Purdue University PURDUE AGRICULTURE
BOTANY AND PLANT PATHOLOGY
Skip Navigation LinksPurdue Agriculture > Botany and Plant Pathology > Rp1-Magic > Project Description

 Project Description

 
Title: Natural genetic variation for innate immunity in maize

  1. Results from Prior NSF Support:
    Johal. NSF-IOB: Plant-Microbe Interactions, Genetic and molecular insights into mechanisms underlying a maize disease (9/06-8/08; $273,000). The major objective of this proposal is to elucidate how HC-toxin, a determinant of disease caused by race 1 of Cochliobolus carbonum, facilitates maize pathogenesis. This is being accomplished by a combination of genetic and genomics approaches. While the genetic approach focuses on maize mutants that no longer succumb to infection by C. carbonum, the genomics approach is based on RNA profiling to identify genes whose expression is specifically impacted by HC-toxin. Additional objectives are to examine the developmental behavior and evolutionary origin of host genes that confer resistance to this pathogen.
         Three publications resulting from this award [1-3] are indicated with an asterisk in the References Cited section).

    Weil
    . NSF-MCB The Genetics of Genetics: Genes controlling meiotic recombination in maize. (5/04-4/07, $700,000, subawards to H. Dooner, Waksman Institute, W. Eggleston, Va. Commonwealth Univ., P. Schnable, Iowa St. Univ. and S. Stack, Colorado St. Univ.); on no-cost extension). The major objective of this project has been to identify mutants that have altered rates and localizations of meiotic recombination events using marker alleles causing visible kernel phenotypes. We have succeeded in finding ~60 putative mutations, have thus far validated 10 of these (more are in progress) and are initiating mapping of these mutations. Once homozygous rec lines are established we will begin characterizing the placement and number of recombination nodules in these lines, whether their effects are global or restricted to specific regions of the genome and whether they influence the positioning of transposition events.

         NSF-DBI The Maize TILLING Project (09/06-08/08, $1,600,000). The objective of this project is to establish the Maize TILLING Project (MTP) a TILLING service for the maize community and those researching other crop and potential crop plants. This award followed an initial award to set up the service in conjunction with the Seattle TILLING Project and the second award is to facilitate making additional mutant populations to screen and to move MTP to a full cost-recovery basis. We have provided over 500 alleles in over 130 genes to the community, have initiated additional projects in soybean, marigold, sorghum, switchgrass and berries. In addition we have established a service analyzing natural variation among maize inbreds for genes of interest (EcoTILLING), and this is also being expanded to the additional crops mentioned above. Finally, we are developing targeted resequencing methodologies for use with short-read, next generation, massively parallel sequencers (Illumina/Solexa and ABI SOLiD) to increase throughput and reduce costs of mutation detection.
         Eight publications resulting from these awards [4-11] are indicated with an asterisk in the References Cited section.


  2. Overview and significance:
    This project seeks to refine/expand the genetic architecture of the HR (hypersensitive cell death reaction) response, one of plant kingdom’s most important immune responses [12]. However, instead of using artificially induced variation to do so, we propose to make use of the variation that is present naturally in the maize germplasm. Our rationale for this derives from the fact that, despite the use of exhaustive mutagenesis screens in many plant species, our understanding of how HR is triggered and executed remains incomplete [12]. New ideas are required to unlock the mystery of genes and mechanisms that underlie the HR response. One way forward is to exploit natural variation, which has been generated and selected over millions of years of evolution. However, a major challenge to this approach is how to sift through the enormous diversity available.
         To this end, we have devised a simple yet effective method to discover and characterize useful alleles. This method, a variation on enhancer/suppressor screening that we have called MAGIC (for Mutant-Assisted Gene Identification and Characterization), makes use of the phenotype of a mutant (for a gene affecting the trait of interest) as a reporter to discover and analyze relevant, interacting genes present naturally in diverse germplasm. The idea underlying MAGIC was conceived during a project that led to the identification of a natural suppressor of cell death associated with a disease lesion mimic mutation [13]. It involves crossing a mutant to diverse germplasm and then evaluating the mutant progeny for transgressive changes (both suppressed and severe) in the mutant phenotype(s). If the mutation is recessive, we need to go to the F2 generation to be able to detect and analyze such variation. However, for a dominant or partially dominant mutant, evaluations can be made immediately in the F1 to discover lines that contain suppressors or enhancers of the trait (mutation) under study. Mutant F1 progenies from such crosses can then be propagated further to identify, map, and clone genes/QTL that impact the trait positively or negatively.

    Rp1-D21 in (H95/Mo17)

    Rp1-D21 in (H95/B73)

    Rp1-D21 in (H95/Mo20W)

    Fig 1. Effect of genetic background on the Rp1-D21 phenotype
         The system we propose to exploit takes advantage of a constitutively active (“autoactive”) allele of the maize Rp1 disease resistance gene, which encodes a canonical NBS-LRR R protein and confers HRassociated innate immunity to common rust [14,15]. This allele of Rp1, designated Rp1-D21, is partially dominant and causes the spontaneous induction of necrotic HR lesions in a pathogen-independent but genetic background-dependent manner [14]. To explore its utility for MAGIC, Rp1-D21 mutants (in the H95 inbred background) were crossed with a small collection of inbred lines, both local and exotic. We observed a large phenotypic range in the F1 progenies of these crosses; different genetic backgrounds had the ability to markedly suppress or enhance the HR phenotype of Rp1-D21 (Fig. 1 and Table 1). Interestingly, B73 and Mo17, two inbreds that have been used extensively in genetic and breeding (heterotic) studies, had opposing effects on HR. While B73 suppressed the necrotic phenotype significantly, Mo17 enhanced it. This observation allowed us to use the intermated B73 x Mo17 recombinant inbred lines (IBM RILs) population [16] to look into the genetic basis for HR suppression/enhancement by these two inbreds. Plants from 174 IBM RILs were test crossed with Rp1-D21(H95) heterozygotes to generate F1 progeny that segregated 1:1 for Rp1-D21 vs. wild-type (WT) siblings. Mutant progeny from each cross were scored for overall severity of the HR phenotype using a scale of 1-10, 10 being most severe [13], as well as for the height ratio of mutant vs. WT siblings. Using publicly available genotypic data, QTL for these traits were mapped using standard techniques. In addition to revealing a few minor QTL, this analysis identified a relatively large effect QTL for both parameters on Chromosome 10 (bin 3) that contributed 25 % of the variation and suppressed the HR phenotypic score of Rp1-D21 by at least 2 units (Fig. 5). We have designated this QTL Hrml1, for HR modulating locus-1.

         From this use of only two diverse inbred backgrounds, we reasoned that the resources now exist to mine the extent and magnitude of HR diversity in the maize germplasm. The work outlined in this proposal will expand on this idea and will lead the elucidation of the genetic network underlying HR in maize. Eventually this approach could be used to define genetic networks for a wide range of maize genes and processes and, as resources develop, for genes of other species as well. The power of this approach to identify key QTL quickly uses existing maize genomic resources that have been compiled as a result of various NSF funded projects. The proposed research will extend these efforts by pursuing the following specific aims:

    1. Survey the entire panel of Maize Diversity Lines (MDL) for their ability to impact HR induced by Rp1-D21.
    2. Clone and characterize Hrml1 by using a combination of linkage and association mapping tools, especially the nested association mapping (NAM) RIL populations.
    3. Evaluate NAM RIL populations for additional Hrml genes/QTL and clone at least three of them.
    Table 1: Diversity of the Rp1-D21 mediate HR in maize
    Cross Severity Height ratio
    (M/WT)
    B73 x Rp1-D21

