Genetic epidemiology of Sarcoptes scabiei - ResearchOnline@JCU

International Journal for Parasitology 34 (2004) 839–849 www.parasitology-online.com

Genetic epidemiology of Sarcoptes scabiei (Acari: Sarcoptidae) in northern Australiaq S.F. Waltona,b,*, A. Dougalla,b, S. Pizzuttoa,b, D. Holtd, D. Tapline, L.G. Arlianf, M. Morganf, B.J. Curriea,b,c, D.J. Kempd a

Menzies School of Health Research, Darwin, NT, Australia b Charles Darwin University, Darwin, NT, Australia c Northern Territory Clinical School, Flinders University, Darwin, NT, Australia d Queensland Institute of Medical Research, Brisbane, Qld, Australia e Department of Dermatology and Cutaneous Surgery, University of Miami, Miami, Fl, USA f Department of Biological Sciences, Wright State University, Dayton, OH, USA Received 16 March 2004; received in revised form 31 March 2004; accepted 1 April 2004

Abstract Utilising three hypervariable microsatellite markers we have previously shown that scabies mites on people are genetically distinct from those on dogs in sympatric populations in northern Australia. This had important ramifications on the formulation of public health control policies. In contrast phylogenetic analyses using mitochondrial markers on scabies mites infecting multiple animal hosts elsewhere in the world could not differentiate any genetic variation between mite haplotype and host species. Here we further analyse the intra-specific relationship of Sarcoptes scabiei var. hominis with S. scabiei var. canis by using both mitochondrial DNA and an expanded nuclear microsatellite marker system. Phylogenetic studies using sequences from the mitochondrial genes coding for 16S rRNA and Cytochrome Oxidase subunit I demonstrated significant relationships between S. scabiei MtDNA haplotypes, host species and geographical location. Multi-locus genotyping using 15 microsatellite markers substantiated previous data that gene flow between scabies mite populations on human and dog hosts is extremely rare in northern Australia. These data clearly support our previous contention that control programs for human scabies in endemic areas with sympatric S. scabiei var. hominis and var. canis populations must focus on human-to-human transmission. The genetic division of dog and human derived scabies mites also has important implications in vaccine and diagnostic test development as well as the emergence and monitoring of drug resistance in S. scabiei in northern Australia. q 2004 Australian Society for Parasitology Inc. Published by Elsevier Ltd. All rights reserved. Keywords: Sarcoptes scabiei; Scabies; 16S rRNA; Cytochrome oxidase; Microsatellite; Phylogeny

1. Introduction The mite Sarcoptes scabiei is the causative organism of scabies, a debilitating skin condition endemic in many q Supplementary data associated with this article can be found, in the online version, at doi:10.1016/S0XXX-XXXX(XX)XXXXX-X. Microsatellite nucleotide sequence data reported in this paper are available in GenBanke, EMBL, DDBJ databases under accession numbers AY322558–AY322569. MtDNA COI sequence data under accession numbers AY493379– AY493398 and 16S rRNA sequence data under accession numbers AY493399 – AY493412. 16S rRNA sequence alignments are available as supplementary files. * Corresponding author. Address: Menzies School of Health Research, P.O. Box 41096, Casuarina, NT 0811, Australia. Tel.: þ 61-8-8922-8928; fax: þ 61-8-8927-5187. E-mail address: [email protected] (S.F. Walton).

disadvantaged populations worldwide and responsible for epizootic mange in many wild and domestic animal populations (Pence and Ueckermann, 2002). In Aboriginal communities in northern Australia conditions of reduced socioeconomic status, inadequate medical facilities, and overcrowding appear to be contributing to high levels of endemic scabies (Munoz et al., 1992; Currie and Carapetis, 2000). Concomitantly, the communities are characterised by large dog populations that often share communal space with people. These dogs also suffer a high prevalence of scabies and other diseases and it has been postulated they present a considerable risk for the transfer of zoonotic disease. Currently host-associated populations of S. scabiei are taxonomically divided into morphologically indistinguishable varieties that have a high degree of host specificity

0020-7519/$30.00 q 2004 Australian Society for Parasitology Inc. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.ijpara.2004.04.002

