Strains, plasmids, and media

U. maydis strains 521 (a1b1) and 518 (a2b2) were used. We also used as control the replicative plasmid pCM54, which is a pUC18 derivative that contains UARS1 and an ~4 kb chimeric hph gene (bacterial hph gene under hsp70 promoter and terminator sequences that confers hygromycin resistance) which was provided by W.K. Holloman, Cornell Weill Medical College, NY, USA. A 2,039 bp cassette flanked by BamHI containing a second version of the chimeric hph gene (bacterial hph gene under hsp70 promoter and short terminator sequences) was excised from the plasmid pCM1007 (also provided by W.K. Holloman) and used for the assembly of the disruption cassette. The plasmid pBluescript KS(-) (Stratagene, La Jolla, CA, USA) was used as the backbone for cloning of the isolated DNA fragments or PCR amplified sequences and to assemble the interruption cassette. Escherichia coli strain XL1-blue MRF’ (Δ(mcrA)183 (mcrCB-hsdSM-mrr)173 recA1 endA1 gyrA96 thi-1 supE44 relA1 lac [F’ proAB lacIq ZΔM15 Tn10(tetr)]; Agilent, Santa Clara, CA, USA) and DH5α (F⁻endA1 glnV44 thi-1 recA1 relA1 gyrA96 deoR nupG purB20 φ80dlacZΔM15 Δ(lacZYA-argF)U169, hsdR17(r_K^–m_K⁺), λ^–; Thermo Fisher, Mexico), were used as hosts for plasmids. The restriction enzymes used were from Thermo Fisher Scientific (Waltham, MA, USA). LB media supplemented with ampicillin sodium salt (100 μg/ml; Sigma-Aldrich Co., St. Louis, MO, USA) was used to select E. coli transformant strains, and bacterial cultures were grown at 37°C overnight. U. maydis strains were grown in YEPS media (1% yeast extract, 1% peptone and 2% sucrose), in potato dextrose broth (PDB) or agar (PDA) from Difco (Becton Dickinson and Company, Sparks, MD, USA), or in complete or minimal medium [38] at 28°C.

Identification of Ustilaginales loci containing TER candidates

The genomic sequences consulted were as follows: Ustilago maydis strain 521, accession GCF_000328475.2; Ustilago bromivora strain UB2112 accession GCA_900080155.1; Ustilago cynodontis strain NBRC 9727; accession GCA_001676635.1; Ustilago esculenta strain MMT, accession GCA_000819925.1; Ustilago hordei strain Uh4857-4, accession GCA_000286035.1; Ustilago trichophora strain RK089, accession GCA_001654535.1; Ustilago tritici strain Utri01, accession GCA_002993085.1; Ustilago vetiveriae strain RK 075, accession GCA_001735935.1; Ustilago xerochloae strain UMa702, accession GCA_001736075.1; Sporisorium iseilematis-ciliati strain BRIP 60887 a, accession GCA_001748505.1; Sporisorium reilianum strain SRZ2, accession GCA_000230245; Sporisorium scitamineum strain SSC39B, accession GCA_001010845.1; Pseudozyma antarctica strain T-34, accession GCA_000334475.1; Moesziomyces aphidis strain DSM 70725, accession GCA_000517465.1; Kalmanozyma brasiliensis strain GHG001, accession GCF_000497045.1; Pseudozyma flocculosa strain PF-1, accession GCA_000417875.1; Pseudozyma hubeiensis strain NBRC 105055, accession GCA_001736105.1; Pseudozyma tsukubaensis strain NBRC 1940, accession GCA_001736125.1; Pseudozyma sp. F8B2, accession GCA_003004685.1; Moesziomyces sp. F5C1, accession GCA_003004715.1; Moesziomyces sp. F16C1, accession GCA_003004725.1; and Melanopsichium pennsylvanicum strain 4, accession ERS346485. Candidate loci encoding the RNA component of the telomerase gene (here named ter1) were identified through a search of the entire U. maydis genome (GCF_000328475.2) according to the method described by Chakrabarti et al. [25]. Briefly, the entire genome sequence of U. maydis was examined using both the Motif Finding Automaton (MFA) from Pérez et al. [39] and SnapGene® v.3.2.1 software (from Insightful Science; available at snapgene.com) to find all regions containing at least one and a half copies of the template sequence (5’-CCCTAA-3’) in any possible permutation. Sequences identified as telomeres were discarded, and sequences identified as ORFs, structural RNAs and nonconserved sequences; target motifs along with 300 bp of flanking sequences on each side were recovered and saved in a local database set. Each sequence was then aligned to all genomes of the Ustilaginales taxon group available on April 30, 2019, using blastn option of BLAST [40] software of NCBI. The highest-scoring single loci conserved in all organisms were sought, and 21 candidates were obtained. Using SnapGene® v.3.2.1 and the genome annotation of U. maydis https://ftp.ncbi.nlm.nih.gov/genomes/all/GCF/000/328/475/GCF_000328475.2_Umaydis521_2.0/GCF_000328475.2_Umaydis521_2.0_genomic.gbff.gz as a reference, the 21 candidate loci were re-examined both to predict the ORFs flanking the intergenic regions and to perform a synteny analysis of nearly 15 kb on both sides of the putative template domain in the probed genomes, even including putative ORFs without annotation. To achieve this aim, approximately 30 kb of each genome of the chosen organisms was cropped, and the cropped sequences were aligned using blastn. After the analysis, the complete intergenic sequences harboring the TERs were recovered, and their lengths were determined (ranging from 784 to 2,669 nucleotides); then, multiple sequence alignment was performed using the online version of MAFFT software (https://mafft.cbrc.jp/alignment/server/index.html) [41] to find conserved sequence motifs. A perfect match of the putative template domain was observed within the putative loci ter1, nucleotide sequence data reported are available in the Third-Party Annotation Section of the DDBJ/ENA/GenBank databases under the accession number TPA: BK059259.

