Kinase activity-dependent nuclear export opposes stress-induced nuclear accumulation and retention of Hog1 mitogen-activated protein kinase in the budding yeast Saccharomyces cerevisiae

Abstract

Budding yeast adjusts to increases in external osmolarity via a specific mitogen-activated protein kinase signal pathway, the high-osmolarity glycerol response (HOG) pathway. Studies with a functional Hog1-green fluorescent protein (GFP) fusion reveal that even under nonstress conditions the mitogen-activated protein kinase Hog1 cycles between cytoplasmic and nuclear compartments. The basal distribution of the protein seems independent of its activator, Pbs2, and independent of its phosphorylation status. Upon osmotic challenge, the Hog1-GFP fusion becomes rapidly concentrated in the nucleus from which it is reexported after return to an iso-osmotic environment or after adaptation to high osmolarity. The preconditions and kinetics of increased nuclear localization correlate with those found for the dual phosphorylation of Hog1-GFP. The duration of Hog1 nuclear residence is modulated by the presence of the general stress activators Msn2 and Msn4. Reexport of Hog1 to the cytoplasm does not require de novo protein synthesis but depends on Hog1 kinase activity. Thus, at least three different mechanisms contribute to the intracellular distribution pattern of Hog1: phosphorylation-dependent nuclear accumulation, retention by nuclear targets, and a kinase-induced export.

Figure 1

Expression of Hog1–GFP and Pbs2–GFP complement hyperosmotic sensitivity of corresponding genomic mutants. (A) Strain K4327 (hog1Δ) was transformed with centromer plasmid pVR50 (HOG1), pVR65–WT (HOG1–GFP), pVR65–T/A (HOG1–GFP T/A), pVR65–Y/F (HOG1–GFP Y/F), or pVR65–K/R (HOG1–GFP K/R). The expression of HOG1 alleles was driven by their endogenous promoter. Transformants were grown on selective medium at 30°C without or with 0.4 M NaCl. (B) Strain VRY 10 (pbs2Δ) was transformed with pVR15 (PBS2), pVR15–GFP (PBS2–GFP), pVR20 (PBS2 S/A, T/A), or pVR15K/M (PBS2 K/M) and YCp22 (control plasmid, pbs2Δ). Expression of PBS2 alleles was driven by their endogenous promoter. Transformants were grown on selective medium at 30°C in the absence or presence of 0.4 M NaCl. (C) Antibody against an active human p38 MAPK specifically recognizes an activated form of yeast Hog1. The protein extracts from strain K4327 (hog1Δ) transformed with a centromer plasmid bearing a WT HOG1 (pVR50), HOG1–GFP fusion gene (pVR65–WT), or control plasmid (YCp111) were tested. Cells were grown in selective medium. For hyperosmotic stress, NaCl was added to a final concentration of 0.4 M for 5 min before preparation of protein extracts, which were analyzed by Western blotting using antibody against active p38 MAPK. Asterisks indicate the positions of a set of proteins from yeast crude extract (e.g., those with slightly higher mobility than Hog1) that are nonspecifically recognized by anti-active human p38 antibody.

Figure 2

Hog1–GFP nuclear accumulation in response to hy-perosmotic stress correlates with its dual phosphorylation status. (A) Hyperosmotic stress induces nuclear accumulation of Hog1. Logarithmically growing strain K4327 (hog1Δ) transformed with pVR65–WT (HOG1–GFP) was examined by fluorescence or light microscopy under iso-osmotic (selective medium) or hyperosmotic growth conditions (5 min after addition of NaCl into selective medium to a final concentration of 0.4 M). In agreement with the observations of others (Stade et al., 1997), DAPI used for staining of nuclei preferentially stains mitochondrial DNA in living cells. Positions of nuclei are indicated by arrows. DIC, differential interference contrast. Bar, 5 μm. (B) The kinetics of Hog1 nuclear accumulation and dual phosphorylation are similar. Strain K4327 (hog1Δ) was transformed with centromer plasmid pVR65–WT (HOG1–GFP). Cells were grown in selective medium to logarithmic phase (iso-osmotic conditions). A control sample of unstressed cells was taken (time 0 min) followed by addition of NaCl to a final concentration of 0.4 M. Then cells were incubated for various times and examined by fluorescence microscopy. To determine the activation profile of Hog1–GFP, samples were taken at the time points indicated, and protein extracts were prepared. Western blots were analyzed with anti-active p38 MAPK antibody and with anti-GFP antibody (C). In a control experiment, no band corresponding to the position of the Hog1–GFP-specific band was detected with anti-GFP antibody when protein extract from hog1Δ strain was used.

