4.1. WP1. Production, characterisation and standardisation of I+Se biofortified lettuce (PL)

4.1.1. Objective: Produce standardized, well-characterised batches of lettuce (Lactuca sativa L.) biofortified with iodine (I) and/or selenium (Se) for use across the project (for in vitro fermentation, organoid exposure, mouse diets). Ensure reproducible mineral content, safe levels, and availability of freeze-dried aliquots for cross-WP use. The PL research lab has extensive experience in the development of functional food products, including lettuce fortified with micronutrients, and the biofortification process will be conducted according to established procedures (Smoleń et al. 2016, Smoleń et al. 2019,  Sularz et al. 2021).

4.1.2. WP1 overview and responsibilities. PL (Polish team) is responsible for WP1 end-to-end: cultivation, biofortification treatments, harvest, processing (freeze-drying, lyophilization, homogenisation, powderization), chemical quantification (I, Se), and distribution of aliquots to CZ (fermentation, preparation of conditioned media) and PL (in vivo, molecular work). PL will coordinate QC and documentation.

4.1.3. Experimental design and groups (lettuce):  L1) Control (deliberate I/Se supplementation, background trace levels only), L2) I only (iodine biofortified), L3) Se only (selenium biofortified), L4) I + Se combined (biofortified with both). Applied varieties may include those rich in dietary fiber, such as romaine (cos) lettuce, which provides approximately 2.1 g of fiber per 100 g. Each treatment will be produced in 3 independent cultivation batches (biological replicates) to control for variability. From each batch produce ≥2 kg fresh leaves (target to obtain ~200–300 g freeze-dried powder per batch).

4.1.3.1. Agronomic setup (greenhouse/growth chamber). Substrate: standard peat-based horticultural substrate or defined potting mix as used in Smoleń et al., with uniform fertilisation background. Environmental control: 16 h photoperiod / 8 h dark; 20–22 °C day / 16–18 °C night; relative humidity 50–70% (documented). Plant density, irrigation, pest control: follow standard horticultural practices (as in Smoleń 2016, 2019). Randomised block design to avoid positional bias. Growth stage for treatment application: apply I/Se at stages used in previous studies (vegetative growth — Smoleń et al., methods).

4.1.3.2. Biofortification chemistry — compounds and dosing. Use forms previously validated in lettuce biofortification literature:
Iodine sources: Potassium iodate (KIO₃) or potassium iodide (KI). Smoleń et al. used KIO₃ at ~5 mg I·dm⁻³ in solution applications as an effective dose (Smoleń 2016). Choose KIO₃ for stability and lower phytotoxicity in some regimes.
Selenium sources: Sodium selenite (Na₂SeO₃) or sodium selenate (Na₂SeO₄). Smoleń et al. used Na₂SeO₃ at ~0.5 mg Se·dm⁻³ for foliar/solution protocols (Smoleń 2016). Use Na₂SeO₃ with caution — Se speciation will be checked.
Agronomic biofortification of vegetables—particularly lettuce—has proven highly effective in enhancing the accumulation and bioavailability of these micronutrients, e.g., through organically bound iodine and selenium, while preserving plant quality and safety (Smoleń et al., 2016a,b).

4.1.3.3. Doses: Start from literature doses: KIO₃ ~5 mg I·dm⁻³, Na₂SeO₃ ~0.5 mg Se·dm⁻³ (Smoleń 2016), adjust after pilot if phytotoxicity or insufficient uptake observed. Keep doses well below plant-toxic ranges; document all amounts applied (mg per plant, frequency, timing).

4.1.3.4. Application schedule (example). Pre-treatment baseline sampling of leaves (random plants) to measure background I/Se.

4.1.3.5. Application window: 2–3 applications across vegetative stage (e.g., week 3, week 4, week 5 post-transplant), or continuous low-dose in nutrient solution for hydroponic setups — follow pilot outcome.

4.1.3.6. Record environmental variables, soil EC, pH, irrigation volumes. 

4.1.3.7. Harvesting and processing. Harvest at defined maturity stage (as in Smoleń 2016) to ensure consistent biochemical background. Post-harvest: wash leaves thoroughly in deionized water (documented rinse steps), blot dry. Freeze immediately in liquid nitrogen or −80 °C freezer, then lyophilise (freeze-dry). Homogenise lyophilised material to fine powder under inert atmosphere (nitrogen) to avoid oxidation. Use cryo-milling if available. Aliquot powder into pre-weighed sterile tubes (e.g., 0.5–1 g per tube) for: a) fermentation substrate; b) organoid exposure; c) mouse diet mixing; d) chemical QC samples.

4.1.3.8. Chemical analysis and QC. Iodine and selenium quantification — ICP-MS or ICP-OES following acid digestion / appropriate sample prep (Smoleń 2016, 2019). Use certified reference material for plant matrices and internal standards. Record total I and total Se (µg/g dry weight).

4.1.3.9. Speciation (optional but recommended) — where feasible, perform Se speciation (SeMet, SeCys) by LC-ICP-MS or HPLC-ICP-MS to confirm organic incorporation (literature shows incorporation into organic forms improves bioavailability) (Smoleń 2019).

4.1.3.10. Accept/reject criteria — batches accepted if I and Se fall within predefined target ranges (set after pilot, e.g. target ±20% of nominal). Document variability (CV) across subsamples; require CV < 15% for acceptance.

4.1.3.11. Microbial contamination check — plate random powder aliquots on general media to check for contamination (mainly for organoid exposures). Sterile handling to produce fermentation substrate; but keep natural microbiota for in vivo diets (no sterilisation).

4.1.3.12.Nutritional composition — basic proximate analysis (moisture, protein, fiber) to ensure diets are macronutrient-matched.

4.1.3.13.Safety and regulatory checks. Ensure that I and Se levels in final mouse diets are within regulatory/nutritional safe margins for mice (consult institutional animal facility and literature for safe Se/I intake for C57BL/6J).

4.1.3.14.Phytotoxicity assessment: monitor plant growth, chlorosis, yield. If excessive phytotoxicity observed, adjust dosing.

4.1.3.15.Record all interventions, batch numbers, chain-of-custody and storage conditions (−80 °C for long-term).

4.1.3.16.Distribution and labelling. Create a tracking sheet with unique batch IDs; include: treatment, date of harvest, measured I and Se content, storage location, aliquot IDs, recipient WP.

4.1.3.17.Ship aliquots for fermentation (CZ) on dry ice with formal sample manifest. Retain QC aliquots at PL.

4.1.3.18. Replication: 3 independent cultivation batches per treatment; within each batch produce ≥10 plants to ensure material yield and biological replication.

4.1.3.19. Data recording and deliverables. Deliverables (D1.x). D1.1 Standardised batches of I only / Se only / I+Se / control lettuce (freeze-dried aliquots) with unique IDs and inventory sheet, D1.2 ICP-MS quantification report (I and Se, µg/g DW) + speciation report if performed, D1.3 QC report: yield, phytotoxicity observations, proximate composition (moisture, fibre), D1.4 Chain-of-custody manifest and sample distribution log (to WP2, WP3, WP5).

4.1.4. Personnel. Personnel: 1 PI (WP lead), 1 agronomist/technician (greenhouse ops), 1 analytical chemist (ICP-MS), 1 lab tech (processing/lyophilisation).

4.1.5. Timeline (example): pilot cultivation and method selection 2 months; scaled cultivation and harvest 2–3 months; processing + QC 1 month. Total WP1 duration: 4–6 months.

4.1.6. Methods and reagents (practical list).

4.1.6.1. Methods. Short SOP extract (example: ICP-MS sample prep). Weigh 0.2–0.5 g lyophilised plant powder into PTFE digestion vessel. Add 5 mL concentrated HNO₃ + 1 mL H₂O₂ (suited for plant matrices). Microwave digestion programme (ramp to 180 °C, hold 15 min) following instrument manual. Cool, transfer digest to volumetric flask, dilute to known volume with ultrapure water. Analyse by ICP-MS with internal standard; run blanks and CRMs. Procedures will be carried out according to the manufacturer’s instructions and the methods listed in the references below.

4.1.6.2. Contingency & pilot steps. Pilot trials to optimise I/Se form and dose (compare KIO₃ vs KI; Na₂SeO₃ vs Na₂SeO₄; foliar vs substrate) — choose method giving reproducible enrichment and minimal phytotoxicity per Smoleń et al. If substrate dosing reduces I uptake (reported interactions between I and Se), adjust to foliar or split dosing (literature: Smoleń 2016/2019 observed interactions between forms).

4.1.6.3. Reagents. KIO₃ (analytical grade) and/or KI — supplier: Sigma-Aldrich (catalogue numbers recorded in lab SOP). Na₂SeO₃ and/or Na₂SeO₄ — analytical grade. Cryo-milling / ball mill for homogenisation, lyophiliser (freeze dryer). ICP-MS (or outsourced service) with appropriate certified reference materials (NIST plant CRMs) and internal standards (e.g., ¹²³Te or ⁷⁵Se depending method). Acid digestion kit (HNO₃/H₂O₂ microwave digestion) for plant samples. Sterile containers, cryovials, desiccant, −80 °C storage.

4.1.6.4. Equipment and Materials.  Seeds/planting substrate & greenhouse. Lettuce seeds (biofortified and control) will be obtained from certified breeders to ensure genetic uniformity. Applied varieties may include those rich in dietary fiber, such as romaine (cos) lettuce. Plants will be grown in sterile, peat-based substrate with controlled nutrient composition in a greenhouse with automated climate control, LED lighting, fertigation systems for precise iodine and selenium delivery, and sensors for soil moisture, EC, and micronutrient levels. Biofortified and control plants will be spatially separated with independent irrigation lines and pre-sterilised equipment to prevent cross-contamination.

4.1.7. Quality control & reproducibility measures. SOPs for application, harvest, processing and storage (versioned). Interbatch controls: include an internal reference batch processed across runs to detect drift. Technical replicates for ICP-MS (triplicates) and use CRMs. Randomisation of plant position and blinded labelling for downstream analyses.

4.1.8. Ethical, safety, legal notes. Use of Se is potentially hazardous at high concentrations — ensure appropriate handling, PPE, waste disposal. All plant trials comply with institutional GMO/plant regulations (non-GMO, standard horticulture). For transport of biological material across labs use appropriate biosafety documentation (fermentation will use human faecal inocula later — separate ethics, see WP2/WP3).

