Biology subjects

Foster, L. J.

Publications and source records attributed to Foster, L. J..

At least 19 recordsLinked to original sources

bioRxiv2026

WrapEvoFS enables auditable feature compression with regret-constrained representative locking

WrapEvoFS provides a deterministic, auditable rule for choosing one feature set from repeated stochastic searches. It retains candidates within a prespecified development-score gap, selects the most representative set by feature overlap, and records the decision before held-out evaluation. A corrected genetic objective preserves score ordering lost through zero truncation. In a fully nested four-class cancer analysis, all 30 locked signatures met the 0.01 tolerance, and 11 differed from the highest-scoring candidate; every changed selection increased mean feature overlap within the eligible pool. Paired differences in area under the receiver operating characteristic curve against recursive feature elimination with cross-validation all had 95% intervals crossing zero. No selector dominated matched comparisons on feature count and predictive performance, and one-time evaluation of frozen signatures at held-out centers yielded heterogeneous effects. WrapEvoFS standardizes selection within a fixed candidate bank under an empirical score-gap constraint, enabling reproducible reporting of performance, compression, and feature-agreement trade-offs without implying predictive superiority or stability under changes in participant composition.

bioinformatics↗

bioRxiv2026

Gap Junction Protein Connexin43 (Cx43) Promotes Glioma Invasion via the Annexin A2/MMP3 Signaling Axis

Glioblastoma is among the most aggressive primary brain tumors, with the invasion of surrounding tissue driven by coordinated remodeling of the extracellular matrix (ECM) by proteolytic enzymes, including the matrix metalloproteinase (MMP) family. The gap junction protein connexin43 (Cx43) has emerged as an unexpected promoter of this invasive phenotype. However, the downstream molecular mechanisms linking Cx43 to ECM proteolysis remain poorly understood. We previously demonstrated selective upregulation of MMP3 within the secretome of Cx43-overexpressing glioma cells, implicating Cx43 in the regulation of ECM protease activity. To identify the molecular intermediaries responsible, we carried out a proteomic survey of gap junction (GJ)-associated signaling networks. Amongst these, we identified a Cx43 protein-protein interaction involving annexin A2 (ANXA2), a known regulator of ECM proteases. SDS-PAGE/Western blot and confocal imaging demonstrated that elevated Cx43 expression drives selective recruitment of ANXA2 to gap junction membranes, forming a junctional signaling platform rather than increasing total ANXA2 abundance. This recruitment coincided with co-localization of S100A10, suggesting assembly of an intact ANXA2/S100A10 complex at gap junctions. Using ANXA2-mimetic peptides targeting the S100A10-binding domain, we demonstrate that disruption of the ANXA2/S100A10 complex leads to a significant reduction in MMP3 activity. Furthermore, in 3D spheroids grown in ECM-replicating gels, ANXA2 peptide treatment significantly reduced the invasive potential of C6-13 (Cx43-overexpressing) cells. Together, these data define a novel Cx43[->]ANXA2[->]S100A10[->]MMP3 signaling axis, advancing our understanding of ECM remodeling in glioblastoma and establishing the ANXA2/S100A10 complex as a promising therapeutic target for disrupting Cx43-driven glioblastoma progression.

cancer biology↗

bioRxiv2026

Cathepsin K Mediates the Formation of Potential Rheumatoid Arthritis-Relevant Cis- and Trans-Spliced Peptides Compatible With HLA-DR4 Presentation

