Storage protein profile and amino acid content in wild rice Oryza glumaepatula

The objective of this work was to determine the total protein profile and the contents of the four major protein fractions (albumin, globulin, prolamin and glutelin) and of the amino acids in the endosperm of the rice wild species Oryza glumaepatula. The experiment was performed with 29 accessions of this species, collected from 13 Brazilian locations, and two commercial cultivars. Protein samples were prepared using dried, polished, and ground grains to obtain homogeneous, dry flour used in the preparation of extracts. Oryza glumaepatula accessions were identified with the highest levels of total protein, albumin and glutelin protein fractions, and amino acids (with the exception of tryptophan) in comparison to the two analized rice cultivars. The albumin and glutelin profiles in SDS-Page were distinct between rice cultivars and O. glumaepatula. This wild species has the potential to increase the nutritional quality of rice storage protein through interspecific crosses.


Introduction
The genus Oryza has two cultivated species (Oryza sativa and O. glaberrima) and 21 wild species, which are distributed in tropical and subtropical regions.In Brazil, there are four species of wild rice -O.glumaepatula (AA), O. alta (CCDD), O. grandiglumis (CCDD), and O. latifolia (CCDD) (Rangel et al., 2007).Among these four species, O. glumaepatula has attracted more interest, especially regarding the possibility of its use as a gene reservoir, since it has a diploid genome, which makes feasible interspecific crossing with the cultivated species O. sativa.In addition, it is adapted to Brazilian soil and climate (Brondani et al., 2005).In Brazil, the development of interspecific lines of O. sativa x O. glumaepatula began in 1995.These strains, derived from two backcrosses using rice cultivars as recurrent parents, have some agronomically interesting characteristics, such as fast vegetative growth and increased yield potential (Rangel et al., 2007).
Analysis of encoded proteins by the genome, also known as proteome, is a powerful molecular tool for studying complex biological processes in organelles, cells, organs, and tissues (Komatsu & Tanaka, 2005).
Rice is a model plant for genomic and proteomic studies, due to its small genome compared with other cereals.In recent years, there has been a significant progress in the identification and cataloging of rice proteins.More than thirteen thousand proteins, expressed in different tissues and organelles, have been detected, out of which 5,755 have been classified and had their functions determined (Komatsu, 2006).Information on the rice proteome is available in databases such as the Rice Proteome Database (National Institute of Agrobiological Sciences, 2002).However, there is little information on grain storage proteins in these databases (Komatsu, 2006).
Storage proteins accumulate in large quantities during seed development and are mainly stored in special organelles called protein bodies (Halford & Shewry, 2002).As shown by Kim et al. (2009), the total storage protein content in O. sativa endosperm varies between 4.3 and 18.2%.These storage proteins are divided into four fractions according to their differences in solubility: albumins (soluble in water), globulins (soluble in salts), prolamin (soluble in alcohol), and glutelin (soluble in acidic or basic solutions) (Shotwell & Larkins 1989).Glutelin fraction is the predominant protein in rice endosperm, and is classified according to the molecular weights of α-glutelin (37 kDa) and β-glutelin (20 kDa) (Katsude-Tanaka et al., 2004).
Rice protein is considered to be of good quality because it contains eight out of ten essential amino acids.Compared with other cereals, such as maize and wheat, rice has a high level of lysine, which provides high digestibility and nutritional quality (Huebner et al., 1990).Determining the contents of amino acids in storage proteins is an excellent source of information for planning the development of rice genotypes with higher grain nutritional quality.In Brazil, rice accounts for 14% of the energy and 10% of the protein consumed daily by the population (Naves & Bassinello, 2006).The identification of high-protein O. glumaepatula genotypes, and the development of interspecific lines from this species, could further increase protein content and nutritional quality of rice.
The objective of this work was to determine the total protein profiles and the contents of four major protein fractions (albumin, globulin, prolamin and glutelin) and of the amino acids in the endosperm of the rice wild species O. glumaepatula.

