Conserved signaling modules regulate filamentous growth in fungi: a model for eukaryotic cell differentiation

Abstract

Eukaryotic organisms are composed of different cell types with defined shapes and functions. Specific cell types are produced by the process of cell differentiation, which is regulated by signal transduction pathways. Signaling pathways regulate cell differentiation by sensing cues and controlling the expression of target genes whose products generate cell types with specific attributes. In studying how cells differentiate, fungi have proved valuable models because of their ease of genetic manipulation and striking cell morphologies. Many fungal species undergo filamentous growth—a specialized growth pattern where cells produce elongated tube-like projections. Filamentous growth promotes expansion into new environments, including invasion into plant and animal hosts by fungal pathogens. The same signaling pathways that regulate filamentous growth in fungi also control cell differentiation throughout eukaryotes and include highly conserved mitogen-activated protein kinase (MAPK) pathways, which is the focus of this review. In many fungal species, mucin-type sensors regulate MAPK pathways to control filamentous growth in response to diverse stimuli. Once activated, MAPK pathways reorganize cell polarity, induce changes in cell adhesion, and promote the secretion of degradative enzymes that mediate access to new environments. However, MAPK pathway regulation is complicated because related pathways can share components with each other yet induce unique responses (i.e. signal specificity). In addition, MAPK pathways function in highly integrated networks with other regulatory pathways (i.e. signal integration). Here, we discuss signal specificity and integration in several yeast models (mainly Saccharomyces cerevisiae and Candida albicans) by focusing on the filamentation MAPK pathway. Because of the strong evolutionary ties between species, a deeper understanding of the regulation of filamentous growth in established models and increasingly diverse fungal species can reveal fundamentally new mechanisms underlying eukaryotic cell differentiation.

Introduction

Cell differentiation is a fundamental process underlying the extraordinary biological diversity in nature. Starting with genetically identical progenitor cells, cells can differentiate into a vast array of cell types with different shapes and sizes that are suited for specific functions. These specialized cells are highly organized in multicellular organisms through the process of development (Glover 2000; Basson 2012; Cook and Genever 2013; Brunet and King 2017). Cell differentiation can also occur in single-celled organisms to promote their survival in specific environments, such as in response to stress, and as cells go through their life cycle. Because cells can differentiate into different cell types, a central question is to understand how cell differentiation is regulated. Because the misregulation of cell differentiation is commonly linked to developmental problems and disease, addressing this question is important to understanding human health and organismal development.

During the process of development, cells differentiate in a predefined program. This generally occurs in response to signals (i.e. stimuli) transmitted by signal transduction pathways. Signaling pathways that regulate differentiation can be activated by the detection of extracellular stimuli through receptors or sensors at the cell surface. Once activated, these sensors control effector relay molecules and second messengers, which govern the activity of transcription factors. Transcription factors activated by signaling pathways regulate the expression of target genes whose products function to produce specialized cell types (Kramer 2016; Martinez-Soto and Ruiz-Herrera 2017). Some of these gene products generate specialized cell types by reconfiguring cell polarity or the composition of proteins at the cell surface.

One way that signaling pathways have been studied is by reductionist approaches with genetically tractable model organisms. One valuable group of models has been fungi, which continue to provide valuable insights into the fundamental processes that generate cell-type diversity. The general and distinct biology of each fungal species, including aspects of cell differentiation, come from their evolution in different niches and environments. These niches are diverse as some fungi are free-living saprotrophs (decomposers), others are pathogens (that cause infections), and yet others are symbionts (mutualist and commensal organisms). Originally nonpathogenic models were commonly studied; however, as genetic tools have improved, fungal pathogens have become important models as well. Additionally, cell differentiation in fungi can occur throughout the life cycle, such as during sexual reproduction where diploid cells undergo sporulation to produce haploid progeny (Neiman 2011; Su et al. 2012; Wendland 2020), and haploid cells can mate to form diploids (Bölker and Kahmann 1993; Bardwell 2005; Du and Yang Zhu 2021; Sieber et al. 2023). Many fungal species also undergo filamentous growth (Riquelme 2013; Kiss et al. 2019)—a specialized cell type where elongated cells expand into new environments. Because filamentous growth is morphologically striking and important for disease, it has taken center stage in the studies of fungal differentiation.

How cells make decisions about whether to undergo filamentous growth and how they accomplish this morphological switch is the focus of this review. We will introduce filamentous growth and discuss how it is used as a foraging strategy in various fungal species. We will also discuss one type of pathway that controls filamentous growth, the evolutionarily conserved mitogen-activated protein kinase (MAPK) pathway. Much has been learned about how MAPK pathways are activated, including by mucin-type signaling proteins. We will also explore how different MAPK pathways send specific signals through shared proteins that function in many pathways. In addition to signal specificity, we will discuss how MAPK and other pathways function in networks to integrate a response to multiple signals. Because these pathways control cell differentiation, we will also describe how some of the key target proteins of these pathways produce the filamentous cell type.

Filamentous growth is a conserved cell type across fungal species

At a cellular level, most fungal species exist in one or more different cell types. For most fungi, cells spend at least a portion of their life cycle forming filaments (Fig. 1; Chen et al. 2020; Kumar 2021). These filaments vary widely in appearance depending on the species and life stage and can be categorized in different patterns (e.g. true hyphae or false hyphae). In true hyphae, cells do not fully separate by cytokinesis leaving cytoplasmic connections between cells (Silva et al. 2004; Mouriño-Pérez 2013; Roberson 2020). In some species, there is no separation of cellular compartments (Fig. 1, coenocytic hyphae), while in other species, the formation of septa allows migration of some materials between cells (Fig. 1, septate hyphae). In false hyphae, also called pseudohyphae, cells fully separate by cytokinesis and do not share a cytoplasmic connection. In this state, cells are held together by cell adhesion molecules to form chains of cells (Fig. 1, pseudohyphae). All these patterns are grouped together under the umbrella terms filamentous or filamentous-form growth.

Filamentous growth is the normal, or default growth form, for many fungal species such as Ashbya gossypii (Schmitz and Philippsen 2011) and Neurospora crassa (Patel and Free 2019). These models have been useful for studying many aspects of filamentous growth including cytoplasmic streaming and nuclear migration, as well as the regulation of the fungal cell wall. Other species include yeasts, which are a large subcategory of single-celled fungi distributed across 2 major subphyla (ascomycetes and basidiomycetes; Shen et al. 2018). In yeasts, filamentous growth is an inducible growth form. For example, in nutrient replete environments yeast cells reproduce by budding, where cells are ovoid and fully separate (Fig. 1, yeast-form growth). However, in response to changes in the environment such as nutrient limitation, yeast cells can also undergo filamentous growth (Fig. 1, arrows; Chen et al. 2020; Kumar 2021). Interestingly, some yeasts (such as the saprotroph Saccharomyces cerevisiae and human pathogen Nakaseomyces glabratus formerly Candida glabrata) only make pseudohyphae (Gimeno et al. 1992; Csank and Haynes 2000), while other yeasts [such as the plant pathogen Ustilago maydis (corn smut)] make septate hyphae (Brefort et al. 2009). The most common human pathogen Candida albicans can exist in multiple filamentous types at its physiological temperature, 37°C (Sudbery 2011; Chen et al. 2020). Contrastingly, another category of human pathogens known as the dimorphic fungi (Histoplasma capsulatum, Blastomyces dermatitidis, and Coccidioides posadasii) exist as yeast-form at physiological temperature (37°C) and filamentous molds at lower temperatures. Therefore, although there is quite a variety in filamentous growth patterns, most species of fungi can undergo filamentous growth during at least part of their life cycle.

Filamentous growth makes fungi excellent scavengers because filaments grow outwards to allow the expansion of nonmotile cells into new environments. Filaments can also play important roles in sexual reproduction. For example, in some species of multicellular fungi, filaments can switch from a branching pattern to forming parallel arrangements to form reproductive structures, such as mushrooms, baskets, and cushions (Fig. 2, Reproduction, Mushroom Formation; Chiu and Moore 1996; Money 2016). Filamentous growth also generates branching mycelial networks in multicellular species (Fig. 2; Mycelial Networks; Heaton et al. 2012; Fricker et al. 2017), which are important for obtaining and transporting nutrients over large distances and in heterogenous environments. In fact, the largest organism in the world is composed of a single mycelial network, identified as the honey mushroom Armillaria ostoyae, which covers 9 km² in the Malheur National Forest in Oregon.

