Assessment of Genetic Diversity in Cucumber (Cucumis sativus L.) Genotypes Using Morphological Characters and AFLP Analysis

For this purpose, 18 cucumber lines were evaluated for their genetic diversity using six morphological characterizations (plant morphology, plant length, length of leaf blade, fruit length, fruit diameter and fruit stem length) and eight AFLP markers. These AFLP primer combinations amplified well and also showed polymorphism. Thus, 1975 AFLP fragments were obtained and 1468 fragments were polymorphic (75.34%). Dendrograms were drawn using UPGMA (Unweighted Pair Group Method) arithmetical averages and according to the UPGMA dendrogram, the cucumber accessions clustered into two main groups. The genetic distances of the dendrogram varied between 0.92 and 0.96. Cluster analysis based on morphological data discriminated all lines into three major clusters in UPGMA dendrogram. The similarity coefficient ranged between 0.888 and 0.982 indicating that the cucumber lines used in the study have a low level of genetic variation. Results obtained from the phylogenetic dendrogram by 8 pairs of AFLP primers were consistent with those from the UPGMA clustering analysis, which were in according with the morphological taxonomy on cucumber. Research Article Article History Received : 28.06.2019 Accepted : 09.01.2020


INTRODUCTION
Cucumber (Cucumis sativus L.) is one of the most economically important and fresh eaten vegetables belonging to the family Cucurbitaceae. Due to its narrow genetic base, the intraspecific genetic diversity in cucumber is relatively low (3-12%) as compared to other Cucumis species as revealed in early studies with marker types such as isozymes, RFLPs (restriction fragment length polymorphisms), AFLPs (amplified fragment length polymorphisms) or RAPDs (randomly amplified polymorphic DNAs), (Knerr et al. 1989;Dijkhuizen et al. 1996;Staub et al. 1997;Cavagnaro et al. 2010). However, all methods except AFLP have been reported to have significant disadvantages for this vegetables (Knerr et al. 1989;Waugh and Power, 1992;Staub et al. 1997). As cucumber has a narrow genetic base, it is desirable to develop populations and transfer genes specific to these species using exotic gene sources that control important characters. The evolutions of genetic diversity, relationships and population structure are very important for vegetable characterization and conservation which enhance agricultural production, leading to sustainable development (El-Esawi et al. 2016).
Morphological markers are not widely used because they can be affected by environmental conditions and biochemical markers are not used much due to their limited number in determining the genetic relationships between plant species and varieties. Genetic diversity studies are currently supported by molecular methods, such as molecular markers (Li et al. 2019). AFLP is an effective method allowing the identification of genotypes, the construction of a high saturation genetic map and gene cloning (Vos et al. 1995;Scott et al. 2000). Sequence information is not needed in the AFLP technique and a high rate of polymorphism is obtained. AFLP could explore variation throughout the entire genome, including both coding and non-coding regions of DNA and therefore genome-wide variation was allowed (Wu et al. 2019). The advantage of the system is also its high reproducibility (Witkowicz et al. 2003).
The determination of the genetic distance between genotypes is very important for the breeders in terms of the control of genetic resources and genetic diversity and the selection of genotypes for crossing. In this study, the morphological, phenological, yield and quality characteristics of selected cucumber genotypes that are valuable for agricultural production and the differences between genotypes were determined by using AFLP molecular markers.

