Developing new RNA interference technologies to control fungal pathogens

Abstract

Current agricultural output is challenged by considerable losses in crop yield and post-harvest storage due to fungal infection. Traditional chemical fungicides used to treat these fungi can be ineffective and harmful to the environment if not used properly. With fungicide resistance increasing in fungal pathogens, new environmentally friendly and sustainable technologies are required to manage diseases on the world’s most important crops. RNA interference (RNAi) is an intrinsic cellular mechanism, mediated by double-stranded RNA (dsRNA), which can suppress protein expression through targeted destruction of mRNAs. With recent advances in dsRNA delivery or expression in plants, this mechanism has the potential to provide alternative disease management strategies. Examples of RNAi-based control to manage pathogenic fungal species are steadily increasing, and the technology offers new options to increase species-specificity and/or potency against fungi for which existing fungicides have been ineffective. RNAi technology can be adapted to provide either robust and multi-crop plant protection using topical sprays or can provide more durable resistance through transgene expression of dsRNAs within susceptible plant tissues. Using RNA sequencing to identify fungal gene targets, RNAi-based control technology continues to show promise as an alternative to traditional agrochemicals for crop protection.

Résumé

Les infections fongiques menacent la production agricole actuelle, et ce, tant à cause des immenses pertes de rendement que de celles survenant durant l’entreposage des récoltes. Lorsqu’ils ne sont pas utilisés correctement, les fongicides de synthèse traditionnels utilisés pour combattre les champignons peuvent être inefficaces, en plus d’être nocifs pour l’environnement. Avec la résistance accrue des agents pathogènes fongiques aux fongicides, nous avons besoin de nouvelles technologies respectueuses de l’environnement et durables pour lutter contre les maladies des cultures les plus importantes sur la planète. L’interférence ARN (iARN) est un mécanisme cellulaire intrinsèque médié par un ARN à double brin (ARNdb) qui peut supprimer l’expression d’une protéine par la destruction ciblée des ARNm. Avec les récents progrès réalisés dans le domaine de la délivrance des ARNdb ou de l’expression dans les plants, ce mécanisme peut fournir des stratégies de rechange pour lutter contre les maladies. Les exemples de lutte contre des espèces d’agents pathogènes fongiques, basée sur l’iARN, sont de plus en plus nombreux, et la technologie offre de nouveaux choix permettant d’accroître la spécificité relative aux espèces ou la puissance à l’égard de champignons là où des fongicides traditionnels s’avèrent inefficaces. La technologie découlant de l’iARN peut être adaptée pour fournir soit une protection énergique s’appliquant à plusieurs cultures à l’aide de pulvérisations topiques, soit une résistance plus durable par l’expression de transgènes des ARNdb dans les tissus végétaux sensibles. En utilisant le séquençage de l’ARN pour identifier les cibles génétiques fongiques, la technologie basée sur l’iARN continue d’être prometteuse en tant que solution de rechange aux produits agrochimiques utilisés à ce jour pour protéger les cultures.

Keywords:

• Fungal pathogens
• green technologies
• RNA interference
• RNA sequencing

Mots clés:

• Agents pathogènes fongiques
• interférence ARN
• séquençage de l’ARN
• technologies vertes

Introduction

Access to safe, healthy and sustainable food sources is one of the defining challenges of our time. Given the rapid increase of the human population, it is estimated that we will require an additional production of 200 billion calories, equating to a 100–110% increase in crop production to meet nutritional needs by the year 2050 (Tilman et al., 2011; Bebber & Gurr, 2015; United Nations, 2017). Currently, 40% of ice-free land is used to grow crops; however, global climatic changes challenge our current production systems by decreasing yield potentials and shrinking arable land resources (Pugh et al., 2016; Myers et al., 2017). Moreover, biotic agents, such as necrotrophic and biotrophic fungi, further complicate agronomic production, causing global losses of up to 30–40% in crop yield in-field and post-harvest (Bebber & Gurr, 2015; Myers et al., 2017). With rapid globalization and migration across the globe, fungal pathogens are predicted to spread rapidly to virgin lands, presenting new challenges for crop production globally (Bebber et al., 2014).

Table 1. Summary of publications utilizing RNA interference technology to target plant pathogens.

