Newly made viral RNA in lysates of Escherichia coli infected with phage R17 was studied as a model messenger RNA. The viral RNA that was associated with groups of 1, 2, 3 and 4 ribosomes banded in caesium chloride at densities of approximately 1·68, 1·65, 1·645 and 1·635 g cm-3 respectively and the single ribosomes of uninfected cells at 1·62. The density of infected ribosomes was restored to 1·62 by treatment with RNase. These values are consistent with the presence of a single molecule of viral RNA in each group.

In uninfected bacteria the mean density of the smaller polysomes exceeded that of actino-mycintreated or RNase-treated ribosomes by about 0·005 g cm−3. Similarly, pulse-labelled RNA associated with single ribosomes or pairs banded at a density only marginally higher than that of the bulk of the single ribosomes. If any RNA was attached to the ribosomes it probably amounted to not more than one fifth of the size of the phage RNA for each ribosome pair.

Polysomes are believed to consist of groups of ribosomes translating, and linked together by, a single molecule of messenger RNA (mRNA) (Schlessinger, 1969). The fact that the groups are dis-aggregated by ribonuclease is often quoted in support of this interpretation although some doubt has been expressed over the validity of such evidence (Fenwick, 1968). Viral RNA is at present the most clearly identifiable mRNA. Bacteriophage RNA in association with ribosomes in vitro can direct the synthesis of viral protein (Nathans, 1965), and the parental and progeny viral RNA sediment with polysomes in lysates of infected bacteria (Godson & Sinsheimer, 1967; Godson, 1968). In actinomycin-treated infected spheroplasts, in which only virus specific polysomes exist and only viral proteins are made, the predominant active ribosomes are monomers and dimers, with decreasing numbers of trimers, tetramers and larger groups (Fenwick, 1971).

In the following experiments newly made viral RNA in lysates of infected bacteria has been studied in order to characterize the mRNA-ribosome complexes. For comparison, studies have also been made with the smaller polysomes of uninfected bacteria.

The growth of phage R17 in spheroplasts of Eschercltia coli K12, strain Hfr1 in the presence of actinomycin D and the analysis of lysates by sucrose and CsCl gradient centrifugation have been described in the previous paper (Fenwick, 1971). In these experiments the infected actinomycin-treated spheroplasts were labelled with [3H]uridine for 10 min, 30–40 min after infection, and lysed and centrifuged in sucrose gradients containing 0·01 M Mg2+, 0·01M sodium phosphate. Polysomes were preserved by adding chloramphenicol (CAP), 100 μg/ml, to the culture 4 min before chilling in ice at the end of the incubation (Fenwick, 1968), except in the case of short pulse-labelling (experiment of Fig. 4) when the incubation was terminated by pouring into an ice-cold flask containing CAP, 100 μg/ml.

Infected cells

The distribution in sucrose gradients of ribosom.es and small polysomes in lysates of normal growing cells and of infected spheroplasts is shown in Fig. 1. After a 10-min labelling period intact viral RNA in lysates of infected spheroplasts sediments in a series of discreet bands (Fenwick, 1971). One is just ahead of the 70-s ribosomes, at about 75 s; others correspond to the ribosome dimers, trimers and tetramers.

Fig. 1.
Distribution of ribosomes in sucrose gradients. A, Growing bacteria were treated with CAP, 100 μg/ml, for 4 min, chilled, converted to spheroplasts in the presence of CAP and lysed. B, Infected spheroplasts incubated with actinomycin were treated with CAP, 100μg/ml, for 4 min 36–40 min after infection, chilled, centrifuged and lysed. Lysates were centrifuged in sucrose gradients at 50000 rev/min for 30 min at 10 °C.
View largeDownload slide

Distribution of ribosomes in sucrose gradients. A, Growing bacteria were treated with CAP, 100 μg/ml, for 4 min, chilled, converted to spheroplasts in the presence of CAP and lysed. B, Infected spheroplasts incubated with actinomycin were treated with CAP, 100μg/ml, for 4 min 36–40 min after infection, chilled, centrifuged and lysed. Lysates were centrifuged in sucrose gradients at 50000 rev/min for 30 min at 10 °C.

