THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 248, No. 15, Issue of August 10, pp. 5502-5511, 1973

Printedin U.S.A.

Bacteriophage T4 Gene Expression

EVIDENCE FOR TWO CLASSES OF PREREPLICATIVE CISTRONS*

(Received for publication, December 22, 1972)

PATRICIA 2. O’FARRELL AND LAWRENCE M. GOLD

From the Department of Molecular, Cellular and Developmental Biology, University of Colorado, Boulder, Colorado 80302

SUMMARY

The antibiotic rifampicin has been used in an investigation of the temporal regulation of gene expression during bac- teriophage T4 infection of Escherichia coli. We have asked at what times after infection the transcription of specific cistrons becomes rifampicin insensitive. Gene expression has been studied using analyses of proteins on sodium do- decyl sulfate-acrylamide gels; thus translation has been used to assay for the appearance of specific transcripts.

Two types of promoters account for most prereplicative transcription; early promoters are recognized immediately after infection, whereas quasi-late promoters are recognized only after a delay of 14 min. Those transcripts previously defined as immediate early and delayed early RNAs initially appear to be under the control of early promoters; however, the quasi-late promoters may provide a second mode of regu- lation for several of those transcripts.

Infection of Bscherichia coli with the bacteriophage T4 is marked by the orderly appearance of specific “classes” of phage- specific RNA (l-4). These classes were defined originally on the basis of the times at which specific transcripts first appeared rela- tive to phage DNA replication; thus pre- and postreplicative RNAs were designated “early” and “late” (l-3). Recently the prereplicative RNAs have been reclassified using criteria in addi- tion to kinetics. Those prereplicative RiKAs which are synthe- sized immediately after infection, even in the complete absence of protein synthesis, are called immediate early RNAs (5) ; those prereplicative RNBs which first appear after a lag of about 2 min and whose synthesis apparently requires protein synthesis are called delayed early RNAs (5). A third prereplicative class, called quasi-late, is defined by low abundance prior to DNA syn- thesis and a high relative abundance at postreplicative times (5). All of the prereplicative RNA is synthesized asymmetrically from the same strand of the T4 DNA (6).

Owing to the striking inhibition of delayed early RNA accumu- lation brought about by chloramphenicol or puromycin (4, 5),

* This investigation was supported by Grant E-624 from the American Cancer Society and by Grant GB-30517 from the Na- tional Science Foundation.

much research has been focused on the transition between imme- diate early and delayed early transcription. In highly purified cell-free systems, T4 DNA may be transcribed by the E. coli RNA polymerase to yield, in the proper temporal order, both immedi- ate early and delayed early RNAs (7, 8); promoter recognition occurs at sites adjacent to the immediate early genes (8). I f the E. coli transcriptional termination factor p is added to these sys- tems, delayed early RNAs are not synthesized (9, 10). These data have led to two alternative (but not mutually exclusive) models for the onset of delayed early transcription in viva. An antiterminator model proposes that the product of an immediate early gene abolishes p factor activity, thereby allowing RNA polymerase to continue to the end of each early transcription unit (5, 10, II). In a model utilizing new promoter recognition, p mediated terminations persist, but the product of an immediate early gene allows initiation of RNA synthesis beyond the p termination site (12-15). In either model the immediate early gene product which allows delayed early transcription is thought to be a protein, since chloramphenicol and puromycin restrict transcription to immediate early genes (4, 5). Subsequently, Brody et al. (11) suggested that most delayed early RNAs arise on polycistronic messages containing immediate early RNA; their data exclude the new promoter model as the primary mechanism for delayed early transcription. We argued also that delayed early RNAs arise in vivo as the promoter distal portions of poly- cistronic messengers (16-19). However, we believe that no T4- encoded protein is required or utilized in the transcription of most delayed early genes, since a coupled cell-free system derived from uninfected E. coli transcribes and translates both immediate and delayed early genes (17-19). A more complete argument in sup- port of this view is developed in the accompanying paper (19).

Two pieces of evidence suggest, that prereplicative RNA syn- thesis may occur under the control of promoters other than those recognized immediately after infection. Firstly, rIIB transcrip- tion starts at about 2 min after infection at 30”; this rIIB-specific message is at the 5’ end of an RNA chain (15, 20). Secondly, a large number of reports document alterations which occur in the host RNA polymerase shortly after T4 infection (21-25); Travers (12-14) and Hager et al. (24) have suggested that these changes may confer altered initiation capacities on the modified enzyme.

In order to investigate the control of prereplicative gene ex- pression, we have utilized the antibiotic rifampicin (26). Rif- ampicin binds to the p subunit of the E. coli RNA polymerase and thereby prevents the initiation of transcription (27). RNA

by guest on July 18, 2018http://w

ww

.jbc.org/D

ownloaded from

chain elongation is not affected by rifampicin (28). Since RNA synthesis in T4-infected E. coli is sensitive to rifampicin through- out the lytic cycle (29), we may ask at which times various classes of T4 transcripts become rifampicin refractory.

In the experiments reported below, we identify specific tran- scripts by analyzing their protein products on SDSJacrylamide gels (30, 31). We will discuss the validity of this approach be- low. Our experiments suggest that prereplicative RNAs may be subdivided into two classes : earlya and quasi-late. The quasi- late RNAs include those specific transcripts which first become rifampicin resistant 134 min after infection. The early RNAs include most of the immediate early and delayed early RNAs defined previously on the basis of experiments with the antibiotic chloramphenicol (5). Among the quasi-late genes are several which initially are controlled byearlypromoters; thusnew promo- ter recognition for some genes provides a second mode of trans- cription.

MATERIALS AND METHODS

Preparation of Infected Cell Extracts-E. coli AS19 (32) and bacteriophage T4D were used for all of the experiments reported except where otherwise noted in the text. E. coli AS19 is a B strain with an increased permeability to rifampicin.3 Infections were carried out in Xl9 media plus tryptophan at 30” (2, 31). Cells were irradiated with ultraviolet light prior to every infec- tion (31). Control experiments indicated that the burst is re- duced in irradiated cells; however, neither the temporal order of transcription and translation nor the relative yield of specific proteins is affected by the irradiation.4 Phage were added in one-fifth the final volume to appropriately concentrated cells such that the final concentration was 3 X lo* cells per ml, in- fected at a multiplicity of eight. 13etween 1 and 5 ml of in- fected cells were used for various experiments as described in the figure legends.

