Microbiome characterization
We used both a culture-based microbial analysis followed by Sanger sequencing and DNA metabarcoding to quantify the microbial communities associated with household arthropods captured by the citizen scientists (Fig. 1). For samples collected between August 2018 and July 2019, we characterized the composite microbiota associated with the arthropods’ morphospecies, including exogenous (exoskeleton) and the endogenous (gut) microbiota together. For a subset of the samples (July to October 2019), we analysed the exogenous and the endogenous microbiota separately. This allowed us to study whether the exogenous microbiota composition varied seasonally and co-varied with environmental variables (e.g., temperature and humidity).
The endogenous and exogenous microbiota were collected from pools of up to 5 arthropods from the same morphospecies to capture the bacterial communities’ diversity. The exogenous microbiota was collected by washing and vortexing the arthropod pools (2850 rpm for 30 seconds) in sterile distilled water (SDW) and pelleting the bacterial community through centrifugation at 13000g for two minutes in an Eppendorf centrifuge (Hamburg, Germany). Following the collection of the exogenous microbiota, the pools were washed in 70% ethanol for 2 minutes, rinsed with SDW, and homogenised by grinding in 1 ml of SDW using a sterile micro pestle (Eppendorf, Hamburg, Germany), to collect the endogenous (gut) microbiota. Both the total microbiota extracted from samples collected between August 2018 and July 2019 and the microbiota collected separately from endogenous and exogenous microbiota (July to October 2019), were split into two aliquots; the first aliquot was serially diluted down to 10-4 of the initial concentration and plated on Nutrient Agar, Mannitol Salt Agar, Violet Red Bile Glucose Agar and Blood Agar N° 2 (Thermo Fisher Scientific Oxoid Ltd, Basingstoke, UK); the second aliquot was used for DNA metabarcoding (Fig. 1).
All the inoculated bacteria on agar media were incubated at 37°C for 48h in aerobic conditions, whereas Blood agar plates were incubated both in aerobic and anaerobic conditions. Following incubation, bacterial colonies were counted, corrected for the number of arthropods in each pool and the serial dilution factor and expressed as the number of Colony Forming Units (CFU/ml/arthropod). A total of 166 colonies across the collected samples was sequenced using Sanger sequencing technology, following gDNA extraction with QIAamp DNA mini kit (Qiagen, Hilden, Germany), and amplification of the 16S rRNA Vi-V2 region [27F (5’-AGAGTTTGATCATGGCTCA-3’) and 1492R (5’-TACGGTTACCTTGTTACGACTT-3’)]. The PCR conditions were as follows: 95°C for 15 minutes, followed by 35 cycles at 95°C for 1 minute, 50°C for 1 minute, and 72°C for 1 minute. A final extension at 72°C for 10 minutes was used. The PCR products were purified using QIAquick PCR Purification Kit (Qiagen, Hilden, Germany) and quantified on a 1.2% agarose gel stained with ethidium bromide (Bio-Rad Laboratories, Hercules, California, USA). The Sanger sequencing was completed by Eurofins Genomics, Ebersberg, Germany.
Genomic DNA for the metabarcoding analysis was extracted from the microbiota community of each arthropod pool using the DNeasy PowerWater kit (Qiagen, Venlo, Netherlands) following filtration on 0.2 μm filter funnels (Thermo Fisher Scientific, Waltham, USA). Extracted DNA was then prepared in paired end 250bp amplicon libraries obtained with the 16S RNA V1-V2 region, using a 2 step PCR protocol with 96x96 dual tag barcoding to facilitate multiplexing and to reduce crosstalk between samples in downstream analyses (MacConaill et al. 2018). Negative controls were used for PCR biases and contamination. They consisted of genomic libraries constructed using the SDW used to wash the arthropods and of libraries without target DNA. PCR1 and PCR2 primers, as well as annealing temperatures per primer pair in PCR1 are in Table S1. QCs were performed at each PCR step. Excess primer dimers and dinucleotides from PCR1 were removed using Thermostable alkaline phosphatase (Promega) and Exonuclease I (New England Biolabs). PCR2 amplicons were purified using High Prep PCR magnetic beads (Auto Q Biosciences) and quantitated using a 200 pro plate reader (TECAN) using qubit dsDNA HS solution (Invitrogen). A standard curve was created by running standards of known concentration on each plate against which sample concentration was determined. PCR2 amplicons were mixed in equimolar quantities at a final concentration of 12 pmol using a Biomek FXp liquid handling robot (Beckman Coulter). Pooled libraries molarity was confirmed using a HS D1000 tapestation screentape (Agilent) prior to 250bp paired end sequencing on an Illumina MiSeq platform to obtain 100,000 reads per sample and amplicon. The genomic libraries were prepared and sequenced by EnviSion, Environmental BioSequencign and BioComputing at the University of Birmingham (https://www.envision-service.com/).
Data analysis
The sequences obtained with Sanger technology were analysed using the software FinchTv (version 1.4.0) and taxonomically assigned using Blast V 2.12.0+ using default parameters, following the removal of PCR primer sequences.
Metabarcoding sequences were analysed using qiime2 (Caporaso et al. 2010). Amplicon sequence variants (ASVs) were obtained by removing adapter primers with the cutadapt plug-in (Martin 2011), denoising with DADA2, trimming low quality reads, merging forward and reverse reads, dereplicating and filtering out chimaeras (Callahan et al. 2016). Taxonomic assignment of the 16S sequences was done using the DAIRY database (Meolaet al. 2019) , following the removal of unassigned reads with the microDecon R package (McKnight et al. 2019).
The R-package Vegan (Oksanen 2020) was used for statistical analysis of both the culture-based and the metabarcoding sequences. We applied a permutation analysis of variance (PERMANOVA) on the weighted Bray-Curtis distance for both metabarcoding and culture-based sequences, using the Adonis function in R (Team 2020) to quantify the effect of Arthropod family, season (month of sampling) and environment (urban/suburban), and their combined effects on the microbiota diversity.
We studied correlations between environmental variables and the compositional changes of the exogenous and endogenous microbiota. The environmental variables studied were average air temperature, precipitation, humidity, and wind speed obtained from the Stourbridge weather station (UK), located within a 20 Km range from the 20 households used in this study (metnet.co.uk). Significant associations between beta diversity (weighted diversity between samples) and environmental variables were established using a Spearman’s rank correlation test (P < 0.05).
The bacterial families identified through the culture-based and high throughput DNA metabarcoding were mapped onto the approved list of biological agents published by the Health and Safety Executive Advisory Committee on Dangerous Pathogens ((ACDP) 2020) to identify pathogens with potential adverse effect on human health.