Materials and Methods
Twenty nine accessions of the wild rice species Oryza glumaepatula, collected from five Brazilian states: Goiás (GO), Mato Grosso (MT), Mato Grosso do Sul (MS), Amazonas (AM), and Roraima (RR) were evaluated (Table 1).A total of 200 panicles were collected from each population, with an average of 40 seeds per panicle.Forty seeds of each population were germinated on paper rolls.After one week, seedlings were transplanted to pots and cultivated in greenhouse, in order to increase the availability of seeds, which were stored later in the genebank of Embrapa (Brazilian Agricultural Research Corporation).
Seed samples of each population were analyzed.As controls, the upland rice cultivars BRS Bonança and Primavera, chosen due to their high yield and to the fact that they are widely cultivated in Brazil, were used.The 29 wild accessions were quantitatively and qualitatively analyzed (SDS-Page) for total protein and major protein fractions.Protein samples were prepared using dried, polished grains, which were ground, and defatted with acetone, to obtain a homogeneous, dry flour for the preparation of extracts.Sequential extraction of total proteins and protein fractions was performed as described by Turley & Ching (1986).Protein levels were quantified as described by Bradford (1976).The quantification of total protein and protein fractions was performed in triplicate, using three independent replicates of each accession, in a completely randomized design.The statistical analysis was carried out using the software Genes (Cruz, 2001).After the analysis of variance, the means were compared using the Scott-Knott test, at 5% probability.
For comparison, a calibration curve with BSA (bovine serum albumin) was done at concentrations from 2.5 to 40 μg L -1 .Physicochemical analysis of the amino acids was performed as described by Hagen et al. (1989).
Preparation of total protein samples for the qualitative analysis was made in denaturing polyacrylamide gel, using 20 mg of flour and 350 μL of sample buffer [10 mmol L -1 Tris HCl (pH 6.8), 1% β-mercaptoethanol (v/v), 2% SDS (w/v), 3% glycerol (v/v) and bromophenol blue].Then, samples were boiled for 10 min to denature proteins, in order to run SDS-Page analysis.
For albumin extraction, we used 100 mg of the flour added to 500 μL of 10 mmol L -1 Tris-HCl solution (pH 7.5) and 1 mmol L -1 EDTA.This solution was shaken for one hour and centrifuged (13,000 rpm for 15 min at 4°C).The supernatant (albumin) was collected and precipitated with 1.5 mL of cold acetone, homogenized by inversion, and stored overnight in a freezer at -20°C.The solution was centrifuged at 15°C for 15 min, and the supernatant (acetone) was discarded.The pellet (albumin) was stored in a freezer for subsequent analysis on SDS-Page.For globulin extraction, 500 μL of 10 mmol L -1 Tris-HCl (pH 7.5), 1 mmol L -1 EDTA, and 0.5 mol L -1 NaCl were added to the initial flour.This solution was shaken and centrifuged, and the supernatant (globulin) was collected, precipitated with acetone, homogenized and stored overnight in a freezer.The solution was centrifuged, and the supernatant (acetone) discarded.The pellet (globulin) was stored in a freezer for subsequent analysis on SDS-Page.For prolamin extraction, 500 μL of isopropanol 60% (v/v) was added to the rice flour, as described by Turley & Ching (1986), with modifications.The solution was shaken and centrifuged, and the supernatant (prolamin) was precipitated, homogenized and stored in a freezer overnight.Then, the solution was thawed and centrifuged, and the supernatant (acetone) was discarded.The pellet (prolamin) was stored in a freezer for subsequent analysis on SDS-Page.For glutelin extraction, 700 μL of extraction buffer was added to the flour, as described by Kawakatsu et al. (2008).This solution was shaken vigorously for two hours and centrifuged for 15 min at 4°C.The supernatant (glutelin) was collected in a new tube and precipitated with 1.5 mL of acetone, homogenized by inversion and stored in a freezer overnight.The solution was centrifuged and the supernatant (acetone) was discarded.The pellet (glutelin) was stored in a freezer for subsequent analysis on SDS-Page.
The analyses of total protein and fraction profile were performed in a denaturing polyacrylamide gel (SDS-Page), in a discontinuous buffer system, with 4.5% stacking gel and 14% resolving gel.Gels were subjected to electrophoresis at 80 mA for approximately three hours, using the Rubi SE 600 system (GE Healthcare, Waukesha, WI, USA).Then, gels were stained with coomassie blue and destained in 5% methanol (v/v), 7% acetic acid (v/v) and 88% water.Finally, gels were photographed using a digital camera Sony Cybershot DSC-P200, (Sony Brasil, São Paulo, SP, Brazil).From the images, the protein profiles were compared using the DNA Simdex 3 beta release software program (Scott Archer and GenetX, cid.Jamestown, CO, USA), which determines the molecular weight of each band from the positions of the marker bands at the low-molecular weight pattern using a Amersham (LMW calibration kit for SDS electrophoresis, GE Healthcare São Paulo, SP, Brazil).

