- Email: [email protected]
Refined molecular weights for phage, viral and ribosomal RNA
Gene, 3 (1978) 353--357
353
© Elsevier/North-Holland Biomedical Press, Amsterdam -- Printed in The Netherlands
REFINED M O L E C ~ RIBOSOMAL RNA
WEIGHTS FOR PHAGE, VIRAL AND
(BAC electron microscopy; nucleotide number; coliphages Q~, MS2, R17 and f2; Pseudomonas phage PP7; NVD virus)
H.J. VOLLENWEIDER*, U. STETTLER, O. K~TBLER, TH. KOLLER and H. WEBER**
lnstitut fiir Zelibiologie, ETH-Z, and **Institut fiir Molekularbiologie I, Universit'~t Zfirich, H6nggerberg, 8093 Z~rieh (Switzerland) (Received and accepted May 5th, 1978) SUMMARY
The RNAs of the Escheriehia coli bacteriophages MS2 and Q~ as well as E. coli 16S ribosomal RNA were examined under identical conditions by electron microscopy using the protein-free benzyldimethylalkylammonium chloride (BAC) spreading technique. From the contour length ratios of the RNAs ~ d the known number of nucleotides for MS2, the chain len~hs for Q~ RNA and 16S RNA were found to be 4790 ± 150 and 1645 + 55 nucleotides. Correcting for the base composition of QB RNA the molecular ,~eight of the Na salt of this RNA is (1.64 +- 0.06) • 10 z daltons. Since published values on the relative lengths of Q~ RNA and several other homogeneous RNAs (K coli 23S rRNA, E. coil bacteriophage R17 and f2 RNAs, Pseudomonas aeruginosa phage PP7 RNA and Newcastle disease virus RNA) are available, we are able to calculate the approximate number of nudeotides for these useful standards. Although sedimentation and electrophoretic analysis are common tools for RNA molecular weight determinations, these methods may remain subject to inaccuracies and artifacts dt~e to aggregation and/or secondary and tertiary structure. In contrast, electron microscopy permits direct examination of individual RNA molecules. However, until recently no reference RNA molecule of accurately known chain length and a large enough size was available, and therefore the determination of nucleotide numbers by electron microscopic length measurements was only an approximate method. Since the complete *present address: McArdle Laboratory for Cancer Research, University of Wisconsin, Madison, WiL 53706 (U.S.A.). Abbreviation: BAC, ~nzyldimetbylaikylammonium chloride (n-alkyl mixture: C12Hzs 60%,
C,,H. 40%).
354
nucleotide sequence of MS2 RNA is now known (Fiefs et al., 1{}76), we have used this molecule as a reference for the determination of the number of nucleotides in Q~ RNA. Ribosomal 16S of E. coli served as further control, since itssequence isalmost completely known (J.P.Ebel, personal communication). It isnot possibleto distinguishaccidentallybroken Q~ R N A strands from reference R N A molecules (MS2 or 16S R N A ) when coprepared and adsorbed to the same electronmicroscope grid.It was therefore necessary to prepare the R N A specimen to be memmred (Q~) and the standard specimens (MS2 and 16S) on separate'grids.In order to estimate the fluctuationin magnification of the electron microscope introduced by changing the specimen, we measured 200 squares from four magnification calibrationreplicagrids (0.463 ~ m -+0.3%, ~ d l A m , N.Y,) which were each i n s e r t ~ into and removed from the microscope (Siemens Elmiskop 101) 5 times. The standard deviation of the