Rates of initiation of protein synthesis by two purified species of vesicular stomatitis virus messenger RNA.

Two vesicular stomatitis virus (VSV) messenger RNAs, those encoding the N (M, = 52,500) and G (M, = 63,500) can be isolated in pure form, and the nucleotide sequence of the 5’ ends and ribosome binding sites of these mRNAs have been determined. Thus, this is a suitable eukaryotic system in which to study structure-function relationships in mRNA. Our present results indicate that, under conditions which obtain in infected cells or in the reticulocyte cellfree system, the rates of polypeptide chain initiation and elongation of these two mRNAs is the same: (a) the molar ratio of G to N protein made in infected cells is very similar to the molar ratio of G to N mRNAs recovered from these cells, as was shown previously by Villarreal et al. ((1976) Biochemistry 15, 1663-1668); (b) in the infected cell, the average size of polyribosomes synthesizing G protein is about 20% larger than that synthesizing N. Thus, the density of ribosomes on the translated part of each mRNA seems to be the same: (c) the rate of binding of G and N mRNAs to reticulocyte ribosomes and formation of 80 S initiation complexes is the same; (d) the steady state size of reticulocyte polysomes containing G and N mRNAs, and the rate at which these polysomes are formed in uitro, is the same. However, two types of experiments show that, under certain conditions of inhibition of the overall rate of chain initiation, G mRNA is translated up to 2-fold less efficiently than is N mRNA. First, as was shown by Nuss and Koch ((1976) J. Viral. 17, 283-286), treatment of infected cells with media containing hypertonic salt results in inhibition of synthesis of all VSV proteins, but synthesis of G is reduced twice as much as is N. Second, we showed that addition to a wheat germ protein synthesis reaction of the inhibitors of polypeptide chain initiation poly(dT) or aurintricarboxylate results in inhibition of initiation of translation on all VSV mRNAs, but that synthesis of G is inhibited twice as much as is N. We conclude that G and N mRNAs do differ in their affinity toward or requirement for some component or factor required for chain initiation, possibly that substance(s) whose function is blocked by the above inhibitors. We postulate that the amount of this substance is not normally rate-limiting for mRNA translation, and


Two vesicular stomatitis virus (VSV) messenger
RNAs, those encoding the N (M, = 52,500) and G (M, = 63,500) can be isolated in pure form, and the nucleotide sequence of the 5' ends and ribosome binding sites of these mRNAs have been determined.
Thus, this is a suitable eukaryotic system in which to study structure-function relationships in mRNA.
Our present results indicate that, under conditions which obtain in infected cells or in the reticulocyte cellfree system, the rates of polypeptide chain initiation and elongation of these two mRNAs is the same: (a) the molar ratio of G to N protein made in infected cells is very similar to the molar ratio of G to N mRNAs recovered from these cells, as was shown previously by Villarreal et al. ((1976) Biochemistry 15, 1663-1668); (b) in the infected cell, the average size of polyribosomes synthesizing G protein is about 20% larger than that synthesizing N. Thus, the density of ribosomes on the translated part of each mRNA seems to be the same: (c) the rate of binding of G and N mRNAs to reticulocyte ribosomes and formation of 80 S initiation complexes is the same; (d) the steady state size of reticulocyte polysomes containing G and N mRNAs, and the rate at which these polysomes are formed in uitro, is the same. However, two types of experiments show that, under certain conditions of inhibition of the overall rate of chain initiation, G mRNA is translated up to 2-fold less efficiently than is N mRNA.
First, as was shown by  J. Viral. 17, 283-286), treatment of infected cells with media containing hypertonic salt results in inhibition of synthesis of all VSV proteins, but synthesis of G is reduced twice as much as is N. Second, we showed that addition to a wheat germ protein synthesis reaction of the inhibitors of polypeptide chain initiation poly(dT) or aurintricarboxylate results in inhibition of initiation of translation on all VSV mRNAs, but that synthesis of G is inhibited twice as much as is N. We conclude that G and N mRNAs do differ in their affinity toward or requirement for some component or factor required for chain initiation, possibly that substance(s) whose function is blocked by the above inhibitors.
We postulate that the amount of this substance is not normally rate-limiting for mRNA translation, and * This work was supported by Grant AI-08814 from the United States National Institutes of Health and Grant NP-180 from the American Cancer Society. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. thus that G and N mRNAs are normally translated at the same efficiency. Different messenger RNAs differ in the rates of attachment to ribosomal subunits or in other stages of initiation of protein synthesis (l-3). For example, in the rabbit reticulocyte each molecule of cY-globin mRNA initiates protein synthesis only 60% as frequently as does each p-globin mRNA (1,(4)(5)(6). In prokaryotes, extreme variations in initiation rates have been observed.
