Purification, characterization, and localization of linamarase in cassava

Linamarase expression in cassava cultivars with roots of low- and high-cyanide content

This paper reports the expression and localization of linamarase in roots of two cassava (Manihot esculenta Crantz) cultivars of low and high cyanide. Two different patterns of linamarase activity were observed. In the low-cyanide type, young leaves displayed very high enzyme activity during the early plant growing stage (3 months), whereas in root peel, the activity increased progressively to reach a peak in 11-month-old plants. Conversely, in the high-cyanide cultivar (HCV), root peel linamarase activity decreased during the growth cycle, whereas in expanded leaves linamarase activity peaked in 11-month-old plants. The accumulation of linamarin showed a similar pattern in both cultivars, although a higher concentration was always found in the HCV. Linamarase was found mainly in laticifer cells of petioles and roots of both cultivars with no significant differences between them. At the subcellular level, there were sharp differences because linamarase was found mainly in the cell walls of the HCV, whereas in the low-cyanide cultivar, the enzyme was present in vacuoles and cell wall of laticifer cells. Reverse transcriptase-PCR on cassava tissues showed no expression of linamarase in cassava roots, thus, the transport of linamarase from shoots to roots through laticifers is proposed.

Cyanide bystander effect of the linamarase/linamarin killer-suicide gene therapy system

BACKGROUND

The killer-suicide system linamarase/linamarin (lis/lin) uses the plant gene linamarase (beta-glucosidase) to convert the cyanogenic glucoside substrate, linamarin, into glucose and cyanide. We have studied the bystander effect associated with this new system mediated by the production of the cyanide ion that diffuses freely across membranes.

METHODS

Immunofluorescent staining of cells treated with an anti-linamarase antibody allowed us to localize the enzyme within the cells. Flow cytometry was used to determine the sensitivity of different mixtures of cells, C6lis and C6gfp (green), to linamarin as a percentage of cell survival.

RESULTS

We demonstrate here that rat glioblastoma C6 cells carrying the linamarase gene (lis), mixed with naive C6 cells and exposed to linamarin, induce generalized cell death. Cells expressing lis efficiently export linamarase, whereas linamarin enters cells poorly by endocytosis; as a result most of the cyanide is produced outside the cells. The study was facilitated by the presence of the green fluorescent protein (gfp) gene in the bystander population. As few as 10% C6lis-positive cells are sufficient to eliminate the entire cell culture in 96 h.

CONCLUSIONS

This bystander mechanism does not preferentially kill toxic metabolite producer cells compared with bystander cells, thus allowing production of sufficient cyanide to cause tumor regression. In this report we confirm the potential of the lis/lin gene therapy system as a powerful tool to eliminate tumors in vivo.

Purification and characterisation of a beta-glucosidase abundantly expressed in ripe sweet cherry (Prunus avium L.) fruit

A beta-glucosidase (beta-D-glucoside glucohydrolase, EC 3.2.1.21) was purified to homogeneity from ripe fruits of sweet cherry (Prunus avium L.) by ammonium sulphate precipitation, ion exchange and size exclusion chromatography. The enzyme is a monomer with a molecular mass of approximately 68 kDa and an acidic isoelectric point. N-terminal sequence analysis indicated that sweet cherry beta-glucosidase is related to other plant cyanogenic beta-glucosidases. Substrate specificity studies revealed that the enzyme is able to attack and hydrolyse several synthetic substrates and total cell walls purified from ripe fruit. Biochemical and immunolocalisation studies showed that sweet cherry beta-glucosidases are mainly localised in the cytosol and in the apoplast, at the unripe stage of ripening; in ripe fruit it is also associated with cell wall.

