Precise phenotypic evaluation is crucial for marker-trait association analysis [24–27], especially for those traits with incomplete penetrance and phenotypic plasticity. Apart from being influenced by two major genes, resistance of maize to SCMV is also affected by genetic background and various environmental factors [13, 17, 18]. To obtain reliable phenotypic data, we developed a F2 mapping population from a near-isogenic pair where the resistance allele was fixed at the Scmv2 locus, and only Scmv1 was segregating . Furthermore, selection of only susceptible individuals from the F2 population for fine-mapping of the Scmv1 locus reduced the possibility of misclassification, as heterozygotes tend to be more plastic in their response to SCMV inoculation.
Advances in multiple genomic platforms and analytical methods allow fine-mapping and cloning of QTL responsible for important agronomic traits . In the current study, development of polymorphic markers in the 112.39-kb Scmv1 region was a challenge due to highly repetitive sequences. We thus sequenced the positive BAC clones from the 1145, FAP1360A, and Huangzao4 BAC libraries to mask the repetitive sequences at the Scmv1 locus, in order to reveal single/low-copy sequences for marker development. Generally, re-sequencing of more maize inbred lines  and the availability of additional publicly accessible sequence information (such as for Mo17: http://www.phytozome.net/maize.php; and Palomero: http://www.palomerotoluqueno.org/index.php) will facilitate the development of markers in targeted genome regions. During the process of fine-mapping, we found that the published order of BAC contigs (http://www.maizesequence.org) for the Scmv1 region is incorrect. The BAC-based B73 whole-genome physical map suggests that the adjacent BAC contigs #260 and #261 should be physically close to one another. However, the genetic distance is so large that none of the recombinants within the umc1018/umc1753 interval could be further resolved by the newly developed markers based on the contig #261. Only the markers developed from the contig #263 could resolve the recombinants. This indicates that contig #263 is located between contigs #260 and #261, and that the correct contig order must be #260-#263-#261-#262-#264 (Figure 3A). This correct contig order has been released (http://www.maizesequence.org).
Searching for recombinants with an unambiguous phenotype is crucial for fine-mapping. Thus, we used only highly susceptible individuals in the F2 mapping population for fine-mapping, which allowed us to fine-map the Scmv1 locus into a 112.39-kb region flanked by R1-2 and STS-11. From the fact that the Scmv1 locus has been repeatedly detected [8, 12, 14, 21], we inferred that Scmv1 is likely present in various resistant inbred lines. Based on this hypothesis, two RIL populations were used to fine-map the Scmv1 locus into a 59.21-kb region flanked by the markers 579P4 and SNP3.
The public B73 sequence suggested that two predicted genes, the CAS1-like homolog and Zmtrx-h gene, are the most likely candidates for Scmv1 owing to their potential roles in disease defense response [31, 32], although the potential influence of hypothetical proteins in the Scmv1 region on SCMV resistance cannot be excluded. Comparison with the complete CAS1 gene from Arabidopsis thaliana, which encodes 759 residues, suggests that the putative CAS1-like gene may be partial gene containing the TERPENE_SYNTHASES motif. CAS1 is involved in biosynthesis of steroids and non-steroidal secondary metabolites, such as campesterol and stigmasterol . Campesterol is the precursor of brassinosteroids, an important class of plant hormones that functions in cellular signal transduction, whereas the accumulation of stigmasterol stimulates an important plant metabolic process that occurs following bacterial infection of leaves . Given that phenotypic variance associated with changes in the Scmv1 region are influenced by the stage of plant development , it remains to be established whether the situation with Scmv1 closely resembles that of Hm2. a truncated duplicate of Hm1. Compared to Hm1, Hm2 preferentially functions in adult plants, conferring resistance against the leaf spot and ear mold disease caused by Cochliobolus carbonumrace 1[33, 34].
Thioredoxin is a master regulator of cellular redox status , with h-type thioredoxin reported to function in defense responses to viruses  and fungi . Given that the Zmtrx-h protein in the Scmv1 region lacks the conserved WC(G/P)PC motif, which is essential for redox protein activities, it seems unlikely that Zmtrx-h affects cellular redox status. Hence, additional investigation of the roles of the three candidate genes is required to uncover the mechanism of SCMV resistance.
Linkage and association analysis are two prevalent approaches to map genes or QTL. Both can be used in a complementary manner to dissect the genetic basis of traits of interest [38, 39], as well as to fine-map causative variants of targeted QTL . Six pairs of primers were used to test the association of specific stretches of sequence in the Scmv1 region with SCMV resistance. General linear model (GLM) analysis highlighted three polymorphisms associated with SCMV resistance. The significant association between the PAV and SCMV resistance validates the role of the Scmv1 region with disease resistance, whereas the association between two other SNPs and SCMV resistance provides statistical support for Zmtrx-h as the primary candidate for Scmv1. Moreover, these two SNP located in the intron and 3′-end of Zmtrx-h, respectively, seemingly having an influence on the gene’s expression. The CAS1-like-2 gene cannot yet be excluded, as there was no primer pair available in the vicinity of it. Therefore, a complementary functional test via transformation is still needed to identify which candidate gene, if any, defines Scmv1. These transgenic experiments are underway in our lab.
Broad-spectrum resistance (BSR), resistance against two or more types of pathogen species or the majority of races of the same pathogen species , is essential to improve the resistance of crops to various diseases. The quantitative resistance gene Lr34 cloned recently confers resistance to leaf rust, stripe rust, powdery mildew and various other diseases of wheat . The evidence for the presence of multiple disease resistance genes in a maize association population  revealed that quantitative BSR can be conferred by a single gene. The Scmv1 region has been reported to associate with resistance to other members of the Potyviridae family, including WSMV and MDMV [44, 45]. The near isogenic line F7RR/RR, which was used as a resistant parental line in this study, was also found to be resistant to MDMV and ZeMV . According to a model proposed by Kou , a single gene conferring BSR may function in either basal-resistance pathways, in overlapping pathways that confer race-specific resistances, or at sites of cross-talk between different defense pathways. Given that both candidate genes have putative roles in basal resistance, Scmv1 likely confers BSR, which would be advantageous in environments where plants are threatened by multiple pathogens. However, this speculation remains to be verified by artificial inoculation of transgenic plants and recombinants with the appropriate range of pathogens.