Cassava Flour Session 4 Bioconversion

8/10/2019 Cassava Flour Session 4 Bioconversion

1/53

SESSION4 :

BIOCONVERSION AND

BYPRODUCTUSE


2/53

187

Fermenta tion in Cassa va Bioconversion

Introduction

Cassava fermentation is traditionallypracticed in the tropics. But bothtechnology and productcharacteristics differ according toregion and sociocultural conditions:gar iin East and West Africa,ch i kw angueor fu fuin Central Africa,and sour starch in Latin America. Butthey have in common the aim toeliminate the poisonous cyanidecomponents and conserve cassava by

lactic acidification.

The essential role of lactic acidbacteria in the three products wasdemonstrated by studies carried outby the Institut franais de recherchescientifique pour le dveloppement encoopration (ORSTOM) through the

CHAPTER2 2

FERMENTATIONINCASSAVA

BIOCONVERSION1

M. Raimba ul t*, C. Ramrez Toro**, E. Girau d***,

C. Soccol., andG. Saucedo

STD2 Program of the European Union(EU), otherwise known as Improvingthe Quality of Traditional FoodsProcessed from Fermented Cassava(Raimbault, 1992; Saucedo et al.,1990).

When producing gar i, lacticacidification of cassava is rapid anddetoxification is sometimesincomplete. Controlling throughinoculation would improve quality.For fu fuor ch i kwangue, retting isessential for texturing and detoxifyingthe cassava. Lactic acid fermentationis heterolactic, operating inassociation with secondary alcoholicand anaerobic fermentation toproduce alcohol and organic acidssuch as butyrate, acetate, andpropionate that develop specialaromatic and organolepticcharacteristics. As for gar i,fermentation for sour starch

(especially in Colombia and Brazil) ishomolactic, but takes 3 or 4 weeks.Amylolytic lactic acid bacteria havebeen isolated from ch i kw anguebyORSTOM scientists and from sourstarch by CIRAD scientists.

A. Brauman isolated a new strain,Lactobaci llus p lantaru m A6, which wasdescribed by Giraud et al. (1991). Itsphysiological and enzymological

characteristics for cultivation oncassava starch media, amylaseproduction, and biochemical

* Institut franais de recherche scientifiquepour le dveloppement en coopration(ORSTOM), stationed in Cali, Colombia.

** Laboratorio de Bioconversin, Departamentode Procesos Qumicos y Biolgicos, Facultadde Ingeniera, Universidad del Valle, Cali,Colombia.

*** ORSTOM, Montpellier, France. Laboratrio de Procesos Biotecnologia,

Departamento de Tecnologia Qumica,Faculdade de Engenharia, UniversidadeFederal de Paran, Brazil.

Departamento de Biotecnologa, Universidad

Autnoma Metropolitana (UAM), Iztapalpa,Mexico.

1. No abstract was provided by the authors.


3/53

188

Cassa va Flour and Starch: Progress in Research and Development

properties have now been described(Giraud et al., 1992; 1993a; 1993b).

ORSTOM scientists have beenresearching solid fermentationcultivation of fungi on cassava andamylaceous components for more than10 years. Soccol et al. (1994) showedthat protein enrichment is possible bycultivating various strains of Rhizopus,even on crude, nongelatinized cassavaflours. Saucedo et al. (1992a; 1992b;1992c) studied, at the ORSTOMLaboratory, Montpellier, the growthand alcohol fermentation of cassavastarch in solid-state fermentation,using a highly promising amylolyticyeast.

Swedish and African researchershave described the beneficial effects oflactic acid fermentation on theprophylactic and keepingcharacteristics of those traditionalfoodstuffs made from fermentedcassava, maize, and mixed cereals,and of baby foods. These foods tendto increase childrens resistance to

diarrhoea.

All these studies are beingcontinued in new projects comprisingthe EU-STD3 Program. Other EUstudies are being conducted oncassava quality, environment, physicalprocessing, and transformation at alow industrial scale to take advantageof the economic and commercialopportunities in Latin America.

Sol id-State Ferme ntat ion of Cassava an d St arch y

Products

For more than 15 years, an ORSTOM

group has worked on a solid-statefermentation process for improvingthe protein content of cassava,potatoes, bananas, and other starchycommodities used for animal feed.Fungi, especially from the Aspergi l lusgroup, are used to transform starchand mineral salts into fungal proteins(Oriol et al., 1988a; 1988b; Raimbaultand Alazard, 1980; Raimbault andViniegra, 1991; Raimbault et al.,1985). Table 1 shows the overallchanges in composition between theinitial substrate and final products.Through such techniques acassava-fermented product with an18%-20% protein content (dry matterbasis) was obtained.

More recently, Soccol et al.(1993a; 1993b), also at the ORSTOMLaboratory, obtained good resultswith the Rhizopusfungi, of special

interest in traditionally fermentedfoods. In particular, they studied theeffect of cooking before fermentationon the availability of starch, proteincontent, and the rate of starchsbioconversion into protein (Table 2).They found that a selected strain ofRhizopus oryzaecould transformuncooked cassava, which containsonly 1.68% protein, into a fermentedcassava containing 10.89% protein.

Table 1. Effects of Aspergil lus nigeron protein and sugar contents of different starches (percentage of drymatter) after 30 h of fermentation in solid-state culture.

Substrate Initial composition Final composition

Proteins Sugar Proteins Sugar

Cassava 2.5 90 18 30

Banana 6.4 80 20 25

Banana waste 6.5 72 17 33Potato 5.1 90 20 35

Potato waste 5.1 65 18 28


4/53

189


Table 2. Growth of Rhizopus oryzaein solid-state cultivation on cassava granules after various cookingtreatments.

Treatmenta Dry matterb Total sugarc Proteinsc

Initial Final Initial Final Initial Final

I 60.90 46.48 80.01 46.78 1.20 11.69 II 59.18 45.35 84.11 60.72 1.61 12.40

III 57.95 42.12 82.44 52.57 1.56 13.93

IV 55.63 43.88 82.49 56.62 1.47 11.89

V 45.57 37.88 82.04 56.62 1.68 10.89

a. Treatment:

I = Cassava autoclaved for 30 min at 120 C, frozen, dried, and ground

II = Cassava flour (40% water) autoclaved for 30 min at 120 C

III = Cassava flour (30% water) autoclaved for 30 min at 120 C

IV = Cassava flour (30% water) vapor cooked for 30 min at 100 C

V = Untreated crude cassava flour

b. g/100 g total weight.

c. g/100 g dry matter.

SOURCE: Soccol et al., 1994.

Table 3 shows results of amylasebiosynthesis in solid or liquidculture, using raw or cookedcassava. The amount ofglucoamylase was 10 to 15 times

higher in solid than in liquid culture,and higher in raw starch mediumthan in cooked cassava.

This work is being continued inthe EU-STD3 Program at theBioconversion Laboratory of theUniversidad del Valle, Cali,Colombia. It focuses on simplifyingcassava processing by learning moreabout the specificity of Rhizopus

strains in degrading the raw starchgranule. But clean flours of rawcassava are needed. The commonflours of cassava contain too muchnatural microflora to allow microbialstudies with fungi; they must first besterilized and (unfortunately)gelatinized. Ramrez et al. (1994)developed raw cassava flour with avery low content of bacteria andfungi, and little gelatinization.

To measure gelatinization, thesimple method of Wotton et al.

(1971) was adopted and a goodcorrelation coefficient for thecalibration curve was obtained.Table 4 shows the effect of thermictreatment and microwaves on starch

gelatinization in cassava flour (watercontent typically lower than 10%).Where water content was very low,gelatinization was also low.

The same thermic treatment ofdry cassava flour eliminated thenatural microflora contained in rawflour, from 109bacteria/g of dry flourto fewer than 103bacteria/g afterheating the flour for 30 min at

90 C. With gelatinization limited toless than 5% under such conditions,obtaining clean, raw cassava flour ispossible in the laboratory.

Figures 1 and 2 show the effectsof various physical and thermictreatments on the bacteria content ofcassava flour. Cassava flour will beused as a solid substrate forcultivating Rhizopusstrains, and to

compare the capacity of selectedstrains to grow on raw or gelatinizedcassava starch.


5/53

190

Table 3. Effect of cooking and type of culture on the growth and amylases of various strains of Rhizopus oryza ecultivated on

Strain of Liquid-state culturea SolidRhizopus

Raw cassava Cooked cassava Raw cassava

Gluco- Protein Gluco- Protein Gluco- Protein

amylase amylase (g/100 g amylase amylase (g/100 g amylase amylase (g/100(U/g DM) (U/g DM) DM) (U/g DM) (U/g DM) DM) (U/g DM) (U/g DM) DM)

28168 42.20 9.60 3.90 157.20 3.10 10.00 39.30 55.30 10.60

34612 40.40 7.30 4.60 168.50 5.70 9.30 55.00 70.00 12.60

28627 76.00 7.80 4.00 145.40 3.30 9.60 98.00 108.00 11.40

a. DM = dry matter; U = enzyme units.

SOURCE: Soccol et al., 1994.


6/53

191


Table 4. Effect of temperature and microwaves on starch gelatinization of cassava flour.

Temperature Time Gelatinization rate (%)a

(min)Exp. 1 Exp. 2 Exp. 3 Mean

Test 1 75.439 84.063 88.911 82.80

(80% gel.)

