JOURNAL OF VIROLOGY, Aug. 1976, p. 359-373 Copyright X) 1976 American Society for Microbiology

Vol. 19, No. 2 Printed in U.S.A.

DNA Strand Specificity of Temporal RNA Classes Produced During Infection of Bacillus subtilis by SP82 JONATHON M. LAWRIE, GEORGE B. SPIEGELMAN, AND H. R. WHITELEY* Department of Microbiology and Immunology, University of Washington, Seattle, Washington 98195 Received for publication 10 February 1976

The DNA of the Bacillus subtilis bacteriophage SP82 has been separated into heavy (H) and light (L) fractions by centrifugation in buoyant density gradients in the presence of polyguanylic acid. Competition-hybridization experiments were performed with these separated fractions using RNAs isolated from cells labeled at intervals which account for 80% of the lytic cycle and unlabeled competitor RNAs isolated from phage-infected cells at 2-min intervals throughout infection. The analysis of temporal RNA classes was facilitated by use of a double reciprocal plot of the data. Five temporal classes binding to the H fraction and three binding to the L fraction were detected; the possible existence of an additional class transcribed from the H fraction is discussed. RNA synthesized in the presence of chloramphenicol contains two of the three classes produced from L-DNA and two of the five classes transcribed from H-DNA.

Several phage transcription programs have been investigated by means of competition-hybridization using labeled RNA isolated at various times during infection and each of the separated complementary strands of viral DNA (2-4, 7, 8, 12, 14, 15, 22). In some instances, such as the infection of Escherichia coli with T7, transcription is completely asymmetric (22). Usually, however, some portion of both strands is transcribed during the infection. For example, investigations of the RNA produced during the T4 infection of E. coli (8) show that late RNA is synthesized predominantly from the light (L) strand, defined as the strand of lower density in CsCl-polyuridylic acid-polyguanylic acid [poly(U,G)] gradients. Initially, L-strand transcription represents only a few percent of the total RNA synthesized, but this value increases to 60% after 2 min of infection and to 85% by 5 min. Similar experiments with the heavy (H) strand of T4 indicate that immediate early and delayed early classes are transcribed from this strand; only 10 to 25% of the truly late RNA is complementary to the H strand. Transcription of the Bacillus phage 029 has also been examined using competition-hybridization with separated DNA strands. Schachtele et al. (14, 15) demonstrated that early RNA hybridizes exclusively to the L strand and is produced at all times during the phage infection, accounting for 30% of the RNA synthesized late in infection. Transcription of late genes from the H strand begins after initiation of DNA replication; when chloramphenicol is added, only early RNAs are produced.

Competition-hybridization experiments with unfractionated denatured DNA from two related Bacillus subtilis phages, SPOl (4, 5) and SP82 (18), disclosed six temporal classes of RNA produced during infection. Gage and Geiduschek (4) separated fractions of SPOl DNA by complexing with poly(I,G) and found that transcription occurs predominantly from L fraction DNA very early in infection but that by 7 to 13 min after infection almost all of the RNA was produced from H fraction DNA. A similar pattern of synthesis was found for SP82 (12), and evidence was presented that the switch to transcription of DNA which separates as the H fraction requires the synthesis of a modified RNA polymerase (12, 17; In R. Losick and M. Chamberlin [ed.], RNA Polymerase, in press). The present paper reports a more complete analysis of transcription from the H and L fractions of SP82 DNA during the infection of B. subtilis by this phage. These experiments show that the synthesis of five temporal classes of RNA from H fraction DNA and three temporal classes of RNA from L fraction DNA can be detected by means of competition hybridization. MATERIALS AND METHODS Phage infection. M medium (17) used for the growth of B. subtilis strain 168 was modified by the addition of 0.1% glucose. Cultures were infected during early log phase by adding 5% (vol/vol) SP82 lysate (a multiplicity of 5 to 10 phage/cell). The time of adding the lysate was determined in preliminary experiments so that cultures, incubated with vigorous shaking at 37°C, had a lytic cycle of 36 to 38 min. To minimize the possibility of variation in the tran359

360

scriptional

program, each series of unlabeled culprepared over a 2- to 3-day period using identical conditions of growth and infection, the same batch of medium, and the same lysate. Isolation of RNA. To obtain labeled RNA,

tures

J. VIROL.

LAWRIE, SPIEGELMAN, AND WHITELEY was

[3H]uridine (43 Ci/mmol, Amersham/Searle Corp.,

Arlington Heights, Ill.) was added at a concentration of 1 to 1.5 /Ci/ml to 100-ml portions of culture at the indicated times. Cells were collected by centrifugation and stored at - 20°C. RNA was extracted only if control cultures lysed in 36 to 38 min. Chloramphenicol RNA (CM-RNA) was extracted from cultures to which chloramphenicol (200 Ag/ml) was added 5 min before the addition of phage; cultures were harvested 10 min postinfection (18). Labeled and unlabeled RNA preparations were extracted by the following procedure: frozen cell pellets (1 g) were suspended in 4 ml of 0.05 M potassium acetate, pH 5.0, containing 0.05 mg of bentonite per ml, and 2 ml of freshly distilled phenol (saturated with the same potassium acetate buffer) were added. The suspension was sonicated for 3 min using a Branson Sonifier (Heat Systems-Ultrasonics, Inc., Plainview, Long Island, N.Y.) equipped with a microtip, extracted at 60°C for 10 min after the addition 0.7 ml of 10% sodium dodecyl sulfate and 4 ml of buffer-saturated phenol and centrifuged. The interphase pad was reextracted with a 1.5 ml of acetate buffer. The aqueous phases were combined, phenolextracted as above at 60°C for 10 min, and precipitated overnight at -20°C by the addition of 2 volumes of 95% ethanol in the presence of 0.1 volume of 30% potassium acetate. The precipitated material was collected by centrifugation, washed with -20°C 95% ethanol, dissolved in 4 ml of 0.04 M Tris, pH 7.00.01 M MgCl2 and 150 Ag of DNase (Worthington Biochemicals, Freehold, N.J., RNase-free) were added. The preparation was dialyzed for 4 h at 37°C against 400 volumes of 0.04 M Tris, pH 7.00-0.01 M MgCl2. The RNA was then extracted twice with phenol at room temperature for 5 min, ethanol precipitated overnight, washed with ethanol, and dissolved in 0.1x SSC (0.15 M NaCl-0.015 M sodium citrate). All RNA preparations were tested for RNase activity by comparing the trichloroacetic acid-precipitable radioactivity of [L HIRNA before and after overnight incubation with 1-mg quantities of the test samples under the conditions used for hybridization. Test samples were reextracted with phenol and reprecipitated with ethanol until free of RNase activity. The specific activities of 3H-labeled RNA preparations used in the experiments presented in the figures are summarized in Table 1. Preparation of separated fractions of SP82 DNA. DNA was extracted from SP82 phages purified by sedimentation through CsCl (17). H and L fractions were prepared as described previously (12). Analytical ultracentrifugation. One-half microgram of native SP82 DNA was added to preparations before centrifugation to serve as a buoyant density marker (1.742 g/cm3; 16). All samples had a total volume of 0.3 ml in 0.04 M Tris-0.001 M EDTA, pH 7.5, before the addition of CsCl (60%, wt/vol). The samples were centrifuged in a model E Beckman ultracentrifuge using a single-sector cell in an ANF

