Guidelines for genome-wide expression profiling after gene silencing with RNAi
Abstract
We have devised protocols for siRNA-mediated gene silencing followed by genome-wide expression profiling. The protocols described here were based on experimental setups developed for siRNA mediated knock-down of cdc-2 in 3 frequently used cell types: one established, adherent cell line (MCF-7), one suspension cell line (K562), and primary HUVEC cells. Following cellular lipofection of synthetic siRNAs in triplicates, testing procedures at the mRNA-, protein-, and phenotypic level are described to show successful knock-down of the target gene before proceeding with GeneChip Analysis. These protocols can be used as starting points for fast and easy optimization of experimental conditions for other cell types.
Introduction
RNA interference (RNAi)1 is the process of targeted, post-transcriptional gene silencing mediated by double-stranded RNA (dsRNA). It is thought that the role of RNAi in nature is for protection of organisms from viruses and to suppress the movement of mobile genetic elements, such as transposons. Numerous studies have shown that dsRNA-induced gene silencing occurs in a number of different mammalian species 2-5. The finding that the size of functional dsRNA fragments is conserved in plants and animals suggests a highly conserved mechanism in nature6.
Initial attempts to silence specific genes by RNAi in mammalian cells were unsuccessful, since the introduction of long dsRNA (>30 bp) leads to strong activation of the dsRNA-dependent protein kinase PKR and 2', 5'-oligoadenylate synthetase (2', 5'-AS). The activation of these two enzymes triggers a nonspecific shutdown of protein synthesis and nonspecific degradation of mRNA. It was not until 2001, as RNAi became better understood, that scientists discovered that double-stranded, short interfering RNA (siRNA) oligos of 23 nt could be used to mediate gene silencing in mammalian cells. Elbashir et al. demonstrated that chemically synthesized 21 nt siRNA duplexes specifically suppress the expression of endogenous and heterologous genes in different mammalian cell lines, including human 293 and HeLa cells 7. A key discovery from these studies was that no nonspecific gene silencing effects were seen in mammalian cells after transfection of short dsRNA sequences (<30 nt). These results showed that 21 nt siRNA duplexes can be used as a tool for the study of gene function in mammalian cells, and could eventually serve as gene-specific therapeutics in the near future11-13. As is the case for other gene knockdown techniques, reliable evaluation of phenotypes and gene regulation patterns is critical for a sound interpretation of the effect of the silenced gene. As gene knockdown experiments employing RNAi have developed into a widespread and efficient molecular tool in many labs 8-10, several reports have described various nonspecific side-effects of RNAi, which can and must be controlled in order to ensure reliable data interpretation. Many aspects of the RNAi mechanism are still not well understood and there are also many challenges to valid data interpretation, including “off-target” effects. For these reasons, it is advised to run state-of-the-art controls during an RNAi experiment and also to assess the amount of system-inherent ambiguity in order to confidently filter out the biologically relevant data.
The use of expression profiling as a phenotypic readout of RNAi-mediated gene silencing offers many advantages. Whole genome expression arrays provide an unbiased view of the transcriptional changes when used to profile the RNA extracted from siRNA-treated cells. Changing transcripts pinpoint affected pathways and can help decipher on-target versus off-target effects when appropriate controls are compared to specific silencing treatments. Standard expression analysis methodologies using the Affymetrix GeneChip® system were utilized in this protocol.
Proper interpretation of the complex biological results gained from an RNAi experiment requires a systematic, rigorous approach. The protocols presented in this paper are based on those developed for a series of experiments in which the effects of cdc2 silencing on multiple cell types was tested. The cdc2 gene was knocked down using 3 validated siRNAs independently. All the experiments were performed in triplicate to allow statistical interpretation of the data. To determine the outcomes of cdc2-dependent phenotypes for different cell types, 3 cell types were chosen: MCF-7, K562, and HUVEC cells. Taken together, thoroughly controlled RNAi experiments in combination with genome-wide microarrays allowed rigorous assessment of the specificity of an experiment and filtering out of irrelevant noise.
Materials
Reagents
• Cell culture reagents:
Suitable cells of mammalian origin (here MCF7, K562, and HUVEC)
Cell culture medium, such as DMEM, RPMI (Invitrogen)
Cell culture medium supplements, such as EGM2 (Cambrex)
Trypsin/EDTA (Sigma)
• siRNA and quantitative RT-PCR related reagents (QIAGEN):
Validated, gene-specific siRNAs (e.g., targeting cdc2)
Nonsilencing control siRNAs
HiPerFect Transfection Reagent
RNeasy® Kit
Suitable primer assays for quantitative RT-PCR (e.g., Hs_CDC2_1_SG)
QuantiTect® Primer Assay
QuantiTect Probe RT-PCR Kit
• Miscellaneous reagents and buffers
Propidium iodide solution [1 µg/µl] (Sigma)
Propidium Iodide staining buffer (50 mM Tris-HCl, pH 7.5; 10mM MgCl2)
RNase A [1 µg/µl] (QIAGEN)
ECL Detection Solution (Amersham Pharmacia Biotech).
