Specific α-arrestins negatively regulate Saccharomyces cerevisiae pheromone response by down-modulating the G-protein-coupled receptor Ste2

Abstract

G-protein-coupled receptors (GPCRs) are integral membrane proteins that initiate responses to extracellular stimuli by mediating ligand-dependent activation of cognate heterotrimeric G proteins. In yeast, occupancy of GPCR Ste2 by peptide pheromone α-factor initiates signaling by releasing a stimulatory Gβγ complex (Ste4-Ste18) from its inhibitory Gα subunit (Gpa1). Prolonged pathway stimulation is detrimental, and feedback mechanisms have evolved that act at the receptor level to limit the duration of signaling and stimulate recovery from pheromone-induced G1 arrest, including upregulation of the expression of an α-factor-degrading protease (Bar1), a regulator of G-protein signaling protein (Sst2) that stimulates Gpa1-GTP hydrolysis, and Gpa1 itself. Ste2 is also downregulated by endocytosis, both constitutive and ligand induced. Ste2 internalization requires its phosphorylation and subsequent ubiquitinylation by membrane-localized protein kinases (Yck1 and Yck2) and a ubiquitin ligase (Rsp5). Here, we demonstrate that three different members of the α-arrestin family (Ldb19/Art1, Rod1/Art4, and Rog3/Art7) contribute to Ste2 desensitization and internalization, and they do so by discrete mechanisms. We provide genetic and biochemical evidence that Ldb19 and Rod1 recruit Rsp5 to Ste2 via PPXY motifs in their C-terminal regions; in contrast, the arrestin fold domain at the N terminus of Rog3 is sufficient to promote adaptation. Finally, we show that Rod1 function requires calcineurin-dependent dephosphorylation.

FIG 1

Specific α-arrestins negatively regulate pheromone signaling and act independently from secreted protease Bar1 and RGS protein Sst2. (A) Pheromone sensitivity of wild-type MATa cells (BY4741) and otherwise isogenic derivatives containing the indicated α-arrestin deletions (9arrΔ, EN60; ldb19Δ, BY4741 ldb19Δ; rod1Δ rog3Δ, JT5858; aly1Δ aly2Δ, D2-6A; rod1Δ rog3Δ ldb19Δ, JT6675) was assessed by the agar diffusion (halo) bioassay for α-factor-induced growth arrest on YPD medium (15 μg α-factor spotted on each filter disk). Data from one representative experiment are shown. (B) Quantification and statistical analysis of the change in halo diameter, determined as described for panel A, for independent replicate experiments (n = 4). The average halo diameter for control cells was set at 100%, and halo sizes for each mutant were normalized to the control. Error bars indicate ± standard errors of the means (SEM); **, P < 0.0001; *, P < 0.05; n.s., value not statistically significant. (C) Pheromone sensitivity of a MATa bar1Δ strain (JT5915) and otherwise isogenic ldb19Δ bar1Δ (JT5916), rod1Δ rog3Δ bar1Δ (JT5917), and rod1Δ rog3Δ ldb19Δ bar1Δ (JT6674) derivatives was determined as described for panel A in response to the indicated amounts of α-factor (150 ng to 15 μg). Values represent the averages from independent replicate experiments (n = 5); errors bars indicate ±SEM. (D) Pheromone sensitivity of a MATa sst2Δ (JT6755) strain and otherwise isogenic ldb19Δ sst2Δ (JT6660), rod1Δ rog3Δ sst2Δ (JT6702), and rod1Δ rog3Δ ldb19Δ sst2Δ (JT6662) derivatives was determined in response to the indicated amounts of α-factor. Values represent the averages from independent replicate experiments (n = 3); errors bars indicate ±SEM. (E) Pheromone sensitivity of a MATa sst2Δ strain (JT5919) carrying the GEV chimera for β-estradiol-induced expression of genes under GAL promoter control and containing either empty vector (high-copy-number URA3-marked 2μm DNA plasmid) or the same vector harboring the indicated α-arrestin (as a fusion to GST) under GAL promoter control was determined on SC-Ura, as described for panel A, using 15 μg of α-factor spotted on the filter disk after induction with β-estradiol (see Materials and Methods). Data from one representative experiment (n = 5) are shown. (F) Confirmation of α-arrestin expression. Proteins from whole-cell extracts of the cells shown in panel E were prepared, resolved by SDS-PAGE, and analyzed by immunoblotting with the indicated antibodies. Data from one representative experiment (n = 5) are shown. MW, molecular weight in thousands.

