Schematic representation of the Set1 and subunit major interactors identified in the systematic yeast two-hybrid screens.

A) Schematic representation of Set1 FL and Set1 fragments 1-754 and 754-1081. B) Set1 FL, 1-754 and 754-1081 major Y2H interactors). C) Interacting regions of Set1 with Swd2, Shg1 and Spp1. D) Swd2, Spp1, Shg1 major Y2H interactors. E) Swd1, Swd3, Bre2, Sdc1 major Y2H interactors. The term “interactor” is used to mean a high confidence two-hybrid interaction, with the limitations that this entails. The color reflects the Predicted Biological Score (see METHODS). Red, highest confidence; Blue, high confidence; Green, good confidence.

Set1 1-754 interacts with RGG proteins and the importin Kap104.

A) RGG proteins and import/export proteins interacting with Set1 1-754, Set1 754-1081 and Spp1, Shg1, and Sdc1. B) Set1 interacting domain (SID) (blue) within Kap104. Heat like repeat 9 and 10 are represented in purple C) PY-NLS in the N-terminal region of Set1. D) SID (blue) and RGG motif (green) within the RGG proteins. The interaction domains indicated represent the minimal overlapping DNA sequence present in multiple independent Y2H interacting clones of the same gene. Each genomic fragment of a Y2H clone was analyzed, and the shared overlapping region for given gene was determined to be the only common element among all interacting clones. As such, this region represents the minimal sequence required for interaction.

Chromatin regulators identified in all the two-hybrid screens.

The lines indicate the individual proteins involved in the Y2H interaction. Red line refers to a very high confidence Y2H interaction. All Y2H interactors are described in Table S2. Interactors are grouped according to the complex to which they belong.

SET1C interacts in vitro with Snf2C-AT-hook.

A) A schematic diagram depicting the domains of Snf2 and the Snf2 fragments used in this study, along with the SDS-PAGE/Coomassie staining of the purified GST-tagged Snf2 fragments. B and C) GST pull-down assay using purified GST-tagged Snf2 fragments. The purified SET1C (B) or SET1C-C762 complex (C) was mixed with GST-tagged Snf2 fragments, followed by GST pull-down, and the bound proteins were analyzed by immunoblotting. D) A schematic diagram illustrating Snf2 fragments with a more detailed breakdown of the AT-hook domain, along with the SDS-PAGE/Coomassie Blue staining of the purified Snf2 fragments. The lysines present in the AT-hook are represented by the letter K. E) GST pull-down assay using purified GST-tagged Snf2 fragments and SET1C.

The SET1C Y2H interactome identifies proteins involved in RNA biogenesis.

All Y2H interactors are described in Table S2. The processes linked to RNA metabolism in which the different interactors are involved are shown in the figure. The green lines linking Prp22 to Set1FL/Set1 754-1080 and Prp8 to Set1 FL and Spp1 indicate interactions with a high degree of confidence.

Set1 is SUMOylated.

6His-SUMO–conjugated proteins were purified from cells transformed (+) or not transformed (−) with a plasmid encoding 6His-SUMO under control of the CUP1 promoter. Cell lysates (Input) and Ni-purified material (Elutes) were analyzed by Western blotting with an anti-MYC antibody (A) or and anti-GAL4 antibody (B-D). Analysis of 6His-SUMO -conjugated forms of (A) genomically MYC-tagged Set1 or (B) GB-Set1 transformed cells was performed (left panels), in both the cases SUMO expression and efficiency of purification were controlled using an anti-SUMO antibody (right panels). C) SUMOylation analysis of Set1 fragment F3+F4 (aa. 351-956) WT and the K769R mutant. D) SUMOylation analysis of Set1 fragment F5 (aa. 956-1080) WT, single mutants K1055R and K1060R, and the double mutant K1055R/K1060R mutant.

Nrm1 is methylated in vitro by SET1C.

A) Schematic representation of the S. cerevisiae Nrm1 protein. Nrm1 carries a D-box sequence at its amino-terminus and a central domain that mediated interaction with the Swd1 subunit of SET1C in a yeast two-hybrid screen. A H3K4 like domain that closely resembles the modification site of SET1C in histone H3 is contained within the Swd1 interaction domain. Identical and similar amino-acid positions are shown in italic. Lysine 118 (K118) of Nrm1 aligns with lysine 4 of histone H3; also indicated are K116R and K118A mutant sequences. B) In vitro methylation reactions using partially purified SET1C, 3H-SAM and the indicated peptides. Radioactive reaction products retained on Whatman P81 filter paper following extensive washes were measured by scintillation counting. Each reaction was done in triplicate and error bars indicate the standard deviation. C) In vitro methylation reactions as in (B), however, core histone (2, 4 and 6 mg) and recombinant MBP and MBP-Nrm1 fusion protein (0.5, 1 and 2 mg) were tested as substrates for methylation. Reaction products were separated on 4 to 12% NuPAGE gels and analyzed by fluorography. D) In vitro methylation reactions as in (C), however, core histone (2 mg) and recombinant MBP-Nrm1 wild-type and K118A and K116R fusion protein (0.5, 1 and 2 mg) were tested as substrates for methylation. (upper panel). Note that the molecular mass marker proteins (MM) have been resolved together with the core histone methylation reaction in the same lane. The asterisk indicates migration of a non-specific background signal of unclear origin. The substrate proteins included in the reactions were analyzed on a parallel gel and stained with colloidal Coomassie G-250 (lower panel). E) H3K4-like proteins. Shown are selected proteins containing sequences similar to the modification site found in histone H3. An exhaustive list can be found in Table S4.

