Permissiveness and competition within and between Neurospora crassa syncytia

Abstract A multinucleate syncytium is a common growth form in filamentous fungi. Comprehensive functions of the syncytial state remain unknown, but it likely allows for a wide range of adaptations to enable filamentous fungi to coordinate growth, reproduction, responses to the environment, and to distribute nuclear and cytoplasmic elements across a colony. Indeed, the underlying mechanistic details of how syncytia regulate cellular and molecular processes spatiotemporally across a colony are largely unexplored. Here, we implemented a strategy to analyze the relative fitness of different nuclear populations in syncytia of Neurospora crassa, including nuclei with loss-of-function mutations in essential genes, based on production of multinucleate asexual spores using flow cytometry of pairings between strains with differentially fluorescently tagged nuclear histones. The distribution of homokaryotic and heterokaryotic asexual spores in pairings was assessed between different auxotrophic and morphological mutants, as well as with strains that were defective in somatic cell fusion or were heterokaryon incompatible. Mutant nuclei were compartmentalized into both homokaryotic and heterokaryotic asexual spores, a type of bet hedging for maintenance and evolution of mutational events, despite disadvantages to the syncytium. However, in pairings between strains that were blocked in somatic cell fusion or were heterokaryon incompatible, we observed a “winner-takes-all” phenotype, where asexual spores originating from paired strains were predominantly one genotype. These data indicate that syncytial fungal cells are permissive and tolerate a wide array of nuclear functionality, but that cells/colonies that are unable to cooperate via syncytia formation actively compete for resources.


Introduction
In many filamentous ascomycete fungi, the main vegetative mycelial growth form is a haploid, multinucleate, interconnected syncytium formed by somatic cell fusion.Somatic fusion between germinated asexual spores (germlings) and hyphae can allow genetically distinct nuclei to occupy the same cytoplasm, a so-called heterokaryon (Supplementary Fig. 1).Syncytia occur across multiple domains of life, including muscle cells, the placenta in mammals, plasmodia of slime molds, and heterokaryotic cells of the angiosperm Utricularia (Mela et al. 2020).A potential advantage of syncytia in filamentous fungi is that nuclei and organelles have an opportunity to exchange resources ('public goods') and potentially complement deleterious mutations either through parasexual mitotic recombination or by being in a shared cytoplasm (Pontecorvo 1956;Anderson et al. 2018).Heterokaryons in filamentous fungi have been postulated to provide advantages of diploid cells, where genetically different haploid nuclei can coexist in a common cytoplasm.Further studies showed that regional differences in spatial arrangement of nuclei and mitotic synchronicity (Rosenberger and Kessel 1967;Clutterbuck 1970;Freitag et al. 2004), as well as transcription rates and nuclear number, can facilitate genotypic and phenotypic plasticity within syncytia (Schuurs et al. 1998;Gladfelter 2006;Gladfelter and Berman 2009).However, a disadvantage of syncytial growth strategies is that organelles (e.g.nuclei or mitochondria) can become non-cooperating entities, or "cheaters', within a colony, which benefit from intracellular or extracellular resources (public goods) produced by the "cooperator", but that do not share the cost of producing said resources (Smith and Schuster 2019).Some examples of such public goods in filamentous fungi could be secreted enzymes used for plant cell wall deconstruction, intracellular metabolites, antibiotic production, and compounds associated with the production of spores and tissues for asexual/sexual reproduction.Non-cooperators can also potentially perpetuate deleterious mutations, thereby posing a detriment to the colony overall.Such competition can lead to the entire colony being overrepresented by a mutant genotype, as is the case with the Neurospora crassa "poky" mutant.The poky mutant contains mutations in its mtDNA (Mannella and Lambowitz 1978;Akins and Lambowitz 1984) that provide the mutant with a higher relative fitness in syncytia due to increased replication rate of mitochondria, despite a range of deleterious pleiotropic phenotypes, most notably a defective growth rate, the presence of aberrant mitochondrial ribosomes, and cytochrome deficiencies.Competition can, in some cases, even result in the breakdown of the syncytium into homokaryons (Mannella and Lambowitz 1978;Rayner et al. 1995;Rayner 1996).An evolutionary study by Bastiaans et al. (2016) in N. crassa and subsequent genotyping of the associated evolutionary lines by Grum-Grzhimaylo et al. (2021) showed that syncytia carried across multiple generations resulted in the consistent recovery of mutants blocked in somatic cell fusion, specifically mutations in so (soft), ham-2, and ham-8 (Xiang et al. 2002;Fleissner et al. 2005;Fleissner and Glass 2007;Dettmann et al. 2013;Fu et al. 2014).The evolved syncytia bearing the fusion mutant nuclei did not exhibit the same growth disadvantages as homokaryotic fusion mutants in monoculture, suggesting that they indeed benefit from excreted and/or leaked public goods.Multicellular organisms can limit the spread of cheaters by the transition to a unicellular, uninucleate state during sexual reproduction, which creates a bottleneck of defective mutations or selfish elements, by reallocating them between haploid progeny (Grosberg and Strathmann 2007).
In N. crassa, there are at least three checkpoints that regulate somatic cell fusion.The first checkpoint regulates chemotropic interactions, the second checkpoint regulates cell wall dissolution upon contact, and the third checkpoint regulates the establishment of a heterokaryon following somatic cell fusion (Gonçalves et al. 2020).In the third checkpoint, if cells have different allelic specificity at vegetative incompatibility or heterokaryon incompatibility (HI) het loci, the fusion compartment is compartmentalized and undergoes a rapid regulated cell death process (Glass and Dementhon 2006;Daskalov et al. 2017;Goncalves et al. 2017).In N. crassa, heterozygosity at het loci has been shown to reduce heterokaryons found in nature (Pandit and Maheshwari 1996).In N. crassa, the mating type locus functions to trigger regulated cell death following somatic cell fusion of mat a and mat A cells (Pittenger 1957;Griffiths 1982;Glass et al. 1990;Glass and Kuldau 1992).Specific mutations in mat a-1 or mat A-1 abolish mating type incompatibility, but not sexual compatibility, while extragenic suppressor mutations at the tol (tolerant) locus suppress mating type incompatibility (Newmeyer 1970;Griffiths 1982;Jacobson 1992;Saupe et al. 1996;Shiu and Glass 1999).Some proteins encoded by het loci have functional and sequence homology to proteins involved in innate immunity in animals (Dyrka et al. 2014;Daskalov et al. 2016Daskalov et al. , 2020;;Dyrka et al. 2020;Daskalov and Glass 2022); HI reduces the transmission of mycoviruses and the transmission of defective genetic elements, including mitochondria associated with senescence (Debets et al. 1994;Liu and Milgroom 1996;van Diepeningen et al. 1997;Debets and Griffiths 1998).
In this study, we examined the dynamic mixing of differentially tagged histone H1 nuclei that differed in auxotrophic markers, intracellular public goods, or heterokaryon incompatibility factors.The relative fitness of each nuclear genotype was analyzed in a large population of asexual spores derived from syncytia via flow cytometry and fluorescence microscopy.Here, we show that the syncytia of N. crassa are permissive toward the propagation of nuclei with auxotrophic requirements or nuclear spacing defects in mycelia and asexual spores, despite the clear disadvantage of producing such asexual progeny.However, allorecognition resulting in cell death and the inability to undergo somatic cell fusion was a sufficient selective pressure to increase fitness of one nuclear genotype over another, resulting in a "winner-takes-all" phenotype, where one nuclear genotype reached near saturation in the harvested population of asexual spores.These results illustrate both the cooperative and competitive nature of fungal syncytia.

