An alternative food source for metabolism and longevity studies in *Caenorhabditis elegans*

Caenorhabditis elegans is a widely used model for genetics, development, metabolism, and aging because of its short life cycle, self-reproduction, and conserved biology with humans. In laboratory settings, worms are typically fed live Escherichia coli (OP50), whose active metabolism can confound studies of worm metabolism and drug effects, as bacteria can metabolize compounds and alter experimental outcomes. Prior observations, such as metformin’s effects mediated through bacterial metabolism, underscore this concern. There is therefore a need for a metabolically inactive bacterial food source that prevents bacterial replication and respiration while maintaining bacterial structure for consumption by worms. The authors hypothesize that paraformaldehyde (PFA) treatment can reproducibly inactivate bacterial replication and metabolism while keeping bacteria edible, enabling more accurate measurements of worm metabolism, behavior, development, fecundity, and lifespan.

Several bacterial inactivation methods have been used in C. elegans research: ultraviolet (UV) irradiation, heat-killing, and antibiotic treatment. UV irradiation is common but low throughput and inconsistent; treated plates often require individual verification for bacterial growth, and metabolic inactivity is rarely confirmed. Heat-killed bacteria can destroy nutrients and reduce palatability, leading to developmental arrest. Antibiotics can prevent replication but may leave bacteria metabolically active and themselves can affect worm growth and metabolism. Due to these limitations, a reliable, high-throughput method that ensures both replicative and metabolic inactivity while preserving bacterial structure is needed.

Overview: The study optimized paraformaldehyde (PFA) treatment of E. coli OP50 and assessed effects on bacterial replication and metabolism and on C. elegans physiology and metabolome. Comparators included live, mock-treated (washed without PFA), and UV-treated OP50. Additional bacteria (HB101 and Enterococcus faecalis) were also tested.

Bacterial culture and treatments:

• Live condition: Single colony inoculated into lysogeny broth (LB), cultured overnight (18 h) at 37 °C shaking. Bacteria pelleted (3000 × g, 20 min), concentrated 5× or 10×, and seeded on NGM plates.
• UV treatment (OP50): Plates seeded with live bacteria dried 24 h, then irradiated in a UV crosslinker (CL-600 UVP) at 9999 × 100 µJ/cm² for 5 or 10 min.
• PFA treatment: 50 mL bacterial culture aliquoted into 250 mL Erlenmeyer flasks. 32% PFA added to reach final concentrations (e.g., 0.25% by adding 390 µL to 50 mL). Shaken at 37 °C, 200 rpm for 1 h. Transferred to 50 mL conical tubes, centrifuged (3000 × g, 20 min), supernatant removed, and washed with 25 mL LB five times to remove residual PFA. Final resuspension at 5× (10 mL) or 10× (5 mL) concentration. Seeded onto NGM and dried 48 h before use. Mock-treated controls underwent identical washing and concentration without PFA.

Bacterial replication and metabolism assays:

• Viable plate count: 10× seeded bacteria from NGM plates collected with M9, pelleted, resuspended to 10×, serially diluted (10-fold series), 50 µL spread on LB agar, incubated at 37 °C overnight, CFUs counted and back-calculated.
• Seahorse XF96 respirometry: For basal OCR, 20 µL resuspended bacteria added to 180 µL M9 (final 1×) in wells; OCR measured at 37 °C. For glycolytic stress, 20 µL bacteria into 112 µL M9; injections: glucose (100 mM, 20 µL, Port A), oligomycin (20 µM, 22 µL, Port B), 2-deoxy-glucose (500 mM, 25 µL, Port C). OCR and ECAR measured with standard Mix/Wait/Measure cycles.

Behavioral assays:

• Food preference (pairwise sensitized assay): 60 mm NGM plates with two bacterial lawns (50 µL each) placed 1 cm apart and 2 cm from center. Synchronized L1s (bleach hatch) placed centrally; worm counts on lawns after 24 h at 20 °C.
• Chemoattraction assays: Plates seeded with 1, 2, 4, or 8 lawns (20 µL each), 2 cm from center; synchronized populations placed in center; worms counted on lawns after 24 h.
• Chemotaxis assay: Young adults pre-cleared from food 30 min, then placed on plates with equidistant LB vs PFA-containing LB or live vs PFA-treated OP50 spots (4 µL). Worms counted at 30 and 60 min; chemotaxis index calculated.

