Integration of Self and Non-self Recognition Modulates Asexual Cell-to-Cell Communication in Neurospora crassa

Cells cooperate, compete, and are attacked in nature, driving the evolution of mechanisms for recognizing self versus non-self. Filamentous fungal cells cooperate to form an interconnected colony while competing with genetically dissimilar colonies...

Abstract

Cells rarely exist alone, which drives the evolution of diverse mechanisms for identifying and responding appropriately to the presence of other nearby cells. Filamentous fungi depend on somatic cell-to-cell communication and fusion for the development and maintenance of a multicellular, interconnected colony that is characteristic of this group of organisms. The filamentous fungus Neurospora crassa is a model for investigating the mechanisms of somatic cell-to-cell communication and fusion. N. crassa cells chemotropically grow toward genetically similar cells, which ultimately make physical contact and undergo cell fusion. Here, we describe the development of a Pprm1-luciferase reporter system that differentiates whether genes function upstream or downstream of a conserved MAP kinase (MAPK) signaling complex, by using a set of mutants required for communication and cell fusion. The vast majority of these mutants are deficient for self-fusion and for fusion when paired with wild-type cells. However, the Δham-11 mutant is unique in that it fails to undergo self-fusion, but chemotropic interactions and cell fusion are restored in Δham-11 + wild-type interactions. In genetically dissimilar cells, chemotropic interactions are regulated by genetic differences at doc-1 and doc-2, which regulate prefusion non-self recognition; cells with dissimilar doc-1 and doc-2 alleles show greatly reduced cell-fusion frequencies. Here, we show that HAM-11 functions in parallel with the DOC-1 and DOC-2 proteins to regulate the activity of the MAPK signaling complex. Together, our data support a model of integrated self and non-self recognition processes that modulate somatic cell-to-cell communication in N. crassa.

CELL-to-cell communication that mediates cell fusion between genetically similar cells is an important aspect of colony establishment and colony maintenance in filamentous fungi (Glass 2004; Fleissner et al. 2009; Richard et al. 2012; Bastiaans et al. 2015). Filamentous fungi that are unable to undergo cell fusion establish a colony slower than strains that are able to communicate and fuse (Richard et al. 2012; Simonin et al. 2012). Growth rate during the colony establishment phase is not correlated with the linear growth rate of a mature hyphal colony, as evidenced by the Neurospora crassa fusion mutant Δsoft, which has slowed growth during the colony establishment phase and a roughly wild-type-like growth rate after the colony is fully established (within 48 hr after inoculation) (Richard et al. 2012). In plant pathogenic Ascomycete species, communication and cell fusion are also important for establishing an infective network (Park et al. 2002; Tsuji et al. 2003; Cho et al. 2009; Rispail and Di Pietro 2010; Sarmiento-Villamil et al. 2018). In contrast, in the mutualistic endophyte Epichloë festucae, strains that lack the ability to communicate and fuse behave as a plant pathogen (Tanaka et al. 2008; Charlton et al. 2012).

The filamentous fungus N. crassa has emerged as a model organism for investigating mechanisms that mediate somatic cell-to-cell communication and cell fusion. Somatic cell fusion can occur between genetically identical germinated asexual spores (germlings) and between hyphae within a single colony. Germlings and hyphae frequently grow chemotropically toward other genetically identical cells, resulting in cell fusion and cytoplasmic mixing (Roca et al. 2005; Fleissner et al. 2009). Over 70 genes involved in mediating chemotropic growth (communication) and somatic cell fusion have been identified in N. crassa (Fu et al. 2011; Leeder et al. 2013; Palma-Guerrero et al. 2013; Dettmann et al. 2014; Fischer et al. 2018). Much of the work on communication and cell fusion in N. crassa and related fungi has focused on two conserved MAP kinase (MAPK) signal transduction pathways. The MAK-2 pathway is necessary for cell-to-cell communication and chemotropic interactions between cells undergoing cell fusion. Core components of the MAK-2 pathway form a protein complex associated with cell tips that dynamically assembles and disassembles at regular ∼8-min intervals during chemotropic growth. MAK-2 complex assembly/disassembly occurs perfectly out-of-phase with the dynamic assembly and disassembly of a second protein complex containing a protein called SOFT (Fleissner et al. 2009; Dettmann et al. 2014; Jonkers et al. 2014, 2016). SOFT functions as a scaffold protein for the MAK-1 Cell Wall Integrity (CWI) MAPK pathway; however, MAK-1 does not oscillate dynamically with SOFT during chemotropic interactions (Dettmann et al. 2013; Teichert et al. 2014; Weichert et al. 2016). The CWI pathway is necessary for communication, and components of the CWI pathway engage in phosphorylation-mediated cross talk with the MAK-2 pathway (Maerz et al. 2008; Dettmann et al. 2012; Maddi et al. 2012; Leeder et al. 2013; Fu et al. 2014; Teichert et al. 2014; Fischer et al. 2018). Both MAK-1 and MAK-2 pathways regulate gene expression via the transcription factors PP-1 and ADV-1, and also by directly phosphorylating several different proteins, a number of which are necessary for cell communication and fusion (Jonkers et al. 2014; Dekhang et al. 2017; Fischer et al. 2018).

The vast majority of cell-fusion mutants in N. crassa fail to initiate any chemotropic interactions or show oscillation of MAK-2 to fusion tips, either in interactions with themselves or when paired with wild-type cells (Fu et al. 2011; Leeder et al. 2013; Dettmann et al. 2014; Jonkers et al. 2014; Fischer et al. 2018). However, germlings of one mutant, Δham-11, fail to communicate and undergo self-fusion, but undergo chemotropic interactions and fusion with the parental wild-type cells at a frequency that is roughly equivalent to wild-type + wild-type germling fusion (Leeder et al. 2013). Furthermore, chemotropic interactions with a wild-type cell restore signaling in Δham-11 cells as evidenced by the dynamic oscillations of the MAK-2 and SOFT proteins in Δham-11 germling cell tips (Leeder et al. 2013).