Plant species	Target pathogen	Gene target	Vector/delivery method into plant	References
Nicotiana tabacum	Sclerotinia sclerotiorum	CHITIN SYNTHASE	pCAMBIASynchit2 / Agrobacterium tumefaciens	Andrade et al. 2016
Zea mays	Diabrotica virgifera virgifera (WCR)	V-ATPASE A	pMON94805 / Agrobacterium tumefaciens	Baum et al. 2007
Phaseolus vulgaris	Bean golden mosaic virus	AC1 REPLICASE	pBGMVRNAiAHAS / Particle bombardment	Bonfim et al. 2007
Musa acuminata	Fusarium oxysporum	VELVET and FUSARIUM TRANSCRIPTION FACTOR 1	pCAMBIA-1301 / Agrobacterium tumefaciens	Ghag et al. 2014
Hordeum vulgare	Fusarium graminearum	CYTOCHROME P450 LANOSTEROL C-14α-DEMETHYLASE	Foliar	Koch et al. 2016
Solanum tuberosum	Leptinotarsa decemlineata	ACTIN	Foliar	Miguel et al. 2015
Musa acuminata	Mycosphaerella fijiensis, Fusarium oxysporum	NUCLEAR CONDENSIN, COATOMER ALPHA, DNA-DIRECTED RNA POLYMERASE, ACTIN CORTICAL PATCH 2/3, COATOMER ZETA, CAP METHYLTRANSFERASE, GTPASE BINDING PROTEIN, PROTEASOME PRE4, RIBOSOMAL RNA, DNA POLYMERASE ALPHA/DELTA SUBUNIT, ADENYLASE CYCLASE, PROTEIN KINASE C, FRQ-INTERACTING RNA HELICASE	Foliar	Mumbanza et al. 2013
Triticum aestivum	Puccinia triticina	CALCINEURIN B REGULATORY SUBUNIT, MAPK1, CYCLOPHILIN1	BSMV Viral Vector/VIGS – Barley stripe mosaic virus	Panwar et al. 2013
Zea mays	Aspergillus flavus	POLYKETIDE SYNTHASE	pTF101.1/Agrobacterium tumefaciens	Thakare et al. 2017
Arabidopsis thaliana Solanum lycopersicum	Botrytis cinerea	DICER-LIKE 1/2	1/2pHELLSGATE8/Agrobacterium tumefaciens	Wang et al. 2016
Gossypium hirsutum	Verticillium dahliae	V. DAHLIAE HYGROPHOBINS1	pBI121/Agrobacterium tumefaciens	Zhang et al. 2016

Traditional chemical treatments used to combat fungal disease epidemics and to control necrotrophic pathogens such as Sclerotinia sclerotiorum have achieved only mixed success (Huzar-Novakowiski et al., 2017). When used at sub-lethal levels, common fungicides like boscalid, iprodine, thiophanate methyl, azoxystrobin and pyracostrobin, were associated with increased mutation rates of up to 60-fold in S. sclerotiorum (Amaradasa & Everhart, 2016). Subsequent treatments of fungicides resulted in reduced fungal sensitivity due to the accumulated genetic mutations, suggesting that inadequate dispersal of fungicides on crops could promote the development of fungicide resistance in pathogens (Amaradasa & Everhart, 2016). Development of resistance can lead to greater or more frequent applications of fungicide, and potentially stronger selection for further resistance, attributing to the rise of fungicide-resistant S. sclerotiorum, Botrytis cinerea and Magnaporthe oryzae (Castroagudín et al., 2015; Penaud & Walker, 2015; Rupp et al., 2017). To break this cycle of developing resistance to our current fungicides, new methods to control disease outbreaks using sustainable technologies are needed.

Despite the suggestion that precise chemical applications could reduce environmental impacts, fungicide use can have deleterious effects on the surrounding agro-ecological landscape due to biocidal non-target effects, as well as its dispersal and persistence within the environment (Smalling et al., 2013; Sabatier et al., 2014; Le Cointe et al., 2016). The use of fungicidal compounds was noted to alter the structure and function of aquatic communities, culminating in severe physiological pathologies, such as increased mortality, reduced reproductive rates and decreased enzyme activity, for zooplankton, gastropods, amphibians and earthworms (Zubrod et al., 2011; McMahon et al., 2012; Rico et al., 2016). Furthermore, fungicidal compounds have been implicated in reduced bee health and abnormal behaviours, such as reduced nest recognition, decreased colony initiation and uncoupled mitochondrial respiration, all of which may contribute to the decline of bee populations (Elston et al., 2013; Simon-Delso et al., 2014; Syromyatnikov et al., 2017). The development of any new species-specific fungicides should provide both environmentally safer control strategies, but also provide effective control of the fungus to ensure improved crop yields.

