The genetic diversity and geographical separation study of Oncomelania hupensis populations in mainland China using microsatellite loci

Background Oncomelania hupensis is the unique intermediate host of Schistosoma japonicum, which plays a crucial role in the transmission of schistosomiasis. The endemic area of S. japonicum is strictly consistent with the geographical distribution of O. hupensis. Methods A total of 24 populations of O. hupensis from four ecological landscapes were selected for analysis of genetic diversity by screening eight microsatellite DNA polymorphic loci. Results The number of alleles per locus ranged from 29 to 70 with an average of 45.625 and that of effective alleles were 18.5 to 45.8 with an average of 27.4. The observed (Ho) and expected (He) heterozygosities varied from 0.331 to 0.57 and from 0.888 to 0.974, respectively. The mean of polymorphism information content (PIC) for all populations was 0.940, appearing polymorphic for all loci. For the fixation index of F-Statistics, Fit and Fst were 54.95 and 37.62 %, respectively. Variation of O. hupensis chiefly exists among individuals, accounting for 60.58 % of the total variation determined by Analysis of Molecular Variation (AMOVA). Variation among individuals within populations, among populations within groups and among groups only accounted for 26.60, 8.04 and 4.78 %, respectively. This distribution of variation suggests that genetic differences principally originate from within-populations rather than among-populations. Moreover, UPGMA cluster analysis showed that the populations spreading within middle and lower reaches of the Yangtze River (HBWH, JSYZ, JXNC, HNHS, JXJJ, AHWW, HBJL, JXDC, HNNX, JSYZJZ, ZJJH, AHNG and AHWJ) clustered together first, then gathered with the populations in the high mountains (SCMS, SCYA, SCPJ, YNEY, SCLS, YNWS and SCXC), coastal hills (FJFQ and FJFZ) and Karst landform (GXBS and GXYZ) successively. Conclusion This study provides novel insight into the theoretical source of genetic differentiation of Oncomelania hupensis in mainland China, which is critical for the epidemiological investigation and surveillance of S. japonicum.


Background
Schistosomiasis, caused by Schistosoma japonicum, remains one of the most prevalent parasitic diseases and effects severe socio-economic and public health losses in China [1,2]. Oncomelania hupensis is the unique intermediate host of S. japonicum, which plays a critical role in the transmission of Schistosomiasis japonica [1,3]. The geographical distribution of O. hupensis coincides with the endemic area of S. japonicum [4], which is mainly found throughout the southern region of the Yangtze River basin [5,6]. As a result, significant genetic differentiation leads to the formation of multiple geographical populations of O. hupensis [3]. Coincident with the endemic area for schistosomiasis, O. hupensis has been mainly found in four types of ecological landscapes giving rise to subspecies including:(1) O. h. hupensis largely in the middle and lower reaches of the Yangtze River (among the provinces of Hunan, Hubei, Jiangxi, Anhui, Jiangsu and Zhejiang) (2) O. h. robertsoni in the mountainous region of Sichuan and Yunnan provinces (3) O. h. guangxiensis in the Karst landscape of Guangxi province and (4) O. h. tangi in the southeastern coastal region of Fujian province [7,8]. Interestingly, obvious morphological differences have been identified among individuals from the same regional population [9][10][11]. For example, O. hupensis from upstream of Miaohe basin, which contains regions of swamps and lakes, have a ribbed shell while those from downstream have a smooth shell [12].
Microsatellite DNA, known as short tandem repeat (STR) or simple sequence repeat(SSR), occurs throughout the eukaryotic genome. Differences in repetitive sequence numbers allow for high polymorphism due to the ubiquitous occurrence, high copy numbers, high heterozygosity and easy detection within population [13]. Along with other genome mark technology, it has been widely applied to research examining genetic diversity and serves as an important molecular marker [14][15][16][17]. At present, microsatellites have been isolated from many different organisms [18][19][20]. Specifically, from 128 molluscs, a total of 3, 284 microsatellite sequences have been identified [21]. Although the microsatellite DNA library of O. hupensis was built recently [22], the microsatellite markers have not been used extensively in population genetic structure studies and genome mapping of O. hupensis in P.R. China [23][24][25]. To deepen our knowledge on the genetic diversity of the intermediate host snail, we developed a novel multiplex PCR method to screen and analyze the genetic diversity of O. hupensis using microsatellites loci among the four various ecological landscape populations in mainland China.

