http://orcid.org/0000-0003-0480-8296
Figures
Abstract
Arthropod vectors are responsible for the transmission of human pathogens worldwide. Several arthropod species are bird ectoparasites, however, no study to date has characterized their microbiota as a whole. We sampled hematophagous ectoparasites that feed on migratory birds and performed 16S rRNA gene metabarcoding to characterize their microbial community. A total of 194 ectoparasites were collected from 115 avian hosts and classified into three groups: a) Hippoboscidae diptera; b) ticks; c) other arthropods. Metabarcoding showed that endosymbionts were the most abundant genera of the microbial community, including Wolbachia for Hippoboscidae diptera, Candidatus Midichloria for ticks, Wolbachia and Arsenophonus for the other arthropod group. Genera including pathogenic species were: Rickettsia, Borrelia, Coxiella, Francisella, Bartonella, Anaplasma. Co-infection with Borrelia-Rickettsia and Anaplasma-Rickettsia was also observed. A global overview of the microbiota of ectoparasites sampled from migratory birds was obtained with the use of 16S rRNA gene metabarcoding. A novel finding is the first identification of Rickettsia in the common swift louse fly, Crataerina pallida. Given their possible interaction with pathogenic viruses and bacteria, the presence of endosymbionts in arthropods merits attention. Finally, molecular characterization of genera, including both pathogenic and symbiont species, plays a pivotal role in the design of targeted molecular diagnostics.
Citation: Cerutti F, Modesto P, Rizzo F, Cravero A, Jurman I, Costa S, et al. (2018) The microbiota of hematophagous ectoparasites collected from migratory birds. PLoS ONE 13(8): e0202270. https://doi.org/10.1371/journal.pone.0202270
Editor: Brian Stevenson, University of Kentucky College of Medicine, UNITED STATES
Received: November 7, 2017; Accepted: July 31, 2018; Published: August 27, 2018
Copyright: © 2018 Cerutti et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: Relevant data can be found in the paper and its Supporting Information files. Sequencing data can be accessed at https://www.ncbi.nlm.nih.gov/ under accession number SRP113690 or directly at https://www.ncbi.nlm.nih.gov/Traces/study/?acc=SRP113690.
Funding: This work was supported by the Italian Ministry of Health [Ricerca Corrente 2012 IZS PLV 16/12 RC]. Irena Jurman, Slobodanka Radovic and Federica Cattonaro are employed by IGA technology services. IGA technology services provided support in the form of salaries for authors IJ, SR and FC, but did not have any additional role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript. The specific roles of these authors are articulated in the ‘author contributions’ section.
Competing interests: We have the following interests: Irena Jurman, Slobodanka Radovic and Federica Cattonaro are employed by IGA technology services. Stefano Costa is employed by Laboratorio Chimico Camera Commercio Torino. There are no patents, products in development or marketed products to declare. This does not alter our adherence to all the PLOS ONE policies on sharing data and materials.
Introduction
Arthropod vectors are responsible of numerous diseases (named vector-borne diseases) worldwide [1]. Mosquitoes, ticks, Phlebotominae and Simuliidae flies are ectoparasites that can transmit viruses (e.g., Dengue virus, Yellow fever virus, West Nile virus (WNV), and Zika virus), bacteria (e.g., Borrelia spp., Rickettsia spp., Francisella tularensis, Coxiella burnetii), and parasites (e.g., malaria Plasmodium spp., trypanosomes, Leishmania spp.) [1]. The 2016 Zika virus pandemic is only the most recent example of a global vector-borne disease emergency among the many other pathogens for which there is an epidemic trend [2]. For example, the hard tick Ixodes ricinus, present throughout Europe, is involved in the transmission of a variety of pathogens of medical and veterinary importance including Borrelia burgdorferi s.l., tick-borne encephalitis virus, Anaplasma phagocytophilum, Francisella tularensis, Rickettsia helvetica and Rickettsia monacensis, Babesia divergens and Babesia microti, Louping ill virus, and Tribec virus [3].
Some of the arthropods responsible for disease transmission share their environment with birds. Mosquitoes belonging to Culex are, in fact, mainly ornithophilic and are the main vectors of WNV and Usutu virus. Moreover, birds physically carrying arthropods (such as ticks or mites) feeding on them can introduce novel species to Europe, as recently recorded for the U.K. [4–6].
Owing to its geographical location, the Italian peninsula is crossed by migratory routes from North and sub-Saharan Africa. To our knowledge, no data have been published on the whole microbiota of ectoparasites collected directly from migratory birds, though a few studies have described the presence and prevalence of specific genera of bacteria in ticks collected from birds or their nests [7, 8]. Since these ectoparasitic arthropods may carry pathogens, it may be relevant to study their microbial communities. Other than bacteria of public health interest, the microbiota of arthropods is complex. It has been described in ticks and mosquitoes [9–12] and the role of symbionts in influencing the microbial composition has been highlighted mainly in its interaction with pathogens. Symbionts like Wolbachia can influence arthropod reproduction, including male-killing, parthenogenesis, feminization, and embryonic mortality [13]. Furthermore, they may evolve the necessary adaptations to parasitize vertebrate cells, as recently demonstrated that the intracellular bacterium Coxiella burnetii evolved from a maternally-inherited endosymbiont of ticks [14]. Adaptation may also occur in the opposite direction, as in the case of the Francisella-like endosymbiont that evolved from Francisella tularensis [15].
For this study, we collected ectoparasites feeding on migratory birds during ringing sessions and then processed the arthropod samples for 16S rRNA gene metabarcoding to characterize their microbial community. Special care was paid to identify the genera commonly associated with pathogens. The samples reporting these bacteria were further tested with genus-specific or species-specific molecular assays.
Materials and methods
Sample collection
The ectoparasites were collected from birds during ringing sessions from November 2012 to October 2014. A total of 35 sessions were carried out at 14 different sites in five regions (Piedmont, n = 105; Lombardy, n = 5; Sicily, n = 4; Latium, n = 1; Liguria, n = 1; S1 Fig). A total of 194 ectoparasites were collected from 115 birds, divided into 120 pools by parasite type [a) Hippoboscidae diptera; b) ticks; c) other arthropods], host species, sampling site, date, and location. The species included in the other-arthropods group were: Anatoecus dentatus, Anaticola, Lucilia caesar, Colpocephalum turbinatum, Anystis, and Aphidiinae spp. Details on host species are reported in S1 Table.
