Phage HK022 Nun protein represses translation of phage λ N (transcription termination/translation repression)

Abstract

The N-terminal arginine-rich motif of phage HK022 Nun protein binds to NUT sequences in phage λ nascent transcripts and induces transcription termination. Interactions between the Nun C terminus and RNA polymerase as well as the DNA template are required for termination. We have isolated Nun C-terminal point and deletion mutants that are unable to block transcription. The mutants bind NUT RNA and inhibit translation of the λ N gene. Thus HK022 excludes λ both by terminating transcription on the phage chromosome and by preventing translation of the essential λ N gene. Like N autoregulation, translation repression by Nun requires host RNaseIII deficiency (rnc) or a mutation in the RNaseIII processing site (rIII) located between NUTL and the beginning of the N coding sequence. Our data support the idea that Nun bound at NUTL causes steric interference with ribosome attachment to the nearby N coding sequence. Two models, Nun acting alone or in complex with host proteins, are discussed.

Prophage HK022 excludes superinfecting λ by aborting transcription elongation on the λ chromosome (1). The responsible HK022 function is the 109-aa Nun protein. The N-terminal region of Nun includes an arginine-rich motif that binds the NUTL and NUTR sequences in λ nascent transcripts (2, 3). The C-terminal region makes contacts with the host NusA protein, RNA polymerase (RNAP), and DNA template (4–6). Nun induces transcription termination at sites just promoter-distal to nutL and nutR (1, 7, 8). Deletion of 13 C-terminal Nun amino acids stimulates NUT binding in vitro but abrogates transcription termination in vivo (6). In vitro, Nun alone induces transcription arrest rather than termination; termination and release of RNA require the host Mfd protein (R. Washburn, Y. Wang, and M.E.G., unpublished work). Arrest in vitro is stimulated by, but does not depend on, the host NusA, NusB, NusE, and NusG factors. In vivo, mutations in these nus genes block Nun termination in the λ pR operon (9).

The NUTL and NUTR sequences are essential for λ development. They constitute the site of attachment of λ N protein and the assembly of the transcription antitermination complex that includes N and the four Nus factors (10–12). Between NUTL and the start of the N gene lies a stem–loop region. This sequence is a substrate for RNaseIII, an endoribonuclease that cleaves double-stranded RNA found in stem loops (13–15). The stem loop interferes with translation of adjacent N mRNA (13). In addition, failure to cleave the RNaseIII site (rIII) induces a negative autoregulatory loop, in which N strongly represses its own translation (16–18). It is likely that the down-regulation of N is important for late stages in the λ lytic cycle as well as for entry into the lysogenic pathway (17, 18).

Among mutations that abrogate antitermination, the nutL mutations boxB44 and boxA5 block N-mediated translation repression, whereas nusA1, nusB5, nusE71, and boxA16 have only marginal effects on N translation (18). This finding raised the possibility that N binding to NUTL might suffice to block ribosome attachment. Alternatively, the nus and boxA16 mutations might allow initial assembly but not persistence of an antitermination complex. This initial unstable antitermination complex might provoke translational repression. Several results, including those with multiple mutations, support the latter model, in which a transcription complex with N is responsible for translation repression. Thus, although nusB5 or nusA1 alone have little or no effect on translational repression by N, repression is abolished in a nusB5 nusA1 double mutant (H.R.W., unpublished work).

We show below that Nun C-terminal mutants defective in transcription termination retain the ability to bind NUTL and inhibit N translation. The requirements for translation repression overlap but are not identical to those for transcription termination. Point mutations in the RNaseIII processing region between nutL and N have been selected that block translation repression.

Materials and Methods

Bacterial Strains and Plasmids.

Standard bacteriological techniques, e.g., transformation, transduction, and media preparation, are as described (19). The bacterial hosts are W3110 or the congenic W3102. Strains carrying a defective λ cI857 Nam prophage in which the pR-nutR-tR1 region is fused to galETK or to galEoc derivatives are described (20). The λ pR promoter is activated by thermal denaturation of the cI857 repressor. Strains carrying a defective λ cI857 prophage in which the pL-nutL-N region is linked to lacZ by either an operon or a protein fusion are described (18), with the exception of the NUTL boxA69 mutant, which was constructed for this work by using the protocol described (18). All plasmids were constructed from pET-21d vector (Novagen). Plasmids expressing wild-type Nun, Nun K59, Nun R78, Nun V96, and Nun K106/107D are described in detail below.

