Global Regulation of a Differentiation MAPK Pathway in Yeast

Abstract

Cell differentiation requires different pathways to act in concert to produce a specialized cell type. The budding yeast Saccharomyces cerevisiae undergoes filamentous growth in response to nutrient limitation. Differentiation to the filamentous cell type requires multiple signaling pathways, including a mitogen-activated protein kinase (MAPK) pathway. To identify new regulators of the filamentous growth MAPK pathway, a genetic screen was performed with a collection of 4072 nonessential deletion mutants constructed in the filamentous (Σ1278b) strain background. The screen, in combination with directed gene-deletion analysis, uncovered 97 new regulators of the filamentous growth MAPK pathway comprising 40% of the major regulators of filamentous growth. Functional classification extended known connections to the pathway and identified new connections. One function for the extensive regulatory network was to adjust the activity of the filamentous growth MAPK pathway to the activity of other pathways that regulate the response. In support of this idea, an unregulated filamentous growth MAPK pathway led to an uncoordinated response. Many of the pathways that regulate filamentous growth also regulated each other’s targets, which brings to light an integrated signaling network that regulates the differentiation response. The regulatory network characterized here provides a template for understanding MAPK-dependent differentiation that may extend to other systems, including fungal pathogens and metazoans.

Cell differentiation is the process by which cells undergo specialization to produce different cell types with different functions. Cell-type specialization can result from execution of an intrinsic developmental program and also in response to extrinsic cues. The process of cell differentiation is one of exquisite precision: cells undergo complete morphogenetic restructuring in a specific spatiotemporal context (Kholodenko et al. 2010). Multiple signaling pathways collaborate to control cell differentiation responses. For example, the activity of the Wnt and Hippo pathways is integrated at multiple levels to coordinate development (McNeill and Woodgett 2010). A critical problem in the field of cell differentiation is to elucidate how signals from different pathways become integrated to produce a cohesive response. This problem is relevant from the standpoint of human health, because misregulation of differentiation pathways is an underlying cause of developmental problems and diseases such as cancer (Wagner and Nebreda 2009).

Depending on ploidy and growth condition, the budding yeast Saccharomyces cerevisiae can differentiate into different cell types. Haploid yeast undergoes morphological changes in response to secreted pheromones to mate and form diploids (Bardwell 2005; Dohlman and Slessareva 2006; Merlini et al. 2013). Diploid yeast starved for carbon and nitrogen initiate a meiotic program known as sporulation (Neiman 2011). Haploid and diploid yeast starved for only carbon or nitrogen undergoes filamentous (or invasive/pseudohyphal) growth (Gimeno et al. 1992; Cullen and Sprague 2000, 2012). During filamentous growth, major changes occur to cell polarity (Gimeno et al. 1992; Roberts and Fink 1994; Pruyne and Bretscher 2000; Cullen and Sprague 2002; Bi and Park 2012), cell-cycle progression (Kron et al. 1994; Edgington et al. 1999), and cell adhesion (Lambrechts et al. 1996; Lo and Dranginis 1998; Guo et al. 2000), which results in formation of branched chains of interconnected invasive filaments. Filamentous cells form complex communities during biofilm formation (Reynolds and Fink 2001; Verstrepen and Klis 2006; Bojsen et al. 2012). Many fungal species undergo filamentous growth. In pathogens, differentiation to filamentous/hyphal cells in biofilms is critical for pathogenicity (Lo et al. 1997; Wendland 2001; Nobile et al. 2006; Sohn et al. 2006). Budding yeast therefore provides a convenient genetic system to define the pathways that regulate filamentous growth and has provided insights into the genetic basis of fungal pathogenesis and eukaryotic differentiation.

Signal transduction pathways regulate filamentous growth and control the changes that occur in response to nutrient limitation (Zhao et al. 2007). Among the pathways that regulate filamentous growth in yeast is a MAPK pathway called the filamentous growth MAPK pathway (Supporting Information, Figure 1A). MAPK pathways are evolutionary conserved signaling modules that regulate diverse responses in eukaryotes (Raman et al. 2007). The filamentous growth MAPK pathway is composed of plasma-membrane sensors (Msb2p, Sho1p, and Opy2p) (O’Rourke and Herskowitz 1998; Cullen et al. 2004; Wu et al. 2006; Yamamoto et al. 2010; Karunanithi and Cullen 2012) that connect to a Rho-type GTPase (Cdc42p; Bi and Park 2012) and a kinase cascade consisting of a p21-activated kinase (Ste20p; Peter et al. 1996; Leberer et al. 1997) and MAPK module (including the MAPKKK Ste11p, MAPKK Ste7p, and MAPK Kss1p; Roberts and Fink 1994). The MAP kinase Kss1p regulates the activity of two transcription factors (Ste12p and Tec1p; Madhani and Fink 1997; Madhani et al. 1997) that induce target genes (Madhani et al. 1999) by binding to well-defined promoter elements (Zeitlinger et al. 2003; Chou et al. 2006).

Figure 1.

Positive and negative regulators of the filamentous growth MAPK pathway. (A) The filamentous growth MAPK pathway is shown with components that were identified by the screen represented as positive regulators (green circles) and negative regulators (red circles). Cdc24p and Cdc42p are essential proteins and were not tested here, and Ste7p was not present in the collection. Opy2p is an established regulator of the filamentous growth MAPK pathway (Yang et al. 2009, 2010; Karunanithi and Cullen 2012) and showed a defect in FRE–lacZ expression, but the levels fell below the range of statistical significance (Table S3). (B) Example of the plate-washing assay. Equal concentrations of cells were spotted onto YEPD media. Cells were grown for 2 days. Plates were photographed (top) and washed in a stream of water to reveal invaded cells (bottom). Examples of hyper- and hypoinvasive growth mutants are listed at right in reference to strains lacking a positive (ste11Δ) and negative (dig1Δ) regulator. (C) Mutants showing elevated pFRE–lacZ expression and hyperinvasive growth. β-Galactosidase assays were performed in at least duplicate. Blue bar, wild-type control strain; red bars, mutants tested. Values are expressed in Miller units (U). Error bars represent standard deviation between independent trials. P-values and raw data provided in Table S3. (D) Mutants showing reduced pFRE–lacZ expression and reduced invasive growth. Blue bar, wild-type control strain; green bars, mutants tested. See C for details. The pFRE–lacZ activity of the kss1Δ did not fall within the 1.5-fold cutoff but is shown as a reference.

In addition to the MAPK pathway, other pathways also regulate filamentous growth. Major nutrient regulatory pathways include the Ras2p–cAMP–protein kinase A (PKA) pathway (Toda et al. 1985; Gimeno et al. 1992; Mosch et al. 1996; Colombo et al. 1998; Robertson and Fink 1998; Mosch et al. 1999; Rupp et al. 1999), the AMP-dependent kinase (AMPK) Snf1p and transcriptional repressors Nrg1p and Nrg2p (Celenza and Carlson 1989; Woods et al. 1994; Lesage et al. 1996; Cullen and Sprague 2000; McCartney and Schmidt 2001; Kuchin et al. 2002), the target of rapamycin (TOR) pathway, which responds to nitrogen availability (Beck and Hall 1999; Cardenas et al. 1999; Bruckner et al. 2011), and the mitochondrial retrograde (RTG) pathway (Sekito et al. 2002; Liu et al. 2003; Liu and Butow 2006), which senses changes in metabolic respiration (Aun et al. 2013). The pH sensing Rim101p pathway (Lamb et al. 2001; Barrales et al. 2008), lipid-responsive transcription factor Opi1p (White et al. 1991; Reynolds 2006), tRNA modification complex Elongator [ELP, (Krogan and Greenblatt 2001; Winkler et al. 2001; Petrakis et al. 2004; Li et al. 2007; Svejstrup 2007)], and chromatin remodeling complex Rpd3p(L) (Carrozza et al. 2005; Barrales et al. 2008; Ryan et al. 2012) also regulate filamentous growth. These proteins represent only a subset of a large collection of regulators identified in S. cerevisiae and Candida albicans by gene expression profiling (Madhani et al. 1999; Roberts et al. 2000; Carlisle and Kadosh 2013), genetic screens (Lorenz et al. 2000; Palecek et al. 2000; Barrales et al. 2008), and systematic genome-wide approaches, including large-scale deletion and overexpression studies (Jin et al. 2008; Bharucha et al. 2011; Shively et al. 2013), mass spectrometry (MASS SPEC) approaches (Xu et al. 2010; Zhang et al. 2013) and analysis of ordered deletion collections made in the filamentous (Σ1278b) background (Dowell et al. 2010; Ryan et al. 2012). A critical challenge is to understand how many different proteins and pathways come together to produce a new cell type.

