Corynebacterium pseudotuberculosis
Susan E. Slade1, Luis G. C. Pacheco2,3, Thiago L. P. Castro2, Pablo M. Moraes2, Fernanda A. Dorella2, Anderson Miyoshi2, Natália B. Carvalho2, Sérgio C. Oliveira2, Roberto Meyer3, Martin
Feelisch1, Vasco Azevedo2, Christopher G. Dowson1, and James H. Scrivens1
University, Coventry, U.K., 2Universidade Federal de Minas Gerais, Belo Horizonte, Brazil, 3Universidade Federal da Bahia, Brazil
Relative protein expression of the extracellular proteome of C. pseudotuberculosis wild-type strain 1002 and
∆sigE mutant in response to nitrosative stress
METHODS
Purpose
• Develop a combined approach for characterising the exoproteome of the pathogenic bacterium Corynebacterium
pseudotuberculosis, the etiological agent of caseous lymphadenitis in sheep and goats.
• Compare quantitatively the exoproteomes of two wild type strains of the bacterium, 1002 isolated from goat and C231
from sheep of varying virulence phenotypes.
• Analyse the exoproteome of a mutant strain of C. pseudotuberculosis, deficient in the extracytoplasmic function (ECF)
sigma factor sigE, compared to the wild-type both under normal growth conditions and nitrosative stress caused by nitric
oxide (NO).
Corynebacterium pseudotuberculosis cell culture and exoproteome preparation by three-phase partitioning
Methods
• An optimized protocol of three-phase partitioning was used to obtain the Corynebacterium pseudotuberculosis
exoproteome and combined with data-independent MSE acquisition for comprehensive protein identification and label-free
quantification.
• Nitric oxide stress was induced in mid-exponential phase in the wild type strain 1002 and ∆sigE mutant using
DETANONOate to produce 500 nM NO in steady state. Following 1 hour nitrosative stress in vitro, extracellular proteins
were extracted and characterised quantitatively.
Results
• A total of 93 soluble exoproteins were confidently identified from two wild type C. pseudotuberculosis strains using a
combined three-phase partitioning extraction with LC-MSE methodology. Seventy and sixty seven extracellular proteins
were identified from strains 1002 and C231 respectively with 44 core proteins common to both.
• On average 15 peptides were identified per protein with an average sequence coverage of around 35%.
• Of the proteins identified, 75% were predicted as containing an active exportation signal.
• The exoproteome of the wild type strain 1002 showed little variation as a result of nitrosative stress.
• The ∆sigE mutant exhibited a significantly increased number of exported proteins compared to the wild type and after
treatment with NO.
• One protein annotated as “putative secreted protein” was upregulated over 100-fold in the ∆sigE mutant .
• A putative dioxygenase protein was exclusively identified in the 1002 wild-type strain submitted to nitrosative stress.
This group of proteins in other bacterial species have been shown to be involved in detoxification of extracellular NO.
INTRODUCTION
Corynebacterium pseudotuberculosis is the etiological agent of caseous lymphadenitis (CLA) in sheep and goats
resulting in formation of abscesses in superficial and internal lymph nodes, and in visceral organs, Figure 1. This Grampositive bacterium can successfully infect and reside in phagocytic cells, and thus needs to be able to resist the
adverse conditions found in this intracellular environment such as high nitrosative stress. Despite the important
economic losses caused by this disease worldwide, no effective treatment exists, and the efficacy of the currently
available vaccines and diagnostic methods is still controversial.
Exported proteins from bacterial species can play important roles in the host-pathogen interaction including adhesion,
invasion, damage to host tissues, subversion of the host’s immune response mechanisms and resistance to
environmental stresses during infection. A novel three-phase partitioning (TPP) extraction protocol combined with LCMSE analysis was implemented in this study for characterising the exoproteome of two wild-type strains of C.
pseudotuberculosis isolated from goat (1002) and sheep (C231) that present different virulence phenotypes. This
proteomic strategy allowed for identification of the core exoproteome of this bacterium and permitted the identification
of variant exoproteins that could be associated with the diverse infective potential of the strains.