Filaments can also penetrate into substrates when invading new territories, by a process known as invasive growth. For symbiotic mycorrhizal fungi, invasive cells establish mutualistic connections with plant roots to exchange nutrients between the plant and the fungal mycelial networks. This has the effect of greatly expanding the plant's root network to benefit both species, providing sugars to the fungus and micronutrients and water to the plant (Dighton 2009; Bonfante and Genre 2010; Figueiredo et al. 2021). Similarly, in lichen-forming fungi, filamentous growth is important to establish symbiotic connections with algae (Fig. 2; symbiosis, lichen formation; Wang, Li et al. 2023). Although some species of pathogens do not undergo filamentous growth in the host, many human (Brand 2012; Chen et al. 2020), insect (Islam et al. 2021; Zhang, Meng, et al. 2021) and plant pathogens (Kahmann and Kämper 2004; Haueisen and Stukenbrock 2016) form filaments that penetrate and weave their way through the host to extract nutrients and complete their life cycles (Fig. 2, pathogenesis).

The interactions between yeasts and multicellular organisms have become a fertile area for new discoveries since invertebrates such as the roundworm Caenorhabditis elegans (Mylonakis et al. 2007; Pukkila-Worley et al. 2011) and larvae of the wax moth Galleria mellonella (Trevijano-Contador and Zaragoza 2018) have been developed into infection models. These models are surprisingly relevant to mammalian infection and have allowed the genetic dissection of the fungal and host pathways that control colonization and infection (Phadke et al. 2018). For example, multispecies interactions in C. elegans showed that probiotic yeasts (and their secondary metabolites) can inhibit infection by pathogens (Peleg et al. 2008; Kunyeit et al. 2019; Kunyeit et al. 2021). Similarly, the human oral cavity is a complex habitat where microbes display complex interactions and show striking spatial heterogeneity (Montelongo-Jauregui et al. 2019; Banerjee et al. 2024; Sulyanto et al. 2024). Yeasts have evolved specific strategies to colonize the host—and the host has developed responses to limit fungal invasion. Pathogens produce toxins that damage mammalian cells, such as the peptide candidalysin from C. albicans (Moyes et al. 2016; Verma et al. 2017), and humans produce antimicrobial compounds, such as the salivary histatin proteins (Puri and Edgerton 2014) that interfere with fungal colonization by disrupting membrane integrity. Recently, candidalysin has been shown to have a major role in establishing C. albicans colonization amongst the complex bacterial microbiota in the mammalian gut (Liang et al. 2024). Further studies of the regulation of fungal responses in these systems have the potential to reveal how interactions between yeasts, other microorganisms, and the host impacts coevolution and virulence. For many of the above examples, a common regulatory element of filamentous growth and fungal behavioral responses is the MAPK pathway (Fig. 2, MAPK). The MAPK pathway is a signal transduction pathway that responds to diverse stimuli to regulate filamentation responses in many contexts.

Filamentous growth has been studied in many different species, and this review will focus on a few model systems. One is the Baker's yeast S. cerevisiae, which has proved a useful genetically amenable model to study the regulatory pathways that control the response. Yeast gives us an ever-expanding genetic, molecular, and functional genomics toolkit that is readily available (Mortimer and Johnston 1986; Roman 1986; Botstein et al. 1997; Borneman et al. 2006; Replansky et al. 2008; Dowell et al. 2010; Botstein and Fink 2011; Ryan et al. 2012). Even though S. cerevisiae has long been considered mostly a tool to study cell biology, yeast research is now capitalizing on the growing knowledge of the organism's ecology and evolution (Hittinger 2013; Goddard and Greig 2015; Liti 2015; Duan et al. 2018; Peter et al. 2018; Bai et al. 2022; De Chiara et al. 2022; O'Donnell et al. 2023; Peris et al. 2023) to better contextualize differentiation events and the signaling pathways that control them. Furthermore, although S. cerevisiae is a free-living organism not widely considered a pathogen, strains have been cultured from immunocompromised individuals, which has allowed exploration of this facet of yeast biology as well (Wheeler et al. 2003; Granek et al. 2013; Raghavan et al. 2019; Ekdahl et al. 2023).

Another important model is C. albicans (Nobile and Mitchell 2006; Noble et al. 2010; Nobile and Johnson 2015; Basso et al. 2019; Chen et al. 2020), a commensal yeast found in most healthy individuals that can become deadly when the immune system is compromised. Several plant pathogens have also been studied, which allows comparison of the similarities and differences between plant and animal colonization (Lanver et al. 2010, 2014; Liu et al. 2011; Perez-Nadales and Di Pietro 2015; Qin et al. 2021; Wang et al. 2021). As a result of evolution, most fungal species share conserved regulatory pathways that govern the filamentous growth response making model systems highly relevant. Interestingly, although the components of these pathways are conserved, what the pathways sense varies based on the specific cues found in species-specific environments. In addition, depending on the species, the same regulatory pathways can induce quite different responses. By exploring strategies that related organisms use to solve similar biological problems, new insights have emerged (Johnson 2017; Wolfe and Butler 2022). For example, S. cerevisiae has undergone a whole-genome duplication (Wolfe and Shields 1997) in which diversification of gene function has occurred through subfunctionalization for some regulatory processes (Hickman et al. 2011), whereas C. albicans maintains a single regulator for some of these similar processes. Because of the strong evolutionary ties between species, lessons learned by studying filamentous growth in 1 species benefits the overall knowledge of the principles underlying developmental cell-type specification in eukaryotes.

Regulation of filamentous growth by a conserved MAPK pathway

Many fungal species regulate cell differentiation through homologous signaling pathways (Ryan et al. 2012; Kiss et al. 2019). One of the main pathways, which regulates filamentous growth, is called the filamentous growth or filamentation MAPK (fMAPK) pathway (Cek pathway in C. albicans; Liu et al. 1993; Roberts and Fink 1994; Kumar 2021). MAPK pathways regulate filamentous growth in saprotrophs (Madhani et al. 1997; Pandey et al. 2004; Xu et al. 2016), human pathogens (Chen et al. 2020), plant pathogens (Xu 2000; Cho et al. 2007), insect pathogens (Jin et al. 2014; Zhao et al. 2023), and lichen-forming fungi (Wang, Li et al. 2023). MAPK pathways also govern cell differentiation and stress responses in mammals and control critical processes including cell proliferation, cell migration, cell survival, and apoptosis (Seger and Krebs 1995; Roberts et al. 2000; Zhang and Liu 2002; Krens et al. 2006; Shaul and Seger 2007; Taj et al. 2010). As we will see below, key regulatory features of the fMAPK pathway apply throughout fungal species (including pathogens) and to eukaryotes in general. Fungal-specific regulatory elements in turn lend themselves to strategies aimed at curbing virulence in fungal pathogens.

A kinase module controlled by the Rho GTPase Cdc42 regulates multiple MAPK pathways: the puzzle of specificity

A defining feature of MAPK pathways is they are composed of multiple kinases that phosphorylate and activate each other in a tandem series (Seger and Krebs 1995). The MAP kinase (MAPK, Kss1 for the fMAPK pathway) is phosphorylated and activated by the MAP kinase kinase (MAPKK, Ste7p), which in turn is phosphorylated and activated by a MAP kinase kinase kinase (MAPKKK, Ste11p) (Fig. 3, Cook et al. 1996, 1997; Bardwell, Cook, Voora, et al. 1998; Lee and Elion 1999; Shock et al. 2009). Ste11p is itself activated by phosphorylation by a member of the p21-activated kinases (PAKs), called Ste20p. Ste20p is activated by binding to the ubiquitous Rho-type GTPase Cdc42p. Cdc42p is a member of the monomeric small GTPase superfamily of proteins (Johnson and Pringle 1990; Johnson 1999; Pruyne and Bretscher 2000) and is highly conserved across eukaryotes sharing 81% protein sequence identity between yeast and humans. As for many members of the GTPase family, in its active or guanosine triphosphate (GTP)-bound conformation, Cdc42p binds effector proteins including Ste20p (Peter et al. 1996; Lee and Elion 1999; Tatebayashi et al. 2006) and other proteins involved in polarity organization (Gic1p, Gic2p, and another PAK, Cla4p). Specifically, Cdc42p is activated by the guanine-nucleotide exchange factor (GEF, Cdc24p), which promotes the exchange of GTP for guanosine diphosphate. GTP-bound Cdc42p binds to and relieves autoinhibition of Ste20p's Cdc42p-binding or CRIB domain, which stands for Cdc42/Rac interactive binding motif, allowing the PAK to phosphorylate and activate Ste11p (van Drogen et al. 2000). Additionally, 14-3-3 proteins (Bmh1p and Bmh2p; Roberts et al. 1997) and an adaptor protein (Bem1p; Basu et al. 2020) facilitate activation of Ste20p. Therefore, a signaling cascade composed of a Cdc42p module and 4 kinases in a tandem series represents the core of the fMAPK pathway (Fig. 3, green).

