Crop Science
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online 20 May 2008
Published in Crop Sci 48:1098-1106 (2008)
© 2008 Crop Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Supplemental Data
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Mallory, M. A.
Right arrow Articles by Maughan, P. J.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Mallory, M. A.
Right arrow Articles by Maughan, P. J.
Agricola
Right arrow Articles by Mallory, M. A.
Right arrow Articles by Maughan, P. J.
Related Collections
Right arrow Other Grain Crops
Right arrow Cell Biology & Molecular Genetics
Right arrow Crop Genetics

Development and Characterization of Microsatellite Markers for the Grain Amaranths

Melanie A. Mallorya, Rozaura V. Halla, Andrea R. McNabba, Donald B. Prattb, Eric N. Jellena and Peter J. Maughana,*

a Dep. of Plant and Animal Sci., Brigham Young University, Dep. of Plant & Animal Sciences, Provo, UT 84602
b Dep. of Biology, Stephen F. Austin State Univ., Nacogdoches, TX 75962

* Corresponding author (Jeff_Maughan{at}byu.edu).


   ABSTRACT
TOP
 NOTES
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
The grain amaranths (Amaranthus hypochondriacus L., A. cruentus L., and A. caudatus L.) are important pseudo-cereals native to the Americas. The objective of this project was to produce and characterize a set of highly informative, reproducible microsatellite markers for the grain amaranths. A total of 1457 clones were sequenced from three microsatellite-enriched libraries. Of these, 353 contained unique microsatellites. An additional 29 microsatellite loci were identified from 728 bacterial artificial chromosome–end sequences. A total of 179 microsatellites were polymorphic across accessions from the three grain amaranths. Among these polymorphic microsatellite loci, a total of 731 alleles were identified with an average of four alleles per locus. Heterozygosity values ranged from 0.14 to 0.83, with a mean value of 0.62. Thirty-seven (21%) of the markers were polymorphic between the parents of a segregating population. Phylogenetic analysis using the marker data placed A. hybridus L. accessions into two of the three grain amaranth clades, suggesting the polyphyletic evolution of the three cultivated species from different A. hybridus ancestors. The transferability of these markers to A. hybridus, A. powellii S. Wats., and A. retroflexus L. is reported and suggests that these markers may be useful in studying other species within the genus Amaranthus, including several economically important weeds and ornamentals.

Abbreviations: AFLP, amplified fragment length polymorphism • BAC, bacterial artificial chromosome • BES, BAC-end sequence • H, heterozygosity • MAX, longest tandem repeat excluding half-repeats • ONA, observed number of alleles • PCR, polymerase chain reaction • SSR, simple sequence repeat


   INTRODUCTION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
THE GRAIN AMARANTHS (A. hypochondriacus L., A. cruentus L., and A. caudatus L.) belong to the genus Amaranthus L., which includes 60 to 70 species (Sauer, 1976). The three grain amaranths are classified along with their putative progenitor species (A. hybridus L., A. quitensis H.B.K., and A. powellii S. Wats.) in what is termed the A. hybridus complex and are thought to be paleo-allotetraploids (2n = 4x = 32), although chromosome counts of both 32 and 34 have been reported for A. cruentus (Pal et al., 1982; Greizerstein and Poggio, 1994, 1995). While the grain amaranths have been cultivated for centuries in the Americas, they have been underutilized since the Spanish conquest, when they were replaced by Old World crops and their cultivation suppressed due to their deeply rooted use in indigenous religious practices (Sauer 1976, 1993; Iturbide and Gispert, 1994). In the last few decades, the grain amaranths have begun to reclaim some of their importance, largely because of the recognition of the nutritional value of their seed for human consumption (Bressani et al., 1992; Tucker, 1986).