The birds were caught with mist nets according to the Euring Ringing System and retrieved by authorized personnel. After capture the birds were ringed with a metal ring on the right leg. In Italy, bird leg rings are supplied by the Institute for Environmental Protection and Research (ISPRA) and they bear a unique, permanent code identifying any ringed bird for life. The birds were then identified by species, sexed, and assigned to age categories according to plumage. They were released after ectoparasite collection by veterinarians with ISPRA authorization. Being a non-invasive procedure, no special permission was needed for collection. Parasites from common swifts, mainly Hippoboscidae diptera, were collected either directly from the birds or from their nests in dedicated stations.
To preserve nucleic acids and obtain good quality material for metabarcoding, live parasites were stocked in RNAlaterTM stabilization solution (Invitrogen, Carlsbad, CA, USA) and stored at -80°C until processed. The parasites collected from each bird were pooled together in a single vial, except for two birds (Apus apus), for which the parasites were stored separately for preliminary evaluation of RNA integrity. Data on sampling site location, bird age, sex, and health status were collected and entered in a database.
RNA extraction
As the rationale of the study was to describe the living bacteria (i.e., synthesizing RNA), we analyzed the total RNA to characterize only the active microbiota and to remove bias from the DNA carried over from dead prokaryotic cells. RNA purification was performed with TRIzol™ (Invitrogen) in combination with a Nucleospin miRNA kit (Macherey-Nagel, Düren, Germany) following the manufacturer’s protocol for RNA purification of small and large RNA in two fractions. The large and small RNA fractions were stored at -80°C for further analysis.
Total RNA concentration and purity was estimated using a spectrophotometer for small volumes (Vivaspec, Sartorius, Göttingen, Germany) and a fluorometer (Qubit 2.0, Thermo Fisher Scientific, Waltham, MA, USA). The quality of total RNA was evaluated using a 2100 BioAnalyzer and an RNA 6000 Nano Kit (Agilent Technologies, Santa Clara, CA, USA). Though it was not possible to calculate the RIN (RNA Integrity Number) values [16], since the 28S rRNA subunit of many arthropods contains two hydrogen-bonded fragments that dissociate and co-migrate with the 18S subunit [17], the graph showed a 28S/18S sharp peak associated with a flat baseline that indicated the absence of degradation.
Reverse transcription and arthropod species identification
Total RNA was reverse transcribed using a High-Capacity cDNA Reverse Transcription Kit (Thermo Fisher Scientific) with 7 μl RNA as input and then stored at -20°C until processed. Vector species was determined by partial amplification and sequencing of the cytochrome c oxidase I (COI) gene, as described by Hebert and colleagues [18]. Briefly, the reaction mix was composed of 12.5 μl SuperMix PCR-UDG 2X (qPCR ProbesMaster, Jena Bioscience, Jena, Germany), 0.38 μl primer LC01490 20 μM, 0.38 μl primer HC02198 20 μM, 11 μl H2O, 0.75 μl cDNA, for a total volume of 20 μl. The thermal profile was: 50°C x 2 m; 95°C x 2 m; 40 cycles {94°C x 30 s, 49°C x 30 s, 72°C x 1 m}; 72°C x 5 m.
Successful amplification was verified using E-Gel® precast agarose gels at 2% (Thermo Fisher Scientific). Amplicons were then purified with a EUROGOLD Cycle-Pure kit (Euroclone, Pero, MI, Italy). The cycling reaction was performed with a BigDye® Terminator v1.1 cycle sequencing kit (Thermo Fisher Scientific): 2 μl BigDye® Terminator v1.1 ready reaction mix, 1 μl 5X sequencing buffer, 0.32 μl primer 100 μM, 4.68 μl H2O, 2 μl purified amplicon. The thermal profile was: 96°C x 1 m; 25 cycles at 96°C x 1 m, 50°C x 5 s, 60°C x 4 m. The reaction was purified with a GE Healthcare Illustra™ AutoSeq G-50 columns kit (GE Healthcare, Chicago, IL, USA) to remove dye terminators, and then submitted to sequencing on an Applied Biosystems AbiPrism 3130 (Foster City, CA, USA).
Chromatograms were analyzed with Sequencing Analysis v5.2 software (Thermo Fisher Scientific) for the base call and with FinchTV (Geospiza, Inc, Seattle, WA, USA) for hand editing. The obtained nucleotide sequences were used as query in a Blastn search on the GenBank nt database and in the BOLD database (Barcoding Of Life Database, www.boldsystems.org).
16S metabarcoding
The 16S rRNA gene metabarcoding of 116 out of 120 samples was performed following the protocol suggested by Illumina (four samples were discarded as they were not of adequate quality for sequencing). Briefly, 22.5 ng of cDNA was used as input for the first PCR using 16S amplicon PCR forward and reverse primers, amplifying V3-V4 regions of the 16S rDNA. After purification and second (index) PCR with a Nextera XT Index kit (Illumina, San Diego, CA, USA), the libraries were normalized according to fragment length and dsDNA molarity. The samples were pooled and processed in four sessions on a MiSeq platform (Illumina) using a MiSeq reagent kit v3-600 for 2x300 paired-end sequencing at the IGA Technology Services facility. The datasets generated and analyzed for this study are available in the BioProject database, with SubmissionID: SUB2898018 and BioProject ID: PRJNA396024.
Bioinformatic and statistical analyses
A first level analysis for all samples was achieved with MiSeq Reporter Metagenomics Workflow (MSR, Illumina) to gain an overview of the microbial community for each pool. The dataset was then analyzed following DADA2 workflow within the R framework, including quality check, error rate estimation, forward/reverse reads merge, chimera removal, ribosomal sequence variants (RSVs, equivalent to OTUs) determination, and taxa assignment to the GreenGenes gg_13_8_train_set_97, RDP Training Set 14 and SILVA version 128 reference databases for comparison [19–23]. Since SILVA performed better than the other two in classifying arthropod bacterial symbionts to appropriate taxa, only these results are presented. Data were also analyzed with the QIIME v.1.9.1 pipeline, and the results were largely consistent with those obtained with DADA2.
