Mem Inst Oswaldo Cruz, Rio de Janeiro, 97 (Suppl.I) October 2002
Identification of Snails within the Bulinus africanus Group from East Africa by Multiplex SNaPshot™ Analysis of Single Nucleotide Polymorphisms within the Cytochrome Oxidase Subunit I
Wolfson Wellcome Biomedical Laboratories, Department of Zoology, The Natural History Museum, Cromwell Road, London SW7 5BD
IFaculty of Agricultural and Environmental Sciences, McGill University, Québec, Canada
IIDanish Bilharziasis Laboratory, Charlottenlund, Denmark
Identification of populations ofu00a0Bulinus nasutusu00a0andu00a0B. globosusu00a0from East Africa is unreliable using characters of the shell. In this paper, a molecular method of identification is presented for each species based on DNA sequence variation within the mitochondrial cytochrome oxidase subunit I (COI) as detected by a novel multiplexed SNaPshotu2122 assay. In total, snails from 7 localities from coastal Kenya were typed using this assay and variation within shell morphology was compared to reference material from Zanzibar. Four locations were found to containu00a0B. nasutus and 2 locations were found to containu00a0B. globosus. A mixed population containing bothu00a0B. nasutus andu00a0B. globosusu00a0was found at Kinango. Morphometric variation between samples was considerable and UPGMA cluster analysis failed to differentiate species. The multiplex SNaPshotu2122 assay is an important development for more precise methods of identification ofu00a0B. africanusu00a0group snails. The assay could be further broadened for identification of other snail intermediate host species.
The freshwater pulmonate snail genus Bulinus is divided into four species groups: B. africanus group, B. forskalii group, B. reticulatus group and the B. truncatus/tropicus complex (Brown 1994). Despite limited morphological divergence within species groups, there is considerable molecular divergence (Jones et al. 2001, Stothard et al. 2001). Within the B. africanus group 10 species are recognised and distributed throughout Afro-tropical regions and Madagascar. Several B. africanus group species are known, or suspected, to act as intermediate snail hosts for Schistosoma haematobium, a trematode parasite that causes urinary schistosomiasis.
Interactions between B. africanus group species and S. haematobium can be complex. Not all snail species act as intermediate hosts e.g. B. ugandae appears refractory to infection, or only certain snail species act as hosts in specific areas (Rollinson et al. 2001). Lack of clear-cut morphological characters hinders identification of natural populations (Mandahl-Barth 1965). For accurate separation of these snail species it is necessary to use biochemical (Rollinson & Southgate 1979) or molecular DNA methods (Rollinson et al. 2001). There are many single nucleotide polymorphisms (SNPs) within the mitochondrial cytochrome oxidase subunit I (COI) gene which may be exploited for species identification (Fig. 1). Taxon specific polymerase chain reaction (PCR) primers for B. globosus and B. nasutus have been designed based on genetic variation within the COI (Stothard et al. in press).
Whilst taxon specific primers are highly discriminatory, the PCR assay is limited within the known scope of detected sequence variation; further sequence variation may lead to false negatives. SNaPshot™ is a commercially available product from PE Biosystems, UK for genotyping SNPs using a fluorescent based, primer extension assay (Rollinson et al. 2001). Makridakis and Reichardt (2002) have taken the SNaPshot™ assay a step further by multiplexing SNaPshot™ primers of differing length coining the terminology 'multiplex automated primer extension analysis' (MAPA). Although this multiplex assay requires a semi-automated DNA sequencer, the assay has certain key advantages; many snails can be individually typed simultaneously for several key SNPs, and detection and precise characterisation of further DNA variation is possible.
This paper reports on the development and implementation of a multiplexed SNaPshot™ assay to type simultaneously four SNPs within the COI (Fig. 2) of Bulinus species. The assay is then used for identification of B. africanus group snails collected from 7 localities within coastal Kenya. The Kenyan shell material is compared to a selection of B. globosus and B. nasutus shells from Zanzibar to ascertain if there is any morphological divergence.
MATERIALS AND METHODS
Snail material - A total of 147 B. africanus group snails from 7 collecting localities from coastal Kenya: Kinango (n = 28), Mbovu (n = 24), Msambweni (n = 23), Mazeras (n = 6), Nguzo (n = 3), Ramisi (n = 25) and Timboni (n = 38), were examined. The snails were kindly provided in 70% ethanol by Dr M Otieno, National Museums of Kenya, Nairobi, Kenya. Genomic DNA was extracted from each snail for PCR analysis (see below) following standard methodology (Stothard et al. 1997) and the shell was retained for morphometric analysis. A collection of 79 dry shells as reported by Stothard et al. (1997) of B. africanusgroup snails from Zanzibar [B. globosus Unguja (n = 18), Pemba (n = 22) and B. nasutus Unguja (n = 18), Pemba (n = 21)] was used as reference material for morphometric comparison with Kenyan material.