Mutagenesis.

Plasmids expressing Nun and Nun derivatives were constructed as follows. C-terminal truncated versions of Nun were generated in the pET21d-Nun vector (Novagen; ref. 6) by using the following primers: 5′ primer (K59, R78, V96), CCT ATA GTG AGT CGT ATT AAT TTC T, 3′ primer (K59) GGT GAA TTC TTA CTT ATA TTC AGG ACC GTC ATC TAT AGC, 3′ primer (R78) GGA GAA TTC TTA GCC AGG AAA TGG TGT TGC TAA TGC, 3′ primer (V96) GGA GAA TTC TTA AAC AAC TCT GTG CAT TAC ATC CTC ATA.

Truncations of the Nun gene were amplified by PCR from a pET21d plasmid carrying the full-length Nun gene (16). The 5′ primer is complementary to the T7 promoter upstream of the Nun gene and includes the NcoI restriction site. The 3′ primers for the truncated genes contain a stop codon (ATT) at the position corresponding to I60, D79, or N97, and also contain a restriction site for EcoRI. The amplified DNA fragments were purified on a 2% agarose gel and isolated by using the Qiaex DNA recovery system (Qiagen, Chatsworth, CA). The purified Nun gene fragments and the pET21d plasmid were then digested with NcoI and EcoRI to generate complementary ends for ligation of the fragment into the plasmid. Digestion was confirmed by analysis on agarose gels stained with ethidium bromide. The full-length and shortened Nun genes were incorporated into the pET21d plasmid by incubation with T4 DNA ligase (Roche Applied Science, Indianapolis) at 10°C overnight. The ligated mixture was used to transform XL1-Blue cells on LB/Amp agar plates for propagation of the newly constructed plasmid. Colonies were screened by PCR amplification of Escherichia coli cells by using the same primers that were used for the above PCR reactions.

The K106/107D mutant was made with the QuikChange site-directed mutagenesis technique (Stratagene) from the pET21d-Nun plasmid. The primers used were GCT CAC CAG CGA AAC CCA AAC GAC GAC TGG TCA TAA AAG CTT GC and GCA AGC TTT TAT GAC CAG TCG TCG TTT GGG TTT CGC TGG TGA GC. The K106/107D mutation was confirmed by sequencing.

Enzyme Assays.

Cultures were grown in LB supplemented with the appropriate antibiotics at 32°C and shifted to 42°C for 7 h for galactokinase measurements or for 1 h for β-galactosidase assays. The temperature shift inactivates the λ cI857 repressor and initiates transcription for the λ pL promoter. Galactokinase and β-galactosidase activities were determined as described (21, 22).

Results

Nun C-Terminal Mutations That Eliminate Transcription Termination.

Within a Nun arrested transcription elongation complex, the C terminus of Nun contacts the DNA template 7- to 8-bp promoter distal to the active center of RNAP. A planar hydrophobic amino acid residue, tryptophan at position 108 of Nun, is essential for this contact and for Nun transcription termination (5). We suspected that lysine residues K106 and K107 of Nun might stabilize this Nun/template interaction by neutralizing the negative charge on phosphate groups of DNA template. We therefore mutated K106 and K107 to aspartate and determined the efficiency of transcription termination by the ability of the mutant Nun to exclude phage λ. Substitution of both lysines with negatively charged aspartate residues completely abolished exclusion. λ plated with an efficiency of <10⁻⁶ on strains expressing wild-type Nun and with an efficiency of 1.0 on the Nun K106/107D mutant (data not shown).

We then compared the efficiency of termination induced by wild-type and Nun K106/107D in a λ pR-nutR-tR1-galK fusion (Table 1). Wild-type Nun almost completely blocked readthrough of the galK reporter gene. Strikingly, the K106/107D mutation enhanced expression of galK. This suggested to us that Nun K106/107D interacts with NUTR RNA and may block the access of Rho factor to the neighboring tR1 terminator. This Rho-dependent site has a termination efficiency of 50–60% (23). The stimulation of readthrough, which is seen as well with nus mutations that block Nun termination (9), does not represent modification of RNAP to an antitermination complex. Thus, Nun K106/107D did not suppress polarity at distal terminators activated by polar ochre mutations in galE (Table 1). In contrast, N fully suppresses polarity in these galEoc mutant fusions (9).

