Genetic Analysis of Diversity within a Chinese Local Sugarcane Germplasm Based on Start Codon Targeted Polymorphism

In-depth information on sugarcane germplasm is the basis for its conservation and utilization. Data on sugarcane molecular markers are limited for the Chinese sugarcane germplasm collections. In the present study, 20 start codon targeted (SCoT) marker primers were designed to assess the genetic diversity among 107 sugarcane accessions within a local sugarcane germplasm collection. These primers amplified 176 DNA fragments, of which 163 were polymorphic (92.85%). Polymorphic information content (PIC) values ranged from 0.783 to 0.907 with a mean of 0.861. Unweighted pair group method of arithmetic averages (UPGMA) cluster analysis of the SCoT marker data divided the 107 sugarcane accessions into six clusters at 0.674 genetic similarity coefficient level. Relatively abundant genetic diversity was observed among ROC22, ROC16, and ROC10, which occupied about 80% of the total sugarcane acreage in China, indicating their potential breeding value on Mainland China. Principal component analysis (PCA) partitioned the 107 sugarcane accessions into two major groups, the Domestic Group and the Foreign Introduction Group. Each group was further divided based on institutions, where the sugarcane accessions were originally developed. The knowledge of genetic diversity among the local sugarcane germplasm provided foundation data for managing sugarcane germplasm, including construction of a core collection and regional variety distribution and subrogation.


Introduction
Sugarcane (Saccharum spp.) is the most important sugar crop in China by producing more than 90% of the total consumable sugar [1,2]. Sugarcane is also an important energy source as an herbaceous plant. It was reported that genetic improvement of sugarcane accounted for 75% of the yield increase in Hawaiian sugar industry in the 1950s and >60% yield increase on Mainland China in recent decades [1,3]. According to the heterosis theory, parental lines with larger genetic distance values must be selected for crossing during cultivar development on the premise that the main attributes of the parental lines are complementary [4][5][6]. Evaluation of sugarcane germplasm with DNA markers helps understand the extent of genetic diversity among parental lines and the difference in genetic background among accessions. The knowledge on phylogenetic relationship among accessions of sugarcane germplasm collection will guide parental selection for the development of new cultivars [1,[5][6][7]. Therefore, genetic diversity analysis of sugarcane germplasm is important to the success of sugarcane breeding.
More recently, Collard and Mackill [32] developed another DNA marker in rice, the start codon targeted (SCoT) marker, based on the short conserved nucleotide sequence that flanks the start codon ATG. Similar to RAPD and ISSR, SCoT marker involves a single oligonucleotide primer and is PCR based. However, due to the simultaneous binding of the primer on both DNA strands, the sequence between the two binding sites is amplified. As a relatively new marker technique, the SCoT marker has the following advantages: simple, low-cost, highly polymorphic, gene-targeted, and abundant in the genome. It has been utilized on different plant species, including rice [32], longan [33], grape [34], potato [35], orange [36], mango [37][38][39], peanut [40], and Cicer [41]. The objective of the present study was to explore the potential utility of the SCoT marker technique in assessing the genetic diversity and phylogenetic relationship within a local sugarcane germplasm collection.

Sugarcane Accessions within a Local Sugarcane Germplasm
Collection. One hundred and seven sugarcane accessions from a local sugarcane germplasm collection were involved in the study ( Table 1

SCoT Primers.
Forty SCoT primers of 18 nucleotides each were designed based on the short conserved nucleotide sequence flanking the start codon ATG. The conserved sequence had the "ATG" codon fixed at positions +1, +2, and +3, "G" at position +4, "C" at position +5, and "A, " "C, " and "C" at positions +7, +8, and +9, respectively. The primers differed from one another by at least one nucleotide at other positions, with an emphasis on variations at the 3 end, which allowed specific annealing and amplification events to occur [42]. The 40 SCoT primers were initially evaluated for PCR robustness on two sugarcane accessions, ROC22 and FN02-3504.