In addition to regulating aspects of cell fusion between genetically identical cells, N. crassa also employs non-self recognition mechanisms to mediate cell-to-cell communication and cell fusion between genetically distinct cells. For example, a non-self recognition system regulated by the DOC proteins was identified using a population sample of N. crassa (Ellison et al. 2011; Heller et al. 2016); five distinct communication groups (CGs) are associated with five distinct doc-1/doc-2/doc-3 haplotypes in this population (Heller et al. 2016). Chemotropic interactions between wild isolates within a CG is frequent and equivalent to communication between genetically identical germlings, but communication between germlings from different CGs is either rare or not observed (Heller et al. 2016). The Fungal Genetics Stock Center (FGSC) 2489 strain (referred to as wild-type throughout) is a CG1 strain, which harbors a single copy of the doc-1 and doc-2 genes, but lacks the doc-3 gene that is present in CG2 and CG4 strains. Importantly, DOC-1-GFP oscillates to the cell-fusion tips coincidentally with the MAK-2 signaling complex, while DOC-2-GFP is restricted to the cell periphery (Heller et al. 2016). The DOC-1/DOC-2 proteins function to negatively regulate chemotropic interactions between genetically dissimilar cells, as evidenced by the phenotype of a Δdoc-1 Δdoc-2 strain, which has wild-type-like levels of self-communication, but shows greatly reduced chemotropic interactions and cell fusion with its isogenic wild-type parent (Heller et al. 2016). Further observations led to a model where DOC-1 and DOC-2 fail to reinforce signaling through the MAK-2 pathway during non-self interactions, thereby reducing cell-to-cell communication, chemotropic interactions, and cell fusion between cells containing alternative doc-1/doc-2 alleles (Heller et al. 2016).

In this study, we examine the germling communication regulatory network, with a focus on the interplay between self and non-self interactions, via the doc-1, doc-2, and ham-11 genes using a Pprm1-luciferase reporter construct. We demonstrate that Δham-11 germlings communicate with wild-type germlings, as previously reported (Leeder et al. 2013), in addition to Δdoc-1 Δdoc-2 germlings and other communication mutant germlings. We further characterize HAM-11 as a plasma membrane protein and show that the triple mutant Δdoc-1 Δdoc-2; Δham-11 has a synthetic cell-fusion phenotype. These data indicate that HAM-11 functions in parallel with DOC-1 and DOC-2 to integrate self and non-self recognition processes through MAK-2 MAPK signaling, thus modulating germling communication and cell fusion in N. crassa.

Materials and Methods

Protein prediction

To identify functional domains in the HAM-11 protein, we used the following protein-prediction software: pfam (Finn et al. 2016), XtalPred (Slabinski et al. 2007), SMART (Letunic and Bork 2018), Pred-TMR2 (Pasquier and Hamodrakas 1999), DAS (Cserzo et al. 1997), HMMTOP (Tusnady and Simon 2001), SCAMPI2 (Peters et al. 2016), TMMOD (Kahsay et al. 2005), Protter (Omasits et al. 2014), SPOCTOPUS (Viklund et al. 2008), TOPCONS (Tsirigos et al. 2015), Philius (Reynolds et al. 2008), PrediSi (http://www.predisi.de/), and Signal-BLAST (Frank and Sippl 2008).

Plasmid and strain construction

A list of strains used is provided in Supplemental Material, Table S1. The Pprm1-luciferase; his-3 plasmid was made by first amplifying the codon-optimized luciferase gene (Gooch et al. 2007) with primers to add PacI and EcoRI cut sites. This PCR product was then purified and ligated into the pMF272 backbone vector (Freitag et al. 2004). The ccg-1 promoter in pMF272 was replaced with the prm1 promoter by amplifying 1750 bp upstream of the prm1 ORF from genomic DNA with primers that added NotI and BamHI cut sites. This PCR product was cut and ligated into the pMF272 vector harboring luciferase. This construct was then transformed into the his-3⁻ strains FGSC 6103 and FGSC 9716 using standard electroporation (Colot et al. 2006). Prototrophic transformants were selected on Vogel’s minimal medium (VMM) (Vogel 1956) without histidine. To avoid off-target effects from the transformation, transformants were backcrossed to either FGSC 6103 or FGSC 9716, and progeny were selected for prototrophy and then screened for luciferase expression.

Deletion mutants containing a his-3 mutation were generated by crossing each deletion mutant (Table S1) with either FGSC 6103 or FGSC 9716. Progeny from these crosses were selected for hygromycin B resistance and screened for histidine auxotrophy. Gene deletions were confirmed via PCR. Each his-3⁻ deletion mutant was then crossed with the wild-type strain (Pprm1-luciferase), and progeny carrying both the deletion and Pprm1-luciferase at the his-3 locus were selected on VMM with hygromycin B, and confirmed for luciferase luminescence (Table S1).

The Ptef1-ham-11-v5 plasmid was made by amplifying the ham-11 gene from genomic DNA with primers that added XbaI and PacI cut sites. This PCR product was ligated into a pCR-Blunt vector (Invitrogen, Carlsbad, CA). We then used the XbaI and PacI sites to replace adv-1 with ham-11 in the Ptef1-adv-1-v5; his-3 vector (Fischer et al. 2018). This vector was then transformed into Δham-11; his-3 mat A conidia via electroporation, and prototrophic transformants were selected on VMM with hygromycin B. To avoid off-target effects of electroporation, the Δham-11 (Ptef1-ham-11-v5) strain was backcrossed to wild-type mat a (FGSC 9716), Δham-11 (Ptef1-ham-11-v5) progeny were selected on VMM with hygromycin B, and genotypes at the ham-11 and his-3 loci were confirmed via PCR.

The Δdoc-1 Δdoc-2; Δham-11 triple mutant was made by first crossing Δham-11 mat A (FGSC 17545) with wild-type mat a (FGSC 4200) to obtain a Δham-11 mat a strain. The Δham-11 mat a strain was crossed with Δdoc-1 Δdoc-2 mat A (Heller et al. 2016). Several progeny were isolated, and their genotypes at the doc-1, doc-2, and ham-11 loci were confirmed via PCR. We successfully obtained six progeny with the Δdoc-1 Δdoc-2; Δham-11 genotype, and all six strains had equivalent flat growth and phenotypes on VMM slant tubes. The Δdoc-1; Δham-11 double mutant was made by crossing Δham-11 mat a (Leeder et al. 2013) with Δdoc-1 mat A (Heller et al. 2016), and progeny were screened via PCR. We successfully obtained two progeny that had the Δdoc-1; Δham-11 genotype. Both progeny had equivalent growth phenotypes on VMM agar. The Δdoc-2; Δham-11 double mutant was made by crossing Δham-11 mat a (Leeder et al. 2013) with Δdoc-2 mat A (FGSC 14644), and progeny were screened via PCR. We obtained three progeny that had the Δdoc-2; Δham-11 genotype. All three progeny had equivalent growth phenotypes on VMM agar.