Plant material and DNA extraction
In the study, 18 cucumber genotypes ( Table 1) that were estimated to be genetically and morphologically different, were determined and numbered, and some of them were planted and grown in the greenhouse (Aybak and Kaygısız, 2004). Measurements and observations were taken on 10 plants from each genotype and the phenological, morphological, yield and fruit properties were determined based on UPOV criteria. Some seeds were planted in small vials, and leaves were collected from the seedlings and stored at -80°C for DNA isolation. For this purpose, DNA was isolated from 100 mg of leaf material using the CTAB extraction protocol (Weising et al. 1991). DNA quality and concentration were controlled by running each sample on 1% agarose gel electrophoresis and read on a spectrophotometer at 260 to 280 nm wavelengths.  (Vos et al. 1995;Roldán-Ruiz et al. 2000). Eight AFLP primer sets were used to analyze polymorphisms. Total genomic DNA was digested using the two restrictive enzymes EcoRI (New England Biolabs Ltd, NEB) and MseI (NEB). DNA fragments were ligated with EcoRI and MseI adapters using T4 DNA ligase (NEB) for PCR amplification. After the adapters were ligated to the DNA, they were pre-selected to amplify the existing DNA fragments and eliminate the components that were not ligated to the adapter. In the pre-amplification step, genomic DNA was amplified with AFLP pre-amplification primers EcoRI (5'-GACTGCGTACCAATTCA-3') and MseI (5'-GATGAGTCCTGAGTAAC-3'). The preselective amplification reactions were prepared in a 25 µl reaction mixture, containing 3 µl of DNA sample, 1.25 µL of each preselective primers (50 ng/µl), 2.5 µL reaction buffer (10x), 0.5 µL dNTP mixtures, 2 units of Taq DNA polymerase (Thermo Scientific) and 16.5 µL sterile-double distilled water. The PCR reactions were conducted for 15 cycles of 94 °C for 3 min, 94 °C for 1 min, 65 °C for 1 min, 70°C for 1 min; and then 20 cycles of 94 °C for 1 min, 55°C for 1 min, 72°C for 30 s, plus a final elongation step of 72°C for 7 min. The samples were diluted for selective amplification and PCR reactions were performed with a touch-down cycle as follows: 12 cycles of 94 °C for 45 s, 65 °C for 30 s (a decrease of 0.7 °C per cycle) and 72 °C for 45 s, and then 25 cycles of 94 °C for 45 s, 55 °C for 30s, 72 °C for 45s, and 72 °C for 5 min for a final elongation step. The PCR products were separated by 8% (w/v) polyacrylamide gel electrophoresis.

Morphological Data Analysis
Overall, 50 days after planting, plant morphology, plant length, the length of leaf blade, fruit length, fruit diameter and fruit stem length were determinned. For each of the 6 morphological characters, the mean and standard deviation values were calculated. These morphological features were determined in 10 randomly chosen plants based on the UPOV criteria (UPOV, 2019).

Data analysis
AFLP data from eight primers were transformed into a binary matrix, scored as present "1", absent "0", for further analyses. The total number of fragments, the number of polymorphic fragments, the percentage of polymorphic loci (%) and the polymorphic information content (PIC) were calculated using the software GenAlEx 6.5 (Peakall and Smouse, 2006). The dendrogram was constructed using UPGMA (unweighted pair group method with the arithmetic average) based on Nei's genetic distance (Nei, 1972) and the NTSYS ver 2.10 software (Staub et al. 2005). Similarity indices and pairwise genetic distance values were calculated from AFLP data using the UPGMA method and NTSYS software. Clustering analysis was performed using SPSS22.0 software. To examine the correlation between six morphological characters, the pearson correlation coefficient was calculated using IBM SPSS Statistics 22.0 program.

