GENETIC DIVERSITY AND RELATIONSHIPS AMONG SOME BARLEY GENOTYPES FOR NET BLOTCH DISEASE RESISTANCE USING RAPD, SCOT AND SSR MARKERS

arley (Hordeum vulgare L.) is the fifth important cereal crop species in crop production world-wide after maize, wheat, rice, and soybean. It is also a model species for genetic studies while it is an annual and diploid self-pollinating species and has a relatively short life cycle. Net blotch of barley, caused by the phytopathogen Pyrenophora teres constitutes one of the most serious problems on barley production world-wide (Shipton et al., 1973). Net blotch disease cause significant yield loss and affect the grain quality negatively. Losses due to net blotch could reach 50% of yield with possible complete loss depending on cultivar susceptibility and environmental conditions (Steffenson et al., 1996). Detection of resistance sources to net blotch and understanding their genetic background are very important in developing new resistant varieties to such disease. Net blotch resistance is controlled by several genes and dependent on the source of resistance, the development stage and the pathotype used for testing (Graner et al., 1996; Svobodova et al., 2011).

1. Genetics Dept., Faculty of Agriculture,Egypt 2. Barley Dept.,Field Crops Res. Institute,ARC,Egypt arley (Hordeum vulgare L.) is the fifth important cereal crop species in crop production world-wide after maize, wheat, rice, and soybean. It is also a model species for genetic studies while it is an annual and diploid self-pollinating species and has a relatively short life cycle. Net blotch of barley, caused by the phytopathogen Pyrenophora teres constitutes one of the most serious problems on barley production world-wide (Shipton et al., 1973). Net blotch disease cause significant yield loss and affect the grain quality negatively. Losses due to net blotch could reach 50% of yield with possible complete loss depending on cultivar susceptibility and environmental conditions (Steffenson et al., 1996). Detection of resistance sources to net blotch and understanding their genetic background are very important in developing new resistant varieties to such disease. Net blotch resistance is controlled by several genes and dependent on the source of resistance, the development stage and the pathotype used for testing (Graner et al., 1996;Svobodova et al., 2011).
Using molecular marker technology in barley offers high efficiency tools for indirect selection and would enhance the efficiency and accuracy of screening for net blotch resistance. Furthermore, quantitative analysis proved to be useful for determining genes controlling complex traits and provides a more accurate estimation of gene location because of its lower sensitivity to even modest numbers of phenotypic mis-scores (Wright, 1998), barley germplasm identification and classification (Struss and Plieske, 1998).
The association between molecular markers and phenotypes is one of the most significant factors in the field of molecular genetics and molecular breeding. It provides substantial landmarks for elucidation of genetic variability and detection of genomic regions that are responsible for the trait, which plays an essential role in the strategic improvement of barely using marker-assisted selection (Adawy et al., 2008).
These molecular markers had been used in barley for detecting genetic diversity and genotype identification. Of these techniques, Random Amplified Polymorphic DNA (RAPD) has several advantages, such as simplicity of use, low B cost, and the use of small amount of plant material. RAPDs were proved to be useful as genetic markers in the case of selfpollinating species with a relatively low level of intraspecific polymorphism, such as cultivated barley (Tinker et al., 1993).
A new molecular marker system called Start Codon Targeted Polymorphism (SCoT) was described by Collard and Mackill (2009), based on the observation that the short conserved regions of plant genes are flanked by the ATG translation start codon. The technique uses single primers designed to anneal the surrounding regions of the ATG initiation codon on both DNA strands. The generated amplicons are possibly distributed within gene regions which contain genes on both plus and minus DNA strands. The utility of primer pairs in SCoTs was described by (Gorji et al., 2011). SCoT markers are reproducible, and it is suggested that primer length and annealing temperature are not the only factors determining reproducibility. They are dominant markers, however, while a number of co-dominant markers are also generated during amplification, and thus they could be used for genetic diversity analysis (Collard and Mackill, 2009).
Microsatellites are widely used as genetic markers because they are codominant, multi-allelic, easily scored and highly polymorphic. However, a major drawback of SSR markers is the time and cost required to characterize them (Fisher et al., 1996). SSRs are tandemly arrayed repetitive sequences that are spread throughout the eukaryotic genomes and shown to be the most variable component of the genome with a high level of molecular evolution (Hemleben et al., 2000). Microsatellites are suitable for determining paternity, population genetic studies and recombination mapping. It is also the only molecular marker to provide clues about which alleles are more closely related (Goldstein and Pollock, 1997).
This study aimed to: (1) determine the relationship between natural net blotch disease and yield-related characters in 20 barley genotypes and (2) recognize new resistant barley sources and identification of reliable molecular genetic markers for such disease resistance that can be applied in breeding programs.