determined square size which reflects the fluctuation in magnification was below + 1%. This error is about 10 times smaller than the standard deviations on the contour lengths of the measured RNA molecules (see below). It is therefore not considered to be important in this particular study. For the determination of the contour lengths micrographs were taken strictly at random, and every unambiguously visible filament longer than about 0.5/Jm for MS2 and Q~ RNA and about 0.3 # m for 16S RNA was measured. The results are summarized in the histograms of Fig.1. The size distributions fall off steeply on the right side of the peaks, but they tail off to the left. This tailing off is a common finding in single stranded RNA preparations and is certainly due to fragmented molecules (Chi and Bassel, 1974). In order to obtain the length distribution of the unbroken strands a least squares fi~) of the data to a Gaussian distn'bution was performed. The mean values determined are given in Table I. Based on the range of possible mean values given by the 3 X TABLE I CONTOUR LENGTHS OF MS2, 16S AND Q# RNA
RNA
MS2 16S Q~
Number of molecules within Gaussian distribution
Mean value
Standard deviation
3 × error of the mean (99.7% confidence level)
(~m)
(~m)
(~m)
565 490 408
1.43 0.66 1.92
± 0.15 ± 0.09 ± 0.23
_+0.02 ± 0.01 +- 0.03
error of the mean, Q~ RNA is found to be 1.30 to 1.38 times longer than MS2 RNA. Since MS2 RNA has 3569 nucleotides (Fiefs et al., 1976) we calculate for Q~ RNA 4652 to 4936 nucleotides with an average of 4792. 16S rRNA is 2.10 to 2.23 times shorter than MS2 RNA. Accordingly we compute for this RNA 1600 to 1700 nucleotides with an average of 1647. Based on the value of 4792 nucleotides for Q# RNA, we are able to indicate the approximate nucleotide number for several other RNA species, whose
355
MS2
16 S
QB n
0
I
I pm
!
Fig.1. Contour lengths of unselected MS2 RNA, 16S rRNA and Q~ RNA molecules. Ordinate: Probability density (normalized) of RNA lengths (~m -1). Abscissa: length of RNA molecules (~m). MS2 RNA and Q~ RNA were extracted from purified phage (Weissmann et al., 1968). Full-length MS2 RNA was obtained by sucrose gradient centrifugation (5--23% sucrom:, 0.05 M Tris--HCl. 0.005 M EDTA, pH "/.5 (5 ml) for 2 h at 60 000 rev./ rain and 4° C in a Beckman SW65 rotor). Q~ RNA was used without further purification. 16S rRNA was a gift of Dr. Ebel. For denaturation the RNA was heated for 10 rain to 63 ° C in 8.5% formaldehyde and 80 raM triethanolamine---HCl buffer (pH 7.9), and rapidly cooled to 0°C. Under these conditions denaturation was complete (Boedtker, 1976) as verified spectrophotometrically at 260 nra in parallel samples heated up to 90 ° C. To 0.1 ml of RNA solution (5.0 ~g/ml) were added 0.2 ml of a solution containing 80 mM triethanolamine--HCl (pH 7.9), 3.5% formaldehyde, 2% formamide and 3.6.10-s% BAC. About 20 ~1 of this mixture was spread onto a hypophase of quartz-distilled water as described by Vollenweider et al. (19"/5). The preparation of the specimens for the electron microscope and the quantitative analysis of the raicrographs were those outlined by Vollenweidar et al. (1!~76). The data were plotted by computer using a class size of 0.025 ~ra. The continuous line represents the least squares fitting of the data.