rpm for 20 h at room temperature. Negatives were obtained using UV optics and were scanned with a Joyce-Loebel densitometer. RNA-DNA hybridization. Reactions were performed according to Nygaard and Hall (13) under conditions approaching DNA excess. Reaction mixtures containing 1.5 to 3 jig of H or L fractions of DNA, 20 to 60 Ag of [3H]RNA (as indicated in Table 1) and 0 to 500 ,ug of competitor RNA were incubated at 65°C for 18 h in 0.25 ml of 2x SSC and then treated for 2 h at 37°C with RNase (a final concentration of 10 units/ml T, and 10 ,4g of RNase A per ml, both from Sigma Chemical Corp., St. Louis, Mo.). The RNase-resistant material was collected by filtration of the reactions mixtures on B-6 membrane filters (Schleicher and Schuell, Keene, N.H.), and each filter was washed with 75 ml of cold 0.5 M KCl-0.01 M Tris buffer, pH 7.5. The filters were dried and the radioactivity was determined in a scintillation counter using a toluene-based scintillant. The percentage of added radioactivity binding to H-DNA was 8.5 for the 0- to 5-min 3H-labeled RNA and increased to 60 for RNAs extracted from cultures labeled in the 9- to 13-min interval and thereafter. The percentage of added radioactivity binding to L-DNA for the same IH-labeled RNAs was 2 to 0.5%. RNA-RNA hybridization. Self-annealing of 3Hlabeled RNAs was determined by hybridizations in the absence of DNA. The RNase-resistant material was precipitated with trichloroacetic acid after addition of carrier RNA (a final concentration of 75 ,g/ ml of yeast RNA, Sigma Chemical Corp., St. Louis, Mo.), collected by filtration on glass fiber filters, and the radioactivity was determined as described above. The same method was used to test for the possible presence of "antimessenger" except that 3Hlabeled RNA was mixed with unlabeled RNA preparations extracted from cells at different times during the infection.

rotor at 42,040

TABLE 1. Properties of labeled RNA and concentration of reactants used in competitionhybridization experiments" Pulse interval (min)

Sp act (cpm/

j.tg)

Fig.

Concn of DNA

(,ug)

Concn of 3H-labeled RNA (,ug)

H/L

0-5 3 2,602 3.0 H 26.5 4 1-5 10 3,332 2.0 L 43.2 4 5-9 4 4,324 2.0 H 51.8 16 5-9 2,954 11 2.0 L 14.2 15 9-13 1,578 5 1.5 H 14.1 21 10-13 2,305 12 2.0 L 45.2 21 12-15 704 6 1.5 H 19.2 32 17-20 3,090 7 3.0 H 17.5 39 21-25 2,620 8 1.5 H 14.2 53 a The interval of labeling is indicated relative to the time of addition of phage. H/L is the ratio of radioactivities bound to isolated H and L fractions of SP82 DNA in hybridizations with 20 A.g of 3H-labeled RNA and 3 ,ug of isolated H- and L-DNA.

RNA CLASSES SYNTHESIZED DURING SP82 INFECTION

VOL. 19, 1976

RESULTS AND DISCUSSION

Separation of H and L fractions of SP82 DNA. Figure 1A shows the separation of two fractions when denatured SP82 DNA was subjected to buoyant density centrifugation. The densities of the H and L fractions, calculated from the density of the native SP82 marker, were 1.760 g/cm3 and 1.751 g/cm:1, respectively. These values agree with the buoyant densities of SP82 DNA fractions determined earlier by Sheldrick and Szybalski (16). Figure 1B demonstrates that centrifugation of denatured SP82 DNA in the presence of poly(G) increases the buoyant densities of the H and L fractions to 1.800 g/cm3 and 1.755 g/cm3, respectively, confirming the earlier results of Truffaut et al. (23). This difference in densities provided the means for separating relatively pure H and L

361

fractions on a preparative scale. When 3 mg of SP82 DNA was fractionated preparatively, and the separated fractions were recentrifuged [after hydrolysis to remove poly(G)], each preparation yielded a single symmetrical band with the expected buoyant density (data not shown). When equal amounts of separated fractions were mixed in the absence of marker DNA and incubated at 65°C to permit reannealing before centrifugation, a single band with the density of native SP82 DNA was formed (Fig. 1E). There was very little or no reannealing when the isolated fractions were incubated separately (Fig. 1C and D). It is unlikely that the H and L fractions represent intact complementary strands of SP82 DNA. However, the above observations demonstrate that the separated fractions contain fragments of DNA bearing complementary sequences and can be obtained in

_LC) (0

O( _\-C

I~0

~-

.

A

D

B

E

FIG. 1. Analytical CsCl gradient centrifugation patterns of fractionated and reannealed SP82 DNA. (A) Two micrograms of denatured SP82 DNA with native SP82 DNA as the marker; (B) 2 pg of denatured SP82 DNA with 2 pg poly(G) and marker DNA; (C) 1.0 pg of hydrolyzed and reannealed L-DNA; (D) 1.0 pg of hydrolyzed and reannealed L-DNA; and (E) 0.75 pg of hydrolyzed H-DNA and 0.75 pg ofhydrolyzed L-DNA after mixing and reannealing. The conditions ofdenaturing, centrifugation, hydrolysis, and reannealing are described in Materials and Methods.

362

LAWRIE, SPIEGELMAN, AND WHITELEY

relatively pure state. These fractions will be called H-DNA and L-DNA. Hybridization of [3H]RNA to H- and LDNA. The specificity of transcription from HDNA and L-DNA during the infection of B. subtilis by SP82 has been described (12). These studies showed that :H-labeled RNA extracted from cells labeled before 3 min after infection hybridized equally well to both fractions. Transcription from H-DNA increased starting at approximately 5 min after infection. By 7 min, and for the remainder of the infection, 90 to 97% of the H-labeled RNA hybridized to H-DNA. The shift to transcription from H-DNA is also documented in Table 1, which presents data on the specific activities and the ratio of H/L fraction hybridization for a series of 3H-labeled RNAs isolated from cells labeled at several intervals during infection. Control experiments on the binding of 'H-labeled RNAs to increasing amounts of DNA (Fig. 2) indicate that measurements of H/L ratios and the competition experiments discussed below were performed under conditions approaching DNA excess.

