- Email: [email protected]
Nonisotopic Quantitation of mRNA Using a Novel RNase Protection Assay: Measurement of erbB-2 mRNA in Tumor Cell Lines
ANALYTICAL BIOCHEMISTRY ARTICLE NO.
242, 214–220 (1996)
0455
Nonisotopic Quantitation of mRNA Using a Novel RNase Protection Assay: Measurement of erbB-2 mRNA in Tumor Cell Lines Samuel D. H. Chan, Kilian Dill, Jayne Blomdahl, and H. Garrett Wada1 Molecular Devices Corporation, Sunnyvale, California 94089
Received May 3, 1996
We have developed a nonisotopic RNase protection assay using RNA probes that are dual-labeled with biotin and fluorescein for detection. This system utilizes capture of the protected RNA probe hybrids to streptavidin-coated membranes attached to plastic dipsticks, complexing of anti-fluorescein–urease conjugate with the labeled RNA probe, and quantitative detection of the membrane-bound complex by a potentiometric silicon sensor. The dual-label RNase protection (RP) assay was capable of measuring b-actin mRNA in cellular RNA samples at the 27- to 45-amol level (10–17 pg) with high precision (%CV õ 7). We have used this method to quantitate the levels of erbB2 mRNA in the human tumor cell lines SKBR-3, SKOV3, and MCF-7. The levels of erbB-2 mRNA in these cells were 105, 190, and 0.9 amol per microgram of cellular RNA, respectively. The dual-label RP method should be useful for measuring the mRNA expression for other erbB-2 homologs such as erbB-3 and erbB-4 in tumor cells and tissues and can be a generally useful mRNA quantitative method for laboratories wishing to minimize radioisotope use. q 1996 Academic Press, Inc.
The detection and quantitation of specific mRNA in tissues and cell cultures play an important role in analyzing gene expression. While Northern or slot/dot blotting hybridization using radiolabeled or nonisotopic labeled nucleic acid probes is the commonly used method, solution hybridization RNase protection (RP)2 assay has proven to be more sensitive and more tolerant of 1
To whom correspondence should be addressed. Fax: 408-7473601. 2 Abbreviations used: RP, RNase protection; ILA, immunoligand assay; CHO, Chinese hamster ovary; DEPC, diethyl pyrocarbonate; LAPS, light-addressable potentiometric sensor; CV, coefficient of variation.
partially degraded RNA. The assay involves hybridization of RNA samples to radiolabeled antisense RNA probe in solution, followed by digestion of unhybridized single-stranded RNA with RNases. The protected radioactive hybrid fragments are recovered by ethanol precipitation and analyzed by denaturing polyacrylamide gel electrophoresis and autoradiography (1). Alternatively, the hybridized nucleic acids are captured onto nitrocellulose or nylon membrane, and radioactivity is then detected by liquid scintillation counting (2). One major drawback of this technique, however, is the requirement for radioactivity. Because of the increasing concerns about the health hazards and high disposal cost of radioactive waste, we have developed a nonisotopic RP assay using RNA probes dual-labeled with biotin and fluorescein, with subsequent detection of the protected probes using the immunoligand assay (ILA) system (3) on a light-addressable potentiometric sensor (LAPS) that has been used extensively for lowlevel detection of various analytes including total DNA and sequence-specific DNA (4, 5). We have used this method to quantitate the levels of the erbB-2 oncogene mRNA in human tumor cells. The erbB-2 mRNA encodes the HER2 tyrosine kinase receptor which is overexpressed in some human neoplasias (7). MATERIALS AND METHODS
Cell Culture and Transfection CHO-K1 (ATCC) and CHO-FE6 (14) cells were cultured in Ham’s F-12 containing 10% fetal bovine serum, 2 mM L-glutamine, and antibiotics (100 units/ml penicillin and 100 mg/ml streptomycin). HEK-293 (ATCC) cells were grown in DME supplemented with 10% fetal bovine serum, 2 mM L-glutamine, 100 units/ ml penicillin, and 100 mg/ml streptomycin, while MCF7 (ATCC), SKBR-3 (ATCC), and SKOV-3 (ATCC) cells were cultured in RPMI 1640 containing 10% fetal bo-
214
AID
0003-2697/96 $18.00 Copyright q 1996 by Academic Press, Inc. All rights of reproduction in any form reserved.
AB 9800
/
6m20$$$501
10-17-96 06:53:48
aba
AP: Anal Bio
QUANTITATION OF mRNA USING A NOVEL RNase PROTECTION ASSAY
vine serum, 2 mM L-glutamine, and antibiotics (100 units/ml penicillin, 100 mg/ml streptomycin). CHO cells were transiently transfected with erbB cDNAs encoded in pCEP4 expression plasmid constructs using lipofectamine as described by the manufacturer (Life Technologies). Cells were harvested for RNA extraction 48–72 h after transfection as described below. Total RNA Preparation Total RNA was purified from confluent cell cultures grown in T75 flasks by a modification of the method of Chomczynski and Sacchi (8). The culture medium was decanted and the cells were lysed by incubating with 5 ml of guanidinium thiocyanate/phenol solution (TRIZOL reagent, Life Technologies) for 5 min. One milliliter of chloroform was added to the lysate and the mixture was vortexed for 10 s and let stand at room temperature for 5 min. After centrifugation at 10,000g for 15 min at 47C, RNA in the upper aqueous phase was extracted with an equal volume of phenol (pH 4.0)/ chloroform. RNA was then precipitated with 1 vol of isopropanol at 0207C for 1 h and pelleted by sedimentation at 10,000g for 15 min. The resulting RNA was dissolved in 20 ml DEPC-treated water and stored at 0207C. The concentration of RNA was determined by absorbance at 260 nm (1 A260 unit Å 40 mg/ml). Templates for RNA Probe Synthesis The plasmid pBS-ActB containing 1.1 kb of human b-actin cDNA in pBluescript SK0 was obtained from ATCC (HHC189). The 1.1-kb insert was excised and subcloned in the opposite orientation into pBluescript II SK0 (Stratagene) to generate the plasmid pBSActBT3. pBS-h-erbB-2 was constructed by cloning a 4.0-kb cDNA fragment containing the complete open reading frame of the human erbB-2 gene into pBluescript II SK0. pBS-ActB and pBS-h-erbB-2 were digested, respectively, with HindIII and BamHI, to generate templates for the synthesis of antisense RNA probes. In both cases, a 1.1-kb probe was prepared using T3 RNA polymerase. The sense b-actin probe was prepared using a HindIII-digested pBS-ActBT3 plasmid and T7 RNA polymerase. The single-stranded sense b-actin cDNA for a b-actin calibration standard was prepared using superinfection with the helper phage R408 (9) of an overnight culture containing the pBS-ActBT3 plasmid in host bacteria XL-1 Blue (Stratagene). ErbB-3 and erbB-4 RNA probes approximately 1 kb in size were prepared as described for erbB-2 from RT-PCR cloned sequences containing the entire open reading frame for these two genes. Synthesis of Dual-Labeled RNA Probes for b-Actin and erbB-2 mRNA Detection Dual-labeled RNA probe was synthesized in a 20-ml reaction containing 40 mM Tris–HCl (pH 8.0), 8 mM
AID
AB 9800
/
6m20$$$501
10-17-96 06:53:48
215
MgCl2 , 2 mM spermidine, 25 mM NaCl, 10 mM DTT, 0.5 mM ATP, 0.5 mM GTP, 0.25 mM CTP, 0.25 mM biotin-14–CTP (Life Technologies), 0.325 mM UTP, 0.175 mM fluorescein-12–UTP (Boehringer-Mannheim), 1 mg of template DNA, 40 units of RNasin (Promega), and 50 units of T3 RNA polymerase (Life Technologies). The reaction mixture was incubated at 377C for 1 h, after which the reaction was stopped by heating to 657C for 10 min. The DNA template was then digested by incubating with 14 units of RNase-free DNAse I (Boehringer-Mannheim) at 377C for 15 min. The RNA probes were purified by centrifugation through DEPC-treated G-100 or G-200 spin columns (Clontech). The concentration of the probes was determined by absorbance at 260 nm, then 40 units of RNasin was added, and the probes were stored frozen at 0207C. Characterization of the RNA Probes The RNA probes were characterized with respect to their size, biotin and fluorescein content, and reactivity in the ILA. Molecular size of the RNA probes was determined using electrophoresis in 6% polyacrylamide, Tris-borate-EDTA pH 8.5 gels containing 6 M urea. Samples of 10 to 100 ng of RNA probe were electrophoresed in parallel with low-molecular-weight (1000–50 bases) DNA standards (BioVentures) after heating to 857C for 5 min to denature the samples in 40% formamide sample buffer. The gels were fixed in 50% methanol/10% acetic acid and nucleic acid bands were detected by silver staining (LabLogix Inc.). The fluorescein content of the RNA probe was determined by fluorimetry using a fluorescein reference curve, and the biotin content was determined using dot blot analysis. In brief, the dot blots containing serially diluted probe and standard (5*-biotinylated 20-mer oligonucleotide, Operon Technologies) were incubated with streptavidin–alkaline phosphatase conjugate followed by detection with a chemiluminescent dioxetane substrate (DNA/RNA Detection Kit, LabLogix Inc.) and X-ray film. The dot blots of probe dilutions were interpolated on the biotinylated oligonucleotide reference curve as a semiquantitative measure of their biotin content. Detection of the Dual-Labeled RNA Probe Using the ILA System The reactivity of the dual-labeled probe in the Threshold ILA System (Molecular Devices Corp.) was determined by titrating the probe using the LAPSbased enzyme detection system (4). The dual-labeled probe was diluted into 100 ml of assay buffer (phosphate-buffered saline containing BSA and nonionic detergent) and mixed with 100 or 200 ml of capture reagent (assay buffer containing 2 or 4 mg of streptavidin). The samples were then filtered onto a prewashed
aba
AP: Anal Bio
216
CHAN ET AL.
