Functional Expression and Properties of the tRNALYs-specific Core Anticodon Nuclease Encoded by Escherichia coli prrC*

Escherichia coli carrying the optional locus prr har- bor a latent, tRNALY*-specific anticodon nuclease, activated by the product of phage T4 stp. Anticodon nuclease latency is ascribed to the masking of prrC, implicated with the enzymatic activity, by flanking, type IC DNA restriction modification genes (prrA, B&D-hsdM, S&R). Overexpression of plasmid-borne prrC elicited anticodon nuclease activity in uninfected E. coli. In vitro, the prrC-coded core activity was indifferent to a synthetic Stp polypeptide, GTP, ATP, and endogenous DNA, effectors that synergistically activate the latent enzyme. Several facts suggested that PrrC is highly labile in the absence of the masking proteins. The core activity decayed with t% below 1 min at 30 “C, and the PrrC portion of a fusion protein was unstable. Moreover, expression of prrC from its own promoter at low plasmid copy number did not allow detection of core activity. Yet, it sufficed for establishment of a latent, T4-inducible enzyme when complemented by the masking Hsd proteins, which were provided by another replicon. Interaction between the antagonistic components of latent anticodon nuclease was also demonstrated immunochemically. The coupling of anticodon nuclease with a DNA restriction modification system may serve to ward off its

tions. Polynucleotide kinase hydrolyzes the 2':3'-cyclic phosphodiester as well as the resulting 3'-phosphomonoester. It also phosphorylates the 5'-OH end. RNA ligase joins the 3'-OH and 5'-P termini thus formed. However, in T4 pnkor rliinfections, cleaved intermediates accumulate, and intact tRNALY8 is depleted (6,8). Loss of tRNALy" in the absence of pnk and rli-mediated repair can account for cessation of late T4 translation and other lesions seen in the abortive infections (1)(2)(3). Suppression of prr restriction by stp lesions is attributed to a failure to activate ACNase (7).
Molecular cloning and mutational analysis has revealed that prr comprises positive and negative ACNase functions. prrC is implicated with core ACNase activity whereas the flanking genes, prrA and prrD and perhaps also prrB mask prrC's activity (9). Except for a GTP binding motif (to be discussed later), the primary sequence of PrrC resembles no known protein. In contrast, prrA and B&D are homologous with plasmid-borne hsd genes encoding type IC restriction modification systems such as EcoR124 (10). Genetic linkage of prr with such a system has been suggested previously (4). More recently latent ACNase has been precipitated by antibodies specific to the EcoR124/3 enzyme (11) and prr has been shown to encode type IC restriction modification activity. ' These facts indicate that Hsd proteins assumed an added role of ACNase masking.
A polypeptide encoded by stp (12) may alleviate the masking of PrrC during T4 infection. This conclusion is inferred from activation of ACNase i n uitro, by synthetic polypeptides of the deduced Stp sequence (11,13), or in uiuo, by expression of stp from a plasmid in uninfected E . coli ~r r + .~ However, participation of Stp in tRNALy" cleavage as an activator or cofactor of core ACNase has not been excluded.
Here we show that transcriptional induction of prrC in the absence of other prr genes causes translation of the PrrC polypeptide and elicits core ACNase activity. The advent of an in uitro core activity permitted us to distinguish between requirements of ACNase activation and tRNALy" cleavage. We also show that PrrC is exceedingly labile in the absence of the cognate Hsd proteins.

EXPERIMENTAL PROCEDURES
Materials-Synthetic Stp (residues 2-29) was ordered from Multiple Peptide Systems and purified by high pressure liquid chromatography as previously described (13). Rabbit anti-EcoR124/3 serum was kindly provided by Dr. Thomas A. Bickle, Biozentrum. Protease inhibitors and immunological reagents were purchased from Sigma. The enhanced chemiluminescence kit and radioactive materials were purchased from Amersham Cow.
Strains-Properties and sources of bacterial strains, phage, plasmids, and cosmids used are listed in Table I.