Several of the pathways that regulate filamentous growth can regulate each other’s activities. A landmark finding comes from the discovery that the Ras2p pathway regulates the filamentous growth MAPK pathway (Mosch et al. 1996). More recently, the ELP (Abdullah and Cullen 2009), Rim101, RTG, Rpd3p(L), and Opi1p pathways have also been shown to regulate the filamentous growth MAPK pathway (Chavel et al. 2010). These pathways control expression of the gene encoding one of the plasma-membrane sensors for the filamentous growth MAPK pathway, Msb2p (Chavel et al. 2010). It has also been shown that the major transcription factors that regulate filamentous growth regulate each other's targets, which creates hubs where signal integration events are coordinated (Borneman et al. 2006). One hub is the FLO11 promoter, where multiple transcription factors converge to fine tune cell adhesion (Rupp et al. 1999). Likewise, the major protein kinases that regulate filamentous growth function in an interdependent network (Bharucha et al. 2008). Therefore, signal coordination occurs at multiple levels to regulate the filamentous growth response.

Here we examine the question of signal integration by performing a genetic screen with an ordered deletion collection in the filamentous (Σ1278b) background (Ryan et al. 2012). This effort, combined with hypothesis-based testing, identified 97 new regulators of the filamentous growth MAPK pathway, which map to known regulatory pathways and provide entirely new connections. Using the screen as a platform, we examin questions related to network connectivity. We show that tuning the activity of the filamentous growth MAPK pathway to the other pathways is critical to producing a coordinated response. We also show that several of the key pathways that regulate filamentous growth also regulate each other's targets. Thus, an integrated network regulates the filamentous growth response. We speculate that similarly coordinated networks coordinate cell differentiation responses in other systems.

Materials and Methods

Strains, plasmids, and microbiological techniques

Filamentous growth was evaluated in the Σ1278b strain background (Liu et al. 1996). The haploid gene deletion collection constructed in the Σ1278b strain background has been described (Ryan et al. 2012) and was generously provided by C. Boone. pFRE–lacZ was provided by H. Madhani (Madhani et al. 1997). The YCp–Cdc12–GFP was provided by J. Pringle (Fares et al. 1996). The pste12::URA3 plasmid was provided from G. Sprague (McCaffrey et al. 1987). p8XCRE–lacZ was provided by H. Saito (Tatebayashi et al. 2006). pCIT2–lacZ was provided by the Liu lab and has been described (Liu and Butow 1999).

Standard laboratory conditions were used to grow yeast and bacterial cultures (Rose et al. 1990). Escherichia coli was grown in LB and 2XYT media. Yeast was grown in rich media YEPD (2% glucose) or YEP–GAL (2% galactose) or synthetic complete media at 30° unless otherwise noted. Yeast strains are listed in Table 1. Gene deletions were constructed using antibiotic resistance markers (Goldstein and McCusker 1999) or auxotrophic markers amplified by PCR and introduced into yeast by lithium acetate transformation by standard methods as described (Chavel et al. 2010). The plate washing (Roberts and Fink 1994) and single-cell invasive growth assays (Cullen and Sprague 2000) were used to measure filamentous growth. Colony morphology was examined by visual inspection on YEPD media (Granek and Magwene 2010; Voordeckers et al. 2012). Functional analysis of the MAPK regulatory genes came from SGD (http://www.yeastgenome.org).

Table 1. Yeast strains used in this study.

