Unique bacterial assembly, composition, and interactions in a parasitic plant and its host

Abstract How plant-associated microbiota are shaped by, and potentially contribute to, the unique ecology and heterotrophic life history of parasitic plants is relatively unknown. Here, we investigate the leaf and root bacterial communities of the root holoparasite Orobanche hederae and its host Hedera spp. from natural populations. Root bacteria inhabiting Orobanche were less diverse, had fewer co-associations, and displayed increased compositional similarity to leaf bacteria relative to Hedera. Overall, Orobanche bacteria exhibited significant congruency with Hedera root bacteria across sites, but not the surrounding soil. Infection had localized and systemic effects on Hedera bacteria, which included effects on the abundance of individual taxa and root network properties. Collectively, our results indicate that the parasitic plant microbiome is derived but distinct from the host plant microbiota, exhibits increased homogenization between shoot and root tissues, and displays far fewer co-associations among individual bacterial members. Host plant infection is accompanied by modest changes of associated microbiota at both local and systemic scales compared with uninfected individuals. Our results are a first step towards extending the growing insight into the assembly and function of the plant microbiome to include the ecologically unique but often overlooked guild of heterotrophic plants.

Measures of α-diversity across plant and soil samples. Hedera root communities. 20 Fig. S11 Principle coordinate analysis of soil bacterial communities. 21 Orobanche and Hedera root communities. 30 Table S9 Results from the leave-one-out Procrustes analysis comparing 31 Orobanche and Hedera root communities. 32 33

Figure S1
Measures of (A) ASV richness, (B) Inverse Simpson's, (C) evenness, and (D) phylogenetic diversity across leaves and roots from Hedera and Orobanche and soil samples. Evenness was calculated as inverse Simpson's diversity/ASV richness and phylogenetic diversity was calculated as the sum of the total phylogenetic branch length for a given sample. For results from statistical analyses see Table S2. Infected host (I) Principle coordinate analyses and corresponding scree plots using (A and C) weighted UniFrac dissimilarity and (B and D) Bray-Curtis dissimilarity. Results were qualitatively similar between measures of dissimilarity.

Figure S3
Venn diagrams displaying the number of shared ASVs among (A) leaf communities, (B) root communities, and (C) leaf and root communities.
Root bacterial networks of Orobanche and Hedera inferred using SparCC. We inferred the bacterial network of (A) infected roots of infected Hedera, (B) uninfected roots of infected Hedera, (C) roots of uninfected Hedera, and (D) Orobanche. Node colour and size represent bacterial phylum classification and abundance (centered log-ratio transformed), respectively. Edge colour and width represent sign (green = positive association, red = negative association), and strength of co-association, respectively. At the whole networklevel, we found large differences in the edge density and betweenness centrality between Hedera and Orobanche, but not across infected and uninfected Hedera roots, as reflected in the (E) degree distribution (number of associations per node) between community types. (F) At the level of individual nodes we found large significant differences in mean betweeness centrality between taxa within the root bacterial networks of Hedera and Orobanche, as well as infected and uninfected Hedera roots. We tested significance using a series of Kolmogorov-Smirnov tests on the distributions of mean node-level betweeness centrality estimated from 10,000 samples with replacement of 50 nodes (see Materials and Methods).
Leaf bacterial networks of Orobanche and Hedera inferred using SPIEC-EASI (qualitatively similar results obtained using SparCC for network inference). We inferred the bacterial network of (A) leaves of infected Hedera, (B) leaves of uninfected Hedera, and (C) leaves of Orobanche. Node colour and size represent bacterial phylum classification and abundance (centered log-ratio transformed), respectively. Nodes are labelled with an arbitrary number assigned to individual ASVs. Edge colour and width represent sign (dark shade = positive association, light shade = negative association), and strength of co-association, respectively. . Colours represent the infected host/parasite sample (three per site) from a given site (site 1 -purple; site 2red; site 3 -blue; site 4 -green). The Procrustes residuals represent the level of discordance between corresponding observations occurring within two bacterial communities. For example, the discordance between the bacterial community occurring within (A) Orobanche roots versus infected Hedera roots was low in infected host/parasite pair 3 from site 1 but comparatively high in infected host/parasite pair 2 from site 2. Lower residual values between corresponding bacterial communities lead to higher values of the overall Procrustes correlationlike statistic (t0). Hierarchical clustering of individual samples within community types reflects the results obtained from the Procrustes analysis. Community types that exhibit higher Procrustes congruence (A, B) also display greater dendrogram concordance (D, E). Hierarchical clustering was performed using the weighted UniFrac distance between all samples within a community type followed by clustering by the unweighted pair group method with arithmetic mean (UPGMA). Dendrograms were created using the 'dendextend' R package version 1.9.0 (Galili, 2015). Parasite root communities vs. Host infected root communities Other Figure S9 The relative abundance of the (A) Burkholderiales, a bacterial order that contributes to the congruence between Orobanche and Hedera root communities (Fig. S8), is highly correlated between host and parasite roots. Note this relationship is still significant even after removal of the point in the top right corner. Conversely, the relative abundance of the (B) Actinomycetales, a bacterial order that contributes to the discordance between Orobanche and Hedera root communities, is uncorrelated between host and parasite roots.

