E. Peter Geiduschek

Transcription during bacteriophage T4 development: Synthesis and relative stability of early and late RNA☆☆☆

The control of RNA synthesis during infection of Escherichia coli B with bacteriophage T4 has been studied. RNA-DNA hybridization-competition experiments show that between one-third and one-half of the bacteriophage-specific RNA present at 20 minutes after infection at 30 °C consists of RNA species—the socalled “late” species—which are on the average several hundredfold less abundant five minutes after infection. The remainder of the 20-minute RNA consists of species—usually referred to as “early”—that are readily detectable in RNA extracted five minutes after infection. The late RNA species include some that are relatively much more abundant than others. We have investigated the relative stabilities of these two classes of RNA during the late stages of infection by comparing the quantitative partition, between early and late species, of a radioactive precursor incorporated into RNA in widely varying regimes of labelling. The proportion of label in early and late RNA species is constant regardless of the length of the labelling period, leading to the conclusion that the early and late species have comparable average decay times. Measurements of the labelling of late messenger have been made at several times during the first five minutes of infection to test the possibility that all T4 genes might be transcribed for a brief period after the injection of the phage genome into the host cell. Late RNA could not be detected at the beginning of infection. These results support the conclusion that the pattern and evolution of late RNA abundances during T4 development has its origin in changing rates of transcription and not in changing stabilities.

Transcription during bacteriophage T4 development: requirements for late messenger synthesis.

Abstract The prerequisites of late T4 messenger synthesis have been investigated in phageinfected Escherichia coli B, using the methods of DNA-RNA hybridizationcompetition. The absolute requirement of some DNA synthesis for viral late messenger synthesis has been demonstrated in three ways: (1) several DNA synthesis-negative (DO) amber mutants located in 3 different segments of the viral chromosome have been shown to be defective in late messenger synthesis in the non-permissive host; (2) when phage DNA synthesis is made thymine dependent by the proper choice of viral mutant and host, viral late messenger synthesis also becomes thymine-dependent, but early messenger synthesis does not; (3) ultraviolet-irradiated T4+ which do not replicate their DNA do not synthesize viral late messenger. However, DNA synthesis and late messenger synthesis are not linearly related. For example, late messenger synthesis can be sustained in cells infected with T4 gene 30 amber mutants in which DNA synthesis has almost ceased. DNA replication is not a sufficient requirement for late messenger synthesis: the function of at least one, and possibly more, “maturation gene” products is also required. Thus, a maturation-defective mutant, lacking the function of gene 55 ( am BL 292), synthesizes DNA but no late messenger. A mutant in T4 gene 33 ( am N134) is of the same type but “leakier” and eventually makes readily detectable proportions of late viral RNA. A viral dCTPase, the product of T4 gene 56, is synthesized early in normal T4 development. This enzyme helps to prevent the synthesis of cytosine-containing viral DNA. dCTPase-negative T4 amber mutants, even two multiple mutants that synthesize appreciable quantities of viral DNA, are defective in late messenger synthesis. Evidently, two mechanisms must be involved in the synthesis of T4 late messenger: (1) a competent form of DNA must be made available in the infected cell; and (2) one or more “maturation” gene products must act on this competent DNA. T4 gene 55 is likely to specify one such product.

On the factors controlling the reversibility of DNA denaturation

The experiments described in this paper distinguish two processes which lead to a reversal of DNA denaturation and to the re-formation of long-stacked arrays of nucleotide pairs. Type I reversibility is a very rapid intramolecular process which occurs in DNA preparations from viral, bacterial and animal sources, in solvents varying greatly in ionic strength. Type II reversibility (“renaturation”, Marmur & Lane, 1960 ; Doty, Marmur, Eigner & Schildkraut, 1960 ; Marmur & Doty, 1961 ) has the properties of an intermolecular process, is much slower, and occurs only in relatively homogeneous DNA preparations at moderately high ionic strength. Data on the dependence of type I reversibility upon DNA size and composition are consistent with the proposal that type I reversibility is controlled by GC-rich nucleotide pair sequences which act as “nuclei” for the re-establishment of long-range order. Under the conditions of these experiments, such nuclei control the re-stacking of extremely long sections of polynucleotide chain. Consequently, type I reversibility is strongly affected by molecular weight changes in the range 0·5 to 10 × 10 6 . GC-rich “nuclei” may, however, be dissociated by heating. The loss of type I reversibility is accordingly associated with the final stages of the dissociation of a double helix. However, if complementary polynucleotide chains of DNA are linked covalently, type I reversibility is permanent. (Such material is called “reversible” DNA.) The interpretation of experiments on the stability of DNA secondary structure, especially those involving bacterial transformation assays, is discussed in the light of these results.

