(C) PLOS One
This story was originally published by PLOS One and is unaltered.
. . . . . . . . . .



Conifer-killing bark beetles locate fungal symbionts by detecting volatile fungal metabolites of host tree resin monoterpenes [1]

['Dineshkumar Kandasamy', 'Department Of Biochemistry', 'Max Planck Institute For Chemical Ecology', 'Jena', 'Max Planck Center For Next Generation Insect Chemical Ecology', 'Ngice', 'Department Of Biology', 'Lund University', 'Lund', 'Rashaduz Zaman']

Date: 2023-02

Outbreaks of the Eurasian spruce bark beetle (Ips typographus) have decimated millions of hectares of conifer forests in Europe in recent years. The ability of these 4.0 to 5.5 mm long insects to kill mature trees over a short period has been sometimes ascribed to two main factors: (1) mass attacks on the host tree to overcome tree defenses and (2) the presence of fungal symbionts that support successful beetle development in the tree. While the role of pheromones in coordinating mass attacks has been well studied, the role of chemical communication in maintaining the fungal symbiosis is poorly understood. Previous evidence indicates that I. typographus can distinguish fungal symbionts of the genera Grosmannia, Endoconidiophora, and Ophiostoma by their de novo synthesized volatile compounds. Here, we hypothesize that the fungal symbionts of this bark beetle species metabolize spruce resin monoterpenes of the beetle’s host tree, Norway spruce (Picea abies), and that the volatile products are used as cues by beetles for locating breeding sites with beneficial symbionts. We show that Grosmannia penicillata and other fungal symbionts alter the profile of spruce bark volatiles by converting the major monoterpenes into an attractive blend of oxygenated derivatives. Bornyl acetate was metabolized to camphor, and α- and β-pinene to trans-4-thujanol and other oxygenated products. Electrophysiological measurements showed that I. typographus possesses dedicated olfactory sensory neurons for oxygenated metabolites. Both camphor and trans-4-thujanol attracted beetles at specific doses in walking olfactometer experiments, and the presence of symbiotic fungi enhanced attraction of females to pheromones. Another co-occurring nonbeneficial fungus (Trichoderma sp.) also produced oxygenated monoterpenes, but these were not attractive to I. typographus. Finally, we show that colonization of fungal symbionts on spruce bark diet stimulated beetles to make tunnels into the diet. Collectively, our study suggests that the blends of oxygenated metabolites of conifer monoterpenes produced by fungal symbionts are used by walking bark beetles as attractive or repellent cues to locate breeding or feeding sites containing beneficial microbial symbionts. The oxygenated metabolites may aid beetles in assessing the presence of the fungus, the defense status of the host tree and the density of conspecifics at potential feeding and breeding sites.

Funding: Financial support was provided by the Max Planck society to DK, RZ, YN, HH, AH and JG. MNA acknowledges funding from the Swedish Research Council FORMAS (grants #217-2014-689 and #2018-01444) the Crafoord foundation, the Royal Physiographic Society in Lund, and the Foundation in Memory of Oscar and Lily Lamm. The funders had no role in study design, data collection and analysis, decision to publish and writing.

In this study, we investigated the volatile compounds emitted when fungal symbionts of I. typographus infect the excised bark of their Norway spruce (Picea abies) host trees. We show that these fungi alter the volatile monoterpene composition of spruce bark and demonstrate, using single sensillum recordings (SSRs), that adult I. typographus can perceive the fungal-produced monoterpenes and are attracted to these compounds in behavioral bioassays. Our results imply that bark beetles respond to fungal symbiont biotransformation products of host tree metabolites and employ them to identify suitable sites for feeding and breeding.

Conifer oleoresins are a formidable defense against insects and pathogens, as they can poison and physically entrap invaders [ 23 – 26 ]. However, the volatile fraction of the resin, especially the monoterpenes, also plays a central role in the colonization of host trees by bark beetles [ 24 , 27 , 28 ]. After locating a suitable tree, pioneer male I. typographus oxidize the dominant host monoterpene α-pinene to cis-verbenol, which is used as an aggregation pheromone in combination with the de novo produced 2-methyl-3-buten-2-ol to attract conspecifics for a mass attack [ 29 – 31 ]. In addition to bark beetle pheromones, several other oxygenated monoterpenes such as terpinen-4-ol, camphor, trans-4-thujanol, and borneol have also been detected at the entrance holes of I. typographus galleries [ 32 – 35 ]. Additionally, the phloem colonized by ophiostomatoid fungi around these galleries also produces large amounts of oxygenated monoterpenes compared to galleries without evident fungal growth [ 33 ]. Only a few studies have investigated the changes in the oleoresin composition of conifers within the necrotic lesions caused by fungi [ 36 , 37 ]. While these studies showed that fungal infection changed the composition of oleoresins, the presence of oxygenated monoterpenes and their ecological roles were not investigated. In our previous work, we showed that I. typographus utilizes de novo synthesized fungal volatiles to maintain its association with specific beneficial symbionts and also to avoid saprophytes [ 38 ]. However, it is unknown which volatiles are produced by these fungi when they colonize their native substrate i.e., the phloem and sapwood of the tree.

Microbial symbioses are especially characteristic of wood-boring insects such as bark and ambrosia beetles. Bark beetles have captured much attention recently because of their large-scale outbreaks in many parts of the world. In Europe, for example, the Eurasian spruce bark beetle (Ips typographus) has killed millions of hectares of spruce stands as a result of major population bursts, likely in response to global warming and management practices that increase forest vulnerability to epidemic outbreaks [ 8 – 12 ]. This bark beetle feeds and reproduces in the phloem tissue of trees, which contains high levels of terpene and phenolic defense chemicals [ 8 , 13 ]. This insect overcomes its unfavorable environment by mass attacks and by introducing a suite of microbes into the host tree, including ectosymbiotic ophiostomatoid fungi, primarily Grosmannia penicillata, Leptographium europhioides, Endoconidiophora polonica, and Ophiostoma bicolor that cause blue staining of infected wood [ 14 – 18 ]. The ecological relationships between these fungal symbionts and I. typographus have not been extensively studied. Although they grow outside the insect, as necrotrophic fungi, they may serve as mutualists by exhausting host tree defenses, metabolizing host defense compounds, and providing nutritional benefits to larvae and adults [ 19 – 21 ]. However, some ophiostomatoid fungal associates of bark beetles may also act as commensals or antagonists to beetle brood development [ 22 ].

