Molecular Targets of Herbicides and Fungicides—Are There Useful Overlaps for Fungicide Discovery?

Abstract

New fungicide modes of action are needed for fungicide resistance management strategies. Several commercial herbicide targets found in fungi that are not utilized by commercial fungicides are discussed as possible fungicide molecular targets. These are acetyl CoA carboxylase, acetolactate synthase, 5-enolpyruvylshikimate-3-phosphate synthase, glutamine synthase, phytoene desaturase, protoporphyrinogen oxidase, long-chain fatty acid synthase, dihydropteroate synthase, hydroxyphenyl pyruvate dioxygenase, and Ser/Thr protein phosphatase. Some of the inhibitors of these herbicide targets appear to be either good fungicides or good leads for new fungicides. For example, some acetolactate synthase and dihydropteroate inhibitors are excellent fungicides. There is evidence that some herbicides have indirect benefits to certain crops due to their effects on fungal crop pathogens. Using a pesticide with both herbicide and fungicide activities based on the same molecular target could reduce the total amount of pesticide used. The limitations of such a product are discussed.

Introduction

A recent paper discussed the potentially useful overlaps between molecular targets of herbicides and insecticides that could be used for the discovery and development of insecticides with new molecular targets.¹ Likewise, such overlaps between herbicide and fungicide molecular targets might provide potential new leads for new molecular targets for fungicides that could be useful in combating the growing problem of the evolution and spread of fungicide-resistant plant pathogens (Figure 1).^2,3 This paper is a sequel to that on herbicide and insecticide target site overlaps.¹

Figure 1.

Rise of fungicide-resistant plant pathogens worldwide. Cases are the first case for a given mode of action for a particular pathogen species. The total minus field cases are those selected in a laboratory. Data are from FRAC.³

Companies involved in the discovery of crop protection compounds routinely evaluate all compounds active against one pest type versus other important types of pests.⁴⁻⁷ This approach has resulted in the discovery and development of pesticides for several pest types (e.g., arthropods, weeds, and fungal pathogens) that were inspired by compounds first noticed because of activity against another major pest type.⁵ However, this approach does not consider the molecular target of the starting compound, and the result of this strategy is often a compound with a molecular target on the second pest type different from that on the starting pest type. For example, the methionine synthesis-inhibiting fungicide cyprodinil (Figure 2) was discovered by derivatization of an acetolactate synthase (ALS) inhibitor herbicide.^8,9 Lamberth discusses other examples of fungicides derived from herbicides, but without the same molecular targets.⁵ However, there is one case of molecular site overlap that led to the development of a herbicide from a fungicide. Structural alteration of a cellulose synthase-inhibiting carboxylic acid amide fungicide produced a structurally similar herbicide that inhibits plant cellulose synthase (Figure 2).^10,11

Figure 2.

Compounds mentioned in the text.

Certain fungi and oomycetes are among the most important plant pathogens damaging crops. While they share similarities with regard to morphology and infection processes, these microbes are not closely related. Even though the fungi and oomycetes are genetically distant and some chemical management products are specific to one group (e.g., metalaxyl and oomycetes), the term fungicide is used in the management of both groups. The true fungi are more closely related to animals than plants, while the oomycetes are only distantly related to plants, animals, and fungi. Although plants, fungi, and oomycetes are not closely related, there are many reports of fungicides being phytotoxic and herbicides being fungicidal. For example, Baibakova et al.¹² reviewed the phytotoxicity of several agricultural fungicides. However, in most cases, the molecular target of the fungicide as a phytotoxin is unknown. Likewise, there is considerable literature on the effects of herbicides on plant pathogens, e.g., (13,14). Again, in many cases, the molecular target of the herbicide as a fungicide is unknown. This problem is complicated by the fact that in some cases, the effect of the herbicide is indirect through the herbicide inducing pathogen resistance mechanisms, such as increasing phytoalexin levels, in the herbicide-treated plant. We briefly discuss this situation for some herbicides, but this paper focuses on common molecular targets in plants and fungi.

A few previous reviews discuss the effects of herbicides on plant pathogens, including effects of herbicides for which the herbicide molecular target does not exist in fungi (e.g., PSII inhibitors).¹⁵ However, there is no comprehensive review on the overlapping plant herbicide targets with targets that could be useful in the discovery of new fungicides which could add to the modes of action (MoA) available for fungicide resistance management. There are several target sites common to both commercial fungicides and herbicides (Tables 1 and 2),^3,16 and there are other common molecular targets for which there are no commercial herbicides or fungicides yet. Moreover, there are strong inhibitors that could be the basis for either fungicide or herbicide development of these common molecular targets. For example, inhibitors of hydroxy-3-methylglutaryl-CoA reductase (HMG-CoA reductase) (often termed statins) are both highly phytotoxic and antifungal.¹⁷⁻²⁰ HMG-CoA reductase as a herbicide and insecticide target is discussed in detail are in a recent paper.¹ However, this review will deal with only commercial herbicide molecular targets that might also become targets for fungicides.

Table 1. FRAC (2023) Classification of Commercial Fungicide Molecular Targets (Those in Bold Are Also Found in Plants, and Those in Both Bold and Italics are Targets of Commercial Herbicides)^a.

classification group	target
1, 10, 22	tubulin polymerization
2	osmotic signal transduction
3	sterol synthesis inhibition (demethylase inhibitor: erg 11 enzyme)
5	sterol synthesis inhibition (erg 2 and erg 24 enzyme)
17	sterol synthesis inhibition (ketoreductase: erg 27 enzyme)
18	sterol synthesis inhibition (squalene-epoxidase)
4	RNA polymerase I
6	phospholipid biosynthesis, methyl transferase
7	respiration complex II at succinate dehydrogenase
8	adensinedeaminase
9	methionine biosynthesis
11	respiration complex II at Qo site
12	MAP/histidine kinase in osmotic sign
16.1	reductase in melanin synthesis
16.2	dehydratase in melanin synthesis
16.3	polyketide in melanin synthesis
21, 45	respiration complex III: cytochrome bc1 (ubiquinone reductase)
23	protein sysnthesis (ribosome termination step)
24, 25	protein synthesis (ribosome initiation step)
26	chitin synthase
29	respiratory uncoupler
30	ATP synthase
31	DNA topoisomerase
38	ATP transporter
39	complex I NADH oxido-reductase
40	cellulose synthase
41	protein synthesis (ribosome elongation step)
47, 50	actin
48	ergosterol binding
51	direct disruption of cell membrane
52	dihydroorotate dehydrogenase

^aSeveral fungicides act by inducing host plant defenses, and the molecular target of several fungicides are unknown.

Table 2. HRAC (2022) Classification of Commercial Herbicide Molecular Targets (Those in Bold Are Also Found in Fungi, and Those in Both Bold and Italics are Targets of Commercial Fungicides)^a.

classification group	target
1	acetyl CoA carboxylase
2	acetolactate synthase
3	microtubule subunits
4	auxin mimics (F-box proteins)
5 and 6	D-1 protein of photosystem II
9	5-enolypyruvylshikimate-3-phosphate synthase
10	glutamine synthetase
12	phytoene desaturase
13	deoxy-D-xyulose-5-phosphate synthase
14	protoporphyrinogen oxidase
15	very long-chain fatty acid synthase
18	dihydropteroate synthase
19	auxin transport inhibitor
22	electron diversion from PSI
23	microtubule organization
24	mitochondrial uncoupler
27	hydroxyphenyl pyruvate dioxygenase
28	dihydroorotate dehydrogenase
29	cellulose synthesis
30	fatty acid thioesterase
31	Ser/Thr protein phosphatase
32	solanyl diphosphate synthase
33	homogentisate solanyltransferase
34	lycopene cyclase

^aThe molecular targets of several commercial chemical herbicides are listed as unknown by HRAC.

