Characterization of the in vivo phosphorylation sites of the mRNA.cap-binding complex proteins eukaryotic initiation factor-4E and p20 in Saccharomyces cerevisiae.

Eukaryotic translation is believed to be regulated via the phosphorylation of specific eukaryotic initiation factors (eIFs), including one of the cap-binding complex proteins, eIF-4E. We show that in the yeast Saccharomyces cerevisiae, both eIF-4E and another cap-binding complex protein, p20, are phosphoproteins. The major sites of phosphorylation of yeast eIF-4E are found to be located in the N-terminal region of its sequence (Ser2 and Ser15) and are thus in a different part of the protein from the main phosphorylation sites (Ser53 and Ser209) proposed previously for mammalian eIF-4E. The most likely sites of p20 phosphorylation are at Ser91 and/or Ser154. All of the major sites in the two yeast proteins are phosphorylated by casein kinase II in vitro. Casein kinase II phosphorylation of cap-complex proteins should therefore be considered as potentially involved in the control of yeast protein synthesis. Mutagenesis experiments revealed that yeast eIF-4E activity is not dependent on the presence of Ser2 or Ser15. On the other hand, we observed variations in the amount of (phosphorylated) p20 associated with the cap-binding complex as a function of cell growth conditions. Our results suggest that interactions of yeast eIF-4E with other phosphorylatable proteins, such as p20, could play a pivotal role in translational control.

The eukaryotic initiation factor (eIF) 1 -4E (eIF-4F␣) is an essential component of the eukaryotic translation apparatus. eIF-4E constitutes part of the so-called cap-binding complex eIF-4F which, in higher eukaryotes, also contains eIF-4A and p220 (1). This complex, together with eIF-4B, promotes binding of the 43S preinitiation complex to mRNA (2,3). The yeast Saccharomyces cerevisiae does not possess a cap-binding complex directly equivalent to mammalian eIF-4F. Instead, yeast eIF-4E has been shown to form a complex with two other proteins, p150 (believed to be the homologue of p220) and p20, whose functions are as yet unknown (4 -7). However, up to now it has been assumed that yeast eIF-4F fulfils the same function(s) as its mammalian counterpart.
The sequencing of several eIF-4E genes has revealed strong homology within the group of known mammalian polypeptide sequences and less extensive, but clearly evident, homology between the yeast eIF-4E sequence and the sequences of the counterpart proteins of mammals and wheat. The mammalian and yeast proteins are immunologically distinct (8). However, mouse eIF-4E (9) can substitute for the yeast protein in vivo. Thus, there is also at least partial functional homology between the various eIF-4E proteins. At the same time, the functional role(s) of eIF-4E is(are) incompletely defined. Certain reports have already indicated that this protein might be directly or indirectly involved in a number of processes other than translation. For example, the yeast eIF-4E gene has been identified as the locus of a cell-cycle mutation (cdc 33) that arrests the mitotic cycle at the "start" stage (10). Moreover, a fraction of the cellular population of eIF-4E in COS-1 cells (11) and in yeast (12) localizes to the nucleus.
One striking property of mammalian eIF-4E is that its overproduction can lead to the transformation of higher cells (13,14). Moreover, increased levels of this initiation factor allow enhanced translation of mRNAs whose leaders bear strongly inhibitory secondary structure (15). On the other hand, the overproduction of yeast eIF-4E has little effect on growth or on translation limited by mRNA structure in S. cerevisiae (12).
Various lines of evidence indicate that the phosphorylation state of eukaryotic initiation factors can influence their activities in translation (2). The most convincing case so far is that of eIF-2. Phosphorylation of the ␣ subunit of this factor by specific kinases at serine 51 in mammals (2) and yeast (16) inhibits the exchange of GTP/GDP and effectively reduces the activity of eIF-2. Phosphorylation of other sites in the C-terminal region of eIF-2␣ by casein kinase II also influences eIF-2 activity, albeit in a more subtle manner (17,18). Increased phosphorylation of eIF-4E, in contrast, seems to correlate with enhanced translational rates (19 -25). Conversely, dephosphorylation has been found to accompany the inhibition of protein synthesis under various conditions (26 -28). However, these are indirect correlations which are insufficient to prove the existence of a causal link between phosphorylation and the activity of eIF-4E. A more direct demonstration of a functional change associated with phosphorylation was reported by Minich et al. (29), who found that the phosphorylated form of mammalian eIF-4E has a 3-fold enhanced affinity for the cap.