840

S.F. Walton et al. / International Journal for Parasitology 34 (2004) 839–849

and low degree of cross infectivity (Fain, 1978). The varieties are named based on their host species, e.g. S. scabiei var. hominis, S. scabiei var. canis. Historically, genetic research on scabies has been extremely limited (reviewed in Walton et al., 2004). This is primarily due to the difficulty in obtaining sufficient quantities of the mite and usable amounts of genetic material. In ordinary scabies, often less than 10 organisms can be identified per host. The mites are very small (, ¼ 0.4 mm), often cannot be found on hosts exhibiting clinical symptoms, and there is no in vitro culture system. Walton et al. (1997) utilised shed skin from the bedding of hospitalised crusted scabies patients as a non-invasive source of up to 4000 mites per gram of skin. This resulted in the first molecular study on over 700 scabies mites (Walton et al., 1999). Utilising three hypervariable microsatellite markers it was found that scabies mites on people are genetically distinct from those on dogs in sympatric populations in northern Australia. This study suggested that interbreeding or cross-infection between S. scabiei from humans and dogs living in northern Australian remote Aboriginal communities appeared to be extremely rare. This had important ramifications on the formulation of public health control policies and successful community control programs were designed targeting human-to-human transmission only (Carapetis et al., 1997; Wong et al., 2001, 2002). Although these data may imply that these sub-species deserve biological species status, other studies suggest monospecificity of the genus Sarcoptes, presumably a result of interbreeding. Zahler et al. (1999), using a 450 bp nuclear ribosomal marker ITS-2, studied 21 mites derived from populations of dog, pig, cattle, fox, lynx, wombat, dromedary, and chamois, but were unable to see any association between mite haplotype and host species. Berrilli et al. (2002) also used the ITS-2 marker, as well as 407 bp of the mitochondrial 16S rRNA gene, in a study looking at genetic variation in 28 mites derived from chamois and fox populations in northern Europe. They were able to distinguish clusters in mite populations as a result of geographic isolation but could see no host association of mites using both mitochondrial and nuclear markers. Another study on 23 mites derived from wombats, dogs, and human host populations in Australia, using a 326 bp fragment of the mitochondrial 12S rRNA gene, could not differentiate any genetic variation between host populations (Skerratt et al., 2002). The polyphyletic associations between host strains of these latter studies appear to be based on short uninformative fragments of mitochondrial or ribosomal DNA spacer regions. Patterns of variation in the DNA of the 12S and 16S rRNA mitochondrial genes, especially the third domain (amplified by universal primers), appear to have limited phylogenetic usefulness for assessing relationships among recently diverged populations within species because few sites vary (Simon, 1991). A number of studies suggest however the 50 region may have more useful information and demonstrate

less homoplasy than the 30 region. Protein coding genes may be more useful than rRNA genes for close levels of divergence due to more accurate alignment of codons compared with ribosomal stems and loops. Although the protein sequence of Cytochrome Oxidase subunit I (COI) is relatively conserved, the coding sequences contain numerous sites where synonymous substitutions can occur. Consequently, in this study we chose to look at the coding sequences of a central fragment of the 16S gene and the complete COI gene. Given the lack of knowledge on divergence times and evolutionary rates, the use of two mitochondrial genes allowed us to examine levels and patterns of variation in each marker and determine their applicability towards understanding the phylogeny of the group. Gene flow between parasite populations on different host species can be limited even if parasites are not 100% host specific. Such physical restrictions can favour the formation of host races, or speciation (Johnson et al., 2002). In phylogenetic analyses bootstrapping support for the closest relationships may be relatively poor due to reduced time to accumulate informative changes in the sequence regions examined. Further resolution is therefore provided in faster evolving hypervariable sequences such as nuclear polymorphic microsatellite sequences. On this basis we developed an extended 15 microsatellite marker system to further study gene flow between sympatric host associated populations of S. scabiei in northern Australia. Due to the frequency of complex mutations reported for microsatellites, suggesting that absolute fragment size (as normally scored in microsatellite studies) is an unreliable indicator of historical affinities among alleles (Anderson et al., 2000) we based our analysis on proportion of shared alleles. This has previously demonstrated strong support for geographically discrete populations and at the population level also showed congruence with evolutionary patterns (Bowcock et al., 1994). It was envisaged evaluation of hypervariable nuclear markers in combination with the phylogenetic analysis would provide additional support for genetic differentiation of S. scabiei.

2. Materials and methods 2.1. Mite collection and DNA extraction Scabies mites were obtained from human and dog hosts in remote Aboriginal communities of the Northern Territory (NT), plus a number of other hosts and localities from around the world and stored at minus 80 8C. Each mite analysed was obtained from a different individual host except the wombat mites which were collected from the same animal (Table 1). Male mites were identified as smaller than females and darker than nymphs. DNA extractions for PCR were carried out using methods described by Walton et al. (1997).


841

Table 1 Host species, country of origin, and mitochondrial haplotype of mites typed for mitochondrial and microsatellite loci

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 a b

Codes of individual mites

Host Species

Location

Haplo-type COI 21/26

Haplo -type 16S 14/24

Haplo -type COI/ 16S

Var. hominis 3 Var. hominis 4 Var. hominis 8 Var. hominis 9 Var. hominis 11 Var. hominis 1 Var. hominis 5 Var. hominis 6 Var. hominis 2 Var. hominis 7 Var. hominis 31 Var. hominis 1117 Var. hominis 208 Var. hominis 13 Var. hominis 20 Var. hominis 14 Var. hominis 15 Var. hominis 16 Var. hominis 932 Var. hominis 607 Var. hominis 10 Var. hominis 205 Var. hominis 200 Var. hominis 320 Var. canis 12 Var. canis 10 Var. canis 202 Var. canis 5 Var. canis 1 Var. canis 2 Var. canis 19 Var. canis 4 Var. canis 17 Var. canis 9 Var. canis 22 Var. canis 55 Var. canis 25 Var. canis 26 Var. canis 31 Var. canis 32 Var. canis 11 Var. canis 12 Var. canis 19 ex chimpanzee 1 Var. wombati 7 Var. wombati 13 Var. wombati 14 Var. wombati 15 Var. wombati 20 Var. wombati 9 ex wallaby 1 Var. vulpes

