Wolbachia

Quantitative PCR (qPCR) was performed to characterize the relative Wolbachia infection level in the S2 cell lines and flies. The protocol was similar to prior qPCR amplification using the single-copy wsp and su(fk)C genes of bacterial and host origin, respectively [65 (link)]. S2 cells were quantified using a hemocytometer to obtain 10⁶ cells. The S2 cells or DSR females were homogenized in 100 μl STE with 0.4 mg/ml proteinase K to extract DNA as previously described [66 (link)].
For qRT-PCR, RNA extractions were performed on groups of 10 ovaries or 10 testes dissected from one-day post eclosion infected and uninfected Drosophila adults using the RNeasy Mini Kit (Qiagen). DNA contamination was removed with RNase-Free DNase Set (Qiagen). RNA quality and quantity was checked with NanoDrop ND-100 spectrophotometer (NanoDrop Technologies, Inc.). Synthesis of cDNA was performed with Superscript II Reverse Transcriptase (Invitrogen) using specific primer for Ance (AnceQ F 5'-CGGTCACGTTCGATTGGTTG-3' and AnceQ R 5'-CTTCGGTTTCCACGTTGGTTC-3') and Actin gene (ActinQ F 5'-GCTGACCGTATGCAAAAGG-3' and ActinQ R 5'-GCTTGGAGATCCACATCTG-3'). Primers were designed based upon D. simulans genbank sequences for Ance and Actin (genbank accession number: NM_057696 and NM_079486, respectively]. qRT-PCR was performed separately with the AnceQ F/R and ActinQ F/R primer pairs using a Miniopticon system (BioRad) with a Platinum SYBR Green qPCR superMix (Invitrogen). qRT-PCR reactions were conducted using a 2 minute step at 50°C, 2 minute step at 95°C and 40 cycles of 15 seconds at 95°C and 30 seconds at 56°C. A fluorescence measurement was made at the end of each cycle. A melting curve analysis was performed at the end of the amplification program to examine for primer-dimers or nonspecific amplification. Assays were performed on two (D. simulans and D. melanogaster wild type) or three (D. melanogaster Ance mutants) independent experiment replicates for each sex and infection type. As an examination for variability, duplicate qRT-PCR reactions were performed for each set of ovaries or testes with both the Ance and Actin primers. Relative expression of Ance gene was calibrated against Actin using the ΔΔC_T calculation method [67 (link)] with:

Expression variation=2−ΔΔCT
For comparisons of males and females, the above was modified as follows:

Large volume serum samples were available from six (3 male and 3 female) domestic short-haired cats experimentally infected at 10 months of age with Dirofilaria immitis by subcutaneous inoculation of third-stage larvae (L₃) and confirmed to be infected by recovery of adult worms at necropsy or confirmation of histological lesions. Briefly, trickle infection of a total of 100 L₃ of D. immitis was performed by subcutaneous inguinal inoculation of each cat a total of four times, on study days 7, 14, 21, and 28, with 25 L₃ (Missouri strain) harvested shortly prior to inoculation from infected Aedes aegypti mosquitoes (Liverpool strain). Whole blood samples were collected from the jugular vein or, rarely, the cephalic vein, of each cat on days 84, 112, 140, 168, 196, and 224 directly into vacuum tubes containing either EDTA or no additive. Cats were cared for through Oklahoma State University’s Association for Assessment and Accreditation of Laboratory Animal Care-accredited animal resources program throughout the study; all research procedures followed a detailed animal care and use protocol approved by Oklahoma State University’s Institutional Animal Care and Use Committee. Anti-coagulated whole blood was assayed for microfilaria by modified Knott’s test and by real-time PCR for Wolbachia spp. as previously described [11 ,12 (link)]. For tubes without additive, blood was allowed to clot, the serum separated by centrifugation, placed into aliquots, and stored at −80°C until further use.