PCR and SNaPshot™ typing of COI - A subregion (~ 450 bp) of the COI was amplified from each snail following conditions described by Stothard and Rollinson (1997). The amplification product was then purified using a Qiagen PCR clean-up spin column and adjusted to an approximate concentration of 0.2 pmol/µl.
Four SNaPshot™ primers were designed to type SNPs at positions 49, 109, 166 and 255 (Fig. 1) in the alignment of COI sequences described by Stothard and Rollinson (1997). The primer sequences were as follows:
Position 255 (29mer) 5' _ TAAAAAGAAAAATAAAYCC TAAYACTCAA
Position 166 (34mer) 5' _ ACACCTTAATTCCTGTTGG TACAGCAATAATTAT
Position 109 (38mer) 5' _ GCTCGAGTATCCACATCTATT CCHACAGTAAATATATG
Position 49 (45mer) 5' _ TAAACCTAAAATTCCAATT GAAACTATWGCATAAATTATTCCTAA
1µl of purified amplification product was added to a SNaPshot™ reaction mixture containing 0.15 pmols of each primer. The reaction mixture (10 µl) was incubated for 25 cycles of 96oC for 10 sec, 50oC for 5 sec and 60oC for 30 sec in a standard thermal cycler. The SNaPshot™ reaction mixture was purified by removal of excess fluorescent dye terminators by incubation with 1unit of calf intestinal alkaline phosphatase according to manu-facturer's instructions.
1.5 µl of purified SNaPshot™ reaction was combined with 1.5 µl deionised formamide/blue dextran and the total volume was loaded onto an ABI 377 machine using a standard 5% long ranger, 6M urea sequencing gel. Electrophoresis was conducted according to standard separation conditions detailed in the PE Biosystems' electrophoresis manual. As the 4 SNaPshot™ primers were of differing length, upon separation by denaturing gel electrophoresis each SNP position could be assigned according to the order in which the now fluorescently labelled primers were separated (Fig. 2). The subsequent gel file was visually inspected and also processed with GeneScan software version 2.1 (PE Biosystems, UK).
DNA sequencing - In addition to SNaPshot™ analysis, the DNA sequence of the COI was determined for a total of 21 snails taken across the 7 sampling localities [Kinango (n = 7), Mbovu (n = 3), Msambweni (n = 3), Mazeras (n = 3), Nguzo (n = 1), Ramisi (n = 2) and Timboni (n = 2)] using direct DNA cycle sequencing of purified amplification product and separated on an ABI377 semi-automated DNA sequencer.
Morphometric analysis - A total of 226 snail shells were examined. Identification of Kenyan B. globosusand B. nasutus was based upon the results of the SNaPshot™ COI profile. Digital images of the shells (aperture facing) at either x8 or x12 magnification were collected using a Leica MZ6 dissecting microscope with an attached digital camera and DIC-E image capture software (World Precision Instruments, UK). A 10 mm scale bar was included within each shell image for size calibration. Shell microsulpture was also noted according to descriptions given by Kristensen et al. (1987). Image analysis was conducted with SigmaScanPro 4.0 software package (Jandel Scientific, UK).
Nine measurements were taken for each shell: 5 linear, 3 area and 1 angular. These were, linear (mm) L: total length of shell, W: width of shell (perpendicular from aperture apex to opposite edge), WSp: basal width of shell spire, LAp: length of aperture, WAp: width of aperture; area (mm2), TA: total shell area, Asp: area of spire and AAp: area of aperture; and angular (degrees): angle subtended from longitudinal axis of shell rotation to maximum width point on the outer body whorl edge.
Before analysis, area measurements were transformed by taking the square root. All measurements were then standardised using the geometric mean of Log10 adjusted ratios following Clarke et al. (1999). Euclidean distances between each individual were calculated using the program SYN-TAX 5.0 (J Podani, Scientia Publishing, Budapest) and a dendrogram was generated using unweighted pair-group arithmetic average (UPGMA) analysis to evaluate phenetic groups.
Identification and COI variation - SNaPshot TM COI reactions from all snails could be readily assigned to the expected profile for either species (Fig. 3), and no further variation within the 4 variant positions was detected. From the SNaPshot TM COI profile B. globosus was encountered at Timboni and Mazeras whileB. nasutus was encountered at Mbouv, Msambweni, Nguzo and Ramisi. A mixed population of B. globosus and B. nasutus was found at Kinango. Of the 21 COI sequences obtained and in comparison to previous COI data from B. africanus group snails, 8 novel COI sequences were encountered. Three B. globosus and 5 B. nasutus sequences have been deposited in GenBank, accession numbers AF507035-AF507042. The majority of B. globosus from Kenya did not have a SspI restriction enzyme cutting site within the COI.