What is(are) the site(s) of phosphorylation in eIF-4E? Ser 53 was identified as a residue whose mutation to alanine nullifies certain properties of the wild-type factor (30). Thus, whereas the overproduction of wild-type eIF-4E induced aberrant growth in rodent and Hela cells, this effect was not observed when the Ala 53 mutant was used (13,14). Similarly, overexpression of the wild-type gene, but not of the Ala 53 mutant form, seemed to allow enhanced translation of mRNAs with structured leaders in mouse cells (15). It was also reported that the Ala 53 mutation prevents incorporation of eIF-4E into the 48 S initiation complex (31). Certain lines of evidence have indi-cated that eIF-4E may be phosphorylated by protein kinase C (21)(22)(23)(24), and this enzyme has therefore been assumed to be involved in the regulation of eIF-4E activity.
However, conflicting evidence has thrown doubt on the significance of phosphorylation at position 53 in the mammalian eIF-4E amino acid sequence. It was observed that mutation of Ser 53 to Ala did not affect the phosphorylation level of transiently overexpressed eIF-4E in COS-1 monkey kidney cells (32). These authors also detected no effects of the overexpression of either wild-type or mutant forms of mammalian eIF-4E on translational initiation in this cell type. This therefore suggests that there may be an alternative explanation of the results obtained by others using Ser 53 mutation(s). For example, the Ser to Ala substitution at position 53 may result in a conformational mutant form of eIF-4E with altered properties that are unrelated to the phosphorylation state of this factor. Indeed, it was more recently reported that the major site of phosphorylation in mammalian eIF-4E is Ser 209 (33).
Up to now, there has been no information available about the phosphorylation sites in the yeast cap-binding complex eIF-4F. Given the widely acknowledged significance of S. cerevisiae as an experimental organism for studies of translation, as well as the uncertainty surrounding the results obtained so far with the mammalian factor, it has become especially important to investigate eIF-4F phosphorylation in this lower eukaryote. We have used biochemical and genetic techniques to show that both eIF-4E and p20 in yeast are phosphoproteins and that they are phosphorylated by the same kinase. We examine the functional significance of these phosphorylation events. The results provide a new perspective on the pathway and role of phosphorylation in the regulation of the eukaryotic mRNA⅐cap-binding complex eIF-4F.
Plasmid Constructs, eIF-4E Expression, and in Vivo Labeling-The plasmid pCYTEXP1 (42) was used for synthesis of eIF-4E in E. coli and for isolating single-stranded DNA for in vitro mutagenesis. The yeast eIF-4E gene was inserted into this vector as an NdeI-BamHI restriction fragment (12). Recombinant eIF-4E was obtained from E. coli TG2 as described previously (12,43). Synthesis of eIF-4E and its mutants in yeast was performed using the GPF promoter plasmid YCpSUPEX1 (44) and the TRP1 promoter plasmid YCp332TRP1. The latter plasmid was constructed by substituting a 1.6-kilobase fragment bearing the TRP1 promoter (45) for the GPF promoter in YCpSUPEX1. The derivatives of these plasmids were introduced into the strain 4 -2, after which plasmid shuffling was performed to allow loss of pMDA101. For in vivo labeling, cells were grown in low phosphate medium to an OD 550 ϭ 0.4, centrifuged, and resuspended in fresh medium containing 200 Ci/ml 32 P i . After incubation for 1.5 h at 30°C, the cells were washed and resuspended in buffer (20 mM HEPES, pH 7.4, 100 mM KCl, 0.5 mM EDTA, 1 mM dithiothreitol, 0.5 mM phenylmethylsulfonyl fluoride, 5 mM NaF, 5 mM glycerol phosphate) and vortexed with glass beads. The cell debris was removed by centrifugation and the cap-binding complex proteins were isolated from the extracts by means of affinity chromatography using a 7-methyl-GTP-Sepharose column (Pharmacia).
