Cloning and DNA Sequence Analysis of Two Abortive Infection Phage Resistance Determinants from the Lactococcal Plasmid pnp40

Transcription

1 APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Dec. 1995, p Vol. 61, No /95/$ Copyright 1995, American Society for Microbiology Cloning and DNA Sequence Analysis of Two Abortive Infection Phage Resistance Determinants from the Lactococcal Plasmid pnp40 P. GARVEY, 1 G. F. FITZGERALD, 1,2 * AND C. HILL 1,2 Department of Microbiology 1 and National Food Biotechnology Centre, 2 University College, Cork, Ireland Received 1 May 1995/Accepted 2 October 1995 The lactococcal plasmid pnp40, from Lactococcus lactis subsp. lactis biovar diacetylactis DRC3, confers complete resistance to the prolate-headed phage c2 and the small isometric-headed phage 712 in L. lactis subsp. lactis MG1614. A 6.0-kb NcoI fragment of pnp40 cloned in the lactococcal Escherichia coli shuttle vector pam401 was found to confer partial resistance to 712. Subcloning and deletion analysis of the recombinant plasmid ppg01 defined a 2.5-kb ScaIHpaI fragment as conferring phage insensitivity. Sequence analysis of this region confirmed the presence of two overlapping open reading frames (ORFs). Further subcloning of pnp40 to characterize the resistance determinant active against c2 identified a 5.6-kb EcoRV fragment of pnp40 which, when cloned in pam401, conferred partial resistance to both c2 and 712. Subcloning and deletion analysis of the recombinant plasmid pcg1 defined a 3.7-kb EcoRV-XbaI fragment as encoding phage insensitivity. DNA sequence analysis of this region revealed the presence of a single complete ORF. The introduction of a frameshift mutation at the unique BglII site within this ORF disrupted the phage resistance phenotype, confirming that this ORF is responsible for the observed phage insensitivity. The mechanisms encoded by ppg01 and pcg1 in L. lactis subsp. lactis MG1614 conformed to the criteria defining abortive infection and were designated AbiE and AbiF, respectively. Analysis of the phage DNA content of 712-infected hosts containing AbiF demonstrated that it inhibited the rate of phage DNA replication, while AbiE had little effect on phage DNA replication, suggesting a later target of inhibition. The predicted protein product of abif shows significant homology to the products of two other lactococcal abortive infection genes, abid and abid1. * Corresponding author. Mailing address: Dept. of Microbiology, University College, Cork, Ireland. Phone: Fax: Bacteriophage interference with the starter culture(s) used in dairy fermentations can cause slow acid production, which can result in an inferior product or, in more extreme cases, complete starter failure. The introduction of selected phage resistance plasmids by conjugation into commercial starter strains is one approach that has proven to be successful in minimizing disruption by phage (33). The prolonged use of specific transconjugants in industrial fermentations has, however, promoted selection of a number of phages which are insensitive to the plasmid-encoded mechanisms (22, 31). The rotation of starters containing alternative phages resistance plasmids or the use of strains containing multiple complementary resistance mechanisms can alleviate the pressure on individual plasmids. This necessitates examination of the underlying mechanisms which contribute to the phage insensitivity encoded by these plasmids, which will, in turn, allow the development of knowledge-based strategies for the combination of multiple defenses in starter strains. This study was undertaken to investigate the phage resistance mechanisms encoded by the 65-kb lactococcal plasmid pnp40, originally identified by McKay and Baldwin (28) in Lactococcus lactis subsp. lactis biovar diacetylactis DRC3, and to analyze, at the DNA sequence level, the gene(s) encoding insensitivity. This conjugative plasmid has been exploited to improve the phage resistance properties of L. lactis subsp. lactis biovar diacetylactis 425A, a strain used in lactic butter fermentation (18). The introduction of pnp40 into the plasmid-free strain L. lactis subsp. lactis MG1614 conferred an impressive level of resistance to phage, with no plaques obtained on a lawn of cells with either prolate- or small isometricheaded types (28). The cloning and sequence analysis of the determinants of two independent Abi mechanisms are described, and the stage in the lytic cycle at which these systems operate is investigated. MATERIALS AND METHODS Bacterial strains, plasmids, and culture conditions. Strains, plasmids, and phages used in this study are listed in Table 1. Lactococcal cultures were grown at 30 C in M17 medium supplemented with 0.5% glucose when necessary (35). Escherichia coli cultures were propagated in Luria-Bertani broth and incubated at 37 C (32). M13 phage was propagated as outlined by Sambrook et al. (32). pam401 (36) and various derivatives were maintained in lactococci with chloramphenicol at 10 g/ml and in E. coli with tetracycline at 10 g/ml and/or chloramphenicol at 20 g/ml. Plasmid and phage DNA preparation. The lysis procedure of Anderson and McKay (2) was used to