Purifying Small RNAs: Using Mature MicroRNAs for Microarray Analysis
| flashPAGE™ Electrophoretic Fractionator Small RNA analysis methods (e.g., hybridization assays, expression profiling) require various RNA input amounts and sizes. Small RNA fractions often include tRNA, 5S rRNA, 5.8S rRNA, small mRNA fragments, and precursor miRNAs. The presence of these RNA species does not affect Northern analysis and solution hybridization; however, they can obscure results of some experiments, particularly microarray studies. To obtain mature miRNA, use the flashPAGE™ Electrophoretic Fractionator, which purifies small RNA species (15–35 nt range) from total RNA samples. Here we show how an RNA fraction that contains only mature miRNA can provide better miRNA profiling results than an enriched small RNA (or total RNA) sample. This is just one example of the increased sensitivity gained by further removing unwanted RNA species from the sample to be analyzed. These sensitive methods may be required for detection and identification of less highly expressed miRNAs. |
In microarray experiments, unwanted small RNA species not only compete for reagents used in labeling reactions but also add to background signals and potentially hybridize to probes. Figure 1 shows two microarrays hybridized with RNA that was from the same RNA source but purified by two different means, so that the enriched small RNA fraction contains RNA <200 nt and the mature miRNA fraction, isolated with the flashPAGE Fractionator, contains mature miRNA (15–35 nt). Duplicate bladder and lung samples (15 µg total RNA input that was enriched or fractionated) were compared to each other to provide a two-color readout and provided consistent results. The higher signal intensity and lower background from the samples that contain only mature miRNA (flashPAGE-purified RNA) are immediately noticeable. The patterns are recognizable between the two arrays, but some information is more apparent, primarily with lower-strength signals. This difference is quantified by assessing the correlation with a standard for the tissues studied (Figure 2). The standard represents the average of many assays from the same tissues, and deviation is minimal from samples using the same (non-diseased) tissues.
Figure 1. Representative Arrays Using Enriched or flashPAGE™-purified Samples. Target RNA samples from human bladder and lung samples prepared from the same parent total RNA (A) RNA enriched small RNA (<200 nt) and (B) RNA fractionated with the flashPAGE Fractionator were labeled using the mirVana™ miRNA Labeling Kit and compared on identical arrays (custom-printed arrays using the mirVana miRNA Probe Set): Red indicates higher expression in bladder, green indicates higher expression in lung.
Figure 2. Correlation of Bladder and Lung Data with Standard Profiles Using Enriched or flashPAGE™-purified Samples. Data such as that shown in Figure 1 were compared to consensus data from many trials of relative expression between the two tissues. The percent of correlation was compared for 10 µg, 5 µg, and 1 µg of input total RNA (prior to either enrichment or flashPAGE purification). The first set of columns in each figure shows the correlation if all spots on the test arrays above background are used (Basic Filters), while the second and third sets show a similar comparison if only array signals with a Sum Intensity >3000 or >4500 are compared.
The correlation stays high as the amount of input is decreased from 10 to 1 µg with samples purified with the flashPAGE Fractionator (Figure 2B), but drops radically for those samples merely enriched for small RNA (Figure 2A). This drop is due to low-intensity signals as indicated by the higher correlation from the enriched samples (Figure 2A) when only those signals with very high intensities over 4500 quanta are compared.
As more microRNAs (miRNAs) are discovered, it is becoming clear that the first wave of miRNAs that were identified are expressed at a higher level than the ones now being discovered. This means that determining expression levels of many (perhaps the vast majority) of miRNAs will require a highly-sensitive method. The flashPAGE Fractionator, especially in conjunction with the mirVana miRNA Microarray System, is designed for that purpose.
Scientific Contributors
Kerri Keiger, Patricia Powers, Rick Conrad • Ambion, Inc.
The correlation stays high as the amount of input is decreased from 10 to 1 µg with samples purified with the flashPAGE Fractionator (Figure 2B), but drops radically for those samples merely enriched for small RNA (Figure 2A). This drop is due to low-intensity signals as indicated by the higher correlation from the enriched samples (Figure 2A) when only those signals with very high intensities over 4500 quanta are compared.
As more microRNAs (miRNAs) are discovered, it is becoming clear that the first wave of miRNAs that were identified are expressed at a higher level than the ones now being discovered. This means that determining expression levels of many (perhaps the vast majority) of miRNAs will require a highly-sensitive method. The flashPAGE Fractionator, especially in conjunction with the mirVana miRNA Microarray System, is designed for that purpose.
Scientific Contributors
Kerri Keiger, Patricia Powers, Rick Conrad • Ambion, Inc.
${comergentUrl}