Strains expressing 6xflag-nrc-1^P451S from the his-3 locus were made by crossing the wild-type strain (Pccg1-6xflag-nrc-1^P451S; his-3) (Dettmann et al. 2012) with the following his-3⁻ mutant strains; Δham-11; his-3 (Leeder et al. 2013), Δadv-1; his-3 (Fischer et al. 2018), and Δste-20; his-3 (this study). Progeny were selected for prototrophy and hygromycin resistance on VMM containing hygromycin, which resulted in the following strains; Δham-11 (Pccg1-6xflag-nrc-1^P451S), Δste-20 (Pccg1-6xflag-nrc-1^P451S), and Δadv-1 (Pccg1-6xflag-nrc-1^P451S). Gene deletions and successful insertion of the 6xflag-nrc-1^P451S construct at the his-3 locus was confirmed by PCR.

Sucrose gradient cell fractionation and western blotting

We used a sucrose gradient cell fractionation protocol adapted from previously described methods (Kaiser et al. 2002; Palma-Guerrero et al. 2014). Briefly, Δham-11 (Ptef1-ham-11-v5) conidia were inoculated at a concentration of 10⁶ conidia/ml in 100 ml of liquid VMM in a 250 ml flask. Flasks were incubated at 30° while shaking at 220 rpm for 2.5 hr to induce germination, then shaking was ceased, and incubation at 30° Continued for 2.5 hr to allow for communication and fusion to occur. Germlings were harvested by vacuum filtration and immediately frozen with liquid nitrogen. Cells were lysed by bead beating with 500 μl STE10 [10% sucrose (w/w), 10 mM Tris-HCl pH 7.5, and 10 mM EDTA] and complete protease inhibitors (Protease Inhibitor Cocktail Set IV, Calbiochem, San Diego, CA). Unlysed cells and cell walls were removed by centrifugation. Next, 300 μl of cell lysate was floated on top of a 5-ml 20–60%(w/w) sucrose gradient and centrifuged at 100,000 × g at 4° for 18 hr. Ten 500-μl fractions were collected from the top of the gradient and the last fraction (#10) contained the pellet. Protein from each fraction was purified using a previously described method (Pandey et al. 2004) and run on a 4–12% Bis-Tris NuPAGE gel (Invitrogen). Proteins were transferred to a PVDF membrane via western blotting and blots were probed with α-PMA-1 (ab4645), α-V5 (R96025; Invitrogen), α-ERV-25 (Starr et al. 2018), and α-actin antibodies (MAB1501, Millipore, Bedford, MA).

Luciferase assay

Conidial suspensions were diluted to 6 × 10⁶ spores/ml in 10 ml of liquid VMM in a 15-ml Falcon tube (Falcon, Lincoln Park, NJ). Tubes were shaken at 220 rpm at 30° for 2.5 hr to allow for germination. Conidial suspensions were then diluted 1:1 with 2× d-Luciferin (VWR International) in VMM, and 200 μl of this conidial-Luciferin solution was added to each well in a black 96-well plate with a clear bottom and lid (catalog number 3603, Corning). Final concentrations were 3 × 10⁶ spores/ml and 100 μM Luciferin. Plates were then wrapped in foil to avoid Luciferin degradation, and incubated at 30° for an additional 2 hr to allow for the induction of communication and fusion. Luciferase luminescence was quantified with a 2014 Perkin-Elmer ([Perkin Elmer-Cetus], Norwalk, CT) Envision Multilabel Plate Reader at University of California (UC), Berkeley’s qb3 High-Throughput Screening Facility.

Quantitative RT-PCR

RNA was extracted from germlings as previously described (Fischer et al. 2018). Quantitative RT-PCR reactions were prepared following the manufacturers’ guidelines for the Bioline SensiFast SYBR no-ROX One-Step kit and Bio-Rad (Hercules, CA) CFX Connect Real-Time PCR Detection System. Each sample was replicated five times within a 96-well plate and total reaction volume was 20 μl. Expression data were normalized to actin following the 2^−ΔCt method (Livak and Schmittgen 2001).

Quantification of germling fusion via flow cytometry

The protocol for germling fusion quantification via flow cytometry was previously described (Heller et al. 2018). In brief, we crossed a wild-type strain carrying an alternative sec-9 allele (sec-9^GRD2) at the native sec-9 locus with each of the following strains: Δham-11; his-3, Δdoc-1; his-3, Δdoc-2; his-3, and Δdoc-1 Δdoc-2; his-3. When a germling with this alternative sec-9^GRD2 allele fuses with a germling harboring the FGSC 2489 sec-9 allele (sec-9^GRD1), cell death occurs within 20 min; sec-9-mediated death frequency is a proxy for cell-fusion frequency (Heller et al. 2018). Next, 10⁷ conidia/ml were grown on 20% Pluronic F-127 (Sigma [Sigma Chemical], St. Louis, MO) VMM plates. After 4.5 hr of cultivation at 30°, plates were moved to −20° for 5 min to liquefy the medium. Germlings were harvested and washed by centrifugation, and then suspended in 1 ml of PBS containing the vital dye 0.1 μM SYTOX Blue (Life Technologies) or the vital dye 0.15 μM propidium iodide (Sigma), and analyzed on a BD LSRFortessa X-20 (BD Biosciences) at UC Berkeley’s Flow Cytometry Core Facility. Two vital dyes were used as a technical control: SYTOX Blue fluorescence was detected with a nondichroic 450/50 filter after excitation using a 405-nm laser and propidium iodide fluorescence was detected with a 685 Long Pass 710/50 filter after excitation using a 488-nm laser. For each sample, 20,000 events were recorded. Ungerminated conidia were used as a negative control in each experiment. These data informed the computational exclusion of conidia from experimental (germinated) samples. Data were analyzed using the Cytobank Community software (community.cytobank.org). Cytobank software outputs the Total Death Rate for each sample. Normalized Fusion Frequency was calculated by subtracting the average Basal Death Rate (Figure S1) from the Total Death Rate for each strain. Results for SYTOX Blue and propidium iodide did not differ significantly [Figure S1 (Heller et al. 2018)].

Data availability

Strains and plasmids are available upon request. One supplemental table and seven supplemental figure files are available at Figshare: https://figshare.com/s/74bd456f8c5c726f3a41.