Amaranth grain is 50 to 60% starch, with higher fiber (8%) and more fat (7–8%) than the grain of most cereals (Pedersen et al., 1987; Breene, 1991). The grain amaranths are noted for their relatively high protein content and balance of essential dietary amino acids. Crude protein content from pale-seeded grain types is substantially higher than most cereal grains and has been reported to range from 12.5 to 22.5% on a dry matter basis, with an average of about 15% (Bressani, 1989; Breene, 1991). Lysine is often the limiting amino acid in other cereal grains; however, amaranth seed protein is rich in this essential amino acid, ranging from 0.73 to 0.84% of the total seed protein content (Bressani et al., 1987). Amaranth oil is also of high nutritional quality, containing a relatively high content of squalene (7–8%; Bressani et al., 1987), and is thought to be effective in reducing cholesterol levels in humans (Berger et al., 2003; Martirosyan et al., 2007). The grain amaranths have also been noted for their ability to thrive under extreme abiotic stress (Brenner et al., 2000). They are impressive producers of biomass under warm and dry conditions, an attribute likely related to their C4 photosynthesis (Kadereit et al., 2003). Thus, several researchers have suggested that amaranth may be a useful alternative crop in developing nations, especially in overpopulated and undernourished areas (Pal and Khoshoo, 1974; Sauer, 1993).

The evolutionary origin of the grain amaranths is still under debate, although two hypotheses have been proposed by Sauer (1967, 1976). The first hypothesis is based on geography and suggests that all three grain amaranths evolved independently, while the second hypothesis is based on morphological features and proposes that all three grains are descended mainly from A. hybridus. Molecular studies, including analyses with isozymes (Chan and Sun, 1997), random amplified polymorphic DNAs (Transue et al., 1994; Chan and Sun, 1997), and amplified fragment length polymorphisms (AFLPs) (Xu and Sun, 2001), have attempted to clarify the relationships among the grain amaranths and their relatives. While these studies support Sauer's second hypothesis of a monophyletic evolution of each of the three grain amaranth species from A. hybridus, they have highlighted the need for new methods with greater resolving power to clarify taxonomic relationships within the A. hybridus complex.

Microsatellites are short repeated nucleotide motifs usually one to four base pairs in length that are flanked by conserved sequences and occur ubiquitously throughout eukaryotic genomes (Tautz and Renz, 1984). They are widely considered the genetic marker system of choice because they are highly reproducible, informative, locus-specific, multiallelic, and codominant (Morgante and Olivieri, 1993; He et al., 2002). Microsatellites have been extremely useful in determining taxonomic relationships among closely related individuals and assessing diversity within a species (Ni et al., 2002; Fukunaga et al., 2005; Ellwood et al., 2006). Sun et al. (1999) noted that among probes designed from various types of repetitive sequences, a probe consisting of microsatellite and minisatellite sequence showed the highest polymorphism across the grain amaranths and their close relatives, suggesting that microsatellites may be extremely valuable for characterizing inter- and intraspecific relationships within the A. hybridus complex. While the initial cost of developing microsatellites markers is high, once developed these polymerase chain reaction (PCR)–based markers are inexpensive to use and require less technical expertise relative to other types of molecular markers. Thus, the goals of this project were (i) to develop a collection of reproducible microsatellite markers for the grain amaranths, (ii) to assess the informativeness of these microsatellite markers by screening them against a panel of grain amaranth accessions, and (iii) to use the markers to characterize the relationships of the grain amaranths and their putative ancestors. Moreover, we show the Mendelian inheritance of 37 of these microsatellites in a segregating F2 population.


   MATERIALS AND METHODS
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Plant Material and DNA Extraction
For microsatellite development and characterization, seeds from a total of 35 diverse amaranth individuals, representing 10 A. hypochondriacus accessions, 8 A. cruentus accessions, 9 A. caudatus accessions, 5 A. hybridus accessions, 2 A. powellii accessions, and 1 A. retroflexus L. accession, were obtained from the USDA collection (USDA, Iowa State University, Ames, IA; Table 1 ). For linkage analysis, an F2 population was developed from a cross of PI 482049 (A. cruentus) and PI 477914 (A. cruentus). The F2 population consisted of 92 plants produced by self-fertilizing a single F1 plant provided by David Brenner (USDA, Iowa State University). All plants were greenhouse grown in Provo, UT, in 15-cm (6-in) pots using Sunshine Mix II (Sun Grow, Bellevue, WA) and supplemented with Osmacote fertilizer (Scotts, Marysville, OH). Plants were maintained at 25°C under broad-spectrum halogen lamps with a 12-h photoperiod.


