J Sci Food Agric 1997, 75, 258È262 A Simple and Rapid Method for Isolating Cassava Leaf Linamarase Suitable for Cassava Cyanide Determination Hock-Hin Yeoh,* J Howard Bradbury” and Sylvia V Egan Division of Botany and Zoology, The Australian National University, Canberra, ACT 0200, Australia (Received 23 July 1996 ; revised version received 5 March 1997 ; accepted 26 March 1997) Abstract : A simple method was developed for isolating cassava leaf linamarase. It involved homogenising the leaves in a bu†er containing polyvinylpolypyrrolidone, followed by ammonium sulphate precipitation, Ðltration and partial puriÐcation of the enzyme by hydrophobic interaction chromatography. All stages of the enzyme preparation were carried out at room temperature and it was completed within 90 min. A linamarin-indicator strip for checking the activity of the enzyme preparation was also developed. This isolation technique which has been developed into a simple kit should be suitable for laboratories in developing countries where the enzyme is needed to determine the cyanogenic potential of cassava and its products. J Sci Food Agric 75, 258È262 (1997) No. of Figures : 0. No. of Tables : 2. No. of References : 21 Key words : cassava, linamarase preparation, simple kit, enzyme isolation, hydrophobic interaction chromatography, cyanide determination existing procedures (Wood 1966 ; Cooke et al 1978 ; Cooke 1979 ; Eksittikul and Chulavanatol 1988 ; Yeoh 1989 ; Mpkong et al 1990) or because the cost of commercial linamarase is prohibitive. One possible solution to this problem was to develop an acid hydrolysis method (Bradbury et al 1991) which removed the need for enzyme. Another approach would be to develop a simple procedure for linamarase (a bglucosidase, EC 184.108.40.206) isolation that could be used by laboratories in developing countries, bearing in mind that they may lack instruments for protein isolation. It would also be useful to develop an instrument-free method for checking the activity of the enzyme preparation. In this paper, we describe a simple method for isolation of linamarase from cassava leaves, which was tested on 10 di†erent varieties of cassava. INTRODUCTION Cassava (Manihot esculenta Crantz) is widely grown for its edible storage roots (Cock 1985). However, the presence of cyanogenic glucosides, mainly linamarin, and their breakdown products in cassava and its processed products, has been a cause of concern from the viewpoint of food safety. Many methods for measuring total cyanide content are available (Cooke 1978 ; Bradbury et al 1991 ; Bradbury and Egan 1992 ; Yeoh and Truong 1993 ; Yeoh 1993 ; Bradbury et al 1994 ; Yeoh and Tan 1994a,b ; Brimer 1994 ; Tatsuma et al 1996 ; Yeoh et al 1996). Several of these methods are based on the enzymatic hydrolysis of linamarin, followed by spectrophotometric, potentiometric or amperometric measurement of the cyanide released. Unfortunately, enzyme-based methods cannot be used in many laboratories in developing countries, either because of lack of the equipment needed for enzyme preparation using EXPERIMENTAL Plant material * On sabbatical leave from School of Biological Sciences, The National University of Singapore, Kent Ridge, Singapore 119260. ” To whom correspondence should be addressed. Leaves of cassava plants were obtained from a collection maintained in the Plant Culture Facility at the Australian National University, Canberra. 258 ( 1997 SCI. J Sci Food Agric 0022-5142/97/$17.50. Printed in Great Britain Isolation of cassava linamarase Enzyme preparation About 2 g of fresh, fully expanded young leaves were used. They were homogenised in a mortar in 10 ml 100 mM Na citrate at pH 6 containing 10 g litre~1 polyvinylpolypyrrolidone, followed by the addition of 12 ml of 2 M (NH ) SO . The e†ective molarity of the crude 42 4 extract was about 1 M (NH ) SO . The homogenate 42 4 was allowed to stand for 30 min at room temperature, then Ðltered through a layer of Ñour-sack cloth on a Ðlter funnel. The Ðltrate became clear (slightly tinted