Plant material
This study was carried out at the molecular genetic laboratory, Genetics Department, Faculty of Agriculture, Kafr  Table (1).

Experimental design
Seeds were hand drilled at the recommended sowing rate of barley in Egypt (50 kg/fed.) in the first week of December. Each plot was sown in (4.2 m 2 ) six rows of 3.5 m long, with 20 cm between rows. This experiment was laid out in randomized complete blocks design with three replications. All cultural practices were applied at the proper time according to Ministry of Agriculture recommendations.

Natural
infection with Pyrenophora teres conidia, the causal of barley net blotch, was conducted under natural field conditions. Records of the disease were denoted after disease on set using the (0-9) scale adopted by Leath and Heun (1990). Disease symptoms were measured at heading stage.

Statistical analysis
The components of the analysis of variance were evaluated for each experiment as described by Kearsey and Pooni (1996). Mean performance for all traits of genotypes and cultivars included in this trial were compared using LSD at 0.05 and 0.01 levels of probability. Simple correlation (r) coefficients among all studied traits were calculated according also to Kearsey and Pooni (1996). All statistical analyses were performed using the computer software Costat Computer Program according to (Snedecor and Cochran, 1969).

DNA isolation and primer selection
DNA was isolated using Cetyl trimethyl ammonium bromide (CTAB)based procedure for plants from fresh leaves of the used twenty genotypes of barley (Murray and Thompson, 1980). Three different types of DNA markers (five RAPD, three SCoT and eight SSR primers) as shown in Table (2), were used to screen genetic polymorphism among the 20 barley genotypes and identification of molecular markers associated with net blotch disease resistance. These primers were synthesized by iNtRON Biotechnology, Inc, Korea.

Amplification condition
Amplification reactions were applied using 20 μl reaction mixture containing the following; 1 μl of template DNA (40 ng/μl), 1.0 μl of primer (10 pmol/ μl) in RAPD and SCoT analysis, 1 μl from each primer (forward and reverse in SSR analysis, 10 μl 2X PCR Master mix solution [(i-Taq TM ) iNtRON Biotechnology] and 7-8 μl of sterile ddH 2 O. The reaction mixtures were overlaid with 20 μl of mineral oil per sample. PCR amplification condition was carried out in thermal cycle (Perkin Elmer Cetus) programed. The reaction was subjected to one cycle at 94C for 2 min. (initial denaturation), followed by 35 cycles of 20 sec. at 94C, 30 sec.,1 min. and 1 min. at 30, 50, 55C (for RAPD, SCoT and SSR, respectively) and 30 sec. at 72C, final extention for 5 min at 72C (one cycle) then at 4C for keeping.
Amplification products were separated by horizontal gel electrophoresis unit using 1.5% agarose gel. Electrophoresis was carried out fewer than 70 volts for 15 min., then 90 volts for 90 min. Bands were detected on Benchtop UV-transilliminator and photographed using photo Doc-It TM imaging system. The molecular size of the amplified products was determined against 1 Kb DNA ladder with stain (SibEnzyme) and 1 Kb plus DNA ladder (TIANGEN, cat.no. MD113).

