Development of SCAR Marker Related to Summer Stress Tolerance in Tall Fescue ( Festuca arundinacea )

Summer stress tolerance (SST) is one of the most important breeding objectives in tall fescue (Festuca arundinacea), an important perennial cool-season grass. However, breeding for better SST is generally complicated by the many environmental factors involved during the growing season. Utilizing the bulked segregant analysis (BSA), we were able to identify one marker related to SST from 100 inter-simple sequence repeat (ISSR) markers and 800 random amplified polymorphic DNA (RAPD) markers, and successfully developed a dominant sequence characterized amplified region (SCAR) marker T_SC856 from the UBC856 sequence. Furthermore, the SCAR marker was tested in different clones of new populations, which were identified under complex summer stress (high temperature and humidity, Pythium blight, and brown patch), and it exhibited relatively high consistency (77%) with the phenotype. We believe that with more markers obtained in the future, better efficiency is likely to be achieved in breeding for improved SST in tall fescue and possibly other species as well. Further studies that analyze the factors relating to the SCAR marker are needed.


Introduction
Tall fescue (Festuca arundinacea) is a common perennial cool-season grass (Poaceae) spread throughout the temperate regions of the world.It is widely used in parks, home lawns, athletic fields, golf courses and soil conservation sites.The optimum temperature for the growth of tall fescus is 16 °C～24 °C (Emmons, 1994).When grown and managed in temperate and transition climatic zones, quality of tall fescue decreases often suffers greatly during summer months as evidenced by stunt growth, severe disease and insect pressure, and dead patches.The comprehensive resistance exhibited by tall fescue against multiple summer abiotic and biotic stresses is referred as summer stress tolerance (SST).Tall fescue genotypes with better SST have the potential to maintain better overall quality through summer time, especially in the transition zone.Therefore the improvement of tall fescue SST resistance is an important target in tall fescue breeding efforts.
Many previous studies evaluating summer performances of cool-season grasses haven focused on the impact of a single specific stress, such as heat (Du et al., 2011;He and Huang, 2007;Kim et al., 2010;Lefsrud et al., 2010;Xu and Huang, 2008;Xu et al., 2006;Xu et al., 2010;Xu et al., 2011), drought (Abraham et al., 2004;Merewitz et al., 2010), brown patch (Rhizoctonia ssp.) (Bonos et al., 2006;Dong et al., 2008;Watkins and Meyer, 2004), and Pythium blight (Pythium ssp.) (Allen et al., 2005), etc.Among these commonly studies stresses, heat and drought are the two major ones limiting the summer growth and quality of coolseason turfgrasses.And reports have been made on the combined the effects of both heat and drought, as well as their impacts on the summer health of cool-season grass species (Abraham et al., 2008;Su et al., 2007;Wang and Huang, 2004).In a study done in New Jersey, USA, Bonos and Murphy (1999) evaluated growth and performance of Kentucky bluegrass (KBG, Poa pratensis) genotypes under summer stresses (heat and drought) and identified genotypes that were relatively more tolerant to these stresses.The growth of KBG is also reported to be limited by drought and heat stress during summer in Northwestern China (Liu et al., 2008).However, there are limited reports on the impacts of combined stress from heat, high humidity, diseases and insects, as commonly encountered in field condition during summer months.
Commercial tall fescue is allohexaploid (2n=6x=42) with a large genome (5.27-5.83×10 9bp) (Seal, 1983).Due to a cross-pollinated habit, each seed or plant of the grass is genetically unique.Therefore, intraspecific genetic variation in tall fescue is commonly observed for individual selection breeding (i.e., recurrent selection and ecotype breeding).However, due to the heterogeneity of soil physiochemical characteristics and the variation in spreading rate of fungal pathogen in field, the selection efficiency of conventional breeding programs targeting at summer stress tolerance is relatively low.
Along with the development of molecular technology, transgenic tall fescue lines have been reported with improved stress resistance.Wu et al. (2006) obtained transgenic tall fescue plants with enhanced salt and osmosis tolerances, containing CBF1 gene from Arabidopsis thaliana via Agrobacterium tumefaciens -mediated transformation.Dong et al. (2007) obtained transgenic plants exhibiting resistance to two major fungal diseases greay leaf spot (Magnaporthe grisea) and brown patch (Rhizoctonia solani).Meanwhile, molecular marker techniques were also applied in tall fescue researches, such as linkage maps construction, genetic diversity analysis, cultivar identification, and so on.The first tall fescue linkage map was reported by Xu et al. (1995), generated from the F2 population of HD28-56 and 'Kentucky-31'.It consisted of 108 restriction fragment length polymorphisms (RFLPs) and covered 1274 cM on 19 linkage groups with a mean marker interval of 17.9 cM.Ten years later, an amplified fragment length polymorphism (AFLP) and simple sequence repeat (SSR) based genetic linkage map of tall fescue was constructed by Saha et al. (2005), who combined the female (HD28-56) map and the male (R43-64) map into an integrated map, which covered 1841 cM on 17 LGs with an average marker interval of 2.0 cM.Furthermore, molecular mapping of locating quantitative trait loci (QTLs) related to forage digestibility in tall fescue was also reported by Saha et al. (2009).Base on RFLP data, the results of genetic diversity showed that variation within tall fescue cultivars was high (Xu et al., 1994).Analyzed by SSR, random amplified polymorphic DNA (RAPD), and inter-simple sequence repeat (ISSR) markers, Mester et al. (1999) classified 30 somaclones into four distinct groups.Using 461 AFLPs from six primer combination, Mian et al. (2002) clustered 16 tall fescue plants in groups.In addition, identification of 12 commercial varieties of tall fescue was successful via RFLPs (Busti et al., 2004), in which seven specific bands were distinguished.
As an effective breeding method, marker-assisted selection (MAS) has been applied in many species successfully.It is especially useful for some complex traits.Unfortunately, there are no reports on MAS of SST in tall fescue.Considering the hereditary characteristics of tall fescue, this study was designed to identify molecular markers (ISSR and RAPD) that are linked to SST trait via pseudo-bulked segregant analysis (BSA) method, using tall fescue clones with varying SST.Also, a sequence characterized amplified region (SCAR) marker developed from polymorphic random molecular marker was tested in other clones.Our purpose of this study is to seek and offer sequence specific markers that could be valuable in assisting breeding for summer stress tolerance in tall fescue, and hopefully render future breeding efforts of this kind more simplicity and reliability.

