Triangularia anserina

THE P. ANSERINA View in CoL SPECIES COMPLEX

Strains were recovered from dung as described ( Silar 2020). To isolate Sordariales strains from soils, M0 plates ( 0.25 g /L KH 2 PO 4, 0.3 g /L K 2 HPO 4, 0.25 g /L MgSO 4 -7H 2 O, 0.5 g /L urea, 0.05 mg /L thiamine, 0.25 µg/L biotin, 2.5 mg /L citric acid, 2.5 mg /L ZnSO 4, 0.5 mg /L CuSO 4, 125 µg/L MnSO 4, 25 µg/L boric acid, 25 µg/L sodium molybdate, 25 µg/L iron alum, 10 g /L agar) were supplemented with 0.5 g of shredded miscanthus by adding the complex biomass onto the plates after the agar rigidified. These were then inoculated with c. 0.5 mL of soil. For each soil sample, three plates were prepared. One was left at room temperature, a second one was incubated at 65°C for 30 minutes and then transferred to room temperature; the third one was incubated overnight at 37°C and then transferred to room temperature. The three plates were then left at room temperature in the presence of light for up to two months. Note that the plates were not treated with antibiotics or chemicals to prevent the growth of bacteria or the presence of small animals (such as mites and collembolans), because the stress brought by their presence and/or grazing may promote fruiting body development of Sordariales (Silar, unpublished observations). The plates were regularly checked (once or twice weekly) for the presence of perithecia typical of Sordariales fungi. If such fruiting bodies developed, ascospores ejected from them were collected onto projection plates as described previously ( Silar 2020), except that the plates were supplemented with three antibiotics (chloramphenicol 25 µg/mL, tetracycline 50 µg/mL and kanamycin 50 µg/mL) to prevent bacterial contaminations. Ascospores were then transferred onto G medium (ammonium acetate 4.4 g /L, bactopeptone 15 g /L and agar 13 g /L) supplemented with 5 g /L of yeast extract and three antibiotics (chloramphenicol 25 µg/mL, tetracycline 50 µg/mL and kanamycin 50 µg/mL). A 30-minutes heat-shock at 65°C was immediately applied to promote ascospore germination. If this failed, an overnight heat-shock at 37°C was then applied to a second batch of freshly collected ascospores. Note that in the case of strains of the P. anserina species complex, the presence of yeast extract in G medium and heat-shock are dispensable, while it is often mandatory for most other species of Sordariales . For species the P. anserina complex, the four ascospores of several asci were collected separately; when possible, one ascus with four germinated ascospores was then selected, and the four thalli obtained from the four ascospores of this F0 generation were stored separately at –80°C ( Silar 2020).

DNA was extracted using a quick method adapted from plant ( Bellstedt et al. 2010) and yeast ( Liu et al. 2011) DNA extraction protocol. A 5 mm × 5 mm × 5 mm plug of agar taken from a plate onto which the fungus grew was put in a 1.5 mL Eppendorf tube along with 200 µL of a DNA extraction buffer that was made as follows. Firstly, 8 mL of 0.2 M stock solution of anhydrous sodium in water was freshly mixed with 17 mL of 0.2 M stock solution of sodium bicarbonate in water into 100 mL final volume of sterile water to make a 0.05 M carbonate buffer with pH= 9.6. The extraction buffer was then produced by mixing 96 mL of the carbonate/bicarbonate solution with 4 mL of 0.5 M of polyvinylpyrrolidone, 200 mg of bovine serum albumin fraction V (Sigma-Aldrich cat#A4503) and 50 µL of tween 20. The mycelium plug was then broken at speed 4.0 for 20 s in a TeSeE Precess 24 (Bio-Rad, Hercules, CA, United States), and the tubes were incubated at 95°C for 15 minutes. The tube was gently shaken with a finger and transferred on melting ice for one minute. The tube was then vortexed for 10 seconds and centrifuged at c. 20 000 g for 10-20 seconds in an Eppendorf centrifuge machine. The ITS (Intergenic Transcribed Spacer from the rDNA cluster) was amplified with the ITS5 and ITS4 or ITS1 and ITS5 primer pairs directly on 5 µL of supernatant. The PCR products were sent to Genewiz from AZENTA (Takeley, UK) for sequencing with ITS1 (or ITS5) and ITS4. Note that for Sordariales, ITS 5 and ITS4 amplifications and sequencing with ITS5 gave more reliable results than ITS1 and ITS4 amplifications and ITS1 sequencing, respectively.

