Pectinolytic enzymes of Aspergillus sojae ATCC 20235: The impact of bioprocessing strategy on solid-state production and downstream processing of polygalacturonase

by

Doreen Heerd

A thesis submitted in partial fulfillment of the requirements for the degree of

Doctor of Philosophy in Biochemical Engineering

Approved, Thesis Committee

Prof. Dr. Marcelo Fernández-Lahore Jacobs University Bremen

Prof. Dr. Matthias Ullrich Jacobs University Bremen

Dr.-Ing. Dirk Holtmann DECHEMA Research Institute

Dr. Sonja Diercks-Horn Jacobs University Bremen

Date of Defense: June 14, 2013 School of Engineering and Sciences

Summary

Since antiquity up to the present Aspergillus spp. like A. oryzae or A. sojae have been used in traditional Japanese fermented food production. The long history of safe use in the production of oriental fermented food favors these microorganisms for their application in industrial enzyme production that are applied in the food industry. This thesis deals with the investigation of A. sojae ATCC 20235 as potential pectinolytic enzyme production organism with focus on polygalacturonase (PG) production under solid-state conditions. Pectinolytic enzymes have been exploited for many industrial applications, e.g. the largest industrial application of these enzymes is in juice and wine production. PGs belong to the pectinolytic enzyme group and are an inherent part of commercial enzyme preparations used for food processing. Recent articles reported about the potential of A. sojae ATCC 20235 to produce PG enzyme in submerged fermentation and via surface cultivation methods. These studies have triggered an interest on the investigation of the potential of this strain for pectinolytic enzyme production in solid-state fermentation (SSF). For this, a microbial screening between A. sojae ATCC 20235, a know pectinase producer A. niger IMI 91881, and two further A. sojae strains was performed, which revealed the potential of A. sojae ATCC 20235 as pectinase producer under solid- state culture conditions. Media design and SSF process optimization applying advanced statistical design tools yielded in 10.9 times increased PG production under optimized SSF conditions. Moreover, SSF facilitated the utilization of agricultural and agro-industrial by-products as cultivation medium, which provided the establishment of an economical process for enzyme production. Simultaneously, an enhancement in PG production was traced by microbial strain improvement. Therefore, a classical mutation and selection strategy was developed in order to generate mutants showing increased exo-PG activity. Mutation was induced by ultraviolet irradiation, and via the chemical agent N-methyl-N’-nitro-N- nitrosoguanidine (NTG). The selection of desired mutants based on a three-step strategy. The first pre-screening focused on morphological parameters regarding the sporulation. The second screening based on the selection of “zone mutants”, which enabled the detection of desired mutants with enhanced pectinase activity measured as clear zones on an agar screening medium containing polygalacturonic acid (PGA) sodium salt as substrate. These mutants were screened in the third step for enhanced exo-PG production in SSF. Repeated cycles of mutagenesis by UV rays and sequential mutagenesis employing both methods generated stable mutants

iii

showing enhanced pectinase activities. However, highest improvement of 72 % was achieved after 3 cycles of exposure to UV irradiation. PG production by the generated mutant under optimized process conditions was successfully transferred into a rotating drum type bioreactor at a scaling ratio of 100, without losses in the enzyme activity. Besides the enzyme production focus was also set on efficient enzyme recovery and PG purification. Optimization of enzyme leaching conditions facilitated sufficient PG recovery in the crude extract. PG purification was traced by combination of various chromatographic techniques. Apparently purity of PG enzyme by means of a single band on SDS-polyacrylamide gel was achieved after separating A. sojae proteins on the basis of their charge by ion exchange chromatography, followed by separation on the basis of their size and shape (size exclusion chromatography), and finally on the degree of hydrophobicity (hydrophobic interaction chromatography). Mass spectrometric characterization of A. sojae proteins led to the identification of a broad enzyme spectrum. Several application studies related to fruit juice production and wine making were performed exploring the efficiency of A. sojae enzyme extracts in comparison to commercial pectinolytic enzyme preparations. Results of this investigation revealed the potential of A. sojae ATCC 20235 as promising enzyme producer in SSF, because of its broad enzyme set and this study showed that the fungal strain is capable of secreting high levels of proteins after optimizing the culture conditions for the specific enzyme production.

iv

Acknowledgements

I am very grateful for given the chance to purse a PhD, as well as for all other things I learned during my PhD time. For this, I am especially indebted to Prof. Dr. Marcelo Fernández-Lahore for his trust in me beyond my expertise.

Special thanks to Dr. Sonja Diercks-Horn, who gave me support throughout the work, kind supervision, and thanks for her encouragement.

Thanks to Prof. Dr. Matthias Ullrich and Dr.-Ing. Dirk Holtmann for accepting to review this thesis and evaluating, for helpful and productive hints and support.

Thanks also to all DSP group members providing congenial atmosphere to work, especially to Dr. Rosa Cabrera for the warm welcome and dedication to my initial lab training. Thanks to Dr. Rami Reddy Vennapusa and to Dr. Poondi Rajesh Gavara for scientific support. Many thanks to Nina Nentwig and Nina Böttcher for the pleasant cooperation, dedicated work in the lab and for sharing their experience.

I am thankful to the members of AG Muskhelishvili, AG Ullrich and all people working in Laboratory 2 at Jacobs University Bremen, especially to Dr. Patrick Sobetzko and Rohan L. Shah for scientific support and helpful hints. Many thanks to Ulf Krause for technical support.

This project would not have been possible without financial support provided by Jacobs University Bremen through the project PGSYS / ETB-2008-44 and PGSYS EXCHANGE.

Finally, I want to express my deepest gratitude to my family and to Stefan Gerlach for their moral support and encouragement.

v

Table of Contents

Summary iii

Acknowledgements v

Chapter 1 1

General Introduction

Chapter 2 18

Pectinase enzyme-complex production by Aspergillus spp. in solid-state fermentation: A comparative study

Chapter 3 41

Statistical media design and SSF process optimization for improved PG production by Aspergillus sojae

Chapter 4 75

Microbial strain improvement for enhanced PG production

Chapter 5 103

Improved PG enzyme bioproducion by Aspergillus sojae mutant M3 and SSF process studies at bioreactor level

Chapter 6 126

Separation, purification and partial characterization of polygalacturonase derived from solid-state culture of Aspergillus sojae

Chapter 7 163

Characterization of Aspergillus sojae enzyme extracts and application studies

vi

Chapter 8 205

Concluding remarks and future prospects

Appendix A 215

General Materials and Methods

Appendix B 220

Bioreactor studies

Appendix C 242

Utilization of NW-Q fiber for PG purification

vii

List of Figures

Chapter 1 Figure 1.1 Asexual reproductive structures used in classification of koji molds 3 Figure 1.2 Mode of action of pectinases 7 Figure 1.3 Scheme of solid-state fermentation system 9 Figure 1.4 Schematic drawing of fungal growth in SSF 10

Chapter 2 Figure 2.1 Cultivation profile of A. niger IMI 91881 26 Figure 2.2 Cultivation profile of A. sojae ATCC 20235 27 Figure 2.3 Cultivation profile of A. sojae IMI 191303 28 Figure 2.4 Cultivation profile of A. sojae CBS 100928 29 Figure 2.5 Analysis of crude extracts from the 5th day of SSF by SDS-PAGE 35 Figure 2.6 Native PAGE for enzyme detection on the electrophoretic gel 36

Chapter 3 Figure 3.1 Replicate plot of 2nd screening step 54 Figure 3.2 Contour plot of the 1st optimization step 58 Figure 3.3 Contour plot of the 2nd optimization step 61 Figure 3.4 Solid-state fermentation profile of A. sojae ATCC 20235 62 Figure 3.5 Contour plot of the optimization with A. sojae CBS 100928 66 Figure 3.6 Analysis of crude extracts by SDS-PAGE 67 Figure 3.7 Native PAGE and zymogram 68

Chapter 4 Figure 4.1 Key steps of culture screening and improvement 79 Figure 4.2 Chemical structure of NTG 80 Figure 4.3 Scheme of repeated and sequential mutagenesis 81 Figure 4.4 Sweet spot plot of the screening with A. sojae ATCC 20235 86 Figure 4.5 Sweet spot plot of the screening with A. sojae CBS 100928 86 Figure 4.6 Replicate plot of results of A. sojae ATCC 20235 88 Figure 4.7 Replicate plot of results of A. sojae CBS 100928 88 Figure 4.8 Contour plot of the optimization with A. sojae CBS 100928 89

Chapter 5 Figure 5.1 Loading plot 113 Figure 5.2 Contour plots with the variation of moisture level and temperature 113 Figure 5.3 Contour plots with the variation of moisture level and time 114

viii

Figure 5.4 Cultivation profile of mutant M3 at SSF bioreactor 118 Figure 5.5 Cultivation profile of A. sojae ATCC 20235 at SSF bioreactor 120

Chapter 6 Figure 6.1 Scheme of filtration process 131 Figure 6.2 Leaching efficiency of solvents 134 Figure 6.3 Contour plots of PG activity 139 Figure 6.4 Contour plots of specific activity 139 Figure 6.5 Effect of contact time on enzyme leaching 141 Figure 6.6 Fractionation of PG enzyme by IEXC 144 Figure 6.7 Scaled up process for separation of PG enzyme by IEXC 146 Figure 6.8 Fractionation of PG with step-wise elution starting at 0.15 M NaCl 147 Figure 6.9 Analysis of IEXC fractions by SDS-PAGE 149 Figure 6.10 Fractionation of PG with step-wise elution starting at 0.10 M NaCl 150 Figure 6.11 Protein profile of IEXC fractions 151 Figure 6.12 Separation of PG in IEXC fraction by SEC 153 Figure 6.13 Protein profile of SEC fractions 154 Figure 6.14 Elution profile obtained by HIC 155 Figure 6.15 Analysis of HIC fraction by SDS-PAGE 156

Chapter 7 Figure 7.1 2D-PAGE obtained utilizing IPG of pH 3 to 10 173 Figure 7.2 Separation of proteins in SEC fraction by SDS-PAGE and 2D-PAGE 174 Figure 7.3 SDS-PAGE analysis of SEC fractions 175 Figure 7.4 Analysis of SEC fractions on 7.5 and 10 % SDS-polyacrylamide gels 176 Figure 7.5 Analysis of commercial enzyme preparations by SDS-PAGE 177 Figure 7.6 2D-PAGE obtained utilizing IPG of pH 4 to 7 178 Figure 7.7 Analysis of proteins by native PAGE and zymogram 179 Figure 7.8 Analysis of proteins on 32:1 native polyacrylamide gel and zymogram 180 Figure 7.9 Pectinolytic enzyme activities of A. sojae ATCC 20235 and mutant M3 186 Figure 7.10 Pectinase profiling in comparison to mixtures used in winemaking 188 Figure 7.11 Pectinase profiling in comparison to mixtures used in juice making 189 Figure 7.12 Clarification of cloudy apple juice 193 Figure 7.13 Trend of sugar and ethanol concentration during vinification 196 Figure 7.14 Total polyphenol content 199

Appendix B Figure 1 Vessel of the SSF bioreactor 222 Figure 2 Effect of in-situ sterilization on the medium 226

ix

Figure 3 Medium after in-situ sterilization 227 Figure 4 Effect of sterilization method on PG production 228 Figure 5 PG and protein production in intermittent mixed SSF processes 229 Figure 6 PG and protein production in continuous mixed SSF processes 230 Figure 7 Effect of mixing rate in intermittent mixed SSF processes 231 Figure 8 Effect of mixing rate in continuous mixed SSF processes 232 Figure 9 Effect of mixing rate on medium structure 233 Figure 10 Effect of aeration rate on PG production 235 Figure 11 Repeated fed-batch process 237 Figure 12 Bioprocess kinetics of PG production in repeated fed-batch process 238 Figure 13 PG production in SSF utilizing in-situ sterilized medium 239 Figure 14 PG production in SSF process with increased water recirculation 240

Appendix C Figure 1 Characteristics of NW-Q fibers during IEXC process 245 Figure 2 Effect of Amberlite® resins on the removal of colored impurities 246 Figure 3 Characteristics of NW-Q fibers in IEXC 249 Figure 4 Purification of PG by IEXC with NW-Q fibers 250

x

List of Tables

Chapter 1 Table 1.1 Classification of pectinases 6

Chapter 2 Table 2.1 Enzyme activity at the 5th day of cultivation 31

Chapter 3 Table 3.1 D-optimal design of the 1st screening step with A. sojae ATCC 20235 51 Table 3.2 Full factorial design of the 2nd screening step with A. sojae ATCC 20235 53 Table 3.3 Comparison of inducer substrates 55 Table 3.4 D-optimal design of the 1st optimization step with A. sojae ATCC 20235 57 Table 3.5 D-optimal design of the 2nd optimization step with A. sojae 60 Table 3.6 Full factorial design of the screening with A. sojae CBS 100928 64 Table 3.7 CCF design of the optimization step with A. sojae CBS 100928 65 Table 3.8 Comparison of enzyme activities 70

Chapter 4 Table 4.1 Screening results of fractional factorial design 85 Table 4.2 Optimization experiments for both strains 87 Table 4.3 Validation experiments 90 Table 4.4 Media design 91 Table 4.5 Comparison of mutants and wild types 94 Table 4.6 Enzyme profiles 97 Table 4.7 Ratio of exo-PG activity to endo-pectinase activities 98

Chapter 5 Table 5.1 CCF design of the optimization step with A. sojae mutant M3 111 Table 5.2 Validation experiments 115 Table 5.3 Comparison of SSF at culture flask and at bioreactor level 121

Chapter 6 Table 6.1 Scale up parameters of IEXC process 132 Table 6.2 Full factorial design of the screening for PG leaching 136 Table 6.3 CCF design of the optimization for PG leaching 138 Table 6.4 Validation experiments 140 Table 6.5 Summary of biochemical properties of certain PGs 157 Table 6.6 Summary of chromatographic PG purification process performance 158

xi

Chapter 7 Table 7.1 Biochemical properties of exo-PGs produced by A. sojae in SmF 182 Table 7.2 Summary of identified proteins 184 Table 7.3 Summary of pectinolytic enzyme activities 187 Table 7.4 Enzymatic extraction of apple juice 190 Table 7.5 Enzymatic extraction of blackthorn juice 192 Table 7.6 Enzymatic grape maceration 192 Table 7.7 Sugar and ethanol concentration in musts and wines 196 Table 7.8 Summary of chromatographic characteristics of red wines 197

Appendix C Table 1 Treatment of PG extract with Amberlite® resins 246 Table 2 Treatment of PG extract with pretreated Amberlite® resins 248 Table 3 Comparison of IEXC with NW-Q fibers and DEAE-Sepharose beads 251

xii

Symbols and Abbreviations

ACN Acetonitrile ATCC American Type of Culture collection cV Column volume DEAE Diethylaminoethyl DTT Dithiothreitol GH Glycoside hydrolases HCCA α-cyano-4-hydroxy cinnamic acid HIC Hydrophobic interaction chromatography IEF Isoelectric focusing IEXC Ion exchange chromatography IUBMB International Union of Biochemistry and Molecular Biology IPG Immobilized pH gradient LoF Lack of fit MALDI Matrix-assisted laser desorption ionization MBTH 3-methyl-2-benzothiozolone-hydrazone-hydrochloride MLR Multiple linear regression MS Mass spectrometry MWCO Molecular weight cut off NW-Q fibers Non-woven fibers functionalized with quaternary ammonium ligands PAGE-SDS Polyacrylamide gel electrophoresis in sodium dodecyl sulfate PG Polygalacturonase PGL Pectate lyase pI Isoelectric point PL Pectin lyase PLS Partial least squares PMG Polymethylgalacturonase PME Pectin methylesterase PPase Protopectinase RAPD Random amplification of polymorphic DNA RFLP Restriction fragment length polymorphism SEC Size exclusion chromatography SmF Submerged fermentation SSF Solid-state fermentation TFA Trifluoro acetic acid TOF Time-of-flight

xiii

VCR Volume concentration ratio Ѵ Viscosity

Units cm Centimeter d Day g Gram g dry WS Gram dry weight of substrate gds Gram dried solids h Hour kDa Kilodalton kg Kilogram L Liter M Molar mg Milligram min Minute mL Milliliter mm Millimeter mM Millimolar rpm Revolutions per minute sec Seconds U Units (enzyme activity) W Watt

xiv Chapter 1

Chapter 1.

General Introduction

1.1 Project background The research presented in this thesis forms part of the larger project “Polygalacturonase systems” (“PGSYS”) which represents an extension of the cooperation project between “Jacobs University Bremen gGmbH” (Bremen, Germany) and the “Izmir Institute of Technology” (Izmir, Turkey): “Production and downstream processing of the polygalacturonase from Aspergillus sojae: The effect of microbial morphology and fluid rheology on system global productivity under conditions close to industrial practice. A theoretical and experimental approach”. The main focus of this project was the performance of an investigation on polygalacturonase (PG) production by A. sojae under submerged culture conditions. Whereat, “PGSYS – The impact of bioprocessing strategy on the production and downstreaming of polygalacturonases by Aspergillus sojae” deals with the development of an efficient route for enzyme biosynthesis and downstream processing. “PGSYS” presents the extension of the cooperation between “Jacobs University Bremen gGmbH” and the “Izmir Institute of Technology”, further project partners included “Senzyme GmbH” (Troisdorf, Germany), “Guserbiot S.L.U.” (Victoria, Spain) and “Antibióticos S.A.” (León, Spain). The four scientific “PGSYS” project goals focused on (I) the comparison of PG production within solid-state and submerged fermentation systems, (II) microbial strain improvement to enhance PG production, (III) optimization of culture conditions for improved enzyme production employing advanced statistical designs, and (IV) the development of an efficient product recovery and purification process. At the start of “PGSYS” project the potential of A. sojae ATCC 20235 for PG enzyme production in SmF was known from the former cooperation project [1]. Furthermore, an optimization of process conditions and media design yielded in 13.5 U/mL exo-PG activity in SmF and 29.1 U/g solid via surface cultivation [2, 3]. These studies have triggered an interest on the investigation of the potential of this strain for pectinolytic enzyme production in solid-state culture, as well as eagerness of knowledge on the characteristics and technical applications of A. sojae enzyme extracts.

1 Chapter 1

1.2 The fungal strain A. sojae ATCC 20235 The genus Aspergillus is classified into the Ascomycetes, combining species with ascomycetous teleomorphs as well as species with no known teleomorphs which were previously classified into the Deuteromycetes. However, the complex issue of fungal taxonomy continues subject to changes [4]. The candidate species belongs to the group of fungi that are generally considered to reproduce asexually by means of conidia (fungi imperfecti or Deuteromycetes). Deuteromycotina include many industrially important filamentous fungi, such as Aspergillus niger and Penicillium notatum-chrysogenum. Within the Deuteromycotina, this species belongs to the class of the Hyphomycetes, which produce conidiophores on any part of the mycelium and more specifically it belongs to the genus Aspergillus [5]. Aspergillus is a genus of approximately 250 fungi species [6]. The genus Aspergillus is further divided into three subgenera, consisting of several sections, which was suggested based on the monophyletic taxonomy: Aspergillus, Fumgati, and Nidulantes. The section Flavi, which is commonly referred to as the A. flavus group, is often divided into two groups of the non-aflatoxigenic species like A. oryzae and A. sojae, and the aflatoxigenic species such as A. flavus and A. parasiticus [4]. It has been considered that A. sojae and A. oryzae are domesticated forms of A. parasiticus and A. flavus, respectively. The homology in total DNA hybridization between A. oryzae and A. flavus has been found to be 100 %, to be 91 % between A. sojae and A. parasiticus, and between these groups demonstrated in the A. flavus – A. parasiticus pairing showed 70 % relatedness in conjunction with similar genome size [7]. Because of the high degree of DNA complementarity, Kurtzmann et al. [7] proposed that A. flavus, A. oryzae, A. parasiticus and A. sojae are varieties of a single species. However, several molecular genetic techniques have been used for the classification of species within the A. flavus group. For instance, Yuan et al. [8] have shown that the very closely related A. sojae and A. parasiticus species can be successfully distinguished from each other by random amplification of polymorphic DNA (RAPD). Moreover, their results also confirmed that the sensitivity of A. sojae and A. parasiticus to the antimicrobial agent bleomycin can be used to differentiate these two species. Moreover, Lee et al. [9] demonstrated that amplified fragment length polymorphism technique is a reliable tool for differentiation of A. oryzae from A. flavus. Recently published results of Godet & Munaut [10] show a six-step strategy using real-time PCR as key tool, complemented if necessary by RAPD and DNA restriction enzyme fragment polymorphism technique for accurate identification and discrimination of nine species within the section Flavi.

2 Chapter 1

As filamentous fungi belonging to the section Flavi, the growing colonies of A. sojae are represented by two major morphologies, the vegetative and reproductive mycelia [11]. Aspergilli are characterized by a typical morphological structure of reproductive mycelium (Figure 1.1‎ ). A very conspicuous morphological character of a fungal colony is the color. Colony color is generated by masses of pigmented conidia. These conidia are produced successively from the tips of the phialides which are spread over the whole surface of the conidial head. Conidia of the A. flavus group are ranging in color from yellow-green to olive brown, while the conidiophores, specialized hyphae, are colorless [11].

Figure 1‎.1 Asexual reproductive structures used in classification of koji molds. From Jørgensen [11].

The appearance of the reproductive structures is generally basis for classical classification methods relying on morphological characters [11]. Ushijima et al. [12] stated that Sakaguchi and Yamada isolated some characteristic yellow green Aspergillus strains, which produced echinulate conidia and smooth-walled conidiophores from koji in the Japanese fermented soy-sauce production in 1944, which had been deposited as A. sojae at the “American Type of Culture Collection” (ATCC). Moreover, Ushijima et al. [12] claimed that the strain A. sojae ATCC 20235 did not meet the requirements to be classified as A. sojae on the basis of morphological parameters, but rather the strain was identified as A. oryzae var. brunneus Murakami. As aforementioned, besides the classification on the basis of morphological parameters, several molecular genetic techniques have been used for the classification of species of the A. flavus group, and it has been found that A. oryzae can be distinguished from A. sojae by comparison of the alpA sequences. Thus, based on the alpA restriction fragment length polymorphism (RFLP) the strain ATCC 20235 was reclassified as A. oryzae [4].

3 Chapter 1

It is reported that the domesticated koji molds, A. sojae and A. oryzae, do not produce sclerotia or aflatoxin [7]. Wei & Jong [13] studied several strains of the Aspergillus section Flavi for their aflatoxin-producing abilities in rice, peanut and semisynthetic medium, and concluded that no strains of the food fungi A. oryzae and A. sojae, including the strain A. sojae ATCC 20235, produced detectable levels of aflatoxins. However, homologues of several aflatoxin biosynthetic genes have been found in A. sojae and A. oryzae [14]. However, Kusumoto et al. [14] found that several strains of A. oryzae have partial deletions in the aflatoxin gene cluster. Moreover, Takahashi et al. [15] suggested that the lack of aflatoxin production by A. sojae might be related to the premature termination defect in aflRs, which deletes the C-terminal transcription activation domain that is critical for the expression of aflatoxin biosynthetic genes. Anyhow, Aspergillus oryzae and Aspergillus sojae have traditionally been used as koji molds for fermented food and beverage production in east Asian countries [11, 16]. The geographical distribution indicates that A. sojae and A. oryzae are tied to a traditional application in the east Asian food preparation, with only few isolates outside the East Asia region [11]. The term koji is Japanese, and means “naturally”, “spontaneously”, or even “artificially” molded cereals, pulses, or beans, and other plant materials used as sources of hydrolytic enzymes for food processing. Koji is also called chu, shui, or qu by the Chinese, whereas chu is differentiated by the colors of the spores of molds. The so-called “yellow robe” derived from yellow Aspergillus molds (presumably A. oryzae), while a “five-color robe” arose from randomly mix-inoculated Rhizopus, Mucor, Aspergillus, Monascus, and or Penicillium. Koji molds are still being used in koji production, such as Saké1, Miso2, Shoyu3 or Mirin4 [16, 17]. The koji molds, Aspergillus oryzae and Aspergillus sojae, are often associated with a long history of safe use in traditional food fermentations and several Aspergillus- derived food additive products had already obtained a GRAS (generally recognized as safe) status from regulatory authorities [4, 11]. Since the species have a long history of safe use, the molds are suitable as cell factories in the food and beverage industries, and in particular for the production of enzymes. Both homologous and heterologous enzymes are produced commercially using fungi. For certain

1 Saké: An alcoholic beverage made by the saccharification of rice by mold and the fermentation by yeast. 2 Miso: A paste mmade by grinding a mixture of rice koji, cooked soybeans and salt, and fermenting and brine. 3 Shoyu: The soy sauce, made by subjecting koji of soy beans and roasted wheat to long fermentation and then to long digestion in brine. 4 Mirin: A sweet alcoholic condiment, made by standing steamed rice, rice koji, and a kind of distilled liquor.

4 Chapter 1 heterologous protein production levels the mold A. sojae has been found to exceed those levels achieved by using A. niger and A. awamori [18].

1.3 Pectinolytic enzymes Pectin is one of the major plant cell wall polysaccharides which are the most abundant organic compounds found in nature. The pectin polysaccharides together with hemicellulose polysaccharides, as well as the aromatic polymer lignin interact with cellulose fibrilles creating a ridge structure that is strengthening the plant cell wall [19]. Pectins are complex heteropolysaccharides which can be divided into two different regions, the so-called “smooth” and “hairy” regions. The “smooth” regions consist of homogalacturonan segments with a backbone of α-1,4-linked D-galacuronic acid residues which can be acetylated or methylated. Two different structures are present in the “hairy” regions, xylogalacuronan which is composed of a D-xylose- substituted galacuronan backbone, and rhamnogalacuronan I in which the D- galacuronic acid residues in the backbone are interrupted by α-1,2-linked L- rhamnose residues, to which long arabinan and galactan chains can be attached. [19, 20] Due to the complex and heterogeneous structure of pectic substances, its degradation requires the combined action of various pectinolytic enzymes [21]. Depending on the substrate preference, reaction mechanism, and action pattern pectinases have been classified in three boarder groups: (I) protopectinases, (II) esterases, and (III) depolymerases [22, 23]. The first group of protopectinases (PPase) catalyzes the solubilization of protopectin to the highly polymerized soluble pectin. Pectin methylesterases (PME) catalyze de-esterification of pectin by the removal of methoxy esters, and formation of pectate. Depolymerizing enzymes are classified as hydrolases and lyases, which either catalyze the hydrolytic cleavage with the introduction of water across the oxygen bridge or break the glycosidic bond by a trans-elimination reaction. Depending on the action pattern, i.e. random or terminal, these enzymes are termed as endo- or exo-enzymes. A brief classification of several pectinolytic enzymes based on the reactions they catalyze with corresponding E.C. numbers according to the Nomenclature Committee of the International Union of Biochemistry and Molecular Biology (IUBMB) (http://brenda- enzymes.org/) is mentioned in Table 1.1‎ . [23]

5 Chapter 1

Table 1‎.1 Classification of pectinases according to Jayani et al. [23], with modifications.

Type of pectinases E.C. no. Substrate Product

Esterase Pectinmethylesterase 3.1.1.11 Pectin Pectic acid + methanol

Depolymerases a. Hydrolases Endo-polygalacturonase 3.2.1.15 Pectic acid Oligogalacturonates Exo-polygalacturonase 3.2.1.67 Pectic acid Monogalacturonates Exo-poly-galacturonosidase 3.2.1.82 Pectic acid Digalacturonates b. Lyases

Endo-pectate lyase 4.2.2.2 Pectic acid Unsaturated oligogalacturonates Exo-pectate lyase 4.2.2.9 Pectic acid Unsaturated digalacturonates Endo-pectin lyase 4.2.2.10 Pectin Unsaturated methyloligogalacturonates

The reaction places of several pectinases and their mode of action on pectic substances are also briefly illustrated in Figure 1.2‎ . The pectin backbones can be degraded by hydrolases and lyases. Most of the fungal glycoside hydrolases (GH) involved in the pectin main-chain degradation belongs to GH family 28, which were cassified into this family on the basis of sequence similarities [24]. Glycoside hydrolases can be divided into two mechanistic classes: they hydrolase their substrates either by inversion or retention of the anomeric configuration [25]. The inverting mechanism operates via a direct displacement of the leaving group by water, whereas the retaining mechanism represents a double-displacement mechanism involving a glycosyl-enzyme intermediate. Both classes employ a pair of carboxylic acids at the active site, where one residue acts as a general acid and the other as a general base (inverting mechanism) or as a general acid/base and a nucleophile/leaving group (retaining mechanism), respectively. Biely et al. [26] reported that endo-polygalacturonases (PG) of A. niger and an exo-PG of A. tubingensis belong to the class of inverting glycosidases.

6 Chapter 1

COOR OH COOR OH

O O O OH + OH A OH OH OH HO

OH COOR OH COOR PMG / PG

PME

COOCH3 OH COOH OH

O O

O O OH OH B OH OH

OH OH COOCH3 COOH

PME

COOR

COOR OH OH

O

O C O OH OH OH OH + OH

OH COOR OH COOR PL / PGL Figure 1‎.2 Mode of action of (A) PMG: polymethylgalacturonase and PG: polygalacturonase, where R=H for PG and CH3 for PMG; (B) PME: pectin methylesterase; and (C) PL: pectin lyase and PGL: pectate lyase, where R=H for PGL and CH3 for PL. The arrow indicates the reaction place of the pectinases on the pectic substances. From Gummadi & Panda [27], with slight modifications.

PGs cleave the α-1,4-glycisidic bonds between the α-galacturonic acids within the homogalacturonan. These enzymes generally prefer the non-methylesterified pectins and their activities decrease with increasing degree of methyl-esterification. According to their action pattern PGs are divided into enzymes catalysing random hydrolytic cleavage of α-1,4-glycisidic bonds producing a number of galacturonic acid oligosaccharides (endo-polygalacturonase EC 3.2.1.15), and those that act from the non-reducing end of galacturonan by releasing of one galacturonic acid residue (exo-polygalacturonase EC 3.2.1.67) (Table 1.1‎ , Figure 1.2).‎ [19, 24, 28] In contrast to PG, polymethylgalacturonase (PMG) acts preferentially on highly esterified pectin forming 6-methyl-D-galacturonate by the hydrolytic cleavage of α- 1,4-glycisidic bonds [29]. Pectin and pecate lyase are also involved in the homoglacuronan degradation by splitting α-1,4-D galacturonan linkages via β-elmination and indroduction of a double bond between C4 and C5 of the newly formed non-reducing end (Figure 1.2).‎ Pectin lyases (PL) prefer to act on towards high methylated pectins, while pectate lyases (PGL) are most active towards polygalacturonic acid [29].

7 Chapter 1

Pectin methylesterase (PME) catalyzes deesterification of the methoxyl group of pectin forming pectic acid and methanol. During degradation of pectin, PME acts before PG and PGL which need non-esterified substrates [29]. As aformentioned, the complete degradation of pectic substances requires further enzyme activities, e.g. enzymes that cleave the rhamnogalacturonan chain and such acting on xylogalacuranan. In brief, enzymes with a specific activity towards the rhamnogalacturonan I part of the pectin molecule have been reported as rhamnogalacuronan hydrolases which randomly hydrolyse the rhamnogalacturonan chain producing oligogalacturonates; rhamnogalacturonan lyases which produce unsaturated galacturonate at the non-reducing end by random transelimination; rhamnogalacturonan rhamnohydrolases, and rhamnogalacturonan galacurono- hydrolases, which are exo-acting enzymes of the GH family 28. Rhamno- galacturonan acetylesterases catalyzes hydrolytic cleavage of acetyl groups from the rhamnogalacturonan chain. Xylogalacturonan hydrolases act between two galacturonate residues in xylose-susbstituted rhamnogalacturonan chain producing xylose-galacturonate dimers. The pectin structures xylogalacturonan and rhamno- galacturonan also require accessory enzymes to remove the side chains and provide access for the main-chain degrading pectinolytic enzymes. Accessory enzymes include such as α-arabinofuranosidases, β-galacosidases, β-xylosidases, arabinanases, β-endogalactanasesand various esterases. [28, 29] Pectinolytic enzyme production occupies about 10 % of the worldwide manufacturing of enzyme preparations [30]. Bacteria, yeast and fungi have been utilized for the production of pectinases under both submerged (SmF) and solid- state fermentation (SSF) [31]. However, the industrial production of microbial pectinolytic enzymes is mainly done by filamentous fungi, especially A. niger [32]. Since the 1930s, pectinases have been exploited for many industrial applications [22]. The elimination of pectic substances is an essential step in many food processing and beverage industries. It has been reported that pectinases have a share of 25 % in the global sales of food enzymes [23]. The largest industrial application of these enzymes is in juice and wine production. Pectinases are added to increase juice yield during pressing and for clarification of clear juices. Moreover, they are added to juices where the intent is to preserve the integrity of the plant cells by selectively hydrolyzing the plant polysaccharides of the middle lamella. In the production of orange juices pectinolytic enzyme preparations with high levels of polygalacturonase activity have been utilized for the stabilization of clouds. [33] Furthermore pectinases are also used for oil extraction, removal of citric fruit peels, animal feed production, liquefaction and saccharification of biomass, and tea fermentation etc. [22].

8 Chapter 1

The commercial available pectinase preparations used in food industry are usually mixtures of pectinolytic enzymes associated with cellulytic, proteolytic and other species of enzymes apart from the main pectinases. Therefore, the observed application results cannot easily be attributed to the action of a single enzyme or a class of enzymes. [29, 34]

1.4 Solid-state fermentation Solid-state fermentation (SSF) has been defined as the cultivation process of microorganisms on a solid matrix occurring in the absence or near absence of free water; however, the substrate must have enough moisture available to support microbial growth and metabolism [35]. Most of the water is absorbed within the moist solid matrix, although droplets of water may be present between the solid particles [36]. Such a system is presented in Figure 1.3‎ .

droplet of water in the inter-particle space

moist solid particle

continuous gas phase

water & nutrients absorbed within particle

fungal hyphae Figure 1‎.3 Scheme of a solid-state fermentation system with the arrangement of moist solid particles and the continuous gas phase involving a filamentous fungus according to Mitchell et al. [36].

The solid matrix is generally a natural raw material, which serves as physical support and source of nutrients for the microorganism. However, also an inert material can be employed as solid matrix in SSF, which is soaked with a nutrient solution [35]. Generally the solid matrix should have a large surface area per unit volume. The microorganisms are growing on and in the solid matrix, which supplies the nutrients to the microbial culture and also serves as an anchorage for the cells. Hence, selection of appropriate substrate is a key aspect of SSF. SSF offers numerous opportunities in processing of agro-industrial residues. Nevertheless, it is difficult to get all required features, e.g. physical support, source of nutrients and inducer for

9 Chapter 1 enzyme production, from a single substrate; however, this could be achieved by combination of different substrates. SSF resembles the natural habitat of microorganisms. Thus it is the preferred choice for microorganisms to grow, especially for fungi [37]. Filamentous fungi are commonly used in SSF, due to their relatively high tolerance to low water activities, their high potential to excrete hydrolytic enzymes and their morphology – they have the ability to penetrate deep into intracellular and intercellular space. Generally two types of mycelia are observed in the vegetative growth of filamentous fungi: submerged and aerial mycelia (Figure 1.4‎ ). The formation of submerged mycelia arises from the growth of a wet mycelia layer on the surface, which is penetrating the solid substrate (penetrative mycelia). Further growth on the surface of the substrate particles leads to the formation of aerial mycelia. Conidia are developed on specialized aerial hyphae; spores constitute the beginning and the end of the fungal development. If a fugal spore is placed in a suitable environment, it will germinate and the submerged mycelia will grow first. Afterwards, the aerial mycelia will develop on the submerged mycelia. [38]

Aerial mycelia

Air–liquid interface Wet mycelia layer

Penetrative mycelia

Substrate Figure 1‎.4 Schematic drawing of fungal growth in SSF showing different mycelia layers. Modified from Rahardjo et al. [39].

Since ancient times, the principle of SSF has been utilized to produce fermented food, such as soy sauce, miso, as well as alcoholic beverages [40]. In Asia, enzymes and metabolites are commonly produced on large scale by SSF in traditional processes, whereas in western countries, SSF has been widely superseded by submerged fermentation (SmF) since the 1940s. Perhaps SmF was stronger developed in western countries due to the necessity to produce antibiotics on large scale at SmF, which was having enormous importance at that time. [41] SSF has unique characteristics and limitations. Compared to SmF, microbial growth and product formation occurs mainly on the surface. SSF appears to posses several biotechnological advantages, such as higher product titers are observed by lower

10 Chapter 1 usage of substrate, and the SSF process uses low volume equipment, too. Higher biomass, high enzyme yield and higher product stability contributes to enhanced productivity in SSF processes. Moreover, lower catabolic repression was observed in SSF. SSF enables the utilization of cost-efficient agricultural and agro-industrial by-products as solid substrate. However, there is a heterogeneous environment since the solid substrate is not uniform and it is not easily agitated, which leads to the formation of gradients within the SSF system. Furthermore, the moisture content in SSF processes is relatively low, which on one hand involves reducing pollution concerns and due to the low water activity, SSF is relatively resistant to bacterial contamination. On the other hand, the lack of water in SSF also causes a serious problem in heat removal. Moreover, the heat derived from the metabolism of the microbial growth in SSF causes loss of moisture. [37, 38, 41] Despite its prospects and the growing interest many biochemical engineering aspects of SSF processes still need to be solved and there is a low amenability of the process to standardization. Therefore, the process control by on-line analytic is difficult, if not impossible, e.g. pH control and moisture level. Moreover, separation of biomass is a big challenge in SSF, which is essential for the kinetic studies. Several indirect methods have been developed such as glucosamine estimation, ergosterol determination, protein measurements, dry weight changes and CO2 evaluation, but all of them have their own weakness. Furthermore, the scale-up represents a particular bottleneck because of the various gradients, which can arise during cultivation and influence the process performance, especially in static systems. [37, 41] Despite difficulties and drawbacks, SSF is gaining more and more attention from researchers and industries. SSF processes offer potential advantages for the production of enzymes, including high volumetric productivity or relatively high concentration of the product, in conjunction with unique product properties. The microorganism involved in SSF, produce and secrete enzymes for the degradation of polymeric substances into smaller and more digestible compounds. Thus, SSF processes possess great potential for the production of hydrolytic enzymes, such as pectinases or cellulases. There are numerous applications, where microbial cultures are employed for enzyme production in SSF. Pandey et al. [42] lists some of them, like the amylase production by strains of Aspergillus sp. and Mucor sp., galactosidase production by A. niger or production of phytase by A. carbonarius. Some researchers studied the production of pectinases in SmF and SSF systems. Acuña-Argüelles et al. [43] observed a difference in enzyme characteristics influenced by the type of culture method used, e.g. pectinases obtained by SSF were more stable at extreme pH and temperature values. Some of these enzymes also differ in their molecular

11 Chapter 1 mass. Martinez et al. [44] reported differences in molecular mass of isoenzymes produced in SSF and SmF. The overall productivities of SSF were reported to be superior to those obtained in SmF [43, 45]. However, a direct comparison between SSF and SmF is difficult to make because the processes differ [41]. Nevertheless, from the economic viewpoint SSF processes can bring direct economical benefits. For instance, Castillho et al. [46] performed an economic analysis on lipase production and reported 78 % higher total capital investment costs for the production of 100 m³ lipase concentrate per year compared to the production in SSF. Moreover they claimed the great advantage of the SSF process to be the extremely cheap raw material used as substrate in SSF. However, the selection of a particular strain, especially in case of obtaining production yields of commercially significant enzymes, remains a complex task. Usually, strains selected for SSF processes are different to those selected for SmF processes, i.e. very seldom are strains efficient in both systems. This reveals the need for generation of over producing strains particularly suited for each process. [47]

1.5 Aim and outline of this thesis Currently, pectinolytic enzyme used in fruit juice industries and wine production are mainly produced by fungal species, especially deriving from A. niger [22, 32]. Recent articles reported about the potential of A. sojae to produce this pectinolytic enzyme in SmF and via surface cultivation methods [2, 3]. A. sojae has a long history of safe use as koji mold and many of the products are used in the food industry, which favors this fungus as potential production organism for the synthesis of enzymes used in food industry. This thesis deals with the investigation of A. sojae ATCC 20235 as potential pectinolytic enzyme production organism with focus on polygalacturonase (PG) under solid-state conditions. Therefore, a screening for pectinolytic enzyme production under solid-state conditions had to be performed comparing enzyme titers with other Aspergillus strains and a known pectinase producer (chapter 2). The results of this investigation are published in Food and Bioproducts Processing (2012) [48]. Moreover, a part of the results was also published on a poster presentation at EFFost Conference (2009) [49]. In order to increase the PG yields, an optimization of culture conditions for improved enzyme production employing advanced statistical design tools had to be conducted as described in chapter 3. A part of the optimization results was published on a poster presentation (2010) [50]. Further publication of the complete optimization results is planned by submission to Process Biochemistry.

