The Effect of Inactivation of Aldehyde Dehydrogenase on Pheromone Production by the Gut Bacteria of an Invasive Bark Beetle, Dendroctonus Valens (LeConte)

Semiochemical-based strategies are important for bark beetle management worldwide. One of the most destructive invasive bark beetles, the causal agent of mass pine (Pinus sp.) mortality in China, is the red turpentine beetle (RTB) Dendroctonus valens (LeConte), originating from North America. For this species, verbenone pheromone regulates the beetles’ attack density in a dosage dependent manner. In addition, we have previously shown that RTB’s gut bacteria is involved in the pheromone production under both anaerobic and micro-oxygen environments. However, although some investigations of anaerobic gut bacteria of bark beetles have been made, functional verication of their role at molecular level is still lacking. To clarify the function of gut bacteria in verbenone production, we investigated the activity of key genes of the primary gut bacteria involved in verbenone production in D. valens under anaerobic environment. These key genes (aldehyde dehydrogenase) identied by transcriptome analysis were then knocked out by homologous recombination to obtain mutant bacteria strains. Our results show that these mutants had signicantly decreased ability to convert the monoterpene precursor to verbenone compared to the wild type bacteria. Our ﬁ ndings provide further evidence of the mechanism of pheromone production of D. valens and a new perspective for functional studies of gut bacteria in general.

. Microbiota is also involveld in other key functions in insects' life, such as pheromone production. For example, gut microbiota is largely responsible for the production of aggregation pheromones of locusts, sex pheromones of cockroaches (Dillon et al. 2002), and in uence the mating preference of Drosophilla meladella (Sharon et al. 2010).
Bark beetles are economically important insect pests of conifer and broadleaf trees, and pheromonebased mass trapping and push-pull techniques are essential for their successful management (Wermelinger 2004;Witzgall et al. 2010). A large proportion of bark beelte pheromones are composed of monoterpenoid oxides (Wood 1982;Blomquist et al. 2010). For example, α-pinene is not only an important host defensive monoterpene, but also a precursor for the bark beetle pheromone verbenone (Blomquist et al. 2010;Gitau et al. 2013). Many micro-organisms have been implicated to be involved in the bioconversion of precursors to terpenes and in the synthesis of terpenoid pheromones. In Ips paraconfusus, for instance, gut bacteria oxidize α-pinene to pheromone verbenol (Brand et al. 1975). Pheromone synthesis by microorganisms is conducted by degrading plant compounds, and some investigation of the mechanisms behind this process have also been made. For example, Merlin et al. (2005) suggested that aldehyde oxidase could degrade aldehyde odorant compounds, such as pheromones or plants volatiles. Comparison of in vivo and in vitro physiological data from Drosophila melanogaster adults con rmed that aldehyde dehydrogenase (ALDH) was responsible for the detoxi cation effect of acetaldehyde in vivo (Leal and Barbancho 1992).
Red turpentine beetle Dendroctonus valens LeConte (Coleoptera: Scolytidae), was introduced to China from North America in the 1980s. Despite D.valens is an innocuous secondary pest in its native range, has become a primary tree-killer of Chinese red pine Pinus tabulaeformis (Carr) in the new invasive range, causing substantial economic losses and recieving its status as China's most destructive forest pest.
Semiochemical-based management strategies, including pheromone traps, has been successfully implemented in D. valens' control (Sun et  Previous studies investigating pheromone conversion of D.valens have shown that oxygen concentration inside the gut affecst the rate of pheromone synthesis (Cao et al. 2018). However, Cao et al. (2018) showed that nine out of ten species of gut facultative anaerobic bacteria were able to convert cisverbenol into verbeone under both anaerobic and micro-oxygen environments, facultative anaerobic gut bacteria Enterobacter xiangfangensis having the strongest ability to synthesize this pheromone.
Althought pheromones of D. valens have been extensively studied, molecular level con rmation of the function of its gut bacteria in pheromone production is still lacking. To address this knowledge gap, we identi ed the key bacterial genes involved in D. valens' verbenone production under anaerobic conditions. More speci cially, we analyzed the ability of the most effective facultative anaerobic gut bacteria E. xiangfangensis of D. valens to covert cis-verbenol to verbenone. We then identi ed the key genes responsible for this task via transcriptome analysis, knocked them out by homologous recombination to obtain mutant strains and determined the ability of these mutants to convert the monoterpene precursor into the pheromone product. Our ndings will shed light on the molecular basis how gut bacteria participate in the synthesis of pheromones, and provide a new scienti c basis for the functional study of gut bacteria and the mechanism of pheromone synthesis of D. valens.