The use of a starter culture in the fermentation of cassava for the production of "kivunde", a traditional Tanzanian food product

Three cassava fermentation methods (spontaneous fermentation, back-slopping and the use of starter culture) for the production of kivunde, executed in three trials at 30 degrees C, were compared in terms of cyanide level reduction, microbiology and product quality improvement. Among the isolates from spontaneously fermented cassava batches, four strains were selected on the basis of their enzymatic activities and acid production. All were identified as Lactobacillus plantarum and were used as starters in this study. Lowest residual cyanide levels were detected after 120 h fermentation time in samples fermented with the starter culture and were below the maximum value of 10 mg/kg recommended by the Codex/FAO for cassava flour. This finding seems to be related to the alpha-glucosidase activity of the inoculated strains of which API-zyme (Bio-Merieux) tests showed activities of between 20 and > or = 40 nmol/4 h. The total residual cyanide levels of the spontaneous and back-slopping fermentations at 96 h were respectively 43.5 and 47.7 mg/kg dry weight of cassava. Extension of the fermentation period to 5 days, lead to further substantial reduction in the residual cyanide level in both these processes, but not below the recommended maximum value as in the case of starter culture fermented products. The spontaneous and back-slopping fermented cassava showed signs of deterioration after 3 days of fermentation. There was a sharp drop of pH and an increase of titratable acidity for all three batches during the first 48 h followed by a slow rise of pH and drop in titratable acidity towards the end of fermentation. The samples fermented with the starter culture had a smooth texture and pleasant fruity aroma, as opposed to the course and dull appearance and more complex flavour of the samples of spontaneous and back-slopping batches. During fermentation with starter culture, Enterobacteriaceae and yeasts and moulds could not be isolated throughout the period of fermentation (detection limit: 10 colony forming units/g). The present findings indicate the suitability of these Lb. plantarum strains as starter cultures for cassava fermentation in the kivunde process. The paper highlights the potential for the improvement of a traditional African fermented food (kivunde) through the use of a starter culture.

Plant cyanogenic glycosides

The cyanogenic glycosides belong to the products of secondary metabolism, to the natural products of plants. These compounds are composed of an alpha-hydroxynitrile type aglycone and of a sugar moiety (mostly D-glucose). The distribution of the cyanogenic glycosides (CGs) in the plant kingdom is relatively wide, the number of CG-containing taxa is at least 2500, and a lot of such taxa belong to families Fabaceae, Rosaceae, Linaceae, Compositae and others. Different methods of determination are discussed (including the indirect classical photometrical and the new direct chromatographic ones). The genetic control of cyanogenesis has no unique mechanism, the plants show variation in the amount of the produced HCN. The production of HCN depends on both the biosynthesis of CGs and on the existence (or absence) of its degrading enzymes. The biosynthetic precursors of the CGs are different L-amino acids, these are hydroxylated then the N-hydroxylamino acids are converted to aldoximes, these are turned into nitriles. The last ones are hydroxylated to alpha-hydroxynitriles and then they are glycosilated to CGs. The generation of HCN from CGs is a two step process involving a deglycosilation and a cleavage of the molecule (regulated by beta-glucosidase and alpha-hydroxynitrilase). The tissue level compartmentalisation of CGs and their hydrolysing enzymes prevents large-scale hydrolysis in intact plant tissue. The actual level of CGs is determined by various factors both developmental and ecological ones, which are reviewed too. The last part of the present work demonstrates the biological roles of CGs in plant physiological processes and in plant defence mechanisms as well. The effect of CGs (HCN) on different animals, the symptoms of poisonings are discussed to cows, sheep, donkeys, horses and chicks. Finally, the poisonous effects of cassava (Manihot esculenta) roots are summarised on experimental animals and on the human organism.

