Report

# Copy Number Variation of Multiple Genes at Rhg1 Mediates Nematode Resistance in Soybean

See allHide authors and affiliations

Science  30 Nov 2012:
Vol. 338, Issue 6111, pp. 1206-1209
DOI: 10.1126/science.1228746

## Abstract

The rhg1-b allele of soybean is widely used for resistance against soybean cyst nematode (SCN), the most economically damaging pathogen of soybeans in the United States. Gene silencing showed that genes in a 31-kilobase segment at rhg1-b, encoding an amino acid transporter, an α-SNAP protein, and a WI12 (wound-inducible domain) protein, each contribute to resistance. There is one copy of the 31-kilobase segment per haploid genome in susceptible varieties, but 10 tandem copies are present in an rhg1-b haplotype. Overexpression of the individual genes in roots was ineffective, but overexpression of the genes together conferred enhanced SCN resistance. Hence, SCN resistance mediated by the soybean quantitative trait locus Rhg1 is conferred by copy number variation that increases the expression of a set of dissimilar genes in a repeated multigene segment.

Soybean (Glycine max) is the world’s most widely used legume crop, providing 68% of world protein meal as well as food oil and a renewable source of fuel, with a farm gate value of more than $35 billion in the United States alone (www.soystats.com). Soybean cyst nematode (SCN; Heterodera glycines) is the most economically damaging pathogen of soybean, causing more than$1 billion in annual losses. SCN has infested most major soybean producing areas worldwide, and there are no practical means of eradication (1).

SCN molts through multiple juvenile and adult life stages, including obligate endoparasitic stages on plant roots, to complete its life cycle (1). Infective J2 juveniles invade roots of both susceptible and resistant soybean hosts, then reprogram host root cells to form feeding sites using highly evolved, secreted nematode effectors (2, 3).

The soybean Rhg1 (resistance to H. glycines) quantitative trait locus on chromosome 18 consistently contributes much more effective SCN resistance than any other known loci (4, 5). Rhg1 disrupts the formation and/or maintenance of most potential nematode feeding sites (1). Roughly 90% of the commercially cultivated soybean varieties marketed as SCN-resistant in the central United States use the rhg1-b allele (haplotype), derived from the soybean line PI 88788, as the main SCN resistance locus. The molecular basis of this SCN resistance has remained unclear.

Genetic mapping has placed rhg1-b in an interval that corresponds to a 67-kb segment carrying 11 predicted genes in the genome of the SCN-susceptible but fully sequenced Williams 82 soybean variety (6, 7). It was recently suggested that an amino acid polymorphism in the Glyma18g02590-encoded α-SNAP protein in this interval contributes to SCN resistance (8), although the authors indicated that this polymorphism does not account for rhg1-b–mediated resistance. None of the gene products within the rhg1-b genetic interval resemble canonical plant immune receptors (9).

In the present study, genes from the rhg1-b interval (6) were silenced to test for impacts on SCN resistance (10). Transgenic soybean roots expressing artificial microRNA (amiRNA) or hairpin (RNA interference) constructs were produced using Agrobacterium rhizogenes (1113). Soybean resistance to SCN was measured 2 weeks after root inoculation by determining the proportion of the total nematode population that had advanced past the J2 stage in each root (Fig. 1A) relative to known resistant and susceptible controls (14). Silencing any one of three closely linked genes at the rhg1-b locus of the SCN-resistant soybean variety Fayette significantly reduced SCN resistance (Fig. 1B). Depletion of resistance was dependent on target transcript reduction (fig. S1). Silencing other genes in and around the locus did not affect SCN resistance (e.g., Fig. 1B, genes Glyma18g02570 and -2620) (10). The three Rhg1 genes that were found to contribute to SCN resistance encode a predicted amino acid transporter (Glyma18g02580), an α-SNAP protein predicted to participate in disassembly of SNARE membrane trafficking complexes (Glyma18g02590), and a protein with a WI12 (wound-inducible protein 12) region but no functionally characterized domains (Glyma18g02610) (1517).