Alpha and beta diversities were estimated with the pyhloseq and vegan packages and visualized with the ggplot2 package, and differential expression was assessed with DeSeq2 [24–27]. Alpha diversity was estimated based on the observed species and Shannon index using the whole dataset after removing RSVs unassigned or assigned to Eukaryota. Beta diversity was estimated based on evenly sampled Bray-Curtis distance after filtering the low-frequency RSVs, pruning the samples with a sample size less than 1,800 and rarefaction to sample size of 1,800, where sample size was the number of individuals observed for each sample. Only data from DADA2 analysis using the SILVA database are presented in Results.
Molecular diagnostics for potential pathogen confirmation
To confirm the presence of the prokariotic genera identified by 16S rRNA metabarcoding, the pools were tested using a genus-specific or species-specific PCR for each of the potential pathogens detected by the MiSeq Reporter Metagenomics Workflow. The molecular tests included the following genera: Rickettsia spp., Anaplasma spp., Borrelia spp., Coxiella spp., Francisella spp., and Bartonella spp. For all PCR assays, amplicons were purified and submitted to Sanger sequencing as described earlier for the PCR targeting the COI gene.
Two different PCR assays to confirm and characterize Rickettsia were used: one targeting the citrate synthase gene, according to Regnery and colleagues [28], and the other targeting the 16S rRNA gene, according to Sprong and colleagues [29]. A PCR amplifying the partial 16S rRNA gene was used to identify A. phagocytophilum according to Stuen and colleagues [30]. Two different approaches were applied to identify Borrelia burgdorferi s. l. Real-time PCR targeting a tract of the 23S rRNA gene, highly conserved in all Borrelia species, was used to confirm positivity, as described in Courtney and colleagues [31]. An end-point PCR targeting a fragment of the flagellin gene, specific for the Borrelia burgdorferi s. l. group, was then used to determine whether the detected strain belonged to the causative agents of Lyme borreliosis [32]. The genomic group of samples positive for Borrelia burgdorferi s. l. was then identified by sequencing. A qualitative PCR targeting the IS1111 repetitive transposon-like region of Coxiella burnetii was performed to confirm Coxiella spp., as recommended by the Manual of Diagnostic Tests and Vaccines for Terrestrial Animals [33, 34]. Francisella spp. was investigated using a real-time PCR TaqMan® Francisella tularensis detection kit (Applied Biosystems) that targets the Tul4 and fopA genes. Francisella characterization was performed by targeting the 16S rRNA gene, according to Forsman and colleagues [35].
Results
Host and parasite species
A total of 115 bird hosts were included. S1 Table presents the species and the number of samples collected. Hosts were classified based on their migratory behavior as: resident, short-distance, mid-distance (North Africa) or long-distance (trans-Sahara) migration. A total of 194 parasites were collected and categorized into three groups based on the type of arthropod: Hippoboscidae diptera (n = 51, pool (n) = 49), ticks (n = 114, pool (n) = 60), other arthropods (OA, n = 29, pool (n) = 7). Hippoboscidae diptera were collected from common swifts (Apus apus) and classified by barcoding as Crataerina pallida. One sample of Hippoboscidae was collected from a goldcrest (Regulus regulus) and was classified as Ornithomya fringillina. The tick group included species of the genera Ixodes, Hyalomma, Ambylomma, and Haemaphysalis. The OA included lice (Mallophaga: Colpocephalum turbinatum, Anatoecus dentatus, and unidentified species), blowflies (Diptera: Calliphoridae: Lucilia caesar), mites (Trombidiformes: Anystis), the parasitoid wasps (Hymenoptera: Braconidae: Aphidiinae). A detailed list of ectoparasite species and number of sampled individuals and pools is given in S2 Table.
16S metabarcoding
A total of 116 pools were processed for V3-V4 16S rRNA gene amplification and massive parallel sequencing on an Illumina MiSeq platform, generating also a taxonomic report with MiSeq Reporter Software. The total amount of reads generated was 45,286,016 (median = 323,508, Q1 = 203,740, Q3 = 510,136); after quality filtering 24,012,404 reads per pair (median = 82,992, Q1 = 48,206, Q3 = 143,460, with read length uniformed to 230 bp) were obtained. Details on raw and filtered read numbers for each sample are reported in S3 Table. The resulting 257,298 dereplicated non-chimeric sequences were assigned to 2,257 RSVs, belonging to Bacteria and Eukaryota domains. Taxa assigned to Eukaryota or unassigned were removed, reducing the final number of RSVs to 2,184, classified in 23 different phyla. The most abundant phyla in the whole dataset were the Proteobacteria (90.72%) and the Firmicutes (5.44%). The most common genera for the whole dataset are reported in Table 1, and the taxa with relative abundance >1% are reported by group in Table 2. Fig 1 presents a graphical overview of the genera identified in each sample. The most abundant taxonomic groups of the bacterial composition at a lower taxonomical scale included symbionts like Wolbachia, Arsenophonus, and Candidatus Midichloria mitochondrii.
Samples are grouped by type of parasite (Hippoboscidae diptera, OA, and Ticks).
To better visualize the distribution of the symbiont genera in the Hippoboscidae diptera, a bar plot of the relative abundance within the family is reported in S2 Fig. Due to the large number of genera detected, Rickettsiales from ticks are reported in a bar plot (S3 Fig). To better represent the diversity of the symbionts within parasite species, the number of RSVs belonging to each genus are summarized in Table 3. According to the DADA2 developers, this algorithm in able to detect true biological sequence variants that might be considered different bacterial strains. The composition in RSVs of the major symbiont genera of Hippoboscidae diptera and ticks are reported in S4 Fig and S5 Fig.