Transcriptional expression

Before ter1 gene disruption, gene expression data were sought from the NCBI database, using the ter1 sequence as a query. To confirm the existence of those mature transcripts derived from either the 3’ of the hypothetical gene UMAG_03168 located upstream of the ter1 locus or the 5’ ter1 transcript, a series of primers was designed to perform the reverse transcription assays (S3 Table), and Oligo dT₁₈ (Thermo Fisher Scientific) was also used. 1 μg of RNA previously extracted as described below, added with 100 μM reverse oligo, and lead to a final volume of 12 μl, was incubated at 65°C for 5 min. The previous mixture was added with 4 μl of 5X Buffer (RevertAid H Minus First Strand cDNA Synthesis Kit, Thermo Fisher Scientific), 20 U of RiboLock RNAse Inhibitor (Thermo Fisher Scientific), 10 mM dNTPs, and 200 U of RevertAid H Minus Reverse Transcriptase (RevertAid H Minus First Strand cDNA Synthesis Kit), the reaction was incubated at 25°C for 5 minutes, followed by incubation at 45°C for 60 minutes and finally at 70°C for 10 minutes. 3 μl of cDNA from the previous reaction or 300 ng of DNA used as a positive control were added with 16 μM forward primer, 16 μM reverse primer. PCR reactions were performed using Platinum® Pfx DNA Polymerase (Thermo Fisher Scientific) according to the manufacturer’s instructions.

Nuclei extraction

For the extraction of nuclear RNA (RNA_N) and total RNA (RNA_T), nuclei were obtained based on Groner and Phillips [42]. Briefly, 100 μl of an optimal preinoculum of U. maydis strain 521 was added to 5 ml of YEPS broth and incubated at 30°C for 12 h with aeration (200 rpm) to obtain the inoculum, from which 100 μl was taken and added to 100 ml of YEPS and incubated as above until OD₆₀₀ = 0.8. Cells were centrifuged at 3,500 rpm/10 min at room temperature. The supernatant was discarded, and the pellet was suspended in 100 ml of prewarmed YEPS broth and incubated at 30°C for 4 h under constant shaking (200 rpm) to obtain exponentially grown cells. The culture was centrifuged as above, and the pellet was recovered and suspended in 5 ml of SCS solution (50 mM sodium citrate, 1 M sorbitol, pH 5.8) and washed twice. For each wash, centrifugation was performed for 7 min at 2,000 rpm, and after the last wash, the pellet was resuspended in 2 ml of SCS, and 50 mg/ml of lysing enzymes from Trichoderma harzianum (L1412, Sigma-Aldrich Co.) were added. After 20 min of incubation at 37°C, protoplasts were centrifuged at 1,000 rpm/10 min, recovered, and washed twice with 20 ml of 1 M sorbitol at 4°C. After the last centrifugation, the pellet was recovered, 2 ml of NIB buffer (17% glycerol, 50 mM MOPS pH 7.5, 150 mM potassium acetate, 2 mM MgCl₂, 500 μM spermidine and 150 μM spermine) was added, and the pellet was left on ice for 30 min. For centrifugation, 3 ml of sucrose cushion (2 M sucrose and 2.5 μl of 1 M MgCl₂) was placed in a tube for each 2 ml of lysate and centrifuged at 10,000 rpm for 15 min. The supernatant was removed and stored at -80°C for other purposes. The pellet was suspended in 3 ml of SDS buffer (0.1 M NaCl, 0.02 M EDTA, 0.02 M Tris-HCl pH 7 and 0.5% SDS) and homogenized, and subsequent steps were performed as indicated by the RNeasy Mini Kit (Qiagen). The samples were evaluated by gel electrophoresis, and the optical density of each sample was read to analyze the sample integrity. Samples were stored at -80°C.

Transcriptome analysis

Analysis of the regions comprising UMAG_03168, ter1 and UMAG_03169 from strain 521 of U. maydis was performed. Briefly, poly(A+) RNA and total (ribo−) RNA were obtained from the nuclear fraction of the wild-type 521 strain of U. maydis. cDNA libraries were prepared using the TruSeq Stranded mRNA Library Preparation kit and Truseq Stranded Total RNA Gold Prep Kit (Illumina Inc.), respectively, and sequenced by DNAlink using the Illumina NovaSeq 6000 platform, generating paired-end reads of 101 bp. The quality control of reads was carried out using FastQC v0.11.9 [43] subsequently, the adapters and low quality reads were eliminated using Trimmomatic v0.36 [44] software with the following parameters: PE, -phred33, ILLUMINACLIP:$TRIMMOMATIC_DIR/adapters/TruSeq3-PE-2.fa:2:30:10:1:true SLIDINGWINDOW:4:15 LEADING:3 TRAILING:3 AVGQUAL:30 MINLEN:25. Then, rRNA and mitochondrial transcript reads were removed using SortMeRNA v4.2.0 [45] software with default settings, using the rRNA and mitochondrial genome sequences of U. maydis. De novo transcriptome assembly was conducted with Trinity software v2.11.0 [46] with the following settings:—seqType fq,—min_kmer_cov = 2,—min_contig_length = 100—jaccard_clip—SS_lib_type RF. Read alignment and abundance quantification were performed using Bowtie2 v2.4.2 [47] and RSEM v1.3.3 [48], respectively, using the align_and_estimate_abundance.pl script from the Trinity pipeline with the following settings:—seqType fq,—est_method RSEM,—aln_method bowtie2. Data visualization was carried out with the Integrative Genomics Viewer [49]. Finally, for the analysis of differential expression, the online platform IDEAMEX (Integrated Differential Expression Analysis MultiExperiment, http://www.uusmb.unam.mx/ideamex [50] was used with the following methods: DESeq2 [51], NOISeq [52], Limma-Voom [53], and EdgeR [54]. Transcripts were considered differentially expressed only if they had a log2FC ≥ 2 and an FDR ≤ 0.01. Raw sequencing reads and expression data were submitted to Gene Expression Omnibus of NCBI (accession number: GSE184051).

