In order to convert the information from a transcript population into a form that can be detected, the transcript population needs to be labeled. Several factors influence labeling efficiency. One factor is bias towards nucleotide sequences. Choosing the appropriate labeling strategy will eliminate bias. Another factor is that the length of labeled fragments influences hybridization kinetics. The optimal length is between 200-500bp. Additionally, yields of the labeled probe influence sensitivity -- quenching occurs as a result of high labeling density. If a high density of fluorescent molecules are in close proximity, a significant proportion of the light energy produced is spent on interactions between different molecules and dissipated as heat.

Most molecular biology strategies rely on the use of enzymes to convert mRNA into new populations of DNA or RNA. Resulting DNA or RNA is then labeled with fluorescent dyes, such as cyanine dyes Cy3 and Cy5. The main benefit of using these dyes is that both dyes can be excited independently from the same slide (Cy3 excites at 532nm and Cy5 at 633nm). They also have the advantage of producing bright signals, characterized by a wide dynamic range of detection. This allows for the detection of both weak and strong signals within the same experiment.

Labeling Principles

One of the first steps in labeling is to convert the mRNA population into a first-strand cDNA population using a reverse transcriptase enzyme. cDNA synthesis can be primed with either random primers or anchored oligo(dT) primers. Most eukaryotic transcripts contain poly-A tails, and this property can be exploited to separate transcripts from other RNA molecules using oligo(dT) primers. Incubation of total RNA with oligo(dT) primers will result in the hybridization between poly-A tails of the transcripts and the oligonucleotides. Priming will result in only one copy of cDNA that contains primarily 3’ sequences synthesized from each transcript. Random priming contains a mixture of oligonucleotides comprised of all sequence variants of a short sequence of defined length. Random priming should only be carried out when using mRNA, since it anneals to all RNA molecules. cDNA synthesis from total RNA with random priming will produce a large quantity of short fragments that lack specificity in hybridization and increase background problems. In addition, the proportion of labeled nucleotides incorporated into cDNA derived from mRNA will be small, and the specific signals from microarrays will be low.

Labeling Strategies

Labeling can be done either during first-strand synthesis or with cDNA. Incorporation of labeled nucleotides is the rate-limiting step in the first-strand labeling method as the polymerases incorporate unlabeled nucleotides more efficiently than the bulkier labeled nucleotides. The ratio between the labeled nucleotides and the corresponding unlabeled nucleotide determines the incorporation efficiency.

When labeling is done with cDNA, the incorporation of the labeled nucleotides is more efficient, and the yields are higher.

Microarray Hybridization

The process of hybridization is typically performed in order to identify and quantitate nucleic acids within larger samples. Hybridization involves annealing of a single-stranded nucleic acid (target) to a complementary strand (probe). In the case of microarrays, the probe is attached to a glass slide.

Hybridization Steps

Pre-hybridization of the slide is necessary to remove loosely bound probes and assures that the probes on the glass slide are ready for the target. Pre-hybridization steps also block any sites on the slide surface that are capable of binding targets nonspecifically.

Hybridization

Hybridization buffer containing a known amount of the labeled target cDNA sample is placed on the spotted slide surface and covered with a cover slip. The slide is then incubated in a humid environment for 16 – 20 hrs. If sequence homology is good, the target will hybridize to the probe.

Stringency Washes

The purpose of stringency washes is to remove all unattached and loosely bound target molecules, helping to prevent background noise. Stringency washes affect the amount of labeled targets retained on the slide for the final analysis. Primary washes have high salt concentrations to remove most of the hybridization buffer components. The secondary washes are performed with low salt buffer, removing loosely bound targets from the array and also removing any leftover salts.