How does Fungicide Resistance evolve?

The term fungicide resistance, as used by FRAC, refers to an acquired, heritable reduction in sensitivity of a fungus to a specific anti-fungal agent (or fungicide).  To manage resistance effectively, scientists study fungicide resistance on many different levels including the cellular, organismal or population/field level.  Reports of "resistance" from the field (i.e. where growers observed reduced efficacy of a product that has been effective against that particular pathogen) must be confirmed by studies at the organismal level showing a reduction in sensitivity of the fungal isolate(s) to the specific fungicide.  Some scientists use the terms reduced sensitivity or tolerance when referring to smaller reductions in sensitivity which may have little to no impact on fungicide usage in the field, and save the term "resistance" for large reductions in sensitivity of individual isolates which are likely to affect the efficacy of a specific fungicide under field conditions if the resistant isolates become widespread in the pathogen population.  The term field resistance may also be used to indicate this loss of control under field conditions.

The development of fungicide resistance is a population evolutionary process.  Fungi, like other organisms, are constantly changing.  Occasionally, under certain conditions, these changes provide an advantage or disadvantage in terms of the progeny’s ability to survive and reproduce. Advantageous changes allow the individual containing the change to survive and reproduce resulting in their progeny constituting a greater percentage of the population over subsequent generations. This can happen relatively rapidly in fungi as their reproductive frequency (i.e. the number of progeny produced from a single individual and the speed with which they complete their life cycle) is high.  For example, a single Phytophthora infestans lesion can produce thousands of spores and a spore can produce a new sporulating lesion in 3-5 days.  The change may be evolutionarily neutral, or even slightly disadvantageous, under most conditions and only be advantageous when certain factors are present.  This is the case with fungicide resistance.  In most cases of fungicide resistance, the change leading to reduced sensitivity is evolutionarily neutral except when the specific fungicide is applied.  The fungicide is exerting selection pressure on the pathogen population since it is killing the initial (or wild type) population but does not kill the changed (or mutant) population.  When changes are slightly disadvantageous under normal conditions (i.e. in the absence of the fungicide), the frequency of the changed population may decrease when the selection pressure is removed.  This disadvantage is termed a fitness penalty.

Fungicides inhibit fungal growth by interfering with critical cellular processes.  Mode of action (MOA) refers to the specific cellular process inhibited by a particular fungicide.  FRAC currently lists 11 modes of action in its Mode of Action Poster and Code List (Table 1, column 1).  Fungicides may have a mode action that is not fully understood at the time of introduction.  Until there is specific evidence of the involved biochemical processes, FRAC list such compounds as “Unknown Mode of Action” in the FRAC code list. 

Within each mode of action there are specific sites of action.  These sites of action or target sites are the specific enzymes in a cellular process to which the fungicides bide.  For example, both strobilurin fungicides and SDHI fungicides share the same MOA (inhibition of respiration) but have different sites of action in the respiratory pathway; SDHIs inhibit complex II while strobilurins inhibit complex III.  In pharmaceutical literature the inhibition of a specific enzyme is referred to as the mechanism of action (some plant pathologists use the term mechanism of action interchangeably with mode of action).  FRAC assigns compounds active at the same target site a number (e.g. the SDHIs are FRAC group #7 on the FRAC Code list). 

To understand the biochemical interaction of the fungicide with a specific target site, the analogy of using a lock and key to open a door is useful.  The lock is the target site on the enzyme and the key is the natural substrate the enzyme interacts with to complete normal cellular processes (analogous to unlocking the door). The fungicide(s) active at that target site are an additional set of key-like objects which can also fit into the lock.  If one of these “artificial” keys is in the lock, the normal key/substrate cannot fit into the lock and the fungal biochemical process is blocked (the door cannot be unlocked). Inside the cell, the fungicide and the substrate compete for the lock/target site.  As the fungicide accumulates in the cell, the normal substrate can no longer access the target site and normal cellular processes will reach such a low level, or may be blocked entirely, that adverse effects are observed.  While the fungicides share similarity with the fungal substrate in terms of their three-dimensional structure, they are not identical.  It is possible that alterations to the lock/target site could occur that will allow the fungal substrate to continue to bind and normal cellular processes to proceed, but not allow the fungicide(s) to bind (i.e. the lock could be changed such the original key still works to unlock the door but the key-like fungicide no longer fits in the lock).   This situation results in one specific type of resistance known as target-site resistance.