Many interactions between insects and their host plants are known to be mediated by volatile organic compounds. In contrast, volatile signals between herbivorous insects and their symbiotic microbes have been less studied, aside from a few well-known examples including ambrosia beetles, termites, and the vinegar fly Drosophila melanogaster [ 1 – 4 ]. Yet, such signals could be as critical for insect fitness as their response to host plant cues. In some insect–microbe symbioses, microbes transform host plant metabolites creating volatile signals that are used by insects for food or brood site selection [ 3 – 6 ]. For example, yeasts vectored by D. melanogaster metabolize dietary phenolic antioxidants and release volatile phenolics that attract both larvae and adults to feed on antioxidant-rich foods [ 7 ]. Nevertheless, there is still little information about how microbial transformation of host plant chemicals influences insect–microbe symbioses, and whether the resulting metabolites represent honest signals of partner benefits.

(A) Schematic drawing of a Petri dish arena used for the no-choice tunneling assay. The dish was filled with monoterpene-enriched spruce bark agar and inoculated with G. penicillata (Gp) (top). Example of the tunneling pattern of an adult beetle within the fungus-colonized diet as pictured from the bottom side of Petri dish (bottom). (B–D) Tunnel lengths (cm) made by adult beetles after 48 h in diet containing G. penicillata only, monoterpenes only, G. penicillata and monoterpenes, or controls with neither G. penicillata nor monoterpenes. Error bars represent SEM (n = 30 (15 ♂, 15 ♀) for each trial). Monoterpenes: (−)-β-pinene (B) , (−)-α-pinene (C) , (−)-bornyl acetate (D) . Different lowercase letters indicate significant differences between treatments (ANOVA, Tukey’s test, P < 0.05). The data underlying this Figure can be found at https://doi.org/10.6084/m9.figshare.21692156.v1 .

The growth of G. penicillata on monoterpene-enriched media resulted in significantly longer beetle tunnels than in the other treatments ( Fig 8B–8D ) ((−)-β-pinene, F (3,99) = 4.95, p = 0.003; (−)-α-pinene, F (3,92) = 14.8, p < 0.001; (−)-bornyl acetate, F (3,106) = 6.6, p < 0.001, ANOVA, Tukey’s test)). However, there were no other significant interactions between monoterpenes and G. penicillata affecting tunneling, i.e., no other differences in tunnel lengths when comparing fungus alone versus fungus plus monoterpenes except for the treatment with (−)-β-pinene ( Fig 8B–8D , Tukey’s test). The sex of the beetle and the interaction of sex with other treatments also had no effect on the tunnel length.

The presence of a symbiotic fungus increased the tunneling of adult I. typographus beetles in medium after 48 h. Multiple logistic regression analysis revealed that successful tunneling odds in (−)-bornyl acetate-amended medium were significantly influenced by the fungus when the beetle sex and the monoterpene treatment remained constant ( Table 1 ) (β = 4.98, χ 2 = 20.99, p < 0.001). The presence of the fungus increased the tunneling probability by 99% compared to in the absence of the fungus. Additionally, males had significantly lower tunneling success compared to females ( Table 1 ) (β = −1.72, χ 2 = 8.78, p = 0.003) with the tunneling probability for males being 15.2% lower than for females. In (−)-β-pinene-amended medium, only the presence of the fungus significantly influenced the tunneling probability of beetles ( Table 1 ) (β = 2.65, χ 2 = 20.45, p < 0.001), with an increase of 93% compared to in the absence of the fungus. Similarly, in (−)-α-pinene-amended medium, only the presence of the fungus significantly influenced the tunneling probability of beetles ( Table 1 ) (β = 2.02, χ 2 = 9.39, p = 0.002), with an increase of 88% compared to the absence of the fungus. Addition of the three monoterpenes without fungus did not have any effect on the tunneling behavior of adult beetles.

(A) Adult females chose traps containing 2-methyl-3-buten-2-ol and cis-verbenol at 10 −2 concentration diluted in mineral oil over control traps containing mineral oil. Females strongly preferred traps containing a binary pheromone blend (cis-verbenol: 2-methyl-3-buten-2-ol, 50:1) diluted in mineral oil over the mineral oil control (bottom). Adult males were unresponsive to these concentrations of individual pheromones and their blend. (B) Adult females preferred pheromone blend in the presence of G. penicillata volatiles over pheromone blend without fungus. (A, B) . Deviation of response indices against zero was tested using Wilcoxon’s test (n = 28 for each experiment). The data underlying this Figure can be found at https://doi.org/10.6084/m9.figshare.21692156.v1 .

Female beetles were significantly more attracted towards the individual aggregation pheromone components cis-verbenol and 2-methyl-3-buten-2-ol (cis-verbenol, z = 2.98, p = 0.003; 2-methyl-3-buten-2-ol, z = 2, p = 0.046, Wilcoxon’s test), and towards a mixture of the two pheromone components (z = 5.19, p < 0.001, Wilcoxon’s test) compared to the mineral oil control ( Fig 7A ). By contrast, adult males did not make a choice between these options in accordance with previous studies [ 43 , 44 ]. However, when beetles had to choose between the pheromone mixture with or without G. penicillata volatiles, females preferred the pheromone mixture together with G. penicillata volatiles ( Fig 7B ) (z = 3.41, p = 0.001, Wilcoxon’s test).

Addition of another host tree monoterpene, (−)-bornyl acetate, to fungal growth medium at 0.05 mg g −1 (in the range of natural concentrations in P. abies bark) and 0.5 mg g −1 resulted in strong attraction of I. typographus adults towards G. penicillata when tested against a fungus-free control after 4-d incubation ( Fig 6C , left panel) (0.05 mg g −1 , z = 3.31, p = 0.001; 0.5 mg g −1 , z = 3.21, p = 0.001, Wilcoxon’s test). This attraction was correlated with the presence of fungal biotransformation products from (−)-bornyl acetate ( Fig 6C , right panel). The major biotransformation product that formed in this period ( Fig 6C ) was camphor, which was significantly more attractive to adult beetles at a 100-μg dose than the mineral oil control ( Fig 6D ) (z = 2.58, p = 0.01, Wilcoxon’s test). However, camphor had no effect at larger or smaller doses. Adult beetles preferred G. penicillata grown on unenriched medium against G. penicillata grown on medium enriched with a high amount of (−)-bornyl acetate (0.5 mg g −1 ) ( Fig 6F ) (z = 2.12, p = 0.03, Wilcoxon’s test). By contrast, in the absence of fungus, beetles did not discriminate between diet enriched with monoterpenes and diet without monoterpenes ( S11 Fig ). Thus, I. typographus is attracted to the major oxygenated fungal metabolite of (−)-bornyl acetate but is not attracted and even repelled by (−)-bornyl acetate itself.