Nonaqueous solutions of DNA. Denaturation in methanol and ethanol

Abstract Solutions of DNA in a variety of nonaqueous solvents can be prepared by dialysis in the absence of electrolyte. The properties of DNA in two of these solvents, ethanol and methanol, have been investigated in some detail. In both solvents, DNA may be essentially molecularly dispersed to give solutions that are completely stable. Both macromolecular and optical criteria—increased sedimentation velocity, lowered viscosity and radius of gyration, high ultraviolet absorbance—indicate that, in these solutions, DNA is denatured. The denaturation may be largely reversed by dialysis back to an aqueous salt-containing solvent medium. Both methanol and ethanol lower the thermal denaturation temperature of DNA in the presence of electrolyte. The significance of these observations for the consideration of factors determining stability of the DNA helix is discussed.

Structure and function of cross-linked DNA: I. Reversible denaturation and Bacillus subtilis transformation

One consequence of the reaction of HNO 2 with DNA is the covalent linkage of complementary polynucleotide chains. The introduction of a single cross-link into a DNA molecule confers the ability to undergo reversible denaturation. The rate of introduction of cross-links into Bacillus subtilis DNA has been measured at pH 4·15, 4·50 and 5·00 at 25°C. Comparison with published data on rates of deamination of cytosine and purines in DNA shows that formation of cross-links is, relative to deamination, a frequent event (approximately 1 cross-link per 4 deaminations at pH 4·15). Heat-stable transforming activity and reversibly denaturable DNA appear simultaneously upon reaction with HNO 2 , establishing the fact that cross-linked B. subtilis DNA is active in transformation. The appearance of heat stability is in competition with, and approximately 10 times faster than, the well-known concomitant inactivation.

RNA synthesis during bacteriophage SPO1 development: six classes of SPO1 RNA.

Abstract An analysis of RNA synthesis during development of Bacillus subtilis phage SPO1 is presented. Host transcription decreases substantially after infection. The inhibition of host transcription is non-co-ordinate, with messenger synthesis much more strongly affected than stable RNA synthesis. Viral RNA synthesis increases concurrently with this shut-down and eventually accounts for at least half of the total transcription. The program of viral transcription has been analyzed by DNA-RNA hybridization-competition. Six classes of viral transcripts can be distinguished on the basis of their times of first appearance and cessation of synthesis. Viral RNA first appears about one minute after formation of phage-bacterial complexes, e and em viral RNA start to be synthesized at that time and e RNA synthesis lasts only until about the fourth minute, while em RNA synthesis is shut-off soon after the start of viral DNA replication, m and m1l, RNA synthesis starts four to five minutes after infection; m transcription ceases soon after the onset of DNA replication whereas m1l, transcription continues throughout the rest of the viral eclipse period. m2l and l transcripts first appear a little before, and several minutes after, the start of replication, respectively; m2l and l RNA continue to be made throughout the rest of the viral eclipse period.

Nonaqueous solutions of DNA. Reversible and irreversible denaturation in methanol

Abstract Deoxyribonucleic acid (DNA) undergoes denaturation in methanol-water mixtures. The region of solvent composition in which this transition occurs depends on the ionic strength and temperature but is little influenced by the average nucleotide composition of the DNA. As judged by both macromolecular and optical criteria, the disruption of secondary structure is almost complete, yet methanol denaturation is rapidly and readily reversible by water. The structural basis of this reversibility differs from that controlling the reversibility of partial denaturation in aqueous solution. An irreversible thermal transition has also been found to occur in methanol-rich solvents. The products of this transition may be reconverted to native DNA by water.

