chapter 1 introduction

CHAPTER 1 INTRODUCTION Among the foods eaten by humans, soybeans contain the highest level of isoflavones. Isoflavones are phytoestrogen (plant-derived phenolic compounds with structural homology to human estrogen), they are common in leguminous plants, especially in soybean. These compounds are currently heralded to offer potential natural alternative therapies for a range of hormone-dependent conditions including post menopausal symptoms,cardiovascular disease as well as prostate,breast and colon cancer.(Ravindranath et al., 2004, Weaver and Cheong, 2005). Asian populations with their high intake (50-70 mg/d) of soy derived isoflavones are known to have the lowest incidence of menopausal symptoms such as (hot flushes, osteoporosis), cardiovascular disease and cancer (Nagata et al. 1998). It is also found that isoflavones may reduce low density lipoproteins and increase high-density lipoproteins, which help to prevent coronary heart disease (Demonty et al., 2003). Soy isoflavones also exhibit anti atherosclerotic, antibacterial, antioxidant, and blood glucose lowering activities. It also leads to improvement in cognitive functions of elderly patients who have Alzheimer's disease (Chung, 2000). Soy isoflavones exist in the form of aglycones (daidzein, genistein and glycitein) and glycosides conjugates, which include the β-glycosides (daidzin, genistin and glycitin), malonyl glycosides, and acetyl glycosides (Kudou et al., 1991). The content of daidzin and genistin are high in soybean while the aglycones (daidzein, genistein) are found in trace quantities. Numerous studies have shown that the biological effects of isoflavones are not due to the glycoside forms but mainly to their aglycones, such as daidzein and genistein. For example, aglycone isomers are able to bind to estrogen receptor and hence mimic estradiol functions in the human body (Setchell and Cassidy, 1999), and thus prevent certain cancers (Fritsche and Steinhart, 1999). In addition, some researchers have shown that isoflavone aglycones in soybean food are absorbed faster and in higher 1

amounts from gastro intestinal tract than that of respective glycoside. The anticancer function of soybean isoflavones was shown to be associated with genestein aglycone, which inhibit protein kinase, DNA topoisomerase and binds weakly to estrogen receptor (Messina and Messina, 2000). (Lamartiniere, 2000) concluded that early exposure to genistein enhances cell differentiation of the mammary gland, and may confer a protective effect against carcinogenesis via this process. Furthermore, in vitro studies using cultured human breast cancer cells indicate that genistein inhibited the growth of both estrogen receptor-negative and estrogen receptor-positive cell lines (Wang and Kurzer, 1997). Furthermore, (Ruiz-Larrea et al., 1997) reported that the antioxidant activity of genistein was much greater than that of genistin. One way to improve the bioavailability and biological activity of soybean isoflavones is simply to increase the concentration of isoflavone aglycones in soybean food by enzymatic transformation. Enzymatic transformation has been known to have several advantages over chemical methods in the transformation process. These advantages include the mild reaction conditions, the high yield of hydrolysis and the reduction of the inhibitory compounds formed (Lee et al,. 1999 and Wymen, 1996). For this reason, we turn our think to the use of β-glucosidase enzyme (EC 3.2.1.21, β-glucoside glucohydrolase) which mainly catalyses hydrolysis of the b-1,4-glycosidic linkage in various disaccharides, oligosaccharides, alkyl- and aryl-b-D-glucosides (Bhatia et al., 2002). β-glucosidase is also involved in many applications. This includes the enzymatic hydrolysis of lignocellulosic biomass and the subsequent fermentation of the released sugars which is considered to be the most important route for transportation fuel generation. It also has applications in pharmaceutical, cosmetic, and detergent industries (Job et al., 2010). In the flavor industry, β-glucosidases are also key enzymes in the enzymatic release of aromatic compounds from glucosidic precursors present in fruits and fermentating products (Shoseyov et al., 1990).

2

1.1.

Objectives and Scope of the Current Study

Screening of potential fungi capable of producing considerable amounts of βglucosidase enzyme. Evaluation of carbon sources for β-glucosidase production by the selected strain. Optimizing the production of the enzyme β-glucosidase by the selected strain using a low-cost minimally controlled batch fermentation process using the response surface methodology. Scale up of β-glucosidase production in stirred tank bioreactor (STR) and rotating fibrous bed bioreactor (RFBB) under different cultivation modes. Partial purification of the β-glucosidase enzyme and characterization of its properties. Optimizing the biotransformation of soy isoflavones glycosides form into aglycones form under different conditions using both β-glucosidase enzyme produced from Aspergillus niger 3122 and commercial β-glucosidase containing enzyme (celluclast BG). Evaluation of the antioxidant activity of the biotransformed soy flour by both βglucosidase produced from Aspergillus niger 3122 and commercial β-glucosidase containing enzyme (celluclast BG) against the non biotransformed soy flour.

3

CHAPTER 2 LITERATURE REVIEW 2.1. β-glucosidases enzyme 2.1.1. Definition and mode of acion of β-glucosidases enzyme β-glucosidases (β-D-glucoside glucohydrolase, EC 3.2.1.21) are well characterized, biologically important enzymes that catalyze the transfer of glycosyl group between oxygen nucleophiles. These transfer reaction results in the hydrolysis of β-glucosidic linkage present between carbohydrate residues in aryl-amino-, or alkyl-β-D-glucosides, cyanogenic-glucosides,

short

chain

oligosaccharides

and

disaccharides

under

physiological conditions, whereas; under defined conditions, synthesis of glycosyl bond between different molecules can occur. It occurs by two modes: reverse hydrolysis and transglycosylation. In reverse hydrolysis, modification of reaction conditions such as lowering of water activity, trapping of product or high substrate concentration leads to a shift in the equilibrium of reaction toward synthesis. This reaction is under thermodynamic control. In transglycosylation approach, a preformed donor glycoside (e.g., a disaccharide or aryl-, amino-, or alkyl -linked glucoside) is first hydrolyzed by the enzyme with the formation of an enzyme-glycosyl intermediate. This is then trapped by a nucleophile other than water (such as a monosaccharide, disaccharide, aryl-, amino-, or alkyl-alcohol or monoterpene alcohol) to yield a new elongated product. This reaction is under kinetic control‖ (Bhatia et al, 2002). β-glucosidases are widely distributed in the living world and they play vital roles in many biological processes. The physiological roles associated with this enzyme are diverse and depend on the location of the enzyme and the biological system in which these occur. In cellulolytic microorganisms, β-glucosidase is involved in cellulose induction and cellulose hydrolysis (Bisaria and Mishra, 1989, Tomme, 1995). In plants, the enzyme is involved in β-glucan synthesis during cell wall development, pigment metabolism, fruit ripening, and defense mechanisms (Esen and Gungor, 1993, Brozobohaty et al, 1993) whereas, in humans and other mammals, BGL is involved in the hydrolysis of glucosyl ceramides(Libermann et al, 2007). Due to their wide and 4

varied roles in nature, these versatile enzymes can be of use in several synthetic reactions as reviewed by Bhatia et al, (2002). 2.1.2. Microbial production of β-glucosidase Microbial sources have been widely exploited for β-glucosidase production by both solid-state fermentation (SSF) and submerged fermentation. There are several reports available for β-glucosidase productions from filamentous fungi such as Aspergillus niger (Gunata and Vallier

1999), A oryzae (Riou et al, 1998), Penicillium brasilianum

(Krogh et al, 2010) P. decumbens (Chen et al, 2010), Phanerochaete chrysosporium (Tsukada et al, 2006), Paecilomyces sp., (Yang et al, 2009) etc., though there are also various reports of β-glucosidase production from yeasts (majority of them from Candida sp.) and few bacteria. β-glucosidases was purified from Agrobacterium fecalis. The purified enzyme was characterized and the corresponding gene was identified and expressed in E. coli. The Agrobacterium enzyme is a dimer of 50 kD monomers and shows high specificity for cellobiose. The enzyme has been well characterized with respect to the residues at the catalytic center and the mechanism of catalysis. The cellulolytic fungal ß-glucosidases have also been the subject of numerous investigations by various research groups. The fungal enzymes are used in several biotechnological processes, including development of novel carbohydrate foods, alcohol based fuels and other commercial products from cellulose. Particularly, glucose production can be achieved from the most abundant biological macromolecule, cellulose, by the extracellular enzyme complex (cellulase) that is derived from various fungal species such as Trichoderma. The cellulase complex isolated from Trichoderma reseei comprises at least three different enzymes that together hydrolyze cellulose to oligosaccharides and glucose (Fowler, 1993). Of these, the endoglucanases and cellobiohydrolases synergistically hydrolyze cellulose into small cellooligosaccharides, mainly cellobiose. Subsequently, cellobiose is hydrolyzed to glucose by ß-glucosidase. The ß-glucosidase gene bgl1 from Trichoderma reseei was cloned and sequenced by

5

Barnet et al, (1991); later, (Fowler, 1993) studied ß-glucosidase from the null strain of Trichoderma reesei to explore its role in the cellulose enzyme system. The hydrolysis of cellulose and induction of other cellulolytic enzyme components led Fowler to the conclusion that extracellular ß-glucosidase is required for the induction of the other cellulose enzymes. In the flavor industry, β-glucosidases are also key enzymes in the enzymatic release of aromatic compounds from glucosidic precursors present in fruits and fermentating products (Guegen, et al., 1996; Shoseyov et al., 1990). Indeed, many natural flavor compounds, such as monoterpenols, C-13 norisoprenoids, and shikimate-derived compounds, accumulate in fruits as flavorless precursors linked to mono- or diglycosides and require enzymatic or acidic hydrolysis for the liberation of their fragrances (Vasserot et al., 1995; Winterhalter and Skouroumounis (1997). Finally, β –glucosidases can also improve the organoleptic properties of citrus fruit juices, in which the bitterness is in part due to a glucosidic compound, naringin (49,5,7trihydroxyflavanone-7- rhamnoglucoside), whose hydrolysis requires, in succession, an arhamnosidase and a β –glucosidase (Roitner et al.,1984). The enzymatic hydrolysis of these compounds requires a sequential reaction, which produce monoglucosides. Subsequently, monoglucosides are hydrolyzed by the action of ß-glucosidases. Endogeneous ß-glucosidases from grape are not sufficient to process the hydrolysis of monoterpenyl-glucosides. The grape enzymes display limited activity towards these glucosides and a large fraction of the aromatic compounds remains unprocessed in mature fruit. The addition of glucose-tolerant exogenous ß-glucosidase isolated from fungi (e.g. Aspergillus oryzae) was shown to improve the hydrolysis of glucoconjugated aromatic compounds and enhance wine quality (Riou et al., 1998). In conclusion, microbial ß-glucosidases have been well-characterized and many of them have been used in biotechnological applications. 2.1.3. β-glucosidase production by Aspergillus species β-glucosidase enzyme was produced from different Aspergillus strains (Aspergillus phoenicis, Aspergillus niger and Aspergillus carbonarius) using different carbon sources. 6

Aspergillus carbonarius has been proved to be less effective in β-glucosidase production while Aspergillus phoenicis was proved to be the best enzyme producer using wheat bran as carbon source (Jager et al ., 2001). High titer of β-glucosidase enzyme (80 U/ml) was obtained from Aspergillus strain SA 58

using pectin as a carbon source. The strain

produced two extracellular enzymes and two intracellular enzymes. For both extracellular enzymes (BGL A and BGL B) was found to have the best activity at 60ᴼC and at pH 4 (Elyas et al., 2010). Through a fungal screening program for β-glucosidase activity, strain AP (CBS 127449, Aspergillus saccharolyticus) showed 10 times greater β-glucosidase activity than the average of all other fungi screened, with Aspergillus niger showing second greatest activity. The potential of a fermentation broth of strain AP was compared with the commercial β-glucosidase-containing enzyme preparations Novozym 188 and Cellic CTec. The fermentation broth was found to be a valid substitute for Novozym 188 in cellobiose hydrolysis. The Michaelis–Menten kinetics affinity constant as well as performance in cellobiose hydrolysis with regard to product inhibition were found to be the same for Novozym 188 and the broth of strain AP. Compared with Novozym 188, the fermentation broth had higher specific activity (11.3 IU/mg total protein compared with 7.5 IU/mg total protein) and also increased thermostability, identified by the thermal activity number of 66.8 vs. 63.4 ᴼC for Novozym 188 (Sorensen et al., 2011). 2.1.4. Scale up of β-glucosidase and cellulase enzymes production in bioreactors The mechanisms of cellobiase formation in Aspergillus japonicus 2092 and Aspergillus heteromorphus 3010 were studied. Formation of cellobiase in both strains was found to be a non-inducible constitutive character. The delay in formation of cellobiase in batch culture with glucose was shown to be determined by catabolic repression. Aspergillus heteromorphus 3010 was more sensitive to catabolic repression than Aspergillus japonicus 2092. To remove the effect of glucose repression two modes of fed-batch processes were developed: moderate and intensive. Both modes used the double algorithm of glucose supply based on computer-aided process control. For Aspergillus 7

heterornorphus 3010 the moderate fed-batch was optimal, while for Aspergillus japonicus 2092 intensive fed-batch was more efficient. These regimes resulted in a three-fold increase in cellobiase in comparison with batch culture and allowed the combination of growth and biosynthesis of the enzyme in one phase (Solovyeva et al .,1997). Ahamed and Vermette, (2008) conducted experiments to enhance cellulase enzyme production from Trichoderma reesi Rut-C30. It was conducted as fed batch culture in 7 L bioreactor using 4 L cellulose- yeast extract media. A mixture of lactose and lactobionic acid was added as cellulase inducer. The volumetric enzyme activity was 69.8 U/ml/hr. The filter paper activity was 5.02 U/ml. The CMCase activity was 4.2 U/ml and the fungal biomass was 14.7 g/L. The effect of mechanical agitation on cellulose production was studied. The fermentation was carried out in 35 L draft tube airlift bioreactor equipped with a mechanical impeller in a cellulos culture media with latose and lactobionic acid as fedbatch. Culture carriedout without mechanical agitation resulted in a higher volumetric enzyme productivity (200 IU/l/h), filter paper activity (17 IU/l/h), carboxy methyl cellulase (11.8 IU/l/h) and soluble protein (3.2 mg/ml) when compared to those with agitation. Stereo and polarized light microscopy analysis reveals that mechanical agitation resulted in shorter mycelial hyphae and larger number of tips (Ahamed and Vermette, 2010). The sequential solid state and submergd culture fermentation for production of cellulose enzyme by Aspergillus niger A12 using sugarcane baggase as substrate was studied. An unconvential preculture with an intial fungal growth phase and under solid sate cultivation was established followed by transition to submerged fermentation by adding the liquid culture media to the mycelia grown on solid substrate. For comparison, control experiments were conducted using convential submerged cultivation. These cuture were carried out in shake flasks and in 5-L bubble column bioreactor. An endoglucanase productivity of 57 IU/L/h was achieved in bubble columns cultivations prepared using he new method, representing 3 fold improvement compared to convential submerged fermentations( Cunha et al .,2012). 8

Trichoderma viride 1763, Trichoderma viride 2535, Aspergillus niger 282 and Aspergillus wentii were screened for large scale cellulase and β-glucosidase production. The cellulase and β-glucosidase were produced under submerged fermentation conditions using Mandel’s basal medium supplemented with one percent cellulose. The strains of Trichoderma viride are a poor source of β-glucosidase enzyme. The enzyme from Aspergillus wentii as compared to Trichoderma viride indicated that the enzyme obtained from Aspergillus wentii source contained less amount of endo- and exo- glucanases involved in cellulose degradation. It was observed that Trichoderma and Aspergillus produced maximum amount of enzyme in 8 days at 30ᴼC. Thus we supplemented the cellulase produced by Trichoderma viride 1763 with necessary amount of cellobiase. This was supplemented by the β-glucosidase prepared separately by the cultures of Aspergillus wentii. By doing so it was possible for us to achieve up to 82% hydrolysis of agro- residue cellulose fiber (Kandari et al., 2013). A bioprocess to produce cellulolytic complex (endoglucanases, exoglucanases, and β-glucosidases), which act synergistically in cellulose breakdown for production of second-generation ethanol from the fungus Penicillium funiculosum ATCC11797 was optimised. A statistical full factorial design (FFD) was employed to determine the optimal conditions for cellulase production. The optimal composition of culture media using Avicel (10 g·L ) as carbon source was determined to include urea (1.2 g·L ), yeast extract −1

−1

(1.0 g·L ), KH2PO4 (6.0 g·L ), and MgSO4·7H O (1.2 g·L ). The growth process was −1

−1

−1

2

performed in batches in a bioreactor. Using a different FFD strategy, the optimised bioreactor operational conditions of an agitation speed of 220 rpm and aeration rate of 0.6 vvm allowed the obtainment of an enzyme pool with activities of 508 IU·L for FPase, −1

9,204 IU·L for endoglucanase, and 2.395 IU·L for β-glucosidase. The sequential −1

−1

optimisation strategy was effective and afforded increased cellulase production in the order from 3.6 to 9.5 times higher than production using nonoptimised conditions (Carvalho et al., 2014).

