PRELIMINARY INVESTIGATIONS ON THE RED PIGMENT IN RICE AND ITS EFFECT ON GLUCOSE RELEASE FROM RICE STARCH

Studies on. the red pigment confirmed that it was a n anthocyanin hut. not rnalridin 3-g1,ucoside ur cyanidin as previously repo~ted. I t s l~max corresponded to pelargonidin. Further the anthocyanill appeared to he associated wit11 protein. 'IYypsinization released the pigment. Sephades gel chmmatographgy also separated the pigment hcrm protein. I t appears possible tha t on parhoilirlg t h e protcin anthocyanln complex which, migrates i n to the starch gel hy some mechanism inllihits enzymatic starch hydrolysis.


INTRODUCTION
Rice (Or.y:yza sativa) is the staple food of many Asian countries. Before rice goes to the market i t is processed in a variety of ways. Paddy is dehulled raw by the process known as milling. The steps of processing may remove a part of the rice pericarp and aleurone layer, which contains a pigment and sig~lificant amounts of protein, fat, fiber, minerals and vitamins of the rice grain. This happens in dry milling and further losses occur in po1ishing.l In an alternate method of processing$ parboiling preceeds milling. Here, paddy is heated in water t o 60°C. This causes gelatinization of starch. During this process the rice pericarp and aleurone layer are incorporated in to the rice grain. Red rice has an ailthocyanin in its pericarp accounting for its colo~r."~ Ayurvedic as well as allopathic medical practitioners recommend raw red rice for diabetics. Even though there are no technical reports available, notable nutritionists of Sri Lanka have mentioned througl~ newspaper reports that 'red' is only a pigment and does not have a nutritional function. Unfortunately, no systematic work has been done on this aspect previously until a recent report (1999) stated that the parboiled red rice gives rise to a lower glycemic index than other rice types i.e. raw red rice, raw white rice and parboiled white rice.3urprisingly little has been done regarding the exact nature of the rice pigment. I11 some studies the red colour is referred to an anthocyanin hlalvidin 3-glucoside" or cyanidin4)). I t I is possible that different rice varieties have different anthocyanins, which incidentally are widely dispersed in plants and their colour affected by co-pigments.h7 The objective of'this study was to obtain some idea on the nature of the rice pigment and its effect. 011 the in-vitro enzymatic release of glucose from boiled rice.

METHODS AND MATERIALS
Sanrplea -Rice samples obtained from the Rice Research Institute station in Batalagoda was processed under standard conditions a t the Rice Processing Research and Development Center, Anuradhapura. Rice of following varieties was used in this study.
White rice: BG 300 (raw), BG 300 (parboiled) Red rice: BW 272.6 B (raw), BW 272.6 BP (parboiled) Shsvetl white rice was used in preference to polished white rice in order to keep variety constant and to ensure that all red pigment was removed whole reta~niug as much as possible of the endosperm. This was done manually by carefbully removing the red layer with a sharp blade.
Ba,qificatioll, mfiith NH, with a11.d with,oz~t trypsin,iza.tion -To the extract was added a few drops of conc. NH,. The precipitate formed was centrifuged a t 2000 rpm for 10 minutes. The absorption spectrum of the supernatant was obtained. The precipitate was redissolved in 0.1N HCl(3m1) and the absorption spectrum of the solution obtained. The pH of the solution was adjusted to 7 and trypsin (100 pl) added and the mixture incubated at 37OC for 15 minutes. Aftper basification and centrifugation as above, the absorption spectrum of this solution too was obtained.
Paper chronzatogru,pky of' th,e trypsin.izcd santplen -Butan01 : Acetone : Water (4:1:5) was used as solvent system. The top layer (butanol saturated with acetone) was separated and used as the non-polar mobile phase. Ascending chromatography was performed using Whatman no-1 paper on the trypsinized sample (5111). The paper was visualized with 1% ninhydrin in acetone spray after drying at 100°C for 2-3 mniutes.
Gel chronza.tography -Sephadex G-100 gel (5g) was soaked in 0.1M phosphate buffer (pH7, 200ml) for. 72h for swelling, and packed in a column (height-16cm and diameter-2cm). The void volume was determined using dextran blue (lmg in 0.5ml). The pigment extract was evaporated to l m l and the pH adjusted to 7 and introduced to the colum~i. After the void volume eluted, 2.5ml fractions were co1lecf;ed in constant intervals with a time-based fraction collector (flow rate -2.5ml per 12min). The column was run for 12h. The absorbance of all fractions was read at 280nm and at 520nm. The coloured samples were combined and freeze-dried.

Folill. Ciocnltenu. pherl,ol reagent test for protein.?-
The test was conducted for both the colourless and coloured fractions.

Estimation of rate of' starch digestion of different types of rice
Ei~zyrt,atct h.ydrol,y,si.s -Rice specimens were boiled until cooked for about 20 minutes Samples ( l g ) were homogenized, and heated for 10 minutes a t 100°C with Ilitelnpase (20p1), (a heat stable cx amylase from Bacillus Licheniformis, Quest, 1140unit,s ml-'1 after adjusting the pH to 7. The pH was then adjusted to 5.0 and Amylo (20~1) was added ( a glucoamylase from Aspergillus niger, from yeast and a 1-6: glucano glucan, Quest, llOOunits ml-l) and kept a t 55OC. Samples (lml) were collect,ed at 45min and then in 111 intervals. To determine the complete hydrolysis, larger. quantities of enzymes were used (0.2ml Hitempase and lml Amylo) and lllcubated at 55°C for a longer time (6h).
Assan, of' reduuilzg sugars b y the Nelson meth.od !' -Using Nelson method, glucose was estirrlat,ed in the samples. The Nelson method gave a linear standard curve wit,l? an R2 value of 0.99 with a C.V. for the various concentrations 5-10'3) (an absorbance of 0.34 works out to lOOpg m l l in the sample tested).