Test 2 25.411 26.184 29.702 27.10

(20% gel.)

80 C 60 3.529 3.444 2.714 3.23

85 C 30 3.529 3.357 3.487 3.46

85 C 3.444 3.486 3.444 3.46

90 C 30 3.572 3.444 3.572 3.53

90 C 60 9.454 9.064 9.107 9.21

95 C 30 6.961 5.546 5.803 6.10

100 C 30 4.965 4.602 4.001 4.52

105 C 30 6.961 5.503 5.301 5.92

120 C 30 4.816 4.730 4.473 4.67

140 C 30 4.773 3.100 3.100 3.66160 C 30 3.529 3.487 4.301 3.77

Autoclaving 15 3.572 3.100 4.301 3.66(121 C)

Microwaves 5 2.886 2.410 2.842 2.71(Pot. 70)

Microwaves 5 2.971 2.242 2.242 2.49(Pot. 100)

Microwaves 15 3.879 3.057 3.915 3.62(Pot. 30)

a. Exp. = Experiment. Mean is across the experiments.

Duration of treatment (minutes)

Figure 1. Total microflora (plate count analysis)

in cassava flour, according totreatment. ( = ultra-violet radiation; = microwaves; = 80 C; = 85 C; = 90 C.)

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3 1 2 3 1 2 3

Viablebacteria(n/gofflour)

0 80 85 90 95 100105120140 160180

Temperature (C)

Figure 2. Effect of temperature on bacterial

population in cassava flour.

1010

109

108

107

106

105

104

103

102

109

108

107

106

105

104

103

102

101

Viablebacteria(n/gofflour)

0 30 60 90 120


7/53

192


Lactic Acid Ferme nt ationof Cass ava

Lactic acid fermentation is importantfor many traditional fermented foods,silage, and animal feed, and forrecycling agroindustrial byproducts.Because of its acid, bacteriostatic,and bactericidal properties,fermentation preventsmicroorganisms, whether parasitic,saprophytic, or pathogenic, frombreaking down vegetable material.

In tropical countries, lacticfermentation not only plays animportant role in the traditionaltransformation of starchy foods,such as cassava, but also in thetransformation and conservation ofother foods, and fish and itsbyproducts. Two types of lacticfermentation exist:

(1) Homolactic, when more than80% of total acidity andmetabolites formed consists oflactic acid, and

(2) Heterolactic, when thepercentage of acetic acid,propionic acid, and ethanol ismore significant, and lactic acidrepresents 50%-80% of totalacidity.

Lactic bacteria produce two typesof lactic acid: L(+) and D(-). Only theL(+) form is assimilated by humans.

Previous studies, realized duringthe EU-STD2 Program in 1988-1991(Raimbault, 1992), consisted ofimproving traditional fermented foodmade from cassava in Africa andLatin America. Three kinds oftraditional foods were considered:gar i, ch i kwangue, and sour starch.We demonstrated the essential roleof lactic acid bacteria in alltraditional processes.

Amylolytic lactic bacteria wereisolated from fermented cassava.

The first strain of Lactobaci l lusp lan ta rumto be described as havingvery high amylolytic capacity wasobtained from fermented cassava byA. Brauman in the Congo. Detailedphysiological and biochemicalcharacterization of this new strainis expected to be published soon byE. Giraud.

Mbugua and Njenga (1991) andSvanberg (1991a; 1991b), workingin Tanzania and at the UppsalaUniversity, respectively, havereported on the effect of lactic acidfermentation on the pathogenmicroflora content of traditionalAfrican foods.

Some of their results arereported in Table 5 and Figure 3,which show how lactic acidbacteria reduce the number offood-poisoning pathogens such asspecies of Staphylococcus,Salmonel la, and Shigella, andEscherichiacoli. High levels of suchpathogens are sometimes found in

traditional foods after processingunder unhygienic conditions,especially those for malting maizeduring the rainy season in parts oftropical Africa.

Lactic fermentation oftraditional foods reduces pathogenicbacteria from 108to 103. The sameauthors also found a significantcorrelation between the resistance

of young children to diarrhoea andeating acidified gruels.

We are bioconverting, throughprobiotics and bactericides, cassavaflour and starch containingamylolytic lactic acid bacteria toisolate new strains from traditionalfoods. At the same time, we arebroadening knowledge on thecultivation of lactic acid bacteria in

starchy substrates. We hope suchinformation will help elaborate newfood and feed products.


8/53

193


Table 5. Effect of lactic acid fermentation on the content of pathogenic bacteria in traditional fermentedfoods in Africa.

Time (h) Log number of bacteria/g food

Control Nonfermented, Fermented foodacidified

food Flour Gruel(nonviable) (viable)

Shigella flexneri

0 6.8 6.7 6.4 6.0

3 6.6 5.8 5.1 4.0

7 7.0 4.2 5.5 3.3

24 7.0 4.1 3.7 2.7

Salmonel la typhimur iu m

0 8.5 8.1 8.3 7.7

3 8.0 6.7 6.0 7.1

7 7.9 5.3 4.4 6.3

24 8.9 4.0 2.0 2.0

SOURCE: Lorri and Svanberg, 1988.

production. This may be because,first, cassava cultivation yieldsrelatively few, commercially significantbyproducts, compared with, forexample, sugarcane which yieldsenormous quantities of bagasse, avaluable source of energy fordistillation. Second, cassava starchneeds to be hydrolyzed into sugar forbioconversion into ethanol by thecommon Saccha romy ces cerevisiae.This implies an additional, costly step.

For cassava to be an economicallyviable energy source, its processingcosts must be reduced. Solid-statefermentation is one, simple, and newmethod of reducing costs: the use of an

amylolytic yeast that eliminateshydrolysis.

At the ORSTOM Laboratory,Saucedo et al. (1992a; 1992c)developed a new process for the solidculture of an amylolytic yeast,Schw an niomy ces castel i i (Figure 4).The main advantage of this techniqueis its continuous recuperation ofethanol in a cold trap condenser. The

gas produced in the reactor is pumpedthroughout the system, thus ensuringits continual removal from the medium

Logviablebact

eria(n/ggruel)

Figure 3. Evolution of pathogenic bacteria duringthe lactic fermentation of u ji, a

fermented cassava gruel (after Mbuguaand Njenga, 1991). ( = Staphylococcusaureus; = Salmonellatyph imur ium; = Escherichiacoli; = Shigelladysenter iae.)

Time (h)

10

9

8

7

6

5

4

3

Alcoholic Fermentation ofCassava and Starch Products

Cassava is a potential producer ofethanol, considering its potentially

high yields and low costs. Yet fewreports concern the industrialapplication of cassava for ethanol

0 10 20 30 40 50 60


9/53

194


Table 6. Comparison of various processes for ethanol production from cassava in liquid or solid substrate.

Process Hydrolysis Sugar Ethanol Recovered Theoretical

(g/L) (g/L) (g/L) (%)

Liquid substrate, using S. cerevisia e a, b + 145 72.50 72.50 83.2

Solid substrate, using S. cerevisia e b, c + 165 41.73 41.73 65.0

Solid substrate, using Rhizopus koji d - 200 110.00 110.00 83.0

Solid substrate, using Schw . casteli i e, f - 300 68.40 212.60 64.0

a. Saraswati, 1988.

b. S.= Saccharomyces.

c. Jaleel et al., 1988.

d. Jujio et al., 1984.

e. Schw. = Schwanniomyces.

f. Saucedo et al., 1992a.

Figure 4. Producing ethanol through solid-substrate fermentation of cassava starch. The reactor containsa solid support impregnated with a starchy suspension and inoculated with the fermentationagent, an amylolytic yeast known as Schw anniomy cescasteli i. The resulting gas is pumped to acondenser where ethanol is extracted. The residual gas is sent to a humidifier.

potential of cassava as a substratefor ethanol production. The solid-state technique has to be carefullyconsidered. Results obtained withthe fungus Rhizopus koj iareparticularly significant. Thepotential of Schw ann iomy cesis alsointeresting because amylolytic yeastwould be easier to control at thesmall-scale industrial level.

Column tohumidify gas

Pump

Continuousextraction ofethanol

Cold trapcondenser

Reactor

Ethanol

and limiting its toxic effects on theyeasts metabolism. The resultsobtained by Saucedo et al. (1992a;1992b; 1992c) were promising, butthe technology and feasibility of theprocess for commercial operationneed further research.

Table 6 shows the resultsobtained by various authors on the


10/53

195


__________; Gosselin, L.; and Raimbault, M.1992. Degradation of cassavalinamarin by lactic acid bacteria.Biotechnol. Lett. 14(7):593-598.

__________; __________; Marn, B.; Parada,J. L.; and Raimbault, M. 1993a.

Purification and characterization ofan extracellular amylase fromlactobaci l lus p lantarumstrain A6.

J. Appl. Bacteriol. 75:276-282.

__________; __________; and Raimbault, M.1993b. Production of aLactobaci l lus p lantarumstarter withlinamarase and amylase activities forcassava fermentation. J. Sci. Food

Agric. 62:77-82.

Jaleel, S. A.; Srikanta, S.; Ghildyal, N. P.;

and Lonsane, B. K. 1988.Simultaneous solid phasefermentation and saccharification ofcassava fibrous residue forproduction of ethanol. Starch/Strke40(2):55-58.

