Identification of yeast genes involved in k homeostasis: loss of membrane traffic genes affects k uptake

Abstract

Using the homozygous diploid Saccharomyces deletion collection, we searched for strains with defects in K(+) homeostasis. We identified 156 (of 4653 total) strains unable to grow in the presence of hygromycin B, a phenotype previously shown to be indicative of ion defects. The most abundant group was that with deletions of genes known to encode membrane traffic regulators. Nearly 80% of these membrane traffic defective strains showed defects in uptake of the K(+) homolog, (86)Rb(+). Since Trk1, a plasma membrane protein localized to lipid microdomains, is the major K(+) influx transporter, we examined the subcellular localization and Triton-X 100 insolubility of Trk1 in 29 of the traffic mutants. However, few of these showed defects in the steady state levels of Trk1, the localization of Trk1 to the plasma membrane, or the localization of Trk1 to lipid microdomains, and most defects were mild compared to wild-type. Three inositol kinase mutants were also identified, and in contrast, loss of these genes negatively affected Trk1 protein levels. In summary, this work reveals a nexus between K(+) homeostasis and membrane traffic, which does not involve traffic of the major influx transporter, Trk1.

Keywords: TRK1; VPS genes.

Figure 1

Representative data showing the three classes of hygromcyin B-sensitive mutants. The homozygous diploid deletion collection of 4653 strains was screened for sensitivity to hygromycin B in the presence or absence of added K⁺ (100 mm or 500 mm). Some strains displayed drug sensitivity that was suppressed by 100 mm KCl (including arl1∆, vps8∆, vps9∆, and pep5∆), others required 500 mm KCl for sensitivity to be suppressed (including arf1∆ and per1∆), and others were not suppressed at all by the inclusion of KCl (including vps16∆ and vps54∆). For the complete data set, see Table S1A, Table S1B, and Table S1C. YPAD = normal rich growth medium, + Hyg B = with the addition of 0.075 mg/ml hygromycin B with no addition, or with the addition of 100 mm KCl or 500 mm KCl. The indicated strains were grown overnight in YPAD medium, then diluted to 1.0 OD₆₀₀/ml and subjected to 1:10 serial dilutions, spotted onto the indicated plates with a replicator tool, then grown for 2 days at 30°C. The total data set is reported in Table S1A, Table S1B, and Table S1C.

Figure 2

Representative data showing ⁸⁶Rb⁺ uptake. Examples of data taken from the ⁸⁶Rb⁺ uptake assay. Most membrane traffic mutants tested showed low levels of ⁸⁶Rb⁺ uptake (arl1Δ, ▴ = vps8Δ, x = vps9Δ, ∆ = vps21Δ, and + = mon2Δ. Each strain was tested in triplicate on at least two different occasions. The data are summarized for the class I membrane traffic strains in Table 3. The total data set for all the membrane traffic mutants from all 3 classes is shown in Table S3.

Figure 3

Steady state levels of Trk1-HA in representative membrane traffic mutant strains. Wild-type (BY4743) and mutant strains were transformed with a 2 μ plasmid containing TRK1-HA. Cells were grown, then lysates were prepared for Western blot analysis with an anti-HA primary antibody followed by an HRP-conjugated secondary as described in Materials and Methods. Equal numbers of cells were prepared for each sample. The majority of strains tested showed levels of Trk1-HA similar to that of wild-type. Here we show one strain, glo3Δ, which overexpresses Trk1-HA, and one strain, ric1Δ, which lacks expression of Trk1-HA. Each of the 29 transformed class I membrane traffic mutant strains was tested at least twice, and the data are summarized in Table 3.

Figure 4

Subcellular fractionation analysis of representative strains. Wild-type (BY4743) and mutant strains transformed with 2 μ plasmid containing TRK1-HA were grown, then lysed and prepared for subcellular fractionation analysis as described in Materials and Methods. The majority of strains tested showed all Trk1-HA in the P13 fraction as for wild-type shown here, consistent with plasma membrane localization. However, a few strains, including sec22Δ, showed a portion of Trk1-HA in the S13 fraction, consistent with the notion that a portion of the protein was contained in an internal membrane pool. The small amount of PGK in the P fractions is due to incomplete removal of the supernatant in an effort to avoid disturbing the pellet. Each of the 29 class I transformed membrane traffic mutant strains was tested at least twice, and the data are summarized in Table 3.

Figure 5

Trk1-HA localization to lipid microdomains in representative strains. Wild-type (BY4743) and mutant strains transformed with 2 μ plasmid containing TRK1-HA were grown and then were lysed, treated with Triton X-100, and prepared for raft analysis as described in Materials and Methods. T = total, S = supernatant from the 100,000 × g spin (Triton-X soluble fraction), and P = pellet from the 100,000 × g spin (the Triton-X insoluble fraction contains the lipid microdomains or rafts). The majority of strains tested localized Trk1-HA exclusively to the Triton-X 100 pellet as shown here for wild-type. However, a few strains, as shown here by cog6Δ, did not localize Trk1-HA exclusively to the raft fraction. Each of the 29 transformed class I membrane traffic mutant strains was tested at least twice and the data are summarized in Table 3.

Figure 6

Effects of K⁺ on CPY secretion in selected membrane traffic mutants. CPY secretion was measured in the presence and absence of 500 mm KCl or 1M sorbitol. Sorbitol served as an osmolarity control, used to test whether K⁺ specifically or increased osmolarity in general affected CPY secretion. We found that in 28 of 56 strains tested, both K⁺ and sorbitol suppressed secretion. In 10 strains, only K⁺ suppressed CPY secretion (as shown in the vps9Δ strain, for example). In 18 strains, neither K⁺ nor sorbitol affected CPY secretion. For the complete data set, see Table S4.

Figure 7

Inositol kinases are involved in the modulation of Trk1 steady state protein levels. (A) Wild-type (BY4743), trk1∆, arg82∆, fab1∆, and kcs1∆ strains were grown in rich medium, then diluted to a concentration of 1 OD/mL. Ten-fold serial dilutions of the cells were spotted onto media containing HB, HB + KCl, and HB + sorbitol at the above listed concentrations, and the plates were incubated for 3 days at 30°C and then photographed. (B) Wild type, arg82∆, fab1∆, and kcs1∆ strains were transformed with a 2 µ plasmid bearing Trk1-HA. Protein lysates were prepared from these transformed strains, and the total lysate was separated by gel electrophoresis as described in Materials and Methods and subjected to Western blotting with the anti-HA antibody. Pgk1 is shown as a loading control.

Figure 7

Inositol kinases are involved in the modulation of Trk1 steady state protein levels. (A) Wild-type (BY4743), trk1∆, arg82∆, fab1∆, and kcs1∆ strains were grown in rich medium, then diluted to a concentration of 1 OD/mL. Ten-fold serial dilutions of the cells were spotted onto media containing HB, HB + KCl, and HB + sorbitol at the above listed concentrations, and the plates were incubated for 3 days at 30°C and then photographed. (B) Wild type, arg82∆, fab1∆, and kcs1∆ strains were transformed with a 2 µ plasmid bearing Trk1-HA. Protein lysates were prepared from these transformed strains, and the total lysate was separated by gel electrophoresis as described in Materials and Methods and subjected to Western blotting with the anti-HA antibody. Pgk1 is shown as a loading control.