In Vitro Transcription of T4 Deoxyribonucleic Acid by Escherichia coli Ribonucleic Acid Polymerase SEQUENTIAL TRANSCRIPTION OF IMMEDIATE EARLY AND DELAYED EARLY CISTRONS IN THE ABSENCE OF THE

Abstract T4 DNA was transcribed in vitro by the DNA-dependent RNA polymerase from Escherichia coli. When the termination factor, rho, was present, all of the RNA made was 2000 nucleotides in length and was completely immediate early. When rho was absent, the entire early region of the T4 chromosome was transcribed as a series of polycistronic messengers, either 4300 nucleotides or 7100 nucleotides long. The product made, in the absence of rho, was 40% immediate early and 60% delayed early. Transcription of the delayed early cistrons, in the absence of rho, was due to continued transcription past the point at which rho causes release. Experiments were done at 18° at which initiation alone was allowed to proceed for several minutes by adding only ATP and GTP. The reaction mixture was then completed and rifampicin was added to inhibit further initiation. RNA synthesis continued for 1400 sec after the brief period of initiation. During the initial 400 sec of reaction, only immediate early sequences were synthesized. Delayed early sequences alone were made at later times. Transcription of the e and rII cistrons was studied. The e cistron was transcribed between 1300 and 1400 sec after initiation. The rII A and B cistrons were transcribed sequentially. Transcription of the rII A cistron occurred between 700 and 1100 sec after initiation but the B cistron was not transcribed until 1100 to 1400 sec of reaction. The e and rII transcripts were parts of RNA molecules that were 7100 nucleotides in length. The rII region was transcribed solely as a polycistronic messenger in vitro but only 20 to 30% of the rII transcripts made in vivo were polycistronic.


SUMMARY
T4 DNA was transcribed in vitro by the DNA-dependent RNA polymerase from Escherichia coli. When the termination factor, rho, was present, all of the RNA made was 2000 nucleotides in length and was completely immediate early. When rho was absent, the entire early region of the T4 chromosome was transcribed as a series of polycistronic messengers, either 4300 nucleotides or 7100 nucleotides long. The product made, in the absence of rho, was 40% immediate early and 60% delayed early. Transcription of the delayed early cistrons, in the absence of rho, was due to continued transcription past the point at which rho causes release. Expertients were done at 18" at which initiation alone was allowed to proceed for several minutes by adding only ATP and GTP.
The reaction mixture was then completed and rifampicin was added to inhibit further initiation. RNA synthesis continued for 1400 set after the brief period of initiation.
During the initial 400 set of reaction, only immediate early sequences were synthesized.
Delayed early sequences alone were made at later times.
Transcription of the e and 7II cistrons was studied.
The e cistron was transcribed between 1300 and 1400 set after initiation. The 71X A and B cistrons were transcribed sequentially. Transcription of the 7ZZ A cistron occurred between 700 and 1100 set after initiation but the B cistron was not transcribed until 1100 to 1400 set of reaction.
The e and rII transcripts were parts of RNA molecules that were 7100 nucleotides in length. The 711 region was transcribed solely as a polycistronic messenger in vitro but only 20 to 30% of the 7IZ transcripts made in viuo were polycistronic.
During the first 10 min of phage infection, two distinguishable families of T4 RNA are synthesized.
The immediate early family * This work was supported by Postdoctoral Fellowship l-F2-GM41,523-01 from the National Institute of General Medical Sciences.
$ Present address, Department of Microbiology, Indiana University School of Medicine, Indianapolis, Indiana 46202. appears during the initial 2 min of infection and does not require the prior synthesis of phage-specific proteins.
The delayed early family requires prior synthesis of phage proteins and does not appear until later than 2 min of infection (1,2).
In vitro studies have shown that only immediate early cistrons are transcribed if the host initiation factor, uB. cozi, andrelease factor, rho, are present (3,4). Early in T4 infection, a new sigma factor appears, uT4, which permits transcription to begin at delayed early cistrons (3,4). The operation of these subunits seemed to explain adequately the control of early T4 RNA synthesis in v&o.
However, evidence has been presented that suggests the existence of another control element instrumental in the in viva transcription of early cistrons.