View this table:
[in this window]
[in a new window]

  		Table 1. Amaranthus accessions used in the microsatellite assays. The microsatellite preliminary screening panel consisted of samples 3, 5, 8, 13, 15, 20, 27, and 32.

 
Total genomic DNA was extracted from 30 mg of freeze-dried leaf tissue according to procedures described by Sambrook et al. (1989) with modifications described by Todd and Vodkin (1996). Extracted DNA was quantified using a Nanodrop (ND 1000 Spectrophotometer, NanoDrop Techonologies, Montchanin, DE) and diluted to 30 ng µL–1 in water.

Microsatellite Enriched Library Construction
Four libraries, enriched for microsatellites consisting of AC, AG, AAC, or AAT motifs, were produced by Genomic Identification Services (Chatsworth, CA) using genomic DNA from A. hypochondriacus cv. Plainsman. Genomic DNA was partially digested with a mixture of seven blunt-end restriction endonucleases (RsaI, HaeIII, BsrB1, PvuII, StuI, ScaI, and EcoRV). Size-separated DNA fragments ranging from 300 to 750 bp were ligated with adapters and enriched for each specific microsatellite motifs using biotinylated capture molecules (CPG, Lincoln Park, NJ). The captured fragments were then amplified and digested with HindIII to remove the adaptors and clone the fragments into pUC19. The resulting plasmids were subsequently transformed into competent Escherichia coli DH5{alpha} cells by electroporation.

Microsatellite Identification and Classification
Enriched libraries were plated on S-gal (Sigma, St. Louis, MO) agar media supplemented with 100 µg mL–1 of ampicillin for blue/white selection of recombinant clones. A total of 1457 recombinant clones were sent to the Arizona Genomics Institute (Tucson, AZ) for plasmid DNA isolation and bidirectional sequencing using M13 primers (forward: 5'-GTA AAA CGA CGG CCA GT; reverse: 5'-CAG GAA ACA GCT ATG AC) and standard ABI Prism Taq dye terminator cycle sequencing methodology. The computer program Contig Express (InforMax, Frederick, MD) was used to determine consensus sequences, eliminate redundant clones, and identify microsatellites. Microsatellites were classified as either simple or compound and perfect or imperfect using the classification system given by Weber (1990) with modifications described by Mason et al. (2005).

Primer Design
Primers flanking each unique microsatellite were designed using the Web-based computer program Primer3 version 2.0 (Rozen and Skaletsky, 2000) according to the program's default parameters, with the following exceptions: preferred product size range equal to 150 to 200 bp; melting temperature differences in forward and reverse primers of no more than 1°C; and max poly-X (maximum allowable length of a mononucleotide repeat) of 3. Oligonucleotide primers were synthesized by Integrated DNA Technologies (Iowa City, IA). Primer pairs were assigned names based on their repeat motif (e.g., AHAAT035, where AH = A. hypochondriacus, AAT = motif type, 035 = clone ID).

Bacterial Artificial Chromosome- Derived Microsatellites
Bacterial artificial chromosome (BAC)–end sequence (BES) microsatellites were identified using the Web-based computer program Tandem Repeats Finder (Benson, 1999) and 728 amaranth BAC-end sequences. The sequences were obtained from clones of an A. hypochondriacus (cv. Plainsman) BAC library developed by Maughan et al. (unpublished data). Only sequences with total repeat lengths greater than 20 bp (n = 10 for dinucleotides, n = 7 for trinucleotides, etc.) were selected for primer design using the program Primer3 version 2.0 (Rozen and Skaletsky, 2000) as previously described.

Microsatellite Amplification
Amplification of microsatellite loci was performed in 10 µL PCR reactions using HotStarTaq Master Mix (Qiagen, Germantown, MD) and 30 ng genomic DNA according to the manufacturer's recommendation. The thermocycling profile was 95°C for 15 min followed by 31 cycles of 94°C for 60 s, 56°C for 30 s, 72°C for 60 s, and a final extension at 72°C for 10 min. Polymerase chain reaction products were separated on 3% Metaphor agarose gels (Cambrex Bio Science, East Rutherford, NJ), run in 0.5X TBE at 120 V for 5 h and visualized using ethidium bromide staining with ultraviolet transillumination. Microsatellite alleles using this protocol were effectively resolved with a resolution of at least four base pairs, as evidenced by the molecular ladders run on each gel.