Because of crop injury issues, the market for compounds that kill both weeds and fungal plant pathogens is low; however, as discussed below, crops can be made resistant to such compounds, so that they can be used for both weed and plant pathogen management. Use for preplant weed management, where the dead or dying plants can act as a bridge for crop pathogens, could be especially effective. Such a product could reduce the total pesticide load in the environment, provided the timing of herbicide and fungicide use coincide. Alternatively, if there is a sufficient difference in the target molecule between fungi and plants, the inhibitor’s chemical structure might be modified to be specific for plant pathogenic fungi.

Target Sites of Commercial Herbicides Found in Fungi That Are Not Commercial Fungicide Targets

The number of commercial herbicide targets that are shared with fungi is smaller than that of commercial fungicide targets shared with plants (Tables 1 and 2), so fungicide targets may provide more leads for new herbicide MoAs than vice versa. However, this is a topic for a later review. We discuss commercial herbicide targets not currently used by commercial fungicides that might be useful for new fungicides, beginning in the order of Herbicide Resistance Action Committee (HRAC) group number.¹⁶ The molecular targets already shared by commercial herbicides and fungicides of commercial interest (tubulin and tubulin function, cellulose synthase, mitochondrial uncouplers, and dihydrooratate dehydrogenase) will not be discussed.

The National Center for Biotechnology Information accession numbers of the molecular targets discussed in this review that are shared by fungi and plants are provided in Table 3, using Botrytis cinerea and Arabidopsis thaliana as a representative fungus and plant, respectively. Presumed orthologous human proteins are also provided in this table in cases where they exist. The predicted protein size and amino acid sequence similarities of the enzymes are provided. Levels of similarity are not decisive for selectivity, but they provide an idea of how conserved the targets are between taxa. A detailed analysis of similarities at the catalytic and inhibitor binding sites is beyond the scope of this paper. Clearly though, the amino acid sequence of the binding domain of the substrate and/or a molecular target shared by plants and fungi may differ, considering the evolutionary separation of plants and fungi. For example, differences in plant and fungal binding sites of acetyl CoA carboxylase are discussed briefly below.

Table 3. Accession Numbers of Most of the Molecular Targets Mentioned in the Text for Protein Sequences of Arabidopsis thaliana (At), Botrytis cinerea (Bc), and Homo sapiens (Hs), Providing the Degrees of Protein Sequence Similarity to That of Plant Proteins.

target	NCBI accession no.	predicted size (amino acid)	similaritya (%)
acetyl CoA carboxylase	AAC41645 (At)	2254	100
	ATZ52203 (Bc)	2278	57.23
	AAC50139 (Hs)	2346	57.49
acetolactate synthase	AAK68759 (At)	670	100
	ATZ54994 (Bc)	699	58.73
	AAC50934 (Hs)c	632	44.26
5-enolpyruvylshikimate-3-phosphate synthaseb	AEE32360 (At)	521	100
	ATZ56606 (Bc)	380	48.79
glutamine synthetase	AEE76011 (At)	354	100
	ATZ57173 (Bc)	331	88.35
	AAB30693 (Hs)	373	71.18
phytoene desaturaseb	AEE83394 (At)	566	100
	ATZ45727 (Bc)	602	33.94
protoporphyrinogen oxidase	BAA11820 (At)	537	100
	ATZ51938 (Bc)	423d	57.97
	AAH02357 (Hs)	477	45.8
hydroxyphenyl pyruvate dioxygenase	AEE28007 (At)	418	100
	ATZ57804 (Bc)	422	49.17
	CAA51082 (Hs)	393	49.32
dihydropteroate synthaseb	AEE34892 (At)	484	100
	ATZ58032 (Bc)	522	55.86
Ser/Thr protein phosphatase	AAA64742 (At)	313	100
	ATZ55832 (Bc)	307	81.52
	CAA52169 (Hs)	323	68.75
lycopene cyclaseb	AEE74874 (At)	501	100
	ATZ45729 (Bc)	610	29.31

^aThe similarity between sequences was obtained by pairwise alignment against the sequence assigned to 100%.

^bSequence is not available in the Homo sapiens genome database.

^cA human protein with unknown function, sharing some sequence similarity with other ALS, is encoded by the ILVBL (ilvB acetolactate synthase like) gene.

^dThe number represents a domain for putative protoporphyrinogen oxidase activity in a predicted protein that consists of 1576 amino acids.

Much of the empirical data on the effect of herbicides on plant pathogens has come from work testing the compatibility of these herbicides with mycoherbicides (fungi used for biocontrol of weeds), e.g., ref (21). Some of the effects of commercial herbicides on the enzymes that we cover below are discussed by Thiour-Mauprivez et al.,²² although their focus is on bacterial rather than fungal enzymes and on off-target effects of herbicides, rather than fungicide discovery.

Acetyl CoA Carboxylase

Acetyl CoA carboxylase (ACCase, HRAC group 1) is a major commercial herbicide target and has been validated as a potential fungicide target. The amino acid sequence similarity of ACCase in Arabidopsis and Botrytis cinerea is moderate (57%, Table 3), the same as the similarity to human ACCase. The myxobacterial polyketide soraphen A (Figure 3) is a potent fungal (Ustilago maydis, corn smut) ACCase inhibitor, but it binds a different enzyme domain than commercial herbicides that bind only the carboxyltransferase of ACCase of monocot plastids and bacteria.²³ Because the ACCase of fungi lacks the herbicidal ACCase binding site, ACCase inhibitor herbicides may not be fungicidal. Nevertheless, a commercial formulation of sethoxydim strongly inhibits mycelial growth and sporulation of the experimental mycoherbicide Dactylaria higginsii.²⁴ Similarly, Wyss et al. found formulated sethoxydim and clethodim to be toxic to Phomopsis amaranthicola, another mycoherbicide,²¹ and Madhavi et al. found formulated quizalofop ethyl to inhibit growth of Trichoderma viride.²⁵ Unfortunately, the use of a commercial formulation made it impossible to know whether these effects are due to the herbicide or a formulation ingredient, much less whether the target is ACCase. Some formulation ingredients are fungitoxic.²¹

Figure 3.

Compounds mentioned in the text.

Soraphen A is not phytotoxic and does not inhibit plant ACCase, but it does inhibit mammalian and yeast (Saccharomyces cerevisiae) ACCase.^26,27 The compound binds ACCase of Phytophthora infestens, Ustilago maydis, and Magnaporthe grisea.²³ It was suggested as a chemical for U. maydis control.²⁸ As reviewed by Naini et al., it was found to be effective at low rates in preventing some crop fungal diseases and was patented as an agricultural fungicide, but development was terminated because of potential mammalian toxicity issues²⁹ Fungicides and herbicides that target ACCase without overlapping the other pesticide-binding domain are known.

