Bacteriophages have been proposed as an alternative to pesticides to kill bacterial pathogens of crops. However, the efficacy of phage biocontrol is variable and poorly understood in natural rhizosphere microbiomes. We studied biocontrol efficacy of different phage combinations on Ralstonia solanacearum infection in tomato. Increasing the number of phages in combinations decreased the incidence of disease by up to 80% in greenhouse and field experiments during a single crop season. The decreased incidence of disease was explained by a reduction in pathogen density and the selection for phage-resistant but slow-growing pathogen strains, together with enrichment for bacterial species that were antagonistic toward R. solanacearum. Phage treatment did not affect the existing rhizosphere microbiota. Specific phage combinations have potential as precision tools to control plant pathogenic bacteria.
Subscribe to Journal
Get full journal access for 1 year
only $20.83 per issue
All prices are NET prices.
VAT will be added later in the checkout.
Rent or Buy article
Get time limited or full article access on ReadCube.
All prices are NET prices.
All sequence data generated in this study has been deposited in NCBI SRA database and the accession numbers are reported in the Methods. Phage sequence accession numbers are: SRR8402465 (NJ-P3), SRR8403229 (NB-P21), SRR8403928 (NC-P34) and SRR8410130 (NN-P42). The rhizosphere microbiome data determined at the end of the greenhouse experiment have accession numbers from SRR8417955 to SRR8417999 (45 samples, paired end sequencing). The rhizosphere microbiome data obtained in a separate laboratory experiment have accession numbers running from SRR8470488 to SRR8501098 (30 samples, paired end reads). All other data has been deposited to Dryad Digital Repository with the following digital identifier: https://doi.org/10.5061/dryad.02v6wwpzq.
All code or algorithms used in this study are published and referenced in the Methods.
Raaijmakers, J. M., Paulitz, T. C., Steinberg, C., Alabouvette, C. & Moenne-Loccoz, Y. The rhizosphere: a playground and battlefield for soilborne pathogens and beneficial microorganisms. Plant Soil 321, 341–361 (2009).
Strange, R. N. & Scott, P. R. Plant disease: a threat to global food security. Annu. Rev. Phytopathol. 43, 83–116 (2005).
Berendsen, R. L., Pieterse, C. M. J. & Bakker, P. A. H. M. The rhizosphere microbiome and plant health. Trends Plant Sci. 17, 478–486 (2012).
Hu, J. et al. Probiotic diversity enhances rhizosphere microbiome function and plant disease suppression. Mbio 7, e01790–16 (2016).
Wei, Z. et al. Trophic network architecture of root-associated bacterial communities determines pathogen invasion and plant health. Nat. Commun. 6, 8413 (2015).
Brodeur, J. Host specificity in biological control: insights from opportunistic pathogens. Evol. Appl. 5, 470–480 (2012).
Kyselková, M. & Moënne-Loccoz, Y. Pseudomonas and other microbes in disease-suppressive soils. in Organic Fertilisation, Soil Quality and Human Health 93–140 (Springer, 2012).
Alvarez, B. & Biosca, E. G. Bacteriophage-based bacterial wilt biocontrol for an environmentally sustainable agriculture. Front. Plant Sci. 8, 1218 (2017).
Buttimer, C. et al. Bacteriophages and bacterial plant diseases. Front. Microbiol. 8, 34 (2017).
Rutherford, S. T. & Bassler, B. L. Bacterial quorum sensing: its role in virulence and possibilities for its control. Cold Spring Harb. Perspect. Med. 2, a012427 (2012).
Levin, B. R. & Bull, J. J. Population and evolutionary dynamics of phage therapy. Nat. Rev. Microbiol. 2, 166–173 (2004).
Hoyland-Kroghsbo, N. M., Maerkedahl, R. B. & Svenningsen, S. L. A quorum-sensing-induced bacteriophage defense mechanism. Mbio 4, e00362–00312 (2013).
Tan, D., Svenningsen, S. L. & Middelboe, M. Quorum sensing determines the choice of antiphage defense strategy in Vibrio anguillarum. Mbio 6, e00627 (2015).
Lenski, R. E. Coevolution of bacteria and phage: are there endless cycles of bacterial defenses and phage counterdefenses? J. Theor. Biol. 108, 319–325 (1984).
Bohannan, B. J. M. & Lenski, R. E. Linking genetic change to community evolution: insights from studies of bacteria and bacteriophage. Ecol. Lett. 3, 362–377 (2000).
