Signaling of human frizzled receptors to the mating pathway in yeast

Abstract

Frizzled receptors have seven membrane-spanning helices and are considered as atypical G protein-coupled receptors (GPCRs). The mating response of the yeast Saccharomyces cerevisiae is mediated by a GPCR signaling system and this model organism has been used extensively in the past to study mammalian GPCR function. We show here that human Frizzled receptors (Fz1 and Fz2) can be properly targeted to the yeast plasma membrane, and that they stimulate the yeast mating pathway in the absence of added Wnt ligands, as evidenced by cell cycle arrest in G1 and reporter gene expression dependent on the mating pathway-activated FUS1 gene. Introducing intracellular portions of Frizzled receptors into the Ste2p backbone resulted in the generation of constitutively active receptor chimeras that retained mating factor responsiveness. Introducing intracellular portions of Ste2p into the Frizzled receptor backbone was found to strongly enhance mating pathway activation as compared to the native Frizzleds, likely by facilitating interaction with the yeast Galpha protein Gpa1p. Furthermore, we show reversibility of the highly penetrant G1-phase arrests exerted by the receptor chimeras by deletion of the mating pathway effector FAR1. Our data demonstrate that Frizzled receptors can functionally replace mating factor receptors in yeast and offer an experimental system to study modulators of Frizzled receptors.

Figure 1. Schematic representation of the GPCR-controlled mating pathway in S. cerevisiae.

Shown is the situation for mating type a haploid yeast cells. Cells normally express the α-factor receptor Ste2p – Frizzled receptors are introduced by ectopic expression. Ste2p activation through its ligand α-factor leads to GDP-GTP exchange in the Gα protein Gpa1p, and dissociation of the heterotrimeric G protein. The resulting free βγ subunits trigger activation of a MAP kinase cascade leading to the transcription of pheromone-responsive genes such as FUS1, as well as to G1 growth arrest via inhibition of the cyclin-dependent kinase inhibitor FAR1 gene product. A mating pathway activation growth assay using wildtype yeast cells (MH272-1da) is shown bottom left. The strain was spotted in serial dilutions (∼1E4, 1E3, 1E2, 1E1 cells in duplicate) onto complete minimal medium, and treated with 1 µg α-factor spotted directly onto the yeast patches as indicated.

Figure 2. Expression strategy and expression tests for human Frizzled receptors in yeast.

A. Expression growth test for Fz1 using the nutritional reporter Ura3p. Schematically shown as two-dimensional snake-plots are Fz1, or Fz1 with the N-terminal signal sequence (s) of the α-factor receptor Ste2p fused upstream of the cysteine-rich-domain (CRD), and carrying a C-terminal Ura3p reporter. The constructs were used to transform URA3-deficient yeast cells, and plated equally onto control plates (C) and minimal medium plates lacking uracil (-U). B. Subcellular localization and Western blot analysis. The upper panel shows a confocal micrograph and the corresponsing bright field view of s-Fz1-GFP expressing yeast cells immobilized by agarose-embedding. The lower panel shows the Western blot analysis using s-Fz1-myc expressing yeast cells. Following galactose-induction, crude protein extracts were prepared and resolved on a 4–12% gradient SDS-PAGE system. Detection was performed using an anti-c-myc monoclonal antibody-peroxidase conjugate. Molecular weight standards in kDa are indicated. The control extract was derived from a yeast strain transformed with the empty expression vector. C. The s-Fz2 construct for the expression of human Frizzled 2 was analogous to s-Fz1.

Figure 3. Growth modulation assay of wildtype yeast cells (MH272-1a) overexpressing human Frizzled receptors and receptor chimeras.

s-Fz1 and s-Fz2 (panel A) as well as receptor chimeras with the Ste2 receptor (B and C) were expressed in a galactose-inducible system, and effects were compared to α-factor mediated growth arrest. Corresponding receptor schemes are shown in Fig. 2 and Fig. 5. Yeast strains transformed with plasmid constructs as indicated were pre-cultured in glucose-containing minimal medium (repressed), and spotted in serial dilutions (∼1E4, 1E3, 1E2, 1E1 cells) onto glucose- or galactose- (induced) containing minimal medium agar plates selective for the presence of the plasmid. Growth effects were recorded after 4 and 7 days. Controls were the empty expression vector and Ste2 and Edg2 receptors. The α-factor - mediated growth arrest assay (right panel) was done using wildtype yeast cells (MH272-1da), as well as isogenic ste2 cells, treated with 1 µg α-factor spotted directly onto the yeast patches as indicated.

Figure 4. Effect of galactose-inducible overexpression of human Frizzled receptors (s-Fz1 and s-Fz2) in the P_FUS1-luciferase- based reporter system.

Basal activities of s-Fz1 and s-Fz2 in absence of added ligand were determined. Schemes for receptor constructs used are shown in Fig. 2. Yeast strain MC18 ste2 sst2 expressing plasmid-encoded Gpa1p was employed. Crude extracts from glucose-repressed or galactose-induced yeast strains were assayed in duplicates for luciferase activity. Normalization was done based on total protein concentration. Signals are expressed relative to the receptor-deficient “OFF”-strain grown under induced conditions. Ste2p- (treated and untreated with 1 µM α-factor), Edg2-overexpressing strains, as well as the GPA1-deficient “ON”-strain were included as further controls. Bars represent average+/−range.