Strain	Genotype	Reference
PC313a	MATa ura3-52	Liu et al. (1993)
PC538	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52	Cullen et al. (2004)
PC539	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 ste12::KLURA3	Cullen et al. (2004)
PC563	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 bud8::KLURA3	Cullen and Sprague (2002)
PC586	MATα ura3-52 leu2	Cullen et al. (2004)
PC622	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 GAL-SHO1	Cullen et al. (2004)
PC949	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 pop2::KanMX6	This study
PC950	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 ccr4::KanMX6	This study
PC999	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 MSB2-HA	Cullen et al. (2004)
PC1083	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 MSB2-HA GAL-MSB2::KanMX6	Cullen et al. (2004)
PC1415	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 bni1::KLURA3	Cullen and Sprague (2002)
PC1516	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 MSB2-HAΔ100-818	Cullen et al. (2004)
PC1558	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 sho1::HYG ssk1::NAT	Pitoniak et al. (2009)
PC1621	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 MSB2-HAΔ100-818 GAL-SHO1::GENT	This study
PC1625	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 GAL-MSB2-HA::NAT GAL-SHO1::GENT	This study
PC1895	MATa ura3-52 leu2::HYG	This study
PC2043	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 FLO11-HA::KanMX6	Karunanithi et al. (2010)
PC2061	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 ssk1::NAT ste11::KLURA3	Pitoniak et al. (2009)
PC2112	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 leu2::HYG lacZ::NAT tec1::LEU2	Vadaie et al. (2008)
PC2360	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 ras2::NAT	Chavel et al. (2010)
PC2362	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 ira1::NAT	Chavel et al. (2010)
PC2511	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 ras2::NAT ste12::KLURA3	This study
PC2515	MATα ura3-52 leu2 flo8::NAT	Chavel et al. (2010)
PC2532	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 flo8::HYG	Chavel et al. (2010)
PC2534	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 pde2::HYG	Chavel et al. (2010)
PC2535	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 gpa2::NAT	Chavel et al. (2010)
PC2537b	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 gpr1::KLURA3	Chavel et al. (2010)
PC2588	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 tpk1::NAT	Chavel et al. (2010)
PC2618	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 grr1::KLURA3	This study
PC2622	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 snf8::HYG	This study
PC2633	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 sdc25::NAT	Chavel et al. (2010)
PC2688	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 MSB2-HA ste12::KLURA3	This study
PC2690	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 MSB2-HA ras2::KLURA3	This study
PC2763	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 elp2::KLURA3	Abdullah and Cullen (2009)
PC2845	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 FLO11-HA::KanMX6 gal11:KLURA3	This study
PC2945	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 MSB2-HA rxt2::KLURA3	Chavel et al. (2010)
PC2952	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 MSB2-HA hda1::KLURA3	Chavel et al. (2010)
PC2953	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 MSB2-HA rim101::KLURA3	Chavel et al. (2010)
PC2954	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 MSB2-HA snf2::KLURA3	Chavel et al. (2010)
PC2955	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 MSB2-HA msn1::KLURA3	Chavel et al. (2010)
PC2956	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 MSB2-HA yak1::KLURA3	Chavel et al. (2010)
PC2957	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 MSB2-HA mss11::KLURA3	Chavel et al. (2010)
PC2980	MATa ura3-52 elp2::KLURA3	This study
PC3016	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 bem4::HYG	A. Pitoniak, C. Chavel, J. Chow, j. Smith, D. Camara, S. Karunanithi, K. Wolfe, K., and P. J. Cullen, (unpublished data)
PC3030	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 MSB2-HA sin3::NAT	Chavel et al. (2010)
PC3031	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 MSB2-HA isw1::NAT	Chavel et al. (2010)
PC3032	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 MSB2-HA cka1::NAT	Chavel et al. (2010)
PC3033	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 MSB2-HA nhp10::NAT	Chavel et al. (2010)
PC3034	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 MSB2-HA isw2::NAT	Chavel et al. (2010)
PC3035	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 MSB2-HA mks1::NAT	Chavel et al. (2010)
PC3037	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 MSB2-HA sds3::KLURA3	Chavel et al. (2010)
PC3038	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 MSB2-HA rpd3::KLURA3	Chavel et al. (2010)
PC3039	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 MSB2-HA dig1::KLURA3	Chavel et al. (2010)
PC3352	MATa ura3-52 ras2::NAT	This study
PC3353	MATa ura3-52 sin3::NAT	This study
PC3362	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 MSB2-HA fkh1::KLURA3	Chavel et al. (2010)
PC3363	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 MSB2-HA nrg1::KLURA3	Chavel et al. (2010)
PC3414	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 FLO11-HA spo14::KLURA3	Karunanithi et al. (2010)
PC3415	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 FLO11-HA dfg16::KLURA3	Karunanithi et al. (2010)
PC3419	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 FLO11-HA ash1::KLURA3	Karunanithi et al. (2010)
PC3421	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 FLO11-HA plb3::KLURA3	Karunanithi et al. (2010)
PC3428	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 MSB2-HA swi4::KLURA3	Chavel et al. (2010)
PC3429	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 MSB2-HA mga1::KLURA3	Chavel et al. (2010)
PC3430	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 MSB2-HA fkh2::NAT	Chavel et al. (2010)
PC3431	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 MSB2-HA sfl1::KLURA3	Chavel et al. (2010)
PC3432	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 MSB2-HA fkh1::KLURA3 fkh2::NAT	Chavel et al. (2010)
PC3435	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 MSB2-HA nrg1::KLURA3 nrg2::NAT	Chavel et al. (2010)
PC3635	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 bud3::KLURA3	This study
PC3637	MATα ura3-52 leu2 ste12::kanMX6	This study
PC3642	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 MSB2-HA rtg3::NAT	Chavel et al. (2010)
PC3643	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 MSB2-HA tco89::NAT	Chavel et al. (2010)
PC3644	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 MSB2-HA gzf3::NAT	Chavel et al. (2010)
PC3652	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 MSB2-HA rtg2::NAT	Chavel et al. (2010)
PC3654	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 MSB2-HA tor1::NAT	Chavel et al. (2010)
PC3657	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 MSB2-HA lsc2::NAT	This study
PC3687	MATα ura3-52 leu2 opi1::NAT	This study
PC3688	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 MSB2-HA opi1::NAT	Chavel et al. (2010)
PC3690	MATα ura3-52 leu2 rim101::NAT	This study
PC3691	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 MSB2-HA rim101::NAT	This study
PC3695	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 MSB2-HA rtg1::NAT	This study
PC3861	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 ste11::NAT	Karunanithi and Cullen (2012)
PC3920	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 MSB2-HA vps8::NAT	This study
PC4006	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 mdh1::KLURA3	This study
PC4007	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 spt3::KLURA3	This study
PC4008	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 spt8::KLURA3	This study
PC4032	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 rim20::KLURA3	This study
PC4035	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 bmh2::KLURA3	This study
PC4038	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 cmk2::KLURA3	This study
PC4039	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 cmk1::HYG	This study
PC4043	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 bmh1::KLURA3	This study
PC4141	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 tpk2::KLURA3	This study
PC4256	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 bud1::NAT	This study
PC4468	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 rga1::KLURA3	This study
PC5071	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 pcl9::NAT	This study
PC5072	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 pcl1::KLURA3 plc9::NAT	This study
PC5073	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 pcl2::HYG plc9::NAT	This study
PC5074	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 pcl1::KLURA3 pcl2::HYG pcl9::NAT	This study
PC5075	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 pcl1::KLURA3 plc2::HYG	This study
PC5084	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 MSB2-HA tpk3::NAT	This study
PC5085	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 gln3::KLURA3	This study
PC5090	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 nte1::NAT	This study
PC5091	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 pho85::NAT	This study
PC5095	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 MSB2-HA sir2::NAT	This study
PC5102	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 MSB2-HA gcn5::KLURA3	This study
PC5108	MATa ura3-52 tpk2::NAT	This study
PC5111	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 stb3::NAT	This study
PC5113	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 stb6::NAT	This study
PC5115	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 pho4::NAT	This study
PC5121	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 pho80::NAT	This study
PC5332	MATa ura3-52 rtg2::NAT	This study
PC5335	MATa ura3-52 pho85::KLURA3	This study
PC5340	MATα ura3-52 leu2 gcn5::LEU2	This study
PC5351	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 MSB2-HA ygr125w::NAT	This study
PC5352	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 MSB2-HA gpb2::KLURA3	This study
PC5354	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 MSB2-HA cup2::KLURA3	This study
PC5360	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 MSB2-HA aca1::KLURA3	This study
PC5362	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 MSB2-HA sko1::KLURA3	This study
PC5364	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 MSB2-HA plb1::TRP	This study
PC5651	MATa ura3-52 ste12::NAT	This study
PC5822	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 vps27::KLURA3	This study
PC5826	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 vps26::KLURA3	This study
PC5831	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 vps35::KKLURA3	This study
PC5856	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 hap4::KLURA3	This study
PC5860	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 mep2::KLURA3	This study
PC5862	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 rim15::KLURA3	This study
PC5865	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 sch9::KLURA3	This study
PC5871	MATa ura3-52 leu2::HYG ssk1::NAT	This study
PC5872	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 crz1::KLURA3	This study
PC5875	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 pho81::KLURA3	This study
PC5876	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 pho84::KLURA3	This study
PC5878	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 rim21::KLURA3	This study
PC5880	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 rim9::KLURA3	This study
PC5881	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 pho5:KLURA3	This study
PC6016c	MATa can1Δ::Ste2pr-spHIS5 lyp1Δ::Ste3pr-LEU2 his3::hisG leu2Δ0 ura3Δ0	Ryan et al. (2012)
PC6093	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 gpb1::NAT gpb2::KLURA3	This study
PC6103	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 MSB2-HA tec1::NAT	This study
PC6135	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 rsr1::HIS3 bni1::KLURA3	This study
PC6136	MATa ura3-52 pde2::NAT	This study
PC6137	MATa ura3-52 ccr4::NAT	This study
PC6138	MATa ura3-52 nte1::NAT	This study
PC6139	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 ino1::KLURA3	This study
PC6140	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 smp1::KLURA3	This study
PC6141	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 flo1::KLURA3	This study
PC6159	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 ccr4::NAT ssk1::KLURA3	This study
PC6161	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 nte1::NAT ssk1::KLURA3	This study
PC6163	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 pde2::NAT ssk1::KLURA3	This study
PC6165	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 pho85::NAT ssk1::KLURA3	This study
PC6166	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 ras2::NAT ssk1::KLURA3	This study
PC6192	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 faa4::KLURA3	This study
PC6193	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 MSB2-HA hac1::KLURA3	This study
PC6197	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 MSB2-HA stb5::NAT	This study
PC6198	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 MSB2-HA stp1::KLURA3	This study
PC6201	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 MSB2-HA skn7::KLURA3	This study
PC6202	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 MSB2-HA yhp1::KLURA3	This study
PC6204	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 MSB2-HA mot3::NAT	This study
PC6206	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 MSB2-HA stb2::NAT	This study
PC6208	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 MSB2-HA ace2::KLURA3	This study
PC6210	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 MSB2-HA azf1::NAT	This study
PC6212	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 MSB2-HA phd1::NAT	This study
PC6218	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 MSB2-HA sok2::KLURA3	This study
PC6222	MATa ura3-52 ras2::NAT	This study
PC6253	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 bcy1::KLURA3	This study
PC6258	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 MSB2-HA mot2::KLURA3	This study
PC6284	MATa can1Δ::Ste2pr-spHIS5 lyp1Δ::Ste3pr-LEU2 his3::hisG leu2Δ0 ura3Δ0 mcy1::kanMX4 ste12::LEU2	This study
PC6285	MATa can1Δ::Ste2pr-spHIS5 lyp1Δ::Ste3pr-LEU2 his3::hisG leu2Δ0 ura3Δ0 pnc1::kanMX4 ste12::LEU2	This study
PC6286	MATa can1Δ::Ste2pr-spHIS5 lyp1Δ::Ste3pr-LEU2 his3::hisG leu2Δ0 ura3Δ0 mnl1::kanMX4 ste12::LEU2	This study
PC6287	MATa can1Δ::Ste2pr-spHIS5 lyp1Δ::Ste3pr-LEU2 his3::hisG leu2Δ0 ura3Δ0 rxt3::kanMX4 ste12::LEU2	This study
PC6288	MATa can1Δ::Ste2pr-spHIS5 lyp1Δ::Ste3pr-LEU2 his3::hisG leu2Δ0 ura3Δ0 cwc27::kanMX4 ste12::LEU2	This study
PC6289	MATa can1Δ::Ste2pr-spHIS5 lyp1Δ::Ste3pr-LEU2 his3::hisG leu2Δ0 ura3Δ0 ssn8::kanMX4 ste12::LEU2	This study
PC6290	MATa can1Δ::Ste2pr-spHIS5 lyp1Δ::Ste3pr-LEU2 his3::hisG leu2Δ0 ura3Δ0 nut1::kanMX4 ste12::LEU2	This study
PC6291	MATa can1Δ::Ste2pr-spHIS5 lyp1Δ::Ste3pr-LEU2 his3::hisG leu2Δ0 ura3Δ0 ela1::kanMX4 ste12::LEU2	This study
PC6292	MATα ura3-52 leu2 flo8::NAT ssk1::KLURA3	This study
PC6293	MATα ura3-52 leu2 gcn5::LEU2 ssk1::KLURA3	This study
PC6294	MATα ura3-52 leu2 opi1::NAT ssk1::KLURA3	This study
PC6295	MATα ura3-52 leu2 rim101::NAT ssk1::KLURA3	This study
PC6296	MATa ura3-52 sin3::NAT ssk1::KLURA3	This study
PC6297	MATa ste4 FUS1-lacZ FUS1-HIS3 ura3-52 leu2::HYG lacZ::NAT tec1::LEU2	This study
PC6298	MATα ura3-52 leu2 elp2::LEU2	This study
PC6299	MATa ura3-52 leu2::HYG ssk1::NAT elp2::LEU2	This study