Phmid Constructs-The prrC expression plasmid pRRC6 ( Fig. 1) C. Tyndall and T. Bickle  The downstream portion of prrC was provided by the XmaI-HpaI fragment spanning bp 1801-2968 of the partial prr insert of plasmid pRR39 (9) and abutted by a 5-bp SmaI-BamHI fragment from the multiple cloning site of plasmid pT7-6. The combined EcoRI-BamHI fragment was moved into the corresponding sites of vector pT7-5 to generate pRRC6. A glutathione S-transferase-prrC fusion was constructed from (i) a prrC containing fragment, (ii) the gst gene, and (iii) a derivative of the expression vector pKK223-3 (Pharmacia LKB Biotechnology Inc.). To fuse the gst and prrC genes, an NdeI site was created at the ATG start codon of prrC in PCR oligonucleotide 3, GGCATCA-TATGGGCAAGACA. A 1.27-kilobase fragment bordered by it and the reverse PCR oligonucleotide 2 was amplified. The upstream 152bp NdeI-XmaI portion of the product was cloned into the NdeI-XmaI site of plasmid pT7-7, and its sequence was verified. The downstream portion of prrC was provided in the same manner as in pRRC6 to generate plasmid pRRC5. gst was amplified from plasmid pGEX-3X (Pharmacia) using PCR oligonucleotide 4, CATATGAGCCCTA-TACTAGG, and the reverse PCR oligonucleotide 5, GGGGATCTCGCGACCTTCGAT. The resulting fragment contained an NdeI site at the ATG start codon, a proteolytic site for the coagulation factor Xa and a 3'-NruI site. Following cleavage at the NdeI site it was cloned between the NdeI and SmaI sites of plasmid pT7-7 to generate plasmid pGST2. The nucleotide sequence of this insert was verified. A prrC fragment from pRRC5 was excised in two steps, first by NdeI cleavage followed by blunting with Klenow polymerase and further cleavage at the ClaI site. The resulting prrC insert was ligated between the NruI and ClaI sites of plasmid pGST2. The resultingplasmid, pGC1, was transformed into E. coli K38pGPl-2 for expressing the fusion protein GST-PrrC upon thermal induction of T7 RNA polymerase. A more efficient clone, pGC2, was generated by moving the XbaI-Hind111 gst-prrC-containing fragment of pGCl between the EcoRI-Hind111 sites of pKKN2, a derivative of the vector pKK223-3. pKKN2 was generated by stepwise deletion of pKK223-3. First a small SmaI-HincII portion of the polylinker, containing a BamHI site was removed. Next, a large BamHI-NdeI fragment, containing the inactive remainder of tep gene as well as the rop gene, was deleted.
Plasmids expressing prrC from its own promoter were generated by cloning a HpaI fragment of pRR39 (1391-2969 bp) ( Fig. 1) into three different vectors. pRRlO was generated by placing the insert into the SmaI site of the multicopy plasmid pBluescript KSf (Stratagene) with prrC oriented opposite to the vector's lac promoter. pACCl was generated by moving the prrC insert of pRR10, using flanking BamHI and SalI sites, into corresponding sites of pACYC184 (18 copies/cell chromosome) (14). pGBCl was generated by moving theprrC insert of pRRlO into the low copy plasmid pGB2 (3-5 copies/ cell chromosome) (15), using flanking EcoRI and SalI sites. Inactivation of prrC within the prr locus of cosmid pW16 (9) was achieved by homologous recombination, using a pRR39 derivative in which a BclI fragment of prrC was replaced by a tetracycline resistance cassette derived from plasmid pNK861 (16). This yielded cosmid pW16prrC:tet.
