The Function and Optimal Sequence of the Phage \Boxa Transcription Antitermination Signal

$The Function and Optimal Sequence of the Phage \Boxa Transcription Antitermination Signal$

Downloaded from genesdev.cshlp.org on September 30, 2021 - Published by Cold Spring Harbor Laboratory Press

Transcription-dependent competition for a host factor: the function and optimal sequence of the phage \boxA transcription antitermination signal

David I. Friedman/ Eric R. Olson,^ Linda L. Johnson,^ Diane Alessi,* and Mark G. Craven* ^Department of Microbiology and Immunology, University of Michigan Medical School, Ann Arbor, Michigan 48109-0620 USA; ^Molecular Biology Research, Upjohn Company, Kalamazoo, Michigan 49001 USA

Ordered development of lambdoid phages relies on systems of transcription termination and antitermination. The phage-encoded N early regulatory proteins, acting with the Nus proteins of Escbericbia coli, modify RNA polymerase to a form that overrides many transcription termination signals. These modifications require cis- acting sites, nut, located downstream of the early phage promoters. The nut sites in phages k, 21, and P22, which share similarities but are not identical, contain two signals, boxA and boxB. We demonstrate that although a consensus sequence for the boxA signal (boxAcon), 5'CGCTCTTTA, is found only in P22, changes to consensus in the nut^ sites of X and 21 create more effective antitermination signals than the wild-type signals. An in vivo competition assay demonstrates that a k nut region with boxAcon outcompetes nut regions with wild-type, as well as other variations of the boxA sequence, for the host NusB protein. This suggests that boxA influences NusB activity in N-mediated antitermination. Successful competition by boxAcon requires transcription of the nut site as well as N activation. Nucleotide replacement further demonstrates that bases at both ends of boxA are important for antitermination. [Key Words: nut; boxA; boxAcon termination-, antitermination-, regulation] Received June 18, 1990; revised version accepted September 4, 1990.

Regulation of gene expression by systems of termination boxB sequence is the recognition element for gpN and antitermination of transcription have been well de (Doelling and Franklin 1989; Lazinski et al. 1989). In scribed for prokaryotes, and recent studies indicate that vitro studies have confirmed that Nus factors are re this mode of regulation is operative in eukaryotes (for quired for N-mediated antitermination (Greenblatt and review, see Piatt 1986; Friedman et al. 1987; Proudfoot Li 1981; Das and Wolska 1984; Ghosh and Das 1984; 1989; Spencer and Groudine 1990). The best described of Das et al. 1985), cooperating with gpN to alter RNA these systems are found in phage X and its Escherichia polymerase in response to the nut signal (Barik et al. coh host. In prokaryotes, signals for conversion of RNA 1987; Horwitz et al. 1987). polymerase into elongation complexes^ capable of The boxA transcription signal was subsequently reading through multiple downstream termination shown to be present in variant forms (Table 1) in the nut signals have been identified both within, and down regions of other lambdoid phages (Olson et al. 1982; stream of, promoters (for review, see Friedman and Got- Friedman and Gottesman 1983; Franklin 1985) as well tesman 1983; Morgan 1986; Friedman 1988a; Roberts as in strategic locations in some operons of E. coh. These 1988). bacterial boxA signals, depending on their location, ap The \ nut regions located downstream of the early pear to influence either transcription termination promoters, PL and p^, contain the signals recognized by (Stewart and Yanofsky 1985; Rosenthal and Calvo 1987) the proteins of the N transcription antitermination or antitermination (Aksoy et al. 1984; Holben and complex (Fig. 1; Salstrom and Szybalski 1978). These Morgan 1984; Li et al. 1984; Morgan 1986; Berg et al. proteins include the phage protein gpN (gene product of 1989). Although all of the boxA sequences exhibit signif N] and the products of the E. coh nus genes (for review, icant homology, most show some variation from the see Friedman et al. 1984). Two signals have been identi consensus sequence. fied in the nut region: (1) boxB, a 16-bp region of hy Two other lambdoid phages with N-like antitermina phenated dyad symmetry (Rosenberg et al. 1978; Soma- tion systems, which are functionally and structurally sekhar et al. 1982); and (2) boxA, which is upstream of similar to that of X, are coliphage 21 and Salmonella ty- boxB in both nut regions (Olson et al. 1982; Fig. 1). The phimurium phage P22. These phages have iV-like genes

2210 GENES & DEVELOPMENT 4:2210-2222 © 1990 by Cold Spring Harbor Laboratory Press ISSN 0890-9369/90 $1.00 Downloaded from genesdev.cshlp.org on September 30, 2021 - Published by Cold Spring Harbor Laboratory Press

BoxA transcription signal

imm21 of these phages were derived from X. Ximmll is particu

immP22 larly useful, because P22 normally does not infect E. coli, whereas Ximmll has the X host range. clll N cl cro IS2 ell O P p-nin 5 -, Q The various N reactions can be functionally distin MAP , 1 1.^ ' ' guished by employing E. coli variants with altered nus genes (Friedman et al. 1984). For example, an E. coli with tL1 "^^^ "^^f^ tR1 tlS2 tR:2 3 4 a chimeric nusA gene, nusAs.t., comprised of the 5' 85% B X from S. typhimuiium and the 3' 15% portion from E. C coli, supports growth of Ximmll at all temperatures and Ximmll at lower temperatures, but fails to support D growth of X at any temperature (Baron et al. 1970; Friedman and Baron 1974; Schauer et al. 1987; A.E. TAAATAACCCCGCTCTTACACATTCCAGCCCTGAAAAAGGGCA A. Granston, M. Craven, A. Schauer, D. Thompson, and 3'end of boxA boxB D. Friedman, in prep.). A X mutant that is able to utilize cro nusAsx. has mutations in the N gene and ^OXAR Figure 1. Relevant genes, regulatory signals, and transcription (Friedman and Olson 1983). The boxA^ mutation, patterns of the \ early regulatory region (map not drawn to boxAl, results in a boxA with three Ts at the 3' end as scale). Shown are the relative placement of representative genes do the P22 and 21 boxA sequences. The boxAl sequence and regulatory signals. Early promoters are indicated by p^ and deviates from the consensus by not having a 3' A (Ta Pu termination signals by ovals, and nut sites by boxes. The ble 1). r32-IS2 insertion is shown above the map, and its termination The importance of boxA in the N-mediated antiter- signal is shown below the map. The region of strong termina mination reaction was underscored further by other tion removed by the ninS deletion is shown with the three dis tinguishable regions of termination identified. [A] Regions of boxA mutations that result in nut regions that are less substitution in the hybrid phages Ximmll and \imm22. [B] active as signals for directing N-mediated antitermina- Early transcription of the wild-type PR and PL operons under tion. A mutation in X boxAi^ (a C to G change at position N-deficient conditions with points of termination indicated. 3; Peltz et al. 1985), and one in X boxA^ (a G to T change Note that in the absence of gpN, 40-50% of transcription at position 2 called boxAS; Olson et al. 1984; Robledo et passes through t^i. (C) Rightward transcription in the absence of gpN when the r32-IS2 element with its strong terminator is present. [D] Readthrough of transcription terminators in the Table 1. boxA sequences from lambdoid phages right and presence of active gpN, Nus proteins, and nut signals. (£) DNA left nut legions and from the rmG leader region sequence of the X. nut^^ region showing the placement of the boxA and boxB sequences in relation to the upstream cio gene. Effect^ The boxA sequence is underlined, and the arms of the boxB hairpin are indicated by the converging arrows. A. Natural boxA Sequences 123456789 \boxA CGCTCTTAC llboxA^ TGCTCTTTA llboxAf^ CGCTCTTTA and cognate nut sequences at positions on their genomes XboxAj^ CGCTCTTAA analogous to the positions of the X N gene and nut llboxAi^ GGCTCTTTA signals (Friedman et al. 1973a; Hilhker and Botstein llboxAi, CGCTCTTTA 1976; Hilliker et al. 1978; Franklin 1985; Lazinski et al. rrnboxA TGCTCTTTA 1989). The amino acid sequences of the various gpNs, however, are significantly different (Franklin 1985; La B. Mutant lambdoid boxA sequences zinski et al. 1989). The nut regions of \, 21, and P22 have boxAf^ XboxAl CGCTCTTTCib + boxB sequences that differ in nucleotide composition XboxAS CTCTCTTAC^b but resemble each other in having hyphenated dyad XboxA16 CGCTATTAC^b 0 symmetries. The boxA sequences of these phages are XboxAcon CGCTCTTTA + similar, but not identical. The following minimal con XboxA^GC CTCTTACAC*'' sensus sequence is evident, 5'-CGCTCTTTA, and re llboxAcon CGCTCTTTA -I- gardless of the surrounding context, it will be called boxAi^ boxAcon {con = consensus). A complete match to the XboxA" CGGTCTTAASb consensus is found only in the nut regions of P22 "' (Friedman and Gottesman 1983; Franklin 1985). Nucleotide positions are numbered above. (A) Naturally occur ring boxA sequences; [B] mutant boxA sequences created in X Lambdoid phages share regions of homology, thus per and 21 nut regions. Changes away from wild type are under mitting the construction of hybrid phages by crossing lined. either P22 or 21 with \ (Liedke-Kulke and Kaiser 1967; *( -I-) Enhances; (-) reduces; (0) no effect (for details, see text). Gemski et al. 1972; Botstein and Herskowitz 1974). Hy •'References: ipriedman and Olson (1983); ^.Qlson et al. (1984); brids Ximmll and Ximmll (see Fig. 1) have acquired the 3Robledo et al. (1990); '^Doelling and Franklin (1989); ^Peltz et central control region, called the imm region, respec al. (1985). tively, from P22 and 21, while the rest of the genomes •^This mutation has not been named.