Extraction of Sugarcane Genomic DNA.
Leaf samples were collected from the top visible dewlap leave blade of each accession without any disease symptom. The Biospin plant genomic DNA extraction kit (Bioer Technology CO., Ltd., Hangzhou, China) was used to extract the genomic DNA from sugarcane leaf tissue. Both agarose gel electrophoresis and ultraviolet spectrophotometer were used to estimate quality and quantity of DNA samples. Initial denaturation was carried out at 94 ∘ C for 5 min, followed by 35 cycles of 94 ∘ C for 1 min, 51 ∘ C for 1 min, 72 ∘ C for 2 min, and final extension at 72 ∘ C for 5 min. The amplification products were separated in 1.2% agarose gels containing 0.5 g/mL of ethidium bromide through electrophoresis in 1X TBE buffer solution at 5 V/cm and visualized under a UVP ultraviolet transilluminator (Spring Scientific, New York, USA).

Statistical
Analysis. PCR products were scored visually.
To minimize errors, only clearly distinguishable bands were scored. Presence of a band was recorded as "1" and absence of a band was recorded as "0". Polymorphism information content (PIC) is a property value of a marker based on its allelic number and distribution frequency in a population. PIC for marker was calculated using PIC = 1−∑ 2 according to Botstein et al. [43], where is the allele frequency at locus . Percentage of polymorphic bands (PPB), number of observed alleles ( ), number of effective alleles ( ), Nei's genetic diversity index (ℎ), Shannon's information index ( ), total genetic diversity index ( ), genetic diversity index within series ( ), coefficient of genetic differentiation ( st ), and gene flow ( ) were calculated using POPGENE 1.31 [44]. Unweighted pair group method of arithmetic averages (UPGMA) was used for cluster analysis using NTSYSpc [45]. Principal component analysis (PCA) was conducted using a Dcenter module [45,46].

SCoT Polymorphism in Sugarcane.
Upon initial evaluation, only 20 out of the 40 designed SCoT primers were able to prime amplification of DNA fragments. Nucleotide sequence and GC content of the 20 SCoT primers are listed in Table 2. The 20 SCoT primers amplified a total of 176 DNA fragments from the 107 sugarcane accessions, with 5 to 11 fragments per primer. Of the 176 fragments amplified, 163 were polymorphic. Primers P1, P29, and P31 amplified the highest number of DNA fragments, with an average of 11 DNA fragments per

Cluster Analysis.
A homology tree is shown in Figure 1. Of the YG-series accessions, YG16 (YN73-204 × CP86-1633) and YG18 (YN73-204 × CP72-1210) shared the same female parent (YN73-204). The similarity coefficients were 0.511 between YG16 and YN73-204 and 0.449 between YG18 and YN73-204, causing YG16 and YG18 to be placed in different cluster (Cluster I) from the female parent YN73-204 (Cluster VI). Since the mid-1980s, the "ROC"-series accessions have been one of the most important sources of parental materials for the sugarcane breeding programs on Mainland China. The pairwise genetic similarity coefficients among ROC1, ROC10, ROC20, ROC24, and ROC25 were all high, resulting in the grouping of these "ROC"-series accessions in Cluster IV. However, two other "ROC"-series accessions, namely, ROC22 and ROC16, which along with ROC10 had been grown in the largest planting areas on Mainland China in the recent 25 years, were placed into Clusters I and II, respectively. The four "India"-series accessions were placed in Subclusters IV-II. The three "Brazil"-series accessions were placed into Cluster I (CI-2003), Cluster II (RB76-5418), and Cluster VI (RB72-454), respectively. The pairwise genetic similarity coefficient was 0.648 between CI-20030 and RB76-5418, 0.420 between CI-2003 and RB72-454, and 0.545 between RB72-454 and RB76-5418, respectively, indicating a high genetic diversity among the three Brazilian accessions. (Figure 2): the Foreign Group and the Domestic Group. In the Foreign Group, the "CP"-series accessions are introduced from the USA. Although the "YC"-series accessions were developed by the Guangzhou Industrial Sugar Research Institute, most parental clones of these "YC"-series accessions were "CP"-series accessions. The "ROC"-series accessions were introduced from Taiwan. The "Indian"-series accessions and most of the "Other"-series accessions were introduced from other foreign countries. In the Domestic Group, both the "FN"-and "MT"-series accessions were released by the two sugarcane breeding programs in Fujian Province, one at the Fujian Agriculture and Forestry University and the other at the Fujian Academy of Agricultural Sciences. The "GT"-, "YT"-, and "YZ"-series accessions were released from the sugarcane breeding institutes in Guangxi, Guangdong, and Yunnan Provinces, respectively. There was a large difference in genetic basis between the Foreign Group and the Domestic Group ( Figure 2). Therefore, the PCA results suggested that it would be important to combine both foreign and domestic germplasm accessions for the improvement of sugarcane genetic diversity in Chinese sugarcane breeding programs.