Anticodon Nuclease Assays-To monitor ACNase in uiuo, E. coli cells were pulse-labeled with ["PIPi. Low molecular weight RNA was extracted, before infection or at indicated times after infection, with T4 pseTA1 (pnk-). The RNA was separated on denaturing polyacrylamide-urea gel, as previously described (5). The in vitro ACNase assay is based on cleavage of tRNALYB labeled with 32P in the cleavageligation junction. This generates labeled fragment 1-33 carrying a cyclic phosphate end group. The preparation of the substrate was described before (13). Crude core ACNase was derived from thermoinduced E. coli K38:pGP1-2:pRRC6 cells as follows. All operations were carried out between 0 and 4 "C. The cells were grown in LB medium at 37 "C and harvested at Am of 1.2. Packed cells (2 g) were washed once in buffer A (0.5 M NaC1, 50 mM Tris-HC1 buffer, pH 8.0; 10 mM EDTA, 5 mM 0-mercaptoethanol, 10% glycerol, and 10% sucrose) and twice in buffer B (similar to buffer A but containing only 10 mM NaC1). The washed cells were suspended in 2 ml of buffer B containing 0.2 mM diisopropyl fluorophosphate, 10 pg/ml aprotinin, 50 pg/ml leupeptin, and 3.5 mM phenylmethylsulfonyl fluoride. The suspension was passed twice in an Aminco pressure cell at 18,000 p.s.i. The cell lysate was centrifuged for 30 min at 20,000 X g, and the supernatant (S-30) was further centrifuged for 4 h at 45,000 rpm in a 50Ti Beckman rotor. The soluble fraction (S-150) was used as source of enzyme. The standard anticodon nuclease reaction mixture (10 pl) contained 5 pl of core ACNase (the S-150 fraction diluted 2-10-fold in buffer B, as indicated), 5,000 cpm of the 32P-labeled tRNALy" (3000 Ci/mmol), 25 mM Tris-HCI (pH 8.0), 5 mM NaC1, 10 mM MgClZ, 5 mM 0-mercaptoethanol, 5 mM EDTA, 5% glycerol, and 5% sucrose. It should be noted that the reaction mixtures contained about 10-100 ng of endogenous tRNALY" partly cleaved in vivo. The standard incubation temperature was 10 "C. The reaction was stopped by precipitating the RNA with ethanol. The pellet was washed with 70% ethanol and dissolved in 10 p1 of 10 M urea containing 0.01% xylene cyanol. An aliquot was separated by electrophoresis on 15% polyacrylamide in 7 M urea, 25 mM Tris borate buffer (pH 8.3), 2.5 mM EDTA. Following autoradiography, intact tRNALy" and fragment 1-33 bands were excised from the gel and counted in liquid scintillation fluid. Activity is expressed as percent tRNALYe 32P label converted into fragment 1-33.
Production of PrrC Polyclonal Antiserum-The GST-PrrC fusion protein was purified from E. coli XL1-Blue: pGC2. The cells were  grown on T B (1.2% Bacto-tryptone, 2.4% Bacto-yeast, 0.4% glycerol) (20) medium a t 37 "C to optical density -1.2 with vigorous aeration and induced for 2 h with isopropyl-P-D-thiogalactopyranoside. Sodium azide was added to 0.02%. After 10 min the cells were harvested and lysed by a pressure cell as above. The fusion protein was purified from the S-30 extract by affinity to glutathione. Analysis by SDSpolyacrylamide gel electrophoresis revealed a series of products (  PrrCN). The band containing it was excised from the gel and injected subcutaneously into rabbits. The resultant antiserum was used for immunological detection of free PrrC and the PrrC. Hsd complex (Fig. 9). Immunization, production of polyclonal antiserum, immunoprecipitation, and Western blotting were carried out as described by Harlow and Lane (17). Detection of secondary antibodies by enhanced chemiluminescence was according to the manufacturer's instructions.
Expression Systems-Thermoinduction and exclusive labeling of the overexpressed protein in the T7 promoter/polymerase system were carried out as described by Tabor and Richardson (18). Overexpression and purification of the GST-PrrC fusion protein were carried out as described by Smith and Johnson (19).
General Procedures-General techniques of gene cloning were adapted from Sambrook et al. (20) and PCR protocols from Innis et al. (21).