yellow) after 10È12 ml of Ðltrate (green) had passed through. At this point, collection of the clear Ðltrate was continued in a measuring cylinder. The initial collection of Ðltrate (green) was gently poured back on the Ñoursack cloth Ðlter. Alternatively, Ðltration through Whatman No 1 Ðlter paper gave a clear Ðltrate and was quite satisfactory. About 10È13 ml of the clear Ðltrate was recovered. A phenyl Sepharose CL-4B (Pharmacia Biotech, Sweden) column was prepared by using a 5 ml syringe barrel which was Ðlled up to the 2 ml mark with the phenyl Sepharose slurry. This was washed with 20 ml water and then equilibrated with 10 ml 1 M (NH ) SO 42 4 in 50 mM Na citrate at pH 6. The clear Ðltrate (10È 13 ml) was then passed through the equilibrated phenyl Sepharose column followed by 6 ml 0É3 M (NH ) SO 42 4 in 50 mM Na citrate at pH 6 then 6 ml water. The eluate from the water wash was collected and contained the linamarase. The phenyl Sepharose CL-4B could be regenerated by soaking in 0É1 M NaOH for about 30 min, then washing it with water until the washings were neutral. For some cassava varieties, we found it better to use 12 ml 0É8 M (NH ) SO instead of 2 M (NH ) SO in 42 4 42 4 the enzyme extraction step (see above) and in this case the phenyl Sepharose CL-4B was equilibrated with 0É4 M (NH ) SO (instead of 1 M (NH ) SO , see 42 4 42 4 above) in 50 mM Na citrate at pH 6. The bound proteins were desorbed as before. Protein determination This was carried out using the Bio-Rad protein assay kit (Bradford, 1976). Enzyme assay For convenience, we monitored the b-glucosidase activity and for the Ðnal preparation we also determined the linamarase activity. One unit (U) of enzyme activity was deÐned as 1 kmol of the substrate hydrolysed in 1 min. The b-glucosidase assay was performed using 5 mM pnitrophenyl-b-D-glucoside in 50 mM Na citrate at pH 6 (Yeoh 1989). The total reaction volume was 600 kl. After incubation at 37¡C for 15 min, 3 ml 100 mM Na CO was added to the reaction mixture and the 2 3 259 absorbance measured at 420 nm. Linamarase activity was determined using 5 mM linamarin in 50 mM Na citrate at pH 6 at 37¡C for 15 min. The total reaction volume was kept at 600 kl (Yeoh 1989). The glucose released was determined using the glucose oxidase method (Raabo and Terkildsen 1960). Preparation of linamarin-indicator paper strip An aliquot (20 kl) of 20 mM linamarin was spotted on a piece of 10 ] 15 mm Whatman chromatography paper grade 17 Chr and left to dry at room temperature. Whatman 3MM chromatography paper (10 ] 15 mm) soaked in alkaline picrate solution (Williams and Edwards) 1980) was also left to dry at room temperature. They were then glued onto a 10 ] 40 mm strip of plastic cut from an overhead transparency sheet. The linamarin-paper was placed at one end of the plastic strip and the picrate paper placed next to it, with a 2È3 mm gap between them. A water-based commercial glue was used for this purpose. To test the enzyme activity 50È100 kl of the enzyme preparation was spotted on the linamarin-paper. It was immediately placed in a clear plastic bag designed to Ðt 10 ] 40 mm paper strips and the bag sealed with tape. Alternatively, an air-tight bottle measuring about 14 mm id ] 75 mm could be used. It was incubated at 37¡C, or at 30¡C. If the picrate paper changed from yellow to brick red in 15 to 30 min, then the enzyme preparation had high linamarase (b-glucosidase) activity. The reaction was observed to be faster at 37¡C than at 30¡C. If the picrate paper changed from yellow to brown or brick-red overnight, the enzyme preparation was low in activity but might still be useful for cassava cyanide determination. RESULTS AND DISCUSSION Several methods for isolating cassava linamarase are available (Wood 1966 ; Cooke et al 1978 ; Cooke 1979 ; Eksittikul and Chulavanatol 1988 ; Yeoh 1989 ; Mpkong et al 1990), but many laboratories in developing countries are unable to prepare the enzyme because they lack the facilities to carry out these procedures. Even in a simple isolation