Optimization of RNA amplification for array analysis

Introduction

Experiments involving glass microarrays typically require as much as 20 μg of total RNA or 2 μg of poly(A) RNA for cDNA labeling and subsequent hybridization. However, obtaining sufficient amounts of RNA for array analysis from small precious samples such as needle biopsies and laser capture microdissections is often impossible. Therefore, protocols have been developed to 'amplify' these rare RNAs for array analysis applications.

Linear amplification of RNA

A widely used protocol for linear RNA amplification is that developed by Van Gelder and Eberwine.⁽¹⁾ (Figure 1.) The amplified antisense RNA (aRNA) procedure begins with total or poly(A) RNA that is reverse transcribed using an oligo(dT) primer containing a T7 RNA polymerase promoter sequence. After first strand synthesis, the reaction is treated with RNase H to cleave the mRNA strand into small fragments. These fragments serve as primers during a second strand synthesis reaction that produces double-stranded cDNA, which will serve as a template for subsequent transcription. RRNA, mRNA fragments, and primers are removed, and the cDNA template is used in an in vitro transcription reaction to produce large amounts of linearly amplified aRNA (Figure 1). Using this method, the mRNA of a total RNA sample will be amplified 1000-4000×. It is possible to further increase yields by using the aRNA in a second round of amplification.

Getting more aRNA

To maximize amplification, this multi-step procedure requires paying meticulous attention to each step. Surprisingly, the transcription/amplification step is usually not the critical factor for good microarray data. MRNA quality, on the other hand, is a major limiting factor. Basic precautions when working with RNA are obviously important, and beginning the procedure with high quality RNA is the easiest way to ensure reproducible amplification. To make troubleshooting easier, reaction products can be checked between each enzymatic reaction on agarose gels to track the outcome of each step before moving ahead. Good first and second strand cDNA synthesis, accompanied by an efficient recovery method, are also critical. Just as important, is preventing sample loss during purification; this will also increase yield and decrease variance between samples. The actual amplification by in vitro transcription is typically very reproducible when high quality cDNA is used as a template. Below we provide tips for some of the steps of aRNA synthesis.

click the image to enlarge

Figure 1: Schematic of aRNA amplification as developed by Van Gelder and Eberwine. Ambion's MessageAmp aRNA Kit is an optimized version of this procedure.

click the image to enlarge

Figure 2: Agilent 2100 bioanalyzer electropherogram of aRNA after one or two rounds of amplification. The aRNA was produced from 2 mg of thymus total RNA by using Ambion's MessageAmp aRNA Kit. The aRNA was loaded onto a LabChip™ and run on an Agilent 2100 bioanalyzer. The size and quantity of the aRNA was determined by comparison to Ambion's RNA 6000 ladder, which produces distinct peaks of 0.2, 0.5, 1.0, 2.0, 4.0, and 6.0 kb and is supplied at precisely 150 ng/ml.

Optimization of aRNA synthesis

Effects of RNA quality on aRNA amplification

Sample preparation. The quality of the RNA starting material will have a significant impact on the yield and quality of the synthesized aRNA. The yield and size distribution of aRNAs generated from intact and degraded RNAs has been analyzed. In a typical experiment, 78 μg of aRNA was generated from 2 μg of intact RNA. However, only 19 μg of aRNA was generated from degraded RNA. Moreover, the average size of the aRNA products decreased from an average of 1.1 kb to 0.7 kb when using the intact and degraded RNA samples, respectively.

To avoid RNA degradation, fast processing of tissue and cell samples for RNA isolation is critical. This may involve disrupting tissues or cells immediately upon collection, or storage by immediate freezing. (Note that frozen samples should never be thawed prior to disruption in a cell lysis buffer, as this will usually lead to RNA degradation. Instead frozen tissues should be powdered and added to the lysis solution for disruption while still frozen.) Alternatively, a tissue-storage/RNA-stabilization reagent can be used to permeate the fresh tissue and inactivate RNases, thus making room temperature or 4 C storage possible. Ambion's RNAlater™ Solution can be used for this purpose.

RNA purification. Small RNA and DNA fragments that could prime cDNA synthesis should be removed from the RNA sample during purification. This can be accomplished with protocols that bind the RNA sample to glass fibers or beads. A recent review of the array analysis literature suggests that researchers are having increasing success with a combination of phenol extraction and glass binding protocols for RNA isolation.^(2-4) Recently Ambion has developed a new series of RNA isolation products — RiboPure™ RNA Isolation Kits — that combines these purification protocols specifically for producing RNA for array analysis applications.

