Availability of Partially Milled Rice As a Daily Source of G-Aminobutyric Acid

Food Sci. Technol. Res., +- (+), .+ῌ.., ,**1

Availability of Partially Milled Rice as a Daily Source of g-Aminobutyric Acid

῍ Keiko IWAKI and Yoshimi KITADA

Faculty of Health Science, Kio University, .ῌ,ῌ, Umaminaka, Koryo-cho, Kitakatsuragi-gun, Nara 0-/ῌ*2-,, Japan

Received January +1, ,**0; Accepted October +*, ,**0

We investigated the use of a high-performance liquid chromatography (HPLC) with o-phthalaldehyde ﬂuorescence detection to simultaneously quantify g-Aminobutyric acid (GABA), glutamate and alanine. Using this method, the column (COSMOSIL /C+2-MS-II, ..0 mm i.d.ῌ+/ cm) was eluted using a linear gradient of acetonitrile and / mM citrate bu#er (pH 0.*). The GABA contents of both milled and cooked rice samples were measured, with a gradual decrease observed as the rice was milled in raw rice. However, GABA content was markedly diminished due to washing in rice milled for a longer period. In less milled rice, the GABA and glutamate contents increased when the rice was cooked without washing. As a result, the GABA content of cooked under-milled rice was twice that of cooked well-milled rice available in the market, and three times greater for under-milled cooked rice without washing. These ﬁndings revealed that partially milled rice could be utilized as a daily source of GABA.

Keywords: g-aminobutyric acid, HPLC with a ﬂuorescence detector, under-milled rice, cooked rice

Introduction ly avoided. However, partially-milled brown rice can ab- GABA is the most widespread inhibitory neurotrans- sorb water over a short period of time and tastes as good mitter (Bronstein and Cummings, ,**+). GABA is ex- as well-milled rice. In ,**/, we reported that the GABA pected to have a hypotensive e#ect on blood pressure and content of rice samples could be measured by HPLC with to be involved in alleviating mental stress. It has been fluorescence detection and that the GABA content of rice shown that the GABA in rice germ could allay anxiety milled for a shorter period was higher than that of types during menopause and in the elderly (Okada et al., ,***). that had been milled for longer. To simultaneously Subjects with high normal blood pressure were given quantify GABA, glutamate, and alanine, we analyzed the fermented milk containing GABA for +, weeks, and it was rice samples using this HPLC method using o-phthalal- demonstrated that their blood pressure was significantly dehyde fluorescence instead of fluorescamine. We found lowered by this treatment (Kajimoto et al., ,**.). Recent- that partially milled rice could be utilized as a daily ly, in Japan, some foods supplemented with GABA have source of GABA. gained popularity. However, GABA is naturally syn- thesized by the action of a decarboxylase on glutamate, Materials and Methods which is reversibly transformed into alanine by a trans- Rice samples Brown rice (Oryza sativa L.; ./* g; Hino- aminase. Rice germ and bran contain glutamate decar- hikari from Nara Prefecture grown in ,**.) was milled for boxylase. Incubation with glutamate resulted in the im- +./, ,, -,and../ min using a milling machine (Tiger Co. mediate production of GABA (Ohtsubo et al., ,***). De- RSD-A), and labeled as under-milled, well-milled +, well- termination of the GABA content of brown rice, ger- milled ,, and over-milled rice, respectively. Rice samples minated brown rice, and germinated unhulled rice re- were stored at .῍C until use. vealed that the GABA content of the germinated types Chemicals o-Phthalaldehyde (OPA) and ,-mercapto- was higher than that of brown rice (Ohisa et al., ,**-). ethanol were purchased from Wako Co. All other chemi- Germinated brown rice, or hatsuga-genmai, has recently cals were of analytical grade. The OPA reagent used for been introduced onto the market. It can be cooked well precolumn derivatization of HPLC was prepared accord- with well-milled rice and has a GABA content that is ing to the method of Goto et al. (+33-). Twenty milli- several times higher than that of well-milled rice. Thus, grams of OPA was dissolved in - ml acetonitrile, and 1 ml it is expected that a daily diet of cooked rice combined *.+ M borate bu#er (pH 3.*) was added. This was kept with +*ῌ hatsuga-genmai can e#ectively increase the under cold storage conditions overnight and filtered GABA intake. Conversely, brown rice also contains through a *../ mmῌ,* mm i.d. RC-membrane filter (Sarto- more GABA than well-milled rice but it needs to be rius AG). soaked in water for more than ,ῌ- hours before boiling, Cooking method The cooking method used in this and its palatability is low, which means that it is general- experiment was in accordance with that reported in a previous paper (Iwaki et al., ,**/, in Japanese). Rice (.0, ῍ To whom correspondence should be addressed. g) was washed thrice with 1** ml water, drained, and E-mail: [email protected] water was again added to give a total weight of ++-, g. 42 K. IWAKI &Y.KITADA