Results

Using a prm1 luciferase construct to investigate cell-fusion regulatory interactions

The prm1 (PRM1-like) gene in N. crassa is involved in cell membrane merging during cell fusion. Δprm1 mutants show a ∼50% decrease in cell fusion of genetically identical cells, with many cells blocked at the plasma membrane merger stage (Fleissner et al. 2008). Furthermore, prm1 expression is dependent on two transcription factors required for cell fusion, PP-1 and ADV-1, and catalytically active MAK-2 (Leeder et al. 2013; Dekhang et al. 2017; Fischer et al. 2018). We adapted a codon-optimized luciferase reporter (Gooch et al. 2007) to assess activation of prm1 in cell-fusion mutants. Optimal detection of Pprm1-luciferase luminescence in wild-type germlings (2.5-hr after germination) was dependent on the initial concentration of conidia (10⁶ spores/ml) (Figure S2). Mutants that were completely defective in communication and cell fusion also had very low-level Pprm1-luciferase expression (Figure 1). These mutants included components of the MAK-1 or MAK-2 signal transduction complex (mak-2, mek-2, nrc-1, ham-5, soft, and mik-1), communication-activated transcription factors (adv-1 and pp-1), as well as components of the striatin-interacting protein phosphatase and kinase (STRIPAK) complex (ppg-1, ham-2, ham-3, and ham-4). Mutants with higher germling fusion frequency than wild-type cells (Δgyp-5, Δspr-7, and Δnik-2) did not show hyper-activation of Pprm1-luciferase. There were four exceptions to the observation that low Pprm1-luciferase expression was correlated with defective germling communication and fusion. The Δamph-1, Δmob-3, Δbem-1, and Δste-20 mutants are all capable of germling fusion at a reduced frequency as compared to wild-type germlings (Schürg et al. 2012; Dettmann et al. 2014; Fu et al. 2014), but all of these mutants had low Pprm1-luciferase expression (Figure 1). In N. crassa, the amph-1, bem-1, and ste-20 genes are necessary for full activation of the MAK-2 pathway, which in turn is necessary for expression of prm1 (Schürg et al. 2012; Dettmann et al. 2014); Fu et al. 2014. The MOB-3 protein is a scaffold for the STRIPAK complex, which facilitates cross talk between the MAK-1 and MAK-2 pathways (Dettmann et al. 2013, 2014; Kabi and McDonald 2014; Fischer et al. 2018).

Figure 1.

Pprm1-luciferase expression in germling communication mutants identifies genes that function upstream or downstream of prm1. Summary plot depicting 42 independent Pprm1-luciferase experiments (one experiment = one 96-well plate). Each experiment contained a balanced design of 7–11 samples per strain and each dot represents a single sample, normalized to wild-type. Raw photon cps for each experiment were analyzed by ANOVA with Tukey’s HSD, cps values for each strain were normalized to the mean wild-type value within each experiment, and these normalized values were pooled to build this plot. Wild-type, Δham-11, and Δmak-2 cells were included in every experiment as controls for each ANOVA group. Strains in the low group (green) were significantly different from Δham-11 and wild-type cells. The medium group (orange) was significantly different from Δmak-2 and wild-type cells, and strains in the high group (purple) were significantly different from Δmak-2 and Δham-11 cells (P < 0.001, ANOVA with Tukey’s HSD, n = 7–11). The Δdoc-2 and Δgyp-5 strains were inconsistent between experiments, which is reflected by their classification into two different groups based on the within-experiment ANOVA tests. Black bars indicate the mean of the pooled samples. The self-fusion phenotype of each mutant (based on previously published data) is indicated by a black (0%), gray (1–80%), or white (> 95%) box above each mutant name. Wild-type germling self-fusion frequency = 87 ± 2%. Bolded strains are a focus of the main text. cps, counts per second; HSD, Honest Significant Difference.

The Δham-11, Δham-14, Δras-2, Δgyp-5, Δdoc-1, and Δdoc-2 mutants each showed a medium level of Pprm1-luciferase expression (Figure 1). The ham-11 gene is currently annotated as a hypothetical protein that is a transcriptional target of PP-1 and ADV-1 (Leeder et al. 2013; Fischer et al. 2018). The ham-14 gene encodes a hypothetical protein that is phosphorylated in a MAK-2-dependent manner and is required for the assembly of MAK-2 protein complexes (Jonkers et al. 2014, 2016). The ras-2 gene encodes a conserved GTPase that physically interacts with STE-50 and NRC-1, and is required for full phosphorylation of MAK-2 (Dettmann et al. 2014). The Δgyp-5 mutant is a hyper-fusion mutant, and the gyp-5 gene encodes a GTPase-activating protein orthologous to Gyp5p in Saccharomyces cerevisiae, which physically interacts with amphiphysins to recruit them to sites of endocytosis and exocytosis in S. cerevisiae (Prigent et al. 2011; Palma-Guerrero et al. 2013). Lastly, the doc-1 and doc-2 genes encode proteins that are required to repress communication between cells with genetic differences at doc-1 and doc-2 (Heller et al. 2016).

The Δham-11 mutant undergoes chemotropic interactions with other communication mutants

The Δham-11 mutant has a unique intermediate phenotype compared to other fusion mutants because it has a wild-type-like macroscopic growth phenotype with abundant aerial hyphae, but Δham-11 germlings to do not engage in self chemotropic interactions or fusion, except when co-cultured with wild-type cells (Leeder et al. 2013). This observation indicates that a germling expressing HAM-11 is sufficient to initiate chemotropic interactions with a germling that is lacking HAM-11. The ham-11 gene is annotated as a hypothetical protein with no predicted functional domains, except for one or two predicted N-terminal transmembrane domains and a long cytosolic tail (Figure 2A). To determine the cellular localization of HAM-11, we performed cell fractionation on a fully complemented Δham-11 (Ptef1-ham-11-v5) strain (Figure 2, B and C). HAM-11-V5 was enriched in a subset of the fractions containing the plasma membrane ATPase PMA-1. The distribution of HAM-11-V5 in the gradient showed minimal overlap with the fractions containing the ER/Golgi protein ERV-25 or the cytoplasmic cytoskeletal protein actin (Figure 2C).

Figure 2.