Reverse transcription. Anchored versus unanchored oligo(dT)-T7 primers. Linear amplification begins with reverse transcription (RT) of total RNA using oligo (dT)-T7 primers, which can be anchored (i.e. contain a 3' A, G, or C) or unanchored. We have found that anchored and unanchored primers give rise to similar aRNA profiles, particularly under conditions where primers are in large excess to the RNA template. Under these conditions, multiple primers will bind to the poly(A) tail. During the RT reaction, primers that are bound to an internal poly(A) region will encounter an RNA-bound primer at their 3' end. Since the RT reaction is carried out under conditions where strand displacement activity is weak, the RT reaction will stop. This ensures that only the (unanchored) oligo(dT) primer bound adjacent to the 3'untranslated region (UTR) can make a cDNA product.

Second strand synthesis. For synthesis of 2nd strand cDNA, E. coli DNA polymerase, RNase H, and dNTPs are typically used in a 2 hr incubation carried out at 16 C. Most protocols recommend the addition of E. coli ligase to the reaction, but no advantage has been observed in adding the enzyme to the second strand synthesis reaction.

Clean up. Column purification of double-stranded cDNA. ARNA protocols typically call for purifying double-stranded cDNAs by phenol extraction or gel filtration columns.⁽⁵⁾ However, a special column-based purification method incorporated in Ambion's MessageAmp™ aRNA Amplification Kit (see below) consistently provides increased cDNA recovery (>90;%) and aRNA yields (3×) as compared to phenol extraction and gel filtration columns. Additionally, leaving out this purification step results in lower aRNA yields and shorter aRNA products. Interestingly, another report suggests that omitting the double-stranded cDNA purification step leads to production of nonspecific high molecular weight products, which are produced in a template-independent manner.⁽⁶⁾ These products presumably result from carryover of the oligo(dT) T7 primers to the in vitro transcription (IVT) reaction.

In vitro transcription. Yield versus time. Transcription initiation is the limiting step of in vitro transcription. We have noticed that increasing the in vitro transcription incubation time from 6 hours to overnight will improve consistency in aRNA yields. (Figure 3.) Using high concentrations of T7 polymerase increases the rate of initiation on rare or limited templates. The combination of excess polymerase and extended incubations results in the maximum possible yields.

click the image to enlarge Figure 3: The effect of IVT time on aRNA yield. 10, 100, or 500 ng of total RNA was amplified by using Ambion’s MessageAmp aRNA Kit. The MEGAscript technology was used for the in vitro transcription reaction up to a 24-hour period.

Putting it all together for exceptional amplification

Ambion scientists have streamlined and optimized each step in the aRNA procedure to develop the MessageAmp aRNA Kit. The first strand cDNA synthesis reaction is optimized to ensure that every cDNA bears a T7 promoter at its 5' end, and that even very limited amounts of mRNA are fully converted to full-length cDNA. The second strand cDNA synthesis reaction is designed for the efficient synthesis of full length, double-stranded cDNAs and the complete conversion of single-stranded cDNA into double-stranded transcription templates. The cDNA purification procedure not only removes enzymes, salts, and unincorporated dNTPs, but it efficiently removes RNA from the cDNA sample. This eliminates the heating or enzymatic digestion step that is commonly used in other procedures to degrade RNA (especially ribosomal RNA). The in vitro transcription reaction features Ambion's patented MEGAscript™ technology for maximal transcriptional amplification and yield of aRNA. The MEGAscript reaction used in the MessageAmp aRNA Kit is optimized to ensure the efficient transcription of limited amounts of template DNA as well as the synthesis of long transcripts. The NTPs for in vitro transcription are provided separately so that modified nucleotides (e.g. biotinylated CTP and UTP, or cyanine 3/cyanine 5 CTP and UTP) can be readily incorporated into aRNA. The rapid and simple aRNA purification procedure prepares the aRNA for downstream applications (reverse transcription or post-labeling reactions).

Typically, 70-80 μg of aRNA can be obtained from 2 μg of total RNA starting material (20-40 ng mRNA). When the MessageAmp aRNA products are analyzed by an Agilent 2100 bioanalyzer, their size typically falls between 0.2 kb to 6 kb, with the majority being around 1-1.5 kb long. After a second round of amplification, the size of aRNA products is usually shorter (in the 0.2 kb to 1 kb range), with the majority being around 0.5 kb long. (Figure 2)

Concluding remarks

The linear RNA amplification method has been an invaluable tool for obtaining sufficient amounts of aRNA for use in DNA microarray experiments. At Ambion, the Agilent 2100 bioanalyzer is routinely used to examine the quality and quantity of aRNA that is generated by the MessageAmp aRNA protocol. It has been found that the quality of starting RNA, inclusion of RNase H in the protocol, purification method of the double-stranded cDNA, and the length of the in vitro transcription reaction, can affect the profile of this aRNA. The MessageAmp aRNA Amplification Kit is the result of extensive optimization experiments and contains reagents optimized for each step of the aRNA procedure. Reagents for 20 reactions and a detailed Instruction Manual are also included.

About the author

Mehdi Banan, Ph.D., is Science Content and Business Development Analyst at Ambion Incorporated. More information about RNA amplification is available from: Ambion Inc., Austin, TX 800-888-8804; ambion.com.