sign in sign up
<<
Home , Root microbiome

news and views

COMPETING FINANCIAL INTERESTS 5. Harding, M.W., Galat, A., Uehling, D.E. & Schreiber, S.L. various . Lundberg et al.5 sampled eight The authors declare no competing financial interests. Nature 341, 758–760 (1989). 6. Bogyo, M., Verhelst, S., Bellingard-Dubouchaud, V., Arabidopsis lines grown in two soils of pH 1. Frei, A.P. et al. Nat. Biotechnol. 30, 997–1001 Toba, S. & Greenbaum, D. Chem. Biol. 7, 27–38 (2000). 5.9–6.4 differing in sand and loam content at (2012). 7. Liu, Y., Patricelli, M.P. & Cravatt, B.F. Proc. Natl. Acad. two growth stages (one per pot), whereas 2. Chen, Y., Fournier, A., Couvineau, A., Laburthe, M. & Sci. USA 96, 14694–14699 (1999). 6 Amiranoff, B. Proc. Natl. Acad. Sci. USA 90, 8. Leonard, S.E., Reddie, K.G. & Carroll, K.S. ACS Chem. Bulgarelli et al. compared two ecotypes in 3845–3849 (1993). Biol. 4, 783–799 (2009). sandy or loamy soils of pH 6.9–7.3 collected 3. Roth, B.L. et al. Proc. Natl. Acad. Sci. USA 99, 9. verdoes, M. et al. Chem. Biol. 19, 619–628 (2012). in both spring and autumn (nine per 11934–11939 (2002). 10. Wollscheid, B. et al. Nat. Biotechnol. 27, 378–386 4. Dalgleish, A.G. et al. Nature 312, 763–767 (1984). (2009). pot, pooled at sampling). Both groups used at least ten replicate pots for each treatment and used similar methods to extract bacterial DNA and separate the different plant compartments. They collected (defined as Who’s who in the plant those adhering to the root surface and inhab­ iting up to 1 millimeter from it; Fig. 1) ? before removing the outer root cortex by sonication, which enabled extraction of both Penny R Hirsch & Tim H Mauchline and -plant DNA. In the two studies the researchers amplified different vari­ reveals the identities of the that make up the able regions of the bacterial 16S rRNA gene, but, as reported previously8, soil endophytic and rhizosphere of the model plant . structure at the level was remarkably similar for all the soil samples from plant- The term ‘rhizosphere’ was coined by Lorenz but in situ hybridization of with fluores­ free pots: dominated by , with Hiltner in 1904 to describe the influence of cently labeled DNA probes has localized some Acidobacteria, Actinobacteria, Bacteroidetes, root exudates on the proliferation of soil micro­ bacterial groups5,6. However, whether micro­ and Gemmatimonadetes forming organisms around and inside roots1. Since then, organisms present in the root microbiome are substantial proportions. much has been learned about the interactions recruited stochastically from the local soil com­ In both studies, comparison of rich­ between soil microorganisms, rhizosphere munity or actively through selection by plant ness in the different compartments required colonists and plant hosts2–4. But to understand exudates remains unclear. Similarly, although normalization of the number of sequences, more fully how rhizosphere microorganisms some are beneficial3,4, the identi­ which entailed pooling replicates of the endo­ are recruited from soil and either benefit or ties and functions of the majority of microbial phytic compartments (where bacteria are much harm plant growth, nutrition and health, we species that inhabit root intercellular spaces less abundant) and using rarefaction to gener­ need to compare complete inventories of the are unknown. ate 1,000 reads per sample. In both studies, microorganisms that are present, which can The discrepancy between the numbers of soil type defined the composition of the rhizo­ be problematic as many