Transcriptome analysis of Burkholderia pseudomallei T6SS identifies Hcp1 as a potential serodiagnostic marker

Microbial Pathogenesis 79 (2015) 47e56

Contents lists available at ScienceDirect

Microbial Pathogenesis journal homepage: www.elsevier.com/locate/micpath

Transcriptome analysis of Burkholderia pseudomallei T6SS identifies Hcp1 as a potential serodiagnostic marker Sylvia Chieng 1, Rahmah Mohamed, Sheila Nathan* School of Biosciences and Biotechnology, Faculty of Science and Technology, Universiti Kebangsaan Malaysia, 43600, Bangi, Selangor, Malaysia

a r t i c l e i n f o

a b s t r a c t

Article history: Received 11 September 2014 Received in revised form 16 January 2015 Accepted 19 January 2015 Available online 20 January 2015

Burkholderia pseudomallei, the causative agent of melioidosis, is able to survive extreme environments and utilizes various virulence factors for survival and pathogenicity. To compete and survive within these different ecological niches, B. pseudomallei has evolved specialized pathways, including the Type VI secretion systems (T6SSs), that have a role in pathogenesis as well as interbacterial interactions. We examined the expression profile of B. pseudomallei T6SS six gene clusters during infection of U937 macrophage cells. T6SS-5 was robustly transcribed while the other five clusters were not significantly regulated proposing the utility of T6SS-5 as a potential biomarker of exposure to B. pseudomallei. Transcription of T6SS regulators VirAG and BprB was also not significant during infection when compared to bacteria grown in culture. Guided by these findings, three highly expressed T6SS genes, tssJ4, hcp1 and tssE-5, were expressed as recombinant proteins and screened against melioidosis patient sera by western analysis and ELISA. Only Hcp1 was reactive by both types of analysis. The recombinant Hcp1 protein was further evaluated against a cohort of melioidosis patients (n ¼ 32) and non-melioidosis individuals (n ¼ 20) sera and the data clearly indicates a higher sensitivity (93.7%) and specificity (100%) for Hcp1 compared to bacterial lysate. The detection of anti-Hcp1 antibodies in patients' sera indicating the presence of B. pseudomallei highlights the potential of Hcp1 to be further developed as a serodiagnostic marker for melioidosis. © 2015 Elsevier Ltd. All rights reserved.

Keywords: Burkholderia pseudomallei Melioidosis T6SS Hcp1

1. Introduction Melioidosis, a disease endemic to South East Asia and Northern Australia, is potentially fatal and accounts for a high percentage of all community-acquired septicaemia [1]. The chronic or acute form of the disease is characterized by the formation of abscesses, pneumonia, and septicaemia as well as multiple organ and soft tissue infection [2,3]. This broad range of clinical manifestations in melioidosis cases is similar to other bacterial diseases and frequently results in diagnostic errors [4]. Currently, isolation of the causative agent, Burkholderia pseudomallei, from bodily fluid or samples remains the gold standard in diagnosis and requires trained laboratory personnel and the use of selective media. Its main drawback is that this approach is time consuming and proves

* Corresponding author. E-mail address: [email protected] (S. Nathan). 1 Current address: Faculty of Applied Science and Foundation Studies, Infrastructure University Kuala Lumpur, Unipark Suria, Jalan Ikram-Uniten, 43000, Kajang, Selangor, Malaysia. http://dx.doi.org/10.1016/j.micpath.2015.01.006 0882-4010/© 2015 Elsevier Ltd. All rights reserved.

to be too late for successful disease management as a high percentage of patients admitted with acute septicaemia die within 24e48 h after admission [5,6]. B. pseudomallei is a Gram-negative bacillus commonly found as an environmental saprophyte that survives in wet soil, agriculture land, ponds, rivers and paddy fields in most tropical countries [7]. This ability to adapt to different ecological niches, including humans and animals, is attributed to its capacity to utilize various carbon sources [1,8]. Pilatz et al. [9] successfully identified genes required for the intracellular life cycle and in vivo virulence of this bacterium. Mutants of the type III secretion system (T3SS), type VI secretion system (T6SS) and biosynthesis pathways display reduced intracellular survival and attenuated virulence. In addition, through an in vivo expression technology (IVET)-based approach, a T6SS locus was induced upon invasion of macrophage cells [10]. Putative type VI secretion system (T6SS)-encoding gene clusters are widely dispersed in almost a quarter of sequenced Gramnegative bacteria genomes of which, many are known pathogens and a role in pathogenesis has been attributed to the T6SS [11e13]. Approximately 15 core genes and a variable number of non-

48

S. Chieng et al. / Microbial Pathogenesis 79 (2015) 47e56

conserved accessory elements constitute these gene clusters [14]. Functional assays and protein localization studies suggest that these proteins assemble into a multi-component secretory apparatus similar to a bacteriophage-like structure, injecting effector proteins into eukaryotic target host cells [11,15,16]. In addition, the T6SS also contains antibacterial properties against competitor bacterial cells upon cell-to-cell contact [17,18]. In B. pseudomallei, six clusters of T6SS (T6SS-1, -2, -3, -4, -5 and -6) have been described [10] although experimental evidence is only available for T6SS-5. Preliminary studies suggest that T6SS-5 is required for B. pseudomallei pathogenesis [9,10,19]. A similar conclusion was reported for Burkholderia mallei where strains lacking the homologous T6SS-5 gene cluster were greatly attenuated in a hamster model of glanders [20]. Moreover, T6SS-5 is positively regulated by VirAG during infection of RAW264.7 cells [21] and negatively regulated by iron and zinc during in vitro growth [22]. Currently, the regulation or expression of the other five clusters is unknown. In the present study, we sought to profile the gene expression pattern of all six B. pseudomallei T6SS gene clusters during intracellular infection of macrophage cells. Based on the in vivo expression profile, selected components of this secretion system were further characterised as potential serodiagnostic markers for melioidosis. 2. Materials and methods 2.1. Bacterial strain, culture conditions and genomic DNA extraction B. pseudomallei D286 clinical isolate [23] was grown on Ashdown agar for 48 h prior to culturing in LuriaeBertani (LB) broth overnight at 37  C. Overnight cultures were diluted 1:50 in LB broth and grown to mid-logarithmic phase (OD600 ¼ 0.4e0.6) at 37  C with shaking at 250 rpm for 3 h. Genomic DNA extraction was performed as previously described [24]. The concentration and quality of the genomic DNA was determined using the Nanodrop™ ND-1000 spectrophotometer (Nanodrop Technologies, USA). 2.2. Cell culture and infection of macrophages Human monocyte-like U937 cells (CRL-1593.2) were maintained in RPMI1640 medium (Gibco) supplemented with 10% fetal bovine serum (Hyclone), 2 mM L-glutamine, 10 mM HEPES and 1 mM sodium pyruvate (Invitrogen). Infection of U937 macrophage cells with B. pseudomallei D286 was performed as previously described [25]. 2.3. RNA isolation and purification RNA from intracellular bacteria and control bacteria grown in RPMI1640 medium was harvested as previously described [25] at 1, 2, 4 and 6 h post-infection (hpi). Briefly, intracellular bacteria were liberated from infected macrophages using 1% saponin in PBS and subjected to differential centrifugation to pellet bacterial cells. Bacterial RNA was then extracted using the Qiagen RNeasy Mini Kit and on-column DNase I digestion was performed. Total RNA obtained was further purified by ethanol/ammonium acetate precipitation. The concentration, quality and integrity of RNA isolated were analysed using the Nanodrop® ND-1000 and Agilent 2100 Bioanalyser. 2.4. Microarray and data analysis

