摘要
•Leaf-associated microbiota confer protection against pathogen Pseudomonas syringae•Degree of protection depends on the initial dose of microbiota being applied•A 12-member community applied at low doses was sufficient to reduce pathogen growth•Fertilizer application to host plants abolished the microbiome-mediated protection Plant-associated microbial communities can promote plant nutrient uptake, growth, and resistance to pathogens [1Suckstorff I. Berg G. Evidence for dose-dependent effects on plant growth by Stenotrophomonas strains from different origins.J. Appl. Microbiol. 2003; 95: 656-663Crossref PubMed Scopus (60) Google Scholar, 2Van Wees S.C. Van der Ent S. Pieterse C.M. Plant immune responses triggered by beneficial microbes.Curr. Opin. Plant Biol. 2008; 11: 443-448Crossref PubMed Scopus (575) Google Scholar, 3Finkel O.M. Castrillo G. Herrera Paredes S. Salas González I. Dangl J.L. Understanding and exploiting plant beneficial microbes.Curr. Opin. Plant Biol. 2017; 38: 155-163Crossref PubMed Scopus (336) Google Scholar]. Host resistance to infection can increase directly through commensal-pathogen interactions or indirectly through modulation of host defenses [4Innerebner G. Knief C. Vorholt J.A. Protection of Arabidopsis thaliana against leaf-pathogenic Pseudomonas syringae by Sphingomonas strains in a controlled model system.Appl. Environ. Microbiol. 2011; 77: 3202-3210Crossref PubMed Scopus (322) Google Scholar, 5Bais H.P. Fall R. Vivanco J.M. Biocontrol of Bacillus subtilis against infection of Arabidopsis roots by Pseudomonas syringae is facilitated by biofilm formation and surfactin production.Plant Physiol. 2004; 134: 307-319Crossref PubMed Scopus (651) Google Scholar, 6Millet Y.A. Danna C.H. Clay N.K. Songnuan W. Simon M.D. Werck-Reichhart D. Ausubel F.M. Innate immune responses activated in Arabidopsis roots by microbe-associated molecular patterns.Plant Cell. 2010; 22: 973-990Crossref PubMed Scopus (392) Google Scholar], the mechanisms of which are best described for rhizosphere-associated bacteria. For example, Arabidopsis plants infected with the foliar pathogen, Pseudomonas syringae pathovar tomato (Pst), increase their root secretion of malate, which attracts Bacillus subtillis to the roots and leads to a stronger host response against Pst [7Rudrappa T. Czymmek K.J. Paré P.W. Bais H.P. Root-secreted malic acid recruits beneficial soil bacteria.Plant Physiol. 2008; 148: 1547-1556Crossref PubMed Scopus (610) Google Scholar]. Although there are numerous examples of individual defensive symbionts (e.g., [8Hopkins S.R. Wojdak J.M. Belden L.K. Defensive symbionts mediate host–parasite interactions at multiple scales.Trends Parasitol. 2017; 33: 53-64Abstract Full Text Full Text PDF PubMed Scopus (32) Google Scholar]), it is less clear whether this type of protection is an emergent property of whole microbial communities. In particular, relatively little is known about whether and how the presence of phyllosphere (above-ground) microbial communities can increase host resistance against pathogens. In this study, we examined the ability of augmented tomato phyllosphere microbiomes to confer resistance against the causal agent of bacterial speck, Pst. Across five independent experiments, the augmented phyllosphere microbiome was found to decrease pathogen colonization. Furthermore, the dose of commensal bacteria applied affected the degree of protection conferred, and although the effect is dependent on microbial composition, it is not clearly related to overall bacterial diversity. Finally, our results suggest that resources available to the phyllosphere microbial community may play an important role in protection, as the addition of fertilizer abolished the observed microbiome-mediated protection. Together, these results have clear relevance to microbiome-mediated protection within agricultural settings and the use of plant probiotics to increase disease resistance. Plant-associated microbial communities can promote plant nutrient uptake, growth, and resistance to pathogens [1Suckstorff I. Berg G. Evidence for dose-dependent effects on plant growth by Stenotrophomonas strains from different origins.J. Appl. Microbiol. 