Progress and gaps in understanding mechanisms of ash tree resistance to emerald ash borer, a model for wood-boring insects that kill angiosperms
Summary
We review the literature on host resistance of ash to emerald ash borer (EAB, Agrilus planipennis), an invasive species that causes widespread mortality of ash. Manchurian ash (Fraxinus mandshurica), which coevolved with EAB, is more resistant than evolutionarily naïve North American and European congeners. Manchurian ash was less preferred for adult feeding and oviposition than susceptible hosts, more resistant to larval feeding, had higher constitutive concentrations of bark lignans, coumarins, proline, tyramine and defensive proteins, and was characterized by faster oxidation of phenolics. Consistent with EAB being a secondary colonizer of coevolved hosts, drought stress decreased the resistance of Manchurian ash, but had no effect on constitutive bark phenolics, suggesting that they do not contribute to increased susceptibility in response to drought stress. The induced resistance of North American species to EAB in response to the exogenous application of methyl jasmonate was associated with increased bark concentrations of verbascoside, lignin and/or trypsin inhibitors, which decreased larval survival and/or growth in bioassays. This finding suggests that these inherently susceptible species possess latent defenses that are not induced naturally by larval colonization, perhaps because they fail to recognize larval cues or respond quickly enough. Finally, we propose future research directions that would address some critical knowledge gaps.
Contents | ||
---|---|---|
Summary | 63 | |
I. | Introduction | 64 |
II. | Emerald ash borer life cycle and host range | 64 |
III. | Mechanisms of ash resistance to emerald ash borer | 65 |
IV. | Nutritional quality and primary metabolites | 71 |
V. | Conclusions and future directions | 72 |
Acknowledgements | 75 | |
References | 75 |
I. Introduction
The invasion of North America by emerald ash borer (EAB), Agrilus planipennis Fairmaire (Coleoptera: Buprestidae), an Asian wood-boring beetle, has caused the widespread mortality of ash trees (Fraxinus spp., Oleaceae), with devastating ecological and economic impacts (Cappaert et al., 2005; Poland & McCullough, 2006; Kovacs et al., 2010, 2011). Since its accidental introduction, which can be traced back to at least the early 1990s in southeast Michigan (Siegert et al., 2014), EAB has been detected in 25 US States and two Canadian Provinces (as of June 2015), from the east coast to Colorado, and from southern Quebec to Georgia and Louisiana (http://www.emeraldashborer.info/files/MultiState_EABpos.pdf). North American populations of EAB are characterized by very low genetic diversity, most likely because they originated from a single introduction from China (Bray et al., 2011). The EAB has also established itself in Russia, where it is causing extensive mortality of European ash (F. excelsior L.) (Baranchikov et al., 2008; Straw et al., 2013; Orlova-Bienkowskaja, 2014). Ash species indigenous to eastern Asia share a coevolutionary history with EAB and are more resistant than evolutionarily naïve hosts (Wei et al., 2004, 2007; Liu et al., 2007; Rebek et al., 2008). Several reviews have focused on the biology, ecology, impact, and management of EAB (Cappaert et al., 2005; Poland & McCullough, 2006; Herms & McCullough, 2014). Here, we review the literature on mechanisms underlying inter- and intraspecific variation in ash resistance to EAB, and identify future research directions that would address a number of remaining knowledge gaps.
II. Emerald ash borer life cycle and host range
EAB adults emerge from larval hosts from mid-spring through mid-summer depending on location. They complete c. 1−2 wk of maturation feeding on foliage before they mature reproductively, during which time they also disperse, occasionally for long distances. Upon mating, females oviposit on bark surfaces or in crevices and eggs hatch within 10–14 d (Cappaert et al., 2005; Wang et al., 2010; Herms & McCullough, 2014). Immediately upon eclosion, larvae chew through the outer bark and feed on vascular tissue at the phloem–xylem interface, producing serpentine galleries that disrupt translocation of water, nutrients, and carbohydrates (Cappaert et al., 2005). Girdling of branches and the trunk leads to canopy decline followed by tree death (Cappaert et al., 2005; Herms & McCullough, 2014). In stressed trees, nearly all EAB develop within 1 yr with larvae feeding from mid-summer into autumn, during which time they complete four instars before overwinter as prepupae in the outer sapwood or bark. Pupation occurs in mid to late spring and adults emerge soon thereafter (Cappaert et al., 2005; Wang et al., 2010; Chamorro et al., 2012). A 2-yr life cycle is common in healthy trees and in northern habitats, as slower developing larvae overwinter as early instars, feed during a second summer, then emerge the following spring (Crook & Mastro, 2010; Tluczek et al., 2011). Although all instars can overwinter, pupation does not occur until after prepupae have overwintered (Cappaert et al., 2005; Herms & McCullough, 2014) (Fig. 1).

Host species indigenous to East Asia have coevolved with the EAB, including Manchurian ash (F. mandshurica Rubr.) and Chinese ash (F. chinensis Roxb.) (Wei et al., 2004, 2007) (the latter species is now considered a species complex that includes F. japonica Blume ex K. Koch and F. rhynchophylla Hance; Wallander, 2008). All North American ash species so far encountered by EAB are suitable larval hosts (Herms & McCullough, 2014). In particular, black (F. nigra Marshall), green (F. pennsylvanica Marshall), and white ash (F. americana L.) are highly vulnerable (Cappaert et al., 2005; Poland & McCullough, 2006; Rebek et al., 2008; Klooster et al., 2014), while blue ash (F. quadrangulata Michaux.) has experienced less EAB colonization or mortality than white and green ash where they co-occur (Anulewicz et al., 2008; Tanis & McCullough, 2012; 2105; Carson, 2014). Pumpkin ash (F. profunda (Bush) Bush) (Knight et al., 2013), Oregon ash (F. latifolia Benth.) (Herms, 2015), velvet ash (F. velutina Torr.) (Wang et al., 2010), and freshly cut logs of Shamel ash (F. uhdei (Wenz.) Lingl.) (Duan et al., 2013) have also been documented as larval hosts.