hybridization were done according to the Agilent one-colour microarray protocols (available at http://www.chem.agilent.com/ Library/usermanuals/Public/G4140-90040_GeneExpression_Onecolor_v6.5.pdf). Arrays were scanned with the Agilent Technologies Scanner model G2505B and spot intensities and other quality control features were extracted with Agilent's Feature Extraction Software version 9.5.3.1. 2.5. Sera specimens Thirty two culture-confirmed melioidosis archived sera samples were obtained from the Institute for Medical Research (IMR), Kuala Lumpur (n ¼ 15) and Kelantan Hospital, Kota Bharu (n ¼ 17), Malaysia. Sera samples (n ¼ 20) from healthy and confirmed HIVnegative university student volunteers as well as CliniPath Laboratory Malaysia Sdn. Bhd were used as the negative control. 2.6. Amplification and cloning of recombinant proteins Three T6SS cluster genes were selected for the production of recombinant proteins. Primer pairs (Table 1) used for the amplification of tssJ-4, hcp1 and tssE-5 from B. pseudomallei D286 genomic DNA were designed based on the B. pseudomallei K96243 annotated ORFs using Primer Premier 4.0. The forward primers contained an additional CACC sequence at the 5’-end to allow direct cloning into the expression vector. Amplification of these genes was carried out using the Expand High Fidelity PCR System (Roche, Germany) under the following conditions: an initial denaturation step at 95  C for 5 min, followed by 30 cycles of denaturation at 95  C for 1 min, annealing at 55.3  C (tssJ-4), 53.2  C (hcp1) or 55.5  C (tssE-5) for 1 min, and extension at 72  C for 1 min. Purified PCR products were cloned into pET 200/D-TOPO® according to the manufacturer's recommendation (Invitrogen, USA). Gene orientations and sequences were verified by routine DNA sequencing. 2.7. BLAST analysis and multiple sequence alignment BLAST analysis was conducted on the tssJ-4, hcp1 and tssE-5 predicted amino acid sequences against the non-redundant protein database of sequenced B. pseudomallei strains (taxid: 28450), B. mallei strains (taxid: 13373), Burkholderia thailandensis strains (taxid: 57975) and other species commonly associated with septicaemia or bacteria that cause disease similar to melioidosis. Multiple sequence alignment of the associated sequences was performed using ClustalX2 [26]. Protein sequences used for the alignment of TssJ-4 were from B. mallei ATCC23344 (YP_105227), B. pseudomallei 1026b (YP_006277247), B. pseudomallei 1710b (YP_337238), B. pseudomallei K96243 (YP_110550), B. thailandensis E264 (YP_440082) and B. thailandensis MSMB121 (YP_007921798). For the alignment of Hcp1, B. mallei NCTC10247 (YP_001078858), B. mallei NCTC10229 (YP_001024532), B. pseudomallei 1026b (YP_006278227), B. pseudomallei 1106a (YP_001076060), B. pseudomallei K96243 (YP_111505), B. pseudomallei 1710b (YP_335694), B. pseudomallei 668 (YP_001063119), B. thailandensis

Table 1 List of primers used for amplification. Gene

Primer

Nucleotide sequence (50 e30 )

tssJ-4

tssJ-4_F tssJ-4_R hcp1_F hcp1_R tssE-5_F tssE-5_R

CACCTCGTCGAAGCCGGAAG TCATTGCCTGCTTCTCCTTG CACCATGCTGGCCGGAATATA TCAGCCATTCGTCCAGTTTG CACCATGGCTGATCGCGAAT TCAGGAAATCGTTCGGATATCG

hcp1

Sample labelling and hybridization, scanning and data analysis were performed as previously described [25]. Briefly, bacterial RNA was polyadenylated and reverse transcribed whilst labelling and

tssE-5

S. Chieng et al. / Microbial Pathogenesis 79 (2015) 47e56

E264 (YP_439065), Klebsiella pneumoniae 342 (YP_002238890) and Pseudomonas aeruginosa PAO1 (NP_248954) were compared. For the alignment of TssE-5, B. thailandensis E264 (YP_439064), B. thailandensis MSMB121 (YP_007920446), B. pseudomallei K96243 (YP_111506), B. pseudomallei 1710b (YP_335696), B. pseudomallei 668 (YP_001063120), B. mallei NCTC10247 (YP_001078859), B. mallei ATCC23344 (YP_338390) and K. pneumoniae 342 (YP_002237895) sequences were utilised.

49

western blot analysis described above. The P/N ratio was determined by dividing the average absorbance at 405 nm of the positive samples by the average absorbance of the negative samples. A positive result was obtained when the P/N ratio was 3.0. A statistical analysis on the reactivity of recombinant proteins was performed using analysis of variance (ANOVA) test within the GraphPad Prism® version 4.0 (GraphPad Software) software package.