Mixed 14C-amino acids were used to label proteins under vari- ous conditions. We used sufficient levels of 14C-amino acids and, when necessary, unlabeled casamino acids, to ensure that radio- active amino acids were not limiting throughout the labeling period. Infected cultures were stopped by pouring the cells onto ice (in the presence or absence casamino acids). The preparation of radioactive samples for electrophoresis is described in the ac- companying paper (31). The radioactive proteins in sample buffer for electrophoresis represent a IO-fold increase in final cell concentration. The volume of sample used for electrophoresis was 25 ~1. Thus all figures containing gel patterns represent equal volumes of cells. Amino acid incorporation into protein was determined by removing an aliquot from the samples and measuring acid-prccipitable radioactivity (19).

SDS-Acrylamide Gel Electrophoresis-The procedure for elec- trophoresis is described in the accompanying paper (30, 31). The percentage of acrylamide used in the separating gel was 10% for all experiments except those shown in Fig. 5 and 8 in which the gel was 7.5y0 acrylamide. All of the figures of gels are auto- radiograms prepared as described (31).

1 The abbreviation used is : SDS, sodium dodecyl sulfate. 2 Early transcriotion units are bounded Pr + TR. where Pr is a

promoter recognized by the host E. coli lLi?A poiimerase T7, 8) and TE is a termination site recognized directly by RNA polymer- ase (10). The promoter proximal portions of early transcription units encode immediate early RNAe, whereas the promoter distal portions encode delayed early RNAs (7,8). Such early transcrip- tion units are arranged in tandem on the T4 genome (8).

3 D. Oppenheim, unpublished experiments. 4 Unpublished results from this laboratory.

Antibiotic Experiments-The experiments using the antibiotic rifampicin were performed at a final concentration of 200 pg per ml. This concentration was sufficient to prevent any T4-specific protein synthesis if added to cells at zero time along with phage. Chloramphenicol was used at a final concentration of 250 pg per ml for the experiment reported in Fig. 6.

Chemicals-Rifampicin (rimactane) was a generous gift from CBA Pharmaceutical Company. Mixed K-amino acids were obtained from Schwarz-Mann (No. 3122-09). [methyL3H]Thy- midine was obtained from New England Nuclear (No. NET- 027X).

Phage Mutants-T4D and the T4 mutants in genes 32,44, and rIII3 were from the collection of R. H. Epstein. Other mutants aTere as described in the accompanying paper (31).

RESULTS

Temporal Patterns of T&peciJic Protein Synthesis-We first asked in what order T4 genes are transcribed and their mRNAs translated after infection. Bacteria were exposed to ultraviolet light to preclude host RNA and protein synthesis (33, 34). Fol- lowing infection, we pulse-labeled the phage proteins with I%- amino acids using I-min pulses at every minute for the first 22 min of infection. Amino acid incorporation was measured and the radioactive proteins were analyzed on SDS-acrylamide gels (Fig. 1, A and B). These experiments arc different from similar ones published by Hosoda and Levinthal (33) only in that we possess a larger category of band assignments (31) and we utilize an electrophoresis system with increased resolution (30). We have diagrammed the times at which specific proteins first appear and the times at which those proteins are no longer synthesized (Fig. 1C). These patterns do not reveal a small number of protein classes; rather, the first 5 to 6 min of infection show the continuous appearance of new species. Furthermore, the time at which a protein first appears does not prejudice the time at which its messenger ceases to be translated. Under our conditions, T4 DNA synthesis is first detectable 7 min after infection. Those proteins which appear before that time, by definition, are products of prereplicative genes. During the interval between the onset of DNA replication at 7 min and the onset of late translation at about 12 min, no new proteins appear. During the first 2 min of infection fewer than 12 proteins are strongly labeled; we assume that those proteins are encoded for by genes which are adjacent to promoters recognized immediately after infection (YE) and are therefore products of immediate early RNAs (5). These assignments are rather premature due to considerations of the lengths of specific genes. We will present in a separate manuscript our detailed attempts at pro- moter mapping.

EJect of Rifampicin on Prereplicative Transcription and Translation-E. coli AS19 was infected with T4 at time zero; at 1 min post-infection, one-half of the culture was transferred into rifampicin. The accumulation of radioactive proteins in each culture was measured (Fig. 2A). Rifampicin addition 1 min after infection decreased the apparent rate of protein syn- thesis to about half the rate measured in the control culture; furthermore, after about 16 min the rifampicin-treated culture no longer accumulated additional radioactive proteins. Cells treated with rifampicin 1 rnin after infection were also pulse- labeled at various times after infection with 14C-amino acids (Fig. 2B). The apparent rate of protein synthesis increased during the first 9 min of infection and subsequently decayed; the decay iu protein-synthesizing capacity occurred slowly in this experiment and may not represent the rate of messenger

by guest on July 18, 2018http://w

ww

.jbc.org/D

ownloaded from

Time After Infection (minutes) Time of Synthesis ( minutes 1

Pre- Replicoti\ le_ Genes

43 rIIA

46 39 52

32 me

42 45’

Internal/ Proteina

Post -Replicative i 34 Genes I.17

FIG. 1. Kinetics of T4-specific protein synthesis. Escherichia cells (0) are shown. coli AN9 were irradiated with ultraviolet light and infected with

Extracts of each sample were prepared for

T4 at time zero. Aliquots of the infected culture were labeled electrophoresis (30,31) ; the autoradiograms are shown in B. The

with ‘Gamine acids (2 $Zi per ml) for 1 min every successive times at which specific proteins are first synthesized and the times

minute for the first 22 min after infection. The labeling was at which those proteins are no longer synthesized are diagrammed (C).

stopped with ice and casamino acids (added to give 10 mg per ml). Clearly this diagram reflects judgments made from a visual

The amino acid incorporation into protein was measured for 0.08 inspection of the gels in B; proteins synthesized at low rates at late

ml of cells (A). The rates for uninfected cells (0) and infected times may be diagrammed as though they were no longer synthe- sized at all.

RNA decay for unperturbed infections (35, 36). The pulse- labeled samples obtained for the experiment shown in Fig. 2B were analyzed as well on SDS-acrylamide gels (Fig. 2C). Sev- eral effects of rifampicin addition may be observed (in compari- son with Fig. 1B) : (a) no postreplicative proteins are synthesized during any of the labeling periods; (b) the sequential appearance of most prereplicative proteins is conserved (note, for example, the proteins corresponding to genes 43, rIIA, 46, 39, and rIIB); and (c) at least one prereplicativc protein (encoded by gene 32) is not synthesized at all. Furthermore, the relative intensities of several bands are affected by rifampicin addition.