Wang, X. F. et al. Parasites and competitors suppress bacterial pathogen synergistically due to evolutionary trade-offs. Evolution 71, 733–746 (2017).
Bohannan, B. J., Kerr, B., Jessup, C. M., Hughes, J. B. & Sandvik, G. Trade-offs and coexistence in microbial microcosms. Antonie Van Leeuwenhoek 81, 107–115 (2002).
Peyraud, R., Cottret, L., Marmiesse, L. & Genin, S. Control of primary metabolism by a virulence regulatory network promotes robustness in a plant pathogen. Nat. Commun. 9, 418 (2018).
Perrier, A. et al. Enhanced in planta fitness through adaptive mutations in EfpR, a dual regulator of virulence and metabolic functions in the plant pathogen Ralstonia solanacearum. PLoS Pathog. 12, e1006044 (2016).
Peyraud, R., Cottret, L., Marmiesse, L., Gouzy, J. & Genin, S. A resource allocation trade-Off between virulence and proliferation drives metabolic versatility in the plant pathogen Ralstonia solanacearum. PLoS Pathog. 12, e1005939 (2016).
Hayward, A. C. Biology and epidemiology of bacterial wilt caused by Pseudomonas Solanacearum. Annu. Rev. Phytopathol. 29, 65–87 (1991).
Jiang, G. F. et al. Bacterial wilt in china: History, current status, and future perspectives. Front. Plant Sci. 8, 1549 (2017).
Addy, H. S., Askora, A., Kawasaki, T., Fujie, M. & Yamada, T. Loss of virulence of the phytopathogen Ralstonia solanacearum through infection by phi RSM filamentous phages. Phytopathology 102, 469–477 (2012).
Chapelle, E., Mendes, R., Bakker, P. A. H. M. & Raaijmakers, J. M. Fungal invasion of the rhizosphere microbiome. ISME J.10, 265–268 (2016).
Wei, Z. et al. Ralstonia solanacearum pathogen disrupts bacterial rhizosphere microbiome during an invasion. Soil Biol. Biochem. 118, 8–17 (2018).
Wright, R. C. T., Friman, V. P., Smith, M. C. M. & Brockhurst, M. A. Cross-resistance is modular in bacteria-phage interactions. PLoS Biol. 16, e2006057 (2018).
Hall, A. R., De Vos, D., Friman, V. P., Pirnay, J. P. & Buckling, A. Effects of sequential and simultaneous applications of bacteriophages on populations of Pseudomonas aeruginosa in vitro and in wax moth larvae. Appl. Environ. Microbiol. 78, 5646–5652 (2012).
Kuntal, B. K., Chandrakar, P., Sadhu, S. & Mande, S. S. ‘NetShift’: a methodology for understanding ‘driver microbes’ from healthy and disease microbiome datasets. ISME J. 13, 442–454 (2019).
Yu, L. et al. A guard-killer phage cocktail effectively lyses the host and inhibits the development of phage-resistant strains of Escherichia coli. Appl. Microbiol. Biotechnol. 102, 971–983 (2017).
Wei, C. et al. Developing a bacteriophage cocktail for biocontrol of potato bacterial wilt. Virologica Sinica 32, 476–484 (2017).
Yen, M. M., Cairns, L. S. & Camilli, A. A cocktail of three virulent bacteriophages prevents Vibrio cholerae infection in animal models. Nat. Commun. 8, 14187 (2017).
Betts, A., Gray, C., Zelek, M., MacLean, R. C. & King, K. C. High parasite diversity accelerates host adaptation and diversification. Science 360, 907–911 (2018).
Betts, A., Gifford, D. R., MacLean, R. C. & King, K. C. Parasite diversity drives rapid host dynamics and evolution of resistance in a bacteria-phage system. Evolution 70, 969–978 (2016).
Hall, A. R., Scanlan, P. D., Morgan, A. D. & Buckling, A. Host-parasite coevolutionary arms races give way to fluctuating selection. Ecol. Lett. 14, 635–642 (2011).
Gomez, P. & Buckling, A. Bacteria-phage antagonistic coevolution in soil. Science 332, 106–109 (2011).
Yamada, T. et al. New bacteriophages that infect the phytopathogen Ralstonia solanacearum. Microbiology 153, 2630–2639 (2007).