To test for the possible presence of antimRNA, :H-labeled RNAs labeled at 1 to 5 min, 5 to 9 min, and 21 to 25 min were incubated with a 100-fold excess of unlabeled 5-, 9-, and 25-min RNA or with a heterologous control (RNA extracted from E. coli). Reaction mixtures containing unlabeled RNA from SP82 infected cultures gave RNA-RNA annealing values which were only slightly higher than those

a

z C)

z

±I 2

10C,

a- 0-

0

J. VIROL.

observed in the absence of unlabeled RNA (i.e., measurements of self-annealing) or with E. coli RNA (Table 2). These observations suggest that there is very little or no synthesis of antimRNA during SP82 infection. Similar results have been obtained with SPOl (4), whereas significantly larger amounts of anti-mRNA have been detected during the T4 infection of E. coli (6). Use of a double reciprocal plot to define temporal classes of RNA. A series of competition-hybridization experiments were performed using pulse-labeled and unlabeled RNA preparations isolated at various times after infection to characterize further the patterns of transcription from the two complementary fractions of SP82 DNA. The data from these experiments are presented in two ways: (i) in the "standard" plot showing the effect of increasing amounts of various unlabeled RNAs on the binding of a given labeled RNA preparation (panel a of Fig. 3 through 8, 10 through 12, and 14 through 16) and (ii) in the form of a double reciprocal plot (9; G. B. Spiegelman, Ph.D. thesis, Univ. of Wisconsin, 1972), which transforms the standard hyperbolic curve into a straight line (b of Fig. 3 through 8, 10 through 12, and 14 through 16). Extrapolation of the straight line defined by data using moderate amounts of competitor to the ordinal intercept gives the level of competition at infinite competitor concentration, thus avoiding the necessity of using extremely large quantities of competitor RNAs. Similar plots have been developed by a number of other investigators (1, 5, 11). According to the plot used in this paper, if the unlabeled RNA contains all of the RNA species which are labeled in the radioactive RNA, the ordinal intercept will be 1.0. Intercepts greater

lo

I

0

,% 0

60 o

TABLE 2. Antimessenger content of RNAs synthesized during infection"

.0

05 0

CL

Pulse Pulse in-

Self-an-

(mmi)

nealing

x I

E

ca.

05

10

15 jg

20 DNA

25

3.0

FIG. 2. DNA saturation curves. Twenty micrograms of 0- to 5-min 3H-labeled RNA (triangles), 5to 7-min 3H-labeled RNA (circles), or 17- to 20-min 3H-labeled RNA (squares) were hybridized with increasing amounts of H-DNA (closed symbols, solid lines) and with increasing amounts of L-DNA (open symbols, dashed lines). Conditions of hybridization are described in Materials and Methods.

1-5 5-9 21-25

0.13 0.14 0.39

Annealing in the presence of RNA from: Cells infected at (min): E. coli

5

9

25

0.72 0.76 0.92

0.66 0.66 0.94

0.65 0.60 0.78

0.50 0.45 0.39

a All values are given as percentage of added radioactivity. Reaction mixtures containing 21.6 ,ug of 1- to 5-min, 25.9 ,ug of 5- to 9-min and 20.6 ,tg of 22to 25-min 3H-labeled RNAs and 2 mg of unlabeled RNA per ml as indicated were incubated and assayed as described in Materials and Methods. The specific activities of the 3H-labeled RNAs are shown in Table 1.

VOL. 19, 1976

RNA CLASSES SYNTHESIZED DURING SP82 INFECTION

than 1.0 indicate that some of the labeled RNA species are absent from the competitor. Differences in the slopes of double reciprocal lines which extrapolate to the same intercept indicate different concentrations of the same populations of RNA molecules, with the slopes inversely related to the relative abundance of the RNA species. Because the radioactive RNA is not uniformly labeled, it is conceivable that experiments with two competitors could yield very similar curves (resulting in the same ordinal intercept) despite the fact that they have different RNA species in common with the labeled preparation. This would probably be a rare occurrence and could be detected by use of mixed competitor experiments. On the other hand, two competitors yielding different intercepts must have different populations of RNA molecules. In addition, the competitor with the higher intercept must lack some RNA species which are found in both the other competitor and the radioactive RNA. Temporal RNA classes represent groups of species synthesized in sufficient quantity to be detected by competition-hybridization. The time of synthesis of these classes can be established by comparing the intercepts obtained from competition experiments with a pulse-labeled RNA and a series of competitor RNAs isolated at successive times in the lytic cycle. As an example, suppose that data from competition experiments between two unlabeled RNA preparations, A and B, isolated 5 min apart, and a pulse-labeled RNA, C, isolated 10 min later, extrapolate to two different intercepts, neither of which is 1.0. If the intercept from experiments with B is lower, then the onset of synthesis of at least one temporal RNA class can be presumed to begin in the 5-min interval between the isolation of A and B and continue up to the labeling interval. Further, since neither A nor B yields complete competition (intercept of 1.0), the synthesis of some temporal RNA classes synthesized in the labeling interval must begin after the time of isolation of B. If A and B yield identical intercepts in competition with a pulse-labeled RNA, D, isolated 15 min after B, then presumably A and B compete with the same labeled RNA species. This implies that the temporal class of RNA whose synthesis began in the 5-min interval separating A and B is no longer synthesized in the labeling interval D. Thus, the shutoff of synthesis of an RNA class can be demonstrated when two competitors whose intercepts differed in competition with one labeled RNA preparation become identical when tested with labeled RNA isolated at a later time.

363

In the present study, we used a series of labeling intervals accounting for the first 26 min of infection under conditions which gave a lytic time of 36 to 38 min. We assumed that the timing of the program of transcription would be the same in cultures grown under identical conditions and having the same lytic curves. The accuracy of the estimated intervals of synthesis of the various RNA classes is approximately + 2 min. Each competitor RNA was first checked for its ability to compete completely labeled RNA synthesized at the same time in the lytic cycle. If such "homologous" experiments did not yield complete competition, the competitor RNA was discarded. The same competitor preparations were used throughout each series of analyses (e.g., data presented in Fig. 3 through 12 and 16). All lines derived from double reciprocal plots represent the best fit as determined by the method of least squares; ordinal intercepts were determined from an average of two or more experiments. RNA preparations which yielded data extrapolating to the same ordinal intercept +0.10 were assumed to have identical RNA populations, whereas those yielding intercepts differing by values greater than +0.10 were assumed to differ. As a result, the number of temporal RNA classes described is the minimum number necessary to explain the changes in intercepts seen in the competition experiments. Furthermore, it should be noted that competition-hybridization measures only major changes in RNA populations and that synthesis of minor classes of RNA might not be detected even in experiments with separated fractions of DNA.