356
lengths relative to Q~ RNA have been determined by electron microscopy: Newcastle disease virus (NDV) RNA, coliphage f2 RNA, K coil 23S rRNA and E. co//16S rRNA (Chi and Bassel, 1974), Pseudom~mas aemginosa phage PP7 and f2 RNA (Edlind and Bassel, 1977), PP7 RNA and coliphage R17 RNA (Benike et al., 1975) and R17 RNA and NDV RNA (Kolakofaky et al., 1974). Furthermore, we calculated the molecular weights for the known base compositions for the Na salt of these RNAs [for reviews see Boedtker, 1976 (16S, 23S, f2, MS2, R17, Q~); Edlind and Bassel, 1977 (PPT); Duesberg and Robinson, I965, (NDV]. The data are summarized in Table II. Ou~ data for K coil 16S rRNA are in agreement with the preliminary sequence data (about 1600 nucleotides, J.P. Ebel, personal communication) as well as with the 1640 +_50 nucleotides determined by Midgley (1965) and the molecular weight of 0.56-106 daltons derived by sedimentation viscosity and light scattering (Kurland, 1960). Furthermore, 1640 nucleotides can also be computed from the data of Reijnders et al. (1973) taking the known number of nucleotides for MS2 RNA into account. However, for Q/3 RNA quite different estimates of its molecular weight are reported in the literature: Overby et al. (1966) obtained a value of 0.9.106 daltons from light scattering measurements, whereas molecular weights of 1.4.106 daltons (Reijnder et al., 1973) and 1.5-106 daltons (Boedtker, 1971) were found by methods using gel electrophoresis in urea and in formaldehyde, respectively. Based on contour length measurements by electron microscopy with respect to unsequenced reference molecules Chi and Bassel (1974) and Edlind and Bassel (1977) reported mol~ cular weights for Q~ RNA of 1.55-106 and (1.53 + 0.14) • 106, respectively. From the work cited above it is obvious that different techniques may lead to quite different results~ We believe that the method applied in the present paper provides the most accurate values so far, for the following reasons: (1) we used a large reference molecule, MS2 RNA, whose accurate chain length follows from its known nucleotide sequence; (2) our method is independent of secondary and tertiary structure and aggregation. TABLE H NUMBERS OF NUCLEOTIDE8 AND MOLECULAR WEIGHTS OF SOME RNAs AS CALCULATED FROM DATA IN THE LITERATURE USING THE DETERMINED NUMBER OF NUCLEOTIDES FOR Q~ RNA AS STANDARD RNA
Nucleotide number
Mol.wt. (daltons) (-10")
RNAb
Nucleotide number
Mol. wt. (daltons) (-10')
NDV R17 PP7 f2 23S rRNA
17900a 4140a 3950 3940a 3320
6.13 1.42 1.36 1.35 1.15
Q~ 16s MS2
4790 + 150 1645 + 55 3569
1.64 ± 0.06 0.57 + 0.02 1.23
aThe results of two reported RNA ratios were averaged. b See this paper and Fiers et al. (1976) for MS2.
357
The accuracy of the chain length ratio obtained for MS2 RNA and 16S RNA (whose length is also quite accurately known from sequence data) indicates that the ~elation between contour length and number of nucleotide~ under the fully denaturing conditions used in this study is most probably linear. ACKNOWLEDGMENTS
We are indebted to Dr. Fiers for the gift of bacteriophage MS2, and to Dr. Ebel for providing us with 16S ribosomal RNA. We thank Dr. L. Estis for critically reading the manuscript. Furthermore, we acknowledge the excellent technical assistance of Mrs. M. Soyka and Mrs. H. Mayer-Rosa. Work supported by Swiss National Science Foundation, grant No. 3.1590.73.