method which involves only ammonium sulfate precipitation (Cooke 1979), a centrifuge was needed. The method developed here eliminates the use of a centrifuge. Separation of leaf debris and other precipitated proteins was achieved by simple Ðltration. We found that Ñour-sack cloth which is available in the tropics, was suitable for this purpose, although Whatman No 1 Ðlter paper was satisfactory. Instead of using ammonium sulphate to precipitate the linamarase as employed by most methods, we used this salt to facilitate the hydrophobic attachment of the enzyme present in the soluble form to the phenyl Sepharose CL-4B. The bound proteins were desorbed from the column using water and H-H Y eoh, J H Bradbury, S V Egan 260 the simple batch elution procedure did not require any instruments associated with column chromatography. Overall, the steps were simple to follow and it took about 90 min to prepare the enzyme. It was found possible to scale up the isolation protocol using 10 g cassava leaves instead of 2 g. However, the Ðltration step and binding of the proteins to the phenyl Sepharose column took a longer time, which increased the total time needed to prepare the enzyme. Leaves of 10 cassava varieties were used to assess if the method would be useful for cassava in general, as many di†erent varieties are grown in the tropics. In this protocol, the homogenate was brought to about 1 M (NH ) SO and the proteins adsorbed to phenyl Sep42 4 harose CL-4B in 1 M (NH ) SO (Table 1). Under these 42 4 conditions, the preparations showed enzyme purities ranging from about 3È9-fold and recoveries of 30È81%, depending on the cassava variety. Preparations from the Ðrst six entries in Table 1, gave high b-glucosidase activity (2É4È14É5 U) and they exhibited high linamarase activities (118È680 U mg~1 ; see Table 1). On the other hand, the last four entries in Table 1 gave much lower enzyme activities ; the total b-glucosidase activity in the crude enzyme Ðltrate was only 0É13 to 0É8 U (see Table 1). In an attempt to overcome this problem of low amounts of enzyme being extracted, we modiÐed the extraction conditions. Changes were made to the amount of polyvinylpolypyrrolidone and/or (NH ) SO 42 4 used. Using variety TMS 60506, it was observed that by increasing the polyvinylpolypyrrolidone to 20 g litre~1 and retaining the e†ective molarity of (NH ) SO at 42 4 1 M, the enzyme activity in the crude preparation was still low (total b-glucosidase activity, 0É13 U ; see Table 2). On the other hand, when the e†ective molarity of (NH ) SO was reduced from 1É0 to 0É4 M and the poly42 4 vinylpolypyrrolidone content retained at 10 g litre~1, the crude enzyme Ðltrate showed an increase of nearly 100 fold in enzyme activity (total b-glucosidase activity, 12É7 U). Correspondingly, the Ðnal preparation was higher in linamarase activity (382 U mg~1 protein) compared to results of earlier extraction procedures (see Table 1). It seems likely that with cv TMS 60506 some of the linamarase was precipitated along with other proteins and lost in 1É0 M (NH ) SO , but not in 0É4 M 42 4 (NH ) SO . Increasing polyvinylpolypyrrolidone to 42 4 40 g litre~1 and retaining the use of 0É4 M (NH ) SO , 42 4 however, decreased the level of extractable enzyme activity (total b-glucosidase activity, 1É8 U). Based on these data, we used a lower (NH ) SO concentration 42 4 to prepare the enzyme from TMS 30572, TMS 71693 and MCOL 1468 (Table 2). The results showed that only TMS 30572 gave a high enzyme activity (total bglucosidase activity in crude extract, 10É3 U ; linamarase activity of Ðnal preparation, 528 U mg~1 protein) while the other two varieties still gave low enzyme activities in their preparation (total b-glucosidase activity, 0É2È 0É3 U). The di†erences observed between the 10 cassava cultivars studied showed that di†erent extraction conditions were necessary to obtain the best results for di†erent varieties. Because of the variability of the linamarase activity obtained from di†erent cassava cultivars, it was important to