Analysis of temporal RNA classes transcribed from H-DNA. Classes HI and H2. The competition experiments shown in Fig. 3 can be explained by the assumption that two temporal classes of RNA are transcribed from H-DNA during the first 5 min after infection. Figure 3a demonstrates that the 3-min and 5-min unlabeled competitor RNA preparations decrease the binding of RNA extracted from cells labeled 0 to 5 min to different extents: 1.25 mg of the 3min competitor RNA per ml reduces the binding of the labeled RNA by approximately 35%, whereas the same concentration of the 5-min RNA preparation reduces the binding by 80%. A plot of these data in double reciprocal form (Fig. 3b) shows that the curve obtained with the 3-min competitor extrapolates to an intercept greater than 1.0. This indicates that the 3min RNA contains only some of the RNA species labeled in the 0- to 5-min interval. On the other hand, competition experiments using RNA extracted at 5 min of infection yield data

364

LAWRIE, SPIEGELMAN, AND WHITELEY

J. VIROL.

3a 0-5 3H-RNA

100 A

-

601

-

-'-A

\\O

0

A-A-3 -3

°-o o5,

20

0.25

0.75

1.25

0.05

0.1 I

4b nf"

6 1UU

v

i H3 e- H2

3.0

0 cD

.

a)

0 10 20. 5

6do -7

, 20

01

1.0

5a 9-13' 3H-RNA 100I A -A A-A-A3

--,~0--o

0.05

0.1

0.1

0.2

-

60 20 025

075 1.25 mg/ml Competitor RNA

03

C+/CT

FIG. 3-5. Competition-hybridization analyses with isolated H and L fractions of SP82 DNA, labeled RNAs, and competitor RNAs isolated at various times postinfection. (a) The results are plotted as percentage of radioactivity bound in the presence of competitor RNA relative to the radioactivity bound in the absence of competitor RNA as a function of the concentration of competitor RNA added to the reaction mixture; the concentrations of 3H-labeled RNA and DNA used in these experiments are shown in Table 1. (b) The results are plotted as 111-F as a function of C'T/CT, where F is the fraction of radioactivity bound relative to the radioactivity bound in the absence of competitor RNA, cT is the concentration of WH-labeled RNA and CT is the concentration of competitor RNA added to the reaction mixture; the lines are fitted by the method of least squares as determined by a Hewlett-Packard 9830A calculator; (b) inset, a graphic representation showing the synthesis of classes present in the 3H-labeled RNA.

which give an intercept of 1.0. We will call the RNA species present at 3 min, class Hi. Since the 5-min unlabeled RNA yields complete competition of the 0- to 5-min labeled RNA and the

3-min RNA does not, it can be assumed that the synthesis of at least one more temporal class, H2, starts between 3 and 5 min. The presence of temporal classes Hi and H2 in the 0- to 5-min

VOL. 19, 1976

RNA CLASSES SYNTHESIZED DURING SP82 INFECTION

interval is schematically represented in the inset of Fig. 3b. The time of onset of the synthesis of class Hi was not determined due to the difficulty in preparing competitor and labeled RNAs before 3 min. Competition-hybridizations with RNA labeled during the 5- to 9-min interval (Fig. 4a) and at later times in the lytic cycle (Fig. 5a and 6a) show that the unlabeled competitor extracted 3 min after infection has no effect on the binding of any of these labeled RNAs. This is most easily interpreted as meaning that class Hi is not synthesized after 5 min. Figure 4b shows that both the 7- and 9-min unlabeled RNAs compete the binding of 5- to 9-min 3Hlabeled RNA to an ordinal intercept of 1.0, whereas the 5-min unlabeled RNA has an intercept of 1.7. This finding suggests that H2 is synthesized until at least 7 min, and an additional new class of RNA is produced in this interval (H3, as described below). Only partial competition (approximately 15%) is observed when RNA isolated at 5 min is tested as a competitor with 3H-labeled RNA isolated from cells labeled in the 5- to 9-min interval (Fig. 4a), 9- to 13-min interval (Fig. 5a), 12- to 15-min interval (Fig. 6a), and 14- to 17-min interval (data not shown), and only slight inhibition was observed with 3H-labeled RNA isolated from cells later in infection. These results are interpreted to mean that the synthesis of H2 continues, albeit in reduced amounts, until 25 min after infection; the low level of competition indicates, however, the presence of additional RNA species in the 3Hlabeled RNA. Class H3. The data in Fig. 4 through 8 indicate that the synthesis of a third temporal RNA class from the H-DNA starts very close to 7 min and continues until at least 25 min. This class is defined by the pattern of competitions seen with unlabeled RNAs extracted at 5, 7, and 9 min tested with five labeled RNA preparations. As shown in Fig. 4a and b, the binding of 5- to 9min 3H-labeled RNA is completely eliminated by addition of either 7- or 9-min competitor RNAs (the ordinal intercept is 1.0) but the 5min RNA shows only partial competition. Thus, the 5- to 9-min labeled preparation contains new RNA species synthesized after 5 min. This class of RNAs is designated H3. Since data from competitions using the 7- and 9-min competitor preparations yield lines which have the same intercept when plotted in double reciprocal form, it is assumed that they contain the same RNA species. The annealing of RNA from cells labeled 12 to 15 min (Fig. 6a), 17 to 20 min (Fig. 7a), and 21 to 25 min (Fig. 8a) is competed 20 to 25% by

365

both 7- and 9-min competitor preparations when tested at a concentration of 2.0 mg/ml. Since the 5-min RNA preparation (class H2) only competes slightly with RNAs extracted from cells labeled at late times (Fig. 7a and 8a) in the lytic cycle, the competition observed with 7- and 9-min RNA can be assumed to be due to the continued synthesis of class H3. Experiments to determine if the production of H3 continues beyond 25 min were not performed due to the difficulty in preparing labeled RNAs at this late time in the latent period. Class H4. The synthesis of a temporal class of RNA in the middle of the infection (class H4) can be deduced from the differences in the amounts of competition observed with 11-min unlabeled RNA and either 7- or 9-min competitor RNA (Fig. 5-7). When tested with 9- to 13min 3H-labeled RNA, the 7- and 9-min competitor preparations are only partially effective, whereas the 11-min RNA is as efficient a competitor as 13-min RNA (Fig. 5a). In the double reciprocal plot of these data (Fig. 5b), both 7and 9-min competitors yield intercepts of 1.4, indicating that they compete only a portion of the RNA species labeled in the 9- to 13-min interval, whereas a similar plot of the data obtained with 11-min competitor yields an ordinal intercept of 1.0. The RNA present in the 11min RNA but absent from the 7- and 9-min RNA preparations is designated as class H4. The continued synthesis of class H4 is indicated by experiments using 12- to 15-min 3Hlabeled RNA and 7-, 9-, or 11-min competitor RNAs. As seen in Fig. 6a, each unlabeled RNA preparation competes partially. The standard plot of the data (Fig. 6a) shows that the 7- and 9-min RNAs yield identical curves and decrease the binding of the labeled RNA by approximately 25% (at a concentration of 1.5 mg/ml), whereas the same concentration of 11-min RNA competes 50% of the binding. The double reciprocal plot of these data (Fig. 6b) shows that experiments with 7- and 9-min competitor RNAs yield a line with an intercept of 2.7, and data with 11-min RNA give a line having an intercept of 1.6. According to the preceding discussion, the 7- and 9-min RNAs represent competition by classes H2 and H3. The li-min RNA contains both class H2 and class H3 and competes additional RNA species which have been designated class H4. Experiments with RNA from cells labeled in the 17- to 20-min interval (Fig. 7) show that competition by il-min RNA approaches that obtained with 7- and 9-min RNAs, whereas the same experiment performed with 21- to 25-min 3H-labeled RNA (Fig. 8) yielded virtually identical levels of competition by 7-, 9-, and 11-min

366

~3

J. VIROL.

LAWRIE, SPIEGELMAN, AND WHITELEY

6b

6a 12-15 3H-RNA 100 0

60

o\U, 20

x_

*--X

H3 H2 1

5.0

LIm

10 20. -7910 .,. II'9

II

x

3.0

131'

1-

ol

°-i 15' 1.0

H5 H4

1.0

0.1

20

7b 100

I~~~5-

0.2 H5

0.3 -

H4 H3 i H2 j, ,13' 0 10 20

5.0

o

60

13-0

II

zo-°

.