REFERENCES
Benike, C., McClements, W. and Davis, J.W., The molecular size of the RNA genome of Pseudomonas aeruginose bacteriophage PP7, Virology, 66 (1975) 625--628. Boedtker, H., Conformation independent molecular weight determination of RNA by gel electrophoresis, Biochim. Bio~hys. Acta, 240 (1971) 448--453. Boedtker, H., Nucleic acids-physical properties of RNA, in Fasman, G.D. (Ed.), Handbook of Biochemistry and Molecular Biology, 3rd ed., Vol.II, CRC Press, Cleveland, 1976, pp. 405--410. Chi, Y.Y. and Bassel, A.R., Electron microscopy of viral RNA: Molecular weight determination of bacterial and animal virus RNAs, J. Viror., 13 (1974) 1194--1199. Duesberg, P.H. and Robinson, W.S., Isolation of the nucleic acid of Newcastle disease virus (NDV), Proc. Natl. Acad. Sci. USA, 54 (1965) 794--800. Edlind, T.D. and Bassel, A.R., Secondary structure of RNA from bacteriophages f2, Q~ and PP7, J. Virol., 24 (1977) 135--141. Fiefs, W., Contreras, R., Duerinck, F., Haegeman, G., Iserentant, D., Merregaert, J., Min Jou, W., Molemans, F., Raeymaekers, A., Van den Berghe, A., Volckaert, G. and Ysebaert, M., Complete nucleotide sequence of bacteriophage MS2 RNA: primary mid secondary structure of the replicase gene, Nature, 260 (1976) 500--507. Kolakofsky, D., de la Tour, E.B. and Delius, H., Molecular weight determination of sendal and Newcastle disease virus RNA, J. Virol., 18 (1974) 261--268. Kurland, C.G., Molecular characterization of RNA from E. coli ribosomes, J. Mol. Biol., 2 (1960) 83-91. Midgley, J.E.M., The estimation of polynucleotide chain length by a chemical method, Biochim. Biophys. Acta, 108 (1965) 340--347. Overby, L., Barlow, G.H., Doi, R.H., Jacob, M. and Spiegelman, S., Comparison of two serologically distinct RNA bacteriophages, J. Bacteriol., 92 (1966) 739-745. Reijnder, L., Sloof, P., Sival, J. and Borst, P., Gel electrophoresis of RNA under denaturing conditions, Biochim. Biophys. Acta, 324 (1973) 320-333. Vollenweider, H.J., Sogo, J.M. and Koller, Th., A routine method for protein-free spreading of double- and single-stranded nucleic acid molecules, Proc. Natl. Acad. Sci. USA, 72 (1975) 88--87. Vollenweider, H.J., Koller, Th., Weber, H. and Weissmann, Ch., Physical mapping of Q~ replicase binding sites on Q~ RNA, J. Mol. Biol., 101 (1976) 367--377. Weissmann, Ch., Colthart, L and Libonati, M., Determination of viral plus and minus ribonucleic acid strands by an isotope dilution assay, Biochemistry, 7 (1968) 865--874. Communicated by A.J. van der Eb.
353
© Elsevier/North-Holland Biomedical Press, Amsterdam -- Printed in The Netherlands
REFINED M O L E C ~ RIBOSOMAL RNA
WEIGHTS FOR PHAGE, VIRAL AND
(BAC electron microscopy; nucleotide number; coliphages Q~, MS2, R17 and f2; Pseudomonas phage PP7; NVD virus)
H.J. VOLLENWEIDER*, U. STETTLER, O. K~TBLER, TH. KOLLER and H. WEBER**
lnstitut fiir Zelibiologie, ETH-Z, and **Institut fiir Molekularbiologie I, Universit'~t Zfirich, H6nggerberg, 8093 Z~rieh (Switzerland) (Received and accepted May 5th, 1978) SUMMARY
The RNAs of the Escheriehia coli bacteriophages MS2 and Q~ as well as E. coli 16S ribosomal RNA were examined under identical conditions by electron microscopy using the protein-free benzyldimethylalkylammonium chloride (BAC) spreading technique. From the contour length ratios of the RNAs ~ d the known number of nucleotides for MS2, the chain len~hs for Q~ RNA and 16S RNA were found to be 4790 ± 150 and 1645 + 55 nucleotides. Correcting for the base composition of QB RNA the molecular ,~eight of the Na salt of this RNA is (1.64 +- 0.06) • 10 z daltons. Since published values on the relative lengths of Q~ RNA and several other homogeneous RNAs (K coli 23S rRNA, E. coil bacteriophage R17 and f2 RNAs, Pseudomonas aeruginosa phage PP7 RNA and Newcastle disease virus RNA) are available, we are able to calculate the approximate number of nudeotides for these useful standards. Although sedimentation and electrophoretic analysis are common tools for RNA molecular weight determinations, these methods may remain subject to inaccuracies and artifacts dt~e to aggregation and/or secondary and tertiary structure. In contrast, electron microscopy permits direct examination of individual RNA molecules. However, until recently no reference RNA molecule of accurately known chain length and a large enough size was available, and therefore the determination of nucleotide numbers by electron microscopic length measurements was only an approximate method. Since the complete *present address: McArdle Laboratory for Cancer Research, University of Wisconsin, Madison, WiL 53706 (U.S.A.). Abbreviation: BAC, ~nzyldimetbylaikylammonium chloride (n-alkyl mixture: C12Hzs 60%,
C,,H. 40%).