Snf2 is methylated in tandem AT-hook domain by reconstituted Set1C.

A) In vitro methyltransferase assay using purified SET1C and Snf2 fragments. H-SAM was used as a methyl-donor and methylated proteins were detected by autoradiography. The band marked with a red star is a degradation product of Snf2C. B) In vitro methyltransferase assay using SET1C and two Snf2-AT-hook fragments with two different tags. C) Schematic diagram of N-terminal truncated SET1 complexes (left) and in vitro methyltransferase assay with GST-Snf2-AT-hook and truncated SET1 complexes. D) Schematic diagram showing the positions of all lysines in Snf2-AT-hook and the further cleaved fragments of Snf2-AT-hook. The red box indicates the two lysines that are acetylated by Gcn5. E) Coomassie staining of purified Snf2 fragments (lower) and an in vitro methyltransferase assay using these fragments with SET1C (upper). F) In vitro methyltransferase assay by SET1C when each of the four lysines in the C-terminal region of the Snf2-AT-hook is substituted with arginine or when both lysines known to be acetylated by Gcn5 are substituted. G) A schematic diagram showing the position of the RG-repeat region and the design of Snf2-AT-hook with RG-repeat truncation. H and I) GST pull-down assay (H) and in vitro methyltransferase assay (I) using purified SET1C and GST-Snf2-AT-hook with or without RG-repeats.

The arginines in the RG-repeat of Snf2 are methylated by reconstituted SET1C.

A) A diagram showing the WT Snf2-B3 fragment and the Mut Snf2-B3 with all four lysines substituted with alanine. B) Coomassie staining of purified WT and Mut Snf2-B3. C) Mass-spectrometry experiment design to identify the methylation sites of Snf2-B3. D) Coomassie staining of Mut Snf2-B3 after methylation reaction and an additional purification step using Ni-NTA. The band corresponding to Mut Snf2-B3 (∼18 KDa) was excised and used for mass spectrometry analysis. The ∼37 KDa band observed in lanes 5–8 appears to be a SET1C subunit that binds non-specifically to Ni-NTA, likely SWD2 based on its size. E) Mass spectrometry analysis result of Snf2-B3 methylation sites revealed that multiple arginines in the RG-repeat were methylated. Amino acid sequence of Snf2 highlighting the arginines methylated by SET1. The sequence of the B3 fragment is shown in bold, and the arginines methylated by SET1 are marked in red. Methylation and demethylation are denoted as M and DM, respectively.

The arginines of the motif ARTSTRGR within the AT hook are methylated in vivo in a Set1-dependent way.

A) Set1 interacts in vivo with Snf2 and Snf2-ΔRG. Myc-tagged Snf2 and Snf2ΔRG were immunoprecipitated with 9E10 Myc antibodies (see Methods) and revealed with either 9E10 (Upper panel) or Set1 antibodies (lower panel). B) Snf2C complex was purified from WT and set1Δ strains, separated on a 4-12% Bis-Tris Gel and Silver Stained (Left); the presence of Snf2-GFP is detected by Western-blotting with anti-GFP antibody (Right). The area corresponding to Snf2-GFP was excised from the gel and used for mass spectrometry analysis. Peptides flanking the RG repeats (1485-1549) with their PTM are shown in Fig. S14. C) Panel C show a focus on the amino sequence flanking the RG repeats. The positions of residues from D1485 to V1549 are indicated on the figure. Peptides identified after digestion of Snf2-GFP are indicated in color with their identified PTM indicated by the color code shown in the top of the panel. The small numbers above the amino acids indicate the probability of localisation according to the MS2 peaks. It should be noted that for wild-type K1488 and R1490, a peptide with dimethylation is detected, but no discriminating MS2 peak allows us to conclude whether K or R are dimethylated. This is also the case for dimethylation on R1528 and R1535. The results presented represent the observed PTMs from two independent experiments, each containing 3 replicates (Table S6).