    4

    177/245=0.71
    Mo17 x Rp1-D21

    6

    105/220=0.47
    B97 x Rp1-D21

    2

    190/235=0.81
    CML52 x Rp1-D21

    5

    145/240=0.6
    CM103 x Rp1-D21

    5

    140/230=0.61
    CML228 x Rp1-D21

    5

    148/235=0.63
    CML322 x Rp1-D21

    7

    100/225=0.44
    CML333 x Rp1-D21

    2

    187/227=0.82
    HP301 x Rp1-D21

    6

    115/222=0.52
    IL14H x Rp1-D21

    6

    125/230=0.54
    Ki3 x Rp1-D21

    5

    150/250=0.6
    Ky21 x Rp1-D21

    7

    85/210=0.4
    M37W x Rp1-D21

    10

    dead/250=0
    M162W x Rp1-D21

    8

    65/245=0.26
    MS-71 x Rp1-D21

    4

    135/200=0.67
    NC350 x Rp1-D21

    10

    dead/230=0
    NC358 x Rp1-D21

    6

    124/236=0.52
    Oh7B x Rp1-D21

    3

    180/240=0.75
    0h43 x Rp1-D21

    2

    155/200=0.79
    P39 x Rp1-D21

    5

    150/225=0.66
    TX303 x Rp1-D21

    7

    85/245=0.34
    Tzi8 x Rp1-D21

    2

    186/229=0.81
    Mo20W x Rp1-D21

    10

    dead/205=0
    A632 x Rp1-D21

    1.5

    190/205=0.92
    The mutant tester was heterozygous for Rp1-D21 and was in the H95 background. Inbreds shown in bold are part of the core ('founder') set of diverse lines used to generate the nested association mapping (NAM) panel of 5000 RILs. Overall severity and height ratio are highly correlated.
         The proposed work will not only enhance our understanding of HR in maize, but will also impact this area of research tremendously in other model systems. For example, in Arabidopsis, where many resources have been generated and efforts are underway to make sense of the natural variation, many mutant R gene alleles exist that, like Rp1-D21 confer an HR phenotype constitutively. Interestingly, some of these alleles cause both cell death and stunting [17,18], while others mediate stunting only [19,20]. We also have evidence to show that MAGIC can be utilized effectively with both dominant and recessive mutants (described below) and, though not within the scope of this proposal, that it can be performed with other traits of interest not associated with HR. While MAGIC does have an enhancing effect on the power and effectiveness of QTL analysis, its unprecedented utility is in the discovery and mining of variation that occurs in diverse germplasm, both indigenous and exotic.

         Our team is experienced in the techniques required to accomplish these goals. Balint-Kurti has expertise in quantitative aspects of disease resistance and uses extensive mapping and genomics techniques in his research [21- 23]. A major part of the work that led to the identification of Hrml1 from B73 was conducted by Balint-Kurti. Dr. Moyer is included on behalf of Balint- Kurti to conform to USDA and NCSU regulations. Weil has been involved in maize genetics research for the past 20 years, and one of his recent contributions to the maize community is the establishment of the TILLING resource at Purdue [8-10]. Weil and Johal have interacted extensively over the years, and one of the outcomes of these interactions is the conceptualization of MAGIC. In fact, it was Weil who first suggested the use of IBM RILs to explore the genetics of HR underlying Rp1-D21. The PI, Johal, is one of the few maize geneticists to have worked extensively and continuously in the field of plant-pathogen interactions over the last 20 years. Besides being responsible for the cloning of the first gene for disease resistance in plants [24], Johal has been a pioneer in advancing our understanding of the phenomenon of disease lesion mimicry in plants [25-28].

         Our proposal addresses several of the stated functional genomics goals of the PGRP. The focus of the entire proposal is the use of ‘new tools and resources to tackle major unanswered questions in plant biology on a genome-wide scale. In so doing, we also fulfill two other goals of the program, namely, to focus on plants of economic importance and plant processes of potential economic value. Two of the proposal participants, including the PI, will be participating for the first time in this program, thereby contributing to PGRP’s commitment to broaden participation. The project offers significant crossdisciplinary opportunities for scientific training at all levels of education and for engaging in outreach activities aimed at increasing the scientific awareness of our undergraduate students as well as of high school and middle school students and teachers.
  3. Background

    Genetics of the hypersensitive cell death response (HR).
    Plants are constantly exposed to numerous potential pathogens endowed with diverse modes of attack. Despite this unrelenting onslaught from parasitic organisms, it is rather rare to see plants succumbing to overt disease. One key reason for this is the presence of a highly effective and inducible defense, a major component of which is the HR response [12,29-31]. It is called into action by the host when all other means to ward off the pathogen fail. Although it was initially coined to mean only the rapid collapse of cells at the site of infection [32,33], over the years the term HR has come to represent both cell death and the associated induction of a number of other defense responses, including the accumulation of phytoalexins and of pathogenesis-related (PR) proteins at the site of infection, to name a few [12,29].

         HR is under the control of a subset of disease resistance genes, commonly referred to as R genes [12,32,33]. These R genes operate on a so-called gene-for-gene basis to specifically recognize matching avirulence (Avr) effectors from the pathogen. When corresponding host and pathogen gene products encounter each other at the infection site, the R-gene product mediates a recognition and signal transduction response and HR is triggered resulting in the containment of the pathogen [12,32]. Depending on the kind of proteins that these R genes encode, they can be divided into five categories, the largest of which encodes proteins that have a nucleotide binding site (NBS) domain in the middle of the protein and a leucine rich repeat (LRR) domain at the C-terminal end [12,32]. Two kinds of NBS-LRR R genes are known in dicotyledonous species, carrying either a Toll-Interleukin Receptor (TIR) domain or a coiled coil (CC) domain at their N-terminus [12,32]. In contrast, all known monocotyledonous species NBS-LRR genes have a CC domain [32,34].

         R proteins seem to have two functions. The first, as mentioned above, is to perceive the pathogen via recognition of a specific AVR effector. This recognition, which can occur either directly or indirectly [35- 37], causes the R protein to become activated, which in turn triggers a rapid HR response. How R proteins remain in a quiescent but “vigilant” state remains to be established. However, mutations in R genes have been found that abolish their dependence on AVR proteins for activation [14,17,19]. Such aberrant R genes mostly behave as dominant or partially dominant alleles and trigger the HR constitutively in the absence of the pathogen. Two consequences of such “autoactive” R genes are a massive induction of cell death and the consequential stunting of the organism [38].

         The details of the HR cell death reaction as well as the pathway(s) that link R gene activation with the HR remain an enigma [12,32]. Despite considerable research over the past decade, only a few components have been found thus far. Some of these, Ndr1, Eds1, Pad4, Rar1, and Sgt1, were identified in mutageneses conducted to identify mutants that failed to undergo an HR reaction in response to infection by a virulent pathogen [32]. A few others, RIN4 for example, were identified in yeast two-hybrid assays using an NBS-LRR protein as bait [39]. Recently, an Arabidopsis gain-of-function mutant that carries a point mutation in a TIR-NBS-LRR gene was used to isolate a few second site suppressors following mutagenesis with EMS [20,40,41]. A problem with these mutagenic approaches is that they fail to uncover genes that are either redundant or have other essential functions. One way to avoid this problem would be to seek mine natural variation to identify allelic variants affecting this process. This approach has the added advantage that much of this variation is likely to be adaptive in nature and result from unique changes not deleterious to the organism [42].

    Current approaches to explore natural variation.
    Natural variation is critical for evolution and adaptation, and thus is pervasive in all species. Although it has served as a constant provider of the R genes, natural variability has not been tapped as a tool for understanding other aspects of the disease resistance response [43]. In part this is because natural variation is genetically complex, governed by the action and interaction of many genes (quantitative trait loci, or QTL) and the environment [42,44]. In addition, it remains mostly hidden, surfacing only under certain environmental conditions or in specific genetic backgrounds [45,46]. However, thanks largely to advances in quantitative genetics and genomic technologies, two tools/methods have emerged that allow QTL underlying natural variation to be explored, identified and mapped.