Bacteria can respond to extracytoplasmic stress through the switching of sigma factors associated with the core RNA
polymerase by alternative sigma factors giving new promoter specificities to the polymerase and transiently activating
the transcription of stress responsive genes. Our method was used to analyse the exoproteome of a mutant strain of C.
pseudotuberculosis deficient in the alternative extracytoplasmic factor (ECF) sigE which exhibits a reduced ability to
survive in the host. The exoproteome of the ∆sigE strain was characterised under normal growth and nitrosative stress
caused by nitric oxide (NO), compared to the wild type 1002 strain. The contribution of this sigma factor to the
regulation of the extracellular proteome of C. pseudotuberculosis in response to stress and virulence of this bacterium
was evaluated.
Overnight cultures (ca. 24 hours) of the C. pseudotuberculosis strains were inoculated (1:100) separately into 500 mL of
pre-warmed chemically defined medium1 and incubated at 37 °C, with agitation until mid-exponential growth phase (OD540
nm = 0.4). When required, nitrosative stress was induced by the addition of 100 µM DETANONOate at OD540 nm = 0.3 and
incubated for a further 1 hour.
At this point, cultures were centrifuged and 400 mL of each supernatant was transferred into new sterile flasks. Protease
Inhibitor Cocktail was added and the supernatants filtered (0.22 μm); ammonium sulphate was added to the samples at
30% (w/v) and the pH of the mixtures reduced to 4.0. Then, n-butanol was added to each sample at an equal volume;
samples were vigorously vortexed and left to rest for 1 h at RT, until the mixtures separated into three phases. The
interfacial precipitate was collected in 1.5 mL microtubes, and re-suspended in 1 mL Tris 20 mM + 10 μL protease
inhibitor. Finally, samples were submitted to diafiltration and buffer exchange with NH4HCO3 (100 mM), using 5 kDa
MWCO spin columns.
Tryptic digestion
Protein samples were resuspended in 1 mL of 0.1% Rapigest (Waters Corporation, Milford, MA) and concentrated using a
5 kDa spin column, heated at 80°C for 15 minutes, reduced with dithiothreitol, alkylated with iodoacetamide and digested
with 1:50 (w/w) sequencing grade trypsin for 16 hours. RapiGest was hydrolysed by the addition of 2 μL of 13 M
trifluoroacetic acid, filtered using a 0.22 μm spin filter and each sample was typically diluted to 1 μg/μL prior to a 1:1
dilution with a 100 fmol/μL glycogen phosphorylase B standard tryptic digest.
NanoLC-MSE data acquisition
Nanoscale LC separations were performed with a nanoACQUITY system (Waters Corporation) using a Symmetry C18
trapping column (180 μm × 20 mm 5 μm) and a BEH C18 analytical column (75 μm × 250 mm 1.7 μm). Each sample
(total digested protein 500 ng) was applied to the trapping column and flushed with 0.1% solvent B for 2 minutes at a flow
rate of 15 μL/min. Sample elution was performed at a flow rate of 250 nL/min by increasing the organic solvent
concentration from 3 to 40% B over 90 min. Solvent A composition was 0.1% formic acid and solvent B acetonitrile
containing 0.1% formic acid (JT Baker).
Accurate mass data were collected in data independent mode of acquisition by alternating the energy applied to the
collision cell/s between a low and elevated energy state (MSE). The spectral acquisition scan rate was typically 0.9 s with
a 0.1 s interscan delay. On the Synapt HDMS instrument in the low energy MS mode, data were collected at constant trap
and transfer collision energies (CE) of 3 eV and 1 eV respectively. In elevated energy MS mode, the trap collision energy
was ramped from 15 eV to 30 eV with the transfer collision energy at 10 eV. On the Ultima Global instrument a low
energy of 6 eV was applied to the collision cell, increasing from 6 eV to 35 eV in elevated MS mode. Human [Glu1]Fibrinopeptide B (doubly charged ion m/z 785.8426) was used for mass correction. All data were acquired in at least
triplicate.
NanoLC-MSE
Proteins identified
10
70
Average peptides/protein
1.9
15
Average sequence coverage
11
35
Single peptide identifications
False discovery rate
6
0
Not determined
0.2%
Figure 2. Image of a 2D gel obtained from the exoproteome of the wild type 1002 strain (left) and a comparison of the
results obtained from quantitative proteomic techniques (right).
