The Common Ancestral Core of Vertebrate and Fungal Telomerase Rnas Xiaodong Qi1, Yang Li1, Shinji Honda2, Steve Hoffmann3,4, Manja Marz5, Axel Mosig6, Joshua D

Nucleic Acids Research Advance Access published October 23, 2012 Nucleic Acids Research, 2012, 1–13 doi:10.1093/nar/gks980

The common ancestral core of vertebrate and fungal telomerase RNAs Xiaodong Qi1, Yang Li1, Shinji Honda2, Steve Hoffmann3,4, Manja Marz5, Axel Mosig6, Joshua D. Podlevsky1, Peter F. Stadler4, Eric U. Selker2 and Julian J.-L. Chen1,*

1Department of Chemistry & Biochemistry, Arizona State University, Tempe, AZ 85287, USA, 2Institute of Molecular Biology, University of Oregon, Eugene, OR 97403, USA, 3LIFE Center, University of Leipzig, D-04107 Leipzig, Germany, 4Interdisciplinary Center for Bioinformatics, University of Leipzig, D-04107 Leipzig, Germany, 5Department of Bioinformatics, Friedrich Schiller University of Jena, D-07743 Jena, Germany and 6Max-Planck Institute for Mathematics in the Sciences, D-04107 Leipzig, Germany Downloaded from

Received June 27, 2012; Revised September 5, 2012; Accepted September 27, 2012

ABSTRACT INTRODUCTION http://nar.oxfordjournals.org/ Telomerase is a ribonucleoprotein with an intrinsic Eukaryotic chromosomes are terminally capped by special telomerase RNA (TER) component. Within yeasts, DNA–protein complexes, called telomeres. Telomeric TER is remarkably large and presents little similarity DNA typically shortens with each cell division owing to in secondary structure to vertebrate or ciliate incomplete end replication, ultimately resulting in TERs. To better understand the evolution of fun- chromosome instability and cellular senescence (1). Counteracting telomere shortening, telomerase synthesizes gal telomerase, we identified 74 TERs from short telomeric DNA repeats onto chromosome termini. Pezizomycotina and Taphrinomycotina subphyla, Human telomerase is highly processive, capable of sister clades to budding yeasts. We initially synthesizing hundreds of telomeric DNA repeats onto a at Arizona State University West 2 on October 23, 2012 identified TER from Neurospora crassa using a given primer. Mutations that impair telomerase function novel deep-sequencing–based approach, and hom- result in stem cell defects and have been linked to a ologous TER sequences from available fungal growing number of human diseases, including genome databases by computational searches. dyskeratosis congenita, aplastic anemia and idiopathic Remarkably, TERs from these non-yeast fungi pulmonary fibrosis (2). Recently, mutations decreasing tel- have many attributes in common with vertebrate omerase repeat addition processivity have been linked to TERs. Comparative phylogenetic analysis of highly familial pulmonary fibrosis (3,4). conserved regions within Pezizomycotina TERs Telomerase functions as a ribonucleoprotein enzyme, revealed two core domains nearly identical in sec- requiring the catalytic telomerase reverse transcriptase ondary structure to the pseudoknot and CR4/5 (TERT) and the intrinsic telomerase RNA (TER) compo- within vertebrate TERs. We then analyzed nent for enzymatic activity. The TER contains a short N. crassa and Schizosaccharomyces pombe tel- template that specifies the telomeric repeat sequence synthesized and conserved structural domains that serve omerase reconstituted in vitro, and showed that as binding sites for telomerase accessory proteins. The the two RNA core domains in both systems can TERT protein is highly conserved, containing four struc- reconstitute activity in trans as two separate tural domains: TEN, TRBD, RT and CTD. In compari- RNA fragments. Furthermore, the primer-extension son, TER is divergent in size and sequence, even among pulse-chase analysis affirmed that the reconstituted closely related clades. Over the past two decades, struc- N. crassa telomerase synthesizes TTAGGG repeats tural studies of TER from ciliates, vertebrates and yeasts with high processivity, a common attribute of verte- revealed two ubiquitous structural domains: the template– brate telomerase. Overall, this study reveals the proximal pseudoknot and a template–distal stem-loop common ancestral cores of vertebrate and fungal moiety [termed CR4/5 in vertebrates, three-way junction TERs, and provides insights into the molecular evo- (TWJ) in budding yeasts and stem-loop IV in ciliates] lution of fungal TER structure and function. (5–8). The template–adjacent pseudoknot structure

*To whom correspondence should be addressed. Tel: +1 480 965 3650; Fax: +1 480 965 2747; Email: [email protected]

ß The Author(s) 2012. Published by Oxford University Press. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by-nc/3.0/), which permits non-commercial reuse, distribution, and reproduction in any medium, provided the original work is properly cited. For commercial re-use, please contact [email protected]. 2 Nucleic Acids Research, 2012 forms a unique triple helix and is essential for telomerase ascomycetes from Pezizomycotina and Saitoella function (9–12). In the template-distal moiety, the verte- complicata from Taphrinomycotina. Phylogenetic com- brate CR4/5 domain contains a highly conserved 4-bp parative analysis combined with functional studies of P6.1 stem with an essential 5-nucleotide (nt) L6.1 loop these newly identiﬁed fungal TERs revealed structural that is not found in either the budding yeast TWJ or the features and biochemical attributes akin to vertebrate ciliate stem-loop IV (13,14). In ciliates and vertebrates, TER and not budding yeast TER. Moreover, the struc- these two structural domains bind independently to tural and functional conservation between the distantly TERT and are both essential for telomerase activity, an related vertebrate and non-yeast fungal TERs indicates attribute not yet demonstrated within yeasts (15,16). that these structural elements and functional attributes Fungal telomerase has been extensively studied are descended from a common ancestral TER. in budding yeasts (Saccharomycotina), such as Saccharomyces cerevisiae and Kluyveromyces lactis, and recently in the ﬁssion yeast Schizosaccharomyces pombe MATERIALS AND METHODS (Taphrinomycotina), both belonging to the Ascomycota

Ascomycete strains Downloaded from phylum. However, little is known of telomerase from Pezizomycotina (commonly known as filamentous asco- Wild-type N. crassa (strain FGSC 2489) and A. nidulans mycetes), which includes genetically tractable model (strain FGSC A4) were obtained from the Fungal organisms Neurospora crassa and Aspergillus nidulans Genetics Stock Center (FGSC). Vegetative growth, trans- (Figure 1). Pezizomycotina is the largest subphylum formation and mating of N. crassa were carried out following standard procedures (26). A. nidulans was grown as within Ascomycota, representing 90% of known asco- http://nar.oxfordjournals.org/ mycete species, while Taphrinomycotina is an early- previously described (19). Mycosphaerella graminicola branching subphylum, which includes S. pombe (17). (strain IPO323) was obtained from Dr Gert Kema and Many species from Pezizomycotina and the distantly grown in yeast glucose broth (1% yeast extract and 3% related Basidiomycota phyla have the vertebrate-type TT glucose) at 18 C. S. complicata (strain Y-17804) was AGGG telomeric repeat (18–23). While budding and obtained from the NRRL Collection and grown in YM fission yeasts are useful model systems for the study of broth (0.3% yeast extract, 0.3% malt extract, 0.5% telomere biology, their telomere repeats are longer than peptone and 1% glucose) at room temperature. those of filamentous ascomycetes and often irregular Recombinant N. crassa strains used and generated in (Figure 1). Moreover, telomerase from most yeasts are this study are listed in Supplementary Table S1. at Arizona State University West 2 on October 23, 2012 non-processive, whereas telomerase from most vertebrates are highly processive—capable of adding multiple repeats Cloning of telomeric DNA to a given primer before complete enzyme–product dis- Telomeric DNA was polymerase chain reaction (PCR) sociation. The processivity of non-yeast fungal telomerase amplified and cloned as previously described, with minor was not known before this study. modifications (27). Genomic DNA was isolated from We herein report the identification of novel TER 100 mg of homogenized N. crassa mycelia tissue using sequences from 74 fungal species, 73 filamentous the Wizard genomic DNA purification kit (Promega) following the manufacturer’s instructions. In a 20-ml reaction, 200 ng of N. crassa genomic DNA was incubated with 6 units of terminal deoxynucleotide transferase (TdT, USB) in TdT buffer [20 mM Tris–HCl (pH 7.8), 50 mM KOAc, 10 mM MgCl2 and 1 mM deoxycytidine triphosphat (dCTP)] at 37C for 30 min. After heat inacti- vation at 70C for 10 min, the poly(C)-extended genomic DNA was PCR amplified in 1Â PCR buffer [67 mM Tris– HCl (pH 8.8), 16 mM (NH4)2SO4, 5% glycerol, 0.01% Tween-20], 0.2 mM of each deoxynucleotide triphosphate (dNTP), 1 unit of Taq DNA polymerase (New England 0 Biolabs), 0.4 mM of oligo-dG primer 5 -CGGGATCCG18- 30 and a primer specific to the subtelomeric region of chromosome IIIF or VR. Nested PCR was performed Figure 1. Evolutionary relationships and telomere repeat sequence using 1 ml of 1/100Â of the aforementioned PCR mixture of major fungal subphyla. The evolutionary relationships of the Ascomycota subphyla (Pezizomycotina, Saccharomycotina and in 1Â HF buffer (Finnzymes), 0.2 mM of each dNTP, 1 Taphrinomycotina) and the Basidiomycota subphyla (Pucciniomycotina, unit of Phusion DNA polymerase (Finnzymes), 0.4 mMof Ustilaginomycotina and Agaricomycotina) are depicted based on oligo-dG primer and a nested subtelomeric primer. The recent phylogenomic studies of fungal species (17,24,25). Branch lengths PCR products were gel purified, cloned into the pZero are not proportional to evolutionary time. The names of subkingdom vector (Invitrogen) and sequenced. Dikarya (D), subphylum Ascomycota (A) and subphylum Basidiomycota (B) are indicated at nodes in the tree. Telomere repeat sequences of representative species from each subphylum are shown. The Cloning of N. crassa TERT (NcrTERT) vertebrate-type sequence TTAGGG is shown in red. The subphyla Pezizomycotina and Taphrinomycotina studied in this work are high- A hypothetical tert gene (GenBank accession no. lighted in blue. XM958854) was found in the N. crassa genome database Nucleic Acids Research, 2012 3

(Broad Institute) by Basic Local Alignment Search Tool Sephacryl S-500 gel filtration column (GE Healthcare) (BLAST) and characterized by 50 and 30 rapid amplifica- equilibrated with column buffer [10 mM Tris–HCl (pH tion of cDNA ends (RACE) using the FirstChoice 7.5), 400 mM NaCl, 5% glycerol]. Five-milliliter fractions RLM-RACE kit (Applied Biosystems). The sequence of were collected, and telomerase activity of each fraction the cloned Ncr-tert cDNA was deposited in GenBank was measured by telomeric repeat amplification protocol (Accession no. JF794773). (TRAP) assay. The peak fractions were pooled and applied to 20 ml of anti-FLAG M2 antibody agarose Generation of N. crassa recombinant strains beads (Sigma-Aldrich) pre-washed three times with 1Â Tris-buffered saline (TBS) buffer [10 mM Tris–HCl (pH The recombinant NcrTERT-3xFLAG N. crassa strain was generated by gene replacement. A 3.2-kb recombinant 7.4) and 150 mM NaCl]. The sample was incubated at 4 C DNA cassette was introduced into the Ku80-deficient for 1 h with gentle rotation, centrifugated at 1000g for 15 s, (Ámus-52) N. crassa strain by electroporation to replace and washed with 300 mlof1Â TBS buffer three times. the endogenous NcrTERT gene through homologous RNA was extracted from the beads using 100 ml of RNA extraction buffer [20 mM Tris–HCl (pH 7.5), 10 mM

recombination (Supplementary Table S1). The Downloaded from Ku80-deficient strain permits efficient homologous recom- ethylenediaminetetraacetic acid (EDTA) and 0.5% bination by eliminating non-homologous end joining (28). sodium dodecyl sulfate (SDS)], followed by acid phenol/ The DNA cassette contained 1 kb of the 30 portion of the chloroform extraction and ethanol precipitation. Ncr-tert genomic sequence, a 3xFLAG tag fused in-frame, Telomerase TRAP assay the hygromycin phosphotransferase (hph) drug-resistant gene and 500 bp of the 30 intergenic sequence The standard two-step TRAP assay was modified for de- http://nar.oxfordjournals.org/ (Supplementary Figure S1A). The Ncr-tert coding sequence tecting N. crassa telomerase activity. In the first reaction, and the 30 intergenic sequence were PCR amplified from 1 ml of lysate or immunoprecipitated enzyme was mixed genomic DNA, and the 3xFLAG-loxP-hph-loxP was with 0.5 mM of TS primer (50-AATCCGTCGAGCAGA PCR amplified from the plasmid p3xFLAG::hph::loxP GTT-30)in1Â primer extension (PE) buffer [50 mM (29), before overlapping PCR. Hygromycin-resistant Tris–HCl (pH 8.3), 2 mM dithiothreitol (DTT), 0.5 mM transformants were selected (29) and screened for the MgCl2 and 1 mM spermidine] containing 50 mMof recombinant NcrTERT-3xFLAG gene by Southern blot dATP, dTTP and dGTP, and incubated at 30C for 1 h.

