Inter-simple sequence repeat ( ISSR ) polymorphism-based analysis of diversity in the freshwater turtle genus Pangshura

Assessment of genetic variability among the freshwater turtles is critical to the development of effective conservation strategies. In the present study, analysis of phenetic relationship of the genus Pangshura was studied using inter simple sequence repeat (ISSR) markers to see the concordance of currently available morphological data with genetic data along with spatial pattern of distribution. Ten microsatellite-based primers amplified 156 ISSR markers among 36 individuals belonging to four turtle species Pangshura sylhetensis, Pangshura tecta, Pangshura simithii and Pangshura tentoria, collected from certain localities of Assam, West Bengal and Uttar Pradesh of India. Estimations of Nei’s genetic diversity (h), Shannon’s Index (I) and Total genetic diversity (Ht) reveal the existence of greater genetic diversity in P. tentoria and P. sylhetensis than in the other two species. However, low values of gene flow (Nm) and of within sample diversity (Hs) indicate prevalence of inbreeding in these species groups. The coefficients of differentiation (GST) values divulge that divergence started in the Pangshura species. Projections on the principal components analysis (PCA) plot reflect the distinct genetic identity of these four species. The consensus Neighbour Joining dendogram depicts P. sylhetensis, P. tecta, and P. tentoria as genetically closer to each other than to P. smithii.


INTRODUCTION
The northern region of India, enriched with wide diversity of freshwater turtles has been designated as priority area of turtle conservation (Shrestha, 2001;Buhlmann et al., 2009).The family Geoemydidae to which the genus Pangshura belongs comprises of many of the highly endangered species (Van Dijk et al., 2000;IUCN, 2011).Most of the Geoemydid turtles are freshwater species and occupy a wide range of habitats, from highly aquatic (Pangshura) to terrestrial (Geoemyda).The genus Pangshura comprises small-sized turtles, having maximum shell length of 20 to 26.5 cm.The species that Genetic data have been used in turtle conservation to evaluate the genetic variability within and among population (Janzen et al., 1997;Souza et al., 2002;Schwartz and Karl, 2005) to recognize the existence of cryptic taxa (Russello et al., 2005), and to reveal migratory patterns (Bowen and Avise, 1996).Similarly, microsatellites have been used to evaluate the genetic consequences of population bottlenecks (Kuo and Janzen, 2004;Waldick et al., 2002), population sizes, migration rates (Nichols and Freeman, 2004), natal dispersal (Berry et al., 2002;2004), hybridization (Burns et al., 2003), diversity (Fritz et al., 2008) and also used in wildlife forensics (Avise, 2004).
DNA-based markers have gained popularity in recent years in the assessment of genetic relationship among species.Of these markers, the ISSR markers (Zietkiewicz et al., 1994) are often used in the phenetic studies (Gupta et al., 1994;Wu et al., 1994;Marmi et al., 2006;Fritz et al., 2008;Guicking et al., 2009).These markers being polymorphic (Bornet and Branchard, 2001) and ubiquitous in the genome (Tautz and Renz, 1984), have the advantages of Simple Sequence Repeat (SSR) markers, while bypassing the major obstacle to the development of SSR markers, that is the need to know the flanking sequences.Generally, the ISSRs are scored as dominant markers which are inherited in Mendelian fashion (Ratnaparkhe et al., 1998).The SSR regions scattered evenly throughout the genome (Condit and Hubbell, 1991) and yielding a large number of polymorphic bands, which are interpreted as band present or band absent (Tsumura et al., 1996).The absence of band signifies primer divergence or loss of a locus through the deletion of the SSR site or chromosomal rearrangement (Wolfe et al., 1998).Friz et al. (2005Friz et al. ( , 2008) ) examined the evolutionary relationships of five species of the genus Testudo and Cyclemys diversity using the data generated from the inter simple sequence repeat-polymerase chain reaction (ISSR-PCR) genomic nuclear fingerprints.Hence, the ISSR markers are suitable for use in species distinction where extensive information on DNA sequences is not readily available (Meloni et al., 2006).We report here the findings of our ISSR polymorphism-based study on the genetic diversity and phenetic relationship of all the four species and the sub-species within the Indian freshwater turtle genus Pangshura.