A drawback of ACCase inhibitor herbicides is that target site resistance evolves relatively easily and quickly.³⁰ Considering the binding site differences between ACCase inhibitor herbicides and soraphen A, facile evolution of target site resistance may not be the case for a fungicide with the same binding site as soraphen A. The efficacy of soraphen A indicates that ACCase is a good target site for agricultural fungicide discovery efforts. However, there is weak evidence that several chemical classes of herbicidal ACCase inhibitors could be the basis for fungicide discovery, although there is no evidence that they affect fungal ACCase. We note that Nimbus Discovery, LLC, and Monsanto Company entered a research collaboration agreement in 2013 on the development of new discoveries to help farmers control crop diseases, and ACCase was one of the fungicide target sites mentioned in this collaboration (https://www.nimbustx.com/2013/06/).

Acetolactate Synthase

Many molecular target sites are involved in amino acid biosynthesis shared by plants, fungi, and oomycetes.^31,32 Three of these are molecular targets of herbicides (Table 2, HRAC group 2), but only one, methionine biosynthesis (not a herbicide target), is a commercial fungicide target (Table 1). Acetolactate synthase (ALS; also called acetohydroxyacid synthase or AHAS) inhibitors are a major class of commercial herbicides (about 50 commercial active ingredients), but no commercial fungicides are ALS inhibitors. The amino acid sequence overlap for ALS in Arabidopsis and B. cinerea is moderate (59%, Table 3). Both plants and fungi require this first enzyme of the branched chain amino acid pathway to produce leucine, isoleucine, and valine. However, the enzyme structures of plant and fungal ALS differ, with the plant ALS having dodecameric complexes of catalytic and regulatory subunits, whereas the fungal enzyme is a hexadecameric subunit enzyme.³³ Wyss et al. reported that formulated imazapyr could inhibit germination and growth of the fungal plant pathogen P. amaranthicola (a mycoherbicide) at high concentrations.²¹ However, whether the effect was due to the formulation ingredients or the herbicide was not determined. Similarly, formulated imazethapyr inhibited spore germination of Colletotrichum gloeosporioides.³⁴

ALS has been identified as a potential target for medicinal fungicides.³⁵⁻³⁸ In one of these studies, Richie et al. reported two commercial herbicides, chlorimuron-ethyl and sulfometuron-methyl, to inhibit yeast (S. cerevisiae) growth and to be less effective in yeast strains with a mutated form of ALS.³⁷ As reviewed by Justrzębowska and Gabriel, all of the published attempts to discover fungicides acting on ALS have been directed toward medicinal fungicides.³¹ Recent work has reported that commercial ALS-inhibiting herbicides are promising leads for pharmaceutical fungicides.³⁸⁻⁴² The natural ALS inhibitor harzianic acid (Figure 3 and 4) is effective on the ALS of the fungal plant pathogens Fusarium oxysporum and Sclerotinia sclerotiorum, but not that of plant (Arabidopsis) ALS,⁴³ suggesting that a fungicidal ALS inhibitor could be designed for agricultural use with no crop injury problems. Alternatively, crops can easily be made resistant to ALS inhibitors by mutation breeding,⁴⁴ allowing for a product that could target ALS in both weeds and plant pathogens, thus lowering the overall pesticide load in the environment. The fact that many of the commercial ALS inhibitor herbicides have been shown to be fungicidal by inhibiting ALS suggests that these herbicides could provide additional benefits as fungicides in ALS herbicide-resistant crops. There have been some efforts to discover both sulfonylurea⁴⁵⁻⁴⁷ and imidazolinone^48,49 ALS inhibitor fungicides for use against crop pathogens, but no field studies have apparently been published, and none of these papers demonstrates that the molecular target is ALS. Treatment of imidazolinone-resistant sugar cane with imazapyr resulted in large changes in endophytic filamentous fungi present in the leaves,⁵⁰ but we have found no reports or the effect of ALS-inhibiting herbicides on fungal diseases in ALS herbicide-resistant crops. Venne et al. speculate that imidizolinone-resistant Fusarium oxysporum f. sp. strigae would have to be used as a mycoherbicide on imidazolinone-resistant maize,⁵¹ indicating that they know that this plant pathogen is susceptible to imidazolinone herbicide, although we have found no publications to support this view.

Figure 4.

Effects of harzionic acid (HA) on growth of Fusarium oxysporum with (red, circles) and without (blue, squares) supplementation with branched chain amino acids (isoleucine, leucine, and valine, ILV). Reproduced with permission from ref (43). Copyright 2021 American Chemical Society.

Like ACCase resistance, target site resistance to ALS inhibitor herbicides evolved quickly.³⁰ This would probably also be the case for fungicides that act by inhibition of ALS, thus requiring rigorous resistance management strategies for such products.

Enolpyruvyl-shikimate-3-phosphate Synthase (EPSPS)

EPSPS is the target of the world’s most successful herbicide, glyphosate (HRAC group 9).⁵² No other commercial herbicide targets this shikimic acid pathway enzyme required for the synthesis of aromatic amino acids and essential products of aromatic amino acids, such as lignin, plastoquinone, and indole acetic acid.⁵³ Fungi also possess this enzyme and are thereby sensitive to glyphosate with in vitro bioassays.^14,53,54 The amino acid sequence similarity of EPSPS of Arabidopsis and B. cinerea is moderate (49%) (Table 3).

Some of the literature on whether glyphosate can be used as a fungicide is confused by the fact that glyphosate inhibits production of plant defenses dependent on aromatic amino acid derivatives (e.g., some phytoalexins, lignin, etc.), making glyphosate-affected plants more susceptible to fungal pathogens, e,g,, refs (55,56). Hammerschmidt reviewed the effect of glyphosate on shikimic acid-derived plant defenses against plant pathogens.⁵⁷ Thus, glyphosate is more active on plants grown in soil with plant pathogens than in sterile soil, apparently due to glyphosate weakening plant defenses.^58,59