Figure 3

Phosphorylation, but not kinase activity is necessary for nuclear accumulation of Hog1. Strain K4327 (hog1Δ) was transformed with centromer plasmids (A) pVR65–Y/F containing HOG1 phosphorylation site mutant allele (HOG1–GFP Y/F, HOG1 with T174F mutation) or (B) pVR65–K/R containing HOG1 catalytic site mutant allele (HOG1–GFP K/R HOG1 with K52R mutation). Transformants were grown to logarithmic phase in selective medium (iso-osmotic). Where indicated, strains were stressed for 5 min by addition of NaCl to a final concentration of 0.4 M (hyperosmotic). Positions of nuclei are indicated by arrows. Bar, 5 μm. (C) Hog1–GFP nuclear accumulation is dependent on active Pbs2 MAPKK. The following strains were analyzed: strain VRY 10 (pbs2Δ) cotransformed with pVR65–WT (HOG1–GFP) and pVR15 (PBS2, endogenous promoter), pVR20 (PBS2S/A, T/A; PBS2 with S174A and T178A mutations), and pVR15K/M (PBS2K/M; PBS2 with K389M mutation). (D) The overexpression of the Ssk2ΔN allele induces nuclear accumulation of Hog1–GFP in nonstressed cells. Strain K4327 (hog1Δ) was cotransformed with centromer plasmid pVR65–WT (HOG1–GFP) and plasmid pGSS21 (2 μm P_GAL1-SSK2ΔN, SSK2 from M1173 to D1579). Cells were grown in appropriate selective medium containing raffinose as carbon source, washed with water, and resuspended in fresh selective medium with galactose or glucose as carbon source (iso-osmotic). Where indicated, cells were stressed with 0.4 M NaCl for 5 min before microscopy (hyperosmotic). glc, glucose; gal, galactose.

Figure 4

The MAPKK Pbs2 is not required for anchoring or basal nucleocytoplasmic cycling of Hog1. (A) Pbs2 appears constitutively localized in the cytoplasm. Strain VRY 10 (pbs2Δ) was transformed with pVR28 (2 μm PBS2–GFP), and transformants were grown in appropriate selective medium (iso-osmotic); where indicated, hyperosmotic stress was applied (0.4 M NaCl, 5 min; hyperosmotic). The positions of nuclei are indicated by arrows. (B) Strain VRY 10 (pbs2Δ) was transformed with pVR65-WT (HOG1–GFP). Cells with nuclear staining were counted after growing the strain to logarithmic phase in selective medium (iso-osmotic) and subjecting them to hyperosmotic conditions as described above. (C) Hog1–NES–GFP does not show basal nuclear staining under nonstress conditions. Strain K4327 (hog1Δ) was transformed with centromer plasmid pVR65–NES (HOG1–GFP fused to PKI NES), and cellular distribution of Hog1 was determined after growing them to logarithmic phase in appropriate selective medium (iso-osmotic). (D) Hog1 basal cycling is not dependent on Msn2 and Msn4 transcription factors. Strain YM24 (msn2,4Δ) was transformed with centromer plasmid pVR65 (HOG1–GFP), and cellular distribution of Hog1 was determined after growing cells to logarithmic phase in appropriate selective medium (iso-osmotic). Bar, 5 μm.