4--

0 a)

C: 0

150 -a20 0.2 0.1 0

20

1.0 1.0

2.0

-

8b H5 100

60 20

5.0

H3 H2 0 10 20

13 3.0

15'

.A

I

1.0 0.1

mg/ml Competitor RNA

0.2

0.3

CT /CT

FIG. 6-8. Competition-hybridization analyses with isolated H and L fractions of SP82 DNA, labeled RNAs, and competitor RNAs isolated at various times postinfection. (a) The results are plotted as percentage of radioactivity bound in the presence of competitor RNA relative to the radioactivity bound in the absence of competitor RNA as a function of the concentration of competitor RNA added to the reaction mixture; the concentrations of 3H-labeled RNA and DNA used in these experiments are shown in Table 1. (b) The results are plotted as Ill-F as a function of c'TcT, where F is the fraction of radioactivity bound relative to the radioactivity bound in the absence of competitor RNA, c;T is the concentration of 3H-labeled RNA and CT is the concentration of competitor RNA added to the reaction mixture; the lines are fitted by the method of least squares as determined by a Hewlett-Packard 9830A calculator; (b) inset, a graphic representation showing the synthesis of classes present in the WH-labeled RNA.

RNAs. These observations could indicate either that synthesis of class 4 is greatly decreased or stopped between 17 and 21 min, or that the similarity in curves reflects a general decrease

in RNA populations which fortuitously yielded similar curves. Mixed competitor experiments were performed to distinguish between these two possibilities. As shown in Fig. 9, when 17-

VOL. 19, 1976

RNA CLASSES SYNTHESIZED DURING SP82 INFECTION

367

tercept of 1.0. However, RNAs extracted from infected cells earlier than 13 min, such as the 100 11-min RNA, show intercepts greater than 1.0. These observations are most readily explained * 9 " on the basis that the synthesis of a new class *-i 80 (H5) begins at approximately 13 min. The same 0* 9+1 pattern of competition is observed in experi-O A9I 0 ments with 3H-labeled RNAs from cells labeled _ b g+ I I 0'- o 60 at 17 to 20 min (Fig. 7) and 21 to 25 min (Fig. 8): la-19+11 addition of 13-min unlabeled RNA yields par0 tial competition (Fig. 7a and 8a), but the double 0 reciprocal plots of these data extrapolate to 1.0 40 F (Fig. 7b and 8b). The slopes of the lines generC) ated by replotting data from experiments using L-O 13-min RNA are very large relative to the line a- 20 obtained from data with later competitors, suggesting that the synthesis of H5 starts close to 13 min and that this class is not present in large amounts until after 15 min. The difference in 2.0 1.0 0 the slopes of the curves obtained with 13-min mg/ml competitor RNA RNA in Fig. 6b and 7b can be explained on the FIG. 9. Mixed competitor experiments. Increasing basis of the relative proportions of the different amounts of11-min RNA were added to reaction mix- RNA populations synthesized during the two tures containing 1 mg of 9-min RNA per ml (trian- labeling periods. Only a portion of the unlagles); data obtained with increasing amounts of 9- beled 13-min competitor is homologous to the min RNA are indicated with circles. Hybridization 17- to 20-min labeled RNA. Thus, it is a less were performed with H-DNA and 17- to 20-min 3Hefficient competitor than when tested with 12labeled (open symbols, dashed lines) and 21- to 25- to 15-min 3H-labeled RNA where all members min 3H-labeled (closed symbols, solid lines) using of RNA population are involved in the competithe concentrations indicated in Table 1. tion. Analysis of temporal RNA classes tranto 20-min 3H-labeled RNA was tested, addition scribed from L-DNA. Approximately one-half of increasing amounts of 11-min competitor in of the RNA produced very early in SP82 infecthe presence of a high concentration of 9-min tions of B. subtilis is transcribed from L-DNA RNA resulted in a more efficient competition but by 7 min and thereafter, only a small prothan in the presence of 9-min RNA alone. In portion (10% to 2%) of the RNA hybridizes to contrast, when the same experiment was done this fraction. with 21- to 25-min 3H-labeled RNA, the mixture Classes Li, L2, and L3. Figure 10a shows of competitors yielded a curve that was only experiments using 1- to 5-min 3H-labeled RNA, slightly lower than obtained with 9-min RNA L-DNA, and increasing amounts of 3-min and alone. These observations indicate that synthe- 5-min competitor RNAs. The double reciprocal sis of class 4 stops (or is greatly decreased) at plot of these data (Fig. 10b) demonstrates comapproximately 19 min (ca., between 17 and 21 plete competition with the 5-min RNA (i.e., an min). The 20% competition observed upon addi- ordinal intercept of 1.0) and slightly less effition of either the 9- or the 11-min competitor cient competition with the 3-min RNA (an ordi(Fig. 8 and 9) could be explained as being due nal intercept of 1.2), suggesting the presence of either largely or solely to the continued synthe- two classes of RNA in the labeled preparation. The RNA present in the 3-min competitor prepsis of class H3. Class H5. Figures 6 through 8 present data aration will be called class Li, and the addiwhich support the conclusion that one other tional RNA present in the 5-min RNA will be temporal RNA class is transcribed from the H- designated L2. Competition experiments using DNA. The existence of this class is indicated by RNAs labeled at times later than 5 min demonthe fact that total competition of any labeled strate more clearly that the 3-min and 5-min RNA prepared after 13 min can be achieved RNAs contain different L-DNA specific temonly with competitors isolated after 13 min. poral classes. Figures 11 and 12 present data This is seen in Fig. 6 in which the binding of from experiments with 5- to 9-min and 10- to 13RNA labeled 12 to 15 min is effectively com- min 3H-labeled RNAs. The 3-min unlabeled peted by 13- or 15-min RNA and the correspond- RNA competes partially in these experiments ing double reciprocal plots yield an ordinal in- (20 to 35% of the control at a concentration of --

-11-

-

#

368

LAWRIE, SPIEGELMAN, AND WHITELEY

J. VIROL.

llb - 100

s-L3 9- L2

> LI 0 10 20

501

0

60

& 3' 5

4o-

A

0

-

0

t 20

1.0

1.0

2.0

r 100

5.0

02 12b ,I 0 10 ,A

1.0

1.0 2.0 mg/ml Competitor RNA

35' A,I5'