9

2.1.5. Response surface methodology (RSM) Response surface methodology is an effective statistical technique for the investigation of complex processes, and optimization of multiple variables to predict the best performance conditions with minimum number of experiments (Mutalik et al., 2008). The main advantage of RSM is the reduced number of experimental runs needed to provide sufficient information for statistically acceptable results, so it was considered a faster and less expensive method for gathering research data than the classical method (Gunawan et al., 2005). It was also a useful statistical technique for designing experiments and analyzing the effects of independent variables (Garrido-Vidal et al., 2003 and Nemukula et al., 2009). In addition, this method is efficient to explain the relationships between different variables and the response when there is interaction between the variables. RSM consists of an empirical modelization technique, which has been used to evaluate the relation between experimental and predicted results (Chowdary et al., 2002). 2.1.6. Response surface methodology (RSM) for β-glucosidase enzyme production Response surface methodology was used to investigate the effects of 4 fermentation parameters (yeast extract concentration, cellobiose concentration, ammonium sulfate concentration, and pH) on β-glucosidase enzyme production from a newly isolated fungus Aspergillus niger SOI017 in shake flask cultures. Production of β-glucosidase was most sensitive to the culture medium, especially the nitrogen source yeast extract. The optimized medium for producing maximum β-glucosidase specific activity consisted of 0.275% yeast extract, 1.125% cellobiose, and 2.6% ammonium sulfate at a pH value of 3 (Vaithanomsat et al.,2011). Solid state fermentation production of a highly glucose tolerant β-glucosidase by a novel isolate of Paecilomyces was optimized using a two step statistical experiment design. In the first step which employed a Plackett–Burman design, the effects of 10

parameters such as moisture, temperature, pH, inoculum concentration, incubation time and different concentrations of (NH4)2SO4, KH 2PO4, NaCl, peptone and cellobiose were evaluated. The parameters with significant influence on the process were selected in the second step using a Box–Behnken design. The model obtained was validated and a peptone concentration of (2 g/ l), inoculum concentration of 1.2× 106 spores/ml and an incubation period of 96 h were found to be optimum for the maximum production of the enzyme. The optimization resulted in a doubling of the enzyme production by the fungus (Job et al., 2010). Sequential optimization strategy based statistical design was employed to enhance the production of cellulase enzyme from Trichoderma reesei RutC30 using agricultural waste rice straw and banana fiber as carbon source through submerged cultivation. A fractional factorial design (26-2) was applied to elucidate the process parameters that significantly affect cellulase production. Temperature, Substrate concentration, Inducer concentration, pH, inoculum age and agitation speed were identified as important process parameters effecting cellulase enzyme synthesis. The concentration of lignocelluloses and lactose (inducer) in the cultivation medium were found to be most significant factors. The steepest ascent method was used to locate the optimal domain and a Central Composite Design (CCD) was used to estimate the quadratic response surface from which the factor levels for maximum production of cellulase were determined (Muthuvelayudham and Viruthagiri 2010). Initial screening of different variables affecting β-glucosidase production from Aspergillus terreus was performed using Plackett-Burman design and the variables with statistically significant effects were identified. The optimal levels of the most significant variables with positive effect and the effect of their mutual interactions on β-glucosidase production were determined using Box-Behnken design. Fifteen variables including temperature, pH, incubation time, inoculum size, moisture content, substrate concentration,

NaNO3,

KH2PO4,

MgSO4 · 7H2O,

KCl,

CaCl2,

yeast

extract,

FeSO4 · 7H2O, Tween 80, and (NH4)2SO4 were screened in 20 experimental runs. Among the 15 variables, NaNO3, KH2PO4 and Tween 80 were found as the most significant 11

factors with positive effect on β-glucosidase production. The Box-Behnken design was used for further optimization of these selected factors for better β-glucosidase production. The maximum β-glucosidase production was 4457.162 IU /g (El-Naggar, 2015). 2.1.7. Purification and characterization of β-glucosidase and cellulase enzymes Purification of enzymes is essential for detailed studies on their properties. Also it is essential when a pure enzyme is needed for a given target application. In general, the method of purification involves concentration of the enzyme sample followed by separation of different proteins which is done by chromatographic techniques. Several researchers have successfully purified β glucosidases from Aspergilli and have established their characteristics. Notably Gunata & Vallier (1999) had reported the production of glucose tolerant β glucosidases from Aspergillus niger and Aspergillus oryzae. They had reported the purification of two BGLs from A. oryzae and the major protein was strongly inhibited by glucose. The minor protein was glucose tolerant with a Ki of 0.95M but was expressed only in minor quantities. The molecular mass determined for the minor BGL was 30 KDa, whereas this was 80 KDa for the major BGL. While Gunata & Vallier (1999) employed ultra-filtration followed by gel permeation chromatography. Riou et al (1998) had used ammonium sulfate fractionation, followed by gel permeation and ion exchange chromatography to purify BGLs from Aspergillus niger. Other methods including affinity chromatography (Watanabe et al, 1992) and aqueous two phase partitioning (Johansson & Reczey, 1998), have also been used for purification of BGLs from Aspergilli. It is now an established fact that Aspergillus produces different BGL and at least two BGLs are secreted by the fungus one which is highly expressed but with lower glucose tolerance and the second which is glucose tolerant but produced in lower quantities. It is also observed that the major BGL has a higher molecular weight while the glucose tolerant enzyme has a low molecular weight (Woodward & Wiseman, 1982, Kwon et al, 1992, Gunata et al, 1993, Riou et al, 1998). Characterization of the glucose inhibition constant (Ki), and the temperature and pH optima of BGLs are important in view of assessing their suitability in biomass hydrolysis. 12

A BGL that is better tolerant to glucose and capable of acting at elevated temperatures is desirable in biomass hydrolysis (Saha & Bothast, 1996, Pandey & Mishra, 1997, Ghosh & Ghose, 2003). β-glucosidase was purified from the cell extracts of Rhizopus oryzae MIBA348 to homogeneity by successive anion exchange and size-exclusion chromatography and characterized. The enzyme has a molecular weight of 105,000. The optimum pH and the optimum temperature was 5.0 and 50ᴼC, respectively. The enzyme was active on p-nitrophenyl-β- D-glucopyranoside, cellobiose and salicine. It hydrolyzed gentiobiose and amygdaline, but was inactive on Avicel, carboxymethylcellulose, maltose and p-nitrophenyl-α- D- glucopyranoside (Takii et al., 2005). A cellulose splitting enzyme, cellulase (extracellular) extracted from A. oryzae ITCC-4857.01 was purified by ion-exchange chromatography using DEAE-cellulose followed by Gel filtration. The purification achieved was 53 fold from the crude extract with a yield of 49 %. The purified enzyme was homogenous as judged by disc gel electrophoresis. The molecular weight as determined by gel filtration and SDSpolyacrylamide gel electrophoresis was 41,500 and 41,000 respectively. The purified extracellular cellulase contained only one subunit. The purified extracellular cellulase in aqueous solution gave absorption maximum at 290 nm and minimum at 260 nm. The enzyme is a glycoprotien as nature and contained 0.72 % neutral sugar. The apparent km value of the enzyme against cellulose was 0.67 %. The affinity of the enzyme with different substrates showed as the highest relative ativities on CMC followed by avicel, salicin and filter paper (Begum et al., 2009). β-glucosidase was extracted from an edible mushrooms L. edodes fruiting body and concentrated 26.5-folds by (NH4)2SO4 precipitation, followed by CM-Sephadex C-50 and Sephacryl S-300 HR chromatography. The purified enzyme showed a single 66 kDa band on SDS-PAGE (Sun et al., 2010). Two novel β-glucosidases (BGHG1 and BGHG2) was purified from the enzyme extract of Aspergillus oryzae HML366 through nondenaturing gel electrophoresis and anion- exchange chromatography. The molecular weights for BGHG1 and BGHG2 were 93 and 138 kDa, respectively. The amino acid sequences were determined by matrix-assisted laser desorption/ionization tandem time of flight. The 13

Mascot and Blast analyses indicated that BGHG1 has the same sequence as the hypothetical protein XP_001816831 from A. oryzae RIB40 (He et al., 2013). 2.1.8. Applications of β-glucosidases enzyme Enzymatic hydrolysis of lignocellulosic biomass and the subsequent fermentation of the realesed sugars is considered to be the most important route for transportation fuel generation. Ethanol can be blended with petrol, which will reduce the dependency of fossil fuels. Moreover by steam reforming, hydrogen gas can be generated from ethanol, which can be used as fuel cell hydrogen. Cellulose being cheap and easily available can replace food crops for ethanol production. Cellulosic ethanol at present is costly owing to lack of efficient biocatalysts for saccharification (Saha and Bothast, 1996(. The enzymatic hydrolysis of cellulosic material into glucose involves the synergistic action of at least three different enzymes: endoglucanase or endo-β-1,4- glucanase, exoglucanase or exocellobiohydrolase, and β-1,4- glucosidase or cellobiase. Endo-β-1,4glucanase catalyzes the hydrolysis of cellulose by randomly splitting the sugar residues within the molecule, whereas exo-β-1,4-glucanase removes monomers and dimmers, from the end of the glucan chain. The β-1,4-glucosidase hydrolyzes glucose dimers and in some cases, cellulose oligosaccharides to glucose. Since cellobiose inhibits the action of endo- and exoglucanases, β-glucosidase contributes to the efficiency of this process. Very strong activity of β-glucosidase is thus needed for the pretreatment step of lignocellulose before a further ethanol conversion (Leite et al., 2008). In addition to the role in cellulose degradation, β-glucosidase has also been attributed to several other applications. This includes the applications in pharmaceutical, cosmetic, and detergent industries (Job et al., 2010). In the flavor industry, β-glucosidases are also key enzymes in the enzymatic release of aromatic compounds from glucosidic precursors present in fruits and fermentating products (Shoseyov et al., 1990). Finally, β glucosidases can also improve the organoleptic properties of citrus fruit juices, in which the bitterness is in part due to a glucosidic compound, naringin (49,5,7-

14

trihydroxyflavanone-7- rhamnoglucoside), whose hydrolysis requires, in succession, an arhamnosidase and a β -glucosidase (Roitner et al.,1984).

2.2. Biotransformation of soy isoflavones 2.2.1. Soy flour Soy flour refers to soybeans ground finely enough to pass through a 100-mesh or smaller. It is the starting material for production of soy concentrate and soy protein isolate.  Defatted soy flour is obtained from solvent extracted flakes, and contains less than 1% oil.  Full-fat soy flour is made from unextracted, dehulled beans, and contains about 18% to 20% oil. Due to its high oil content, a specialized Alpine Fine Impact Mill must be used for grinding rather than the more common hammer mill.  Low-fat soy flour is made by adding back some oil to defatted soy flour. The lipid content varies according to specifications, usually between 4.5% and 9%.  High-fat soy flour can also be produced by adding back soybean oil to defatted flour at the level of 15%.  Lecithinated soy flour is made by adding soybean lecithin to defatted, low-fat or high-fat soy flours to increase their dispersibility and impart emulsifying properties.

The

lecithin

content

varies

up

to

15%.

http://en.wikipedia.org/wiki/Soybean 2.2.2. Phytoestrogens Phytoestrogens occur naturally in many plants, and have structural and functional similarities to the human estrogen, 17 β-estradiol (Cassidy, 1996). Other plant compounds with reported estrogenic properties include lignans, coumestans, and resorcyclic acid lactones (Fig. 2.1) (Knight and Eden, 1994). Phytoestrogens are classified according to their chemical structure and mainly fall into the class of flavonoids which include coumestans, lignans and isoflavones. 15

Fig. 2.1: Chemical structures of the human estrogen, 17β-estradiol, and the classes of

phytoestrogens compounds, isoflavones, lignans, coumestans, and resorcylic acid lactones(Knight & Eden, 1994)..

2.2.3. Soy isoflavones Similar to soy proteins, isoflavones are plant derived phytoestrogens and belong to a class of compounds known as flavonoids. They are phenolic compounds with structural homology to human estrogens (Tsangalis et al., 2002). Isoflavones are predominantly found in soybeans and non-fermented soy foods as biologically inactive glucoside conjugates comprising 80% to 95% of the total isoflavone concentration (King and Bignell, 2000). They are well known to provide beneficial effects for the prevention and treatment of many aging diseases, as well as a protective role in a range of conditions including cardiovascular diseases, high cholesterol, osteoporosis, and breast, colon and prostate cancers. There are 4 chemical forms of isoflavones found in soybeans (Fig. 2.2) namely malonyl- acetyl-, β-glycoside conjugates and aglycones (King and Bignell, 2000). The biologically active estrogen-like isoflavone compounds are the aglycone configuration of 16

genistein, daidzein, and glycitein (Setchell and Cassidy, 1999). Soybeans were first recognised to contain isoflavones more than 70 years ago, when genistin was isolated in crystalline form from a 90% methanol extract of soybeans and acid hydrolysis was shown to yield its aglycone, genistein (Walter, 1941). Since then several investigators have shown the abundance of isoflavones in soybeans and other legumes. Studies have shown that there is a large variability in concentration and composition among different soybeans or soy-proteins products and that this is a function of species differences, geographic and environmental conditions, and the extent of industrial processing of the soybean (Setchell et al. 1987). Isoflavones, of which there are approximately 230 individual types, are most commonly found in legumes with the highest amounts found in soybeans (Knight and Eden, 1994). Although there is a large variability of isoflavone composition among soybeans or soy-based food products, most dietary sources contain a mixture of derivatives based on three isoflavone aglycones with the common names genistein, daidzein, and glycitein (Fig. 2.3) (Song et al. 1999

Fig. 2.2: Four chemical forms of three analogues of isoflavones found in soybeans. Adapted from King & Bignell (2000).