Fungicides active at the same target site (i.e. that is within the same FRAC code # on the FRAC Code List) are generally considered to be cross-resistant to each other.  Cross-resistance is a phenomenon that occurs when resistance arises to one fungicide that also results in resistance to another fungicide.  Occasionally, cross-resistance can occur between compounds active at different target sites (see multi-drug resistance under mechanisms of resistance below). The actual target site is not completely understood for some fungicides so the target site description remains rather generic.  For example, the target site description of the azanapthalenes is signal transduction.  The two azanapthalenes, quinoyxfen and proquinazid, are grouped together in the same FRAC group since cross resistance was observed in Erysiphe necator.  Interestingly, no cross-resistance was observed in another powdery mildew species, Blumeria graminis  (Genet and Jaworska, 2009).  Negative cross-resistance can also occur.  Negative cross resistance is when a change results in a reduction in sensitivity to one fungicide and an increase in sensitivity to another fungicide.  For example, isolates of Botrytis cinerea with reduced sensitivity to the benzimidazole fungicides (FRAC group #1 on the FRAC code list) have an increased sensitivity to the N-phenyl carbamates (FRAC group #10 on the FRAC code list; Leroux et al 1989).

There are 4 main mechanisms by which fungi can become resistant to fungicides.

1. Alteration of the target site so that sensitivity to the fungicide is reduced:  By far the most common way that fungi can become resistant to a specific fungicide is via a change at the target site. As fungi grow their DNA is replicated when new cells are created.  This process of replication is imperfect and errors can occur.  These errors are known as mutations.  Since DNA is the code used to produce enzymes in the cell, some mutations result in changes to the amino acid sequence of the target site which in turn alters the shape of the lock/target site.  The fungicide/key may not fit as well anymore or may not fit at all in the target site/lock.  This results in a reduction in sensitivity that may range from small to very large.

2. Detoxification or metabolism of the fungicide:  The fungal cell contains a vast array of metabolic machinery for normal cellular processes.  This metabolic  machinery may be able to modify the fungicide to a non-toxic form that is no longer harmful to the cell.  Some fungicides are applied as inactive pro-fungicides which require further metabolism by the fungal cell to become the active form.  If fungal metabolism is altered such that the activation step does not occur the active form of the fungicide is not produced. 

3. Overexpression of the target:  As discussed above, the fungicide is “competing” with the natural substrate for the target site.  As more and more fungicide enters the cell, it out-competes the natural substrate for the target and as a result shuts down critical cellular processes. The production of additional target site enzyme (i.e. overexpression of the target) may increase the likelihood that enough of the fungal substrate will be able to bind with the target site enzyme such that cellular processes such as respiration can occur to some degree.  Higher doses of the fungicide in in vitro experiments may restore the balance in favor of the fungicide, but higher doses may not always be practical under field conditions.

4. Exclusion or expulsion from the target site: Efflux pumps exist naturally within the cell to exclude or expel foreign substances or to export endogenous substances.  In fungi, the most common efflux pumps are ABC and MFS transporters.  Despite these efflux pumps, most fungicides can reach effective concentrations inside the cell and inhibit cellular processes.  Occasionally, these transporters are successful in expelling enough of the fungicide such that an isolate has reduced sensitivity. The fungicides expelled from the cell by a specific transporter may or may not be active at the same target site; i.e. there is not a direct relationship between the transporter that expels a specific fungicide and the target site of the fungicide.  Multidrug resistance (MDR) develops when a specific transporter is able to exclude multiple fungicides from different target site groups.  Application of the fungicides in question may exert enough selection pressure that isolates containing these fungicide-exporting transporters become more prevalent in the population as is the case in Botrytis cinerea (Kretschmer et al. 2009).