A o,oI

0

0

30

20

^

A

AA 1-

60

0.4 IX

"~

~~~

0.2

13'

0.4

CY/CT

FIG. 10-12. Competition-hybridization analyses with isolated H and L fractions of SP82 DNA, labeled RNAs, and competitor RNAs isolated at various times postinfection. (a) The results are plotted as percentage of radioactivity bound in the presence ofcompetitor RNA relative to the radioactivity bound in the absence of competitor RNA as a function of the concentration of competitor RNA added to the reaction mixture; the concentrations of H-labeled RNA and DNA used in these experiments are shown in Table 1. (b) The results are plotted as 111-F as a function of CT/CT, where F is the fraction of radioactivity bound relative to the radioactivity bound in the absence of competitor RNA, c T is the concentration of 3H-labeled RNA and CT is the concentration of competitor RNA added to the reaction mixture; the lines are fitted by the method of least squares as determined by a Hewlett-Packard 9830A calculator; (b) inset, a graphic representation showing the synthesis of classes present in the 3H-labeled RNA.

1.75 mg/ml), whereas the 5 min and later RNAs efficient competitors. This suggests that class Li is synthesized at least up to 13 min after infection and that additional temporal classes are produced from L-DNA starting at 5 are more

min.

The synthesis of class L2, which starts rather abruptly at 5 min and continues until 13 min, is documented in Fig. 11 and 12. Five-minute RNA is a better competitor than 3-min RNA (representing L1) in experiments with either 5to 9-min or 10- to 13-min 'H-labeled RNA. We

r,

VOL. 19, 1976

RNA CLASSES SYNTHESIZED DURING SP82 INFECTION

369

at a concentration of 1.75 mg/ml; these data yield an ordinal intercept of 1.0 in the double reciprocal plots shown in Fig. 14b and 15b. This indicates that the temporal classes present in the 0- to 5-min 3H-labeled RNA (classes Hi, H2, Li, and L2) are produced when protein synthesis is inhibited. These same classes are produced in vitro from SP82 DNA by RNA polymerase extracted from uninfected cells, whereas in vitro synthesis by the modified RNA polymerase extracted from SP82-infected cells yields RNA classes characteristic of later stages of infection (In Losick and Chamberlin [ed.], RNA Polymerase, in press). Figures 14a and b show that the binding of 5to 9-min 3H-labeled RNA to the H-DNA is competed less efficiently by CM-RNA, whereas this competitor has little effect on the binding of 9to 13-min 3H-labeled RNA and only a slight effect on the binding of 17- to 20-min 3H-labeled RNA to the H fraction of DNA. Correlation with analyses of temporal RNA classes detected in experiments with unfractionated SP82 DNA. Earlier analyses using unfractionated filter-bound DNA from SP82 (18) lead to the conclusion that there are six temporal RNA classes synthesized during these infections: class 1, synthesized from 0 to 3 min to 7 min after infection; class 2, from 4 to 12 min; class 3, from 4 to 17 min; and classes 4, 5, and 6 from 4, 7, and 15 min after infection, respectively, to the same time of lysis. These classes were defined on the basis of experiments utilizing competitor preparations isolated during, and subsequent to, the labeling interval. The six classes found in these experiments can be related in part to the temporal classes detected with isolated H and L fractions of DNA. For example, since the times of onset and duration of synthesis are the same for classes 4 and rzzz L3 L22ZZZZZ L - DN A H2, classes 5 and H3, and classes 6 and H5, it seems reasonable that they correspond. In addiL YZZZZZ tion, class 1 and class Hi are close in time of 5 10 15 20 25 30 35 synthesis; however, class 1 probably includes a min large contribution from classes Li and L2. Classes H4 and L3, both of which contribute very little to the total RNA synthesized during the lytic cycle, would not have been detected in the previous study. rzzzzzzz. H5 None of the classes measured in the present FZZ,ZA H4 - D N A experiments corresponds to classes 2 and 3 reH H3 ported earlier (18). The existence of these 1.I5~ 5k3 vzzzz H2 was postulated from experiments using classes H RNAs extracted from cells labeled 0 to 5 min m~~m and 5 to 10 min after infection (Fig. 1 and 2, min reference 18). Comparisons of the levels of comFIG. 13. The temporal classes of RN,A produced petition attained when these two 3H-labeled from the H and L fractions. Time intervalIs are shown RNAs were annealed in the presence of 10-, 15-, and 20-min RNAs were interpreted to mean with an estimated accuracy of +2 min.

define the RNA present in the 5-miin preparation which is responsible for this incrreased efficiency of competition as temporal cIEass L2. On the other hand, only unlabeled RN]ks isolated after 5 min yield complete competiti on with 5to 9-min and 10- to 13-min 3H-labeled RNA (i.e., ordinal intercepts of 1.0, Fig. 111b and 12b). This indicates the presence of an additional temporal class whose synthesis begiris between 5 and 7 min and continues until at le,ast 13 min (class L3). The analysis of classes of RNA pro'luced from L-DNA was not continued beyond 13 min after infection because of two considerati ons. First, earlier studies as well as data pr(esented in Table 1 show that very little of the tiotal phage RNA (less than 2%) is transcribed firom the LDNA from approximately 15 min afte,r infection to lysis. Secondly, this level of biinding approaches the antimessenger contenLt of RNA preparations extracted from cells lab eled at late times in infection (Table 2) making it difficult to determine if the competition obserrved is due to RNA-DNA or RNA-RNA duplex Iformation. The results of the H and L fractioin analyses are summarized in Fig. 13. CM-RNA. Figure 14 and 15 compaIre the patterns of competition observed when RNAs labeled at various times during the infection are hybridized to H-DNA or L-DNA in tble presence of RNA extracted from cells infected with SP82 after addition of chloramphenicol (OCM-RNA). The bindings of 0- to 5-min 3H-label ed RNA to either the H fraction (Fig. 14a) or the^ L fraction (Fig. 15a) can be reduced 80 to 90% b3y CM-RNA

r-

370

LAWRIE, SPIEGELMAN, AND WHITELEY

J. VIROL.

14b H fraction DNA

x

5.0 x

3.0

x x xX,X

I

/~~~~ I

1.0 01

02

03

15b L fraction DNA S6100

5.0

0 r-

60

*-K--x- x 5-9' 1.0 --0o-000 -o 0-5'

a)

0 "

10-13'

Z,__ ,

20 1.0

2.0

QI1

0.2

0.1

0.2

5.0 3.0 1.0

10 20 mg/ml Competitor RNA

CIT/CT

0.3

