Discovery of Rickettsia and Rickettsiella Intracellular Bacteria in Emerald Ash Borer Agrilus planipennis by Metagenomic Study of Larval Gut Microbiome in European Russia

Abstract

Emerald ash borer Agrilus planipennis (Coleoptera: Buprestidae) is a quarantine pest posing a threat to ash trees all over Europe. This wood-boring beetle native to Asia is quickly spreading in North America and European Russia, and approaching the European Union and the Middle East. It is important to study microorganisms associated with this pest, because the knowledge of its “natural enemies” and “natural allies” could be potentially used for the control of the pest. All previously published information about the A. planipennis microbiome was obtained in North America and China. We present the first study on procaryotes associated with A. planipennis in Europe. Alive larvae were sampled from under the bark of Fraxinus pennsylvanica in the Moscow Oblast and the gut microbiome was studied using metagenomic methods. Next-generation Illumina-based amplicon sequencing of the v3-v4 region 16S-RNA gene was performed. In total, 439 operational taxonomic units from 39 families and five phyla were detected. The dominant families in our samples were Pseudomonadaceae, Erwiniaceae and Enterobacteriaceae, in accordance with the published information on the larval gut microbiome in North America and China. We detected intracellular bacteria in A. planipennis for the first time, namely Rickettsia (Rickettsiaceae) and Rickettsiella (Diplorickettsiaceae). Representatives of the genus Rickettsia are known to be in mutualistic symbiosis with some phytophagous insects, while Rickettsiella bacteria are pathogenic to many arthropods. The finding of Rickettsia and Rickettsiella opens perspectives for future research on the interactions between these bacteria and A. planipennis and the possible use of these interactions for the control of the pest.

1. Introduction

Emerald ash borer Agrilus planipennis (Fairmaire, 1888) (Coleoptera: Buprestidae) is the most serious invasive pest of ash trees in the world [1]. The larvae of this wood-boring beetle develop in the cambium region and mostly disrupt the tree’s phloem, thereby impeding the translocation of nutrients and triggering canopy dieback and tree decline [1].

The native range of A. planipennis occupies a restricted territory in East Asia, namely in northeastern China, Japan, the Korean Peninsula and the southern Russian Far East [2]. Emerald ash borer was a little-known species and regarded as a minor pest before its first record in North America in 2002 [3] and in Europe, namely in Moscow, in 2003 [4]. Since that time, the pest has been spreading over both continents and killing millions of ash trees. By now its range in North America occupies about a half of the continent [5], and the range in Europe occupies 19 regions of European Russia and part of Ukraine [6,7,8]. Now the southern border of A. planipennis range in Russia is approaching the Caucasus and Kazakhstan [9], and the western border is approaching the European Union (EU) [7,10]. Since ash trees (Fraxinus spp.) are very usual in European forests [11], and the climatic factors could probably not limit the spread of A. planipennis in the most part of Europe [12,13], A. planipennis is included into the list of the priority quarantine pests of the EU [14]. The potential spread of A. planipennis to the Caucasus region and the Middle East could threaten not only ash trees, but also olive trees [9,15]. Therefore, A. planipennis is included into the lists of quarantine pests in Turkey and the countries of Transcaucasia [16].

It is important to study bacteria associated with the emerald ash borer because procaryotes play an important role in the physiology and ecology of wood-boring insects and their impact on the host trees [17]. Bacterial symbionts of insects have been shown to function as reproductive manipulators, nutritional mutualists and defenders of their hosts [18]. Since the larva is the most destructive stage of this pest, the study of the larval microbiome is especially important.

Surprisingly, just a few articles on A. planipennis’s microbiome are available. The bacterial microbiome of the A. planipennis larva was studied in the native range of the pest in China via next-generation sequencing of the v3-v4 fragment of the 16S-RNA gene [19]. It was found that the dominant bacteria in the larval midgut of A. planipennis are Pseudomonadaceae, Xanthomonadaceae and Enterobacteriaceae, and that the resistance of the ash tree F. velutina to A. planipennis is connected to the induced increase in lignans, which suppress these bacteria families [19]. This example shows that there are complex interrelationships between A. planipennis, its microbiome and host plants. Therefore, the ecology of A. planipennis should be studied in the context of these interrelationships.

