Development of Microfluidics Based Biosensors in Cancer Detection and Therapy

Abstract

Cancer can be broadly classified as either solid tumors, which grow in organ tissue and form a mass, or liquid tumors, which circulate throughout the body via the bloodstream. Both types are characterized by the uncontrolled growth of abnormal cells. Solid tumors can occur in any part of the body, such as breast cancer, while liquid tumors are found in blood, bone marrow, or lymph nodes, including leukemia. The application of microfluidic biosensors has shown tremendous promise in cancer detection and therapy in the past three decades. Microfluidics is widely used in personalized medicine, point-of-care testing, and even microfabrication, making them attractive alternatives to routine diagnostic testing. Recent advances in the design and development of microfluidics devices have made it possible to miniaturize conventional biochemical laboratory protocols into a microchannel system, which has emerged as an efficient and cost-effective tool.

In this thesis, I detail the development of novel microfluidic biosensors for metastatic risk detection and combination cancer treatment.

Metastasis is a complex process in which primary tumor cells migrate and establish secondary tumors by invading specific tissues or disseminating blood and lymphatic systems. Early detection of metastatic disease significantly reduces mortality and improves overall survival to facilitate timely treatment and intervention. As a result, developing techniques for early detection of cancer metastasis is critical. Most patients' treatment, however, is frequently delayed due to the complex procedures required for routine diagnostic analyses, which necessitate extensive and highly specialized equipment and highly trained personnel.

Firstly, to investigate the application of microfluidic biosensors in cancer detection, we demonstrated the development of a smart nematode microfluidic-based biosensor (SmartC^M) that could rapidly detect and assess metastatic status, enabling a low-cost, non-invasive approach to personalized medicine. The ultrasensitive smart biosensor (SmartC^M) uses only optical microscopy to visualize nematode movement, allowing for real-time (< 60 min) and high-throughput screening with low sample volume requirements (1 ml). We found the potential of the nematode to differentiate conditioned media from metastatic and non-metastatic breast cancer cell lines. SmartC^M was then validated in the urine samples of non-metastatic (n = 14) and metastatic (n = 16) breast cancer patients, as well as healthy controls (n = 6). We demonstrated that the SmartC^M platform has high sensitivity (93.3%) based on our unique chemotaxis index thresholds (CI _MCF-7; metastasis > 2). The CI value in urine samples from metastatic breast cancer patients was significantly higher than in non-metastatic breast cancer patients (2.9-folds; CI _MCF-7). Raman spectroscopy profiling revealed significant differences in the urea and L-proline metabolic profiles of metastatic and non-metastatic urine samples. The SmartC^M platform was enhanced with an amplification system based on a Z stage with a chip holder, droplet lens, and smartphone to promote automated operations for portable and high throughput analysis. Metastasis risk was indicated by a CI _MCF-7 value greater than 2. We concluded that nematode preference for metastatic urine samples was associated with lower urea and L-proline concentrations, implying that these compounds could be biomarkers for metastatic disease. Our low-cost, label-free biosensor allows for rapid evaluation of patient cohorts during routine metastatic risk screening, supplementing clinical assays for immediate intervention or promoting remote testing and monitoring.

Cancer is controlled by sophisticated mechanisms that are intrinsically difficult to understand since there is a limitation in direct observation of interference of biological molecules. Hence, methods for cancer modeling are of considerable interest for understanding cancer pathophysiology and developing advanced therapeutic strategies. The two-dimensional cell culture method has some drawbacks, including the possibility of cell morphology and polarity changes, which might cave to interruption in cellular-extracellular communication. Moreover, the monolayer structure of two-dimensional cell culture leads to unrestricted availability to reach the optimum medium, oxygen, and signal molecules. Significantly, the accessibility of the nutrients, oxygen, or/and signal molecules for cancer cells in a living organism could be changeable due to the inherent structure of the tumor. While existing models are limited to 2D systems, recent research has shown that 3D models can potentially influence cell phenotype and gene expression, especially regarding targeted signaling pathways. The microenvironment has been shown to influence blast cell drug sensitivity, as previously reported for anti-leukemia agents.

Secondly, to investigate the application of microfluidic biosensors in cancer combination treatment, we evaluated the effect of short-term fasting, which is a technique to reduce nutrient intake for a specific period on blast cell aggregates in acute lymphoblastic leukemia (ALL) patients using a microfluidic-based blast cell aggregate assay (MA^B) to mimic the environment infiltration of blast cells into the bone marrow. The bone marrow microenvironment can hinder the efficacy of chemotherapy, leading to treatment resistance and relapse. The MA^B platform creates a compact, porous microcavity that better reflects the constraints of the blast cell microenvironment, allowing for the study of drug sensitivity. The MA^B platform provides a high-density growth state, increasing blast cell viability. It is a suitable platform for evaluating the impact of dietary interventions and combination treatment on cancer cells. Our results show that fasting selectively kills blast cell aggregates within 72 hours without affecting healthy leukocytes. We also demonstrated that fasting significantly reduces biomarkers associated with cell metastasis and drug resistance, including CD44, Vimentin, and cell deformability. When combined with doxorubicin, fasting reduced the proportion of phenotypic leukemia stem cells more than doxorubicin alone. Mechanistically, fasting regulates leukemia relapse and remission through epithelial-mesenchymal transition (EMT) biomarkers, inhibiting blast cell development and self-renewal through TWIST1/2.

Additionally, we observed a significant increase in the expression level of ZEB-1, a biomarker associated with reduced stemness of leukemic cells, with fasting. Further analysis revealed that the regulation of CD44, TWIST1, and Vimentin expression in blasts was linked to the signal transducer and activator of the transcription 3 (STAT3) signaling pathway based on siRNA transduction of STAT3 knockdown. Finally, we confirmed that patients with lymphoblastic leukemia and low baseline glucose levels had more prolonged overall survival and a lower relapse rate than those with higher glucose levels (n = 58). Our findings suggest that fasting-based therapies have broad and differential effects and could be a viable treatment option for selectively killing blast cells to improve patient outcomes in ALL. The modulation of EMT biomarkers through the STAT3 signaling pathway can be a potential mechanism underlying the anti-leukemic effects of fasting-based therapies.

In summary, our study showed that microfluidic devices are used in diagnostics and disease modeling, making microfluidic systems a potent tool, making the demand for unique small biomedical tools expand and benefit more patients.

Date of Award	25 Aug 2023
Original language	English
Awarding Institution	City University of Hong Kong
Supervisor	Bee Luan KHOO (Supervisor)