Nucleic acids extraction and manipulation

DNA extraction from U. maydis cells was achieved as follows: 50 ml of YEPS broth was inoculated with the chosen fresh strains. The cell cultures were grown overnight to an OD₆₀₀ = 1.0–1.2 at 28°C with aeration; cells were harvested in a 5810 Eppendorf centrifuge with an A-4-62 rotor (Eppendorf North America, Inc., Hauppage, NY, USA) at 2,000 rpm. The culture medium was discarded, and the pellet was frozen and ground in liquid nitrogen with a mortar and pestle. Powder was poured into a 50 ml tube containing 5 ml of lysis buffer (200 mM Tris-HCl, pH 8.5; 250 mM NaCl; 25 mM EDTA; and 0.5% SDS) and mixed thoroughly. DNA was extracted by a phenol-chloroform procedure, conducted in two rounds with 15-minute centrifugation intervals. The supernatant was washed once with chloroform and precipitated with 0.6 v/v isopropanol. The DNA was pelleted by centrifugation at 12,000 rpm in a microfuge; the aqueous phase was discarded, and the pellet was suspended in 100 μl of sterile distilled water and re-extracted with phenol once again. After centrifugation, the DNA pellet was suspended in 50 μl of sterile distilled water. For Southern blot assays, the DNA was quantified by spectrophotometry using a GENESYS 10S UV-Vis (Thermo Fisher Scientific) spectrophotometer according to the manufacturer’s instructions prior to and after restriction enzyme digestion. Approximately equal amounts of DNA were used for each lane in the TRF assays.

RNA was extracted as follows: fresh colonies were inoculated into tubes containing 5 ml of YEPS medium and incubated at 28°C overnight. When necessary, hygromycin (200 μg/ml, (Sigma-Aldrich Co.) was added to the medium; 50 μl of the preculture was added to 5 ml of YEPS medium and incubated for 10 h. Then, an aliquot of 0.3 to 0.5 ml of strain was inoculated into a flask containing 200 ml of YEPS medium (with or without hygromycin for selection) and incubated until it reached OD₆₀₀ = 0.6–0.8. The cultures were centrifuged at 2,000 rpm for 10 min, the supernatant was discarded, and the pellet was frozen in liquid nitrogen and ground to powder with a mortar and pestle. The powder was recovered, and 2 ml of lysis buffer R (0.02 M sodium acetate, 0.5% SDS, and 1 mM EDTA, pH 5.5) was added and mixed thoroughly with the powder by vortexing. Samples were then partitioned into 500 μl aliquots in microtubes. Each sample was mixed with 500 μl of acid phenol (phenol equilibrated with 0.02 M sodium acetate pH 5.5) and placed at 65°C for 5 min mixing occasionally. The samples were cooled and centrifuged at 10,000 rpm for 10 min at 4°C, and the supernatant was transferred to another tube. Then, 500 μl of acid phenol:chloroform:iso-amyl alcohol (125:24:1) was added. The mixture was homogenized and centrifuged at 10,000 rpm for 10 min at 4°C. The aqueous phase was washed with 500 μl chloroform twice and centrifuged at 10,000 rpm for 5 min at 4°C; then, 7.5 M ammonium acetate and 2 volumes of absolute ethanol were added, and the sample was mixed vigorously and left at -70°C for 1 h. RNA was recovered by centrifugation at 10,000 rpm for 10 min at 4°C. The pellet was washed twice with 70% cold ethanol at -20°C. Finally, the isolated RNA was dissolved in 100 μl of sterile water. Samples were treated with RNase-free DNase I (Merck KGaA, Darmstadt, Germany) according to the manufacturer recommendations.

Disruption cassette

Approximately 1 kb of the DNA sequences flanking the putative template domain in the ter1 locus were amplified by PCR and cloned in pBluescript KS(-), oligonucleotide primer pairs (Integrated DNA Technologies Inc., San Diego CA, USA) were Cr8S1-5Upp (genomic location, 124,977–125,009) and Cr8S1-5Low (126,126–126,157), which amplify 1,181 bp of the hanging 5’ side of the disruption cassette. The amplified DNA fragments contained NotI and BamHI cut sites on their 5’ and 3’ sides, respectively. For the 3’ hanging DNA fragment of the cassette, Cr8S1-3Upp (126,791–126,819) and Cr8S1-3Low (127,937–127,965) were used to amplify a 1,175 bp DNA fragment downstream of the putative template domain and bordered by BamHI and NotI cut sites. The two fragments were cloned into pBluescript KS(-) to generate plasmids pTer5 and pTer3, respectively. A 1,170 bp DNA fragment was excised from the pTer5 plasmid by NotI/BamHI digestion and was subcloned into pTer3 in the NotI/BamHI sites. The resulting plasmid was pTerSub5, which contained a BamHI site between the upstream and downstream fragments. Finally, the 2,039 bp DNA BamH1 fragment from pCM1007 containing the hph chimeric gene was subcloned into the BamHI site of pTerSub5 to obtain the 4,373 bp disruption cassette in the plasmid pTer1, which was excised from the plasmid by complete NotI digestion. The excised fragment was resolved by electrophoresis on an 0.8% agarose gel and recovered from the agarose by the freeze–squeeze method.

Transformation and confirmation of ter1Δ mutants

Transformation was performed as described in Bautista-España et al. [32] following the Tsukuda et al. [55] protocol, using the 518 (a2b2) strain, the linearized disruption cassette, and pCM54 as a positive control. The bottom layer of the plates of selective medium was supplemented with 400 μg/ml hygromycin such that the medium reached a final concentration of 200 μg/ml after the top layer without hygromycin was added. Transformants appeared after five days of incubation on selective medium at 28°C, and individual clones were streaked on plates of PDA medium without selective pressure and then grown on YEPS plates and in YEPS broth, both containing hygromycin at 200 μg/ml. These cultures were used for observations of telomere shortening after a passage series, for cryopreservation of early-stage transformants in glycerol (50% v/v, -70°C) and for total DNA extraction as described above. ter1 disruption in transformants was confirmed by PCR using the oligonucleotide primer pairs Template-Upp (genomic location 126,268–126,287 [irp 188–207 nt]) and Template-Low (genomic location 126,492–126,511 [irp 412–431 nt]) to check for the absence of the template domain. Its replacement by the hph cassette was confirmed by PCR using the primer pair 5H-Upp (genomic location 124,812–124,835) and 5H-Low (position on the selection marker 396–419) and the primer pair 3H-Upp (position on the selection marker 1,768–1,788) and 3H-Low (genomic location 129,503–129,524) to amplify DNA fragments spanning from the 5’ and 3’ ends of the selective marker to DNA regions outside the homologous regions used for the disruption cassette assembly.