Human Human Human Human Human Human Human Human Human Human Human Human Human Human Human Human Human Human Human Human Human Human Human Human Dog Dog Dog Dog Dog Dog Dog Dog Dog Dog Dog (ex rabbit)b Dog (ex rabbit) Dog Dog Dog Dog Dog (ex rabbit) Dog (ex rabbit) Dog (ex rabbit) Chimpanzee Wombat Wombat Wombat Wombat Wombat Wombat Wallaby Fox

Panama Panama Panama Panama Panama Panama Panama Panama Panama Panama NT, Australia NT, Australia NT, Australia NT, Australia NT, Australia NT, Australia NT, Australia NT, Australia NT, Australia NT, Australia NT, Australia NT, Australia NT, Australia NT, Australia NT, Australia NT, Australia NT, Australia NT, Australia NT, Australia NT, Australia NT, Australia NT, Australia NT, Australia Ohio, USA Ohio, USA Ohio, USA Ohio, USA Ohio, USA Ohio, USA Ohio, USA Ohio, USA Ohio, USA Ohio, USA Tanzania SA, Australia SA, Australia SA, Australia SA, Australia SA, Australia SA, Australia NT, Australia Sweden

1 2 3 NAa

1 2 3 3

1 2 3 NA

Geno -typed for 15 Micro-satellite loci

Yes Yes Yes Yes Yes Yes 4 4 4 5 NA 6 7 7 8 9 10 11 12 NA 13 13 14

NA 4 4 4 4 5 NA 6 NA 7 NA 8 9 10 11 11 11

NA 4 4 5 NA 6 NA 7 NA 8 NA 10 11 NA 12 12 13

15 NA 16

12 12 12

14 NA NA

Yes Yes Yes Yes Yes Yes Yes Yes Yes

Yes Yes Yes Yes 17 15 NA

NA NA 12

NA NA NA Yes Yes Yes Yes Yes Yes Yes

18 19

13 12

15 16

20 21

14 NA

17

Yes Yes Yes Yes Yes Yes Yes

NA ¼ not amplified. ex rabbit ¼ passaged on rabbits.

2.2. PCR amplification, cloning and sequencing We sequenced 749 bp of the CO1 gene from 25 mites collected from 15 humans and seven dogs from five

locations, one chimpanzee, one wallaby and one wombat. Sequences were amplified using primers CytoF, 50 -AAGATTTATTGTACCATTAGA-30 and CytoR, 50 ATTTTTATATCAACATTTA-3 0 . The products were

842


cloned into pBluescript II KS vectors (Stratagene, Integrated Sciences, NSW, Australia) and sequenced in both directions using standard methods. Multiple overlapping S. scabiei var. vulpes ESTs were obtained from GenBank (accession numbers BG817648, BG817696, BG817769, BG817791, BG817818, BG817877, BG817941, BM522091, BM522113, BM522174, BM522237, BM522257). The fox mite sequences were aligned using ClustalW (accurate) (Thompson et al., 1994) and a consensus sequence determined. We sequenced 375 bp of the 16S rRNA gene corresponding to the region of the Ixodes hexagonus 16S gene from 520 to 900 bp (Genbank accession number AF081828). 16S was amplified from a total of 24 individual S. scabiei mites collected from 14 human hosts, seven dog hosts, one chimpanzee, one wallaby and one wombat host, originating from five geographically distinct localities (Table 1). Sequences were amplified using primers 16S D1, 5 0 -CTAGGGTCTTTTTGTTCTTGG-30 and 16S D2, 50 -GTAAGTATACGTTGTTATAAC-30 . The products were cloned into pGEM-T Easy Vector System (Promega, NSW, Australia) and sequenced in both directions using standard methods. In order to discover microsatellites an expressed sequence tag (EST) database of 8544 S. scabiei var. hominis sequences, established through the Scabies Gene Discovery Project (Fischer et al., 2003), was BLAST searched for CT repeats. More than 50 repeat sequences were identified. Twenty-four primer pairs were designed for sequences comprising a complete and uninterrupted repeat region, with a minimum of 10 repeat units. The final fingerprinting system was based on 13 highly polymorphic dinucleotide repeats; Sarms 1, 15 and 20 from Walton et al. (1999) and Sarms 23, 33, 34, 35, 36, 37, 40, 41, 44, 45 identified here. Two slightly polymorphic microsatellite loci, Sarms 31 (tri-nucleotide repeat) and 38 (dinucleotide repeat) were also employed. Primers were synthesised by Applied Biosystems (ABI) Custom Oligo Synthesis Service, (Melbourne, Australia) with one primer from each set 50 labelled with FAMe, VIC, NEDe (ABI PRISMw filter set D) fluorescent dye tag. Amplification was achieved in 25 ml reactions using 1 unit HotStarTaq DNA Polymerase and PCR Buffer (Qiagen, Australia), 1.5 mM MgCl2, 0.5 mM primers, 0.2 mM dNTP’s and 1 ml template genomic DNA. Microsatellite loci were amplified from a total of 33 individual S. scabiei mites collected from 14 human hosts, seven dog hosts, one wallaby and one wombat host, originating from four geographically distinct localities (Table 1). Allele lengths were determined by the Australian Genome Research Facility (Melbourne, Australia) using an ABI 377 DNA Sequencer. Electrophoresed samples were tracked using Prism Genescan version 3.1.2 software (Applied Biosystems) and then exported to Genotyper version 2.1 (Applied Biosystems).