biotinylated nitrocellulose membrane (mounted on a plastic dipstick) at low vacuum setting, whereby the streptavidin–mRNA–RNA probe complexes were captured and bound to the membrane. The urease conjugate of a rabbit anti-fluorescein antibody (Ç1.5 mg) in 200 ml of assay buffer was filtered through the membrane at low vacuum setting followed by 0.5 ml of wash buffer (phosphate-buffered saline containing nonionic detergent) at high vacuum setting. The quantity of bound enzyme conjugate is directly proportional to the amount of hybridized RNA probe complex bound to the membrane. The enzyme activity is measured in the presence of urea substrate by a potentiometric sensor and expressed in mV/s which is equivalent to 1 millipH unit/min (5). The Threshold ILA System provides for internal calibration of each assay by comparing a negative and positive control to a reference curve that can be generated by serial dilution of the dual-labeled probe (6). Dual-Label RNase Protection Assay Protocol The dual-labeled RNA probes were hybridized with specific mRNA present in total RNA prepared from cell extractions. Typically, 10–20 mg of cellular RNA (or varying amounts if preparing a reference curve) was mixed with yeast tRNA to give 20 mg of total RNA and 3–10 ng of RNA probe in 30 ml of a hybridization buffer containing 80% formamide, 40 mM Mes buffer (pH 6.5), 400 mM NaCl, and 1 mM EDTA. The mixture was heated to 857C for 10 min and then incubated at 427C overnight. In some experiments hybridization time and temperature were varied, and alternatively 557C for 4 h could be used. As a negative control, the RNA probe was mixed with 20 mg of yeast tRNA and the above protocol was followed. The next day, the samples were mixed with 350 ml of the RNase digestion buffer composed of 10 mM Tris–HCl (pH 7.6), 300 mM NaCl, 5 mM EDTA, 125 unit/ml RNase T1, and 6.25 mg/ml RNase A (Ambion). Samples were digested at 377C for 1 h. Following this, the samples were extracted with 400 ml of phenol/chloroform and centrifuged. The aqueous layer was removed and 20 mg of yeast tRNA, 100 ml of 7.5 M ammonium acetate, and 500 ml of isopropanol were added. The mixture was thoroughly mixed and placed at 0207C for 1 h. The samples were then centrifuged at 47C for 15 min at 10,000g. The supernatant was carefully removed and the pellet was allowed to air dry for a few minutes. The samples were then redissolved in 20 ml of TE buffer (10 mM Tris–HCl, 1 mM EDTA, pH 8.0) for ILA detection of hybridized RNA probe as described above. RESULTS
Characterization of Dual-Labeled RNA Probes for b-Actin The RNA probes for b-actin and erbB-2 were designed to be complementary to a 1100-base stretch of
AID
AB 9800
/
6m20$$$501
10-17-96 06:53:48
FIG. 1. Polyacrylamide gel electrophoretic analysis of dual-labeled RNA probes. Dual-labeled RNA probes for b-actin and erbB-2 were analyzed by electrophoresis in 6% acrylamide gels containing 6 M urea as described under Materials and Methods. (A) 25 ng of erbB2 RNA probe was applied in lane 1 and 25 ng of b-actin RNA probe in lane 2. The electrophoresed samples were silver stained to detect the RNA bands. (B) erbB-2 and b-actin RNA probes were digested with RNase A/RNase-T1 prior to electrophoretic analysis as described for A. 25 ng of digested erbB-2 RNA probe was applied in lane 1 and 25 ng of digested b-actin RNA probe in lane 2.
the 3* terminal coding sequence for the b-actin and erbB-2 mRNA. When analyzed by electrophoresis in denaturing (6 M urea) polyacrylamide gels, a major band was observed at 1.1 kb and small amounts of nucleic acid with progressively smaller size were observed (Fig. 1). The RNA composition of these silverstained bands was confirmed by their lability to RNase treatment. The fluorescein content of the two RNA probes was determined by fluorimetry and indicated that, under the conditions employed for synthesis, approximately 11 fluorescein–UTPs were incorporated per 1.1 kb of RNA probe. The biotin content of the RNA probes was also determined by dot blot to be Ç10 biotin–CTPs incorporated per 1.1 kb of b-actin RNA probe and Ç22 biotins per 1.1 kb of erbB-2 RNA probe. The reactivity of the RNA probe with the aforementioned dual-label composition was characterized by titration in the ILA system. By virtue of their duallabeled nature, the RNA probes cause anti-fluoresceinconjugated urease to bind to biotinylated membrane via streptavidin/biotinylated RNA probe complexes. The membrane-bound urease is then detected by the pH-sensitive, potentiometric silicon sensor in the Threshold instrument. Figure 2 shows the dose response curve obtained by testing dilutions of the bactin and erbB-2 RNA probes in the Threshold instrument. The signal unit is mV/s which is equivalent to the milli-pH unit/min rate of pH change. The slope of the dose–response curves indicates that 1 fmol (10015 mol) of b-actin probe (363 pg) generates a signal of 461 mV/s, approximately 1 mV/s of signal per picogram of
aba
AP: Anal Bio
217
QUANTITATION OF mRNA USING A NOVEL RNase PROTECTION ASSAY
FIG. 2. Dose response of b-actin and erbB-2 dual-labeled RNA probes using the ILA method. The RNA probes for b-actin and erbB2 dual-labeled with biotin and fluorescein were serially diluted in ILA assay buffer and detected in the ILA assay protocol as described under Materials and Methods. The slopes of the response curves were calculated using a least-squares regression equation. The erbB2 curve is plotted with a solid line (slope Å 1797 mV/s/ng) and the b-actin curve is plotted with a dashed line (slope Å 999 mV/s/ng).
