Protein arginine methylation has emerged as a key regulator in protein function in eukaryotes. This modification alters the biochemical properties of the arginine side chain and is catalyzed by members of the protein arginine methyltransferase (PRMT) family of enzymes (Bedford and Richard, 2005, Lorton and Shechter, 2019). These enzymes are categorized into four subtypes based on the specific type of methylarginine that is formed. Type I PRMTs transfer one or two methyl groups from S-adenosyl-L-methionine (AdoMet) to a single guanidino nitrogen on a protein-incorporated arginine residue, producing either monomethylarginine (MMA) or asymmetric dimethylarginine (aDMA), respectively. Type II PRMTs also catalyze monomethylation, but subsequently add a second methyl group from AdoMet to the opposite guanidino nitrogen, thereby forming symmetric dimethylarginine (sDMA) (Morales et al., 2016). In the budding yeast Saccharomyces cerevisiae, the enzyme Hmt1 is the homologue of mammalian PRMT1 and is the only known PRMT that can catalyze ADMA formation in the organism (Low and Wilkins, 2012). Hmt1 is responsible for the formation of 89 % of aDMA and 66 % of MMA in vivo (Gary et al., 1996, Tang et al., 2000). Hmt1 preferentially methylates arginine residues in the context of RXR/RG/RGG motifs (Hamey et al., 2018), but can methylate arginyl residues in other sequences (Wooderchak et al., 2008). In the crystalline state, Hmt1 adopts a hexameric structure composed of three dimeric units (Weiss et al., 2000). Dimerization of Hmt1 is essential for its function, as disruption of dimerization leads to a loss of methylation activity (Weiss et al., 2000). Based on structure function studies, it has been suggested that transitions from dimer to hexamer may be critical in regulating the activity of Hmt1 (Weiss et al., 2000). Previous large-scale and targeted studies have identified approximately 40 substrates of Hmt1 in vivo, with RNA-binding proteins representing the largest class of identified substrates (Jackson et al., 2012, Low et al., 2016, Yagoub et al., 2015).

Given the broad spectrum of substrates modified by Hmt1, it is important to understand the mechanisms that regulate Hmt1 activity. Post-translational modification (PTM) represents a widespread mechanism for regulating protein function. A previous in-depth proteomic characterization of Hmt1 identified an N-terminal PTM “hotspot” harboring acetylation and phosphorylation sites (S2, K3, T4, K7, and S9) (Winter et al., 2018). In a separate study by Messier et al., serine phosphorylation at position 9 (S9) on Hmt1 was implicated in a nutrient-response pathway that promotes proper entry into M-phase during the cell cycle (Messier et al., 2013). In this model, phosphorylation of Hmt1 S9 by the kinase Dbf2 is required for maintaining the oligomeric state of Hmt1, which allows for methylation of the SR-/hnRNP-like protein Npl3. The methylated state of Npl3 is critical for stabilizing the B-cyclin mRNA CLB2, as CLB2 mRNA homeostasis governs the timing of M-phase entry during the cell cycle. However, an earlier study demonstrated a truncated version of Hmt1, in which the first 20 amino acids are removed, remains enzymatically active in methylating Npl3 in vivo (Weiss et al., 2000). Furthermore, additional studies have shown that Hmt1 lacks methyltransferase activity when expressed to include a C-terminal TAP-tag (Lacoste et al., 2002), the construct used in the nutrient-response pathway study. These conflicting observations warranted further investigation into the function of Hmt1 S9 phosphorylation, which was the first reported study on how phosphorylation may control Hmt1 function and activities.

In the work presented here, we employ various experimental approaches that bypass the need to express C-terminal epitope-tagged Hmt1 in yeast to elucidate the functional role of Hmt1 S9 phosphorylation. Using two classes of Hmt1 S9 substitution mutants, one being a phosphomimetic (S-to-E; Hmt1-S9E) and the other being non-phosphorylatable (S-to-A; Hmt1-S9A), we demonstrated that these mutations do not impact the oligomeric state nor the ability of Hmt1 to methylate Npl3 in vitro. Additionally, methylation of a second substrate, R3, was not enhanced by the S9E mutation. Utilizing a well-characterized antibody that specifically recognizes only the methylated form of Npl3 (Siebel and Guthrie, 1996, Wilson et al., 1994), we showed that Npl3 remains methylated in dbf2Δ cells, albeit at a reduced level compared to the wild-type. The aDMA banding profile in the dbf2Δ mutant was virtually identical to those observed in the wild-type. A decrease in Npl3 methylation was observed in the rapamycin-treated Hmt1-S9E mutant when compared to the wild-type. The aDMA banding pattern in both Hmt1 S9 mutants closely resembles those in the wild-type. In synchronized yeast cells, no significant alteration was detected in the peak timing of Clb2 protein after release into the media in either Hmt1 S9E or S9A. In vivo, we demonstrated that C-terminally TAP-tagged Hmt1 abolishes its ability to methylate Npl3 and possesses an aDMA banding profile resembling those observed with hmt1Δ. Lastly, we found that the S9 phosphorylation status influences proper Hmt1 levels in vivo. These results support a role for Hmt1 S9 phosphorylation in maintaining the Hmt1 protein levels rather than altering Hmt1 oligomeric state and methyltransferase activity in vivo.