Quantitation of Ribonucleic Acids Using 18O Labeling and Mass

Pseudouridine: Still mysterious, but never a fake (uridine)!. Felix Spenkuch , Yuri Motorin , Mark Helm. RNA Biology 2014 11, 1540-1554 ...
0 downloads 0 Views 87KB Size
Anal. Chem. 2005, 77, 1891-1895

Quantitation of Ribonucleic Acids Using Labeling and Mass Spectrometry

18

O

Zhaojing Meng and Patrick A. Limbach*

Rieveschl Laboratories for Mass Spectrometry, Department of Chemistry, P.O. Box 210172, University of Cincinnati, Cincinnati, Ohio 45221

A previous limitation in the analysis of ribonucleic acids (RNAs) by mass spectrometry (MS) has been the inability to obtain quantitative information relating to total RNA, RNA subunits, and undermodified nucleosides in a straightforward manner. Here, a simple and rapid method has been developed for the relative quantitation of small RNAs using 18O labeling and matrix-assisted laser desorption/ionization mass spectrometry (MALDI-MS). One RNA sample is digested with RNase T1 in 18O-labeled (“heavy”) water with the 18O being incorporated at the 3′phosphate end of oligonucleotides upon hydrolysis. A second RNA sample is digested with RNase T1 in normal (“light”) water. The two samples are then combined and analyzed by MALDI-MS. Relative ion abundances of the light- and heavy-water digestion products, which are separated by 2 Da due to the isotopic mass of 18O, reveal relative quantitation information from the two RNA samples. The accuracy and reproducibility of this approach were tested on 18 known RNA samples and 4 unknown RNA samples. The coefficients of variation for quantitation were found to be generally below 15% when using MALDI-MS. The approach yields accurate quantitative information for heavy-to-light ratios greater than 1:2. This method should prove useful for quantitatively characterizing variations in RNA production and variations in the amount of posttranscriptionally modified nucleosides. The identification and characterization of ribonucleic acids (RNAs) and their posttranscriptional modifications is essential for fully understanding their structural and functional roles.1-3 RNA production as well as the level of posttranscriptional modification are known to be affected by a number of factors.3-14 For example, * To whom correspondence should be addressed. Phone: (513) 556-1871. Fax: (513) 556-9239. E-mail: [email protected]. (1) Agris, P. F. Nucleic Acids Res. 2004, 32, 223-238. (2) Doudna, J. A. Nat. Struct. Biol. 2000, 7, 954-956. (3) Bjork, G. R.; Jacobsson, K.; Nilsson, K.; Johansson, M. J. O.; Bystrom, A. S.; Persson, O. P. EMBO J. 2001, 20, 231-239. (4) Kowalak, J. A.; Dalluge, J. J.; McCloskey, J. A.; Stetter, K. O. Biochemistry 1994, 33, 7869-7876. (5) Jayabaskaran, C.; Hande, S. Plant Growth Regul. 1995, 16, 73-81. (6) Dong, H.; Nilsson, L.; Kurland, C. G. J. Mol. Biol. 1996, 260, 649-663. (7) Kanduc, D. Biochem. Biophys. Res. Commun. 1996, 221. (8) Berg, O. G.; Kurland, C. G. J. Mol. Biol. 1997, 270, 544-550. (9) Helm, M.; Florentz, C.; Chomyn, A.; Attardi, G. Nucleic Acids Res. 1999, 27, 756-763. (10) Madore, E.; Florentz, C.; Giege, R.; Sekine, S.-i.; Yokoyama, S.; Lapointe, J. Eur. J. Biochem. 1999, 266, 1128-1135. 10.1021/ac048801y CCC: $30.25 Published on Web 02/11/2005