Genet JL, Jaworska G.  Baseline sensitivity to proquinazid in Blumeria graminis f. sp. tritici and Erysiphe necator and cross‐resistance with other fungicides. Pest management science 2009; 65(8): 878-884.

Leroux P, Gredt, M. Negative cross-resistance of benzimidazole-resistant strains of Botrytis cinerea, Fusarium nivale and Pseudocercosporella herpotrichoides to various pesticides.  Netherlands Journal of Plant Pathology 1989; 95(1): 121-127.

Kretschmer M, Leroch M, Mosbach A, Walker AS, Fillinger S, Mernke, D, Schoonbeek HJ, Pradier JM, Leroux P, De Waard M,  Hahn M. Fungicide-driven evolution and molecular basis of multidrug resistance in field populations of the grey mould fungus Botrytis cinerea. PLoS pathogens 2009; 5(12);e1000696.

Apply it as a mixture with one or more fungicides of a different type, or as one component in a rotation or alternation of different fungicide treatments.

The ‘companion’ or ‘partner’ compounds applied in either of these ways will dilute the selection pressure exerted by the at-risk fungicide and inhibit the growth of any resistant biotypes that arise. The companion compound can be a multi-site compound known to have a low risk of inducing resistance. Alternatively, it can be a single-site fungicide that is not cross-resistant to or related to its partner (in the absence of known resistance) by a similar mode of action. Use of a mixture of two single-site fungicides must carry some element of risk of selecting dual-resistant strains. However, the chances of two mutations occurring simultaneously will be very small compared to that of a single mutation (e.g. 10-18 instead of 10-9). Consecutive development of double resistance could occur, but the likelihood is much lower than if the two fungicides were used separately and repeatedly.

This type of strategy is widely recommended by industry and also by advisory bodies. The use of formulated (‘pre-packed’) mixtures of two different fungicides has often been favoured by manufacturers. If an at-risk fungicide is not sold alone, the mixture is the only use option open to the farmer and implementation of the strategy is ensured. Also, the control of many pathogens only requires one or two treatments per annum so that the rotational approach is not applicable. Mixtures are also marketed for other purposes, such as broadening the range of pathogens which can be controlled or enhancing control by increasing the duration of protection. Questions of what application rate is appropriate for each mixture component are difficult and have been debated many times. Some reduction relative to the full recommended separate rates has often been made, to keep down costs. This may reduce selection pressure for the ‘at risk’ fungicide, but clearly it is vitally important to maintain the companion compound at a level where it can still exert an effective independent action against the target pathogens.

Theoretical and empirical research as well as practical experience suggest that both mixture and rotation strategies have delayed resistance development (examples are discussed in the FRAC-Monograph 1). However, fully conclusive evaluations of com mercial-scale strategies are difficult to make because comparable ‘non-strategy’ areas have seldom existed.

This approach, like rotation, reduces the total number of applications of the at-risk fungicide and therefore must slow down selection to some extent.  It can also favor decline of resistant strains that have a fitness penalty. However, the treatments, which are still applied consecutively, generally coincide with the most active stages of epidemics when selection pressures are highest. Thus, any delay in resistance may not be proportional to the reduction in the number of sprays.  On the other hand a substantial break in use at a time when the pathogen is still multiplying can allow a beneficial resurgence of more sensitive forms. Examples are considered later.

For many years farmers have often applied reduced doses of fungicides, mainly to reduce costs. The practice is common especially in conditions where disease pressure is usually low, or where the risk of financial loss from reduced performance is not high. Also, advisory services in pursuing lower-input approaches for economic and environmental reasons, have recommended use of smaller doses for certain situations. FRAC's view is that recommended doses must be maintained, not only because they will retain the built-in safety factor and secure the claimed levels of performance under a wide range of conditions, but more importantly because it is possible that reducing the dose could enhance the development of resistance.

