Correlation between grain nutritional content and pasting properties of pre-gelatinized red rice flour

Assim como as de qualquer variedade de arroz, as características do arroz-vermelho estão sujeitas a diferenças varietais, condições de cultivo e forma de processamento, propriedades nutricionais e reológicas. Neste trabalho, determinaram-se as características nutricionais (composição centesimal e de minerais) e as propriedades de viscosidade da pasta dos grãos crus de quatro genótipos de arroz-vermelho (Tradicional, MNAPB0405, MNACE0501 e MNACH0501) e as propriedades de viscosidade de pasta de farinhas pré-gelatinizadas, obtidas em diferentes tempos de cocção (20, 30 e 40 min). Foram correlacionadas as principais propriedades nutricionais com as propriedades de pasta das farinhas pré-gelatinizadas. A amostras apresentaram diferenças nutricionais e de viscosidade de pasta. MNAPB0405 e MNACE0501 mostraram maiores teores de fibras e lipídios e forneceram maior energia calórica que Tradicional e MNACH0501, os quais, por sua vez, apresentaram maiores teores de amilose. MNACH0501 apresentou


INTRODUCTION
Red rice, also known as "arroz-de-Veneza" and "arrozda-terra" (Pereira et al., 2010), and in English as weedy rice and red yeast rice, belongs to the same botanical species of the Asian rice (Oryza sativa L.), but in commercial rice crops, it is considered an invasive plant for causing considerable damage, mainly by reducing the quality of the final product (Santos et al., 2007).
Although there is abundant literature on its control, in an attempt to eliminate it from white rice crops, the genetic viability of red rice should be preserved to maintain rice quality and promote its genetic improvement.That is, the white rice actually originated from a mutation that changed the red seed of wild rice into the white seeds of modern rice and has finally become a characteristic of great commercial interest, and therefore a priority in most rice breeding programs (Pereira et al., 2007, Pereira et al., 2008, Pereira et al., 2010).In addition, new red rice genotypes have emerged to supply the demands for products that improve the eating pattern of its consuming population, meeting the concepts of productivity and quality.
Little is known about its nutritional quality, but in Asia and Oceania, the red rice is valued for its antioxidant properties and is used in the preparation of breads, colored pasta, vinegar, alcoholic beverages, drugs and cosmetics (Pantindol et al., 2006).In Brazil, it is cultivated in small areas of semi-arid Northeast, predominantly by family farmers, and it is estimated that the currently acreage does not reach 10,000 hectares (Pereira, 2004); however, its consumption is appreciated for the nutritional value and flavor.
The nutritional composition is considered a key factor affecting the technological properties of rice.During cooking, proteins and lipids can interact with the starch, affecting the viscosity, volume and stickiness of cooked rice (Martin & Fitzgerald, 2002).Proteins form a network of starch-protein that restricts uptake of water during cooking.Lipids tend to form complexes with starch, also preventing water uptake.Rice with high levels of lipids and proteins tends to absorb less water during cooking, resulting in a looser and less sticky rice (Saleh & Meullenet, 2006).
The amylose content is considered one of the most important variables to assess the cooking behavior of rice (Xie et al., 2007), directly affecting the firmness of grain and, conversely, the stickiness of cooked rice.Pereira et al., (2009) observed that the red rice has a lower amylose content than white rice and requires more time for effective cooking, even because it is traditionally consumed as partially milled (or barely polished).
Palavras-chave: Oryza sativa, cereais, propriedades físico-químicas e reológicas.Ascheri et al., (2010) submitted five cultivars of upland rice to different cooking times and found that the maximum viscosity and the retrogradation tendency of the flours prepared depended on the interaction of these factors and were reduced as a function of the cooking time.
Therefore, it is believed that grains of different genotypes of red rice have different properties and produce pre-gelatinized flours also with different properties.Thus, in this study, we determined the nutritional and viscoamylographic properties of four genotypes of red rice in the form of raw and pregelatinized flours obtained at different cooking times and correlated the main nutritional properties with the pasting properties.