BC Assay Protein Quantification Kit (Uptima),
Protein Lysis Buffer (50 mM Tris/HCl pH 7.4, 150 mM NaCl, 5 mM EDTA, 5 mM EGTA, 1 g/l SDS, 1 g/l Na-Desoxycholat, 1% NP40)
5x Sample Loading Buffer (0.225 M Tris/HCl pH6.8, 50% glycerol, 5% SDS, 0.05% bromophenol blue, 0.25 M DTT)
1x Electrophoresis Buffer (14.4 g/l Glycine, 1 g/l SDS, 3.04 g/l Tris)
Protein Blot Buffer (0.025 M Tris, 0.192 M Glycine, 20% (v/v) methanol).
cdc2-specific primary antibody (Becton-Dickinson)
Alpha-tubulin antibody (Sigma, T-5168)
Peroxidase-coupled secondary antibody (Sigma)
Cell Suspension Buffer (50 mM Tris-HCl, pH 7.5; 10 mM MgCl2)
• GeneChip® hybridization reagents (Affymetrix):
One Cycle cDNA Kit
GeneChip IVT Labeling Kit
Sample Cleanup Module
Eukaryotic Poly-A RNA Control Kit
Hybridization Controls
Equipment
• GeneChip Scanner 3000
• GeneChip Fluidic Station 450
• GeneChip Hybridization Oven 640
• Real-time PCR device
• Flow Cytometer
• BioRad Protean II device
• Nitrocellulose membrane (Schleicher & Schuell)
Procedure
h3. Transfection of cells with siRNA
Optimization experiments for determination of the best experimental conditions
Before starting with the actual experiment, titration of the transfection reagent and siRNA for the gene of interest (here: cdc2) should be optimized for each cell type under investigation. Use of 3 functional siRNA species against the gene of interest in individual transfections is highly recommended. The gene silencing effect should be tested at the mRNA and protein level and, if possible, phenotypic analysis should be performed. In the experiments described here, individual cell cycle states were estimated by FACS analysis, after cell fixation and Propidium Iodide staining, to confirm cdc2-siRNA–mediated cell cycle arrest in the G2 phase.
Cells transfected with a nonsilencing siRNA and an untreated cell sample should always be set up in parallel to assess the specificity of knockdown. For later statistical analysis, it is advised to perform all experiments in triplicate.
For every cell type, optimized experimental conditions should be selected which result in the most pronounced phenotype and strongest knockdown effect for all siRNAs used at their lowest possible concentration.
Since many functional RNAi studies focus on the detection of downstream effects caused by the disappearance of the target protein, GeneChip® array analysis is typically performed 3 days after transfection, when the silenced protein (such as the cdc2 protein) is almost completely absent for at least one day. However, although the 3-day time point is frequently used for phenotypic analysis, knockdown studies may display their phenotype at differing time points depending on the gene targeted. This should be taken into account when using these protocols for siRNAs targeting genes other than cdc2.
The following table shows the optimization strategy with the final experimental conditions printed in bold.
The standard experimental procedures described focus mainly on MCF-7 cells. Notable differences in the protocol for the other cell types used are indicated.
Transfection protocol for MCF-7 cells:
1 Carefully detach cells using trypsin and wash in PBS. Count the cells and resuspend in complete culture medium at a density of 1.2 × 105 cells/ml.
2 Transfer 500 µl cell suspension to each well of a 24-well plate. For the short time until transfection, incubate cells under normal growth conditions.
3 Dilute 37.5 ng siRNA per well in 100 µl culture medium without serum.
4 Add suitable transfection reagent, such as 3 µl HiPerFect Transfection Reagent, to the siRNA and mix by vortexing.
5 Incubate the samples for 5–10 min at room temperature (15–25°C) to allow the formation of transfection complexes.
6 Add the complexes drop-wise onto the cells. Gently swirl the plate to ensure uniform distribution of the transfection complexes.