2003; 95: 656-663Crossref PubMed Scopus (60) Google Scholar, 2Van Wees S.C. Van der Ent S. Pieterse C.M. Plant immune responses triggered by beneficial microbes.Curr. Opin. Plant Biol. 2008; 11: 443-448Crossref PubMed Scopus (575) Google Scholar, 3Finkel O.M. Castrillo G. Herrera Paredes S. Salas González I. Dangl J.L. Understanding and exploiting plant beneficial microbes.Curr. Opin. Plant Biol. 2017; 38: 155-163Crossref PubMed Scopus (336) Google Scholar]. Host resistance to infection can increase directly through commensal-pathogen interactions or indirectly through modulation of host defenses [4Innerebner G. Knief C. Vorholt J.A. Protection of Arabidopsis thaliana against leaf-pathogenic Pseudomonas syringae by Sphingomonas strains in a controlled model system.Appl. Environ. Microbiol. 2011; 77: 3202-3210Crossref PubMed Scopus (322) Google Scholar, 5Bais H.P. Fall R. Vivanco J.M. Biocontrol of Bacillus subtilis against infection of Arabidopsis roots by Pseudomonas syringae is facilitated by biofilm formation and surfactin production.Plant Physiol. 2004; 134: 307-319Crossref PubMed Scopus (651) Google Scholar, 6Millet Y.A. Danna C.H. Clay N.K. Songnuan W. Simon M.D. Werck-Reichhart D. Ausubel F.M. Innate immune responses activated in Arabidopsis roots by microbe-associated molecular patterns.Plant Cell. 2010; 22: 973-990Crossref PubMed Scopus (392) Google Scholar], the mechanisms of which are best described for rhizosphere-associated bacteria. For example, Arabidopsis plants infected with the foliar pathogen, Pseudomonas syringae pathovar tomato (Pst), increase their root secretion of malate, which attracts Bacillus subtillis to the roots and leads to a stronger host response against Pst [7Rudrappa T. Czymmek K.J. Paré P.W. Bais H.P. Root-secreted malic acid recruits beneficial soil bacteria.Plant Physiol. 2008; 148: 1547-1556Crossref PubMed Scopus (610) Google Scholar]. Although there are numerous examples of individual defensive symbionts (e.g., [8Hopkins S.R. Wojdak J.M. Belden L.K. Defensive symbionts mediate host–parasite interactions at multiple scales.Trends Parasitol. 2017; 33: 53-64Abstract Full Text Full Text PDF PubMed Scopus (32) Google Scholar]), it is less clear whether this type of protection is an emergent property of whole microbial communities. In particular, relatively little is known about whether and how the presence of phyllosphere (above-ground) microbial communities can increase host resistance against pathogens. In this study, we examined the ability of augmented tomato phyllosphere microbiomes to confer resistance against the causal agent of bacterial speck, Pst. Across five independent experiments, the augmented phyllosphere microbiome was found to decrease pathogen colonization. Furthermore, the dose of commensal bacteria applied affected the degree of protection conferred, and although the effect is dependent on microbial composition, it is not clearly related to overall bacterial diversity. Finally, our results suggest that resources available to the phyllosphere microbial community may play an important role in protection, as the addition of fertilizer abolished the observed microbiome-mediated protection. Together, these results have clear relevance to microbiome-mediated protection within agricultural settings and the use of plant probiotics to increase disease resistance. The microbiomes of both plants and animals have been shown to provide beneficial host functions [1Suckstorff I. Berg G. Evidence for dose-dependent effects on plant growth by Stenotrophomonas strains from different origins.J. Appl. Microbiol. 2003; 95: 656-663Crossref PubMed Scopus (60) Google Scholar, 3Finkel O.M. Castrillo G. Herrera Paredes S. Salas González I. Dangl J.L. Understanding and exploiting plant beneficial microbes.Curr. Opin. Plant Biol. 2017; 38: 155-163Crossref PubMed Scopus (336) Google Scholar, 9Lugtenberg B. Kamilova F. Plant-growth-promoting rhizobacteria.Annu. Rev. Microbiol. 2009; 63: 541-556Crossref PubMed Scopus (2254) Google Scholar, 10Clemente J.C. Ursell L.K. Parfrey L.W. Knight R. The impact of the gut microbiota on human health: an integrative view.Cell. 2012; 148: 1258-1270Abstract Full Text Full Text PDF PubMed Scopus (2263) Google Scholar], including conferring resistance against pathogens either directly through microbe-microbe interactions [4Innerebner G. Knief C. Vorholt J.A. Protection of Arabidopsis thaliana against leaf-pathogenic Pseudomonas syringae by Sphingomonas strains in a controlled model system.Appl. Environ. Microbiol. 