Drug Effects on Cancer Cells Attachment

More than 10 million new cancer cases will be diagnosed worldwide over the next year. Although this illness is better understood every minute, intensive research is still ongoing to find more efficient treatments with less side effects and toxicity.


Anti-cancer therapies are aimed at two major targets: killing cancer cells in tumors, mainly by inducing apoptosis, and preventing cancer propagation by creating metastases. Metastasis is a complex phenomenon believed to involve many steps, which are targeted by drug discovery strategies: the escape of the cancer cells from the primary tumor, their transport through the body’s vessels, and their attachment to an organ where a secondary tumor may be formed.

It is this final step that has interested IME’s BioElectronics Program in the article published in the Biosensors and Bioelectronics journal (Chen Yu et al., vol. 23, pages 1390-1396, 2008). During this process, the extra cellular matrix (ECM) and its constitutive proteins such as laminin, fibronectin, and collagens, play a crucial part. The interaction between the cancer cells and the ECM will decide the arrest of the tumor cells in a target organ to spawn metastases.

Techniques to study cancer cells adhesion to the ECM, and how a particular drug candidate can influence this process are readily available in cell biology laboratories. They are mainly based on reporter cell-based assays using fluorescent labels to estimate under the microscope, the number of cells that have been able to attach to a surface coated with various constituents of the ECM, depending on the characteristics studied (Fig 1).

These assays can be easily be implemented in the high-throughput automatic facilities of drug discovery labs. However, the use of labels to identify the interaction means that the cells or proteins are modified from their natural states, pushing the cellular models and the in vitro assay farther away from the relevant in vivo conditions faced by the tumor cells. Moreover, although great progress has been seen in live cell imaging microscopy recently, it is still not easy to obtain dynamic output and behavior out of standard optical adhesion assay. It usually means having to stop the experiment at different times to collect the data.

Label-free sensing
In order to overcome the shortcomings inherent to the use of reporters, a number of label-free methods have been developed (reviewed by Matthew A. Cooper, Drug Discovery Today, vol, 11, N.23/24, 2006, pages 1068-1074). Among those, Quartz Crystal Microbalance (QCM) can quantify the number of cells that have attached to the coated sensor through their influence on its resonating frequency. Surface Plasmon Resonance (SPR) spectroscopy or microscopy is also able to detect cell attachment through an optical, yet label-free means. However, those two systems are still at a development phase as far as cell-based assays are concerned.

On the contrary, electrical cell-surface impedance sensing (ECIS) is now well-developed and commercially available in microfabricated products. It is based on the fact that living cells adhering to an electrode change its impedance, which can be recorded in real-time. Products by Applied Biosystems Inc., ACEA Biosciences Inc. or MDS Inc. for example have been widely used to study various aspects of living cells, such as cell motility, spreading, adhesion and growth. Most of these systems comply with the standard microtitre plate format, inherent in drug discovery and can perform real-time monitoring in situ, enabling dynamic kinetic studies of cell attachment behaviors. However, these designs are limited to the study of large cell populations.

Populations at “a few cells” resolution
In the last decades, the microtechnologies pioneered by the microelectronics industry have spread the gospel of miniaturization to a wide variety of areas. Biology is no stranger to this trend, where automated high-throughput screening became a new paradigm in drug discovery. From the large volumes of a Petri dish, biology labs have turned to microlitre samples in multi-well plates. Therefore precious samples such as primary cultures and stem cells can now be used in automated screening platforms, reaching single cell sizes for some particular studies.

However, decreasing the number of cells in the sample of a test is limited by the fact that the response of a cell population is computed through various cell-cell interaction and signaling. In order to obtain a significant and meaningful response, a sample of a few tens of thousands of living cells is usually used.

However, the fact that cell populations are highly heterogeneous has attracted a lot of attention recently. A tumor, for example, is still a “black box” and only a few specific cells inside it may create a metastasis. Therefore biology and drug discovery would greatly benefit from the use of systems that are able to study single cells, or a few cells, but in the context of a large cell population and in real-time.

Heterogeneity could then be apprehended, and the impact of the drugs on the cells of interest assessed in greater details. High-Content Screening (HCS) for example (reviewed by Haney, S.A et al. in Drug Discovery. Today, vol.11 pages 889–894), the multiparametric image-based analysis of cell populations, enables the study of cell populations at the resolution of a single cell, characterized by multiple end points. It has become a very popular platform in biology and drug screening strategies. Implementing such a scheme within label-free methods has been the driving force of the developments on impedance measurement systems carried out at IME.

