Mochizuki K, Gorovsky MA. 2004. Histone H3 lysine 9 methylation is required 552

for DNA elimination in developing macronuclei in Tetrahymena. Proceedings of the 553

National Academy of Sciences of the United States of America 101:1679-1684. 554

26. Liu Y, Taverna SD, Muratore TL, Shabanowitz J, Hunt DF, Allis CD. 2007. RNAi-555

dependent H3K27 methylation is required for heterochromatin formation and DNA 556

elimination in Tetrahymena. Genes Dev 21:1530-1545. 557

27. Cheng CY, Vogt A, Mochizuki K, Yao MC. 2010. A domesticated piggyBac 558

transposase plays key roles in heterochromatin dynamics and DNA cleavage during 559

programmed DNA deletion in Tetrahymena thermophila. Molecular biology of the cell 560

21:1753-1762. 561

28. Coyne RS, Nikiforov MA, Smothers JF, Allis CD, Yao MC. 1999. Parental expression 562

of the chromodomain protein Pdd1p is required for completion of programmed DNA 563

elimination and nuclear differentiation. Mol Cell 4:865-872. 564

on May 29, 2021 by guest

http://ec.asm.org/

Dow

nloaded from

23

29. Schwope RM, Chalker DL. 2014. Mutations in Pdd1 reveal distinct requirements for its 565

chromodomain and chromoshadow domain in directing histone methylation and 566

heterochromatin elimination. Eukaryotic cell 13:190-201. 567

30. Vogt A, Mochizuki K. 2013. A domesticated PiggyBac transposase interacts with 568

heterochromatin and catalyzes reproducible DNA elimination in Tetrahymena. PLoS 569

genetics 9:e1004032. 570

31. Nikiforov MA, Smothers JF, Gorovsky MA, Allis CD. 1999. Excision of micronuclear-571

specific DNA requires parental expression of pdd2p and occurs independently from DNA 572

replication in Tetrahymena thermophila. Genes Dev 13:2852-2862. 573

32. Lin IT, Chao JL, Yao MC. 2012. An essential role for the DNA breakage-repair protein 574

Ku80 in programmed DNA rearrangements in Tetrahymena thermophila. Molecular 575

biology of the cell 23:2213-2225. 576

33. Jacob NK, Skopp R, Price CM. 2001. G-overhang dynamics at Tetrahymena 577

telomeres. The EMBO journal 20:4299-4308. 578

34. Jacob NK, Stout AR, Price CM. 2004. Modulation of telomere length dynamics by the 579

subtelomeric region of tetrahymena telomeres. Molecular biology of the cell 15:3719-580

3728. 581

35. Wenkert D, Allis CD. 1984. Timing of the appearance of macronuclear-specific histone 582

variant hv1 and gene expression in developing new macronuclei of Tetrahymena 583

thermophila. The Journal of cell biology 98:2107-2117. 584

36. Nelson JD, Denisenko O, Sova P, Bomsztyk K. 2006. Fast chromatin 585

immunoprecipitation assay. Nucleic Acids Res 34:e2. 586

37. Allis CD, Colavito-Shepanski M, Gorovsky MA. 1987. Scheduled and unscheduled 587

DNA synthesis during development in conjugating Tetrahymena. Dev Biol 124:469-480. 588

on May 29, 2021 by guest

http://ec.asm.org/

Dow

nloaded from

24

38. Yin L, Gater ST, Karrer KM. 2010. A developmentally regulated gene, ASI2, is required 589

for endocycling in the macronuclear anlagen of Tetrahymena. Eukaryotic cell 9:1343-590

1353. 591

39. Nikiforov MA, Gorovsky MA, Allis CD. 2000. A novel chromodomain protein, pdd3p, 592

associates with internal eliminated sequences during macronuclear development in 593

Tetrahymena thermophila. Molecular and cellular biology 20:4128-4134. 594

40. Rexer CH, Chalker DL. 2007. Lia1p, a novel protein required during nuclear 595

differentiation for genome-wide DNA rearrangements in Tetrahymena thermophila. 596

Eukaryotic cell 6:1320-1329. 597

41. Pan WC, Blackburn EH. 1981. Single extrachromosomal ribosomal RNA gene copies 598

are synthesized during amplification of the rDNA in Tetrahymena. Cell 23:459-466. 599

