After identifying SLC17A2 with its high liver-specific expression, and considering our foundational SynNotch system, we needed to further investigate whether this protein has a transmembrane structure and possesses an ideal extracellular domain for antibody development.

For this purpose, we performed TMHMM-2.0 prediction for transmembrane helix structure and AlphaFold structure prediction for murine SLC17A2. TMHMM-2.0 prediction (Figure 1) indicated that murine SLC17A2 has 10 transmembrane helices. AlphaFold structure prediction (Figure 2) revealed it possesses an ideal extracellular structure, with amino acids 48-84 being relatively conserved, representing an ideal region for antibody development

Combined predictions from TMHMM-2.0 and AlphaFold indicated that the amino acid sequence spanning residues 48 to 84 of SLC17A2 constitutes its extracellular domain, representing an ideal region for scFv development. Subsequently, we provided this sequence to Xi'an Haina Biotechnology Co., Ltd. for the generation of an anti-SLC17A2 scFv. Our Syn-M INPUT module was precisely built upon this anti-SLC17A2 scFv, ultimately enabling successful liver-specific activation.

After constructing Syn-M according to our final design, we co-cultured it with AML12 cells at a 1:2 ratio for 24 hours to activate Syn-M. Subsequently, activated Syn-M cells were isolated using magnetic bead sorting. The collected cells were lysed with Trizol and sent to Beijing Biomarker Biotechnology Co., Ltd. for transcriptome sequencing to analyze the differences between activated Syn-M and unmodified macrophages.

Principal Component Analysis (PCA) revealed that samples from activated Syn-M and the control group each clustered tightly together and were clearly separated on the PCA plot (Figure 3A), indicating minimal intra-group variation but substantial inter-group differences, suggesting a significant alteration in the gene expression profile of the cell population. Volcano plot analysis of all relevant genes identified differentially expressed genes (Figure 3B), with numerous genes upregulated (red) or downregulated (blue), and 21 genes showing particularly significant differences (yellow), indicating that co-culture with AML12 induced substantial changes in the gene expression of Syn-M.

Enrichment analysis of these significantly different genes (Figure 4) revealed notable differences in pathways related to TNF, TGF, NF-κB, and other inflammatory pathways, suggesting that Syn-M was effectively activated and promoted the expression of genes associated with inflammatory pathways.

To further validate the expression of downstream effector molecules in these pathways, we analyzed relevant factors (Figure 5A). The data showed significantly increased expression of pro-inflammatory cytokines in activated Syn-M, which can exert effective anti-tumor activity. To verify whether activated Syn-M retains the phagocytic activity characteristic of M1 macrophages, we examined phagocytosis-related factors (Figure 5B). The results indicated a moderate upregulation of these factors, suggesting that Syn-M can be activated to exhibit M1-like phagocytic function.

Additionally, to verify whether activated Syn-M possesses the ability to recruit immune cells, we analyzed factors associated with T cells and neutrophils . Significant changes were also observed, indicating that this function has been successfully activated. In summary, compared to unedited macrophages, activated Syn-M exhibits distinct pro-inflammatory and anti-tumor properties.

To quantitatively evaluate the infiltration capacity and targeting efficiency of Syn-M within the liver cancer microenvironment, we established a spatial computational model based on the following hypotheses and methods:

We primarily employed three main modeling methodologies.

1.Spatial Modeling

Liver tissue was simplified into a uniform two-dimensional grid, ignoring the three-dimensional branching structure of blood vessels. Macrophages, oxygen, and other substances were introduced via boundary injection to simulate capillary input.

2.Chemical Field Modeling

• The oxygen field was constructed based on Fick’s diffusion law, incorporating diffusion terms, linear oxygen consumption by cancer cells, and distance-based terms from blood vessels.
• Chemokine fields accounted for spatial distributions of both tumor cells and macrophages.
• Regional transitions (normal, growth, hypoxic, necrotic) were smoothly approximated using Sigmoid functions.

3.Cell Behavior Modeling

• Macrophage migration was modeled using a random walk mechanism biased by chemical gradients. Migration probabilities were computed for eight possible directions.
• Upon SynNotch activation, immediate behavioral feedback—P65 overexpression and SIRPα knockdown—was implemented, ignoring biological delays.
• SLC17A2 expression was restricted to hepatocyte-enriched regions (normal and growth zones), and Syn-M activation was modeled as a switch-like threshold response.

Simulation Results

1.Microenvironment Field Simulation

We successfully simulated physiologically realistic oxygen and chemokine fields

• Oxygen concentration was highest near blood vessels and decreased significantly in cancer-dense regions.
• Chemokine gradients were centered around the tumor, effectively guiding macrophage migration. Regional partitioning achieved smooth transitions through continuous field modeling, consistent with the spatial characteristics of hepatocellular carcinoma lesions.

2.Migration and Activation Dynamics

Simulations demonstrated that Syn-M cells originating from vascular regions efficiently migrated along chemokine gradients into tumor regions. Post-SynNotch activation, behavioral switching (P65↑, SIRPα↓) immediately influenced migration paths, confirming the feasibility of the engineered design.

Conclusion

This study employed a multi-field coupled grid model and gradient-based migration algorithms to dynamically simulate the behavior of Syn-M within the hepatocellular carcinoma microenvironment. The model demonstrated both rationality and scalability in chemical field construction, regional partitioning, and cellular response mechanisms, providing a robust foundation for further mechanistic investigations and optimization of macrophage-based targeted therapies