Induction of prrC Elicits ACNase Activity in Uninfected E.
coli-To examine the function of the prrC ORF we investigated the consequences of its transcriptional induction in the absence of other prr genes ( p r o background). prrC was placed under control of the T 7 promoter in plasmid pRRC6 ( Fig. 1 and "Experimental Procedures"). It was expressed in the T 7 polymerase/promoter system in which transcription of target genes is turned on by thermoinduction of T 7 RNA polymerase from the coresident plasmid pGP1-2 (18). The induction elicited ACNase activity, indicated by appearance of typical cleavage products migrating with tRNALY" fragments 34-76 and 1-33 (Fig. 2, lane 5). Such fragments are seen in T 4 pseTA1 (pnk-)-infected E. coEi prr+ cells (Fig. 2, lane I ) .
Manifestation of ACNase activity in the uninfected transformant cells depended on the prrC insert and thermoinduction of T7 RNA polymerase. Cells containing the vector plasmid (Fig. 2, lanes 2 and 4 ) or pRRC6 but not thermo-induced (Fig.  2, lane 3) did not express detectable activity.
The thermo-induced cells were metabolically labeled with [3sS]Met and their proteins separated by denaturing SDSpolyacrylamide gel electrophoresis to monitor the polypeptide expressed from prrC. An extract from cells containing the vector plasmid served as a control. This analysis revealed a prrC-specific translation product with an apparent molecular mass of 45 kDa (Fig. 3, lane 2), close to the size predicted for PrrC (45,404 daltons). Thus, transcriptional induction ofprrC over the prro background elicits concomitant appearance of a PrrC-like polypeptide and overt ACNase activity.
I n Vitro Core ACNase Actiuity-Core ACNase activity was assayed in uitro, in S-150 fractions derived from the thermoinduced K38pGP1-2:pRRC6 transformants. It should be noted that the crude S-150 fraction used in this study contained an undetermined quantity of endogenous tRNALy" and tRNALy8 fragments generated by ACNase cleavage in uiuo.
Instability of the core activity (see below) impeded further purification of the enzyme. The endogenous tRNALy8 was mixed with tracer amounts of [32P]tRNALy' labeled in the cleavage-ligation junction. ACNase generates from this substrate fragment 1-33 as the only labeled product (13). Due to uncertainties about the effective concentration of the substrate, the activity of core ACNase was expressed as percent of input tRNALYn radioactivity converted into fragment 1-33. Fig. 4 (lanes 3-7) shows a time course of tRNALy" cleavage at 10 "C with an S-150 supernatant fraction derived from thermo-induced E. coli K38:pGP1-2:pRRC6 cells. The cleav-  core ACNase. A , core ACNase was assayed at the indicated temperatures, essentially as described in Fig. 4, except that the enzyme was diluted 1:5 in buffer B. B, core ACNase was preincubated at 30 "C in the absence of radioactive substrate. Aliquots withdrawn at the indicated times were assayed at 10 "C in the presence of the radioactive substrate.

TABLE I1
Agents that activate latent ACNase do not stimulate the core enzyme The S-150 fraction from thermoinduced K38:pGP1-2:pRRC6, containing core ACNase, was diluted 5-fold in buffer B and was assayed for 10 min as described in the leeend to Fie. 4.