Bioinformatic and statistical analysis
The reports generated by the integrated pipeline of the Miseq Reporter were analyzed. Based on these results, the samples were considered positive or negative for the detected potentially pathogenic genera. Table 4 reports the number of samples in which the MSR found the corresponding genus (MSR Hits). The microbiota of the three groups of ectoparasites differed in microbial composition by both abundance and represented taxa. Box-plots illustrating alpha diversity are reported in Fig 2. The Shannon index showed no statistically significant difference between the groups (box-plot Fig 2). There was a statistically significant difference in the number of species for the three groups. The box-plot represents the low number of species in the Hipposboscidae diptera group as compared to the other two groups. The principal coordinates analysis based on the Bray-Curtis distance highlighted a difference in microbial composition in the microbiota of the Hipposboscidae diptera as compared to that of the ticks and the OA groups (PERMANOVA test implemented in the adonis function in vegan) (Fig 3).
Indexes are observed species and Shannon index.
As reported by the axis label, the axis 1 shows 36.1% of variation in the samples.
Hits (OTUs matching a given genus) were obtained by MiSeq Reporter (MSR) analysis. Prevalence was calculated on 116 samples. The 95% confidence interval (CI) of the prevalence is reported in brackets.
Molecular diagnostics for potential pathogen confirmation
As summarized in Table 4, Rickettsia and Ehrlichia were highly represented among the samples, with more than 80% of the samples having reads corresponding to these two genera. The diagnostic PCR for Rickettsia spp. confirmed only 47 positive samples (42% of hits, 40.5% of total samples), further characterized as R. aeschlimannii, R. helvetica, and R. monacensis. Some of the sequences were similar to unclassified endosymbionts, 6 of which were were close to R. bellii, a species found only in the Americas (USA, Brazil, Argentina, Costa Rica, Colombia, El Salvador, Peru) [36–42]. As described for C. burnetii [43], identification of Rickettsia endosymbionts by means of a PCR used for pathogen detection in routine work shows that these species may interfere with the correct diagnosis of pathogenic rickettsial species.
Borrelia spp. were confirmed in 10 samples, four of which were further confirmed by the PCR specific for Borrelia burgdorferi s.l. and classified as B. valaisiana. The two samples positive for Francisella spp. were Francisella-like endosymbionts, while the two Anaplasma positive samples belonged to A. phagocytophilum. Ehrlichia required a different approach, since the MSR identified E. ovina in 98 samples, a species poorly described in the literature. On the basis of this unexpectedly high prevalence, and in contrast to only three sequences registered in the NCBI database, we suspected a misclassification issue, so we randomly chose one sample and retrieved the reads assigned to this species. A Blast search against the NCBI 16S prokariotic rRNA database was performed and the results were then plotted in MEtaGenome ANalyzer. The output is reported in S6 Fig The reads were classified as Anaplasmataceae, and at a lower taxonomic level as Candidatus Midichloria mitocondrii, Anaplasma spp., and Wolbachia spp. For this reason, the samples were not tested for Ehrlichia spp.
Among the PCR-confirmed samples, the following co-infections were observed: Borrelia-Rickettsia (n = 9), two of which occurred in individual samples, and Anaplasma-Rickettsia (n = 2). Considering only the ticks, the prevalence of confirmed Rickettsia spp. in the tick-only group was 60.32% (95% CI: 47.98–71.47), 15.87% (95% CI 8.86–26.81) for Borrelia spp., and 18.37% (95% CI 9.98–31.36) for Ixodes. Rickettsia spp. was present in numerous samples and detected in parasites collected from resident (n = 2), short-distance (n = 32), mid-distance (n = 2), and long-distance (n = 11) migratory birds. Borrelia spp. was detected only in ticks from short- and mid-distance migratory birds.
Discussion
With this study we wanted to describe the microbiota of ectoparasites collected from migratory birds since they constitute a route of introduction for exotic vector-borne diseases. The parasites were divided into three groups based on taxonomical features and sample size: Hippoboscidae diptera, ticks, and other arthropods. The microbiota of the Hippoboscidae diptera was composed of a limited number of species, as expressed by the low value of the observed species. Considering the evenness (Shannon index), the diversity was comparable to the tick and the OA groups. The low number of species may be explained by the predominance of symbiont species among the most abundant genera observed, such as Wolbachia, Arsenophonus, and Sodalis (Sodalis endosymbionts were also detected in Craterina melbae) [44–47]. The difference in microbial composition by number and taxonomy of the RSVs in the three groups is supported by the significant difference in the alpha and beta diversity, suggesting that the bacterial communities are heavily influenced by the parasite they live with. Briefly, the most abundant symbionts were Wolbachia, Rickettsia, Arsenophonus, and C. Midichloria. They were closely associated with the type of arthopod: Hippoboscidae were mainly colonized by Wolbachia and Arsenophonus and ticks by Rickettsia and C. Midichloria.
Regarding the distribution of the main genera in the Hippoboscidae, the microbial population of all but three samples was almost totally composed of Wolbachia. The three Wolbachia-free samples were totally colonized by Arsenophonus, while Sodalis was present only together with other symbionts. Only in three samples Wolbachia was the unique genus, for a total of 6 individuals with a single symbiont. Although the relative abundance is based on the family, the majority of the Hippoboscidae was colonized by at least two symbiont genera (S2 Fig).
Wolbachia strains seemed to be closely connected to host species (S4 Fig); C. pallida was mainly colonized by one variant and O. fringillina by another. In contrast, Arsenophonus in C. pallida had a slight diversity (with RSVs similar by 99.30–99.53%), while in O. fringillina it was present only as one RSV (similar to other Arsenophonus symbionts of Ornithomya species). The high homogeneity of Wolbachia suggests that it may be an obligate symbiont vertically inherited by maternal lineage. Differently, the diversity of Arsenophonus within samples suggests that it may have been transmitted horizontally or by other ways. The high presence of Wolbachia define this genus as the predominant symbiont. While it might be an obligate symbiont within C. pallida, the presence of other genera suggests that they still play an important role in the survival of Hippoboscidae, but further data are needed.
