Distribution of pathogenicity islands among Colombian isolates of Salmonella

Introduction: Salmonella pathogenicity islands (SPIs) are regions scattered along the bacterial chromosome, with an acknowledged pivotal role during gastrointestinal and systemic infection. The distribution of SPIs has been investigated in reference strains. However, there is a lack of studies on their presence and/or assortment within the genomes of Salmonella enterica (S. enterica) serovars that circulate in different geographical regions. Therefore, in this study, we aimed to determine the presence of genes of the pathogenicity islands 1 to 5 (SPI-1 to 5), in Salmonella clinical isolates from Colombian patients with systemic and enteric outcomes. Methodology: A total of 125 strains of S. enterica belonging to different serovars were isolated from various clinical samples. Strains were identified and screened for the presence of various genes located in pathogenicity islands. The genes tested were selected according to the attributed pathogenic function and detected by PCR for the SPI-1 hilA and invA; for SPI-2 spiC and ttrC; for SPI-3 misL and mgtC; for SPI-4 orfL and SPI-4R; and for SPI-5 pipD and sopB. Results: Salmonella pathogenicity islands 1 to 5 were detected in isolates from patients with systemic and gastrointestinal infection. All the systemic isolates possessed all the genes tested; in contrast, 16 isolates from stool samples lacked one or more sequences encoded by the SPI3 and SPI-4 (p < 0.000001). Conclusions: These results describe the heterogeneous distribution of SPIs-encoded sequences within the genomes of Colombian clinical isolates, and reveal important differences among systemic and stool sample isolates.


Introduction
Salmonellosis is responsible for a large number of infections in both humans and animals [1].Salmonella enterica comprises approximately 2,579 serovars that cause a variety of clinical syndromes [2].
Virulence properties of S. enterica are associated with many molecular determinants.Among the other proteins encoded by genes located on Salmonella, pathogenicity islands (SPIs) have a very important role during the infection process, since they are either structural components of the secretion system III (SST), or effectors injected through this supramolecular device, and allow bacteria to colonize and invade the host mucosa and to establish the subsequent systemic infection [3].
SPIs are portions of DNA that have been acquired from other microorganisms by horizontal transfer, and are absent in non pathogenic strains [4].To date, 12 SPIs have been identified in different serovars [5], but only five are well characterized.
The assessment of whether a different assortment of SPI genes occurs in serovars causing systemic versus enteric infection in humans could help to explain the function of this group of genes, and the variability of clinical syndromes produced by S. enterica in humans.
The aim of the present work was to determine the distribution of pathogenicity islands SPI-1 to SPI-5 among Colombian isolates of Salmonella from patients with systemic and enteric clinical outcomes, inhabitants of an endemic region of salmonellosis.This is the first report on the distribution of SPIs among Colombian S. enterica clinical isolates.

Bacterial Strains
A total of 125 strains belonging to different serovars of Salmonella enterica, were isolated from various clinical samples.Out 125 strains, 91 were isolated from stool samples collected from patients who showed gastrointestinal symptoms, including diarrhoea, vomiting, and fever.Thirty-four strains were isolated from clinical samples of patients having different systemic symptoms such as enteric fever (n = 24), osteomyelitis (n = 3), hepatic abscess (n = 1), meningitis (n = 2), and intestinal perforation (n = 1).Samples included blood, cerebrospinal fluid (CSF) and peritoneal fluid.

Bacterial identification
Salmonella strains were isolated from clinical samples and Xylose Lactose Desoxycholate agar (XLD) (Difco, Becton Dickinson, New South Wales, Australia).All isolates were classified as Salmonella enterica by standard biochemical methods using API20E (Biomerieux, Cambridge, MA, USA) strips.Serotyping was performed using mono and polyclonal antibodies (Bio-Rad Hercules, CA, USA) and Pasteur Institute, Paris, France).Bacterial cultures were further grown in Luria Bertani broth (Sigma, St. Louis, MO, USA) for DNA extraction.
Genomic DNA was extracted by boiling and used for PCR assays.