^aAll strains are in the Σ1278b background unless otherwise indicated.

^bKLURA3 refers to the Kluyveromyces lactis URA3 cassette.

^cΣ1278b ordered deletion collection control strain MATa can1Δ::Ste2pr-spHIS5 lyp1Δ::Ste3pr-LEU2 his3::hisG leu2Δ0 ura3Δ.

Halo assays were performed as described (Jenness et al. 1987). Specifically, wild-type and mutant strains were grown to saturation in YEPD (2% glucose). Cell density was determined by OD A₆₀₀. Cultures were serially diluted such that ∼10,000 cells were spread onto YEPD plates. After the cell suspension had dried, four spots of 1 µg/µl alpha factor were spotted (10 µl). Plates were incubated at 30° for 2 days and photographed.

Invasive growth screen of the filamentous deletion collection

The MATa Σ1278b deletion collection (Ryan et al. 2012) was pinned in 96-well format to YEPD media omnitrays (Thermo Scientific, Waltham, MA). Each plate was pinned independently using a pinning tool (V&P Scientific, VP408, San Diego, CA)—sterilized with a 10% bleach solution, 95% ethanol, 70% ethanol and flame—and a pinning guide tray (V&P Scientific, VP381). Plates were pinned in duplicate and incubated for 5 and 12 days. Plates were photographed, washed in a stream of water, and photographed again. Each plate was scored visually for colonies that showed changes in morphology and invasive growth. Scores were tabulated to produce a single score called the invasive growth index. The results of the screen and details of the scoring system are presented in Table S2.

Evaluating filamentous growth MAPK pathway activity

The activity of the filamentous growth MAPK pathway was evaluated with a transcriptional reporter [pFRE–lacZ (Madhani and Fink 1997)]. Strains that showed clear-cut invasive growth phenotypes (hyper- and hypoinvasive) were transformed with the pFRE–lacZ plasmid. Strains were grown in media lacking uracil to maintain selection for the plasmid and the nonpreferred carbon source galactose (S-GAL–URA) to induce pathway activity (Pitoniak et al. 2009). Mutants were induced in S-GAL–URA for 4 hr. Mutants were grown in batches of ∼20 alongside control strains (tec1Δ and dig1Δ) to minimize batch-to-batch variation. Cell extracts were prepared, and β-galactosidase assays were performed as described (Chavel et al. 2010). The average values of at least two independent experiments were reported. Statistical significance was determined by comparing the difference between wild-type and experimental pFRE–lacZ expression averages in a z-test score (Freedman et al. 1998). The z-test score was converted to the P-value (http://www.graphpad.com). Samples with a P-value ≤0.0001 and ≥1.5-fold change from wild type were considered statistically significant. Raw data for the β-galactosidase assays can be found in Table S3. For some experiments, the filamentous growth MAPK pathway was evaluated with a growth reporter [FUS1–HIS3, (McCaffrey et al. 1987)]. In Σ1278b cells lacking an intact mating pathway (ste4Δ), growth on SD–HIS is dependent on the filamentous growth MAPK pathway (O’Rourke and Herskowitz 1998; Cullen et al. 2004; Pitoniak et al. 2009; Chavel et al. 2010; Karunanithi and Cullen 2012). Growth assays are shown in Figure S1. As a separate test, 26 genes identified by the screen were disrupted in a wild-type Σ1278b strain and checked for invasive growth and pFRE–lacZ; 77% passed a preliminary test.

Budding pattern analysis

Patterns of bud-site selection were based on established principles (Chant and Pringle 1995). Budding pattern was determined in two ways. In one method, budding pattern was based on visual inspection of connected cells. Σ1278b cells grown in liquid YEPlowD (0.2% glucose) media undergo filamentous growth and exhibit Ste12p-dependent changes in cell length, cell–cell adhesion, and distal-unipolar budding. Cells were grown to midlog phase in YEPlowD (0.2% glucose) liquid medium for 12–14 hr and examined by microscopy at 100× magnification. Buds were assigned as proximal, equatorial, or distal depending on their position relative to mother cells. At least 150 cells were counted for each experiment.