To our knowledge, this report is the first identification of Rickettsia bellii and R. monacensis in the Hippoboscidae C. pallida. Strains of R. bellii have been reported only for the Americas; similar strains have been detected in Australia, Thailand, Réunion Island, and Japan [44–47]. Our report is the first identification of R. bellii in Italy and Europe. This finding raises the question as to whether C. pallida behaves as an accidental vector for rickettsiosis or, if not competent for transmission, whether it might play a role as a sentinel parasite for the spread of arthropod-borne pathogens.
In ticks, the endosymbiont Candidatus Midichloria accounted for half of the RSV abundance in the samples, followed by Rickettsia spp. When Rickettsia spp. was present, it had the highest prevalence in almost all samples; only in one sample, Rickettsia was present but the predominant genus was Midichloria. As reported elsewhere, Wolbachia symbionts in ticks are rare [48]. Unlike a recent study on ticks in France, our study noted no relevant presence of Acinetobacter, but we did observe co-infection with pathogens and symbionts in our samples [49]. In the samples with only one tick, we observed co-infection mainly between Rickettsia and Midichloria; also Wolbachia, Rickettsiella, Neoehrlichia, and Spiroplasma were present together with other symbionts. Candidatus Midichloria was present mainly in Ixodes ricinus ticks, where it was represented by a unique variant. Unfortunately, the sample size was too small to make further observations for other species like I. arboricola and Hyalomma spp.
Candidatus Midichloria was first described in 2006 as an endosymbiont of Ixodes ricinus, and was later also detected in other hard ticks (Ixodidae) in Italy [50–51]. The detection of circulating DNA and the presence of antibodies against an antigen against M. mitochondrii in humans and mammals suggest that it might represent a novel group of vector-borne agents [52, 53].
The role of endosymbionts in arthropods has been partially described and a strong correlation with pathogen replication and transmission has been shown in some cases. For example, infection of Wolbachia(+) and Wolbachia(-) Culex quinquefasciatus colonies with WNV revealed a greater proportion of Wolbachia(-) infected mosquitoes developing high virus titers in saliva, which is necessary for virus dissemination and transmission [54]. This observation led to the suggestion that the difference in susceptibility to WNV infection between Cx. quinquefasciatus and Culex tarsalis might be partially explained by the difference in Wolbachia infection between these two species, since Cx. tarsalis is not infected with Wolbachia [54].
By applying the metabarcoding approach, we were able to detect several pathogenic species and to confirm several of them by species-specific or genus-specific PCRs. As for Rickettsia and Borrelia genera, the prevalence in our data set is shared by similar studies in Italy and Europe [55–59]. In addition, our findings show that Rickettsia seems to be widespread among residential and migratory birds, while Borrelia was detected only in short- and mid-distance migratory birds, suggesting different patterns in its transmission.
The observation of bacterial genera in the metabarcoding results not confirmed by the species-specific or genus-specific PCR tests may be explained by the presence of Rickettsia-, Coxiella-, and Francisella-like symbionts. The primer pairs used for the diagnostic tests were retrieved from published studies on the detection of pathogenic species of these genera. Most likely, these tests fail to detect a symbiont species of the targeted genus. As reported for Coxiella, the genetic diversity of symbiont organisms is very high, and little is known about their spread in arthropods, which may explain the discordance between the results of 16S rRNA gene metabarcoding and diagnostic PCRs [60]. Alternatively, it has been shown that some molecular tests that are specific for C. burnetii also detect Coxiella-like bacteria, leading to overestimation of the pathogen species. Indeed, the molecular characterization of bacterial endosymbionts plays a pivotal role in the design of targeted molecular tests for the sole detection of pathogenic species.
Recent studies have shown that C. burnetii could have originated from a tick-associated ancestor, while the Francisella-like endosymbiont of the hard tick Ambylomma probably evolved from a pathogenic strain of Francisella, indicating that tick endosymbionts can evolve from mammalian pathogens [14,15]. Little is known about these recently uncovered symbionts, perhaps because the research was biased towards the pathogenic species. Such is the case of the Coxiella genus, which only has two species (i.e., burnetii and cheraxi). The majority of studies have described C. burnetii because most isolates were collected from humans or domestic ruminants during Q fever outbreaks. More information on novel Coxiella-like organisms in non-vertebrate species like ticks has been acquired via 16S rRNA gene metabarcoding [14].
Finally, we observed a critical point in the bioinformatics analysis of our data. The first point is the erroneous identification of E. ovina by the MSR, not confirmed by deeper analysis. This issue may concern the used database, since, as reported in the Illumina manual, the Metagenomics workflow uses an Illumina-curated version of the Greengenes database. The choice of database may also lead to different results in taxa assignment. In our analyses, the SILVA database allowed us to assign RSVs to Wolbachia, C. Midichloria, and Arsenophonus. Since these three genera represent the majority of the microbial community, being symbionts, correct taxonomic assignment is very relevant for these kind of studies. We suggest the use of the SILVA database for future projects investigating arthropod microbiota by 16S rRNA metabarcoding.
Our metabarcoding analysis showed that the microbiota living with (and within) arthropods is complex, closely related to the host species, and that its major component comprises endosymbiont-related species. This approach provides a global overview of the bacteria present in/on ectoparasites collected live from migratory birds. Because it employs a universal primer set for prokaryotic metabarcoding, this approach was also useful for identifying in one shot the genera that include pathogen species. Since the method does not often discriminate beyond the genus level, a second-level, genus- or species-specific investigation was required to confirm the presence of the pathogen species in some samples. Without an overview provided by the metabarcoding method, multiple tests for each pathogen in all the samples would have been needed.
Supporting information
S1 Fig. Map of sampling sites.
Sites in Liguria, Latium, and Sicily were located on the islands of Palmaria, Ventotene, and Ustica, respectively. Each of these map tile sets are Stamen Design, under a Creative Commons Attribution (CC BY 3.0) license.
https://doi.org/10.1371/journal.pone.0202270.s001
(PDF)
S2 Fig. Bar plot of the relative abundance by family of bacteria in Hippoboscidae diptera.
The abundance is relative only to the family considered and not to the total microbiota.
https://doi.org/10.1371/journal.pone.0202270.s002
(PDF)
S3 Fig. Bar plot of the relative abundance of Rickettsiales in ticks.
