Background
Members of the genus
Shigella, namely
S. flexneri,
S. dysenteriae,
S. sonnei and
S. boydii have caused and continue to be responsible for mortality and/or morbidity in high risk populations such as children under five years of age, senior citizens, toddlers in day-care centres, patients in custodial institutions, homosexual men and, war- and famine-engulfed people. Yearly episodes of shigellosis globally have been estimated to be 164.7 million and of these, 163.2 million were in developing countries and the remaining in industrialized nations. The mortality rate was approximately 0.7% [
1]. A recent study by Lee & Puthucheary [
2] on bacterial enteropathogens in childhood diarrhoea in a Malaysian urban hospital showed that
Shigella spp. was the third most common bacteria isolated.
S. flexneri and
S. dysenteriae type 1 infections are usually characterized by frequent passage of small amounts of stool and mucus or blood. At times, watery stool followed by typical dysenteric stool maybe present with
S. dysenteriae type 1 infection.
S. sonnei and
S. boydii infections are less severe with watery faeces but little mucus or blood.
Shigellosis is usually a self-limiting infection, however when it subsides, the intestinal ulcers heal with scar tissue formation. Uncomplicated recovery is usual and the organisms rarely cause other types of infections. Adversely, in 3 to 50% of cases, depending on the virulence of the strain, the nutritional and immune status of the host, the initial infection maybe followed by neurological complications or kidney failure. Serious complications do occur at greatest frequencies in malnourished infants, toddlers, older adults and immunocompromised individuals [
3,
4].
Virulence genes responsible for the pathogenesis of shigellosis may be located in the chromosome or on the
inv plasmid borne by the organism. They are often multifactorial and coordinately regulated, and the genes tend to be clustered in the genome. Previously reported PCR-based detection methods concentrated mainly on the
ipaH gene alone [
5,
6] or on
ipaH and
ial genes in two separate PCR assays [
7,
8]. As
ial is found on the large
inv plasmid which is prone to loss or deletions, this gene-based detection may give false negative results.
ipaH, on the other hand, is present on both the
Shigella chromosome and on a large plasmid and hence, it is a more stable gene to detect. However, the sole presence of
ipaH is not an absolute indicator of virulence as loss or deletion of the plasmid renders the bacterium noninvasive and therefore, avirulent.
set1A and
set1B are chromosomal genes encoding
Shigella enterotoxin 1 (ShET1), which cause the watery phase of diarrhoea in shigellosis [
9,
10].
ial and
ipaH are responsible for directing epithelial cell penetration by the bacterium and for the modification of host response to infection, respectively [
11‐
13].
Here, we describe the application of a multiplex PCR (mPCR) design for simultaneous detection of four virulence genes (set1A, set1B, ial and ipaH) in Shigella spp. and to determine the prevalence of these virulence genes in a random selection of Malaysian Shigella strains.
Methods
Bacterial strains and growth conditions
A total of 110 Shigella strains of S. flexneri (n = 84), S. sonnei (n = 15), S. dysenteriae (n = 10) and S. boydii (n = 1) were used in this study. These strains were isolated from patients with diarrhoea in Peninsular Malaysia from 1997–2000, and were provided by the Institute for Medical Research (IMR), Malaysia. Serotyping of the strains (Shigella antisera from Mast Diagnostics, UK) was carried out by the Bacteriological Unit, IMR. All the strains were checked on Salmonella-Shigella (SS) agar before being transferred to Luria Bertani (LB) agar plate, incubated overnight at 37°C for subsequent screening of virulence-associated genes. All strains were stored at -20°C in LB broth containing 15% glycerol.
Development of mPCR
Boiled suspension of bacterial cells was used as DNA template. Previously described primers, obtained from Integrated DNA Techs, USA, for detection of the four virulence genes were applied to the template [
8,
14,
15] (Table
1). Prior to combining all the four primer sets in an mPCR, each pair of primers was optimized singly in separate PCR assays. A typical 25-μl PCR reaction mixture for every primer set consisted of 1x PCR buffer B (Promega, USA), 4 mM MgCl
2, 130 μM of each deoxynucleotide (dNTP), 0.5 μM of each primer, 1 U of
Taq DNA polymerase (Promega, USA) and 2 μl of DNA template. Amplifications were carried out using a Robocycler Gradient 40 Temperature Cycler (Strategene Cloning Systems, USA). The cycling conditions were an initial denaturation at 95°C for 5 min, template denaturation at 95°C for 50 s, annealing at 55°C for 1.5 min, and extension at 72°C for 2 min for a total of 30 cycles, with a final extension at 72°C for 7 min.
Table 1
Primers used to identify various virulence-associated genes of Shigella spp.