analysis as previously described (26). Transformants The TS primer was purified by phenol/chloroform extrac- at Arizona State University West 2 on October 23, 2012 lacking the endogenous TERT gene were crossed with the tion and ethanol precipitation. The telomerase-extended wild-type strain, generating a homokaryotic NcrTERT- products were PCR amplified in a 25-ml reaction consist- 3xFLAG::hph+ strain expressing the FLAG-tagged ing of 1Â Taq buffer (NEB), 0.1 mM of dNTP, 0.4 mMof NcrTERT protein (Supplementary Table S1). 32P end-labeled TS primer, 0.4 mM of ACX primer (50-GC NcrTER template mutants were generated by site- GCGGCTTACCCTTACCCTTACCCTAACC-30), directed PCR mutagenesis and cloned into the 0.4 mM of NT primer (50-ATCGCTTCTCGGCCTTTT pCCG::C-Gly::3xFLAG vector (29) with BamHI and -30), 4 Â 10À13 M of TSNT primer (50-CAATCCGTCGA EcoRI sites. Mutant NcrTER plasmids were linearized GCAGAGTTAAAAGGCCGAGAAGCGATC-30) and 1 by StuI digestion, introduced into N. crassa strain NC1 unit of Taq DNA polymerase (NEB). Samples were ini- by electroporation and integrated into a non-functional tially denatured at 94C for 2 min, followed by 25–28 hisÀ locus. Selection of his+ transformants was performed cycles at 94C for 25 s, 50C for 25 s and 68C for 60 s, in Vogel’s minimum medium with 2% sorbose in place and a final extension at 68C for 5 min. PCR products of sucrose. were resolved on a non-denaturing 10% polyacrylamide/ 2% glycerol gel. The gel was dried, exposed to a phosphor Purification of N. crassa telomerase storage screen and analyzed with a Bio-Rad FX-Pro Nuclei were isolated from 2 g of N. crassa mycelium Molecular Imager. homogenized in a bead beater using five 10-s pulses at High-throughput sequencing and computational analysis 10-s intervals in ice-cold solution of 1 M sorbitol, 7% Ficoll (w/v), 20% glycerol (v/v), 5 mM MgCl2,10mM RNA extracted from purified N. crassa telomerase was CaCl2 and 1% Triton X-100. The cell slurry was used for cDNA library construction and sequenced centrifuged at 1500g for 10 min, and the supernatant using Illumina HiSeq2000 at Cofactor Genomics (St was then centrifuged at 15 000g for 20 min. The nuclear Louis, MO, USA). The 55-bp sequencing reads were pellet was washed once with wash buffer [50 mM Tris– analyzed by a computational pipeline including mapping HCl (pH 7.5), 5 mM MgCl2,10mMb-mercaptoethanol, the reads to the N. crassa reference genome and screening 20% glycerol and 0.6 M sucrose], resuspended and lysed in the assembled loci for putative template sequences. The CHAPS lysis buffer [10 mM Tris–HCl (pH 7.5), 400 mM sequencing reads were mapped with Segemehl (30), NaCl, 0.5% CHAPS, 5% glycerol, 1 mM ethylene glycol requiring a minimum accuracy rate of 85% and a tetraacetic acid (EGTA), 1 mM MgCl2, supplemented with maximum seed E-value of 10. Each read could maximally 5mM b-mercaptoethanol and 1Â protease inhibitor map to 100 distinct loci in the genome, with scoring cocktail (Roche)]. The nuclear extract was cleared by cen- only for maximal or multiple identical maximal hits. trifugation at 14 000g for 10 min. Two milliliters of Ambiguous reads were corrected by hit normalization, nuclear extract was immediately applied to an XK26/60 whereby each read is divided by the number of matching 4 Nucleic Acids Research, 2012 loci. Candidate loci with an average hits-per-base more Telomerase in vitro reconstitution and direct than five were subjected to the BLAST and open- primer-extension assay reading-frame (ORF) analyses. Recombinant NcrTERT protein was in vitro synthesized Multiple sequence alignment and phylogenetic in rabbit reticulocyte lysate (RRL) using the TnT Quick comparative analysis Coupled Transcription/Translation System (Promega) as previously described (37). In vitro transcribed full-length Filamentous ascomycete genome sequences obtained from or truncated fragments of NcrTER were gel purified and the DOE Joint Genome Institute and Broad Institute were assembled in RRL with the synthetic NcrTERT protein searched for NcrTER homologues by BLAST, with the at concentrations of 0.1 and 1 mM, respectively. The À3 E-value set between 1 Â 10 and 1. The TER sequences mixture was incubated at 30C for 30 min. The telomerase were aligned with ClustalW (31) and manually refined primer-extension assay was carried out with 2 mlof using BioEdit v. 7.1.3. Initial sequence alignment was per- in vitro-reconstituted telomerase in a 10-ml reaction in formed within each class and then expanded to include 1Â PE buffer [50 mM Tris–HCl (pH 8.3), 2 mM DTT, neighboring classes using highly conserved regions as 0.5 mM MgCl2 and 1 mM spermidine], 1 mM dTTP, Downloaded from anchor sites. Identification of nucleotide co-variations 1 mM dATP, 5 mM dGTP, 0.165 mM a-32P-dGTP and inference of a secondary structure model were per- (3000 Ci/mmol, 10 mCi/ml, PerkinElmer) and 1 mM telo- formed as previously described (5,32). The TER sequences meric primer (TTAGGG)3. The reactions were incubated identified from genome databases are available at the at 30C for 60 min and terminated by phenol/chloroform Telomerase Database (http://telomerase.asu.edu) (33). extraction, followed by ethanol precipitation. Telomerase- extended products were resolved on a denaturing 8 M http://nar.oxfordjournals.org/ Computational identification of S. complicata TER urea/10% polyacrylamide gel. The dried gel was exposed S. complicata TER (ScoTER) was identified by searching to a phosphor storage screen and analyzed with a Bio-Rad for putative template and sequences conserved in the 73 FX Pro Molecular Imager. filamentous ascomycete TERs. First, putative template se- The S. pombe telomerase was reconstituted in vitro and quences were identified in the S. complicata genome with analyzed as described previously in the text, with minor Fragrep (34). For each occurrence, an additional 1-kb modifications. A recombinant S. pombe TERT protein sequence upstream and 3-kb sequence downstream of gene was synthesized de novo, with codons optimized for the putative template were extracted. These 4-kb spans mammalian expression system (GenScript), and subcloned at Arizona State University West 2 on October 23, 2012 were then searched with Infernal 1.0 (35) for the TWJ into the pCITE-4a vector (Promega). The recombinant region structure P6/6.1 (Supplementary Figure S2B) and S. pombe TERT protein was synthesized in RRL and with HMMer for a conserved pattern close to the 30 end assembled with 1 mM full-length or truncated fragments of the fungal TER sequences. The query patterns were of S. pombe TER1 transcribed in vitro. The telomerase obtained as follows: A co-variance model (CM) was con- direct primer-extension assay was carried out with 2 mL structed and calibrated with Infernal starting from the of in vitro-reconstituted S. pombe telomerase in a 10 -ml sequence alignment and the manually annotated consen- reaction mixture containing 1Â telomerase reaction sus structure of the P6/6.1 region. In brief, CMs are gen- buffer [50 mM Tris–HCl (pH 8.0), 100 mM KCl, 1 mM eralizations of hidden Markov models (HMM) that MgCl2, 1 mM spermidine, 2 mM DTT], 0.2 mM dATP, evaluate not only sequence patterns but also complemen- dCTP and dTTP, 2 mM dGTP, 0.33 mM a-32P-dGTP tary base pairings consistent with the prescribed consensus (3000 Ci/mmol, 10 mCi/ml, PerkinElmer) and 5 mM structure. The Pezizomycotina TER alignment of the DNA primer 50-GTTACGGTTACAGGTTACG-30. highly conserved pseudoknot region was used to train an HMM using HMMer (36). A single sequence matching Telomerase Pulse-chase analysis both the P6/6.1 CM and the pseudoknot HMM was Telomerase pulse-chase analysis was carried out using identified. in vitro-reconstituted N. crassa telomerase and the telomerase direct primer-extension assay as described previously in Rapid amplification of cDNA ends (RACE) the text. Before the pulse reaction, in vitro-reconstituted The 50 and 30 ends of TERs from N. crassa, A. nidulans, telomerase was incubated with 2 mM telomeric primer M. graminicola and S. complicata were determined using (TTAGGG)3 in 1Â PE buffer at 20 C for 10 min. The the FirstChoice RLM-RACE kit following manufacture’s pulse reaction was initiated by the addition of an instructions (Applied Biosystems). For the 30 RACE, total equal-volume nucleotide mixture containing 1 mM dATP, RNA was pre-treated with yeast poly(A) polymerase. 1 mM dTTP, 2 mM dGTP and 0.33 mM a-32P-dGTP Eight mg of total RNA was incubated with 500 units of (3000 Ci/mmol, PerkinElmer) and incubated at 20C for yeast poly(A) polymerase (US Biological) in 1X reaction 20s. The chase reaction was initiated by the addition of buffer [20 mM Tris-HCl (pH 7.0), 0.6 mM MnCl2, non-radioactive dGTP to 100 mM and the competitive 0 0.02 mM EDTA, 0.2 mM DTT, 100 mg/ml acetylated primer (TTAGGG)3 with a 3 -blocking amine to 10 mM. BSA and 10% glycerol], 0.5 mM ATP and 1 unit/ml Aliquots of the chase reactions were removed at various SUPERase IN (Life Technology) at 37C for 15 min, time points (2, 4, 6 and 8 min) and terminated by phenol/ followed by phenol/chloroform extraction and ethanol chloroform extraction, followed by ethanol precipitation. precipitation. The PCR-amplified cDNAs were cloned The products were analyzed by polyacrylamide gel and sequenced. electrophoresis. Nucleic Acids Research, 2012 5

RESULTS N. crassa TER (NcrTER) mutant genes were integrated Identification of N. crassa TER into the his-3 locus for ectopic expression in the cells (Figure 3A). Telomeric DNA from the template mutant The filamentous ascomycete, N. crassa, is an important strains was cloned and sequenced (see ‘Materials and model organism with established genetics tools and Methods’ section). For each template mutant, 20% of a sequenced genome (26,38). We identified TER from the telomeric DNA clones contained one or more corres- N. crassa using a novel approach that combined a two- ponding mutant repeats. For example, the insertion of an step purification of telomerase, deep-sequencing of additional cytosine residue in the template resulted in the co-purified RNA and computational analysis of insertion of an additional guanine residue in the telomeric sequencing data (Figure 2A). For telomerase purification, DNA (Figure 3B). The mutant repeats were found at or we generated a recombinant N. crassa strain that ex- near the 30 end of the telomeric DNA, suggesting these pressed NcrTERT protein with a C-terminal 3xFLAG mutant repeats were newly synthesized. tag (Supplementary Figure S1A and S1B). Nuclear extract prepared from the NcrTERT-3xFLAG strain Secondary structure of NcrTER core domains Downloaded from was fractionated by gel filtration and assayed for telomer- To predict the secondary structure of NcrTER by phylo- ase activity by TRAP (Supplementary Figure S1C). genetic comparison analysis, we identified homologous Fractions with the highest telomerase activity were TER sequences from 72 Pezizomycotina species by combined, and the NcrTERT protein was immunopurified BLAST of available fungal genome databases using the with an anti-FLAG antibody (Supplementary Figure NcrTER sequence as query (Figure 4A and supplementary