2.3. Sequence and phylogenetic analysis Sequences were aligned, compared and edited using the EditSeq, Megalign and SeqMan algorithms from Lasergene (DNASTAR Inc., Madison, WI). Each mutation was confirmed on the chromatograms from sequencing results. Individual mite consensus sequences were then manually trimmed of primer sequence and aligned using ClustalW (Thompson et al., 1994). As an outgroup we used the published COI and 16S sequences for I. hexagonus, a species of tick (Genbank accession number AF081828) (Black and Roehrdanz, 1998). These were taxonomically the closest sequences to S. scabiei published in Genbank for the complete COI and 16S genes but were later shown to be too distantly related and removed from further analyses as they had no effect on the outcome. The alignments were bootstrapped using Seqboot (Felsenstein, 1989) and phylogenetic analyses including maximum likelihood (DNAml), maximum parsimony (DNApars), and distance methods (DNAdist) were performed using the Australian National Genomic Information Service (ANGIS) suite of programs (Felsenstein, 1989) and MEGA version 2.1 (Kumar et al., 2001). Distance analyses were estimated using the Tamura3-parameter as this model corrects for multiple hits taking into account the differences in transitional and transversional rates and G þ C-content bias (Tamura, 1992). Trees were generated using minimum evolution and neighbourjoining algorithms. Phylogenies were estimated for each gene independently and in combination, with support for the distance and parsimony algorithms measured by bootstrapping over 500 replicates. 2.4. Microsatellite data As observed in previous studies (Walton et al., 1999) a number of individual mites failed to amplify a product at some markers indicating the presence of null alleles. In total null alleles were present in nine out of 15 loci, thus estimating unbiased allele frequencies without also assuming Hardy Weinberg Equilibrium (HWE) was not possible. This is frequently observed in studies with microsatellites (Pemberton et al., 1995). GENEPOP version 3.3 (ftp://ftp.cefe.cnrs.mop.fr/pub/pc/msdos/genepop/) was used to compute non-parametric exact tests for deviations from HWE, allelic and genotypic population differentiation and genotypic disequilibrium (Raymond and Rousset, 1995a). These were assessed for the five hostassociated populations and six loci in which no null alleles were displayed (even though this was a biased selection). The coancestry coefficient Fst was calculated for population pairs using the same six loci. Genetic distance based on the proportion of shared alleles was used to measure the relationship between multilocus genotypes of individual mites using methods described by Walton et al. (1999). Matrices were constructed from pairwise distances using MICROSAT


(Stanford University, Stanford, CA), with dendrograms based on hierachial clustering computed with STATA version 7. Two distance matrices were compared, one assuming all single alleles were homozygous, the other assuming all single alleles were heterozygous-nulls. Population assignment analysis was also used to identify the population of origin of an individual (Paetkau et al., 1995; Rannala and Mountain, 1997). The program GENECLASS (http://www.montpellier.inra.fr/CBGP/softwares) provides a method for assigning individuals of unknown origin to populations based on the likelihood of multilocus genotypes. We used both the ‘direct’ method, as well as the simulation-based ‘Bayesian’ method that provides a probability value to each assignment (Cornuet et al., 1999). Direct and simulation tests were carried out among all five mite populations using the ‘leave one out’ procedure, and for assigning individual mites to host-associated populations.

3. Results 3.1. Mitochondrial DNA sequence analysis In total the 26 COI sequences analysed comprised 21 unique haplotypes. A total of 11.9% (89/747) nucleotide positions were variable, and 8.8% (66/747) were parsimonyinformative sites, with 5% demonstrating substitutions at four-fold degenerate sites. Alignments of the predicted amino acid sequences revealed 11/249 (4.4%) variable sites among the haplotypes, indicating some of the substitutions were non-synonymous. Base composition was biased, averaging 73% A þ T, similar to that observed previously in Sarcoptes and other mites (Navajas et al., 1996; Zahler et al., 1999; Skerratt et al., 2002). The transition/ transversion ratio was 5. To demonstrate relationships between haplotypes, several trees were constructed using distance approaches, maximum parsimony and maximum likelihood. All trees constructed were very similar in topology with S. scabiei mtDNA haplotypes phylogenetically partitioned into three main groups consisting of the human derived mites from Panama (Group A), human derived mites from NT (Group B), and a mixed group of animal and human derived mites (Group C). The latter result apparently contradicted divergence observed in previous studies using microsatellite loci (Walton et al., 1999). The average nucleotide pairwise distance between groups was between 5.3 and 7.3%. The Neighbour-Joining tree is shown in Fig. 1. Fourteen 16S rRNA sequence haplotypes were obtained from 24 mite sequences. Within this fragment a total of 9% (29/310) nucleotide positions were variable, of which 7% (22/310) were parsimony-informative sites and 2% were singleton substitutions (nucleotide sequence alignment provided as Supplement 1). Base composition was biased, averaging 87% A þ T with a transition/transversion ratio of