         One of these tools is QTL mapping, which is based on detecting associations between genetic markers and the trait of interest. It involves the use of statistical techniques and DNA marker technology to identify regions of the genome that correlate with variations in the trait phenotype [44,47,48]. The quickest way to discover QTL is to use F2 or backcross (BC) populations derived from a bi-parental cross. However, plant geneticists favor the use of immortalized populations, especially recombinant inbred lines (RILs) [16,48-51]. RILs are derived from an F2 population by single seed descent through at least 5 rounds of selfing. Typically kept as a set of ~100-300 lines, RILs collectively have the genomes of the two parents that gave rise to them represented nearly equally but in a randomly fragmented form. A key advantage of RILs is that multiple replications of the same population can be scored over many environments for any trait of interest, providing more accurate measurement of traits and consequently higher confidence in the QTL identified. In addition, QTL impacting multiple traits can be identified using the same population without additional genotyping. A key disadvantage of RILs (and all other biparentally-derived populations) is that they represent only the two parent genotypes, tapping only a small fraction of the functional diversity present in the germplasm as a whole. Making multiple RIL populations between many combinations of diverse, inbred parents is one way to alleviate this problem, and it has proven to be a powerful alternative approach in maize and in Arabidopsis [51,52].

         Another kind of immortalized population for QTL mapping is near isogenic lines (NILs) or introgression lines (ILs), which are produced by repeated backcrossing of an F1 hybrid to one of the parents [53-56]. An extension of the NILs idea is the advanced backcross QTL analysis, which can be used to specifically find and introgress agronomic QTL (from exotic lines) into elite backgrounds quickly [57].

         An alternative approach to map QTL that has become available recently to plant geneticists is association analysis also called “association mapping”. Association studies are based not on linkage mapping but on linkage disequilibrium (LD), which refers to the nonrandom association of alleles at different loci [58]. Association analysis evaluates whether certain gene or marker alleles within a population are found associated with specific phenotypes more frequently than expected by chance [59]. Originally developed for use in mapping human disease genes, association analysis relies on surveying natural genotypic variation to exploit the rich history of alleles and recombination that have accumulated throughout evolution [60]. Because of this, even a collection of inbred lines can be used for QTL mapping using association mapping. In fact, a population of 302 lines, called maize diversity lines (MDLs), has been established that comprise ~85% of the known variation in maize [61]. Statistical approaches to control for population structure have successfully been applied to this population [62,63].

         Association mapping can be conducted either to scan the whole genome or for candidate-gene testing [60,61]. In a genome scan, SNP markers are identified across the genome at an appropriate marker density, while candidate-gene testing involves sequencing only the candidate gene across a panel of diversity lines. Success of either method depends on population size and the degree of LD, which varies from species to species as well as population to population within a species [61]. Genome scans are most useful in species or populations with moderate to extensive LD, because species with low LD require prohibitively large number of markers to cover the genome [58]. The candidate-gene testing is more effective for populations with low LD. This is the case with the Maize Diversity Lines (MDLs), both the entire collection and the founder set, in which LD decays rather quickly, typically within a few kb [58]. The low LD of these lines is ideal for fine mapping of QTL that are initially identified by linkage analysis (conventional QTL mapping) [58].

         However, all of these approaches discussed above for mapping QTL take substantial amounts of time, expense and effort, and all of them hinge inevitably, at some stage in any protocol, on having extensive phenotypic data for hundreds of plants or populations. This remains a formidable challenge because of the difficulties of measuring phenotypes effectively and precisely [48,60].

    MAGIC.
    The need for a rapid means of assessing genes that affect a trait incrementally across a wide diversity of germplasm in some focused way led us to consider an additional approach for identifying QTL. Using mutant phenotypes associated with genes of interest as an indicator to reveal QTL, MAGIC has the potential to alleviate some of the problems discussed above.

         It is well known that genetic background influences the penetrance (whether or not a given phenotype is detected) and expressivity (the strength of a given phenotype) of mutations. As a result, when one seeks to identify a gene underlying a trait or phenotype of interest, the mutation is introgressed into an inbred background to help clarify the inheritance pattern of the mutation. Indeed, it is generally regarded as useful to cross a mutation into several inbred backgrounds and look for the background that gives the most consistent and pronounced phenotype, then discard the rest. We suggest that rather than being a confounding problem, the variation observed for a mutation among different inbred backgrounds (so-called “background effects”) is instead a valuable asset for discovering interacting genes and QTL.

         The principle underlying MAGIC is akin to the mutagenesis that geneticists often use to isolate extragenic suppressors or enhancers of individual mutations in genetic screens [64]. However, the genes (alleles) identified with MAGIC have existed in natural populations for millions of years and have been defined and refined by natural selection [46]. Alleles in natural populations may also have more complex changes than single base mutations (or other consequences of conventional mutagenesis) [48].

         As mentioned earlier, MAGIC can also be utilized starting with recessive mutations and we have applied the approach successfully on a small scale to a study of genetic background factors impacting the disease lesion mimic phenotype of the lesion mimic23 (les23) mutant. These mutants show spontaneous cell death resembling what is caused by pathogen attack, but do so in the absence of the pathogen [25,26,65]. Some maize les mutants have phenotypes so extreme that they are lethal in one genetic background, yet have a nearly undetectable, benign phenotype in another background [25] (Fig. 1). This easily discerned, visible phenotype manifests in a developmentally specific, cell-autonomous manner, and the sensitivity of les mutations to background effects can be exploited to determine genetically the nature of these background factors.

         An F2 population was developed between les23, a recessive lesion mimic mutant characterized by formation of yellowish-brown cell death patches on leaves, and the inbred Mo20W, known to suppress a number of les mutations [25,66]. Approximately 3,000 F2 plants were grown in the field, and ~900 were les23/les23 homozygotes. We phenotyped these plants and grouped them into ten discrete classes ranging from very severe to highly suppressed. In addition, we noted the timing of lesion onset for all 900 les23 segregants. Genotypes of these les23 segregants were determined using 103 SSR markers and QTL mapping established associations between the genotype and phenotype. We identified a major suppressor of cell death (Slm1) that accounted for ~70% and ~90% of the phenotypic variation in overall symptom severity and the timing of lesion initiation, respectively [13]. The use of les23 as a reporter thus facilitated the identification of a novel QTL, which our recent results suggest protects plants from a number of abiotic stresses, including excess light and heat.

         While recessive alleles can be used in MAGIC analyses, dominant or partially dominant mutations are even more well-suited because they can provide information in the F1 populations. F2 populations (F1 wild type X mutant sib mating) can also be generated to uncover recessive suppressors or enhancers, if any. MAGIC can also be used on cytoplasmic traits, identifying modifiers of these traits that are nuclear genes. In fact, restorers of fertility (Rf) in cytoplasmic male sterile maize and other plants were identified using essentially this approach [67]. Similarly, transgenic variants, made either by overexpression or RNAi knockdown of a gene, could also be used to reveal QTL in diverse germplasm for the trait of interest in a single generation.


  4. Rp1-D21 and the identification of Hrml1:

    Fig 2. A 3-week-old Rp1-D21(H95) heterozygote in the H95 background
    As mentioned earlier, Rp1-D21 is an aberrant allele at the Rp1 locus which confers resistance to specific races of the common rust pathogen, Puccinia sorghi [14]. Rp1 is a complex locus consisting of a family of tandemly-repeated, tightly linked and closely related resistance genes or paralogs whose number/organization (haplotype) differs from gene to gene [15]. For instance, the Rp1-D haplotype is composed of Rp1-D plus eight paralogs that exhibit 91-97% DNA identity with Rp1-D. The Rp1-D21 allele was derived from a crossover event between paralog 2 and paralog 9 (the Rp1-D gene itself), such that it acquired the NBS domain and the first 13 of the 15 repeats of the LRR region from paralog 2 and the remainder of the LRR region from paralog 9 [68]. It still contains intact paralog 1, but paralogs 3 through 8 are missing from Rp1-D21.