Taylor, V. L., Fitzpatrick, A. D., Islam, Z. & Maxwell, K. L. The diverse impacts of phage morons on bacterial fitness and virulence. Adv. Virus Res. 103, 1–31 (2019).
Khokhani, D., Lowe-Power, T. M., Tran, T. M. & Allen, C. A single regulator mediates strategic switching between attachment/spread and growth/virulence in the plant pathogen Ralstonia solanacearum. Mbio 8, e00895–00817 (2017).
Friman, V. P., Dupont, A., Bass, D., Murrell, D. J. & Bell, T. Relative importance of evolutionary dynamics depends on the composition of microbial predator-prey community. ISME J. 10, 1352–1362 (2016).
Hiltunen, T. & Becks, L. Consumer co-evolution as an important component of the eco-evolutionary feedback. Nat. Commun. 5, 5226 (2014).
Brunner, F. S., Anaya-Rojas, J. M., Matthews, B. & Eizaguirre, C. Experimental evidence that parasites drive eco-evolutionary feedbacks. Proc. Natl Acad. Sci. USA 114, 3678–3683 (2017).
Gu, Y. et al. Pathogen invasion indirectly changes the composition of soil microbiome via shifts in root exudation profile. Biol. Fertil. Soils 52, 997–1005 (2016).
Li, M. et al. Facilitation promotes invasions in plant-associated microbial communities. Ecol. Lett. 22, 149–158 (2018).
Balogh, B., Jones, J. B., Iriarte, F. B. & Momol, M. T. Phage therapy for plant disease control. Current Pharm. Biotechnol. 11, 48–57 (2010).
Wei, Z. et al. Efficacy of Bacillus-fortified organic fertiliser in controlling bacterial wilt of tomato in the field. Appl. Soil Ecol. 48, 152–159 (2011).
Schonfeld, J., Heuer, H., van Elsas, J. D. & Smalla, K. Specific and sensitive detection of Ralstonia solanacearum in soil on the basis of PCR amplification of fliC fragments. Appl. Environ. Microbiol. 69, 7248–7256 (2003).
Elphinstone, J., Hennessy, J., Wilson, J. & Stead, D. Sensitivity of different methods for the detection of Ralstonia solanacearum in potato tuber extracts. EPPO Bull. 26, 663–678 (1996).
Ji, P. & Wilson, M. Assessment of the importance of similarity in carbon source utilization profiles between the biological control agent and the pathogen in biological control of bacterial speck of tomato. Appl. Environ. Microbiol. 68, 4383–4389 (2002).
Cardenas, E. et al. Significant association between sulfate-reducing bacteria and uranium-reducing microbial communities as revealed by a combined massively parallel sequencing-indicator species approach. Appl. Environ. Microbiol. 76, 6778–6786 (2010).
Caporaso, J. G. et al. QIIME allows analysis of high-throughput community sequencing data. Nat. Methods 7, 335 (2010).
Macdonald, C. A. et al. Long-term impacts of zinc and copper enriched sewage sludge additions on bacterial, archaeal and fungal communities in arable and grassland soils. Soil Biol. Biochem. 43, 932–941 (2011).
Margesin, R., Płaza, G. A. & Kasenbacher, S. Characterization of bacterial communities at heavy-metal-contaminated sites. Chemosphere 82, 1583–1588 (2011).
Fierer, N., Jackson, J. A., Vilgalys, R. & Jackson, R. B. Assessment of soil microbial community structure by use of taxon-specific quantitative PCR assays. Appl. Environ. Microbiol. 71, 4117–4120 (2005).
Eden, P. A., Schmidt, T. M., Blakemore, R. P. & Pace, N. R. Phylogenetic analysis of Aquaspirillum magnetotacticum using polymerase chain reaction-amplified 16S rRNA-specific DNA. Int. J. Syst. Evol. Microbiol. 41, 324–325 (1991).
Friman, V. P. & Buckling, A. Phages can constrain protist predation-driven attenuation of Pseudomonas aeruginosa virulence in multienemy communities. ISME J. 8, 1820–1830 (2014).
Yang, T. et al. Resource availability modulates biodiversity‐invasion relationships by altering competitive interactions. Environ. Microbiol. 19, 2984–2991 (2017).
Dixon, P. VEGAN, a package of R functions for community ecology. J. Veg. Sci. 14, 927–930 (2003).
Bastian, M., Heymann, S. & Jacomy, M. Gephi: an open source software for exploring and manipulating networks. in Third International AAAI Conference on Weblogs and Social Media (2009).