Prediction of sub-cellular localisation of the identified proteins
IdentityE
The raw data files were processed using ProteinLynx Global Server™ (PLGS) v2.4 with
and
informatics (Waters) using default parameters for MSE data. The database search parameters used the following variable
modifications, N-terminal acetylation, deamidation of N/Q and oxidation of M residues. The C. psudotuberculosis
database (NCBI Genome Project ID: 40687 and 40875), released in November 2009, to which the glycogen
phosphorylase B and trypsin sequences had been appended was randomised within PLGS generating a new
concatenated database consisting of the original sequences plus one additional randomised sequence for each entry with
identical composition. This database contained a total of 4314 entries. A fixed modification of carbamidomethyl-C was
specified, and variable modifications included were acetyl N-terminus, deamidation N, deamidation Q and oxidation M.
One missed trypsin cleavage site was permitted and was used for all subsequent interrogations.
The protein tables from PLGS IdentityE were compiled in Excel and pivot tables were used to identify proteins observed in
a minimum of 2 replicate analyses from each sample. The protein abundance (as a % of the total loading) was then
calculated for these identifications. The number of random entries in the confident protein table was used to determine
the false positive rate for the analyses. The ExpressionE algorithm was used to determine relative differences in protein
levels between the samples for those proteins observed in at least two technical replicates with a score > 250 and
likelihood of regulation value greater than 0.95 for upregulation and lower than 0.05 for downregulation as determined by
the PLGS quantification algorithm.
Exoproteome analysis of wild type strains of Corynebacterium pseudotuberculosis
The TPP-extracted exoproteomes from strains 1002 (goat) and C231 (sheep) were analysed quantitatively by nanoLCMSE. Proteins were only identified if they were observed in a minimum of two technical replicates. Seventy proteins
were identified from strain 1002, 67 proteins from C231 and 44 proteins were common to both strains giving a total of
93 proteins in total, Figure 3. These results are in good agreement with the observation of 80 visualised spots on a 2D
gel of the exoproteome of strain 1002, Figure 2. The confidence in the results from the nanoLC-MSE analysis was
significantly higher with an average 15 peptides/protein and an average sequence coverage of 35% obtained from
replicate injections containing 500 ng of trypsinised exoproteome. The protein identifications from the 2D gel were
based on three replicate gels, sample loading of 150 µg producing 10 identified proteins, 6 by a single peptide with an
average sequence coverage of 11% and 1.9 peptides/protein, Figure 2. The numbers of corynebacterial exoproteins
identified in the literature vary from as few as 2 in strain R of C. glutamicum2 to 74 in C. diphtheriae C7s(-)tox- .
In addition to permitting us to compare the exoprotein profiles of the different groups, the LC-MSE methodology also
allowed for a comparison of the relative level of expression of the proteins identified in both wild-type 1002 and ∆sigE
strains. There was virtually no alteration in the expression of the proteins commonly found in the extracellular proteomes
of the wild-type C. pseudotuberculosis strain treated by nitric oxide. Whereas in the ∆sigE strain a protein previously
annotated as ‘putative secreted protein’ was up-regulated more than 100-fold following treatment with nitric oxide, Figure
8.
Strain 1002 (goat) ∆sigE
under normal growth
conditions
Figure 4. Venn diagram indicating proteins identified in the wild type strain 1002 and the ∆sigE mutant grown under
normal conditions.
Comparison of the extracellular proteome of C. pseudotuberculosis wild type strain 1002 and ∆sigE mutant in
response to nitrosative stress
Nitrosative stress was induced in the cultures by the addition of 100 µM DETANONOate (500 nM NO) to the growth
medium monitored by an Iso-NO electrode, Figure 5 (left). Cell viability studies indicated that the ∆sigE mutant was more
sensitive to this stress in vitro than the wild type strain 1002, Figure 5 (right).
The exoproteomes of the wild type strain 1002 and its ∆sigE mutant, deficient in the extracytoplasmic sigma factor sigE,
were analysed under normal and nitrosative stress conditions caused by nitric oxide. An increased number of exoproteins
were observed as a result of the exposure to NO, Figure 6 including general stress responses, such as chaperones and
oxidoreductases.