Strains, media, culturing methods, and sexual crosses
All N. crassa strains used in the study are listed in Supplementary Table 1.The wild-type N. crassa genetic background for tagged and crossed strains were OR74A (FGSC 2489 or FGSC 4200) (Colot et al. 2006).All deletion strains used in this study were obtained from the Fungal Genetics Stock Center (http://www.fgsc.net/)singlegene Neurospora deletion collection (McCluskey 2003;Colot et al. 2006).NCU numbers for the loci in this study are as follows: his-3 NCU04393; csr-1 NCU00726; arg-5 NCU05410; arg-12 NCU01667; ro-3 NCU03483; ro-10 NCU10696; tol NCU03378; cwr-1 NCU03180; cwr-2 NCU03182; rcd-1 NCU05712; so NCU02794; and fl NCU08726.Culturing and crossing, using Vogel's media (Vogel 1964) and Westergaard's media (Westergaard et al. 1947), respectively, were performed as previously described, with modification (Davis and De Serres 1970;Perkins 1986).All deletion strains from single-gene deletion collection were backcrossed once to either Wild Type (WT), histone H1 (H1)-eGFP or WT, H1-mCherry strains listed in Supplementary Table 1, to obtain histone-tagged homokaryons; the genotype of progeny was verified by PCR or phenotypic characterization.Fluffy mat A and fluffy mat a strains (BF and A7, respectively, Supplementary Table 1) are routinely crossed to mat A strains (FGSC 2489) to maintain fertility and which were used to determine mating type of progeny.Auxotrophic supplements were prepared in 100× and 1,000× stock solutions of 60-mg/mL L-arginine (Sigma-Aldrich) and 10-mg/mL calcium pantothenate (Fisher Biotech), respectively, filter sterilized, and stored at 4°C.Supplements were added to autoclaved media precooled to ∼5°C before pouring in plates or race tubes.Crosses were performed on Westergaard's medium (Westergaard et al. 1947).Ascospores were activated by incubating in PCR tubes at 6°C for 30 minutes in a Thermo Fisher Scientific MiniAmp thermal cycler.Ascospores were individually excised from plates onto slants containing 1-mL media (with supplements as needed), and progeny were subsequently verified by PCR for the proper genotype, tested for mating type, auxotrophic mutation and for fluorescence (Perkins 1986).Primers used in this study are listed in Supplementary

Cloning and transformation
Spheroplast preparation and transformation by electroporation were performed as previously described (Schweizer et al. 1981;Vollmer and Yanofsky 1986), except for selecting for resistance to hygromycin B (250 units/mL; Calbiochem).The H1-mCherry-tagged parent strain was constructed with a modified pMF272 his-3 targeting vector (Pccg-1-histone H1-5x Gly linker-eGFP or mCherry-Tccg-1), (Freitag et al. 2004).The eGFP sequence was removed from pMF272 by restriction digestion in universal buffer at 37°C with NEB XbaI and EcoRI restriction enzymes.An ∼1,174-bp product containing the native histone H1 sequence, along with a 5x Gly linker sequence was amplified from FGSC WT 2489 gDNA (Supplementary Table 2).A ∼1,083-bp product containing the mCherry gene and Tccg1 terminator sequence was amplified from plasmid pNLY1-4.Using Q5 NEB Polymerase, fusion PCR was conducted to combine fragments, followed by gel purification using a Qiaquick Qiagen Gel Extraction Kit, following the manufacturer's instructions.Fragments were cloned into digested vector pMF272 between the Pccg1 promoter and downstream his-3 flanking region, using NEB HiFi DNA Assembly Cloning Kit following manufacturer's instructions.
FGSC 6103 histidine auxotrophic conidia were harvested in 30-mL ice-cold 1 m sorbitol and filtered through 4 sheets of cheesecloth.Samples were centrifuged at 4°C and 3,400 rpm for 10 minutes, and the pellet was washed 3× with 15-30-mL ice-cold sorbitol.The conidial pellet was gently resuspended in as small volume of sorbitol, and spore concentration was counted and normalized to 5 × 10 9 conidia/mL.The plasmid was linearized using NdeI/SspI restriction site, and transformations were carried out as previously described (Colot et al. 2006).Transformants were plated with precooled recovery media (1.4-mL FIGS (10X) + 14-mL top agar [VMM Nitrate Salts (1X) with 1% agar)] onto thin layer of bottom agar (VMM Nitrate Salts (1X) with 2% agar, FIGS (1X), and hygromycin), and incubated at 3°C for 2-3 days.Primary transformants were selected for histidine prototrophy and hygromycin resistance and analyzed by PCR for insertion at the proper locus, and nuclear fluorescence was verified by microscopy.A selection of transformants with correct genotype was backcrossed to FGSC 2489 strain to obtain homokaryons with desired genotype (Ebbole and Sach 1990).

Microscopy
Each strain was harvested in ddH20, filtered through cheesecloth, and normalized to 1.5 × 10 7 conidia/mL before plating.All microscopy was conducted on a Zeiss Axioskop 2 MOT fluorescence microscope equipped with a CoolLED pe-300 microscope illuminator.Imaging was conducted using a QImaging Cool 12-bit Mono digital camera equipped on the fluorescent microscope and scale bars were inserted in iVision imaging software.Micrographs were analyzed, and brightness/contrast changes were uniformly applied to entire image in ImageJ; figures were prepared in Adobe Illustrator 2022 v26.5.

Methylene blue vital dye staining
For methylene blue (MB) experiments, strains were normalized to 1.5 × 10 7 conidia/mL, mixed 1:1, and added to 10-mL Vogel's minimal liquid media with 2% sucrose in a small petri dish on top of a flame-sterilized glass coverslip.Petri dishes were incubated in dark at 3°C for 5-6 hpi, and media was removed/discarded and immediately replaced with staining solution.Methylene blue trihydrate (Sigma; Lot # 99H3618) stock solution was prepared fresh each experiment by adding 0.01 to 1-mL ddH20 in small Eppendorf tube, and diluted 200× for working solution in 10-mL sterile ddH20.The surface of the mycelium-side-up coverslip was covered with 3 mL of working solution in petri plate and allowed to incubate for 3 minutes.The dye was removed from the petri dish, and 3× subsequent washes were done with 5-mL ddH20 or until the stain was visibly removed.Coverslips were immediately mounted hyphae-side-down onto a clean glass microscope slide and viewed with a 20× objective lens.Within 20-30 minutes, brightfield images were taken to visualize the blue staining of dead/dying hyphal compartments and/or clear viable hyphal compartments, and red and green fluorescence channels were observed to verify the presence of nuclei from each partner of the germling fusion.

Race tube experiments
Race tubes were 39 cm in length with a 1.2 cm internal diameter, and solidified media occupied approximately 30 cm across the bottom of each tube.Bird's Media was prepared fresh as previously described with 1.8% sucrose as the primary carbon source and 1.5% BTS Superpure Agar as the solidifying agent; 13 mL was aliquoted into sterile race tubes from the same media batch in each experiment (Metzenberg 2004).Strains were harvested in 2-mL ddH20 from slants and filtered through multiple layers of cheesecloth to remove mycelial fragments.Fresh spores were normalized to 1.95 × 10 7 conidia and mixed in 1:1 ratio prior to inoculation in a 10-μL droplet at the "Start" of each race tube in at least triplicate.In spore pairings with ropy mutants, the initial inoculum was a 1:1.3 ratio (1.95 × 10 7 conidia/mL:2.54× 10 7 conidia/ mL, prototrophic:ropy spores), to normalize the number of Δro-3 and Δro-10 spores with detectable fluorescence signal.Inoculum ratios for each spore pairing were quantified by flow cytometry on day 0 of each independent experiment.Race tubes were placed in an incubator at 3°C in the dark O/N to ensure inoculum was dry, and initial growth had started before flipping race tubes to avoid potential pooling of CO 2 , which could potentially affect conidiation (Sargent and Kaltenborn 1972).Race tubes were incubated ∼7 days.To harvest the spores from race tubes, 1 mL of ddH20 was added to the "start" and "end" of each tube, and gently pipetted up and down to free spores into solution, followed by filtration through sterile cheesecloth.Race tubes and controls were harvested together and analyzed on the same day to ensure consistent results.
For linear growth rate determinations, conidia were harvested, filtered, normalized to 1.95 × 10 7 conidia/mL, and inoculated in 10-μL droplets from the same inoculum in at least triplicate onto race tubes containing fully supplemented Bird Media, and incubated in the dark at 3°C.Mycelial growth was measured (mm) each day on race tubes at the same time for 3 days.