Some have claimed that glyphosate-treated, glyphosate-resistant (GR) crops are also more susceptible to plant pathogens.⁶⁰ The resistance factor for glyphosate in transgenic GR soybean and canola is about 50 and is more than 100 for GR maize.^61,62 At the recommended application rates of glyphosate for weed management, there is no phytotoxicity to GR crops. Thus, the shikimate-pathway-derived compounds needed for pathogen tolerance are not affected. The claims of glyphosate treatments increasing susceptibility of GR crops to pathogens have largely been debunked by several papers and reviews,^57,63 and most research has found that glyphosate has either no effect or reduces fungal plant diseases of GR crops. For example, a recommended glyphosate application rate one day before inoculation of GR wheat with wheat leaf rust (Puccinia triticina) prevented infection (Figure 5).⁶⁴ Similar results were found with glyphosate and wheat stem rust (P. graminis f.sp. tritici) on GR wheat.⁶⁵ Likewise, glyphosate reduced infection of GR soybean with Asian soybean rust (Phakopspora pachyrhizi),^64,66,67 GR alfalfa with alfalfa rust (Uromyces striatus),⁶⁸ and GR cotton with Rhizoctonia solani.⁶⁹ Results were more variable with GR rapeseed and three diseases, two fungal and one oomycete: white leaf spot (Neopseudocercosporella capsellae), Alternaria leaf spot (Alternaria brassicae and Alternaria japonica), and downy mildew (Hyaloperonospora brassicae).⁷⁰ Einhardt et al. insinuated that nickel ion and glyphosate act synergistically in reducing Asian soybean rust in GR soybean, but a more than additive effect was not proven.⁷¹ Glyphosate applied at the V3 or V6 growth stages of GR soybean suppressed charcoal rot (Macrophomina phaseolina) in a tilled site and had no effect in a no-till site.⁷² Glyphosate was fungicidal to this microbe in an in vitro assay. Commercially formulated glyphosate was recently shown to be as effective as a commercial fungicide in suppressing rice blast (Magnaporthe oryzae) in GR rice, and the authors indicated that the effect was solely through the fungicidal effects of glyphosate.⁷³ Low doses of glyphosate even reduced infection of glyphosate-sensitive eucalyptus by Austropuccinia psidii.⁷⁴ Several field studies have reported no effect of glyphosate used for weed management in GR crops on crop diseases.⁵³ For example, in a multisite study carried out in GR soybeans in five U.S. states and one Canadian province over several years, no effect of glyphosate was found on root rot and damping off caused by Rhizoctonia solani.⁷⁵ However, the timing of glyphosate use was focused on weed management, which would rarely have coincided with the optimal application timing for disease management. In summary, the preponderance of the literature indicates that glyphosate can act as a fungicide in GR crops when applied at the right time. Thus, glyphosate has probably provided significant, unintended, and unmeasured plant protection from some fungal phytopathogens in GR crops.

Figure 5.

Effect of glyphosate on leaf rust (Puccina triticina) on GR wheat 13 days after inoculation. (A) No spray. (B) Formulated glyphosate (0.84 kg ae/ha) applied 14 days before inoculation. (C) Formulated glyphosate applied 1 day before inoculation. Reproduced with permission from ref (64). Copyright 2004 National Academy of Sciences USA.

As with any fungicide, glyphosate could potentially reduce mycorrhizal infections as well as the presence of other beneficial fungi in GR crops. However, Savin et al.⁷⁶ found no effect of glyphosate on mycorrhizal infection in GR soybean, maize, and cotton, and Hart et al.⁷⁷ found no effects of glyphosate on rhizosphere-denitrifying fungal communities in GR maize.

The highly phloem-mobile property of glyphosate could add to its utility as a fungicide. However, the fact that it is a high use rate, foliar-applied herbicide that can be used only directly in GR crops makes its use as a fungicide less attractive. Glyphosate is the only commercial herbicide with EPSPS as a molecular target, suggesting that there are no other suitable chemical analogues or classes with the same target. The protein structures of fungal and plant EPSPS could differ sufficiently for a fungal EPSPS-specific inhibitor to exist that would not be active on plant EPSPS. Despite the heavy use of glyphosate, it took more than 20 years for glyphosate-resistant weeds to evolve after the introduction of glyphosate as a herbicide in 1974. However, due to the enormous selection pressure over decades there is a rapidly growing number of species with evolved resistance and more mechanisms of evolved resistance than for any other herbicide.⁷⁸

A glyphosate-resistant Aspergillus oryzae was isolated from sludge near a glyphosate production plant.⁷⁹ The mechanism of resistance was not determined from this apparently evolved case of glyphosate resistance in a fungus. Resistance to glyphosate was quickly selected for in several citrus postharvest fungal pathogens,⁸⁰ and glyphosate-resistant Fusarium verticillioides was isolated from an agricultural soil that had been treated with glyphosate for multiple years.⁸¹ Thus, one can expect resistance to evolve if a fungicide that targets EPSPS is widely used.

In summary, although glyphosate has been the most successful of all herbicides, finding an EPSPS inhibitor fungicide with a new chemistry would be challenging.

Glutamine Synthetase

Glutamine is produced by glutamine synthetase (GS) (HRAC, group 10) in both fungi and plants. GS is the target of only one herbicide, glufosinate (Figure 6). Glufosinate has become a major nonselective herbicide, partly because of the increased use of transgenic, glufosinate-resistant crops after glyphosate-resistant weeds became a major problem. Several natural and a few synthetic GS inhibitors are known.⁸² Bialaphos (Figure 6) was a minor herbicide product produced by fermentation of Streptomyces hygroscopicus that breaks down in planta to l-phosphinothricin, the active enantiomer of glufosinate, thus acting as a pro-herbicide. Glufosinate is a chemically synthesized racemic mixture of l- and d-phosphinothricin.

Figure 6.

Commericial inhibitors (glufosinate and bialaphos) of glutamine synthase (GS) and one natural product (tabtoxin) GS inhibitor.

Two reviews on fungicide targets in amino acid biosynthesis pathways do not discuss GS.^31,32 Despite this, GS is involved in several aspects of fungal development. For example, Wang et al.⁸³ reported it is involved in growth, conidiation, sclerotia development, and resistance to oxidative stress in Aspergillus flavus. Furthermore, the expression of the gene encoding GS is highly upregulated during the pathogenesis of C. gloeosporioides.⁸⁴ Glufosinate acts as a fungicide in GR creeping bentgrass (Agrostis palustris)^85,86 and GR rice⁸⁷⁻⁹⁰ that have been made resistant to glufosinate with a transgene that encodes an enzyme that detoxifies the herbicide. An example of such results is seen in Figure 7, in which glufosinate provided clear protection from two fungal pathogens (M. grisea and Cochliobolus miyabeanus) in glufosinate-resistant rice.⁹⁰ The amino acid sequence of GS in Arabidopsis and B. cinerea is high (88%) (Table 3), supporting the hypothesis that good inhibitors of plant GS would also be effective against fungal GS.

Figure 7.

Effects of glufosinate (+) applied 24 h before inoculation on the development of rice blast (Magnaporthe grisea) and brown leaf spot (Cochliobolus miyabeanus) in nontransgenic (NC) and glufosinate-resistant (bar) rice 10 days after inoculaton. Reproduced with permission from ref (90). Copyright 2008 American Society of Plant Biologists.

Even though GS is apparently essential to fungi, the fungicidal effect of glufosinate was concluded to be primarily indirect by inducing host defense mechanisms via caused oxidative stress by a low-level herbicidal effect on the host plant.⁹⁰ Inhibition of the chloroplast form of GS by glufosinate causes rapid production of reactive oxygen species, resulting in lethal oxidative stress.⁹¹ However, glufosinate is fungitoxic in in vitro bioassays at low concentrations with several plant pathogens,⁹² and concentrations of glufosinate that do not harm grapevines inhibited growth of the oomycete Plasmopara viticola, the cause of downy mildew on grapevines.⁹³ Unfortunately, only formulated glufosinate was used in this study, so formulation ingredients could account for some or all of the activity. Formulated glufosinate was sufficiently fungitoxic to the fungal bioinsecticide Beauveria bassiana in an in vitro assay to indicate that they are incompatible.⁹⁴ Hoagland et al. found that 0.25 mM unformulated glufosinate was toxic to the fungal microbial bioherbicide Colletotrichum truncatum in an in vitro bioassay.⁹⁵ Wang et al. found in vitro exposure to glufosinate at 1.9 and 2.5 mM (concentrations that are used for weed control) to completely inhibit the growth of the plant pathogens Sclerotinia homeocarpa (recently reclassified as Clarireedia spp.) and Rhizoctonia solani, respectively, leading them to recommend glufosinate for pathogen control in glufosinate-resistant turf grass.⁸⁶In vitro studies with Aspergillus flavus found that 1.1 mM glufosinate inhibited growth by 60% and aflatoxin B production by 80%.⁹⁶ Black et al. reported that the recommended field rate of glufosinate (impossible to convert to molarity from their information) reduced the growth of R. solani in vitro by 58–78%, depending on the fungal isolate.⁹⁷

