Validating RNA Quantity and Quality: Analysis of RNA Yield, Integrity, and Purity

Obtaining high-quality, intact RNA is the first and often the most critical step in performing many fundamental molecular biology experiments. Most RNA isolation products use the powerful chaotropic salt solution guanidinium isothiocyanate for sample lysis and homogenization, followed by organic extraction and alcohol precipitation or solid-phase purification. Organic extraction using acidified phenol and chloroform removes proteins, lipids, and DNA from the RNA sample, which is then recovered by alcohol precipitation. Solid-phase, column-based procedures utilize glass-fiber filters that bind RNA; proteins and DNA are removed by washing them through the filter. RNA is then eluted from the filter with RNase-free water. An alternative to column-based procedures is magnetic beads, which also bind RNA very efficiently under selective conditions.

Usually the first step after RNA isolation is to measure how much was recovered — the yield. There are also several methods for assessing RNA integrity and purity, both components of RNA quality, that may affect downstream applications. The following sections discuss aspects of measuring quantity and quality of isolated RNA.

Figure 1. Total RNA Yield (µg) from Different Tissues (mg) * g total RNA/mg tissue. This is intended as a general guide only and may vary depending on physiological state or organism.

CALCULATING RNA YIELDS

RNA yields vary widely depending on tissue or cell type, physiological state at the time of tissue removal, method of tissue disruption, and efficiency of RNA recovery (i.e., the RNA isolation method used, and the proficiency of the researcher performing the test). Figures 1 and 2 provide information about estimated yields of RNA (total + messenger RNA [mRNA]) from different sample types.

Figure 2. Yield of Total and mRNA from Different Tissues and Cells *Actual values may vary depending on tissue or cell type, physiological state, etc.

UV Spectroscopy

UV spectroscopy is a commonly used and easy method for quantification of RNA. The absorbance of a diluted RNA sample is measured at wavelengths of 260 nm and 280 nm. The nucleic acid concentration is calculated using the Beer-Lambert law, which predicts a linear change in absorbance with concentration. It is simple to perform, and UV spectrophotometers are available in most laboratories. Some of the more recently developed spectrophotometers (e.g., the NanoDrop ND-3300 Fluorospectrometer) make it possible to use only nanoliter to microliter volumes of sample.

Removing DNA Interference Because UV spectroscopy does not discriminate between RNA and DNA, it is advisable to first treat RNA samples with RNase-free DNase to remove contaminating DNA. Contaminants such as residual proteins and phenol can also interfere with absorbance readings, so care must be taken to remove them during purification steps following RNA isolation.

Factors Affecting Absorbance The A₂₆₀/A₂₈₀ ratio is dependent on both pH and ionic strength. As pH increases, the A₂₈₀ decreases while the A₂₆₀ is unaffected, resulting in an increasing A₂₆₀/A₂₈₀ ratio.¹ Water often has an acidic pH, it can lower the A₂₆₀/A₂₈₀ ratio. The remedy is to use a buffered solution with a slightly alkaline pH — such as Tris-EDTA (TE) at pH 8.0 — as a diluent, and as a blank to ensure accurate and reproducible readings.

Appropriate RNA Dilution The RNA dilution selected should be within the linear range of the spectrophotometer. Generally, absorbance values fall between 0.1 and 1.0 (representing 4 µg/ml to 40 µg/ml), but the spectrophotometer manual should be consulted to determine the linear range of the particular instrument. Solutions outside the linear range cannot be measured accurately, with the greatest error occurring at low concentrations. Because an A₂₆₀ of 0.1 corresponds to ~4 µg/ml RNA, it is impractical to use UV spectroscopy to quantitate RNA isolated from small samples that have lower concentrations once diluted.

Fluorescent Dyes

Certain fluorescent dyes, such as RiboGreen (from Molecular Probes, Inc.), exhibit a large fluorescence enhancement when bound to nucleic acids. As little as 1 ng/ml of RNA can be detected and quantitated using RiboGreen with a standard fluorometer, fluorescence microplate reader, or filter fluorometer. To accurately quantitate RNA, unknowns are plotted against a standard curve produced with a sample of known concentration, usually based on its absorbance at 260 nm. The linear range of quantitation with RiboGreen can extend three orders of magnitude (1 ng/ml — 1 µg/ml) when two different dye concentrations are used.