2011; 77: 3202-3210Crossref PubMed Scopus (322) Google Scholar, 5Bais H.P. Fall R. Vivanco J.M. Biocontrol of Bacillus subtilis against infection of Arabidopsis roots by Pseudomonas syringae is facilitated by biofilm formation and surfactin production.Plant Physiol. 2004; 134: 307-319Crossref PubMed Scopus (651) Google Scholar, 11Kamada N. Kim Y.G. Sham H.P. Vallance B.A. Puente J.L. Martens E.C. Núñez G. Regulated virulence controls the ability of a pathogen to compete with the gut microbiota.Science. 2012; 336: 1325-1329Crossref PubMed Scopus (422) Google Scholar] or indirectly through alteration of host defenses [2Van Wees S.C. Van der Ent S. Pieterse C.M. Plant immune responses triggered by beneficial microbes.Curr. Opin. Plant Biol. 2008; 11: 443-448Crossref PubMed Scopus (575) Google Scholar, 6Millet Y.A. Danna C.H. Clay N.K. Songnuan W. Simon M.D. Werck-Reichhart D. Ausubel F.M. Innate immune responses activated in Arabidopsis roots by microbe-associated molecular patterns.Plant Cell. 2010; 22: 973-990Crossref PubMed Scopus (392) Google Scholar, 12Montalvo-Katz S. Huang H. Appel M.D. Berg M. Shapira M. Association with soil bacteria enhances p38-dependent infection resistance in Caenorhabditis elegans.Infect. Immun. 2013; 81: 514-520Crossref PubMed Scopus (88) Google Scholar, 13Morella N.M. Koskella B. The value of a comparative approach to understand the complex interplay between microbiota and host immunity.Front. Immunol. 2017; 8: 1114Crossref PubMed Scopus (6) Google Scholar]. Microbe-mediated disease resistance in plants has primarily focused on the rhizosphere and the surrounding soil microbial community [5Bais H.P. Fall R. Vivanco J.M. Biocontrol of Bacillus subtilis against infection of Arabidopsis roots by Pseudomonas syringae is facilitated by biofilm formation and surfactin production.Plant Physiol. 2004; 134: 307-319Crossref PubMed Scopus (651) Google Scholar, 14Wei Z. Yang T. Friman V.P. Xu Y. Shen Q. Jousset A. Trophic network architecture of root-associated bacterial communities determines pathogen invasion and plant health.Nat. Commun. 2015; 6: 8413Crossref PubMed Scopus (256) Google Scholar], where soils that suppress pathogen growth (i.e., disease-suppressive soils) do so by increasing microbe-pathogen competition for resources in the rhizosphere [15Schlatter D. Kinkel L. Thomashow L. Weller D. Paulitz T. Disease suppressive soils: New insights from the soil microbiome.Phytopathology. 2017; 107: 1284-1297Crossref PubMed Scopus (252) Google Scholar]. Although rhizosphere bacteria have been shown to increase resistance to infection [2Van Wees S.C. Van der Ent S. Pieterse C.M. Plant immune responses triggered by beneficial microbes.Curr. Opin. Plant Biol. 2008; 11: 443-448Crossref PubMed Scopus (575) Google Scholar, 14Wei Z. Yang T. Friman V.P. Xu Y. Shen Q. Jousset A. Trophic network architecture of root-associated bacterial communities determines pathogen invasion and plant health.Nat. Commun. 2015; 6: 8413Crossref PubMed Scopus (256) Google Scholar], little has been done to study how the phyllosphere community can impact host resistance, even though the phyllosphere is likely to be important for pathogens infecting aerial portions of plants, such as Pseudomonas syringae pathovar tomato (Pst). We therefore set out to examine the ability of augmented tomato phyllosphere microbiomes to confer resistance against Pst, the causal agent of bacterial speck. To examine how phyllosphere (above ground) microbiota of tomato plants can affect Pst colonization of leaves, we first infected plants sprayed with different doses of “leaf wash” containing phyllosphere microbes from field-grown tomato leaves or with sterile buffer. Plants were spray inoculated with either undiluted or 85% diluted concentrations (referred to herein as “dose”) of each of six independently generated leaf wash inocula one week before infection (n = 3 plants per treatment). We then pressure inoculated three leaves per plant with Pst and measured bacterial densities after 24 hr in both infected and uninfected leaves using