In the case of T4, the e gene, which is not under the control of an early promotor (5), can be transcribed during the early period of phage infection (5). This premature transcription of the e gene requires prior synthesis of phage protein (6). In the case of phage X, the N gene does not appear to code for a new u factor but the product of this gene allows transcription to extend further into the left arm of X DNA (7). Therefore, it appears as if an antiterminator protein is synthesized soon after phage infection and that this protein somehow obviates the host cell chain termination mechanism.
Since rho cannot mediate release in the case of T4 delayed early cistrons,' the apparent function of the antiterminator is to allow RNA polymerase molecules that have initiated at immediage earg cistrons to transcribe beyond the point at which rho would normally cause release to take place.
Conceptually, at least, in vitro transcription of T4 DNA by E. coli RNA polymerase, in the absence of rho, should mimic the in vivo situation for these enzyme molecules when the antiterminator is present.
To this end, it has been shown that E. coli RNA polymerase, in the absence of rho, transcribes two discrete size classes of T4 RNA in vitro (8). Roughly 70 y0 of the product is 4300 nucleotidee long and the remainder is 7100 nucleotides in length. We now ask to what extent the early region of the T4 chromosome is transcribed when rho is absent.
The data indicate that (a) probably the entire early region of the T4 chromosome is transcribed, (b) immediate early and delayed earZy cistrons are adjacent on the T4 DNA, (c) the order Issue of September 10,1971 H. J. Witmer 5221 of appearance is immediate early + delayed early, and (d) each RNA molecule contains no more than one immediate early transcript but either one or several delayed early transcripts.
In most respects, these results confirm already published data (9-11).
We also demonstrate that the e and rII cistrons are transcribed in the absence of rho. All three cietrons appear to be associated with those RNA molecules that are 7100 nucleotides in length.

MATERIALS AND METHODS
Purification of E. coli DNA-dependent RNA Polymerase and Rhc+-DNA-dependent RNA polymerase (EC 2.7.7.6) was purified to greater than 90% purity by the glycerol gradient method (12). Rho was purified according to the method of Roberts (13) and was estimated to be greater than 80% pure.
PuriJication of DNA-dependent RNA Polymerase from T4injected Cells-E.
coli BB was grown in 7-liter batches to 1 X lo9 cells per ml. The growth medium was 3 x D (14) and the temperature was 30". Phage T4 was added to give a multiplicity of five viruses per cell. Ten minutes after the phage were added, 7 liters of finely crushed frozen 3 x D were added. The cells were harvested by centrifugation.
DNA-dependent RNA polymerase was purified by the glycerol gradient method (12). The final enzyme preparation was judged to be greater than 80% pure. This enzyme had no affinity for T7, T5, and E. coli DNA.
Roughly 40% of the product was immediate early in nature.
All transcription was from the 1 strand, suggesting that lute T4 RNA was not transcribed (15).
Puri$cation of Phage of T4 DNA-Phage T4 D was prepared in IO-liter batches with E. coli BB grown on 3 x D as host. The phage were purified by differential centrifugation. Phage DNA was isolated by the standard phenol procedure (16). DNA used as template was routinely monitored for the presence of single strand breaks. When more than 10% of the strands were broken, the DNA was no longer used as template.
Single strand breaks caused false initiation with the resultant production of many short chained RNAs.~ In Vitro Synthesis of T4 RNA-The reaction mixture contains 5 mu MgC12, 50 mM Tris-Cl (pH 7.9), 50 mM KCl, 0.5 mM [U-14C]ATP, GTP, CTP, and UTP, 25 pg per ml of T4 DNA, 10 pg per ml of enzyme, and, when present, 3 pg per ml of rho. Fractions containing RNA were pooled The RNA was then precipitated overnight at -20" by adding 2 volumes of 95% ethanol.
The RNA solution was stored at -80". Conditions of Phage Infection and Pulse Labeling---E.
coli K-12 was grown to 1 X lo8 cells per ml at 30" in M9 salts-glucose (17). The cells were harvested by centrifugation and resus-Q H. J. Witmer, unpublished data. pended in previously warmed (30") MS-glucose to 1 x 109 cells per ml. T4 was added to give a multiplicity of 5 and 2 min were allowed for phage adsorption.
