Sequence analysis of Leuconostoc mesenteroides bacteriophage Phi1-A4 isolated from an industrial vegetable fermentation

Abstract

Vegetable fermentations rely on the proper succession of a variety of lactic acid bacteria (LAB). Leuconostoc mesenteroides initiates fermentation. As fermentation proceeds, L. mesenteroides dies off and other LAB complete the fermentation. Phages infecting L. mesenteroides may significantly influence the die-off of L. mesenteroides. However, no L. mesenteroides phages have been previously genetically characterized. Knowledge of more phage genome sequences may provide new insights into phage genomics, phage evolution, and phage-host interactions. We have determined the complete genome sequence of L. mesenteroides phage Phi1-A4, isolated from an industrial sauerkraut fermentation. The phage possesses a linear, double-stranded DNA genome consisting of 29,508 bp with a G+C content of 36%. Fifty open reading frames (ORFs) were predicted. Putative functions were assigned to 26 ORFs (52%), including 5 ORFs of structural proteins. The phage genome was modularly organized, containing DNA replication, DNA-packaging, head and tail morphogenesis, cell lysis, and DNA regulation/modification modules. In silico analyses showed that Phi1-A4 is a unique lytic phage with a large-scale genome inversion ( approximately 30% of the genome). The genome inversion encompassed the lysis module, part of the structural protein module, and a cos site. The endolysin gene was flanked by two holin genes. The tail morphogenesis module was interspersed with cell lysis genes and other genes with unknown functions. The predicted amino acid sequences of the phage proteins showed little similarity to other phages, but functional analyses showed that Phi1-A4 clusters with several Lactococcus phages. To our knowledge, Phi1-A4 is the first genetically characterized L. mesenteroides phage.

FIG. 1.

Transmission electron micrograph of phage Φ1-A4.

FIG. 2.

Genome atlas of phage Φ1-A4. The linear map was created using Genewiz, developed by Pederson et al. (60a; Center for Biological Sequence Analysis), and software developed in house. The legend at the bottom describes the individual panels corresponding to their respective designations (A to G). (A) Gapped BLASTP results obtained using the nonredundant database. (B) Gapped BLASTP results obtained using the custom phage genome database. In both panels A and B, blue regions represent unique proteins in Φ1-A4 and highly conserved features are red. The color intensity corresponds to the level of similarity. Individual E values were grouped into trust interval levels as depicted in the legend. (C) G+C content deviation: green shading, low-GC regions; orange shading, high-GC islands. (D) Annotation shows the absolute positions of functional features as indicated. (E) ORF orientation. ORFs in the sense orientation (ORF +) are blue; ORFs oriented in the antisense direction (ORF −) are red. The origin of DNA replication is indicated by a red inverted triangle. The cos site, depicting the phage genome boundaries, is indicated by two purple inverted triangles on the genome line (gray line). (F) COG classification. COG families were assembled into five major groups: 1, information storage and processing; 2, cellular processes and signaling; 3, metabolism; 4, poorly characterized; 5, ORFs with uncharacterized COGs or no COG assignment. (G) GC skew. ORFs with predicted functional annotation are shown at the top of panel A with bars indicating their absolute sizes. The color coding corresponds to predicted functional phage modules. Red, DNA replication; purple, DNA packaging; dark blue, phage head; sky blue, tail structure; green, lysis; greenish yellow, DNA regulation and modification; black, not assigned to a specific phage module.

FIG. 3.

Determination of the cos site. Two runoff Sanger electropherograms on linear phage DNA prepared from infectious particles identified the cos site boundaries. (a and c) Runoff electropherograms. (b) Manually enhanced traces identified the cos sequence. The figure is drawn to scale, and individual boxes are placed in relation to their expected sequence positions. Trace files were visualized using Chromas (http://www.technelysium.com.au/).

FIG. 4.

HindIII restriction digestion analysis of Φ1-A4 DNA. Lane M, 1-kb DNA ladder; lane Φ, Φ1-A4.

FIG. 5.

pACT alignment of Φ1-A4 against phages of its functional cluster (F-Cluster 1 in Fig. 7), as well as against Leuconostoc phage KM20 and L. mesenteroides ATCC 8293 prophage. Based on the FGD algorithm, predicted phage ORFeomes were subjected to an all-versus-all analysis. Results were stored in pMSP data files (http://www.sanger.ac.uk/Software/ACT/v7/manual/start.html#COMPARISON-FILE-FORMAT) and visualized using the Artemis comparison tool (10). Red lines show direct similarities between amino acid sequences of predicted ORFs, and blue lines represent inverted similarity hits. Color intensities are proportional to the respective increasing levels of sequence similarity. Phage genomes are represented by two gray lines (showing sense and antisense strands), and light blue arrowed boxes indicate individual predicted ORFs, their relative lengths, and their orientations. Genomes are drawn to scale.

FIG. 6.

Analysis of phage Φ1-A4 structural proteins. After the purification of Φ1-A4 by CsCl density gradient centrifugation, the phage-containing band was analyzed by 4 to 12% SDS-PAGE. Lane M, molecular mass marker; lane Φ, proteins from Φ1-A4. The N-terminal amino acid sequences of the five structural proteins were determined. The position of the N-terminal sequence in the corresponding Φ1-A4 ORF and its proposed function are indicated.

FIG. 7.

Functional genome distribution of 112 bacteriophages. A total of 112 bacteriophage genomes were analyzed using compACTor. A distance matrix was calculated and subsequently imported into Mega4. The functional history was inferred using the UPGMA. The optimal tree with a branch length sum of 2,115.25304686 is shown. The tree is drawn to scale.