Recently, a new generation of species-specific control methods, taking advantage of a cellular defence mechanism called RNA interference (RNAi), demonstrated successful control of insects, nematodes, viruses and parasitic plants (Whyard et al., 2009; Alakonya et al., 2012; Papolu et al., 2013; Schmitt-Engel et al., 2015). Pioneering studies utilizing RNA interference to control plant pathogens are summarized in Table 1. The first commercially approved, transgenic plants carrying RNAi constructs against corn rootworm (Diabrotica virgifera virgifera) and Bean golden mosaic virus were approved for cultivation in the USA and Brazil, respectively (Tollefson, 2011; United States Environmental Protection Agency, 2017). Despite the successes of RNAi approaches, few studies have applied this revolutionary technique for the management of fungal phytopathogens, likely due to the lack of reasonable target identification tools and poor fungal genomic annotation. Here, we describe how RNAi technology exploits intrinsic cellular pathways in eukaryotes for the development of novel fungal control strategies. A brief summary of aspects relating to safety of this new technology will be explored, as well as how integrating modern genomics techniques could help guide the development of next-generation RNAi-based control of fungal pathogens.

The mechanism of RNA interference

RNAi is a conserved pathway in eukaryotes that protects cells from viruses and controls transposon activity. The mechanism utilizes short interfering RNAs (siRNAs) to guide the targeted degradation of transcripts using sequence homology (Torres-Martínez & Ruiz-Vázquez, 2017). Prior to the discovery of RNAi, Rothstein et al. (1987) originally described an ‘antisense effect’ in tobacco plants through the silencing of a nopaline synthase transgene using the expression of antisense nopaline synthase. Subsequently, the description of RNAi in Caenorhabditis elegans by Andrew Fire and Craig Mello earned them the Nobel Prize for Medicine in 2006 (Fire et al., 1998; Nobel Media AB, 2017). However, similar phenomena had been noted in other organisms. Romano & Macino (1992) earlier had described a post-transcriptional silencing phenomenon, which they termed quelling, in the ascomycete Neurospora crassa. The core proteins in the quelling pathway are the same proteins implicated in the RNAi pathway: ARGONAUTE (AGO), QUELLING DEPENDENT-2 (QDE-2), DICER-like (DCL) and RNA-dependent RNA Polymerase (RdRP) (Torres-Martínez & Ruiz-Vázquez, 2017).

Since RNAi is an intrinsic cellular defence process against invading double stranded RNA (dsRNA) viruses, introducing in vitro synthesized dsRNAs or producing the molecules in planta exploits this cellular reaction as a crop management strategy naturally (Wang et al., 2016) (Fig. 1a). Transport of long dsRNA and shorter, small interfering RNA (siRNA) into B. cinerea spores was observed using fluorescein-labelled nucleotides (nt) (Wang et al., 2016); however, the mechanism of transport remains undefined in fungi. In both humans and C. elegans, dsRNA diffuses passively through a dsRNA-specific channel, SID1 (Duxbury et al., 2005; Whangbo et al., 2017). Homologues of the SID1 channel do not exist in fungi, and thus, dsRNA transport must occur via an alternative mechanism (Wang et al., 2016). In various insects, exogenous dsRNA is transported using receptor-mediated clathrin endocytosis (reviewed in Huvenne & Smagghe, 2010). An ‘RNAi of RNAi’ approach provided evidence for endocytic involvement in dsRNA transport. Initially, cells were treated with dsRNA targeting a specific component implicated in clathrin-mediated endocytosis, before a following treatment of GFP-dsRNA was applied. Using a GFP-reporter system, components involved in uptake of the molecules could be elucidated by observing a diminished fluorescent signal. Thus, the components of receptor mediated endocytosis were inferred to be clathrin heavy chain, clathrin adaptor protein 50, vacuolar H⁺ ATPase and ADP ribosylation factor ARF72A (Saleh et al., 2006). Similarly, application of chemical inhibitors confirmed endocytosis as a secondary pathway of dsRNA uptake in C. elegans (Xiao et al., 2015). Without SID1, clathrin-mediated endocytic pathways may play an integral function in the transport of dsRNA in fungi.

