The intestine plays a crucial role in nutrient digestion and absorption, making it a key site for assessing the biological effects of food. Traditional animal-based food safety and efficacy studies are costly, time-consuming, ethically questionable, and low-throughput. Microphysiological models, such as organ-on-a-chip technologies, offer a promising alternative. These models more accurately mimic the physiological microenvironment of an organ compared to standard in vitro models, which often lack specialized epithelial cell types. The researchers had previously developed a bioengineered intestinal tubule containing various intestinal cell types, better representing the intestinal heterogeneous cell population than traditional Transwell™ models. This tubule mimics the three-dimensional structure of the intestine, including villi, and is suitable for transepithelial transport assessments. This bioengineered model has been validated for assessing the safety and efficacy of nutrition, examining parameters like epithelial barrier integrity (using zonula occludens-1 (ZO-1) as a marker), cell viability, brush border enzyme (alkaline phosphatase) activity, and immune markers (nitric oxide (NO) and cytokines). Plant breeding plays a vital role in improving crop yield and quality. Lettuce (*Lactuca sativa*) is a globally significant leafy vegetable, with breeding efforts focused on disease resistance, yield, and other marketable traits. Wild lettuce relatives (*L. serriola*, *L. saligna*, *L. virosa*) represent valuable genetic resources for improving cultivated lettuce, offering potential benefits in terms of disease resistance and other traits. These wild relatives often contain higher levels of bitter-tasting compounds. While metabolome-assisted selection combined with machine learning is emerging as a useful tool in plant breeding, assessing the potential toxicity of new breeding material remains a challenge. This study aimed to utilize the bioengineered intestinal tubule to evaluate extracts from commercially available lettuce and compare them with extracts from different *Lactuca* species, including wild relatives, to assess potential intestinal effects.

The existing literature highlights the limitations of traditional animal models for food safety and efficacy testing. Organ-on-a-chip technology is presented as a superior alternative, providing a more physiologically relevant in vitro system. Several studies demonstrate the successful application of similar microphysiological models in biomedical and drug research. The specific bioengineered intestinal tubule used in this study has been previously validated for nutritional assessments. The literature also emphasizes the importance of lettuce breeding and the potential benefits of utilizing wild relatives to improve cultivated varieties. Previous research has characterized the metabolite profiles of different *Lactuca* species, revealing species-specific patterns in their chemical composition, including bitter compounds. The challenges associated with assessing the toxicity of breeding material are discussed, emphasizing the need for alternative methods.

The study employed a bioengineered intestinal tubule, previously developed by the authors, composed of human Caco-2 cells seeded on a hollow fiber membrane within a 3D-printed chamber. This model incorporates various cell types representative of the intestinal epithelium. To validate the system, extracts from commercially available lettuce varieties (butterhead, lollo rosso, red iceberg, and stalk lettuce) were initially tested. Subsequently, extracts from laboratory-grown *L. sativa* cultivars (Salinas and Olof), *L. serriola*, *L. saligna*, and *L. virosa* were evaluated. Plant materials were snap-frozen, ground, and extracted. Bioengineered intestinal tubules were exposed to extracts for 24 hours. Several parameters were assessed: inulin-FITC leakage (epithelial barrier integrity), ZO-1 expression (tight junction integrity), cell viability (PrestoBlue assay), cell attachment (nuclear count), alkaline phosphatase activity (brush border enzyme function), and immune markers (IL-6, IL-8, and NO) via ELISA and Griess reaction. Immunofluorescent staining was performed to visualize goblet cells (MUC2) and ZO-1. Data analysis involved outlier removal, statistical testing (t-test and one-way ANOVA), and comprehensive cluster analysis (K-means and hierarchical clustering) to group the *Lactuca* lines based on their intestinal biological efficacy profiles.

Extracts from commercially available lettuce did not significantly affect the intestinal epithelial barrier integrity, cell viability, cell attachment, or alkaline phosphatase activity. However, extracts from the wild species *L. saligna* and *L. virosa* caused a significant decrease in epithelial barrier integrity, cell viability, cell attachment, and alkaline phosphatase activity. *L. sativa* cv. Salinas extracts showed a minor reduction in cell attachment and alkaline phosphatase activity. While IL-6, IL-8, and NO levels were detected, background levels in the extracts themselves made it difficult to definitively interpret their changes as responses from the intestinal cells. Cluster analysis, based on the scaled mean values of multiple parameters, revealed three distinct clusters: 1) commercially available lettuce (negligible effect); 2) *L. sativa* cultivars and *L. serriola* (minimal effect); and 3) *L. saligna* and *L. virosa* (adverse effects on cell viability and barrier integrity). This clustering aligned with the degree of domestication.

The bioengineered intestinal tubule model successfully differentiated epithelial cells morphologically resembling the in vivo organ and showed low levels of immune markers. The findings highlight the potential of the bioengineered intestinal tubule as a screening tool for assessing the intestinal safety and efficacy of lettuce germplasm. The contrast between the effects of cultivated and wild lettuce species suggests that breeding practices may have unintentionally reduced the potential for intestinal toxicity. The absence of a clear effect on immune markers (IL-6, IL-8, NO) may be due to cross-reactivity of antibodies with plant proteins. Future studies incorporating gene expression analysis could provide further insights into the immunomodulatory effects of lettuce. The differences in intestinal efficacy are likely linked to differences in phytochemical profiles, with potentially protective compounds like fumarate and myo-inositol being more abundant in cultivated varieties. Conversely, sesquiterpene lactones, prevalent in wild species, may be responsible for the observed toxicity. Further research is needed to establish a direct link between the metabolic profiles of *Lactuca* species and their intestinal effects.

This study demonstrates the successful application of a bioengineered intestinal tubule model to evaluate the intestinal biological efficacy of lettuce extracts. Wild relatives of cultivated lettuce exhibited adverse effects on intestinal parameters, whereas commercially available and domesticated lines showed minimal effects. This approach offers a valuable tool for evaluating the safety and nutritional properties of lettuce breeding materials, facilitating the development of varieties with improved nutritional properties and reduced potential toxicity. Future research should focus on further characterization of the phytochemicals responsible for the observed effects and the detailed mechanism of action.

The study focused on a limited number of lettuce accessions and environmental conditions. The assessment of immune markers was complicated by the presence of these markers in the plant extracts themselves. Further studies with a larger number of accessions and more controlled growth conditions are needed to validate the findings and explore the influence of environmental factors on the observed effects. The study primarily focused on acute effects of lettuce extracts; long-term effects were not investigated.