Flow cytometry
Nuclear fluorescence within spores were recorded for ∼10,000 to ∼20,000 events for remaining samples on an LSR Fortessa Analyzer using PE-Texas Red 600LP | 610/20 filter and FITC 505LP | 525/50 filter.Voltages are calibrated with control samples of each individual strains used in the experiment.Events from the sample of interest are counted by the flow cytometer for each replicate sample per spore pairing.Cells are gated for spore size/ granularity and the presence of cell stuck together (doublets) based on [side scatter area × forward scatter area] and [forward scatter height × forward scatter area] measurements, respectively, to avoid counting spores too large/small/granular, cell debris, and cells stuck together.Intial data showed that cytoplasmic GFP and cytoplasmic dsRed could be used to assess heterokaryotic spore formation, but the fluorescence signal was much brighter using histone H1-eGFP and histone H1-mCherry.The data were further grouped in a two-parameter density plot, based on fluorescence signal in 4 quadrants (Supplementary Fig. 2).The PE-Texas Red filter detects hH1-mCherry fluorescence, and FITC filter detects hH1-GFP fluorescence.Spore size, nuclear number, and diffusion of histones are not likely to affect the majority of the spores counted within each quadrant, although a fraction of spores recorded to be on the cusp between quadrants could be over/under counted as "mixed" spores, meaning that they are showing both GFP and RFP fluorescence signals.BDFACs Diva software was used to analyze samples, and the same experimental parameters were used for every flow cytometry experiment in this study, with manual adjustments to voltage for proper discrimination of single fluorophores between runs.Graphing and statistical tests for each experiment were conducted using GraphPad Prism 9.4.1.681.The statistical significance within each dataset was determined by running a two-way ANOVA with replication.Multiple comparisons test was done using parameters ensuring all cell means were compared regardless of rows and columns.Corrections for multiple comparisons were conducted using statistical hypothesis testing with Tukey's pairwise post hoc t-tests (family-wise alpha threshold-0.05;95% confidence interval; N = 3-6).

Prototrophic vs auxotrophic nuclear cooperation in fungal syncytia
Within the fungal multinucleate syncytium, the number of nuclei within a hypha can vary, with the large trunk hyphae (∼10 μm in diameter) containing large numbers of nuclei, while other hyphae contain fewer numbers of nuclei (Turian and Bianchi 1972;Freitag et al. 2004).To assess nuclear composition of syncytia under different genetic scenarios in a N. crassa, we analyzed frequency of nuclear genotypes in asexual spores using flow cytometry.Asexual spores in N. crassa contain, on average, between 2 and 5 nuclei that are compartmentalized within a spore during conidiation (Turian and Bianchi 1972; Ruger-Herreros and Corrochano 2020) (Supplementary Fig. 1).To facilitate these experiments, we introduced a histone H1-GFP allele to mark nuclei of one strain and H1-mCherry allele to mark nuclei in a second strain.We first assessed the linear growth rate of strains in monoculture (Supplementary Fig. 3).For spore pairing experiments, spore counts from individual strains were normalized to 1.95 × 10 7 conidia/mL and subsequently co-inoculated in a 1:1 ratio onto 30-cm race tubes in triplicate.Conidia at the beginning of the race tube were harvested on day 7 ("start"), which represented ∼1-2 days of growth, and at the "end" of the race tube on day 7, which represented 7 days of growth.
We first assessed pairings of strains that were of identical mating type, but where nuclei were differentially marked (hH1-GFP mat A or hH1-mCherry mat A); both these strains had identical linear growth rates (92.4 mm/day) (Supplementary Fig. 3).The inoculum for h1-GFP mat A and H1-mCherry mat A spores was normalized on day 0 in a near 1:1 ratio by flow cytometry (Fig. 1b), and conidial suspensions were inoculated in race tubes at the "inoculation site" (blue circle) of a 30-cm race tube (Fig. 1a).After 7 days, conidial suspensions at the "start" and "end" of the race tube were analyzed by flow cytometry (at least 10,000 spores).Fluorescence in the spores from the race tube was either green (presence of only GFP fluorescence), magenta (presence of only mCherry fluorescence), or both magenta and green (presence of both GFP and mCherry fluorescence) (Supplementary Fig. 2).From the h1-GFP mat A and H1-mCherry mat A conidial pairings, we observed a significant number of homokaryotic conidia, containing only H1-GFP or H1-mCherry signal, as well as ∼34% of heterokaryotic conidia containing both H1-GFP and H1-mCherry in conidial suspensions at both the "start" and "end" of the race tubes (Fig. 1b and c and Supplementary Fig. 2).
We then paired prototrophic conidia with conidia bearing an auxotrophic mutation; we postulated that on non-supplemented media, the population of auxotrophic nuclei would be outcompeted by wild-type nuclei in the syncytia and that nuclei containing the specific H1-tag associated with the auxotrophic mutation would be underrepresented in asexual spores.We also predicted that when the required exogenous supplementation was provided, the auxotrophic nuclei would increase in frequency in the spore suspensions compared to prototrophic nuclei.For these experiments, we used two auxotrophic mutants, one blocked in the ability to synthesize the vitamin B 5 (pantothenic acid; pan-2) and a second mutant unable to synthesize the amino acid arginine (arg-5); both mutants were unable to grow on non-supplemented media.The linear growth rate of the pan-2 mutant on supplemented media (93 mm/day) was not significantly different from WT (92.4 mm/day), but the arg-5 mutant on supplemented media showed a moderately reduced linear growth rate (79.4 mm/day) (Supplementary Fig. 3).
To examine how auxotrophic genotypes behave in a heterokaryon, spores of the Δpan-2 or Δarg-5 mutants were mixed in a 1:1 ratio with prototrophic spores and inoculated onto race tubes with or without supplementation (arginine + pantothenic acid) (Fig. 1a and b).Flow cytometry analysis of spore suspensions at the "start" and "end" of the race tubes on non-supplemented media showed that Δpan-2 and Δarg-5 nuclei were maintained throughout the colony as shown by the production of heterokaryotic spores, and homokaryotic Δpan-2 or Δarg-5 spores (Fig. 1c  and d).However, in these pairings, prototrophic spores (Fig. 1c; H1-mCherry and Fig. 1d; H1-GFP) were significantly overrepresented in spore suspensions at both "start" and "end" of the race tube, particularly in comparison to arg-5 spores.Homokaryotic auxotrophic spores were nevertheless regularly observed (∼22-29%) as well as heterokaryotic spore populations (∼26-35%) (Fig. 1b and c).These data indicated that nuclei unable to support colony growth due to their auxotrophic mutations, and thus dependent on the presence of WT nuclei for production of essential nutrients for survival, could still be compartmentalized in both homokaryotic and heterokaryotic spores.
We predicted that in pairings between Δpan-2 conidia and Δarg-5 conidia, that Δpan-2 nuclei would be overrepresented in spore suspension at the end of the race tube due to the lower requirement for pantothenic acid for growth as compared to arginine, and that this result would be abolished with supplementation.To test this prediction, equal ratios of Δpan-2 and Δarg-5 conidia were inoculated into race tubes, and spore suspensions were evaluated by flow cytometry at the beginning and end of the race tubes.Homokaryotic Δpan-2 spores showed a higher frequency as compared to Δarg-5 spores at the start of the race tube experiments on non-supplemented media, but this discrepancy was abolished by the time syncytia grew to the "end" of race tubes (Fig. 1e).Under supplementation experiments in the prototrophic + Δpan-2 pairings, the percentages of homokaryotic prototrophic and Δpan-2 conidia were similar (∼25-28%) with a significant shift to heterokaryotic spores (∼45-48%) (Fig. 1f).Also, as predicted, under supplementation conditions, we observed a significant increase in the frequency of homokaryotic Δarg-5 spores in prototrophic + Δarg-5 pairings (Fig. 1g).To determine if this phenotype was particular to the arg-5 mutation or was more broadly applicable to strains containing mutations in the arginine biosynthesis pathway, a second experiment was conducted by pairing prototrophic conidia with Δarg-12 conidia (also an arginine auxotroph) (Supplementary Fig. 4).Similar to results with prototrophic + Δarg-5 pairings on non-supplemented media (Fig. 1d), in prototrophic + Δarg-12 pairings on non-supplemented media, ∼55% of spores were prototrophic, while Δarg-12 nuclei were present in ≤10% of the homokaryotic spore population; ∼40% of the spores were heterokaryotic (Supplementary Fig. 4b).When arginine was added to the media, heterokaryotic spores increased in frequency to ∼62% in prototrophic + Δarg-12 pairings; differences in the percentages of prototrophic vs Δarg-12 homokaryotic conidia was abolished (Supplementary Fig. 4c).Thus, the arginine supplementation shift toward Δarg-5 homokaryotic spores in prototrophic + Δarg-5 pairings was not observed in prototrophic + Δarg-12 pairings, although the overall effect of increasing the abundance of auxotrophic nuclei in asexual spores upon supplementation remained consistent.These data indicate that the nuclear ratios within a syncytium can be modified by external nutritional supplementation and mutational spectrum.Overall, these results showed that N. crassa syncytia are highly permissive toward harboring nuclei that "cheat" for nutrient supplementation and allowed the production of auxotrophic homokaryotic conidia despite these conidia being unable to germinate and grow on the media from which they were derived.

