Send to printer »

Tutorials : Jan 15, 2009 (Vol. 29, No. 2)

Researchers Tackle Tumors One Cell at a Time

Single-Cell Molecular Analysis Platform Helps to Measure and Interpret Heterogeneity
  • Frank Feist

Cancer is characterized by genetic heterogeneity that is caused by tumorigenesis. Several research groups have begun to measure the degree of genetic heterogeneity in tumors to establish the impact of heterogeneity on cancer diagnostics, prognosis, and selection and monitoring of treatment. This tutorial shows how researchers at the Children’s Hospital in Boston, the Klinikum Rechts der Isar in Munich, and The University of Texas M.D. Anderson Cancer Center use Advalytix’ AmpliGrid single-cell molecular analysis platform in their quest to improve cancer diagnostics, prognostics, and treatment.

Laurie Jackson-Grusby, Ph.D., of the Children’s Hospital Boston, is working to improve diagnostic methods and treatment of brain cancer. Her research aims to uncover the mechanisms of epigenetic control that permit self-renewal of cancer stem cells, and to describe phenotypes that characterize cancer stem cells. Ultimately, her research could lead to the development of drugs that specifically target stem cells to stop a tumor’s ability to spread.

Using flow cytometry, Dr. Jackson-Grusby tests potential stem cell markers by sorting stem cells into groups that are enriched for markers of interest vs. groups that do not express those markers. She then tests the ability of either group to form a tumor by injecting those cells into mouse brains, and also tests the stem cell self-renewal properties in cell culture. The idea that tumors originate from rare, individual cells is tested in these experiments by using the AmpliGrid platform (Figure) for mutation and epi-mutation analysis of individual cells. 

Assuming stem cells differ in gene expression from nonstem cells, and that only a sub-population of enriched cells has the ability to self-renew, Dr. Jackson-Grusby’s team uses single-cell genetic analysis on marker-enriched cell fractions along with population-based qPCR to identify candidate genetic and epigenetic changes that might identify tumor-initiating or cancer stem cells.

Since gene-expression analysis destroys the cells, geno-phenotype cannot directly be associated with cell function. Therefore, the group translates expression hypothesis using immuno-enrichment to refine cell populations in each successive experiment. Using this iterative approach, the group expects to define more and more accurate phenotypes of stem cells.

Pancreatic Cancer

Marc Martignoni, M.D., is a surgeon and clinical researcher with Klinikum Rechts der Isar in Munich. As part of a team led by Professor Helmut Friess, director of surgery at the Klinikum Rechts der Isar, Dr. Martignoni’s research focuses on improving outcomes for patients with pancreatic cancer. Pancreatic cancer has an abysmal prognosis, therefore targets for early detection, prevention, and therapy are desperately needed.

Prior research directed by Professor Friess identified several genes of interest in the quest to better understand the genetics of pancreatic cancer. To date, the team has analyzed gene expression using tumor samples that consisted of 1,000 cells or more in order to extract the quantities of genetic material needed for microarray analysis or conventional qPCR.

Dr. Martignoni now wants to investigate how gene expression for those genes of interest differs between individual tumor cells, and how the differences help explain prognosis and response to treatment. Differences between gene expression in tumor cells—tumor heterogeneity—are a direct result of the genetic mutations that occur during tumorigenesis. However, due to the methodological limitations of conventional microarray or PCR technologies, only a few research groups have directly measured tumor heterogeneity, let alone studied the associations between degree of heterogeneity and clinical outcomes. 

In order to achieve the sensitivity needed to obtain robust multiplex gene-expression data from individual tumor cells, Dr. Martignoni is using the AmpliGrid single-cell molecular analysis platform. Initially, he will be working with circulating tumor cells. First he wants to deliver a molecular proof that circulating EPCAM-positive cells from patients with pancreatic cancer that are isolated via flow cytometry do indeed originate from the pancreatic tumor. This process will be accomplished by comparing their expression profile with cells extracted from the primary tumor.