droplet digital PCR (see STAR Methods). In our uninfected leaves, spray-inoculated plants had higher densities of bacteria compared to plants sprayed with sterile buffer (Figure 1A; Mann-Whitney test comparing 16S copy number; Z = −2.138; p = 0.032), confirming that our inoculation methods were successful in establishing a phyllosphere community. For leaves subsequently inoculated with Pst, we observed an effect of phyllosphere augmentation on Pst density, where spray-inoculated plants hosted lower densities of Pst than those inoculated with buffer (Figure 1B; Z = −2.084; p = 0.035). More fine-grained analysis uncovers a significant interaction effect between leaf wash source and dose on Pst density in leaves (general linear model [GLM] with source and dose as fixed effects; F5,23 = 9.037; p < 0.001) but no main effect of either source (F5,23 = 1.068; p = 0.404) or dose (F1,23 = 0.750; p = 0.395). This interaction effect was not observed for overall bacterial densities (16S copy number) of uninfected leaves (F5,23 = 1.585; p = 0.204), where we observed a main effect of leaf wash source (F5,23 = 10.115; p < 0.001) and no effect of dose (F1,23 = 0.194; p = 0.664). The observed interaction effect between dose and source on Pst density led us to repeat this experiment in a more controlled setting using a constructed phyllosphere community comprised of bacteria previously isolated from tomato leaves. To directly test the dose-dependent effect observed in the first experiment, we constructed a phyllosphere community (CC) composed of 12 bacterial strains isolated from field-grown tomato leaves (Table S1). We sprayed plants with one of four concentrations (doses) of the CC inoculum: 100% diluted (sterile buffer); 75% diluted (75% sterile buffer; 25% CC); 50% diluted; or undiluted, followed a week later by Pst inoculation. The undiluted spray was at a concentration similar to what we found in our field-collected leaf washes (approximately 1,000 colony-forming units [CFUs]/mL). We again found that the phyllosphere-sprayed plants had lower Pst densities than buffer-sprayed plants 24 hr after infection (F1,10 = 4.957; p = 0.05). We also detected a significant difference in protection among the doses (Figure 2A; F3,8 = 6.129; p = 0.018), with the largest difference being between 75% diluted and undiluted sprays (Dunnett’s T3; p = 0.019). Dose-dependent responses have been demonstrated in other plant-microbe systems. In tomato plants, Colorado potato beetle larvae utilize their oral bacteria to suppress the plant’s anti-herbivore defenses, with higher oral bacteria doses exhibiting a stronger degree of suppression [16Chung S.H. Rosa C. Scully E.D. Peiffer M. Tooker J.F. Hoover K. Luthe D.S. Felton G.W. Herbivore exploits orally secreted bacteria to suppress plant defenses.Proc. Natl. Acad. Sci. USA. 2013; 110: 15728-15733Crossref PubMed Scopus (294) Google Scholar]. In strawberry plants, an optimal concentration of Stenotophomonas was found to promote root growth, with lower or higher concentrations resulting in a reduction of root growth [1Suckstorff I. Berg G. Evidence for dose-dependent effects on plant growth by Stenotrophomonas strains from different origins.J. Appl. Microbiol. 2003; 95: 656-663Crossref PubMed Scopus (60) Google Scholar]. Our observation that Pst density in plants sprayed with undiluted CC was not significantly different than buffer-sprayed plants (Figure 2A) led us to wonder whether the pattern was driven by microbial density or the diversity of the phyllosphere community. We tested this by generating ten spray inocula that varied in the number of isolates they contained, starting with our full community of 12 isolates and then randomly, but sequentially, dropping one isolate until we had a spray inoculum containing only two isolates (Table S2; Figure S1; note that this does not necessarily represent taxonomic or functional diversity). Plants were sprayed with each inoculum at the undiluted concentration used above, followed by Pst inoculation. Although there was no correlation between number of isolates and Pst colonization (linear regression, F1,18 = 0.043, p = 0.838; Pearson correlation, N = 20, −0.049, p = 0.838), there was evidence for a significant nonlinear relationship (r2 = 0.477; F2,17 = 