Signaling pathways are not separate entities from each other. Most pathways share common factors, even in the same cell type. In S. cerevisiae, the fMAPK pathway shares components with at least 2 other MAPK pathways: The mating and high osmolarity glycerol (HOG) pathways (Fig. 3; Roberts and Fink 1994; Posas and Saito 1997). The mating (or pheromone response, blue) pathway allows haploid cells to recognize each other by secreted peptide pheromones that leads to cell fusion and the formation of diploids (Herskowitz 1995). The HOG pathway (red) regulates tolerance to osmotic and other stresses (Saito and Tatebayashi 2004; de Nadal and Posas 2022) and modulates filamentous growth (Shock et al. 2009; Yang et al. 2009; Adhikari and Cullen 2014) through 2 redundant branches (Ste11p and Sln1p). Although each pathway senses a different stimulus and induces a distinct set of targets genes, they all utilize the same core module (Fig. 3, gray box). Moreover, any 2 pathways also utilize some of the same components. For example, the mating and fMAPK pathways both require the same MAPKK, Ste7p. How the same core module induces different responses to different stimuli is a longstanding question that extends to signaling modules in general. Studies of MAPK pathways in yeast have revealed fundamental mechanisms that cells have developed to keep their signals straight.

Below, we point out features of the fMAPK pathway in reference to the other pathways. We compare the receptors/sensors between the pathways that contribute to specificity, including the mating G-protein-coupled receptors (GPCRs) and the mucin sensors of the fMAPK and HOG pathways. We discuss how scaffolds maintain insulation between pathways, and the cross talk that occurs when specificity is lost. Finally, we show how different target genes are induced by shared transcription factors. In this way, as we learn about the fMAPK pathway, we will also learn how specificity is maintained between MAPK pathways and address open questions in this area.

A mucin sensor at the head of the fMAPK pathway

Signaling pathways can be activated by many different types of receptors. The sensor for the fMAPK pathway is a member of the mucin family of proteins (Fig. 4, Msb2p; Cullen et al. 2004). Mucins differ from the more common GPCRs, whose ligands are well defined (Lemaire et al. 2004; Xue et al. 2008), structures resolved (Velazhahan et al. 2021, 2022), and connections to effector proteins firmly established (Lee and Dohlman 2008). Mucins by comparison are broadly similar to a diverse group of glycoprotein sensors, such as Notch (Zhou et al. 2022), cadherins (Fulford and McNeill 2020), and dystroglycan (Endo 2015). As a group, these proteins are interesting because of their mechanosensory properties, the varied effects of glycosylation and posttranslational processing on signaling, and the challenges associated with studying large glycoproteins by biochemical and structural methodologies (Pei and Grishin 2017).

Mucins are conserved in fungi, invertebrates, and vertebrates and are highly glycosylated proteins that are secreted to provide a layer of protection, and a layer of lubrication for epithelium tissue in animals. Because of their presence in the gut and oral epithelium, mucins play important roles in the digestive and immune systems (Johansson and Hansson 2016). Mucins that regulate signaling pathways have a single-pass transmembrane helix and cytosolic domain and are called signaling mucins (Corfield 2015). In humans, the signaling mucin MUC1 regulates the RAS-MEK-ERK pathway, which controls key cellular responses including cell proliferation (Theodoropoulos and Carraway 2007) and when misregulated is a main cause of cancer. MUC1 is itself up-regulated in many cancers and is therefore a target for immunotherapies (Singh and Bandyopadhyay 2007; Sousa et al. 2016; Supruniuk and Radziejewska 2021; Qing et al. 2022). One defining feature of mucins is a tandem repeat region rich in serine, threonine, and proline (S/T/P) residues (Desseyn et al. 1998; Cao et al. 2012). Because S/T residues are targets of O-linked glycosylation (in yeast typically poly-mannosylation), mucins can be abundantly glycosylated; so much so that the carbohydrate content outweighs the protein content of the protein. Variation in the number of repeats can occur and have phenotypic consequences. For example, increases in the repeat number of MUC1 is associated with kidney disease (Dvela-Levitt et al. 2019). The roles mucin signalers play in regulatory pathways and their prevalence in disease make them important molecules of interest.

Msb2p shares many defining features of signaling mucins (Cullen et al. 2004). It is a single-pass transmembrane protein that is highly glycosylated, resides at the cell surface, and is required for activity of the fMAPK pathway. Msb2p also contains an S/T/P-rich region of tandem repeats of 17 amino acids in length (Fig. 4, repeats). The S/T/P region is part of a larger portion of the extracellular domain (from residues 100 to 950 amino acids) that inhibits the fMAPK pathway (Vadaie et al. 2008). Moreover, this large portion of the extracellular domain is shed from cells, which is a typical feature of mucins and other signaling glycoproteins. Shedding of the inhibitory domain of Msb2p requires members of a family of glycosylphosphatidylinositol anchored aspartyl proteases called yapsins (Krysan et al. 2005). Processing and release of the inhibitory domain of Msb2p leads to cleavage-dependent activation of the fMAPK pathway. Cleavage-dependent activation is common to the activation of other glycoprotein sensors, including Notch (van Tetering and Vooijs 2011), adhesion GPCRs (Araç et al. 2012), and even several bacterial proteases (Jeong et al. 2020), including one that when shed induces virulence (Dong et al. 2004).

What is Msb2p sensing? An established inducer of filamentous growth is nutrient limitation. The limitation of a preferred carbon source (e.g. glucose) and/or nitrogen induces filamentous growth in S. cerevisiae (Gimeno et al. 1992; Roberts and Fink 1994; Cullen and Sprague 2000), human pathogens (Brown et al. 1999; Csank and Haynes 2000; Biswas and Morschhäuser 2005; Morschhäuser 2011; Ng et al. 2016), and plant pathogens (Banuett and Herskowitz 1994; Van den Ackerveken et al. 1994). Moreover, the expression of genes related to filamentous growth is linked in some cases to glucose repression, where adequate glucose inhibits the expression of the metabolism of other carbon sources (Carlson 1999).

Msb2p may respond to glucose limitation in a noncanonical way. Upon glucose limitation, the glycosylation of Msb2p becomes reduced, presumably because the substrates for its glycosylation are glucose and mannose moieties (Fig. 4, Adhikari, Vadaie et al. 2015). Msb2p is glycosylated in the endoplasmic reticulum (ER), and problems with protein folding/glycosylation in the ER can trigger the unfolded protein response (UPR, Fig. 4), a regulatory pathway in the ER that is responsible for dealing with protein folding stress in eukaryotes including fungi and mammals (Walter and Ron 2011). The UPR induces the levels of yapsin proteases resulting in elevated cleavage of Msb2p (Adhikari, Vadaie et al. 2015). In this way, Msb2p functions as an unconventional glucose sensor that is activated by cleavage when carbon levels become limited (Fig. 4). In support of this idea, problems in protein glycosylation, especially O-linked glycosylation by the mannoyltransferase Pmt4p (Yang et al. 2009), also result in elevated Msb2p-dependent fMAPK pathway activity (Cullen et al. 2004).