Morphological characterization
Morphological plant characters including plant length, length of leaf blade, fruit length, fruit diameter and fruit stem length were analyzed in 18 selected cucumber genotypes (Fig. 1). Şekil 1. Çeşitlilik analizi için kullanılan hıyar genotiplerinin genel büyüme şekilleri, yaprak ve çiçekleri (A) ile meyve morfolojisi The CV (coefficient of variation) values for three of the morphological characters including internodes length, fruit length and fruit stem length indicated a high level of variation (i.e.>10%), ( Based on Pair-wise comparisons among the 18 genotypes, an Euclidean distance matrix was obtained and an unweighted pair-group method using arithmetic mean (UPGMA) dendrogram was constructed. At an Euclidean distance 19.1, three major clusters were obtained based on the morphological characters (Fig. 2), (plant morphology, plant length, length of leaf blade, fruit length, fruit diameter and fruit stem length). A large number of genotypes (12 genotypes) grouped together in cluster I.
Cluster II grouped two genotypes. Four genotypes grouped together in cluster III (Fig. 2). At the 0.05 level of confidence, the results from Pearson correlation analysis appeared that plant height was positively correlated with internodes length, and leaf blade length. Internodes length was positively correlated with leaf blade length but negatively correlated with fruit diameter and fruit stem length (Table 3).
The cucumber has a very narrow gene pool that limits the development of new cucumber varieties (Innark et al. 2013). Genetic diversity data in cucumber genotypes are used in cucumber breeding programs to help determine parental lines. For morphological data, most characters showed an extremely narrow range, indicating that the cucumber genotypes used in the study have a low level of genetic variation. From the correlation analysis of six morphological characters, plant height was positively correlated with internodes length, and leaf blade length. This result was logical because internodes length and leaf blade length were considered as the yield components affecting the plant height. The UPGMA cluster analysis of morphological measurements were effective in distinguishing 18 cucumber ecotypes. The results from this study were consistent with the previous results reported by Innark et al. (2013), which evaluated the morphological data correlation in switchgrass (Cortese et al. 2010) and Chrysanthemum morifolium (Shao et al. 2010).   Kong et al. (2006) but lower than that detected by genomic SSRs (Watcharawongpaiboon and Chunwongse, 2008). The observed PIC values of AFLP markers in this study varied from 0.27 to 0.58, however, the PIC values were estimated for 45 SSR primer pairs as ranging from 0.11 to 0.78, with an average of 0.47 (Watcharawongpaiboon and Chunwongse, 2008). Similar results were observed from the cucumber germplasm collection containing a range of ecotypes distributed in China sources by Hu et al. (2010). In their study, PIC values ranged from 0.185 to 0.642 with an average of 0.374.

Similarity indices and genetic distance
Similarity coefficients between cucumber genotypes ranged from 0.888 to 0.982 (Table 5), which was higher than those of reported by Hu et al. (2010) for EST-SSR markers from the cucumber. Similarity coefficients calculated from EST-SSR data varied from 0.542 to 0.941 with a mean value of 0.792 (Hu et al. 2010). The differences in all of these data could be attributed to the differences in the EST analysis. The highest degree of similarity indices (0.982) and the lowest genetic distance (0.018) was observed between 10226 GY and 10222 MO, while the lowest degree of similarity (0.888) and the highest genetic distance (0.112) was observed between 1095 MO and 529 GY. The second highest similarity indices (0.964) were between 1085 MO and 224 GY, while the second lowest similarity indices (0.893) were between 1082 GY and 1102 MO. Most genotypes evaluated in this study were shown to be very closely related and shared a high degree of genetic similarity. Relatively large genetic distances were observed by RAPD profiling in cucumber (between 0.01 and 0.58) (Horejsi and Staub, 1999) suggesting that these genotypes were more unrelated and RAPD analysis was useful in genotypic differentiation for cucumber. On the contrary, the study performed by Park et al. (2000) determined that AFLP analyses created more polymorphisms than either RFLPs or RAPDs. The selected 37 primer combinations produced approximately 3000 bands, of which 339 bands (11%) were polymorphic more than RFLPs or RAPDs. Furthermore, they suggested converting the AFLP marker to a dual-primer PCR based marker to enhance its usefulness in cucumber breeding. All these results indicate that both AFLP markers and genotypes used are the most important determinants of the similarity indices and genetic distances.

Cluster analysis
The similarity matrix for genotypes in cucumber was calculated by using the Dice coefficient method with the NTSYS program. According to the dendrograms, the minimum genetic similarity was 92% while the maximum similarity between cucumber accessions was 96%. The UPGMA analysis of the marker data resulted in two main groups.  Figure 3).

CONCLUSIONS
In this study, we selected eight pairs of AFLP informative primers to assess the genetic diversity and relationships among cucumber genotypes collection. AFLP fragments generated by the 8 AFLP primer pairs assayed in this study were 1975 of which 1468 (75.34%) were polymorphic. The overall mean similarity index calculated based on AFLP fragments amplified using Nei's similarity index ranged from 0.888 to 0.982 with an average of 0.936. There are only a few studies related the characterization of genetic diversity of cucumber accessions by AFLP. The study revealed a low molecular diversity among the cucumber accessions. Morphological traits of the cucumber accessions were used in conjunction with molecular data to determine germplasm collections. These results may help in the selection of accessions as breeding materials for the development of new cultivars.

Statement of Conflict of Interest
Authors have declared no conflict of interest.

Author's Contributions
The contribution of the authors is equal.