ANALYTICAL BIOCHEMISTRY 163, 5 13-5 16 ( 1987)
A Simple and Rapid Solid-Phase
RNA Sequencing
Method
YAN ZHANG, WANGYI LIU,* YAXIONG FENG, AND T. P. WANG* Department of Bioscience and Technology, Shanghai Jiao-Tong University, 1954 Hua-San Road, Shanghai, China, and *Shanghai Institute of Biochemistry, Academia Sinica, 320 Yue-Yang Road, Shanghai, China Received December 15, 1986 A simple and rapid solid-phase RNA sequencing method has been developed based on Peattie’s direct chemical method. 3’-Terminally labeled RNA was immobilized on DEAE-cellulose sheets and followed by specific modification with dimethyl sulfate, diethylpyrocarbonate, hydroxylamine (at pH 10 for the uridine and pH 5.5 for the &dine reaction), and cleavage reaction with aniline. RNA fragments were washed from the DEAE-cellulose sheets using salt solutions, precipitated with ethanol, and separated by 15% polyacrylamide gel electrophoresis. Due to the complete removal of the impurities normally present in the solution method, the higher resolution of the sequencing bands and lower background on the autoradiograph make this solid-phase technique more efficient. This solid-phase technique is much faster and more convenient than the original method. 0 1987 Academic press, IX. KEY WORDS: solid-phase; RNA sequencing.
The direct chemical RNA sequencing method (1) developed by Peattie has been widely used in many laboratories. The merit of this method is that expensive specific ribonucleases are not needed, but the procedure involves many cycles of washing, centrifugation, and lyophilization which are tedious and time consuming. Furthermore, incomplete removal of salts, chemical reagents, and by-products of specific reactions usually gives rise to artificial sequencing bands and excessive background on the autoradiograph, which make it difficult to read the nucleotide sequence accurately. These problems also occurred in chemical DNA sequencing. Recently, Chuvpilo and Kravchenko (2) designed a solid-phase technique for sequencing DNA on DEAE-cellulose paper to overcome the problems mentioned above. This work stimulated us to investigate the possibility of sequencing RNA on a solid support. We describe a simple and rapid solidphase method which enables one to get clearer RNA sequencing bands in much less time than with the original method. 513
MATERIALS
AND METHODS
Posterior silk gland 5 S RNA of silkworm, Philosamia Cynthia ricini, was prepared and labeled at its 3’ end with 5’-[32P]pCp by Tq RNA ligase according to the methods described previously (3). The labeled 5 S RNA was purified by 15% polyacrylamide gel electrophoresis. The labeled 5 S RNA in the gel was located by autoradiography, released by elution buffer, and finally precipitated with ethanol and dried in a freeze dryer (4). [y-32P]ATP was purchased from New England, Inc., and Tq RNA ligase was from Boehringer. The NH20H HCl reagent (I) for the C reaction was prepared as a 2.5 M stock solution adjusted to pH 5.5 with diethylamine; the NH20H HCl reagent (II) for the U reaction was prepared as a 2.5 M stock solution and adjusted to pH 10 with ammonium hydroxide. A sample of 2 X lo5 cpm of 3’-terminally labeled RNA dissolved in 5 ~1 of distilled water containing 30 pugof carrier tRNA was 0003-2697187 $3.00 Copyright 0 1987 by Academic Press, Inc. All rights of reproduction in any form reserved
514
ZHANG
ET AL.
TABLE 1 PROCEDURESFOR SOLID-PHASE RNA SEQUENCING G Add 300 ~1 G buffer” 0.5 /.d DMS 90°C 1 min
ArG 200 pl A bufferb 1 /.d DEP 90°C 5 min
C
U
7d 2.5 M NHzOH pH 5.5 90°C 6 min
7 PI 2.5 M NHzOH pH 10 90°C 3 min
Wash with ethanol, 70% ethanol, water, and ethanol in turn, and dry at room temperature 10 /.d 1 M Tris-HC1 buffer, pH 8.2; 10 pl 0.2 M NaBH4 0°C 30 min Wash and dry as above I 5 ~1 of 1 M aniline acetate buffer, pH 4.5,6O”C in dark for 20 min Wash 5 times in turn with ethanol and water; 100 ~1 eluting solution (I M NaCl, 10 mM EDTA, 10 pg tRNA); 65”C, 30 min Precipitate with 250 pl ethanol Wash 3 times with 80% ethanol Dissolve in loading buffer and load on 15% polyactylamide gel * G buffer: 50 mM sodium cacodylate, pH 5.5, 1 mM EDTA. ‘A buffer: 50 mM sodium acetate, pH 4.5, 1 mM EDTA.
