Reliability and comparison of the polymorphism revealed in amaranth by amplified fragment length polymorphisms ( AFLPs ) and inters simple sequence repeats ( ISSRs )

The present study reported the effectiveness of two PCR-based molecular techniques, inters simple sequence repeats (ISSRs) and amplified fragment length polymorphisms (AFLPs), for genetic assessment of amaranth. The polymorphic loci ranged from 110 among A. caudatus to 228 among A. cruentus and 16 among A. tricolor to 56 among A. hypochondriacus for AFLP primer combinations and ISSR primers, respectively. Among the two marker systems used, ISSR fingerprinting detected the highest number of alleles per locus (1.83) compared to AFLPs (1.63). However, the assay efficiency index for AFLP was 14.49, more than five-fold higher than ISSR (1.75). The study also revealed that ISSR primers with di-nucleotide repeats gave a good fingerprint, indicating that di-nucleotide repeats are more frequent in amaranth genome. The reproducibility of the two marker systems was confirmed by the narrow gene diversity (0.03 ± 0.11 to 0.07 ± 0.17) observed between the controls. Bayesian consensus and neighbor-joining trees were constructed to describe the cluster arrangement among the Amaranthus spp. The cluster pattern was similar for both markers, though the cluster order in the trees was slightly different. The results of this study confirm the usefulness of AFLPs and ISSRs for the genetic assessment of amaranth.


INTRODUCTION
The species within the Amaranthus are very closely related and literature shows that misclassifications among the grains, vegetable as well as their weedy and wild relatives occur frequently.Comprehensive genetic diversity studies have been conducted in major crops, using passport, agro-morphological (Ben-Har et al., 1995), and biochemical data obtained by analyses of isozymes (Hamrick and Godt, 1997) or storage proteins (Smith et al., 1987).However, their usefulness for obtaining reliable estimates of genetic similarity is limited because of the small number of marker loci available and the low degree of polymorphism generally found in improved local breeding materials (Messmer et al., 1991).The advantage in the use of molecular markers technique present or absent.Fingerprinting techniques have the their ability to detect genetic variation at levels of resolution that exceed those achievable with other, previously applied methods (Karp, 2002).Owing to the great number of polymorphic marker loci and nature, DNA-assays are more robust and independent of environmental conditions.PCR-based DNA markers are less labour-and time-consuming, and provide an estimate of genetic similarity by direct sampling from the entire genome with unprecedented precision (Peleman and van der Voort, 2003).However, the nature of the marker system, genome coverage and the crop determines the extent of their utility.Most widely applied DNA marker techniques differ not only in principle, but also in the type and amount of polymorphism detected.Techniques such as non-PCR based restriction fragment length polymorphisms (RFLPs) (Botstein et al., 1980) and PCR-based microsatellites or simple sequence repeat polymorphisms (SSRs) (Tautz, 1989) possess the ability to distinguish multiple bands (alleles) per locus, thus giving more information on a single locus.By contrast, individual bands detected with PCR-based fingerprinting techniques, such as randomly amplified polymorphic DNA (RAPDs) (Williams et al., 1990) and amplified fragment length polymorphisms (AFLPs) (Vos et al., 1995), are scored on a biallelic basis, as marker band ability to generate multiple marker bands in a single assay.
Molecular markers are also employed for the genetic characterization of amaranth germplasm.They have been used to differentiate genotypes under environmental conditions that confounded their phenotypes (Costea et al., 2006).Simple sequence repeats (SSR's) are one of the frequently used molecular markers for genotyping crops (Tautz, 1989).A number of research studies have demonstrated the use of SSRs and ISSRs to detect polymorphism and diversity in amaranth (Mallory et al., 2008;Xu and Sun, 2001;Ray and Roy, 2007), and quinoa (Mason et al., 2005;Jarvis et al., 2008;Fuentes et al., 2009).However, inter-simple sequence repeat (ISSR) markers are simpler to use than SSR technique (Ray and Roy, 2007;Nolan et al., 2010).The use of ISSRs does not require prior knowledge of the target sequences flanking the repeat regions, is not expensive and is relatively easy to score manually compared to SSR.This aim of this study was to evaluate the reliability and compare the application of ISSRs and AFLPs to reveal genetic polymorphism in amaranths.