Removing Contaminants RiboGreen assays are relatively insensitive to non-nucleic acid contaminants commonly found in nucleic acid preparations, and linearity is maintained. However, as in UV spectroscopy, RiboGreen does not discriminate between RNA and DNA, and RNA samples must be treated with RNase-free DNase to remove contaminating DNA.

Preventing Nonspecific Adsorption and Photodegradation The RiboGreen reagent can adsorb to the sides of tubes. Adsorption can be minimized by preparing solutions in non-stick, nuclease-free polypropylene plastic ware. Nucleic acids will adsorb to the sides of even non-stick tubes with repeated freeze-thaw cycles. This is more noticeable at lower concentrations and will result in inflated RNA concentration estimates. The RiboGreen reagent can be protected from photodegradation by wrapping the container with foil and using the reagent within several hours of preparation. In addition, it is important to make a fresh set of standard for each curve and to carefully dilute the standard.

ASSESSING RNA QUALITY

Isolation of intact RNA is essential for many techniques used in gene expression analysis. Northern analysis, complementary DNA (cDNA) library construction, and RNA sample labeling for microarray analysis all require RNA-intact mRNA. Because reverse-transcriptase polymerase chain reation (RT-PCR) and ribonuclease-protection assays usually involve analysis of smaller regions of RNA (<1 kb), they usually are more tolerant of partially degraded RNA.

Assays for RNA Integrity

Agarose Gel Electrophoresis The most common method used to assess the integrity of total RNA is to electrophorese an aliquot of the RNA sample on a denaturing agarose gel stained with ethidium bromide. While native (non-denaturing) gels can be used, the results are difficult to interpret. The secondary structure of RNA alters its migration pattern in native gels so that it will not migrate according to its true size. Non-denaturing conditions also result in bands that are not sharp, and possibly in multiple bands representing different structures of a single RNA species.

Denaturing agarose gel systems include either formaldehyde and MOPs (3-[N-Morpholino]-propanesulfonic acid) electrophoresis buffer, or glyoxal in the loading buffer, to denature the RNA so the molecules will migrate according to size. The 28S and 18S ribosomal RNA (rRNA) bands are visualized by ethidium-bromide staining.

Intact eukaryotic RNA electrophoresed on a denaturing gel will have sharp, clear 28S and 18S rRNA bands. The 28S rRNA band should be approximately twice as intense as the 18S rRNA band (Figure 3). This 2:1 ratio (28S:18S) is a good indication of intact RNA. As RNA integrity decreases, the intensity of the 28S:18S band will decrease. Partially degraded RNA will have a smeared appearance and will lack the sharp rRNA bands. Completely degraded RNA will appear as a very low molecular-weight smear (Figure 3). Inclusion of RNA size markers on the gel is important to determine the molecular weight of RNA species and is an important control to ensure the gel was run properly and the RNA was not inadvertently degraded during gel preparation.

Figure 3. Intact vs. Degraded RNA Degraded total RNA and intact total RNA (2µg) were run along with RNA Millennium Markers (Ambion) on a 1.5% denaturing agarose gel. The 18S and 28S rRNA bands are clearly visible in the intact RNA sample. The degraded RNA appears as a lower molecular weight smear.

Lack of Sensitivity The amount of RNA needed for visualization using agarose gel electrophoresis makes it a less-than-attractive option, particularly when RNA yields are low, as is the case with RNA preparations from needle biopsies or laser-capture microdissected samples. Generally, at least 1 µg total RNA or mRNA must be loaded to be seen by ethidium-bromide staining. Alternative nucleic acid stains, such as SYBR Gold and SYBR Green II RNA gel stain (from Molecular Probes, Inc.), offer a significant increase in sensitivity compared to the traditional ethidium-bromide stain in agarose gels. Using a 300 nm transilluminator (6 x 15-watt bulbs) and a special filter, as little as 1 ng RNA can be detected with SYBR Gold and 2 ng with SYBR Green II RNA gel stain, allowing less sample to be used for pre-experimental integrity assessment.