42. Lewis KA, Wuttke DS. 2012. Telomerase and telomere-associated proteins: structural 600

insights into mechanism and evolution. Structure 20:28-39. 601

43. Lei M, Podell ER, Cech TR. 2004. Structure of human POT1 bound to telomeric single-602

stranded DNA provides a model for chromosome end-protection. Nature structural & 603

molecular biology 11:1223-1229. 604

44. Croy JE, Podell ER, Wuttke DS. 2006. A new model for Schizosaccharomyces pombe 605

telomere recognition: the telomeric single-stranded DNA-binding activity of Pot11-389. 606

Journal of molecular biology 361:80-93. 607

45. Smothers JF, Mizzen CA, Tubbert MM, Cook RG, Allis CD. 1997. Pdd1p associates 608

with germline-restricted chromatin and a second novel anlagen-enriched protein in 609

developmentally programmed DNA elimination structures. Development 124:4537-4545. 610

46. Taverna SD, Coyne RS, Allis CD. 2002. Methylation of histone h3 at lysine 9 targets 611

programmed DNA elimination in tetrahymena. Cell 110:701-711. 612

47. Wu L, Multani AS, He H, Cosme-Blanco W, Deng Y, Deng JM, Bachilo O, Pathak S, 613

Tahara H, Bailey SM, Deng Y, Behringer RR, Chang S. 2006. Pot1 deficiency initiates 614

on May 29, 2021 by guest

http://ec.asm.org/

Dow

nloaded from

25

DNA damage checkpoint activation and aberrant homologous recombination at 615

telomeres. Cell 126:49-62. 616

48. Hockemeyer D, Daniels JP, Takai H, de Lange T. 2006. Recent expansion of the 617

telomeric complex in rodents: Two distinct POT1 proteins protect mouse telomeres. Cell 618

126:63-77. 619

49. Surovtseva YV, Shakirov EV, Vespa L, Osbun N, Song X, Shippen DE. 2007. 620

Arabidopsis POT1 associates with the telomerase RNP and is required for telomere 621

maintenance. EMBO J 26:3653-3661. 622

50. Shakirov EV, Surovtseva YV, Osbun N, Shippen DE. 2005. The Arabidopsis Pot1 and 623

Pot2 proteins function in telomere length homeostasis and chromosome end protection. 624

Molecular and cellular biology 25:7725-7733. 625

51. Shakirov EV, Song X, Joseph JA, Shippen DE. 2009. POT1 proteins in green algae 626

and land plants: DNA-binding properties and evidence of co-evolution with telomeric 627

DNA. Nucleic Acids Res 37:7455-7467. 628

52. Cifuentes-Rojas C, Kannan K, Tseng L, Shippen DE. 2011. Two RNA subunits and 629

POT1a are components of Arabidopsis telomerase. Proceedings of the National 630

Academy of Sciences of the United States of America 108:73-78. 631

53. Bednenko J, Noto T, DeSouza LV, Siu KW, Pearlman RE, Mochizuki K, Gorovsky 632

MA. 2009. Two GW repeat proteins interact with Tetrahymena thermophila argonaute 633

and promote genome rearrangement. Molecular and cellular biology 29:5020-5030. 634

635

636

637

on May 29, 2021 by guest

http://ec.asm.org/

Dow

nloaded from

26

FIGURE LEGENDS 638

Table 1. List of oligonucleotides used in experiments. 639

640

Figure 1. Cell growth and telomere maintenance are unaffected by POT2 macronuclear 641

gene disruption. (A) Schematic of the macronuclear POT1 and POT2 gene loci and the 642

encoded proteins. Black boxes, exons; arrows, transcription start sites; grey boxes, predicted 643

OB-folds. (B) Growth curves for wild type cells (black line) and two POT2 knockout clones 644

(dashed lines). PD, population doubling. (C) Southern blot showing rDNA telomere length in wild 645

type cells (WT) and three different POT2 knockout clones (POT2 KO). The probe was to the 646

subtelomeric region of the rDNA. M indicates size markers. (D-E) Ligation and primer extension 647

assay to measure G-overhang length. (D) Top diagram illustrates the assay. Bottom diagram 648

illustrates the products expected for a 14 nt overhang. (E) Polyacrylamide gel showing reaction 649

products obtained with DNA from wild type (lane 1), POT1 KO (lane 2) or POT2 KO cells (lane 650