Plant materials and DNA extraction
Seeds of seventeen genotypes of Amaranthus spp., representing A. caudatus, A. cruentus, A. hypochondriacus and A. tricolor were obtained from the genetic resource unit of Asian Vegetable Research and Development Centre (AVRDC), Tanzania, and Research Institute and Crop Production (RICP), Czech Republic (Table 1).Two control genotypes were included, a genotype sown twice under coded numbers (blind check) and a randomly chosen genotype duplicated after DNA extraction and re-duplicated in consecutive steps of the analyses (laboratory duplicate).Total genomic DNA was extracted from fresh young leaves of each genotype following the manufacturer protocol of the DNeasy Plant Mini Kit (QIAGEN, Hilden, Germany).

AFLP protocol
AFLP analysis was performed as described in Vos et al. (1995) with modification (Baránek et al., 2009).Genomic DNA (0.30 μg) was digested with restriction enzymes EcoRI and MseI (10 units each) at 37°C for 6 h in 1 × NEB buffer, and EcoRI and MseI adapters were ligated to both ends of the restriction fragments with 1 unit of T4 DNA ligase at 16°C overnight.The adapter and primer sequences used in this study are given in Table 2. PCR was carried out with specific, commercially produced primers exactly complementary to the adaptors, but whose 3´ ends are extended for selective nucleotides into the fragments.
The pre-amplification products were diluted five times and used as templates for subsequent selective amplifications.Nine primer pair combinations were used for selective amplifications (Table 3).Three EcoRI primers with three selective nucleotides (E-AGG, E-ACT, E-AGC) were each combined with one of three MseI primers with two selective nucleotides (M-CT, M-GC, M-AG).EcoRI primers were end-labeled with fluorescence T4 polynucleotide kinase (Invitrogen Life Technologies GmbH, Karlsruhe, Germany).Selective PCR amplification was performed following the protocol as suggested in the AFLP Analysis System I AFLP starter Primer Kit (GibcoBRL) with modification: first cycle at 94°C for 30 s, 65°C for 30 s, and 72°C for 1 min; the annealing temperature was then lowered 0.7°C each cycle during the following 9 cycles, and the optimal annealing temperature of 56°C was reached after a touchdown phase of 10 cycles; the amplification was then continued for an additional 24 cycles (94°C for 30 s, 56°C for 30 s, and 72°C for 1 min).
The AFLP products were analyzed using an automated ABI PRISM 310 genetic analyzer (Applied Biosystems).GeneScan software (Applied Biosystems) was used to overlap the signals from all samples making it possible to evaluate the peak and intensity of each sample, and to translate the product into a descriptive computer files.Detailed manual evaluation was undertaken by two ISSCR-3(CA)8GG, ISSCR-5(CA)8AC, ISLA-(AGC)4G, ISLA-(CA)6GT, ISLA-(CT)8AC and ISLA-(CT)8TG, produced polymorphic bands and were subsequently used in the study.PCR reactions for ISSR analysis were done using the method of Ray et al. (2006).PCR mixtures were carried out in 25 µl volumes containing of 18.75 µl dH2O, 0.5 µl primer, 2.5 µl Finnzymes Buffer (1x), 0.5 µl MgCl2, 0.25 µl dNTP, 0.5 µl Taq DNA polymerase and 2 µl of template DNA (10 ng/µl).The amplification was performed in a T-Gradient thermocycler (Biometra).The amplification reaction involved an initial denaturation at 94°C for 2 min followed by 45 cycles of denaturation at 94°C for 60 s, hybridization of primers at 50°C for 60 s and polymerization by Taq at 72°C for 2 min.Final extension at 72°C for 7 min followed by 4°C hold.The annealing temperature was adjusted according to specification.The amplification products were mixed with DNA gel loading buffer and fragments were separated by horizontal electrophoresis on a 1.4% resolute agarose gel, using 1 × TAE buffer (pH 8.0) at 110V for 1 h.The bands were detected under UV light and digitized with trans-illuminator ECX-20.M (VILBER LOURMAT).