Concurrent study of the physical structure of the rhg1-b locus revealed an unusual genomic configuration. A 31.2-kb genome segment encoding the above-noted genes is present in multiple copies in SCN resistant lines (Figs. 2 and 3). The DNA sequence of fosmid clone inserts carrying genomic DNA from the rhg1-b genetic interval identified a unique DNA junction, not present in the published Williams 82 soybean genome, in which the intergenic sequence downstream of (centromeric to) Glyma18g02610 is immediately adjacent to a 3′ fragment of Glyma18g02570 (Fig. 2A, fosmids 3, 4, and 5). The genomic repeat contains full copies of Glyma18g02580, -2590, -2600, and -2610, as well as the final two exons of Glyma18g02570. Whole-genome shotgun sequencing of a line containing rhg1-b revealed greater depth of coverage of this interval by a factor of 10 relative to the surrounding chromosomal region or homeologous regions on other chromosomes (Fig. 2B), suggesting the presence of multiple repeats. Further polymerase chain reaction (PCR) and sequencing tests confirmed the presence of the Glyma18g02610-2570 junction in DNA from multiple SCN-resistant soybean accessions, whereas the junction was not detected in four tested SCN-susceptible varieties including Williams 82 (Fig. 2C and fig. S2). The shared identity of the junction sites in disparate sources of SCN resistance suggests a shared origin of the initial resistance-conferring event at Rhg1.

Gene expression analysis using quantitative PCR (qPCR) determined that the three genes found to significantly affect SCN resistance show higher transcript levels in roots of SCN-resistant varieties than in susceptible lines (Fig. 2D and fig. S2). This suggested that elevated expression of one or more of the SCN-affecting genes could be a primary cause of elevated SCN resistance. Full-length transcripts were confirmed for Glyma18g02580, -2590, and -2610, no transcript was detected for Glyma18g02600, and no hybrid repeat-junction transcript was detected for Glyma18g02570 (fig. S2).

Fiber-FISH (fluorescence in situ hybridization) was used to directly view the arrangement and copy number of the 31-kb repeat segment in different haplotypes of the Rhg1 locus. The hybridization pattern and DNA fiber length estimates (Fig. 3 and table S1) indicate a single copy of the repeat in Williams 82, as in the reference soybean genome sequence (7). In Fayette, fiber-FISH revealed 10 copies of the repeat segment per DNA fiber, in the same configuration throughout the multiple nuclei sampled, in a pattern consistent with 10 direct repeats abutting head-to-tail (Fig. 3 and table S1). In samples from Peking, another common source of SCN resistance, three copies per DNA fiber were present in direct repeat orientation (Fig. 3). No additional copies (e.g., at other loci) were evident. Rhg1 repeat copy number expansion is likely to have occurred by unequal-exchange meiotic recombination events between homologous repeats.

Amino acid polymorphism or overexpression of any one of the three identified rhg1-b genes did not account for SCN resistance. From all available rhg1-b sequence reads (across multiple repeat copies), no predicted amino acid polymorphisms relative to Williams 82 were identified for Glyma18g02580, Glyma18g02600, or Glyma18g02610. Some copies of Glyma18g02590 from rhg1-b resembled the Williams 82 sequence, whereas others contained a set of polymorphisms, notably at the predicted C-terminal six amino acids of the predicted α-SNAP protein (table S2, confirmed by cDNA sequencing). However, expression of this non–Williams 82–type Glyma18g02590 downstream of a strong constitutive promoter or native promoter sequence did not increase the SCN resistance reaction of Williams 82 transgenic roots (Fig. 4 and fig. S3), which suggests that rhg1-b SCN resistance requires more than this 2590–amino acid polymorphism. Overexpression of Glyma18g02580 or Glyma18g02610 also failed to increase SCN resistance (Fig. 4).

Given the above, simultaneous overexpression of genes within the 31-kb repeat segment was tested as a possible source of SCN resistance. In two separate experiments that together tested >25 independent transgenic events for each DNA construct, resistance to SCN was significantly increased in SCN-susceptible Williams 82 by simultaneous overexpression of the set of genes (Fig. 4; see also fig. S4A). A DNA construct overexpressing Glyma18g02580, -2590, and -2610 (but not -2600) also conferred enhanced SCN resistance (fig. S4B). The collective findings indicate that Rhg1-mediated SCN resistance is attributable to elevated expression of Glyma18g02580, -2590, and -2610.

