Plant-associated Acinetobacter spp. have plant growth-promoting properties that include antagonism toward fungi and the ability to solubilize phosphate and produce the plant hormone gibberellic acid . Acinetobacter spp. are highly abundant in the floral nectar microbiome of Citrus paradisi and other plant species and were identified as a core member of the grapevine xylem sap microbiome . Their significant increase in relative abundance in the leaf microbiome at the time of flowering in citrus suggests a potential synergy between the foliar and floral microbiomes. Acinetobacter was also predicted to be a keystone taxon and was a major link between the flowering community and fruit set/development community clusters in our network. This enrichment in Acinetobacter may be simply due to selection imposed on the microbial community by the local plant environment, but it is tempting to speculate that Acinetobacter spp. provide an exogenous service to the plant by producing gibberellic acid and biologically available phosphorus to promote flowering that is in phase with its host’s phenological development. This hypothesis that the plant environment selects for taxa within its foliar microbiome that, in turn, promote its own reproductive growth warrants future inquiry. Specific bacterial enrichments also occurred at bloom time in grapevine, square plant pots further supporting the speculation of a microbial role in plant growth development that is in sync with plant developmental stage . We also observed signatures that indicate that specific taxa were depleted in relative abundance during flowering but enriched during fruit set.
Phylogenetic reconstruction of these taxa indicated that the majority of the taxa belonging to the Actinobacteria phylum were significantly depleted during flowering but subsequently enriched when trees begun to set fruit. This phylogenetic conservation of depletion/enrichment patterns within the Actinobacteria clade indicates that this is a nonrandom fluctuation within the microbiome structure associated with the transition from flowering to fruit production. As citrus trees set fruit, the fruits themselves begin exporting and importing hormones, such as indole acetic acid and cytokinins, respectively . This results in a change in hormone levels in leaves as well. Phytohormones can affect microbiome composition by being directly utilized as a carbon source or through other undefined mechanisms . We speculate that these hormonal shifts may place selective pressure on the foliar microbial community and lead to significant enrichments of Actinobacteria during fruit set and development. Specific differentially abundant taxa within the Actinobacteria clade that followed this pattern included Corynebacterium, Dietzia, Georgenia, and Ornithinimicrobium. Members of these genera can fix nitrogen and produce IAA, both of which are important supporters of fruit development . Thus, it is tempting to speculate that these taxa could play a role in coregulating fruit development in a manner that is synergistic with the host’s production of reproductive hormones. The biological role of Actinobacteria in the foliar microbiomes of plants is not well understood, but overall species richness was conserved across all phenological stages, except for fruit set, indicating that this phenological stage allows for microbial enrichments of specific taxa, particularly those belonging to the Actinobacteria phylum.
Genera outside the Actinobacteria clade were also depleted in leaves during floralbud development and full flowering, including Bacillus, Methylobacterium, Romboutsia, and Sphingomonas. Notably, Romboutsia and Sphingomonas were predicted to be keystone taxa in our microbe-microbe interaction network analysis, and all are in the top 20% highest betweenness centrality scores. Depletion of these taxa during flowering may have cascading effects that influence microbial species turnover by allowing other taxa, such as the Actinobacteria, to flourish during subsequent developmental stages, like fruit set. This suggests that microbial turnover in the foliar microbiome is mediated by selective pressures imposed by the plant developmental stage in conjunction with microbemicrobe interactions to modulate community diversity and composition. Further experimentation is needed to confirm these interactions. Microbial dispersal events can drive microbial turnover and influence the relative abundance of endogenous taxa in the community. Full flowering is a dynamic phenophase in plant development where there are frequent interactions between plants and pollinator species that rely on floral resources, like nectar and pollen. These macro-level interactions can also have effects at the microorganismal level. Pollinator visitation alters flower surface, nectar, and subsequent seed microbial community composition . During flowering, we observed striking microbial enrichments of bacteria taxa belonging to the Betaproteobacteria and Gammaproteobacteria clades that include Gilliamella, Snodgrassella, Bifidobacterium, and Frischella.
