ARTICLE IN PRESS

0733-5210/$ - se

doi:10.1016/j.jc

Abbreviations

printing metho

Ex6C, allele C

SNP, intron 1 S

intron 1 SNP; I

variation; In5C

morphism; SSC$Mention of

guarantee or wa

and does not im

also can be suit�CorrespondE-mail addr

Journal of Cereal Science 47 (2008) 536–545

www.elsevier.com/locate/jcs

Waxy gene haplotypes: Associations with apparent amylose contentand the effect by the environment in aninternational rice germplasm collection$

Ming-Hsuan Chena,�, Christine Bergmanb, Shannon Pinsona, Robert Fjellstroma

aUnited States Department of Agriculture, Agricultural Research Service, Rice Research Unit, 1509 Aggie Drive, Beaumont, TX 77713, USAbDepartment of Food and Beverage, University of Nevada-Las Vegas, Las Vegas, NV 89154, USA

Received 16 May 2007; accepted 13 June 2007

Abstract

Apparent amylose content (AAC), the key determinant of rice end-use quality attributes, is primarily controlled by the Waxy gene

which codes for granule bound starch synthase (GBSS). We examined the combination of sequence variation in the Waxy gene and

environmental effects, and their associations with AAC using 171 rice accessions originating from 43 countries. The combination of two

single-nucleotide-polymorphism (SNP) markers in the Waxy gene allows for the identification of three marker haplotypes in this gene.

The first SNP is at the leader intron splice site (In1 SNP), and the second polymorphism is in exon 6. The haplotypes explained 86.7% of

the variation in AAC and discriminated the three market classes of low, intermediate and high AAC rice from each other. The

environment affected the AAC of all haplotypes. Higher air temperature during grain development associated with a decrease in AAC of

low and intermediate AAC-types, but with an increase in AAC of high AAC-type. The association of AAC with several Waxy RM190

microsatellite-(CTn) alleles in combination with the In1 SNP was also examined. In conclusion, the Waxy haplotypes studied appear to

be useful markers for selecting the AAC of breeding lines developed from the world’s rice germplasm.

r 2007 Elsevier Ltd. All rights reserved.

Keywords: Amylose; Waxy gene; SNP; Single-nucleotide-polymorphism; SSR; Microsatellite marker; Oryza sativa L.

1. Introduction

Consumer preference for rice around the world, wheneaten as an intact grain, is largely dependent on a desire forits cooked texture to be either firm-and-not-sticky or soft-and-sticky. For the food processing industry, different

e front matter r 2007 Elsevier Ltd. All rights reserved.

s.2007.06.013

: AAC, apparent amylose content; ddF, dideoxyfinger-

d; Ex6 SNP, exon 6 SNP; Ex6A, allele A of Ex6 SNP;

of Ex6 SNP; GBSS, granule bound starch synthase; In1

NP; In1G, allele G of intron 1 SNP; In1T, allele T of

n5 InDel, intron 5 insertion and deletion sequence

, allele C of In5 InDel; SNP, single-nucleotide-poly-

P, single-strand conformation polymorphism

a trademark or proprietary product does not constitute a

rranty of a product by the US Department of Agriculture,

ply its approval to the exclusion of other products that

able.

ing author. Tel.: +1409 752 5221; fax: +1 409 752 5720.

ess: ming.chen@ars.usda.gov (M.-H. Chen).

functional properties such as freeze–thaw stability or flourthickening power are desired depending on the application.Apparent amylose content (AAC) is the key determinant ofthese different rice (Oryza sativa) cooking, sensory andprocessing properties (Bergman et al., 2004). Consequently,rice breeders develop cultivars with improved agronomictraits, while maintaining a targeted AAC that fits thedesired end-use quality characteristics.Milled rice samples are commonly categorized into an

AAC-class or quality type similar to the following:glutinous/waxy- (AAC range: 0–5%), low AAC- (range:6–18%), intermediate AAC- (range: 19–23%), or highAAC- (423%) types. In the USA, the high AAC-class isfurther divided into two sub-classes: those quality typessuitable for commercial thermal processing and those thatare not (Bergman et al., 2004).The Waxy gene, located on rice chromosome 6, encodes

the enzyme—granule bound starch synthase (GBSS),

ARTICLE IN PRESSM.-H. Chen et al. / Journal of Cereal Science 47 (2008) 536–545 537

which plays a key role in apparent amylose synthesis(Smith et al., 1997). Two Waxy gene alleles, Wxa and Wxb,have traditionally been associated with the contents ofGBSS and AAC in rice endosperm, with the Wxa allelesynthesizing higher contents of GBSS, and thus AAC, thanthe Wxb allele (Sano, 1984; Sano et al., 1985). Wang et al.(1995) demonstrated that the AAC of rice is stronglyassociated with the ability of the enzymes to excise theintron 1 from the Waxy pre-mRNA to different degreesbetween the two Waxy alleles, suggesting the geneticcontrol of this starch fraction is at least in-part at thepost-transcriptional level. Subsequent studies identifiedthat the single nucleotide change of G-T at the +1position of the intron 1 consensus cleavage site isresponsible for the characteristics of the Wxb allele: thepresence of pre-mRNA (containing intron 1), low level ofmature Waxy transcript, GBSS and AAC (Bligh et al.,1998; Cai et al., 1998; Hirano et al., 1998; Isshiki et al.,1998; Larkin and Park, 1999). Rice waxy mutants, whichhave no detectable AAC and GBSS, were reportedto contain a premature translation termination codon(Wanchana et al., 2003). In a study of 89 non-glutinousUSA cultivars, an intron 1 single-nucleotide polymorphism(In1 SNP) discriminated low AAC-type cultivars fromintermediate and high AAC-type cultivars (Ayres et al.,1997). However, the In1 SNP was unable to distinguishbetween intermediate and high AAC-type rices.

