Volume 28, Issue 7
NEWS AND VIEWS
Free Access

Importance of plant‐ and microbe‐driven metabolic pathways for plant defence

Anna M. O'Brien

Corresponding Author

E-mail address: anna.obrien@utoronto.ca

Department of Ecology and Evolutionary Biology, University of Toronto, Toronto, Ontario, Canada

Correspondence

Anna M. O'Brien, Department of Ecology and Evolutionary Biology, University of Toronto, Toronto, ON, Canada.

Email: anna.obrien@utoronto.ca

Search for more papers by this author
First published: 10 April 2019

Abstract

Expression of plant phenotypes can depend on both plant genomes and interactions between plants and the microbes living in, on and near their roots. We understand a growing number of the mechanistic links between plant genotypes and phenotypes, such as defence against herbivory (see brief review in Hubbard et al., 2019), yet the links between root microbiomes and the comprehensive swathe of plant phenotypes they affect (Friesen et al., 2011) remain less clear. In this issue of Molecular Ecology, Hubbard et al. (2019) follow microbe‐ and plant‐driven changes in plant defence against hervibory from molecular underpinnings to ecological consequences, contrasting both the metabolites affected and the magnitude of defensive impact. Naively, we might expect plant genomes to drive more variation in phenotype than the root microbiome, but Hubbard et al. (2019) find the opposite, implying profound consequences for plant trait evolution and ecological interactions.

Hubbard et al. (2019) grew Boechera stricta (Figure 1a) plants in different microbial contexts and measured effects on concentrations of defensive and primary metabolites, and on realized plant defence. The authors used successive filtration to remove soil particles and then microbes from soil–water slurries, and inoculated populations of B. stricta with either fully filtered water (disrupted communities) or water without microbe filtration (intact communities). By surveying chemical defence and primary plant metabolites, Hubbard et al. (2019) disentangled pathways to defence that are driven by plant variation from pathways driven by microbial variation. Higher concentrations of induced or constitutive defensive chemicals, such as the glucosinolates previously shown to be important in B. stricta (see Hubbard et al., 2019), can reduce the ability of insects to feed, survive or grow on plants. Similar effects may also be achieved by changing the nutritional quality of plant tissues for insects, or the physical barriers to feeding (Agrawal & Fishbein, 2006). These distinct antiherbivore defence mechanisms may rely on separate metabolic pathways (Zhou, Lou, Tzin, & Jander, 2015), and the influence of plant and microbial variation on these pathways may, or may not, overlap.

image
Boechera stricta and its natural habitat. (a) Rosette and root system of one season of growth; (b) natural habitat and sampling field site Road 234 of B. stricta. Image credit Charley Hubbard

Hubbard et al. (2019) found that plant and microbial variation shifted the relative abundances of a number of metabolites, some of which are involved in known herbivore resistance pathways, and that these changes were probably responsible for knock‐on shifts in resistance to herbivory both in the greenhouse and in the field (Figure 1b). Most intriguingly, microbes and plant genotype shift plant defence via different metabolic pathways. Variation in aphid herbivory attributable to plant genotype probably reflects inter‐population differences in glucosinolate compounds, although shifts in ascorbate metabolism may also contribute (further discussed in Hubbard et al., 2019). However, microbe‐driven shifts in plant defensive phenotypes had almost no effect on glucosinolate production: instead, microbiome disruption appeared to decrease physical defences through greater pentose and glucoronate interconversion and to reduce overall defence allocation by increasing plant stress (citrate cycle upregulation, see Hubbard et al., 2019). Essentially, two different subphenotypes make up the combined phenotype of herbivory defence in B. stricta, and these in turn have two separate molecular bases: the microbe‐independent phenotype of chemical defence and the microbe‐dependent phenotype of physical defence and overall allocation. Whether separation of microbe‐dependent and microbe‐independent subphenotypes into distinct molecular bases is common across plant phenotypes will be an exciting avenue of future research.