For pulse labeling of in z&o RNA, [GJH]uridine (20 &i per ml) was added. PuriJication of RNA from Phage-infected Cells-After various times of infection or pulse labeling, the culture (175 to 200 ml) wae podred over an equal volume of finely crushed frozen medium. The cells were centrifuged at 10,000 X g for 5 min and resuspended in protoplasting buffer (50 mM Tris (pH 7.9), 500 mM sucrose, 10 mM EDTA, and 150 pg per ml of bentonitetreated lysozyme).
After 10 min, the lysate was warmed to 37" for 5 min and then chilled to 0". Saturated NaCl (0.5 volume) was added.
The resultant precipitat,e was removed by centrifupation at 10,000 x g for 45 min and discarded. The remainder of the procedtire was identical with the one outlined for in vitro RNA.
Purijkalion of rII and e RNA-Specific messengers were purified to greater than 98% purity by three passages through T4 DNA-nitrocellulose columns (19). DNAs containing deletions rid%' and r6S8 (20,21) were used to purify rII and rIIB RNA, respectively.
Molecular Weight Determination-[14C]T4 RNA was added to a solution of E. coli 4 S, 5 S, 16 S, and 23 S RNA (total unlabeled RNA concentration = 500 pg per ml) and precipitated with ethanol as described above. The precipitate was dissolved to its original concentration in 0.1 M sodium phosphate (pH 7.7), 1.1 M HCHO and heated at 65" for 5 min. Aliquots (0.2 ml) of the formaldehyde-treated RNA were layered onto 5-ml linear sucrose gradients (5 to 20% in 0.1 M sodium phosphate (pH 7.7), 1.1 M HCHO) and centrifuged at 50,000 rpm (SW 65L Ti rotor) at 4". Samples were collected from the bottom of the tube into 1.0 ml of distilled water.
After the absorbance at 260 nm had been determined, the samples were individually precipitated with cold 10% trichloroacetic acid. Molecular weights were calculated with the formula of Boedtker (22).
DNA-RNA Hybridization-Unless otherwise stated, each assay tube contained 5 pg of labeled RNA, varying amounts of unlabeled competitor RNA, and 10 pg of alkali-denatured reneutralized T4 DNA (23). The ionic condition was 5 x SSC.3 The total volume was 0.75 ml.
After incubation at 65" for 5 hours, the contents of each tube were filtered through 13-mm Schleicher and Schuell B6 nitrocellulose filters. The filters were washed with 100 ml of 5 x SSC (50 ml on each side of filter) and incubated for 90 min with boiled pancreatic RNase (30 pg per ml in 2 x SSC), dried, and counted.
For the experiments reported here, unless otherwise noted, the efficiency of hybridization was 200/,. In the absence of competitor RNA, 30,000 cpm of [14C]RNA were retained on the filter and 10,000 cpm of [3H]-labeled in vivo RNA were retained.
a The abbreviations used are: SSC, 150 mM N&I, 15 mM Nar citrate, pH 7.5; 5 X SSC, 2 X SSC, 5 $nd 2 times, respectively, the standard concentration of SSC: GG-RNA. RNA svnthesized in vitro by the glycerol gradient-pkfied enzy'me; GG-;-RNA, RNA synthesized in vitro by the glycerol gradient-purified enzyme in the presence of the release factor rho. After 2 min of infection, all of the T4 RNA sedimented as a narrow band at approximately 11 S. Later in infection (5 and 10 min), the majority of the T4 RNA molecules migrated as a broad band between 15 and 20 S. Of the T4 RNAs made in vitro, GG-RNA sedimented at 25 to 30 S whereas GG-p-RNA moved at 11 S (Fig. lb).
The molecular weights of in vitro RNA were determined and are given in Table I. Judging from the sedimentation coefficient of in vivo RNA vis-h-vis in vitro RNA, it can be estimated that 2-min in vivo RNA was 2000 nucleotides in length and the majority of 5-and IO-min RNA was 2500 to 3500 nucleotides in length. Therefore, the bulk of the T4 RNA present during the early period of phage infection was considerably shorter than GG-RNA.

Effect of Rho on Regions of T4 Chrmosorne Transcribed in
Vi&- Fig.  2A shows hybridization competition experiments between GG-p-RNA and GG-RNA. Unlabeled GG-RNA competed completely with labeled GG-p-RNA but the reciprocal experiment showed that unlabeled GG-p-RNA competed only 40 to 41% with labeled GG-RNA.
Consequently, the absence of rho allowed transcription of much larger regions of the T4 chromosome. Fig. 2B gives hybridization experiments between GG-p-RNA and 15 S-GG-RNA and 18 S-GG-RNA.