In addition to nutrient limitation, recent studies suggest that the fMAPK pathway responds to other stimuli connected to the cell's ability to colonize fruit-based ecological niches (Vandermeulen and Cullen 2023). One stimulus is galactose, a nonpreferred carbon source found in some plant-rich environments (Botha 2011; Gunina and Kuzyakov 2015) and certain fruits (Folsom et al. 1974; Asgar et al. 2003; Qi and Tester 2019). Galactose stimulates the activity of the fMAPK pathway (Karunanithi and Cullen 2012; Adhikari and Cullen 2014) beyond glucose limitation and requires galactose metabolism (Vandermeulen and Cullen 2023). Another stimulus is pectin (Vandermeulen and Cullen 2023), a major component of plant cell walls (JARVIS 1984; Willats et al. 2001; Voragen et al. 2009; Zdunek et al. 2021). Pectin may be sensed by Msb2p at the cell surface directly, as some mammalian mucins expressed in gut tissue interact with pectin (Thirawong et al. 2008; Sriamornsak et al. 2010). Alternatively, pectin molecules are large and may bring Msb2p and other sensors (Sho1p, discussed below) into a complex. Similarly, pectin and galactose induce the secretion of pectolytic enzymes in the absence of glucose in other fungi, such as N. crassa (Polizeli et al. 1991; Crotti et al. 1996; Crotti, Terenzi, Jorge, de Lourdes et al. 1998; Crotti, Terenzi, Jorge, Polizeli 1998).

The third molecule that induces the fMAPK pathway is the metabolic byproduct ethanol (Vandermeulen and Cullen 2023). Ethanol and other alcohols in S. cerevisiae (Dickinson 1996; Lorenz et al. 2000; Chen and Fink 2006; Wuster and Babu 2009) and C. albicans (Chen et al. 2004; Nickerson et al. 2006; Albuquerque and Casadevall 2012) are soluble metabolites and indicators of cell density, which in microbes is known as quorum sensing (Ng and Bassler 2009; Albuquerque and Casadevall 2012; Whiteley et al. 2017; Zhao et al. 2020; Tian et al. 2021). In S. cerevisiae, ethanol can act as an inhibitor of microbial competitors (Piskur et al. 2006; Dashko et al. 2014) and an attractant for insect vectors (Goddard et al. 2010; Stefanini et al. 2012; Palanca et al. 2013; Buser et al. 2014; Liti 2015).

When combined with galactose, ethanol induces the fMAPK pathway to high (i.e. near maximal) levels, suggesting quorum and nutrient-sensing mechanisms can be additive to pathway activity (Vandermeulen and Cullen 2023). Ethanol may be detected through its ability to disrupt the structure of cell membranes (which is where the sensors/receptors are located) or its ability to denature proteins and activate heat shock proteins (HSPs; Toth et al. 2014; Cohen 2018). Alternatively, the ethanol signal may come through a separate signaling pathway that regulates fMAPK pathway activity (RAS-PKA, discussed below) as Ras2p is more critical to fMAPK pathway activity than Msb2p in response to ethanol (Vandermeulen and Cullen 2023). The fMAPK pathway may also sense cell density through cell-to-cell contact, as Msb2p has been proposed to function as a pressure sensor in some contexts (Delarue et al. 2017) although this possibility has not been tested in the context of fMAPK pathway signaling.

Msb2p interacts with the tetraspan protein Sho1 and cysteine-rich protein Opy2

Msb2p functions with 2 other proteins at the plasma membrane (Fig. 4). One is Sho1p (Maeda et al. 1995; O'Rourke and Herskowitz 1998; Cullen et al. 2004), a tetraspan protein important for filamentous growth signaling in fungi (Lambou et al. 2008). Sho1p has 4 transmembrane helices and a cytosolic region containing an SH3 domain. The cytosolic domain of Sho1p functions as an adaptor that scaffolds together multiple proteins that regulate the MAPK pathway, including the PAK Ste20p and MAPKKK Ste11p (Tatebayashi et al. 2006). By comparison, the SH3 domain binds to the MAPKK of the HOG pathway, Pbs2p (Maeda et al. 1995). Loss of Pbs2p results in crosstalk, where salt induces the activation of the fMAPK and mating pathways (O'Rourke and Herskowitz 1998). Msb2p interacts with Sho1p and may activate the protein. The activatable part of Sho1p is connected to an external loop between the second and third transmembrane helices, which when mutated (typically P120L) causes hyperactivity (Tatebayashi et al. 2007; Vadaie et al. 2008). Msb2p and Sho1p also interact with Opy2p, a single-pass transmembrane protein that contains an extracellular region with 8 cysteines that form disulfide bonds in a predicted knot structure (Wu et al. 1999, 2006; Yamamoto et al. 2010). Opy2p by its Ras-associated domain binds to an adaptor protein (Ste50p) that recruits and activates the MAPKKK Ste11p at the plasma membrane (Fig. 4; Wu et al. 1999; Ramezani-Rad 2003). Msb2p, Sho1p, and Opy2p form a complex that assembles in predictable hetero-oligomeric ratios (Tatebayashi et al. 2015). Puzzlingly, each member of the complex is localized and turned over independently of the other members. Msb2p is turned over rapidly by the ubiquitination of lysines in the cytosolic region of the protein (Adhikari, Vadaie et al. 2015), whereas Sho1p and Opy2p are stable proteins turned over at lower rates (Adhikari, Caccamise et al. 2015). It will be interesting to understand how these proteins cooperate in signaling with their unique turnover and localization properties.

Conserved and unique features of mucins in different fungal species

Msb2p, Sho1p, and Opy2p have homologs in other species, although the sensing mechanisms of the complex may be tailored for specific ecological niches. In C. albicans, Msb2p and Sho1p homologs induce filamentous growth by activation of the homologous Cek1p MAPK pathway (Csank et al. 1998; Román et al. 2009). During activation, Msb2p is cleaved by the aspartyl-type protease Sap8p [(Puri et al. 2012) similar to S. cerevisiaeYps1p] and shed into the extracellular matrix. The aspartyl-type proteases are members of large families: there are 5 cell wall anchored yapsins in S. cerevisiae (Krysan et al. 2005), 11 yapsins in C. glabrata (Kaur et al. 2007), 10 secreted aspartyl proteases (8 secreted, and 2 cell-wall anchored) in C. albicans (Monod et al. 1994; Albrecht et al. 2006), and 7 in Candida auris (Kim et al. 2023). The amplification of gene families can produce functional diversity. For example, by regulating the expression, activity, and specificity of these proteases, cells can tailor protease activity and hence penetration into specific environments.

In the extracellular space, the large glycol-domain of Msb2p functions to protect cells against antimicrobial peptides (Szafranski-Schneider et al. 2012). Cleavage of the C. albicansMsb2p occurs in nutrient-limiting environments and elevated temperatures corresponding to host conditions (37°C). Moreover, in this species, Msb2p may specifically play a role in temperature sensing (Saraswat et al. 2016). Interestingly, in another human pathogen, Aspergillus fumigatus, the Msb2p homolog induces a related MAPK pathway controlling cell integrity (Gurgel et al. 2019).

Msb2p and Sho1p homologs also regulate MAPK pathways involved in host recognition, appressorium formation, filamentous growth, and invasion in the plant pathogens. These include U. maydis (Lanver et al. 2010, 2014), Fusarium oxysporum (Perez-Nadales and Di Pietro 2015), Magnaporthe oryzae (Liu et al. 2011), Colletotrichum gloeosporioides (Wang et al. 2021), and Aspergillus flavus (Qin et al. 2021). Host recognition by Msb2p and Sho1p occurs by sensing hydrophobicity and by chemical signals associated with the cuticle of the plant surface. Msb2p regulates a related MAPK pathway (i.e. HOG) in Arthrobotrys oligospora that induces a different type of differentiation program for trap formation by nematode predators (Kuo et al. 2020). Therefore, Msb2p-type mucins have functional homologs in other fungal species, although their sensing mechanisms and effector MAPK pathways differ, perhaps as a result of selective pressure due to diverse ecological niches.