S1D). RNAs co-purified with the immunopurified Table S2). In addition to Pezizomycotina, we identified the http://nar.oxfordjournals.org/ NcrTERT-3xFLAG protein were extracted and analyzed TER in S. complicata from the distantly related Taphrino- by high-throughput Illumina next-generation sequencing, mycotina subphylum through an advanced bioinformatics generating 15 million 55-bp reads. Nearly 90% of the approach that searches for both sequence and structural reads were successfully assembled onto the N. crassa ref- patterns (see ‘Materials and Methods’ section). The erence genome, and the assembled loci were screened for standard BLAST search failed to identify S. complicata putative template sequences. The locus with the highest TER from the genome owing to the great evolutionary average coverage was a precursor ribosomal RNA distance between Pezizomycotina and Taphrinomycotina. (Figure 2B). This was not unexpected because ribosomal The length of representative Pezizomycotina and at Arizona State University West 2 on October 23, 2012 RNA is the most abundant RNA in the cell. The second Taphrinomycotina TERs (A. nidulans, M. graminicola highest ranked locus, our candidate TER, was 1900 nt and S. complicata) were determined by RACE analysis long and lacked an apparent ORF. The exact 50 and 30 (Supplementary Table S2). Notably, the M. graminicola ends of this putative TER transcript were determined by TER with a size of 2425 nt was found to be the largest the RNA ligase-mediated RACE (RLM-RACE) assays, TER yet identified. which revealed the candidate TER to be 2049 nt in Secondary structures of two NcrTER core domains length, only slightly larger than the length of the locus were inferred by phylogenetic comparative analysis of covered by sequencing reads. The remaining loci were the newly identified Pezizomycotina TER sequences and hypothetical protein-coding genes with insignificant read supported by nucleotide co-variations (see ‘Materials and coverage (Figure 2B). Methods’ section). The two NcrTER core domains, the This 2049-nt N. crassa RNA was confirmed to be a template–pseudoknot and the TWJ, are highly conserved functional TER by expression of template–mutated across Pezizomycotina species and presumably homolo- RNA genes in vivo resulting in the synthesis of corres- gous to vertebrate and budding yeast TER structures ponding mutant telomeric DNA sequences. Three (Figure 4B). The NcrTER template–pseudoknot domain

Figure 2. Identification of NcrTER by a deep-sequencing approach. (A) The multi-step strategy for N. crassa telomerase purification and NcrTER identification. Telomerase was purified from nuclear extract by size-exclusion chromatography and anti-FLAG immunoprecipitation (IP). Telomerase activity was followed by TRAP assay through the purification steps. The RNA extracted from purified telomerase was sequenced using next-generation sequencing, followed by bioinformatics screening for putative template sequences. (B) The average read coverage, length, putative template sequence and gene annotation for the five top-ranking TER candidates. The candidates are ranked by the average read coverage. 6 Nucleic Acids Research, 2012

B Downloaded from http://nar.oxfordjournals.org/

Figure 3. Validation of the NcrTER gene. (A) Schematic for generating NcrTER template mutants. The parental N. crassa strain NC1 had a loss-of-function mutation (red cross) in the his-3 allele. Mutant NcrTER genes driven by the ccg-1 promoter were integrated into the genome at the his-3 locus via homologous recombination. The template sequences of the three NcrTER mutants are presented, with the inserted nucleotides shown in red. (B) Expression of the NcrTER template mutants promotes synthesis of mutant telomeric repeats. Sequences of mutant telomeric DNA repeats (blue) are shown, with the inserted nucleotides (red) highlighted. at Arizona State University West 2 on October 23, 2012 consists of the template, a template boundary element, a N. crassa telomerase requires both template–pseudoknot pseudoknot structure and two core-enclosing helices that and TWJ domains for activity bring the pseudoknot and template close to each other N. crassa telomerase reconstituted in vitro was analyzed by (Figure 4B). The pseudoknot structure is formed by two direct primer-extension assay and exhibited typical tel- helices, termed PK1 and PK2, with an additional helix omerase activity, adding multiple repeats to the primer 50- PK2.1 inserted in the loop between. From the 0 (TTAGGG)3-3 to generate a 6-nt ladder pattern with core-enclosing helix 2 to the pseudoknot, two additional major bands at positions +9, +15, +21, +27 nt etc. helices are predicted by mfold, but these base-paired struc- (Figure 5A, lane 2). This activity required both NcrTER tures lack strong nucleotide co-variation support. The and NcrTERT components, and was RNase sensitive NcrTER TWJ domain is highly conserved across (Figure 5A, lanes 3–5). From the same primer, the human Pezizomycotina species and shares striking similarity to telomerase generated a pattern with major bands at pos- the vertebrate CR4/CR5 domain (Figure 4B). Based itions +4,+10,+16,+22 nt etc. (Figure 5A, lane 1). The on the structural homology to vertebrate TER, we offset patterning arises from the NcrTER template 30-AAU named the three helices of N. crassa TWJ domain P5, CCCAAU-50 having a 50-boundary one nucleotide offset P6 and P6.1. Compared with the budding yeast TWJ, compared with the human TER template 30-CAAUCCC which usually contains long helices, the Pezizomycotina AAUC-50. This is the first fungal telomerase successfully TWJ contains short P6 and P6.1 stems, similar to the reconstituted in vitro from full-length synthetic TER. fish P6 and P6.1 stems, which are only 9 bp and 4 bp Vertebrate and yeast telomerases differ in the RNA long, respectively (39). The regions beyond these two domains required for enzymatic function. In vertebrate core domains are too variable in sequence to be properly telomerases, both the TER template–pseudoknot and the aligned outside individual classes and are thus not CR4/5 domains are essential for enzymatic activity, and included in the phylogenetic sequence analysis. can assemble in trans with the TERT protein (15,16,40). In Preliminary structure prediction in these variable regions contrast, S. cerevisiae telomerase can be reconstituted using the mfold software did not reveal any structures from a miniature TER lacking TWJ, indicating that the similar to the Est1 or Ku protein-binding sites reported yeast TWJ is dispensable for enzymatic activity (41). previously in budding yeast TERs. Examination of Unlike S. cerevisiae and more similar to vertebrates, the TER primary sequences also did not reveal any canon- N. crassa required both the TER template–pseudoknot ical Sm protein-binding site. However, direct experimen- and TWJ domains for enzymatic activity and these two tation is necessary to determine the presence or absence TER domains assembled in vitro with the TERT protein in of these accessory proteins in N. crassa telomerase trans as two separate RNA fragments (Supplementary holoenzyme. Figure S3). We identified an NcrTER minimal core Nucleic Acids Research, 2012 7

B Downloaded from http://nar.oxfordjournals.org/ at Arizona State University West 2 on October 23, 2012

Figure 4. Secondary structure model of the NcrTER core domains. (A) Number of fungal TER sequences identified. In this study, 73 new TER sequences were identified from six Pezizomycotina classes (red) and one Taphrinomycotina TER sequence from S. complicata. The number of TER sequences identified in each class is indicated. Branch length is not proportional to evolutionary distance. (B) Secondary structure model of the NcrTER core domains: the template–pseudoknot and the TWJ that includes the P5, P6 and P6.1 stems. Invariant nucleotides (red) or nucleotides with 80% identity (blue) as well as base pairings supported by co-variations (solid gray boxes) are indicated. Other major structural features shown include the core-enclosing helices 1 and 2, template boundary element, template and the pseudoknot helices PK1, PK2 and PK2.1. The regions without secondary structure determined are indicated by dotted lines, with the number of omitted residues indicated. essential for activity by serial truncations and functional G1856C (m3), into the minimal P6/6.1 NcrTER assay. Full activity can be reconstituted from a miniature fragment (nt 1826–1864) and assayed the reconstituted 295-nt template–pseudoknot core, named T-PK3, and a mutant enzymes for activity (Figure 5C). All three muta- 39-nt TWJ fragment of only the joined P6 and P6.1 tions dramatically reduced telomerase activity (Figure 5D, stem-loops without P5 (Figure 5B and 5C; lanes 3–6), suggesting that the N. crassa L6.1 is crucial Supplementary Figure S3C, lanes 6–9). This is similar to to telomerase activity and functionally homologous to the vertebrate CR4/5 domain, of which only P6 and P6.1 vertebrate L6.1. stem-loops, but not the P5 stem, are required for activity (42,43). Thus, N. crassa telomerase preserves the N. crassa telomerase is processive vertebrate-like functional attribute, requiring two To determine whether the N. crassa telomerase is truly distinct RNA domains for activity. processive, adding multiple repeats to the same primer The NcrTER P6.1 stem-loop, identical to vertebrate by the same enzyme, we performed a pulse-chase time P6.1, contains a conserved 4-bp stem and a 5-nt loop course assay with the N. crassa telomerase reconstituted 0 0 with the consensus sequence 5 -NUNGN-3 (Figure 4B in vitro. The pulse primer (TTAGGG)3 pre-bound to the and Supplementary Figure S2B). In vertebrate telomerase, telomerase enzyme was first ‘pulsed’ in the presence of the conserved uridine and guanosine residues in the loop radioactive a-32P-dGTP for 20 s and then ‘chased’ in the are crucial for telomerase function (13,44). To examine presence of 50-fold non-radioactive dGTP and 10-fold the importance of the NcrTER L6.1 loop for telomerase chase primer for 2, 4, 6 or 8 min. This assay detects activity, we introduced single mutations, U1854A (m1) processive elongation of only the initial ‘pulsed’ products and G1856C (m2), and a double mutation, U1854A/ by the same enzyme during the chase reaction. During the 8 Nucleic Acids Research, 2012

A BD

Δ Δ Downloaded from

C http://nar.oxfordjournals.org/

EF at Arizona State University West 2 on October 23, 2012

Figure 5. Functional characterization of N. crassa telomerase activity. (A) Direct primer-extension assay of in vitro-reconstituted N. crassa telomerase. Recombinant NcrTERT protein was synthesized in RRL and assembled with 0.1 mM full-length 2049-nt NcrTER. The reconstituted N. crassa telomerase was analyzed by direct primer-extension assay using the telomeric primer (TTAGGG)3. The major bands (+9, +15, +21 and +27) are denoted on the right of lane 2, indicating the number of nucleotides added to the primer. Addition of RNase A (lane 3), the absence of NcrTER (lane 4) and the absence of NcrTERT (lane 5) is indicated above the gel. Human telomerase (H) served as a positive control (lane 1). A 32P-end-labeled 15-mer oligonucleotide was included as loading control (l.c.) before puriﬁcation and precipitation of telomerase-extended products. (B) Secondary structure of the minimal 295-nt NcrTER template–pseudoknot fragment T-PK3. The T-PK3 fragment contains nucleotides 225–1515, with two internal deletions (nt 256–433 and 463–1288) replaced with tetraloops (bold, lower case), GAAA and GGAC, respectively. The template region is denoted by an open box. (C) The 39-nt NcrTER P6/6.1 RNA fragments with L6.1 mutations. Point mutations, U1854A (m1), G1856C (m2) and U1854A/G1856C (m3), in the L6.1 loop are indicated by solid circles (D). Mutations in loop L6.1 severely reduced telomerase activity. N. crassa (continued) Nucleic Acids Research, 2012 9 pulse reaction, 3–4 repeats were added to the primer at the +4 position, a pattern similar to the activity from (Figure 5E, lane 1). These ‘pulsed’ products were progres- in vivo-reconstituted telomerase previously reported by sively elongated during the chased reaction time course, Baumann group (46). We then tested whether the indicating processive repeat addition by the same enzyme S. pombe TWJ domain (nt 1029–1090), which included (Figure 5E, lanes 2–5). Therefore, we concluded that helices P5, P6 and P6.1, could reconstitute telomerase N. crassa telomerase is processive in repeat addition. activity in trans with the template–pseudoknot domain N. crassa telomerase appeared to have lower repeat (nt 83–957). Similar to N. crassa and vertebrate TERs, addition processivity than human telomerase, possibly S. pombe telomerase required both the template– from the shorter alignment region in the NcrTER pseudoknot and TWJ domains for reconstituting template (Figure 5A, lanes 1 and 2). In vertebrate telomer- wild-type level of activity (Figure 6B, lanes 6–8). ase, extending the alignment region in the template in- Therefore, the vertebrate-like P6/6.1 motif of the TWJ is creases repeat addition processivity (45). We thus crucial for telomerase activity and is conserved across ver- generated and analyzed two template-extended NcrTER tebrates, Pezizomycotina and Taphrinomycotina. mutants, NcrTER-t1 and -t2, to determine the effect of the extended alignment region on repeat addition Downloaded from processivity. As seen previously with human and mouse DISCUSSION telomerases (45), the repeat addition processivity of the TER is evolutionarily divergent in size, sequence and NcrTER template mutants was increased 2- to 3-fold, structure, posing an obstacle in discerning structural compared with the wild-type enzyme (Figure 5F). commonalities and evolutionary connections among or-