Data analysis
DNA banding patterns generated from RAPD and SSR techniques were analyzed by GelAnalyzer 3 program. Amplification with some arbitrary RAPD primer was repeated three times, and consistent bands for each primer were selected for data generation. Only consistent and reproducible bands were used to run the corresponding statistical analysis. DNA polymorphic bands were registered as discreet variables considering "1" for presence and "0" for absence to construct a binary data matrix. From this matrix, the genetic similarity (GS) was estimated using Nei & Li coefficient's (Nei and Li, 1979) by computational package MVSP 3.1. Also, depending on this matrix, clus-ter analysis was applied using the same program. The resulting matrix was analyzed on the basis of the Unweighted Pair Group Method with Arithmetic Mean (UPGMA).
The informational certainly of primers to differences between genotypes was analyzed by means of the estimation of their Polymorphic Information Content (PIC) and Resolving Power (RP). PIC was calculated using the formula reported by Roldan-Ruiz et al. (2000) as follow: where PICi is the polymorphic information content of the locus i, fi is the frequency of the present bands, and (1-fi) represents the frequency of the absent bands. The PIC of each primer was calculated using the average PIC value from all loci of each primer. Resolving Power was calculated according to Prevost and Wilkinson (1999) using the formula (Rp = ∑ I b ), where I b represents the informative bands, which was calculated with: I b = 1-[2 x (0.5 -p)] where p is the proportion of genotypes containing the bands.

RESULTS AND DISCUSSION
Mean squares of all traits of the studied genotypes in two seasons are presented in Table (3). Results pointed out those mean squares of genotypes were highly significant for all traits in both seasons.
Mean performances of the twenty genotypes for eleven different characters under study are presented in Table ( Table ( Varied response by barley lines confirms that they were genetically diverse and that their response to disease may be under the control of several resistantce genes (Liu et al., 2011;Owino et al., 2014) which may have conditioned their response to the disease. Low temperatures coupled with higher relative humidity at El-Hosainia plain Agricultural Research Station may have favored spore production and multiple infections of genotypes (Agrios, 2005;Kosiada, 2008). Maximum spore produc-tion has been reported to occur at 25C and at a high relative humidity (Kosiada, 2008). These conditions may have contributed to the observed significant variations in disease response in the two seasons. Higher amounts of rainfall observed in both seasons at early growth stages may caused a rise in moisture levels in the host plants thus causing increased infection by the pathogens (Agrios, 2005).
Pyrenophora teres has the ability to undergo sexual reproduction and this may cause an increase in frequency of pathotypes that have the ability to adapt to the changes in the genetic makeup of the host population (Statkevičiūtė et al., 2010). Such pathotypes can also be increased in frequency due to the influence of selection pressure from growing resistant varieties (McDonald and Linde, 2002).

Correlation coefficient
Correlation coefficient is important in plant breeding where it measures the degree of association between two or more characters. The correlation coefficients among the studied characters of barley genotype are shown in Table (5). Significant positive correlation was observed between days to heading and days to maturity, also significant positive correlation was observed between days to maturity and each of biological yield and grain yield. Significant and positive correlation was detected between plant height and each of spike length, number of grains/spike, number of spikes/m 2 , 1000- Also grain yield showed positive correlation with harvest index. While harvest index showed negative and significant correlation with net blotch disease. From the previous results, it could be concluded that, it is logically presence of positive correlation between grain yield and one or more traits of its components and this happened in this study. Also, negative correlation between net blotch and each of 1000-grain weight, biological yield, grain yield and harvest index was due to the negative effect of net blotch on plant leaf area. Kashif and Khaliq (2004), Saleem et al. (2006) and Muhammad et al. (2010) found positive correlation between grain yield and most of its components. Riggs et al. (1981) reported that a high meaningful and positive correlation was existed between harvest index and grain yield in barley. Kiflu (2009) also reported significant and positive correlation between days to heading and days to maturity.