FIG 2

Pheromone signaling is more persistent in cells lacking Ldb19, Rod1, and Rog3. (A) Cultures of a MATa bar1Δ strain (JT5915) and an otherwise isogenic ldb19Δ rod1Δ rog3Δ bar1Δ derivative (JT6674) were grown to mid-exponential phase and then treated with 15 nM α-factor for the indicated times. Samples were withdrawn at the indicated time points and rapidly chilled on ice. The cells were collected by centrifugation and extracted, and proteins in the resulting whole-cell lysates were resolved by SDS-PAGE and analyzed by immunoblotting with the indicated antibodies. Data from one representative experiment (n = 3) are shown. MW, molecular weight in thousands. (B) Values represent the mean pixel intensities for the phosphorylated and total Fus3 bands, determined as described for panel A, from the three independent replicate experiments. Error bars indicate ±SEM; *, P < 0.01. (C) Cultures of a MATa bar1Δ FUS1_prom-eGFP strain (JT6686) and an otherwise isogenic ldb19Δ rod1Δ rog3Δ bar1Δ FUS1_prom-eGFP derivative (JT6668) were grown to mid-exponential phase. Samples of these cultures were withdrawn, and the distribution of fluorescent cells was determined using a fluorescence-activated cell sorter (model FC500; Beckman-Coulter) at the Flow Cytometry Facility of the UC Berkeley Cancer Research Laboratory. The remainder of each culture was treated with 15 nM α-factor for 2 h, and the profile of fluorescent cells in each culture was redetermined. (D) The average fold change in the level of GFP fluorescence determined from the ratio of the areas under the curves of uninduced and pheromone-induced cells of the indicated genotypes for independent replicate experiments (n = 3) performed as described for panel C. Error bars indicate ±SEM.

FIG 3

Increased abundance of Ste2 in the plasma membrane in cells lacking Ldb19 and/or Rod1 and Rog3. (A) A MATa strain expressing an integrated copy of Ste2-mCherry (as the sole copy of this receptor) from the native STE2 promoter at the endogenous STE2 locus on chromosome VI (JT6677) and otherwise isogenic ldb19Δ (JT6678), rod1Δ rog3Δ (JT6679), and rod1Δ rog3Δ ldb19Δ (JT6680) derivatives were examined by fluorescence microscopy. Representative images were recorded as described in Materials and Methods. (B) The same as panel A, except the cells expressed an integrated copy of Ste2-GFP as the sole source of the receptor, and 9arrΔ (JT6757) and aly1Δ aly2Δ (JT6762) derivatives were also visualized. (C) Mean intensity of PM fluorescence was quantified for each of the indicated strains (≥50 cells each) using ImageJ and plotted in arbitrary units (a.u.). Values significantly different from those of the control cells were assessed using a one-way analysis of variance test with Tukey's post hoc comparison (165). *, P < 0.0001.

FIG 4

P/VPXY motifs in Ldb19, Rod1, and Rog3 are required for Rsp5 binding and Rsp5-mediated ubiquitinylation. (A) Schematic depiction of the primary structures of Ldb19, Rod1, and Rog3. Residues (numbers below each bar) comprising the arrestin fold (blue) in Ldb19 according to reference and in Rod1 and Rog3 as predicted by the Phyre2 modeling algorithm (166) are shown, and positions of the consensus Rsp5-binding motifs, PPXY and VPXY, and reported ubiquitinylated Lys residue(s) (black lines) are indicated. (B) Cultures of a GEV derivative of vacuolar protease-deficient strain BJ5459 (160) expressing the indicated α-arrestin or the derived P/VPXY substitution mutant (as a GST fusion from the GAL promoter) were grown to mid-exponential phase. Protein expression was induced with β-estradiol, and the cells were harvested by centrifugation and ruptured by vigorous vortex mixing with glass beads. GST fusions in the resulting extracts (pink dots) were captured by binding to glutathione-agarose beads. After washing, the bound proteins were resolved by SDS-PAGE and analyzed by immunoblotting with the indicated antibodies. MW, molecular weight in thousands; Long Exp., long exposure. (C) The Rsp5-catalyzed and time-dependent ubiquitinylation of an [³⁵S]Met-labeled α-arrestin or its cognate PPXY substitution mutant (pink dots), prepared by coupled in vitro transcription-translation, was performed and analyzed using a phosphorimager as described in Materials and Methods. add., addition; Ub, ubiquitinylation.