Once in the cytoplasm, the presence of a dsRNA molecule is recognized by the dsRNA binding domain of DCL1 or DCL2 (Lee et al., 2010; Li et al., 2010) (Fig. 1b). Upon recognition of dsRNA, DCL recruits the SAGA complex with histone acetyltransferase activity to increase transcription from the DCL and AGO promoters, and mobilize RNAi machinery within the fungus (Andika et al., 2017). To cleave dsRNA molecules, the 5ʹ end of the dsRNA anchors in the PAZ (Piwi-Argonaute-Zwille/Pinhead) domain within DCL, allowing two consecutive RNase III domains to cleave the ribose-phosphate backbone, resulting in siRNAs of 21–25 nt in length with a 5ʹ monophosphate and a 3ʹ 2 nt overhang (Kandasamy & Fukunaga, 2016) (Fig. 1c). Once the double stranded siRNA is generated, AGO complexes with the siRNA to recruit QIP (QDE-2-INTERACTING PROTEIN) and form the RNA-induced silencing complex (RISC) (Dang et al., 2011) (Fig. 1d). Once bound, AGO nicks the siRNA duplex; QIP recognizes and degrades the nicked passenger strand with exonuclease activity (Maiti & Lee H-C, 2007; Cheng et al., 2016) (Fig. 1e). The RISC complex then becomes activated to seek transcripts with complementary sequences to the remaining siRNA strand, termed the guide strand (Fig. 1f). When a messenger RNA (mRNA) base pairs to the guide strand in RISC, exonuclease activity is activated to degrade complementary RNA, resulting in a reduction of mRNA accumulation within the fungal hyphae (Dang et al., 2011) (Fig. 1g).

Fig. 1 (Colour online) Overview of the mechanism of RNAi within fungal hyphae. (Colour) Upon encountering double stranded RNA (dsRNA), the molecules are transported into the cytoplasm through an undefined mechanism (a). Once in the cytoplasm, the molecules are recognized by DICER-LIKE (DCL) (b), which cleaves the molecules into small interfering RNA (siRNA) 21–25 nucleotides in length (c). The siRNA molecules then complex with ARGONAUTE (AGO) (d), which nicks the siRNA and recruits QUELLING DEFICIENT-2-INTERACTING PROTEIN (QIP) to degrade the passenger strand (e). With the removal of the passenger strand, RNA induced silencing complex (RISC) becomes activated to seek messenger RNA (mRNA) transcripts with complementary sequences (f) for degradation (g). The degraded mRNA and the siRNA can function as primers in secondary dsRNA synthesis using RNA-dependent RNA Polymerase (RdRP) (h) to further amplify gene silencing.

In some organisms, siRNAs produced during long dsRNA processing can act as primers to initiate RdRP activity and temporarily sustain silencing (Fig. 1h). siRNAs anneal to complementary mRNA transcripts to act as a primer for second strand synthesis through recruited RdRPs. The secondary dsRNAs produced by RdRPs invoke further DCL activity and RISC complex formation to amplify RNAi-mediated silencing (Ghildiyal & Zamore, 2009; Villalobos-Escobedo et al., 2016). The number of RdRPs vary amongst fungi, with Fusarium graminearum having five, while N. crassa has two (Zong et al., 2009; Chen et al., 2015). RdRP can be essential for potent silencing, as in Mucor circinelloides, but for many fungi, RdRPs may not be essential for RNAi signal amplification, but are involved in miRNA (microRNA) biogenesis and transposon silencing (Dang et al., 2011; Calo et al., 2012).

The safety of RNA interference technology

The sequence-specific mechanism of RNAi provides unparalleled opportunities for RNAi-based technologies to offer safe and environmentally friendly alternatives to more traditional agrochemicals. DsRNA can be designed to avoid sequences of other organisms within the environment. Recently, transgenic corn expressing dsRNA targeting corn rootworm was approved for commercial use (United States Environmental Protection Agency, 2017). The transgenic plant material and in vitro synthesized dsRNA molecules were used to assess cross-reactivity amongst a variety of invertebrates (e.g. Apis mellifera, Eisenia adrei, Coleomegilla maculata), vertebrates (e.g. Gallus domesticus, Ictalurus punctatus) and soil microorganisms. No observable changes in physiology, nutrient assimilation or reproduction in the tested organisms was observed, regardless of the dsRNA source (Bachman et al., 2016). The presence of the corn rootworm-specific dsRNA in the pollen did not adversely affect honeybees (A. mellifera) consuming and distributing the pollen within the hives (Bachman et al., 2016). Larval or adult honeybees fed the dsRNA at doses exceeding 10 times the environmentally relevant exposure levels showed no adverse effects on their growth, development or longevity (Tan et al., 2016). The findings from these studies suggest that exogenously delivered dsRNAs, whether derived from either transgenic plants or foliar applications, may not pose serious threats to these important pollinators. Further studies will likely be required to satisfy government regulators that any dsRNA applied to a crop does not affect honeybees or any other species within the surrounding environment.