Table 1.

Termination properties of Nun C-terminal lysine mutants

	VMHRVVNHAHQRNPNK106K107WS
FUSION: pR-cro-nutR-tR1-galETK
gal	nun
	−	+	K106/107D
+	2.3	0.3	5.9
Eoc95	0.1	0.1	0.2
EocB4	0.1	0.1	0.2

Cells were diluted from cultures grown overnight at 32°C and incubated at 42°C for 7 hr. Values represent galactokinase units (21); a background value of 0.4 units, representing a galK control, has been subtracted. The Nun C-terminal amino acid sequence, including the adjacent (K106K107) lysines, is shown. The strains used are described (20). W108 is required for interaction with the λ DNA template and transcription termination. K106/107D represents a double mutant of Nun in which both lysines are replaced by aspartate residues. The galEoc95 and galEocB4 mutations are polar on galK. The wild-type and mutant fusions are described (29). 

Inhibition of N Translation by Nun K106/107D.

The λ pL transcript is cleaved between NUTL and the N ribosome-binding site by RNaseIII (14, 15). When cleavage does not occur, either because of RNaseIII deficiency or deletion of the RNaseIII site (rIIIΔ), N inhibits its own translation (17, 18). It was suggested that the folding of pL RNA to form the rIII structure brings NUTL close to the beginning of N. If NUTL is bound to N and the antitermination complex, the ribosome-binding site of N is occluded. The isolation of Nun K106/107D, which does not terminate transcription efficiently, allowed us to ask whether Nun bound to NUTL would likewise inhibit N translation. Table 2 shows the results of these experiments. Using a pL-nutL-N-lacZ operon fusion (13), we showed that unlike Nun⁺, NunK106/107D allows readthrough expression of the reporter gene in either a wild-type or an rnc mutant host. In contrast, lacZ expression from a pL-nutL-N∷lacZ protein fusion (18), was completely inhibited by Nun K106/107D in the rnc but not the wild-type host. We conclude that the C-terminal Nun mutant inhibits N translation when the pL-nutL-N transcript remains intact. As was demonstrated for translation repression by N, Nun represses translation only of the gene fusion.

Table 2.

Nun K106/107D inhibits N translation in an rnc⁻ strain

Fusion	nun
	rnc	−	+	K106/107D
Operon	+	288	<1	82
	rnc14∷Tn10	618	<1	103
Protein	+	350	5	162
	rnc14∷Tn10	190	1	2

Both fusions include λ pL, the N leader, and the first 33 codons of N fused to lacZ at residue nine. With respect to the operon fusion, λ pL-nutL-N*lacZ, a synthetic sequence (*) containing translational stop codons in the N reading frame and the ribosome-binding site and 5′-end of the lacZ structural gene is inserted between the N leader sequence and lacZ (23). Cells were grown in LB supplemented with the appropriate antibiotic at 32°C to early log and then shifted to 42°C for 1 hr. Units represent β-galactosidase activity (22). 

Comparing Nun K106/107D and Nun C-Terminal Deletion Mutants.

We next asked whether C-terminal deletions of Nun affected translation repression. Several such deletions were constructed. As expected, Nun V96 (1–96), R78 (1–78), and K59 (1–59) all failed to provoke transcription termination at nutL (data not shown). Nun V96 does not bind RNAP in vitro, although the effect of Nus factors on this binding has not been tested (5). In contrast, all of the Nun deletion mutants supported translation repression in an rnc strain (Table 3; data not shown). Recall that the RNA-binding arginine-rich motif is present in all of the mutant proteins.

Table 3.

Effect of rIII deletion on translation repression by Nun and Nun deletion mutants

Nun	rnc	rIII	β-galactosidase	% expression
−	+	+	1081	100
1–59	+	+	1016	94
K106/107D	+	+	500	46
−	−	+	758	100
1–59	−	+	83	11
K106/107D	−	+	7	1
−	+	Δ	1860	100
1–59	+	Δ	1122	60
K106/107D	+	Δ	80	4
−	−	Δ	3662	100
1–59	−	Δ	1747	47
K106/107D	−	Δ	123	3

Induction is at 42°C for 1 hr. β-galactosidase levels are in Miller units (22). 