These enrichments in these anaerobic taxa were unique to the flowering phenophase and quickly declined following flowering, suggesting they are immigrants to the community and not endogenous members of the native microbiome. Moreover, these anaerobic taxa are prevalent in the bee gut microbiome . We hypothesize that these taxa immigrated into the citrus microbiome via a dispersal event during bee visitation. We consider this external influence host phenology-associated because phenophase-specific plant morphology regulate this diffuse interaction. Bacteria can be introduced to plants by bees and potentially migrate from the flower to the vascular bundles resulting in systemic movement within the plant . Leaf carbohydrate content is highest during flowering, which may promote the growth of these fermenting bacteria . Notably, Bifidobacterium was the only core leaf genus from the Actinobacteria phylum that was enriched during flowering, whereas the other 19 Actinobacteria taxa were depleted during flowering. This further supports the hypothesis that Bifidobacterium was introduced via a dispersal event and is not part of the endogenous microbiota like the other taxa in the plant-associated Actinobacteria clade. Nectar-inhabiting bacteria can influence nectar volatile profiles that, in turn, influence pollinator visitation preferences , and it would be interesting to determine if these putative immigrants contribute to shifts in nectar volatile profiles that affect bee feeding behaviors. The next frontier in microbiome research is to determine the functional roles that microbes play in microbe-microbe and host-microbiome interactions. Martiny et al. found that conservation of microbial traits was linked more strongly to vertical phylogenetic relatedness of the microorganisms within a microbiome than to traits that are shared among taxa by horizontal gene transfer . Similarly, we also observed phylogenetic conservation within microbial enrichments, suggesting that those groups of related organisms play similar functional roles during specific phenophases or across several phenophases. We speculate that taxa with high stability across phenophases may serve a community-stabilizing function, while low stability or phenophase-specific core microbes likely have more specialized, transient roles in the community. Because microbes can alter host phenology , which is a critical factor in plant health and productivity, incorporation of microbial presence/absence and patterns of enrichment into plant phenological models may improve phenophase timing predictions once the functional roles of these microbes are determined. This information could also lead to the commercialization of biofertilizers for horticultural purposes that could be applied at specific plant life stages to enhance crop productivity.Various genetically encodable reporters have been developed to monitor gene expression, protein subcellular localization, protein stability, hormonal signaling, and impacts of environmental signals. The green fluorescent protein and its derivatives such as RFP, mCherry, and YFP have many applications as reporters for gene expression or as fusion proteins. Although GFP is easy to use, plastic potting pots it needs light sources to visualize the fluorescence signals. The β-glucuronidase reporter has been widely used in plants for monitoring gene expression patterns and as a reporter for hormonal signaling. For example, DR5-GUS transgenic lines are commonly used to monitor auxin distribution and auxin signaling. Luciferase is another broadly used reporter in both animals and plants. Both GUS and luciferase require the addition of expensive substrates X-Gluc and luciferin, respectively. Whereas the traditional reporters have been very useful, they have limitations. Fluorescent proteins are often monitored under a microscope, rendering it less useful in analyzing plants in natural growing fields or analyzing large samples such as a tree. GUSstaining is invasive and often requires sacrifice of the plants. Luciferase can be used noninvasively, but it requires a special camera and spraying the expensive substrate. It is also not very practical to use them in fields.