12 Chapter 1

Pectinolytic enzyme titers should be enhanced by microbial strain improvement applying classical mutation and screening methods as described in chapter 4. Results of this study were partially published on a poster presentation at EFFost Conference (2009) [51] and in Turkish Journal of Biology (2012) [52]. Moreover, a complete publication of these results is planned by submission to Applied Microbiology and Biotechnology. A combination of microbial strain improvement with process optimization strategy should yield in further enhanced PG production. Moreover, the scale up of optimized process conditions had to be performed to explore the potential for industrial application. Results of this investigation are presented in chapter 5. These results are in preparation for submission. Besides enzyme production, focus was also set on the development of a downstream processing strategy for product recovery and purification (chapter 6). Purification results were partially published on a poster presentation (2010) [50]. PG is one of the major food enzymes extensively used in fruit juice production and wine making. Commercial enzyme preparations generally contain other cell wall digesting enzymes in addition to PG to obtain greater juice yield or clarity. Hence, the efficiency of A. sojae enzyme extracts had to be tested in several applications related to the food industry and compared to commercial preparations. Moreover, identification and characterization of A. sojae enzymes should be realized by mass spectrometric characterization and investigating pectinolytic enzyme activities towards different pectic substances (chapter 7). A combination of these results with results of chapter 6 is in preparation for submission.

References [1] Gögus N, Tari C, Oncü S, Unluturk S, Tokatli F. Relationship between morphology, rheology and polygalacturonase production by Aspergillus sojae ATCC 20235 in submerged cultures. Biochemical Engineering Journal 2006; 32:171-178. [2] Tari C, Gögus N, Tokatli F. Optimization of biomass, pellet size and polygalacturonase production by Aspergillus sojae ATCC 20235 using response surface methodology. Enzyme and Microbial Technology 2007;40:1108-1116. [3] Ustok FI, Tari C, Gogus N. Solid-state production of polygalacturonase by Aspergillus sojae ATCC 20235. Journal of Biotechnology 2007; 127:322-334. [4] Heerikhuisen M, Van den Hondel C, Punt P. Aspergillus sojae. In: Gellissen G. Production of recombinant proteins. Novel microbial and eukaryotic expression systems. Weinheim: WILEY-VCH; 2005. 191 - 214. [5] Pitt JI, Hocking AD. Fungi and food spoilage. 1997.

13 Chapter 1

[6] Geiser DM, Klich MA, Frisvad JC, Peterson SW, Varga J, Samson RA. The current status of species recognition and identification in Aspergillus. Studies in Mycology 2007; 59:1-10. [7] Kurtzman CP, Smiley MJ, Robnett CJ, Wicklow DT. DNA relatedness among wild and domesticated species in the Aspergillus flavus group. Mycologia 1986; 78(6):955-959. [8] Yuan G-F, Liu C-S, Chen C-C. Differentiation of Aspergillus parasiticus from Aspergillus sojae byrandom amplification of polymorphic DNA. Applied and Environmental Microbiology 1995; 61(6):2384-2387. [9] Lee C-Z, Liou G-Y, Yuan G-F. Comparison of Aspergillus flavus and Aspergillus oryzae by amplified fragment lengh polymorphism. Botanical Bulletin of Academia Sinica 2004; 45:61-68. [10] Godet M, Munaut F. Molecular strategy for identification in Aspergillus section Flavi. FEMS Microbiology Letters 2010; 304:157-168. [11] Jørgensen TR. Identification and toxigenic potential of the industrially important fungi, Aspergillus oryzae and Aspergillus sojae. Journal of Food Protection 2007; 70(12):2916-2934. [12] Ushijima S, Hayashi K, Murakami H. The current taxonomic status of Aspergillus sojae used in Shoyu fermentation. Agric. Biol. Chem. 1982; 46:2365- 2367. [13] Wei D-L, Jong S-C. Production of aflatoxins by strains of Aspergillus flavus group maintained in ATCC. Mycopathologia 1986; 93:19-24. [14] Kusumoto K-I, Nogata Y, Oha H. Directed deletions in the aflatoxin biosynthesis gene homolog cluster of Aspergillus oryzae. Curr. Genet. 2000; 37:104-111. [15] Takahashi T, Chang P-K, Matsushima K, Yu J, Abe K, Bhatnagar D, Cleveland TE, Koyama Y. Nonfunctionality of Aspergillus sojae aflR in a strain of Aspergillus parasiticus with a dirupted aflR gene. Applied and Environmental Microbiology 2002: 3737-3743. [16] Wood BJB. Oriental food uses of Aspergillus. In: Smith JEPateman JA. The British Mycological Symposium Series No 1. Genetics and Physiology of Aspergillus. London: Academic Press; 1977. 481-498. [17] Murakami H. Classification of the koji mold. Journal of General and Applied Microbiology 1971; 17:281-309. [18] Heerikhuisen M, Van den Hondel C, Punt P, Van Biezen N, Albers A, Vogel K. Novel means of transformation of fungi and their use for heterologous protein production. 2001; (WO 01/09352 A2). [19] de Vries RP, Visser J. Aspergillus enzymes involved in degradation of plant cell wall polysaccharides. Microbiology and Molecular Biology Reviews 2001; 65(4):497-522.

14 Chapter 1

[20] Schols HA, Vorhagen AGJ. Pectic polysaccharides. In: Whitaker JR, Vorhagen AGJWong DWS. Handbook of Food Enzymology: CRC Press; 2002. [21] Lara-Márquez A, Zavala-Páramo MG, López-Romero E, Camacho HC. Biotechnological potential of pectinolytic complexes of fungi. Biotechnol Lett 2011; 33:859-868. [22] Kashyap DR, Vohra PK, Chopra S, Tewari R. Applications of pectinases in the commercial sector: a review. Bioresource Technology 2001; 77:215-227. [23] Jayani RS, Saxena S, Gupta R. Microbial pectinolytic enzymes: A review. Process Biochemistry 2005; 40:2931-2944. [24] Markovic O, Janecek S. Pectin degrading glycoside hydrolases of family 28: sequence-strucural features, specificities and evolution. Protein Engineering 2001; 14(9):615-631. [25] McCarter JD, Withers SG. Mechanisms of enzymatic glycoside hydrolysis. Current Opinion in Strucural Biology 1994; 4:885-892. [26] Biely P, Benen J, Heinrichová K, Kester HCM, Visser J. Inversion of configuration during hydrolysis of alpha-1,4-galacturonidic linkage by three Aspergillus polygalacturonases. FEBS Letters 1996; 382:249-255. [27] Gummadi SN, Panda T. Purification and biochemical properties of microbial pecinases - a review. Process Biochemistry 2003; 38:987-996. [28] van den Brink J, de Vries RP. Fungal enzyme sets for plant polysaccharide degradation. Applied Micriobiology and Biotechnology 2011; 91:1477-1492. [29] Pedrolli DB, Monteiro AC, Gomes E, Carmona EC. Pectin and pectinases: production, characterization and industrial application of microbial pectinolytic enzymes. The Open Biotechnology Journal 2009; 3:9-18. [30] Semenova MV, Sinitsyna OA, Morozova VV, Fedorova EA, Gusakov AV, Okunev ON, Sokolova LM, Koshelev AV, Bubnova TV, Vinetskii YP, Sinitsyn AP. Use of preparation from fungal pectin lyase in the food industry. Applied Biochemistry and Microbiology 2006; 42(6):598-602. [31] Favela-Torres E, Volke-Sepúlveda T, Viniegra-González G. Production of hydrolytic depolymerising pectinases. Food Technol. Biotechnol. 2006; 22:221- 227. [32] Naidu GSN, Panda T. Production of pectolytic enzymes - a review. Bioprocess Eng. 1998; 19:355-361. [33] Whitaker JR. Pectic substances, pectic enzymes and haze formation in fruit juices. Enzyme and Microbial Technology 1984; 6:341-349. [34] Del Cañizo AN, Hours RA, Miranda MV, Cascone O. Fractionation of fungal pectic enzymes by immobilized metal ion affinity chromatography. J. Sci. Food Agric. 1994; 64:527-531. [35] Pandey A. Solid-state fermentation. Biochemical Engineering Journal 2003; 13:81-84. [36] Mitchell DA, Krieger N, Berovic M. Solid-state fermentation bioreactors. 2006.

15 Chapter 1

[37] Singhania RR, Patel AK, Soccol CR, Pandey A. Recent advances in solid-state fermentation. Biochemical Engineering Journal 2009; 44:13-18. [38] Sato K, Sudo S. Small-scale solid-state fermentations. In: Demain ALDavies JE. Manual of industrial mcrobiology and biotechnology. Washington D.C.: ASM Press; 1999. [39] Rahardjo YSP, Tramper J, Rinzema A. Modeling conversion and ransport phenomena in solid-state fermentation: A review and perspectives. Biotechnology Advances 2006; 24:161-179. [40] Campbell-Platt G. Fermented foods - a world perspective. Food Research International 1994; 27(3):253-257. [41] Hölker U, Lenz J. Solid-state fermentation - are there any biotechnological advantages? Current Opinion in Microbiology 2005; 8:301-306. [42] Pandey A, Soccol CR, Mitchell D. New developments in solid state fermentation: I-bioprocesses and products. Process Biochemistry 2000; 35(10):1153-1169. [43] Acuña-Argüelles ME, Gutiérrez-Rojas M, Viniegra-González G, Favela-Torres E. Production and properties of three pectinolytic activities produced by Aspergillus niger in submerged and solid-state fermentation. Applied Micriobiology and Biotechnology 1995; 43:808-814. [44] Martinez MJ, Böckle B, Camarero S, Guillén A, Martinez T. MnP isoenzymes produced by two Pleurotus species in liquid culture and during wheat-straw solid-state fermentation. ACS Symposium Series 1996; 655:183-196. [45] Solís-Pereira S, Favela-Torres E, Viniegra-González G, Gutiérrez-Rojas M. Effects of different carbon sources on the synthesis of pectinase by Aspergillus niger in submerged and solid state fermentations. Applied Micriobiology and Biotechnology 1993; 39:36-41. [46] Castilho LR, Polato CMS, Baruque EA, Sant' Anna Jr. GL, Freire DMG. Economic analysis of lipase production by Penicillium restrictum in solid-state and submerged fermentations. Biochemical Engineering Journal 2000; 4:239-247. [47] Barrios-González J. Solid-state fermentation: Physiology of solid medium, its molecular basis and applications. Process Biochemistry 2012; 47:175-185. [48] Heerd D, Yegin S, Tari C, Fernandez-Lahore M. Petinase enzyme-complex production by Aspergillus spp in sold-state fermentation: A comparative study. Food and Bioproducts Processing 2012; 90:102-110. [49] Heerd D, Yegin S, Tari C, Fernandez-Lahore M. A comparative study on pectinase production by Aspergillus sojae strains in solid-state fermentation. EFFost Conference, Budapest, Hungary 2009; P330. [50] Heerd D, Gavara PR, Diercks-Horn S, Tari C, Fernandez-Lahore M. Pectinolytic enzyme production by Aspergillus sojae and purification of the pectinase complex. IBS, Rimini, Italy 2010.

16 Chapter 1

[51] Demir H, Bayasal N, Tari C, Heerd D, Fernandez-Lahore M. Optimization of polygalacturonase production from Aspergillus sojae by using orange peel. EFFost Conference, Budapest, Hungary 2009; P331. [52] Demir H, Gögus N, Tari C, Heerd D, Fernandez-Lahore M. Optimization of the process parameters for the utilization of orange peel to produce polygalacturonase by solid-state fermentation from an Aspergillus sojae mutant strain. Turkish Journal of Biology 2012; 36:1104-1123.

17 Chapter 2

Chapter 2.

Pectinase enzyme-complex production by Aspergillus spp in solid-state fermentation: A comparative study

Abstract

A comparative evaluation of three Aspergillus species according to their pectinase production in solid-state fermentation was performed. Solid-state fermentation offers several potential advantages for enzyme production by fungal strains. Utilization of agricultural by-products as low-cost substrates for microbial enzyme production resulted in an economical and promising process. The pectinolytic enzyme activities of two Aspergillus sojae strains were compared to a known producer, A. niger IMI 91881, and to A. sojae ATCC 20235, which was re-classified as A. oryzae. Evaluation of polymethylgalacturonase and polygalacturonase activity was performed as well as exo- vs. endo-enzyme activity in the crude pectinase enzyme-complex of the mentioned strains. Furthermore, a plate diffusion assay was applied to determine the presence and action of proteases in the crude extracts. A. sojae ATCC 20235 with highest polymethylgalacturonase activity and highest polygalacturonase activity both exo- and endo-enzyme activity, is a promising candidate for industrial pectinase production, a group of enzymes with high commercial value, in solid-state fermentation processes. Beside the enzymatic assays a protein profile of each strain is given by SDS-PAGE electrophoresis and in addition species-specific zymograms for pectinolytic enzymes were observed, revealing the differences in protein pattern of the A. sojae strains to the re-classified A. oryzae.

KEYWORDS Aspergillus sojae, pectinase, polygalacturonase, solid-state fermentation

18 Chapter 2

2.1 Introduction The middle lamella and the primary cell wall of higher plants contain a complex heteropolysaccharide called pectin. These carbohydrate polymers support the cohesion of the other cell wall polysaccharides and proteins. Pectin is composed mainly of galacturonic acid residues [1]. Pectinases include a number of related enzymes involved in the breaking down of pectic substances. Therefore, they can cause plant tissue maceration, cell lysis, and modification of cell wall structures, allowing other depolymerases to further degrade their product of decomposition [2]. Pectinolytic enzymes are extensively used in the food industry, e.g. as processing aids, the largest industrial application being in fruit juice extraction and clarification. Break down of pectin reduces the viscosity of pectin-rich crude juice and thus increases juice flow and reduces the press-time. Pectinases are also involved in clarification of wine, oil extraction, removal of citric fruit peels, and degumming fibres [3-5]. Depolymerizing enzymes like polygalacturonases are distinguished according to their substrate preference, whether they have preference for poly[α(14)-D-methylgalacturonic acid] (pectin-like substrates), which are termed as PMG in this study or poly[α(14)-D-galacturonic acid] (pectic acid-like substrates), which are termed as PG [6]. Furthermore these enzymes are termed as exo- or endo-enzymes depending on the action pattern. Endo-PGs randomly attack the [14]-α-glycosidic linkages of the polysaccharide chain producing a number of galacturonic acid oligomers, while exo-PGs specifically hydrolyses at the non-reducing end of polygalacturonic acid. Commercial pectic enzymes used in food industry normally contain a mixture of enzymes that split pectic compounds; traditionally mixtures consist of PG, PL (pectin lyase) and PME (pectin methylesterase), and are associated with cellulytic, proteolytic and other species of enzymes apart from the main pectinases [7]. In some food processes, it is convenient to use only one type of pectinolytic enzymes, e.g. preparation of instant potato flakes and carrot juice for baby food requires the maceration, where vitamins, colour and aroma have to be preserved and for these applications preparations that mainly contain PG activity are preferred [8]. Filamentous fungi especially Aspergillus niger (A. niger), are the major producers of acidic pectic enzymes used in fruit juice industries and wine production [9, 10]. Products of A. niger as well as A. sojae and A. oryzae have obtained a GRAS (General Regarded As Safe) status, which has approved their use in the food industry. Usually pectolytic microorganisms produce a multiplicity of pectinolytic enzymes. The production of these enzymes is carried out in solid-state (SSF) and submerged fermentation (SmF).

19 Chapter 2

The utilization of SSF processes is interesting for pectinase production by fungi, because of its capability to grow in low water activity (aw), which is a dimensionless quantity used to estimate the amount of free water that is readily available for the microorganisms. This SSF process offers several potential advantages in comparison to SmF, e.g. higher product concentration, simpler fermentation technology, and reduced waste-water output [11]. SSF also holds a tremendous potential for enzyme production. In several comparative studies on fungal pectinases production in solid- state and submerged fermentation, SSF gave superior results compared to submerged conditions and the protease production was also extremely lower [12- 14]. Moreover, simple and economic agricultural by-products like wheat bran and orange peel could be utilized so as to provide both nutritional and physical support during solid substrate cultivation. Wheat bran, which is composed predominantly of non-starch carbohydrates like arabinoxylans or cellulose, starch and crude proteins, has been a preferred substrate for the production of pectinolytic enzymes [15]. Previous attempts to produce the pectinolytic enzyme polygalacturonase (PG) by A. sojae ATCC 20235 have included submerged and surface cultivation [16, 17]. These studies have demonstrated the presence of enzymes with exo-PG activity in the crude (or partially purified) fermentation broths. These studies have triggered an interest in understanding the potential of A. sojae for the production of PG, as well as an eagerness of knowledge on the characteristics and technical applications of pectinases. On the other hand, SSF has been successfully employed for the production of the pectinase-complex from other Aspergillus strains, notably A. niger [11, 13]. The aim of this study focused on the comparative evaluation of three Aspergillus species including A. niger IMI 91881, as a known producer, A. sojae ATCC 20235, which was reclassified as A. oryzae [18], A. sojae CBS 100928, and A. sojae IMI 191303 for the production of pectinolytic enzymes. This general screening was performed in order to identify potential pectinase producers. The focus of this work was the evaluation of pectinase production of exo- vs. endo-PG activity, as well as exo- vs. endo-polymethylgalacturonase (PMG) activity in the crude extract obtained in SSF of the mentioned strains. To the best of our knowledge, there is no available information in the open literature related to the production of both pectinases by the two A. sojae strains employed in this study.

20 Chapter 2

2.2 Materials and methods

2.2.1 Materials All chemicals were purchased from AppliChem GmbH (Darmstadt, Germany), except

FeSO4·7H2O, dithiothreitol (DTT) and polyvinylpyrrolidone 360 which were obtained from Carl Roth GmbH & Co. KG (Karlsruhe, Germany). Microbial substrates like wheat bran, orange peel, and molasses were obtained from local suppliers (Bremer Rolandmühle Erling GmbH & Co. KG, Bremen, Germany; Freeze-Dry Foods GmbH, Greven, Germany; Golden Sweet, Meckenheim, Germany). Substrates for detection of pectinolytic activities, e.g. pectin, polygalacturonic acid and polygalacturonic acid sodium salt were purchased from Sigma-Aldrich Chemie GmbH, Steinheim, Germany.

2.2.2 Microbial strains and propagation The Aspergillus strains used throughout this study were all purchased in lyophilized form from different culture collections and propagated on agar plates according to the specifications given by the culture collections. A. sojae ATCC 20235 was obtained from Procochem Inc (Teddington, United Kingdom), an international distributor of the American Type of Culture Collection (ATCC) in Europe. Yeast Malt Extract (YME) agar medium, containing malt extract (10 g/L), yeast extract (4 g/L), glucose (4 g/L) and agar (20 g/L), was used for the propagation of this culture. Plates were incubated at 30 °C until abundant sporulation (1 week). It has to be noted that A. sojae ATCC 20235, which is still deposited as A. sojae at the ATCC, did not meet the requirements to be classified as A. sojae on the basis of morphological parameters [19] and has been recently reclassified as A. oryzae based on the alpA restriction fragment length polymorphism (RFLP) [18]. A. sojae CBS 100928 was purchased from the Centraalbureau voor Schimmelcultures (CBS) (Utrecht, Netherlands). The propagation of this culture was done on Malt Extract Agar (MEA) plate medium containing, malt extract (20 g/L), peptone (1 g/L), glucose (20 g/L) and agar (15 g/L), and incubated at room temperature (22°C) in darkness until sporulation (1 week). A. sojae IMI 191303 and A. niger IMI 91881 were obtained from CABI Bioscience / United Kingdom National Culture Collection (UKNCC) (Egham, United Kingdom). The propagation of these cultures was done on Czapek agar plate medium containing per liter: sucrose (30 g), K2HPO4 (1 g), agar (15 g), and Capek concentrate

(10 mL) containing per 100 mL: NaNO3 (30 g), KCl (0.5 g), MgSO4·7H2O (0.5 g) and

FeSO4·7H2O (0.01 g). The cultures were incubated at room temperature for 1 week until sporulation.

21 Chapter 2

Stock cultures from spores of all mentioned strains were preserved in 20% glycerol and stored at -80°C.

2.2.3 Inoculum A pre-activation step was performed on agar plates, containing the mentioned medium (see section 2.2.2)‎ for propagation for each culture, using the stock cultures. Spores from these plates were used as inoculum for molasses slants. The spores for inoculation of the main culture were extracted from molasses agar slants containing: glycerol (45 g/L), molasses (45 g/L), peptone (18 g/L), NaCl (5 g/L), KCl

(0.5 g/L), FeSO4·7H2O (15 mg/L), KH2PO4 (60 mg/L), MgSO4 (50 mg/L), CuSO4·5H2O

(12 mg/L), MnSO4·H2O (15 mg/L) and agar (20 g/L). Spores were harvested from the slants using 5 mL of 0.02% (w/v) Tween 80 and counted microscopically.

2.2.4 Culture medium and growth conditions Erlenmeyer flasks (250 mL) containing 10 g of wheat bran and fine-particle granulate of dried orange peel in the ratio 70:30, wetted at 120% with 0.2 N HCl solution (sterilized at 121 °C for 20 min) were inoculated with the spore suspensions and incubated at room temperature for 1 to 6 days. Each Erlenmeyer flask (250 mL) was inoculated with the total number of 2 × 107 spores.

2.2.5 Enzyme leaching At each 24 h interval of cultivation, the enzyme recovery was obtained by adding 100 mL distilled water into the Erlenmeyer flask of each strain and mixed in an incubator shaker (Innova 4230, New Brunswick Scientific) at 350 rpm, 30 °C, for 30 min. The mycelium and solid medium were separated by filtration through cheese cloth and the filtrate was centrifuged at 4 °C, 3200×g, for 20 min. Enzyme activities, pH, soluble carbohydrate from crude extract and total protein concentration were determined in the supernatant.

2.2.6 Analytical methods

2.2.6.1 Total protein determination Total extracellular protein was measured according to the modified Bradford´s method, using Coomassie PlusTM Protein Assay Kit (Pierce, Fischer scientific, Schwerte, Germany). The assay was performed in a microplate by determining the absorbance at 595 nm using bovine serum albumin (BSA) as a standard. Determinations were performed in duplicate.

22 Chapter 2

2.2.6.2 Exo-pectinolytic activity measurement

2.2.6.2.1 Polymethylgalacturonase assay PMG activity was determined by measuring the release of reducing groups from citrus pectin using a modification of the DNS method [20]. Galacturonic acid was used as standard. The reaction mixture contained an appropriate dilution of the supernatant along with 0.5% (w/v) pectin dissolved in 0.1 M acetate buffer (pH 5.0), which was incubated at 45°C for 10 min according to Blandino et al. [21]. One unit of exo-enzyme activity was defined as the amount of enzyme that catalyses the release of 1 µmol of galacturonic acid per unit volume of supernatant per unit time at standard assay conditions. Enzyme activity was expressed as unit per gram dry substrate (U/g). Determinations were performed in duplicate.

2.2.6.2.2 Polygalacturonase assay PG activity was assayed according to the Nelson-Somogyi method adapted by Panda et al. [22], using 2.4 g/L polygalacturonic acid as substrate at pH 4.0. Crude extract (0.086 mL) containing enzyme was added to 0.4 mL substrate and incubated for 20 min at 26 °C. The corresponding galacturonic acid content was determined from the standard galacturonic calibration curve. As to the assay conditions mentioned above, one unit of exo-enzyme activity was defined as the amount of enzyme that catalyses the release of 1 µmol of galacturonic acid per unit volume of supernatant per unit time at standard assay conditions. Enzyme activity was expressed as unit per gram dry substrate (U/g). Determinations were performed in duplicate.

2.2.6.3 Endo-pectinolytic activity measurement Endo-enzyme activity was determined by measuring the decrease in viscosity of a substrate solution, either 2% (w/v) pectin for endo-PMG or 3.2% (w/v) polygalacturonic acid (sodium salt) for endo-PG. Pectinolytic activity was assayed by adding 0.2 mL of a 1/20 dilution of crude extract containing the enzyme, to 0.2 mL of 0.2 M acetate buffer (pH 5.0) and 1.6 mL substrate. The mixture was incubated in a water bath for 1 h at 40°C with shaking. After incubation, the viscosity of the samples was determined. The later was done indirectly by measuring the time required for the reaction mixture to elute through a 1.0 mL glass pipette. Samples were measured in triplicate. Viscosity was calculated from time measurements to pass the standard solution through the pipette according to a calibration curve obtained utilizing PVP 360 standard solutions at 25 C [23]. Controls for non- enzymatically treated substrate solutions were included, utilizing inactivated

23 Chapter 2 samples. One unit of Endo-PG activity was defined as the quantity of enzyme which caused a 50% reduction in viscosity of the reaction mixture per minute, under the conditions of the assay [21].

2.2.6.4 Plate assay for proteolytic activity Dual-substrate assay plates were prepared according to the procedure given by Montville [24], containing 1% (w/v) casein and 1% (w/v) gelatin as substrates. Wells of 5 mm diameter were cut into the solid media and filled with 30 µL crude extract or commercial protease solution (as control). After 24 h incubation at 30 °C the diameters of the zones formed were measured. Zone diameters (D) were converted to log10 adjusted zone area by the following expression:

2 2 log10 adjusted zone area = log10 [(D/2) π – (5.0/2) π] (2.1)

Proteolytic activity was reported in this manner and referred to as zone area (log10 mm2). Determinations were performed in duplicate.

2.2.6.5 Total soluble carbohydrate assay Soluble carbohydrates in crude extract were determined by the phenol–sulfuric acid method according to Dubois et al. [25], using D-glucose as standard. For the assay, 1 mL of 5% (w/v) phenol solution and 5 mL 96% H2SO4 were added to 1 mL of an appropriate dilution of the supernatant. Samples were incubated at room temperature for 20 min. The absorbance of each sample was spectrophotometrically determined at 490 nm. Soluble carbohydrate content was expressed as mg per gram substrate.

2.2.6.6 Protein fractionation

2.2.6.6.1 One-dimensional electrophoresis The supernatant of the crude extract obtained by SSF was first dialyzed over night at 4 °C, using SnakeSkin® pleated dialysis tubing, 3,500 MWCO (Thermo Scientific, Rockford, USA). Samples were centrifuged at 4 °C, 6000×g, for 20 min and concentrated 15 times, using a freeze-dryer. SDS-PAGE was performed according to Laemmli [26]. Briefly: 12.5% SDS-PAGE gels with an approximately 2 cm stacking buffer zone were cast and samples run in constant current mode at 20 mA/gel, at 4 °C. Samples were mixed with sample buffer in the ratio 2:1. Sample load add up to 10 µL per lane. Protein bands were visualized, using colloidal Coomassie (G-250) staining [27].

24 Chapter 2

2.2.6.6.2 Native polyacrylamide gel electrophoresis / zymogram Native PAGE was performed by excluding SDS and DTT from the electrophoresis protocol described above. The “sandwich” method was used to detect the activity of pectinases acting on polygalacturonic acid as substrate [28]. Briefly, proteins were separated and subsequently the gel was a) incubated for 20 min in 0.1 M citrate phosphate buffer (pH 5) and b) contacted with an (solid) agar substrate containing 0.25% (w/v) polygalacturonic acid for 80 min at 30°C (80% humidity chamber). The polygalacturonic acid agar plate was treated with 1% (w/v) cetyltrimethylammonium bromide which revealed pectinases activity as translucent bands on an opaque background.

2.3 Results and Discussion

2.3.1 Culture profiles on solid substrates Solid substrate cultivation experiments were performed to evaluate the production of pectinases by several fungal strains which belong to the genus Aspergillus. The cultivation procedure for the different strains was performed under the same conditions, utilizing wheat bran moistened with a hydrochloric acid solution [29]. Grinded (dehydrated) orange peel was added to the solid substrate as an inducer for the production of pectinolytic enzymes [30]. Thus, the opportunity for a direct comparison between different Aspergillus strains in terms of pectinase production was set. Information of this kind is seldom available in the literature since most comparisons of pectinase activity between several strains are done under very different cultivation conditions and using different analytical assays to quantify the enzymatic activity [10, 13]. Experiments were carried out in duplicate and the corresponding standard variation was below 15 %.

25 Chapter 2

2.3.1.1 Cultivation of Aspergillus niger IMI 91881

10 350 35 9 300 30 8 7 250 25

6 200 20 5 15

4 150 pH (-) pH 3 100 10 2

Soluble protein (mg/g) protein Soluble 5 50 exo-PMGactivity (U/g)

1 (mg/g) carbohydrate Soluble 0 0 0 0 1 2 3 4 5 6

Time (d) Figure 2‎ .1 Cultivation profile of A. niger IMI 91881: ( ) exo-PMG activity, ( ) soluble carbohydrate content, ( ) soluble protein content and ( ) pH in the crude extract.

The cultivation profile of A. niger IMI 91881 is shown in Figure 2.1‎ . From the mentioned graph, the total soluble carbohydrate content, the total soluble protein content, the pH, and the PMG activity of the crude extract obtained from the fermented mass can be observed. The SSF process was run for six days. In this case, the peak of PMG activity production ( 23.5 U/g) was observed on the 4th day of cultivation. Hence, a space-time-yield of PMG production of 5.9 U/g per day can be calculated. The total protein content decreased at the beginning of the fermentation but from the 3rd day of cultivation the observed protein secretion was higher than the degradation of soluble proteins. The increasing protein concentration could indicate the production of extracellular enzymes. During the course of SSF, soluble carbohydrate were degraded from 320 mg/g to 60 mg/g indicating extensive utilization of the carbon and energy (C/E) source available for biomass and product synthesis. The pH in the crude extract evolved from 4.1 to 4.6 thus, confirming the appropriate selection of the acidic conditions to favor mold growth while avoiding bacterial contamination of the system. A. niger is well known for its ability to produce pectinases in both solid substrate and submerged fermentation systems [31]. Pectinolytic enzymes from this species are industrially produced [32].

26 Chapter 2

2.3.1.2 Cultivation of Aspergillus sojae ATCC 20235

10 350 35 9 300 30 8 7 250 25

6 200 20 5 15 4 150 pH (-) pH 3 100 10 2 5

50 exo-PMGactivity (U/g) Soluble protein (mg/g) (mg/g) protein Soluble 1 (mg/g) carbohydrate Soluble 0 0 0 0 1 2 3 4 5 6

Time (d) Figure 2‎ .2 Cultivation profile of A. sojae ATCC 20235: ( ) exo-PMG activity, ( ) soluble carbohydrate content, ( ) soluble protein content and ( ) pH in crude extract.

Figure 2.2‎ depicts the process parameters measured during the solid substrate cultivation of A. sojae ATCC 20235. This strain has been extensively studied before, particularly in relation to its ability to produce enzymes with exo-polygalacturonase activity in submerged and surface cultivation [16, 17, 33]. When subjected to fermentation on solid substrates in the absence of free water, A. sojae ATCC 20235 produced a maximum PMG activity 33.4 U/g at the 5th day of cultivation. Thus, the calculated space-time yield for PMG production amounts to 6.7 U/g per day. While PMG production was negligible during the early cultivation stage, it was observed that between the 2nd and 5th day of cultivation the PMG activity sharply increased. A similar trend was observed in the total protein content of the extracts; a higher secretion than degradation of soluble protein was noticed from the 2nd day of fermentation onwards. This could indicate that A. sojae ATCC 20235 is a good protein exporter under SSF conditions. Total soluble carbohydrate levels decreased abruptly after the 2nd day of cultivation indicating utilization for biomass and protein production. The pH of the extracts only increased slightly during the time course of the cultivation process.

27 Chapter 2

2.3.1.3 Cultivation of Aspergillus sojae IMI 191303

10 350 35 9 300 30 8 7 250 25

6 200 20 5 15

4 150 pH (-) pH 3 100 10

2 exo-PMG activity (U/g) activity exo-PMG

Soluble protein (mg/g) protein Soluble 50 5

1 (mg/g) carbohydrate Soluble 0 0 0 0 1 2 3 4 5 6

Time (d) Figure 2‎ .3 Cultivation profile of A. sojae IMI 191303: ( ) exo-PMG activity, ( ) soluble carbohydrate content, ( ) soluble protein content and ( ) pH in crude extract.

The fermentation profile of A. sojae IMI 191303 on solid substrate is presented in Figure 2.3‎ . It can be observed that the peak of exo-PMG production (19.2 U/g) is reached on the 4th day of cultivation. This translates into a space-time yield of 4.8 U/g per day. While pectinolytic activity decreased after the 4th day of fermentation, the degradation of carbohydrates stagnated after that fermentation time, too. This would suggest that the product synthesis has been limited by the unavailability of the C/E source. Total protein was also degraded during the first 3 days of cultivation and remained at a relatively low level afterwards (5 mg/g). The pH of the extracts varied only slightly and remained within acidic range, as observed in the case of the other strains mentioned before. Previous studies on pectinase production by A. sojae were done in the 1970ies and focused on the enzyme activity of pectin transeliminase [34]. Recent publications about the strain A. sojae explore the production of recombinant proteins [18]. To the knowledge of the authors no other publications about pectinase production by A. sojae of other research groups are published.

28 Chapter 2

2.3.1.4 Cultivation of Aspergillus sojae CBS 100928

10 350 35 9 300 30 8 7 250 25

6 200 20 5 15

4 150 pH (-) pH 3 100 10

2 exo-PMG activity (U/g) activity exo-PMG

Soluble protein (mg/g) protein Soluble 50 5

1 (mg/g) carbohydrate Soluble 0 0 0 0 1 2 3 4 5 6

Time (d) Figure 2‎ .4 Cultivation profile of A. sojae CBS 100928: ( ) exo-PMG activity, ( ) soluble carbohydrate content, ( ) soluble protein content and ( ) pH in crude extract.

The cultivation profile of A. sojae CBS 100928 is shown in Figure 2.4‎ . The production of enzymes with exo-PMG activity can be observed after the 2nd day of fermentation. However, maximum values are reached at the 3rd day (14 U/g) and the 5th day (16.6 U/g). This has coincided with a moderate increase in total protein concentration which was noticed after 4 days of cultivation. The presence of a plateau at the total protein concentration between 2 and 4 days incubation time – or eventually a bimodal activity profile – may indicate the presence of various types of enzymes presenting distinct synthesis kinetics. Considering the values found at the 5th day of cultivation, a space-time yield of 3.3 U/g per day for PMG production can be calculated. Comparing the two A. sojae strains a higher exo-PMG activity was observed by A. sojae IMI 191303 under the growth conditions used in this study. Total soluble carbohydrate content decreased gradually, particularly after the 3rd day of cultivation. The pH values measured in the extracts was comparable to the other observed cultivations reported here. As reported by Sardjono et al. [35] A. sojae and A. oryzae are both producers of microbial proteases, this aspect is further explored in the next sections.

2.3.2 Enzymatic activities of the crude extracts The depolymerizing hydrolases of the pectinase group are distinguished according to their substrate preference and can be further subdivided depending on the action pattern as mentioned above. Usually the combined action of pectinolytic

29 Chapter 2 enzymes in combination with cellulytic, proteolytic and other species of enzymes is used in food industries. Following results reveal an overview of some enzymatic activities present in the crude extracts of the studied Aspergillus species.

2.3.2.1 Exo-PMG and endo-PG enzymatic activities Table 2.1‎ compares the types of enzymatic activity found in the crude extracts obtained by harvesting the fermented mass of the various strains studied at the 5th day of cultivation. This time was fixed to standardize the cultivation conditions for a better comparison of all strains. At the 5th day of cultivation the peak of pectinase production or a plateau of enzyme production was observed. Hence, it was possible to assure high enzyme activity for all strains. It can be observed that all the cultures were able to produce enzymes with reducing-sugar liberation capability on both pectin (exo-PMG) and polygalacturonic acid (exo-PG). According to the exo-PMG potency, the mentioned extracts can be ordered as follows:

A. sojae ATCC 20235 > A. niger IMI 91881 > A. sojae IMI 191303 > A. sojae CBS 100928

While according to their exo-PG activity, the extracts can be ordered as follows:

A. sojae ATCC 20235 > A. sojae CBS 100928 > A. niger IMI 91881 > A. sojae IMI 191303

Interestingly, the total amount of protein found in the corresponding extracts follows a similar pattern, indicating that the total reducing-sugar liberating activity may be linked to the protein exportation capability of the strains under study. Moreover, the exo-PMG specific activity for A. sojae ATCC 20235 was higher (3.9 U/mg) than the observed for the other species, which may indicate that pectinases are a main component of the secreted proteome of the mentioned microorganism.

30 Chapter 2

Table 2‎ .1 Enzyme activity at the 5th day of cultivation.

PMG activity (U/g) PG activity (U/g) Specific activity Protease Total protein Microorganism Ratio Ratio (U/mg) activity exo endo exo endo (mg/g) 2 exo-PMG (log10 mm ) A. niger IMI 91881 21.7 14.3 1.52 19.7 2.3 8.57 5.7 3.8 1.5

A. sojae ATCC 20235 33.4 32.9 1.02 18.3 30.1 0.94 8.6 3.9 2.4

A. sojae IMI 191303 16.7 4.8 3.48 14.9 18.8 0.79 4.6 3.6 1.8

A. sojae CBS 100928 16.6 28.8 0.57 23.0 23.0 1.00 5.2 3.2 2.1

31 Chapter 2

The results presented in Table 2.1‎ confirm previous findings by Ustok et al. [17] who were able to produce exo-PG by A. sojae ATCC 20235 via surface cultivation on a complex media containing crushed maize, maize meal, corncob, and molasses broth. Since this strain was re-classified as A. oryzae, these results confirm previous findings by Malvessi and da Silveira [36]; they studied the production of exo-PG activity and endo-PMG activity by A. oryzae CCT3940 in a complex media based on wheat bran, salts, and pectin as an inducer in SmF. Utilizing this media resulted in a maximum exo-PG activity of 45 U/mL after 64 h and a maximum endo-PMG activity of 159 U/mL after 83 h of cultivation at 28°C and 300 rpm. Furthermore, they observed that acidic initial pH values close to 4 favored both mycelia development and enzyme production. Comparing the endo-pectinase activity to different strains of A. niger resulted in higher activities for A. oryzae CCT3940. On the other hand, Galiotou-Panayotou et al. [37] produced a maximum exo-PG activity of 14.5 U/mL by A. niger NRRL-364 (also known as A. niger IMI 91881) in submerged fermentation with a space-time yield of 4.8 U/mL per day. Employing this strain in SSF, an exo-PG activity of 19.7 U/g was observed after 5 days. According to the results of this study A. niger produced an extract characterized by the predominant presence of exo- pectinase activity, which implies the degradation of pectic substances into mono- or digalacturonic acid.

