Transitions in symbiosis: evidence for environmental acquisition and social transmission within a clade of heritable symbionts

The study investigates how transmission modes shift within symbiont clades, focusing on Arsenophonus associated with the Western honey bee (Apis mellifera). While many bacterial symbionts of arthropods, including Arsenophonus, are considered primarily vertically transmitted, transitions to or inclusion of horizontal routes can reshape host–microbe interactions, influencing dependence, virulence, and genome evolution. Eusociality in honey bees presents unique ecological and behavioral contexts—high host density, relatedness, homeostatic nest environments, trophallaxis, and cooperative brood care—that can strongly shape symbiont transmission. The research question is whether Arsenophonus in honey bees is maintained by vertical transmission or instead follows an infectious, horizontally acquired lifestyle, and how this relates to broader transitions in symbiosis within the Arsenophonus clade. The purpose is to resolve the epidemiology and transmission ecology of the honey bee–Arsenophonus association, providing insight into evolutionary transitions in symbiotic lifestyles.

Background literature outlines a continuum of symbiotic relationships, from facultative to obligate and from vertical to horizontal transmission. Heritable symbionts like Wolbachia and Arsenophonus are predominantly vertically transmitted, though horizontal transfer can occur over evolutionary timescales. Exclusively horizontal symbionts often rely on environmental reservoirs or social contacts and may exhibit looser alignment with host fitness. Transitions in transmission are linked to changes in virulence and genome evolution. Prior work within Arsenophonus shows diverse lifestyles—reproductive parasitism, facultative mutualism, and obligate endosymbiosis with genomic reduction—yet documented strains typically combine vertical transmission with occasional horizontal transfer. Notable comparative cases include the emergence of Coxiella burnetii from maternally inherited endosymbionts, and the presence of opportunistic human-infective Sodalis, which inform how infectious lifestyles can arise from heritable ancestors. Honey bees have been previously reported to harbor Arsenophonus and associations with poor health have been suggested, but transmission routes and epidemiology in this eusocial host remain unclear.

Phylogenomics and comparative genomics: A draft genome of the honey bee-associated Arsenophonus (bioproject PREJB39047) was assembled from Illumina paired-end reads derived from haemolymph purified via Nycodenz gradient (per Gauthier et al.). Contigs were annotated with Prokka v1.14.0 and completeness assessed using BUSCO v4.1.4 with 124 universal bacterial markers. Phylogenetic placement within Arsenophonus was inferred using alignments of 53 ribosomal proteins and a concatenated set of 155 single-copy orthologous proteins, with Proteus mirabilis and Providencia stuartii as outgroups. Predicted metabolic potential was compared across Arsenophonus strains, with detailed functional and synteny comparisons to Arsenophonus nasoniae. Spatial and seasonal dynamics: Adult workers were collected from 159 colonies across 45 apiaries in ten counties in England, sampled April–November from 2014–2018, yielding 230 sampling events (some colonies repeatedly sampled). Bees were preserved in 70% ethanol at −20 °C. For each colony, posterior legs from 12 workers were pooled (groups of four), UV-crosslinked (10 min), soaked in 1 M KOH (60 min), and DNA extracted via high-throughput Chelex. Extraction quality was verified by EF1-α amplification; Arsenophonus presence was assayed by PCR targeting rpoB (adapted from Duron et al.) with Sanger sequencing for strain identity. Assay sensitivity was validated by serial dilution over two orders of magnitude. Overwintering persistence: Twenty-five colonies (from seven apiaries) with known autumn status were tracked into spring. In autumn, 15 workers per colony (five pools of three) were screened: colonies with ≥80% positive pools (≥4/5) formed infected cohort A (n=19); those with all negative pools formed cohort B (n=6). In spring, surviving colonies were resampled (24 bees per colony; eight pools of three). DNA was extracted from whole bees (heads removed) using Promega Wizard purification; Arsenophonus detected by PCR. Localisation in the gut: Whole guts from live workers were dissected and fixed in Carnoy’s fixative (24–48 h), washed in 100% ethanol, incubated in hybridization buffer (~15 h), and probed by FISH with an Arsenophonus-specific oligonucleotide (CTAGTACACCAACTCCCA) labeled with 5' Alexa Fluor 647. Post-hybridization washes were performed at ~48 °C. Samples were DAPI-counterstained, mounted, cured (24 h), and imaged by confocal microscopy (ZEISS LSM 88, 40x), assembling Z-stacks in ImageJ. Guts from Arsenophonus-negative colonies served as negative controls. Additionally, material samples from workers (n=22; pooled two per sample) were extracted and tested by PCR for Arsenophonus. Assessment of heritability and acquisition timing: Brood frames from field colonies (N=8; Arsenophonus-positive A+ = 6, negative A− = 2) were sampled across life stages: eggs (n=29), larvae (n=67), pupae (n=49), newly emerged workers (NEWs; n=36), adult workers (n=45), and drones where possible (n=22). To detect latent heritable association emerging later, an additional 64 NEWs were isolated with age-matched NEWs only for two weeks (25 ± 2 days post-emergence) before screening. DNA extractions used Qiagen DNeasy for eggs/early larvae and Promega Wizard for other stages; Arsenophonus detection by PCR as above. Maintenance off-colony: To test infection persistence absent the colony/foraging environment, adult workers were collected from two infected colonies (total N=76; colony A=36, B=40) with high Arsenophonus prevalence (>90% in 15 tested workers) and from colonies with no detectable infection (N=32). On day 1, infection status was determined via PCR on posterior tarsus tissue (leg snip). A no leg-snip control group from infected colonies (N=40) was included to check for procedural bias. Marked individuals were maintained in groups of four; subsamples were culled on days 4, 8, and 15, and infection status reassessed using the opposite tarsus. Horizontal transmission assays: Donor adult workers from six colonies with high Arsenophonus prevalence (>85% of 15 adults infected) were mixed with uninfected NEW recipients in 340 ml pots (10 donors + 5 recipients per pot). Two treatments were imposed: (1) general contact (social interactions permitted with shared sucrose feed), and (2) trophallaxis-only, where donors had access to 50% sucrose and recipients were separated by fine mesh to force feeding via trophallaxis while preventing other contact. Interactions lasted 5 days; survival status (alive/dead) at endpoint was recorded. Controls paired uninfected donors with recipients. DNA was extracted individually from whole bees (test n=300; control n=165) and screened by PCR. Statistical analysis: GLMs/GLMMs with binomial errors and logit link were fitted in R 3.3.1 using glm and lme4. Model selection used AIC; significance of fixed and random effects was evaluated by likelihood ratio tests. Overdispersion was checked using Blmeco. Supplementary tables detail model selection.