Polymorphism as detected by RAPD analysis
Five RAPD primers were used to study the genetic diversity and relationships among the 20 barley genotypes.
These primers produced multiple band profiles ( Fig. 1) with different amplified DNA bands (Table 6). The molecular size of the amplified DNA bands ranged from 161 bp to 1656 bp. A total of 48 amplified fragments (loci) were obtained, out of them 34 (70.83%) were polymorphic. The total number of polymorphic DNA fragments ranged from high that was scored by the primer OPH-03 (12), while the lowest number was recorded by primer OPH-01 (3). The polymorphism percentage ranged from 37.5% (OPH-01) to 87.5% (OPH-04). These variations in the number of bands amplified by different primers are influenced by variable factors such as primer structure and number of annealing sites in the genome (Kernodle et al., 1993). Results showed also that five fragments out of the 34 polymorphic ones were genotype-specific markers. The re-solving power (RP) ranged from (7.7 to 15.7). The average polymorphic information content (PIC) was 0.23, ranging from 0.04 (OPH-01) to 0.33 (OPH-04).
The percentage of polymorphic bands (70.83) expressed by random primers is in the range of other reports on other RAPD studies of barley which were 74% (Karim et al., 2010) and 88% (Ciulca et al., 2010). These results agree with those reported by Sosinski et al. (2000), Saker (2005 and Zaki and Al-Masry (2008).
The RAPD based-dendrogram ( Fig. 4) was divided into two clusters at the genetic similarity percentage 76.2% and each cluster was divided into two subclusters. The first cluster was separated in 81.7% genetic similarity percentage, the first subcluster included the most resistant genotypes and located together such as (Line 77, Line 38, Line 81, Line 26 and Line 15), while the second subcluster consisted of the genotypes (Giza 136, Giza 135 and Line 91). On the other hand, the second cluster which was separated in 79.1% of genetic similarity percentage included most of the susceptible genotypes according to morphological data (Giza 117 and Giza 2000) but they were found in two different subclusters. The other genotypes ranged from moderately resistance or moderately susceptible (Giza 123, Giza 124, Giza 126, Giza 131, Giza 132, Giza 133, Giza 134, Giza 129, Line 9 and Line 46) also found in this cluster. These results agreed with those of Peltonen et al. (1996) and Zaki and Al-Masry (2008).

Polymorphism as detected by SCoT analysis
Three SCoT primers were used to study the genetic differences and relationships among the 20 barley genotypes as shown in Fig. (2) and Table (6). The molecular sizes of the amplified bands ranged from 169 bp to 2277 bp. A total of 31 major SCoT amplified fragments were obtained, out of them 24 (77.42%) were polymorphic and the polymorphism percentage ranged from 55.56% (SCoT-8) to 91.67% (SCoT-9). The total number of polymorphic DNA fragments ranged from high scored by the primer SCoT-9 (11), to low scored by the primer SCoT-8 (5). These variations in the number of bands amplified by different primers are influenced by variable factors such as primer structure and number of annealing sites in the genome (Kernodle et al., 1993).
Results showed that two fragments out of the 24 polymorphic ones were unique (genotype-specific markers). The resolving power (RP) ranged from 9.6 to 12.7 for SCoT-7 and SCoT-8, respectively. The polymorphic information content (PIC) ranged from 0.18 to 0.33 for SCoT-8 and SCoT-9, respectively. The percentages of polymorphic bands expressed by SCoT primers were compared to earlier reports of other SCoT studies on barley. These results agree with those of Amirmoradi et al. (2012) who detected a total of 112 bands among 38 accessions belonging to eight annual Cicer species using nine SCoT markers, of which 109 were polymorphic. The number of bands ranged from 7 to 17 with an average of 12.4 per primer. The overall size of amplified products ranged from 220 to 2250 bp. Polymorphism percentage ranged from 86.6% to 100% with average polymorphism of 97% across all accessions.
The dendrogram constructed based on SCoT markers (Fig. 4) was separated at 69.8% similarity percentage into two clusters. The first cluster was divided into two subclusters at 75.3% similarity percentage. The first subcluster contained the most resistant ICARDA genotypes (Line 77, Line 81 and Line 91), while the second subcluster contained the Egyptian genotypes in two groups, Giza 135 in the first group, while Giza 126 and Giza 124 was found in the second group. On the other hand, the second cluster was separated at 75.7% similarity percentage into two subclusters, the first included the genotypes (Giza 136, Line 38, Line 15, Line 26 and Line 46), while the second subclusters included most of susceptible genotypes (Giza117 and Giza 2000) and other genotypes ranged from moderately resistance to moderately susceptible (Gi-za123, Giza131, Giza132, Giza133, Gi-za134, Giza129 and Line 9). These results were agreed with those of Karim et al. (2010), Adawy et al. (2013) and Diab et al. (2013).