Isolation of Casein Kinase II and in Vitro Phosphorylation-Casein kinase II was purified from strain 4 -2 (carrying the wild-type eIF-4E gene in YCpSUPEX1) using a protocoll adapted from Padmanabha and Glover (46). Yeast cell extracts were fractionated on a Q-Sepharose column. The fractions possessing casein kinase II activity were pooled and run through a heparin-Sepharose affinity column. Finally, the kinase preparation was concentrated on a fast protein liquid chromatography MonoQ column (Pharmacia). The kinase activity assay was performed using ␤-casein as acceptor according to Glover et al. (47), except that the incubation temperature was 30°C. Phosphorylation assays were also performed using recombinant eIF-4E or cap-binding proteins isolated from yeast instead of ␤-casein.
Phosphoamino Acid Analysis, Phosphopeptide Mapping, and HPLC Analysis-Phosphoamino acid analysis and two-dimensional peptide analysis were performed on thin layer cellulose plates (Kodak) (48,49). For HPLC analysis, 200 l (about 2 nmol) of eIF-4E tryptic peptides were applied to a reverse-phase C 18 column (Waters-Millipore) and eluted over a period of 80 min using a 5-50% acetonitrile gradient (flow-rate ϭ 0.5 ml/min).

RESULTS AND DISCUSSION
eIF-4E and p20 in Yeast Are Phosphoproteins-In order to determine whether yeast eIF-4E is phosphorylated in vivo, we added 32 P i to shaking cultures of log-phase yeast. After a period of incubation of 90 min, cell extracts were prepared and subjected to affinity chromatography. Separation of the cap-binding complex proteins using SDS-polyacrylamide gel electrophoresis and analysis using autoradiography revealed that both p20 and eIF-4E are phosphoproteins, whereby the former was approximately 10 times more strongly labeled than the latter (Fig. 1, A and B). Isoelectrofocussing of labeled extracts yielded various bands in the pH range 5.6 to 6.0 detectable by immunoblotting using eIF-4E-antibody (Fig. 1E). Only one of these bands, slightly more acidic than the major one, could be detected by autoradiography. We therefore conclude that only a small fraction of the total eIF-4E is phosphorylated in vivo. The additional bands to the basic side of the major one may correspond to eIF-4E degradation products, which were routinely observed in preparations of this protein, or to heterogeneity caused by other types of protein modification (compare Ref. 31). Phosphoamino acid analysis showed that the amino acids phosphorylated in p20 and eIF-4E are in both cases serine residues ( Fig. 2A). Mapping the Phosphorylation Sites in eIF-4E-32 P-Labeled eIF-4E was isolated from strain 20B-12, which has a wild-type chromosomal eIF-4E gene, or from strain 4 -2 transformed with the wild-type gene in the plasmid YCpSUPEX1, and subjected to tryptic digestion. The resulting peptides were analyzed using two-dimensional peptide mapping and HPLC. Twodimensional peptide mapping revealed two major and several weakly labeled phosphopeptides (Fig. 3, WT). The more strongly labeled peptide (peptide 1, Fig. 3) migrates slightly further toward the cathode than the neutral marker ⑀-dinitrophenyl-lysine (dnp-lysine). This behavior would be expected of either a positively charged peptide or, in the case of a neutral peptide, might be explained on the basis of endoosmotic effects (48). The second major phosphopeptide (peptide 2, Fig. 3) moves about the same distance as xylene-cyanol toward the anode. This marker has a negative charge at pH 3.5. Other weakly labeled peptides can be seen close to peptide 1. A further peptide migrates close to dnp-lysine (peptide 3, Fig. 3) and is evidently relatively hydrophobic.
HPLC analysis of the tryptic peptides derived from in vivo phosphorylated eIF-4E also yielded two major radioactive peaks, plus four weaker ones (Fig. 4), consistent with the data obtained from two-dimensional peptide mapping. Thus, the above analytical methods indicated the existence of at least two major (serine) phosphorylation sites in yeast eIF-4E. We attempted to identify the phosphorylated peptides in the two major fractions via peptide sequencing. However, the low level of labeling combined with the incomplete separation of the peptides prevented us from obtaining unequivocal data, and we turned to an alternative approach.