4.1.9. Budget (see the project)

4.1.9.1. Seeds/planting substrate & greenhouse costs.

4.1.9.2. KIO₃, KI, Na₂SeO₃, Na₂SeO₄ reagents.

4.1.9.3. Lyophilisation and cryo-milling access (or purchase).

4.1.9.4. ICP-MS analysis (instrument time or outsourcing costs per sample).

4.1.9.5. Consumables, vials, packaging, shipping to CZ.

4.1.10.  References (methodology and SOP sources) at the end of the document.

4.2.1. Objective. The aim of WP2 is to determine how diets enriched with iodine- and/or selenium-biofortified lettuce modulate gut microbiota, SCFA production, micronutrient metabolism, organism-level physiological parameters in vivo.  WP2 will establish a rigorously controlled 8-week dietary intervention in C57BL/6N mice, followed by systematic collection of fecal samples (for ex vivo fermentation in WP3 and microbiome profiling in WP4) and tissues for downstream transcriptomic, epigenomic and multiomic analyses (WP5). WP2 ends strictly at biological material biobanking; no omics assays are performed in this Work Package. Microbiome-metabolome studies will be shown in the next WP below. Final plan and methods can be adjusted/selected based on WP1 results.

4.2.2. WP2 overview and responsibilities. The Polish team (PL) is responsible for all in vivo procedures, animal husbandry, dietary intervention, physiological monitoring, sample acquisition (feces, urine, blood, tissues), necropsy, and standardized biobanking. Fecal samples will be shipped to CZ for fermentation assays (WP3). Tissues will remain in PL for extraction and preparation for single-cell, nuclei, and bulk multiomics. All procedures follow FELASA recommendations, EU Directive 2010/63, and institutional IACUC approval.

4.2.3. Experimental animals. Female C57BL/6N mice, 4 weeks of age, sourced from a certified breeder (Charles River), specific pathogen-free (SPF), weight 13–15 g at arrival. Animals acclimate for 7 days before any intervention. All procedures comply with Polish Ministry of Agriculture Regulation of 14 December 2016 and EU Directive 2010/63/EU.

4.2.4. Housing and husbandry conditions. Animals will be housed in the Rodent Facility of OMEiI under controlled conditions: polycarbonate cages (37 × 21 × 15 cm), 8 mice per cage, temperature 22 ± 2 °C, humidity 55 ± 10 %, 12:12 light–dark cycle, 15 air exchanges per hour, access to normocaloric diet and autoclaved water ad libitum, environmental enrichment (bedding, nesting material, manipulable objects). Veterinary oversight is provided by Dr. Urszula Bracha, DVM.

4.2.5. Microbiota pre-standardisation and stress reduction. To reduce baseline inter-individual variation in gut microbiota and mitigate stress-induced confounders:

4.2.5.1. Handling habituation: 7 days of daily gentle handling using cupping rather than tail picking (Hurst & West 2010).

4.2.5.2. Microbiota optimisation: 21 days of group-specific cohousing with cage randomisation every 3–4 days, homogenising microbiota profiles via coprophagy (Ericsson et al. 2018).

4.2.5.3. Pre-intervention sampling: fecal pellets collected at the end of cohousing (T0) for baseline microbiota composition, SCFA levels, and baseline I/Se excretion.

4.2.6. Dietary intervention groups (8 weeks). Mice will be randomly assigned (random number generator, stratified by starting body weight) to seven dietary groups. All diets are based on a purified AIN-93M normocaloric formulation.

4.2.6.1. STD (baseline control): standard AIN-93M diet without lettuce or supplements.

4.2.6.2. SAL-CTR (control lettuce): diet supplemented with non-biofortified lettuce lyophilizate (L1 from WP1).

4.2.6.3. SAL-I+Se: diet supplemented with lettuce biofortified with both iodine and selenium (L4).

4.2.6.4. SAL-I: diet supplemented with iodine-biofortified lettuce (L2).

4.2.6.5. SAL-Se: diet supplemented with selenium-biofortified lettuce (L3).

4.2.6.6. INU (positive fibre control): diet containing inulin (5–10 % w/w) known to increase SCFA production (Roberfroid 2010).

4.2.6.7. BUT (butyrate mechanistic control): diet containing sodium butyrate (1–2 % w/w), a classical HDAC inhibitor (Davie 2003), serving as an epigenomic reference independent of microbiome modulation.

This design allows dissection of matrix-only effects, micronutrient-specific effects (I vs Se), synergy (I+Se), microbiota-driven SCFA modulation (inulin), and direct epigenetic modulation (butyrate).

4.2.7. Diet preparation, I/Se dose calculation and justification. All diets prepared in collaboration with a certified laboratory feed producer. Lyophilised lettuce powders from WP1 will be incorporated based on validated ICP-MS measurements of I and Se.

4.2.7.1. Human RDA used to estimate physiologically meaningful doses: I = 150 µg/day, Se = 55 µg/day.

4.2.7.2. Upper tolerable levels: I = 600–1100 µg/day, Se = 400 µg/day (WHO/FAO).

4.2.7.3. Typical human lettuce consumption: 50–100 g/day fresh weight.

4.2.7.4. Allometric conversion from human→mouse: factor = 12.3 (Reagan-Shaw et al. 2008).

4.2.7.5. Target mouse intake: I target ≈ 100–250 µg/kg diet; Se target ≈ 0.25–0.5 mg/kg diet. These fall within historically safe ranges for C57BL/6 mice (Schomburg & Köhrle 2008).

4.2.7.6. Final I/Se concentrations adjusted based on actual lettuce ICP-MS quantification from WP1.

4.2.8. Monitoring of physiological parameters. Body weight, food intake, water intake, clinical observation, grooming, posture, and mobility monitored daily. Stool consistency and signs of GI upset documented. Any animal exhibiting >15 % body-weight loss or signs of distress triggering humane endpoints will be removed.

4.2.9. Sampling schedule and sample processing. Sampling occurs at T0 (post-cohousing, pre-diet), T1 (week 1 of intervention), T2 (end of week 8). Collection performed in metabolic cages for 24 h to isolate feces and urine.

4.2.10. Fecal sample handling (WP2 endpoint for fecal material).

4.2.10.1. Fresh fecal pellets collected on dry ice every 2 h to minimise SCFA degradation.

4.2.10.2. Split into aliquots for: a) microbiota profiling (WP4), b) SCFA quantification (HPLC/GC-FID), c) ICP-MS for I/Se, d) shipment to CZ for ex vivo fermentation (WP3).

4.2.10.3. Storage: snap-frozen in liquid nitrogen, then −80 °C.

4.2.10.4. Chain-of-custody documentation identical to WP1.

4.2.11. Longitudinal sample collection (continuous monitoring of microbiome and I/Se, T3/T4 and global methylation levels),

4.2.11.1. Urine collection and handling. Urine collected from metabolic cages over 24 h. Volumes recorded. Samples centrifuged at 4000 g, 10 min, 4 °C, aliquoted and stored at −80 °C for ICP-MS analysis of I/Se excretion and retention.

4.2.11.2. Longitudinal fecal collection for microbiome profiling.

4.2.11.2.1. Feces collected at three time-points: TP1 (Week 0), TP2 (Week 4), TP3 (Week 8).

4.2.11.2.2. Immediately snap-frozen in liquid nitrogen and stored at −80 °C.

4.2.11.2.3. Material shipped to external sequencing center (Czech Republic): TP1 and TP3.

4.2.11.2.4. TP2 retained as backup for optional exploratory sequencing or metabolomics.

4.2.11.3. Longitudinal tail blood collection for monitoring of thyroid hormone axis and global methylation.

4.2.11.3.1. Tail-vein blood sampling at Week 0, Week 4, Week 8 (20–30 µL per time-point; serum obtained after 10 min at 2000 g, 4 °C).

4.2.11.3.2. Quantification of TSH, total T4, free T4, total T3 using ultrasensitive mouse ELISA.

4.2.11.3.3. Body weight measurement twice weekly.

4.2.11.3.4. Core body temperature assessment once weekly using a calibrated rectal thermometer (5–10 s per animal).

4.2.11.3.5. Global DNA methylation surrogate markers using: 5-mC ELISA performed on leukocyte DNA or plasma cell-free DNA (cfDNA), 5-hmC ELISA to assess DNA-demethylation dynamics or LINE-1 methylation quantified using bisulfite-qPCR as a proxy of global genomic methylation.

4.2.12. Necropsy and tissue collection (WP2 endpoint for tissues). At T2 animals are anesthetised with 3 % isoflurane, euthanised by cervical dislocation.

4.2.12.1. Blood via cardiac puncture → plasma separation (15 min, 2000 g, 4 °C).

4.2.12.2. GI tract dissection: duodenum, jejunum, ileum, colon, cecum.

4.2.12.3. Liver, kidney (metabolicaly active tissues driven by THRB) harvested for micronutrient metabolism reference and thyroid-hormone axis.

4.2.12.4. Intestinal epithelial cells isolated by EDTA scraping according to standard methods (Weber et al. 2011), then snap-frozen for omics.

4.2.12.5. Adjacent segments fixed in 4% PFA for histology.

4.2.12.6. All tissues labelled, cryopreserved at −80 °C or −150 °C depending on planned WP protocols.

4.2.13. SOP extracts for key assays. Short SOP format identical to WP1.

4.2.13.1. SCFA quantification (HPLC): Feces homogenised in 0.5 % phosphoric acid, centrifuged, supernatant filtered (0.22 µm), injected into HPLC fitted with an organic acid column; quantification relative to authentic standards (Zhao et al. 2006).

4.2.13.2. Microbiota DNA preservation: Fecal aliquots preserved in DNA/RNA Shield (Zymo) or dry-frozen; extraction later in WP4 with Qiagen PowerSoil Pro Kit.

4.2.13.3. ICP-MS for I/Se (urine and feces): Acid digestion HNO₃/H₂O₂, microwave digestion, analysis as described in WP1.

4.2.14. Materials and reagents.  AIN-93M purified diet ingredients, sodium butyrate (Sigma), inulin (chicory-derived, >90 % purity), lyophilised lettuce powders from WP1, metabolic cages (Tecniplast), SCFA standards, HPLC solvents, DNA collection tubes, isoflurane.

4.2.15. Equipment. Metabolic cages, HPLC system (UV or RID), GC-FID (optional), centrifuges, −80 °C freezers, isoflurane anesthesia system, necropsy instruments, biosafety hood for sample handling.

4.2.16. Replication and statistical power. Based on prior microbiome dietary studies (Thaiss et al. 2016), n = 8–10 mice per group ensures power to detect ≥20–30 % shifts in SCFA and microbiota beta-diversity. Total animals ≈ 64–70.

4.2.17. Quality control measures. Randomisation, blinding during sample processing, consistent time-of-day sampling, strict SOPs, internal controls for SCFA and ICP-MS, microbiota extraction batch controls, and cross-WP sample manifest shared with CZ and PL teams.

4.2.18. Ethical, safety, regulatory notes. All procedures follow FELASA, Polish animal law, ARRIVE guidelines. Chemical safety procedures applied for Se handling. Isoflurane used according to veterinary SOPs. Euthanasia by cervical dislocation is approved for small rodents under anesthesia.

4.2.19. Deliverables (WP2 endpoints). D2.1. Complete biobank of fecal samples (T0, T1, T2) with metadata and manifest. D2.2. Urine samples with I/Se quantification report. D2.3. Plasma and organ samples archived for downstream omics. D2.4. Necropsy report including body composition, organ weights, histological preservation. D2.5. Shipment of fecal fermentation batches to CZ (WP3). D2.6. Internal QC documentation.