MATERIALS AND METHODS
Red rice seeds of genotypes Tradicional (Tr), MNAPB0405 (PB), MNACE0501 (EC) and MNACH0501 (CH) were kindly supplied by the Red Rice Breeding Program -Embrapa Meio-Norte (Teresina/PI).Sowing was carried out in an experimental field with irrigation system belonging to Embrapa Meio-Norte in July 2007 and harvested in November of the same year.
After harvest, 5 kg of paddy rice of each genotype were stored in plastic bags and taken to the Food Analysis Laboratory at Embrapa Arroz e Feijão (Santo Antônio de Goiás-GO) for processing in a Suzuki TM-95 milling machine (Suzuki Co., Sao Paulo, Brazil).Semi-polished grain rice was obtained based on the comparison with a commercial sample (polishing defined at 15s).The degree of polishing was determined by the Satake Milling Meter MM1B.
The analysis of nutritional properties was carried out using semi-polished red rice dried in a FANEM 520 oven (Fanem, Sao Paulo, Brazil) at 105 °C and ground in a hammer mill Perten Laboratory Mill 3100 (Petern Instruments, Huddinge, Sweden) to pass a sieve with 0.63 mm mesh opening.Moisture, ash, protein and lipids were determined according to the methods described by the Association of Official Analytical Chemists (AOAC, 2000).The crude fiber content was determined as described by Gonçalves (1998).
Total carbohydrate content was calculated by subtraction of the sum of the moisture, total fat, ash and crude protein from 100; carbohydrate values include the total dietary fiber (Taco, 2006).The total energy was calculated by multiplying the levels of lipids, proteins and carbohydrates in each sample by their respective caloric values: 9.4 and 4 kcal, according to the conversion factor Atwarter.
The amylose content was determined by the colorimetric method of Martinez & Cuevas (1989), using as indicator the iodine/potassium iodide solution and the formed complex was measured by a spectrophotometer (Femto, 432, São Paulo) at 620 nm.
Ca, Mg, Zn, Cu, Mn, Fe and K were determined using the method of Battaglia et al., (1978).Preliminarily, unbroken red rice grains were washed in distilled water and dried at 50 ºC in an air-forced oven (Nova Ética, 400/ 5ND 300°, São Paulo) for 48 h.Then, the material was ground in a zirconium ball mill (Retsch, MM200, Germany), weighed (500 mg) in parchment paper, using an analytical balance (Kern 410, Germany) and transferred into digestion tubes for nitroperchloric digestion.The elements were quantified by an atomic absorption spectrophotometer (Perkin Elmer, Aanalyst 100, Massachusetts) using the instrumental parameters (lamp wavelength, lamp current and slit width) specific for each element in atomic absorption.Potassium was determined by the intensity of radiation emitted, and readings were held in a flame photometer (D-igimed, DM-61, São Paulo).
The changes that occur in the pasting properties of pre-gelatinized red rice flours were studied as a function of genotype and cooking time.The experiment was arranged in a completely randomized design as a 3 x 4 factorial, with three cooking times (20, 30 and 40 min) and four red rice genotypes (Tr, PB, CE, and CH), with two replicates.
Rice cooking was analyzed by the method described by Bassinello et al., (2004) with modifications to the cooking time.Rice and water were added in an aluminum pan and allowed to boil until before the water dried up, covering the pan and then reducing the gas flame for slow cooking; 150 g of rice and 300 mL of distilled water were used for each cooking time of 20, 30 and 40 min.During the cooking, 75, 150 and 225 mL of water were added to complete these times, respectively, otherwise, the samples would burn and would not be boiled sufficiently.The raw and cooked samples were dried in an Fanem 520 oven (Fanem, Sao Paulo, Brazil) at 45 °C for 16 h, and ground in a hammer mill Perten Laboratory Mill 3100 (Petern Instruments, Huddinge, Sweden) to obtain the red rice flours.
The pasting properties of flour from red rice were obtained with a RVA (Rapid Visco Analyzer 4, Newport Scientific PTY LTD, Sydney, Australia): 3 g of flour (14% wet basis) sieved through a 250 µ mesh were suspended in distilled water to complete 28 g.The viscosity profile was obtained according to the following time/temperature regime: 25 °C for 2 min, heating from 25 to 95 °C at a rate of 14 °C/min; holding the paste at 95 °C for 3 min (heating cycle); cooling from 95 to 25 °C at a rate of 12 °C/min (cooling cycle); and holding the paste at 25 °C for 4 min.From the viscosity profile generated, the following parameters were detected: the temperature peak was detected at the beginning of the increase in flour viscosity; the peak time and peak viscosity were detected at the maximum peak viscosity; and the values for viscosity breakdown and the tendency to retrogradation (setback) were calculated by the difference between peak viscosity and final viscosity (Vfim, detected at the time 20 min of the RVA reading) and hot paste viscosity (Vmin was detected during the cooling cycle), respectively.
The statistical analysis of the experimental data was performed according the method described by Gomes (2000).The measurements (means of three replicates) were expressed as mean value ± standard deviation.The effect of rice genotype (independent variable) on the red rice physicochemical properties and the effects caused by the factor genotype x time on the pasting properties of the flours were examined by the analysis of variance at 5% probability and significant responses were tested using the Tukey's test at 5% probability level.
The relationship between the main nutritional properties and paste viscosity of raw and pre-gelatinized flours was obtained with the degree of correlation, applying the Pearson correlation matrix (Pearson's product-moment correlation coefficient, r) at 5 and 10% probability.The r values were interpreted as follows (Callegari-Jacques 2003): -If 0.00 < r <0.30, weak linear correlation exists; -If 0.30 < r <0.60, moderate linear correlation exists; -If 0.60 < r <0.90, there is a strong linear correlation; The statistical analysis and graphs were performed using the software Statistica 8.0 (Statsoft, 2007).
Tables 1 and 2 shows that the levels of ash, fat, fiber, amylose, energy value, calcium, magnesium, potassium, copper, manganese and iron, differed between genotypes, indicating that the improvement of red rice brought about variation in the grain chemical composition.
The protein contents were similar to those described by FAO (2010) for red rice (7.0%), the Brazilian Table of Food Composition (Taco, 2006), for long grain brown and polished (raw) rice (7.3 and 7.2%, respectively), and Lumen & Chow as cited in Walter et al., (2008) for polished white rice (4.3-18.2%).Juliano & Bechtel (1985) stated that this variation is due to the genetic characteristics of rice and other factors such as soil (nitrogen), environmental conditions (solar radiation and temperature), etc..However, Liu et al., (2005) found, not only difference in the total protein concentration (7.35 to 11.47%) between genotypes of Japanese rice, but also variation in the types of proteins.
The contents of ash and lipids were higher than those reported for polished white rice (Storck et al., 2005;Taco, 2006, Walter et al., 2008) and are close to the contents of ash (1.2%) and lipids (1.9%) of raw brown rice (Taco, 2006), even because red rice is classified as barely polished, and as such, during the polishing part of the bran is not removed, remaining a reddish color which is related to the concentration of phenols, polyphenols, anthocyanins (Pantindol et al., 2006) and other fatty substances.
The red rice genotypes have the same energy value of brown and polished (raw) rices, which provide between 357 and 364 kcal per 100 g grains, respectively (Taco, 2006).However, the Tukey test (Table 1) showed that the genotypes PB and CE had higher fiber and fat contents and provided higher caloric energy than the genotypes Tr and CH.On the other hand, Tr and PB had higher amylose content, followed by the genotypes CE and CH.