A sequence change was reported in exon 6 of the Waxy

gene, comprised of an A-C substitution, that results in anamino acid substitution (Larkin and Park, 2003). Theintermediate AAC-type rice accessions studied containedcytosine (Ex6C) at this Ex6 SNP while both adenine(Ex6A) and cytosine (Ex6C) were identified in the highand low AAC-types. The association of Ex6 SNP to AACwarrants further investigation.

A polymorphic DNA microsatellite marker having adinucleotide (cytosine–thymine) repeat (CTn), located atexon 1 in the non-translated region of the rice Waxy genewas identified by Bligh et al. (1995). The CTn alleles of thismicrosatellite, subsequently renamed RM190 (Temnykhet al., 2000), in an historical collection of 89 non-glutinousUSA rice cultivars explained 82.9% of the variation inAAC (Ayres et al., 1997). This association dropped to 68%when 170 non-glutinous rice accessions from 53 countrieswere examined (Bergman et al., 2000). Both of these studieswere performed using samples of rice accessions that werenot grown under the same conditions. Bergman et al.(2001) studied 198 non-glutinous USA cultivars andbreeding lines of diverse parentage, grown in fourlocations, and determined that 88% of the variation inAAC was explained by the RM190 alleles. Thus, theassociation of this microsatellite marker with AAC across agenetically diverse germplasm collection, grown in multiplelocations is yet to be determined.

In addition to the major effect of the Waxy gene, minorgenes and the environment are also thought to influencerice AAC (Bollich and Webb, 1973; Juliano and Pascual,

1980; McKenzie and Rutger, 1983). The AAC is therefore aquantitative trait that requires rice breeders to grow outtheir breeding material during multiple years and multiplelocations to ensure that advanced lines have the desiredAAC level. The molecular markers discussed above arecapable of more quickly identifying the AAC-class ofbreeding lines in some crosses. However, the utility ofthese markers individually or in combination for cultivardevelopment using non-US germplasm has not beenstudied. The purpose of this research effort was to examinethe association of Waxy sequence variants with the AAC ofrice accessions from many rice-growing regions of theworld. The effects of these sequence polymorphisms alone,and in combinations (as haplotypes), versus environmentaleffects were also studied.

2. Materials and methods

2.1. Materials

A collection of 171 rice accessions were grown in 2000and 2001 under field conditions in Beaumont, TX, usingcultural management practices common for the region.This collection included historical and present-day USAcultivars as well as Asian, European, South American,African, and Australian accessions. Rough rice sampleswere obtained from the National Small Grains Collectionof the US Department of Agriculture, the InternationalRice Research Institute Genebank and the Rice ResearchUnit of the USDA ARS. The accessions were grown insingle plots arranged in blocks. Accessions were assigned toone of three blocks depending on their previouslydocumented days to anthesis. The blocks were planted atdifferent times in order to synchronize anthesis and thusgrain filling of the accessions. The plots consisted of sixrows, 3.5m long, spaced 15 cm apart. Average within-rowspacing was approximately 10 cm. The plants were keptcontinuously flooded after good tiller initiation in about10 cm standing water. Mature seeds were harvested andthreshed by hand, the grains were dehulled, and then allbroken, diseased and immature kernels were removed byhand. Approximately, 50 g of dehulled kernels were milledfor 30 s using a McGill Mill ]1 with an 858-g weight inpositions 12 and 6 for long- and medium-grain types,respectively (Chen and Bergman, 2005). The milled ricewas ground using a Cyclone Sample mill (UDY Corp.,Boulder, CO) and sieved through a 100-mesh screen. Allthe rice accessions were genotyped when first received fromthe original sources using the whole kernel alkali DNAextraction method (Bergman et al., 2001) and RM190microsatellite CTn markers (Ayres et al., 1997). The grainsof each accession collected from both years were verifiedwith the same methods to ensure the correct materials werecarried out through planting, harvesting and grainchemical analysis. Multiple plots of control cultivars,Lemont (7–15 plots/block) and Bengal (1–3 plots/block),were planted at random interspersed among the diverse rice

ARTICLE IN PRESSM.-H. Chen et al. / Journal of Cereal Science 47 (2008) 536–545538

accessions and harvested at maturity. Since the blockswere planted at different times and the days-to-anthesiswas invariant, the control cultivars were exposed todifferent environments during grain development. Thesesamples were evaluated on how the air temperature affectsthe AAC.

The heading and harvesting dates of each plot wererecorded. The averages of maximum and minimum airtemperatures from heading to harvest were calculated foreach plot.

2.2. Chemical analysis

All chemicals used were ACS reagent grade andpurchased from Sigma-Aldrich (St. Louis, MO) unlessotherwise specified. The AAC was determined using themodified iodine spectrophotometric method of Perezand Juliano (1978) with a continuous-flow analyzer(AutoAnalyzer 3 Seal Analytical, Mequon, WI) (Webb,1972). The analyses were performed with two replications.The data is expressed on a 12% (w/w) moisture basis.

2.3. Genotyping methods

The genomic DNA of each accession was extracted usingthe CTAB extraction procedure of Fjellstrom et al. (2004)from leaf material collected from the field plots grownin 2000.

The G/T polymorphism in intron 1 of the Waxy gene(In1G and In1T alleles) was genotyped by restrictionenzyme digest of a polymerase chain reaction (PCR)fragment generated from this genomic region. Specifically,the DNA fragment containing the polymorphic sequenceof In1 SNP was amplified from genomic DNA usingforward primer RM-190F (50-CTTTGTCTATCTCAA-GACAC-30) and reverse primer GBSS-W2R (50-TTT-CCAGCCCAACACCTTAC-30) (Ayres et al., 1997). ThePCR reaction was as follows: first, DNA denatured at94 1C for 5min; then amplified for 35 cycles of 94 1C for45 s, 54 1C for 45 s, and 72 1C for 1min; then extended at72 1C for 7min. The Tfl DNA polymerase (MasterAmpTM