One somewhat surprising advantage of RNAi-based control is the relative stability of the dsRNA molecule within the phyllosphere (Miguel & Scott, 2015). Synthetic dsRNA molecules can be readily made in the laboratory and are more thermodynamically stable than single stranded RNA (Nicholson, 2014; Wang & Jin, 2017). Due to the double stranded structure of the molecules, dsRNAs are also more resistant to nuclease degradation than mRNA (Hoerter et al., 2011; Aryani & Denecke, 2015). As a topical application, dsRNA molecules were bioactive against Colorado potato beetle (Leptinotarsa decemlineata) on the surface of potato (Solanum tuberosum L.) leaves for over 28 days under greenhouse conditions (Miguel & Scott, 2015). Using natural chemistries, such as clay nanosheets (BioClay; Mitter et al., 2017), dsRNA efficacy was improved against Cucumber mosaic virus and Pepper mild mottle virus under adverse environmental conditions. The clay nanosheets shielded dsRNAs from environmental RNase III degradation and improved adhesion to the leaf surface (Ladewig et al., 2009; Mitter et al., 2017). With suitable formulations, it is entirely possible that dsRNAs could be used in topical applications against a variety of foliar pathogens.

Another advantage of RNAi-based pathogen management is the lack of persistence of dsRNAs within the pedosphere (Dubelman et al., 2014). Regardless of dsRNA length, these molecules rapidly degrade within 24 hours in all soil types examined thus far (Dubelman et al., 2014; Fischer et al., 2016). Similarly, dsRNAs are almost fully degraded within 96 hours upon entering natural water systems (Albright et al., 2017). Currently, there is a lack of evidence describing the timing of dsRNA degradation in the phyllosphere versus the pedosphere, although differences may exist due to distinct microbial communities (Bodenhausen et al., 2013). Bacterial nucleases, and in particular, RNase III enzymes are most likely responsible for much of the dsRNA degradation in the soil and aquatic environments (Urich et al., 2008; Cho, 2017). Therefore, dsRNA derived from plant material or from foliar sprays are unlikely to spread far from the point of application through the soil. Coupled with its sequence specificity, RNAi-based approaches are unlikely to have environmentally adverse effects on non-target species and therefore represent an attractive alternative to chemical fungicides.

With respect to food production, RNAi technology should not pose any additional risks to our food supply. All of our food already contains a diversity of small RNAs (sRNAs) including siRNAs, miRNAs, piRNAs and endogenous dsRNAs, produced from a variety of sources, including naturally occurring viruses, transposons or the host genome itself (Ivashuta et al., 2009; Frizzi et al., 2014). In silico analyses predict endogenous siRNAs with 100% complementarity to human mRNA transcripts are present in many non-GM crops, yet no pathological effects have been associated with daily consumption of the siRNAs derived from these dsRNAs (Jensen et al., 2013). Higher animals have evolved many barriers that would prevent or limit the transport of siRNAs, such as nucleases in the saliva and gastrointestinal tract, acid in the stomach, and the unfavourable transport of large, polar molecules across a hydrophobic membrane (Juliano et al., 2009; O’Neill et al., 2011). The lack of dietary siRNA efficacy has been noted in mouse studies; for example, mice dosed daily with either siRNA or long dsRNA targeting an essential vacuolar ATPase over a 28-day period showed no evidence of RNAi-mediated knockdown of the target transcripts and no adverse cellular pathologies (Petrick et al., 2015). The lack of any observable RNAi following ingestion of dsRNA in mammals could be attributed to instability of the molecules passing through the gastrointestinal tract or rapid metabolism in the bloodstream (Christensen et al., 2013; Dickinson et al., 2013). It therefore seems unlikely that any consumed dsRNA or siRNA molecules will elicit adverse effects on higher organisms.