The amylose contents are in agreement with those reported by Pereira et al., (2009) for the same red rice genotypes studied.Tr, PB and CE showed intermediate levels of amylose (between 24.1 and 26.1%), which characterizes their grain as dry and loose after cooking, and soft during rewarming.Genotype CH had low amylose content (14.6%) and its grains are characterized as soft, watery and sticky during cooking.Red rice genotypes are characterized by sticky grains and, therefore, they usually have low contents of amylase (Pereira, 2004).But, other factors affect the cooking quality of rice; for example, the distribution of long chains of amylopectin in the starch, climatic and growing conditions.Furthermore, the cooking method of red rice, traditionally done in a large quantity of milk rather than water, can produce a more soggy rice.In this study, we used the optimum amount of water, not in excess, for the cooking.
The red rice genotypes in this study showed different patterns of mineral composition (Table 2).Although the contents of zinc were statistically identical, CE and CH were rich in magnesium and potassium, but had lower calcium contents.CH and Tr had the highest copper and lowest manganese contents, and genotypes CH, CE and PB had iron contents statistically equivalent.
Rev. Ceres, Viçosa, v. 59, n.1, p. 16-24, jan/fev, 2012 However, it has been observed, as in Table 2, that the red rice has approximately two, three and eight times the content of magnesium, potassium and zinc contained in the polished rice, as reported by Juliano (1985) and Taco (2006).Although the genotypes PB and CE had higher manganese contents, the other nutrients (calcium, copper, and iron) had levels that were intermediate to those reported by Juliano (1985) and Taco (2006).These results show that these red rice genotypes provide one or more times the amount of minerals in relation to the white polished rice.
Figure 1 shows the paste viscosity profile obtained from flour of raw grains of the red rice genotypes studied.It also highlights the points of peak viscosity, paste temperature, breakdown and setback.
Figure 1 shows that during the initial heating phase, the viscosity starts to increase because of the swelling of starch and the leaching of amylose and other lower molecular weight polymers, forming a paste.Pasting started between 5 and 6 min at 70-80 ºC.Genotype PB spent less energy (70 ºC) and time (5 min) to form a paste, while the other genotypes spent 6 min, at a temperature of approximately 80 ºC.
Maximum peak viscosity was reached during pasting and occurred between 8 and 9 min at 95 ºC.According to Tsai et al., (1997), this phenomenon occurs when most of the starch granules are fully swollen and other granules and dissolved polymers remain molecularly non-aligned within the field of friction of the Rapid Visco Analyzer (RVA).
Genotype CH formed the highest peak viscosity (2402 cP), faster than the others, with a sharp shape.According to Matuda (2004), the sharp shape of peaks of viscosity indicates the presence of starch granules with expansion and break properties more uniform during the hydrothermal treatment.This fact was not observed in the other genotypes.The flat shape of peak viscosity of genotypes PB, Tr and CE, wich had viscosity values lower than genotype CH, indicates that the starch granules have low resistance to expansion/break during the hydrothermal treatment.
After the peak, genotype CH showed a more precise profile, with a steep decline in viscosity (696.0 cP break), until it reached a viscosity of 1706.0 cP, within 11 min at 81 ºC.However, the profile of genotypes Tradicional and CE formed a platform, indicating a very small decrease in viscosity (70 and 26 cP, respectively).The platform was formed due to remaining starch granules which were not gelatinized during heating and therefore, they have a certain resistance to expansion/break during the hydrothermal treatment.Although we have considered the viscosity peak at 1137 cP for genotype PB, this profile has not stopped risen, reaching a negative value in viscosity breakdown (-117 cP).
During the cooling cycle, pastes of all genotypes showed a similar profile, increasing with time of RVA reading, and the highest retrogradation tendency was recorded for genotype CH.Although the genotypes Tr and CE had the same final viscosity (2281 cP), their retrogradation tendency differed in 94 cP; however, they were higher than that of PB, 1206 cP.
The viscosity profile in Figure 1 shows that different red rice genotypes have different rheological properties.Although the genotype CH has the same viscosity profile of the commercial white rice (Ascheri et al., 2006), their paste viscosity properties differ considerably, and these differences are even greater with the other genotypes.In general, the maximum viscosity peak of the red rice flour, the breakdown and tendency to retrogradation were 1-5, 1-40 and 2-2.5 times lower than those of the commercial white rice.That is, the pastes made with red rice are less sticky, both heated and cooled, which also allows inferring that the cooked grains can absorb less water when hot, and become soft and loose when cooled, in relation to the commercial white rice.
Table 3 shows the pasting properties of pre-gelatinized flours of red rice, as a function of genotype and cooking time and Table 4 shows the results of the variance analyzes.
The statistical analysis (Table 4) shows that genotype CH differs from all others in cooking times, except for viscosity breakdown at 20 and 40 °C.Conversely, in the Table 3, the factor cooking time caused changes in the starch structure of all genotypes of red rice studied, altering significantly the rheological behavior of the pregelatinized flours, reducing the peak viscosity and retrogradation, with irregular behavior of the viscosity breakdown.Pinkrová et al., (2003) and Yu & Wang (2007) also noted this type of viscous behaviour, but for rice starch irradiated in microwave oven and with gamma rays, respectively.
The pastes of pre-gelatinized flours of genotype CH are slimy and have higher setback viscosity than the flours of the other red rice genotypes studied.The peak viscosity, breakdown and setback tendency had values of 948-1725.5,931.5-1640 and 21-47.5 cP, respectively, indicating that the pre-gelatinized flours from cooked grains of genotype CH can absorb more water when hot and become sticky when cold.
The low values of viscosity peak of pre-gelatinized flours of genotypes Tr (127.5-80.5 cP), PB (199.5-88.5 cP) and CE (89-67.5 cP) show that the crystalline structure of the starch granules that remain intact after cooking is resistant to gelatinization and retrogradation, suggesting that these flours can absorb less water when hot and are less sticky when cold.
The relationship between the main nutritional components of red rice, peak viscosity and retrogradation tendency of the raw and pre-gelatinized flours are shown in Table 5.
The lipid content showed strong positive correlation with the fiber content (r = 0.91) and negative correlation with the protein content (r = -0.92).Correlation was also found, although not significant, between the fiber and protein contents (r = -0.75);however, it suggests a decreasing trend between the degree of polishing and the protein content in red rice, ie, the greater the degree of polishing, the greater the fiber removal and hence the greater the protein concentration in red rice.Therefore, the greater the retention of bran in the polished red rice, the higher the lipid content and the lower the protein content.Nuwamanya et al., (2010) found a linear relationship between the fiber content and pasting properties in cassava starch.The increase in fiber content decreases the peak viscosity.In this study, this relationship was not observed; the peak viscosity was not linearly dependent on the fiber content (r = ~ 0).
The positive correlation between the protein content and pasting properties (0.52 < r < 0.65) suggests that proteins may contribute to the increase in viscosity; however, this correlation was not significant, contrary to the study of Moorthy (2000), which showed that a high protein content influences the properties of starchy materials, adversely affecting the pasting properties.
The contents of lipids, fiber and protein did not affect the amylose content of red rice genotypes, with very low correlations of 0.23, -0.09 and -0.58, respectively.
The amylose content is an important parameter to study the changes that occur in the starch pasting properties.In the genotypes studied, the amylose content

