Most scientists would agree that the most physiologically relevant systems are the best models for experimentation. And, using the same reasoning, if these are the best models for studying disease, then they will also be the most accurate models for the identification of therapeutic targets and drugs specific for those targets. This is exactly the undisputed reason for using primary cells in the laboratory. Primary cells, however, are more than just a sound cell model in the laboratory. The advantages are particularly clear when studying diseases such as cancer, diabetes, diseases involving the skin, bone marrow and liver. This far, we have discussed the differences between primary cells and cell lines in general terms. Below we examine specific examples of the differences between primary cells and established cell lines and then outline the direction that primary cell use is taking academia and biotechnology.
Primary Cells vs Cell Lines
The fact that the number of reports comparing primary cells to cell lines has risen over the last few years testifies that scientists are not only noticing these changes but are actively looking for alternatives.
A recent report by Alge et al undertook the task of comparing primary human retinal pigment epithelial (RPE) cells to immortalized RPE cell lines. Using 2-D electrophoresis and MALDI-TOF peptide mapping they compared the protein expression profile of the two most prominent PRE cell lines, ARPE-19 and hTERT-RPE to primary early passage retinal pigment epithelial (ep RPE) cells.
This is an interesting example since users of RPE cells understand that once RPE cells are passaged in culture they differentiate and thereby, display different morphologies, lose cell specific markers and resemble mesenchymal cell types. Researchers tried to control for such differentiation by establishing the immortalized cell lines, ARPE-19 and hTERT-RPE in hopes of maintaining cell cultures that closely resembled primary RPE cells. Since primary human RPE cells are difficult to obtain due to limited tissue availability from human donor eyes, and since research heavily depends on ARPE-19 and hTERT-RPE cell lines, Alge et al made a comprehensive comparison and documented some interesting findings. ARPE-19 cells showed differences in the microtubule cytoskeleton and also in the proteins that are involved in proliferation and cell death. hTERT-RPE, on the other hand, showed differences in cell migration, adhesion, and extracellular matrix formation and cell polarization.
While some differences were anticipated, the specific differences outlined in this study give some insight as to how these cell lines respond to different stimuli and the potential functional implications this might have on experimentation. Furthermore, if primary human RPE cells are limited, and the immortalized cell lines are the next most feasible alternative, knowing these differences should enhance the interpretations of any data gained from experiments.
Cancer cell lines provide another example where established cell lines deviate from their primary tumor samples. This is a more complex issue in that tumor cells carry and inherent genetic instability that promotes variation, both in the body and in culture. The National Cancer Institute (NCI) initiated a drug-screening program using human transformed cell lines with this notion in mind. They reasoned that using a large array of human transformed cell lines in a high-throughput screening system would increase the identification of viable drug molecules. The fact that the transformed cell lines contained substantial differences than native tumors was considered a positive since native tumors are also unstable and prone to changes in therapeutic targets. This approach, then, would allow all the variations to be represented in the cell line platform. While the theory sounds reasonable most drug screening programs are moving away from this approach since it’s usefulness has been challenged.
We can use prostate cancer cell lines as an example. While numerous prostate cancer cell lines exist, they are all derivatives of the three main established prostate cancer cell lines, PC-3, DU 145 and LNCaP. These cell lines represent the standard lines used in most laboratories. Aside from the differences we already discussed, we are starting to understand that human cancers of the same tumor type can vary widely among individuals. In other words, spontaneously generated cell lines from the excision of metastatic human tumors represent only a subset of tumor characteristics from that particular individual patient. This makes generalizations more difficult and less accurate especially when we rely on these generalizations for drug development. This could very well be one reason why drugs with proven efficacy for a given tumor type only work in small patient populations with the same tumor type.
This has been very elegantly shown from microarray analysis of breast cancers. That body of work has demonstrated that breast cancer is term that covers a variety of breast malignancies that can be distinguished according to their genetic profile. The differences are so striking that a patient that is diagnosed with breast cancer today will first undergo identification of their tumor type to determine the therapy required. A HER+ breast cancer will mostly likely be treated with tranztuzumab while ER/PR+ breast cancers will be treated using anti-estrogen therapies. We would not realize these differences among breast tumors had scientists not taken efforts to make the necessary comparisons.
Current trends using Primary Cells
Primary cells are at the edge of gaining quite a following among laboratories both in the academic and biotechnology/pharmaceutical sectors. Specifically, primary cells are being used in cancer research to validate the results obtained from cell lines, predict the response and resistance based on individual patient primary samples to select the most likely drug candidates, test for the most selective/toxic compounds against the tumor, and finally as models for identifying new drug targets and validating new drugs. Aside from cancer, primary keratinocytes are also being used to explore options for skin diseases such as psoriasis, toxicology testing for cosmetics, and most recently to evaluate drugs against biological warfare. The fact that human keratinocytes are easy to isolate and also behave very well in culture makes this a great primary cell model to study skin physiology and toxicology.
The use of primary cells is also extending to areas of research that focus on the bone marrow, bone development and diabetes. Specifically, the Cell Culture System was developed and used to screen the available glitazones that act against diabetes. This system features primary adipocyte samples from 10 different donors which are cultured in the same well to allow for simultaneous screening with limited well-to-well variability. The first experiments allowed the identification of two patient subgroups, the high responders and the low responders.
These are just a few brief examples of how research is evolving to incorporate more relevant and more accurate models using primary cells. The reasons for these efforts are not for purely altruistic since the cost of developing drugs is astronomical with conservative estimates around 800 million dollars for every drug that makes it to the marketplace. However, drug development cost also covers all the targets that never made it to the market. The reality is that these costs are finally absorbed by the consumers and insurance companies. The current thinking is eager to use primary cells as models to quickly determine which drug candidates carry false hope.