In a separate approach, cells were stained by FITC-ConA and TRITC-ConA based on published protocols (Matheos et al. 2004; Gao and Bretscher 2009) with the following modifications. Cells were grown in YEPlowD for 16 hr. FITC-ConA (0.1 mg/ml) was added to 1 ml cells. Cells were incubated in the dark for 15 min, washed three times, and resuspended in YEPlowD for 4 hr. Cells (1 ml) were then stained with TRITC-ConA (0.1 mg/ml), washed three times in water, and examined by fluorescence microscopy to visualize the position of buds. At least 200 buds were recorded for each condition.

Quantitative PCR analysis

Quantitative PCR (qPCR) analysis was performed as described (Pfaffl 2001). To prepare total RNA, cells were grown in 50-ml aliquots in YEP–GAL medium to midlog phase (∼6 hr). Total RNA was isolated by hot acid phenol extraction. cDNA synthesis and real-time PCR reactions were performed as described (Chavel et al. 2010). qPCR and melt curve data collection was performed as described (Chavel et al. 2010) with the following alterations to the amplification cycles: initial denaturation for 3 min at 95°, followed by 35× cycle 3 (denaturation for 30 sec at 95°, annealing for 30 sec at 60°, and extension for 30 sec at 72°). Gene expression was quantified using the ΔΔC_t method as described (Livak and Schmittgen 2001). All reactions were performed in triplicate, and average values are reported. Primers used are as follows: ACT1 (forward 5′-GGCTTCTTTGACTACCTTCCAACA-3′ and reverse 5′-GATGGACCACTTTCGTCGTATTC-3′), NRG1 (forward 5′-CTAATGATGCATATAATAAGATGGC-3′ and reverse 5′-ATGACCCGATGTAGTGAATCCT-3′), PHO5 (forward 5′-ACATCACCTTGCAGACTGTCA-3′ and reverse 5′-AAGTACTAGCGTCAGTTGAGG-3′), INO1 (forward 5′-CTAATCAAGATGAGAGAGCCAAT-3′ and reverse 5′-ATACTTCTACGTACCTCTCAGTA-3′), and SMP1 (forward 5′-AGTCAAGATTCCTCCAGTGTAC-3′ and reverse 5′-ATCCGCTCGTGATATTGCTC-3′).

Fluorescence microscopy

Actin staining by rhodamine phalloidin has been described (Amberg 2000). Differential interference contrast (DIC) and fluorescence microscopy using rhodamine and GFP filter sets were performed using an Axioplan 2 fluorescent microscope (Zeiss, Jena, Germany) with a PLAN-APOCHROMAT 100X/1.4 (oil) objective (N.A. 0.17). Digital images were obtained with the Axiocam MRm camera (Zeiss). Axiovision 4.4 software (Zeiss) was used for image acquisition and analysis.

Results

Identification of filamentous growth MAPK pathway regulators

An ordered collection of 4072 deletion mutants constructed in the filamentous (MATa Σ1278b) background (provided by the Boone Lab, Toronto, ON; Ryan et al. 2012) was screened for changes in colony morphology (Granek and Magwene 2010; Voordeckers et al. 2012) and invasive growth based on the plate-washing assay (Roberts and Fink 1994) to identify regulators of filamentous growth. These assays provide a readout of filamentous growth that correlate with the activity of the filamentous growth MAPK pathway (Figure 1B; Roberts and Fink 1994).

Screens were performed at two time periods (5 and 12 days), which allowed evaluation of the progression of invasive growth. The 5-day screen was better suited to identify hyperfilamentous growth mutants, and the 12-day screen enriched for hypofilamentous growth mutants. The two screens also provided independent validation of the mutants identified (Table S2). A scoring system incorporated colony morphology and agar invasion from both screens into a single value called the invasive growth index that was used to rank the mutants by strength-of-phenotype (Table S2).

Of the 4072 mutants represented in the collection, 220 showed hyperfilamentous growth and 478 showed hypofilamentous growth (Table S2). Many of these mutants have been identified in other screens (Table S1) (Lambrechts et al. 1996; Lo and Dranginis 1998; Pan and Heitman 1999; Lorenz et al. 2000; Ryan et al. 2012; Shively et al. 2013). The screen uniquely identified new regulators of filamentous growth (Table S1), which may have been due to the specific incubation times or differences in scoring systems.

To identify those regulators of filamentous growth that also regulate the filamentous growth MAPK pathway, mutants identified in the screen were examined for changes in the activity of a transcriptional reporter, pFRE–lacZ, which provides a readout of filamentous growth MAPK pathway activity (Madhani and Fink 1997; Pitoniak et al. 2009). Control strains verified that loss of negative regulators showed elevated pFRE–lacZ activity (Figure 1, A and C, asterisks, and Table S3), which included the transcriptional repressor Dig1p (Cook et al. 1996), the mating pathway MAP kinase Fus3p (Bruckner et al. 2004), and the HOG pathway MAP kinase kinase Pbs2p (Figure 1, A and C and Table S3; asterisks in panel C refers to pathway components). Loss of pathway components showed reduced pFRE–lacZ activity (Figure 1, A and D, asterisks; msb2Δ, sho1Δ, ste50Δ, tec1Δ, ste20Δ, ste11Δ, kss1Δ, and ste12Δ; Table S3).

Mutants identified by the invasive growth screens were transformed with pFRE–lacZ reporter and evaluated for β-galactosidase activity. For the hyperinvasive growth mutants, 41 of 110 showed elevated pFRE–lacZ expression (37%, Figure 1C). For the hypoinvasive growth mutants, 43 of 116 tested showed a defect in pFRE–lacZ expression (37%, Figure 1D). Not all candidates were examined, because mutants with weaker phenotypes showed differences that fell below the statistical cutoff employed (≥1.5-fold, P-value ≤ 0.0001). Thus, the screen was not saturating.

To validate the results of the screen, and/or extend connections of known pathways to the filamentous growth MAPK pathway, ∼100 genes were disrupted in a wild-type Σ1278b strain, and gene disruptants were evaluated for invasive growth and MAPK activity (Table S1 and Figure S1). The analysis was facilitated by a cross-talk reporter that in a mating-deficient strain (ste4Δ FUS1–HIS3) provides a readout of the filamentous growth MAPK pathway. The analysis eliminated ∼15% of the candidates as false positives (Table S1). The analysis also identified several new components. In total, 97 proteins were identified by the screen and gene disruption analysis that regulate the filamentous growth MAPK pathway and play a corresponding role in the regulation of filamentous growth.

Evaluating filamentous growth MAPK pathway regulators by the change in budding pattern and cell elongation

Candidate regulators were examined for morphological phenotypes that are controlled by the filamentous growth MAPK pathway. The filamentous growth MAPK pathway regulates changes in budding pattern (Gimeno et al. 1992; Roberts and Fink 1994; Cullen and Sprague 2000, 2002) and the cell cycle that results in an increase in cell length (Kron et al. 1994; Madhani et al. 1999). The change in budding pattern is visually striking in haploid cells that switch from axial to distal-unipolar budding (Cullen and Sprague 2002). A recent study showed an abundance of filamentous cells in MAPK pathway mutants, raising the question of whether and to what extent the MAPK pathway regulates this aspect of the response (Chen and Thorner 2010). To clarify this issue, the budding pattern of filamentous cells was examined by two approaches. In one approach, filamentous growth was examined in liquid culture by microscopy (Figure 2A, left columns). In a second approach, cells were stained with FITC-ConA and TRITC-ConA at different times to visualize bud position (Figure 2A, right columns). The latter approach had the advantage of determining bud position without the assumption of which cell the parent was (Figure 2B). The approaches were in close agreement for wild-type cells and control strains lacking axial [bud3Δ (Chant et al. 1991)], distal [bud8Δ (Harkins et al. 2001; Schenkman et al. 2002; Kang et al. 2004), and core [rsr1Δ (Park et al. 1997, 2002; Kang et al. 2010)] bud-site-selection markers (Figure 2A).