These results reveal a novel mechanism for disease resistance: an expression polymorphism for multiple disparate but tightly linked genes, derived through copy number variation at the Rhg1 locus. This suggests future approaches to enhance Rhg1-mediated quantitative resistance against the globally important SCN disease of soybean—for example, through isolation of soybean lines that carry more copies of the 31-kb Rhg1 repeat. Transgenic overexpression of the native or altered genes may improve SCN resistance and/or be applicable in other species for resistance to other endoparasitic nematodes.

The biochemical mechanisms of Rhg1-mediated resistance remain unknown. Other sequenced plant genomes do not carry close homologs of the predicted Glyma18g02610 protein, although a wound-inducible protein in ice plant with 55% identity has been studied (17). Modeling of the Glyma18g02610 predicted tertiary structure suggested (10) that Glyma18g02610 may participate in the production of phenazine-like compounds that are toxic to nematodes. The Glyma18g02590 α-SNAP protein is likely involved in vesicle trafficking and may influence exocytosis of products that alter feeding site development or nematode physiology (18). Because it is one of at least five α-SNAP homologs encoded in the soybean reference genome, Glyma18g02590 may have undergone subfunctionalization or neofunctionalization (19). The Glyma18g02580 protein and its most closely related plant transporter proteins are not functionally well characterized, but Glyma18g02580 contains a predicted tryptophan/tyrosine permease family domain. Tryptophan shares structural similarity with and is a precursor of the auxin hormone indole-3-acetic acid, which suggests the intriguing possibility that Glyma18g02580 may affect functionally important auxin levels or distribution (2, 3). Together, these genes create an unfavorable environment at nematode feeding sites.

Growing evidence from metazoa and plants suggests that genome structural variation is a frequent and powerful driver of phenotypic diversity (20, 21). Copy number variation of chromosomal subsegments (beyond simple duplication) can affect gene expression levels (22), and single-gene copy number variation contributes to a number of adaptive traits in humans, plants, and insects (2326). Recent analyses of genome architecture in sorghum, rice, and soybean have reported high levels of copy number variation and a tendency for these genomic regions to overlap with postulated biotic and abiotic stress-related genes (2729).

Our work provides a concrete example of copy number variation in which the repeat encodes multiple gene products that contribute to a valuable disease resistance trait. Single-copy clusters of genes that are functionally related but nonhomologous are highly unusual in multicellular eukaryotes, but such clusters have been reported in association with plant secondary metabolism (30, 31). Given the repetitive and plastic nature of plant genomes and the relatively underexplored association between copy number variation and phenotypes, it seems likely that a number of other complex traits are controlled by this type of structural variation.

## Supplementary Materials

www.sciencemag.org/cgi/content/full/science.1228746/DC1

Materials and Methods

Figs. S1 to S4

Tables S1 to S3

References (3348)

## References and Notes

1. See supplementary materials on Science Online.
2. Acknowledgments: Supported by the United Soybean Board (A.F.B.), USDA CSREES (M.E.H.), Illinois Soybean Association (M.E.H. and B.W.D.), NSF grant DBI-0922703 (J.J.), Wisconsin Experiment Station Hatch Award (A.F.B.), and the Pioneer Fellowship in Plant Pathology awarded to D.E.C. by the American Phytopathological Society through a gift from Pioneer Hi-Bred. We thank A. E. MacGuidwin for suggestions regarding SCN experiments and J. M. Palmer for assistance with RNA blots. DNA sequence data are available at NCBI GenBank under accession codes JX907804 to JX907808 and at the NCBI Short Read Archive, study no. SRA059285. A.F.B., M.E.H., B.W.D., D.E.C., T.G.L., X.G., S.M., T.J.H., and J.W. are listed as inventors on a provisional patent application filed by Wisconsin Alumni Research Foundation that covers the findings reported herein.
View Abstract