7 Incubate the cells for the desired time (e.g., 72 h). Change the medium 2 days after transfection.
RNA preparation
8 Isolate total RNA from the cells by standard methods, for example using an RNeasy Kit (QIAGEN) according to the handbook. Lyse the cells from 6 wells of a 24-well plate directly in the well by addition of 350 µl buffer RLT.
PAUSE POINT
The plates containing cell lysates can be stored at -20°C until further processing.
Troubleshooting
The integrity of purified total RNA is crucial for further procedures. Therefore, careful quality analysis of the isolated RNA is advised. This can be performed using the Agilent Bioanalyzer 2100 which should provide an “RNA Integrity Number” (RIN) of between 8 and 10. We have used RNA with RIN numbers ranging from 9.4 to 10. This analysis should show a flat line between the 28S and 18S ribosomal RNA, indicating the absence of degradation products of the 28S ribosomal RNA. If analyzing the RNA on an agarose gel, the 28S:18S ratio should be 2:1.
For microarray analysis, it is essential to correlate the final genome-wide expression profiles with biological data in order to ensure that RNA from a successful gene knockdown experiment is used for GeneChip hybridization. To gain confidence that siRNA-treated cells are in fact reflecting the successful knockdown of the target gene and subsequently the protein (cdc2 in our case), and that phenotypic analysis is indicating the predicted phenotype (in this case cdc2-dependent arrest of cells in the G2 phase of the cell cycle), the following experimental quality controls on 3 levels should be performed whenever possible:
• Estimation of siRNA-targeted mRNA knockdown.
• Estimation of target-protein reduction.
• Phenotype analysis (e.g., determination of cell cycle profiles) of the treated cell populations.
Measurement of gene silencing efficiency by real-time qRT-PCR
Real-time, quantitative RT-PCRs are an ideal tool for knockdown analysis of siRNA-targeted mRNA. A PCR protocol using a MJ Research Opticon® 2 real-time PCR device is described in which 2 µl of total RNA from Step 8 is used in a one-step qRT-PCR with a PCR primer mix specific for the gene of interest.
9 Thaw 2x SYBR Green RT-PCR Master Mix, 10x Primer Assay, template RNA, and RNase-free water (all provided in the QuantiTect SYBR Green RT-PCR Kit). Mix the individual solutions and place them on ice. The RT Mix should be taken from –20ºC immediately before use, always kept on ice, and returned to storage at –20ºC immediately after use.
10 Prepare a reaction mix according to the following table.
Keep samples on ice while preparing the reaction mix.
Note: The final Mg2+ concentration of 2.5 mM provided by the 2x SYBR Green RT-PCR Master Mix gives optimal results.
11 Mix the reaction mix thoroughly and dispense appropriate volumes into PCR tubes or plates. Keep the tubes or plates on ice.
12 Add template RNA (≤10 ng/reaction) to the individual PCR tubes or wells containing the reaction mix.
13 Program the real-time cycler according to the following table.
Data acquisition should be performed during the extension step.
Troubleshooting: The PCR parameters indicated above worked very well for the MJ Research Opticon 2 real-time PCR device. Other PCR devices may require different parameters for optimal performance.
14 Keep the samples on ice until the real-time cycler is programmed.
Place the PCR tubes or plates in the real-time cycler and start the cycling program.
Note: The Tm of an RT-PCR product depends on buffer composition and salt concentration. Tm values obtained when using QuantiTect SYBR Green RT-PCR reagents may differ from those obtained using other reagents.
Detection of cdc2 knockdown at the protein level by Western blot
15 Three days after transfection, lyse cells from one 24-well using 100 µl of Protein Lysis Buffer.
16 Measure the protein concentration using the BC Assay Protein Quantification Kit.
17 For SDS-PAGE, mix 10 µg protein lysate with 5x Sample Loading Buffer, boil for 5 min at 95°C, and load on a 12.5% PAA gel.
18 Run the gel for 1 h at 100V (when using the BioRad Protean II device) in 1x Electrophoresis Buffer.
19 Transfer protein to a nitrocellulose membrane using protein blot buffer.
PAUSE POINT
20 Perform immunostaining with antibodies specific for cdc2, followed by incubation with a peroxidase-coupled secondary antibody. Carry out detection with ECL Detection Solution. Confirm equal loading of the blots by reprobing with an alpha-tubulin antibody.
Detection of cdc2-knockdown phenotype by cell-cycle analysis
21 Three days after transfection, treat cells thoroughly with trypsin to detach them from the culture plates and to disrupt all cellular aggregates.
Note: Cellular aggregates that are not disrupted in this step will stick together after fixation and throughout the whole procedure and will interfere with data analysis.
