Accelerating cancer drug discovery: the advent of Raman spectroscopy

Three-dimensional imaging allows for faster analysis of candidate compounds

By Michael A. Miller and Jian Ling     image of PDF button

Michael A. Miller (right) is manager of Materials Development in SwRI's Materials Engineering Department. A specialist in materials chemistry, Miller has contributed significantly to the development of advanced analytical methods and theoretical models to determine the kinetic and structural disposition of matter in complex systems. Jian Ling is a senior research engineer in the Bioengineering Department. At SwRI, Ling has evaluated a number of medical devices including a noninvasive blood pressure monitoring system and an artificial heart device and has been involved in advanced biomedical signal processing technology.

Scientists at Southwest Research Institute (SwRI) are attempting to alleviate the bottleneck of cancer-fighting compounds competing for federal approval by applying three-dimensional cellular imaging techniques to assess the pharmacological value of candidate drugs more quickly.

The creation of an estimated 15,000 new compounds per year is increasing steadily as computational methods for drug design and chemical synthesis become more sophisticated. However, only one compound every two to four years is deemed effective enough to promote to clinical trials.

Drug discovery today is a multiphase operation. First, a candidate drug must be engineered to act upon a specific cellular target, starting with a minimal understanding of how the drug's hypothetical chemical structure might interact with the biology and chemistry of the target. Often molecular engineering of this sort leads to the creation of a family of theoretically potent new compounds.

Initial synthesis or isolation of a new family of agents typically is achieved only in milligram quantities. These agents are tested against tumor cell cultures in vitro. If an agent inhibits a specific tumor cell line, it is then tested in small experimental animals to determine its ability to differentiate between normal and cancer cells and to inhibit tumor growth. The next phase of testing is done in larger animals to determine and quantify the toxicity of the new drug. If the toxicological profile is favorable, only then is the drug considered for clinical trials.

This approach to cancer drug development has been astonishingly successful even though the molecular target upon which a new agent acts is most often speculative. The cost is also high, in excess of $100 million to develop a single clinically viable anti-cancer drug. Nevertheless, several types of malignancies now can be cured or at least mitigated with new drug therapies.

Efforts to accelerate drug discovery must focus on screening methods that can rapidly assess whether an investigational drug has a desirable effect on malignant cell lines in vitro. An early opportunity to determine the drug's affinity to the target and its uptake, metabolism and distribution within the cell permits researchers to make informed decisions about what, if any, chemical or structural modifications are needed to improve its potency.

A light-scattering technique called Raman spectroscopy lies at the core of the Institute's research. Discovered by C.V. Raman in 1928, Raman spectroscopy has evolved into a powerful tool to analyze the structure and composition of matter. The information obtained from Raman spectroscopy complements that obtained from infrared spectroscopy, long used in molecular assessment.

However, the instrumental arrangement, and the rules that govern the transition of light quanta from one vibrational (and rotational) state to another, are distinct. These distinctions lead to the practical experimental advantages for Raman spectroscopy over infrared spectroscopy in specific applications - like direct imaging in this case.

In the global illumination instrumental arrangement used in modern Raman imaging, a beam of laser light is expanded and focused onto a target object, whereupon the Rayleigh component of light scattered back is rejected by a holographic filter. A specific Raman component is selected and the image that remains is projected onto a two-dimensional charged coupled device detector.

Application to cellular imaging

Understanding how new drugs metabolize and distribute at the level of a single cell is important to drug development and evaluation. This level of understanding, if it is convenient to attain, can be used early in the drug discovery process to define the efficacy or potency of the drug and, ultimately, to decide whether animal testing is warranted.

Cultivating the desired abnormal cells outside the organism and analyzing the fate of the drug within individual cellular compartments is the ideal method for cellular evaluation of drug action. For example, discovering that an investigational drug that was thought to target a specific cellular component, such as the cell membrane, has instead entered the cell nucleus, would be an important and potentially beneficial discovery. A fundamental requirement of this sort of spatially resolved in vitro evaluation is that the analytical tool or device must be minimally invasive to the extent that the response of the cell or the cell itself is not altered or destroyed in the process.

Three-dimensional Raman imaging fulfills many analytical requirements by providing a non-invasive and chemically specific analysis of a heterogeneous, multicompartment structure such as a living cell. Conventional microscopy or fluorescent imaging cannot accomplish this.

The principal limitation of fluorescent imaging is that the drug of interest often must be chemically modified with a fluorescent marker or probe, altering its behavior and chemical characteristics. No such modifications are required in the Raman technique, since the unique combination of vibrational states of the drug molecule itself is interrogated, not the electronic transitions of a chemically attached marker.