The effects of rifampicin addition on protein synthesis (implicit in Fig. 2C) may be shown more dramatically through the use of

continuously labeled cultures. Two T4-infected cultures were labeled with r4C-amino acids for the first 12 min after infection; one culture received rifampicin at 1 min post-infection. The radioactive proteins from these two cultures were analyzed on SDS-acrylamide gels (Fig. 3). At least 10 bands have greatly decreased intensities in the preparation from rifampicin-treated T4-infected cells. Among the T4 prereplicative genes whose expression is selectively inhibited by rifampicin addition at 1 min post-infection are the genes 43, 46, 45, rIIB, 32, and internal protein III. Bautz and his co-workers have presented evidence that the rIIB cistron is under the control of an independent promoter which is not recognized until about 2 min post-in- fection at 30” (15). Thus the class of cistrons represented by

by guest on July 18, 2018http://w

ww

.jbc.org/D

ownloaded from

E 4 25 z :: f

.- .E E E 20

a a

3 x 15

E 4

- IO

Time After InfectIon (minutes)

$‘k;O-2 2-4 4-6 6-8 8-10 15-18 18-21 21-26 26-31 j

Time dfter Infection (mh~tes)

FIG. 2. The effect of rifampicin addition 1 min after infection on T4specific protein synthesis. Cells were infected with T4 and simultaneously labeled with ‘%-amino acids (1 &i per ml; 10 pg of casamino acids per ml). At 1 min post-infection the cells were divided into a control culture (0) and a rifampicin-treated culture (0). At various times post-infection, aliquote of 1.0 ml were re- moved and the incorporation of W-amino acids into proteins was determined (A). In a separate experiment cells were infected with T4, and rifampicin was added to the entire culture at 1 min post-infection. Aliquots of this culture were pulsed with 14C- amino acids (0.8 pCi per ml) at the following times: 0 to 2, 2 to 4, 4 to 6,6 to 8,s to 10, 10 to 12, 12 to 15, 15 to 18, 18 to 21, 21 to 26, and 26 to 31 min post-infection. The rates of amino acid incor- poration for an uninfected aliquot (0) and the various pulses after infection (0) are given as counts per min per min for 0.5 ml of cells (B). Several samples from B were run on SDS-gels and are shown in C.

5505

., ::52. _ _ _

Control ‘Infeciion Rifampicin Added One Minute After

Infection

FIG. 3. The effect of rifampicin addition at 1 min post-infection on T4specific protein synthesis during continuous labeling. Cells were infected at zero time and rifampicin was added to one-half of the culture at 1 min post-infection. Radioactive amino acids (0.2 PCi per ml) were added successively to the cultures at zero time, 3,6, and 9 min after infection; the infections were stopped at 12 min post-infection. The radioactive proteins were analyzed on SDS-gels. The arrows indicate those bands which are missing or present in reduced quantities in the culture to which rifampicin was added at 1 min post-infection.

the 10 rifampicin-sensitive bands in Fig. 3 may share a common regulation with the rIIR cistron. For the remainder of this

paper we assume that these rifampicin experiments define a

second class of promoters which we designate quasi-late (Pa). Under “Discussion” we will outline a defense for this specific interpretation.

Kinetics of Quasi-late Promoter Recognition-We now ask at

what time after infection the second class of promoters is recog-

nized. Since we know that no new prereplicative proteins are synthesized later than 6 min post-infection (Fig. l), we can be certain that all prereplicative promoters are first recognized before that time. Therefore we have infected E. coli and added rifampicin to the infected cultures at various times after in- fection. Each of these cultures was continuously labeled for the first 12 min of the infection. The data for amino acid incorporation in such cultures and the SDS-polyacrylamide gel analyses are shown (Fig. 4, A and B) . These preliminary experi- ments show that a short period of transcriptional initiations is sufficient to saturate the translational apparatus; that is, for rifampicin addition at 1. min and 2 min post-infection, amino acid incorporation is approximately 60% and 80% of the untreated control, respectively (see below). As a function of the time of

by guest on July 18, 2018http://w

ww

.jbc.org/D

ownloaded from

3

2 1-

Time of Rifampicin Addition

(minutes after infection)

FIG. 4. T4specific protein synthesis after rifampicin addition at various times post-infection. Cells were infected at zero time in the presence of %-amino acids (0.8 $Zi per ml). At the times indicated, aliquots were transferred into rifampicin. The infec- tions were stopped at 12 min after infection by pouring onto ice in the presence of casamino acids (10 mg per ml). Amino acid in- corporation into protein is shown for two separate experiments (0, n ); the data are for 0.25 ml of infected cells (A). Samples taken from the cultures which received rifampicin at every min- ute after infection (H) were run on SDS-gels (B). The autoradio- gram also shows a control culture labeled for the first 12 min of infection in the absence of rifampicin.

rifampicin addition to each infected culture, there is a progres-

sive increase in the relative intensities of those bands regulated by quasi-late promoters (note in Fig. 4B, for example, the pro-

teins encoded by genes 43, 46, 32, rIIB, 45, and internal protein III). We note as well that simultaneous addition of rifampicin

and T4 does not permit any T4-specific protein synthesis (as in 0’ in Fig. 4B).

In order to understand the implications of Fig. 4B, the results were re-examined quantitatively. T4-infected cells were again treated with rifampicin at various times post-infection; the labeling period was again continuous but was extended to 20 min (which allows complete translation of available mRNA (see Fig. 2B)). The radioactive proteins were analyzed on SDS- acrylamide gels; the band densities in the resulting autoradio- grams were determined. We then plotted the relative accumu-

Time of Rifampicin Addition (minutes after infection)

FIG. 5. Accumulation of specific prereplicative proteins after rifampicin addition at various times post-infection. In three separate experiments, cells were infected with T4 in the presence of ‘*C-amino acids (1 pCi per ml; 10 pg of casamino acids per ml). Aliquots of each culture were transferred into rifampicin at vari- ous times after infection. The incubations were stopped at 20 min post-infection with ice and casamino acids. The ‘Gproteins were analyzed on 7.5% SDS-acrylamide gels, and the resulting autoradiograms were scanned on the Joyce-Loebl microdensitome- ter. Peaks corresponding to specific proteins were cut out from the scan and weighed. The precision of this method of determin- ing areas is ~3%. The data shown represent the average of three experiments. The eight prereplicative proteins illustrated re- side in unique positions of the SDS gel (rIIB was analyzed in a genetic background of 3%, 44-; the gene 32 protein was analyzed in a background of rIIB-, 44-; see Fig. G as well). We normalized the data for each protein to that time of rifampicin addition which gave the maximum accumulation of that specific protein. A, proteins from gene 32 (0), gene rIIB (O), gene 46 (A), gene 43 (A); B, proteins from gene 39 (0), gene rIIA (O), gene 52 (A), gene 293 (A).s