Fig. 1.
Distribution of ribosomes in sucrose gradients. A, Growing bacteria were treated with CAP, 100 μg/ml, for 4 min, chilled, converted to spheroplasts in the presence of CAP and lysed. B, Infected spheroplasts incubated with actinomycin were treated with CAP, 100μg/ml, for 4 min 36–40 min after infection, chilled, centrifuged and lysed. Lysates were centrifuged in sucrose gradients at 50000 rev/min for 30 min at 10 °C.
View largeDownload slide

Distribution of ribosomes in sucrose gradients. A, Growing bacteria were treated with CAP, 100 μg/ml, for 4 min, chilled, converted to spheroplasts in the presence of CAP and lysed. B, Infected spheroplasts incubated with actinomycin were treated with CAP, 100μg/ml, for 4 min 36–40 min after infection, chilled, centrifuged and lysed. Lysates were centrifuged in sucrose gradients at 50000 rev/min for 30 min at 10 °C.

Infected spheroplasts were labelled in their viral RNA by a 10-min pulse of [3H]uridine. The lysate was fractionated on a sucrose gradient like that of Fig. 1B, collecting the optically dense peaks of ribosomes from monomers to tetramers separately. Each class was fixed with formaldehyde and then mixed with a sample of fixed 70s ribosomes obtained in the same way from uninfected cells labelled in their ribosomal RNA with32P and chilled in the absence of CAP to allow accumulation of single ribosomes. The results of centrifugation in CsCl are shown in Fig. 2. The viral RNA in each case was found at a higher density than the 70-s ribosomes from uninfected cells, although the difference was progressively less with increasing polysome size. The viral RNA associated with single ribosomes had a mean buoyant density of 1-68 while that obtained from the dimers, trimers and tetramers banded at approximately 1·65, 1·645 an d 1·635 respectively. These differences in density were confirmed by mixing singles with dimers or with trimers from infected cells (the singles were labelled in their viral RNA by a 10-min pulse of [3H]uridine and the dimers and trimers by a 15-min pulse of32P in a different infected culture).

Fig. 2.
Density in CsCI of polysome-associated viral RNA compared with normal ribosomes. Uninfected bacteria were labelled for 90 min with 32P, 1 μCi/ml and chilled. They were made into spheroplasts, lysed and fractionated on a sucrose gradient. The 70-s peak was collected and fixed with formaldehyde. Infected spheroplasts were labelled with [3H]uridine (5 μCi/ml, uridine 5 μg/ml) for 10 min, treated with CAP and the lysate fractionated on a sucrose gradient. The ribosome monomers (A), dimers (B), trimers (c) and tetramers (D) were collected, fixed and each mixed with a sample of normal 32P-labelled 70-s ribosomes. CsCI was added and the mixture centrifuged at 35000 rev/minfor 15 hat 10 °C. ○ — ○ 32P; • — •,3H; —, density of CsCI.
View largeDownload slide

Density in CsCI of polysome-associated viral RNA compared with normal ribosomes. Uninfected bacteria were labelled for 90 min with 32P, 1 μCi/ml and chilled. They were made into spheroplasts, lysed and fractionated on a sucrose gradient. The 70-s peak was collected and fixed with formaldehyde. Infected spheroplasts were labelled with [3H]uridine (5 μCi/ml, uridine 5 μg/ml) for 10 min, treated with CAP and the lysate fractionated on a sucrose gradient. The ribosome monomers (A), dimers (B), trimers (c) and tetramers (D) were collected, fixed and each mixed with a sample of normal 32P-labelled 70-s ribosomes. CsCI was added and the mixture centrifuged at 35000 rev/minfor 15 hat 10 °C. ○ — ○ 32P; • — •,3H; —, density of CsCI.