Furthermore, many food products undergo many processing techniques before consumption (Chemat et al., 2017; Misra et al., 2017). The majority of processing techniques (baking, microwaving, solvent extraction, thermal treatment, fermentation, acidification, alkalization and bleaching) result in effective nucleic acid destabilization and degradation prior to the final food product (Vijayakumar et al., 2009; Gryson, 2010). For example, edible oils undergo multiple steps involving heat, pressure and solvent treatments, which exclude polar molecules, such as nucleic acid, and/or result in molecule fragmentation (Mba & Dumont M-J, 2015; Belur et al., 2017). Similarly, heating and purification in sugar production eliminates DNA by a factor of 10¹⁴ (Klein et al., 1998). Based on the chemical and physical similarities, dsRNA would likely have a similar fate to DNA during food processing (Forbes & Peppas, 2012; Lipfert et al., 2014). Thus, any nucleic acid introduced from RNAi technology would not resist food processing. Taken together, RNAi technology, due to both the chemistry of the RNA molecules and the sequence-specificity of the molecule, could be considered a safe, green technology, which can be expressed as a novel trait (transgene) or through topical formulations.

Development of novel traits through host-induced gene silencing

Host-induced gene silencing (HIGS) is an emerging biotechnology in which plants are engineered to produce siRNAs capable of silencing target genes of a target organism. HIGS utilizes the RNAi pathway by equipping the host plant with hairpin RNAs (hpRNAs) containing sequence homology to target genes. Upon transcription, these hpRNA molecules mimic dsRNA and initiate the inherent cellular RNAi pathway. Export of hpRNA from the plant nucleus to the cytoplasm is hypothesized to be facilitated by the binding of an exportin protein HASTY to guide successful nucleocytoplasmic transport (Bollman, 2003). Once in the cytoplasm, hpRNA initiates the host RNAi pathway leading to the generation of approximately 21 nt siRNA molecules. For siRNAs to function in targeted gene silencing for plant protection, they must undergo successful transfer from host plant cell to the pest or pathogen. While evidence suggests target gene silencing in pathogenic fungal species may operate through host derived siRNAs (Panwar et al., 2013), the mechanism of siRNA transfer from host to pathogen remains unclear. Studies in animal systems show that secreted miRNAs may be associated with host-derived exosomes and lead to successful transfer between organisms (Valadi et al., 2007). Exosomal uptake by the receiving cell is hypothesized to utilize exosome-mediated endocytosis, where the vesicular membrane of the sRNA-containing exosome fuses with the receiving plasma membrane, leading to the release of sRNA into the pathogen’s cytoplasm (Valadi et al., 2007). Alternatively, sRNA transfer may involve transmembrane transporter-mediated uptake without utilization of host-derived vesicles. Membrane-free sRNAs have been found within the extracellular space and associated with high density lipoproteins. These lipoproteins may facilitate successful transfer of extracellular sRNA to recipient cells (Vickers et al. 2011). With a more complete understanding of dsRNA transport in and between cells, it may be possible to enhance both HIGS technology and topical dsRNA delivery formulations to maximize the degree of RNAi-mediated protection of crop plants to a broad range of pathogens and pests.

HIGS technology for crop protection has been approved in two RNAi crops for commercial production in the USA and Brazil. The Brazilian National Technical Commission approved RNAi pinto beans (Phaseolus vulgaris) for commercial production in 2011 (Tollefson, 2011). The plants were engineered to disrupt early viral replication of the Bean golden mosaic virus by targeting the viral gene AC1 (Bonfim et al., 2007). Similarly, the Environmental Protection Agency approved maize plants expressing dsRNA targeting DvSnf4, a component of the ESCRT-III complex involved in endosomal sorting and lysosomal degradation in corn rootworms, Diabrotica virgifera virgifera, for production in the USA (Bolognesi et al., 2012; United States Environmental Protection Agency, 2017). DvSnf4 dsRNA was potent at low doses, leading to accumulation of ubiquinated proteins in the midgut cells of larvae. Since autophagy was impaired, the cells malfunctioned, the gut’s digestive processes ceased, and the insects failed to grow and eventually died (Baum et al., 2007; Ramaseshadri et al., 2013). While targeted genes are commonly selected due to their essential role in pathogenicity, an alternative strategy that has demonstrated success is the targeting of host susceptibility genes. In a study by Sun et al. (2016), the silencing of six previously identified susceptibility genes in potato ‘Desiree’ led to significant reductions in susceptibility against potato late blight (Phytophthora infestans). Given the flexibility and utility of the technology, more RNAi-HIGS crops will undoubtedly be developed and commercialized in the near future, including plants with RNAi-mediated protection from fungal pathogens.

