| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Dep. of Plant and Animal Sciences, Brigham Young Univ., Provo, UT 84602
b The Foundation for the Promotion and Investigation of Andean Products (PROINPA), La Paz, Bolivia
* Corresponding author (Jeff_Maughan{at}byu.edu)
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
0.70). These microsatellites markers are an ideal resource for use in managing quinoa germplasm, trait mapping and marker-assisted breeding strategies. The wide cross-species transportability of these markers may extend their value to research involving other Chenopodium species.
Abbreviations: H, heterozygosity value • MAX, longest tandem repeat excluding half-repeats • ONA, observed number of alleles • PRO, expected PCR product size
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Quinoa is especially important as a source of protein for subsistence farmers on the Altiplano of Bolivia and Peru (Wilson, 1988), an area that covers 255000 km2 at an elevation of 3500 to 3850 m above sea level and is generally characterized as cold and arid. Quinoa is one of only a few crop plants adapted to the extreme conditions that characterize much of this region (Cusack, 1984; Risi and Galwey, 1984; Prado et al., 2000; Vacher, 1998). Although increasing quinoa productivity is a primary food-security issue in the Andean region, limited research on quinoa genetics and plant breeding has been conducted (see reviews by Fleming and Galwey, 1995; Risi and Galwey, 1984). Fortunately, a new awareness of the importance of quinoa, as well as a growing health food export market for quinoa, has led to the establishment of several new quinoa breeding programs throughout the Andean region of South America. In addition to germplasm management, principal objectives of these programs include enhancing grain yield, disease resistance, drought tolerance, and modulating saponin content (Ochoa et al., 1999). These programs recognize that the development and use of molecular markers are critical to meeting these objectives (A. Gandarillas, personal communication).
Molecular markers are an effective way of enhancing breeding efficiency (Lande, 1991, Patterson et al., 1991; Staub et al., 1996). To date, only a few researchers have reported the development and use of molecular markers in quinoa. Wilson (1988) used allozyme data to confirm the genetic difference between Altiplano and coastal quinoa ecotypes. Fairbanks et al. (1990) used electrophoretic variation in seed proteins for quinoa germplasm characterization. Ruas et al. (1999) used random amplified polymorphic DNA markers (RAPDs), to detect polymorphism among several C. quinoa cultivars and Chenopodium weed species. Maughan et al. (2004) generated the first genetic map of quinoa, primarily on the basis of amplified fragment length polymorphism (AFLP) markers. Unfortunately, the AFLP, RAPD, seed protein, and allozyme markers employed by these researchers are of only limited utility to laboratories based in the Andean region because of the inherent problems of cost, reproducibility, and technology transfer.
In an attempt to make quinoa marker technology more widely available, we herein report the development of quinoa microsatellite markers. Microsatellite loci consist of short tandemly repeated nucleotide motifs flanked by conserved sequences (Tautz, 1989). Polymorphism is detected by standard polymerase chain reaction (PCR) techniques as variation in the number of repeats among individuals (Weber and May, 1989). Microsatellites are multiallelic and generally more informative and are based on heterozygosity values, than RAPD or AFLP markers (Powell et al., 1996). Microsatellite loci are ubiquitous in eukaryotic genomes, with estimates of one microsatellite per 33 kb in plants (Chawla, 2000). While the initial cost associated with development of a microsatellite marker is high because of the requirement of sequence information, once developed, they are easily maintained and shared among laboratories (Maughan et al., 1995). Microsatellites can be assayed with basic laboratory equipment available in most laboratories throughout the world. The ease of use, high reproducibility, low cost, and abundance of microsatellites in plant genomes makes them an ideal marker system for genetic analysis, especially for under-researched crop plants in developing countries where only limited resources are available (Groben and Wricke, 1998; Akkaya et al., 1992; He et al., 2002).
The objective of this study was to develop a collection of reproducible and highly informative microsatellite markers for use in quinoa breeding and germplasm management programs.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
Library Construction
Four microsatellite-enriched libraries were created by Genomic Identification Services, Inc. (Chatsworth, CA) as described by Jones et al. (2002). Briefly, genomic DNA from the Bolivian quinoa variety ‘Surimi’ was partially digested with a mixture of seven blunt-end cutting enzymes (RsaI, HaeIII, BsrB1, PvuII, StuI, ScaI, and EcoRV). Adapters were ligated to size separated DNA fragments ranging from 300 to 750 bp. Enrichment for fragments containing specific microsatellite motifs (CA, AAT, ATG, and TAGA) was accomplished with biotinylated capture molecules (CPG Inc., Lincoln Park, NJ). Captured fragments were amplified and digested with HindIII to remove adapters. The resulting fragments were cloned into the HindIII site of pUC19, which was subsequently transformed into competent Escherichia coli DH5
cells via electroporation.