    Plants carrying Rp1-D21 exhibit HR constitutively under both greenhouse and field conditions, albeit in a developmentally-programmed fashion (Fig. 2). Developing Rp1-D21 lesions accumulate both superoxide and hydrogen peroxide, two hallmarks of the HR response (Fig. 3). They also induce a number of defense markers (Fig. 4), suggesting the HR lesions induced by Rp1-D21 are identical to the bona fide HR lesions triggered at the site of infections. It has been reported that Rp1-D21 requires some biotic stimulus to form HR lesions [14]. This conclusion was based on data from Rp1-D21 seedlings that were grown axenically in 2L bottles. However, the higher humidity and lower incident light of these growing conditions, both of which are known to dampen the HR response [18,33,69], could also explain these results and we have not detected any requirement for a biotic stimulus.

    Fig 3. In situ staining of developing Rp1-D21 lesions with nitroblue tetrazolium (A) and di-amino benzadine known to dampen the HR response (B), showing the production of superoxide and H2O2.













         We identified the background-dependent nature of Rp1- D21 because of Slm1, which we had found earlier in the Mo20W inbred background as a QTL able to suppress cell death associated with les23, as well as a number of other les mutants. However, Slm1 did not suppress the HR associated with Rp1-D21, if anything, it appeared to enhance it (Fig. 1). Control crosses were also made between Rp1-D21 and B73 and Mo17. Interestingly, in the F1, B73 suppressed Rp1-D21, while Mo17 enhanced the Rp1-D21 phenotype (Fig. 1). The HR phenotype of Rp1-D21 was so severe in the H95/Mo17 background that we were unable to propagate the plants further. This raised the question of how to look into the genetic basis of suppression in B73 and of enhancement in Mo17. Weil’s suggestion that the IBM population could be used worked well and led to the identification of Hrml1.
    Fig. 5. A QTL scan of chromosome 10 showint the size and location of Hrml1. The Rp1 locus is also located on ch 10, but many cM away.
         Hrml1 was mapped as follows. Rp1-D21(H95) was crossed to 174 of the 302 IBM lines. Since the Rp1-D21(H95) plants were heterozygous for the Rp1- D21 gene, each of the 174 resulting F1 families segregated 1:1 for Rp1-D21 but were otherwise isogenic. This population of 174 families was assessed in three different environments: in the greenhouse (scored at six times between 17 and 44 days after planting) in the field in West Lafayette, IN (one replication) and in the field at Clayton, NC (two randomized replications). In each case at least 8 plants from each F1 family were scored for 1) average lesion severity of the Rp1-D21 segregants, scored on a 1-10 scale, and 2) average height of the Rp1-D21 plants divided by average height of the wild type segregants from the same F1 family. The Windows QTL cartographer version 2.5 software package was used to detect the QTL using standard methodology [22]. A moderate size QTL was found to map on chromosome 10, bin 3, somewhere between 144-192 map units on the IBM2 map (Fig. 5). (http://www.maizegdb.org) (Fig. 5). Since it was identified in all three experiments (replications), we considered it to be genuine and have named it Hrml1.

         In summer 2007, we generated an 100 additional Rp1-D21(H95) x IBM F1 families. We will therefore run field trials using at least 270 F1 families in 2008. This should give us some extra precision and power for QTL mapping. We are also in the process of introgressing Rp1-D21 into B73. Four backcrosses have been made and the 5th will be finished this winter in the greenhouse. Plans have been made to cross the entire MDL (maize diversity lines) population with both Rp1-D21 in both H95 and B73 backgrounds.

  5. Experimental Plan:

    Specific goals: This proposal aims to identify the extent of HR-impacting variation that exists naturally in maize and then clone at least 4 genes with the ability to suppress or enhance the HR response. An innovative idea will be used to unveil diversity of HR in maize.

    Objective 1. Survey Maize Diversity Lines (MDLs) for Rp1-D21-mediated HR diversity

    As noted earlier, the MDLs consist of 302 diverse inbreds that have been compiled to capture ~85% of the diversity present in public sector maize breeding programs worldwide . This germplasm set is comprised of current breeding lines as well as historically important lines from both temperate and tropical programs, including 8 popcorn and 7 sweetcorn lines with genetically distinct breeding histories [61]. Because these lines have also been characterized with regard to their population structure and linkage disequilibrium they represent a high-resolution platform for QTL dissection and association mapping. A subset of 25 of these MDLs, called the “founder” or core set, represent ~65% of the worldwide diversity in a more manageable number of lines. As described in detail later (objective 3), each of these ‘founders’ were crossed with B73 generate 25 separate recombinant inbred line (RIL) populations of 200 RILs. The 5000 RILs thus developed have been named NAM, for nested association mapping.

         As mentioned earlier, we have identified a great deal of HR-impacting diversity in the founder MDLs (Table 1). Balint-Kurti and colleagues have developed an extensive dataset on the resistance of the full 302 line population to three necrotrophic diseases (southern leaf blight, northern leaf blight and gray leaf spot) and have observed a large amount of phenotypic variation in disease resistance, well beyond that displayed amongst the 25 lines constituting the core set. Johal has assessed the population for resistance to common rust and has observed several lines with high levels of susceptibility including one which even allows rust to sporulate on leaf sheaths, which to our knowledge has never been reported before (Figure 6). This implies that the larger population of 302 lines may contain significant diversity for innate immunity traits.

         To characterize MDLs with respect to HR diversity, they will each be crossed with Rp1-D21 heterozygotes in both the H95 and the B73 backgrounds. There are three reasons for crossing with Rp1-D21 in both backgrounds. First, the HR phenotype of Rp1-D21 F1s in certain MDL x Rp1- D21(H95) crosses is so severe that they die before reaching anthesis. As a result, they cannot be propagated and used for further genetic studies. Since B73 has a suppressive effect on the Rp1-D21 phenotype, Rp1-D21(B73) crosses will allow us to recover Rp1-D21 crossed to these MDLs. Second, including both B73 and H95 backgrounds in these crosses may provide a more complete understanding of the effect of each MDL on the Rp1-D21 phenotype as in one case it will be interacting with a genome that has an enhancing effect on the Rp1-D21 phenotype (H95) and in the other it will be interacting with a genome that has a suppressive effect (B73). It is well known that different QTL can have different effects in different backgrounds [70]. Thus, crossing with B73 may reveal additional QTL that will be missed if crossed only to H95, and vice versa. Several of the MDL lines are very early flowering while several others flower very late. We will try to obtain crosses with each of these lines by planting delays as appropriate and by attempting some of the crosses in the winter nursery, where shorter day lengths tend to reduce disparities in flowering time. Nevertheless, it is likely that it will not be possible to make all the crosses.

         Crossing each MDL with lines heterozygous rather than homozygous for Rp1-D21 allows both mutant and wild type siblings to be compared side by side, thereby providing an ideal way to obtaining the height ratio between mutants and WT siblings.

    F1 Phenotypic surveys: To survey F1 progeny from crosses between each MDL and our two Rp1-D21 heterozygotes, families of 15 F1 kernels will be planted in a complete randomized block design with two replications per location at two locations (Indiana and N. Carolina) in each of two years and will be assessed for the following phenotypic parameters.

    1. The overall severity of the HR phenotype using the 1-10 scale described earlier.
    2. Height ratio of mutants compared to wt siblings after anthesis. This is an easy measurement to make and keeping the comparison within families serves to cancel out differences that may arise among different F1 crosses because of different levels of hybrid vigor.
    3. The date of lesion initiation. Thus far, we have only seen HR lesions forming on Rp1-D21 seedlings after they are at least 2 weeks old. What prevents earlier expression of Rp1-D21-mediated HR and is there genetic variation for this aspect of the HR? Given the extent of diversity we have seen in the MDLs, it is likely that we will see variation in the timing of lesion initiation. If this timing becomes earlier, Rp1-D21 plants may die as soon as they emerge, or may not germinate altogether. We will observe these seedlings at daily intervals once they emerge and then every few days after 3 weeks of age to document this parameter rigorously.
    4. The nature of HR lesions. Are they discrete or spreading? Are they necrotic, chlorotic, or of some other hue? We have observed a variety of lesion types previously in different backgrounds (Fig. 7). Since this is an important aspect of the HR response, we will document the Rp1-D21 lesion phenotypes in detail. In this effort we will use both visual observation and computational image analysis approaches. For image analysis we will expand our existing collaboration with the group of Dr. Chi-Ren Shyu (see letter of collaboration) who have extensive expertise in the image analysis of complex leaf phenotypes in maize.
    5. Electrolyte loss. This can be done fairly easily and generates high quality data for QTL and association studies. This type of assay is a standard assay used for quantitative analysis of cellular response and damage caused by events such as the hypersensitive response [71,72]. We will measure electrolytes using leaf punches from equivalent leaves and leaf areas.