Identification of the Phosphorylated Serines in eIF-4E-Analysis of the eIF-4E protein sequence reveals the existence of several potential phosphorylation sites in the N-terminal region of the protein. Ser 7 is located in a theoretical consensus site for both casein kinase I and protein kinase C-like kinases, while Ser 2 lies in a consensus sequence for casein kinase II and Ser 15 in a sequence expected to be recognized by both casein kinase I and II (Fig. 6). An initial crude test of phosphorylation in this region was to make use of two N-terminal deletion mutants which have lost two or more of the initial serines in the amino acid sequence. 2 In vivo labeling was performed using derivatives of 4 -2 bearing each of the deletion mutants. Both of these deletion mutant proteins were able to support translation in vivo. 2 eIF-4EDel7 was found to be less strongly labeled than the wild-type protein, and tryptic peptide analysis revealed that phosphopeptide 2 was no longer present (Fig. 3, Del7). Phosphorylation of the eIF-4EDel19 mutant, moreover, was hardly detectable (Figs. 1, C and D, and 5). In order to define the phosphoserine residues precisely, we constructed point mutants in which one or more of the N-terminal serines has been mutated to alanine(s). Mutation of Ser 7 to Ala (S7A) did not change the labeling pattern observed (Fig. 3, S7A). However, further phosphopeptide mapping analyses of the single mutants S2A and S15A, as well as the double mutant S2ϩ15A, clearly showed that Ser 2 and Ser 15 are the major phosphorylation sites of eIF-4E in S. cerevisiae grown under the described conditions. The major labeled peptides were removed only in the double mutant S2ϩ15A (Fig. 1, C and D and Fig. 3,  S2ϩ15A). Moreover, phosphorylation of the double mutant form of eIF-4E was reduced by more than 80% in vivo (Fig. 5). The remaining spot in the area of peptide 1 observed with the S15A mutant (Fig. 3) was one of the weakly labeled phosphopeptides (see previous section and Figs. 3 and 4) generated by tryptic digestion with a similar mobility to peptide 1. It is also evident in the map of the double mutant (S2ϩ15A).
Identification of the two phosphorylated serines at positions 2 and 15 via mutational analysis also provides an explanation of the observed mobilities of the phosphopeptides in two-dimensional mapping (Fig. 3; compare Fig. 6). These mobilities are consistent with the expected cleavage sites of trypsin in the eIF-4E amino acid sequence (compare Ref 48). For example, phosphopeptide 1 could be generated by tryptic cleavage at Lys 24 and at either Lys 8 (expected net charge ϩ1 at pH 3.5) or  FIG. 2. Phosphoamino acid analysis. A, in vivo labeled eIF-4E and p20 proteins were subjected to partial acid hydrolysis and the phosphoamino acids analyzed by electrophoresis on cellulose plates using a buffer system at pH 3.5. B, phosphoamino acid analysis of proteins labeled in vitro with casein kinase II. R-WT and R-S2ϩ15A, recombinant wild-type and double mutant serine 2 and 15 to alanine eIF-4E, both isolated from E. coli. The lanes labeled with the prefix Y are analyses of proteins isolated from yeast: WT and S2ϩ15A eIF-4E wildtype and double mutant; and p20. The positions of the phosphoamino acids used as standards (P-Ser (P-S), P-Thr (P-T), and P-Tyr (P-Y)) as well as phosphate (P i ) and the origin (OR) are indicated. Partial hydrolysis products are visible in certain lanes. at Lys 9 (neutral). Phosphopeptide 2, on the other hand, could be generated by cleavage at Lys 8 (expected net charge Ϫ1).
We investigated whether there was a growth phenotype associated with the double mutation in eIF-4E. Expression of the mutant gene in strain 4 -2 lacking the wild-type eIF-4E gene allowed growth at a rate reduced by no more than 5% relative to wild-type under standard growth conditions both in YEPD and minimal medium. We also investigated the rates of total protein synthesis in whole cells (measured as [ 35 S]L-methionine incorporation) containing only the S2ϩ15A mutant (data not shown). A decrease of less than 10% compared to the equivalent strain bearing wild-type eIF-4E was detectable during log-phase growth at 30°C or under heat shock conditions (39°C).
It might be argued that the described mutations induce changes in the phosphorylation state of eIF-4E via indirect conformational changes. However, this is unlikely for at least three reasons. First, each of the single mutations, S2A and S15A, eliminates the phosphorylation of one specific peptide, and the combination of both mutations results in an additive effect (Fig. 3), whereas S7A has no effect on the phosphorylation pattern. Second, the results obtained with the deletion mutants Del7 and Del19 are fully consistent with the former observations (Figs. 3 and 5), despite the fact that deletions of this kind would be more likely to induce significant conformational alterations in the overall structure of the protein. Third, the fact that none of the amino acid substitutions have a readily detectable phenotype under log-phase growth conditions indicates that if they do cause conformational changes in the protein these are probably small.