Morphometric variation - The Table details the mean value with 1 standard deviation for shell measurements. A significant difference (p < 0.05) was detected between B. globosus and B. nasutus with an unpaired student t test for ASp and AAp. Generally, B. globosus had larger aperture and smaller spire areas than B. nasutus. Plotting a combination of shell variables in bivariate plots against shell length did not reveal the existence of two discrete distributions of points (Fig. 4).
UPGMA analysis of morphometric variation within Kenyan and Zanzibarian snails resulted in a dendrogram containing 8 major clusters, designated A to H (Fig. 5). The largest clusters, B and H, contained 64 and 65 snails respectively. No single cluster contained exclusively B. globosus or B. nasutus. Both species were distributed across clusters although the cluster B was predominately composed of Kenyan B. globosus. Following from the identification equations proposed by Kristensen et al. (1987) using presence of microsculpture and aperture banding, a specificity of 53% for identification of B. globosus was found. Approximately 1 in 2 B. globosus snails would have been identified correctly within this sample.
Multiplex SNaPshot™ assay - The multiplex SNaPshot™ assay is a robust methodology for typing SNPs within the COI from individual snails. Analysis and interpretation of the resultant SNaPshot™ profile is less labour intensive than inspection of a corresponding DNA sequencing chromatogram as only 4 fragments are produced (Fig. 2). Each characteristic colour bar-code is immediately apparent from the corresponding gel file picture (Fig. 3). As such identification of many snails can be quickly accomplished and as this technology requires minimal liquid handling steps, snails can be processed in microtitre plate format. This multiplex assay has great potential for regular, high-through-put DNA typing in a convenient single tube reaction. Multiplex SNaPshot™ assays may also simultaneously screen for variation within other candidate genes e.g. ribosomal 18S and Internal Transcribed Spacer, and once sufficient DNA sequence data has been accrued for representatives of Bulinus, the assay could no doubt be extended. The methodology might also be of use for identification of problematic populations of Lymnaea (Bargues et al. 1997)or Biomphalaria (Morgan et al. 2001) or even in the detection of Schistosoma sequences from infected snails.
Snail populations - Accurate identification of B. glo-bosus and B. nasutus populations is important. Recent work on Zanzibar has shown that B. nasutus plays no role in transmission of local S. haematobium and the distribution of urinary schistosomiasis in schoolchildren is clearly linked with the distribution of B. globosus(Stothard et al. in press). A similar transmission picture might also exist in Kenyan and Tanzanian coastal regions nearby.
Both B. globosus and B. nasutus have been encountered within the Kenyan material. While 6 of the 7 samples contained a single species, the sample from Kinango contained a mixed population of both species. This situation contrasts somewhat with Zanzibar where mixed populations have yet to be encountered (Stothard et al. 1997). In Zanzibar, B. globosus appears to be associated with harder water while B. nasutus is associated with softer water but the range of water hardnesses between species does overlap (Stothard et al. in press); potentially, mixed populations could occur in nature. As water conductivity values were not recorded at the time of collection of this Kenyan material, it might be expected that water conductivity at Kinango could be within this overlapping range.
The DNA divergence within the COI between Kenyan and Zanzibarian species is confined to several further point mutations. It is worthy of note that the PCR-RFLP methodology designed by Stothard and Rollinson (1997) to differentiate B. globosus and B. nasutus in Zanzibar on the presence and absence of aSspI cutting site respectively (recognition sequence AATATT) does not work consistently on Kenyan B. globosus. Coastal Kenyan B. globosus has two predominant sequence types present within this selected 6 bp region: AATATT (SspI site) and GATATT (non-SspI site), while all B. nasutus have AATATA. Whilst this assay can positively identify certain B. globosus by the presence of an SspI site, in this case the snails from Kinango, the absence of a SspI restriction site cannot now be interpreted as exclusive to B. nasutus. Confident typing of the 3' T in B. globosus and 3' A in B. nasutus which differentiates the species, is best performed by SNaPshot™ or taxon specific PCR primers assays.
Shell morphology - It is known that the shell of B. africanus group snails from East Africa can be highly variable (Brown 1994). The problem was identified by Mandahl-Barth (1965) as long-spired and short-spired forms of either species `overlapped' making separation of populations and individuals of either species problematic. Morphometric analysis presented here has confirmed that it is difficult to separate either species on shell characters alone. The bivariate plots are particularly illustrative of the overlapping nature of variation. Whilst the mean aperture area (AAp) of each species is significantly different, for example aperture area is greater in B. globosus than B. nasutus for a typical shell, this measurement is not capable of precisely differentiating species (Fig. 4).