Reaction mixture
Extent of tRNALp Activity of standard cleavage" reaction mixture  for 30 min at 0 "C. In this experiment the standard mixture was preincubated similarly but with buffer devoid of DNase I. position of fragment 1-33 (Fig. 4, lanes 1 and 2 ) . The specificity of cleavage was ascertained by hydrolysis of the labeled 2':3'-cyclic phosphate end group of the product in a reaction catalyzed by 3'-phosphatase-polynucleotide kinase (not shown). Incubation of core ACNase above 10 "C resulted in fast decay of the activity, in a lower extent of tRNALys cleavage and degradation of the reaction product (Fig. 5A). To determine the stability of the core enzyme it was preincubated at 30 "C in the absence of the radioactive substrate, which was added at the onset of the assay at 10 "C. The preincubation caused rapid decay of core ACNase, with tnh below 1 min (Fig.  5B). Core ACNase Is Not Stimulated by Activators of Latent ACNase-ACNase activity in extracts of prr+ cells is increased manyfold by adding a synthetic Stp polypeptide, ATP, and GTP, while removal of the endogenous DNA abolishes the activity (11). In contrast, the requirements of the prrCencoded core ACNase were simple. Optimal core activity was obtained in a buffer-salt solution at pH 8.0 and 5-10 mM M e . However, about one-quarter of the activity was retained without added M e and with 5 mM EDTA (data not shown). Core ACNase was not stimulated by addition of 2 mM ATP, 0.1 mM GTP, both nucleotides, 33 PM synthetic Stp, or by a combination of all three components (Table 11). In contrast, omission of any of these components from the latent enzyme assay mixture drastically reduces ACNase activity (11). Moreover, digestion of endogenous DNA with DNase I did not reduce core ACNase activity (Table 11), whereas the same treatment abolishes activation of the latent enzyme (11).
Effect of prrC Gene Dosage on Core ACNase Activity-The prrC ORF is preceded by a sequence resembling the E. coli d o consensus promoter (9). When prrC was placed in plasmids under control of this sequence and expressed over the prr' background, ACNase activity was seen only at medium to highprrC gene dosage. Three clones containing the same prrC insert embedded in different vectors were compared. pGBCl was derived from the low copy plasmid pGB2 (3 copies/cell chromosome) (15), pACCl from pACYC184 (18 copies) (14), and pRRlO from pBluescript KS' (-200 copies) (22). Only pRRlO and pACC1 elicited detectable ACNase activity (Fig.  6A, compare lanes 1 with 2 and 3 ) . When pRRlO was moved to a pcnB-host background that reduces the copy number as much as 15-fold (23), core ACNase activity was abolished (Fig. 6B, compare lanes 4 and 5). However, pRRlO rescued from the pcnBcells expressed ACNase activity when moved back to a pcnB+ background (not shown). If the activity of the indigenous prrC promoter was not altered by vector sequences or by the pcnB lesion, a threshold gene dosage, less than 18 copies/cell chromosome, may be required for detecting ACNase activity when the prrC ORF is expressed from its own promoter and in the absence of other prr genes. Yet, a single copy of prrC of the chromosomal prr locus suffices to deplete the cellular pool of tRNALy" upon phage T4 infection (8). This discrepancy could be attributed to a positive effect exerted by other prr genes over the expression of prrC or stability of its product(s).
Reconstitution of Latent ACNase from Core and Masking Components-The effect of the other prr genes on prrC's activity was examined in a complementation experiment aiming to reconstitute latent ACNase. The antagonistic components of the latent enzyme were provided by different replicons. PrrC was encoded by the low copy plasmid pGBC1, which did not elicit detectable core ACNase activity, neither before nor after T 4 infection (Fig. 6C, lanes 6-8). The masking elements were contributed by pW16prrC:tet, a derivative of the hnR,prr cosmid pW16 (9). This construct was deficient in prrC due to replacement of an internal BclI fragment with a tetracycline resistance cassette.
As shown in Fig. 7, the original cosmid pW16 expressed ACNase activity after T4 infection (lanes [1][2][3]. Cells containing pW16prrC:tet lacked it (lanes 4-6). The combination of pGBCl and pW16prrC:tet restored a weak but significant level of latent ACNase (lanes  7-9). Namely, tRNALy8 fragments appeared in the double transformants during infection later than fragments of tRNAb"' that characterize T4-infected E. coli, irrespective of The reconstitution of latent ACNase was due to complementation in trans rather than restoration of the cosmid's prrC gene by homologous recombination. This conclusion is based on the (i) unaltered restriction pattern of the cosmid and plasmid rescued from the double transformants, (ii) similar level of latent ACNase activity seen with six independent double transformants, and (iii) absence of izanR,tetS clones among 120 progeny cosmids rescued from three independent double transformants. The reconstitution of latent ACNase by pGBCllpW16prrC:tet complementation is consistent with a positive effect of the masking genes over prrC's activity, either by augmenting prrC expression or by stabilizing its product(s) or both.