lation of eight prereplicative proteins as a function of the time of rifampicin addition (Fig. 5, A and B). Clearly two classes of prereplicative proteins may be distinguished: some proteins (encoded by genes 52, rIIA, 39, and 2935) accumulate in de- creasing amounts if rifampicin is added after 2 min (Fig. 5B), whereas some proteins (encoded by genes 43, 46, 32, and rIIB) accumulate in increasing amounts if rifampicin is added at 2 min

post-infection or later (Fig. 5A). Included in the second class is the gene 32 protein whose synthesis is completely inhibited by rifampicin addition during the first 135 min of infection. We will interpret these data (under “Discussion”) by suggesting t.hat some genes (such as rIIA, 39, 52, and 293) are regulated uniquely by promoters recognized immediately after infection @‘n’s), that some genes (such as 43,46, and rIIB) may be regula- ted both by Pn’s and quasi-late promoters @Q’S), and that some genes (such as 32) may be uniquely regulated by Po’s. In

5 For simplicity we designate the prereplicative gene whose product is missing during infections with T4 293 as gene 293 (31).

by guest on July 18, 2018http://w

ww

.jbc.org/D

ownloaded from

5507

addition, we will argue that quasi-late promoter recognition first occurs at 135 min (see also below) post-infection, but that quasi-late transcripts probably account for only a small fraction of the prereplicative RNAs which accumulate in the first minutes after infection (5, 11). Lastly, we must include in any model building the possibility that T4 development occurs at mRNA excess.

We next examined the regulation of the rIIB cistron. The rIIB cistron may be transcribed polycistronically along with the rIIA cistron or under the cont,rol of a promoter site which lies between the rIIA and rIIB genes (15, 20). The rIIA cistron is regulated by a promoter recognized immediately after infection (Fig. 4B and 5B). Therefore, the ratio of the synthetic capaci- ties for the rIIB and rIIA prot,eins, if rifampicin is added at various times after infection, should reflect the relative number of rIIB messengers initiated at each promoter by the time of rifampicin a,ddition. To quantitate the ratio of rIIB protein to rIIA protein, we constructed a mutant lacking genes 32 and 44 (since the products of genes 32 and 44 have Gmilar molecular weights to the rIIB protein (31)). Fig. 6A khows that in the

FIG. 6. Accumulation of the rIIA and rIIB proteins after rifam- acids per ml) (B). Aliquots of each culture were transferred into picin addition at various times post-infection. The double T4 rifampicin at various times after infection. The incubations were mutant amHl8 amE (32-, 44-) and the triple mutant amH18 stopped at 20 min post-infection with ice and casamino acids umE2059 r638 (32-, 44-, rIIB-) were constructed and used to infect (added to 10 mg per ml). The radioactive proteins were run on Escherichia coli BE (su-) ; the proteins resulting from a continuous SDS-gels (as above) and the resulting autoradiograms were labeling with ‘%-amino acids (0.375 &i per ml) for the first 12 scanned. The relative amounts of rIIA and rIIB specific proteins min of infection were analyzed on SDS-gels (A). These gels were were determined as described in Fig. 5. These amounts were run for 9 hours (rather than 6 hours (31)) to increase the resolution corrected for molecular weights (molecular weight of rIIB, 33,000 in the rIIB region of the gel. The autoradiograms were scanned (31) ; molecular weight of rIIA, 95,000 (31)), and the ratios of rIIB using a Joyce-Loebl microdensitometer (A). In three separate to rIIA were calculated for each time of rifampicin addition. experiments, cells were infected with each T4 mutant (as above) Each point on the curve is the average of three separate experi- in the presence of ‘%-amino acids (1 pCi per ml; 10 pg of casamino ments; the standard errors were calculated and are also shown (*).

absence of the gene 32 and 44 proteins, the rIIB protein resides in a unique position on an SDS gel. Using the double mutant (32-, 447, we analyzed promoter recognition for the rIIA and rIIB cistrons. Rifampicin was added to infected cells at various times; each culture was labeled with 14C-amino acids from the time of infection until 20 min post-infection. The radioactive proteins were run on SDS gels and the amounts of rIIA and rIIB proteins were determined from densitometer tracings. The result of this experiment is quite striking. The ratio of rIIB to rIIA protein begins to rise steeply in cultures to which rifam- picin is added at times later than 1 34 min post-infection (Fig. 6B). We suggest that this result vindicates the use of rifampicin as a probe for the timing of promoter recognition, since 2 min post- infection is precisely the time at which rIIA-independent rIIB mRNA can first be detected (15). Thus the ratio of rIIB:rIIA rises at the same moment that gene 32 expression becomes insensitive to rifampicin (Fig. 5A). We note that rifampicin addition during the 1st min of infection allows the rIIB protein to be synthesized in 2-fold molar excess of the rIIA protein. Thus the polycistronic transcript of the rIIA and rIIB region may be translated in a manner rather different from the equi- molar translation of the polycistronic tryptophan operon of E. COG (37).

Rifampicin Sensitivity of T4 DNA Synthesis-We have used T4-specific DNA synthesis as an additional diagnostic probe for the recognition of quasi-late promoters. Since at least three quasi-late genes are required for efficient DNA replication (genes 45, 43, and 32 (38)), the time at which DNA synthesis becomes rifampicin insensitive could reflect the initiation of transcription at quasi-late promoters. Infected cells were treated with rifam- picin at various times after infection; control cultures were treated with chloramphenicol at similar times. The cultures

I I I I 1 I I 2 3 4 5 6

Time of Rifompicin Addition (minutes after infection)

by guest on July 18, 2018http://w

ww

.jbc.org/D

ownloaded from

Time of Antibiotic Addition (minutes after infection)

I I 1 1 1 1

I 2 3 4 5 6

FIG. 7. T4 DNA synthesis after addition of rifampicin or chlor- amphenicol at various times post-infection. Cells were infected in the presence of [methyPH]thymidine) (4 #Zi per ml; 4 rg of thymidine per ml). At the times indicated, aliquots of 1.0 ml were transferred into rifampicin (O ) or chloramphenicol (A). The infections were stopped at 30 min post-infection by the addi- tion of an equal volume of 1 N NaOH. The samples were allowed to sit in NaOH overnight; incorporation of tritiated thymidine into DNA was measured by standard techniques (3).

were labeled with [3H]thymidine for the first 30 min of infection, and the incorporation of radioactivity into DNA was measured. DNA synthesis first became rifampicin insensitive at about 3 min post-infection (Fig. 7). In the chloramphenicol-treated cultures, DNA synthesis did not become antibiotic insensitive until later than 6 min post-infection. Apparently 6 or 7 min of protein synthesis are required for T4 DNA synthesis to begin. In a variety of experiments designed to measure when T4 DNA synthesis first begins, we have been unable to detect incorpora- tion of tritiated thymidine into T4 DNA before about 7 min post-infection4 (3). We suspect that in these experiments DNA synthesis reflects quasi-late promoter recognition; we have not reversed cause and effect, since the absence of DNA synthesis during infections with the appropriate mutants does not prevent quasi-late promoter recognition (31). However, the possible direct involvement of RNA synthesis with DNA replication (39-41) suggests that these data should be interpreted cautiously.