Another important aspect of the work by Hubbard et al. (2019) is the direct comparison of effect sizes across plant and microbiome variation. Understanding the relative importance of a plant's genome versus microbiome for trait expression matters whenever we consider plant responses to selection pressures, ranging from climate change mitigation to breeding for crop improvement. The vast functional diversity contained within root‐colonizing microbes has raised the possibility that microbial community composition, or even individual species, responds to selection on plant traits faster and with greater effectiveness than plants themselves, and could thus contribute more to plant traits than plant genomes (Mueller & Sachs, 2015). Despite acting through less studied pathways, the effects of microbiome disruption in Hubbard et al. (2019) dwarfed the trait variation due to plant genotype. Disruption caused a seven‐fold difference in flea beetle susceptibility and a three‐fold difference in aphid susceptibility. In contrast to the major effect of disruption, effects of natural microbiome variation (two‐fold, aphids only) appeared more similar in magnitude to effects of plant genotype variation (60%, aphids only, see figures 3 and 4 in Hubbard et al., 2019).

The greater effect on herbivory of microbial disruption than natural microbial variation may reflect the respective scale of differences in microbial communities. Hubbard et al. (2019) used 16S rRNA amplicon sequencing to characterize intact and disrupted root microbiomes. All plants accumulated microbes throughout the experiment: those inoculated with intact communities differed slightly between the two sites of origin, but intact compared to disrupted communities showed dramatic compositional differences (Hubbard et al., 2019, Figures 1 and 2), linking the larger phenotypic effects to larger shifts in microbial communities. If increasing community turnover generally results in increasing effects on plant traits, then ecological drivers that simultaneously alter selection pressures on plant traits and cause microbial community turnover could have consequences for plant trait evolution (e.g., Lau & Lennon, 2012). Global changes are already altering root and soil microbial communities: climate change is expected to produce gradual, smaller shifts in composition (Castro, Classen, Austin, Norby, & Schadt, 2010), but land use activities can generate large shifts (e.g., Verbruggen et al., 2010). As conditions change, microbial turnover could have immediate impacts on plant traits and may make plant evolutionary responses to global change less predictable.

In addition to evolutionary implications, microbially driven shifts in plant metabolites and defensive traits will have ecological consequences. Variation in plant genotype can have dramatic effects on herbivore communities and higher trophic levels (e.g., in poplars, Whitham et al., 2006). Likewise, Hubbard et al. (2019) found a strong effect of B. stricta genotype on aphid abundances. However, while plant genotype affected only aphids, dramatic disruption of the microbial community shifted interactions with both aphids and flea beetles in the field, implying that microbiome variation has the potential to simultaneously alter whole communities of herbivorous insects. Moreover, individual microbial associates of plants affect not only growth rates of herbivorous insects, but can also positively or negatively affect the recruitment and fitness of their parasitoids (Gadhave & Gange, 2018), potentially magnifying or diluting direct effects on defence. Additionally, microbially induced defences may drive generalist herbivorous insects to consume alternative plant hosts. Such a phenomenon could contribute to competitive outcome differences between plants that host contrasting root microbes (Anacker, Klironomos, Maherali, Reinhart, & Strauss, 2014).

The work by Hubbard et al. (2019) cements the emerging view that plant–microbe interactions shape variation in plant defence and herbivory. Much previous discussion of defence in B. stricta has been plant‐centric and focused on chemical defences, and here, the authors again link plant‐driven defence to the well‐studied glucosinolate pathway. However, by broadening ecological context, the authors uncovered microbial influences on entirely different pathways to plant defence, linked by metabolites to physical barrier development and overall defence allocation. Yet, complicating matters, plant and microbe changes may feed back on each other. Plants can themselves shift microbial community composition and may generate genetically deterministic or plastic plant–soil feedbacks on defence, and impacts on plant herbivory can in turn affect microbial fitness (Hu et al., 2018). Furthermore, trade‐offs inherent in some ecological consequences (i.e., deterring both herbivores and their parasitoids) may weaken the benefits of microbe‐mediated defence. Indeed, future work should seek to understand the potential feedbacks in evolutionary and ecological implications of plant–microbe interactions.

One major question future work could address is whether the multiple genetic sources of plant and microbe variation for herbivory cause plant trait evolutionary trajectories to shift, and whether such shifts are adaptive for both plants and microbes. Another road of investigation, in contrast, might measure field ecological consequences for herbivores and competitors, and whether microbially‐driven plant defence is more often mitigated or exacerbated by effects on higher trophic levels. Subsequent investigations should keep in mind the clarity gained here (in Hubbard et al., 2019) through linking larger scale ecological and evolutionary processes to their origins on the molecular scale.

AUTHOR CONTRIBUTIONS

A.M.O'. is the sole author of this News and Views.

    The full text of this article hosted at iucr.org is unavailable due to technical difficulties.