The rates of initiation of translation of the three cistrons of bacteriophage f2 RNA differ more than 20-fold (71, and a similar situation obtains for several cistrons in bacteriophage h mRNA (8 . (13); also, the sequence of the ribosome binding sites of the G, N, and NS mRNAs have been determined (14).
Two studies suggest that all VSV mRNAs initiate protein synthesis at the same relative rate in the infected cell. In one, the relative synthesis of the VSV proteins and mRNAs was compared (151, but this type of study is subject to many uncertainties: All of the mRNA encoding the G protein is bound to intracellular membranes (10, 16,18)  Cells -A modification of an earlier procedure was used (24,25). One hundred milliliters of Chinese hamster ovary (CHO) cells in spinner culture at 4 x lo" cells/ml were concentrated by centrifugation (1500 x g, 5 min), washed twice in PO,-free medium, and then resuspended in 10 ml of PO,-free medium. Actinomycin D (final concentration 5 pg/ ml) and VSV (multiplicity of infection, 6) were then added, and incubation at 3'7" was continued. At 30 min. an additional 10 ml of PO,-free medium was added, and at 60 min after addition of virus, 10 mCi of carrier-free .12P0, (New England Nuclear Corp., Boston, Mass.). Incubation was continued for an additional 3 h.
To prepare polysomes, emetine (final concentration 100 @g/ml) was added to the cells to block polypeptide chain elongation and the cultures were chilled on ice. All subsequent steps were done at 4". Cells were recovered by centrifugation, washed once in phosphatebuffered saline (0.9% NaCl solution) containing 100 pg/ml of emetine, once with Buffer A (0. The optical density at 260 nm was recorded continuously; collection tubes contained 0.15 ml of 10% sodium dodecyl sulfate. To quantitate acid-precipitable radioactivity, one aliquot of 50 ~1 was added to a tube containing 100 pg of carrier RNA and 1 ml of a 10% solution of trichloroacetic acid. The precipitate was collected on fiberglass filters, washed, dried, and counted in a scintillation counter.
with an equal volume of chloroform.
To the aqueous layer was added sodium acetate (pH 4.7) to 0.4 M, then 2 volumes of ethanol. After sitting at -20" for 1 h, the RNA precipitate was recovered by centrifugation.
It was dissolved in 0.4 M sodium acetate (pH 4.7) and reprecipitated with ethanol. Finally, the RNA pellet was dried at room temperature and dissolved in 0.1 ml of water.
Gel Electrophoresm of RNA -Five-microliter samples of RNA were denatured and subjected to electrophoresis on a 3.75% polyacrylamide gel containing 98% formamide, as detailed previously (11,24). The wet gel was subjected to radioautography with Kodak No-Screen x-ray film. A Joyce-Loebl microdensitometer was used to scan the autoradiographs, using a maximum pen deflection of 1.16 A. The height of the peaks was proportional to the radioactivity providing the optical density was below 1.0 A. Areas under the peaks were determined with a planimeter. VW mRNAs-Labeling and isolation of VSV mRNAs from infected cells has been described, as was the procedure for preparing radiochemically pure "2P-labeled G and N mRNAs (11,13).
Cell-free Protein Synthesis -Conditions for protein synthesis in cell-free extracts of rabbit reticulocytes (26) and wheat germ (27) have been described in detail. Conditions for studying binding of labeled VSV mRNAs to ribosomes and incorporation into polyribosomes have also been described (24).

Gel EZectrophoresis of Protein -Pancreatic
RNase (50 pg/ml) was added to all reactions, followed by incubation at 37" for 5 min. Generally, 5 ~1 of the reaction were analyzed on a 13% polyacrylamide slab gel as detailed by Laemmli (28). Samples of labeled, infected cells were analyzed in a similar manner (29). The gels were fixed (301, dried, and subjected to radioautography with Kodak Royal Blue single-side emulsion x-ray film. As described above, the radioautograms were scanned with a Joyce-Loebl microdensitometer and the peaks quantitated by planimetry. (not shown) indicated that this mixture of detergents had no effect on the sedimentation profile of rabbit reticulocyte polyribosomes. Following centrifugation through a sucrose gradient to resolve polvribosomes, RNA was ex-In cases where total cullular RNA was to be extracted ( dodecyl sulfate (final concentration 1%) was added. The solution fractionated by gel electrophoresis (Fig. 2). As is apparent was incubated at 37" for 10 min. from Fig. 2  were labeled with 32P0,, disrupted, and fractionated by centrifugation through a sucrose gradient as described under "Materials and Methods." a, profile of optical density (solid line); and acid-precipitable radioactivity (circles) per 50-4 aliquot. Centrifugation is from right to left; monosomes are in Fraction 16, disomes in 12, trisomes in 9, and tetrasomes in 6. b, ration of 32P radioactivity in Bands II (G mRNA) to III (N mRNA) in pooled gradient fractions. As detailed under "Materials and Methods," RNA was extracted from pooled gradient fractions and analyzed by electrophoresis through polyacrylamide gels (Fig. 2). The ratio was determined from the areas under the peaks in the scans shown in Fig. 2.