Species indigenous to Europe confirmed as hosts for EAB larvae include European ash (Orlova-Bienkowskaja, 2014), flowering ash (F. ornus L.), and raywood ash (F. oxycarpa Willld. cv ‘Raywood’) (Herms, 2015). To date, the only non-Fraxinus species shown to support the complete development of EAB is white fringetree (Chionanthus virginicus L., Oleaceae) (Cipollini, 2015). The EAB was not able to complete development in other species in the Oleaceae that have been tested, including Japanese tree lilac (Syringa reticulata Bl.), Chinese privet (Ligustrum sinense Lour.), glossy privet (L. lucidum Ait.), and swamp privet (Forestiera acuminata (Michx.) Poir.) (Anulewicz et al., 2006, 2008).
Common garden studies and observations in Asia have shown that Manchurian and other Asian ash species that have coevolved with EAB are more resistant (but not immune) to attack than North American and European species (Liu et al., 2003, 2007; Wei et al., 2004, 2007; Rebek et al., 2008; Herms, 2015). In fact, significant colonization of Asian ash species by EAB has been reported consistently only when trees are stressed and dying (Wei et al., 2004, 2007; Liu et al., 2007), which further suggests that they are inherently resistant. Thus, it appears that EAB has evolved as a secondary colonizer of stressed trees, as is the case with many species of Buprestidae (Muilenburg & Herms, 2012). Hence, resistance of Asian ash species to EAB probably results from targeted defenses selected for by natural selection imposed over their coevolutionary history.
III. Mechanisms of ash resistance to emerald ash borer
While mechanisms of conifer resistance to bark beetles have been intensively studied (Raffa & Berryman, 1983; Franceschi et al., 2005; Kolosova & Bohlmann, 2012), fewer investigations focused on mechanisms of resistance of deciduous trees to wood-borers before the discovery of EAB in North America (Dunn et al., 1990; Hanks, 1999; Morewood et al., 2004; Muilenburg et al., 2011, 2013). Recent advances in understanding interactions between EAB and ash could serve as a model for other tree-killing buprestids and their hosts.
Tree resistance to wood-borers is a function of female oviposition preference and larval performance (Hanks, 1999). Hence, we find it useful to classify potential resistance mechanisms of ash to EAB as antixenosis, antibiosis, and tolerance (Painter, 1951; Kogan & Ortman, 1978). Antixenosis results from traits that decrease behavioral preferences for feeding and/or oviposition. Antibiosis is the outcome of traits that negatively affect insect growth, survival, and/or fecundity. Finally, tolerance is the ability to endure levels of herbivory with less impact on growth, yield, or fitness, than a more susceptible plant experiencing an equivalent amount of damage. Tolerance to larval feeding is probably an important component of resistance to wood-borers, especially because of the extensive disruption of water and nutrient transport caused by removal of vascular tissue (Nielsen et al., 2011). However, the ability of trees to tolerate wood-borers is difficult to quantify because of logistical challenges associated with measuring a form of herbivory that is often cryptic and demonstrating heritable variation in host response to larval feeding, and we are not aware of any studies that have quantified tolerance of ash to EAB.
Both antixenosis and antibiosis can result from physical and chemical traits. Physical (or mechanical) defenses consist of structural elements, such as lignified and/or suberized cell walls, which provide a barrier against insect feeding and oviposition (Franceschi et al., 2005), while chemical defenses consist of secondary metabolites (henceforth, specialized metabolites; Pichersky & Lewinsohn, 2011) that can repel, deter feeding, or otherwise be detrimental to the insect (Chen, 2008; Howe & Jander, 2008). Resistance traits can further be categorized as either constitutive (occurring pre-attack) or inducible (produced in response to attack) (Karban & Myers, 1989; Agrawal, 2007; Eyles et al., 2010). Some authors also classify plant traits that result in food limitation (pre-ingestion) and/or those that decrease nutrient value (post-ingestion) as antinutritive defenses (Chen, 2008).
1. Antixenosis: adult maturation feeding and oviposition preferences
EAB adults express variation in host preference for maturation feeding and oviposition. Oviposition preference is likely the most relevant for explaining variation in host resistance, as defoliation caused by maturation feeding is negligible. The same tree or species may not be the optimal or preferred host for maturation feeding, oviposition preference by females, or larval performance. Hence, adult behavioral preferences might be driven by different host traits at different times. During the maturation feeding period, adults may select foliage based on traits that maximize their survival and fertility, while ovipositing females may select hosts on which their offspring have the highest probability of survival when feeding on vascular tissue. The relationship between adult feeding preferences and oviposition preference has yet to be fully characterized for EAB (Pureswaran & Poland, 2009).
1.1. Maturation feeding
EAB adults can discriminate between ash species as hosts for maturation feeding. In a choice test, Pureswaran & Poland (2009) found that adults preferred to feed on green, black, and white ash relative to Manchurian, blue, and European ash. This pattern of feeding preferences corresponds broadly with patterns of ash mortality observed in the field (Anulewicz et al., 2007; Rebek et al., 2008; Tanis & McCullough, 2012; Klooster et al., 2014), with the most preferred species for maturation feeding also experiencing highest rates of mortality. An exception is European ash, which was not as highly preferred by feeding adults (Pureswaran & Poland, 2009), but has experienced widespread mortality in forests (Orlova-Bienkowskaja, 2014) and a common garden planting (Herms, 2015). This suggests that adult maturation feeding preferences and performance might generally correspond with female oviposition preference and larval performance. However, this remains to be demonstrated experimentally, and may not necessarily be the case. For example, consumption and adult longevity of bronze birch borer (Agrilus anxius Gory) females were higher on cottonwood (Populus deltoides Bartr) and pin oak (Quercus palustris Münchh), which are not larval hosts, than on European white birch (Betula pendula Roth) (Akers & Nielsen, 1990), which is a highly suitable larval host (Muilenburg & Herms, 2012). Furthermore, fecundity of females was just as high on the two non-larval hosts as on the highly susceptible larval host (Akers & Nielsen, 1990).
EAB adults can also exhibit variation in host preference for maturation feeding within a host species. For example, EAB adults preferred feeding on mature over young leaves, sun-exposed over shaded plants, and girdled over non-girdled plants, with patterns associated with environmental effects on foliar concentrations of defensive compounds, sugars, amino acids, and proteins (Chen & Poland, 2009a,b). Adults survived longer on mature leaves, but survival was not otherwise consistently affected by host nutritional quality (measured as protein, amino acids, and nonstructural carbohydrate content), perhaps because of compensatory feeding on hosts with lower nutritional quality (Chen & Poland, 2009b).