Growth kinetics

Individual colonies of clones and the wild-type 518 strain were inoculated into tubes containing 5 ml YEPS broth and grown at 28°C with aeration to OD₆₀₀ = 1; 1 ml from these precultures was inoculated into 100 ml of YEPS medium and incubated at 28°C with aeration (200 rpm). A series of 1 ml samples were taken every 2 h for 18 h, and the OD₆₀₀ was determined and recorded. Serial dilutions were made to determine the CFU at each time point, plated in YEPS medium, and incubated at 28°C for 24 h. The data were graphed using GraphPad Prism v. 8.4.3. Linear regressions of the log10 CFU vs. time were performed, and the exponential phase data were used to calculate the generation time of the wild-type strain and ter1Δ mutants.

TRF analysis in U. maydis

TRF analysis was carried out as previously described by Bautista-España et al. [32]. Four micrograms of total DNA from each strain or clone was digested to completion with PstI (Thermo Fisher Scientific), and 500 ng of the digested DNA per lane was electrophoresed in a 1.0% agarose gel and transferred to a nylon membrane (Hybond-N, Invitrogen, Carlsbad, CA, USA) as described by Sambrook et al. [56]. Hybridization was carried out in Church and Gilbert [57] solution (0.5 M Na₂HPO₄ pH 7.2, 7% SDS, 1 mM EDTA and 1% BSA; Sigma-Aldrich Co.). The probe used in this work was the 232 bp HinCII/XhoI DNA fragment from the pUT2 plasmid [30] (accession X77242.1) containing 37 copies of the telomere repeat TTAGGG and 9 bp of the TAS sequence, labeled with digoxigenin-11-uridine-5’-triphosphate (Roche Diagnostics GmbH, Mannheim, Germany, provided by Sigma-Aldrich Co.). Hybridization was performed at 59°C, followed by washes with 0.2X SSPE and 0.1% SDS at 59°C. TRF length was estimated using the ImageJ 1.46r program (Wayne Rasband, NIH, USA).

Microscopy

U. maydis cells freshly cultured in YEPS medium at 2.0x10⁷ cells/ml were harvested by centrifugation at 2,500 rpm for 5 min at 30°C. The medium was discarded, and the cells were washed twice in 200 μl of TBS (50 mM Tris HCl pH 7.5, 150 mM NaCl), with centrifugation for 5 min at 2,500 rpm between washes. The cells were suspended in 100 μl of TBS and mixed thoroughly with 100 μl of fixative solution (50% v/v ethanol, 10% v/v CH₃COOH). The suspension was kept at RT for 5 min and centrifuged again to remove the fixative solution. Cells were washed twice with 200 μl of TBS, and the pellet was suspended in 100 μl of the same solution supplemented with 0.2 μg/ml RNase A from bovine pancreas (Sigma-Aldrich Co.). The cell suspension was kept at 4°C overnight, moved to 37°C for 1 h, centrifuged and washed once in 200 μl of TBS. The cell pellet was then suspended in 50 μl TBS with propidium iodide (2 μg/ml; Sigma-Aldrich Co.) and allow to absorb the dye for 6–8 h at 4°C. Stained cells were washed twice with 100 μl of TBS and resuspended in 50 μl of TBS. Then, calcofluor (Sigma-Aldrich Co.) was added at 25 ng/ml TBS, and after 5 minutes of treatment, the excess dye was removed by centrifugation and suspended in 25 μl of buffer for observation under a microscope.

Plant infection assays

Infection assays in five days-old maize-seedlings line MCS-02 were carried out according to the method previously described by Bautista-España [32], using mixtures of compatible strains wild-type and ter1^– to achieve the heterozygous cross in planta.

PFGE

Fresh inoculum of U. maydis strains (50 μl) was inoculated into 20 ml of YEPS and incubated overnight at 20–30°C with aeration. When the culture reached OD₆₀₀ = 1.0, cells were collected by centrifugation at 2,000 rpm for 7 min, the supernatant was discarded, and the cells were washed twice in 3 ml SCS buffer (20 mM sodium citrate pH 5.8, 1 M sorbitol) and finally suspended in 500 μl of SCS buffer. Then, 100 μl of lysing enzymes from Trichoderma harzianum (L1412, Sigma-Aldrich Co.) at 50 μg/ml in CFS were added and allowed to stand for 15 min at 28–30°C without agitation. Protoplasts were centrifuged again for 7 min at 1,000 rpm and resuspended in 500 μl of CFS; then, the suspension was added to 800 μl of 1% low melting point agarose (UltraPure Low Melting Point agarose; Thermo Fisher Scientific) that had been dissolved in 125 mM EDTA, melted, and cooled to ∼37°C; the mixture was homogenized, poured into molds, and allowed to gel on ice for 10 min before being released into 50 ml sterile conical tubes, covered with LET buffer (0.5 M EDTA, 10 mM Tris pH 7.0, 5 μl β2-mercaptoethanol), and incubated at 30°C for 12 h. After this time, the LET was replaced with approximately 4 ml of NDS (0.15 M EDTA, 0.01 M Tris-HCl pH 8.5, 1% N-lauryl sarcosine pH 9.5, 2 mg/ml proteinase K (Sigma P-8044, Sigma Chemical Co.) and incubated overnight at 50°C; then, the NDS was discarded, and the plugs were washed in 50 mM EDTA solution and stored at 4°C until use. For the PFGE run, the plugs were rinsed in 100 mM EDTA solution, placed in the wells of a 1% agarose ultrapure (chromosomal-grade) gel, melted in 0.35X TBE and gelled in the PFGE mold. The wells were covered with LMP agarose melted in 125 mM EDTA. Pulsed field electrophoresis was performed in a Bio-Rad CHEF DRII unit with the following conditions: running buffer TBE 0.35 X, temperature 14°C, agarose gel 1% in TBE 0.35X. Block 1: Pulse 160 s, 2.5 V/cm, 16 h; block 2: Pulse 200 s, 3.5 V/cm, 16 h the running conditions varied as described in the images; running buffer TBE 0.35 X.