- Phylogenomics: The honey bee-associated Arsenophonus falls robustly within the Arsenophonus clade and shares genomic architecture and predicted metabolic features with the male-killing symbiont Arsenophonus nasoniae. - Transmission mode: Multiple independent lines of evidence indicate deviation from a heritable, vertically transmitted model. Arsenophonus was rarely detected in early life stages (eggs, larvae, pupae, newly emerged workers), arguing against maternal vertical transmission. - Spatial and seasonal epidemiology: Field sampling across 159 colonies from 45 apiaries (230 sampling events, 2014–2018) revealed spatial and seasonal fluctuations in prevalence, inconsistent with stable vertical maintenance. - Infection loss: Rapid loss of infection was observed both at the colony level (across the overwinter period in tracked colonies; 25 colonies from seven apiaries) and at the individual level when infected bees were removed from colonies and maintained in the laboratory for up to 15 days, indicating dependence on ongoing exposure for maintenance. - Localization: FISH localized Arsenophonus to the gut, consistent with an environmentally or socially acquired enteric association; negative controls from uninfected colonies lacked signal. - Horizontal transmission: Direct acquisition by uninfected recipients from infected donors occurred under both general social contact and trophallaxis-only conditions in the lab (test n=300; controls n=165), demonstrating efficient social transmission pathways. - Overall: Evidence supports a mixed acquisition pathway in honey bees, combining environmental exposure and social transmission, rather than vertical inheritance.

The findings address whether Arsenophonus in honey bees is vertically inherited or infectiously acquired. The rarity of detection in eggs and early developmental stages, seasonal and spatial variability in colony prevalence, infection loss off-colony and over winter, gut localization, and experimentally demonstrated social transmission together support a predominantly horizontal transmission ecology. In a eusocial context, behaviors such as trophallaxis and close contact provide efficient routes for spread. These results challenge assumptions that all insect-associated Arsenophonus are maintained primarily by vertical transmission and illustrate a transition within a clade best known for heritable symbioses. This has implications for symbiont evolution: reduced alignment with host fitness can facilitate different selective pressures on virulence, and mixed environmental/social transmission may influence genome dynamics differently from strict vertical inheritance. The system offers a comparative framework to study how transmission mode shifts arise and are maintained in social hosts and how they shape symbiont–host coevolution.

This study demonstrates that the Arsenophonus associated with Western honey bees is primarily horizontally acquired via environmental exposure and social interactions (including trophallaxis), rather than vertically inherited. The strain is phylogenomically allied to A. nasoniae and localizes to the gut. Field and laboratory data reveal dynamic prevalence, infection loss when off-colony or over winter, and direct social transmission. These results provide a key example of a transition in symbiotic lifestyle within a clade of typically heritable symbionts and establish honey bees as a valuable model to study how sociality shapes transmission modes. Future work should identify specific environmental reservoirs and transmission foci, quantify transmission rates in natural settings, assess fitness effects on hosts and symbionts, resolve genome features underlying transmissibility, and evaluate generality across regions, seasons, and other eusocial hosts.

- Geographic and temporal scope: Sampling was restricted to apiaries in England across 2014–2018, which may limit generalizability to other regions or years. - Detection methods: Reliance on PCR-based detection from specific tissues and pooling strategies can miss low-titer infections and does not quantify bacterial load. - Environmental sources unresolved: While environmental acquisition is inferred, specific reservoirs and pathways in nature were not directly identified. - Health impact not tested: The study did not directly measure fitness or health effects on bees associated with infection dynamics. - Genomic status: Analyses were based on a draft genome assembly; complete genome closure could refine metabolic and synteny inferences.