Permissiveness for nuclear distribution mutants in syncytia
Previous studies elucidated how bulk flow (or cytoplasmic streaming) contributes significantly to nuclear mixing within heterokaryotic hyphal compartments in N. crassa (Roper et al. 2013).In addition to bulk flow, dynein-dynactin motors also contribute to nuclear transport (Ramos-Garcia et al. 2009).Based on shared nutritional resources in syncytia shown by experiments in the previous section (Fig. 1), we predicted that if one nucleus were to be defective in motor-driven nuclear transport, when paired with WT nuclei, there would be little or no decrease in the fitness of either nucleotype.To test this hypothesis, we used two different (ropy) mutants, Δro-3 and Δro-10; the former encodes for a subunit of dynein/dynactin motor protein complex, while the latter is responsible for localization/integrity of cytoplasmic dynein/dynactin (Bruno et al. 1996;Minke, Lee, and Plamann 1999;Minke, Lee, Tinsley et al. 1999;Xiang and Plamann 2003).Both mutants showed a severe nuclear spacing defect phenotype (Fig. 2b and c and Supplementary Fig. 5); bulk flow was previously shown to be sufficient to maintain some level of retrograde and anterograde nuclear movement (Ramos-Garcia et al. 2009).Consistent with their nuclear spacing defects, both the Δro-3 and Δro-10 mutants grew significantly slower than WT (28.5 mm/day and 26.9 mm/day, respectively, as compared to 92.4 mm/day for WT) (Supplementary Fig. 3).The ro-3 and ro-1 mutants showed a terminal linear growth phenotype on the race tubes at 8.1 and 8.6 cm of growth, respectively.However, WT + Δro-3 and WT + Δro-10 heterokaryons showed restored growth, as compared to the Δro-3 and Δro-10 strains in monoculture.
It was previously shown that ropy mutants exhibit ∼90% anucleate hyphal tips and ∼50% anucleate hyphal compartments (Plamann et al. 1994).Consistent with these findings, no nuclear fluorescence was observed in ∼15% of homokaryotic spores from the Δro-3 or Δro-10 mutants in monoculture (Fig. 2d); by comparison, homokaryotic WT strains displayed a small fraction (≤0.5%) of spores with no fluorescence.In pairings between Δro-3 or Δro-10 mutants and WT, the initial inoculum ratios evaluated by flow cytometry showed that ∼5-8% of spores lacked nuclear fluorescence (Fig. 2d).In contrast, conidial suspensions from the end of the race tube of the WT + Δro-3 or WT + Δro-10 pairings showed a low percentage of spores with no fluorescence (1.1 and 0.35%, respectively).The percentage of heterokaryotic spores in the WT + Δro-3 and WT + Δro-10 pairings was ∼30-35% (Fig. 2e).There was a significant increase in both ro-3 and ro-10 homokaryotic spores at both the "start" and "end" of the race tube relative to the initial inoculum levels (Fig. 2e).These data indicate that the Neurospora syncytia is highly permissive toward sharing of intracellular public goods, such as molecular motors, and that even nuclei with severe nuclear spacing defects have access to asexual spore formation.

Mating type incompatibility leads to strong competitive selection and a "winner-takes-all" phenotype
It has been previously shown that opposite mating type strains (mat A and mat a) are required for mating and meiosis during the sexual cycle, but somatic cell fusion between mat a + mat A hyphae leads to HI, cell death, and breakdown of heterokaryotic colonies (Gross 1952;Metzenberg and Glass 1990;Glass and Kuldau 1992).Thus, we predicted that asexual spore pairings between mat A + mat a strains would result in a drastic reduction in heterokaryotic spores, and result in mat A or mat a homokaryotic spores in ratios representative of the initial inoculum.
Individual compatible mat A + mat A or mat a + mat a spore pairings from single race tubes showed 28-56% heterokaryotic spores at the "start" and "end" of race tubes, as well as homokaryotic spores (Fig. 3).In contrast, mat A + mat a spore pairings showed a very small fraction (≤3%) of heterokaryotic spores, and a near "saturation" (≥92%) of one mating type nucleotype in homokaryotic spores (Fig. 3).The dominance, or near saturation of a particular mat A or mat a genotype in spore pairings, was termed as a "winner-takes-all" phenotype.Importantly, the nucleotype that showed near saturation in replicates was independent of mating type (although the mat a genotype "won" more often than the mat A genotype).Thus, contrary to our hypothesis, these results show that mating type incompatibility creates a clear selective advantage for one nuclear genotype in pairings of opposite mating type asexual spores.
Although mating type incompatibility has been investigated in colonies, it has not previously been evaluated in germlings undergoing somatic cell fusion.To visualize germling fusion events occurring at the "start" of the race tubes, we conducted vital dye staining using MB and fluorescence microscopy of mating type compatible (mat A + mat A) vs incompatible (mat A + mat a) spore pairings (Fig. 4a and b).We observed evidence of mating type incompatibility via vacuolization and dark blue MB staining of fused hyphal compartments between opposite mating type germlings, starting at ∼6 hpi (Fig. 4b).Mating type compatible germlings (mat A + mat A) showed no evidence of MB staining post-fusion (Fig. 4a).

TOL, required for mating type incompatibility, is not nuclear-limited and suppresses the "winner-takes-all" phenotype
Previous findings showed that loss-of-function mutations at the tol (for "tolerant") locus are sufficient to suppress mating type incompatibility in heterokaryons (Newmeyer 1970); tol mutations have no effect on sexual reproduction (Griffiths 1982;Saupe et al. 1996;Shiu and Glass 1999).The tol locus encodes a protein containing a HET domain, which is of unknown biochemical function, but is a motif commonly observed in proteins associated with HI (Paoletti and Clave 2007).We first assessed the phenotype of fused germlings by fluorescence microscopy on Δtol mutant spore pairings to determine if mat A Δtol + mat a Δtol pairings showed any evidence of death, as observed with mat A + mat a pairings (Fig. 4b).Pairings of germlings with identical mating type (mat A + Δtol mat A and Δtol mat A + Δtol mat A) showed no evidence of cell death following fusion, as expected (Fig. 4d and e).Pairings between Δtol mat A + Δtol mat a germlings also did not show any evidence of cell death, consistent with the ability of mutations at tol to suppress mating type incompatibility (Fig. 4f).However, vital dye staining of Δtol mat a + mat A spore pairings revealed that fused cells showed vacuolization and death (dark blue staining pattern in "Brightfield Channel') starting at approximately 6-8 hpi (Fig. 4c), with a phenotype similar to fusion between mat A + mat a germlings (Fig. 4b).These data are consistent with previous results showing that mutations in tol are recessive (Newmeyer 1970;Jacobson 1992).
We predicted that if both partners in a mating type incompatible heterokaryon also carried a Δtol mutation, spore pairings of identical mating type (Δtol mat A + Δtol mat A) or opposite mating type (Δtol mat A + Δtol mat a) would show similar frequencies of H1-GFP, H1-mCherry, and heterokaryotic spores as assessed by flow cytometry.Consistent with this prediction, mat A + mat A, Δtol mat A + mat A, Δtol mat A + Δtol mat A, and Δtol mat A + Δtol mat a spore pairings did not display a winner-takes-all phenotype.For the Δtol mat A + Δtol mat a spore pairing in particular, both nucleotypes were found in homokaryotic spores and also showed a significant fraction (38-54%) of heterokaryotic spores (Fig. 5b).These results show that the tolerant mutation was sufficient to suppress the "winner-takes-all" phenotype observed in opposite mating type pairings and that the incompatibility function of mating type was essential for this phenotype.In contrast, all pairings between spores of opposite mating type, but where the tol mutation was in only one partner, showed a similar winner-takes-all phenotype as opposite mating type spore pairings (Fig. 5c and d).Importantly, in some cases, the wild-type genotype won, while in others, the tol mutant genotype won, showing that there was no bias for nuclear origin of tol.These data support our finding that mating type incompatibility leads to partitioning of nuclear genotypes into individual homokaryons by failure to maintain viable heterokaryotic syncytia.Furthermore, these results support the hypothesis that mating type incompatibility mediated by TOL produces a barrier to nuclear mixing of opposite mating types, which not only serves the function of allorecognition but consequently produces a distinct competitive advantage for one nucleotype.