Snail sampling
A total of 24 populations of O. hupensis were sampled from four ecological landscape populations in mainland China covering: (1) the region of swamps and lakes in the middle and lower reaches of the Yangtze River, (2) the mountainous region of the Sichuan and Yunnan provinces, (3) the littoral hill part of the Fujian province and (4) the karst landscape of Guangxi autonomous region ( Fig. 1, Table 1).

DNA preparation
Ten to 20 O. hupensis samples were randomly chosen from each site, fed for 1 week and identified as infected or non-infected with S. japonicum by observation of cercariae emerging from the snails. Only non-infected snails were used in this study. After removal of the gut and digestive glands from the soft parts of the snails, the 30 mg muscle tissues from the pleopod of a single snail were digested for 3 hours at 56°C with proteinase K (Amresco Inc. Solon, OH, USA) followed by the standard DNA extraction procedure [26] using mollusc DNA Kit (Omega, USA).

PCR amplification and detection of PCR products
The microsatellite DNA polymorphic loci were selected and evaluated from previous microsatellite loci library [22]. Two rounds of multiplex PCR reaction were developed including four microsatellite loci in each one, which were identified by different lengths and fluorescence peaks of 6-FAM, VIC, NED and PET labeled by (Sigma-aldrich London, UK). Primer sequences and information are summarized in Table 2.
The multiplex PCRs were developed using the Type-it Microsatellite PCR Kit (QiaGen, London, UK) with a 25 μl reaction system, including 2x Type-it Multiplex

Analysis of microsatellite diversity
The accurate length of amplified fragments of microsatellite DNA loci were determined using Geneious software (Version 7.0.6) and subsequently exported as an Excel  [27]. The correlation between genetic distance and geographical distance were tested with Mantel regression. Analysis of molecular variance (AMOVA) was processed through Popgene software, clustering analysis was determined by unweighted pair group method with arithmetic means (UPGMA) and the phylogenetic tree was modified with TreeView [28]. The polymorphism information content (PIC) was calculated according to the formula previously described [28].

Gene scan
From the 24 populations of O. hupensis sampled, 396 specimens were scanned at the genetic level across eight polymorphic loci of microsatellite DNA. The lengths of amplified fragments for a total of 6,196 microsatellite DNA loci were obtained.     possible single exact locus tests (P < 0.01).No significant linkage disequilibrium was found between all pairs of the eight loci examined (P < 0.01), which indicated the independent behaviour of all loci. Analysis with Genepop software showed the possible occurrence of null alleles, which may lead to deviations from HWE and result in exaggerated levels of genetic differentiation [26,29,30]. Null alleles may be due to flank sequence variation decreasing primer annealing efficiency, allele drop out or DNA quality [23,31].

Genetic differences among individuals
Fit and Fst values were 54.95 and 37.62 %, respectively. This suggests that genetic differences mainly exist within populations rather than among those with unbalanced differentiation degrees (Table 6). Mantel ' s test of regression showed that the correlation (41.97 %) between geographic distance and genetic distance among populations is positive (R 2 = 0.1011, P < 0.05) and genetic distribution of all populations accorded with the Isolation-by-distance Model (Fig. 2, Tables 7 and 8).
Genetic parameters of the four groups from different landscapes (i.e. lakes and marshes, high mountains, Karst and coastal Hills) showed that Na ranged from 2.063 to 11.452, He from 0.465 to 0.852 and Ho from 0.274 to 0.492. The group from the Karst landscape had the lowest value in all three indices, which indicated its low   (Table 9). This suggests that there is no significant genetic differentiation among groups. UPGMA cluster analysis for the 24 O. hupensis populations based genetic distance showed that the populations spread in the landscape of middle and lower reaches of Yangtze River (HBWH, JSYZ, JXNC, HNHS, JXJJ, AHWW, HBJL, JXDC, HNNX, JSYZJZ, ZJJH, AHNG and AHWJ) clustered together first and then gathered with the populations of high mountains (SCMS, SCYA, SCPJ, YNEY, SCLS, YNWS and SCXC), coastal hills (FJFQ and FJFZ) and Karst land form (GXBS and GXYZ) successively (Fig. 3).