GENES & DEVELOPMENT 2211 Downloaded from genesdev.cshlp.org on September 30, 2021 - Published by Cold Spring Harbor Laboratory Press

Friedman et al. al. 1990) reduce N-mediated antitermination. Moreover, pressed more effectively by boxAcon are nusB5 and the deletions of the GC, as well as the GCT, at positions 2, hybrid nusAsx. in the presence of a second mutation 3; and 4 also reduce this activity (Doelling and Franklin sneA16. The sneA16 mutation that maps in the rplP 1989). An apparently contradictory result has been re gene partially suppresses the failure in gpN action ported by Zuber et al. (1987), M^ho found that a deletion caused by some nus mutations (A.T. Schauer, D.L. of boxA did not significantly influence N-imposed anti- Thompson, D. Alessi, and D.I. Friedman, in prep.). termination. However, other studies using the same de These differences were examined in more detail by letion reveal that it causes a fivefold reduction in N-me following the phage bursts through one round of growth. diated antitermination (D. Lazinski and A. Das, pers. The results of one such experiment are shown in Figure comm.). 2, where growth of X derivatives with the three boxAJ^ To further assess the activity of boxA in gpN action, sequences was compared in a nusBS host at 37°C. Note site-directed mutagenesis was employed to change nat that the burst of KboxAcon occurs —45 min earlier than ural boxA sequences to conform to the consensus se that of kboxAl, but eventually both phages produce the quence called boxAcon. Studies reported demonstrate same final burst. \boxA '*' has a longer delay and a sub the biological consequences of these changes. stantially lower final burst. We then examined the effect of a change to boxAcon in the nutR region of a Kimmll derivative (Table 2). The Results success of the selection scheme for KimmllboxAcon Effect of altered boxA sequences on I growth demonstrates a priori that boxAcon confers a growth ad vantage for Xjn2m21 in a nus variant, that is, the The growth of \ and kimmll phages with different boxA immllboxAcon recombinant grows at 42°C in an E. coli sequences was compared in E. coli derivatives carrying hybrid that has a nusAs,^ gene. Extension of this study to mutant nus genes using efficiency of plating (eop) as the other 12 us mutants showed that the efficiency of plating assay (see Table 2). of KimmllboxAcon is either greater than (observed in We first examine the effect of changes in the X^OXAR. hosts carrying one of the following alleles: nusA^x. or The boxAl mutation had been shown previously to sup nusAl or nusB5 alleles) or similar to Ximm21 (observed press the inhibitory effect of the nusAl and nusEJl mu in a host carrying the nusEll allele); of those tested, tations on \ growth (Friedman and Olson 1983; Schauer there is no nus mutant in which Ximm21 grows better et al. 1987). Since boxAl still differs from the derived than KimmllboxAcon. consensus sequence, boxAcon, by having a TTTC-3', rather than TTTA-3', we tested whether a nut^ region with boxAcon conferred any advantage for X growth in Antitermination measured by expression from a galK hosts with variant nus alleles known not to be effec tively suppressed by the boxAl mutation. fusion A comparison of the growth of X derivatives with wutR To directly assess the role of the boxA sequence in N- regions containing boxAcon, boxAl, or boxA"^ in E. coli mediated transcription antitermination, we employed a having mutant or variant nus genes revealed a hierarchy chromosomally located operon fusion that has the X pR of activity for boxA sequences. The data in Table 2 show promoter with its associated nut^ region and two down that in every E. coli tested, kboxAcon grows either as stream Rho-dependent terminators fused to the gal well as or better than XboxAl which, in turn, grows as operon (Fig. 3; Reyes et al. 1979; Dambly-Chaudiere et well as or better than XboxA*. The nus alleles sup al. 1983). The galK gene can thus be used as a reporter

Table 2. Efficiency of plating of X derivatives with different boxA sequences Bacteria and relevant alleles K4087 boxA K95 K450 K556 nusAst K4092 Phage sequence nusAl nusB5 nusEJl sneA16 nusAs.t. [Arc] (40°C) (42°C) (37°C) X CGCTCTTAC

2212 GENES & DEVELOPMENT Downloaded from genesdev.cshlp.org on September 30, 2021 - Published by Cold Spring Harbor Laboratory Press

BoxA transcription signal

10000

1000

Figure 4. Expression of galactokinase from pj^-gal fusions in a nusAl host. Following a shift from 32°C to 42°C, aliquots of logarithmically growing cells were removed at the indicated times and assayed for galactokinase. Galactokinase units are defined as nmoles of galactose-1-phosphate produced per min Time (min) per OD420iim of cells. The temperature shift inactivated the clts857 repressor, permitting transcription to initiate atp^ (for Figure 2. Effect of boxAj^ sequence on X growth in a nusBS reference, see Materials and methods). The boxA signals in the mutant host. Phage growth was followed through one round at X. nut^ regions were boxAcon (•), boxAl (•), and boxA+ (A). 37°C according to the method referenced in Materials and methods. (•] XboxA""-. (•) XboxAl; (•] \boxAcon.

Competition between boxA sequences with expression of its product^ galactokinase, repre senting a measure of transcription antitermination. Fu We compared the activities of boxA sequences by asking sions with three variations of boxAa—t>oxA+, boxAl, if any of the variations of boxA conferred an advantage and boxAcon—were tested. Expression of galK from the to a nut^ region in competition with another nutR region fusions was studied in the presence of the nusAl muta containing a different box A. The design of the experi tion at 42°C, a condition shown above to favor growth of ment was based on two components, Xr32 and plasmids phages with the boxAl or boxAcon variations in their with cloned nut^ regions. The strategy of using a nutR region. The kinetics of galK expression by the three plasmid to titrate functions required by k has been em fusions following the shift of log-phase cultures from ployed successfully in the study of N-mediated action 32°C to 42°C are shown in Figure 4. The temperature (Friedman and Yarmolinsky 1972; Lieb 1972) and in shift places the cultures under N-limiting conditions be demonstrating that an E. coli rrn operon, which has an cause of the nusAl mutation while simultaneously re antiterminator, competes with \ for a common factor moving repression from the X. PR promoter by heat-dena (Sharrock et al. 1985). turing the cI857 repressor of the prophage. Consistent The KI32 phages were used because the IS2, with its with the studies measuring phage growth^ we again find strong Rho-dependent terminator (De Crombrugghe et a hierarchical order of activity of boxA sequences: The al. 1973) located upstream of cll (Fig. 1; Brachet et al. fusion of the nut region with boxAcon expresses galK 1970), imposes additional termination in the p^ operon earlier and reaches a higher steady state than the fusion and, thus, increases the level of antitermination required with boxAl. The fusion with the wild-type boxA se for effective transcription of downstream genes, in quence, as expected, fails to express significant levels of cluding the Q late operon activator. The effect of this galactokinase under the conditions of these experi increased termination is easily observed using phage ments. growth in nus mutants (Tomich and Friedman 1977). Because Xr32 does not form plaques on lawns of nus mu tants at 32°C while X does, growth of AJ'32 derivatives boxA boxB provides a more sensitive assay of the effectiveness of gal 152 the assembly of the antitermination complex at nut^ pL cits pR cro nut tRl T than does growth of the wild-type phage.