Δham-11 germlings communicate with each other when co-cultured with wild-type cells. (A) Cartoon of the HAM-11 protein. The most N-terminal transmembrane domain (striped) is predicted to be either a signal peptide or a transmembrane domain. Phosphorylation sites in HAM-11 as identified in Jonkers et al. (2014). (B) The germling self-fusion phenotype of Δham-11 (Ptef1-ham-11-v5), wild-type, and Δham-11 cells. Black arrows indicate fusion points. (C) Western blots of protein extracted from Δham-11 (Ptef1-ham-11-v5) germlings subjected to sucrose gradient (0–60% sucrose) centrifugation. Input lane contains whole-cell lysate. Fractions were probed with either anti-plasma membrane ATPase antibodies (α-PMA-1), anti-ERV-25 antibodies, a ER/Golgi-localized protein (α-ERV-25), anti-actin antibodies (α-actin), a cytoplasmic protein, or anti-V5 antibodies, which detect HAM-11-V5 protein. (D) Δham-11 (Pccg1-soft-gfp) germlings co-cultured at a 1:1 ratio with FM4-64-stained wild-type germlings. Top panel shows the Δham-11 mutant fusing with wild-type, while the bottom panel shows two Δham-11 germlings fusing near a wild-type germling. Both sets of images were taken from the same ∼1-cm² agar slab. (E) Δham-11 (Pccg1-gfp) co-cultured at a 1:1 ratio with FM4-64-stained wild-type germlings. Top panel shows a Δham-11 (Pccg1-gfp) germling fusing with a wild-type germling, while the bottom panel shows two fused Δham-11 (Pccg1-gfp) germlings. Both sets of images were taken from the same ∼1-cm² agar slab. (F) Wild-type (Pccg1-gfp) co-cultured at a 1:1 ratio with FM4-64-stained Δham-11 germlings. Top panel shows a wild-type (Pccg1-gfp) germling fusing with a Δham-11 germling, while the bottom panel shows two fused Δham-11 germlings. Both sets of images were taken from the same ∼1-cm² agar slab. All germling communication assays started with a concentration of 10⁷ cells/ml. For microscopy images, arrows indicate communication and fusion events. Bars, 5 μm.

As previously reported, we observed Δham-11 + wild-type chemotropic interactions and cell fusion (Figure 2D). Surprisingly, we also observed Δham-11 + Δham-11 cell fusion in instances where Δham-11 cells were co-cultured at a 1:1 ratio with wild-type cells (Figure 2D). In contrast, chemotropic growth or cell fusion between Δham-11 cells in the absence of wild-type cells was not detectable (Figure 2B). Germling communication and cell fusion between pairs of Δham-11 germlings in co-culture with wild-type cells was independent of fluorescent constructs used to visually identify each strain [co-cultures of wild-type + Δham-11 (Pccg1-soft-gfp), wild-type + Δham-11 (Pccg1-gfp) or wild-type (Pccg1-gfp) + Δham-11; Figure 2, D–F]. Cell fusion events between Δham-11 + Δham-11 cells were also frequent and easily observable when Δham-11 was co-cultured with the hyper-fusion mutant Δspr-7, the Δcse-1 mutant, which shows reduced fusion frequency (Palma-Guerrero et al. 2013), and Δprm1 or Δlfd-1 germlings, which are affected in membrane merger (Fleissner et al. 2008; Palma-Guerrero et al. 2014) (Figure S3). We made varied attempts to induce communication in Δham-11 cells, and to activate Pprm1-luciferase in Δham-11 (Pprm1-luciferase) monocultures, by adding conditioned media from wild-type or Δspr-7 cultures, but observed no effect from the addition of conditioned media on the growth, behavior, or luciferase activity of Δham-11 germlings.

In addition to Δham11 + wild-type cell fusion, we observed Δham-11 (Pccg1-soft-gfp) germlings communicating and fusing with germlings of strains that were self-communication competent, but showed altered fusion frequencies or behavior as compared to wild-type cells; Δdoc-1 Δdoc-2, Δdoc-1, Δdoc-2, Δste-20, Δgyp-5, Δmob-3, Δnik-2, Δarg-15, and Δsec-22 (Figure 3A and Figure S4A). Chemotropic interactions or cell fusion were not observed in co-cultures of Δham-11 (Pccg1-soft-gfp), or in mutants that are completely defective in germling communication and fusion (Figure S4B), including Δpp-1, Δmak-2, Δadv-1, Δmak-1, and Δham-14 (Figure 3B). However, we observed Δham-11 (Pccg1-soft-gfp) germlings chemotropically growing toward Δsoft germlings, although Δsoft germlings never responded chemotropically to Δham-11 cells; cell fusion never occurred in these interactions (Figure 3A, top panel). In contrast, wild-type (Pccg1-h1-gfp) and wild-type (Pccg1-mak-2-gfp) germlings ignored Δsoft germlings (Fleissner et al. 2005) (Figure S5).

Figure 3.

Δham-11 germlings chemotropically interact with germlings of other communication mutants. Δham-11 (Pccg1-soft-gfp) co-cultured at a 1:1 ratio with each communication mutant (Δsoft, Δdoc-1 Δdoc-2, Δdoc-1, Δdoc-2, Δprm1, Δste-20, Δgyp-5, Δpp-1, Δmak-2, Δadv-1, Δmak-1, and Δham-14), which were stained with FM4-64. (A) Fluorescent images of co-cultures of mutants paired with Δham-11 (Pccg1-so-gfp) that engaged in chemotropic interactions. (B) Fluorescent images of co-cultures of mutants paired with Δham-11 (Pccg1-so-gfp) where chemotropic interactions and communication did not occur. Bar, 5 μm.

The Δham-11 mutant maintains CG specificity

Surprisingly, Δham-11 germlings communicated with otherwise isogenic Δdoc-1 Δdoc-2 germlings (Figure 2 and Figure 3), but Δdoc-1 Δdoc-2 germlings showed very low levels of communication and fusion with otherwise isogenic wild-type germlings [Figure 4 and Heller et al. (2016)]. We quantified germling fusion frequency for all pairwise interactions between wild-type, Δham-11, Δdoc-1, Δdoc-2, and Δdoc-1 Δdoc-2 germlings by adapting a flow cytometry method for quantifying germling fusion based on a robust postfusion death response mediated by genetic differences at sec-9 (Heller et al. 2018) (Figure 4). For this assay, Δham-11, Δdoc-1, Δdoc-2, and Δdoc-1 Δdoc-2 strains were constructed containing incompatible sec-9 alleles.

Figure 4.