(A) Adult beetles preferred spruce bark agar enriched with 0.1 and 0.5 mg/g of (−)-β-pinene and inoculated with G. penicillata over spruce bark agar enriched with 0.1 and 0.5 mg/g (−)-β-pinene alone, but without fungus (left). GC–MS traces depict the headspace volatiles of (−)-β-pinene-enriched agar left for 4 d with and without G. penicillata (right) and show several oxygenated monoterpenes produced by the fungus from (−)-β-pinene. Numbers refer to the identities of compounds. (B) Adult beetles chose trans-4-thujanol (left) at a 100-μg dose diluted in mineral oil, when tested against a mineral oil control (Bonferroni adjusted significance level, α = 0.007). Adult beetles showed indifferent responses to (+)-isopinocamphone (middle), and terpinen-4-ol (right), applied in various doses in mineral oil. (C) Adult beetles preferred spruce bark agar enriched with various amounts of (−)-bornyl acetate and inoculated with G. penicillata over spruce bark agar enriched with (−)-bornyl acetate but without fungus (left). GC–MS traces depict the headspace volatiles of 0.5 mg/g (−)-bornyl acetate-enriched agar left for 4 d with and without G. penicillata (right) and show several oxygenated monoterpenes produced by the fungus from (−)-bornyl acetate. Numbers refer to the identities of compounds. (D) Adult beetles preferred (±)-camphor at a 100-μg dose against a mineral oil control, but not at other doses (Bonferroni adjusted α = 0.012). (E) Adult beetles did not discriminate between G. penicillata on agar with three different concentrations of (−)-β-pinene and G. penicillata without (−)-β-pinene. (F) Adult beetles chose G. penicillata on agar without any administrated (−)-bornyl acetate vs. G. penicillata on agar with 0.5 mg/g (−)-bornyl acetate. (A–F) Deviation of response indices against zero was tested using Wilcoxon’s test. n = 25 or 50 for each trial. The data underlying this Figure can be found at https://doi.org/10.6084/m9.figshare.21692156.v1 .

G. penicillata growing on SBA with added 0.1 mg g −1 or 0.5 mg g −1 (−)-β-pinene (within the range of natural concentrations in P. abies bark [ 42 , 43 ] was much more attractive to adult I. typographus than a fungus-free control medium supplemented with (−)-β-pinene ( Fig 6A , left) (0.1 mg g −1 , z = 2.22, p = 0.02; 0.5 mg g −1 , z = 2.54, p = 0.01, Wilcoxon’s test). This attraction was correlated with the presence of oxygenated monoterpenes produced by the fungus from (−)-β-pinene ( Fig 6A , right). However, supplementation of agar with 1 mg g −1 (−)-β-pinene did not lead to a significant choice between fungus and fungus-free (−)-β-pinene-containing medium ( Fig 6A , left). To further understand this concentration-dependent response, individual (−)-β-pinene biotransformation products that had been shown to be electrophysiologically active (see last section) were used in trap bioassays at various doses against a mineral oil control. Adult beetles were significantly attracted to 100 μg trans-4-thujanol ( Fig 6B , left panel) (z = 2.78, p = 0.005, Wilcoxon’s test), but other concentrations of trans-4-thujanol and all concentrations of (+)-isopinocamphone and terpinen-4-ol tested had no significant effect. Adult beetles did not discriminate between G. penicillata grown on (−)-β-pinene enriched medium and G. penicillata grown on nonenriched medium ( Fig 6E ). Based on these results, we conclude that I. typographus shows concentration-specific responses to some oxygenated fungal metabolites of (−)-β-pinene, but not to (−)-β-pinene itself ( S11 Fig ).

A few previously characterized OSN classes for host tree monoterpenes, including the classes with primary responses to α-pinene, p-cymene, and Δ3-carene, respectively [ 41 ], showed varying secondary responses to some of the fungal-derived compounds tested here for the first time. For example, the α-pinene OSN class responded also to (+)-isopinocamphone, (−)-isopinocamphone, and (±)-pinocarvone and weakly to (±)-camphor, (−)-myrtenol, trans-pinocarveol, carvone, borneol, and (−)-fenchone ( S10A Fig ). The p-cymene OSN class showed intermediate responses to (+)-trans-4-thujanol and carvone ( S10B Fig ). Although the Δ3-carene OSN class showed high specificity towards Δ3-carene, ligands such as camphor and (−)-isopinocamphone also elicited weak responses from this neuron class ( S10C Fig ).

(A) Mapping of three classes of olfactory sensory neurons (OSN) selective for oxygenated monoterpenes (isopinocamphone; verbenone; (−)-trans-4-thujanol) on the antenna. Olfactory sensilla housing these OSN classes are distributed in medial (subscript “m”) and lateral (subscript “l”) regions of sensillum areas a and b (but not in area c) of the antenna. (B) Representative spike trains from an isopinocamphone-responsive neuron stimulated with 1 μg of (+)-isopinocamphone (top left) and (±)-camphor (bottom left); a verbenone-responsive neuron stimulated with 1 μg (−)-verbenone (top right); a (+)-trans-4-thujanol-responsive neuron stimulated with 1 μg (+)-trans-4-thujanol (bottom right). Black horizontal bars indicate the 0.5-s odor puffs. (C) Response spectra of OSN classes responding predominantly to oxygenated monoterpenes produced by fungi at the 10-μg screening dose. The average number of spikes/second was recorded from the isopinocamphone-tuned OSN class (left) (n = 6 except (+) and (−)-pinocamphone (n = 3)), trans-4-thujanol-tuned OSN class (middle) (n = 5), and verbenone and isophorone-tuned OSN class (right) (n = 4) after neurons were stimulated with a panel of 97 odors. Error bars represent SEM. (D) Dose–response curves of the OSNs stimulated with their most active ligands: isopinocamphone-tuned OSN class (left) ((n = 9) except for (+) and (−)-pinocamphone (n = 3)), trans-4-thujanol-tuned OSN class (middle) (n = 3), and the OSN class tuned to isophorone and verbenone (right) (n = 3). Error bars represent SEM. The data underlying this Figure can be found at https://doi.org/10.6084/m9.figshare.21692156.v1 .