Polymorphism as detected by SSR analysis
Data in Table (7) were obtained from eight microsatellite primer pairs which were screened against 20 barley genotypes to detect polymorphic markers.
The eight SSR primers selected in this study generated a total of 40 major SSR alleles and the number of polymorphic alleles was 29, representing a level of polymorphism of 72.5% as presented in Fig. (3) and Table (7). The number of alleles per primer ranged from 2 in (Bmag0344a) to 8 in (GBM1215 and Bmag0496). The number of polymorphic alleles generated by individual primer pairs ranged from 1 in (Bmac0040 and Bmag0344a) to 8 in (Bmag0496). The average of the total alleles per primer was 5, while the average of polymorphic alleles per primer was 3.63.
The resolving power (RP) of the eight SSR primers ranged from 3.6 to 8.7. Similarly, polymorphic information content (PIC) values ranged from 0.05 to 0.40 demonstrating uniform polymorphism rate among all the eight SSR primers. Polymorphic information content (PIC) refers to the values of a marker for detecting polymorphism within a population or set of genotypes by taking into account not only the number of alleles that are expressed but also the relative frequencies of alleles per locus. As evident, SSR marker Bmag0496 showed the highest level of polymorphism with PIC value of 0.40, whereas the PIC values for the rest of SSR markers were in the range of 0.05-0.32 (Table 7). In this regard, Sipahi (2011) differentiated and identified 34 Turkish barley genotypes using barley SSR markers. Amplification of SSR loci was generated using 17 SSR primers. These SSR primers totally produced 67 alleles ranging from two to six alleles per locus with a mean value of 3.94 alleles per locus. Also, Khodayari et al. (2012) evaluated the genetic diversity of 32 individuals of tworowed and six-rowed Iranian landraces barley using 17 microsatellite markers. A high level of polymorphism information content (PIC; average = 0.651) and an average of 8.1 allele per locus were observed.
Regarding the SSR-based dendrogram according to (Nei & Li's Coefficient) the dendrogram showed two clusters at similarity percentage of 71.3% as shown in Fig. (4). In the first cluster, Line 38 genotype was separated in a single subcluster from all the other barley genotypes in this cluster. However, the second subcluster has 14 genotypes and was divided into two groups. The first group was separated into two subgroups, the first containing six Lines of ICARDA genotypes and the other containing four Egyptian genotypes. Also, the second group contained four of Egyptian barley genotypes. Meanwhile the second cluster which was divided into two subclusters at similarity percentage of 73.5% containing the genotypes Giza117, Giza124, Gi-za135, Giza136 and Line 9. Most of the resistant genotypes are located together such as (Line 91 and Line81) and (Line 26 and Line15). On the other hand; susceptible genotypes and moderately resistant or moderately susceptible such as (Giza 117, Giza124, Giza135, Giza136 and Line 9) also were located together. These results confirmed the conclusion mentioned in the performance of the genotypes tested and are in accordance with those reported by Abu Qamar et al. (2008) and Svobodova (2011).
Good results could be obtained if we crossed these twenty genotypes because there is a wide diversity among them. It is noteworthy that cluster analysis is a valuable tool for subdividing genotypes into groups including similar and dissimilar lines and has a great value from the breeder's point of view for initiating barley hybrid program. These findings are in line with those obtained earlier by Svobodova et al. (2011) and Maniruzzaman (2014).

Comparison of RAPD, SCoT and SSR data
From results presented in Tables (6  and 7), 20 barley genotypes were characterized by nine genotype-specific markers (four positive and five negative) as shown in Table (8). These marker loci were classified as five genotype-specific markers (two positive and three negative) by RAPD primers, two genotype-specific markers (one positive and one negative) by SCoT and SSR primers. Among the 20 genotypes of barley, six showed genotypespecific markers; Line 77 had the highest number of negative markers (three negative markers) using all types, while Line 9 had the highest number of positive markers (two positive markers). These results indicated that all types applied in this study succeeded in showing different molecular marker patterns which can be relied upon in distinguishing among the studied barley genotypes. Although, SCoT marker type had the highest percentage of polymorphism (77.42%), while RAPD primers were the best in terms of the average of resolving power RP (11.78) and the average number of genotype-specific markers / primer (1.0) as shown in Table  ( 9). These findings were in harmony with that illustrated previously by Fernández et al. (2002) in barley.

