Roll-Your-Own-Tumor Technology for Cancer Research May Boost Personalized Treatments

December 1, 2015
Colin G. Evans, PhD

Details of a new technology for studying tumor metabolism have been published in Nature Materials, whereby a cellulose strip containing tumor cells was wrapped around a metal core and incubated, mimicking conditions consistent with in vivo tumors.

Alison P. McGuigan, PhD

Details of a new technology for studying tumor metabolism have been published inNature Materials, whereby a cellulose strip containing tumor cells was wrapped around a metal core and incubated, mimicking conditions consistent within vivotumors. If the technology proves successful, then it could facilitate drug screening and the identification of new combination therapies, as well as boost the design of personalized treatments for patients with cancer.

The research team was led by Alison P. McGuigan, PhD, professor of chemistry in the Department of Chemical Engineering and Applied Chemistry, Institute of Biomaterials and Biomedical Engineering, University of Toronto in Canada.1This team collaborated with Christian Frezza, PhD, program leader, Medical Research Council Cancer Unit, University of Cambridge, UK, and Bradly G. Wouters, PhD, professor, Departments of Radiation Oncology and Medical Biophysics, director, Hypoxia and Microenvironment Program, Ontario Cancer Institute. According to a press release, this collaboration ensured that the tool enables the types of experimental tests that are most needed by cancer biologists to ask cutting questions and translate their findings into benefits for patients.2

The teams developed this new technology because of the well-known drawbacks of studying tumor cells in petri dishes, where the two-dimensional culture cannot replicate the 3D consequences of tumor structure, namely the impacts of differential diffusion. Even 3D models are difficult to analyze. The team decided to develop a model system that enables very rapid (less than 10 seconds) isolation of cells from known locations within a tumor model. They developed a strip, within which tumor cells grow in a collagen medium. This cellulose strip is then wrapped around a metal core and incubated. The strip can be unrolled safely and quickly leaving the cells intact and available for detailed analysis. “It’s simple enough that one could teach an undergrad to do it in a week,” said McGuigan.1,2

Structure of the Strip

The researchers took porous cellulose scaffold strips with a thickness of 34.19 µm ± 4.38 µm and infiltrated them with tumor cells (human ovarian cancer cell line SK-OV-3, 1 x 108cells mL-1) suspended in a type 1 collagen gel (3.2 mg mL-1). This collagen was selected because it is abundant in tumors and other work has demonstrated that when grown in this gel, tumor cells behave as they doin vivo.3The strips were then incubated in the medium for 24 hours to allow tumor cells to become established. There was no difference in morphology between cells grown in the collagen gel alone versus those in the strip. Cellulose was chosen because tumor cells do not interact with it or grow along its fibers, and it exerts no impact on tumor cell survival. The cell matrix phase interlocks with the cellulose skeleton, ensuring integrity of the strip, allowing rolling and unrolling of the strip with no damage to the cellular matrix. The authors emphasizing that this means it is possible to confidently target specific populations of cells for analysis taken from distinct regions of the engineered tumor. The strip can be unrolled in less than 1 second for snap freeze-drying or fixation of cells for snap shot analysis or live cells can be removed from different regions for secondary assays.1,2

On a Roll

Following the incubation, the strips were rolled around a metal cylinder, 6 mms in diameter that was impermeable to oxygen. They wrapped the strips to generate 6 layers, typically with a total thickness of about 200 µm. They selected this becausein vivowork has shown that oxygen gradients will be generated over length scales of 100 µm to 200 µm.4The resulting “tumor roll for analysis of cellular environment and response” (TRACER), was then submerged in culture medium. Tumor cells thus received oxygen and nutrients by diffusion, and the authors predicted this would lead to the generation of gradients with higher levels in the outer layers than the inner layers.1