The abundance is relative only to the Rickettsiales family and not to the total microbiota.
https://doi.org/10.1371/journal.pone.0202270.s003
(PDF)
S4 Fig. Bar plot representing the relative abundance of the RSVs by genus in the Hippoboscidae diptera.
Abundance is relative to the total microbiota. To improve color-coding readability, RSV numbering is assigned by genus, so that RSV 1 in Wolbachia is not the same as RSV 1 in Arsenophonus.
https://doi.org/10.1371/journal.pone.0202270.s004
(PDF)
S5 Fig. Bar plot representing the relative abundance of the RSVs in ticks for the most common genera.
Abundance is relative to only the genera considered. To improve color-coding readability, RSV numbering is assigned by genus, so that RSV 1 in Wolbachia is not the same as RSV 1 in Rickettsia.
https://doi.org/10.1371/journal.pone.0202270.s005
(PNG)
S6 Fig. Blast output of the reads classified as E. ovina by MSR and visualized in MEGAN.
https://doi.org/10.1371/journal.pone.0202270.s006
(PDF)
S1 Table. List of bird species caught during ringing sessions and hosting the ectoparasites sampled.
https://doi.org/10.1371/journal.pone.0202270.s007
(DOC)
S2 Table. List of the ectoparasites with their scientific name as obtained by partial sequencing of the COI gene.
The number of sampled individuals and the number of pools are reported.
https://doi.org/10.1371/journal.pone.0202270.s008
(DOC)
S3 Table. Number of reads before and after filtering for each sample and its relative parasite group.
https://doi.org/10.1371/journal.pone.0202270.s009
(DOC)
Acknowledgments
The authors wish to thank Carla Lo Vecchio for laboratory assistance with the PCR tests, Giovanni Savini (Istituto Zooprofilattico Sperimentale dell'Abruzzo e del Molise Giuseppe Caporale, Teramo), Riccardo Orusa (Istituto Zooprofilattico Sperimentale del Piemonte, Liguria e Valle d'Aosta, Aosta), Piergiovanni Piatti (Camera di Commercio, Torino), and Santo Caracappa (Istituto Zooprofilattico Sperimentale della Sicilia, Palermo) for their support in the study.
The authors are grateful to the two anonymous reviewers for their helpful suggestions in improving the manuscript.
References
- 1. Saldaña MA, Hegde S, Hughes GL. Microbial control of arthropod-borne disease. Mem Inst Oswaldo Cruz. 2017;112: 81–93.
- 2. Fauci AS, Morens DM. Zika virus in the Americas—yet another arbovirus threat. N Engl J Med. Massachusetts Medical Society; 2016;374: 601–604. pmid:26761185
- 3.
European Centre for Disease Prevention and Control (ECDC). Ixodes ricinus—Factsheet for experts [Internet]. 2014 [cited 13 Dec 2017]. Available: https://ecdc.europa.eu/en/disease-vectors/facts/tick-factsheets/ixodes-ricinus
- 4. Hoogstraal H, Kaiser MN, Traylor MA, Guindy E, Gaber S. Ticks (Ixodidae) on birds migrating from Europe and Asia to Africa 1959–61. Bull World Health Organ. World Health Organization; 1963;28: 235–262. pmid:13961632
- 5. Lindeborg M, Barboutis C, Ehrenborg C, Fransson T, Jaenson TGT, Lindgren P-E, et al. Migratory birds, ticks, and Crimean-Congo hemorrhagic fever virus. Emerg Infect Dis. 2012;18: 2095–2097. pmid:23171591
- 6. Jameson LJ, Morgan PJ, Medlock JM, Watola G, Vaux AGC. Importation of Hyalomma marginatum, vector of Crimean-Congo haemorrhagic fever virus, into the United Kingdom by migratory birds. Ticks Tick Borne Dis. 2012;3: 95–99. pmid:22300969
- 7. Duron O, Jourdain E, McCoy KD. Diversity and global distribution of the Coxiella intracellular bacterium in seabird ticks. Ticks Tick Borne Dis. Urban & Fischer; 2014;5: 557–563. pmid:24915875
- 8. Duron O, Cremaschi J, McCoy KD. The high diversity and global distribution of the intracellular bacterium Rickettsiella in the polar seabird tick Iixodes uriae. Microb Ecol. 2016;71: 761–770. pmid:26573831
- 9. Gofton A, Oskam C, Lo N, Beninati T, Wei H, McCarl V, et al. Inhibition of the endosymbiont “Candidatus Midichloria mitochondrii” during 16S rRNA gene profiling reveals potential pathogens in Ixodes ticks from Australia. Parasit Vectors. BioMed Central; 2015;8: 345. pmid:26108374
- 10. Finney CAM, Kamhawi S, Wasmuth JD, Janelle J, Patrel D. Does the arthropod microbiota impact the establishment of vector-borne diseases in mammalian hosts? Bliska JB, editor. PLOS Pathog. Public Library of Science; 2015;11: e1004646. pmid:25856431
- 11. Narasimhan S, Fikrig E. Tick microbiome: The force within. Trends Parasitol. Elsevier Ltd; 2015;31: 315–323. pmid:25936226
- 12. Novakova E, Woodhams DC, Rodríguez-Ruano SM, Brucker RM, Leff JW, Maharaj A, et al. Mosquito microbiome dynamics, a background for prevalence and seasonality of West Nile virus. Front Microbiol. Frontiers; 2017;8: 526. pmid:28421042
- 13. Mohanty I, Rath A, Mahapatra N, Hazra RK. Wolbachia: A biological control strategy against arboviral diseases. J Vector Borne Dis. 2016;53: 199–207. pmid:27681542
- 14. Duron O, Noël V, McCoy KD, Bonazzi M, Sidi-Boumedine K, Morel O, et al. The recent evolution of a maternally-inherited endosymbiont of ticks led to the emergence of the Q fever pathogen, Coxiella burnetii. PLoS Pathog. 2015;11: 1–23. pmid:25978383
- 15. Gerhart JG, Moses AS, Raghavan R. A Francisella-like endosymbiont in the Gulf Coast tick evolved from a mammalian pathogen. Sci Rep. Nature Publishing Group; 2016;6: 33670. pmid:27645766
- 16. Schroeder A, Mueller O, Stocker S, Salowsky R, Leiber M, Gassmann M, et al. The RIN: an RNA integrity number for assigning integrity values to RNA measurements. BMC Mol Biol. BioMed Central; 2006;7: 3. pmid:16448564
- 17. Winnebeck EC, Millar CD, Warman GR. Why does insect RNA look degraded? J insect Sci. 2010;10: 159. pmid:21067419
- 18. Hebert PDN, Cywinska A, Ball SL, deWaard JR. Biological identifications through DNA barcodes. Proc R Soc B Biol Sci. 2003;270: 313–321. pmid:12614582
- 19. Callahan BJ, McMurdie PJ, Rosen MJ, Han AW, Johnson AJA, Holmes SP. DADA2: High-resolution sample inference from Illumina amplicon data. Nat Methods. NIH Public Access; 2016;13: 581–583. pmid:27214047
- 20.