There is evidence that volatile compounds emitted by foliage play a role in host location and adult feeding preferences of the EAB (Rodriguez-Saona et al., 2006; de Groot et al., 2008; Pureswaran & Poland, 2009; Crook & Mastro, 2010). Pureswaran & Poland (2009) hypothesized that adult feeding preferences may be inversely related to total emission of whole host volatiles, because they found that susceptible green ash emitted lower levels than resistant Manchurian ash. However, the quantitative relationship between total volatile emission and host preference was not consistent with this hypothesis for the other ash species tested (Pureswaran & Poland, 2009). This indicates that qualitative variation in volatile profiles is likely also to be important.
Rodriguez-Saona et al. (2006) reported that adult virgin females (but not males) were attracted to whole plant volatiles from Manchurian ash that had been induced by adult feeding or application of methyl jasmonate (MeJA), and suggested that host volatiles may play a role in long-range host finding by females. EAB adults are also responsive to green leaf volatiles (de Groot et al., 2008). Lures baited with (Z)-3-hexenol were attractive to EAB adults, with males more strongly attracted than females. Antennae of males also responded more strongly to green leaf volatiles than antennae of females (de Groot et al., 2008), and Crook & Mastro (2010) suggested that green leaf volatiles may play a role in mate location by males, as mating occurs on host plants. Abiotic and biotic factors, including EAB adult and larval feeding, can also alter the volatile emissions of ash foliage (Rodriguez-Saona et al., 2006; Crook & Mastro, 2010; Chen et al., 2011a,b), perhaps affecting host preferences of feeding adults.
1.2. Oviposition preferences
Because the devastating host impact of EAB is due to larval feeding, female oviposition preferences could be a key determinant of inter- and intraspecific variation in ash resistance to EAB. Rigsby et al. (2014) examined oviposition preferences of the EAB in common garden studies and found that susceptible North American ash species consistently received more eggs than resistant Manchurian ash, which is generally consistent with patterns of adult landing and feeding preferences (Anulewicz et al., 2008; Pureswaran & Poland, 2009), larval density (Tanis & McCullough, 2015), exit hole numbers (Rebek et al., 2008; Whitehill et al., 2014), and host mortality (Rebek et al., 2008) observed in other studies. Despite the fact that the EAB and North American ash species lack a coevolutionary history, the findings of Rigsby et al. (2014) are consistent with optimal oviposition theory, which predicts a positive correlation between adult oviposition preference and larval performance (Jaenike, 1978; Gripenberg et al., 2010). This correspondence is particularly important for wood-boring insects because it is impossible for larvae to disperse if they find themselves in a low quality host (Hanks, 1999). Anulewicz et al. (2008) found that females preferred to oviposit on green and white ash relative to blue ash, which may contribute to the lower levels of mortality observed for blue ash (Anulewicz et al., 2007; Tanis & McCullough, 2012), despite it being an acceptable host for EAB larvae (Peterson, 2014). Collectively, these studies suggest that oviposition preference is an important determinant of interspecific variation in ash mortality and decline observed in field studies.
Experimental studies also show that EAB prefers to oviposit on stressed relative to healthy trees within a species. In field experiments, the landing rates of adults were higher on girdled trees, as were larval densities (McCullough et al., 2009; Tluczek et al., 2011). Jennings et al. (2014) found that females preferred to oviposit on declining ash trees that were previously infested by the EAB relative to healthy trees. Cappaert et al. (2005) and Tluczek et al. (2011) have also shown that larvae develop faster in girdled or previously infested trees. Hence, oviposition preferences of EAB females for stressed hosts is consistent with predictions of optimal oviposition theory (Jaenike, 1978), and the hypothesis that the EAB evolved as a secondary colonizer of weakened trees.
Crook et al. (2008) found that girdling green ash trees increased bark emission of volatile sesquiterpenes, which elicit an EAB antennal response, with mated females more sensitive than virgin females and males (Crook & Mastro, 2010). Traps baited with Phoebe oil containing these sesquiterpenes captured significantly more adults than unbaited traps (Crook et al., 2008). Collectively, these observations suggest that the EAB may utilize trunk volatiles as cues for locating stressed host trees (Crook & Mastro, 2010) and perhaps even as oviposition cues, although this has yet to be demonstrated.
2. Antibiosis, specialized metabolism, and insect counter adaptations
The specialized metabolism of Fraxinus is very complex (Kostova & Iossifova, 2007), and analysis of traits contributing to interspecific variation in larval performance have focused mainly on the phenolic and defensive protein chemistry of bark (in its extended meaning – including periderm, cortex (in early stages of growth), phloem, cambium, and the outer layer of xylem) (Eyles et al., 2007; Cipollini et al., 2011; Whitehill et al., 2011, 2012, 2014; Chakraborty et al., 2014). However, interspecific comparisons can be confounded by genetic variation in unrelated traits, making it important to control for phylogeny (Agrawal, 2011). Based on the most parsimonious explanation for diversification, blue ash (section Dipetalae) diverged early in North America from green and white ash (section Melioides). Early ash progenitors spread from North America to Eurasia, where the section Fraxinus diverged, including into Manchurian and European ash, then ultimately migrated back to North America where black ash speciated (Hinsinger et al., 2013). Hence, while North American green, white, blue, and black ash are sympatric, they are phylogenetically divergent. Therefore, comparison of resistant Manchurian ash with closely related but susceptible black and European ash may be the most profitable for identifying defensive traits (Whitehill, 2011; Whitehill et al., 2012).
2.1. Phenolic compounds
Phenolic compounds, which can act as feeding deterrents, toxins, and digestion inhibitors (Schultz, 1989), are the prevailing class of specialized metabolites in ash, with coumarins, secoiridoids, and phenylethanoids particularly characteristic of the genus (Kostova & Iossifova, 2007). Consistent with this, Bai et al. (2011) found that many of the gene transcripts of ash they sequenced mapped to the phenylpropanoid biosynthetic pathway. Other classes of phenolic compounds isolated from ash include phenolic acids, simple phenolics, monolignols, lignans, and flavonoids (Eyles et al., 2007; Kostova & Iossifova, 2007; Cipollini et al., 2011; Whitehill et al., 2012, 2014; Chakraborty et al., 2014).