Scaffolds promote specificity between MAPK pathways that share components

How do proteins at the plasma membrane (Msb2p, Sho1p, and Opy2p) connect to and regulate the Cdc42p-dependent MAPK cascade? One clue toward answering this question comes from protein scaffolds. In signaling pathways, scaffolds bind to general and pathway-specific proteins to induce a pathway-specific response (DiRusso et al. 2022). In the mating pathway, the scaffold Ste5p links the GPCR, namely the G_β protein Ste4p to Ste20p, Ste11p, Ste7p, and the MAPK Fus3p (Fig. 3; Zalatan et al. 2012). In this way, signals emanating at the receptor are selectively and efficiently directed toward a specific MAPK pathway. Furthermore, Ste5p binding to Fus3p catalytically unlocks the protein, enabling it to be phosphorylated and activated by its MAPKK Ste7p (Good et al. 2009; Coyle et al. 2013). As mentioned above, Pbs2p is the MAPKK and scaffold for the HOG pathway that binds to the tetraspan protein Sho1p. Again, this has the effect of linking a protein that operates at the plasma membrane (Sho1p) to a pathway-specific MAPK (Hog1p, by way of Pbs2p, Fig. 3).

For the fMAPK pathway, a different type of scaffold functions specifically in the fMAPK but not mating or HOG pathways, called Bem4p (Figs. 3 and 4; Pitoniak et al. 2015). Bem4p is a Cdc42p-binding protein with homology to SMG GDS-type regulators of GTPases (Brandt et al. 2021). Bem4p is required for activation of the fMAPK pathway but not the mating or HOG pathways. Moreover, Bem4p binding to Cdc42p protects the protein from turnover (González et al. 2023). Specifically, the ubiquitin ligase Rsp5p and HSPs Ssa1p and Ssa2p mediate turnover of GTP-bound Cdc42p, and this effect is countered by Bem4p. Although many GTPases are known to be regulated by turnover, here the turnover of the active or GTP-bound conformation of Cdc42p leads to attenuation of the fMAPK pathway. The current model is that Bem4p in some manner protects Cdc42p to sustain fMAPK pathway activity. Cdc42p protection does not impact the mating pathway because that pathway is gated: Fus3p cannot be phosphorylated by Ste7p without the activation of Ste5p (Good et al. 2009; Coyle et al. 2013). Therefore, Bem4p functions at a distinct point in the pathway to amplify signals generated by activated Cdc42p.

Bem4p not only interacts with Cdc42p but also with its major activator, the GEF Cdc24p. Two other proteins also bind Cdc24p and promote Cdc42p function in the fMAPK pathway (Fig. 4, Rsr1 and Bem1). One of these proteins is the Ras-type GTPase Rsr1p. Interestingly, Rsr1p is itself activated by proteins that control the direction in which cells grow or bud, called bud-site-selection proteins, a mechanism of polarity establishment (Park et al. 1997). In haploid cells, cells grow back toward the mother in an axial pattern when nutrients are plentiful and away from the mother to make pseudohyphae in a distal pattern when nutrients become limiting (Chant and Pringle 1995). Bud-site-selection proteins located at either pole recruit Rsr1p, which directs GEF-dependent activation of Cdc42p at the growth site. Interestingly, Rsr1p contributes to fMAPK pathway activation (Basu et al. 2016), presumably as a way of integrating positional information into the regulation of MAPK pathway signaling.

In addition to Bem4p and Rsr1p, the third protein that binds Cdc24p and regulates Cdc42p is the main polarity scaffold Bem1p. Bem1p regulates Cdc42p in multiple contexts—it is required for symmetry breaking during polarity establishment (Woods and Lew 2019), and it regulates the mating, fMAPK, and HOG pathways (Leeuw et al. 1995; Elion 2000; Tanaka et al. 2014). Nevertheless, a direct comparison shows that different domains of Bem1p play different roles in different pathways (Fig. 4, mark on Bem1; Basu et al. 2020). Similarly, specific residues of the adaptor Ste50p mediate pathway-specific functions (Sharmeen et al. 2019). Therefore, even proteins that are shared between pathways can have pathway-specific determinants. It will be interesting to learn how Msb2p and Sho1p connect to cytosolic signaling proteins (Fig. 4, top question mark), and how Bem4p, Rsr1p, and Bem1p coordinate GEF activation in the specific context of filamentous growth.

Much remains to be learned by additional studies on the fMAPK pathway. One example comes from the fact that 2 mucins regulate MAPK pathways in yeast. One is Msb2p and the other is Hkr1p (Fig. 3). Msb2p regulates the fMAPK pathway, while Msb2p and Hkr1p regulate the HOG pathway (Tanaka et al. 2014). Disentangling specificity between these mucins is an area of intense interest. A second example comes from evidence for a Ste20p-independent branch of the MAPK pathway (Fig. 4, curved arrow, bottom question mark). In the HOG pathway, Ste20p can be bypassed by the addition of salt (Raitt et al. 2000). Similarly, conditions that hyperactivate the fMAPK pathway (galactose with ethanol; Vandermeulen and Cullen 2023) or some alleles that hyperactivate the fMAPK pathway (González et al. 2023), lead to a signal that requires Ste11p but not Ste20p. Future work can clarify this mechanism to understand how it works and why it evolved.

Regulation of transcription factors and gene expression by the fMAPK pathway

In yeast, the MAPK Kss1p functions mainly in the fMAPK pathway, although it can modulate mating in some circumstances (Ma et al. 1995; Cherkasova et al. 1999; Farley et al. 1999). Unlike other MAPKs, Kss1p functions in an inhibitory capacity until it is phosphorylated [Fig. 4, block arrow at Kss1, (Cook et al. 1997; Gustin et al. 1998)]. Relief of the inhibitory role of Kss1p by phosphorylation by Ste7p leads to the activation of several transcription factors. One of these is Ste12p, a homeodomain protein that regulates the fMAPK and mating pathways. Here, the puzzle of specificity surfaces again, this time at the transcription factor level. How is Ste12p directed to a specific gene set? The answer to this question comes from the discovery of a TEA/ATTS family transcription factor, named for the members of the family, Tec1p, AbaA, TEF1/TEAD1, and Scalloped. Tec1p can hetero-dimerize with Ste12p (Fig. 4; Laloux et al. 1990; Madhani and Fink 1997; Borneman et al. 2006; Chou et al. 2006). Ste12p and Tec1p each recognize different DNA-specific binding motifs and are conserved across fungi (Schweizer et al. 2000; Vallim et al. 2000; Wong Sak Hoi and Dumas 2010; León-Ramírez et al. 2022). During mating, Ste12p binds pheromone response elements as a homodimer. Tec1p does not interfere with mating in this setting because Fus3p phosphorylates Tec1p, which targets the protein for ubiquitination and destruction by the 26S proteosome (Bao et al. 2004; Chou et al. 2004). In this way, Ste12p is directed solely to the mating pathway upon exposure to pheromone. In filamentous growth inducing environments, Ste12p forms a Ste12p–Tec1p heterodimer, leading to these transcription factors binding to distinct sites containing TCS elements, which contain adjacent pheromone (Ste12p) and filamentation (Tec1p) binding sites (Zeitlinger et al. 2003; Zhou et al. 2020), to regulate filamentation related target genes (Madhani et al. 1999; Roberts et al. 2000; Heise et al. 2010; Adhikari and Cullen 2014; van der Felden et al. 2014; Chow, Starr, et al. 2019; Zhou et al. 2020).

The transcription factors Ste12p and Tec1p are regulated by transcriptional modulators, Msa1p and Msa2p (van der Felden et al. 2014), and are inhibited by the transcriptional repressors, Dig1p and Dig2p (Cook et al. 1996, 1997; Bardwell, Cook, Zhu-Shimoni et al. 1998). Ste12p and Tec1p also have independent regulatory mechanisms (Kohler et al. 2002; Heise et al. 2010; van der Felden et al. 2014) some of which depend on the environment (Vandermeulen and Cullen 2023), which may be due to their regulation by unique mechanisms. For example, Ste12p can also associate with Mcm1p, an essential transcription factor (Zeitlinger et al. 2003) and is phosphorylated by Cdk8p, a cyclin-dependent kinase (CDK; Nelson et al. 2003). Tec1p can be independently modulated by SUMOylation (Wang et al. 2009). Preferences in binding sites may also influence transcription factor specificity (Dorrity et al. 2018).