17

Fig.2.3: Chemical structures of the isoflavones, daidzein, genistein and glycitein, in aglycone form (Song et al. 1999).

2.2.4. Health benefits of soy Interest in the possible health benefits associated with eating soy foods has concentrated in three main areas - cardiovascular disease, cancer, and postmenopausal symptoms and consequences, although other diseases and conditions have been examined.

a. Cardiovascular disease This is probably the area where there is strongest evidence for a health benefit. Food and Drug Administration (FDA) in the USA accepted that there was sufficient evidence that soy foods reduce cholesterol for it to allow food manufacturers to label foods which meet certain guidelines with a health claim to indicate that they may assist in reducing cardiovascular disease (Dotzel, 1999). This was supported by an official statement from the American Heart Association (AHA) in 2000 (Erdman, 2000), and a similar health claim was approved in the UK in 2002 (Gayer, 2002). The FDA concluded that ―there is significant scientific agreement that soy protein, included at a level of 25 g/day in a diet low in saturated fat and cholesterol, can help reduce total and LDL cholesterol levels, and that such reductions may reduce the risk of coronary heart disease (CHD). Recent research, have suggested that the isoflavones may assist in maintaining healthy arteries (van der Schouw et al. 2002; Walker et al. 2001). Oxidised LDL is an important early 18

initiator of formation of precursors of atherosclerotic plaques in the walls of blood vessels (Griffin, 1999). Substances that inhibit the oxidation of LDL may therefore protect against plaque formation. Many studies have shown the ability of genistein to inhibit oxidation of LDL in the test tube (Kerry and Abbey, 1998) b. Cancer and the risk of cancer There is considerable evidence supporting a decreased risk of developing breast cancer in countries with increased consumption of soy-based foods (Wiseman, 1997). Cross-cultural studies have demonstrated lower mortality from breast cancer in Japan than in the United States (Kelsey and Horn, 1993). Case-control studies suggest that breast cancer patients have lower soy intake than the control (Ingram et al. 1997). In an epidemiologic study, Wu et al. (1996) correlated increased risk of breast cancer with low intakes of soy in Asian-Americans and determined that soy intake was correlated with birthplace; people born in Asia consumed more soy than those of Asian descent born in the United States. In the presence of physiologic doses of both genistein and estrogen, the former was reported to competitively inhibit estrogen binding and slightly inhibit cellular proliferation (De la Rochefordiere et al. 2001). Lamartiniere, (2000) concluded that early exposure to genistein enhances cell differentiation of the mammary gland, and may confer a protective effect against carcinogenesis via this process. Furthermore, in vitro studies using cultured human breast cancer cells indicate that genistein inhibited the growth of both estrogen receptor-negative and estrogen receptor positive cell lines (Wang and Kurzer, 1997). In vitro studies, genistein inhibited the growth of a cultured human prostate cancer cell line, possibly through inhibition of focal adhesion kinase (Kyle et al. 1997).

c. Post-menopausal symptoms i.

Osteoporosis

Osteoporosis is a serious and potentially debilitating consequence of the menopause in many women, particularly in the West (Cooper et al. 1992). Whilst hormone 19

replacement therapy is effective in preventing bone loss in early menopause, there may be side effects (Morabito et al. 2002) and as with other menopausal symptoms women are therefore seeking other treatments for this condition. A synthetic isoflavone, ipriflavone (which interestingly gives rise to the soy isoflavone daidzein as one of a number of intermediates upon metabolism) has been studied for at least three decades (Gennari, 1997). It has been marketed since the late 1980s in at least twenty countries for the clinical treatment of osteoporosis (Gennari, 1997) with generally positive results (Maugeri et al. 1994). Doses are high, in the range 500 - 1000 mg/d (Maugeri et al. 1994) and whilst these are not achievable for isoflavones from soy foods, they may be relevant if daidzein is a major active metabolite of ipriflavone. Studies of the effects of soy and isoflavones have used in vitro methods, animal models and human studies (Anderson and Garner, 1998). The animal studies have generally used ovariectomised rodents (Ohta et al. 2002), or in some cases ovariectomised primates (Lees and Ginn 1998) and these have provided some supportive evidence

ii.

Estrogen replacement therapy

Estrogen replacement therapy (ERT) has long been used in postmenopausal women to aid in treating the symptoms of menopause, such as hot flushes and sweats, as well as a possible preventative treatment for bone loss and cardiovascular disease. Because of structural similarities of phytoestrogens to endogenous estrogen, the possibility for use of phytoestrogens, like soy isoflavones, as an alternative to traditional ERT for the treatment of menopausal symptoms has been investigated, but overall, the results are equivocal. In a 12-weeks study. For women consuming the soy supplemented diet, a significant increase in urinary daidzein excretion over the 12-week period was reported, accompanied by a rapid decrease in the number of hot flushes in the first 6 weeks, and a total reduction of 40% in the number of hot flushes in the 12 weeks of the study. Additionally, consumption of a soy extract (providing 50 mg total aglycone isoflavones per person per day) alone or in combination with ERT in the form of conjugated equine estrogens, was reported to decrease the number of hot flushes in postmenopausal women with no reported effects on 20

vaginal cytology, endometrial thickness, uterine artery pulsatility index, or other metabolic and hormonal parameters tested (Scambia et al. 2000).

d.

Antioxidant Activity

Further reports of isoflavone’s antioxidant and antifungal activity; also lend credibility to their health benefits. According to Naim et al. (1976), isoflavones inhibit lipoxygenase action and prevent peroxidative hemolysis of sheep erythrocytes in vitro. However, the extent of this effect depends on the structure of the isoflavones. Pratt and Birac (1979) also reported that soybeans, defatted soy flour, soy protein concentrates, and soy isolates have appreciable antioxidant activity detected by the rate of β-carotene bleaching in a lipid – aqueous system, which was due to phenolic compounds. Fleury et al. (1992) also showed that malonyl isoflavones are good antioxidants in a storage test carried out at 37ᴼC and in UV light-induced oxidation of a β- carotene / linoleic acid system. 2.2.5. Quantitative determination of soy isoflavones According to (Verbruggen, 2002), the AOAC international in 2001 developed a method for measuring isoflavone levels within foods using HPLC and UV detector. The overall resulting isoflavone level from this method is a little less than for most HPLC and UV methods because isoflavone calculations do not account for the malonyl and acetyl isoflavone forms. HPLC is the popular instrumentation used for identification of soy isoflavones in soybeans and soy food products or matrices. Yang et al., (2009) established HPLC system used for quantification of isoflavones which was a Shimadzu HPLC (Kyoto, Japan), consisting of an LC-10AT pump, a UV detector (SPD-10AVVP), and a Dikma Diamonsil C18 column (4.6 _ 250 mm) (Dima Co., Ltd., Orlando, FL). The mobile phase for HPLC consisted of solvent (A) 0.1% (v/v) acetic acid in filtered MilliQ water, and solvent (B) solvent 0.1% (v/v) acetic acid in acetonitrile. The following gradient for solvent B was applied: 15–25% over 35 min, 25– 26.5% over 12 min, and 26.5–50% over 30 s, followed by isocratic elution for14.5 min. 21

The flow rate was 1.0 ml/min. The column temperature was 40 ᴼC and the absorbance was measured at 254 nm For the linear HPLC gradient, solvent (A) was 0.1% glacial acetic acid in H2O, and solvent (B) was 0.1%glacial acetic acid in acetonitrile. An injection of a 20-μL ofsample was followed by the increase of solvent (B) from 15% to35% and the decrease of solvent (A) from 85% to 65% in 26 min. The solvent flow rate was 1.0 mL/min. A TSK-Gel Super-ODS HPLC column (4.6 H 100 mm) was used. The eluting components were detected from their absorbency at 254 nm. Each sample was spiked to confirm identification of genistin and genistein. The concentrations of genistin and genistein were calculated from the standard curves (Pandijtan et al., 2000). Other instrumentation used in the analysis of volatile compounds from soymilk includes

electrospray

ionization

mass

spectrometry,

gas

chromatography-mass

spectrometry to detect daidzin and genistin after solid-phase extraction and to confirm isoflavone conjugates in biological samples. line liquid chromatography–atmospheric pressure chemical ionisation mass spectrometry (LC–APCI-MS) also was found to be very accurate in quantitative determination of daidzein, genistein, glycitein, daidzin, glycitin, 6-Oacetyldaidzin,6-O-acetylglycitin and 6-O-acetylgenistin contents in selected high and low isoflavones in nutrition supplements. Improved extraction and hydrolysis methods of the isoflavones from three nutrition supplements were also studied and a rapid extraction method was developed. Comparison of different MS2 and MS3 spectra of isoflavones and some unknown compounds were also explored and proposed pathway fragments of nine isoflavones were first systematically suggested (Chen et al., 2005). 2.2.6. Isolation and identification of soy isoflavones . High-speed countercurrent chromatography, a special type of liquid–liquid partition chromatography, was applied for the preparative isolation of isoflavones in nutraceuticals and foods. By using this technique the major monoglucosylated and acetylated isoflavones from soy extracts were obtained after a cleaning-up step on Amberlite XAD-7 material. Furthermore, it was possible to isolate isoflavone aglycones as well as 22

glycosides from a red clover extract. Purity and identity of the isolated isoflavones were confirmed by HPLC with DAD, HPLC-ESI multiple-MS, and NMR spectroscopy, (Sturtz, 2006). Defatted soy flakes (DSF) was used as the starting material for the preparative isolation of isoflavones. First, the crude DSF extracts were prepared using an extractor with solvent refluxing, operated under optimal extraction conditions (using 500 mL 80% ethanol aqueous for 3 h per 100 g DSF). The extraction yield was about 10% and the purity of isoflavones was about 2.0-2.5 wt%. Before the isolation operations, the extracts were dissolved in deionized water. The isolation procedures included the method of liquid-liquid extraction and the method of column chromatography. For the method of liquid-liquid extraction using a mixed solvent of 10% n-butanol and 90% ethyl acetate operated under the optimal extraction conditions, the purity and yield of isoflavones were 35% and 70%. For the method of column chromatography, XAD-7HP and XAD-4 adsorbents with different polarities were used as the packing materials. For the XAD-4 column, a part of non-polar impurities was efficiently separated with the majority of isoflavones by a proper step gradient elution, which resulted in an efficient isolation: the purity and yield of isoflavones were 58% and 89%. In comparison, the method of column chromatography using XAD-4 adsorbents achieved both the highest purity and yield, and was found to be the best isolation method in the current isolation stage (Wu et al., 2007)

2.2.7. Biotransformation of soy isoflavones glycosides in soy products using βglucosidase enzyme The conversion of genistin in the glucoside form to genistein in the aglucone form by a β-glucosidase enzyme hydrolytic activity during soy protein concentrate (SPC) preparation was evaluated. The optimum conditions for the conversion of genistin to genistein were pH 5.0, 50 ᴼC, 2 IU /g of soy flour, 1 h incubation period, and 1:10 (w/v) defatted soy flour to water ratio. Under these conditions, 1.214 mg/g genistein were obtained in SPC and is significantly (p < 0.05) higher than the genistein content in SPC 23

prepared under similar conditions without enzyme addition (0.845 mg/g). (Pandjaitan et al., 2000). According to study conducted by (Yang et al., 2009) the β-glucosidase from Paecilomyces thermophila J18 was found to be able to hydrolyse daidzin and genistin. He evaluated the thermostability and hydrolysis of soybean isoflavone glycosides. The enzyme was found to be very stable at 50 ᴼC, and retained more than 95% of its initial activity after 8 h at 50ᴼC. It converted isoflavone glycosides, in soybean flour extract and soybean embryo extract, to their aglycones, resulting in more than 93% of hydrolysis of three isoflavone glycosides (namely, daidzin, genistin and glycitin) after 4 h of incubation. Also, addition of the β-glucosidase greatly increased the contents of isoflavone aglycones in the suspended soybean flour and soymilk. The results indicate that the thermostable β-glucosidase may be used to increase the isoflavone aglycones in soy products. This is the first report on the potential application of fungal β-glucosidases for converting isoflavone glycosides to their aglycones in soy products. Enzyme concentration, substrate concentration, pH, incubation temperature, and incubation time are among the important factors for the conversion of isoflavone glucosides to aglycones in soy germ flour was. The incubation temperature / time most significantly affected aglycone yield. A full 5 (35, 40, 45, 50, and 55) ᴼC × 6 (1, 2, 3, 4, 5, and 6) h factorial design and response surface methodology was employed to attain an optimal incubation time/temperature condition. The optimum condition producing soy germ flour with a high concentration of daidzein, glycitein, and genistein was as follows: soy germ flour:deionized water (1:5, w/v), β-glucosidase at 1 IU/g of soy germ flour, pH 5, and incubation temperature/time of 45 ᴼC/5 h. Under this optimal condition, most isoflavone glucosides were converted to aglycones with daidzein, glycitein, and genistein of ≥15.4, ≥6.16, and ≥4.147 μmol/g, respectively. In contrast, the control soy germ flour contained 13.82 μmol/g daidzin, 7.11 μmol/g glycitin, 4.40 μmol/g genistin, 1.56 μmol/g daidzein, 0.52 μmol/g glycitein, and 0.46 μmol/g genistein , (Tipkanon et al., 2010).

24

The hydrolysis of isoflavone glycosides in soy molasses by D.hansenni β-glucosides leads to increase in genestein content from (1.28 to 6.73) mg/g, in daidzein content from (0.05 to 0.23) mg/g and in glycetein conten from (0.3 to 1.68) mg/g.The hydrolysis conditions were as follow: 40ᴼC, pH 5.5, 10 IU β-glucosidase enzyme, 100 rpm for 120 min (Maitan-Alfenas et al., 2014)

25

CHAPTER 3 PRODUCTION OF β-GLUCOSIDASE BY ASPERGILLUS NIGER ON WHEAT BRAN AND GLYCEROL IN SUBMERGED CULTURE: FACTORIAL EXPERIMENTAL DESIGN AND PROCESS OPTIMIZATION 3.1. Abstract The goal of this study was to develop a fermentation process for the production of β-glucosidase, an important enzyme in the hydrolysis of lignocellulose and has many applications in food and flavor industries, using low-cost agricultural residues as substrates. Based on statistical experimental design, high-titer production of β-glucosidase on wheat bran and glycerol by Aspergillus niger in a submerged culture was achieved. A 2-level Plackett-Burman design was first used to screen the bioprocess parameters affecting β-glucosidase production. Among the tested parameters, the concentrations of wheat bran, glycerol, corn steep liquor and KCl showed significant effects on βglucosidase production. These four medium components were further optimized using a 3-level Box-Behnken design, and their optimal levels were found to be: wheat bran, 3.5 g/L; glycerol, 5 g/L; KCl, 0.1 g/L and corn steep liquor, 7.5 g/L, giving a high βglucosidase titer of 9.37 IU/ml, 2.24 folds of the maximum level obtained in the screening experiment.