Tfl DNA polymerase, Epicentre Technologies, Madison,WI) was used. The PCR reactions were performed in a20-ml volume containing 0.2mM dNTPs, 2.5mM MgCl2,0.1 mM of forward and reverse primers, approximately 5 nggenomic DNA, and 0.8U Tfl polymerase. After PCR, 2 mlof the amplification product was digested with 2U of therestriction enzyme AccI (New England BioLabs) in 15 mltotal volume for 1 h, electrophoresed on a 2% agarose gelin 1� TAE buffer at 100V (AMRESCO Agarose SFSmall Fragment Isolation, E950-100G), and visualized bystaining with 1� Gel Star (CAMBREX Bio ScienceRockland Inc., Rockland, ME). Each restriction-enzymedigested experimental sample was compared alongsidea buffer-only digest of cognate control. To verify theefficiency of the restriction enzyme activity, a checksample known to be cleaved by AccI was included with

each batch of samples subjected to digest. This set ofprimers amplified through the microsatellite CTn motif inexon 1. The size of the amplified PCR product, forexample, for rice accessions containing microsatellite CT20

allele is 257 bp.For genotyping the RM190 microsatellite CTn alleles,

the two primers used were RM-190F (50-CTTTGTCT-ATCTCAAGACAC-30) and RM-190R (50-TTGCAGA-TGTTCTTCCTGATG-30) (Ayres et al., 1997; Temnykhet al., 2000). The PCR reaction was: first, DNA denaturedat 94 1C for 4min; then, followed by 30 cycles of ampli-fication at 94 1C for 45 s, 55 1C for 45 s, and 72 1C for 1min;and then extended at 72 1C for 5min. The PCR reactionswere performed in a 20-ml volume containing 0.2mMdNTPs, 2.5mM MgCl2, 0.1 mM of forward and reverseprimers, approximately 5 ng genomic DNA, and 0.8UTfl polymerase. The fluorescently labeled amplificationproduct was analyzed by a capillary electrophoresisinstrument (3100 Genetic Analyzer, ABI) to genotype.Three control samples were included for each 96-well platerun. Each comprises a mixture of rice genomic DNA ofknown CTn: CTn of 11+17+20, CTn of 8+16+19, andCTn of 10+14+18.A dideoxyfingerprinting method was designed to identify

sequence variants in the region of intron 5 to intron 6 ofthe Waxy gene. Rice genomic DNA was the template forPCR amplification of Waxy gene sequences. The forwardprimer was 21633-1F (50-GATGGTTGGAAGCATCA-CGAG-30), and the reverse primer was 21633-1Rb(50-TTGGTATGGCAAGAACAAGC-30). The PCR reac-tion was as follows: DNA denaturation at 94 1C for 5min;followed by 30 amplification cycles of 94 1C for 45 s, 62 1Cfor 45 s, and 72 1C for 1min; and then extension at 72 1Cfor 7min. The PCR reactions were performed in a 20-mlvolume containing 0.2mM dNTPs, 2.5mM MgCl2, 0.1 mMof forward and reverse primers, approximately 5 nggenomic DNA, and 0.8U Tfl polymerase. The amplifiedPCR product was then diluted 1:5 with water, and 1 mlof it was used as a DNA template for dideoxy-GTPtermination reaction with an IRD800 labeled reverseprimer (50-AGTCGTTGCAGACGAACACAAC-30). Eachtermination reaction contained: 0.65 pmol IRD-labeledprimer, 2.7U of Thermo Sequenase DNA Polymerase,3 ml ddCGTermination Mix (Thermo Sequenase CycleSequencing Kit, USB Corp.), 0.1� reaction buffer(Thermo Sequenase Cycle Sequencing Kit, USB Corp.).The thermocycle of the termination reaction was: dena-turation at 95 1C for 2min; then, 30 amplification cycles of95 1C for 30 s, 57 1C for 30 s, and 72 1C for 45 s; and thenextension at 72 1C for 5min. After the terminationreaction, 3 ml of IR2 stop solution (Li-Cor, Inc.) was added.The mixture was denatured for 3min at 92 1C andelectrophoresed on a 0.5� MDE gel (CAMBREX BioScience Rockland Inc., Rockland, ME). The electropho-resis and detection of the labeled bands was performed ona Li-Cor IR2 DNA Sequencer (Li-Cor, Inc.). The runsetting of the Licor system was 2000V, current of 25A,

ARTICLE IN PRESSM.-H. Chen et al. / Journal of Cereal Science 47 (2008) 536–545 539

power of 12W and run time of 16 h. The running bufferwas 0.6� TBE.

2.4. Statistics

All statistical analyses were performed using SAS version7 (SAS Institute Inc., 1999). The PROC GLM was used foranalysis of variance determination, and mean separationanalysis was carried out using the Ryan–Einot–Gabriel–Welsch multiple range test. The PROC CORR was used toperform Pearson’s Product Moment Correlation analysis.The PROC MEANS was used to perform two-samplepaired t-test for means.

3. Results

3.1. Origin of rice accessions

Table 1 lists the geographic origin of the 171 riceaccessions used for this study, coming from 43 differentcountries. Roughly equal numbers of rice accessions fellinto the low, intermediate and high AAC-classes. Acces-sions in each of these AAC-classes came from approxi-mately 20 countries. The countries with the top five largestnumbers of accessions were USA, China, the Philippines,India, and Japan. The origins of the glutinous types werethe USA, China, Japan, Korea, Taiwan, India, Laos, thePhilippines, and Thailand.

3.2. AAC and RM190 variation

Nine CTn alleles of the Waxy microsatellite RM190 wereidentified from the 171 rice accessions (Fig. 1). For the non-glutinous accessions (AAC45%), CT8, CT10, and CT11

alleles were associated with high AAC varieties(AACX23%), and CT14 and CT16 were intermediateAAC-types (18%oAACo23%). The CTn of 17, 18, and19 were found primarily in low AAC-types (AACp18%)and a few intermediate AAC varieties. The CT20 allele wasfound predominantly in intermediate and a few high AACrice varieties. The RM190 alleles explained 75.6% of thevariation in AAC of the 152 non-glutinous accessions.Among the 15 glutinous accessions (AACo5%), 11 had

Table 1

Geographic distributions of rice accessions

Amylose class (AACa, %) Africa Australia Europe Americas

Northern Central Sou

High (o23) 2 0 1 6 2 3

Intermediate (19–23) 5 2 0 27 3 2

Low (5–18) 1 2 9 25 1 2

Glutinous (45) 0 0 0 2 0 0

Total 8 4 10 60 6 7

aAAC is apparent amylose content.

the CT17 allele, and the remainder had either a CTn of 8,16, 19, or 20.