Fig. 2.
Density in CsCI of polysome-associated viral RNA compared with normal ribosomes. Uninfected bacteria were labelled for 90 min with 32P, 1 μCi/ml and chilled. They were made into spheroplasts, lysed and fractionated on a sucrose gradient. The 70-s peak was collected and fixed with formaldehyde. Infected spheroplasts were labelled with [3H]uridine (5 μCi/ml, uridine 5 μg/ml) for 10 min, treated with CAP and the lysate fractionated on a sucrose gradient. The ribosome monomers (A), dimers (B), trimers (c) and tetramers (D) were collected, fixed and each mixed with a sample of normal 32P-labelled 70-s ribosomes. CsCI was added and the mixture centrifuged at 35000 rev/minfor 15 hat 10 °C. ○ — ○ 32P; • — •,3H; —, density of CsCI.
View largeDownload slide

Density in CsCI of polysome-associated viral RNA compared with normal ribosomes. Uninfected bacteria were labelled for 90 min with 32P, 1 μCi/ml and chilled. They were made into spheroplasts, lysed and fractionated on a sucrose gradient. The 70-s peak was collected and fixed with formaldehyde. Infected spheroplasts were labelled with [3H]uridine (5 μCi/ml, uridine 5 μg/ml) for 10 min, treated with CAP and the lysate fractionated on a sucrose gradient. The ribosome monomers (A), dimers (B), trimers (c) and tetramers (D) were collected, fixed and each mixed with a sample of normal 32P-labelled 70-s ribosomes. CsCI was added and the mixture centrifuged at 35000 rev/minfor 15 hat 10 °C. ○ — ○ 32P; • — •,3H; —, density of CsCI.

Control CsCl gradients showed that3H-labelled phage RNA added to 70-s ribosomes before fixation did not attach to them but was found in the sediment at the bottom of the centrifuge tube. Intact phage was found near the top of a gradient in which 70-S ribosomes formed a band near the middle at a density of 1·62 g cm-3. The buoyant density of the phage (1·46) was unaffected by treatment with formaldehyde.

In order to determine what proportion (if any) of the ribosomes in each class of polysome was actually associated with viral RNA, cells were labelled in their ribosomal RNA by a long incubation with 32P. They were then infected and the new viral RNA labelled by a 10-min pulse of [3H]uridine. The lysate was centrifuged in a sucrose gradient and the peaks of ribosomes collected as before, fixed and centrifuged separately in CsCl. In the singles (Fig. 3 A) the viral RNA has a higher density than the bulk of the ribosomes. If it is associated with ribosomes, only a minority of them is involved. In contrast, in the twos, threes and fours (Fig. 3B, C, D) the viral and ribosomal RNA have the same density, indicating that all the ribosomes in these peaks are attached to viral RNA. Thus more ribosomes are associated with viral RNA in the dimer form than in any other.

Fig. 3.
Density of viral RNA and ribosomes from the same infected cells. Bacteria were pre-labelled for 1 h with32P (0·5 μCi/ml), incubated a further 30 min in non-radioactive medium, and then infected. Spheroplasts were labelled for 10 min, 30-40 min after infection with [3H]uridine (5 μCi/ml, uridine 5μg/ml), and finally treated with CAP. The lysate was centrifuged in a sucrose gradient and the ribosome monomers (A), dimers (B), trimers (c) and tetramers (D) collected, fixed and centri fuged in CsCl. ○ — ○, 31P; • — •,3H.
View largeDownload slide

Density of viral RNA and ribosomes from the same infected cells. Bacteria were pre-labelled for 1 h with32P (0·5 μCi/ml), incubated a further 30 min in non-radioactive medium, and then infected. Spheroplasts were labelled for 10 min, 30-40 min after infection with [3H]uridine (5 μCi/ml, uridine 5μg/ml), and finally treated with CAP. The lysate was centrifuged in a sucrose gradient and the ribosome monomers (A), dimers (B), trimers (c) and tetramers (D) collected, fixed and centri fuged in CsCl. ○ — ○, 31P; • — •,3H.

