what you need to know beforehand
Special DNA regions (STREs) promote specific gene transcription under stress
- "STRE" regions in gene promotors
- list of genes?
- transcription factors: Msn2, Msn4
Stefan Hohmann: es gibt keinen unspezifischen "general" stress response, die Antwort über STREs ist spezifisch!
- Aktin agglomeriert in whi2-Zellen
- ...whi2 mutants were defective in fluid-phase endocytosis...
- ...The actin clumps in whi2 and siw14 mutants are strikingly similar to the actin clumps seen in ark1 prk2 cells (Cope, 1999)...
- ...suggests that Whi2 may be a positive regulator of endocytosis...
- ...The actin cytoskeleton participates in endocytosis...
- Jpred3 sagt eine Reihe von Helizes und Faltblättern voraus
MDDIITQVSPDNAESAPILQEQQQQQNSQYEGNEEDYGDSLIHLNIQENHYFITRDQLMSLPESLLLCLFPSGVFLDRCGQVITNLTRDDEVYIVNFPPDCFEYIMEIYTKAHDDLYNHPVEKFFDR ---EEEEEE-------------------------------EEEEE----EEEE-HHHH---HHHHHHH----------------------EEEE------HHHHHHHHHHHHHHHHHH--------- PSSSFVSNAKGFFGLSSNNSISSNNEQDILHQKPAIIVLREDLDYYCVPQEEFQFDSTNEENNEDLLRHFMAQVKMAAGSYLTSKTSIFQGLYSSNRLKQQQQQQKIEKGSNSSSNTKSTSKKLGPA -----------------------------------EEEEE----EEE----EEE---------HHHHHHHHHHHHHHHHHHHHHH--HHHH-----------------------------------H EQHLMDMLCSSGFTKETCWGNRTQETGKTVISSLSLCRLANETTEGFRQKFNEAKAKWEAEHKPSQDNFITPMQSNISINSLSASKSNSTISTARNLTSGSTAPATARDKRKSRLSKLADNVRSHSS HHHHHHHHH-------------------EEEHHHHHHH-----HHHHHHHHHHHHHHHHHH-------------------------------------------------HHHHHHHH----E---- SRHSSQTRSKPPELPKLYDLVPKPNINAKLLLFWRKPARKCWWGEEDIELEVEVFGSWKDESKKIIELILPTNVDPEAELHKIIVPVRLHIRRVWTLELSVIGVQ --------------------------------E-------------EEEEE-------------EEEE--------------EEEEEEEEEEEEE-EEEEEEE--
Whi2 is a 55 kDa cytoplasmatic, globular scaffold protein of bakers' yeast Saccharomyces cerevisiae. It's protein sequence shows no homologies to other yeast proteins (data not shown) and it's structure is unknown.
Protein sequence analysis reveals neither transmembrane domains, nor signal peptides, but two distinct POZ domains of different length. The POZ domain is a motif, that conferres interaction with either DNA or protein and is capable of conferring protein dimerization and tetramerization.
In yeast, cell size homeostasis after proliferation is ensured by arrest at the G1/S cell cycle phase boundary ("START"), which prevents cell division until a critical cell size is attained (Zhang et al, 2002). Rahmann (1988) as well as Auberson and von Stockar (1992) showed, that cells lacking Whi2 continue proliferating after diauxic shift from fermentable to non-fermentable carbon sources (glucose to ethanol). This behavior is inappropriate given the expected depletion of nutrients. Accordingly, it results in a cell density three times higher, each cell having only half the volume, compared to wild-type cells, suggesting Whi2 to play a role in nutrient-dependent cell cycle arrest. Auberson and von Stockar pointed out, that this specific whi2 mutant phenotype was exclusively visible during fully oxidative growth on ethanol, neither during aerobic nor anaerobic growth on glucose.
Kelly et al (1988) confirmed these results, adding, that mutant whi2 cells fail to arrest in G0, a state of physiological adaption involving cease of cell division (Hartwell, 1974), resistance to a variety of environmental stresses, such as heat shock (Schenberg-Frascino 1971), and accumulation of the storage carbohydrates glycogen and trehalose (Lillie and Pringle, 1988). Instead, whi2 cells arrest randomly in the cell cycle, presumably due to energy exhaustion. Interestingly a similar inability to arrest was found in ard1 mutant cells (Whiteway \& Szostak, 1985), however, we could not establish a link between the behavior of Whi2 and Ard1 in the context of this work.