Figure 3 (Left) Venn diagram indicating exoproteins identified in wild type strains 1002 (goat) and C231 (sheep).
Predicted localisation of the identified proteins in this study by the SurfG+ tool (right).
Key: SE=secreted; PSE=potentially surface exposed; C=cytoplasmic; M=membrane; NCS=non-classically secreted.
Exoproteome variation and differential virulence
Wild type strain 1002 (goat)
exposed to nitric oxide
stress
Strain 1002 (goat) ∆sigE
exposed to nitric oxide
stress
Of the 93 exoproteins observed, 49 were only identified in one of the two wild type strains and may account for the
differing virulence phenotypes. Highly variant exoproteomes have been reported for other Gram-positive bacterial
pathogens and such a variation may be considered an important factor leading to the observable phenotypic
dissimilarities.
Figure 6. Venn diagram indicating proteins identified in the wild type strain 1002 and the ∆sigE mutant under 500 nM
nitric oxide stress caused by the addition of DETANONOate for one hour.
Phospholipase D (PLD) is an exotoxin considered as the major virulence factor of C. pseudotuberculosis. It possesses
sphingomyelinase activity that contributes to endothelial permeability and spreading of the bacterium within the host.
Mutation of the pld gene in C. pseudotuberculosis rendered strains no longer capable of causing caseous lymphadenitis in
sheep and goats. PLD was only observed in the more virulent C231 exoproteome in our study.
A putative dioxygenase protein was exclusively identified in the 1002 wild type strain submitted to NO stress, Figure 7.
Dioxygenase proteins have been shown to be involved in detoxification of nitric oxide in other bacterial species making
this protein a good candidate to explain the augmented resistance of this strain to nitrosative stress, compared to the
ΔsigE mutant.
• A three-phase partitioning methodology combined with a nanoLC-MSE quantitative proteomic strategy provided
confident and comprehensive coverage of the extracytoplasmic proteins from Corynebacterium pseudotuberculosis.
• Comparative exoproteome analysis of two wild type strains of different virulence status allowed us to detect
considerable variations of the C. pseudotuberculosis extracellular proteome.
• In our study we did not observe phospholipase D in the cultured wild type 1002 exoproteome and this may contribute
as one of the main factors responsible for the lowered virulence of this strain. Conversely we did observe the FagD
and Cp40 proteins in the C231 strain.
• Several novel targets for future work on molecular determinants of virulence can be identified. Such proteins may
represent promising new candidates for composing a vaccine more effective than the ones currently available.
• A mutant deficient in the alternative extracytoplasmic sigma factor sigE from strain 1002 produced a more diverse
exoproteome both under normal and nitric oxide stress than its wild type counterpart.
• The exoproteome of the 1002 wild type strain showed little variation under NO stress indicating the reproducibility of
the TPP and nanoLC-MSE approach for the study of the exoprotein complement. The observation of a dioxygenase
protein in this strain under NO stress suggests that additional studies are needed to confirm whether the expression of
this species is really dependent on the alternative sigma factor sigE and, most importantly, whether this protein plays
any role in detoxification of extracellular NO in C. pseudotuberculosis.
• It was not possible to identify known conserved domains in the putative secreted protein upregulated over 100-fold in
the ∆sigE mutant under nitrosative stress, thus its function remains elusive. The fact that it is highly expressed in the
mutant strain in response to NO ensures that it warrants further investigation.
• This study is the first to demonstrate that the alternative sigma factor sigE may participate in the specific bacterial
response triggered by NO and its contribution to resistance to nitrosative stress in vitro and in vivo and we propose a
model for its participation in the alteration of C. pseudotuberculosis exoproteome in response to nitric oxide. After
entering a host cell C. pseudotuberculosis is exposed to nitrosative stress generated by nitric oxide synthase (iNOS)
within the phagosomal environment. Then, the wild-type strain of this bacterium secretes a few additional proteins that
include a putative dioxygenase, which might aid in detoxification of intraphagosomal nitric oxide. The ΔsigE mutant
strain in turn, secretes more proteins normally involved in general stress responses, indicating a compensatory
response. A highly up-regulated protein (indicated in red) is secreted only in the mutant strain following NO exposure,
but its function is still unknown.