2.8. Expression and purification of recombinant proteins The expression host Escherichia coli BL21 (DE3) Star™ (Invitrogen, USA) was transformed with the individual recombinant clones according to the manufacturer's protocol. For overexpression of recombinant proteins, 1 L LB broth containing 50 mg/ml kanamycin was inoculated with overnight cultures at 1:100 and grown at 37  C with shaking (250 rpm) until an OD600nm reading of 0.4e0.6 was achieved. Protein expression was induced with 1 mM isopropyl b-D-1-thiogalactopyranoside (IPTG) and incubation was continued for 5 h. Expression cultures were harvested and the bacterial pellets obtained were resuspended in BugBuster®Protein Extraction Reagent (Novagen®, USA). The suspension was incubated on a rotating platform at room temperature for 20 min and then centrifuged at 12,000  g for 20 min at 4  C. The supernatant was collected and the residual cell pellet was resuspended in binding buffer with 6 M urea and incubated at 4  C overnight for extraction of inclusion body protein. The lysate was centrifuged at 12,000  g for 30 min at 4  C and the supernatant was collected. Both soluble and inclusion body fractions were purified by affinity chromatography using a HisTrap HP 1 mL column € on the AKTApurifier (GE Healthcare, Sweden) according to the manufacturer's recommended protocol. Eluted fractions containing the purified protein of interest were pooled, desalted and concentrated using an Amicon Centricon 10 kDa (Millipore, USA). The purified proteins were analysed by SDS-PAGE and western blot against pooled melioidosis human sera (n ¼ 5). 2.9. ELISA A 96-well microtitre plate (Greiner Bio One) was coated with each purified antigen (0.25 mg/well) in coating buffer (0.1 M sodium carbonate, pH 9.6) in a final volume of 50 ml and incubated at 4  C overnight. Following the overnight incubation, wells were washed 5 times with PBS-Tween 0.1% and blocked with 100 ml blocking buffer (5% skimmed milk in PBS) for 1 h at 37  C. After 10 washes with 0.1% PBST, 50 ml sera diluted in blocking buffer (1:1600) was added to each well. Each sera sample was analysed in triplicate. The plate was incubated at 37  C for 1 h and washed as described above. Anti-human IgG-conjugated with horseradish peroxidase (Sigma Chemical Co., USA) (50 ml of a 1:10,000 dilution) was added to the wells and incubation was continued for 1 h at 37  C. Wells were then washed 10 times with 0.1% PBST and the enzyme reaction was detected by the addition of 100 ml ABTS® Peroxidase substrate (KPL, USA) for 30 min at room temperature. Absorption at 405 nm was determined using the Sunrise™ Microplate Reader (Tecan Group Ltd., Switzerland). The mean and standard deviations of repeated measurements were calculated for each tested sample. 2.10. Determination of recombinant protein reactivity Positive-to-negative absorbance ratios (P/N) were used to determine the reactivity of the recombinant proteins towards the sera samples [27]. The proteins were evaluated with the five sera samples from melioidosis patients (positive samples) and five sera samples from healthy blood donors (negative samples) used in the

Fig. 1. Gene expression profile of B. pseudomallei T6SS gene clusters. The heat-maps show the expression profile of intracellular B. pseudomallei within U937 cells for 1e6 h relative to in vitro grown bacteria. Expression values are determined from the SAM analysis with red representing up-regulation (ratio of þ5.0) and green representing down-regulation (ratio of 5.0) on a log2 scale.

50

S. Chieng et al. / Microbial Pathogenesis 79 (2015) 47e56

2.11. Determination of ELISA specificity and sensitivity A total of 32 human melioidosis sera and 20 healthy blood donor sera samples (negative controls) were analysed to determine specificity and sensitivity of the antigens. When the average absorbance at 405 nm of the tested sample was greater than that of the negative controls plus 2 standard deviations, the tested sample was considered positive for B. pseudomallei antigen specific antibodies [28]. 2.12. Statistical analysis Statistical analysis on human sera reactivity was performed using the Mann Whitney test within the GraphPad Prism® version 4.0 (GraphPad Software) software package. 3. Results 3.1. T6SS expression during intracellular infection In our earlier study, we reported on the transcriptome profile of intracellular B. pseudomallei during infection of U937 macrophage cells [25]. Guided by those findings, in this study, we determined the expression of the T6SS gene clusters during an active infection of U937 macrophages. Hierarchical clustering of gene expression levels revealed two different patterns of gene regulation over the period of 1e6 h post-infection (hpi) (Fig. 1). The majority of the genes involved in the T6SS cluster were down-regulated throughout the infection period compared to bacteria grown in control medium. Fluctuating expression levels were observed in all clusters except for T6SS-5. Genes from cluster T6SS-1 (BPSL3111BPSL3097), T6SS-2 (BPSS0095-BPSS0116), T6SS-3 (BPSS0185BPSS0167), T6SS-4 (BPSS0515-BPSS0533) and T6SS-6 (BPSS2093BPSS2109) were down-regulated at 1 hpi and increased to a level

Fig. 3. Relative expression ratio of T6SS regulators. The horizontal axis represents fold change in log2 scale obtained from microarray analysis. Data are mean ± SD of triplicate measurements from three independent experiments.

lower or similar to control bacteria at 2 hpi before decreasing again at 4 hpi. At 6 hpi, most of the genes were not significantly expressed. This expression profile is similar to the overall expression profile of the B. pseudomallei genome during intracellular infection [25], indicating a rapid adaptation of the bacterium to the intracellular milieu. This observation also suggests the dispensability of these five T6SS gene clusters during intracellular infection

Fig. 2. Regulation of the T6SS-5 gene cluster. The heat-map shows the expression ratio of each gene in the T6SS-5 gene cluster during intracellular growth in host macrophages relative to in vitro growth. Expression values are determined from the SAM analysis with red representing up-regulation (ratio of þ5.0) and green representing down-regulation (ratio of 5.0) on a log2 scale. Classification of COGs and description of each gene component of T6SS-5 is included.

Fig. 4. Multiple sequence alignment of B. pseudomallei D286 T6SS predicted amino acid sequences with protein sequences from Burkholderia spp. and other pathogens. (A) TssJ-4 (B) Hcp1 (C) TssE-5. BM e B. mallei; BP e B. pseudomallei; BT e B. thailandensis; KP e Klebsiella pneumoniae; PA e Pseudomonas aeruginosa.

