Gene Expression Analysis in Cancer Cells: A Literature Review

Introduction

Perhaps one of the most critical advances in cancer research was the discovery that certain cancers have a strong genetic component and that cancer is not simply a random occurrence. It is now well established that the presence of certain genes can be highly predictive of whether a person will develop cancer, and this fact is considered in some prevention, detection, and treatment guidelines. However, genetic predisposition is only one factor to examine when assessing whether someone will develop cancer and what the prognosis will be if cancer does develop. In addition to genetic predisposition, exposure to certain environmental factors — most notably carcinogens — can promote the mutations underlying cancer development. The relationship between environmental exposure and genetic predisposition is a complicated one, but understanding how environmental factors impact gene expression in cancer cells may be the key to understanding who develops particular types of cancers and under what circumstances.

Gene expression analysis provides a way of examining the genes in a cell and, more importantly, which of those genes are expressed. Gene expression is not solely a cancer concern. For example, most cells in a person contain that person's entire genome, but not all of that genome is activated in each cell. Instead, the function of the cell determines which parts of the genome are activated, and as a result, cells containing the same genetic code can engage in vastly different functions. The messenger RNA (mRNA) helps determine which genes are expressed in each cell, and, though the exact mechanism is not yet fully understood, it appears that environmental factors can influence how mRNA affects a cell.

Genetic information can also be useful in the preventive treatment of cancers. When a person carries a harmful mutation in either their BRCA1 or BRCA2 gene, that mutation can indicate a higher risk of developing breast cancer as well as other cancers, most notably ovarian cancer. Combined with the patient's family and personal history, the presence of such a mutation may lead a patient to choose a preventive mastectomy or oophorectomy in order to reduce the risk of developing breast cancer. In other cases, the use of hormones, hormone blockers, or preventive drugs such as tamoxifen may be appropriate. Even when preventive surgeries are not available and lifestyle risk factors cannot be altered, knowing that a patient is at increased risk can guide screening guidelines. For example, there is a genetic component to many colorectal cancers, and patients in high-risk groups may begin routine colonoscopies at a younger age than general screening guidelines recommend and undergo them more frequently than low-risk individuals.

One of the most fascinating aspects of cancer and genetic risk is that the presence of high-risk genes does not mean a patient will necessarily develop cancer. Genes are not a determining factor, although they have a tremendous influence on the likelihood of development. Some people with high-risk genetic profiles never develop cancer within the course of a normal lifetime. This has led to the conclusion that the mere presence of a gene is not sufficient to predict cancer development; rather, a combination of genetic predisposition and some type of environmental triggering factor that activates the gene appears to be necessary. Cancer appears, in most cases, to be a combination of environment and genetics, although the environment in question may be the internal environment of the body and not within the patient's control.

Perhaps even more compelling than the role of environmental triggers in largely genetic cancers is the notion that genetics may predispose people to cancers traditionally considered non-genetic. Exposure to the carcinogens in cigarettes is known to be a significant risk factor for various cancers, most notably lung cancer. However, the correlation between smoking and lung cancer is not one-to-one. Some people can smoke heavily their entire lives without developing lung cancer, while others may develop smoking-related lung cancer after relatively minimal exposure to cigarette smoke. The apparent natural resistance of some individuals to particular carcinogens may be explained by the presence or absence of certain genes and the environmental triggers required to activate those genes.

The purpose of this overview is to examine current research on gene expression analysis in cancer cells. The research addresses several ways that environmental factors can determine gene expression in cancer cells or potential cancer cells. Some research focuses on the microenvironment and how laboratory growth conditions can affect a cancer cell's resistance to radiation therapy and chemotherapy. Other research links chemosensitivity not only to the likelihood of genetic expression of cancer, but also to other factors that greatly affect prognosis at diagnosis, including cell adhesion, tumor growth, tumor progression, and invasiveness. Taken together, these factors can help predict who will develop cancer, guide treatment regimens, and help predict metastasis and recurrence rates. Eventually, this research should allow clinicians to more accurately determine when aggressive, high-risk therapies are appropriate and when less aggressive treatments are sufficient — increasing survival rates while reducing the severe side effects that accompany cancer treatment.

