Alpha-cyano-4-hydroxycinnamic acid

Tubes and tips were purchased from Eppendorf. All chemicals were of highest purity (HPLC-grade) and purchased from Sigma-Aldrich.
The Erwinia strains came from various strain collections, others (described in ref. 3) and the E. coli K12-strains 1100 and W3350 were obtained from the collection of the JKI Dossenheim. The bacteria used for the reference database and the dendrogram were the following: E. amylovora CFBP1232 (T), E. amylovora Ea1/79 DSM 17948, E. amylovora 273 ATCC 49946, E. amylovora IL6 (rubus) (Lab collection JKI Dossenheim, isolated in Illinois, USA), E. amylovora MR1 (rubus) (Lab collection JKI Dossenheim, isolated in Michigan, USA), E. pyrifoliae 16/96 (T) DSM 12163, E. pyrifoliae 1/96 DSM 12162, E. pyrifoliae Ejp557 (Japan) (Lab collection JKI Dossenheim, isolated from Nashi pear, Hokkaido, Japan, 1994, A. Tanii), E. tasmaniensis 1/99 (T) DSM 17950, E. tasmaniensis 2/99 DSM 17949, E. billingiae Eb 660 (T) NCPPB660 and Eb 661 (T) NCPPB661, E. persicina CFBP3622 (T), E. rhapontici CFBP3618 (T), E. psidii CFBP3627 (T), Pectobacterium cacticida CFBP3628 (T), Brenneria quercini CFBP3617 (T), E. mallotivora CFBP2503 (T), E. toletana CFBP6631 (T), E. papayae CFBP5189 (T), E. tracheiphila CFBP2355 (T), E. coli 1100 (E. coli/ K-12, Lab collection JKI Dossenheim), E. coli W3350 (E. coli/ K-12, Lab collection JKI Dossenheim). Erwinia type strains are indicated (T); CFBP = Collection Française des Bactéries Phytopathogènes; DSMZ = German Collection of Microorganisms and Cell Cultures; NCPPB = National Collection of Plant Pathogenic Bacteria (UK); ATCC = American Type Culture Collection; JKI = Julius Kuehn Institute.
Infection of in vitro pear plants (micro-propagated plants): Pear leaves were wounded and inoculated with cells of the German E. amylovora strain Ea1/79. After incubating for 5 days, the infected pear plantlets displayed symptoms typical of fire blight infection, such as water soaking and necrosis accompanied by the production of bacterial ooze. We washed the bacteria from the plant surface with 1.5 ml water, centrifuged the samples at 1000× g for a minute, and decanted the liquid. Afterwards we suspended and inactivated the bacteria as is described below.
Isolation of Erwinia spp. from necrotic wood of pear trees (from Carinthia): Fifty milligram of dark bark slices contaminated with bacteria were immersed in 1 ml water. After soaking for 15 minutes, samples were diluted, and 200 µl of that were plated on LB agar with cycloheximide (50 µg/ml). White colonies were assayed on semi-selective agar for E. amylovora by using PCR and DNA sequencing at the JKI Dossenheim. E. amylovora colonies were processed as blind samples for MALDI analysis at the Max-Planck-Institute for Molecular Genetics (Berlin).
Cell culturing on agar: For cell culturing on agar plates, all dilutions were incubated for 2 days at 28°C. Bacteria were suspended from cell lawns in 1 ml water to a density of approximately 1 (light absorption at 600 nm) and centrifuged. The pellets were washed with 1 ml water and then the liquid was discarded. The presence of culture medium adhering to the bacterial colonies cells from agar had no visible effect on the mass signal patterns. The bacteria were inactivated as described below.
Culturing in liquid media: Bacteria grown on agar were inoculated into LB liquid medium with 1% glucose for the generation of reference spectra and in many cases for identification of unknown samples. The medium was autoclaved and then filtrated through a 0.2 µm nitrocellulose filter to remove particles. Replacement of LB-glucose by LB-glycerol showed little effect in the peak pattern distribution. Identification of bacteria grown on different media was reliably achieved as shown in Table 1. Once the bacteria have entered the stationary phase, the method is robust against growth times. However, other (minimal) media might have a stronger influence on the mass peak patterns. For the generation of reference spectra, we used LB-glucose as standard medium because most Erwinia bacteria grew well in this medium and resulted in very good mass spectra in terms of sensitivity and resolution.
Inactivation of bacteria: The bacteria were suspended in 300 µl water and inactivated by the addition of 800 µl ethanol at room temperature. The samples could be stored at room temperature for several days or at 4–8°C for several weeks. To assay for viability, we applied dilutions to agar plates and found no surviving E. amylovora cells already after an hour of storage in ethanol.