Transcriptional targets of the fMAPK pathway control aspects of the filamentation response

The main function of the transcription factors for the fMAPK pathway is to control the expression of genes whose products mount the filamentous growth response. The specific gene set based on multiple studies contains more than fifty genes that regulate cell differentiation, cell adhesion, and enzyme secretion (Madhani et al. 1999; Zeitlinger et al. 2003; Adhikari and Cullen 2014; Chow, Starr et al. 2019; Vandermeulen and Cullen 2020). In addition, the fMAPK pathway up-regulates several of its own components (e.g. MSB2, KSS1, STE12, and TEC1) to induce positive feedback, presumably to amplify the initial signal. The fMAPK pathway also up-regulates negative regulators of the pathway (i.e. NFG1/YLRO42c, RGS2, RPI1, and TIP1) to modulate pathway activity and filamentous growth (Vandermeulen and Cullen 2020). Collectively, these feedback-control mechanisms allow the cell to precisely set the level of its filamentous growth response.

Cell differentiation by polarity and cell cycle target genes

Perhaps the most striking feature of filamentous growth is the ability of cells to elongate and begin moving away from parent cells in search of nutrient-rich environments. The apparent complexity of this growth pattern is caused by induction of a few target genes. The change in polarity or growth direction is particularly clear in haploid cells, which switch from growth toward the mother cell in an axial pattern to growth away from the mother in a distal pattern. Transcriptional induction of the distal pole marker, BUD8 (Fig. 5; Chant and Pringle 1995; Zahner et al. 1996; Taheri et al. 2000; Ni and Snyder 2001; Cullen and Sprague 2002) promotes this outward growth. In addition, reduced levels of an axial-specific protein, Axl1p (Cullen and Sprague 2002) solidifies the distal pattern.

The change in the length of the cell is caused by a delay in cell cycle progression. MAPK pathways commonly regulate progression through the cell cycle by affecting cyclins that bind to and activate CDKs. In this case, the fMAPK pathway induces expression of CLN1, which encodes a G₁ cyclin (Hadwiger et al. 1989), whose induction leads to a delay in progression through the cell cycle resulting in an elongated cell morphology (Fig. 5; Madhani et al. 1999). Cell cycle extension is also mediated by a second mechanism. The fMAPK pathway induces expression of SFG1, which encodes a transcription factor that contributes to cell elongation (Vandermeulen and Cullen 2020) presumably by causing a delay in the cell cycle at the G₂/M phase (White et al. 2009). In other fungi, cell polarization depends on steep phosphatidyl inositol gradients (Vernay et al. 2012), Cdc42 regulation by multiple control proteins (Bassilana et al. 2005; Brand et al. 2014), and the RAM or regulation of Ace2 and morphogenesis pathway (Bharucha et al. 2011; Chadwick et al. 2022).

Cell adhesion

MAPK pathways also up-regulate the expression of genes that control cell adhesion, which is a major aspect of the filamentous growth response. Many fungi have numerous genes that encode cell adhesion molecules that serve different purposes depending on context, which have been extensively reviewed (Hoyer 2001; Bruckner and Mosch 2012; Lipke 2018; Essen et al. 2020). In S. cerevisiae, mechanisms underlying the regulation of cell adhesion have been worked out, which includes the fMAPK pathway-dependent up-regulation of the expression of FLO11 and SFG1 (Fig. 5), 2 genes that encode proteins that regulate separate cell adhesion mechanisms.

A large amount of research has been done on FLO11, which encodes a cell surface mucin-type protein and major cell adhesin involved in filamentous growth in S. cerevisiae (Fig. 5, Lambrechts et al. 1996; Lo and Dranginis 1996, 1998; Guo et al. 2000). For example, extensive structure and function analysis has been performed on Flo11p (Verstrepen et al. 2005; Christiaens et al. 2012; Meem and Cullen 2012; Kraushaar et al. 2015), revealing hydrophobic and homotypic binding sites. Additionally, studies on the regulation of FLO11 gene expression reveal extensive regulation at its unusually large promoter (Rupp et al. 1999; Palecek et al. 2000; Borneman et al. 2006; Barrales et al. 2008; Bumgarner et al. 2012), and evolutionary studies have shown strain and functional variation (Fidalgo et al. 2006; Barua et al. 2016; Brückner et al. 2020). Moreover, Flo11p has been shown to be shed into the extracellular space to attenuate adhesion and contribute to formation of a mucus layer (Karunanithi et al. 2010). Flo11p is also critical for biofilm/mat formation (Reynolds 2018), and the protein can be expressed in a variegated manor among individuals in a population (Halme et al. 2004). Other adhesin genes exist in S. cerevisiae (e.g. FLO1, FLO10, FLO9, FLO5; Guo et al. 2000; Veelders et al. 2010; Bruckner and Mosch 2012; Christiaens et al. 2012); however, they are thought to be silenced at the gene level in some strain backgrounds (Halme et al. 2004) and only play a minor role compared to FLO11. These adhesins are involved in promoting the formation of adhesion clusters in liquid environments (flocs when suspended, flors when on the surface; Lindquist 1952; Mill 1964; Bruckner and Mosch 2012).

Beyond Flo11p, the inhibition of cell-wall-degrading enzymes that regulate cell separation also plays a role in regulating cell adhesion (Vandermeulen and Cullen 2020). A family of cell-wall-degrading enzymes (e.g. glucanases and endo-chitinases; King and Butler 1998; Doolin et al. 2001; Baladrón et al. 2002; Roncero and Sanchez 2010) promote cell separation between mother and daughter cells to reduce adhesion. The fMAPK pathway up-regulates expression of SFG1 (Vandermeulen and Cullen 2020), which encodes a transcription factor that inhibits the expression of cell-wall-degrading enzymes (Fujita et al. 2005; White et al. 2009; Fig. 5). Ace2p is another transcription factor that also inhibits expression of these enzymes (King and Butler 1998; Doolin et al. 2001).

For human pathogens, such as C. albicans, adhesion molecules can play even more critical roles in modulating the strength and specificity of cellular interactions. One key example is the recognition of host cells (Hoyer 2001; Martin et al. 2021). A second is forming fungicidal resistant multicellular biofilms (Chandra et al. 2005; Nobile and Mitchell 2006; Flemming and Wingender 2010; Desai et al. 2014; Silva-Dias et al. 2015) that adhere to other microbes or abiotic surfaces, including plastics found in medical equipment (Kennedy et al. 1989; Pereira et al. 2021; Ponde et al. 2021). Therefore, it is not surprising that there are more cell adhesion molecules in pathogens, and that they are extensively regulated. C. albicans has more than a dozen adhesion molecules in the ALS gene family (Hoyer 2001; Lipke et al. 2012), which have structural homology to the FLO proteins, as well as cell-wall-degrading enzymes that include an ACE2 homolog (Kelly et al. 2004). Many of the adhesion molecules in C. albicans create strong adhesive forces to form biofilms through amyloid-type aggregation (Garcia et al. 2011; Lipke et al. 2018). Different combinations of adhesins show different properties, suggestive of an adhesion code, which could promote redundancy or phenotypic plasticity depending on the environment (Rosiana et al. 2021). Future research may allow analysis of the function of each adhesin, which at present is hampered by the presumed redundancy in these protein families.

For plant-associated fungi (Tucker and Talbot 2001; Taylor et al. 2022) or insect pathogens (Shang et al. 2024), adhesion molecules play critical roles on the host surface. This includes specific cell adhesion molecules found on the adhesive structure called the appressorium (Mendgen et al. 1996; Tucker and Talbot 2001; Ryder and Talbot 2015; Chethana et al. 2021), a specialized structure that generates turgor pressure for breaking through the cuticle of the plant surface for plant pathogens (Kahmann and Kämper 2004), the insect cuticle for insect pathogens (Ortiz-Urquiza and Keyhani 2013), or the plant root for symbiotic mycorrhizal fungi (Demoor et al. 2019). Recently, the insect pathogen Metarhizium robertsii was found to contain a histone lysine methyltransferase called Ash1p, which is up-regulated upon exposure to the insect cuticle to promote the production of peroxisomes that promote lipid hydrolysis (through Pex16p) to produce large amounts of glycerol for turgor generation (Wang, Lai et al. 2023). Additional types of adhesion molecules have been discovered in the plant-associated fungus U. maydis where a novel adhesin promotes hyphal aggregation in plant tumors (Lep1p; Fukada et al. 2021). Likewise, proteins similar to FAS-domain containing fasciclins have been found in U. maydis (Mueller et al. 2008) and the Shiitake mushroom forming fungus Lentinula edodes (Miyazaki et al. 2007).