Figure 5

Intracellular distribution and phosphorylation of enlarged Hog1 kinases. (A) Fluorescence images of hog1 cells transformed with a plasmid expressing Hog1 fused to three tandemly repeated GFP polypeptides. Cells were treated and processed as in Figure 8. GFP images were taken under normal growth conditions 5 min after osmotic stress and 5 min after return to iso-osmotic medium. (B) Western analysis from yeast cells containing a HOG1 fusion with one, two, and three copies of GFP. The top panel was been probed with anti-active p38 antibody; the bottom panel was probed with antibodies against GFP. The apparent sizes for marker proteins are indicated.

Figure 6

Duration of Hog1 nuclear accumulation is reduced in cells lacking Msn2 and Msn4 transcription factors. Logarithmically growing strain W303-1A (WT) and YM24 (msn2,4Δ) transformed with pVR65–WT (HOG1–GFP) were examined by fluorescence microscopy under iso-osmotic (selective medium) or hyperosmotic growth conditions (5 min after addition of NaCl into selective medium to a final concentration of 0.4 M) at the indicated time points (A) or by determining the phosphorylation profile of Hog1–GFP (B). Samples were taken at the time points indicated, and protein extracts were prepared. Western blots were analyzed with anti-active p38 antibody (asterisks indicate the positions of proteins that are nonspecifically recognized by antibody) or with anti-GFP antibody (bottom panel) to determine the protein level of Hog1–GFP.

Figure 7

Cellular distribution of Hog1 during stress and adaptation is independent of de novo protein synthesis. Logarithmically growing strain K4327 (hog1Δ) transformed with centromer plasmid pVR65–WT (HOG1–GFP) was preincubated with cycloheximide (0.1 mg/ml) for 2 h before microscopy. Pictures were taken from cells incubated under iso-osmotic (selective medium) (A) or hyperosmotic conditions (5 min after addition of NaCl to a final concentration of 0.4 M) (B). (C) Cells from B were left to adapt for 60 min and then reexposed to hyperosmotic stress (5 min, final concentration of NaCl 0.8 M) (D). The positions of nuclei are indicated by arrows. Bar, 5 μm.

Figure 8

Hog1 nuclear export and dephosphorylation depend on Hog1 kinase activity. Strain K4327 (hog1Δ) was transformed with centromer plasmids containing pVR65–WT (HOG1–GFP) or HOG1–GFP K/R (pVR65–K/R) and strain W303-1A (WT) with HOG1–GFP K/R (pVR65–K/R). After exposure to hyperosmotic conditions (selective media with 0.4 M NaCl, 5 min), cells were returned rapidly to iso-osmotic conditions (selective medium without NaCl), and localization of Hog1–GFP derivatives was determined after 5 min (A; positions of nuclei are indicated by arrows; bar, 5 μm), or samples were taken at the time points indicated, and protein extracts were prepared (B). Western blots were analyzed with anti-active p38 MAPK antibody. Asterisks indicate the positions of proteins that are nonspecifically recognized by anti-active human p38 antibody.

Figure 9

Model proposing how different regulatory mechanisms cooperate to determine Hog1 intracellular distribution. (A) Under favorable growth conditions Hog1 travels between cytoplasm and nucleus. This cycling is independent of activity and the presence of upstream components of the MAPK pathway. (B) During acute stress Hog1 rapidly accumulates in the nucleus. Hog1 phosphorylation by Pbs2 is essential for this process. Phosphorylated Hog1 is either more competent for nuclear import or less competent for nuclear export (as indicated by the thickness of the arrows). At the same time, some nuclear retention factors accumulate in the nuclear compartment. At this step, Hog1 kinase might modify stress response–specific targets. (C) Late phase of adaptation induces kinase activity-dependent nuclear export of Hog1. There are three potential mechanisms for which kinase activity is necessary. Hog1 might activate nuclear phosphatases; that might convert Hog1 into a more efficient export cargo. Alternatively, nuclear substrates phosphorylated by Hog1 might be exported, decreasing the concentration of nuclear retention factors. Finally, Hog1 might be required to enhance the activity of an export system. P, phosphate; PP, protein phosphatase.