Cyanogenesis in cassava. The role of hydroxynitrile lyase in root cyanide production

In the cyanogenic crop cassava (Manihot esculenta, Crantz), the final step in cyanide production is the conversion of acetone cyanohydrin, the deglycosylation product of linamarin, to cyanide plus acetone. This process occurs spontaneously at pH greater than 5. 0 or enzymatically and is catalyzed by hydroxynitrile lyase (HNL). Recently, it has been demonstrated that acetone cyanohydrin is present in poorly processed cassava root food products. Since it has generally been assumed that HNL is present in all cassava tissues, we reinvestigated the enzymatic properties and tissue-specific distribution of HNL in cassava. We report the development of a rapid two-step purification protocol for cassava HNL, which yields an enzyme that is catalytically more efficient than previously reported (Hughes, J., Carvalho, F., and Hughes, M. [1994] Arch Biochem Biophys 311: 496-502). Analyses of the distribution of HNL activity and protein indicate that the accumulation of acetone cyanohydrin in roots is due to the absence of HNL, not to inhibition of the enzyme. Furthermore, the absence of HNL in roots and stems is associated with very low steady-state HNL transcript levels. It is proposed that the lack of HNL in cassava roots accounts for the high acetone cyanohydrin levels in poorly processed cassava food products.

Subsite structure of the beta-glucosidase from Aspergillus niger, evaluated by steady-state kinetics with cello-oligosaccharides as substrates

The beta-glucosidase from a commercially available preparation from Aspergillus niger was highly purified. The Michaelis constant Km and the molar activity K0 for cello-oligosaccharide substrates Gn (n = 2-6) were obtained by steady-state kinetic analysis on the beta-glucosidase-catalyzed hydrolysis at 25 degrees C and pH 5.0. Stoichiometric production of Gn-1 by the beta-glucosidase reaction for Gn was confirmed by HPLC techniques. Based on Km and K0 for Gn, subsite affinities (Ai, i = 1-6) were estimated as follows (kcal/mol): A1 = 1.3, A2 = 5.2, A3 = 0.65, A4 = -0.10, A5 = -0.65, and A6 = -0.26, of which A1-A3 are much higher than those of the beta-glucosidase of Candida wickerhamii. The subsite structure is quite similar to that of the alpha-glucosidase of A. niger, whereas the dependence of k0 on n is highly characteristic for beta-glucosidase, and decreases with n, suggesting some interaction between the particular subsites.

Expression of a zeatin-O-glucoside-degrading beta-glucosidase in Brassica napus

A beta-glucosidase was purified from seeds of Brassica napus L. (oilseed rape). The 130-kD native enzyme consisted of a disulfide-linked dimer of 64-kD monomers. Internal amino acid sequences were used to construct degenerate primers for polymerase chain reaction-mediated cloning of cDNA for the enzyme. One nearly full-length and one partial beta-glucosidase-encoding cDNA clone were isolated and sequenced. Southern hybridization showed that beta-glucosidase is encoded by a small gene family in B. napus. Northern hybridization showed that the genes are expressed in the seed, with a low degree of expression in other tissues. In the seed, the expression started at 30 days after pollination (DAP), with the highest expression at 40 DAP. The size of the transcript was approximately 1900 nucleotides. In situ hybridization to developing seeds of B. napus showed that the beta-glucosidase expression started at 30 DAP around the provascular tissue in the embryo axis. In the cotyledons, mRNA initially accumulated around the provascular tissues but was detected first at 35 DAP. At 40 DAP, expression occurred in most parts of the seed. In situ hybridization also detected beta-glucosidase mRNA in shoots, young roots, and the basal part of the hypocotyls. Zeatin-O-glucoside was identified as a natural substrate for B. napus beta-glucosidase.

Cyanide detoxification in cassava for food and feed uses

Cassava (Manihot esculenta Crantz) is an important tropical root crop providing energy to about 500 million people. The presence of the two cyanogenic glycosides, linamarin and lotaustralin, in cassava is a major factor limiting its use as food or feed. Traditional processing techniques practiced in cassava production are known to reduce cyanide in tubers and leaves. Drying is the most ubiquitous processing operation in many tropical countries. Sun drying eliminates more cyanide than oven drying because of the prolonged contact time between linamarase and the glucosides in sun drying. Soaking followed by boiling is better than soaking or boiling alone in removing cyanide. Traditional African food products such as gari and fufu are made by a series of operations such as grating, dewatering, fermenting, and roasting. During the various stages of gari manufacture, 80 to 95% cyanide loss occurs. The best processing method for the use of cassava leaves as human food is pounding the leaves and cooking the mash in water. Fermentation, boiling, and ensiling are efficient techniques for removing cyanide from cassava peels.