Rheumatoid arthritis (RA) is characterized by a loss of immunological tolerance to synovial self-proteins, yet the initial triggers generating novel neo-antigens remain incompletely defined. Here, we demonstrate that human cathepsin K (hCatK), a key cysteine protease driving joint degradation in RA, catalyzes covalent cis- and trans-splicing of peptides from major RA-associated self-proteins and foreign antigens, including type II collagen, fibrinogen, and SARS-CoV-2 Spike protein. Using high-resolution LC-MS/MS and database-assisted de novo sequencing, we identified over 90 unique spliced peptides. Splicing efficiency peaked at near-neutral pH (6.5-7.5), contrasting with classic hydrolytic profiles. Biochemical profiling revealed strong subsite selectivity, with a striking enrichment for small, aliphatic and/or hydroxyl-containing residues (Gly, Thr, Ser) at the P1 position. Furthermore, splicing preferentially targeted flexible, intrinsically disordered protein regions, with 81% of fibrinogen splicing events clustering within its C domain. In silico binding predictions for the RA-susceptibility allele HLA-DRB1*04:01 harboring the shared epitope revealed that numerous hCatK-generated spliced peptides exhibit predicted affinities exceeding those of established immunogenic and genomic sequences, uncovering protease-mediated transpeptidation as a novel post-translational modification capable of generating potent MHC class II autoantigens in RA.

biochemistry↗

bioRxiv2026

Cardiomyocyte pyruvate dehydrogenase kinase 1 knockout activates Rho-mediated remodeling and decreases fatty acyl availability in female mice

Pyruvate dehydrogenase kinase 1 (PDK1) inhibits pyruvate dehydrogenase (PDH) and therefore may regulate the balance between glucose and fatty acid oxidation in cardiomyocytes. We assessed metabolism and function in tamoxifen-inducible, cardiomyocyte-specific Pdk1 knockout (Pdk1cKO) mice, including both males and females fed a high fat diet, to resolve the role of PDK1 in the heart. Female, but not male, Pdk1cKO hearts showed the expected increase in cardiac PDH activity. Though cardiac function was largely preserved, female Pdk1cKO hearts had smaller left ventricles. Proteomics and phosphoproteomics revealed no change in classical hypertrophic markers but downregulated aerobic respiration alongside upregulated Ca2+ handling and Rho signalling, suggesting metabolic remodeling with altered cardiomyocyte contractility and cytoarchitecture in Pdk1cKO hearts. Concordantly, lipidomics revealed membrane remodeling marked by elevated saturated phosphatidylcholine in female Pdk1cKO hearts and decreased triacylglycerols, free fatty acids and fatty acid uptake proteins (CD36, FABP3). These findings reveal novel, female-predominant cardiometabolic roles for PDK1.

molecular biology↗

bioRxiv2026

Honey bee (Apis mellifera) gut microbiome associations with viruses and pesticides across Canadian agroecosystems

The honey bee (Apis mellifera) gut microbiome plays a central role in host health, yet its variation across agricultural landscapes remains poorly resolved. This study investigates how major environmental stressors, particularly pesticide exposure and RNA virus loadings, shape the honey bee gut microbiome in a large-scale field study conducted across Canada, spanning diverse agroecosystems from British Columbia to Quebec. We identify consistent associations between specific bacterial taxa and major RNA viruses, including enrichment of Serratia marcescens with SBV and depletion of Bombella intestini with BQCV. Pesticide exposure is likewise linked to reproducible shifts in key microbial taxa. Together, these findings reveal that interacting stressors jointly shape the bee gut microbiome and enable prediction of microbiome responses in agroecosystems. HighlightsDistinct associations identified between gut bacteria and major bee RNA viruses (BQCV, SBV, LSV, IAPV) Pesticide exposure is linked to reproducible shifts in key microbial taxa Combined virus-pesticide effects form coordinated clusters that predict microbiome variation and specific bacterial responses Integrated modeling demonstrates that environmental stressors can jointly explain microbiome structure beyond crop effects Graphical abstractSchematic overview of potential links between pesticide exposure and RNA virus infection and their effects on the bee gut bacterial community. Solid arrows indicate associations supported by the present study, whereas dashed arrows indicate hypothesized or unresolved interactions. Associations between the presence of specific bee RNA viruses (left) or pesticide residues (right) and changes in the relative abundance of particular gut taxa (pink {uparrow}, increased; blue {downarrow}, decreased). The pesticide subtype is indicated by the icon in the cell (leaf - herbicide, hyphae - fungicide and insect - insecticide). Several bacterial taxa showed reproducible associations with specific viral or pesticide variables, including Bombella intestini, Serratia marcescens, Melissococcus plutonius, Paenibacillus alvei, Apibacter sp. wkB309, and Gilliamella sp. A7. Abbreviations: BQCV Black queen cell virus; LSV, Lake Sinai virus; SBV, Sacbrood virus; IAPV, Israeli acute paralysis virus. (p/n/b) indicate the sample matrix in which the pesticide was detected, namely pollen, nectar, and bee tissue, respectively. O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=133 SRC="FIGDIR/small/731697v1_ufig1.gif" ALT="Figure 1"> View larger version (40K): org.highwire.dtl.DTLVardef@cb92a4org.highwire.dtl.DTLVardef@1087045org.highwire.dtl.DTLVardef@102cabforg.highwire.dtl.DTLVardef@4ce2f1_HPS_FORMAT_FIGEXP M_FIG C_FIG