In planta expression of hpRNA overcomes fungal pressure

The development of stably transformed plants expressing dsRNA molecules could impart full plant protection from fungal infections. HIGS has been implemented to reduce fungal pressure and toxin biosynthesis for a variety of phytopathogenic fungi (Koch et al., 2013; Zhang et al., 2016; Thakare et al., 2017). Engineering novel traits like fungal resistance using genetic techniques is an efficient strategy to build resistance in crop plants. While traditional germplasm screening and breeding strategies have achieved limited success in providing resistance against S. sclerotiorum (Disi et al., 2014), the use of HIGS in Nicotiana tabacum expressing hpRNA targeting fungal chitin synthase (SSCHS) proved effective in controlling S. sclerotiorum. Expression of hpRNA reduced the level of SSCHS mRNA within the fungus, indicating S. sclerotiorum could readily take up dsRNA from the host (Andrade et al., 2016). Since other phytopathogens are likely capable of taking up environmental dsRNA from host cells, HIGS could provide a functional strategy to reduce fungal pressure on the plant.

The root–pathogen interface could also be protected from many fungal pathogens that initiate root infections from the soil, such as Verticillium sp., Fusarium sp. and Rhizoctonia solani (Tedersoo et al., 2014; De Coninck et al., 2015). While topical formulations of dsRNA would likely degrade rapidly within the soil (Dubelman et al., 2014), GM plants could deliver dsRNAs to the pathogen at the root–pathogen interface. Roots of cotton (Gossypium sp.) expressing hpRNA targeting V. dahliae hygrophobins1 gene were able to resist severe root infection (Zhang et al., 2016). Additionally, plants engineered with RNAi constructs expressing hpRNA could convey resistance throughout the plant life cycle. Transgenic banana plants (Musa spp.) expressing hpRNA directed at either F. oxysporum genes VELVET or FUSARIUM TRANSCRIPTION FACTOR 1 resisted infection for 8 months post-inoculation (Ghag et al., 2014). Thus, engineered plants expressing novel RNAi traits conferring resistance against economically important plant pathogens represents an additional level of durability that would benefit growers interested in sustainable crop protection technologies.

Foliar applications of long dsRNA reduce fungal disease

Foliar dsRNA applications offer shorter-term protection from fungal infections, relative to transgene mediated resistance, but nevertheless, they could be particularly useful to protect agri-food products during post-harvest storage and protecting plants without defined or efficient transformation protocols for HIGS (Wang & Jin, 2017). Despite recent advances to control insect pests and viral pathogens, few studies have implemented in vitro synthesized molecules for fungal control. In a pioneering study by Mumbanza et al. (2013), 14 different genes involved in processes such as transcription, RNA modification, DNA replication and intracellular transport were tested in vitro, using nutrient plates rather than plants, for spore germination inhibition in F. oxysporum f. sp. cubense and Mycosphaerella fijiensis. Treatment of fungal spores with dsRNA molecules inhibited germination of the two banana pathogens by up to 95.9% and 65.8%, respectively. In vitro testing provided compelling evidence for the potential for fungal suppression, despite not having tested these molecules on their plant hosts.

Recently, topical application of dsRNA to the surface of treated barley (Hordeum vulgare L.) leaves reduced F. graminearum growth. By targeting three fungal cytochrome P450 lanosterol C-14α demethylases, required for fungal ergosterol synthesis and a common fungicide target, reductions over 50% of both target transcript and fungal DNA accumulation were achieved (Koch et al., 2016). Similarly, using long dsRNA targeting DCL1 and DCL2 simultaneously in B. cinerea, Wang et al. (2016) demonstrated remarkable levels of protection in Arabidopsis thaliana, tomato, grape, strawberry, onion, and rose petals from infection, with average lesion sizes, transcript levels and fungal DNA all being reduced by more than 80%. The observed protection in a wide-range of host tissues suggests broad utility of the technology in the protection of food and ornamental species. The protection of barley leaves and tomato leaves demonstrates the potential utility for field management of diseases, while reduced fungal presence on various fruits and vegetables could be useful after harvest, protecting food in transport, storage, or on store shelves (Koch et al., 2016; Wang et al., 2016). Taken together, these studies provide evidence of the flexibility of RNAi-based molecular sprays that could confer protection against diseases in food production.