Loss-of-function mutations of Whi2 as-well-as Whi3 are known to upregulate the transcription of the cyclins Cln1 and Cln2, which results in abnormally small cells (Sudbery, 1980; Radcliffe, 1997). Overexpression of Whi2 triggers polar budding in haploids, hinders the completion of cytokinesis, and causes hyperpolarized growth which results in the production of cells with an elongated shape (Radcliffe, 1997).
Recent findings suggest, that Whi2 plays a special role in the induction of mitophagy, the specific degradation of one cell's mitochondria by transport to and lysis inside the vacuole. Up to our knowledge, this is currently unexplained.
In the present work, the signalling network of all proteins and functional modules, known to interact with Whi2, has been modelled. Since little is known about Whi2 and many of it's interaction partners, the type of the modelled network is Boolean, allowing all network nodes to take only two possible states, on or off respectively. In order to confirm al infered interactions as well as to test their downstream effects, the created model has also been subject to simulation.
Siniossoglou et al first identified Psr1 and Psr2 in 2000 as two functionally homologous protein phosphatases in yeast, required for adaption to sodium osmostress via transcriptional activation of ATP-driven sodium pump Ena1. Remarkably, in this function it is independent of the HOG and Calcineurin pathways.
They identified a conserved DxDxT motif at position 263 (Psr1) and 233 (Psr2), required for functionality. Meanwhile, the whole C-terminal region of both Psr1 and Psr2 is recognized by protein sequence analysis (yeastgenome.org) to be a nuclear import factor recognizing domain (NIF NLI).
Protein sequence comparison confirms the redundancy of Psr1 and Psr2 (51% identity, 62% similarity, only 13% gaps).
Psr is strictly localized to the plasmamembrane due to a conserved, amino-terminal sequence motif, suggesting a close functional relationship with plasmamembrane-bound environmental sensors.
Estruch et al first reported in 1993, that two homologous Cys2His2 zinc finger proteins Msn2 and Msn4 act as multicopy-suppressors in an Snf1 mutant. They already found, that deletion of Msn2 caused no phenotype, while double deletion with Msn4 caused defects related to carbon utilization, indicating a redundant function.
Protein sequence comparison partly confirms this finding (32% identity, 45% similarity, but also 29% gaps).
Meanwhile it is the established opinion, that Msn2/4 are transcriptional activators, localized to the cytoplasm under stress-free conditions, translocated to the nucleus upon a variety of stresses, where they specifically bind stress-response elements (STREs) in gene promotors, thereby inducing gene expression. Translocation is regulated via Msn2/4 (de)phosphorylation and depends on their nuclear import/export sequences.
Model of induction of stress response
Whi2 binds Psr1 and Psr2
As we already stated, Psr1 and Psr2 fulfill redundant functions in yeast.
- protein sequence comparison Psr1 vs Psr2
Yu H found Whi2 bound to Psr1 (2008), while Ito H found Whi2 bound to Psr2 (2001), both in Two-hybrid assays. Although we don't want to defraud, that in Affinity-Capture MS, Whi2 only captured Psr1, not Psr2 (Ho 2002; Krogan 2006; Collins 2007), we believe it is safe, to assume Psr1/2 redundancy.
|Kaida et al (2002)|
|Yeast Whi2 and Psr1-phosphatase form a complex and regulate STRE-mediated gene expression|
|PubMed (title) PubMed (ID) Google||Vorlage:Paper|
Kaida et al found, that binding of Whi2 to Psr1 is required for STRE activation. Upon deletion of Whi2, STRE promotor containing gene expression was reduced to half under various stress conditions. While this finding hints to Whi2 acting in parallel with at least one other STRE activating pathway, we will show later, that this is not necessarily the case.
It may be of interest, that the importance of e.g. Psr1 for stress response is visible on the corresponding BioGRID page. Many genetic interactions are listed there, associated e.g. with chromatin remodeling and metabolically important proteins, localized to or in the mitochondria, as such, when interpreted correctly, already suggesting, that Psr1 is linked mitochondrial quality control.