4.2.20. Timeline (descriptive Gantt-aligned narrative). Month 1: Animal arrival, 1-week habituation, 3-week cohousing and microbiota stabilisation. Month 2–3: 8-week dietary intervention including T0/T1/T2 sampling. Month 4: Necropsy, tissue biobanking, data logging, shipment of fecal samples to CZ. Total WP2 duration: approximately 4 months.

4.2.21. References (methodology and SOP sources) at the end of the document.

4.3.1. Objective: 1) Microbiome profiling, and generation of 2) microbiota-metabolite–enriched conditioned media. Produce reproducible, well-documented anaerobic fermentations of control and I/Se-biofortified lettuce using pooled human and mouse faecal inocula for microbiome profiling. Generate sterile SCFA-enriched conditioned media for downstream organoid and cell assays, produce parallel reference-butyrogenic strain supernatants, and provide comprehensive QC datasets. Half of each culture biomass will be retained for microbiome genotyping (shipped frozen to PL) and half processed into conditioned media by re-suspension/incubation in cell culture medium for ≥4 hours, sterile clarification and storage for WP4 cell/organoid experiments.Final plan and methods will be adjusted/selected based on WP2 results.

4.3.2. WP3 overview and responsibilities. CZ leads end-to-end WP3: receipt and QC of WP1 lettuce aliquots, receipt and cold-chain QC of mouse faeces from WP2, human donor recruitment/screening, inoculum preparation, anaerobic cultivation (batch and semi-continuous as needed), reference strain culture, conditioned media production (bacterial growth medium → transfer to cell medium incubation), sterile processing, analytical characterisation (GC-FID SCFAs, LC-MS/MS targeted metabolites, optional HPLC-ICP-MS Se speciation), microbiome profiling (16S ± shotgun), documentation and shipment of conditioned media and biomass aliquots. CZ maintains versioned SOPs, chain-of-custody logs, QC database and metadata files tied to unique sample IDs

4.3.3. Experimental design — substrates, inocula, controls and overall replication.

4.3.3.1. Substrates (WP1 batches): SAL-CTR (control lettuce), SAL-I (iodine enriched), SAL-Se (selenium enriched), SAL-I+Se (combined enrichment) + SAL(-)-CTR (just medium).

4.3.3.2. Inocula: pooled human faecal inoculum generated from ≥6 screened healthy donors (no antibiotics in prior 3 months), and pooled mouse faecal inocula received frozen from WP2 (one pool per dietary group or per cage as agreed with PL)

4.3.3.3. Reference strains: Faecalibacterium prausnitzii, Roseburia intestinalis, Eubacterium rectale obtained from DSMZ/ATCC for strain-specific conditioned media

4.3.3.4. Controls: medium-only blank, heat-killed substrate control, inulin (positive fermentable substrate) control, SCFA-spiked media (defined acetate/propionate/butyrate mix) as assay benchmarks

4.3.3.5. Replication: human donor pools n≥6 × technical triplicates; mouse pooled fermentations n≥4 independent biological replicates × technical triplicates; reference strains biological n≥3; timepoints 0, 6, 12, 24, 48 h (optionally 72 h for selected runs)

4.3.4. Donor recruitment, fecal collection and inoculum preparation.

4.3.4.1. Human donor recruitment under approved ethics: inclusion criteria healthy adults 20–50 y, no antibiotics/probiotics in prior 3 months, no chronic GI or endocrine disease; record metadata (age, sex, BMI, diet, medication history)

4.3.4.2. Faecal collection: use anaerobic stool collection kits with reducing buffer (e.g., AnaeroGen/BD GasPak style kits or commercial anaerobic stool collection devices). Transport at 4 °C and process ≤2 h from defecation wherever feasible.

4.3.4.3. Inoculum stock: prepare 10% w/v faecal slurry in pre-reduced anaerobic PBS + 0.05% L-cysteine HCl inside anaerobic chamber; homogenise by stomacher or vortex, filter through sterile gauze to remove large particulates. Aliquot working stocks and cryopreserve backup stocks in 20% glycerol at −80 °C

4.3.4.4. Mouse feces: receive frozen T2 pellets on dry ice from PL; log origin metadata; pool per dietary group as agreed; thaw minimally and prepare slurry as for human material

4.3.4.5. Pre-fermentation QC: small aliquot for rapid 16S profiling (shallow sequencing) and anaerobic CFU plating on nonselective media to document baseline composition and viability

4.3.5. Fermentation platform, media formulations and operating conditions.

4.3.5.1. Reactor types and layout: Coy anaerobic workstation with N₂/CO₂ (80:20) headspace for handling; batch fermentations in sterile serum bottles (120–250 mL) with butyl rubber septa for throughput; larger stirred anaerobic reactors (0.5–2 L) for scaled or continuous/semi-continuous where required

4.3.5.2. Basal bacterial growth medium (recipe based on YCFA / Macfarlane formulations), according to the manufacturer.

4.3.5.3. Cell culture medium used later for conditioning: organoid/epithelial culture medium (Advanced DMEM/F-12 + supplements used by CZ organoid lab) prepared sterile and pre-equilibrated to 37 °C and reduced O₂ if needed

4.3.5.4. Substrate loading: initial substrate 1% w/v freeze-dried lettuce powder in bacterial medium; pilot optimise (0.5–2% w/v range) to reproduce in vivo SCFA profiles while avoiding medium saturation

4.3.5.5. Inoculation ratio: 1:10 faecal slurry to medium (final faecal equivalent 1% w/v) or 5–10% depending on pilot biomass and metabolite output

4.3.5.6. Physical conditions: incubate at 37 °C with gentle agitation (100–150 rpm); monitor and record pH at set intervals; maintain pH 6.5–6.8 via manual titration (1 M NaOH/HCl) for batch runs or automated titration for prolonged runs

4.3.5.7. Time course sampling: remove 2–5 mL aliquots at 0, 6, 12, 24, 48 h (and 72 h if used); aliquots split for (a) metabolomics/SCFA, (b) DNA extraction (biomass pellet), (c) sterility/viability plating; freeze metabolite aliquots at −80 °C immediately (snap freeze on dry ice)

4.3.6. Bacterial pre-culture on bacterial medium and subsequent conditioned media production in cell medium.

4.3.6.1. Two-step workflow rationale: grow bacteria first in optimized bacterial growth medium to achieve biomass and metabolite production then transfer portion of culture biomass to cell culture medium for incubation ≥4 h to produce cell-compatible conditioned media containing microbially produced metabolites but minimising carryover of bacterial medium constituents

4.3.6.2. Step A — growth phase: inoculate bacterial medium with faecal slurry or reference strains and incubate under conditions in 4.3.5 until target timepoint (e.g., late exponential at 12–24 h). Harvest culture and split immediately 1:1 into two streams

4.3.6.3. Step B1 — genotyping aliquot: take 50% of culture biomass, centrifuge 5000 g 10 min 4 °C, remove supernatant, freeze pellet at −80 °C and ship to PL for microbiome genotyping (this documents community after growth on bacterial medium)

4.3.6.4. Step B2 — conditioned media production: take remaining 50% culture, centrifuge 4,000–10,000 g 10 min 4 °C, discard supernatant, gently resuspend pellet in pre-warmed, pre-reduced cell culture medium (e.g., Advanced DMEM/F-12 + 10 % FBS or organoid base medium without antibiotics) to achieve a defined ratio (e.g., biomass equivalent to 5–10% v/v of original culture in final conditioned media). Incubate this mixture under anaerobic or microaerophilic conditions at 37 °C for minimum 4 h (user requirement) to allow transfer/diffusion of small microbial metabolites into cell medium

4.3.6.5. Clarification: after ≥4 h incubation, centrifuge 10,000 g 10 min 4 °C, sterile filter supernatant through low-binding 0.22 μm PES filters under aseptic/anaerobic conditions into sterile vials; perform endotoxin (LAL) and sterility checks; aliquot and store at −80 °C as conditioned media for direct use in WP4 without further dilution of cell media

4.3.6.6. Documentation: record biomass equivalent, initial and final SCFA concentrations, pH, osmolarity and any deviations. Retain a QC aliquot of conditioned media for metabolomics and sterility plating

4.3.7. Sterile processing, optional heat treatment and storage. 

4.3.7.1. Filtration: 0.22 μm PES low-binding syringe/inline filters; perform filtration in laminar flow or anaerobic chamber where possible to avoid oxidation of labile metabolites

4.3.7.2. Heat treatment: optional controlled mild heat inactivation (e.g., 56 °C for 30 min) may be applied to some conditioned media lots to assess effect of heat-labile components; document effects on SCFA/metabolite profile before and after heat

4.3.7.3. Storage: aliquot 0.5–1 mL into cryovials, avoid >1 freeze–thaw for experimental lots, store −80 °C; prepare shipment logs and ship on dry ice to PL with QC metadata

4.3.8. Conditioned media compatibility testing (cell safety pilot; SOP).

4.3.8.1. Pilot panel: test dilution series of conditioned media (2.5%, 5%, 10%, 20% v/v) on Caco-2, HT-29 and small organoid pilot cultures

4.3.8.2. Assays: CellTiter-Glo viability, LDH release, ROS (DCFDA), morphological assessment under phase-contrast/IF after 24 and 72 h 4.3.8.3. Acceptance criteria: select working concentration giving ≥90% viability of control and LDH ≤1.5× control; if endotoxin >0.5 EU/mL consider endotoxin removal or additional cleanup

4.3.9. Analytical assays and QC panels. 

4.3.9.1. SCFA quantification: GC-FID after derivatisation (propylation or methylation methods); use internal standard 2-ethylbutyric acid; calibration curves for absolute quantification of acetate, propionate, butyrate, valerate, isobutyrate, isovalerate (Zhao et al. 2006)

4.3.9.2. Targeted metabolomics: LC-MS/MS MRM panels for indoles, phenolic acids, lactate and redox metabolites with isotope-labelled standards when available; include QC pooled samples and technical replicates

4.3.9.3. pH, osmolarity and conductivity measurement for each conditioned media lot

4.3.9.4. Microbiome profiling: extract DNA from inocula, pre-culture and post-culture biomass pellets using QIAamp PowerFecal Pro Kit; 16S V3–V4 amplicon sequencing on Illumina MiSeq and shotgun metagenomics for selected representative runs; process with DADA2/QIIME2 for ASV tables and HUMAnN3 for functional profiling

4.3.9.5. Viability/sterility: plate pre- and post-filtration aliquots anaerobically on Gifu anaerobic agar, incubate 48–72 h; document absence of growth for sterile lots

