Elf1 is a conserved member of a zinc finger family of uncharacterized proteins. Homologues of Elf1 were identified by standard protein BLAST analysis. Full-length homologous protein sequences from several eukaryotic organisms were then aligned using ClustalW. Identical residues are noted by an asterisk, conserved substitutions are noted by a colon, and semiconservative substitutions are noted by a period. The four conserved cysteines of the zinc finger domain are noted by arrows above the alignment. Abbreviations: YEAST, Saccharomyces cerevisiae; SCHPO, Schizosaccharomyces pombe; HUMAN, Homo sapiens; MUS, Mus musculus; CAEEL, Caenorhabditis elegans; DROME, Drosophila melanogaster; ARATH, Arabidopsis thaliana.

The elf1Δ mutation causes mild sensitivity to 6AU and genetically interacts with dst1Δ. Yeast strains were spotted onto plates in a series of 10-fold dilutions. Strains were grown on SC-ura or SC medium for 2 days, medium containing 6AU (50 μg/ml) for 5 days, or medium containing MPA (20 μg/ml) for 3 days.

elf1Δ genetically interacts with mutations in the genes encoding the transcription elongation factors Spt4, Spt5, Spt6, and the Paf1 complex. (A) Deletion of elf1Δ is lethal in combination with mutations in SPT4, SPT6, or Paf1 complex genes. elf1Δ strains were transformed with plasmid pDP1 containing ELF1⁺ on a CEN URA3 plasmid and mated to spt4Δ, spt6-1004, cdc73Δ, or paf1Δ strains. Diploids were sporulated and dissected, and representative spores from the resulting complete tetrads were patched onto YPD. Patches were replica plated to medium lacking uracil for 1 day and medium containing 5FOA for 2 days. (B) Deletion of elf1Δ suppresses spt5-242 phenotypes. spt5-242 is a Cs⁻ Spt⁻ mutation that can be suppressed by mutations in Pol II or other elongation factors that impair elongation (21). elf1Δ strongly suppresses the Cs⁻ phenotype and modestly suppresses the Spt⁻ phenotype of spt5-242. The Spt⁻ phenotype was scored by observing growth on medium lacking lysine and using the lys2-128δ allele. Yeast strains were spotted in a series of 10-fold serial dilutions onto YPD for 2 days at 30°C, YPD at 14°C for 9 days, SC for 2 days, or SC-lys for 4 days.

Elf1 is preferentially localized to transcribed regions of active genes. (A) Elf1 is localized to the transcribed region of PMA1. Yeast cells containing the Elf1-13Myc and Rpb3-HA epitope-tagged proteins were grown at 30°C to mid-log in YPD medium. Cells were cross-linked with 1% formaldehyde and lysed, and the chromatin was sheared by sonication to an average size of 400 bp. Elf1-13Myc was immunoprecipitated using the A14 anti-Myc antibody, and Rpb3-HA was immunoprecipitated using the 12CA5 anti-HA antibody. Radiolabeled PCR products were obtained with primers specific to the PMA1 UAS, TATA, 5′ ORF, 3′ ORF, or 3′ UTRs and were visualized on 6% polyacrylamide gels. As a control for DNA binding specificity, each PCR contained primers to amplify a nontranscribed region of chromosome V that is not contained in any open reading frames (28). ChIP values are represented as the fold enrichment of the specific PCR product compared to this control region. The locations of primer sets used for ChIP analysis are represented as lines below the schematic of the PMA1 gene. Results were obtained for at least three independent experiments, and the average and standard error are plotted on the graph. (B) Elf1 is localized across the GAL1 transcribed region upon transcriptional induction. Yeast cells containing Elf1-13Myc and Rpb3-HA epitope-tagged proteins were grown at 30°C to mid-log in medium containing 2% raffinose. Transcription of GAL1 was induced by the addition of 2% galactose for 2 hours. Chromatin immunoprecipitations were performed as described for panel A in at least three independent experiments, with the resulting averages and standard errors plotted on the graph.

Elf1 localization to PMA1 is partially dependent on the Spt4 and Spt6 elongation factors. (A) Elf1 localization at PMA1 in several elongation mutants. ChIP experiments were performed as described for Fig. 4A for wild-type, spt4Δ, rtf1Δ, paf1Δ, spt6-1004, or spt16-197 strains at 30°C. ChIP results for Elf1-13Myc (top panel) and Rpb3-HA (bottom panel) are shown. The results from at least three independent experiments are shown. (B) Localization of Spt4 and Ctr9 to PMA1 is not affected in an elf1Δ mutant. Wild-type or elf1Δ strains containing Spt4-Flag, Ctr9-Myc, and Rpb3-HA were subjected to ChIP analysis as described for Fig. 4A. Ctr9-Myc was immunoprecipitated using the A14 anti-Myc antibody, and Spt4-Flag was immunoprecipitated using the M2 anti-Flag antibody. The results from at least three independent experiments are shown.

Elf1 is necessary for maintaining proper chromatin structure in actively transcribed regions. (A) elf1Δ mutations allow use of a cryptic TATA element within the GAL1-FLO8-HIS3 reporter. Wild-type and elf1Δ strains containing the GAL1-FLO8-HIS3 reporter construct were patched onto YPD medium. Strains were then replica plated to synthetic complete medium, medium lacking histidine and containing glucose, and medium lacking histidine and containing galactose as the carbon source. Strains were scored for growth at 30°C for either 1 or 3 days. (B) Mutations in elf1Δ are Ts⁻ in combination with rtf1Δ or hir1Δ mutations. elf1Δ strains were crossed to rtf1Δ and hir1Δ mutants. Diploids were sporulated and dissected, and representative spores from the resulting complete tetrads were spotted onto YPD medium at 30°C or 39°C for 2 days. (C) elf1Δ hir1Δ mutants allow transcription to initiate from within the endogenous FLO8 coding region. Wild-type, elf1Δ, hir1Δ, elf11Δ hir1Δ, rtf1Δ, and elf1Δ rtf1Δ strains were grown in YPD to mid-log at 30°C. Strains were shifted to 39°C by addition of an equal volume of YPD medium prewarmed to 48°C. Strains were then incubated at 39°C, and cells were collected at 2 and 4 h after the temperature shift. Northern analysis was performed using probes to the 3′ region of FLO8 (+1515 to +2326) and also to the SCR1 loading control.

Purification of Elf1. Purification of Elf1 was carried out with strains containing either an untagged protein or a TAP-tagged version of Elf1. The protein complex was purified and was then analyzed by SDS-PAGE and silver staining. Elf1 and the two largest subunits of CKII (Cka1 and Ckb1) were identified by trypsin digestion and MALDI-TOF. In the mass spectrum, there were 6 (35% coverage), 16 (43% coverage), and 10 (39% coverage) peptides used to identify Elf1, Cka1, and Ckb1, respectively. The two smaller subunits of CKII, which comigrate with Elf1-TAP, were identified (along with Cka1, Ckb1, and Elf1) by using gel-free LC-MS/MS. The + indicates the position of TEV protease. Other detectable bands are present in both lanes, including those at around 70 kDa that are likely the heat shock proteins Ssa1 and Ssb1, present in most purifications (N. Krogan and J. Greenblatt, unpublished results).