Journal of Scientific & Industrial Research

\101. 61 , August 2002. pp 625-629

Vaginal Pheromone and Other Compounds in Mung-Bean Aroma

R L Brahmachary'

21 B, Motijheel , Calcutta 700 074, India

and

Mahua Ghosh (Basu)

Department o f Chemical Technology, University of Cal cutta, 92. A PC Road. Calcutta 700 009, Indi a

Revised : 25 January 2002; accepted: 12 April 2002

Thi s paper reports probabl y a first atte mpt to identify the molecul es which impart a most desirable, characteristic aroma to certain varieti es of mung bean of Bengal - a valuable indi geno us natural product that may be on the verge of extinction . In boiled, or fried mung bean of the two best strains , strong evidence for the presence of 2-acetyl-l-pyrroline. the molecu le responsible for the aroma of Bas mati rice l and one component o f tiger pheromone. has been obtained. GCMS analysis and computeri zed library search further suggested the presence o f six volatile compounds in the CCI4 extracts of even the second grade mung bean samples. Two of these were a mal e- ::tllracting vaginal pheromone and a puberty-delaying pheromone in rodents.

Introduction

Mung bean, Vigna radiata, is of considerable commercial interest in India. Certain strains of unusually good aroma-bearing mung bean (sona cultivar) were famous in Bengal. Unfortunate ly, these are rapidly facing extinction or may have a lready vanished.

From the scientific point of view the aroma components deserve to be ana lysed or identified. As is well-known, 2-acetyl-l-pYIT0Iine (2AP), the single largest component of basmati rice aroma was identified I in 1982. It was then traced in Palldallus foetoedui and then in the tiger pheromone3

-5 and

Maduca indicalBassia latiolila f1ower6 as well. The good mung bean aroma may well be due to some interesting, little known chemicals. Several varieties of mung bean were therefore collected and subjected to chemical analysis. As the aroma components are produced by boiling and frying, heat-dependent

--... Maillard reaction? was taken into consideration.

* Author for correspondence; he is not affi li ated to the Department of Chemical Technology. Calcutta University.

With such a perspective the possibility of N­containing mo lec ules (such as 2 AP) was explored. Because of two consecuti ve yearly floods the best quality mung bean was not avai lable for GC-MS analysi s after pre liminary investigations, yet certain interesting aroma compounds were detected even in the second-grade samp les such as those used in this study .

Materials and Methods

Materials

Seve ral mung bean samples were collected from sources at the point of c ulti vat ion the most well ­known of which is the sona moong of MaIda; another nearly equal ly good strain was also available at Lalgola, both in West Bengal. Other samples were purchased from local market in Calcutta.

Extractioll and Separatioll of Aroma Components

Boiling o f fried , or non-fried mung seeds was fo ll owed by co llec tion of steam distillate and this was shaken with e ther, hex ane and carbon tetrachloride to extract aroma co mponents in these respecti ve solvents. In pre liminary experiments CCI4 appeared

626 J SCI IND RES VOL 61 AUGUST 2002

advantageous, therefore, spectroscopic grade CCI.j was used in subsequent experiments.

Tests for the Aroma Components

Some al iquots of steam distillate were treated with hydrochloric, citric, and picric acids, and tes ted for disappearance of aroma; and then aga in treated with alkali for reappearance, if any, of the aroma.

Aqueous and organic solvent-extracts of aroma were treated with NaI04 or KI04 to detect the loss of aroma, if any. Likewise, tonsil powder was also used.

The aqueous aroma-hydrochloride, or cit rate was air dried under fan in a cold room and concentrated; the citrate sample was further concentrated by partial Iyophi li sation.

Paper Chrlllalographic Separation and Idell/ification of Aroma Impact COlllponents

Aqueous aroma sa lts were run in paper chromatograms with solven t systems li ke II -butanol­acetic ac id-water, ethanol-acetic acid-water (4: I: 1), and water saturated butanol. The posi ti on of the aroma-sa lt in chromatograms was determined after culting out the paper into several pieces, moistening with alkali, and sni ffi ng. In add ition, at least two (out of a panel of five) unbiassed sniffers (who were not told which pieces were expected to bear the aroma) were asked for their opinion. In every case, all of them clearl y identified the same pieces. Pai r-wise and triple sample compa ri sons were made in paper chromatograms, spotting samples of mung bean aroma citrate, 2-acetyl-l-pyrroline-citrate, Bassia flower aroma-citrate and Pandanlls Joe/oedlls aroma­citrate (2 tests with 1II11ng and Pandanlls aroma, 2 with 1II/11lg and standard 2 AP, 3 with mu ng and bassia, one compari son with 3 aromas, each, namely 1II1Ing, 2 AP and Bassia ).