4.3.9.6. Endotoxin: LAL assay on conditioned media; record EU/mL

4.3.9.7. Se speciation (optional): HPLC-ICP-MS for Se species (SeMet, SeCys, MeSeCys) in I+Se conditions

4.3.10. Data recording, deliverables and sample flow. 

4.3.10.1. D3.1 SOP package for fermentation, two-step bacterial→cell-medium conditioning SOP and QC criteria

4.3.10.2. D3.2 Time-series raw SCFA and metabolomics files and processed tables tied to unique sample IDs

4.3.10.3. D3.3 Sterile conditioned media lots with QC sheets (SCFA, pH, osmolarity, sterility, endotoxin) shipped to PL

4.3.10.4. D3.4 Microbiome pre/post sequencing datasets (raw reads, ASV/OTU tables, taxonomic and functional profiles)

4.3.10.5. D3.5 Reference strain conditioned media, stocks and profiles

4.3.10.6. Maintain chain-of-custody manifests and sample tracking database for all shipped biomass and conditioned media lots

4.3.11. Replication, sample sizes and statistical rationale.

4.3.11.1. Human donor fermentation: ≥6 donor pools with technical triplicates per substrate × timepoint to capture inter-individual variance and permit differential abundance/metabolite association testing

4.3.11.2. Mouse pooled fermentations: n≥4 independent biological replicates per dietary group × technical triplicates

4.3.11.3. Reference strains: biological n≥3 replicates

4.3.11.4. Power rationale: designed to detect medium effect sizes in metabolite concentration and taxa relative abundance; finalize power calculations in pilot using observed variance

4.3.12. Timeline (descriptive Gantt-aligned narrative).

4.3.12.1. Month 0–1 Pilot optimisation: substrate load, inoculum prep, bacterial→cell medium conditioning parameters, filtration and compatibility testing

4.3.12.2. Month 2–4 Main fermentation campaigns: human and mouse sets, SCFA/targeted metabolomics runs, microbiome sequencing library prep

4.3.12.3. Month 4–4.5 Conditioned media QC, sterile filtration, packaging and shipment to PL

4.3.12.4. Month 5 Data delivery, SOP finalisation and WP3 closeout

4.3.13. Personnel and responsibilities.

4.3.13.1. WP3 leader (CZ PI) overall coordination, QC sign-off and data delivery

4.3.13.2. Senior microbiologist (1 FTE) inoculum handling, fermentations and SOP maintenance

4.3.13.3. Culture technician (1 FTE) biomass handling, filtration, aliquoting and shipping

4.3.13.4. Analytical chemist (1 FTE) SCFA GC-FID and LC-MS/MS metabolomics

4.3.13.5. Bioinformatician (0.5 FTE) 16S/shotgun processing and reporting

4.3.13.6. Lab technician (0.5 FTE) sample logging, sterility checks and inventory

4.3.14. Materials, reagents and equipment (detailed list). 4.3.14.1. Anaerobic workstation (Coy or equivalent) with N₂/CO₂/H₂ capability, oxygen monitors and incubator stage 4.3.14.2. Serum bottles, Hungate tubes, stirred anaerobic reactors (0.5–2 L), gas-tight septa and magnetic stir plates 4.3.14.3. Basal media reagents (peptone, yeast extract, salts, vitamin mixes, hemin), L-cysteine HCl, resazurin and sterile anaerobic buffers 4.3.14.4. Centrifuges with refrigerated rotors for 50–250 mL tubes capable of 10,000 g 4.3.14.5. 0.22 μm low-binding PES sterile filters, syringe filters, sterile filtration manifold compatible with anaerobic use 4.3.14.6. GC-FID or GC-MS with SCFA derivatisation kit, autosampler and SCFA standards (acetate, propionate, butyrate, valerate, isobutyrate, isovalerate) and internal standard 2-ethylbutyric acid 4.3.14.7. LC-MS/MS system with targeted MRM capability and isotope-labelled standards for indoles/phenolics 4.3.14.8. HPLC-ICP-MS access for Se speciation (optional) 4.3.14.9. DNA extraction kits QIAamp PowerFecal Pro, 16S library prep kits, Illumina MiSeq/NextSeq access or outsourcing 4.3.14.10. LAL endotoxin assay kits, media for sterility plating (Gifu anaerobic agar/BHI), cryovials, dry ice shipping containers and temperature loggers 4.3.14.11. Cell culture supplies: Advanced DMEM/F-12, Matrigel (if organoid pilots), supplements (EGF, Noggin, R-spondin), antibiotic-free reagents for conditioned media compatibility testing

4.3.15. SOP extracts (short formats for grant).

4.3.15.1. Fermentation setup SOP extract: pre-reduce basal medium ≥24 h under N₂ in anaerobic chamber; prepare 10% w/v faecal slurry in pre-reduced PBS + 0.05% L-cysteine; add freeze-dried lettuce substrate to 1% w/v; inoculate at 1:10 final slurry:medium; incubate 37 °C with gentle stirring; sample at defined timepoints, snap freeze metabolite aliquots and pellet biomass for DNA

4.3.15.2. Two-step conditioned media SOP extract: at harvest centrifuge culture 4,000–10,000 g 10 min 4 °C; split biomass 1:1 — pellet half freeze for genotyping; resuspend other half in pre-warmed cell culture medium and incubate ≥4 h at 37 °C; centrifuge 10,000 g 10 min 4 °C; sterile filter supernatant through 0.22 μm PES filters into sterile vials; test endotoxin and sterility; aliquot and store −80 °C

4.3.15.3. Compatibility SOP extract: expose Caco-2/HT-29/organoid pilots to conditioned media dilution series 2.5–20% v/v for 24–72 h; measure CellTiter-Glo and LDH; accept working dilution where viability ≥90% and LDH ≤1.5× control

4.3.16. Quality control and reproducibility measures.

4.3.16.1. Versioned SOPs and training records

4.3.16.2. Technical triplicates and adequate biological replication

4.3.16.3. Inclusion of positive/negative controls and medium blanks in each run

4.3.16.4. Internal standards and pooled QC samples in metabolomics; sequencing blanks and mock communities for microbiome QC

4.3.16.5. Chain-of-custody and manifest for all shipped conditioned media and frozen biomass

4.3.17. Safety, ethics and regulatory notes.

4.3.17.1. Human stool work under approved ethics with informed consent; BSL-2 containment and institutional biosafety approval for handling

4.3.17.2. Mouse feces shipped on dry ice under institutional and IATA guidelines; MTAs in place between PL and CZ

4.3.17.3. Waste decontamination by autoclaving and chemical disinfection; reagent safety (Se compounds) handled per MSDS and institutional hazardous waste rules

4.3.18. Contingencies and alternatives.

4.3.18.1. If human pooled inocula are too heterogeneous, create stratified pools representing major enterotypes (two pools)

4.3.18.2. If 0.22 μm filtration causes unacceptable loss of metabolites, evaluate centrifugal filtration with low-binding membranes and quantify recovery using spiked standards

4.3.18.3. If strict anaerobes (e.g., F. prausnitzii) are hard to culture, partner with specialist labs or obtain pre-prepared strain supernatants; document provenance

4.3.18.4. If conditioned media endotoxin incompatible with organoids, apply endotoxin removal (polymyxin B columns) and re-test

4.3.19. Replication and statistical power (summary).

4.3.19.1. Human donor fermentations ≥6 donor pools, technical triplicates per condition; mouse pools ≥4 biological replicates; reference strains ≥3 replicates

4.3.19.2. Designed to detect medium effect sizes for SCFA/metabolite shifts and taxa changes; finalize power calculations in pilot using observed variances

4.3.20. Deliverables (WP3 endpoints).

4.3.20.1. D3.1 Complete SOP package including two-step bacterial growth → cell-medium conditioning SOP and QC acceptance criteria

4.3.20.2. D3.2 Batch-wise sterile conditioned media lots with complete QC sheets shipped to PL

4.3.20.3. D3.3 SCFA quantification datasets and processed metabolomics tables

4.3.20.4. D3.4 Microbiome pre/post-fermentation sequencing datasets (raw reads, ASV tables, taxonomic and functional profiles)

4.3.20.5. D3.5 Reference strain conditioned media stocks and documentation

4.3.20.6. D3.6 Final WP3 report including pilot optimisation results, compatibility assays and shipment manifests

4.3.21. Timeline summary.

4.3.21.1. Month 0–1 Pilot optimisation: bacterial medium recipe confirmation, substrate load, two-step conditioning parameters, filtration and cell-compatibility pilots 4.3.21.2. Month 2–4 Main fermentation campaigns and analytics (human and mouse sets) 4.3.21.3. Month 4–4.5 Conditioned media QC, sterile filtration, aliquoting and shipment to PL 4.3.21.4. Month 5 Data delivery, SOP finalisation and WP3 closeout

4.3.22. Personnel and budget notes. 4.3.22.1. Roles and FTE as in 4.3.13; budget covers anaerobic workstation consumables, GC-FID/LC-MS consumables, sequencing, sterile filtration supplies, donor compensation, shipping and contingency for outsourced Se speciation where needed

4.3.23. References (selected methodological precedent in NCBI citation style)

4.4.1. Objective. The aim of WP4 is to determine how metabolites produced by gut microbiota exposed to iodine- and/or selenium-biofortified lettuce (WP3 conditioned media) modulate human intestinal epithelial epigenetic regulation, chromatin status, thyroid-hormone-receptor–associated pathways, and global transcriptional programs. WP4 will use 2D cultures of two human colorectal epithelial cell lines (Caco-2 and HT-29) and 3D primary human colon spheroids. Cells will be treated with WP3-derived conditioned media (CM), obtained from bacteria pre-grown on lettuce-supplemented bacterial media and subsequently transferred for ≥4 h into cell-culture media. This taks is also designed to evaluate how SCFA-enriched conditioned media influence human intestinal epithelial epigenome via SAM/SAH balance, histone deacetylase (HDAC) activity, DNA methylation enzyme activities (DNMT, TET), chromatin accessibility, transcriptional output of thyroid-related genes, inflammatory pathways, epithelial barrier integrity and cellular phenotypes. WP4 will collect cellular material for epigenomics, transcriptomics, targeted RT-qPCR, and chromatin assays, and will collect spent media for metabolic, immunological, and epigenetic marker analyses. All epigenomic and genomic material is shipped to PL; metabolomics and functional assays are performed in CZ. Final plan and methods will be adjusted/selected based on WP3 results.

4.4.2. WP4 overview and responsibilities. The Czech team (CZ) is responsible for all in vitro cell culture, intestinal spheroid culture, conditioned-media preparation (from WP3), exposure experiments, metabolic and immunological assays, light- and fluorescence-microscopy imaging, morphometric quantification, and primary data processing. PL receives all cellular DNA/RNA/chromatin extracts for next-generation sequencing and epigenetic profiling. CZ performs RT-qPCR panels, ELISA panels, metabolomics of media and cell lysates, and phenotypic assays. All procedures follow GCCP (Good Cell Culture Practice), OECD in vitro system recommendations, and institutional biosafety approval.

