Activation of mRNA translation by phage protein and low temperature: the case of Lactococcus lactis abortive infection system AbiD1

Abstract

Background: Abortive infection (Abi) mechanisms comprise numerous strategies developed by bacteria to avoid being killed by bacteriophage (phage). Escherichia coli Abis are considered as mediators of programmed cell death, which is induced by infecting phage. Abis were also proposed to be stress response elements, but no environmental activation signals have yet been identified. Abis are widespread in Lactococcus lactis, but regulation of their expression remains an open question. We previously showed that development of AbiD1 abortive infection against phage bIL66 depends on orf1, which is expressed in mid-infection. However, molecular basis for this activation remains unclear.

Results: In non-infected AbiD1+ cells, specific abiD1 mRNA is unstable and present in low amounts. It does not increase during abortive infection of sensitive phage. Protein synthesis directed by the abiD1 translation initiation region is also inefficient. The presence of the phage orf1 gene, but not its mutant AbiD1R allele, strongly increases abiD1 translation efficiency. Interestingly, cell growth at low temperature also activates translation of abiD1 mRNA and consequently the AbiD1 phenotype, and occurs independently of phage infection. There is no synergism between the two abiD1 inducers. Purified Orf1 protein binds mRNAs containing a secondary structure motif, identified within the translation initiation regions of abiD1, the mid-infection phage bIL66 M-operon, and the L. lactis osmC gene.

Conclusion: Expression of the abiD1 gene and consequently AbiD1 phenotype is specifically translationally activated by the phage Orf1 protein. The loss of ability to activate translation of abiD1 mRNA determines the molecular basis for phage resistance to AbiD1. We show for the first time that temperature downshift also activates abortive infection by activation of abiD1 mRNA translation.

Figure 1

Transcriptional analysis of the abiD1 gene. (A) Schematic organization of the abiD1 gene region. Bent arrows and circles denote promoter and terminator sequences, respectively. Position of oligonucleotides used for Northern hybridization (1) and quantitative RT-PCRs are indicated by tail-less arrows. Transcripts initiated at the abiD1 promoter are shown by solid (50 b) and broken (3.6 kb) lines. (B) Northern hybridization results. RNA was extracted from IL1403 AbiD1+ cells grown with or without Cm. Hybridization was performed with oligonucleotide n°1. (C) Quantitative RT-PCRs. abiD1 transcript was followed in IL1403 AbiD1+ cells. Samples were taken before (time 0), 10 and 20 min after infection with bIL66 phage. Amount of abiD1 transcript was normalized to L. lactis tuf transcript level. Values shown are means of 9–15 measurements, expressed in arbitrary units.

Figure 2

Translation of the abiD1 mRNA. (A) MFOLD [44] predicted secondary structure of the abiD1 translation initiation region (TIR). The sequence is presented from the transcriptional start (+ 1) to the abiD1 AUG start codon. Ribosome binding site (RBS) and the poly-U stretch are boxed. Arrows show the 5' ends of different TIRs used. The start of luxAB gene in fusion constructs is indicated. (B) Translation of abiD1 TIR:luxAB fusions. The background level of luciferase activity was measured in the presence of plasmid pJIM1715 [50]. Experiments were performed at 30°C. Results are means of 6–10 independent experiments. Luciferase activity is shown in arbitrary light units (lux/OD unit at an OD₆₀₀ of 0.4).