3.3. AAC and Waxy In1 SNP variation

For the non-glutinous rices, 103 rice accessions con-tained an In1G allele, and their AAC ranged from 16.6%to 26.0% (Fig. 2). The accessions’ AAC values in Fig. 2 arethe average of the 2 years. The other 53 rice accessions hadan In1T allele and their AAC ranged from 8.0% to 19.7%.Among the glutinous varieties, 12 had an In1T and 3 hadan In1G allele. Analysis of variance demonstrated thatboth the In1 SNP and the growing year had a significanteffect (Po0.0001) on AAC among the non-glutinous riceaccessions. The In1 SNP explained 76.6% of the variationin AAC. The dividing point in AAC between the In1G andIn1T alleles was approximately 18%.

3.4. AAC and Waxy Ex6 SNP variation

The dideoxyfingerprinting (ddF) method was applied toassay the DNA sequence variation around Waxy exon 6.The principle of the ddF method is to electrophorese thedenatured, single-stranded, dye-labeled DNA on a non-denaturing gel. Sequence variation within a dye-labeledamplification product results in folding and confor-mational changes differing from the control sample(‘Nipponbare’) thus yielding a display of multiple single-stranded, conformation-polymorphic bands (SSCPs)compared with the control. In addition to the use ofSSCPs as one way to genotype sequence variation, thechoice of dideoxyguanine as the dye terminator labeling thenon-sense strand enables the identification of cytosinemutations in the sense strand. When compared to a controlsequence, any mutations of cytosine will result in missingdye-labeled bands, and vice versa, if another nucleotideis mutated to cytosine then additional bands would bedetected (Gasser et al., 1998; Liu and Sommer, 1994).We determined there were three nucleotide changes near

exon 6 by examining the alignments of the nucleotidesequences of four rice Waxy gene sequences deposited inthe NCBI GenBank (http://www.ncbi.nlm.nih.gov/Genbank):Nipponbare (gi:55775016), Lemont (gi:7798550), Rexmont

Asia Total Countries

thern Eastern Southern Southeastern Southwestern

11 14 5 2 46 20

1 6 7 1 54 21

11 1 4 0 56 16

8 1 4 0 15 10

31 22 20 3 171 43

ARTICLE IN PRESS

0 4 8 12 0 4 8 12 0 4 8 120 4 8 120 4 8 120 4 8 120 4 8 120 4 8 120 4 8 12

Appare

nt A

mylo

se C

onte

nt (%

)

0

2

4

6

8

10

12

14

16

18

20

22

24

26

CT8 CT10 CT11 CT14 CT16 CT17 CT18 CT19 CT20

No. of Rice Accessions

Fig. 1. Frequency distributions of rice accessions’ AAC in association with the alleles of RM190 microsatellite CTn of Waxy gene. The dotted lines

provide information delineating the categories of high (423%), intermediate (19–23%), low (5–18%), and glutinous- (o5%) AAC-classes.

M.-H. Chen et al. / Journal of Cereal Science 47 (2008) 536–545540

(gi:8101496), and Jodon (gi:8101498). PCR primers weredesigned to amplify this DNA region including the 30-endportion of intron 5, the entire exon 6, and the 50-endportion of the intron 6. One of the sequence variants was inthe coding region of the exon 6 (Ex6 SNP) (relative tonucleotide sequence of Lemont2389), an adenine (Ex6Aallele) to cytosine (Ex6C allele) transversion that alteredthe amino acid coding sequence from serine to tyrosine.Another sequence variant was in intron 5 (In5 InDel),specifically, a cytosine nucleotide deletion relative to thenucleotide of Lemont2322 (In5C and In5Del alleles). Thethird sequence variant was a cytosine/thymine transition inintron 6 (In6C and In6T alleles) relative to the nucleotideof Lemont2400. These three sequence changes span a78-nucleotide segment which we genotyped in all theaccessions. The results of the ddF banding patterns wereevaluated based on the presence and absence of bands andthe mobility-shift-patterns of the bands relative to theNipponbare (NPBR) or Jodon (these two have the samebanding patterns), Lemont, and Rexmont controls. A totalof four ddF patterns were identified among all the riceaccessions. Three of the ddF patterns were the same asthose of Nipponbare, Lemont and Rexmont, and thefourth one had an identical mobility-shift banding patternas the Rexmont’s except that it had one extra labeled bandat In6 SNP. Based on the ddF banding patterns andavailable sequence information of the four cultivars, wewere able to determine the nucleotide at each sequencevariation site for each rice accession.

Neither In5 SNP nor In6 InDel discriminated AAC-classes. The Ex6C allele associated with the intermediateAAC-class and the Ex6A allele associated with low andhigh AAC-classes (Fig. 3). The strength of the associationsamong In5, Ex6, and In6 was evaluated by the Cramer’s V

statistic. The Cramer’s V statistic values range from �1to 1, with values further away from 0 indicating thepresence of a relatively strong association. The strongestassociation was between In5 and In6 polymorphic sites(Cramer’s V value ¼ 0.78), and the lowest association wasbetween Ex6 and In6 SNPs (Cramer’s V value ¼ �0.37).