Current commercial systems and published developments in impedance-based cellular assays use relatively large electrodes (250µm-2mm), thus tackling large cell populations. This is due to the fact that the change in the impedance detected by an electrode is linked to the area of that electrode that is covered by the cells. Single cells are much smaller than the electrode used, and a few cells adhering to the surface would lead to very small or even undetectable signal.

In order to be able to study a few cells per electrode, IME has opted to use smaller electrodes (60µm in diameter). By creating an array of 22 electrodes inside one chip (microelectrode array chip or MEA) that can all be recorded individually, it is possible to greatly increase the sensitivity of the approach. The signal from each electrode is correlated to the number of cells that have attached on it. Adding together the signals recorded by many electrodes relates the number of cells attached on the selected surface of the chip, giving a statistically significant and meaningful response of the cell population.

This approach allows getting multiple end points, out of the response of each small group of cells, in one chip. It corresponds to achieving statistically significant results for one drug concentration for example, in one well, whereas conventional strategies can provide only one value per well and need a few wells to gather sufficient data. This reduces further the amount of precious samples that are used. Finally, to enable in situ and real-time measurements on the cells in their best environment at 37°C in a humid incubator, a compact system was designed.

Compact system
The real-time impedance monitoring system (Fig 2.) comprises three main elements. A computer (a) controls an electrochemical workstation (b) that measures impedance on the MEA chip located inside the incubator (c), to provide the best culture conditions to the living cells. The entire system can be fitted to a bench top size. Located inside the incubator, a test jig electrically links the individual electrodes of the MEA chip to the measuring workstation.

To ease chip handling and operations such as cell loading on the electrode array or drug addition to the culture media of the cells, the test jig presents an array of pins that directly corresponds to an array of conducting pads on the chip, leading to the individual electrodes. When the chip is loaded inside the jig, the pins are lowered into contact with the pads, bridging the connection to the workstation. This enables rapid connections and positioning, and only the chip is disposable, which greatly reduces operating costs compared to PCB wire-bonded systems.

The chip is fabricated through standard silicon microtechnologies. The MEA contains 22 individual electrodes of 60µm in diameter separated by a 160µm pitch, completed by bigger reference and counter electrodes. The surface of the chip is tuned to mimic an ECM environment by coating it with a layer of collagen I to which the cells attach. A biocompatible plastic chamber is taped onto the surface to provide a rectangular well containing 100µl of cell suspension, which is equivalent to a 96-well plate. A biocompatible gas-permeable membrane is taped to the top of the chamber cell incubation inside the incubator,

Attachment studies are aimed at identifying the effect of drugs on the adhesion of living cancer cells present in solution to the ECM. In this compact system, the experiments involved mixing the drug with the cells in a tube, loading them in the chip, which was then inserted in the test jig inside the incubator. In a matter of a few minutes, the cells could be monitored in real-time, via the change of the impedance signal at the time of measurement versus the first measurement, corresponding to a situation where no cell is attached to the surface, over a few hours. The total sensor response is calculated by adding together the impedance change of every electrode (six in this case).

Sticky integrins
Integrins are transmembrane receptors that are implicated in various aspects of cell attachment to the ECM, such as adhesion to ECM proteins, ECM organization or cell motility. Antibodies, which bind specifically to different integrins, can be used as model systems of a drug targeting this important family of proteins. Moreover they can help to identify the specific integrins involved in the attachment of cancer cells. The NCI-H460 cells used in this study are non-small lung cancer cells, well known in cancer research and used especially to trigger tumor growth in in vivo mouse models.

Dose-response data (Fig 3.) has been obtained for 2 antibodies (against β1- and α2β1-integrin). Both proteins seem to be involved in cancer cell attachment as the impedance signal decreased with antibody concentration, which was reported for the first time for the sub-unit α2β1. Moreover, the individual effects of the antibodies have been compared and show that anti α2β1-integrin would be more potent than its counterpart anti-β1 in preventing attachment. The IC50 value, defined here as the concentration of antibody to decrease the impedance signal by half, was lower for α2β1-integrin.

Completing those results, the publication in Biosensor and Bioelectronics reports a thorough investigation of different parameters that characterize the impedance monitoring system and its use as a drug discovery tool, bringing more insight into how label-free high content assays can fulfill their great potential to help biologists and researchers in drug discovery.