Materials And Methods
Bacterial strains, plasmids, and growth conditions Strains used in this study are listed in Table 1. E. xiangfangensis were cultured in Luria-bertani medium (LB medium, per liter, 10 g of trypton, 5 g of yeast extract) at 30 ℃. E. coli strains were cultured in LB medium at 37 ℃.

RNA-Seq library construction and sequencing
We selected the E. xiangfangensis strain as the key strain. After the activation, E. xiangfangensis were cultured in 5 mL LB medium at 30 ℃, 180 rpm for 24 hours until the light absorption value reached 0.5 at the wavelength of 600 nm. Then, 40 ng/L cis-verbenol was added to the nal concentration as the treatment group, and DMSO was added to control.
The bacteria were cultured at 30 ℃, 180 rpm for another 16 hours and then were collected in a 1.5 mL tube by centrifugation at 12000 rpm for 3 min. Three replicates were conducted for each group. RNA extraction was performed using the RNeasy Mini Kit (QIAGEN, USA Brie y, the upstream and downstream genomic sequences of the ALDH coding sequence were separately ampli ed using primers listed in Table 2. After ligating using restriction enzyme and digesting pRE112 using ClonExpress II One Step Cloning Kit (Vazyme, China) (Liu et al. 2020), the recombinant plasmid was transferred to the E.coli S17-1 (λpir) component cells. The positive clones were con rmed by PCR using primers in Table 2 and sequenced.
The positive E.coli S17-1 (λpir) colony was cultured in 5 mL LB medium at 37 ℃, 180 rpm for 16 hours, meanwhile the wild type E. xiangfangensis were cultured in 5 mL LB medium at 30 ℃, 180 rpm. Then, we mixed 1mL of each of the above two strains with 3 mL LB medium and cultured at 30℃, 180 rpm for 24 hours, after which the bacteria were collected and spread in the LB plate (50mg/L chloramphenicol, 50mg/L Ampicillin). After culturing overnight, the putative single-cross-mutant clones were con rmed by PCR, and cultured in LB medium for second-round homologous cross-over. The ALDH mutants were obtained by spreading the media in the LB agar plate containing 10% sucrose and con rmed by PCR with two pairs of primers.

Conversion experiments
Wild type E. xiangfangensis and its ALDH gene-de cient mutants were cultured in LB medium and incubated for 24 h. A dilution of 1:100 of each isolate was made when cultures were adjusted to an optical density (OD 600 ).0.5. concentration (40 ng/µl and 200 ng/µl ) of cis-verbenol was then added into 4 ml bacterial suspensions and shaken for a further 36 h. Both the wild (control) and mutant E.
xiangfangensis bacteria suspensions contained an equivalent amount of cis-verbenol. All solutions were extracted with hexane and then stored for later chemical analysis to determine verbenone concentration.
The conversion experiments followed previously described methods (Cao et al. 2018), with slight modi cation of incubation and shaking times.
Statistical analysis R software (version 3.0.3) was used to for pearson's correlation analysis, and signi cant correlations were declared at r > 0.8 or < −0.8. Conversion experiment results were analyzed using Dunnett's T 3 test, and signi cances were determined at P<0.05.

Results
Transcriptome data analysis Statistics of sequencing data of E. xiangfangensis shown in Table 3. These sequence data have been submitted to the GenBank databases under accession number PRJNA798447. The genome of the E. xiangfangensis was used as a reference genome (reference genome completed by Beijing Novo Zhiyuan Sci-Tech Company Limited) shown in Table 4.  Overall distribution of expressed genes showed that 262 genes were differentially expressed in the control group (DMSO was added) compared to treatment (cis-verbenol was added), whereby 199 DEGs were upregulated and 63 DEGs were downregulated (Fig.1). The expression of control and treatment also displayed a clear separation based on PCA analysis, and the samples in the group were clustered together (Fig.2). These results show that D. valens gut strains have a complex oxidative defense mechanism against monoterpenes.