Data Analysis
The number of alleles and the informativeness for each microsatellite locus was determined by calculating heterozygosity (H). For a multiallele system, heterozygosity values can be estimated using the following equation:

Formula
where Pi is the frequency of the ith allele and k is the number of alleles (Nei, 1978).

Phylogenetic analysis of marker data was performed using the distance (neighbor-joining) method and the full heuristic search option (TBR branch swapping, random addition with 10 replications) of the computer program PAUP*4.0 (beta version 4.0b10; Swofford, 2002). Robustness of the topology of the cladogram was evaluated by bootstrap analysis (10,000 replicates) of the data set. Only groups with frequencies > 50% were retained.

Marker segregation was analyzed for conformation to Mendelian ratios expected in an F2 population using a chi-squared test, with two and one degrees of freedom for codominant and dominant markers, respectively. Linkage groups were constructed with a minimum LOD score of 3.0 using the default mapping parameters (LOD > 1.0, recombination threshold = 0.4, ripple value = 1, jump threshold = 5, Kosambi mapping function) of the computer program JoinMap, version 3.0 (Van Ooijen and Voorrips, 2001).

Statistical Models
Statistical analysis of factors contributing to the polymorphism of individual microsatellite markers, measured as the observed number of alleles (ONA) per locus, was performed using stepwise forward selection using the computer program NCSS97 (Hintze, 1997). Microsatellites were classified as described by Mason et al. (2005) according to (i) complexity (simple or compound), (ii) type (perfect or imperfect), (iii) motif (AAC, AAT, AC, other), (iv) total complete repeats, (v) longest uninterrupted stretch of tandem repeat (excluding partial repeats) (MAX), (vi) total length of repeat including nonrepeat bases, (vii) nonrepeat and half-repeat bases, (viii) size of motif (e.g., dinucleotide, trinucleotide), (ix) number of terminal repeats, (x) number of microsatellites amplified per primer pair, and (xi) expected PCR product size. The numerical classifiers were measured in base pairs. Motif, complexity, and type were determined on the basis of the repeat with the largest MAX. Motifs observed less than five times were grouped together in the category "other."


   RESULTS AND DISCUSSION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
CONCLUSIONS
 REFERENCES
 
Library Analysis
To determine the success of the enrichment process for each of the four microsatellite libraries developed, a small number of clones from each library were sequenced and scanned for microsatellites motifs. The initial estimates of enrichment were 67, 22, 50, and 78% for the AC, AG, AAC, and AAT libraries, respectively. Due to the low enrichment estimate of the AG library, it was excluded from further analysis. A total of 1457 clones were sequenced from the remaining libraries, of which 487 clones corresponded to the AAT library, 482 to the AAC library, and 488 to the AC library. Of the total clones sequenced, only 31 (2%) of clones failed to produce high-quality reads (Phred quality > 20). A high rate of redundancy was observed among the sequences, with 938 (64%) sequences being redundant with at least one other sequence in the collection. This redundancy was likely due to the enrichment process that utilized an amplification process after the affinity capture and before cloning (Jones et al., 2002). Thus, after accounting for redundancy and including 29 microsatellites identified in the BES library, a total of 382 unique microsatellite-containing sequences remained, including 201, 69, and 83 from the AAT, AAC, and AC libraries, respectively. The high frequency of AAT microsatellite sequences derived from the AAT library may be an artifact of the enrichment process, or it may be that AAT repeats are characteristically more frequent in the amaranth genome than are the other types of repeat motifs. Evidence for the latter case is seen when searching nonenriched amaranth DNA sequence (e.g., BES) for repeat motifs. Of the 29 microsatellite repeats identified from the 728 amaranth BAC-end sequences (563 kb), the most common microsatellite motifs identified were all AT-rich, with the most frequently observed motifs being AT and AAT. Elevated numbers of AT-rich microsatellites have also been observed in many of the other species of the Amaranthaceae family, including sugar beet (Beta vulgaris L.) (Mörchen et al., 1996), spinach (Spinacia oleracea L.) (Groben and Wricke, 1998), and quinoa (Chenopodium quinoa Willd.) (Mason et al., 2005), as well as in several unrelated plant species. Morgante et al. (2002) reported that AT repeats are particularly frequent in nongenic regions of Arabidopsis thaliana L., soybean [Glycine max (L.) Merr.], and maize (Zea mays L.). Indeed, in a recent large-scale development of microsatellites from BAC-end sequences in soybean, the AT motif was the most common of all motif classes observed, while the AAT motif was the most common trinucleotide repeat observed (Shultz et al., 2007).