Genetic Diversity within the Local Sugarcane Germplasm
Collection. The number of SCoT polymorphic bands (NPB) varied from 74 to 164 across the 12 series of sugarcane accessions ( Table 3). The highest NPB value (164) was found in the "Other"-series. The "Q"-series ranked the second to the highest. The "CP"-and "ROC"-series had similar numbers of polymorphic bands. The least number of polymorphic bands of 74 was observed in the "YZ"-series, which also had the lowest PPB value. The "Co"-series had the second lowest PPB value (43.75%). The extent of variability among NPB, PPB, , , ℎ, and indices also indicated a high level of genetic diversity among the 12 series (Table 3). The values of these 12 series ranged from 1.4205 to 1.9318, while the values ranged from 1.2844 to 1.6440. The observed percentages of effective alleles were from 1.2844 to 1.6440. The "YZ"-series had the lowest observed percentage of effective alleles of 1.2844. These results suggested that the 20 SCoT primers had high amplification efficiencies and thus could be an effective method for the genetic diversity analysis of sugarcane germplasm collections.
The genetic diversity index ℎ reflected the diversity and differentiation among the germplasm collections. Shannon's index was used to evaluate the genetic diversity within and between the series. The higher the index, the higher the genetic diversity. The ℎ values of these 12 series were from 0.1635 to 0.3619. Shannon's index ( ) varied from 0.2411 to 0.5298. The "Other"-series had the highest ℎ (0.3619) and the highest (0.5298) values, because the 10 accessions were from different breeding institutes in other countries. Therefore, the genetic distances among accessions within the "Other"series were farther and the differences in their genetic basis were larger. If different accessions within the same series were  FY0901  FY0904  FY0903  FY0906  FY0909  FY0908  FY0907  FY0905  FY0916  FY0902  Q138  Q190  FY0912  Q208  FY0913  FY0914  FY0915  FY0917  GT97-40  H56-752  CI-2003  YG24  MT96-1027  YG26  YZ03- II   III   IV   V   VI   I   VI-II   VI-I   IV-I   IV-II   III-I  III-II   II-I   II-II   I-I I-II Figure 1: Cluster analysis dendrogram of 107 accessions from a local sugarcane germplasm collection based on SCoT marker data. crossed, one would expect a higher genetic diversity among the cross-progeny. The ℎ and values of the "ROC"-series were 0.3496 and 0.5122, respectively, which were similar to those of the "CP"-series. The "CP"-and "ROC"-series had high genetic diversity, ranking the second and the third after the "Other"-series. Lower genetic diversity and Shannon's index were observed within both the "MT"-series (0.1785 and 0.2654) and "YZ"-series (0.1635 and 0.2411).
Since the "Other"-series accessions were from different breeding institutes of countries other than China, the genetic diversity among accessions of the "Other"-series was not compared in this study. However, the genetic diversity among the remaining 11 series was analyzed. The NPB and PPB were 175 and 99.43% for the 11 series, which were higher than those within any series, including the "Other"-series (  to that within each series. The ℎ value (0.3592) and value (0.5343) were significantly higher than those of any series except the "Other"-series. The results indicated that the selection of cross-parents released from different breeding institutes would be beneficial in developing new sugarcane cultivars because of higher genetic diversity. The total genetic diversity index ( ) (0.3640) among the 107 accessions was not only higher than between series (0.3592), but higher than within series (0.2526) as well. The genetic differentiation coefficient ( st ) between series was 0.3060, and the gene flow ( ) was 1.1340, indicating that gene flow and genetic differentiation occurred between the series as well.