Microsatellite Identification
Each library was plated on selective LB medium containing 100 µg mL–1 of ampicillin. Recombinant clones were identified by standard blue/white screening methods with IPTG and X-Gal. The Arizona Genomic Institute (Tucson, AZ) performed plasmid DNA isolation and bidirectional (5' and 3') sequencing with M13 primers and standard ABI Prism Taq dye-terminator cycle-sequencing methodology. The computer program Vector NTI (InforMax, Inc., Frederick, MD) was used to remove vector sequence, remove redundant clones, and identify microsatellite repeats. Microsatellites were identified as motifs repeated at least five times. Microsatellites were then categorized according to type, with slight modifications to the definitions by Weber (1990). A perfect repeat was defined as a run of tandem repeats without interruption. Imperfect repeats were defined as two or more runs of uninterrupted repeats where the terminal repeats consisted of at least three full repeat lengths for dinucleotide repeats and two full repeats lengths for repeats of all larger motifs. Internal repeats had to be at least 1.5 repeats in length and be separated by no more than the equivalent of 1.5 repeats in length. Compound repeats were defined as runs of repeats separated by no more than three consecutive nonrepeat nucleotides from a perfect or imperfect repeat of a different motif or 10 uninterrupted mononucleotides.
Primer Design
Flanking primers were designed for each unique microsatellite for which sufficient flanking sequence data were available with the web-based computer program Primer3 version 0.2 (Rozen and Skaletsky, 2000). Primers were designed according to the default parameters of the program with the following exceptions: product size = 150 to 200, max Tm difference = 1°C, and max poly-X = 3. Oligonucleotide primers were synthesized by Integrated DNA Technologies Inc. (Iowa City, IA). Primer-pairs were assigned a locus name on the basis of the repeated motif (e.g., QCA112, where Q = quinoa, CA = motif type, 112 = clone ID).
Primer Screening
All primer-pairs were initially screened on a panel containing seven quinoa accessions and one accession of C. berlandieri (Table 1). Results of the prescreening were used to classify primer-pairs as polymorphic, monomorphic, or nonamplified. All primer-pairs classified as polymorphic markers were screened against a larger panel of 31 quinoa accessions to calculate the marker's informative values (Table 1). Amplification of microsatellite regions was performed in 15-µL PCR reactions containing 90 ng genomic DNA, 0.2 mM of each dNTP, 2.5 mM MgCl2, 1x PCR buffer, 0.1 mM cresol red and 2% (w/v) sucrose, 0.5U Taq polymerase, 1.0 mM forward primer, and 1.0 mM reverse primer. All amplifications were performed with a touchdown amplification protocol as follows: 94°C for 1 min, followed by 5 cycles of 94°C for 30 s, 55°C for 30 s (decreasing 1°C every cycle), 72°C for 1 min; 10 cycles of 94°C for 30 s, 50°C for 30 s, 72°C for 1 min; 5 cycles of 94°C for 30 s, 50°C for 30 s (decreasing 1°C every cycle), 72°C for 1 min; 10 cycles of 94°C for 30 s, 45°C for 30 s, 72°C for 1 min; hold at 72°C for 5 min. PCR products were separated on 3% (w/v) Metaphor agarose gels (Cambrex Bio Science Inc., East Rutherford, NJ) run in 0.5x TBE at 150 V for 5 to 6 h and visualized with ethidium bromide staining and UV transillumination. This protocol effectively identified polymorphic microsatellites alleles with a resolution of at least 4 bp, as evidenced by the molecular ladders run on each gel (data not shown).
Data Analysis
Informativeness is defined as the probability that a marker will distinguish between two randomly selected individuals in a population. Two measures of marker informative values were calculated for each polymorphic marker: (i) heterozygosity (H) and (ii) observed number of alleles (ONA). Heterozygosity, a measure of allelic diversity, was estimated by:
 |
Genetic similarity of accessions was calculated on the basis of similarity matrices with the unweighted pair group method arithmetic average (UPGMA) and Jaccard matching coefficients (Nei, 1978). Phenetic analysis was performed by the computer program NTSYS-pc version 2.1 (Rohlf, 2000).
Statistical Models
Stepwise forward selection models using generalized linear model (GLM) analysis of covariance were used to identify which characteristics of a microsatellite contributed significantly (p < 0.05) to observed polymorphism level, reported as the observed number of alleles (ONA). All statistical analyses were conducted using the computer program SAS 9.1 (Cary, NC). All microsatellites were characterized with the following descriptors: (i) complexity (simple/compound); (ii) motif (AAT, ATG, CA, etc.); (iii) type (perfect/imperfect); (iv) total repeats excluding non-repeat and half-repeat bases (TOTAL); (v) longest tandem repeat excluding half-repeats (MAX); (vi) repeat length including nonrepeat bases (LENGTH); (vii) non-repeat and half-repeat bases (NON); (viii) motif size (dinucleotide, trinucleotide, etc.); (ix) number of terminal repeats (TER); (x) number of microsatellites amplified per primer-pair (NML); and (xi) expected PCR product size (PRO). Numerical descriptors MAX, TOTAL, LENGTH, NON, LENGTH, and PRO were reported in base pairs (bp) to allow comparison across motifs of different sizes (di-, tri-, and tetranucleotide repeats). Motif, type, and complexity were reported for the primary repeat, defined as the repeat with the largest MAX. All motifs represented less than five times were grouped together under the classification "other."