The two size classes of GG-RNA competed nearly 100% with GG-p-RNA.
However, GG-p-RNA competed with 15 S-and 18 S-GG-RNA only 46 and 27%, respectively. Fig. 2C shows that 15 S-and 18 S-GG- The RNAs all were treated with 0.1 N NaOH at 20" until their average size was the same as GG-p-RNA.
RNA cross-competed only to the extent that GG-p-RNA competed with either size class. As shown below, GG-p-RNA is equivalent to immediate early RNA.
Therefore, 15 S-and 18 S-GG-RNA molecules may share only immediate early transcripts.
To obtain an idea of when in the phage infection process RNA molecules were made that were identical with either GG-p-RNA or GG-RNA, phage-infected cells were pulse-labeled at various times and the RNA was isolated.
Hybridization competition experiments were then performed between the pulse-labeled in uivo RNA and unlabeled in vitro RNA.
Sequences identical with GG-p-RNA were the only ones made in vivo during the initial 2 min of infection.
Later, these sequences constituted progressively less of the total T4 RNA synthesized in vivo. Little detectable synthesis of GG-p-RNAequivalent sequences was observed after 10 min of infection (Fig. 3).
On the other hand, GG-RNA competed completely with all of the T4 RNA synthesized in vivo during the initial 10 min of infection.
Detectable synthesis of GG-RNA-equivalent sequences was still observed 35 to 40 min after infection (Fig. 3).
The fact that both in vitro RNAs competed with T4 RNA made during the first 2 min of infection suggested that both contained all of the immediate early transcripts.
To confirm this, immediate early T4 RNA was isolated from cells 5 min after infection in the presence of chloramphenicol (1,2). The data in Table II clearly show that both GG-p-RNA and GG-RNA totally competed with immediate early RNA.
Whereas immediate early RNA competed nearly 100% with GG-p-RNA, it competed only 41 y. with GG-RNA.
Consequently, even though GG-RNA contained all of the immediate early transcripts, these constituted but a portion of the total transcripts present in GG-RNA.
The data in Table II indicate that most of the remaining transcripts present in GG-RNA were delayed early.

In Vitro Time of Appearance of Immediate Early and Delayed
Early RNA-Most likely, the transcription of delayed early cistrons in vitro, in the absence of rho was due to continued transcription beyond the point at which rho would normally cause release. If such were the case, one would expect a time lag between the initiation of transcription and the appearance of delayed early sequences. To illustrate this, experiments were done in which imitation alone was allowed to proceed by incubating the DNA and enzyme with only ATP and GTP.
After several minutes, [W]ATP, GTP, and UTP were added to complete the reaction mixture along with rifampicin to inhibit further initiation (26)(27)(28)(29). The RNA was extracted at frequent intervals and tested for the presence of immediate early and delayed early RNA sequences. The RNA was purified, formaldehyde-treated, and centrifuged.
These controls showed that short polynucleotide chains, 50 to 100 nucleotides in length, were made during the initiation period.
Generally, stocks of nucleoside triphosphates, as supplied by the manufacturer, are cross-contaminated. This probably explains the slight amount of propagation that occurred. tion were observed.
For the first 900 set, the rate of incorporation was 44.4 pmoles per ml per see; after 900 set, the rate changed abruptly to 15.0 pmoles per ml per sec. The reaction stopped at 1400 sec. Fig. 5 shows that 70% of the RNA chains stopped growing after they were 4300 nucleotides long but the remainder continued to grow until a size of 7100 nucleotides was achieved.
This probably explains the change in linear rate of incorporation observed in Fig. 4. The change in sedimentation coefficient with time indicated a chain elongation rate of five nucleotides per sec.
Hybridization competition experiments (Fig. 6) showed that only immediate early RNA sequences were present during the initial 400 set of reaction.
Past 400 set, delayed early sequences appeared.
When the reaction stopped, 59% of the product was delayed early and 41% was immediate early. Considering the elongation rate and the time of appearance of delayed early sequences, it was estimated that delayed early cistrons were no more than 2000 nucleotides from an immediate early promotor.