It was cooked in a rice cooker (Tiger Co. IH rice cooker sample was prepared by mixing ,* ml of the extract solu- JKA-A+**HU, Japan) under conditions similar to those tion, ,** ml of the OPA reagent, and ,** mlof,-mer- used for hakumai-futsudaki and was labeled as washed captoethanol solution [/* mlof,-mercaptoethanol in *.+ and cooked rice. Rice that was cooked without washing M borate bu#er (pH 3.*)] in a phial; +* ml of the mixture was labeled “non-washed” and “cooked” rice. After cook- was then immediately injected into the HPLC column ing, a sample of cooked rice was immediately collected (COSMOSIL /C+2-MS-II, ..0 mm i.d.ῌ+/ cm). The method from the central portion of the cooked rice. used a linear acetonitrile gradient (solution A) and / mM Measurement of the moisture content Before cooking, citrate bu#er (pH 0.*) (solution B). This operation was the rice was ground into pieces using a mill (Iwatani Co. performed mechanically (Shimadzu Co. LC-+*AD with IFM-1**G, Japan). Subsequently, / g was weighed, placed RF-+*A, Japan). The mobile phase comprised +/ῌ solu- within an aluminum foil receptacle, and dried at +-/῍Cfor tion A and 2/ῌ solution B at * min, +0./ῌ solution A and - h to measure the moisture content. A similar proce- 2-./ῌ solution B at +/ min, -*ῌ solution A and 1*ῌ dure was followed for the cooked rice sample, except that solution B from -/ to .* min, and +/ῌ solution A and 2/ῌ +* g of cooked rice was weighed into an aluminum foil solution B for +* min. The flow rate and column temper- vessel and wrapped within the aluminum foil. The alu- ature were set to *.2 ml/min and .*῍C, respectively. The minum foil was then unfolded to dry at +-/῍Cfor- h. fluorescence detector was set at Ex. -.* nm and Em. ./* nm. Analyses of several rice characteristics The weight of +*** rice grains was estimated using the method em- Results and Discussion ployed by the National Food Research Institute (Ministry Elution pattern of the rice extract In order to investi- of Agriculture and Forestry, +303). The color of the sur- gate the e#ect of concentration and pH of the citrate face of the rice grains was measured using a spectropho- bu#er on peak height, the concentration and pH were tometer (Nippondenshoku Co. SE ,***, Japan), as reported adjusted to / to ,* mM and /.* to 3.*, respectively. The previously. Rice grains were placed within a transpar- peaks were most pronounced when the concentration of ent acrylic cylinder (-+ mm diameter, +- mm height) to citrate bu#er was / mM; however, the peaks were inde- measure the reflected light (Lab value) when the gauge pendent of the pH. Based on these findings, / mM citrate diameter was -* mm. bu#er (pH 0.*) was used as the mobile phase. The elution Extract preparation Rice powder (/ g) was homoge- patterns of standard GABA (+ mg/ml) and the under- nized (+/*** rpmῌ, min) with ,/ ml of 1/ῌ ethanol in a milled rice extract are shown in Fig. +. The major peaks homogenizer (IKA Works T,/ Basic 2+), and the homo- from the under-milled rice extract were confirmed to genate was filtered through a No. /B filter paper. The correspond to L-glutamate (1.,1 min), L-asparagine (++.1. residue was extracted thrice with ,/ ml of 1/ῌ ethanol, min), L-arginine (+-./3 min), L-alanine (,3.1, min), GABA and the volume of the collected filtrate was made up to (-+.31 min), and L-tyrosine (--./+ min). The contents of +** ml with 1/ῌ ethanol. In the case of cooked rice, the other amino acids were comparatively lower. These approximately +2 ml of 33./ῌ ethanol was added to a +* g results agreed with previously reported results on the sample, which is equivalent to / g of raw rice (Iwaki et al., composition of amino acids in cooked rice (Oryza sativa L., ,**/). The other procedures were the same as those Koshihikari, harvested from Kyoto Prefecture in ,**-) performed in the case of rice powder. that had been obtained using an amino acid analyzer Analysis of GABA content The HPLC conditions (Hitachi Co. L-2/**) (Iwaki, unpublished). L-Glutamate were the same as those reported by Goto et al. (+33-). The and L-alanine are closely related to GABA biosynthesis

