Linkage Mapping in a watermelon Population Segregating for Fusarium Wilt Resistance

Isozyme, randomly amplified polymorphid DNA (RAPD), and simple sequence repeats (SSR) markers were used to generate a linkage map in an F, and F3 waterjnelon [Citrullus funatus (Thumb.) Matsum. & Nakai] population derived from a cross between the fusarium wilt (Fcrsar& o.rysporum f. sp. nivemr) susceptible ‘New Hampshire Midget’ and resistant PI 296341-FR. A 112.9 CM RAPD-base? map consisting of 26 markers spanning two linkage groups was generated with FI data. With F,data,a 139 CM RAPDtbased map consisting of 13 markers covering five linkage groups was constructed. Isozyme and SSR markers were unli/lked. About 40% to 48% of the RAPD markers were significantly skewed from expected Mendelian segregation ratios in both generations. Bulked segregant analysis and single-factor analysis of variance were employed to identify RAP$ markers linked to fusarium wilt caused by races 1 and 2 of F. (Ixvs4orunl f. SD. niveunt. Current linkage estimates b4tween the resistance trait and the marker loci were too large for _ . effective use in a marker-assisted selection program. The diploid watermelon (Citrdlw lcmatus; 2n = 2x = 22; Shimotsuma, 1963) suffers from a number of serious fun lal. bacterial, and viral diseases that reduce yield and quality (Bru t”on . 1998; Nagel et al . , 1992). One of the most economically imdortant of these is fusarium wilt caused by the soilbome funkus Fusarium oxysporum f.sp. niveum (FON) (Martyn, 19196; Purseglove, 1987). FON has been separated into three pathogeinic races: 0, I, and 2 (Martyn, 1987; Netzer, 1976). Control of this disease relies primarily on use of resistant cultivars and c!op rotation. While many commercial cultivars have resistan& to races 0 and 1 of the pathogen. the more aggressive ract 2 overcomes all cultivars and has great potential for spread in watermelon production areas in the southeastern United St&es because it can be seedborne (Bruton. 1998; Hopkins et al., 19~92; Martyn and Netzer. 1991). Resistance to the various races has been identified in plant accessions from Africa (Dane et al., 19b8; Martyn, 1987). A single dominant gene, designated Fo-f, con$ers resistance to race I of FON in watermelon (Netzer and Weintlall. 1980), while resistance to race 2 in PI 296341-FR (Citr&rs lanat~s var. citroidees) is thought to be conferred by a recesslive gene with interactions with some minor genes (Martyn and Netzer. I99 1; Zhang and Rhodes, 1993). Variations among the cultivated watenneion are low with respect to isozymes and economically important characterisqics and this has hindered construction of a detailed genetic map (Hashizume et al., 1996; Zhang et al.. 1994). In a survey of26 Kecewcd for publmtion II August 2000. Accepted for pubhcatmn I7 Feb. 2t)Ol. The cost of publishing this paper was defrayed 111 part by the payment of age 4, charges. Under postal regulations. this paper therefore must be hereby mar ed dlw’lrseme,rf solciy to mhcate this tict. allozymic loci in 550 cultivated watermelon accessions, very little variation was found, but significant divergence was detected between cult ivated and wild Citrullus Schrad. sp. forms (Navot and Zamir, 1987). Navot et al . ( 1990) constructed an isozymebased map containing seven l inkage groups spanning 354 CM using a backcross population derived from C. lanatus x C. colocynthis (L.) Schrad. Hashizume et al. (1996) constructed a 524 CM l inkage map spanning 11 l inkage groups in a backcross population of a cult ivated Japanese C. Ianatus l ine and a wild African form. Other cucurbit genomes have been studied more extensively than the Citrdus genome. Genetic linkage maps have been created using molecular, isozymic. morphological, and disease resistance markers in intraspecific cucumber (Cucumis sativr~ L.) populations (Kennard et al . , i 994; Megl ic and Staub, 1996), and in intraand interspecific melon (Qrcumis melo L.) populations (Danin-Poleg et al., 1998; Oliver et al., 1998; Wang et al., 1997). Highlysaturatedgeneticmapsfacil i tate identif icationofgenes control l ing both qual i