Lorri, W. S. M. and Svanberg, U. 1988.Improved protein digestibility incereal based weaning foods by lacticacid fermentation. Harare, Zimbabwe.

Mbugua, S. K. and Njenga, J. 1991.

Antimicrobial properties of fermentedUJI as a weaning food. In: Westby, A.and Reilly, P. J. A. (eds.). Traditional

African foods: quality and nutrition.International Foundation of Science(IFS), Sweden. p. 63-67.

Oriol, E.; Raimbault, M.; Roussos, S.; andViniegra-Gonzlez, G. 1988a. Waterand water activity in the solid statefermentation of cassava starch byAspergi l lus niger. Appl. Microbiol.Biotech. 27:498-450.

__________; Schetino, B.; Viniegra-Gonzlez,G.; and Raimbault, M. 1988b. Solidstate culture of Aspergi l lus nigeronsupport. J. Ferment. Technol.66:1-6.

Raimbault, M. 1992. Etudes physiologiqueset gntiques des bactrieslactiques dans les fermentationstraditionnelles du manioc. Finalreport CEE/STD2, no. TS2A-00226.Institut franais de recherchescientifique pour le dveloppement encoopration (ORSTOM), Montpellier,France. p. 1-53. (Internaldocument.)

Conclusions onBioc onve rting Cassava and

Poten tial Product s

To bioconvert cassava starch and

flour to elaborate new products,ORSTOM, CIRAD, and collaboratinginstitutes are emphasizing twoapproaches: solid-state fermentation,and lactic acid fermentation.

The first is of great interestbecause of its potential to simplifyprocesses and reduce costs, and itslarge reactor volume. Both Rhizopusand Schw ann iomy ces(or otheramylolytic) yeasts can be used in asolid-state cultivation process. Thisimplies a three-phase reactor with asolid fiber support, a liquid phasecontaining the substrate insuspension and salts, and a gaseousphase for exchanging volatilecomponents, that is, oxygen, water,and ethanol.

In lactic acid fermentation, weare investigating the culture control

of amylolytic lactic acid bacteria inmixed and composite starters able toremain competitive in a natural,nonaxenic environment. Theprophylactic role of lactic acidbacteria is also of great interest.

Finally, we are studyingmicroorganisms able to degradenative cassava starches withoutneed of gelatinization, as in in

natural biotransformation andbiodegradation. We will alsostudy the amylolytic capacity ofRhizopusspp., yeasts, and lactic acidbacteria.

References

Giraud, E.; Brauman, A.; Klke, S.; Lelong,B.; and Raimbault, M. 1991.Isolation and physiological study ofan amylolytic strain of Lactobaci l lusp lan ta rum. Appl. Microbiol.Biotechnol. 36:379-383.


11/53

196


__________ and Alazard, D. 1980. Culturemethod to study fungal growth insolid fermentation. Eur. J. Appl.Microbiol. Biotechnol. 9:199-209.

__________ and Viniegra, G. 1991. In:Chahal, D. S. (ed.). Modern and

traditional aspects of solid statefermentation in food, feed and fuelfrom biomass. p. 153-163.

__________; Revah, S.; Pina, F.; and Villalobos,P. 1985. Protein enrichment of cassava

by solid substrate fermentation usingmolds isolated from traditional foods.

J. Ferment. Technol. 63(4):395-399.

Ramrez, C.; de Stouvenel, A.; and Raimbault,M. 1994. Effect of physical treatmentson microflora content in cassava flour.

Poster presented at the InternationalMeeting on Cassava Flour and Starch,held in January 1994 at Cali,Colombia.

Saraswati. 1988. The experience of pilot plantof ethanol from cassava in Indonesia.Regional Workshop on Upgrading ofCassava/Cassava Wastes by

Appropriate Biotechnologies, Bangkok,Thailand, 1987. Thailand Institute ofScientific and Technological Research,Bangkok, Thailand. p. 41-49.

Saucedo, G.; Gonzlez, P.; Revah, S.; Viniegra,G.; and Raimbault, M. 1990. Effect ofLactobaci l l iinoculation on cassava(Man ihot esculenta) silage:fermentation pattern and kineticanalysis. J. Sci. Food Agric.50:467-477.

__________; Lonsane, B. K.; Navarro, J. M.;Roussos, S.; and Raimbault, M.1992a. Potential of using a singlefermenter for biomass build-up, starchhydrolysis and ethanol production:

solid state fermentation systeminvolving Schw anniomyces castel i i.

Appl. Biochem. Biotechnol. 36:47-61.

__________; __________; __________; __________;and __________. 1992b. Importance ofmedium pH in solid state fermentationsystem for growth of Schw anniomy cescastel i i. Lett. Appl. Microbiol.15:164-167.

__________; __________; and Raimbault, M.1992c. Maintenance of heat and

water balance as scale-up criterionfor production of ethanol bySchw anniomy ces castel i iin solidstate fermentation system. ProcessBiochem. 27:97-107.

Soccol, C.; Iloki, I.; Marn, B.; andRaimbault, M. 1994. Comparativeproduction of alpha-amylase,glucoamylase and proteinenrichment of raw and cookedcassava by Rhizopusstrains insubmerged and solid statefermentations. J. Food Sci. Technol.31:320-332.

__________; Marn, B.; Roussos, S.; andRaimbault, M. 1993a. Scanning

electron microscopy of thedevelopment of Rhizopus arrh izuson raw cassava by solid statefermentation. Micol. Neotrop. Apl.6:27-39.

__________; Rodrguez, J.; Marn, B.;Roussos, S.; and Raimbault, M.1993b. Growth kinetics of Rhizopusarrh izusin solid state fermentationof treated cassava. Biotechnol.

Tech. 7(8):563-568.

Svanberg, U. 1991a. Lactic fermentation ofcereal-based weaning gruels andimproved nutritional quality. In:

Westby, A. and Reilly, P. J. A. (eds.).Traditional African foods: qualityand nutrition. InternationalFoundation of Science (IFS),Sweden. p. 53-60.

__________. 1991b. The potential role offermented cereal gruels inreduction of diarrhoea among

young children. In: Westby, A. andReilly, P. J. A. (eds.). Traditional

African foods: quality and nutrition.International Foundation of Science(IFS), Sweden. p. 33-38.

Wotton, M.; Weedon, D.; and Munck, N.1971. A rapid method for estimationof starch gelatinization in processedfoods. Food Technol. Aust.23:612-614.


12/53

197

Cassa va Lactic Fermenta tion in Centr al Africa:...

CHAPTER2 3

CASSAVALACTICFERMENTATIONIN

CENTRALAFRICA:MICROBIOLOGICALANDBIOCHEMICALASPECTS

A. Brauma n*, S. Klke**, M. Malonga***,

O. Mavoungou***, F. Ampe, andE. Miam bi***

cyanogenic compounds (e.g.,concentration decreased from400 ppm in fresh cassava to 20 ppmin fermented mash); (2) a significantlysis of cassava cell walls due to thesimultaneous action of endogenouspectin methylesterase and bacterialpectin lyase; and (3) the production oforganic acids (C

2to C

4), mainly lactate

and butyrate, that contribute to thetypical flavors of ch i kwangueandf u fu.

In the study, most microflorainvolved in retting were facultative,anaerobic, fermentative bacteria,among which lactic bacteria werepredominant. From the second day offermentation, endogenousLactobaci l lusspecies were totallysupplanted by Leuconostocmesenteroides and La ctococcus la ctis.Anaerobic bacteria such asClos t r id ium buty r i cum were also found

and seemed responsible for initiatingbutyrate production. Yeasts playedno significant role, but theirincreasing number at the end of theprocess (Cand idaspecies) probablyinfluenced the conservation of endproducts.

Despite the significant number ofamylolytic bacteria (105-106b/ml), theamylase activity found in the retting

juice came from the roots anddisappeared after 48 h offermentation. The main enzymes of

Summary

Retting is a lactic fermentation duringwhich cassava roots are soaked forlong periods in water. Despite theimportance of this fermentation, nokinetic study of it has beenundertaken. Our study thereforeexamined the biological and physicalchanges of cassava roots duringretting to provide a basis for itspossible mechanization.

The study was carried out to(1) enumerate and characterize themain microorganisms of the process;(2) determine the evolution ofphysicochemical parameters duringretting; and (3) measure theproduction of organic products andsome principal enzyme activities.

Retting can be characterized bythree essential transformations of the

roots: (1) a degradation of endogenous

* Institut franais de recherche scientifiquepour le dveloppement en coopration(ORSTOM), Paris, France.

** Laboratoire de microbiologie, Directiongnrale de la recherche scientifique ettechnique (DGRST), Brazzaville, Congo.

*** Laboratoire de biologie cellulaire, Facult dessciences, Universit Marien-NGouabi,Brazzaville, Congo.

Laboratoire de microbiologie et debiotechnologie, ORSTOM, stationed inBrazzaville, Congo.


13/53

198


this process were cassava pectinmethylesterase, bacterial pectinase,and endogenous linamarase.

The pH became stable at about4.5 after 48 h and the partial oxygenpressure dropped to 0.2 mg/L after10 h.

These results suggested thatretting is a typical heterolacticfermentation with a significantproduction of butyrate.