GC alld GC-MS Studies

GLC (Hewlett Packard 5890) ana lys is was carried out with 20 per cent sq ualane column at a temperature of 65 and 90°C with oven and FID temperature at 150 °C. Anhyd rous sodi um su lfate­treated CCI.j extract was used for GC-MS .

GC-MS (Fison's In strument) ana lysis was executed in a DB 5 column (column temperature 40 °C), 1 min holding to 60°C at 5°C/mi n, total ru n time 30 min . Library search and matching of

molecular ion fragments was effected with the help of a computerised library. A blank was run with CCI4

fo r comparison.

Results

Of all the mung bean samples tested by boiling and fryin g, namel y from Bihar and Maida, Lalgola , Ranaghat, (a ll three in West Bengal) and several packets purchased from loca l market in Calcutta, the most pleasant aroma was that of Maida and Lalgola samples. The aqueous extract of the aroma was treated with di ethyl ether, hexane and carbon tetrac hl oride. The turbid material, espec ially after cru shing mung bean seeds in a blender, as well as pigments cont am in ated some of the solvents, though the aroma or at least a part of it was transferred to them. Steam di still ate of the material col lected in CCI.j turned out to be a sati sfactory techn ique. The lower CC I.j phase contained aroma components and it was not contaminated with pigments.

Ma ida and Lalgola strains treated with HCI immed iately produced an altered tone of aroma , namely that of sweet-sour, unripe mango. It happened with Bihar and Ranaghat sampies also. The acid­wa ter, spotted on fi lter paper and rendered alkaline with NaOH or KOH , yielded the mung bean smell again . Ci tri c acid or picric acid also led to loss of aroma that reappcared on treatment with alkalies. When NaJO.) or KIO.j was added to the aroma ex tract, they very quickl y destroyed the aroma, as did the trea tment with tUll si I.

An aliquot or the srcam distillate of Ma ida and Lalgo la strains or II I illig bean concentrated by air drying and then by part ial Iyophilisation was spotted on paper and rendered alkaline; it emitted a very character is ti c, intcnse aroma , resembling very concentrated 1-AP or the sweet, fresh smell of the just husked i nd ige nous rice va riet ies. With passage of timc . the quality of smell altered in a manner exactly mimicki ng 2 AP. i e. changing from sweet to the ricy tone. Spotted 011 Whatman paper 3 in acidic condition (aroma citrate) alld developed in the three solvent systems (methanol : acetic acid: water:: 4 : 1 : 1; n­butanol : acetic acid: water :: 4 : 1 : 1, v/v/v) and water-saturated butanol, the aroma material could be isolated in a specific region by sniffing the pieces of chromatogram Clit out in a consecutive seri es and rendered alkaline. Arter prelimi nary results obtained with samples rrom both Maida and Lalgola, detailed

BRAHMACHARY: VAGINAL PHEROMONES & OTHER COMPOU DS IN MUNG-BEAN AROMA 627

Solvent Front

7

6

5

4

,IS +\ :\ +

2 ++ ++ ++

I \

'-l + J ~ \.+..J -B ' 2 AP M'

Figure 1- Comparison of the three chromatograms run in watcr saturated butanol, B-MIIllg bean, 2 AP-2 acctyl- l -pyrrolinc. M-Madllca (Bass ia ) flower; +--indicates smell. ++-- indicates strong smell

study was carried out with samples from the latter. The chromatograms shown are based on this material. In all the three solvents, the Rr values of thi s region were very close to those of the standard 2 AP and Maduca illdicalBassia latifolia flower extract. figure I shows the three chromatograms run in water saturated butanol, the amount spotted was relatively large. But a narrow strip above the base line was aromaless. A smaller amount and a longer run (Figure 2) resulted into a sharper segregation of aroma material.

Figure 3 shows the comparative chromatograms of Mung beall and Maduca flower aroma run in n­butanol: acetic acid: water (4 : 1 : I, v/v) . PandallllS joetoeduslalllaryllifolius leaf extracts in citrate water was also compared in the same PC ex periments. Standard 2 AP, P. joe/oedlls, Maduca fl ower and mung bean aroma (of two strains) Maida and Lalgola were seen to ha ve the same Rr val ue.