ANTITERhlNATION The 22utR regions with the various box A sequences NO - were placed downstream of the p^ac promoter (de Boer et al. 1983) in pKK223-3 (de Boer 1984) in the same orienta YES tion relative to PR as they are found in the phage. Tran scription of the cloned nut^ regions can be controlled by Figure 3. Chromosomally located fusion of pJ^ to the gal operon. The map shows the essential parts of the truncated placing the plasmids in a host that overproduces lac re prophage with the r32-IS2 element and the fusion to the gal pressor {lacl'^] (Muller-Hill et al. 1968). Hence, high operon. The nutj^ region is expanded above the map, while the levels of transcription of the I2U£R insert can occur only fate of pR transcripts when N-mediated antitermination is inac when an inducer, such as IPTG, is added to the culture. tive and when it is active is shown below the map. The test for nut^^ site competition relies on the potential

GENES & DEVELOPMENT 2213 Downloaded from genesdev.cshlp.org on September 30, 2021 - Published by Cold Spring Harbor Laboratory Press

Friedman et al. for the plasmid to compete for a limiting component of or boxA5) (data not shown) has any effect on the time or the antitermination reaction during transcription. size of the burst. Failure of the induced plasmid pKBAl Figure 5 shows the results of these experiments. Phage (with boxAl) to compete demonstrates the specificity of production was measured by following one round of the competition by the boxAcon plasmid; the single- growth at 40°C (a single-step growth experiment). The base change differentiating boxAl and boxAcon is suffi effect of the plasmid-borne boxA sequences was assessed cient to interfere with the burst. Second, interference is by examining \r32 growth in the presence of the various only observed if the nut region is transcribed. In the ab plasmids either induced or not induced with IPTG. sence of IPTG, neither pKBAcon nor any of the other There are two essential findings. First, only plasmid plasmids interferes with \r32 growth. Moreover, a deriv pKBAcon (containing a nut^ region with the boxAcon ative of TpKBAcon deleted for Ptac (pKBAconAP) failed to sequence) has any effect on Xr32 growth (Fig. 5A). In the compete in the presence of IPTG (data not shown). presence of IPTG, pKBAcon delays the appearance of the To determine whether boxAcon in the context of a ki32 burst. In contrast, neither the IPTG-induced parent nonfunctioning nut region competes with XJ32, we used pKK223-3 plasmid nor derivatives with cloned frag derivatives of pKK223-3 with cloned fragments con ments containing other box A variations [boxA'*', boxAl, taining a 21 nut^ region with either the boxAcon

1000

E 100

•i^—r A u 4 .0001 50 100 150 50 100 Time (min) Time (min)

1000 lUUU

100

.01

.0011 ^ D

.0001 1 . 1 . 1 .1 20 40 60 Time (min) Figure 5. Phage production in competition experiments. Single rounds of growth were assayed as outlined in Materials and methods. The bacteria used in these experiments are derivatives of K37 that have the lacl''^ repressor mutation and carry either pKK223-3 or pKK223-3-derivatives with nutj^ regions and, when indicated, pGB2 or derivative plasmids. Table 4 lists the relevant details about the plasmids. Infected bacteria were divided into two aliquots, one in which Pt^c remained repressed and one in which it was induced by addition of IPTG. Infections were allowed to proceed at 40°C. Aliquots were removed at indicated times and assayed for phage production. {A) Burst of Xr32 in K3093 derivatives with either pKBAcon which has a nut region with boxAcon (D, uninduced; •, induced) or the control plasmid pKK223-3 |0, uninduced; •, induced). (B) Effect of gpN on plasmid competition with \r32. Bacteria (derivatives of K4461) were grown at 40°C prior to infection with Xr32 to induce the defective X prophage and gpiV production. Phage-infected bacteria, which were induced with IPTG, were allowed to burst as in A. The bacteria contained pKBAcon (•), pKBAl, which has a nut region with boxAl (•), pKBAL, which has a nut region with boxA'^ (A), or pKK223-3 (•). (C) Role of NusB in competition. The phage and the conditions weie the same as in A, except the bacteria carried either pGB2r2usB with pKBAcon (ffi, uninduced; -I-, induced) or pGB2 with pKBAcon (D, uninduced; •, induced). (D) Role of NusA in competition. The phage and the conditions were the same as in A, except the bacteria carried either pGB2 with pKBAcon (D, uninduced; •, induced) or pGB2r!usA with pKBAcou (B, iminduced; x, induced).

2214 GENES & DEVELOPMENT Downloaded from genesdev.cshlp.org on September 30, 2021 - Published by Cold Spring Harbor Laboratory Press

BoxA transcription signal

(pKBA21con) or boxA wild-type (pKBA21) sequence in tion that overexpression of nusB results in a general en competition experiments in the nus'*' host. Since the hancement of X growth. only N product available in the infection is gpN^, the 21 We then determined whether interference with com nut region cannot function in antitermination, even petition was specific to the nusB gene product or could though it is transcribed (Friedman et al. 1973a; Lazinski merely be due to the excess expression of any nus gene. et al. 1989). When induced, neither pKBA21 nor To test for possible involvement of NusA, we used an pKBA21 con causes significant interference with the other pGB2 derivative, pGB222usA, which contains a growth of \i32 (data not shown). cloned 5-kb fragment that includes the nusA gene. As shown in Figure 5D, pGBlnusA does not interfere signif icantly with pKBAcon competition. Role of gpN The role of gpN in this plasmid-mediated competi Competition assayed by expression from a galK fusion tion was assessed using a defective prophage The chromosomally located p^-gal fusion was em (X.cI857AB(amAHI) to provide additional gpN. This pro ployed to directly test whether N-mediated antitermina phage produces high levels of gpN at temperatures tion directed through the nut^ signal was the basis for >40°C (Gottesman et al. 1980). If gpAT is a limiting the plasmid competition. K5319, the J2us+ bacteria used factor, an increased amount of it would be expected to in these studies, also carries the lacl'^ mutation. relieve the effect of plasmid competition on phage In the first set of experiments, competition was as growth. sessed using a fusion with a nut^ region containing box Competition between \r32 and the i2ut-containing Acon (Fig. 6A). Three plasmids were employed, pKK223- plasmids was re-examined in the presence of prophage- 3, the boxA^ derivative (pKBAL), and the boxAcon de supplied gpN. Consistent with the findings reported rivative (pKBAcoiz). When transcription of the nut region above, significant competition was observed only with of pKBAcon is induced, competition is observed by a sig the plasmid containing the nutR boxAcon, pKBAcon, in nificant reduction in galK expression from the fusion. the presence of IPTG (Fig. 5B). However, instead of re The plasmid-based nutR with wild-type boxA fails to lieving competition, the additional gpN, if anything, compete even when induced. supported slightly more effective competition (Fig. In the second set of experiments, the effect of nusB 5A,B). With or without extra gpN, there is a delay in the expression on competition was assessed using the same start of the Xr32 burst. Without prophage-supplied gpN, fusion. Three plasmids were employed, pKBAcon, pGB2, in the presence of IPTG-induced pKBAcon, the burst ap and pGB2n usB. As shown in Figure 6B, the induced proaches the size seen in the absence of the plasmid (Fig. pKBAco22 plasmid reduces galK expression from the fu 5A). With gpN expressed from the prophage, however, sion. In the presence of the nusB plasmid, the level of pKBAcon inhibits the size of the burst for the entire galK expression goes up. Note that pGB2nusB has no ef time course of the experiment (Fig. 5B). fect on the expression from the fusion if pKBAcon is not induced, offering further evidence that overexpression of nusB does not cause a nonspecific increase in transcrip Role of Nus factors tion from PR. To determine whether the limiting factor(s) in the com petition assay is an E. coli Nus protein, we employed Plasmid copy number plasmids with cloned nus genes as the source of addi tional nus protein production. These plasmids were de We determined whether plasmid copy number was an rivatives of pGB2 (Churchward et al. 1984), a low-copy- explanation for the observed difference in growth of \r32 number plasmid (~10 per cell) compatible with in the plasmid competition experiments by measuring pKK223-3. These pGB2i2us derivatives express their re plasmid DNA. Plasmid DNA isolated from bacteria with spective nus genes as shown by complementation with the pKK223-3 set of boxA plasmids grown under the appropriate nus mutations (data not shown). conditions used in the competition experiments was A pGB2 derivative with a cloned 2.5-kb fragment con compared with controls for DNA loss (see Materials and taining the nusB gene [pGBlnusB] significantly reduces methods). No significant differences in plasmid copy competition by the induced pKBAcon plasmid (Fig. 5C). number were found under any of the conditions of our The reduction in competition is due to the insert con experiments (data not shown). taining the nusB gene, since a control pGB2 plasmid has no effect on the competition (Fig. 5C). The role of the nusB gene was directly tested using a derivative of Discussion pGBlnusB containing a deletion and substitution within The experiments reported here address the question of the nusB gene, pGBlAnusB. Unlike the parent plasmid, the importance of the boxA sequence in \ N-mediated pGB2AnusB has no effect on the competition by antitermination by demonstrating that (1) the full extent pKBAco22 (data not shown). There was no effect of of the consensus boxA sequence defined previously by pGB2nusB on the growth of Xr32 in the presence of a sequence comparison is functionally important, (2) tran control such as pKK223-3, ruling out the trivial explana scription is required for the boxA signal to be active, (3)