Each new sequence is first filtered for quality, a process that excludes any record with less than 500 bp coverage for the barcode region of COI or with more than 1% ambiguous bases. If a sequence meets these quality requirements, it is then checked for reading frame shifts as indicated by stop codons or improbable peptides given the COI profile [44] (link). Because sequences showing these attributes are likely to derive from pseudogenes, they are excluded. Sequences are then screened to ensure that they do not derive from bacterial (e.g. Wolbachia) or certain external (e.g. human, mouse) contaminants by matching the sequence recovered from each specimen against a reference library of bacterial and selected vertebrate sequences. Finally, when a sequence record originates from the assembly of two or more shorter sequences, the Bellerophon package [47] (link) is utilized to check for possible chimeras that would arise if the component sequences inadvertently (e.g. contamination, laboratory error) derived from two different taxa.

To estimate the abundance of the different endosymbionts in O. gibbosus individuals we used a dataset consisting of 16 short-read (Illumina) sequencing libraries obtained from male O. gibbosus individuals from six different sampling locations (Damvallei, n=4; Sevendonck, n=2; Pollismolen, n=2; Honegem, n=2; Overmeren, n=2; Walenbos, n=4) described previously [37 (link)]. Detailed information about the sequencing libraries is provided in Table S1. Illumina reads were right-tail clipped using a minimum quality of 20, and all reads shorter than 75 after this treatment were discarded using FASTX-Toolkit (http://hannonlab.cshl.edu/fastx_toolkit/). Reads with undefined nucleotides or left without a pair were removed using PRINSEQ [48 (link)]. Afterwards, the reads were mapped to the genome of O. gibbosus (GCA_019343175.1) using bbmap (v37.61; ‘-minid=99’) (sourceforge.net/projects/bbmap/). The unmapped reads were then mapped to the endosymbiont genomes, and the mapped reads to the single-copy gene elongation-factor alpha (EF1alpha) from O. gibbosus [37 (link)] (JTE90_028130) using Bowtie2 (v2.3.5.1) [44 (link)]. The median coverage for each endosymbiont genome and the mean coverage for EF1alpha was calculated per sample using samtools (v1.12) [49 (link)]. For ‘Ca. Tisiphia’, Cardinium and Rhabdochlamydia the coverage across the whole genome was used, while for Wolbachia A and Wolbachia B only the coverage of the respective accessory genes (n=319 Wolbachia A, n=353 Wolbachia B) was used. We used the coverage of the accessory genes for Wolbachia as the two genomes share several genes which would bias the abundance estimations as short reads from Wolbachia A would also map to Wolbachia B and vice versa. To calculate the abundance of each endosymbiont per sample, we normalized the median coverage of the endosymbionts by the mean coverage of EF1alpha. To get the number of endosymbionts by host cell we divided the abundance by the number of host chromosomes (n=2).

For the phylogenetic reconstructions of Wolbachia and Cardinium we downloaded representative genomes belonging to different groups from NCBI (Table S8). In the case of Wolbachia, Anaplasma phagocytophilum strain HZ and Ehrlichia canis strain Jake were used as outgroups, and for Cardinium, Amoebophilus asiaticus strain 5a2 was used. rRNA proteins were extracted, and universally conserved ones selected. The protein sequences were individually aligned using MAFFT (v7.453; ‘--localpair’ ‘--maxiterate 1000’) [50 (link)]. Divergent and ambiguously aligned blocks were removed using Gblocks (v0.91b) [51 (link)]. The alignments were then concatenated using custom perl scripts and Bayesian inference was performed using MrBayes (v3.2.7) [52 (link)], using the JTT+I+G4 substitution model. We ran two independent analyses with four chains each for 300 000 generations and checked for convergence (convergence diagnostic≤0.01).
For Rickettsiaceae we also downloaded representative genomes from NCBI and used Megaira sp. strain MegNEIS296, Megaira sp. strain MegCarteria, Occidentia massiliensis strain Os18, Orientia tsutsugamushi strain Boryong and Orientia chuto strain Dubai as outgroups (Table S8). We extracted rRNA proteins and selected universally conserved ones. The protein sequences were individually aligned using muscle (v3.8.31) [53 (link)] and the alignments were concatenated using AMAS [54 (link)]. The phylogeny was calculated with IQ-TREE 2 (v2.1.2; ‘-bnni’ ‘-alrt 1000’ ‘-m TESTNEW’ ‘--madd LG4X’ ‘-bb 1000’) using the JTTDCMut+F+R3 substitution model [55 (link)]. In order to verify if the endosymbionts belong to an existing clade or represent a new one, we calculated the average amino acid (AAI) and average nucleotide identity (ANI) for the endosymbiont and selected representative genomes using the method and thresholds described previously [56 (link)]. We did not exclude transposase genes for the ANI and AAI calculations. While a large amount of transposase genes might influence the ANI value, the AAI is regarded to be unbiased as only homologous genes are used for the calculation.