Introduction

Processed cassava (Maniho tesculentaCrantz) is eaten in Westand Central Africa in such forms asgar i, l a f un, fu fu, ch i kwangue, andtapioca. In the Congo, the worldssecond largest cassava consumerafter Zare (Trche, n.d.), cassavaroots account for 47% of the calorieintake (Trche and Massamba,n.d.b).

The two main products associatedwith fermented cassava are fu fuandch i kwangue. The former is a flourobtained from sun-dried cassavamash that is pulverized. This flourmay be mixed with boiling water andserved in bowls with sauce and fish ormeat. Ch i kwangue, a cassava bread,is obtained after multiplepostfermentation steps, includingdefibering and pugging (Trche and

Massamba, n.d.a).

Both products require afermentation in which the roots soakfor 3 to 6 days in tap water. Duringthis process, cyanogenic compoundsare eliminated, flavor compoundsare elaborated, and the roots soften(Okafor et al., 1984; OladeleOgunsa, 1980). Softening isindispensable for further root

processing but the mechanismsinvolved are not yet fullyunderstood.

Significant differences exist inretting processes throughout CentralAfrica and even in the Congo. Peeledor unpeeled roots are retted in rivers,standing water, large barrels of water,or even buried in soil. Thefermentation temperature varies withseason and location. Suchdifferences, combined with the lowreproducibility of the local processors,lead to a variability in quality andtaste of cassava foods (Trche andMassamba, n.d.a).

To increase the quality of thesetraditional products and provide abasis for the possible mechanizationof the process, the European Union(EU) Program-STD2, known asImproving the Quality of TraditionalFoods Processed from FermentedCassava was set up in 1990 inCentral Africa and South America.Our laboratory was to describe themechanisms of root transformationduring retting with a view tooptimizing product quality andfermentation speed.

In this paper, we present the mainresults obtained during this EUprogram, describe the microbiologicaland biochemical evolution throughoutthe process, and define the origin(vegetal or microbial) of the mainenzymes.

Mate rial and Met ho ds

Or i g i n o f p l a n t m a t er i a l

Cassava roots (Man ihot esculentavar.MM 86, or Ngansa) were harvestednear Brazzaville, Congo, 18 monthsafter planting.

Ret t i n g p r o cedu r e s

About 100 kg of washed and peeled

roots were placed in a barrel and thevolume made up to 50 L with rainwater. A second barrel, filled only


14/53

199


with rain water, was used as controlfor physicochemical measurements(T C, pH, pO

2). Samples were taken

every 12 h for the first 2 days andthen every 24 h until retting wascompleted.

Samp l e p r e p a r a t i o n f o r ba c t er i a l

e n ume r a t i o n

Sampling was carried out byrandomly selecting six root sections,which were then cut into 0.5-cmcubes and mixed under sterileconditions. Of this mixture, 60 gwere extracted and diluted in 540 mlof sterile, peptonized water (dilution10-1). The solution was then mixed ina Blendor (Turnmix ME 88,SOFRACA, France) and seriallydiluted in sterile, peptonized water foraerobic counts and in anaerobicHungate tubes containing sterile,reduced water, flushed with 20% CO

2

and 80% N2for anaerobic counts.

Met h o d s of b a c t er i a l q u a n t i f i c a t i o n

Two types of enumeration wereperformed: most probable number(MPN) enumeration and plate countson solid medium. The MPN methodwas used to either ascertain thegrowth of fermentative andpectinolytic bacteria or count themetabolites produced during growthon appropriate media for anaerobic,lactate-using bacteria. For each MPNdetermination, four successive

dilutions of root samples wereinoculated in three or four tubes perdilution. Results were calculatedaccording to the McCready tables(McCready, 1918).

For plate counts, 0.1 ml samplesof appropriate dilutions wereinoculated in triplicate on agarmedium in plates. All the plateswere incubated at 30 C and the

number of colony-forming unitsdetermined after 48 or 72 h ofincubation.

Ba c t er i a l e n ume r a t i o n

La c t i c a c i d ba c t e r i a (l . a . b .).

The l.a.b. were enumerated on MRSagar medium (de Man et al., 1960),supplemented with 0.1% of anilineblue. In each petri dish, 0.1 ml ofappropriate root sample dilution wascovered with medium and kept at45 C. Enumeration was carried outafter a 48-h incubation at 30 C.Subcultures were further purified byrepeated plating.

Strains were differentiated intovarious bacterial groups by thefollowing tests: microscopyexamination, gram reaction, catalasetest, and oxygen metabolism(fermentative or oxidative) test in softMRS agar. Strains which were grampositive, catalase and oxidasenegative, nonmotile rods or cocci, andcolored by aniline blue wereconsidered as lactic bacteria.

Gl u c o se- an d l a c t a t e-f e rm en t i n g

b a c t e r i a . These bacteria (g.f.b. and

l.f.b., respectively) were enumeratedon a basal medium that contained theequivalent of 2 g/L glucose or 5 g/Lof lactate (used as a carbohydratesource); 0.5 g/L of trypticase andyeast extract; 0.5 g/L of cysteine HCl(used as a reductive agent); 0.1 g/Lof sodium acetate; 0.005 g/L ofresazurine; 20 ml of Widdel mineralsolution (Widdel and Pfennig, 1984);and 1 ml of Widdel trace element

solution (Widdel and Pfennig, 1984).

The Hungate technique (Hungate,1969), modified for using syringes(Macy et al., 1972), was usedthroughout the study. After boiling,the medium was cooled under acontinuous flow of oxygen-free N

2,

adjusted to a pH of 7.2 with NaOHsolution, and distributedanaerobically into Hungate tubes.

The medium was sterilized for 35 minat 110 C. Before inoculation, 1% ofNa

2S-9H

2O (5%) was added as a


15/53

200


reductive agent to each tube.Inoculations were performed withsyringes filled with oxygen-free N

2,

using a gas manifold.

Yeas t. A potato-dextrose agarmedium (PDA, DIFCO Laboratory) wasprepared, containing 0.05 g/L ofchloramphenicol and with a final pHof 3.5, adjusted with tartaric acid(10%). The agars surface was thendried. From an appropriate microbialdilution, 0.1 ml was spread, intriplicate, on plates containing themedium. The plates were thenincubated for 72 h at 30 C.Subcultures were further purified byrepeated plating on PDA. Isolateswere characterized to the genus level,and Api tests (API 5030 stripsBiomerieux, France) were used todetermine fermentation carbohydratesources.

Ph y s i c oc h em i c a l p a r am et e r s

Pen e t r om et r y i n d ex .

Penetrometry was used to indicate

root softening during retting. Aprevious study showed that apenetrometry index of 15 mm/5 scorresponded to the end of retting asit is traditionally evaluated (Braumanet al., n.d.). A penetrometer (PNR10-SUR, Berlin) was used to measurethe consistency of the roots. Every10 h, and for each experiment, sixroot sections were randomly chosen.Penetrometry depth was estimated

with six repetitions for each rootsection.

T h e pH a n d p a r t i a l o x y gen

p r es su r e of t h e r e t t i n g j u i c e.Every10 h, 50 ml of retting juice wasextracted to test the pH (measuredwith CG 838 pH-meter from SCHOTTGerte, Germany) and estimatepartial oxygen pressure (measuredwith OXI 91 from WTW, Germany).

T h e pH a n d p a r t i a l o x y gen

p r es su r e o f t h e r o ot s . A 20-g

sample was added to a Waringblender and mixed with 120 mldistilled water at low speed for 15 sand at high speed for 1 min. Themixture was then filtered through aGF/A filter and the volume made upto 200 ml with distilled water.Extracts were taken in duplicate at0 h, 48 h, and at the end of retting.Acidity was titrated with 0.01 MNaOH.

B i o ch em i c a l a n a l y si s

Enz ym e as s a y s . A sample of40 g of cassava mash was added to aWaring blender, together with 80 mlof 0.1 Mcitrate buffer (pH = 6.5) andthe mixture homogenized. Themixture was held overnight at 4 Cand centrifuged at 12,000 gfor30 min. The supernatant waslyophilized and resuspended in 1/10volume of citrate buffer.

-g l u c o si d a s e a c t i v i t y . Thiswas measured with a chromogen,p-nitrophenol--d-glucopyranoside,

at 20 mMin 0.1 Mof Na-phosphatebuffer (pH = 6.8) for 1 h at 25 C.The reaction was stopped by addingan equal volume of 0.2 Msodiumborate (pH = 9.8), and p-nitrophenolwas determined with aspectrophotometer at 400 nm (Hoseland Bartz, 1975).

L i n am a r a s e . This was assayedwith linamarin as substrate and by

measuring the appearance of CN-(Giraud et al., 1992). To 400 l ofextract, 100 l of 50 mMlinamarinin 0.1 Mcitrate buffer (pH = 6.0)were added. At regular intervals,50 l aliquots were added to 50 l of0.1 MNaOH to stop the reaction,and stored at 4 C. Cyanide wasliberated by adding 50 l of 0.1 MH

2SO

4and 850 l distilled water to

each aliquot, and was measured

with a spectroquant kit (Merck,Darmstadt, Germany). One unit oflinamarase was defined as the


16/53

201


amount of enzyme that released1 mol of CN-per minute.