For N autoregulation, deletion of the rIII site is phenotypically equivalent to inactivation of rnc (17). N represses its own translation in either an rnc⁻ host or a pL transcript deleted for rIII (17, 18). Table 3 shows that Nun K106/107D efficiently repressed N translation in a rnc⁺ host in a pL-nutL-N∷lacZ fusion lacking the rIII site (rIIIΔ). Interestingly, Nun K59 was inactive on the rIIIΔ fusions (Table 3). Nun R78 and Nun V96 similarly failed to down-regulate the rIIIΔ fusion (data not shown). The loss of activity of the Nun C-terminal deletion mutants was due to deletion of the rIII site rather than presence of active RNaseIII in the host. Introduction of a rnc mutation into the rIIIΔ strain did not restore the activity of the Nun C-terminal deletion mutants (Table 3; data not shown).

Effect of NUTL and Host Mutations on Translation Repression.

Nun termination at NUTR is ablated by mutations in the host nusA, nusB, nusG, or rpoC genes (20, 24). These mutations have little effect at NUTL. We believe this difference reflects the fact that NUTL is closer to λ pL than NUTR is to its cognate promoter, λ pR, rather than a difference between the two NUT sequences. Thus converting NUTL to NUTR did not block Nun termination in nusG mutants (24). In contrast, a nusA mutation inhibited Nun termination at a NUTL located ≈1 kb distal to a ptac promoter (ref. 24; R. Washburn, personal communication). We asked whether nus and rpoC mutations affected translation inhibition by Nun K106/107D. As shown in Table 4, Nun K106/107D was partially active in repression of a N∷lacZ protein fusion in the nusA E132K and rpoC mutant backgrounds. However, Nun K106/107D was fully active in a nusB deletion strain. We conclude that transcription termination and translation repression share some, but not all, host requirements.

Table 4.

Translation repression in nus and rpoC mutants

Mutation	nun		% expression
	−	K106/107D
Experiment I
None	535	27	5
nusA E132K	456	85	19
rpoC D329G	1597	488	31
rpoC D264G	1485	256	17
rpoC R322H	1089	227	21
Experiment II
None	444	8	2
nusBΔ	992	17	2

Strains in Experiment I, rIIIΔ; Experiment II, rnc14∷Tn10. Induction is for 1 hr at 42°C. Values indicate β-galactosidase activity in Miller units (22). NusBΔ strains are cs. In Experiment II, therefore, strains were grown to mid-log at 37°C prior to shifting to 42°C for 1 hr. 

A NUTL deletion or the NUTL BOXB44 mutation strongly reduced translation repression by Nun K106/107D (Table 5, Experiment II). The NUTL BOXB44 mutation is known to abolish Nun binding in vitro (2). Of particular interest were mutations in boxA, the site of organization of the NusB/NusE/NusG complex (ref. 10 and Fig. 1A). The boxA16 and boxA5 mutations eliminate Nun transcription termination at nutR (20). boxA5 but not boxA16 abrogated termination at nutL (Table 5, Experiment I). boxA69, a transversion of all boxA nucleotides, likewise suppressed transcription termination at nutL. In contrast, the boxA5, boxA16, and boxA69 mutations still allowed translation repression (Table 5, Experiment II). This is consistent with the ability of Nun K106/107D to inhibit translation in strains lacking NusB. We conclude that, whereas transcription termination by Nun at NUTR entails assembly of the full Nus-RNAP complex, including NusB, NusE, and NusG, the requirements for transcription termination at NUTL and translation repression of N are less strict.

Table 5.

Effect of nut mutations on transcription termination (Experiment I) and translational inhibition (Experiment II)

Mutation	nun		% expression
	−	+
Experiment I
None	790	7	<1
boxA69	1040	964	93
boxA5	1299	988	76
boxA16	1164	126	11

Mutation	nun		% expression
	−	K106/107D
Experiment II
nutLΔ	590	508	86
boxB44	365	146	40
boxA69	228	21	9
boxA5	315	9	3
boxA16	612	35	6

Strains in Experiment I, rnc⁺; Experiment II, rnc14∷Tn10. Induction is for 1 hr at 42°C. Values indicate β-galactosidase activity in Miller units (22). 

Figure 1.

(A) Wild-type and mutant λ NUTL sequences. (B) Part of the λ pL transcript showing NUTL, the RNaseIII-sensitive structure in mature λ pL transcript, and the beginning of the N gene. The nun1 and nun3 mutations are indicated. (C) An RNaseIII-sensitive site thought to form during λ pL transcription (9, 10). The nun3 mutation is indicated.