To reconstruct the endosymbiont genomes, we used long-read (PacBio; W744xW766) and short-read (Illumina; W791, W815) sequencing libraries from the host O. gibbosus described previously [37 (link)]. Detailed information on the sequencing libraries is provided in Table S1. As references for the long-read alignments, we used metagenome-assembled genomes (MAGs) from the endosymbionts that were reconstructed from the short-read libraries. In addition, we used the contigs from the host-assembly that were classified as endosymbiont contigs using a custom Kraken database as described previously [37 (link)].
For Cardinium sp. cOegibbosus-W744×776, Wolbachia sp. wOegibbosus-W744×776A and Wolbachia sp. wOegibbosus-W744×776B we mapped long reads to the respective MAGs and contigs using minimap2 (v2.17) [38 (link)]. Finally, all mapped reads were merged, and duplicates were removed. As the coverage was too high for Wolbachia, the mapped reads were subsampled to a coverage of 70×. The final set of reads was then assembled using Unicycler (v0.4.6) [39 (link)]. The quality of the assemblies was checked by checkM (v1.0.18) [40 (link)] and visually inspecting the assembly graph [41 (link)]. The Wolbachia genomes were closed manually, and the assembly of Wolbachia sp. wOegibbosus-W744×776B was manually curated to remove misassembled regions.
For Candidatus Tisiphia sp. Oegibbosus-W744×776 we mapped long reads to the respective MAG and contigs using minimap2 (v2.17) [38 (link)] and short reads using bbmap (v37.61) (sourceforge.net/projects/bbmap/). Both long and short reads were merged and deduplicated afterwards. The final set of long and short reads was then used for a hybrid assembly using Unicycler (v0.4.6) [39 (link)]. The quality of the assembly was checked by checkM (v1.0.18) [40 (link)] and visually inspecting the assembly graph [41 (link)].
Rhabdochlamydia oedothoracis MAG OV001 was assembled using the paired-end (2×150 bp) short-read library ‘OV001_G’ (SRX9657682, see details of sequencing libraries used in this study in Table S1). ‘OV001_G’ is the ID of the individual used for sequencing, with ‘OV001’ being unique to the sequencing library, and thus used for strain designation. We first carried out a de novo assembly using MEGAHIT (v1.1.2) [42 (link)]. Afterwards, the contigs were binned using MetaBAT 2 (v2.15) [43 (link)]. The required bam file for binning was created using bowtie2 (v2.3.5.1) [44 (link)]. Finally, the qualities of all bins were checked with checkM (v1.0.18) [40 (link)]. The bin belonging to Rhabdochlamydia was used for a reassembly. For that purpose, we mapped the short-read library from Overmere_OV001 to the Rhabdochlamydia bin and the reference genome R. oedothoracis W744xW776 (GenBank accession number: CP075587-CP075588) using bbmap (v37.61) (sourceforge.net/projects/bbmap/). Afterwards all mapped reads were merged, and duplicates were removed. This set of reads was then used for a second round of de novo assembly and binning as described before. The quality of the final MAG was again checked with (v1.0.18) [40 (link)].
Origins of replication were detected using originx [45 (link)] and Ori-Finder [46 (link)]. The orientation of genome sequences was adjusted to previously sequenced Wolbachia and Rickettsia genomes [47 (link)].