Computational search for the TER gene in U. maydis

Candidate loci harboring one and half copies of the 5’-CCCTAA-3’ motif in a nine-nucleotide window were identified and compiled, yielding 248 regions in the U. maydis genome, which we then searched for TER subunits. After BLAST searching these candidates against all genome sequences from Ustilaginales species, two candidates sharing a putative template domain were recovered. These candidates occupied orthologous genomic positions and had the highest scores. The first candidate matched sequences of the 18S ribosomal RNA gene, which is highly conserved among the Basidiomycota; therefore, it was discarded. The second candidate sequence consisted of 1,626 bp located on chromosome 8 (position 126,081–127,706 nt) that matched an intergenic region (ir) containing the putative template domain that was found toward the 5’ ends of intergenic regions among Ustilaginales species. Intergenic regions from these 22 fungal species, comprising the syntenic regions around putative TERs, were selected (Fig 1); our analysis revealed that the additional ORFs and intergenic sequences in all those syntenic regions were matched, which also coincided with the genome annotations of the sequences in the U. maydis strain 521 data bank. After manual inspection of putative TER subunits from 21 additional species of Ustilaginales, the potential template domain that once transcribed would be 5’ UAACCCUAA 3’ was identified (Fig 2).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Synteny map around the ter1 locus in Ustilaginales species.

Representation of the genes that encode proteins that are located in a widely preserved region around the intergenic area where the ter1 gene is located (indicated by a blue rectangle). The region shows high conservation, where the same two ORFs flank ter1 in every examined organism (except of the 3’ ORF in P. flocculosa).

https://doi.org/10.1371/journal.pone.0281251.g001

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Preserved elements identified within the ter1 locus.

(A) Representation of the chromosomal context of ter1. The image shows the region corresponding to both isoforms of the UMAG_03168 locus, the deleted region and the location of preserved motifs that include essential elements for TER activity obtained from multiple sequence alignment. (B) Multiple alignments of preserved motifs within orthologous intergenic regions of ter1. The intergenic regions contain characteristic TER sequences. The sequence of the putative template domain is conserved in all examined organisms. At the 5’ end of the region, three preserved elements were identified (CE1–3). Toward the 3’ end of the region, the consensus sequence AAU_5-6CY was identified, which resembles the Sm-binding site of TLC1 in S. cerevisiae and thus could correspond to the end of the mature ter1 transcript.

https://doi.org/10.1371/journal.pone.0281251.g002

Once we drew up the boundaries of the intergenic region, multiple sequence alignment of the genome orthologs was performed using MAFFT software. The putative template domain of U. maydis lies in a position between 385–393 nt from the intergenic ncRNA gene, here named ter1. These domains perfectly matched all putative TERs of Ustilaginales included in this study. Other well-conserved elements (CE) at the end 5’ of the putative TER of Ustilaginales were CE1 (TTCGATCY/RGATCGAA), which, along with CE2 (CGCCCGCC/GGCGGGCG) and CE3, conserve slight variants of CCTTGACT/AGTCAAGG in its sequence, the domains comprise three sequence stretches of unknown function within of the TER subunit (Fig 2). Downstream of ter1, a less conserved T+A-rich DNA sequence spanning 730–780 harbored a pseudoknot domain, which was found in all the species analyzed (S1C Fig). A brief visual inspection allowed us to identify a preserved region (S2A Fig) that resembles the AAGAGTTGGGCTCTG sequence, which creates the P6.1 helix of the CR4-CR5 domain [13,58] that is essential for catalytic activity [59,60]. According to the RNAalifold [61] program, in addition, this sequence has the potential to form a stem-loop P6.1-like structure (S2B Fig) and contains some residues within uracil and guanine loops, in the same way as the P6.1 of vertebrates and filamentous fungi [12]. Toward the 3’ end, a consensus transcribed sequence AAU_5-6CY that resembles the Sm (AAU₅GG) site of TLC1 in S. cerevisiae [24] was found in ter1 at position 1,208–1,218 (Fig 2). Two repeats of 90 bp in length, here named TRR1 (Telomerase RNA Repeat 1) and TRR2 (Telomerase RNA Repeat 2), that share 96.67% identity were found in ter1; their positions spanned from 1,380 nt to 1,469 nt and from 1,472 nt to 1,561 nt and were exclusive to the U. maydis TER sequence (Fig 2A). Copies of the polyadenylation sequence signal 5’-AGTAAA-3’ reported by Doyle et al. [62] were found in the TRR2 repeat embedded in positions 1,537–1,542 and downstream of TRR2 at positions 1,578–1,583 of the ter1 region (T1 and T2). Because of the findings presented below, we consider the second site to be the main polyadenylation site for the ter1 TER gene (Figs 3A and S3C). This site is moderately to poorly conserved in the intergenic region of most Ustilaginales analyzed here, a feature previously documented by Ji et al. [63] for intergenic regions of yeasts and plant genomes. The search for transcription factor binding sites of U. maydis previously described by Basse and Farfsing [64] yielded no target sites nor constitutive splicing consensus motifs were found in the entire ter1 gene.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. RNA-seq data analysis of the ter1 locus.

(A) Genomic context and conserved elements (upper panel) and density of RNA-seq reads mapped to the ter1 locus (lower panel). (B) Transcripts of the ter1 locus originating upstream of UMAG_03168 (represented by white rectangles) reconstructed in this work. (C) Analysis of the relative expression of the isoforms of ter1. Quantification of normalized counts was carried out with the EdgeR method for each isoform of ter1 and of isoforms of housekeeping genes.

https://doi.org/10.1371/journal.pone.0281251.g003

Characterization of the ter1 transcript

To confirm its identity as a gene, the transcriptional expression of ter1 needed to be corroborated. First, we examined the dbEST database of NCBI and the transcriptome of the WT strain previously reported [65,66]. To further characterize ter1, we analyzed the transcriptomes of both mRNA nuclear transcripts and nuclear RNA-ribodepleted transcripts of strain 521, which were assembled as described previously. From the reads of both nuclear fractions, we found four reconstructed transcripts of the ter1 locus, which represent four isoforms containing the structural domains of the TERs. All four isoforms of ter1 start 122 bp upstream of the start codon of the uncharacterized gene UMAG_03168 (https://fungi.ensembl.org/Ustilago_maydis/), which has a length of 859 bp and is located on chr 8 (spanning 125,222–126,080 nt); the gene encodes two protein isoforms, 134 aa (KIS68597, accession XP_011389625.1) and 142 aa (KIS68596, accession XP_011389624.1), respectively, which differ according to the splicing of the two exons of 241 and 188 bp and one intron of 430 bp composing the uncharacterized gene (Fig 2A).