The constitutive phenolic profiles of Manchurian ash cv ‘Mancana’, green ash cv ‘Patmore’, and white ash cv ‘Autumn Purple’ were characterized for dormant stems by Eyles et al. (2007), and for stems harvested during the growing season (when EAB larvae actively feed) by Cipollini et al. (2011). Both studies detected phenylethanoid glycosides (calceolariosides A and B) and the hydroxycoumarin fraxidin hexoside in Manchurian ash, but not in green and white ash, and suggested that these compounds may be determinants of resistance. However, the two studies also reported somewhat discordant lists of compounds unique to Manchurian ash, a discrepancy that may be attributed to the different vegetative states of the trees at the time they were analyzed. Eventually, Whitehill et al. (2012) conducted a more comprehensive analysis, including six ash species, in which a total of 66 phenolics displayed a pattern of variation that strongly tracked phylogeny (Fig. 2). The authors observed similar or even greater concentrations of some compounds that were initially considered Manchurian ash-specific by either Eyles et al. (2007) or Cipollini et al. (2011), in susceptible black and European ash (which are phylogenetically closely related to Manchurian ash). This suggests that those compounds are probably not involved in resistance of Manchurian ash to the EAB, unless they participate in synergistic interactions with other undetermined compounds or processes unique to the resistant species (Whitehill et al., 2012).

The authors also concluded that a specific derivative of the lignan pinoresinol (a dihexoside) and a coumarin derivative, both identified for the first time in their study, were unique to Manchurian ash and hypothesized that they contributed to host resistance (Whitehill et al., 2012). The potential role of lignans in Manchurian ash resistance is supported by (1) the higher constitutive expression in Manchurian ash of a phenylcoumaran benzylic ether reductase, an enzyme involved in lignan biosynthesis (see later; Table 1) (Whitehill et al., 2011); (2) the higher concentrations of other pinoresinol related compounds in Manchurian ash than in green, white, and black ash (Eyles et al., 2007; Cipollini et al., 2011; Whitehill et al., 2012; Chakraborty et al., 2014); and (3) the fact that these compounds have broad biological activities against insects, including as antifeedants and as inhibitors of larval molting and growth (Cabral et al., 1999; Garcia et al., 2000).
NCBI accession numbera | Protein match ID | GOb annotation for Biological Process |
---|---|---|
gi¦886683 | Major allergen (Malus × domestica) | Defense response |
gi¦7578895 | Phenylcoumaran benzylic ether reductase Fi1 (Forsythia × intermedia) | Metabolic process |
gi¦13897888 | Putative aspartic protease (Ipomoea batatas) | Proteolysis |
gi¦25992557 | Thylakoid-bound ascorbate peroxidase (Triticum aestivum) | Hydrogen peroxide catabolic process |
- a Searching the NCBI Peptidome Archive (ftp://ftp.ncbi.nih.gov/pub/peptidome/studies/PSEnnn/PSE148/PSM1314/PSM1314_manchurianAsh_resultTable.txt) using the NCBI accession number of the matched protein will lead to detailed information about the peptides identified in the study. Manchurian ash peptide information can be obtained through Peptidome sample accession number PSM1314 (ftp://ftp.ncbi.nih.gov/pub/peptidome/studies/PSEnnn/PSE148/).
- b GO, gene ontology.
Whitehill et al. (2012) also found that the phenolic composition of blue ash bark differed substantially from green, black, and white ash. They suggested that the high concentrations of the hydroxycoumarin esculin found in blue ash may contribute to the greater EAB resistance of this host species relative to the other North American species (Anulewicz et al., 2008; Tanis & McCullough, 2012, 2015; Carson, 2014).
It is noteworthy that all three studies examined clonal cultivars that may not be representative of the species as a whole. Hence, it is significant that Cipollini et al. (2011) and Whitehill et al. (2012, 2014) compared the metabolic profile of the green ash cultivar to that of an open-pollinated population of green ash, and found them to be very similar, suggesting that the phenolic profile of the green ash cv ‘Patmore’ is characteristic of the species (see for example Fig. 2).
All ash species examined thus far contain secoiridoid glycosides, most of which have phenolic moieties (Eyles et al., 2007; Kostova & Iossifova, 2007; Cipollini et al., 2011; Whitehill et al., 2012). Secoiridoid glycosides are structurally related to the monoterpene iridoids, a well-known class of anti-herbivore defenses (Bowers, 1991). The secoiridoid glycoside oleuropein, for instance, is a primary defense of privet (Ligustrum spp.) and its antifeedant activity may derive from its strong protein-complexing alkylating activity after it is hydrolyzed by plant β-glucosidases (Konno et al., 1999; Spadafora et al., 2008). Indeed, β-glucosidase activity, rather than oleuropein content per se, differentiated susceptible and resistant varieties of privet (Konno et al., 1999) and olive (Spadafora et al., 2008). Constitutive concentrations of oleuropein were lower in Manchurian ash than in susceptible ash species (Eyles et al., 2007; Cipollini et al., 2011; Whitehill et al., 2012). However, oleuropein may still be an important defense if specific β-glucosidases are constitutively present or induced in resistant phenotypes, which would release more toxic alkylating aglycones (Konno et al., 1999).
In addition to imparting rigidity to cell walls, lignin (a phenolic polymer) is involved in plant defense against pathogens and insects (Vance et al., 1980; Franceschi et al., 2005), both constitutively (Wainhouse et al., 1990) and as an inducible trait (Vance et al., 1980; Nicholson & Hammerschmidt, 1992). Against wood-borers, it can act as a physical defense associated with stone cells (i.e. sclereids) in bark (Wainhouse et al., 1990; King et al., 2011). However, constitutive lignin concentration of susceptible and resistant ash species generally did not differ, suggesting that it does not contribute to interspecific variation in ash resistance to EAB (Cipollini et al., 2011; Whitehill et al., 2012). Nevertheless, lignin may contribute to intraspecific variation in resistance, as concentrations increased in white and green ash in response to application of MeJA relative to an insecticide-treated control, and this was associated with induced resistance to EAB (Whitehill et al., 2014). When incorporated into an artificial diet, lignin alkali (from spruce) was associated with marginally higher survival of larvae, although their growth was reduced in a dose-dependent fashion (Whitehill et al., 2014). If lignin has any role in EAB resistance of ash, it is probably related to an antinutritive mechanism, rather than to direct toxic effects on larvae, or its defensive capacity is exerted through physical modification of host cell walls (Whitehill et al., 2014).