52

S. Chieng et al. / Microbial Pathogenesis 79 (2015) 47e56

Fig. 6. Recombinant protein reactivity towards melioidosis patient sera. (A) Western blot analysis with pooled melioidosis patient sera (n ¼ 5). (B) Western blot analysis with pooled normal sera (n ¼ 5). Lane 1: TssJ-4; 2: Hcp1; 3: TssE-5; 4: B. pseudomallei D286 lysate.

Fig. 5. Expression of B. pseudomallei T6SS recombinant proteins within soluble and inclusion body fractions in E. coli BL21 (DE3) Star™. (A) SDS-PAGE analysis of expression products stained with Coomassie Blue. (B) Western blot analysis with antiHis antibody. Lane 1: TssJ-4; 2: Hcp1; 3: TssE-5. Proteins expressed from pET200/DTOPO® have an extra ~4.1 kDa corresponding to the N-terminal His-tag.

or survival. In contrast, a prominent expression profile was obtained for T6SS-5 (BPSS1493-BPSS1511). Most of the genes in this cluster were up-regulated as early as 1 hpi and peaked at 6 hpi (Fig. 2). Compared to bacteria grown in culture medium, expression levels of major genes within T6SS-5 such as iglA (BPSS1496) and its secreted effector proteins hcp1 (BPSS1498) and vgrG (BPSS1503) were all significantly higher in intracellular bacteria harvested from infected U937 macrophages. Additionally, genes encoding lysozyme, chaperone, lipoprotein and hypothetical proteins were also expressed at high levels throughout the infection period. However, we observed that the transcript levels of the T6SS-5 response regulator, virAG (BPSS1494-1495), were similar to control bacteria. These data suggest that T6SS-5 is activated by the host signals during infection as previously proposed by Chen et al. [21]. We further explored the expression levels of a different regulator, bprC.

BprC is an AraC regulator within the T3SS-3 cluster and controlled the expression of T6SS-5 in bacteria grown in culture medium [29]. We found that it was also not significantly expressed when compared to control bacteria (Fig. 3), suggesting that BprC might not play a role in regulating T6SS-5 during infection. Interestingly, we noticed the elevated transcript levels of bprB (BPSS1522) and bprD (BPSS1521) in intracellular B. pseudomallei. BprB is predicted to be a regulator of the T3SS-3 cluster [30], while BprD shares weak homology with a sigma-54-dependent transcriptional activator [31]. The data propose that these regulators may encompass a complex regulatory system in controlling T6SS-5 expression in an intracellular niche.

3.2. Production of recombinant proteins Based on the expression levels of all T6SS gene clusters throughout the infection period, we selected three highly expressed T6SS genes i.e. tssJ-4 (BPSS0529), tssD-5/hcp1 (BPSS1498) and tssE-5 (BPSS1499) for further characterization. Whilst both hcp1 and tssE-5 are members of the T6SS-5 cluster, tssJ4 was selected, as its transcript level was similar to the T6SS-5 gene cluster. A BLASTP analysis on these amino acid sequences showed a high degree of identity to Burkholderia spp. (80e100% identity),

S. Chieng et al. / Microbial Pathogenesis 79 (2015) 47e56

particularly to B. pseudomallei (95e100%) and its homologues in B. mallei (98e99%) (Fig. 4), displaying a high degree of sequence conservation among Burkholderia family. On the contrary, no identity or low identity (10e60%) was observed with proteins from other bacterial pathogens regularly associated with septicaemia such as P. aeruginosa and K. pneumoniae. Of note was the low identity with proteins from Mycobacterium tuberculosis and Legionella pneumophila, two bacterial pathogens that cause disease symptoms akin to melioidosis. The complete open reading frames (ORFs) of tssJ-4 (GenBank accession number JF798394), hcp1 (JF798396), and tssE-5 (JF798397) were successfully amplified. SignalP 3.0 analysis of the amplified ORF sequences indicated the presence of a signal peptide sequence and possible cleavage site in TssJ-4 at positions 27e28 (probability of 0.674). Thus, a second round of amplification was conducted to exclude the signal sequence. All three proteins were over-expressed as His-tagged recombinant proteins in E. coli BL21 (DE3) Star™ and their predicted molecular weights were confirmed by SDS-PAGE (Fig. 5A) and western analysis with mouse anti-His antibody (Fig. 5B). Both analyses confirmed the expression of each recombinant protein as either soluble or inclusion body fractions with the corresponding molecular weights of His-tagged proteins as follows: TssJ-4, 22.5 kDa; Hcp1, 22.9 kDa and TssE-5, 21 kDa.

53

Fig. 7. Evaluation of T6SS recombinant proteins antigenicity by ELISA. The bars indicate absorbance values at 405 nm for each recombinant protein tested against melioidosis patients' sera (melioidosis positive), and negative sera (melioidosis negative). Detection of anti-Hcp1 antibodies in positive sera was significant. Values represent the mean of triplicate measurements and error bars indicate standard deviations (SD).

3.4. Hcp1 as a potential serodiagnostic marker

3.3. Recombinant proteins demonstrate diverse immuno-reactivity As the level of conservation between these three selected proteins with other Gram-negative non-Burkholderia spp. pathogens was low, we proposed that these recombinant proteins could be evaluated as serodiagnostic markers for melioidosis. The proteins could be utilised as antigens to detect the presence of antibodies towards these proteins in suspected melioidosis sera; a positive reaction would be indicative of the presence of the bacteria. To validate this proposal, the proteins were assessed with human melioidosis and control sera samples through western blot and ELISA. The western blot screen of these antigens against pooled sera from five melioidosis patients demonstrated that only Hcp1 was immuno-reactive against sera from B. pseudomallei cultureconfirmed infected individuals (Fig. 6A). None of the recombinant proteins were reactive towards pooled negative (control) sera (Fig. 6B), indicating a reactivity specific to sera from melioidosis culture-confirmed individuals. A further evaluation using ELISA was carried out to confirm the western screen and to determine the P/N value for each recombinant protein. ELISA was performed using pooled melioidosis sera and B. pseudomallei D286 lysate to define the optimal antigen concentration and sera dilution. As shown in Table 2, Hcp1 demonstrated the highest P/N value (P/N > 27) compared to the other two recombinant antigens (P/N  1.2). Screening of individual sera samples by ELISA demonstrated the potential of Hcp1 as a serodiagnostic marker for B. pseudomallei infection (Fig. 7). Significantly, low-level detection of anti-TssD-5/ Hcp1 antibodies in melioidosis negative sera confirmed the specificity of these screens. Table 2 P/N value of recombinant proteins based on ELISA data. Recombinant protein