Ac t i v i t y o f pec t i n est e r ase (PE ;

p ec t i n p ec t y l h y d r o l a s e,

EC 3 .1 . 1 . 11 ).This was assayed bytitrating 1 ml of extract in 1% pectinat 30 C (Grindsted RS400-DM 74%),and in 0.1 MNaCl and 1 mMNaN

3.

pH was increased to 7.0 with 0.01 MNaOH. One unit corresponds to theneutralization of 1 mol of COO-/min.

Po l y g a l a c t u r o na t e l y a s e (PGL )

a c t i v i t y . PGL activity was assayedby the Starr et al. (1977) procedure.This assay does not differentiatebetween endo-PGL (poly (1,4--d-galacturonide) lyase, EC 4.2.2.2) andexo-PGL (poly (1,4--d-galacturonide)exolyase, EC 4.2.2.9). One unit ofPGL corresponds to the formation of1 mol of one unsaturated bond ingalacturonide between C4 and C5.

Po l yga l a c t u r ona se (PG ; po l y

(1 ,4 -- d - g a l a c t u r o n i d e ) g l y c a noh y d r o l a s e , EC 3 . 2 . 1 . 1 5 ).

This was assayed by viscometry. To40 ml of 1% pectin in 100 mMofacetate buffer (pH = 4.7), 0.5 ml ofextract was added. The rate ofreduction in viscosity was measuredat 25 C in a viscometer (Haakemodel; VT 500, rotation: 150.93 s-1

and system MV-MV1). One unitcorresponds to the release of 1 molof hexose/min. Total activities areexpressed as units per 100 g of

cassava.

Ac t i o n o f pe ct i c en z ym es i n

v i vo .Sterilized slices of cassava wereinoculated with 50 l of enzymeextract or 5 l of purified pectolyticenzymes (endopolygalacturonaseP-5146 from Aspergi l lus n iger;pectolyase P-3026 from A. japonicum;and pectinesterase P-0764 fromorange peel) (Sigma, Saint-Quentin

Fallavier, France). The inoculatedslices were placed in sterile beakerscontaining 10 ml of 0.01 Mof citrate

buffer (pH = 5.0). Penetrometerreadings were estimated after 24 hand 48 h at 30 C.

Cel l u l a s e, am y l a s e, and

x y l a n a s e a ct i v i t i e s. These activitieswere also assayed at 37 C and pH of5.8, using the Somogyi procedure(Somogyi, 1945). The substrates weremicrocrystalline cellulose (100 mg)and xylan (18 mg/ml).

Ot h er a n a l y t i c a l met h o d s

Total and free cyanides were assayedby the Cooke et al. method (1978).Proteinwas determined with amodified Lowry procedure (Bensadounand Weinstein, 1976).

Or g a n i c c ompo u n d s

Sugars, volatile fatty acids (VFA), andlactate and ethanol concentrations inthe roots were determined byhigh-performance liquidchromatography (HPLC) of thesupernatant, as described by Giraud

et al. (1991). The resulting columns(BioRad Laboratories, Richmond,California) were:

(1) Fast carbohydrate column formonosugars analysis (100 x 7,8 min) with 0.6 ml flow of milliQwater (pH = 6.0) at 70 C;

(2) Aminex HP 42 A (300 x 7.8 minBiorad) for polyosides analysiswith 0.3 ml flow of milliQ water

(pH = 6.0) at 70 C;(3) Aminex HP x 87H column with

0.8 ml/min flow of H2SO

46 mMat

60 C.

Results and Discus s ion

K i n et i c s t u d i es of r et t i n g

We now present the results of ourglobal study of lactic fermentation.

Kinetic parameters such as total andfermentative microflora,physicochemical parameters, and


17/53

202


substrates and metabolites producedhave been measured throughout theprocess. These results are the meanof seven rettings performed in barrelsunder the same conditions.

Evo l u t i o n o f ph y s i c o chem i c a l

p a r am e t e r s

The main physicochemicalparameters were assayed throughoutthe process (Figure 1). The partialoxygen pressure dropped to wellbelow 1 mg/L after 10 h and the pHbecame stable (at 4.5) within 48 h.Conversely, root softening, indicated

by the penetrometry index, appearedafter 2 days of fermentation andevolved exponentially. This processseems to require anaerobic and acidicconditions to proceed. Microscopicexamination shows that the cassavacell walls were extensively disruptedat the end of the process,demonstrating the attack ofdepolymerizing enzymes.

The concentration of endogenouscyanogenic compounds decreasedfrom 300 mg/kg as HCN (dry matter

ppm

300

200

100

0

0 1 2 3 4 5

Figure 2. Total cyanide evolution.( = linamarin; = cyanhydrines +free cyanides; = free cyanides.)

Time (days)

basis) in fresh cassava to 20 in thefermented mash (Figure 2). In allassays, total cyanogens werealmost eliminated (90%). Theseresults demonstrated that, underthe standard conditions of localtransformations in Central Africa,

detoxification occurred normallywithout need of an additionalprocess.

Figure 1. The evolution of physicochemical parameters during retting. ( = pH; = pO2;

= penetrometry index.)

12

10

8

6

4

2

0

Penetrometry

index

7

6

5

4

5

4

3

2

1

0

Time (days)

0 1 2 3 4

pH

pO

2(mg/L)


18/53

203


Evo l u t i o n o f s ub st r a t es and

me t a b o l i t e s

The main substrates degraded(Figure 3) were oligosaccharides(fructose, glucose, and saccharose).The low level of polyosides generatedby starch degradation (e.g.,maltotriose and maltose) underlinethe weak degradation of the starchymass during retting. Saccharoseseems to be the main substratedegraded by the fermentativemicroflora.

The main organic acid producedwas lactate. However, significantlevels of ethanol, acetate, andbutyrate were also found (Figure 4).They seem to be generated mostly bythe heterolactic fermentation of theoligosaccharides present in thecassava roots, except for butyrate,which could have come from ananaerobic fermentation mediated byClostr id iumspecies. Butyrateconcentration could vary from 0.4 to2.5 g/100 of dry matter in differentfermentations carried out under the

Figure 4. Organic acids and alcohol evolution during retting. (1 2 3

1 2 3

1 2 3 = butyrate;1 2 3

1 2 3= ethanol;1 2 3

1 2 3

1 2 3

= acetate; = lactate.)

4.5

4.0

3.5

3.0

2.5

2.0

1.5

1.0

0.5

0

Time (hours)

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3

1 2 3

1 2 3

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3

1 2 3

1 2 31 2 3

1 2 3 4

1 2 3 41 2 3

1 2 3 4

1 2 31 2 3

1 2 3 4

1 2 3

1 2 3

1 2 3

1 2 3

Conc.gper100gofdrymatter

Figure 3. Oligo- and monosaccharide evolution during retting. ( = maltotriose; = maltose;1 2 3

1 2 3 = saccharose;1 2

1 2 = glucose;1 2

1 2 = fructose.)

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2

1 2

1 2

1 2

1 2

1 2

1 2

1 2

1 2

1 2

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2

1 2

1 2

1 2

1 2

1 2

1 2 3

1 2 3

1 2 3

1 2

1 2

1 2

1 2

1 2

1 2

1 2

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

4.5

4.0

3.5

3.0

2.5

2.0

1.5

1.0

0.5

00 12 24 36 48 62

Time (hours)

Conc.gper100gofdrymatter

0 12 24 36 48 62


19/53

204


same conditions. Because of theirorganoleptic qualities, butyrate andlactate seem to be the most typicalproducts of this process.

M i c r o f l o r a ev ol u t i o n

F erm en t a t i v e a n d l a c t i c

m i c r o f l o r a . In the enumerations,only fermentative bacteria werecounted because retting was seen aslargely anaerobic (Figure 1). Thefermentative microflora evolvedduring the first 2 days of fermentationand remained stable to the end. Thetotal fermentative microflorarepresented by the glucose-fermentingbacteria was dense, reaching 1012b/gafter 48 h of fermentation. The nextmost predominant flora were thel.a.b. (Figure 5), reaching 104to108b/g of DM on fresh roots. Thevariation of endogenous l.a.b.,composed mainly of Lactococcusandheterolactic Lactobaci l lusspecies, didnot influence the evolution of l.a.b.during fermentation.

L a c t a t e-f e rme n t i n g b a c t e r i a .

One metabolite formed duringfermentation is butyrate (Figure 4).This compound is a typical product ofcarbohydrate fermentation by

anaerobic spore formers (Clostr id iumspecies). To evaluate this population,enumeration was done anaerobicallyon lactate because (1) lactate is themajor substrate found in retting; and(2) it is not used as a substrate by thel.a.b. Surprisingly, the results of thisenumeration showed that thepopulation of lactate-fermentingbacteria remained constant and atlow levels (103b/g of DM) throughoutthe retting (Figure 5). The presence ofbutyrate and acetate in the positivetubes, and the isolation of strictlyanaerobic, sporulating, gram-positiverods with the same fermentationpattern as Clos t r id ium buty r i cum,confirmed that Clostr id iumspecies arepresent in retting. However, their rolein the process remains to be studiedbecause of their reduced numbers inthe enumeration and lactate does notseem to be their natural substrate inretting.