843

0.8. Phylogenetic trees again identified three main groups consisting of the human derived mites from Panama, human derived mites from the NT, and a mixed group of animal and human derived mites (data not shown). The average pairwise distance between groups was between 5.6 and 6.6%. 3.2. Microsatellite data Allele counts for each of the 15 microsatellite loci ranged from five to 27, with an average of 12 alleles per loci. Sarms 31 was monomorphic for allele 145 in all populations except the Panama human mite population in which alleles 151 and 154 were observed. HWE estimates were assessed for host-associated mite populations at the six loci displaying no null alleles and a highly significant deviation from HWE was only observed for the human-derived mites from Panama (Table 2). The all-locus, all-population estimate indicated highly significant departure from HWE. Genotypic disequilibrium was evaluated for each pair of loci in all host-associated populations and across all populations and no significant deviations were observed. Genotypic differentiation using a G-like exact test over all host-associated populations, assuming statistical independence over all 15 loci, again suggested highly significant differences between mites collected from different host-associated populations and at individual loci, P , 0:00001 (Chi 2 ¼ infinity, df ¼ 30) (Fisher’s method). In addition, allelic differentiation (using loci displaying no null alleles) was highly significant between all population pairs including sympatric NT dog and NT human mites, P , 0:00001 (Fisher exact test). Population differentiation for NT dog and NT human mite populations (again using loci displaying no null alleles) gave an overall Fst comparison of 0.1665 (Weir and Cockerham, 1984). As these detailed studies on a smaller number of mites support our previous microsatellite study based on over 700 mites (Walton et al., 1999) further analyses on a larger numbers of mites were deemed unnecessary. Multi-locus relationships of the 31 individual mites in the microsatellite study were determined using the distance algorithm of Bowcock et al. (1994). Trees were formed using STATA 7 hierarchical clustering via single, average, and complete linkage. All clustering methods produced the same result with individual mites grouping into distinct host-associated groups showing a high degree of dissimilarity. Fig. 2A displays the dendrogram compiled when all single alleles are assumed to be homozygous whereas Fig. 2B assumes all single alleles are heterozygous-nulls. Neither assumption alters the overall clustering and supports the robustness of the analysis. Furthermore, analyses undertaken using only the loci displaying no null alleles resulted in comparable genetic differentiation. This result is clearly analogous with the mitochondrial phylogenetic analysis and our previous microsatellite study with three markers and over 700 mites (Walton et al., 1999). The analyses separate the human-derived individual

844


Fig. 1. Neighbour-joining tree clustering Sarcoptes scabiei Cytochrome Oxidase subunit I sequence haplotypes into three distinct groups. Group A (human derived mites from Panama), Group B (human derived mites from Northern Territory) and C (human and animal derived mites). The sequences were aligned using ClustalW and the distance matrix constructed using the Tamura 3-parameter index.

mites into two distinct geographical clusters and also further divide the animal-derived mites into distinct host-associated groups. Interestingly, the NT Wallaby mite clusters with the NT Dog mites suggesting that this infestation was derived from the regional dog population. Assignment tests on the microsatellite data using the direct method and leave one out procedure of Cornuet et al. (1999) demonstrated 100% correct assignment of individual mites to their population of origin (using the five populations and 15 loci). Additionally, the direct method and ‘assignation of unknown data to a reference’ was able to

assign the three NT human-derived mites (var. hominis 1117, 13, and 14) with haplotypes that clustered in Group C in the mitochondrial analysis, into the NT human-derived population of mites. The exclusion-simulation significance test correctly assigned 69% of individuals when using a threshold P of 0.01 for each individual assignment. 3.3. Elimination of haplodiploidy as a possible confounder Finally, we screened 22 S. scabiei var. hominis mites, collected from the NT, for the presence of haplodiploidy.


845

Table 2 Hardy–Weinberg Equilibrium for host-associated mite populations at the six microsatellite loci displaying no null alleles Population

Number of individuals

Sarms 35 P

Sarms 37 P

Sarms 38 P

Sarms 40 P

Sarms 41 P

Sarms 44 P

All (Fishers method) P

NTDog NTHuman Wombat PanamaHuman USADog

5 8 6 5 7

0.0476 0.9424 1.0 0.0426 –

0.0476 0.0046 0.7576 0.0707 0.7762

1.0 0.2822 0.2727 0.1877 1.0

1.0 0.8787 – 0.0101 0.1608

0.0476 0.8937 0.2035 0.3420 1.0

0.0476 0.2332 1.0 0.0101 1.0

0.0182 0.1570 0.7861 0.0004 0.9398

0.1294

0.0100

0.5825

0.1083

0.3096

0.0518

0.0001

All (Fishers method) P Ho, random union of gametes.

Haplodiploidy is observed in populations in which haploid males develop from unfertilised eggs and diploid females develop from fertilised eggs. The presence of this characteristic has been observed in certain other mite

species and has important implications for population analyses due to differences in the degree of relatedness between individuals (White, 1973; Wrensch and Ebbert, 1993). The haplodiploid genetic system, as with inbreeding

Fig. 2. Multilocus microsatellite clustering analysis of individual Sarcoptes scabiei using a similarity matrix based on the proportion of shared alleles. (A) analysis assumes single alleles are homozygous. (B) Analysis assumes single alleles are heterozygous-nulls.