GENES & DEVELOPMENT 2215 Downloaded from genesdev.cshlp.org on September 30, 2021 - Published by Cold Spring Harbor Laboratory Press

Friedman et al.

Defining the optimal boxA sequence P22 served as the prototype for designing alterations in boxAftj because its boxA^ corresponds to the consensus, boxAcon (Table 1), and its growth is relatively unaf >^_ L O /^^"^ O fected by the presence of mutations in nus genes (Hil- liker and Botstein 1976; Friedman and Olson 1983; Z^^.-'- ° D Schauer et al. 1987). This suggested that one or more elements of the termination-antitermination system of P22 (i.e., gpiV, terminators, or nut] might either enhance antitermination or reduce termination. //// • • Fortuitously, natural variants of the boxA sequence - //// J^ • provided the means to test whether the consensus boxA contributes to this enhanced action of P22 gpN. i^ A Changing both the 21 and X sequences to match box Acon resulted in an increased ability to grow in nus

20 40 60 100 variants under conditions where parental phages with Time (min) nonconsensus (wild-type or altered) boxA sequences failed to grow. In a previous study, we demonstrated that a change at position 8 of \boxAJ^ to a T, resulting in TC 3' (boxAl), enhances the ability of \ to grow in nus mutants 6 (Friedman and Olson 1983), demonstrating that the X^ / ° wild-type sequence, as tested by this assay, does not //^ • _^—- have optimal activity. We now show that a second change at nucleotide 9 (resulting in boxAcon] extends // /'''' * the range of X growth to include additional 22us mutants. • / / /^ Thus, as boxAp^ is changed to conform more closely to • 7/ / • // / " the consensus sequence, the range of nus mutants in which the resulting X grows is increased, giving the fol z/ / • lowing hierarchical order of functional activity, box Acon > boxAl > boxA'^. A p^—gal fusion was used to directly and quantita tively assess gpiV—nutR action in promoting antitermin // B ation. Results obtained with the fusions correspond

0 20 40 60 80 exactly to the those obtained with the phages. This sup ports our contention that the growth patterns we ob Time (min) serve with the phages are direct consequences of the Figure 6. Competition assayed from a p^-gal fusion con strength of the antitermination reaction at nut^ and also taining a nutjt^ with boxAcon (for details, see Fig. 4). The host confirms that the plasmid competition assay assesses ef strain, K5319, is nus-^ and lacl'i. Bacteria grown at 32°C in LB fects on antitermination. (containing appropriate antibiotics to maintain plasmids) were diluted into LB with the appropriate antibiotics and, where in dicated, IPTG. The diluted cells were shifted to 40°C, and aliquots were removed at the indicated times and assayed for ga- Involvement of boxA in antitermination lactokinase activity. [A] Role of boxA in competition. The bac teria contained one of the following plasmids: pKBAcon, which The results of the plasmid competition studies argue has a nutR region with boxAcon (•, uninduced; •, induced); that boxA, or a part of that sequence, recognizes an anti- pKBAL, which has a nutR region with boxA wild type (A, unin termination factor. Using the pKK223-3 plasmids with duced; A, induced); or pKK232-3 (O, uninduced; •, induced). nut^ regions under p^ac control, we find that transcrip (B) Role of NusB in competition. The bacteria contained tion of a nutR-containing boxAcon reduces growth of in pGB2nusB with pKBAcon (O, uninduced; •, induced) or pGB2 fecting Xr32 phages and expression from the p-R-gal fu with pKBAcon (D, uninduced; •, induced). sion. Plasmids with boxA'^ or boxAl do not compete. Because boxAcon enhances nut-directed antitermina tion in cis but reduces it in trans, we conclude that box Acon competes more avidly than other boxA sequences the nature of the boxA signal influences the affinity of for a factor required in N-mediated antitermination. the antitermination complex for the host-encoded nusB This conclusion may seem surprising in light of the sug gene product, (4) this competition is only observed when gestion that boxA is a component of the rutA site re the boxA signal is part of a nut region activated by gpiV, quired for action of Rho at tRi (Chen et al. 1986; Chen and (5) the optimally active boxA sequence may not and Richardson 1987). Perhaps boxAcon only appears to always have been selected. increase antitermination, but actually reduces the af-

2216 GENES & DEVELOPMENT Downloaded from genesdev.cshlp.org on September 30, 2021 - Published by Cold Spring Harbor Laboratory Press