4.4.3. Cell culture models. Three intestinal epithelial models will be used. Caco-2 (ATCC HTB-37) maintained in DMEM high glucose + 10 % FBS, 1 % penicillin/streptomycin, 1 % non-essential amino acids; cultured at 37 °C, 5 % CO₂. HT-29 (ATCC HTB-38) maintained in McCoy’s 5A + 10 % FBS, 1 % P/S. Human colon spheroids derived from commercially available or biobanked human colonic crypts, expanded in Matrigel domes using standard intestinal organoid medium (WRN-conditioned medium or commercial IntestiCult medium). Spheroids maintained in 3D Matrigel matrix, with medium changed every 48 h. All cultures routinely passaged at 70–80 % confluence.

4.4.4. Bacterial conditioned media (CM) used in WP4. Conditioned media are received from WP3, prepared by growing bacteria on bacterial medium supplemented with lettuce powders from WP1 (groups identical to WP3 ), followed by transfer for ≥4 h into sterile cell-culture medium containing the same lettuce extracts to maintain matrix-specific exposure. After ≥4 h incubation, bacterial suspensions are centrifuged (4000 g, 10 min, 4 °C), supernatants filtered through 0.22 µm PES filters, heat-treated when required (60 °C, 30 min), aliquoted under sterile conditions, snap-frozen and stored at −80 °C. A non-conditioned control medium (base cell-culture medium with lettuce additives but without bacteria) is included as an additional reference. CM groups correspond exactly to WP3: CTR-CM, I-CM, Se-CM, I+Se-CM, INU-CM, BUT-CM, and non-conditioned medium (NC-CM).

4.4.5. Experimental design. Cells and spheroids will be seeded at defined densities. Caco-2 and HT-29: 24-well plates, 1×10⁵ cells/well; 6-well plates for omics (4×10⁵ cells/well). Spheroids: Matrigel domes (25 µL), 500–1500 spheroids/dome. After 24 h stabilisation, cultures receive 20–30 % CM dilution (testing range validated in pilot assay to avoid cytotoxicity unrelated to biological metabolites). Treatments last 4 h, 12h, 24 h, 48 h, and 72 h depending on downstream assay. For each treatment, parallel wells are harvested for DNA, RNA, chromatin, microscopy, cell viability, and metabolomics.

4.4.6. Treatment groups. Each model receives: a) NC-CM (negative control), b) CTR-CM, c) I-CM, d) Se-CM, e) I+Se-CM, f) INU-CM (SCFA-enhancing control), g) BUT-CM (HDAC-inhibition control). This design allows distinguishing matrix effects, micronutrient-specific effects, synergy, microbiota-dependent SCFA effects, and direct epigenomic effects (butyrate).

4.4.7. Conditioned-media preparation quality control. Each CM batch undergoes: sterility check on blood agar plates; endotoxin quantification (LAL assay); SCFA panel (GC-FID); total protein concentration; metabolite fingerprint (LC-MS). pH adjusted to 7.3–7.4 before cell exposure; osmolarity validated using freezing-point osmometer.

4.4.8. Cell viability and proliferation assays. MTT assay and IC50 of conditioned media performed according to standard protocols: 0.5 mg/mL MTT, incubation 4 h, formazan dissolved in DMSO, absorbance at 570 nm. BrdU incorporation assay: 10 µM BrdU added for last 2 h of treatment, detected via ELISA-based BrdU kit. LDH cytotoxicity quantified from supernatants. All assays performed at 4h, 12h, 24 h, 48 h, and 72 h.

4.4.9. Morphological, cytoskeletal, and proliferation markers. Immunocytochemistry: cells fixed in 4 % PFA, permeabilised with Triton X-100, blocked with 3 % BSA. Primary antibodies: Ki-67, ZO-1, E-cadherin, β-catenin, acetyl-H3K9, H3K27ac or γH2AX. Fluorescent secondary antibodies applied; nuclei stained with DAPI. Imaging on CZ confocal systems (Zeiss LSM or Leica SP). Morphometric quantification using ImageJ: cell size, polarity indices, tight-junction continuity, spheroid diameter and lumen integrity. Barrier function in 2D monolayers will be assessed by TEER and FITC-dextran permeability. Spheroids will be analysed for morphology, size, polarity, lumen formation and survival. Parallel assays will quantify proliferation (Ki-67), apoptosis (caspase-3), oxidative stress and mitochondrial function.

4.4.10. Intestinal spheroid-specific functional endpoints. Spheroid swelling assay as surrogate for ion-transport activity: forskolin-stimulated swelling recorded every 10 min for 1 h. Viability via CellTiter-Glo 3D assay. Spheroid differentiation markers quantified by RT-qPCR: MUC2, LGR5, OLFM4, KRT20. Spheroid morphology scored using standard metrics (lumen formation, budding, crypt-like structures).

4.4.11. Sampling and material handling. Supernatants collected at 24, 48, and 72 h, centrifuged (1500 g, 10 min, 4 °C), aliquoted, snap-frozen at −80 °C. Cells harvested via TrypLE or mechanical dissociation (for spheroids), washed in PBS, split into: a) DNA isolation, b) RNA isolation, c) chromatin preparation, d) metabolomics pellets, e) microscopy slides. All materials barcoded and traced through cross-WP documentation.

4.4.12. Cellular DNA/RNA extraction for omics (PL). All samples shipped to PL on dry ice. DNA extracted using Qiagen AllPrep DNA/RNA Mini Kit; RNA integrity confirmed by Agilent Bioanalyzer (RIN > 8). DNA and RNA stored at −80 °C. Chromatin prepared using truChIP cross-link kit. Chromatin accessibility will be determined by ATAC-seq in selected conditions to associate metabolite-driven epigenetic reprogramming with accessible regulatory regions.

4.4.13. Preliminary assessement of transcriptomic responses (CZ). RT-qPCR will be performed using SYBR-based kits, analyzing the following panels: epigenetic regulators (DNMT1, DNMT3A, DNMT3B, TET1, TET2, TET3, HDAC1, HDAC2, HDAC3 and/or SIRT1); thyroid-hormone–related genes (THRA, THRB, DIO1, DIO2, DIO3); inflammatory cytokines (IL-6, IL-8, TNF-α and/or  IL-10); metabolic regulators (PPARG and/or SLC-family transporters such as SLC5A5); transcriptional control and proliferation markers (MKI67 and/or  E2F1); translational control genes, including those extracted from the butyrate-responsive gene list (EEF2, RPL28, RF00100, MT-ND5, MT-ATP6, MT-CO2, MT-CO3, RPS3A, RPL13A, RPL3, RPL15, EEF1G, RPS11, RPL18A, GAPDH, RPS6, FTL, RPS4X, PCBP2, YBX1, HNRNPA1, HSPD1, EIF4G2, EEF1A1, PPIA and/or  RPL4); and/or  the remaining butyrate-responsive genes (EPPK1, EGR1, KRT19, AGRN, SLC2A3, FLNB, PLEC, FAT1, ACTN4, ACTB, RNU2-2P, RN7SL2, RMRP, AL355075.4, NKD1, HUWE1, APCDD1, SLC1A5, KRT8, CTNNB1, PROX1, CANX, ENO1and/or  KRT5). At least 3 housekeeping genes: GAPDH, ACTB, HPRT1. ΔΔCt analysis applied.

4.4.14. Preliminary assessement of proteins and metabolites (CZ). ELISA panels might be used for epigenetic and inflammatory markers. Supernatants analysed for: 5-mC, 5-hmC, HDAC activity surrogate markers, IL-6, IL-8, TNF-α, TGF-β, IFN-γ and/or T3/T4/FSH elisa. Kits from Abcam, Cayman, or R&D Systems. Samples run in duplicates.

4.4.5 Preliminary assessement of chromatin and DNA methylation assays (CZ): Global 5-mC and 5-hmC levels will be assessed by ELISA-based quantification. Locus-specific methylation patterns for eg. THRB, THRA, DIO1, DIO2, NCOR1, and regulatory enhancers will be evaluated using targeted bisulfite sequencing and/or dedicated RTPCR kits.

4.4.15. Metabolomic assays (CZ).

4.4.15.1. LC-MS/MS targeted metabolomics on CZ mass spectrometers (Thermo Q-Exactive or Agilent 6545 Q-TOF and/or GC–FID (Shimadzu Nexis GC-2030) will be used to analyze SCFAs, BCFAs, indoles, bile acids, polyamines, tryptophan metabolites, redox metabolites. Microwave digestion + FAAS / ETAAS and/or LC–ESI–MS/MS (Ultimate 3000 RS) will be used for Inorganic trace elements: iodine, selenium. Organic extraction: methanol precipitation (80 % MeOH), centrifugation (14,000 g, 10 min). Chromatographic separation on C18 reversed-phase column.

4.4.15.2. Intracellular SAM and SAH concentrations will be quantified using LC–MS metabolomics to determine methylation capacity (SAM/SAH ratio). C–ESI–MS/MS (Ultimate 3000 RS) can be also used to analyze cellular SAM, SAH, MTA, folate species, amino-acid derivatives, polyamines. Global HDAC activity will be measured fluorometrically. DNMT and TET activities will be quantified using enzyme-specific activity assays to determine shifts in DNA methylation and demethylation potential driven by bacterial metabolites derived from I+Se lettuce fermentations.

4.4.7 Data integration: Epigenetic enzyme activities, SAM/SAH balance, DNA methylation, chromatin accessibility and transcriptional output will be integrated with WP3 microbial metabolomic profiles to determine mechanistic links between I+Se-induced SCFA and non-SCFA metabolites and host epithelial epigenome reprogramming. Results will be prepared for cross-WP mechanistic modelling in WP6.

4.4.16. Chromatin accessibility and histone-mark assays (PL). ATAC-qPCR on selected loci using Omni-ATAC adapted for cell lines. Histone modifications (H3K27ac, H3K4me3, H3K9ac, H3K9me3) assessed via ChIP-qPCR. Chromatin aliquots shipped to PL for ChIP-seq and ATAC-seq.

4.4.17. Quality control measures. Mycoplasma tests every 2 weeks (PCR). Authentication of cell lines via STR profiling. Batch-to-batch CM variation monitored by SCFA and metabolomic fingerprints. Blinded sample coding for imaging and RT-qPCR. Standardised exposure time windows and CM dilutions.

4.4.18. Safety and biosafety notes. All CM treated as BSL-2 material. Work performed in Class II biosafety hoods. All waste autoclaved. Matrigel handled on ice. Organic solvents for metabolomics handled in fume hood with PPE.

4.4.19. Deliverables (WP4 endpoints). D4.1. Complete library of treated cells and spheroids for omics. D4.2. DNA/RNA/chromatin shipments to PL. D4.3. Metabolomic datasets from media and cell lysates. D4.4. ELISA and RT-qPCR datasets. D4.5. High-resolution microscopy images with morphometric quantification. D4.6. Cross-WP metadata and QC documentation.