Fig. 3.
Density of viral RNA and ribosomes from the same infected cells. Bacteria were pre-labelled for 1 h with32P (0·5 μCi/ml), incubated a further 30 min in non-radioactive medium, and then infected. Spheroplasts were labelled for 10 min, 30-40 min after infection with [3H]uridine (5 μCi/ml, uridine 5μg/ml), and finally treated with CAP. The lysate was centrifuged in a sucrose gradient and the ribosome monomers (A), dimers (B), trimers (c) and tetramers (D) collected, fixed and centri fuged in CsCl. ○ — ○, 31P; • — •,3H.
View largeDownload slide

Density of viral RNA and ribosomes from the same infected cells. Bacteria were pre-labelled for 1 h with32P (0·5 μCi/ml), incubated a further 30 min in non-radioactive medium, and then infected. Spheroplasts were labelled for 10 min, 30-40 min after infection with [3H]uridine (5 μCi/ml, uridine 5μg/ml), and finally treated with CAP. The lysate was centrifuged in a sucrose gradient and the ribosome monomers (A), dimers (B), trimers (c) and tetramers (D) collected, fixed and centri fuged in CsCl. ○ — ○, 31P; • — •,3H.

In the experiments above it was seen that the majority of the single ribosomes of uninfected slowly cooled cells (Fig. 2) or of infected actinomycin-treated sphero-plasts (Fig. 3 A) had a considerably lower density than ribosomes bearing phage RNA. It seemed likely that the former lacked mRNA. Attempts were made to distinguish RNA-ribosome complexes in uninfected cells by their higher density.

  • In the hope of labelling preferentially mRNA (Friesen, 1968), cells were given a short (30-s) pulse of [3H]uridine. Single ribosomes and pairs were then isolated from a sucrose gradient and fixed and their associated tritium compared in a CsCl gradient with single ribosomes obtained from slowly cooled cells (Fig. 4). The pulse-labelled RNA had an average density about one gradient fraction higher, a difference of approximately 0·005 gcm-3. This contrasts with the experiment of Fig. 2 A and B in which the viral mRNA was 9 and 6 fractions respectively heavier than the single ribosomes.

  • Polysomal ribosomes were compared with 70-S ribosomes from spheroplasts that had been incubated with actinomycin with the object of depleting them of mRNA. Fig. 5 shows that ribosomes in groups of 1, 2, 3 and 4 were all heavier by between one and two fractions than actinomycin-treated single ribosomes. This treatment caused the disappearance of polysomes and the accumulation of 70-s ribosomes.

  • The effect of pancreatic ribonuclease (RNase) on ribosomal density was measured and compared with the effect on ribosomes bearing viral RNA. The pairs of ribosomes from normal cells and from infected spheroplasts, labelled with3H in their ribosomal RNA, were collected and treated with RNase before fixing with formaldehyde. Formaldehyde inactivated the RNase so that32P-labelled untreated dimers could be safely added for comparison in a CsCl gradient. The infected ribosomes (Fig. 6A) were reduced in density by 6-7 fractions from 1·66 to 1·62, the same as the difference seen in Fig. 2B. The ribosome pairs from uninfected cells, however, were shifted only one fraction or less to lower density by RNase. This is equivalent to a decrease from 1·62 to 1·615 g cm-3.

Fig. 4.
Density of pulse-labelled RNA from normal cells. Normal 70-s ribosomes labelled with32P were isolated as in Fig. 2. A separate culture of bacteria was labelled for 30 s with [3H]uridine, 2μCi/ml, and then poured into an iced flask containing CAP to make 100μg/ml. Spheroplasts were made and lysed and ribosome monomers (A) and dimers (B) were collected from a sucrose gradient, fixed and mixed with 32P-labelled ribosomes before centrifuging in CsCl. ○ — ○,32P; • — •, 3H.
View largeDownload slide

Density of pulse-labelled RNA from normal cells. Normal 70-s ribosomes labelled with32P were isolated as in Fig. 2. A separate culture of bacteria was labelled for 30 s with [3H]uridine, 2μCi/ml, and then poured into an iced flask containing CAP to make 100μg/ml. Spheroplasts were made and lysed and ribosome monomers (A) and dimers (B) were collected from a sucrose gradient, fixed and mixed with 32P-labelled ribosomes before centrifuging in CsCl. ○ — ○,32P; • — •, 3H.