7.768; p = 0.004), demonstrated by both the 12- and 2-isolate spray inoculums having the lowest Pst colonization (Figure 2B; pairwise comparisons between each community were not evaluated, as identifying specific members of the microbial community that conferred greater resistance was not the goal of our study). The observed nonlinear relationship (peak in Pst density with 5- or 6-isolate sprays, compared to the 12-isolate spray) suggests the protective effect cannot be explained simply by the presence or absence of particular species but may be due to plant-priming effects as more Pseudomonas isolates were dropped from the community. Our CC was comprised of multiple Pseudomonas isolates, and some Pseudomonads have been shown to prime the plant immune system [17Bakker P.A.H.M. Pieterse C.M.J. van Loon L.C. Induced systemic resistance by fluorescent Pseudomonas spp.Phytopathology. 2007; 97: 239-243Crossref PubMed Scopus (372) Google Scholar] and to reduce rhizosphere pathogen density due to both direct inhibition with secondary metabolites and resource competition [18Hu J. Wei Z. Friman V.-P. Gu S.-H. Wang X.-F. Eisenhauer N. Yang T.J. Ma J. Shen Q.R. Xu Y.C. Jousset A. Probiotic diversity enhances rhizosphere microbiome function and plant disease suppression.MBio. 2016; 7 (e01790-16)Crossref Scopus (174) Google Scholar]. Examining this directly will require testing of each isolate independently, which is beyond the scope of the current work. Interestingly, one of our CC isolates includes a known tomato pathogen (Xanthomonas campestris; Table S1), although the Moneymaker tomato plants used in this study have been shown to be resistance to this pathogen [19Ballvora A. Pierre M. van den Ackerveken G. Schornack S. Rossier O. Ganal M. Lahaye T. Bonas U. Genetic mapping and functional analysis of the tomato Bs4 locus governing recognition of the Xanthomonas campestris pv. vesicatoria AvrBs4 protein.Mol. Plant Microbe Interact. 2001; 14: 629-638Crossref PubMed Scopus (77) Google Scholar]. However, this isolate was dropped out when creating the 7-isolate community, which is not associated with any major change in Pst density, and therefore the presence or absence of this specific isolate is unlikely to be the cause for the nonlinear relationship observed (Figure 2B). Due to low plant growth in our experiments, which we attributed to use of autoclaved soil, we added fertilizer (Miracle-Grow Water Soluble All Purpose Plant Food; N:P:K 24-8-16) to the soil and repeated the phyllosphere dose experiment. Surprisingly, the dose-dependent effect we previously observed was abolished after fertilization (Figure S2; p = 0.899). The lack of a fully factorial experiment meant we were unable to rule out experimental effects underlying our observed difference, so we next sought to directly test for interaction effects between fertilization and phyllosphere augmentation on pathogen colonization using our CC approach. We uncovered a main effect of fertilizer application on Pst density (Figure 3; F1,14 = 14.041; p = 0.002) and an interaction effect between fertilizer application and phyllosphere augmentation (F1,14 = 4.635; p = 0.049; note that both dilutions were combined for this analysis). Furthermore, although we did not observe an effect of spray dose on Pst colonization in fertilized plants (F2,6 = 0.221; p = 0.808), we again observed a strong dose effect in unfertilized plants (F2,6 = 13.617; p = 0.006). This surprising observation led us to consider the importance of resources available to microbes on the leaf. Previous studies have demonstrated that fertilizer use can increase plant size and decrease leaf sugar concentration [20Tetsuo H. Effects of nitrogen, phosphorus and potassium in culture solution on the head yield and free sugar composition of cabbage.J. Jpn. Soc. Hortic. Sci. 1989; 58: 595-599Crossref Google Scholar, 21Takebe M. Ishihara T. Matsuno K. Fujimoto J. Yoneyama T. Effect of nitrogen application on the contents of sugars, ascorbic acid, nitrate and oxalic acid in spinach (Spinacia oleracea L.) and komatsuna (Brassica campestris L.).Japan J. Soil Sci. Plant Nutr. 1988; 66: 238-246Google Scholar]; further, leaf surface sugar concentration was previously shown to positively correlate with phyllosphere bacteria density [22Mercier J. Lindow S.E. Role of leaf surface sugars in colonization of plants by bacterial epiphytes.Appl. Environ. Microbiol. 