Literature Review

The idea that the extracellular matrix (ECM) could have a significant impact on gene expression is not new; it has been established for more than 30 years. The premise is that ECM molecules may provide a means of regulating cells within the context of their microenvironment. This has proven to be true, and it is now well established that the ECM is linked to changes in both tissue and organ structure in both malignant and non-malignant cells (Spencer et al., 2007). In malignant progression, structure is affected on multiple levels: chromatin, nuclear, cellular, and tissue (Spencer et al., 2007). Spencer et al. verified that the ECM controls, or at least significantly influences, both gene expression and tissue function. When cells transition from the 2D monolayer to a 3D environment, their structure and shape change, and this change in structure affects genetic expression (Spencer et al., 2007). Furthermore, these changes are cumulative; the longer cells are exposed to ECM, the greater the changes in cell shape, and the cells undergo morphogenic events and form acinar structures (Spencer et al., 2007). These results make clear that tissue structure affects genetic expression, which in turn affects tissue function and health.

Zschenker et al. began their research from the fundamental premise that the microenvironment is critical to regulating cell behavior and that the nature of the ECM can impact gene expression (2012). Cells grown in 3D growth conditions had previously been shown to have greater resistance to radiation and chemotherapy than cells grown in 2D conditions. Zschenker et al. investigated whether differences in gene expression due to 3D versus 2D growth conditions were linked to DNA repair pathways by comparing basal gene expression of cancer cell lines under both 3D ECM scaffold conditions and 2D monolayer conditions (Zschenker et al., 2012). The experiment was repeated with two human cancer cell lines from different origins to ensure that results were not specific to a single cell line. Their results did not support the hypothesis that 3D growth conditions affected DNA repair pathways, but did reveal a significant difference between cells grown under the two conditions. They found that growth conditions did impact gene expression and that 3D cell cultures could make cancer cells more resistant to both radiation therapy and chemotherapy. Their findings suggested that cell shape and cell-to-cell contact help determine cellular responsiveness to external stress, which can in turn provide cues for therapy resistance. However, the transition to a 3D structure does not appear to inhibit or promote the creation of DNA repair pathways (Zschenker et al., 2012).

Martin et al. did not compare 2D and 3D models, but instead used the 3D culture model to examine breast cancer outcomes across different datasets (2008). Breast cancer is not only prevalent among women but, even with aggressive treatment, remains the second deadliest cancer for women in the United States. It is well established that early detection improves survival rates, though screening guidelines may fail to identify the individuals most at risk. Recurrence is also a serious problem for breast cancer patients. The goal of their research was to examine how a set of genes — which they refer to as a gene signature — could be used to predict the prognosis of breast cancer patients (Martin et al., 2008). They used a 3D cell culture model of non-malignant human mammary epithelial cell morphogenesis as the study baseline (Martin et al., 2008). When focusing on single-gene links to cancer, individual subgroups can dramatically affect the efficacy of genetic predictive power (Martin et al., 2008). However, when studies are expanded to include gene sets rather than specific genes, predictive power grows and appears more fully translatable across populations (Martin et al., 2008).

The 3D signature was able to accurately predict prognosis across different groups, even given the natural variance introduced by different estrogen receptor (ER) status. Both ER-positive and ER-negative groups could receive prognostic information from the 3D model, but each required a different gene set. In ER-positive patients, the prognostic gene set comprised AURKA, CEP55, RRM2, EPHA2, FGFBP1, and VRK1 (Martin et al., 2008). In ER-negative patients, the prognostic gene set comprised ACTB, FOXM1, and SERPINE2 (Martin et al., 2008). Therefore, knowing the genetic makeup of cancer cells can be helpful in determining how to analyze the remainder of a patient's genetic profile in order to develop treatment plans specific to that cancer subtype.