Sizes of Polyribosomes
90% of the labeled material in Fractions 1 to 21 is VSV RNA (data not shown). By multiplying the total radioactivity in each Fraction A to E by the fraction of radioactivity in each fraction that is in G mRNA (calculated by determining the area under the peaks in Fig. 21, one can determine the distribution of G mRNA in these polysome fractions. Calculations similar to those detailed in Table I  Using very different procedures for isolation of RNA from the entire cell (see "Materials and Methods") and different estimates of RNA and protein molecular weights (11, 121, we have been able to reproduce their results. Infected cells were labeled, in parallel, with 32P04 ( Fig. 3) or with a mixture of 'C-amino-acids (Fig.  4). Radioactive RNA and proteins extracted from the entire cell were fractionated by polyacrylamide gel electrophoresis, and the amounts of labeled material were determined by integrating the area under the peaks from the microdensitometer scans of the radioautograms (Figs. 3 and 4). Using values we previously determined for the sizes of G and N RNA, for the N polypeptide (12) and for the polypeptide (GJ which is the unglycosylated translation product of G mRNA (29), we calculate ( Table I)  In all cases, one-twentieth of the RNA recovered from each fraction was analyzed.
proteins made in the infected cell is about 0.68, and the molar ratio of G:N mRNAs is about 0.55. It is apparent that the yield of polypeptide per molecule of G and N mRNAs are very nearly the same. The major problems with this experiment concern the uncertainties in the molecular weights of these RNAs and proteins, and in the unproven assumption that all mRNAs and proteins are extracted from the cell at the same efficiency.
Polysomes Containing G and N mRNAs in Reticulocyte Cell-free System -Crude cell-free extracts from rabbit reticulocytes translate endogenous globin mRNA and exogenous mRNAs at high efficiency (31); each mRNA is translated 10 to 50 times during a 30-min incubation period. As was shown previously, a large fraction of total VSV mRNA is incorporated into polyribosomes in such a lysate under conditions of protein synthesis (24,32). These polysomes contained an average of four ribosomes per mRNA (c.f. Fig. 5, Table II) similar to that obtained in intact cells (Fig. 1) 180, 210, 225, 240, and 270 min of infection, and the ratios of G/N mRNAs were averaged from these five samples. The absolute value of G/N mRNAs (shown 2 the standard deviation) is the ratio of the areas under the peaks in the scans of the radioautograms; the molar ratio is calculated from this value assuming molecular weight for G and N mRNAs of 700,000 and 500,000, respectively (11). "C-proteins were isolated after 8, 12, 15, 20, and 25 min of labeling, and the average absolute ratio of G/N polypeptides was averaged from these five samples. The molar ratio was RNA was isolated from the cell pellet as detailed under "Materials and Methods," and 10% was analyzed by polyacrylamide gel electrophoresis.
Shown is a scan of the radioautogram. As described in Table I, RNA samples were prepared and analyzed after other times of labeling.
in intact reticulocytes (5). Thus, these extracts seem appropriate for in vitro investigations of structure-function relationships of mRNA. Fig. 5 and Table II show that 50% of added G mRNA or N mRNA can be bound to reticulocyte ribosomes; the extent of binding is maximal at 1.5 min of incubation (at 30") and does not change during the next 3.5 (Table II) Table I) of further incubation, cells from an aliquot of the culture were recovered by centrifugation.