Both eIF-4E and p20 Are Phosphorylated by Casein Kinase II-The two phosphorylated serines identified lie in potential casein kinase II sites. Casein kinase II was isolated from yeast and used for in vitro phosphorylation reactions. Recombinant wild-type and S2ϩ15A eIF-4E proteins were isolated from E. coli and used as substrates. The mutant was weakly labeled compared to the wild-type protein (Fig. 7A). Phosphoamino acid analysis revealed that both proteins were phosphorylated at serine residues (Fig. 2B, R lanes). One of the phosphopeptides (data not shown) generated by tryptic digestion corresponds to the major site of phosphorylation in vivo (peptide 1; see Fig. 3) and is not present in the peptide map of the S2ϩ15A mutant. The other major phosphopeptide observed, which is present in the maps of both proteins, is located at the same position as the weakly labeled peptide obtained from the in vivo phosphorylated protein which is located close to dnp-lysine (see peptide 3, Fig. 3). Unexpectedly, no phosphopeptide derived from wild-type recombinant eIF-4E was observed at the position corresponding to peptide 2.
In light of the above discrepancy between the in vitro and in vivo labeling, we isolated cap-binding complexes from yeast cells containing either the wild-type or the S2ϩ15A mutant and phosphorylated the resulting preparations in vitro using casein kinase II. The two major phosphopeptides 1 and 2 observed in the maps prepared from in vivo labeled wild-type eIF-4E could again be observed in the wild-type, but not in the S2ϩ15A mutant (Fig. 8, panels WT and S2ϩ15A). In conclusion, casein   Fig. 3) are compared to wild-type. The measured values of protein and 32 P incorporation for wild-type eIF-4E (averages of two independent experiments) were each normalized to 1, and the equivalent data obtained with the three mutants were expressed as fractions of the wild-type values. See Fig. 1 for an explanation of the strain nomenclature. Light shaded, protein; dark shaded, radioactivity.
FIG. 6. Confirmed and potential phosphorylation sites in the S. cerevisiae cap-complex proteins eIF-4E and p20. Shown are the N-terminal (indicated by the subscript N) amino acid sequence of eIF-4E, and the two regions of the now corrected version of the p20 amino acid sequence containing potential CK-II sites. The serines that were mutated to alanines in the eIF-4E mutants are numbered, and the positions of the start codons of the deletion mutants eIF-4EDel7 and eIF-4EDel19 are indicated by arrowheads. The serines in the revised p20 reading frame suspected to be phosphorylated by CK-II are also numbered.
kinase II can phosphorylate both Ser 2 and Ser 15 in vitro. The difference in substrate behavior of recombinant eIF-4E from E. coli might be due to an altered conformation relative to that of the natural yeast protein. Alternatively, the association of other yeast proteins, such as p20, with eIF-4E isolated from yeast might influence the accessibility of site 2. The relative strengths of peptides 2 and 3 are reversed in the in vitro and in vivo labeling of yeast eIF-4E (compare Figs. 8 and 3, WT). Ser 2 is evidently poorly phosphorylated in vitro, whereas the third site (peptide 3) seems to be a better substrate.
A striking aspect of the in vitro phosphorylation of the capbinding complex isolated from yeast is that p20 is very rapidly labeled, apparently becoming fully phosphorylated after 5 min of incubation under the conditions described (Fig. 7B). In the process of analyzing the potential sites of phosphorylation in p20, we discovered a discrepancy in the gene sequence compared with that published previously. Our sequencing analysis revealed the existence of an additional A nucleotide in the series of As (370 -374) in the published reading frame (51). The revised reading frame encodes a protein of 161 amino acids and contains two apparent casein kinase II consensus sequences at serines 91 and 154 (Fig. 6). Phosphoamino acid analysis of p20 (Fig. 2), together with peptide mapping of this protein labeled in vivo and in vitro (Fig. 8), indicates that casein kinase II is the major enzyme responsible for p20 phosphorylation in vivo. The in vivo peptide mapping data reveal the existence of at least one major (peptide A), and one minor (peptide B), phosphorylation site in p20. The same major phosphopeptide (A) was also generated by tryptic digestion of p20 that had been phosphorylated by casein kinase II in vitro (Fig. 8).
