Large-scale phenotyping of 1,000 fungal strains for the degradation of non-natural, industrial compounds

Abstract

Fungal biotechnology is set to play a keystone role in the emerging bioeconomy, notably to address pollution issues arising from human activities. Because they preserve biological diversity, Biological Resource Centres are considered as critical infrastructures to support the development of biotechnological solutions. Here, we report the first large-scale phenotyping of more than 1,000 fungal strains with evaluation of their growth and degradation potential towards five industrial, human-designed and recalcitrant compounds, including two synthetic dyes, two lignocellulose-derived compounds and a synthetic plastic polymer. We draw a functional map over the phylogenetic diversity of Basidiomycota and Ascomycota, to guide the selection of fungal taxa to be tested for dedicated biotechnological applications. We evidence a functional diversity at all taxonomic ranks, including between strains of a same species. Beyond demonstrating the tremendous potential of filamentous fungi, our results pave the avenue for further functional exploration to solve the ever-growing issue of ecosystems pollution.

Fig. 1. Diversity and enrichment strategy of the CIRM-CF fungal collection.

a Overview of the taxonomic and geographic diversity of the 2824 (as of May, 2021) available fungal strains preserved in the CIRM-CF. b Workflow of strains enrichment in the CIRM-CF collection from 2006 to 2020, with fungi coming from mainly France and its overseas territories. The first step consists in collecting fungi in their natural habitat. Second, each collected sample is registered and identified by expert mycologists. The strain isolations are directly performed in a field laboratory, by the CIRM-CF or by associated mycologists from Universities and French learned societies. Finally, the viability, the purity, and the identity of each strain are checked. When the strains are viable, pure, and when the molecular information is in accordance with the morphological identification, the strains can enter the CIRM-CF collection.

Fig. 2. Functional screening methodology.

a The growth phenotype was scored by comparing the mycelium growth in a given condition (i.e., agar-YNB + tested compound) vs the two following controls setting the boundaries of the growth scale: agar-YNB plate (score 0) and malt agar plate (score 4). Note that for RB5 and BB41 the agar plates contained 0.5% malt (vs 1.5% malt in positive control plate) to sustain growth. The white dotted circle indicates the fungal growth diameter (which is not always visible on pictures). b Functional phenotype scores (ranging from 0 to 4) reflect the extent of decolorization (on the industrial dyes Reactive Black 5 (RB5) and Basic Blue 41 (BB41)), browning upon phenol oxidation (on lignosulfonate (LGS)), clearing halo formation (on the soluble polyurethane Impranil® DLN (IMP)) or growth (on Avicel (AVI)). A score of 0 means no decolorization/degradation/oxidation. The pictures show examples for the different scores.

Fig. 3. Global phenotype scoring of the 1031 filamentous strains.

The figure shows for each substrate the number of strains displaying growth and either decolorization (for RB5 and BB41), or oxidation (for LGS) or clearing (for IMP), clustered by score intensity (0, 1–2, and 3–4), see Fig. 2 legend for details on score definition. The pie charts show the relative distribution of scores categories (100% = 1031 strains). AVI Avicel, BB5 Basic Blue 41, IMP Impranil, LGS lignosulfonate, na not applicable, RB5 Reactive Black 5.

Fig. 4. Substrate-wise and phylogeny-wise distribution of the phenotypic scores of the 1031 tested fungal strains.

a UpSet plots illustrating quantitative intersection of the sets of strains showing the highest score (score 4) for each functional phenotype (the colored bars on the left-hand side indicate the total number of strains that scored a 4 for each compound). For instance, looking at IMP, out of 233 “score 4” strains, 172 scored a 4 exclusively on IMP (and not on any other compound), while 22 scored a 4 on both IMP and BB41. The figure shows that high scoring strains with a single phenotype represent the majority. b Mean scores of multi-phenotyping of 1031 natural strains belonging to 26 fungal orders mapped onto the phylogenetic tree. The numbers of analyzed strains (n samples) are indicated in the first column. Horizontal histograms show the mean scores for each fungal order for each phenotype, ranging from 0 to 4. Note that these scores reflect a functional phenotype (i.e., decolorization (RB5, BB41), oxidation (LGS), degradation (IMP), or growth (AVI)). Growth phenotypes for all strains/compounds couples are provided in Supplementary Data 2. Histograms colored in pink indicate orders for which more than ten strains were analyzed (otherwise shown in gray). The topology of the tree was built according to recent phylogenies of Fungi^–.

Fig. 5. Overview of the functional diversity at different taxonomic levels.

a Distribution of functional profiles at the order level. The two-level categorization of non-active (N) and active (A) substrate–strain couples evaluated on 5 different substrates yields a theoretical maximum of 32 profiles. For each fungal order, the numbers provided in the matrix represent the number of strains with a given profile (for each fungal order, a color gradient from white (only one strain) to red (maximum number of strains) has been applied to show the distribution of functional profiles in the order). As an example, 46 strains from the Agaricales order displayed the profile #3 (LGS oxidation only). b–e Logarithmic regression between the number of functional profiles observed (x axis) within a taxonomic rank (indicated in the figure) and the number of analyzed strains (y axis) in this rank, at the order (b), family (c), genus (d), and species (e) levels.

Fig. 6. Box plots showing the fungal family-dependent distribution of phenotypic scores.

For each fungal family, and each target compound (a–e), the median score (thick gray bars) and mean score (red or green diamonds) are shown. For a given target compound, green diamonds show families with a mean score significantly higher than other families. Outliers are represented by dots. Families of brown-rot fungi are marked by an asterisk. For each target compound, non-parametric Kruskal–Wallis test indicates (p value and χ²) significative differences between fungal families. Wilcoxon signed-rank test were computed to compare paired data (i.e., family to family; p values are shown in Supplementary Data 3).