Skip to main content
. 2021 Aug 11;10:e67106. doi: 10.7554/eLife.67106

Figure 4. Application of colon intestine-Chip as model of CEA-TCB-mediated adverse effects.

(A) IHC of human colon and duodenum tissue stained with anti-CEA (brown coloration) demonstrating difference in regional expression. (B) Representative immunofluorescent micrograph depicting CEA expression in the epithelial compartment of the Colon Intestine-Chip and Duodenum Intestine-Chip. (C) Average cell surface expression of CEA within 3D organoids and chips at day 8 of culture (n=3). (D) Colon-Chip epithelial channels were administered with PBMC treated with/without low and high-affinity (CEA(Lo) and CEA(Hi)) TCB (0.1–10 µg/mL), or Non-targeting (NT) TCB (10 µg/mL). Co-culture was maintained under flow for 72 hr. Quantification of immunofluorescent images collected live indicate multiple clusters of PBMC settled throughout epithelial structures. Statistical analysis was conducted by one-way ANOVA and was defined as *p<0.05 and ***p<0.001. Errors bars represent ± SEM. (E) CD69+ Activation of CD8+ T cells of harvested PBMC measured by flow-cytometry (n=3± SEM). (F) Heat map of multiplex cytokine panel from epithelial channel supernatants. Data (D–F) from terminal endpoint 72 hr after administration (n=3). (G) Colon- and Duodenum-Chips were administered with PBMC with low and high-affinity (CEA(Lo) and CEA(Hi)) TCB treatment from 0 to 10 µg/mL, along with Non-targeting (NT) control. Flow cytometry analysis of harvested PBMC from chips 72 hr post-treatment to measure levels of activated CD69+CD8+ T cells (n=3± SEM). (H) Multiplex cytokine analysis of supernatant collected from epithelial channels of Colon and Duodenum-Chips after 72 hr of treatment (n=3± SEM).

Figure 4.

Figure 4—figure supplement 1. Anti-tumor potency and animal cross-reactivity of CEA-targeted TCBs.

Figure 4—figure supplement 1.

(A) Both TCB molecules displayed concentration-dependent binding to human CEA-expressing gastric cancer cell line MKN45. CEA(Hi) TCB showed stronger binding, consistently with its higher affinity for CEA. Treatment with both TCBs led to concentration-dependent (B) MKN45 cancer cell killing and (C) T-cell activation, with the higher affinity molecule producing a stronger effect. (D) Effect of CEA-targeted TCBs on tumor progression in CD34+ HSC humanized NSG mice (HSC-NSG mice), engrafted with tumor-forming MKN45 cells. (E) Assessment of binding of CEA(Lo) TCB to CEA derived from humans or cynomolgus monkeys. Data are represented as mean values, with SEM.

Figure 4—figure supplement 2. Experimental outline of Intestine-Chip model.

Figure 4—figure supplement 2.

Colon Intestine-Chips were cultured following the organ-chip protocol until day 4. On day 4, frozen PBMCs were thawed and rested overnight at 37°C prior to introduction to the chip. On day 5, PBMC-TCB dosing solutions were prepared and administered to the epithelial channel of the chips. For days 6–8, the epithelial channel was administered with media containing TCB but without additional PBMC. Apical and basal outflow collection and PBMC (Cell Tracker) live imaging was performed at 24 hr, 48 hr, and 72 hr post PBMC-TCB dosing. At the 72 hr terminal time point PBMCs were collected from chips for T-Cell analysis using flow cytometry. Chips were later fixed with 4% PFA and stored for further immunofluorescence imaging.

Figure 4—figure supplement 3. Intestine-Chip CEA expression, comparison to conventional models and target-independent PBMC activation of CEA-targeted TCBs.

Figure 4—figure supplement 3.

(A) Diagram of Colon and Duodenum-Intestine chip seeding, beginning with fragmented primary human organoids seeded into the epithelial channel of the chip. Primary intestinal endothelial cells, either colon or small-intestinal depending on corresponding epithelial tissue, are seeded into the vascular channel and the chip is cultured to maturity under flow and mechanical deformations. (B) Immunofluorescence staining of CEA and nuclei in colon or dudodenum chips versus matched colon or duodenum organoids. (C) Immunohistochemistry analysis of CEA (brown coloration) in a conventional, static model of the intestinal barrier: organoid-derived intestinal cell seeded on ECM-coated transwell membranes. (D) Flow cytometry-based quantification of CEA binding sites expressed by intestinal barriers cultured in transwells. High CEA-expressing cancer cell lines MKN5 and A549 serve as positive controls. (E) Treatment of PBMC with CEA-targeted TCBs in the absence of target does not induce activation, confirming target-dependent mode of toxicity observed in the Intestine-Chips.