The production of pediocin PA-1, a small heat-stable bacteriocin, is associated with the presence of the 9.4-kbp plasmid pSRQii in Pediococcus acidilactici PAC1.0. It was shown by subcloning of pSRQll in Escherichia coli cloning vectors that pediocin PA-1 is produced and, most probably, secreted by E. coli cells. Deletion analysis showed that a 5.6-kbp SalI-EcoRI fragment derived from pSRQ11 is required for pediocin PA-1 production. Nucleotide sequence analysis of this 5.6-kbp fragment indicated the presence of four clustered open reading frames (pedA, pedB, pedC, and pedD). The pedA4 gene encodes a 62-amino-acid precursor of pediocin PA-1, as the predicted amino acid residues 19 to 62 correspond entirely to the amino acid sequence of the purified pediocin PA-1. Introduction of a mutation in pedAl resulted in a complete loss of pediocin production. The pedB and pedC genes, encoding proteins of 112 and 174 amino acid residues, respectively, are located directly downstream of the pediocin structural gene. Functions could not be assigned to their gene products; mutation analysis showed that the PedB protein is not involved in pediocin PA-1 production. The mutation analysis further revealed that the fourth gene, pedD, specifying a relatively large protein of 724 amino acids, is required for pediocin PA-1 production in E. coli. The predicted PedD protein shows strong similarities to several ATP-dependent transport proteins, including the E. coli hemolysin secretion protein HlyB and the ComA protein, which is required for competence induction for genetic transformation in Streptococcus pneumoniae. These similarities suggest strongly that the PedD product is involved in the translocation of pediocin PA-1.

No abstract

A novel plasmid-based expression strategy, exploiting two features of lytic bacteriophages, was developed in Lactococcus lactis. Components of this system include a phage origin of replication and phage expression signals, which were induced to high efficiency upon phage infection of the host. Phage-specific expression signals were cloned from phi 31 in a promoter-screening strategy using the lacZ gene from Streptococcus thermophilus. One clone exhibited a significant induction in beta-galactosidase production and concomitant increase in lacZ mRNA during the phi 31 infection cycle of the host. Molecular characterization of the cloned insert revealed 888 bp positioned near the phi 31 cos site. Primer extension analysis showed that transcription was induced approximately 20 min following phi 31 infection at four points, apparently organized in two sets of tandem promoters on the cloned phage insert. One of these middle phage promoters also showed a basal level of activity prior to phage infection. The phi 31 promoter lacZ cassette was cloned into a low-copy-number vector plasmid containing the phi 31 origin of replication (ori31) and the resulting low-copy-number plasmid exhibited negligible beta-galactosidase production in L. lactis. However, > 2,000 units were detected following a deliberate infection with phi 31. A control expression plasmid without ori31 could only be induced to 85 units. The combination of these phage-inducible expression signals together with ori31 functioned synergistically to drive rapid and high efficiency expression of a heterologous gene in L. lactis.

Lactococcus lactis W-37 is highly resistant to phage infection. The cryptic plasmids from this strain were coelectroporated, along with the shuttle vector pSA3, into the plasmid-free host L. lactis LM0230. In addition to pSA3, erythromycin- and phage-resistant isolates carried pSRQ900, an 11-kb plasmid fromL. lactis W-37. This plasmid made the host bacteria highly resistant (efficiency of plaquing <10−8) to c2- and 936-like phages. pSRQ900 did not confer any resistance to phages of the P335 species. Adsorption, cell survival, and endonucleolytic activity assays showed that pSRQ900 encodes an abortive infection mechanism. The phage resistance mechanism is limited to a 2.2-kbEcoRV/BclI fragment. Sequence analysis of this fragment revealed a complete open reading frame (abiQ), which encodes a putative protein of 183 amino acids. A frameshift mutation within abiQ completely abolished the resistant phenotype. The predicted peptide has a high content of positively charged residues (pI = 10.5) and is, in all likelihood, a cytosolic protein. AbiQ has no homology to known or deduced proteins in the databases. DNA replication assays showed that phage c21 (c2-like) and phage p2 (936-like) can still replicate in cells harboring AbiQ. However, phage DNA accumulated in its concatenated form in the infected AbiQ+ cells, whereas the AbiQ− cells contained processed (mature) phage DNA in addition to the concatenated form. The production of the major capsid protein of phage c21 was not hindered in the cells harboring AbiQ.

A novel system that leaks ␤-galactosidase (␤-gal) without a requirement for secretion or export signals was developed in Lactococcus lactis by controlled expression of integrated phage holin and lysin cassettes. The late promoter of the lytic lactococcal bacteriophage 31 is an 888-bp fragment (P 15A10 ) encoding the transcriptional activator. When a high-copy-number P 15A10 ::lacZ.st fusion was introduced into L. lactis strains C10, ML8, NCK203, and R1/r1t, high levels of the resultant ␤-gal activity were detected in the supernatant (approximately 85% of the total ␤-gal activity for C10, ML8, and NCK203 and 45% for R1/r1t). Studies showed that the phenotype resulted from expression of Tac31A from the P 15A10 fragment, which activated a homologous late promoter in prophages harbored by the lactococcal strains. Despite the high levels of ␤-gal obtained in the supernatant, the growth of the strains was not significantly affected, nor was there any evidence of severe membrane damage as determined by using propidium iodide or transmission electron microscopy. Integration of the holin-lysin cassette of phage r1t, under the control of the phage 31 late promoter, into the host genome of MG1363 yielded a similar "leaky" phenotype, indicating that holin and lysin might play a critical role in the release of ␤-gal into the medium. In addition to ␤-gal, tetanus toxin fragment C was successfully delivered into the growth medium by this system. Interestingly, the X-prolyl dipeptidyl aminopeptidase PepXP (a dimer with a molecular mass of 176 kDa) was not delivered at significant levels outside the cell. These findings point toward the development of bacterial strains able to efficiently release relevant proteins and enzymes outside the cell in the absence of known secretion and export signals.Lactococcus lactis is best known for its role in mesophilic dairy fermentations, including those used in production of cheddar cheese, buttermilk, and sour cream. Its long history of safe use in the food industry and generally recognized as safe (GRAS) status provide new opportunities for using L. lactis in important roles in food biotechnology, particularly in the presentation of vaccines, antimicrobial agents, or intracellular peptidases involved in cheese ripening, outside of the cell. Three major mechanisms have been exploited in these studies. First, signal sequences of the lactococcal secreted protein Usp45 (43), the lactococcal proteinase (46), and the S-layer protein (encoded by slpA) (45) of Lactobacillus brevis have been employed to secrete heterologous proteins from L. lactis via the secretory pathway. ATP-binding cassette-transporter export systems have also been used to export heterologous bacteriocins from L. lactis (for a review, see reference 1). Lastly, induction of cell autolysis can result in efficient release of homologous and heterologous proteins from the cell. Interest in naturally occurring autolytic strains has focused on the release of intracellular enzymes, namely peptidases, into the cheese medium to enhance fl...