Source of variation df
showed inversely proportional correlations with peak viscosity (-0.99 < r < -0.96).Tukomane et al., (2007) reported that starches with low amylose content showed a higher degree of crystallinity, resulting from the reduction of amorphous regions in their granules, and tend to have higher peak viscosity.Sarmento et al., (1999) found a positive correlation between the retrogradation of cassava starch pastes and amylose.Rodríguez-Sandoval et al., (2008) discuss that this phenomenon is caused by the aggregation of amylose, after cooling of the pastes, which acts as a nucleus during the rearrangement of amylopectin.The amylose tends to reassociate, through hydrogen bonds, with other adjacent amylose molecules forming doublehelical crystal structures when the solution cools down and remains intact for a long period of time (Denardin & Smith, 2009) .
Nevertheless, retrogradation not always has this behavior, as it happened in this study; the amylose content and retrogradation tendency of the red rice samples were negatively correlated (-0.99 < r < -0.76), for some cause that has yet to be explained.However, Denardin & Silva (2009) reported that retrogradation is a complex phenomenon and varies with several factors such as temperature, pH, starch source and presence of other components (lipids, electrolytes and sugars).Some studies have shown that as in gelatinization, some populations of amylopectin chains may promote or inhibit the retrogradation of amylopectin because of the greater or lesser formation of double helices during retrogradation (Vandeputte & Delcour, 2004).However, the exact influence of the amylose retrogradation is still obscure.