Study area
The field work and sample collection were carried out during January, 2008 to March, 2011 in different parts of Assam (including bordering areas of Bangladesh and Bhutan), West Bengal, Uttar Pradesh of India (Table 1, Figure 1).Almost all the target species and their phenotypic variations were captured and used for the analysis of molecular data.Taxonomy and nomenclature were followed after Das (1995Das ( , 2002) ) and Praschag et al. (2007).Baruah et al. 239

Sample collection and DNA extraction
Thirty-six (36) Pangshura samples belonging to four species as presented in Table 1 were sampled by clipping off a tiny piece of the webbing of toes.Tissues were preserved in 95% ethanol and stored at -20°C until DNA extraction.DNA was extracted using the method of Ausubel et al. (1995) with minor modifications and stored in TE buffer at -20°C until use.

Inter-simple sequence repeat-polymerase chain reaction (ISSR-PCR) amplification
Polymerase chain reactions (PCRs) were performed using ten ISSR primers (Table 2), after an initial screening of 19 primers.Primers P4 to P9 were based on turtle microsatellites reported by Aggarwal et al. (2004), Edwards et al. (2003) and Schwartz et al. (2003), with some modifications.Primers P1, P2, P3 and P10 were designed and used by Chen et al. (2009) in grass carp.PCRs were carried out in 20 μL of reaction volume containing 20 ng of genomic DNA, 1.0 U Taq polymerase, 200 μM dNTP, 1 μM primer, and 1X amplification buffer (containing 2 mM MgCl 2 ).The condition for amplification was an initial denaturation temperature 94°C for five min, followed by 35 cycles of 45 s at 94°C, then by 45 s at appropriate annealing temperature (Table 2) followed by 2 min at 72°C, and then by a final extension step for 7 min at 72°C.Amplicons were resolved at 100 V on 1.8% agarose gels containing 0.5 μg/ml ethidium bromide.Low range DNA ruler was used as size marker.

Data analysis
The molecular sizes of the amplicons were measured using the gel documentation software UVItec ver 12.8.Lanes representing the turtle samples were scored for the presence (1) or absence (0) of bands.POPGENE ver.1.32 (Yeh and Boyle, 1997) was utilized to estimate genetic variability based on Nei's (1973) coefficient of gene differentiation (G ST ).The software was also used to calculate observed alleles, effective alleles, Shannon's information index (I), total genetic diversity (Ht), sample genetic diversity (Hs) and estimated gene flow (Nm).The relatedness among turtle samples was analysed based on the Neighbour Joining algorithm (Saitou and Nei, 1987).The matrix of binary characters was bootstrapped using the software Phyltools (Buntjer, 2001).Re-sampling was done 1000 times using 35% bootstrap values.Phytools was then used to generate distance matrices based on Nei's coefficient (Nei and Li, 1979); S= 2a/ (a+b) + (a+c), where, a= band(s) shared by individual x and y; b= bands amplified in individual x but not in y; c= band(s) amplified in individual y but not in x.Distance was estimated as 1-S.The Consensus Neighbour Joining Tree was prepared using the software Phylip ver 3.1 (Felsenstein, 2004).Geneious Pro ver 4.8 (Drummond et al., 2010) was used to obtain a presentable consensus tree.
Principal components analysis (PCA) of the original binary matrix was also performed sequentially by using the modules of NTSYSpc ver 2.20r (Rohlf, 2005): stand, simint, eigen and project.