Phylogenic relationship among 20 barley genotypes as detected by genetic similarity (GS) and cluster analysis using RAPD, SCoT and SSR combined data
The similarity matrix resulting from the combined DNA markers RAPD, SCoT and SSR data were performed to generate correct relationships based on large and different genome regions as shown in Table ( From our results of SSR, SCoT and RAPD-based phylogenetic relationship study of the selected barley genotypes, it is evident that Egyptian barley genotypes are genetically very close and originated from closely related genotypes. This molecular evidence is confirmed based on data extracted from the historical genetic background of these genotypes. The majority of these genotypes have a common ancestor, at least for one of the parents. For instance, as previously reported by Afiah and Abdel-Hakim (1999) and Saker et al. (2005) the ancestors of Giza 124 are Bahteem and Giza 117, the ancestors of Giza 126 are Bahteem and SD 729 and the ancestors of Giza 123 are Giza 117 and FAO 86. It is also evident that RAPD analysis misplaced some of the ICARDA genotypes as well as revealed conflicting and unexpected genetic similarities among other genotypes. Similar observations were reported by Virk et al. (2000). In this context, Lombard (2001) concluded that molecular markers can be used as a tool and a convenient method for application of combine information from a large number of markers. Herein, we could conclude that it is possible to tag the breeding history and the origin of different barley genotypes using a combination of different molecular systems. Previously published data on barley by Russell (1997) and Sakar (2005) indicated that correlations between the relationships revealed by different polymorphism assays can vary widely both within and between species.
Results of this study are considered as the starting point needed to identify the valuable Egyptian barley net blotch resistance germplasm at both the phenotype and genotype levels and draw the attention of breeders and banks of natural plant genetic resources towards this valuable yet neglected germplasm. This is especially significant since molecular analysis combined with biological evaluation has proved to be a promising strategy in the selection of disease resistant germplasm (Haley, 1993).
On the basis of observed responses, it can be concluded that screened barley genotypes and groups contain a number of genes conferring resistance to P. teres. These genotypes could be incorporated into breeding program. The expression of such gene(s) is usually dependent on environment and barley genotypes containing such gene(s) are likely to vary in their response to net blotch under different environments. There is a need to establish molecular basis of the observed responses under field conditions. Multiple location studies using the same genotypes are also required to confirm the responses of the genotypes in other environments since environment was found to play a major role in the reaction in some number of screened genotypes.

SUMMARY
To evaluate the resistance of some barley genotypes for net blotch disease and grain yield and its related traits, twenty genotypes (12 local varieties and 8 exotic lines) of barley were used. Expression of severity to foliar infection varied between the evaluated genotypes, Giza 117 and Giza 2000 appeared the highest infection response, Giza 123, Giza 124, Giza 126 and Giza 131 were moderately susceptible, while the other genotypes ranged between resistant to moderately resistant. Line 81 and Line 91 proved to be most resistant genotypes for net blotch. Moreover, Giza 133 and line 91 showed superiority in grain yield values over all the tested barley genotypes and high resistance reaction for net blotch disease. Genetic variability and relationships among the used barley genotypes were evaluated by using five RAPD primers, three SCoT primers and eight SSR primer pairs. A high degree of polymorphism was detected with the three types of DNA markers which recorded 70.83, 77.42 and 72.5%, respectively. Alleles number ranged from 8 to 15, 9 to 12 and 2 to 8 per primer, with averages of 9.6, 10.33 and 5 per RAPD, SCoT and SSR primers, respectively. The highest percentage of genetic similarity as revealed by combined RAPD, SCoT and SSR data was found between line 81 and line 91 (90.7%), while the lowest similarity percentage was detected between Giza 124 and line 46 (65.2%). Giza 134 and Line 9 genotypes were resistant for net blotch disease while they gave positive genotype-specific markers with RAPD and SCoT analyses. Only Giza 123 genotype gave a positive genotype-specific marker using SSR analysis. Therefore, these genotype-specific markers could be considered as a molecular marker for net blotch disease response under similar conditions.