2000; 66: 369-374Crossref PubMed Scopus (244) Google Scholar]. However, our measurements on uninfected leaves did not show a strong relationship between bacterial abundance and fertilizer treatment (data not shown); therefore, it is unlikely that increased leaf sugar concentration due to fertilization can solely explain the overall lower densities of Pst we observed in the unfertilized plants. Interestingly, the initial experiment using natural microbial communities (where we did observe microbiome-mediated protection, although the effect was dose dependent) was carried out with fertilized soil. The difference among experiments may be explained by the strength of microbiome-pathogen competition across communities. Importantly, all of our CC isolates were cultured on KB media, on which Pst grows well, which may have led us to select for bacteria more likely to directly compete with Pst for resources. Coexistence of leaf epiphytes and Pst has previously been shown to inversely correlate with the degree of similarity in carbon source utilization profiles [23Wilson M. Lindow S.E. Coexistence among epiphytic bacterial populations mediated through nutritional resource partitioning.Appl. Environ. Microbiol. 1994; 60: 4468-4477Crossref PubMed Google Scholar]. Similarly, microcosm experiments suggest that resource availability affects microbial community dynamics and invader success, with catabolically similar residents out-competing invaders in low-resource environments and fast-growing residents competing with invaders in high-resource environments [24Yang T. Wei Z. Friman V.-P. Xu Y. Shen Q. Kowalchuk G.A. Jousset A. Resource availability modulates biodiversity-invasion relationships by altering competitive interactions.Environ. Microbiol. 2017; 19: 2984-2991Crossref PubMed Scopus (41) Google Scholar]. Our six naturally derived leaf wash communities almost certainly had higher functional and metabolic diversity than did our CC, so a change in leaf surface sugars likely would not have had as strong of an effect on these communities as it would on the CC. For an overview of designs across experiments, please see Figure 4. The observed interaction between microbiome protection and fertilization treatment (Figure 3) suggests that the host’s nutritional environment (e.g., fertilizer application) in agricultural settings could fundamentally alter how selection acts on the host-microbiome interaction over time, especially in the presence of pathogens. Another study using a synthetic community similarly demonstrated that the root-associated microbiome can enhance the plant’s immune system when grown under limited phosphate [25Castrillo G. Teixeira P.J.P.L. Paredes S.H. Law T.F. de Lorenzo L. Feltcher M.E. Finkel O.M. Breakfield N.W. Mieczkowski P. Jones C.D. et al.Root microbiota drive direct integration of phosphate stress and immunity.Nature. 2017; 543: 513-518Crossref PubMed Scopus (434) Google Scholar]. And recent work from Arabidopsis suggests that soil composition can alter plant susceptibility to P. syringae [26Hassan J.A. de la Torre-Roche R. White J.C. Lewis J.D. Soil mixture composition alters Arabidopsis susceptibility to Pseudomonas syringae infection.Plant Direct. 2018; 2: e00044Crossref Scopus (6) Google Scholar]. Results from these and the present studies add support to the importance of host nutritional state on host-microbiome interactions The observed effect of fertilizer on microbiome-mediated protection could be the result of changes occurring at the plant level, such as altered hormones or expression of resistance genes, and/or of changes in competition among microbiota due to altered resource availability. We took an in vitro approach to gain a better understanding of the processes at work by growing monocultures of Pst, 85% diluted CC, or undiluted CC in filtered leaf homogenate generated from fertilized or unfertilized plants that were sprayed with each of the three CC doses. Our results indicate a significant interaction effect between fertilizer application and CC dose of donor plants on growth of both Pst and CC in vitro, with leaf homogenate from fertilized