OSN classes tuned to fungal-produced oxygenated monoterpenes were identified in both the A m and B m regions on the antennae ( Fig 5A ). One of these OSN classes responded most strongly to (+)-isopinocamphone, and this class was highly specific for oxygenated monoterpenes, especially ketones ( Fig 5C , left panel). Apart from (+)-isopinocamphone, relatively strong responses were also elicited by (+)-pinocamphone, (−)-isopinocamphone, (±)-pinocarvone, (±)-camphor, and (−)-pinocamphone ( Fig 5C , left panel). Dose–response tests showed that this OSN class was the most sensitive to (+)-isopinocamphone of all the compounds tested with responses evident at a dose of 100 pg. The responses to (+)-pinocamphone, (−)-isopinocamphone, (±)-pinocarvone, and (±)-camphor all appeared between 1 ng and 10 ng doses ( Fig 5D , left panel). Another OSN class with specific responses to fungal-derived compounds responded most strongly to (+)-trans-4-thujanol and weakly to (±)-3-octanol, (±)-1-octen-3-ol, (+)- and (−)-terpinen-4-ol, and (+)- and (−)-α-terpineol ( Fig 5C , middle panel). This OSN showed a 1,000-fold lower response threshold to (+)-trans-4-thujanol compared to the next best ligands, the C 8 alcohols ( Fig 5D , middle panel). Finally, an OSN class responding strongly to verbenone, α-isophorone and β-isophorone, followed by weaker responses to (−)- and (+)-trans-verbenol, pinocarvone, and (−)-cis-verbenol ( Fig 5C , right panel) was also found. Dose–response tests revealed that this neuron class responded the strongest to α-isophorone across most tested doses, followed by slightly weaker and similarly strong responses to both verbenone and β-isophorone ( Fig 5D , right panel).

To test if I. typographus antennal olfactory sensilla contain OSNs that detect the biotransformation products of monoterpenes, we challenged 231 olfactory sensilla with a test panel comprising 92 ecologically relevant compounds diluted in paraffin oil (1 μg μl −1 ) using single cell recordings ( S1 Table ). Only 23 (approximately 10%) of the sensilla housed neurons that did not respond to any of the compounds from the odor panel, although their OSNs showed spontaneous firing. We obtained odor-evoked responses with strong excitation (>120 Hz) from 198 OSNs and weak excitation (<50 Hz) from 10 OSNs, allowing the grouping of these neurons into different classes based on their response profiles. From initial screening experiments at a 10-μg dose on filter paper (to determine the maximum receptive range of OSNs), we identified 20 classes of OSNs. Three OSN classes responded primarily to fungal-produced oxygenated monoterpenes. We also identified neurons belonging to previously described OSN classes tuned to pheromones, host tree volatiles, and nonhost odorants [ 41 ] that are not further considered here.

Analysis of the microbes present in each treatment revealed that galleries constructed by fungus-free beetles were indeed free from symbiotic ophiostomatoid fungi and yeasts but contained many saprophytes and (or) endophytes and bacteria ( S10 Table ). Microbes present in the galleries of untreated and G. penicillata-reinoculated beetles were similar to those found on beetles that were used to infest the bark ( S9 and S10 Tables).

Next, we quantified the amount of oxygenated monoterpenes present in the bark galleries constructed by untreated, fungus-free, and G. penicillata-reinoculated beetles separately using GC–MS analyses of gallery extracts. The concentration of (−)-camphor was significantly higher in the galleries of reinoculated beetles than in bark samples without beetles at all (F (3,23) = 15.6, p < 0.001, ANOVA), while the concentration of terpinen-4-ol was significantly higher in untreated beetles than in both fungus-free beetles or in bark samples without beetles (H (3) = 18.6, p < 0.001, Kruskal–Wallis) ( Fig 4B ). On the other hand, the concentrations of these two oxygenated monoterpenes were similar when comparing galleries of G. penicillata-reinoculated beetles to those of untreated beetles. For (+)-isopinocamphone, the concentration in galleries constructed by untreated beetles was significantly higher compared to bark samples without beetle galleries, but similar to that in galleries constructed by fungus-free and G. penicillata-reinoculated beetles (F (3,23) = 5.7, p = 0.005, ANOVA) ( Fig 4B ). On the other hand, the concentration of (−)-borneol in fungus-free galleries was significantly higher compared to that in bark samples without galleries and was similar to that in the galleries of G. penicillata-reinoculated and untreated beetles (F (3,23) = 5.3, p = 0.006, ANOVA) ( Fig 4B ). There were no significant differences in the concentrations of pinocamphone, α-terpineol, and (−)-verbenone among treatments.

To determine the primary source of oxygenated monoterpenes in beetle galleries, unsterilized pupae were reared into fungus-free mature adults using a diet fortified with the fungistatic agent sodium benzoate ( S1 Method ). Sterility tests and scanning electron microscopy of beetle body parts revealed that beetles reared on this diet were free from fungi and the elytral pits were devoid of fungal spores and yeasts ( S15C Fig and S9 Table ), although bacteria were abundant. In comparison, elytral pits of beetles reared on diet without sodium benzoate contained abundant spore masses, with structural similarity to those of O. bicolor and yeasts ( S15B Fig and S9 Table ) [ 40 ]. Reinoculation of fungus-free beetles with G. penicillata resulted in acquisition of G. penicillata spores in the elytral pits ( S15D Fig ).

The volatile blend of the co-occurring, but nonbeneficial fungus, Trichoderma sp., produced from a synthetic spruce monoterpene mixture ( S8 Table ) was distinct from that of G. penicillata ( S9 Fig ). Trichoderma sp. produced mainly the oxygenated monoterpenes, (−)-borneol and (−)-verbenone, while G. penicillata produced mainly (−)-camphor and terpinen-4-ol. These results collectively show that symbiotic fungi can alter the volatile profile of spruce bark by increasing the emission of oxygenated monoterpenes, and the emission profile of symbiotic fungi is distinct from that of a co-occurring fungal saprophyte.