846


Fig. 2 (continued )

populations, has a higher probability for sharing and fixing alleles and consequently this may have a significant effect on the spread of resistance. Previous microsatellite studies undertaken by our group had indicated a high proportion of null alleles at some loci which could have been the result of haploid males in the population (Walton et al., 1999). Ten female and 12 male mites were genotyped for variation at two polymorphic microsatellite loci, Sarms 20 and 36. Heterozygous genotypes from 6/12 male and 4/10 female mites were amplified, clearly showing the existence of diploid male S. scabiei. While mixed systems, involving cyclic or facultative switching has been rarely documented in other organisms and cannot be ruled out (Normark, 2003), the fact that the frequency of males heterozygous at these loci is comparable to that of females suggests that most or all males are diploid and haplodiploidy is not a significant confounder.

4. Discussion A primary issue in evolutionary biology is how much genetic isolation is required for incipient species to diverge from a common ancestor. A secondary question is what is the influence of geographical isolation in comparison to other mechanisms. Previously, microsatellite data analysis resulted in clear genetic separation between the mites from humans and those from dogs, as well as additional geographic separation between host-associated populations (Walton et al., 1999). In contrast the use of mitochondrial and ribosomal markers have previously been unable to demonstrate any divergence between host-associated populations of scabies mites, and only limited geographical separation (Zahler et al., 1999; Berrilli et al., 2002). Significantly, these latter studies did not include scabies mites derived from humans. Here we have extended our


previous investigations on both animal and human derived mites in a study using a smaller number of mites but with a considerably larger number of polymorphic microsatellite markers. Further, we have sequenced two mitochondrial markers from a set of mites overlapping the same individual mites genotyped with microsatellites. We have also excluded haplodiploidy as a possible confounder. Phylogenetic analysis of S. scabiei mtDNA COI and 16S genes resulted in haplotypes clustering into three distinct groups that differed by 5– 7%. Within each of these groups sequences were less than 1% divergent. Interestingly some of the mtDNA COI substitutions were non-synonymous between groups. Although unrecognised pseudogenes in phylogenetic studies may produce spurious results (Williams and Knowlton, 2001), this is unlikely to be the situation here as primer sequence design was based on the transcribed copy of the protein coding genes amplified from a S. scabiei cDNA library (Fischer et al., 2003). Additionally, cloned sequences aligned from the same individual mite were identical and stop codons, commonly found in pseudogenes, were not observed. As a result the nonsynonymous changes observed may be a result of selection or founder effects. Molecular clocks for mitochondrial genes in arthropods give a substitution rate of around 2% per million years (Anstead et al., 2002). High A þ T rich mtDNA poses several problems for phylogeny reconstruction as it effectively provides fewer possible character states, moreover sites saturate more quickly and reduce resolution when distance or parsimony methods are used (Blouin et al., 1998). This problem is minimised in studies based on closely related taxa. Our study indicates there is substantial divergence between human-associated mite populations and other animal-associated mite populations and indicates they may not have shared a common mitochondrial ancestor since 2 – 4 million years ago. Due to the maternal inheritance of mtDNA this divergence could result from different mechanisms, geographical isolation or isolation on separate hosts, or both. The mites from Panama analysed in this study were obtained from the Indigenous Kuna Indians living on the San Blas Islands, Republic of Panama. These island communities have few dogs, with no mange observed (Walton, personal observations). As such the inclusion of human-derived mites from Panama in this genotyping analysis could only inform us on the possible affects of geographical isolation. Scabies was not previously recognised in the Kuna until 1973 – 74 (Taplin et al., 1991). From introduction the previously unknown disease spread rapidly in the Kuna Indians and by 1976 approximately 66 communities were affected. Lee (1975) cites 1868 as the earliest record of scabies in Australia, appearing in the first shipment of Chinese people at the beginning of the gold rush in Victoria in 1854. According to Kettle (1991) the first medical recording of scabies in the Northern Territory (over 3000 km from Victoria) was in 1944 by Lt Col

847

Kirkland, who recorded six cases of scabies. Because of the relatively recent introduction of scabies to Panama and the NT the occurrence of different groups in these locations may be the result of founder effects. As noted already, these groups must have diverged elsewhere in the world long before the introduction to these countries, possibly in primates before the evolution of Homo sapiens. Further resolution of the epidemiology of scabies in humans and its introduction into Australia would require sampling and analysis of mite populations from Europe and Asia. The sub-clustering of haplotypes derived from human hosts in Group C is unlikely to be the result of recent interbreeding. If genetic exchange was occurring between groups the haplotypes of host-associated mite populations would not be expected to group together in the phylogenetic tree; rather we would expect them to be randomly scattered among the groups. However there are no haplotypes derived from non-human hosts observed in Groups A or B. Instead, the observed sub-clustering of human-associated mites in Group C is what would be expected as the result of one or a few rare gene exchange events that occurred relatively recently in evolutionary history. To examine whether the sub-clustering observed in Group C based on mitochondrial data was a result of ongoing interbreeding or of rare hybridisation and backcrossing in the past, patterns of variation in the same populations of mites were also observed with a high resolution microsatellite marker system. If ongoing interbreeding was occurring this would be reflected in the nuclear genetic structure of host-associated populations. This latter analysis, using a number of the same individual mites and higher resolution, indicated that interbreeding between host-associated populations is extremely rare. Individual mites clustered into host-associated populations congruent with the mitochondrial analysis but with no evidence of current gene flow between animal-derived mites and human derived mites. Significantly the microsatellite analysis clustered the three human-derived mites with Group C haplotypes, into the NT S. scabiei var. hominis population (Fig. 2, highlighted in bold). This discordance observed between mtDNA and nuclear data has been reported in other closely related but non interbreeding mite species (Navajas and Boursot, 2003), as well as freeliving organisms and some parasitic taxa (Anderson, 2001). Assignment tests on the microsatellite data again placed the three human-derived mites with Group C haplotypes into the NT S. scabiei var. hominis reference population. Additionally genotype frequencies between host-associated populations were significantly different, adding emphasis to the proposal that there is limited gene flow between populations. The sub-clustering observed in the mitochondrial analysis is therefore most likely a relic of past gene exchange. The mitochondrial analysis reveals a Group C mitochondrial genome with a Group B nuclear background. If these were simple F1 hybrids, the microsatellite alleles would be half Group B and half Group C; this is not