Solvent Front

7

6

5

4

3

o 2

x _____ x

Figurc 2 - Basl/w li ri ce aroma run in water saturatcd butanol

On inj ec ting 2-3 ~lL CCI4 extrac t of ordinary quality mung bean (from Cacutta) into 20 per cent squ;:t1ane column in GLC experiments at column temperature of 6S and 90 °C, four peaks were recorded with R, less than that of the CCI4.

CCI.j_extr;l ctcd aroma was subjected to GCMS analysis. DB I <lnu 2 and DB S columns were tried and only DB S clearly separated several aroma components - six of which were good peaks . Computerised library sea rch suggested the following identi fications (Tab le I).

628 J SCIIND RES VOL 61 AUGUST 2002

Table I-The aroma compounds of second gradc Mung bcan as inLli catcd by GCMS.

Peak No. mle ILlcntificatioll

108.9

2 122.0

1st choicc

2-5, dimethyl pyrazi ne

2-ethyl-5-methyl pyrazinc

2nLl choicc

2-6. L1illlcthyl pyrazinc

2-cthyl-6-mclhyl pyrazine

3 136.4

4 *

2-cthyl-3 . 6-di-mcthyl pyrazinc

2-cthyl-3 . 6-di-mcthylamino ethanol

2. J-di Illcthyl-5-clhyl pyrazi ne

5 94.02

Additional 6

dimcthyl disulphidc

benzaldehydc

L1 illlclhyllclrasulphide

* Sec di scuss ion . Compounds (1-4) wcrc not availablc, anLi thcrcfore. thcsc cou ld not bc used:.ls standards

Discussion

Several indirec t ev idences, viz, peri odote treatment immedi ate ly remov ing the aroma, loss of aroma fo llowing tonsil powder treatment (which absorbs N-contain ing molec ules), loss of aroma on acidification and then its rea ppearance on a lka li ­treat ment, extremely sweet, utte rl y non-mung and 2 AP-li ke aroma of concentrated aqueous citrate extrac t, nature of decay-dependent alterati on of the aroma tone (from extreme sweet to ri cy), and a number of comparative PC runs with standard 2AP and P. foetoeduslamaryllifolius and Madllen illdicaJBassia latiJolia aroma in different solven ts strong ly suggested that the fragrance of mung beans of Maida and Lalgola was at least partly due to 2AP. This apparentl y occurred in a much lesser quantity, if at all, in the second grade mung bean samples exam ined. In some indigenous frag ran t rice varieti es , characteristic aroma is perceptible before boiling, even in the ripening or pre-ripened stage. But rice. on boili ng, aga in yields aroma due to Maillard reacti on

7.

Unboiled rice or even rice plant may develop the aroma ei ther through an enzymatic react ion or Maillard reaction which occurs at a slower rate at a lower temperature7 bu t catches up over time. The biosynthetic pathway of 2 AP in the ri ce plant is not known . but in mung bean there was no aroma before boiling or frying and so it was most likely ca used by the Maillard reaction7

. Some other aroma compounds were also absorbed by tonsi l, suggesting the presen e of -atom (s) containing components.

After removing the putati ve 2AP as citrate or some other salt, another comp nent of the frag rance, sweet-sour-mango smell, very unlike mung bean aroma, was perceptible. Apart from the Maida and

Solvent Front

B 1\'1

Figure J - COlllp;Jn ~OIl of M llllg hcan and Madllcu nowcr aroma run in n-hiita llol:accli c acid:watcr (4: I: I). B - M ll llg bean. M -MadllCil (lJass ia) !lowcr. +--indicales presence of

smcll

Lalgola sa mples , thi s wa so to a lesser degree with two samples of other stra in s obtained from West Benga l and Bihar. All these samples were not availahle aga in. the re fore. the further study could not be carri e out. S ince 2 AP has already been detected in more than one vegetal source, the interest was focused on ot her aroma molecules. GLC on 20 per cent squalane sho\\'cd that the first grade //lung bean aroma component has also more than one molecule, even arart froIll the swcet-sour mango smell.