The monoterpene metabolites of G. penicillata grown on either (−)-α-pinene- or (−)-β-pinene-enriched agar were similar, with the oxygenated monoterpene terpinen-4-ol being the major product from both spruce precursors ( Fig 3B ). Other oxygenated monoterpenes formed by G. penicillata in the presence of (−)-α-pinene and (−)-β-pinene included (+)-isopinocamphone, (+)-trans-4-thujanol and α-terpineol, with (−)-verbenone being detected only from (−)-α-pinene ( S5 and S6 Figs). From (−)-bornyl acetate, the dominant oxygenated monoterpenes of G. penicillata were camphor and endo-borneol ( Fig 3B , bottom panel; S8 Fig ), and their production coincided with the decrease of the precursor ( Fig 3A ). Similar results were obtained for the three other fungal symbionts tested ( S4 – S8 Figs).

(A, B) Metabolism of major spruce monoterpenes by G. penicillata after fungal-infected vs. uninfected PDA was supplemented with 0.5 mg/g of (−)-β-pinene, (−)-α-pinene, and (−)-bornyl acetate. (A) The amounts of starting monoterpenes remaining after 4 days. Error bars represent SEM (n = 5 to 13). Asterisks indicate a significant difference between the PDA-inoculated G. penicillata and the fungus-free control (Welch’s t test) with ns = not significant, **P < 0.01. (B) The most abundant metabolites of administered monoterpenes are depicted with their percentages relative to the total amounts of metabolites detected for each compound (derived from (−)-β-pinene in red; derived from (−)-α-pinene in purple; derived from (−)-bornyl acetate in green). Amounts were determined by headspace collection of volatiles from the agar (n = 3). The data underlying this Figure can be found at https://doi.org/10.6084/m9.figshare.21692156.v1 .

Over a time course of 4, 8, 12, and 18 d post-inoculation with G. penicillata on detached bark plugs ( Fig 2B and S3 and S2 , S3 , and S5 Tables), 79 compounds comprising host tree and fungal volatiles were detected and classified into different groups, namely, aliphatic hydrocarbons (17 compounds), aromatics (2 compounds), monoterpene hydrocarbons (15 compounds), oxygenated monoterpenes (26 compounds), sesquiterpenes (17 compounds), and spiroketals (3 compounds). The proportion of total oxygenated monoterpenes gradually increased to dominate the volatile profile of G. penicillata-infected bark reaching a maximum at 18 d post-inoculation during our sampling period (F (3,14) = 5.51, p = 0.01, ANOVA), while the proportion of total oxygenated monoterpenes was unchanged in mock-inoculated controls ( Fig 2B ). Camphor was the major contributor to the overall increase of oxygenated monoterpenes, with the highest relative abundance after 18 d and a significant difference between time points (F (1,16) = 13.06, p = 0.002, ANOVA) ( S5 Table ). The proportion of monoterpene hydrocarbons gradually decreased over time in both treatments (control, F (3,7) = 11.6, p = 0.004; G. penicillata, F (3,14) = 11.5, p < 0.001, ANOVA). The proportion of total sesquiterpenes also decreased significantly over time in G. penicillata-infected bark (F (3,15) = 4.4, p = 0.02, ANOVA), but not in mock-inoculated control bark. When measuring emission rate (ng mg DW −1 h −1 ), the emission of total monoterpene hydrocarbons in the control and the G. penicillata-treated bark plugs was not significantly different at 4 d post-inoculation ( Fig 2C and S2 Table ). However, there was a dramatic increase in the emission rate of total oxygenated monoterpenes at this time point in spruce bark inoculated with G. penicillata compared to the mock-inoculated control ( Fig 2D ) (9-fold increase, t(3.2) = 7.38, p = 0.004, Welch’s t test). Nineteen oxygenated monoterpenes were identified in the bark inoculated with G. penicillata, and a total of 15 compounds significantly increased in fungus infected bark compared to the control ( Fig 2E ) ( S2 Table ) including camphor (51-fold increase, t(3) = 7.7, p = 0.004, Welch’s t test), endo-borneol (18-fold increase, t(6) = 6.5, p < 0.001), isopinocamphone (3-fold increase, t(6) = 6.8, p < 0.001), verbenone (2-fold increase, t(6) = 4.3, p = 0.005), and bornyl acetate (3-fold increase, t(6) = 3.7, p = 0.01). Measurements conducted on four other I. typographus fungal symbionts also showed differences in the volatile composition of fungus-inoculated versus control bark ( S2 Fig ), with increases in the proportion of oxygenated monoterpenes in bark inoculated with L. europhioides and O. bicolor over time, but not for E. polonica ( S3 Fig and S4 , S6 , and S7 Tables).

(A) Volatile emission pattern differed between spruce bark inoculated with G. penicillata (Gp) and uninfected bark 4 days after inoculation, as depicted in a partial least squares discriminant analysis (PLS-DA ). Complete volatile emission data by compound for G. penicillata and other I. typographus fungal symbionts are given in S2 Table . (B) Changes in volatile emission profiles of spruce bark due to G. penicillata infection over an 18-d time course. Compounds are classified into six groups by their chemical structures (n = 3 to 5). Each circle displays the relative amounts of each class of compounds at a single time point and does not indicate changes in the total amounts of volatiles. Complete volatile emission data by compound and time point for G. penicillata and other symbionts are given in S3 – S6 Tables. ( C , D ) Emission of specific monoterpenes from fresh spruce bark inoculated with G. penicillata at 4 days post-inoculation. Identified compounds were classified into monoterpene hydrocarbons (C) and oxygenated monoterpenes (D). The individual compounds are stacked within a single bar representing the total emission. Significant differences in the total emission levels induced by G. penicillata are denoted by asterisks above the bars (n = 4, Welch’s t test with ns = not significant *P < 0.05, **P < 0.01, ***P < 0.001). The numbers denote the identities of the compounds in the stacked bars. Complete volatile emission data are given in S2 Table . (E) Emission rate of major oxygenated monoterpenes from spruce bark inoculated with G. penicillata 4 d post-inoculation. Asterisks indicate significant differences between the spruce bark-inoculated G. penicillata and the fungus-free control (Welch’s t test). The data underlying this Figure can be found at https://doi.org/10.6084/m9.figshare.21692156.v1 .