microbiology↗

bioRxiv2026

Comparative Proteomics Across Tissues and Crop Agroecosystems Reveals Agricultural Stressor Responses in the Western Honey Bee

Maintaining honey bee health in crop production systems is increasingly difficult because worker bees encounter multiple chemical and biological pressures from pesticides and pathogens. How these field-realistic pressures affect molecular physiology across functionally distinct tissues remains poorly understood. Here, we tested whether tissue-resolved proteomics could separate stable tissue-specific patterns from crop-associated molecular changes. To do this, we profiled abdomen, gut, and head proteomes from honey bees collected across four Canadian crop ecosystems over two consecutive years, and integrated these data with pesticide-residue and pathogen-load measurements. Proteomic variation was structured by both tissue identity and crop environment. Tissue-specific proteomic profiles were characterized across samples, whereas crop-associated effects were detected in both years and were stronger in 2021, the second year of the study. Tissue-specific enrichment and network analyses linked the abdomen to lipid catabolism and ubiquitin-proteasome proteostasis, the gut to central carbon metabolism, membrane transport, vesicle trafficking, and cytoskeletal organization, and the head to neurosensory and mitochondrial functions, together with amino-sugar metabolism and vesicle-associated quality-control modules. Among the measured pesticide residues, boscalid was the most reproducible chemical correlate of proteomic variation, with the strongest signal in the gut. Cross-year validation associated boscalid exposure with reduced abundance of gut proteins involved in mitochondrial metabolism, protein quality control, vesicle trafficking, nutrient transport, and biosynthetic pathways. Additionally, integrated proteome-transcriptome-microbiome factor analysis further identified gut-centered components associated with measured stressor variables and linked protein-level variation to coordinated transcriptomic and microbial shifts. Independent-year validation showed that compact crop-associated protein signatures detected in 2020 were also present in 2021. Together, these results show that honey bee tissues maintain stable proteomic identities while showing tissue- and year-specific responses to pesticide and pathogen pressures encountered in crop ecosystems. The gut proteome may specifically provide a sensitive molecular indicator of pesticide-associated perturbation under field conditions.

bioinformatics↗

bioRxiv2026

Recovering biological structure in sparse single-cell proteomics with GIRAFI

Single-cell proteomics (SCP) based on liquid-chromatography mass-spectrometry resolves protein-level cellular heterogeneity, but interpretation remains limited by detection-linked sparsity. SCP profiles continuous, peptide-derived intensities and has lower throughput than single-cell RNA sequencing, making denoising methods for large-scale, count-based transcriptomics difficult to apply. Here we present GIRAFI, a graph-informed statistical learning framework that imputes missing values and reveals reproducible cell states by constraining inference to dataset-aware, prior-knowledge-informed protein neighborhoods. We evaluated GIRAFI across SCP datasets spanning diverse biological/technical contexts. In masking-based recovery experiments and cell-type-specific protein-protein interaction inference, GIRAFI outperformed existing methods, and matched bulk proteomics comparisons corroborated recovery accuracy and ablations supported the graph-informed design. Beyond reduced replicate- and source-associated technical structure, GIRAFI recovered ground-truth cell-type annotations, improved cell state-resolved pathway analysis, and enabled trajectory inference consistent with known time courses. These results establish graph-constrained imputation as an effective strategy for improving SCP robustness, biological structure, interpretation, and cross-dataset comparability.