Yeas ts . The only flora thatappeared after 48 h of fermentationand still developed until the end of

retting were yeasts. Theirmetabolisms allow them to grow atthe low pH imposed by the l.a.b.Their numbers remained low duringthe fermentation (about 103b/g of

Figure 5. Evolution of fermentative microflora during retting. (1 2 3

1 2 3

1 2 3

= glucose-fermenting bacteria;

1 2

1 2

1 2

= lactic acid bacteria;1 2 3

1 2 3

1 2 3

= lactate-fermenting bacteria; = yeast.)

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

14

12

10

8

6

4

2

0

Log.

b/gofdrymatter

Time (hours)

0 24 48 60 72


20/53

205


DM), suggesting that they do notplay a significant role in retting.When the retting finished, the yeastscovered the entire water surface andbecame the main flora of thepostretting stage. Their increasingnumbers at the end of the process(mostly Cand idaspecies) maytherefore influence the conservationof end products.

Or i g i n o f en z ym es i n v o l v ed i n

r e t t i n g . The main enzymes foundin this process were pectinase andlinamarase, and to a lesser extent,amylase (data not shown). Nocellulase or xylanase activities werefound in retting. To elucidate theorigin of cyanogen elimination andthe mechanism of root softening,two fermentations were carried outsimultaneously: one natural, usedas a control (CF), and one sterile(SF). pH and oxygen pressure of SFwere set on those of CF. Pectinaseand linamarase activities wereassayed throughout the experiment.For SF, cassava roots were sterilized

with HgCl2and soaked in sterilewater.

Or i g i n o f s of t e n i n g . Nosoftening was obtained in sterilefermentation (Figure 6). Highendogenous pectin methyl esteraseactivities were found in cassavaextracts from both fermentations(Figure 7). Depolymerizing enzymes,endopolygalacturonase (active atlow pH), and pectate lyase werefound only in the naturalfermentation (Figures 8 and 9). Noother depolymerizing enzymes, suchas cellulase or xylanase, nor otherhydrolases were found. Moreover,softening could be performed byinoculating commercialpectinesterase and depolymerizingpectolytic enzymes on fresh andsterile cassava roots.

We suggest, therefore, that rootsoftening is a result of the combinedaction of both endogenous pectinmethyl esterase and exogenousbacterial depolymerizing enzymes.But further studies are needed toshow the precise contribution ofeach pectic enzyme to root

softening.

Penetrometryindex(mm/s)

15

10

5

00 20 40 60 80

Time (hours)

molofCOO-permin/m

gofproteins

1.0

0.8

0.6

0.4

0.2

0

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

1 2 3 4

0 9.5 20 27 44

Time (hours)

Figure 6. Comparative evolution of softeningbetween a sterile ( ) and a naturalretting ( ).

Figure 7. Pectinesterase activity during retting.(

1 2 3

1 2 3 = sterile fermentation;1 2

1 2 = controlfermentation.)


21/53

206


50

40

30

20

10

0molofCOO-permin/mgofproteins

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7 8

1 2 3 4 5 6 7 8

1 2 3 4 5 6 7 8

1 2 3 4 5 6 7 8

1 2 3 4 5 6 7 8

1 2 3 4 5 6 7 8

1 2 3 4 5 6 7 8

1 2 3 4 5 6 7 8

1 2 3 4 5 6 7 8

1 2 3 4 5 6 7 8

1 2 3 4 5 6 7 8

1 2 3 4 5 6 7 8

1 2 3 4 5 6 7 8

1 2 3 4 5 6 7 8

1 2 3 4 5 6 7 8

1 2 3 4 5 6 7 8

1 2 3 4 5 6 7 8

1 2 3 4 5 6 7 8

1 2 3 4 5 6 7 8

1 2 3 4 5 6 7 8

1 2 3 4 5 6 7 8

1 2 3 4 5 6 7 8

1 2 3 4 5 6 7 8

1 2 3 4 5 6 7 8

1 2 3 4 5 6 7 8

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

0 9.5 20 27 44

Time (hours)

Figure 8. Pectate lyase activity during natural

fermentation.

Or i g i n o f c y a n o g en el i m i n a t i o n .

Of total cyanogenic compounds, 50%were eliminated in SF and 97% inCF (Figure 10). Enzyme assaysfurther confirmed endogenouslinamarase activity (Table 1).Linamarase activity (measured as-glucosidase activity) in CF wassignificant in fresh roots (specificactivity 9.4 units/mg protein). Thistotal activity then decreased after afew hours. In SF, total activityremained constant, but at a low level.The difference in -glucosidaseactivity in the fresh roots between SFand CF may be attributed to theinhibitory effect of the HgCl

2used to

sterilize the roots. However, as nearly25% (Table 1) of the total-glucosidase activity present in thesterile roots can degrade more than50% of the total cyanide content ofthe fresh roots, we can assume thatthe level of linamarase activitypresent in the intact roots wassufficient to detoxify the roots.

Or i g i n o f t h e am y l o l y t i c a c t i v i t y .

The amylase activity remainedconstant in SF, but disappeared after36 h of fermentation in CF (Figure 11).Our data suggest that the amylaseactivity detected in retting does nothave a bacterial origin as suggestedby different authors (Collard and Levi,1959; Oyewole and Odunfa, 1992;Regez et al., 1987).

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

1 2 3 4 5 6 7

0.3

0.2

0.1

0

Time (hours)

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5 6

1 2 3 4 5 6

1 2 3 4 5 6

1 2 3 4 5 6

1 2 3 4 5 6

1 2 3 4 5 6

1 2 3 4 5 6

1 2 3 4 5 6

1 2 3 4 5 6

1 2 3 4 5 6

1 2 3 4 5 6

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

Percentageoftotalcyanide 120

100

80

60

40

20

00 9.5 20 27 44

Time (hours)

Figure 10. Total cyanide evolution in control (1 2

1 2

1 2

) and sterile (1 2 3

1 2 3

1 2 3

) fermentations.

molofgalacturonicacidper

mi

n/mgofproteins

0 9.5 20 27 44

Figure 9. Endopolygalacturonase activity duringnatural fermentation.


22/53

207


20,000

10,000

0

communication) suggest thatClostr id iumspecies (such asClostr id iu m buty r icum) could beinvolved with Baci l lusspecies (suchas Bac i l lus po lymy xa) in rootsoftening as pectinase producers. Wedid not see any involvement ofGeotrichumspp. or Corynebacter iumspp., as have other authors (Collardand Levi, 1959; Okafor et al., 1984;Regez et al.,1987). Yeasts (mostlyCand idaspecies) were more involvedin postretting.

Our biochemical analyses showedthat retting is a fermentation in whichboth endogenous and microbialenzymes coact to soften the roots anddegrade cyanogenic, endogenouscompounds. Our results suggestedthat cell-wall degradation is initiatedby endogenous pectinesterase, locatedin intercellular spaces and releasedby pH decrease. This is followed bymicrobial polygalacturonase and lyasedepolymerizing pectic chains. Thepresence of pectic enzymes in cassavaretting has previously been reported

(Okafor et al., 1984; Oyewole andOdunfa, 1992). But this work givesthe first evidence of the vegetal originof pectinesterase and of the in vivoactivity of depolymerizing enzymes.

The amylase activity measured inretting seems to be of vegetal origin.But its low level of activity anddisappearance within the first 30 h ofretting suggest that it is not

important to the retting process.

Results of cyanide measurementsindicate that endogenous linamarase(measured as -glucosidase activity) isthe main enzyme responsible fordetoxification. We can assume, asMaduagwu (1983) suggested, that thelevel of linamarase activity present inintact roots is sufficient to detoxifythem of their cyanogen content

without help from any microbiallinamarase. Nevertheless, if bacteriado not directly detoxify cassava roots,

Conclusions

These results suggest that retting is acomplex heterolactic fermentation,with an interaction between lacticbacteria, Clostr id iumspecies, and

possibly Baci l lusspecies. Heterolacticbacteria (such as Leuconostocmesenteroides) are the mostimportant and numerous microflorain the process; they are responsiblefor the physicochemical properties ofretting (e.g., pO

2and pH) and the

production of the main organic acids(acetate and lactate). Clostr id iumspecies seem to be involved inbutyrate formation, which is essential

for the organoleptic properties of thefinal products. Moreover, recentresults (S. Klke, 1994, personal

Table 1. -glucosidase activities in control andsterile fermentations. (Activities areexpressed in mmol per min/100 g ofdry matter).

Time (h) Fermentation

Control Sterile

0 9.12 2.15

9.5 5.58 2.55

20.0 6.10 1.75

27.0 7.68 2.30

44.0 7.24 1.38

Figure 11. Amylase activity in control ( ) andsterile ( ) fermentations.

0 20 40 60 80

Totalactivity(U/L)

Time (hours)


23/53

208

Cassava Flour and Starch: Progress in Research and D evelopment

they could help degrade linamarin bydestroying cell walls.

Findings from our study havehelped other researchers:

(1) Isolate and characterize the firstamylolitic Lactobaci l lus plantar um(strain A6) (Giraud et al., 1991);

(2) Improve fu fuprocessing bysignificantly reducing retting time,and increase the organolepticqualities of the final product(Ampe et al., 1994);

(3) Adapt the process for areas withlow water availability (Miambi etal., n.d.).

References

Ampe, F.; Brauman, A.; Trche, S.; andAgossou, A. 1994. The fermentationof cassava: optimization by theexperimental research methodology.