Figure 2.

Role of filamentous growth MAPK pathway regulators in bud-site selection and filament formation. (A) Bar graph, the percentage of cells exhibiting distal (blue), equatorial (red), or proximal (black) budding pattern. Left column: assignment based on visual inspection. Right column: assignment based on fluorescence microscopy. (B) Example of cells costained with FITC/TRITC-ConA. Top: wild-type cells. Bottom left: the ste20Δ mutant. Bar, 5 μm. Bottom right: cartoon representing budding cells. D, distal; P, proximal. (C) Single-cell assay shows hyperpolarized growth and distal-pole budding pattern of the fus3Δ and dig1Δ mutants in medium containing 2% glucose (HIGH GLU). Bar, 20 μm. (D) Single-cell assay of select hyperinvasive growth mutants. Cells were grown on S-GLU (-GLU) medium at low density for 16 hr and photographed at 100×. Bar, 20 μm. Representative microcolonies are shown; other examples are in Figure S2. (E) Single-cell assay of select hypoinvasive growth mutants. Bar, 20 μm. Other mutants are shown in Figure S2. (F) Budding pattern analysis of hypofilamentous growth mutants indicated. Scoring system is the same as in A.

Wild-type haploid cells showed a characteristic change in budding pattern from axial to distal-unipolar budding when grown in glucose-limited medium (Figure 2A) (Chant and Pringle 1995; Cullen and Sprague 2002)]. The ste12Δ mutant showed a 15% reduction in distal-unipolar budding, and the ste20Δ mutant showed a 20% reduction (Figure 2A). Many cells retained distal-pole budding (60%), which can account for the conclusion that the filamentous growth MAPK pathway is dispensable (Chen and Thorner 2010). Therefore, the filamentous growth MAPK pathway regulates the change in polarity during filamentous growth. Other signaling pathways probably also regulate the change in budding pattern. Under this condition, the ras2Δ mutant did not play a role (Figure 2A). The rsr1Δ mutant did not show a completely random budding pattern, but retained the propensity to bud at the distal pole. This may be due to increased polarized growth of filamentous cells, which bias bud-site-selection to the distal pole (Sheu et al. 2000; Cullen and Sprague 2002). Disruption of the gene encoding the formin Bni1p, which reduces the polarized growth of filamentous cells (Cullen and Sprague 2002), conferred random budding to the rsr1Δ mutant (Figure 2B, bni1Δ rsr1Δ).

The single-cell invasive-growth assay provides a convenient measure of the changes in budding pattern and cell length that occur during filamentous growth (Cullen and Sprague 2000) and was used to examine mutants identified in the screen. Most hyperinvasive growth mutants showed hyperelongated morphology by the single-cell assay. Glucose suppresses the filamentous morphology (Cullen and Sprague 2000) and effectively suppressed hyperelongated morphology and distal-pole budding pattern in all but two hyperinvasive growth mutants, dig1Δ and fus3Δ (Figure 2C). These mutants regulate the filamentous growth MAPK pathway; thus, a hyperactive filamentous growth MAPK pathway can bypass the inhibition of cell elongation and distal-pole budding induced by growth of cells in high-glucose environments.

Other hyperinvasive growth mutants showed hyperelongated cell morphology (Figure 2D and Figure S2). A subset of these were dependent on Ste12p for invasive growth and morphology (Figure S3, A and B). Not all mutants showed Ste12p dependence (Figure S3), which might reflect a role for these proteins in regulating filamentous growth outside the MAPK pathway. Hypoinvasive growth mutants were similarly examined. Most hypoinvasive growth mutants, like the ste12Δ mutant, showed a defect in unipolar budding and cell elongation (Figure 2E and Figure S2). Similarly, a subset of mutants showed defects in distal-pole budding (Figure 2F). Whereas no mutant was completely defective, several (like the ste12Δ mutant) showed minor differences. It is possible that distal-pole budding during filamentous growth results from the additive contribution of multiple pathways. Therefore, the budding pattern and single-cell analysis corroborated a role for many of these proteins in regulating the filamentous growth MAPK pathway.

Functional analysis of candidate genes and pathways

The genes identified by the screen and directed gene deletion analysis were classified by GO annotation terms for biological process, cellular compartment, and molecular function (Ashburner et al. 2000). Genes were also overlaid onto known protein and genetic interaction maps (Uetz et al. 2000; Drees et al. 2001; Ho et al. 2002; Miller et al. 2005; Costanzo et al. 2010). As a result, the genes were found to comprise functional categories that were explored in detail below (Figure 3).

Figure 3.

Functional classification of filamentous growth MAPK pathway regulators. Genes identified by the screen (Figure 1, C and D) or by hypothesis-based testing (Figure S1) are shown according to their established roles in pathways or protein complexes. Lines refer to functional connections. Green, positive regulator; red, negative regulator; gray, no phenotype; and white, not tested. (A) The Ras2p–cAMP–PKA pathway; (B) Rpd3p(L) chromatin remodeling complex (Table S3); (C) the ELP complex; (D) the Pho85p–Pho80p pathway; (E) the Rim101 pathway. (F) Lipid biosynthesis; (G) the RTG pathway and proteins associated with mitochondrial function; and (H) Functional classification of other proteins. (I) Plate-washing assay for pho85Δ mutant alongside controls. (J) Single-cell assay for the pho85Δ mutant alongside controls. Scraped refers to cells scraped from an invasive scar. (K) FUS1–HIS3 reporter activity for the pho85Δ mutant alongside controls.

A subset of Ras2p pathway regulators was found to regulate the filamentous growth MAPK pathway. These include Ras2p (Mosch et al. 1996), the alternative GEF Sdc25p (Damak et al. 1991; Jones et al. 1991), the phosphodiesterase Pde2p, and PKA subunit Tpk2p (Figure 3A; Chavel et al. 2010). Tpk2p regulates the transcription factor Flo8p (Robertson and Fink 1998) and negatively regulates the transcriptional repressor Sfl1p (Conlan and Tzamarias 2001; Pan and Heitman 2002). Flo8p and Sfl1p were identified by the screen (Figure 3A). Sfl1p-interacting proteins Ssn8p and Nut1p are required for Sfl1p to carry out its role as a transcriptional repressor (Conlan and Tzamarias 2001) and were identified by the screen (Figure 3A). Ccr4p, a component of the Ccr4p–NOT deadenylase complex, which is an effector of Tpk2p (Lenssen et al. 2002) and target of the MAP kinase Kss1p (Fasolo et al. 2011) was also identified.

The screen uncovered components of the chromatin remodeling complex Rpd3p(L) (Figure 3B, Ash1p, Sap30p, Ume1p, Cti6p, Pho23p, and Rxt3p; Rpd3p, Sin3p, Rxt2p, and Sds3p were previously identified; Chavel et al. 2010). The ELP complex regulates the filamentous growth MAPK pathway (Figure 3C; Abdullah and Cullen 2009; Elp2p, Elp6p, Iki3p, Kti12p). The screen identified members of the ELP complex (Figure 3C, Elp3p, Iki3p, and Elp4p). Thus, the screen was effective at identifying regulatory connections to the filamentous growth MAPK pathway and supports the idea that the Ras2p pathway, Rpd3p(L), and ELP are major regulators of the filamentous growth MAPK pathway.