2.3.2.2 Endo-PMG and endo-PG enzymatic activities The crude extracts obtained from the SSF processes employing various Aspergillus strains were assayed for endo-PMG and endo-PG activity. In this study, the ability of the enzymes present in the crude extract to reduce the viscosity of substrate solutions (2% pectin and 3.2% polygalacturonate, respectively) was observed. According to the results listed in Table 2.1‎ , the endo-enzymatic potency of the extracts can be ordered as follows:

Endo-PMG: A. sojae ATCC 20235 > A. sojae CBS 100928 > A. niger IMI 91881 > A. sojae IMI 191303

Endo-PG: A. sojae ATCC 20235 > A. sojae CBS 100928 > A. sojae IMI 191303 > A. niger IMI 91881

The highest endo-pectinase activities were obtained by A. sojae ATCC 20235. Using a 1/20 dilution of this crude extract resulted in a viscosity reduction of 90.3% for pectate and 98.6% for pectin under the conditions mentioned above. In comparison to this, the commercial pectinase of A. niger (Sigma) had to be diluted 5000 times to reduce the viscosity of polygalacturonic acid by 90.4% and of pectin by 96.2% under

32 Chapter 2 the same conditions. To better understand the characteristics of the crude extracts the ratio PMG-exo-to-endo and the ratio PG-exo-to-endo were calculated. A. sojae ATCC 20235 showed a very well balanced ratio in both cases (~ 1), showing that the enzymes produced by this microorganism have a broad substrate action pattern. Besides that, the mentioned strain also provided the best production levels of pectinolytic enzymes, which promise an extensive degradation of pectic substances (Table 2‎ .1). A. sojae CBS 100928 demonstrated to be a good producer of endo- pectinases. However, the PMG ratio for this strain was 0.57 while the PG ratio was 1. This indicates a preferential endo-PMG enzymatic activity. On the contrary, A. sojae IMI 191303 showed a remarkable PMG-ratio of 3.48, indicating a significantly higher exo-enzymatic activity, while on consideration of the PG-ratio (0.79) the endo- enzymatic activity is dominant. A. niger IMI 91881, as mentioned before, presented a marked exo-PMG activity (ratio 1.52) and extraordinary exo-PG activity (ratio 8.57) due to the low endo-enzymatic activity when acting on pectic substrates. Summarizing, it was shown that A. sojae ATCC 20235 produced a balanced pectinase-complex, while A. niger IMI 91881 produced preferential exo-pectinases and low endo-PG activity. A. sojae CBS 100928 delivered a predominant endo-PMG profile compared to the endo-PG activity, and A. sojae IMI 191303 delivered higher endo-enzymatic activity on polygalacturonate than on pectin. However, since the effect on viscosity is dependent on location of the glycosidic bonds, and the hydrolysis of a glycosidic bond near the middle of the polymer chain has far more effect on viscosity reduction than hydrolysis near the end of the chain, the endo-enzymatic activities are the crucial factor for applications of viscosity reduction. Therefore endo-PGs have especially been applied for maceration [8]. Nevertheless, there is no direct correlation between viscosity reduction and number of glycosidic bonds hydrolyzed. Furthermore, Baker and Bruemmer [38] demonstrated the effect of endo-depolymerizing enzymes to stabilize cloud in orange juice, at which the important fact was the ratio of high endo-PG activity to low endo-PMG activity. Concerning this, A. sojae IMI 191303 produced a preferable extract for such applications.

2.3.2.3 Proteolytic activity The crude extracts were tested for total proteolytic activity. The presence of proteases in the extracts may render positive or negative consequences, depending upon the final application thought. As mentioned before, commercial pectinase mixtures also contain proteolytic enzymes, which is similar to the markets for microbial proteinases for the baking and leather industries, i.e. usually a processing aid in a proprietary blend with other ingredients is purchased, not just an enzyme

33 Chapter 2

[39]. For example, mixtures with proteinases from A. oryzae are used for leather bating and blends containing fungal acidic proteinases are predominantly used for modification of high gluten doughs and for cracker and biscuit production in the bakery. On the other hand, protease-deficient strains are of use for the production of secreted recombinant proteins [18]. According to Table 2.1‎ , the proteolytic potency of the extracts can be ordered as follows:

A. sojae ATCC 20235 > A. sojae CBS 100928 > A. sojae IMI 191303 > A. niger IMI 91881

A. niger IMI 91881 produced the lowest proteolytic activity, which could be an advantage regarding the stability of the enzyme concentrate e.g. protease attack on pectinases is less likely to occur or may occur in a lesser extent. The other three strains, however, showed a tendency to produce higher levels of proteases. A. sojae ATCC 20235 (re-classified as A. oryzae) appeared to be the strain producing a larger amount of extracellular protein, as well as, the higher levels of the enzyme types explored in this work. Therefore, an application of the SSF-derived extract to cases were extensive action on complex structures is required can be envisioned for this strain. Oda et al. [40] demonstrated the ability of A. oryzae to secrete large amounts of a wide range of different enzymes into its environment and a higher extracellular protein production in SSF than under submerged culture conditions for this species.

2.3.3 Protein fractionation studies

2.3.3.1 Fractionation by SDS-PAGE electrophoresis Samples of the crude extracts obtained by extraction of the solid substrate cultures were analyzed by polyacrylamide-gel electrophoresis under denaturing conditions. Results are presented in Figure 2.5‎ . The observed protein patterns differed in complexity and the number of bands present in each species. Analysis of SDS-PAGE of whole-cell protein profiles has turned out as a useful tool for classification and identification of bacteria and fungal species [41-43]. According to the crude extracts obtained in this study there is an appropriate clear difference in the protein pattern of the A. niger strain, the A. sojae strain which was reclassified as A. oryzae and the two A. sojae strains. The extracts of A. sojae IMI 191303 and A. sojae CBS 100928 produced similar profiles. The profile of the commercial pectinase solution (SIGMA) presented three remarkable bands in the range 35 – 44 kDa and a number of thin bands with molecular masses in the range 50 – 120 kDa. Purifying pectinolytic enzymes from a commercial enzyme preparation derived from A. niger (Pectinase K2B 078, Rapidase) Kester and Visser [32] isolated two most abundant endo-PGs with molecular masses of 55 and 38 kDa.

34 Chapter 2

The molecular weights of some members of the pectinases complex produced by A. niger in SSF have been determined [44]: PG (36 kDa), PME (42 kDa), PeL I (pectate lyase) (42 kDa), PeL II (30 kDa), PeL III (36 kDa).

A B C D E F

250 kD 150 kD

100 kD

75 kD

50 kD

37 kD

25 kD

20 kD

Figure 2‎ .5 Analysis of crude extracts from the 5th day of SSF by SDS-PAGE: lane A: dual color prestained precision plus protein standards (BIO-RAD); lane B: commercial pectinase from A. niger (Sigma) diluted 1:3; lane C: A. niger IMI 91881; lane D: A. sojae ATCC 20235; lane E: A. sojae IMI 191303; lane F: A. sojae CBS 100928.

2.3.3.2 Native electrophoresis and zymogram The protein content of the extracts was also fractionated by native electrophoresis so as to preserve enzyme function. Results are shown in Figure 2.6‎ -1. The polyacrylamide gel slab was subsequently contacted with a solid media surface that contained polygalacturonate as substrate. Figure 2.6‎ -2 depicts the results obtained after revealing the substrate plate chemically. It can be observed that the different Aspergillus species developed a distinctive pattern. Nasuno [45] differentiated the species A. sojae and A. oryzae on the basis of species-specific mobility of alkaline proteinases in polyacrylamide gel disc electrophoresis under non-denaturing conditions. It has to be recalled that the active fractions (bands or zones) observed

35 Chapter 2 in this case are not directly related to the molecular mass of the enzymes since separation in the native gel is related to charge / size. This makes the information gathered in Figure 2.6‎ (native PAGE) difficult to compare with the profiles observed in Figure 2.5‎ (SDS-PAGE). Nevertheless, it can be clearly observed that A. niger presented a single active component, as is the case for A. sojae ATCC 20235, although both bands radically differed in electrophoretic mobility. While the commercial pectinase mixture (SIGMA) depicts at least 4 active zones, whereof one of them is outstanding. A. sojae IMI 191303 and A. sojae CBS 100928 presented noteworthy similar profiles: two main active zones well differentiated one from the other in terms of migration distances. Hence, similar to the SDS-PAGE profiles, the obtained zymograms for pectinolytic enzymes of the fungi were characteristic for each species like a kind of fingerprint to distinguish peculiar characteristics in the extracts of the used Aspergillus species. Nealson and Graber [46] already distinguished species in the genus Aspergillus by means of esterase and phosphatase zymograms. Further studies are required to fully understand the nature and properties of the A. sojae extracellular enzymes.

A B C D E F A B C D E F

1 2

Figure 2‎ .6 Native PAGE for enzyme detection on the electrophoretic gel. (1) Substrate-containing agar plate after precipitation. (2) Electriphorized gel; lane A: commercial pectinase from A. niger (SIGMA) diluted 1:5; lane B: A. niger IMI 91881; lane C: A. sojae ATCC 20235; lane D: A. sojae IMI 191303; lane E: A. sojae CBS 100928; lane F: commercial pectinase from A. niger (SIGMA) diluted 1:6.

2.4 Conclusions This work demonstrated the use of three Aspergillus species for pectinolytic enzyme production in SSF. All strains produced pectinases with the highest yield reached

36 Chapter 2 between the fourth and fifth day of cultivation. Two new A. sojae strains were identified to express enzymes of this group. The zymogram for pectinolytic enzymes of that species presented two separated zones with activity towards polygalacturonic acid. Nevertheless, the highest exo-pectinolytic activity with 33.4 U/g PMG and 28.3 U/g PG, as well as the highest endo-enzyme activity of 32.9 U/g PMG and 30.1 U/g PG was observed by A. sojae ATCC 20235 (re-classified as A. oryzae). The high protein secretion in combination with the GRAS status of this strain seems to be promising with regard to enzyme production for industrial applications. Further optimization experiments using this strain for the production of pectinases may provide auspicious results in the future. Moreover, it was shown that the use of complex media in SSF also holds a great potential for the production of these enzymes. The yields might be greatly increased by manipulating some growth conditions. On the basis of this work further studies on the pectinase- complex obtained by A. sojae ATCC 20235 in SSF will follow.

Acknowledgements Financial support of Jacobs University Bremen gGmbH through the project PGSYS / ETB-2008-44 and Scientific and Technological Research Council of Turkey (TUBITAK) through the project 107O602 is gratefully acknowledged. Furthermore, the authors are indebted to Dr. S. Diercks-Horn of Jacobs University Bremen gGmbH for proof reading the article and helpful suggestions.

References [1] Willats, WGT, McCartney, L, Mackie, W, Knox, JP (2001). Pectin: cell biology for functional analysis. Plant Molecular Biology 47, (9-27). [2] Collmer, A, Ried, JL, Mount, MS (1988). Assay methods for pectic enzymes. In: W.A. Wood; S.T. Kellogg, eds. Methods in Enzymology. Academic Press: San Diego; Vol. 161, pp. 329-335. [3] Jayani, RS, Saxena, S, Gupta, R (2005). Microbial pectinolytic enzxmes: A review. Process Biochemistry 40, 2931-2944. [4] Mutlu, M, Sarioglu, K, Demir, N, Ercan, MT, Acar, J (1999). The use of commercial pectinase in fruit juice industry. Part I: viscosimetric determination of enzyme activity. J. Food Eng. 41, 147-150. [5] Silva, D, Da Silva Martins, E, Da Silva, R, Gomes, E (2002). Pectinase production by Penicillium viridicatum RFC3 by solid state fermentation using agricultural wastes and agro-industrial by-products. Brazilian Journal of Microbiology 33, (318-324). [6] Whitaker, JR (1984). Pectic substances, pectic enzymes and haze formation in fruit juices. Enzyme and Microbial Technology 6, 341-349.

37 Chapter 2

[7] Del Cañizo, AN, Hours, RA, Miranda, MV, Cascone, O (1994). Fractionation of fungal pectic enzymes by immobilized metal ion affinity chromatography. J. Sci. Food Agric. 64, 527-531. [8] Lang, C, Dörnenburg, H (2000). Perspectivies in the biological function and the technological application of polygalacturonases. Applied Micriobiology and Biotechnology 53, 366-375. [9] Kashyap, DR, Vohra, PK, Chopra, S, Tewari, R (2001). Applications of pectinases in the commercial sector: a review. Bioresource Technology 77, 215-227. [10] Naidu, GSN, Panda, T (1998). Production of pectolytic enzymes - a review. Bioprocess Eng. 19, 355-361. [11] Pandey, A, Soccol, CR, Mitchell, D (2000). New developments in solid state fermentation: I-bioprocesses and products. Process Biochemistry 35, (10) 1153-1169. [12] Díaz-Godínez, G, Soriano-Santos, J, Augur, C, Viniegra-González, G (2001). Exopectinases produced by Aspergillus niger in solid-state and submerged fermentation: a comparative study. J. Ind. Microbiol. Biot. 26, 271-275. [13] Favela-Torres, E, Volke-Sepúlveda, T, Viniegra-González, G (2006). Production of hydrolytic depolymerising pectinases. Food Technol. Biotechnol. 22, 221-227. [14] Patil, SR, Dayanand, A (2006). Optimization of process for the production of fungal pectinases from deseeded sunflower hesd in submerged and solid- state conditions. Bioresource Technology 97, 2340-2344. [15] Sun, X, Liu, Z, Qu, Y, Li, X (2008). The effects of wheat bran composition on the producion of biomass-hydrolyzing enzymes by Penicillium decumbens. Appl. Biochem. Biotechnol. 146, 119-128. [16] Tari, C, Gögus, N, Tokatli, F (2007). Optimization of biomass, pellet size and polygalacturonase production by Aspergillus sojae ATCC 20235 using response surface methodology. Enzyme and Microbial Technology 40, 1108- 1116. [17] Ustok, FI, Tari, C, Gogus, N (2007). Solid-state production of polygalacturonase by Aspergillus sojae ATCC 20235. Journal of Biotechnology 127, 322-334. [18] Heerikhuisen, M, Van den Hondel, C, Punt, P (2005). Aspergillus sojae. In: G. Gellissen, ed. Production of recombinant proteins. Novel microbial and eukaryotic expression systems. WILEY-VCH: Weinheim, pp. 191 - 214. [19] Ushijima, S, Hayashi, K, Murakami, H (1982). The current taxonomic status of Aspergillus sojae used in Shoyu fermentation. Agric Biol Chem 46, 2365- 2367. [20] Miller, GL (1959). Use of dinitrosalicylic acid reagent for determnation of reducing sugar. Analytical Chemistry 31, 426-428.

38 Chapter 2

[21] Blandino, A, Iqbalsyah, T, Pandiella, SS, Cantero, D, Webb, C (2002). Polygalacturonase production by Aspergillus awamori on wheat in solid-state fermentation. Appl. Microbiol. Biotechnol. 58, 164-169. [22] Panda, T, Naidu, GSN, Sinha, J (1999). Multipleresponse analysis of pectinolytic enzymes by Aspergillus niger: a statistical view. Process Biochemistry 35, 187-195. [23] Yeh, HM, Cheng, TW, Wu, HH (1998). Membrane ultrafiltration in hollow- fiber module with the consideration of pressure declination along the fibers. Separation and Purification Technology 13, 171-180. [24] Montville, TJ (1983). Dual-substrate plate diffusion assay for proteases. Applied Environmental Microbiology 45, 200-204. [25] Dubois, M, Gilles, KA, Hamilton, JK, Rebers, PA, Smith, F (1955). Colorimtric method for determination of sugars and related substances. Analytical Chemistry 28, 350-356. [26] Laemmli, UK (1970). Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680-685. [27] Neuhoff, V, Arnold, N, Taube, D, Ehrhardt, W (1988). Improved staining of proteins in polyacrylamide gels including isoelectric focusing gels with clear background at nanogram sensitivity using Coomassie Brillant Blue G-250 and R-250. Electrophoresis 9, 255-262. [28] Manchenko, GP (1994). Handbook of Detection of Enzymes on Electrophoresis Gels. CRC Press: Boca Raton, FL. [29] Fernandez-Lahore, MH, Gallego Duaigües, MV, Cascone, O, Fraile, ER (1997). Solid state production of a Mucor bacilliformis acid protease. Revista Argentina de Microbiologia 29, 1-6. [30] Patil, SR, Dayanand, A (2006). Exploration of regional agrowastes for the production of pectinase by Aspergillus niger. Food Technol. Biotechnol. 44, 289-292. [31] Minjares-Carranco, A, Trejo-Aguilar, BA, Aguilar, G, Viniegra-González, G (1997). Physiological comparison between pectinase-producing mutants of Aspergillus niger adapted either to solid-state fermentation or submerged fermentation. Enzyme and Microbial Technology 21, 25-31. [32] Kester, HC, Visser, J (1990). Purification and characterization of polygalacturonases produced by the hyphal fungus Aspergillus niger. Biotechnology and Applied Biochemistry 12, (2) 150-160. [33] Gögus, N, Tari, C, Oncü, S, Unluturk, S, Tokatli, F (2006). Relationship between morphology, rheology and polygalacturonase production by Aspergillus sojae ATCC 20235 in submerged cultures. Biochemical Engineering Journal 32, 171-178. [34] Ishii, S, Yokotsuka, T (1972). Purification and properties of pectin trans- eliminase from Aspergillus sojae. Agric. Biol. Chem. 36, 146-153.

39 Chapter 2

[35] Sardjono, Zhu, Y, Knol, W (1998). Comparison of fermentation profiles between lupine and soybean by Aspergillus oryzae and Aspergillus sojae in solid-state culture systems. J. Agr. Food Chem. 46, 3376-3380. [36] Malvessi, E, da Silveira, MM (2004). Influence of medium composition and pH on the production of polygalacturonases by Aspergillus oryzae. Brazilan Archives of Biology and Technology 47, (5) 693-702. [37] Galiotou-Panayotou, M, Rodis, P, Kapantai, M (1993). Enhanced polygalacturonase production by Aspergillus niger NRRL-364 grown on supplemented citrus pectin. Letters in Applied Microbiology 17, 145-148. [38] Baker, RA, Bruemmer, JH (1972). Pectinase stabilization of orange juice cloud. J. Agr. Food Chem. 20, (6) 1169-1173. [39] Outtrup, H, Boyce, COL (1990). Microbial proteinases and biotechnology. In: W. Fogarty; M.C.T. Kelly, eds. Microbial enzymes and biotechnology. Elsevier Science Publishers LTD, England, pp. 227-274. [40] Oda, K, Kakizono, D, Yamada, O, Iefuji, H, Akita, O, Iwashita, K (2006). Proteomic analysis of extracellular proteins from Aspergillus oryzae grown under submerged and solid-state culture conditions. Applied Environmental Microbiology 72, (5) 3448-3457. [41] Bent, KJ (1967). Electrophoresis of proteins of 3 Penicillium species on acrylamide gels. J. gen. Microbiol. 29, 195-200. [42] Merquior, VLC, Peralta, JM, Facklam, RR, Teixeira, LM (1994). Analysis of electrophoretic whole-cell protein profiles as a tool for characterization of Enterococcus species. Curr. Microbiol. 28, 149-153. [43] Vancanneyt, M, Van Lerberge, E, Berny, J-F, Hennebert, GL, Kersters, K (1992). The application of whole-cell protein electrophoresis for the classification and identification of basidiomycetous yeast species. Antonie van Leeuwenhoek 61, (1) 69-78. [44] Dinu, D, Nechifor, MT, Stoian, G, Costache, M, Dinischiotu, A (2007). Enzymes with new biochemical properties in the pectinolytic complex produced by Aspergillus niger MIUG 16. Journal of Biotechnology 131, (2) 128- 137. [45] Nasuno, S (1972). Differentiation of Aspergillus sojae from Aspergillus oryzae by polyacrylamide gel disc electrophoresis. Journal of General Microbiology 71, 29-33. [46] Nealson, K, Graber, E (1967). An electrophoretic survey of esterases, phosphatases, and leucine aminopeptidases in mycelia extracts of species Aspergillus. Mycologia 59, (2) 330-336.

40 Chapter 3

Chapter 3.

Statistical media design and SSF process optimization for improved PG production by Aspergillus sojae

Abstract

Previously identified pectinase producers of Aspergillus sojae were used for optimization of polygalacturonase production in solid-state fermentation by utilization of Design of Experiment (DoE). Several fermentation parameters were studied by applying a multiple step experimental setup of screening and optimization. The effects of media composition and several process parameters, like inoculum size, moisture level, incubation time and temperature on polygalacturonase activity were studied. Utilization of agricultural and agro- industrial by-products provided the establishment of an economical process for enzyme production. A comparison of two fungal strains resulted in similar optimal conditions for maximal polygalacturonase production by A. sojae ATCC 20235 and A. sojae CBS 100928. Highest enzyme yield (909.5 ± 2.7 U/g) was obtained by A. sojae ATCC 20235 under optimized conditions after 8 days at 30 °C applying 30% sugar beet pulp as inducer substrate in combination with wheat bran as medium wetted at 160% by 0.2 M HCl. This optimization produced 10.9 times increased polygalacturonase yield. Comparing both strains under optimized conditions between each other resulted in a 6.9 times increased polygalacturonase production by A. sojae ATCC 20235. Furthermore, the effect of changing solid-state fermentation parameters on protein pattern was presented and hence the influence of culture conditions on enzyme production. An overview of pectinolytic activities present in the extracts of both strains obtained under the same cultivation conditions is also given in this chapter. High fungal pectinolytic enzyme production by A. sojae in combination with the cost-efficient substrate indicated the high potential of the optimized process for large scale production and industrial application.

41 Chapter 3

3.1 Introduction The biotechnological potential of pectinolytic enzymes is well known due to their various industrial applications wherever degradation of pectic substances is required. This includes food related processes like fruit juice clarification, tissue maceration, wine clarification, coffee and tea fermentation and many others [1]. Polygalacturonases (PG) are hydrolases of the pectinolytic enzyme group, which catalyze the hydrolytic release of D-galacturonate from polygalacturonic acid [2]. Filamentous fungi are known to produce pectinolytic enzymes either via submerged fermentation (SmF) or solid-state fermentation (SSF). They are capable of synthesizing and secreting large quantities of certain proteins into the extracellular medium. Species of the genera Aspergillus are good producers of pectinases [3]. SSF permits the utilization of natural agricultural products, as well as agro-industrial residues and by-products as low cost substrates for microbial enzyme production [4]. Degradation and utilization of diverse biopolymers such as starch, cellulose or pectin enables cultivation of Aspergillus species on agricultural and agro-industrial residues. The application of agricultural and agro-industrial by-products, such as apple pomace derived from processing of apples in fruit juice industries or sugar beet pulp as residue after the sugar has been extracted, offers a wide range of alternative substrates and helps to solve disposal problems of these by-products. The apple pomace is mainly composed of insoluble carbohydrates such as cellulose, hemicelluloses and lignin. As solid by-product from processing apples it is a heterogeneous mixture consisting of peel, core, seed, calyx, stem and soft tissue, with a high water content [5]. Traditionally, apple pomace and citrus peels are used as raw materials for pectin production. Alternatively, high content in pectins (20-25%), its availability and low cost make sugar-beet pulp a potential source of pectins [6]. Sugar beet pulp is valuable as a fiber and energy source after the sugar has been extracted. It is mainly composed of (% on dry basis) pectin, 28.7; cellulose, 20; hemicellulose, 17.5; protein, 9.0; and lignin 4.4 [3]. Since the pectin content of sugar beet pulp or apple pomace is high it can be used for the microbial production of pectinolytic enzymes without adding pectinaceous material as inducer. The most commonly used agro-industrial residue in biotechnological processes is wheat bran [3]. Wheat bran is a by-product which accumulates during flour production. It is composed predominantly of non-starch carbohydrates, starch and crude proteins [7]. Wheat bran is an economic and readily available by-product and offers therefore a cost-effective microbial enzyme production. A classical approach for the design of media has been used for centuries to improve fermentation. More systematic approaches to media design have been applied in

42 Chapter 3 modern times, e.g. defining a nutrient profile that mimics the elemental composition of the cells to be grown in terms of C: N ratio, balancing phosphate, sulfate and other salts present in the cell [8]. Besides the medium, there are many variables that can influence the biotechnological process in solid-state fermentation. Hence, optimization is essential for improving the medium and fermentation process. The traditional (classical) method of medium and process optimization applying the “one-variable-at-a-time” approach, which involves changing one parameter at a time while keeping the other entire parameters constant, is costly, labor and time intensive [9]. At the same time, the factors have no more than 4 or 5 different levels of variation as the total number of trials is particularly big, e.g. testing the effect of 5 factors at 5 levels would end up in 55= 3125 different combinations of factor-trials for the complete testing of the research subject. This is hard to realize in practice, so that their number is reduced, which decreases the confidence of conclusions [10]. Moreover, it does not consider the interaction between variables. An alternative and more efficient approach is the use of statistical methods. Statistical experimental design, also called design of experiment (DoE), is a well-established concept for planning and conducting of informative experiments, analyzing the response data so that valid and objective conclusions are obtained [11]. Hereby, the following is essential:

 minimization of total number of trials  simultaneous varying of all factors over a set of planned experiments  results are connected by means of a mathematical model, which then is used for interpretations, predictions and optimization

To increase polygalacturonase production it was decided to utilize DoE to achieve a survey of important factors for the solid-state process related to PG production and to explore the optimal settings of these factors for increased enzyme yield. Optimization of the fungal production process included improving medium performance as well as identifying optimal environmental process parameters, such as pH, temperature, and moisture content. The aim of this study was to determine the optimal conditions of a solid-state process using an economic medium composed of agricultural or agro-industrial by-products for maximal PG production by A. sojae. Since a complex medium was applied the enzyme product also comprised various enzyme activities and optimization was conducted in order to increase especially polygalacturonase activity in the crude extract. A comparative evaluation of Aspergillus species identified A. sojae ATCC 20235 as highest pectinolytic enzyme producer in previous work [12]. Therefore this strain was

43 Chapter 3 chosen for further optimization of pectinase production in SSF. Furthermore, one of the newly identified pectinolytic enzyme producing A. sojae strains, A. sojae CBS 100928, was also chosen for optimizing PG production in SSF.

3.2 Materials and Methods

3.2.1 Materials The investigation to increase the amount of polygalacturonase production in SSF was performed on different agricultural and agro-industrial residues obtained from different suppliers. Nordzucker AG (Uelzen, Germany) provided two different kinds of sugar beet pulp pellets (by-products of the sugar industry): molassed with approximately 30% molasses and unmolassed. Pelletized pulp (dry matter > 89%) was grinded to small particles due to easier handling in SSF in laboratory scale. Dried bitter orange peel was purchased from Heinrich Klenk GmbH & Co. KG (Schwebheim, Germany) and was also grinded to a fine-particle granulate in a coffee mill. Grinding of substrates resulted in a heterogeneous mixture of small particles of varied size. Apple pomace, a by-product from processing apples, was a heterogeneous mixture of different kinds of apples, like Elstar, Jonagold, Jonagored or Braeburn and was obtained from Döhler GmbH (Neuenkirchen, Germany). It had a moisture content of approximately 73 % (wet-basis moisture content), determined by drying at 105°C until constant weight. All by-products in this study were used in combination with wheat bran which was obtained as fine bran (90% < 630 µm) from Bremer Rolandmühle Erling GmbH & Co. KG (Bremen, Germany). All chemicals were purchased from AppliChem GmbH (Darmstadt, Germany), except dithiothreitol (DTT) was purchased from Carl Roth GmbH & CO. KG (Karlsruhe, Germany). Substrates for detection of pectinolytic activities, e.g. pectin, polygalacturonic acid and polygalacturonic acid sodium salt, as well as the chemical sodium arsenate dibasic heptahadrate, and pectinase from A. niger were purchased from Sigma-Aldrich Chemie GmbH (Steinheim, Germany). Fructozym P was obtained from ERBSLÖH Geisheim AG (Geisheim, Germany).

3.2.2 Microorganism The Aspergillus strains used throughout this study were all purchased in lyophilized form from different culture collections and propagated on agar plates according to the specifications given by the culture collections [12]. The Aspergillus strain A. sojae ATCC 20235 was purchased from Procochem Inc (Teddington, United Kingdom), an international distributor of the American Type of Culture Collection (ATCC) in Europe. It has to be noted that A. sojae ATCC 20235, which is still deposited as A.

44 Chapter 3 sojae at the ATCC, did not meet the requirements to be classified as A. sojae on the basis of morphological parameters [13] and has been recently reclassified as A. oryzae based on the alpA restriction fragment length polymorphism (RFLP) [14]. A. sojae CBS 100928 was obtained from the Centraalbureau voor Schimmelcultures (CBS) (Utrecht, Netherlands).

3.2.3 Inoculum The spore suspensions used as inoculum were obtained from molasses agar slants containing: glycerol (45 g/L), molasses (45 g/L), peptone (18 g/L), NaCl (5 g/L), KCl

(0.5 g/L), FeSO4·7H2O (15 mg/L), KH2PO4 (60 mg/L), MgSO4 (50 mg/L), CuSO4·5H2O

(12 mg/L), MnSO4·H2O (15 mg/L) and agar (20 g/L). Slants were incubated at 30 °C for 1 week. Spores were harvested from the slants using sterile Tween 80 water (0.02%) and counted in a Thoma counting chamber to adjust the spore concentrations.

3.2.4 Culture medium and growth conditions SSF was performed in 300-mL culture flasks containing 10 g solid media wetted with diluted HCl at the respective concentrations according to the experimental design given in section 3.2.10.‎ The described moisture levels in all experimental set-ups were calculated as dry basis moisture content according to the following equation:

Moisture content (%) = (weightwet – weightdry) / weightdry *100 (3.1) where weightwet is the weight of solid media together with diluted HCl and weightdry means weight of solid media (natural moisture content in solid media has not been considered for calculation). Culture flasks containing the wetted media were sterilized at 121 °C for 20 min. After cooling to room temperature (R.T.) flasks were inoculated with 1 mL of spore suspension containing the desired amount of total spores and incubated as needed (section 3.2.10).‎

3.2.5 Enzyme leaching, protein and soluble carbohydrate estimation At the end of cultivation, the enzyme recovery was obtained by adding 50 mL distilled water into each flask and mixing in an incubator shaker (Innova 4230, New Brunswick Scientific) at 350 rpm, 24 °C, for 60 min. The extract was separated from the fermented substrate by centrifugation at 4 °C, 3220×g, for 30 min. Enzyme activity, soluble carbohydrate content and total protein concentration present in the crude extracts were determined in the supernatant. Total extracellular protein was measured according to the modified Bradford´s method [15] as described in Appendix A. Soluble protein content was expressed as mg per gram dry substrate.

45 Chapter 3

Soluble carbohydrates in the crude extract were determined by the phenol–sulfuric acid method according to Dubois et al. [16], following the instructions given in Appendix A. Soluble carbohydrate content was expressed as mg per gram dry substrate.

3.2.6 Exo-pectinolytic activity measurement

3.2.6.1 Polygalacturonase assay Exo-PG activity was assayed according to the procedure of Panda et al. [17], which was further optimized as described in Appendix A.

3.2.6.2 Polymethylgalacturonase assay Exo-PMG activity was determined according to the method provided by Blandino et al. [18] with slight modifications, as previously described [12]. In brief: samples of 0.25 mL containing an appropriated dilution of the PMG enzyme were mixed with 0.5 mL of 0.5% (w/v) pectin dissolved in 0.1 M acetate buffer (pH 5.0). Reaction mixture was incubated at 45°C for 10 min. The release of reducing groups from citrus pectin was measured by using the DNS method [19], and galacturonic acid was used as standard for calibration. One unit of exo-PMG activity was defined as the amount of enzyme that catalyses the release of 1 µmol of galacturonic acid per unit volume of supernatant per unit time at standard assay conditions mentioned above.

3.2.7 Endo-pectinolytic activity measurement Endo-enzyme activity was determined by measuring the decrease in viscosity of a substrate solution, either 2% (w/v) pectin for endo-PMG or 3.2% (w/v) polygalacturonic acid (sodium salt) for endo-PG. Pectinolytic activity was assayed by adding 0.2 mL of enzyme sample containing 1 U/mL exo-PG activity, to 0.2 mL of 0.2 M acetate buffer (pH 5.0) and 1.6 mL substrate. The mixture was incubated in a water bath for 5 min at 40°C. After incubation, the mixture was cooled down for 30 sec in ice-cold water and viscosity of the samples was determined. The later was done indirectly by measuring the time required for 0.9 mL of reaction mixture to elute through a 1.0 mL glass pipette. Samples were measured in duplicate. Viscosity was calculated from a calibration curve obtained by time measurements to pass polyvinylpyrrolidone (PVP) 360 aqueous standard solutions through the pipette at 25 °C, which were previously passed through an Ostwald viscometer at 25 C. Controls for non-enzymatically treated substrate solutions were included, utilizing 0.2 mL water instead of enzyme samples. One unit of endo-pectinase activity was

46 Chapter 3 defined as the quantity of enzyme which caused a 50% reduction in viscosity of the reaction mixture per minute, under these assay conditions described above [20].

3.2.8 Plate assay for proteolytic activity Dual-substrate assay plates were prepared according to the procedure given by Montville [21], containing 1% (w/v) casein and 1% (w/v) gelatin as substrates. The assay was performed as previously described [12]. Zone diameters were converted to log10 adjusted zone area. Proteolytic activity was reported in this manner and 2 referred to as zone area (log10 mm ).

3.2.9 Protein pattern studies The crude extract was dialyzed over night at 4 °C, using SnakeSkin® pleated dialysis tubing, 10,000 MWCO (Thermo Scientific, Rockford, USA). Samples were concentrated to 1.5 mg/mL total protein concentration, using a freeze-dryer.

3.2.9.1 One-dimensional electrophoresis SDS-PAGE was performed according to the method of Laemmli [22], following the procedure described in the technical manual for protein electrophoresis [23]. Briefly: 12.5% SDS-PAGE gels with an approximately 2 cm stacking buffer zone were cast and samples run in constant current mode at 15 mA/gel, at R.T. Samples were mixed with 2x treatment buffer in the ratio 2:1. Sample load add up to 10 µL per lane. Protein bands were visualized, using colloidal Coomassie (G-250) staining [24].

3.2.9.2 Native polyacrylamide gel electrophoresis / Zymogram Native PAGE was performed by excluding SDS and DDT from the electrophoresis protocol described above. The “sandwich” method was used to detect the activity of pectinases acting on polygalacturonic acid sodium slat as substrate [25]. Briefly, proteins were separated on a native PAGE and subsequently the gel was first incubated for 20 min in 0.1 M citrate phosphate buffer (pH 5.0) and afterwards contacted with an (solid) agar substrate containing 0.25% (w/v) polygalacturonic acid sodium salt for 50 min and for 90 min at 30°C (80% humidity chamber). The agar plate was then treated with 1% (w/v) cetyltrimethylammonium bromide which precipitated the substrate and revealed pectinases activity as translucent bands on an opaque background.

3.2.10 Designing experiments and experimental designs The experiments were planned according to DoE applying different designs at different steps. Data analysis was performed by using multiple linear regression

47 Chapter 3

(MLR) and partial least squares projections to latent structures (PLS) applying the MODDE 9.0 software package, supplied by Umetrics AB, Umeå, Sweden. At the beginning screening experiments were conducted to explore many factors in order to reveal whether they have a significant influence on the response and to identify their appropriate ranges. Numerous variables may affect the SSF process. Since it is practically impossible to identify the effects of all parameters, it was necessary to select these ones having major effects. In the next step response surface methodology (RSM) was used for optimization. RSM consists of a group of mathematical and statistical techniques useful for the modeling and analysis of problems in which the response is influenced by several variables and the objective is to optimize the response [26]. During this study optimization was performed to understand in more detail how the factors influence the response, to make predictions and finally to optimize the process for PG production and identify the region of operability. In both cases, screening and optimization, an experimental design had to be chosen, involving the consideration of sample size, selection of a suitable run order and so on. The software package MODDE supported this phase of experimental design. After entering information about the number of factors, levels, ranges, and choosing the type of investigation (screening or RSM), the software presented a selection of designs for consideration and also recommended a particular design of the selection. Detailed information on applied designs during this study is given below. Experiments were conducted according to the selected experimental matrix followed by statistical analysis of the data. Also during this step, the software package assisted in data analysis. The data collected by the experimental design was used to estimate the coefficients of the model. The model represented the relationship between the response and the factors. The mathematic-statistical treatment of the obtained experimental data through the fit of a polynomial function was followed by the evaluation of the model’s fitness. Three model evaluation tools have been applied which were useful for giving guidance of how to formulate the most valid models [27]. The two companion statistics, goodness of fit (R2) and goodness of prediction (Q2) are one of these tools. R2 is a measure of how well the regression model can be made to fit the raw data and Q2 estimates the predictive power of the model. Both R2 and Q2 should be high (close to 1.0) and preferable not be separated by more than 0.2 – 0.3, for a model to pass this diagnostic test. The second tool, analysis of variance (ANOVA), is useful for checking the adequacy of a regression model in terms of a lack of fit (LoF) test. This LoF test implies that the residual response sum of squares is separated into the components model error and pure error, and their sizes are compared by an F-test. The third

48 Chapter 3 applied tool is the evaluation of the model residuals using a normal probability plot (N-plot) for detecting deviating experiments [11, 27]. The last step was the graphical presentation of the model equation and determination of optimal operating conditions. The visualization of the predicted model equation was obtained by contour plots, showing the relationship between the response and the independent variables. Finally, conformation testing was performed for the validation of the conclusions.

3.2.10.1 Screening steps for increased PG production by A. sojae ATCC 20235 In the first step five factors (kind of inducer substrate, temperature, inoculum size, time and HCl concentration) were screened with respect to their main effects on PG production in SSF by a complemented D-optimal design (Table 3.1‎ ). D-optimal design was used due to complementation of the screening design by two repetitions of the center points for each of the 4 inducer substrates, which resulted in 24 experimental runs, consisting of 16 runs in design and 8 center points. In a second screening step a full factorial design was applied to study the main effect and the interaction effect of two significant inducer substrates, inducer substrate concentration, a narrowed HCl concentration range and incubation time on exo-PG activity (Table 3.2‎ ). The design matrix consisted of 20 experimental runs, including 16 design runs and 2 center points for each inducer substrate.

3.2.10.2 Optimization experiments for increased PG production by A. sojae ATCC 20235 Final medium composition was optimized by response surface methodology, using D-optimal design (Table 3.4‎ ). The initial central composite face-centered (CCF) design with two factors (inducer concentration and moisture level) consisted of 11 experimental runs having 3 replicates at the central point. It did not include the optimal region and had to be completed by 7 more experiments to a D-optimal design with 18 experimental runs. In a second optimization step, RSM was applied to explore optimal settings for the 2 process parameters cultivation time and temperature. The initial design had to be also completed by 3 additional experiments and changed to a D-optimal design, due to further increase of the response values at longer incubation times (Table 3.5‎ ). Final experimental set-up consisted of 14 experimental runs including 3 center points of the initial design.

3.2.10.3 Screening step for increased PG production by A. sojae CBS 100928 In the screening part a full factorial design was applied to explore the effects of 3 factors (moisture level, cultivation time and temperature) on PG production by A.

49 Chapter 3 sojae CBS 100928 (Table 3.6‎ ). The center points were performed at the low and high levels of temperature in duplicate. Therefore, the design included 12 experimental runs.

3.2.10.4 Optimization of PG production by A. sojae CBS 100928 Optimization of PG production was performed by RSM, using CCF design (Table 3.7‎ ). The factors moisture level and cultivation time were investigated in 11 experiments including 3 repetitions of the central point.