848


manifested in the analysis. Instead it appears there have been one or a few hybridisation events in the past, between females from Group C and males from Group B. Simulated data computed by Kalinowski (2002) showed that highly polymorphic loci provide superior estimates of genetic distance, or more specifically, the greater the number of independent alleles the more precise the genetic distance will be. The microsatellite data as seen in Fig. 2 provide evidence that mites from dogs in USA and Australia form a cluster that is distinct from the human-derived mites in both Australia and Panama. This supports previous microsatellite analysis demonstrating evidence of a strong association between scabies mites derived from the same host but different geographical locations. This data clearly support our previous contention that control programs for human scabies in endemic areas with sympatric humans and dogs must focus on human-to-human transmission. Furthermore, the genetic division of dog and human derived scabies mites will have important implications in the future design and development of vaccines and diagnostic reagents, as well as the monitoring and development of drug efficacy and resistance.

Acknowledgements We thank Dr David Blair of the School of Tropical Biology, James Cook University for helpful comments. This work was supported by National Health and Medical Research Grant (NHMRC) 137206, NHMRC Medical Genomics grant 219175, NHMRC Program grant 290208 and The Channel 7 Children’s Research Foundation of SA Inc. The sequencing program was carried out by contract at the Australian Genome Research Facility, Brisbane, Australia.

References Anderson, T.J., 2001. The dangers of using single locus markers in parasite epidemiology: Ascaris as a case study. Trends Parasitol. 17, 183– 188. Anderson, T.J., Su, X.Z., Roddam, A., Day, K.P., 2000. Complex mutations in a high proportion of microsatellite loci from the protozoan parasite Plasmodium falciparum. Mol. Ecol. 9, 1599–1608. Anstead, J.A., Burd, J.D., Shufran, K.A., 2002. Mitochondrial DNA sequence divergence among Schizaphis graminum (Hemiptera: Aphididae) clones from cultivated and non-cultivated hosts: haplotype and host associations. Bull. Entomol. Res. 92, 17–24. Berrilli, F., D’Amelio, S., Rossi, L., 2002. Ribosomal and mitochondrial DNA sequence variation in Sarcoptes mites from different hosts and geographical regions. Parasitol. Res. 88, 772– 777. Black, W.T., Roehrdanz, R., 1998. Mitochondrial gene order is not conserved in arthropods: prostriate and metastriate tick mitochondrial genomes. Mol. Biol. Evol. 15, 1772–1785. Blouin, M.S., Yowell, C.A., Courtney, C.H., Dame, J.B., 1998. Substitution bias, rapid saturation, and the use of mtDNA for nematode systematics. Mol. Biol. Evol. 15, 1719–1727.