R Development Core Team. R: A language and environment for statistical computing. 2008.
- 21. McDonald D, Price MN, Goodrich J, Nawrocki EP, DeSantis TZ, Probst A, et al. An improved Greengenes taxonomy with explicit ranks for ecological and evolutionary analyses of bacteria and archaea. ISME J. BioMed Central; 2012;6: 610–618. pmid:22134646
- 22. Quast C, Pruesse E, Yilmaz P, Gerken J, Schweer T, Yarza P, et al. The SILVA ribosomal RNA gene database project: improved data processing and web-based tools. Nucleic Acids Res. 2013;41: D590–6. pmid:23193283
- 23. Cole JR, Wang Q, Fish JA, Chai B, McGarrell DM, Sun Y, et al. Ribosomal Database Project: data and tools for high throughput rRNA analysis. Nucleic Acids Res. 2014;42: D633–42. pmid:24288368
- 24. McMurdie PJ, Holmes S. Phyloseq: An R package for reproducible interactive analysis and graphics of microbiome census data. Watson M, editor. PLoS One. Public Library of Science; 2013;8: e61217. pmid:23630581
- 25. Oksanen J, Blanchet FG, Friendly M, Kindt R, Legendre P, McGlinn D, et al. Vegan: community ecology package. 2017.
- 26.
Wickham H. ggplot2: elegant graphics for data analysis. Springer-Verlag New York; 2009.
- 27. Love MI, Huber W, Anders S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 2014;15: 550. pmid:25516281
- 28. Regnery RL, Spruill CL, Plikaytis BD. Genotypic identification of rickettsiae and estimation of intraspecies sequence divergence for portions of two rickettsial genes. J Bacteriol. 1991;173: 1576–1589. pmid:1671856
- 29. Sprong H, Wielinga PR, Fonville M, Reusken C, Brandenburg AH, Borgsteede F, et al. Ixodes ricinus ticks are reservoir hosts for Rickettsia helvetica and potentially carry flea-borne Rickettsia species. Parasit Vectors. 2009;2: 41. pmid:19732416
- 30. Stuen S, Nevland S, Moum T. Fatal cases of Tick-borne fever (TBF) in sheep caused by several 16S rRNA gene variants of Anaplasma phagocytophilum. Ann N Y Acad Sci. 2003;990: 433–4. pmid:12860670
- 31. Courtney JW, Kostelnik LM, Zeidner NS, Massung RF. Multiplex real-time PCR for detection of Anaplasma phagocytophilum and Borrelia burgdorferi. J Clin Microbiol. 2004;42: 3164–8. pmid:15243077
- 32. Skotarczak B, Wodecka B, Cichocka A. Coexistence DNA of Borrelia burgdorferi sensu lato and Babesia microti in Ixodes ricinus ticks from north-western Poland. Ann Agric Environ Med. 2002;9: 25–8. pmid:12088393
- 33.
World Organisation for Animal Health, Cossio MLT, Giesen LF, Araya G, Pérez-Cotapos MLS, VERGARA RL, et al. Manual of diagnostic tests and vaccines for terrestrial animals. In: OIE. 2015 pp. 1–25. https://doi.org/10.1007/s13398-014-0173-7.2
- 34. Berri M, Laroucau K, Rodolakis A. The detection of Coxiella burnetii from ovine genital swabs, milk and fecal samples by the use of a single touchdown polymerase chain reaction. Vet Microbiol. 2000;72: 285–293. pmid:10727838
- 35. Forsman M, Sandström G, Sjöstedt A. Analysis of 16S ribosomal DNA sequences of Francisella strains and utilization for determination of the phylogeny of the genus and for identification of strains by PCR. Int J Syst Bacteriol. Microbiology Society; 1994;44: 38–46. pmid:8123561
- 36. Tomassone L, Nuñez P, Ceballos LA, Gürtler RE, Kitron U, Farber M. Detection of “Candidatus Rickettsia sp. strain Argentina” and Rickettsia bellii in Amblyomma ticks (Acari: Ixodidae) from Northern Argentina. Exp Appl Acarol. 2010;52: 93–100. pmid:20186466
- 37. Barbieri ARM, Romero L, Labruna MB. Rickettsia bellii infecting Amblyomma sabanerae ticks in El Salvador. Pathog Glob Health. 2012;106: 188–189. pmid:23265378
- 38. Ogrzewalska M, Literak I, Cardenas-Callirgos JM, Capek M, Labruna MB. Rickettsia bellii in ticks Amblyomma varium Koch, 1844, from birds in Peru. Ticks Tick Borne Dis. 2012;3: 254–256. pmid:22809734
- 39. Miranda J, Mattar S. Molecular detection of Rickettsia bellii and Rickettsia sp. strain colombianensi in ticks from Cordoba, Colombia. Ticks Tick Borne Dis. 2014;5: 208–212. pmid:24378078
- 40. Troyo A, Moreira-Soto RD, Calderon-Arguedas ?lger, Mata-Somarribas C, Ortiz-Tello J, Barbieri ARM, et al. Detection of rickettsiae in fleas and ticks from areas of Costa Rica with history of spotted fever group rickettsioses. Ticks Tick Borne Dis. 2016;7: 1128–1134. pmid:27592065
- 41. Hecht JA, Allerdice MEJ, Krawczak FS, Labruna MB, Paddock CD, Karpathy SE. Development of a Rickettsia bellii- specific taqman assay targeting the citrate synthase gene. J Med Entomol. Oxford University Press; 2016;53: 1492–1495. pmid:27473178
