Detection and differentiation of Entamoeba histolytica and Entamoeba dispar in clinical samples through PCR-denaturing gradient gel electrophoresis

Amebiasis is one of the twenty major causes of disease in Mexico; however, the diagnosis is difficult due to limitations of conventional microscopy-based techniques. In this study, we analyzed stool samples using polymerase chain reaction-denaturing gradient gel electrophoresis (PCR-DGGE) to differentiate between Entamoeba histolytica (pathogenic) and E. dispar (non-pathogenic). The target for the PCR amplification was a small region (228 bp) of the adh112 gene selected to increase the sensitivity of the test. The study involved 62 stool samples that were collected from individuals with complaints of gastrointestinal discomfort. Of the 62 samples, 10 (16.1%) were positive for E. histolytica while 52 (83.9%) were negative. No sample was positive for E. dispar. These results were validated by nested PCR-RFLP (restriction fragment length polymorphism) and suggest that PCR-DGGE is a promising tool to differentiate among Entamoeba infections, contributing to determine the specific treatment for patients infected with E. histolytica, and therefore, avoiding unnecessary treatment of patients infected with the non-pathogenic E. dispar.


Introduction
One of the major health problems in developing countries is amebiasis. In 1997, the World Health Organization declared this disease as the third leading cause of death due to parasitic infections (1,2). Currently, amebiasis is still a serious public health problem because parasitic infections are commonly neglected, particularly in populations that lack hygienic measures and clean drinking water (1,3). In Mexico, amebiasis is one of the twenty major causes of disease; its incidence rate in 2000 was 1,353.43 per 100,000 (1,4,5).
Amebiasis is caused by the parasite Entamoeba histolytica, including both intestinal and extra-intestinal infections. This parasite can be present in sewage and contaminated water. According to its cell cycle, it can exist in two forms: trophozoites and cysts (6,7). There is a second species with identical morphological characteristics to those described for E. histolytica called Entamoeba dispar; however, the biochemical, immunological and genetic data indicate that E. dispar is non-pathogenic (8). The life cycle in both species is the same. The infection begins with the ingestion of cysts from water or food contaminated with fecal matter. In the small intestine occurs the excystation and the trophozoites emerge. The trophozoites colonize the large intestine and adhere to the colonic mucosa (6,9). Only the not encysted trophozoites of E. histolytica acquire invasiveness. By the action of proteases, hyaluronidases and mucopolysaccharidases E. histolytica erodes the mucosa producing ulcers and may even reach the submucosa. The adhesive interaction of the trophozoites with the surface of host cells is determinant for the invasion of human tissues, cytotoxic activity, and severity of the disease (10). Primary molecules involved in the intestinal invasion process of E. histolytica are the Gal/ Gal NAc lectin and EhCPADH112 (124 kDa) complex (11,12). This complex is formed by the genes: i) Ehcp112, encoding a cysteine protease (50 kDa), and ii) Ehadh112, encoding an adhesin (75 kDa). Some studies have analyzed the molecular role of EhCPADH112 in E. histolytica, but it has not been identified in E. dispar (13).
The laboratory diagnosis of amebiasis is usually based on microscopy, immunological methods and polymerase chain reaction (PCR). The occurrence of non-pathogenic species (particularly E. dispar) causes a confusing scenario for a correct diagnosis of intestinal amebiasis and E. histolytica is often inaccurately reported or diagnosed (8,(14)(15)(16). The denaturing gradient gel electrophoresis (DGGE) is a well-established tool for molecular microbiology; this method makes possible the electrophoretic separation of DNA fragments on the basis of differences in nucleotide composition rather than their size (17,18).
To address the need for a reliable diagnostic test of amebiasis caused by E. histolytica in human stools, we developed a polymerase chain reaction-denaturing gradient gel electrophoresis (PCR-DGGE), which differentiates E. histolytica from E. dispar, as it is highly specific and sensitive to these two species (more specific than other techniques such as ELISA). The PCR primers were based on a conserved portion of the adh112 gene spanning a region with substitutions that allowed the differentiation between pathogen and non-pathogen species. To our knowledge, there are no studies using DGGE as a molecular identification technique to determine the prevalence of E. histolytica in Mexico.