Whi2 binds Msn2 and probably also Msn4
Msn2 and Msn4 are the only transcription factors, that up to our knowledge are capable of inducing STRE gene expression. It has been shown, that Msn2/4 activity depends on Rim15, which integrates several upstream signals, inhibiting induction of general stress response. While from this findings, it could be concluded, that Rim15 interacts with Msn2/4 to induce STRE activation, we will show later, that a different mechanism is more likely to be responsible for this. Still, co-immunoprecipitation of Whi2 and Msn2 strongly hints to Psr1/2-Whi2-Msn2 complex formation.
We didn't find a reference, that Msn4 also binds Whi2, so we cannot rule out, that Msn4 is activated in a Whi2-independent fashion, but we assume it.
Contradictingly, Sadeh et al. report a decrease in stress response upon deletion of Msn2, while deletion of Msn4 showed little to no decrease of stress reporter gene expression. While this may be an artifact due to unfortunate choice of reporter genes, it still demonstrates, that there must be slight differences between the target genes (STRE sequences?) of Msn2 and Msn4.
Psr1/2-Whi2-Msn2/4 complex formation is responsible for STRE activation
The delta whi2 mutant and also the double mutant delta psr1 psr2 show similar phenotypes, failing to elicit an appropriate stress response. Since the reduced stress response can be accounted to failure in Msn2/4 confered STRE activation, Whi2's function to activate Msn2/4 must hence be dependent on Psr1/2, confirming the redundancy of the latter.
Whi2 strongly binds Msn2 (in the cytoplasm, since Whi2 is a cytoplasmatic protein), a binding which is abolished after heat shock. Since Msn2 (and also Msn4) are transcription factors and after heat shock, they are supposed to induce STRE genes, they must be transported to the nucleus after activation.
Psr1/2 is responsible for dephosphorylation of Msn2
Msn2 activation correlates with a decrease in phosphorylation of several sites in its nuclear localization signal (NLS). This dephosphorylation must occur via stress-responsive phosphatase activity. If we combine, that Whi2 binds Msn2/4 as well as phosphatase Psr1/2, it would be safe to assume, that Whi2 mediates Msn2/4 dephosphorylation via recruitment of Psr1/2 after some kind of stress-induced activation of Whi2 and/or Psr1/2. Surprisingly, in this context nutrition conditions do not seem to play a role in Whi2 function, as it was suggested by the results linking Whi2 to Ras signalling and nitrogen starvation. The assumption of Psr1/2 being the phosphatases inducing Msn2/4 nuclear import, goes together with the above finding of STRE gene expression being reduced to half in delta whi2 cells, since Psr1/2 may still become activated upon an upstream stress signal, but being less efficient in finding it's downstream dephosphorylation targets Msn2/4. It also resembles the findings, that delta whi2 and delta psr1/2 mutants exhibit the same phenotype of defective stress response, since both Whi2 and at least one Psr would be necessary to induce (efficient) Msn2/4 dephosphorylation, and thus appropriate stress response. The finding is further supported by the fact, that Psr1 and also Psr2 protein sequence analysis reveals a known nuclear import sequence dephosphorylating domain (NIF phosphatase), indicating, that the phosphatase activity of Psr1/2 is capable of acting directly on a transcription factor.
- is there evidence, that NIF domains recognize the NLS present in Msn2/4 ?
Activity regulation of the Psr1/2-Whi-Msn2/4 complex
Garmendia-Torres et al identified Msn2 amino acids 575 to 642 as the nuclear import sequence, being recognized by both karyopherins Kap121 and Kap123 in-vivo. They also identified 4 distinct PKA phosphorylation sites in this sequence, suggesting that PKA is acting to suppress stress response in a two-fold manner, by suppressing chromatin remodelling (Rim15) and also Msn2/4 translocation, both of which it confers via direct interaction. Also other kinases seem to inhibit Msn2/4 translocation via phosphorylation of it's nuclear import sequence (see Rim15 below).
Prediction of Psr1/2 autodephosphorylating behaviour
The fact, that stress signal transmission is still at 50% in delta whi2 cells, although the necessary interaction conferer is missing, leaves the possibility, that one or more parallel pathways rescue the cell from from full susceptibility to stress, but may also hint to a pathway, that creates a very strong signal, since it seems unlikely to us, that dephosphorylation of the crucial stress-inducing factor Msn2/4 occurs spontaneously.