22 Resuspend cells in 10 ml of ice-cold PBS and wash for 10 minutes (1000 rpm, 4°C).
23 Discard supernatant, resuspend cells in 100 µl PBS and add the cells drop-wise into 10 ml of 70% ethanol on ice (prechilled at -20°C before use) for fixation. Incubate the cells for at least 2 h at -20°C.
PAUSE POINT
Fixed cells can be stored at -20 °C for up to 2 months after this step.
24 Centrifuge the fixed cells for 10 min at 1000 rpm and wash once in PBS (5 ml, 10 min, 1000 rpm).
25 Carefully discard as much of the supernatant as possible without aspirating the cell pellet. Resuspend the cells in 400 µl Cell Suspension Buffer.
26 Add RNase A (stock solution 1 µg/µl) to the cells at a final concentration of 10 µg/ml and incubate for 30 min at 37°C.
27 Add Propidium Iodide to the cells (final concentration of 70 µg/ml) and stain the cells overnight in the dark at 4°C.
28 Measure the cell-cycle distribution by FACS analysis. Count 40,000 cells per sample to obtain statistically valid results.
(Analysis of PI in channel 2, linear scale)
Note: When all the tests described above indicate successful knockdown of the cell samples which were transfected with the cdc2 siRNAs, but not with the controls, total RNA samples may be used for GeneChip array analysis.
Affymetrix Gene expression analysis
Note: Detailed protocols may also be found on the Affymetrix website at www.Affymetrix.com
29 Set up a 12 µl reaction for each sample to prime for first strand cDNA synthesis.
Mix the reaction by flicking the tube.
30 Incubate in a PCR thermocycler (e.g., MJ-PTC 200) with a heated lid for 10 min at 70°C. Afterwards, immediately place the reaction on ice for approx. 2 min.
31 Prepare the first-strand mastermix as detailed in the following table.
Make sure that the reactions are mixed properly. Incubate for 1 h at 42°C in the thermocycler with a heated lid. Afterwards, immediately place the reaction on ice.
32 Prepare sufficient second-strand cDNA synthesis master mix as detailed in the following table.
Mix the master mix by gently flicking the tube several times.
33 Add 130 µl of second-strand master mix to each first-strand synthesis sample (results in a total volume of 150 µl). Gently flick the tube to mix and centrifuge briefly to collect the solution at the bottom of the tube. Incubate for 2 h at 16°C without a heated lid.
34 Add 2 µl of T4 DNA Polymerase to each sample and incubate for 5 min at 16°C
35 After incubation with T4 DNA Polymerase, add 10 µl of 0.5 M EDTA and proceed to the cleanup of the double-stranded cDNA
36 Clean up the double-stranded cDNA using the Sample Cleanup Module according to the manufacturer’s recommendations. Binding of the cDNA is a critical step. Therefore, the cDNA Binding buffer must be mixed very well with the double-stranded cDNA synthesis preparation by vortexing for 3 s.
At the end of the cleanup procedure, the eluate volume should be 12 µl.
37 Proceed with IVT reaction setup as detailed in the following table.
Carefully mix the reagents and collect the mixture at the bottom of the tube with a brief centrifugation. Incubate for 16 h 37°C in the thermocycler with a heated lid.
PAUSE POINT
Store the labeled cRNA at -20°C or -70°C if not purifying immediately.
38 Proceed with cleanup of the biotin-labeled cRNA using the GeneChip Sample Cleanup according to the manufacturer’s recommendations.
39 Quantify the cRNA by using spectrophotometric analysis to determine yield and quality. Apply the convention that 1 absorbance unit at 260 nm equals 40 µg/ml RNA. Check the absorbance at 260 nm and 280 nm to determine sample concentration and purity. Maintain the A260/A280 ratio close to 2.0 for pure RNA (ratios between 1.9 and 2.1 are acceptable).
Troubleshooting
For quantification of cRNA when using total RNA as starting material, an adjusted cRNA yield must be calculated to reflect carryover of unlabeled total RNA. Using an estimate of 100% carryover, use the formula below to determine adjusted cRNA yield:
Adjusted cRNA yield = RNAm - (total RNAi) (y)
RNAm = amount of cRNA measured after IVT (μg)
Total RNAi = Starting amount of total RNA (μg)
y = fraction of cDNA reaction used in IVT
Example: Starting with 10 μg total RNA, 50% of the cDNA reaction is added to the IVT, giving a yield of 50 μg cRNA. Therefore, adjusted cRNA yield = 50 μg cRNA - (10 μg total RNA) (0.5 cDNA reaction) = 45 μg.
If cRNA concentration is above 1 µg, dilute RNA to 1 µg prior to hybridization.
