Genomic distribution of SINEs in Entamoeba histolytica strains: implication for genotyping

Abstract

Background: The major clinical manifestations of Entamoeba histolytica infection include amebic colitis and liver abscess. However the majority of infections remain asymptomatic. Earlier reports have shown that some E. histolytica isolates are more virulent than others, suggesting that virulence may be linked to genotype. Here we have looked at the genomic distribution of the retrotransposable short interspersed nuclear elements EhSINE1 and EhSINE2. Due to their mobile nature, some EhSINE copies may occupy different genomic locations among isolates of E. histolytica possibly affecting adjacent gene expression; this variability in location can be exploited to differentiate strains.

Results: We have looked for EhSINE1- and EhSINE2-occupied loci in the genome sequence of Entamoeba histolytica HM-1:IMSS and searched for homologous loci in other strains to determine the insertion status of these elements. A total of 393 EhSINE1 and 119 EhSINE2 loci were analyzed in the available sequenced strains (Rahman, DS4-868, HM1:CA, KU48, KU50, KU27 and MS96-3382. Seventeen loci (13 EhSINE1 and 4 EhSINE2) were identified where a EhSINE1/EhSINE2 sequence was missing from the corresponding locus of other strains. Most of these loci were unoccupied in more than one strain. Some of the loci were analyzed experimentally for SINE occupancy using DNA from strain Rahman. These data helped to correctly assemble the nucleotide sequence at three loci in Rahman. SINE occupancy was also checked at these three loci in 7 other axenically cultivated E. histolytica strains and 16 clinical isolates. Each locus gave a single, specific amplicon with the primer sets used, making this a suitable method for strain typing. Based on presence/absence of SINE and amplification with locus-specific primers, the 23 strains could be divided into eleven genotypes. The results obtained by our method correlated with the data from other typing methods. We also report a bioinformatic analysis of EhSINE2 copies.

Conclusions: Our results reveal several loci with extensive polymorphism of SINE occupancy among different strains of E. histolytica and prove the principle that the genomic distribution of SINEs is a valid method for typing of E. histolytica strains.

Figure 1

Analysis of EhSINE1: 393 EhSINE1 copies (length > 450 bp) of Entamoeba histolytica HM-1:IMSS were taken for analysis. 1.0 kb from both 5′- and 3′-ends of each EhSINE1 element were extracted using a perl code wherever possible. The flanking sequences were separately mapped to the contigs of the Rahman strain using BLAST. Only when both flanking sequences of a specific SINE element mapped to a single contig was it used for further analysis.

Figure 2

Analysis of EhSINE2. All the copies of EhSINE2 fulfilling the above mentioned criteria, were extracted from the whole genome of Entamoeba histolytica HM-1:IMSS. These were compared with the Rahman database using BLAST. Loci identified were analyzed for their occupancy as described.

Figure 3

Detection and validation of EhSINE1 polymorphic loci 13, 17, 19 and 42. (A) Schematic representation of primers designed from different loci. The hollow box represents the EhSINE1 element; flanking genes have not been shown for simplicity. (B) PCR was performed using genomic DNA of HM-1:IMSS (H) and Rahman (R) strains as template, using primers from sequences flanking the EhSINE1 copy as shown in the schematic representation. The size of amplicons was determined by electrophoresis in 1% agarose gels (Top panel). Of 4 SINE1 unoccupied sites found computationally two were tested (13 and 42). Two more (17 and 19) were evaluated by PCR and Southern Blotting in Rahman. The sizes of amplicons obtained are indicated on the right, with arrows. The amplicon from strain Rahman was shorter by ~550 bp (the size of EhSINE1) at loci 13, 17 and 19, but was longer at locus 42 (explained in the text). The absence of EhSINE1 was further confirmed by Southern blotting with EhSINE1 probe, which failed to hybridize with the amplicons from strain Rahman (Bottom panel).

Figure 4

Validation of EhSINE 2 polymorphic loci 18 and 50. PCR was performed using genomic DNA of E. histolytica HM-1:IMSS (H) and Rahman (R) as template, using primers from sequences flanking the EhSINE2 copy (number as shown on top). The size of amplicons was determined by electrophoresis in 1% agarose gel. EhSINE2 was missing in Rahman at the two loci, as the amplicon from Rahman was shorter by ~700 bp (the size of EhSINE2) at these loci (Panel (i)). The sizes of amplicons obtained are indicated on the right (arrows). The absence of EhSINE2 was further confirmed by Southern blotting with EhSINE2 probe, which failed to hybridize with the amplicons of strain Rahman (Panel (ii)). The specificity of the amplicon in Rahman was checked by Southern blotting with locus specific probe (Panel (iii)), which hybridized with the amplicons in both strains.

Figure 5

Sequence alignment of EhSINE1 Loci 17 and 19. Genomic DNA of HM-1:IMSS and Rahman was used to obtain amplicons of the two loci, which were cloned and sequenced. Underlined sequences correspond to the target site (site of EhSINE1 insertion), which is duplicated in HM-1:IMSS and present as single copy in Rahman. SINE1 has been represented by solid box, dotted line shows conserved sequence and broken line represent missing sequence of Rahman with respect to HM-1:IMSS.

Figure 6

Categorization of strains based on EhSINE1 loci 13, 17 and 19: (A). Schematic representation of primer positions in each locus. Solid boxes represent the flanking genes, hollow box represents EhSINE1 element and the arrow inside it shows the orientation of EhSINE1 with respect to the locus. (B) PCR was performed using the two primer pairs indicated in the Tables below, with the genomic DNA of different strains of E. histolytica as template. For each locus and strain PCR reactions using the two primer sets were mixed and resolved on a 1% agarose gel (upper panel); the gel was subjected to Southern blotting and hybridized with the locus-specific probe to check the specificity of the band pattern (middle panel). Hybridization was then performed with the EhSINE1 probe to check for the presence or absence of EhSINE1 in these loci in different strains (lower panel). The expected size of the amplicons for each locus is given in the tables at the bottom of the figure.

Figure 7

Strain identification in xenic cultures based on locus 13, 17 and 19: PCR was performed using genomic DNA of 16 different xenic cultures of E. histolytica for each locus, as described in Figure 6. PCR reactions were resolved on a 1 % agarose gel and subjected to Southern blotting with the locus specific probes 13 (Panel A), 17 (Panel B), or 19 (Panel C). Samples which did not give a product at locus 13 were amplified using an alternate reverse primer 13.1 R instead of 13.2 R followed by Southern blotting and hybridization with a locus specific probe. The expected size of the amplicon with each primer set is mentioned in the table below each locus panel.

Figure 8

Classification of EhSINE2. 119 EhSINE2 copies were extracted from the E. histolytica HM-1:IMSS database (length > 400 bp and similarity >70% with the consensus EhSINE2) and analyzed for IR; 111 could be categorized according to number of internal repeats, represented by the bars in Blue. The rest were excluded due to having a single copy in the database or having only a fraction of an IR in the SINE2. Correlation of TSD length and number of IR. Out of 119 SINE2s analyzed, TSDs were found in 92 cases (77.31%). All 92 examples with a TSD were analyzed for the number of IRs and average TSD length was plotted against IR number, represented by the bars in Red.