Significant hybridization between the invasive North American fungal plant pathogen Heterobasidion irregulare and its Eurasian sister species H. annosum is ongoing in Italy. Whole genomes of nine natural hybrids were sequenced, assembled and compared with those of three genotypes each of the two parental species. Genetic relationships among hybrids and their level of admixture were determined. A multi-approach pipeline was used to assign introgressed genomic blocks to each of the two species. Alleles that introgressed from H. irregulare to H. annosum were associated with pathways putatively related to saprobic processes, while alleles that introgressed from the native to the invasive species were mainly linked to gene regulation. There was no overlap of allele categories introgressed in the two directions. Phenotypic experiments documented a fitness increase in H. annosum genotypes characterized by introgression of alleles from the invasive species, supporting the hypothesis that hybridization results in putatively adaptive introgression. Conversely, introgression from the native into the exotic species appeared to be driven by selection on genes favoring genome stability. Since the introgression of specific alleles from the exotic H. irregulare into the native H. annosum increased the invasiveness of the latter species, we propose that two invasions may be co-occurring: the first one by genotypes of the exotic species, and the second one by alleles belonging to the exotic species. Given that H. irregulare represents a threat to European forests, monitoring programs need to track not only exotic genotypes in native forest stands, but also exotic alleles introgressed in native genotypes.

Biological invasions, fungi, Heterobasidion, hybridization

Introgression, i.e. the exchange of genetic material between interfertile species through hybridization, is recognized as a significant catalyst of evolution of eukaryotes (Arnold et al. 2008; Edelman et al. 2019). Thanks to increasing world trade, to advances in genetic characterization and to the democratization of genomic research technologies, the frequency and estimates of hybridization events across all phyla have risen exponentially in recent years (Rieseberg et al. 2000; Barton 2001; Seehausen 2004; Mallet 2005; Le Gac and Giraud 2008; Dasmahapatra et al. 2012; Clarkson et al. 2014).

Although the interspecific transfer of genetic material can be a stochastic process not necessarily involving genes responsible for phenotypic adaptive variation (Suarez-Gonzalez et al. 2018), there are many cases in which introgressive events have been documented to provide novel genes and alleles or to generate favorable allelic combinations that confer adaptive advantages to the recipient organism (Suarez-Gonzalez et al. 2018). Introgression may also increase the fitness of hybrids, allowing them to be competitive against parental species, particularly in novel ecological niches (Mallet 2007), thus facilitating successful invasions (Schierenbeck and Ellstrand 2009; Lee 2002). In plants, it has been well documented that hybridization may result in evolutionary changes increasing invasiveness (Schierenbeck and Ellstrand 2009; Abbott et al. 2003). It is also recognized that hybridization can lead to the emergence of new pathogenic fungal species (Brasier 2000, 2001; Fisher et al. 2012; Giraud et al. 2008; Gladieux et al. 2014; Stuckenbrock 2016).

Stukenbrock (2016) suggested that studies integrating genomic and experimental data are pivotal to understand the evolution of fungal hybrids in nature. Whereas fungal hybrid genotypes can be generated and studied in vitro (Olson and Stenlid 2001; Giordano et al. 2018), hybrids resulting from natural selection are difficult to find in nature, given that only a few cases of ongoing hybridization are known (Garbelotto et al. 1996; Gonthier and Garbelotto 2011; Hughes et al. 2013; Pryszcz et al. 2014; Menardo et al. 2016; Sillo et al. 2019).