3.5. AAC and Waxy gene haplotypes

By splitting the RM190 CTn alleles between the twopossible In1 allele present within them, 12 differentRM190-In1 haplotypes were identified (Fig. 4). Thehaplotypes 8G (having 8 CT repeats in RM190 and In1Gallele), 10G, and 11G are associated with the highAAC-types; haplotypes 14G, 16G, 17G, 18G, 19G, and20G are found in the intermediate AAC-types; andhaplotypes 17T, 18T, and 19T are in the low AAC-types.The amount of variation in AAC explained by thesehaplotypes was 86.4%. Haplotypes 18T and 20G hadespecially high variation in AAC compared to the otherhaplotypes. A few rice accessions with 20G hadAAC424.5% and were classified as high AAC-type.The combination of the In1 and the Ex6 SNPs generated

four haplotypes: In1G-Ex6C, In1G-Ex6A, In1T-Ex6A,

ARTICLE IN PRESS

0 5 10 15 20 0 5 10 15 20

0

2

4

6

8

10

12

14

16

18

20

22

24

26

No. of Rice Accessions

Appare

nt A

mylo

se C

onte

nt (%

)

allele G allele T

Fig. 2. Frequency distributions of rice accessions’ AAC in association with the alleles G and T of In1 SNP of Waxy gene. The dotted lines provide

information delineating the categories of high, intermediate, low, and glutinous AAC-classes.

M.-H. Chen et al. / Journal of Cereal Science 47 (2008) 536–545 541

and In1T-Ex6C. Haplotype In1T-Ex6C (n ¼ 6) was onlyfound in glutinous rices. Other glutinous accessionscontained haplotypes In1G-Ex6C (n ¼ 2) and In1G-Ex6A(n ¼ 1). For the non-glutinous accessions, the haplotypesIn1G-Ex6A, In1G-Ex6C, and In1T-Ex6A associatedwith high, intermediate, and low AAC rices, respectively(Fig. 5). The haplotypes explained 86.7% of the variationin AAC.

3.6. Environmental effects on AAC of Waxy gene

haplotypes

A paired t-test (difference of AAC ¼ AAC of2000�AAC of 2001) was conducted to examine howenvironment may have affected AAC of each haplotype(Table 2). The results demonstrated that there was asignificant difference in AAC between the 2 years(Po0.0001), and the t-values were 4.6, �11.8, and �8.8for haplotypes In1G-Ex6A, In1G-Ex6C, and In1T-Ex6A,

respectively. The high AAC-haplotype had higher AACin 2000 than in 2001; while the intermediate and lowAAC-haplotypes had lower AAC in 2000. The environ-ment had an opposite effect on the high AAC-haplotype(In1G-Ex6A) compared to the intermediate (In1G-Ex6C)and low (In1T-Ex6A) AAC-haplotypes.Each haplotype group was sub-divided by year, and the

mean AAC of sub-classes compared (Table 2). The resultsdemonstrate that even though the environment has someinfluence on AAC, the sequence variations of In1 and Ex6have the major effect in determining the AAC variationunder the conditions of this study.During the grain filling period, both the maximum and

minimum average daily air temperatures were higher in2000 (33.1 and 20.1 1C) than in 2001 (29.6 and 18.5 1C). Foreach of the control cultivars, the AAC of the Lemont andBengal plots across all blocks in the 2 years was highlycorrelated with the average maximum air temperature(r ¼ �0.90 and �0.93, respectively). These results suggest

ARTICLE IN PRESS

0

0

2

4

6

8

10

10

12

14

16

18

20

20 0 10 20

22

24

26

No. of Rice Accessions

allele C

Ap

pa

ren

t A

mylo

se C

onte

nt

(%)

allele A

Fig. 3. Frequency distributions of rice accessions’ AAC in association

with the alleles A and C of Ex6 SNPs. The dotted lines provide

information delineating the categories of high, intermediate, low, and

glutinous AAC-classes.

M.-H. Chen et al. / Journal of Cereal Science 47 (2008) 536–545542

that the 81% and 86% variations in AAC of Lemont andBengal, respectively, were largely explained by maximumair temperature during grain development. Lemont is anintermediate AAC-type rice with a Waxy haplotype ofIn1G-Ex6C. Bengal is a low AAC Waxy In1T-Ex6Ahaplotype.

4. Discussion

The RM190 microsatellite CTn alleles in the Waxy genewere first identified by Bligh et al. (1995). Eight RM190CTn alleles were reported by Ayres et al. (1997) and one inaddition to this by Bergman et al. (2000). Most of theaccessions studied in the latter study were included in thepresent report. Thus, it was not surprising that no addi-tional RM190 alleles were identified. Studying primarily

Thai accessions, Prathepha (2003) reported two additionalalleles thus bringing the total RM190 CTn identified todate to 11.The association of AAC with the Waxy RM190 CTn

alleles in the present study was lower than that reported byAyres et al. (1997), but more than Bergman et al. (2000)found. This can be understood by considering that theAyres et al. (1997) study focused on a genetically narrowset of USA accessions, while the Bergman et al. (2000)materials contained greater genetic diversity. The presentstudy contained most of the accessions Bergman et al.(2000) evaluated and had a similar degree of geneticdiversity. The variation in AAC explained by the RM190alleles was greater in the present study compared toBergman et al. (2000) by almost 10 percentage points. Theaccessions in the present study were grown under condi-tions designed to control environmental effects. The knowninteraction between genetics and temperature on AAC(Asaoka et al., 1984, 1985; Resurreccion et al., 1977) waslimited by the experimental design used in this study whichcaused all accessions to pass through anthesis within a fewdays of each other. The importance of this experimentaldesign in eliminating the environmental effect and thusmaximizing the detection of genetic variation was reflectedin the AAC of the control cultivars. Lemont and Bengalwere planted in multiple plots in each block in 2 years.The planting dates of each block were different, so the plotsin each block were exposed to different environmentsduring grain filling. The results demonstrated that thevariation in AAC explained by the maximum air tempera-ture was 481%.Ayres et al. (1997) and Bergman et al. (2001) suggested

the RM190 microsatellite was suitable for use as a markerfor AAC during the breeding process. Unlike the studiesfocused only on USA germplasm, in the present studyevaluating diverse germplasm the individual RM190 allelesdo not associate with only one AAC-class. For example, inour collection, the rice accessions containing the CT17,CT18, and CT19 alleles were associated with the low andintermediate AAC-types; whereas in most of the USAgermplasm, these alleles were found only in low AAC-types(Ayres et al., 1997; Bergman et al., 2001).Although the CT17, CT18, and CT19 alleles cannot

discriminate between the low and intermediate AAC-typerices, the Waxy In1 SNP alleles can. All the low AAC riceaccessions, excluding glutinous rice, have an In1T alleleand were associated only with the RM190 CTn alleles 17,18, and 19. The In1 SNP divided the low AAC-class fromthe intermediate and high AAC-classes at approximately18% AAC in both USA germplasm (Ayres et al., 1997)and in this international rice collection. However, thisIn1 SNP is unable to discriminate between intermediateand high AAC-classes, since both classes have an In1Gallele.The sequence analysis of Waxy exon 6 identified one

single nucleotide change from adenine to cytosine thatresulted in an amino acid change in GBSS from serine to