A global illumination scheme is one of several methods being further developed at SwRI to image Raman signatures of anti-cancer agents at the level of a single tumor cell. In contrast to point illumination, which is too slow to accurately capture intracellular drug actions, global illumination interrogates the entire field of view of the cell using a near-infrared laser.

The scattered photons from each volume element of the cell are filtered simultaneously to remove the Rayleigh scattering component, radiation caused by elastic collisions between laser photons and molecules of matter. A second filtering device then selects a single Raman vibrational mode of the drug to project its intensity onto a two-dimensional detector, resulting in a direct two-dimensional image of drug distribution in the cell.

The anti-cancer drug Taxol® is shown undergoing a molecular vibration that appears at 1,002 cm-1 in the Raman spectrum. Each frame shows the minimum and maximum displacements of molecular vibrations computed for this mode. Vectors indicate the atoms involved and the magnitude and direction of their motion. Computations at this level enable researchers to predict and correlate precisely structural variances in the molecule with the observed spectral features.

A case study

Taxol® (paclitaxel) is undergoing clinical trials for treatment of a variety of malignancies, including breast, ovarian, non-small-cell lung, gastric and melanoma cancers. Taxol is antimitotic; in other words, it prevents the usual method of cell division. Taxol promotes the assembly of ultrafine cylindrical structures found in the cytoplasm of cells. These structures, known as microtubules, are involved in the shape of a cell and in the transport of molecules through the cell membrane. Promoting uncontrolled microtubule assembly leads to cell death because the dynamic reorganization of the microtubule network normally needed for cell survival and mitosis (cell division) is inhibited. These alterations to the microtubule network are so profound that global deformations to cell shape are commonly observed.

A vibrational mode appearing at 1,002 cm-1 was selected for imaging from the detailed Raman spectrum of Taxol. This mode constitutes the most intense peak in the spectrum and reflects carbon-carbon stretching vibrations of the backbone structure. A Raman image of the cell recorded at the selected mode in the absence of drug served as the control image.

Breast cancer cells (MDA 435) were cultivated in a normal saline solution on a specially coated petri dish. The coating was designed by SwRI to facilitate Raman cellular imaging and to passively immobilize the cells without inducing death. A drug delivery system was used to expose the cultivated cancer cells to a Taxol solution. After exposing the cells to the Taxol solution, the extracellular solution was exchanged with normal saline. Raman images of a single cell were immediately recorded at the selected vibrational mode, at different depths, using an optical sectioning technique.

Image enhancement of spatially resolved Raman signatures is a crucial aspect of these cellular imaging studies. The spatial heterogeneity of the laser light source and the peculiarities associated with how accurately the spectrometer hardware gathers and transforms the Raman signal contribute to the quality and precision of the image. Laser-induced light emissions, such as luminescence or fluorescence, from the cellular matter itself can further compromise the weak Raman component of the image.

SwRI researchers have placed a considerable emphasis on developing and applying image processing techniques especially suited to correct the imperfections of the instrumentation and to remove unwanted background emissions. These techniques were applied to restore the Raman signatures of Taxol from the cellular images.

As expected for a multicompartment structure such as a living tumor cell, the most salient lesson from these images is that the distribution of Taxol within cellular matter is not homogeneous. The highest concentrations are found in and slightly beyond the cell membrane, with minimal distribution into the nuclear compartment. These observations are consistent with the accepted action of Taxol because the microtubule structures to which Taxol binds occur near the cell membrane.


This study and others like it indicate that Raman cellular imaging may lead to an advanced method to determine the pharmacologic effect of drugs at the single-cell level. Before it can be considered for routine use in medicinal studies, a number of obstacles must be overcome, the most significant being improvement in the detection sensitivity of the method.

In the present case study, cancer cells were exposed to Taxol concentrations one order of magnitude greater than what can be achieved safely in practice, either in animal models or in human clinical trials. Imaging the Raman signatures of a drug at physiologic concentrations within the microscopic compartments of a cell is particularly demanding because the scattering cross section of drug molecules at these sparse levels is very weak compared with the cellular matter itself. However, incremental refinements to the experimental techniques and instrumentation are being explored at SwRI, as are sophisticated visualization algorithms. The use of a more powerful near-infrared laser is one example of modifications to improve detection.

Molecular computational methods also play an essential role in advancing the field of Raman cellular imaging. By integrating this experimental approach with additional advancements in molecular computations, it will become possible to correlate accurately variances in the drug molecule with variances in the corresponding Raman spectral characteristics. Thus, theoretical assumptions about the pharmacologic action of new drugs can be rapidly validated or invalidated.

Comments about this article? Contact Miller at (210) 522-2189 or or Ling (210) 522-3953 or

Published in the Summer 2001 issue of Technology Today®, published by Southwest Research Institute. For more information, contact Maria Stothoff.

Technics Summer 2001 Technology Today
SwRI Publications SwRI Home