Physiological assays:

• Fecundity: 6-h egg lays on test food; L4s singled (10 per condition) to 35 mm plates; daily transfers for 7 days; F1 progeny counted; censored if animals left plate.
• Development: ~1000 bleach-prepped eggs plated on 100 mm NGM with test foods; after ~2.5 days, worms washed to remove bacteria; COPAS Biosorter used to analyze L4s by time-of-flight (TOF). Development time from egg to egg-laying quantified.
• Microscopy and size: Synchronized worms imaged at young adult, day 2, day 4 adults; area quantified via ImageJ; immobilization with 30 mM sodium azide.
• Lifespan: Synchronized cohorts (~75 worms/plate; ≥2 plates/condition/replicate) maintained at 20 °C; transferred daily for 7 days; scored every 2–3 days; death defined by lack of response to prodding; crawlers censored. Three independent experiments performed.

Metabolomics:

• Sample prep: ~1000 eggs plated per condition; collected at late L4; washed with M9 and 150 mM ammonium acetate; pellets flash-frozen.
• Extraction: 500 µL ice-cold 9:1 methanol:chloroform; probe sonication 30 s; incubate on ice 5 min; centrifuge 10 min at 4000 × g, 4 °C; supernatant to vials.
• LC-MS: HILIC-LC-ESI-MS (negative mode) on Agilent 1200 + 6220 TOF; Phenomenex Luna NH2, 150 × 1.0 mm, 0.07 mL/min, 10 µL injection. Untargeted peaks detected/aligned with XCMS. Data processed in MetaboAnalyst 4.0: median normalization, autoscaling; PCA and PLS-DA. Targeted panel of 95 metabolites quantified with MassHunter; normalization by pooled sample, log transform, range scaling; one-way ANOVA with Tukey’s HSD; significant metabolites used for pathway analysis (nematode pathway library; KEGG IDs required).

Statistics: One-way ANOVA with Tukey post-hoc for fecundity and attraction assays; log-rank tests for lifespan; error bars typically SEM (SD in specified figure).

• Optimal PFA treatment: 0.5% PFA for 1 h consistently prevented bacterial growth and respiration in OP50, while preserving structure for worm feeding.
• Replicative death: Viable plate counts showed UV-treated OP50 exhibited variable inhibition of replication across replicates, whereas 0.25% and 0.5% PFA-treated OP50 consistently showed no replication. Mock-treated and live had similar CFUs, controlling for washing effects.
• Metabolic inactivity: Seahorse respirometry indicated minimal OCR and ECAR in both UV- and PFA-treated OP50 compared with live and mock-treated controls; UV showed inconsistent basal OCR between replicates. PFA treatment also eliminated respiration in HB101 and Enterococcus faecalis.
• Table 1 summary (growth in CFU; OCR relative to mock-treated):
  • 0% PFA, 24 h: Growth ~10^7; OCR High.
  • 0.25% PFA, 30 min: Growth ~10^5; OCR High.
  • 0.25% PFA, 1 h: Growth ~10^2; OCR Low.
  • 0.25% PFA, 2 h: Growth 0; OCR None.
  • 0.5% PFA, 30 min: Growth ~10^2; OCR Low.
  • 0.5% PFA, 1 h: Growth 0; OCR None.
  • 0.5% PFA, 2 h: Growth 0; OCR None.
  • 0.75% PFA, 30 min: Growth 0; OCR Low.
  • 0.75% PFA, 1–2 h: Growth 0; OCR None.
• Metabolomics: PCA separated worm metabolomes by bacterial treatment. PLS-DA showed extensive alterations: among 3284 features, 1349 had VIP >1 in component 1 comparing PFA-treated vs controls. Pairwise VIP >1 features: live vs mock (1211), mock vs PFA (1426), live vs UV (1172), indicating high sensitivity to bacterial condition changes. Targeted analysis (95-metabolite panel) showed 25 significantly altered metabolites (0.25% PFA vs mock) and 27 (0.5% PFA vs mock). Enriched pathways included phenylalanine, tyrosine and tryptophan biosynthesis, phenylalanine metabolism, aminoacyl-tRNA biosynthesis, and arginine and proline metabolism. Lysine abundance was significantly reduced in worms fed PFA-treated OP50, consistent with formaldehyde crosslinking chemistry.
• Food attraction and preference: Worms remained on PFA-treated lawns (no repulsion to residual PFA). In pairwise preference assays, worms preferred live and mock-treated OP50 over PFA-treated OP50 (p-values ranging from <0.05 to <0.0001), consistent with loss of chemo-attractants in non-metabolizing bacteria.
• Development: PFA-treated bacteria caused a small but significant developmental delay (~4–5 hours from egg to egg-laying) and reduced L4 size (lower TOF), but worms caught up in size by young adulthood; sizes at day 2 and day 4 adults were comparable across conditions.
• Fecundity: No significant difference in brood size between worms on live vs PFA-treated OP50; a small but significant decrease compared with mock-treated OP50.
• Lifespan: Across three independent experiments, PFA-treated conditions generally did not significantly alter lifespan compared with live or mock-treated bacteria. Table 2 highlights: Experiment 1 median 22 days across conditions with no significant differences; Experiment 2 showed a significant difference for 0.5% PFA vs live (p=0.005) and vs mock (p=0.020), while other comparisons were not significant; Experiment 3 showed no significant differences. Overall, PFA treatment did not substantially affect lifespan.
• Generalizability: PFA treatment reliably eliminated replication and respiration in both nonpathogenic (OP50, HB101) and pathogenic (E. faecalis) bacteria.