The Phosphorylated eIF-4F Complex in Vivo: Potential Regulatory Pathways-The phosphorylation state of eIF-4E was analyzed under different culture conditions. The amount of 32 P-labeled eIF-4E in the isolated cap-binding complex was unaffected by heat shock (Figs. 1, A and B, and 9) or the addition of cycloheximide (data not shown). The level of eIF-4E phosphorylation in the stationary phase appears, at first sight, to have increased. However, this change in the specific labeling intensity was primarily due to the depletion of unlabeled phosphate observed under these conditions. More striking was the observation that the amount of phosphorylated p20 associated with eIF-4E was increased under conditions of heat shock or cycloheximide treatment, whereas reduced amounts were observed in extracts from cells in the stationary phase (Figs. 1, A  and B, and 9). The true reduction in the degree of phosphorylation of p20 in stationary phase, for the reason mentioned above, must be significantly greater than that indicated in Fig.  9. We do not know which of the potential phosphorylation sites in p20 is(are) involved in the overall changes in phosphorylation level.  WT and S2ϩ15A, eIF-4E wild-type and double mutant, respectively. p20 in vivo, peptide mapping of p20 labeled in vivo; p20 in vitro, peptide mapping of p20 phosphorylated in vitro using casein kinase II. The positions of the eIF-4E phosphopeptides 1-3, the major (A) and minor (B) phosphopeptides of p20, the standards ⑀-dinitrophenyl-lysine (dnp-K) and xylene-cyanol (XC), and of the origin (f) are indicated. The amounts of eIF-4E or p20 and of 32 P incorporation into these proteins are expressed as values relative to the log-phase growth data (normalized to 1.0).
The results described in this paper show that eIF-4E and p20 in yeast are at least partially phosphorylated in vivo. The level of phosphorylation of eIF-4E detected was particularly low, with most of this factor being present in the non-phosphorylated form. The pattern of phosphorylation was unexpected. The major positions of phosphorylation of eIF-4E are not protein kinase C sites and are not equivalent to the sites reported to be present in the mammalian counterpart protein. Instead, casein kinase II, a serine/threonine kinase ubiquitously present in eukaryotic cells, has been found to phosphorylate both proteins. This kinase has a wide range of substrates, including proteins linked to the cell-cycle, cell growth, and cell differentiation (52). In yeast, casein kinase II is essential for cell viability, and there is evidence for its involvement in cell-cycle regulation (53). For example, casein kinase II phosphorylation of yeast topoisomerase II is cell-cycle-dependent (54). Here, as in yeast eIF-2␣ (17,18), the phosphorylation sites are close to the C terminus of the protein. The main casein kinase II phosphorylation sites of yeast eIF-4E are close to the N terminus, and one of the p20 sites is predicted to be close to the C terminus (Fig. 6). Mutation of the two major phosphorylation sites in eIF-4E results in only a minimal phenotype in yeast under standard growth conditions. Indeed, yeast can survive with an eIF-4E protein that has no detectable phosphorylation (see Fig. 5). In this context, it is significant that the mutation of Ser 209 to Ala in mammalian eIF-4E has no effect on this factor's association with the 48 S initiation complex (33). However, as with eIF-2␣ (17), our data do not rule out that there are conditions under which such sites are functionally important. Of at least equal importance is the observed variation in binding of phosphorylated p20 to eIF-4E, which will need to be investigated in considerable detail.
In conclusion, this study establishes the necessity to reassess the pathway and regulation of phosphorylation of yeast (and possibly mammalian) eIF-4E. Indeed, the variations in the amount of phosphorylated p20 we have described suggest that the possible involvement of other proteins that interact with eIF-4E may play important regulatory roles in translational initiation. We see here a potential parallel to the recently described regulatory system for mammalian eIF-4E. The binding of PHAS-I to mammalian eIF-4E is regulated via phosphorylation of Ser 64 in the regulatory protein by mitogen-activated kinase (50,55). Thus, the mitogen-induced stimulation of protein synthesis is thought to be mediated via this regulator binding mechanism. On the basis of our new data on the yeast cap-binding proteins, future work will need to explore analogous paths for potential regulatory mechanisms in this lower eukaryote.