Kelch-repeat proteins interacting with the Galpha protein Gpa2 bypass adenylate cyclase for direct regulation of protein kinase A in yeast

Abstract

The cAMP-PKA pathway consists of an extracellular ligand-sensitive G protein-coupled receptor, a G protein signal transmitter, and the effector, adenylate cyclase, of which the product, cAMP, acts as an intracellular second messenger. cAMP activates PKA by dissociating the regulatory subunit from the catalytic subunit. Yeast cells (Saccharomyces cerevisiae) contain a glucose/sucrose-sensitive seven-transmembrane domain receptor, Gpr1, that was proposed to activate adenylate cyclase through the G(alpha) protein Gpa2. Consistently, we show here that adenylate cyclase binds only to active, GTP-bound Gpa2. Two related kelch-repeat proteins, Krh1/Gpb2 and Krh2/Gpb1, are associated with Gpa2 and were suggested to act as G(beta) mimics for Gpa2, based on their predicted seven-bladed beta-propeller structure. However, we find that although Krh1 associates with both GDP and GTP-bound Gpa2, it displays a preference for GTP-Gpa2. The strong down-regulation of PKA targets by Krh1 and Krh2 does not require Gpa2 but is strictly dependent on both the catalytic and the regulatory subunits of PKA. Krh1 directly interacts with PKA by means of the catalytic subunits, and Krh1/2 stimulate the association between the catalytic and regulatory subunits in vivo. Indeed, both a constitutively active GPA2 allele and deletion of KRH1/2 lower the cAMP requirement of PKA for growth. We propose that active Gpa2 relieves the inhibition imposed by the kelch-repeat proteins on PKA, thereby bypassing adenylate cyclase for direct regulation of PKA. Importantly, we show that Krh1/2 also enhance the association between mouse R and C subunits, suggesting that Krh control of PKA has been evolutionarily conserved.

Fig. 1.

Gpa2 and Krh1/2 act antagonistically and independently on several PKA targets. (A) The strains BY4742, Y10152, TP261, TP675, TP676, and TP677 were grown to stationary phase on YPD; samples were taken after 1, 2, and 3 days (white, gray, and dark gray bars, respectively); and trehalose content was determined (OD₆₀₀ and glucose consumption were followed to verify that all strains reached saturation at approximately the same time). (B) The strains were grown and samples were taken as in A. Expression of HSP12 and ACT1 (as an internal standard) were measured by quantitative PCR. (C and D) The strains S18-1D, TP626, and TP701 were grown to stationary phase on yeast extract/peptone glycerol. cAMP (C) and trehalase (D) activity were measured after addition of 100 mM glucose.

Fig. 2.

Both adenylate cyclase and Krh1 preferentially bind to active Gpa2. (A) GST and GST-Gpa2 were purified from bacteria, preloaded with the indicated nucleotide, and incubated with yeast extracts from the strains MV601 and MV600, expressing HA₃-Cyr1 and Krh1-HA₃, respectively. Quantification indicates that the intensity of the GDP-bound fraction of Krh1-HA is 61% of the GTP-bound fraction. The input fraction represents 10% of the total amount added to each binding reaction. (B) (Left) Extracts from yeast strains expressing GST-Gpa2 and GST-Gpa2^Q300L were mixed with extract from strain MV601 containing HA₃-Cyr1 and incubated with glutathione-agarose. (Right) GST-Gpa2 and GST-Gpa2^Q300L were coexpressed with Krh1-HA₃ (using strains MV605 and MV606) and incubated with glutathione-agarose. Quantification indicates that the intensity of the Gpa2-bound fraction of Krh1-HA is 68% of the Gpa2^Q300L-bound fraction. The input fraction represents 10% of the total amount added to each binding reaction.

Fig. 3.

Absence of Krh1 and Krh2 or activation of Gpa2 reduces the cAMP requirement for growth of an adenylate cyclase mutant. The strains with indicated genotypes (TP433 and TP390 in A; TP433, TP506, and TP519 in B) were grown on YPD supplemented with 3 mM cAMP to midexponential phase, spotted in serial 2-fold dilutions on YPD plates with the indicated cAMP concentrations, and incubated at 30°C for 3 days.

Fig. 4.

Both the catalytic and the regulatory subunits of PKA are required for Krh control. (A) Trehalose determination in strains of the indicated genotypes (TP573, TP575, TP572, and TP571) grown to stationary phase on YPD. Samples were taken during growth on YPD after 1, 2, and 3 days (white, gray, and dark gray bars, respectively; OD₆₀₀ and glucose consumption were followed to verify that all strains reached saturation at approximately the same time). The data shown in the two graphs were measured in a single experiment but are illustrated separately because of the large difference in trehalose levels in the yak1Δ tpk1–3Δ background compared with the yak1Δ mutant. (B) Trehalose determination in strains of the indicated genotypes (S18-1D, TP408, TP626, RS13-58A-1, TP415, and TP627).

Fig. 5.

Krh1 binds to PKA by means of the catalytic subunits. (Left) The indicated GST fusions were purified from bacteria and incubated with cell extracts from strain MV600 expressing Krh1-HA₃. (Right) His-6-Tpk1 expressed alone in bacteria or coexpressed with untagged Bcy1 was purified and incubated with cell extracts from strain MV600 expressing Krh1-HA₃. The input fraction represents 10% of the total amount added to each binding reaction.

Fig. 6.

Krh1 and Krh2 stimulate the interaction between Tpk1 and Bcy1 in vivo. Two-hybrid reporter strains PJ69-4A, TP636, and TP637 transformed with the indicated GAL4 fusion constructs (R, Bcy1; C, Tpk1), were grown overnight in liquid selective medium (SD-Leu-Trp) and analyzed for β-gal activity and total protein content (A) or spotted on a selective SD-Leu-Trp-Ade plate (B). The values in A are an average of two independent cultures; the error bars represent the range between those values.

Fig. 7.

A model for the role of Gpa2 and Krh1/2 in PKA signaling. Upon activation of Gpa2, PKA is stimulated by means of two distinct mechanisms: direct activation of adenylate cyclase and inhibition of Krh1/2-mediated down-regulation of PKA. Krh1/2 regulate PKA activity by stimulating the association between Tpk1 and Bcy1. As explained in Discussion, other (nutrient) signals besides the availability of sugars may affect Krh regulation of PKA.