The screen identified the cyclin-dependent kinase Pho85p (Figure 3, D and I–K and Figure S1) (Measday et al. 1997; Huang et al. 2002, 2007; Shemer et al. 2002; Moffat and Andrews 2004). Loss of cyclins Pcl1p, Pcl2p, and Pcl9p, and the triple pcl1Δ pcl2Δ pcl9Δ mutant, showed no defect in filamentous growth MAPK pathway activity (Figure 3D and Figure S1), whereas loss of Pho80p, the cyclin responsible for environmental responses controlled by Pho85p (Liu et al. 2000), showed reduced filamentous growth MAPK pathway activity (Figure 3D). Deletion of the transcription factor PHO2 (Kaffman et al. 1994; O’Neill et al. 1996; Liu et al. 2000) showed reduced filamentous growth MAPK pathway activity (Figure 3D).

Several proteins and pathways were not identified by the screen but were shown to regulate the filamentous growth MAPK pathway by direct testing. These may have been missed by the screen for several reasons. One is that they were not represented in the collection. This was true for components of Rpd3p(L), including tod6Δ and ume6Δ. A second reason is that the pathways may play a conditional role in regulating the filamentous growth MAPK pathway. For example, the Ras2p pathway does not constitutively regulate the filamentous growth MAPK pathway (Figure S4), which complicated assessment of the roles of Flo8p and Tpk2p. A third reason is that phenotypes may have fallen below threshold of statistical significance applied to the data (e.g., tpk2Δ, rim101Δ, kss1Δ, opy2Δ, rim9Δ, dfg16Δ, rim13Δ, rim8Δ, ume1Δ; Table S3).

Components of the Rim101p pathway were found to regulate the filamentous growth MAPK pathway (Figure 3E, Chavel et al. 2010). Nrg1p and Nrg2p (Lamb and Mitchell 2003) did not regulate the filamentous growth MAPK pathway (Figure 3E). The lipid regulatory transcription factor Opi1p (Greenberg et al. 1982) regulates the filamentous growth MAPK pathway (Chavel et al. 2010). The serine esterase Nte1p, which serves as the phospholipase B of yeast (Fernandez-Murray et al. 2009) was also found to regulate the filamentous growth MAPK pathway (Figure 3F). Ncr1p, which regulates sphingolipid biosynthesis (Malathi et al. 2004) and has not been shown to interact with Opi1p or Nte1p genetically or physically, negatively regulated the filamentous growth MAPK pathway.

The RTG pathway regulates the filamentous growth MAPK pathway (Figure 3G, Chavel et al. 2010). The screen did not identify this pathway but did uncover proteins that influence mitochondrial function (Figure 3G). A main target of RTG is aconitase (Liu and Butow 1999), an enzymatic component of the TCA cycle necessary to generate α-ketoglutarate, a precursor in glutamate biosynthesis (Magasanik and Kaiser 2002). Aco1p positively regulated the filamentous growth MAPK pathway (Figure 3G). Rtg2p is also incorporated into the histone-acetyl transferase (HAT) complex SLIK, with the HAT Gcn5p as its catalytic component (Pray-Grant et al. 2002). Deletion of GCN5 reduced filamentous growth MAPK pathway activity (Figure 3G). Npr1p, a kinase that stabilizes amino acid transporters at the membrane (De Craene et al. 2001) and is negatively regulated by the TORC1 complex (Schmidt et al. 1998), was found to positively regulate the filamentous growth MAPK pathway.

The above analysis accounted for nearly half the proteins identified by the screen and involved the Ras2p, Rpd3p(L), ELP, Opi1p, Rim101p, RTG, and Pho85p pathways. The remaining proteins represent new connections to the filamentous growth MAPK pathway. These included proteins that regulate transcription [including components of the THO complex and chromatin remodeling proteins that are separate from Rpd3p(L)], protein transport and trafficking (including components of the signal recognition complex, SRP), protein translation, prefoldin, metabolism, sporulation, the cytoskeleton, post-translational modification, and genes whose functions remain to be characterized (Figure 3H).

Although it is not clear how these pathways regulate the filamentous growth MAPK pathway, the majority of regulators tested did not influence the activity of the mating pathway (Figure S5) or the HOG pathway (Figure S6A), although a HOG pathway reporter p8XCRE–lacZ was modestly induced in some hyperinvasive growth mutants (Figure S6B). These pathways share components with the filamentous growth MAPK pathway (Chen and Thorner 2007; Saito 2010; Saito and Posas 2012). Thus, it would appear that these factors by and large play a specific role in regulating the filamentous growth MAPK pathway. We previously showed that many pathways converge on the expression of the MSB2 promoter (Chavel et al. 2010). Perhaps these regulators regulate the filamentous growth MAPK pathway in a similar manner.

Unregulated filamentous growth MAPK pathway activity is detrimental to invasive growth and proper morphogenesis

It is not entirely clear why the regulation of the filamentous growth MAPK pathway is so extensive. One possibility is that the activity of the filamentous growth MAPK pathway may be adjusted to that of other pathways that regulate filamentous growth. Coordination of morphogenetic pathways that regulate cell-cycle progression and cell polarity might be critical, for example, for proper growth. To test this possibility, the activity of the filamentous growth MAPK pathway was genetically separated from other regulatory pathways using gain-of-function alleles and by driving expression of pathway regulators with inducible promoters. Hyperactive alleles of MSB2 (Cullen et al. 2004), SHO1 (Vadaie et al. 2008), and STE11 (Stevenson et al. 1992) were examined. In addition, overexpression of pathway components was assessed with the strong inducible promoter (pGAL; Longtine et al. 1998). Preliminary observations with these strains showed that hyperactivation or overexpression of Sho1p or Msb2p did not induce hyperinvasive growth but rather caused a reduction in invasive growth (Figure 4A). Microscopic examination showed that the cells had morphological defects (Figure 4, B–D).

Figure 4.

Overexpression/hyperactivation of pathway components does not enhance filamentous growth. (A) Plate-washing assay was performed for the indicated strains on YEP–GAL. GAL–MSB2 and GAL–SHO1 strains where the GAL1 promoter has replaced the native promoter. MSB2^Δ100–818 is a hyperactive allele of MSB2. (B) Septin staining. Strains containing the pCdc12p–GFP plasmid were grown on SCD or S medium lacking glucose. Bar, 5 μm. (C) Indicated strains were stained with rhodamine phalloidin. Bar, 5 μm. (D) Colocalization of Sho1p and Cdc24p in cells carrying a hyperactive allele of MSB2. Bar, 5 μm.

To further explore this question, cell-polarity markers for the mother-bud neck (with the septin Cdc12p–GFP) and the cytoskeleton (with rhodamine phalloidin, which stains actin) were examined in these mutants. Septin staining showed defects in cytokinesis (Figure 4B). Actin staining showed irregular patterns, with polymerized actin at multiple surface sites on the plasma membrane (Figure 4C, File S1, and File S2). Moreover, the localization of polarity control proteins, Sho1p and Cdc24p, were localized to aberrant structures in these mutants (Figure 4D). The prevalence of these phenotypes varied among the mutants tested and ranged from ∼10% irregular morphologies for the STE11-4 mutant to >90% irregular cells in the GAL–MSB2 GAL–SHO1 mutant. Prolonged overexpression of SHO1 resulted in growth defects (not shown). Equivalent morphologies were observed in other mutants that exhibited filamentous growth MAPK pathway hyperactivation (not shown). However, these phenotypes stood out from most of the hyperfilamentous growth mutants identified in the screen, possibly because pGAL-driven and hyperactive proteins have higher pathway activity. Therefore, hyperactivation of the filamentous growth MAPK pathway causes problems with normal cell morphogenesis. A likely explanation for this phenotype is that it results from that pathway’s critical roles in cell-polarity and cell-cycle control. Therefore, coordination of the activity of the filamentous growth MAPK pathway may be necessary not only to promote a coherent filamentation response but also to maintain proper cell growth.