These were dissolved in buffer containing sodium dodecyl sulfate and analyzed by polyacrylamide gel electrophoresis (see "Materials and Methods"). Shown are scans of the radioautograms of the dried gel. from this profile, due predominantly to the endogenous polysomes containing globin mRNA, the positions of the monoribosomes and polysomes of different sizes were ascertained (numbers in Panels a and c). The entire contents of each fraction (about 1.0 ml) was counted, following addition of Aquasol (N. E. Nuclear Corp.) and water, in a liquid scintillation counter. Analysis of this experiment is in  The polysome gradients depicted in Fig. 5 are those labeled peak. Hence, the total number of ribosomes translating the mRNAs Experiment I; Experiment II was a similar study using different is proportional to the sum of this product over all regions of the preparations of RNA. In calculating the data given in Column 5, polysome gradient sedimenting with or faster than monoribosomes. the amount of radioactivity in any region of a polysome gradient, Since the total radioactivity in polysomes (Column 4) is proportional C,, was multiplied by the number of ribosomes per polysome (9, to the number of mRNAs being translated, the average number of that is, by the number of ribosomes translating each mRNA in that ribosomes per bound mRNA is given by Column 6. Reactions and method of analysis have been described in detail (24). Reticulocyte cell-free reactions (300 ~1) containing 1 mg/ml of anisomycin were incubated at 25" for 30 s. Twenty microliters of water containing about lo6 cpm of G or N ln2PlmRNA were added. After further incubation for 15, 30, or 240 s, an aliquot of 100 ~1 was added to 1.2 ml of ice-cold Buffer D (see legend to Fig. 51. Gradient analysis was as in Fig. 5, except centrifugation was at 16,000 rpm for 20 h. Shown is the fraction of 32P radioactivity sedimenting under the 80 S monosome peak. In the absence of incubation, between 0.02 and 0.04 of the added labeled RNA sedimented with 80 S monosomes. This background has been subtracted from all values shown. The method for calculating the time for half-maximal binding was described in detail (24).  Effects of inhibitors of polypeptide chazn initiation on synthesis of G and N protans

Seconds of incubation
Reactions identical with those of Fig. 6 but containing the indicated concentrations of ATA or poly(dT) were employed. The ratio of G to N polypeptide produced (last column) was calculated from the areas under the respective peaks in scans of the radioautograms. The values in parentheses are the ratios normalized to a value of 1.0 for the control (uninhibited) reaction. an amount of VSV mRNA one-fourth that which results in maximum synthesis, other reagents required for protein synthesis (28,33), and in addition, 3 x 10-j M aurintricarboxylic acid (Panel b) or 1.1 x 10m4 M poly(dT). Incubation was at 25" for 120 min. Acid-precipitable protein radioactivity in a 5~1 aliquot of the reaction was (a) 198,000 cpm; (b) 78,000 cpm; (c) 45,000 cpm; sample with no VSV mRNA, 4,000 cpm. A 5~1 aliquot of each reaction was analyzed by polyacrylamide gel electrophoresis.
Shown are scans of radioautograms of dried gels which were exposed to x-ray film for 20 h (Panel a) or 80 h (Panels b and c).

translation
of VSV G (and M) mRNAs. Extrapolation of the data in Fig. 6 and Table IV Rates of all subsequent stages of chain initiation, binding to 40 S.Met-tRNA, subunits, addition of 60 S subunits, are assumed to be the same. Thus, the relative rates of translation of the two mRNAs will be proportional to the relative concentrations of the (mRNA.F) complex. If Gt and NI represent the total amount of G and N mRNAs, then the amounts of the two complexes will be given by  which could be related to these differences in rates of chain initiation.
All VSV mRNAs have the same nucleotide sequence at the 5' end: m'G(5')pppA(m)pApCpApGp, and a poly(A) sequence of about 100 bases at the 3' end (131, so these regions are unlikely to be involved in these differences in function. The AUG initiation codon in N mRNA is 13 nucleotides from the 5' end, and the m7G-containing "cap" sequence is contained within the region protected by 80 S initiation complexes from nuclease digestion. By contrast, the AUG initiation codon in G mRNA is greater than 19 nucleotides from the 5' end, and the 5' "cap" is not protected by the 80 S ribosomes (14). The presence of the m7G residue does enhance the rate of chain initiation by VSV mRNAs by mammalian ribosomes (24,32,39), and it is possible that the closer proximity of the 5' cap to the AUG initiator codon in N mRNA enhances its affinity for some initiation factor, possibly one which binds to, in part, the m7GpppX sequence on mRNA.
In the infected cell, polysomes containing G mRNA are localized to the endoplasmic reticulum, while N mRNA is found on free polysomes (10, 16,17). Attachment of the G mRNA.ribosome complex to membranes is mediated by the nascent protein chain (25), and initiation of translation of G mRNA occurs with ribosome subunits which need not be attached to membranes (40). The localization of G mRNA to membranes thus seems unlikely to affect, directly, the process of polypeptide chain initiation, although initiation on G mRNA could be facilitated by a pool of 40 S ribosome subunits bound to the endoplasmic reticulum (41).