check the level of enzyme activity in the preparation. In this respect, we developed a linamarin-indicator paper strip in which about 50 kl TABLE 1 Preparation of cassava leaf linamarase (b-glucosidase) from cassava varieties Cassava variety Crude enzyme Ðltrate T otal protein (mg) SM1 150 M AUS 7 TMS 4(2) 1425 TMS 91934 TMS 63397 TMS 50395 TMS 71693 TMS 30572 TMS 60506 MCOL 1468 3É5 2É4 1É3 1É9 0É7 0É9 0É6 0É6 0É4 0É5 b-Glucosidase activity T otal activity (U) SpeciÐc activity (U mg~1) 19É5 10É9 10É6 11É5 3É2 6É1 0É8 0É4 0É13 0É13 5É6 4É5 8É2 6É1 4É6 6É8 1É3 0É7 0É3 0É3 Preparation from Phenyl Sepharose CL -4B T otal protein (mg) 0É7 0É6 0É2 0É4 0É1 0É1 0É05 0É05 0É04 0É02 b-Glucosidase activity T otal activity (U) SpeciÐc activity (U mg~1) 14É5 8É6 8É6 9É1 2É4 4É6 0É6 0É17 0É06 0É04 20É7 14É3 43É0 22É8 24É0 46É0 12É0 3É4 1É5 2É0 Puritya (Fold) Recoverya (%) 3É7 3É2 5É2 3É7 5É2 6É8 9É2 4É9 5É0 6É7 74 79 81 79 75 75 75 50 46 30 L inamarase activity (U mg~1)b 320 118 495 348 530 680 159 26É4 1É8 10É8 (37É3) (11É7) (19É0) (21É0) (6É0) (13É0) (1É4) (0É2) (0É01) (0É04) a Purity and recovery were based on b-glucosidase activity, purity \ col 7/col 4, recovery (%) \ col 6/col 3 ] 100. b Values in parentheses refer to U ml~1 of Ðnal enzyme preparation. Isolation of cassava linamarase 261 TABLE 2 Preparation of cassava leaf linamarase (b-glucosidase) following some modiÐcations of the extraction conditions Cassava Crude enzyme Ðltrate T otal protein (mg) TMS 60506c TMS 60506d TMS 60506e TMS 30572d TMS 71693d MCOL 1468d a b c d e 0É4 1É3 0É6 0É8 0É8 0É4 b-Glucosidase activity T otal activity (U) SpeciÐc activity (U mg~1) 0É13 12É7 1É8 10É3 0É3 0É2 0É3 9É8 3É0 12É9 0É4 0É5 Preparation from Phenyl Sepharose CL -4B T otal protein (mg) 0É04 0É3 0É07 0É14 0É04 0É04 b-Glucosidase activity T otal activity (U) SpeciÐc activity (U mg~1) 0É08 9É3 1É4 7É6 0É06 0É04 2É0 31É0 20É0 76É0 1É5 1É0 Puritya (Fold) Recoverya (%) 6É7 3É2 6É7 5É9 3É8 2É0 80 73 78 74 20 20 L inamarase activity (U mg~1)b 19É2 382 208 528 3É6 1É2 (0É1) (19É1) (3É5) (8É8) (0É02) (0É01) Purity and recovery were based on b-glucosidase activity, purity \ col 7/col 4, recovery (%) \ col 6/col 3 ] 100. Values in parentheses refer to U ml~1 of Ðnal enzyme preparation. Extraction medium contained 20 g litre~1 polyvinylpolypyrrolidone, followed by addition of 2 M (NH ) SO . 42 4 Extraction medium contained 10 g litre~1 polyvinylpolypyrrolidone, followed by addition of 0É8 M (NH ) SO . 42 4 Extraction medium contained 40 g litre~1 polyvinylpolypyrrolidone, followed by addition of 0É8 M (NH ) SO . 42 4 (one drop) of the enzyme preparation was added to the linamarin on the paper. Using the enzymes prepared in Table 1, it was observed that preparations with linamarase activity greater than 6 U ml~1 changed the picrate paper from yellow to brick-red within 15 min at 37¡C. Enzyme preparations from the four varieties that gave low linamarase activity required an overnight incubation (about 20 h) to turn the picrate paper drown. We also compared this response with their ability to hydrolyse the linamarin in cassava Ñour. Here, we added 50 kl of the enzyme preparation to 100 mg cassava Ñour in 450 kl phosphate bu†er at pH 8 in an air-tight bottle in which a piece of picrate paper was suspended. After incubation at 30¡C for about 20 h, the colour of the picrate paper was compared against the control which contained sufficient linamarase to hydrolyse the linamarin in the Ñour sample. It was observed that the six enzyme preparations with high linamarase activity were able to hydrolyse completely the linamarin in the Ñour. For preparations with low enzyme activity, namely those from varieties TMS 60506, TMS 71693, TMS 30572 and MCOL 1468, 200 kl of the enzyme preparation was required to give the same result. Although it was possible to use enzyme preparations with low linamarase activity, we would suggest using those with high enzyme activity on the basis of productivity, since a 6 ml preparation from 2 g leaves could be used for about 120 analyses, assuming 