Thus, glufosinate has direct and, under some circumstances, indirect fungicidal effects on plant pathogens. Whether these effects are additive, synergistic, or antagonistic has not been determined. In a glufosinate-resistant crop, the herbicide would have to cause stress to the crop at the application rate used for induced pathogen resistance to be in play. Glufosinate use in glufosinate-resistant crops has undoubtedly contributed to fungal disease management in these crops, although, to the best of our knowledge, this has not been documented. Although no commercial fungicide targets GS, current knowledge suggests that GS could be a fungicide target site. Bialaphos and glufosinate have strong antifungal activity against R. solani, Sclerotinia homeocarpa (Clarireedia spp.), and Pythium aphanidermatum to support this view.⁸⁵ There are little published data on the fungicidal activity of the several noncommercial GS inhibitors. Exceptions are the apparent effect of the synthetic plant GS inhibitor l-methionine sufoximine on GS of yeast, in which it greatly reduces growth and glutamine levels and on ecotomycorrhizal fungi.^98,99 The natural product GS inhibitor tabtoxin which is a good inhibitor of plant GS, apparently also inhibits yeast GS.¹⁰⁰ Whether any of the other GS inhibitors would be more fungicidal than glufosinate is unknown.

How quickly resistance might evolve to a GS inhibitor fungicide is unknown. Glufosinate resistance has not evolved quickly in weeds,¹⁰¹ but use of glufosinate is rising,¹⁰² increasing the selection pressure. Target site resistance in weeds has evolved to glufosinate by both evolution of glufosinate-resistant GS (e.g., (103)) and gene amplification by increased expression of the chloroplastic GS gene,¹⁰⁴ but resistance is not yet widespread. Metabolic, nontarget site resistance to glufosinate has also been reported in a weed.¹⁰⁵ The short generation time of fungi might contribute to rapid evolution of glufosinate resistance if a GS-inhibiting fungicide were used. The plant pathogen Verticilium albo-atrum is apparently naturally highly tolerant to phosphinothricin (the active enantiomer of glufosinate),¹⁰⁶ but there is no evidence that evolved glufosinate-resistance of a fungus has occurred in an agricultural field due to selection by glufosinate.

Phytoene Desaturase

Bacteria, fungi, and green plants synthesize carotenoids, which require the desaturation of phytoene via a phytoene desaturase (PDS) to produce lycopene as one of the steps in the pathway. Carotenoids are an absolute requirement for chlorophyll and chloroplast stability in light, and there are seven commercial herbicides (HRAC group 12) that inhibit its synthesis by directly inhibiting PDS. Fungi also synthesize carotenoids, and PDS is involved in desaturation of phytoene to produce lycopene.^107,108 The amino acid sequence of PDS in Arabidopsis and B. cinerea is low (34%) (Table 3). Different fungi have different forms of PDS that desaturate phytoene to different degrees, yielding neurosporene, lycopene, or 3,4-didehydrolycopene.^109,110 However, the fungal PDS enzyme differs significantly from that of cyanobacteria and plants, which only desaturates phytoene to ζ-carotene.¹¹¹

PDS from the Gram-negative bacterium Erwinia uredovora was unaffected by the highest soluble concentrations of the PDS inhibitor herbicides norflurazon, fluridone, and flurtamone, but was weakly inhibited by diphenylamine (not a commercial herbicide).¹¹² Plastoquinone (PQ) is a required cofactor for plant PDS activity, and PDS inhibitor herbicides compete for the PQ binding site,¹¹³ whereas fungal PDS uses NADP as a cofactor instead of PQ.¹⁰⁸ A transgene encoding Erwinia uredovora PDS confers strong norflurazon resistance to the cyanobacterium Synechococcus PCC7942 and to several commercial PDS inhibitors to tobacco.^114,115 These results suggest that chemically different PDS inhibitors from those used for weed management might be developed as fungal PDS inhibitors.

How fungitoxic will an efficient fungal PDS inhibitor be? Some mutant fungi that do not produce carotenoids grow normally in the laboratory, apart from a lack of pigmentation.¹¹⁶ However, evidence exists that carotenoids play a role in the field in protection from UV exposure and oxidative stress as well as in the production of photoreceptors needed as triggers for physiological responses to light. Almost ubiquitous, unnecessary biosynthetic pathways seem improbable in nature. A study to show the effects of a strong fungal PDS inhibitor on a plant pathogen under field conditions is needed to determine whether PDS could be a good molecular target for a fungicide.

Target site resistance to some PDS inhibitors has evolved in the aquatic weed hydrilla (Hydrilla verticullata) that was selected by fluridone,^117,118 but not in terrestrial plants. This is because the way the critical codon change is coded in hydrilla requires only one base pair change for an amino acid that provides resistance, whereas in terrestrial plants studied, different coding is used for the same amino acid, requiring two base pair changes to produce the amino acid needed for resistance. Only PDS inhibitor resistance due to enhanced metabolic degradation is found in terrestrial plants. We do not know what amino acid changes in fungal PDS will provide resistance to a fungal PDS inhibitor.

Protoporphyrinogen Oxidase

All organisms that synthesize heme, including plants and fungi, require protoporphyrinogen oxidase (PPO) to produce protoporphyrin IX (ProtoIX), the last compound before iron is inserted in this cyclic tetrapyrrole by a ferrochelatase. PPO is also essential for chlorophyll synthesis, and 22 commercial herbicides from six chemical groups (HRAC group 14) inhibit this enzyme. In plants, high levels of ProtoIX accumulate outside the plastid when PPO is inhibited.^119,120 ProtoIX is a photosensitizer, producing high levels of reactive oxygen species (ROS) in the presence of light and molecular oxygen. PPO inhibitor herbicides act more rapidly than almost all other herbicides because of this mechanism of action. ProtoIX does not have a molecular target, and thus it is toxic to all life forms in the presence of light and molecular oxygen, including fungi.¹²¹ Ironically (pun intended), hemoproteins are involved in the detoxification of ROS, so the toxicity of accumulated ProtoIX in the light and in an aerobic environment is exacerbated by reductions in these enzymes by PPO inhibitors. In fungi, PPO is found only in the mitochondrion (see Labbé et al.¹²² for a review of heme biosynthesis in fungi). The amino acid sequence of PPO in Arabidopsis and B. cinerea is moderate (58%) (Table 3). Fungal PPO is inhibited by herbicidal PPO inhibitors,^123,124 but we are unaware of attempts to determine their efficacy as a fungicide. Camadro et al. found yeast PPO to have less requirement for lipophilicity of inhibitors in vitro than that of plant PPO.¹²⁵ We are unaware of any efforts to develop PPO inhibitors as fungicides that act directly on fungal PPO.