P/N value (average A405nm positive sample/average A405nm negative sample)

TssJ-4 Hcp1 TssE-5

0.98 27.74 1.06

The specificity and sensitivity of Hcp1 as a serodiagnostic marker was analysed with an additional 32 melioidosis cultureconfirmed sera samples and 20 healthy blood donor sera. To compare the efficiency of each antigen in detecting the presence of antibodies towards B. pseudomallei proteins, the B. pseudomallei D286 lysate was also included in the screen. The comparative ELISA values for each antigen tested in an ELISA-IgG based screen are shown in Fig. 8. For Hcp1, 30 out of 32 culture-confirmed melioidosis sera samples tested positive, giving a sensitivity of 93.7% and specificity of 100% (Table 3). The sensitivity of D286 bacterial lysate was relatively low at 65.6%, wherein 21 out of 32 samples tested positive although specificity of the lysate was high at 90%. 4. Discussion The Mekalanos group first described the Type VI secretion system in detail in 2006. The seminal study reported on the extracellular export of Vibrio cholerae Hcp and three related VgrG proteins needed for contact-dependent cytotoxicity in amoebae and macrophages [32]. These secreted proteins lacked signal sequences and did not display any similarity to the type III or type IV secretion systems, leading the authors to coin the term VAS (virulence-associated secretion) for the genes and suggested that they encoded a prototypical type VI secretion system. From the discovery of this new secretion system, many more reports describing the existence of T6SS in Gram-negative pathogenic bacteria have emerged (reviewed in Bingle et al. [11] and Filloux et al. [33]). A genome wide in silico analysis revealed that the T6SSs are encoded by large variable gene clusters found mainly in Proteobacteria [14]. This system is composed of core proteins conserved in both pathogenic and non-pathogenic bacteria. In addition to the core genes, a set of conserved accessory proteins, for example structural components, regulatory proteins, chaperones and transcriptional activators were identified in the T6SS gene clusters [14]. Shalom and colleagues discovered a total of six T6SS gene clusters in B. pseudomallei [10]. Four of these gene clusters are conserved among B. pseudomallei, B. mallei and B. thailandensis. A phylogenetic analysis of these T6SSs demonstrated that T6SS-5, the only Burkholderia system previously associated with virulence, clustered within the subtree of eukaryotic cell-targeting systems,

54

S. Chieng et al. / Microbial Pathogenesis 79 (2015) 47e56

Fig. 8. Sensitivity and specificity of Hcp1 as determined by ELISA. (A) ELISA assay using recombinant Hcp1. (B) ELISA assay using B. pseudomallei D286 whole cell lysate. Each symbol represents the average of triplicate readings of one serum sample. The cut-off value was determined based on the mean of 20 sera samples from healthy controls þ2 SD against each antigen. The dashed line indicates cut-off values. A dilution of 1:1600 was used for all sera samples. Significance determined using the Mann Whitney test (***P-value < 0.0001).

Table 3 Serological results for melioidosis confirmed and control sera using ELISA. Protein

Cutoff

Sensitivity (%) (no. positive/ total no. of melioidosis confirmed sera)

Hcp1 0.097 93.7 (30/32) B. pseudomallei 0.168 65.6 (21/32) lysate

Specificity (%) (no. positive/total no. of control sera) 100 (0/20) 90 (2/20)

while four of the remaining systems (T6SS-2, 3, 4 and 6) clustered within the bacteria-specific systems subtree, and the final system (T6SS-1) was found in a neighbouring subtree [18]. Moreover, mutants lacking the T6SS-5 were attenuated in a mice model of acute pneumonia of melioidosis, while mice infected with wild type or mutants bearing deletions of the other T6SSs succumbed by the third day post-infection. This observation demonstrated that T6SS-5 is required for virulence, while the other systems are dispensable. In our study, we noted that T6SS-1, -2, -3, -4 and -6 were not significantly expressed in intracellular B. pseudomallei, whereas the transcript levels of T6SS-5 were greatly elevated throughout the infection period, similar to that previously described by Schwarz et al. [18]. T6SS-5 is known to be induced upon invasion of macrophages or direct contact with host cells and is highly expressed during infection but not when grown in culture medium [10,21]. Transcript levels of the secreted proteins Hcp1 and VgrG were significantly high during infection (p < 0.01). Hcp1 and VgrG were up-regulated as high as 182-fold and 27-fold, respectively, at 6 h post-infection relative to control bacteria. These proteins exhibit co-dependency for export and together constitute the extracellular portion of the T6SS apparatus [15,17]. Both Hcp1 and VgrG form a bacteriophage-like structure and given the relative position of these proteins, it is speculated that Hcp1 polymerization is triggered by VgrG recruitment to the apparatus [16]. This concurs with our observation that both genes are highly expressed during an active infection confirming the mutual dependence for optimal function. Hcp1 was first identified in P. aeruginosa [34] and homologues of Hcp1 can be found in various bacteria including several pathogenic bacteria [14,35]. In the closely related species, B. mallei, Hcp1 is reportedly expressed in vivo during infection. Mutation of the Hcp1 encoding gene BMAA0742 led to attenuation in a hamster model of infection [20]. Additionally, an hcp mutant of the pathogenic strain E. coli ExPEC was shown to display reduced attachment, invasiveness and actin rearrangement of epithelial cells [35,36]. In B. pseudomallei, Burtnick et al. [19] demonstrated