CONCLUSIONS
The genetic improvement influences the chemical composition of red rice, causing significant variation in the contents of fiber, lipids, amylose and minerals.Red rice is richer in fiber, fat and minerals than the commercial white rice.
The pasting properties of genotype CH indicate that their cooked grains can absorb more water when hot and are sticky when cold, while the genotypes Tr and CE (89-67.5 cP) may absorb less water when hot, and are less sticky when cold.
The protein and amylose contents are correlated with the pasting properties and can be used as indicators of these characteristics for different genotypes of red rice, either for raw grains or processed into pre-gelatinized flours.* Correlation significant at 5% probability level; ¥ correlation significant at 10% probability level; ns non-significant correlation.Lip = lipids, Prot = protein, Fib = fiber; Amyl = amylose, P0, P20, P30 and P40 and R0, R20, R30 and R40 = peak viscosity and retrogradation tendency, respectively, of raw and pre-gelatinized red rice flours cooked for 20, 30 and 40 min.

Figure 1 .
Figure 1.Viscosity profile of pastes obtained from flour of raw grains of different red rice genotypes.

Table 1 .
Centesimal composition, energy value and amylose content in processed grains of different red rice genotypes 1

Table 2 .
Profile of main minerals in processed grains of different red rice genotypes 1 (mg/100g)

Table 3 .
Pasting properties of different red rice genotypes at different times of cooking 1 Means + standard deviation of two replicates.Means followed by different small letters in the columns and different capital letters in the rows are significantly different by the Tukey's test at 5% probability. 1

Table 4 .
Summary of the analysis of variance of pasting properties of pre-gelatinized flours of red rice genotypes (G) obtained at different cooking times (T) MS: mean square, F: F test, *, ** and ns : significant at 5%, 1% and not significant by the F test, respectively.