The headspace volatiles of P. abies bark inoculated with I. typographus fungal symbionts were distinct from those of fungus-free bark 4 d after inoculation when analyzed using gas chromatography with flame ionization detection (GC-FID) and gas chromatography–mass spectrometry (GC–MS) ( S2 – S7 Tables). For G. penicillata-inoculated bark, PCA analysis revealed that nearly 68% of the variation in the volatile profiles was explained by the first two principal components ( Fig 2A ).

(A) Arena used for trap bioassays to study the behavior of I. typographus to volatiles emitted by a symbiotic fungus. Cups containing agar (with and without fungus) were offered to beetles as odor samples. Holes on each side of the cup allowed the beetles to smell but not touch the agar until they entered the cup and then could no longer escape. (B, C) Adult beetles chose Gp-colonized agar medium over fungus-free medium (n = 25, with two beetles per replicate). (D) Adult beetles chose Gp-inoculated spruce bark agar over Gp-inoculated agar without spruce bark (n = 25). (B–D) Deviation of response indices against zero was tested using Wilcoxon’s test. The data underlying this Figure can be found at https://doi.org/10.6084/m9.figshare.21692156.v1 .

Adult I. typographus beetles were strongly attracted to volatiles emitted by their symbiotic fungus G. penicillata grown on both potato dextrose agar (PDA) or spruce bark agar (SBA) compared to the respective fungus-free agar controls in laboratory trap bioassays ( Fig 1 ) (PDA, z = 3.34, p = 0.001; SBA, z = 2.83, p = 0.005, Wilcoxon’s test). Further, bark beetles showed a stronger attraction towards G. penicillata grown on SBA over the same fungus grown on PDA ( Fig 1D ) (z = 4.28, p < 0.001, Wilcoxon’s test). Volatiles from several other bark beetle fungal symbionts, such as E. polonica and L. europhioides, grown on SBA were also attractive to adult beetles, although not all bark beetle-associated fungi, nor an antagonistic, nonsymbiotic saprophyte (Trichoderma sp.), tested in this way emitted attractive volatile blends ( S1 Fig ) [ 39 ].

Discussion

The successful attack of bark beetles on their host trees is invariably associated with free-living fungal symbionts. These symbionts may detoxify the terpene-rich defensive resin of the host tree, hasten host tree death, provide nutritional benefits, and increase resistance to pathogens [16,17,21]. Here, we document the ability of I. typographus fungal symbionts to metabolize host tree monoterpenes to oxygenated derivatives based on experiments in which we cultured fungi with and without host tree monoterpenes. Several oxygenated monoterpenes have been previously identified as volatiles released from trees that were attacked by I. typographus [32,33,35]. We showed that these are fungal metabolites that become dominant components of the volatile profile making up over half of the total bark volatiles emitted 12 d following infection. Bark beetle OSNs can detect many of these oxygenated monoterpenes, and some of these compounds attract adult beetles to diets inoculated with fungal symbionts. We hypothesize that oxygenated monoterpenes may assist adult beetles in locating suitable breeding and feeding sites [34,45].

Bark beetle symbiotic fungi metabolize host tree monoterpenes to oxidized derivatives Fungi are well known to metabolize monoterpenes as well as many other chemical compounds they encounter in their natural environments [46,47]. In particular, fungi associated with conifer trees are reported to transform different types of resin monoterpenes to oxidized derivatives [48,49]. Previous literature describes this capability for a symbiotic fungus of the North American bark beetle Dendroctonus ponderosae [50]. Here, we showed that fungal symbionts of the European spruce bark beetle I. typographus also oxidize monoterpenes of their host trees by adding monoterpene standards to the culture medium and by growing symbionts on spruce bark with natural levels of monoterpenes. Oxidation of monoterpenes can be considered a detoxification process since these lipophilic compounds have antifungal properties [49,50], and oxidation may facilitate excretion by increasing polarity. However, individual oxidative transformations may not necessarily decrease toxicity but could serve as entry points to reaction cascades that allow monoterpenes to be degraded and used as carbon sources for intermediary metabolism [50,51]. Here, we only searched for the volatile monoterpene metabolites of fungal symbionts and so cannot speculate on whether or not monoterpenes are employed as carbon sources. Although we did not use isotopically labeled substances to confirm that fungal symbionts metabolize host tree monoterpenes to oxygenated derivatives, the de novo synthesis of oxygenated cyclic monoterpenes by these fungi can be almost ruled out. In previous work, we collected volatiles from G. penicillata and the other I. typographus-associated fungi studied here growing under many conditions without spruce bark or supplemental monoterpenes [38]. While we detected a large range of volatile compounds, we never found any trace of oxidized cyclic monoterpenes like those reported here, which all have the bornane, p-menthane, pinane, or thujane carbon skeletons characteristic of spruce monoterpenes. Some fungi do produce cyclic monoterpenes of their own, but none have yet been reported to synthesize the variety of cyclic skeletons produced by conifers, and so are very likely to obtain their precursors for oxygenated monoterpenes from conifer trees.