isolate plasmid DNA from lactococcal strains. E. coli plasmid DNA was obtained according to the method of Birnboim and Doly (4), and large volumes were purified by cesium chloride-ethidium bromide density gradient ultracentrifugation in a Beckman VTi65 rotor. Lactococcal phage DNA was isolated by the method of Fitzgerald et al. (12) with the modifications described by Coveney et al. (10). Restriction endonucleases and molecular cloning techniques. Restriction enzymes, the Klenow fragment of DNA polymerase I, and T4 DNA ligase were obtained from Boehringer Corp. (Dublin, Ireland). DNA digestions and ligations were performed as outlined by Sambrook et al. (32). DNA fragments were isolated from agarose gels with the GeneClean kit II (BIO 101, La Jolla, Calif.). Electroporation of bacteria. The electroporation of lactococcal strains was executed according to the procedure of Holo and Nes (24) with the Bio-Rad Gene Pulser apparatus (Bio-Rad Corp., Richmond, Calif.). E. coli transformations were performed under the conditions outlined in the Bio-Rad manual. Nucleotide sequence analysis. Relevant DNA fragments were cloned into the M13mp18 and M13mp19 vectors (37). The nucleotide sequences were determined with both single-stranded M13 and alkali-denatured ppg01 and pcg1 templates and the Sequenase 2.0 kit (U.S. Biochemical, Cleveland, Ohio) or the Taq Track sequencing system (Promega Corp., Madison, Wis.). Sequencing was initiated with commercial M13 primers and continued with specific synthetic 17-mer primers prepared with a DNA synthesizer (PCR-MATE; Applied Biosystems, Foster City, Calif.). Each strand of DNA was sequenced at least once. 4321

2 4322 GARVEY ET AL. APPL. ENVIRON. MICROBIOL. TABLE 1. Bacterial strains, plasmids, and phages used in this study Strain, plasmid, or phage Relevant characteristic Reference or source Strains MG1614 Plasmid-free derivative of L. lactis subsp. lactis MG1614/pNP40 MG1614 transconjugant containing pnp40 DPC collection a PG001 MG1614 transformant containing ppg01 This study PG020 MG1614 transformant containing pcg1 This study Plasmids pam401 Lactococcal E. coli shuttle vector 36 pnp40 65-kb plasmid from L. lactis subsp. lactis biovar diacetylactis DRC3 28 ppg kb NcoI fragment from pnp40 cloned into pam401 This study ppg03 ppg01 with 2.5-kb EcoRI fragment deleted This study ppg05 ppg01 with two small ScaI fragments deleted This study ppg kb XbaI fragment from ppg01 cloned into pam401 This study ppg kb EcoRV fragment from ppg01 cloned into pam401 This study ppg09 ppg05 with NcoI-HpaI fragment deleted This study pcg1 5.6-kb EcoRV fragment from pnp40 cloned into pam401 This study ppg21 pcg1 with BamHI-BglII fragment deleted This study ppg22 pcg1 with two PstI fragments deleted This study ppg kb EcoRV-XbaI fragment from pcg1 cloned into pam401 This study ppg kb XbaI fragment from pcg1 cloned into pam401 This study ppg kb BamHI-BglII fragment from pcg1 cloned into pam401 This study ppg26 pcg1 containing frameshift mutation at BglII site This study Phages c2 Prolate-headed phage for MG Small-isometric-headed phage for MG1614 a DPC, Dairy Products Centre, Moorepark, Fermoy, County Cork, Ireland. Sequencing gels were run as outlined by Bio-Rad. Sequence data were analyzed by the Gene Jockey and DNAstar database software programs (Apple Computers, Inc., Cupertino, Calif.). Phage assays. The efficiency of center of infection (ECOI) was measured by the method of Sing and Klaenhammer (34). Cells were infected with phage at a multiplicity of infection of 0.1 and incubated for 10 min to permit phage to adsorb to hosts. The ECOI was calculated as (PFU of the infected resistant host) (PFU of the infected sensitive host) 1. Cell survival was assayed as described by Behnke and Malke (3). Surviving cells were enumerated as CFU. The percentage of cell death was calculated as 100 [(CFU per milliliter in cultures without phage) (CFU per milliliter in cultures with phage)][cfu per milliliter in cultures without phage] 1. Burst sizes were measured as described by Klaenhammer and Sanozky (25). One-milliliter samples were removed at time zero and at increasing time intervals and assayed directly for phage with plasmid-free L. lactis MG1614 as a sensitive host. Phage DNA replication. Intracellular phage DNA replication was monitored by the method of Hill et al. (20). DNA samples were digested with HindIII ( 712) or EcoRI ( c2) and electrophoresed on 0.7% agarose gels. Nucleotide sequence accession number. The nucleotide sequence data reported in this paper have been submitted to GenBank (accession no. U36837). RESULTS Cloning of two phage resistance mechanisms from pnp40. L. lactis subsp. lactis MG1614 containing pnp40 displays complete resistance to c2 and 712, as manifested by the absence of plaques when the titer of an undiluted phage stock (10 9 PFU/ml) on a lawn of cells is determined (Table 2). In order to localize the phage resistance determinant(s) within pnp40, a number of EcoRI and NcoI fragments were cloned into the lactococcal E. coli shuttle vector pam401 (36). Following the introduction of the various recombinant derivatives into L. lactis subsp. lactis MG1614, the clones were examined