PPO-inhibiting herbicides cause mild oxidative stress in crops in which they are used, and this can induce the production of phytoalexins. For example, the diphenyl ether PPO inhibitor herbicide acifluorfen increases phytoalexin levels in several crops,¹²⁶ and it is mentioned on the label of a similar PPO inhibitor herbicide, lactofen, that it can prevent white mold (Sclerotinia sclerotiorum) in soybean.¹²⁷ Soybeans are tolerant of both acifluorfen and lactofen by rapid metabolic degradation of the herbicides, but these herbicides cause mild oxidative stress in the crop. Lactofen also reduces Sclerotinia stem rot (S. sclerotiorum)¹²⁸ and sudden death syndrome (Fusarium virguliforme)¹²⁹ infections in soybeans. Whether the effects on severity of pathogen infection are entirely due to lactofen induction of biosynthesis of the phytoalexin glyceollin¹³⁰ or is also partly due to direct effects on fungal PPO is unknown. Several other PPO inhibitor herbicides also promote glyceollin biosynthesis in soybean, as does rose Bengal, a photodynamic compound that causes ROS formation.¹³¹In vitro studies of PPO inhibitor effects on fungal plant pathogens are needed to determine whether PPO inhibitors can act as fungicides through effects on fungal PPO.

Evolution of resistance to PPO inhibitor herbicides has evolved by both target site and nontarget site mechanisms.³⁰ The target site resistance is due to a codon deletion (a unique target site resistance mechanism), and resistance to these herbicides did not evolve quickly. A fungicide that acts on fungal PPO might be used with a PPO inhibitor-resistant crop¹³² or with crops that are naturally tolerant to some PPO inhibitor herbicides via rapid metabolic degradation, such as soybeans.

Very Long Chain Fatty Acid Synthase

HRAC (2022, group 15) lists 35 commercial herbicides belonging to eight chemical classes that kill plants by inhibiting very long chain fatty acid synthases (VLCFASs). Biosynthesis of VLCFAs involves multiple fatty acid elongase complexes that consist of four core enzymes and other proteins, making it difficult to decide which proteins to compare between fungi and plants. For this reason, this enzyme is left out of Table 3. VLCFAs are found in fungi.^133,134 Much of the early work on VLCFAS was done with S. cerevisiae.¹³⁵ In yeast, VLCFAs are found mainly in sphingolipid synthesis, which is essential for yeast growth.¹³⁴ They are also found as storage lipids, phospholipids, and glycosylphosphatidylinositol anchors in yeast.¹³⁵ VLCFAs are required for spore production in red yeast (Sporobolomyces spp.).¹³⁶ Yeast VLCFAS is not affected by some herbicides that inhibit plant VLCFASs (e.g., napropamide), but others (e.g., alachlor and flufenacet) do inhibit it.¹³⁷ Similarly, not all herbicide VLCFAS inhibitors inhibit all plant VLCFASs. Several herbicidal chloroacetanilide VLCFAS inhibitors (e.g., metolachlor and alachlor) are weakly fungicidal to Phytophthora megasperma f. sp. medicagninis compared to the acylalanine fungicide metalaxyl.¹³⁸ This study did not link the activity to inhibition of VLCFAS. There do not appear to be any definitive studies of any of the commercial VLCFAS inhibitor herbicides as fungicides against plant pathogens.

Considering that these herbicides have been used extensively for many years, there are relatively few cases (13) of evolved resistance,¹⁰¹ and none of the cases have been reported to be target site resistance. Thus, resistance to a fungicide with this target might also develop slowly.

Dihydropteroate Synthase

Asulam is the only commercial herbicide that inhibits dihydropteroate synthase (DHPS, HRAC group 18). This enzyme is required for folate synthesis in fungi and plants, but is not present in animals,¹³⁹ making it an appealing target for both fungicides and herbicides. The amino acid sequence of DHPS of Arabidopsis and B. cinerea is moderate (56%) (Table 3). In S. cerevisiae, DHPS is encoded in a polycistronic gene that also encodes dihydroneopterin aldolase.¹⁴⁰ Bacterial DHPS is the target of sulfonamide and sulfone antibiotics. Acrylamide–sulfisoxazole conjugates and sulfonamide analogues that inhibit DHPS have been designed to treat human fungal pathogens.^141,142 Several of the compounds of Nasr et al. were as fungicidal as sulfisoxazole and amphotericin B to the three fungi of medicinal concern that were tested.¹⁴¹ One of the compounds studied by Othman et al. was more active against Fusarium oxysporum and Candida albicans than amphotericin B.¹⁴² No in vitro enzyme studies were performed in either study, but molecular docking studies indicated DHPS as the molecular target. We have found no evidence that the DHPS inhibitor compounds from either study have been tested as agricultural fungicides. Cai et al. tested 34 novel sulfonamides against B. cinerea in a quest for new agricultural fungicides.¹⁴³ Most of the compounds studied were highly effective against the fungus, and some were better than commercial fungicides. No cross resistance to these compounds with other fungicides was found. However, the molecular target(s) of these compounds were not determined, even though they are likely to be DHPS inhibitors. The Cai al al. paper lists patents and earlier papers on similar compounds that are likely to be DHPS inhibitors.¹⁴³

Although, selection of asulam resistance in plants can be done rapidly (e.g., Gifford et al.¹⁴⁴), evolved resistance to asulam in weeds is not yet recorded in the International Herbicide-Resistant Weed Database,¹⁰¹ even though it is an older herbicide. Asulam resistance was claimed for Panicum dichotomiflorum, although the mechanism of resistance was not reported.¹⁴⁵ Resistance to DHPS inhibitor antibiotics in bacteria is widespread,¹⁴⁶ and thus, overuse of an agricultural DHPS inhibitor fungicide might promote rapid evolution of fungal resistance. The same amino acid changes in DHPS that have evolved in Pneumocystis jiroveci provide resistance to the S. cerevisiae form of the enzyme,¹⁴⁷ so resistance to a fungicide that targets DHPS of fungal plant pathogens might evolve similarly. Asulam-resistant potatoes have been produced to kill parasitic weeds that infest this crop through use of a bacterial gene for an asulam-resistant DHPS.¹⁴⁸ This technology could be used to make crops resistant to a fungicide that targets DHPS.

Asulam is fungitoxic to some plant pathogens such as Pythium arrhenomanes,¹⁴⁹ so it is possible that asulam use in some crops has provided some level of plant disease protection, in addition to weed control.

p-Hydroxyphenylpyruvate Dioxygenase Inhibitors

p-Hydroxyphenypyruvate dioxygenase (HPPD, HRAC group 27) is required for the conversion of tyrosine to plastoquinone (PQ) and tocopherols in green plants.¹⁵⁰ PQ is required for photosynthetic electron transport and PDS (see above) activity, and tocopherols contribute to the quenching of ROS generated by the photosynthetic apparatus. At least part of the phytotoxicity of HPPD inhibitors is due to the accumulation of toxic levels of tyrosine.¹⁵¹ Inhibition of HPPD is the last major MoA of herbicides introduced, with six triketones, six pyrazoles, and isoxaflutole commercialized herbicides and other HPPD inhibitor herbicides under development.¹⁵² Isoxaflutole and several of the pyrazole HPPD inhibitors are pro-herbicides.