Keywords: Aspergillus niger, β-glucosidase, glycerol, Plackett-burman design, Box-Behnken, wheat bran.

26

3.2. Introduction Enzymatic hydrolysis of lignocellulosic biomass and the subsequent fermentation of the sugars released is0 considered to be the most important route for generation of transportation fuel. Ethanol can be blended with petrol, which will reduce the dependency of fossil fuels. Moreover by steam reforming, hydrogen gas can be generated from ethanol, which can be used as fuel cell hydrogen. Cellulose being cheap and easily available can replace food crops for ethanol production. Cellulosic ethanol at present is costly owing to lack of efficient biocatalysts for saccharification (Saha and Bothast, 1996(. The enzymatic hydrolysis of cellulosic material into glucose involves the synergistic action of at least three different enzymes: endoglucanase or endo-β-1,4- glucanase, exoglucanase or exocellobiohydrolase, and β-1,4- glucosidase or cellobiase (Leite et al.,2008). Endo-β-1,4-glucanase catalyzes the hydrolysis of cellulose by randomly splitting the sugar residues within the molecule, whereas exo-β-1,4-glucanase removes monomers and dimmers, from the end of the glucan chain. The β-1,4-glucosidase hydrolyzes glucose dimers and in some cases, cellulose oligosaccharides to glucose. Since cellobiose inhibits the action of endo- and exoglucanases, β-glucosidase contributes to the efficiency of this process. Very strong activity of β-glucosidase is thus needed for the pretreatment step of lignocellulose before a further ethanol conversion. In addition to the role in cellulose degradation, β-glucosidase has also been attributed to several other applications. This includes the applications in pharmaceutical, cosmetic, and detergent industries (Job et al., 2010). In the flavor industry, β-glucosidases are also key enzymes in the enzymatic release of aromatic compounds from glucosidic precursors present in fruits and fermentating products (Guegen, et al., 1996; Shoseyov et al., 1990). Indeed, many natural flavor compounds, such as monoterpenols, C-13 norisoprenoids, and shikimatederived compounds, accumulate in fruits as flavorless precursors linked to mono- or diglycosides and require enzymatic or acidic hydrolysis for the liberation of their 27

fragrances (Vasserot et al., 1995; Winterhalter and Skouroumounis (1997). Finally, β -glucosidases can also improve the organoleptic properties of citrus fruit juices, in which the bitterness is in part due to a glucosidic compound, naringin (49,5,7trihydroxyflavanone-7- rhamnoglucoside), whose hydrolysis requires, in succession, an arhamnosidase and a β -glucosidase (Roitner et al.,1984). Many bacteria, yeast and filamentous fungi have been shown to produce βglucodidase. Fungi like Aspergillus oryzae and Aspergillus niger are be reported to be the best enzyme producers (Dhake and Patil, 2005 and Zhang et al, 2007). Among those microbes, the most particular of industrial fungi with GRAS (Generally regarded as safe) status as enzyme producer is the A. niger due to its easy cultivation with agroproducts such as wheat bran. The goal of this study was to develop a fermentation process for the production of βglucosidase from low-cost agricultural residues such as wheat bran and glycerol. Here, the focus is on screening potential fungi capable of producing considerable amounts of βglucosidase and optimizing the production of β-glucosidase enzyme by the selected strain using the response surface methodology.

3. 3. Materials and Methods 3.3.1. Strains Aspergillus niger NRRL 3122, Aspergillus oryzae NRRL 2217, Rhizopus oryzae NRRL 3562, Rhizopus oryzae NRRL 3563 and Pacilomyces variotii NRRL 1115 used in this study were obtained from Food Technology Research Institute,Agricultural Research Service (ARS) Culture Collection (Peoria, Illinois, USA) (Fig.3.1).

3.3.2. Substrate  Wheat bran containing (w/w) ~50% fiber (cellulose and hemicellulose), ~15% starch, ~16% protein, vitamins and minerals obtained from supermarket was used un-altered.

28

 Glycerol (analytical grade), microcrystalline cellulose and cellobiose was purchased from Acros-Organics (Fisher Chemical, USA).glucose, xylose, soluble starch and sucrose were from (Sigma-aldrich, USA).  Corn Steep liquor (CSL, 50% dry matters, composition (wt %): crude protein and amino acids 25 to 45, lactic acid 10 to 25, reducing sugars (as dextrose) 2.5, and ash (oxide) 17) was obtained from Cargill (Eddyville, IA).

3.3.3. Culture media Unless otherwise noted, the medium (M1) used in the initial culture screening for βglucosidase production in shake-flasks composed of (g/L): cellobiose, 10; corn steep liquor, 20 ; NaNO3, 3; K2HPO4, 1; KCl, 0.5; MgSO 4·7H2O, 0.5; FeSO 4·7H2O, 0.01. The initial pH of the medium was adjusted to 5.6 with 1 N NaOH.and sterilized by autoclaving at 121 ᴼC, 15 psig for 15 to 30 min depending on the medium volume.

3.3.4. Preparation of inoculum These cultures were maintained on potato-dextrose-agar (PDA) slants at 4 °C and sub-cultured every 4 weeks. Spore suspension of the different strains were prepared by scraping spores from 7 days PDA slants incubated at 30ᴼC with 10 ml of distilled water, giving a final concentration of approximately 2×107 spores/ml.

3.3.5. Culture screening for enzyme production in shake-flasks Each 250-ml flask containing 50 ml of M1 medium was inoculated with 0.4 ml of spore suspension (2×107 spores/ml) and incubated at 30ᴼC with agitation at 200 rpm for 14 days. Broth samples were taken on the 4th, 6th, 8th, 11th, and 14th days, and assayed for their β-glucosidase activities after removing the mycelia by centrifugation.

29

3.3.6. Multifactorial experiments for optimizing β-glucosidase production a. The Plackett-Burman Design. The Plackett-Burman experimental design was used to reflect the relative importance of various factors on the enzyme production. The Plackett-Burman design is based on the following first-order model: Y = βo + Σ βiXi Where Y is the response (β-glucosidase production level), βo is the model intercept, βi is the linear coefficient associated with Xi, the level of the independent variable. This model does not describe the interaction among the factors, but it can be used to screen and evaluate the important factors affecting the response. For screening purpose, various medium components and culture conditions were evaluated. Based on a Plackett-Burman factorial design, each factor was examined at 2 levels: -1 for the low level, and +1 for the high level. This design is especially practical in the case of a large number of factors and when it is unclear which settings are likely to be nearer to the optimum responses. In the present work, nine assigned independent variables (pH, concentrations of glycerol, bran, corn steep liquor, NaNO3, KCl, MgSO4·7H2O, FeSO4·7H2O, K2HPO4) were screened in 12 experimental trials. Table (3.1) illustrates the factors examined, as well as the levels of each factor used in the experimental design. Table (3.2) represents the design matrix. All experiments were carried out for 8 days in duplicate and the average of the β-glucosidase production level was taken as the response (dependant variable). The main effect of each variable was calculated as the difference between the average of measurements made at the high value (+) and low value (-).

30

Table 3.1: Plackett-Burman experimental design used for β-glucosidase production optimization Factor

Symbol

Low level (-1)

High level (+1)

Glycerol

X1

5 g/L

15 g/L

Bran

X2

5 g/L

15 g/L

Corn steep liquor

X3

10 g/L

30 g/L

NaNO3

X4

1 g/L

4 g/L

KCl

X5

0.1 g/L

1 g/L

MgSO4·7H2O

X6

0.1 g/L

1 g/L

FeSO4·7H2O

X7

0.01 g/L

0.05 g/L

K2HPO4

X8

0.3 g/L

2 g/L

pH

X9

4.5

6.0

b. Box-Behnken Design The variables with significant effects on enzyme production, as identified by the Plackett–Burman design were further optimized using a response surface Box- Behnken design (Box and Behnken 1960). The design composed of 27 experiments, where each variable was tested at three levels and in multiple combinations with the other parameters (Table 3.4). The whole set of experiments was performed in duplicate and mean response was used for analyses. The data was then fitted to a second-order polynomial equation

using the Design-Expert software. The optimal value for β- glucosidase production was then estimated using SAS JMP 8 NULL program.

31

Rhizopus oryzae (3562)

Aspergillus oryzae (2217)

Aspergillus niger (3122)

Pacilomyces variotti (1115)

Rhizopus oryzae (3563)

Fig. 3.1: Morphology of the screened strains

32

3.3.7. Analytical methods a. Enzyme Assay The activity on cellobiose was determined according to (Zaldivear et al., 2001) where 0.5 ml of the substrate which is 0.4% cellobiose in 0.05M citrate phosphate buffer (pH 4.8) is added to 0.5 ml of appropriately diluted enzyme solution and then incubated for 30 min at 50ᴼC. The reaction mixture was placed in boiling water for 5 min to stop the reaction and then immediately cooled in an ice bath. This mixture assayed By HPLC to determine glucose concentration. A unit of β-glucosidase activity was defined as the amount of enzyme that produced 1 µmol of glucose per min from cellobiose. b. Protein Asay

Protein concentration was measured by commercial protein assay kit (Bioradlaboratories, USA laboratories, USA) Bovine serum albumin was used as standard protein (Bradford ,1976). c. HPLC Glucose present in the hydrolysate were analyzed with HPLC with Aminex® HPX87H carbohydrate column (300 mm×7.8 mm) at 45 ᴼC using 0.005 N H2SO4 as mobile phase at 0.6 ml/min and the samples were detected using a refractive index detector (Shimadzu, USA) maintained at 45 ᴼC. Authentic chromatographic grade glucose was used as standards for identification and quantification of glucose in the hydrolysates

(Suwannakham and Yang, 2005). 3.4. Results 3.4.1. Screening for β-Glucosidase Producing Strains Table (3.2) shows the screening results of the five fungal strains for the cell biomass, total protein, and β-glucosidase production at different times in shake-flask fermentations. It is clear that A. niger NRRL 3122 on the 8th day gave the highest β-glucosidase activity (3.84 IU/ml), followed by A. oryzae (1.58 IU/ml), P. varioti (0.92 IU/ml), R. oryzae 33

NRRL 3562 (0.177 IU/ml) and R. oryzae NRRL 3563 (0.08 IU/ml). The results are in

good agreement with previous studies by Sternberg et al. (1977), who found that black Aspergilli were generally superior in terms of β-glucosidase production, and Soloveyva et al. (1997), who reported that the most promising fungi with respect to cellobiase activity were Aspergillus species. 3.4.2. Effect of carbon source on β-glucosidase production by Aspergillus niger β-Glucosidase production by A. niger grown on different carbon sources (10 g/L) was studied and the results are compared in Fig. 3.2. It can be seen that the highest βglucosidase production was obtained with wheat bran (4.18 IU/ml) followed by glycerol (3.72 IU/ml) and cellobiose (3.69 IU/ml) as carbon source. The higher activity produced with wheat bran might be attributed to the fact that wheat bran is a crude substrate containing proteins, cellulose, starch and minerals that apparently can enhance growth and enzyme production (Reeta,2011). High β-glucosidase production could also be due to the dual role of wheat bran as a nutrient source and a support matrix for fungal adherence (Reeta, 2011). In the shake-flask cultures, A. niger mycelia had grown as spherical pellets with various sizes depending on the carbon sources. Wheat bran was present as insoluble particulates in the medium and could increase mixing turbulence, resulting in better aeration as well as breakage of larger pellets. Smaller pellets improved mass transfer, cell growth, and enzyme production due to increased surface area (Reeta, 2011). The lowest β-glucosidase activity (2.4 IU/ml) was obtained with cellulose as the carbon source. This might be due to poor cell growth, perhaps because of a lack of adequate endo- and exo-glucanases to break down cellulose to soluble sugars. Compared to cellobiose, β-glucosidase production from glucose (3.60 IU/ml) and sucrose (3.62 IU/ml) was only slightly lower, indicating that the enzyme is constitutive in A. niger (Srivastava et al, 1987). The somewhat lower enzyme production from glucose could be attributed to glucose being a catabolite repressor and inhibitor to cellulases (Gulati and Mahadevan, 2000).

34

Table 3.2: Screening of cultures for β-glucosidase production.

Strain

Aspergillus niger NRRL 3122

Aspergillus oryzae NRRL 2217

Rhizopus oryzae NRRL 3562

Rhizopus oryzae NRRL 3563

Pacilomyces variot NRRL 1115

Incubation period (day)

Cell dry weight (g)

pH

Protein (mg/ml)

β-glucosidase (IU/ml)

4

0.333

6.42

1.97

3.10 ± 0.07

6 8 11 14 4 6 8 11 14 4 6 8 11 14 4 6 8 11 14 4 6 8 11 14

0.257 0.251 0.247 0.291 0.478 0.352 0.290 0.311 0.320 0.450 0.423 0.379 0.363 0.353 0.476 0.354 0.323 0.365 0.357 0.347 0.320 0.242 0.222 0.180

6.74 6.48 6.63 6.77 8.36 8.48 8.07 8.70 8.59 7.31 7.16 6.88 8.24 7.54 7.5 7.91 7.23 7.93 7.63 6.05 3.71 3.48 3.89 4.22

2.07 2.45 2.19 1.90 1.25 1.84 2.31 1.59 1.65 1.32 1.35 1.47 1.80 1.37 1.41 1.39 1.99 1.65 1.37 1.29 1.39 2.68 2.32 1.48

3.62 ± 0.07 3.84 ± 0.11 3.11 ± 0.06 2.93 ± 0.05 1.02 ± 0.09 1.58 ± 0.03 1.29 ± 0.09 0.94 ± 0.02 0.66 ± 0.02 0.003 ± 0.001 0.012 ± 0.001 0.066 ± 0.005 0.177 ± 0.003 0.088 ± 0.004 0.0 ± 0.0 0.0 ± 0.0 0.036 ± 0.002 0.08 ± 0.01 0.056 ± 0.003 0.24 ± 0.01 0.57 ± 0.01 0.92 ± 0.04 0.60 ± 0.003 0.43 ± 0.01

35

Fig. 3.2: Production of β-glucosidase on various carbon sources by Aspergillus niger in shakeflasks.