The biological activity of phenolics may derive from their oxidation, quinylation, and/or deglycosylation (Felton et al., 1992; Appel, 1993). Accordingly, research on potential counter adaptations of EAB larvae to host defenses has focused on a suite of antioxidant defense genes, especially expressed during the larval stages in the midgut and Malpighian tubules, which may protect the insect from reactive oxygen species (ROS) that may be generated by oxidized phenols (Treutter, 2006; Rajarapu et al., 2011). Rigsby et al. (2015) found higher activities of ortho-quinone reductase, catalase, superoxide dismutase, and glutathione reductase activities in larvae that fed on Manchurian ash than in larvae that fed on white and green ash. Such responses indicate that larvae feeding on Manchurian ash are under greater amount of stress due to ROS and quinones than larvae feeding on more benign hosts. In fact, EAB larvae are capable of metabolizing phenolic compounds of susceptible green ash, either directly or indirectly through gut microflora (Rajarapu, 2013), as was demonstrated for Asian longhorned beetle (Anoplophora glabripennis Motschulsky) (Scully et al., 2014).
Potential counter adaptations of EAB larvae to other host defenses include genes encoding detoxification enzymes such as cytochrome P450s and glutathione-S-transferases (Mittapalli et al., 2010; Rajarapu et al., 2011; Rajarapu & Mittapalli, 2013). For example, EAB larvae that fed on green ash accumulated higher numbers of midgut transcripts coding for digestive enzymes, amino acid transporters, cytochrome P450s, and glutathione-S-transferases than larvae that fed on Manchurian ash. Conversely, larvae that fed on Manchurian ash accumulated more midgut transcripts coding for peritrophic membrane synthesis/repair, sugar transporters, serine proteases, carboxylesterases, sulfotransferase, and lactoylglutathione lyases than larvae that fed on green ash, suggesting a differential ability to detoxify defenses of susceptible and resistant ash species (Rajarapu, 2013). However, Rigsby et al. (2015) found no variation in activities of cytochrome P450, glutathione-S-transferase, carboxylesterase, sulfotransferase, and trypsin in larvae feeding on resistant and susceptible hosts, although the profile of functional trypsins varied.
2.2. Defensive proteins
Plant defensive proteins, including chitinases, trypsin inhibitors, polyphenol oxidases, and peroxidases, as well as related processes such as the rate of tissue browning, can have direct and indirect detrimental effects on insect herbivores (Ryan, 1990; Zhu-Salzman et al., 2008). Effects range from inhibition of digestive enzymes such as proteases and amylases, oxidation of ingested proteins and phenolics, deamination of essential amino acids, and disruption of the peritrophic membrane, which can cause mortality (Felton et al., 1989; Ryan, 1990; Vandenborre et al., 2011). Antinutritive proteins include protease and amylase inhibitors (Ryan, 1990), oxidative enzymes (Felton et al., 1992; Appel, 1993), and amino acid deaminases (Chen, 2008). Toxic proteins, which are involved in the disruption of the peritrophic membrane, include cysteine proteases, chitinases, lectins, and leucine aminopeptidases (Chen, 2008; Zhu-Salzman et al., 2008).
Studies of defensive proteins in ash bark have focused both on their constitutively expressed (Cipollini et al., 2011; Whitehill et al., 2011) and MeJA-induced (Whitehill et al., 2014) enzymatic activities. Constitutive activity of chitinases, polyphenol oxidases, and peroxidases of Manchurian ash did not differ from susceptible white and green ash (Cipollini et al., 2011). Furthermore, none of these proteins were affected by application of MeJA in any of the species tested (Whitehill et al., 2014). However, even in cases where polyphenol oxidases and peroxidases appear to be expressed at similar levels, they may still help explain variation in resistance to EAB through differential substrate usage, rather than their enzymatic activity, per se (Cipollini et al., 2011). Interestingly, although Manchurian ash does not appear to have higher constitutive polyphenol oxidase and peroxidase activities than white and green ash, it does have a faster browning reaction than white ash, with green ash intermediate (Cipollini et al., 2011). The browning of tissue in response to wounding occurs when endogenous soluble phenolics are oxidized by constitutive phenol oxidases and peroxidases, which can convert them to toxic and indigestible quinones and polyphenolics (Appel, 1993). Elevated peroxidase activity and a rapid browning reaction have been correlated with plant resistance to insects in several systems (Dowd et al., 1999). Because the browning reaction depends on both the presence of phenolics and the activity of phenol oxidizing enzymes, and because activities of oxidizing enzymes in Manchurian ash were not different from those in susceptible species, with the exception of ascorbate peroxidases (Whitehill et al., 2011), the rapid wound browning reaction in Manchurian ash may be driven mainly by its particular phenolic composition (Eyles et al., 2007; Cipollini et al., 2011; Whitehill et al., 2012). However, Whitehill et al. (2014) were not able to confirm the correlation between browning reaction and resistance of the host, and the molecular events involved in this reaction require further clarification (Chakraborty et al., 2011).
Constitutive activity of trypsin inhibitors was highest in green ash, intermediate in white ash, and lowest in Manchurian ash (Cipollini et al., 2011). Furthermore, application of MeJA-induced higher trypsin inhibitor activity in EAB susceptible green, white, and black ash relative to an insecticide-treated control (Whitehill et al., 2014). Trypsin inhibitors (be they protein-based or other kinds of inhibitors, such as low molecular weight compounds; Polya, 2003) decrease protein assimilation and insect growth by inhibiting digestive enzymes such as serine proteases (Ryan, 1990; Howe & Jander, 2008), which occur in the midgut of EAB larvae (Mittapalli et al., 2010; Rajarapu, 2013). The inhibition of serine proteases, for example, has been suggested as a potential defense strategy of oak against red oak borer (Enaphalodes rufulus Haldeman) (Crook et al., 2009). As expected, when soybean trypsin inhibitor was used as a proxy for unavailable ash trypsin inhibitors, it suppressed growth of EAB larvae fed on amended artificial diet in a dose-dependent manner (Whitehill et al., 2014).