for their abilities to confer resistance to phage. One clone designated PG001 harbored ppg01 (Fig. 1A and Table 1), a recombinant plasmid containing a 6.0-kb NcoI fragment from pnp40, which conferred partial resistance to 712 (Table 2). No resistance was evident against c2. A detailed restriction map of ppg01 was generated, and subsequent deletion and subcloning analyses permitted the construction of the derivatives ppg03, ppg05, ppg06, and ppg07 (Fig. 1A and Table 1). An examination of the effect on the plaquing efficiency of 712, when these plasmids were introduced into L. lactis subsp. lactis MG1614, allowed the phage resistance locus to be defined within a 3.1-kb ScaI-NcoI region on ppg05 (Fig. 1A). Further deletion analysis of ppg05, resulting in the plasmid ppg09, identified a 2.5-kb ScaI-HpaI fragment as the smallest region capable of encoding insensitivity. The ability of pnp40 to mediate resistance to c2, in addition to 712, suggested the presence of a second phage resistance locus on this replicon. In order to localize this determinant, a number of EcoRV fragments of pnp40 were cloned into pam401, and the resulting recombinant plasmids, when introduced into L. lactis subsp. lactis MG1614, were screened for their abilities to confer resistance to phage. A single recombinant plasmid, pcg1, harboring a 5.6-kb insert, which conferred partial resistance to both c2 and 712 was identified in clone PG020 (Table 2). A detailed restriction map of pcg1 was generated, and subcloning and deletion analysis of pcg1 enabled the construction of several derivatives, ppg21 to ppg25 (Table 1 and Fig. 1A), which were subsequently TABLE 2. EOPs of 712 and c2 on L. lactis subsp. lactis MG1614, PG001, PG020, and MG1614/pNP40 Strain EOP a 712 Plaque size (mm) EOP c2 Plaque size (mm) MG MG1614/pNP40 NP b NA c NP NA PG PG Pinpoint a (PFU on test strain) (PFU on sensitive strain) 1. b NP, no plaques. c NA, not applicable.

3 VOL. 61, 1995 LACTOCOCCAL RESISTANCE TO PHAGE BY ABORTIVE INFECTION 4323 FIG. 1. (A) Restriction map of pnp40-derived inserts in ppg01 and pcg1. Resistance (R) or sensitivity (S) to 712 or c2 is indicated in the right-hand column. (B) Relative sizes and orientations of ORFs, inferred from sequence analysis of pnp40 insert of ppg09 and ppg23. introduced into L. lactis subsp. lactis MG1614. An assessment of their effects on the efficiencies of plaquing (EOPs) of c2 and 712 allowed the phage resistance gene to be localized to a 3.7-kb XbaI-EcoRV fragment on ppg23 (Fig. 1A). In addition, the introduction of an additional 4 bp at the unique BglII site on pcg1 by the activity of the Klenow fragment of DNA polymerase I on BglII-digested pcg1 DNA, generating the derivative ppg26, disrupted the phage resistance phenotype. Although the pnp40 DNA inserts of ppg01 and pcg1 were shown by restriction mapping and hybridization to be overlapping, subcloning and phenotypic analysis demonstrated that they encoded two independent phage resistance systems (Fig. 1A). Localization of phage resistance determinants on pnp40. A restriction map of pnp40 was generated with a number of restriction enzymes chosen for their ability to restrict pnp40 infrequently (Fig. 2). Hybridization analyses, with the pnp40 fragments from ppg01 and ppg23 as probes, localized the determinants for both phage resistance mechanisms close to one another on pnp40 (Fig. 2). Additional hybridization analyses with ISS1, IS981, and IS904 DNA as probes revealed the presence of two copies of ISS1 and a single copy of IS981 on pnp40. No homology to IS904 was detected. The approximate locations of these insertion sequence elements and of the fragment identified by Froseth et al. (14) as harboring the nisin resistance gene and the plasmid origin are indicated in Fig. 2. Insensitivity encoded by ppg01, pcg1, and pnp40. A number of assays were performed to assess the nature of the resistances encoded by ppg01 and pcg1. Plaque assays suggested that the pcg1-mediated resistance corresponded to a classic abortive infection phenotype: both 712 and c2 formed reduced-size plaques on L. lactis subsp. lactis PG020 at a lower frequency than they did on a sensitive host (Table 2). In the case of ppg01, plaque assay results indicated that, although the EOP of 712 was reduced relative to the EOP on the plasmid-free L. lactis subsp. lactis MG1614 (Table 2), no accompanying reduction in plaque size was observed. As this phenotype could reflect the operation of either restriction modification (R/M) or an Abi system, host-dependent plaquing efficiency was assayed. Phage from plaques obtained on PG001 was found to plaque more efficiently on PG001 in a second round of plaquing (unpublished data). However, this increased level of efficiency was not reversible by passage through MG1614, indicating that the ability of these phage to plaque on PG001 is host independent. ppg01 does not there- FIG. 2. Restriction map of pnp40 indicating the locations of the determinants for AbiE and AbiF. The approximate locations of the two copies of ISS1, the copy of IS981, the nisin resistance determinant, and the origin of replication identified by Froseth and McKay (15) are also indicated.

