Sunday, March 11, 2007
Microarray and Cancer Research
For those of you who are going to IADR meeting in New Orleans in the next two weeks, feel free to stop by and check out my poster (Abstract #2568, "Role of Calprotectin in Head and Neck Squamous Cell Carcinoma").
In this post I'm going to talk a bit about microarray technology in cancer research. Please feel free to give any comments or suggestions.
It is clear that DNA microarray technology is an extremely powerful tool in studying gene expression profiles of a single cell or a set of cells. Unlike gene microarray, where you're studying the presence or absence of gene(s) or changes in the gene copy numbers, or exon microarray, where you're studying splice variances, expression microarray studies changes in the gene expressions at the mRNA level.
Gene expression microarray is a bit more complicated in terms of experimental design, which can become costly and even useless if not properly controlled. For those of you who are new with this technology, it is important to remember that expression microarray cannot answer all of your project's questions. Microarray is just an assay and not an experiment in itself. It is important that when designing an experiment for a microarray study you need to keep things simple. Complicated design will result in complicated data that can make your data analysis a nightmare. The most important step in a microarray study is the result, which depends entirely on the data mining and data analysis steps, which further depends on the accuracy of the experimental design.
When designing an expression microarray study you need two comparative conditions: control condition and the experimental condition. Otherwise, your expression data is meaningless. Next, you have to make sure that there are exactly two variables in your comparison, such as normal and cancer samples. That means, in the case of comparing normal and cancer samples, you need to make sure that both the normal and cancer samples come from the sample tissue type and the same anatomical origin. This is highly critical because gene expressions are highly regulated in a tissue-specific manner. That is, the same gene can express differently in different cell or tissue types. So, when studying gene expressions of the oral squamous cell carcinoma your normal samples should not come from skin, etc.
Now, some of you may wonder why study gene expressions in cancer? A simple answer to this questions is that, it is the easiest and the least expensive method to start with given our current knowledge and available technology. It is true that cancers stem from mutation(s) in a gene or a set of genes and that the proteins are what controlling the cancerous activities, but in terms of biomarkers for diagnostic and treatment purposes it would be too expensive and impractical to perform genomic sequencing on individual patients to identify "potential" cancerous mutations. As far as proteomic profiling of individual patients, no known technology has yet to be developed, except for certain small sets of well characterized proteins. Furthermore, just because a carcinogenic (cancerous) mutation is found in a gene it doesn't mean that that mutation will result in cancer development. The mutant gene has to be expressed in order to create a dysfunctional protein that result in the development of cancer. In some cells or even some individuals, that gene may be completely silenced and the associated cancer may never occur.
What makes gene expression important in cancer study is that, although not all expressed genes will result in proteins development, most of the time changes in gene expression levels usually result in changes in the corresponding protein levels. This is when biomarker(s) discovery usually begins -- identification of certain genes expressed in unique pattern(s) corresponding to the disease will allow us to narrow down our search for the correct protein(s).
What are biomarkers, you may ask? Biomarkers are biological markers or proteins that can uniquely identify the state of the cell(s) or tissue sample. Cancer biomarkers are normally proteins that are expressed or present in only cancer samples and not in normal samples. Cancer markers may also be present in normal samples but absent in cancers, but this is more difficult to validate for the diagnosis of cancer because we don't know if the inability to detect the marker is due to the presence of the cancer or technical errors. Another method for defining a biomarker is changes in the protein levels that can uniquely identify cancer from normal, normally measured in terms of fold-change (whether increased or decreased level).
Since there are thousands of genes in our genome it is difficult to try to figure out which set of genes may be involved in cancer development (tumorigenesis), expression microarray comes in very handy. Once a set of genes that were expressed (increased or decreased, present or absent) in a unique manner to cancer compared to normal tissue, the next step normally involves validation by a reverse transcription polymerase chain reaction (RT-PCR), which then is further validated by protein assays.
The nice thing about expression array is that the result can lead to discovery at both expression (mRNA) and protein levels. Biomarkers may be important for cancer detections and treatments, but once a cancer-associated protein is detected it normally means that the cancer is already in an advanced stage. Because of the limitations in our current technology, cancer markers can only be reliably detected at higher levels that can be distinguishable from the background. At this level, the cancer cells are normally fully developed and the treatments are normally harder to implement.
The goal in cancer research is not just the treatment in itself, but more importantly the early detection of the cancer. The detection of pre-cancer cells or tissues will give the patients and physicians more options to consider. Furthermore, the treatment used to prevent the formation of cancer will most likely be less toxic and less risky than radiation, chemotherapy or surgery.
Using expression microarray we are able to identify unique gene or a set of genes that may act as a gene signature unique to cancer or even cancer type at the mRNA level. With the ability to detect aberrations in gene expression patterns unique to cancer we will in fact be able to reliably have a detection of the early onset of cancer development even in the absence of physical signs of the disease. This will be a significant breakthrough in cancer or disease management.
However, to be practical the cells with cancer gene signature must be detectable in the bloodstream of the patients since tissue biopsies will not be available without a physical evidence of the cancer itself. Since not all cancerous or pre-cancer cells will enter the bloodstream, detections will be limited to only certain types of cancer. Despite of this limitation, microarray technology still plays an essential role in the discovery of molecular markers of the disease at both protein and gene signature levels.
Thanks for reading. Any comments are appreciated.
