Get Complete Project Material File(s) Now! »
Agrobacterium identification using whole-cell MALDI-TOF mass spectrometry4
Fine typing of agrobacteria strains is an absolute prerequisite to assess the environmental diversity of agrobacteria for epidemiological and agro-ecological investigations especially to determine the epidemic or endemic status of crown gall outbreaks. As a reminding point, current techniques were divided into those who that grossly identify agrobacteria at the biovar level and those that operate very fine discriminations at the strain/isolate level. Methodologies delineating bacteria at the biovar level are not accurate enough for that purposes. Indeed, except for biovar 3 (i.e. A. vitis) that is infeodated to vine and therefore found only in association with this plant, biovars 1 and 2 are so common that they are consistently found in almost all environments (Chapter 1-3). Biovar typing is therefore of limited interest to elucidate factors that shape agrobacterial population genetic structures. At the opposite, the identification at the strain/isolate level was found to be very relevant for epidemiological investigations because finding the same strains is different crown gall outbreaks is a clue that those outbreaks are epidemiologically related (Chapter 1-3; Ponsonnet et al., 1994; Pionnat et al., 1999). However, in contrast, identification at the strain/isolate level is often too detailed for agro-ecological studies. Indeed, the environmental diversity agrobacterial strain is generally so huge that finding the same strains in different environment is a rare event except of course in the case of epidemics as described above. In this respect, the MLSA approach developed by Costechareyre (thesis 2009) for agrobacteria is relevant methodology to define the strain, species and genus status of any agrobacterial isolates. However, sequencing of housekeeping genes is labor-intensive, time-consuming, and impractical for high-throughput investigations. Thus, there is still a valuable need of simplified and rapid approach to perform mass agrobacterium identification.
Whole-cell matrix-assisted laser desorption ionization-time of flight mass spectrometry (MALDI-TOF MS) is an emerging technology (Fig. 2-3-1) in microbial diagnostics (Krader and Emerson, 2004; Sauer et al., 2008). Identity is based on unique mass/charge ratio (m/z) fingerprints of proteins, which are ionized using short laser pulses directed to bacterial cells obtained from a single colony embedded in a matrix. After desorption, ions are accelerated in vacuum by a high electric potential and separated on the basis of the time taken to reach a detector, which is directly proportional to the mass-to-charge ratio of an ion. This technique has been shown to deliver reproducible protein mass fingerprints starting from an aliquot of a single bacterial colony within minutes and without any prior separation, purification, or concentration of samples. Whole-cell MALDI-TOF MS is a reliable technique across broad conditions (e.g., different growth media, cell growth states), with limited variability in mass-peak signatures within a selected mass range (2,000 < m/z < 20,000) that does not affect reliability of identification (Lay, 2001). MALDI-TOF MS profiles primarily represent ribosomal proteins, which are the most abundant cellular proteins and are synthesized under all growth conditions (Ryzhov et al., 2001). However, this approach is not informative as it relies upon the availability of a reference database. Current databases are weighted to clinical taxa (Bessède et al., 2011; Stephan et al., 2010) and only a few references are available for plant-associated and environmental bacteria (Sauer et al., 2008; Rezzonico et al., 2010; Wensing et al., 2010) .
Figure 2-1-1. Sequence of identification by MALDI-TOF mass spectrometry of a bacterial colony in a clinical microbiology laboratory. (Photographs: Olivier Gaillot, Laboratoire de bactériologie-hygiène, CHRU de Lille)
In this chapter, we evaluated the usefulness of MALDI-TOF MS technology as a high throughput tool to characterize and classify agrobacteria. For this aim, strains covering all Agrobacterium species known so far as well as outgroup strains belonging to the Rhizobiaceae family were analyzed using the whole cell MALDI-TOF MS approach. To evaluate the quality of the taxonomic assignation done by the method, the MALDI-TOF MS data clustering was compared to the recA phylogeny done with the same strains. In addition, we evaluated whether the technique was able to provide information about strain pathogenicity by looking for consistant and remarkable differences between MALDI-TOF MS patterns of a standard pathogenic strain (C58) and its plasmid free derivatives.