Table S4
Bacterial ASVs unique to host tissues. We list the full ASV sequence as well as the prevalence in the given tissue type (see supplemental excel sheet).

Table S5
Full differential abundance. For each taxonomic rank (different coloured rows), we report the differential abundance results for each taxon across each of the six contrasts. Positive values of differential abundance indicate that a bacterial taxon was found at higher abundance in samples corresponding to the first term in the listed contrast. Differential abundance is estimated as the log2-fold change in read count (DESeq2: LFC) or as the relative difference in read count between experimental factors versus within experimental factors (ALDEx2: Effect). Both DESeq2 and ALDEx2 yielded qualitatively similar results. All P values were corrected using the False Discovery Rate. We also list the average relative abundance of each taxon in leaf and root tissues across both host plant species (see supplemental excel sheet).

Table S7
Results from a series of Procrustes analyses, in which individual sample scores along the first two principle coordinate axes of separate ordinations were matched between different bacterial communities. For example, in the first row of the following table the sample scores from a PCoA of IIR and a PCoA of IUR were matched with a Procrustes analysis. The fit between two sets of scores is given by the Procrustes correlation-like statistic (t0), which ranges from 0 (no fit) to 1 (perfect fit). Numerous permutations of PCoA scores among individual samples followed by recalculation of t0 provides a test of the statistical significance of the observed fit between two sets of PCoA scores. We also include three tests to serve as a 'sanity check', which we expect to exhibit nonsignificant congruence (e.g. Orobanche root microbiota should not exhibit any congruence with those of uninfected Hedera).

Table S8
Results from our leave-one-out approach at the level of bacterial phylum for both parasite root and leaf microbial communities. For parasite root and leaf communities separately, we removed all bacterial ASVs from the parasite dataset classified to a given bacterial phyla, re-calculated weighted UniFrac distances among all samples, obtained sample scores from a new PCoA, and re-calculated t0 with the original Hedera infected root PCoA. The effect of excluding a particular bacterial phylum on the fit between host and parasite microbiota is given by Δt = (texcluded -t0).  Table S9 Results from our leave-one-out approach at the level of bacterial order for both parasite root and leaf microbial communities. For parasite root and leaf communities separately, we removed all bacterial ASVs from the parasite dataset classified to a given bacterial order, re-calculated weighted UniFrac distances among all samples, obtained sample scores from a new PCoA, and re-calculated t0 with the original Hedera infected root PCoA. The effect of excluding a particular bacterial phylum on the fit between host and parasite microbiota is given by Δt = (texcluded -t0).