         The F1 survey will begin to tell us about the extent of variation in the maize germplasm with regard to natural genes/alleles capable of impacting HR. The information at this stage reveals, on a genome-wide scale, the loss of any recessive interactors present in either H95 or B73 and the presence of any dominant interactors in the MDLs. In addition, this step is also important for association studies that we will undertake as part of fine mapping and cloning Hrml1 (in Objective 2) and other HR-modulating loci (Objective 3).

    F2 surveys: We will also advance each F1 cross to an F2. However, in this case rather than selfing the F1, we need to cross F1 Rp1-D21 mutants with nomutant siblings. There are two important reasons for this approach. First, many Rp1-D21 F1 plants produce pollen, but no ear. Second, by always scoring Rp1-D21 heterozygotes, the variation we observe cannot be due to changes in copy number of the Rp1-D21 allele.

         F2 progeny (40 plants for each family) will be planted at both the Indiana and North Carolina locations, and they will be surveyed for all of the parameters listed above except the height ratio between Rp1-D21 plants and nonmutant sibs. Plant height will be harder to interpret meaningfully in the F2 because a large number of modifiers controlling plant height will be segregating in these plants that we were able to ignore because they were uniformly heterozygous in the F1.

         We have three goals in generating these F2 populations and surveying them for diversity in HR. The first is that such F2 populations are needed to search for recessive suppressors or enhancers of the HR response. Second, F1s by themselves may be misleading with regard to the gene action underlying a trait. For example, if a particular MDL has both an enhancer and a suppressor of HR, then the mutant plants in the F1 of this cross might show little change in phenotype because the suppressor and the enhancer cancel each other out. However, in an F2 population, we will be able to distinguish these Hrml genes as they segregate away from each other and their individual effects are revealed (assuming they are not completely linked). Third, these F2 populations may facilitate high-resolution mapping and/or validation of Hrml loci that will be identified under objectives 2 and 3.

    Objective 2. Clone and characterize Hrml1

         Hrml1 was mapped using only 174 RILs of the available 302 IBM lines. We have now generated F1 crosses between Rp1-D21(H95) and ~100 of the remaining IBM lines (giving us a total mapping population of >270 F1 families). In 2008, this expanded population will be phenotyped in the field in Indiana and North Carolina (two replications, complete randomized block design). Combining these data with additional markers from the region underlying Hrml1, we anticipate localizing this QTL to a <3cM region.

         We will further improve the precision with which Hrml1 is mapped by exploiting the NAM population and its extraordinary potential to define the genetic architecture of quantitative traits in maize. The design of the NAM population takes advantage of both recent and ancient recombination events in the same population, allowing linkage mapping and association studies to be integrated [52]. All of the NAM components, the common parent (CP), which is B73, 25 founders, and 5000 RILs have been genotyped with 1536 B73-specific rare SNP loci (common parent specific (CPS) markers). The 26 founders (but not the RILs) are also being genotyped to an extremely high density (at least 1 million SNPs) [52]. To fine map a QTL, it is first identified by conventional association between chromosomal segments and phenotype in the 5000 lines. Fine-mapping to within a gene or few genes can then be achieved by projecting the expected SNPs in each of the RILs derived from the dense genotyping of the parents.

    To identify the gene underlying Hrml1, we will also take advantage of additional NAM RILs, which we plan to phenotype for Rp1-D21 (objective 3). With all these data , combined with the fact that Mo17 and the IBM population are also being integrated with the NAM population the above mentioned CPS markers (Holland, personal communication), we should be able to locate Hrml1 to an extremely precise region on chromosome 10 by NAM analysis.

    Identifying the gene underlying Hrml1. Our fine-mapping should reduce the number of candidate genes that might underlie the Hrml1 phenotype to at most ~20, but likely less. We will then focus on sorting through these genes, taking advantage of the soon-to-be available B73 genomic sequence to identify potential candidates similar to known genes. Further prioritization will be based on 1) Any prior knowledge about the functions of genes in the interval; 2) Expression analysis using RT-PCR of replicated, growth chamber-derived samples (genes in the region that are transcribed differentially in Rp1- D21 vs. nonmutant leaves will be prioritized), and 3) Limited sequencing as needed (genes that show potential amino acid sequence polymorphisms between the suppressing B73 and H95 parents will be prioritized).

         Validation of the Hrml1 gene will be accomplished as follows. We will start by using the wellcharacterized, EMS-mutagenized B73 TILLING populations that are already at Purdue [8,9]. We will identify non-silent point mutations in the TILLING populations and cross these mutants to Rp1- D21(H95), looking for failure to suppress the HR phenotype. Silent alleles of the same gene will serve as controls. We will also sequence our candidate gene in the 302 lines of the association mapping population and use these data combined with the phenotypic data for the MDL x Rp1-D21(H95) F1 families generated under objective 1 to perform association mapping to validate a positive association between the Rp1-D21 phenotype and specific alleles of our candidate gene.

         As an alternative, we will use directed mutagenesis to generate loss-of-function alleles of Hrml1. Rp1-D21 is being introgressed into B73 for this purpose. After the fifth backcross, mutant plants (these are heterozygous for Rp1-D21) will be selfed to generate Rp1-D21 homozygotes in the B73 background. These homozygotes will be propagated either by self-pollination where possible—recall, Rp1-D21 homozygotes often make no ear (see above)-- or by sib mating with Rp1-D21 heterozygotes. In the latter case they will segregate 1:1 along with Rp1-D21 heterozygotes. To knock out Hrml1 by EMS, pollen from Rp1-D21 homozygotes will be collected, treated with EMS according to established protocols, and used to pollinate ears of H95. We will generate at least 5,000 M1 seed, which will be planted and screened for plants that have a more severe (so, less suppressed) Rp1-D21 phenotype than the rest of the M1 population. These mutants will be sampled for DNA and RNA, and propagated if possible. The candidate gene will be sequenced from a maximum of 5 of these mutants and compared with the WT gene(s) from B73. Polymorphisms between the mutant vs WT genes will serve to validate that mutant gene as the Hrml1 gene. As indexed populations of transposon-mutagenized B73 become publicly available we will screen these resources as well; however, existing resources are not in this uniform background as of this writing.

         In the event that we have difficulty getting enough pollen from Rp1-D21(B73) homozygotes to carry out an effective mutagenesis, we will use pollen from Rp1-D21(B73) heterozygotes for treatment instead. This approach will require that we generate twice as many M1 seed to identify the same number of mutants as with the homozygous pollen because half of the M1 progeny will not carry the Rp1-D21 allele and will, therefore, be uninformative.

         Once completed, this work will demonstrate and provide methodology for the isolation of QTL from throughout the B73 genome, creating a powerful functional genomics pipeline from an existing, invaluable and rapidly improving resource.

    Objective 3. Evaluate NAM RIL populations for additional Hrml genes/QTL and clone at least three of them.

         A key goal of this proposal is to broaden the methodology described in the first two objectives and demonstrate that collections of naturally diverse germplasm can be mined for networks of interacting genes. As a proof of this principle, we will identify additional Hrml genes/QTL present in diverse maize inbreds using additional NAM RIL populations. The NAM population will inevitably become a standard mapping set in the maize community, the mapping of disease resistance response phenotypes on this population increases its value, enabling the genetic architecture of other traits that are investigated with this population (e.g. other stress responses) to be compared and contrasted with the architecture of the disease resistance response.