4.4.20. Timeline. Month 1: Pilot optimisation of CM dilutions, cell viability, spheroid establishment. Month 2: Large-scale exposure experiments, sample collections at 24/48/72 h. Month 3: Metabolomics, ELISA, microscopy analyses. Month 4: RT-qPCR, final QC, shipment of omics materials to PL. Total WP4 duration: approx. 4 months.

4.4.21. References (methodology and SOP sources).

4.5.1. Objective. Perform systematic, high-throughput multi-omics characterisation and integrative computational analysis to (1) genotype microbiome samples received from CZ (pre- and post-fermentation), (2) generate high-quality bulk and single-nucleus multi-omic datasets (snRNA-seq, snATAC-seq) from mouse colon tissues (WP2 endpoint) and human cell/organoid exposures (WP4 endpoint), (3) quantify metabolome bridges (SCFAs, SAM/SAH, Se-metabolites generated by CZ) and link metabolite changes to chromatin/ DNA methylation/transcriptome shifts, and (4) deliver mechanistic models that prioritise SCFA- and micronutrient-responsive gene networks, cell states and candidate regulators for validation. WP5 stops at data products, integrated models and prioritized targets; it returns processed libraries, raw & processed data, and analysis pipelines to consortium repositories. Key conceptual axis to be tested: I+Se lettuce → microbiome composition/function → microbial metabolome (SCFA + non-SCFA + Se species) → one-carbon metabolism (SAM/SAH) & HDAC/DNMT/TET activity → chromatin accessibility & DNA methylation → transcriptional responses (THRB, THRA, DIOs) → epithelial phenotypes. Mechanistic literature and multiome methods guide experimental choices. Final plan and methods can be adjusted/selected based on WP4 results.

4.5.2. WP5 overview and responsibilities. PL leads WP5 end-to-end: receipt and QC of sequencing inputs (microbiome reads from CZ; tissue/nuclei/ RNA from PL and metabolome from CZ), library preparation for bulk RNA-seq and bisulfite/targeted methylation assays, single-nucleus multiome library prep (Chromium single-nucleus Multiome workflow — procured/outsourced as needed), sequencing coordination (in-house MiSeq for 16S, outsourced Illumina NextSeq/NovaSeq for bulk and multiome as required), metabolomics run coordination with CZ (C–ESI–MS/MS) for SCSF, SAM/SAH and related metabolites, bioinformatics pipelines (raw QC → mapping → quantification → differential testing → pathway enrichment → multi-omics integration using MixOmics and mediation models), versioned code, and delivery of processed datasets, visualisations and mechanistic reports. PL will track sample provenance, ensure metadata completeness, and coordinate data deposition to controlled repositories. Key in-scope tasks are strictly computational/analytical; wet-lab multiome library prep may be performed at PL if Chromium 10x Multiome kit and instrument access are secured, otherwise outsourced to a certified genomics provider; sequencing will follow vendor and institutional QC standards.

4.5.3. Samples and upstream inputs. WP5 will process: a) microbiome sequencing inputs from CZ: pre- and post-fermentation biomass for human and mouse experiments (raw reads for 16S and shotgun where available); b) conditioned media metabolomics panels (SCFA tables from CZ GC-FID and LC-MS non-SCFA panels) linked to fermentation sample IDs; c) mouse colon epithelial nuclei and bulk tissue RNA/DNA from WP2 (T2 prioritized animals representing STD, SAL-I+Se, INU, BUT and selected contrasts); d) human cell line and organoid lysates and nucleic acids from WP4 exposures (control media, conditioned media from SAL-CTR, SAL-I, SAL-Se, SAL-I+Se, INU, BUT and medium blanks). For each sample, WP5 requires metadata spreadsheets (sample ID, WP origin, animal ID/donor ID, treatment, timepoint, storage conditions, freeze-thaw history). Microbiome datasets will be sequenced/processed before host multiome integration to allow taxon→function mapping used in mediation models.

4.5.4. Microbiome genotyping and functional profiling (inputs from CZ). Microbiome raw data (16S amplicons on MiSeq and shotgun where available) will be processed using QIIME2/DADA2 pipelines for amplicons (denoising, ASV generation, taxonomy with SILVA) and KneadData, MetaPhlAn/HUMAnN for shotgun to obtain species-level and functional pathway abundances. QC includes negative controls, mock communities and read depth thresholds. Outputs: ASV/OTU tables, taxonomic assignments, alpha/beta diversity metrics, differential abundance tests (ANCOM/DESeq2), functional pathway tables (MetaCyc/KEGG), and correlation matrices linking taxa/pathways to SCFA/non-SCFA metabolite abundances from CZ. All microbiome deliverables will be versioned and linked to conditioned-media lots used in WP4 experiments.

4.5.5. Bulk transcriptomics and targeted epigenomics on mouse and human samples. Bulk RNA-seq: library prep (ribo-depletion or polyA depending on sample) per Illumina protocol; sequencing depth: ~30–50M paired reads/sample for bulk tissue, 20–30M for organoid bulks; alignment (STAR), quantification (featureCounts), DE analysis (DESeq2), pathway enrichment (GSEA/Reactome). Targeted promoter methylation: PyroMark Q48 assays (eg. THRB, DIO1, DIO2, DNMTs, TETs) using bisulfite conversion kits and pyrosequencing per vendor SOPs; global 5-mC quantification by ELISA and LC-MS validation on subset. Outputs: normalized expression matrices, lists of DE genes per contrast, promoter-level methylation % per CpG site, and global methylation summaries. Quality thresholds and replicate consistency checks applied.

4.5.6. Single-nucleus multiome (snRNA + snATAC) — library prep overview and selection strategy. Select representative mice (T2) and organoid samples based on prior QC (SCFA, microbiome shift, histology) to run single-nucleus Multiome for cell-type resolved chromatin+transcriptome. Nuclei isolation: fresh frozen colon tissue processed by Dounce homogenisation with nuclei buffer + sucrose cushion as per published single-nucleus protocols, filtration through 40 µm, nuclei counting and viability staining. Chromium Single-Nucleus Multiome library prep (10x Genomics Chromium Next GEM) will follow manufacturer protocols for nuclei QC, transposition, GEM generation, and dual library construction; target 5–10k nuclei per sample. Sequencing: paired ATAC and gene expression libraries, recommended read depths per 10x guidance. Computational pipeline: CellRanger-Arc or equivalent for initial processing, downstream integration with Seurat/ArchR for clustering, label transfer, differential accessibility and peak-to-gene linkage, motif enrichment and co-accessibility analyses. Candidate regulatory regions near THRB, DIOs, epigenetic enzyme genes and diet-microbiome-metabolite-epigenome responsive genes will be inspected for accessibility changes correlated with expression and methylation. If Chromium Multiome is not immediately available in-house, library prep will be outsourced to certified provider; PL will retain raw nuclei and aliquots for repeat runs.

4.5.7. Microbiota SCSF - regulated cellular SAM/SAH and metabolite integration. Targeted LC-MS/MS quantification of SAM, SAH, MTA, folate species and selected amino-acid derivatives will be run at CZ (Ultimate 3000 RS C–ESI–MS/MS). Measurements will be performed on matching tissue extracts and selected organoid lysates; normalization to protein or tissue mass. These data will be used to compute methylation potential (SAM/SAH ratio) and correlate with global 5-mC, promoter methylation and DNMT/TET expression. SCFA concentrations measured by GC-FID (CZ) will be integrated with SAM/SAH and gene regulation data to test mediation hypotheses: microbial SCFA production → altered SAM/SAH or HDAC inhibition → chromatin accessibility and methylation changes → transcriptional outcomes. Statistical mediation and causal inference models (e.g., structural equation modelling, causal mediation via bootstrapping) will be used to prioritise candidate causal chains.

4.5.8. Data integration, modelling and hypothesis testing. Multi-omics integration will proceed in stages: per-omics QC and normalization, pairwise correlation matrices (taxa ↔ metabolites, metabolites ↔ SAM/SAH, metabolites ↔ accessibility/expression), multi-block integration using MixOmics (sPLS, DIABLO) to find correlated multi-omic signatures, network inference to identify regulatory modules and mediator analyses to test causal paths linking diet → microbiome → metabolome → epigenome → expression. Machine-learning models (random forest, elastic net) will rank predictive features for epigenetic shifts; cluster-level heatmaps and trajectory analyses will visualise responsive cell states. All code will be versioned in GitLab with containers for reproducibility; final models and biomarker lists delivered with confidence metrics and rationale for validation in downstream targeted assays.

4.5.9. Sample allocation, replication and statistical power. For bulk assays plan n=6–8 biological replicates per contrast (STD, SAL-I+Se, INU, BUT) for robust DE detection. For single-nucleus multiome select 3 biological replicates per key group with 5–10k nuclei each to ensure power to detect cell-type specific accessibility changes. Microbiome genotyping uses all donor/group samples (pre/post fermentation) with ≥6 human donor pools and ≥4 mouse pooled fermentations as input to CZ pipelines (WP3), matched to host samples. Formal power calculations will be finalised in pilot runs using observed variance from WP2/WP3 QC.

4.5.10. SOP extracts (short formats for grant).
4.5.10.1. snRNA/snATAC nuclei prep (extract): mince 50–100 mg colon tissue on ice, dounce homogenize in chilled nuclei isolation buffer with 0.1% NP-40, filter through 40 µm, cushion through 30% iodixanol if needed, count nuclei (Trypan exclusion for integrity), adjust concentration to 1,000 nuclei/µL and proceed to Chromium Multiome following 10x Genomics protocol.

 4.5.10.2. Bulk RNA extraction (extract): homogenise tissue in QIAzol or Trizol, perform column cleanup (RNeasy) with on-column DNase, QC by Bioanalyzer RIN >7 for library prep. 4.5.10.3. Pyrosequencing promoter assay (extract): bisulfite convert 200–500 ng gDNA (EpiTect), PCR amplify target promoter region using PyroMark assays, run on PyroMark Q48 and compute % methylation per CpG per manufacturer SOP. 4.5.10.4. SAM/SAH extraction (extract): homogenise tissue in cold 80% methanol with internal standards, centrifuge 20,000 g 10 min, dry supernatant and reconstitute for LC-MS/MS; quantify using calibration curves and isotope-labelled standards.

4.5.11. Materials, reagents and equipment (detailed list). 4.5.11.1. In-house PL instruments: Illumina MiSeq (07-053) for 16S amplicons, ChemiDoc Imaging System, Bio-Plex 200 System (BioRad), PyroMark Q48 Autoprep (Qiagen), BioRad CFX96 RealTime PCR system; liquid-handling/thermocyclers, Qubit/BioAnalyzer instruments. 4.5.11.2. CZ instruments (to coordinate runs): GC-FID Shimadzu Nexis GC-2030 for SCFAs, C–ESI–MS/MS Ultimate 3000 RS for SAM/SAH and targeted metabolomics, microwave digestion + FAAS/ETAAS for elemental I/Se when required. 4.5.11.3. Required/outsourced instrument: Chromium 10x Single-Nucleus Multiome kit + controller (to procure or outsource library prep and sequencing on NextSeq/NovaSeq). 4.5.11.4. Reagents: library prep kits (Illumina), PyroMark bisulfite kits, ELISA kits for 5-mC/5-hmC, SAM/SAH stable isotope standards, sequencing consumables, nuclei buffers, QC spike-ins and mock community standards. Instruments on hand are explicitly acknowledged and will be used per above; Multiome will be acquired or outsourced depending on procurement timelines.