Fig. +. Chromatograms of GABA standard (+ mg/ml) (a) and under-milled rice extract (b). Availability of Partially Milled Rice as a Daily Source of g-Aminobutyric Acid 43

Table +. Some characteristics and GABA content of milled rice.

and were measured with GABA in order to plot the fol- lowing calibration curves: L-glutamate (y῍010.1/xῌ ,3.++,R,῍*.333,), L-alanine (y῍1/30/,xῌ/1010,R,῍ *.3323), and GABA (y῍+,+2/**xῌ+3,-0,R,῍*.333.). The limits of detection for GABA and the other amino acids were *.*/ mg/ml and *.+ mg/ml, respectively; whereas, the detection limits of were *.*, mg/ml and *.*/ mg/ml, respectively. Some characteristics and the GABA content of rice samples A few characteristics and the GABA content of four rice samples are shown in Table +. In a previous paper, the milled rice samples were over-milled due to the use of a large milling machine. However, a smaller milling machine was used in this study and the milling yields of the four samples were 3..*ῌ, 3,..ῌ, 3+.-ῌ,and22.2ῌ, respectively. In general, the milling yield of well-milled rice derived from brown rice was approximately 3,ῌ. Further, the samples were labeled as being “under-milled”, “well-milled +”, “well-milled ,”, and “over-milled” rice, while the identically labeled samples reported in the pre- Fig. ,. Correlation between GABA content and weight of vious paper were labeled as “more milled”. +*** grains of Hinohikari harvested in ,**. (῎)and,**- in ῏ The weight of +*** grains was observed to decrease in the previous paper ( ). proportion to the milling yield. The residual amounts of the germs present in all the samples were significantly Table ,. Contents of water and GABA in cooked rice. di#erent. The color of the grain surface indicated that the L-value increased in proportion to the milling yield, whereas the a- and the b-values decreased. The color of rice gradually became brighter and tinged with green. However, it changed to pale yellow in proportion to the milling yield. The GABA content was significantly higher in under-milled than in well-milled + rice, and it was also higher in well-milled + than in over-milled rice. There was no significant di#erence between well-milled + and well-milled , rice. The relationship between GABA than in well-milled + rice, and it was also higher in well- content and the weight of +*** grains is shown in Fig. , milled + than in well-milled , rice. The GABA content of along with additional results from a previous paper. Al- under-milled rice was ,.0 times that of over-milled rice. though the harvest year of the two samples was di#ered, A comparison of the GABA contents of washed and the GABA content was in direct proportion to the weight cooked rice with that of raw rice indicated that it was of +*** grains. This showed that the GABA content barely altered by washing and cooking in under-milled gradually decreased as the rice was milled. rice. Conversely, however, it was reduced to /2ῌ and Water and GABA content of cooked rice The water 0.ῌ in well-milled , and over-milled rice, respectively. and GABA contents of cooked rice are shown in Table ,. On the other hand, the GABA content was considerably Although there was no significant di#erence in the water higher in non-washed and cooked rice than it was in content of cooked rice, the water content of washed, washed and cooked rice. A comparison of the GABA under-milled rice was slightly higher than that of the content in non-washed and cooked rice with that of raw other samples. In washed and cooked rice samples, the rice suggested that it was remained relatively constant in GABA content was significantly higher in under-milled well-milled , and over-milled rice but increased markedly 44 K. IWAKI &Y.KITADA

der-milled rice was twice that of cooked well-milled rice available on the market, that of cooked under-milled rice without washing was three times higher still. These ﬁndings revealed that partially milled rice without washing could be utilized as a daily source of GABA.

