How to Find a New BCA

Two fundamentally different approaches are commonly used in attempts to identify novel BCAs. These are, first, the indirect screening of microbial libraries for antagonistic properties in planta or in silico and, secondly, isolation of organisms from the habitat where the product would be used and then screening them directly for activity in planta (Collinge et al., 2019; Knudsen et al., 1997; Köhl et al., 2011; Teperi et al., 1998).

The in vitro approach has been used for high-throughput screening of existing collections of strains for activities against one or more pathogens. We do not know of documented examples of successful products for plant protection from this approach. Direct screening is less suited to high throughput but facilitates the identification of organisms where the mode of action involves plant responses, for example, induced resistance or the ability of an organism to colonize and compete in plant niches (e.g., rhizosphere, phyllosphere, endosphere or in wounded tissue). The advantage of the in vitro approach is that many strains can be tested for the production of antimicrobial metabolites and, for example, mycoparasitic (also termed hyperparasitic) activity. However, both positive and negative results may be misleading as one cannot be sure that the mechanisms would be active in the plant, nor, conversely, that useful mechanisms are not activated in vitro. The latter has led in many cases to discarding promising BCAs based on in vitro screening. There have been many disappointments but a few promising BCAs have been identified. The in planta approach, in its extreme form, involves testing potential BCAs under field conditions. This has been successful for selecting some isolates that are now commercialized. However, in practice, it is necessary in most cases to develop tests on plants in growth chambers or greenhouses, or even, in a few examples, on leaves or wheat heads. Although these methods are a compromise, they can simulate conditions that are comparable to the field and can often be carried out throughout the year instead of depending on a brief growth season. Thus, they can give a reasonable level of throughput to select promising candidates for extensive tests in production systems.

Recently, the availability of next-generation sequencing tools has allowed research on biocontrol agents to take a directly functional approach. In C. rosea and species of Trichoderma, for example, genomics and metabolomics are currently allowing the discovery and investigation of a vast repertoire of specialized metabolic pathways. Study of the roles these metabolites play in the environmental and biotic relations of these organisms may represent a new route to development of BCAs. However, genomic or metabolomics screens are necessarily restricted to looking for signatures derived from study of organisms known to have biocontrol activity. Therefore, such screens should follow on from in planta selection of potential organisms and be a stand-alone approach.

Useful organisms are not only found by targeted searches. For example, a C. rosea strain (IK726), originally found in the rhizosphere of a barley root, is effective against many diseases of diverse organs in a wide range of hosts ranging from brassicas to strawberry, oak and cereals. Similarly, Serendipita indica (syn. Piriformospora indica), a plant growth-promoting organism, was found in the root of a desert shrub, but has positive effects for protection against both abiotic stress and attack by certain pathogens in many plant species in very different environments. It is commercially available both for biological control and as a biofertilizer. In both cases, several mechanisms of action may be operating. Another example is the isolate Trichoderma gamsii T6085, isolated from an uncultivated soil in Crimea but effective, when applied on spikes at anthesis, in reducing the incidence of Fusarium head blight on wheat. Like several examples quoted here, it also possesses several quite diverse modes of action, from mycoparasitism to induction of plant defence responses.

Different pathogen lifestyles may necessitate different strategies for identifying and isolating appropriate BCAs. For example, biotrophic parasites of a fungal (or bacterial) pathogen would benefit from the development of methods for isolating and subsequently cultivating them on bait organisms. This is especially true for viral BCAs that can only live as parasites, for example, bacteriophage on bacteria and mycoviruses and other biotrophic hyperparasites on fungi. It can also be a challenge to isolate specialized organisms that may be slow growing or require a host to grow at all; equally, it may be difficult to exploit a slow-growing BCA.

IMPROVED EFFICACY – A KEY TO IMPLEMENTATION?

One of the challenges of biological control is reliable efficacy. Biological control is often considered to be less reliable and efficient than chemical control or host resistance, probably because exposure to the external environment is largely an uncontrollable variable. A counterargument is that some types of biological control (unlike some mechanisms of host resistance) may have an effect against multiple diseases, especially where induced resistance or resistance priming is an underlying mechanism. In addition, it has been shown that, for example, C. rosea can be a mycoparasite of diverse fungal plant pathogens such as Fusarium graminearum and Botrytis cinerea. This seems to rely on the response of both general-purpose and specific gene expressions in C. rosea depending on which fungal species it parasitizes, indicating that the BCA can work through different modes in biocontrol interactions.

Most successful BCAs are effective competitors in the harsh biotic environment of soil and in the plant holobiome (the combination of the plant and its associated microbiome), as they have evolved mechanisms for tolerating toxins from other organisms and are adapted to stressed conditions in those environments, which include growth on roots, stems, leaves and flowers and in wounded tissue. Endophytes, defined as microorganisms colonizing the interior of plants (the endosphere) without causing disease, are adapted to the ecological niche of the endosphere and are also partly protected from the external environment (and colonize the same niche as pathogens). Therefore, it is suggested that endophytes have the potential to be more consistent as BCAs than purely epiphytic organisms, especially those in the phyllosphere.

However, this hypothesis is speculative, based on knowledge that many plant pathogens compete poorly, with an advantage only inside the plant. The hypothesis remains to be demonstrated experimentally for potential endophytic BCAs. One example is the use of endophytic fungi associated with the invasive weed Japanese knotweed (Fallopia japonica). Some endophytes can increase the effectiveness of the rust Puccinia polygoni-amphibii var. tovariae as a potential control agent of F. japonica. Another example concerns grass endophytic fungi of the genera Neotyphodium and Epichloë that can produce alkaloid mycotoxins (e.g., ergovaline) affecting ruminants (especially cattle and sheep). However, some Neotyphodium and Epichloë isolates can provide a very high level of protection of the host plant against insect pests (e.g., Argentinian weevil) or fungal pathogens of grasses including Rhizoctonia spp., Bipolaris sorokiniana and Curvularia lunata; Sclerotinia homoeocarpa; and Fusarium oxysporum. This appears to be mediated through priming of defences.