bioinformatics↗

bioRxiv2026

Bridging Simplicity and Depth in Single-Cell Proteomics: A Cost-Effective Workflow and Expanded Framework for Data Evaluation

Single-cell proteomics (SCP) offers direct insight into functional protein states that drive cellular heterogeneity, complementing genomic and transcriptomic analyses. Although recent reports have demonstrated improved proteome coverage, their reliance on specialized instrumentation limits broader adoption. Additionally, current evaluation practices remain largely centered on protein and peptide identification counts, which alone do not fully reflect data quality or biological interpretability. Here, we describe an accessible, label-free SCP workflow which implements easily accessible laboratory equipment: a single-cell dispenser, conventional multiwell plates, and an incubator with water-bath-based humidity control. Using trapped ion mobility spectrometry-time-of-flight mass spectrometry (timsTOF), we systematically optimize key sample preparation variables, including trypsin concentration, incubation time, reduction/alkylation, digestion conditions, and plate types, which together maximize data quality and reproducibility. We further introduce a data-quality framework that moves beyond identification counts, emphasizing quantitative consistency and biological interpretability via individual protein coverage completeness across cells, coefficients of variation across technical replicates, peptide-to-protein ratios, and single-cell-to-bulk correlations. Collectively, our approach lowers technical barriers to accessing SCP while enabling more rigorous, interpretable, and scalable SCP analysis across diverse research contexts.

biochemistry↗

bioRxiv2025

SPIN: Inkjet-Driven Nanowell Workflow for Scalable and Sensitive Single-Cell Proteomics

Single-cell proteomics is emerging as a powerful approach to resolve cellular heterogeneity, yet sample processing remains challenging due to limited input material and the absence of protein amplification. We present a protocol centered on an image-guided, machine-learning-driven inkjet single-cell printer integrated with a dew-point-controlled nanowell chip to reduce loss, increase throughput, and improve reproducibility. The system dispenses single cells at >1 Hz into sealed nanoliter wells with minimal surface contact, virtually eliminating evaporation; a high-thermal-conductivity aluminum substrate and precise environmental control further ensure exceptional reproducibility. Relative to a commercial dispenser, the workflow yields significantly higher protein and peptide recovery without bias toward high-abundance species, delivering uniformly deep coverage. Biological pathway analysis emphasizes the robustness of this workflow, as there is a near 100% protein completeness detected among the enzymes in the Kreb cycle in both A549 and astrocytes, suggesting the consistency across all samples evaluated. This platform addresses core processing bottlenecks and enables reliable, scalable single-cell proteomics.

systems biology↗

bioRxiv2025

Framework for analyzing MAE-derived immunopeptidomes from cell lines with shared HLA haplotypes

BACKGROUNDThe goal in vaccinology is to identify candidate antigens for clinical trials that will elicit an immune response for a significant portion of the target population. Unfortunately, promising data generated at the preclinical level often cannot be replicated in larger sample sizes. The goal of this project was to develop a methodology for processing MAE-generated data to identify MHC epitopes, minimize non-specific contaminants, find binding motifs, and utilize genetic connections among donors to determine which peptides were presented by specific MHC alleles. RESULTSOur approach demonstrated that mild acid elution of peptides from seven consanguineous B-lymphocyte lines accurately reflects the HLA genotypes within family members, highlighting the specificity of MAE. Additionally, the data successfully reproduced known MHC binding motifs and partially deconvoluted the originating HLA alleles of the epitopes. CONCLUSIONSThese findings suggest that our approach could be applied to numerous cell lines globally to evaluate a wide array of HLA haplotypes. This may help to reveal candidate vaccine antigens that induce immune protection for a wider population.

