Plant Soil (2010) 335:423–437 DOI 10.1007/s11104-010-0431-1
REGULAR ARTICLE
Inheritance of resistance to iron deficiency and identification of AFLP markers associated with the resistance in mungbean (Vigna radiata (L.) Wilczek) Peerasak Srinives & Ratanakorn Kitsanachandee & Thitaporn Chalee & Warunee Sommanas & Sontichai Chanprame
Received: 5 January 2010 / Accepted: 11 May 2010 / Published online: 29 May 2010 # Springer Science+Business Media B.V. 2010
Abstract Iron deficiency chlorosis (IDC) is a major problem reducing yield of mungbean in many countries. In this study, we crossed “KPS1”, the most popular Thai mungbean cultivar susceptible to IDC with “NM10-12”, a mungbean line from Pakistan resistant to IDC. Segregation analysis of the F2 population revealed that the resistance is controlled by a major gene (IR) with dominant effect. Two AFLP markers, E-ACT/M-CTA and E-ACC/M-CTG were identified closely linking with the IR gene. The frequencies of these markers were assessed in 241 mungbean accessions from several countries. The accessions could be divided, in relative to total chlorophyll content of the resistant check (NM10-
Responsible Editor: Jian Feng Ma. P. Srinives : W. Sommanas : S. Chanprame Department of Agronomy, Faculty of Agriculture at Kamphaeng Saen, Kasetsart University, Kamphaeng Saen, Nakhon Pathom 73140, Thailand P. Srinives (*) : R. Kitsanachandee : T. Chalee : S. Chanprame Center for Agricultural Biotechnology, Kasetsart University, Kamphaeng Saen, Nakhon Pathom 73140, Thailand e-mail:
[email protected] R. Kitsanachandee : T. Chalee Center for Agricultural Biotechnology: (AG-BIO/PERDO-CHE), Bangkok, Thailand
12) and the susceptible check (KPS1), into the resistant group with 125 accessions and the susceptible group with 116 accessions. Among 125 resistant accessions, E-ACT/M-CTA and E-ACC/M-CTG were present in 119 (95%) and 109 (87%) accessions, respectively. Both markers can identify all resistant accessions from England, Indonesia and Pakistan, but only E-ACT/M-CTA linked to all resistant accessions from Australia, India, Iraq, Taiwan and Thailand. Understanding the inheritance and identifying molecular markers linking to the IR gene can help plant breeders to improve this crop for growing in irondeficient soils. Keywords Mungbean . Vigna radiata . Iron-deficiency chlorosis . Chlorophyll content . AFLP markers . Alkaline soil Abbreviations IDC iron-deficiency chlorosis IR gene controlling resistance to IDC Chl a chlorophyll a Chl b chlorophyll b Total chl total chlorophyll
Introduction Mungbean (Vigna radiata (L.) Wilczek) is high protein legume grown widely in South and Southeast
424
Asian countries. The most popular foods made of mungbean seed include dahl, bean sprouts and many kinds of sweet. Mungbean is a self-pollinated crop and thus one accession is theoretically composed of plants of the same homozygous genotype. Irondeficiency chlorosis (IDC) is a major factor decreasing yield of mungbean grown in the upper central and lower northern regions of Thailand. These 2 regions are major mungbean production areas where the high pH soil scatters in several hundred thousand hectares. Ohwaki et al. (1996 and 1997) and Ohwaki and Sugahara (1997) and Oonkasem and Thavarasook (1991) reported that the two most popular mungbean cultivars in Thailand, “Kamphaeng Saen 1” (KPS1) and “Kamphaeng Saen 2” (KPS2) grown on alkaline soils showed iron-deficiency (chlorotic symptoms) and died prematurely or gave very low yield. This is because iron is a major constituent of ferredoxin, which is required for formation of chlorophyll in plant cells and activation of respiration and photosynthesis. However, soil heterogeneity and heavy rains can hinder clear expression of this character. Rain water replacing air in soil pores creates reduction state in the soil and induces temporary iron availability to the mungbean. This is a major obstacle for field screening of germplasm and selecting segregating progenies. The use of DNA markers to screen and select in the laboratory can reduce time and money required for field growing and can prevent false identification of the susceptible genotypes. So far, there has been no report on inheritance of resistance to IDC in mungbean while there were only a few studies available in the other field crops. In dry bean, iron deficiency resistance is controlled by two complementary dominant genes (Zaiter et al. 1987). In soybean, Weiss (1943) reported that reaction to IDC was controlled by a major gene with a recessive allele controlling the inefficiency. Later, Cianzio and Fehr (1980) and Cianzio (1999) reported that this trait is also conditioned by several modifying genes. In chickpea, resistance to IDC is controlled by a single dominant gene (Saxena et al. 1990). Resistance to IDC in oats was reportedly controlled by a major dominant gene with modifiers (McDaniel and Brown, 1982). Dansgan et al. (2002) reported that iron deficiency resistance in tomato is polygenically inherited with a relatively high additive effect. Evaluation of IDC symptoms by spectrometric chlorophyll determination is more sensitive than by