J. Sci. Food Agric. 65:355-361.

Bensadoun, A. and Weinstein, D. 1976.Assay of protein in the presence of

interfering materials. Anal. Biochem.70:241-250.

Brauman, A.; Klke, S.; Mavoungou, O.;Ampe, F.; and Miambi, E. n.d. Etudesynttique du rouissage traditionneldes racines de manioc en Afriquecentrale (Congo). In: Agbor, E.;Brauman, A.; Griffon, D.; and Trche,S. (eds.). Cassava food processing.Institut franais de recherchescientifique pour le dveloppement encoopration (ORSTOM) Editorials,Paris, France. (In press.)

Collard, P. and Levi, S. 1959. A two-stagefermentation of cassava. Nature(Lond.) 183:620-621.

Cooke, R. D.; Blake, G. G.; and Battershill,J. M. 1978. Purification ofcassava linamarase. Phytochemistry(Oxf.) 17:381-383.

de Man, J. C.; Rogosa, M.; and Sharpe, M. E.1960. A medium for the cultivation ofLactobaci l l i. J. Appl. Bacteriol.

23:130.

Giraud, E.; Brauman, A.; Klke, S.; Lelong,B.; and Raimbault, M. 1991. Isolationand physiological study of anamylolitic strain of Lactobaci l lus

p lan ta rum. Appl. Microbiol.Biotechnol. 36:379-383.

__________; Gosselin, L.; and Raimbault, M.1992. Degradation of the cassavalinamarin by lactic acid bacteria.Biotech. Lett. 14(7):593-598.

Hosel, W. and Bartz, W. 1975. DFglucosidases from Cicer a r ientumL.Eur. J. Biochem. 57:607-616.

Hungate, R. E. 1969. A roll tube method forthe cultivation of strict anaerobes. In:Norris, J. R. and Ribbons, D. W.(eds.). Methods in microbiology,

vol. 3B. Academic Press, NY.

McCready, M. H. 1918. Tables for rapidinterpretation of fermentation tuberesults. Can. J. Public Health 9:201.

Macy, J. M.; Snellen, J. E.; and Hungate,R. E. 1972. Use of syringe methodsfor anaerobiosis. Am. J. Clin. Nutr.25:1318-1323.

Maduagwu, E. N. 1983. Differential effects onthe cyanogenic glycoside content of

fermenting cassava root pulp by-glucosidase and microbialactivities. Toxicol. Lett. (Amst.)15:335-339.

Miambi, E.; Machicout, M.; Trche, S.; andBrauman, A. n.d. Le rouissage sanseau, une nouveau procd detransformation des racines demanioc. In: Agbor, E.; Brauman, A.;Griffon, D.; and Trche, S. (eds.).Cassava food processing. Institutfranais de recherche scientifiquepour le dveloppement en coopration(ORSTOM) Editorials, Paris, France.(In press.)

Okafor, N.; Ijioma, B.; and Oyolu, C. 1984.Studies on the microbiology ofcassava retting for fufu production.

J. Appl. Bacteriol. 56:1-13.

Oladele Ogunsa, A. 1980. Changes in somechemical constituents during thefermentation of cassava roots(Man ihot esculentaCrantz). FoodChem. 5:249.


24/53

209

Cassava Lact ic Fermentat ion in Centra l Afr ica:. ..

Oyewole, O. B. and Odunfa, S. A. 1992.Extracellular enzyme activitiesduring cassava fermentation forfufu production. World J. Microbiol.& Biotechnol. 8:71-72.

Regez, P. F.; Ifebe, A.; and Mutinsumu, M. N.

1987. Microflora of traditionalcassava foods during processing andstorage: the cassava bread(chikwangue) of Zaire. Microb.

Aliment. Nutr. 5:303-311.

Somogyi, M. 1945. Determination of bloodsugar. J. Biol. Chem. 160:61-68.

Starr, M. P.; Chatterjee, A. K.; Starr, P. B.;and Buchanan, G. E. 1977.Enzymatic degradation ofpolygalacturonic acid by Yersin iaand

Klebsie l laspecies in relation toclinical laboratory procedures. J.Clin. Microbiol. 6:379-386.

Trche, S. n.d. Importance du manioc enalimentation humaine dansdifferentes rgions du monde. In:

Agbor, E.; Brauman, A.; Griffon, D.;and Trche, S. (eds.). Cassava foodprocessing. Institut franais derecherche scientifique pour ledveloppement en coopration(ORSTOM) Editorials, Paris, France.

(In press.)

__________ and Massamba, J. n.d.a. Laconsommation du manioc au Congo.In: Agbor, E.; Brauman, A.; Griffon,D.; and Trche, S. (eds.). Cassavafood processing. Institut franais derecherche scientifique pour ledveloppement en coopration

(ORSTOM) Editorials, Paris, France.(In press.)

__________ and __________. n.d.b. Les modesde transformation traditionnels dumanioc au Congo. In: Agbor, E.;Brauman, A.; Griffon, D.; and Trche,S. (eds.). Cassava food processing.Institut franais de recherchescientifique pour le dveloppement encoopration (ORSTOM) Editorials,Paris, France. (In press.)

Widdel, F. and Pfennig, N. 1984.Dissimilatory sulfate- orsulfur-reducing bacteria. In: Krieg,N. R. and Holt, J. G. (eds.). Bergeysmanual of systematic bacteriology,

vol. 1. Williams and Wilkins, MD,USA. p. 663-679.


25/53

210


CHAPTER2 4

A LACTICACIDBACTERIUMWITH

POTENTIALAPPLICATIONINCASSAVAFERMENTATION

E. Giraud*, A. Braum an**, S. Klke***,L. Gosselin*, andM. Raimba ul t

* Laboratoire de biotechnologie, Institutfranais de recherche scientifique pour ledveloppement en coopration (ORSTOM),Montpellier, France.

** ORSTOM, Paris, France.

*** Laboratoire de microbiologie, Directiongnrale de la recherche scientifique ettechnique (DGRST), Brazzaville, Congo.

ORSTOM, stationed in Cali, Colombia.

Introduction

Lactic microflora play an importantrole in the preparation of traditionalfoods based on fermented cassava,such as gar i, ch i kw angue, fu fu, andsour starch. But this microflorasfunction in preserving foods,eliminating cyanogenic compounds,and improving organoleptic qualities isnot yet clear. Traditional technologiesare still used to manufacture thesefoods. As fermentation occursnaturally with lactic microflora, thequality of the food products is notuniform.

The mass inoculation of cassavaroots with one or several selectedstrains would permit a better controlover natural fermentation, thusresulting in a product of improvedquality. Because cassava containsmainly starch (more than 80% of drymatter), the selection of a lactic acid

bacterium capable of metabolizingstarch (i.e., amylolytic) is essential.

But few lactic acid bacteria canconvert starch into lactic acid.Examples of amylolytic lactic acidbacteria are Streptococcus bovis, S.equinus, Lactobaci llus amy lophi lus, L.amylovorus, L. acidophi lu s, L.cellobiosus, and others isolated fromanimal digestive tracts and plant

wastes (Champ et al., 1983; Cotta,1988; Nakaruma, 1981; Nakarumaand Crowell, 1979; Sen and

Abstract

An amylolytic lactic acid bacterium,identified as Lactobaci llus plantar um,was isolated from cassava roots(Manih ot esculentavar. Ngansa) duringretting. Cultured on starch, the straindisplayed a growth rate of 0.43 perhour, a biomass yield of 0.19 g/g, anda lactate yield of 0.81 g/g. The growthkinetics were similar on starch andglucose. Enough enzyme wassynthesized, and starch hydrolysiswas not a limiting factor for growth.The synthesized amylolytic enzymewas purified by fractionatedprecipitation with ammonium sulfateand by anion exchangechromatography. It was identified asan -amylase with an optimal pH of5.5 and an optimal temperature of65 C. The use of such a strain as acassava fermentation starter for gar iproduction had the following effects: achange from a heterofermentative

pattern observed in naturalfermentation to a homofermentationone, a lower final pH, a faster pHdecline rate, and a greater productionof lactic acid (50 g/kg of dry matter).


26/53

211

A Lactic Acid Ba cterium...

Chakrabarty, 1986; Sneath, 1986).Almost no information exists on thephysiology of these microorganisms.

Below we describe how we isolatedand identified a new amylolytic lacticacid bacterium from fermentingcassava roots. We also investigated thephysiology of this bacterium and theproperties of the amylase produced.

Methods

I so l a t i n g a n d i d en t i f y i n g st r a i n s

Peeled roots were immersed in rainwater. Sampling was carried out4 days after fermentation by randomlyselecting six roots cut into 0.5-cmcubes and mixed under sterileconditions. A sample of 60 g wasdiluted in 540 ml of sterile peptonesolution. Then 0.1 ml of decimaldilutions were spread on JP2 medium(see below) in petri dishes. Afterincubation for 48 h at 30 C, thedishes were exposed to iodine vapor to

detect the starch hydrolysis areas.Isolated strains were then purified bythree successive transfers on JP2medium, and cultures routinelychecked for purity by microscopicobservation.