BoxA transcription signal finity in the nut region for a protermination factor such able sequestration of NusB. However, it is possible that as Rho. According to this scenario, the reduced level of the sequestration of NusB is a secondary effect resulting termination factor loading at boxAcon would diminish from increased activity of another component of the downstream termination which, in turn, would reduce N-Nut interaction. According to this model, boxAcon the requirement for antitermination at the downstream would increase the action of another Nus factor at nut, terminators. Thus, it would appear that antitermination enhancing N-Nus complex formation there. The avail had been improved. able supply of NusB would then be depleted by its in Although the studies showing good growth of \box- creased utilization in the formation of those complexes. Acon in nus mutants are consistent with this idea, the Whether or not boxA is a primary site for NusB, our re results of the plasmid competition studies argue against sults demonstrate that NusB is a limiting factor in the it. Even if the transcribed plasmid failed to sequester any N-mediated antitermination reaction. of the hypothetical termination factor, the concentra tion of that factor in those bacteria should be no dif ferent than in the control bacteria without the plasmid. The N-Nus complex Therefore, there should be no difference in termination The demonstration that transcription of the nut region factor available to act on the infecting phage and the is required for boxA competition suggests that the boxA burst of phage should be the same in the two classes of signal is read from the RNA and not the DNA. Although bacteria, that is, boxAcon should only influence antiter previous studies have suggested that transcription of the mination in CIS. However, the transcribed plasmid-based nut region is required for N-mediated antitermination boxAcon signal acts in tians to reduce Xr32 growth, sug (Olson et al. 1984; Warren and Das 1984; Zuber et al. gesting that the sequestered factor is not required for ter 1987), the work reported here offers the first conclusive mination, but is for antitermination. evidence that activation of the boxA signal, per se, re quires transcription of the nut region. Thus far, the only antitermination factor shown to bind to RNA is NusA boxA—A signal for NusB^ (Tsugawa et al. 1985). In those studies NusA binding ap Our observation that the competition by the plasmid- pears to be specific for RNA containing a nut region, but borne boxAcon is reversed in the presence of NusB ex the binding occurred upstream of nut^ in the cio gene pressed from a compatible plasmid demonstrates a sequences that obviously are not conserved in nut sites. linkage between boxA and NusB action. That this re Our studies, employing the nut region of 21, showed versal is specifically due to NusB action is proved by that merely transcribing a. nut region with boxAcon is the failure of a derivative plasmid with a deletion and not sufficient to activate that boxA signal for competi substitution in the nusB gene to alleviate competition. tion in the absence of the proper gpN. This suggests that Since competition is still observed in the presence of a competition results from the formation of a larger com similar construct expressing nus A, the reversal of com plex activated by the full nut signal and containing, at a petition must not result merely from increased expres minimum, gpN. sion of Nus products in general. What is particularly striking is the role of a single nucleotide change at the 3' end of boxA in effecting this competition for NusB. Why not the bestl These results are consistent with studies showing that If boxAcon is the optimal signal, why do X and 21, un antitermination directed by the leader region of nn like their P22 cousin, have a less-than-optimal signal? operons that is dependent on an intact boxA sequence To begin, we note that a number of X-encoded factors (Berg et al. 1989) is reduced in the presence of the nusB5 apparently have less than optimal activity. As in the mutation (Sharrock et al. 1985). It has been estimated case of boxA, the lowered activities of the naturally oc that there are 6000-7000 molecules of NusB per cell in curring factors were demonstrated by isolating mutants fast growing E. coli (Swindle et al. 1988), corresponding that appeared to function more effectively. Interestingly, to —50-80% of the number of RNA polymerase mole all of these phage-encoded proteins or signals interact cules. As pointed out by Swindle et al. (1988), much of with host-encoded proteins. Examples are gpNul (acts in the NusB could be tied up in transcription of nn formation of mature phage genome and subsequent operons, which account for 27-67% of transcription in packaging; Feiss et al. 1988; Cranston et al. 1988), gpint E. coli. Our finding of competition for NusB is consis (required for integrative recombination; Miller et al. tent with this idea and suggests that either most of the 1980), and gpiV (Friedman and Ponce-Campos 1975; NusB is sequestered or the N complex has a low affinity Franklin 1985; Schauer et al. 1987). Action of gpNul is for NusB. enhanced by and gpint requires the histone-like protein It is attractive to conclude that our results present in IHF (for review, see Friedman 1988b). Action of gpN re vivo evidence that boxA, as originally suggested by D. quires the Nus proteins. Thus, we suggest that evolu Court (pers. comm.), is a signal for NusB. Indeed, Hor- tionary pressure selects for the less-than-optimal se witz et al. (1987) have suggested that NusB is necessary quence to maintain regulatory pathways. In the case of specifically for the functioning of box A. Certainly, such the N-antitermination system, the time during X devel a conclusion might seem warranted from our observa opment when terminators are transcended influences tion that changing one base in boxA causes an observ the coordination of functions involved in the lysis-ly-

GENES & DEVELOPMENT 2217 Downloaded from genesdev.cshlp.org on September 30, 2021 - Published by Cold Spring Harbor Laboratory Press

Friedman et al. sogeny decision (Wulff and Rosenberg 1983). Indeed, a \ buffer (50 mM Na2HP04, 20 mM KH2PO4, and 70 mM NaCl) and mutant with a more active N-mediated antitermination then used to infect cultures of K3929 at a multiplicity of five system at nut^ is channeled toward lysogeny (D. phage per bacterium. After adsorption, infected bacteria were Friedman, unpubl.). Thus, what appears to be a less ac allowed to grow in LB broth. Based on the assumption that the tive sequence may actually be optimal for its biological mutant boxA sequences should suppress the effect of the nusAl mutation, candidate recombinants were identified as role. We cannot explain why P22 fimctions with the op Gal+ colonies at 42°C. Backcrosses to X.r32 proved that these timal boxA sequence, except to note that the mecha Gal+ fusions did have the mutant boxA sequences. nism for deciding between lysis and lysogeny differs sig nificantly in X and P22 (Susskind and Youderian 1983). Plasmids Materials and methods Constructs were made by employing standard cloning tech niques (Maniatis et al. 1982). Relevant plasmid characteristics Media are listed in Table 4. A promotorless derivative of pKBAcon, The media used in these studies have been described (Miller pKBAconAP, was constructed by deleting a BamHl-EcoRl frag and Friedman 1980). LBM is LB broth containing 0.2% maltose. ment containing Ptac- pGB2AnusB was constructed by removing IPTG was used at a final concentration of 1 niM. a 240-bp internal fragment from the nusB gene in pGRlnusB and inserting a 1.35-kb Seal fragment from pBR325 that con tains the cam^ gene (Bolivar 1978). Bacteria and phage Plasmid DNA preparation and transformation procedures A list of bacteria and their relevant characteristics are given in were performed according to Maniatis et al. (1982). pGB2 deriv Table 3. Phages and sources are as follows: XcI857 (NIH collec atives were prepared essentially according to the published pro tion); \i31, which carries the r32 insertion; an IS2 inserted be cedure (Churchward et al. 1984). tween the nut region and the cll gene (Brachet et al. 1970) (NIH collection); kboxAl (Friedman and Olson 1983); Ximmll (hy5) Single-step growth experiments and site-directed mutagenesis (Liedke-Kulke and Kaiser 1967); and M13mp8 and M13mp9 (Yanisch-Perron et al. 1985). XboxAcon, krSlboxAcon, and Single-step growth experiments were performed as outlined in )dmm21boxAcon were constructed during this work. Friedman etab (1973b). The method employed for site-directed mutagenesis was that used in Olson et al. (1984). The nutR region with the boxAl Construction of chromosomal p^-gal fusions variation (Friedman and Olson 1983) was cloned in M13mp9, The altered kboxA sequences were crossed from derivatives of and the sequence of the mutagenic oligonucleotide primer was ki3>l by homologous recombination into a Pj^-lS2-gal fusion 5'-CCCCGCTCTTTAACATTCC yielding nutRX.boxAcon. The (see Fig. 3; Reyes et al. 1979) in strain K3929, which carries the nutf^2l region on an £coRI-HindIII fragment from pRZrjut-21 nusAl allele and has the cro62 mutarion (Olson et al. 1982). was cloned in M13mp8, and the sequence of the mutagenic oli K3929 is phenotypically Gal" at 42°C even though repression is gonucleotide primer was 5'-GGCCAGAACTGTTAAAGAGC- off, because the N-antitermination reaction is not active. This GATTTGC, yielding nutJ^2lboxAcon. recombination was possible because Xr32 has the same genetic arrangement as the Pj^-gal fusion from the left of immunity through the 5' two-thirds of the IS2. Lysates of \r32 derivatives Construction of phage with mutant boxA sequences with either boxAl or boxAcon were UV-irradiated (General Method An ovemight culture (0.1 ml) of an E. coli carrying a Electric lamp no. G8T5) at a dose of 630 ergs/mm^ in phosphate plasmid with the desired boxA sequence was inoculated into 10

Table 3. Bacteria Strain Relevant genotype Parent* Source K37 nus'^ NA National Institutes of Health K95 nusAl K37 University of Michigan K450 nusB5 K37 University of Michigan K556 nusEJl K37 University of Michigan K1102 nusAs.t. NA L.S. Baron K1227 Pf^lcro62/gal fusion OR1150 University of Michigan K3093 lacl'i K37 University of Michigan K3929 p^/cro62/gal/nusAl fusion K1227 University of Michigan K3930 Ps^/boxAl/gal/nusAl fusion K3929 University of Michigan K4087 nusAs,t/sneA16 K37 University of Michigan K4092 nusAs_t, K37 University of Michigan K4461 lacl'i \cI857ABamAHI K3093 University of Michigan K5310 Pj(/boxAcon/gal/nusAl fusion K3929 University of Michigan K5319 p-n/boxAcon/gal/lacI'i nus'^ fusion K5310 University of Michigan K5693 p^lgal/nusAl fusion OR1150 University of Michigan JMlOl lacl'i, lacZAmlS NA J. Messing DH5a lacZHimlS, recAl NA Bethesda Research Labs "(NA) Not applicable.