Promoter Recognition during Postreplicatiwe Transcription- Experiments were designed to ask at what time the late pro- moters first become rifampicin insensitive; these experiments are conceptually identical with those shown earlier. Rifampicin is added to infected cells at various times post-infection; these cultures are labeled continuously with ‘4C-amino acids for the first 30 min of infection (Fig. 8A). At about the 11th min after infection, late protein synthesis becomes insensitive to rifam- picin (Fig. 8B). These facts are compatible with DNA-RNA hybridization data (5, 6), which suggest that late transcription may be detected at about the 11th min post-infection at 36”. Sheldon has measured the time at which the synthesis of the gene 34 protein becomes insensitive to rifampicin (42) ; our data support her conclusions. Since her experiments were performed in cells made permeable to rifampicin by the procedure of Leive (43), the similarities in our data suggest that the permeability of rifampicin is not a complication in our experiments.

DISCUSSION

Rifampicin as a Probe for Classes of Gene Expression-Most of the experiments in this paper are dependent on the action of the

I I II II III I 2 4 6 8 IO 12 14 16 18

Time of Rifompicin Addition (minutes after infection)

FIG. 8. T4specific protein synthesis after rifampicin addition at both early and late times after infection. Cells were infected and 14C-amino acids (1 &i per ml; 20 pg of casamino acids per ml) were added simultaneously. At the times indicated, aliquots were transferred into rifampicin. The infections were stopped at 30 min post-infection by pouring onto ice in the presence of casamino acids (10 mg per ml). Amino acid incorporation into protein was determined for 0.08 ml of culture (A). The rates for uninfected cells (0) and infected cells (0) are shown. These samples were analyzed on SDS-gels using 7.5% acrylamide in the separating gel (B). The products of genes 34, 7, and 37 were identified by R. Vanderslice and C. Yegian using appropriate mutants; the product of gene 23 has been identified previously (30).

antibiotic rifampicin (26). For our interpretations to be valid, we must know that rifampicin rapidly prevents the subsequent initiation of RNA synthesis in T4-infected E. coli AS19. In fact, we do not know precisely to what extent the rifampicin has blocked initiation events. We used as controls cultures to which rifampicin and bacteriophage were added simultaneously; never did such cultures transcribe and translate any bacterio- phage genes (see for example, Fig. 4B and SB). Since irradiated hosts were used for these experiments (31), we would have ob- served easily a tiny fraction of the normal synthetic rate in autoradiograms obtained after long exposures. These results suggest that by the time adsorption and DNA penetration have occurred, the RNA polymerase of the host must be completely inhibited by the antibiotic.

We assume for the remainder of the discussion that the sensi- tivity of RNA polymerase to rifampicin during T4 infection is not altered by changes in the uptake of the drug. Support for

by guest on July 18, 2018http://w

ww

.jbc.org/D

ownloaded from

this assumption may be derived from a comparison of our data erase must be responsible for the recognition of early promoters for rifampicin-refractive synthesis of the gene 34 protein and (7, 8, 48). similar data obtained by Sheldon (42). Sheldon’s experiments The experiments reported in this paper, the experiments with were performed with cultures made permeable to rifampicin cell-free systems (16-19), and the data demonstrating read- using the procedure of Leive (43); in her experiments the synthe- through from immediate early cistrons as the primary source of sis of the gene 34 protein became rifampicin insensitive at the delayed early RNA in &JO (11) all provide strong confirmation same time as in our experiments (Fig. 8B). Lastly, the available of the topological model for the prereplicative portions of the data suggest that T4-induced modifications to the RNA polym- T4 genome (proposed originally by Milanesi et al. (7, 8)). Tan- erase (12-14, 21-25) do not decrease significantly the sensitivity demly arranged early transcription units (Pn-+T,) span the of the enzyme to rifampicin (29). prereplicative regions; these transcription units must account

T4 Development i@ay Occur under Conditions of Excess mRNA- for the bulk of the prereplicative transcripts which accumulate A striking observation may be made by examining the gel pat- during the first 4 to 5 min of infection (5, 8, 19). I f initiations terns and amino acid incorporation data shown in Figs. 4 and 8. at early promoters occur rapidly (5, 6), if the average early If, as in Fig. 8, rifampicin is added at various times post-in- transcription unit is not longer than 8000 base pairs (10, 49), fection to cultures labeled continuously for 30 min, the transla- and if the RNA chain growth rate is between 1000 and 1500 tional apparatus is nearly saturated with that mRNA synthesized nucleotides per min at 30” (15, 50, 51), all of the early RNAs via initiations which occur during the first 2 min. Furthermore, would first appear within the initial 5 min of infection (5, 11). those proteins which are translated under the regulation of This is in fact the case (5). promoters recognized later in infection (quasi-late and late) Lastly, early transcription units are frequently thought to appear to be translated at the expense of proteins under the contain termination signals dividing those units into promoter control of previously available promoters. Thus, in a 30-min proximal (immediate early) and promoter distal (delayed early) labeling, substantially more protein encoded by genes 39, 52, segments (4, 5). The primary experiment in support of this and rIIA accumulates if rifampicin is added 2 min after infection concept is the exclusive transcription of immediate early regions than if the antibiotic is added much later. Quantitative data during T4 infection of chloramphenicol-treated E. coli (4, 5). which are consistent with this interpretation are shown in Fig. We have argued previously that chloramphenicol, by virtue of 5, A and B. Although complicated interpretations for this its ability to cause polarity (52, 53), can restrict apparent phenomenon may be considered (44), we favor the notion that transcription to promoter proximal genes (17, 19, 54, 55). Until the addition of new transcripts to an excess pool of mRNA can further experiments are reported which deal directly with the cause “shut-off” of some gene expression. Thus new mRNAs putative regulatory regions between immediate early and delayed could successfully compete with old transcripts for the limiting early genes, we will assume that passive transcription of early translational apparatus. We mention this aspect of T4 develop- transcriptive units provides an adequate description of most ment here because I’, recognition against a background of excess prereplicative RNA synthesis (19). mRNA must have a profound effect on the subsequent accumu- Rijampicin Experiments Dejine a Second Class of Prereplicative lation of quasi-late proteins (as in Fig. 5). We plan to explore Promoters-Of the many possible interpretations for our data, this phenomenon further. the simplest is that the sensitivity of specific gene expression to