probe, and the erbB-2 probe generates approximately 1.8 mV/s of signal per picogram of probe. Hybridization Kinetics and Assay Performance Using the b-Actin RNA Probe When 10 mg of RNA from HEK-293 human kidney cells was hybridized with anti-sense b-actin RNA probe at 427C for 24 h, RNase protection was observed which corresponded to 2.8 fmol of b-actin RNA probe and 0.025% of the original cellular RNA tested (Fig. 3A). The same hybridization experiment using dual-labeled sense b-actin RNA probe showed no protection from RNase when hybridized to HEK-293 RNA, indicating that the RNase protection is sequence specific (Fig. 3A). The kinetics of hybridization were examined at 427C and indicated approach to equilibrium in about 4–8 h as shown in Fig. 3B. The hybridization kinetics were also similar for erbB-2 RNA probe hybridizing to target mRNA extracted from CHO-FE6 cells that are stably transfected with erbB-2 (14). These results indicated that the hybridization time could be shortened to 8 h without loss of sensitivity and that the hybridization kinetics do not appear to be sequence specific. The effect of temperature on hybridization was also investigated. By elevating the temperature from 42 to 557C a large increase in signal was observed after a 2-h hybridization (data not shown). This suggested that elevating the temperature may further shorten the hybridization time. Assay precision for b-actin RP assay was determined by testing 0, 2.5, 5, and 10 mg of HEK293 RNA in quadruplicate using the 427C overnight hybridization (Table 1). The precision was excellent, ranging from 1.4 to 6.9% CV. The lower limit of detection calculated from the 21 standard deviation range
AID
AB 9800
/
6m20$$$501
10-17-96 06:53:48
FIG. 3. Protection of hybridized dual-labeled RNA probes. Total RNA prepared from HEK-293 cells was tested in the dual-labeled RP assay for b-actin mRNA. The hybridization was conducted at 427C unless otherwise specified. (A) The three levels of total RNA were hybridized overnight in quadruplicate along with a negative control (0, 2.5, 5.0, and 10 mg total RNA) using anti-sense b-actin RNA probe (solid circles). The hybridized samples were RNAse digested and processed as described under Materials and Methods. The error bars indicate the mean value {1 SD. The sense b-actin RNA probe (open circles) was also used to test these same samples in duplicate. (B) Total cellular RNA was hybridized in duplicate for 1, 2, 4, 8, and 24 h with dual-labeled RNA probe, after which the samples were stored at 0207C. The samples were all tested the next day in the dual-labeled RP assay. Ten-microgram samples of RNA from HEK293 cells were tested using b-actin RNA probe (open circles), and 20-mg samples of RNA from CHO-FE6 cells were tested using erbB-2 RNA probe (solid circles). The assay signal was background subtracted and adjusted to a percentage of the maximum signal obtained during the time course and plotted versus time. The assay background for the b-actin assay was 196 mV/s and for the erbB-2 assay 80 mV/s.
of the negative control and the 1 mV/s signal gain per picogram of RNA probe (1.1 kb) was 26 amol (10018 mol or 9.5 pg) of b-actin (1.75 kb) which represents 0.0001% of a 10-mg sample of cellular RNA. TABLE 1
Precision of the b-Actin mRNA RPA Measurement in HEK-293 Cellsa Cell RNA (mg)
mRNA ng
% of total RNA
fmol/mg
% CV
0 2.5 5.0 10.0
0.400 0.781 1.506
0.025 0.025 0.024
0.27 0.27 0.26
2.6 3.4 6.9 1.4
a
Mean values from four replicates (N Å 4) from data in Fig. 3A. % CV is coefficient of variation of the signal in mV/s.
aba
AP: Anal Bio
218
CHAN ET AL.
FIG. 4. Dose response of b-actin single-stranded cDNA tested in dual-label RNase protection assay. Single-stranded b-actin cDNA prepared as described under Materials and Methods was serially diluted, and duplicate samples were tested in the dual-labeled RP assay protocol using ILA detection. The slope of the b-actin cDNA response curve was calculated to be 1352 mV/s/ng (788 mV/s/fmol) using a least-squares regression equation. The error bars represent the {11 SD range about the mean.
The sensitivity of the assay for b-actin was demonstrated using a single-stranded sense target cDNA produced in the pBluescript phagemid system (Stratagene). Figure 4 shows a dose–response curve using the b-actin sense cDNA that gave 1.1 mV/s signal per picogram of target cDNA that is equivalent to 788 mV/ s signal per femtomole of target. The high precision at the low end of the dose–response curve made the calculated lower limit of detection 27 amol (10 pg) of cDNA at the 99% confidence interval (31 standard deviation above the mean negative control value). Spike recovery experiments were run using HEK-293 total RNA. This entailed testing 5 mg of the RNA with and without the addition of 380 and 760 pg of b-actin singlestranded sense cDNA standard. The spiked standard DNA gave a recovery of signal equivalent to 90% of the theoretical signal, indicating that the effective sensitivity in cell samples is approximately 30 amol of target sequence (data not shown). This value was similar to the detection limit calculated from the HEK-293 total RNA dose–response curve for b-actin mRNA and the mV/s signal per picogram of RNA probe value as described above. Assay for erbB-2 mRNA Transcripts in Mammary Carcinoma Cell Line Extracts Several tumor cell lines of mammary carcinoma origin (SKBR-3, SKOV-3, and MCF-7) were tested for the production of erbB-2 mRNA using the dual-labeled RNA probe RP assay. The total RNA samples prepared from the cell lines were also tested as described above for b-actin mRNA as an internal control for RNA degra-
AID
AB 9800
/
6m20$$$501
10-17-96 06:53:48
dation. The results from the erbB-2 and b-actin assays are shown in Table 2, which presents the erbB-2 mRNA level in mV/s, fmol/mg RNA, and percentage of total RNA. The amount of erbB-2 mRNA in samples was determined from the erbB-2 RNA probe reference curve using a linear regression analysis to calculate the value for each sample after subtracting the background signal from the negative control HEK-293 cells. The values were corrected for the relative molecular size of the message (4.8 kb) vs the probe (1.1 kb) and the quantity of sample RNA tested (20 mg). The SKOV-3 and SKBR3 cells contained erbB-2 mRNA at 0.017–0.03% total RNA, levels three- to fourfold lower than that present for b-actin mRNA. The MCF-7 cells contained two log lower levels of the erbB-2 message, 0.00015% total RNA. The b-actin mRNA levels in the three cell lines were on the same order of magnitude, and the mammary cells contained about twice as much b-actin message as the HEK-293 cells that were used as negative controls for erbB-2. As an indication of assay specificity, CHO cells stably transfected with erbB-2 (CHO-FE6) and CHO cells transiently transfected with erbB-3 and erbB-4 cDNA were tested using the erbB-2 RNA probe in the duallabeled RP assay. Figure 5 shows that only the CHO cells transfected with erbB-2 generate a signal with erbB-2 RNA probe. When erbB-3 or erbB-4 RNA probes were incorporated into the assay, the mRNA was detected in the erbB-3- and erbB-4-transfected cells, respectively.
DISCUSSION
The quantitation of specific mRNA is a valuable tool in assessing the tissue-specific expression of somatic genes in cultured cells and tissues. The traditional techniques for mRNA detection and quantitation are the Northern blotting methods which require electrophoresis of samples and are subject to operator variations and low sensitivity. The more recently developed radioisotopic RNase protection methods are more sensitive but have the disadvantage of radioactive material use and disposal. The RT-PCR amplification method (10) is clearly the most sensitive mRNA detection method available, theoretically capable of singlecopy sequence detection. However, direct measurement techniques are more reproducible for quantitation of mRNA expression levels than amplification techniques that multiply minor sample to sample variations and are dependent on the consistency of reverse transcription reactions for first-strand cDNA synthesis (11). Here, we have described a novel nonisotopic RP method that uses a RNA probe that is dual-labeled with fluorescein and biotin to directly measure mRNA using the ILA method and a silicon sensor (3, 4).