Plant materials
Selection population: In the autumn of 2007, 130 tall fescue clone lines (genotypes) from different varieties and lines (Tab. 1) were planted in Funing County, Jiangsu Province, China.Each clone line included twenty-two plants (4-5 tillers per plant).Funing County is located at 33'N latitude and 119'E longitude, and it has four distinct seasons featuring high temperature and plenty of rainfall in summer.Soil type of this area is sandy loam.Till the end of the experiment in 2009, we only fertilized twice (late autumn: 88 kg N•ha -1 ; early spring: 34 kg N•ha -1 ).No pesticide was used for insects and diseases control and weeds were handpicked twice a year in late autumn and early June.Technology and Service Co. Ltd, Shanghai.PCR reactions for ISSR and RAPD markers were carried out in 10-μl mixture [20 ng genomic DNA, 10 pmol primer, 100 μmol•L -1 dNTPs, 1×Taq Buffer, 1.5mmol•L -1 MgCl2, and 0.5 unit of Taq DNA polymerase (TaKaRa, Talien, China)].The reaction conditions were as follows: 94 °C for 3min, followed by 35 cycles at 94 °C for 10s，an annealing temperature depending different primers for 40s，and 72 °C for 45s and a nal extension at 72 °C for 6 min.The amplification products were separated on 1.5% agarose gels, and then soaked in a 1.0 μg•mL -1 of ethidium bromide water solution for 10 min.Then, the gels were illuminated by UV light, and photographed with a Tanon3500 imaging system (Tanon Science & Technology Co., Ltd., Shanghai, China).
Cloning and sequencing of the ISSR fragment: Sequence-related amplified polymorphism fragment was retrieved with the Gel Extraction Kit (DV805A, TaKaRa, Talien), and the quality of fragment extracted was checked by 0.8% agarose gels.The target band was cloned into the PMD18-T vector (D101A, TaKaRa, Talien).Positive colonies bearing DNA of the expected size were sequenced with ABI 3700 Sequencer (Sangon, China).
Conversion of ISSR marker to SCAR (sequence characterized amplified region) marker: According to the sequence of the fragment, SCAR primers were then designed using the Primer Premier5.0.The SCAR primers were tested in two pools and 22 clones.The condition of the SCAR amplification was as follows: 94 °C for 3 min, followed by 29 cycles at 94 °C for 10s，63 °C for 40s，and 72 °C for 30s and a nal extension at 72 °C for 6 min.We obtained one polymorphic SCAR marker, named as T_SC856 (Forward primer sequence: 5'ACACACACACACACACCAATTG3'; Reverse primer.sequence:5'ACACACACACACACACTACCTC3').