3.3 Results and Discussion Previous attempts revealed already the potential of PG production in SSF by A. sojae ATCC 20235 and A. sojae CBS 100928 [12]. Screening and optimization experiments were conducted in order to increase the yield of exo-PG in SSF utilizing low-cost substrates. Substrates used as inducer for pectinase production (dried bitter orange peel and pelletized sugar beet pulp) were grinded, which resulted in a heterogeneous mixture of small particles of varied size. Particle size is very important in SSF systems as it determines the amount of substrate surface available as well as the void space which is occupied by air (oxygen). Furthermore it affects the packing density within the substrate mass [28]. However, Pandey [29] obtained enzyme yields utilizing substrates with mixed particle sizes comparable to the highest yields achieved using a defined particle size, and noted since the rate of oxygen transfer into the void space affects growth, the substrate should contain particles of suitable sizes to enhance mass transfer. Besides medium composition, the effect of process parameters like temperature or incubation time on pectinase production was investigated by measuring the exo-PG activity.

3.3.1 Optimization of PG production by A. sojae ATCC 20235 in SSF

3.3.1.1 First screening step for PG production by A. sojae ATCC 20235 First, a screening of suitable pectinase inducer substrates was done in combination with wheat bran at the ratio 30:70 for maximum PG activity. The aim of the first step was to determine the effect of various agricultural substrates on PG production in SSF. Therefore, agricultural byproducts such as orange peel, apple pomace, sugar beet pulp and sugar beet pulp supplemented with molasses were used as inducer substrates. Media were wetted at 120 % dry basis moisture content by different concentrations of diluted hydrochloric acid (Table 3.1‎ ). Tolerance of fungi towards low pH-values is minimizing the risk of contamination due to acidic pH treatment. Moreover, many fungi secrete PG in acidic media and this is also the pH range where majority of their PGs show optimum catalytic activity

50 Chapter 3

[30]. Furthermore, the effect and ranges of process factors like temperature, inoculum size and cultivation time affecting PG synthesis were explored using a complemented D-optimal design (Table 3.1‎ ).

Table 3‎ .1 D-optimal design and experimental results of exo-PG activity in the first screening step of A. sojae ATCC 20235.

Experimental factors Response

Exp. No. HCl exo-PG Inducer Temperature Inoculum Time conc. activity substrate (°C) (Total spores) (d) (mM) (U/g)

1 Orange peel 32 3 × 107 3 50 30.3

2 Orange peel 32 3 × 107 5 50 21.5

3 Orange peel 24 104 3 300 -

4 Orange peel 24 104 5 300 -

5 Apple pomace 24 104 3 50 20.0

6 Apple pomace 32 104 3 50 20.1

7 Apple pomace 24 3 × 107 5 300 216.3

8 Apple pomace 32 3 × 107 5 300 223.8

9 Sugar beet pulp 24 3 × 107 3 50 58.5

10 Sugar beet pulp 32 104 5 50 94.9

11 Sugar beet pulp 24 3 × 107 3 300 -

12 Sugar beet pulp 32 104 5 300 - Sugar beet with 13 24 104 5 50 76.7 molasses Sugar beet with 14 24 3 × 107 5 50 30.5 molasses Sugar beet with 15 32 104 3 300 - molasses Sugar beet with 16 32 3 × 107 3 300 - molasses 17 – 18* Orange peel 28 1,5 × 107 4 175 140.3 ± 3.8

19 – 20* Apple pomace 28 1,5 × 107 4 175 243.0 ± 10.3

21 – 22* Sugar beet pulp 28 1,5 × 107 4 175 407.5 ± 13.9 Sugar beet with 23 – 24* 28 1,5 × 107 4 175 291.2 ± 1.7 molasses *The standard variation between the values of each center point repetition was below 5%.

51 Chapter 3

From Table 3.1‎ can be seen that experiments performed at high HCl concentrations resulted in no activity due to repression of fungal growths under these conditions. This observation will be discussed more precisely below. Nevertheless, for analysis of the data and evaluation of the model the experimental runs without fungal growth have been excluded. Prior to the data analysis, the distribution of the response (exo-PG activity) was investigated using a histogram (no plot shown). Because of the appearance of strong positive skewness logarithm transformation was used. The MLR modeling of the screening data gave a model with R2 = 0.88 and Q2 = 0.71, suggesting a sound model. Nevertheless, due to the exclusion of several experiments resulting in no growth and consequently in no PG production, this first screening investigation provided only an indication for identification of significant factors and their ranges. Therefore, another screening investigation was performed before conducting optimization experiments. According to the evaluation of the present screening experiments the inducer substrates apple pomace and sugar beet pulp significantly affected exo-PG activity as well as the concentration of diluted HCl. Therefore, these two agro-industrial byproducts were chosen for further screening investigation on increase of PG production in SSF. It was found that an especially useful type of sugar beet pulp pellets is un-molassed sugar beet pulp. SSF processes are known to overcome the effect of catabolite repression [31-33]. The concentration of molassed sugar beet pulp as inducer substrate might be even too low to cause catabolite repression. Rather, the presence of an additional carbon source might trigger other metabolic processes than stimulating exo-PG production. Maximal PG activity was obtained at the center points of the experimental setup using sugar beet pulp as inducer substrate. Besides center points, high enzyme activity was measured after 5 days SSF utilizing apple pomace as inducer compound and substrate wetted with 0.3 M HCl. Though, high level of acid concentration at 0.3 M HCl inhibited fungal growth totally applying other agricultural byproducts as inducer substrate except for apple pomace. Since substrates were used in their original state without any drying step, the total HCl concentration in apple pomace was diluted due to its high moisture content (section 3.2.1).‎ The effect of different moisture levels of inducer substrate was further explored (section 3.3.1.2).‎ However, total inhibition of fungal growth showed the significance of this factor and HCl concentration needed to be decreased for further experiments. The factor inoculum size did not affect PG production significantly and was therefore fixed at 2 × 107 spores per flask for optimization experiments. A survey of temperature ranges was obtained by the first screening step. Since highest enzyme activity was measured at the center point,

52 Chapter 3 temperature was fixed at 28°C at the next step and further specification of this factor was delayed for the optimization of process parameters.

3.3.1.2 Second screening step for PG production by A. sojae ATCC 20235 In the next step, two significant medium components apple pomace and un- molassed sugar beet pulp were chosen as inducers for PG production. Their effect at different concentrations on exo-PG activity was explored as well as the effect of incubation time and HCl concentration using full factorial design (Table 3.2‎ ).

Table 3‎ .2 Full factorial design and experimental results of exo-PG activity in the second screening step of PG production by A. sojae ATCC 20235.

Experimental factors Response

Exp. No. Inducer HCl exo-PG Inducer Time concentration concentration activity substrate (d) (%) (M) (U/g)

1 Apple pomace 10 0.1 3.5 28.8

2 Sugar beet pulp 10 0.1 3.5 68.2

3 Apple pomace 40 0.1 3.5 57.5

4 Sugar beet pulp 40 0.1 3.5 104.2

5 Apple pomace 10 0.22 3.5 209.5

6 Sugar beet pulp 10 0.22 3.5 300.4

7 Apple pomace 40 0.22 3.5 165.5

8 Sugar beet pulp 40 0.22 3.5 -

9 Apple pomace 10 0.1 5.5 4.3

10 Sugar beet pulp 10 0.1 5.5 12.5

11 Apple pomace 40 0.1 5.5 38.1

12 Sugar beet pulp 40 0.1 5.5 88.8

13 Apple pomace 10 0.22 5.5 91.7

14 Sugar beet pulp 10 0.22 5.5 374.5

15 Apple pomace 40 0.22 5.5 225.2

16 Sugar beet pulp 40 0.22 5.5 -

17 – 18* Apple pomace 25 0.16 4.5 118.2 ± 5.4

19 – 20* Sugar beet pulp 25 0.16 4.5 277.5 ± 1.7 *The standard variation between the values of each center point repetition was below 5%.

53 Chapter 3

Since the significance of HCl concentration and the need of reducing this factor range were already demonstrated in the first screening experiment, the range of HCl concentration was explored from 100 mM to 220 mM in the second step. The combination of HCl concentration at high factor level of 0.22 M with high sugar beet pulp concentration of 40 % also inhibited fungal growth totally and therefore no PG activity could be produced. Hence, tow experiments (Exp. No. 8 and 16) applying this factor combination were excluded from the screening investigation. Prior data analysis, the distribution of the response was investigated using histogram (plot not shown). Because the distribution was skewed the logarithm transformation was used to obtain the desired “bell shaped” normal distribution. In general, normally distributed responses will give better model estimates and statistics. The replicate plot in Figure 3.1‎ represents a graphical version of the “signal-to-noise-ratio”, and shows the variation among the replicates (last two sticks) is low compared to the variation across the entire data material. 400 14 400 14 350 350 6 300 20 6 19 300 1920 250 15 5 250 200 15 5 7 200 150 7 17 4 18

exo-PG exo-PG activity (U/g) 150 100 12 13 17 2 4 18 exo-PG activity (U/g) 3 12 13 50 100 11 1 2 3 9 10 50 1 11 0 10 2 4 6 89 10 12 14 0 2 4 6 Experiment8 number10 12 14 16 Figure 3‎ .1 Replicate plot, showing the variationReplicate in results Index for all experiments of the 2nd screening MODDE 9 - 2013-02-06 13:57:11 (UTC+1)

investigation. Repeated experiments of each center point appear on the sameMODDE 9respective - 2013-02-06 14:34:33 (UTC+1) stick (the two last sticks).

The MLR modeling of the exo-PG data from the second screening investigation gave a preliminary model with R2 = 0.935 and Q2 = 0.73, i.e. suggesting a sound model. This model showed a slight lack of fit (p = 0.01), but normally distributed residuals (normal probability plot not shown). The obtained LoF might result from some nonlinear blending behavior, which the screening model is unable to cope with. However, the model was still used for indicating where to continue. Evaluation of results showed that both inducer substrates as well as their concentration and as previously also determined the concentration of HCl significantly influenced exo-PG activity in the crude extracts. Furthermore, the interactions of the factors amount inducer/time and HCl concentration/time also

54 Chapter 3 had a significant effect on enzyme activity. Highest PG activity was obtained applying 10 % sugar beet pulp as inducer substrate in the medium wetted at 120 % with 0.22 M HCl after 5.5 days incubation at 28°C. Increase of this inducer substrate to 40 % caused total inhibition of fungal growth in combination with 0.22 M HCl concentration. The combination of high amount of apple pomace and high HCl concentration resulted in lower acid concentration compared to sugar beet pulp due to the higher moisture content of apple pomace as mentioned above. Since the substrates were used how they were obtained from the different providers, the moisture content differed strongly. This affected the total amount of substrate in the SSF process as well as other significant process parameters, e.g. moisture level. The moisture level in SSF systems is very important [34]. High moisture levels decrease the substrate porosity and hence, reduce the oxygen transfer. While low moisture contents may limit the bioavailability of nutrients and increase the accumulation of heat. Therefore, experiments were performed comparing these two substrates at similar moisture levels (Table 3.3‎ ).

Table 3‎ .3 Comparison of inducer substrate for PG production by A. sojae ATCC 20235.

Amount HCl exo-PG Inducer HCl Time Exp. No. inducer concentration activity substrate (mL) (d) (g) (mM) (U/g)

1 Apple pomace 3 12 175 4 294.2 (freeze-dried) 2 11 3.89 539 4 235.6 Apple pomace 3 3 12 175 4 143.6 Apple pomace 4* 0.81 12 175 4 295.8 Sugar beet pulp 5 3 12 175 4 321.7 Sugar beet pulp 6 3 12 175 5 372.3 Sugar beet pulp 7 3 12 175 6 468.1 Sugar beet pulp 8 3 12 190 4 407.3 Sugar beet pulp 9 3 12 200 4 405.0 Sugar beet pulp

* Addition of 2.19 mL dH2O to the media.

The apple pomace used during these experiments was a mixture of different kinds of apples and had a moisture level of approximately 73 % (section 3.2.1).‎ The moisture content in the dehydrated and pelletized sugar beet pulp was negligible marginal. Process conditions like temperature (28 °C), inoculum size (1.5×107 spores), or mainly 30 % of inducer substrate were kept similar to the center point

55 Chapter 3 conditions of the first screening experiment to establish a better comparability. Thus, HCl concentration was mainly kept at 175 mM. Besides the inducer substrate all flasks contained 7 g wheat bran. In experiment number (Exp. No.) 1 apple pomace was freeze-dried to reduce the moisture content and to establish comparable conditions to Exp. No. 5. Utilizing sugar beet pulp as inducer instead of freeze-dried apple pomace increased PG activity slightly by 9 %. In Exp. No. 2 were 11 g of apple pomace used to obtain a similar dry-weight of the inducer substrate like with freeze-dried apple pomace (Exp. No. 1). Therefore, the volume of HCl solution had to be reduced and the concentration had to be increased to obtain a media with a moisture level of 120 % and a final acid concentration of 0.2 M. This decreased enzyme activity by 20 % compared to Exp. No. 1. Furthermore, the total amount of sugar beet pulp in Exp. No. 4 was reduced to obtain the same dry-weight like with 30 % of fresh apple pomace (Exp. No. 3) and remaining 2.19 g were filled up with dH2O. Comparing these experiments 2.1 times higher PG activity was obtained utilizing sugar beet as inducer substrate. Based on the results of this media comparison sugar beet pulp was chosen as inducer substrate for further optimization experiments. In addition to the media comparison the results of Exp. No. 5 – 7 indicated a further increase of PG activity at longer cultivation times. Exp. No. 8 and 9 presented similar activities which were slightly higher than in Exp. No. 5. Based on these results and the judgment based of experience of previous experiments the HCl concentration was fixed at 0.2 M. The combination of a cultivation medium including predominantly wheat bran with applying the moisture level by diluted HCl at a concentration of 0.2 M was already demonstrated for enhanced microbial enzyme production in SSF system by filamentous fungus [35].

3.3.1.3 First optimization step of PG production by A. sojae ATCC 20235 During optimization sugar beet pulp was chosen as inducer which resulted previously in maximum enzyme activity. The concentrations of this ingredient were optimized together with different moisture levels in the medium using a complemented D-optimal design (Table 3.4‎ ). The moisture level was applied by 0.2 M HCl. Hence, varying the moisture level influences also water activity and pH of the fermentation system. Generally, varying the moisture level as independent parameter should be seen disadvantageously, because changing the moisture level means a change in the total reaction volume, which changes both substrate volume in solid-state cultivation and enzyme concentration obtained after enzyme leaching. The optimal setting of the factor moisture content for the cultivation of microorganisms in solid-state cultivation

56 Chapter 3 processes is highly dependent upon water-binding properties of the substrate [36]. Besides growth of the microorganism, also the formation of products, such as enzymes, is markedly affected by the moisture content in SSF systems [36]. Hence, this factor had to be optimized for PG production by A. sojae in the present SSF system.

Table 3‎ .4 D-optimal design and experimental results of exo-PG activity in the first optimization step of A. sojae ATCC 20235.

Experimental factors Response

Exp. No.

Inducer concentration (%) Moisture level (%) Enzyme activity (U/g)

1 10 85 124.1

2 35 85 270.1

3 10 135 371.2

4 35 135 552.3

5 10 110 252.4

6 35 110 413.5

7 22.5 85 221.9

8 22.5 135 493.6

9* 22.5 110 347.5

10* 22.5 110 328.5

11* 22.5 110 325.4

12 65 165 -

13 65 145 -

14 45 165 512.4

15 45 145 418.7

16 55 155 39.5

17 65 135 -

18 35 165 557.2 *The standard variation between the center point repetitions was below 3 %.

57 Chapter 3

The original experimental set-up (Table 3‎ .4, Exp. No. 1 – 11) of the response surface modeling included an inducer concentration range from 10 – 35 % and a moisture level range from 85 – 135 %. This design had to be complemented by Exp. No. 12 – 18 due to further increase in PG activity to all appearances by further increasing the amount of inducer substrate and the moisture level as well. Therefore, Exp. No. 12 – 18 were complemented including higher factor ranges up to 65 % inducer concentration and 165 % moisture level, which resulted in the D-optimal design for response surface modeling. All cultivations were inoculated with a total amount of 2×107 spores of A. sojae ATCC 20235 and incubated at 28 °C for 5 days. The MLR analysis of the data shown in Table 3.4‎ gave a model with R2 = 0.90 and Q2 = 0.84, suggesting that the quadratic model is well founded and valid. Nevertheless, exploring the N-plot (plot not shown) Exp. No. 16 seemed to be an outlier. Hence, this value was excluded from the optimization investigation. Upon exclusion and refitting the model, an increase in R2 to 0.97 and in Q2 to 0.94 was obtained. Also the LoF test (p = 0.087) pointed in the direction of a valid model. Results in Table 3.4‎ represented growth inhibition at high inducer concentration in combination with high moisture levels supplied with 0.2 M HCl. High PG activities were obtained at Exp. No. 4 and 18 using 35 % sugar beet pulp as inducer substrate. Prediction for high exo-PG activity values was obtained for inducer concentration of 15 – 34 % in the medium and moisture levels of 157 – 165 % applied by 0.2 M HCl (Figure 3.2‎ ).

exo-PG activity (U/g)

Figure 3‎ .2 Contour plot presenting the interaction between amount of inducer substrate and moisture level on PG activity produced by A. sojae ATCC 20235 during the first optimization step.

58 Chapter 3

Highest exo-PG activity of approximately 613 U/g was predicted at an inducer concentration of 25% and 160% moisture content level. Therefore, an experimental trial including the predicted optimized conditions for PG production, as well as a slightly higher inducer concentration of 30% close to the highest results of the optimization experiments were tested. The media were wetted at 160% with 0.2 M HCl and cultivation was performed at 28 °C for 5 days. Under the predicted optimized conditions 634.0 ± 44.6 U/g PG activity was obtained which is very close to the predicted result. This represented a very good correlation between the experimental data and the predicted value of the exo-PG activity indicating a good fit of the model. Nevertheless, using 30% of inducer substrate in the medium resulted in a slightly higher exo-PG activity of 648 ± 22.3 U/g. Therefore, the media composition was fixed at wheat bran and sugar beet pulp in the ratio 70:30, wetted at 160% with 0.2 M HCl. Hence, the first optimization step provided the final medium composition.

3.3.1.4 Second optimization step of PG production by A. sojae ATCC 20235 Previous results demonstrated already further increase of exo-PG activity at longer incubation times (section 3.3.1.2).‎ Also the factor incubation temperature needed to be further optimized after the first screening step (section 3.3.1.1).‎ In a second optimization step the SSF process parameters incubation time and temperature were explored. Due to the tendency of further increase of exo-PG activity at longer cultivation times during the optimization experiments, the experimental set-up was complemented by 3 additional runs ( Exp. No. 12 – 14) applying a D-optimal design (Table 3.5‎ ). The MLR modeling of the data gave a model with R2 = 0.95 and Q2 = 0.869, suggesting that the quadratic model was valid. Upon removal of insignificant model terms, which were determined by exploring the plot of the regression coefficients (plot not shown), the value of Q2 increased to 0.896. Also the LoF test (p = 0.73) and the N-plot of the model residuals strongly pointed in the direction of a valid model.

59 Chapter 3

Table 3‎ .5 D-optimal design and experimental results of exo-PG activity in the second optimization step of A. sojae ATCC 20235.

Experimental factors Response Exp. No. Cultivation time (d) Temperature (°C) exo-PG activity (U/g)

1 5 26 496.0

2 7 26 603.1

3 5 34 306.7

4 7 34 532.4

5 5 30 586.2

6 7 30 678.2

7 6 26 614.0

8 6 34 390.5

9* 6 30 619.4

10* 6 30 705.3

11* 6 30 680.3

12 8 26 693.9

13 8 30 847.3

14 8 34 651.0 *The standard variation between the center point repetitions was below 10%.

Additional experiments presented in Table 3.5‎ obtained at the 8th day of cultivation resulted in high exo-PG activity values. Maximal exo-PG activity of 847.3 U/g was obtained after 8 days at 30°C. Comparing all values obtained at the same incubation time but at different temperatures, highest PG activity was always achieved at 30 °C. This is also demonstrated in Figure 3.3‎ , where high PG activity at different cultivation times is given at the temperature scale around 30 °C.

60 Chapter 3

exo-PG activity (U/g)

Figure 3‎ .3 Contour plot of the second optimization step presenting interaction of cultivation time and temperature on PG activity produced by A. sojae ATCC 20235.

The results also strongly indicated further increase of exo-PG activity at longer incubation times. Therefore, further investigation on the factor cultivation time was done to determine the optimal conditions for PG production by A. sojae ATCC 20235 in SSF. Incubation temperature was fixed at 30 °C and cultivation was performed over a period of nine days harvesting daily samples from day 5 till day 9 (Figure 3.4‎ ). According to the results obtained between five to nine days SSF the maximum of PG activity was achieved at the 8th day of cultivation with 909.5 ± 2.7 U/g. Comparing this value with PG activity obtained during optimization at Exp. No. 13 with 847.3 U/g the variation is approximately 7 % which indicated also a good reproducibility. At the peak of enzyme production also a high specific activity of 180.0 ± 6.8 U/mg protein was obtained. The maximal productivity of 128.9 U/g/d was achieved after six days and it slightly decreased to 112.3 ± 0.3 U/g/d at the 8th day of SSF. Only the amount of total protein in the crude extract was further increasing over the time. With regard to previous work [12], where the potential of this strain as PG producer in SSF was presented utilizing a mixture of wheat bran and dried orange peel wetted at 120% with 0.2 M HCl, the maximum in exo-PG activity was 10.9 times increased by the present optimization, measuring both extracts under the same enzyme assay conditions described in Appendix A. Productivity determined at the peak of enzyme activity was 6.7 fold increased after optimization.

61

200 150 100 50 0 200 150 100 50 0

200 150 100 50 0 200 150 100 50 0

10 8 6 4 2 0 10 8 6 4 2 0 10 8 6 4 2 0 10 8 6 4 2 0 9 9 9 9 8 8 8 8

Chapter 3 7 7 7 7 10 200 900

800 8 6 6 6 700 150 6

600 6 500

100 5

5

5 5 0 0 100 0 200 700 500 300 400 0 900 400 800 600 100 200 700 500 300 400 900 800 600

PG activityPG (U/g) 4 100 200 700 500 300 100 400 900

800 600 - 200 700 500 300 400 900

800 600

Productivity(U/g/d) exo

300 (U/g) activity exo-PG Protein (mg/g) (U/mg) activity Specific (U/g/d) Productivity Proteincontent(mg/g)

exo-PG activity (U/g) activity exo-PG Protein (mg/g) (U/mg) activity Specific (U/g/d) Productivity Specificactivity(U/mg); exo-PG activity (U/g) activity exo-PG Protein (mg/g) (U/mg) activity Specific (U/g/d) Productivity exo-PG activity (U/g) activity exo-PG 50 Protein (mg/g) (U/mg) activity Specific (U/g/d) Productivity 200 2

100

0 0 0 5 6 7 8 9 Time (d) Figure 3‎ .4 Solid-state fermentation profile of A. sojae ATCC 20235 utilizing optimal conditions for exo- PG production.

Additionally, to demonstrate the importance of sugar beet pulp as inducer substrate for pectinase production, A. sojae ATCC 20235 was cultivated under optimized conditions on the mixture of wheat bran and sugar beet pulp in the ratio 70:30 and on pure wheat bran. In both cases 10 g of dry substrate was utilized. The obtained enzyme activity after 8 days SSF was 197 times higher if the inducer substrate was present. This result strongly favored utilization of sugar beet pulp in SSF process for pectinolytic enzyme production. Hence, vegetable material composing pectic substances, has an impact on the enzyme product produced by A. sojae. Besides exo-PG activity the total protein concentration in the crude extract obtained with fermented wheat bran was 1.9 times higher compared to the extract from the fermented substrate mixture, which suggested the stimulation of pectinase production by the presence of sugar beet pulp as inducer substrate in SSF by A. sojae ATCC 20235. Hence, specific activity was 373 times higher applying the mixture containing the inducer substrate. Interesting seemed to be also the difference in protein concentration obtained from the extracts of the media without cultivation. Higher soluble protein content with 7.02 ± 0.02 mg/g was measured in the extract obtained from the wheat bran and sugar beet pulp mixture while the extract of the wetted pure wheat bran contained 5.52 ± 0.08 mg/g soluble protein. The total soluble carbohydrate concentration was also higher in the mixtures containing the inducer substrate. Hence, medium supplemented with sugar beet pulp as inducer substrate increased the amount of soluble protein and carbohydrates available for the fungal metabolism. Nevertheless, optimization

62 Chapter 3 results also indicated the imperative of wheat bran for the SSF (Table 3.4‎ ). At a sugar beet pulp concentration of 55% fungal growth was strongly inhibited and there was no fungal growth observed at a sugar beet pulp concentration of 65% in combination with wetting the medium with 0.2 M HCl. Therefore, the combination of both substrates in optimized ratio is essential for a productive SSF by A. sojae ATCC 20235. Wheat bran is a suitable substrate for cultivation of many microorganisms, which provides sufficient nutrients and is able to remain loose even in moist conditions, thereby providing a large surface area. Moreover, as wheat bran is a by-product in enormous quantities of the cereal industry, it is an attractive substrate for various biotechnological processes and it used for pectinolytic enzyme production by various microbial cultures [3].

3.3.2 Optimization of PG production by A. sojae CBS 100928 in SSF The optimization of the SSF process was performed to increase exo-PG production by this strain and to compare the optimal conditions for enzyme production with optimized conditions for exo-PG activity by A. sojae ATCC 20235. Prior to optimization a small screening on the different inducer substrates was performed and similar to A. sojae ATCC 20235 maximal exo-PG activity was obtained with grinded sugar beet pulp (data not shown). Hence, the previously optimized medium containing wheat bran and sugar beet pulp in the ratio 70:30, wetted with 0.2 M HCl was also used as SSF medium for this strain.

3.3.2.1 Screening for PG production by A. sojae CBS 100928 During the screening step investigations on the factors moisture level, temperature and cultivation time were performed using full factorial design (Table 3.6‎ ). The center points were conducted at the minimum and maximum temperature level due to the narrow temperature range. Hence, two center points applying lowest temperature level of 24 °C (Table 3.6‎ , Exp. No. 9 - 10) and two center points applying highest temperature level of 30 °C (Table 3.6‎ , Exp. No. 11 - 12) were chosen. Due to the implementation of two center points at the low and high temperature level in the full factorial design the analysis of data was performed by PLS modeling of the screening data. The distribution of the response was investigated using histogram (plot not shown), which identified that the distribution was skewed. Hence, logarithm transformation was used to obtain a distribution closer to the desired “bell shape” distribution. Upon removal of insignificant coefficient terms PLS modeling gave a model with R2 = 0.97 and Q2 = 0.84, suggesting a sound model. Also the LoF test (p = 0.094) and the N-plot of the model residuals (plot not shown) pointed in the direction of a valid model.

63 Chapter 3

Table 3‎ .6 Full factorial design and experimental results of exo-PG activity in the screening step of A. sojae CBS 10928.

Experimental factors Response Exp. No.

Moisture level Temperature Cultivation time exo-PG activity (%) (°C) (d) (U/g)

1 80 24 4 23.6

2 160 24 4 36.4

3 80 30 4 31.6

4 160 30 4 39.6

5 80 24 8 33.9

6 160 24 8 106

7 80 30 8 34.1

8 160 30 8 104.2

9 – 10* 120 24 6 69.8 ± 1.8

11 – 12* 120 30 6 108.1 ± 5.9

*The standard variation between the values of the center point repetition was below 6%.

Evaluation of the screening data presented in Table 3.6‎ indicated that high values of enzyme activity were obtained at incubation times of six to eight days in combination with higher moisture levels and incubation temperature of 30 °C. Maximum exo-PG activity of 108.1 ± 5.9 U/g was achieved at the center points at 30 °C. Therefore, the temperature was fixed at 30 °C. The tendency of optimal parameters for enhanced PG production was similar to the previous findings utilizing A. sojae ATCC 20235.

3.3.2.2 Optimization of PG production by A. sojae CBS 100928 During optimization of SSF process for increased PG production by A. sojae CBS 100928 the moisture level was increased to the range of 130% to 170% and the cultivation time was explored in the range from six to eight days based on the screening results (Table 3‎ .7). Analysis of the data using the regression technique MLR gave a model with R2 = 0.97 and Q2 = 0.797. Upon removal of an insignificant coefficient term Q2 even increased to 0.815. The LoF test (p = 0.485) supported also the conclusion that the quadratic model was valid.

64 Chapter 3

Table 3‎ .7 CCF design and experimental results of exo-PG activity in the optimization step of A. sojae CBS 100928.

Experimental factors Response Exp. No.

Moisture level (%) Cultivation time (d) exo-PG activity (U/g)

1 130 6 87.5

2 170 6 49.7 3 130 8 95.5

4 170 8 95.7

5 130 7 94.7

6 170 7 63.2

7 150 6 94.0

8 150 8 120.7

9 – 11* 150 7 99.7 ± 4.2

*The standard variation between the center point repetitions was below 5%.

Maximal PG activity was obtained at Exp. No. 8, wetting the substrate at 150% with 0.2 M HCl and incubating for 8 days at 30°C, which was also predicted as optimal settings for exo-PG production by A. sojae CBS 100928 in SSF. Evaluating the data presented high activity could be achieved at moisture levels of 138 - 160% and incubation times from 7.5 - 8 days (Figure 3.5‎ ). Hence, the optimized conditions for PG production in SSF by A. sojae CBS 100928 are similar to the optimized conditions of A. sojae ATCC 20235. Therefore and in order to validate the model constructed for exo-PG activity in the optimization step, two experiments at the predicted optimal point and two experiments at a moisture level of 160%, which represented the optimized conditions for the other strain, were conducted. The predicted enzyme activity under optimal conditions was 116.3 U/g and achieved was a PG activity of 131.9 ± 6.9 U/g, which indicated a good compatibility of the model with the experimental results. PG production at a moisture level of 160% was slightly lower with 118.1 ± 1.2 U/g. Hence, optimized conditions for exo-PG production by A. sojae CBS 100928 in SSF utilizing a mixture of wheat bran and sugar beet pulp in the ratio 70:30 as substrate were determined at a moisture level of 150% applied by 0.2 M HCl after 8 days incubation at 30°C.

65 Chapter 3

exo-PG activity (U/g)

6 6.5 7 7.5

Figure 3‎ .5 Contour plot presenting the interaction between moisture level and cultivation time on exo-PG activity produced by A. sojae CBS 100928 during optimization step.

3.3.3 Comparison under optimized conditions Comparing enzyme activities obtained under optimized conditions for both strains 6.9 times higher exo-PG activity was achieved by A. sojae ATCC 20235. This result confirmed previous findings of highest pectinase production by A. sojae ATCC 20235 in SSF [12]. So far the leading fungal pectinase producer was A. carbonarius with an enzyme yield of 480 U/g dry fermented bran utilizing wheat bran as substrate in SSF [3]. This enzyme yield was 1.9 times increased using A. sojae ATCC 20235 in SSF under optimized conditions in this study. Crude extracts of A. sojae ATCC 20235 and A. sojae CBS 100928 obtained under optimized conditions for pectinolytic enzyme production were analyzed by polyacrylamide gel electrophoresis under denaturing conditions. Results are presented in Figure 3.6.‎ Interestingly, the change of inducer substrate and setting of SSF process variables for optimized pectinase production caused a significant change in protein pattern of both crude extracts compared to previous results ([12], chapter 2). Previous cultivation conditions resulted in protein patterns of the crude extracts, which clearly differed of both strains. Optimizing pectinolytic enzyme production in SSF applying the same substrate and similar process conditions produced two increased protein bands between the protein standards of 46 kD and 58 kD present in the extracts of both strains.

66 Chapter 3

A B C D E F G

175 kD

80 kD

58 kD

46 kD

30 kD

23 kD

Figure 3‎ .6 Analysis of crude extracts obtained under optimized conditions for PG production by SDS- PAGE: lane A: ColorPlus prestained protein marker (New England BioLabs); lane B: commercial pectinase (SIGMA); lane C: Fructozym P (Erbslöh); lane D: A. sojae CBS 100928; lane E: A. sojae ATCC 20235; lane F: Fructozym P (7.5 µg protein); lane G: commercial pectinase (7.5 µg protein).

The protein content of the extracts was also fractionated by native electrophoresis so as to preserve enzyme activity. Results are shown in Figure 3.7‎ (1). The polyacrylamide gel slab was subsequently contacted with a solid agar medium surface that contained polygalacturonic acid sodium salt as substrate. Figure 3.7‎ (2) and Figure 3.7‎ (3) depict the results obtained after treating the substrate plate chemically. It has to be recalled that the active fractions (bands or zones) observed in this case are not directly related to the molecular mass since separation in native gel is related to charge/size. This makes the information gathered in Figure 3.7‎ difficult to compare with the profiles observed in Figure 3.6‎ . Nevertheless, it can be observed that the extracts of the different strains induced a distinctive pattern in the zymogram. The commercially available enzyme preparations from Sigma-Aldrich and Erbslöh Geisenheim AG were produced by Aspergillus spp. and presented different protein patters with several active zones. In contrast to previous findings [12], only one active zone was observed in the extract produced under optimized conditions by A. sojae CBS 100928. Recent studies already discussed the difference in enzyme and secondary metabolite production between SSF and SmF [37, 38]. Based on the results of this study SSF process conditions and media design

67 Chapter 3 significantly influence protein pattern produced in SSF and hence enzyme production.

A B C D A B C D A B C D

1 2 3 Figure 3‎ .7 Native PAGE. (1) Electrophorized gel and (2) substrate containing agar plate after 50 min pre-incubation, (3) substrate containing agar plate after 90 min pre-incubation. Lane A: Fructozym P (Erbslöh); lane B: A. sojae CBS 100928; lane C: A. sojae ATCC 20235; lane D: commercial pectinase (SIGMA).

3.3.4 Comparison of enzyme profiles Since optimization of SSF resulted in similar optimized condition for increased exo- PG production by A. sojae ATCC 20235 and A. sojae CBS 100928, the fermentation conditions were kept constant for a comparison of enzyme production. SSF was performed with 10 g medium, utilizing a mixture of wheat bran and sugar beet pulp in the ratio 70:30 as substrate at a moisture level of 160 % applied by 0.2 M HCl. Substrate was inoculated with the total number of 1 × 107 spores and incubated for 8 days at 30 °C. The obtained extracts were screened for different enzyme activities (Table 3.8‎ ). Comparing the values for exo-PG activity with the results obtained under optimized conditions for A. sojae ATCC 20235, where 2 × 107 spores were applied, the produced enzyme activity decreased by 30 %. While decreasing the inoculum size of A. sojae CBS 100928 increased the produced PG activity by 13 %. Nevertheless, this variation of enzyme activity compared to previous results obtained under optimized conditions might be caused by utilization of sugar beet pulp from another harvest batch. This could imply that the beet root was harvested from another producer

68 Chapter 3 region and also at another season. These factors would strongly effect the composition of beet roots, e.g. by varying the sugar content, and hence the type of inducer substrate. Exo-PMG activity obtained by cultivation of A. sojae ATCC 20235 was higher compared to PMG activity in the extract of A. sojae CBS 100928. In comparison to results of previous work [12], exo-PMG activity in the extract of A. sojae ATCC 20235 slightly decreased by 10 %, while the PMG activity in the extract of A. sojae CBS 100928 increased by 25 %. These changes are minimal in contrast to the change of exo-PG activity, which is confirming that primarily the exo-pectinase activity degrading polygalacturonic acid was increased by this optimization. Furthermore, the obtained extracts of both strains where diluted to 1 U/mL of exo- PG activity to compare the ratio of exo-PG activity and endo-pectinolytic activities. The values given in Table 3.8‎ show an almost balanced ratio of exo- to endo-PG activity for both strains with a slightly higher endo-PG activity. The ratio of exo-PG activity to endo-PMG activity is also balanced in the extract of A. sojae CBS 100928. The endo-PMG activity in the extract of A. sojae ATCC 20235 is slightly lower than the exo-PG activity. This indicates, that besides exo-PG activity also the endo-enzyme activities increased by this optimization in comparison to previous results [12]. Comparing endo-PG with endo-PMG activities the enzyme activity degrading polygalacturonic acid increased more in the extracts of both strains. In addition to pectinolytic activities also the proteolytic activity was determined. Activity of proteases in the extract of A. sojae ATCC 20235 is higher than in the extract of A. sojae CBS 100928. In comparison to previous results, no significant changes in proteolytic activity were observed [12].

69 Chapter 3

Table 3‎ .8 Comparison of enzyme activities obtained under the same cultivation condition

PG activity Specific activity endo-activity (U/mL) at PMG activity Proteolytic activity Microorganism (U/g) (U/mg) 1 U/mL exo-PG activity (U/g) (log mm2) exo exo-PG PG PMG exo 10

A. sojae ATCC 20235 632.7 ± 61.7 133.7 ± 10.2 1.08 ± 0.07 0.87 ± 0.08 30.1 ± 2.4 2.50 ± 0.05

A. sojae CBS 10928 133.2 ± 12.1 54.6 ± 7.7 1.13 ± 0.01 0.99 ± 0.01 20.7 ± 1.9 2.12 ± 0.02

70 Chapter 3

Comparing present results with reported PG activity in the literature, previous optimization studies with A. sojae ATCC 20235 applying crushed maize wetted with a nutrient solution, which resembled a surface cultivation, yielded in 29.1 U/g exo-PG activity [39]. Hence, PG production by A. sojae ATCC 20235 was significantly increased under solid-state conditions in the present study optimizing the SSF process and utilizing a cost-efficient medium. Optimization of PG production in SmF by A. sojae ATCC 20235 resulted in maximum PG activity of 13.5 U/mL using synthetic medium [40]. Utilization of agro-based products in SmF and using a mutant of this strain, generated by treatment with UV radiation, yielded in an enzyme activity of 145.4 U/mL in shaking flask cultures [41]. PG production in SmF on agro-based products by the mutant of A. sojae was also investigated in bioreactor studies, where the maximum PG activity slightly decreased to 120.0 U/mL [41]. The combination of optimizing the fermentation process utilizing agricultural substrates with microbial strain improvement increased dramatically the enzyme yield in SmF, which might be also a promising step for SSF. Optimization of the PG production in SSF utilizing agricultural products developed already an efficient process for enzyme production by the wild type. The combination with microbial strain improvement might generate a process of PG production for industrial applications.

3.4 Conclusions PG production by A. sojae was optimized applying crude plant compounds, such as wheat bran or sugar beet pulp, as substrate in SSF. The present study demonstrated the great potential for cost-efficient pectinolytic enzyme production by Aspergillus sojae ATCC 20235. Production under optimized conditions in laboratory scale yielded in highest reported exo-PG activity produced by filamentous fungi [3]. Utilization of agricultural and agro-industrial by-products makes the optimized process attractive for application in industrial enzyme production. Furthermore, the influence of a pectinase inducer substrate on PG production was demonstrated comparing exo-PG activities obtained with pure wheat bran and a mixture of wheat bran and sugar beet pulp. Sugar beet pulp contains a high content of pectin and was used as inducer substrate for pectinolytic enzyme production [3]. Hence, the significance of SSF process optimization for increased PG production in combination with utilization of agricultural and agro-industrial residues as cost-efficient substrates was presented in this work, as well as importance of a pectinase inducer substrate. Furthermore, a comparison between two fungal strains was provided. Optimization of SSF conditions for A. sojae ATCC 20235 and A. sojae CBS 100928 resulted in similar optimized conditions for increased PG production. However, pectinolytic enzyme

71 Chapter 3 yield was considerably higher by A. sojae ATCC 20235. This property, as well as its GRAS status (generally recognized as safe) favors this fungal strain to be the preferred production organism for industrial and food enzymes. The high enzyme yield obtained in optimized SSF will be great promising starting point for polygalacturonase purification. Furthermore, future studies are aimed at the scale-up operation in combination with strain improvement, which should be helpful in faster commercialization of the process.