In fungi, HPPD is involved in conversion of tyrosine to fumarate and acetoacetate, which are necessary for metabolism, as well as in production of melanin-type pigments (e.g., Bolognese et al.¹⁵³) The amino acid sequence homology of HPPD in Arabidopsis and B. cinerea is moderate (49%) (Table 3). Production of these pigments that are implicated in pathogenicity of fungal plant pathogens has been proposed as a target for new fungicides.¹⁵⁴ Deletion of HPPD by mutation in Aspergillus nidulans increases tyrosine and hydroxyphenylpyruvic acid levels and stops its growth.¹⁵⁵ However, inhibition of HPPD by the HPPD-inhibitor herbicide sulcotrione did not inhibit growth, but completely inhibited pyomelanin synthesis and caused p-hydroxyphenylpyruvic acid (HPPA) levels to increase in tyrosine-fed Aspergillus fumigatus.¹⁵⁶ Earlier studies found sulcotrione to inhibit heterologously expressed wheat leaf-spot fungus Mycosphaerella graminicola HPPD involvement in pigment production on E. coli.¹⁵⁷ The mesotrione analogue, nitisinone (NTBC), a HPPD inhibitor pharmaceutical for treating hereditary defects in human tyrosine metabolism, inhibits growth and differentiation of the fungus Paracoccidioides brasiliensis(158) and mycelial growth of the mammalian fungal pathogen Coccidiodies immitis.¹⁵⁹

Growth and pathogenicity of rice blast (Magnaporthe oryzae) is inhibited by 9-phenanthrol (Figure 8), and overexpression of HPPD protects the fungus from this compound.¹⁶⁰ This group also found this chemical to be fungitoxic to F. oxysporum, Exserohilum turcicum, Bipolaria maydis, and Rhizoctonia solani and to be phytotoxic to rice, cabbage, and tomato, but not to maize. In M. oryzae, levels of HPPA were reduced by 9-phenanthrol treatment, indicating that it might act as an inhibitor of tyrosine amino transferase (TAT), the enzyme immediately before HPPD in the tyrosine degradation pathway. No enzyme assays were performed to determine which enzyme is inhibited. This paper indicates that interruption of this pathway with inhibitors of either TAT or HPPD in fungi should be fungitoxic. Because major crops are also affected, crops with either a resistant TAT or HPPD would be desirable for use with a fungicide that functions as 9-phenanthrol.

Figure 8.

Compounds mentioned in the text.

We have found no reports on the effects of commercial HPPD inhibitor herbicides on fungal diseases in field situations. However, the existing literature on the effects of HPPD inhibitors on plant pathogenic fungi indicates that such a phenomenon occurs at some level. No effect of soil-applied isoxaflutole on colonization of maize roots by the arbuscular mycorrhizal fungus Glomus intraradices was found in a greenhouse study,¹⁶¹ indicating that there may not be any adverse effects of HPPD inhibitors on mycorrhizal associations with crops.

Transgenic, HPPD herbicide-resistant soybeans are available, and other HPPD inhibitor-resistant crops (particularly cotton) are being developed. Some crops are naturally tolerant to some HPPD inhibitors (e.g., maize and mesotrione).¹⁵² Thus, a fungicide with HPPD as its target site could probably be found that would not affect the crop, depending on the compound and the mechanism of resistance or tolerance of the crop. Weed resistance has evolved to some HPPD inhibitor herbicides,^30,152 but no target site-based resistance has been reported. Mutations that cause specific amino acid changes in bacterial HPPD provide resistance to the commercial triketone HPPD inhibitor topramezone,¹⁶² suggesting that target-site-based resistance can be expected with enough selection pressure. Thus, target site resistance may occur with a HPPD-targeted fungicide unless a resistance prevention/mitigation strategy is used with such a product.

Fatty Acid Thioesterase

Acyl-ACP thioesterase (fatty acid thioesterase, FAT or fatty acid synthase, FAS), an enzyme involved in midchain length lipid synthesis, is found in both plants and fungi. Acyl-ACP thioesterases belong to a large protein family. The Arabidopsis genome encodes several genes that have been annotated as potential FATs. Because FATs have not been annotated in the B. cinerea genome yet, this target is not included in Table 3. Six commercial herbicides (HRAC group 30), including cinmethylin, target this enzyme.^163,164 FATs are also required for fatty acid production in fungi.^165,166 We have found no literature on the effects of these herbicides on fungi. However, the natural microbial metabolite cerulenin inhibits this enzyme in both fungi (S. cerevisiae) and plants.^167,168 Cerulenin was first mentioned as an antifungal antibiotic and later reported to be phytotoxic.¹⁶⁹ Analogues have been produced in a search for more effective fungicides and herbicides.^170,171

Heap (2023) does not record any weeds evolving resistance to any herbicides with this molecular target, perhaps because they have been used very little.¹⁰¹

Ser/Thr Protein Phosphatase

The herbicide endothall (Figure 8) acts by inhibiting the activity of serine/threonine protein phosphatases (PPP) by binding to the active site (HRAC group 31).¹⁷² This family of phosphoprotein phosphatases dephosphorylates amino acids serine and threonine, which are found in all eukaryotes. These essential enzymes partner with kinases in signal transduction cascades. The distinct origins, and hence the variability, of the nucleotide sequences of some PPP types are a reason why all types of PPP might not be sensitive to this herbicide. Not surprisingly those PPPs affected by endothall have an exceptionally conservative catalytic domain.^173,174 The amino acid sequence of A. thaliana and B. cinerea homologues share a relatively high (82%) level of similarity (Table 3). Endothall is the only herbicide with this molecular target. Cantharidin (Figure 8), a well-known serine/threonine PPP inhibitor derived from insects, is more phytotoxic than endothall to some plant species.¹⁷² Cantharidin is a close analogue of endothall and may have been the inspiration for the discovery of endothall.¹⁷⁵ No in vitro studies have been done comparing the activities of endothall and cantharidin on serine/threonine PPP of plants and fungi.

This enzyme has been well studied in fungi, especially in S. cerevisiae.¹⁷³ It is required for the pathogenesis of S. sclerotiorum, as its inhibition by cantharidin inhibits growth and sclerotia production of the fungus.¹⁷⁶ Cantharidin was found to be fungitoxic to six fungal plant pathogens.¹⁷⁷ An effort to produce a PPP inhibitor fungicide produced a cantharidin analogue that is fungicidal to S. sclerotiorum.¹⁷⁸ Low rates of endothall are compatible with a fungal plant pathogen (Mycolelptodiscus terrestris) used as an experimental mycoherbicide for management of the aquatic weed Hydrilla verticillata,¹⁷⁹ indicating that the weed is more sensitive to endothall than the fungus. Earlier work with endothall and the mycoherbicide Colletotrichum gloeosporioides on another aquatic weed, Myriophyllum spicatum, gave similar results.¹⁸⁰ A PPP inhibitor that requires higher doses for crop damage than that used as an effective fungicide would be needed. No information is available on PPPs of the two fungi and plant species, but S. cerevisiae genome contains 13 PPPs presumably sensitive to cantharidin. This number also includes type PP2B/calcineurin (not sensitive to cantharidin) also present in animal genomes but absent in plants.¹⁸¹ There are Arabidopsis genes for as many as 26 different catalytic subunits of this family of enzymes.¹⁸² Not only could the number of genes encoding the target protein be the reason for the distinct responses to the same treatment but the existence of other physiological differences between the two kingdoms such as cell wall composition, a barrier that restricts the movement of the compound into the cell, could play a role. In general, the fungal cell wall contains mainly polysaccharides like chitin, β-(1,3) glucan and cell wall proteins that reach about 50% of the dry weight of walls.¹⁸³ While plant cell walls contain about 90% polysaccharides like cellulose and pectin.¹⁸⁴

Cantharidin is highly toxic to mammals¹⁸⁵ even though it has been used as a pharmaceutical for hundreds of years and continues to be studied as a therapeutic, especially in anticancer research.^186,187 Thus, finding an inhibitor of serine/threonine PP that is sufficiently specific for this enzyme in fungi may be required for regulatory approval of a fungicide with this MoA. This may be difficult, because the active site of this family of enzymes is highly conserved.