50 kl were needed for each analysis. Overall, a simple and rapid method of preparing cassava leaf linamarase was developed. The linamarinindicator paper could be used to check the activity of the enzyme preparation and from its response one could use a suitable amount for the determination of linamarin or cyanogenic potential in cassava. Since the procedure does not require any equipment such as a centrifuge or fraction collector, it should be suitable for laboratories in developing countries. This should enable researchers and food analysts to use the enzyme-based methods of linamarin and cyanogenic potential determination, thus making safer the consumption of cassava and cassava products. A simple kit has been developed for the preparation of linamarase from cassava leaves using this method. ACKNOWLEDGEMENTS HHY is a Visiting Fellow at the Australian National University on study leave from the National University of Singapore. The authors thank Mr Chew Boon Swee of Exclusive Mark (M) Sdn Bhd, Malaysia for kindly donating the water-base commercial glue. Thanks also to the Australian Centre for International Agricultural Research (ACIAR) for Ðnancial support of this work. REFERENCES Bradbury J H, Egan S V 1992 Rapid screening assay of cyanide content of cassava. Phytochem Anal 3 91È94. Bradbury J H, Egan S V, Lynch M J 1991 Analysis of cyanide in cassava using acid hydrolysis of cyanogenic glucosides. J Sci Food Agric 55 227È290. Bradbury J H, Bradbury M G, Egan S V 1994 Comparison of methods of analysis of cyanogens in cassava Acta Hort 375 87È96. Bradford M M 1976 A rapid and sensitive method for the quantiÐcation of microgram quantities of protein utilising the principle of proteinÈdye binding. Anal Biochem 72 248È 254. 262 Brimer L 1994 Quantitative solid-state detection of cyanogens : from Ðeld test kits to semi-automated laboratory systems allowing kinetic measurements. Acta Hort 375 105È 116. Cock J H 1985 Cassava New Potential for a Neglected Crop. Westview Press, Boulder, CO, USA. Cooke R D 1978 An enzymatic assay for the total cyanide content of cassava. J Sci Food Agric 29 345È352. Cooke R D 1979 Enzymatic assay for determining the cyanide content of cassava and cassava products. Centro Internacional de Agricultura (CIAT), Cali, Colombia, Series 05EC-6, pp 1È14. Cooke R D, Blake G G, Battershill J M 1978 PuriÐcation of cassava linamarase. Phytochemistry 17 381È383. Eksittikul T, Chulavanatol M 1988 Characterization of cyanogenic b-glucosidase (linamarase) from cassava (Manihot esculenta Crantz). Arch Biochem Biophys 266 263È 269. Mpkong O E, Yan H, Chism G, Sayre R T 1990 PuriÐcation, characterization and localization of linamarase in cassava. Plant Physiol 93 176È181. Raabo E, Terkildsen T C 1960 On the enzymatic determination of blood glucose. Scand J Clin L ab Invest 12 402È 407. H-H Y eoh, J H Bradbury, S V Egan Tatsuma T, Tani K, Oyama N, Yeoh H H 1996 Linamarin sensors : amperometric sensing of linamarin using linamarase and glucose oxidase. J Electroanal Chem (in press). Williams H J, Edwards T G 1980 Estimation of cyanide with alkaline picrate. J Sci Food Agric 31 15È22. Wood T M 1966 The isolation, properties and enzymatic translocation of linamarin from cassava. J Sci Food Agric 17 85È90. Yeoh H H 1989 Kinetic properties of b-glucosidase from cassava. Phytochemistry 28 721È724. Yeoh H H 1993 Nylon-linamarase electrode for rapid determination of linamarin. Biotechnol T echnol 7 761È764. Yeoh H H, Tan C K C 1994a An enzyme-immobilized microplate determination of linamarin for large number of samples. Biotechnol T echnol 8 337È338. Yeoh H H, Tan C K C 1994b Determination of linamarin in cassava using enzyme-sensitised microcentrifuge tubes. J Sci Food Agric 66 31È33. Yeoh H H, Truong V D 1993 Quantitative analysis of linamarin in cassava using a cassava b-glucosidase electrode. Food Chem 47 295È298. Yeoh H H, Lim S L, Woo H C 1996 An enzyme-bound linamarin indicator paper strip for the semi-quantitative estimation of linamarin. Biotechnol T echnol 10 319È322.