Topology of Early Transcription Units-The early transcrip- rifampicin 1 min after T4 infection (see Fig. 3) reflects a delay tion units (PE+TE) must encode the majority of the prereplica- in promoter recognition for those genes. Schmidt et ul. (15) tive proteins. When rifampicin is added to infected cells after reported that rIIB transcription first occurs under the control 1 min of infection, all but a few of the prereplicative proteins are of a promoter adjacent to the rIIB cistron, and that rIIB- synthesized (Figs. 2 and 3) ; these include the delayed early specific mRNA can first be detected 2 min after infection. In proteins (5, 19). Among the T4 proteins synthesized in large support of those data, we find that the rIIB promoter is first quantities in a culture treated with rifampicin at 1 min post- recognized 1% min post-infection (Fig. 6B), and that the rIIB infection is the rIIA protein; since the rIIA cistron is in the protein is first detectable during a pulse of K-amino acids from delayed early class (15, 20, 45), the rifampicin does not inhibit 2 to 236 min post-infection.4 The data in Figs. 3, 5, and 6B selectively transcription of delayed early genes.6 This is con- suggest that many cistrons display a rifampicin sensitivity sistent with data which suggest that the major delayed early analogous to the rIIB cistron; we believe that these cistrons are species are generated via promoter recognitions that occur im- regulated as well by new promoters. mediately after infection; most delayed early RXAs are the 3’. Support for this straightforward interpretation may be derived OH termini of polycistronic messengers containing immediate from two independent sources. Travers has shown that the early species at the 5’ end (11, 17). RNA polymerase ‘extracted from T4-infected cells during the

Our experiments utilizing a cell-free system derived from un- prereplicative period has the capacit.y to initiate transcription infected E. coli suggest that the majority of immediate early and at sites other than those recognized by the E. coli holoenzyme delayed early genes can be expressed in vitro (16-19, 46, 47). (12-14). In addition, Salser et al. described a prereplicative Early gene expression requires no components from T4-infected class of transcripts which has a low relative abundance during cells other than T4 DNA; therefore, we may argue that the the first 5 min of infection and an increased relative abundance complete expression of early transcription units occurs without at later times (5). Each of the prereplicative proteins (the the mediation of transcriptional or translational factors altered products of genes 43, 45, 46, 32, rIIB, internal protein III, as during T4 infection (17, 19). Therefore the host RNA polym- well as each of the four remaining unidentified bands of Fig. 3)

6 If rifampicin is added to cultures at very early times post-in- which may be under the control of new promoters is synthesized

fection, very little T4-specific protein synthesis occurs (see Fig. at a higher rate 20 min post-infectlion than at any time during

4A); nevertheless, no restriction against rIIA transcription and the first 5 min of infection (see Fig. 1) ; conversely, most of the translation (relative to other T4 proteins) has ever been observed. prereplicative proteins arc synthesized at decreased ratps at

by guest on July 18, 2018http://w

ww

.jbc.org/D

ownloaded from

5510

such late times. If we argue that no lengthy delays occur between the transcription of a gene and it.s translation (56) and that the amount of translation of a specific mRNA (at different times after infection) is approximately proportionate to its relative abundance (17, 57), we may use these measure- ments of translational capacities to equate the 10 rifampicin- sensitive cistrons of Fig. 3 to the quasi-late genes of Salser et al. (5). There are likely to be additional prereplicative cistrons which utilize quasi-late regulation; interpretation of our gels is still hampered by the complexity of the patterns. We obviously focus on bands which either represent major proteins or reside in a unique position of the autoradiogram. Additional pre- replicative cistrons which might show quasi-late regulation include internal proteins I and II (58), genes 40 and 41 (59), and lysozyme (45).

Thus, our notions of prereplicative T4 development include the opening of new promoters and the constraint of a limiting translational apparatus. Our data easily fall within such con- structs. However, it is clear that other, more fanciful, interpre- tations are possible. For example, our data could be explained by the rapid synthesis after infection of any post-transcriptional “factor” which increases the translational yield of the quasi- late cistrons. Such complicated models seem premature at this time. Furthermore, the DNA-RNA hybridization data (concerning the rIIB cistron specifically (15), and the quasi- lates in general (5)) support a transcriptional explanation for our experiments.

Topology of Prereplicative Transcription Units-Our terminol- ogy may now be reclarified and set into proper perspective. Early transcription units designate prereplicative regions bounded by a promotor and termination site, PE+TE; imme- diate early genes are proximal to Ps sites, whereas delayed early genes are distal to those same PE sites (8). Quasi-late promoters may regulate genes which are included within or excluded from early transcription units; if a quasi-late promoter regulates a gene included within an early transcription unit, that gene may be an immediate early or delayed early gene (Fig. 9). Of the 10 protein bands under the control of quasi-late promoters, perhaps three do not accumulate at all if P, recognition is blocked by rifampicin (Figs. 2C and 3). Of these three bands, only one has been identified (the product of gene 32). Jayaraman has demonstrated that the prereplicative genes 40 and 41 are not transcribed in vitro by the E. coli RNA polymerase (59) ; perhaps these three genes are essentially dependent on P, recognition for any significant expression. Conversely, the rIIB, 43, and internal protein III cistrons are encoded within early transcrip- tion units; the rIIB protein and internal protein III are efficiently synthesized as well under the control of early promoters (17, 19). The rIIB cistron is promoter distal with respect to a PE site and is a delayed early gene (15, 20) ; the internal protein III gene is nearly adjacent to a PE site and is an immediate early gene (17, 45). Our experiments with rifampicin suggest strongly that the rIIB and internal protein III cistrons are regulated by quasi-late promoters in addition to early promoters.7 Thus, as depicted in Fig. 9, the early transcription units and the quasi- late transcription units must overlap to some (unknown) extent. I f chloramphenicol functions as described above, T4 infection of chloramphenicol-treated E. coli would result in the accumulation

7 A cistron is considered to be under the control of a PE if the protein encoded by that cistron appears when rifampicin is added to the culture 1 min after infection. These conclusions are verified by performing an identical infection using an appropriate mutant in that gene.