G ENOME SEQUENCING OF STRAINS FROM THE P. ANSERINA SPECIES COMPLEX.

We sequenced the genome of mat1-1 (aka mat+) and mat1-2 (aka mat-) homokaryotic isolates.To obtain them, a self-fertile dikaryotic F 0 thallus was used to produce homokaryotic selfsterile F 1 progenies. These were confronted to each other to determine their mating compatibility, and the DNA from two sexually compatible progenies were extracted using the NucleoSpin® Soil from Machery Nagel (Düren, Germany) and submitted to 2×150 bp Novaseq illumina sequencing by Novogene (Cambridge, United Kingdom). The resulting FastQ files were then assembled using Unicycler ( Wick et al. 2017), as previously described for other strains of the P. anserina species complex ( Lalanne & Silar 2025). Mining the genomes with BLAST and manual annotation of the mating type loci allowed to determine whether the isolates were mat1-1 or mat1-2.

ANI CALCULATION

ANIs were calculated with FungANI ( Lalanne & Silar 2025) using the default parameters on the mat1-2 genome assemblies.

D ETERMINATION OF THE REGION LACKING RECOMBINATION

To compute synonymous divergence (dS) values between the mat1-1 and mat1-2 genomes of PSN 1158, we first performed SNP calling against the P. anserina S mat+ genome ( Grognet et al. 2014) as described previously ( Hartmann et al. 2021). For SNP calling, we used the version of the P. anserina S mat+ assembly available from the Joint Genome Institute MycoCosm website (https://mycocosm.jgi.doe.gov/mycocosm/home, last accessed November 15, 2018) as “Podan2” ( Espagne et al. 2008; Grognet et al. 2014) and annotations improved by ( Vogan et al. 2019), available from the GitHub repository https://github.com/johannessonlab/SpokBlock-Paper (last accessed June 1, 2020). Briefly, we mapped Illumina trimmed reads against the P. anserina S mat+ genome assembly with the bowtie2 v2.3.4.1 program ( Langmead et al. 2009), with the following software options: – very-sensitivelocal – phred33-X 1000. To remove PCR duplicates, we used the MarkDuplicates tool of Picard tools version 1.88 (http://broadinstitute.github.io/picard, last accessed October 30, 2019). We used the RealignerTargetCreator and Indel-Realigner tools of the Genome Analysis Toolkit (GATK) to locally realign the mapped reads and improve alignment accuracy in indel regions. To perform SNP calling, we run the HaplotypeCaller tool of GATK version 3.7 ( McKenna et al. 2010) in the haploid mode on each genome individually. Then, we performed joint variant calls with GenotypeGVCFs on a merged gvcf variant file. SNP calls were filtered for quality with VariantFiltration, in accordance with GATK Good Practice for the hard filtering of variants (QUAL<250; QD<2; MQ<30.0; – 10.5>MQRankSum>10.5; –5>ReadPosRank-Sum>5; FS>60; SOR>3). We performed additional filtering steps with vcftools to retain only biallelic and polymorphic SNPs. For each genome, we used a customized script (available on request) to build pseudogenome sequences, replacing the nucleotide bases of the P. anserina S mat+ genome with the base pair ID present in the focal genome as inferred from SNP calling and we retrieved gene coding sequences based on P. anserina S mat+ gene models with the gffread program (available from https://ccb.jhu.edu/software/stringtie/gff. shtml#gffread, last accessed April 15, 2020). We performed pairwise sequence comparisons using the codon-based approach implemented in translatorX, with default parameters ( Abascal et al. 2010) and used the nucleotide alignment as input for the yn00 program in the PAML package, to calculate dS values ( Yang & Nielsen 2000; Yang 2007).

M ORPHOLOGICAL AND PHYSIOLOGICAL ANALYSES

All strains, including those of the new species PSN 1158, grew and produced sexual fruiting bodies as previously described for species of the P. anserina complex ( Silar 2020). Morphological and physiological analysis of PSN 1158, including the ability to produce microsclerotic-like and appressorium-like structures as well as to display Hyphal Interference towards Penicillium chrysogenum , was performed as previously described ( Boucher et al. 2017). For PSN 1158, ascospore spore head and spermatia measurements were made on 50 ascospores and 50 spermatia, respectively. Perithecia were measured on 10 fruiting bodies and ascospore primary and secondary appendages on 15 ascospores. The perithecia analyzed for their peridium and centrum were those obtained on the M 2 medium. Comparisons of fruiting body production and morphology of PSN 1158 and P. comata T were made in triplicate with the same batches of media and in parallel to avoid the influence of medium and growth condition differences on fruiting body production and repartition, as well as on ascospore morphology.