PCR protocol
Briefly, 25 μl volume of reaction mixture was made, containing 1.25 μl 10 μM of each primer; 0. all experiments, each PCR amplification was repeated three times.

Statistical analysis
Statistical analysis was performed by the calculation of frequencies by percentages.The comparison of variables was done through the proportion comparison using Epi Info 6.0, CDC Atlanta, GA, USA statistical program.
The significance of 0.05 was used to establish differences between variables.
All S. enterica serovar Typhi strains were isolated from blood and other systemic sources.Table 2 shows serotype distribution according to source of sample

Statistical analysis
Statistical differences were found comparing the percentages of positive isolates from systemic samples versus positive isolates from stool samples (p = 0.000001).Significant differences among presence or absence of the genes tested according to Salmonella serovars were not found (p > 0.05).

Discussion
During the last decade, knowledge about the biology of Salmonella enterica has become more extensive and identification of some structural components such as virulence factors has been made.However, information about the presence of virulence genes in clinical isolates is scarce, especially for those that come from geographical regions where Salmonella enterica is endemic, such as Colombia.In this country, the prevalence of disease is not well known; the information that is available is from the National Institute of Health of Colombia, which reports the identification of 3,173 isolates between 1997 and 2008,, mainly S. Typhimurium with 1,131 isolates (35.6%) [9].
We tested the presence of virulence genes located in SPI-1 to 5 (SPI-1 to 5) in local clinical isolates.A number of studies focusing the same approach have been conducted in other countries including Italy [10], France [11], Brazil [12] and the United States [13].To our knowledge, however, this is the first study reporting the distribution of SPIs in Colombian isolates.
In this study, all clinical isolates were found positive for SPI-1 genes hilA and invA, regardless of serovars and clinical outcome.These results are in agreement with those of Ginocchio et al. [14], who reported the occurrence of SPI-1 gene deletions solely in a few environmental isolates of Salmonella serovars, which are consistently present in isolates from human source.Ammari et al. [15] also reported the occurrence of invA genes in fifteen isolates of Salmonella Enteritidis analyzed from Morocco.
Accordingly, we showed that Salmonella serovars isolated from systemic infection samples (i.e., blood, bone, cerebral spinal fluid, and peritoneal fluid) were positive to all genes tested.In contrast, the selected panel of SPI-encoded genes was not complete in 16 isolates from stool samples.In particular, SPI-3 and SPI-4 encoded genes were missing (misL and orfL respectively).This result could be explained by insertions or deletions within SPI-3 and SPI-4 regions, to be attributed to differences in the evolutionary process of acquisition and loss of corresponding genes.
A number of studies reported that misL and orfL genes play an important role in Salmonella pathogenicity.For example, MisL, an autotransporter protein, is an extracellular matrix adhesin involved in intestinal colonization [16], and orfL is required for macrophage survival [17].The function of both genes is involved directly with the invasion process, which is necessary to produce systemic infection [18].SPI-3 and SPI-4 encoded genes have been associated with Salmonella pathogenicity, due to the impaired virulence of their specific deleted derivatives, as demonstrated with experimental evidence in animal or cellular models of infection [16,18].In this respect, our data provide epidemiological evidence that misL and orfL, and possibly more virulence SPI-encoded genes, already known to play a "secondary" role during Salmonella systemic infection are, indeed, required to establish a systemic infection in humans.
The present study on Salmonella enterica isolates from Colombia constitutes the first attempt to learn about the distribution of virulence loci in clinical report demonstrate that molecular epidemiology might effectively contribute to define the clinical isolates in this region.The data collected in this impact of already known SPI-encoded virulence genes, and point to the need for further studies in different epidemiological and geographic contexts.

Figure 1 .
Figure 1.Agarose gel for detection of invA gene of SP1 by PCR.

Figure 2 .
Figure 2. Agarose gel for detection of spi4R gene of SPI4 by PCR.

Table 1 .
Sequences of primers and products sizes.