Pairwise fusion frequency between Δham-11, Δdoc-1, Δdoc-2, and Δdoc-1 Δdoc-2 mutants and wild-type cells. Germling fusion frequency quantified via flow cytometry using a cell death assay that is activated upon cell fusion. Strains were combined at a 1:1 ratio and a total of 20,000 events were counted per co-culture. Data corresponding to conidia were removed from the data set and germling fusion frequency was determined by the proportion of cells that were stained with propidium iodide (germlings that fused and died) vs. cells that were not stained with propidium iodide (non-fused cells or fusion between isogenic germlings). Letters indicate significantly different levels of fusion; for example, “a” values are significantly different from “b” values and “c” values are significantly different from both “a“ and “b” values. “ab” values are not significantly different from either “a” or “b” values, and “bc” values are not significantly different from either “b” or “c” values (* P < 0.005, ANOVA with Tukey’s Honest Significant Difference, n = 6). Error bars indicate SD.

The flow cytometry data recapitulated previous microscopy-based quantification of the self-communication frequency of each strain (Leeder et al. 2013; Heller et al. 2016). For example, the self-fusion frequency of Δdoc-1 Δdoc-2 + Δdoc-1 Δdoc-2 cells was similar to wild-type + wild-type cells, but fusion frequency between Δdoc-1 Δdoc-2 + wild-type parental cells was very low (Figure 4). The frequency of fusion between Δham-11 + wild-type cells was not significantly different from wild-type + wild-type cell-fusion frequencies [P = 0.033, n = 6, ANOVA with Tukey’s Honest Significant Difference (HSD) post hoc test, Figure 4]. However, the frequency of Δham-11 + Δdoc-1 Δdoc-2 cell-fusion events was significantly lower than wild-type self-fusions (P = 0.000032, n = 6, ANOVA with Tukey’s HSD, Figure 4), but was still significantly higher than wild-type + Δdoc-1 Δdoc-2 fusion events (P = 0.001, n = 6, ANOVA with Tukey’s HSD, Figure 4). The Δdoc-1 germlings and Δdoc-2 germlings showed greatly reduced fusion frequencies with Δdoc-1 Δdoc-2 germlings, but only Δdoc-2 germlings fused with Δham-11 germlings at a significantly higher frequency (P = 2.11 × 10⁻¹¹, ANOVA with Tukey’s HSD, Figure 4).

Since Δham-11 germlings fused with both wild-type and Δdoc-1 Δdoc-2 germlings, but wild-type germlings showed greatly reduced fusion frequency with Δdoc-1 Δdoc-2 germlings (Figure 4), we reasoned that the Δham-11 mutant might be impaired in its ability to distinguish between CGs. To test this hypothesis, we co-cultured Δham-11 (Pccg1-soft-gfp) germlings with N. crassa strains representative of CGs CG1, CG2, CG3, and CG5 (Figure 5). Strains in CG4 did not have a discrete communication phenotype and were therefore eliminated (Heller et al. 2016). In contrast to our prediction, we did not observe any communication between Δham-11 (Pccg1-soft-gfp) germlings and germlings from CG2, CG3, or CG5 (Figure 5). Thus, the Δham-11 strains still retain the capacity to distinguish CG, with the exception of the Δdoc-1 Δdoc-2 germlings. Our data also suggests that the Δdoc-1 Δdoc-2 mutant represents a unique CG, as the Δham-11 mutant was able to distinguish between CG5 and Δdoc-1 Δdoc-2 genotypes; the Δdoc-1 Δdoc-2 mutant had previously been placed in CG5 (Heller et al. 2016).

Figure 5.

Δham-11 germlings maintain communication group (CG) specificity. Δham-11 (Pccg1-soft-gfp) germlings were co-cultured with FM4-64-stained representative strains from each communication group. Designation of strains within each CG was defined in Heller et al. (2016). (A) Communication and cell fusion between a Δham-11 (Pccg1-soft-gfp) germling and a Fungal Genetics Stock Center 2489 (CG1) germling. (B) Lack of communication and cell fusion between a ham-11 (Pccg1-soft-gfp) germling and a JW258 (CG2) germling. (C) Lack of communication and cell fusion between a Δham-11 (Pccg1-soft-gfp) germling and a P4471 (CG3) germling. (D) Lack of communication and cell fusion between a Δham-11 (Pccg1-soft-gfp) germling and a JW220 (CG5) germling. (E) Communication and cell fusion between a Δham-11 (Pccg1-soft-gfp) germling and a Δdoc-1 Δdoc-2 germling. Bar, 5 μm.

The triple mutant Δdoc-1 Δdoc-2; Δham-11 has a synthetic phenotype

The data above indicated that ham-11 and doc-1 doc-2 have a genetic interaction that regulates cell-fusion identity. Therefore, we hypothesized that a synthetic phenotype with respect to fusion or communication may occur in a Δdoc-1 Δdoc-2; Δham-11 triple mutant; we constructed double and triple ham-11 mutants to test this hypothesis. The Δdoc-1 Δdoc-2; Δham-11 triple mutant, the Δdoc-1; Δham-11 double mutant, the Δdoc-2; Δham-11 double mutant, and the parental Δham-11 mutant were all defective in germling self-fusion (Figure 2B and Figure 6A). Hyphal fusion was also lacking in the Δdoc-1 Δdoc-2; Δham-11 triple mutant, the Δdoc-1; Δham-11 double mutant, and the Δdoc-2; Δham-11 double mutant, in contrast to observed hyphal fusion events in both the Δham-11 and Δdoc-1 Δdoc-2 parental strains [Figure S6 and Leeder et al. (2013)]. All parental strains and double mutants produced aerial hyphae equivalent to the wild-type parental strain, but only the triple mutant Δdoc-1 Δdoc-2; Δham-11 produced significantly shorter aerial hyphae (P < 0.0001, ANOVA with Tukey’s HSD, n = 6, Figure S6), a phenotype associated with strains that show a complete absence of cell fusion (Fu et al. 2011).

Figure 6.