ShET1A |
set1A
| TCA CGC TAC CAT CAA AGA TAT CCC CCT TTG GTG GTA | 309 | 14 |
ShET1B |
set1B
| GTG AAC CTG CTG CCG ATA TC ATT TGT GGA TAA AAA TGA CG | 147 | 14 |
ial |
ial
| CTG GAT GGT ATG GTG AGG GGA GGC CAA CAA TTA TTT CC | 320 | 15 |
Shig1 |
ipaH
| TGG AAA AAC TCA GTG CCT CT | 423 | 8 |
Shig2 | | CCA GTC CGT AAA TTC ATT CT | | |
Based on the results of individual priming, an mPCR was designed. Various parameters such as concentrations of primers (0.5–0.8 μM), MgCl2 (2 to 4 μM), Taq DNA polymerase (0.6 to 4 U) and dNTPs (100–150 μM) and buffer strength (1.4X to 2.4X) were tested. The simultaneous gene amplifications were performed in a reaction volume of 25 μl consisting of 1.8X PCR buffer B (Promega, USA), 4 mM MgCl2, 130 μM of each dNTP, 0.3 μM of each ShET1B primer, Shig1 and Shig2 primers, 0.5 μM of each ShET1A and ial primers, 1 U of Taq DNA polymerase (Promega, USA) and 2 μl of DNA template. All the reaction mixtures were overlaid with 20 μl of sterile mineral oil. Amplifications were similarly carried out as above.
After initial screening, strain TH13/00 (S. flexneri 2a) was chosen as a positive control for PCR assays. A negative control using sterile distilled water as template was included in every PCR assay. The DNA fragments were separated in 2% agarose gel.
Reproducibility test
The mPCR assay was repeated at least twice with 28 strains to determine the reproducibility of the results, whereby the DNA template of a particular strain was freshly prepared for each repeat.
Specificity test
The specificity of the mPCR assay was tested with 12 other non-Shigella pathogens: Enterobacter cloacae, Salmonella Paratyphi A (ATCC 9281), S. Paratyphi C, S. Typhimurium, S. Enteritidis, S. Typhi (ATCC 7251), Listeria monocytogenes, Pseudomonas aeruginosa, Klebsiella pneumoniae, Citrobacter freundii, Escherichia coli O157:H7 and E. coli O78:H11.
Faecal spiking and sensitivity test
This was based on a modification of that described by Chiu and Ou [
16]. Approximately 0.2 g of faeces from a healthy individual was suspended in 1 ml of brain heart infusion (BHI) (Oxoid Ltd., UK) and diluted 10-fold. Then, 1 ml of the diluted faecal suspension was inoculated into 4 ml of BHI and vortexed to obtain a homogenous mixture of broth-faecal suspension. Meanwhile, an overnight culture of
S. flexneri 2a was harvested and serially diluted 10-fold with BHI. Then, 250 μl of each dilution of cell culture was mixed with 250 μl of the broth-faecal suspension and 500 μl of BHI in a new eppendorf tube. The tubes were vortexed and preincubated at 37°C for 4 h without shaking. Simultaneously, 100 μl of each diluted culture was plated on LB agar (Oxoid Ltd., UK) to determine the number of viable bacteria in each dilution. After preincubation, mPCR assay was performed on the boiled lysates of each diluted culture. A pure culture of
S. flexneri 2a (TH13/00) and an unspiked faecal sample served as positive and negative controls.. The test was repeated with a spiked faecal sample of another healthy individual, and the average detection limit was reported.
Screening of clinical specimens
0.2 g of each faecal sample from 10 patients suffering from diarrhoea in a local tertiary University Hospital was suspended in 1 ml of BHI and diluted 10-fold. A volume of 250 μl of broth-faecal suspension was inoculated into 5 ml of BHI and preincubated at 37°C for 4 h without shaking. Concurrently, 100 μl of the suspension was plated onto MacConkey and SS agar plates and incubated overnight at 37°C. mPCR assay was performed on the boiled lysate of the broth-faecal suspension after preincubation. A pure culture of strain TH13/00 and a Shigella-spiked faecal sample served as a positive control, whilst a PCR reaction mixture without bacterial DNA template and an unspiked faecal sample from a healthy individual acted as a negative control.
Discussion
Numerous studies had been performed to detect virulence genes in
Shigella by monoplex PCRs [
8,
17,
18]. Studies involving the combination of chromosomal- and plasmid-encoded virulence genes in a single assay for
Shigella detection, on the other hand, are scarce. Although the optimization of mPCR is more tedious and difficult to achieve than monoplexes, the ease of screening a large number of specimens, once the system is optimized, far outweighs the initial problems. The present mPCR system encompasses the presence of virulence genes found in the
Shigella chromosome and on the large
inv plasmid. Hence, it can determine if the pathogenesis of a particular strain is attributable to its chromosome or the plasmid, or if the strain is still invasive or otherwise, in a single reaction.