Mutarelli et al. also examined breast cancer cells, specifically hormone-responsive human breast cancer cells (2008). They employed the microarray experimental process because of its ability to provide simultaneous measurements of the transcripts present in cells at various stages (Mutarelli et al., 2008). This measurement can provide a snapshot of cellular activity — not just as a whole, but at each individual stage of development — which could yield insight into eventual disease treatment. They used the Illumina Bead Array platform to examine gene expression in 30–40 replicates on the chip surface (Mutarelli et al., 2008) and identified the genes most likely to affect hormonal treatment. While their research had significant detailed applications, the overarching conclusion was that estrogen impacts genetic expression in ER-positive breast cancer cells. The role a hormone can play in cancer is important, because hormone therapy may be a more appropriate line of treatment for some patients with cancer variants that are resistant to radiation and chemotherapy.

Liu et al. also examined chemosensitivity in cancer cells (2006), focusing on brain tumors in both adult and pediatric patients and building on emerging research demonstrating that CD133 was a marker for some leukemia and glioblastoma cancer stem cells. They began by confirming a high prevalence of the CD133 marker in the glioblastoma cells in their study, verifying earlier research suggesting that CD133 was a marker for certain subtypes of glioblastoma (Liu et al., 2006). They then examined whether CD133-positive cells exhibited greater chemoresistance than other cells. CD133-positive cells were found to be more resistant than CD133-negative cells to a number of chemotherapy agents, including temozolomide, carboplatin, paclitaxel, and etoposide (Liu et al., 2006). In addition, when cancer cells came from patients with verified recurrences of glioblastoma, the cells were far more likely to contain CD133-positive cells, suggesting a link between CD133 expression and cancer recurrence (Liu et al., 2006). Even more significant was the overlapping relationship between CD133-positive status and other genetic factors: CD133-positive cells also show higher expression levels of other factors previously linked to cancer, including BCRP1, MGMT mRNA, and genes inhibiting apoptosis (Liu et al., 2006). These conclusions not only yielded laboratory findings but also suggested a varying course of treatment for some glioblastoma patients, focused on eradicating CD133-positive cancer stem cells.

Hsu et al. focused specifically on chemosensitivity and tumor metastasis, as treatment resistance and the ability to spread throughout the body are both highly linked to the overall danger of a particular cancer (2013). Using the NCI-60 panel — a panel of 60 human cancer cell lines developed by the National Cancer Institute for cancer research — Hsu et al. sought to examine the relationship between invasiveness (IA) and chemosensitivity (2013). They first identified invasion-associated genes in the NCI-60 panel and then tested drugs previously used in chemosensitivity trials. They identified 633 IA genes and, from those, narrowed the set to eight genes correlated with drug-sensitivity profiles. Those gene variants were less likely to relapse after successful cancer treatment. The eight genes identified were EGFR, ITGA3, MYLK, RAI14, AHNAK, GLS, IL32, and NNMT (Hsu et al., 2013). These genes were linked in many ways to both metastasis and recurrence, as they were associated with cell adhesion, tumor growth, tumor progression, migration, and invasiveness (Hsu et al., 2013). Practically, their results showed two distinct genetic signatures linked to cancer cell survival in both lung and breast cancers (Hsu et al., 2013). In a clinical environment, those signatures could help determine which patients are at high risk of metastasis — dictating how aggressive treatment should be — or at risk for recurrence — dictating clinical guidelines for follow-up care. For breast cancer patients, for example, this marker might influence the recommendation of a lumpectomy versus a double mastectomy upon diagnosis.

Not all gene expression research focuses on individual genes or markers. Some newer research focuses on analyzing gene sets. Subramanian et al. examined Gene Set Enrichment Analysis (GSEA) (2005). Rather than focusing on the presence or absence of particular genes or whether certain genes have been activated, GSEA focuses on the fact that genes frequently work in combination, and therefore examines sets of genes that share common traits. What they found is that GSEA analysis can reveal significant information about genetic links to cancer that is not revealed — or may appear less significant — when concentrating on single-gene analysis (Subramanian et al., 2005). Their results suggest that there are certain biological pathways for specific cancers and that genes sharing common traits can have a cumulative impact on the course of development, metastasis, treatment, and recurrence of those cancers. This focus on gene sets helps examine how environmental factors affect genetic expression by looking beyond the presence of individual genes and examining biological processes present across entire gene networks — such as transcriptional programs, metabolic pathways, and stress responses — that might have only a minimal impact on a single gene (Subramanian et al., 2005). GSEA is a powerful and flexible tool; Subramanian et al. created an initial database with 1,325 gene sets, but any number of gene sets could be analyzed using the GSEA approach (2005).

