Outline

General definition of gene expression microarray

A technology using a high-density array of nucleic acids, protein, or tissue for examining complex biological interactions simultaneously which are identified by specific location on a slide array. A scanning microscope detects the bound, labeled sample and measures the visualized probe to ascertain the activity of the genes of interest in genotyping, cellular studies, and expression analysis.

Principles of microarray technology

Central dogma

Discovered by Francis Crick and published in a Nature paper in 1970, the central dogma gives the flow of information as first inscribed by DNA, then to RNA, and finally to protein with the impossibility of reverse flow. For more on the Central Dogma, (or a general overview of molecular biology) see the previous entry on central dogma. Gene expression microarray fundamentally depends on detecting amounts of either RNA or DNA present. However, we always keep in mind that the amount of mRNA is not necessarily one-to-one with the protein levels.

Measuring gene expression through protein and mRNA expression

The neat part about microarray is its high output - previously, methods to measure gene expression included N Northern blots, S1 nuclease protection, differential display, etc. , each of which must be done gene by gene and was very time consuming. Nowadays, we can use expression profiling using microararay data, which although noisy, can provide expression information on many thousands of genes in parallel. How? The different formats are discussed below.

Formats

Serial Analysis of Gene Expression (SAGE)

SAGE is the earliest example of a high output gene expression experiment. "Serial analysis of gene expression (SAGE) is an open platform for monitoring the expression patterns of thousands of transcripts in one sample and can lead to the discovery of novel genes. The technique relies on the generation of a library of short cDNA 'tags' each corresponding to a sequence near the 3' end of every transcript in a cell or tissue sample. The tags are sequenced to reveal the identity and quantity of the corresponding transcripts." (Nature 440, 1233-1237 (27 April 2006)

It is based on two principles

(1) A short nucleotide sequence of 9 - 10 bp contains enough information to uniquely identify a transcript

(2) The stringing together of short sequences tags allows efficient analysis of transcripts via the sequencing of the multiple tags within a single clone (Velculescu, Science, Vol 270, 1995).

See below for diagram of this process.

Synthesized double-stranded cDNA is cleaved with a restriction endonuclease that is expected to cleave most transcripts at least once. The most 3' portion of the cleaved cDNA was then isolated using streptavidin beads. cDNA is then divided into half, and ligated via anchoring restriction site to one of the two linkers containing a type IIIS restriction site.

Microarrays

General Procedure

• Grow cells at certain condition, collect mRNA
• Microarray has high density sequence specific probes with known location for each gene or RNA
• Sample will be hybridized to microarray probes by DNA complementarity
• Wash away non-specific binding
• Check sample mRNA value by labeled signals at each probe location

Currently, there are two main types of microarray with similar but significantly different features

cDNA microarray

This was first developed by Patrick Brown and his colleagues at Stanford University and published in a Science paper in 1995. cDNA microarray is essentially a set of cloned cDNAS spotted on a membrane or a glass slide, and hybridized to labeled RNA or DNA. There are several steps to gene expression using cDNA microarrays:

Production of arrays:

1) Select "probes" to be printed on array. Generally, probes are available from databases such as GenBank, dbEST and UniGene. Also, cDNAs of any type from the library of interest can be used. As the array production process is expensive, it is a good idea to select distinct arrays to be printed, as to survey the broadest set of genes.

2) The DNA fragments to be arrayed (spotted) are amplified by PCR to gain sufficient quantity: the reactions should largely yield single bands in the 400—2000 bp size range.

3) Printing is carried by a high precision and high speed robot that spots a sample of each gene product onto a number of matrices in a serial operation. The plain glass microscope slides are either coated with silane-derivative or with poly-L-Lysine to give the surface DNA binding capacity

4) The “print tip” is a needle-like steel pin with a split tip that takes up several hundred nanoliters of fluid by capillary action

5) The tip dips into a well, takes up DNA solution, and then releases a nanoliter sized drop each time it is touched down on a slide. The droplet dries in a few seconds, leaving the DNA spot behind as a “stain”. The spots are laid on the slides in a rectangular grid pattern.