R Core Team. R: A Language and Environment for Statistical Computing (R Foundation for Statistical Computing, 2015).
Hair, J. F., Ringle, C. M. & Sarstedt, M. Partial least squares structural equation modeling: rigorous applications, better results and higher acceptance. Long Range Plann. 46, 1–12 (2013).
This research was financially supported by the National Natural Science Foundation of China (41922053, 41671248; Z.W.), the National Key Basic Research Program of China (2015CB150503; Q.S.), the National Key Research and Development Program of China (2018YFD1000800; Z.W.) and the 111 project (B12009; Q.S.). A.J. is supported by the Netherlands Organization for Scientific Research (NWO) project ALW.870.15.050. V.P.F. is supported by the Wellcome Trust (ref. 105624) through the Center for Chronic Diseases and Disorders (C2D2) and Royal Society Research Grants (RSG\R1\180213 and CHL\R1\180031) at the University of York.
The authors declare no competing interests.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Integrated supplementary information
Supplementary Figure 1 A neighbor-joining phylogenetic tree showing the relatedness of the four phage types used in the experiments with similar phages publicly available in NCBI.
A neighbor-joining phylogenetic tree based on phage head-tail connector protein similarity between the phages used in this experiment (NJ-P3, NB-P21, NC-P34, NN-P42; red colour) and other similar phages that are publicly available at NCBI (black colour). All four phages belonged to order Caudovirales and the family Podoviridae, were highly similar with each other and a Ralstonia phage DU_RP_II. The scale bar represents 0.2 amino acid substitutions per site and values next to the nodes show bootstrap values based on 500 samples.
In all circular genome maps, the outermost black circle represents the full length of the genome, the second outermost multicolored circle represents the forward and reverse reading frames, the third outermost green circle represents GC content and the innermost purple circle represents GC skew.
Supplementary Figure 3 Genetic distance between the four phage types used in our experiment and similar phages publicly available in NCBI with reported information on lytic and lysogenic activity.
Phylogenetic tree showing the relatedness of lytic and lysogenic phages based on genomic similarity. Each genome was first compared to another and to themselves to account and normalize for differences in phage genomes sizes and a neighbor joining tree was then constructed based on the complete distance matrix based on 500 bootstrap sample values as described previously (Mizuno, C.M., Rodriguez-Valera, F., Kimes, N.E. & Ghai, R. Expanding the Marine Virosphere Using Metagenomics. PLoS Genetics. 9 (2013)). Branch colours show different phage families and known lytic and lysogenic phages are shown in black and red, respectively. The phage types used in the experiments (NJ-P3, NB-P21, NC-P34, NN-P42) were most closely related to a lytic phage DU_RP_II.
Supplementary Figure 4 The mean infectivity of phage types to Ralstonia solanacearum clones isolated from China.
The average infectivity of four phage types against 96 R. solanacearum bacterial isolates originating from the tomato rhizosphere of four independent tomato fields in China (24 independent R. solanacearum isolates from each field: Nanjing, Nanchang, Ningbo and Nanning). Bars represent standard deviation between four different fields.
The R2 and P value refer to the most parsimonious model fitted based on linear regression analysis (n=3 for ancestral and control treatment, n=12 for one- and three-phage treatments and n=18 for two-phage treatment; all n are biologically independent samples).
Supplementary Figure 6 Relationships between pathogen growth rate, phage combination treatment and phage resistance at the end of the greenhouse experiment.
Panel A shows the growth rate for evolved pathogen clones isolated from different phage combination treatments at the end of the greenhouse experiment. Lowercase letters above boxplots denote for significant differences between phage combination treatments (multiple comparisons were conducted by Tukey test, FDR adjusted P<0.05) and R2 and red dashed line represents the maximum growth rate of ancestral pathogen strain. Box-plots show interquartile range (25 to 75% of the data), the median as lines and outliers as dots. Panel B shows the correlation between pathogen maximum growth rate and mean phage resistance to all ancestral phages at the end of the greenhouse experiment. The R2 and P values refer to the most parsimonious model, black line shows the mean regression based on all the data points and the small inset in top right corner shows the change in regression coefficient between phage combination treatments (n=3 for ancestral and control treatment, n=12 for one- and three-phage treatments and n=18 for two-phage treatment; all n are biologically independent samples). N0 to N3 denote for no-phage, single-phage, two-phage and three-phage treatments, respectively.