Data scoring and genetic analysis
AFLP and ISSR fragments were used for each individual and primer combination to score the presence (1) or absence (0) of bands.Fragment sizes were estimated based on GS 500 ROX size standard and 1-kb DNA ladder size (according to the BioMax ID software algorithm) for AFLP and ISSR, respectively.This information generated the binary matrix that was used for analysis.Only bands that could be scored consistently among the genotypes were used.It was interpreted as dominant markers and was scored as diallelic regardless of band intensity.The binary matrix was used to determine the allele frequency, number of polymorphic loci and percentage of polymorphism using POPGEN Version 1.32 software program (Yeh et al., 1997).
The efficiency of AFLP and ISSR marker systems was compared by computing the assay efficiency index (Ai).The index combines the effective number of alleles identified per locus (calculated as 1  2 , where fi is the frequency of i th marker allele) and the number of polymorphic bands detected in each assay.Ai =    , where   is the total number of effective number of alleles detected over all loci and P is the total number of assay performed for their detection (Pejic et al., 1998).FREETREE program (Pavlicek et al., 1999) was used to calculate the degree of similarity using the Nei and Li/dice similarity index (Nei and Li, 1979).Similarity matrices obtained were used to calculate the average intra-and interspecific similarity between the amaranth genotypes, and also used to construct a Bayesian tree using the unweighted pair group mathematical average (UPGMA), and a dendrogram using neighbor-joining.The bootstrap re-sampling method was used to evaluate the reliability of phylogenetic groupings, with bootstrap support values obtained over 500 replications.

RESULTS
The number of polymorphic loci generated with the nine AFLP primer combinations varied among intraspecfic genotype and was higher compared to the number of polymorphic loci detected by ISSR primers (   of fragments generated varied from 17 to 35% for AFLP primer combinations and 17 to 61% for ISSR primers.At intraspecific level, higher gene diversity was detected by ISSR primers compared to AFLP primer combinations.Furthermore, low gene diversity estimates was observed within laboratory duplication, 0.03 ± 0.09 for AFLP and 0.05 ± 0.16 for ISSR, and blind check, 0.03 ± 0.11 for AFLP and 0.07 ± 0.17 for ISSR.The average number of alleles per locus was 1.63 ± 0.48 for AFLP and 1.83 ± 0.38 for ISSR while the assay efficiency index of these alleles was 14.49 and 1.75 for AFLP and ISSR, respectively. ISSRs gel electrophoresis profile generated by ISLA-(CT) 8 (di-nuclotide repeats) and UBC-866 (CTC) 6 (trinucleotide repeats) is presented in Figure 1.Generally, ISSR primers composed of di-nucleotide repeats motif Nei and Li/Dice's similarity coefficients ranged from 0.77 to 0.86 and 0.64 to 0.69 with AFLP primer combinations and ISSR primers, respectively.Both primers indicated the highest similarity coefficient between A. cruentus and A. tricolor and the lowest between A. cruentus and A. caudatus (Table 5).
The Bayesian consensus trees (BCT) of the Amaranthus spp.based on AFLP primer combinations and ISSR primers using Nei and Li/Dice's genetic similarity matrix are presented in Figure 2. In both molecular technique, the Amaranthus spp., Z151 and Z150, were closely associated with the blind check (Z151) and laboratory duplicate (Z150), respectively.The same result was also observed with the neighbor-joining tree (NJT) obtained from the same primer set (Figure 3).NJT produced a better resolution of species relationship compared to the BCT.Although minor differences exist among the NJ and BC trees, similar clustering pattern was generated by AFLP and ISSR primer sets.Generally, however, the trees from AFLP data set had stronger bootstrap support values than ISSR-based trees.