ta t ive and quant i ta t ive t rai ts of interest (Wang et al . , 1998). Molecular markers l inked to disease resistance would accelerate the t ime-honored, though t ime-consuming method of artificial inoculation, by easing the process of screening large numbers of individual plants to evaluate the introgression of resistance (Wechter et al . , 1995). Near-isogenic lines (NILS) and various pooling strategies have a great potential for rapid characterization of a trait of interest (Weising et al., 1995). Bulked segregant analysis (BSA) can be used to simulate a pairofNILs by pooling DNA from a population segregating for a specific trait (Wang et al., 1998). BSA has been used to identify markers l inked to genes for disease resistance e.g. fusarium wilt resistance in melon (Qrcumis melo, Wechter et al. , 1995; 1998). Therefore, the objectives of this research were to 1) identifyland characterize polymorphic molecular markers in an F1 population as well as F; l ines derived from a cross between the suscepdible ‘New Hampshire Midget’ (NHM) (C. 1nnaftl.r var. lanatrls)~and the resistant PI 29634 1 -FR (C. Irlnatusvar. citroides), 2)use these markers to construct a genetic l inkage map. and 3) examine the association between the inheritance ofspecific molecular marker alleles and resistance to race I and 2 of FON. Materials and Methods PLANT MATERIAL. The susceptible ‘New Hampshire Midget (female parent) and the resistant PI 296341-FR (Martyn land Netzer, 1991) were used as parents. Individual F, and F1 plbnts were advanced to produce F? l ines . FUSARIUMWILTSCREENINC ASSAY. American typeculturecollection (ATCC) strains of the pathogen Race I (ATCC 18467; Armstrong and Armstrong, 1978; Biles and Martyn, 1989) land Race2 (ATCC 62939; Martyn, 1987)) wereused to screenNHM, PI 296341-FR, F,, and 72 F, lines. Following DNA extracnon, eight, 2-week-old seedlings from each line per race were inoculated by root dip in an inoculum of 1 x 106 microconidia{mL (Martyn. 1987). The plants were evaluated under greenh use conditions without supplemental lighting. Individual plants % ere rated on a scale of I to 5 with 1 = healthy, no evidence ofwilt v e r a 3-week period, 2 = beginning signs of wilt, 3 = slightly wi 1 ted and stunted plants, 4 = wilted and stunted, and 5 = heavily wi ted and death of the seedlings (Dane et al . , 1998). Disease ratmgs were conducted over 3 weeks. Crown (stem-root junction) $ections of selected plants, both susceptible and resistant, were surface steri l ized and plated on quarter-strength potato dextrose agar to verify FON infection (Wechter et al. , 1995). ISOZYME .ANALYSIS. Cotyledonary tissue from four ofthe e i ght individuals from each of 72 F? lines ( 125 mgiline) was bulked and homogenized by hand in a pre-chilled mortar and pestle with 500 mL phosphate extraction buffer (HyPure, Isolab, Perkin Elmer, Branchburg, N.J.). Crude extracts were microcentrifuged fok 10 min at 14,OOOg, in a microcentrifuge (Eppendorf54 I 5C; Brinkman Instruments, Westbury. N.Y.). Supematant (8.5 mL per sam was electrophoresed on a Hypure agarose horizontal i soe leB le) tric focusing gel (FS-5080, pH 4-5) using the Multiphor Electrophoresis System (Amersham Pharmacia Biotech, Piscateway. NJ.) . Samples were assayed for acid phosphatase [ACP, Enzyme Commiss ion (EC) 3 I .3.2], aconitase (AK; EC 4.2.1.3), alcohol dehydrogenase (ADH; EC I. I. I. I), diaphorase (DIA; EC 1.6.99) esterase (EST; EC 3. I. I), glutamate-oxaloacetate transaminase (GOT; EC 2.6.1. I ), isocitrate dehydrogenase(IDH; EC 1.1.1.42), malate dehydrogenase (MDH; EC 1.1. I .37), malic enzyme (ME; EC 1. I. 1.40), peroxidase(PRX; EC 1.11.1.7). 