Figure 9.
Proposed model for the participation of
the alternative extracytoplasmic function
sigma factor sigE in the alteration of C.
pseudotuberculosis
exoproteome
in
response to nitric oxide.
Proteins believed to be associated with the virulence of C. pseudotuberculosis were also identified exclusively in the
exoproteome of the C231 strain, namely FagD and Cp40. FagD is a component of an iron uptake system, whose coding
sequences are clustered immediately downstream of the pld gene in the C. pseudotuberculosis genome. Cp40 is a
secreted serine protease shown to be protective against CLA when used to vaccinate sheep.
Comparison of the extracellular proteome of C. pseudotuberculosis wild type strain 1002 and ∆sigE mutant under
normal growth conditions
The TPP-extracted exoproteome from strain 1002, deficient in the extracytoplasmic sigma factor SigE was analysed by
nanoLC-MSE and compared to the control wild type strain. An increased number of proteins were observed in the mutant,
a total of 87 proteins, of which 63 were common to the wild type strain. Twenty four proteins were only observed in the
∆sigE mutant indicating that there is a differential response in the exoproteome under normal growth conditions due to the
absence of the transcriptional regulator sigE, Figure 4.
Strain 1002 (goat) ∆sigE
CONCLUSIONS
Figure 5. The amount of NO released after addition of 100 µM DETANONOate to the growth medium (CDM) was
monitored by an Iso-NO electrode (World Precision Instruments, Inc.), at 37°C (left). Cell viabilities were evaluated by
colony-forming units expressed as percentage survival, compared to a non-treated control (right).
Wild-type strain
C231 (sheep)
Strain 1002 (goat) wild-type
Figure 8. Relative protein expression changes in the exoproteome of wild-type strain 1002 (left) and ∆sigE mutant
(right) under nitrosative stress plotted on a natural log scale.
Wild-type strain
1002 (goat)
ExpressionE
Confident identifications, quantitation and comparative protein expression
Wild type strain 1002 (goat)
under normal growth
conditions
The SurfG+ tool3 was used to classify the identified C. pseudotuberculosis proteins from nanoLC-MSE analysis into four
categories; secreted, potentially surface exposed (PSE), membrane and cytoplasmic. The software brings together the
predictions of global protein localisations performed by well known algorithms, and innovates by allowing for an accurate
prediction of PSE proteins. For the proteins identified in both wild type strains, 75% (70) were predicted to be actively
secreted or PSE which constitutes 50% and 15% of the predicted secreted and PSE proteins from the genome
respectively.
Data processing and database interrogation
RESULTS
Figure 1 Corynebacterium pseudotuberculosis cultured on a blood agar plate (left), internal abscesses typical of
visceral caseous lymphadenitis (middle) and goat showing symptoms of superficial caseous lymphadenitis (right).
2D-GE MALDITOF/TOF
Loge ratio nitric oxide stress:control
OVERVIEW
Loge ratio nitric oxide stress:control
1Warwick
REFERENCES
Figure 7. Three-dimensional structure of the C. pseudotuberculosis ‘putative dioxygenase’, predicted by homology
modeling using SWISS-MODEL (left) and structure-based prediction of protein function, performed using ProFunc
(right) identified in the wild-type strain under nitrosative stress.
1Moura-Costa
LF, Paule BJA, Azevedo V, Freire SM, Nascimento I, Schaer R, Regis LF, Vale VLC, Matos DP, Bahia RC, Carminati R, Meyer R
Chemically defined synthetic medium for Corynebacterium pseudotuberculosis culture. Rev. Bras. Saúde e Produção Animal 2002, 3:1-9.
2Suzuki N, Watanabe K, Okibe N, Tsuchida Y, Inui M, Yukawa H
Identification of new secreted proteins and secretion of heterologous amylase by C. glutamicum. Appl Microbiol Biotechnol 2009, 82:491-500.
3Barinov A, Loux V, Hammani A, Nicolas P, Langella P, Ehrlich D, Maguin E, van de Guchte M
Prediction of surface exposed proteins in Streptococcus pyogenes, with a potential application to other Gram-positive bacteria. Proteomics 2009, 9:61-73.