Transcription factors regulate gene expression in the nucleus in response to a variety of external stimuli. This level of control is necessary in order for the organism to adapt quickly to changing environments. The mechanisms of such control can be complex and involve several other proteins which can act in concert with the transcription factor during functions including nuclear import/export, and binding to gene promoter sequences. Phosphorylation of transcription factors has been frequently observed in response to signaling events.
Pho4 is a transcription factor found in yeast which activates the PHO5 acid phosphatase gene when the organism is exposed to a phosphate-deficient environment. It has been demonstrated that (1) phosphorylated Pho4 is primarily localized to the cytoplasm, while unphosphorylated Pho4 is localized to the nucleus (2) Pho4 is phosphorylated in the nucleus by a complex consisting of a cyclin and a cyclin-dependent kinase, Pho80-Pho85 (3) Pho4 is phosphorylated at five separate sites which consist of Ser-Pro dipeptides (4) phosphorylated Pho4 is exported from the nucleus to the cytoplasm by Msn5, a member of the importin b protein receptor family; and (5) that Pse1, another importin b receptor protein, binds to and transports unphosphorylated Pho4 from the cytoplasm to the nucleus.
The authors of this paper seek to carry previous research in this area a step further by attempting to identify what roles the five sites of phosphorylation of Pho4 play in nuclear export, nuclear import, and transcription regulation. A series of experiments are performed which examine the behavior of mutant strains in which the expression of Pho4 has been modified by alteration of Pho4 phosphorylation sites using single amino acid substitutions at the dipeptide in question. The authors also investigated the roles of Pho4 phosphorylation sites in regulating Pho4 binding to the export receptor Msn5, the import receptor Pse1, and transcription factor Pho2 by using SDS-PAGE gels to estimate the degree of binding by various Pho4 strains to the receptor in question under either phosphorylated or mock-phosphorylated conditions. Finally, Pho5 acid phosphatase enzyme activity was measured for various Pho4 mutant yeast strains when exposed to high-phosphate or low-phosphate conditions.
The authors conclude that phosphorylation of sites 2 and 3 together promote Pho4 export from the nucleus and is required for proper binding to Msn5; that phosphorylation of site 4 blocks nuclear import and interferes with Pse1/Pho4 binding; and that phosphorylation of site 6 retards Pho4/Pho2 binding, inhibiting transcription of Pho5. The sum of their findings is that Pho4, through its phosphorylation sites, exerts a detailed and multi-layered regulatory mechanism for transcriptional response to variations in the organism’s environment.
1. Regulation of Pho4 nuclear export
Pho4 mutants in which Pho4 was expressed containing Ser6Ala substitutions at various Pho4 phosphorylation sites and fused to green fluorescent protein (GFP) were produced. Paired cultures were then exposed to high-phosphate and low-phosphate environments, respectively, and examined by fluorescent microscopy (FM). Mutants which contained Ser6Ala substitutions at sites 1,4, or 6 experienced nuclear export similar to that seen in wild-type Pho4-GFP strain, which serves as the control. However, Pho4SA2-GFP and Pho4SA3-GFP cultures demonstrated significant defect in nuclear export upon exposure to high-phosphate conditions as seen under FM (figure 1A). From this, and the demonstration that Pho4SA146-GFP was localized primarily to the cytoplasm under high-phosphate conditions, the authors conclude that phosphorylation of sites 2 and 3 in concert is the requisite condition in order for Pho4 to be properly exported from the nucleus under high-phosphate conditions.
The data tends to support the author’s conclusion. However, FM microscopy is not the most accurate method of determining cellular localization; in particular, it appears that it is at least possible that there has been a mild degree of export of Pho4SA2 -GFP and Pho4SA3-GFP from the nucleus in the photos demonstrating localization under high-phosphate conditions. It is likely that this is an artifact introduced by the intense fluorescence of the nucleus, but we cannot be certain. Moreover, while the authors conclude that phosphorylation of both sites 2 and 3 together is necessary for proper export, it is interesting that they do not further elaborate on this by creating a Pho4SA1246 or a Pho4SA1346 mutant and demonstrating the localization pattern when just one of sites 2 and 3 is phosphorylated.
The authors then examine the role of Pho4 phosphorylation with respect to binding interactions with the nuclear export receptor Msn5. Pho4SA23 and Pho4SA146 were fused to “zz” binding domains of Immunoglobulin G, were phosphorylated or mock phosphorylated in vitro, rendered stationary by addition to IgG-Sepharose beads, and treated with Msn5-His6 in concert with Gsp1Q71L, a form of Ran which is only able to manifest as the GTP-bound form, thus fully mimicking the Msn5 nuclear export complex. SDS-Page electrophoresis was performed in order to quantify the nature of Pho4-zz (wild type control and mutant forms) binding to the Msn5 nuclear export complex. The authors conclude from their data (fig.1B) that Pho4SA23-zz fails to bind to the Msn5 complex while Pho4SA146-zz experiences normal binding; thus, they determine that the phosphorylation of sites 2 and 3 is necessary for nuclear export of Pho4.