Discussion
Oncomelania hupensis is the sole intermediate host for transmitting Schistosoma japonicum in mainland China [32], and it is widely distributed in the southern region of the Yangtze River valley. Significant genetic variations have developed in O. hupensis from different geographic populations due to their distribution range, complexity of breeding environment and geographical location.
In this research, The genetic differentiation of four different landscape groups of O. hupensis were studied through eight screened polymorphic microsatellite DNA loci. This information is pertinent because it further improve our understanding on the effect of genetic diversities on the distribution of O. hupensis. This will ultimately help boost our surveillance activities and also strengthen the control of schistosomiasis transmission in China. genetic indices were tested aross eight  microsatellite DNA loci. The mean Fis value for the 24 populations examined was 0.272, indicating a deficiency of heterozygotes and frequent inbreeding within populations, which is likely due to the small range of activity of O. hupensis. A total of 47 microsatellite DNA loci deviated from the Hardy Weinberg Equilibrium demonstrating a serious lack of heterozygotes. Possible explanations that may account for this include: activities of migration and inbreeding, drug pressure, gene mutation and null alleles. However, it is currently unclear which one is the dominant factor contributing to this phenomenon [33]. No significant linkage disequilibrium was found between all pairs of the eight loci, clearly showing the independent behaviour of all loci. Null alleles were found at all eight polymorphic loci. This may be due to: 1) mismatching of primer pairs: mutations in microsatellite DNA sites critical for binding with primers leads to abnormal amplification 2) losses of large alleles: the superiority of short alleles restrict amplification of long fragments or 3) differences in DNA quality: unevenness of templates character obstruct amplification in some loci [26,31,34]. Null alleles could implicate genetic diversity parameters for populations such as excess of homozygote individuals, reduction of Ho and He and increase of genetic distance and Fis; moreover, it leads to inaccuracy of parent analysis [30][31][32][33][34][35][36][37].
The abundance of the number of heterozygotes and the amount of genetic information in a population is directly proportional to the PIC value [38,39]. Result shows that PIC was greater than 0.5 at every locus, and the mean  [23]. This signifies that all the eight loci screened were highly polymorphic. Furthermore, this study reveals that the average Fst for all loci was 0.376, which means that 37.6 % of genetic variation was among populations and 72.4 % was among individuals within populations. The analysis of AMOVA displayed that genetic variation among individuals (60.58 %) were far higher than that within populations (26.60 %), while among populations and among groups are (8.04 %) and (4.78 %) respectively. This implies that, genetic diversity is strongly derived from among-individuals rather than amongpopulations. However, the average Fst (0.376) and genetic variation among populations (8.04 %) were higher than values obtained from the previous results (0.048 and 4.8 %) respectively, revealing genetic variation among populations increased along with geographical distance [23]. The Mantel test demonstrated an apparent positive correlation between genetic distance and geographical distance. The genetic structure between geographical populations is embodied with some degree of independence. For example, the geographical distance between the HBWH and JSYZ populations located in the lake region was far, but with low degree of variation. This could possibly be related to the genetic differentiation principally being among individuals within populations rather than among geographic locations for the populations in Lakes and Marshes landscape.
The phylogenetic tree constructed by UPGMA also showed that populations in neighboring geographical locations generally cluster together, which was consistent with the Mantel test results. The cluster sequence of geographical populations showed us that the population from the karst landscape of Guangxi autonomous region maybe the most original one, then the population from the littoral hill part of the Fujian province, the population from the mountainous region of the Sichuan and Yunnan provinces and the population from the region of swamps and lakes in the middle and lower reaches of the Yangtze River, respectively. Regarding as the largest population spread throughout the middle and lower reaches of the Yangtze River [7], the populations from different provinces also crossed cluster, these include, between Hubei and Jiangsu, Hunan and Jiangxi, and Zhejiang and Anhui, which may be as a result of O. hupensis spreading along the river within the large population, or gene drifting for surged water flow in the lakes and marshes landscape [34]. Then this branch clustered with the populations of Sichuan and Yunnan province successively. Furthermore, the major branch clustered with the populations of Fujian and Guangxi province in turn, this agrees with the conclusion of four landscape populations relationships from previous studies using SSR-PCR [40] and DNA sequence markers [7,41,42].

Conclusion
This study has shown that the genetic diversity of O. hupensis, an important snail intermediate host of S. japonicum in China mainly originates from among-individuals rather than among-populations. It also reveals that the populations within subspecies have closer consanguinity than between subspecies in the mass, nevertheless, genetic variations exist within subspecies. These findings further provide important information on genetic structure of O. hupensis and strengthen our knowledge about diffusion trend and tracking to the source of Oncomelania in mainland China. Ultimately, these findings will help us develop more effective guidelines for controlling the spread and distribution of Oncomelania and consequently prevent the transmission of Schistosomiasis in China. Our data offers a better understanding of the genetic differentiation of Oncomelania hupensis, enhancing our ability to effective and efficient surveillance of Schistosomiasis.