2218 GENES & DEVELOPMENT Downloaded from genesdev.cshlp.org on September 30, 2021 - Published by Cold Spring Harbor Laboratory Press

BoxA transcription signal

Table 4. Plasmids and M13 derivatives Parent Relevant fragment Plasmid vector or characteristic* Donor Source pCG69 pCRl EcoRI with immll (from Ximmll C. Georgopoulos red to O) pMClOO pBR322 EcoRI with immll pCG69 this work pRZ-nutji pRZ152 Cifll-EcoRI with flutR2i pMClOO this work pBA7-6 pBR322 EcoKL-Hindlll nutji^)^ with MliboxAl this work boxAl pBAcoiJ pBR322 EcoRI-Hindlll mztju with M13 boxAcon this work boxAcon pT21 pUC18 EcoKL-HindUl nutj^i M13nutR2i this work wild-type pC21 pUC18 £coRI-HindIII nutRji v^^ith M13nutR2i this work boxAcon boxAcon pKK223-3 pBR322 cloning site downstream de Boer (1983) of Ptac pKBAL PKK223-3 \boxAJ^^^ M13-boxA + this work pKBAl PKK223-3 kboxAl pBA7-6 this work pKBA5 pKK223-3 \boxA5 Ml3-boxA5 this work pKBAcon pKK223-3 \boxAcon pBAcon this work pKBA21 pKK223-3 21 boxAJ^+ Clal-HincU pT21 this work pKBA21 con pKK223-3 21 boxAcon Clal-Hincll pC21 this work pKBAconAP pKBAcon kboxAcon Aptac this work pGB2 pSClOl Churchward etal. (1984) pQBlnusB pGB2 nusB F. Kappel (via C. Georgopolous) pGB2AnusB PGB2HUS5 nus^.B this work ipGBlnusA pGB2 nusA pNAG2010 this work M13nutR2i M13mpl8 £coRI-HindIII nutjui pRZ-nutR2i this work M.l3nutJ^2lboxAcon M13mpl8 JScoRI-Hindlll nu£R2i M13mp8nutR2i this work nutjai mutagenesis MUboxAl M13mp9 nutJ^-boxAl M13mp9HutR Friedman and mutagenesis Olson (1983) *The \ inserts contain DNA from coordinates 38214 to 38350 (Daniels et al. 1983). The 21 CM and Hindi sites are in the cro and cll genes, respectively (Schwarz 1980). Other sites are in X or the donor vectors.

ml of LB broth made 0.01 M in CaClj and containing 30 |xg/ml 21 at 42°C. Recombinant P21 derivatives that grew under these ampicillin. The bacteria were infected with phage at a multi conditions were found at a frequency of 10~^. Control crosses plicity of 5 and incubated at 37°C for 2 hr. The resulting lysate with plasmids containing wild-type nut regions yielded phages was treated with chloroform, and recombinants were selected that grew in the selecting bacteria at much lower frequencies using appropriate bacterial lawns (see text). (<10"^). DNA sequencing confirmed that the indicated plasmids and phages contained the boxAcon sequence. Strategy We have shown previously that X and \r32 deriva tives with the boxAl mutation in nutj^ grow in nusAl mutants Competition infections under conditions where the nutj^'*' derivatives fail to grow (Friedman and Olson 1983; Schauer et al. 1987). We assumed Bacteria were diluted from an overnight culture into LBM con that X. and 21 derivatives with boxAcon in their nut^^ regions taining ampicillin at 30 |xg/ml (spectinomycin at 50 jig/ml was might also grow in hosts with mutant or variant nusA genes included if pGB2 derivatives were also used) and grown to a under conditions where the parent phages with wild-type nut^ concentration of ~10^/ml. The bacteria were sedimented by regions fail to grow, because of a failure in N-mediated antiter- centrifugation, resuspended in 0.1 volume of 0.01 M MgS04 mination. Recombinants derived from A, and Xr32 with box and, where indicated, 1 mM IPTG, mixed with the phage at a Acon were isolated from a cross with pBAcozi using K95 multiplicity of 0.1, and incubated at room temperature for 20 [nusAl] as the selective host at 42°C. Putative derivatives of \ min to allow phage to adsorb. Infected bacteria were diluted and krSl with boxAcon were found, respectively, at a frequency into LB made 0.01 M in MgS04 and 30 |xg/ml in ampicillin and, of 10~* and 10"^. Recombinants derived from 21 with boxAcon where indicated, 1 mM in IPTG and/or spectinomycin at 50 were isolated from a cross with pC21. The selecting bacterium, |jLg/ml, and incubated at the appropriate temperature. Aliquots K1102, contains a hybrid nusA gene with the 5' 85% from S. were removed at the indicated times, treated with chloroform typhimurium and the 3' 15% from E. coli (Friedman and Olson to lyse bacteria, and titered for viable phage. Burst represents 1983; Schauer et al. 1987) and does not support growth of phage phage output/phage input.

GENES & DEVELOPMENT 2219 Downloaded from genesdev.cshlp.org on September 30, 2021 - Published by Cold Spring Harbor Laboratory Press

Friedman et al.