While the data once again tend to support the author’s conclusions, it is notable here that the SDS-PAGE gel for Pho4SA23-zz appears to demonstrate weak bands in the Msn5 binding assay; we would be more comfortable here if there was no banding evidence at all given the author’s conclusions. This is especially interesting given the criticism leveled above. Furthermore, only two mutants are examined for Msn5 binding: SA146 and SA23. SA23 fails to bind (for the most part, anyway) regardless of phosphorylation state. SA146 binds only in its phosphorylated state. While this is strong evidence that either site 2 or 3 is important in binding to Msn5, it fails to demonstrate that both are necessary together for binding. If, for instance, the case was that only site 3 was important for Msn5 binding, and site 2 was not involved, the author’s experimental results could be expected to be similar to what is seen here. No results are presented which indicate that both sites 2 and 3 are important, only that at least one of the two sites is important. The distinction is relevant because in addition to this, the binding assay was carried out in vitro, which calls into question whether other factors present in vivo are being excluded. Perhaps site 2 or 3 is involved in another activity necessary for nuclear export but not involved directly with Msn5 binding.
2. The role of phosphorylation in Pho4 nuclear import
The authors examine the role phosphorylation exerts upon the interaction between Pho4 and its nuclear import receptor Pse1 by scrutinizing Pho4 phosphorylation site 4, utilizing the same technique they used in examining Msn5 binding detailed above. Pho4SA4 -zz and Pho4SA1236-zz were both phosphorylated and mock phosphorylated and tested for degree of Pse1 binding by using an SDS-PAGE gel (fig. 2A). The authors conclude that SA1236 does not bind Pse1 when it is phosphorylated and that SA4 does bind Pse1, and that the phosphorylation status of site 4 is the crucial determinant.
While the data once again tend to support the author’s conclusion, the evidence, as in the Msn5 binding assay, is not conclusive; a weak band can be seen on the gel lane which represents phosphorylated SA1236 binding to Pse1-His6, whereas it would be preferred that there be no banding at all if we were to suppose the authors’ conclusions to be absolutely correct. This experiment took place in vitro, so the question of other factors present in vivo again also arises here.
In order to investigate the role of Pho4 binding site 4 in nuclear import in vivo, the authors create a special mutant Pho4 strain Pho4SA1236SD4 in which site 4 has been modified by a Ser6Asp substitution (which simulates phosphorylation at site 4) and in which all other sites have been blocked by a Ser6Ala substitution. These modifications are necessary in order to avoid error introduced by the fact that export of Pho4 and blocking import of Pho4 will both contribute to buildup of Pho4 in the cytoplasm. Pho4SA1236SD4 is fused to a GFP3 domain and its localization in the cell was compared to that of Pho4SA1236-GFP3. Pho4SA1236SD4 shows good localization in the cytoplasm under FM (fig.2B), whereas Pho4SA1236 appears to be confined to the nucleus. This is good evidence to support the author’s conclusion that the phosphorylation of site 4 blocks Pho4 nuclear import, although it is interesting that the SDS-PAGE gel (fig2A) lane for the SA1236SD4 mutant appeared to have a weak band in both phosphorylated and mock-phosphorylated conditions. The use of the S6D substitution to mimic phosphorylation (as well as other amino acid substitutions) could possibly be introducing some minor error into these experiments, for instance by slightly altering the conformation of the protein, but it is difficult to see how to get around this obstacle.
3. Pho4 phosphorylation sites and transcription regulation
Having elucidated the roles of Pho4 phosphorylation sites in nuclear export and nuclear import, the authors then turn to analyzing Pho4 regulation of transcription. Utilizing the techniques previously outlined using single amino acid substitutions at phosphorylation sites to produce Pho4 mutants and evaluating Pho5 acid phosphatase production in mutant strains under high phosphate and phosphate-deficient conditions, the authors determine that binding sites 1,2,3, and 4 are not significantly involved in Pho4 transcriptional regulation. However, they also noted that acid phosphatase production was somewhat increased in Pho4SA1234 strains cultured in high-phosphate mediums as opposed to wild-type Pho4 cultures(fig.3B), which is interesting because as seen below they postulate that phosphorylation of site 6 prevents proper binding of Pho4 to transcription factor Pho2 which binds to promoter sequence PHO5 and induces acid phosphatase production. This suggests that there are other factors involved in transcription control other than site 6 phosphorylation, although clearly site 6 phosphorylation is the most important event.