Marker Characterization in the Grain Amaranths
Of the 382 microsatellite loci identified, we successfully designed flanking primer pairs for 319, including 157, 66, 76, and 20, corresponding to the AAT, AAC, AC, and BES libraries, respectively. Flanking primers could not be designed for the remaining 63 microsatellite sequences because of Tm constraints and/or the lack of flanking sequence. All primer pairs were initially screened on an exploratory panel of eight diverse amaranth lines (Table 1). Primer pairs that produced monomorphic banding patterns or failed to amplify on this panel were eliminated from further consideration. A total of 179 microsatellites produced strong amplification products that showed simple, polymorphic banding patterns among the grain amaranth accessions in the exploratory panel. All 179 microsatellite sequences have been submitted to GenBank under accession numbers EU094479 to EU094654. Among these were 97, 30, 39, and 13 markers from the AAT, AAC, AC, and BES libraries, respectively. Interestingly, 19 of the polymorphic microsatellites identified in this study amplified two distinct polymorphic bands that appear to represent two independent loci. The amplification of duplicate loci from a single microsatellite marker has been reported for several polyploid plant species (Röder et al., 1998; Han et al., 2004), where it was suggested that each locus represented orthologous loci derived from the independent ancestral genomes of the polyploid. Such loci in the amaranths, once confirmed via segregation analysis, should prove to be valuable tools in elucidating the paleo-polyploidy event that led to the evolution of allotetraploid amaranths (Pal et al., 1982; Greizerstein and Poggio, 1994, 1995).

To characterize the informativeness of these markers, we screened all 179 markers on a larger and more diverse panel of 35 grain and wild Amaranthus accessions (Table 1; Fig. 1 ). Marker informativeness was quantified by calculating the ONA amplified per marker and by calculating the H value associated with each marker (Supplementary Table 1). ONA and H values were calculated for each grain species separately, for the grain species combined, and for the A. hybridus accessions alone (Table 2 ). Limiting the data set to the three grain species (n = 27), a total of 731 alleles were observed with an average of 4 alleles per locus and a range of 2 to 8 alleles observed per locus. Using the H value calculated for the grain species and the thresholds given by Ott (1992), where a marker is considered polymorphic if H ≥ 0.1 and highly polymorphic if H ≥ 0.7, all 179 markers were considered polymorphic, with 59 (33%) of the microsatellite loci being highly polymorphic (H ≥ 0.7). Heterozygosity values ranged from 0.14 to 0.83, with an average H value of 0.62 per locus, and are similar to those obtained from microsatellite development studies in cultivated relatives of amaranth, including sugar beet (0.61; Rae et al., 2000) and quinoa (0.57; Mason et al., 2005).


Figure 1
View larger version (49K):
[in this window]
[in a new window]

  		Figure 1. Amplification of a microsatellite markers (A) AHAAC046 and (B) AHAC016 across 35 diverse accessions, including individuals from all three grain species (Amaranthus hypochondriacus = lanes 2–11; A. cruentus = lanes 12–19; A. caudatus = lanes 20–28), A. hybridus (lane 29–33), A. retroflexus (lane 34), and A. powellii (lanes 35–36). Standards are loaded in the outside lanes (lanes 1 and 37).