4 4324 GARVEY ET AL. APPL. ENVIRON. MICROBIOL. TABLE 3. Percentages of cell death, ECOI values, and burst sizes of 712 on L. lactis subsp. lactis MG1614, PG001, PG020, and MG1614/pNP40 Strain Resistance % Cell death ECOI Burst size MG1614 None PG001 AbiE PG020 AbiF MG1614/pNP40 Total NP a a NP, not possible to evaluate. fore encode an R/M system, and it probably encodes an Abi system. These phage from normal-sized plaques on PG001 presumably represent mutants which are not affected by the ppg01-encoded mechanism. As previously stated, no plaques were obtained on cells containing pnp40. To confirm that ppg01 and pcg1 encoded Abi mechanisms, a number of further assays were performed. For the purpose of this study, 712 was used as a model system since both systems conferred resistance to this phage. An examination of the number of cells which die as a result of infection is a valuable indicator of whether the resistance mechanism acts early, e.g., producing adsorption inhibition or R/M, or late in the lytic cycle, i.e., producing abortive infection. Cell death results indicated that 85% of the ppg01-containing cells and 90% of the pcg1-containing cells died as a result of 712 infection, confirming that both mechanisms operate by abortive infection. When pnp40 was present in the host, the level of cell death remained relatively high at 42% (Table 3). ECOI assays, which indicate the number of infected cells which give rise to productive infections, were carried out on sensitive and resistant hosts. By convention, it is assumed that 100% of the infected sensitive cells result in progeny phage (ECOI 1.0). Only 15% of the cells containing ppg01 and 80% of the cells containing pcg1 resulted in productive 712 infections. This was reduced to 3% for cells containing pnp40 (Table 3). The burst sizes for 712 propagated on strains PG001 and PG020 were considerably reduced, with only 7 and 6 viable progeny per cell, respectively, compared with 54 per cell following infection of the plasmid-free host. Burst size could not be evaluated in cells containing pnp40 because of the low ECOI value (Table 3). In addition to the high-level cell death results, which are characteristic of Abi, reduced ECOI and burst size results are also typical of abortive infection phenotypes. The mechanisms encoded by ppg01 and pcg1 were designated AbiE and AbiF, respectively, in keeping with the nomenclature suggested by Coffey et al. (8). Effects of AbiE and AbiF on 712 DNA replication. The replication of phage DNA in sensitive and resistant hosts was examined at 15-min intervals following 712 infection (Fig. 3). Normal phage development was monitored in plasmid-free hosts, with the infecting phage genomes detected within the cell after 15 min. The intracellular phage DNA concentration increased over time until 60 min had elapsed, at which time a considerable reduction was observed. This reduction presumably reflected phage DNA packaging and the release of progeny. In the cells expressing AbiE, the DNA of the infecting particles entered the cells in the normal manner during the first 15 min and the intracellular phage DNA concentration increased over time, reflecting phage genome replication. However, there was no evidence of DNA packaging or phage release, as was discerned in the sensitive host after 60 min, suggesting that AbiE operates late in the lytic cycle. When AbiF was present in the host, phage DNA entered the cells normally but replication was significantly retarded relative to that of the plasmid-free host, suggesting that AbiF acts at the level of phage DNA replication. In MG1614/pNP40, phage DNA replication was completely retarded, with no increase in phage DNA concentration despite normal DNA internalization as visualized by DNA hybridization (data not shown). DNA and deduced amino acid sequence analysis. M13 clones were generated in both orientations, encompassing the entire pnp40-derived DNA inserts of ppg09 and ppg23. Where overlapping clones were not obtained, ppg01 and pcg1 DNAs were used as sequencing templates. Sequencing was initiated with commercial M13 primers and was continued with synthetic 17-mer oligonucleotides. Computer analysis of the sequence data suggested the presence of two open reading frames (ORFs) of 861 and 894 bp on ppg09, which overlap by 1 bp (Fig. 1B and 4). The smallest ppg01 derivative which confers insensitivity to phage contains both ORF1 and ORF2 (i.e., ppg09; Fig. 1A). ppg07, which contains all of ORF1 but is missing 69 bp at the C terminus of ORF2, does not confer resistance to 712, and ppg06, which encodes all but the C-terminal amino acids of ORF2 but lacks the promoter sequences preceding ORF1 and the N-terminal 10 amino acids of ORF1, is also sensitive to 712. This implies that either both ORFs are required for the expression of the resistance phenotype or ORF2 alone is sufficient but requires DNA sequences upstream of ORF1 for expression. FIG. 3. DNA content of L. lactis subsp. lactis MG1614, PG001, PG020, and MG1614/pNP40 hosts following infection with 712. Lanes 1 to 6 show the HindIII-digested total DNA isolated at 0, 15, 30, 45, 60, and 75 min, respectively, from 712-infected MG1614, PG001, PG020, and MG1614/pNP40 (A, B, C, and D, respectively).