3). Products corresponding to overhangs of 14, 20, 26, and 32 nucleotides are marked at the 651

right; positions of the marker oligonucleotides are shown on the left. 652

653

Figure 2. Pot2 expression and localization during macronuclear development. A. Cartoon 654

depicting stages of Tetrahymena conjugation and macronuclear development. After 2-3 hours of 655

starvation cells pair and the micronuclei undergo two rounds of meiosis. Three of the four nuclei 656

are degraded and the remaining nucleus undergoes mitosis to form two haploid pronuclei. One 657

of the pronuclei is exchanged with the partner cell and the pronuclei then fuse to form a diploid 658

zygotic nucleus. After two rounds of mitosis (Anlagen I), two of the mitotic products begin to 659

develop into new macronuclei (Anlagen II) while two remain as micronuclei. The cells separate 660

and complete macronuclear development before degrading one of the new micronuclei. The 661

on May 29, 2021 by guest

http://ec.asm.org/

Dow

nloaded from

27

cells remain with one micronucleus and two macronuclei until nutrients become available at 662

which point they resume vegetative growth. (B) POT1 and POT2 mRNA expression profile. RT-663

PCR was performed with RNA collected from vegetatively growing cells (V), starved cells (S), or 664

mated cells harvested at the indicated times after the initiation of mating. Cells were re-fed after 665

24 hours and RNA was collected two hours later (26 hours). PCR with genomic DNA (G) 666

monitored DNA contamination, RT-PCR with U1 snRNA controlled for RNA quality and loading. 667

(C) Chromosome breakage analysis using the 11 kb PCR assay. The assay was performed with 668

DNA isolated at the indicated time points from the mated cells used in (B). PCR of the U1 669

snRNA gene controlled for DNA quality and loading. (D) Immunolocalization of Pot2. Cells at 670

the indicated stages in conjugation and macronuclear development were stained with Pot2 671

antibody and counterstained with DAPI. Arrow heads, micronuclei; arrows, developing 672

macronuclei; stars, parental macronuclei. 673

674

Figure 3. Pot2 localizes to sites of chromosome breakage. (A) Panel I, Representative 675

images of acridine orange stained cells showing stages of macronuclear development used for 676

scoring. Panel II, Percent of cells at the indicated stages of macronuclear development at each 677

time point. Values are the average from the three ChIP experiments shown in (B-D), error bars 678

represent S.E.M. Panel III, 11 kb PRC assay for chromosome breakage. Samples were from 679

one of the time courses used to generate data in panel II and (B-D). (B-D) ChIP analysis of 680

chromatin association throughout macronuclear development. Chromatin was collected at 8, 10, 681

12 and 14 hours after the initiation of mating. Input DNA and precipitated DNA were quantified 682

by real time PCR, n = 3 independent experiments, error bars represent S.E.M. ChIP was 683

performed with Pot2 (B), Pot1 (C) and Pdd1 (D) antibody. 684

685

on May 29, 2021 by guest

http://ec.asm.org/

Dow

nloaded from

28

Figure 4. Pot2 lacks binding specificity for CBS or telomeric DNA. Mobility shift assays 686

using purified Pot1 and Pot2 and oligonucleotides corresponding to the telomeric G-strand 687

overhang (A) or the CBS 4L-6 duplex and the two strands of CBS 4L-6 DNA (B). The bracket in 688

(A) marks DNA-protein complexes formed by a contaminant in the Pot2 preparation (see Fig. 689

S4). 690

691

Figure 5. Pat1 and Tpt1 localize to telomeres but not sites of chromosome breakage. (A) 692

mRNA expression profile for POT2, POT1, PAT1 and TPT1 during conjugation and 693

macronuclear development. RNA was collected and analyzed by RT-PCR as described for Fig. 694

2B. (B) ChIP analysis of Pat1 and Tpt1 association with CBS 4L-6, IES M-element (M-El) and 695

telomeres throughout macronuclear development. Chromatin was from the mating time courses 696

shown in Fig. 2. ChIP was with antibody to Pat1 or Tpt1, n = 3 experiments, error bars represent 697

S.E.M. 698

699

Figure 6. H3K27Me3 and H3K9Me3 do not accumulate at sites of chromosome breakage. 700