The pigment
It was fo~md that the pi-grnent of red parboiled rice was very difficult to extract in ethanolic HCI (anthocyanin extracting solvent) compared to raw red rice. This was probably due to either binding or the texture of rice. Homogenized parboiled red I rice (EW 272.6BP) released only 20-25% of the pigment released by raw red rice of the same variety on extraction with ethanolic HC1. Absorbance was measured a t I the hmax of 520nm.
On extraction of the pigment from raw red rice with methanolic HC1 and obtaining UV spectrum, a hmax of 520nm -as observed. On basification with NH, the coloured substance precipitated with a large amount of other components. The supernatant gave a slight yellow colour with a hmax of 470nm. The acidification of the precipitate gave back the red colour with a hmax of 520nm. Addition of trypsin to the precipitate a t pH7 resulted in the colour being released from the solid material. On basification the supernatant still gave a red colour with a hmax of 580nm without giving any precipitate. The red colour became lighter when sodium nletabisulphite (5mg) was added as expected for anthocyanins being irreversible with excess metabi~ulphite.'~ -' "These absorption maxima do not correspond to cyanidin (hmax-535nm) or malvidin 3-glucoside (hmax-550nm).lo The solution resulting from trypsinization, paper chromatography and ninhydrin spray yielded a streak of ninhydrin positive purple with a dark purple spot just moving above the origin.
Following gel chromatography the red colour was obtained only a t eluent volume of 40ml after the void volume. The fractions before this showed a slight consistent absorption at 280nm and mildly positive for the Folin Ciocalteau test. The red pi,gment gave a strong Folin Ciocalteau positive reaction.

Location of pigment in processed rice
In raw red rice the pigment is in the outer pericarp and can easily be extracted in ethanolic -HC1. In parboiled red rice the pi-pnent has impregnated the gelatinized starch which has been subsequently dried. The grain is glassy with the red pigment being found in the starchy gel. I t is not possible to extract the red pigment quantitatively even after powdering. Therefore the extent of red pigment in raw red and parboiled red cannot be compared. However this does not affect the final conclusion.

Rate of starch digestion of different types of rice
The glucose release from parboiled red rice is much slower initially when compared t,o raw red rice, raw white rice and parboiled white rice (Fig. 1 1. Shaving the red layer from raw red rice had no effect compared to raw red rice but both raw types showed higher initial glucose release (1-2h) compared to parboiled red rice. This shows that the red pigment had no effect on glucose release if there was no parboiling to transfer it into the starch gel (Fig. 2). Using larger amounts of enzyme, total glucose release was found to be similar a t the end of 6h indicating comparable glucose release in the 4 types of rice. parboiled red rice, shaved raw red rice

DISCUSSION
The abo-ire experiments confirm that the red pigment is an anthocyanin. B~zt hmax values from this rice variety indicate that it is not cyanidin or malvidin 3-glucoside and possibly pelargonidin because its hmax was 5201lm.~."' Its colour also ciisappeared with treatment with sodium metabisulphite.1° Precipitation in allraline rxledium initially was a puzzle as anthocyanins are more soluble in basic mediurxi. From the hull< of' the precipitate it was assumed that the anthocyanin was 111 some way attachetl to a macro-component.This appears valid as trypsinizc~tion liberated the anthocy:~ni~i, which was then so1ubl.e in basic medium. Paper chronlatography of the trypsinized sample showing degradation prod~zcts of proteins further supportfed this finding. Gel chromatography indicated that the binding of the pigment to protein was weak as the pigment elut,ed separately from protein. The positive Folin Ciocalteau can be due to reaction of the red pigment with phenolic groups of the anthocyanin.
It has not been possible to compare the relative content of red pigment in raw and parboiled red rice.
In-vitro enzyme hydrolysis experiments clearly ( fig. 1 and Fig. 2 ) show that a conll~jnation of parboiling and presence of red pigment causes a slower release of glucose. This is consistent with the finding that parboiled red rice has the lowest gl.ycemjc index of the rice types studied.This can be due either t,o simple nonspecj:fic ~,olyphenolic i~ihibition'~ of rx amylase or by protein-anf;hocyanin linlreii steric factors (recal.ling that rice starch contains a protein coat). It, is possible that parboiled led rice has more anthocyanin than raw red rice but the authors d.o not lwlieve that thi.s is aiajor factor. These studies are prelinliilary and i t could. be argued t,l?at if nlaninlaliarl enzyIries were used the results could be different. However recent studies'show that a simi1.a~ result was manifested using pancreatic a -amylase. This st~zdy shows t.liat the red pigment inhibits even the more rugged microbinl enzymes. I t is specu'lnteti that the protein-pigment complex, by a niechallisln not cotn]>letely ~111.d erst,ood, permeates the starch gel, in parboiling and inhibits in-vitro enzymatic l~.ydro:l.ysis. The t'act that the same variety of rice was used i11 the case of parboiled and raw milled rj.ce coni.irnled that the change occ~zrred. as a result of processing tt:chniq,ues and not variety. F~lrther that both the red pjgrnent and parboiling is necessary t,o slow gl.ucose release.
The inhibition of enzyme hydrolysis of starch in the presence of rice pigment of parboiled rice could provide the biochemical basls for the beneiicial effects of pan'bolled red rice. However, clea~ly more work has to be done.