The Δdoc-1 Δdoc-2; Δham-11 mutant has a synthetic germling fusion phenotype. (A) Germling fusion phenotype of Δdoc-1 Δdoc-2; Δham-11, Δdoc-1; Δham-11, Δdoc-2; Δham-11, and the parental Δdoc-1 Δdoc-2 strain. (B) Germling fusion phenotype of Δdoc-1 Δdoc-2; Δham-11 (“ΔΔΔ”), Δdoc-1; Δham-11 (“Δd1Δh11”), and Δdoc-2; Δham-11 (“Δd2Δh11”) grown in co-culture with either Δdoc-1 Δdoc-2 (Pccg1-gfp) (“Δd1Δd2”) or wild-type (Pccg1-gfp) (“wild-type”) germlings. Δdoc-1 Δdoc-2; Δham-11, Δdoc-1; Δham-11, and Δdoc-2; Δham-11 conidia were stained with FM4-64 prior to mixing with parental wild-type GFP-expressing strains. Arrows indicate chemotropic interactions and fusion. Bar, 5 μm.

Since the Δdoc-1 Δdoc-2; Δham-11 germlings did not engage in self-communication, we hypothesized that the Δdoc-1 Δdoc-2; Δham-11 germlings would communicate with Δdoc-1 Δdoc-2 germlings. However, communication or cell fusion between Δdoc-1 Δdoc-2; Δham-11 and Δdoc-1 Δdoc-2 germlings was not observed (Figure 6B). Communication or fusion was also not observed between Δdoc-1 Δdoc-2; Δham-11 and wild-type germlings. In contrast, the cell-fusion phenotype of the double mutants mirrored that of the parental strains: Δdoc-1; Δham-11 germlings communicated and fused with wild-type germlings, but not with Δdoc-1 Δdoc-2 germlings, while the Δdoc-2; Δham-11 germlings communicated and fused with both wild-type and Δdoc-1 Δdoc-2 germlings (Figure 6B). These data showed that the Δdoc-1 Δdoc-2; Δham-11 mutant was completely defective at somatic cell-to-cell communication and fusion, in contrast to the fusion-competent phenotype of its parental strains Δdoc-1 Δdoc-2 and Δham-11. Like other cell-fusion-defective mutants in N. crassa, the Δdoc-1 Δdoc-2; Δham-11 mutant was also female sterile but male fertile (Fu et al. 2011) (Figure S7).

HAM-11 functions upstream of the MAK-2 pathway

To gain a better understanding of how the HAM-11, DOC-1, and DOC-2 proteins modulate communication-related gene expression, we crossed the Pprm1-luciferase construct into the Δdoc-1 Δdoc-2; Δham-11, the Δdoc-1; Δham-11, and the Δdoc-2; Δham-11 mutants (Figure 7A). The Δdoc-1; Δham-11 and the Δdoc-2; Δham-11 double mutants showed intermediate levels of Pprm1-luciferase output, similar to the Δham-11 mutant, but the triple Δdoc-1 Δdoc-2; Δham-11 mutant abolished expression of Pprm1-luciferase (Figure 7A), consistent with its lack of chemotropic interactions and cell fusion.

Figure 7.

ham-11, doc-1, and doc-2 function upstream of the MAK-2 pathway. (A) Germlings of wild-type or each mutant expressing Pprm1-luciferase were grown from a starting concentration of 10⁶ spores/ml in liquid VMM with Luciferin at 30° in the dark. Luciferase luminescence was quantified after 5 hr of growth. Letters indicate significantly different groups (P < 0.001, ANOVA with Tukey’s HSD, n = 24). Error bars indicate SD. (B) Quantitative RT-PCR quantification of the levels of prm1 mRNA in each mutant and wild-type germlings. Expression of prm1 was normalized to the level of actin mRNA in each strain. nrc-1P451S is a constitutively active allele of nrc-1, which encodes the MAPK kinase directly upstream of MAK-2. The STE-20 protein is a well-characterized conserved activator directly upstream of NRC-1, thus ste-20 (Pccg1-3xflag-nrc-1^P451S) is a positive control for upstream mutants. Error bars indicate SD (*** P < 10^-26, Welch’s t-test, n = 5). cps, counts per second; HSD, Honest Significant Difference; VMM, Vogel’s minimal medium.

Since prm1 transcription is dependent on signaling through the MAK-2 pathway (Fu et al. 2011), we hypothesized that the HAM-11, DOC-1, and DOC-2 proteins may function upstream of the MAK-2 pathway. The DOC-1 and DOC-2 proteins were previously described as being part/upstream of the MAK-2 signaling complex (Leeder et al. 2013; Fischer et al. 2018). To determine whether the HAM-11 protein functions upstream of the MAK-2 pathway, we measured levels of prm1 transcription in strains containing a constitutively active nrc-1^P451S allele (Dettmann et al. 2012) (Figure 7B). The nrc-1 gene encodes the MAPKKK (MAPK kinase kinase) that is part of the MAK-2 signaling complex and functions upstream of the MAK-2 protein; strains containing the nrc-1^P451S allele constitutively activate signaling through the MAK-2 cascade (Dettmann et al. 2012). Δste-20 germlings served as a control for genes that function upstream of NRC-1, and Δadv-1 germlings controlled for genes that function downstream of MAK-2. As shown with the luciferase data (Figure 1), the Δham-11, Δste-20, and Δadv-1 germlings all had reduced levels of prm1 transcription as compared to wild-type cells (Figure 7B). In an nrc-1^P451S background, prm-1 transcription was significantly increased in the Δste-20 germlings, as predicted, but also in the Δham-11 germlings, consistent with the hypothesis that HAM-11 functions upstream of the MAK-2 signaling complex (P < 10⁻⁶, Welch’s t-test, n = 5, Figure 7B). In contrast, transcription of prm1 was similar in the Δadv-1 and Δadv-1 (Pccg1-flag-nrc-1^P451S) germlings (P = 0.01, Welch’s t-test, n = 5, Figure 7B).

Discussion

An important aspect of development and survival is the ability for cells to accurately recognize and respond to other nearby cells based on their genetic identity. Most of the cell-to-cell communication research in filamentous fungi has focused on elucidating mechanisms that result in cell fusion between genetically identical cells. Among the > 70 identified cell-fusion mutants, the hypothetical signal(s) and receptor(s) involved in chemotropic interactions and cell fusion have not yet been described (Fu et al. 2011; Leeder et al. 2013; Palma-Guerrero et al. 2013; Fischer et al. 2018). In an effort to evaluate how the >70 cell-fusion genes fit into a regulatory network, we crossed a Pprm1-luciferase reporter into 41 strains that showed altered cell-fusion frequencies compared to wild-type. The prm1 gene is a direct target of the cell-fusion transcription factors PP-1 and ADV-1 (Dekhang et al. 2017; Fischer et al. 2018), and the PRM1 protein functions at the membrane merger stage of cell fusion (Fleissner et al. 2008). Our data recapitulate the cell-fusion phenotype of most of the 41 mutants, with a few exceptions. Hyper-fusion mutants (spr-7, gyp-5, and nik-2; Palma-Guerrero et al. 2013) did not show a higher level of Pprm1-luciferase activity. Similarly, Δham-12, a mutant with decreased fusion frequency (Leeder et al. 2013), did not show decreased Pprm1-luciferase activity. Lastly, Δham-11 and Δham-14 (Leeder et al. 2013; Jonkers et al. 2016) fail to undergo self-germling fusion, but showed a moderate level of Pprm1-luciferase activity when cultured alone.