Many endophytes only enter the apoplast, but may still have a control function there, either directly inhibiting the pathogens or indirectly by inducing or priming defence responses in the plant. However, these organisms might also be adapted to function outside the plant, as is known for Trichoderma spp. and Clonostachys spp. As good root colonizers, these fungi are also adapted to the harsh environment of the rhizosphere. The isolation of an organism from the rhizosphere or endosphere does not necessarily mean that it can only colonize that environment. Nevertheless, most endophytes will be specialized to some extent to survive inside a plant and would be predicted to compete poorly with microbes outside the plant endosphere. Despite this, there is a continuum in lifestyle, and the same organism may behave as an endophyte, epiphyte or pathogen under different environmental conditions. Of course, this must be considered in the selection of potential BCAs to prevent accidental selection of plant pathogens.

Consortia, that is, mixtures of microorganisms, are receiving increasing attention as a way of addressing multiple problems. For example, in experiments to control both the fungal pathogen Fusarium culmorum and root-feeding insects of wheat, the insect pathogen Metarhizium brunneum was combined with the fungal BCA C. rosea; effects were observed on both the insect pests and F. culmorum, although the efficacy was reduced compared to treatment with either M. brunneum or C. rosea separately. It is tempting to assume that a mixture of BCAs would be more effective than a single agent. However, modelling suggests that, depending on exactly how the organisms compete and act, this may often be untrue. Different associations can have opposite or antagonistic effects; thus, the ability of S. indica to control R. solani or F. oxysporum infections depended on associated bacteria. It has also been difficult, except in a few cases, to demonstrate significant additional or synergistic biocontrol efficacy by combining different BCAs in consortia. One challenge is to ensure that the different agents can operate together under variable environmental conditions and do not have incompatible modes of action. For example, two BCAs acting mostly by bulk nutrient competition would be expected to counter each other's activity. Thus, the idea of forming complex consortia—synthetic biomes or synthetic communities, abbreviated SynComs—consisting of several different microorganisms with biocontrol effects that could be used as mixtures does not seem to be the most promising route. It can be predicted that there will be selection within a consortium to favour the best adapted to a particular environment and that the dominant consortium members will change following treatment in response to the local environment. This could be an advantage by making consortia more adaptable to different environments. Nevertheless, in special cases, several products comprising bacteriophage consortia have been released for combating bacterial disease.

BCAs are an attractive component in management of postharvest disease, by application at harvest or shortly before. An example is Alfasafe and similar products for controlling aflatoxin contamination using nontoxigenic Aspergillus flavus strains to compete with the toxigenic forms. Consumer sensitivity over the use of artificially synthesized chemical application is greater for applications made postharvest than during crop growth; however, the environment is usually less variable or much less variable than in the field, and doses applied can be much more uniform, assisting the use of BCAs that act by resource competition or breakdown of mycotoxins produced by other microbial species. Nevertheless, biological control using applications of BCAs postharvest is currently not allowed in the EU. In contrast, postharvest BCAs have been used for many years in the United States, for example, to protect soft fruit from postharvest decay before they reach the consumers. Examples include use for postharvest control of various fungal pathogens of citrus fruit, pome fruits, cherries and potatoes. Postharvest BCA treatment of soft fruit for controlling Penicillium and Aspergillus species and other spoilage pathogens like B. cinerea and Rhizopus spp. therefore seems to be an important way forward in the EU in view of its successful commercial use in the United States.

Product spoilage can, in some cases, also be avoided by BCA treatments before harvest, depending on the epidemiology of the pathogen–host association. Postharvest problems with mycotoxin production may also be addressed long before harvest by reducing the populations of toxin-producing organisms or the rate at which they produce the toxins, and by increasing the rate and extent that mycotoxins are degraded. For example, mycotoxin production by ear-inoculated F. graminearum and F. culmorum in wheat was greatly reduced in outdoor (but pot-grown) wheat inoculated with S. indica at sowing. This must have been caused by an indirect effect on host resistance because S. indica remained restricted to the roots. The doses of BCA culture used in that study were very large (equivalent to 60 g/m2 or 600 kg/ha), but the effect is intriguing. There are interesting examples of beneficial fungi able to degrade mycotoxins, for instance, the ability of C. rosea to detoxify the mycotoxin zearalenone (ZEA) through the enzyme zearalenone lactonohydrolase has been demonstrated and there are promising results from the field where C. rosea has reduced the deoxynivalenol content in harvested wheat grain. Similarly, the ability of some Trichoderma isolates to degrade mycotoxins has recently been studied. In the case of Trichoderma aggressivum, its zearalenone lactonohydrolase was expressed in Escherichia coli BL21 (DE3) and successfully purified.

Postharvest pathogens on soft fruit such as mycotoxin-producing species of Aspergillus and Penicillium are not likely to be controlled efficiently preharvest, even though it is often suggested that application of beneficial organisms preharvest can reduce mycotoxin accumulation postharvest. There are exceptions. This is the case for beneficial yeasts such as A. pullulans, whose preharvest application on grape resulted in a reduction of ochratoxin A contamination by around 95%. Another interesting example is Kluyveromyces thermotolerans, able to control the growth of Aspergillus carbonarius and Aspergillus niger in the field by up to 100% and to reduce mycotoxin accumulation by up to almost 80%.