The fungus Heterobasidion irregulare Garbel. & Otrosina is a pathogen of pines in North America that was accidentally introduced into central Italy in 1944, during World War II (Gonthier et al. 2004). The pathogen has become invasive, (Gonthier et al. 2014), and is now included in the A2 list of organisms recommended for regulation by the European and Mediterranean Plant Protection Organization (EPPO 2015). In Central Italy, where H. irregulare has been reported so far, its Eurasian sister species H. annosum (Fr.) Bref. is also present; however the invasive exotic species is significantly more abundant than the native one (Gonthier et al. 2007). The invasiveness of the exotic species has been experimentally shown to depend not on pathogenicity, which is similar between the two species (Garbelotto et al. 2010), but on its greater ability to saprobically colonize wood and to produce more fruiting bodies, thus increasing its sporulation rate compared to the native sister species (Giordano et al. 2014). Primary infection by Heterobasidion is indeed effected by basidiospores colonizing exposed woody surfaces of stumps and wind-thrown trees (Garbelotto and Gonthier 2013), hence an enhanced ability to saprobically colonize wood would confer significant advantages to individuals carrying them. Genomic analyses have corroborated the above hypothesis by showing that genes involved in saprobic growth are significantly more divergent between the two species than expected by genetic drift, while, on the contrary, genes involved in pathogenicity appear to be rather conserved between the two, having undergone purifying selection (Sillo et al. 2015). A recent phenotypic and transcriptomic characterization of artificial Heterobasidion hybrids has also determined that H. irregulare nuclear genes related to nucleus-mitochondrion communication may confer an advantage to hybrid genotypes compared to H. annosum genes involved in the same function (Giordano et al. 2018).

The two species have evolved allopatrically for 34–41 millions of years, but have remained interfertile (Garbelotto and Gonthier 2013). Their recent sympatry has resulted in one of the few known current examples of fungal hybrid swarms (Gonthier and Garbelotto 2011). A study of over 260 genotypes from the hybrid zone in Italy using AFLP and multi gene phylogenies has indicated that, depending on site, up to 42% of the genotypes had an admixed genome, with levels of genomic admixing varying between 5% and 45% (Gonthier and Garbelotto 2011). Sequencing of 11 loci has indicated that in 85% of cases, introgression involved alleles of one species being absorbed by the other, while intragenic recombination was detected only in 15% of cases (Gonthier and Garbelotto 2011). Thus, it is expected that the introgression of whole alleles, and not recombination, may be the dominant process reshaping genomes through hybridization between Heterobasidion species. This study may confirm or disprove such expectations.

Two studies using different genetic markers (Gonthier and Garbelotto 2011; Garbelotto et al. 2013) have shown that there is no significant genetic structuring within the invasive and the native species (Gonthier et al. 2014, 2015). Thus, despite the relatively large area of the hybrid zone, it can be assumed that a single population of H. irregulare is hybridizing with a single population of H. annosum. The lack of population structure further simplifies assignment of alleles to either species, given that a genome-wide comparative study has determined that interspecific DNA divergence between the two taxa is high and ranges between 20.0 and 40.0 SNPs/Kbp, while intraspecific divergence within either species ranges only between 1.0 and 4.0 SNPs/Kbp (Sillo et al. 2015).

The model system represented by these interbreeding species with a comparable biology and epidemiology provides a unique opportunity to study the genomics of hybridization and introgression in a natural habitat. The main overarching goal of this study was to investigate a fungal invasion at the gene, rather than at the species level (Petit 2004). Theory predicts that the number of introgression events should be significantly larger from the native into the exotic species, if the latter is in demographic expansion (Currat et al. 2008; Excoffier et al. 2009). Despite its relatively recent introduction, the exotic fungus is overwhelmingly dominant and its geographic range is increasing, hence, it can be safely assumed that it is in demographic expansion. Our first prediction, thus, is that the number of chromosomal blocks introgressed from the native into the exotic species should be larger than the number of chromosomal blocks introgressed from the exotic into the native species.

Theory predicts that the introgression of alleles between species may be driven by the adaptive advantages such alleles may provide (Clarke et al. 2002; Fitzpatrick et al. 2010). In our case, invasiveness of the exotic species has been associated with genes involved in saprobic wood decay and in nucleus-mitochondrion communication, hence our second prediction is that alleles of genes involved in these two functions would be introgressed from the exotic into the native species, and that, vice versa alleles involved in these two functions would not be lost by H. irregulare when hybridizing with H. annosum.