5 VOL. 61, 1995 LACTOCOCCAL RESISTANCE TO PHAGE BY ABORTIVE INFECTION 4325 Downloaded from FIG. 4. DNA sequence of the 2.5-kb insert of ppg09. The deduced amino acid sequences of abiei and abieii are indicated above the DNA sequence with the three-letter code designations. The putative promoter sequences and RBSs are underlined, and the putative terminator is indicated with arrows. These ORFs have the capacity to encode proteins of 287 and 298 amino acids with predicted molecular masses of 33.8 and 35.4 kda, respectively. Each is preceded by a putative ribosome binding site (RBS) (AAAGGAG and GGAG, respectively) with G values of 14.0 and 9.4 kcal (ca and 39.3 kj)/mol, respectively. A consensus 10 sequence (TATAAT) is present 38 bp upstream of the RBS of ORF1 and is separated by 17 bp from a 35 sequence (TTGTGT) which partially resembles the 35 sequence of the E. coli consensus promoter (TTGACA). No consensus 10 or 35 region was identified upstream of ORF2, suggesting that both ORFs may be transcribed as a polycistronic operon with the promoter sequences preceding ORF1. The nearest structure FIG. 5. DNA sequence of the 3.7-kb insert of ppg23. The deduced amino acid sequence of AbiF is indicated above the DNA sequence with the three-letter code designations. The putative promoter and RBS are underlined. The 23-bp sequence, which is repeated three and one-half times (DR1 to DR4), a 7-bp inverted repeat downstream, and a putative transcriptional terminator are indicated with arrows. which could function as a transcriptional terminator is a 14-bp inverted repeat sequence ( G 25.2 kcal [ca kj]/ mol) which is located 135 bp downstream. These ORFs were designated abiei and abieii. Computer analysis of the sequence data for ppg23 suggested the presence of a single complete ORF which crosses the BglII site (Fig. 1B and 5). This ORF, designated abif, is 1,026 bp in length and has the capacity to encode a protein of 342 amino acids with a predicted molecular mass of 41.2 kda. Six base pairs upstream of the ATG start codon, the ORF is preceded by a putative RBS (AAAGG) with a G value of 9.6 kcal (ca kj)/mol. A 10 sequence (TTTAAT) that is separated by 17 bp from a 35 sequence (TTGAAT) was identified 85 bp upstream of the RBS; these sequences partially resemble the 10 and 35 sequences (TATAAT and TTGACA, respectively) of the E. coli promoter. A TG dinu- on June 19, 2018 by guest

6 4326 GARVEY ET AL. APPL. ENVIRON. MICROBIOL. cleotide found in many lactococcal promoters is located 1 bp upstream from the 10 sequence, and an 8-bp inverted repeat, which has a G value of 13.4 kcal (ca kj)/mol, 2 bp downstream of the ORF has the potential to form a stem-loop structure and, thus, could serve as a transcriptional terminator. The GC content of both abiei and abieii is 29%, and the GC content of abif is 26%, values which are exceptionally low compared with the 37% average for lactococcal DNA; however, low GC content appears to be a feature of lactococcal abi genes (16). Comparative analysis with known sequences in the EMBL and GenBank databases (release 87), with the DNAstar software program, revealed no DNA sequence homology to any known sequence. In addition, comparative analysis of the deduced amino acid sequences with known sequences in the Swiss and PIR and Translated databases (release 87) revealed no significant amino acid sequence homology. A search of all three deduced amino acid sequences for functional motifs from the Prosite database (release 11) with the MacPattern and DNAstar (DNAstar Inc., Madison, Wis.) programs showed no evidence of homology, offering no indication of possible modes of action for these mechanisms. However, the abif gene product showed significant homology with the predicted products of two recently sequenced Abi systems, AbiD (29) and AbiD1 (1). The predicted product of the abif gene shows a high level of relatedness to that encoded by abid1 (47% identity, along the length of the protein) and is also somewhat related to the recently identified abid gene product (26% identity). This is similar to the degree of relatedness between abid and abid1 (28% identity) (1). An alignment of the three related gene products reveals a number of highly conserved regions, with 18% of the residues being common to all three proteins (Fig. 6). Interestingly, downstream of abif on ppg23, an imperfect 23-bp sequence is directly repeated three and one-half times (Fig. 5). This motif bears a strong resemblance to a feature found in the replication origin of many theta-replicating plasmids. All pci305-type plasmids, for example, contain a conserved 22-bp sequence repeated three and one-half times in the noncoding region of their minimal replicon (19). Hence this region could potentially encode a replication origin for pnp40. The 23-bp sequence of pcg1 is not homologous to the 22-bp direct repeat of the pci305 family. It is worth noting that this putative origin is not located on the 7.6-kb EcoRI pnp40 fragment identified by Froseth et al. (14) as harboring a plasmid origin. FIG. 6. Alignment