The unique communication phenotype of the Δham-11 mutant allowed us to explore novel aspects of germling communication. In contrast to quorum sensing in bacteria (Papenfort and Bassler 2016), germling communication is not initiated uniformly in a switch-like manner across a spatial population of germlings. We hypothesize that N. crassa cells secrete a competency factor that initiates signaling when cells are in close proximity, and subsequently produce a second factor involved in mediating chemotropic interactions. This second factor could be a different molecule from the competency factor or it could be the same molecule with distinct spatiotemporal dynamics. The MAK-2 and SOFT proteins display remarkably specific oscillatory dynamics that are necessary for chemotropic interactions (Fleissner et al. 2009; Serrano et al. 2018), which is consistent with the spatiotemporal signaling hypothesis. The fact that conditioned media from either wild-type cells or conditioned media from cells that show increased fusion frequency failed to induce communication among Δham-11 germlings, or activate Pprm1-luciferase, suggests that the signaling molecule is highly reactive, short-lived, and/or that spatiotemporal dynamics are critical for function.

Here, we show that HAM-11 functions upstream of NRC-1, which is the MAPKKK of the MAK-2 signaling complex. The DOC-1 and DOC-2 proteins have also been described to function either with or upstream of the MAK-2 signaling complex, and genetic differences at doc-1 and doc-2 have been shown to negatively regulate the reinforcement of signaling through the MAK-2 pathway, which regulates chemotropic interactions (Heller et al. 2016). However, our data show that DOC-1 and DOC-2 also play a positive role in mediating chemotropic interactions and cell fusion in the absence of ham-11, where Δham-11 germlings chemotropically interact with wild-type germlings. These data are consistent with the observation that DOC-1 oscillates dynamically to cell tips with the MAK-2 signaling complex during chemotropic interactions between genetically identical cells (Heller et al. 2016). These observations support the hypothesis that positive chemotropism toward a potential fusion partner must be balanced with avoiding incompatible fusion partners. Here, we demonstrate a model for integrating self and non-self signals in fungal cells that guide chemotropic growth (Figure 8). The HAM-11 protein is required to initiate chemotropic interactions because a germling expressing HAM-11 is sufficient to initiate communication with a germling lacking HAM-11 (i.e., wild-type + Δham-11). After initiation, signaling through the MAK-2 pathway is reinforced depending on the genetic identity of the communication partner (specifically at the doc locus). Reinforcement of signaling facilitates MAK-2 complex formation and chemotropic growth, whereas blocked signaling mediated by allelic differences at the doc loci prevents MAK-2 complex formation, chemotropic growth, and cell fusion (Heller et al. 2016). Our data indicate that HAM-11 functions in parallel with DOC-1 and DOC-2 to modulate signal reinforcement or suppression, which defines an early checkpoint in the process of germling communication.

Figure 8.

HAM-11 functions in parallel with DOC-1 and DOC-2 to modulate self/non-self recognition, and signaling reinforcement. During interactions between cells with genetic identity at doc-1 and doc-2 [identical communication group (CG)], chemotropic interactions are initiated via the action of HAM-11 and/or DOC-1/DOC-2 to activate the MAK-2 MAPK signaling complex, which is essential for chemotropic interactions. DOC-1 is also part of the MAPK signaling complex and shows oscillatory localization to cell tips with MAK-2 during chemotropic interactions between cells with identical DOC-1/DOC-2 specificity (Heller et al. 2016). Cells that lack ham-11 and doc-1/doc-2 are completely fusion-deficient, both in self-fusion and in pairings with fusion-competent cells. During interactions between cells of different CGs (and thus different allelic specificity at doc-1/doc-2), communication is initiated via the action of HAM-11, but reinforcement of MAPK signaling is prevented by DOC-1/DOC-2, resulting in a lack of MAK-2 complex formation, and a significant reduction in chemotropic growth and cell fusion between cells of different CGs.

Two MAPK cascades (the MAK-1/CWI pathway and the MAK-2 pathway) together with the STRIPAK complex form a conserved signaling hub that integrates perceived information and directs appropriate cellular responses (Kück et al. 2016; Fischer et al. 2018). We used a Pprm1-luciferase reporter to identify which genes likely function upstream of the MAPK signaling hub and which genes have a downstream function. These data included a group of mutants (Δham-11, Δham-14, Δdoc-1, Δdoc-2, Δgyp-5, and Δras-2) that showed a moderate level of Pprm1-luciferase expression (Figure 1). Given these data, we hypothesized that the regulation of germling communication involves a positive feedback loop and that more than one pathway may function in parallel, upstream of the MAK-2 signaling cascade. The ham-11, ham-14, doc-1, doc-2, and gyp-5 genes may represent parallel pathways that converge on the MAK-2 pathway via ras-2, which encodes a conserved GTPase and is a well-characterized upstream regulator of the MAK-2 pathway (Kana-uchi et al. 1997; Dettmann et al. 2014). Transcription of ham-11 and doc-1 is regulated by MAK-2, PP-1, and ADV-1 (Leeder et al. 2013; Fischer et al. 2018), which is consistent with the positive feedback loop hypothesis. Additionally, the HAM-11, HAM-14, and DOC-2 proteins are phosphorylated in a MAK-2-dependent manner (Jonkers et al. 2014). The synthetic phenotype of the triple mutant Δdoc-1 Δdoc-2; Δham-11 supports the hypothesis that ham-11 functions in parallel with doc-1 and doc-2, and that these three genes together regulate aspects of asexual communication and cell fusion. Future work is necessary to further elucidate the specific relationship between ham-11, doc-1, and doc-2, in addition to how these three genes integrate with the rest of the molecular machinery that controls germling communication and cell fusion.