plants leading to lower growth than unfertilized plants when sprayed with diluted CC (p = 0.009; Figures S3A and S3B), supporting previous work linking nitrogen nutrition to resistance against P. syringae in tobacco [27Gupta K.J. Brotman Y. Segu S. Zeier T. Zeier J. Persijn S.T. Cristescu S.M. Harren F.J. Bauwe H. Fernie A.R. et al.The form of nitrogen nutrition affects resistance against Pseudomonas syringae pv. phaseolicola in tobacco.J. Exp. Bot. 2013; 64: 553-568Crossref PubMed Scopus (96) Google Scholar]. Conversely, leaf homogenate from fertilized plants supported greater bacterial growth than unfertilized plants when sprayed with a high dose of CC, suggesting that the unfertilized leaf homogenate had fewer available resources. The plants were spray inoculated one week before leaves were homogenized, so it is possible that the higher densities of phyllosphere bacteria from the undiluted spray exhausted many of the resources over the course of the week, leading to a reduction in bacterial density by sampling time. Interestingly, we did not see any differences in Pst growth in leaf homogenate from unfertilized plants sprayed with buffer compared to leaf homogenate from unfertilized plants sprayed with the CC (Figure S3B), suggesting that our observed loss of protection in planta is not solely due to changes in the host plant due to fertilization. Rather, the loss of protection is likely a result of an interaction between fertilization and the phyllosphere microbiome. Importantly, when we grow Pst in the same leaf homogenates, but this time in the presence of our CC (i.e., as a co-culture), we do observe an interaction effect between fertilization and dose, as observed in planta (F2,12 = 24.24; p < 0.001; Figure S3C). Moreover, for unfertilized plants, Pst densities were lowest when grown in leaf homogenate that had been sprayed with a diluted CC, suggesting that, along with fertilizer-induced changes in the plant, microbial interactions are a key factor driving the observed phyllosphere-mediated protection. Our results suggest that augmentation of the phyllosphere does increase infection resistance, as plants sprayed with either a natural (Figure 1) or constructed (Figure 2A) microbial community had significantly less pathogen colonization than plants sprayed with sterile buffer. We additionally showed that phyllosphere spray dose has a more nuanced effect on protection, which is influenced by the source of augmented microbiota (Figure 1B), but not directly related to the number of isolates within the microbial community (Figure 2B). Intriguingly, we did not find that higher doses conferred greater protection; in our CC experiments, we found that the more diluted microbiome spray conferred the greatest protection (a result that was also observed for half of the “natural” microbiome communities; Figure 1B). Finally, our CC results suggest that microbe-microbe competition (resource community dynamics) may play an important role in protection, as the addition of fertilizer abolished the observed microbiome-mediated protection (Figure 3). The robust protection observed across different communities suggests that microbiome-mediated protection is most likely not caused by one single microbial species or strain but rather by the presence of the community itself. This result is also supported by our finding that protection is not linearly correlated with microbiome diversity in our CC (Figure 2B). However, our observed interaction effect between dose of the spray inoculum and leaf wash origin suggests that the presence of the community itself is also not the sole factor in mediating resistance. Work from the rhizosphere suggests that some specific members of the community can have more direct effects on promoting disease resistance than others, for example, by sequestering nutrients, such as iron, and restricting the growth of potential pathogens [4Innerebner G. Knief C. Vorholt J.A. Protection of Arabidopsis thaliana against leaf-pathogenic Pseudomonas syringae by Sphingomonas strains in a controlled model system.Appl. Environ. Microbiol. 2011; 77: 3202-3210Crossref PubMed Scopus (322) Google Scholar, 18Hu J. Wei Z. Friman V.-P. Gu S.-H. Wang X.-F. Eisenhauer N. Yang T.J. Ma J. Shen Q.R. Xu Y.C. Jousset A