RNA sequencing as an informative guide for RNAi

Despite the advances in the development of RNAi-based phytopathogen control, the selection of target genes still represents a significant challenge in the design and effectiveness of fungal suppression. Previously, RNAi targets were selected using gene deletions or chemical inhibition. For example, Koch et al. (2016) chose a common fungicidal target, while Wang et al. (2016) developed foliar applications based on previous genetic deletions of B. cinerea strains (Weiberg et al., 2013). Fungal transformation protocols are lengthy and at times inefficient, thereby limiting the effective size of target gene identification screens (Meyer, 2008). However, dual RNA-sequencing of both fungal pathogen and host plant now provides unprecedented opportunities to identify novel RNAi targets based on the transcriptome atlas of the pathogen life cycle, infection state, tissue type, host defence response or treatment conditions (Westermann & Barquist, 2017). Genes implicated in plant resistance response provide useful information for plant breeding and manipulation, while specific fungal responses essential for pathogenesis could eventually be used as targets for RNAi-based fungal control (Girard et al., 2016). For example, McLoughlin et al. (2018) utilized RNA-seq to develop foliar antifungal dsRNAs that inhibit disease pressure on two plant species. By comparing global transcriptomic changes of S. sclerotiorum infection on susceptible and tolerant B. napus cultivars, as well as growth in vitro, hundreds of genes involved in fungal pathogenesis were uncovered. Subsequently, topical dsRNAs were developed to target genes implicated in various processes such as respiration, toxin biosynthesis, protein modification, translation and transcription, which reduced fungal infection on both B. napus and A. thaliana significantly. This suggests that dsRNAs could impart protection for many host species, and shows potential in controlling devastating crop pathogens. Furthermore, homologous dsRNAs were also successful in supressing the related pathogen, B. cinerea (McLoughlin et al., 2018). Consequently, homologous targets could be exploited to provide broad levels of control. Taken together, RNA-seq identification of host–pathogen interfaces could provide useful insight for the development of future RNAi target genes for fungal control.

The lack of development of RNAi-based fungal strategies is likely due to poor genomic annotation of many fungi and a lack of accessible bio-computational platforms. User-friendly programs, such as FungiFun2 (Priebe et al., 2015; elbe.hki-jena.de/fungifun/fungifun.php) and SeqEnrich (Becker et al., 2017; belmontelab.com), have overcome many of the problems associated with both a lack of functional annotation and complicated bioinformatics analyses. Once gene ontology (GO) is resolved using FungiFun2, researchers could use SeqEnrich, to identify significantly upregulated GO terms and genes. Melding both programs together would provide a solid basis for the execution of RNAi-based technologies. Upregulated transcripts and processes could then be selected for the development of either foliar implementations or HIGS technology. RNA-seq provides greater depth and efficiency for the rapid identification of hundreds of critical, upregulated transcripts and processes within infecting pathogens. Moreover, RNA-seq offers a marked improvement over previous searches reliant on labour-intensive fungal transformation studies. The use of RNA sequencing technology has the potential to revolutionize and expedite the next generation of species-specific fungal management strategies.

Outlook

RNAi technology provides a flexible and environmentally friendly solution to combat an array of devastating pathogens. Both HIGS and foliar dsRNAs represent potent and sustainable next-generation strategies capable of controlling pathogens that affect global food production. While the advent of RNAi technologies holds significant promise, concerns over pathogen resistance will continue to exist. Furthermore, while RNAi machinery is known to be conserved in eukaryotic organisms, those that lack the RNAi machinery like Ustilago maydis, the causal agent of corn smut (Billmyre et al., 2013) would not be a candidate for host resistance through RNAi. Although further research exploring the interaction of dsRNA within food should be explored, RNAi strategies have proven to be safe through its species-specificity and molecule degradation during food processing. Fundamental aspects of small molecule transport have yet to uncover the underlying molecular mechanism of RNAi-based control of fungal species. The identification of new targets for pathogen control may provide additional clues into conserved pathogenesis genes or common targets for broad levels of efficacious fungal control. Furthermore, RNA sequencing could provide an effective and efficient method for recognizing targets for RNAi, accelerating the establishment of additional novel, species-specific fungicides. The field of RNAi technology is developing for fungi and RNAi is a sustainable solution to deal with fungal diseases and food security challenges.

Acknowledgements

We would like to thank the Canadian Phytopathological Society for the invitation to present this review.

Additional information

Funding

This work was generously supported by grants from the province of Manitoba Agricultural Rural Development Initiative, the Canola Council of Canada, and the Western Grains Research Foundation to SW and MFB. AM and NW were supported by a Natural Sciences and Engineering Research Council of Canada (NSERC) Graduate Masters Scholarship and the Manitoba government Tri Council Top-Up award.