Pathways that regulate filamentous growth control each other’s targets

The filamentous growth MAPK pathway may be the terminal pathway at which many pathways converge. An alternative possibility that has not been explored is that many of the pathways that regulate filamentous growth may also regulate each other’s activities. To test this possibility, transcriptional targets of several of the major pathways that regulate filamentous growth were evaluated in a panel of pathway mutants by qPCR analysis and/or transcriptional reporters. NRG1 and SMP1 were selected as targets of the Rim101p pathway, which are downregulated by that pathway and upregulated in that mutant (Lamb and Mitchell 2003); PHO5, which is downregulated by the Pho85p pathway and upregulated in the mutant (Kaffman et al. 1994); CIT2, which is a target of the RTG pathway (Rothermel et al. 1995); and INO1, which is downregulated by the lipid/Opi1p pathway (White et al. 1991). qPCR and/or lacZ analysis was performed in mutants lacking the major filamentation control pathways (ras2Δ, nte1Δ, sin3Δ, elp2Δ, ste12Δ, rim101Δ, pho85Δ, rtg2Δ, flo8Δ, and opi1Δ). With the exception of Opi1p/INO1, each pathway regulated the expression of its own target (Figure 5, A–D). Many pathways similarly influenced the expression of each other’ targets. For example, INO1’s expression was upregulated in the nte1Δ, sin3Δ, elp2Δ, rim101Δ, rtg2Δ, and flo8Δ mutants (Figure 5A). Thus, in some manner, the Nte1p, Sin3p, Epl2p, Rim101p, Rtg2p, and Flo8p proteins contribute to the coregulation of a lipid pathway target. Similar results were found for PHO5 (Figure 5B), SMP1 (Figure 5C), and NRG1 (Figure 5D). Several pathways were also found to regulate the RTG pathway, including Ras2p and ELP, based on CIT2–lacZ reporter (Figure S7). DNA microarray analysis previously identified a major transcriptional target of the filamentous growth MAPK pathway as the gene encoding Rim8p, a component of the Rim101p pathway (Chavel et al. 2010), and we confirmed that the filamentous growth MAPK pathway contributed to RIM8 expression by qPCR analysis (data not shown). These results indicate that a subset of the major pathways that regulate filamentous growth regulate at least one of each other’s key targets.

Figure 5.

Signal integration between pathways during filamentous growth. (A–D) Quantitative PCR analysis of labeled genes in the indicated mutants. The bar indicates change in expression using the ΔΔC_t quantitation method. Wild-type expression levels were set to 1. The error bars represent standard deviation between experiments. Cells were grown to midlog phase in YEP–GAL media for 6 hr. (E) Diagram showing connections between pathways that regulate filamentous growth. Arrows refer to positive regulation. Bars, negative regulation.

The mechanisms by which such regulation occurs is not clear and may occur through diverse means such as pathway-to-pathway connections or the modulation of transcription factors that serve as master regulators of signaling outputs. Moreover, not all of the possible regulatory connections were observed. For example, the filamentous growth MAPK pathway does not appear to regulate Ras2p–cAMP–PKA (Chavel et al. 2010) or the RTG pathways (Figure S7). Nevertheless, the results are striking from the perspective that each arrow represents in principle a regulatory connection that occurs between two pathways (Figure 5E).

Discussion

Cell differentiation involves the combined action of many different proteins and pathways. How multiple signals become integrated into a cohesive response is an important biological problem that in many cases remains unclear. Here, we explore the question of signal integration by identifying, from a global perspective, regulators of the MAPK pathway responsible for controlling differentiation to the filamentous cell type. Using a genetic screen and direct testing, we identify >95 proteins that when absent influence the activity of the filamentous growth MAPK pathway. This number likely represents an underestimate because the screen was not saturating and because a rigorous statistical cutoff was used to establish regulators. In addition, the screen was performed under a single condition where some pathways and complexes may not be required. A conservative estimate is that >35% of the major regulators of filamentous growth regulate the activity of the filamentous growth MAPK pathway.

One consequence of these regulatory connections is to sensitize MAPK activity to different stimuli. Many of the major nutrient-regulatory pathways in yeast, such as TOR (Bruckner et al. 2011), Snf1p (Karunanithi and Cullen 2012), Ras2p (this study; Mosch et al. 1996; Chavel et al. 2010), and RTG (this study; Chavel et al. 2010) impinge on the activity of the filamentous growth MAPK pathway. We also show that pathways that sense and respond to diverse stimuli, such as pH (Rim101p) and other environmental stimuli (Pho85p) also regulate the filamentous growth MAPK pathway. The connection between Pho85p and the filamentous growth MAPK pathway is particularly relevant as Pho85p has been shown in C. albicans to be required for temperature-dependent filamentation (Shapiro et al. 2012).

A second reason that the filamentous growth MAPK pathway is extensively regulated might be to coordinate its activity with other pathways that regulate the same response. In this way, the MAPK-dependent changes to budding pattern, the cell cycle, and cell adhesion can be tuned to the global network. Hence, multiple pathways can tap into these major regulatory events (instead of each pathway making changes directly). The filamentous growth MAPK pathway regulates the change in budding that occurs during filamentous growth. Mutants that reduce pathway activity (ste20 and ste12) show a decrease in distal-pole budding, and mutants with elevated pathway activity (dig1 and fus3) show an increase in distal-pole budding, even under high-glucose conditions. Given that multiple pathways regulate distal-pole budding, it is likely that the role of the filamentous growth MAPK pathway may be as significant as the contributions of other pathways. We show that unregulated filamentous growth MAPK pathway activity is detrimental to proper morphogenesis and cell growth.

We also show that extensive cross-regulation occurs among several of the pathways that regulate filamentous growth. Our study highlights a degree of signal integration that has not been previously appreciated. The qPCR performed here (Figure 5) was under conditions in which the filamentous growth MAPK pathway is activated (e.g., poor carbon sources, like galactose). The network described in Figure 5A was examined under a single growth condition. It is possible that other connections between pathways were missed if they occur under a condition that was not tested. Genetic buffering between the pathways may obscure connections. Similarly, loss of one pathway may or may not induce loss of the filamentous growth phenotype. The ways by which this regulation is accomplished is not clear. Multiple pathways could feed into a central metabolite or small molecule (cAMP) that regulates a master regulatory transcription factor, from which multiple filamentation targets are controlled. Indeed, transcriptional “hub” proteins globally regulate filamentous growth (Borneman et al. 2006). At least one connection may be direct. RIM8 is a major target of the filamentous growth MAPK pathway (Chavel et al. 2010). As not all pathways regulate each other’s targets, the connections between the pathways are presumably specific.

In conclusion, filamentous growth results from a highly coordinated and integrated signaling network. A single MAPK pathway regulates filamentous growth, which is controlled by many different pathways to integrate various signals and coordinate the response. Signal integration is commonly seen in similar differentiation responses in higher eukaryotes, where multiple stimuli activate an interconnected set of signaling pathways (Cuenda and Rousseau 2007; Katz et al. 2007; Raman et al. 2007). Perhaps the connections identified here extend to related pathways in other systems.