Next time, I will talk about bioinformatics in cancer research and what are the implications of the molecular markers in terms of disease management.
In this post I'm going to talk a bit about microarray technology in cancer research. Please feel free to give any comments or suggestions.
It is clear that DNA microarray technology is an extremely powerful tool in studying gene expression profiles of a single cell or a set of cells. Unlike gene microarray, where you're studying the presence or absence of gene(s) or changes in the gene copy numbers, or exon microarray, where you're studying splice variances, expression microarray studies changes in the gene expressions at the mRNA level.
Gene expression microarray is a bit more complicated in terms of experimental design, which can become costly and even useless if not properly controlled. For those of you who are new with this technology, it is important to remember that expression microarray cannot answer all of your project's questions. Microarray is just an assay and not an experiment in itself. It is important that when designing an experiment for a microarray study you need to keep things simple. Complicated design will result in complicated data that can make your data analysis a nightmare. The most important step in a microarray study is the result, which depends entirely on the data mining and data analysis steps, which further depends on the accuracy of the experimental design.
When designing an expression microarray study you need two comparative conditions: control condition and the experimental condition. Otherwise, your expression data is meaningless. Next, you have to make sure that there are exactly two variables in your comparison, such as normal and cancer samples. That means, in the case of comparing normal and cancer samples, you need to make sure that both the normal and cancer samples come from the sample tissue type and the same anatomical origin. This is highly critical because gene expressions are highly regulated in a tissue-specific manner. That is, the same gene can express differently in different cell or tissue types. So, when studying gene expressions of the oral squamous cell carcinoma your normal samples should not come from skin, etc.
Now, some of you may wonder why study gene expressions in cancer? A simple answer to this questions is that, it is the easiest and the least expensive method to start with given our current knowledge and available technology. It is true that cancers stem from mutation(s) in a gene or a set of genes and that the proteins are what controlling the cancerous activities, but in terms of biomarkers for diagnostic and treatment purposes it would be too expensive and impractical to perform genomic sequencing on individual patients to identify "potential" cancerous mutations. As far as proteomic profiling of individual patients, no known technology has yet to be developed, except for certain small sets of well characterized proteins. Furthermore, just because a carcinogenic (cancerous) mutation is found in a gene it doesn't mean that that mutation will result in cancer development. The mutant gene has to be expressed in order to create a dysfunctional protein that result in the development of cancer. In some cells or even some individuals, that gene may be completely silenced and the associated cancer may never occur.
What makes gene expression important in cancer study is that, although not all expressed genes will result in proteins development, most of the time changes in gene expression levels usually result in changes in the corresponding protein levels. This is when biomarker(s) discovery usually begins -- identification of certain genes expressed in unique pattern(s) corresponding to the disease will allow us to narrow down our search for the correct protein(s).
What are biomarkers, you may ask? Biomarkers are biological markers or proteins that can uniquely identify the state of the cell(s) or tissue sample. Cancer biomarkers are normally proteins that are expressed or present in only cancer samples and not in normal samples. Cancer markers may also be present in normal samples but absent in cancers, but this is more difficult to validate for the diagnosis of cancer because we don't know if the inability to detect the marker is due to the presence of the cancer or technical errors. Another method for defining a biomarker is changes in the protein levels that can uniquely identify cancer from normal, normally measured in terms of fold-change (whether increased or decreased level).
Since there are thousands of genes in our genome it is difficult to try to figure out which set of genes may be involved in cancer development (tumorigenesis), expression microarray comes in very handy. Once a set of genes that were expressed (increased or decreased, present or absent) in a unique manner to cancer compared to normal tissue, the next step normally involves validation by a reverse transcription polymerase chain reaction (RT-PCR), which then is further validated by protein assays.
The nice thing about expression array is that the result can lead to discovery at both expression (mRNA) and protein levels. Biomarkers may be important for cancer detections and treatments, but once a cancer-associated protein is detected it normally means that the cancer is already in an advanced stage. Because of the limitations in our current technology, cancer markers can only be reliably detected at higher levels that can be distinguishable from the background. At this level, the cancer cells are normally fully developed and the treatments are normally harder to implement.
The goal in cancer research is not just the treatment in itself, but more importantly the early detection of the cancer. The detection of pre-cancer cells or tissues will give the patients and physicians more options to consider. Furthermore, the treatment used to prevent the formation of cancer will most likely be less toxic and less risky than radiation, chemotherapy or surgery.
Using expression microarray we are able to identify unique gene or a set of genes that may act as a gene signature unique to cancer or even cancer type at the mRNA level. With the ability to detect aberrations in gene expression patterns unique to cancer we will in fact be able to reliably have a detection of the early onset of cancer development even in the absence of physical signs of the disease. This will be a significant breakthrough in cancer or disease management.
However, to be practical the cells with cancer gene signature must be detectable in the bloodstream of the patients since tissue biopsies will not be available without a physical evidence of the cancer itself. Since not all cancerous or pre-cancer cells will enter the bloodstream, detections will be limited to only certain types of cancer. Despite of this limitation, microarray technology still plays an essential role in the discovery of molecular markers of the disease at both protein and gene signature levels.
Thanks for reading. Any comments are appreciated.
Next time, I will talk about bioinformatics in cancer research and what are the implications of the molecular markers in terms of disease management.
Labels: Fun Technology