- 42. Blanco CM, Teixeira BR, da Silva AG, de Oliveira RC, Strecht L, Ogrzewalska M, et al. Microorganisms in ticks (Acari: Ixodidae) collected on marsupials and rodents from Santa Catarina, Paran and Mato Grosso do Sul states, Brazil. Ticks Tick Borne Dis. 2017;8: 90–98. pmid:27769655
- 43. Jourdain E, Duron O, Barry S, Gonzalez-Acuna D, Sidi-Boumedine K. Molecular methods routinely used to detect Coxiella burnetii in ticks cross-react with Coxiella-like bacteria. Infect Ecol Epidemiol. Co-Action Publishing; 2015;5: 29230. pmid:26609691
- 44. Vilcins IME, Old JM, Deane E. Molecular detection of Rickettsia, Coxiella and Rickettsiella DNA in three native Australian tick species. Exp Appl Acarol. 2009;49: 229–242. pmid:19296229
- 45. Sumrandee C, Hirunkanokpun S, Doornbos K, Kitthawee S, Baimai V, Grubhoffer L, et al. Molecular detection of Rickettsia species in Amblyomma ticks collected from snakes in Thailand. Ticks Tick Borne Dis. 2014;5: 632–640. pmid:25027232
- 46. Dietrich M, Lebarbenchon C, Jaeger A, Le Rouzic C, Bastien M, Lagadec E, et al. Rickettsia spp. in seabird ticks from western Indian Ocean Islands, 2011–2012. Emerg Infect Dis. 2014;20: 838–842. pmid:24751287
- 47. Hayashi M, Watanabe M, Yukuhiro F, Nomura M, Kageyama D. A nightmare for males? A maternally transmitted male-killing bacterium and strong female bias in a green lacewing population. Bourtzis K, editor. PLoS One. R Foundation for Statistical Computing; 2016;11: e0155794. pmid:27304213
- 48. Varela-Stokes AS, Park SH, Kim SA, Ricke SC. Microbial communities in North American ixodid ticks of veterinary and medical importance. Front Vet Sci. 2017;4: 179. pmid:29104867
- 49. Moutailler S, Valiente Moro C, Vaumourin E, Michelet L, Tran FH, Devillers E, et al. Co-infection of ticks: the rule rather than the exception. PLoS Negl Trop Dis. 2016;10: 1–17. pmid:26986203
- 50. Sassera D, Beninati T, Bandi C, Bouman EAP, Sacchi L, Fabbi M, et al. Candidatus Midichloria mitochondrii’, an endosymbiont of the Ixodes ricinus with a unique intramitochondrial lifestyle. Int J Syst Evol Microbiol. 2006;56: 2535–2540. pmid:17082386
- 51. Epis S, Sassera D, Beninati T, Lo N, Beati L, Piesman J, et al. Midichloria mitochondrii is widespread in hard ticks (Ixodidae) and resides in the mitochondria of phylogenetically diverse species. Parasitology. 2008;135: 485–494. pmid:18205982
- 52. Mariconti M, Epis S, Gaibani P, Dalla Valle C, Sassera D, Tomao P, et al. Humans parasitized by the hard tick Ixodes ricinus are seropositive to Midichloria mitochondrii: is Midichloria a novel pathogen, or just a marker of tick bite? Pathog Glob Health. 2012;106: 391–6. pmid:23265610
- 53. Bazzocchi C, Mariconti M, Sassera D, Rinaldi L, Martin E, Cringoli G, et al. Molecular and serological evidence for the circulation of the tick symbiont Midichloria (Rickettsiales: Midichloriaceae) in different mammalian species. Parasites and Vectors. 2013;6: 1–7.
- 54. Glaser RL, Meola MA. The native Wolbachia endosymbionts of Drosophila melanogaster and Culex quinquefasciatus increase host resistance to West Nile virus infection. PLoS One. 2010;5. pmid:20700535
- 55. Hornok S, Kováts D, Csörgő T, Meli ML, Gönczi E, Hadnagy Z, et al. Birds as potential reservoirs of tick-borne pathogens: first evidence of bacteraemia with Rickettsia helvetica. Parasit Vectors. BioMed Central; 2014;7: 128. pmid:24679245
- 56. Martello E, Selmi M, Ragagli C, Ambrogi C, Stella MC, Mannelli A, et al. Rickettsia slovaca in immature Dermacentor marginatus and tissues from Apodemus spp. in the northern Apennines, Italy. Ticks Tick Borne Dis. 2013;4: 518–21. pmid:24120274
- 57. Amore G, Tomassone L, Grego E, Ragagli C, Bertolotti L, Nebbia P, et al. Borrelia lusitaniae in immature Ixodes ricinus (Acari: Ixodidae) feeding on common wall lizards in Tuscany, central Italy. J Med Entomol. 2007;44: 303–7. pmid:17427701
- 58. Pintore MD, Ceballos L, Iulini B, Tomassone L, Pautasso A, Corbellini D, et al. Detection of invasive Borrelia burgdorferi strains in north-eastern Piedmont, Italy. Zoonoses Public Health. 2015;62: 365–74. pmid:25220838
- 59. Poupon M-A, Lommano E, Humair P-F, Douet V, Rais O, Schaad M, et al. Prevalence of Borrelia burgdorferi sensu lato in ticks collected from migratory birds in Switzerland. Appl Environ Microbiol. 2006;72: 976–9. pmid:16391149
- 60. Machado-Ferreira E, Vizzoni VF, Balsemão-Pires E, Moerbeck L, Gazeta GS, Piesman J, et al. Coxiella symbionts are widespread into hard ticks. Parasitol Res. 2016;115: 4691–4699. pmid:27595990