of the amino acid sequences of AbiD, AbiD1, and AbiF. DISCUSSION Bacteriophage resistance mediated by abortive infection is common in Lactococcus spp. This communication describes the cloning of two abortive infection mechanisms from the lactococcal plasmid pnp40: AbiE, which confers resistance to 712 by interfering in phage development late in the lytic cycle, and AbiF, which affects the level of phage DNA replication. The AbiE system is unusual in that two ORFs are apparently required for the expression of the resistance phenotype. The phage resistance phenotype mediated by ppg01 conforms to the definition of abortive infection. Phage adsorbed to the host normally (data not shown), their DNA was injected, and the host cells were killed. Phage DNA replication occurred, but little evidence of packaging of phage DNA or phage release was observed. In fact, 85% of the infected cells did not produce progeny phage and the burst size in the remainder was reduced eightfold. The majority of the 712 phage did not form plaques on hosts expressing AbiE; however, mutant phage were detected at a frequency of 10 4 and formed normal-sized plaques. These phage plaqued with equal levels of efficiency on MG1614 and PG001, regardless of the host used for propagation. These observations suggest a late target of inhibition for AbiE, as there is little if any effect on phage DNA replication. Possible mechanisms include interference in transcription and translation processes, as in the case of AbiC (30), defects in packaging, or release of phage particles. In gram-negative bacteria, deficiencies in late protein synthesis have been observed with a number of Abi systems. The Pif system encoded by the F factor in E. coli induces ribosomal misreading, which results in the suppression of ochre stop codons (26). This gives rise to the inaccurate translation of late phage genes and aborts the infection. In Vibrio cholerae biotype El Tor hosts, deficiencies in transcription and translation were shown to be due to the destabilization of phage DNA concatemeric replication intermediates. The association of phage DNA with the cell membrane is a requirement for late protein synthesis in permissive Vibrio hosts, and this association is hindered in El Tor cells by a small hydrophobic protein which inserts itself into the cell membrane (6). The effect, if any, of AbiE on phage protein synthesis remains to be established. The resistance encoded by pcg1 also bears the phenotypic hallmarks of an abortive infection-type mechanism. The EOPs and plaque sizes were reduced for both prolate- and isometric-

7 VOL. 61, 1995 LACTOCOCCAL RESISTANCE TO PHAGE BY ABORTIVE INFECTION 4327 headed phages on strains containing AbiF. Phage adsorbed normally to PG020 (data not shown) and their DNA was internalized, but phage DNA replication was significantly retarded. This seemingly modest effect of lowering the number of phage genomes generated within the cell can result in a reduced production of phage gene products (30). In this case, the outcome was a considerable reduction in the number of phage progeny produced by infected PG020 cells and even the failure of some of the cells (20 to 23%) to give rise to any progeny. AbiF has a broader spectrum of activity than AbiE in that it functions against the prolate-headed phage c2 and the smallisometric-headed phage 712. It is likely that AbiF acts by a general antiphage mechanism, with AbiE having a more specific target of action. It also differs from AbiE in that it acts at the level of phage DNA replication. The parental plasmid pnp40 confers complete resistance to 712, with no plaques detected even when phage titers of 10 9 PFU were employed. Phage 712 adsorbed to MG1614/ pnp40, and phage DNA entry into the cell was normal. The presence of pnp40, however, completely inhibited phage DNA replication, reducing the ECOI value and burst size to below the ECOI value and burst size obtained for either Abi mechanism alone. This enhanced level of resistance may reflect the combined effect of both AbiE and AbiF when resident on the parental plasmid, but it is also possible that pnp40 encodes additional resistance mechanisms. pnp40 is the only lactococcal phage resistance plasmid isolated to date on which two distinct Abi mechanisms have been identified. In fact, the determinants for AbiE and AbiF lie in close proximity to one another on pnp40. Analyses of many other abortive infection plasmids have shown that they can encode additional mechanisms. ptr2030 (23), ptn20 (11), pkr223 (27), and pbf61 (13) were all found to encode R/M in addition to Abi. In the case of pci528, Abi was found in association with adsorption inhibition (9). These examples of stacking of complementary phage resistances demonstrate the effectiveness of this strategy of phage defense. DNA sequence analysis of abortive infection determinants has, in general, provided little information regarding the nature of the resistance. However, there are a number of interesting features concerning the analysis of these genes. First, the