FIG. 9. The topology of a prereplicative region of the T4 ge- nome. This figure depicts the possible relationships between quasi-late promoters and the early transcription units bounded PE + Tn. The tandem arrangement of early transcription units is described by Milanesi et al. (7, 8). Quasi-late transcription units are not bounded in this figure by termination sites, since no information about such terminations is presently available.

of immediate early RNAs as well as (owing to overlap) RNAs subsequently controlled by P&As. Chloramphenicol would restrict against the accumulation of delayed early RNAs and RNAs subsequently controlled by P&Bs and Po/Cs. These facts will make the analysis of those components required for quasi-late promoter recognition in vitro a difficult task (12-14) ; in fact, only cistron-specific analyses of the products of cell-free RNA synthesis may be used to demonstrate rigorously quasi-late promoter recognition.

An additional aspect of P, recognition is implicit in Fig. 9. The quasi-late promoters may be used both to increase the transcriptional yield of specific prereplicative cistrons and also to insure that those transcripts continue to be synthesized if initiations at early promoters become infrequent. We note that each of the identified quasi-late cistrons (except gene 32) is thought to encode a catalytic function (60); therefore it is not clear what evolutionary constraints demanded the generation of this second transcriptional mode. However, the notion of overlapping transcription units is not original and has been discussed previously (61). In particular, we call attention to an extremely thoughtful discussion pertaining to such overlap in the prereplicative regions of the T4 genome (48) ; that discus- sion, although addressed to a slightly different point, could well have been written for the data presented in this manuscript.

We emphasize that we do not understand in detail what alterations of the host RNA polymerase are brought about by T4 infection, nor do we understand what role such alterations play in controlling promoter recognition. The host enzyme is altered soon after infection; the (Y subunit is adenylated (21, 25), the /3 and /3’ subunits are altered in an unknown manner (22), and the cr factor no longer is bound to the core during purification (23). These changes occur quickly enough to cause quasi-late promoter recognition (21, 25), although no direct evidence has been presented that the two phenomena are related. We note that a new T4 mutation has been described which, under re- strictive conditions, does not appear to allow the expression of the quasi-late cistrons.8 The RNA polymerase undergoes further alterations at later times, becoming associated after the

8 T. Mattson, personal communication.

by guest on July 18, 2018http://w

ww

.jbc.org/D

ownloaded from

5511

9th min with the protein products of genes 33 and 55 (62). Since the products of genes 33 and 55 are required for post- replicative transcription (3, 63), it is reasonable that the addition of these proteins to the RNA polymerase is involved directly in late promoter recognition. We do not know whether RNA polymerase is altered at the complete expense of the previously functioning enzyme. If the enzyme is only partially altered at each conversion, then the infected cell continuously acquires additional capacities to interact with DNA, without losing (completely) previous promoter options. If these ideas are even partially correct, the description of altered promoter recog- nition in molecular terms seems quite distant. As the phage develops, the complexity of its genetic regulation continuously increases. We hope that a description of the temporal control of promoter recognition in viva will be of value to those attempt- ing cell-free transcription using components found in T4-in- fected cells.

Acknowledgme7ats-We thank Dr. Joyce Silver for reading this manuscript and Tom Hill for helping with the experiments. We thank many of our colleagues in Boulder, especially the other members of our laboratory, Dr. David Hirsh’s laboratory, and Dr. Charles Yegian’s laboratory for helpful and enjoyable discussions. One of us (L.M.G.) thanks Drs. Richard Epstein and Ed Brody for their hospitality during pleasant visits to their laboratories.

1. 2.

3.

4. 5.

6.

7

8.

9. 10.

11.

12. 13. 14.

15.

16.

17. 18.

19.

20.

21.

22.

23.

REFERENCES NYGAARD, A. P., AND HALL, B. D. (1964) J. Mol. Biol. 9,125 BOLLE, A., EPSTEIN: R. H., SALSER, W., AND GEIDUSCHEK,

E. P. (1968) J. Mol. Biol. 31, 325 BOLLE, A., EPSTEIN, R. H., SALSER, W., AND GEIDUSCHX,

E. P. (1968) J. Mol. Biol. 33, 339 GRASSO,.R. J., AND BUCHANAN, J. M. (1969) Nature 224,882 SALSF,R. W.. BOLLE. A., AND EPSTEIN. R. (1970) J. Mol. BioZ.

49,271 ’ ’ ’ I ~ I

GUHA, A., SZYBALSI<I, W., SALSER, W., BOLLE, A., GEIDU- SCHEK, E. P., AND PULITZER, J. F. (1971) J. Mol. Biol. 69, 329

MILANESI, G., BRODY, E. N., AND GEIDUSCHEK, E. P. (1969) Nature 221, 1014

MILANESI, G., BRODY, E. N., GRAU, 0.. AND GEIDUSCHER, E. P. (1970) F%oc: Nat. A&d. Sk. U. S. A. 66, 181

ROBERTS. J. W. (1969) Nature 224. 1168 RICHARDSON, J. ‘P. (i970) Cold Spring Harbor Symp. Quant.

Biol. 36, 127 BRODY, E., SEDEROFF, R., BOLLE, A., AND EPSTEIN, R. H.

(1970) Cold Spring Harbor Symp. Quant. Biol. 36, 203 T~Av&s, A. A: (1969) Nature i2i, 1107 TRAVERS. A. A. (1970) Nature 226. 1009 TRAVERS; A. (1970) Cold Spring ‘Harbor Symp. Quanf. Biol.

36, 241 SCHMIDT, D. A., MAZAITIS, A. J., KASAI, T., AND BAUTZ, E. K.

F. (1970) Nature 226, 1012 GOLD, L. M., AND SCHWEIGER, M. (1970) J. BioZ. Chem. 246,

2255 BLACK, L. W., AND GOLD, L. M. (1971) J. Mol. BioZ. 60, 365 BRODY, E. N., GOLD, L. M., AND BLACK, L. W. (1971) J. Mol.

BioZ. 60, 389 O’FARRELL, P. Z., AND GOLD, L. M. (1973) J. Biol. Chem. 243,

5512-5519 SEDEROFF, R., BOLLE, A., GOODMAN, H. M., AND EPSTEIN,

R. H. (1971) Virology 46,817 GOFF, C. G., AND WEBER, K. (1970) Cold Spring Harbor Symp.

Quant. Biol. 36, 101 SCHACHNF,R, M., AND ZILLIG, W. (1971) Eur. J. Biochem. 22,

513 SEIFER, W., QASBA, P., WALTER, G., PALM, P., SCHACHNI’:R,

M., AND ZILLIG, W. (1969) Eur. J. Biochem. 9,319

24.