Vasanthakumar et al. [20] studied the gut microbiome of all life stages of the emerald ash borer (larva, prepupa and adult) in the USA by sequencing clones (pGEM-Tvector) of 16S-RNA genes amplified from total DNA. They figured out that the microbiome associated with the emerald ash borer is quite different at different life stages. Only 16% of the detected species were the same in all stages. Additionally, these authors detected that the microbiome of the emerald ash borer is much more diverse than the microbiomes of other wood-boring insects studied by them, including Anoplophora glabripennis Motschulsky, Saperda vestita Say, Dendroctonus frontalis Zimmerman and Ips pini (Say). Some microorganisms found by Vasanthakumar et al. [20] had not been recorded in other invertebrates before. According to Vasanthakumar et al. [20], the diverse, dynamic and presumably multifunctional microbial community associated with the emerald ash borer guts suggests that invasive insects should be viewed as multispecies complexes and that such an interpretation can improve our ability to develop more effective management approaches.

There were also several studies on the bacterial microbiome associated with the adults of the emerald ash borer in Canada [21,22,23]. All of them were based on metagenomic methods. Bergeron [21] studied the general composition of this microbiome and identified which genera belong to the core microbiome. Mogouong et al. [22,23] found that the structure of the microbiome is related to the density of the A. planipennis population [22], and that the phyllosphere microbiome appears to be a strong predictor of the microbial community structure in adult guts [23].

The microbiome associated with A. planipennis in its current European range of Russia and Ukraine has never been studied. Here, we present the first study of the bacteria associated with A. planipennis in Europe by studying the larval gut microbiome.

2. Materials and Methods

2.1. Site Location and Sampling

The Moscow Oblast is the epicenter of A. planipennis invasion. Shakhovskaya Town in the Moscow Oblast was chosen as the region of study because of the severe outbreak of A. planipennis recorded there in the last few years. Fraxinus pennsylvanica Marsh. trees were chosen for sampling, because most of the infestations in the center of European Russia refer to this species, which was introduced from North America and is widely planted in cities, while the only native ash species F. excelsior L. is rare in the center of European Russia [6,24,25].

We chose three trees of Fraxinus pennsylvanica infested with A. planipennis in a roadside planting in Shakhovskaya Town (56.036931 N, 35.502851 E), and sampled four larvae on 1 June and six larvae on 17 June 2020.

The larvae were collected from under the bark with sterile instruments and placed separately into 1.5 mL Eppendorf tubes. The collected material was transported to the laboratory on the same day. In the lab, each larva was superficially sterilized with 75% ethanol and then well-rinsed in sterile water. The guts were extracted with sterile scissors and fine tip forceps under a binocular microscope. The extracted guts were combined into two integrated samples (four larvae on 1 July and six larvae on 17 July). Here, 50 mg from each sample was used for DNA extraction. In addition, we collected one dead prepupa covered with mycelium and one dead adult from under the bark.