PrrC Interacts with Hsd Proteins-Latent ACNase can be precipitated by anti-EcoR124/3 antibodies that cross-react with the homologous masking components of Prr (11). However, the presence of PrrC in this complex has not been investigated. To address this issue we prepared PrrC-specific antibodies. The antigen employed for this purpose was a glutathione S-transferase-PrrC fusion protein (see "Experimental Procedures"). However, only trace amounts of the full-sized 71-kDa fusion protein could be recovered (Fig. 8,  lane 1 ), probably due to instability of the PrrC portion of the fusion protein. An abundant, partial degradation product estimated to contain the N-terminal 30-40 residues of PrrC (GST-PrrCN) was used instead.
Immunoprecipitation followed by Western analysis, both performed with GST-PrrCN rabbit antiserum, revealed PrrC as a specific band in the thermo-induced, pRRC6-containing cells expressing the core ACNase (Fig. 9A, lane 2, the upper band in the faster migrating doublet). This band was not seen with an extract of control cells containing the vector plasmid pT7-5 (lane 1 ) or when the extract of the pRRC6-containing cells was treated with preimmune serum (lane 3). This confirmed the existence of the PrrC polypeptide in the core prr (5).  When immunoprecipitations were conducted with the EcoR124/3-specific antiserum and followed by Western analysis with PrrC-specific antiserum, a band corresponding to PrrC was detected in cells expressing the ACNase holoenzyme (Fig. 9A, compare lanes 7 and 8). Controlprr' cells (lane 9) and cells expressing the prrC-encoded core enzyme (lane 10) did not feature this band. Thus, PrrC was precipitated indirectly by the anti-Hsd antibodies, probably by virtue of a physical interaction with the Hsd components of ACNase holoenzyme. Conversely, immunoprecipitation of the holoenzyme with PrrC-specific antiserum followed by Western analysis with EcoR124/3-specific serum detected the masking factor PrrD as a faint but significant band over the background (Fig. 9B, compare lanes 11 and 12). PrrD was lighted up more intensely when the anti-EcoR124/3 antibodies were used both for immunoprecipitation and immunoblotting. Under these conditions a PrrA band was detected as well (compare lanes 14 and 15). These bands were not seen when the extract of holoenzyme-containing cells was treated with preimmune serum (lane 13). The absence of a PrrB band is attributed to the weak homology of PrrB and the EcoR1241 3-HsdS counterpart (10) and to the low immunogenicity of the latter (24).

DISCUSSION
Is PrrC the Core Anticodon Nuclease Enzyme?-Mutational analysis implicatedprrC with core ACNase activity (9). This assumption was confirmed by the coincident appearance of the PrrC polypeptide and core ACNase activity upon transcriptional induction of prrC (Figs. 2, 3, and 9). Catalysis of the ACNase reaction by the prrC transcript itself is improbable because small C-terminal or internal deletions in the ORF inactivate core ACNase (9): We conclude that the 45-kDa PrrC polypeptide or a derivative is needed for core ACNase activity. However, whether PrrC suffices for core ACNase activity or constitutes only a subunit or auxiliary factor is not known. The pursuit of an answer to this question was impeded by the unusual lability of core ACNase ( 1 and 9 ) , K38pGPl-2pRRC6 (lanes 2,3, and lo), or S-30 fractions of strains C600 (lanes 4, 7, 11,  and 1 4 ) and C6OO:pW16 (lanes 5, 6, 8, 12, 13, and 1 5 ) were treated with PrrC-specific serum (lanes I, 2, 4, 5, 11, and 121, EcoR124/3antiserum (lanes 7-10.14, and 15), or preimmune serum (lanes 3,6, and 13). The immunoprecipitates were separated on SDS-polyacrylamide gel electrophoresis, transferred to nitrocellulose membrane, and analyzed by Western blotting, using as probes the PrrC-specific antiserum (panel A ) or EcoR124/3-specific antiserum (panel B). and 8). Nonetheless, several facts hint that PrrC itself may PrrA-D? Still, PrrC could act upon a common (non-prr) E. be the core enzyme. Latent ACNase can be immunoprecipi-coli gene product that is needed