Sequence Homologies
The flanking sequence of each polymorphic microsatellite marker was compared with GenBank entries at the nucleotide level with BLASTN and the amino acid level with BLASTX. Both types of queries were issued through the National Center for Biotechnology Information (NCBI, http://www.ncbi.nlm.nih.gov, verified 22 March 2005) with default settings. Matches with an E value 
1.0E-4 were considered significant.
|
RESULTS AND DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
After accounting for redundancy, a total of 457 unique microsatellite-containing clones were identified, including 149 (33%), 192 (42%), and 116 (25%) from the CA, AAT, and ATG libraries, respectively. Interestingly, several microsatellite motifs other than the enriched motifs were present in clones from each of the three libraries. The most common of these motifs were GA (50 identified) and CAA (25 identified) motifs. This suggests that GA and CAA repeats may be present at a high frequency in the quinoa genome. Indeed, other researchers previously suggested that GA repeats are consistently more abundant than CA repeats in plant genomes (Powell et al., 1996), including soybean [Glycine max (L.) Merr.] (Akkaya et al., 1992), oilseed rape (Brassica napus L.) (Uzunova and Ecke, 1999), and sunflower (Helianthus annuus L.) (Paniego et al., 2002).
Primer Classification
Flanking primers were designed for 397 of the 457 unique microsatellite loci identified, with preference given to microsatellites having repeat lengths 20 bp (Table 2). These loci were derived from 385 clone sequences, including 12 clones with two unique microsatellites repeats that were amplified using separate flanking primers. Thirteen of the microsatellite loci that had extensive motif replication were not true microsatellites as defined by Weber (1990). All primer pairs were initially screened against a panel containing seven diverse quinoa accessions and one accession of C. berlandieri (Table 1). Seventy-two (18%) of the 397 primer pairs failed to amplify fragments in genotypes included in the screening panel. Two hundred eight (52%) of the microsatellite markers were polymorphic among quinoa accessions (Fig. 1). An additional 25 (6%) were polymorphic only when the C. berlandieri accession was included in the analysis; these markers were monomorphic across the quinoa accessions. In only nine instances did a microsatellite amplify in quinoa and not in C. berlandieri, reaffirming the close genetic relationship between the two species (Kadereit et al., 2003).
|
|
Informative values were calculated for each polymorphic marker as ONA and H (Table 2). One quinoa accession, ‘Baer II’, consistently failed to amplify, or exhibited an aberrant banding pattern, and therefore, was not included in any calculations of informativeness (see Fig. 1; lane 32). The observed number of alleles (ONA) ranged from 2 to 13, with an average of four alleles detected per microsatellite locus. This is similar to the reported ONA for other species in the subfamily Chenopodioideae, including spinach (Groben and Wricke, 1998) and sugar beet (Cureton et al., 2002). Heterozygosity (H) values for the microsatellite loci ranged from 0.20 to 0.90 with a mean value of 0.57. These results are similar to the average H reported in other chenopods such as sugarbeet, where mean H was calculated at 0.61 for 25 microsatellites scored among 10 diverse sugarbeet accessions, one sea beet [Beta vulgaris subsp. maritima (L.) Arcang.] and one fodder beet (Beta vulgaris L. subsp. vulgaris) (Rae et al., 2000). According to the definition of Ott (1992), where a marker is considered polymorphic if H 0.1 and highly polymorphic if H 0.7, all of the polymorphic markers reported in this study were considered polymorphic and 67 (32%) were highly polymorphic. These informative values may be slightly underestimated since this study utilized 3% Metaphor agarose (Cambrex Bio Science, Inc.) and not polyacrylamide gel electrophoresis to resolve the microsatellite alleles. In our experiments, 3% Metaphor agarose resolved the PCR fragments to 4-bp resolution. Two-base pair resolution with 3.5% Metaphor agarose has been reported (Groben and Wricke, 1998). The use of Metaphor agarose parallels electrophoresis technology available in the developing countries and is a cost-effective way to analyze microsatellite markers in terms of technical expertise and reagent cost.