applied to four marked DEAE-cellulose strips (each 0.2 X 1 cm) and then washed three times with water and twice with ethanol. The strips were then air-dried. The specific reactions for G, A, C, and U’ are summarized in Table 1. RESULTS AND DISCUSSION
The simplified solid-phase method for RNA sequencing is much faster and less laborious than the solution method of Peattie (1). Many washing, centrifugation, and lyophilization steps have been omitted in this method. The total time needed for adsorption, specific reactions, cleavage by aniline, and desorption of RNA from DEAE-cellulose strips is only 4 h. Figure 1 shows the ’ Abbreviations used: G, guanosine; A, adenosine; C, cytidine; U, uridine; DMS, dimethyl sulfate; DEP, diethylpyrocarbonate.
autoradiographic pattern of the silkworm 5 S RNA sequence worked out by the solidphase method. This pattern shows that the specificities of chemical reactions and cleavage with aniline on the DEAE-cellulose strips are the same as those in the solution method. The higher resolution of the bands and the lower background on the autoradiograph are the obvious advantages. Chuvpilo and Kravchenko (2) found that approximately one-quarter of the starting radioactivity of DNA was lost during the solidphase sequencing. The main loss occurs due to incomplete desorption of DNA from DEAE paper rather than to the incomplete chemical reaction and piperidine cleavage. Radioactivity loss can be reduced by prewashing DEAE paper with 1 M NaCl solution and water. This situation also occurred in solid-phase RNA sequencing. Tables 2 and 3 show the radioactivity recoveries in
SOLID-PHASE G
A
C
515
RNA SEQUENCING
U
FIG. 1. The autoradiographic pattern of 5 S RNA of posterior silk gland of Philosamia Cynthia ricini obtained by the solid-phase sequencing method. Separation of bands was carried out by 15% polyacrylamide gel elecreophoresis at 1600 V for 2.5 h.
both the solid-phase method and the solution method for RNA sequencing. In the beginning, we tried to use 3 M sodium chloride in anhydrous hydrazine for the C reaction and 50% hydrazine for the U reaction on the DEAE-cellulose strips. Good results for U were obtained. In the C reaction, however, many artificial bands were observed. Furthermore, in 3 M sodium chloride, hydrazine induced cellulose to peel off from the sheet. So instead of hydrazine, we used hyroxylamine for the C and U reactions. Hydroxylamine is a nucleophilic reagent which has been frequently used in functional studies of nucleic acids. Recently, Rubin and Schmid (5) used hydroxylamine at pH 6 for C-specific reactions in chemical DNA sequencing; Feng (6) used it at pH 10 for Uspecific reactions in chemical RNA sequencing. In this communication, we use hydroxylamine to selectively modify cytosine at pH 5.5 and 90°C and uracil at pH 10 and 90°C. The resulting partially modified RNA can be easily broken by P-elimination with aniline treatment through the opening of the ribose moiety and cleavage of the phosphodiester bond. As compared with hydrazine, hydroxylamine is more specific and stable for modifications of both cytosine and uracil at different pH values. It should be pointed out that the migration rate of C bands on the gel is approximately
TABLE 2 RADIOA~IWITY
RECOVERED
DURING
SOLID-PHASE
SEQUENCING
RNA
Type of specific reactions G 03-W Before modification After modification and cleavage by aniline After washing from DEAEcellulose sheets and precipitation Left on DEAE-cellulose sheets Recovery %
C @pm)
U bm)
214,260
238,460
292,180
214,960
177,790
217,300
206,660
132,050
120,440 12,680 56
151,530 23,464 64
150,100 15,836 51
94,146 11,954 44
516
ZHANG
ET AL.
TABLE 3
Type of specific reactions
Before modification Before loading on sequencing gel Recovery %
G (cpm)
A @pm)
C (cpm)
U @pm)
112,330 90,473 81
102,590 81,726 81
68,540 56,110 83
99,939 81,766 82
two nucleotides slower than that of 3 M sodium chloride anhydrous hydrazine. The chemical mechanism for this is unknown at present, but the sequence position of C may be read accurately if one keeps this fact in mind. ACKNOWLEDGMENTS We thank Hans J. Gross for his critical reading of this manuscript. We are also indebted to the Alexander Von Humboldt Foundation for scientific equipment support.
REFERENCES 1. Peattie, D. A. (1979) Proc. Natl. Acad. Sci. USA 76, 1760-1764. 2. Chuvpilo, S. A., and Kravchenko, V. V. (1985) FEBS Lett. 179, 34-36. 3. Feng, Y. X., Krupp, G. and Gross, H. J. (1982) Nucleic Acids Rex 10, 6383-6387. 4. Krupp, G., and Gross, H. J. (1979) Nucleic Acids Res. 6, 348 l-3490. 5. Rubin, C. M., and Schmid, C. W. (1980) Nucleic Acids Res. 8, 46 13-46 19. 6. Feng, Y. X. (1985) Acta Biochim. Biophys. Sinica 17,302-307.