Visual assessment of the 28S:18S rRNA ratio on agarose gels can be subjective because the appearance of rRNA bands is affected by electrophoresis conditions, amount of RNA loaded, and saturation of fluorescence. In the case of low yields, it may be difficult to spare 1 µg RNA to assess integrity. One option is to use the Agilent 2100 bioanalyzer, which can analyze the integrity of picogram to nanogram amounts of RNA, depending on the assay selected. The most widely used assay is the RNA 6000 Nano assay, which has a linear range of 5-500 ng/ml RNA. For limiting samples, Agilent Technologies has developed the RNA 6000 Pico assay, which has a linear range of 200-5000 pg/µl.

Agilent 2100 Bioanalyzer The Agilent 2100 Bioanalyzer uses a combination of microfluidics, capillary electrophoresis, and fluorescent dyes that bind to nucleic acid, and it simultaneously evaluates both RNA concentration and integrity. As RNA moves through the separation channel of the LabChip, intercalating dye within the sieving matrix binds the RNA, and the fluorescence of these molecules is measured as they pass the detector. The output is a scan of mass versus size. The 28S:18S rRNA ratio is calculated by integrating the areas of 18S and 28S rRNA peaks and then dividing the area of the 18S rRNA peak into the area of the 28S rRNA peak. As each RNA sample is analyzed, the software generates both a gel-like image and an electropherogram (Figure 4).

Figure 4. Agilent 2100 Bioanalyzer Data Electropherogram of a high quality, eukaryotic, total RNA sample. The 18S and 28S rRNA peaks are clearly visible at 39 and 46 seconds, respectively.

The RNA Integrity Number The RNA Integrity Number (RIN) is a software tool designed by Agilent Technologies to estimate the integrity of total RNA samples. Sample integrity is no longer determined by the ratio of the ribosomal bands, but by the entire electrophoretic trace of the RNA sample, including the presence or absence of degradation products. The algorithm assigns a RIN score of 1 to 10, where level 10 RNA is completely intact. Because the analysis of the electropherogram is automated and not subject to individual interpretation, universal and unbiased comparison of samples is enabled and experimental precision is improved.²

The rRNA Ratio The 28S:18S rRNA ratio has traditionally been viewed as the primary indicator of RNA quality, with a ratio of 2.0 considered to indicate high-quality, intact RNA. However, it has become increasingly clear that the long-time standard of a 2.0 rRNA ratio is difficult to meet, especially for RNA derived from clinical samples. In high-quality RNA (rRNA ratio near 2.0), the base line above and below the 18S and 28S rRNA peaks will be relatively flat. As 28S rRNA breaks down, the degradation products cause the base line between and below the 18S and 28S rRNA peaks to rise. Some total RNA samples might contain classes of small RNAs that appear to be breakdown products but are actually abundant tissue-specific mRNAs (Figure 5). These abundant small RNAs are most often found in RNA isolated from reproductive and intestinal tissues. Their fluorescence can sometimes dwarf that of the rRNA. Generally, they can be distinguished from degradation products because the rRNA ratio of the sample will be greater than 1.0 and the base line between the small RNAs and the 18S RNA will be relatively low.

Figure 5. Different Types of Genomic DNA Contamination In Total RNA Preparations Agilent 2100 bioanalyzer electropherograms of RNA contaminated with high molecular weight genomic DNA (A), partially sheared genomic DNA (B), and small DNA fragments generated by incomplete DNase I digestion (C).

ASSESSING RNA PURITY

Because most RNAs are used in enzymatic applications, residual contaminants can critically affect the quality of downstream assays. The most intact RNA will not perform well if the sample contains residual organics, metal ions, or nucleases that compromise integrity or interfere with enzymatic reactions.

Assay for RNA Purity To ensure that RNA is free of contaminants, one should perform a stability test by incubating a small amount of RNA at 37°C for several hours and comparing it to a duplicate sample stored at -20°C. The sample stored at 37°C should show minimal decrease in the 28S:18S rRNA ratio compared to the one stored at -20°C. Samples with >20% change in rRNA ratio may not perform well in downstream applications.