The first isoform of ter1, here named ter1-i1, is a transcript of 2,475 nt that spans from 125,100 to 127,574; the start site position is consistent for all isoforms reported here (Fig 3B). In agreement with the transcriptome analysis, ter1-i1 is a nonspliced continuous transcript whose abundance is similar in the polyadenylated RNA fraction and the total RNA fraction (Table 1). The second isoform, ter1-i2, is a transcript of 2,347 nt produced from two spliced exons, the first located from 125,100 to 125,764, and the second located from 125,893 to 127,574 (Fig 3B). The reconstructed transcript suggests the use of the noncanonical alternative donor site (D2) TTTCCCCAATgtgagcacca (motif conserved among Ustilaginales is underlined, position 125,761–125,770) for splicing (Figs 3A and S3A); the branch (B) site AGCTCACCG (125,863–125,871) and acceptor site AG (125,891–125,892) are the same as those used for the splicing of the KIS68596 transcript. The third isoform reconstructed, ter1-i3, is a 2,332 nt transcript produced by the splicing of two exons; the first exon (664 nt, encoded by 125,100–125,764) and the second exon (1,666 nt, 125,908–127,574). This isoform seems to be the product of alternate splicing using the same donor and branching sites as in ter1-i2 with an alternative acceptor site located 13 nt downstream of the ter1-i2 acceptor site. Similar to the ter1-i2 isoform, its abundance is increased in the total fraction relative to the polyadenylated fraction, with logFC values approximately 3 and 2.4, respectively (Table 1). Finally, the fourth reconstructed transcript, ter1-i4, has a length of 2,045 nt, composed of two exons of 363 bp (125,100–125,462) and 1,682 bp (125,893–127,574) and one intron of 430 bp, making this the largest intron of the three interrupted isoforms (125,463–125,892); it is reported by the program as having the same donor, branching and acceptor sites for splicing as the KIS68596 transcript (Fig 3A and 3B) and has notably higher expression in the total fraction than in the polyadenylated fraction, with logFC values of approximately 6.5 (Table 1).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Differential expression analysis of ter1 isoforms.

https://doi.org/10.1371/journal.pone.0281251.t001

Regarding the transcription start sites of the isoforms, it is tempting to say that they share the same site upstream of the start codon of the uncharacterized gene UMAG_03168. Read mapping and differential expression analysis shows that the processed isoforms, mainly ter1-i4, could be included among the large quantities of nonpolyadenylated transcripts, which is observed as a nonuniform coverage of reads aligned in the ribodepleted fraction, with a notable increase in a region of approximately 1,280 nt (Fig 3A); and these continuous isoforms are probably reconstructed as a consequence of the existence of both polyadenylated and nonpolyadenylated populations of transcripts as in TLC1 [67], that coincide with the patterns observed in stable introns excised from mRNAs [68], so that the increase in abundance suggests that this region corresponds to the stable/mature nonpolyadenylated transcript, and the presence of CE1–CE3 in the reconstructed transcripts (Fig 3A and 3B), reinforces the idea that these elements are surely structural domains. The ter1-i4 could correspond to the precursor transcript of the mature form of TER, this idea and the importance of this particular isoform are reinforced by the notable increase in its expression relative to the levels of various housekeeping genes (Fig 3C).

To confirm the finding of polycistronic-transcript isoforms, a series of primers for reverse transcription and PCR assays were designed, all aligning upstream of the putative Sm site; likewise, to distinguish transcript size changes among polyadenylated mature and non-polyadenylated transcripts, Oligo(dT)₁₈ was used to accomplish the reverse transcription from total RNA samples from 521 strain (Fig 4A and S3 Table). We used either Oligo(dT)₁₈, random primers, or specific reverse primers WT-ter1-Low and 5ter-rev to synthesize the first cDNA strand, then the subsequent PCR assay were attained using ter-ALL-5a and Template-Low primers produced the expected 469 bp fragment, corroborating the transcriptional activity of the region (Fig 4B and 4B’, upper panel). Moreover, despite we did not find an oligo primer-pair quite useful to analyze qRT-PCR isoform abundances, we observe slight decreases in amplification products when Oligo(dT)₁₈ was used for reverse transcription; that is why, it was tempting to correlate those results with a poor abundance of polyadenylated TER RNAs as described elsewhere before [67]. Using the ter-i4D-5a forward primer, which was designed to align only the ter1-i4 spliced site, we obtained the expected 637 bp fragments, and no significant amplification was detected when genomic DNA was used as a template. These experiments confirm that the ter1 transcription start site is in the UMAG_03168 locus and that the synthesized transcripts are alternatively spliced. When Oligo(dT)₁₈ was used for first-strand cDNA synthesis, the amount of amplified DNA was reduced (Fig 4B’, lower panel), suggesting that only a minor fraction of the spliced transcripts was polyadenylated; the results back up the transcript would be considered polycistronic in nature.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 4. Confirmation of transcription from the ter1 locus occurs.