Contribution of selected fungi to the reduction of cyanogen levels during solid substrate fermentation of cassava

The effect of six individual strains of the dominant microflora in solid substrate fermenting cassava on cyanogen levels was examined. Six out of eight batches of disinfected cassava root pieces were incubated for 72 h after inoculation with either of the fungi Geotrichum candidum, Mucor racemosus, Neurospora sitophila, Rhizopus oryzae and Rhizopus stolonifer, or a Bacillus sp., isolated from on-farm fermented cassava flours from Uganda. One non-inoculated batch was incubated as a reference. Levels of initial and final moisture and cyanogens were assayed. The experiment was done in quadruplicate. Incubation of disinfected root pieces reduced cyanogenic glucoside levels significantly to 62.7% (SD 2.8) of the initial value. Microbial growth resulted in significant additional reduction of the cyanogenic glucoside levels to 29.8% (SD 18.9) of the levels which were obtained after non-inoculated incubation. Among the tested strains, N. sitophila reduced cyanogenic glucoside levels most effectively, followed by R. stolonifer and R. oryzae. Of all fermented samples, both Rhizopus spp. showed highest proportion of residual cyanogens in the cyanohydrin form. Flours showed similar patterns of cyanogens as the batches they were prepared from. Cyanogenic glucoside level reduction was significantly correlated (r = 0.86) with the extent of root softening. It is concluded that both incubation and microbial activity are instrumental in reducing the potential toxicity of cassava during the solid substrate fermentation and that effectiveness varies considerably between the species of microorganisms applied.

Reducing cassava toxicity by heap-fermentation in Uganda

Processing of cassava roots by the Alur tribe in Uganda includes a stage of solid substrate fermentation in heaps. Changes in cyanogen levels during the process, microflora involved, and protein levels, amino acid patterns and mycotoxin contamination of the final products were studied. Processing was monitored at six rural households and repeated at laboratory site, comparing it to sun-drying. Flour samples from rural households were analysed for residual cyanogens, mutagenicity, cytotoxicity and aflatoxins. Mean (+/- SD) total cyanogen levels in flours collected at rural households were 20.3 (+/- 16.8) mg CN equivalents kg-1 dry weight in 1990 (n = 23) and 65.7 (+/- 56.7) in 1992 (n = 21). Mean (+/- SD) levels of cyanohydrins plus HCN were 9.1 (+/- 8.7) in the 1992 flours. Total cyanogen levels in the village monitored batches were reduced considerably by heap-fermentation from 436.3 (+/- 140.7) to 20.4 (+/- 14.0) mg CN equivalents kg-1 dry weight cassava. Residual cyanogen levels were positively correlated with particle size of the resulting crumbs. Heap-fermentation was significantly more effective in reducing cyanogen levels than sun-drying alone, but did not always result in innocuous levels of of cyanogens. Dominant mycelial growth was from the fungi Neurospora sitophila, Geotrichum candidum and Rhizopus oryzae. No mutagenicity, cytotoxicity nor aflatoxins could be detected in the flours. Protein quantity and quality were not significantly reduced. Cassava gel viscosity pattern was modified to the consumers' preference by this method. As the removal of cyanogens was more efficient and we found no new obvious health risk, heap-fermentation can be regarded as an improvement compared to sun-drying alone in areas where cassava varieties with higher cyanogen levels prevail, but we recommend optimisation of the process for ensuring still safer products.