genetic organization of the determinants for AbiE is striking in that it appears that two overlapping ORFs may encode the abortive infection phenotype. Overlapping genes are not unusual in lactococci, in which they have been found extensively among ORFs in amino acid biosynthesis operons, among other operons (5). The overlapping is a unique arrangement, however, among the lactococcal abortive infection genes sequenced to date, all of which involve a single ORF. The overlapping of genes may induce translational coupling, which is believed to direct the stoichiometric synthesis of proteins by making translation of the second gene dependent on that of the first, thus ensuring a proportional production of proteins. The absence of an identifiable promoter for abieii substantiates the possibility of cotranscription for these ORFs. Second, while the sequence of the abif gene is novel, it is obvious that the predicted protein is related to those of AbiD and AbiD1. It is notable that all three systems confer broadly similar degrees of resistance, and we can conclude that AbiF, AbiD, and AbiD1 form a family of abortive infection genes. Another striking feature of the pnp40 Abi systems is the low GC content of abiei, abieii, and abif. This is also true of abia (21), abib (7), abic (11), abid (29), and abid1 (1), all of which have GC contents of 26 to 29%. It is not known if these atypical values reflect the function performed by abi genes or if it reflects their origin. Downstream of abif, a 23-bp sequence repeated three and one-half times was identified. This interesting motif bears a strong resemblance to the genetic organization of the plasmid origin of many theta-replicating plasmids. The lactococcal pci305-type replicons, for example, contain a conserved 22-bp repeat repeated three and one-half times (19). While the 23-bp direct repeat exhibited no sequence homology to this 22-bp sequence, this region could nevertheless potentially be a replication origin for pnp40. Intriguingly, this region does not correspond to the 7.6-kb EcoRI fragment of pnp40 identified by Froseth et al. (14) as encoding the pnp40 origin of replication which was exploited by Froseth and McKay (15) in the construction of a food-grade vector for Lactococcus spp. This observation, in addition to the fact that pnp40 encodes two copies of ISS1, suggests that the plasmid may have originated as a consequence of a cointegration event, possibly during conjugative transfer, as pnp40 is also self-transmissible. The cloning of overlapping fragments from pnp40 into a replication probe vector should establish if this region encodes an operational plasmid origin. This study has identified two distinct Abi systems, the determinants for which have been cloned and sequenced. Although the molecular basis of the Abi-phage interactions has not yet been established, AbiE and AbiF represent further examples of phage resistances which may be used alone or in combination with other phage resistance mechanisms in the construction of improved starter cultures. ACKNOWLEDGMENTS This work was supported by a European Community BRIDGE grant to G. F. Fitzgerald (contract number BIO-CT ) and by funds from Irish Dairy Farmers. We also acknowledge the contribution of Ciaran Geoghegan in the generation of the original AbiF clone, pcg1. REFERENCES 1. Anba, J., E. Bidnenko, A. Hillier, D. Ehrlich, and M.-C. Chopin Characterization of the lactococcal abid1 gene coding for phage abortive infection. J. Bacteriol. 177: Anderson, D. G., and L. L. McKay Simple and rapid method for isolating large plasmid DNA from lactic streptococci. Appl. Environ. Microbiol. 46: Behnke, D., and H. Malke Bacteriophage interference in Streptococcus pyogenes. 1. Characterisation of prophage-host systems interfering with the virulent phage A25. Virology 85: Birnboim, H. C., and J. Doly A rapid alkaline extraction procedure for screening recombinant plasmid DNA. Nucleic Acids Res. 7: Chopin, A Organisation and regulation of genes for amino acid biosynthesis in lactic acid bacteria. FEMS Microbiol. Rev. 12: Chowdury, R., S. K. Biswas, and J. Das Abortive replication of choleraphage 149 in Vibrio cholerae biotype el tor. J. Virol. 63: Cluzel, P.-J., A. Chopin, S. D. Ehrlich, and M.-C. Chopin Phage abortive infection mechanism from Lactococcus lactis subsp. lactis, expression of which is mediated by an Iso-ISS1 element. Appl. Environ. Microbiol. 57: Coffey, A. G., G. F. Fitzgerald, and C. Daly Cloning and characterisation of the determinant for abortive infection from the lactococcal plasmid pci829. J. Gen. Microbiol. 143: Costello, V Characterisation of bacteriophage-host interaction in Streptococcus cremoris UC503 and related streptococci. Ph.D. thesis. University College, Cork, Ireland. 10. Coveney, J. A., G. F. Fitzgerald, and C. Daly Detailed characterization and comparison of lactic streptococcal bacteriophages based on morphology, restriction mapping, DNA homology, and structural protein analysis. Appl. Environ. Microbiol. 53: Durmaz, E., D. L. Higgins, and T. R. Klaenhammer Molecular characterization of a second abortive phage resistance gene present in Lactococcus lactis subsp. lactis ME2. J. Bacteriol. 174: Fitzgerald, G. F., C. Daly, L. R. Brown, and T. R. Gingeras ScrFI: a new sequence specific endonuclease from Streptococcus cremoris. Nucleic Acids Res. 10: Froseth, B. R., S. K. Harlander, and L. L. McKay Plasmid-mediated reduced phage sensitivity in Streptococcus lactis KR5. J. Dairy Sci. 71:275