(A) Upper panel. Schematic representation of ter1 locus and alignment sites of primers used for RT-PCR assays located upstream of the putative Sm site (red asterisk) and polyadenylation signals (green asterisks), reverse primers used for cDNA synthesis are represented by blue arrows. (B–D). Combination of primer pairs used and expected amplicon sizes for each combination. (B’–D’) Agarose gel electrophoresis of PCR products obtained from cDNA synthesis with the reverse primers indicated at the top of each lane, genomic DNA was used as a positive control. (B and B’). Confirmation of transcription of the ter1 locus containing at least one fragment from the 3’ end of UMAG_03168 and of the existence of the ter1-i4 isoform, evidenced by the presence of a 637 bp fragment unable to amplify in the positive control, top panel, and bottom panel, respectively (C, C’ and D, D’). The long ter1-i1 isoform allowed the amplification of the 418 bp fragment with the ter-i1A-5a and ter-i4D-3b primer pair and revealed at least two isoforms processed by alternative splicing. (E) A 405 bp amplified fragment of α-Tubulin transcript was used as a gene expression control.

https://doi.org/10.1371/journal.pone.0281251.g004

To bring out additional isoforms produced from the ter1 transcripts, Oligo(dT)₁₈ and 5ter-rev1 were used for first cDNA strand synthesis; then, a series of DNA polymerization reactions using swapped primer pairs was achieved (Fig 4C and 4C’). The combination of the forward ter-i1A-5a, which anneals to a ter1-i2 and ter1-i3 left splicing border, and the reverse primers 5ter-rev2 and 5ter-rev1 reveals the existence of at least one spliced isoform within the expected size for ter1-i2 and ter1-i3; again, the polyadenylated fraction seems to be scarce (Fig 4C’ bottom panel). The ter-i1A-5a and ter-i1A-3b primer pair is capable of amplifying fragments either from the nonspliced intron of UMAG_03168 transcripts or from the polyadenylated ter1-i1 isoform; the predicted 134 bp products were amplified when Oligo(dT)₁₈ was used for first strand DNA synthesis (Fig 4D’ first lane). The same fragment was amplified when specific oligo 5ter-rev1 was used for the first strand synthesis of cDNA, but the predicted fragment could be amplified from both polyadenylated and nonpolyadenylated isoforms ter1-i1 (Fig 4D’ second lane). Finally, using the ter-i1A-5a and ter-i4D-3b primer pair, we amplified three cDNA fragments, one with the predicted size for the long ter1-i1 isoform transcribed from the UMAG_03168 locus to downstream and two from spliced isoforms with the alternative acceptor sites (Fig 4D’ bottom panel).

ter1 gene disruption

One-step gene disruption of ter1 was carried out by homologous recombination in the WT 518 strain using a conventional cassette assembled as described previously. The strategy is depicted in S4A Fig: a 649 bp fragment comprising the template domain was replaced by a 2,045 bp BamHI fragment of pCM1007 containing the chimeric hph (S4B Fig). Four hygromycin-resistant transformants were selected, as they recapitulate the features of trt1Δ telomerase-negative mutants, i.e., small, dry, slim dark-pigmented colonies with defined, slightly wavy and raised borders [32]. Mitotic stability was reached at ∼72 generations, and total DNA was extracted from each transformant clone, and the replacement of the template sequence was confirmed by PCR assay.

To verify the gene disruption, the oligonucleotide primer-pair WT-ter1-Upp/WT-ter1-Low were employed to amplify the 3,758 bp fragment spanning the selectable marker gene plus adjacent sequences in ter1-disrupted strains; a 2,362 bp DNA fragment matching the endogenous WT locus was found in strain 518 (S4A and S4C Fig). Stable insertion of the hph gene was further confirmed by PCR amplification of 1,757- and 3,004-bp DNA segments spanning from the middle of the hph gene to 5’ and 3’ boundaries, respectively, outside of the flanking sequences used to construct the gene disruption cassette (S4B Fig). The expected sizes were amplified from ter1Δ mutants, here after named ter1-02, ter1-24, ter1-35, ter1-40 strains, but not from WT 518 and negative controls (S4C Fig). Finally, using the primer pair Template-Upp/Template-Low, amplification of a DNA fragment of 244 bp comprising the template domain was obtained from the wild-type strain but not from the negative control or disrupted transformants (S4A and S4D Fig). Thus, we conclude that the ter1 locus was disrupted.

Characterization of ter1Δ mutants

Once the identity of the selected mutants was verified, the occurrence of progressive telomere shortening was examined by extracting DNA samples from a series of continuous subcultures. Every sample was taken after approximately 24 duplication rounds from the prior and used to analyze the terminal restriction fragment (TRF) of chromosomes. As seen in Fig 5A, we compared wild-type vs. mutant strains at approximately 108 replication rounds (lanes 1 and 2). Both samples corresponded to a 6^th subculture. At that timepoint, in the mutant strain, telomeres were noticeably shortened compared with those of the control. Samples were also taken at the 7^th and 8^th subcultures (132 and 156 replication rounds, respectively; Fig 5A, lanes 3 and 4, respectively), and at replication round 252, a PstI DNA fragment of 350 bp that matched the rumT repeat plus short telomere sequences could be amplified from Type I survivors; in addition, radiography revealed stronger hybridization signals from sequences of ∼700 bp and ∼1,200 bp that could correspond to multiple series of rumT or even rearrangements of UTASa sequences (currently under study).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 5. Phenotypic alterations caused by the interruption of ter1 in U. maydis.

(A) TRF analysis of the WT 518 strain and ter1 mutants. The loss of TER triggers a progressive shortening of TRF and eventually leads to the emergence of surviving cells showing the amplification of a 350 bp fragment. (B) Serial 10-fold dilutions of cultures were spotted on YEPS or MC plates and incubated for 2 days at 28°C; ter1 mutants showed dramatic decreases in colony formation and colony size, consistent with the loss of telomerase activity described in other organisms. (C) Cell morphology of strain WT 518 and mutants ter1-02 and ter1-24. The interruption of ter1 causes alterations in cell morphology, and the mutants have irregularly edged cells and an increase in the presence of elongated cells with alterations in the budding angle, suggesting defects in cell cycle regulation. (D and E) Growth kinetics of the WT strain and ter1 mutants. Strains were grown in YEPS broth (28°C) and sampled every two hours. Samples were used for measurements of OD₆₀₀ or were diluted and plated. Colonies were counted after 24 h of incubation at 28°C and plotted. The results of triplicate assays are shown.