Although we have yet to identify H. annosum alleles that may provide an advantage to H. irregulare genotypes acquiring them, we can hypothesize that alleles of regulatory genes evolutionarily adapted to the Italian landscape, or that genes putatively related to hybrid genome stability may be good candidates as alleles introgressing into H. irregulare. The latter group of alleles may enhance the survival of these hybrids by counteracting the effects of genomic instability caused by the large number of introgression events, as has been shown in other cases (e.g. in animals; Walsh et al. 2018),

Finally, our last prediction is that H. annosum genotypes that have absorbed H. irregulare alleles involved in saprobic decay and nucleus-mitochondrial communication should display an increased saprobic ability when grown on woody substrates. Conversely, given that virulence does not differentiate the two species, there should be no clear association between genomes of hybrids and this trait. Nonetheless, because all genotypes tested in this study were established in nature and thus naturally viable, we predict pathogenicity should be within the range displayed by pure parental genotypes. In addition, H. irregulare genotypes that have absorbed a large number of chromosomal blocks from H. annosum may display decreased saprobic growth, pathogenicity or both, due to the effects of large-scale structural genomic rearrangements caused by the large extent of genetic introgression from H. annosum.

As stated above, the main overarching goal of this study was to investigate a fungal invasion at the gene level, rather than at the species level (Petit 2004). Are exotic alleles introgressed in the native species, and do these alleles confer an adaptive advantage to genotypes of the native species that acquire them? Given that the exotic fungal species is a forest pathogen that has been identified as a serious threat to the European region and that a strict monitoring program has been recommended for the timely detection of the enlargement of its range (Gonthier et al. 2014; EPPO 2015), this gene centric approach may also indicate whether the monitoring should be expanded to include the detection of specific exotic H. irregulare alleles as they introgress in native H. annosum populations both within and outside the current zone of infestation by the exotic fungus.

The introgression of both exotic genotypes and exotic alleles in native populations, where exotic is defined here as pertaining to an exotic species, is knowingly associated with biological invasions. During interspecific hybridization, the interspecific introgression of alleles bears significant evolutionary implications, shifting the focus from the more traditional “species-centric” to a “gene-centric” concept of invasion (Crispo et al. 2011). In this study, we have shed light on the dynamics of hybridization-mediated allelic introgression between an invasive and a native fungal species. We performed the study by sampling pure and admixed genotypes in an area in Italy where the North American forest pathogen H. irregulare has been introduced and is hybridizing with the native Eurasian forest pathogen H. annosum (Gonthier et al. 2007; Gonthier and Garbelotto 2011). This system is particularly interesting as the precise date and location of introduction are known (Gonthier et al. 2004), the biology and population structure of pathogens are also known, the two species are fully interfertile, and 15 years of research have identified some of the likely mechanisms behind the clear demographic dominance of the exotic species compared to the native one.

For this study, we selected a set of 15 natural genotypes with levels of admixture ranging between zero (pure genotypes) and 45% (fully admixed) based on a large number of anonymous AFLP markers. The fact that all studied genotypes had successfully established themselves on wood in a natural setting suggested they were all sufficiently fit to survive in nature. Our expectation that these genotypes would display variably admixed genomic regions (Elgvin et al. 2017) and variable phenotypic fitness proved correct, allowing us to associate genomic variability and phenotypic fitness with specific introgression histories.

Phylogenomic analyses confirmed that the selected natural hybrids varied in the level of genomic admixture. Three hybrid genotypes (118NB, 150EA and 49OE) had a balanced level of admixture, suggesting either a recent hybridization event between two pure parental genotypes or balanced backcrosses with both parental species. The remaining six genotypes had signs of introgression in a relatively small portion of their genome, indicative of multiple backcrosses with a single parental species, to which they could be assigned. Some scaffolds (e.g. scaffolds 6, 12 and 14) showed significantly lower levels of introgression, suggesting these scaffolds contain genes regulating essential functions or species-specific genes (Olson et al. 2012).

