Joseph Plurad and Stéphane Mabic
Introduction
Highly purified water has become a standard tool in molecular biology. Techniques such as reverse osmosis, electrodeionization, ultrafiltration, and chemical treatments provide water of varying quality and utility. Water quality specifications differ for applications such as liquid chromatography,
in situ hybridizations, single nucleotide polymorphism analysis, and DNA amplification. Experiments involving reverse-transcriptase PCR (RT-PCR), plasmid preparation, construction of cDNA libraries, and Northern blotting minimally require RNase-free water. However, RNA preparations from cells and tissues are complicated by the presence of ribonucleases (RNases).
Water treatment
Autoclaving and heat treatment do not satisfactorily inactivate RNase, which can survive heating to 180 C for 4 hours. By killing bacteria, autoclaving promotes release of bacterial RNases, which may regain significant activity at experimental temperatures.
Treatment with diethylpyrocarbonate (DEPC), an irreversible RNase inhibitor, has evolved as the standard method for preparing RNase-free water for PCR. DEPC reacts with an active-site histidine residue on RNase, forming a carbamate and liberating a molecule of carbon dioxide and ethanol. Since the chemical reaction is irreversible, RNase inactivation by DEPC is presumed to be complete and permanent. Although vendors and end-users employ DEPC to produce PCR-quality nuclease-free water, the DEPC method suffers from several drawbacks.
As a chemical treatment, DEPC merely inactivates RNase without removing it. Inactivated enzyme remains in solution. After a one-hour incubation, excess DEPC is destroyed by being heated to 121 C, hydrolyzing pyrocarbonate to carbon dioxide and generating ethanol as the side-product. Most of the ethanol evaporates from solution. What remains of side products and spent reagent contributes significantly to the water's total organic carbon (TOC) content.
By contributing carbonate/bicarbonate, DEPC decreases resistivity and lowers the solution's pH, which may contribute to RNA instability. As we have observed by monitoring DEPC-treated water, inactivated RNase may regain some activity over time.
TOC content and conductivity value of various purified waters from Millipore illustrates the potential unsuitability of DEPC-treated water for sensitive experiments. DEPC-treated water from a major vendor contained a TOC of 122,800 ppb and exhibited a conductivity of 2.9 μS/cm. Milli-Q water treated with DEPC was somewhat better, with a TOC of 15,700 ppb but even higher conductivity than commercial DEPC-treated water, 3.4 μS/cm. Although two other commercial RNase-free waters contained much lower levels of TOC (821 and 361 ppb), they showed high conductivities of 2.1 and 2.8 μS/cm, respectively.
Ultrafiltration (UF), the non-chemical alternative to DEPC treatment, removes RNase from water to below detectable limits and thereby removes, rather than contributing to, TOC. Typical UF methods employ polysulfone UF membranes embedded in acrylonitrile butadiene styrene polymer (ABS) housings. Neither membrane nor housing materials contribute to product water's ion or organic content. Milli-Q water treated by ultrafiltration contained less than 15 ppb of TOC and conductivity of 0.055 μS/cm corresponding to 18.2 M·cm resistivity.
Examples
Quantitative RNase recovery at high flowIn one experiment, a polysulfone Millipore Pyrogard UF membrane with a nominal molecular weight cutoff of 13,000 Da was challenged with 200 liter solutions containing 0.1 ng/ml of RNaseA (Roche Diagnostics, 109 142) at a throughput of 500 ml/min. This concentration of RNase is between 10-100 fold higher than is normally encountered in most PCR experiments. Total weight of collected RNase A was 22 mg, which was essentially quantitative (200 liters × 0.1 ng/ml).
RNase in sample waters
RNase activity was measured at various stages of water purification with the RNaseAlert test kit (Ambion), which claims a lower detection limit of 0.003 ng/ml. RNaseAlert uses a cleavable, RNase substrate standard labeled with a green fluorescent probe. After generating a calibration curve using an RNase standard (Ambion), we tested samples on an SFM 25 Konitron fluorometer, excited at 520 nm and read at 490 nm.
Table 1 lists RNase concentrations in various water samples as measured by the RNaseAlert assay. The table suggests that RNase is ubiquitous in both treated and untreated water samples. Even highly purified water (reverse osmosis, Elix water systems, unattended Milli-Q Gradient system) contains significant RNase activity that could affect PCR results. As expected, RNase-free, DEPC-treated waters, and all UF-treated samples (Millipore Pyrogard D UF cartridge; 13,000 Da NMWL cutoff) showed the lowest RNase levels, <0.003 ng/ml.
RNA stability
Rat thymus total RNA (Ambion) stability was measured for two purified waters using agarose gel microelectrophoresis with fluorescence detection through an Agilent Bioanalyzer 2100 (Figure 1). The top two electrophoregrams in Figure 1 show fluorescence peaks representing 18S and 28S rRNAs in DEPC-treated (left; BIO-101) and UF-treated waters after one hour. At this time point significant RNA sample remained intact in both DEPC- and UF-treated waters, although more RNA was present in UF water. The bottom electropherograms tell a different story. At 24 hours, significant intact RNA was evident in UF-treated water. However, significant degradation had occurred in the DEPC-treated water. DNA may have degraded due to the buildup of carbonate generation (and lower pH) resulting from DEPC degradation. Another possibility is that DEPC inactivation of RNase may not be as irreversible as is generally believed. After initial inhibition, the enzyme may regain significant activity over time.
Reverse-transcriptase polymerase chain reaction
To assess the suitability of UF-purified water for an RNase-sensitive PCR amplification, Millipore assessed PCR at five different initial RNA concentrations in water treated by reverse osmosis followed by ultrafiltration. To generate an amplification curve, 625, 1250, 2500, 5000, or 10,000 copies of pAW109 RNA (Applera, Monza, Italy) were added to water samples. RT-PCR reagents were purchased from Applera and amplified using an Applera GeneAmp PCR System 9700. After amplification, samples were analyzed on an agarose gel. DNA bands were visualized on a Bio-Rad Gel Doc instrument and quantified using Bio-Rad Quantity One software.
Figure 2 shows results of amplification for the five RNA samples. In the figure, data points represent the expected weight (in ng) of the sequential amplifications, while the vertical bars represent measured values. The high correlation between theoretical and experimental values indicates that PCR carried out in UF-treated water results in quantitative, reproducible amplification.
Summary
Although DEPC treatment has been the standard preparation method for molecular biology-grade water, including water used in PCR, chemical treatment suffers from serious drawbacks. Contribution to TOC, the potential reversibility of RNase inhibition, and lower pH may add to RNA degradation, resulting in the lowering of DNA amplification in PCR. In addition DEPC is toxic and has a measurable shelf-life.
By removing RNase rather than merely inactivating it, ultrafiltration eliminates all the drawbacks of DEPC treatment. Ultrafiltration is a rapid, reliable, and easy to use platform for quantitative, reproducible molecular biology experiments.
About the authors
Joseph Plurad, is product marketing manager, and Stéphane Mabic is worldwide applications support manager, both with the Lab Water Division of Millipore Corporation in Billerica, Mass.; joseph_plurad@millipore.com.
Additional information is also available from: