646 - Modeling a Physiological Relevant in vitro System for Nutritional Studies in the Preterm Infant

Publication Number: 1623.646

Omar Hayatt, Weill Cornell Medicine, New York, NY, United States; Liliette I. Quintana, Weill Cornell Medicine, New York, NY, United States; Divya Mandalaywala, Weill Cornell Medicine, New York, NY, United States; Joanne E. Brown, Beth Israel Deaconess Medical Center, Dracut, MA, MA, United States; Steven Freedman, Beth Israel Deaconess Medical Center, Boston, MA, United States; Camilia R. Martin, Weill Cornell Medicine, New York, NY, United States

Background: Animal studies have advanced understanding of nutrient digestion and host interactions, but physiological differences limit translation to humans. In vitro intestinal models offer relevance yet often exclude digestion, applying undigested lipids directly to cells and risking misinterpretation of hydrolysis and response. A physiologically based digestion-hydrolysis model may better represent neonatal Docosahexaenoic Acid (DHA) and Arachidonic Acid (ARA) processing and cellular recovery under clinically relevant stress.

Objective: To develop an in vitro predigestion model capable of achieving over 60% DHA/ARA hydrolysis within 24 hours and to evaluate cellular confluence and viability following exposure to predigested and unhydrolyzed supplements. We further aim to assess cellular recovery after a scratch-induced stress test over 24 hours.

Design/Methods: A predigested (PD) DHA/ARA supplement was generated under relevant intestinal conditions (pH 7.5, porcine lipase, bile) and incubated at 37 °C for 24h; and Undigested (UD) without. Thin Layer Chromatography (TLC), determined hydrolysis. Lipid meals were sonicated with lipid free BSA and combined with de-lipidized FBS (dFBS) MEM media. Caco-2 cells were treated with dFBS control (CN), PD or UD meal media for 0-48h. IncuCyte wide field live cell imaging was used quantify percent confluency (n=12) and caspase-3/7 expression for percent viability (n=12). Scratch-wound assays were used as a stress recovery factor (n=8) and used 0h measurement as baseline for relative wound closure. Differences between CN, PD, and UD groups were evaluated using the Kruskal-Wallis test, post-hoc.

Results: The predigestion model achieved 71% total lipid hydrolysis, including FFA, MAG, and DAG fractions after 24h incubation (Fig. 1), and maintained >70% confluence and viability suitable for downstream assays (Fig. 2). Confluence and viability were comparable across conditions through 24h, though PD and UD showed reduced viability versus CN at later timepoints (Fig. 2). Under scratch-wound stress, PD showed significantly greater fold change in wound confluency than CN and UD at 2-12h (p < 0.05-0.0001; Fig 3A). No differences were observed until 24h, when CN showed slightly higher viability (p < 0.05).

Conclusion(s): This predigestion model achieved high hydrolysis and maintained sufficient viability for subsequent comparative nutrient processing in vitro studies. Non-predigestion of nutrient substrates results in disparate biological outcomes that may not demonstrate in vivo human physiology, emphasizing predigestion in vitro is needed to reflect preterm intestinal lipid responses.

Percent Hydrolysis of Predigested ARA/DHA Supplement Meal
Figure 1: Modeling of Predigestion System Prior to in vitro Experiments. Hydrolysis confirmed by bands of Free Fatty Acids (FFA), Monoglycerides (MAG), Diglycerides (DAG) over all components and Triglyceride Base (TAG). Hydrolysis determined by Thin Layer Chromatograph (TLC), timepoints between 0-24 Hours.

Percent Confluence and Viability of Control (CN) de-Lipidized Culture Media, Predigested (PD) meal and Undigested (UD) meal on Caco-2 Cells.
Figure 2: Threshold for Predigestion Cell Culture Experiments. Caco-2 cells were incubated with CN, PD, and UD media to assess whether ≥70% confluence (Panel A) and viability (Panel B) could be maintained for downstream stress assays. Box-and-whisker plots demonstrate that all groups remained above the 70% threshold, confirming suitability for subsequent experiments.

Relative Fold Change in Confluence of Scratch Wound in Caco-2 Cells
Figure 3. Wound Gap Closure in Scratch Assay. Panel A: Box and whisker plot of relative wound closure with control and predigested/undigested meal in MEM media during scratch assay. Caco-2 cells exposed to various media from timepoints between 0-24h. Panel B: Cellular recovery after scratch-wound was visible through wound closure in 10x wide field imaging of Caco-2 cells in dFBS control (CN), Predigested (PD) and Undigested (UD). Blue Mask indicates wound, yellow mask indicates epithelial cell growth after initial scratch.