Inter-simple sequence repeat (ISSR) amplification
The primers generated 156 clear and distinguishable ISSR bands, where sizes ranged from 0.2 to 2.836 kb (Table 2).The number of bands per primer ranged from 9 to 24 with a mean of 15.6 bands per primer.Primers 1, 4, 6 and 9 amplified monomorphic markers (Table 2).Primer 3 revealed three markers of 0.615, 0.666 and 0.722 kb sizes and Primer 5 revealed two markers of 0.52 and 0.592 kb sizes that were common to both the P. smithii subspecies (Figure 2).Primer 5 also amplified two markers of 0.541 and 0.949 kb sizes that were present in all the three P. tentoria subspecies.
The highest number of bands was generated from the primer P10.The highest percentage of polymorphism (100%) was generated from the primers P2, P3 and P5 (Table 2).The number of polymorphic bands produced per primer ranged between 6 (P9) to 23 (P10).Out of the 145 polymorphic bands, P. sylhetensis had 22 (highest) numbers of unique polymorphic bands followed by 20 in P. tentoria, 18 in P. tentoria circumdata, 15 each in P. smithii and P. tentoria flaviventer, 11 in P. tecta and 7 in P. smithii pallidipes (Table 3).The ISSR banding patterns were amplified by seven primers (Figure 2).
The band profiles generated by ISSR primers showed polymorphism among the Pangshura samples.The extent of polymorphism varied.The percentage polymorphic loci were calculated to be 38.59,24.46, 27.72 and 47.83 for P. sylhetensis, P. tecta, P. smithii and P. tentoria respectively.

Genetic variability revealed through inter-simple sequence repeat (ISSR) markers
A summary of genetic variation statistics for all loci are presented in Table 4. Average numbers of alleles observed in P. sylhetensis, P. tecta, P. smithii and P. tentoria were 1.34 ± 0.49, 1.24±0.43,1.26 ± 0.45 and 1.48 ± 0.5 respectively, while it was 2.0 ± 0.0 when all four Pangshura species were taken together (Table 5).The genetic diversity in the four Pangshura species is presented in Tables 5 and 6.

Phenetic relationship among the test species
Genetic similarity and distance values based on Nei's original (Nei, 1972) and unbiased (Nei, 1978) measures are presented in Table 6.In the present study, each species of Pangshura is considered as a single population and the genetic distances among the four   7).
The Neighbor-Joining (NJ) consensus tree is depicted in Figure 3.The species were grouped into three clusters.Cluster I comprised of P. tecta and P. sylhetensis.Topology of the cluster indicates that genetic diversity exists between the two species.
Both the species formed two distinct sub-clusters with 81 and 100% bootstrap support for P. tecta and P. sylhetensis respectively.Diversity existed among the P. sylhetensis samples.P. sylhetensis from Bishwanathgath (T28) and Jia Bharali (T8) from the state of Assam were clustered together with 83% bootstrap support.The P. sylhetensis samples from Kuruwaghat (C10) of northern Assam and Kushiara River (T 52) bordering Bangladesh were in the same cluster.However, P. sylhetensis of Buxa Wildlife Sanctuary (T3) of West Bengal formed a distinct clade within the P. sylhetensis group with 81%, bootstrap separation.In the P. tecta sub-cluster of Hajo (T7) and Gomerighat (C7), samples were grouped together with 100% bootstrap support, while the Kushiara river sample (C4) formed a monophyletic branch at 100% bootstrap separation.
All the three currently recognized subspecies of P. tentoria were clubbed in the clusters II with bootstrap separation 64%.P. tentoria sample collected from Hajo (T49) was on a monophyletic branch of the clade.P. tentoria (T 26) and P. tentoria flaviventer (T27) samples from Lahorighat were clustered together with 78% bootstrap support.Samples of P. tentoria circumdata collected from the Ganga (T12, T13) and Yamuna rivers (T15, T16) at different sites of Uttar Pradesh (Table 1) reflected their geographical distance from the Pangshura samples of Assam.One of the P. tentoria circumdata samples (T13) collected from the Ganga River at Farrukhabad, Uttar Pradesh was on a monophyletic branch (Figure 3).
Two subspecies of P. smithii were clustered together in cluster III.The P. smithii samples from Depor beel (T50) and Tengatoli (T1) were closer than the other samples.The P. smithii samples collected from Burachapari (T2) and Biswanathghat (C12) formed monophyletic branches.On the other hand, the samples of P. smithii pallidipes from Farrukhabad, Uttar Pradesh also formed monophyletic branch with 53% bootstrap separation.
The genetic interrelationships among Pangshura samples were projected on the PCA plot (Figure 4) and those depicted by the consensus NJ tree were similar; but were better represented on the plot.The subspecies of P. tentoria and that of P. simithii were well-separated by the first and the second principal components, however, their phenetic relationship with the subspecies of P. tecta, and P. sylhetensis was less defined.The subspecies of P. tecta and those of P. sylhetensis are projected as being closer to one another than they are either to the subspecies of P. tentoria or P. smithii.