Results and Discussion
Significant differences between the accessions of the wild rice species O. glumaepatula were found for total protein and protein fractions (Table 2).Total protein levels were divided into four groups, with ranges from 14.94% (wild genotype BGA14280) to 9.07% (BGA14179).The control cultivars BRS Bonança and Primavera, together with the wild accessions BGA14210, BGA14232, BGA14233, and BGA14179, showed the lowest levels of total protein (Table 2).
Kennedy & Burlingame (2003) analyzed the protein contents of 2,869 genotypes of rice (2,674 O. sativa and 195 O. glaberrima), and found 8.8% as the mean for O. sativa, ranging from 4.5 to 15.9%.The mean score for O. glaberrima was 13.6%, ranging from 10.2 to 15.9%, which was similar to the values found here for the wild species O. glumaepatula.Cao et al. (2009) found a wide variation in the levels of storage protein content (7.38-15.41%) in Chinese varieties of rice (O.sativa).Silveira et al. ( 2010) found a range of 4.4-20.2%for storage protein contents of 550 accessions, in the rice Core Collection of Embrapa, with an average of 10.31%, lower than the average found here (12.17%) for O. glumaepatula.These authors proposed a classification as to the protein contents: high (≥12%), medium (11.9-9%), and low (≤8.9%).From the 29 genotypes evaluated in the present study, 16 genotypes had high-total protein contents and 13 had medium-total protein contents (Table 2), according to this classification.
Oryza glumaepatula accessions showed higher levels of albumin (BGA14210) and glutelin (BGA14208, BGA14197, BGA14137, BGA14207, BGA14205, BGA14202 and 14269) than the two evaluated cultivars.However, no difference was found between the high levels of globulin found in the accession BGA14162 and the ones determined in the 'BRS Bonança'.Similarly, 'Primavera' did not differ from the 12 accessions of O. glumaepatula (Table 2) with the highest prolamin contents.Other studies reported similar results for protein fractions contents in these O. sativa cultivars (Agboola et al., 2005;Cao et al., 2009).These results denote that O.glumaepatula equalize, at least, the highest levels of protein fractions found in cultivated rice.
All accessions of O. glumaepatula showed a differential α-polypeptide of 40 kDa, which was not found in BRS Bonança and Primavera (Figure 1).When SDS-Page was performed with protein fractions, the results were similar to those found in  (1) .
Accessions of O. glumaepatula showed higher levels of all measured amino acids, in comparison with the two cultivars of O. sativa, except for tryptophan.Considering the average, only methionine and tryptophan showed higher values in O. sativa, while isoleucine had the same mean values between the two species.In both species, there was a predominance of glutamic acid and aspartic acid.A relevant fact is that lysine levels were higher in O. glumaepatula.This amino acid is considered to be important for the digestibility of food and for its nutritional quality (Huebner et al., 1990).The accessions of O. glumaepatula that had the highest levels of amino acids were BGA14207 (for seven amino acids), BGA14208 (for eight amino acids), and BGA14280 (for two amino acids) (Table 3).BGA14207 and BGA14208 came from two distinct populations collected in Roraima state, Brazil; while BGA14280 came from Goiás state.These three accessions can be used in rice breeding programs to increase grain nutritional quality.

Conclusions
1. Oryza glumaepatula accessions have high levels of total protein, albumin and glutelin fractions, and amino acids (excepting for tryptophan), equalizing, at least, the highest levels found in the commercial cultivars.
2. Albumin and glutelin profiles are distinct between rice and O. glumaepatula 3. Oryza glumaepatula has potential to increase the nutritional quality of rice storage protein through interspecific crosses.   1) In brackets is the identification of genotype in which the value was found.B, 'BRS Bonanza'; P, 'Primavera'.

Table 1 .
Collecting locations and germplasm storage sites of the 29 Oryza glumaepatula accessions analyzed.

Table 2 .
Contents of storage protein and of protein fractions in 29 Oryza glumaepatula accessions and in two O. sativa cultivars

Table 3 .
Levels of total amino acids (g per 100 g) determined in grains of Oryza glumaepatula (OG) and O. sativa (OS).