6-phosphoglucanate dehydrogenase (6-PGD; EC I. I. I .49), phosphohexose isomeq’ase (PHI; EC 5.3.1.9), shikimic acid dehydrogenase (SKD; IEC I. I. 1.25) superoxide dismutase (SOD; EC I. 15. I. I), and triose phosphate isomerase (TPI; EC 5.3.1.1). Gels were stained as described by Wendel and Weeden ( 1989). DNA ISOLATION. DNA was extracted from 100 mg freeze-d ied leaf material from 98 F1 watermelon plants, and from 500 m r of 5 fresh cotyledon tissue from 2-week-old seedlings of PI 2963,4 IFR, NHM, F,, and four individuals from each of72 Fj lines using a modified CTAB extraction technique (Wagner et al., 1992) with additional purification steps (Kubisiak et al., 1997) or using~the Phytopure plant DNA extraction kit (Amersham Pharmacia Biotech.). BULKED SEGREGANT ANALYSIS. Disease resistance rating means were calculated for each Fi l ine for each race of the pathogen. Data were subjected to analysis ofvariance (ANOVA) using Proc GLM of SAS and means were separated by Duncan’s mult iple range test (SAS Inst., Inc., 1996). Two bulked DNA samples were prepared for each race ofthe pathogen. one resistant bulk(R) and one susceptible bulk (S). Bulked samples consisted of equal volumes ofstandardized DNA (25 ng.mL--I) from IO resistant and IO susceptible l ines, respectively. RAPD ,ANALYSIS. To identify segregating polymorphisms, one hundred fourteen lo-mer primers from Operon Technologies (Alameda, Calif.) and the University of British Columbia (Vancouver. Brit ish Columbia, Canada), along with five 12-mer primers identified by Hashizume et al., ( 1996) were screened on a panel of DNAs consisting of NHM, PI 29634 1 -FR, F,, and 5 F1 progenies. Primers that amplif ied bands that were polymorphic between NHM and PI 29634 1 -FR, present in the F,, and present in at least one F: progeny were classif ied as potential ly useful markers. In the Fi populat ion. three hundred twenty nine IO-mer primers from Operon Technologies and from the University ofBritish Columbia were screened for their ability to detect polymorphisms between NHM and PI 29634 I -FR as wel l as the R and S bulks . Potential ly useful polymorphisms were further characterized on each ofthe individuals from the F1 lines composing the respective bu lks . Polymerase chain reaction (PCR) was performed in a 25 l.tL volume containing 10 mM Tris-HCl, 3.0 mM MgC&, 10 mM KC1 (pH 8.3), 0.2 mM of each nucleotide (Perkin Elmer), 5 pmol IOmer or 12-mer oligonucleotide primers, IO ng genomic DNA and 1 unit offfnzplitnypolymerase (Perkin Elmer). The thermocyclers were programmed as follows: initial denaturation at 94 “C for IO min. 35 cycles of 94 “C for 45 s. 36 “C for 45 s, and 72 “C for 2 min, followed by a final extension at 72 “C for 5 min. Amplification products were electrophoresed on 1.2% agarose gels and detected by staining with ethidium bromide. The gels were photographed under ultraviolet light with Polaroid film 667 (Polaroid Corporation, Cambridge, M.A.). Lambda (h) DNA digested with PstI was used as a molecularsizemarker. Each band was named by the primer used and its size in basepairs (bp): OPF 16,,(,(, was the 1600 bp band amplified by Operon primer F 16. SSR .ANALYSIS. A total of 107 simple sequence repeats (SSR) primers from Jarret et al. (1997. and unpublished results) and Katzir et al . (1996) were screened against the parents and the F, to detect polymorphisms. Analysis was petformed on four indi viduals bulked from each of 72 F, lines. PCR was performed in a 20 yL volume containing 10 rnbi Tris-HCI, 3.0 mM MgCl*, 10 mM KCI (pH 8.3) 0.2 mM ofeach nucleotide (Perkin Elmer), I ,UM of each primer, I .6% bovine serum albumin, 50 ng genomic DNA and I unit ofAmplify polymerase (Perkin Elmer). Thermocyclers were programmed as fol lows: ini t ial denaturat ion at 94 “C for 5 min, 35 cycles of 94 “C for 1 min, 55 “C for I min and 72 “C for 2 min, followed by a final extension at 72 “C for 2 min. The amplified products were separated on a 0.4 mm thick, 6% denaturing polyacrylamide gel. For molecular weight size determinations, 25 bp and/or 100 bp DNA ladders (GibcoBRL) were used. The gel was silver-stained using the method of Bassam et al. (1991). LINKAGE A NALYSIS. Each RAPD band was tested for goodness of fit to the expected 3: I segregation ratio in the Fz or 5:3 ratio in the F, by chi-square analysis (P < 0.01). Those markers experiencing segregation distort ion were excluded from linkage analyTable I, Mean fusarium wilt race I disease resistance ratings in F, lines derived from a cross between fusarium wilt-susceptible ‘New Hampshire Midget’ x resistant PI 296341 -FR.