DISCUSSION
Amplification of a large number of polymorphic ISSR bands (77.78 to 100%) indicates the existence of genetic diversity in the four Pangshura species.The average values of the observed and effective alleles, percentage polymorphic loci, Nei's gene diversity (h), Shannon's information index (I), total genetic diversity (H t ) (Table 5) all reveal that the highest polymorphism is present in P. tentoria followed by P. sylhetensis.
The high G ST values and very low Nm values (Table 5) for all species indicate that they are in a high state of differentiation; G ST > 0.25 and Nm < 1.0 are threshold values beyond which significant population differentiation occurs (Kar et al., 2005).The PCA plot (Figure 4) also indicates the genetic distinctness of the four species.Differentiation might have occurred probably only due to genetic divergence through the evolutionary courses.The Neighbour Joining dendogram (Figure 3) presents the genetic interrelatedness of the Pangshura samples.The 100% bootstrap separation of P. smithii samples (cluster III) implies this species is more distant from the other two species.The G ST values presented in Table 5 supports this result.Nei's measures of genetic identity and genetic distance (Table 7) are also in broad agreement of phenetic relationship that the NJ tree presents.
The PCA result reveals that a close relation within the respective P. tentoria could not be expected.From the P. sylhetensis group, samples from West Bengal and Kushiara River bordering to Bangladesh are placed near to the P. tecta group, representing a close genetic relationship between the two species (Figure 4).The analysis of species specific unique ISSR amplicons of the present study (Table 2) along with the genetic parameters (Table 5), the correlation of PCA plot and neighbour joining tree reveal that the genetic diversity among the Pangshura could be well inferred.Praschag et al. (2007) sampled P. sylhetensis and P. tentoria from Assam (Northeast India) to construct phylogeny of endangered Southeast Asian turtles including Pangshura.However, their study was based on patchy taxon sampling and could not establish any distinctness between the subspecies of P. smithii and P. tentoria with the help of cyt b gene dataset.The present study has established that in order to establish the variation in the genome of Pangshura species, ISSR appear as powerful tool.These variations support the validity of the morphologically weakly defined subspecies P. tentoria and P. tentoria circumdata as well as P. smithii and P. smithii pallidipes.Although few cross-species microsatellites could be used for genetic diversity of Pangshura, the availability of species-specific markers is highly desirable for population structure assessments.These microsatellites thus provide efficient genetic markers to understand the population structure, phylogeography and species relationships of Pangshura and other freshwater turtle species.

Conclusion
Our study reveals the existence of a narrow genetic base for all the four turtle species, which we attribute to inbreeding.It is necessary to formulate conservation strategies in terms of protection of the habitat sites to minimize the reduction of gene pool, which might be the probable cause of inbreeding.
In-situ egg hatching of eggs followed by headstarting of those species can also be undertaken as a part of conservation strategy.Future work should focus on finer scale genetic analyses in order to gain a better baseline  Government of India for the financial support to establish the Bioinformatics Infrastructure Facility (BIF).The authors are very grateful to Asutosh Triapthi of TSA-India for providing the samples.

Figure 1 .
Figure 1.Map showing the location of sample collection sites in Assam (Including Assam-Bangladesh Border, Assam-Bhutan Border and part of West Bengal) and Uttar Pradesh.

Table 1 .
Pangshura samples used in the present studied.
Pangshura species ranged from 0.14 to 0.23.The shortest distance (0.14) existed between P. smithii and P. tentoria.The longest genetic distance (0.23) existed between P. smithii and P. tecta.The genetic identity value ranged from 0.79 to 0.87 (Table

Table 4 .
Observed and effective number of alleles in the four Pangshura species for all primers.

Table 5 .
Genetic parameters for the four species of genus Pangshura.

Table 6 .
Nei's F-statistics and Shannon's information index in Pangshura species for all ISSR primers.

Table 7 .
Nei's original (1972) and unbiased (1978)measures of genetic identity and genetic distance among the four Pangshura species.