Next he will determine if circulating tumor cells (CTC) can be substratified by their individual expression differences, and whether the degree of heterogeneity is associated with clinical parameters such as prognosis or response to treatment.

If such subpopulations exist, and they can be shown to drive clinical outcomes, genetic analysis of CTCs would present a significant refinement to approaches that rely solely on the number of CTCs detected. Another aspect of Dr. Martignoni’s work will be to determine whether and how chemotherapy changes the expression patterns in CTCs. To what extent are apoptosis markers, which are downregulated in cancer cells, upregulated in CTCs following chemotherapy? Ultimately, the insight gained from measuring heterogeneity may lead to a more personalized approach to selecting the course of chemotherapy most likely to be successful for a given patient.

Breast Cancer

Two groups at The University of Texas M.D. Anderson Cancer Center also work on CTCs, but focus on breast cancer.  Researchers led by James Reuben, M.D., associate professor, and Massimo Cristofanilli, M.D., an associate professor in the department of breast medical oncology, are working to establish the clinical relevance of heterogeneity. This requires studying a large number of clinical specimens of circulating tumor cells in breast cancer patients. “Using AmpliGrid, we plan to genetically fingerprint subpopulations of tumor cells with a convenient workflow,” says Dr. Reuben. 

Using the CellTracks system by Veridex, Dr. Cristofanilli and colleagues were the first to report that in metastatic breast cancer more than five CTCs per sample are associated with a worse prognosis than those samples with fewer than five cells. More recently, in collaboration with Marianna Alunni-Fabbroni, Ph.D., and colleagues at Advalytix in Germany, Drs. Reuben and Cristofanilli have successfully sorted breast cancer cell lines onto an AmpliGrid slide based on a number of surface antigens (EpCAM, CD44, and CD24), the intracellular marker ALDH1, or side population using high-throughput flow cytometry. This was followed by qPCR to identify as many as six genes of interest in a single cell. This technology will be used to identify targets of interest on tumor cells and breast cancer-initiating stem cells, which can be used in the development of targeted therapies for breast cancer.

The M.D. Anderson clinical chemistry diagnostic lab, led by Herbert Fritsche, M.D., has been pioneering the diagnostic use of CTCs, along with Drs. Cristofanilli and Reuben. Using the CellTracks system, which the group had helped validate, they enumerate the number of CTCs for nearly 100 patients per month.

The CellTracks system flags patients’ cells from blood samples as CTCs when they express the EpCAM surface protein and cytokeratin, markers for epithelial cells. Pathologists then evaluate an immunohistochemical image of those cells to count the number of CTCs per sample. Dr. Fritsche and his team have observed that the images of these CTCs are anything but homogeneous: they differ significantly in size, morphology, and fluorescent intensity. 

Now, using AmpliGrid single-cell PCR, Dr. Fritsche wants to test his hypothesis that morphologically defined subpopulations correspond to differences in gene expression similar to genetic heterogeneity observed in breast cancer cell lines. The workflow proposed for these studies includes enrichment and identification of EpCAM positive cells on the CellTracks platform, sorting cells onto AmpliGrid slides via flow cytometry, running a multiplex PCR assay on each cell, and finally analyzing the results on one of the genetic analysis platforms at M.D. Anderson, such as an Agilent Bioanalyzer or Beckman Coulter GeXP. 

The M.D. Anderson teams expect that the single-cell sensitivity of the AmpliGrid platform will produce robust data to measure heterogeneity by comparing the genetic profiles of individual tumor cells. From a clinical perspective, AmpliGrid meets these throughput, automation, and cost requirements with a simple workflow that seamlessly integrates the steps needed for each clinical specimen, namely: cell sorting, sample preparation, and PCR.

“The ultimate goal is to develop powerful tools to monitor the response to treatment, improve early detection, and personalize treatment for our patients,” says Dr. Cristofanilli.