However, relationships between fungicide doses and risk of resistance are not yet fully established, and it seems likely that they may vary according to the fungicide in question. Some of the models referred to above indicate that lowering the dose of the at-risk fungicide (but retaining normal spray frequency) can delay build-up of major gene resistance by decreasing the overall effectiveness, increasing the numbers of sensitive survivors and hence slowing down the selection of resistant forms that can survive the full dose. For multi-step resistance, it has been argued that lowering dose can enhance resistance development by favouring the survival of low-level resistant genotypes which would be inhibited by the full dose. The low-level resistant strains could then mutate further or recombine sexually to give higher levels of resistance. In practice, however, the doses that actually reach the target organisms vary greatly over space and time, giving very complex mixes of different exposure sequences (examples are discussed in the FRAC-Monograph 1).

It is now widely accepted, on theoretical grounds, limited experimental data and practical experience, that risks of major-gene (single-step) resistance are unlikely to increase, and may well decline as dose is lowered. The situation with regard to polygenic resistance is still not clear and more experimental work is required to justify recommendations for lowering doses. Some of the published data refer specifically to ‘split’ schedules, in which dose is lowered but frequency of application is correspondingly increased, to give the same total mount applied each season. It is important to distinguish these from reduced-dose applications made on normally timed schedules so that the total dose per season is decreased. Indeed, the use of more frequent ‘split’ applications could increase resistance risk and should be avoided.

One of the advantages of systemic fungicides is that they can eradicate or cure existing infections. This property greatly assists their use on a ‘threshold’ basis, where application is made only when a certain, economically acceptable, amount of disease has already appeared, in order to prevent further spread. However, avoidance of the use of systemic fungicides in this way has been recommended in two different situations as an anti-resistance strategy:

1. FRAC has recommended that eradicant use of e.g. of phenylamides should be avoided. This is because they are now always applied for control of foliage diseases as a mixture with a multi-site companion fungicide. The latter does not work as an eradicant, so that the systemic fungicide, e.g. phenylamide, is acting alone when the mixture is applied to existing infections.
2. Opportunity for selection could be much greater than if the fungicide had been applied as a preventive treatment to keep populations permanently low. When forecast models and threshold populations are used to determine the spray timing, then a threshold population of the pathogen is present. Usually this means that many sporulating lesions (occupying up to 5% of the foliar area) are exposed to the fungicide.

This is a particular aspect of the concept generally referred to as IPM (Integrated Pest Management). The integrated use of all types of countermeasures against crop disease is not only highly desirable on economic and environmental grounds, but is also a major strategy for avoiding or delaying fungicide resistance. The use of disease resistant crop varieties, biological control agents, and appropriate hygienic practices, such as crop rotation and removal of diseased parts of perennial crop plants, reduces disease incidence and permits the more sparing use of fungicides, and in both these ways decreases selection of fungicide-resistant strains. Equally, the application of fungicides reduces the risk of build-up of pathotypes with changed virulence and the consequent ‘breakdown’ of disease-resistant varieties.
Unfortunately, non-chemical methods of disease control are often weak or not available, so that fungicide application is the predominant or even the sole countermeasure for many diseases (e.g. potato late blight, grape downy mildew, Sigatoka disease of bananas, wheat bunt, stripe (yellow) rust of wheat, to name a few).

The availability of a number of different types of fungicides for the control of each major crop disease is highly beneficial to both the environment and to mitigate resistance problems. The continued use of one or a very few types of compound over many years presents a much greater risk of side-effects and favours resistance in the target organisms. Thus it is crucial that chemical invention and new product development are sustained. Fortunately, registration authorities now accept the need for diversity, in terms of pesticide chemistry and mechanisms of action, provided that the new compounds maintain safety standards. A new fungicide does not necessarily have to be superior to existing ones in order to be of value. It has to be effective, and, in the resistance context, it should work against strains that are resistant to existing fungicides.
This latter property is usually associated with a new mode of action, and ideally there should be more than one site of action to decrease the risk of evolution of resistance to the new fungicide. However, the development of new, highly active members of an existing fungicide class, which retain the same primary mechanism of action, may also be of some use in resistance management. The withdrawal of fungicides for safety reasons has been necessary from time to time, but it has reduced options for resistance avoidance strategies. Hopefully, plans for future withdrawal of fungicides consider the impact on the diversity of modes of action in order to sustain the means of control of important fungal diseases in crop production.