In Vitro Transcription of rII and e Cistrons-One of the lines of evidence pointing to the existence of the antiterminator protein is the early transcription of the e cistron (6), a late function (5). Therefore, transcription of the e cistron was expected under the present in vitro conditions, provided that transcription in the absence of rho truly reflected the in vivo situation when rho is inhibited by the antiterminator.
Since the rII cistrons are delayed early, their transcription was also expected since GG-RNA appeared to contain transcripts of all of the delayed early cistrons. The hybrid was dissociated by heating at 85" for 2 min. Three-milliliter portions of the heated solution are layered onto 30-ml linear sucrose gradients (10 to 30% in 0.1 M sodium phosphate (pH 7.7), 1.1 M HCHO) and centrifuged at 20,000 rpm (SW 25.1 rotor) for 20 hours at 4". The RNA-containing fractions were then pooled and dialyzed extensively against 10 mM Tris-Cl (pH 7.9), 10 mM KC1 to remove the formaldehyde.
This RNA had no detectable competing power against e RNA but competed normally with immediate early RNA and rZZ RNA.
This RNA was then hybridized to T4+ DNA.
The hybrid was treated with mitomycin C and NaBH4 to cross-link the DNA and RNA strands (31). The time sequence of e gene transcription is shown in Fig. 6. The e cistron was not transcribed until 1300 to 1400 set after initiation.
Therefore, the e cistron transcript seemingly constituted the last 500 nucleotides of an 18 S-GG-RNA molecule. Transcription of rII region began at 700 set and continued until the end of the reaction (Fig. 7). Between 700 and 1100 set, all of the rII RNA belonged to the A cistron.
The B cistron was not transcribed until later than 1100 sec. The relative amount of rIIA RNA decreased after 1100 set, implying that the rIIA cistron was not transcribed after that time.
The results also indicated that the A and B cistrons were transcribed sequentially.
From these data, it was estimated that the entire rII region was 3500 nucleotides long. The A cistron was 2000 nucleotides in length and the B cistron was 1500 nucleotides long. These estimates agreed reasonably well with other estimates (6,32).
Like the e cistron transcript, rII RNA apparently was associated with an 18 S-GG-RNA molecule. To confirm the assignments of rII and e RNAs to 18  0, rZZA; @, raze%; l , rZZ. The RNAs were all degraded to the size of e RNA by mild alkali treatment. (Fig. 8). The results show that only 18 S-GG-RNA competed with either rII or e RNAs.
Size Distribution of e RNA and I-II RNA Synthesized in Vitro and in Viv+When T4 RNA was hybridized with eG19 DNA, the remaining RNA sedimented as a single homogeneous boundary at 5 S (Fig. 9). This was equivalent to a molecular weight of 1.75 x lo5 (500 nucleotides) and agreed well with the size estimated from the data in Fig. 6.  I  I  I  I  I  I  I  I  III  I  II  TOP  8  16  24  32  40  48  56  64  72  80  88  96  FRACTION NO. FIG. 10. Zone centrifugation of formaldehyde-treated rZZ RNA. In vitro T4 RNA (0) and in vivo T4 RNA (0) after hybridization with rl,$?r$ DNA; in vitro T4 RNA (X) and in vivo T4 RNA (c)) after hybridization with r698. Centrifugation was for 5 hours. Each tube received 6000 to 8500 cpm of radioactivity.
The recovery was 88%.
On the other hand, rII RNA showed a more complicated picture.
Peaks I and II each represented 35 to 40% of the total rII RNA while Peak III constituted 20 to 30%.
Only Peak III was observed if in vitro T4 RNA was hybridized with rf 272 DNA. Peak I alone was obtained when either in viva or in vitro T4 RNA is hybridized with r6% DNA.
These data, plus the experiments shown in Table III, allowed the following assignments to be made. Peaks I and II were monocistronic transcripts of the rZIB and rIZA cistrons, respectively.
Peak III contained transcripts of the entire rII region in which the A and B cistrons were present in equimolar quantities.
The size estimates of the rIZA cistron obtained by zone centrifugation and the data in Fig. 7 were identical. However, zone centrifugation provided a slightly lower estimated size for the r1IB cistron and the rll region than did the data in Fig. 7  Peaks I, II, and III are defined in Fig. 10. After formaldehyde treatment, 3.0-ml aliquots. of rZZ RNA were layered onto 30.ml sucrose gradients (10 to 30y0 in 0.1 M sodium phosphate (pH 7.7), 1.1 M HCHO) and centrifuged at 25,000 rpm at 4" for 20 hours. The peak fractions were pooled separately and dialyzed into 10 mM Tris-Cl (pH 7.9), 10 mM KC1 to remove the formaldehyde. RNA designated rZZ or rZZB was purified as described under "Materials and Methods." Unlabeled competitor RNA was present at 2.0 mg. The values are averages of five determinations. reproducible to within 1 unit, the differences should not be taken too seriously.