Bowcock, A.M., Ruiz-Linares, A., Tomfohrde, J., Minch, E., Kidd, J., Cavalli-Sforza, L.L., 1994. High resolution of human evolutionary trees with polymorphic microsatellites. Nature 368, 455 –457. Carapetis, J., Connors, C., Yarrmirr, D., Krause, V., Currie, B., 1997. Success of a scabies control program in an Australian Aboriginal community. Pediatr. Infect. Dis. J. 16, 494 –499. Cornuet, J.M., Piry, S., Luikart, G., Estoup, A., Solignac, M., 1999. New methods employing multilocus genotypes to select or exclude populations as origins of individuals. Genetics 153, 1989–2000. Currie, B., Carapetis, J., 2000. Skin infections and infestations in Aboriginal communities in northern Australia. Aust. J. Dermatol. 41, 139 –143. Fain, A., 1978. Epidemiological problems of scabies. Int. J. Dermatol. 17, 20 –30. Felsenstein, J., 1989. PHYLIP—Phylogeny Inference Package (Version 3.2). Cladistics 5, 164–166. Fischer, K., Holt, D., Harumal, P., Currie, B., Walton, S., Kemp, D., 2003. Generation and characterisation of cDNA clones from Sarcoptes scabiei var. hominis for an expressed sequence tag library: identification of homologues of house dust mite allergens. Am. J. Trop. Med. Hyg. 68, 61 –64. Johnson, K.P., Williams, B.L., Drown, D.M., Adams, R.J., Clayton, D.H., 2002. The population genetics of host specificity: genetic differentiation in dove lice (Insecta: Phthiraptera). Mol. Ecol. 11, 25 –38. Kalinowski, S.T., 2002. How many alleles per locus should be used to estimate genetic distances? Heredity 88, 62 –65. Kettle, E., 1991. Health Services in the Northern Territory—a history 1824–1970, vol. 1. Australian National University, Australia. Kumar, S., Tamura, K., Jakobsen, I.B., Nei, M., 2001. MEGA2: molecular evolutionary genetics analysis software. Bioinformatics 17, 1244–1245. Lee, D.J., 1975. Arthropod bites and stings and other injurious effects (School of Public Health and Tropical Medicine, University of Sydney, Sydney). Munoz, E., Powers, J., Nienhuys, T., Mathews, J., 1992. Social and environmental factors in 10 Aboriginal communities in the Northern Territory: relationship to hospital admissions of children. Med. J. Aust. 156, 529–533. Navajas, M., Boursot, P., 2003. Nuclear ribosomal DNA monophyly versus mitochondrial DNA polyphyly in two closely related mite species: the influence of life history and molecular drive. Proc. R. Soc. Lond. B, (Suppl.) 270, 124–127. Navajas, M., Fournier, D., Lagnei, J., Gutierrez, J., Boursot, P., 1996. Mitochondrial COI sequences in mites: evidence for variations in base composition. Insect Mol. Biol. 5, 281 –285. Normark, B.B., 2003. The evolution of alternative genetic systems in insects. Annu. Rev. Entomol. 48, 397 –423. Paetkau, D., Calvert, W., Stirling, I., Strobeck, C., 1995. Microsatellite analysis of population structure in Canadian polar bears. Mol. Ecol. 4, 347 –354. Pemberton, J., Slate, J., Bancroft, D., Barrett, J., 1995. Nonamplifying alleles at microsatellite loci: a caution for parentage and population studies. Mol. Ecol. 4, 249 –252. Pence, D., Ueckermann, E., 2002. Sarcoptic mange in wildlife. Rev. Sci. Tech. 21, 385–398. Rannala, B., Mountain, J.L., 1997. Detecting immigration by using multilocus genotypes. Proc. Natl Acad. Sci. USA 94, 9197–9201. Raymond, M., Rousset, F., 1995a. An exact test for population differentiation. Evolution 49, 1280–1283. Simon, C., 1991. Molecular systematics at the species boundary: exploiting conserved and variable regions of the mitochondrial genome of animals via direct sequencing from amplified DNA. In: Hewitt, G.M., (Ed.), Molecular Techniques in Taxonomy, Springer, Berlin, pp. 33 –71. Skerratt, L.F., Campbell, N.J.H., Murrell, A., Walton, S., Kemp, D., Barker, S.C., 2002. The mitochondrial 12S gene is a suitable marker of populations of Sarcoptes scabiei from wombats, dogs and humans in Australia. Parasitol. Res. 88, 376 –379.

S.F. Walton et al. / International Journal for Parasitology 34 (2004) 839–849 Tamura, K., 1992. Estimation of the number of nucleotide substitutions when there are strong transition–transversion and G þ C-content biases. Mol. Biol. Evol. 9, 678–687. Taplin, D., Porcelain, S.L., Meinking, T.L., Athey, R.L., Chen, J.A., Castillero, P.M., Sanchez, R., 1991. Community control of scabies: a model based on use of permethrin cream. Lancet 337, 1016–1018. Thompson, J.D., Higgins, D.G., Gibson, T.J., 1994. CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids Res. 22, 4673–4680. Walton, S.F., Currie, B.J., Kemp, D.J., 1997. A DNA fingerprinting system for the ectoparasite Sarcoptes scabiei. Mol. Biochem. Parasitol. 85, 187–196. Walton, S., Low Choy, J., Bonson, A., Valle, A., McBroom, J., Taplin, D., Arlian, L., Mathews, J., Currie, B., Kemp, D., 1999. Genetically distinct dog-derived and human-derived Sarcoptes scabiei in scabies-endemic communities in northern Australia. Am. J. Trop. Med. Hyg. 61, 542–547. Walton, S.F., Holt, D.C., Currie, B.J., Kemp, D.J., 2004. Scabies: new future for a neglected disease. Adv Parasitol in press.

849

Weir, B.S., Cockerham, C.C., 1984. Estimating F-statistics for the analysis of population structure. Evolution 38, 1358–1370. White, M., 1973. Haplodiploidy, in: Animal Cytology and Evolution, Cambridge University Press, London. Williams, S.T., Knowlton, N., 2001. Mitochondrial pseudogenes are pervasive and often insidious in the snapping shrimp genus Alpheus. Mol. Biol. Evol. 18, 1484– 1493. Wong, L., Amega, B., Barker, R., Connors, C., Ninnal, M.D., Cumaiyi, A., Kolumboort, M., Currie, B., 2002. Factors supporting sustainability of a community-based scabies control program. Austalas. J. Dermatol. 43, 274– 277. Wong, L.C., Amega, B., Connors, C., Barker, R., Dulla, M.E., Ninnal, A., Kolumboort, L., Cumaiyi, M.M., Currie, B.J., 2001. Outcome of an interventional program for scabies in an Indigenous community. Med. J. Aust. 175, 367 –370. Wrensch, D., Ebbert, M., 1993. Evolution and Diversity of Sex Ratio in Insects and Mites, Chapman & Hall, New York. Zahler, M., Essig, A., Gothe, R., Rinder, H., 1999. Molecular analyses suggest monospecificity of the genus Sarcoptes (Acari: Sarcoptidae). Int. J. Parasitol. 29, 759–766.