Fig. 4.
Density of pulse-labelled RNA from normal cells. Normal 70-s ribosomes labelled with32P were isolated as in Fig. 2. A separate culture of bacteria was labelled for 30 s with [3H]uridine, 2μCi/ml, and then poured into an iced flask containing CAP to make 100μg/ml. Spheroplasts were made and lysed and ribosome monomers (A) and dimers (B) were collected from a sucrose gradient, fixed and mixed with 32P-labelled ribosomes before centrifuging in CsCl. ○ — ○,32P; • — •, 3H.
View largeDownload slide

Density of pulse-labelled RNA from normal cells. Normal 70-s ribosomes labelled with32P were isolated as in Fig. 2. A separate culture of bacteria was labelled for 30 s with [3H]uridine, 2μCi/ml, and then poured into an iced flask containing CAP to make 100μg/ml. Spheroplasts were made and lysed and ribosome monomers (A) and dimers (B) were collected from a sucrose gradient, fixed and mixed with 32P-labelled ribosomes before centrifuging in CsCl. ○ — ○,32P; • — •, 3H.

Fig. 5.
Density of normal polysomes compared with actinomycin-treated ribosomes. Separate cultures of bacteria were labelled for 1 h with3IP (0·5 μCi/ml) or [3HJuridine (1 μCi/ml, uridine 2·5 μg/m\). Both were converted to spheroplasts and incubated for 30 min with (31P) or without C’H) actinomycin 0·5 μg/ml. CAP (100 μg/ml) was added to each for the final 4 min before chilling. The lysates were fractionated in sucrose gradients. After fixing with formaldehyde a sample of actinomycin-treated 70-s ribo somes was mixed with3H-labelled monomers (A), dimere (B), trimers (c) or tetramers (D) and centrifuged in CsCl. ○ — ○ 32P; • — •, 3H.
View largeDownload slide

Density of normal polysomes compared with actinomycin-treated ribosomes. Separate cultures of bacteria were labelled for 1 h with3IP (0·5 μCi/ml) or [3HJuridine (1 μCi/ml, uridine 2·5 μg/m\). Both were converted to spheroplasts and incubated for 30 min with (31P) or without C’H) actinomycin 0·5 μg/ml. CAP (100 μg/ml) was added to each for the final 4 min before chilling. The lysates were fractionated in sucrose gradients. After fixing with formaldehyde a sample of actinomycin-treated 70-s ribo somes was mixed with3H-labelled monomers (A), dimere (B), trimers (c) or tetramers (D) and centrifuged in CsCl. ○ — ○ 32P; • — •, 3H.

Fig. 5.
Density of normal polysomes compared with actinomycin-treated ribosomes. Separate cultures of bacteria were labelled for 1 h with3IP (0·5 μCi/ml) or [3HJuridine (1 μCi/ml, uridine 2·5 μg/m\). Both were converted to spheroplasts and incubated for 30 min with (31P) or without C’H) actinomycin 0·5 μg/ml. CAP (100 μg/ml) was added to each for the final 4 min before chilling. The lysates were fractionated in sucrose gradients. After fixing with formaldehyde a sample of actinomycin-treated 70-s ribo somes was mixed with3H-labelled monomers (A), dimere (B), trimers (c) or tetramers (D) and centrifuged in CsCl. ○ — ○ 32P; • — •, 3H.
View largeDownload slide

Density of normal polysomes compared with actinomycin-treated ribosomes. Separate cultures of bacteria were labelled for 1 h with3IP (0·5 μCi/ml) or [3HJuridine (1 μCi/ml, uridine 2·5 μg/m\). Both were converted to spheroplasts and incubated for 30 min with (31P) or without C’H) actinomycin 0·5 μg/ml. CAP (100 μg/ml) was added to each for the final 4 min before chilling. The lysates were fractionated in sucrose gradients. After fixing with formaldehyde a sample of actinomycin-treated 70-s ribo somes was mixed with3H-labelled monomers (A), dimere (B), trimers (c) or tetramers (D) and centrifuged in CsCl. ○ — ○ 32P; • — •, 3H.