that the Hcp1 mutant is attenuated in a Syrian hamster infection model and exhibits reduced cytotoxicity in RAW 264.7 mouse macrophages when compared to the wild type. The growth of this mutant within macrophage cells was also retarded and the mutant failed to induce the formation of multi-nucleated giant cells (MNGCs) [19]. Like other virulence factors, the expression of T6SS is tightly regulated for the benefit of the pathogen. As extensively reviewed by Bernard et al. [16] and Silverman et al. [37], the regulation of T6SS depends on both environmental signals and transcriptional regulation. Chen and colleagues [21] demonstrated that the expression of B. pseudomallei T6SS-5 within host cells was dependent on the two-component sensor-regulator VirAG, whilst under in vitro conditions, expression was completely dependent on BprC. Both virAG and bprC are critical for T6SS-5 function in macrophages and mutants of these genes were avirulent in a mouse model of infection, indicating the absolute dependence of T6SS-5 expression on these regulators in vivo. In our study, we observed that both regulators were not significantly expressed in the intracellular environment. The similar transcript levels observed in both intracellular and control bacteria suggests that the bacteria were able to rapidly adapt to the intracellular niche. Interestingly, elevated expression levels for bprB and bprD during B. pseudomallei infection of macrophages was observed. BprB, a predicted regulator of T3SS3, is similar to Pseudomonas viridiflava response regulator GacA, while BprD shares weak homology with a sigma-54-dependent transcriptional activator. In Pseudomonas sp., the GacS/GacA signal transduction system controls metabolic and physiological processes including transcription of T6SS-encoding genes [38,39]. In addition, the sigma-54-dependent transcriptional activator regulates T6SS in several other bacteria including V. cholerae and Aeromonas hydrophila [37]. Specifically, sigma-54 positively regulates the expression of the hcp genes [40]. Therefore, the overexpression of these genes in our study suggests a role within a complex regulatory network that controls the expression of T6SS-5 during hostepathogen interaction. Recently, Hopf et al. [41] reported that BPSS1504, another component of the T6SS1 cluster affected Hcp1 secretion although transcriptional expression of hcp1 was independent of bpss1504 expression. Our expression profile of bpss1504 shows a moderate up-regulation of gene expression in the first 6 h post-infection of macrophages (Fig. 2). As the transcript levels of hcp1, tssJ-4 and tssE-5 were significantly high in bacteria undergoing an active infection, we postulated that similarly high levels of these three proteins would be available during infection. Hence, there was a strong likelihood that

S. Chieng et al. / Microbial Pathogenesis 79 (2015) 47e56

antibodies towards these proteins would be present in B. pseudomallei infected individuals. Genes encoding these proteins are highly conserved in B. pseudomallei strains and are also present in B. mallei and B. thailandensis. Infection with B. mallei, the causative agent of glanders rarely occurs [42], while B. thailandensis, a closely related species of B. pseudomallei, is generally avirulent in humans [43,44]. Therefore, the expression of these proteins by both B. mallei and B. thailandensis would not affect the specificity of these proteins as diagnostic agents for melioidosis. Additionally, these proteins also demonstrated low identity to proteins of other bacterial pathogens associated with septicaemia and diseases that mimic melioidosis, again highlighting the potential of these proteins as serodiagnostic markers for melioidosis. To demonstrate proof-of-concept, we expressed the recombinant form of the proteins and screened against melioidosis confirmed sera samples. The reactivity of each recombinant protein towards melioidosis patient sera samples was evaluated through western blot and ELISA. From the initial screen, only Hcp1 was highly reactive against pooled melioidosis patients' sera. The P/N value obtained for Hcp1 was also significantly higher (p < 0.001) than that of the other two recombinant proteins. The failure of recombinant TssJ-4 and TssJ-5 proteins to react against melioidosis confirmed sera could be attributed to a number of reasons. Although the respective genes were highly expressed at the transcript level during infection, this may not necessarily reflect or correlate with protein expression levels in vivo [45e47]. Failure to detect antibodies towards these B. pseudomallei proteins may perhaps denote the non-immunogenic nature of these proteins and hence, the failure to trigger the host immune response to stimulate B cells for specific antibody production [48]. Additionally, a heterogeneous immune response towards individual antigens following bacterial infection of different individuals is also well documented [49,50]. An enzyme-linked immunosorbent assay was carried out to evaluate the sensitivity and specificity of Hcp1 in comparison to B. pseudomallei whole cell lysate. When recombinant Hcp1 was used as the reacting antigen, high sensitivity (93.7%) and specificity (100%) was obtained, while moderate sensitivity (65.6%) but equally high specificity (90%) was demonstrated for bacterial lysate. Whilst the data suggests that Hcp1 and bacterial lysate are highly specific for B. pseudomallei, Hcp1 is far superior as an exposure marker. A pronounced genome level plasticity and heterogeneity for different B. pseudomallei strains is well established and the low sensitivity observed for bacterial lysate could be attributed to the use of only a single bacterial strain antigen preparation. Additionally, a concurrent screen of a larger cohort of melioidosis confirmed sera was performed to evaluate the potential of Hcp1 and three other B. pseudomallei recombinant proteins as diagnostic antigens [51]. Hcp1 demonstrated the highest specificity and sensitivity level compared to the other recombinant antigens tested. In addition, TssD-5/Hcp1 specifically excluded culture-negative sera from individuals living in Malaysia, a highly endemic region for melioidosis. Taken together, Hcp1 is a more valuable serodiagnostic marker to routinely screen suspected melioidosis individuals especially in endemic regions. The use of this marker could usurp the current time consuming gold standard of culture confirmation or immunohemagglutination (IHA). In conclusion, through transcriptome level analysis, we have demonstrated changes in the expression profile of B. pseudomallei T6SS gene clusters during intracellular infection of human macrophage cells. Further characterization of selected T6SS genes identified Hcp1 as a reliable efficient exposure marker. This antigen could be further developed into various rapid assays to allow for fast and sensitive serodiagnosis of melioidosis leading to early remedial intervention.

