Department of Pharmacology, University of California, Davis
Diaz, E., Barisone, G. A. DNA Microarrays: Sample Quality Control, Array Hybridization and Scanning. J. Vis. Exp. (49), e2546, doi:10.3791/2546 (2011).
Microarray expression profiling of the nervous system provides a powerful approach to identifying gene activities in different stages of development, different physiological or pathological states, response to therapy, and, in general, any condition that is being experimentally tested1. Expression profiling of neural tissues requires isolation of high quality RNA, amplification of the isolated RNA and hybridization to DNA microarrays. In this article we describe protocols for reproducible microarray experiments from brain tumor tissue2. We will start by performing a quality control analysis of isolated RNA samples with Agilent's 2100 Bioanalyzer "lab-on-a-chip" technology. High quality RNA samples are critical for the success of any microarray experiment, and the 2100 Bioanalyzer provides a quick, quantitative measurement of the sample quality. RNA samples are then amplified and labeled by performing reverse transcription to obtain cDNA, followed by in vitro transcription in the presence of labeled nucleotides to produce labeled cRNA. By using a dual-color labeling kit, we will label our experimental sample with Cy3 and a reference sample with Cy5. Both samples will then be combined and hybridized to Agilent's 4x44 K arrays. Dual-color arrays offer the advantage of a direct comparison between two RNA samples, thereby increasing the accuracy of the measurements, in particular for small changes in expression levels, because the two RNA samples are hybridized competitively to a single microarray. The arrays will be scanned at the two corresponding wavelengths, and the ratio of Cy3 to Cy5 signal for each feature will be used as a direct measurement of the relative abundance of the corresponding mRNA. This analysis identifies genes that are differentially expressed in response to the experimental conditions being tested.
1. Quality control analysis of RNA sample with 2100 Bioanalyzer
Before you start,
2. Amplification and labeling
To prepare for microarray analysis, RNA samples are amplified and labeled, usually in a reaction based on T7 RNA polymerase3. Double stranded cDNA is generated by reverse transcription. cDNA is used in an in vitro transcription reaction to generate what is known as cRNA. This reaction is performed in the presence of labeled ribonucleotides, producing microgram quantities of labeled RNA for array hybridization. The choice of amplification/labeling methods depends on the subsequent microarray platform to be used. In this section, we describe the generation of fluorescently labeled RNA using Agilent's Quick Amp labeling kit.
3. Microarray hybridization
4. Microarray washing
5. Representative Results:
Good quality total RNA samples should produce only two major peaks when run in a Bioanalyzer for quality control, corresponding to the 2 major ribosomal RNA species. Some RNA degradation will show as a smear before the first ribosomal RNA peak. Severely degraded, low quality RNA will show a broad peak or a series of peaks at low retention times, while the 2 ribosomal RNA peaks will be of very low intensity or not identifiable at all. For examples of the 2100 Bioanalyzer output, see figure 1.
The Expert 2100 software will calculate an RNA Integrity Number, or RIN, as a quantitative measurement of RNA quality. High quality samples, with RIN values higher than 9, are obviously best for microarray applications. However, we have used samples with RIN values as low as 5.2 in generating microarray data of reasonable quality.
Good quality arrays should produce high signal at relatively low PMT values. In our experiment, most of the transcripts are expected to be present at similar levels in the experimental and reference sample; massive, widespread changes in gene expression will probably lack any biological significance. Therefore, most of the array should look yellow rather than green or red. Good quality signals should also be in a dynamic range such that the signal histograms fully overlap. For examples, see figure 2.
Figure 1. Example of four RNA samples isolated from human brain tumor tissue. RNA was isolated using the protocol presented in 3.1.1. Sample quality was assessed using the 2100 Bioanalyzer on an RNA 6000 chip as described in 3.1.3. Samples are representative of 4 distinct RNA qualities: A, very good RNA quality (RIN>9); B, partially degraded RNA (RIN 5-6, note the significantly lower peak for 28S rRNA); C, highly degraded RNA (RIN<3); D, almost completely degraded RNA (RIN=2). We have successfully used samples with qualities similar to or better than B (RIN>5.2) for downstream microarray applications. Solid and open arrowheads indicate the position of the 18S and 28S rRNA species, respectively.
Figure 2. Examples of array images and scatter plots. A, whole slide preview, showing the four arrays in the 4X44K Agilent format; B, high resolution scan of one of the array areas, note that neither of the signals (red or green) is predominant; C, scatter plot of the image in B, note that the signals fully overlap and no more than 1x10-6 features are at saturating intensity.
Expression profiling with DNA microarrays provides a straightforward approach to identify differentially expressed genes between two biological samples. Successful expression profiling experiments require high quality RNA samples, robust labeling and hybridization. In our experience, the commercial arrays and labeling kits from Agilent provide high quality data at a reasonable cost. While Agilent also offers their own hybridization and scanning machinery, we favor the hybridization system from Roche-Nimblegen and the GenePix 4000B microarray scanner from Molecular Devices. Note that the Roche-Nimblegen hybridization unit was formerly known as the MAUI hybridization system. After the recent corporate acquisition by Roche, the A4 mixers have been discontinued. As of the moment of this writing, they can still be found through other sellers such as Kreatech Diagnostics (www.kreatech.com; this website also offers a convenient array compatibility tool), but long-term stock availability is uncertain. Other hybridization systems are available (for example, from Agilent); however, if compatible mixers can be found for the array being used, we recommend the Roche-Nimblegen's system for its quality and reproducibility.
No conflicts of interest declared.
This work was supported by grants to E. D. from the James S. McDonnell Foundation 21stCentury Award Program, and a NIH Director's New Innovator Award. G. A. B. was supported in part by a postdoctoral fellowship from the California Institute of Regenerative Medicine.
|2100 Bioanalyzer||Agilent Technologies||G2943CA||Includes chip priming station, vortexer, software and laptop computer|
|2100 Bioanalyzer electrophoresis set||Agilent Technologies||G2947CA|
|RNA 6000 Nano kit||Agilent Technologies||5067-1511||includes size ladder, marker, gel matrix, dye, electrode cleaners, spin filters and nanochips|
|RNase Zap||Applied Biosystems||AM9780M|
|Quick Amp labeling kit, two-color||Agilent Technologies||5190-0444|
|RNeasy mini kit||Qiagen||74104|
|Gene Expression Hybridization Kit||Agilent Technologies||5188-542||Includes Blocking agent, Fragmentation buffer and Hybridization buffer|
|Gene Expression Wash Buffer Kit||Agilent Technologies||5188-5327||Includes Wash Buffers 1 and 2 and Triton X-102|
|Hybridization Station||Roche Group||05223652001||4-slide model|
|Hybridization System Accesory Kit||Roche Group||05327695001||Contains verification assembly for testing mixing, replacement O-rings, disassembly tool, forceps and brayer|
|A4 mixer hybridization chambers|
|Whole Human Genome (4x44) Oligo Microarray||Agilent Technologies||G4112F|
|Positive displacement pipette||Gilson, Inc.||F148504|
|Capillary pistons for positive displacement pipette||Gilson, Inc.||F148415|
|Microarray fluorescent scanner, Axon GenePix 4000B||Molecular Devices||GENEPIX4000B|
|GenePix Pro 6.0 software||Molecular Devices|