The winner-takes-all phenotype is not limited to mating type incompatibility
A major aspect of HI mediated by allelic differences at het loci is cell permeabilization and death of the fusion compartment after somatic cell fusion (Saupe 2000;Glass and Kaneko 2003).Our data indicates that mating type incompatibility results in a winner-takes-all phenotype when opposite mating type spores are co-inoculated.We hypothesized that the induction of cell death drives this "winner-takes-all" phenotype.We therefore assessed whether genetic differences at a different het locus, the regulator of cell death-1 (rcd-1) locus (Daskalov et al. 2019;Daskalov et al. 2020), also result in a winner-takes-all outcome in incompatible pairings.
If regulated cell death is the major driver of the nuclear dominance phenotype, we predicted that pairings between incompatible rcd-1 strains would also exhibit a winner-takes-all phenotype.To test this hypothesis, we paired equal ratios of otherwise isogenic wild-type rcd-1-1 strain with a Δrcd-1 strain with the rcd-1-2 allele targeted to the csr-1 locus and of identical mating type onto race tubes and evaluated the frequency of each genotype in asexual spores at the start and the end of the race tube.Pairings between rcd-1-1 + rcd-1-2 strains showed a very similar winner-takes-all phenotype as mating type incompatible spore pairings (Fig. 3), with a clear dominance of one nuclear genotype (rcd-1-2) (Fig. 6a).Spore pairings between strains that were of opposite mating type and different at rcd-1 (rcd-1-1 hH1-mCherry mat a + rcd-1-2; cytoplasmic-GFP mat A) also showed a winner-takes-all phenotype that was not significantly different than pairings between rcd-1 + rcd-1-2 of identical mating type (Fig. 6b).These data show that regulated cell death (RCD) mediated by allelic differences at het loci such as mating type or rcd-1 is a major driver that enables one genotype to outcompete another in spore pairings between incompatible strains.

A block in hyphal fusion partially recapitulates the "winner-takes-all" phenotype
RCD triggered by allelic differences at het loci during somatic growth in filamentous fungi first requires cell fusion (Saupe 2000;Fischer and Glass 2019).We predicted that a block in cell fusion, which would eliminate cell death, would abolish the "winner-takes-all" phenotype of mating type incompatible and rcd-1 incompatible pairings.To test this hypothesis, we used two strategies.The first strategy used strains with allelic differences at cwr (cell wall remodeling), a locus that encodes two linked genes (cwr-1 and cwr-2) that regulate cell wall dissolution and germling/hyphal fusion in N. crassa (Gonçalves et al. 2019;Detomasi et al. 2022); post-fusion phenotypes (such as cell death) are not regulated by genetic differences at the cwr locus.cwr-1 alleles from one of the six CWR haplogroups found in N. crassa populations result in a block in somatic cell fusion when paired with germlings bearing cwr-2 from any of the other CWR haplogroups.For example, paired germlings expressing incompatible cwr-1/ cwr-2 alleles from different haplogroups (for example, cwr-1 HG1 + cwr-2 HG6 ) show a block in cell wall dissolution during somatic cell fusion events (Gonçalves et al. 2019;Detomasi et al. 2022).In a second strategy, we used a mutant termed soft (so), which has a drastic defect in chemotropic interactions and somatic cell fusion with itself and with wild-type cells (Fleißner et al. 2005).Fig. 5. Flow cytometry analysis of spore pairings with mating type compatible and incompatible Δtol strains.Spores were inoculated into race tubes in approximately equal ratios and grown for 7 dpi at 3°C. a) Inoculum ratios for all spore pairings used in this flow cytometry experiment.b) Identical mating type his-3::hH1-GFP (G) mat A + his-3::hH1-mCherry (R) mat A (10BI + 74DM) spore pairings; identical mating type (G) mat A + Δtol (R) mat A (10BI + 76BM) spore pairings; identical mating type Δtol (G) mat A + tol (R) mat A (63FB + 76BM) spore pairings; and opposite mating type Δtol (G) mat A + Δtol (R) mat a (63FB + 76BI) spore pairings.c) Individual biological replicates from opposite mating type spore pairings of (G) mat A + (R) mat A (10BI + 74DM).d) (G) mat A + Δtol (R) mat a (10BI + 76BI).Flow cytometry was conducted to analyze the relative percentage of each fluorescently tagged nuclear genotype in asexual spore populations derived from syncytia."Start" denotes that spores were collected from the opening of the race tube most proximal to the inoculation site on day 7, and "End" denotes that the spore sample was collected from the portion of the race tube most distal from the site of the inoculation on day 7. Left (magenta) bars in graphs denote fluorescence signal from homokaryotic spores with the presence of only histone H1-mCherry-tagged nuclei, right (inoculum) or middle (green) bars in graphs denote fluorescence signal from homokaryotic spores with the presence of only histone H1-eGFP-tagged nuclei, and black bars in graphs denote spores with signal from both histone H1-mCherry and histone H1-GFP nuclear tags in the same cell.All five biological replicates for each spore pairing are shown individually in panels c and d to clearly present which genotype nearly reached saturation in each race tube trial of the experiment.Statistical analysis was conducted using two-way ANOVA followed by Tukey's HSD post hoc tests.*P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001; ns = not significant.N = 5.Error bars = SEM.
Spore pairings between cwr-1 HG1 cwr-2 HG1 + cwr-1 HG2 Δcwr-2 showed no significant differences in the percentages of homokaryotic or heterokaryotic spores from fully compatible cwr spore pairings (cwr-1 HG1 cwr-2 HG1 + cwr-1 HG1 Δcwr-2) (Fig. 7b).However, cwr-1 HG1 cwr-2 HG1 + cwr-1 HG6 Δcwr-2 spore pairings, which have a more severe reduction in cell fusion (∼95%) (Detomasi et al. 2022), showed a significant reduction in the percentage of heterokaryotic spores in spore suspensions from both the start and end of the race tubes (Fig. 7b).From these data, we predicted that pairings between cwr-1 HG1 cwr-2 HG1 mat A + cwr-1 HG6 Δcwr-2 mat a strains would not exhibit the drastic winner-takes-all phenotype associated with mating type incompatibility (Fig. 3), due the reduction in cell fusion, cell death, and production of heterokaryotic spores.However, in spore pairings of opposite mating type, the cwr-1 HG1 cwr-2 HG1 nucleotype was dominant when paired with spores of cwr-1 HG2 Δcwr-2 or cwr-1 HG6 Δcwr-2 genotypes (Fig. 7c), although a significant percentage of homokaryotic cwr-1 HG6 Δcwr-2 spores was present in spore suspensions at the start of the race tube.These data show that the significant reduction in cell fusion observed in incompatible CWR pairings was not sufficient to abolish the "winner-takes-all" phenotype of mating type incompatible pairings and suggests that the post-fusion cell death function of het loci is an important driver for this phenotype.
The incompatible cwr strains showed a significant reduction, but not a complete block in cell fusion, and did not suppress the winner-takes-all phenotype of opposite mating type strains.We therefore chose to also assess interactions using a strain that is blocked in cell fusion, soft (so).Null mutants in so are unable to undergo chemotrophic interactions resulting in ∼300-fold reduction in somatic cell fusion (Fleissner et al. 2005).The SOFT protein oscillates at the tips of cells undergoing chemotropic interactions (Fleissner et al. 2009;Fleissner and Herzog 2016) and has been shown to be a scaffold for the upstream components of the Cell Wall Integrity MAPK signaling pathway (Teichert et al. 2014;Weichert et al. 2016).We reasoned that there should be an extremely low proportion of heterokaryotic spores produced in Δso + Δso pairings; moreover, without the cell death associated with post-fusion HI, we predicted that homokaryotic nuclear populations from race tube spore suspensions would reflect initial spore genotype frequencies in the inoculation samples.Flow cytometry of Δso mat A (R) + Δso mat A (G) spore pairings of identical mating Fig. 6.Flow cytometry analysis of spore pairings between rcd-1-1 + rcd-1-2 of identical and opposite mating type.a) Individual biological replicates of mating type compatible csr-1::rcd-1-2 his-3::eGFP Δrcd-1 (G) mat A + rcd-1-1 his-3::hH1-mCherry (R) mat A (5AB + 74DM) spore pairings.b) Individual biological replicates of mating type incompatible csr-1::rcd-1-2 (G) Δrcd-1 mat A + rcd-1-1 (R) mat a (5AB + 74ED) spore pairings.Spores were inoculated in approximately equal ratios and grown in race tubes for 7 dpi at 3°C.Flow cytometry was conducted to analyze the relative percentage of each fluorescently tagged nuclear genotype in asexual spore populations.Spore pairings in each biological replicate were derived from a single inoculum sample, and the ratios of each partner in the inoculum are shown in the first "Inoculum" column of each graph."Start" denotes that spores were collected from the opening of the race tube most proximal to the inoculation site on day 7, and "End" denotes that the spore sample was collected from the portion of the race tube most distal from the site of the inoculation on day 7. Left (magenta) bars in graphs denote fluorescence signal from homokaryotic spores with the presence of only histone H1-mCherry-tagged nuclei, right (inoculum) or middle (green) bars in graphs denote fluorescence signal from homokaryotic spores with the presence of only histone H1-eGFP-tagged nuclei, and black bars in graphs denote spores with signal from both histone H1-mCherry and histone H1-GFP nuclear tags in the same cell.All six biological replicates for each spore pairing are shown individually to clearly present which genotype nearly reached saturation in each race tube trial of the experiment.type showed nearly equal ratios of both spore genotypes at the start of the race tube, but, unexpectedly, showed only one nucleotype at the end of the race tube (Fig. 8a).These results were similar to mating type incompatible Δso mat a (R) + Δso mat A (G) pairings (Fig. 8b).However, in replicates 5 and 6 of the Δso mat a + Δso mat A pairings, neither nuclear genotype reached saturation, albeit the nucleotype in the slight majority in the inoculum was more represented [≥75%; Δso mat A (G)] in the final spore suspension.As Fig. 7. Flow cytometry analysis of spore pairings with mating type compatible and incompatible cwr-1 HG1 , cwr-1 HG2 , and cwr-1 HG6 in pairings with a cwr-1 HG1 cwr-2 HG2 strain.(a) Inoculum ratios for all spore pairings used in this flow cytometry experiment.b) Mating type compatible spore pairings cwr-1 HG1 cwr-2 HG1 his-3::hH1-mCherry (R) mat A (74DM) with cwr-1 HG1 Δcwr-2 his-3::eGFP (G) mat A (336a) or cwr-1 HG2 Δcwr-2 (G) mat A (88a) or cwr-1 HG6 Δcwr-2 (G) mat A (385a) strains.c) Individual replicates of the mating type incompatible spore pairings of cwr-1 HG1 cwr-2 HG1 (R) mat a (74ED) + cwr-1 HG1 Δcwr-2 (G) mat A (336a) or cwr-1 HG2 Δcwr-2 (G) mat A (88a), or cwr-1 HG6 Δcwr-2 (G) mat A (385a).Each strain was inoculated in approximately equal ratios and grown in race tubes for 7 dpi at 3°C.Flow cytometry was conducted to analyze the relative percentage of each fluorescently tagged nuclear genotype in asexual spore populations derived from syncytia.Spore pairings in each biological replicate were derived from a single inoculum sample."Start" denotes that spores were collected from the opening of the race tube most proximal to the inoculation site on day 7, and "End" denotes that the spore sample was collected from the portion of the race tube most distal from the site of the inoculation on day 7. Left (magenta)  predicted, a very small proportion (≤5%) of the spore population was heterokaryotic in the Δso mat A (R) + Δso mat A (G) pairings and the opposite mating type Δso mat a (R) + Δso mat A (G) pairings.
Previous data showed that Δso and nuclei containing mutations in other loci required for somatic cell fusion mutants are present in evolved strains that "cheat" their way into spores (Grum-Grzhimaylo et al. 2021).The Δso mutant showed a minor growth defect (76.4 mm/ day) in linear growth as compared to wild-type cells (92.4 mm/day) (Supplementary Fig. 3).We hypothesized that the wild-type germlings/colonies would be able to fuse with themselves, and potentially with adjacent Δso germlings/colonies, while Δso germlings/colonies would be unable to fuse with themselves or wild-type cells, resulting in a selective advantage for wild type as compared to Δso nucleotypes.Consistent with this prediction, in both mating type compatible and incompatible WT + Δso spore pairings, we observed that the WT genotype showed a "winner-takes-all" phenotype (Fig. 8c and d).
Post hoc t-tests showed that the percentage of heterokaryotic spores in either mating type compatible or incompatible (WT + Δso) or (Δso + Δso) spore pairings, or with mating type incompatible (mat A + mat a) spore pairings, was not statistically different.The WT + Δso pairings showed a clear bias for the WT genotype at the "start" of the race tubes, whereas in experiments with Δso + Δso pairings, the spore ratios at the "start" of the race tubes were very similar to the initial starting inoculum.These data show that the winner-takes-all phenotype can be driven not only by post-fusion cell death but can also occur when cells of different genotypes are unable to undergo somatic cell fusion.