ganisms. In this study, we showed that fungal TERs http://nar.oxfordjournals.org/ S. pombe telomerase also requires two RNA from Pezizomycotina and the early-branching domains for activity Taphrinomycotina retain vertebrate-like structural In addition to Pezizomycotina TER, both S. pombe and features and functional attributes: a common RNA core S. complicata TERs also preserve the same pseudoknot that includes the vertebrate-like pseudoknot and P6/6.1 and TWJ core domains (Figure 6A). Specifically, the stem-loop, as well as the requirement of both RNA S. pombe and S. complicata TWJ structures contain the elements for enzymatic activity (Figure 7). In contrast, vertebrate-like 4-bp P6.1 stem and a 4- to 5-nt L6.1 loop budding yeast TER has diverged substantially from (Figure 6A, Right). To determine whether the S. pombe other ascomycete TERs, exhibiting structural variations at Arizona State University West 2 on October 23, 2012 TWJ domain is functionally homologous to vertebrate in both the pseudoknot and TWJ domains, requiring TWJ, we first established an in vitro reconstitution only the template–pseudoknot domain for enzymatic system for S. pombe telomerase core enzyme and later activity. examined whether a physically separate S. pombe TWJ The TER pseudoknot structure, although universally RNA fragment added in trans can reconstitute telomerase present, shows substantial variations in size and complex- activity. The S. pombe telomerase core enzyme was ity among ciliates, fungi and vertebrates (47,48). The TER reconstituted in RRL from the recombinant S. pombe pseudoknot is minimally defined by two quasi-continuous TERT protein and the synthetic full-length 1207-nt helices, PK1 and PK2, connected by two loops (Figure 7). S. pombe TER1 (see ‘Materials and Methods’ section), Within the vertebrate and non-yeast pseudoknot, the PK1 and analyzed for activity by primer-extension assay and PK1 helices are contiguous with no unpaired residues (Figure 6B). The activity reconstituted required S. pombe at the junction point. In contrast, budding yeast TER1 and was sensitive to RNase A treatment, suggesting pseudoknots contain several unpaired residues at the a ribonucleoprotein enzyme (Figure 6B, Left, lanes 1–3). helical junction. In the human TER pseudoknot, the The nucleotide addition reaction terminated at expected upstream loop is highly conserved, consisting of 4–5 in- positions when dideoxyadenosine triphosphate (ddATP) variant uridine residues that form a triple helix with the or dideoxythymidine triphosphate (ddTTP) was present, PK2 helix through base triple interactions (9). Similar base indicating a template-dependent reaction (Figure 6B, triples have also been reported in ciliate and budding yeast Right). The products extended by the in vitro- pseudoknots, suggesting a universal and crucial role of reconstituted S. pombe telomerase showed a strong band this unique triple-helix pseudoknot for telomerase

Figure 5. Continued telomerase was reconstituted in RRL from the recombinant NcrTERT protein and two separate RNA fragments, T-PK3 and P6/6.1, and were analyzed by direct primer-extension assay. The RNA fragments assayed within each reaction are denoted above the gel. Relative telomerase activity is shown under the gel (lane 4–6) and was determined by normalizing the total intensities of all bands within each lane to wild-type activity in lane 3. The relative activity was not determined (n.d.) for lanes 1 and 2. A 32P-end-labeled 15-mer oligonucleotide was included as the loading control (l.c.). (E) Pulse-chase time course analysis of N. crassa telomerase. In vitro-reconstituted telomerase was incubated with telomeric primer (TTAGGG)3 for 20 s at room temperature in the presence of radioactive a-32P-dGTP, dTTP and dATP. An aliquot of the reaction was removed and terminated after 20 s to determine the initial length of the product (lane 1). Chase reaction was initiated by adding 50 folds of non-radioactive dGTP and 10 folds of 0 0 competitive DNA oligonucletide 5 -(TTAGGG)3-3 -amine. Aliquots of the chase reaction was removed and terminated after 2, 4, 6 or 8 min from the start of the chase reaction (lane 2–5). A 15-mer 32P-end-labeled oligonucleotide was added as loading control (l.c.). (F) Effect of template length on repeat addition processivity. Full-length NcrTER (1–2049) with either the wild-type 9-nt template (wt), 10-nt template (t1) or 11-nt template (t2) was reconstituted with NcrTERT protein in RRL and analyzed by direct primer-extension assay. The NcrTER template sequence is denoted by an open box, with the alignment region shaded. Relative processivity was determined by normalization to wild-type processivity and is shown below the gel. A 15-mer 32P-end-labeled oligonucleotide was included as loading control (l.c.). 10 Nucleic Acids Research, 2012

A Downloaded from

B http://nar.oxfordjournals.org/ at Arizona State University West 2 on October 23, 2012

Figure 6. Reconstitution of S. pombe telomerase activity from two separate RNA fragments. (A) The secondary structure model of S. pombe TER1 core domains. The template–pseudoknot (T-PK) is composed of nt 83–957 and the TWJ is from nt 1029–1090. (B) Left, direct primer-extension assay of in vitro-reconstituted S. pombe telomerase. Telomerase was reconstituted in RRL from recombinant S. pombe TERT protein and 1 mM T7-transcribed TER1 RNA (Full-length: 1–1207; T-PK: 83–957 or TWJ: 1029–1090). The addition of ddATP and ddTTP in place of dATP and dTTP, respectively, in the reactions is indicated. RNase A (RNase) was added to the reaction in lane 2. A DNA size marker (M) was generated by labeling the 30 end of the DNA oligonucleotide (50-GTTACGGTTACAGGTTACG-30) with a-32P-dGTP using TdT. A 15-mer 32P-end-labeled oligonucleotide was included as loading control (l.c.). The expected sequence of nucleotides incorporated is shown on the right of the gel. Right, Alignment of the DNA primer sequence with the template sequence of S. pombe TER1. The previously determined template region is denoted by an open box (46). Nucleotides added to the DNA primer are shown in bold, with the radioactive 32P-dGTP underlined. function (11,12). The downstream loop is highly variable variable P5 stem appears to function merely to connect in sequence, often harboring large insertions that can po- the P6/P6.1 stem-loops to the remainder of the TER. In tentially form additional helical elements. In vertebrate the vertebrate telomerase ribonucleoprotein (RNP), the and budding yeast pseudoknots, no conserved helical P6/6.1 stem-loop binds tightly and specifically to the structures have been determined by phylogenetic compari- TRBD of TERT protein, potentially facilitating TERT son or mutagenesis analysis in this variable region. The folding (42). Although lacking a vertebrate-like L6.1 N. crassa pseudoknot is unique in that a conserved helix loop, the K. lactis TWJ contains an asymmetric U-rich PK2.1 was identified in the variable loop by phylogenetic internal loop that has been proposed to function similarly comparative analysis (Figure 4B). The functional signifi- to L6.1 (14). However, budding yeast telomerase has cance of the N. crassa helix PK2.1, however, remains to be evolved to not rely on the TWJ for enzymatic function determined. (10,41). The binding interface between the vertebrate P6/ The P6/6.1 stem-loop element in the TWJ is essential 6.1 element and the TRBD of vertebrate TERT has been for telomerase enzymatic activity and is highly conserved mapped using ultraviolet (UV) cross-linking analysis (42). in vertebrates, the filamentous fungus N. crassa and the fis- Analysis of the yeast TWJ–TERT binding interface would sion yeast S. pombe, but not in budding yeasts (Figure 7). help determine the extent of structural and functional In both vertebrates and N. crassa, the TERT protein rec- homology between vertebrate and yeast TWJ structures. ognizes minimally the two linked P6/6.1 stem-loops, rather Repeat addition processivity is a telomerase-specific than requiring the complete TWJ structure (42,43). The mechanistic attribute conserved widely in ciliates and Nucleic Acids Research, 2012 11 Downloaded from http://nar.oxfordjournals.org/ at Arizona State University West 2 on October 23, 2012

Figure 7. Conservation and diversiﬁcation of the common ancestral core of vertebrate and fungal TERs. Left, a minimal consensus structure of TER depicts the core domains common to both vertebrate and fungal TERs. Right, secondary structure models of TER pseudoknot and TWJ cores from human, N. crassa, S. cerevisiae, S. pombe and S. complicata are shown, with their evolutionary relationships depicted. Nucleotides shown in red indicate 80% identity within vertebrates, Saccharomyces and ﬁlamentous ascomycete, or identical between S. pombe and S. complicata. Sequences omitted are denoted as dashed lines. The helical regions in the pseudoknot structure and the highly conserved loop L6.1 in the TWJ domain are shaded gray. The predicted and experimentally determined base triples in the pseudoknot are indicated by gray and green dashed lines, respectively. The RNA core domains required for reconstituting telomerase activity in vitro are indicated for individual species.

