Molecular characterization of soybean cultivars by microsatellite markers with universal tail sequence

The objective of this work was to standardize a semiautomated method for genotyping soybean, based on universal tail sequence primers (UTSP), and to compare it with the conventional genotyping method that uses electrophoresis in polyacrylamide gels. Thirty soybean cultivars were genotypically characterized by both methods, using 13 microsatellite loci. For the UTSP method, the number of alleles (NA) was 50 (2–7 per marker) and the polymorphic information content (PIC) ranged from 0.40 to 0.74. For the conventional method, the NA was 38 (2–5 per marker) and the PIC varied from 0.39 to 0.67. The genetic dissimilarity matrices obtained by the two methods were highly correlated with each other (0.8026), and the formed groups were coherent with the phenotypic data used for varietal registration. The 13 markers allowed the distinction of all analyzed cultivars. The low cost of the UTSP method, associated with its high accuracy, makes it ideal for the characterization of soybean cultivars and for the determination of genetic purity.


Introduction
Soybean [Glycine max (L.) Merr.] has been the fastest growing crop in Brazil for the last three decades and accounts for 52% of the cultivated area of the country (Companhia Nacional de Abastecimento, 2013).More than 1,000 varieties of soybean are registered in the National Cultivar Registration Service, and more than 590 cultivars are protected by the National Cultivar Protection Service (Brasil, 2013).
Distinguishability analysis is a basic requirement for DUS tests (distinguishability, uniformity and stability) used for cultivar protection.In the case of soybean, these tests have been hampered by the high phenotypic similarity among cultivars, which is mainly due to the narrow genetic base of cultivated genotypes (Priolli et al., 2004).
Molecular descriptors have been used for cultivar protection purposes by the United States Department of Agriculture (USDA) (Diwan & Cregan, 1997).In Brazil, the molecular analysis is not yet recognized as an official method to aid the cultivar protection process, which is still based on phenotypic characteristics.
Microsatellite markers are widely used for the determination of genetic diversity, paternity tests, varietal purity analysis, and genetic mapping (Schuster et al., 2004;Song et al., 2004;Garcia et al., 2007;Rodrigues et al., 2008;Oliveira et al., 2010;Kalia et al., 2011).With the advent of automatic techniques for genotyping, the use of fluorescent primers for DNA fragment analysis has become quite common (Lazaruk et al., 1998).Although this technique is remarkably practical, synthesis of fluorescently labeled primers is still an expensive step when compared to the synthesis of unlabeled ones (Hayden et al., 2008;Arruda et al., 2010).
Variations of the automated technique for microsatellite genotyping have been developed in order to match the speed and automation of the process with the quality and low cost of the analysis.Among them is a method first described by Oetting et al. (1995), which involves the joint use of three primers (unlabeled forward and reverse primers, and a fluorescently labeled M13 primer).This technique was first described in studies to evaluate genotypes in humans (Oetting et al., 1995).It was also used for genotyping animals (Diniz et al., 2007;Arruda et al., 2010) and several plant species, such as common beans (Oblessuc et al., 2009), rice (Goulart et al., 2011), apricot, cherry, barley, wheat (Hayden et al., 2008), Eucalyptus (Missiaggia & Grattapaglia, 2006), and soybean (Priolli et al., 2010).
Based on Oetting et al. (1995), the universal tail sequence primers (UTSP) technique is proposed for the distinguishability of soybean cultivars using the fluorescently labeled universal primers M13, T3, and T7.In general, these primers do not match sequences present in plant genomes; therefore, they can be used for different plant species.Only three primers are fluorescently labeled.There is no need to label specific primers for each species to be analyzed.This is a great advantage if the high prices of fluorescently labeled primers are considered (Missiaggia & Grattapaglia, 2006).
The objective of this work was to standardize a semiautomated method for genotyping soybean, based on UTSP, and to compare it with the conventional genotyping method that uses electrophoresis in polyacrylamide gels.2008), using a bulk of 50 seeds for each cultivar.Information concerning the genealogy of the cultivars was obtained from the website CultivarWeb (Brasil, 2013).