Discussion
Genetic cooperation and conflict on an organismal level have been an important topic of study in biology, starting most notably with Fig. 8. Flow cytometry analysis of mating type compatible and incompatible Δso + Δso spore pairings.a) Individual biological replicates of mating type compatible spore pairings with Δso his-3::hH1-eGFP (G) mat A + Δso his-3::hH1-mCherry (R) mat A (19C + 80AS).b) Mating type incompatible spore pairings with Δso (G) mat A + Δso (R) mat a (19C + 80BB).c) Mating type compatible spore pairings with Δso (R) mat A + (G) mat A (80AS + 10BI).d) Mating type incompatible spore pairings of Δso (R) mat a + (G) mat A (80BB + 10BI).Conidia from these strains were inoculated in approximately equal ratios in race tubes and grown for 7 dpi at 3°C.Flow cytometry was conducted to analyze the relative percentage of each fluorescently tagged nuclear genotype in asexual spore populations.Spore pairings in each biological replicate were derived from a single inoculum sample, and the ratios of each partner in the inoculum are shown in the first "Inoculum" column of each graph."Start" denotes that spores were collected from the opening of the race tube most proximal to the inoculation site on day 7, and "End" denotes that the spore sample was collected from the portion of the race tube most distal from the site of the inoculation on day 7. Left (magenta) bars in graphs denote fluorescence signal from homokaryotic spores with the presence of only histone H1-mCherry-tagged nuclei, right (inoculum) or middle (green) bars in graphs denote fluorescence signal from homokaryotic spores with the presence of only histone H1-eGFP-tagged nuclei, and black bars in graphs denote spores with signal from both histone H1-mCherry and histone H1-GFP nuclear tags in the same cell.N = 6.
Hamilton's studies of social insects (Hamilton 1964).More recently, investigations on microbes have revealed mechanisms of cooperation and conflict (West et al. 2007).Organelles, such as nuclei, mitochondria, and chloroplasts, may act as semiautonomous units within an organism, encoding distinct genetic information and thus providing additional levels of genomic conflict within cells and syncytia, beyond inter-organismal social structures (Dobrogojski et al. 2020;Kouvelis and Hausner 2022).Here, we show that the network of N. crassa is permissive to the generation and maintenance of defective nuclei within a syncytium, as long as WT nuclei were present.WT nuclei essentially "cooperate" with defective nuclei by providing resources that the defective nucleus cannot produce.Syncytia facilitate dispersion of these nuclei and WT intracellular public goods primarily via bulk flow, and are independent of nuclear genotype, resulting in the production of both homokaryotic and heterokaryotic spores.This capacity is a type of "bet hedging", such that beneficial mutations within the syncytium can be carried to the next generation either as homokaryotic spores or as heterokaryotic spores, and where mutant nuclei could potentially be subjected to further evolution via mutational events within the resulting heterokaryotic syncytia.Bet hedging has been previously used to explain heterogeneity in asexual spore size and germination rates in Aspergillus fumigatus based on fluctuations in environmental conditions from which the spores were derived (Kang et al. 2021).This aspect has also been explored in artificial evolution experiments in N. crassa, by systematically analyzing sexual spore germination and viability under alternating time periods of favorable and unfavorable environmental conditions (Graham et al. 2014).
In contrast to the highly permissive nature of sharing intracellular goods in spore pairings of auxotrophic and ropy mutants with WT, here we show that permissiveness is limited to syncytia that have genetic identity at post-fusion het loci.Allorecognition mediated by genetic differences at het loci and which result in cell death following somatic cell fusion has been shown to be important for the reduction in transfer of mycoviruses and defective mitochondria between genetically different strains in a number of filamentous fungal species (Debets et al. 1994;Liu and Milgroom 1996;van Diepeningen et al. 1997;Debets and Griffiths 1998;Zhang et al. 2014;Wu et al. 2017), and which has been postulated to be a type of fungal innate immunity (Paoletti and Saupe 2009;Daskalov and Glass 2022;Gaspar and Pawlowska 2022).In N. crassa, there are at least 14 allorecognition loci that result in cell death following somatic cell fusion (Zhao et al. 2015), thus potentially generating >2 million incompatible genotypes in segregating populations (Goncalves and Glass 2020).Here, we show that RCD mediated by genetic differences at the mat and rcd-1 loci not only causes death of fusion compartments but results in a winner-takes-all phenotype in pairings of incompatible strains.A study assessing fitness effects in N. crassa showed a similar trend of competitive exclusion in pairings between opposite mating type strains (Kronholm et al. 2020).In addition, we show that the cell-death inducing trigger mediates this competitive advantage, as mutations at a locus (tol) required for mating-type mediated cell death abolished the winner-takes-all phenotype.In mating type incompatible pairings, the mat a strain often won (Figs.3c, 5c and 7c).It is unclear what the basis of the mat a strain's capacity to triumph over the mat A strain in these pairings; a similar dominance was also observed for the rcd-1-2 strain in pairings with rcd-1-1 (rcd-1-2 mat A).It is possible that slight differences in germination rate, germling/hyphal fusion frequencies, or the strength of the death response in compartments surrounding the fusion cell could play a role.Interestingly, the bias of mat a over mat A was not observed in the Δso mat A + Δso mat a pairings (Fig. 8).It would be of interest to determine if the bias of mat a strains over mat A strains (and rcd-1-2 over rcd-1-1) in competitive pairings also occurs under conditions more akin to those found in nature.Nevertheless, in population samples, the frequencies of mat A, mat a, rcd-1-1, and rcd-1-2 alleles are nearly equal (Daskalov et al. 2019).
To further test whether fusion-mediated cell death or a block in cell fusion was essential for the winner-takes-all phenotype, we tested pairings between strains that were incompatible at the cwr-1 and cwr-2 loci, which significantly reduces somatic cell fusion frequencies, but where cell death is not triggered post-fusion; these pairings showed significant proportions of heterokaryotic spores (Fig. 7).These data indicate that HI resulting in cell death has a here-to-fore underappreciated role in competition for resources between genetically incompatible strains and which may also be involved in contributing to the evolution of allelic diversity observed in these systems (Saupe 2000;Muirhead et al. 2002;Goncalves and Glass 2020).
The characterized het loci in filamentous fungi do not affect sexual fertility; indeed, N. crassa is an obligate outbreeder and crosses with strains that are genetically different at many het loci.Post-fusion incompatible interactions in outbreeding species mediated by genetic differences at het loci must somehow be suppressed during sexual reproduction, as cell fusion and proliferation of opposite mating type nuclei in a common cytoplasm are a prerequisite for karyogamy and subsequent meiosis (Raju 1980;Berteaux-Lecellier et al. 1998).HI mediated by opposite mating types during vegetative growth is not unique to Neurospora species, but has also been reported in distantly related species, such as Ascobolus stercorarius (Bistis 1994).In the pseudohomothallic species, Podospora anserina, in which ascospores are dikaryotic, genetic differences at het loci can result in meiotic drive and ascospore abortion (Dalstra et al. 2003;Debets et al. 2012) as well as reproductive isolation (Ament-Velásquez et al. 2022).
Previous studies have investigated the existence of "nucleuslimited" genes in several fungal model systems, where a nucleus containing the WT allele of a gene cannot complement a null mutant allele when paired in a heterokaryon (Newmeyer 1970;Demeter et al. 2000;Czaja et al. 2013;Kasbekar 2014).Here, we show that the Δtol mutation suppressed mating type incompatibility in heterokaryons only where both partners lacked a functioning copy of tol.If one partner was tol+, regardless of mating type, this was sufficient to complement the null mutation and induce mating type incompatibility, thereby recapitulating the "winner-takes-all" phenotype (Fig. 5).These data support the hypothesis that TOL is not a nucleus-limited protein.
The soft mutants, which are defective in chemotropic interactions and somatic cell fusion and are blocked in cell death in mating type incompatible spore pairings, also showed a similar dominance of one nuclear genotype in spore populations.The severely reduced proportion of heterokaryotic spores in mating type compatible and mating type incompatible Δso + Δso pairings suggests that competition occurred predominantly between homokaryotic Δso colonies.In WT + Δso spore pairings, the dominant nuclear genotype was consistently WT.The growth advantage of WT over Δso (Supplementary Fig. 3) and the ability to undergo anastomosis resulting in more rapid spore production are presumably the cause of WT consistently prevailing over Δso strains.In a number of filamentous fungal species, such as in the human pathogen A. fumigatus, somatic cell fusion is rare (Macdonald et al. 2019), thus, these strains would mimic Δso mutants in nature.We predict that pairings between fusion deficient A. fumigatus strains would also show a winner-takes-all phenotype that is likely independent of genetic differences at het loci.The interplay in the regulation of somatic cell fusion vs post-fusion allorecognition likely impacts life history evolution in filamentous ascomycete species.
Here, we show that N. crassa syncytia are highly permissive to replication and maintenance of mutant nuclei, allowing these nucleotypes to compete for access to asexual spores.One interesting finding was that Δarg-5 homokaryotic spores had a higher representation than WT in syncytia when arginine was provided, even though Δarg-5 strains with supplementation had a slightly lower growth rate than WT (Supplementary Fig. 3).However, when a strain carrying a different arg mutation (Δarg-12) was paired with WT on supplemented media, we did not see a similar increase in Δarg-12 homokaryotic spores, rather an increase in mixed spores, reflective of an increase in the Δarg-12 nucleotype.In N. crassa, arginine is largely sequestered to the vacuoles, which regulates the levels of cytosolic arginine (Weiss and Davis 1977;Davis 1986).Vacuolar distribution in syncytia follows similar heterogeneous patterns of localization as nuclei (Bowman et al. 2009(Bowman et al. , 2015)), where cytoplasmic flow and microfluidic eddies result in uneven distribution of these organelles throughout the mycelial network (Pieuchot et al. 2015).Our data suggests that utilization of resources within heterokaryotic syncytia is complex.Localization of nuclei, as well as whether syncytial public goods are accessed cytoplasmically, within organelles (e.g.vacuoles, the endoplasmic reticulum, and mitochondria), or even extracellularly by leakage or secretion, may affect both competition and sharing of resources within a syncytia.