immunology↗

bioRxiv2025

Beyond Degradation: How Reverse Proteolysis Creates Disease-Relevant Antigens

Proteases are conventionally regarded as degradative enzymes, yet their catalytic machinery also permits peptide bond formation through reverse proteolysis, a process that remains poorly characterized. Here, we show that lysosomal cysteine cathepsins catalyze iterative cycles of hydrolysis and ligation to generate multi-generational fusion peptides, including hybrids derived from host-viral protein substrates. Quantitative analysis demonstrates that peptide ligation can account for up to 4.5% of proteolytic turnover. This activity is strongly influenced by pH, substrate sequence, and post-translational modification, with citrullination and neutral pH favoring fusion peptide formation and the accumulation of more stable higher-order products. Using full-length protein substrates, we provide direct evidence that cathepsins can generate a hybrid insulin peptide previously identified as a Type 1 Diabetes (TID) autoantigen in patients. Moreover, several identified fusion peptides show effective binding to TID-associated HLA class II molecules. To examine whether ligation products can be captured under cellular conditions, we developed a click-based targeted transpeptide retrieval and purification strategy (CT-TRAP), which enabled detection of probe-derived cis/transpeptides in cell-based systems under controlled conditions. These findings establish reverse proteolysis by cysteine cathepsins as a quantifiable enzymatic pathway for generating non-genomically templated peptides, revealing an unrecognized dimension of lysosomal protease activity and peptide diversification.

biochemistry↗

bioRxiv2025

Virus infection of honey bee queens alters lipid profiles and indirectly suppresses a retinue pheromone component via reducing ovary mass

Virus infections reduce honey bee (Apis mellifera) ovary mass (in part due to resource-allocation trade-offs with immunity) and are linked to the presence of supersedure cells in colonies, indicating that workers are attempting to replace their queen. Pheromones, lipids, and lipid transport proteins could mediate relationships among virus infection, ovary mass, and supersedure. When we infected honey bee queens in the laboratory and profiled their queen retinue pheromone (QRP) components from head extracts, we found that virus infections specifically reduced the QRP component methyl oleate. Data from an observational field study were consistent with this pattern. Lipidomics analysis of the same extracts suggests that virus infection decreases triacylglycerol abundances (major sources of stored energy). Reducing ovary mass via laying restriction was sufficient to lower methyl oleate abundance [-] suggesting that virus infection reduces methyl oleate indirectly via ovary effects [-] but was insufficient to reduce abundance of most triacylglycerols or stimulate immune effector expression. Abundance of circulating apolipophorin-III, a lipid transporter and putative potentiator of immune effectors, was lower in queens with restricted laying, suggesting that its expression may be controlled by nutrient availability while its immune-stimulating capacity is governed by other mechanisms. Prior research has shown that queen pheromone blends lacking methyl oleate are less attractive to workers; therefore, diminishing methyl oleate could result in a less desirable pheromone bouquet and possibly stimulate supersedure. The mechanism of methyl oleate reduction is yet to be determined, but is clearly tied to ovary size and possibly resource availability or mobilization.

physiology↗

bioRxiv2025

Western red cedar (Thuja plicata) beehives have no impact on honey bee (Apis mellifera) overwintering colony survival or detoxification enzyme expression