vertebrates. In this study, we show that N. crassa telomer- Telomerases from specific vertebrate and fungal species ase is highly processive in vitro, synthesizing the vertebrate- appear to have lost repeat addition processivity during the type TTAGGG from a 1.5-repeat template in the TER course of evolution from alterations in either the template (Figure 5E). However, despite the significant repeat add- or the TERT motifs. For example, mouse telomerase is ition processivity observed with in vitro-reconstituted tel- non-processive owing to the shortened template, which omerase, the telomeric DNA cloned from N. crassa strains is unable to undergo efficient primer–template realignment expressing template mutants contains up to three consecu- during template translocation (45,51). Increasing the tive mutant repeats (Figure 3B). This limited addition TER template length significantly increases repeat add- of mutant repeats could result from telomere length regula- ition processivity as previously shown with mouse tel- tion restricting processive repeat addition in vivo or poten- omerase (45) and in this study with N. crassa telomerase tially deleterious effects of long mutant repeats to the cell (Figure 5F). Conversely, telomerases from many budding survival. While the detailed mechanism remains unknown, yeast models, such as S. cerevisiae, K. lactis and Candida telomerase repeat addition processivity requires highly albicans, synthesize irregular or long telomere repeats coordinated actions of numerous telomerase-specific non-processively (7,52,53). For example, K. lactis telomer- TERT motifs, such as motif 3, IFD, CTD and TEN ase synthesizes the TGGTGTACGGATTTGATTAGGT domain, and a realignment region in the template sufficient ATG repeat from a 30-nt-long template, which possibly for template translocation. Impairment of any of these hinder template translocation. Nonetheless, the budding elements has been shown to drastically reduce repeat yeast Saccharomyces castellii telomerase appears to have addition processivity (37,45,49,50). retained repeat addition processivity, suggesting the loss 12 Nucleic Acids Research, 2012 of processivity in S. cerevisiae telomerase is a relatively 6. Gunisova,S., Elboher,E., Nosek,J., Gorkovoy,V., Brown,Y., recent evolutionary event (53). Lucier,J.F., Laterreur,N., Wellinger,R.J., Tzfati,Y. and Tomaska,L. (2009) Identification and comparative analysis of The conservations of the vertebrate-like pseudoknot telomerase RNAs from Candida species reveal conservation of and an essential P6/6.1 element and processive synthesis functional elements. RNA, 15, 546–559. of TTAGGG repeats suggest that N. crassa telomerase 7. Tzfati,Y., Knight,Z., Roy,J. and Blackburn,E.H. (2003) A novel preserves more structural features and functional attri- pseudoknot element is essential for the action of a yeast telomerase. Genes Dev., 17, 1779–1788. butes of vertebrate telomerases, than budding yeast tel- 8. Lingner,J., Hendrick,L.L. and Cech,T.R. (1994) Telomerase omerases. While our understanding of telomerase and RNAs of different ciliates have a common secondary structure telomere biology has benefited greatly from yeast models, and a permuted template. Genes Dev., 8, 1984–1998. N. crassa is an exciting new model for the study of telomer- 9. Theimer,C.A., Blois,C.A. and Feigon,J. (2005) Structure of the ase function and telomere biology. Understanding the human telomerase RNA pseudoknot reveals conserved tertiary interactions essential for function. Mol. Cell, 17, 671–682. common mechanisms of telomerase action shared 10. Qiao,F. and Cech,T.R. (2008) Triple-helix structure in telomerase between distantly related organisms will potentially RNA contributes to catalysis. Nat. Struct. Mol. Biol., 15, 634–640. provide new insights into telomere-mediated diseases, 11. Shefer,K., Brown,Y., Gorkovoy,V., Nussbaum,T., Ulyanov,N.B. cancer and potential therapies. and Tzfati,Y. (2007) A triple helix within a pseudoknot is a Downloaded from conserved and essential element of telomerase RNA. Mol. Cell. Biol., 27, 2130–2143. 12. Ulyanov,N.B., Shefer,K., James,T.L. and Tzfati,Y. (2007) ACCESSION NUMBERS Pseudoknot structures with conserved base triples in telomerase RNAs of ciliates. Nucleic Acids Res., 35, 6150–6160. GenBank Accession number: JF794773, JQ793886, 13. Chen,J.-L., Opperman,K.K. and Greider,C.W. (2002) A critical JQ793887, JQ793888, JX173807. stem-loop structure in the CR4-CR5 domain of mammalian http://nar.oxfordjournals.org/ telomerase RNA. Nucleic Acids Res., 30, 592–597. 14. Brown,Y., Abraham,M., Pearl,S., Kabaha,M.M., Elboher,E. and SUPPLEMENTARY DATA Tzfati,Y. (2007) A critical three-way junction is conserved in budding yeast and vertebrate telomerase RNAs. Nucleic Acids Supplementary Data are available at NAR Online: Res., 35, 6280–6289. Supplementary Tables 1 and 2 and Supplementary 15. Mitchell,J.R. and Collins,K. (2000) Human telomerase activation Figures 1–3. requires two independent interactions between telomerase RNA and telomerase reverse transcriptase. Mol. Cell, 6, 361–371. 16. Mason,D.X., Goneska,E. and Greider,C.W. (2003) Stem-loop IV of tetrahymena telomerase RNA stimulates processivity in trans. at Arizona State University West 2 on October 23, 2012 ACKNOWLEDGEMENTS Mol. Cell. Biol., 23, 5606–5613. 17. Liu,Y., Leigh,J.W., Brinkmann,H., Cushion,M.T., Rodriguez- We are grateful to Dr Stephen Marek for providing the Ezpeleta,N., Philippe,H. and Lang,B.F. (2009) Phylogenomic genomic DNA of Phymatotrichum omnivorum (OKAlf8), analyses support the monophyly of Taphrinomycotina, and Dr Gert Kema for providing M. graminicola.We including Schizosaccharomyces fission yeasts. Mol. Biol. Evol., acknowledge the Fungal Genetic Stock Center and the 26, 27–34. 18. Schechtman,M.G. (1987) Isolation of telomere DNA from ARS Culture Collection (NRRL) for providing materials. Neurospora crassa. Mol. Cell. Biol., 7, 3168–3177. 19. Bhattacharyya,A. and Blackburn,E.H. (1997) Aspergillus nidulans maintains short telomeres throughout development. Nucleic Acids FUNDING Res., 25, 1426–1431. 20. Underwood,A.P., Louis,E.J., Borts,R.H., Stringer,J.R. and National Science Foundation (NSF) [CAREER award Wakefield,A.E. (1996) Pneumocystis carinii telomere repeats are MCB0642857 to J.J.-L.C.]; National Institutes of Health composed of TTAGGG and the subtelomeric sequence contains a [GM025690 to E.U.S.]; DFG [MA5082/1-1 to M.M., gene encoding the major surface glycoprotein. Mol. Microbiol., STA-850/7-2 to P.F.S.]. Funding for open access charge: 19, 273–281. 21. Edman,J.C. (1992) Isolation of telomerelike sequences from NSF. Cryptococcus neoformans and their use in high-efficiency transformation. Mol. Cell. Biol., 12, 2777–2783. Conflict of interest statement. None declared. 22. Aimi,T., Kano,S., Iwasaki,Y. and Morinaga,T. (2003) Telomeric fingerprinting of the violet root rot fungus, Helicobasidium mompa: a useful tool for karyotype estimation. Mycol. Res., 107, REFERENCES 1055–1059. 23. Perez,G., Pangilinan,J., Pisabarro,A.G. and Ramirez,L. (2009) 1. de Lange,T. (2009) How telomeres solve the end-protection Telomere organization in the ligninolytic basidiomycete Pleurotus problem. Science, 326, 948–952. ostreatus. Appl. Environ. Microbiol., 75, 1427–1436. 2. Armanios,M. (2009) Syndromes of telomere shortening. Annu. 24. Hibbett,D.S., Binder,M., Bischoff,J.F., Blackwell,M., Rev. Genomics Hum. Genet., 10, 45–61. Cannon,P.F., Eriksson,O.E., Huhndorf,S., James,T., Kirk,P.M., 3. Alder,J.K., Cogan,J.D., Brown,A.F., Anderson,C.J., Lawson,W.E., Lucking,R. et al. (2007) A higher-level phylogenetic classification Lansdorp,P.M., Phillips,J.A. 3rd, Loyd,J.E., Chen,J.J.-L. and of the Fungi. Mycol. Res., 111, 509–547. Armanios,M. (2011) Ancestral mutation in telomerase causes defects 25. Hedges,S.B. (2002) The origin and evolution of model organisms. in repeat addition processivity and manifests as familial pulmonary Nat. Rev. Genet., 3, 838–849. fibrosis. PLoS Genet., 7, e1001352. 26. Davis,R.H. (2000) Neurospora: Contributions of a Model 4. Tsang,A.R., Wyatt,H.D., Ting,N.S. and Beattie,T.L. (2012) Organism. Oxford University Press, Oxford, New York. hTERT mutations associated with idiopathic pulmonary fibrosis 27. Forstemann,K., Hoss,M. and Lingner,J. (2000) affect telomerase activity, telomere length, and cell growth by Telomerase-dependent repeat divergence at the 3’ ends of yeast distinct mechanisms. Aging cell, 11, 482–490. telomeres. Nucleic Acids Res., 28, 2690–2694. 5. Chen,J.-L., Blasco,M.A. and Greider,C.W. (2000) Secondary 28. Ninomiya,Y., Suzuki,K., Ishii,C. and Inoue,H. (2004) Highly structure of vertebrate telomerase RNA. Cell, 100, 503–514. efficient gene replacements in Neurospora strains deficient for Nucleic Acids Research, 2012 13

nonhomologous end-joining. Proc. Natl Acad. Sci. USA, 101, telomerase ribonucleoprotein. Proc. Natl Acad. Sci. USA, 108, 12248–12253. 20333–20338. 29. Honda,S. and Selker,E.U. (2009) Tools for fungal proteomics: 43. Kim,N.K., Theimer,C.A., Mitchell,J.R., Collins,K. and Feigon,J. multifunctional Neurospora vectors for gene replacement, protein (2010) Effect of pseudouridylation on the structure and activity of expression and protein purification. Genetics, 182, 11–23. the catalytically essential P6.1 hairpin in human telomerase RNA. 30. Hoffmann,S., Otto,C., Kurtz,S., Sharma,C.M., Khaitovich,P., Nucleic Acids Res., 38, 6746–6756. Vogel,J., Stadler,P.F. and Hackermuller,J. (2009) Fast mapping of 44. Robart,A.R. and Collins,K. (2010) Investigation of human short sequences with mismatches, insertions and deletions using telomerase holoenzyme assembly, activity, and processivity index structures. PLoS Comput. Biol., 5, e1000502. using disease-linked subunit variants. J. Biol. Chem., 285, 31. Larkin,M.A., Blackshields,G., Brown,N.P., Chenna,R., 4375–4386. McGettigan,P.A., McWilliam,H., Valentin,F., Wallace,I.M., 45. Chen,J.-L. and Greider,C.W. (2003) Determinants in mammalian Wilm,A., Lopez,R. et al. (2007) Clustal W and Clustal X version telomerase RNA that mediate enzyme processivity and 2.0. Bioinformatics, 23, 2947–2948. cross-species incompatibility. EMBO J., 22, 304–314. 32. Pace,N.R., Smith,D.K., Olsen,G.J. and James,B.D. (1989) 46. Box,J.A., Bunch,J.T., Zappulla,D.C., Glynn,E.F. and Baumann,P. Phylogenetic comparative analysis and the secondary structure of (2008) A flexible template boundary element in the RNA subunit ribonuclease P RNA—a review. Gene, 82, 65–75. of fission yeast telomerase. J. Biol. Chem., 283, 24224–24233. 33. Podlevsky,J.D., Bley,C.J., Omana,R.V., Qi,X. and Chen,J.J.-L. 47. Lin,J., Ly,H., Hussain,A., Abraham,M., Pearl,S., Tzfati,Y., (2008) The telomerase database. Nucleic Acids Res., 36, D339–D343. Parslow,T.G. and Blackburn,E.H. (2004) A universal telomerase Downloaded from 34. Mosig,A., Chen,J.J.-L. and Stadler,P.F. (2007) Homology search RNA core structure includes structured motifs required for with fragmented nucleic acid sequence patterns. Lect. Notes binding the telomerase reverse transcriptase protein. Proc. Natl Bioinformatics, 4645, 335–345. Acad. Sci. USA, 101, 14713–14718. 35. Nawrocki,E.P., Kolbe,D.L. and Eddy,S.R. (2009) Infernal 1.0: 48. Chen,J.-L. and Greider,C.W. (2004) An emerging consensus for inference of RNA alignments. Bioinformatics, 25, 1335–1337. telomerase RNA structure. Proc. Natl Acad. Sci. USA, 101, 36. Eddy,S.R. (2011) Accelerated profile HMM searches. PLoS