ARTICLE IN PRESS

0

5

10

15

20

25

30

8G 10G 11G 14G 16G 17G 18G 19G 20G 17T 18T 19T

R2 = 86.4%

Haplotypes of (CTn–In1 allele)

Ap

pa

ren

t A

mylo

se

Co

nte

nt (%

)

Fig. 4. Association of AAC of non-glutinous rice accessions with the Waxy gene haplotypes, the combination of the alleles of RM190 microsatellite (CTn)

and In1 SNP. The n ¼ 8, 10, 11, 14, 16, 17, 18, 19, and 20. The alleles of In1 SNP are G and T. The dotted lines provide information delineating the

categories of high, intermediate, and low AAC-classes. The R2 value is the percent variance explained by the haplotypes. The box plot elements: the box

displays the AAC of accessions between the 25th and 75th percentiles; the horizontal line inside the box is the AAC at the median; the horizontal lines

above and below the box are the AAC at 90th and 10th percentiles, respectively; the closed circles above and below the box are rice accessions beyond and

below 90th and 10th percentiles, respectively.

0

5

10

15

20

25

30

Ap

pa

ren

t A

mylo

se

Co

nte

nt

(%)

G-A G-C T-A

Haplotypes of (In1 allele - Ex6 allele)

R2 = 86.7%

Fig. 5. Association of AAC of non-glutinous rice accessions with the

Waxy gene haplotypes, the combination of alleles of In1 and Ex6 SNPs.

The alleles of In1 SNP are G and T. The alleles of Ex6 SNP are A and C.

The dotted lines provide information delineating the categories of high,

intermediate, and low AAC-classes. The R2 value is the percent variance

explained by the haplotypes. For explanation of box plot elements, see

Fig. 4.

M.-H. Chen et al. / Journal of Cereal Science 47 (2008) 536–545 543

tyrosine (Larkin and Park, 2003). This A-C sequencechange was found in all 13 intermediate AAC riceaccessions studied including ‘Lemont’, a tropical japonica

rice. The high AAC rice cultivars Jodon and L202 have thenucleotide adenine at this Ex6 SNP. In the study of USAcultivars, the CT20 microsatellite allele associated with boththe intermediate AAC- and some high AAC-types. Thus,this Waxy Ex6 SNP has been suggested to be a markercapable of discriminating intermediate AAC-types fromhigh AAC-types when both parents have CT20 alleles(Larkin and Park, 2003).

Two Waxy gene alleles, Wxa and Wxb, have traditionallybeen associated with the contents of GBSS and AAC in riceendosperm, with the Wxa allele synthesizing highercontents of GBSS, and thus AAC, than the Wxb allele(Sano, 1984; Sano et al., 1985). Hirano et al. (1998)demonstrated that Wxa alleles from O. sativa indica have aG at the In1 SNP site and Wxb alleles from O. sativa

japonica have a T at the In1 SNP site. We examined theeffect of the Ex6 SNP on AAC for separating riceaccessions having In1G and In1T alleles, thus, Wxa andWxb alleles, respectively. For all the accessions with anIn1G allele, the Ex6A allele associated with the highAAC-class and Ex6C allele associated with the intermedi-ate AAC-class. This suggested that the sequence changeof A-C in Ex6 SNP might affect the GBSS enzyme at apost-translational level (e.g., enzyme activity or stability),thus decreasing amylose synthesis. The effect of the Ex6SNP on rice accessions having an In1T allele was difficultto assess. Even though the haplotype In1T-Ex6Awas associated with the low AAC-type and the haplotypeIn1T-Ex6C was found only in glutinous rice in thiscollection, sequence analysis has demonstrated that theglutinous rice contained a pre-mature termination codon inwaxy gene (Wanchana et al., 2003). Thus, the reasonglutinous rice is glutinous may be due to a nucleotidesequence mutation resulting in the synthesis of non-active,truncated-GBSS rather than being due to the Ex6C allele.The DNA sequence analysis of the coding region of thewaxy mutant accessions would be required to prove thispoint. More studies will be needed to understand how thisEx6 SNP affects amylose synthesis.The three haplotypes defined by combining the two

SNPs, In1 and Ex6, were able to discriminate allthree amylose classes from each other, and explained86.7% variation in AAC of the non-glutinous accessions.

ARTICLE IN PRESS

Table 2

Statistical analyses of the variation in AAC (%) of the three Waxy gene haplotypes in two growing seasons (2000 and 2001) of the non-glutinous rice

accessions

Haplotypea No of accessions Year Air temperature ( 1C)b Paired t-test valuec AAC7SDd (%)

Max Min

In1G-Ex6A 47 2000 33.1 20.1 4.61*** 24.770.8a

2001 29.6 18.5 24.170.8a

In1G-Ex6C 56 2000 33.1 20.1 �11.80*** 20.271.2c

2001 29.6 18.5 21.870.9b

In1T-Ex6A 53 2000 33.1 20.1 �8.75*** 13.671.8e

2001 29.6 18.5 15.871.9d

aCombination of In1G/T alleles and Ex6A/C alleles of Waxy gene.bThe maximum (Max.) and minimum (Min.) air temperatures are the average of maximum and minimum air temperatures, respectively, from heading

to harvest of each plot.cTest of the mean difference in AAC of (AAC of 2000�AAC of 2001); ***Po0.0001.dMean comparison of AAC (%) is determined by Ryan–Einot–Gabriel–Welsch multiple range test. Different letters indicate significant difference at

a ¼ 0.05.