Fig. 6.
Effect of RNase on the density of infected and normal ribosome dimers. Separate cultures of bacteria were labelled for 1 h with either32P or [3HJuridine. Half of each culture was treated with CAP for 4 min, made into spheroplasts and lysed. The other half of each was infected and lysed at 40 min after infection with the usual actinomycin and CAP treatment. Ribosome dimers were isolated from the 4 lysates by sucrose gradient cencrifugation. The3H-labelled dimers were treated with RNase, 0·1 μg/ml for 10 min at 20°C, before fixing with formaldehyde. They were then mixed with the corresponding 32P-labelled untreated dimers and centrifuged in CsCl. A, Infected, B, uninfected; c, uninfected, from a different experiment. ○ — ○,32P; • — •3H.
View largeDownload slide

Effect of RNase on the density of infected and normal ribosome dimers. Separate cultures of bacteria were labelled for 1 h with either32P or [3HJuridine. Half of each culture was treated with CAP for 4 min, made into spheroplasts and lysed. The other half of each was infected and lysed at 40 min after infection with the usual actinomycin and CAP treatment. Ribosome dimers were isolated from the 4 lysates by sucrose gradient cencrifugation. The3H-labelled dimers were treated with RNase, 0·1 μg/ml for 10 min at 20°C, before fixing with formaldehyde. They were then mixed with the corresponding 32P-labelled untreated dimers and centrifuged in CsCl. A, Infected, B, uninfected; c, uninfected, from a different experiment. ○ — ○,32P; • — •3H.

Fig. 6.
Effect of RNase on the density of infected and normal ribosome dimers. Separate cultures of bacteria were labelled for 1 h with either32P or [3HJuridine. Half of each culture was treated with CAP for 4 min, made into spheroplasts and lysed. The other half of each was infected and lysed at 40 min after infection with the usual actinomycin and CAP treatment. Ribosome dimers were isolated from the 4 lysates by sucrose gradient cencrifugation. The3H-labelled dimers were treated with RNase, 0·1 μg/ml for 10 min at 20°C, before fixing with formaldehyde. They were then mixed with the corresponding 32P-labelled untreated dimers and centrifuged in CsCl. A, Infected, B, uninfected; c, uninfected, from a different experiment. ○ — ○,32P; • — •3H.
View largeDownload slide

Effect of RNase on the density of infected and normal ribosome dimers. Separate cultures of bacteria were labelled for 1 h with either32P or [3HJuridine. Half of each culture was treated with CAP for 4 min, made into spheroplasts and lysed. The other half of each was infected and lysed at 40 min after infection with the usual actinomycin and CAP treatment. Ribosome dimers were isolated from the 4 lysates by sucrose gradient cencrifugation. The3H-labelled dimers were treated with RNase, 0·1 μg/ml for 10 min at 20°C, before fixing with formaldehyde. They were then mixed with the corresponding 32P-labelled untreated dimers and centrifuged in CsCl. A, Infected, B, uninfected; c, uninfected, from a different experiment. ○ — ○,32P; • — •3H.

An estimate can be made of the expected density of a ribosome with an attached molecule of viral RNA, based on the following assumptions: the molecular weight of a 70-s ribosome is 2·7 × 109 Daltons (Petermann, 1964) and its density 1·62 g cm-3 (observed here); the molecular weight of R17 RNA is 1·1 × 106 (Gesteland & Boedtker, 1964) and its buoyant density in CsCl 1·9 gcm-3 (1·62 in Cs2SO4 (Erikson, 1966) +0-28, the density difference for DNA in CsCl and Cs2SO4 (Sober, 1968)). Calculation of the mean density gives the values shown in Table 1.

Table 1.

Mean density values of ribosomes vrith an attached molecule of viral RNA

Mean density values of ribosomes vrith an attached molecule of viral RNA
Mean density values of ribosomes vrith an attached molecule of viral RNA
View Large

The observed buoyant densities of the ribosomes bearing viral RNA (Fig. 2) are in good agreement with these estimates, and so support the idea that each group of ribosomes is associated with a single strand of viral RNA. Godson (1968) detected more infective RNA in a phenol extract of the 80-s region than could be accounted for by the presence of intact phage. He suggested that the excess may be in unfinished phage particles or attached to single ribosomes. The density of the viral RNA in this region (1·68, that of phage being 1·46 g cm-3), as well as its association with a protein-synthesizing structure (Fenwick, 1971), favour the latter explanation.