14683–14684. http://nar.oxfordjournals.org/ Comput. Biol., 7, e1002195. 49. Lue,N.F., Lin,Y.-C. and Mian,I.S. (2003) A conserved telomerase 37. Xie,M., Podlevsky,J.D., Qi,X., Bley,C.J. and Chen,J.J.-L. (2010) motif within the catalytic domain of telomerase reverse A novel motif in telomerase reverse transcriptase regulates transcriptase is specifically required for repeat addition telomere repeat addition rate and processivity. Nucleic Acids Res., processivity. Mol. Cell. Biol., 23, 8440–8449. 38, 1982–1996. 50. Zaug,A.J., Podell,E.R. and Cech,T.R. (2008) Mutation in TERT 38. Selker,E.U. (2011) Neurospora. Curr. Biol, 21, R139–R140. 39. Xie,M., Mosig,A., Qi,X., Li,Y., Stadler,P.F. and Chen,J.J.-L. (2008) separates processivity from anchor-site function. Nat. Struct. Structure and function of the smallest vertebrate telomerase RNA Mol. Biol., 15, 870–872. from teleost fish. J. Biol. Chem., 283, 2049–2059. 51. Prowse,K.R., Avilion,A.A. and Greider,C.W. (1993) Identification 40. Lai,C.K., Miller,M.C. and Collins,K. (2003) Roles for RNA in of a nonprocessive telomerase activity from mouse cells. Proc. Natl Acad. Sci. USA, 90, 1493–1497. telomerase nucleotide and repeat addition processivity. Mol. Cell, at Arizona State University West 2 on October 23, 2012 11, 1673–1683. 52. Hsu,M., Yu,E.Y., Singh,S.M. and Lue,N.F. (2007) Mutual 41. Zappulla,D.C., Goodrich,K. and Cech,T.R. (2005) A miniature dependence of Candida albicans Est1p and Est3p in telomerase yeast telomerase RNA functions in vivo and reconstitutes activity assembly and activation. Eukaryot. Cell, 6, 1330–1338. in vitro. Nat. Struct. Mol. Biol., 12, 1072–1077. 53. Cohn,M. and Blackburn,E.H. (1995) Telomerase in yeast. Science, 42. Bley,C.J., Qi,X., Rand,D.P., Borges,C.R., Nelson,R.W. and 269, 396–400. Chen,J.J.-L. (2011) RNA-protein binding interface in the Qi_Supplementary Fig. S1

Xho I Xho I A Chromosome 1 Ncr-tert 1.4kb

F Ncr-tert-3xFLAG Ncr-tert-3'UTR Ptrpc hph R recombinant 3xFLAG cassette 3.2kb Xho I Xho I

Hygromycin-Resistant Clones B kb M wt 1 2 3 4 5 6 71098 11 12 4 3 NcrTERT-3xFLAG (3.2kb) 2 1.6

NcrTERT (1.4kb) 1 Xho-I digested DNA

C 400 D

300 partially-purified telomerase

FLAG-IP 200

100 input sup. pellet Absorbance 280nM (mAu)

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 Fraction# 0 50 100 150 (ml)

input 10 11 12 13 14 15 16 17 18 19 20

Internal control 1 2 3 Internal control Qi_Supplementary Fig. S2A

Neurospora crassa AGAGCACGGCGCCUCCCCCACAGUUCUCCGA178UCGGA-ACUUAACCCUAA---GUAUACA-CUGGU826GCCAG108GCACGG--UUU Neurospora discreta AGAGCACGGCGCCUCCCCCACAGUUCUCCGA174UCGGA-ACUUAACCCUAA---GUAUACA-CUGGU821GCCAG105GCACGG--UUU Neurospora tetrasperma AGAGCACGGCGCCUCCCCCACAGUUCUCCGA178UCGGA-ACUUAACCCUAA---GUAUACA-CUGGU825GCCAG108GCACGG--UUU Sordaria macrospora -GAGCACGACGCCUCCCCUACAGUUCUCCGA174UCGGA-ACUUAACCCUA---CGUAUACG-CUGGU830GCCAG109GCACGG--UUU Podospora anserina --CGCACGGCGCCUCCUCCACAGUUCUCCGA163UCGGA-ACUUAACCCUAA---GUUUUUAUUUGGU679GCCAA-71GCGCGG--UUU Colletotrichum graminicola --CGCACGACACCUCCCGUUCAGUUCUCCGA169UCGGA-ACUUAACCCUAA---GCAGCCUUUUGGU712GCCAA-71GCGCGG--UUU Colletotrichum higginsianum --CGCACGACACCUCCCCUUCAGUUCUCCGA169UCGGA-ACUUAACCCUAA---GCAGCCCUUUGGU722GCCAA-68GCGCGG--UUU Fusarium graminearum --CGCGUGACGCCUCCCCCACAGU---CCGC159UCGGA-ACUUAACCCUAA---GAAGAUAUCUGGU676GCCAG-70GGGUGG--UUU Fusarium verticillioides --CGCGUGACGCCUCCCCCACAGUCCUCCGG164UCGGA-ACUUAACCCUAA---GAAGAUAUCUGGU795GCCAG-69GAGUGG--UUU Fusarium oxysporum --CGCGCGACGCCUCCCCCACAGUCCUCCGG164UCGGA-ACUUAACCCUAA---GAAGAUAUCUGGU702GCCAG-69GGGUGG--UUU Metarhizium acridum -CUUCGCGUCGCCUCCCGCACAGUUUACCGA144UCGGAAACUUAACCCUAA---GAUAAGAUCUGGU642GCCAG-74GCGCGG--UUU Metarhizium anisopliae -CUUCGCGUCGCCUCCCACACAGUUUACCGA-55UCGGA-ACUUAACCCUAA---GAUAAGAUCUGG-650-CCAG-74GCGCGG--UUU Verticillium dahliae --CGCACGACGCCUCCUCCACAGUCCUCCGG169UCGGA-ACUUAACCCUAA---GCAUCACCUUGGU711GCCAA-70GCGCGG--UUU S Verticillium albo-atrum --CGCACGACGCCUCCUUCACAGUCCUCCGG169UCGGA-ACUUAACCCUAA---GCAUCACCUUGGU712GCCAA-69GCACGG--UUU Trichoderma atroviride ---GCACGUCGCCUCCCCCACAGUUUCCCGA173UCGGA-ACUUAACCCUAA---GCAUACAUCUGGU722GCCAG-64ACGCGG--UUU Trichoderma reesei --UGCACGUCGCCUCCCCUACAGUUUCCCGA173UCGGA-ACUUAACCCUAA---GCAUACAUCUGGU717GCCAG-64ACGCGU--UUU Trichoderma virens --UGCACGUCGCCUCCCCCACAGUUUCCCGA174UCGGA-ACUUAACCCUAA---GCAUACAUCUGGU677GCCAG-62ACGCGG-CUUU Acremonium alcalophilum --CGCGCGACGCCUCCACCACAGUUCUCCGA169UCGGA-ACUUAACCCUAA---GCUUGCUAUUGGU715GCCAA-76GCACGG--UUU Nectria haematococca --CGCGCGACGCCUCCCCCACAGUUCUCCGA165UCGGA-ACUUAACCCUAA---GAAGAUAUCUGGU699GCCGG-91GGGUGG--UUU Chaetomium globosum --CGCACCACGCCUUCCCCACAGUUCUCCGA173UCGGA-ACUUAACCCUAA---AUACACAUUUGGU706GCCAA-77GCACGG--UUU Sporotrichum thermophile --CGCACCGCGCCUCCCCCACAGUUCUCCGA178UCGGA-ACUUAACCCUAA---GCAGAUCUUUGGU731GCCAA-78GCGCGG--UUU Cryphonectria parasitica --CGCACGACGCCUCCCCCACGGUGCUCCGA177UCGGA-ACUUAACCCUA-GCCCUAAAUGCCGGGU700ACUCG-76GCGCAA--UUU Thielavia terrestris --CGCACGGCGCCUCCCCCACAGUUCUCCGA171UCGGA-ACUUAACCCUAA---GUAUACCUUUGGU702ACCAA-88GCGCGG--UUU Magnaporthe oryzae --UGCACCUCGCCUCCCCCACAGUCUACCGU186UCGGA-ACUUAACCCUAA---GUUCUAUCGUGGU803GCCAC-40UCGGAG--UUU Magnaporthe poae --UGCGCUUCGCCUCCCCUACAGUUAUCCGU202UCGGA-ACUUAACCCUAA---GUUCGACCGUGG-873-CCAC152--GGGGACUUU Gaeumannomyces graminis ---GCGCCACGCCUCCCCUACAGUCAUCCGU193UCGGA-ACUUAACCCUAA---GUUUGAUCGUGG-798-CCAC136--GGGGACUUU Epichloe festucae -CUUCGCGUCGCCUGCCUCACAGUUUCCCGA172UCGGA-ACUUAACCCUAA---GAUAGAAAUUGGU716GCCAG-74GCGUGG-CUUU Geomyces destructans --AGCAACGCGCCUUCUCCACAGUCAUCCGG176UUGGA-ACUUAACCCUAA---GUAA-UCGUUGGU932GCCAA-72--ACGG--UUU Leo Botrytis cinerea AAGUAUA-UCGUCUACCCUACAG-ACUCUGG-28UCAGA-ACUUAACCCUAA---GAAUACUGCCGGU859GCCGG-67--ACGG-CUUU Sclerotinia sclerotiorum AAGUAUA-UCGUCUACCCCACAG-ACUCUGG-28UCAGA-ACUUAACCCUAA---AUAAAUCGCCGGU858GCUGG-61--ACGG-CUUU Saitoella complicata ------GCAGUUUCCCGG125CCGGA-ACUUAACCCUAA----GUGUUCCAUCGA734UCGAU-89AGACGG--UUU Structure ((((((( ((( ((( (((( )))) )))UAACCCUAA ((((())))) {{{{{{ Template Core enclosing Template boundary Core enclosing pseudoknot helix 1 helix helix 2 helix 1 (PK1)