Discussion and Conclusions
Improvement of sugarcane through genetic manipulation based on sexual crossing has been a directed, ongoing process since 1888 [5]. However, conventional breeding technology generally takes 12 to 15 years to develop a sugarcane cultivar with the first selection cycle on about 0.3 million seedlings. Because sugarcane is a clonally propagated crop, creation of new genotypes is only done through sexual crossing. The seedling and ratoon crops of new genotypes are subjected to several cycles of evaluation and selection under various environments in comparison to concurrent elite cultivars as checks [1]. Choosing parental accessions is the most crucial step in any sugarcane improvement program. There has never been single incidence of developing a sugarcane cultivar out of a poor cross [1,[5][6][7]. Therefore, genetic diversity analysis of sugarcane germplasm based on molecular evaluation and characterization is the basis for effective germplasm utilization. A high genetic diversity and complementarity between two parental accessions are crucial for producing high quality seedling populations of hybrid progeny [1,[4][5][6][7].
SCoT is a new gene-targeted technique based on the nucleotide sequences at the translational start site ATG. The technique has been validated in several plant species [32][33][34][35][36][37][38][39][40][41]47]. In this study, the utility of 20 SCoT primers was explored through evaluation of genetic polymorphism among 107 accessions of a local sugarcane germplasm collection. The percentage of polymorphic bands (PPB) detected reached 92.85%. The polymorphic information content (PIC) values of these SCoT bands ranged from 0.783 to 0.907 with an average of 0.861, which was much higher than that of the SSR markers (0.57) reported by Filho et al. [48]. The average observed percentage of effective alleles was 85.49%, indicating the highly polymorphic and robust nature of these SCoT markers.
One of the modern sugarcane breeding objectives is to broaden the genetic basis of cultivars [1]. Genetic similarity analyses in sugarcane suggested that cross-progeny from the same parental combination could display large genetic differences due to the complex polyploid genome of sugarcane [1,6,7], which was the basis for seedling selection [1][2][3]. UPGMA clustering and principal component analyses of the SCoT marker data indicated that the extent of genetic diversity among the three most popular "ROC" accessions in China, namely, ROC22, ROC16, and ROC10, was relatively high. The average genetic similarity coefficient among the 22 newly released accessions from the Chinese National Yield Trials Program was only 0.593, indicating a fairly abundant genetic diversity among these accessions. The principal component analysis divided the 107 sugarcane accessions into distinct domestic and foreign groups. Crossing between domestically bred sugarcane accessions with foreign introductions may help enhance the genetic diversity level of sugarcane germplasm in China.
Based on the geographic origin, the 107 sugarcane accessions were sorted into 12 series, namely, "Brazil"-, "CP"-, "FN"-, "GT"-, "India"-, "MT"-, "Q"-, "ROC"-, "YC"-, "YT"-, "YZ"-, and "Other"-series. The "Other"-series included 10 accessions that belonged to miscellaneous breeding institutes. The genetic diversity (ℎ) indices among these 12 series ranged from 0.1635 to 0.3619. The highest ℎ value was found in "Other-series" (0.3619), followed by ROC-(0.3496) and CP-(0.3466) series, respectively. The lowest ℎ value existed in the YZ-series (0.1635). It was noteworthy that the genetic diversity between any two series was much greater than among the accessions within the same series. A previous report concluded that a gene flow index of > 1 would be indicative of no significant differentiation among populations [49]. The gene flow index was moderate ( = 1.1340), indicating a high level of genetic diversity within populations that were not prone to genetic drift. The mode of pollen dispersal, which determined the gene flow among populations, might partly account for this moderate differentiation [1]. This was further confirmed by the low level of interpopulation genetic differentiation manifested by the low gene differentiation coefficient ( st ) among populations (0.3060). Therefore, we deduced that the genetic diversity among the 107 accessions had existed mainly between different series. From all the above, the knowledge of genetic diversity among the local sugarcane germplasm collection would help direct future sugarcane cross-breeding programs in China. It would also provide foundation data for managing sugarcane germplasm resources, including the construction of a core collection and regional variety distribution and subrogation.