aba
AP: Anal Bio
QUANTITATION OF mRNA USING A NOVEL RNase PROTECTION ASSAY
219
TABLE 2
Quantitative Analysis of Specific mRNA in Mammary Carcinoma Cell Linesa b-Actin mRNAb Cells
% total RNA
tRNA HEK293 SKBR3 SKOV3 MCF7
0.036 0.063 0.080 0.060
{ { { {
0.0004 0.0092 0.0046 0.0018
erbB-2 mRNAc (mV/s)d
% total RNA
(mV/sec)d
fmol/mg RNA
(103) (475) (675) (965) (825)
Neg. control 0.0169 { 0.0098 0.0302 { 0.0044 0.0002 { 0.0001
(41) (79) (2391) (4800) (98)
0.0 0.105 { 0.062 0.190 { 0.078 0.001 { 0.001
a
Values are means { SD of two separate determinations made using two passages of cells. Amount of RNA tested Å 2 mg. c Amount of RNA tested Å 20 mg. d Potentiometric rate signal, average signal adjusted as described under Materials and Methods. b
b-Actin Model Assay Using b-actin as the model target sequence and a labeled RNA probe reference curve, we have seen that the sensitivity of the dual-label RNA probe assay is on the order of 30 amol (10 pg) or 0.0001% of a 10-mg sample of total RNA from cell extracts. Similar sensitivity and signal gain (Ç1 mV/s/pg) were observed for the b-actin single-stranded sense strand cDNA as target sequence, indicating that RNA–RNA and DNA– RNA hybrids appear to behave comparably in this assay. The singled-stranded cDNA may therefore be used as a convenient secondary standard for the RP assay. The level of sensitivity achieved by the RP assay is two to three logs more sensitive than that reported for the colorimetric microplate methods for labeled
DNA detection (12) and one log less sensitive than reported for the isotopic RP methods (2). Since the bactin level in our reference RNA extracts prepared from HEK-293 cells was 0.025 ({0.01)% of the total RNA, we should be able to detect specific mRNA expression at levels on the order of less than 1% of the b-actin expression level. Because this assay uses filtration to concentrate the hybridized RNA probe, sensitivity could be increased by testing larger samples, as was done for the erbB-2 assay when 20 mg of RNA was tested. Therefore, the sensitivity of the dual-label RP assay should be adequate to directly quantitate mRNA expression for many of the naturally occurring tissuespecific genes and all of the artificially transfected gene expression systems that normally result in high levels of specific mRNA. The specificity of the dual-labeled RP assay appears to be very high as evidenced by the total lack of any signal generated by the sense b-actin RNA probe when testing total cell RNA. The specificity of this assay therefore appears to be more a function of the probe sequence. The addition of a sense, dual-labeled RNA probe as a negative control for sample interference during the RNAse digestion step is a possibility. We, however, have not observed the need for such a control. Measurement of erbB-2 mRNA
FIG. 5. The specificity of the dual-label RP assay for erbB-2 in transfected CHO cells. Total RNA was prepared by TRIZOL extraction of CHO-K1 cells that were transiently transfected with erbB-3 or erbB-4. RNA was also prepared from untransfected CHO-K1 cells and CHO-FE6 cells that are stably transfected with erbB-2. The RNA preparation from each cell type was tested in duplicate using the dual-label RP assay with RNA probes for b-actin, erbB-2, erbB-3, and erbB-4. The % Total RNA values shown are calculated from three experiments testing for erbB-2, -3, and -4. The b-actin test was repeated for all three experiments and expressed as an average in this figure.
AID
AB 9800
/
6m20$$$501
10-17-96 06:53:48
The utility of the dual-label RP method was illustrated by the quantitation of the erbB-2 mRNA present in SKBR-3, MCF-7, and SKOV-3 cells that are mammary carcinoma cells reported to express the erbB-2 oncogene product, HER2 (13). ErbB-2 mRNA was easily detected in the SKBR-3 and SKOV-3 samples (2300 to 4720 mV/s above the HEK negative control), and very low levels that were close to the detection limit were measured in the MCF-7 samples (19 mV/s above the HEK negative control). The amount of erbB-2 mRNA in 1 mg of total MCF-7 RNA as determined by the RP
aba
AP: Anal Bio
220
CHAN ET AL.
method was Ç1 amol (see Table 2). This is equivalent to 20 amol per 20 mg of RNA tested in the assay. Twenty attomoles is equal to Ç6.6 pg of erbB-2 RNA probe (or 28.8 pg of erbB-2 mRNA). Because the erbB-2 RNA probe gave Ç2 mV/s signal per picogram, twice the bactin RNA probe signal per picogram, it is not unreasonable to believe that the detection limit for erbB-2 is below the 20 amol measured for the MCF-7 sample, when the detection limit was shown to be 26 amol for b-actin. Using Northern blot analysis in an earlier study, we had shown that SKBR-3 and SKOV-3 cells were positive for erbB-2 message, and MCF-7 was negative (14). The presence of erbB-2 message in MCF-7 cells was, however, detected by the exquisitely sensitive RT-PCR method in this earlier study, confirming that the dual-label RP assay had correctly identified the presence of the erbB-2 transcripts and was more sensitive than Northern blots for detecting erbB-2 message. A demonstration of the specificity of the erbB-2 duallabel RP assay was obtained by testing the RNA from CHO cells that were transfected with the homologous receptors erbB-3 (15) and erbB-4 (16). No significant signal was obtained from these cells using the erbB-2 RNA probe. Tests using specific erbB-3 and erbB-4 RNA probes did show specific detection of the respective transcripts (17), indicating the presence of these transcripts in the transfected cells. Since the values for erbB-2 mRNA were calculated from the dose–response curve for the lot of erbB-2 RNA probes used in the assay, some assay-to-assay variation was observed. A higher degree of run-to-run precision and test accuracy should be possible if a singlestranded DNA standard is used to calibrate the reference curve for each run. The dual-label RP assay is therefore a nonisotopic method with the advantages of direct quantitation of mRNA, adequate sensitivity for many applications, and a convenient protocol using probes that have an increased shelf-life over 32P-labeled probes.
AID
AB 9800
/
6m20$$$501
10-17-96 06:53:48
ACKNOWLEDGMENTS The authors thank Ma Wah and Liisa Alajoki for providing the cells for this study.
REFERENCES 1. Durnam, D. M., and Palmiter, R. D. (1983) Anal. Biochem. 131, 385–393. 2. Paul, M., Wagner, D., Metzger, R., Ganten, D., Lang, R. E., Suzuki, F., Murakami, K., Burbach, J. H. P., and Ludwig, G. (1988) J. Hypertension 6, 247–252. 3. Olson, J. D., Panfili, P. R., Armenta, R., Femmel, M., Merrick, H., Gumperz, J., Goltz, M., and Zuk, R. F. (1990) J. Immunol. Methods 134, 71–79. 4. Owicki, J. C., Bousse, L. J., Hafeman, D. G., Kirk, G. L., Olson, J. D., Wada, H. G., and Parce, J. W. (1994) Annu. Rev. Biophys. Biomol. Struct. 23, 87–113. 5. Hafeman, D. G., Parce, J. W., and McConnell, H. M. (1988) Science 240, 1182–1185. 6. Briggs, J., and Panfili, P. R. (1991) Anal. Chem. 63, 850–859. 7. Scott, G. K., Dodson, J. M., Montgomery, P. A., Johnson, R. M., Sarup, J. C., Wong, W. L., Ullrich, A., Shepard, H. M., and Benz, C. C. (1991) J. Biol. Chem. 266, 14300–14305. 8. Chomczynski, P., and Sacchi, N. (1987) Anal. Biochem. 162, 156– 159. 9. Russell, M., Kidd, S., and Kelley, M. R. (1986) Gene 45, 333– 338. 10. Wang, A. M., Doyle, M. V., and Mark, D. F. (1988) Proc. Natl. Acad. Sci. USA 86, 9717–9721. 11. Repp, R., Borkhardt, A., Gossen, R., Kreuder, J., Hammermann, J., and Lampert, F. (1995) Biotechniques 19, 84–90. 12. Rothschild, C. B., Triscott, M. X., Bowden, D. W., and Doellgast, G. (1995) Anal. Biochem. 225, 64–72. 13. Peles, E., Ben-Levy, R., Tzahar, E., Liu, N., Wen, D., and Yarden, Y. (1993) EMBO J. 12, 961–971. 14. Chan, S. D. H., Antoniucci, D. M., Fok, K. S., Alajoki, M. L., Harkins, R. N., Thompson, S. A., and Wada, H. G. (1995) J. Biol. Chem. 270, 22608–22613. 15. Plowman, G. D., Whitney, G. S., Neubauer, M. G., Green, J. M., McDonald, V. L., Todaro, G. J., and Shoyab, M. (1990) Proc. Natl. Acad. Sci. USA 87, 4905–4909. 16. Plowman, G. D., Bulouscou, J.-M., Whitney, G. S., Greem. J. M., Carlton, G. W., Foy, L., Neubauer, M. G., and Shoyab, M. (1993) Proc. Natl. Acad. Sci. USA 90, 1746–1750. 17. Blomdahl, J., Chan, S. D. H., Dill, K., and Wada, H. G. (1996) FASEB J. 10, A757 (abstract No. 4378).