Identification of summer stress tolerant clones in selection population
Throughout the whole summer season, the 130 clone lines of tall fescue planted exhibited very distinct summer hardiness.Some clones (i.e., TF18) almost lost all above ground tissues (Fig. 1a), while some other clones (i.e., TF02) were still growing well, with abundant green and dense leaves (Fig. 1b).Obviously, the former type was summer stress sensitive clones, and the latter was summer stress tolerant ones.Careful examination of the dead plants revealed large populations of millipedes (Spirobolus bungii) nearby.
Finally, eleven summer stress tolerant clone lines (TF01-TF11) and twenty-five summer stress sensitive clone lines (TF12-TF28, TF31-TF38) were selected from the selection population (Tab.1).All the summer stress tolerant ones and 11 of the summer stress sensitive ones (TF12-TF22) were selected randomly to construct the DNA bulks.
Testing population: In 2005, plants of tall fescue variety 'Shanghai Evergreen Dwarf' were grown in Sanjiagang area, Shanghai, China.The seeding rate was 20 g•m -2 .Clone lines were constructed from fifty plants (including summer stress tolerant and sensitive plants) selected under summer stress (mainly insect and disease pressure), and transplanted to research farm at School of Agriculture and Biology, Shanghai Jiao Tong University, Shanghai, China.Subsequently, twelve clones performed well under rainy and high temperature conditions and thirteen other clones performed poorly under same conditions were selected in autumn of 2008 and 2009.All these plants were used as testing population.
Shanghai is located at 31'N latitude and 121'E longitude.The climate here is similar to Funing County, with an average temperature of 27.9 °C in July (recorded highest temperature of 40.7 °C) and an annual precipitation of 1087 mm (He et al., 1997).The Soil type at testing location is loam.Management program was similar to that of selection population.
Identification of summer stress tolerance: We identified SST from clone lines of selection population and clones of testing population.The identification was operated after the full duration of summer stress and before recovery started to occur.Based on the status of plants (survival rate, green leaf index, degree of leaf curling, disease incidence, and so on), three breeders marked each clone by visual inspection independently.Scoring is based on a scale of 1 to 9, with 9 being the best growing state (He et al., 2001).Summer stress tolerant and summer stress sensitive clone lines (Tab. 1) were selected, respectively, for future work.Some of the summer stress sensitive clones lost green shoots due to summer stress, but they were able to recover from crown in the next year.
DNA extraction and pseudo-BSA: The same amounts of leaves were collected from each plant per clone line, and genomic DNA mixture of each clone lines was extracted with the CTAB method (Clark 1998).Extracted DNA samples were visualized after electrophoresis on 0.8% agarose gels in 1×TAE.DNA concentration and purity was measured with a UV spectrophotometer.The DNA was adjusted to a final concentration of 20 ng•μL -1 with TE buffer (pH 8.0) and stored at -20 °C until use.
Equal amounts of DNA from eleven summer stress tolerant clone lines and eleven summer stress sensitive clone lines were pooled to construct two DNA bulks for BSA (Michelmore et al., 1991), summer stress tolerant pool (STP) and summer stress sensitive pool (SSP).Obviously, the polymorphism between these two DNA pools was affected by multiple factors rather than only one.To indicate the difference with the traditional BSA approach, we refer our approach as pseudo-BSA.
RAPD and ISSR analyses: A total of 100 ISSR primers and 800 RAPD primers were used to screen polymorphisms in the two bulks.The primers that were able to amplify polymorphic bands between the two pools were tested in the 22 individual clone lines that made up the pools for further check of the polymorphism.RAPDs and ISSRs were all according to sequences of NAPS Unit standard primers (University of British Columbia, Canada).All the primers were synthesized by Sangon Biological Engineering Identification of markers related to the summer stress tolerance trait Of the 100 ISSR primers tested, 13 primers were polymorphic between the two bulks (Fig. 2).Unfortunately, none of the 800 RAPD primers showed polymorphism.From the successful reaction system, about 5.5 marker bands were amplified clearly per primer.Upon further testing with 22 clone lines DNA that made up the pools, only one ISSR marker (UBC856) was able to generate the same polymorphic bands in the clone lines and two bulks (Fig. 3a).It was a dominant marker appearing brightly in SSP, and the size of the band was about 1300bp.

Conversion of ISSR marker to SCAR marker
The ISSR marker related to summer stress tolerance were converted into SCAR marker (T_SC856).The polymorphic product was 411bp, appeared brightly in SSP and almost none in STP (Fig. 3b).Among the 22 individuals of the two pools, the polymorphic bands amplified by the SCAR marker were identical to those of its corresponding ISSR marker.In the future, this anchor marker could be used for large-scale screens for summer stress tolerance of MAS.
Anchor marker verified in the testing population Summer stress tolerant clones and summer stress sensitive clones were selected from "Shanghai Evergreen Dwarf" testing population.Combined with the clone lines (TF23-TF28, TF31-TF38) identified in selection population (not include in the bulk construction), they were all tested by the SCAR maker.According to the comparison between polymorphism of SCAR amplification and phenotype of SST in the field, 4 of 13 WT clones (WT3, WT5, WT6, and WT11) were not consistent (Fig. 4b); 2 of 12 ST clones (ST08 and ST12) were not consistent (Fig. 4a).Besides, 3 of 14 clone lines (TF31, TF35, and TF36) were inconsistent (Fig. 4c).Taken together, accuracy of the molecular marker selection reached 77%.