M.-H. Chen et al. / Journal of Cereal Science 47 (2008) 536–545544

These haplotypes appear to be suitable for use in selectinglines with the desired AAC in breeding programs, nomatter the source of the germplasm. Alternatively, the12 haplotypes of RM190 combined with the In1 SNP,that explained 86.4% of variation in AAC, could alsobe used.

The environment had a significant effect on the AAC(Po0.001). Low AAC-type rice accessions with In1T allelehave been documented to accumulate a lower amount ofmature Waxy transcripts and AAC when grain fillingoccurs at higher temperatures (Larkin and Park, 1999). Thehaplotype In1T-Ex6A of this diverse germplasm alsoshowed the effect by environment, the mean AAC wassignificantly higher in 2001 than 2000. The averagemaximum and minimum daily air temperatures in 2000were higher than in 2001 during grain filling period. TheAAC of the control cultivars, Lemont and Bengal, whichwere grown multiple plots across all blocks in 2 years, werehighly, negatively correlated with the average maximum airtemperature (r ¼ �0.90 and �0.93, respectively). Thissuggested that the maximum air temperature might bethe key environmental factor contributing to much of thevariation in AAC between the 2 years. The highertemperature probably contributed to the lower AAC inhaplotype In1T-Ex6A. Interestingly, we found the inter-mediate AAC-type rice containing haplotype In1G-Ex6Cwas also influenced by the environment in a similar patternas the low-AAC haplotype (In1T-Ex6A): high temperatureapparently resulting in a lower accumulation of AAC in themature grain. Unexpectedly, the paired t-statistics revealeda significant difference in AAC change between the 2 yearsof the high AAC haplotype (In1G-Ex6A). The effect by theenvironment is the reverse of the two other haplotypes:with In1G-Ex6A the higher temperature apparentlyresulted in higher AAC. Further studies are required toconfirm and understand how temperature influences AACin these haplotypes.

The glutinous rice accessions were neither linked toneither the SNP polymorphisms of In1 or Ex6, nor to theRM190 alleles. This is not surprising because the sequenceinformation of waxy mutants, both the naturally selectedand chemical-mutagenized waxy mutants, reported thus farcontain a premature translation stop codon in variousexons. A 23-bp duplicated sequence in exon 2 of the Waxy

gene has been found in several tropical glutinous riceaccessions. This duplication creates a premature transla-tion stop codon at 78 bp downstream of the repeatedsequence. This prevents the formation of a functionalGBSS protein (Wanchana et al., 2003). The glutinous riceEM21 mutant from a Kinmaze background had a sequencemutation in exon 7 and generated a premature translationstop codon downstream, with no other nucleotide sequencechanges identified between the mutant line and theKinmaze (Isshiki et al., 2001). In our study, 12 out of 15glutinous accessions are T variants at the In1 SNP, and 11out of 15 have RM190 CT17 alleles. Bao et al. (2002) alsofound the CT17 to be the most common waxy microsatellitein a collection of 56 glutinous rice accessions. Anotherstudy (Wanchana et al., 2003) evaluated 24 glutinous riceaccessions and found them to have the CT17 or CT18

RM190 alleles and a T variant at the In1 SNP, which arethe typical nucleotide sequences present within thetemperate japonica accessions. The higher frequency ofglutinous rice derived from temperate japonica accessionscould possibly be due to the preferential selection for itsend-use quality by many Asian people during thedomestication process of rice. It would be interesting toknow whether this 23-bp duplicated sequence is alsopresent in the glutinous indica rice accessions.In conclusion, three haplotypes of the Waxy gene, formed

by combining two single-nucleotide-polymorphisms (SNP)were examined in a genetically diverse germplasm collec-tion. One SNP is a guanine-thymine sequence change atthe leader intron splice site (In1 SNP) and the second SNP

ARTICLE IN PRESSM.-H. Chen et al. / Journal of Cereal Science 47 (2008) 536–545 545

is in exon 6. The three haplotypes formed from thesesequence variants explained 86.7% of the variation inAAC and discriminated the low, intermediate, and highAAC-groups of rice from each other. The environmentsignificantly affected all three haplotypes: higher tempera-ture apparently reduced the AAC of the low and inter-mediate AAC-types, but apparently increased AAC in thehigh AAC-types.

Acknowledgments

The authors thank Janis Delgado, Naomi Gipson, EricChristensen, and Piper Roberts for their excellent technicalsupport. This research was partially funded by the USARice Foundation.

References

Asaoka, M., Okuno, K., Sugimoto, Y., Kawakami, T., Fuwa, H., 1984.

Effect of environmental temperature during development of rice plants

on some properties of endosperm starch. Starch/Starke 36, 186–193.

Asaoka, M., Okuno, K., Fuwa, H., 1985. Effects of environmental

temperatures at the milky stage on amylose content and fine structure

of amylopectin of waxy and nonwaxy endosperm starches of rice

(Oryza sativa L.). Agricultural and Biological Chemistry 49, 373–379.

Ayres, N.M., McClung, A.M., Larkin, P.D., Bligh, H.F.J., Jones, C.A.,

Park, W.D., 1997. Microsatellites and a single-nucleotide polymorph-

ism differentiate apparent amylose classes in an extended pedigree of

US rice germ plasm. Theoretical and Applied Genetics 94, 773–781.

Bao, S., Corke, H., Sun, M., 2002. Microsatellites in starch-synthesizing

genes in relation to starch physicochemical properties in waxy rice

(Oryza sativa L.). Theoretical and Applied Genetics 105, 898–905.

Bergman, C.J., Fjellstrom, R.G., McClung, A.M., 2000. Association

between amylose content and a microsatellite across exotic rice

germplasm. In: Fourth International Rice Genetics Symposium.

International Rice Research Institute, Manila, Philippines.