The genomes of H. irregulare hybrids were profoundly affected by introgression from H. annosum, and only approximately 2 Mbp appeared to have been left untouched, while the remaining 30 Mbp were subject to allelic replacement. A significantly larger amount (about 7 Mbp) of the genome of H. annosum hybrids, instead, had been left untouched by the introgression of H. irregulare alleles, suggesting a smaller scale process in terms of genomic rearrangement. Regions not affected by alleles introgression might be subjected to intra locus recombination, however, in this study, these regions were estimated to be only approximately 24% and 7% of the genome in H. annosum and H. irregulare, respectively. The effects of intralocus recombination, despite their significance, are not treated in this work and these regions were excluded from the quantification of allelic introgression.

The size of blocks introgressed from the exotic H. irregulare into the native H. annosum was on average one order of magnitude larger than the size of those introgressed from H. annosum into H. irregulare. Small block dimensions (about 85 bp on average) and high number of genomic blocks (more than 14 000) introgressed into the exotic species may result in the excessive fragmentation of chromosomal blocks. As a result, exons may be partial, non-functional and may be more likely to be lost during recurrent recombination events (Sachdeva and Barton 2018). On the other hand, a low number (about 6 000) of larger blocks (on average 890 bp in size) introgressed from H. irregulare may suggest that putatively adaptive and functional alleles may be truly integrated and preserved into the recipient H. annosum genomes. It is known that large genomic blocks have a better chance to be maintained through multiple generations when compared to small blocks (Baird et al. 2003; Sachdeva and Barton 2018).

Alleles contained in the larger blocks introgressed from H. irregulare into H. annosum showed a GO term enrichment related to catabolic pathways, i.e. peroxidases, heme binding proteins, oxidoreductases, metal ion binding proteins and to vesicle trafficking. Oxidoreductases and heme peroxidase genes have been identified among those that best differentiate the highly saprobic H. irregulare from the less saprobic H. annosum (Sillo et al. 2015), and some of them, e.g. manganese peroxidases, are known to play a role in lignin degradation during saprobic colonization of wood by Heterobasidion species (Yakovlev et al. 2013). All experimental evidence has indicated that higher saprobic potential explains much of the invasiveness of H. irregulare (Giordano et al. 2014; Giordano et al. 2019). Another category of over-represented GO terms among H. irregulare alleles introgressed into H. annosum was related to mitochondrial transport. Interestingly, this is another category of genes that well differentiates the two species at the genomic level (Sillo et al. 2015), with H. irregulare nuclear genes better at regulating mitochondrial functions in admixed genomes than H. annosum nuclear genes (Sillo et al. 2015). Based on the preexisting literature, the introgression of alleles belonging to these gene groups above into H. annosum had been predicted. Alleles involved in clathrin-mediated vesicular trafficking were also significantly enriched among alleles introgressed in H. annosum hybrids: this catabolic process plays a role in the breakdown of host defenses and in the development of polarized fungal hyphal growth (Steinberg 2007; Shultzhaus et al. 2017). To explain the advantages provided by the presence of these alleles, it should be noted that mutants of the fungus Candida albicans (C.P. Robin) Berkhout with modifications in genes affecting vesicular trafficking are plagued by defects in the cytoskeleton (Epp et al. 2010). Finally, alleles related to “iron-sulphur cluster binding” and “membrane” were found to be introgressed in H. annosum. Iron-sulphur cluster enzymes are known to play different roles in cells, including a role in regulating genomic integrity (Stehling et al. 2012) and in detoxification (Haas et al. 2008). Maintaining genomic integrity is obviously an important function in admixed hybrid genomes, while detoxification may be relevant, for example, during delignification processes (Sista Kameshwar and Qin 2020). Moreover, genes related to membrane may affect the activity of extracellular enzymatic activities, required by wood decay fungi to delignify the substrate in which they live (Wu et al. 2020).