(A) ChIP analysis to detect methylated histone H3 at CBS 4L-6, IES M-element (M-El) and 701

telomeres during macronuclear development. Chromatin was from the mating time courses 702

shown in Fig. 2. ChIP was with antibody to H3K27Me3 (A) and H3K9Me3 (B), n = 3 703

experiments, error bars represent S.E.M. (B) Model showing possible roles for Pot2 in recruiting 704

the endonuclease responsible for chromosome cleavage and telomerase. MDS, macronuclear 705

destined sequence, RF, recruitment factor. 706

on May 29, 2021 by guest

http://ec.asm.org/

Dow

nloaded from

on May 29, 2021 by guest

http://ec.asm.org/

Dow

nloaded from

on May 29, 2021 by guest

http://ec.asm.org/

Dow

nloaded from

on May 29, 2021 by guest

http://ec.asm.org/

Dow

nloaded from

on May 29, 2021 by guest

http://ec.asm.org/

Dow

nloaded from

on May 29, 2021 by guest

http://ec.asm.org/

Dow

nloaded from

on May 29, 2021 by guest

http://ec.asm.org/

Dow

nloaded from

Oligo Name Sequence Used for POT2 R5 5'-CTAGTTCTCATATACATGTATTTG RT-PCR of POT2

POT2 F13 5'-TGACTACACAAATGTGCAAAGAC RT-PCR of POT2 POT1 F1 5'- GCCTTGACCCTATTCCAATG RT-PCR of POT1

POT1 R1 5' ACGTGGTGCTTTTCTCTTTCT RT-PCR of POT1

U1 snRNA F1 5'-CTTACCTGGCTGGAGTTTGCTATC RT-PCR of U1 snRNA

U1 snRNA R1 5’-GCGGAGACAGCACTAAGTGCACG RT-PCR of U1 snRNA

R3c 5'-CGCAGTTTCGCTGTTCAATA 11 kb PCR

F3 5'-CCCCAACCCCAACCCCAACCCCA 11 kb PCR

CBS 4L-6 F1 5’-TCTTTTACAAAAATTGGCTTTATTGA ChIP of CBS 4L-6

CBS 4L-6 R1 5’-TTTCAGAACACATCACCGATATTT ChIP of CBS 4L-7

CBS 3R-2 F1 5'- AACCAACCTCTTTCGACTTAGGAAT ChIP of CBS 3R-2

CBS 3R-2 R1 5'-CAGCAATCAGGAAATAACATACCAC ChIP of CBS 3R-3

IES M-El F 5’-GTGTGGTACAATAGGTTGTCGTAG ChIP of IES M-element

IES M-El R 5’-TTGAAAGCTAAGTAGCCTTCTTGC ChIP of IES M-element

26S rRNA subtel 5'GAACTTCAATCTTTGACTAGC ChIP of rDNA subtel

26S rRNA subtel 5'AATTTCTTTGACATTGAGTAAAAGTTATTTATT

ChIP of rDNA subtel

rDNA internal 5'TGAAATTGCAAGGTAGGTTTC ChIP of rDNA int

rDNA internal 5'CATAGTTACTCCCGCCGTT ChIP of rDNA int

CBS 4L-6 F 5'GAATTGCATATAAACCAACCTCTTTTTAAATATCGGTG

Pot1/Pot2 EMSA

CBS 4L-6 R 5'CACCGATATTTAAAAAGAGGTTGGTTTATATGCAATTC

Pot1/Pot2 EMSA

Tel G-strand 5'TTGGGGTTGGGGTTGGGGTT Pot1/Pot2 EMSA

Tel C 5'CCCAACCCCAACCCCAACCC Pot1/Pot2 DBD EMSA

Tel G (20) 5'GTTGGGGTTGGGGTTGGGGT Pot1/Pot2 DBD EMSA

Tel G (14) 5'GTTGGGGTTGGGGT Pot1/Pot2 DBD EMSA

Tet G (12) 5'TGGGGTTGGGGT Pot1/Pot2 DBD EMSA

CBS 20 F 5'TATAAAGAGGTTGGTTTATT Pot1/Pot2 DBD EMSA

CBS 20 R 5'AATAAACCAACCTCTTTATA Pot1/Pot2 DBD EMSA

hTel 5'GGTTAGGGTTAGGGTTAGGG Pot1/Pot2 DBD EMSA

Tel Dup G-strand 5’-GGCTTAAGC(GGGGTT)7GGGGT Pot1/Pot2 DBD EMSA

Tel Dup C-strand 5’-ACCCC(AACCCC)4GCTTAAGCC Pot1/Pot2 DBD EMSA

Cranert et al. Table 1

on May 29, 2021 by guest

http://ec.asm.org/

Dow

nloaded from