© 2005 American Chemical Society

quantitative differences in isoaccepting tRNA levels and posttranscriptional modifications to tRNAs have been found between chloroplasts and etioplasts of cucumber cotyledons, and these differences are thought to result from the light-induced transformation from etioplast to chloroplast.5 Other researchers have noted that point mutations to human mitochondrial tRNA genes lead to differences in the posttranscriptional modification of mitochondrial isoaccepting tRNAs.9,11 Because these point mutations are known to be associated with various neuromuscular disorders,15 it is believed that differential posttranscriptional modification of these mitochondrial tRNAs have significant functional consequences. Most current methods for RNA quantitation target total RNA or mRNA and generally involve reverse transcription to DNA, which is then analyzed.16 A limitation of such methods is that after reverse transcription, qualitative and quantitative information regarding any posttranscriptional modifications is lost. Mass spectrometry (MS) offers a number of advantages for the characterization of nucleic acids arising from its ability to provide mass and sequence information.17,18 To date, quantitative information relating to RNA or posttranscriptionally modified nucleosides obtained via MS-based approaches relies upon hydrolysis of RNA into constituent nucleosides that can then be analyzed by GC/ MS or LC/MS.19-21 No MS-based method has been developed yet that would allow one to obtain quantitative information from RNA at the oligonucleotide or higher level. Stable isotope labeling has been used successfully in proteomics for determining quantitative information related to protein (11) Yasukawa, T.; Suzuki, T.; Suzuki, T.; Ueda, T.; Ohta, S.; Watanabe, K. J. Biol. Chem. 2000, 275, 4251-4257. (12) Noon, K. R.; Guymon, R.; Crain, P. F.; McCloskey, J. A.; Thomm, M.; Lim, J.; Cavicchioli, R. J. Bacteriol. 2003, 185, 5483-5490. (13) Wittenhagen, L. M.; Kelley, S. O. Trends Biochem. Sci. 2003, 28, 605-611. (14) Leipuviene, R.; Qian, Q.; Bjork, G. R. J. Bacteriol. 2004, 186, 758-766. (15) Schon, E. A.; Bonilla, E.; DiMauro, S. J. Bioenerg. Biomembr. 1997, 29, 131-149. (16) Ding, C.; Cantor, C. R. J. Biochem. Mol. Biol. 2004, 37, 1-10. (17) Limbach, P. A. Mass Spectrom. Rev. 1996, 15, 297-336. (18) Nordhoff, E.; Kirpekar, F.; Roepstorff, P. Mass Spectrom. Rev. 1996, 15, 67-138. (19) Richardson, F. C.; Zhang, C.; Lehrman, S. R.; Koc, H.; Swenberg, J. A.; Richardson, K. A.; Bendele, R. A. Chem. Res. Toxicol. 2002, 15, 922-926. (20) Dalluge, J. J.; Hashizume, T.; McCloskey, J. A. Nucleic Acids Res. 1996, 24, 3242-3245. (21) Peters, G. J.; Noordhuis, P.; Komissarov, A.; Holwerda, U.; Kok, M.; Van Laar, J. A.; Van der Wilt, C. L.; Van Groeningen, V.; Pinedo, H. M. Anal. Biochem. 1995, 231, 157-163.

Analytical Chemistry, Vol. 77, No. 6, March 15, 2005 1891

production.22-24 Among the various stable isotope labeling approaches, enzyme-catalyzed 18O-labeling is a simple and effective method which has been successfully applied for relative quantification of proteins, comparative proteomics, protein sequencing, and C-terminal identification, as well as absolute quantitation of proteins and peptides.25-30 18O labels are incorporated into the carboxy terminus during protease digestions in the presence of 18O-labeled water. One or two 18O labels can be incorporated, depending on the protease used and experimental conditions, which subsequently result in a 2- or 4-Da mass shift, as compared to digestion products in the presence of normal water.31 Even though the approach has potential problems, such as possible complex data processing due to incomplete label incorporation and back exchange of the isotopic label, it remains an attractive and universal approach for quantitative proteomics because the label is introduced without perturbing the biological system. Recently, we demonstrated that the combination of 18O labeling and mass spectrometry is an effective approach for RNase mapping of intact RNAs and for localizing posttranscriptional modifications by MS and MS/MS.32,33 An 18O label can be incorporated into the final endonuclease digestion product if the enzymatic hydrolysis of RNAs takes place in the presence of 18Olabeled water. In a manner similar to that already demonstrated for proteomics, here we demonstrate the applicability of 18O labeling and matrix-assisted laser desorption/ionization time-offlight mass spectrometry (MALDI-TOF MS) for the relative quantitation of transfer RNAs (tRNAs). EXPERIMENTAL SECTION Materials. Escherichia coli tRNAVal, diammonium hydrogen citrate (DAHC), and 2,4,6-trihydroxyacetophenone (THAP) were purchased from Sigma-Aldrich (St. Louis, MO) and were used as received. RNase T1 was obtained from Roche Molecular Biochemicals (Indianapolis, IN). The 18O-labeled water (99.4% purity) was purchased from Isotec (Miamisburg, OH). Sep-Pak C18 cartridges were obtained from Waters (Milford, MA). Nanopure water (18 MΩ) was filtered using a Barnstead (Dubuque, IA) Nanopure System and autoclaved before use. Enzyme Purification. RNase T1 was precipitated from its original suspension by use of acetone. The precipitate was resuspended and then eluted in 1 mL of 75% aqueous acetonitrile from a Sep-Pak C18 cartridge. The eluent was divided into two equal volumes, dried, and reconstituted in 16O or 18O water, (22) Gygi, S. P.; Rist, B.; Gerber, S. A.; Turecek, F.; Gelb, M. H.; Aebersold, R. Nat. Biotechnol. 1999, 17, 994-999. (23) Ong, S.-E.; Foster, L. J.; Mann, M. Methods 2003, 29, 124-130. (24) Tao, W. A.; Aebersold, R. Curr. Opin. Biotechnol. 2003, 14, 110-118. (25) Shevchenko, A.; Chernushevich, I.; Ens, W.; Standing, K. G.; Thomson, B.; Wilm, M.; Mann, M. Rapid Commun. Mass Spectrom. 1997, 11, 10151024. (26) Yao, X.; Freas, A.; Ramirez, J.; Demirev, P. A.; Fenselau, C. Anal. Chem. 2001, 73, 2836-2842. (27) Mirgorodskaya, O. A.; Kozmin, Y. P.; Titov, M. I.; Korner, R.; Sonksen, C. P.; Roepstorff, P. Rapid Commun. Mass Spectrom. 2000, 14, 1226-1232. (28) Kosaka, T.; Takazawa, T.; Nakamura, T. Anal. Chem. 2000, 72, 1179-1185. (29) Wang, Y. K.; Ma, Z.; Quinn, D. F.; Fu, E. W. Anal. Chem. 2001, 73, 37423750. (30) Stewart, I. I.; Thomson, T.; Figeys, D. Rapid Commun. Mass Spectrom. 2001, 15, 2456-2465. (31) Schnolzer, M.; Jedrzejewski, P.; Lehmann, W. D. Electrophoresis 1996, 17, 945-953. (32) Meng, Z.; Limbach, P. A. Int. J. Mass Spectrom. 2004, 234, 37-44. (33) Berhane, B. T.; Limbach, P. A. J. Mass Spectrom. 2003, 38, 872-878.