Isolation of λ Mutants Resistant to Nun K106/107D.

Repression of λ N translation by Nun K106/107D results in exclusion of infecting λ. As shown in Table 6, λ failed to form plaques in a rnc mutant strain expressing Nun K106/107D from a multicopy plasmid (N8993). Interestingly, λ deleted for the pL operon rnc site (λ rIIIΔ) propagated in this host. Presumably, Nun K106/107D requires the rIII site to repress N expression in an infecting phage, although deletion of rIII has little effect on N translation repression in a fusion assay (Table 3).

Table 6.

Plating of λ and λ rIIIΔ mutants

rnc	nun	λ	λ rIIIΔ
+	−	+	+
+	+	−	−
+	K106/107D	+	+
−	−	+	+
−	+	−	−
−	K106/107D	−	+

Phages were plated at 42°C on strain WJW244 (18), carrying plasmid pET21d-Nun⁺ or pET21d-Nun K106/107D. 

The difference in plating efficiency of λ and λ rIIIΔ suggested a means to select mutations in the rIII site that alleviate translation repression. Accordingly, we selected five λ mutants that grew on N8993 by passing wild-type λ through a host carrying mutations in mutS, mutD, and mutT (XL1-Red, Stratagene). The pL to N regions (nucleotides 35,239–35,641) of each were sequenced, and mutations were detected in three of the five mutant phages. One mutation, a G to T transversion at nucleotide 35,559, lies in the −10 region of the pL promoter and may enhance its activity. Two mutations lie between nutL and N, each in a potential base pair structure (Fig. 1 B and C). To verify that the phenotype of the mutant phage was due to a single mutation, three oligonucleotides, each carrying one of the three mapped mutations, were recombined into wild-type λ (25, 26). As predicted, if a single mutation was responsible for the phenotype, the recombinant phage propagated on WJW244/pNun K106/107D (data not shown).

Discussion

Phage HK022 Nun protein blocks phage λ growth by terminating transcription at sites just distal to the NUTR and NUTL sites in the nascent λ transcript. The reaction entails contacts of the Nun C terminus with λ DNA template and RNAP. These interactions are aided by the binding of Nun N terminus arginine-rich motif to NUTR or NUTL and association with the host Nus factors.

We have isolated Nun C-terminal point and deletion mutants that fail to terminate λ transcription but can still bind λ NUT (H.C.K. and M.E.G., unpublished work; this work). The Nun point mutant carries lysine to aspartate mutations at residues 106 and 107. The inability of Nun K106/107D to terminate transcription effectively may be due to electrostatic repulsion between Nun and the λ DNA template, preventing the Nun C terminus from contacting DNA and anchoring RNAP. In vivo, Nun K106/107D increases readthrough of λ pR transcription past the Rho-dependent λ tR1 terminator cluster. We suggest that Nun binding to NUTR interferes with the access of Rho to its RNA-binding site, which overlaps BOXB (27). Binding of Nun K106/107D at NUTL has little effect on the λ pL operon and N expression in a wild-type background but represses λ N translation in a rnc⁻ strain. Even Nun mutants deleted for C-terminal residues required for in vitro interactions with RNAP and DNA effectively block N translation in a rnc⁻ host.

Translation repression of N by Nun resembles the negative autoregulatory loop that controls the translation of N in a rnc⁻ strain or with a fusion that lacks the rIII between NUTL and the beginning of N (18). A mechanism for the N autoregulatory loop has been proposed (18). In this model, λ NUTL is approximated to the N ribosome-binding site by the intervening stem–loop structure of rIII. The proximity of NUTL and N is normally transient, because rIII is cleaved by RNaseIII, separating the two parts of the λ pL transcript. However, in a rnc⁻ strain or with a λ pL transcript that lacks rIII, binding of the N antitermination complex at NUTL occludes ribosome entry and represses N translation. The notion that formation of the entire antitermination complex, RNAP, N, NusA, NusB, NusE, and NusG, is required for translation repression is supported by some observations but not by others. Thus, the nutLΔ, boxA5, and boxB44 mutations eliminate both antitermination and translation repression. However, boxA16 and the nusA1, nusB5, and nusE71 mutations, which inhibit antitermination, have little effect on N autoregulation. It was suggested that boxA16 and the individual nus mutations permit initial formation of the antitermination complex, thus occluding ribosome entry, and cause the complex to be unstable for long range antitermination.