         As mentioned earlier, many lines of the core set of MDLs exhibit extensive diversity for Rp1-D21 mediated HR (Table 1). Some of these inbreds significantly suppressed the HR phenotype of Rp1-D21, while others enhanced it and still others had minimal impact (at least in the F1 generation). To make optimal use of our resources for discovering additional Hrml genes, we will first cross Rp1-D21(H95) heterozygotes with the complete RIL populations derived from NAM founders that show either suppressor or enhancer effects in the F1 (or the F2 in Objective 1 (see above)). Based on our data so far from the F1 of these crosses, the first three sets of 200 will be those derived from crosses between B73 and B97, CML333, and Tzi8, respectively, all of which markedly suppress HR when crossed with Rp1- D21(H95) (Table 1). The second set will be those derived from crosses between B73 and M162W, NC350 and TX303, all of which significantly enhanced Rp1-D21 mediated HR. A third, control set of three complete RIL populations will be those derived from crossing B73 with Ki3, IL14H, and HP301, all of which caused little or no change in HR when crossed with Rp1-D21(H95). To take advantage of the full power of the NAM population for cloning genes underlying additional Hrmls, we will also cross 100 RILs from each of the remaining 16 populations to Rp1-D21(H95). The total number of RILs that we will phenotype thus amounts to 3400.

         The resulting F1 progeny will be evaluated for the HR phenotypes using the lesion severity scale of 1 to 10 and the height ratio of mutant to nonmutant siblings, both described earlier. Using the publicly available genotypic data for the NAM population we will perform Nested Association Mapping (in collaboration with ED Buckler, Cornell, and Dr. Jim Holland, NCSU, see attached letters), as well as conventional QTL mapping to estimate the location of Hrml genes segregating in these populations. We will analyze these data by both including and excluding RILs that inherit Hrml1 from B73, which will be roughly half of them. The advantage of doing this kind of differential analysis is that it will allow us to find QTL regardless of whether they are masked by Hrml1 or not.

         As in Objective 2, candidate gene approaches will then be used to identify these genes precisely and validate them. Validation for the suppressing Hrmls will be done by generating the TILLING knockout alleles of these genes as described above. Validation of the enhancer loci will be carried out by using the cloned Hrml allele to transform maize at the NSF-funded Maize Transformation facility at Iowa State University. The transgenic line will then be crossed with Rp1-D21(H95). Both sets of genes will also be validated by association analysis as described in objective 2 above.

    Major outcomes:

    1. A comprehensive screen for natural genes/alleles that contribute to the HR response. Maize is especially well suited for this screen because, being an outbred species, maize is likely to retain much higher levels of allelic diversity than self-pollinated species.
    2. A proof-of-concept for MAGIC, which we believe can be used to mine and exploit natural variation underlying any trait, thereby providing new opportunities to gain insights into the networks of genes and mechanisms that underlie all traits agricultural and scientific importance in plants. The aims of this project are thus perfectly aligned with the program goals of research supported by the PGRP.


  6. Outreach and Training

    The outreach component of this proposal focuses largely on expanding the science curriculum of Grades 6-12. An understanding and demonstration (in understandable terms) of how biodiversity can be useful goes a long way towards reinforcing the message that biodiversity is important and in need of preservation. In addition, we propose that there are ways to introduce continuous and quantitative variation at the earliest levels of introducing genetics that are neither confusing nor statistically intensive. These early introductions to the material can help students have a clearer understanding of the continuous variation they will encounter repeatedly, both in their continuing educations and in the popular press (for example, with regard to human genetics).

         Purdue currently has outreach activities in plant sciences with area schools for Grades 3-4, Grades 6-8 and Grades 9-12, and NCSU has a program established through multiple NSF Plant Genome Research Programs awards to Dr. Ralph Dean (see attached letter). We will build upon these existing infrastructures. This program will have a direct impact on students from underrepresented groups at both levels proposed. Purdue has one of the largest international communities of any university in the U.S. and the local schools all reflect that diversity directly. In addition, schools that we will target through the NCSU program are located in the rural counties around Raleigh, NC and are multiracial (45% percent Caucasian, 40% African American, 9% Hispanic, and 6% Asian or other ethnic group).

         In efforts to expose students in rural counties of North Carolina to the science and technology being pioneered in North Carolina, Dr. Dean and associates created engagement activities that they present in high schools north and east of Raleigh. During the past several years, researchers from their program visited high schools in each of these counties over 20 times presenting their popular DNA Sequencing activity along with discussions on biotechnology and ethics. The response to these visits by students and teachers alike has been overwhelmingly positive. We will work with Dr. Dean to extend the supplemental curriculum they offer.

    Biological Diversity

    We propose two Outreach activities related to our research. The first is a Middle School (grades 6-8) level program demonstrating the value and the usage of genetic diversity using corn as the major example but introducing other species as well. The lower level materials emphasize food, where it comes from and the role plants play in that. The middle school materials add in how the nutritional aspects of plants can be evaluated, introducing the idea of enzyme assays for evaluating starch. The exercise takes varieties that differ in their starch content and demonstrates that starch digestive enzymes are still present even in lines that have little starch.

         Current curricula at the Grade 6-8 level often introduce the idea of biodiversity, particularly emphasizing biomes such as the rainforest, together with some of the problems associated with destruction of the rainforest. However, we are not aware of any curricula that go into how that diversity is utilized beyond saying that it can be and, perhaps, giving a superficial description of an example. We suggest that a closer look at how genetic information can be adapted from a storehouse of diverse germplasm can better introduce students early in their studies to the importance of both preserving biodiversity and to creating collections of and using that diversity.

         Briefly, these teaching modules will start with using a wide range of diverse types of corn to emphasize that two inbreds of maize differ as much at their DNA level as humans and chimpanzees. Using analogies to a person interacting with different groups of people with different sets of expertise, the idea of a gene will be introduced (or re-introduced!) as well as the notion that each gene acts in a context of lots of other genes that impact how it functions. Then, using corn and tomato, two familiar food crops, emphasis will be on genes introduced from wild relatives that improve disease resistance and nutritional quality in domesticated germplasm even if their effects in the wild species may be less noticeable.

         The materials will be assembled and tested in demonstrations at local middle schools in Indiana and in North Carolina. Feedback from both the students and the teachers will then be sought through written surveys on the materials and the presentation methods. We anticipate two iterations of these trials, and then materials will be made generally available, statewide and nationally, through the Purdue K-12 Outreach programs of both the College of Agriculture and the College of Science, the NCSU K-12 Outreach Programs and through the Hoosier Association of Science Teachers, Inc (HASTI), the North Carolina Science Teachers Association (NCSTA) and the National Science Teachers Association (NSTA).

    Quantitative Genetics

    Efforts are underway to add students performing the starch enzyme assays to the material described above for high school level. High school level outreach, through the Purdue Botany and Plant Pathology Dept., has also included lectures on the importance of agriculture, the debate over GMOs and on bioterrorism. We propose adding new materials for use in the high school curriculum.

         In addition to introducing the Dean DNA sequencing high school activity to Indiana, specifically, we propose to add to this activity by developing and then offering teachers an Analysis Module that will use the student’s sequencing data to demonstrate basic tenets of bioinformatics. We will develop a module based on analysis of allelic sequence diversity within maize and between maize and rice. We will provide computers running the sequence analysis software “Sequencher” (or similar). Students will take part in an exercise to assemble contigs and identify polymorphisms. Issues such as sequence quality controls and silent vs non-silent and conservative vs non-conservative mutations will be addressed. We anticipate visiting eight high school classes a semester, which will allow us to engage over 350 students and teachers a year. While it will be difficult to measure the long-term effectiveness of this program over the duration of this project we will gauge our success though teacher and student responses to a questionnaire.