In North America, wooden honey bee hives are most often constructed from pine, but some companies also produce and sell boxes made of western red cedar (Thuja plicata) as a result of its local availability and desirable properties. However, there is debate within the beekeeping community about whether cedar is a safe hive material for bees, since resins within the wood are known to be insecticidal or insect deterrents. There is very little empirical evidence to support or refute these arguments. Here, we recorded health metrics of honey bee nucleus colonies hived in western red cedar and pine boxes (n = 10 each) to determine if the type of wood affects colony outcomes. Colonies were produced and introduced into these boxes in late July, with monitoring continued until the following spring. We found no significant differences in adult bee populations, brood areas, or Varroa mite prevalence among colonies hived in cedar versus pine boxes at either the end of summer (September 1st) or spring (April 1st) assessments. Overwintering survival was identical in the two groups at 90%. Hemolymph detoxification enzyme expression differed strongly between callow (day-old) workers and foragers but did not differ with hive material. Overall, we did not find evidence that hiving honey bee colonies in boxes constructed of western red cedar has any negative or positive effect on bee physiology or colony outcomes.

physiology↗

bioRxiv2024

Language model-guided anticipation and discovery of unknown metabolites

Despite decades of study, large parts of the mammalian metabolome remain unexplored. Mass spectrometry-based metabolomics routinely detects thousands of small molecule-associated peaks within human tissues and biofluids, but typically only a small fraction of these can be identified, and structure elucidation of novel metabolites remains a low-throughput endeavor. Biochemical large language models have transformed the interpretation of DNA, RNA, and protein sequences, but have not yet had a comparable impact on understanding small molecule metabolism. Here, we present an approach that leverages chemical language models to discover previously uncharacterized metabolites. We introduce DeepMet, a chemical language model that learns the latent biosynthetic logic embedded within the structures of known metabolites and exploits this understanding to anticipate the existence of as-of-yet undiscovered metabolites. Prospective chemical synthesis of metabolites predicted to exist by DeepMet directs their targeted discovery. Integrating DeepMet with tandem mass spectrometry (MS/MS) data enables automated metabolite discovery within complex tissues. We harness DeepMet to discover several dozen structurally diverse mammalian metabolites. Our work demonstrates the potential for language models to accelerate the mapping of the metabolome.

bioinformatics↗

bioRxiv2024

Pyruvate dehydrogenase kinase 1 controls triacylglycerol hydrolysis in cardiomyocytes

Pyruvate dehydrogenase kinase (PDK) 1 is one of four isozymes that inhibit the oxidative decarboxylation of pyruvate to acetyl-CoA via pyruvate dehydrogenase. PDK activity is elevated in fasting or starvation conditions to conserve carbohydrate reserves. PDK has also been shown to increase mitochondrial fatty acid utilization. In cardiomyocytes, metabolic flexibility is crucial for the fulfillment of high energy requirements. The PDK1 isoform is abundant in cardiomyocytes, but its specific contribution to cardiomyocyte metabolism is unclear. Here we show that PDK1 regulates cardiomyocyte fuel preference by mediating triacylglycerol turnover in differentiated H9c2 myoblasts using lentiviral shRNA to knockdown Pdk1. Somewhat surprisingly, PDK1 loss did not affect overall PDH activity, basal glycolysis, or glucose oxidation revealed by oxygen consumption rate experiments and 13C6 glucose labelling. On the other hand, we observed decreased triacylglycerol turnover in H9c2 cells with PDK1 knockdown, which was accompanied by decreased mitochondrial fatty acid utilization following nutrient deprivation. 13C16 palmitate tracing of uniformly labelled acyl chains revealed minimal acyl chain shuffling within triacylglycerol, indicating that the triacylglycerol hydrolysis, and not re-esterification, was dysfunctional in PDK1 suppressed cells. Importantly, PDK1 loss did not significantly impact the cellular lipidome or triacylglycerol accumulation following palmitic acid treatment, suggesting that effects of PDK1 on lipid metabolism were specific to the nutrient-deprived state. We validated that PDK1 loss decreased triacylglycerol turnover in Pdk1 knockout mice. Together, these findings implicate a novel role for PDK1 in lipid metabolism in cardiomyocytes, independent of its canonical roles in glucose metabolism.