FIG. 14. Competition-hybridization experiments with the H fraction of SP82 DNA, a series of 3H-labeled RNAs and CM-RNA as the competitor RNA. The concentrations of 3H-labeled RNA and DNA used in these experiments are shown in Table 1. Data are plotted as described in the legend to Fig. 3 through 8, 10 through 12, and 16. FIG. 15. Competition-hybridization experiments with the L fraction of SP82 DNA, a series of 3H-labeled RNAs and CM-RNA as the competitor RNA. The concentrations of WH-labeled RNA and DNA used in these experiments are shown in Table 1. Data are plotted as described in the legend to Fig. 3 through 8, 10 through 12 and l6. FIG. 16. Competition-hybridization analyses with isolated H and L fractions of SP82 DNA, labeled RNAs, and competitor RNAs isolated at various times postinfection. (a) The results are plotted as percentage of radioactivity bound in the presence of competitor RNA relative to the radioactivity bound in the absence of competitor RNA as a function of the concentration of competitor RNA added to the reaction mixture; the concentrations of 3H-labeled RNA and DNA used in these experiments are shown in Table 1. (b) The results are plotted as 1i1-F as a function of cT/cT, where F is the fraction of radioactivity bound relative to the radioactivity bound in the absence ofcompetitor RNA, c' is the concentration of 3H-labeled RNA and CT is the concentration of competitor RNA added to the reaction mixture; the lines are fitted by the method of least squares as determined by a Hewlett-Packard 9830A calculator; (b) inset, a graphic representation showing the synthesis of classes present in the WH-labeled RNA.

VOL. 19,. 1976

RNA CLASSES SYNTHESIZED DURING SP82 INFECTION

that the synthesis of both classes 2 and 3 started before 5 min, and that the synthesis of class 2 stopped between 10 and 15 min, whereas that of class 3 stopped between 15 and 20 min. For this reason, competition experiments were performed with H-DNA and RNA preparations isolated subsequent to the labeling interval. While different levels of competition were observed when 9-, 15-, and 25-min competitor preparations were tested with 5- to 9-min 3H-labeled RNA (Fig. 16a), the double reciprocal plot (Fig. 16b) indicates that data obtained with both the 9- and 15-min RNAs extrapolates to an ordinal intercept of 1.0 (i.e., complete competition). The fact that the 25-min RNA does not yield complete competition in experiments with 5- to 9-min 3H-labeled RNA can be interpreted in two ways: the cessation of the synthesis of some RNA species between 9 and 25 min and their subsequent degradation, or a change in the relative abundance of RNA species due to the onset of synthesis of other RNAs. If the competition pattern in Fig. 16 were due to the first of these possibilities, it could be concluded that a class of RNA populations analogous to class 3 is produced from H-DNA. However, as shown earlier, class H5, which accounts for a large proportion of the RNA present late in infection (Fig. 13), is synthesized beginning at 13 min. Therefore, the results in Fig. 16 could be explained equally well as reflecting a decrease in the relative abundance of the earlier RNA. The ambiguity of this result is inherent in competition experiments using competitor RNAs isolated later than the labeling interval. Thus, the existence of a class analogous to class 3 cannot be proven or excluded on the basis of the present data. Other experiments in progress in this laboratory analyzing the proteins synthesized during SP82 infection may help to resolve this uncertainty. Finally, the data shown in Fig. 16 differ from those found with unfractionated DNA in that the 15-min unlabeled RNA yields complete competition with RNA labeled early in infection. The class of RNA (class 2) previously postulated on the basis of incomplete competition with 15-min RNA cannot be detected in experiments with H-DNA. One possible explanation for the difference in these results is that experiments utilizing total DNA yielded results that reflect the strand switch as a temporal class of RNA. The percentage of total RNA binding to L-DNA drops from approximately 20% in the 0to 5-min interval to less than 3% by 15 min after infection. Assuming reasonable rates of degradation of mRNA, the L-DNA transcripts would disappear and experiments using unfractionated DNA would show incomplete competition,

371

suggesting the existence of an additional temporal RNA class. Comparison of temporal RNA classes produced in SP82 and in T4 infections. Studies of the infection of E. coli with the phage T4 have provided much of the framework for ideas concerning temporal regulation of gene expression (reviewed in reference 24). It is of interest, therefore, to compare the regulation seen in T4 with that presented here for the infection ofB. subtilis with SP82. Competition-hybridization experiments with the RNA produced during T4 infection have detected four temporal classes: immediate early, delayed early, quasi-late, and truly late RNAs. Immediate early and delayed early RNA, whose synthesis is initiated from the L-strand before 3 min after infection (8), are distinguished by the fact that only immediate early RNA is synthesized in the presence of chloramphenicol. Considerable evidence has been presented that transciption from immediate early promoters can be extended to delayed early genes in the presence of a phage-coded protein which functions as an antiterminator (10). Synthesis of quasi-late RNA is initiated before the onset of DNA replication whereas production of truly late RNA, which occurs from the H strand (8), is coupled with DNA replication. The synthesis of both late classes requires the action of two phage genes coding for proteins which bind to RNA polymerase isolated from T4-infected cells (19). Transcription of late genes results from initiations which the host enzyme is unable to perform; however, it is not known which of the several modifications of the host polymerase are required for these initiations. In attempting to correlate the classes detected in the T4 infection with those presented here for the SP82 infection, it is useful to consider the classes of RNA synthesized in vitro by RNA polymerases isolated from phage-infected and uninfected cells. As described earlier (17; Losick and Chamberlin [ed.], RNA Polymerase, in press), this enzyme is substantially modified after infection of B. subtilis with SP82. RNA polymerase isolated 6 min after infection contains approximately one-half the amount of sigma found in purified host polymerase and has an additional polypeptide with a molecular weight of 28,000 (Spiegelman and Whiteley, unpublished observations). At least two other phage-induced polypeptides are added to the modified enzyme as infection proceeds since three major small molecular weight polypeptides (28,000, 26,000, and 13,000) co-purify with the polymerase extracted from cells 10 min after infection. Different forms of the modified enzyme containing predominantly the 28,000-

372

LAWRIE, SPIEGELMAN, AND WHITELEY