Conversely, directional allelic introgression from H. annosum to H. irregulare, although larger in scale, was limited to DNA-methylation and transcription factor genes. Genes related to DNA-repair have been reported in hybridization events in animals, e.g. sparrows, and it has been hypothesized that they may be selected by nature to modulate epistatic interactions following genomic alterations caused by hybridization (Walsh et al. 2018). The re-patterning of DNA methylation has been also recently documented in plants (Li et al. 2019). This phenomenon is a process affecting gene expression and silencing of mobile elements in the genome (Martienssen and Colot 2001). The introgression of these alleles can be regarded as the result of a selection of H. annosum alleles involved in epistatic regulation within the highly mosaic-like H. irregulare hybrid genotypes. In other words, given that the consequence of mosaic architecture of hybrid genomes can be the alteration of transcriptomic profiling and/or proliferation of mobile elements, methylation and regulation of genes may contribute to the stability of the hybrid genome. Regulatory genes may also be important in optimizing gene expression and adaptation in a microbial species facing a new environment (Elena and Lenski 2003), and it has been hypothesized that evolution may promote adaptations through already available and even acquired genetic tools rather than generating novel catabolic components (Lavoie et al. 2010). In addition, H. annosum GO terms involved in tRNA aminoacylation processes were identified as being overrepresented in H. irregulare hybrids. It is worth noting that tRNA aminoacylation is not only involved in protein translational fidelity that is indispensable for cell survival, but also allows for the tolerance of low levels of mistranslation, a predicament that may occur during stressful exposure to new environments (Wiltrout et al. 2012; Pan 2013). Thus, it could be inferred that the introgression of these alleles may be beneficial to H. irregulare hybrids constantly facing exposure to novel environments. A further transcriptomic analysis (i.e. RNA-sequencing) will be pivotal to clarify the functional role of introgressed alleles during different life phases of the hybrids (e.g. wood colonization, infection).

The sabrobic in vitro assay revealed that H. annosum hybrids containing H. irregulare alleles involved in saprobic decay had an increased saprobic ability, comparable to that of pure H. irregulare genotypes. Based on the results of the saprobic assay, it is reasonable to hypothesize these H. annosum hybrid genotypes may be competitive against pure H. irregulare genotypes. By contrast, H. irregulare hybrid genotypes showed a decrease in saprobic growth when compared to pure H. irregulare genotypes. We believe the impact of large-scale introgression is costly to hybrid H. irregulare genotypes, resulting in a reduction in fitness. We suggest that viability may be maintained only in those H. irregulare hybrids that: first, absorb transcription and methylation related genes from the native species stabilizing their genomic instability caused by large genomic alterations, and, second, maintain their alleles involved in production of peroxidase and heme-binding proteins, mitochondrial transport and vesicle trafficking. This last point is corroborated by the fact that the alleles involved in catabolic pathways were not affected by introgressive events, i.e. these alleles were never replaced by H. annosum alleles in any of the H. irregulare hybrids. Negative epistatic interactions and the breakdown of advantageous gene complexes (Lynch 1991; Guerrero et al. 2017) in H. irregulare recipient genomes, highly modified by large scale introgression events, are two plausible explanations for this detrimental effect on fitness.

Virulence assays in the past have not been able to differentiate the two species (Garbelotto et al. 2010; Pepori et al. 2019). The results of the virulence assay performed in this study further confirmed that virulence may be a genotype-specific trait, rather than a species-level trait as previously suggested (Garbelotto et al. 2010). We note, though, that all hybrids fell within the range of virulence identified for the pure parental genotypes used in the study, and so all can be regarded as ecologically viable. This result confirms that by using genotypes that have established themselves on wood in a natural setting, we obtained a representative sample of functionally and ecologically viable genotypes. In addition, the result also confirms that hybrid swarms represent a viable genetic bridge for the transfer of genetic material between the two species (Brasier 2001).

