Affymetrix microarray Probe


Studies have demonstrated that the brightness level is correlated with the absolute amount of RNA in the original sample, and by extension, the expression level of the gene associated with this RNA [157]. At a coarse level, one microarray experiment may be thought of as N many Northern blots that simultaneously assay a total RNA sample on a small common medium or substrate for as many different mRNA species—N being the total number of unique RNA species probes located on the physical chip. Note that the amount of total RNA required for a typical Northern blot is more than sufficient for one microarray experiment in the current technologies. However, this analogy breaks down in that only a single hybridization condition (e.g., temperature, time) is used in hybridizing all N assays, and unless the probes are carefully chosen, this may not be the optimal condition for the assay of all RNA species.

A prominent characteristic of microarray technologies is that they enable the comprehensive measurement of the expression level of many genes simultaneously on a common substrate. In this regard, typical applications of microarrays include the comprehensive quantification of RNA expression profiles of a system under different experimental conditions, or expression profile comparisons of two systems under one or several conditions. The former embraces the comparison of expression profiles of a system under a control and a test condition, whereas the latter includes contrasts between different strains of organisms, as, for instance, between a normal (e.g., wild-type), and a constructed (e.g., knockout) organism. Another intriguing use of microarrays is to compare expression levels between neighboring cells within the same microscopic field, as demonstrated in [123]. Aside from their widespread utility in functional genomics, oligonucleotide microarrays have also been used for single nucleotide polymorphism (SNP) analysis since many probe SNPs can be placed on the microarray for comprehensive parallel SNP detection, and in much the same way one can also perform DNA sequence analysis.

Regardless of the gene expression technology to be adopted, almost all of them have performance factors that depend critically on the general validity of certain fundamental biological assumptions outlined below:

1. There is a close correspondence between mRNA transcription and its associated protein translation. As noted by Brown and Botstein [34], one would ideally like to measure the final products of every gene, such as proteins, or even better, the biochemical activity of these products, which are more directly related to biological functionality. Such quantitation would provide a link between chemical DNA bases at microscopic levels with biological aspects that are manifest at macroscopic scales such as phenotype and physiology. However, there is no practical generic tool to do this yet. The assumption that there is a principle of parsimony—akin to Hamilton's principle of least action in the physical sciences—which drives the close relationship between gene expression and biological function was most clearly articulated by P. O. Brown and D. Botstein in [34]:

"The second reason [for using DNA microarrays to study gene expression on a genomic scale] is the tight connection between the function of a gene product and its expression pattern. As a rule, each gene is expressed in the specific cells and under the specific conditions in which its product makes a contribution to fitness. Just as natural selection has precisely tuned the biochemical properties of the gene product, so it has tuned the regulatory properties that govern when and where the product is made and in what quantity. The logic of natural selection, as well as experimental evidence, provides part of the basis for our belief that there is a sensible link between the expression pattern and the function of its gene product. Thirty years of molecular biology have provided numerous examples of genes that function under specific conditions and whose expression is tightly restricted to those conditions."

Biologists can quickly find exceptions to this assumption. For example, proteins that make up the cellular matrix can considerably outlast the lifetime of their associated mRNA, or conversely, they may be metabolized or degraded much more rapidly than the mRNA from which they were transcribed. Nevertheless, the initial successes in the applications of gene expression microarrays in investigations of expression and function suggest that this assumption holds true more frequently than not.

2. All mRNA transcripts have identical lifespans. Again, there are several well-known exceptions. For instance, we know that length of the 3' poly-A tail of an mRNA appears to be related to its stability. Furthermore, there are examples of mRNA that have longer- or shorter-term stability within specific cells or after its transcriptional event such as dystrophin mRNA from patients with certain types of muscular dystrophies. As a tangential comment on temporal effects: The probe-sample probe hybridization rate is known to be a function of the guanine-cytosine (GC) content of a transcript. In general, this rate is proportional to GC richness.

3. All cellular activities and responses are entirely programmed by transcriptional events. At a meta-systems level, this assumption may indeed be true, but in terms of direct mechanistic coupling, there exist many examples in which external stimuli cause changes in the biochemical program within the cell without engaging the transcriptional machinery. Figure 3.1 is an illustration of a response of free intracellular calcium to aldosterone exposure. Aldosterone is a steroid hormone that typically acts through binding with receptors that are translocated to the nucleus and which then initiate or modify a transcriptional program. Here, the time scale of the acute response suggests a nongenomic mechanism; in other words, the response does not require transcriptional activation. This example demonstrates how a molecule that usually works via modulation of transcription (i.e., steroid hormones) may also affect bioprocesses at the nongenomic level. There is also a much larger class of biological processes that do not primarily operate at the transcriptional level. These include muscular contraction, nerve excitation, and hormonal release. Eventually, all these events will cause some change in transcriptional activity, e.g., replenishing stores of neurotransmitter, but the patterns of gene expression would probably not reveal the control processes that govern them at the subgenomic time scale.

03 13


03 13


120 240 360 480 600

time (sec)

Figure 3.1: The nongenomic time scale response in aldosterone exposure. The response (in seconds) shown here is much faster than any known receptor-to-transcription response that steroid hormones are usually thought to act through. (Derived from Gamarra et al. [74].) There are presently two types of chip technologies in common usage: robotically spotted and oligonucleotide microarrays. It is important to note that these two microarray technologies are the very earliest and are currently the most widely used. Several competing technologies are emerging that may prove to be more cost-effective, reliable, and versatile (see chapter 7). Next, we briefly describe these two most popular microarray technologies.

Was this article helpful?

0 0

Post a comment