Sample details
The study involved 62 stool samples collected from patients who presented gastrointestinal complaints and attended the "Maximiliano Dorantes" Health Center. They were all examined for intestinal parasites using coproparasitoscopic studies (for multiple ova and parasites); we used the method of Faust with subsequent staining with Lugol solution in order to find cysts (19,20). We used E. histolytica DNA as a positive control, which was obtained from a monogenic culture of strain HM-1 IMSS donated by the Molecular Diagnostic Laboratory, Department of Experimental Pathology CINVESTAV. Likewise, DNA of E. dispar was also used as a positive control; this was donated by the Department of Experimental Medicine, Faculty of Medicine UNAM. The coproparasitoscopic tests were performed in triplicate.

DNA extraction
The DNA of controls (E. histolytica and E. dispar) as well as the 62 studied samples were obtained directly from stools stored at -20°C, by means of Entamoeba cysts mechanical lysis, using zircon beads of 0.01 mm in diameter and then using the Wizard s Genomic DNA Purification Kit (Promega, USA) following the manufacturer's recommendations. As negative controls, we used human DNA extracted from whole blood of volunteer donors. We also utilized the Wizard s Genomic DNA Purification Kit to isolate DNA from white blood cells. The DNA samples were stored at -20°C until analyzed.

Bioinformatic analysis for primer design
In order to differentiate the orthologous adh112 gene of E. dispar from the adhesin of E. histolytica, a basic local alignment search tool (BLAST) was performed using the sequence reported by Garcia-Rivera et al. (21) obtained from GenBank database with accession number: AF127375 (E. histolytica). Taking this as a target sequence, it was aligned using the BLAST tool in the genome database of all organisms to find highly conserved regions. E. dispar showed 93% sequence identity with the adhesin gene from E. histolytica (accession No.: AANV02000421).
According to the high similarity between the two sequences, it was necessary to select a part of the sequence containing five differences for at least one nucleotide, because the DGGE is sensitive enough to detect differences of a single base between two sequences. The BioEdit (22) and GeneDoc (www.psc.edu/biomed/genedoc) softwares were used to interpret multiple alignments and manual adjustments. The alignment of E. histolytica and E. dispar sequences allowed the selection of a 228 bp-region containing five single base differences throughout the sequence (from base 916 to 1144) to be further amplified by PCR (Figure 1).
An in silico analysis to examine the complementarity within the sequence of each primer was performed using the Oligonalyzer program (www.idtdna.com/Home/Home.aspx). The designed primers also amplified another species of amoeba: E. nuttalli; however, its selected 228 bp-region did not contain the five single base differences observed in E. histolytica and E. dispar, so it was not included in our analysis. Moreover, E. nuttalli is non-pathogenic for humans and therefore, not important for our study (23). In order to achieve greater sensitivity in the detection by DGGE between E. histolytica and E. dispar, a GC-clamp was added to the forward primer: 5 0 -CGCCCGCCGCGCGGCCGCGG CCGGCCGGGGGCACGCGGCG-3 0 (24), changing the total size of the fragment to be amplified from 228 to 268 bp.

PCR amplification of adh111 gene
We tested various quantities of DNA (25, 50, and 100 ng) for PCR amplification and decided to use 100 ng of DNA.

Denaturing gradient gel electrophoresis
The PCR products were subjected to DGGE with 10 and 30% linear denaturing gradients of urea and formamide in a 10% polyacrylamide gel (Promega, USA). DGGE was performed with a 10-30% denaturing gradient adding 210 mL of 10% ammonium persulfate and 10 mL of tetramethylethylenediamine (TEMED).  instructions. The RFLP-digested product was visualized by loading 5 mL of sample on a 1.6 % agarose gel containing ethidium bromide.

Coproparasitoscopic exam of stool samples
All samples were tested for E. histolytica and E. dispar using the Faust coproparasitoscopic exam. Of the 62 stool samples screened, 18 were positive for E. histolytica, 22 were positive for either Escherichia coli, E. nana, Giardia lamblia, or Ascaris lumbricoides cysts and negative for Entamoeba. Finally, 22 samples were negative for any parasite. No sample was positive for E. dispar.

Specificity and sensitivity of PCR methods
We found that the designed primers were specific for the expected fragment of 268 pb. This amplicon was observed in stool samples and positive controls (E. histolytica and E. dispar). Figure 2 shows a visible amplicon of 268pb in the analyzed samples. Although other amplicons were detected in the stool samples and negative controls, it should be noted that they were not found in the positive controls.