Neurospora crassa UU--GAGGACGCAU-UUUGUCCGUGCGGUAAU--GCCCUCC30GGAGAGG--AAA-CAAAAGCUAGUCUCAGUUGGCAGUGUUCU Neurospora discreta UU--GAGGACGCAU-UUUGUCCGUGCGGUAUU--GCCCUCC29GGAGAGG--AAA-CAAAAGCUAGUCUCAGUUGGCAGUGUUCU Neurospora tetrasperma UU--GAGGACGCAU-UUUGUCCGUGCGGUAAU--GCCCUCC30GGAGAGG--AAA-CAAAAGCUAGUCUCAGUUGGCAGUGUUCU Sordaria macrospora UU--GAGGACGCAU-UUUGUCCGUGCGGAAAU--GCCCUCC30GGAGAGG--AUA-CAAAAGCUAGUCUCAGUUGGCCGUGCUC- Podospora anserina UU--AAGGAUGUCU-UUUGUCCGCGCGUUAAU--GCCCUCC26GGAGAGG--AAU-CAAAAACUAGUCCCUGUUGGCAGUGCG-- Colletotrichum graminicola UUU-GUGGAUGUCU-UUUGUCCGCGCGG-AAU--GCCCUCC17GGAGAGG--AAA-CAAAAACUAGUCCAAGGAGGUCGUGCG-- Colletotrichum higginsianum UUU-AUGGAUGUCU-UUUGUCCGCGCGG-AAU--GCCCUCC16GGAGAGG--AAA-CAAAAACCAGUCCAAGGAGGUCGUGCG-- Fusarium graminearum UU--AAGGAUGUAU-UUUGUCCGCCCGUUAAU--GCCCUCC24GGAGAGG--AAU-CAAAAACUUGUCCCAGGGGGCAGCGCG-- Fusarium verticillioides UU--AUGGAUGUCU-UUUGUCCACCCGUCAAA--GCCCUCC24GGAGAGG--AAU-CAAAAACUAGUCCUAGUUGGCAGCGCG-- Fusarium oxysporum UU--AUGGAUGUCU-UUUGUCCACCCGUCAAC--GCCCUCC24GGAGAGG--AAU-CAAAAACUAGUCCUAGUUGGCAGCGCG-- Metarhizium acridum UU--GUGGAUGUCU-UUUGUCCGCGCGUCGAU--GCCCUCC24GGAGAGG--AUA-CAAAAACUAGUCCUUGUUGGCAGCGAAG- Metarhizium anisopliae UU--AUGGAUGUCU-UUUGUCCGCGCGUCGAU--GCCCUCC24GGAGAGG--AUA-CAAAAACUAGUCCUUGUUGGCAGCGAAG- Verticillium dahliae UU--GAGAAUGUCU-UUUGUCCGCGCG-CAAU--GCCCUCC24GGAGAGG--AAA-CAAAAACGAGUCCUUGUUGGCAGUGCG-- S Verticillium albo-atrum UU--GAGAAUGUCU-UUUGUCCGCGCG-CAAU--GCCCUCC24GGAGAGG--AAA-CAAAAACGAGUCCUUGUUGGCAGUGCG-- Trichoderma atroviride UU--AUGGAUGUAU-UUUGUCCGCGUGUCAAC--GCCCUCC16GGAGAGG--AAA-CAAAAACAAGUCUCCGUUGGCAGUGC--- Trichoderma reesei UU--GCGGAUGUGU-UUUGUGCGCGUGUAAAG--GCCCUCC16GGAGAGG--AAA-CAAAAACAAGUCCUAGUUGGCAGUGCG-- Trichoderma virens U---ACGGAUGUGU-UUUGUGCGCGUGUUGAU--GCCCUCC16GGAGAGG--AAA-CAAAAACAAGUUCACGUUGGCAGUGCG-- Acremonium alcalophilum UUU--CGGAUGU-U-UUUGAUCGUGCGC-GAC--GCCCUCC23GGAGAGG--AAACCAAAAACAA-UCCUUGUUGGCAGCGCG-- Nectria haematococca UU--AUGGAUGUCU-UUUGUCCACCCGUUCAC--GUCCUCC19GGAGAGA--UUA-CAAAAACUGGUCCUUGUUGGCAGUGCG-- Chaetomium globosum UUUUGAGGAUGUCU-UUUGUCCGUGCGGUACCC-GCCCUUC35GAAGAGG--AAA-CAAAAACUAGUCCCUGUUGGCAGUGCG-- Sporotrichum thermophile UUU-GAGGAUGUAU-UUUGUCCGUGCGGUUACCUGCCCUCC40GGAGAGG--AAA-CAAAAACUAGUCCCCGUUGGCAGUGCG-- Cryphonectria parasitica UU--GAGAAUGUAU-UUUGUUUGUGCGAUUUU--GCCCUCC22GGAGAGG--AUA-CAAAAACUAGUCCUAGUUGGCAGUGCG-- Thielavia terrestris UU--GAGGACGUCU-UUUGUCCGUGCGGUACC--GCCCUCC33GGAGAGG--AAU-CAAAAACUAGUCCCAGUUGGCAGUGCG-- Magnaporthe oryzae UU--GUGGAUGU-U-UUU-U-UCCGG-UUUGU------22------AAACGAAAAACUAGUCCCUGUUGGCAGUGCA-- Magnaporthe poae UU--GUGGAUGUAU-UU-CACCCC--C------GUCC17GGAC---CCAAACGAAAAACUAGUCCUUGUUGGC-G-GCA-- Gaeumannomyces graminis UU--GUGGAUGUAU-UU-CACCCC--C------GUCC35GGAC---CCAAACGAAAAACUAGUCCCUGUUGGCAGCGC--- Epichloe festucae U---AAGGAUGUGU-UUUGUCCACGCGUCAAC--GCCCUCC26GGAGAGG--AAA-CAAAAACGAGUCCUAGUGGGUAGCGAAG- Geomyces destructans UU--GUUGAUGUCU-UUUGUCCGU----UUGU--GCCC-GC13GC-GAGG--AAUCGAAAAACUUAUCUCUGAGGGCAUUGCU-- Leo Botrytis cinerea U---GUUGAUGUCU-UUUGUCCGU----UUGU--GCCCUGC-6GCAGAGG--AAU-CAAAAACUUAUCUUCGUGGAC-UAUACUU Sclerotinia sclerotiorum U---GUUGAUGUCU-UUUGUCCGU----UUGU--GCCCUGC-4GCAGAGG--AAU-CAAAAACUUAUCUCCGUGGAC-UAUACUU Saitoella complicata UU--GCGGAUGU-CGUUUGUCCGUUUGUCCAC------10------AAA-CAAAGUCUAGUCCCUGAC------Structure <<<<< < <<<<<}}}}}} (((((()))) )) > >>>>>>> >>> ))) ))))))) pseudoknot pseudoknot pseudoknot pseudoknot Core enclosing helix 2 (PK2) helix 1 (PK1) helix 2.1 (PK2.1) helix 2 (PK2) helix 1 Qi_Supplementary Fig. S2B

Neurospora crassa GGACCCA--UCUUGAGU--UGG-AG-UCAAAGUUUCGUACUUGUC Neurospora discreta GGACCCA--UCUCGAGU--UGG-AG-UCAAAGUUUCGUACUUGUC Neurospora tetrasperma GGACCCA--UCUUGAGU--UGG-AG-UCAAAGUUUCGUACUUGUC Sordaria macrospora GGACCCA--UCUCGAGU--UGG-AG-UCAAAGUUUCGUACUUGUC Podospora anserina GGACCCA--UCUAGAGU--UGG-AG-UCAAAGUUUCGUACUUGUC Colletotrichum graminicola GGACCCA--UCUCGUGU--UGG-AG-UCAAAGUUUUGUGCUUGUC Colletotrichum higginsianum GGACCCA--UCUCGUGU--UGG-AG-UCAAAGUUUUGUGCUUGUC Fusarium graminearum GGACCCA--UCUUGAGU--UGG-AG-UCAAAGUUUCGUGCUUGUC Fusarium verticillioides GGACCCA--UCUCGAGU--UGG-AG-UCAAAGUUUCGUACUUGUC Fusarium oxysporum GGACCCA--UCUCGAGU--UGG-AG-UCAAAGUUUCGUGCUUGUC Metarhizium acridum GGACCCA--UCCCGAGU--UGG-AG-UCAAAGUUUCGUGCUUGUC Metarhizium anisopliae GGACCCA--UCCUGAGU--UGG-AG-UCAAAGUUUCGUACUUGUC Verticillium dahliae GGACCCA--UCUUGAGU--UGG-UG-UCAAAGUUCCGCGCUUGUC S Verticillium albo-atrum GGACCCA--UCUUGAGU--UGG-UG-UCAAAGCUCCGCGCUUGUC Trichoderma atroviride GGACCCA--UCUCGAGU--UGG-AG-UCAAAGUUUAGUGCUUGUC Trichoderma reesei GGACCCA--UCUCGAGU--UGG-AG-UCAAAGUUCAGUACUUGUC Trichoderma virens GGACCCA--UCUUGAGU--UGG-UG-UCAAAGUUUAGCACUUGUC Acremonium alcalophilum GGACCCA--UCUAGAGU--UGG-AG-UCAAAGCUUCGUGCUUGUC Nectria haematococca GGACCCA--UCUCGAGU--UGG-AG-UCAAAGUUUCGUGCUUGUC Chaetomium globosum GGACCCA--UCUUGAGU--UGG-AG-UCAAAGCUUCGUGCUUGUC Sporotrichum thermophile GGACCCA--UCUCGAGU--UGG-AG-UCAAAGCUUCGUGCUUGUC Cryphonectria parasitica GGACCCA--UCUCGAGU--UGG-AG-UCAAAGUUUUGUACUUGUC Thielavia terrestris GGACCCA--UCUCGAGU--UGG-AG-UCAAAGCUUUGUGCUUGUC Magnaporthe oryzae GGACCCA--UCUCGAGU--UGG-AG-UCAAAGUCUUGUACUUGUC Magnaporthe poae GGACCCA--UCUAGAGU--UGG-AG-UCAAAGCUUCGCGCUUGUC Gaeumannomyces graminis GGACCCA--UCUAGAGU--UGG-AG-UCAAAGCUUCGCGCUUGUC Epichloe festucae GGACCCA--UCUCGUGU--UGG-AG-UCAGAGUUUCGUACUCGUC Geomyces destructans GGACCCA--UCUCGAGU--UGG-AG-UCAAGGCUUAGUGCCUGUC Leo Botrytis cinerea GGACCCA--UCUCGUGU--UGG-UG-UCAAAGCUUAGUGCUUGUC Sclerotinia sclerotiorum GGACCCA--UCUCGUGU--UGG-UG-UCAAAGCUUAGUGCUUGUC Lec Xanthoria parietina GGGCCUC--UCUCGUGU--GAG--GAUCAAAGUUUCGUGCUUGUC Cladonia grayi GGACCUC--UCUAGAGU--GAG--GAUCAAAGUUUAG-GCUUGUC Aspergillus nidulans GGACCUC--UCUCGUGU--GAG--GCUCAAAGUUUAG-GCUUGUC Neosartorya fischeri GGACCUCAC---AAUG-GUGGG--GAUCAAAGCUUAG-GCUUGUC Aspergillus fumigatus GGACCUCAC---AAUG-GUGGG--GAUCAAAGCUCAG-GCUUGUC Aspergillus clavatus GGACCUCAC--GAAGG-GUGGG--GAUCAAAGCUCAG-GCUUAUC Aspergillus terreus GGA-CUCCC--GAAC--GGGAG--GAUCAAAGCCCAG-GCUUGUC Aspergillus aculeatus GGACCUC----GAAA----GGG--GAUCAAAGCUCAGUGCUUGUC Aspergillus flavus GGACCUC--UCUAGAGU--GAG--GAUCAAAGCUUAG-GCUUGUC Aspergillus oryzae GGACCUC--UCUAGAGU--GAG--GAUCAAAGCUUAG-GCUUGUC Aspergillus niger GGACCCC----UUGAU---GGG--GAUCAAAGUUUAG-GCUUGUC Aspergillus carbonarius GGACCUC----UUAAUU--GGG--GAUCAAAGCUUAG-GCUUGUC Penicillium chrysogenum GGACCUC--UCCCGAGU--GAG--GAUCAAAGCUUUG-GCUUGUC Penicillium marneffei GGACCUC--UCUCGAGU--GAG--GAUCAAAGCUUCG-GCUUGUC Paracoccidioides brasilliens GGACCUC--UCUAGAGU--GAG--GAUCAAAGCUUAG-GCUUGUC E Blastomyces dermatidis GGACCUC--UCUCGAGU--GAG--GAUCAAAGCUUAG-GCUUGUC Histoplasma capsulatum GGACCUC--UCUCGAGU--GAG--GAUCAAAGCUUCG-GCUUGUC Coccidioides immitis GGACCUC--UCUUGAGU--GAG--GCUCAAAGUUUCG-GCUUGUC Coccidioides posadasii GGACCUC--UCUUGAGU--GAG--GCUCAAAGUUUCG-GCUUGUC Uncinocarpus reesii GAGCCUC--UCUCGAGU--GAG--GAUUAAAGCUUUU-GCUUAUC Microsporum canis GGACCUC--UCUCGGGU--GAG--GAUCAAAGCUCCG-GCUUGUC Microsporum gypseum GGACCUC--UCUCGAGU--GAG--GAUCAAAGCCCUG-GCUUGUC Trichophyton rubrum GGACCUC--UCUCGAGU--GAG--GAUCAAAGCCUUG-GCUUGUC Trichophyton equinum GGACCUC--UCUUGAGU--GAG--GAUCAAAGCCUUG-GCUUGUC Trichophyton tonsurans GGACCUC--UCUUGAGU--GAG--GAUCAAAGCCUUG-GCUUGUC Arthroderma benhamiae GGACCUC--UCUCGAGU--GAG--GAUCAAAGCCUUG-GCUUGGC Trichophyton verrucosum GGACCUC--UCUCGAGU--GAG--GAUCAAAGCCUUG-GCUUGGC Talaromyces stipitatus GGGCCUC--UCUAGAGU--GAG--GAUCAAAGCUUUG-GCUUGUC Stagonospora nodorum GGACCUC--UCAAGAGU--GAG--GCUCAAGGUCUAG-ACCUGUC Pyrenophora tritici-repentis GGACCUC--UCUCGAGU--GAG--GAUCAAGGUCUAG-GCCUGUC Pyrenophora teres f. teres GGACCUC--UCUCGAGU--GAG--GAUCAAGGUCUAG-GCCUGUC Setosphaeria trucica GGACCUC--UCUAGAGU--GAG--GCUCAGGGUCUAG-ACCUGUC Cochliobolus heterostrophus GGACCUC--UCAAGAGU--GAG--GCUCAAGGUCUAG-ACCUGUC Alternaria brassicicola GGACCUC--UCAUGAGU--GAG--GCUCAAGGUCUAG-ACCUGUC D Leptosphaeria maculans GG-CCUU--UCCAGUGU--GAG--GUGC-AGGCUUAG-ACCUAUC Mycosphaerella graminicola GGACCUC--UCUCGAGU--GAG--GAUCAAAGCUUUG-GCUUGUC Mycosphaerella fijensis GGACCUC--UCUAGAGU--GAG--GAUCAAAGCUUUG-GCUUGUC Septoria musiva GGACCUC--UCUAGAGU--GAG--GAUCAAAGUUUUGUGCUUGUC Dothistroma septosporum GGACCUC--UCUCGUGU--GAG--GAUCAAAGCUUUG-GCUUGUC Hysterium pulicare GGACCUC--UCUAGAGU--GAG-CG-UCAAAGCUUCG-GCUUGUC Rhytidhysteron rufulum GGACCUC--UCUAGAGU--GAG-CG-UCAAAGUUUCG-GCUUGUC Tuber melanosporum GGACACAC---ACAU---GUGCAAG-UCAGAGCUCCGUGCUCGUC P Phymatotrichum omnivorum GGAC------CAUGC------G-UCAGAGUUCCGUACUCGUC Saitoella complicata GGACCUUCC---CAUU-GGAAC-CG-UCAAAGCUCCGUGCUUGUC Schizosaccharomyces pombe AGAUC------CAUG-----G--A-UCAAAGCUUUU-GCUUGUU Structure [(((((((()))))) )) (((()))) ] P6 P6.1 Qi_SupplementaryFig.S3