Materials and Methods
For the UTSP method, a multiplex system was used, as described by Oetting et al. (1995), which included three kinds of primers: 1, a sense microsatellite primer, which carries, besides its specific sequence, a 17 base pair (bp) tail at its 5' end identical to sequences of universal primers commonly used in molecular biology, such as: M13 (GTAAAACGACGGCCAGT), T3 (ATTAAC CCTCACTAAAG) or 7 (AATACGACTCACTATAG); 2, a regular antisense microsatellite primer; and 3, a fluorescently labeled primer, with the same sequence as the sense primer tail (Figure 1).The tails were tagged with either 6-FAM (M13 tail), HEX (T3 tail) or NED (T7 tail) (Life Technologies do Brasil Ltda., São Paulo, SP, Brazil).
PCR reactions (20 μL) contained 30 ng DNA, PCR buffer 1X (10 mmol L -1 Tris-HCl, pH 8.8, 50 mmol L -1 KCl, 0.08% v v -1 Nonidet P40), 1.5 mmol L -1 MgCl 2 , 0.2 mmol L -1 dNTP, 0.2 mmol L -1 of each primer (sense and antisense), and 1.0 U Taq DNA polymerase.For UTSP PCR reactions, the concentration of the tailed sense primer was five times lower (0.04 mmol L -1 ) than the other two primers.Reactions were carried out separately for each microsatellite locus.After optimization, 13 markers were selected based on their amplification profile, the smaller amount of stutter products, and the higher number of alleles (Table 1).All loci used were described by Cregan et al. (1999), except for AF162284 and Satt612, which were obtained from the integrated genetic linkage map constructed by Song et al. (2004).The amplification program consisted of an initial denaturation step at 94°C for 5 min, followed by 35 Pesq. agropec. bras., Brasília, v.48, n.3, p.270-279, mar. 2013 DOI: 10.1590/S0100-204X2013000300005 cycles with denaturation at 94°C for 40 s, annealing at 50°C for 40 s, and extension at 72°C for 40 s.After the thirty-fifth cycle, a final extension step was performed at 72°C, for 25 min.
The cultivars were genotyped by both methods, i.e., conventional PCR evaluated in polyacrylamide gel and PCR with UTSP evaluated by capillary electrophoresis.The fluorescent genotyping (fragment size analysis) was performed in duplicate, on two different days (proof and counter proof).In the conventional method, PCR products were separated by electrophoresis in a 10% polyacrylamide gel, using 1X TBE buffer (89 mmol L -1 Tris-borate, 2 mmol L -1 EDTA, pH 8.0) for approximately 3 hours at 120 V, and stained with ethidium bromide (0.2 g mL -1 ) for 10 min.A vertical electrophoresis tank was used with a 15 cm dual glass plate, separated by 1.0 mm spacers, with a capacity of 60 samples per gel.The fragment sizes were estimated by comparison with a 50 bp DNA ladder (Promega, Madison, WI, USA).In the UTSP method, markers with distinct sizes and tagged with different fluorochromes were grouped in order to simultaneously detect by fluorescence the amplified alleles.The sample used for the electrophoretic analysis consisted of 0.4 µL of the three independent PCR reactions, 8.55 µL Hi-Di formamide, and 0.25 µL ROX 500 molecular size standard.Capillary electrophoresis was performed in the ABI 3130xl equipment, according to the manufacturer's instructions (Life Technologies do Brasil Ltda., São Paulo, SP, Brazil).Data were automatically collected and coded using the software Gene Mapper 4.0 (Life Technologies do Brasil Ltda., São Paulo, SP, Brazil).
To check the effectiveness of the construction of the tailed sense microsatellite primer, a test was performed to compare the sizes (17 bp) of the products amplified by conventional PCR (sense primer with no tail) and by UTSP PCR.The PCR reactions were performed separately using cultivars BMX Magna RR and CD 202, which were analyzed in duplicate for four microsatellite loci (Satt233, Satt005, Sat_085, and Satt540) in both methods, and the electrophoresis was performed as described above for the conventional method, in 10% polyacrylamide gel (Figure 2).The polymorphic information content (PIC) was calculated for each marker according to the expression: in which, P ij is the frequency of the j th allele of the i th marker (Anderson et al., 1993).
The probability of random identity was calculated as described by Schuster et al. (2009), according to the formula: in which P i is the frequency of allele A i , and n is the number of the evaluated marker.For those eventually heterozygous loci, the frequencies used in the formula were that of the summed individual alleles.
The genetic relationships between genotypes were assessed by the complement of the similarity index for codominant and multiallelic variables from the unweighted index used to calculate the similarity matrix.The unweighted index is given by: , in which L is the total number of loci analyzed, and C j is the number of common alleles between cultivar pairs i and i'.
The cophenetic correlation was obtained from the dissimilarity matrix, and cultivar clustering was performed by the unweighted pair group method with arithmetic mean (UPGMA).The correlation between the dissimilarity matrices obtained by the two genotyping methods was performed by the Mantel test with 10,000 simulations.The Genes software, version 2009, was used for these analyses (Cruz, 2006).