Protein extraction: This step was performed at room temperature. The solution was centrifuged at 25,000×g for 2 minutes and the supernatant was discarded. Again, centrifugation was performed for 2 minutes at 25,000×g and residual supernatant was discarded. Five to 20 µl of 70% formic acid were added to the “pellet” (1 to 5 mg, or less bacterial material), and mixed to re-suspend the bacteria. Then 5–20 µl acetonitrile were added, accordingly, and the sample was mixed carefully. The solution was centrifuged at 25,000×g for 2 minutes. The supernatant (∼5–20 µl) was transferred to a new tube immediately.
MALDI preparation: This step was performed at room temperature and at 20–80% air humidity. One microliter of the supernatant was placed onto a stainless steel target plate and led dry in air. Then, 1 µl of matrix (3 mg/ml solution of alpha-cyano-4-hydroxycinnamic acid in 50% acetonitrile/2.5% trifluor acetic acid) was overlaid onto the dried sample and led dry in air. This simple preparation method provided homogenous samples to enable automated measurements and sufficiently reproducible mass spectra. To increase data reliability, we applied each bacterial sample six times onto the target plate.
Mass spectrometry detection: Mass spectra were acquired using an Ultraflex I MALDI-TOF mass spectrometer (Bruker Daltonics, Bremen, Germany). Alternatively, a simpler MALDI-TOF instrument such as the benchtop Microflex (Bruker Daltonics) can be used without loosing data quality. We performed measurements in linear positive ion detection mode, using a Nd:YAG laser at maximum frequency of 66 Hz. Pulsed ion extraction (PIE) was set to zero. Acceleration voltage (IS1) was set to 20 kV. The mass range of spectra was from 2,000 to 20,000 m/z. The final resolution in the mass range of 7,000–10,000 m/z was optimized to be higher than 600 and absolute signal intensities were about 10³. Automated spectrum acquisition was performed using the Auto Execute software with fuzzy control of laser intensity. At least 10⁷ bacterial cells were required for high quality mass spectra. For reference spectra we measured six spots on the MALDI target. On each spot, four spectra with 10 times 100 laser shots were accumulated. Twenty spectra were stored for the reference spectra library. For identification we generally acquired spectra by accumulating 1000 laser shots in ten 100 shot portions.
Factors influencing the intensities of signal peaks comprise concentration and location of proteins in the bacterial cell and biophysical properties of proteins such as solubility, hydrophobicity, basicity, and compatibility with MALDI. In general, most of the proteins detected by MALDI protein bacterial profiling derive from highly abundant, basic ribosomal proteins [11] (link).
Data analysis: Mass spectra were analyzed with Flex Analysis software 2.4 (Bruker Daltonics). Further bacterial data analysis was performed by software developed and tested by us that we termed BioTyper. The mass spectral input data can be listed in generic data formats such as the extensible markup language (XML) to make them independent from the hardware used. Spectra were pre-processed using default parameters for reference spectra libraries that we call main spectra libraries (MSPs). A maximum of 100 peaks with a signal-to-noise (S/N) ratio of 3 were selected in the range of 3,000–15,000 Da. Afterwards the main spectra were generated as a reference using all spectra given for a single microorganism. In general, 75 peaks were picked automatically, which occurred in at least 25% of the spectra and with a mass deviation of 200 ppm.
For the evaluation of mass spectra reproducibility, we loaded the spectra into the ClinProTools 2.1 software (Bruker Daltonics). Through this process mass spectra were firstly normalized before we applied baseline subtraction, peak detection, realignment, and peak-area calculation. The optimal settings resulted in an S/N ratio of 5, a Top Hat baseline subtraction with 10% as the minimal baseline width, and a 3-cycle Savitsky-Golay smoothing with a 10 Da-peak width filter. For the example shown in Figure 3 the coefficient of variation (CV) of each of the individual peak areas was determined; 100 peaks were taken for intra run assessment detected in 18 measurements and 75 peaks for inter run detected in 5 biological replicates. The mean CV for all of the signals from the same replicate sample was calculated to provide a measure of intra- and inter-run reproducibility.
Using the bacterial analysis software (Biotyper) and taking a list of mass signals and their intensities into consideration, dendrograms were generated by similarity scoring of a set of mass spectra. Dendrograms shown in this article had graphical distance values between species constructed from their reference spectra. A correlation function was used for calculating distance values. For graphical correlations an average statistical algorithm was applied as implemented in our software package. The maximal number of top level nodes was set to 2. As mentioned in the Figure legend 1, the arrangement of spectra on the left site of the dendrogram is arbitrary. Species with distance levels under 500 are reliably classified. DNA-based phylogenetic analysis (Figure 2) was done using the Mega (Molecular Evolutionary Genetics Analysis) program, version 3.1 (http://www.megasoftware.net/) [15] .