BRAHMACHARY: VAGINAL PHEROMONES & OTHER COMPOUNDS IN MUNG-BEAN AROMA 629

GCMS study of a second grade sample (from Calcutta) was attempted after successful separation with CCI4 from the aqueous fraction remained as a clear phase which did not contain any pigment. DB 5 column is suitable for separating the aroma molec ules and this indeed, was borne out by the trials with DB 1,2 and 5 columns.

Of the molecules identified by GCMS analysis (Table 1), only benzaldehyde and dimethyl disulfide are well-known. Others were not available in the market, and inquiry at Alfred Bader library, Aldrich, also failed to elicit any further information. Dimethyl disulfide was also difficult to procure in Calcutta, consequently, the comparison with these standards could not be made. Benzaldehyde, relatively a simple compound, is a possible Strecker degradation product in the Maillard reaction and such Strecker aldehydes are important in food-f1avour8 and, therefore , use of a standard was not considered necessary . Benzaldehyde also occurs in bitter almonds in natural condition8

.

Dimethyl pyrazine is also a common Maillard reaction product8

, formed from glutamine-fructose reactions. However, it deserves attention as 2-5-dimethyl pyrazine has also been recorded as a puberty-delaying pheromone in femal e mice9

.

Dimethyl disulfide is one amongst 43 sui fur­containing compounds in the durian fruit8 and it has also been reported to be a male-attracting pheromone

. h . 10 present tn amster vagtna .

The presence of these mammalian pheromones in boiled/fried mung bean is apparently co-incidental products of the high-temperature Maillard reaction. The presence of a mammalian pheromone in unboiled vegetables has earl ier been reported II .

As for 2-phenylmethylaminoethanol, mw 151 , (Table 1) the rnle is 120 because the terminal-CH20H (mw - 31) immediately breaks away. The compound may be helpful against bacterial and fungal attacks l 2

It is customary to store certain mung beans after frying. Dehydration coupled with 2-phenylmethyl-

aminoethanol may be responsible for the storage protec ti on gained.

Acknowledgments

The authors are indebted to Electron Microscope Centre, Prof. D K Bhattacharyya and Dr Santinath Ghosh, Chemical Technology Department, Calcutta Univers ity; Ecolab, UIRA, Calcutta; and Prof. J Dutta, Dr. E Ali, and Dr P Bhattacharyya for their co-operation in carrying out the study. The authors also thank Dr G Schieberle for a gift of GCMS-tested 2 AP.

References

Bullery R G. Ling C & Juliano B 0 , 2-acetyl-1 pYlToline, an important aroma cO IllPonent of cooking rice, Chem Ind, (1982) 958.

2 BUllery R G, Ling C & Juliano B 0, 2-acetyl-1 pYlToline in PalldallllS allla rilli/'o IiIlS leaves, Chellllnd, (1983) 478.

3 Brahmachary R L. Poddar-Sarkar M & Dutta J, Chemical signals in the ti gcr. Chelllical Sigllals ill Vertebrates , VI (Plenum Prcss) 1992. ·n I.

4 Brahmachary R L, Poddar-Sarkar M & Dulla J, The aroma of rice and tige r. Natll re. 244 (1990) 26.

5 Brahmachary R L, The expanding world of 2-acetyl-1 pyrroline. ClIrr Sci. 71 (1996) 257.

6 Midya S & I3rallinachary R L, A method of producing baslllati rice aroilla i'rom Bassia latifolia flowers, lilt Rice Res Ill st Nell 'slell. 22 ( 1997) 21.

7 Ledl F & Schl eicher E, New aspects of Maill ard reaction in food s and in hum~l n body, Allgew Chelll (lnt Edn. Engl). 29 (6) (1990) 565.

8 Teranishi R, Wi ck E L & Hornslein I, Flavor chelllistry: 30 years a/prog ress (Kluwer, Dordrecht) 1999.

9 NOVOIIllY M. Jcmiolo B, Harvey S, Wiesler 0 & Marchlewska- Koj A. Adrenal mediated endogenous metabolites inhibit puberty in female mice. Science. 231 ( 1986) 722.

10 Singer A G. Manides F & Agosta WC, Chemical studies of hamster reprod ll ctive pheromones, Chemical Signals ill Vertebrates. vol 2 (Pl enum Press, London & New York) 1980, pp. 365.

I I Claus R & Hoppen H 0 , The hoar pheromone steroid identifi ed in vegetabl es . E.rperielliia, 35 ( 1979) 1674.

12 Chemical substance index; Chelll Abstr, 120, ( 1994) CS 3. 4942 CS .