References [1] Kashyap DR, Vohra PK, Chopra S, Tewari R. Applications of pectinases in the commercial sector: a review. Bioresource Technology 2001; 77:215-227. [2] Pedrolli DB, Monteiro AC, Gomes E, Carmona EC. Pectin and pectinases: production, characterization and industrial application of microbial pectinolytic enzymes. The Open Biotechnology Journal 2009; 3:9-18. [3] Jacob N. Pectinolytic Enzymes. In: Nigam PSPandey A. Biotechnology for agro- industrial residues utilisation: Springer Netherlands; 2009. 383-396. [4] Nigam PS, Pandey A. Biotechnology for agro-industrial residues utilisation. 2009. [5] Mamma D, Topakas E, Vafiadi C, Christakopoulos P. Biotechnological potential of fruit processing industry residues. In: Nigam PPandey A. Biotechnology for agro-industrial residues utilisation: Springer; 2009. 273-292. [6] Levigne S, Ralet M-C, Thibault J-F. Characterization of pectins extracted from fresh sugar-beet roots under different conditions using experimental design. In: Vorhagen F, Schols HVisser R. Advances in pectin and pectinase research. Dordrecht: Kulwer Academic Publishers; 2003. 419-430. [7] Sun X, Liu Z, Qu Y, Li X. The effects of wheat bran composition on the producion of biomass-hydrolyzing enzymes by Penicillium decumbens. Appl. Biochem. Biotechnol. 2008; 146:119-128. [8] Parekh S, Vinci VA, Strobel RJ. Improvement of microbial strains and fermentation processes. Applied Micriobiology and Biotechnology 2000; 54:287-301. [9] Bas D, Boyaci IH. Modeling and optimization I: Usability of response surface methodology. Journal of Food Engineering 2007;78:836-845. [10] Lazi´c ZR. Design of experiments in chemical engineering. 2004. [11] Eriksson L, Johansson E, Kettaneh-Wold N, Wikström C, Wold S. Design of Experiment. Priciples and Applications. 2008. [12] Heerd D, Yegin S, Tari C, Fernandez-Lahore M. Petinase enzyme-complex production by Aspergillus spp in sold-state fermentation: A comparative study. Food and Bioproducts Processing 2012; 90:102-110.

72 Chapter 3

[13] Ushijima S, Hayashi K, Murakami H. The current taxonomic status of Aspergillus sojae used in Shoyu fermentation. Agric Biol Chem 1982; 46:2365- 2367. [14] Heerikhuisen M, Van den Hondel C, Punt P. Aspergillus sojae. In: Gellissen G. Production of recombinant proteins. Novel microbial and eukaryotic expression systems. Weinheim: WILEY-VCH; 2005. 191 - 214. [15] Bradford MM. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Analytical Biochemistry 1976; 72:248-254. [16] Dubois M, Gilles KA, Hamilton JK, Rebers PA, Smith F. Colorimtric method for determination of sugars and related substances. Analytical Chemistry 1955; 28:350-356. [17] Panda T, Naidu GSN, Sinha J. Multipleresponse analysis of pectinolytic enzymes by Aspergillus niger: a statistical view. Process Biochemistry 1999; 35:187-195. [18] Blandino A, Iqbalsyah T, Pandiella SS, Cantero D, Webb C. Polygalacturonase production by Aspergillus awamori on wheat in solid-state fermentation. Appl. Microbiol. Biotechnol. 2002; 58:164-169. [19] Miller GL. Use of dinitrosalicylic acid reagent for determnation of reducing sugar. Analytical Chemistry 1959; 31:426-428. [20] Patil SR, Dayanand A. Optimization of process for the production of fungal pectinases from deseeded sunflower hesd in submerged and solid-state conditions. Bioresource Technology 2006; 97:2340-2344. [21] Montville TJ. Dual-substrate plate diffusion assay for proteases. Applied Environmental Microbiology 1983; 45:200-204. [22] Laemmli UK. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 1970; 227:680-685. [23] AmershamBiosciencesAB. Protein electrophoresis. Technical manual. 1999. [24] Neuhoff V, Arnold N, Taube D, Ehrhardt W. Improved staining of proteins in polyacrylamide gels including isoelectric focusing gels with clear background at nanogram sensitivity using Coomassie Brillant Blue G-250 and R-250. Electrophoresis 1988; 9:255-262. [25] Manchenko GP. Handbook of Detection of Enzymes on Electrophoresis Gels. 1994. [26] Montgomery DC. Design and analysis of experiments. 2013. [27] Eriksson L, Johansson E, Wikström C. Mixture design - design generation, PLS analysis, and model usage. Chemometrics and Intellgent Laboratory Systems 1998; 43:1-24. [28] Mitchell DA, Targonski Z, Rogalski J, Leonowicz A. Substrates for processes. In: Doelle HW, Mitchell DARolz CE. Solid substrate cultivation. London: Elsevier Applied Science; 1992. 29-52.

73 Chapter 3

[29] Pandey A. Effect of particle size of substrate on enzyme production in solid- state fermentation. Bioresource Technology 1999;37:169-172. [30] Niture SK. Comparative biochemical and structural characterizations of fungal polygalacturonases. Biologia 2008; 63(1):1-19. [31] Hölker U, Höfer M, Lenz J. Biotechnological advanteges of laboratory-scale solid-state fermentation with fungi. Applied Micriobiology and Biotechnology 2004; 64:175-186. [32] Nandakumar MP, Thakur MS, Raghavarao KSMS, Ghildyal NP. Studies on catabolite repression in solid state fermentation for biosynthesis of fungal amylases. Letters in Applied Microbiology 1999 ;29(6):380-384. [33] Viniegra-González G, Favela-Torres E. Why solid-state fermentation seems to be resistant to catabolite repression? Food Technol. Biotechnol. 2006; 44(3):397-406. [34] Raimbault M. General and microbiological aspects of solid substrate fermentation. Electronic Journal of Biotechnology 1998; 1(3):1-15. [35] Fernandez-Lahore MH, Gallego Duaigües MV, Cascone O, Fraile ER. Solid state production of a Mucor bacilliformis acid protease. Revista Argentina de Microbiologia 1997; 29:1-6. [36] Prior BA, Du Preez JC, Rein PW. Environmental parameters. In: Doelle HW, Mitchell DRolz CE. Solid substrate cultivation. Essex, England: Elsevier Science Publishers LTD; 1992. [37] Barrios-González J. Solid-state fermentation: Physiology of solid medium, its molecular basis and applications. Process Biochemistry 2012;47:175-185. [38] Kumar S, Sharma HK, Sarkar BC. Effect of substrate and fermentation conditions on pectinase and cellulase production by Aspergillus niger NCIM 548 in submerged (SmF) and solid state fermentation (SSF). Food Science and Biotechnology 2011; 20(5):1289-1298. [39] Ustok FI, Tari C, Gogus N. Solid-state production of polygalacturonase by Aspergillus sojae ATCC 20235. Journal of Biotechnology 2007; 127:322-334. [40] Tari C, Gögus N, Tokatli F. Optimization of biomass, pellet size and polygalacturonase production by Aspergillus sojae ATCC 20235 using response surface methodology. Enzyme and Microbial Technology 2007; 40:1108-1116. [41] Buyukkileci AO, Tari C, Fernandez-Lahore M. Enhanced production of exo- polygalacturonase from agro-based products by Aspergillus sojae. BioResources 2011; 6(3):3452-3468.

74 Chapter 4

Chapter 4.

Microbial strain improvement for enhanced PG production

Abstract

Strain improvement is a powerful tool in the development of a commercial fermentation process. Strains of Aspergillus sojae which were previously identified as polygalacturonase producers were utilized for random mutagenesis. Therefore, the cost-effective mutagenesis and selection method, the so-called “random screening”, was chosen. Physical (ultraviolet irradiation at 254 nm) and chemical mutagens (N-methyl-N´-nitro-N-nitrosoguanidine) were used in the development and implementation of a classical mutation and selection strategy for the improved production of pectic acid degrading enzymes. Mutagenic procedures induced by the chemical mutagen were optimized by design of experiment in order to obtain 50 - 60 % survival rates. Three mutation cycles of both mutagenic treatments and also the combination of them were performed in order to generate mutants descending from Aspergillus sojae ATCC 20235 and from Aspergillus sojae CBS 100928. Polygalacturonase production of mutants was compared to their wild types in submerged and solid-state fermentation. Highest polygalacturonase activity (1087.2 ± 151.9 U/g) in solid-state culture was obtained by mutant M3, which was 1.7 times increased in comparison to the wild strain Aspergillus sojae ATCC 20235. Additionally, further mutation of mutant M3 for two more cycles of UV treatment generated mutant DH56 with highest polygalacturonase activity (98.9 ± 8.7 U/mL) in submerged culture. This corresponded to 2.4 fold enhanced polygalacturonase production in comparison to the wild strain. Highest polygalacturonase activity (203.7 ± 17.9 U/g) in solid-state culture of a mutant descending from Aspergillus sojae CBS 100928 was obtained after repeated treatment by the chemical mutagen, which corresponded to 1.5 fold increased enzyme production. However, polygalacturonase activity produced by mutant M3, which was generated by three repeated cycles of UV treatment from Aspergillus sojae ATCC 20235, was 5.3 times higher. Subjecting Aspergillus sojae CBS 100928 to three repeated cycles of UV irradiation generated a mutant, which yielded with 19.4 ± 0.1 U/mL in 1.8 enhanced enzyme production in submerged culture. The results of this study indicated the development of a classical mutation and selection strategy as promising tool to improve the enzyme production in few mutation cycles.

75 Chapter 4

4.1 Introduction Nowadays, microbes are routinely used in large-scale processes of fermentation for the commercial production of enzymes such as proteases, cellulases and pectinases. Economics of such processes might be improved by overproduction of natural products, assimilation of inexpensive and complex raw materials or reduction of fermentation time. Microbial strain improvement is one of the methods used to target improvement of fermentation economics. Success in making and keeping a fermentation industry competitive depends greatly on continuous improvement of the production strain [1]. Development of industrial strains is based on changes in the microbial DNA sequence, which is achieved by mutation, genetic recombination, or genetic engineering techniques. Traditional strain improvement programs employ mutagenesis followed by screening or selection. These classical strain improvement methods have historical use. Hence, strains derived from non- recombinant methods or classical strain improvement are widely accepted as less significant process changes, assuming that product specifications are met and regulatory notification is completed [2]. Mutagenesis is generating novel genotypes either unintentionally (spontaneous mutation) or intentionally (induced mutation). Induced mutations are achieved by subjecting the genetic material to physical or chemical agents called mutagens. Conventional mutagens employed for strain improvement include N-methyl-N’-nitro-

N-nitrosoguanidine (NTG), ethyl methanesulfonate (EMS), hydroxylamine (NH2OH), nitrous acid (HNO2) and ultraviolet rays (UV) [2, 3]. Each mutagen includes DNA alterations in a specific manner. A mutagen may also induce more than one type of lesion. Most mutagenic agents cause some damage to the DNA through deletion, addition, transversion, or substitution of bases or breakage of DNA strands. For example, mutation by UV irradiation induces pyrimidine dimerization and cross-links in DNA [2]. A common method used, is treating cells with a mutagen until a certain “desired” kill is obtained. Often high doses of mutagen are applied to kill most microorganisms in order to increase frequency of generated mutants [3-5]. However, the optimum dose of mutagen is that which gives the highest proportion of desirable mutants in the surviving population. There is no simple linear relationship between frequency of mutants and mutagen dosage or survival rate. It was even reported that low doses of mutagen (20 – 50 % survival) yielded in highest mutant frequency for different types of mutants [6]. Furthermore, high mutagen doses can produce chromosome rearrangements and in general disturb the genetic

76 Chapter 4 background by an enhanced load of undesirable mutations especially utilizing recurrent mutagenic treatment [6]. Besides modifying a strain, the identification and selection of improved mutants is an important part in strain improvement. Screening strategies may be divided into two basic types: (a) the non-selective random screening, where randomly picked isolates are tested for desired qualities, and (b) rationalized selection, based on some form of pre-selection [2]. Besides the isolation of a strain with high yield or productivity, the development of a commercially viable process requires also media design and process optimization as well as an efficient product recovery [2, 7]. The production of pectinolytic enzymes is carried out in solid-state (SSF) and submerged fermentation (SmF) systems [8-10]. Filamentous fungi are the major producers of pectinolytic enzymes used in food industry [11, 12]. The koji molds Aspergillus sojae and A. oryzae have a long history of safe use in food industry, which has proved their safety [13, 14]. Previous studies demonstrated already the potential of pectinolytic enzyme production by A. sojae [15, 16]. Based upon a microbial screening for pectinase production in SSF [16], two strains were chosen for exploring the enhancement of their polygalacturonase (PG) production by microbial strain improvement. This study aimed at the enhanced PG production by A. sojae applying classical strain improvement methods. Spores of A. sojae ATCC 20235 and A. sojae CBS 100928 were treated with different mutagens (NTG and UV rays) to generate mutants with increased exo-PG activity. Both mutagens have been already successfully applied for the generation of mutants with enhanced pectinolytic activities [17-20]. Hence, mutation and selection strategies had to be developed and implemented applying the different mutagens. Mutants were generated utilizing repeated and sequential mutagenesis. PG production by generated mutants of both strains was explored under submerged and solid-state conditions.

4.2 Materials and Methods

4.2.1 Materials All chemicals were purchased from AppliChem GmbH (Darmstadt, Germany), except substrates for detection of pectinolytic activities, e.g. polygalacturonic acid, polygalacturonic acid sodium salt and pectin, and mutagen N-methyl-N’-nitro-N- nitrosoguanidine were obtained from Sigma-Aldrich Chemie GmbH (Steinheim, Germany). Microbial substrates like wheat bran, sugar beet pulp pellets and molasses were obtained from local suppliers (Bremer Rolandmühle Erling GmbH &

77 Chapter 4

Co. KG, Bremen, Germany; Nordzucker AG, Uelzen, Germany; Golden Sweet, Meckenheim, Germany).

4.2.2 Microorganism Fungal strains of A. sojae ATCC 20235, purchased from Procochem Inc (Teddington, United Kingdom), and A. sojae CBS 100928, obtained from the Centraalbureau voor Schimmelcultures (CBS) (Utrecht, Netherlands), were propagated on agar plates according to the specifications given in Heerd et al. [16]. Spores from these plates were used as inoculum for molasses agar slants containing: glycerol (45 g/L), molasses (45 g/L), peptone (18 g/L), NaCl (5 g/L), KCl

(0.5 g/L), FeSO4·7H2O (15 mg/L), KH2PO4 (60 mg/L), MgSO4 (50 mg/L), CuSO4·5H2O

(12 mg/L), MnSO4·H2O (15 mg/L) and agar (20 g/L). High amount of spore production was achieved by cultivation on molasses agar slants at 30 °C for 1 week.

4.2.3 Strain improvements Spores of A. sojae were treated with UV radiation or NTG to obtain PG-hyper- producing mutants, as described below. Key steps of the mutagenesis and screening procedure are presented in Figure 4.1‎ . After mutagenic treatment, spore suspensions were diluted stepwise to 1×103 spores/mL with phosphate buffered saline (PBS) buffer (pH 7.4) in order to achieve the final spore concentration for plating. From these dilutions, 100 µL were inoculated into Petri dishes containing Yeast Malt Extract (YME) agar medium, composed of malt extract (10 g/L), yeast extract (4 g/L), glucose (4 g/L) and agar (20 g/L), and incubated at 30 °C for 48 h, followed by colony count and expressed as survival rate (%). Applying this amount of diluted spore suspension resulted in growth of well countable colonies on agar plates. Untreated spore suspensions were used for comparison to determine 100 % survival rate. Colonies were further incubated at 30 °C and mutants were morphologically selected for the presence of sporulation. The isolation of mutants consisted in the isolation of colonies originated from single conidia. Spores from single colonies were replica-plated on agar plates containing screening medium (section 4.2.4‎ ) using sterile wood toothpicks. Based upon the size of clearing zones surrounding the colonies observed after 2 – 3 days growth at 30 °C, “zone mutants” showing increased pectinase activity on screening medium were selected and were further screened for their PG production in SSF. The best mutant strain was selected for further mutation treatment.

78 Chapter 4

Spore suspension

Mutagenesis

Colony isolation Plate screening

Selection of strains with increased activity

Screening for PG production in SSF

Selection of strains with increased PG activity Slant culture

Figure 4‎ .1 Key steps in the progression of culture screening and improvement.

4.2.3.1 UV-Mutagenesis Mutagenesis was performed using a modification of the method given by De Nicolás-Santiago et al. [21]. Spores produced on molasses slants were harvested using 0.02 % (v/v) tween 80 solution and their concentration was adjusted to 1×107 spores/mL adding PBS buffer (pH 7.4). Five mL of spore suspension were poured into a sterile small Petri dish (Ø 35 mm). Spores were treated with UV radiation under agitation using a magnetic bar, during the whole procedure. The UV radiation source, a 254-nm-wavelength germicidal lamp, was placed 17 cm above the spore suspension. Mutation was performed under semi-darkened conditions, transmitting UV light in a box containing the spores. Exposure time to UV radiation was for 5 min for mutation of both strains due to adequate inactivation of spores during this time (section 4.3).‎ Directly after UV treatment, samples of 100 µL were taken and were 10 times diluted with PBS buffer (pH 7.4). Diluted samples were first placed on ice in the dark for 5 min, followed by 30 min incubation in darkness at room temperature to avoid photoreactivation repair. After incubation, spore suspensions were further diluted stepwise to 1×103 spores/mL with PBS buffer (pH 7.4) in order to achieve the final spore concentration for plating. Further proceeding of spore treatment was done as described in section 4.2.3.‎

79 Chapter 4

4.2.3.2 Mutagenesis by N-methyl-N’-nitro-N-nitrosoguanidine (NTG)

NTG is an N-nitrosamine alkylating agent, NH existing as yellow crystal at room temperature. O2N C CH3 It is soluble in water, dimethyl sulfoxide, and N N H organic solvents [22]. NTG is heat, light and N moisture sensitive. Hence, in order to keep O stability during experimental set up a stock Figure 4‎ .2 Chemical structure of solution was prepared and stored in aliquots as N-methyl-N’-nitro-N-nitrosoguanidine described in Appendix A. (C2H5N5O3, CAS No. 70-25-7) [22]. Spores were harvested from molasses slants using sterilized distilled water and diluted with PBS buffer (pH 7.4) to 5×107 spores/mL. Treatment of spores by NTG was performed in dark Eppendorf test tubes. NTG stock solution of 0.3 % was added into the tubes and mixed with PBS buffer and spore suspension to give the desired concentration. Spores were added to the final concentration of treated spores of 1×107 spores/mL in all experiments. Procedure for treatment with NTG was optimized to obtain a survival rate of 50 – 60 % (section 4.3.1).‎ Mutation was performed in Thermomixer R (Eppendorf) at 600 rpm. After incubation, spore suspensions were further diluted stepwise to 1×103 spores/mL with PBS buffer in order to achieve the final spore concentration for plating. Further proceeding of spore treatment was done as described in section 4.2.3.‎

4.2.3.3 Sequential and repeated mutagenesis Three strategies of mutagenesis were employed for strain improvement of A. sojae ATCC 20235 and A. sojae CBS 100928 (Figure 4.3‎ ). In all cases mutants were generated in three cycles of mutation, except the generated mutant M3 descending from A. sojae ATCC 20235 was further exposed to two more cycles of UV irradiation. After generating all mutants of A. sojae ATCC 20235 and A. sojae CBS 100928 PG production of mutants in comparison to the wild types was evaluated in SmF and SSF. Sugar beet pellets were used as inducer substrate for enzyme production due to their significant effect on enhanced PG production (chapter 3).

80 Chapter 4

Aspergillus sojae

U.V. NTG 1st mutation cycle

U.V. U.V. NTG 2nd mutation cycle

U.V. NTG. NTG 3rd mutation cycle Mutant M3 Mutant II-8 Mutant 2b-18 Mutant 60 Mutant 2b-12 Mutant 3b-4 Mutant 84

U.V. 4th mutation cycle

U.V. 5th mutation cycle Mutant DH56

1st Method 3rd Method 2nd Method

Figure 4‎ .3 Scheme of repeated and sequential mutagenesis for improvement of PG production by A. sojae including generated mutants descending from A. sojae ATCC 20235 and mutants descending from A. sojae CBS 100928.

In the first method, spores of A. sojae were repeatedly exposed for three cycles to UV irradiation as described in section 4.2.3.1.‎ Based upon morphological appearance of spores on YME agar medium, clearing zones surrounding the colonies on screening medium and finally PG production in SSF, mutants which were potential over-producers of PG were selected. Mutant M3 descending from A. sojae ATCC 20235 was treated by UV radiation for two more cycles and the mutant A. sojae DH56 was generated. During the second procedure, spores of A. sojae were repeatedly treated with three cycles NTG solution under optimized conditions for each strain (section 4.3.1)‎ as described in section 4.2.3.2‎ . The third Method combined NTG treatment and UV irradiation. Selected fungal strain obtained after the first mutation treatment of the second method was further exposed to UV radiation in the second cycle. Generated mutants with increased PG activity were subsequently treated for another cycle with NTG solution.

81 Chapter 4

4.2.4 Screening medium Screening medium was used for the selection of pectinase over-producing mutants of Aspergillus spp. The media was a modification of the selection media described by Durrands & Cooper [23]. 250 mL of 0.1 M acetate buffer (pH 5.0) containing 1 g

NaNO3, 0.5 g KH2PO4, 0.25 g MgSO4·7 H2O and 7.5 g agar was sterilized by autoclaving. Warm solution was blended with equal volume of separately sterilized solution of 0.5 % (w/v) polygalacturonic acid sodium salt, and poured into Petri dishes. Pectinase activity was detected as clear zone around colonies in the background of the precipitated substrate after treating plates with 1 % (w/v) cetyltrimethyl ammonium bromide solution.

4.2.5 Solid-state fermentation (SSF) SSF was performed in culture flasks (250 mL) under optimized SSF conditions for the wild types (chapter 3), utilizing 10 g of wheat bran and grinded sugar beet pellets in the ratio 70:30, wetted at 160% with 0.2 N HCl solution (sterilized at 121 °C for 20 min). The inoculum was adjusted to 1×106 spores per gram dry weight, which corresponds to 3.8×105 spores per gram moist mass of medium. Flasks were incubated at 30°C for 8 days and shaken manually twice at the day of inoculation and the 1st cultivation day. Enzyme leaching was performed with 80 mL tap water for 20 min at 250 rpm agitation and 25 °C. Mycelium and solid medium were separated by filtration through cotton cheese cloth and the filtrate was centrifuged at 4 °C, 3220 × g, for 20 min. Enzyme activities (section 4.2.8‎ and section 4.2.9)‎ and total protein content (section 4.2.7)‎ were determined in the supernatant.

4.2.6 Submerged fermentation (SmF) Fermentation was carried out in 250-mL Erlenmeyer flasks containing 30 mL of medium (sterilized at 121 °C for 20 min) at 30 °C and 250 rpm agitation for 4 days. Medium K was designed for enhanced PG production in SmF (section 4.3.2).‎ Culture flasks were inoculated with 3.8×105 spores per mL of medium. Enzyme activities were determined in the supernatant obtained after centrifugation of the fermentation broth at 4 °C, 3220 × g, for 20 min.

4.2.7 Total protein determination Soluble protein content was determined according to the modified Bradford’s method [24] as described in Appendix A.

82 Chapter 4

4.2.8 Exo-pectinolytic activity measurement

4.2.8.1 Polygalacturonase assay PG activity was assayed according to the procedure of Panda et al. [25], which was further optimized as described in Appendix A.

4.2.8.2 Polymethylgalacturonase assay PMG activity was determined according to the method provided by Panda et al. [25] with slight modifications, using 0.5 g/L pectin as substrate dissolved in 0.1 M acetate buffer (pH 5.0). Crude extract (0.086 mL) containing enzyme was added to 0.4 mL substrate and incubated for 10 min at 40 °C. The reducing sugar released was measured using the Nelson-Somogyi method [26] as adapted by Panda et al. [25]. First reaction was terminated by adding 0.5 mL copper reagent and placing the mixture in boiling water for 10 min. After cooling down, 1 mL of arsenomolybdate reagent was added, followed by intensive vortexing and centrifugation at 3220 × g at 22 °C for 5 min. The absorbance of the supernatant was read on a spectrophotometer at 500 nm. Blanks were in-cooperated containing all the reagents and the enzyme, but the enzyme was not allowed to react with the substrate. Standard solutions of galacturonic acid were used for calibration. One unit of exo-PMG activity was defined as the amount of enzyme that catalyzes the release of 1 µmol of galacturonic acid per unit volume of supernatant per unit time under standard assay conditions mentioned above.

4.2.9 Endo-pectinolytic activity measurement Endo-enzyme activities were determined by measuring the decrease in viscosity of a substrate solution, either 2 % (w/v) pectin for endo-PMG or 3.2 % (w/v) polygalacturonic acid (sodium salt) for endo-PG. Reduction in viscosity was determined according to a modified method of Mill & Tuttobello [27], utilizing a graduated glass pipette as viscometer.

4.2.9.1 Endo-pectinase activity in crude extracts Pectinolytic activity was assayed by adding 0.2 mL of an appropriate diluted enzyme sample, to 0.2 mL of 0.2 M acetate buffer (pH 5.0) and 1.6 mL substrate. The mixture was incubated in a water bath for 1 h at 40°C. After incubation, the mixture was cooled down for 30 sec in ice-cold water and viscosity of the samples was determined. The later was done indirectly by measuring the time required for 0.9 mL of reaction mixture to elute through a 1.0 mL glass pipette. Samples were measured in duplicate. Viscosity was calculated from a calibration curve obtained by time measurements to pass polyvinylpyrrolidone (PVP) 360 aqueous standard

83 Chapter 4 solutions through the pipette at 25 °C, which were previously passed through an Ostwald viscometer at 25 C. Controls for non-enzymatically treated substrate solutions were included, utilizing 0.2 mL water instead of enzyme samples. One unit of endo-PG activity was defined according to Patil & Dayanand [8], as the quantity of enzyme which caused a 50% reduction in viscosity of the reaction mixture per minute, under the assay conditions described above.

4.2.9.2 Endo-pectinolytic activity at 1 U/mL of exo-PG activity Pectinolytic activity was assayed by adding 0.2 mL of enzyme sample containing 1 U/mL exo-PG activity, to 0.2 mL of 0.2 M acetate buffer (pH 5.0) and 1.6 mL substrate. The mixture was incubated in a water bath for 5 min at 40°C. After incubation, the mixture was cooled down for 30 sec in ice-cold water and viscosity of the samples was determined. The later was done indirectly following the procedure described in section 4.2.9.1.‎ According to Patil & Dayanand [8], one unit of endo-PG activity was defined as the quantity of enzyme which caused a 50% reduction in viscosity of the reaction mixture per minute, under these assay conditions mentioned above.

4.3 Results and Discussion Mutants of A. sojae were generated by repeated and sequential mutagenesis applying the mutagens NTG and UV light. The exposure time to UV radiation under the described conditions was previously explored by taking samples at different time intervals. Applying an exposure time of 5 min resulted in a strong decrease of survival rate for spores of A. sojae ATCC 20235 and also a high inactivation of A. sojae CBS 100928 spores. The percent of mortality as compared to control (untreated) was calculated by serial dilution plating. Due to the high variation in survival rate (10 – 50 %) achieved by several repetitions it was not possible to determine the exact lethality rate reached after 5 min treatment. Nevertheless, following this procedure a sufficient amount of spores from both strains was inactivated. Hence exposure time to UV radiation was fixed at 5 min following the procedure described in section 4.2.3.1.‎ In addition to UV treatment, also a procedure for the exposure to the chemical agent NTG had to be established. In order to achieve a survival rate of 50 – 60 % for both strains the method for mutagenesis by NTG was optimized by applying statistical tools (section 4.3.1).‎ Both mutagenic treatments were successfully applied in repeated and sequential mutagenesis for fungal strain improvement. After generating all mutants of A. sojae ATCC 20235 and A. sojae CBS 100928 PG production of mutants in comparison to the wild types was evaluated in SmF and SSF.

84 Chapter 4

4.3.1 Optimization of mutagenesis by NTG The investigation to optimize mutagenesis by NTG comprised of defining the survival rate between 50 – 60 %. This rate was chosen since a low mutagen dose should be considered taking the disturbance of the genetic background into account [6]. Optimization procedure was performed in two steps, comprising a screening of factors and optimization of their settings in the second step. All experiments planned according to Design of Experiment (DoE) were evaluated using Multiple Linear Regression (MLR) utilizing the MODDE 9.0 software package, supplied by Umetrics AB, Umeå, Sweden. In the first step screening experiments were planned according to DoE as a fractional factorial design with three factors. The effect of NTG concentration (0.005 – 0.05 %), incubation time (0.5 – 3 h) and temperature (16 – 30 °C) on the survival rate during mutagenesis by NTG was studied. This resulted in 14 experiments including a replicate and three center points (Table 4.1‎ ).

Table 4‎ .1 Factors and responses of screening results applying fractional factorial design.

Factors Response: survival rate (%) Exp. Temperature A. sojae A. sojae No NTG (%) Time (h) (°C) ATCC 200235 CBS 100928

1 0.005 0.5 30 100 60

2 0.05 0.5 16 46 44

3 0.005 3 16 89 60

4 0.05 3 30 16 2

5 0.0275 1.75 23 64 44

6 0.0275 1.75 23 73 44

7 0.0275 1.75 23 70 39

8 0.005 0.5 30 100 62

9 0.05 0.5 16 66 36

10 0.005 3 16 100 85

11 0.05 3 30 10 3

12 – 14* 0.0275 1.75 23 72.7 ± 4.7 41.0 ± 2.0 * Three repetitions of the center point with a standard variation below 7 %.

The responses for survival rate of both strains were modeled using MLR. The model quality was sound for the survival rate of spores from A. sojae ATCC 20235 (R2/Q2

85 Chapter 4

0.95/0.88) as well as for spores of A. sojae CBS 100928 (R2/Q2 0.94/0.80). The linear models showed no lack of fit (LoF) (A. sojae ATCC 20235: p = 0.205; A. sojae CBS 100928: p = 0.588), because their p-values were larger than the critical reference value of 0.05, and normally distributed residuals (plot not shown). Hence the models had small model error and good fitting power. All three factors had significant effect on the survival rate of spores from both strains. Looking at the sweet spot plots of the screening step (Figure 4.4‎ and Figure 4‎ .5), it is clear that the desired survival rate of 50 – 60 % was achieved for both strains in the chosen factor range.

Sweet spot = 50 – 60 % survival rate Temperature: 16 C Temperature: 23 C Temperature: 30 C

0.045 0.045 0.045 0.04 0.04 0.04 0.035 0.035 0.035 0.03 0.03 0.03 0.025 0.025 0.025 0.02 0.02 0.02

0.015 0.015 0.015

NTG concentration (%)concentrationNTG (%)concentrationNTG NTG concentration (%)concentrationNTG 0.01 0.01 0.01 0.005 0.005 0.005 0.5 1 1.5 2 2.5 0.5 1 1.5 2 2.5 0.5 1 1.5 2 2.5 Time (h) Time (h) Time (h) Figure 4‎ .4 Sweet spot plot of desired survival rate of A. sojae ATCC 20235 in screening step. The green zone represents the conditions yielding in 50 – 60 % survival rate.

Sweet spot = 50 – 60 % survival rate Temperature: 16 C Temperature: 23 C Temperature: 30 C

0.045 0.045 0.045 0.04 0.04 0.04 0.035 0.035 0.035 0.03 0.03 0.03 0.025 0.025 0.025 0.02 0.02 0.02

0.015 0.015 0.015

NTG concentration (%)concentrationNTG (%)concentrationNTG NTG concentration (%)concentrationNTG 0.01 0.01 0.01 0.005 0.005 0.005 0.5 1 1.5 2 2.5 0.5 1 1.5 2 2.5 0.5 1 1.5 2 2.5 Time (h) Time (h) Time (h) Figure 4‎ .5 Sweet spot plot of desired survival rate of A. sojae CBS 100928 in screening step. The green zone represents the conditions yielding in a survival rate of 50 – 60 %.

Nevertheless, spores of A. sojae ATCC 20235 seemed to be more resistant to the mutagenic treatment comparing the survival rates of both strains under the same conditions. Compared to A. sojae CBS 100928 a higher NTG concentration, longer

86 Chapter 4 incubation time or higher incubation temperature is needed to obtain the survival rate of 50 – 60 %. There is a stronger influence of temperature for the mutagenesis of A. sojae CBS 100928. Therefore, the temperature range was decreased for the optimization step using this fungus. Furthermore, NTG concentration had to be increased for the optimization experiments at lower temperature to get a survival rate of 50 – 60 % for A. sojae ATCC 20235. Hence, the settings of all three factors were optimized independently for each strain. Respective factor ranges for optimization experiments are given in Table 4.2‎ . The experimental setup comprised of 34 experiments including a replicate and three center points for each strain using central composite face-centered (CCF) design.

Table 4‎ .2 The investigated factors and their levels during optimization experiments for both strains.

Actual factor levels Microorganism Factor (unit) -1 0 + 1

NTG concentration (%) 0.01 0.055 0.1

A. sojae ATCC 20235 Incubation time (h) 0.5 1.75 3

Temperature (°C) 15 25 35

NTG concentration (%) 0.005 0.0275 0.05

A. sojae CBS 100928 Incubation time (h) 0.2 1.6 3

Temperature (°C) 14 22 30

Looking at the replicate plots of the optimization experiments (Figure 4.6‎ and Figure 4.7‎ ), it is clear that the replicates of each optimization investigation obtained equal values, indicating a good reproducibility of the mutation experiments. A stronger variation was obtained only in few cases, e.g. experiment no. 9 and 26 for survival rate of A. sojae ATCC 20235 (Figure 4.6‎ ).

87 Chapter 4

1 522 100 1 522 9 18 19 100 28 3 9 30 80 18 19 2 28 13 3 30 80 16 2 11 13 60 26 16 6 11 32 60 24 26 17 6 343332 40 23 15 Survival rate Survival (%) 24 17 27 3433 40 23 7 10 3114 15

Survival rate Survival (%) 421 27 20 7 10 3114 421 20 825 2912 0 825 2912 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Experiment number Experiment number Figure 4‎ .6 Plot of replications for survival rate of A. sojae ATCC 20235 with experimentMODDE 9 - 2013-02-08 17:36:58 number (UTC+1) labels during optimization. MODDE 9 - 2013-02-08 17:36:58 (UTC+1)

118 320 5 30 100 118 320 5 9 30 100 22 9 22 26 80 13 26 32 80 19 421 28 13 21 7 11 3215 192 4 28 1733 60 7 11 15 2 1733 60 6 27 1634 24 12 14 6 2710 1634 40 24 1229 1431

Survival rate Survival (%) 10 40 23 29 31 Survival rate Survival (%) 23 20 20 825 0 825 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 1 2 3 4 5 6 Experiment7 8 number9 10 11 12 13 14 15

Figure 4‎ .7 Plot of replications for survival rateExperiment of A. sojae number CBS 100928 with experimentMODDE 9 - 2013-02-08 17:44:00 number (UTC+1) labels during optimization. MODDE 9 - 2013-02-08 17:44:00 (UTC+1)

Furthermore, it can be seen from the replicate plots that experimental conditions close to the center points (last stick) favor the achievement of the desired survival rate of 50 – 60 %. Evaluation of the data of A. sojae ATCC 20235 utilizing the optimizer of MODDE 9.0 for prediction of the optimized region to obtain the desired survival rate resulted in a clear identification of the settings for NTG concentration and temperature with 0.1 % and 20 °C, respectively. Therefore, the NTG concentration and temperature was fixed at these settings during the validation experiments and the factor incubation time was tested at the suggested values of 1 h and 1.25 h. Validation experiments were performed in triplicate and each trial was plated in duplicate to determine the survival rate. The incubation time of 1 h using 0.1 % NTG at 20 °C yielded in a survival rate of 64 ± 7 %. Increasing the time to 1.25 h resulted in 52 ± 4 % survival rate under the same conditions. Hence, the optimal conditions for mutagenesis by NTG to obtain a survival rate of 50 – 60 %

88 Chapter 4 were fixed at 0.1 % NTG concentration at 20 °C for 1.25 h incubation time. These settings were used for the mutagenesis procedure of A. sojae ATCC 20235. Evaluation of data from A. sojae CBS 100928 did not result in a clear identification of factor settings as compared to A. sojae ATCC 20235. Contour plots presented in Figure 4.8‎ show the effect of NTG concentration and incubation time at fixed temperature levels on the survival rate of A. sojae CBS 100928 spores. According to the graph, the desired survival rate of 50 – 60 % can be obtained at all temperature levels applying diverse factor settings for NTG concentration and incubation time. The lower the temperature, the more harsh conditions for NTG concentration and time is needed to achieve the desired survival rate.

Temperature: 14 C Temperature: 22 C Temperature: 30 C Survival rate (%) 0.045 0.045 0.045

0.04 0.04 0.04 0.035 0.035 0.035

0.03 0.03 0.03

0.025 0.025 0.025

0.02 0.02 0.02

NTG concentrationNTG (%) concentrationNTG (%) NTG concentrationNTG (%) 0.015 0.015 0.015

0.01 0.01 0.01

0.005 0.005 0.005 0.5 1 1.5 2 2.5 0.5 1 1.5 2 2.5 0.5 1 1.5 2 2.5 Time (h) Time (h) Time (h) Figure 4‎ .8 Contour plots presenting the survival rate of A. sojae CBS 100928 spores at the fixed temperatures applying the low, medium and high temperature level of the optimization investigation.

Predicting a survival rate of 50 – 60 % utilizing the optimizer of MODDE 9.0 resulted in several experiments at wide ranges of factor settings for temperature (14 – 30 °C) and incubation time (0.7 – 1.6 h). Validation experiments were conducted at the suggested optimal settings, which are given as run I - III in Table 4.3‎ . Since the desired survival rate of 50 – 60 % was not obtained under these conditions and the strong influence of temperature on the survival rate of A. sojae CBS 100928 was known from previous screening experiments (see above), further validations experiments were added with increased temperature and adapted incubation times (run IV – VI in Table 4.3‎ ). The settings of mutagenesis parameters of run IV resulted in almost 50 % survival rate. Hence, procedure of mutagenesis of A. sojae CBS 100928 was established at 0.04 % NTG concentration at 25 °C for 0.7 h. Comparing the settings for mutagenic procedure of A. sojae CBS 100928 to the those of A. sojae ATCC 20235 confirmed previous findings, that a conspicuous lower mutagenic dose is required to obtain the same survival rate.

89 Chapter 4

Table 4‎ .3 Validation experiments and results at the optimal point for 50 – 60 % survival rate of A. sojae CBS 100928.

NTG concentration Temperature Incubation time Survival rate Run (%) (°C) (h) (%)

I 0.04 30 0.7 37 ± 3

II 0.03 22 1.6 63 ± 3

III 0.05 14 0.76 67 ± 1

IV 0.04 25 0.7 49 ± 1

V 0.03 30 0.8 48 ± 10

VI 0.05 25 0.6 46 ± 4

4.3.2 Media design for SmF PG production by A. sojae ATCC 20235 in SmF was already optimized by Ustok et al. [15], utilizing a face-centered central composite design for addition of maltrin and corn steep liquor to a basal medium consisting of glucose (25 g/L), peptone (2.5 g/L), disodium phosphate (3.2 g/L) and monosodium phosphate (3.3 g/L). According to their results, highest PG activity of 8.4 U/mL was obtained at high maltrin concentration (75 g/L) and low corn steep liquor concentration (2.5 g/L) with an inoculum of 2.5 × 105 total spore at 300 rpm agitation and cultivation at 30 °C for 96 h. PG production in SmF was further optimized applying statistical tools, utilizing mutant DH56, which was called mutant M5/6 [28]. An optimized medium composition of dried orange peel (10 g/L), sugar beet syrup (60 g/L) and (NH4)2SO4 (8 g/L) resulted in highest exo-PG activity of 145 U/mL in shake flask cultures. Based on these findings a basic media screening was performed utilizing sugar beet pulp as inducer substrate in order to use the same pectinase inducer substrate like under solid-state conditions (section 4.2.5).‎ Galiotou-Panayotou et al. [29] demonstrated already the inducing effect of sugar beet for PG production by Aspergillus sp. under submerged conditions.