Initially, the monoplex PCRs were carried out following reaction conditions as proposed by a previous report [
17]. However, we could not reproduce their results and hence had to modify the PCR conditions. Our failure to reproduce identical results despite using similar reagent concentrations and amplification conditions maybe attributed to the different makes of PCR reagents and primers used. Broude
et al. [
19] had compared amplification efficiencies of two commercial
Taq DNA polymerases and found that they displayed different specificity in PCR.
Preferential amplification of one target sequence over another is a known phenomenon in mPCRs and it is usually overcome by increasing the amount of primers for the weaker amplification simultaneously with a decrease of primer concentrations for the stronger amplification. Buffer concentration may also affect mPCR amplifications despite it being seldom considered during monoplex optimization works [
20]. Upon adjustment of primers and buffer concentrations, specific and consistent amplification of all the genes in the multiplex combination was achieved.
Although other studies have demonstrated the presence of
ial and
ipaH in strains of enteroinvasive
Escherichia coli (EIEC) [
11,
13,
21], we had not applied the mPCR assay to EIEC strains. It is unfortunate that these strains were not available for our study as EIEC gives rise to similar illness as Shigellosis.
Our study supported the observations of Noriega
et al. [
9] and Vargas
et al. [
17] in local
Shigella strains. Their studies showed that both
set1A and
set1B were present exclusively in
S. flexneri 2a. The complete correlation between the presence of both
set1A and
set1B showed that both genes are indeed found in tandem in the
Shigella genome. In this study, almost all the
Shigella strains positive for the presence of
set1A and
set1B (41/45 strains) belonged to
S. flexneri 2a, thus confirming previous works that both genes are highly conserved in this particular serotype [
14].
Both the prevalence of
ial and
ipaH were independent of the four different species of
Shigella tested. Though both
ial and
ipaH are responsible for invasion-related processes and are found on the
inv plasmid, the
ial gene cluster resides near a region of the plasmid, which is a hot spot for spontaneous deletions [
22]. This probably explains the lower prevalence of
ial (45/110 strains) than
ipaH (110/110 strains) in the Malaysian
Shigella strains. Since invasiveness is a prerequisite for virulence in shigellae and since most of these virulence genes are located on the large plasmid, these strains would have possessed the plasmid when first isolated from patients. Due to storage/subculturing, the plasmid might have been lost together with the virulence-associated genes. By virtue of multiple copies being present on both the chromosome and the
inv plasmid [
23],
ipaH seemed to be less compromised by plasmid loss and/or deletions. As the sole presence of
ipaH is not indicative of the invasive phenotype, our mPCR design, which incorporated three other virulence genes, could determine the invasiveness of
Shigella strains in epidemiological studies.
Dilution of the faecal sample with BHI was performed to lower the levels of PCR inhibitors such as bilirubin, bile salts and heme in the faeces [
16]. An additional step of preincubating the spiked faecal samples also helped to eliminate the natural inhibitors [
24]. The short 4-h enrichment step would increase the total number of target sequences caused by more bacterial growth and the overall detection sensitivity of the assay. Although PCR cannot differentiate between dead and viable bacteria, enrichment helped to dilute the concentrations of dead bacteria, thus reducing the probability of detecting them by the subsequent mPCR assay. The sensitivity level achieved in the study was found to be comparable to other studies. Houng
et al. [
25] detected up to 7.4 × 10
4 cfu
shigellae ml
-1 by amplifying the IS 630 sequences in
shigella spp.. Yavzori
et al. [
24] reported a detection level of 10
4 cfu
shigellae per gram of faeces with the use of
virF primers. Although it has been reported that ingestion as low as 100
shigellae resulted in clinical disease [
26], the highest percentage of volunteers having diarrhoea were administered doses of at least 10
4 viable organisms. Thus, the average detection limit of mPCR described in this study (5.0 × 10
4 cfu/ml) is within the common infectious dose for
shigellae.
Results from the preliminary clinical screening were promising. Nevertheless, the consideration of other diarrhoeal pathogens being present in the clinical samples cannot be negated. More patient samples are warranted to thoroughly vet the robustness and applicability of the developed mPCR in clinical environments.
One limitation of the present mPCR system is its inability to differentiate
Shigella spp., unlike the multiplex reactions based on specific
rfc genes developed by Houng
et al. [
25]. For future research, either
set1A or
set1B may be omitted from the multiplex system as both genes are shown to exist tandemly.
rfc primers of different
Shigella origins maybe incorporated to enable the discrimination of
Shigella spp. as well as the identification of virulent strains in one assay.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
SLLH carried out the experiments, data analysis and wrote the manuscript. RMY provided the bacterial strains. SDP contributed to the writing of the manuscript. TKL conceived and co-designed the study, provided input for writing and supervision of the study. All authors read and approved the final manuscript.