Plant Soil (2010) 335:423–437
visual scoring. Cianzio et al. (1979) showed a negative correlation between chlorophyll content and visual score for IDC in soybean. Lin et al. (1997) reported that visual score data of two populations in 2 years were statistically the same while chlorophyll concentration data were different, due to chlorophyll concentration was more sensitive to environment in different years than visual scoring. However, both methods can complement each other, as Samdur et al. (2000) used both chlorophyll content and visual scoring to classify the degree of resistance to iron deficiency in field-grown groundnuts. On the other hand, Azia and Stewart (2001) investigated the relationship between extractable chlorophyll and SPAD value in muskmelon leaves and found that SPAD reading was significantly related to extracted chlorophyll, both on fresh weight and leaf area bases. However, our earlier study showed that SPAD index in a large collection of field-grown mungbeans was rather sensitive to environment (data not shown), since the leaf surface of same accessions deformed from pest infestation and thus affected the reading. Chlorophyll content, although more tedious to measure, gave a better relative IDC among accessions. Chlorophyll content has been used to investigate the relationship between iron concentration change in the soil and IDC symptoms in the plants. Jacobson and Oertli (1995) found that chlorosis correlated with iron concentration and chlorophyll contents in sunflower leaves. Change in iron concentration can be examined physiologically and morphologically in deciduous fruit tree plants (Cinelli et al. 2003), tomato (Dansgan et al. 2004), and chickpea (Ohwaki and Sugahara, 1997). Physiological root response to IDC was reported in tomato. The resistance “Roza” line had higher number of lateral roots, Fe+3 reduction capacity, uptake rate and translocation of iron, and thicker root tips than the susceptible 227/1 line. The resistant line also had the ability to reduce pH of the nutrient solution from 6.0 to 4.6–5.3, making the iron more available in the Fe+2 form (Dansgan et al. 2004). Ohwaki and Sugahara (1997) reported in resistant chickpea K-850 that the period of acidification induced by iron deficiency and the rate of extrusion of protons were much higher than the rate of exudation of carboxylic acids from the roots. Molecular markers in mungbean have been poorly developed so far. Two restriction fragment length
Plant Soil (2010) 335:423–437
polymorphism (RFLP) maps were constructed by Young et al. (1992, 1993) and Chaitieng et al. (2002), but have never been used by plant breeders, due mainly to high amount of DNA required and availability of the primers. A few SSR markers have been developed into a partial linkage map with rather limited polymorphism among the mungbean germplasm (Sangsiri et al. 2007, Somta et al. 2009a, b). Thus this experiment was designed to use AFLP markers to identify the resistance to IDC. AFLP method developed by Vos et al. (1995) is a highly reproducible marker system (Vuylsteke et al. 1999). The evidence on usefulness of simplified AFLP systems based on EcoRI and MseI was further given in self-pollinated and cross-pollinated plant species (Ranamukhaarachchi et al. 2000). Powell et al. (1996) and Pejic et al. (1998) convinced that AFLP is 10-fold higher efficiency than RFLP, RAPD and SSR. Sommanas (2000) visually scored on 199 recombinant inbred lines from the cross between KPS1 and NM10-12 which are susceptible and resistant to IDC, respectively. She found two AFLP makers, E-CAG/M-TAC and E-CGT/M-CTG locating at 2.9 and 3.0 cM away from the gene controlling the resistance. The objectives of this study were (1) to examine the inheritance of resistance to IDC, (2) to identify AFLP markers associating with gene(s) controlling the trait in an F2 mungbean population, and (3) to detect the presence/absence of the markers in 241 diverse mungbean assessions.