1892

Analytical Chemistry, Vol. 77, No. 6, March 15, 2005

respectively. The enzyme recovery was assumed to be 50%, and minimal loss of enzyme activity was noted after the purification process. RNase T1 Hydrolysis of RNA. Initially, from 2.7 to 27 µg of tRNAVal was used to prepare the stock tRNA solutions. Aliquots of 10 µL of RNase T1 reconstituted in 18O or 16O water after purification were added to tRNA samples as needed. The minimum enzyme-to-substrate ratio was estimated to be ∼50 units/µg RNA. The reaction mixtures were incubated for 1 h at 37 °C. The 16Oand 18O-labeled digestion products were then combined at specified ratios prior to MALDI-MS analysis. Mass Spectrometry. All mass spectrometry experiments were performed on a Bruker Reflex IV MALDI-TOF instrument (Bruker Daltonics, Billerica, MA) having a 3-m effective flight path, a twostage gridless ion reflector, pulsed ion extraction, and a nitrogen laser (λ ) 337 nm). All MALDI spectra were acquired in negative polarity and in reflectron mode. A two-layer sample spotting approach was used. The matrix components were 300 mM THAP in acetonitrile and 250 mM DAHC in water, with each being prepared fresh before use. Approximately 0.5 µL of the THAP matrix was first spotted onto the MALDI sample target and allowed to dry. Then 2 µL of THAP and DAHC combined together in equal volumes was mixed with 2 µL of sample. Approximately 1 µL of this sample/matrix mixture was then spotted on top of the previously dried THAP, and three sample spots were spotted for each sample. Vendor-supplied Flex control and X-mass software were used for data acquisition and processing, respectively. All MALDI-MS analyses were performed using the same operating parameters. Typically, 100 laser shots were co-added per spectrum. Each spectrum was smoothed using a five-point Savitsky-Golay algorithm and background-subtracted. The monoisotopic peak abundances of the 18O- and 16O-labeled digestion products were then exported into a Microsoft Excel spreadsheet for further processing. Calibration Curve and Data Reduction. The calibration curve was generated by analyzing mixtures of tRNAVal prepared and digested in light and heavy water. The tRNA digestion products were combined in ratios of 10:1 to 1:10 ([heavy]/[light] digestion products) prior to MALDI-TOFMS. Each mixture was spotted three times, and three separate spectra were acquired from each sample spot for a total of nine mass spectra per ratio. The arithmetic mean, standard deviation, and coefficient of variation were calculated for each ratio. A calibration curve representing the mean values was then generated from a best-fit linear leastsquares fit of the data. The resulting linear equation was then used for determining the relative quantities of all other tRNA samples. RESULTS AND DISCUSSION Correlation and Precision of MALDI-MS Measurements. The basic principle of the present approach for relative RNA quantitation is shown in Figure 1. One RNA sample is hydrolyzed in the presence of 18O-labeled (heavy) water while a second RNA sample is hydrolyzed under the same experimental conditions but in the presence of normal (light) water. After enzymatic digestion, aliquots from both RNA samples are combined and analyzed by MALDI-MS. Ion abundances of the digestion product isotopic pair

Figure 1. General overview of the described 18O labeling and MALDI-MS approach for quantitation of RNA samples.