6) The slides are post processed. In most cases, UV light is used to crosslink DNA to glass which is then chemically treated to “finish” the surface i.e. bind the DNA and make the surface non-sticky so that it does not non-specifically bind labeled DNA during the later hybridization.

7) Finally, return plates to –20C for long term storage. Plates can typically be arrayed from 2-4 times.

Probe labeling and hybridization

1) Total RNA and mRNA from tissues or cell lines are extracted and purified.

2) Labeling is carried out by performing reverse transcription of the RNA in the presence of dye-labeled nucleotides.

3) The resulting cDNA copy has incorporated fluorescently labeled bases the red (Cye5 - dUTP) and green (Cye3 - dUTP) labeled probes must be produced in separate labeling reactions so that each incorporates one color of dye.

4) Red and green labeled probe reactions are combined.

5) “Blocking DNA” added and then applied to the surface of the DNA slides (repetitive and non-specific DNA fragments, to block such hybridization sites on the array before they can attach to labeled probe fragments, Poly-A fragments.

6) Microscope cover slip is applied to spread the solution uniformly over the array and prevent rapid evaporation.

7) Hybridization: about 8—16 hours suffice to produce a reasonable amount of appropriate hybridization

8) The slide is washed to remove buffer and unhybridized probe DNA.

9) The finished slide is scanned by a fluorescent scanner.

Scanning

There are commonly two types of scanners.

1. Confocal Microscopy Scanners: The most common approach is to scan across the slide with a laser that is tuned/filtered to predominantly excite a single dye
2. CCD Camera “Scanners”: An alternative approach is to simultaneously excite both dyes over the entire slide surface with bright white tungsten filament light (no lasers)

Image Processing and Normalization

Images must be quantified numerical measurements of the brightness of the spots (normalized -- more on this in next wiki entry)

Normalization includes

• subtract background
• adjust for variable DNA
• adjust for variable amounts of green and red dyes

Summary

Preparation of microarray
Viewing of Image

Here's a simple cDNA array animation

Oligonucleotide microarray

An oligonucleotide is a short sequence of nucleotides. Hence, an oligonucleotide microarray is a microarray whose probes consist of synthetically created DNA oligonucleotides. One of the most famous examples of oligonucleotide microarrays is the GeneChip by Affymetrix:

Here, we detail in brief the steps for oligonucleotide microarray according to Affymetrix.

Genechip probe construction

1) Light-directed synthesis is combined with photolightography and solid-phase DNA synthesis

2) Synthetic linkers modified with photochemically removable protecting groups are attached to a glass substrate

3) Light is directed through a photolightographic mask to produce localized photodeprotection

4) The first of a series of chemical building blocks, hydroxyl-protected deoxynucleosides is incubated with the surface

5) Chemical coupling occurs at those sites that have been illuminated in the next step

6) Light is directed to different regions of substrate by a new mask, and the chemical cycle is repeated.

cRNA sample preparation

cRNA sample preparation follows these steps:

(1) primer hybridization

(2) reverse transcription

(3) second strand cDNA synthesis

(4) cleanup of double-stranded DNA

(5) application and biotin labeling of antisense cRNA

(6) cleanup of biotinylated DNA

(7) fragmentation

(8) washing/staining

(9) scanning. The labeled cRNA samples hybridize to DNA probes on GeneChip and shining laser lights caused tagged fragments to glow.

Expression probe and array design

As a basis, we need to have sequence information of genome, whether partial or complete.

• Key to design is choosing a probe that will target the desired gene expression.

25-mer oligonucleotides are designed using the photolithographic method described above. The feature size is usually about 5 microns.