Microorganisms were identifiedby:

(1) the configuration of the lactic

acid produced after treatment(Ivorec-Szylit and Szylit, 1965)with the enzymes dehydrogenasel and d (Boerhinger Mannheim);

(2) the microorganisms homolactic orheterolactic character, asdetermined by acetic acid or

(3) presence or absence of catalase;(4) microscopic and macroscopic

examination of morphology,mobility, and spores;

(5) Gram stain;(6) arginine dissemination;(7) growth at 15 and 45 C; and

(8) fermentation of different carbonsources (API 50CH #5030 strips,Biomrieux, France).

Bergeys Manual (Sneath, 1986)was used to evaluate results andidentify the different strains.

S t r a i n s a n d c u l t u r e m ed i a

Three strains were used as reference:Lactobaci llus pla ntaru m (Lacto Labo,France), Streptococcus equinu sCNCM 103233, and Lactobaci l lusamy loph i lusCNCM 102988T.

JP2 m ed i um (g /L ). Thisconsisted of:

M66 universal peptone 2.5

Soya peptone obtainedby papain digestion 5

Casein peptone obtainedby pancreatic digestion 2.5

Yeast extract 5

Meat extract 2.5

MgSO4,7H

2O 0.1

NaCl 3

(NH4)2SO

42

K2HPO

40.2

Prolabo soluble starch 3

Tween 80 (in ml) 0.4

The pH was adjusted to 6.75before sterilization.

Physiological studies were

performed, using a deMan-Rogosa-Sharpe (MRS) basalmedium (de Man et al., 1960) andchanging the carbon sources to 5%glucose and 5% starch.

Cu l t u r e c ond i t i o n s . Strains werecultured in a 2-L bioreactor(Biolafitte, France) at 30 C andagitated at 200 rpm. The pH wasadjusted to 6.0 by adding NaOH (5 N).

Inoculation at 10% v/v was performedwith a 20-h pre-culture in the samemedium used for fermentation.


27/53

212


An a l y t i c a l m et h o d s

The biomass concentration wasdetermined by measuring the opticaldensity (OD) at 540 nm related to thedry weight measured after twowashing and centrifugation cycles anddrying at 105 C for 24 h. For starchcultures, hydrolysis of residual starchwas performed with a mixture ofamylases (thermamyl + dextrosyme,supplied by Novo). The dry weightand OD were then determined asabove. Lactic acid, glucose, aceticacid, and ethanol concentrations inthe supernatant were assayedby high-performance liquidchromatography (HPLC). Compoundswere separated by using an AminexHPX 87H column (Bio RadLaboratory) with a 0.8 ml/min flow(pump LDC 3200) of H

2SO

4(0.012 N)

solution at 65 C. Analyses werecarried out with a refractive indexdetector (Philips PU 4026). Totalsugars in media containing starchwere also determined with anthrone,using the Dubois et al. (1956)

method.

Amy l a s e a s sa y . The -amylaseactivity was measured by incubating0.1 ml of appropriately dilutedenzyme solution with 0.8 ml of asolution containing 1.2% of Prolabosoluble starch in 0.1 mol/Lcitrate-phosphate buffer (pH = 5.5) at55 C. The reaction was stopped byadding 0.1 ml of 1 mol/L H

2SO

4. After

incubation, residual starch contentswere determined colorimetrically afterdifferent periods at 620 nm by adding0.1 ml of the reaction mixture to2.4 ml of an iodine solutioncontaining 30 g/L of KI and 3 g/L of I

2

and diluted to 4% with distilled water.

An enzyme unit is defined as theamount of enzyme that permits thehydrolysis of 10 mg of starch in

30 min under the conditionsdescribed above. Proteinconcentration was estimated with the

Bradford (1976) method, using aBiorad Kit (Cat No. 500-0001,Ivry-sur-Seine, France) and bovineserum albumin as standard.

Pu r i f i c a t i o n o f am y l a s e.

Fermentation was stopped afterculture for 9 h. Cells were removed bycentrifugation (at 15,000 g for 15 minat 4 C), and the supernatant fluid(750 ml) filtered through a cellulosefilter (0.45 m pore size, HAWP type,Millipore, Saint Quentin les Yvelines,France) to remove cell debris.

Powdered ammonium sulfate wasthen slowly added to the supernatantfluid under constant stirring at 4 C.Most of the amylase activity wasprecipitated at between 50% and 70%saturation.

After the ammonium sulfatefractionation, the precipitatedprotein collected by centrifugation(at 15,000 gfor 30 min at 4 C) wasresuspended in 50 mmol/L KH

2PO

4/

Na2HPO

4standard buffer (pH = 6.8).

The enzyme solution was washed andconcentrated with a PM-10 Amiconultrafiltration membrane. It was thenloaded onto a diethylaminoethyl(DEAE) cellulose column (DE-52;Whatman Laboratory Sales, Hillsboro,Oregon, USA). The column(25 x 250 mm, flow rate 2.5 ml/min,25 C) was previously equilibratedwith the standard buffer. The enzymewas eluted, using a concave, sodium

chloride gradient (0-1.0 mol/L).Fractions (5 ml) were collected. Thefractions that were enzymatically themost active were pooled, dialyzedovernight at 4 C against the standardbuffer, and used for further studies.They were kept at -30 C. No activitywas lost for at least 3 months undersuch conditions.

Po l y a c r y l am i d e gel

e l e c t r opho res i s . This was carried outaccording to Laemmlis method (1970),with a 10% running gel and 4%


28/53

213

A Lactic Acid Ba cterium...

stacking gel. Electrophoresis undernondenaturating conditions wasperformed in the absence ofsodium dodecyl sulfate (SDS) and-mercaptoethanol in any buffer. Gelswere run at a constant 150 V for 1 h at25 C. Proteins were stained by thesilver method (Oakley et al., 1980).

Amy l a se st a i n . Afterelectrophoresis, gel was incubatedfor 1 h at 30 C in 0.1 mol/Lcitrate-phosphate buffer (pH = 5.5),containing 1% of soluble starch. Aftertwo washes with distilled water, lightlanes (representing starch hydrolysisareas of amylase activity) were detectedby immersing the gel in Lugolssolution.

Mol ec u l a r ma s s d et e rm i n a t i o n .

SDS-PAGE electrophoresis was used todetermine the approximate molecularmass of amylase. Marker proteins(Biorad, Cat. No. 161-0315) used weremyosin (200,000), galactosidase(116,250), phosphorylase-b (97,400),bovine serum albumin (67,000), and

ovalbumin (45,000).

Assa y s on ga r i . Fresh importedcassava roots from Cameroon wereobtained from Anarex (Paris, France).Gariwas prepared from peeled, washedcassava roots, which were chopped andminced in a food mixer (SEB). Thepulp obtained was packed tightly intoplastic, sterile, screw-cappedcontainers (60 ml; OSI, A12.160.56)

and placed at 30 C.

Three batches were prepared:(1) natural fermentation, using theendogenous microflora present;(2) fermentation after inoculation withL. plantar umA6 (108cfu/g of driedcassava), which had been cultured inbioreactors on cellobiose MRS medium;(3) fermentation after inoculation withL. plantar umLactolabo (108cfu/g of

dried cassava), which had beencultured in bioreactors on MRScellobiose. Cells were washed in

physiological solution before cassavainoculation.

A container from each batch wasmonitored every day to test the followingparameters:

(1) The pH was measured on a 10-gsample and homogenized indistilled water (20 ml). Moisturewas measured by drying a 10-gsample at 105 C for 24 h.

(2) The number of lactic acid bacteria(l.a.b.) was estimated on a 10-gsample homogenized in90 ml of physiological sterilesolution. Colonies werecounted on MRS agar, using aspread-plate technique on petridishes and after incubation at30 C and 48 h.

Results and Discus s ion

I so l a t i o n a n d i d en t i f i c a t i o n of

Lactobacillus plantarum A6

Seven amylolytic microorganisms wereisolated on JP2 medium from rettedcassava roots. Two were revealedby HPLC to have a capacity toproduce lactic acid from starch.Table 1 lists their morphological,physiological, and biochemicalcharacteristics. The ability of thesecultures to use 49 differentcarbohydrates was studied withAPI 50CH #5030 strips. The results

were compared, by computer, with thepercentage of positive reactions ofdifferent Lactobaci l lusspecies as perAPI. A 99.9% rate of similarity withL. planta rumwas observed and henceidentifying these cultures as strains ofL. planta rum. The two strains, A6 andA43, displayed precisely the same sugardegradation profiles, which suggeststhat they are probably the same.

The amylolytic activities on JP2medium of L. plantar umA6, S. equinu s,and L. amylophi lus indicated that the


29/53

214


Table 1. Characteristics of Lactobacil lusp lan tarumstrains A6, A43, and Lacto Labo (check).

Strain A6 A43 Check

Ratio of d:l lactic acid 69:31 66:34 73:27

Homolactic + + +

Catalase - - -

Bacterium shape Short rod Short rod Short rod

Gram stain + + +

Spore - - -

Mobility - - -

Dissemination of arginine - - -

Growth at 15 C + + +

Growth at 45 C - - -

starch hydrolysis zone was largest forL. plantar umA6. It was thereforeselected for further studies.

Lactobacillus plantarum A 6 g r o w t h

k i n e t i c s

The growth of L. plantar umA6 onglucose MRS medium (Figure 1) is fullycomparable with that of L. planta rum(Lacto Labo). The growth rate (0.

Documents

Cassava Flour Session 4 Bioconversion