Acknowledgements This study was partly supported by Tiger Co., Osaka.

References Bronstein Y.I. and Cummings J.L. (,**+). Neurochemistry of Fron- Fig. -. Contents of GABA, glutamate and alanine in raw tal ῌSubcortical Circuits. In “Frontal-Subcortical Circuits in rice and non-washed cooked rice. Psychiatric and Neurological Disorders,” ed. by Lichter D.G. and Cummings J.L., The Guilford Press, New York and Lon- don, 0-ῌ00. in under-milled and well-milled + rice. This suggested Goto T., Horie H. and Muakai T. (+33-). Analysis of major amino that the GABA content decreased primarily due to wash- acids in green tea by high-performance liquid chromatogra- ing in well-milled rice such as over-milled and well-milled phy coupled with OPA precolumn derivatization. Chagyo 11 ,3ῌ-- , rice. However, in less well milled rice such as under- kenkyu hokoku, , (in Japanese). Kajimoto O., Hirata H., Nakagawa S., Kajimoto Y., Hayakawa K. milled and well-milled + rice, the loss of GABA due to and Kimura M. (,**.). Hypotensive e#ect of fermented milk washing was either low or the production of GABA while containing g-aminobutyric acid (GABA) in subjects with high cooking was equal to the amount lost. The contents of normal blood pressure. Nippon Shokuhin Kagaku Kogaku GABA, glutamate, and alanine in raw and non-washed Kaishi, /+, 13ῌ20 (in Japanese). cooked rice are shown in Fig. -. Glutamate was con- Iwaki K., Matsumura Y. and Kitada Y. (,**/). Determination of verted to GABA by a glutamate decarboxylase, and ala- g-aminobutyric acid in partially milled and cooked rice sam- nine was produced by a transfer reaction of amine with ples by high-performance liquid chromatography with ﬂuore- scence detection. J. Cookery Science of Japan (Nippon chori- pyruvate (Sanga T., ,**,). In under-milled and well- kagakukai-shi), -2, ,-+ῌ,-/ (in Japanese). + milled rice, the GABA and glutamate contents markedly Ohisa N., Ohno T. and Mori K. (,**-). Free amino acid and g- increased, while that of alanine increased only slightly. aminobutyric acid contents of germinated rice. Nippon Shoku- This indicated that more glutamate was produced than hin Kagaku Kogaku Kaishi, /*, -+0ῌ-+2 (in Japanese). was lost due to conversion to GABA by a glutamate Ohtsubo S., Asano S., Sato K. and Matsumoto I. (,***). Enzymatic decarboxylase in under-milled and well-milled + rice. In production of g-aminobutyric acid using rice (Oryza sativa) 0 ,*2ῌ,++ over-milled rice, the glutamate and alanine contents in- germ. Food Sci. Technol. Res., , (in Japanese). Okada T., Sugishita T., Murakami T., Murai H., Saikusa T., Horino creased markedly but that of GABA changed only slight- T., Onoda A., Kajimoto O., Takahashi R. and Takahashi T. ly. This indicated that the glutamate produced during (,***). E#ect of the defatted rice germ enriched with GABA for cooking was converted to alanine. sleeplessness, depression, autonomic disorder by oral adminis- In conclusion, the GABA content in raw rice gradually tration. Nippon Shokuhin Kagaku Kogaku Kaishi, .1, /30ῌ0*- (in decreased in proportion to the milling yield. With re- Japanese). gard to washed and cooked rice, the GABA content de- Sanga T. (,**,). Production and metabolism of g-aminobutyric ,/th creased primarily during cooking in more milled rice. acid. In “Harper’s Biochemistry the edition,” ed. by Murray R.K., Granner D.K., Mayes P.A. and Rodwell V.W., Maruzen Inc., The GABA content of non-washed less milled and cooked Tokyo, -21ῌ-22 (in Japanese). rice was signiﬁcantly higher than that of the more milled varieties. As a result, the GABA content of cooked un-