aba
AP: Anal Bio
242, 214–220 (1996)
0455
Nonisotopic Quantitation of mRNA Using a Novel RNase Protection Assay: Measurement of erbB-2 mRNA in Tumor Cell Lines Samuel D. H. Chan, Kilian Dill, Jayne Blomdahl, and H. Garrett Wada1 Molecular Devices Corporation, Sunnyvale, California 94089
Received May 3, 1996
We have developed a nonisotopic RNase protection assay using RNA probes that are dual-labeled with biotin and fluorescein for detection. This system utilizes capture of the protected RNA probe hybrids to streptavidin-coated membranes attached to plastic dipsticks, complexing of anti-fluorescein–urease conjugate with the labeled RNA probe, and quantitative detection of the membrane-bound complex by a potentiometric silicon sensor. The dual-label RNase protection (RP) assay was capable of measuring b-actin mRNA in cellular RNA samples at the 27- to 45-amol level (10–17 pg) with high precision (%CV õ 7). We have used this method to quantitate the levels of erbB2 mRNA in the human tumor cell lines SKBR-3, SKOV3, and MCF-7. The levels of erbB-2 mRNA in these cells were 105, 190, and 0.9 amol per microgram of cellular RNA, respectively. The dual-label RP method should be useful for measuring the mRNA expression for other erbB-2 homologs such as erbB-3 and erbB-4 in tumor cells and tissues and can be a generally useful mRNA quantitative method for laboratories wishing to minimize radioisotope use. q 1996 Academic Press, Inc.
The detection and quantitation of specific mRNA in tissues and cell cultures play an important role in analyzing gene expression. While Northern or slot/dot blotting hybridization using radiolabeled or nonisotopic labeled nucleic acid probes is the commonly used method, solution hybridization RNase protection (RP)2 assay has proven to be more sensitive and more tolerant of 1
To whom correspondence should be addressed. Fax: 408-7473601. 2 Abbreviations used: RP, RNase protection; ILA, immunoligand assay; CHO, Chinese hamster ovary; DEPC, diethyl pyrocarbonate; LAPS, light-addressable potentiometric sensor; CV, coefficient of variation.
partially degraded RNA. The assay involves hybridization of RNA samples to radiolabeled antisense RNA probe in solution, followed by digestion of unhybridized single-stranded RNA with RNases. The protected radioactive hybrid fragments are recovered by ethanol precipitation and analyzed by denaturing polyacrylamide gel electrophoresis and autoradiography (1). Alternatively, the hybridized nucleic acids are captured onto nitrocellulose or nylon membrane, and radioactivity is then detected by liquid scintillation counting (2). One major drawback of this technique, however, is the requirement for radioactivity. Because of the increasing concerns about the health hazards and high disposal cost of radioactive waste, we have developed a nonisotopic RP assay using RNA probes dual-labeled with biotin and fluorescein, with subsequent detection of the protected probes using the immunoligand assay (ILA) system (3) on a light-addressable potentiometric sensor (LAPS) that has been used extensively for lowlevel detection of various analytes including total DNA and sequence-specific DNA (4, 5). We have used this method to quantitate the levels of the erbB-2 oncogene mRNA in human tumor cells. The erbB-2 mRNA encodes the HER2 tyrosine kinase receptor which is overexpressed in some human neoplasias (7). MATERIALS AND METHODS
Cell Culture and Transfection CHO-K1 (ATCC) and CHO-FE6 (14) cells were cultured in Ham’s F-12 containing 10% fetal bovine serum, 2 mM L-glutamine, and antibiotics (100 units/ml penicillin and 100 mg/ml streptomycin). HEK-293 (ATCC) cells were grown in DME supplemented with 10% fetal bovine serum, 2 mM L-glutamine, 100 units/ ml penicillin, and 100 mg/ml streptomycin, while MCF7 (ATCC), SKBR-3 (ATCC), and SKOV-3 (ATCC) cells were cultured in RPMI 1640 containing 10% fetal bo-
214
AID
0003-2697/96 $18.00 Copyright q 1996 by Academic Press, Inc. All rights of reproduction in any form reserved.
AB 9800
/
6m20$$$501
10-17-96 06:53:48
aba
AP: Anal Bio
QUANTITATION OF mRNA USING A NOVEL RNase PROTECTION ASSAY
vine serum, 2 mM L-glutamine, and antibiotics (100 units/ml penicillin, 100 mg/ml streptomycin). CHO cells were transiently transfected with erbB cDNAs encoded in pCEP4 expression plasmid constructs using lipofectamine as described by the manufacturer (Life Technologies). Cells were harvested for RNA extraction 48–72 h after transfection as described below. Total RNA Preparation Total RNA was purified from confluent cell cultures grown in T75 flasks by a modification of the method of Chomczynski and Sacchi (8). The culture medium was decanted and the cells were lysed by incubating with 5 ml of guanidinium thiocyanate/phenol solution (TRIZOL reagent, Life Technologies) for 5 min. One milliliter of chloroform was added to the lysate and the mixture was vortexed for 10 s and let stand at room temperature for 5 min. After centrifugation at 10,000g for 15 min at 47C, RNA in the upper aqueous phase was extracted with an equal volume of phenol (pH 4.0)/ chloroform. RNA was then precipitated with 1 vol of isopropanol at 0207C for 1 h and pelleted by sedimentation at 10,000g for 15 min. The resulting RNA was dissolved in 20 ml DEPC-treated water and stored at 0207C. The concentration of RNA was determined by absorbance at 260 nm (1 A260 unit Å 40 mg/ml). Templates for RNA Probe Synthesis The plasmid pBS-ActB containing 1.1 kb of human b-actin cDNA in pBluescript SK0 was obtained from ATCC (HHC189). The 1.1-kb insert was excised and subcloned in the opposite orientation into pBluescript II SK0 (Stratagene) to generate the plasmid pBSActBT3. pBS-h-erbB-2 was constructed by cloning a 4.0-kb cDNA fragment containing the complete open reading frame of the human erbB-2 gene into pBluescript II SK0. pBS-ActB and pBS-h-erbB-2 were digested, respectively, with HindIII and BamHI, to generate templates for the synthesis of antisense RNA probes. In both cases, a 1.1-kb probe was prepared using T3 RNA polymerase. The sense b-actin probe was prepared using a HindIII-digested pBS-ActBT3 plasmid and T7 RNA polymerase. The single-stranded sense b-actin cDNA for a b-actin calibration standard was prepared using superinfection with the helper phage R408 (9) of an overnight culture containing the pBS-ActBT3 plasmid in host bacteria XL-1 Blue (Stratagene). ErbB-3 and erbB-4 RNA probes approximately 1 kb in size were prepared as described for erbB-2 from RT-PCR cloned sequences containing the entire open reading frame for these two genes. Synthesis of Dual-Labeled RNA Probes for b-Actin and erbB-2 mRNA Detection Dual-labeled RNA probe was synthesized in a 20-ml reaction containing 40 mM Tris–HCl (pH 8.0), 8 mM
AID
AB 9800
/
6m20$$$501
10-17-96 06:53:48
215
MgCl2 , 2 mM spermidine, 25 mM NaCl, 10 mM DTT, 0.5 mM ATP, 0.5 mM GTP, 0.25 mM CTP, 0.25 mM biotin-14–CTP (Life Technologies), 0.325 mM UTP, 0.175 mM fluorescein-12–UTP (Boehringer-Mannheim), 1 mg of template DNA, 40 units of RNasin (Promega), and 50 units of T3 RNA polymerase (Life Technologies). The reaction mixture was incubated at 377C for 1 h, after which the reaction was stopped by heating to 657C for 10 min. The DNA template was then digested by incubating with 14 units of RNase-free DNAse I (Boehringer-Mannheim) at 377C for 15 min. The RNA probes were purified by centrifugation through DEPC-treated G-100 or G-200 spin columns (Clontech). The concentration of the probes was determined by absorbance at 260 nm, then 40 units of RNasin was added, and the probes were stored frozen at 0207C. Characterization of the RNA Probes The RNA probes were characterized with respect to their size, biotin and fluorescein content, and reactivity in the ILA. Molecular size of the RNA probes was determined using electrophoresis in 6% polyacrylamide, Tris-borate-EDTA pH 8.5 gels containing 6 M urea. Samples of 10 to 100 ng of RNA probe were electrophoresed in parallel with low-molecular-weight (1000–50 bases) DNA standards (BioVentures) after heating to 857C for 5 min to denature the samples in 40% formamide sample buffer. The gels were fixed in 50% methanol/10% acetic acid and nucleic acid bands were detected by silver staining (LabLogix Inc.). The fluorescein content of the RNA probe was determined by fluorimetry using a fluorescein reference curve, and the biotin content was determined using dot blot analysis. In brief, the dot blots containing serially diluted probe and standard (5*-biotinylated 20-mer oligonucleotide, Operon Technologies) were incubated with streptavidin–alkaline phosphatase conjugate followed by detection with a chemiluminescent dioxetane substrate (DNA/RNA Detection Kit, LabLogix Inc.) and X-ray film. The dot blots of probe dilutions were interpolated on the biotinylated oligonucleotide reference curve as a semiquantitative measure of their biotin content. Detection of the Dual-Labeled RNA Probe Using the ILA System The reactivity of the dual-labeled probe in the Threshold ILA System (Molecular Devices Corp.) was determined by titrating the probe using the LAPSbased enzyme detection system (4). The dual-labeled probe was diluted into 100 ml of assay buffer (phosphate-buffered saline containing BSA and nonionic detergent) and mixed with 100 or 200 ml of capture reagent (assay buffer containing 2 or 4 mg of streptavidin). The samples were then filtered onto a prewashed
aba
AP: Anal Bio
216
CHAN ET AL.