S TRAIN AVAILABILITY

The F 0 heterokaryotic isolates for all strains listed in Table 1 View TABLE and their F 1 homokaryotic progenies whose genomes were sequenced ( Table 2 View TABLE ) can be obtained upon request to the corresponding author. The type specimen of P. reunionensis Silar , sp. nov. PSN 1158 was deposited in the Herbarium of the Museum national d’Histoire naturelle (MNHN, Paris, France; PC), and an ex-type living heterokaryotic F 1 culture of PSN 1158 was deposited in the “ Centre International de Ressources Microbiennes-Champignons Filamenteux” ( CIRM-CF, Inrae, France).

RESULTS AND DISCUSSION

I DENTIFICATION OF STRAINS OF THE P. ANSERINA SPECIES COMPLEX AND SPECIES DISTRIBUTION RANGES

To understand the geographical distribution in France of the different species of the P. anserina complex, we isolated new strains not only dung but also from soil. For the latter substrate, a new method that proved efficient in isolating Sordariales fungi was designed (see Material and Methods). Among the isolates gathered from the soil, some corresponded to P. anserina sensu lato ( s.l.) (i.e., belonged to the P. anserina species complex). Dung and soil samples were collected from different regions in metropolitan and overseas France for several years, and 86 isolates with morphologies corresponding to P. anserina s.l. were recovered ( Table 1 View TABLE ). To identify the species to which they belonged, we first sequenced their ITS barcodes and compared them with those of the seven species of the complex ( Boucher et al. 2017). We could identify five P. comata, ten P. pauciseta , two P. pseudoanserina and 67 P. anserina s.s. strains. Another strain, PSN 1303, presented a difference with the reference P. anserina s.s. ITS from strain S at position 460 where an additional C was present in PSN 1303. The genome sequence of PSN 1303 (see below) showed that this strain also belonged to P. anserina s.s., showing that two different ITS barcode sequences exist for this species. In total, 68 P. anserina s.s. strains were thus isolated. Another strain, PSN 1158, had one difference with the P.comata ITS sequence, having a G instead of an A at position 47 and two differences with the reference ITS sequence of the P. anserina strain S, with an A instead of G at position 22 and the presence of two additional Cs after position 467. Genome sequence and morphological analyses of PSN 1158 (see below) showed that it belonged to a species new to science.

As seen in Figure 1 View FIG , the different species had distinct distribution ranges in metropolitan and overseas France. Podospora anserina s.s. is found all over metropolitan France, including Corsica, while P. comata was found restricted so far to the Northern part of continental France and P. pauciseta to the southern part. P. comata has been isolated from the Netherlands ( Ament-Velásquez et al. 2024), suggesting that this species may prefer colder climates. This is confirmed by our own isolation of a strain from Britany and one from the Alps mountains, both of which are regions colder than the rest of France. On the contrary, P. pauciseta may prefer hotter areas, as it seems fairly common near the Mediterranean Sea; note that we also found this species in Greece (Silar, unpublished), confirming its preference for the southern part of Europe.

In overseas France, we found members of the species complex only in Guadeloupe, that is located in the Caribbean’s near Martinique, and in La Réunion island, located in the Indian ocean near Madagascar, although we also analyzed samples from Martinique (but not from Mayotte or Guyane). Guadeloupe hosts P. pseudoanserina , and La Réunion a new species, here described as P. reunionensis Silar , sp. nov.

G ENOME SEQUENCING FOR

ACCURATE SPECIES IDENTIFICATION

To clarify which species some strains actually belonged to, we sequenced the genomes of mat1-1 and mat1-2 homokaryotic isolates for five strains. These were: 1) PSN 1303 and PSN 1158 because of their differences with the previously known ITS barcodes; 2) PSN 1705 because it originated from Corsica, an island distant from continental France; and 3) PSN 1871 and PSN 1899, because the two strains presented a different mycelium morphology and genome sequences were available for only two strains of P. pseudoanserina . The main features of genome assemblies are given inTable 2. All genomes were around 35 MB, a size similar to the genomes of strains previously sequenced ( Espagne et al. 2008; Grognet et al. 2014; Silar et al. 2019; Ament-Velásquez et al. 2024) and moderately fragmented (505 to 999 contigs).