Bergman, C.J., Delgado, J.T., McClung, A.M., Fjellstrom, R.G., 2001. An

improved method for using a microsatellite in the rice waxy gene to

determine amylose class. Cereal Chemistry 78, 257–260.

Bergman, C.J., Bhattacharya, K.R., Ohtsubo, K., 2004. Rice end-use

quality analysis. In: Champagne, E.T. (Ed.), Rice Chemistry and

Technology. American Association of Cereal Chemists, Inc., St. Paul,

Minnesota.

Bligh, H.F.J., Till, R.I., Jones, C.A., 1995. A microsatellite sequence

closely linked to the Waxy gene of Oryza sativa. Euphytica 86, 83–85.

Bligh, H.F.J., Larkin, P.D., Roach, P.S., Jones, C.A., Fu, H., Park, W.D.,

1998. Use of alternate splice sites in granule-bound starch synthase

mRNA from low-amylose rice varieties. Plant Molecular Biology 38,

407–415.

Bollich, C.N., Webb, B.D., 1973. Inheritance of amylose in two hybrid

populations of rice. Cereal Chemistry 50, 631–636.

Cai, X.-L., Wang, Z.-Y., Xing, Y.-Y., Zhang, J.-L., Hong, M.-M., 1998.

Aberrant splicing of intron 1 leads to the heterogeneous 50 UTR and

decreased expression of waxy gene in rice cultivars of intermediate

amylose content. Plant Journal 14, 459–465.

Chen, M.-H., Bergman, C.J., 2005. Influence of kernel maturity, milling

degree, and milling quality on rice bran phytochemical concentrations.

Cereal Chemistry 82, 4–8.

Fjellstrom, R., Conaway-Bormans, C.A., McClung, A.M., Marchetti,

M.A., Shank, A.R., Park, W.D., 2004. Development of DNA markers

suitable for marker assisted selection of three Pi genes conferring

resistance to multiple Pyricularia grisea pathotypes. Crop Science 44,

1790–1798.

Gasser, R.B., Zhu, X., McManus, D.P., 1998. Dideoxy fingerprinting:

application to the genotoyping of Echinococcus. International Journal

for Parasitology 28, 1775–1779.

Hirano, H.-Y., Eiguchi, M., Sano, Y., 1998. A single base change altered

the regulation of the Waxy gene at the posttranscriptional level during

the domestication of rice. Molecular Biology and Evolution 15,

978–987.

Isshiki, M., Morino, K., Nakajima, M., Okagaki, R.J., Wessler, S.R.,

Izawa, T., Shimamoto, K., 1998. A naturally occurring functional

allele of the rice waxy locus has a GT to TT mutation at the 50 splice

site of the first intron. Plant Journal 15, 133–138.

Isshiki, M., Yamamoto, Y., Satoh, H., Shimamoto, K., 2001. Nonsense-

mediated decay of mutant waxy mRNA in rice. Plant Physiology 125,

1388–1395.

Juliano, B.O., Pascual, C.G., 1980. Quality characteristics of milled rice

grown in different countries. IRRI Research Paper Series 48: 1–25.

International Rice Research Institute, Los Banos, Laguna, Philippines.

Larkin, P.D., Park, W.D., 1999. Transcript accumulation and utilization

of alternate and non-consensus splice sites in rice granule-bound starch

synthase are temperature-sensitive and controlled by a single-nucleo-

tide polymorphism. Plant Molecular Biology 40, 719–727.

Larkin, P.D., Park, W.D., 2003. Association of waxy gene single

nucleotide polymorphisms with starch characteristics in rice (Oryza

sativa L.). Molecular Breeding 12, 335–339.

Liu, Q., Sommer, S.S., 1994. Parameters affecting the sensitivities of

dideoxy fingerprinting and SSCP. PCR Methods and Applications 4,

97–108.

McKenzie, K.S., Rutger, J.N., 1983. Genetic analysis of amylose content,

alkali spreading score and grain dimensions in rice. Crop Science 23,

306–313.

Perez, C.M., Juliano, B.O., 1978. Modification of the simplified amylose

test for milled rice. Starch/Starke 30, 424–426.

Prathepha, P., 2003. Characterization of Waxy microsatellite classes that

are closely linked to the rice Waxy gene and amylose content in Thai

rice germplasm. Songklanakarin Journal of Science and Technology

25, 1–8.

Resurreccion, A.P., Hara, T., Juliano, B.O., Yoshida, S., 1977. Effects of

temperatures during ripening on grain quality of rice. Soil Science and

Plant Nutrition 23, 109–112.

Sano, Y., 1984. Differential regulation of waxy gene expression in rice

endosperm. Theoretical and Applied Genetics 68, 467–473.

Sano, Y., Katsumata, M., Amano, E., 1985. Correlations between the

amounts of amylose and Wx protein in rice endosperm. Sabrao

Journal 17, 121–127.

Smith, A.M., Denyer, K., Martin, C., 1997. The synthesis of the starch

granule. Annual Review of Plant Physiology and Plant Molecular

Biology 48, 67–87.

Temnykh, S., Park, W.D., Ayres, N., Cartinhour, S., Hauck, N., Lipovich,

L., Cho, Y.G., Ishii, T., McCouch, S.R., 2000. Mapping and genome

organization of microsatellite sequences in rice (Oryzan sativa L.).

Theoretical and Applied Genetics 100, 697–712.

Wanchana, S., Toojinda, T., Tragoonrung, S., Vanavichit, A., 2003.

Duplicated coding sequence in the waxy allele of tropical glutinous rice

(Oryza sativa L.). Plant Science 165, 1193–1199.

Wang, Z.-Y., Zheng, F.-Q., Shen, G.-Z., Gao, J.-P., Snustad, D.P., Li,

M.-G., Zhang, J.-L., Hong, M.-M., 1995. The amylose content in rice

endosperm is related to the post-transcriptional regulation of the waxy

gene. Plant Journal 7, 613–622.

Webb, B.D., 1972. An automated system of amylose analysis in whole-

kernel rice. Cereal Science Today 30, 284.