Results and Discussion
The expected difference of 17 bp between the amplicons generated by the two genotyping methods was confirmed (Figure 2).The DNA bands obtained by the UTSP method were less intense; however, fewer unspecific bands were obtained by this method.This aspect can be easily observed in the amplification profiles obtained with the marker Sat_085.
The fragment sizes estimated for each marker in both genotyping methods for the 30 evaluated soybean cultivars are shown in Tables 2 and 3.Although the allele size obtained by the UTSP method was 17 bp larger than the corresponding one genotyped by the conventional method, small differences could be observed between the sizes of some of the alleles when the two methods were compared.In the evaluated electrophoresis conditions, fragments differing by less than 4 bp had their identification hampered when genotyped by the conventional method in polyacrylamide gels.
For the UTSP genotyping method, the total number of alleles was 50, ranging from 2 (Satt045) to 7 (Satt005) per marker with an average of 3.85.The PIC ranged from 0.40 (Satt045) to 0.74 (Satt005), with an average of 0.62.For the conventional method, the number of alleles was 38, ranging from 2 (Satt045, Satt070, and AF162283) to 5 (Satt005) alleles per marker, with an average of 2.92.The PIC ranged from 0.39 (Satt045) to 0.67 (Satt079), with an average of 0.56 (Tables 2 and  3).Several authors have used microsatellite markers to characterize soybean genotypes with different goals: wide genome genotyping (Sayama et al., 2011), genetic diversity for different traits (Mian et al., 2009;Singh et al., 2010), and germplasm characterization (Wang et al., 2008;Mulato et al., 2010).In these studies, the number of alleles per marker and PIC values ranged from 3 to 18 and from 0.507 to 0.815, respectively.The results obtained in the present study are in accordance with those of Priolli et al. (2002), who characterized 186 Brazilian soybean genotypes using 12 SSR loci and detected an average of 5.3 alleles per locus and a mean PIC value of 0.64.
The differences found between the two tested methods, regarding number of alleles and PIC values, can be explained by the greater accuracy of the UTSP Table 1.Selected microsatellite primers for the analysis of soybean cultivars, using the conventional and the universal tail sequence primers (UTSP) methods.
the "slippage" of DNA polymerase during DNA synthesis, producing fragments sizes similar to the real size of the allele (Hauge & Litt, 1993).
The stutter can overlay heterozygote alleles and interfere with the interpretation of the generated amplification profile (Miller & Yuan, 1997).Higher precision in the detection of alleles can be obtained under electrophoretic systems with high resolution capabilities.This explains why a higher number of heterozygotes could be detected in the UTSP method.
An important parameter to be considered for cultivar identification is the probability of random identity (PRI), which corresponds to the probability that an individual taken at random and analyzed with a specific set of markers will be equal to any other individual in the population (Schuster et al., 2009).PRI is calculated by the product of individual allele frequencies at each locus.In the case of the UTSP method, PRI varied from 1.41x10 -8 (CD 240RR) to 9.54x10 -6 (BRS 184), whereas for the conventional method, PRI varied from 1.12x10 -6 (BRS Pampa RR) to 8.45x10 -5 (CD 219RR).These numbers indicate that method, which uses an automatic DNA sequencer, as opposed to gel electrophoresis in the conventional method (Wenz et al., 1998).These differences can be overcome if longer polyacrylamide gel is used under denaturing conditions (Sánches-Pérez et al., 2006).At least two extra alleles could be detected for markers Satt005, GMABAB, and AF162283, using the UTSP method.Amplification artifacts may also have influenced the number of alleles detected by either method.It is a well-known fact that Taq DNA polymerase adds an extra nucleotide (usually adenosine) to the 3' end of the newly synthesized fragment.This modification may affect the precise determination of the fragment size (Pompanon et al., 2005).A larger extension time (more than 20 min) after the final PCR cycle increases the chance of all fragments being adenylated, minimizing the effect of the extra nucleotide on fragment size determination, therefore, the final extension time was increased in both genotyping methods.Another common artifact that can influence data interpretation is the stutter, which is explained by Conclusions 1.The universal tail sequence primers (UTSP) method is very effective to detect polymorphisms among soybean cultivars, when compared to the conventional genotyping method using polyacrylamide gels.
2. The low cost of the UTSP method associated with its high accuracy, possibility for automation, and high throughput analysis makes it ideal for the characterization of soybean cultivars and for the determination of genetic purity.
3. A highly robust and informative set of SSR markers is presented here, which can be used for diversity analysis and for identification of soybean cultivars.

Figure 1 .
Figure 1.Universal tail sequence primers (UTSP) method: A, primers used for amplification of the fluorescent fragment, including a sense microsatellite primer with universal tail sequence (A1), a fluorescently labeled primer with the same universal tail sequence (A2), and an antisense microsatellite primer (A3); B, the first steps of the amplification cyclethe specific regions of the genome complementary to the microsatellite primers are represented by black boxes; C, pairing of a fluorescently labeled primer to a complementary region (hatched gray box) synthesized in the early stages of PCR -the gray box represents the sequence of the tail incorporated to the amplicon due to annealing of the primer shown in B; and D, fluorescently labeled amplicons in the final stages of PCR.

Figure 2 .
Figure 2. Comparison between the amplification profiles obtained with the conventional (*Conv.)method and with the universal tail sequence primers (UTSP) method.The markers are identified at the bottom portion of the figure.Cultivar BMX Magna RR is indicated by number 1, and CD 202, by number 2. "M" indicates a 50 bp DNA ladder.Analyses were done in duplicate.The sizes of the DNA ladder fragments (bp) are indicated by arrows on the left side of the figure.The other arrows indicate the specific amplification product for each marker.