bacteria, and bacterial cells observed by direct microscopy sphere microbiome, and the endophytes were fungi cannot be cultured. In recent articles in of stained soil samples and of those grow­ a smaller, distinct group (50% fewer species © 2012 Nature America, Inc. All rights reserved. America, Inc. © 2012 Nature Nature, two groups have used metagenomics to ing on agar plates has long been recognized. identified than in the rhizosphere) dominated describe in the most detail reported to date the The much less abundant, slow-growing soil by Actinobacteria followed by Proteobacteria; core rhizosphere and endophytic microbiomes archaea have not been detected using agar plate Firmicutes, Bacteroidetes and npg of the model plant Arabidopsis thaliana5,6. culture, and there are obvious inaccuracies were also present. The families and genera The 19th century pioneers of using culture-based methods to assess the detected have been described previously as identified the importance of soil bacteria in the abundance of fungal mycelium and spores7. effective root colonizers; some, including major nutrient cycles and were aware of special It was unclear to soil microbiologists whether Streptomyces, have been detected in seeds9. The relationships between plant roots and micro­ the excess bacterial cells are moribund relatives relative abundance of sequences identified as organisms. They studied mycorrhizal fungi and of the community that can be cultured or of Streptomyces sp. was surprising but possibly the nitrogen-fixing symbioses of root nodule– different phyla that cannot be cultured. The use biased by the extraction method as these acti­ inducing with legumes and of actino­ of DNA-based and RNA-based metagenomic nobacteria have robust spores. Bulgarelli et al.6 bacteria with various shrubs. Subsequently, sequencing for analyzing soil microbial com­ also investigated the colonization of non-live, the importance of the whole munities en masse has revealed that the excess exudate-free plant material by inserting wooden (below and above ground, including seeds) was cells most likely represent uncultured bacteria, sticks into soil. They found that some members demonstrated2–4. constrains with the 109 cells per gram in typical soil com­ of the Proteobacteria and Bacteroidetes colo­ ­bacterial movement, and rhizosphere bacteria prising up to 106 taxa7. Notably, archaeal spe­ nized both Arabidopsis roots and wooden sticks, are recruited from the immediate vicinity of the cies and recently described bacterial phyla, indicating that the lignocellulose nature of the root2, although some are bequeathed from pre­ including Acidobacteria, Gemmatimonadetes surface may be more important than root exu­ vious generations via the seed. Spatial aspects and Verrucomicrobia, are ubiquitous in soil8. dates in selecting these particular groups. of the plant microbiome are difficult to resolve, Lundberg et al.5 and Bulgarelli et al.6 used Both Lundberg et al.5 and Bulgarelli et al.6 454 sequencing (Roche) of 16S rRNA gene conclude that there is a typical ‘core’ microbiome Penny R. Hirsch and Tim H. Mauchline are at amplicons to compare soil, rhizosphere and for Arabidopsis that is recruited from common Rothamsted Research, Harpenden, Herts, UK. endophytic bacterial communities (Fig. 1) soil bacteria and selected by the ability of com­ e-mail: [email protected] or from different Arabidopsis genotypes grown munity members to grow in root exudates; a [email protected] in pots under controlled conditions in distinct subset of the bacteria that colonize the