FIG 5

Ldb19 and Rod1, but not Rog3, require Rsp5 binding to downregulate pheromone signaling. (A) Pheromone sensitivity of a MATa bar1Δ strain (JT5915) and an otherwise isogenic ldb19Δ bar1Δ derivative (JT5916) carrying either empty vector (HIS3-marked CEN plasmid) or the same vector expressing wild-type LDB19 or derivatives containing point mutations in each or both of its PPXY motifs binding was determined as described for Fig. 1A, except that the medium was SC-His. (B) Quantification and statistical analysis of the change in halo diameter, determined as described for panel A, for independent replicate experiments (n = 3). The average halo diameter for control cells was set at 100%, and the halo sizes for each mutant were normalized to that of the control. Error bars indicate ±SEM; *, P < 0.0001. (C) Pheromone sensitivity of cultures of MATa sst2Δ GEV cells (JT5919) overexpressing either Rod1 or Rog3, as indicated, or the derived PPXY point mutants under the control of the GAL promoter on a high-copy-number URA3-marked 2μm DNA plasmid was determined as described in the legend to Fig. 1E. Data from one representative experiment (n = 5) are shown. (D) Proteins from whole-cell extracts of the cells shown in panel C were prepared, resolved by SDS-PAGE, and analyzed by immunoblotting with the indicated antibodies. Data from one representative experiment (n = 5) are shown. MW, molecular weight in thousands. (E) Pheromone sensitivity of cultures of MATa sst2Δ GEV cells (JT5919) or an ldb19Δ rod1Δ rog3Δ sst2Δ GEV derivative (JT6716) overexpressing either Rog3 or the derived cognate P/VPXY point mutants, under the control of the GAL promoter on a high-copy-number URA3-marked 2μm DNA plasmid, was determined as described for panel C. Data from one representative experiment (n = 3) are shown. (F) Confirmation of protein expression, as described for panel D. Data from one representative experiment (n = 3) are shown.

FIG 6

Ubiquitinylation of Ldb19, Rod1, and Rog3 is not required for downregulation of pheromone signaling. (A) For analysis of in vivo ubiquitinylation of Ldb19, cultures of a GEV derivative of BJ5459 (JT6743) were grown to mid-exponential phase. Expression of either Ldb19 or the derived K486R substitution mutant (as a GST fusion from the GAL promoter) was then induced with β-estradiol for 3 h, and lysates were immediately prepared and analyzed as described for Fig. 4B using the indicated antibodies. For analysis of in vivo ubiquitinylation of Rod1, cells were grown to mid-exponential phase in 4% raffinose. Expression of either Rod1 or the derived 4K-to-R mutant (as a GST fusion from the GAL promoter) was induced by addition of galactose (2% final concentration) for 3 h and then shifted to dextrose medium (2% final concentration) for 5 min, and lysates were immediately prepared and analyzed as described for Fig. 4B. MW, molecular weight in thousands. (B) Pheromone sensitivity of a MATa bar1Δ (JT5915) strain and an isogenic ldb19Δ bar1Δ derivative (JT5916) carrying either empty vector (HIS3-marked CEN plasmid) or the same vector expressing wild-type LDB19 or the K486R substitution mutant was determined as described in the legend to Fig. 1A, except that the medium was SC-His. (C) Quantification and statistical analysis of the change in halo diameter, determined as described for panel B, from independent replicate experiments (n = 3). The average halo diameter for control cells was set at 100%, and the halo sizes for each mutant were normalized to the control. Error bars indicate ±SEM; *, P < 0.0001. (D) Pheromone sensitivity of cultures of MATa sst2Δ GEV cells (JT5919) expressing either GST-Rod1 or GST-Rog3, as indicated, or the derived 4K-to-R substitution mutants under the GAL promoter on a high-copy-number URA3-marked 2μm DNA plasmid was determined as described for Fig. 1E. Data from one representative experiment (n = 3) are shown.

FIG 7

Ldb19, Rod1, and Rog3 bind preferentially to the C-terminal tail of Ste2. (A, lower) GST and GST-Ste2(297-431) (the tail), which was constructed, expressed in E. coli, and purified to apparent homogeneity as described previously (21), were used to coat glutathione-agarose beads to an equivalent level. (Upper) In this representative experiment, samples of the beads were incubated in duplicate with equivalent amounts (cpm) of [³⁵S]Met-labeled molecules of the indicated α-arrestins (pink dots) and prepared by coupled in vitro transcription and translation. The amount of bound radioactive protein detected was quantified using a phosphorimager as described in Materials and Methods. (B) Average fold increase in the level of radioactivity bound to the GST-Ste2^tail construct relative to that bound to the GST control for the indicated α-arrestins in independent replicates (n = 3), each performed essentially as described for panel A. Error bars indicate ±SEM. The dashed line indicates behavior expected for a negative control [i.e., no increase in binding to GST-Ste2(297-431) compared to that of GST alone, yielding a ratio of 1]. (C) Same as described for panel A, except that binding of in vitro-transcribed and -translated Rog3 and Rog3^Δ400 are compared. (D) The average fold increase in binding, determined in panel B for independent replicates (n = 3), each performed as described for panel C.

