Mosaic fungal individuals have the potential to evolve within a single generation

Within-generation HGM

After matings of compatible hyphal tips grown from spores, haploid dikaryotic nuclei (n + n) of A. gallica fuse to produce diploid monokaryons (2n). As monokaryons are persistent in vegetative stages and often possess two distinct molecular-marker alleles, the model of vegetative heterozygous diploidy is widely accepted. But since other studies show vegetative stages can possess recombinant, haploid nuclei, an alternative hypothesis has been advanced. This hypothesis proposes a life cycle in which a vegetative-stage haploidization produces HGM^6,7,17,18. Our analyses confirm that vegetative-stage hyphae can be haploid (Fig. 1, Supplementary Table S1), while still possessing two different molecular-marker alleles (Supplementary Table S2).

Although RFLP data are consistent with both heterozygous diploid and haploid genetic mosaic models, DNA content data and EF1α sequence data both argue against the heterozygous diploid model. Since EF1α is a single-copy gene, multiple cloned sequences isolated from a single hyphal filament should have only 1 haplotype if the filament is a diploid homozygote or 2 haplotypes if it is a diploid heterozygote; but it could have 1, 2, 3 or more haplotypes if it is a haploid genetic mosaic. The upper limit on the number of haplotypes detected in a hyphal filament is set by the number of hyphal compartments recovered during cell-line isolation. We estimate that, on average, six contiguous compartments were harvested each time we isolated a hyphal filament line; and there were 26 instances in which 3 or more clones were successfully sequenced from within a single hyphal filament line. In these 26 lines, we detected 1 or 2 haplotypes 11 times and 3 or 4 haplotypes 15 times (Table 1, Supplementary Table S3a–c). The 11 instances in which 1 or 2 haplotypes were detected are compatible with either model; but the 15 instances in which 3 or 4 haplotypes were detected are compatible with only the haploid genetic mosaic model. In conjunction with the finding of haploidy in vegetative stages, this finding argues against the heterozygous diploid model and supports the haploid genetic mosaic model. We define a haploid genetic mosaic as a mycelium with haplotypes that vary within and among hyphae. As an example, Fig. 6 depicts two haploid genetic mosaic rhizomorph hyphal filament lines that were isolated from the Raynham genet.

Haplotype designations hap 1, hap 3…hap 13 refer to EF1α haplotypes listed in rows 1, 3, 5, 6, 8, 12, and 13 of Table 1. Note that (1) haplotype 13 is the only haplotype shared by both filament lines; (2) the order of the nuclei in the filaments is not known, so it is arbitrarily shown as numerical; (3) the spacers are hypothetical, as usually a maximum of 6 nuclei were included in an isolate.

We are not the first to propose HGM in Armillaria. Ullrich and Anderson¹⁹ considered stable diploidy as the most likely explanation for prototrophy in mated auxotrophs of Armillaria mellea. However, they also presented an alternative hypothesis that they considered a less likely but possible explanation for their results: “Alternatively, it is possible that an unusual (unprecedented) type of heterokaryon is present, i.e., one that is vegetatively stable in a filamentous fungus with uninucleate cells and intact septa.” Our results appear to be an example of Ullrich and Anderson’s alternative model.

Because hyphal extension requires mitosis, contiguous compartments within growing hyphal tips should contain a series of identical nuclei. How then, in rhizomorphs capable of undergoing mitosis for decades, can within-hyphal filament HGM persist? Korhonen²⁰ was the first to document nuclear migration through cytoplasmic bridges in Armillaria. We found cytoplasmic bridges to be common in monokaryotic rhizomorph hyphae collected in nature (Fig. 3) and hyphae grown in culture (Fig. 4). Because nuclei were frequently found in or near bridges, we propose nuclear exchange through bridges as a mechanism that maintains within-line and among-line HGM (Fig. 7).

Growth

Gallic acid growth experiments revealed significant line effects, treatment effects, and line × treatment effects for all 4 sets of Raynham and Bridgewater cell-lines (ANOVA P < 0.0001, Fig. 5, Supplementary Table S6). Because spores and rhizomorphs each possess genetic variation for growth and phenotypic plasticity, selection has the potential to affect their growth and phenotypic plasticity in nature.