Statistically Significant Factors Affecting Polymorphism
Statistical analysis was conducted on marker data to identify key characteristics of the microsatellite that significantly contributed to marker polymorphism, measured as ONA. Stepwise forward selection using a generalized linear model (GLM) indicated that MAX, motif and the interaction of MAX by motif were the most informative combination of significant factors, explaining 34% (R2 = 0.34) of the variation in ONA. The most polymorphic motifs in this study were AAT, CAA, and CA. Future endeavors to identify potentially polymorphic microsatellite markers in quinoa should carefully consider motif as well as MAX.
Moriguchi et al. (2003) observed that microsatellites with high tandem repeat numbers (MAX) have higher variability, measured as ONA, in Cryptomeria japonica D. Don, and is consistent with reports that microsatellites with large MAX values have higher degrees of mutation in Drosophila melanogaster (Goldstein and Clark, 1995; Schug et al., 1998). Empirical observations by several researchers suggest that an appropriate threshold for identification of polymorphic microsatellite loci from sequence data is a minimum repeat length of 20 bp (MAX 10 for dinucleotide repeats or MAX 7 for trinucleotide repeat units) (Beckman and Weber, 1992; Lagercrantz et al., 1993; Cregan et al., 1999). Our data confirm that a definite change in the percentage of polymorphic versus monomorphic markers occurs at a MAX of approximately 20 bp. This change persists up to a MAX of 30 bp, where 87% of the markers are polymorphic. We also note that the percentage of polymorphic markers with an average number of alleles greater than four (the mean ONA for this study) also peaks at a MAX of approximately 30 bp. Even though microsatellites with repeat length nearer to 20 bp may be polymorphic, development of future quinoa microsatellites with high heterozygosity values should focus on repeats with a MAX > 30 bp.
The significance of the motif by MAX interaction suggests that not all motifs responded equally to increases in the MAX value. This can be observed by plotting MAX by ONA for individual motifs. While the overall correlation (r2 = 0.20) between MAX and ONA for all motifs combined was significant, the correlation (r2 = 0.001) between the factors for the motif AAT was not significant, suggesting that MAX is not predictive of ONA when the primary repeat is AAT.
Relatedness of Quinoa Accessions
The 31 accessions of quinoa studied in this report belong to two distinct groups of quinoa germplasm based on ecotype (coastal and Altiplano). Microsatellite data were used to phenetically analyze the quinoa lines by cluster analysis (Fig. 2). The similarity coefficients ranged from 0.22 to 0.65, with the least genetic similarity detected between the Bolivian accession ‘0654’ and the Chilean accession ‘G-205-95DK’, and the greatest similarity occurring between the two sister-line Chilean coastal accession, ‘KU-2’ and ‘RU-2’. These findings agreed well with previous morphological and isozyme studies, which separated the quinoa germplasm into two distinct groups, Chilean coastal ecotypes and Altiplano ecotypes (Risi and Galwey, 1989; Wilson, 1988). Three separate subgroups, encompassing the Salares, Andean, and valley ecotypes, were observed within the Altiplano group. A single accession labeled Baer II, failed to cluster with any of the quinoa lines (Fig. 2). Baer II is a commercial variety from coastal Chile that was expected to cluster with the Chilean coastal ecotypes; however, this DNA sample often failed to amplify or showed aberrant banding patterns suggesting that the DNA sample used was likely not isolated from Baer II.
|
|
One hundred thirty-six of the microsatellite markers (67%) amplified successfully in all of the species in the panel. The most notable PCR conservation was seen in C. berlandieri nuttaliae, where 99.5% of the primer pairs amplified specific PCR products, of which 164 (81%) were polymorphic between the two accessions examined. The lowest level of conservation was seen in C. pallidicaule, where 150 (74%) of the markers amplified, of which 21 (10%) were polymorphic between the two C. pallidicaule accessions. In addition, 164 (81%) of the primer pairs amplified specific PCR products in C. giganteum, of which 68 (34%) were classified as polymorphic (Table 2). As suggested by several researchers, successful amplification of microsatellite loci across related species declines with increased genetic distance (Gaitán-Solís et al., 2002). Thus, these data confirm the close ancestry between quinoa and C. berlandieri and demonstrate the utility of these quinoa microsatellites as new molecular tools for genetic mapping and germplasm characterization across the genus.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
Received for publication May 13, 2004.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  M. A. Mallory, R. V. Hall, A. R. McNabb, D. B. Pratt, E. N. Jellen, and P. J. Maughan Development and Characterization of Microsatellite Markers for the Grain Amaranths Crop Sci., May 1, 2008; 48(3): 1098 - 1106. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| The SCI Journals | Agronomy Journal | Vadose Zone Journal | |||
| Journal of Natural Resources and Life Sciences Education |
Soil Science Society of America Journal | ||||
| Journal of Plant Registrations | Journal of Environmental Quality |
The Plant Genome | |||