354
nucleotide sequence of MS2 RNA is now known (Fiefs et al., 1{}76), we have used this molecule as a reference for the determination of the number of nucleotides in Q~ RNA. Ribosomal 16S of E. coli served as further control, since itssequence isalmost completely known (J.P.Ebel, personal communication). It isnot possibleto distinguishaccidentallybroken Q~ R N A strands from reference R N A molecules (MS2 or 16S R N A ) when coprepared and adsorbed to the same electronmicroscope grid.It was therefore necessary to prepare the R N A specimen to be memmred (Q~) and the standard specimens (MS2 and 16S) on separate'grids.In order to estimate the fluctuationin magnification of the electron microscope introduced by changing the specimen, we measured 200 squares from four magnification calibrationreplicagrids (0.463 ~ m -+0.3%, ~ d l A m , N.Y,) which were each i n s e r t ~ into and removed from the microscope (Siemens Elmiskop 101) 5 times. The standard deviation of the determined square size which reflects the fluctuation in magnification was below + 1%. This error is about 10 times smaller than the standard deviations on the contour lengths of the measured RNA molecules (see below). It is therefore not considered to be important in this particular study. For the determination of the contour lengths micrographs were taken strictly at random, and every unambiguously visible filament longer than about 0.5/Jm for MS2 and Q~ RNA and about 0.3 # m for 16S RNA was measured. The results are summarized in the histograms of Fig.1. The size distributions fall off steeply on the right side of the peaks, but they tail off to the left. This tailing off is a common finding in single stranded RNA preparations and is certainly due to fragmented molecules (Chi and Bassel, 1974). In order to obtain the length distribution of the unbroken strands a least squares fi~) of the data to a Gaussian distn'bution was performed. The mean values determined are given in Table I. Based on the range of possible mean values given by the 3 X TABLE I CONTOUR LENGTHS OF MS2, 16S AND Q# RNA
RNA
MS2 16S Q~
Number of molecules within Gaussian distribution
Mean value
Standard deviation
3 × error of the mean (99.7% confidence level)
(~m)
(~m)
(~m)
565 490 408
1.43 0.66 1.92
± 0.15 ± 0.09 ± 0.23
_+0.02 ± 0.01 +- 0.03
error of the mean, Q~ RNA is found to be 1.30 to 1.38 times longer than MS2 RNA. Since MS2 RNA has 3569 nucleotides (Fiefs et al., 1976) we calculate for Q~ RNA 4652 to 4936 nucleotides with an average of 4792. 16S rRNA is 2.10 to 2.23 times shorter than MS2 RNA. Accordingly we compute for this RNA 1600 to 1700 nucleotides with an average of 1647. Based on the value of 4792 nucleotides for Q# RNA, we are able to indicate the approximate nucleotide number for several other RNA species, whose
355
MS2
16 S
QB n
0
I
I pm
!