Additional roles for cell adhesion molecules outside of filamentous growth include aiding the formation of traps in predatory nematode-trapping fungi (Jiang et al. 2017; Lipke 2018), promoting adherence of sexual partners (i.e. agglutination) before the fusion of cells (de Nobel et al. 1995; Zhao et al. 2001; Muller et al. 2003), and acting as a mechanism to recognize “same” individuals or individuals of different species to promote kin selection (Dranginis et al. 2007; Smukalla et al. 2008; Shinn-Thomas and Mohler 2011; Kraushaar et al. 2015; Brückner et al. 2020). Overall, fungi contain numerous mechanisms to regulate cell adhesion with a repertoire of different adhesion strategies that allows for context-specific cell adhesion regulation and nuanced responses when invading new environments and performing other critical tasks.

Secreted enzymes

In addition to the morphological and adhesive changes that occur during filamentous growth, many fungi also secrete molecules into the extracellular milieu (Fig. 5). Broadly speaking, these molecules act to break down extracellular materials and facilitate entry into new environments. Fungi can secrete proteases, like the yapsins and Saps mentioned above, that break down host tissues for human pathogens (Hube 1996; Naglik et al. 2004; Kaur et al. 2007; Tobouti et al. 2016; Safiya et al. 2023), cuticle-degrading enzymes that break down the insect exoskeleton in insect pathogens (Charnley and St. Leger 1991; Zhang, Meng, et al. 2021), and pectinases and cellulases that break down pectin and cellulose, respectively, in the plant cell wall for plant pathogens and saprotrophic fungi (Collmer and Keen 1986; Jayani et al. 2005; Benz et al. 2014; Panchapakesan and Shankar 2016; Lange et al. 2019; Berbee et al. 2020).

In S. cerevisiae, the fMAPK pathway up-regulates expression of the PGU1 gene that encodes a secreted endo-polygalacturnase (Blanco et al. 1998), which breaks down pectin in the environment (Madhani et al. 1999; Gognies et al. 2001, 2006; Gognies and Belarbi 2002) including in the rind of citrus fruit (Vandermeulen and Cullen 2023). It is likely that the fMAPK pathway evolved to regulate pectinase production as the pathway also senses pectin in the external environment as discussed above. The fMAPK pathway also up-regulates the secreted enzyme encoded by SUC2, which is an invertase for hydrolyzing sucrose (Carlson et al. 1981) and contributes to social behaviors (Greig and Travisano 2004; Maclean and Brandon 2008; Koschwanez et al. 2011), such as by the formation of invasive aggregates that allow cells to invade in large groups by making gouges into surfaces (Chow, Dionne et al. 2019).

From pathways to networks: the integration of signals by multiple pathways

Besides the fMAPK pathway, other pathways sense and integrate signals into the filamentous growth response. In pathogens such as C. albicans (Kadosh and Johnson 2005; Hall et al. 2010; Shapiro and Cowen 2010; Du et al. 2012; Su et al. 2018; Pentland et al. 2021) and Na. glabratus (C. glabrata; Sasani et al. 2016; Hassan et al. 2021), these signals include serum, human body temperature (37°C), and elevated CO₂ concentrations. The host immune system can also suppress filamentous growth in pathogens (Gow et al. 2011; Kavanaugh et al. 2014; Vila et al. 2017; Takagi et al. 2022). In plant pathogens, signals include the plant surface (Lanver et al. 2010; Liu et al. 2011; Leroch et al. 2015; Perez-Nadales and Di Pietro 2015; Wang et al. 2021), the plant cell wall (Polizeli et al. 1991; Gognies et al. 2001; Gognies and Belarbi 2002; Vandermeulen and Cullen 2023), and plant hormones, including indoleacetic acid (Prusty et al. 2004) and ethylene (Kolattukudy et al. 1995).

These signals are sensed by a variety of signaling pathways that can be as important for filamentous growth as the fMAPK pathway. One of these is the Ras-cAMP-regulated Protein Kinase A (RAS-PKA) pathway (Kronstad et al. 1998; Borges-Walmsley and Walmsley 2000; Fortwendel 2015; Kayikci and Magwene 2018; Dautt-Castro et al. 2021; Hong et al. 2024). The RAS-PKA pathway induces filamentous growth in response to nitrogen- and/or glucose-limited environments (Gimeno et al. 1992; Pan and Heitman 1999; Hogan and Sundstrom 2009; Cullen and Sprague 2012). The RAS-PKA pathway also regulates cell growth and metabolism (Morishita et al. 1995; Schmelzle et al. 2004; Kunkel et al. 2019) and depending on the species can play roles in other differentiation-type responses. These include responses central to the life cycle, such as sporulation and mating (Fillinger et al. 2002; Lee and Kronstad 2002; McDonald et al. 2009; Fortwendel 2015; Wendland 2020), and appressorium formation (Zhou et al. 2014; Li et al. 2017; Zhu et al. 2017; Qu et al. 2021). A long history of evolutionary conservation links mating, sporulation, and filamentous growth to nutrient starvation in many species, which may explain the requirement for a master nutrient-sensing pathway (RAS) pathway in overseeing these responses.

Many other regulatory proteins and pathways beyond the fMAPK and RAS-PKA pathways have been identified by comprehensive genetic screens in S. cerevisiae (Fig. 6a; Jin et al. 2008; Xu et al. 2010; Ryan et al. 2012; Kiss et al. 2019). These include the Pal/Rim pathway, which positively regulates filamentous growth, pathogenicity, and stress tolerance across fungi in response to pH changes through a related group of transcription factors (e.g. PacC in Aspergillus nidulans and Rim101p in S. cerevisiae and C. albicans; Li and Mitchell 1997; Peñalva Miguel and Arst Herbert 2002; Davis 2003; Lamb and Mitchell 2003; Barrales et al. 2008; Mira et al. 2009; Selvig and Alspaugh 2011; Du and Huang 2016; Li et al. 2022; Vandermeulen and Cullen 2022). Neutral to alkaline pH stimulates filamentous growth in S. cerevisiae (Li and Mitchell 1997; Davis, Edwards, et al. 2000; Davis, Wilson et al. 2000; Lamb and Mitchell 2003; Barrales et al. 2008; Mira et al. 2009) and C. albicans (Nobile et al. 2008; Du and Huang 2016), whereas an acidic pH stimulates filamentous growth in U. maydis (Mayorga and Gold 1999; Martínez-Espinoza et al. 2004), likely because auxins produced by the fungus cause a reduction in pH in plant tissue (Guevara-Lara et al. 2000; Martínez-Espinoza et al. 2004).

Other pathways include the UPR, as mentioned above; the retrograde (RTG) pathway, a nuclear pathway that responds to mitochondrial stress (Bui and Labedzka-Dmoch 2024) and filamentous growth (Aun et al. 2013; Gonzalez et al. 2017; Rollenhagen et al. 2020); the AMP-dependent protein kinase (AMPK) Snf1p, which regulates glucose repression (Carlson 1999; Simpson-Lavy and Kupiec 2023) and filamentous growth (Cullen and Sprague 2000); the target of rapamycin (TOR) pathway, a conserved pathway in eukaryotes that regulates cell metabolism, survival, ribosome biogenesis, and growth (Laplante and Sabtini 2009; Laplante and Sabatini 2012; Gutiérrez-Santiago and Navarro 2023; Wang, Zheng et al. 2023) and controls many aspects of filamentous growth (Cutler et al. 2001); the cell wall integrity protein kinase C pathway (Levin 2005; Birkaya et al. 2009; Zhang, Wang et al. 2021; Yoshimi et al. 2022); a transcriptional regulator of phospholipid biogenesis, Opi1p (Reynolds 2006; Chen et al. 2015; Vandermeulen and Cullen 2022); the alternative CDK Pho85p (Carroll and O'Shea 2002; Huang et al. 2007; Lee et al. 2007; Chavel et al. 2014; Vandermeulen and Cullen 2022); and the general yeast stress response transcription factor Msn2p (Estruch and Carlson 1993; Martínez-Pastor et al. 1996; Gasch et al. 2000). Some pathways, such as the RTG, TOR, and Pho85p have been shown to exhibit conditional role reversals in S. cerevisiae in that they switch between positive and negative regulators depending on the environment (Vandermeulen and Cullen 2022).