4.5.12. Bioinformatics pipelines, software and compute. PL will run reproducible pipelines via Nextflow/Snakemake inside containers (Docker/Singularity). Tools: QIIME2/DADA2 for 16S, KneadData/MetaPhlAn/HUMAnN for shotgun, STAR/featureCounts/DESeq2 for bulk RNA, CellRanger-Arc or equivalent + Seurat/ArchR for multiome, Picard/FastQC/MultiQC for QC, MixOmics for integration, limma/edgeR for targeted analyses, and mediation packages for causal testing. Compute resources: institutional HPC with ≥1 TB storage, GPUs for ML tasks as needed. All pipelines and parameter files archived in GitLab with sample manifests.

4.5.13. Quality control and reproducibility measures. Versioned SOPs, training logs, negative/positive controls, mock community spikes for sequencing, pooled QC samples for metabolomics, technical replicates for key assays, cross-lab sample exchange to control batch effects (e.g., shared QC aliquots between PL and CZ), blinded analysis where applicable, and metadata completeness checks. Data release will include raw data, processed matrices, code and documentation to allow independent reproduction.

4.5.14. Ethical, safety and regulatory notes. All analyses performed on anonymised human samples per ethics approvals; animal sample processing follows EU Directive 2010/63. Sequencing of human-derived samples complies with data protection and consent for secondary analyses. Export/import of human data handled under consortium agreements and institutional data-transfer policies. Biosafety level for incoming materials is maintained as per origin WP protocols.

4.5.15. Personnel and responsibilities. WP5 PI (PL): overall coordination and final QC sign-off. Senior bioinformatician (1 FTE): pipeline development and integration. Wet-lab lead (0.5–1 FTE): library prep for bulk/targeted assays and nuclei QC. Metabolomics liaison (0.5 FTE): coordinate SAM/SAH and targeted panels with CZ. Data manager (0.5 FTE): sample manifest, metadata and repository deposition. Statistical modeller (0.5 FTE): mediation/causal models and ML ranking. External service coordination officer (0.2 FTE): manage outsourced multiome library prep/sequencing.

4.5.16. Replication, sample sizes and statistical power. Bulk assays: n≥6–8 per group for key contrasts; single-nucleus: 3 biological replicates per contrast with 5–10k nuclei each; microbiome sequencing: include all pooled donor/group samples from WP3 (≥6 human pools, ≥4 mouse pools). Power reassessed after pilot variance estimates; aim to detect medium effect sizes (Cohen’s d ≈0.6) in expression/accessibility with FDR control.

4.5.17. Deliverables (WP5 endpoints). D5.1 Processed microbiome genotyping datasets (pre/post fermentation) with differential abundance & functional pathway tables. D5.2 Bulk RNA-seq expression matrices and DE gene lists for all contrasts. D5.3 Promoter methylation pyrosequencing tables and global 5-mC summaries. D5.4 Single-nucleus Multiome processed matrices (snRNA + snATAC) with annotated clusters, marker genes, differential accessibility peaks and peak-to-gene links for prioritized contrasts. D5.5 SAM/SAH and targeted metabolite tables linked to sample IDs. D5.6 Integrated multi-omics models, ranked regulatory modules, candidate causal chains and heatmaps; code and pipelines for reproducibility. D5.7 WP5 final report with mechanistic interpretation and prioritized targets for validation.

4.5.18. Timeline (descriptive Gantt-aligned narrative). Month 0–1 Pilot QC on WP2/WP3 inputs and nuclei prep optimisation; Month 1–3 Bulk RNA/DNA library prep and sequencing; Month 2–4 Microbiome re-analysis and metabolomics integration; Month 3–5 Single-nucleus Multiome library prep/sequencing (outsourced/procured) and primary processing; Month 5–7 Multi-omics integration, modelling and report generation; Month 7–8 Data deposition and WP5 closeout. Total WP5 duration: ~8 months (overlaps allowed with upstream WP completions).

4.5.19. Contingencies and alternatives. If Chromium Multiome access delayed, perform snRNA and snATAC separately (single-assay) or outsource entire Multiome library prep; if sequencing budget limits deep sequencing, prioritise key contrasts and increase nuclei per sample for cell-type resolution; if SAM/SAH panels limited in-house, perform targeted assays at CZ or external metabolomics core; if microbial shotgun depth insufficient, rely on 16S + HUMAnN imputed pathways with caution and flag in the report.

4.5.20. References (selected, NCBI style).

4.6.1. Objective. Confirmation studies will aim to functionally validate top candidate genes and pathways identified in WP5, focusing on a) the thyroid-axis (THRB–THRA) pathway via DNMT3A and/or HDAC3, which participate in epigenetic feedback loops within the axis. b) one microbiome metabolite-responsive epigenetic pathway selected from WP5 ranking; perform gain- and loss-of-function assays, targeted epigenetic editing, promoter methylation manipulation and protein-level confirmation in cell lines and/or organoid models, and confirm metabolite sensitivity and rescue experiments with SCFA/Se/I manipulations. Final plan and methods can be adjusted/selected based on WP5 results.

4.6.2. WP6 overview and responsibilities. PL coordinates in vitro genetic and epigenetic perturbations, protein and functional assays, and final mechanistic reporting. CZ provides conditioned media, metabolite panels and phenotypic readouts for rescue experiments. WP6 ends with validated causal evidence for 2 pathways (THRB/THRA and one metabolite-responsive target) with SOPs for perturbation and readouts.

4.6.3. Selection and validation strategy for the two pathways and rationale.
4.6.3.1. Pathway A (THRB/THRA axis): validate ligand-dependent regulation, epigenetic state dependence (promoter methylation, chromatin accessibility) and downstream gene expression network. Conduct perturbation matrix: THRB KO, THRB overexpression, THRA KO, THRA overexpression, and combined perturbations. Readouts: TH target gene panel (including DIO1/DIO2/DIO3), promoter methylation and histone marks at target loci, TRE-reporter activity, DIO enzymatic activity, and downstream cytokine/metabolic markers. Evaluate rescue experiments (e.g., overexpression of THRB in THRB KO cells) to demonstrate causality, in the context of diet-microbiome-metabolites responsive epigenetics.

4.6.3.2. Pathway B: top metabolite-responsive gene/module from WP5 (select based on cross-omic score, effect size and feasibility of assays). Select Target-B using criteria: highest metabolite responsiveness (magnitude and reproducibility), biological plausibility (link to thyroid/epigenome or barrier/metabolism), and tractability (available antibodies/assays and successful perturbation in vitro). Validation pipeline mirrors Pathway 1 (expression, epigenetics, protein, functional readouts, genetic perturbation).

4.6.4. Experimental models.
4.6.4.1. Cell lines: Caco-2 and HT-29 cultured under standard conditions; use differentiated monolayers for barrier and transcriptional assays.
4.6.4.2. Human colon organoids/spheroids: use organoids supplied from CZ biobank; validate viability and baseline marker expression prior to manipulation.
4.6.4.3. Conditioned media/metabolite treatments: use WP3 conditioned media and reference SCFA/metabolite mixes for dose–response and rescue experiments.
4.6.4.4.  T4/T3 -suplemented and -deprived media experiments.

4.6.5. Genetic perturbation strategies. Methods will be selected based on target genes (WP5 results)
4.6.5.2. siRNA knockdown: transient knockdown for pathway B targets when rapid assays preferred; use pool of 3 siRNAs per gene and non-targeting control.
4.6.5.1. CRISPR/Cas9 knockout and CRISPRi: design sgRNAs targeting THRB and THRA coding exons; transfect/viral deliver CRISPR/Cas9 constructs into Caco-2/HT-29; select clonal knockouts where feasible; validate by Sanger sequencing and Western blot.
4.6.5.3. Epigenetic editing: dCas9-TET1 and dCas9-DNMT3A fusions targeted to promoter CpG islands of THRB (or chosen B target) to demethylate or methylate specific loci; validate methylation changes by bisulfite sequencing (PyroMark) and assess downstream expression.
4.6.5.4. Overexpression and rescue: cDNA overexpression constructs for THRB/THRA or mutant constructs (ligand-binding mutants) to rescue phenotype in knockouts/knockdowns.
4.6.5.5. Gene-specific enhancement of native protein translation by targeting 5′-UTRs with dGoligos/eRNA — a strategy complementary to siRNA-mediated gene silencing that increases translation efficiency.

4.6.6. Epigenetic assays and SOPs. 4.6.6.1. Bisulfite sequencing: bisulfite conversion, targeted PCR and PyroMark Q48 quantification for CpGs in promoter regions; acceptance criteria and triplicate technical replicates. 4.6.6.2. If needed: Chromatin accessibility: ATAC-seq on treated cell lines and organoids (50k cells per sample minimum for bulk ATAC); compare accessibility near validated promoters and enhancers. 4.6.6.3. If needed Histone marks: ChIP-qPCR for H3K27ac, H3K4me3 and H3K27me3 at promoter/enhancer loci of interest; use validated antibodies and spike controls.

4.6.7. Functional readouts.
4.6.7.1. RT-qPCR panels: measure THRB/THRA, DIO1/2/3, DNMTs, TETs, HDACs and pathway B marker genes; normalise on housekeeping genes and include technical triplicates.
4.6.7.2. Protein validation: Western blot and ELISA for THRB/THRA and/or selected downstream effectors.
4.6.7.3. Phenotypic assays: proliferation (BrdU or EdU), viability (MTT/CellTiter-Glo), barrier integrity (TEER, paracellular flux of FITC-dextran) in monolayers and organoids, cytokine release (Bio-Plex/ELISA).
4.6.7.4. Rescue experiments: apply defined SCFA mixtures, butyrate, or Se/I forms to test whether metabolite exposure rescues or mimics genetic perturbation effects; include dose–response and time course.
4.6.7.5. Enzymatic DIO activity: use established assays measuring T4→T3 conversion in cell lysates or media (radioactive or LC-MS based) as available in lab SOP; include Selenoprotein activity controls.
4.6.7.6. Reporter assays (THRE luciferase): transiently transfect cells with TH response element luciferase reporter ± Renilla control; treat with conditioned media / T3; measure luciferase activity 24–48 h post-treatment.
4.6.7.7. Organoid functional assays: treat organoids with conditioned media or with genetically modified organoid lines; assess growth, budding, barrier markers, TEER (if applicable), proliferation (Ki67) and apoptosis (cleaved caspase-3), plus molecular readouts above.