Material and Method
Bacterial strains and growth conditions
A total of 114 strains covering all the genomic species richness of the genus Agrobacterium known so far as well as 4 outgroup strains belonging to the Rhizobiaceae family were investigated with whole-cell MALDI-TOF MS with 1 to 19 strains per species (Table 2-1-1). A special emphasis was given to A. vitis, with a selection of 19 strains done by P. Portier on the basis of their recA diversity. Along to the standard pathogenic strain C58, C58 derivative cured of pTiC58 and pAtC58 plasmids or containing other Ti plasmids were included in the analysis (Table 2-1-2). Bacteria were grown on MG agar (Ophel and Kerr, 1990) and LPG agar medium (yeast extract, 5 g/L; Bacto peptone, 5 g/L; glucose, 10 g/L) at 28°C for 42-50 h.
IC MALDI-TOF analysis
Colonies harvested on LPG plates were directly smeared onto target spots of a polished ground steel 48-position MALDI-TOF MS sample target (Industrietechnik mab AG, Basel, Switzerland) with two distinct spot per strain using a blind random spotting design. Each spot was then overlaid with 1 l of a saturated solution of sinapinic acid (SA – 49508, Sigma-Aldrich, Buchs, Switzerland) in 60% acetonitrile (154601, Sigma-Aldrich, Buchs, Switzerland) – 0.3% trifluoroacetic acid (T6508, Sigma-Aldrich, Buchs, Switzerland), and air-dried for some minutes at room temperature.
Protein mass fingerprints were obtained using an AXIMA Confidence MALDI-TOF Mass Spectrometry (Shimadzu-Biotech Corp., Kyoto, Japan), with a detection in the linear, positive mode at a laser frequency of 50 Hz and within a mass range from 2,000-20,000 Daltons. The acceleration voltage was 20 kV, and the extraction delay time was 200 ns. MALDI-TOF MS spectra were the average of at least four replicate measurements using slides prepared at least in two different instances. Protein mass fingerprints were determined using 100 different laser direction and at least 10 laser shots per direction, then averaged and processed to provide a single raw spectrum using the Launchpad v2.8 software (Shimadzu-biotech, Kyoto, Japan). The software was also used for peak processing using the following settings: the Advanced Scenario was chosen from the Parent Peak Cleanup menu, Peak Width was set to 80 channels, Smoothing Filter Width to 50 channels, Baseline Filter Width to 500 channels and Threshold Apex was chosen as the peak detection method. For the Threshold Apex Peak Detection, the Threshold Type was set to Dynamic, the Threshold Offset to 0.020 mV and the Threshold Response Factor to 1.2. Each target plate was externally calibrated using spectra of reference strain Escherichia coli DH5 alpha. Processed spectra were reduced to the 150 most intensive peaks with an ad hoc PERL script written by J.F. Pothier, V. Pflüger, B. Duffy (unpublished), and exported as peak lists with m/z values and signal intensity in the ASCII format. A binary matrix was then generated using the SARAMIS™ (Spectral ARchive And Microbial Identification System, AnagnosTec, Potsdam-Golm, Germany) SuperSpectrum tool. Peak lists were trimmed to a mass range of 2-20 kDa. Peak lists were binned and average masses were calculated using the SARAMIS™ SuperSpectrum tool with an error of 800 ppm. Final consensus spectra for each strain were generated with a customized VBA script (J.F. Pothier, V. Pflüger, B. Duffy, unpublished) by eliminating masses present within fewer than half of the replicate measurements. Multivariate cluster analysis using the UPGMA algorithm with Dice coefficient was performed in PAST v2.14 (Hammer et al., 2001) and the resulting dendrogram was visualized with FigTree v1.3.1.
For clustering, a dendogram was constructed using strain consensus spectra selecting peaks that were found in at least half of 4 independent measurements, with the weighted pair-group average (UPGMA) clustering algorithm with the Dice coefficient. Similarity distances are expressed as percentage. The tree topology robustness was evaluated using the 1000 bootstrap resembling of peak values.
Table of contents :
Part 1: Bibliography
1-1: Rapid and efficient methods to isolate, type strains and determine species of Agrobacterium spp. in pure culture and complex environment
1-2: Detection and identification methods and new tests as developed and used in the framework of COST873 for bacteria pathogenic to stone fruits and nuts: Tumorigenic Agrobacterium spp.
1-3: Agrobacterium biodiversity studies
Part 2: Experimental
2-1: Agrobacterium identification using whole-cell MALDI-TOF mass spectrometry
2-2: Rapid and accurate species identification and exhaustive population diversity assessment of Agrobacterium spp. using recA-based PCR
2-3: Assessing the presence of agrobacteria and Ti plasmids in soil at a country scale
2-4: Assessing the pyrosequencing approach to analyze Agrobacterium microdiversity
Discussion and perspectives
Annexes