Methods (molecular, IHC, frozen)
Microarray-basic procedures

Topic Completed: 1 July 2010

Minor changes: 8 July 2020

Copyright: 2008-2021,, Inc.

PubMed Search: Microarray[TI] basic procedures free full text[sb]

Rodney E. Shackelford, D.O., Ph.D.
Page views in 2020: 197
Page views in 2021 to date: 14
Cite this page: Shackelford RE. Microarray-basic procedures. website. Accessed January 18th, 2021.
Definition / general
  • Although there are many different materials and variations commonly used in microarray, the basis of the technique involves the binding of nucleic acid "probes" or "capture molecules" to a solid support
  • These nucleic acids are often chemically synthesized oligonucleotides or cDNAs derived from mRNA, which are labeled with an appropriate fluorescent or luminescent labeled probe, such as Cy3, Cy5 or dye doped silica nanoparticles
  • cDNA probes are usually 25 - 65 base pairs and often derived from the 3' end of mRNAs, as many of these sequences were captured from transcripts using polyT primers
  • Oligonucleotide probes are usually larger, often 1,000 nucleotides
Measuring gene expression
  • Measuring gene expression via cDNA microarray is referred to as "expression analysis" or "expression profiling"
  • Labeled nucleic acid is bound via nucleic acid derivatives attached to the solid support by S-carboxymethyl-L-cystein, polyacrylamide, 3-aminopropyltrimethoxysilane, 3' glycidoxy propyltrimethoxysilane, phenylisothiocyanate, 1-ethyl=3-(3-dimethylaminopropyl)-carbodiimide hydrochloride or other compounds
  • Linkages themselves must be freely soluble in the hybridization and washing buffers, chemically stable and long enough to prevent inhibition of probe target nucleic acid hybridization by steric hindrance
  • Microarray can use up to four different labels; gene expression profiles use one or two probes
  • When one probe in employed, the sample (treated) and control (untreated) are hybridized to separate probe sets and compared; when the sample and control are differently labeled, a single slide is used
  • Solid support is commonly a glass slide, silicon biochip or nylon membrane, often called a gene chip
  • Bound nucleic acid probes may be synthesized oliogonucleotides, cDNAs or more rarely RNA
  • Typically the amount of each bound probe is a few picomoles
  • Each probe hybridizes with a different fluorescently labeled target nucleic acid species, such as an mRNA derived cDNA
  • Hybridization occurs via complementary Watson-Crick hydrogen bond formation between the A:T and G:C pairs
  • Binding specificity is increased with increased bp length of the probe target hybridization sequence, a high probe target G:C content and bp complementarity and if possible, the same / similar probe target Tms (sometimes be difficult with shorter cDNA derived probes, usually not a problem with longer oligonucleotide probes which usually have similar Tms)
  • In some cases mixing is employed
  • Mixing can be done by magnetic bar stirring, air driven bladders or centrifugal / shear mixing
  • Mixing often requires some sample dilution, however it reduces hybridization time, increases signal, lowers background and provides homogeneous hybridization conditions
  • Ratio of labeled target to probe is very high in microarray, so high that the amount of target that actually hybridizes with the probe is less than 1% of the total target molecules
  • Thus probe target hybridization does not result in significant target dilution which could alter assay results
Maintaining high sensitivity and specificity
  • Once binding is completed, the chips are extensively washed under stringent conditions, resulting in only exact or near exact Watson-Crick base pairing
  • High specificity is further achieved by having the hybridization step occur at a relatively high temperature, with a low buffer salt concentration
  • High stringency prevents the binding of noncomplementary strands, hairpin formation (probe or target self hybridization) and the disassociation of strands with very high complementary
  • High sensitivity is achieved by the target nucleic acid sequences being highly concentrated
  • Hybridization of a labeled target sequence reveals target binding only
  • Size, sequence and composition of the target are unknown
  • Labeled target sequences are measured via scanning, with images analysis performed by specialized software
  • Fluorescent intensity is measured and quantified, with the darkest areas (pixels) equal to no signal, the brightest or "whitest" areas recorded as maximum intensity and intermediate "gray" areas given corresponding signal intensity
  • Defining the areas to be quantified is difficult and can be done by the user circling the spot or by the software program
  • Once a spot is defined, the total signal intensity is summed and divided by the number of pixels within the spot to give the total signal intensity
  • Average background of the slide is often subtracted from the value of each spot to give initial data
  • Final data is derived from various statistical analytic methods
  • Microarray results show only the relative gene expression levels, not the absolute amount of a gene expression
  • It is also a limited "snapshot" of gene expression patterns and multiple microarray studies or other molecular methods are required to examine changes in gene expression patterns over time
  • Additionally, it is often prudent to verify microarray results with other techniques (PCR, Northern blotting or RNase protection assays are often used)
Common errors
  • Like any other technique, microarray is subject to several common errors

Assay complexity
  • Cloning and PCR steps required to create and process up to one million different sequences, combined with printing these sequences on the microarray chip, is extremely complex
  • Any error in this process will result in the misidentification of an expressed sequence, giving false data

Signal variation and analysis
  • Hybridization step, washing and pixel quantification steps are complicated by many factors, including background fluorescence, uneven hybridization, fluorophore inactivation by ozone and light exposure, temperature variation, cover slip positioning, hybridization time, uneven hybridization and dye leaking giving a false signal
  • Many other small but important details can significantly alter microarray results
  • Locating and troubleshooting these problems can be complex, expensive and time consuming

Incomplete oligonucleotide and cDNA synthesis
  • Extreme care must be taken to insure the quality and sequence integrity of the microarray probes because unrecognized incomplete or altered probes will drastically alter the hybridization step, invalidating assay results
  • Sheer number of probes required in many microarray assays can make this insuring the integrity of the probes very difficult

Data analysis and evaluation
  • Each microarray data set can consist of several million data points
  • Additionally, each microarray result is typically repeated multiple times, giving an enormous amount of raw data to be analyzed
  • Microarray experimental validity rests on multiple complex steps, such as defining the areas to be pixilated and counted, setting and subtracting out the appropriate background level, setting the boundaries of expression significance, compiling the results of multiple experiments and choosing the most appropriate method of statistical analysis
  • Any error at one of these steps will compromise the final data
  • Background intensity analysis can be especially complex in high density microarrays, compared to spotted microarrays, as the former lacks spaces between the probes
Back to top
Image 01 Image 02