Whitehill et al. (2011) focused on de novo discovery of potential defense-related proteins in ash by comparing the constitutive proteome of whole bark tissues of susceptible and resistant species (Fig. 3), and identified four proteins putatively involved in ash resistance to EAB whose expression was more than five-fold higher in Manchurian ash than in susceptible species. These include a major allergen (PR-10), an aspartic protease, a phenylcoumaran benzylic ether reductase, and a thylakoid-bound ascorbate peroxidase (Table 1) (Whitehill et al., 2011). The major allergen (PR-10) shares homology with the major allergen from Malus domestica L. Borkh., Mal d 1, which belongs to the Bet v 1 family of allergens (birch family allergens), and is classified as a pathogenesis-related (PR) protein (Breiteneder et al., 1989; Liu & Ekramoddoullah, 2006). PR-10 proteins function in defense and can be induced in response to microbial infection (Liu & Ekramoddoullah, 2006), and Whitehill et al. (2011) hypothesized that their high constitutive expression may also contribute to resistance of Manchurian ash to EAB.

Phenylcoumaran benzylic ether reductases participate in the biosynthesis of lignans (Gang et al., 1999), high constitutive concentrations of which have been found in Manchurian ash (Eyles et al., 2007; Cipollini et al., 2011; Whitehill et al., 2012; Chakraborty et al., 2014). These enzymes also belong to a class of cross-reactive allergens related to PR-proteins (Karamloo et al., 2001), and thus could have detrimental effects on insects (Whitehill et al., 2011).
Aspartic protease enzymes are involved in numerous processes, including antimicrobial defense (Davies, 1990; Ge et al., 2005). A cysteine protease in corn confers resistance to Spodoptera frugiperda (J. E. Smith) by degrading its peritrophic membrane (Pechan et al., 2002). Hence, Whitehill et al. (2011) hypothesized that the highly expressed aspartic protease may also contribute to EAB resistance of Manchurian ash. This hypothesis is strengthened by the observation that genes coding for peritrophic membrane synthesis and repair were more highly expressed in midguts of EAB larvae that had fed on Manchurian ash than on green ash (Rajarapu, 2013), suggesting that peritrophic membranes of EAB larvae may be damaged when feeding on Manchurian ash.
Ascorbate peroxidases are scavengers of ROS that reduce hydrogen peroxide (H2O2) generated during photosynthesis to water (Asada, 1999). ROS are involved in the metabolic cascade (oxidative burst) associated with induction of plant resistance to insects and pathogens (Doke et al., 1996; Apel & Hirt, 2004). Ascorbate peroxidases may also contribute to resistance of Manchurian ash to EAB (Whitehill et al., 2011) by oxidizing phenolics to quinones, which are less digestible by insects (Felton et al., 1992; Appel, 1993). However, functional analyses are required to clarify the exact roles of these proteins in EAB resistance.
3. Induced resistance
While constitutive defenses have been the primary focus of work on ash resistance to EAB, more recent studies have also explored the role of induced defenses. Chakraborty et al. (2014) quantified larval performance and the induction of phenolics in response to EAB larval feeding in Manchurian ash and black ash grown under high and low water availability and found that drought stress increased larval growth, which is consistent with the observation that stress predisposes coevolved ash species to colonization by EAB. However, drought had no effect on constitutive or induced levels of bark phenolics in either species. Furthermore, concentrations of most (but not all) phenolics decreased in response to larval feeding (Fig. 4). Collectively, these results suggest that traits other than phenolic concentration are primarily responsible for drought stress-induced susceptibility of black and Manchurian ash to EAB (Chakraborty et al., 2014). Larval feeding induced higher concentrations of the pinoresinol derivative pinoresinol A in both ash species, but to a higher degree in Manchurian ash (Chakraborty et al., 2014), which is consistent with the previously hypothesized role for selected lignans in resistance of Manchurian ash to EAB (Eyles et al., 2007; Cipollini et al., 2011; Whitehill et al., 2011, 2012).

Whitehill et al. (2014) exploited the capacity of MeJA to elicit defense responses in plants (Erbilgin et al., 2006; Koo & Howe, 2009) to investigate induced resistance of ash. They found that application of 1 M MeJA to the outer bark of ash reduced the number of EAB adult exit holes to levels comparable with those of insecticide-treated trees in otherwise susceptible ash species (i.e. white, green, and black ash), but had no effect on the constitutively resistant Manchurian ash (Whitehill et al., 2014). In addition to the effects on lignin and trypsin inhibitors discussed above, MeJA-induced accumulation of the phenylethanoid glycoside verbascoside in white and green ash. Verbascoside, which has been shown to have insecticidal properties (Muñoz et al., 2013), decreased growth and survival of EAB larvae feeding on an amended artificial diet (Fig. 5a,b).

While MeJA can successfully induce resistance in white, green, and black ash, these species experience nearly 100% mortality in forests (Klooster et al., 2014), which suggests that their inducible responses are not expressed strongly enough under natural conditions to be ecologically relevant, possibly because susceptible species fail to recognize cues associated with EAB colonization or respond to elicitors rapidly enough. However, the presence of latent resistance traits in susceptible North American ash species warrants further investigation as a part of resistance breeding programs.
The lack of induction of trypsin inhibitor activity and other defense-related compounds in Manchurian ash in the study by Whitehill et al. (2014) suggests that its high resistance to EAB may be based primarily on constitutive expression of its defenses, results from induced responses that were not examined, or that induction occurs on a faster time scale than measured in the study. For example, EAB larval feeding induced accumulation of the lignan pinoresinol A (Chakraborty et al., 2014), which belongs to a class of compounds that had previously been implicated in EAB resistance (Eyles et al., 2007; Cipollini et al., 2011; Whitehill et al., 2012). However, this occurred to a faster or greater extent in Manchurian ash than in black ash (Chakraborty et al., 2014). Furthermore, protein kinase domains are among the most abundant protein domains in the Fraxinus transcriptome (Bai et al., 2011), and two calcium dependent protein kinases, an ethylene response factor, and a lipoxygenase, were more highly expressed, constitutively, in Manchurian ash than black and green ash (Bai et al., 2011). Protein kinases participate in recognition of invading agents and signal transduction and thus play a key role in induced resistance (Romeis, 2001; Tena et al., 2011), while ethylene response factors may participate in the insect-mediated ethylene burst (von Dahl & Baldwin, 2007). Finally, lipoxygenases are potentially involved in the early phases of plant defense responses (Hwang & Hwang, 2010). These findings suggest that inducible as well as constitutive defenses may contribute to resistance of Manchurian ash to EAB. Time course studies involving a wider array of potential defensive compounds would be useful for testing this hypothesis.