 

View this table:
[in this window]
[in a new window]

  		Table 2. Summary analysis of marker results for Amaranthus, including total number of polymorphic microsatellite loci, observed number of alleles (ONA), total alleles observed, heterozygosity (H) value range and average, and total highly polymorphic microsatellites.

 
When the three grain species were analyzed separately, 129, 123, and 136 microsatellite markers were polymorphic among A. hypochondriacus, A. cruentus, and A. caudatus accessions, respectively (Table 2). Amaranthus caudatus showed the highest total number of polymorphic microsatellite markers and the highest number of alleles observed (371), while A. cruentus showed the lowest genetic diversity of the grain species, with only 123 polymorphic markers and 327 total alleles observed. The lower degree of genetic diversity observed in A. cruentus is consistent with observations using other types of genetic markers including restriction fragment length polymorphism, isozyme, and AFLP markers (Chan and Sun, 1997; Xu and Sun, 2001). Chan and Sun (1997) suggested that the decrease in genetic diversity observed in A. cruentus may be a result of the domestication process, where only a small subset of the wild population was initially subjected to artificial selection for specific agronomic characteristics followed by inbreeding to produce true breeding types. In the case of A. cruentus, this domestication process coupled with a limited and uniform cultivation range (Central America) may have further reduced the level of intraspecies variation. Conversely, the varied topography (high plateaus and mountain valleys) and niche cultivation zones characterized by extreme abiotic stresses (drought, frequent frost, and saline soils) of the Andes may account for the increased genetic diversity seen in A. caudatus.

Statistically Important Factors Affecting Microsatellite Polymorphism
To evaluate the factors that influence the informativeness of a potential marker in amaranth, we used a stepwise forward selection model and found that the factors motif (AAC, AAT, AC, etc.) and MAX (base pair length of the longest uninterrupted tandem repeat) were the most significant predictors (P < 0.01) of marker polymorphism, measured as ONA. The model explained 32% of the variation of ONA, and a t test analysis showed that AAT repeats have a significantly higher (P < 0.0001) ONA than other types of repeats, especially when the tandem repeat length is greater than 20 bp. These observations correlate well with those observed by others, including Moriguchi et al. (2003), who observed that microsatellites with high tandem repeat numbers have higher polymorphism (ONA), and Mason et al. (2005), who reported that a definite change in the percentage of polymorphic versus monomorphic markers occurs when the tandem repeat length is greater than 20 bp. The high rate of polymorphism for the AAT motif compared with the other repeat types has also been observed in other plant species. In wheat (Triticum aestivum L.), for example, Song et al. (2002, 2005) reported that among trinucleotide repeats, the AAT motif is the most polymorphic. These observations and the relative abundance of the AAT motif observed in the amaranth genome (see above) suggest that future development of microsatellite markers for amaranth species with high polymorphic content should focus on AAT repeats with a tandem repeat greater than 20 bp.

Genetic Diversity within Weedy Amaranthus Species
Amaranthus hybridus, a putative wild progenitor species of the grain amaranths, showed the most genetic diversity of all the species included in the complete screening panel. One-hundred and sixty microsatellite markers were polymorphic, and 472 total alleles were observed. That >99% of the microsatellite markers (developed from A. hypochondriacus) amplified in A. hybridus is notable because it confirms the close ancestry between the grain amaranths and A. hybridus at the DNA level. The higher genetic diversity observed among the A. hybridus accessions is consistent with an expectation that a wild progenitor species should be more diverse than a derived domesticated species due to genetic drift and selection (Hilu, 1995).

