Co-Occurrence of Francisella, Spotted Fever Group Rickettsia, and Midichloria in Avian-Associated Hyalomma rufipes

Abstract

The migratory behavior of wild birds contributes to the geographical spread of ticks and their microorganisms. In this study, we aimed to investigate the dispersal and co-occurrence of Francisella and spotted fever group Rickettsia (SFGR) in ticks infesting birds migrating northward in the African-Western Palaearctic region (AWPR). Birds were trapped with mist nests across the Mediterranean basin during the 2014 and 2015 spring migration. In total, 575 ticks were collected from 244 birds. We screened the ticks for the species Francisella tularensis, the genus Francisella, and SFGR by microfluidic real-time PCR. Confirmatory analyses and metagenomic sequencing were performed on tick samples that putatively tested positive for F. tularensis during initial screenings. Hyalomma rufipes was the most common tick species and had a high prevalence of Francisella, including co-occurrence of Francisella and SFGR. Metagenomic analysis of total DNA extracted from two H. rufipes confirmed the presence of Francisella, Rickettsia, and Midichloria. Average nucleotide identity and phylogenetic inference indicated the highest identity of the metagenome-assembled genomes to a Francisella-like endosymbiont (FLE), Rickettsia aeschlimannii, and Midichloria mitochondrii. The results of this study suggest that (i) FLE- and SFGR-containing ticks are dispersed by northbound migratory birds in the AWPR, (ii) H. rufipes likely is not involved in transmission of F. tularensis in the AWPR, and (iii) a dual endosymbiosis of FLEs and Midichloria may support some of the nutritional requirements of H. rufipes.

Genomic characterization of Francisella tularensis and other diverse Francisella species from complex samples

“…12,086,278) sequence reads per sample, with an average of 80% of the reads classified as Francisellaceae (range: 67.6-93.9%; S3 Table). Reads from all five samples mapped to >72% of the 759 FLE-specific probes (S4Table) and grouped into one Francisella MAG per sample, which ranged in size from 1.21-1.45Mb (S3 Table); these five MAGs clustered with other FLE genomes in the dendrogram (S2 Fig).Note that some initial enrichment results for tick samples D14IT15.2 and D14IT20 have been previously described[20] but they are included here for completeness and because additional analyses were conducted on these samples in this study.The Francisella reads generated from the five enriched complex tick samples facilitated high-resolution phylogenetic analyses of FLEs from different as well as the same tick species.Although we generated and included genomic data for FLEs from three additional tick species(D. albipictus, D. variabilis, H. rufipes), some of the general phylogenetic patterns (Fig 7) remain similar to those previously described [26, 27, 74]: all 14 FLE genomes group together in a monophyletic clade within the Francisellaceae phylogeny and belong to the genus Francisella (S2 Fig), the FLE in A. arboreus (F. persica) is basal to all other FLEs, FLEs from hard ticks (Family , with the caveat that D. albipictus and D. variabilis have not been previously analyzed with that system. However, our targeted genome enrichment approach provides significantly increased resolution among FLE samples because we can use much more of the shared core genome of these samples; the FLE phylogeny in Fig 7 is based upon 53,377 shared positions among the five FLE samples enriched in this study and the nine previously published 34 FLE genomes.…”

“…These five samples were DNA extractions of whole ticks or tick cell lines containing FLEs (S3 Hyalomma rufipes ticks both collected on the island of Capri, Italy on 1 May 2014 [20]. As previously described, DNA was extracted from these surface-sterilized whole ticks and detection of Francisella DNA in the extracts was determined by PCR [20]. Sample DALBE3 is a tick cell line of Dermacentor albipictus that was previously documented to contain an FLE [21];…”

“…rufipes [74] but a total of 1,922 SNPs were identified between the Francisella MAGs generated from the two H. rufipes samples enriched and sequenced in this study (D14IT15.2, D14IT20), which were collected on the same day at the same location from two different birds [20]. This increased genetic resolution, and the ability to generate targeted FLE genomic data directly from whole, wild-caught ticks, offers the opportunity for future detailed studies of the natural population structure of these FLEs within single tick species.…”

“…These five samples were DNA extractions of whole ticks or tick cell lines containing FLEs (S3 [19], DNA was extracted from these whole ticks using a 96-well DNeasy Tissue Kit (Qiagen, Valencia, CA, USA) modified for use with the QIAvac vacuum filtration system, and presence of FLE DNA in the resulting extracts was determined via PCR and targeted gene sequencing. Samples D14IT15.2 and D14IT20 were extracted from two separate Hyalomma rufipes ticks both collected on the island of Capri, Italy on 1 May 2014 [20]. As previously described, DNA was extracted from these surface-sterilized whole ticks and detection of Francisella DNA in the extracts was determined by PCR [20].…”

“…Loss and disruption of FPI CDSs in the FLE genomes is not surprising given that, in F.Midichloria contain intact biotin pathways that are assumed to compensate for the disruption of this pathway in their associated FLE counterparts[20,30]. Of note, previous attempts[20] to examine biotin gene content in metagenomic sequences of DNA extracts from samples D14IT15.2 and D14IT20 were unsuccessful due to low FLE sequence signal. It was recently demonstrated that the Coxiella-like endosymbiont (CLE) of the of the Asian longhorned tick (Haemaphysalis longicornis), but not the tick itself, contains an intact shikimate pathway that synthesizes chorismate.…”