Clinical evaluation of PCR-DGGE-adh112 assay
Once the DNA was isolated from stool samples, the primers were selected and concentrations standardized, the denaturant conditions were fully optimized, then DGGE gels were run (Figure 3). Out of the 62 samples, 10 (16.1%) were positive for E. histolytica, while 52 (83.9%) were negative. No sample was found to be positive for E. dispar.

Nested PCR-RFLP of SSU rRNA gene
The RFLP pattern for E. histolytica showed 563 bp and 311 bp fragments and an undigested 874 bp fragment, whereas that for E. dispar showed 743 bp and 131 bp fragments (Figure 4). The nested PCR-RFLP was positive for E. histolytica in 10 (16.1%) stool samples and 52 (83.9%) samples were negative. These results are similar to what we observed with the PCR-DGGE technique, that is, the positive samples for E. histolytica by PCR-DGGE were also positive by PCR-RFLP. No sample was found to be positive for E. dispar. The sensitivity and specificity of both, PCR-DGGE and PCR-RFLP were 100%. The 95% CI values were also estimated and used to evaluate the sensitivity (95%CI=65.55-99.08) and specificity (95% CI=91.43-99.82) of PCR-DGGE technique.

Discussion
The identification of Entamoeba spp. has always been controversial and microscopy is usually used to diagnose protozoa in stool samples. However, this method is unable to differentiate E. histolytica from the morphologically identical non-pathogenic species such as E. dispar (8,14). Therefore, the WHO recommends the development and application of new methods for a specific diagnosis of E. histolytica infection (2). The present study describes a new PCR-DGGE strategy for species-specific detection and differentiation of E. histolytica and E. dispar DNA using stool samples. Several methods including isoenzyme analysis, antibody or antigen detection tests, immunochromatographic assays and real-time PCR have been used for an accurate detection of E. histolytica and E. dispar However, high cost limits their use in underdeveloped countries and, additionally, some yield false negative results (25)(26)(27)(28).
Another alternative for differentiating the species is using PCR. This technique demonstrates exceptional sensitivity and specificity compared with microscopy (8,14). This research is the first of its kind in Mexico, and was designed to detect and differentiate E. histolytica from E. dispar using a fragment of the adh112 gene, which presents five differences in single bases when comparing both species. The advantage of the PCR-DGGE-adh112 method over the PCR and restriction enzyme digestion method has to do with the reliability of the results. The electrophoretic pattern obtained in PCR and restriction enzyme digestion are based on the nucleotides size, and it can be difficult to see the differences and similarities that could exist among numerous samples (29). In contrast, DGGE is sensitive enough to detect differences of a single base. With the PCR-DGGE-adh112 the DNA fragments move through polyacrylamide gels containing a linear gradient of denaturing agents, resulting in a partially denatured molecule observed in the gel with distinctive electrophoretic pattern (the change in the sequence causes a change in the pattern of run), which gives greater effectiveness to the test and no likelihood of false positives or false negatives (17,24).
In Mexico, there is a lack of knowledge about the epidemiology of E. histolytica infection, although amebiasis has been considered for many years as a major health problem in the country (4,30). By using PCR-DGGE-adh112 we found the mono infection with E. histolytica to be 16.1%. No sample was found to be positive for E. dispar. These results were validated using another technique (nested PCR-RFLP of SSU rRNA gene) considered by several authors as the gold standard for the differential diagnosis between E. histolytica and E. dispar (31)(32)(33)(34). Our results were consistent with those observed by PCR-RFLP. Our PCR-DGGE-adh112 results showed sensitivity and specificity of 100%, indicating that it is a useful and reliable test to specifically detect E. histolytica in stool samples. PCR-DGGE has the advantage of identifying and differentiating E. histolytica and E. dispar, which is not possible using microscopy or ELISA (8,35). It should be noted that the differentiation of pathogenic E. histolytica from the morphologically identical E. dispar is important for the clinical management of patients.
In conclusion, the present study reports a new PCR-DGGE technique for species-specific detection and differentiation of E. histolytica and E. dispar DNA in stool samples. This technique could become an alternative or a complementary diagnostic tool.