Fig.1. Contour lengths of unselected MS2 RNA, 16S rRNA and Q~ RNA molecules. Ordinate: Probability density (normalized) of RNA lengths (~m -1). Abscissa: length of RNA molecules (~m). MS2 RNA and Q~ RNA were extracted from purified phage (Weissmann et al., 1968). Full-length MS2 RNA was obtained by sucrose gradient centrifugation (5--23% sucrom:, 0.05 M Tris--HCl. 0.005 M EDTA, pH "/.5 (5 ml) for 2 h at 60 000 rev./ rain and 4° C in a Beckman SW65 rotor). Q~ RNA was used without further purification. 16S rRNA was a gift of Dr. Ebel. For denaturation the RNA was heated for 10 rain to 63 ° C in 8.5% formaldehyde and 80 raM triethanolamine---HCl buffer (pH 7.9), and rapidly cooled to 0°C. Under these conditions denaturation was complete (Boedtker, 1976) as verified spectrophotometrically at 260 nra in parallel samples heated up to 90 ° C. To 0.1 ml of RNA solution (5.0 ~g/ml) were added 0.2 ml of a solution containing 80 mM triethanolamine--HCl (pH 7.9), 3.5% formaldehyde, 2% formamide and 3.6.10-s% BAC. About 20 ~1 of this mixture was spread onto a hypophase of quartz-distilled water as described by Vollenweider et al. (19"/5). The preparation of the specimens for the electron microscope and the quantitative analysis of the raicrographs were those outlined by Vollenweidar et al. (1!~76). The data were plotted by computer using a class size of 0.025 ~ra. The continuous line represents the least squares fitting of the data.
356
lengths relative to Q~ RNA have been determined by electron microscopy: Newcastle disease virus (NDV) RNA, coliphage f2 RNA, K coil 23S rRNA and E. co//16S rRNA (Chi and Bassel, 1974), Pseudom~mas aemginosa phage PP7 and f2 RNA (Edlind and Bassel, 1977), PP7 RNA and coliphage R17 RNA (Benike et al., 1975) and R17 RNA and NDV RNA (Kolakofaky et al., 1974). Furthermore, we calculated the molecular weights for the known base compositions for the Na salt of these RNAs [for reviews see Boedtker, 1976 (16S, 23S, f2, MS2, R17, Q~); Edlind and Bassel, 1977 (PPT); Duesberg and Robinson, I965, (NDV]. The data are summarized in Table II. Ou~ data for K coil 16S rRNA are in agreement with the preliminary sequence data (about 1600 nucleotides, J.P. Ebel, personal communication) as well as with the 1640 +_50 nucleotides determined by Midgley (1965) and the molecular weight of 0.56-106 daltons derived by sedimentation viscosity and light scattering (Kurland, 1960). Furthermore, 1640 nucleotides can also be computed from the data of Reijnders et al. (1973) taking the known number of nucleotides for MS2 RNA into account. However, for Q/3 RNA quite different estimates of its molecular weight are reported in the literature: Overby et al. (1966) obtained a value of 0.9.106 daltons from light scattering measurements, whereas molecular weights of 1.4.106 daltons (Reijnder et al., 1973) and 1.5-106 daltons (Boedtker, 1971) were found by methods using gel electrophoresis in urea and in formaldehyde, respectively. Based on contour length measurements by electron microscopy with respect to unsequenced reference molecules Chi and Bassel (1974) and Edlind and Bassel (1977) reported mol~ cular weights for Q~ RNA of 1.55-106 and (1.53 + 0.14) • 106, respectively. From the work cited above it is obvious that different techniques may lead to quite different results~ We believe that the method applied in the present paper provides the most accurate values so far, for the following reasons: (1) we used a large reference molecule, MS2 RNA, whose accurate chain length follows from its known nucleotide sequence; (2) our method is independent of secondary and tertiary structure and aggregation. TABLE H NUMBERS OF NUCLEOTIDE8 AND MOLECULAR WEIGHTS OF SOME RNAs AS CALCULATED FROM DATA IN THE LITERATURE USING THE DETERMINED NUMBER OF NUCLEOTIDES FOR Q~ RNA AS STANDARD RNA
Nucleotide number
Mol.wt. (daltons) (-10")
RNAb
Nucleotide number
Mol. wt. (daltons) (-10')
NDV R17 PP7 f2 23S rRNA
17900a 4140a 3950 3940a 3320
6.13 1.42 1.36 1.35 1.15
Q~ 16s MS2
4790 + 150 1645 + 55 3569
1.64 ± 0.06 0.57 + 0.02 1.23
aThe results of two reported RNA ratios were averaged. b See this paper and Fiers et al. (1976) for MS2.