DISCUSSION
Molecular marker approaches are considered efficient in fingerprinting plant genome.This study investigated the usefulness and effectiveness of two PCR-based molecular techniques, ISSRs and AFLPs in detecting polymorphism in amaranth.The number of alleles per locus detectable by ISSRs was higher compared to AFLPs.Such high level of polymorphism is to be expected with molecular techniques that are based on replication slippage (Tautz et al., 1986).Similarly, studies have shown that when ISSRs are compared to other marker systems they revealed the highest level of polymorphism (Ray and Roy, 2007;Xu and Sun, 2001).The study also revealed that the assay efficiency index was more than five-fold higher for AFLPs than ISSRs.Therefore, ISSRs have the ability of revealing the highest level of information per single marker while AFLPs can detect the highest number of polymorphisms in a single assay.The high assay efficiency index is a reflection of the efficiency of AFLPs to simultaneously analyze a large number of bands rather than the levels of polymorphism detected at each locus (Pejic et al., 1998).This principle account for the high similarity coefficients observed among the amaranth species for AFLP relative to ISSR.The assay efficiency index for ISSRs, however, can be considerably higher if multiplex PCR and gel-running procedures are adopted, where several microsatellites are simultaneously amplified using multicolour fluorescent technologies (Lindqvist et al., 1996;Heyen et al., 1997;Fuentes et al., 2009).
Di-nucleotide repeat ISSR primers produced the highest average number of bands and generally gave a clear fingerprint pattern compare to tri-nucleotide repeats.This suggested that di-nucleotide repeat ISSR primers are more frequent in amaranth genome compared to trinucleotide repeats.However, using di-nucleotide repeats alone may not be efficient and sufficient to differentiate between amaranth genotypes.Similarly, di-nucleotide repeats ISSR primers yielded the highest amount of polymorphic bands in rice (Blair et al., 1998) and Diplotaxis (Brassicaceae) (Martin and Sanchez-Yelamo, 2000).
Both molecular markers used in the study revealed narrow gene diversity between the laboratory duplicate and the blind check, thus, indicating the reproducibility of the two marker systems in genetic assessment of amaranth.The slightly high polymorphism observed in the blind check relative to laboratory duplicate may be due to heterogeneity present in the amaranth genotype.The distribution pattern of the Amaranthus spp.into different clusters was similar for AFLPs and ISSRs, although the cluster order in the trees was slightly different.Furthermore, the tree robustness was lower nearly for all relationships in the ISSR-based trees than the AFLP-based trees.This could result from a smaller ISSR data set compared to the AFLP data set.
In conclusion, this study may not be sufficient to justify that both markers used in the study strongly support the phylogenetic assessment of amaranth species because fewer number of species were used.However, the phylogenetic trees obtained from these marker systems were related, even though AFLPs and ISSRs differ in nature and principles of mechanisms.The current study has shown that AFLPs and ISSRs are highly reproducible and can generate informative characters useful for  phylogenetic assessment of amaranth.The use of these molecular markers can be valuable for efficient germplasm management and breeding programmes of amaranth.

Figure 2 .
Figure 2. Bayesian consensus tree of the Amaranthus spp.based on AFLP primer combinations and ISSR primers; bootstrap support values are given above branches.A. hypo.-A.hypochondriacus.

Table 1 .
List of amaranth accessions used in the study.
AVRDC, Asian Vegetable Research and Development Centre, Tanzania; NACGRAB, National Center for Genetic Resources and Biotechnology, Nigeria.

Table 2 .
Oligonucleotide adapters and primer combinations used for AFLP analysis.

Table 3 .
Twenty-one ISSR primers screened using five randomly selected amaranth species R = A, G; Y = C, T.

Table 4
).The polymorphic loci ranged from 110 among A. caudatus to 228 among A. cruentus and 16 among A. tricolor to 56 among A. hypochondriacus for AFLP primer combinations and ISSR primers, respectively.However, the percentage of polymorphic loci detected relative to the total number

Table 4 .
Gene diversity and polymorphism at the intraspecific level detected with 14 ISSR primers and 9 AFLP primer combinations.

Table 5 .
Average genetic similarity matrix between four Amaranthus spp. with AFLP primer combinations (above diagonal) and ISSR primers (below diagonal) based on Nei and Li/Dice coefficients.