https://doi.org/10.1371/journal.pone.0281251.g005

The growth kinetics of the two survivor strains were calculated as described previously. The growth of sporidia in culture in at least three batches was plotted as OD₆₀₀ vs. time and as CFU/ml vs. time (Fig 5D and 5E). A significant (p = 0.05) delay in cell proliferation was noted in the ter1-2 and ter1-24 cultures in comparison with 518 cultures; the wild-type strain reached stationary phase after 12 h at 30°C, whereas the ter1-02 and ter1-24 mutants continued to replicate (Fig 5D). When log₁₀ (CFU/ml) was plotted vs. time, the decrease in cell viability (p = 0.01) from the start point was clearer in both disrupted mutants than in strain 518 (Fig 5E). Taking into account the log₁₀ (CFU/ml) vs. time from 2 h to 12 h, linear regression analysis gave R² > 0.98, allowing us to determine the generation times in our experiments as 117 min for strain 518, 176 min for strain ter1-02, and 205 min for strain ter1-24. The combination of growth delay and deficiency in replicative potential should cause a dramatic diminishing of the colony appearance. Indeed, after 1:10 serial dilutions of cultures grown to OD = 1.0, differences in both the number and size of the colonies were observed for both mutants, with the most dramatic diminution of growth potential observed in the ter1-24 strain, regardless of the medium used (for example, YEPS or MC) (Fig 5B). The cellular morphology of strain 518 and in its two ter-disrupted mutants, ter1-02 and ter1-24, was also examined under a light microscope. Mutants showed elongated sporidia with slender irregular contour, vacuolation and, to a lesser extent, alterations in the growth polarity and angle of the bud and mother cells (Fig 5C). The elongated shape and perpendicular buds suggested to us that cell cycle activity was defective in these telomerase-negative ter1Δ mutants. Other infrequent alterations could also be observed.

Loss of the ter1 gene also causes chromosome instability

The electrophoretic karyotype of wild-type strain 521 was obtained and compared with that of both types of telomerase-negative mutants of U. maydis. Whereas the TERT mutant trt1-1 (Type I survivor) showed increased chromosome size possibly to extensive TAS + telomere overamplification (currently under study), the trt1-2 (Type II survivor) exhibited moderate lengthening of some chromosomes and visible rearrangements of others (Fig 6A). On their own, TER mutants gave rise to only Type I survivors; of these, ter1-02 shows gross chromosomal rearrangements or even extensive smearing such that individual chromosome bands were poorly distinguishable; the ter1-24 karyotype was very similar to wild-type strain (Fig 6A). Additionally, the cells were stained with propidium iodide and calcofluor white as described in the Materials and Methods and examined under optical and epifluorescence microscopy (Labophot-2, Nikon, Japan). The staining of nuclei with propidium iodide also allowed us to record the percentages of budding sporidia at 90, 45 and 0° as well as the percentage of nuclear aberrations occurring in telomerase-negative trt1-disrupted and ter1Δ strains compared with the same events occurring in wild-type cells. The results are plotted in Fig 6B and 6C. As in trt1Δ mutants, we observed two phenotypes that, as will be discussed, are related to the cells’ telotypes. The WT sporidia showed a rod-like shape (13 μm x 3 μm) with a smooth contour, budding at ∼30–45° from the mother length axis, and poor vesicle content, as has been described elsewhere [69]. The ter1Δ cells included a significantly elevated percentage (p = 0.5) of lengthened and mononucleated cells (Fig 6B) and even septate sporidia, with some septum lacking nuclei, which suggested possible cell cycle arrest or even nondisjunction of chromosomes. The TERT-disrupted mutants described by Bautista-España et al. [32] showed elongated cells, with minor disturbances in budding angle compared with ter1Δ mutants (Fig 6B), while ter1 mutants show an increase in the formation of nuclear aberrations and chromosome bridges (Fig 6C).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 6. Electrophoretic karyotype and percentage of cells exhibiting nuclear aberrations and alterations in the budding angle.

(A) Electrophoretic karyotype of WT 521 strain and four telomerase negative mutants trt1-1 and trt1-2 are TERT-disrupted mutants (trt1 Δ) and ter1-02 and ter1-24 are TER-disrupted mutants (ter1Δ); trt1-1 and ter1-02 strains (lanes 2^nd and 4^th respectively) show increases in chromosome size, possibly due to UTASa amplification, both strains also exhibit fuzzy bands along the entire lane possibly associated to DNA DSBs, which coincided with the formation of nuclear aberrations; in contrast, the trt1-2 and ter1-24 strains had a similar karyotype to that of the WT strain. (B) trt1-disrupted mutants also show slight changes in the budding angle, which in the ter1-disrupted strains is noticeable, mainly as polar growth, perhaps as consequence of alterations in cell cycle regulation. (C) Interruption of ter1 triggers an increase in the formation of nuclear aberrations and chromosome bridges, suggesting DNA damage response system; in contrast, trt1 mutants exhibit an only slight increase in the formation of these structures.

https://doi.org/10.1371/journal.pone.0281251.g006

Plant infection assays

The time course of the infection in maize seedlings was monitored for 21 days, and the results are shown in Fig 7A. In both crosses, chlorosis appeared in the leaves, and at approximately 5 dpi, chlorosis percentages of 48%, 36% and 37% were recorded for the positive control and for heterozygous crosses with strains ter1-02 and ter1-24, respectively (Fig 7A). At 21 dpi, 89% of the controls presented symptoms of infection, in contrast to 45% and 47% for heterozygous ter1^– crosses, in which only the appearance of chlorosis was observed, accompanied in some cases by a slight twisting of the leaves (Fig 7A and 7B). The anthocyanin formation during the development of the infection was infrequent and subtly perceptible in the heterozygous crosses (Fig 7D). Tumor development was not observed, and approximately 47% of plants in both cases did not develop symptoms of infection, in contrast to the plants inoculated with the crosses of wild strains, which at that time showed tumors developed on all the aerial tissues of the plants, 35% of which corresponded to mature tumors with teliospore content (Fig 7A and 7E). Thus, it is concluded that TER plays a crucial role in the development of U. maydis infection and the completion of the life cycle.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 7. Development of infection in maize plants.

(A) Quantification of symptoms in maize plants infected with the indicated crosses at 21 dpi. The mean values of three independent experiments are shown. (B) Details of the leaves and the stem at 21 dpi of the plants inoculated with the wild crosses and heterozygous crosses ter1⁺/ter1^–. (C and D) Stem detail and anthocyanin formation at 21 dpi of plants inoculated with the wild crosses and heterozygous ter1^+/ter1^– crosses, respectively. (E) tumor developed by wild crosses at 21 dpi.

https://doi.org/10.1371/journal.pone.0281251.g007