3.4.3. Multifactorial designs for optimizing β-glucosidase production The optimization of medium and fermentation process by selecting the best nutritional and environmental conditions is important as it can greatly increase the βglucosidase production. In this study, we applied a sequential optimization strategy, where the first phase dealt with screening and identifying the nutritional and environmental factors affecting β-glucosidase production by A. niger. Once the significant factors affecting β-glucosidase production were determined, the second phase involved ascertaining the combination leading to the maximum β-glucosidase activity. In the first phase, a Plackett-Burman experimental design was applied to reflect the relative importance of various fermentation factors. Data shown in Table 3.3 illustrates the wide variation of β-glucosidase activity from (4.25 to 5.84) IU/ml, thereby reflecting the importance of studying the medium composition for attaining a higher productivity. The main effects of the examined factors on β-glucosidase production were identified and presented graphically (Table 3.4). Since it is a 2-level experimental design, it involves a linear polynomial correlation model that describes the correlation between the nine 36

factors and the β-glucosidase activity as follows: Y = 8.383 - 0.0173 Glycerol - 0.0243 Bran - 0.036 CSL - 0.00778 NaNO3 - 0.919 KCl + 0.0176 K2HPO4 + 0.0815 MgSO4 + 2.833 FeSO4·7H2O - 0.109 pH. On analyzing the regression coefficients for the tested nine variables, it was found that KCl, CSL, wheat bran and glycerol in the concentration ranges studied all showed a significant effect on decreasing β-glucosidase production at the higher concentration level. The results suggested that the optimal concentrations would be closer to the lower levels. It is noted that CSL is widely used as rich organic nitrogen source in many industrial fermentations to produce, such as, vitamins, peptides, antibiotics, and organic acids as it is relatively inexpensive and effective in promoting cell growth and product formation (Akrinel,1970, Atlas,1995). However, too much of CSL could promote excessive cell growth, instead of enzyme production, which usually started after cells had entered the stationary phase.

37

Table 3.3: Plackett-Burman design of 9 factors and their effects on β-gluosidase production

Glycerol

#

(g/L)

Bran (g/L)

CSL (g/L)

NaNO3 (g/L)

KCl (g/L)

K2HPO4 (g/L)

MgSO4

FeSO4·

·7H2O

7H2O

(g/L)

(g/L)

pH

β-glucosidase (IU/ml)

1

15

5

30

4

1

0.3

0.1

0.01

6

4.25

2

5

15

30

1

1

2

1

0.01

4.5

4.42

3

5

5

30

1

1

2

0.1

0.05

6

4.60

4

5

5

10

1

0.1

0.3

0.1

0.01

4.5

5.84

5

15

5

10

1

1

0.3

1

0.05

4.5

5.12

6

15

15

10

4

1

2

0.1

0.01

4.5

4.86

7

15

5

30

4

0.1

2

1

0.05

4.5

5.28

8

15

15

30

1

0.1

0.3

1

0.01

6

4.87

9

5

5

10

4

0.1

2

1

0.01

6

5.72

10

15

15

10

1

0.1

2

0.1

0.05

6

5.42

11

5

15

30

4

0.1

0.3

0.1

0.05

4.5

5.09

12

5

15

10

4

1

0.3

1

0.05

6

4.98

38

Table 3.4: Analysis of the effects of 9 variables on β-glucosidase production by Aspergillus niger using Plackett-Burman design. Sorted Parameter Estimates Term

Estimate

Std Error

t Ratio

t Ratio

Prob>|t|

KCl(0.1,1)

-0.560833

0.040457

-13.86

0.0052*

Corn steep(10,30)

-0.489167

0.040457

-12.09

0.0068*

Bran(5,15)

-0.164167

0.040457

-4.06

0.0557

glycerol(5,15)

-0.1175

0.040457

-2.90

0.1009

pH(4.5,6)

-0.110833

0.040457

-2.74

0.1114

FeSO4.7H2O(0.01,0.05)

0.0775

0.040457

1.92

0.1955

MgSO4.(0.1,1)

0.0491667

0.040457

1.22

0.3483

K2HPO4(0.3,2)

0.0208333

0.040457

0.51

0.6579

NaNO3(1,4)

-0.015833

0.040457

-0.39

0.7333

To improve the pre-optimization formula for the subsequent optimization step, the variables with a negative-effect value obtained from the Plackett-Burman design were fixed at their (-1) coded values, while the variables with a positive-effect value were fixed at their (+1) coded values. To identify the optimum response region for β-glucosidase production, the significant independent variables (KCl, CSL, wheat bran, and glycerol) were further explored at three levels. Table 3.5 presents the design matrix for the variables, given in both coded and natural units, and the experimental results of βglucosidase production. To predict the optimal point, within the experimental constraints, a second-order polynomial equation was fitted to the experimental β-glucosidase activity data, and the best model is shown below: Y = 3.666 + 1.067 KCl + 0.042 CSL + 0.59 Bran + 0.407 Glycerol + (KCl 0.075)[3.105e-15 (CSL - 7.5)] + (KCl - 0.075)[2.591e-15 (Bran - 3.5)] + (KCl -

39

0.075)[-5.017e-15 (Glycerol - 3.5)] + (CSL - 7.5)[0.0053 (Bran - 3.5)] + (CSL - 7.5)[1.263e-16 (Glycerol - 3.5)] + (Bran - 3.5) [-0.222 (Glycerol - 3.5)] The main effects of the examined factors on β-glucosidase production were identified and are presented in (Table 3.6). The optimal levels were estimated using SAS JMP 8 NULL program tools and found as follows: bran, 3.5 g/L; glycerol, 5 g/L; KCl, 0.1 g/L ; CSL, 7.5 g/L. The optimal condition realized from the optimization experiment was verified experimentally and compared with the prediction calculated from the model (Fig. 3.3). The experimentally obtained β-glucosidase yield of 9.37 U/ml was comparable to the polynomial model predicted value of 8.68 U/ml; confirming the high accuracy of the model (R2 = 0.79; RSME error = 0.825) for the investigated conditions. The R2 value of 0.79 for β-glucosidase production indicated a high degree of correlation between the experimental and predicted values. It also showed the relative effects of any two variables when the third is maintained constant.

40

Table 3.5: Box-Behnken factorial experimental design and responses of β- glucosidase production as affected by KCl, CSL, bran and glycerol concentrations in the medium.

Pattern

KCl (g/L)

CSL (g/L)

Bran (g/L)

Glycerol (g/L)

00−+ 0000 0+0− 0–0− 0−−0 −00− −+00 −−00 +0+0 0+−0 0000 +0–0 0+0+ −0–0 0–0+ −00+ 00+− 00−− 0++0 −0+0 0000 0−+0 ++00 +00− +−00 00++ +00+

0.075 0.075 0.075 0.075 0.075 0.05 0.05 0.05 0.1 0.075 0.075 0.1 0.075 0.05 0.075 0.05 0.075 0.075 0.075 0.05 0.075 0.075 0.1 0.1 0.1 0.075 0.1

7.5 7.5 10 5 5 7.5 10 5 7.5 10 7.5 7.5 10 7.5 5 7.5 7.5 7.5 10 7.5 7.5 5 10 7.5 5 7.5 7.5

2 3.5 3.5 3.5 2 3.5 3.5 3.5 5 2 3.5 2 3.5 2 3.5 3.5 5 2 5 5 3.5 5 3.5 3.5 3.5 5 3.5

5 3.5 2 2 3.5 2 3.5 3.5 3.5 3.5 3.5 3.5 5 3.5 5 5 2 2 3.5 3.5 3.5 3.5 3.5 2 3.5 5 5

41

βGlucosidase (IU/ml) 6.64 8.11 7.15 6.93 5.75 7.37 7.89 7.74 8.04 5.97 8.04 6.27 8.85 6.19 8.63 9.22 8.26 5.53 7.82 7.96 8.19 7.52 7.82 7.52 7.67 7.37 9.37

Predicted βglucosidase (IU/ml) 7.77 7.54 7.04 6.83 6.57 6.91 7.62 7.41 8.46 6.74 7.54 6.69 8.26 6.63 8.05 8.13 8.32 5.55 8.55 8.4 7.54 8.3 7.68 6.96 7.47 8.54 8.68

Table. 3.6: Analysis of the effects of 4 most significant variables (Bran, Glycerol, CSL, and KCl) on β-glucosidase production by Aspergillus niger using Box-Benken design. Sorted Parameter Estimates Term Bran Glycerol (Bran-3.5)*(glycerol-3.5) CSL KCl (CSL-7.5)*(Bran-3.5) (CSL-7.5)*(glycerol-3.5) (KCl-0.075)*(CSL-7.5) (KCl-0.075)*(glycerol-3.5) (KCl-0.075)*(Bran-3.5)

Estimate 0.59 0.4066667 -0.222222 0.042 1.0666667 0.0053333 -1.26e-16 3.105e-15 -5.02e-15 2.591e-15

Std Error t Ratio t Ratio 0.158718 3.72 0.158718 2.56 0.183272 -1.21 0.095231 0.44 9.523101 0.11 0.109963 0.05 0.109963 -0.00 6.597798 0.00 10.99633 -0.00 10.99633 0.00

2

R (RSq) 2 R Adj Root Mean Square Error (RSME) Mean of Response Observations

Prob>|t| 0.0019* 0.0209* 0.2429 0.6651 0.9122 0.9619 1.0000 1.0000 1.0000 1.0000

0.7864 0.516915 0.824725 7.548889 27

Fig. 3.3: Comparison of the experimental values versus predicted values for β-glucosidase production in shake-flasks.

42

3.5. Discussion Many bacteria, yeast and filamentous fungi have been shown to produce βglucosidase. Filamentous fungi such as A. oryzae and A. niger are among the best enzyme producers [Dhake and Patil, 2005 and Zhang et al., 2007) with GRAS (generally regarded as safe) status and can grow on inexpensive agro-products such as rice straw and wheat bran (Table 3.7). In general, a higher amount of β-glucosidase can be produced by Aspergillus than by other cellulase-producing species such as Trichoderma (Lan et al., 2013). Our initial screening found that A. niger NRRL 3122 was a good β-glucosidase producer, which was then further studied to optimize the fermentation conditions in shake-flasks. Enzyme production by filamentous fungi in submerged fermentation is highly dependent on the medium composition and other culture conditions, and is usually susceptible to induction or inhibition by the carbon source. Therefore, identifying the proper medium components and optimize their compositions to maximize the enzyme production in a minimally controlled batch fermentation is important. Traditionally, fermentation processes have been optimized by changing one independent variable or factor at a time while keeping the others at some fixed values. This single dimensional search is slow and laborious, especially if a large number of independent variables are involved. Furthermore, the traditional optimization does not reflect the interaction effects among the variables and does not depict the net effect of the various factors on the enzyme activity. Consequently, statistical methods are increasingly preferred for fermentation optimization because they reduce the total number of experiments needed and provide a better understanding of the interactions among factors on the outcome of the fermentation [Revankar and Lele 2006). Statistical techniques such as the response surface methodology (RSM) have gained broad acceptance in fermentation optimization. RSM allows calculation of the optimum levels of various process parameters based on a few sets of experiments.

43

In this study, 9 variables including pH and 8 medium components were screened for their effects on β-glucosidase production by A. niger based on the Plackett-Burman experimental design. The 4 variables (KCl, CSL, bran and glycerol) with significant effects on enzyme production identified by the Plackett–Burman design were further optimized using a response surface Box-Behnken design, which gave the optimal medium composition for β-glucosidase production in shake-flasks. It is noted that higher titer and productivity could be obtained with further process optimization and different organisms such as the marine Aspergillus strain SA 58 (Elyas et al. 2010). Although higher βglucosidase production has been reported from xylan (Khisti et al 2011), malt extract (Srivastav et al. 1987), and pectin (Elyas et al.2010) (Table 3.7), those processes were not economical because of the expensive substrates used in the fermentation. In our study, we achieved high β-glucosidase production from low-cost agricultural residue wheat bran and glycerol, an abundant byproduct from biodiesel production. Furthermore, we found a synergistic effect of wheat bran and glycerol when used as co-substrates in the fermentation that gave a much higher β-glucosidase production compared to bran or glycerol alone as the substrate. Wheat bran contains ~50% (w/w) fiber (cellulose and hemicellulose), ~ 15% starches, ~ 16% proteins, and additional nutrients including vitamins and minerals that are good for cell growth and enzyme production. Using glycerol as the additional carbon source allowed the fermentation to have a quick start with good initial cell growth, while the slow release of reducing sugars from bran during the fermentation gave sustained enzyme production throughout the extended fermentation period to reach a high enzyme titer. Therefore, the process using wheat bran and glycerol as co-substrates is advantageous over those using wheat bran alone (Jager et al 2001) or other single carbon source such as Papurus paper (Ismail et al. 2007) and rice straw (Kim et al. 1997).

44

Table 3.7: Comparison of β-glucosidase production from various substrates in fermentation by Aspergillus and other species. βMicroorganism

Substrate

Glucosidase (IU/ml)

A. niger KKS

1-2% Rice straw

Productivity (IU/ml/day)

References

4

0.5

Kim et al., 1997

8.3

1.03

Kerns et al., 1986

19

1.35

Khisti et al., 2011

9.4

1.17

This study

4

0.446

Jager et al., 2001

16.5

1.17

Srivastav et al., 1987

2% Lactose + A .niger ZIMET 1% Starch A. niger NCIM

Xylan + Urea

1207

+ Glycerol

A. niger NRRL

0.35% Bran

3122

+ 0.5% Glycerol

A. phoenicis

Wheat bran

A. wentii

3% Malt extract

A. wentii

3% Cellulose

10

1.25

Srivastav et al., 1981

Pectin

80

20

Elyas et al., 2010

2.32

0.116

Ismail et al., 2007

0.261

0.016

Lan et al., 2013

Aspergillus SA 58 Fusarium oxysporum

Papyrus paper

Trichoderma

1% Bagasse

viride

+ 1% Bran

45

3.6. Conclusion Although many microbes can produce β-glucosidase, an efficient fermentation process for industrial application still needs to be developed using low-cost feedstock. Optimization of a fermentation process is often done with classical approaches with one factor tested at a time. The identification of important process variables by PlackettBurman experiments and the optimization of their levels by the Box-Behnken design can improve β-glucosidase production from (4.184 to 9.37) IU/ml, a 2.24-folds increase. The process using low-cost agricultural residue (wheat bran) and biodiesel waste (glycerol) as co-substrates should have good potential for β-glucosidase enzyme production.