Figure 3

Orf1 activates translation of abiD1 TIR: luxAB fusions. (A) Translation of abiD1 TIR I:luxAB, abiD1 TIR II:luxAB and aldB TIR:luxAB fusions in the absence (control) or in the presence of orf1, orf1M1 and orf1M2 genes in L. lactis IL1403 cells. Experiments were performed at 30°C. Results are means of 6 independent experiments. (B) Transcription of abiD1 TIR:luxAB fusions. The amount of abiD1 TIR I: luxAB and abiD1 TIR II: luxAB transcripts was measured by quantitative RT-PCR in the absence (control) and in the presence of orf1 or orf1M1 genes in L. lactis cells grown at 30°C. The aldB TIR: luxAB fusion was used as a control. Amount of transcript was normalized to the L. lactis tuf transcript level. Values are means of 7 – 15 measurements, expressed in arbitrary units. (C) Translation of abiD1 TIR II:luxAB fusion in the presence of plasmid- or chromosome- integrated orf1 and orf1M2 in B. subtilis 168 cells. Experiments were performed at 30°C. Results are means of 4–5 independent experiments. Luciferase activity is shown in arbitrary light units (lux/OD unit at an OD₆₀₀ of 0.4).

Figure 4

Effect of temperature on expression of the abiD1 TIRI : luxAB fusion. (A) Translation of abiD1 TIR I:luxAB fusion in L. lactis IL1403 cells at 30°C and 18°C. The background level of luciferase activity at 30°C and 18°C was measured in the presence of the control plasmid pJIM1715 [50]. Results are means of 5 independent experiments. Luciferase activity is shown in arbitrary light units (lux/OD unit at an OD₆₀₀ of 0.4). (B) Expression of FLAG-tagged LuxAB at 30°C and 18°C. L. lactis cells carrying FLAG tagged abiD1 TIR I:luxAB fusion plasmid were grown to an OD₆₀₀ of 0.4 at either 30°C or 18°C in M17 medium. Twenty mg of total cell protein were separated on an 8% SDS-PAGE gel. Western blot was developed with anti-FLAG tag M2 monoclonal antibody. The control panel shows cell lysates without specific FLAG-tagged protein. Molecular masses of standard proteins (Prestained Protein Marker, Broad Range, New England BioLabs) are indicated on the right. (C) Transcription of abiD1 TIR:luxAB fusion. The amount of abiD1 TIR I: luxAB transcript in L. lactis cells grown at 30°C and 18°C was quantified by ImageQuant software after dot-blot Northern hybridization and by quantitative RT-PCR. Values are means of 3–7 measurements, expressed in arbitrary units.

Figure 5

Orf1-RNA binding activity. (A) Schematic organization of the abiD1 gene. Arrows indicate position of abiD1 transcripts (1 to 7) used for binding experiments. (B) Binding of purified Orf1, OrfM1 and Orf1M2 to the abiD1 mRNA 1. Ten ng of radiolabelled abiD1 mRNA (transcript 1) were incubated with increasing amounts of Orf1, OrfM1 or OrfM2 protein (0; 0.03 μM; 0.05 μM; 0.11 μM; 0.22 μM), followed by separation of the nucleoprotein complex in an 8% non-denaturing polyacrylamide gel at room temperature. (C) Binding of Orf1 to the abiD1 mRNA 6 and abiD1 mRNA 7. (D) Binding of Orf1 to trpA mRNA and aldB mRNA. Experiments (C and D) were performed with increasing amounts of Orf1 protein (0; 0.11 μM; 0.22 μM; 0.33 μM) in the same conditions as above.

Figure 6

Putative Orf1-binding mRNA motif. (A) Predicted mRNA secondary structures reveled by FOLDALIGN [67,68] in abiD1 mRNA (nucleotide positions are 7341–7362; 7518–7539; GenBank Accession Number AF116286) and (B) phage bIl66 mid-infection induced M-operon mRNA (nucleotide positions are 566–591; GenBank Accession Number L35175) and L. lactis IL1403 cold shock inducible genes osmC, hslA, clpX, llrC and pgmB mRNAs (nucleotide positions are 68675–68698; 502344–502363; 1163846–1163866; 403740–403763 and 442024–442049, respectively; GenBank Accession Number NC_002662). Gene names are indicated. (C) Binding of Orf1 to M-operon and osmC RNAs. Experiments were performed with increasing amounts of Orf1 (0; 0.05 μM; 0.11 μM; 0.22 μM) in the same conditions as above.