In addition to being the putative progenitor of the grain amaranths, A. hybridus (smooth pigweed) along with several other members of the Amaranthus genus, including A. retroflexus (redroot pigweed) and A. powellii (Powell amaranth), are particularly notorious weeds (Wassom and Tranel, 2005). Various studies have already demonstrated the utility of molecular markers for correcting taxonomic misclassifications among the weedy species within the Amaranthus genus (Wetzel et al., 1999; Wassom and Tranel, 2005); however, taxonomic problems still exist, especially for closely related species, and highly polymorphic markers are needed to resolve these taxonomic questions. Such markers would also be beneficial in intraspecies population studies and for establishing the first genetic maps in these species. To determine the transferability and utility of these microsatellite markers to related weedy species, we evaluated the level of amplification for the 179 polymorphic microsatellite markers in three additional weedy species (A. hybridus, A. powellii, and A. retroflexus). As previously noted, 177 (>99%) of the markers amplified in the A. hybridus accessions, while 158 (88%) and 141 (78%) of the microsatellite markers amplified in the A. powellii and A. retroflexus accessions, respectively. Between the two A. powellii accessions included in the large screening panel, 97 (52%) markers were polymorphic (Supplemental Table 1). The high transferability observed in this study demonstrates the utility of these markers as new molecular tools for use across the Amaranthus genus.

Evolutionary Origins of the Grain Amaranth Species
Neighbor-joining analysis reveals that A. caudatus, A. cruentus, and A. hypochondriacus are monophyletic, while A. hybridus is polyphyletic (Fig. 2 ). Sauer (1967, 1976) proposed two hypotheses for the evolutionary origins of the grain amaranths. The first hypothesis is based on geography and suggests that all three grain amaranths evolved independently: A. caudatus from A. quitensis in the Andean region of South America; A. cruentus from A. hybridus in Central America; and A. hypochondriacus from A. powellii in Mexico. The second hypothesis is based on morphological features and proposes that A. hybridus gave rise to A. cruentus, that introgression of A. cruentus and A. powellii gave rise to A. hypochondriacus, and that introgression of A. cruentus and A. quitensis produced A. caudatus.


Figure 2
View larger version (25K):
[in this window]
[in a new window]

  		Figure 2. Neighbor-joining analysis of Amaranthus accessions based on microsatellite data set. Bootstrap support values are given above branches. Individuals in the tree are identified by their abbreviated species (A. hypo = A. hypochondriacus, A. caud = A. caudatus, A. hybr = A. hybridus, A. crue = A. cruentus, A. retro = A. retroflexus, A. pow = A. powellii), panel number, and geographic origin.

 
Our results support the designation of A. hybridus as the progenitor species of all three grain amaranths but suggest an alternative hypothesis to explain their origins: multiple, independent domestication events from geographically diverse populations of A. hybridus, specifically, a domestication event in central Mexico corresponding to A. hypochondriacus, a domestication event in southern Mexico–northern Central America corresponding to A. cruentus, and a South American Andean domestication event corresponding to A. caudatus. The progenitor status of A. hybridus is supported by the observation that A. hybridus forms hybrids with all of the other species in the complex (Pal and Khoshoo, 1974); the polyphyletic placement of A. hybridus accessions within the dendrogram; the failure of either of the other two proposed progenitors to group with the grain amaranths (Chan and Sun, 1997); and the high genetic diversity observed in A. hybridus.

We also note that in the neighbor-joining tree, accessions of A. caudatus and A. cruentus split into monophyletic subclades corresponding to New and Old World accessions, whereas New World A. hypochondriacus are paraphyletic with respect to Old World accessions. The paraphyletic result is not unexpected considering the biogeography of the grain amaranths, which are all native to the New World and were spread to Asia, Europe, and Africa during post-Colonial American times (Sauer, 1967). With increased sampling, we expect that New World accessions will be paraphyletic and that Old World accessions will constitute a subset of the genetic diversity found in New World populations.

A larger investigation with a wider sampling of Amaranthus species, including numerous accessions of A. hybridus, needs to be conducted to further evaluate our proposed alternative hypothesis of the origins of the grain amaranths, as well as the relationship between New and Old World accessions of the grain species. The transferability and highly polymorphic nature of microsatellite markers across the genus make them ideal for such an investigation.