GUS and luciferase may not be optimal for sterile conditions such as tissue culture because addition of substrates increases the chance for contamination of microbes. Therefore, there is a need to develop new reporter systems that can be widely used to monitor cellular activities noninvasively, continuously, and costeffectively. For the past few years, gene editing has been widely used in basic research and crop improvement. A visible marker for transgenes will greatly accelerate the isolation of edited plants that no longer harbor the gene editing machinery. Plants produce many colorful compounds that potentially can serve as reporters. For example, anthocyanins display bright red-blue colors and anthocyanin-producing rice plants have been used to generate interesting patternsin rice field. However, synthesis of anthocyanins requires multiple enzymes and varies greatly among different plants. It is difficult to use anthocyanin biosynthesis pathways as a universal visible reporter. Betalains are a class of plant natural products derived from the aminoacid tyrosine. The bright red color seen in beets, dragon fruit, Swiss chard, and other plants is resulted from accumulation of betalains. Biosynthesis of betalains has been well studied and only needs three enzymatic reactions to convert tyrosine into betalain. Tyrosine is first hydroxylated on the benzene ring, resulting in L-3,4-dihydroxyphenylalanine . The reaction is catalyzed by the P450 oxygenase CYP76AD1 . L-DOPA can be further oxidized into cyclo-DOPA by CYP76AD1 . Alternatively, LDOPA is catalyzed by L-DOPA 4,5-dioxygenase into betalamic acid, which is subsequently condensed with cyclo-DOPA into betanidin. The condensation reaction does not require an enzyme . Finally, a sugar moiety is added to betanidin by a glucosyltransferase to generate the colorful betalain . Betalain has a very bright red color, which potentially can serve as a reporter to track gene expression or to visualize transgenic events. Because every cell contains the amino-acid tyrosine, exogenous application of tyrosine to tissues may not be required. We hypothesized that betalain would be a more convenient reporter than the aforementioned reporters. It is visible to naked eyes without any needs for special equipment. It does not require processing samples and it allows continuously monitoring events throughout the life cycle of an organism. Moreover, it is applicable to large plants grown under normal field conditions. Herein, we synthesize an artificial open reading frame named RUBY that when expressed can produce all of the enzymes required for betalain biosynthesis. We show that RUBY is a very effective marker for noninvasively selecting transformation events in both rice and Arabidopsis. Moreover, we show that RUBY can be used to visualize gene expression without any chemical treatments or special equipment, providing useful tools for visualizing gene expression in large plants under natural field growth conditions.We transformed the 35S:RUBY construct into Arabidopsis using Agrobacterium-mediated floral dipping. Two days after floral dipping, we noticed that the transformed plants displayed patches of red color , indicating that the RUBY cassette was functionally expressed and that RUBY may be used to monitor transient Arabidopsis transformation. Once the seeds from the Agrobacterium-dipped plants were harvested, transgenic seeds could be easily differentiated from non-transgenic seeds . The transformed seeds had a dark red color , demonstrating that RUBY can be used as a visual selection marker for transgenic events in Arabidopsis. We previously used mCherry as a very effective marker to select transgenic events , which requires a dissecting microscope with fluoresence capability. RUBY is a better option because it does not require special equipment. The 35S:RUBY plants produced sufficient amount of betalain to become visually evident . Consistent with previous reports that CaMV 35S promoter is constitutively active, we observed red color in all tissues throughout the plant life cycle . We also expressed RUBY reporter under the control of the Maize UBIQUITIN promoter, which has been widely used to overexpress genes in monocots. Similar to 35S:RUBY plants, UBQ:RUBY plants were also visibly red in leaves, stem, and flowers . These results clearly demonstrated that RUBY could be expressed in Arabidopsis and that our RUBY reporter was able to functionally re-constitute the betalain biosynthetic pathway. We expressed RUBY using the seed specific At2S3 promoter, which we previously used to drive mCherry expression in Arabidopsis to facilitate the selection of transgenes. As shown in Fig. 2c, the transgenic plants were indistinguishable from wild type plants. When we checked the seeds in a silique from an At2S3:RUBY T1 plant, RUBY-expressing seeds displayed strong red color, whereas the non-transgenic seeds were green . RUBY can be conveniently used to select single T-DNA insertion events by analyzing the ratio of red seeds to green seeds, which should be ~3:1 for single insertions.