Fig. 1 .
Fig. 1.Flow cytometry analysis of prototrophic and auxotrophic spore pairings with and without supplementation.a) Schematic of race tubes labeling the media (blue) with no supplementation or arginine + pantothenic acid supplementation, along with the inoculation site (magenta circle).b) Inoculum ratios of spore pairings used in this race tube experiment.c) his-3::hH1-mCherry (R) mat A + Δpan-2 his-3::hH1-eGFP (G) mat A (74DM + 11U).d) mat A (G) + Δarg-5 (R) mat A (10BI + 102CC).e) Δarg-5 (R) mat A + Δpan-2 (G) mat A (102CC + 11U) spore pairings inoculated at approximately equal ratios and grown in race tubes for 7 dpi at 3°C with no media supplementation; or f-h) with arginine + pantothenic acid supplementation.Flow cytometry was conducted to analyze the relative percentage of each fluorescently tagged nuclear genotype in asexual spore populations derived from syncytia.Spore pairings in each biological replicate were derived from a single inoculum sample, and the ratios of each partner in the inoculum are shown in the "Inoculum" graph (panel b)."Start" denotes that spores were collected from the opening of the race tube most proximal to the inoculation site on day 7, and "End" denotes that the spore sample was collected from the portion of the race tube most distal from the site of the inoculation on day 7. Left bars (magenta) in graphs denote fluorescence signal from homokaryotic spores with the presence of only histone H1-mCherry-tagged nuclei, right (green) bars (panel b) or middle (green) bars (panels c-h) in graphs denote fluorescence signal from homokaryotic spores with the presence of only histone H1-eGFP-tagged nuclei, and black bars in graphs denote spores with signal from both histone H1-mCherry and histone H1-GFP nuclear tags in the same cell.Statistical analysis was conducted using two-way ANOVA followed by Tukey's honestly significant difference (HSD) post hoc tests.*P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001; ns = not significant.N = 3. Error bars = SEM.