The study addressed the need for a reliable, high-throughput method to eliminate bacterial metabolic activity in C. elegans studies, which is critical for accurate metabolomics and drug response measurements. Paraformaldehyde treatment at 0.5% for 1 hour consistently abrogated bacterial replication and respiration, outperforming UV irradiation in reproducibility and throughput, and avoiding the nutritional and palatability issues of heat-killed bacteria and the confounds of antibiotics. Although worms display a preference for live over PFA-treated bacteria, they still consume PFA-treated food without major adverse effects; only a minor developmental delay was observed, brood size was largely unaffected relative to live, and lifespan was generally unchanged. Metabolomic analyses revealed extensive changes when the food source is altered, including PFA treatment, indicating that any modification to bacterial conditions can have systemic effects on the worm metabolome. This underscores the importance of controlling and standardizing bacterial preparation in metabolic and pharmacological studies. By maintaining bacterial structural integrity while ensuring metabolic inactivity, PFA-killed bacteria provide a practical and effective food source for experiments focused on worm-intrinsic metabolism, drug effects, and longevity, reducing confounding from bacterial metabolism.

This work establishes paraformaldehyde-killed bacteria—optimally 0.5% PFA for 1 hour, followed by thorough washing—as a consistent, scalable, metabolically inactive food source for C. elegans. The method eliminates bacterial replication and respiration across multiple bacterial strains while preserving edibility, producing only a small developmental delay without substantive impacts on fecundity or lifespan. Metabolomic profiling highlights that worm metabolism is highly sensitive to bacterial condition, emphasizing the value of standardized, metabolically inert food for metabolic and pharmacological studies. Future work includes applying this PFA-based approach to high-throughput drug and toxicity assays and further dissecting how specific bacterial-derived metabolites influence worm physiology.

• Altered worm metabolome: PFA treatment of bacteria induces widespread changes in the worm metabolome, comparable in scope to other food preparation changes (UV, washing). These systemic shifts must be considered when interpreting metabolomics data.
• Food preference: Worms prefer live/mocked-treated bacteria over PFA-treated food, likely due to loss of bacterial chemo-attractants; this could influence behavioral outcomes in certain assays.
• Developmental delay: A modest 4–5 hour delay in development was observed on PFA-treated food, which may affect timing-sensitive experiments.
• Lifespan variability: While overall lifespan effects were minimal, one experiment showed a significant difference for 0.5% PFA vs live and mock-treated conditions, suggesting potential context-dependent effects that warrant replication and careful control.
• Method specificity: Findings were generated under specific culture conditions (e.g., strain OP50, treatment parameters); while HB101 and E. faecalis results support generalizability, other bacterial strains or growth conditions may require re-optimization.