Although we recognize that the number of genotypes characterized as pure H. annosum and H. irregulare used in our phenotypic tests is limited, the genotypes employed here represented a select sample representative of each species, and results were in agreement with our expectations. In a previous comparative genomic study, genes involved in virulence have been reported to be conserved between the two species (Sillo et al. 2015), resulting in indistinguishable virulent phenotypes (Garbelotto et al. 2010; Pepori et al. 2019). Conversely, genes involved in catabolic processes, including wood decay, and genes regulating nucleus-mitochondrion communication have been shown to diverge (Sillo et al. 2015) and to be associated with different phenotypes (Giordano et al. 2014; 2018). Results of this study are in agreement with a role of selection driving introgression: alleles from divergent loci such as those involved in catabolic processes were significantly enriched as a result of introgression and caused a measurable phenotypic change in the species receiving them by increasing its saprobic ability. On the other hand, we could not document introgression-related enrichment of genes directly involved in virulence of live plants and no measurable change in the associated phenotype could be measured.

It is known that competition with parental species living in the same environment may negatively affect the success of hybrids (Stukenbrock 2016). In the pathosystem studied here, the exotic Heterobasidion species may outcompete or even replace the native species, possibly due to its higher saprobic and sporulation potential (Gonthier et al. 2014). Our results suggest that H. annosum genotypes hybridizing with H. irregulare may become as competitive as those belonging to the exotic species, by receiving alleles that provide them with adaptive phenotypic advantages, including an increased saprobic wood decay potential. If this suggested scenario was correct, these alleles introgressed from H. irregulare into H. annosum, including those related to the saprobic ability among others, may move into H. annosum populations both inside and outside the hybrid zone. The rapid absorption of advantageous or adaptive exotic alleles through introgressive hybridization has been previously documented for many organisms (Fitzpatrick et al. 2010; Crispo et al. 2011) including the fungi (Hessenauer et al. 2020). Thus, H. annosum hybrids, with some specific H. irregulare alleles introgressed in their genome, may themselves represent a threat to the stability of European forests. On the other hand, introgression from the native into the invasive species affected a larger portion of the genome and involved alleles exclusively associated with gene regulation and transcription processes, resulting in lower saprobic growth, but ensuring the viability of hybrid genotypes.

Overall, our results show that mating events leading to viable and fertile Heterobasidion hybrids occur frequently in the Latium region of Italy where the exotic forest pathogen H. irregulare is now sympatric with the native forest pathogen H. annosum. The resulting allelic introgression has identified specific classes of adaptive H. irregulare alleles that confer an advantage to recipient native H. annosum genotypes. Given the phenotypic outcomes of these introgression events including a measurable increase in saprobic potential of H. annosum individuals, we propose an operational shift in the monitoring of the invasion process to include not only the detection of exotic H. irregulare genotypes, but also the presence of adaptive exotic H. irregulare alleles or genes spreading into native H. annosum populations. A gene centric approach to the study of invasions has been previously advocated (Petit, 2004) and clear adaptive implications of genic introgression between tree pathogens have been previously documented for the two fungi causing Dutch elm disease (Hessenauer et al. 2020).

In conclusion, despite the limited number of sequenced hybrid genotypes, this study suggests that horizontal allelic movement occurred as a result of interspecific hybridization, but it was qualitatively and quantitatively different when comparing the two directions of introgression. A large population genomics survey in and near the invasion area in Central Italy targeting introgressed alleles identified in this study will be pivotal to further validate these results.

This work was supported by the Italian Ministry of Education, University and Research, within the FIRB program (grant number RBFR1280NN). This work used the Vincent J. Coates Genomics Sequencing Laboratory at UC Berkeley, supported by NIH S10 Instrumentation Grants S10RR029668 and S10RR027303.

Sequence data were submitted to the EMBL database ENA - European Nucleotide Archive as a project under the PRJEB36378 accession number. All other data were available within the article or its supplementary materials.