Tissue Level Compartmentation of (R)-Amygdalin and Amygdalin Hydrolase Prevents Large-Scale Cyanogenesis in Undamaged Prunus Seeds

Plum (Prunus domestica) seeds, which contain the cyanogenic diglucoside (R)-amygdalin and lesser amounts of the corresponding monoglucoside (R)-prunasin, release the respiratory toxin HCN upon tissue disruption. Amygdalin hydrolase (AH) and prunasin hydrolase (PH), two specific [beta]-glucosidases responsible for hydrolysis of these glucosides, were purified to near homogeneity by concanavalin A-Sepharose 4B and carboxymethyl-cellulose chromatography. Both proteins appear as polypeptides with molecular masses of 60 kD upon sodium dodecyl sulfate-polyacrylamide gel electrophoresis, but they exhibit different isoelectric points (PH, 5.6-6.0; AH, 7.8-8.2). AH and PH were localized within mature plum seeds by tissue printing, histochemistry, and silver-enhanced immunogold labeling. As was previously shown in black cherry (Prunus serotina) seeds (E.Swain, C.P. Li, J.E. Poulton [1992] Plant Physiol 100: 291-300), AH and PH are restricted to protein bodies of specific procambial cells and are absent from the cotyledonary parenchyma, bundle sheath, and endosperm cells. In contrast, the cyanogenic glycosides in both plum and black cherry seeds, which were detected by tissue printing, occur solely in the cotyledonary parenchyma and are absent from the procambium and endosperm. It is concluded that tissue level compartmentation prevents large-scale cyanoglycoside hydrolysis in intact Prunus seeds.

Tissue and Subcellular Localization of Enzymes Catabolizing (R)-Amygdalin in Mature Prunus serotina Seeds

In black cherry (Prunus serotina Ehrh.) homogenates, (R)-amygdalin is catabolized to HCN, benzaldehyde, and d-glucose by the sequential action of amygdalin hydrolase, prunasin hydrolase, and mandelonitrile lyase. The tissue and subcellular localizations of these enzymes were determined within intact black cherry seeds by direct enzyme analysis, immunoblotting, and colloidal gold immunocytochemical techniques. Taken together, these procedures showed that the two beta-glucosidases are restricted to protein bodies of the procambium, which ramifies throughout the cotyledons. Although amygdalin hydrolase occurred within the majority of procambial cells, prunasin hydrolase was confined to the peripheral layers of this meristematic tissue. Highest levels of mandelonitrile lyase were observed in the protein bodies of the cotyledonary parenchyma cells, with lesser amounts in the procambial cell protein bodies. The residual endosperm tissue had insignificant levels of amygdalin hydrolase, prunasin hydrolase, and mandelonitrile lyase.

A molecular and biochemical analysis of the structure of the cyanogenic beta-glucosidase (linamarase) from cassava (Manihot esculenta Cranz)

The cyanogenic beta-glucosidase (linamarase) of cassava is responsible for the first step in the sequential break-down of two related cyanoglucosides. Hydrolysis of these cyanoglucosides occurs following tissue damage and leads to the production of hydrocyanic acid. This mechanism is widely regarded as a defense mechanism against predation. A linamarase cDNA clone (pCAS5) was isolated from a cotyledon cDNA library using a white clover beta-glucosidase heterologous probe. The nucleotide and derived amino acid sequence is reported and five putative N-asparagine glycosylation sites are identified. Concanavalin A affinity chromatography and endoglycosidase H digestion demonstrate that linamarase from cassava is glycosylated, having high-mannose-type N-asparagine-linked oligosaccharides. Consistent with this structure and the extracellular location of the active enzyme is the identification of an N-terminal signal peptide on the deduced amino acid sequence of pCAS5.