biochemistry↗

bioRxiv2024

Honey bee egg composition changes seasonally and after acute maternal virus infection

Honey bee (Apis mellifera) colonies depend on the reproductive output of their queens, which in turn is contingent on the care provided by worker bees. Viral infections in queens can compromise reproductive output, while worker infections can inhibit the successful functioning of a colony and its ability to care for the queen. Transgenerational immune priming (TGIP) occurs when queens transfer immune-related compounds or immune elicitors to their offspring, enhancing the ability of subsequent generations to resist infections. These maternal effects on offspring could positively impact colony health and resilience to viral infections, but little is currently known about TGIP for viruses. In this study, we investigate how viral infections affect the proteomic composition of eggs laid by virus-challenged queens (injected with a mixture of black queen cell virus and deformed wing virus B), both in controlled experimental settings and natural field conditions. Our results showed that virus-challenged queens upregulated immune effectors in their eggs and ovaries. In contrast, naturally infected queens from field surveys did not; there were no significant differences in egg protein, lipid, or metabolite composition related to maternal viral load or ovary size. However, egg collection date strongly influenced egg composition, likely reflecting seasonal variations in pollen resources. These findings highlight that while viral infections can induce transgenerational effects on egg proteomes under short-term experimental conditions, such effects are less apparent in natural settings and can be overshadowed by seasonal and other ecological factors.

zoology↗

bioRxiv2024

Proteomic Investigation of Neurotrophic trans-Banglene Reveals Potential Link to Iron Homeostasis

In an effort to gain insight into cellular systems impacted by neurotrophic trans-banglene (t-BG), global proteomic profiling and Western blot analyses were employed. Expression level changes in response to t-BG treatment were compared to those observed with nerve growth factor (NGF), a natural neurotrophic protein and functional analog to t-BG. Findings from these studies did not point to direct interception of NGF/TrkA signaling by t-BG. Instead, significant alterations in iron-binding and iron-regulating proteins were observed. Intracellular iron measurements by FerroOrange indicate lower ferrous (Fe2+) iron levels in t-BG treated cells but not in NGF treated cells. These results highlight a potential connection between iron regulation and neurotrophic activity.

molecular biology↗

bioRxiv2024

Pollen foraging mediates exposure to dichotomous stressor syndromes in honey bees

Recent declines in the health of honey bee colonies used for crop pollination pose a considerable threat to global food security. Foraging by honey bee workers represents the primary route of exposure to a plethora of toxins and pathogens known to affect bee health, but it remains unclear how foraging preferences impact colony-level patterns of stressor exposure. Resolving this knowledge gap is crucial for enhancing the health of honey bees and the agricultural systems that rely on them for pollination. To address this, we carried out a national-scale experiment encompassing 456 Canadian honey bee colonies to first characterize pollen foraging preferences in relation to major crops, then explore how foraging behaviour influences patterns of stressor exposure. We used a metagenetic approach to quantify honey bee dietary breadth and found that bees display distinct foraging preferences that vary substantially relative to crop type and proximity, and the breadth of foraging interactions can be used to predict the abundance and diversity of stressors a colony is exposed to. Foraging on diverse plant communities was associated with increased exposure to pathogens, while the opposite was associated with increased exposure to xenobiotics. Our work provides the first large-scale empirical evidence that pollen foraging behaviour plays an influential role in determining exposure to dichotomous stressor syndromes in honey bees. Significance StatementInsect-mediated pollination is an important ecological process that is crucial for food production. Managed honey bee colonies are one of the most important insect pollinators, but their health has been under threat from a variety of stressors. Bee workers are primarily exposed to stressors while foraging and understanding how bee foraging preferences are related to exposure risk could provide pivotal information to improve management efforts. Here, we studied honey bee foraging preferences in relation to prominent Canadian crops and across a gradient of modified environments. We found that honey bees show distinct, measurable foraging preferences and that dietary diversity is a strong predictor of the stressors that colonies are exposed to.

ecology↗