• These probes need to be sensitive, unique, and non-overlapping, if possible.
• Specifically, probes will be complimentary to target gene, and not bind to other RNAs.
• A probe set is used to measure mRNA levels of a single gene.
• When creating the GeneChip, the base has approximately 100,000 cells on each GeneChip array, and then millions of DNA strands build up in each cell to yield one individual strand totaling about 25 base pairs.
• Probe cells in a probe set are arranged in probe pairs.
• Each probe pair contains a perfect match (PM) probe cell and a mismatch (MM) probe cell.
• A PM oligo perfectly matches part of a gene sequence.
• A MM oligo is identical to a PM oligo except that the middle nucleotide (13th of 25) is intentionally replaced by its complementary nucleotide.

Key Features

• Probe redundancy: that is, the use of multiple oligonucleotides of different sequences designed to hybridize to different regions of the same RNA. The use of multiple independent detectors for the same molecule greatly improves signal-to-noise ratios.
• Use of MM and PM probes, which allows one to determine with good confidence whether or not a signal generated by hybridization of the intended RNA molecule
• Oligonucleotide probes are designed using the same set of composition rules, regardless of organism or set of RNAs.

Summary

Synthesis of array
Hybridization
Excitation
Final image

See this Oligonucleotide array animation for a visualization of stated information.

Comparisons

Other commercial array platforms include

• Agilent
• ABI
• Amgen
• NimbleGen: cool mirror-tilting array, can go up to 70 bp...
• Many use long oligo probes of 30 - 70 nt

Applications of Microarray in Disease

In the fight against cancer, microarray technology has been highly influential. In the case of cancer, it is a bit of a misnomer to name cancer by the location (such as lung cancer, kidney cancer, etc.) as the cancers with the same biological placement in the body may be different in appearance, behavior and most importantly, source.

Breast Cancer

Breast cancer has a long history, dating back to Egypt in the 1600 B.C., as described in a medical statement called the Edwin Smith Papyrus (www.cancer.org). Since then, treatment has come a long way from cauterization. We know now that cancer results from DNA abnormalities; in the 1970s, it was discovered that 70% of breast cancers were ER (estrogen-receptor) positive. Then, in 1984, it was discovered that 20% of breast cancers have abnormal HER-2 gene. In the mid-1990s, it was discovered that 5% of breast cancers were an inherited defect in gene BRCA1 or BRCA2. Hence, 85% of breast cancer can be categorized and treated accordingly. However, the serious issue of diagnosing breast cancer type still remained. Current research allows microarrays to capture different patterns of genes that are switched on and off. Different patterns themselves may constitute distinct form of the disease, in terms of rate of growth, potential to spread and responsiveness to treatment. Understanding diversity can better match therapies with patients. Hence,

Leukemia

Acute leukemia can be classified into those arising from lymphoid precursors (acute lymphoblastic leukemia, ALL) or those from myeloid precursors (acute myeloid leukemia, AML). Appearance-wise, AML and ALL look very similar. Furthermore, in 1999, it took many types of tests and confirmations in order to classify acute leukemia. Hence, DNA microarrays provide a potential way to predict the class of leukemia. In the test-run using between 10 to 200 genes, predictors were found to be 100% accurate, reflecting a strong correlation between genes and type of class. Hence, microarrays prove a potentially facile way for potential gene classification and prediction.

References:

• Duggan et al, "Expression profiling using cDNA microarrays", Nature Genetics 1999, Vol 21: 10-14.
• Golub, Slonim, Lander et al. (1999) Molecular Classification of Cancer: Class Discovery and Class Prediction by Gene Expression Monitoring. Science, Vol 286, 531-537.
• Lipshutz et al, "High density synthetic oligonucleotide arrays", Nature Genets 1999, Vol 21: 20 - 24.
• Schena et al, "Quantitative Monitoring of Gene Expression Patterns with a Complementary DNA Microarray", Science 1995, Vol 270: 467-470.