Interestingly, Psr1 binds Psr2 and vice-versa, hinting to a possible Psr dimerization or to some kind of autodephosphorylation mechanism. With respect to the expected strong signal, autodephosphorylation seems to be the favorable explanation, since it could quickly enhance and distribute a possible stress signal, once it has been induced. However, according to the signal's strength, the phosphorylation of Psr1/2 is then expected to be confered by a strong mechanism under stress-free conditions, since phosphorylation is required to suppress inappropriate stress response.
Autodephosphorylation, of course, would require Psr1 and also Psr2 to contain the phosphorylation motif, they are by themselves capable of dephosphorylating. Since in our modell Psr1 and Psr2 are responsible for dephosphorylation of Msn2 and presumably Msn4, which we showed above, contain PKA target motifs, we looked for PKA target motifs in Psr1 and Psr2. In fact, Psr1 contains a phosphorylation site for PKA. Therefore we expect, Psr1 activity to rise exponentially upon activation, limited only by diffusion of Psr1 and strong PKA activity. In this context, it must be noted, that we could not find a PKA target motif in Psr2.
Protein sequence alignment using the EMBOSS needle tool reveals, that although there is a high similarity between Psr1 and Psr2 (51% identity, 62% similarity, 13% gaps), the PKA phosphorylation site in Psr1 is deleted in Psr2 (lies inside an alignment gap). We are currently unable to explain this finding and admit, that it could contradict previous findings claiming Psr2 to act redundant with Psr1.
Kinase activity on Psr1/2, Whi2 and Msn2/4
The importance of Rim15 for the Psr-Whi2-Msn pathway has been previously addressed. Since the transcriptional activators Msn2/4 require Rim15-confered chromatin remodeling and on the other hand stress induced chromatin remodeling does not make sense without activation of the propper transcriptional activators, it is safe to assume, that similar upstream signals control both Rim15 and Psr1/2-Whi2-Msn2/4 regulation. Rim15 is (at least) under the control of glycolysis-sensing kinase Sch9, TCA- & glucose-sensing kinase PKA, nitrogen-sensing kinase TORC1 (via inactivation of phosphatase(s) PP2A) and phosphate-sensing kinase Pho85-80.
PhosphoGrid predicted, that Psr1 is under the control of PKA, however Psr2 does not contain a recognized kinase target sequence, Whi2 is controlled by GSK-3, and Msn2 as well as Msn4 are controlled by PKA, CK-1 and also GSK-3.
It may be, that Sch9-, PP2A- and Pho85-80-motifs are not found in this context, because they are not known.
Interpretation of the distribution of kinase targets
We would like to note, that a stress signal induced via Psr1/2-Whi2-Msn2/4, could be considered to be of a recessive nature, since it can easily be revoked downstream by efficient inhibition of Rim15, e.g. due to efficient nutrient-sensory kinase activity. In a scenario, where the translocation of Msn2/4 to the nucleus has taken place with the STREs still being heterochromatized, the Psr1/2-Whi2-Msn2/4 pathway activity could also be considered a "precautionary" stress signalling: In case of e.g. a drop in nutrient-sensory kinase activity, Rim15 would shutdown and the already translocalized transcription factors could immediately induce stress response.
It may also not be a coincidence, how upstream kinase activity is distributed among Psr1, Psr2, Whi2, Msn2 and Msn4. The more downstream their inhibitory signals occur, the more they also influence the impact of upstream inhibitions or activations. Also, possible different amounts of phosphorylation and also the different protein structures could modulate individual phosphorylation's impact on downstream activity. Thus signals from different pathways may quantitatively modulate each other on platform of the Psr1/2-Whi2-Msn2/4 pathway. While in the context of this work, only qualitative studies (Boole'an networks) were possible, quantitative studies and further experiments are required to elucidate this topic.
other Msn2 interaction partners
Sadeh et al have identified numerous physical and genetic interactions of Msn with various kinases, phosphatases, transporters, and chromatin remodelers.
As expected, stress response is decreased or abolished in psr1, psr2, whi2, msn2, rim15, rpd3 and hda3 mutants. Interestingly also msn5 mutants show decreased stress response, but it should not be confused with Msn2 or Msn4, since Msn5 is a karyopherin, involved in nuclear protein import and export. We suggest, that Msn5 plays a role in Rim15 translocation.