Following the findings of Buyukkileci et al. [28] (NH4)2SO4 was utilized as nitrogen source. The media screening was performed in 100-mL culture flasks utilizing 20 mL of medium per flask. Flasks were inoculated with 4 × 105 total spore of wild strain A. sojae ATCC 20235 and incubated at 30 °C and 250 rpm agitation for 96 h. Exo-PG activity was determined in the supernatant obtained after centrifugation of the

90 Chapter 4 crude broth at 4 °C and 6,000 × g for 30 min. The media screening was performed in 3 steps. All combinations of applied media designs are presented in Table 4.4‎ together with the respective exo-PG activity. First, media A – C were utilized for PG production. Highest enzyme activity was produced with medium C, applying sugar beet syrup as carbon source. According to Buyukkileci et al. [28], highest PG activity was obtained at low pH values, and phosphates in the medium composition had an insignificant effect on PG activity and have been excluded. Based on these findings, the pH was decreased to 4.0 by addition of diluted HCl in the second step of this media design (media D – G). Comparing medium E and F, high enzyme activity was also obtained without phosphates. Comparing medium E and G, decrease in ammonium sulfate concentration decreased also enzyme activity. Increase of sugar beet syrup concentration (medium D) also decreased PG activity. In the third step, media H – K were screened for pectinase production applying an increased ammonium sulfate concentration and a decreased sugar beet syrup concentration. This resulted in an increased enzyme activity, comparing medium E with media I. Further increase of ammonium sulfate concentration (medium H) influenced PG activity insignificantly. Increased inducer substrate concentration influenced enzyme production negatively (medium J). Decrease of sugar beet pulp concentration significantly increased PG activity to 64.2 ± 2.6 U/mL (medium K). Applying the same inducer substrate used under solid-state conditions, unmolassed sugar beet pulp, slightly decreased PG activity by 13 % (data not shown). Nevertheless, for the comparison of PG production in SmF and SSF the same inducer substrate was applied and molassed sugar beet pulp was replaced by unmolassed sugar beet pulp in the medium composition of medium K. This simple media screening remarkable increased PG production by A. sojae ATCC 20235 in comparison to previous findings of Ustok et al. [15]. Hence the medium used for comparison of mutants and wild type in SmF contained sugar beet pulp (30 g/L), ammonium sulfate (10 g/L) and sugar beet syrup (95 g/L). The pH of this medium was adjusted to 4.0 by adding diluted HCl before autoclaving.

Table 4‎ .4 Media design for increased PG production in SmF.

Ingredients (g/L) exo-PG activity (U/mL) Medium Sugar beet pulp 34

(NH4)2SO4 8

A Na2HPO4 3.2 10.1

NaH2PO4 3.3 Maltrin 142

91 Chapter 4

Molassed sugar beet pulp 70

(NH4)2SO4 6

B Na2HPO4 3.2 10.7

NaH2PO4 3.3 Maltrin 70 Molassed sugar beet pulp 70

(NH4)2SO4 6

C Na2HPO4 3.2 14.7

NaH2PO4 3.3 Sugar beet syrup 70 Molassed sugar beet pulp 50 * D (NH4)2SO4 6 7.9 Sugar beet syrup 150 Molassed sugar beet pulp 40 * E (NH4)2SO4 8 21.3 Sugar beet syrup 120 Molassed sugar beet pulp 40

(NH4)2SO4 8 * F Na2HPO4 3.2 15.6

NaH2PO4 3.3 Sugar beet syrup 120 Molassed sugar beet pulp 40 * G (NH4)2SO4 6 13.0 Sugar beet syrup 120 Molassed sugar beet pulp 40 * H (NH4)2SO4 12 29.7 Sugar beet syrup 95 Molassed sugar beet pulp 40 * I (NH4)2SO4 10 28.3 Sugar beet syrup 95 Molassed sugar beet pulp 60 * J (NH4)2SO4 10 7.9 Sugar beet syrup 95 Molassed sugar beet pulp 30 * K (NH4)2SO4 10 64.2 Sugar beet syrup 95 * pH of medium C was adjusted to 4.0 by adding diluted HCl before autoclaving

4.3.3 Comparison of mutants for pectinase production in SmF and SSF The wild strains and selected mutants descending from A. sojae ATCC 20235 and A. sojae CBS 100928 were evaluated for PG production under solid-state and submerged conditions. The medium for solid-state cultivation was previously optimized for increased PG production by the wild types A. sojae ATCC 20235 and

92 Chapter 4

A. sojae CBS 100928 (chapter 3). Medium for submerged fermentation was also designed for enhanced PG production by A. sojae (section 4.3.2).‎ The aim of this study was to compare pectinase production of generated mutants and wild strains in SmF and SSF (Table 4.5‎ ). PG activities produced by the wild strains in SmF and SSF were taken as 100 % for the comparison to the generated mutants. Highest PG activity of 1087.2 ± 151.9 U/g in SSF was obtained by mutant M3 descending from A. sojae ATCC 20235. This corresponded to a 1.7 fold increase of enzyme activity obtained after 3 cycles UV irradiation. Regarding PG activity, mutant M3 produced also highest specific activity in SSF, which was 2.1 fold increased compared to the wild strain. Mutation by UV irradiation slightly enhanced exo-PMG activity by 12 %, too. Further UV treatment of this mutant by two more cycles generated mutant DH56 with increased pectinase activity on screening medium agar plate, but decreased PG activity in SSF. The high protein content produced by this mutant in SSF is conspicuous in contrast to the low PG activity. Exo-enzyme activity degrading pectin is also low compared to the wild strain, which indicated that more proteins, which are not involved in the degradation of pectic substances, were secreted by mutant DH56 in SSF. The results produced by mutant 2b-18, which was generated by 3 cycles of treatment with NTG, were not reproducible. Furthermore, PG activity decreased by each new generation cycle, indicating that the generated mutant was not stable. Hence, mutant 2b-18 was excluded from further considerations. Increased pectinolytic activities of mutants generated by a combination of NTG treatment and UV irradiation were stable over repeated generation cycles. Mutant II-8 showed 1.4 fold increased PG activity in SSF, while mutant 2b-12 achieved a similar PG activity in SSF like the wild strain. Their specific PG activities were higher compared to the wild type. Looking at their exo-PMG activities, these mutants produced even higher pectin degrading activities than mutant M3. Highest exo-PMG activity in SSF was obtained by mutant 2b-12, which was 21 % increased compared to the wild type. Hence, a combination of these two mutagenic treatments in sequential mutagenesis might be promising for the generation of mutants producing various pectinolytic enzyme activities.

93 Chapter 4

Table 4‎ .5 Comparison of mutant and wild types for the production of PG in SSF and SmF.

Solid-state culture Submerged culture Cycle No./ MO mutagen/ exo-PG Specific activity Total protein exo-PMG exo-PG Specific activity Total protein exo-PMG mutant code (U/g) (%) (U/mg) (mg/g) (U/g) (U/mL) (%) (U/mg) (mg/mL) (U/mL)

wild type 632.7 ± 61.7 100 133.7 ± 10.2 4.72 ± 0.10 18.5 ± 1.6 40.4 ± 3.8 100 80.4 ± 1.2 0.50 ± 0.02 7.1 ± 0.4

3/ UV/ M3 1087.2 ± 151.9 172 284.0 ± 75.6 3.95 ± 0.44 20.8 ± 0.8 38.7 ± 6.5 96 84.6 ± 12.9 0.46 ± 0.01 7.1 ± 0.5

5/ UV/ DH56 16.1 ± 1.4 3 2.1 ± 0.1 7.58 ± 0.43 3.1 ± 0.8 98.8 ± 8.7 245 254.5 ± 18.1 0.39 ± 0.01 8.8 ± 0.2

ATCC20235

3/ NTG & UV/ II-8 890.3 ± 22.3 141 170.1 ± 19.4 5.31 ± 0.68 22.2 ± 1.5 24.2 ± 1.2 60 62.7 ± 1.0 0.39 ± 0.01 6.7 ± 0.3

A. sojae A. 3/ NTG & UV/ 2b-12 651.9 ± 47.4 103 144.3 ± 19.7 4.59 ± 0.61 22.4 ± 1.6 31.6 ± 0.4 78 62.4 ± 1.3 0.51 ±0.01 6.6 ± 0.2

3/ NTG/ 2b-18 348.2 ± 227.4 55 66.8 ± 46.7 5.48 ± 0.39 21.3 ± 1.1 41.2 ± 2.5 102 82.2 ± 7.6 0.50 ± 0.02 7.1 ± 0.2

wild type 133.2 ± 12.1 100 54.6 ± 7.7 2.46 ± 0.14 12.5 ± 0.8 10.6 ± 1.2 100 35.8 ± 3.4 0.30 ± 0.01 1.9 ± 0.2

3/ NTG/ 3b-4 203.7 ± 17.9 153 107.7 ± 18.4 1.92 ± 0.15 13.0 ± 0.6 6.6 ± 1.1 62 21.2 ± 1.2 0.31 ± 0.03 2.0 ± 0.7

CBS100928

3/UV/ 60 57.7 ± 6.9 43 30.0 ± 0.6 1.93 ± 0.23 10.9 ± 1.1 13.7* 129* 29.1* 0.47* 1.53*

A. sojae A. 3/ UV/ 84 60.2 ± 8.0 45 29.1 ± 1.6 2.07 ± 0.21 11.8 ± 0.9 19.4 ± 0.1 183 38.5 ± 2.5 0.51 ± 0.03 2.5 ± 0.2

* Cultivation of mutant UV-60 in SmF created the formation of two morphologies. The formation of a big clump was observed twice and once mutant UV-60 was grown as fine mycelia. The clump structure resulted in a decreased PG activity of 1.5 ± 0.4 U/mL compared to the fine mycelia. Since the observed morphology of the other mutants and wild types was closer to the mycelia structure, only the value obtained by mutant UV-60 growing as fine mycelia was considered.

94 Chapter 4

Evaluation of mutants descending from A. sojae ATCC 20235 in SmF gave reverse results than in SSF. Mutants generated after three cycles UV irradiation or a combination of NTG and UV treatment produced lower PG activity in SmF than the wild strain. This might be explained by the screening method used to select mutants with enhanced PG activity (section 4.2.3).‎ Since the focus of this work was placed on SSF, mutants with increased PG production were selected after comparing their enzyme production in SSF. Mutant DH56 was also tested in SmF due to the low PG production in SSF but simultaneous showing a high pectinase activity on screening medium on agar plates. The PG production by mutant DH56 in SmF was 2.4 fold increased compared to the wild type, while the specific PG activity increased even by factor 3.2. The soluble protein content in SSF was very high, while the amount of secreted proteins in the supernatant in SmF was lower compared to the wild strain. Besides exo-PG activity, mutant DH56 was also the only mutant producing higher exo-PMG activity in SmF than the wild strain. The wild strain A. sojae ATCC 20235 produced an exo-PG activity of 40.4 U/mL, which is lower compared to the previous screening results for media design (section 4.3.2).‎ This might be affected by the scale up utilizing more medium in a bigger flask and also by bigger inoculum size used for the comparison of mutants. A comparison of exo-PG production in SmF between mutant DH56, mutant M3 and A. sojae ATCC 20235 was already done by Buyukkleci et al. [28], applying a complex medium containing pectin or corn meal as pectinase inducer. The cultivation was performed for 120 h. Wild strain produced only around 2 U/mL of exo-PG activity in the presence of both inducer substrates. Activities obtained by mutant M3 were 50.1 U/mL and 7.7 U/mL in corn meal and pectin, respectively. Highest exo-PG activity of 98.6 U/mL was obtained with corn meal by mutant DH56, while this mutant produced 56.8 U/mL of enzyme activity with pectin. In contrast to these findings wild strain A. sojae ATCC 20235 produced huge increased PG activity utilizing the economical medium described in the present study. Furthermore, PG activity obtained by mutant M3 under submerged conditions was slightly lower compared to the wild type. Regarding mutant DH56 similar PG activity was obtained in the present study. Utilizing statistical tools for optimization Buyukkleci et al. [28] enhanced exo-PG production by mutant M5/6, also known as mutant DH56, to 145 U/mL in shake flask cultures. This seems to be promising for media design using the economic media components of the present study and optimization studies utilizing mutant DH56 for pectinolytic enzyme production. The combination of this high enzyme yield with the GRAS status of the strain might be interesting for scale up applications.

95 Chapter 4

Considering mutants descending from A. sojae CBS 100928 for PG production in SSF, only mutant 3b-4 showed increased exo-PG and exo-PMG activity compared to the wild type. Interestingly, this mutant, generated by treating spores with NTG, was stable in contrast to the mutant 2b-18 descending from A. sojae ATCC 20235. Mutant 3b-4 produced 53 % increased exo-PG activity compared to its wild strain. Specific PG activity produced in SSF was even two times higher. Pectinase activity produced in SSF by mutants of A. sojae CBS 100928 generated by treatment with UV irradiation was lower compared to the wild strain. Considering mutants descending from A. sojae CBS 100928 for exo-PG production in SmF, higher PG activity compared to the wild strain was obtained from mutant 60 and 84, generated by mutation with UV irradiation. Highest exo-PG activity of 19.4 U/mL and also highest exo-PMG activity of 2.5 U/mL was produced by mutant 84. Hence, three cycles of UV radiation resulted in 1.8 fold increased PG activity in SmF. Similar to the observations gained from comparison of mutants descending from A. sojae ATCC 20235, mutants with increased PG production in SSF produced lower enzyme activity in SmF and vice versa. These findings support the information given by Barrios-Gonzáles [1], that enzyme or secondary metabolites over-producing strains, generated for SmF, generally do not perform well in SSF, and very seldom are strains efficient in both systems. Furthermore, it has been found that protein production is controlled in response to solid- or liquid-culture conditions [1]. Hence, over-producing strains have to be generated particularly for the respective fermentation system. Previous comparisons of PG production by A. sojae ATCC 20235 and A. sojae CBS 100928 in SSF demonstrated already the higher PG production by A. sojae ATCC 20235 (chapter 2 and chapter 3). According to the present results A. sojae ATCC 20235 produced also higher exo-pectinase activity in SmF. PG activity obtained by A. sojae ATCC 20235 was 4.8 fold higher in SSF and 3.8 fold higher in SmF. Besides PG activity, A. sojae ATCC 20235 produced also higher PMG activity in both fermentation systems. Nevertheless, comparing the obtained exo-PG activity under submerged conditions by A. sojae CBS 100928 to PG production by Aspergillus sp. under optimized conditions utilizing sugar beet as sole carbon source [29], higher enzyme activity was obtained by A. sojae CBS 100928 under the conditions of the present study. Mutants with highest exo-PG activity in SmF and SSF were further explored for their endo-pectinase production in both fermentation systems. Results are presented in Table 4‎ .6.

96 Chapter 4

Table 4‎ .6 Enzyme profile present in crude extracts of SSF and SmF.

Solid-state culture Submerged culture Cycle No./ MO mutagen/ exo-PG endo-activity (U/g) exo-PG endo-activity (U/mL) mutant code (U/g) PG PMG (U/mL) PG PMG

wild type 632.7 ± 61.7 869.4 ± 75.8 126.6 ± 1.3 40.4 ± 3.8 44.40 ± 2.25 7.43 ± 0.38

3/ UV/ M3 1087.2 ± 151.9 1029.5 ± 33.7 129.1 ± 0.2 38.7 ± 6.5 42.71 ± 3.77 7.74 ± 0.06 A. sojae A.

ATCC 20235 ATCC 5/ UV/ DH56 16.1 ± 1.4 2.4 ± 0.3 3.5 ± 1.0 98.8 ± 8.7 61.39 ± 2.06 7.95 ± 0.03

wild type 133.2 ± 12.1 109.1 ± 3.4 55.3 ± 1.5 10.6 ± 1.2 10.49 ± 0.44 3.00 ± 0.05

3/ NTG/ 3b-4 203.7 ± 17.9 114.9 ± 2.8 60.0 ± 1.8 6.6 ± 1.1 9.47 ± 0.58 2.80 ± 0.14 A. sojae A.

CBS100928 3/ UV/ 84 60.2 ± 8.0 75.9 ± 8.3 29.8 ± 2.2 19.4 ± 0.1 11.09 ± 0.13 3.02 ±0.02

97 Chapter 4

Mutants with increased exo-PG activity in SSF showed also increased endo-PG activity, while utilizing these mutants in SmF resulted in lower exo-PG and lower endo-PG activity, and vice versa. The profile of endo-PMG activities was similar to endo-PG activities, but at a considerable lower distinctive extent. In order to achieve a better comparison of endo-pectinase activities between the mutants of the two strains all SSF crude extracts were diluted to 1 U/mL exo-PG activity before measuring the endo-enzyme activities (Table 4.7‎ ). The crude extract of A. sojae ATCC 20235 presented a balanced ratio of exo- to endo-PG activity and a lower endo-PMG activity. Besides the ratio of exo- to endo- PG activity, mutant M3 produced also a balanced exo-PG to endo-PMG ratio. Endo- pectinase activities produced by mutant DH56 in SSF were even considerable lower compared to the already low exo-PG activity. A. sojae CBS 100928 and its mutant 3b-4 produced higher endo-PG than exo-PG activity and a balanced exo-PG to endo-PMG ratio. Mutant 84 produced a balanced ratio of both endo-pectinase activities.

Table 4‎ .7 Ratio of exo-PG activity to endo-pectinase activities in SSF crude extracts.

endo-activity (U/mL) at Cycle No./ exo-PG activity MO mutagen/ 1 U/mL exo-PG activity

mutant code (U/g) PG PMG

5 wild type 632.7 ± 61.7 1.08 ± 0.07 0.87 ± 0.08

3/ UV/ M3 1087.2 ± 151.9 1.09 ± 0.06 0.94 ± 0.07 A. sojae A.

ATCC 2023 5/ UV/ DH56 16.1 ± 1.4 0.13 ± 0.02 0.09 ± 0.03

wild type 133.2 ± 12.1 1.13 ± 0.01 0.99 ± 0.01

3/ NTG/ 3b-4 203.7 ± 17.9 1.13 ± 0.03 1.04 ± 0.12 A. sojae A.

CBS100928 3/ UV/ 84 60.2 ± 8.0 1.02 ± 0.03 0.91 ± 0.05

4.4 Conclusions Development and implementation of a mutation and selection strategy for the improved production of extracellular pectic acid degrading enzymes resulted in enhanced pectinolytic activities of A. sojae mutants. The pre-selection focused on morphological aspects regarding sporulation. Hence, generation of mutants producing sufficient amount of spores for inoculation was assured. Selection of “zone mutants” in the second step of the screening procedure enabled the

98 Chapter 4 detection of desired mutants with enhanced pectinase activity measured as clear zones on screening medium. Utilization of polygalacturonic acid (sodium salt) in the screening medium preferred identification of mutants with increased PG activity. Furthermore, screening on “zone mutants” did not distinguish between increased enzyme production in SmF or SSF system. Hence, in order to specifically increase PG production in SSF another culture flask screening under SSF conditions had to be performed. Performance of repeated and sequential mutagenesis generated desired mutants of A. sojae. Considering A. sojae ATCC 20235 repeated cycles of UV irradiation produced stable mutants in contrast to repeated cycles of NTG treatment. Sequential mutagenesis of A. sojae ATCC 20235 combining both methods generated also stable mutants. According to the results of the pectinolytic enzyme profile a combination of these methods in sequential mutagenesis might be promising for the generation of mutants producing various pectinolytic enzyme activities. However, the substrate of the screening medium should be changed to pectin for the enhancement of multitude of pectinases. Regarding A. sojae CBS 100928 both mutagens can be applied for repeated mutagenesis. Nevertheless, comparison of PG production in both fermentation systems was much lower compared to A. sojae ATCC 20235. Hence, this characteristic indicates the potential of the strain A. sojae ATCC 20235 and its mutants as production organism for industrial applications. Especially the high pectinase activities obtained under solid-state conditions are promising for further scale-up studies.

References [1] Barrios-González, J (2012). Solid-state fermentation: Physiology of solid medium, its molecular basis and applications. Process Biochemistry 47, 175- 185. [2] Parekh, S, Vinci, VA, Strobel, RJ (2000). Improvement of microbial strains and fermentation processes. Applied Micriobiology and Biotechnology 54, 287-301. [3] Demain, AL, Adrio, JL (2008). Strain improvement for production of pharmaceuticals and other microbial metabolites by fermentation. Progress in Drug Research 65, 253-289. [4] Meireles, LA, Guedes, AC, Malcata, FX (2002). Increase of the yields of eicosapentaenoic and docosahexaenoic acids by the microalga Pavlova lutheri following random mutagenesis. Biotechnology and Bioengineering 81, (1) 50-55.

99 Chapter 4

[5] Smith, DC, Wood, TM (1991). Isolation of mutants of Aspergillus awamori with enhanced production of extracellular xylanase and ß-xylosidase. World Journal of Microbiology and Biotechnology 7, 343-354. [6] Bos, CJ (1987). Induction and isolation of mutants in fungi at low mutagen doses. Current Genetics 12, 471-474. [7] Savergave, LS, Gadre, RV, Vaidya, BK, Narayanan, K (2011). Strain improvement and statistical media optimzation for enhanced erythritol production with minimal by-products from Candida magnoliae mutant R23. Biochemical Engineering Journal 55, 92-100. [8] Patil, SR, Dayanand, A (2006). Optimization of process for the production of fungal pectinases from deseeded sunflower hesd in submerged and solid- state conditions. Bioresource Technology 97, 2340-2344. [9] Díaz-Godínez, G, Soriano-Santos, J, Augur, C, Viniegra-González, G (2001). Exopectinases produced by Aspergillus niger in solid-state and submerged fermentation: a comparative study. J. Ind. Microbiol. Biot. 26, 271-275. [10] Solís-Pereira, S, Favela-Torres, E, Viniegra-González, G, Gutiérrez-Rojas, M (1993). Effects of different carbon sources on the synthesis of pectinase by Aspergillus niger in submerged and solid state fermentations. Applied Micriobiology and Biotechnology 39, 36-41. [11] Kashyap, DR, Vohra, PK, Chopra, S, Tewari, R (2001). Applications of pectinases in the commercial sector: a review. Bioresource Technology 77, 215-227. [12] Naidu, GSN, Panda, T (1998). Production of pectolytic enzymes - a review. Bioprocess Eng. 19, 355-361. [13] Chang, P-K, Matsushima, K, Takahashi, T, Yu, J, Abe, K, Bhatnagar, D, Yuan, G-F, Koyama, Y, Cleveland, TE (2007). Understanding nonaflatoxigenicity of Aspergillus sojae: a windfall of aflatoxin biosynthesis research. Applied Micriobiology and Biotechnology 76, 977-984. [14] Machida, M, Asai, K, Sano, M, Tanaka, T, Kumagai, T, Terai, G, Kusumoto, K-I, Arima, T, Akta, O, Kashiwagi, Y, Abe, K, Gomi, K, Horiuchi, H, Kitamoto, K, Kobayashi, T, Takeuchi, M, Denning, DW, Galagan, JE, Nierman, WC, Yu, J, Archer, DB, Bennett, JW, Bhatnagar, D, Cleveland, TE, Fedorova, ND, Gotoh, O, Horikawa, H, Hosoyama, A, Ichinomiya, M, Igarashi, R, Iwashita, K, Juvvadi, PR, Kato, M, Kato, Y, Kin, T, Kokubun, A, Maeda, H, Maeyama, N, Maruyama, J-i, Nagasaki, H, Nakajima, T, Oda, K, Okada, K, Paulsen, I, Sakamoto, K, Sawano, T, Takahashi, M, Takase, K, Terabayashi, Y, Wortman, JR, Yamada, O, Yamagata, Y, Anazawa, H, Hata, Y, Koide, Y, Komori, T, Koyama, Y, Minetoki, T, Suharnan, S, Tanaka, A, Isono, K, Kuhara, S, Ogasawara, N, Kikuchi, H (2005). Genome sequencing and analysis of Aspergillus oryzae. Nature 438, 1157-1161.

100 Chapter 4

[15] Ustok, FI, Tari, C, Gogus, N (2007). Solid-state production of polygalacturonase by Aspergillus sojae ATCC 20235. Journal of Biotechnology 127, 322-334. [16] Heerd, D, Yegin, S, Tari, C, Fernandez-Lahore, M (2012). Petinase enzyme- complex production by Aspergillus spp in sold-state fermentation: A comparative study. Food and Bioproducts Processing 90, 102-110. [17] Antier, P, Minjares, A, Roussos, S, Raimbault, M, Viniegra-González, G (1993). Pectinase-hyperproducing mutants of Aspergillus niger C28B25 for solid-state fermentation of coffee pulp. Enzyme and Microbial Technology 15, 254-260. [18] Loera, O, Aguirre, J, Viniegra-González, G (1999). Pectinase production by a diploid construct from two Aspergillus overproducing mutants. Enzyme and Microbial Technology 25, 103-108. [19] Solís, S, Flores, ME, Huitrón, C (1990). Isolation of endoploygalacturonase hyperproducing mutants of Aspergillus sp. CH-Y-1043. Biotechnology Letters 12, (10) 751-756. [20] Molina, SMG, Pelssari, FA, Vitorello, CBM (2001). Screening and genetic improvement of pectinolytic fungi for degumming of textile fibers. BRAZILIAN JOURNAL OF MICROBIOLOGY 32, 320-326. [21] De Nicolás-Santiago, S, Regalado-González, C, García-Almendárez, B, Fernández, FJ, Téllez-Jurado, A, Huerta-Ochoa, S (2006). Physiological, morphological, and mannase production studieson Aspergillus niger uam-gs1 mutants. Journal of Biotechnology 9, (1) 51-60. [22] (2004). In Report on Carcinogens, Eleventh Edition; U.S. Department of Health and Human Services - Public Health Service. [23] Durrands, PK, Cooper, RM (1988). Selection and characterization of pectinase-deficient mutants of the vascular wilt pathogen Verticillium albo- atrum. Physiological and Molecular Plant Pathology 32, 343-362. [24] Bradford, MM (1976). A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Analytical Biochemistry 72, 248-254. [25] Panda, T, Naidu, GSN, Sinha, J (1999). Multipleresponse analysis of pectinolytic enzymes by Aspergillus niger: a statistical view. Process Biochemistry 35, 187-195. [26] Nelson, N (1944). A photometric adaption of the somogyi method for the determination of glucose. Journal of Biological Chemistry 153, (2) 375-380. [27] Mill, PJ, Tuttobello, R (1961). The pectic enzymes of Aspergillus niger. Biochemical Journal 79, 57-64. [28] Buyukkileci, AO, Tari, C, Fernandez-Lahore, M (2011). Enhanced production of exo-polygalacturonase from agro-based products by Aspergillus sojae. BioResources 6, (3) 3452-3468.

101 Chapter 4

[29] Galiotou-Panayotou, M, Kapantai, M, Kalantzi, O (1997). Growth conditions of Aspergillus sp. ATHUM-3482 for polygalacturonase production. Applied Micriobiology and Biotechnology 47, 425-429.

102 Chapter 5

Chapter 5.

Improved PG enzyme bioproduction by Aspergillus sojae mutant M3 and SSF process studies at bioreactor level

Abstract

The development of a fermentation process involves many steps, which include amongst others the selection of an appropriate microorganism, optimization of process parameters as well as standardization of process unit operations at a laboratory scale and scale up studies. The previously performed microbial screening identified Aspergillus sojae ATCC 20235 as potential pectinase producer and the microbial strain improvement generated mutant M3 with enhanced polygalacturonase production in solid-state culture. Furthermore, the previously conducted media design utilizing the wild strain provided a substrate mixture of wheat bran and sugar beet pulp in the ratio 70:30, which acts as physical support, source of nutrients and appropriate inducer for polygalacturonase production. This was taken as basis for the solid-state fermentation process optimization applying mutant M3 before performing the scale up of the optimized process. However, the substrate pretreatment procedure was changed with regard to the scale up of the fermentation process. Optimization experiments identified highest polygalacturonase production (1009.4 ± 89.9 U/g) by mutant M3 at a moisture level of 160 % after 8 days solid-state culture at 30 °C. Under the applied optimized conditions mutant M3 produced 2.3 fold increased polygalacturonase activity at laboratory culture flask level in comparison to the wild type. The optimized process was scaled up to bioreactor level utilizing a rotating drum type solid-state bioreactor with a scaling factor of 100. Comparison of enzyme production at culture flask level and at bioreactor level indicated similar production levels obtained by mutant M3. Comparing polygalacturonase production between mutant M3 and the wild strain at bioreactor level presented 1.4 fold increased enzyme production by the mutant. The results of this study demonstrated the successful solid-state fermentation process scale up without loss in the total amount of enzyme. Moreover, the combination of cost-efficient substrate and high productivity by the applied strain are promising for large scale production processes.

103 Chapter 5

5.1 Introduction Enzyme production is a growing field of biotechnology with increasing number of patents and research articles [1]. Enzymes are working as effective catalysts under mild conditions resulting in significant savings in resources such as energy and water for both the industry and the environment [2]. Pectinases have a share of 25 % in the global sales of food enzymes, and polygalacturonases (PG) are the most abundant among all pectinases [3]. Various commercial preparations of pectinases are produced from fungal sources [4], especially Aspergillus niger is used for industrial pectinase production [5]. Classically, industrial important enzymes have been obtained from submerged fermentation (SmF), because of the better control of environmental factors such as temperature and pH during SmF, facilitating scaling up of this technique [1, 6, 7]. However, solid-state fermentation (SSF) is a promising technique with tremendous potential for the production of enzymes [6, 8]. This fermentation process has developed in eastern countries over many centuries and has been considered superior to SmF in several aspects, such as the generally natural materials as solid substrates, which are more cost effective. The possibility of processing agro- industrial residues particularly offers numerous opportunities in SSF [9]. In addition, the low moisture content involves reducing pollution concerns and due to the low water activity, SSF is relatively resistant to bacterial contamination. Another advantage of SSF is presented by the product extraction which requires less solvent and lower recovery costs than the extraction from SmF. Furthermore, higher product titers are observed by lower usage of substrate, and the SSF process uses low volume equipment, too [6]. Filamentous fungi are the most suitable microorganisms for SSF, because of their capability to grow well in the absence or near absence of free water, their morphology and their high potential to secret hydrolytic enzymes [9, 10]. Furthermore, the capability of most filamentous fungi to grow well at low pH is advantageous since pH control is difficult in SSF [11]. The process operation at low pH-values is minimizing the risk of bacterial contamination. Furthermore, many fungi secrete PG in acidic media and this is also the pH range where majority of their PGs show optimum catalytic activity [12]. Additionally, the ability of filamentous fungi to degrade macromolecular substrates, especially carbohydrates favors the use of these microorganisms in SSF. SSF has been extensively used in eastern countries to produce oriental food applying koji molds, such as A. sojae and A. oryzae, which has proved the safety of these filamentous fungi [13, 14].

104 Chapter 5

SSF conditions resembles the natural habitat of most higher filamentous fungi, which favors growths and production of bio-products, such as enzymes under SSF conditions [6, 15]. Nevertheless, there are major challenges that need to be addressed for the successful implementation of SSF processes, such as the process regulation, e.g. providing heat and mass transfer within the substrate bed; and on- line monitoring of key parameters [6]. Biomass determination is also challenging in SSF, which is essential for kinetic studies. Since, it is often not possible to separate biomass from solid medium in SSF, several indirect methods were applied for biomass estimations having all of them their own weakness. Indirect biomass determination methods rely for example on glucosamine estimation, ergosterol estimation, protein estimation, DNA estimation, dry weight changes, oxygen uptake rate and carbon dioxide evolution rate [15]. Besides that, the scale up of the SSF process is a major bottle neck [15]. Many different bioreactors have been used in SSF processes. Mitchell et al. [16] divided SSF bioreactors into four groups based on the type of aeration or the mixing system employed. This grouping is based on the operation. Hence, a strictly distinction is not always clear. Bhargav et al. [17] distinguished SSF bioreactors similar like Durand [18], whether they are used in small scale or large scale processes. According to the division tray bioreactors, packed-bed bioreactors, rotating drum bioreactors and fluidized bed reactors belong to the group of large scale bioreactors. Nevertheless, there are still challenges of SSF and continuous efforts are needed, e.g. in the direction of biomonitoring improvement and higher standardization in SSF processes [6, 15]. Generally development of commercially viable processes is favored by: (a) strain with high yield or productivity, (b) optimized fermentation medium composition having low-cost nutrients, (c) optimized process parameters and (d) ease of downstream processing for product recovery [19]. After identifying the potential of the koji mold A. sojae for pectinolytic enzyme production in SSF ([20] and chapter 2), and successfully enhancing PG production by microbial strain improvement (chapter 4), the present study was initiated to optimize PG production by mutant M3, descending from A. sojae ATCC 20235, and transfer optimized process conditions from culture flask into the bioreactor level. The impact on PG production was explored in laboratory fermenter level utilizing a rotation drum type solid-state bioreactor.

105 Chapter 5

5.2 Materials and Methods

5.2.1 Materials All chemicals were purchased from AppliChem GmbH (Darmstadt, Germany), except the substrates for detection of PG activity, polygalacturonic acid, and the chemicals sodium arsenate dibasic heptahadrate and D-(+)-glucosamine hydrochloride, were obtained from Sigma-Aldrich Chemie GmbH (Steinheim, Germany). Microbial substrates like wheat bran and sugar beet pulp pellets were obtained from local suppliers (Bremer Rolandmühle Erling GmbH & Co. KG, Bremen, Germany; Nordzucker AG, Uelzen, Germany).

5.2.2 Microorganism A. sojae ATCC 20235 was purchased from Procochem Inc (Teddington, United Kingdom). Mutant M3, an enhanced producer of PG, was generated by repeated mutagenesis applying UV irradiation. This mutant derived from wild strain A. sojae ATCC 20235 (chapter 4). Fungal strains were propagated on yeast malt extract (YME) agar plates according to the specifications given in Heerd et al. ([20] and chapter 2). Spores from these plates were used as inoculum for molasses agar slants containing: glycerol (45 g/L), molasses (45 g/L), peptone (18 g/L), NaCl (5 g/L), KCl

(0.5 g/L), FeSO4·7H2O (15 mg/L), KH2PO4 (60 mg/L), MgSO4 (50 mg/L), CuSO4·5H2O

(12 mg/L), MnSO4·H2O (15 mg/L) and agar (20 g/L). High amount of spore production was achieved by cultivation on molasses agar slants at 30 °C for 1 week. Spores suspensions obtained from harvesting molasses agar slants with sterile Tween 80 water (0.02%) were adjusted to the desired spore concentration by counting spores of the suspension in a Thoma counting chamber and diluting with sterile distilled water. Inoculation of culture flasks and bioreactor was done applying 2×106 spores per gram dry WS (weight of substrate).

5.2.3 PG production at culture flask level, RSM design and enzyme leaching SSF was performed in 250-mL Erlenmeyer flasks containing 10 g of wheat bran and sugar beet pulp in the ratio 70:30, wetted with tap water and diluted HCl to a final concentration of 0.2 M HCl according to the experimental design given in Table 5.1‎ . Culture flasks containing the wetted medium were sterilized in the autoclave (121 °C for 20 min). After cooling to room temperature flasks were inoculated and incubated as needed (section 5.3.1).‎ Flasks were shaken manually twice at the day of inoculation and the first cultivation day.

106 Chapter 5

At the end of cultivation, enzyme leaching was performed with 50 mL distilled water for 60 min at 350 rpm agitation and 24 °C. Enzyme activity (section 5.2.6)‎ and total protein content (section 5.2.7)‎ were determined in the supernatant obtained after centrifugation at 4 °C, 3220 × g, for 30 min. Previously optimized medium for enhanced PG production by the wild strain A. sojae ATCC 20235 (chapter 3) was used as basal medium for optimization studies applying mutant M3. With regard to the scale up of the fermentation process the substrate pretreatment changed from grinding the sugar beet pulp pellets to soaking in tap water. After swelling, sugar beet pulp was mixed with wheat bran and 0.5 M HCl to adjust the final HCl concentration to 0.2 M and the final moisture content to the amounts given in Table 5.1‎ . The described moisture levels in the experimental set- ups were calculated as dry basis moisture content according to the following equation:

Moisture content u (%) = (weightwet – weightdry) / weightdry *100 (5.1) where weightwet is the weight of solid media together with water and diluted HCl and weightdry means weight of solid media (natural moisture content in solid media has not been considered for calculation). Besides moisture content, the process parameters temperature and cultivation time were optimized for exo-PG production by mutant M3, applying central composite face-centered (CCF) design. Optimization of these three factors was done in an experimental setup of 17 experiments including three center points and the response, exo-PG activity, was recorded (Table 5.1‎ ). Factor levels were defined with regard to previous optimization of the PG production by wild strain A. sojae ATCC 20235 (chapter 3). Hence, moisture level should be high, but must not exceed a maximum of 170 % under the applied conditions. Previous optimized temperature and time of cultivation have been 30 °C and 8 d, respectively. Consequently, the applied factor levels of the present optimization of PG production by mutant M3 were fixed including the previously optimized ranges. The experiments were evaluated using PLS applying the MODDE 9.0 software package, supplied by Umetrics AB, Umeå, Sweden.

5.2.4 Bioreactor studies Scale up of PG production from 10 g dry substrate at culture flask level to 1 kg dry substrate at laboratory bioreactor level was carried out. Bioreactor studies for PG production were conducted in a 15-L rotating drum type solid-state bioreactor, Terrafors-IS (Infors HT, Switzerland). Operating conditions for PG production by A. sojae in SSF were previously adapted to bioreactor scale (Appendix B). Operating

107 Chapter 5 temperature was maintained at 30 °C and compressed air passed through a sterile filter (Novasip™ capsules, Pall GmbH, Dreieich, Germany) was used for forced- aeration of the bed. Aeration was supplied according to the standard aeration procedure increasing the air flow rate from 2 L/min to 5 L/min at the first day of SSF (Appendix B). An exit gas cooler minimized evaporation losses. The condensate was re-circulated into the vessel through the air inlet with a peristaltic pump, applying a setpoint entry of 50 (%). Levels of oxygen and carbon dioxide in the exit gas stream were detected by an exit gas analyzer. Data logging of on-line parameters was done with Iris V5 control software. SSF was performed in an intermittent mixed process. Related to the manual shaking of culture flasks (section 5.2.3),‎ the drum was rotated at 1 rpm for 10 min clockwise and for another 10 min anticlockwise twice on the day of inoculation and twice during the first day of cultivation. The bioreactor was in-situ sterilized via a double jacket and direct steam injection directly before use and filled with separately sterilized medium. In order to apply the previously optimized medium wetted at 160 % (section 5‎ .3.1), the medium had to be sterilized in the autoclave (121 °C, 20 min); further explanations are given in Appendix B. For this purpose were 0.3 kg sugar beet pellets soaked in 0.96 L of tap water. To this 0.7 kg wheat bran and 0.64 L of 0.5 M HCl were added to achieve a final HCl concentration of 0.2 M HCl and a moisture level of 160 %. After sterilization the medium was filled into the sterilized vessel of the bioreactor. At 24 h intervals samples of approximately 100 g were taken and allocate a sample. Aliquots of fermented biomass were used directly to determine the moisture content (section 5.2.5)‎ and estimate the glucosamine content (section 5.2.9),‎ while other aliquots were used for enzyme leaching. Enzyme leaching was performed utilizing 80 mL distilled water for 20 g biomass, in order to keep the ratio 1:4 of biomass to eluent (chapter 6). The mixture was incubated for 20 min at 250 rpm at 25 °C. Supernatant was separated from biomass by centrifugation for 30 min at 3220 × g at 4 °C. Exo-PG activity (section 5.2.6),‎ soluble protein content (section 5.2.7),‎ total soluble carbohydrate content (section 5.2.8)‎ and pH were determined in the supernatant. The pH was recorded using pH-meter equipped with glass electrode.