Supplementary Figure 7 Relationship between pathogen competitiveness and phage combination treatment at the end of the greenhouse experiment.
Pathogen competitive ability was measured as the relative growth against the phage-susceptible ancestral pathogen in terms of reduction in the ancestral pathogen abundance. All box-plots show interquartile range (25 to 75% of the data) and median as line. Lowercase letters above boxplots denote for significant differences between phage combination treatments (multiple comparisons were conducted by Tukey test, FDR adjusted P<0.05, n=3 for ancestral and control treatment, n=12 for one- and three-phage treatments and n=18 for two-phage treatment; all n are biologically independent samples).
Supplementary Figure 8 Effects of phage combination treatments on microbial OTU richness at the end of the greenhouse experiment.
Microbial OTU richness correlated positively with the number of phages present in the phage combination treatment. The R2 and P value refer to the most parsimonious model fitted based on linear regression analysis and red dashed line shows the OTU richness of no-phage control treatment (n=3 for control treatment, n=12 for one- and three-phage treatments and n=18 for two-phage treatment; all n are biologically independent samples).
Supplementary Figure 9 Effects of phage combination treatments on relative abundance of different bacterial phyla at the end of the greenhouse experiment.
Correlations between relative abundances of different bacterial phyla and the number of phages present in the phage combinations at the end of the greenhouse experiment. Panels A-H show the correlation for Ralstonia, Proteobacteria, Bacteroidetes, Chloroflexi, Acidobacteria, Planctomycetes, Firmicutes and Actinobacteria, respectively. Red dashed lines in all panels show the no-phage control treatment. The R2 and P value refer to the most parsimonious model fitted based on linear regression analysis (n=3 for control treatment, n=12 for one- and three-phage treatments and n=18 for two-phage treatment; all n are biologically independent samples).
Supplementary Figure 10 Effects of phages on bacterial co-occurrence networks at the end of the greenhouse experiment.
The bacterial co-occurrence networks associated with single-phage and three-phage combinations treatments. Each node represents a bacterial OTU, and each edge represents a Spearman correlation with negative (blue) and positive correlations (red). The node colours represent taxa classification at the phylum level (n=12 biologically independent samples for one- and three-phage treatments).
Supplementary Figure 11 The resistance of pathogenic and non-pathogenic bacterial isolates to phage types used in the experiments.
The average infectivity of four phage types against 96 R. solanacearum bacterial isolates originating from the tomato rhizosphere of four independent tomato fields in China (24 independent R. solanacearum isolates from each field: Nanjing, Nanchang, Ningbo and Nanning) and 400 non-pathogenic bacterial isolates from the soil used in the greenhouse experiment. Bars represent standard deviation between four different field for pathogen isolates.
Panel A shows significant positive (red) and negative (blue) correlation coefficients between bacterial genera abundances and number of phages present in the phage combinations. Correlations were determined using Spearman method and all correlations were FDR adjusted P<0.05. Panel B shows to which phyla positively (top pie chart) and negatively (bottom pie chart) correlated bacteria belonged to.
Supplementary Figure 13 Simplified structural equation models explaining pathogen density, microbiome diversity and disease incidence.
Structural equation model path diagrams showing the phage combination-mediated effects on the pathogen density (a), microbiome diversity (b) and the disease incidence (c). Red, blue and grey arrows denote for positive, negative and non-significant pathways, respectively, and the numbers beside arrows denote for the magnitude of these effects. Numbers within the circles show how much of the variance of each variable was explained by the other variables and χ2 and NFI values denote for the fit of the models.
Supplementary Figure 14 Phylogenetic relationship between Ralstonia solanacearum clones isolated from four different tomato fields.
Phylogenetic neighbour-joining tree based on partial endoglucanase (egl) gene sequences of Chinese Ralstonia solanacearum strains and most closely related reference strains (red circles), that were used to test the phage type infectivity ranges (Supplementary Fig. 4). The tips of the tree show 96 Chinese R. solanacearum isolates with sample location abbreviation and sample ID. Chinese isolates are assigned to specific sequevar types based on their closest relative reference strains and sequevar type numbers are shown in boxes over each clade.
About this article
Cite this article
Wang, X., Wei, Z., Yang, K. et al. Phage combination therapies for bacterial wilt disease in tomato. Nat Biotechnol 37, 1513–1520 (2019) doi:10.1038/s41587-019-0328-3