(T0930-06-32) Tumor Chip to Study Dynamics between Immunotherapeutics, Tumor Stroma, and Immune Cells

Purpose: Immunotherapy has emerged over the past decade since the success in treating hematopoietic malignancies. Massive race in clinical translation of immunotherapeutics is currently taking place. For screening and optimizing these immunotherapeutic candidates in the pipeline, however, there is a need for proper models that can recapitulate human tumor architecture and pathophysiological microenvironment. Patient-derived xenograft (PDX) and high-throughput 2D screening are commonly used in vivo and in vitro models for this purpose, respectively. Despite their popularity, PDX mouse models only offer a murine physiological microenvironment to the transplanted human tumors,¹ while conventional 2D monolayer culture model for drug screening provides very limited mechanistic insights.² To address these hurdles, we previously developed a multi-layered tumor chip to reconstruct the spatial configuration of both tumor microvessel and tumor architecture to better mimic the tumor microenvironment for high-throughput cancer drug screening.³ Here, we further develop a new tumor chip platform integrated with the recirculating circuit, so that we can create on-chip immune cell circulation to study how both tumor stroma and immunotherapeutics affects immune cell’s infiltration to the tumor and its efficacy.  As a proof-of-concept, we choose triple negative breast cancer (TNBC) as a model in our study as it is the most intractable cancer type in women, and the outcome of currently approved drugs remains unsatisfactory.⁴

Methods: Figures 1A and 1B present the configuration of our tumor chip setup. Both the tumor chip and recirculating circuit were made by lithography and PDMS molding by following our established protocol.³ Four check valves were designed in our recirculating circuit (Figure 1B) to generate unidirectional flow to the tumor chip unit using a simple syringe pump. Flow pattern generated through the recirculating circuit and its effect on circulating cells were validated using fluorescent beads and human monocytes, respectively. To study T cell infiltration, GFP-expressing human TNBC cell line (MDA-MB-231) and normal fibroblasts or cancer associated fibroblasts were first loaded in Matrigel in the bottom compartment in our tumor chip. On the second day, endothelial cells were then seeded on the top channel to generate tumor microvessel. The recirculating circuit was connected to the tumor chip on the fourth day. The CD8+ T cells were stained with CellTracker CM-DiI and then introduced to the chip. At the endpoint, cells were incubated with 2 µM of NucView®405 Caspase-3 substrate. Fluorescent microscope (Zeiss AxioObserver) or confocal microscope (Zeiss LSM800) was used to characterize T cell infiltration and caspase-3 activity.

Results: From our chip validation, both fluorescent beads and monocytes could follow the flow directed by the four check valves in our recirculating circuit. Recirculation for three days in chip did not compromise the viability of monocytes. To study how tumor stroma and immunotherapeutics would affect the immune cell infiltration to the tumor, human CD8+ T cells were introduced and recirculated in the chip with different stromal compositions (with tumor endothelium, normal fibroblasts or cancer associated fibroblasts) and with different therapeutic treatments (with or without anti-PD-L1 antibody). As shown in Figure 2, the presence of endothelial cells showed negative effects in T cell infiltration, while anti-PD-L1 treatment boosted T cell infiltration. Cancer associated fibroblasts in tumor stroma impeded the T cell infiltration, leading to reduced T cell signal when compared with its counterpart with normal fibroblasts. In addition, caspase-3 activity of the TNBC cancer cells correlated with T cell infiltration trend. In all the groups with normal fibroblasts, cancer cell’s caspase-3 activity was higher than that in their cancer-associated fibroblast+ counterparts. Similarly, anti-PD-L1 antibody treatment enhanced the cancer cell toxicity in all the groups, but this enhancement in the cancer-associated fibroblast+ microenvironment was still lower than what we observed in the conditions with normal fibroblast.

Conclusion: Collectively, we have developed a tumor chip that recapitulates human tumor-stroma microenvironment and enables CD8+ T cell recirculation for studying CD8+ T cell infiltration and cytotoxicity. From this system, impacts of tumor stroma on CD8+ T cell infiltration and cytotoxic activity can be replicated in our tumor chip.  For instance, the reconstructed endothelium in the tumor chip hampers the T cell infiltration to the tumor compartment. Moreover, cancer-associated fibroblasts attenuate T cell infiltration, thereby resulting in reduced cancer cell apoptosis. Treatment of anti-PD-L1 antibody could partially reverse these negative effects and therefore enhances T cell’s infiltration and efficacy. These results confirm that our tumor chip could serve as an in vitro tool potentially suitable for optimizing cancer immunotherapeutics.

References: 1. D. S. Chulpanova, K. V. Kitaeva, C. S. Rutland, A. A. Rizvanov and V. V. Solovyeva, Int J Mol Sci, 2020, 21.
2. C. W. Chi, A. H. R. Ahmed, Z. Dereli-Korkut and S. Wang, Bioanalysis, 2016, 921.
3. C. W. Chi, Y. H. Lao, A. H. R. Ahmed, E. C. Benoy, C. Li, Z. Dereli-Korkut, B. M. Fu, K. W. Leong and S. Wang, Adv Healthc Mater, 2020, e2000880.
4. L. Li, F. Zhang, Z. Liu and Z. Fan, Cancers (Basel), 2023, 15.

Acknowledgements: The authors acknowledge the instrumental support from the nanofabrication facility at CUNY-Advanced Science Research Center and UB Science& Engineering Shared Facilities. This work is supported by the Pershing Square Sohn Cancer Research Alliance (S. W.), the American Society of Cell & Gene Therapy (Career Development Award, ASGCT2021003, Y.-H. L.), the NIH (R21NS133635, Y.-H. L., C.-W. C.; UH3TR002151, K. W. L.) and UB Center for Protein Therapeutics (1181131, Y.-H. L., C.-W. C.).

Figure 1. Design and configuration of our tumor chip and recirculating circuit. (A) A representative image showing the setup of this system. (B) Schematic illustration of connection between the recirculating circuit, reservoir, syringe pump and tumor chip (included an image showing one unit of tumor chip)

Figure 2. CD8+ T cell infiltration and cancer cell’s apoptotic activity in different stromal and immunotherapeutic conditions. (A) A representative image of the tumor compartment with the microenvironment composed of cancer cell (MDA-MB-231), normal fibroblast, endothelial cell and CD8+ T cells, treated with anti-human PD-L1 antibody. MDA-MB-231 and T cell were labeled with GFP and DiI, respectively, and cancer cell’s apoptotic activity was monitored using NucView®405 Caspase-3 substrate (blue). (B) Quantitative analysis of the amount of DiI-stained T cell signals in the tumor compartment after 2-day recirculation. (C) Quantitative analysis of T cell-induced cancer cell apoptosis under different stromal conditions. Results are presented as average ± S.D. (n = 4). Significance was determined using one-way ANOVA with Tukey’s Multiple Comparison post-hoc test.