The situation is different in HeLa cells infected with poliovirus. Although the viral RNA is only a little more than twice the size of that of phage R17 (Tannock,Gibbs & Cooper, 1970), it is found associated with 35 or more ribosomes (Summers, Maizel & Darnell, 1967). The buoyant density of these large polysomes is no higher than that of normal single ribosomes (Baltimore & Huang, 1968; Huang & Baltimore, 1970), as expected with such a low ratio of viral RNA to ribosomes.

Measurement of buoyant density is a considerably more sensitive method of detecting the presence of mRNA attached to a ribosome than is measurement of sedi-mentation rate. However, attempts to detect such complexes in uninfected cells have had equivocal results. Differences of the order of 0·005 g cm-3 between supposed mRNA-bearing and mRNA-free ribosomes were observed. The use of actinomycin yielded slightly lighter-than-normal ribosomes. Actinomycin has been used to estimate the rate of decay of mRNA (Yudkin, 1965) and it caused the breakdown of polysomes (Fenwick, 1971), but an effect on ribosome structure cannot be excluded. Pulse-labelled RNA was found in slightly heavier structures, but its identification as mRNA (Friesen, 1968; Sedat, Lyon & Sinsheimer, 1969; Phillips, Hotham-Iglewski & Franklin, 1969) is still tentative. Mild treatment with RNase would seem to be a good method of removing externally bound mRNA without affecting the gross structure of the ribosome, although some simultaneous breakage of the 23-s ribosomal RNA itself is known to occur (Fenwick, 1968).

It can be concluded that the distributions of mRNA are quite different in infected and in uninfected bacteria. In uninfected cells there were few or no polycistronic mRNA molecules comparable in size to phage R17 RNA (which codes for 3 proteins) associated with small polysomes. The buoyant densities indicate that mRNA of not more than about one fifth of the size of phage RNA may be attached to the small polysomes. For example, dimers had a density not more than one gradient fraction (0·005gcm-3) greater than RNase-treated dimers. Their maximum mRNA content was therefore equivalent to 1 molecule of viral RNA per 10 ribosomes or 0·2 viral RNA molecules per pair of ribosomes. The density of the pulse-labelled RNA attached to monomers suggested that it was about one tenth of the size of phage RNA, which is consistent with its low sedimentation rate (Friesen, 1968; Phillips et al. 1969). These observations suggest that in normal cells either small proteins only are made in small polysomes, or alternatively the messages axe not detached from ribosomes by RNase treatment which does remove the viral message.

The able assistance of Barbara Livett and a research grant from the Medical Research Council are gratefully acknowledged.

Baltimore
,
D.
 &
Huang
,
A. S.
 (
1968
).
Isopycnic separation of subcellular components from poliovirus-infected and normal HeLa cells
.
Science, N. Y
.
163
,
572
–
574
.
https://doi.org/10.1126/science.162.3853.572
Google Scholar
Crossref
Search ADS
 
Erikson
,
R. L.
 (
1966
).
Fractionation of viral-specific RNA by cesium sulfate equilibrium density-gradient centrifugation
.
J. molec. Biol
.
18
,
372
–
381
.
https://doi.org/10.1016/S0022-2836(66)80254-1
Google Scholar
Crossref
Search ADS
 
Fenwick
,
M. L.
 (
1968
).
The effect of ribonuclease on polysomes and ribosomes of bacteria and animal cells
.
Biochem. J
.
107
,
481
–
489
.
https://doi.org/10.1042/bj1070481
Google Scholar
Crossref
Search ADS
PubMed
 
Fenwick
,
M. L.
 (
1971
).
On the function and fate of phage progeny RNA in infected bacteria
.
J. CellSci
.
8
,
631
–
647
.
Google Scholar
 
Friesen
,
J. D.
 (
1968
).
A study of the relationship between polyribosomes and messenger RNA in Escherichia coli
.
J. molec. Biol
.
32
,
183
–
200
.
https://doi.org/10.1016/0022-2836(68)90003-X
Google Scholar
Crossref
Search ADS
 