Pheromone related gene identi cation
Gene expression patterns and phylogenetic analysis from transcriptomic data of gut bacteria revealed that aldehyde dehydrogenase (ALDH, Gene ID: GM002204) genes were predominantly expressed in oxidoreductase activity (Fig. 3), and clustered with genes involved in pheromone synthesis and detoxi cation in other species. Furthermore, the transcription abundance of ALDH and aldehyde-alcohol dehydrogenase (ADH) was signi cantly upregulated when verbenol was added, and the expression level of ALDH was signi cantly higher than that of ADH. We therefore considered ALDH as the main candidate gene for D. valens pheromone synthesis. ALDH knock-out ALDH mutant E. xiangfangensis showed a shorter PCR product than wild type when using primer T1/T2 or T1/T3 (Fig. 4).

Conversion experiments
Comparison of cis-verbenol conversion results between the mutant and the wild-type strains of E. xiangfangensis at 40 ng/µl and 200 ng/µl concentrations showed that the wild-type strains (control group) had the ability to convert cis-verbenol, whereas this ability was signi cantly decreased in the ALDH-de cient mutant strain (Fig. 5).

Discussion
In our previous work, our group has demonstrated that the gut bacteria of D. valens have the ability to convert monoterpenes to pheromone products. For example, under aerobic environment, intestinal bacteria isolated from D.valens converted cis-verbenol to verbenone (Xu et al. 2015), and nine out of ten intestinal facultative anaerobic bacteria were able perform this function under different oxygen environments, but the conversion rate increased with increasing oxygen concentration (Cao et al. 2018).
In order to investigate the mechanisms of phernomone conversion ability at molecular level, we selected a bacterial strain with the strongest conversion ability (E. xiangfangensis, a facultative anarobe) to analyze its transcriptome with or without cis-verbenol, screened the target genes responsible for the pheromone conversion and investigated the converision ability of the bacteria in the absence of function of these genes.
To our knowledge, our current study is the rst one to provide molecular evidence to verify the function of intestinal bacterial in the conversion of monoterpenes to pheromone products under anaerobic conditions. When identifying candidate genes responsible for cisverbenol conversion, our transcriptome analysis showed that the highest expression level was detected in ALDH gene, which is a functional gene commonly found This result is in contrast with other studies where gut bacteria does not enhance insect's pheromone synthesis (Hunt et al. 1989;Bell et al. 2003). These inconsistencies might result from different experimental settings, such as use of diverse bacterial communities and nutritional media, which in turn in uence both microbial communities and host physiology. Another potential reason for differences among current and previous studies investigating pheromone converision of insects is that the cultivation and validation of intestinal facultative anaerobes in this study was carried out in an anaerobic environment, whereas some earlier studies have used atmospheric oxygen concentration (Keeling et al. 2016;Nadeau et al. 2017). Therefore, our simulated intestinal microenvironment may be more accurately re ect the oxygen conditions in insect integtines.
Previous studies on the function of intestinal bacteria mainly focused on their ecological roles (Cao et al. 2018), whereas our current study provides a new methodology for molecular veri cation of the functions of insects' intestinal microbes. Future work should consider external factors that in uence intestinal facultative anaerobic bacteria, such as insect gut hydrogen ion concentration (pH). To better understand the complex symbiotic relationships among bark beetles and their micro-organisms, future studies should also address the functions of these bacteria to D. valens development, detoxi cation, and chemical communication. Overall distribution of differently expressed genes of Enterobacter xiangfangensis, the primary bacteria invoveld in verbanone conversion of D. valens in the presense and absence of the pheromone precursor cis-verbenol. Signi cantly differentially expressed genes marked with red (elevated) and green (degreased) dots. Non-signi cant differences are indicated by blue dots. ExCK vs ExV= The difference genes of control group (DMSO was added) and treatment group (cis-verbenol was added) were compared.   Construction of the ALDH gene-de cient Enterobacter xiangfangensis mutants. PCR identi cation of mutants with T1/T2 or T1/T3; Marker. DL5000 DNA Ladder; 1, wild-type; 2, mutant.