• Incubate at 94°C for 35 minutes. Place on ice following the incubation.
• Save an aliquot for analysis on the Bioanalyzer.
The standard fragmentation procedure should produce a distribution of RNA fragment sizes from approximately 35 to 200 bases.
PAUSE POINT
• Store undiluted, fragmented sample cRNA at –20°C (or –70°C for long-term storage) until ready to perform the hybridization.
40 Perform hybridization, probe array washing, staining, and scanning according to Affymetrix GeneChip Expression Analysis Technical Manual.
41 Analyze the array images first by visual inspection.Inspect the quality measurements including:
• the presence or absence of the B2 control oligo
• 3’/5’ ratio of control genes
• Background
• Noise
• Present call rates
Use this inspection to decide whether arrays should be used for further analyses.
42 Overview of Data Analysis:
Following basic array QC, raw array data may be analyzed using the PLIER algorithm14 which uses an affinity-model to summarize a signal value from PM and MM probe pair data. In these experiments, Multi-way ANOVA was used to look for significant changes due to various experimental factors such as time, transfection status, siRNA type, and the interaction of factors, such as Time + Treatment. Pair-wise tests were used to test specific conditions against each other at each time point, calculate a median Signal Log Ratio (change estimate), and compute a T-test p-value for every pair-wise comparison.NetAffx (annotations resource on Affymetrix.com) and Ingenuity Pathways Analysis may be used to look for significant ‘undesired’ changes, such as in immune networks that involve interferon, and to look for siRNA specific changes in target genes, such as for cdc2 knockdowns in cell-cycle networks.
To proof silencing specificity of the siRNA sequence, an in silico method may be applied. Thus, probeset signal data are mined for biological results first through statistical filtering, then by examining the significant changes between two conditions in the context of gene networks, using tools such as Ingenuity Pathway analysis ( www.ingenuity.com).
References
1. Fire, A. et al. Potent and specific genetic interference by double-stranded RNA in Caenorhabditis elegans. Nature 391, 806-11 (1998).
2. Ngo, H., Tschudi, C., Gull, K. & Ullu, E. Double-stranded RNA induces mRNA degradation in Trypanosoma brucei. Proc Natl Acad Sci U S A 95, 14687-92 (1998).
3. Li, Y. X., Farrell, M. J., Liu, R., Mohanty, N. & Kirby, M. L. Double-stranded RNA injection produces null phenotypes in zebrafish. Dev Biol 217, 394-405 (2000).
4. Wianny, F. & Zernicka-Goetz, M. Specific interference with gene function by double-stranded RNA in early mouse development. Nat Cell Biol 2, 70-5 (2000).
5. Caplen, N. J., Parrish, S., Imani, F., Fire, A. & Morgan, R. A. Specific inhibition of gene expression by small double-stranded RNAs in invertebrate and vertebrate systems. Proc Natl Acad Sci U S A 98, 9742-7 (2001).
6. Hammond, S. M., Bernstein, E., Beach, D. & Hannon, G. J. An RNA-directed nuclease mediates post-transcriptional gene silencing in Drosophila cells. Nature 404, 293-6 (2000).
7. Elbashir, S. M. et al. Duplexes of 21-nucleotide RNAs mediate RNA interference in cultured mammalian cells. Nature 411, 494-8 (2001).
8. Ovcharenko, D., Jarvis, R., Hunicke-Smith, S., Kelnar, K. & Brown, D. High-throughput RNAi screening in vitro: from cell lines to primary cells. RNA 11, 985-93 (2005).
9. Echeverri, C. J. & Perrimon, N. High-throughput RNAi screening in cultured cells: a user's guide. Nat Rev Genet 7, 373-84 (2006).
10. Pelkmans, L. et al. Genome-wide analysis of human kinases in clathrin- and caveolae/raft-mediated endocytosis. Nature 436, 78-86 (2005).
11. Nekhai, S. & Jerebtsova, M. Therapies for HIV with RNAi. Curr Opin Mol Ther 8, 52-61 (2006).
12. Rondinone, C. M. Therapeutic potential of RNAi in metabolic diseases. Biotechniques Suppl, 31-6 (2006).
13. Pai, S. I. et al. Prospects of RNA interference therapy for cancer. Gene Ther 13, 464-77 (2006).
14. http://www.affymetrix.com/support/technical/technotes/plier_technote.pdf
Acknowledgements
Keywords
RNAi, siRNA, gene silencing, transfection, gene expression, genome-wide expression profiling, arrays
Figure 1
Flowchart of experimental design
