Spatiotemporal expression of the putative MdtABC efflux pump of Phtotorhabdus luminescens occurs in a protease-dependent manner during insect infection

Abstract

Photorhabdus luminescens is an enterobacterium establishing a mutualistic symbiosis with nematodes, that also kills insects after septicaemia and connective tissue colonization. The role of the bacterial mdtABC genes encoding a putative multidrug efflux system from the resistance/nodulation/cell division family was investigated. We showed that a mdtA mutant and the wild type had similar levels of resistance to antibiotics, antimicrobial peptides, metals, detergents and bile salts. The mdtA mutant was also as pathogenic as the wild-type following intrahaemocoel injection in Locusta migratoria, but had a slightly attenuated phenotype in Spodoptera littoralis. A transcriptional fusion of the mdtA promoter (PmdtA) and the green fluorescent protein (gfp) encoding gene was induced by copper in bacteria cultured in vitro. The PmdtA-gfp fusion was strongly induced within bacterial aggregates in the haematopoietic organ during late stages of infection in L. migratoria, whereas it was only weakly expressed in insect plasma throughout infection. A medium supplemented with haematopoietic organ extracts induced the PmdtA-gfp fusion ex vivo, suggesting that site-specific mdtABC expression resulted from insect signals from the haematopoietic organ. Finally, we showed that protease inhibitors abolished ex vivo activity of the PmdtA-gfp fusion in the presence of haematopoietic organ extracts, suggesting that proteolysis by-products play a key role in upregulating the putative MdtABC efflux pump during insect infection with P. luminescens.

Fig 1. Genomic organisation of the mdtABC and baeSR loci in Photorhabdus luminescens TT01 and several enterobacteria.

Photorhabdus luminescens TT01 (BX470251), Salmonella typhimurium LT2 (AE006468), and Escherichia coli K-12 (U00096). kb, kilobases.

Fig 2

Survival plot of Spodoptera littoralis insect larvae after the injection of Photorhabdus luminescens TT01 wild-type (black line) and mdtA mutant (dotted line). Each experiment was performed with 20 larvae for each bacterium tested and the results shown are the survival functions obtained in four independent experiments. Statistical analyses with SPSS showed that the survival curve of the mdtA mutant was significantly different from that of the wild type (P < 0.0001). The calculated median survival times of larvae infected with WT and mdtA mutant were 34.9 h and 43.5 h, respectively.

Fig 3. Copper significantly enhances the P_mdtA-gfp[AAV] activity in vitro in Photorhabdus luminescens.

Overnight cultures of P. luminescens carrying the P_mdtA-gfp[AAV] fusion were diluted (1/500) in fresh LB medium in black-sided clear-bottomed 96-well plates (Greiner) and incubated at 28°C, with shaking, in a microplate reader system (TECAN). 2mM Copper was added to the bacterial culture at mid log phase (O.D₆₀₀ equivalent to 0.4). Changes in A_600nm and GFP fluorescence were monitored after copper supplementation and the data shown are maximum specific fluorescence obtained by FU/A_600nm ratio value, calculated at six hours after the addition of copper. The results shown are the means of three experiments and asterisks indicate statistically significant differences (*, P < 0.05) in paired Student’s t tests.

Fig 4. Comparison of the in vivo expression of the Photorhabdus luminescens mdtABC operon and constitutive GFP expression in Locusta migratoria and Spodoptera littoralis haemolymphs.

Insects were injected with wild-type P. luminescens harbouring a transcriptional fusion between E. coli P_lac (A, B, E, and F) or the mdtABC (C, D, G, and H) promoter and the gfp[AAV] reporter gene. Haemolymph was regularly extracted from infected larvae during infection, observed under a light microscope (A, C, E, and G) and fluorescence was determined (B, D, F, and H). The observations shown were made at 20 to 24 hours post-injection for S. littoralis larvae and at 30 hours post-injection for L. migratoria, and correspond to the results of at least three independent experiments. Scale bar represents 10 μm.

Fig 5. In vivo site-specific expression of the Phototrhabdus luminescens mdtABC operon during colonization of Locusta migratoria HO and Spodoptera littoralis midgut.

Insects were injected with recombinant P. luminescens-P_lac-gfp[AAV] (A-E and K-O) or with recombinant P. luminescens-P_mdtA-gfp[AAV] (F-J and P-T). HO tissues were observed under a light microscope (B, D, G, I; L, N, Q, and S), or by fluorescence microscopy (A, C, E, F, H, and J). Nodules formed after the injection of P. luminescens TT01 or its derivatives into the L. migratoria haemocoel are indicated with white arrows (A, F, B, and G). Note the presence of bacterial aggregates within the connective tissues of the HO in L. migratoria (A and F) and in the vicinity of the midgut in S. littoralis (K and P), each injected with Photorhabdus. Pieces of HO in which nodule structures were visible (B and G) or not visible (D and I) were observed under the epifluorescence microscope (C and E, H and J, respectively). Brightly fluorescent green bacteria were detected in both cases, demonstrating the HO site-specific expression of P_mdtA-gfp[AAV] within closed bacterial aggregates, independently of nodule structures. The circle (G) indicates the limit of a bacterial aggregate and arrowheads (H) indicate isolated bacteria. All these observations were made 20 to 28 hours post-injection for S. littoralis and at 30 hours post-injection for L. migratoria, and correspond to results of at least three independent experiments. ECM, extracellular matrix. Red scale bar represents 10 μm. White scale bar represents 50 μm.

Fig 6. Level of expression of Photorhabdus luminescens mdtABC operon in Locusta migratoria plasma and medium containing HO extract in vitro.

Overnight cultures of P. luminescens carrying the P_mdtA-gfp[AAV] fusion were diluted (1/500) in black-sided clear-bottomed 96-well plates containing one of the following media: fresh LB medium, plasma or fresh LB medium supplemented with L. migratoria HO extract. A_600nm and GFP fluorescence were monitored, in real time, in triplicate for each set of conditions, at 28°C, with orbital shaking, in an Infinite M200 microplate reader (TECAN). Specific fluorescence was obtained by dividing GFP fluorescence units (FU) by the A_600nm value at 20 hours of growth. Relative Fluorescence Units (RFUs) were determined by the ratio of the P_mdtA-gfp[AAV] specific fluorescence value to that of the P_lac-gfp[AAV]. Asterisks indicate statistically significant differences (*, P <0.05) in the paired Student’s t test. The expression rate is the ratio between specific activity obtained after incubation with HO extracts and that obtained after incubation with insect plasma.

Fig 7. Cysteine and serine protease inhibitors prevent expression of the mdtABC operon.

TT01 strains carrying the P_mdtA–gfp[AAV] (A) or P_lac–gfp[AAV] (B) fusions were grown in LB medium containing HO extracts and supplemented (× and □) or not supplemented (● and ♦) with complete mini EDTA-free protease inhibitor. Specific fluorescence is expressed as the ratio of GFP fluorescence units (FU) to A_600nm value. The results are representative of three independent assays.