Fig. 2 .
Fig. 2. Flow cytometry analysis of spore pairings with nuclear spacing mutants.a) Normal nuclear spacing and morphology of the his-3::hH1-eGFP mat A (10BI) germlings.Adjacent cells in panel a without GFP fluorescence are his-3::hH1-mCherry mat A (74DM) germlings.b) Nuclear spacing defects of Δro-3, his-3::hH1-eGFP mat A (71AO) germlings.c) Nuclear spacing defects of Δro-10, his-hH1-eGFP mat A (72AO) germlings grown from spores at 3°C for approximately 6 hpi in liquid VMM on coverslips.Scale bar = 10 μm.d) Fluorescence signals (%) of each inoculum sample for every spore pairings used in this flow cytometry experiment (left side), and fluorescence signals (%) of monocultures from each strain used in this experiment, as well as an untagged mat A strain (FGSC 2489) (right side).e) his-3::hH1-eGFP (G) mat A + his-3::hH1-mCherry (R) mat A (10BI + 74DM), Δro-3 (G) mat A + (R) mat A (71AO + 74DM), and Δro-10 (G) mat A + (R) mat A (72AO + 74DM) spore pairings were grown in race tubes for 7 dpi at 3°C.Flow cytometry was conducted to analyze the relative percentage of each fluorescently tagged nuclear genotype in asexual spore populations derived from syncytia.Spore pairings in each biological replicate were derived from a single inoculum sample, and the ratios of each partner in the inoculum are shown in the "Inoculum" graph (panel d)."Start" denotes that spores were collected from the opening of the race tube most proximal to the inoculation site on day 7, and "End" denotes that the spore sample was collected from the portion of the race tube most distal from the site of the inoculation on day 7. Left (magenta) bars in graphs denote fluorescence signal from homokaryotic spores with the presence of only histone H1-mCherry-tagged nuclei, right (panel d) or middle (panel e) (green) bars in graphs denote fluorescence signal from homokaryotic spores with the presence of only histone H1-eGFP-tagged nuclei, black bars in graphs denote spores with signal from both histone H1-mCherry and histone H1-GFP nuclear tags in the same cell, and gray bars denote spores with no detectable fluorescence signal.Black and white arrows highlight the abnormal nuclear spacing in ropy mutants (panels b and c).Statistical analysis was conducted using two-way ANOVA followed by Tukey's HSD post hoc tests.*P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001; ns = not significant.N = 4. Error bars = SEM.Pairwise post hoc t-tests between heterokaryotic spore populations of hH1-eGFP (G) mat A, + hH1-mCherry (R) mat A (10BI + 74DM) from the "END" and Δro-3 (G) mat A + (R) mat A (71AO + 74DM) from the "END" were significantly different (P = 0.0003); Pairwise post hoc t-tests between heterokaryotic spore populations of hH1-eGFP (G) mat A, + hH1-mCherry (R) mat A (10BI + 74DM) from the "END", and Δro-10 (G) mat A + (R) mat A (72AO + 74DM) from "END" were significantly different (P = 0.0135).

Fig. 3 .
Fig. 3. Individual flow cytometry runs of identical or opposite mating type spore pairings.Mating type compatible his-3::hH1-eGFP (G) mat A + his-3:: hH1-mCherry (R) mat A spore pairings (10BI + 74DM) (left side), mating type compatible (G) mat a + (R) mat a spore pairings (8BH + 74ED) (middle), or mating type incompatible (G) mat A + (R) mat a (10BI + 74ED) spore pairings (right side) were inoculated in approximately equal ratios and grown in race tubes for 7 dpi at 3°C.Flow cytometry was conducted to analyze the relative percentage of each fluorescently tagged nuclear genotype in asexual spore populations derived from syncytia.All four biological replicates for each spore pairing are shown individually to clearly present which genotype nearly reached saturation in each race tube trial of the experiment.Spore pairings in each biological replicate were derived from a single inoculum sample, and the ratios of each partner in the inoculum are shown in the left-hand "Compatible Inoculum" and "Incompatible Inoculum" columns of each dataset in the graph."Start" denotes that spores were collected from the opening of the race tube most proximal to the inoculation site on day 7, and "End" denotes that the spore sample was collected from the portion of the race tube most distal from the site of the inoculation on day 7. Left (magenta) bars in graphs denote fluorescence signal from homokaryotic spores with the presence of only histone H1-mCherry-tagged nuclei, right (inoculum) or middle (green) bars in graphs denote fluorescence signal from homokaryotic spores with the presence of only histone H1-eGFP-tagged nuclei, and black bars in graphs denote spores with signal from both histone H1-mCherry and histone H1-GFP nuclear tags in the same cell.
Fig.7.Flow cytometry analysis of spore pairings with mating type compatible and incompatible cwr-1 HG1 , cwr-1 HG2 , and cwr-1 HG6 in pairings with a cwr-1 HG1 cwr-2 HG2 strain.(a) Inoculum ratios for all spore pairings used in this flow cytometry experiment.b) Mating type compatible spore pairings cwr-1 HG1 cwr-2 HG1 his-3::hH1-mCherry (R) mat A (74DM) with cwr-1 HG1 Δcwr-2 his-3::eGFP (G) mat A (336a) or cwr-1 HG2 Δcwr-2 (G) mat A (88a) or cwr-1 HG6 Δcwr-2 (G) mat A (385a) strains.c) Individual replicates of the mating type incompatible spore pairings of cwr-1 HG1 cwr-2 HG1 (R) mat a (74ED) + cwr-1 HG1 Δcwr-2 (G) mat A (336a) or cwr-1 HG2 Δcwr-2 (G) mat A (88a), or cwr-1 HG6 Δcwr-2 (G) mat A (385a).Each strain was inoculated in approximately equal ratios and grown in race tubes for 7 dpi at 3°C.Flow cytometry was conducted to analyze the relative percentage of each fluorescently tagged nuclear genotype in asexual spore populations derived from syncytia.Spore pairings in each biological replicate were derived from a single inoculum sample."Start" denotes that spores were collected from the opening of the race tube most proximal to the inoculation site on day 7, and "End" denotes that the spore sample was collected from the portion of the race tube most distal from the site of the inoculation on day 7. Left (magenta) bars in graphs denote fluorescence signal from homokaryotic spores with the presence of only histone H1-mCherry-tagged nuclei, right (panel a) or middle (panels b and c) (green) bars in graphs denote fluorescence signal from homokaryotic spores with the presence of only histone H1-eGFP-tagged nuclei, and black bars in graphs denote spores with signal from both histone H1-mCherry and histone H1-GFP nuclear tags in the same cell.All six biological replicates for each mating type incompatible spore pairing are shown individually to clearly present which genotype nearly reached saturation in each race tube trial of the experiment.CWR haplogroup 1 (HG1); haplogroup 2 (HG2); and haplogroup 6 (HG6) naming from Gonçalves et al. (2019) and Detomasi et al. (2022).Statistical analysis was conducted using two-way ANOVA followed by Tukey's HSD post hoc tests.*P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001; ns = not significant.N = 6.Error bars = SEM.

Table 2 .
Strains used in this study are available at the Fungal Genetics Stock Center (https://www. fgsc.net/); FGSC numbers are listed in Supplementary Table 1.