Materials and methods Plant materials A cross was made between a mungbean cultivar “KPS1” and a line “NM10-12”. KPS1 is the most widely-grown cultivar in Thailand but highly susceptible to IDC when cultivated in alkaline soils. NM1012 is an introduced line from Pakistan and highly resistant to IDC. KPS1, the female parent has white hypocotyl while NM10-12, the male parent has purple hypocotyl, which is dominant to the white one. The F1 plant showed purple hypocotyl, proving that it is a true hybrid. Seeds set on an F1 plant were sown to produce a population of 160 F2 plants used in this study. In addition, a diverse set of 241 mungbean
425
accessions from Afghanistan, Australia, China, England, Guam, Guatemala, India, Indonesia, Iran, Iraq, Korea, Nigeria, Pakistan, the Philippines, Taiwan, Thailand and USA was obtained from AVRDC–The World Vegetable Center, Taiwan. The population and accessions were tested in an alkaline soil field with irondeficiency, as well as in the Half-Hoagland nutrient solution to observe their chlorotic reaction. Field testing Field testing was conducted at a selected field of Nakhon Sawan Field Crops Research Center, the Department of Agriculture, Thailand where alkaline soil is uniformly prevailing. F3 seeds from each F2 plants were sown in 2 replicates while the accessions were sown in a separated experiment without replication. Each entry was hill-planted in rows of 1 m long, one accession per row. In both experiments, NM10-12 and KPS1 were planted alternately as check genotypes every 10 rows. The seeds were sown at spacings of 12.5 cm between plants and 50 cm between rows. Chlorosis symptom in each entry was evaluated at 21 days after planting by visual scoring, SPAD index and chlorophyll determination. Visual observation was made using the scale of 1 for no yellowing, 2 for slight yellowing, 3 for moderate yellowing, 4 for intense yellowing, and 5 for severe yellowing with some necrosis (Cianzio et al. 1979). A SPAD index of each plant was determined using a chlorophyll meter (SPAD-502; MINOLTA®, Tokyo), measured on 3–5 spots on the middle leaf of the second full-grown trifoliate leaf from the top of the plants. The average SPAD reading of a plant was used to characterize it, and the average of three plants (accessions or F2 line) per genotype was used for providing the score of IDC symptoms. Since each mungbean accession is a pure line developed from a series of natural selfing, leaves sampled from three healthy plants are sufficient to represent the accession. Chlorophyll content was determined from leaves of the sampled individuals following the method of Moran (1982). Briefly, the middle leaflet of the second full-grown trifoliate leaf was detached to prepare for leaf sections. If the leaflet was abnormal or damaged, one of the 2 side leaflets of that trifoliate leaf was used instead. Chlorophyll was extracted from three leaf sections of 0.198 cm2 each, using 4 ml of DMF (N, N-dimethyformamide)
426
for 48 h at 4°C in a dark room. The solution was measured with spectrophotometer at the wave length
Plant Soil (2010) 335:423–437
of 664 and 647 nm. The chlorophyll contents (g/cm2) were calculated from the formulas:
Chlorophyll a content ¼ ð2:99A647 þ 12:64A664 ÞVol=ðX» Area» 100Þ Chlorophyll b content ¼ ð23:26A647 5:60A664 ÞVol=ðX» Area» 100Þ Total chlorophyll content ¼ ð20:27A647 þ 7:04A664 ÞVol=ðX» Area» 100Þ
Hydroponic solution testing The hydroponic solution was formulated following Hoagland and Arnon (1950). The composition in the solution was 3 mM KNO3, 2 mM Ca (NO3)2, 0.05 mM KH2PO4,1 mM MgSO4, 1.5 µM H3BO3, 0.25 µM MnSO4.H2O, 0.1 µM CuSO4.5K2O, 0.2 µM ZnSO4.7H2O, 0.025 µM Na2MoO4, 2 µM Fe-EDTA and 5 g/l CaCO3, then adjusted to pH 9 by NaOH. Due to limited number of F3 seeds, only 115 F2 plants were tested in the solution. F3 seeds from each F2 plants and their parents were sown in a sand box until the first true leaf stage (∼7 days after germination). Three seedlings from each F2 plant were transferred into a 1 m x 1 m hydroponic tank filled at 0.1 m high with the Half-Hoagland nutrient solution and treated as one replicate. Chlorosis of the apical leaves was evaluated by visual scoring, the chlorophyll meter and chlorophyll content at 14 days after transferring (i.e. 21 days old). One run of the nutrient solution experiment was conducted, with two replications in a greenhouse of Kasetsart University, Kamphaeng Saen, Thailand. The temperature during the experiment varied between 30–36°C with the relative humidity 60–90%. The average scores from 3 seedlings were used to represent the chlorosis score of each F2 plants and the parents in each replicate.