The FungANI analysis confirmed that PSN 1303 and PSN 1705 belonged to P. anserina s.s., because the genomes of these two strains displayed high similarity to the one of the type P. anserina strain s.s. S mat+ (ANI> 99.5%; Fig. 2 View FIG ). In addition, their morphology, especially the repartition of the perithecia on the M 2 medium as a ring onto the mycelium was typical of P. anserina s.s. (( Boucher et al. 2017); Fig. 3 View FIG ). The FungANI analysis also confirmed that PSN 1871 and PSN 1899 belonged to P. pseudoanserina (ANI> 99.5% with the P. pseudoanserina type strain CBS253.71; Fig. 2 View FIG ); their morphology further corresponded to that of P.pseudoanserina , perithecia being formed on M 2 as a disk in the center of the mycelium ( Boucher et al. 2017; Fig. 3 View FIG ). Note that PSN 1871 and PSN 1899 displayed different mycelium morphologies, PSN 1899 producing more aerial hyphae.

FungANI analyses of PSN 1158 showed that it did not correspond to any of the seven previously known species ( Fig. 4 View FIG ). Indeed, the PSN 1158 genome was only about 97% identical to those of the other species, except P. comata T, with which it had 99.17% identity ( Fig. 4 View FIG ). PSN 1158 and P. comata strain T genomes share 48% of sequences with similarity between 99.5% and 99.9% (magenta bar on the FungANI graphic comparing PSN 1158 with T ofFig. 4), 25% with 99.0% and 99.5% similarity (dark blue bar on the same graphic) and 11% with 98.0% and 99.0% (light blue bar on the graphic). They, however, had only about 6% of highly similar sequences (i.e., genomic region with a percentage identity> =99.9% as calculated by FungANI; Fig. 4 View FIG ), suggesting little recent gene flow between these lineages. Importantly, although ANI showed that PSN 1158 was closely related to the P. comata strain T, they displayed contrasting cultural features (see below) and ascospore morphology, indicating that PSN 1158 belong to a new species ( Figs 5 View FIG ; 6 View FIG ).

On M 2, PSN 1158 produced a mycelium pigmented in dark green with a center nearly colorless, while P. comata formed a mycelium with a more homogeneous pigmentation (( Boucher et al. 2017); Fig. 5 View FIG ). PSN 1158 perithecia were produced along a ring (wider than the one produced by P. anserina ), while P. comata produced perithecia in a disk in the center of the colony (( Boucher et al. 2017); Fig. 5 View FIG ). Additionally, P. comata was particularly fertile on M 0 supplemented with Guibourtia demeusii L. wood shavings ( Boucher et al. 2017), while PSN 1158 sexual reproduction was delayed by three days on this medium and produced much fewer fruiting bodies ( Fig. 6 View FIG ). It however rapidly produced perithecia on M 0+ miscanthus like all the species of the P. anserina complex ( Fig. 5 View FIG ). The ascospores produced by PSN 1158 had clear morphological differences with those produced by P. comata strain T, which is the ex-type for this species ( Fig. 6 View FIG ). The ascospore spore head was plumpier (35.0+/–1.8×22.7 +/–0.8 µm for PSN 1158 vs 35.7 +/– 4.1 × 17.8 +/– 1.5 µm for P. comata T), presented a longer primary appendage (32.3 +/– 2.5 µm for PSN 1158 vs 26.3 +/– 2.8 µm for P. comata T) and a much more prominent secondary appendage(s) at the junction between the spore heard and the primary appendage.

Another argument supporting that PSN 1158 belonged to a new species was that, as previously stated, P. comata appears to like colder climates, and PSN 1158 came from a humid tropical island. Finally, crosses between P. comata T and PSN 1158 were much less fertile than PSN 1158 × PSN 1158 and T × T selfing crosses. Indeed, only three fruiting bodies, including only one bearing ascospores, were obtained in conditions where hundreds of spore-bearing perithecia were produced for intra-species crosses.