Organic Contaminants Organic contaminants are most often a problem when RNA is isolated using single-step organic extraction protocols. Although relatively fast and easy, single-step extraction may not be sufficient to remove contaminants from some tissues, especially if one exceeds the recommendation for input sample amount. Also, organic contaminants are often carried over into samples during aqueous phase transfer. To avoid this, one should leave some of the aqueous phase behind during phase separation. A combination of phenol-based and solid-phase extraction methods is recommended to avoid inadvertent carry-over of contaminants. Several studies have shown that RNA isolated by a combination of both techniques provides superior array data than RNA isolated by either method alone.^3,4,5

Genomic DNA Contamination Residual genomic DNA contamination is most problematic in tissues with high cell densities, such as spleen or tissue-culture cells. Figure 6 shows examples of gross contamination with genomic DNA. High-molecular-weight genomic DNA typically migrates as a broad, larger molecular-weight peak that is well-separated from rRNA peaks (Panel A). Note, also, that the base line is high in this electropherogram — generally a signature of underlying genomic DNA contamination. Genomic DNA that has been partially sheared can sometimes migrate between the 18S and 28S rRNA ribosomal bands (Panel B), making it difficult to accurately determine the rRNA ratio. Incomplete DNase I digestion can generate small-molecular-weight DNA fragments of 50 to 200 bases in size (Panel C).

Figure 6. Low Baseline Fluorescence Indicates Good RNA Quality (A) Partially degraded total RNA.The 28:18S rRNA ratio is 0.9 and the baseline fluorescence is elevated both between the 28S and 18S rRNA and below the 18S rRNA, making the rRNA peaks appear to be riding on top of the baseline. (B) Human intestinal RNA containing abundant small RNAs. This sample can be distinguished from partially degraded RNA by the relatively high 28S:18S rRNA ratio, and the relatively low baseline both between the 28S and 18S rRNA peaks, and immediately below the 18S rRNA peak.

Small RNA Contamination Small RNAs, such as 5S rRNA and transfer RNAs (tRNAs), are efficiently recovered in organic extraction methods but are depleted in column-based purification methods. Total RNA samples in which small RNAs have been removed by solid-phase extraction (following organic extraction) show superior performance in RNA amplification, compared to RNA isolated by organic extraction only. Because small RNAs can comprise as much as 15% of total RNA, their removal effectively increases the percent of mRNA within the total RNA sample and decreases the potential for interference during cDNA synthesis of mRNA.

Small RNA Enrichment Small RNAs can be considered a contaminant in some situations and are not captured by most traditional RNA isolation methods. However, as more is learned about RNA interference and microRNA, scientists have begun focusing their research on small RNAs. Several isolation methods have been developed that specifically enrich small RNAs and can separate precursor microRNAs from mature, active microRNAs.^6,7

CONCLUSION

With the emergence of more sophisticated techniques for gene expression analysis (e.g., microarray analysis, real-time RT-PCR), the ability to quantitate smaller samples and assess the quality of the RNA samples studied is critical. Evaluation of the RNA sample by different assays to quantify RNA yield and to validate RNA quality — including mRNA integrity, genomic DNA contamination, presence/absence of enzymatic inhibitors, and ribosomal RNA contamination in poly(A) preparations — is critical for future results.

Sapna Chacko, Ph.D., technical writer/editor 17820 Windflower Way, Unit 502 Dallas, Texas 75252 USA, Phone and Fax: 469.828.4695, sapnachacko@gmail.com

REFERENCES

1. Wilfinger WW, Mackey K, Chomczynski P. Effect of pH and ionic strength on the spectrophotometric assessment of nucleic acid purity. Biotechniques. 1997;22:474--481.

2. RNA Integrity Number (RIN). Agilent Technologies web site. Available at: http://www.chem.agilent.com/scripts/generic.asp?lpage=14975&indcol=Y&prodcol=Y.

3. Hakak Y, Walker JR, Li C, et al. Genome-wide expression analysis reveals dysregulation of myelination-related genes in chronic schizophrenia. PNAS. 2001;98: 474-6-51.

4. Jost JP, Oakeley EJ, Zhu B, et al. 5-Methylcytosine DNA glycosylase participates in the genome-wide loss of DNA methylation occurring during mouse myoblast differentiation. Nucleic Acids Research. 2001;29:445-2-61.

5. Wang E, Miller LD, Ohnmacht GA, Liu ET, Marincola FM. High-fidelity mRNA amplification for gene profiling. Nature Biotechnology. 2000;18: 457-479.

6. The RNA Interference Resource. Ambion, Inc. web site. Available at: www.ambion.com/RNAi.

7. miRNA Resource. Ambion, Inc. web site. Available at: www.ambion.com/miRNA.