Although reaction-norm shapes are similar in all four sets of Fig. 5 curves, the vertical spread of curves appears to be greater for spores than for rhizomorphs, and a posteriori paired t-tests show this is true at all 8 gallic acid concentrations (Supplementary Table S5). Variance differences could reflect different selection histories. Armillaria gallica spores disperse over long distances of up to 2 km¹³ and have the potential to land near hosts that produce very different concentrations of gallic acid. Selection in spores may therefore favor wide ranges of abilities to grow in the presence of gallic acid. In contrast, rhizomorphs assimilate nutrients and grow for long periods through soils and near hosts where conditions are comparatively stable. Environmental stability may have given rhizomorph genets in Bridgewater and Raynham opportunities to approach local adaptive norms. The immortal-strand²¹ hypothesis posits that asymmetric cell divisions preserve template DNA strands within stem cells. The dancing-genome hypothesis² proposes that fungal nuclei can be distributed non-randomly to daughter cells in some fungi. Taken together these models suggest mechanisms that could maintain different populations of alleles in spore vs. rhizomorph stages of the life cycle. It would be interesting to see whether a priori comparisons of variances in other genets support this hypothesis.

Phenotypic plasticity

Cytoplasmic bridges permit exchanges of nuclei, nutrients, and other materials among hyphae^{20,22,23,24,25}. Every intersection between graph lines in Fig. 5 represents a reversal in relative growth of 2 rhizomorph lines grown on media containing different gallic acid concentrations. In the Raynham rhizomorph graph, there are at least 70 intersections. Consider the intersection of Raynham rhizomorph hyphal filament lines r12 and r13 between gallic acid concentrations of 8 mM and 16 mM. In the lab, rhizomorph line r12 grows larger than line r13 on 8 mM media; r13 grows larger than r12 on 16 mM media (95% CIs do not overlap). In nature, if a rhizomorph containing hyphal lines r12 and r13 extends between areas where one host produces lower (8 mM) concentrations of gallic acid and another host produces higher (16 mM) concentrations, cytoplasmic bridges might favor movement of nuclei (and potentially nutrients) in one direction near the first host and in the opposite direction near the second host.

In other fungi, nuclei travel considerable distances along hyphal filaments^26,27. If migrating nuclei in A. gallica cross cytoplasmic bridges, enter different hyphae, and undergo differential rates of mitosis, this process might create new populations of interacting haplotypes better suited to overcoming host defenses or utilizing resources of different hosts. The resulting reversible partitioning of nuclei among interconnected cells could help mosaic individuals deal with exposure to diverse environments over their long lives. This within-generation process could be seen as effectively equivalent to adaptive evolution that usually takes place between generations in species with unitary individuals. Environmentally-dependent synergisms within mycelia might also apply to interactions between mycorrhizal fungi and plants²⁸, and therefore contribute to phenotypes that impart selective advantages in other types of haploid genetic mosaic organisms^29,30. If organisms other than A. gallica respond to environmental conditions in this way, haploid genetic mosaics may be more common than is currently thought.

Growth-study evidence of rhizomorph within-line genetic variance

ANOVA analyses of gallic acid growth trials suggest spores lack a within-line variance component that rhizomorphs possess. When averaged over 8 gallic acid concentrations, among-line variance accounts for an average of 95% of all spore-line growth variance (Table 2). Spore residual terms are presumed to be low (average = 5%) because they are affected by only plate-to-plate environmental variation. This is expected, given that spore mycelia lack cytoplasmic bridge connections among hyphae; and EF1α sequences show that spore within-line haploid genetic mosaic variation approaches zero. In contrast, rhizomorph residual terms (average = 19%) are approximately 4 times higher than spore residual terms (average = 5%) even though in comparison to spores their experimental plates are not expected to have higher levels of plate-to-plate environmental variation. We propose instead that rhizomorph residual terms are higher because they include a within-line genetic variance component that spores lack. Genetic variance is expected within rhizomorph hyphal filaments because they typically have multiple, among-hyphae cytoplasmic-bridge connections and have been shown to possess as many as 4 different EF1α sequences within individual hyphal filaments.

Haploid genetic mosaic organisms vs. diploid or haploid organisms

Because diploids have more mutation targets than haploids, they may have advantages in environments where adaptation is limited by total genetic diversity³. Haploid genetic mosaic A. gallica individuals, however, potentially have even more mutation targets than diploids; and differential nuclear replication and migration within individuals may allow beneficial alleles to increase in frequency within as well as between generations. Selection may eliminate harmful alleles more efficiently in haploids because their fitness effects are not masked. Haploid genetic mosaic A. gallica individuals may have advantages over strict haploids though, because nutrient flow within and among hyphal filaments may temporarily protect nuclei containing deleterious alleles so that they will remain available to be selected for in the event that environmental conditions change. The extreme longevity and size some A. gallica individuals attain suggest that life cycle features allow them to adapt to a wide range of environmental conditions over time and space, and we propose HGM may have contributed to this success.

Source: Ecology - nature.com