IV. Nutritional quality and primary metabolites
The nutritional quality of plants is also central to understanding interactions between insects and their hosts (Berenbaum, 1995). Whether or not low nutritive quality of plant tissues results from natural selection remains an interesting evolutionary question (Haukioja et al., 1991; Augner, 1995), but it clearly has a detrimental effect on insect performance (Mattson, 1980). In some cases, however, low nutritive quality can be overcome by compensatory feeding (Simpson & Simpson, 1990).
A multivariate analysis of the constitutive nutritional characteristics of ash bark (i.e. water content, free amino acids, total protein, soluble sugars and starch, percent carbon and nitrogen, and macro- and micronutrients) revealed a clear separation of green and white ash from Manchurian and black ash (Hill et al., 2012). Manchurian ash was also distinguished from black ash on the basis of water content and concentrations of histidine, lysine, methionine, ornithine, proline, sarcosine, tyramine, tyrosol, total soluble sugars, and the mineral nutrients aluminum, iron, potassium, sodium, and vanadium (Fig. 6). However, only concentrations of proline, tyramine, and tyrosol were higher in Manchurian ash than in black ash, and of these, only proline and tyramine were higher in Manchurian than in green and white ash (Hill et al., 2012).

The amino acid proline is a plant osmoregulator (Delauney & Verma, 1993; Hare & Cress, 1997) with the ability to scavenge free radicals, and hence stabilize subcellular structures, protect enzymes (Hare & Cress, 1997), and shield plant tissues from photooxidation (Saradhi et al., 1995). For these reasons, Hill et al. (2012) proposed that proline may contribute to EAB resistance of Manchurian ash by alleviating stressors that can predispose it to larval colonization.
Manchurian ash also contained c. 9-, 30-, and 150-fold higher constitutive levels of the non-amino acid monoamine tyramine than black, green, and white ash, respectively (Hill et al., 2012). Tyramine hydroxycinnamates are important structural components of lignin (Boerjan et al., 2003) and suberin (Hagel & Facchini, 2005; Graça, 2010), which in turn are key components of wound-induced periderm (Ginzberg, 2008). Hence, Hill et al. (2012) suggested that tyramine may be involved in formation of wound periderm, which has been hypothesized to play an important role in tree resistance to wood-borers and bark beetles (Matson & Hain, 1985; Dunn et al., 1990; Muilenburg et al., 2013), and thus may contribute to tolerance of ash to EAB by facilitating reestablishment of integrity of injured vascular tissue.
Tyramine may also be directly toxic to EAB, as it is a biosynthetic precursor of the invertebrate neurotransmitter octopamine (Hiripi et al., 1994), and itself is a neuroactive compound that can have negative physiological and behavioral effects on insects (Lange, 2009). Tyramine can also interact with the Malpighian tubules of insects as a paracrine/autocrine hormone that induces a dose-dependent increase in Cl− conductance and stimulates diuresis (Blumenthal, 2003), perhaps promoting insect desiccation. It is noteworthy that tyramine is also a biosynthetic precursor of the phenylethanoid glycoside verbascoside (Saimaru & Orihara, 2010), which is toxic to EAB (Whitehill et al., 2014). Interestingly, Rigsby et al. (2015) assayed activity of monoamine oxidase, an enzyme capable of metabolizing tyramine, using tyramine as a substrate, and found that EAB larvae that had fed on Manchurian ash had higher activities than larvae that had fed on white or green ash.
V. Conclusions and future directions
Studies reviewed here have shown that: (1) Manchurian ash was less preferred as a host for oviposition than North American species (Rigsby et al., 2014); (2) development of EAB larvae was slower in Manchurian ash relative susceptible species such as black ash (Chakraborty et al., 2014), which confirms the importance of antibiosis; and (3) drought stress increased susceptibility of Manchurian ash to EAB (Chakraborty et al., 2014) (Fig. 7), which is consistent with the hypothesis that the EAB is a secondary colonizer of coevolved hosts.

Compared with susceptible North American species, Manchurian ash bark was characterized by higher constitutive concentrations of lignans (Eyles et al., 2007; Cipollini et al., 2011; Whitehill et al., 2012), a coumarin derivative (Whitehill et al., 2012), the amino acid proline, and the monoamine tyramine (Hill et al., 2012). Manchurian ash was also characterized by higher expression of four putative defensive proteins (Whitehill et al., 2011), and a faster browning reaction (Cipollini et al., 2011; but see Whitehill et al., 2014) (Table 2; Fig. 8). Multiple qualitative and quantitative phytochemical differences between resistant and susceptible species highlight the importance of investigating the full metabolome of ash, rather than individual compounds, as is also true for many other systems (e.g. Herms & Mattson, 1992; Ossipov et al., 2001).
Taxonomic section1 | Fraxinus spp. | Geographic distribution | Induced resistance | |||||||
---|---|---|---|---|---|---|---|---|---|---|
Constitutive resistance | MeJA2 | Larval feeding3 | ||||||||
Low water availability | Normal water availability | |||||||||
Phenotype | Defense trait | Phenotype | Defense trait | Phenotype | Defense trait | Phenotype | Defense trait | |||
Dipetalae | quadrangulata (blue) | North America | I4,5 | Esculin6 | na | na | na | na | na | na |
Melioides | americana (white) | North America | S2,5,7,8,9,10 | None | R | TIs*Lignin*Verbascoside** | na | na | na | na |
pennsylvanica (green) | North America | S2,5,7,8,9,10 | None | R | TIs*Lignin*Verbascoside** | na | na | na | na | |
Fraxinus | nigra (black) | North America | S2,10 | None | R | TIs* | S | None;No effect of drought on phenolics | S | None |
excelsior (European) | Europe | S11 | None | na | na | na | na | na | na | |
mandshurica (Manchurian) | Asia | R2,9 | Proteins12,13Lignans3,6,12,13,14Browning reaction12Coumarins6 Proline15Tyramine15 | R (No additional effects of MeJA) | No effects | S | None; No effect of drought on phenolics | R | Lignans |
- Phenotype: R, resistant; S, susceptible; I, intermediate. na, data not available. Tis, Trypsin inhibitors. *A proxy of the compound was tested directly against EAB larvae in vitro and had antibiotic effects. **Compound was tested directly against EAB larvae in vitro and had antibiotic effects.