55

Conflict of interest The authors declare that they have no competing interests. Acknowledgements This project was funded by the Ministry of Science, Technology and Innovation of Malaysia 07-05-16-MGI-GMB06 under the Genomics and Molecular Biology R&D Initiatives Grant Program awarded to S.N. We are grateful to the Institute for Medical Research, Hospital Besar Kelantan and CliniPath Laboratory Malaysia Sdn. Bhd. for generously providing the sera samples. We also thank Mei-Perng Lim, RuieRui Wong and Pei-San Yee for their technical assistance. References [1] N.J. White, Melioidosis, Lancet 361 (2003) 1715e1722. [2] A.C. Cheng, B.J. Currie, Melioidosis: epidemiology, pathophysiology, and management, Clin. Microbiol. Rev. 18 (2) (2005) 383e416. [3] D.M. Phuong, T. Trung, K. Breitbach, N.Q. Tuan, U. Nubel, G. Flunker, D.D. Khang, N.X. Quang, I. Steinmetz, Clinical and microbiological features of melioidosis in Northern Vietnam, Trans. R. Soc. Trop. Med. Hyg. 102 (Suppl. 1) (2008) S30eS36. [4] J.F. Ellis, R.W. Titball, Burkholderia pseudomallei: medical, veterinary and environmental aspects, Infect. Dis. Rev. 1 (3) (1999) 174e181. [5] S. Sirisinha, N. Anuntagool, T. Dharakul, P. Ekpo, S. Wongratanacheewin, P. Naigowit, B. Petchclai, V. Thamlikitkul, Y. Suputtamongkol, Recent developments in laboratory diagnosis of melioidosis, Acta Trop. 74 (2e3) (2000) 235e245. [6] S.D. Puthucheary, J. Vadivelu, Human Melioidosis, Singapore University Press, Singapore, 2002. [7] H.S. Bruce, Melioidosis: an important emerging infectious disease e a military problem? ADF Health 3 (2002) 13e21. [8] T.J. Inglis, J.L. Sagripanti, Environmental factors that affect the survival and persistence of Burkholderia pseudomallei, Appl. Environ. Microbiol. 72 (11) (2006) 6865e6875. [9] S. Pilatz, K. Breitbach, N. Hein, B. Fehlhaber, J. Schulze, B. Brenneke, L. Eberl, I. Steinmetz, Identification of Burkholderia pseudomallei genes required for the intracellular life cycle and in vivo virulence, Infect. Immun. 74 (6) (2006) 3576e3586. [10] G. Shalom, J.G. Shaw, M.S. Thomas, In vivo expression technology identifies a type VI secretion system locus in Burkholderia pseudomallei that is induced upon invasion of macrophages, Microbiology 153 (2007) 2689e2699. [11] L.E.H. Bingle, C.M. Bailey, M.J. Pallen, Type VI secretion: a beginner's guide, Curr. Opin. Microbiol. 11 (1) (2008) 3e8. [12] O.P. Persson, J. Pinhassi, L. Riemann, B.I. Marklund, M. Rhen, S. Normark, lez, A. Hagstrӧm, High abundance of virulence gene homologues in J.M. Gonza marine bacteria, Environ. Microbiol. 11 (6) (2009) 1348e1357. [13] S. Pukatzki, S.B. McAuley, S.T. Miyata, The type VI secretion system : translocation of effectors and effector domains, Curr. Opin. Microbiol. 12 (1) (2009) 11e17. [14] F. Boyer, G. Fichant, J. Berthod, Y. Vandenbrouck, I. Attree, Dissecting the bacterial type VI secretion system by a genome wide in silico analysis: what can be learned from available microbial genomic resources? BMC Genomics 10 (2009) 104. [15] S. Pukatzki, A.T. Ma, A.T. Revel, D. Sturtevant, J.J. Mekalanos, Type VI secretion system translocates a phage tail spike-like protein into target cells where it cross-links action, Proc. Natl. Acad. Sci. U. S. A. 104 (39) (2007) 15508e15513. [16] J.M. Silverman, Y.R. Brunet, E. Cascales, J.D. Mougous, Structure and regulation of the type VI secretion system, Annu Rev. Microbiol. 66 (2012) 453e472. [17] R.D. Hood, P. Singh, F. Hsu, et al., A type VI secretion system of Pseudomonas aeruginosa targets a toxin to bacteria, Cell Host Microbe 7 (1) (2010) 25e37. [18] S. Schwarz, T.E. West, F. Boyer, W.C. Chiang, M.A. Carl, R.D. Hood, L. Rohmer, T. Tolker-Nielsen, S.J. Skerrett, J.D. Mougous, Burkholderia type VI secretion systems have distinct roles in eukaryotic and bacterial cell interactions, PLoS Pathog. 6 (8) (2010) e1001068. [19] M.N. Burtnick, P.J. Brett, S.V. Harding, S.A. Ngugi, W.J. Ribot, N. Chantratita, A. Scorpio, T.S. Milne, R.E. Dean, D.L. Fritz, S.J. Peacock, J.L. Prior, T.P. Atkins, D. Deshazer, The cluster 1 type VI secretion system is a major virulence determinant in Burkholderia pseudomallei, Infect. Immun. 79 (4) (2011) 1512e1525. [20] M.A. Schell, R.L. Ulrich, W.J. Ribot, et al., Type VI secretion is a major virulence determinant in Burkholderia mallei, Mol. Microbiol. 64 (6) (2007) 1466e1485. [21] Y. Chen, J. Wong, G.W. Sun, Y. Liu, G.Y.G. Tan, Y.H. Gan, Regulation of type VI secretion system during Burkholderia pseudomallei infection, Infect. Immun. 79 (8) (2011) 3064e3073. [22] M.N. Burtnick, P.J. Brett, Burkholderia mallei and Burkholderia pseudomallei cluster 1 type VI secretion system gene expression is negatively regulated by iron and zinc, PLoS One 8 (10) (2013) e76767.