To examine the behavior of site 6, the authors create a Pho4 mutant which contains a Pro6Ala substitution at site 6. Ser6Ala was not used because the Ser6Ala mutation interfered with transcription. Tagging mutant Pho4 with GFP and examining FM patterns of localization(fig.3A) revealed that mutants Pho4SA1234 and Pho4SA1234PA6 remained isolated in the nucleus regardless of the phosphate levels present in the culture environment, while Pho4PA6 disseminated to the cytoplasm when culture was subjected to a high phosphate environment, as expected from previous results. Strains expressing Pho4SA1234PA6 secreted acid phosphatase at a high level in phosphate-rich culture as opposed to Pho4SA1234(fig.3B), indicating that the phosphorylation of site 6 effectively inhibits transcription, which is interesting because site 6 lays in the region in which Pho4 binds to transcription factor Pho2.
The authors then examine the role of site 6 in Pho4/Pho2 binding by assaying Pho4-zz tagged mutant and wild type proteins for binding to Pho2-His6 using SDS-PAGE electrophoresis as previously demonstrated in their analysis of Msn5 and Pse1 binding to Pho4 (fig.3C). The gel reveals that wild-type Pho4- zz binds to Pho2 only in its unphosphorylated state; that Pho4PA6-zz binds to Pho2 regardless of phosphorylation status; and that Pho4SA1234-zz binds to Pho2 only in its unphosphorylated state. The authors interpret these results as demonstrating that the phosphorylation status of site 6 is the critical determinant in Pho2/Pho4 binding and thus transcription regulation.
The data presented here is largely in agreement with the authors’ conclusion: the SDS-PAGE gel (fig.3C) does not exhibit weak aberrant banding as seen in the Msn5 and Pse1 binding assays. Could this be due to their modification of proline at site 6 as opposed to serine on dipeptide phosphorylation sites as used in other experiments? The measurement of acid phosphatase production by different mutant strains (fig.3B) under varying phosphate conditions reveals that there must be factors involved in transcription regulation other than simply site 6 phosphorylation status. In particular, mutants with pho4Δpho3Δ deletions expressing Pho4SA1234 and PhoPA6 using low copy plasmids both exhibit elevated acid phosphatase activity under high phosphate conditions as compared to wild-type Pho4, which exhibits virtually none. Thus, while site 6 is important, it is not the only factor involved in transcription regulation.
In spite of the mostly minor caveats mentioned above, the authors largely accomplished their goal of demonstrating the variable roles of phosphorylation sites of Pho4 in regulating nuclear export, nuclear import, and transcription. The technique of utilizing different mutant strains of Pho4 which contained single amino acid substitutions and comparing cellular localization and receptor binding under different phosphate conditions was impressive and quite logical.
Possible flaws introduced by substituting amino acids (which might slightly alter the structure of Pho4) could be addressed by examining Pho4 wild type control and mutant Pho4 proteins using X-ray crystallography and comparing their various conformations under phosphorylated and unphosphorylated conditions.
Perhaps another type of microscopy should be performed, utilizing different tags than those used in this experiment, in order to determine what if any residual nuclear export exists when sites 2 and/or 3 are blocked. The use of a Pro →Ala substitution at site 6 resulted in excellent banding on the SDS-PAGE gel assay for Pho4/Pho2 binding. Perhaps the same substitution could be tried at binding sites 2,3, and 4 when investigating Pho4 binding to Pse1 and Msn5, as there were weak bands present in lanes that should have been totally clear given the author’s conclusions. Different binding assays than those used in this experiment could also be tried to test Pho4 binding to Msn5 and Pse1.
A more in-depth analysis of Pho4 nuclear export could be performed, utilizing mutant strains which can only be phosphorylated at site 2 or site 3. This experiment would remove any remaining doubt as to the authors’ conclusion that both sites are involved at once in nuclear export with the Msn5 receptor complex. Phosphorylation sites 2 and 3 are located adjacent to each other on Pho4; how does the phosphorylation of one site effect the conformation of the protein? Are cooperative binding effects involved? Here again x-ray crystallography could provide some answers.
The probable involvement of multiple phosphorylation sites in the same process also brings to mind the possibility of ordering. Is one phosphorylation site favored over the other when sites 2 and 3 are phosphorylated in the nucleus? What about other phosphorylation sites? The existence of 5 separate sites causes one to wonder whether there is a specific hierarchy in nuclear phosphorylation involving the pho80/pho85 complex. Time-dependent phosphorylation experiments using isotopic phosphorous in combination with the amino acid substitution method at Pho4 phosphorylation sites could be performed to identify any patterns which may exist here.
Finally, it is fairly obvious from the author’s data that there is more to transcriptional regulation than the phosphorylation status of site 6. Future experiments should be directed at further elucidating the roles of other possible factors. The function of site 1 has yet to be determined. Is it also involved in transcription somehow?