Photorhabdus adhesion modification protein (Pam) binds extracellular polysaccharide and alters bacterial attachment

Abstract

Background: Photorhabdus are Gram-negative nematode-symbiotic and insect-pathogenic bacteria. The species Photorhabdus asymbiotica is able to infect humans as well as insects. We investigated the secreted proteome of a clinical isolate of P. asymbiotica at different temperatures in order to identify proteins relevant to the infection of the two different hosts.

Results: A comparison of the proteins secreted by a clinical isolate of P. asymbiotica at simulated insect (28 degrees C) and human (37 degrees C) temperatures led to the identification of a small and highly abundant protein, designated Pam, that is only secreted at the lower temperature. The pam gene is present in all Photorhabdus strains tested and shows a high level of conservation across the whole genus, suggesting it is both ancestral to the genus and probably important to the biology of the bacterium. The Pam protein shows limited sequence similarity to the 13.6 kDa component of a binary toxin of Bacillus thuringiensis. Nevertheless, injection or feeding of heterologously produced Pam showed no insecticidal activity to either Galleria mellonella or Manduca sexta larvae. In bacterial colonies, Pam is associated with an extracellular polysaccharide (EPS)-like matrix, and modifies the ability of wild-type cells to attach to an artificial surface. Interestingly, Surface Plasmon Resonance (SPR) binding studies revealed that the Pam protein itself has adhesive properties. Although Pam is produced throughout insect infection, genetic knockout does not affect either insect virulence or the ability of P. luminescens to form a symbiotic association with its host nematode, Heterorhabditis bacteriophora.

Conclusions: We studied a highly abundant protein, Pam, which is secreted in a temperature-dependent manner in P. asymbiotica. Our findings indicate that Pam plays an important role in enhancing surface attachment in insect blood. Its association with exopolysaccharide suggests it may exert its effect through mediation of EPS properties. Despite its abundance and conservation in the genus, we find no evidence for a role of Pam in either virulence or symbiosis.

Figure 1

Two-dimensional gel electrophoresis of the secreted proteome of P. asymbiotica ATCC43949. Proteins were obtained from supernatants of cultures grown at 28°C and 37°C and separated first by isoelectric point, along a 3-10 pH gradient, and then by mass. The 2D gel identifies several proteins with differential levels of production in these conditions, including S1 and S15 (circled) which are only secreted at 28°C.

Figure 2

Detection of Pam in infected G. mellonella. Each insect was injected with (A) P. luminescens TT01 or (B) P. asymbiotica ATCC43949, and was frozen and crushed in 1 ml of buffer at days 1 to 10 and 13 post injection. 10 μl of each sample was used per lane for SDS-PAGE, and Western blot analysis using anti-Pam antibody showed production from the second day after infection. The arrow indicates that Pam is not produced by Photorhabdus in the first day of G. mellonella infection or that it is below the detection limit of the assay.

Figure 3

In vitro Pam production. (A) Western blot confirmation of the temperature-dependent secretion of Pam in P. asymbiotica ATCC43949: the protein is present in supernatants from cultures grown at 28°C but not at 37°C. (B) Western blot confirmation of the pam knock-out in P. luminescens TT01 (left) compared with the wild-type strain (right). Note that figures A and B share the same molecular markers. (C) Pam heterologous production in E. coli. The arrow shows high levels of the recombinant Pam protein from P. asymbiotica ATCC43949 produced in E. coli. (D) Pam was purified by two steps of anion-exchange chromatography and the eluted fractions were analysed by SDS-PAGE. Lane 1: Proteins from overnight culture, lanes 2-5: elution fractions from the second ion-exchange column. The estimated purity of recombinant Pam was 95%.

Figure 4

Pam localization on bacterial cells. (A) Micrograph of a cross-section from a P. luminescens TT01 colony observed by TEM. Note the presence of an extensively folded extracellular matrix (black arrow) between the bacterial cells (indicated with P). (B) Immunolocalization of Pam using the anti-Pam antibody and a conjugated-gold secondary antibody. Gold particles extensively decorate the fibrillar EPS-like matrix (black arrow). (C) The TT01pam strain shows no anti-Pam antibody signal but the fibrillar matrix is still present. Scale bars are 0.2 μm. (D) Western blot confirming the presence of Pam in preparations of crude EPS. Lane 1: crude EPS extracted from TT01rif, lane 2: EPS from TT01pam and lane 3: purified recombinant Pam.

Figure 5

Comparison of bacterial attachment to surfaces in presence of insect hemolymph by fluorescence microscopy between TTO1rif and the pam mutant. Cells were grown on glass coverslips at 28°C for 8 hours. Planktonic bacteria were washed off and adherent bacteria were fixed and stained with DAPI. The adherence of TT01pam (B) is presented as a percentage of the data determined for the corresponding parental strain TT01rif (A). Bacterial counts were performed at 60× magnification and the data represent the mean values of 12 fields from triplicate experiments (± St.Dev) (C).

Figure 6

Surface plasmon resonance analysis of Pam-mediated adhesion on gold-coated glass probes. (A and B) Presence of the protein on the cell surface (washed cells) showed decreased adhesion to untreated gold surfaces in both TT01rif and E. coli pBADpam (+Pam), when compared with the correspondent strains lacking Pam, TT01pam and E. coli pBAD respectively (-Pam). (C and D) Supernatants from cultures expressing pam, TT01rif and E. coli pBADpam (+Pam), showed more adhesion than those lacking the protein TT01pam and E. coli pBAD (-Pam), indicating the ability of free Pam to adhere to surfaces.

Figure 7

Structural properties of Pam. (A) Graphical output of far-UV CD data for Pam reveals that experimental data (green crosses) and calculated spectrum (blue boxes), derived from the calculated output secondary structure, show agreement. The difference spectrum (purple lines) is very close to zero throughout the wavelength range, indicating the goodness of fit of the structural predictions. The CD data indicate that Pam is largely helical (58%), with only a small fraction of residues forming β-strands. (B) Thermal stability of Pam measured by differential scanning calorimetry. The normalised thermal transition curve (red line) shows energy uptake by Pam reached a peak (Tm) at 77.4°C, representing the temperature at which 50% of the protein molecules are unfolded. This was almost identical after cooling the sample and repeating (black line).