46

CHAPTER 4 PRODUCTION OF β-GLUCOSIDASE BY ASPERGILLUS NIGER IN STIRRED TANK AND ROTATING FIBROUS BED BIOREACTORS 4.1. Abstract Aspergillus niger NRRL 3122 can produce considerable amounts of β-glucosidase when grown on wheat bran and glycerol as co-substrates. β-glucosidase production was first investigated in a stirred-tank bioreactor (STR) at 450 rpm and 2 vvm. About 5.41 IU/ml β-glucosidase was obtained using spore suspension as inoculum, whereas a higher production of 9.33 IU/ml was obtained using precultured cell pellets, which was comparable to that obtained in shake-flasks. The production of β-glucosidase in batch, fed-batch and repeated batch modes in a rotating fibrous bed bioreactor (RFBB) was also studied and compared to the STR. The highest β-glucosidase productivity of 1.78 IU/ml/day was obtained in the RFBB operated at a repeated batch mode, which was about 2.6-folds of that (0.68 IU/ml/day) for the free-cell batch fermentation in STR inoculated with spores. This work demonstrated that the RFBB could provide an efficient process for β-glucosidase production from low-cost wheat bran and glycerol.

Keywords: Aspergillus niger, β-glucosidase, fed-batch fermentation, repeated batch, rotating fibrous bed bioreactor

47

4.2. Introduction Commercial production of β-glucosidase is often achieved by use of species of Aspergilli. Aspergilli are known to produce higher titers of the enzyme. Nevertheless, reports on large scale production of β-glucosidase are limited. Some work has been directed to find suitable β-glucosidase and cellulase-producing microorganisms through strain selection and development (Zhang et al., 2006; Montenecourt et al., 1979; Janbon et al., 1994). Other work has been focused on improvements of fermentation processes. This has included extensive studies of not only the conventional batch but also fed-batch systems (Esterbauer et al., 1991; Persson et al, 1991). Although many of the prior studies have focused on cellulase production using different bioreactor types (Mandel and Weber, 1969; Wase et al., 1985), little has been done on β-glucosidase production in a rotating fibrous bed bioreactor (RFBB), which offers many advantages in culturing filamentous fungi for chemicals and enzymes production (Tay and Yang, 2002; Xu and Yang, 2007; Lan et al, 2013). The stirred-tank reactor (STR) widely used in β-glucosidase production is known to have high shear stress causing mycelial rupture and cellulase deactivation (Domingues et al.,2000). This shear damaging effect was observed with the more sensitive Aspergillus fumigatus but less with Trichoderma reesei (Domingues et al., 2000). In addition, the fermentation process with filamentous fungi also suffers from poor mass transfer and product secretion due to the difficulty in controlling the fungal morphology during fermentation (Talabardon and Yang, 2005). So extensive research has been done on the immobilization of fungal spores and mycelia on solid supports, including cotton towel, polyester fabrics, and alginate beads, for morphology control (Shen and Xia, 2004; Wang et al., 2010). The objective of this research was to study β-glucosidase production by Aspergillus niger using different bioreactor types and fermentation modes. Following our previous batch-culture experiments with wheat bran and glycerol as major media components in a STR (Abdella et al., 2014), attempts were made to further increase β-glucosidase 48

production in fed-batch and repeated batch modes in the present study. We also studied the RFBB under different fermentation modes as an alternative and potentially more efficient process for β-glucosidase production by A. niger from low-cost wheat bran and glycerol, which could lower the production costs for cellulosic ethanol and other chemicals (Singhania et al, 2013).

4.3. Materials and Methods 4.3.1. Strains Aspergillus niger (NRRL 3122, from Agricultural Research Service (ARS) Culture Collection (Peoria, Illinois USA) were used. The organism was maintained on potato-dextrose-agar (PDA) slants at 4ᴼC and sub cultured every 4 weeks 4.3.2. Materials  Wheat bran containing (w/w) ~50% fiber (cellulose and hemicellulose), ~15% starch, ~16% protein, vitamins and minerals obtained from supermarket was used un-altered  Glycerol (analytical grade) was purchased from Acros-Organics (Fisher Chemical, USA).  Corn Steep liquor (CSL, 50% dry matters, composition (wt%): crude protein and amino acids 25-45, lactic acid 10-25, reducing sugars (as dextrose) 2.5, and ash (oxide) 17) was obtained from Cargill (Eddyville,IA)  Silicone antifoam 204 from Sigma-aldrich  Dinitro salicylic acid  Sodium potassium tartarate  Protein assay kit from Biorad laboratories,USA

4.3.3. Inoculum preparation The stock culture of A. niger was maintained on potato-dextrose-agar (PDA) slants at 4 °C and sub-cultured every 4 weeks. Spore suspension was prepared by scraping

49

spores from PDA slants (7 days incubated at 30 °C) with 10 ml distilled water to give a final concentration of approximately 2×107 spores/ml

4.3.4. Culture media The seeding cell pellets for the fermentation kinetics studies in a stirred-tank bioreactor were prepared by inoculating concentrated spore suspension (2×108 per ml) into 150 ml preculture medium containing 10 g/L glucose, 7.5 g/L corn steep liquor, 1 g/L NaNO3, 0.3 g/L K2HPO4, 0.1 g/L KCl, MgSO4·7H2O, and 0.01 g/L FeSO4·7H2O in a shake-flask and incubated at 30 °C on a rotatory shaker at 200 rpm for 48 h. The fermentation medium was the same as the preculture medium except that glucose was replaced with wheat bran (3.5 g/L) and glycerol (5 g/L) as carbon source. Unless otherwise noted, all media were adjusted to pH 5.6 with 1 N NaOH, and sterilized by autoclaving at 121 °C, 15 psig for 15−30 min, depending on the medium volume. 4.3.5. Fermentation kinetics studies in stirred-tank bioreactor Batch fermentation kinetics was first studied in a 3-L stirred-tank reactor (STR) containing 1.5 L of the fermentation medium. After autoclaving at 121 ᴼC for 30 min, the bioreactor was inoculated with 12 ml of spore suspension (2 × 107) or 150 ml of preformed cell pellets and operated at 30ᴼwith aeration at 1–2 vvm and agitation at 450 rpm for 10 days, unless otherwise noted. The medium pH was not controlled during the fermentation. Silicone antifoam 204 (Sigma-Aldrich) was added as needed to control foaming. For fed-batch fermentation, the bioreactor initially contained 1 liter of the medium and was fed with 500 ml of the basic medium containing 1.5 g/L wheat bran and 7.5 g/L glycerol at 120 h. For the repeated-batch culture, 66% of the fermentation broth was removed and replaced with the same volume of fresh medium containing 3.5 g/L wheat bran and 7.5 g/L glycerol at the end (~240 h) of each batch. Unless otherwise noted, Samples were taken once per day, centrifuged at 13000 rpm for 10 min, and stored at -20 ᴼC for further analysis.

50

4.3.6. Fermentation kinetics studies in rotating fibrous bed bioreactor The rotating fibrous bed bioreactor was made of a 3-L stirred-tank bioreactor with a perforated stainless steel cylinder affixed with polypropylene cloth for cell immobilization mounted on the impeller shaft in the bioreactor. The bioreactor with 1.5 L of the fermentation medium was autoclaved at 121 ᴼC for 30 min. After cooling, the bioreactor was inoculated with 12 ml of spore suspension (2 × 107/ml) and then operated at 30 ᴼC with aeration at 2 vvm and agitation initially at 90 rpm and then increased to 150 rpm when wheat bran and mycelia had been adsorbed on the polypropylene cloth. The medium pH was not controlled during the fermentation. For the fed-batch culture, 500 ml of the basic medium containing 1.5 g/L wheat bran and 7.5 g/L glycerol was fed when enzyme production had leveled off at 288 h. For repeated batch culture, 66% of the fermentation broth was removed and replaced with the same volume of fresh medium containing 3.5 g/L wheat bran and 7.5 g/L glycerol at the end of the batch fermentation (~288 h) as indicated by the cessation of enzyme production. .

Fig. 4.1.:Stirred Tank Bioreactor(STR)

Fig. 4.2.:Rotating Fibrous Bed Bioreactor (RFBB) 51

4.3.7. Analytical methods a. Enzyme Assay The activity on cellobiose was determined according to Zaldivear et al., 2001 where 0.5 ml of the substrate which is 0.4% cellobiose in 0.05M citrate phosphate buffer (pH 4.8) is added to 0.5 ml of appropriately diluted enzyme solution and then incubated for 30 min at 50ᴼC. The reaction mixture was placed in boiling water for 5 min to stop the reaction and then immediately cooled in an ice bath. This mixture assayed By HPLC to determine glucose concentration. A unit of β-glucosidase activity was defined as the amount of enzyme that produced 1 µmol of glucose per min from cellobiose.

b. Protein Assay Protein concentration was measured by commercial protein assay kit (Bioradlaboratories, USA laboratories, USA) Bovine serum albumin was used as standard protein (Bradford,1976).

c. Reducing Sugar Determination Estimation of total reducing sugars was done by DNS method according to Miller, 1959 and was expressed as g/L.

d. HPLC Glucose present in the hydrolysate were analyzed with HPLC with Aminex® HPX-87H carbohydrate column (300 mm × 7.8 mm) at 45 ᴼC using 0.005 N H2SO4 as mobile phase at 0.6 ml/min and the samples were detected using a refractive index detector (Shimadzu, USA) maintained at 45 ᴼC. Authentic chromatographic grade glucose was used as standards for identification and quantification of glucose in the hydrolysates. (Suwannakham and Yang, 2005).

52

4.3.8. Morphological Identification of Aspergillus species Aspergillus niger was identified using manual about the genus aspergilli (Raper and Fennell, 1965; McClenny, 2005; Diba et al., 2007; Domsch et al., 1980; Samson and Pitt, 2000 and Gams et al., 1985). 4.3.9. Microscopic characteristics We diluted 20 µl of fermentation broth with 1 ml of PBS (phosphate buffered saline). Then 20 µl of the previous solution was incubated for 10 min with 4% paraformaldehyde. Then centrifuge at 1000 rpm for 5 min. The supernatant was discarded and the pellets were diluted with 1 ml PBS. Then100 µl of the previous solution is placed on microscopic slide to examine under microscope. The microscpe used to analyze the slides is Nikon Eclipse 80i epifluorescence microscope with intensilight C-HCFi mercury vapor lamp and DS-Qi 1 Mc digital CCD camera and an Olympus FV1000 specral confocal system.

4.4. Results 4.4.1. Free cell fermentation in stirred-tank bioreactor Batch fermentation kinetics was first studied in STR inoculated with either spores or preformed cell pellets. Fig. 4.3 shows the kinetics for batch fermentation inoculated with spores. The pH was not controlled and gradually decreased from 5.6 to ~2.78, then went back up to ~3.12. The total reducing sugars decreased from 2.7 g/L at zero time to ~0.23 g/L at 120 h, but then increased slightly to 0.41 g/L at 192 h due to the continuous saccharification of wheat bran (Lan et al.,2013; Abdella et al.,2014). Meanwhile, protein and β-glucosidase production started after 24 h and increased to reach the maximum values of 0.76 mg/ml and 5.41 IU/ml, respectively, at 192 h. It was noted that most wheat bran particles were covered by a thick and dense layer of mycelia, which led to poor mass transfer and kept wheat bran undegraded until the end of the fermentation. In addition, there were large amounts of freely dispersed mycelia, which increased the broth viscosity and further hampered mass transfer and β-glucosidase production. Compared to the

53

shake-flask fermentation, which produced 9.37

IU/ml β-glucosidase (Abdella et

al,2014), significantly lower enzyme production was obtained in the STR, which could be attributed to the unfavorable cell morphology (large mycelial clumps and compact pellets) produced directly from spores in the STR, resulting in poor mass transfer, and hence limiting cell growth and enzyme production. To overcome the mass transfer limitation caused by the poor cell morphology, preformed cell pellets were used in the fermentation and the results are shown in Fig.4.4. The pH gradually decreased from 5.5 to ~1.6, and then went back up to ~3.3. The total reducing sugars decreased from the initial 2.3 g/L to ~0.21 g/L at 144 h, then increased to 0.52 g/L at 240 h due to the continuous saccharification of wheat bran. Meanwhile, protein and β-glucosidase production reached the maximum values of 1.5 g/L and 8.6 IU/ml, respectively, at 216 h. Enzyme production with the preformed pellets was comparable to that obtained in shake-flasks, confirming the importance of cell morphology in affecting the fermentation performance.

54

Fig. 4.3.: Batch fermentation kinetics of Aspergillus niger in a stirred-tank bioreactor inoculated with spores at 450 rpm and 1 vvm RS: reducing sugar

Fig. 4.4.: Batch fermentation kinetics of Aspergillus niger in a stirred-tank bioreactor inoculated with preformed cell pellets at 450 rpm and 1 vvm (B). RS: reducing sugar

55

4.4.2. Effect of Different Aeration Rates . The reactor was operated at various agitation (300, 450 and 600 rpm) and aeration rates (0.5, 1 and 2 vvm) to study their effects on the fermentation and enzyme production. Fig. 4.5 shows the fermentation kinetics at 450 rpm at different aeration rates (0.5, 1 and 2 vvm). In general, increasing the aeration rate also increased the fermentation rate as indicated by the faster changes in the pH, reducing sugars, and protein and β-glucosidase production. At the increased aeration rate of 2 vvm, the pH rapidly dropped to 1.6 in 48 h before climbing back to ~3.4 by the end of the fermentation (240 h). Meanwhile, both protein and enzyme production reached the maximum levels of 1.63 mg/ml and 9.33 IU/ml, respectively, in 216 h.

10 1 vvm 2 vvm

β-Glucosidase

8

0.5 vvm 6 4 2 0 0

50

100

150

Time (h)

200

250

Fig. 4.5.: Effects of aeration rate on fermentation of Aspergillus niger in a stirred-tank bioreactor inoculated with cell pellets with agitation at 450 rpm.

56

4.4.3. Effect of Different Agitation Rates The effects of agitation rate on the fermentation was then studied at 2 vvm for 3 different agitation rates (300, 450 and 600 rpm), and the results are shown in Fig.4.6. It appeared that 450 rpm was the optimal agitation rate. The lower agitation rate of 300 rpm might cause poorer mixing and mass transfer due to lowered dissolved oxygen level, while the higher rate of 600 rpm imposed too much shear stress that damaged cells (Reeta et al., 2011) and greatly reduced protein and enzyme production to 0.94 mg/ml and 2.73 IU/ml, respectively.

7 300 rpm 450 rpm 600 rpm

6

pH

5 4 3 2

1 0 0

50

100

150

200

250

2.5

β-Glucosidase(IU/ml)

Reducing Sugar (g/L)

3.0 300 rpm

2.0

450 rpm

1.5 600 rpm 1.0 0.5 0.0 0

50

100

150

200

10 9 8 7 6 5 4 3 2 1 0

300 rpm 450 rpm 600 rpm

0

250

50

100

150

Time (h)

200

250

Time(h) Fig. 4.6.: Effects of agitation rate on fermentation of Aspergillus niger in a stirred-tank bioreactor inoculated with cell pellets with aeration at 2 vvm.

57

Fig. 4.7 compares the enzyme production in shake-flasks and STR operated at various agitation and aeration rates. It can be concluded that 450 rpm and 2 vvm gave the best enzyme production comparable to that from shake-flasks. Further optimization and scale up of the STR should lead to an industrially viable process for β-glucosidase

β-glucosidase(Iu/ml)

production.