biotinylated nitrocellulose membrane (mounted on a plastic dipstick) at low vacuum setting, whereby the streptavidin–mRNA–RNA probe complexes were captured and bound to the membrane. The urease conjugate of a rabbit anti-fluorescein antibody (Ç1.5 mg) in 200 ml of assay buffer was filtered through the membrane at low vacuum setting followed by 0.5 ml of wash buffer (phosphate-buffered saline containing nonionic detergent) at high vacuum setting. The quantity of bound enzyme conjugate is directly proportional to the amount of hybridized RNA probe complex bound to the membrane. The enzyme activity is measured in the presence of urea substrate by a potentiometric sensor and expressed in mV/s which is equivalent to 1 millipH unit/min (5). The Threshold ILA System provides for internal calibration of each assay by comparing a negative and positive control to a reference curve that can be generated by serial dilution of the dual-labeled probe (6). Dual-Label RNase Protection Assay Protocol The dual-labeled RNA probes were hybridized with specific mRNA present in total RNA prepared from cell extractions. Typically, 10–20 mg of cellular RNA (or varying amounts if preparing a reference curve) was mixed with yeast tRNA to give 20 mg of total RNA and 3–10 ng of RNA probe in 30 ml of a hybridization buffer containing 80% formamide, 40 mM Mes buffer (pH 6.5), 400 mM NaCl, and 1 mM EDTA. The mixture was heated to 857C for 10 min and then incubated at 427C overnight. In some experiments hybridization time and temperature were varied, and alternatively 557C for 4 h could be used. As a negative control, the RNA probe was mixed with 20 mg of yeast tRNA and the above protocol was followed. The next day, the samples were mixed with 350 ml of the RNase digestion buffer composed of 10 mM Tris–HCl (pH 7.6), 300 mM NaCl, 5 mM EDTA, 125 unit/ml RNase T1, and 6.25 mg/ml RNase A (Ambion). Samples were digested at 377C for 1 h. Following this, the samples were extracted with 400 ml of phenol/chloroform and centrifuged. The aqueous layer was removed and 20 mg of yeast tRNA, 100 ml of 7.5 M ammonium acetate, and 500 ml of isopropanol were added. The mixture was thoroughly mixed and placed at 0207C for 1 h. The samples were then centrifuged at 47C for 15 min at 10,000g. The supernatant was carefully removed and the pellet was allowed to air dry for a few minutes. The samples were then redissolved in 20 ml of TE buffer (10 mM Tris–HCl, 1 mM EDTA, pH 8.0) for ILA detection of hybridized RNA probe as described above. RESULTS
Characterization of Dual-Labeled RNA Probes for b-Actin The RNA probes for b-actin and erbB-2 were designed to be complementary to a 1100-base stretch of
AID
AB 9800
/
6m20$$$501
10-17-96 06:53:48
FIG. 1. Polyacrylamide gel electrophoretic analysis of dual-labeled RNA probes. Dual-labeled RNA probes for b-actin and erbB-2 were analyzed by electrophoresis in 6% acrylamide gels containing 6 M urea as described under Materials and Methods. (A) 25 ng of erbB2 RNA probe was applied in lane 1 and 25 ng of b-actin RNA probe in lane 2. The electrophoresed samples were silver stained to detect the RNA bands. (B) erbB-2 and b-actin RNA probes were digested with RNase A/RNase-T1 prior to electrophoretic analysis as described for A. 25 ng of digested erbB-2 RNA probe was applied in lane 1 and 25 ng of digested b-actin RNA probe in lane 2.
the 3* terminal coding sequence for the b-actin and erbB-2 mRNA. When analyzed by electrophoresis in denaturing (6 M urea) polyacrylamide gels, a major band was observed at 1.1 kb and small amounts of nucleic acid with progressively smaller size were observed (Fig. 1). The RNA composition of these silverstained bands was confirmed by their lability to RNase treatment. The fluorescein content of the two RNA probes was determined by fluorimetry and indicated that, under the conditions employed for synthesis, approximately 11 fluorescein–UTPs were incorporated per 1.1 kb of RNA probe. The biotin content of the RNA probes was also determined by dot blot to be Ç10 biotin–CTPs incorporated per 1.1 kb of b-actin RNA probe and Ç22 biotins per 1.1 kb of erbB-2 RNA probe. The reactivity of the RNA probe with the aforementioned dual-label composition was characterized by titration in the ILA system. By virtue of their duallabeled nature, the RNA probes cause anti-fluoresceinconjugated urease to bind to biotinylated membrane via streptavidin/biotinylated RNA probe complexes. The membrane-bound urease is then detected by the pH-sensitive, potentiometric silicon sensor in the Threshold instrument. Figure 2 shows the dose response curve obtained by testing dilutions of the bactin and erbB-2 RNA probes in the Threshold instrument. The signal unit is mV/s which is equivalent to the milli-pH unit/min rate of pH change. The slope of the dose–response curves indicates that 1 fmol (10015 mol) of b-actin probe (363 pg) generates a signal of 461 mV/s, approximately 1 mV/s of signal per picogram of
aba
AP: Anal Bio
217
QUANTITATION OF mRNA USING A NOVEL RNase PROTECTION ASSAY
FIG. 2. Dose response of b-actin and erbB-2 dual-labeled RNA probes using the ILA method. The RNA probes for b-actin and erbB2 dual-labeled with biotin and fluorescein were serially diluted in ILA assay buffer and detected in the ILA assay protocol as described under Materials and Methods. The slopes of the response curves were calculated using a least-squares regression equation. The erbB2 curve is plotted with a solid line (slope Å 1797 mV/s/ng) and the b-actin curve is plotted with a dashed line (slope Å 999 mV/s/ng).