DISCUSSION
The data presented in this paper show that essentially the entire early region of the T4 chromosome was accessible to DNA-dependent RNA polymerase molecules that initiated at immediate early sites in the absence of rho. Since GG-RNA actually contained two discrete size classes of RNA, the early region was transcribed as a series of polycistronic messengers either 4300 or 7100 nucleotides long. These two size classes seemingly contained the same immediate early sequences but different delayed early sequences. The observed value was 40 to 41% so it was concluded that each GG-RNA molecule contained no more than one immediate early transcript. An estimate of the number of delayed early transcripts per GG-RNA molecule was obtained by assuming that all delayed early cistrons were comparable in size to the rII cistrons.
With this assumption in mind, each 15 S GG-RNA molecule contained only one to two delayed early transcripts and each 18 S-GG-RNA molecule had three to four delayed early transcripts.
The early genes of T4 occupy nearly 25% of the genetic map (32,33). Assuming that they occupy a similar proportion of the T4 DNA, as many as 54,000 base pairs could be devoted to the early genes. Considering the relative contributions of 15 S and 18 S species in GG-RNA, seven 15 S-GG-RNA molecules and three 18 S-GG-RNA molecules would be required to transcribe the entire early region.
With the estimates made above for the number of delayed early and immediate early transcripts per GG-RNA molecule, the early region of T4 DNA could contain as many as 10 immediate early cistrons and 16 to 26 delayed early cistrons. Should an antiterminator mechanism account for a sizeable proportion of delayed early transcription, at some time during infection, such a high number of immediate early cistrons relative to the number of delayed early cistrons may be necessary.
Obviously, these numbers must be viewed as tentative until gene titration studies have been done. We are currently engaged in such studies.
Milanesi et al. (11) reported that no more than 38% of their in vitro product was delayed early. They suggested that some of their enzyme molecules traversed more than 1 unit of transcription and re-entered immediate early cistrons. Our data were inconsistent with such a mechanism. It would appear that none of the enzyme molecules re-entered immediate early cistrons under the experimental conditions used here. The above-mentioned authors also suggested that immediate early and delayed early cistrons are interdigitated on the T4 chromosome. Our estimate of the numbers of immediate early and delayed early cistrons would allow one to speculate that generally immediate early cistrons are separated by no more than two delayed early cistrons.
It is not definitely known whether product release occurs in viva while the termination mechanism of the host is nullified by the antiterminator.
During T4 infection, only a portion of the rII transcripts were polycistronic ( Fig. 10 and Reference 6). Therefore, it would appear as if termination can occur at the junction separating the A and B cistrons.
Since the rZI region was transcribed only as a polycistronic messenger in vitro, perhaps a phage release mechanism is established during T4 infection. Product release can occur in vitro in the absence of p (8, 34), so release points do exist on the T4 chromosome that do not require the action of any subunit not present in the core enzyme (35,36). Perhaps similar "passive" release points function in vivo. Since the host core enzyme did not release at the junction between the rII A and B cistrons, perhaps the known modifications of the core enzyme upon T4 infection (37) are associated with recognition of different "passive" release points.
It would be of interest to examine in vitro transcription of T4 DNA by using the T4 core enzyme.
It is not known what relative contributions uT4 and the antiterminator make to delayed early transcription. The data in Fig. 1 revealed no in z&o REAs that were the same size as GG-RNA.
There are three plausible explanations for this. (a) The antiterminator mechanism actually accounts for little delayed early transcription.
(6) A phage-specific termination mechanism is established soon after infection.
(c) RNA molecules the same size as GG-RNA cannot be made in viva because degradation of these polycistronic messengers commences prior to their completion.
One of the predictions of the antiterminator mechanism is that the resultant polycistronic messengers be translatable into functional proteins.
It will be of interest to examine the in vitro translation of GG-p-RNA and GG-RNA in systems known to give functional products (38, 39).