         Furthermore, many high school biology classes now introduce basic Mendelian genetics but do not generally branch out into continuous variation, quantitative traits or how these traits can be studied. Part of this tendency away from such topics is a combination of the notion that students cannot handle the material, together with a lack of understanding of such material by the teachers. Genetics is thought to be difficult enough, without adding intensive statistics on top of it. However, basic discussions of quantitative traits can be handled with some basic statistical concepts (a mean and a standard deviation), some basic genetics and some basic introductions to the molecular biology involved in mapping, all of which are well within the grasp of both the students and the teachers.

         Given that continuously variable traits and their inheritance are becoming increasingly prominent in the popular press, particularly with respect to human genetics, we propose a program introducing the basics of these topics at the high school level. Again, corn can serve as the key model and the materials can draw on the studies we are proposing, introducing these students to the use of quantitative analysis for plant biology. In addition, examples and connections to analysis of human traits woven into the discussion can keep the material tightly linked to the interests of high school students. This approach will also emphasize how easily the analyses can be translated from one system to another. Again, the methodologies and materials will be assembled and tried out on local high school classes, both in Indiana and in North Carolina, which include schools with differing degrees of emphasis on science and mathematics and a broad diversity of student backgrounds, allowing us to sharpen the presentations to both types of audiences and adapt them for use by teachers in both types of environments. Feedback from both the students and the teachers will then be sought through written surveys on both the materials and the presentation methods and, again, we anticipate two iterations of these trials. As mentioned above, the materials and approaches developed will be made widely available through Purdue and NCSU’s existing Outreach distribution resources and through HASTI, NCSTA and NSTA.