nature biotechnology volume 30 number 10 OCTOBER 2012 961 news and views

a was only one factor in controlling disease, and Endophytic Rhizosphere microbiome microbiome the authors concluded that the presence of a complex community was also important in Bacteria protecting roots. This illustrates the advan­ tage to the plant in attracting and cultivating a Root xylem hair diverse rhizosphere community and explains the observation that it is difficult to intro­ Bulk duce root-colonizing bacteria and fungi with soil 2–4 Rhizosphere Fungi ­beneficial properties . Unless they are espe­ cially fast-growing and aggressive or occupy a specific niche (as do the root-nodule bacteria),

Inner root Cortex Epidermis they cannot compete with the existing commu­ nity adapted to the receiving soil, influenced by b factors including climate, local mineralogy and soil pH (the major influence on the composi­ tion of soil communities)11. Sonication The work of Lundberg et al.5 and Bulgarelli et al.6 provides the most comprehensive picture to date of the Arabidopsis root microbiome and its relationship with the soil microbial com­ munity. It also provides a set of protocols for examining the rhizosphere in Arabidopsis and other small-rooted plants, which will facilitate future comparisons with other genotypes— for example, to study the interaction of endo­ phytes and plant gene expression. However, full metagenomic sequencing will be required Washing Washing to pinpoint which bacterial genes are selected Rhizosphere in surface and internal root colonization of compartment Arabidopsis, as well as transcriptomics and Unplanted soil Root compartment proteomics to indicate when and where these compartment genes are expressed. Such studies may reveal Figure 1 Separating out root-associated microbiomes for metagenomic analyses. (a) The rhizosphere the contribution of rhizosphere bacteria to microbiome includes bacteria and fungi that are recruited from bulk soil and colonize the root surface. both disease resistance and nutrient cycling in The endophytic microbiome includes species that infiltrate the root cortex and live as endophytes until this model plant system. Ultimately, we look their release back into the soil upon root senescence. Some microorganisms can colonize the seed. forward to the application of these approaches © 2012 Nature America, Inc. All rights reserved. America, Inc. © 2012 Nature (b) The fractionation protocols used in both studies (adapted from ref. 6). Briefly, roots are shaken to to crop plants, leading to a full understanding remove adherent soil, and rhizosphere microorganisms are collected by washing. Sonication is used of the plant--soil system that to remove epidermal cells, leaving the root compartment, which contains the endophytic microbiome. DNA is extracted from soil, rhizosphere and endophytic compartments for metagenomic sequencing. will enable us to optimize plant health, nutri­ npg tion and yields in sustainable agriculture.

root surface can enter the interior of the root been subjected to such intensive sequencing, so COMPETING FINANCIAL INTERESTS The authors declare no competing financial interests. and survive as endophytes. As both studies it is difficult to compare the influence of dif­ compared only 16S rRNA genes, the relative ferent genotypes in one species with the selec­ 1. Hartmann, A., Rothballer, M. & Schmid, M. Plant Soil 312, 7–14 (2008). importance of the ability to grow rapidly in tive ability of different plant species overall, 2. Dennis, P.G., Miller, A.J. & Hirsch, P.R. FEMS Microbiol. common components of root exudates, catabo­ but many studies attest to gross differences in Ecol. 72, 313–327 (2010). lize unusual substrates or respond to specific the composition of exudates and also of rhizo­ 3. Friesen, M.L. et al. Ann. Rev. Ecol. Evol. Syst. 42, 2–4 23–46 (2011). signal molecules can only be partially dis­ sphere communities . 4. Berendsen, R.L., Pieterse, C.M.J. & Bakker, P.A.H.M. cerned. However, from a principal coordinate A recent study of sugar beet grown in soil Trends Plant Sci. 17, 478–486 (2012). analysis of their sequence data, Lundberg et al.5 suppressive or conducive for the fungal patho­ 5. Lundberg, D.S. et al. Nature 488, 86–90 (2012). 6. Bulgarelli, D. et al. Nature 488, 91–95 (2012). 10 concluded that plant genotype and age were gen Rhizoctonia revealed similar assemblages 7. Hirsch, P.R., Mauchline, T.H. & Clark, I.M. Soil Biol. less important than soil type in defining the of rhizosphere bacteria to those reported in Biochem. 42, 878–887 (2010). 6 8. Roesch, L.F. et al. ISME J. 1, 283–290 (2007). assemblage of species present. Bulgarelli et al. Arabidopsis, using a different method (phlyo­ 9. Coombs, J.T. & Franco, C.M.M. Appl. Environ. estimated that only 60% of root colonists were chip microarrays rather than 16S rRNA Microbiol. 69, 4260–4262 (2003). stimulated by root exudates, with the remain­ amplicon sequencing). Although some of the 10. Mendes, R. et al. Science 332, 1097–1100 (2011). der responding to the lignocellulose surface. proteobacteria present in disease-suppressive 11. Fierer, N. & Jackson, R.B. Proc. Natl. Acad. Sci. USA To date, no other plant species microbiome has soils produced an antifungal compound, this 103, 626–631 (2006).

962 volume 30 number 10 OCTOBER 2012 nature biotechnology