8 4328 GARVEY ET AL. APPL. ENVIRON. MICROBIOL Froseth, B. R., R. E. Herman, and L. L. McKay Cloning of nisin resistance determinant and replication origin on a 7.6-kilobase EcoRI fragment of pnp40 from Streptococcus lactis subsp. diacetylactis DRC3. Appl. Environ. Microbiol. 54: Froseth, B. R., and L. L. McKay Development and application of pfm011 as a possible food-grade cloning vector. J. Dairy Sci. 74: Garvey, P., D. van Sinderen, D. P. Twomey, C. Hill, and G. F. Fitzgerald. Molecular genetics of bacteriophage and natural phage defence systems in the genus Lactococcus. International Dairy Journal, in press. 17. Gasson, M. J Plasmid complements of Streptococcus lactis NCDO 712 and other lactic streptococci after protoplast-induced curing. J. Bacteriol. 154: Harrington, A., and C. Hill Construction of a bacteriophage-resistant derivative of Lactococcus lactis subsp. lactis 425A by using the conjugal plasmid pnp40. Appl. Environ. Microbiol. 57: Hayes, F., P. Vos, G. F. Fitzgerald, W. M. devos, and C. Daly Molecular organisation of the minimal replicon of novel, narrow host-range, lactococcal plasmid pci305. Plasmid 25: Hill, C., I. J. Massey, and T. R. Klaenhammer Rapid method to characterize lactococcal bacteriophage genomes. Appl. Environ. Microbiol. 57: Hill, C., L. A. Miller, and T. R. Klaenhammer Nucleotide sequence and distribution of the ptr2030 resistance determinant (hsp) which aborts bacteriophage infection in lactococci. Appl. Environ. Microbiol. 56: Hill, C., L. A. Miller, and T. R. Klaenhammer In vivo genetic exchange of a functional domain from a type II A methylase between lactococcal plasmid ptr2030 and a virulent bacteriophage. J. Bacteriol. 173: Hill, C., K. Pierce, and T. R. Klaenhammer The conjugative plasmid ptr2030 encodes two bacteriophage defence mechanisms in lactococci, restriction modification (R /M ) and abortive infection (Hsp ). Appl. Environ. Microbiol. 55: Holo, H., and I. F. Nes High-frequency transformation, by electroporation, of Lactococcus lactis subsp. cremoris grown with glycine in osmotically stabilized media. Appl. Environ. Microbiol. 55: Klaenhammer, T. R., and R. B. Sanozky Conjugal transfer from Streptococcus lactis ME2 of plasmids encoding phage resistance, nisin resistance and lactose-fermenting ability: evidence for a high-frequency conjugative plasmid responsible for abortive infection of virulent bacteriophage. J. Gen. Microbiol. 131: Kruger, D. H., and T. A. Bickle Abortive infection of Escherichia coli F cells by bacteriophage T7 requires ribosomal misreading. J. Mol. Biol. 194: Laible, N. J., P. L. Rule, S. K. Harlander, and L. L. McKay Identification and cloning of plasmid deoxyribonucleic acid coding for abortive infection from Streptococcus lactis ssp. diacetylactis KR2. J. Dairy Sci. 70: McKay, L. L., and K. A. Baldwin Conjugative 40-megadalton plasmid in Streptococcus lactis subsp. diacetylactis DRC3 is associated with resistance to nisin and bacteriophage. Appl. Environ. Microbiol. 47: McLandsborough, L. A., K. M. Kolaetis, T. Requena, and L. L. McKay Cloning and characterization of the abortive infection genetic determinant abid isolated from pbf61 of Lactococcus lactis subsp. lactis KR5. Appl. Environ. Microbiol. 61: Moineau, S., E. Durmaz, S. Pandian, and T. R. Klaenhammer Differentiation of two abortive mechanisms by using monoclonal antibodies directed towards lactococcal bacteriophage capsid proteins. Appl. Environ. Microbiol. 59: Moineau, S., S. Pandian, and T. R. Klaenhammer Evolution of a lytic bacteriophage via DNA acquisition from the Lactococcus lactis chromosome. Appl. Environ. Microbiol. 60: Sambrook, J., E. F. Fritsch, and T. Maniatis Molecular cloning: a laboratory manual, 2nd ed. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y. 33. Sing, W. D., and T. R. Klaenhammer Conjugal transfer of bacteriophage resistance determinants on ptr2030 into Streptococcus cremoris strains. Appl. Environ. Microbiol. 51: Sing, W. D., and T. R. Klaenhammer Characteristics of phage abortion conferred in lactococci by the conjugal plasmid ptr2030. J. Gen. Microbiol. 136: Terzaghi, B. E., and W. E. Sandine Improved medium for lactic streptococci and their bacteriophage. Appl. Microbiol. 29: Wirth, R., F. Y. An, and D. B. Clewell Highly efficient protoplast transformation system for Streptococcus faecalis and a new Escherichia coli-s. faecalis shuttle vector. J. Bacteriol. 165: Yanisch-Perron, C., J. Vieira, and J. Messing Improved M13 phage cloning vectors and host strains: nucleotide sequences of the M13mp18 and puc19 vectors. Gene 33: Downloaded from on June 19, 2018 by guest