(A and A + 2) are extracted from the mass spectrum as shown in Figure 1. Ion abundance ratios of the 18O- and 16O-labeled digestion products were calculated by use of eq 1,

I18O /I16O ) (IA+2 - b*IA)/IA

(1)

where IA represents the monoisotopic peak abundance of the 16O product, IA+2 represents the combination of the monoisotopic peak abundance of the 18O digestion product and the A + 2 isotopic peak abundance of the 16O digestion product, and b represents the A + 2 isotopic peak abundance contribution from the 16O digestion product. The ratios of the isotopic pair ion abundances provide a measure of the relative quantity of the RNase digestion products, which also provide the relative quantitation of the original two RNA samples. To apply 18O labeling with MALDI-MS for the relative quantitation of RNAs, a correlation between the ion abundance ratios of the 18O- and 16O-labeled digestion products measured using MALDI-MS to the original RNA quantities must exist. To establish such a correlation, 11 RNA samples of known heavy-to-light ratios ranging from 1:10 to 10:1 were analyzed by MALDI-MS. Figure 2 is a representative MALDI mass spectrum obtained for a sample prepared at a 2:1 (heavy:light) ratio. Ions denoted with an asterisk represent expected RNase T1 digestion products from tRNAVal. The ion abundances of the isotopic pair for the RNase T1 digestion product 5′-CUCAGp-3′ were chosen for quantitation, extracted using the data processing software, and output into a spreadsheet to calculate the ion abundance ratios according to eq 1, where b is 0.2594 for 5′-CUCAGp-3′. Table 1 lists the calculated average ion abundance ratios from 9 separate spectra for the tRNAVal RNase T1 digestion product 5′-CUCAGp-3′ analyzed at 11 different heavy-to-light ratios. From these data, a calibration curve can be generated by plotting the measured ion abundance ratios versus the concentration of heavyto-light digestion products. A representative calibration curve for the 5′-CUCAGp-3′ RNase T1 digestion product of tRNAVal is shown in Figure 3. The correlation coefficient (i.e., slope) of the calibration curve is 0.937 ( 0.027 when using the method of least

Figure 2. Representative MALDI mass spectrum of RNase T1 digestion products obtained from a tRNAVal mixture prepared at a heavy-to-light RNA ratio of 2:1. The asterisk (*) denotes expected RNase digestion product pairs. Inset: expanded view of the RNase T1 digestion product 5′-CUCAGp-3′ used for quantitating RNA levels for tRNAVal with overlaid calculated isotopic distribution assuming a 2:1 ratio. Table 1. Average Ion Abundance Ratiosa Calculated Using Eq 1 for Denoted Heavy/Light Mixtures of the tRNAVal RNase T1 Digestion Product 5′-CUCAGp-3′ and the Coefficients of Variationb [heavy]/[light]

calcd av ion abundance ratio

% CV

10.0 8.00 6.00 4.00 2.00 1.00 0.500 0.250 0.167 0.125 0.100

9.21 ( 0.63 7.18 ( 0.94 6.15 ( 0.57 4.40 ( 0.38 1.98 ( 0.19 0.817 ( 0.058 0.349 ( 0.030 0.197 ( 0.032 0.170 ( 0.036 0.121 ( 0.019 0.106 ( 0.018

6.84 13.1 9.27 8.64 9.60 7.10 8.60 16.2 21.2 15.7 17.0

a ( SD, n ) 9. b The corresponding calibration curve is shown in Figure 3.

Figure 3. Calibration curve generated from the data presented in Table 1 of the RNase T1 digestion product 5′-CUCAGp-3′ from tRNAVal.

squares. The intercept for the curve is calculated to be 0.051 ( 0.122. Good linearity (R2 ) 0.993) is achieved over the range of 1:10-10:1 (heavy/light). Thus, for this particular RNase T1 digestion product, a linear correlation exists between the original RNA quantities and the ion abundance ratios of the 18O- and 16OAnalytical Chemistry, Vol. 77, No. 6, March 15, 2005

1893

labeled digestion products measured using MALDI-MS. Coefficients of variation for these analyses are all 0.25 yielding lower CVs than those having ratios