5.2.5 Determination of moisture content Samples from SSF were dried at 60 °C until constant weight (48 h) to determine the moisture content. The moisture content of biomass was calculated as wet-basis moisture content (relative to total weight) according to the following equation:

Wet-basis moisture w (%) = (weightwet – weightdry) / weightwet *100 (5.2)

108 Chapter 5

where weightwet is the weight of biomass and weightdry means weight of dried biomass. A conversion of the aforementioned dry basis moisture content (formula 5.1) to the wet-basis moisture is calculated as follows:

100*Moisture content u (%) Wet-basis moisture w (%) = (5.3) 100 + Moisture content u (%)

5.2.6 Exo-polygalacturonase assay Exo-PG activity was assayed according to the procedure of Panda et al. [21], which was further optimized as described in Appendix A. Exo-PG activity values obtained at culture flask level were calculated per g dry WS, whereas values obtained at bioreactor level were calculated per g biomass. In order to achieve a comparison between these values, the values achieved at bioreactor level were also calculated per gds (dried solids) (section 5.3.2.3).‎

5.2.7 Soluble protein content Total extracellular protein was measured according to the modified Bradford´s method [22] as described in Appendix A. Soluble protein content was expressed as mg per g dry WS in culture flask studies and as mg per g biomass at bioreactor level, which were also calculated per gds for comparison.

5.2.8 Soluble carbohydrate content Soluble carbohydrates in the crude extract were determined by the phenol–sulfuric acid method according to Dubois et al. [23], following the procedure given in Appendix A. Soluble carbohydrate content was expressed as mg per g biomass at bioreactor level.

5.2.9 Biomass determination The biomass estimation was done based on determining the glucosamine released by acid hydrolysis of chitin present in the cell wall of the fungi according to the method provided by Zamani et al. [24] with slight modifications. Dried and grinded samples (section 5.2.5)‎ of 100 mg were mixed with 0.3 mL of 72 % (v/v) H2SO4 and incubated for 90 min at room temperature with occasional mixing using a glass bar. Then, 8.4 mL distilled water was added into each screw cap centrifuge tube, tubes were closed tightly and samples were placed in the autoclave (20 min at 121 °C) for hydrolysis. After autoclaving, two samples of each 0.5 mL were taken from each tube, while solutions were still warm. When samples were cooled to room temperature, 0.5 mL of 1 M NaNO2 or 0.5 mL of distilled water was added to sample

109 Chapter 5 and sample blank, respectively. All tubes were closed tightly, mixed and left for 3 h at room temperature. Then, tubes were opened and left overnight under the hood to complete the depolymerization – deamination reaction and to remove the NO2 that arose as a byproduct in the reaction mixture of the sample tube. After incubation, 0.5 mL of 12 % (w/v) NH4SO3NH2 was added into each tube, followed by 4 min mixing. Then, 0.5 mL of 0.5 % (w/v) 3-methyl-2-benzothiozolone-hydrazone- hydrochloride (MBTH) was added. Tubes were incubated for 5 min in boiling water and cooled to room temperature in another water bath for 5 min. It was followed by addition of 0.5 mL of o.5 % (w/v) FeCl3. After addition of FeCl3 tubes were vortexed and incubated at room temperature for 30 min. Samples and sample blanks were diluted 10 times with distilled water, and the absorbance of all samples was measured at 650 nm. Solutions of glucosamine hydrochloride in 2.84 % (v/v)

H2SO4 were used as standards (0 – 800 µg/mL) and in-cooperated in the assay starting with the addition of 1 M NaNO2 and followed the assay procedure like samples. The glucosamine content was expressed as mg per gds.

5.3 Results and Discussion The development of a microbial fermentation process involves many steps [25], which include amongst others the isolation, screening and selection of the appropriate microorganism; optimization of physic-chemical and nutritional parameters as well as standardization of process unit operations at a laboratory scale; and scale up studies. A microbial screening for pectinolytic enzyme production identified the potential of A. sojae as PG producer in SSF ([20] and chapter 2). Microbial strain improvement generated mutant M3 with enhanced PG production in SSF (chapter 4). Hence, the appropriate microorganism was selected and previous media design utilizing the wild strain (chapter 3) provided a substrate mixture of wheat bran and sugar beet pulp in the ratio 70:30, which acts as physical support, source of nutrients and also as appropriate inducer for the production of PG. This was taken as basis for process optimization applying mutant M3 to further improve enzyme yield in SSF. Moreover, a scale up of PG production from 10 g dry substrate used in culture flasks to 1 kg dry substrate (2.6 kg moist solids) applied in batch processes at bioreactor level was performed.

5.3.1 Optimization of PG production by A. sojae mutant M3 in SSF Previously optimized medium for enhanced PG production by the wild strain A. sojae ATCC 20235 (chapter 3) was used as basal medium for optimization studies applying mutant M3. The factors moisture content, incubation temperature and cultivation time were optimized for PG production applying central composite face-centered

110 Chapter 5

(CCF) design. The design of experiments and respective experimental results are given in Table 5.1‎ . Results of the runs based on the CCF design showed a wide variation in PG production ranging from 1.3 U/g above 1000 U/g. This variation showed the importance of optimizing these parameters for improving PG production by mutant M3.

Table 5‎ .1 Central composite face-centered design and experimental results of PG activity, specific activity and protein content.

Factors Response Exp. Specific Protein activity content No. Moisture Cultivation Tempera- PG activity (U/mg) (mg/g) level (%) time (d) ture (°C) (U/g)

1 110 6 25 713.1 136.4 5.23

2 170 6 25 967.4 425.8 2.27

3 110 10 25 91.3 9.3 9.77

4 170 10 25 1072.0 325.4 3.29

5 110 6 35 5.6 0.7 7.62

6 170 6 35 1166.9 591.7 1.97

7 110 10 35 1.3 0.2 7.84

8 170 10 35 23.0 5.0 4.55

9 110 8 30 63.0 7.2 8.70

10 170 8 30 984.7 473.9 2.08

11 140 6 30 939.5 360.6 2.61

12 140 10 30 258.9 34.0 7.62

13 140 8 25 1098.9 238.0 4.62

14 140 8 35 778.1 218.0 3.57

15 – 17* 140 8 30 905.5 ± 33.4 196.8 ± 6.9 4.61 ± 0.31

* The standard variation between the center point repetitions of the response values was below 4 %.

For model evaluation the two companion statistics goodness of fit (R2) and goodness of prediction (Q2) were examined. The PLS analysis of the data shown in

111 Chapter 5

Table 5.1‎ gave a model with R2 = 0.825 and Q2 = 0.201, i.e., a clearly invalid model. The reason for this was found in a plot of the regression coefficients and their 95 % confidence intervals (plot not shown). The interaction terms which were small and insignificant were removed. For evaluation of the model residuals the normal probability plot (N-plot) was used for detecting deviating experiments. When looking at the N-plot (plot not shown) Exp. no. 6 was identified as outlier, which degraded the predictive ability of the model. This seemed to be reasonable looking at the response values of Exp. no. 5 – 8 obtained at a cultivation temperature of 35 °C. With exception of Exp. no. 6 a relatively low PG activity was obtained at high cultivation temperature. Hence, this experimental run was excluded for further evaluation. Upon removal of insignificant interaction terms, exclusion of Exp. no. 6 and refitting the model, the statistics R2 = 0.834 and Q2 = 0.561 were obtained. These statistics point to a model with some imperfection, because R2 substantially exceeds Q2. This model showed also a slight lack of fit (p = 0.027). But the F-test, which assessed the significance of the regression model, had a p-value of 0.001. The value p < 0.05 indicated that the model was statistically significant. Since similar optimization of these factors was already performed utilizing the wild strain A. sojae ATCC 20235 (chapter 3), optimized conditions for enhanced PG production by the wild strain were already known. Hence, the model was used for predictions, but always with regard to the previous findings. The loading scatter plot in Figure 5.1‎ shows the correlation structure between all factors and the response at the same time. Points that lie close to each other and far from the origin indicate variables (factors and response) that are highly correlated. Hence, factor moisture level and response PG activity were strongly correlated, which means moisture level had a strong positive influence on enzyme activity – an increase in moisture level should result in an increased value of PG activity. Nevertheless, previous optimization studies for PG production by the wild strain indicated also an upper limit for the moisture level (chapter 3). Hence, maximal moisture level was set at 170 % for the present study. Generally, varying the moisture level as independent parameter should be seen disadvantageously, because changing the moisture level means a change in the total reaction volume, which changes both substrate volume in solid-state cultivation and enzyme concentration obtained after enzyme leaching. The optimal setting of the factor moisture content for cultivation of microorganisms in solid-state cultivation processes is highly dependent upon water-binding properties of the substrate [26]. Therefore, utilization of the substrate mixture containing wheat bran and sugar

112 Chapter 5

beet pulp in the ration 70:30 was essential for this study. Wheat bran provides sufficient nutrients and is able to remain loose even in moist conditions, thereby providing a large surface area.Investigation: While Opti No sugar 6 excluded (PLS,beet comp.=2) pulp, with its high pectin content, acted as inducer substrate for pecLoadingtinase Scatter: wc[1] production. vs wc[2] for PG activity Factors Responses Interaction Terms 1.01,0 Time * Time 0.50,5 Moisture level * Moisture level Moisture level

[2] 0.0 PG

0,0 Temperature wc[2]

Investigation: Opti No wc 6 excluded (PLS, comp.=2) Time Loading Scatter: wc[1] vs wc[2] for PG activity -0.5-0,5 Factors Responses Interaction Terms -1.0-1,0 -1.0 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 0,8 Tim*Tim -1,0 -0,8 -0,6 -0,4 -0,2 -0,0 0,2 0,4 0,6 0,8 1,0 wc [1] 0,6 wc[1] Figure 5‎ .1 Loading plot of the PLS model showing X- and Y-weights (w and c), thus the correlation HCl*HCl HCl 0,4 structure between factors and the response.N=16 Cond. no.=3,48 DF=10

MODDE 9 - 2013-01-20 18:43:00 (UTC+1) 0,2 Furthermore, the factors temperature and time are plotted relatively distant from

PGa wc[2] -0,0 the response in Figure 5.1‎ , indeed on the opposite side of the origin. Hence, these Temp factors were expected to have a lesser, and negative, influence on PG activity. -0,2 Regarding the temperature, an increase to the maximal factor level of 35 °C resulted -0,4 Tim in low PG activity values (Table 5.1‎ , Exp. No. 5, 7 and 8). -0,5 -0,4 -0,3 -0,2 -0,1 -0,0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 The relationship between the response and the independent variables is presented in the contourwc[1] plots (Figure 5.2‎ and Figure 5.3‎ ).

N=16 Cond. no.=3,48 DF=10 exo-PG activity MODDE 9 - 2013-01-19 14:50:17 (UTC+1) (U/g)

Figure 5‎ .2 Contour plots illustrating the effect of moisture level and temperature on PG activity at specific cultivation times.

113 Chapter 5

exo-PG activity (U/g)

Figure 5‎ .3 Contour plots showing the effect of moisture level and time on PG activity at specific cultivation temperatures.

One must not forget that the contours represent contours of estimated response and the general nature of the system arises as a result of a fitted model. According to these results, high PG activity was estimated at low cultivation temperature (25 °C) in combination with medium incubation times (around 7 days) and high moisture levels (around 156 %). For further predictions to determine conditions yielding in maximum PG activity the optimizer of the MODDE 9.0 software package was used, which suggested a cultivation temperature of 25 °C, incubation times between 6 to 8.5 days and moisture levels ranging from 146 to 158 %. On the basis of optimization results for increased PG production by the wild strain (chapter 3) and also obtained enzyme activities of the mutant M3 during comparison experiments in SSF (chapter 4), a temperature of 30 °C was considered for enzyme production, too. Performing a small screening varying moisture levels between 140 – 170 %, cultivation times between 6 – 8 d and temperature at 25°C and 30 °C, strongly favored a moisture level of 160 % for enhanced PG production by mutant M3 (data not shown). Furthermore, the higher temperature of 30 °C and cultivation times of 7 – 8 d seemed to be preferable for PG production. Thus, moisture level was fixed at 160 % and a comparison of PG production under the mentioned conditions was performed (Table 5.2‎ ). Comparing PG activities presented in Table 5.2‎ strongly favored the temperature of 30 °C and 8 days cultivation time for PG production by mutant M3. Interestingly, higher specific activity was observed at a shorter incubation time of 7 days, due to the lower protein content obtained in the supernatant after enzyme leaching. Similar exo-PG activity of 1087.2 ± 151.9 U/g after 8 days SSF was obtained by mutant M3 in previous culture flask experiments (chapter 4), which proved neither the

114 Chapter 5 change of inducer substrate pretreatment nor the usage of tap water instead of distilled water for adjusting the moisture level significantly affected the PG production by mutant M3. Thus, the changed pre-treatment improved process economics without losses in enzyme yield and productivity.

Table 5‎ .2 Validation experiments to identify optimal factor settings for enhanced PG production.

Time Temperature exo-PG activity Specific Protein content MO (d) (°C) (U/g) activity (U/mg) (mg/g)

7 25 839.4 ± 1.5 246.3 ± 1.4 3.41 ± 0.03

8 25 820.8 ± 14.1 194.0 ± 9.9 4.24 ± 0.14 Mutant M3 7 30 905.9 ± 152.2 294.6 ± 40.4 3.21 ± 0.96

8 30 1009.4 ± 89.9 254.6 ± 7.5 3.96 ± 0.24

A. sojae ATCC 8 30 447.9 ± 11.1 74.9 ± 3.2 5.99 ± 0.11 20235

Factor settings under optimized conditions for enhanced PG production by mutant M3 correspond to previous findings utilizing grinded sugar beet pulp for PG production by wild strain A. sojae ATCC 20235. Under these conditions an exo-PG activity of 909.5 ± 2.7 U/g was obtained after 8 days cultivation of A. sojae ATCC 20235 at 30 °C (chapter 3). Taking this value into account, PG production by mutant M3 increased only by 11 %, which is significant lower compared to the difference obtained during previous comparison studies (chapter 4). Nevertheless, one has to consider the use of whole sugar beet pellets at a stretch during this optimization. With regard to planned bioreactor studies, instead of grinding the inducer substrate to fine particles, pellets were soaked in tap water before use. Therefore, wild strain A. sojae ATCC 20235 had to be cultivated under the same conditions utilizing the same type of inducer substrate for a meaningful comparison of PG activities (Table 5.2‎ ). This decreased PG production of the wild strain by factor two in comparison to previous results, which corresponds to a 2.3 fold increased enzyme activity of the mutant. With regard to process economics the changed substrate pretreatment in combination with PG production by the mutant M3 seemed to be promising for enzyme production in scale up studies.

115 Chapter 5

5.3.2 Scale up of PG production PG enzyme production was scaled up to laboratory bioreactor level based on the previous optimization study at culture flask level (section 5.3.1).‎ Accordingly, the exo-PG production was scaled up to 1 kg dry substrate level in a rotating drum type solid-state bioreactor. At this level experiments were conducted for the selection of operating strategies for medium sterilization, inoculum development, aeration and agitation (Appendix B). In order to perform experiments, which were comparable to culture flask experiments, the applied medium at bioreactor level had to be separately sterilized in the autoclave that the previously optimized moisture content of 160 % could be used (Appendix B). Furthermore, with regard to the manual shaking of culture flasks at the beginning of the cultivation (section 5.2.3)‎ an intermittent mixed process was performed at bioreactor level with similar mixing periods at the beginning. Besides enzyme production by mutant M3, PG production by the wild strain was also explored at bioreactor level utilizing the same cultivation conditions.

5.3.2.1 PG production by mutant M3 at bioreactor level The exo-PG production trend during an intermittent mixed solid-state process by mutant M3 is presented in Figure 5.4‎ -A. Maximal exo-PG activity of 420.4 ± 17.4 U/g was achieved after 8 days fermentation, which corresponded to a productivity of 52.6 U/g/d. At the same time a soluble protein content of 2.3 ± 0.1 mg/g and thus, a specific activity of 182.8 U/mg was obtained. It has to be recalled that the mentioned values are calculated per g biomass and can’t be compared directly with the aforementioned optimized values at culture flask level, which have been calculated per g dry WS. In order to achieve comparable values at bioreactor level, SSF samples were dried until constant weight to determine the dry weight and values were calculated per g dried solids (section 5.3.2.3).‎ In Figure 5.4‎ can be seen that it took approximately 24 h before significant changes in the parameters have been measured. During this period the microorganism was adapting to the growth conditions. In terms of inoculation with a spore suspension the conidia begin the process of germination when they encounter a suitable substrate, forming mature hyphae that will develop into mycelium. After one day an increase in carbon dioxide and a decrease in oxygen concentration in the exhaust gas was measured, which indicated fungal metabolism. These changes reached their maximum at the end of the first cultivation day and gradually changed back during further fermentation time. At the same time visible formation of mycelium in the bioreactor was noticed. After one day SSF, an uptake of soluble carbohydrates and proteins in the medium was detected. At the same time an increase in

116 Chapter 5 glucosamine content was measured, indicating fungal growth. Since it was not possible to separate biomass from solid medium in samples taken during SSF, an indirect approach had to be used to monitor the fungal growth. Among indirect methods relying on measurement of biomass compounds, such as glucosamine, ergosterol or protein, glucosamine measurement was chosen, since it was considered as a good biomass indicator [27]. Moreover, glucosamine content of fungal biomass in liquid cultures is not applicable to fungal biomass in solid-state culture [28]. Hence, the amount of glucosamine was used for representing biomass, without converting it into mycelia dry weight. While oxygen consumption and carbon dioxide production already decreased at the end of the first fermentation day the detectable glucosamine formation just started. Maximum glucosamine content was measured after 6.3 days. Further cultivation yielded in a decrease of the glucosamine content. The visible formation of spores was observed after 3 days SSF. Like the production of glucosamine for fungal growth, the exo-PG production started after one cultivation day. While glucosamine content decreased after 6.3 days of fermentation, the enzyme production further increased until the 8th day of SSF. A higher protein production then uptake from the medium by the fungal strain was observed after 2 days. The increase in protein content lasted till the end of cultivation and reached a total soluble protein content of 3.44 ± 0.10 mg/g after 9 days. The total soluble carbohydrate content in the medium of 57.9 ± 0.8 mg/g was determined. Intensified carbohydrate consumption to a carbohydrate concentration of 15.3 ± 0.1 mg/g was followed between 1.3 and 3.0 days SSF. After this time only a slight decline in soluble carbohydrate concentration was measured in the crude extracts, that reached its minimum of 6.6 ± 0.1 mg/g soluble carbohydrate concentration after 7 days. The initial pH of 3.4 ± 0.1 in the crude extract increased especially in the second half of the fermentation period to pH 5.7 ± 0.1 after 9 days.

117

60 50 40 30 20 10 0

Chapter 5 7 6 5 4 3 2 1 0

0

100 200 300 500 400

500 5 80 10

70 9 400 4 60 8

50 7 300 3 6

40

5

200 2 (%) moisture basis -

MW PG activity (U/g) activity PG 30

5 4 3 2 1 0

Wet

4 Soluble proteins (mg/g) proteins Soluble MW MW MW MW MW

20

0

100 1 3 100 200 500 300 400 60 50 10 40 30 20 10 0 2

70 60 50 40 30 20 10 0

9 7 6 5 4 3 2 1 0

0 0 0 1 0 5 4 3 2 1 0 60 50 40 30 20 10 0 8 100 60 50 40 30 20 10 0 200 300 500 0 1 2 3 4 5 6 7 8 9 400 Time (d) 10 7 6 5 4 3 2 1 0 0 0 7 6 5 4 3 2 1 0 A 7 100 200 300 500 400 0 9 0 100 200 300 500 400 100 200 300 500 60400 6 9 8 10

10 6

50 5 9 8 7 9 5 4 8 6 8 40 7 4 7 5 7

3

) -

30 6

6

4 pH ( pH MW MW MW MW MW 6

3 2 5 3 5 20 5 2 1 2 4 4 MW MW MW MW MW

4 Soluble carbohydrates (mg/g), (mg/g), carbohydrates Soluble

MW MW MW MW MW Glucosamine content (mg/gds) content Glucosamine

10 1 0 1 3 3 3 0 2 0 2 0

0 1 2 3 4 5 6 7 8 9 2 1 Time (d)1 B 1 0 Figure 5‎ .4 Cultivation profile of mutant M30 at rotating drum type solid-state bioreactor; A: Time courses of process parameters, total soluble protein content (——), exo-PG activity (——) and 0 wet-basis moisture (——); B: Time courses of soluble carbohydrate content (——), glucosamine content (——) and pH (——).

The initial moisture content of 160 %, which was adjusted before autoclaving was determined as 65.3 % wet-basis moisture after medium sterilization. Thus, the initially adjusted wet-basis moisture of 61.5 % was increased to 65.3 % during sterilization. This was also determined as an increase in medium mass after

118 Chapter 5 autoclaving. At the beginning of the fermentation process a slight increase in wet- basis moisture was detected, which might result from the formation of biomass. After the second half of the fermentation process the wet-basis moisture started to decrease which intensified after 7.3 days SSF. After 9 days the wet-basis moisture decreased to a value of 55.0 %, which corresponded to a dry basis moisture content of approximately 122 %.

5.3.2.2 PG production by A. sojae ATCC 20235 at bioreactor level To compare the exo-PG production at bioreactor level between mutant M3 descending from A. sojae ATCC 20235 and the wild strain (section 5.3.2.3),‎ both strains had to be cultivated under the same conditions. Thus, A. sojae ATCC 20235 was used for pectinase production in the rotating drum type solid-state bioreactor. The exo-PG production trend during an intermittent mixed solid-state process by the wild strain is presented in Figure 5.5‎ -A. Maximal exo-PG activity of 266.0 ± 37.3 U/g was achieved after 8 days fermentation, which corresponded to a productivity of 33.3 U/g/d. At the same time a soluble protein content of 1.82 ± 0.01 mg/g and thus, a specific activity of 146.2 U/mg was obtained. These values are calculated per g biomass. Values calculated per g dried solids are presented in section 5.3.2.3.‎ Similar to the profile obtained by cultivation of mutant M3 was a change in parameters recognized after one day SSF. The oxygen consumption and carbon dioxide production reached their maximum also at the end of the first day, but oxygen consumption and carbon dioxide concentration measured in the exhaust gas were slightly higher compared to mutant M3. Similar trend was also observed in total soluble protein concentration, but with a significant lower maximal value of 2.00 ± 0.3 mg/g after 9 days SSF. An intensified total soluble carbohydrate consumption of 11.7 mg/g/d was followed between 1.3 to 5.3 days of cultivation. Afterwards it remained almost constant with its minimum of 8.3 mg/g at the 8th day of SSF. Production of glucosamine was measured after the first fermentation day and till 6.1 days SSF. Then it seemed to be at steady state with a maximum value of 39.2 ± 0.1 mg/gds at the 8th fermentation day. Similar values of glucosamine concentration were obtained in the cultivation of mutant M3. The pH value in the crude extracts also increased in the second half of fermentation period to pH 4.8 ± 0.4 at the 9th SSF day.

119

60 50 40 30 20 10 0

Chapter 5 7 6 5 4 3 2 1 0

0 100 200 300 500 400 500 5 80 10

70 9 400 4

60 8

50 7 300 3

40 6

5

200 (%) moisture basis 2 -

MW PG activity (U/g) activity PG 30

5 4 3 2 1 0 Wet

Soluble proteins (mg/g) proteins Soluble 4

MW MW MW MW MW

20

100 0 3 1 100 200 500 300 400 60 50 10 40 30 20 10 0

2 9 70 60 50 40 30 20 10 0

7 6 5 4 3 2 1 0

0 0 0 1 0 5 4 3 2 1 0 8 60 50 40 30 20 10 0

0 1 2 3 4 5 6 7 8 9 100 60 50 40 30 20 10 0 200 300 500 400 Time (d) 0 10

0 7 6 5 4 3 2 1 0 7 7 6 5 4 3 2 1 0

A 100 200 300 500 400 0 9 60 0 100 6 200 300 500 400 100 200 300 500 400 9 6 8 10 10 50 5 8 9 7 9 5 4 8 7 6

40 8 4 3 7 5 7

6 )

30 -

3 2 6

4 pH ( pH MW MW MW MW MW 6

5 1 3 5

20 5 2 4 2 4 0 MW MW MW MW MW

4 Soluble carbohydrates (mg/g), (mg/g), carbohydrates Soluble

MW MW MW MW MW Glucosamine content (mg/gds) content Glucosamine

10 1 1 3 3 3

0 0 0 2 2 0 1 2 3 4 5 6 7 8 9 2 1 Time1 (d) B 1 0 Figure 5‎ .5 Cultivation profile of A. sojae ATCC0 20235 at rotating drum type solid-state bioreactor; A: Time courses of process parameters, total soluble protein content (——), exo-PG activity (——) 0 and wet-basis moisture (——); B: Time courses of soluble carbohydrate content (——), glucosamine content (——) and pH (——).

Furthermore, an increase in the initially adjusted moisture content was recognized after medium sterilization, too. During fermentation the wet-basis moisture also increased at the beginning, but it did not decrease intensified at the end of

120 Chapter 5 cultivation like during SSF of mutant M3. A deepening comparison in product yield between A. sojae ATCC 20235 and its mutant M3 is given below.

5.3.2.3 Comparison of SSF at culture flask and bioreactor level Previous comparison in exo-PG production between the wild strain and the descending mutant M3 utilizing grinded inducer substrate at culture flask level resulted with 1087.2 ± 151.9 U/g dry WS in 1.7 fold increased enzyme activity by the mutant M3 (chapter 4). Also the comparison at culture flask level with changed substrate pretreatment with regard to the study at bioreactor level clearly favored mutant M3 as production organism (section 5.3.1‎ ). The results of the culture flask comparison with changed substrate pretreatment and the yields obtained at bioreactor level after 8 days SSF are summarized in Table 5‎ .3.

Table 5‎ .3 Comparison of SSF at culture flask and at bioreactor level.

Culture flask Bioreactor

Substrate (g dry WS) 1 10 1000

Scaling ratio 1 100

Agitation Manual shaking 2 Intermittent mixing 2

3 Forced aeration (L/min/kg dry WS)1 No 2  5

Products at 8 d SSF per g dry WS 1 per g biomass per g dried solids

Soluble protein (mg) 5.99 1.82 5.68 A. sojae ATCC 20235 PG activity (U) 447.9 266.0 829.7

Soluble protein (mg) 3.96 2.29 6.51 Mutant M3 PG activity (U) 1009.4 420.4 1194.2

1 dry WS – dry weight of substrate 2 Agitation twice at the day of inoculation and twice at the first day of cultivation; static conditions at further fermentation periods 3 Shift of the aeration rate at the first cultivation day

In order to compare the values achieved at culture flask level, which were calculated per g dry WS, with the yielded activities at bioreactor level, the values from bioreactor studies were calculated per gds (gram dried solids). Even if these values are not identical, they are similar and thus, they were used for the comparison.

121 Chapter 5

Comparing the values obtained by cultivation of A. sojae ATCC 20235, which are presented in Table 5.3‎ , similar total soluble protein content values were achieved at culture flask and at bioreactor level. However, the produced exo-PG activity at bioreactor level was approximately 1.9 fold higher and thus, indicating a predominant PG enzyme production at bioreactor level under the described conditions. While the PG yield obtained by mutant M3 was similar at both levels, but the total soluble protein content was approximately 1.6 times higher at bioreactor level. Hence, cultivation of mutant M3 in the bioreactor increased the protein production. Commercially available pectinolytic enzyme preparations are usually mixtures of several pectinases, and are associated with cellulytic, proteolytic and other species of enzymes apart from the main pectinases [29]. Thus, the higher protein production by mutant M3 might be even beneficial if other useful enzyme activities were produced. Nevertheless, the exo-PG yield obtained by cultivation of mutant M3 was significantly higher at all levels in comparison to the wild strain and no loss in enzyme activity was observed during the scale up of the PG production process. Thus, indicating the potential of this strain for enzyme production at large scale application.

5.4 Conclusions Optimization of SSF parameters for enhanced PG production by mutant M3 utilizing a changed substrate pretreatment procedure, which is more cost-efficient in large scale applications, resulted in significant higher product titers in comparison to the enzyme production by the wild strain. However, except the substrate pretreatment optimized parameters did not differ from previously optimized process conditions for exo-PG production by the wild strain (chapter 3). Moreover, the scale up of the optimized process was successfully implemented at laboratory bioreactor level without losses in PG yield or productivity. Hence, utilization of the previously optimized medium applying agricultural and agro- industrial residues for pectinase production was effectively proved at bioreactor level. The combination of cost-efficient substrate and high productivity by the applied strain are promising for large scale production processes. Besides that, the application of the rotating drum type solid-state bioreactor was successfully used for PG enzyme production by A. sojae and thus, the scale up of the SSF process at this type of bioreactor.

References [1] Viniegra-González, G, Favela-Torres, E, Aguilar, CN, Rómero-Gomez, SdJ, Díaz-Godínez, G, Augur, C (2003). Advantages of fungal enzyme production

122 Chapter 5

in solid state over liquid fermentation systems. Biochemical Engineering Journal 13, 157-167. [2] Kirk, O, Borchert, TV, Fuglsang, CC (2002). Industrial enzymes applications. Current Opinion in Biotechnology 13, 345-351. [3] Jayani, RS, Saxena, S, Gupta, R (2005). Microbial pectinolytic enzxmes: A review. Process Biochemistry 40, 2931-2944. [4] Singh, SA, Ramakrishna, M, Appu Rao, AG (1999). Optimisation of downstream processing parameters for the recovery of pectinase from the fermented bran of Aspergillus carbonarius. Process Biochemistry 35, 411-417. [5] Naidu, GSN, Panda, T (1998). Production of pectolytic enzymes - a review. Bioprocess Eng. 19, 355-361. [6] Hölker, U, Lenz, J (2005). Solid-state fermentation - are there any biotechnological advantages? Current Opinion in Microbiology 8, 301-306. [7] Hölker, U, Höfer, M, Lenz, J (2004). Biotechnological advanteges of laboratory-scale solid-state fermentation with fungi. Applied Micriobiology and Biotechnology 64, 175-186. [8] Pandey, A, Selvakumar, P, Soccol, CR, Nigam, P (1999). Solid-state fermentation for the production of industrial enzymes. Curr. Sci. 77, (1) 149- 162. [9] Pandey, A (2003). Solid-state fermentation. Biochemical Engineering Journal 13, 81-84. [10] Pandey, A (1992). Recent process developments in solid-state fermentation. Process Biochemistry 27, 109-117. [11] Doelle, HW, Mitchell, DA, Rolz, CE (1992). Solid substrate cultivation. Elsevier Applied Science: England. [12] Niture, SK (2008). Comparative biochemical and structural characterizations of fungal polygalacturonases. Biologia 63, (1) 1-19. [13] Chang, P-K, Matsushima, K, Takahashi, T, Yu, J, Abe, K, Bhatnagar, D, Yuan, G- F, Koyama, Y, Cleveland, TE (2007). Understanding nonaflatoxigenicity of Aspergillus sojae: a windfall of aflatoxin biosynthesis research. Applied Micriobiology and Biotechnology 76, 977-984. [14] Machida, M, Asai, K, Sano, M, Tanaka, T, Kumagai, T, Terai, G, Kusumoto, K-I, Arima, T, Akta, O, Kashiwagi, Y, Abe, K, Gomi, K, Horiuchi, H, Kitamoto, K, Kobayashi, T, Takeuchi, M, Denning, DW, Galagan, JE, Nierman, WC, Yu, J, Archer, DB, Bennett, JW, Bhatnagar, D, Cleveland, TE, Fedorova, ND, Gotoh, O, Horikawa, H, Hosoyama, A, Ichinomiya, M, Igarashi, R, Iwashita, K, Juvvadi, PR, Kato, M, Kato, Y, Kin, T, Kokubun, A, Maeda, H, Maeyama, N, Maruyama, J-i, Nagasaki, H, Nakajima, T, Oda, K, Okada, K, Paulsen, I, Sakamoto, K, Sawano, T, Takahashi, M, Takase, K, Terabayashi, Y, Wortman, JR, Yamada, O, Yamagata, Y, Anazawa, H, Hata, Y, Koide, Y, Komori, T, Koyama, Y, Minetoki, T, Suharnan, S, Tanaka, A, Isono, K, Kuhara, S,

123 Chapter 5

Ogasawara, N, Kikuchi, H (2005). Genome sequencing and analysis of Aspergillus oryzae. Nature 438, 1157-1161. [15] Singhania, RR, Patel, AK, Soccol, CR, Pandey, A (2009). Recent advances in solid-state fermentation. Biochemical Engineering Journal 44, 13-18. [16] Mitchell, DA, Krieger, N, Berovic, M (2006). Solid-state fermentation bioreactors. Springer-Verlag Berlin Heidelberg. [17] Bhargav, S, Panda, BP, Ali, M, Javed, S (2008). Solid-state fermentation: An overview. Chemical and Biochemical Engineering Quarterly 22, (1) 49-70. [18] Durand, A (2003). Bioreactor design for solid state fermentation. Biochemical Engineering Journal 13, 113-125. [19] Savergave, LS, Gadre, RV, Vaidya, BK, Narayanan, K (2011). Strain improvement and statistical media optimzation for enhanced erythritol production with minimal by-products from Candida magnoliae mutant R23. Biochemical Engineering Journal 55, 92-100. [20] Heerd, D, Yegin, S, Tari, C, Fernandez-Lahore, M (2012). Petinase enzyme- complex production by Aspergillus spp in sold-state fermentation: A comparative study. Food and Bioproducts Processing 90, 102-110. [21] Panda, T, Naidu, GSN, Sinha, J (1999). Multipleresponse analysis of pectinolytic enzymes by Aspergillus niger: a statistical view. Process Biochemistry 35, 187-195. [22] Bradford, MM (1976). A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Analytical Biochemistry 72, 248-254. [23] Dubois, M, Gilles, KA, Hamilton, JK, Rebers, PA, Smith, F (1955). Colorimtric method for determination of sugars and related substances. Analytical Chemistry 28, 350-356. [24] Zamani, A, Jeihanipour, A, Edebo, L, Niklasson, C, Taherzadeh, MJ (2008). Determination of glucosamine and N-acetyl glucosamine in fungal cell walls. Journal of Agricultural and Food Chemistry 56, 8314-8318. [25] Lonsane, BK, Saucedo-Castaneda, G, Raimbault, M, Roussos, S, Viniegra- González, G, Ghildyal, NP, Ramakrishna, M, Krishnaiah, MM (1992). Scale-up strategies for solid state fermentation systems. Process Biochemistry 27, 259-273. [26] Prior, BA, Du Preez, JC, Rein, PW (1992). Environmental parameters. In: H.W. Doelle; D. Mitchell; C.E. Rolz, eds. Solid substrate cultivation. Elsevier Science Publishers LTD: Essex, England. [27] Desgranges, C, Vergoignan, C, Georges, M, Durand, A (1991). Biomass estimation in solid state fermentation I. Manual biochemical methods. Appl. Biochem. Biotechnol. 35, 200-205.

124 Chapter 5

[28] Sardjono, Zhu, Y, Knol, W (1998). Comparison of fermentation profiles between lupine and soybean by Aspergillus oryzae and Aspergillus sojae in solid-state culture systems. J. Agr. Food Chem. 46, 3376-3380. [29] Del Cañizo, AN, Hours, RA, Miranda, MV, Cascone, O (1994). Fractionation of fungal pectic enzymes by immobilized metal ion affinity chromatography. J. Sci. Food Agric. 64, 527-531.

125 Chapter 6

Chapter 6.

Separation, purification and partial characterization of polygalacturonase derived from solid-state culture of

Aspergillus sojae

Abstract

Polygalacturonase belongs to the group of pectic substances degrading enzymes. Most microbial pectinases have been purified by means of chromatographic methods. However, the downstream processing procedure in solid-state fermentation systems starts with a leaching step to separate the desired product from the fermented solids. During this study the efficiency of the leaching process was optimized for polygalacturonase extraction by testing the extraction efficiency of several solvents and optimizing the leaching process parameters with advanced statistical design tools. Water was identified as most efficient leaching solvent. Optimization studies revealed optimized leaching conditions at a solvent-to-solid ratio of 4:1 in combination with a mixing rate of 250 rpm for 20 min contact time at room temperature for enhanced enzyme leaching. Polygalacturonase purification from the crude extract was conducted by combination of different chromatographic techniques using unique combinations of protein characteristics for the enzyme isolation, the so-called chromatographic “fingerprint”. Separation of proteins derived from solid-state culture of Aspergillus sojae ATCC 20235 by combination of ion exchange chromatography, size exclusion chromatography and hydrophobic interaction chromatography yielded in the successful isolation of polygalacturonase by means of a single band on SDS-polyacrylamide gel with a molecular weight of about 40 kDa. Additionally, the scale up of the first downstream processing step of ion exchange chromatography was demonstrated applying a scaling factor of 25 and developing a step-wise elution procedure for the recovery of polygalacturonase.

126 Chapter 6

6.1 Introduction Polygalacturonases (PG) belong to the group of pectic substances degrading enzymes, pectinases, which are industrially interesting for a growing number of different application, e.g. as processing aids for extraction, clarification, and maceration purposes [1]. It is reported that pectinases hold a share of 25 % in the global sales of food enzymes, which is expected to increase over time with the invention of new application areas [2]. Pectinolytic enzymes used in food processing predominantly derive from fungal source, especially from Aspergillus species [3, 4]. Filamentous fungi are known to produce pectinolytic enzymes either via submerged fermentation (SmF) and solid-state fermentation (SSF) [5-7]. Whereas, SSF offers various potential advantages in comparison to SmF, such as higher product concentration, simpler fermentation technology, and reduced waste-water output [8]. In contrast to SmF processes, the downstream processing procedure starts with a leaching step to separate the desired product from the fermented solids in SSF systems. The efficiency of the leaching process is affected by a number of factors such as efficiency of the solvent, diffusivity of solute and solvent, retention of solvents by solids, mixing of solids and solvents, solid to solvent ratio, contact time, contact temperature and effect of pH on the system [9]. Lonsane et al. [9] described a number of leaching techniques including percolation, multiple-contact counter-current leaching, pulsed plug-flow extraction in column, hydraulic pressing, and supercritical fluid extraction; and advised to the importance of selection the leaching method in terms of practicability and economics. The extract from leaching may be subjected to further downstream processing. Generally, this crude extract contains various soluble matters including the desired product and co-metabolites; and it also contains suspended solids such as finer particles, microbial cells and spores [10]. The characterization of the biotechnologically produced microbial enzymes requires the isolation of the desired protein. Purification prior characterization of the enzyme of interest is essential to avoid stabilizing effects of compounds, which might be present in the crude extract and thus, significantly influence enzyme properties [11]. Most microbial pectinases have been purified by means of chromatographic methods [12]. The isolation of pectinases was conducted by combinations of different chromatographic techniques, such as ion exchange (IEXC), hydrophobic interaction (HIC) and size exclusion chromatography (SEC) [12- 16], while few purification methods of pectinases contained a single step like three- phase partitioning [17] or affinity adsorption chromatography [18]. Whereat the method of affinity chromatography (AC) applied by Camperi et al. [18] was specially

127 Chapter 6 designed for the removal of pectinesterase from a commercial pectinolytic enzyme preparation. The separation of proteins applying various chromatographic techniques is based on unique combinations of protein characteristics, the so-called chromatographic “fingerprint”. IEXC separates proteins on the basis of their charge, SEC separates on the basis of size and shape, HIC separates on the degree of protein hydrophobicity, and AC is based on a reversible interaction between the protein and a specific ligand. Although chromatographic techniques effectively separate various proteins, however the recovery of the target protein might be significantly less than the loaded amount. Hence, as few as possible chromatographic steps should be applied for the purification of the desired protein. Moreover, chromatographic steps significantly increase downstream processing costs and thus, influences the overall process economics [19]. The objective of the present study was the recovery of PG enzyme produced by A. sojae in SSF. A fractionation of major exo-PG activity from crude extract with good recovery was achieved by combination of various chromatographic techniques. Moreover, the scale up of IEXC process was investigated studying the effect of elution procedure on enzyme recovery.