Discussion
It generally takes turfgrass breeders more than ten years to breed a new variety, from single plant selection and cloning to quality evaluation via field tests.Through efforts in conventional phenotypic selection, we have indeed obtained some summer stress tolerant varieties and lines of tall fescue in Shanghai, China (He et al., 2001;He et al., 2002).However, released varieties would be usually replaced by more recent ones after five years or so.
Combining phenotyping with the molecular marker polymorphism selection could certainly accelerate the breeding process, especially for some complex traits.In order to do so, abundant molecular markers closely linked to target traits have to be available to breeders first.Several applications of turfgrass SCAR markers have been reported.For example, SCAR markers created by RAPD bands were used to differentiate some species of bentgrass, and identify progenies derived from hybridization among bentgrass species (Scheef et al., 2003).
Identification and detection of Phoma sclerotioides was operated by sequence-characterized DNA markers in the brown root rot of alfalfa (Larsen et al., 2002).Humaid et al. (2004) detected the genetic variation and Fusarium resistance in seven tufgrass genotypes with RAPD and SCAR markers.Our work is the first report of the development of SCAR marker relating to summer stress tolerance trait.And it showed the relatively high accuracy of the molecular marker selection in different clones.It is undoubtedly a worthwhile effort using MAS for summer stress tolerance in tall fescue and could be potentially extended to other species.
With continued trend in global warming, grass breeders have been paying more attention to SST in tall fescue.Of course, this is not an easy breeding target to work with.SST in turfgrass species is a very practical trait, but is also very complex indeed.Each involving factor (i.e.heat, disease, insect, drought, water-logging tolerance, etc.) in SST is a complex trait itself controlled by multiple genes.For example, several QTLs related to these factors have been located in diploid ryegrass, which is a main research object in grasses.Pfender et al. (2011) detected three stem rust resistance QTLs by SSR, sequence-tagged site (STS), and restriction-site associated DNA (RAD).The most prominent QTL accounted for more than 30% of the phenotypic variance.Pearson et al. (2011) identified 37 QTLs for morphological traits influencing water logging tolerance.Certainly, heat tolerance is one of the most important SST components, and some studies on heat tolerance genes have been reported.In C3 Agrostis grass species, Xu et al. (2007) identified and characterized an expansin gene AsEXP1 associated with heat tolerance.Kim et al. (2010) overexpressed an Arabidopsis 2-Cys Prx in transgenic tall fescue plants to enhance heat tolerance.
Factors involved with the combined summer stress could offer different levels of impact depending on the specific climate of a given region or a specific year.In a wet years or regions with ample summer precipitation, summer stress would be dominated by high temperature, high humidity, and brown patch/Pythium blight promoted by these environmental condition.In a dry years and regions with less scarce rainfall, the main stress would be heat and drought.
In this study, the polymorphism of SCAR markers corresponds well with the phenotypic identification in the field, between Funing County and Shanghai.As a result of the environmental factors analysis of the two locations in these years, they were both characterized by high temperature, ample rainfall, and Pythium blight promoting in summer.Later, we identified the polymorphism of the SCAR in another tall fescue population in 2011.However, the result was not ideal (data not shown).The analysis of 2011 found lower temperature, less rainfall, and better field management than those in 2008 and 2009.Under the summer stress of 2011, there were only some plants wilting in the afternoon, and of course, resulted in no brown shoots.It was determined that there were no significant correlations (p＞0.05) between T_SC856 and wilting trait.Instead, the SCAR marker maybe linked to heat, high humidity or Pythium blight tolerance.Yet, it is hard to single out which factor relates to the SCAR marker developed indeed.Further studies that specifically look at heat, high humidity tolerance, etc., under carefully controlled environment are needed to verify the factor.
Although this study reported a SST related SCAR marker with an accuracy closed to 80%, it is impractical to base MAS for SST on this one marker alone.Phenotype, affected by many uncertain factors, is unlikely to express consistently if plant materials were selected using one molecular marker, even though they are linked tightly.As a result, if locations and years change, MAS for SST based on one molecular maker is likely to fail.Currently, we are testing the SCAR marker in more clones from different locations and years.Next, many more types of molecular markers related to SST trait need to be developed.A combination use of two or more molecular markers for summer stress tolerance selection could be more accurate and stable.Development of multiple markers linked to SST will certainly be good news for MAS breeding in tall fescue and other species as well.

Fig. 1 .
Fig. 1.Identification of tall fescue clone lines in Funing County, Jiangsu, China.Summer stress sensitive clone line TF18 (a) and summer stress tolerant clone line TF02 (b)