Conclusion

Although it is not classified this way, cancer is, by definition, a genetic disease. Normal cells mutate through oncogenesis to become cancer cells. This is why doctors can examine cells for precancerous changes that indicate the cells are mutating toward malignancy. The process is not spontaneous or immediate, even in aggressive cancers. What these changes do is reprogram cells so that they no longer engage in their normal functions but instead engage in rapid cell division that leads to tumor growth. As tumors develop, the hardiest cells gain a greater share of space, so that most cells in a tumor come to carry the same genes. Cancer is therefore a disease involving genetics, even if it does not fall under the rubric of diseases conventionally considered genetic.

It should come as no surprise, then, that gene expression analysis is proving to be an invaluable tool in the study of cancer. Gene expression analysis has helped identify which genes predispose individuals to cancer and how different gene clusters may promote the expression of those genes. Already, genetic analysis has helped shape screening guidelines and preventive protocols in high-risk groups. As gene expression analysis continues to advance, one can only expect continued progress in treatment regimens and prevention protocols.

References

Hsu, Y., Hsuan, Y., Yuan, S., Yu, S., Lin, C., W., G., Yang, P., & Li, K. (2013). Genome-wide analysis of three-way interplay among gene expression, cancer cell invasion and anti-cancer compound sensitivity. BMC Medicine, 11(106). Retrieved October 28, 2013 from BioMed Central website: http://www.biomedcentral.com/1741-7015/11/106

Liu, G., Yuan, X., Zeng, Z., Tunici, P., Ng, H., Abdulkadir, I., Lu, L., Irvin, D., Black, K., & Yu, J. (2006). Analysis of gene expression and chemoresistance of CD133 cancer stem cells in glioblastoma. Molecular Cancer, 5, 67. doi:10.1186/1476-4598-5-67. Retrieved October 28, 2013 from Molecular Cancer website: http://www.molecular-cancer.com/content/5/1/67

Martin, K., Patrick, D., Bissell, M., & Fournier, M. (2008). Prognostic breast cancer signature identified from 3D culture model accurately predicts clinical outcome across independent datasets. Retrieved October 26, 2013 from PLOSOne.org website: http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0002994

Mutarelli, M., Cicatiello, L., Ferraro, L., Grober, O., Ravo, M., Facchiano, A., Angelini, C., & Weisz, A. (2008). Time-course analysis of genome-wide gene expression data from hormone-responsive human breast cancer cells. BMC Bioinformatics, 9, S12. doi:10.1186/1471-2105-9-S2-S12.

Spencer, V., Xu, R., & Bissell, M. (2007). Extracellular matrix, nuclear and chromatin structure, and gene expression in normal tissues and malignant tumors: A work in progress. Advances in Cancer Research, 97, 275–294. doi:10.1016/S0065-230X(06)97012-2

Subramanian, A., Tamayo, P., Mootha, V. K., Mukherjee, S., Ebert, B., Gillette, M., Paulovich, A., Pomeroy, S., Golub, T., Lander, E., & Mesirov, J. (2005). Gene set enrichment analysis: A knowledge-based approach for interpreting genome-wide expression profiles. PNAS, 102(43), 15545–15550. doi:10.1073/pnas.0506580102

Zschenker, O., Streichert, T., Hehlgans, S., & Cordes, N. (2012). Genome-wide gene expression analysis in cancer cells reveals 3D growth affects ECM and processes associated with cell adhesion but not DNA repair. Retrieved October 28, 2013 from PLOSOne.org website: http://www.plosone.org/article/info:doi/10.1371/journal.pone.0034279