Additionally, several general-purpose protein complexes also play a role in filamentous growth (Fig. 6a; Abdullah and Cullen 2009; Chavel et al. 2014; Chow, Starr et al. 2019; Vandermeulen and Cullen 2022). These factors include the chromatin-remodeling complex Rpd3p(L) (Kurdistani and Grunstein 2003); the elongator complex (ELP; Fellows et al. 2000; Krogan and Greenblatt 2001; Dong et al. 2015); and the chromatin-remodeling complex and transcription coactivator Spt-Ada-Gcn5 acetyltransferase (SAGA; Roberts and Winston 1997; Kohler et al. 2010; Hirsch et al. 2015). How these general-purpose factors function during filamentous growth to promote the response is not clear.

Complex networks like the one described above exist in many eukaryotes and appear to be the norm. The filamentation regulatory network in C. albicans is even more complicated than in S. cerevisiae and in some cases is wired differently. For example, in S. cerevisiae, glucose limitation is an inducer of filamentous/invasive growth through AMPK Snf1p (Cullen and Sprague 2000). Snf1p regulates filamentous growth by relieving the inhibitory effects of the transcriptional repressor Nrg1p (Kuchin et al. 2002), which is a negative regulator of glucose-repressed genes (Zhou and Winston 2001). In C. albicans, nutrient deprivation is also a trigger of filamentous growth (Huang 2012; Chow et al. 2021). However, in C. albicans, Snf1p is an essential protein (Petter et al. 1997). As in S. cerevisiae, Nrg1p also negatively regulates the filamentation response in C. albicans (Braun et al. 2001; Murad et al. 2001), although not by Snf1p. Rather, in C. albicans, Nrg1p is inhibited in a sequential manner by the RAS-PKA pathway and reduced TOR activity (Lu et al. 2011). Therefore, the RAS-PKA and TOR pathways are equally important for signaling and filamentation responses in C. albicans but in some ways work through different regulatory mechanisms (Shapiro et al. 2009; Inglis and Sherlock 2013; Huang et al. 2019; Qi et al. 2022). This is just 1 example showing how the same type of protein can be regulated by different upstream pathways in different organisms to achieve the same outcome. Another example can be seen when exploring how the RAS-PKA pathway controls pH-dependent responses in C. albicans (Hollomon et al. 2016), whereas pH-dependent responses are more exclusive to the Pal/Rim pathway in S. cerevisiae.

How do multiple pathways control a biological response in a coordinated manner? For filamentous growth, 1 way is that regulatory pathways coregulate shared target genes. For example, numerous transcription factors from different signaling pathways converge to regulate the same gene (e.g. FLO11; Rupp et al. 1999; Borneman et al. 2006, 2007). The fMAPK pathway, RAS-PKA pathway, SAGA complex, Pal/Rim pathway, RTG pathway, Opi1p transcriptional regulator, ELP complex, and chromatin-remodeling complex Rpd3(L) also regulate the expression of a subset of the same genes that induce filamentous growth (Chavel et al. 2014; Chow, Starr et al. 2019). Additionally, the RAS-PKA and TOR pathways regulate the same genes based on different temporal contexts after glucose induction (Kunkel et al. 2019; Plank 2022). Recently, studying the up-regulation of 2 different target genes that regulate cell adhesion (FLO11 or SFG1) has revealed insight into how an integrated signaling network can generate phenotypic plasticity based on the environment. This is achieved because the expression of FLO11 and SFG1 are both conditionally and differentially regulated by several different signaling pathways (e.g. RAS-PKA, Pal/Rim, Opi1p, fMAPK, Fig. 6b;Vandermeulen and Cullen 2022). The fact that numerous pathways can fine-tune the level of gene expression of filamentous targets allows for a highly nuanced and coordinated response.

A second way that signaling pathways control filamentous growth in a coordinated manner is through major transcription factors regulating each other's gene expression (Borneman et al. 2006). In many cases, this can create a type of cross feedback and has been shown to amplify and in some cases modulate transcription factor activity. Similarly, the transcription factor of 1 pathway has been found at the promoter of a gene encoding the components of another pathway, referred to as cross-pathway feedback. For example, the RAS-PKA pathway up-regulates MSB2 gene expression (Fig. 6c; Chow, Starr et al. 2019).

A third way is that pathways can regulate each other's activities (Mosch et al. 1996; Reinders et al. 1998; Chavel et al. 2010; Brückner et al. 2011). For example, the kinases of 1 pathway can regulate the activity and localization of the kinases in another (Bharucha et al. 2008). A classic example discovered in yeast is the fact that the RAS-PKA pathway can regulate fMAPK pathway activity by Ras2p (Mosch et al. 1996, 1999). It has since been shown that more than 5 pathways/protein complexes regulate the activity of the fMAPK pathway (Chavel et al. 2010). However, different pathways regulate fMAPK pathway activity depending on which stimuli are present and some even show opposing effects (i.e. role reversals) across environments (Fig. 6d; Vandermeulen and Cullen 2022, 2023). Therefore, a highly integrated network of protein complexes and signaling pathways regulate filamentous growth. This allows for the integration of multiple signals and coordination of cellular responses, including perhaps to modulate phenotypic plasticity.

In C. albicans, these combinatorial control mechanisms are evident during biofilm growth, which is regulated by a huge transcriptional network overseeing more than 1,000 genes (Nobile et al. 2012). These genes are coregulated by a core group of 9 transcription factors, the promoters of which are directly bound by most of the other regulators in a an intricate spiderweb of regulatory interactions that may allow integrated control by multiple environmental stimuli. Mammalian differentiation pathways similarly operate in vast uncharted networks. Addressing network-related questions in a simple model is relevant given the relatively small number of proteins and interactions compared to metazoans. Deciphering how signals become integrated into a cohesive response is critical to comprehensively understand the regulation of cell differentiation.

Conclusions and future directions

Much of the vast diversity in nature can be explained by the specialization of cells into differentiated cell types that perform highly specific functions. Progress toward understanding cell-type specialization has come in part from studying fungal model systems. In a handful of yeast models, genetic and molecular tools have been developed that have elucidated the molecular basis of the striking morphological diversity observed in nature. Remarkably, a few common pathways, such as MAPK pathways, control aspects of cell differentiation across many fungal species, and this logic extends to eukaryotes in general. Therefore, a satisfying feature of studying signaling pathways in fungal models is that molecular insights can be broadly applied across eukaryotes to understand how cell differentiation is regulated. Likewise, fungal-specific elements identified in these models provide potential road maps toward curbing fungal pathogenesis in plants and animals, including emerging fungal pathogens, such as C. auris (Yue et al. 2018; Egger et al. 2022).

Many questions remain to be answered surrounding MAPK pathway regulation during filamentous growth. It remains unclear what mucin sensors are sensing and how different mucins control different pathways. In addition, the way that mucins regulate the Cdc42p module is an open question. How the MAPK pathway operates in different modes (Ste20p dependent and Ste20p independent) in different contexts is not well understood. Equally fascinating are questions surrounding the broader network. How do signaling pathways regulate each other's activities? Is the network static, or does it change in different settings to modulate the filamentation response? Given that genetic screens have identified hundreds of proteins that impact filamentous growth, understanding the global picture of the response will be a major undertaking. New model systems, now approachable by genetic tools, such as clustered regularly interspaced short palindromic repeats or CRISPR may provide new insights into longstanding questions surrounding pathogenesis, symbiosis, and multispecies interactions. It will be interesting to “listen in” on the conversations that fungal cells have with their hosts and neighboring microbes at the molecular level. Future work aimed at addressing these important questions is highly related to human health from the perspective of fungal pathogenesis. Future work can also reveal fundamental insights into the regulation of eukaryotic cell differentiation.

Data availability

Genes and proteins are linked to Saccharomyces Genome Database (SGD) https://www.yeastgenome.org/.

Acknowledgments

Apologies to labs whose work was not cited due to page limitations. Thanks to laboratory members for reading the manuscript.

Funding

The work was supported by grants from the National Institutes of Health to P.J.C. (R01GM098629) and M.C.L. (R01AI143304).

Literature cited

Author notes