Figure 5. Construction schemes and subcellular localization of chimeras between the human Frizzled and S. cerevisiae Ste2 receptors.

Receptor homology modeling was used to predict the transmembrane segments (see Fig. 6). A. This chimera type replaces the predicted intracellular portions of Ste2p with the corresponding ones of either Fz1 or Fz2, resulting in chimeras termed Ste2-Fz1(i) and Ste2-Fz2(i), respectively, with “(i)” designating intracellular portions. B. The second chimera type replaces the predicted intracellular portions of the human Frizzled receptors with respective ones of Ste2p, and uses the signal sequence of Ste2p (“s”) upstream of the cysteine-rich-domain (CRD) to give rise to s-Fz1-Ste2(i) and s-Fz2-Ste2(i). The fully synthetic gene sequences and respective amino acid sequences are given in the Supporting Information section of this article (Text S1). The subcellular localization of constructs tagged C-terminally with GFP is shown in the confocal micrographs.

Figure 6. Transmembrane segment predictions for human Fz1, Fz2, and S. cerevisiae Ste2 receptors.

The helices one to seven are shown as amino-acid sequences with indicated position numbers of the full sequence. The predictions were derived by receptor modeling. The transmembrane region assignment was based on a hydrophobicity analysis followed by manual adjustment of the overlap of some key amino acids (e.g., W and P in TM4, TM5, and TM6) with those found in rhodopsin and other reference receptors.

Figure 7. Quantification of receptor expression by Western blot.

Following galactose-induction, crude protein extracts were prepared from the MH272-1da wildtype yeast strain transformed with the indicated plasmids, and resolved on a 4–12% gradient SDS-PAGE system. Detection of the myc-tagged receptors was performed using chromogenic indirect anti-c-myc monoclonal detection. Equal loading was monitored using a polyclonal anti-Kss1p antibody. The control extract was derived from a yeast strain transformed with the empty expression vector. The asterisks indicate the specific signals. Molecular weight standards in kDa are indicated.

Figure 8. Effect of galactose-inducible overexpression of the receptor chimeras between human Frizzled receptors (Fz1 and Fz2) and the S. cerevisiae Ste2 receptor in the P_FUS1-luciferase- based reporter system.

Yeast strain MC18 ste2 sst2 expressing plasmid-encoded Gpa1p was employed. Schemes for receptor constructs used are shown in Fig. 5. A. and B. Basal activities of Ste2-Fz1(i) and s-Fz1-Ste2(i) receptors in absence of added ligand, and comparison to the s-Fz1 overexpressing strain. Crude extracts from glucose-repressed or galactose-induced yeast strains were assayed in duplicates for luciferase activity after 6 h (A.) or 24 h (B). Normalization was done based on total protein concentration. Signals are expressed relative to the receptor-deficient “OFF”-strain grown under induced conditions. Bars represent means+/−range. C. α-factor concentration - response in the MC18 ste2 sst2 strain expressing plasmid-encoded Gpa1p and Ste2-Fz1/2(i) receptor chimeras or Ste2p. The galactose-induced yeast strains were incubated for 4 hours with indicated concentrations of α-factor, and crude extracts were subsequently assayed in quadruplets for luciferase activity. As negative control served a strain transformed with empty vector (“OFF”-strain). Responses are expressed as treated/control values; error bars indicate S.E.M. (N = 4).

Figure 9. Growth behaviour and morphologies of wildtype yeast cells (MH272-1da) overexpressing galactose-inducible s-Fz1 or receptor chimeras between Fz1 and Ste2p in liquid culture.

Schemes for receptor constructs used are shown in Fig. 2 and Fig. 5. As control served a strain transformed with empty vector. Growth curves in glucose- (repressed; (A)), or galactose-containing (induced; (B.)) minimal medium, following washout of glucose, as determined using optical density measurement, or microscopic counting in a hemocytometer slide at the indicated timepoints, respectively. C. Budding indices (ratio of budded to total yeast cells) of the induced cultures were determined using microscopic counting in a hemocytometer slide.

Figure 10. Mating pathway activation growth assay of wildtype yeast cells (MH272-1da), as well as of the isogenic strains carrying far1 or ste2 - alleles.

(A) Yeast strains were spotted in serial dilutions (∼1E4, 1E3, 1E2, 1E1 cells in duplicate) onto complete minimal medium, and treated with 1 µg α-factor spotted directly onto the yeast patches. Growth effects were recorded after 2 and 7 days. (B) Comparative growth modulation assay of wildtype yeast cells (MH272-1a) and respective far1 isogenic strain overexpressing human Frizzled receptors (s-Fz1 and s-Fz2), as well as chimeras with the S. cerevisiae Ste2 receptor in a galactose-inducible expression system. The corresponding receptor schemes are shown in Fig. 2 and Fig. 5, respectively. Yeast strains transformed with plasmid constructs as indicated were pre-cultured in glucose-containing minimal medium (repressed), and spotted in serial dilutions (∼1E4, 1E3, 1E2, 1E1 cells) onto glucose- or galactose- (induced) containing minimal medium agar plates selective for the presence of the plasmid. Growth effects were recorded after 4 to 7 days. Controls were the empty expression vector, as well as overexpressed Ste2 and Edg2 receptors.