Oxygenated monoterpenes are widespread volatile cues for tree-feeding insects Various forest insects that are associated with fungal symbionts, such as bark beetles, ambrosia beetles, and wood wasps, live in host trees producing large quantities of monoterpene volatiles. These insects are often attracted to their fungal symbionts through volatiles [2,38,52], and, hence, fungus-produced monoterpene metabolites could be critical components of the attractive volatile blends. For bark beetles, oxygenated monoterpenes derived from either the beetles, fungi, or host trees are already known to play important roles in their life history [53]. During mass attacks, the host-derived oxygenated monoterpene cis-verbenol acts together with 2-methyl-3-buten-2-ol as an aggregation pheromone for I. typographus to promote mass attack on individual trees [28,54]. Beetles also utilize oxygenated monoterpenes to restrict the density of attack. Microbes lining the gallery walls or living in the beetle gut oxidize cis- and trans-verbenol into verbenone, which inhibits the attraction of both sexes to densely colonized trees [55–57]. In addition, mated male I. typographus produce ipsenol and ipsdienol, of which ipsenol acts as an anti-attractant [57]. Furthermore, an oxygenated monoterpene from host trees, 1,8-cineole, which is produced in higher amounts in resistant or MeJA-primed trees, inhibits attraction of beetles to their pheromones [42,58,59]. Oxygenated monoterpenes are also used as reliable cues by parasitoids of bark beetles to locate their prey [35,60]. Here, we discovered that a blend of oxygenated monoterpenes emitted by I. typographus-associated fungi growing on agar medium amended with spruce bark or specific spruce monoterpenes is highly correlated with beetle attraction (Fig 6A and 6C). These same fungi were shown in a previous study [38] to produce mixtures of non-terpene aliphatic and aromatic volatiles when grown without sources of monoterpenes, and a blend of these aliphatic and aromatic compounds attracted newly emerged (callow) adult I. typographus [38]. These non-terpene compounds were also detected in the present study as major components of the volatile blend at later phases of fungal growth (S3–S7 Tables). Although we focused principally on the symbiont G. penicillata in our work, volatiles from other fungal symbionts were also investigated. L. europhioides and E. polonica emitted volatiles on SBA that were also attractive to adult I. typographus (S1 Fig), but the volatiles of O. bicolor and saprophytes such as O. piceae and Trichoderma sp. were not attractive. The production of oxygenated monoterpenes by other microbes coinhabiting I. typographus galleries, such as fungal saprophytes, yeasts, and bacteria, has been reported before, but these microorganisms produce different metabolites of host tree monoterpenes (or the same metabolites in different ratios) [61–63] than those produced by symbiotic fungi as reported here. In addition, the blend of oxygenated monoterpenes emitted by I. typographus alone was distinct from that emitted by the fungal symbionts [64] (S12 Fig). Since oxygenated monoterpenes are major components of the volatile profiles of I. typographus fungal symbionts that are attractive to beetles (S2–S7 Tables and S9 Fig), it can be suggested that the oxygenated monoterpenes themselves are important in I. typographus attraction. Bioassays of individual oxygenated monoterpenes showed that two were attractive to beetles at specific concentrations, and two had no effect (Fig 6B and 6D). I. typographus was recently reported to be attracted to the fungi associated with the North American spruce bark beetle D. rufipennis when attacking a different species of spruce than I. typographus normally does in its home range. This attraction could also be due to oxygenated monoterpene profiles, although it is unclear if an association with these fungi would be beneficial to I. typographus [65].

Oxygenated monoterpenes signal the presence of fungi to I. typographus and may modulate beetle colonization The first chemical signals reported to mediate bark beetle colonization of their hosts were aggregation pheromones. Yet even in the presence of these pheromones, a large proportion of aggregating beetles that land on trees leave without tunneling into the bark [45,66,67]. This suggests that other cues may be needed to induce beetles to stay and bore into the bark. Indeed, bark beetles have been shown to respond to visual and chemical cues from host and nonhost species when selecting trees for colonization [28,38,44,68–70]. Based on our results, fungus-produced oxygenated monoterpenes might also belong to the list of colonization cues. For bark beetles, fungal metabolites can serve as indicators of host tree sectors where their fungal symbionts are already established. These compounds could also provide evidence for the ongoing metabolism of host tree defenses or signal that tree defenses have been overwhelmed, which could improve the success of bark beetle colonization. Fungal volatiles also enhance the attraction of I. typographus to aggregation pheromones. Female I. typographus is known to use the aggregation pheromone mixture (2-methyl-3-buten-2-ol and cis-verbenol) to locate trees suitable for mating and oviposition [45,71]. Here, we showed that female I. typographus were more attracted at short range to a combination of the pheromone mixture plus fungal volatiles than to the pheromone mixture alone. Similarly, a recent study reported that the presence of fungal volatiles increased the attraction of dispersing beetles to their pheromones in the field compared to traps with pheromones only [72]. Oxygenated monoterpenes and other fungal volatiles could provide information about the presence of fungal symbionts, which promote the successful development of their offspring. Similarly, the pheromone component cis-verbenol, itself an oxygenated monoterpene produced by I. typographus from the host tree precursor α-pinene [73], provides information about the presence of other beetles, especially males. The lack of response of males to pheromones in our experiments is not unexpected, as male I. typographus have been reported to be less responsive than females to high doses of pheromones in walking bioassays [71,74]. This behavior may help them avoid dense colonies of male conspecifics within a tree to reduce competition for mates and food. Depending on how fast symbiotic fungi colonize the tree bark, fungal cues could benefit both epidemic and endemic beetle populations. During the endemic phase, beetles generally colonize stressed trees in a gradual fashion over months [66], and in this case, volatiles emitted from fungal symbionts introduced into the tree by early colonizers, in combination with beetle pheromones, could increase the probability for conspecifics to locate attacked trees, present at low densities in spruce stands. In contrast, during epidemics, vigorous trees are typically colonized by large numbers of bark beetles within a few days, which overwhelms their defenses. Since ophiostomatoid fungi have been reported to emit volatiles within 1 d following the infection of detached bark plugs in the laboratory [75,76], these chemicals might function in attraction during epidemics as well. However, during rapid natural attacks, it is unclear if fungi can colonize a tree sufficiently to emit behaviorally active volatiles. More research in this area is needed. The oxygenated metabolites of host monoterpenes produced by fungal symbionts may not only attract bark beetles, but also stimulate them to tunnel in a fungus-colonized medium. I. typographus tunneled more in a diet with G. penicillata and β-pinene than with either fungus or monoterpene alone, perhaps because the fungus metabolized β-pinene to behaviorally active metabolites. However, this effect was not seen for α-pinene or bornyl acetate. Interestingly, both sexes showed similar tunneling preferences to fungi and monoterpenes in contrast to their different response to fungi and pheromones. The lack of sex-specific differences in tunneling could be due to the fact that the nutritional advantage of feeding on fungus-colonized spruce bark is beneficial to both male and female adults [17,18]. No other bioassays performed in this study showed sex-specific differences. The proportion of oxygenated monoterpenes to total monoterpenes in the volatile blend of G. penicillata increased over the time course studied to nearly 50% at 12 d and nearly 70% at 18 d post-inoculation, a trend also observed for L. europhioides and O. bicolor. Thus, higher proportions of oxygenated monoterpenes may indicate older fungal infection sites and, hence, older beetle invasion sites that may be less attractive to newly arriving beetles due to crowding. The lack of attraction and even repellency of certain individual oxygenated monoterpenes, as seen in laboratory bioassays in our own and in previous studies [77] is consistent with this interpretation. Among the oxygenated monoterpenes already reported to inhibit I. typographus attraction, verbenone, produced by microbial or autooxidation of the pheromone cis-verbenol (or by direct oxidation of α-pinene), reduces aggregation of I. typographus in later phases of the attack on the tree [57,63,78]. Oxygenated monoterpenes are signals not only for bark beetles, but also for their natural enemies. Both beetle predators and parasitoids employ these compounds and other volatiles to locate bark beetle larvae hidden under the bark [79,80]. Specifically, a three-component blend comprising camphor, isopinocamphone, and terpinen-4-ol, all fungal metabolites of host tree monoterpene hydrocarbons, was reported to attract a coleopteran predator and several hymenopteran parasitoids of I. typographus in the presence of host tree background signals [60,81,82]. Furthermore, the bark beetle predator, Thanasimus formicarius, possesses OSNs to detect oxygenated monoterpenes such as camphor and pinocamphone as well as several additional OSN classes detecting various oxygenated bark beetle pheromone compounds [83]. Thus, any benefit to the beetle arising from oxygenated monoterpene production by its symbiotic fungi may come at the cost of revealing its presence to natural enemies that employ these same volatiles to locate bark beetles.