2.2. DNA Extraction, Sequencing and Data Processing

DNA isolation was performed with a NucleoSpin Soil Kit from Macherey-Nagel (Germany) in accordance with the manufacture’s instruction. The V3–V4 region of the 16S rRNA gene was amplified in duplicate using the F (341) 5′-CCT ACG GGN GGC WGC AG-3′/R (805) 5′-GAC TAC HVG GGT ATC TAA TCC-3′ primer set. The libraries were sequenced using Illumina MiSeq 2 × 300 bp sequencing; 280,000 total reads were obtained. The research was performed using equipment of the core centrum of the Genomic Technologies, Proteomics and Cell Biology of the All-Russian Research Institute for Agricultural Microbiology of the Russian Academy of Sciences (Sankt-Petersburg, Russia). Raw sequence data were processed to remove adapters, primers and primer-free reads using Qiime cutadapt. The DADA2 pipeline [26] was applied to denoise the reads, trim the reads based on quality scores (Q25 ≥ 100%) and merge the paired-end reads. Subsequently, the denoised sequences were checked for chimeric sequences using Qiime-Vsearch; the latter were excluded from the downstream analyses. The remaining sequences were clustered into operational taxonomic units (OTUs) using the open_reference algorithm against the Silva132 reference database based on a 97% consensus threshold. All annotations were manually checked via Blast PubMed. De novo OTUs were also manually assigned using Blast Genbank PubMed. OTUs with identical annotations at the species level were demultiplicated. Obsolete taxonomic data from Silva132 were checked.

The basic statistics (median, mean ± SD) and a non-parametric ANOVA were computed using Statistic 8.0 (Statsoft, Tulsa, AK, USA), while the indices of alpha-diversity were computed using Qiime2.3.

3. Results and Discussion

3.1. General Composition of the Larval Gut Microbiome

More than 220,000 sequences passed the filtration process. They formed 439 OTUs, containing more than 10 sequences. The annotated OTUs were attributed to 56 genera from 39 families from five phyla (

Table 1. Phyla, families and genera of bacteria detected in the metagenomic study of the Agrilus planipennis larval gut microbiome in European Russia.

Phylum	Family	Genus
Acidobacteria	Acidobacteriaceae	Terriglobus
Actinobacteria	Actinomycetaceae	Actinomyces
Actinobacteria	Brevibacteriaceae	Brevibacterium
Actinobacteria	Corynebacteriaceae	Corynebacterium 1
Actinobacteria	Lawsonellaceae	Lawsonella
Actinobacteria	Microbacteriaceae	Curtobacterium
Actinobacteria	Microbacteriaceae	Microbacterium
Actinobacteria	Micrococcaceae	Kocuria
Actinobacteria	Micrococcaceae	Micrococcus
Actinobacteria	Micrococcaceae	Rothia
Actinobacteria	Micromonosporaceae	Micromonospora
Actinobacteria	Nocardioidaceae	Nocardioides
Actinobacteria	Propionibacteriaceae	Cutibacterium
Actinobacteria	Pseudonocardiaceae	Actinomycetospora
Bacteroidetes	Chitinophagaceae	Segetibacter
Bacteroidetes	Dysgonomonadaceae	Dysgonomonas
Bacteroidetes	Flavobacteriaceae	Myroides
Bacteroidetes	Hymenobacteraceae	Hymenobacter
Bacteroidetes	Sphingobacteriaceae	Sphingobacterium
Firmicutes	Aerococcaceae	Globicatella
Firmicutes	Bacillaceae	Bacillus
Firmicutes	Bacillaceae	Virgibacillus
Firmicutes	Lactobacillaceae	Lactobacillus
Firmicutes	Paenibacillaceae	Paenibacillus
Firmicutes	Peptoniphilaceae	Peptoniphilus
Firmicutes	Planococcaceae	Rummeliibacillus
Firmicutes	Staphylococcaceae	Macrococcus
Firmicutes	Staphylococcaceae	Staphylococcus
Firmicutes	Streptococcaceae	Streptococcus
Proteobacteria	Acetobacteraceae	Craurococcus
Proteobacteria	Alcaligenaceae	Paenalcaligenes
Proteobacteria	Aurantimonadaceae	Aureimonas
Proteobacteria	Brucellaceae	Paenochrobactrum
Proteobacteria	Comamonadaceae	Acidovorax
Proteobacteria	Diplorickettsiaceae	Rickettsiella
Proteobacteria	Enterobacteriaceae	Citrobacter
Proteobacteria	Enterobacteriaceae	Enterobacter
Proteobacteria	Enterobacteriaceae	Kluyvera
Proteobacteria	Enterobacteriaceae	Lelliottia
Proteobacteria	Erwiniaceae	Erwinia
Proteobacteria	Erwiniaceae	Pantoea
Proteobacteria	Hyphomicrobiaceae	Pedomicrobium
Proteobacteria	Methylobacteriaceae	Methylobacterium
Proteobacteria	Moraxellaceae	Acinetobacter
Proteobacteria	Morganellaceae	Providencia
Proteobacteria	Neisseriaceae	Roseomonas
Proteobacteria	Oxalobacteraceae	Massilia
Proteobacteria	Pseudomonadaceae	Pseudomonas
Proteobacteria	Rickettsiaceae	Candidatus Megaira
Proteobacteria	Rickettsiaceae	Rickettsia
Proteobacteria	Sphingomonadaceae	Sphingomonas
Proteobacteria	Weeksellaceae	Chryseobacterium
Proteobacteria	Xanthomonadaceae	Stenotrophomonas
Proteobacteria	Yersiniaceae	Serratia
Proteobacteria	Neisseriaceae	Neisseria
Verrucomicrobia	Chthoniobacteraceae	Candidatus Udaeobacter