respectively 10 resistant and 10 susceptible F2 plants. These two bulks were screened by AFLP markers. They were digested with EcoRI and MseI, then ligated to EcoRI and MseI adaptors with T4 DNA ligase. The restriction-ligation products were used as primary template DNA for the first PCR step (preamplification) with no selective EcoRI and MseI primers. Pre-amplifications were controlled by the following program conditions: denaturation of 2 min at 94°C, 26 cycles with 60 s at 94°C, 60 s at 56°C, and 60 s at 72°C and 5 min at 72°C. The initial PCR products were used in the second PCR (selective PCR) with 200 primer combinations of EcoRI and MseI with three selective nucleotides at the 3´end. The thermal profile in the selective PCR was more stringent. Selective amplification was run following PCR profile: denaturation of 30 s at 94°C, 13 cycles with 30 s at 94°C, 30 s at 65°C (decreasing at 0.7°C per cycle) and 60 s at 72°C followed by 25 cycles of 30 s at 94°C, 60 s at 56°C, 60 s at 72°C (increasing time at 1 s per cycle). After that, the denatured second PCR products were run on 6% denaturing polyacrylamide gel (19:1) and stained with silver nitrate. In screening the mungbean germplasm, the AFLP markers E-ACC/M-CTG, E-ACT/M-CTA from our experiment and E-CAG/M-TAC and E-CGT/M-CTG reported by Sommanas (2000) were used in the selective amplification step.
DNA extraction and AFLP analysis Data analysis Total genomic DNA was isolated from the youngest full-grown leaf of the same sample field-grown plants mentioned above using the method suggested by Doyle and Doyle (1990). The concentration of DNA was adjusted to 100 ng/µL for AFLP analysis. The method of bulk segregant analysis (Michelmore et al. 1991) was employed to identify markers associating with genomic DNA of the F2 plants. Briefly, the resistant and susceptible DNA bulks were prepared by pooling equal amount of genomic DNA samples from
In the F2 population, the SPAD index was used in the inheritance study. A chi-square (χ2) test was performed to examine the goodness-of-fit between the observed and expected number of plants as well as the number of co-segregating AFLP markers against the Mendelian ratio of 3:1. Demarcation between the resistant and susceptible groups was set based on the average between the lowest SPAD reading of the resistant parent (NM10-12) and the highest SPAD
Plant Soil (2010) 335:423–437
reading of the susceptible parent (KPS1). Segregation testing of the AFLP markers and linkage analysis were performed on individual F2 plants. Linkage between the markers and the gene controlling resistance to IDC was constructed using the MAPMAKER 3.0 program (Lander et al. 1987). The “group” command was used to determine the linkage groups, pair-wise comparisons, and grouping of the markers. The minimum LOD score of 3.0 and a maximum recombination frequency of 30% were specified. The “compare” command was employed to determine the most likely order within a linkage group, and the best order was accepted based on a log-likelihood difference of two or more. The map distance unit was in centiMorgan (cM) of the Kosambi mapping function. The mapping of QTL was performed by the method of interval mapping using MAPMAKER/ QTL version 1.1 based on the phenotypic and linkage map data. The “scan” command at the LOD threshold of 3.0 was used to identify putative QTL in the linkage map. By fixing the strongest QTL, the other possible QTLs were progressively searched. Finally, the “try” command was used to evaluate the genetic models.