Measurement of plasmid copy numbei Botstein, D. and I. Herskowitz. 1974. Properties of hybrids be tween Salmonella phage P22 and coliphage lambda. Nature Ten milliliters of culture from burst experiments was com 251: 584-589. bined with 1 ml of overnight culture of DH5a with pUC8 and Brachet, P., H. Eisen, and A. Rambach. 1970. Mutations of coli sedimented. Plasmid DNAs were isolated and treated with a phage lambda affecting the expression of replicative func restriction enzyme that cleaved both plasmids once. Aliquots tions O and P. Mol. Gen. Genet. 108: 266-276. of plasmid preparations were examined on an agarose gel using Chen, C.Y. and J.P. Richardson. 1987. Sequence elements es ethidium bromide staining to visualize bands. A comparison sential for rho-dependent transcription termination at with the pUC8 standard provided a measure of the DNA con lambda tRl. /. Biol. Chem. 162: 11292-11299. centration of the experimental plasmid. Chen, C.Y., G.R. Galluppi, and J.P. Richardson. 1986. Tran scription termination at lambda tRl is mediated by interac DNA sequencing, galactokinase assay, and enzymes tion of rho with specific single-stranded domains near the 3' end of cro mRNA. Cell 46: 1023-1028. DNA sequencing was performed using the methods in Biggin et Churchward, G., D. Belin, and Y. Nagamine. 1984. A pSClOl- al. (1983), with slight variations. Galactokinase assays were derived plasmid which shows no sequence homology to performed according to the published method (Adhya and other commonly used cloning vectors. Gene 31: 165-171. Miller 1979). All enzymes were purchased from standard com Dambly-Chaudiere, C, M. Gottesman, C. Debouck, and S. mercial sources. Adhya. 1983. Regulation of the pR operon of bacteriophage lambda. /. Mol. Appl. Genet. 2: 45-56. Acknowledgments Daniels, D.L., J.L. Schroeder, W. Szybalski, F. Sanger, and F.R. Blattner. 1983. A molecular map of coliphage lambda. In We thank Jack Weber and Susan Ellis for help in constructing Lambda II (ed. R.W. Hendrix, J.W. Roberts, F.W. Stahl, and the boxAl- and hoxAcon-p^^-gal fusions. Jay Kilpatrick, Wes R.A. Weisberg), pp. 469-676, Cold Spring Harbor Labora Dunnick, and Max Gottesman are gratefully acknowledged for tory, Cold Spring Harbor, New York. editorial assistance. We also thank Mike Imperiale, Don Court, Das, A. and K. Wolska. 1984. Transcription antitermination in Robert Weisberg, Sankar Adhya, Andrew Granston, and Asis vitro by lambda N gene product: Requirement for a phage Das for helpful discussion. Don Court, WilHam Whalen, David nut site and the products of host nusA, nusB, and nusE Lazinski, and Asis Das are thanked for providing unpublished genes. Ce77 38: 165-173. information. We are grateful to Costa Georgopoulos for pro Das, A., B. Ghosh, S. Barik, and K. Wolska. 1985. Evidence that viding pGB2flusB. We thank Angle Gross and Jana Gilbert for ribosomal protein SIO itself is a cellular component neces help in preparing the manuscript. M.G.C. was supported by sary for transcription antitermination by phage lambda N U.S. PubUc Health Service training grant 5-T32-GM07315-13. protein. Proc. Natl. Acad. Sci. 82: 4070-4074. Work at the University of Michigan was supported by U.S. de Boer, H.A. 1984. A versatile plasmid system for the study of Pubhc Health Research grant 5 ROl All459-10 from the Na prokaryotic transcription signals in Escherichia coli. Gene tional Institutes of Health. 30:251-255. The publication costs of this article were defrayed in part by de Boer, H.A., L.J. Comstock, and M. Vasser. 1983. The tac pro payment of page charges. This article must therefore be hereby moter: A functional hybrid derived from the trp and lac pro marked "advertisement" in accordance with 18 USC section moters. Proc. Natl. Acad. Sci. 80: 21-25. 1734 solely to indicate this fact. De Crombrugghe, B., S. Adhya, M. Gottesman, and I. Pastan. 1973. Effect of Rho on transcription of bacterial operons. References Nat. New. Biol. 241: 260-264. Doelling, J.H. and N.C. Franklin. 1989. Effects of all single base Adhya, S. and W. Miller. 1979. Modulation of the two pro substitutions in the loop of boxB on antitermination of moters of the galactose operon of Escherichia coh. Nature transcription by bacteriophage lambda's N protein. Nucleic 279: 492-494. Acids Res. 17: 5565-5577. Aksoy, S., C.L. Squires, and C. Squires. 1984. Evidence for anti- Feiss, M., S. Fogarty, and S. Christiansen. 1988. Bacteriophage termination in Escherichia coh irnA transcription. /. Bac- lambda DNA packaging: A mutant terminase that is inde teiiol. 159: 260-264. pendent of integration host factor. Mol. Gen. Genet. Barik, S., B. Ghosh, W. Whalen, D. Lazinski, and A. Das. 1987. 212: 142-148. An antitermination protein engages the elongating tran Franklin, N.C. 1985. Conservation of genome form but not se scription apparatus at a promoter-proximal recognition site. quence in the transcription antitermination determinants of Cell 50: 885-899. bacteriophages lambda, phi 21, and P22. /. Mol. Biol. Baron, L.S., E. Penido, I.R. Ryman, and S. Falkow. 1970. Be 181: 75-84. havior of coliphage lambda in hybrids between Escherichia Friedman, D.I. 1988a. Regulation of phage gene expression by coh and Salmonella, f. Bacteriol. 102: 221-233. termination and antitermination of transcription. In The Berg, K.L., C. Squires, and C.L. Squires. 1989. Ribosomal RNA bacteriophages 2 (ed. R. Calendar), pp. 263-319, Plenum operon anti-termination. Function of leader and spacer re Press, New York. gion box B-box A sequences and their conservation in di . 1988b. Integration host factor: A protein for all reasons. verse micro-organisms. /. Mol. Biol. 209: 345-358. Cell 55: 545-554. Biggin, M.D., T.J. Gibson, and G.F. Hong. 1983. Buffer gradient Friedman, D.I. and M.B. Yarmolinsky. 1972. Prevention of the gels and 35S label as an aid to rapid DNA sequence determi lethality of induced lambda prophage by an isogenic lambda nation. Proc. Natl. Acad. Sci. 80: 3963-3965. plasmid. Virology 50: 472-481. Bolivar, F. 1978. Construction and characterization of new Friedman, D.I. and L.S. Baron. 1974. Genetic characterization of cloning vehicles. III. Derivatives of plasmid pBR322 carrying a bacterial locus involved in the activity of the N function of unique £coRI sites for selection of £coRI generated recombi phage lambda. Virology 58: 141-148. nant DNA molecules. Gene 4: 121-136. Friedman, D.I. and R. Ponce-Campos. 1975. Differential effect

2220 GENES & DEVELOPMENT Downloaded from genesdev.cshlp.org on September 30, 2021 - Published by Cold Spring Harbor Laboratory Press