Gesteland
,
R. F.
 &
Boedtker
,
H.
 (
1964
).
Some physical properties of bacteriophage R17 and its RNA
.
J. molec. Biol
.
8
,
496
–
507
.
https://doi.org/10.1016/S0022-2836(64)80007-3
Google Scholar
Crossref
Search ADS
 
Godson
,
N. G.
 (
1968
).
Site of synthesis of viral RNA and phage assembly in MS2-infected Escherichia coli
.
J. molec. Biol
.
34
,
149
–
163
.
https://doi.org/10.1016/0022-2836(68)90241-6
Google Scholar
Crossref
Search ADS
 
Godson
,
N. G.
 &
Sinsheimer
,
R. L.
 (
1967
).
The replication of phage MS2. VI. Interaction between phage RNA and cellular components in MS2-infected Escherichia coli
.
J. molec. Biol
.
23
,
495
–
521
https://doi.org/10.1016/S0022-2836(67)80121-9
Google Scholar
Crossref
Search ADS
 
Huang
,
A. S.
 &
Baltimore
,
D.
 (
1970
).
Initiation of polyribosome formation in poliovirus-infected HeLa cells
.
J. molec. Biol
.
47
,
275
–
291
.
https://doi.org/10.1016/0022-2836(70)90302-5
Google Scholar
Crossref
Search ADS
 
Nathans
,
D.
 (
1965
).
Cell-free protein synthesis directed by coliphage MS2 RNA: synthesis of intact viral coat protein and other products
.
J. molec. Biol
.
13
,
521
–
531
.
https://doi.org/10.1016/S0022-2836(65)80114-0
Google Scholar
Crossref
Search ADS
 
Petermann
,
M.
 (
1964
).
In The Physical and Chemical Properties of Ribosomes
, p.
118
.
Amster dam
:
Elsevier
.
Google Scholar
 
Phillips
,
L. A.
,
Hotham-Iglewski
,
B.
 &
Franklin
,
R. M.
 (
1969
).
Polyribosomes of Escherichia coli. II. Experiments to determine the in vivo distribution of polysomes, ribosomes and ribosomal subunits
.
J. molec. Biol
.
45
,
23
–
38
.
https://doi.org/10.1016/0022-2836(69)90207-1
Google Scholar
Crossref
Search ADS
 
Schlessinger
,
D.
 (
1969
).
Ribosomes: development of some current ideas
.
Bad. Rev
.
33
,
445
–
543
.
Google Scholar
 
Sedat
,
J.
,
Lyon
,
A.
 &
Sinsheimer
,
R. L.
 (
1969
).
Purification of E. coli pulse-labelled RNA by benzoylated DEAE-cellulose chromatography
.
J. molec. Biol
.
44
,
415
–
434
.
https://doi.org/10.1016/0022-2836(69)90370-2
Google Scholar
Crossref
Search ADS
 
Sober
,
H. A.
 (Editor). (
1968
).
In Handbook of Biochemistry
, p.
11
–
12
.
Cleveland
:
Chemical Rubber Co
.
Google Scholar
 
Summers
,
D. F.
,
Maizel
,
J. V.
 &
Darnell
,
J. E.
 (
1967
).
The decrease in size and synthetic activity of poliovirus polysomes late in the infectious cycle
.
Virology
31
,
427
–
435
.
https://doi.org/10.1016/0042-6822(67)90222-X
Google Scholar
Crossref
Search ADS
PubMed
 
Tannock
,
G. A.
,
Gibbs
,
A. J.
 &
Cooper
,
P. D.
 (
1970
).
A re-examination of the molecular weight of poliovirus RNA
.
Biochem. biophys. Res. Commun
.
38
,
298
–
301
.
https://doi.org/10.1016/0006-291X(70)90712-6
Google Scholar
Crossref
Search ADS
PubMed
 
Yudkin
,
M. D.
 (
1965
).
Amino acid incorporation in Bacillus megaterium treated with actinomycin
.
Biochim. biophys. Acta
103
,
705
–
707
.
https://doi.org/10.1016/0005-2787(65)90093-6
Google Scholar
Crossref
Search ADS
PubMed
 
Copyright © 1971 The Company of Biologists Ltd.
1971