the 26,000-molecular-weight polypeptides (the 13,000-molecular-weight polypeptide appears in both forms) can be isolated by column chromatography. We have demonstrated (Losick and Chamberlin [ed.], RNA Polymerase, in press) that each of these forms of the modified enzyme can synthesize all RNA classes except Hi in vitro after very short periods of incubation (30 s and 2 min). The in vitro synthesis is asymmetric and has good fidelity since at least one gene transcript (mRNA for dCMP deaminase) can be translated in vitro into an active protein (Losick and Chamberlin [ed.], RNA Polymerase, in press). Isolated polymerase from uninfected cells can synthesize only Hi, H2, Li, and L2 in vitro. For comparison, E. coli host polymerase produces both immediate and delayed early classes in vitro from T4 DNA in the absence of the terminator protein rho; if rho is present, transcription is restricted to immediate early sequences (11). Analysis of RNAs produced in vitro by the enzyme extracted from T4-infecting cells is complicated by the fact that transcription is symmetrical (20). Class Hi and Li RNAs made in SP82 infections are clearly analogous to the immediate early RNA produced in T4 infections. Temporally, classes H2 and L2 resemble delayed early T4 RNAs, but differ from the latter because they can be synthesized in the presence of chloramphenicol. However, it seems unlikely that H2 is produced by the host enzyme solely by read-through from Hi promoters since transcription of these sequences continues long after production of Hi ceases. Since H2 is also synthesized in vitro by the modified polymerase, it is conceivable that it is produced both by read-through from Hi promoters by the host polymerase and by recognition of H2 promoters by the modified enzymes. Thus, all four classes produced by the host enzyme in vitro resemble immediate early T4 classes and none of the SP82 classes appears analogous to the delayed early T4 classes. The absence of RNAs of the latter type may reflect differences in the mechanisms of termination functioning in the two infections. Since synthesis of the remaining SP82 classes in vivo requires phage protein synthesis and these classes are produced in vitro by the modified polymerases after very short periods of incubation, it seems likely that they result from recognition of new initiators although an antitermination mechanism is not excluded. Temporally, the H3, L3, and H4 classes are analogous to quasi-late RNA in T4 infections, whereas the time of appearance of H5 is similar to truly late T4 RNA. The production of this latest class also occurs in vitro suggesting that or

J. VIROL.

late gene expression in vivo may not be coupled to DNA replication (21). Thus, transcription of genes coding for these middle and late classes may be regulated solely by the sequential modification of the polymerase. Additional mechanisms of regulation may be involved, however, in the decrease or shutoff in the synthesis of classes H2 and H4, whereas the shutoff in the synthesis of class Hi may be a consequence of enzyme modification. ACKNOWLEDGMENTS This research was supported by grant VC-46B from the American Cancer Society and by Public Health Service grant GM-20784 from the National Institute of General Medical Sciences. J.M.L. gratefully acknowledges support from the Developmental Biology Training Program, grant no. HD 00266 of the National Institute of Child Health and Human Development. H.R.W. is a recipient of Research Career Award K6-GM-442 from the National Institute of General Medical Sciences. LITERATURE CITED 1. Bishop, J. O., F. W. Robertson, J. A. Burns, and M. Melli. 1969. Methods for the analysis of deoxyribonucleic acid-ribonucleic acid hybridization data. Biochem. J. 115:361-370. 2. Chakraborty, P. R., P. Bandyopadhyay, H. H. Huang, and U. Maitra. 1974. Fidelity of in vitro transcription of T3 deoxyribonucleic acid by bacteriophage T3-induced ribonucleic acid polymerase and by Escherichia coli ribonucleic acid polymerase. J. Biol. Chem. 249:6901-6909. 3. Esche, H., and H. Ch. Spatz. 1973. Asymmetric transcription of SPP1 in vivo. Mol. Gen. Genet. 124:57-63. 4. Gage, L. P., and E. P. Geiduschek. 1971. RNA synthesis during bacteriophage SPOl development: six classes of SPOl RNA. J. Mol. Biol. 57:279-300. 5. Gage, L. P., and E. P. Geiduschek. 1971. RNA synthesis during bacteriophage SPOl development. II. Some modulations and prerequisites of the transcription program. Virology 44:200-210. 6. Geiduschek, E. P., and 0. Grau. 1970. T4 anti-messenger, p. 190-203. In L. Silvestri (ed.), First International Lepetit Colloquium. American Elsevier Publishing Co., Inc., New York. 7. Guha, A., and W. Szybalski. 1968. Fractionation of the complementary strands of coliphage T4 DNA based on the asymmetric distribution of the poly U and poly U,G binding sites. Virology 34:608-616. 8. Guha, A., W. Szybalski, W. Salser, A. Bolle, E. P. Geiduschek, and J. R. Pulitzer. 1971. Controls and polarity of transcription during bacteriophage T4 development. J. Mol. Biol. 59:329-349. 9. Hansen, J. N., G. Spiegelman, and H. 0. Halvorson. 1970. Bacterial spore outgrowth: its regulation. Science 168:1291-1298. 10. Jayaraman, P. 1972. Transcription of bacteriophage T4 DNA by Escherichia coli RNA polymerase in vitro: identification of some immediate-early and delayedearly genes. J. Mol. Biol. 70:253-263. 11. Lavalle, R., and G. De Hauwer. 1968. Messenger RNA synthesis during amino acid starvation in Escherichia coli. J. Mol. Biol. 36:269-288. 12. Lawrie, J. M., G. B. Spiegelman, and H. R. Whiteley. 1975. DNA strand specificity of transcripts produced in vivo and in vitro by RNA polymerase from SP82infected Bacillis subtilis. J. Virol. 15:1286-1288. 13. Nygaard, A. P., and B. D. Hall. 1963. A method for the detection of RNA-DNA complexes. Biochem. Biophys. Res. Commun. 12:98-104.

VOL. 19, 1976

RNA CLASSES SYNTHESIZED DURING SP82 INFECTION

14. Schachtele, C. F., D. V. DeSain, and D. L. Anderson. 1973. Transcription during the development of bacteriophage 029: definition of "early" and "late" 029 ribonucleic acid. J. Virol. 11:9-16. 15. Schachtele, C. F., C. V. DeSain, L. A. Hawley, and D. L. Anderson. 1972. Transcription during the development of bacteriophage 029: production of host- and 029-specific ribonucleic acid. J. Virol. 10:1170-1178. 16. Sheldrick, P., and W. Szybalski. 1967. Distribution of pyrimidine "clusters" between the complementary DNA strands of certain Bacillus bacteriophages. J. Mol. Biol. 29:217-228. 17. Spiegelman, G. B., and H. R. Whiteley. 1974. Purification of ribonucleic acid polymerase from SP82-infected Bacillus subtilis. J. Biol. Chem. 249:1476-1482. 18. Spiegelman, G. B., and H. R. Whiteley. 1974. In vivo and in vitro transcription by ribonucleic acid polymerase from SP82-infected Bacillus subtilis. J. Biol. Chem. 249:1483-1489. 19. Stevens, A. 1972. New small polypeptides associated with DNA dependent RNA polymerase of Escherichia

20. 21.

22.

23. 24.

373

coli after infection with bacteriophage T4. Proc. Natl. Acad. Sci. U.S.A. 69:606-607. Stevens, A. 1974. Deoxyribonucleic acid dependent ribonucleic acid polymerases from two phage-infected systems. Biochemistry 14:493-503. Stewart, C. R., B. Click, and M. F. Tole. 1972. DNA replication and late protein synthesis during SP82 infection of Bacillus subtilis. Virology 50:653-663. Summers, W. C., and W. Szybalski. 1968. Totally asymmetric transcription of coliphage T7 in vivo: correlation with poly G binding sites. Virology 34:9-16. Truffaut, N., B. Revet, and M. 0. Soulie. 1970. Etude comparative des DNA de phages 2C, SP8*, SP82, 0e, SPOl, et SP50. Eur. J. Biochem. 15:391-400. Wu, R., E. P. Geiduschek, D. Rabussay, and A. Cascino. 1973. Regulation of transcription in bacteriophage T4-infected E. coli-a brief review and some recent results, p. 181-204. In C. F. Fox and W. S. Robinson, (ed.), Virus research, ICN-UCLA Symposium on Molecular Biology. Academic Press, New York.