and Table S1).

Our samples were too small to make quantitative conclusions or reliable comparisons with the data obtained by Vasanthakumar et al. [20] in the USA and by Liu et al. [19] in China. Additionally, the methods used in our study and those two studies were different. However, there were at least 13 families detected in the A. planipennis gut microbiome in all three continents: Pseudomonadaceae; Erwiniaceae; Enterobacteriaceae; Xanthomonadaceae; Bacillaceae; Burkholderiaceae; Caulobacteraceae; Lactobacillaceae; Micrococcaceae; Moraxellaceae; Propionibacteriaceae; Sphingomonadaceae; Streptococcaceae (Figure 1).

The dominant families in our samples were Pseudomonadaceae, Erwiniaceae and Enterobacteriacea. The same families were also dominant in the USA samples according to Vasanthakumar et al. [20] and in the Chinese samples according to Liu et al. [19].

According to the experimental data obtained in China, Pseudomonadaceae, Enterobacteriaceae and Xanthomonadaceae play an important role in the life of A. planipennis [19]. The suppression of these bacteria in the larval midgut by the addition of lignans to the diet is correlated with a significant reduction in larva weight [19].

Our Pseudomonas OTU sequence has 100% identity with the sequence described in the wood-boring beetle, Hylecoetus dermestoides L. (Lymexylidae) (GenBank, PubMed MT605322.1). Another dominant OTU (Erwinia) has 100% identity with the sequence, described in the healthy leaves of Fraxinus excelsior (GenBank, PubMed MN989056.1).

3.2. The First Record of Rickettsia

Bacterial intracellular symbionts often occur in insects and play an important role in their life, causing different negative and positive effects [27]. Almost nothing is known about bacterial intracellular symbionts in A. planipennis. Only a single sequence in the larval midgut transcriptome similar to a Wolbachia sequence has been detected in the USA [28].

We found Rickettsia (Rickettsiaceae) in the gut of the A. planipennis larva. This is the first record of Rickettsia in A. planipennis. This finding is not surprising, because Rickettsia bacteria are recorded in beetles of different families, including Curculionidae, Cerambycidae, Chrysomelidae, Apionidae, Coccinellidae, Carabidae and Buprestidae [29]. As for the representatives of the family Buprestidae, to which A. planipennis belongs, Rickettsia infestation is known in Anthaxia nitidula (L.) in Europe [29] and Brachys tessellatus (Fabricius) in North America [30].