Purification and Partial Characterization of Maize (Zea mays L.) beta-Glucosidase

Maize (Zea mays L.) beta-glucosidase (beta-d-glucoside glucohydrolase, EC 3.2.1.21) was extracted from the coleoptiles of 5- to 6-day-old maize seedlings with 50 millimolar sodium acetate, pH 5.0. The pH of the extract was adjusted to 4.6, and most of the contaminating proteins were cryoprecipitated at 0 degrees C for 24 hours. The pH 4.6 supernatant from cryoprecipitation was further fractionated by chromatography on an Accell CM column using a 4.8 to 6.8 pH gradient of 50 millimolar sodium acetate, which yielded the enzyme in two homogeneous, chromatographically different fractions. Purified enzyme was characterized with respect to subunit molecular weight, isoelectric point, amino acid composition, NH(2)-terminal amino acid sequence, pH and temperature optima, thermostability, and activity and stability in the presence of selected reducing agents, metal ions, and alkylating agents. The purified enzyme has an estimated subunit molecular mass of 60 kilodaltons, isoelectric point at pH 5.2, and pH and temperature optima at 5.8 and 50 degrees C, respectively. The amino acid composition data indicate that the enzyme is rich in Glx and Asx, the sum of which approaches 25%. The sequence of the first 20 amino acids in the N-terminal region was H(2)N-Ser-Ala-Arg-Val-Gly-Ser-Gln-Asn-Gly-Val-Gln-Met-Leu-Ser-Pro-(Ser?) -Glu-Ile-Pro-Gln, and it shows no significant similarity to other proteins with known sequence. The enzyme is extremely stable at 0 to 4 degrees C up to 1 year but loses activity completely at and above 55 degrees C in 10 minutes. Likewise, the enzyme is stable in the presence of or after treatment with 500 millimolar 2-mercaptoethanol, and it is totally inactivated at 2000 millimolar 2-mercaptoethanol. Such metal ions as Hg(2+) and Ag(+) reversibly inhibit the enzyme at micromolar concentrations, and inhibition could be completely overcome by adding 2-mercaptoethanol at molar excess of the inhibitory metal ion. The alkylating agents iodoacetic acid and iodoacetamide irreversibly inactivate the enzyme and such inactivation is accelerated in the presence of urea.

Immunocytochemical Localization of Mandelonitrile Lyase in Mature Black Cherry (Prunus serotina Ehrh.) Seeds

Mandelonitrile lyase (MDL, EC 4.1.2.10), which catalyzes the reversible dissociation of (R)-(+)-mandelonitrile to benzaldehyde and hydrogen cyanide, was purified to apparent homogeneity from mature black cherry (Prunus serotina Ehrh.) seeds by conventional protein purification techniques. This flavoprotein is monomeric with a subunit molecular mass of 57 kilodaltons. Glycoprotein character was shown by its binding to the affinity matrix concanavalin A-Sepharose 4B with subsequent elution by alpha-methyl-d-glucoside. Upon chemical deglycosylation by trifluoromethanesulfonic acid, the molecular mass was reduced to 50.9 kilodaltons. Two-dimensional gel analysis of deglycosylated MDL revealed the presence of several subunit isoforms of similar molecular mass but differing slightly in isoelectric point. Polyclonal antibodies were raised in New Zealand white rabbits against deglycosylated and untreated MDL. Antibody titers were determined by enzyme linked immunosorbent and dot immunobinding assays, while their specificities were assessed by Western immunoblot analysis. Antibodies raised against untreated lyase recognized several proteins in addition to MDL. In contrast, antisera raised against deglycosylated MDL were monospecific and were utilized for developmental and immunocytochemical localization studies. SDS-PAGE and immunoblotting analysis of seed proteins during fruit maturation showed that MDL first appeared in seeds shortly after cotyledons began development. In cotyledon cells of mature seeds, MDL was localized primarily in the cell wall with lesser amounts in the protein bodies, whereas in endosperm cells, this labeling pattern was reversed. N-terminal sequence data was gathered for future molecular approaches to the question of MDL microheterogeneity.