Results Inheritance of the resistance to IDC Visual scoring of the field-grown F2 lines showed that 19 lines had no yellowing (no chlorosis), 43 with a Fig. 1 Histogram of SPAD index of KPS1, NM10-12 and their 160 F2 progenies grown in calcareous soil at Nakhon Sawan Field Crops Research Center, Thailand. The distribution of F2 progeny was divided into two groups based on the average SPAD values of KPS1 and NM10-12 (arrow)
427
slight yellowing, 27 with moderate yellowing, 66 with intense yellowing and 5 with necrosis. SPAD reading values and total chlorophyll contents displayed between 5.6 to 49.0 and 0.2 to 2.27 g/cm2 (data not shown), respectively. Correlation coefficients (r) between SPAD index and chlorophyll content, SPAD index and visual score, and chlorophyll content and visual score were all highly significant (p0.72). In the nutrient solution study, the SPAD reading values of F2 population were also significantly different between the parents. The readings ranged from 22.5 to 42.5 for NM10-12 and between 2.5 to 12.5 for KPS1, giving the demarcation value of 17.5 (Fig. 2). Mungbean plants grown in the solution looked less green (less chlorophyll) than the normal field-grown plants. The F2 SPAD indices fell within 2.5 to 42.5 classes without transgressive segregation, giving the number of resistant and susceptible plants at 86:29. The number fitted well to the 3:1 ratio (χ 2 = 0.0029, p>0.96). Thus, resistance to IDC is controlled
428
Plant Soil (2010) 335:423–437
Fig. 2 Histogram of SPAD index of KPS1, NM10-12 and their 115 F2 progenies grown in Half-Hoagland nutrient solution. The distribution of F2 progeny was divided into two groups based on the average SPAD values of KPS1 and NM1012 (arrow)
by a single locus of major gene designated herewith as gene controlling resistance to IDC (IR) in both field and hydroponic experiments. The segregation patterns in Figs. 1 and 2 also showed that there are modifying genes with minor effect conditioning the degree or level of the resistance. Correlation coefficients (r) of SPAD reading between nutrient and field experiment in the F2 population showed highly significant (p< 0.01, df=113), at r=0.92. Identification of AFLP markers linked to the gene controlling resistance to IDC A total of 200 EcoRI/MseI primers were used to test the parents and bulk. The polymorphic bands from the AFLP markers were identified as the ones present in the resistant parent (P1) and bulk versus the ones missing in the susceptible parent (P2) and bulk. The initial screening of DNA from P1, P2 and bulked DNA from 10 resistant and 10 susceptible plants showed polymorphic bands among 50 primer combinations. Five bands from five primer combinations, E-ACC/ CTG, E-AAC/AAC, E-ACT/CTA, E-AAC/CAA1 and E-ACT/CAA1 corresponded with resistance to IDC as determined by SPAD indices. They amplified bands in the resistant parent and resistant F2 bulks, but not in the susceptible parent and susceptible F2 bulks. Thus these five primers were further employed to determine the segregation in IDC among individual F2 plants. Single regression analysis was used to detect the association between presence/absence of AFLP
markers and IDC. E-ACC/M-CTG marker explained up to 72.97% of the variation in IDC. The markers with smaller effect were E-ACT/M-CTA, E-ACT/MCAA E-AAC/M-AAC and E-AAC/M-CAA that explained 60.21%, 46.55%, 45.07%, and 21.14% of the variation, respectively. Analysis of the F2 individuals using the five primer combinations showed that E-ACC/M-CTG and E-ACT/M-CTA were dominant primers segregating according to a 3:1 ratio with 114 and 113 plants showing the markers of NM10-12, respectively (Table 1). Whereas E-ACT/M-CAA, EAAC/M-AAC and E-AAC/M-CAA were found scattered in both resistance and susceptible plants and thus did not fit to the expected ratio. A linkage analysis performed by MAPMAKER 3.0 (Lander et al. 1987) showed that all the markers were located on the same linkage group with the order as shown in Fig. 3. QTL analysis revealed that qIR controlling resistance to IDC located at 26.4 cM of a partial linkage map at LOD score of 27.36, with the additive gene effect (a) of 7.49 and dominance deviation (d) of 2.38. This quantitative trait locus accounted for 76.39% of the variance for IDC. We found that EACT/M-CTA and E-ACC/M-CTG markers flanked respectively at 3.1 and 10 cM away from qIR. Detection of AFLP markers and IDC in mungbean germplasm The 241 mungbean accessions exhibited various degrees of chlorosis that could be visually classified
Plant Soil (2010) 335:423–437
429
Table 1 Segregation of AFLP markers in the F2 mungbean population derived from the cross KPS1×NM10-12 χ2 value
Markers
Number presenta
Number absentb
Expected ratio
E-ACC/M-CTG
114
46
3:1
1.2
0.2733
E-ACT/M-CTA
113
47
3:1
1.633
0.2012
E-ACT/M-CAA
106
54
3:1
6.533
0.0106
E-AAC/M-AAC
97
63
3:1
17.633