From their findings it is also visible, that oxidative stress is modulating stress response differently, than heat or osmotic stress. Stress response significantly increases after deletion of components of the PKA as well as the TORC1 pathway under conditions of oxidative, but not heat or osmotic stress, except for deletion of Ras, which drastically increases stress response in all scenarios and to the most extent measured. While the disruption of Rim15 inhibitors understandably enhances stress response, the latter finding clearly demonstrates, that Ras is playing an important and special role in induction of stress-response, which seems to be independent of the PKA pathway, but is still confered via Msn. If Ras acts on Msn via PKA, we would expect the tpk1 and tpk3 mutants to show a phenotype similar to ras2, since we would expect the Rim15 and/or Msn phosphorylation to become abolished. Surprisingly, they show a different phenotype. This might indicate, that tpk1 and tpk3 are both independently active components of PKA, separately acting on Rim15 and Msn respectively. Since deletion of tpk3 increases stress response similar to ras2 under no and oxidative stress, Tpk3 presumably acts on a mitochondrial quality controller. tpk1 deletion on the other hand shows little to no increase in all scenarios, suggesting it to fulfill a rather general mechanism of function, while still being required for full efficiency of the PKA.
Tpk3 seems to inhibit stress response in response to a mitochondrial quality control mechanism.
It has previously been shown, that Whi2 is required for degradation of Ras.
Additionally it has been shown, that PKA activity mainly responds to presence of glucose, while recent findings uncovered, that Ras activity is not responsive to glucose presence but cytosol acidification. These findings apparently contradict the established opinion of the "glucose-sensing Ras-cAMP-PKA pathway" and favor the second possible Cyr1 activation via Gpr1-Gpa2 to be the actual glucose-responsive upstream step of PKA signalling, while Ras is linked via a different pathway.
It is still to be elucidated, why deletion of PKA and TORC pathway components increase stress response under oxidative, but not osmotic or heat stress.
We expect inactivation of one Rim15 upstream inhibitor to be sufficient for stress response induction, since we believe, that similar kinases account for Msn cytoplasmatic retention and Rim15 inhibition cooperatively, as it is a general, upstream signal-integrating mechanism. If we assume, that Rim15 inhibition in the above mutants is abolished, as our model predicts, suppression of stress response must be due to inhibition of Psr or a dominant kinase activity on Msn.
- required for correct maturation of the 20S proteasome, degraded by proteasome upon completion of its assembly
- interestingly, deletion affects Msn translocation especially in the absence of environmental stresses
- Ump1 binds Msn2 and Msn4
- Ump1 is phosphorylated by Pho85
- Slt2 overexpression rescues delta ump1 mutant
Whi2's database interactions
The BioGRID is an online database, collecting interaction data of proteins, DNA and RNA, found in high- and low throughput experiments (unspecific whole-genome/proteome assays vs. experimental data about single proteins). Available results are sorted into various categories, allowing a rough estimation of the observed interaction strength.
The BioGRID dataset for Whi2 interactions lists 16 physical and 277 genetic interactions (June 2012).
Among the many inferred genetic interactions (277 in June 2012), some are rather interesting:
- Aim4 mutant is viable, but displays elevated frequency of mitochondrial genome loss and is sensitive to freeze-thaw stress; simultaneous deletion with Whi2 causes severe fitness defects
- deletion of Cdc25 rescues the lack of stress response, seen in the whi2 deletion mutant
- simultaneous mutation of Whi2 and Hsp82 chaperone is lethal
- simultaneous mutation of Whi2 and Lsm3 (part of Lsm2-7 complex) reduces the severeness of the fitness defects
- overexpression of Whi2 rescues mutations in Msn2
- overexpression of Msn4 enhances the phenotype of Whi2 mutations (!!)
- simultaneous mutation of Whi2 and Pde2 causes severe fitness defects
- simultaneous mutation of Whi2 and Pho80 causes severe fitness defects
- simultaneous mutation of Whi2 and Tor1 enhances the individual mutation's phenotype
- simultaneous mutation of Whi2 and Ras2 enhances the individual mutation's phenotype
- simultaneous mutation of Whi2 and Sch9 causes severe fitness defects
- simultaneous mutation of Whi2 and Snf4 (part of Snf1 kinase complex) enhances the individual mutation's phenotype
- deletion of Whi2 rescues the lethality of Tpk3 deletion
- simultaneous mutation of Whi2 and Vhs1 causes severe fitness defects