10 9 8 7 6 5 4 3 2 1 0

Fig.4.7.: Comparison of β-glucosidase production in shake-flask and stirred-tank Bioreactor (STR) inoculated with spores (450 rpm and 1 vvm) or cell pellets at various agitation and aeration rates.

4.4.4. Immobilized cell fermentation in rotating fibrous bed bioreactor Spores were immobilized via adsorption to the polypropylene cloth and after germination, formed a biofilm in the RFBB (Fig. 4.2). At the beginning the medium was turbid due to suspended wheat bran and spores, but became clear after several days. There were no suspended mycelia or wheat bran after 10 days of fermentation and all cells were attached to the fibrous matrix forming a homogenous biofilm. Fig.4.8 shows 58

the fermentation kinetics in the RFBB for 14 days after inoculating with spores. Since the pH was not controlled, it gradually decreased from 5.5 to ~2.56 at 96 h and then back up to ~4.2 at 265 h. The total reducing sugars decreased from the initial 2.36 g/L to ~0.24 g/L at 192 h and then increased slightly to 0.35 g/L at 265 h. Meanwhile, protein and β glucosidase production started at ~72 h and reached the maximum levels of 1.35 g/L and 8.1 IU/ml, respectively, at 240 h. It is noted that the biofilm was made of relatively porous layer of mycelia, allowing efficient mass transfer, cell growth, and enzyme production. Unlike the free cell fermentation, there were no freely dispersed mycelia, mycelial clumps or wheat bran in the fermentation broth after 10 days. The RFBB with the biofilm operated under low viscosity and low shear environment thus allowed efficient cell growth and enzyme production (Xu and Yang, 2007; Lan et al, 013; Talabardon and Yang, 2005). Compared to free-cell fermentation in STR inoculated with spores, which produced 5.41 IU/ml β-glucosidase, significantly higher enzyme production was obtained in the RFBB. 9

3

7 6

2.5

2

5

1.5 4 3

1

2 0.5 1 0

0 0

50

100

150

200

250

Time (h) Fig. 4.8: Batch fermentation kinetics of Aspergillus niger in a RFBB inoculated with spores at 150 rpm and 2 vvm. RS: reducing sugar. 59

Protein (g/L), RS (g/L)

pH, β-glucosidase (U/ml)

RFBB

pH β-glucosidase RS protein

8

4.4.5. Fed-batch fermentation Fed-batch fermentation was studied in STR and RFBB by adding fresh medium containing wheat bran and glycerol which might be limiting the enzyme production in batch fermentation. Fig. 4.9 shows the fermentation profiles for 14 days in STR inoculated with cell pellets. The pH decreased from 5.5 to 2.43 at 48 h and then increased to ~3.14 at 96 h. After the addition of fresh medium at 120 h, it increased to 3.69 but declined again to 1.66 at 142 h and then increased gradually to 3.39. The total reducing sugars decreased from the initial 2.7 g/L to ~0.35 g/L at 96 h. After the addition of fresh medium at 120 h, it increased to 1.1 g/L but then decreased rapidly to 0.25–0.37 g/L. Meanwhile, protein and β-glucosidase production started after 24 h and reached 0.57 g/L and 3.66 IU/ml, respectively, at 96 h. After the addition of fresh medium at 120 h, protein and β-glucosidase production continued to reach their maximum levels of 1.73 g/L and 11.28 IU/ml, respectively, at 288 h. Similar profiles were observed with the RFBB fermentation (Fig. 4.10), although the total time was longer for 17 days because the additional time required for spore germination. The pH decreased from 5.5 to 2.71 and then back up to ~4.39 at 264 h. After the addition of fresh medium at 288 h, it increased to 5.0 and then declined again to 3.7 before increasing gradually to 4.1. The total reducing sugars decreased from the initial 2.9 g/L to ~0.28 g/L then increased to 0.44 g/L. After the addition of fresh medium at 288 h, it increased to 1.47 g/L then decreased to 0.4 g/L. Meanwhile, protein and enzyme production started after 72 h and increased to 1.04 g/L and 8.08 IU/ml, respectively, by 240 h. After the addition of fresh media at 288 h, they continued to increase to reach their maximum levels of 1.30 g/L and 10.6 IU/ml, respectively, at 360 h. Compared to batch fermentation, higher protein and β-glucosidase production was obtained in fed-batch fermentation in both STR and RFBB due to the replenish of limiting substrates (wheat bran and glycerol) allowing continued cell growth and enzyme production. Fed-batch fermentation is commonly used to extend production phase and

60

increase the final product titer when the limiting substrate is toxic to cells or has a low solubility (Jong et al., 1995).

3

pH β-glucosidase RS protein

10 8

STR, Fed-batch 2.5 2

6

1.5

4

1

2

0.5

0

Protein (g/L), R S (g/L)

pH, β-glucsidase (U/ml)

12

0 0

50

100

150

200

250

300

Time (h)

Fig. 4.9: Fed-batch fermentation kinetics of Aspergillus niger in a stirred-tank bioreactor inoculated with cell pellets at 450 rpm and 2 vvm .RS: reducing sugar 3.6 pH β-glucosidase protein RS

10

RFBB, Fed-batch 3.0

8

2.4

6

1.8

4

1.2

2

0.6

0

0.0 0

50

100

150

200

250

300

350

400

Time (h)

Fig. 4.10.: Fed-batch fermentation kinetics of Aspergillus niger in RFBB inoculated with cell pellets at 150 rpm and 2 vvm .RS: reducing sugar 61

Protein (g/L), RS (g/L)

pH, β-glucosidase (U/ml)

12

4.4.6. Repeated batch fermentation Repeated batch fermentation with cell recycle has the advantage of higher reactor productivity due to increased cell density in the subsequent batches. Fig. 4.11 shows the repeated batch fermentation in STR inoculated with pellets operated for 16 days. In the first batch, the pH decreased from 5.32 to 2.39 then increased to ~3.87 at 216 h. Meanwhile, the total reducing sugars decreased from the initial 3.5 g/L to ~0.21 g/L, and protein and β-glucosidase production reached 1.59 g/L and 8.58 IU/ml, respectively. The second batch started at 240 h by replacing 66% of the fermentation broth with fresh medium. The medium pH rapidly decreased from the initial value of 4.72 to 2.91 and then increased gradually to 5.32 at 380 h. Meanwhile, total reducing sugars decreased from 2.82 g/L to 0.38–0.45 g/L, and protein and β-glucosidase production increased from 0.45 g/L to 1.74 g/L and 2.59 IU/ml to 10.83 IU/ml, respectively, by 360 h. Similarly, repeated batch fermentations were also studied in RFBB (Fig.4.12). In the first batch, the pH decreased from 5.5 to 2.6 and then back up to ~4.25 at 264 h, the total reducing sugars decreased from the initial 2.73 g/L to ~0.24 g/L, and protein and β-glucosidase production reached 0.9 g/L and 8.1 IU/ml, respectively, at 240 h. After starting the second batch at 288 h, the medium pH decreased from 5.2 to 3.6 and then increased gradually to 4.3 at 408 h. Meanwhile, total reducing sugars decreased from 2.04 g/L to 0.27 g/L, and protein and β-glucosidase production increased from 0.32 g/L to 0.95 g/L and 2.48 IU/ml to 9.6 IU/ml.

62

12

3.6

pH, β-glucosidase (U/ml)

10

STR, Repeated batch 3

8

2.4

6

1.8

4

1.2

2

0.6

0

Protein (g/L), RS (g/L)

pH β-glucosidase proein RS

0 0

50

100

150

200

250

300

350

400

Time (h)

Fig. 4.11: Repeated batch fermentation kinetics of Aspergillus niger in a stirred-tank bioreactor

inoculated with cell pellets at 450 rpm and 2 vvm.RS: reducing sugar. 12

3

pH, β-glucosidase (U/ml)

10

8

RFBB, Repeated batch 2.5

2

6

1.5

4

1

2

0.5

0

Protein (g/L), RS (g/L)

pH β-glucosidase RS protein

0 0

50

100

150

200

250

300

350

400

Time (h)

Fig. 4.12: Repeated batch fermentation kinetics of Aspergillus niger in RFBB inoculated with cell

pellets at 150 rpm and 2 vvm .RS: reducing sugar

63

4.4.7. Microsopic examination of Aspergillus niger morphology Fig.4.13 demonstrates the microscopic morphology of Aspergillus niger showing large, globose, dark brown conidial heads, which become radiate, tending to split into several loose columns with age.

Fig. 4.13:Morphology of Aspergillus niger under microscope

4.5. Discussion 4.5.1. Comparison of fermentation performance under different modes Table 4.1 summarizes and compares the results obtained in STR and RFBB under various operation modes. It is clear that better fermentation performance was attained in the immobilized-cell fermentation with either cell pellets in STR or biofilm in the RFBB as compared to the free-cell fermentation with spores as inoculum.

The better

performance can be attributed to the better controlled cell morphology, resulting in better mixing and mass transfer favorable to cell growth and enzyme production (Lan et al., 2013; Talabardon and Yang, 2005). Fed-batch fermentation extended the production period to reach a higher final product titer, while repeated batch fermentation increased volumetric productivity due to increased cell density and the elimination of lag phase in subsequent batches. Although comparable fermentation performance was obtained with cell pellets in STR and biofilm in the RFBB, the latter was easier to operate under the repeated batch mode since all cells were attached to the rotating fibrous bed and the cell-

64

free broth allowed for easier medium replacement and sampling without clogging of sample port and tubing by large mycelia clumps or pellets.

4.5.2. Comparison to other studies Many bacteria, yeasts and filamentous fungi can produce β-glucosidase (Persson et al, 1991, Zhang et al, 2007; Dhake and Patil 2005; Kandari et al, 2013; Carvalho et al, 2014). Among them, filamentous fungi with GRAS (generally regarded as safe) status, such as Aspergillus, are among the best enzyme producers ((Persson et al., 1991; Khisti et al., 2011; Kandari et al, 2013). The results from this study also compare favorably with those reported in the literature (Table 4.2). With cells immobilized in the RFBB operated under the repeated batch mode, a high β-glucosidase productivity of 1.78 U/ml/day was attained, which was higher than most of the reported values in the literature. Although a higher β-glucosidase productivity of 2.5 U/ml/day was reported (Solovyeva et al, 1997), it was obtained with a different species, A. heterornorphus grown on glucose, which is a more expensive substrate than wheat bran and glycerol used in the present study. It should be noted that the fermentation performance was also affected by the substrate, in addition to the species and fermentation mode, used in the process. Further improvements in the final product titer and productivity can be achieved with medium and process optimization. Table 4.1: Comparison of β-glucosidase production in various fermenters in different Aspergillus niger

Titer (IU/ml) Productivity (IU/ml/day)

Batch STR RFBB Spor Pellet e 5.4 9.3 8.1 0.68 1.04 0.81

Fed batch STR RFBB (Pellet) 11.3 0.94

10.6 0.71

65

modes by

Repeated batch STR (pellet) RFBB First Second First Second batch batch batch batch 8.6 10.8 8.2 9.6 0.95 1.65 0.82 1.78

Table 4.2: Comparison of β-glucosidase production from various substrates in different modes of fermentation by Aspergillus and other species Mode and βProductivity Microorganism Substrate type of Glucosidase Reference (IU/ml/day) fermentation (IU/ml) T. viride Sugarcane Batch, STR 0.261 0.016 (Lan et bagasse + al.,2013) wheat bran Penicillium simplicissimum H-11

Wheat bran, rice straw, bean cake powder

STR, batch

0.15

0.037

(Bai et al.,2013)

Penicillium funiculosum

Avicel

Batch, STR

2.4

0.29

(Carvalho et al.,2014)

Fusarium oxysporum A. phoenicis

Papyrus paper Wheat bran

Shake flask

2.32

0.116

Shake flask

4.0

0.446

(Ismail et al.,2007) (Jager et al.,2001)

A. oryzae CBS 12559 A. niger KKS

quercetin

Shake flask

2.2

0.16

1-2% Rice straw

external-loop air-lift reactor, fedbatch

4

0.5

(Riou et al.,1998) (Kim et al.,1997)

A. niger 1207

NCIM Xylan + Urea + Glycerol

Shake flask

19

1.35

(Khisti et al.,2011)

A. niger 3112

NRRL Wheat bran + glycerol

Repeated batch, RFBB

9.6

1.78

This study

A. wentii

3% Cellulose

Batch, STR

10

1.25

(Srivastava et al.,1981)

A. wentii

CMcellulose 0.5% glucose

STR, batch

13.8

1.7

Batch, STR

17.6

2.5

(Kandari et al.,2013) (Solovyeva et al.,1997)

A. heterornorphus F- 3010

66

4.6. Conclusion In general, the cultivation of filamentous fungi, including A. niger, at a large-scale STR needs to be approached with caution. As found in the present study, inoculation with preformed cell pellets of proper size, instead of spores, could have major impact on enzyme production in STR because of improved mass transfer inside the size-controlled pellets, which could enhance both nutrients transport into cells and the release of produced β-glucosidase into the medium. Similarly, immobilized cells as biofilm in the RFBB also enhanced mass transfer and enzyme production. It also facilitated medium exchange and sampling as there was no freely suspended mycelium or cell in the fermentation broth.

Consequently, the highest productivity of 1.78 IU/ml/day was

obtained in the RFBB operated at the repeated batch mode, which was about 2.6-fold of that for the free-cell batch fermentation in STR inoculated with spores.

67

CHAPTER 5 PARTIAL PURIFICATION AND CHARACTERIZATION OF THE β-GLUCOSIDASE ENZYME FROM ASPERGILLUS NIGER 3122 5.1. Abstract β-glucosidase enzyme produced from Aspergillus niger 3122 has been partially purified using ammonium sulfate fractionation. The molecular weight of the β- glucosidase was estimated to be about 180 KDa. The optimal pH and stable pH range of the enzyme activity were determined as 3.98 and (3-7), respectively. The optimal temperature and temperature stability was 55ᴼ C and up to 70ᴼC respectivly. The enzyme was strongly inhibited by 5 mM of Fe+2, Fe+3, Co+2, Cu

+2

, Zn+2, K+ and Na+ ions ,citric acid,cysteine

DMSO , Na2 EDTAand SDS and stimulated by Ca+2, Mg+2 and Mn+2,mercaptoethanol and urea. The understudy β-glucosidase had a high specific affinity to cellobiose substrate with Km and Vmax were 2 mM and 200 IU/mg respectively toward cellobiose. Key words: Purification, characterization, β-glucosidase, molecular weight.

68

5.2. Introduction Purification of enzymes is crucial for the meaningful interpretation of the results. The presence of other enzymes of the same group may interfere while studying the kinetic parameters, regarding stability against various denaturants and activators (Esen and Guangor, 1993). Also it is essential when a pure enzyme is needed for a given target application. Furthermore monoclonal antibody structure analysis and chemical modification studies need the enzyme purity up to homogeneity level (Sanayl et al., 1988; Heupel et al., 1993). In general, the method of purification involves concentration of the enzyme sample followed by separation of different proteins which is done by chromatographic techniques. Several researchers have successfully purified β-glucosidases from Aspergilli and have established their characteristics. While (Gunata and Vallier, 1999) employed ultra-filtration followed by gel permeation chromatography, (Riou et al., 1998) had used ammonium sulfate fractionation, followed by gel permeation and ion exchange chromatography to purify BGLs from

Aspergillus niger.