4.6.8. Experimental flow and sample allocation.
4.6.8.1. Generate KO/knockdown and matched controls in Caco-2/HT-29.
4.6.8.2. Expose cells to conditioned media (WP3) and defined metabolite mixes for 4–72 h.
4.6.8.3. Collect samples for RNA, DNA (bisulfite), protein and ATAC.
4.6.8.4. For organoids, perform transient perturbations (siRNA or CRISPRi via lentivirus) where possible and run phenotypic assays; send aliquots of processed nucleic acids to PL sequencing cores if required.

4.6.9. SOP extracts (short).
4.6.9.1. CRISPR KO: transfect px459/SpCas9 sgRNA vectors, puromycin select 48 h, single-cell clone, expand, genotype by PCR and Sanger, validate protein loss by Western blot.
4.6.9.2. dCas9 epigenetic editing: transduce dCas9-TET1/dCas9-DNMT3A constructs with sgRNAs targeting promoter CpG island, select populations, assay methylation at 7–14 days post-transduction.
4.6.9.3. ATAC-seq: follow Buenrostro protocol adapted for epithelial cells; transposase QC, library amplification minimised to avoid PCR bias, sequencing depth ~50M reads/sample.

4.6.10. Controls and acceptance criteria.
4.6.10.1. Genetic controls: non-targeting sgRNA/siRNA and empty vector controls.
4.6.10.2. Epigenetic editing: include scrambled sgRNA and mock transduced controls.
4.6.10.3. Functional acceptance: rescue defined as partial or full restoration of expression/phenotype to control levels with metabolite or gene re-expression; statistical significance p<0.05 after multiple testing correction where applicable.

4.6.11. Replication and sample sizes.
4.6.11.1. Cell line experiments: 3 independent biological replicates per condition, technical triplicates for qPCR and ELISA.
4.6.11.2. Organoids: n≥4 biological organoid lines per condition where possible.
4.6.11.3. Validation throughput focuses on depth for two pathways; power calculations done per assay based on effect sizes from WP5.

4.6.12. Data analysis and interpretation.
4.6.12.1. Differential expression and methylation analyses as in WP5 pipelines.
4.6.12.2. Causality tests: perturbation effect size compared to predicted effect from multi-omic models; use mediation framework to test if methylation change mediates expression change following metabolite exposure.
4.6.12.3. Prepare mechanistic diagrams and provide evidence scoring for causal links.

4.6.13. Deliverables.
4.6.13.1. D6.1 Validated causal evidence for THRB/THRA axis role in microbiome/metabolite response (RT-qPCR, bisulfite, ATAC, protein).
4.6.13.2. D6.2 Validated metabolite-responsive target pathway (same assays).
4.6.13.3. D6.3 SOPs for genetic and epigenetic perturbations and assay readouts.
4.6.13.4. D6.4 Final mechanistic report integrating WP5 candidate selection and WP6 validation.

4.6.14. Personnel and responsibilities. 4.6.14.1. WP6 PI (PL) — experimental design and validation sign-off. 4.6.14.2. Molecular biologist (1 FTE) — CRISPR/siRNA/overexpression work.
4.6.14.3. Epigenetics specialist (0.5 FTE) — PyroMark, bisulfite and ATAC assays.
4.6.14.4. Proteomics/assay technician (0.5 FTE) — Westerns, ELISAs.
4.6.14.5. Bioinformatician (0.5 FTE) — data analysis and causality testing.

4.6.15. Equipment and reagents.
4.6.15.1. In-house PL instruments: PyroMark Q48, Bio-Rad CFX96, ChemiDoc, Bio-Plex 200. 4.6.15.2. Cell culture, transfection and viral packaging facilities; selection antibiotics; CRISPR plasmids, dCas9 epigenetic editors, validated sgRNA libraries; bisulfite kits and PyroMark consumables; ATAC reagents and Tn5 transposase; antibodies for ChIP/Western.

4.6.16. Timeline. 4.6.16.1. Month 0–1: final target selection from D5. 4.6.16.2. Month 1–3: generate genetic perturbants and pilot epigenetic editing. 4.6.16.3. Month 3–5: perform functional assays, rescue experiments and ATAC/ChIP. 4.6.16.4. Month 5–6: final analyses, report and SOP deposit.

4.6.17. Ethical and safety notes. 4.6.17.1. Genetic manipulations performed under institutional GMO policies; viral vectors handled per biosafety regulations; relevant approvals documented. 4.6.17.2. Waste and chemical safety per institutional lab safety.

4.6.18. Contingencies. 4.6.18.1. If epigenetic editing fails to produce robust methylation changes, use alternative approaches (targeted methyltransferase recruitment or pharmacologic DNMT/HDAC inhibitors) and document limitations. 4.6.18.2. If organoid transduction efficiency low, prioritise cell line validation and complementary organoid pharmacologic tests.

4.6.19. References (NCBI citation style).

4.7.1. Objectives. This Work Package will include final integration of the nutriome–microbiome–epigenome datasets and publication strategy. The task ensures high-quality coordination between Polish and Czech partners, full compliance with FAIR principles, harmonised cross-WP data governance, transparent reporting, and strategic dissemination of scientific outcomes. WP7 supports the entire project by establishing a unified infrastructure for storing, curating, processing and interlinking all datasets generated across the nutriome–microbiome–metabolome–epigenome axis. It also prepares the final integrated database for long-term reuse, defines the publication strategy, and secures field-leading dissemination pathways to maximise scientific, societal and translational impact.Final plan and methods can be adjusted/selected based on WP4 results.

4.7.2. Project coordination and communication.
Coordination activities include continuous operational communication between laboratories, supervisors, and technical teams in PL and CZ, establishment of shared experimental calendars, and synchronisation of sample transfer between WP1–WP6. Internal review checkpoints ensure adherence to protocols, ethical approvals, timeframes and reporting obligations. Regular online meetings and two in-person workshops (mid-term and final) support critical evaluation, troubleshooting and alignment of analytical strategies, particularly for multi-omics integration and mechanistic validation. Administrative coordination includes maintaining financial transparency, documentation of consumables and sequencing runs, and preparation of deliverables for the funding body.

4.7.3. Data management and FAIR compliance.
4.7.3.1. This WP oversees a unified Data Management Plan covering all raw and processed data generated in WP1–WP6, including plant biofortification data, microbial genotypes, metabolomics, SAM/SAH measurements, bulk RNA-seq, DNA methylation, chromatin accessibility, single-nucleus multiome datasets, and functional validation outputs. All data will be catalogued in a central server (PL host with mirrored CZ node), with version-controlled repositories, traceable metadata, sample identifiers and harmonised ontologies. Data will be formatted according to field standards such as MIxS (microbiome), MINSEQE (RNA-seq), MINQE (qPCR), and established reporting standards for methylation and ATAC-seq. Storage will comply with FAIR rules ensuring findability, accessibility via controlled layers, interoperability through standardised metadata, and reusability via curated documentation. Long-term archiving will be provided using institutional repositories with DOI assignment. Sensitive human data are not collected; all materials remain fully anonymised.
4.7.3.2. AI-Driven and Big Data Analytical Framework. The use of AI and big-data analytics in this project includes integrated processing of large-scale omics datasets, covering transcriptomics, metabolomics, and microbiome profiles. Machine-learning algorithms will reduce noise, detect key biomarkers, and identify nonlinear relationships between diet, metabolites, and cellular responses. Network-based models will map gene–metabolite–microbe interactions, while feature-importance analyses will highlight diet-sensitive regulatory pathways. Model interpretability will be ensured using SHAP values and integrated-gradients–based explainable AI approaches. All algorithms will be benchmarked on cross-validated datasets, and results will be cross-referenced with mechanistic experiments (WP4–WP6) to ensure biological consistency. AI-driven integration of multi-platform data will accelerate analysis and improve the precision of biological interpretation.

4.7.4. Construction of the final interoperable nutriome–microbiome–epigenome database.
A key outcome of WP7 is the assembly of a structured multi-layer database connecting diet composition (I, Se, plant matrix), microbial community structure, microbial metabolite outputs (SCFA and non-SCFA metabolites), host multi-omics datasets and mechanistic validation readouts. Each data tier will be harmonised using common sample IDs, cross-references and integrated annotations. Database architecture will follow modular design so that plant-level, microbiome-level, metabolome-level and host-level modules can be queried independently or joined to reconstruct full causal chains. A dedicated section will store regulatory signatures such as THRB–THRA pathways, DIO1/2/3 response, SCFA-responsive epigenetic clusters, and microbiome-driven chromatin accessibility states. The compiled resource will enable external researchers to interrogate mechanistic modules linking functional foods to microbiome activity and host epigenome.

4.7.5. Dissemination, outreach and visibility of results. Dissemination activities include presentations at international conferences in nutrition, microbiome science, epigenomics and functional foods, as well as contributions to themed symposia and invited seminars. A project website will host non-confidential summaries, graphical overviews of workflows, updates, and links to deposited datasets. Consortium members will disseminate achievements through institutional and public communication channels to increase societal awareness of biofortified foods and microbially mediated epigenetic responses. WP7 will also support early-career training, promoting methodological competencies in multi-omics and mechanistic modelling.

4.7.6. Publication strategy and scientific impact. The project aims to deliver a coherent publication package covering (1) plant biofortification efficacy and micronutrient bioaccessibility; (2) microbial fermentation and metabolite signatures induced by I+Se lettuce; (3) multi-omics characterisation of microbial and epithelial responses; (4) single-cell and single-nucleus epigenomic mapping of diet-responsive pathways; (5) mechanistic validation of priority targets such as THRB/THRA and the top metabolite-responsive gene clusters; and (6) an integrative synthesis describing how functional foods modulate the diet–microbiome–epigenome axis. High-impact journals in nutrition, microbiology, metabolism and epigenetics (e.g., Nature Food, Nature Metabolism, Gut, Microbiome, Cell Reports, iScience, eLife) will be targeted. Emphasis will be placed on the novelty of linking biofortified foods to epigenomic programming via microbe–metabolite signalling, the development of integrated cross-kingdom datasets, and the demonstration of mechanistic causality using organoids and gene-editing approaches.

4.7.7. Summary of WP7 contribution to the project goals.
This WP ensures that the full project arc—from functional food production to microbial fermentation, metabolomics, multi-omics analysis, and mechanistic validation—is captured in a structured, reusable knowledge framework. WP7 maximises scientific visibility, secures long-term value of datasets, and provides robust documentation of how I+Se biofortification modulates microbiome-derived metabolites and reshapes the host epigenome. By integrating all outputs into a coherent system and disseminating them strategically, WP7 strengthens the project’s impact on the fields of nutrition science, functional food biotechnology, microbiome–host interactions and epigenetic regulation.

For further procedural detail, please refer to Section - Methodology and Last Section - Literature.