probe, and the erbB-2 probe generates approximately 1.8 mV/s of signal per picogram of probe. Hybridization Kinetics and Assay Performance Using the b-Actin RNA Probe When 10 mg of RNA from HEK-293 human kidney cells was hybridized with anti-sense b-actin RNA probe at 427C for 24 h, RNase protection was observed which corresponded to 2.8 fmol of b-actin RNA probe and 0.025% of the original cellular RNA tested (Fig. 3A). The same hybridization experiment using dual-labeled sense b-actin RNA probe showed no protection from RNase when hybridized to HEK-293 RNA, indicating that the RNase protection is sequence specific (Fig. 3A). The kinetics of hybridization were examined at 427C and indicated approach to equilibrium in about 4–8 h as shown in Fig. 3B. The hybridization kinetics were also similar for erbB-2 RNA probe hybridizing to target mRNA extracted from CHO-FE6 cells that are stably transfected with erbB-2 (14). These results indicated that the hybridization time could be shortened to 8 h without loss of sensitivity and that the hybridization kinetics do not appear to be sequence specific. The effect of temperature on hybridization was also investigated. By elevating the temperature from 42 to 557C a large increase in signal was observed after a 2-h hybridization (data not shown). This suggested that elevating the temperature may further shorten the hybridization time. Assay precision for b-actin RP assay was determined by testing 0, 2.5, 5, and 10 mg of HEK293 RNA in quadruplicate using the 427C overnight hybridization (Table 1). The precision was excellent, ranging from 1.4 to 6.9% CV. The lower limit of detection calculated from the 21 standard deviation range
AID
AB 9800
/
6m20$$$501
10-17-96 06:53:48
FIG. 3. Protection of hybridized dual-labeled RNA probes. Total RNA prepared from HEK-293 cells was tested in the dual-labeled RP assay for b-actin mRNA. The hybridization was conducted at 427C unless otherwise specified. (A) The three levels of total RNA were hybridized overnight in quadruplicate along with a negative control (0, 2.5, 5.0, and 10 mg total RNA) using anti-sense b-actin RNA probe (solid circles). The hybridized samples were RNAse digested and processed as described under Materials and Methods. The error bars indicate the mean value {1 SD. The sense b-actin RNA probe (open circles) was also used to test these same samples in duplicate. (B) Total cellular RNA was hybridized in duplicate for 1, 2, 4, 8, and 24 h with dual-labeled RNA probe, after which the samples were stored at 0207C. The samples were all tested the next day in the dual-labeled RP assay. Ten-microgram samples of RNA from HEK293 cells were tested using b-actin RNA probe (open circles), and 20-mg samples of RNA from CHO-FE6 cells were tested using erbB-2 RNA probe (solid circles). The assay signal was background subtracted and adjusted to a percentage of the maximum signal obtained during the time course and plotted versus time. The assay background for the b-actin assay was 196 mV/s and for the erbB-2 assay 80 mV/s.
of the negative control and the 1 mV/s signal gain per picogram of RNA probe (1.1 kb) was 26 amol (10018 mol or 9.5 pg) of b-actin (1.75 kb) which represents 0.0001% of a 10-mg sample of cellular RNA. TABLE 1
Precision of the b-Actin mRNA RPA Measurement in HEK-293 Cellsa Cell RNA (mg)
mRNA ng
% of total RNA
fmol/mg
% CV
0 2.5 5.0 10.0
0.400 0.781 1.506
0.025 0.025 0.024
0.27 0.27 0.26
2.6 3.4 6.9 1.4
a
Mean values from four replicates (N Å 4) from data in Fig. 3A. % CV is coefficient of variation of the signal in mV/s.
aba
AP: Anal Bio
218
CHAN ET AL.
FIG. 4. Dose response of b-actin single-stranded cDNA tested in dual-label RNase protection assay. Single-stranded b-actin cDNA prepared as described under Materials and Methods was serially diluted, and duplicate samples were tested in the dual-labeled RP assay protocol using ILA detection. The slope of the b-actin cDNA response curve was calculated to be 1352 mV/s/ng (788 mV/s/fmol) using a least-squares regression equation. The error bars represent the {11 SD range about the mean.
The sensitivity of the assay for b-actin was demonstrated using a single-stranded sense target cDNA produced in the pBluescript phagemid system (Stratagene). Figure 4 shows a dose–response curve using the b-actin sense cDNA that gave 1.1 mV/s signal per picogram of target cDNA that is equivalent to 788 mV/ s signal per femtomole of target. The high precision at the low end of the dose–response curve made the calculated lower limit of detection 27 amol (10 pg) of cDNA at the 99% confidence interval (31 standard deviation above the mean negative control value). Spike recovery experiments were run using HEK-293 total RNA. This entailed testing 5 mg of the RNA with and without the addition of 380 and 760 pg of b-actin singlestranded sense cDNA standard. The spiked standard DNA gave a recovery of signal equivalent to 90% of the theoretical signal, indicating that the effective sensitivity in cell samples is approximately 30 amol of target sequence (data not shown). This value was similar to the detection limit calculated from the HEK-293 total RNA dose–response curve for b-actin mRNA and the mV/s signal per picogram of RNA probe value as described above. Assay for erbB-2 mRNA Transcripts in Mammary Carcinoma Cell Line Extracts Several tumor cell lines of mammary carcinoma origin (SKBR-3, SKOV-3, and MCF-7) were tested for the production of erbB-2 mRNA using the dual-labeled RNA probe RP assay. The total RNA samples prepared from the cell lines were also tested as described above for b-actin mRNA as an internal control for RNA degra-
AID
AB 9800
/
6m20$$$501
10-17-96 06:53:48
dation. The results from the erbB-2 and b-actin assays are shown in Table 2, which presents the erbB-2 mRNA level in mV/s, fmol/mg RNA, and percentage of total RNA. The amount of erbB-2 mRNA in samples was determined from the erbB-2 RNA probe reference curve using a linear regression analysis to calculate the value for each sample after subtracting the background signal from the negative control HEK-293 cells. The values were corrected for the relative molecular size of the message (4.8 kb) vs the probe (1.1 kb) and the quantity of sample RNA tested (20 mg). The SKOV-3 and SKBR3 cells contained erbB-2 mRNA at 0.017–0.03% total RNA, levels three- to fourfold lower than that present for b-actin mRNA. The MCF-7 cells contained two log lower levels of the erbB-2 message, 0.00015% total RNA. The b-actin mRNA levels in the three cell lines were on the same order of magnitude, and the mammary cells contained about twice as much b-actin message as the HEK-293 cells that were used as negative controls for erbB-2. As an indication of assay specificity, CHO cells stably transfected with erbB-2 (CHO-FE6) and CHO cells transiently transfected with erbB-3 and erbB-4 cDNA were tested using the erbB-2 RNA probe in the duallabeled RP assay. Figure 5 shows that only the CHO cells transfected with erbB-2 generate a signal with erbB-2 RNA probe. When erbB-3 or erbB-4 RNA probes were incorporated into the assay, the mRNA was detected in the erbB-3- and erbB-4-transfected cells, respectively.
DISCUSSION
The quantitation of specific mRNA is a valuable tool in assessing the tissue-specific expression of somatic genes in cultured cells and tissues. The traditional techniques for mRNA detection and quantitation are the Northern blotting methods which require electrophoresis of samples and are subject to operator variations and low sensitivity. The more recently developed radioisotopic RNase protection methods are more sensitive but have the disadvantage of radioactive material use and disposal. The RT-PCR amplification method (10) is clearly the most sensitive mRNA detection method available, theoretically capable of singlecopy sequence detection. However, direct measurement techniques are more reproducible for quantitation of mRNA expression levels than amplification techniques that multiply minor sample to sample variations and are dependent on the consistency of reverse transcription reactions for first-strand cDNA synthesis (11). Here, we have described a novel nonisotopic RP method that uses a RNA probe that is dual-labeled with fluorescein and biotin to directly measure mRNA using the ILA method and a silicon sensor (3, 4).