Other methods including

affinity chromatography (Watanabe et al., 1992) and aqueous two phase partitioning (Johansson and Reczey, 1998) have also been used for purification of BGLs from Aspergilli. β-glucosidase have been purified and characterized from a variety of fungi like Sclorotium rolfsi(Sadana et al.,1988); Trichoderma reesi (Chirico and Brown,1987); Hwnicola grisea(Filho,1996), A.japonius (Sanayl,1988), A.nidulans (Hoh et al.,1993), A.fumigatus (Xiemens et al.,1996), and different local strains of A.niger (Himmel et al.,1993,Unno et al.,1993). The properties of β-glucosidase from many different species of Aspergilli have been established in review by (Bennet and Klich, 1992). In the chapter a comprehensive scheme for partial purification and characterization of β-glucosidase from Aspergills niger NRRL 3122 has been reported. 69

5.3. Materials and Methods 5.3.1. Chemicals Amicon Ultra centrifugal filters MWCO 100 KDa (Sigma-Aldrich, USA), blue stain protein ladder, (20-245) KDa (Gold Biotechnology, USA), ammonium sulphate, methyl umbelferyl glucopyranoside (MUG) (Bioworld, USA), dextran, microcrystalline, cellulose (Sigma-Aldrich,USA), soluble starch, maltodextrin, pullulan, sucrose, maltose, cellobiose, Na2EDTA, SDS, urea, dimethylsulfoxide (DMSO), citric acid,mercaptoethanol, metals (Zn+2, Ca+2, K+, Mg+2, Mn+2, Cu+2, Na+, Co+2, Zn+2, Fe+2, Fe+3). 5.3.2. Enzyme purification a. Ultrafiltration Amicon Ultra centrifugal filters MWCO 100 kDa b. Ammonium Sulfate Precipitation Fine and dried powder of ammonium sulfate was added slowly with stirring to the crude enzyme extracts to give 60% saturation at 4ᴼC, allowed to stir for 60 min, and then allowed to stand for 24 h at 4ᴼC. After centrifugation at 10,000 rpm for 20 min, supernatant was decanted and the precipitate was discarded. Ammonium sulfate was added to bring supernatant to 80% saturation under the same conditions. After centrifugation at 10,000 rpm for 20 min, supernatant was decanted and the precipitate was dissolved in 10 ml, 0.05 M citrate phosphate buffer (pH 4.8), and then dialyzed against the same buffer for 48 h. 5.3.3. Native poly acrylamide gel electrophoresis (Native-PAGE) and zymogram analysis. Standard protocol for SDS and Native PAGE were employed to prepare gels with 10% strength and were used throughout the study. Samples were concentrated using amicon ultra centrifugal filter 100 KDa before loading on to the gels. Protein was 70

estimated by Bradford method and samples were normalized to contain equal protein concentration before loading the gel in duplicates. Gels were loaded as two halves with each half containing the same samples exactly in the same order and concentration. After completion of the electrophoresis, the gels were washed once in distilled water and were divided into two parts each corresponding to a half containing all the samples as the other one. One of the halves was incubated with 10mM MUG (Methylumbeliferyl glucopyranoside) solution in citrate buffer (0.05M, pH 4.8) for 10 min at room temperature (28 ᴼC). The second half stained with Coomassie Brilliant Blue staining. BGL activity was visualized as blue –green fluorescence under long wavelength UV trans-illumination. Then it was photographed using an imaging system (Red Cell Bioscience, USA). 5.3.4. Characterization of the β-glucosidase from Aspergillus niger 3122. a. Determination of the molecular weight of β-glucosidase from Aspergillus niger 3122. The partially purified β-glucosidase enzyme was run on native PAGE (Laemmli, 1970) along with standard protein markers (Blue stain protein ladder, (20-245) KDa, Gold Biotechnology, USA). β-glucosidase activity band was visualized by Methylumbeliferyl glucopyranoside (MUG) (Bioworld, USA) staining and the Zymogram was photographed. The position of β-glucosidase enzyme was confirmed by comparison with the standard protein markers of known molecular weights. b. Determination of the optimal temperature and pH for activity of the βglucosidase from Aspergillus niger 3122. 0.5 ml of the substrate (0.4% cellobiose in citrate–phosphate buffer) was added to 0.5 ml of appropriately diluted enzyme solution (fermentation broth filtrate) and incubated under the conditions indicated in the experimental designs for 30 min. The reaction mixture was placed in boiling water for 5 min to stop the reaction and then

71

immediately cooled in an ice bath. The glucose concentration in the mixture was determined by using a glucose analyzer (YSI Biochemical Analyzer). c. Temperature stability of the β-glucosidase from Aspergillus niger 3122. The thermal stability of enzyme was done by incubating the enzyme in citrate– phosphate buffer (0.05 M, pH 4.8) at (30 to 90) ᴼC for 30 min. Then, the residual enzymatic activity was measured. d. pH stability of the β-glucosidase from Aspergillus niger 3122 The pH stability of enzyme was done by incubating the enzyme in series of buffer at pH range of (3 to 9) at 4ᴼC for 24 hr. Then the enzyme solution was adjusted to pH 4.8. Then, the residual enzymatic activity was measured. e. Substrate specificity of the β-glucosidase from Aspergillus niger 3122 0.5 mL of appropiatley diluted enzyme solution was incubated with 0.5 mL of each substrate in citrate–phosphate buffer (0.05 M, pH 4.8) at 500C for 30 min.The total amount of reducing sugars from 1% poly

sacchrides (dextran,soluble starch,

maltodextrin, and microcrystalline cellulose,pullulan ) were determined by DNS. The glucose released from 10 mM (sucrose, maltose, and cellobiose) was determined. One U of enzyme activity was defined as the amount of the enzyme that released 1 µmol glucose. f. Effect of metals ions on the β-glucosidase from Aspergillus niger 3122 The partially purified β -glucosidase were incubated with Zn+2, Ca+2, K+, Mg+2, Mn+2, Cu+2, Na+, Co+2, Zn+2, Fe+2, Fe+3 ions at final concentration of 5 mM in the 0.05 M sodium citrate buffer (pH 4.8) for 60 min at room temperature (30ᴼC). Then, the residual enzymatic activity was measured. The control test was performed without any addition of the metal ions or reagents.

72

g. Effects of chemicals on activity of β-glucosidase from Aspergillus niger 3122 The partially purified β -glucosidase were incubated with Na2EDTA, SDS, urea, dimethylsulfoxide (DMSO), mercaptoethanol and citric acid at final concentrations of 5 mM in the 50 mM sodium citrate buffer (pH 5.0) for 60 min at room temperature (30ᴼC). Then, the residual enzymatic activity was measured. The control test was performed without any addition of the metal ions or reagents. h. Determination of Kinetic Parameters of β-glucosidase from Aspergillus niger 3122 The effect of cellobiose (1-20 mM) on the reaction rate was determined at 50°C and pH 4.8. The values of the Michaelis constant (Km) and the maximum velocity (Vmax) were determined from Lineweaver–Burk plots.

5.4. Results 5.4.1. Enzyme purification

Table 5.1: Chart for the purification of β-glucosidase from Aspergillus niger 3122. Total

Protein

Total

Specific

Purification

Recovery

activity

conc

protein

activity

folds

%

(IU)

(mg/ml)

(mg)

(IU/mg)

Crude enzyme

4185

0.101

44.1

94.89

1

100

Ultra filtration

3719.7

0.359

16.15

230.32

2.42

88.8

2814

0.667

6.67

422

4.44

67.2

80% Ammonium sulphate/dialysis

A summary of the partial purification of the β-glucosidase enzyme is given in Table 5.1. The protein precipitated in the range of 70-80 % saturation contained the bulk of enzyme activity. After the purification steps mentioned above β-glucosidase enzyme was purified by 4.44-folds and specific activity increased to 422 U/mg 73

5.4.2. Characterization of the β-glucosidase from Aspergillus niger 3122. a. Determination of the molecular weight of β-glucosidase from Aspergillus niger 3122.

A

B

C

245 180 135 100 75

63 45

35 25 20

Figure 5.1.: β-glucosidase zymogram of Aspergillus niger NRRl 3122 using native polyacrylamide gel electrophoresis. Lane A: .Denaturing SDS-PAGE of the partially purified β-glucosidase enzyme Lane B: Standard protein marker(245,180,135,100,75,63,45,35,25,20). Lane C: Zymogram of β-glucosidase enzyme visualized under ultraviolet (UV) light

The presence of β-glucosidase enzyme was verified by native PAGE and Zymogram of partially purified β-glucsidase enzyme which revealead one band of activity (hydrolysis of MUG and release of 4-methylumbelliferone which gives blue fluorescence under UV), with apparent molecular weight of 180 KDa by comparison with standard protein marker (Fig. 5.1). 74

b. Determination of the optimal temperature and pH for activity of β-glucosidase

from Aspergillus niger 3122 i.Full factorial design

The optimization of temperature and pH for activity of β-glucosidase is important as it can greatly increase the enzyme activity. In the Full factorial design we applied a sequential optimization strategy, where the independent variables (pH and temperature) were tested at seven different levels. All trials were performed in duplicates and final data was the mean of the duplicate data. The averages U/ml production results were used as the responses. Data shown in Table 5.2 illustrates the wide variation of β-glucosidase activity from 0 to 288 U/ml, thereby reflecting the importance of temperature and pH for attaining the maximum enzyme activity. On analyzing the regression coefficients for the tested variables, it was found that pH showed a significant effect on β-glucosidase activity at the low level. The results suggested that the optimal activity would be closer to the lower pH levels (Table 5.3). .

Table 5.2.: Full factorial design for determination of the optimal temperature and pH for activity of β-glucosidase from Aspergillus niger 3122

Pattern 16 21 67 53 72 25 22 64 54 57 15 32 63 17

Temp.ᴼC 30 40 80 70 90 40 40 80 70 70 30 50 80 30

pH 8 3 9 5 4 7 4 6 6 9 7 4 5 9 75

IU/ml 0 32.9 0 64.9 0 0 140.33 15.66 20.81 0 6.04 210.16 44.54 0

75 73 52 77 14 41 13 51 0 24 35 56 33 36 66 65 34 0 76 31 11 61 12 43 47 46 44 71 45 62 37 27 23 74 55 26 42

90 90 70 90 30 60 30 70 60 40 50 70 50 50 80 80 50 60 90 50 30 80 30 60 60 60 60 90 60 80 50 40 40 90 70 40 60

7 5 4 9 6 3 5 3 6 6 7 8 5 8 8 7 6 6 8 3 3 3 4 5 9 8 6 3 7 4 9 9 5 6 7 8 4

76

0 0 96.68 0 25.51 67.59 59.08 6.94 90.9 39.84 4.92 3.58 165.399 2.46 0 4.25 72.29 90.9 0 72.74 17.9 15.44 55.73 239.48 0 0 90.9 0 9.4 49.01 0 0 113.69 0 2.014 0 288

Table.5.3. Pareto chart rationalizing the effect of each variable on β-glucosidase enzyme activity by full factorial design Parameter Estimates

Term

Estimate

Std Error

t Ratio

Prob>|t|

Intercept

171.80223

36.33077

4.73

|t|

pH(4,6)

-112.4247

12.43779

-9.04

|t|

pH (4,7)

-1.448256

0.118991 -12.17

0.0012*

Time (0.5, 3 h)

0.624744

0.118991 5.25

0.0135*

MnCl2 (0.5, 1)

0.1380893

0.118991 1.16

0.3298

agitation (100, 200 rpm)

-0.149464

0.132576 -1.13

0.3416

CaCl2 (0.5, 1)

-0.127256

0.118991 -1.07

0.3633

Temp. (40, 60 ˚C)

0.1119107

0.118991 0.94

0.4163

Buffer strength (0.05, 0.1 M) 0.0600893

0.118991 0.50

0.6483

-0.053744

0.118991 -0.45

0.6822

Substrate conc. (0.05, 0.1 w/v)

98

b. Optimization of the factors affecting enzymatic deglycosylation of genistin

The variables with significant effects on enzyme production, as identified by the Plackett–Burman design were further optimized using a response surface Box–Behnken design (Box and Behnken, 1960). In the case of commercial enzyme, the design comprised 25 experiments where pH, reaction time, enzyme conc. and agitation rate were tested at three levels and in multiple combinations with the other parameters (Table 6.7). In the case of enzyme produced from Aspergillus niger 3122 the design comprised 15 experiments where time and pH were tested at three levels and in multiple combinations with each other (Table 6.9). The whole set of experiments was performed in triplicate and the mean response was used for analyses.

99

Table 6.7.: Box-Behnken factorial experimental design, representing response of genistin deglycosylation by commercial β-glucosidase containing enzyme as influenced by enzyme conc., reaction time, agitation rate and pH

Pattern

Enzyme

Agitation

conc.

rate

(IU)

(rpm)

Time (h)

pH

Genistein conc. mg/g

1

0+0+

1

250

5

4

7.93

2

+0+0

0.75

300

4

5

4.88

3

-0+0

0.75

300

4

3

5.4

4

0000

0.75

250

4

4

4.73

5

--00

0.5

250

4

3

4.299

6

-00+

0.75

250

5

3

7.016

7

00+-

0.75

300

3

4

2.383

8

-+00

1

250

4

3

6.64

9

+0-0

0.75

200

4

5

5.66

10

-0-0

0.75

200

4

3

4.483

11

0—0

0.5

200

4

4

4.5

12

++00

1

250

4

5

5.45

13

00++

0.75

300

5

4

7.506

14

+00+

0.75

250

5

5

6.11

15

0+0-

1

250

3

4

3.58

16

+00-

0.75

250

3

5

2.92

17

-00-

0.75

250

3

3

3.528

18

0-0+

0.5

250

5

4

5.3

19

0++0

1

300

4

4

5.908

20

0-+0

0.5

300

4

4

3.946

21

0+-0

1

200

4

4

7.308

22

+-00

0.5

250

4

5

3.348

23

00--

0.75

200

5

4

6

24

0-0-

0.5

250

3

4

2.182

25

00--

0.75

200

3

4

3.73

N.B: Starting genistein conc. is 0.8 mg/g soy flour

100

Table 6.8.: Sorted parameter estimates the effect of each variable on genistin deglycosylation by commercial β-glucosidase containing enzyme using Box-Behnken design Term

Estimate

Std

t Ratio

t Ratio

Prob>|t|

Error Time (3, 5 h) Enzyme conc. (0.5, 1 IU/ml) Agitation rate (rpm) *Time

1.79475

0.137411 13.06

1.1034167 0.137411