Variation within other shell variables has been inspected with the hope of finding clear-cut divisions between species. Kristensen et al. (1987) developed a simple, powerful morphometric test which could differentiate B. africanus group species from the Lake Victoria region of Kenya. Using these equations, however, for the Kenyan and Zanzibarian material there is incongruence between molecular and morphological methods of identification. This has also been recently reported by Raahauge and Kristensen (2000) upon analysis of B. africanus group snails from Kisumu region Kenya. The presence and absence of shell microsculpture has been used to differentiate B. nasutus and B. globosus respectively from Lake Victoria but upon inspection of further snails from different areas it seems that this character becomes less species specific. In Zanzibar shell microsculpture does not help to differentiate the species. Certain populations confidently typed by both genetic and biochemical methods as B. globosus do have microsulpture, as well as, aperture bands whereas others do not. Inspection of the clusters generated by UPGMA analysis of shell measurements (Fig. 5) adds further weight to the suggestion made by Stothard et al. (1997) that there may be few, if any, conchological measurement useful for reliable identification.
Conclusion - Correct identification of populations of B. africanus group snails requires the use of biochemical or molecular DNA methods. The multiplex SNaPshot™ assay is a promising new tool for species identification. New data are required concerning the distribution and transmission status of potential intermediate hosts in East Africa in order to gain a fuller understanding of the distribution of urinary schistosomiasis.
Bargues MD, Mangold AJ, Muñoz-Antoli C, Pointier JP, Mas-Coma S 1997. SSU rDNA characterisation of lymnaeid snails transmitting human fascioliasis in South and Central America. J Parasitol 83: 1086-1092.
Brown DS 1994. Freshwater Snails of Africa and their Medical Importance, 2nd ed., Taylor & Francis Ltd., England, 609 pp.
Clarke RK, Grahame J, Mill PJ 1999. Variation and constraint in the shells of two sibling species of intertidal rough periwinkles (Gastropoda: Littorina spp). J Zool Lond 247: 145-154.
Jones CS, Rollinson D, Mimpfoundi R, Ouma J, Kariuki HC 2001. Molecular evolution of freshwater intermediate snail hosts within the Bulinus forskalii group. Parasitology 123: S277-S292.
Kristensen TK, Frandsen F, Christensen AG 1987. Bulinus africanus-group snails in East and South East Africa, differentiated by use of biometric multivariate analysis on multivariate characters (Pulmonata: Planorbidae). Rev Zool Afric 101: 55-66.
Makridakis NM, Reichardt JKV 2002. Multiplex automated primer extension analysis: simultaneous genotyping of several polymorphisms. Biotechniques 31: 1374-1380.
Mandahl-Barth G 1965. The species of Bulinus, intermediate hosts of Schistosoma. Bull WHO 33: 33-44.
Morgan JAT, DeJong RJ, Synder S, Mkoji GM, Loker ES 2001. Schistosoma mansoni andBiomphalaria: past history and future trends. Parasitology 123: S211-S228.
Raahauge P, Kristensen TK 2000. A comparison of Bulinus africanus group species (Planorbidae: Gastropoda) by the use of the internal transcribed spacer 1 region combined by morphological and anatomical characters. Acta Trop 75: 85-94.
Rollinson D, Southgate VR 1979. Enzyme analyses of Bulinus africanus group snails (Mollusca: Planorbidae) from Tanzania. Trans R Soc Trop Med Hyg 73: 667-672.
Rollinson D, Stothard JR, Southgate VR 2001. Interactions between intermediate snail hosts of the genus Bulinus and schistosomes of the Schistosoma haematobium group. Parasitology 123: S245-S260.
Stothard JR, Rollinson D 1997. Partial DNA sequences from the mitochondrial cytochrome oxidase subunit I (COI) gene can differentiate the intermediate snail hosts Bulinus globosus and B. nasutus(Gastropoda: Planorbidae). J Nat Hist 31: 727-737.
Stothard JR, Brémond P, Andriamaro L, Sellin B, Sellin E, Rollinson D 2001. Bulinus species on Madagascar: molecular evolution, genetic markers and compatibility with Schistosoma haematobium.Parasitology 123: S261-S275.
Stothard JR, Mgeni AF, Alawi KS, Savioli L, Rollinson D 1997. Observations on shell morphology, enzymes and random amplified polymorphic DNA (RAPD) in Bulinus africanus group snails (Gastropoda: Planorbidae) in Zanzibar. J Moll Stud 63: 489-503.
Stothard JR, Mgeni AF, Khamis S, Seto E, Ramsan M, Hubbard SJ, Kristensen TK, Rollinson D. New insights into the transmission biology of urinary schistosomiasis in Zanzibar. Trans R Soc Trop Med Hyg, in press.