FIG 8

Rod1-mediated desensitization requires calcineurin-dependent dephosphorylation. (A) A single PXIXIT motif mediates CN-Rod1 interaction. Cultures of strain JRY11, which produces Cna1-GFP from the native CNA1 promoter at the endogenous CNA1 locus on chromosome XII, and also expressing, as indicated, either GST alone, GST-Rod1, or GST-Rod1^AQAKAA, were grown to mid-exponential phase, harvested, and lysed, and proteins in the resulting extracts were captured on glutathione-agarose beads, resolved by SDS-PAGE, and analyzed with the indicated antibodies. MW, molecular weight in thousands. (B) Rod1 is phosphorylated at CN-sensitive sites. Cultures of strain BY4741 (WT) or otherwise isogenic cnb1Δ (BY4741 cnb1Δ) and cna1Δ cna2Δ (JT5574) derivatives, as indicated, expressing either GST-Rod1 or GST-Rod1^AQAKAA were grown to mid-exponential phase and stimulated with 200 mM CaCl₂ to activate CN in either the absence or presence (+) of the CN inhibitor FK506 (FK). After harvesting and lysis, proteins in the resulting extracts were purified by capture on glutathione-agarose beads (lanes 1 to 4). Samples of the material shown in lanes 1 to 4 then were either left untreated treated or were treated (+) with lambda phosphatase (λ) in either the absence or presence (+) of phosphatase inhibitors (PPase inhibitors), and the resulting products were separated under SDS-PAGE conditions that permit resolution of phospho-isoforms and analyzed with anti-GST antibodies. exp., exposure. (C) Lack of CN binding reduces Rod1-mediated adaptation. Pheromone sensitivity of MATa sst2Δ GEV (JT5919) cells expressing either GST-Rod1 or GST-Rod1^AQAKAA, as indicated, under the GAL promoter on a high-copy-number URA3-marked 2μm DNA plasmid was determined as described for Fig. 1E. Data from one representative experiment (n = 3) are shown. (D) Absence of CN eliminates Rod1-mediated adaptation but not Rog3-mediated adaptation. Pheromone sensitivity of cultures of MATa sst2Δ GEV (JT5919) cells and isogenic cnb1Δ (JT6694) and cna1Δ cna2Δ (JT6695) derivatives, as indicated, expressing either GST-Rod1 or GST-Rog3 under the GAL promoter on a high-copy-number URA3-marked 2μm DNA plasmid was determined as described for Fig. 1E. Data from one representative experiment (n = 3) are shown. (E) Confirmation of protein expression. Whole-cell extracts of the cells used in panels C and D were prepared, resolved by SDS-PAGE, and analyzed by immunoblotting with the indicated antibodies. Here, phospho-isoforms were not separated, because different SDS-PAGE conditions were used. Data from one representative experiment (n = 3) are shown.

FIG 9

Distinct mechanisms of Ste2 downregulation by the α-arrestins Ldb19, Rod1, and Rog3. The α-factor receptor (Ste2), a polytopic integral membrane protein, exists primarily in three conformational states. In näive cells, Ste2 undergoes spontaneous stochastic dissociation from its cognate heterotrimeric G protein (not shown for clarity) at a certain rate and thereby becomes destabilized. When it does so, it may misfold. Current evidence suggests that Ldb19/Art1 has a primary role in a PM quality-control pathway that mediates Rsp5-dependent ubiquitinylation and endocytic removal of such misfolded PM proteins. In the presence of α-factor, Ste2 undergoes a ligand-induced conformational change that activates and dissociates its cognate G protein; however, in this case, the receptor is stabilized by bound pheromone. Because Rog3-imposed inhibition of pheromone signaling does not obligatorily require its association with or modification by Rsp5, it may act similarly to classical arrestin or β-arrestin by binding to the C-terminal tail of the receptor and sterically preventing additional rounds of G-protein activation by the pheromone-bound receptor. Later during response to pheromone, Ca²⁺ influx will stimulate the CN-dependent dephosphorylation of Rod1, making Rod1 competent to mediate Rsp5-dependent ubiquitinylation and endocytic removal of the pheromone-bound receptor, a prime example of a stimulus-induced, late-stage, negative-feedback control.