Only one case of evolved resistance to endothall has been reported.¹⁰¹ The mechanism of the resistance has not been determined. Target site resistance may be unlikely because plants have a high number of PPP targets, and indirect evidence indicates that endothall inhibits all of them and that the efficacy between different isozymes varies.¹⁷² The same may be true for fungal PPPs. Hence, engineering a crop to be resistant to endothall or another PPP inhibitor would probably have to be based on a transgene for an enzyme that rapidly degrades endothall.

Lycopene Cyclase

A relatively old herbicide, amitrole, is the only herbicide that acts by the inhibition of lycopene cyclase (HRAC group 34), an enzyme required for the final step in β-carotenoid synthesis. Its inhibition leads to in vivo accumulation of lycopene. Fungi, such as Phycomyces and Rhodosporidium diobovatum, also have this enzyme, which is encoded by a gene that also encodes phytoene synthase in some fungi.^188,189 The amino acid sequence of lycopene cyclase in Arabidopsis and B. cinerea is low (29%) (Table 3). Amitrole is highly fungitoxic to five fungal species,¹⁹⁰ but there is no proof that its fungitoxicity is due to inhibition of lycopene cyclase. Its fungitoxic activity on Phytophthora parasitica was attributed to inhibition of histidine synthesis,¹⁹¹ at one time a proposed MoA of amitrole as a herbicide.¹⁹² Lycopene synthase inhibitors have been studied to increase lycopene accumulation in fungi,^193,194 but we find no literature to indicate that this enzyme has been studied as a potential target for fungicides, despite the reports of fungitoxicity to some fungi.

There is relatively little (five species) evolved resistance to amitrole after decades of use,¹⁰¹ and none of it is target site based.³⁰ Inhibition of both lycopene cyclase and histidine biosynthesis would make the evolution of target site-based resistance highly unlikely. Such a situation could exist for both plants and fungi.

Future Directions

All new crop protection compounds, including herbicides and fungicides, require favorable human safety profiles, physiochemical properties, and environmental safety. These requirements have led to “likeness” concepts as a means to identify crop protection compunds that will have the appropriate properties and environmental profiles needed for a successful commercial crop protection compound.^195−197 These same likeness concepts can also be employed in the repurposing of compounds from one therapeutic area to another. The present review has identified a few prospective overlaps between molecular targets of some herbicides that may potentially apply to fungicides. At present, analyses are lacking that can potentially define overlaps in molecular targets between crop protection therapeutic areas (fungicides, herbicides, insecticides). While such information and systems to exploit this information may exist internally in some crop protection companies, these types of analyses do not appear to currently exist in the published literature. As such, efforts to repurpose herbicides, and other chemical classes, to fungicides and vice versa remain challenging.

Five of the molecular targets (ACCase, GS, HPPD, FAT, and PPP) discussed in this article that are shared by plants and fungi are also shared by mammals. This has not been a safety issue for the commercial herbicides that act on these targets for reasons unique to each target. Nine of the 23 (39%) molecular targets of commercial herbicides listed by the HRAC are shared by mammals. Each of these herbicides has met rigorous safety requirements from regulatory agencies of the countries in which they are used. The fact that these products are considered safe as herbicides suggests that similar products that have these targets in fungi would also be safe as fungicides; however, they would still be subject to rigorous safety testing.

Computational tools, including quantitative structure–activity relationships (QSAR) and virtual screening, have long been a focus in the discovery of crop protection compounds.^6,198,199 The recent growing interest in artificial intelligence (AI) provides another computational avenue to exploit opportunities for chemistries or targets in one therapeutic area to another by identifying potential overlaps. These potential overlaps between therapeutic areas (i.e., herbicides to fungicides) could point to new fungicide targets and/or chemical scaffolds for fungicides as has been done for pharmaceuticals and recently suggested for herbicides.^200−202 Recent studies highlight the potential of this approach, wherein a machine learning-based cheminformatics analysis of chemical similarities between pharmaceuticals and herbicides led to the recognition of the pharmaceultical HPPD inhibitor nitisinone and the antimicrobial triclocarban as herbicide-like molecules (Figure 9).²⁰³ Indeed, both molecules were shown to have a high herbicidal activity. AI and machine learning has been used to determine the kinase binding sites of a fungicide that is active on B. cinera.²⁰² Thus, we suggest that future studies aimed at identifying new fungicide chemistries should consider employing AI-based tools to identify the overlap of molecular spaces of binding sites and physiochemical properties of commercialized molecules to increase the rate of discovery and development for novel fungicides.

Figure 9.

Pharmaceuticals with herbicidal activity.

A key consideration for any new crop protection product is resistance management. In light of the increasing cases of resistance to fungicides (Figure 1),^2,204 it is important to protect the long-term efficacy and investment through the judicious use of any new product. As noted above, many of the proposed fungicide MoAs have the potential for target site resistance, which may influence the rate of resistance development if a fungicide with that MoA is used exclusively for an extended period of time. As such, resistance management will likely be an important consideration, which in the cases of fungicides commonly involves the use of fungicide rotations and especially fungicide mixtures to minimize the chances for resistance development.^2,205 Likewise, the potential new MoAs for fungicides outlined above could provide a source of new mixing partners for currently available fungicide MoAs.

A major problem with compounds that can be both herbicidal and fungicidal is the difficulty in using them in crop protection because of differences in application times, doses, and growth cycles. There are essentially no agrochemicals that are clearly sold for both weed and plant pathogen management. The only exception is lactofen, sold as a PPO inhibitor herbicide, which mentions on its use label for soybeans that it controls white mold (S. sclerotiorum) and sudden death syndrome (F. virguliforme) is applied prior to infection and before the soybean full bloom growth stage (https://www3.epa.gov/pesticides/chem_search/ppls/083529-00078-20170726.pdf). However, in this lone case, the chemical is not acting directly as a herbicide but as an inducer of phytoalexins by the crop. Producing herbicide-resistant crops can potentially overcome this limitation,⁴⁴ but the cost and complexity of launching a transgenic or otherwise genetically altered crop along with a new herbicide/fungicide might be daunting. New approaches to the alteration of crop genetics (e.g., CRISPR-Cas9) might reduce the cost of such an endeavor. As recently reviewed by Jeschke,²⁰⁶ another approach to this problem is to produce a pro-pesticide form for the chemistry, such that only fungicidal activity is realized by metabolic activation of the compound by the targeted fungal pathogen.

The authors declare no competing financial interest.

Special Issue

Published as part of Journal of Agricultural and Food Chemistryvirtual special issue “AGRO Division 50th Anniversary”.