The complete set of reference spectra compiled in the database of our software package is linked to the NCBI taxonomy database (http://www.ncbi.nlm.nih.gov/Taxonomy/).
For identity scoring, the algorithm implemented in the Biotyper software counted mass signals in experimental spectra that matched with reference spectra and vice versa. Furthermore, the algorithm applied correlates signal intensities of matched signals. Together, three scores obtained from such a procedure are multiplied and normalized to a value of 1000 and then converted in its common logarithm (3). Log scores over 2 indicated a reliable identification of species; log scores over 1.7 generally meant a reliable identification of bacterial genera. Log scores of 3 were obtained when spectra matched with themselves. For the identification of bacterial species, this pattern matching algorithm was routinely applied. For the distinction of highly similar mass spectra of closely related sub species, we used a weighted pattern matching algorithm. In practice, we assigned additional values to informative mass signals that were found in the reference spectra of these sub species. For the application of weighted pattern matching we used the masses and settings listed in Figure 5 and in Table 3. For more details on the BioTyper software the reader is referred to a handbook that is available from the authors as a hardcopy or an electronic version (CD of the complete analysis software package that is freely available for reproducing the results of this study and for testing the procedure shown in this article for additional bacterial genera).
SNP genotyping: Approximately 5 mg of bacterial pellet was re-suspended in 1 ml 0.1% Tween-20 and heated up to 65°C for 15 min. One micoliter was used as template for subsequent PCR. PCR was carried out in 10 µl volume. The PCR buffer consisted of 20 mM (NH₄)₂SO₄, 75 mM Tris-HCl (pH 9.0), 0.01% Tween-20, 2.5 mM MgCl₂, 0.5 M betaine solution, 0.3 mM dNTPs, 1 U conventional Taq polymerase (produced in-house), 0.025 U proofreading Taq polymerase (produced in-house), 0.3 µM forward primer (5′-CGATGACGTGGTGATACTGG-3′), 0.3 µM reverse primer (5′-TCGACTCCCCTACAGCCTTA-3′). After denaturating the PCR samples for 5 minutes at 95°C, amplifications were carried out at 94°C for 30 seconds, 65°C for 30 seconds, and 72°C for 30 seconds for 35 cycles. Finally, the samples were incubated at 72°C for 5 minutes. SNP genotypes were detected by mass spectrometry with the standard GOOD assay in negative ion mode [13] (link), [16] (link) as described in full detail in ref. 13 by using 2 µl of the PCR products generated from 10 µl reactions. The extension primer used for the GOOD assay was 5′-GCGACTTTCTTCGAAGGGG*AC-3′ (* indicates a phosphorothioate linkage). The reference sequences of the galE gene of two E. amylovora strains are shown in Figure S1.

MALDI-TOF MS analysis was performed with a MicroFlex LT mass spectrometer (Bruker Daltonik) as reported [5 (link),6 (link),12 (link)]. Each isolated colony was deposited on a MALDI-TOF MS target Microflex (Bruker Daltonik) as above. Then, each colony was overlaid with 2 μL of matrix solution (saturated solution of alpha-cyano-4-hydroxycinnamic acid in 50% acetonitrile and 2.5% tri-fluoracetic-acid), and the matrix-sample was crystallized by air-drying at room temperature, as previously described [5 (link),6 (link),12 (link)]. Two spots were systematically created for each colony. For each assay, a strain of Escherichia coli (DH5 alpha, Bruker Daltonik) was also analyzed for quality control.
The analyses of the obtained spectra were performed using our personal database [5 (link),6 (link)], the Bruker database updated with a laboratory collection of spectra from clinical isolates identified using molecular sequencing (primarily 16S rRNA sequencing) [6 (link)].
The criteria for identification were previously reported [5 (link)]. An isolate was considered correctly identified by MALDI-TOF MS if both spectra had a score ≥1.9 for species identification or ≥1.7 for genus identification [5 (link)].
Finally, positive controls are used in a routine manner for both MALDI-TOF MS systems. Besides, colonies are systematically tested in duplicate. The E. coli positive controls must be correctly identified and a same identification for the duplicate spots of each colony must be obtained using the score of each software package in order to conclude to the identification of microorganism.

The molecular mass and homogeneity of recombinant Trx-Ac-Var-1 and the active Ac-Var-1 peptide were determined using matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS). For the Trx-Ac-Var-1, a linear time-of-flight mode was performed and for the acid-digested Ac-Var-1 peptide both linear and reflector mode was performed for high resolution using Applied Biosystems/MDS SCIEX 4800 MALDI TOF/TOF™ Analyzer (Waltham, MA, USA). The samples were run in positive detection mode using alpha-cyano-4-hydroxycinnamic acid as the matrix. Internal mass calibration of the instrument with known standards was used.