The finding of Rickettsia in A. planipennis raises many questions and opens prospects for future research because interactions between these bacteria and insects are very tight and complex. Rickettsia are obligate intracellular parasites. They are known to be in mutualistic symbiosis with some phytophagous insects. It has been experimentally proven that Rickettsia-positive adults of the tobacco whitefly Bemisia tabaci (Gennadius) (Hemiptera: Aleyrodidae) have a significantly higher fecundity than Rickettsia-negative adults [31]. The proliferation of Rickettsia may be involved in B. tabaci’s ability to defend against natural enemies [32] and may influence the thermotolerance [33]. In the southwestern United States, the range expansion of B. tabaci B biotype is apparently facilitated by the rapid spread of Rickettsia sp. nr. Bellii [34]. Whiteflies infected with Rickettsia had a significantly higher fitness level [34]. Rickettsia protect aphids against a fungal pathogen [35]. Rickettsia can manipulate the sex ratio of whitefly hosts by producing female-biased offspring [34]. This phenomenon is also known in Buprestidae beetles; Rickettsia bacteria are associated with sex ratio distortion and selective killing of male embryos in Brachys tessellatus [30].

One more interesting question is the geographic origin of Rickettsia discovered in European Russia. Rickettsia are the heritable bacterial endosymbionts, and the horizontal transmission of these microorganisms from one insect species to another through the host plant is known [36]. Therefore, theoretically, Rickettsia detected by us could originate from Asia, North America or European Russia. If the detected Rickettsia taxa are native to Asia, they could be brought with A. planipennis from its native range. If they are native to North America, they could be brought with F. pennsylvanica trees. If they are native to European Russia, A. planipennis could be infected from some native insects.

The OTU detected by us has been previously detected in different arthropods in 200 different parts of the world, for example in Sigara striata L. (Hemiptera) in the United Kingdom (99.75% identity with LR812279.1, GenBank, PubMed), in Hymenoptera in the USA (99.75% identity with LR800047, GenBank, PubMed), in Ricaniidae (Hemiptera) in Madagascar (99.5% identity with LR800144.1, GenBank, PubMed) and even in the crustaceans of the family Paraleptamphopidae in New Zealand (99.5% identity with MT507674.1, GenBank, PubMed). Therefore, we are unable to determine what is the geographical origin of our Rickettsia.

3.3. The First Record of Rickettsiella

We also found OTUs of another obligate intracellular bacteria genus—Rickettsiella (Diplorickettsiaceae). Rickettsiella DNA was detected in alive larvae and was abundant in one dead prepupa and one dead adult. Our Rickettsiella sequence has 100% identity with the sequence, described in the mites of the family Ixodidae, Haemaphysalis concinna (C. L. Koch) (LC388767.1, GenBank, PubMed) and Ixodes uriae White (KT697673, GenBank, PubMed). The authors who discovered them in these mites suggest that the mites act as a vector of Rickettsiella [37,38].

Rickettsiella bacteria are intracellular pathogens of a wide range of arthropods [39]. The Rickettsiella diseases are deadly for beetles and other insects. Since Rickettsiella are virulent and able to survive in the soil for years, it was proposed to use them for the biological control of pests [40]. However, as mammals may be susceptible to Rickettsiella, this control strategy needs further investigation to characterize Rickettsiella–host interactions [40]. Since the DNA of Rickettsiella was abundant in two dead specimens, and as Rickettsiella are pathogenic for insects, it could be hypothesized that Rickettsiella could be pathogenic for A. planipennis. On the other hand, it was shown that Rickettsiella can be not only pathogens, but also are able to increase the survival rates of some arthropods, such as aphids [41].

4. Conclusions

• The dominant families in our samples from European Russia were Pseudomonadaceae, Erwinaceae and Enterobacteriacea. Since the same families are known to be dominant in the A. planipennis larval gut in the USA and China, these families probably belong to the core microbiome typical for A. planipennis;
• Obligate intracellular parasites were first recorded in A. planipennis, namely representatives of the genus Rickettsia, which are known to be in mutualistic symbiosis with some phytophagous insects, and Rickettsiella, which are known to be pathogens of a wide range of arthropods;
• The finding of Rickettsia and Rickettsiella opens perspectives for future research on the interactions between these bacteria and A. planipennis. There is hope that if “natural enemies” or “natural allies” of the pest can be found, it could be used for the development of biological control measures.