Mendelian Inheritance of Microsatellite Markers
To evaluate the utility of these markers for future linkage map construction, we investigated the inheritance of the microsatellite loci in a segregating F2 cross (PI 482049 x PI 477914). In total, 40 (22%) loci were polymorphic between these parents and screened on 92 individuals of the F2 population. Thirty-four (85%) of the markers were scored codominantly (1:2:1), while the remaining six (15%) were scored in a dominant fashion (3:1). Three (AHAAT143, AHAAT144, and AHAC008) of the loci deviated significantly (P < 0.01) from their expected Mendelian pattern of inheritance based on chi-squared analysis. All three distorted markers had a significantly higher than expected frequency of the paternal allele (P < 0.0001) and may reflect marker loci linked to genes affecting gametic or zygotic viability (Xu et al., 1997). Linkage analysis, performed using the program JoinMap (Van Ooijen and Voorrips, 2001), identified nine linkage groups consisting of 21 linked microsatellite loci spanning 108 cM (data not shown). While most linkage groups consisted of only two linked markers, two groups with three and four linked markers were observed. The largest linkage group spanned 29 cM. While this is only an exploratory linkage analysis, it highlights two important observations regarding the grain amaranths: (i) the disomic marker inheritance patterns observed herein suggest that if these species are of polyploid origin, they are most likely allotetraploids and are thus amenable to linkage map construction and marker assisted breeding; and (ii) the level of intraspecies polymorphism is limiting (e.g., in this cross, only 22% of the markers were polymorphic), suggesting that either additional markers will be needed to develop saturated intraspecies maps or interspecies populations will be needed to augment the level of polymorphism within a single cross. One such interspecific mapping population currently being constructed at the University of Illinois (Urbana) by Pat Tranel (personal communication, 2007) is between Plainsman (A. hypochondriacus) and 21605-16 (A. hybridus). Preliminary analysis of our data set shows that nearly a threefold increase in the total number of markers (119 microsatellite markers) should be segregating in this interspecies population compared with the intraspecific cross described above.


   CONCLUSIONS
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
REFERENCES
 
We report the first large-scale development of microsatellite markers for the grain amaranths, an important regional crop of Central and South America. These markers will have particular importance in ongoing efforts to characterize germplasm banks, including the development of core collections for use in emerging breeding programs (Diwan et al., 1995). Thirty-seven of the markers were shown to be inherited in a normal Mendelian fashion, demonstrating the utility of these markers for linkage map development—an important first step toward the development of genetic maps and marker-assisted breeding programs. Current efforts are aimed at mapping the microsatellite markers in an interspecies recombinant-inbred line population developed at the University of Illinois. Phylogenetic analysis using these microsatellite markers suggested a new polyphyletic evolution hypothesis for the three cultivated species and suggests that A. hybridus may be a rich source of genetic variation for improving all three of the grain amaranths. Furthermore, the transferability of these markers to weedy Amaranthus (A. powellii and A. retroflexus) species, as reported here, suggests that these markers will undoubtedly be important in clarifying taxonomic relationships among the grain amaranths and their wild relatives.


   ACKNOWLEDGMENTS
 
This research was funded by a Stephen F. Austin State University faculty research grant, as well as grants from the Ezra Taft Benson Agriculture and Food Institute and the Holmes Family Foundation. We gratefully acknowledge Dr. David McClellan (BYU) for his assistance with phylogenetic analyses. We also express our appreciation to Jimena Alvarez Soto and Mary N. Karanu for technical assistance in developing the microsatellite protocols.


   NOTES
TOP
 NOTES
ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
All rights reserved. No part of this periodical may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. Permission for printing and for reprinting the material contained herein has been obtained by the publisher.

Received for publication August 16, 2007.


   REFERENCES
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 




This article has been cited by other articles:


Home page
 The Plant GenomeHome page
P. J. Maughan, S. M. Yourstone, E. N. Jellen, and J. A. Udall
SNP Discovery via Genomic Reduction, Barcoding, and 454-Pyrosequencing in Amaranth
The Plant Genome, November 1, 2009; 2(3): 260 - 270.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Supplemental Data
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Mallory, M. A.
Right arrow Articles by Maughan, P. J.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Mallory, M. A.
Right arrow Articles by Maughan, P. J.
Agricola
Right arrow Articles by Mallory, M. A.
Right arrow Articles by Maughan, P. J.
Related Collections
Right arrow Other Grain Crops
Right arrow Cell Biology & Molecular Genetics
Right arrow Crop Genetics

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