aba
AP: Anal Bio
QUANTITATION OF mRNA USING A NOVEL RNase PROTECTION ASSAY
219
TABLE 2
Quantitative Analysis of Specific mRNA in Mammary Carcinoma Cell Linesa b-Actin mRNAb Cells
% total RNA
tRNA HEK293 SKBR3 SKOV3 MCF7
0.036 0.063 0.080 0.060
{ { { {
0.0004 0.0092 0.0046 0.0018
erbB-2 mRNAc (mV/s)d
% total RNA
(mV/sec)d
fmol/mg RNA
(103) (475) (675) (965) (825)
Neg. control 0.0169 { 0.0098 0.0302 { 0.0044 0.0002 { 0.0001
(41) (79) (2391) (4800) (98)
0.0 0.105 { 0.062 0.190 { 0.078 0.001 { 0.001
a
Values are means { SD of two separate determinations made using two passages of cells. Amount of RNA tested Å 2 mg. c Amount of RNA tested Å 20 mg. d Potentiometric rate signal, average signal adjusted as described under Materials and Methods. b
b-Actin Model Assay Using b-actin as the model target sequence and a labeled RNA probe reference curve, we have seen that the sensitivity of the dual-label RNA probe assay is on the order of 30 amol (10 pg) or 0.0001% of a 10-mg sample of total RNA from cell extracts. Similar sensitivity and signal gain (Ç1 mV/s/pg) were observed for the b-actin single-stranded sense strand cDNA as target sequence, indicating that RNA–RNA and DNA– RNA hybrids appear to behave comparably in this assay. The singled-stranded cDNA may therefore be used as a convenient secondary standard for the RP assay. The level of sensitivity achieved by the RP assay is two to three logs more sensitive than that reported for the colorimetric microplate methods for labeled
DNA detection (12) and one log less sensitive than reported for the isotopic RP methods (2). Since the bactin level in our reference RNA extracts prepared from HEK-293 cells was 0.025 ({0.01)% of the total RNA, we should be able to detect specific mRNA expression at levels on the order of less than 1% of the b-actin expression level. Because this assay uses filtration to concentrate the hybridized RNA probe, sensitivity could be increased by testing larger samples, as was done for the erbB-2 assay when 20 mg of RNA was tested. Therefore, the sensitivity of the dual-label RP assay should be adequate to directly quantitate mRNA expression for many of the naturally occurring tissuespecific genes and all of the artificially transfected gene expression systems that normally result in high levels of specific mRNA. The specificity of the dual-labeled RP assay appears to be very high as evidenced by the total lack of any signal generated by the sense b-actin RNA probe when testing total cell RNA. The specificity of this assay therefore appears to be more a function of the probe sequence. The addition of a sense, dual-labeled RNA probe as a negative control for sample interference during the RNAse digestion step is a possibility. We, however, have not observed the need for such a control. Measurement of erbB-2 mRNA
FIG. 5. The specificity of the dual-label RP assay for erbB-2 in transfected CHO cells. Total RNA was prepared by TRIZOL extraction of CHO-K1 cells that were transiently transfected with erbB-3 or erbB-4. RNA was also prepared from untransfected CHO-K1 cells and CHO-FE6 cells that are stably transfected with erbB-2. The RNA preparation from each cell type was tested in duplicate using the dual-label RP assay with RNA probes for b-actin, erbB-2, erbB-3, and erbB-4. The % Total RNA values shown are calculated from three experiments testing for erbB-2, -3, and -4. The b-actin test was repeated for all three experiments and expressed as an average in this figure.
AID
AB 9800
/
6m20$$$501
10-17-96 06:53:48
The utility of the dual-label RP method was illustrated by the quantitation of the erbB-2 mRNA present in SKBR-3, MCF-7, and SKOV-3 cells that are mammary carcinoma cells reported to express the erbB-2 oncogene product, HER2 (13). ErbB-2 mRNA was easily detected in the SKBR-3 and SKOV-3 samples (2300 to 4720 mV/s above the HEK negative control), and very low levels that were close to the detection limit were measured in the MCF-7 samples (19 mV/s above the HEK negative control). The amount of erbB-2 mRNA in 1 mg of total MCF-7 RNA as determined by the RP
aba
AP: Anal Bio
220
CHAN ET AL.
method was Ç1 amol (see Table 2). This is equivalent to 20 amol per 20 mg of RNA tested in the assay. Twenty attomoles is equal to Ç6.6 pg of erbB-2 RNA probe (or 28.8 pg of erbB-2 mRNA). Because the erbB-2 RNA probe gave Ç2 mV/s signal per picogram, twice the bactin RNA probe signal per picogram, it is not unreasonable to believe that the detection limit for erbB-2 is below the 20 amol measured for the MCF-7 sample, when the detection limit was shown to be 26 amol for b-actin. Using Northern blot analysis in an earlier study, we had shown that SKBR-3 and SKOV-3 cells were positive for erbB-2 message, and MCF-7 was negative (14). The presence of erbB-2 message in MCF-7 cells was, however, detected by the exquisitely sensitive RT-PCR method in this earlier study, confirming that the dual-label RP assay had correctly identified the presence of the erbB-2 transcripts and was more sensitive than Northern blots for detecting erbB-2 message. A demonstration of the specificity of the erbB-2 duallabel RP assay was obtained by testing the RNA from CHO cells that were transfected with the homologous receptors erbB-3 (15) and erbB-4 (16). No significant signal was obtained from these cells using the erbB-2 RNA probe. Tests using specific erbB-3 and erbB-4 RNA probes did show specific detection of the respective transcripts (17), indicating the presence of these transcripts in the transfected cells. Since the values for erbB-2 mRNA were calculated from the dose–response curve for the lot of erbB-2 RNA probes used in the assay, some assay-to-assay variation was observed. A higher degree of run-to-run precision and test accuracy should be possible if a singlestranded DNA standard is used to calibrate the reference curve for each run. The dual-label RP assay is therefore a nonisotopic method with the advantages of direct quantitation of mRNA, adequate sensitivity for many applications, and a convenient protocol using probes that have an increased shelf-life over 32P-labeled probes.
AID
AB 9800
/
6m20$$$501
10-17-96 06:53:48
ACKNOWLEDGMENTS The authors thank Ma Wah and Liisa Alajoki for providing the cells for this study.
REFERENCES 1. Durnam, D. M., and Palmiter, R. D. (1983) Anal. Biochem. 131, 385–393. 2. Paul, M., Wagner, D., Metzger, R., Ganten, D., Lang, R. E., Suzuki, F., Murakami, K., Burbach, J. H. P., and Ludwig, G. (1988) J. Hypertension 6, 247–252. 3. Olson, J. D., Panfili, P. R., Armenta, R., Femmel, M., Merrick, H., Gumperz, J., Goltz, M., and Zuk, R. F. (1990) J. Immunol. Methods 134, 71–79. 4. Owicki, J. C., Bousse, L. J., Hafeman, D. G., Kirk, G. L., Olson, J. D., Wada, H. G., and Parce, J. W. (1994) Annu. Rev. Biophys. Biomol. Struct. 23, 87–113. 5. Hafeman, D. G., Parce, J. W., and McConnell, H. M. (1988) Science 240, 1182–1185. 6. Briggs, J., and Panfili, P. R. (1991) Anal. Chem. 63, 850–859. 7. Scott, G. K., Dodson, J. M., Montgomery, P. A., Johnson, R. M., Sarup, J. C., Wong, W. L., Ullrich, A., Shepard, H. M., and Benz, C. C. (1991) J. Biol. Chem. 266, 14300–14305. 8. Chomczynski, P., and Sacchi, N. (1987) Anal. Biochem. 162, 156– 159. 9. Russell, M., Kidd, S., and Kelley, M. R. (1986) Gene 45, 333– 338. 10. Wang, A. M., Doyle, M. V., and Mark, D. F. (1988) Proc. Natl. Acad. Sci. USA 86, 9717–9721. 11. Repp, R., Borkhardt, A., Gossen, R., Kreuder, J., Hammermann, J., and Lampert, F. (1995) Biotechniques 19, 84–90. 12. Rothschild, C. B., Triscott, M. X., Bowden, D. W., and Doellgast, G. (1995) Anal. Biochem. 225, 64–72. 13. Peles, E., Ben-Levy, R., Tzahar, E., Liu, N., Wen, D., and Yarden, Y. (1993) EMBO J. 12, 961–971. 14. Chan, S. D. H., Antoniucci, D. M., Fok, K. S., Alajoki, M. L., Harkins, R. N., Thompson, S. A., and Wada, H. G. (1995) J. Biol. Chem. 270, 22608–22613. 15. Plowman, G. D., Whitney, G. S., Neubauer, M. G., Green, J. M., McDonald, V. L., Todaro, G. J., and Shoyab, M. (1990) Proc. Natl. Acad. Sci. USA 87, 4905–4909. 16. Plowman, G. D., Bulouscou, J.-M., Whitney, G. S., Greem. J. M., Carlton, G. W., Foy, L., Neubauer, M. G., and Shoyab, M. (1993) Proc. Natl. Acad. Sci. USA 90, 1746–1750. 17. Blomdahl, J., Chan, S. D. H., Dill, K., and Wada, H. G. (1996) FASEB J. 10, A757 (abstract No. 4378).
aba
AP: Anal Bio