Within the TRACER, the researchers found spatial features known to be present inin vivotumors. After incubation for 72 hours, the innermost layer contained fewer viable cells with an increase in cell death. Proliferation of cells was also reduced in these layers, all probably due to limits on oxygen and nutrient availability. Similarly, when they treated the TRACER with doxorubicin, the responses of the cells in the layers were comparable with what occurs in tumors. The known penetration limitations of the drug were reflected in the decreasing concentrations of the drug from the outermost to innermost regions. The concentration of doxorubicin reached a maximum between layers 3 and 4, a depth (approximately 100 µm -150 µm) similar to its penetration in mouse xenografts. The response to radiotherapy showed a similar gradient of effectiveness, with tumor cells in the innermost layer experiencing significantly less cell death, mimicking therapy resistance. The authors stated that, “Together these data demonstrate that cells in the TRACER system exhibit relevant behaviors consistent with in vivo tumors.”1

Oxygen and Metabolite Dynamics

Using the quantification of EF5 (2 — [2-nitro-1H-imidazol-1 yl]-N- [2,2,3,3,3 – pentafluoropropyl] acetamide) binding, to detect oxygen gradients, they found no regions of hypoxia in the nonrolled configuration, but when rolled up into TRACERS, gradients were observed, due to cellular consumption (varied according to cell type), from the outer to innermost layers. By 6- and 12-hour time points, the deepest layers (4-6) were extremely hypoxic (O2<0.1%). However, between 12 and 24 hours, oxygenation increased in all layers and were stable between 24 and 48 hours. The authors commented that this is consistent with the known adaptive hypoxia response mediated by hypoxia inducible factor (HIF), which decreases oxygen consumption. They demonstrated this as the mechanism using cells expressing a short hairpin RNA (sh-RNA) against HIF1a. Analysis of these TRACERS showed significantly higher levels of hypoxia in layers 4 to 6 after 12 hours. There was no cellular adaptation to hypoxia at 12 hours. Overall this again showed TRACER properties resemble those found in tumors.1

Metabolite concentrations and their relationship to the oxygen gradients (ie, spatial variations) were investigated using layer-specific liquid chromatography—mass spectrometry. They analyzed 88 metabolites in each layer. They found that 23 metabolites whose concentrations across the various TRACER layers significantly correlated with levels of hypoxia. Specifically, cells in the TRACER maintained metabolic signatures according to layer. Cells in the innermost layers had signatures typical of hypoxic cells such as increased glycolysis, deregulation of the tricarboxylic acid cycle, and a reduction of fatty acid oxidation. The authors emphasized, “Of note, the metabolic changes in glycolysis and mitochondrial metabolism seem to be, by and large, HIF independent, suggesting that the underpinning pathways might be under biochemical, rather than genetic control.” They suggest their results also show that HIF controls oxygen availability by regulating the activity of oxygen-dependent enzymes (eg, xanthine oxidase, indoleamine dioxygenase).1

Future and Applications

This new technology is easily constructed and disassembled and very versatile. It allows rapid isolation of cells from different locations for analysis. If cell migration needed to be studied, cells could be labeled for monitoring. Conversely if cell migration was not needed, barrier layers may be incorporated. Different cell types can be added, different gradients for molecules such as lactate and glucose may be studied to assess the impact of additional molecules on cell metabolism in cancer. It will also facilitate drug screening and the identification of new combination therapies.1“It’s very translatable and transferable to other labs,” said McGuigan. “We definitely want others to use it, because the larger the community, the more applications we will discover.”2There are also implications for personalized medicine, “The idea would be to take a patient’s own cells and create copies of their tumor,” said McGuigan. These copies could then be subjected to various treatments and analysed by the simple unrolling process, providing information about what is likely to work best for that specific patient.2


  1. Rodenhizer D, Gaude E, Cojocari D, AP. A three-dimensional engineered tumour for spatial snapshot analysis of cell metabolism and phenotype in hypoxic gradients.Nat Mater. 2015 doi:10.1038/nmat4482.
  2. A tumor that can unroll: Engineers create new technology for understanding cancer growth November 23, 2015.
  3. Paszek MJ, Zahir N, Johnson KR, et al. Tensional homeostasis and the malignant phenotype.Cancer Cell. 2005;8(3):241-254.
  4. Carmeliet P, Jain RK. Angiogenesis in cancer and other diseases.Nature.2000;407(6801):249-257.