Formation of oxygenated derivatives may reduce monoterpene toxicity for I. typographus The conversion of host tree monoterpene defenses by symbiotic fungi to oxygenated products may alleviate toxicity to bark beetles. Terpene-rich resins are a general defense of P. abies and other conifers against herbivores and pathogens [84–86]. Thus, it is not surprising that monoterpenes have exhibited toxicity to bark beetles in many studies [87–89]. Monoterpene hydrocarbons, such as α-pinene, are typically more toxic to beetles than host tree-produced oxygenated monoterpenes, such as bornyl acetate [90]. Hence, the oxidative transformations carried out by fungal symbionts described in this study could reduce toxicity to I. typographus through conversion to less poisonous derivatives. Such detoxification of host tree defenses could represent a significant benefit of fungal symbionts [91]. By contrast, oxygenated monoterpenes may be more toxic to fungi compared to monoterpene hydrocarbons [48,49]. Thus, the initial oxidation of monoterpene hydrocarbons may not constitute a detoxification unless it is a step towards further metabolism. The potential toxicity of oxygenated monoterpenes may explain why these substances are degraded further after initial oxidation by fungi specialized on conifers such as Heterobasidion parviporum and Seridium cardinale [48,49]. Another class of terpenes present in P. abies bark and other conifers host trees of bark beetles are diterpene resin acids, which were reported to be toxic to associated fungi in a North American bark beetle system [92].

Other sources of oxygenated monoterpenes in spruce bark beetle interactions Oxygenated monoterpenes emitted from trees attacked by I. typographus may arise from sources other than fungal symbionts. The host tree P. abies itself synthesizes large amounts of bornyl acetate [93] and small amounts of 1,8-cineole [94]. In these compounds, the oxygen functions are incorporated during biosynthesis from basic precursors, whereas the products from fungal symbionts are presumably formed by oxidative modification of a previously formed monoterpene hydrocarbon skeleton. The compound trans-4-thujanol belongs to the latter group. We identified it as a G. penicillata metabolite of α- and β-pinene, but trans-4-thujanol may also be synthesized by the tree, although at low levels in P. abies bark [77]. As an alternative, this and other oxygenated monoterpenes could be produced via autoxidation of monoterpene hydrocarbons [95]. The oxidation of monoterpenes upon exposure to air could explain the release of low but readily detectible amounts of these compounds from uninfected control bark plugs in our and other studies. In the field, oxygenated monoterpenes other than bornyl acetate and 1,8-cineole have been detected from damaged P. abies trees when monoterpenes were exposed to air [96,97]. However, in the present study, the emission rate from uninfected bark plugs was much lower than from fungus-infected plugs, suggesting that microbial metabolism is a much more significant source of oxygenated monoterpenes than autoxidation (S9 Fig) [95]. Nevertheless, the tree itself cannot be ruled out as a source of any of the detected oxygenated monoterpenes, since P. abies cell suspension cultures have been reported to oxidize added monoterpenes [94,95]. If P. abies produced monoterpenes as a defense response only when infected by fungi, it might be difficult to determine which organism was producing these compounds—the tree or the fungus—without identifying specific biosynthetic genes in one of the two genomes. More research is needed to assess the contribution of host trees to the oxygenated monoterpenes emitted in response to beetle tunneling or fungal infection. Among microbial sources of oxygenated monoterpenes are several yeast species including Hansenula holstii, H. capsulata, and Candida nitratophila, which were isolated from I. typographus and produce terpinen-4-ol, α-terpineol, borneol, and trans-pinocarveol when grown in phloem medium or in α-pinene-supplemented medium [63]. In addition, several bacterial symbionts of bark beetles are capable of metabolizing monoterpenes into oxygenated derivatives, such as verbenols and verbenone [62]. Some bacteria coinhabit the beetle gallery together with ophiostomatoid fungi, but it is not known if these bacteria induce the production of oxygenated monoterpenes, either by producing them themselves or by interacting with fungi to form these compounds [98]. Another bark beetle species, Polygraphus poligraphus, which is sometimes found together with I. typographus, was shown to emit large amounts of terpinen-4-ol [99,100]. Intermediate amounts of α-terpineol, cis- and trans-4-thujanol were also identified from the hindgut as well as the entrance holes of this beetle’s gallery and could be formed by microorganisms associated with this beetle species from host tree monoterpenes or by the beetle itself. Our experiments with fungus-free I. typographus, however, demonstrated that live beetles themselves did not produce high concentrations of any of the fungal biotransformation products identified in this study. Instead, beetles formed a range of other oxidation products (S12 Fig) that have been identified in previous studies [64,101]. In summary, our study brings together several lines of evidence showing that the oxygenated derivatives of spruce resin monoterpenes, which appear during I. typographus attack, are produced mainly by the metabolism of symbiotic fungi. The fungal origin of these compounds had been suggested previously [32] but not supported by experimental evidence until now.

[END]
---
[1] Url: https://journals.plos.org/plosbiology/article?id=10.1371/journal.pbio.3001887

Published and (C) by PLOS One
Content appears here under this condition or license: Creative Commons - Attribution BY 4.0.

via Magical.Fish Gopher News Feeds:
gopher://magical.fish/1/feeds/news/plosone/

You are viewing proxied material from magical.fish. The copyright of proxied material belongs to its original authors. Any comments or complaints in relation to proxied material should be directed to the original authors of the content concerned. Please see the disclaimer for more details.