25.

26.

27.

28.

29.

30. 31.

32.

33. 34. 35. 36. 37.

38.

39.

40.

41.

42. 43. 44.

45. 46.

47.

48.

49.

50. 51.

52. 63. 54. 55.

56.

57.

58.

59. 60. 61.

62. 63.

HAGER, G., HALL, B. D., AND FIELDS, K. L. (1970) Cold Spring Harbor Symp. Quant. BioZ. 36,233

ZILLIG, W., ZECHEL, K., RABUSSAY, D., SCHACHNF,R, M., SETHI, V. S., PALM, P., HEIL, A., AND SEIFERT, W. (1970) Cold Spring Harbor Symp. Quant. BioZ. 36,47

DI MAURO, E., SNYDER, L., MARINO, P., LAMBERTI, A., COPPO, A., AND TOCCHINI-VALENTINI, G. P. (1969) Nature 222,533

HEIL, A., AND ZILLIG, W. (1970) Fed. Eur. Biochem. Sot. Lett. 11, 165

LILL, H., LILL, U., SIPPEL, A., AND HARTMANN, G. (1969) in RNA-Polymerase and Transcription (SILVESTRI, L., ed) p. 55, American Elsevier Publishing Co., New York

HASELKORN, R., VOGEL, M., AND BROWN, R. D. (1969) Nature 221,833

LAEMMLI, U. K. (1970) Nature 227, 680 O’FARRELL, P. Z., HUANG, W. M., AND GOLD, L. M. (1973)

J. Biol. Chem. 248, 5499-5501 SEKIGUCHI, M., AND IIDA, S. (1967) Proc. Nat. Acad. Sci.

u. s. A.‘66, i315 ’ HOSODA. J.. AND LEVINTHAL. C. (1968) Virology 24, 709 STUDIE$, F: W. (1972) Science 176,~367 GREENE, R., AND KORN, D. (1967) J. Mol. BioZ. 28,435 MATHEWS, C. K. (1968) J. Biol. Chem. 243, 5610 MORSE, D. E., BAKER, R. F., AND YANOFSKY, C. (1968) Proc.

Nat. Acad. Sci. U. S. A. 60, 1428 EPSTEIN, R. H., BOLLE, A., STEINBERG, C. M., KELL~~:N-

BERGER, E., BOY DE LA TOUR, E., CHEVALLEY, R., EDGAR, R. S., SUSSMAN, M., DENHARDT, G. H., AND LEILAUSIS, A. (1963) Cold Spring Harbor Symp. Quant. Biol. 26, 375

SCHEKMAN, R., WICKNER, W., WESTERGAARD, O., BRUTLAG, D., GEIDER, K., BERTSCH, L. L., AND KORNBERG, A. (1972) Proc. Nat. Acad. Sci. U. S. A. 69, 2691

SUGINO, A., HIROSE, S., BND OKAZAKI, R. (1972) Proc. Nat. Acad. Sci. U. S. A. 69, 1863

BUCKLEY, P. J., KOSTURI~O, L. D., AND KOZINSKI, A. W. (1972) Proc. Nat. Acad. Sci. U. S. A. 69, 3165

SHELDON, M. V. (1971) PhD. thesis, University of Chicago LEIVE, L. (1965) Biochem. Biophys. Res. Commun. 18, 13 WIBERG, J. S., DIRKSEN, M.-L., EPSTEIN, R. H., LURIA, S.

E., AND BUCHANAN, J. M. (1962) Proc. Nut. Acad. Sci. U. S. A. 46,293

KASAI, T., AND BAUTZ, E. K. F. (1969) J. Mol. BioZ. 41,401 GOLD, L. M., AND SCHWEIGER, M. (1969) Proc. Nat. Acad. Sci.

U. S. A. 62, 892 GOLD, L. M., AND SCHWEIGER, M. (1971) Methods Enzymol. 20,

535 BRODY, E. N., AND GEIDUSCHEK, E. P. (1970) Biochemistry 9,

1300 SAUERBIER, W., MILLETTE, R. L., AND HACKETT, P. B., JR.

(1970) Biochim. Biophys. Acta 209, 368 BREMER, H., AND YUAN, D. (1968) J. Mol. Biol. 34,527 MANOR, H., GOODMAN, D., AND STENT, G. S. (1969) J. MoZ.

Biol. 39, 1 MORSE, D. E., AND YANOFSKY. C. (1969) Nature 224, 329 ZIPSER; D. (1967) J. Mol. BioZ: 29, ;141 MORSE. D. E.. AND PRIMAICOFF. P. 11970) Nature 226, 28 VARMU~, H. I&, PERLMAN, R. i., .& ?ASTAN, I. (1971) Na-

ture New Biol. 230. 41 MORSE, D. E., MOSTELLER, R. D., AND YANOFSKY, C. (1969)

Cold Spring Harbor Symp. Quant. BioZ. 34,725 YOUNG, E. T., II, AND VAN Houws, G. (1970) J. Mol. Bio2.

61,605 HOWARD, G. W., WOLIN, M. L., AND CHAMPE, S. P. (1972)

Trans. N. Y. Acad. Sci. 34, 36 JAYARAMAN, R. (1972) J. Mol. BioZ. 70, 253 SNUSTAD, D. P. (1968) Virology 36, 550 SZYBALSIU, W., BOVR~, K., FIANDT, M., HAYES, S., HRADEIONB,

Z., KUMAR, S., LOZERON, H. A., NIJKAMP, H. J. J., AND STEVENS, W. F. (1970) Cold Spring Harbor Symp. Quant. BioZ. 36, 341

STEVANS, A. (1972) Proc. Nat. Acad. Sci. U. S. A. 69, 603 CASCINO, A., RIVA, S., AND GIOIDUSCHEIC, E. P. (1970) Cold

Spring Harbor Symp. Quant. Biol. 36,213

by guest on July 18, 2018http://w

ww

.jbc.org/D

ownloaded from

Patricia Z. O'Farrell and Lawrence M. GoldPREREPLICATIVE CISTRONS

Bacteriophage T4 Gene Expression: EVIDENCE FOR TWO CLASSES OF

1973, 248:5502-5511.J. Biol. Chem. 

  http://www.jbc.org/content/248/15/5502Access the most updated version of this article at

 Alerts:

  When a correction for this article is posted• 

When this article is cited• 

to choose from all of JBC's e-mail alertsClick here

  http://www.jbc.org/content/248/15/5502.full.html#ref-list-1

This article cites 0 references, 0 of which can be accessed free at

by guest on July 18, 2018http://w

ww

.jbc.org/D

ownloaded from