357
The accuracy of the chain length ratio obtained for MS2 RNA and 16S RNA (whose length is also quite accurately known from sequence data) indicates that the ~elation between contour length and number of nucleotide~ under the fully denaturing conditions used in this study is most probably linear. ACKNOWLEDGMENTS
We are indebted to Dr. Fiers for the gift of bacteriophage MS2, and to Dr. Ebel for providing us with 16S ribosomal RNA. We thank Dr. L. Estis for critically reading the manuscript. Furthermore, we acknowledge the excellent technical assistance of Mrs. M. Soyka and Mrs. H. Mayer-Rosa. Work supported by Swiss National Science Foundation, grant No. 3.1590.73.
REFERENCES
Benike, C., McClements, W. and Davis, J.W., The molecular size of the RNA genome of Pseudomonas aeruginose bacteriophage PP7, Virology, 66 (1975) 625--628. Boedtker, H., Conformation independent molecular weight determination of RNA by gel electrophoresis, Biochim. Bio~hys. Acta, 240 (1971) 448--453. Boedtker, H., Nucleic acids-physical properties of RNA, in Fasman, G.D. (Ed.), Handbook of Biochemistry and Molecular Biology, 3rd ed., Vol.II, CRC Press, Cleveland, 1976, pp. 405--410. Chi, Y.Y. and Bassel, A.R., Electron microscopy of viral RNA: Molecular weight determination of bacterial and animal virus RNAs, J. Viror., 13 (1974) 1194--1199. Duesberg, P.H. and Robinson, W.S., Isolation of the nucleic acid of Newcastle disease virus (NDV), Proc. Natl. Acad. Sci. USA, 54 (1965) 794--800. Edlind, T.D. and Bassel, A.R., Secondary structure of RNA from bacteriophages f2, Q~ and PP7, J. Virol., 24 (1977) 135--141. Fiefs, W., Contreras, R., Duerinck, F., Haegeman, G., Iserentant, D., Merregaert, J., Min Jou, W., Molemans, F., Raeymaekers, A., Van den Berghe, A., Volckaert, G. and Ysebaert, M., Complete nucleotide sequence of bacteriophage MS2 RNA: primary mid secondary structure of the replicase gene, Nature, 260 (1976) 500--507. Kolakofsky, D., de la Tour, E.B. and Delius, H., Molecular weight determination of sendal and Newcastle disease virus RNA, J. Virol., 18 (1974) 261--268. Kurland, C.G., Molecular characterization of RNA from E. coli ribosomes, J. Mol. Biol., 2 (1960) 83-91. Midgley, J.E.M., The estimation of polynucleotide chain length by a chemical method, Biochim. Biophys. Acta, 108 (1965) 340--347. Overby, L., Barlow, G.H., Doi, R.H., Jacob, M. and Spiegelman, S., Comparison of two serologically distinct RNA bacteriophages, J. Bacteriol., 92 (1966) 739-745. Reijnder, L., Sloof, P., Sival, J. and Borst, P., Gel electrophoresis of RNA under denaturing conditions, Biochim. Biophys. Acta, 324 (1973) 320-333. Vollenweider, H.J., Sogo, J.M. and Koller, Th., A routine method for protein-free spreading of double- and single-stranded nucleic acid molecules, Proc. Natl. Acad. Sci. USA, 72 (1975) 88--87. Vollenweider, H.J., Koller, Th., Weber, H. and Weissmann, Ch., Physical mapping of Q~ replicase binding sites on Q~ RNA, J. Mol. Biol., 101 (1976) 367--377. Weissmann, Ch., Colthart, L and Libonati, M., Determination of viral plus and minus ribonucleic acid strands by an isotope dilution assay, Biochemistry, 7 (1968) 865--874. Communicated by A.J. van der Eb.