U U G A A U C C G Truncated Ncr TERfragments(1 µMeach) G C 1-2049 C G _ U G Template-Pseudoknot T-PK1 T-PK1 T-PK2 T-PK3 C G _ TWJ1TWJ1TWJ1TWJ1TWJ2TWJ3TWJ4 UU A UC G A U GC UA U U UC G C G U A C GA C G C AUG GA 1515 A A UA U ACAAAAGC GUC UCAGUU GGCAGUGUUC C G G UUUUGUCCGUGC GACGC CCG A GAGCC G A CGG GCACGG G U U A C 225 U U U U G 1289 1402 C C acGCCAG C g A C g GUAUACACUGGU UAACCCUAA U A C C G 462 Template A UU A C G CU G C 826nt ∆463-1288 434 C G 255 U A ∆256-433 a g aa

178nt T-PK1: 190-1532nt T-PK2: 190-1532nt( ∆463-1288+GGAC ) T-PK3: 225-1515nt( ∆463-1288+GGAC /∆256-433+GAAA )

B TWJ1 TWJ2 TWJ3 TWJ4 UG G UG UG U A UU A U A U A C G C G C G C G U U U U U U U U A U A U A U A U C G C G C G C G UUC UUC UUC U C GA C GA C GA C GA U C C G U G U G U U G G C G G C G G C G G A U A AU A U A AU A U A AU 1826 A U AG AU G CAA C G CAA C G CAA C G C A C U U U A U U U U G U G CUG G CUG G CUG 5' G G U G U G U U 1864 C G C G C G C U G U G U G 3' G C G C G C 1821 G C 1869 1821 G C 1869 1821 G C 1869 5' 3'

1891 3' 1783 2049 1783 5' 3' 5' l.c.- 1 2 34 56789 Relative 100 30512217921360612 Activity(%) Qi et al.

Supplementary Table S1. Recombinant Neurospora crassa strains used in this study Strain Genotype Reference NC1 mat A his-3 Δmus-52::bar+ Honda & Selker (2009) NC2 mat A his-3 Δmus-52::bar+, tert-3xFLAG::hph::loxP this study NC3 mat a, tert-3xFLAG::hph::loxP this study NC4 mat A his-3+::Pccg-1::ncrTER(template:5’-UAACCCCUAA-3’), this study Δmus-52::bar+ NC5 mat A his-3+::Pccg-1::ncrTER(template:5’-UAAGCCCUAA-3’), this study Δmus-52::bar+ NC6 mat A his-3+::Pccg-1::ncrTER(template:5’-UAAACCCUAAA-3’), this study Δmus-52::bar+ Honda, S., and Selker, E.U. (2009). Genetics 182, 11-23.

1 Qi et al.

Supplementary Table S2. Species with TER identified in this study Template to Full Class Species Source TWJ (nt) length (nt) Neurospora crassa 1423 (442-1864) 2049a JGI, Broad Neurospora discreta 1417 ~2032b JGI Neurospora tetrasperma 1442 ~2065b JGI Sordaria macrospora 1458 NCBI Podospora anserina 1176 NCBI Colletotrichum graminicola 1120 Broad Colletotrichum higginsianum 1134 Broad Fusarium graminearum 1113 Broad Fusarium verticillioides 1132 Broad Fusarium oxysporum 1137 Broad Metarhizium acridum 1033 NCBI Metarhizium anisopliae 1061 NCBI Verticillium albo-atrum 1131 Broad Sordariomycetes Verticillium dahilae 1130 Broad Trichoderma atroviride 1143 JGI Trichoderma reesei 1124 JGI Trichoderma virens 1083 JGI Acremonium alcalophilum 1157 JGI Nectria haematococca 1150 JGI Chaetomium globosum 1210 JGI, Broad Sporotrichum thermophile 1243 JGI Cryphonectria parasitica 1189 JGI Thielavia terrestris 1211 JGI Magnaporthe oryzae 1274 Broad Magnaporthe poae 1520 Broad Gaeumannomyces graminis 1407 Broad Epichloe festucae 1159 OU Geomyces destructans 1427 Broad Leotiomycetes Botrytis cinerea 1343 Broad Sclerotina sclerotiorum 1335 Broad Cladonia grayi 1478 JGI Lecanoromycetes Xanthoria parietina 1834 JGI Aspergillus nidulans 1162 (270-1431) 1584a JGI, Broad Neosartorya fischeri 1494 Broad Aspergillus fumigatus 1498 Broad Aspergillus clavatus 1491 Broad Aspergillus terreus 1513 Broad Aspergillus aculeatus 1472 JGI Eurotiomycetes Aspergillus flavus 1477 Broad Aspergillus oryzae 1479 Broad Aspergillus niger 1476 JGI, Broad Aspergillus carbonarius 1476 JGI Penicillium chrysogenum 1509 NCBI Penicillium marneffei 1327 NCBI Paracoccidioides brasiliensis 1703 NCBI

2 Qi et al.

Blastomyces dermatidis 1717 Broad Histoplasma capsulatum 1687 Broad Coccidioides immitis 1642 Broad Coccidioides posadasii 1640 Broad Uncinocarpus reesii 1605 Broad Microsporum gypseum 1558 Broad Microsporum canis 1609 Broad Trichophyton rubrum 1559 Broad Trichophyton equinum 1561 Broad Trichophyton tonsurans 1561 Broad Arthroderma benhamiae 1557 NCBI Trichophyton verrucosum 1557 NCBI Talaromyces stipitatus 1343 NCBI Stagonospora nodorum 1470 JGI Pyrenophora tritici-repentis 1471 JGI Pyrenophora teres f. teres 1468 NCBI Setosphaeria trucica 1471 JGI Cochliobolus heterostrophus 1476 JGI Alternaria brassicicola 1466 JGI Dothidiomycetes Leptosphaeria maculans 1481 INRA Mycosphaerella graminicola 1772 (376-2147) 2425a JGI Mycosphaerella fijiensis 1634 JGI Septoria musiva 1742 JGI Dothistroma septosporum 1632 JGI Hysterium pulicare 1616 JGI Rhytidhysteron rufulum 1548 JGI Tuber melanosporum 1275 INRA Pezizomycetes Phymatotrichum omnivorum 1190c OU Saitoella Saitoella complicata 1313 (380-1692) 1893a JGI a: Full-length mature TER confirmed by 5’- and 3’-RACE and deposited in the GenBank (N. crassa TER=JQ793886, A. nidulans TER=JQ793887, M. graminicola TER=JQ793888 and S. complicata TER=JX173807) b: Size of the full-length TER was predicted by alignment with NcrTER c: The genomic sequence of TER was determined using the Universal GenomeWalker Kit (ClonTech) and the genomic DNA was obtained from Dr. Stephen Marek. JGI: DOE Joint Genome Institute, URL: www.jgi.doe.gov Broad: Broad institute of MIT and Harvard, URL: www.broadinstitute.org NCBI: National center for Biotechnology Information, URL: www.ncbi.nlm.nih.gov INRA: Institut national de la recherché agronomique, URL: www.inra.fr OU: Oklahoma University, URL: www.genome.ou.edu/fungi.html

3 Qi et al.

Supplementary Figure Legends

Supplementary Fig. S1. Purification of Neurospora crassa telomerase holoenzyme. (A) Generation of the NcrTERT-3xFLAG strain. Schematic drawing shows XhoI sites at the N. crassa tert locus before and after insertion of the recombinant NcrTERT-3xFLAG DNA cassette. The blue color indicates the endogenous tert locus, whereas the orange color shows the inserted DNA cassette. Digestion of the genomic DNA with XhoI results in a 1.4 kb band from the endogenous tert locus which increases to 3.2 kb with the inserted cassette. Positions of the forward (F) and reverse (R) PCR primers for generating the recombination DNA cassette are shown. The DNA cassette includes a Hygromycin B phosphotransferase gene (hph) driven by a TrpC promoter (Ptrpc). (B) Southern blot analysis of XhoI digested genomic DNA from wild-type (wt) and 12 transformants (lanes 1 to 12) with a radiolabeled probe against tert 3’ region. (C) Gel filtration of N. crassa NcrTERT-3xFLAG strain nuclear extracts. Five mililiter fractions were collected as indicated. Fractions 10-20 were assayed for telomerase activity together with the original nuclear extract. The PCR internal control is shown. Fractions with peak activities (#15-17) were combined for IP. (D) Telomerase activity detected by TRAP assay after IP with anti-FLAG antibody conjugated beads. One microliter of input and supernatant after IP together with 5% of the beads (pellet) were subjected to the telomerase TRAP assay.

Supplementary Fig. S2. Sequence alignment of Pezizomycotina and S. complicata TERs. (A) Sequence alignment of the template-pseudoknot domain. The TER sequences from Sordariomycetes (S), Leotiomycetes (Leo) and Saitoella complicata are included. Nucleotides with ≥80% identity are shaded yellow. Numbers within the sequences indicate nucleotides omitted. Base-pairings are denoted below the sequences. Structural features are labeled under the sequence alignment and colored the same as in Fig 3B. (B) Sequence alignment of the P6/6.1 region. The TER sequences from Sordariomycetes (S), Leotiomycetes (Leo), Lecanoromycetes (Lec), Eurotiomycetes (E), Dothidiomycetes (D), Pezizomycetes (P) and Saitoella complicata (shown in red) are included. Nucleotides with ≥80% identity are shaded yellow. Base-pairings are denoted below the sequences.

Supplementary Fig. S3. The minimal functional core of NcrTER. (A) The template-pseudoknot fragments, T-PK1, T-PK2 and T-PK3, contained the template-pseudoknot domain with deletions Δ256-433 and Δ463-1288. For the Δ256-433 and Δ463-1288 deletions, the truncated core-closing helix 2 and template-boundary helix were capped with GGAC or GAAA, respectively, for structural stability. The minimal fragment T-PK3 contained both deletions, while the T-PK2 fragment contained only the Δ463-1288 deletion. Residue numbers are based on the mature NcrTER (1-2049). Predicted NcrTER

4 Qi et al. structure of nucleotides 225-1515 is presented, with truncations indicated in blue and green. (B) TER fragments TWJ1, TWJ2, TWJ3 and TWJ4, contained the 3’ portion of NcrTER with various truncations removing regions flanking the P6/6.1 element. NcrTER TWJ3 and TWJ4 secondary structure predictions are shown. (C) Telomerase activity assay using the longer fragments T-PK1 and TWJ1 showed that both fragments were required to reconstitute telomerase activity in trans (lanes 2-4). The full-length NcrTER activity is shown in lane 1. Single fragment pseudoknot T-PK1 and TWJ1 reconstituted with NcrTERT in RRL respectively were shown in lanes 2 and 3. Various two-fragment reconstitutions with NcrTERT in trans were shown from lane 4 to lane 9. A 15 mer 32P end-labeled oligonucleotide served as the loading control (l.c.). Relative activities (%) are shown under the gel.