56

S. Chieng et al. / Microbial Pathogenesis 79 (2015) 47e56

[23] S.H. Lee, C.E. Chong, B.S. Lim, S.J. Chai, K.K. Sam, R. Mohamed, S. Nathan, Burkholderia pseudomallei animal and human isolates from Malaysia exhibit different phenotypic characteristics, Diagn. Microbiol. Infect. Dis. 58 (3) (2007) 263e270. [24] Y.C. Su, K.L. Wan, R. Mohamed, S. Nathan, A genome level survey of Burkholderia pseudomallei immunome expressed during human infection, Microbes Infect. 10 (12e13) (2008) 1335e1345. [25] S. Chieng, L. Carreto, S. Nathan, Burkholderia pseudomallei transcriptional adaptation in macrophages, BMC Genomics 13 (2012) 328. [26] M.A. Larkin, G. Blackshields, N.P. Brown, et al., ClustalW and Clustal X version 2, Bioinformatics 16 (2007) 944e945. [27] D.A. Martin, B.J. Biggerstaff, B. Allen, A.J. Johnson, R.S. Lanciotti, J.T. Roehrig, Use of immunoglobulin M cross-reactions in differential diagnosis of human flaviviral encephalitis infections in the United States, Clin. Diagn. Lab. Immun. 9 (3) (2003) 544e549. [28] Y.S. Chen, D. Shiuan, S.C. Chen, S.M. Chye, Y.L. Chen, Recombinant truncated flagellin of Burkholderia pseudomallei as a molecular probe for diagnosis of melioidosis, Clin. Diagn. Lab. Immunol. 10 (3) (2003) 423e425. [29] G.W. Sun, Y. Chen, Y. Liu, G.Y. Tan, C. Ong, P. Tan, Y.H. Gan, Identification of a regulatory cascade controlling type III secretion system 3 gene expression in Burkholderia pseudomallei, Mol. Microbiol. 76 (2010) 677e689. [30] M.P. Stevens, M.W. Wood, L.A. Taylor, P. Monaghan, P. Hawes, P.W. Jones, T.S. Wallis, E.E. Galyov, An inv/mxi-spa-like type III protein secretion system in Burkholderia pseudomallei modulates intracellular behaviour of the pathogen, Mol. Microbiol. 46 (3) (2002) 649e659. [31] M.T.G. Holden, R.W. Titball, S.J. Peacock, et al., Genomic plasticity of the causative agent of melioidosis, Burkholderia pseudomallei, Proc. Natl. Acad. Sci. U. S. A. 101 (39) (2004) 14240e14245. [32] S. Pukatzki, A.T. Ma, D. Sturtevant, B. Krastins, D. Sarracino, W.C. Nelson, J.F. Heidelberg, J.J. Mekalanos, Identification of a conserved bacterial protein secretion system in Vibrio cholerae using the Dictyostelium host model system, Proc. Natl. Acad. Sci. U. S. A. 103 (2006) 1528e1533. [33] A. Filloux, A. Hachani, S. Bleves, The bacterial type VI secretion machine: yet another player for protein transport across membranes, Microbiology 154 (2008) 1570e1583. [34] J.D. Mougous, M.E. Cuff, S. Raunser, et al., A virulence locus of Pseudomonas aeruginosa encodes a protein secretion apparatus, Science 312 (2006) 1526e1530. [35] S. Shrivastava, S.S. Mande, Identification and functional characterization of gene components of type VI secretion system in bacterial genomes, PLoS One 3 (8) (2008) e2955. [36] F. de Pace, G. Nakazato, A. Pacheco, J.B. de Paiva, V. Sperandio, W.D. da Silveria, The type VI secretion system plays a role in type 1 fimbria expression and pathogenesis of an avian pathogenic Escherichia coli strain, Infect. Immun. 78 (12) (2010) 4990e4998.

[37] C.S. Bernard, Y.R. Brunet, E. Gueguen, E. Cascales, Nooks and crannies in type VI secretion system, J. Bacteriol. 192 (15) (2010) 3850e3860. [38] K.A. Hassan, A. Johnson, B.T. Shaffer, et al., Inactivation of the GacA response regulator in Pseudomonas fluorescens Pf-5 has far-reaching transcriptomic consequences, Environ. Microbiol. 12 (4) (2010) 899e915. [39] X. Wei, X. Huang, L. Tang, Y. Xu, Global control of GacA in secondary metabolism, primary metabolism, secretion systems, and motility in the rhizobacterium Pseudomonas aeruginosa M18, J. Bacteriol. 195 (15) (2013) 3387e3400. [40] T. Ishikawa, P.K. Rompikuntal, B. Lindmark, D.L. Milton, S.N. Wai, Quorum sensing regulation of the two hcp alleles in Vibrio cholerae O1 strain, PLoS One 4 (2009) e6734. [41] V. Hopf, A. Gohler, K. Eske-Pogodda, A. Bast, I. Steinmetz, K. Breitbach, BPSS1504, a cluster 1 Type VI secretion gene, is involved in intracellualr survival and virulence of Burkholderia pseudomallei, Infect. Immun. 82 (2014) 2006e2015. [42] A. Srinivasan, C.N. Kraus, D. Deshazer, P.M. Becker, J.D. Dick, L. Spacek, J.G. Bartlett, W.R. Byrne, D.L. Thomas, Glanders in a military research microbiologist, N. Engl. J. Med. 345 (2001) 256e258. [43] P.J. Brett, D. Deshazer, D.E. Woods, Burkholderia thailandensis sp. Nov., a Burkholderia pseudomallei-like species, Int. J. Syst. Bacteriol. 48 (Pt 1) (1998) 317e320. [44] W.J. Wiersinga, A.F. de Vos, R. de Beer, C.W. Wieland, J.J.T.H. Roelofs, D.E. Woods, T. van der Poll, Inflammation patterns induced by different Burkholderia species in mice, Cell. Microbiol. 10 (2007) 81e87. [45] M. Gry, R. Rimini, S. Stromberg, A. Asplund, F. Ponten, M. Uhlen, P. Nilsson, Correlations between RNA and protein expression profiles in 23 human cell lines, BMC Genomics 10 (2009) 365. [46] S.P. Gygi, Y. Rochon, B.R. Franza, R. Aebersold, Correlation between protein and mRNA abundance in yeast, Mol. Cell. Biol. 19 (3) (1999) 1720e1730. [47] Y. Taniguchi, P.J. Choi, G.W. Li, H. Chen, M. Babu, J. Hearn, A. Emili, X.S. Xie, Quantifying E. coli proteome and transcriptome with single-molecule sensitivity in single cells, Science 329 (5991) (2010) 533e538. [48] T.W. Mak, M.E. Saunders, The Immune Response: Basic and Clinical Principles, Elsevier Inc., Oxford, England, 2006. €llby, Levels of [49] P. Colque-Navarro, G. Jacobsson, R. Andersson, J.I. Flock, R. Mo antibody against 11 Staphylococcus aureus antigens in a healthy population, Clin. Vaccine Immunol. 17 (7) (2010) 1117e1123. [50] N.J. Verkaik, H.A. Boelens, C.P. de Vogel, M. Tavakol, L.G.M. Bode, H.A. Verbrugh, A. van Belkum, W.J.B. van Wamel, Heterogeneity of the humoral immune response following Staphylococcus aureus bacteremia, Eur. J. Clin. Microbiol. Infect. Dis. 29 (2010) 509e518. [51] Y. Hara, C.Y. Chin, R. Mohamed, S.D. Puthucheary, S. Nathan, Multiple-antigen ELISA for melioidosis e a novel approach to the improved serodiagnosis of melioidosis, BMC Infect. Dis. 13 (2013) 165.