Nun termination, like N antitermination, entails the assembly of RNAP, NusA, NusB, NusE, and NusG complex at the NUT sites. However, unlike N, Nun acts at or close to its site of recognition, obviating the requirement for a persistent complex between Nun, NUT, and the Nus factors. It was, therefore, of interest to test the roles of the components of the Nun termination complex in translation repression. Deletion of nutL or boxB44 mutation eliminates both termination and translation repression. boxB44 mutation eliminates Nun binding to NUTL in vitro (2). However, mutations in boxA do not restore N translation, whether they abrogate termination (boxA5) or not (boxA16). NusB binds BOXA and recruits NusE and NusG into the N antitermination complex and, presumably, into the Nun termination complex as well. This suggests that these Nus factors do not play a role in translation repression. Consistent with this idea is the fact that deletion of nusB did not affect translation repression. We also tested nusA and rpoC mutations that eliminate termination at nutR. These mutations reduced, but did not abrogate, translation repression.

Although the formation of a termination complex at NUTR is supported by genetic evidence, a similar complex at NUTL is quite problematic. Thus, depletion of NusG blocks termination at NUTL, whereas deletion of nusB does not (ref. 24; data not shown). Furthermore, nusA E132K and the rpoC mutations described above inhibit transcription termination at NUTR but not at NUTL (20). The relevant distinction between the two nut sites is the distance from their cognate promoters (R. Washburn and M.E.G., unpublished work). Taken together, these data suggest that the Nun termination complex that forms at NUTR plays, at best, a marginal role in transcription termination or translation repression at NUTL. The distinctive ability of boxA5 to block transcription termination is not understood but may be related to an inhibitor thought to compete with NusB for binding to BOXA (28).

Whereas the point mutant Nun K106/107D represses N translation in either a rnc⁻ strain or a rIIIΔ fusion, Nun C-terminal deletion mutants were effective only in the former. It is possible that the full-length Nun interacts with the rIII site via C-terminal residues, although alternative explanations have not been ruled out. For example, the cellular concentration of Nun K59 is significantly lower than wild-type Nun (A. Stuart, personal communication). Higher Nun levels may be required to repress N translation in a fusion that lacks the rIII site.

Finally, we found that λ failed to propagate on a rnc⁻ strain expressing Nun K106/107D, although λ rIIIΔ grew with an efficiency of plating of one. Two λ mutants that formed plaques on the nonpermissive strain carried mutations that altered the λ pL transcript between NUTL and the N ribosome-binding site (Fig. 1B). λ nun3 is a C→U transition at position 67, whereas λ nun1 carries a C→A transversion at position 206 (Fig. 1B). Nascent λ pL transcripts are cleaved by RNaseIII between positions 71 and 72, in a RNA duplex region (refs. 13 and 14 and Fig. 1C). The C67U mutation might be expected to disrupt the duplex and eliminate cleavage. Cleavage at C71, however, is seen only during transcription, suggesting that this structure is transient. Mature λ pL transcripts are cleaved at positions 88 and 197, and a structure consistent with this pattern has been observed (ref. 15 and Fig. 1C). In this structure, C206A disrupts a predicted duplex region immediately promoter-proximal to N and might likewise inhibit cleavage. However, inactivation of the rIII site(s) clearly does not account for the ability of the mutations to support λ growth in a rnc⁻ strain. It is possible that both mutations eliminate RNA contacts that stabilize Nun attachment to the λ pL transcript. Alternatively, the mutations may increase the expression of N, which is inhibited directly by the RNA stem. The increased N would compete with Nun for binding to NUTL (2). With small amounts of λ pL transcription, as in the case of single copy fusions, the effect of a rIII mutation on translation repression by Nun K106/107D is not significant. Replication by infecting λ, however, would reduce the ratio of Nun K106/107D to λ pL transcript and would, in addition, express N. This would then amplify the effect of a rIII mutation, allowing N translation and phage growth.

Wilson et al. (17) have demonstrated that strains growing in poor media express low levels of RNaseIII and exhibit N autoregulation. Under these conditions, we think it is likely that Nun, in addition to terminating transcription, would repress N translation. HK022 may, therefore, possess two mechanisms for blocking λ growth.

Abbreviations

RNAP

RNA polymerase

rIII

RNaseIII processing site