for ACNase activity. Against tated by antibodies directed against the Hsd-masking corn-this possibility argues the ability to elicit apparent ACNase ponents and then be activated by of a prrO extract activity in human HeLa cells by infection with a recombinant (11), suggesting that all the prr-specific components needed Prrc-vaccinia Thus, PrrC may suffice to elicit ACNase for core activity reside in the immunoprecipitated complex. activity in the heterologous environment. That PrrC is included in this complex was demonstrated Separation of Anticodon Nuclease Activation and tRNALy8 immunochemically, using Hsd-and PrrC-specific antibodies activity with the latent holoenzyme, including a synthetic Cleavage-A number of cofactors are needed to elicit ACNase ( Fig. 9). The probability that another core ACNase component exists in the complex, beside PrrC, is deemed low, because immunoprecipitations of the ACNase holoenzyme unpublished results. Stp-like polypeptide, ATP, GTP, and endogenous DNA (11, 13). Because these agents did not stimulate the prrC-coded core ACNase (Table 11), they probably alleviate the masking effect. ATP and DNA, known Hsd-ligands (25), could alter the conformation of Hsd subunits of the latent enzyme (11) and thus unleash the core. The role of GTP is hinted at by a GTP binding motif, albeit imperfect, present in the deduced sequence of PrrC (Fig. 10). Such a motif is found in a diverse array of proteins that are involved in signal transduction and translation, using GTP in conjunction with protein-protein dissociation-association cycles (26). Because GTP has no effect on core ACNase (Table II), the GTP binding motif of PrrC could portend roles in ACNase latency and activation, interfacing the Hsd components of the holoenzyme. By default, a carboxyl domain of PrrC could harbor the tRNALYE recognition and cleavage activities. These possibilities can be addressed by mutational analysis aiming to associate PrrC sequences with ACNase functions.

J.
Significance of PrrC's Instability-Core ACNase activity decayed at 30 "C with t% less than 1 min (Fig. 5). The PrrC polypeptide seemed likewise unstable because it was preferentially degraded in a fusion protein (Fig. 8). In contrast, the ACNase holoenzyme is relatively stable both in uitro' and in vivo. The in vivo stability is indicated by delayed early schedule of the induction of ACNase activity during T4 infection and its persistence throughout infection ( 5 , 7). Yet, translation of host mRNAs ceases abruptly at the onset of infection (27). Hence, a preexising ACNase must sustain this time interval. Further credence to a stabilizing effect the Hsd proteins exert over PrrC was lent by reconstitution of latent ACNase from its separate components. The low copy prrC plasmid used in this experiment did not elicit detectable core activity, neither before nor after T4 infection (Fig. 6). Nevertheless, it furnished sufficient PrrC to establish latent AC-Nase when Hsd components were provided in trans (Fig. 7).
A simple explanation of this result is that PrrC is labile in the absence of the cognate masking proteins, which normally interact with it in latent ACNase. In fact, this interaction is indicated by the immunochemical analyses (Fig. 9). An alternative explanation, that prrA,B&D stimulate prrC transcription and/or translation, is refuted by a high level of prrC mRNA expressed from the low copy plasmid pGBC1, relative to prrC-mRNA transcribed from the chromosomal or cosmidborne prr locus. Furthermore, other prr genes fail to augment the reporter activity of a prrC-lacZ gene f u~i o n .~ Stabilization of PrrC by the masking elements could qualify them as effector subunits of ACNase if the PrrC-Hsd interaction sustains the activation, an issue not addressed yet. Although the molecular basis of PrrC's instability is not known, it is conceivable that this property evolved to safeguard ACNase latency. That is, free PrrC may be promptly removed to ward off inadvertent toxicity. However, shielded by the Hsd subunits, it may be kept as an antiviral contingency.