- 1Wallander (2008); 2Whitehill et al. (2014); 3Chakraborty et al. (2014); 4Anulewicz et al. (2008); 5Tanis & McCullough (2012); 6Whitehill et al. (2012); 7Cappaert et al. (2005); 8Poland & McCullough (2006); 9Rebek et al. (2008); 10Klooster et al. (2014); 11Orlova-Bienkowskaja (2014); 12Cipollini et al. (2011); 13Whitehill et al. (2011); 14Eyles et al. (2007); 15Hill et al. (2012).

The phenolic profiles of susceptible black ash and resistant Manchurian ash are very similar (Whitehill et al., 2012), which reflects their close phylogenetic relationship. Furthermore, contrasting variations in resistance phenotypes in these two species, due to either drought stress (Chakraborty et al., 2014) or exogenous MeJA application (Whitehill et al., 2014), were not associated with significant variation in phenolic profiles (Table 2; Figs 7, 8). This suggests that phenolic composition may not contribute to the observed variation in resistance of these two species. Alternatively, Manchurian ash may have a faster browning reaction than black ash that results in more rapid and comprehensive oxidation of phenolic compounds. The browning reactions of black and Manchurian ash have yet to be compared directly, but Cipollini et al. (2011) did find that Manchurian ash had a faster browning reaction than susceptible green and white ash.
Induced resistance of ash to EAB has also been investigated. Application of MeJA-induced resistance of generally susceptible North American white, green, and black ash to EAB, and this was associated with accumulation of verbascoside and lignin in green and white ash, and higher trypsin inhibitor activity in all three species (Whitehill et al., 2014) (Table 2, Fig. 8). These findings suggest that green, white, and black ash possess potential for resistance that is not expressed under natural conditions, possibly because susceptible species fail to recognize cues associated with EAB colonization or respond rapidly enough to its elicitors. Evidence suggests that Manchurian ash resistance to EAB may rely primarily on constitutive mechanisms (Whitehill et al., 2014). However, EAB larval feeding induced accumulation of pinoresinol A to a greater extent in Manchurian ash than in black ash (Chakraborty et al., 2014), which suggests that inducible responses may also be important.
Many knowledge gaps remain to be addressed. For example, it is noteworthy that most interspecific comparisons of ash resistance to EAB have examined only one genotype per species. The few studies that have compared clonal cultivars with open-pollinated populations of green (Cipollini et al., 2011; Whitehill et al., 2012, 2014) and Manchurian ash (Herms, 2015) have found no differences in resistance or metabolic profiles. However, wider screening of genotypes is required. Additionally, while 788 transcripts involved in seven biosynthetic pathways for alkaloids were found in the transcriptome of Fraxinus (Bai et al., 2011), no study has yet identified alkaloids from ash bark, and little is known about the alkaloid chemistry of ash in general (Wall et al., 1959; Barbosa & Krischik, 1987). And in spite of their hypothesized role in tree resistance to wood-borers and bark beetles (Matson & Hain, 1985; Dunn et al., 1990; Muilenburg et al., 2013), little is known about physical defenses and induced wound-periderm formation in ash.
A very intriguing finding, which could aid in understanding the interaction between North American ash and the EAB, is that a very small proportion (<1%) of green and white ash trees has survived in heavily EAB-attacked stands (Koch et al., 2010; Knight et al., 2011). These ‘lingering ash’ may have higher resistance, provide material to study resistance traits in North American ash, and serve as a source of resistance genes in native ash populations that could be used in programs to breed or select EAB resistant ash. Such efforts would complement ongoing attempts to hybridize susceptible North American species with resistant Asian species (Koch et al., 2012).
Finally, the finding that white fringetree is a larval host (Cipollini, 2015) suggests that studies incorporating this species will be particularly useful for isolating volatile attractants of adult oviposition and chemical determinants of larval performance.
Confirmation of putative defense mechanisms using bioassays with artificial diets developed for EAB (Keena et al., 2015) have been a useful contribution (Whitehill et al., 2014), but further verification is still required. The development of genomic and transcriptomic resources would also contribute to increased understanding of resistance mechanisms (Parent et al., 2015). Currently, only a few published studies have explored the constitutive transcriptome of ash (Bai et al., 2011) and EAB (Mittapalli et al., 2010; Rajarapu, 2013). Targeted experiments would generate fundamental insights into the interaction at the molecular level. For example, genomic or more extensive transcriptomic profiling of the insect, as has been done for Asian longhorned beetle (Scully et al., 2013, 2014), would help elucidate EAB counter adaptations to host resistance mechanisms. Comparing transcriptomes of resistant and susceptible ash species after larval induction might lead to identification of relevant host genes and metabolites responsible for resistance. Targeted transcriptomic studies on the host are currently in progress (Showalter et al., 2015), but results are not yet available. These studies are made possible by the existence of resources that include annotated ash reference genomes and transcriptomes. For example, the genome of European ash, a close relative of resistant Manchurian ash, has recently been sequenced by the British Ash Tree Genome Project – http://www.ashgenome.org/ – (Sollars et al., 2014) and the genome of 35 other Fraxinus species is underway (Sollars et al., 2015). Similarly, the Hardwood Genomic Project (http://www.hardwoodgenomics.org/) is using assemblies of expressed sequence tags (ESTs) of green ash under a variety of conditions to build the whole transcriptome of the species. Sequencing the genome or transcriptome of ash species would also strongly facilitate the development of markers for marker-assisted selection (Yang et al., 2015). Once genes of interest are identified, it may be possible to verify their role in a given resistance mechanism by genetically engineering them into susceptible ash, which will be facilitated by the recent development of transformation tools (Du & Pijut, 2009), or, alternatively, their disruption in resistant ash. The feasibility of conducting functional studies has also been boosted by the recent report that EAB possesses functional RNAi systems, which could be exploited to knock down specific detoxifying enzymes in order to investigate their role in EAB counterdefense (Zhao et al., 2015).
Acknowledgements
We thank D. Showalter and C. Rigsby for providing valuable comments and insights on the manuscript. We also thank S. Chakraborty and A. Hill for providing data from previously published studies that was used to compose Figs 4 and 6, respectively. This review, including genesis, development, and writing was funded by the USDA APHIS Accelerated Emerald Ash Borer Research Program, USDA Forest Service Northern Research Station, USDA ARS, USDA AFRI, Horticultural Research Institute, Tree Research and Education Fund, and by state and federal funds appropriated to the Ohio Agricultural Research and Development Center, The Ohio State University, and Wright State University.