Controls and polarity of transcription during bacteriophage T4 development

Abstract Hybridization and hybridization-competition experiments, using the separated l and r strands of bacteriophage T4 DNA and 3H-labeled RNAs coded by T4 and its mutants, permit one to study the transcriptional controls for various phagespecific RNA species. In the present study, the T4-coded RNAs have been classified according to: (i) orientation of transcription; (ii) time of appearance; (iii) requirement for protein and DNA synthesis and for expression of the gene 33 and 55 functions; and (iv) true-early or quasi-late behavior, as defined by the respective decrease or increase in the early appearing RNA species during the second half of the viral eclipse period. The “immediate-early” messenger RNAs, produced after phage infection in the absence of protein synthesis, are transcribed counterclockwise from the l strand of T4 DNA. The “delayed-early” RNA is transcribed predominantly from the l strand, whereas the “true-late” RNA, the synthesis of which starts later than five minutes after infection, is mainly transcribed clockwise from the r strand, i.e. with the same orientation as T4 DNA replication. There is only a small difference in the 20-minute l-strand transcription between the wild-type T4 and its mutants blocked in DNA synthesis (D0; genes 42 and 43) or in the late functions (MD, maturation defective) controlled by genes 33 and 55, as revealed by hybridization-competition experiments. Also, 5-minute unlabeled T4-coded RNA competes with almost 90% of the labeled 20-minute RNA complementary to the l strand, indicating that, at most, 10 to 12% of the 20-minute l-specific RNA belongs to the true-late class. Experiments employing r strands show that the 20-minute RNA of the D0 and MD mutants is quite deficient in RNA species coded by wild-type T4 at 20 minutes after infection. Similarly, under non-permissive conditions, mutations in gene rII and in genes 46 and 47, the latter two controlling nuclease activity, depress the r-strand transcription, but mutation in gene 30, which controls T4 ligase, does not have such an effect. The influence of chloramphenicol inhibition at various stages of T4 development on the l and r strand transcription was also evaluated. The activity of the gene 55 product and of DNA replication are continuously required for efficient r strand transcription.

Essential Roles of Bdp1, a Subunit of RNA Polymerase III Initiation Factor TFIIIB, in Transcription and tRNA Processing

ABSTRACT The essential Saccharomyces cerevisiae gene BDP1 encodes a subunit of RNA polymerase III (Pol III) transcription factor (TFIIIB); TATA box binding protein (TBP) and Brf1 are the other subunits of this three-protein complex. Deletion analysis defined three segments of Bdp1 that are essential for viability. A central segment, comprising amino acids 327 to 353, was found to be dispensable, and cells making Bdp1 that was split within this segment, at amino acid 352, are viable. Suppression of bdp1 conditional viability by overexpressing SPT15 and BRF1 identified functional interactions of specific Bdp1 segments with TBP and Brf1, respectively. A Bdp1 deletion near essential segment I was synthetically lethal with overexpression of PCF1-1, a dominant gain-of-function mutation in the second tetracopeptide repeat motif (out of 11) of the Tfc4 (τ131) subunit of TFIIIC. The analysis also identifies a connection between Bdp1 and posttranscriptional processing of Pol III transcripts. Yeast genomic library screening identified RPR1 as the specific overexpression suppressor of very slow growth at 37°C due to deletion of Bdp1 amino acids 253 to 269. RPR1 RNA, a Pol III transcript, is the RNA subunit of RNase P, which trims pre-tRNA transcript 5′ ends. Maturation of tRNA was found to be aberrant in bdp1-Δ253-269 cells, and RPR1 transcription with the highly resolved Pol III transcription system in vitro was also diminished when recombinant Bdp1Δ253-269 replaced wild-type Bdp1. Physical interaction of RNase P with Bdp1 was demonstrated by coimmunoprecipitation and pull-down assays.

Transcription during bacteriophage SPO1 development: mutations affecting the program of viral transcription.

Abstract Bacteriophage SPO1 host-dependent mutants in three cistrons show defects in the program of viral transcription (Gage & Geiduschek, 1971a). Mutants in two eistrons ( sus F4 and sus F14) are analogous to T4 maturation defective mutants: they replicate viral DNA in the non-permissive host but are unable to make viral late l or m 2 l , RNA and do not form protein subassemblies of the virion. These mutants are also unable to shut off viral m transcription. A mutant in a third cistron ( sus F21) only transcribes that viral RNA ( e , em ) for which no protein synthesis in the infected cell is required. Non-permissive sus F21 infected cells fail to make m , m 1 l , or m 2 l , RNA and almost entirely cease viral transcription eight to ten minutes after infection (at 37 °C). These transcriptive defects suggest a simple hierarchical ordering of the normal temporal sequence of wild-type viral transcription. That ordering, in turn, suggests a model of transcription during viral development that involves not less than three control elements of viral transcription, two of which are positive and one of which is negative.