  7. References Cited

    1. *Chintamanani S, Multani DS, Ruess H, Johal GS: Distinct mechanisms govern the dosagedependent and developmentally regulated resistance conferred by the maize hm2 gene. Mol Plant Microbe Interact 2008, 21:79-86.
    2. *Balint-Kurti P, Johal GS: Maize disease resistance. In The Maize Handbook, 2nd ed.,. Edited by Bennetzen SHaJ: Springer-Verlag; 2008.
    3. *Sindhu A, Chintamanani S, Brandt A, Zanis M, Scofield S, Johal GS: A guardian of grasses: specific origin and conservation of a unique disease resistance gene in the grass lineage. Proceedings National Academy of Sciences, U.S.A. 2008, In Press.
    4. *Dooner HK, Weil CF: Give-and-take: interactions between DNA transposons and their host plant genomes. Curr Opin Genet Dev 2007, 17:486-492.
    5. *Weil C: Single base hits score a home run in wheat. Trends Biotechnol 2005, 23:220-222.
    6. *Weil C: Transposons in the modern age. Maydica 2005, 50:339-348.
    7. *Weil C: Single base hits score a home run in wheat. Trends in Biotech. 2005, 23:220-222.
    8. *Weil C, Monde R: Getting the point-mutations in maize. Crop Sci. 2007, 47S:60-67.
    9. *Weil C, Monde R: EMS mutagenesis and point mutation discovery. In In Molecular Genetic Approaches to Maize Improvement. Edited by A.Kriz BLa: Springer-Verlag; 2008.
    10. *Weil C, Monde R: Induced mutations in maize. Israeli J. of Pl. Sci. 2008, (in press).
    11. *Weil C, Martienssen R: Epigenetic interactions between transposons and genes: lessons from plants. Curr. Opin. Gen. Dev 2008, (accepted).
    12. Jones JD, Dangl JL: The plant immune system. Nature 2006, 444:323-329.
    13. Penning BW, Johal GS, McMullen MD: A major suppressor of cell death, slm1, modifies the expression of the maize (Zea mays L.) lesion mimic mutation les23. Genome 2004, 47:961-969.
    14. Hu G, Richter TE, Hulbert SH, Pryor T: Disease Lesion Mimicry Caused by Mutations in the Rust Resistance Gene rp1. Plant Cell 1996, 8:1367-1376.
    15. Collins N, Drake J, Ayliffe M, Sun Q, Ellis J, Hulbert S, Pryor T: Molecular characterization of the maize Rp1-D rust resistance haplotype and its mutants. Plant Cell 1999, 11:1365-1376.
    16. Lee M, Sharopova N, Beavis WD, Grant D, Katt M, Blair D, Hallauer A: Expanding the genetic map of maize with the intermated B73 x Mo17 (IBM) population. Plant Mol Biol 2002, 48:453-461.
    17. Shirano Y, Kachroo P, Shah J, Klessig DF: A gain-of-function mutation in an Arabidopsis Toll Interleukin1 receptor-nucleotide binding site-leucine-rich repeat type R gene triggers defense responses and results in enhanced disease resistance. Plant Cell 2002, 14:3149-3162.
    18. Zhou F, Mosher S, Tian M, Sassi G, Parker J, Klessig DF: The Arabidopsis Gain-of- Function Mutant ssi4 Requires RAR1 and SGT1b Differentially for Defense Activation and Morphological Alterations. Mol Plant Microbe Interact 2008, 21:40- 49.
    19. Zhang Y, Goritschnig S, Dong X, Li X: A gain-of-function mutation in a plant disease resistance gene leads to constitutive activation of downstream signal transduction pathways in suppressor of npr1-1, constitutive 1. Plant Cell 2003, 15:2636-2646.
    20. Goritschnig S, Zhang Y, Li X: The ubiquitin pathway is required for innate immunity in Arabidopsis. Plant J 2007, 49:540-551.
    21. Jines MP, Balint-Kurti P, Robertson-Hoyt LA, Molnar T, Holland JB, Goodman MM: Mapping resistance to Southern rust in a tropical by temperate maize recombinant inbred topcross population. Theor Appl Genet 2007, 114:659-667.
    22. Balint-Kurti PJ, Zwonitzer JC, Wisser RJ, Carson ML, Oropeza-Rosas MA, Holland JB, Szalma SJ: Precise mapping of quantitative trait loci for resistance to southern leaf blight, caused by Cochliobolus heterostrophus race O, and flowering time using advanced intercross maize lines. Genetics 2007, 176:645-657.
    23. Gao X, Shim WB, Gobel C, Kunze S, Feussner I, Meeley R, Balint-Kurti P, Kolomiets M: Disruption of a maize 9-lipoxygenase results in increased resistance to fungal pathogens and reduced levels of contamination with mycotoxin fumonisin. Mol Plant Microbe Interact 2007, 20:922-933.
    24. Johal GS, Briggs SP: Reductase activity encoded by the HM1 disease resistance gene in maize. Science 1992, 258:985-987.
    25. Johal GS: Disease lesion mimic mutants of maize: APSnet Feature Story July 2007, American Phytipathological Society. http://www.apsnet.org/online/feature/mimics/. 2007.
    26. Johal GS, Hulbert SH, Briggs SP: Disease Lesion Mimics of Maize - a Model for Cell- Death in Plants. Bioessays 1995, 17:685-692.
    27. Hu G, Yalpani N, Briggs SP, Johal GS: A porphyrin pathway impairment is responsible for the phenotype of a dominant disease lesion mimic mutant of maize. Plant Cell 1998, 10:1095-1105.
    28. Gray J, Close PS, Briggs SP, Johal GS: A novel suppressor of cell death in plants encoded by the Lls1 gene of maize. Cell 1997, 89:25-31.
    29. Dangl JL, Jones JD: Plant pathogens and integrated defence responses to infection. Nature 2001, 411:826-833.
    30. Jones AM: Programmed cell death in development and defense. Plant Physiol 2001, 125:94-97.
    31. Chisholm ST, Coaker G, Day B, Staskawicz BJ: Host-microbe interactions: shaping the evolution of the plant immune response. Cell 2006, 124:803-814.
    32. Bent AF, Mackey D: Elicitors, effectors, and R genes: the new paradigm and a lifetime supply of questions. Annu Rev Phytopathol 2007, 45:399-436.
    33. Mur LA, Kenton P, Lloyd AJ, Ougham H, Prats E: The hypersensitive response; the centenary is upon us but how much do we know? J Exp Bot 2007.
    34. Bai J, Pennill LA, Ning J, Lee SW, Ramalingam J, Webb CA, Zhao B, Sun Q, Nelson JC, Leach JE, et al.: Diversity in nucleotide binding site-leucine-rich repeat genes in cereals. Genome Res 2002, 12:1871-1884.
    35. Dodds PN, Lawrence GJ, Catanzariti AM, Teh T, Wang CI, Ayliffe MA, Kobe B, Ellis JG: Direct protein interaction underlies gene-for-gene specificity and coevolution of the flax resistance genes and flax rust avirulence genes. Proc Natl Acad Sci U S A 2006, 103:8888-8893.
    36. Ellis JG, Dodds PN, Lawrence GJ: Flax rust resistance gene specificity is based on direct resistance-avirulence protein interactions. Annu Rev Phytopathol 2007, 45:289-306.
    37. Dangl JL, McDowell JM: Two modes of pathogen recognition by plants. Proc Natl Acad Sci U S A 2006, 103:8575-8576.
    38. Howles P, Lawrence G, Finnegan J, McFadden H, Ayliffe M, Dodds P, Ellis J: Autoactive alleles of the flax L6 rust resistance gene induce non-race-specific rust resistance associated with the hypersensitive response. Mol Plant Microbe Interact 2005, 18:570- 582.
    39. Mackey D, Belkhadir Y, Alonso JM, Ecker JR, Dangl JL: Arabidopsis RIN4 is a target of the type III virulence effector AvrRpt2 and modulates RPS2-mediated resistance. Cell 2003, 112:379-389.
    40. Palma K, Zhang Y, Li X: An importin alpha homolog, MOS6, plays an important role in plant innate immunity. Curr Biol 2005, 15:1129-1135.
    41. Zhang Y, Li X: A putative nucleoporin 96 Is required for both basal defense and constitutive resistance responses mediated by suppressor of npr1-1,constitutive 1. Plant Cell 2005, 17:1306-1316.
    42. Mitchell-Olds T, Schmitt J: Genetic mechanisms and evolutionary significance of natural variation in Arabidopsis. Nature 2006, 441:947-952.
    43. Holub EB: Natural variation in innate immunity of a pioneer species. Curr Opin Plant Biol 2007, 10:415-424.
    44. Mauricio R: Mapping quantitative trait loci in plants: uses and caveats for evolutionary biology. Nat Rev Genet 2001, 2:370-381.
    45. Rutherford SL: From genotype to phenotype: buffering mechanisms and the storage of genetic information. Bioessays 2000, 22:1095-1105.
    46. Tanksley SD, McCouch SR: Seed banks and molecular maps: unlocking genetic potential from the wild. Science 1997, 277:1063-1066.
    47. Doerge RW: Mapping and analysis of quantitative trait loci in experimental populations. Nat Rev Genet 2002, 3:43-52.
    48. Koornneef M, Alonso-Blanco C, Vreugdenhil D: Naturally occurring genetic variation in Arabidopsis thaliana. Annu Rev Plant Biol 2004, 55:141-172.
    49. Mackey D, Holt BF, Wiig A, Dangl JL: RIN4 interacts with Pseudomonas syringae type III effector molecules and is required for RPM1-mediated resistance in Arabidopsis. Cell 2002, 108:743-754.
    50. Alonso-Blanco C, Koornneef M, van Ooijen JW: QTL analysis. Methods Mol Biol 2006, 323:79-99.
    51. Koornneef M, Alonso-Blanco C, Stam P: Genetic analysis. Methods Mol Biol 2006, 323:65- 77.
    52. Yu J, Holland JB, McMullen MD, Buckler ES: Genetic design and statistical power of nested association mapping in maize. Genetics 2008, 178:539-551.
    53. Eshed Y, Zamir D: An introgression line population of Lycopersicon pennellii in the cultivated tomato enables the identification and fine mapping of yield-associated QTL. Genetics 1995, 141:1147-1162.
    54. Keurentjes JJ, Bentsink L, Alonso-Blanco C, Hanhart CJ, Blankestijn-De Vries H, Effgen S, Vreugdenhil D, Koornneef M: Development of a near-isogenic line population of Arabidopsis thaliana and comparison of mapping power with a recombinant inbred line population. Genetics 2007, 175:891-905.
    55. Monforte AJ, Tanksley SD: Development of a set of near isogenic and backcross recombinant inbred lines containing most of the Lycopersicon hirsutum genome in a L. esculentum genetic background: a tool for gene mapping and gene discovery. Genome 2000, 43:803-813.
    56. Szalma SJ, Hostert BM, Ledeaux JR, Stuber CW, Holland JB: QTL mapping with nearisogenic lines in maize. Theor Appl Genet 2007, 114:1211-1228.
    57. Tanksley S, Nelson JC: Advanced backcross QTL analysis: a method for the simultaneous discovery and transfer of valuable QTLs from unadapted germplasm into elite breeding lines. Theor. Appl. Genet. 1996, 92:191-203.
    58. Yu J, Buckler ES: Genetic association mapping and genome organization of maize. Curr Opin Biotechnol 2006, 17:155-160.
    59. Whitt SR, Buckler ES: Using natural allelic diversity to evaluate gene function. Methods Mol Biol 2003, 236:123-140.
    60. Flint-Garcia SA, Thuillet AC, Yu J, Pressoir G, Romero SM, Mitchell SE, Doebley J, Kresovich S, Goodman MM, Buckler ES: Maize association population: a highresolution platform for quantitative trait locus dissection. Plant J 2005, 44:1054- 1064.
    61. Yu J, Pressoir G, Briggs WH, Vroh Bi I, Yamasaki M, Doebley JF, McMullen MD, Gaut BS, Nielsen DM, Holland JB, et al.: A unified mixed-model method for association mapping that accounts for multiple levels of relatedness. Nat Genet 2006, 38:203-208.
    62. Pritchard JK: Deconstructing maize population structure. Nat Genet 2001, 28:203-204.
    63. Buckler ES, Gaut BS, McMullen MD: Molecular and functional diversity of maize. Curr Opin Plant Biol 2006, 9:172-176.
    64. Page DR, Grossniklaus U: The art and design of genetic screens: Arabidopsis thaliana. Nat Rev Genet 2002, 3:124-136.
    65. Kolkman JM, Conrad LJ, Farmer PR, Hardeman K, Ahern KR, Lewis PE, Sawers RJ, Lebejko S, Chomet P, Brutnell TP: Distribution of Activator (Ac) Throughout the Maize Genome for Use in Regional Mutagenesis. Genetics 2004.
    66. Buckner B, Johal GS, Janick-Buckner D: Cell death in maize. Physiologia Plantarum 2000, 108:231-239.
    67. Duvick DN: Allelism and Comparative Genetics of Fertility Restoration of Cytoplasmically Pollen Sterile Maize. Genetics 1956, 41:544-565.
    68. Sun Q, Collins NC, Ayliffe M, Smith SM, Drake J, Pryor T, Hulbert SH: Recombination between paralogues at the Rp1 rust resistance locus in maize. Genetics 2001, 158:423-438.
    69. Henikoff S, Till BJ, Comai L: TILLING. Traditional mutagenesis meets functional genomics. Plant Physiol 2004, 135:630-636.
    70. Holland JB: Genetic architecture of complex traits in plants. Curr Opin Plant Biol 2007, 10:156-161.
    71. Liu Y, Ren D, Pike S, Pallardy S, Gassmann W, Zhang S: Chloroplast-generated reactive oxygen species are involved in hypersensitive response-like cell death mediated by a mitogen-activated protein kinase cascade. Plant J 2007, 51:941-954.
    72. Pike SM, Zhang XC, Gassmann W: Electrophysiological characterization of the Arabidopsis avrRpt2-specific hypersensitive response in the absence of other bacterial signals. Plant Physiol 2005, 138:1009-1017.