BoxA transcription signal

of phage regulator functions on transcription from various Lieb, M. 1972. Properties of polylysogens containing dere- promoters: Evidence that the P22 gene 24 and the lambda N pressed lambda N-prophages II. The number of prophages af products distinguish three classes of promoters. /. Mol. Biol. fects bacterial viability. Virology 49: 818-820. 98: 537-549. Liedke-Kulke, M. and A.D. Kaiser. 1967. The c-region of coH- Friedman, D.I. and M. Gottesman. 1983. Lytic mode of lambda phage 21. Virology 32: 475-481. development. In Lambda II (ed. R.W. Hendrix, J.W. Roberts, Maniatis, T., E.F. Fritsch, and J. Sambrook. 1982. Molecular F.W. Stahl, and R.A. Weisberg), pp. 21-51, Cold Spring cloning, A laboratory manual. Cold Spring Harbor Labora Harbor Laboratory, Cold Spring Harbor, New York. tory, Cold Spring Harbor, New York. Friedman, D.I. and E.R. Olson. 1983. Evidence that a nucleotide sequence, "boxA," is involved in the action of the NusA Miller, H.I. and D.I. Friedman. 1980. An E. coli gene product protein. Cell 34: 143-149. required for lambda site-specific recombination. Cell Friedman, D.I., G.S. Wilgus, and R.J. Mural. 1973a. Gene N reg 20:711-719. ulator function of phage A. immll: Evidence that a site of N Miller, H.I., M.A. Mozola, and D.I. Friedman. 1980. int-h: An action differs from a site of N recognition. /. Mol. Biol. int mutation of phage lambda that enhances site-specific re 81: 505-516. combination. Cell 20: 721-729. Friedman, D.I., C.T. Jolly, and R.J. Mural. 1973b. Interference Morgan, E.A. 1986. Antitermination mechanisms in rRNA with the expression of the N gene function of phage X in a operons of Escherichia coli. f. Bacterial. 168: 1-5. mutant of Escheiichia coli. Virology 51: 216-226. MuUer-Hill, B., L. Crapo, and W. Gilbert. 1968. Mutants that Friedman, D.I., E.R. Olson, C. Georgopoulos, K. Tilly, I. Hers- make more lac repressor. Proc. Natl. Acad. Sci. 59: 1259- kowitz, and F. Banuett. 1984. Interactions of bacteriophage 1264. and host macromolecules in the growth of bacteriophage Olson, E.R., E.L. Flamm, and D.I. Friedman. 1982. Analysis of lambda. Microbiol. Rev. 48: 299-325. nut^: A region of phage lambda required for antitermination Friedman, D.I., M.J. Imperiale, and S.L. Adhya. 1987. RNA 3' of transcription. Cell 31: 61-70. end formation in the control of gene expression. Annu. Rev. Olson, E.R., C.S. Tomich, and D.I. Friedman. 1984. The nusA Genet. 21: 453-488. recognition site. Alteration in its sequence or position rela Gemski, P., Jr., L.S. Baron, and N. Yamamoto. 1972. Formation tive to upstream translation interferes with the action of the of hybrids between coliphage lambda and Salmonella phage N antitermination function of phage lambda. /. Mol. Biol. P22 with a Salmonella typhimuiium hybrid sensitive to 180: 1053-1063. these phages. Proc. Natl. Acad. Sci. 69: 3110-3114. Peltz, S.W., A.L. Brown, N. Hasan, A.J. Podhajska, and W. Szy- Ghosh, B. and A. Das. 1984. nusB: A protein factor necessary balski. 1985. Thermosensitivity of a DNA recognition site: for transcription antitermination in vitro by phage lambda Activity of a truncated nutL antiterminator of coliphage N gene product. Proc. Natl. Acad. Sci. 81: 6305-6309. lambda. Science 228: 91-93. Piatt, T. 1986. Transcription termination and the regulation of Gottesman, M.E., S. Adhya, and A. Das. 1980. Transcription gene expression. Annu. Rev. Biochem. 55: 339-372. antitermination by bacteriophage lambda N gene product. /. Mol Biol. 140: 57-75. Proudfoot, N.J. 1989. How RNA polymerase II terminates tran scription in higher eukaryotes. Trends. Biochem. Sci. Granston, A.E., D.M. Alessi, L.J. Eades, and D.I. Friedman. 14: 105-110. 1988. A point mutation in the Nul gene of bacteriophage Reyes, O., M. Gottesman, and S. Adhya. 1979. Formation of lambda facilitates phage growth in Escherichia coli with lambda lysogens by IS2 recombination: gal operon-lambda himA SLndgyrB mutations. Mol. Gen. Genet. 212: 149-156. pR promoter fusions. Virology 94: 400-408. Greenblatt, J. and J. Li. 1981. The nusA gene protein of Esche Roberts, J.W. 1988. Phage lambda and the regulation of tran richia coli. Its identification and a demonstration that it in scription termination. Cell 52: 5-6. teracts with the gene N transcription anti-termination pro Robledo, R., M.E. Gottesman, and R.A. Weisberg. 1990. \ nut-g_ tein of bacteriophage lambda. /. Mol. Biol. 147: 11-23. mutations convert HK022 Nun protein from a transcription Hilliker, S. and D. Botstein. 1976. Specificity of genetic ele termination to a supressor of termination. /. Mol. Biol. ments controlling regulation of early functions in temperate 212: 635-643. bacteriophages. /. Mol. Biol. 106: 537-566. Rosenberg, M., D. Court, H. Shimatake, C. Brady, and D.L. Hilliker, S., M. Gottesman, and S. Adhya. 1978. The activity of Wulff. 1978. The relationship between function and DNA Salmonella phage P22 gene 24 product in Escherichia coli. sequence in an intercistronic regulatory region in phage Virology 86: 37-47. lambda. Nature 272: 414-423. Rosenthal, E.R. and J.M. Calvo. 1987. Transcription termina Holben, W.E. and E.A. Morgan. 1984. Antitermination of tran tion sites at the distal end of the leu operon of Salmonella scription from an Escheiichia coli ribosomal RNA pro typhimurium. f. Mol. Biol. 194: 443-452. moter. Proc. Natl. Acad. Sci. 81: 6789-6793. Salstrom, J.S. and W. Szybalski. 1978. Coliphage lambda nutL": Horwitz, R.J., J. Li, and J. Greenblatt. 1987. An elongation con A unique class of mutants defective in the site of gene N trol particle containing the N gene transcriptional antiter product utilization for antitermination of leftward tran mination protein of bacteriophage lambda. Cell 51:631- scription, f. Mol. Biol. 124: 195-221. 641. Schauer, A.T., D.L. Carver, B. Bigelow, L.S. Baron, and D.I. Lazinski, D., E. Grzadzielska, and A. Das. 1989. Sequence-spe Friedman. 1987. Lambda N antitermination system: Func cific recognition of RNA hairpins by bacteriophage antiter- tional analysis of phage interactions with the host NusA minators requires a conserved arginine-rich motif. Cell protein. /. Mol. Biol. 194: 679-690. 59: 207-218. Schwarz, E. 1980. Sequenz analyse des DNA lambdoide bacte Li, S.C., C.L. Squires, and C. Squires. 1984. Antitermination of riophages. Ph.D. thesis. University of Freiburg, Freiburg, E. coli rRNA transcription is caused by a control region seg Federal Republic of Germany. ment containing lambda nut-like sequences. Cell 38: 851- Sharrock, R.A., R.L. Course, and M. Nomura. 1985. Inhibitory 860. effect of high-level transcription of the bacteriophage

GENES & DEVELOPMENT 2221 Downloaded from genesdev.cshlp.org on September 30, 2021 - Published by Cold Spring Harbor Laboratory Press

Friedman et al.

lambda nutL region on transcription of rRNA in Escherichia coh. f. Bacteiiol. 163: 704-708. Somasekhar, G., D. Drahos, J.S. Salstrom, and W. Szybalski. 1982. Sequence changes in coliphage lambda mutants af fecting the nutL antitermination site and termination by tLl and tL2. Gene 20: 477-480. Spencer, C.A. and M. Groudine. 1990. Transcription elongation and eukaryotic gene regulation. Oncogene 5: 777-785. Stewart, V. and C. Yanofsky. 1985. Evidence for transcription antitermination control of tryptophanase operon expression in Escherichia coh K-12. /. Bacteriol. 164: 731-740. Susskind, M.M. and P. Youderian. 1983. Bacteriophage P22 an- tirepressor and its control. In Lambda II (ed. R.W. Hendrix, J.W. Roberts, F.W. Stahl, and R.A. Weisberg), pp. ^47-363, Cold Spring Harbor Laboratory, Cold Spring Harbor, New York. Swindle, J., M. Zylicz, C. Georgopoulos, J. Li, and J. Greenblatt. 1988. Purification and properties of the NusB protein of Escherichia coh. J. Biol. Chem. 263: 10229-10235. Tomich, P.K. and D.I. Friedman. 1977. Isolation of mutations in insertion sequences that relieve IS-induced polarity. In DNA insertion elements, plasmids and episomes (ed. A.I. Buk- hari, J.A. Shapiro, and S.L. Adhya), pp. 99-107, Cold Spring Harbor Laboratory, Cold Spring Harbor, New York. Tsugawa, A., T. Kurihara, M. Zuber, D.L. Court, and Y. Naka- mura. 1985. E. coh NusA protein binds in vitro to an RNA sequence immediately upstream of the boxA signal of bacte riophage lambda. EMBO. /. 4: 2337-2342. Warren, F. and A. Das. 1984. Formation of termination-resis tant transcription complex at phage lambda nut locus: Ef fects of altered translation and a ribosomal mutation. Proc. Natl. Acad. Sci. 81: 3612-3616. Wulff, D.L. and M. Rosenberg. 1983. Establishment of repressor synthesis. In Lambda II (ed. R.W. Hendrix, J.W. Roberts, F.W. Stahl, and R.A. Weisberg), pp. 53-73, Cold Spring Harbor Laboratory, Cold Spring Harbor, New York. Yanisch-Perron, C, J. Vieira, and J. Messing. 1985. Improved M13 phage cloning vectors and host strains: Nucleotide se quences of the M13mpl8 and pUC19 vectors. Gene 33: 103-119. Zuber, M., T.A. Patterson, and D.L. Court. 1987. Analysis of nutR, a site required for transcription antitermination in phage lambda. Proc. Natl. Acad. Sci. 84: 4514-4518.

2222 GENES & DEVELOPMENT Downloaded from genesdev.cshlp.org on September 30, 2021 - Published by Cold Spring Harbor Laboratory Press

Transcription-dependent competition for a host factor: the function and optimal sequence of the phage lambda boxA transcription antitermination signal.

D I Friedman, E R Olson, L L Johnson, et al.

Genes Dev. 1990, 4: Access the most recent version at doi:10.1101/gad.4.12a.2210

References This article cites 68 articles, 18 of which can be accessed free at: http://genesdev.cshlp.org/content/4/12a/2210.full.html#ref-list-1

License

Email Alerting Receive free email alerts when new articles cite this article - sign up in the box at the top Service right corner of the article or click here.