6.2 Materials and Methods

6.2.1 Materials All chemicals were purchased from AppliChem GmbH (Darmstadt, Germany), except the substrate for detection of exo-PG activity, polygalacturonic acid, and the chemical sodium arsenate dibasic heptahydrate was obtained from Sigma-Aldrich Chemie GmbH (Steinheim, Germany). Dithiothreitol (DTT) and methanol were supplied by Carl Roth GmbH & CO. KG (Karlsruhe, Germany). Chromatographic resins for IEXC and HIC, as well as the pre-packed columns used for SEC and HIC were purchased from GE Healthcare (Freiburg, Germany).

6.2.2 Microorganism and cultivation conditions A. sojae ATCC 20235, purchased from Procochem Inc (Teddington, United Kingdom), was cultivated on wheat bran and sugar beet pulp in the ratio 70:30, wetted at 160 % with 0.2 M HCl, at 30 °C for PG production under optimized conditions as described in chapter 3. Fermented substrate was harvested after 8 days for enzyme leaching.

128 Chapter 6

6.2.3 Improvement of PG leaching Biomass obtained from cultivation of A. sojae was homogenized with a hand-held blender. Portions of 10 g biomass were weight into 125-mL culture flasks for the improvement of PG leaching conditions.

6.2.3.1 Extraction efficiency of different solvents Several solvents including water, acetate buffer (pH 4.8, 0.1 M); salts like sodium chloride and sodium sulfate; nonionic surfactants like Triton X-100 and Brij 35; and isopropyl alcohol were used to extract the enzyme from fermented substrate (Figure 6.2‎ ). Therefore, 10 g biomass was mixed with 20 mL solvent and incubated for 50 min at room temperature (23 °C) at 400 rpm. Supernatant was separated from biomass by centrifugation for 30 min at 3220 × g at 4 °C and analyzed for PG enzyme recovery and total protein content.

6.2.3.2 Optimization of leaching conditions The improvement of leaching conditions for enhanced PG extraction was performed by applying statistical tools. During the screening experiment, the effect of solid-solvent contact time, contact temperature, solid to solvent ratios, and the mixing of solids and solvents were explored utilizing a full factorial design at two levels (Table 6.2‎ ). The design matrix consisted of 19 experimental runs, including 16 design runs and 3 center point repetitions. In the next step response surface methodology (RSM) was used for optimization. Hence, the optimal factor settings of significant factors affecting the enzyme leaching process were investigated, using central composite face-centered (CCF) design. The factors solid-solvent contact time, solid to solvent ratio, and the mixing of solids and solvents were explored in 17 experiments including 3 repetitions of the center point. Data analysis was performed by using multiple linear regression (MLR) applying the MODDE 9.0 software package, supplied by Umetrics AB, Umeå, Sweden.

6.2.3.3 Scale up of PG leaching Biomass used for scaling up of the leaching conditions was produced at a rotation drum type solid-state bioreactor (Terrafors-IS, Infors HT), utilizing 1 kg dry substrate of wheat bran and sugar beet pellets in the ratio 70 : 30 for PG enzyme production. Cultivation conditions are described in chapter 5 and Appendix B. The enzyme leaching was performed keeping the previously optimized solvent to solid ratio of 4:1 and using a stirrer (RZR 2051 control, Heidolph) with an impeller (PR 39 pitched blade, Heidolph) for suspension and homogenization at 700 rpm and at room

129 Chapter 6 temperature. Incubation time was optimized by analyzing samples taken at intermittent time intervals over a period of 2 h.

6.2.4 Separation of solids from crude extract and enzyme concentration Crude extracts obtained from leaching of biomass produced at culture flask level were separated from the fermented substrate by centrifugation at 4 °C, 3220×g, for 30 min. The supernatant was dialyzed against distilled water over night at 4 °C, using SnakeSkin® pleated dialysis tubing, 10000 MWCO (Thermo Scientific, Rockford, USA). Samples were centrifuged at 4 °C, 6000×g, for 20 min and the supernatant was freeze-dried for storage until use. The removal of solids at larger volumes of crude extract, which were produced at bioreactor level, was achieved by decanting and filtration through cotton cloth using a dead-end filtration technique. The filtrate, which still contained fine solid particles, was subjected to cross flow filtration, where the fluid feed stream ran tangential to a membrane, utilizing the hollow fiber cartridge UFP-500-C-6A (GE Healthcare, Life Sciences). This was followed by ultrafiltration (UF) and diafiltration (DF) using hollow fiber cartridge UFP-10-C-6A (GE Healthcare, Life Sciences) to concentrate the sample and to transfer the PG enzyme into 0.02 M acetate buffer (pH 5.0). The scheme of filtration processing is illustrated in Figure 6‎ .1. The material to be filtrated by membrane filters was applied to the filtration system under pressure in order to achieve satisfactory flow rates by use of a suitable pump. The performance parameters were calculated as follows:

Microfiltration (MF) C V Recovery (%) = P P × 100 (6.1) CF VF

C V – C V Elimination (%) = F F P P × 100 (6.2) CF VF

Ultrafiltration (UF) C V Recovery (%) = R R × 100 (6.3) CF VF

C V – C V Elimination (%) = F F R R × 100 (6.4) CF VF

V Volume concentration ratio (VCR) = F × 100 (6.5) VR

130 Chapter 6

Where CF, CP and CR are the enzyme activity (U/mL) or protein content (mg/mL) in feed, permeate and retentate. VF, VP and VR are the volume of feed, permeate and retentate, respectively. Overall enzyme and protein recovery were estimated with reference to the crude extracts.

Crude extract

Dead-end filtration Solid particles, through cotton cloth larger impurities

Filtrate containing PG and impurities (fine solid particles) MF 500 kDa Fine solid particles, other impurities

0.02 M acetate buffer (pH 5.0) MF filtrate containing PG and (DF) impurities (non-PG proteins etc)

UF 10 kDa Concentrated PG

Smaller impurities Figure 6‎ .1 Scheme of filtration process for recovery and concentration of PG enzyme from SSF crude extract.

6.2.5 Enzyme purification Concentrated PG extracts produced by A. sojae were subjected to a diversity of chromatographic procedures, including IEXC, SEC and HIC, for the fractionation of PG enzyme. Column chromatography was performed at room temperature using a GE Healthcare ÄKTA FPLC and ÄKTA explorer system.

6.2.5.1 Ion exchange chromatography (IEXC) Concentrated PG extract was loaded on a Pharmacia C 10/10 column (GE Healthcare) containing 5 mL diethylaminoethyl (DEAE) sepharose Fast Flow resin (GE Healthcare) equilibrated in 0.02 M sodium acetate buffer (pH 5.0). After washing with 20 column volumes (cV) of equilibration buffer, the bound PG was eluted by a linear concentration gradient of 0.0 – 0.5 M NaCl in the buffer in 40 cV. Column chromatography was performed at a volumetric flow rate of 2 mL/min. Samples

131 Chapter 6 were collected in 2 mL fractions and analyzed for exo-PG recovery (section 5.2.6‎ ) and protein content (section 5.2.7‎ ).

6.2.5.2 Scale up IEXC The performance of anion-exchange chromatography at larger scale was conducted increasing the column diameter while keeping the bed height constant (Table 6‎ .1). The expansion of column diameter by factor 5 resulted in an enlargement of column volume by factor 25:

Column volume cV = π * r² * h (6.6)

Where r is the inner column radius and h represents the bed height. In order to assure the same linear flow (Y) at both scales the volumetric flow rate (Z) had to be increased by factor 25 at the larger column volume [20]:

Y π * d² Volumetric flow rate Z (mL/min) = * (6.7) 60 4

Where Y means linear flow (cm/h) and d is the inner diameter of the column.

Table 6‎ .1 Scale up parameters of the anion-exchange chromatographic process.

Parameter C 10/10 column XK 50 column

Column volume (mL) 5 125 Scaling ratio 1 25

inner diameter (cm) 1 5

Bed height (cm) 6.37 6.37

Volumetric flow rate (mL/min) 2 50

Linear flow (cm/h) 153 153

6.2.5.3 Size exclusion chromatography (SEC) Concentrated sample of 1.1 % cV was loaded onto a Superdex 200 10/300 GL (GE Healthcare) column and proteins were eluted by flushing with 0.05 M sodium acetate buffer (pH 5.0) containing 0.2 M L-arginine hydrochloride at a volumetric flow rate of 0.1 mL/min. Fractions of 1 mL volume were collected and analyzed for exo-PG recovery (section 5.2.6‎ ) and protein content (section 5.2.7‎ ).

6.2.5.4 Hydrophobic interaction chromatography (HIC) Sample subjected to HIC was previously adjusted to contain 1.7 M ammonium sulfate. It was then applied to a Tricorn column (GE Healthcare), which was packed

132 Chapter 6 with 2 mL phenyl sepharose FF (GE Healthcare). HIC medium was equilibrated with 0.05 M sodium acetate buffer (pH 5.0) containing 1.7 M ammonium sulfate. Elution was carried out with a linear concentration gradient of ammonium sulfate from 1.7 to 0.0 M in the same equilibration buffer in 40 cV with a volumetric flow rate of 1 mL/min. Fractions of 2 mL volume were collected and analyzed for exo-PG recovery (section 5.2.6‎ ) and protein content (section 5.2.7‎ ).

6.2.6 Exo-polygalacturonase assay Exo-PG activity was assayed according to the procedure of Panda et al. [21], which was further optimized as described in Appendix A.

6.2.7 Soluble protein content Total extracellular protein was measured according to the modified Bradford´s method [22] as described in Appendix A.

6.2.8 Electrophoresis SDS-PAGE was performed according to the method of Laemmli [23], following the procedure described in the technical manual for protein electrophoresis [24]. Briefly: 12.5% SDS-PAGE gel with an approximately 2 cm stacking buffer zone was cast and samples run in constant current mode at 15 mA/gel, at room temperature. Samples were mixed with 2x treatment buffer in the ratio 2:1. Sample load add up to 10 µL per lane. Protein bands were visualized, using colloidal Coomassie (G-250) staining [25].

6.3 Results and Discussion Downstream processing involves isolation of the product and conversion into a marketable form. In SSF systems the product, e.g. enzymes, must be initially leached from the fermented solids and recovered in a suitable solvent. Therefore, several parameters affecting the leaching process have been investigated in order to enhance the leaching of exo-PG from the fermented solids.

6.3.1 Improvement of enzyme leaching Leaching is an important unit operation in downstream processing, where the soluble product will be recovered from the fermented mass in form of a crude extract using an appropriate solvent. It is essential to leach out as much as possible of the product from the solids, as this increases the economic feasibility of the SSF process. In terms of proteins the solubility depends on pH and ionic strength. Generally, the pH of leaching solvent should be different from the isoelectric point of the desired protein to maintain the proteins charge, and salts present in the

133 Chapter 6

solvent increases the solubility of the charged protein [26]. Different solvents were

Protein content (mg/mL) content Protein Protein content (mg/mL) content Protein tested for their leaching efficiency(mg/mL) for thecontent extractionProtein of PG (Figure 6.2‎ ). 6 5 4 3 2 1 0 6 5 4 3 2 1 0 6 5 4 3 2 1 0

220 6

5%T + 3%P + 5%T 5%T + 3%P + 5%T 200 3%P + 5%T

180 5 5%T + 1%P + 5%T 5%T + 1%P + 5%T

160 1%P + 5%T

3%T + 1%P + 3%T 3%T + 1%P + 3%T 3%T + 1%P + 3%T 4

140

3%T + 3%P + 3%T 3%T + 3%P + 3%T 120 3%P + 3%T

3

100 2-Propanol 3% 3% 2-Propanol 3% 3% 2-Propanol 3%

80

2 Triton 3% 3% Triton 3%

60 Triton 3%

B F D

Protein content (mg/mL)

B F D Ttriton 1%

B F D

exo-PG activity (U/mL), exo-PGactivity Specific activity (U/mg) Specificactivity 1% Ttriton 1% Ttriton 1%

40 1

1% Brij 1%

1% Brij 1% 20 Brij 1%

0 0

Triton X Na2SO4 1%

Triton X

Triton X

Brij

Triton X Triton X

Sodium sulfate (1 sulfate Sodium

Acetate buffer (pH 4.8, 0.1 M) 0.1 4.8, (pH bufferAcetate chloride Sodium

Water %) (3 Isopropyl alcohol Triton X 1% Na2SO4 1%

1% Na2SO4 1%

35 (1 %) 35(1

water NaCl 1%

1% NaCl 1%

1% Brij NaCl 1%

-

-

-

-

-

1% NaCl

-

100 (3%) 100

100 (1%) 100

100 (3 %) + Isopropyl alcohol (1 %) (1 alcohol Isopropyl + (3%) 100

100 (3 %) + Isopropyl alcohol (3 %) (3 alcohol Isopropyl + (3 %) 100 %) (3 alcohol Isopropyl + (5%) 100

100 (5 %) + Isopropyl alcohol (1 %) (1 alcohol Isopropyl + (5%) 100

3% Triton

1% Ttriton

3%T + 3%P 3%T + 1%P 5%T + 1%P 5%T + 3%P 1% Na2SO4 acetate buffer acetate

acetate buffer acetate buffer acetate

3% 2-Propanol buffer acetate

(1 %) (1 water water %) water 0 0 0 20 40 80 60

20

120 40 80 60 140 180 160 100 220 20 40 200 80 60

120

140 180 160 100 220 120

200 140 180 160 100 220

200 (U/mg) activity Specific

Specific activity (U/mg) activity Specific Specific activity (U/mg) activity Specific

exo-PG activity (U/mL), activity exo-PG exo-PG activity (U/mL), activity exo-PG exo-PG activity (U/mL), activity exo-PG

Figure 6‎ .2 Leaching efficiency of various solvents for the extraction of PG enzyme.

Highest PG activity of 175.4 ± 0.7 U/mL was measured in the crude extract leached out with the solvent containing 5 % Triton X-100 in combination with 1 % isopropyl alcohol. Nevertheless, the overall variation of exo-PG activity values measured in the different solvents after enzyme leaching was relatively low. Ideally, the solvent should selectively leach out the desired product completely at room temperature within minimal contact time and preferably at the pH of the moist solids [27]. Besides PG activity, also the total protein content and the resulting specific activity were taken into account for choosing a suitable solvent. In contrast to the PG activity values a wide variation of specific activity values was observed in the different crude extracts. The selective recovery of the desired product is an important aspect in downstream processing. Leaching out of high amounts of impurities, which were measured as high protein content in combination with low

134 Chapter 6 specific activity might strongly complicate further purification steps. Hence, besides high PG activity values also high specific activity was desired. Considering the solvent yielding in highest PG activity leached out also highest amount of total protein content and resulted therefore in lowest specific activity. Generally, solvents containing nonionic surfactants showed an increased leaching efficiency for proteins, but seemed not to be suitable for selective extraction. Highest specific activity of 207.1 ± 10.6 U/mg was obtained applying water as solvent. This resulted also in high exo-PG activity of 158.9 ± 5.4 U/mL. Since water represented also the most cost-efficient solvent, water was utilized as solvent in further enzyme leaching studies. Further PG leaching studies were conducted applying design of experiment (DoE) in order to optimize the enzyme leaching conditions. Exo-PG activity values and protein content were calculated per gram utilized fermented solids. Calculation of values per milliliter of supernatant would lead to wrong conclusions since the volume of the applied solvent was varied in the following experiments and had to be taken into account. During a screening for the effect of the leaching parameters contact time, incubation temperature, solvent to solid ratio and mixing rate on the PG activity obtained in the crude extracts a full factorial design was applied (Table 6.2‎ ). Besides the response, exo-PG activity, also the total protein content and resulting specific activity were determined in the different crude extracts. The mixing of solids and solvent was achieved at laboratory scale with an incubator shaker, since agitation during leaching ensures good contact of substrate with solvent thereby increasing the efficiency of leaching [9]. Moreover, the better performance of agitated systems over static systems was demonstrated for leaching of several enzymes from solid particles [28, 29]. The MLR modeling of the screening data gave a model with R2 = 0.811 and Q2 = 0.596, suggesting a sound model. The linear model showed no lack of fit (LoF) (p = 0.105), because its p-value was larger than the critical reference value of 0.05, and normally distributed residuals (plot not shown). Furthermore, the regression model was statistical significant (p = 0.015), which is satisfied when p < 0.05. Hence the model had a small model error and good fitting power. Looking at the exo-PG values presented in Table 6‎ .2 a wide variation in the response values was observed ranging from 0 U/g above 330 U/g. This variation showed the importance of optimizing these parameters for improving PG leaching from fermented solids. Furthermore, it can be seen that long contact time in combination with high incubation temperature resulted in decreased exo-PG activity values, while longer incubation time at lower incubation temperature yielded in higher response values. Whereof, high temperature level for short incubation times yielded

135 Chapter 6 high enzyme activities in the crude extracts. Thus, with regard to the findings of Castilho et al. [30], the combined effect of temperature and contact time significantly influenced the pectinase extraction. Moreover, this indicated that the stability of the enzyme is influenced at the higher temperature level. Highest enzyme activity of 332.1 U/g was leached at 15 °C at high mixing rate and long incubation time (Table 6‎ .2, Exp. No. 6).

Table 6‎ .2 Full factorial design and experimental results of the screening step for enhanced PG leaching conditions.

Factors Response Protein Specific Solvent exo-PG content activity Exp. No. Time Tempera- Mixing to solid activity (mg/g) (U/mg) (min) ture (°C) (rpm) ratio (U/g) 1 10 15 40 1.5 175.7 2.07 84.9

2 60 15 40 1.5 227.1 2.18 104.1

3 10 45 40 1.5 224.8 2.24 100.4

4 60 45 40 1.5 30.2 2.27 13.3

5 10 15 400 1.5 320.6 3.17 101.2

6 60 15 400 1.5 332.1 3.60 92.4

7 10 45 400 1.5 308.7 3.06 101.0

8 60 45 400 1.5 25.4 2.97 8.5

9 10 15 40 5.0 200.7 2.30 26.2

10 60 15 40 5.0 275.2 2.89 95.2

11 10 45 40 5.0 182.0 2.13 85.7

12 60 45 40 5.0 91.0 2.92 31.2

13 10 15 400 5.0 183.6 3.16 58.1

14 60 15 400 5.0 238.8 3.64 65.5

15 10 45 400 5.0 205.6 3.06 67.2

16 60 45 400 5.0 0.0 3.53 0.0

17 – 19* 35 30 220 3.25 319.8 ± 6.6 3.52 ± 0.02 90.7 ± 2.4 * Three repetitions of the center point experiments.

136 Chapter 6

The variation in total protein content was low in comparison to the observed variation by changing the solvent type. Hence, specific activity values strongly depended on the observed exo-PG activity. Generally, high response values were obtained at the high mixing level except under the mentioned combination of high temperature and long contact time. Nevertheless, the PG activities obtained in the center point experiments were also very high, which were conducted at 30 °C at medium mixing rate and incubation time. Hence, the influence of temperature seemed not to be important as long as the enzyme stability is not affected. Therefore, the incubation temperature was fixed at room temperature for further leaching studies, which might be the optimum with regard to process economics. Furthermore, values from the center point experiments indicated that the levels of the factors mixing rate and incubation time could be decreased. Hence, optimal settings for enhanced PG leaching of these factor levels and also of the factor solid to solvent ratio was investigated applying response surface technology (Table 6‎ .3). The MLR modeling of the response surface data gave a model with R2 = 0.862 and Q2 = 0.612, upon removal of insignificant interaction terms. The quadratic model showed no lack of fit (p = 0.068), and normally distributed residuals (plot not shown). The regression model was statistical significant with p < 0.05, hence the model had a small model error and good fitting power. In comparison to the previous screening experiments was the distribution of obtained exo-PG activity values in a smaller range and except Exp. No. 14 were all values above 200 U/g exo-PG activity. Hence, the chosen factor levels facilitated PG leaching from the fermented solids. Highest exo-PG activity of 308.8 U/g was slightly below the highest result of the screening experiments. However, it has to be taken into account that the fermented solids used for this study were obtained from another solid-state fermentation batch. Hence, even applying exactly the same fermentation conditions as described in section 6.2.2‎ might result in low variations from batch to batch in cultivation experiments. Similar to the results of the screening experiments was the variation in total protein content relatively low and the specific activity mainly depended from the observed exo-PG activity in the crude extracts. These findings were confirmed comparing the contour plots of exo-PG activity and specific activity (Figure 6‎ .3 and Figure 6‎ .4).

137 Chapter 6

Table 6‎ .3 CCF design and experimental results of the optimization step for enhanced PG leaching conditions.

Factors Response Protein Specific content activity Exp. No. Time Mixing Solvent to exo-PG (mg/g) (U/mg) (min) (rpm) solid ration activity (U/g)

1 10 50 1.4 212.4 1.98 107.0

2 30 50 1.4 247.2 2.24 110.4

3 10 250 1.4 263.1 2.40 109.5

4 30 250 1.4 297.7 2.48 120.1

5 10 50 5.0 212.5 2.50 85.0

6 30 50 5.0 266.7 2.67 99.9

7 10 250 5.0 296.8 2.94 100.9

8 30 250 5.0 308.6 3.15 97.9

9 10 150 3.2 288.7 2.75 104.9

10 30 150 3.2 295.6 2.80 105.7

11 20 50 3.2 220.2 2.53 87.0

12 20 250 3.2 275.1 2.91 94.7

13 20 150 1.4 230.0 2.40 95.6

14 20 150 5.0 174.0 2.82 61.7

15 – 17* 20 150 3.2 236.1 ± 3.9 2.92 ± 0.08 81.1 ± 3.1

* Three repetitions of the center point experiments.

138 Chapter 6

exo-PG Mixing: 50 rpm Mixing: 150 rpm Mixing: 250 rpm activity (U/g) 4.5 4.5 4.5

4.0 4.0 4.0

3.5 3.5 3.5

3.0 3.0 3.0

2.5 2.5 2.5

Solventtosolid ratio

Solventtosolid ratio Solventtosolid ratio 2.0 2.0 2.0

1.5 1.5 1.5

Figure 6‎ .3 Contour plots illustrating the interaction of solvent to solid ratio and contact time at specific mixing rates on the leaching of PG from fermented solids.

According to the contour plots presented in Figure 6‎ .3, leaching of PG would be enhanced at high mixing rate, long contact time and over a broad range of solvent to solid ratio. Similar findings were observed in Figure 6‎ .4 for enhanced specific activity, while solvent to solid ratio should be preferable at the low to medium level.

Specific Mixing: 50 rpm Mixing: 150 rpm Mixing: 250 rpm activity (U/g)

4.5 4.5 4.5

4.0 4.0 4.0

3.5 3.5 3.5

3.0 3.0 3.0

2.5 2.5 2.5

Solventtosolid ratio

Solventtosolid ratio Solventtosolid ratio 2.0 2.0 2.0

1.5 1.5 1.5

Figure 6‎ .4 Contour plots showing the effect of solvent to solid ratio and contact time at specific mixing rates on obtained specific PG activity during enzyme leaching.

Predicting maximal exo-PG activity values in the crude extracts utilizing the optimizer of MODDE 9.0 (Umetrics AB, Umeå, Sweden) resulted in several experiments at a mixing rate of 250 rpm. Validation experiments were conducted at the suggested optimal factor settings, which are given in Table 6‎ .4.

139 Chapter 6

Table 6‎ .4 Validation experiments and results for enhanced PG leaching from fermented solids.

exo-PG Protein Specific Time Mixing Solvent to activity content activity (min) (rpm) solid ratio (U/g) (mg/g) (U/mg)

30 250 2.1 367.0 ± 6.3 2.53 ± 0.01 144.8 ± 1.8

30 250 4.3 405.7 ± 7.2 3.17 ± 0.06 127.9 ± 0.1

20 250 2.0 364.1 ± 7.6 2.55 ± 0.03 142.8 ± 4.4

20 300 2.0 363.6 ± 10.2 2.56 ± 0.05 142.1 ± 6.8

30 300 2.0 360.7 ± 12.7 2.58 ± 0.04 140.2 ± 7.1

20 250 4.0 411.8 ± 0.6 3.26 ± 0.00 126.3 ± 0.3

20 300 4.0 415.1 ± 5.2 3.40 ± 0.03 122.3 ± 2.5

Experiments were performed in duplicate.

Additionally, experiments including higher mixing rates of 300 rpm were conducted to examine the effect of high share rates during enzyme leaching. The applied biomass was obtained from another batch fermentation conducted according to the specifications given in section 6.2.2‎ . Highest PG activity of 415.1 ± 5.2 U/g was obtained at high mixing rate of 300 rpm, over a contact time of 20 min and with a solvent to solid ratio of 4:1. However, this only slightly differed from the value of 411.8 ± 0.6 U/g obtained under similar conditions at lower mixing rate of 250 rm. Furthermore, the increased mixing rate yielded in an increased content of soluble proteins in the extract and hence, in a decreased specific activity. Therefore, leaching conditions for extraction of PG from fermented solids were set at room temperature for 20 min contact time, at 250 rpm mixing rate and a solvent to solid ratio of 4:1, utilizing water as solvent. Optimization studies targeted enhanced PG leaching applying water as solvent. With regard to the scale up of the fermentation process into a laboratory fermenter (chapter 5) and process economics the impact of utilizing tap water compared to distilled water as solvent for PG leaching was studied. Therefore, 10 g fermented solids were leached out under optimized conditions applying distilled water and tap water as solvents. Utilizing distilled water and tap water as leaching solvent yielded in 85.6 ± 1.1 U/mL and 89.6 ± 2.7 U/mL PG activity, respectively. Hence, use of tap water as solvent for enzyme leaching did not negatively influence PG activity and could be applied as leaching solvent in the scaled up process.

140 Chapter 6

Furthermore, at bioreactor level experiments had to be conducted for the selection of a downstream processing strategy starting with the enzyme leaching at a bigger scale. The selection of a leaching method influences the economics of a SSF process and is also critical in terms of practicability. In order to keep the previously optimized solvent to solid ratio of 4:1, the leaching of bigger amounts of biomass was no more convertible at a laboratory shaker. Thus, the enzyme leaching was performed utilizing a stirrer with an impeller for homogenization and suspension. The incubation time was optimized for PG leaching from biomass by taking samples at intermittent time intervals during 2 h incubation period Figure 6‎ .5.

1.21,21,2 5050 4545 1,01,01.0 4040

3535 0,80.80,8 3030 2525 0,60.60,6 2020 0,40,40.4 1515 PG Specific Protein

PG Specific Protein PG Specific Protein

Protein content (mg/mL) content Protein

Protein content (mg/mL) content Protein

exo-PG activity (U/mL), exo-PGactivity

Specific activity (U/mg) Specificactivity

exo-PG activity (U/mL), exo-PGactivity Specific activity (U/mg) Specificactivity

1010 (mg/mL) content Protein

Protein content (mg/mL) content Protein Protein content (mg/mL) content Protein 0,20.20,2 55 1,2 1,0 0,8 0,6 0,4 0,2 0,0 1,2 1,0 0,8 0,6 0,4 0,2 0,0 1,2 0 1,0 0,8 0,6 0,4 0,2 0,0 0,00,00.0 0 20 40 60 80 100 120120 120 120 Time (min) 120 Figure 6‎ .5 Effect of contact time on enzyme leaching with agitator. 100 100 In Figure 6.5‎ can be seen that 1 h enzyme leaching yielded100 in maximal PG activity in the extract. Further incubation did not increase the enzyme activity in the extract, 80 80 but increased the extraction of other proteins from80 the biomass and thus the impurity of the crude extract. Hence, maximal specific activity in the crude extract 60 60 Time (min) Time was obtained after 1 h contact time under the described60 conditions. This time was Time (min) Time Time (min) Time used for the PG enzyme extraction from biomass obtained at bioreactor level. 40

As previously mentioned, a number of different leaching40 techniques have been 40 applied as methods for leaching the solutes from the fermented solids obtained by

SSF processes, and the transfer of the product from the solid phase into the 20 20 20 leaching solution is driven by concentration gradients [9]. Thus, the loss of solute

retained in the solids is observed in solid-liquid extraction [28]. Castilho et al. [28] 0 0 5 0 0 15 10 5 25 35

0 20 45 50 30 40

15

10

5 25 35 0 20 45 50 30 40

15 (U/mg) activity Specific

studied the recovery of exo-PG and endo-10 PMG enzymes obtained in solid-state 25 35 20 45 50 30 40 Specific activity (U/mg) activity Specific

Specific activity (U/mg) activity Specific

exo-PG activity (U/mL), activity exo-PG

culture by A. niger in a fixed bed(U/mL), system,activity stirredexo-PG tank system, repeated extraction exo-PG activity (U/mL), activity exo-PG and multiple stage counter-current extraction. They obtained 105 % exo-PG and 15 %

141 Chapter 6 endo-PMG activities in the stirred tank system superior the fixed bed extractor. Furthermore, they performed five successive washings of the solids with fresh solvent and recovered 84.2 % exo-PG activity in the first stage, 9.1% in the second, 3.6 % in the third and 2 % or less in the last two stages. Based on these results they studied counter-current extraction employing two and three stages. With two stages, an increase of 51.6 % exo-PG activity and 27.2 % endo-PMG activity was observed and three stages allowed an increase in exo-PG and endo-PMG of 81.3 % and 19.6 %, respectively, when compared to a system with only one stage. During the present study the aim was the maximal enzyme recovery in one leaching step. The application of counter-current extraction applying the optimized PG leaching conditions seems to be promising for increasing the PG yield during enzyme leaching. With regard to industrial application, counter-current extraction might improve process economics.

6.3.2 Performance of membrane filtration

6.3.2.1 Microfiltration of crude extract The crude extracts used for the study were mixtures of several crude extracts obtained from SSF at bioreactor level. These extracts were decanted and filtered through cotton cloth before subjecting to membrane filtration. The utilized crude extracts contained 65.5 and 39.4 U/mL exo-PG activity. Taking the total protein content into account, these extracts showed a specific activity of 99.2 and 44.3 U/mg, respectively. These extracts were subjected to membrane filtration with a membrane of 500 kDa MWCO. MF processing of the mentioned crude extracts yielded in 63 and 67 % recovery of exo-PG activity, and 57 and 65 % total protein concentration, respectively. Thus, the specific activity in the crude extract mixture was slightly increased to 108.0 and 45.3 U/mg respectively. Nevertheless, the loss of PG, indicated by the elimination of 37 and 33 % in the membrane filtration process, resulted in some extent by the separation of the retentate from the extract. The retentate, which did not pass through the membrane contained solid particles and larger impurities (Figure 6‎ .1). With regard to the total applied extract volumes, the retentate amounted to 7 and 14 %.

6.3.2.2 Ultrafiltration of crude extract The performance of UF process was initially tested with centrifuged crude extract before coupling this step after MF as presented in the scheme of the filtration process (Figure 6‎ .1). Therefore, centrifuged crude extract containing 57.1 U/mL exo- PG activity and 0.47 mg/mL total protein content was subjected to UF with a membrane of 10 kDa MWCO. The volume concentration ratio (VCR) of

142 Chapter 6 approximately 18 was reached. The retentate fraction was further subjected to discontinuous diafiltration in three cycles for removal of salts and transfer of PG enzyme into the buffer system applied for protein purification by ion-exchange chromatography. The final concentrate contained 676.2 U/mL of exo-PG activity and a protein content of 4.72 mg/mL. The recovery of PG was 65 % and 55 % for total protein. Hence, the specific activity was increased from 121.4 U/mg to 143.3 U/mg by UF processing. Moreover, PG activity and protein content were concentrated approximately 12 and 10 times, respectively.

6.3.3 Enzyme purification The objective of an enzyme purification is to get rid of as much unwanted impurities in the crude extract as possible while retaining the desired enzyme activity. The success of each step depends both on the retention of activity and the extent of improvement in specific activity. Separation of PG enzymes secreted by A. sojae was carried out by using different types of chromatography. In the first step the concentrated enzymatic extract was purified by ion exchange chromatography (IEXC). This downstream processing step was scaled up increasing the applied matrix volume by factor 25 and changing the gradient type from a linear gradient to a step gradient in order to concentrate the PG activity elution peak in the collected fractions and to reduce the process time. Partially purified IEXC fractions containing the elution peak of exo-PG activity were pooled together and subjected to size exclusion chromatography (SEC) and hydrophobic interaction chromatography (HIC) for further separation of PG enzyme from crude extract proteins. The summary of PG enzyme purification is presented in Table 6‎ .6.

6.3.3.1 Ion exchange chromatography IEXC is based on the reversible electrostatic attraction of a charged molecule to a solid matrix which contains covalently attached side groups of opposite charge. The elution of proteins can be caused by altering the pH, or by increasing the salt concentration of the buffer system [19]. IEXC was performed on an ion-exchange matrix which contained covalently attached positive groups and thus, is termed as anion exchanger. Hence, it will adsorb anionic proteins, i.e. proteins with a net negative-charge. Therefore, the concentrated PG extract (in 0.02 M acetate buffer, pH 5.0), which was obtained by MF, UF and DF as described in section 6.2.4,‎ containing 43 mg protein was loaded on 5 mL DEAE-Sepharose Fast Flow resin (GE Healthcare) pre-equilibrated with the

143 Chapter 6 same buffer. The elution profile, which was achieved by increasing the NaCl concentration in the buffer system, is presented in Figure 6.6‎ . 2011 11 11 UF crude Run6 Anion deae006:1_UV 2011 11 11 UF crude Run6 Anion deae006:1_Cond 2011 11 11 UF crude Run6 Anion deae006:1_Conc 2011 11 11 UF crude Run6 Anion deae006:1_Fractions PG 2011 11 11 UF crude Run6 Anion deae006:1_Inject PG mAU 350 30 1400

300 1200 25

1000 250

20 800 200

15

30 25 20 15 10 5 0

600 150 PG PG

PG activity (U/mL) activity PG -

30 25 20 15 10 5 0

Absorbance at 280 nm 280 at Absorbance Relative activity (%) activity Relative

0 exo 50 PG PG 150 100

250 350 200 10 300

400 0 350 50 150

100 100 250 350 200 300 350

200 5 50 300 300

0 A1 A2 A3 A4 A5 A6 A7 A8 A9 A11 B1 B3 B5 B7 B9 B11 C1 C3 C5 C7 C9C11 D1 D3 D5 D7 D9D11 E1 E3 E5 E7 E9 E11 F1 F3 F5 F7 F9 F11 G1 G3 G5 G7 G9G11H1 H3 H5 Waste 0 0 0 50 100 150 200 250 300 ml

0 50 100 150 200 250 300 350 250 250 Elution volume (mL) Figure 6‎ .6 Fractionation of PG enzyme from A. sojae by anion-exchange chromatography on DEAE- 200 Sepharose Fast Flow resin with 5 mL cV: () exo-PG activity; () relative activity; (—) absorbance at 280 nm; (—) elution by a linear NaCl gradient (0.0 – 0.5 M in 40 cV) and washing with 1 M NaCl at the 200 end; (—) conductivity. 150 150

Approximately 87 % of exo-PG activity was detected in the enzyme activity peak 100 ranging from 144 to 184 mL elution volume presented in Figure 6.6‎ . Fractions 50 containing highest PG activity, which were eluted between 169 and 174 mL elution 100 volume composed 68 % of the loaded enzyme activity and were 6.2 times more concentrated considering the volume than taking the whole peak of exo-PG activity 0 50 into account (Table 6.6‎ ). Moreover, this peak fraction mixture had a protein content of 0.63 mg/mL. Hence, the specific activity increased from the loaded 65.9 U/mg to

410.1 U/mg by IEXC, indicating a purification of 6.2 fold with a yield of 68 %. Several 0 repetitions of this chromatographic process loading the concentrated PG extract produced similar chromatograms (data not shown), which proved the reproducibility of this purification step. The analogous fractions containing highest exo-PG activity of 4 repetitions were pooled together, dialyzed against distilled water at pH 5.0 (pH adjusted adding acetic acid) and concentrated by freeze-drying for further processing by size exclusion chromatography (section 6.3.3.3‎ ). Mill [31] fractionated an exo-PG produced by A. niger by eluting from DEAE-cellulose with increasing concentration of sodium acetate buffer (pH 4.6). This yielded in 3

144 Chapter 6 fold purification of loaded protein solution, but only 33 % of loaded enzyme activity was recovered within this chromatographic step. Enzyme purification of an exo-PMG to apparent homogeneity, showing a single band of 35.5 kDa at SDS-PAGE, was achieved by a combination of IEXC and SEC procedures [16]. Fractionation by IEXC was performed by eluting from DEAE- Sepharose with increasing NaCl concentration in sodium acetate buffer. The chromatographic technique of IEXC yielded in 1.9 fold purification of a protein solution, which was subjected to SEC prior IEXC. Moreover, 67 % of loaded enzyme activity was recovered during IEXC.

6.3.3.2 Scale up of the anion-exchange chromatographic process The aim of this study was to scale up the IEXC process so as to get more partially purified PG enzyme. Ideally, the cross section is increased without changing the column length [32]. Thus, the scale up of the anion-exchange chromatographic process was performed as described in section 6.2.5.2‎ by increasing the column volume by factor 25 with increase of the column diameter. In order to compare the process to the small scale experiment a concentrated PG extract containing 352 mg protein was loaded on 125 mL DEAE-Sepharose Fast Flow resin (GE Healthcare) pre- equilibrated with 0.02 M sodium acetate buffer (pH 5.0). Hence the applied protein to matrix volume was 3 times lower compared to the small scale experiment. The washing step using the equilibration buffer was reduced to 9 cV, since absorbance values detected at 280 nm indicated the flushing of unbound material only at the beginning of the washing step at the small scale experiment. In order to compare the elution profile, protein elution was achieved by increasing the salt concentration from 0.0 to 0.5 M NaCl in 40 cV. The elution profile obtained under this described conditions is presented in Figure 6.7‎ . Majority of unbound material was flushed out within the 9 cV of the washing step. The washing fraction contained about 6 % of loaded protein. The majority of exo-PG activity, approximately 65 % of loaded amount of enzyme activity, was eluted between 1902 to 2352 mL of elution volume. Fractions containing highest PG activity within this elution peak were eluted between 1952 and 2102 mL of elution volume and composed 56 % of the loaded enzyme activity. Considering enzyme activity concentration within these fractions, it was 2.7 times more concentrated than taking the whole peak of exo-PG activity into account. Moreover, the specific activity in the fractions, eluted in the range from 1952 to 2102 mL of elution volume, was with 255.4 U/mg 1.6 times higher than taking the whole peak of exo-PG activity. Hence, the loaded PG extract was purified by 3.4 fold with 56 % recovery.

145 Chapter 6 PG Protein

PG 0.6 24 120 22 0.5 20 100 18

0.4

16 80 14

0.3 30 25 20 15 10 5 0

12

60 PG PG

PG activity (U/mL) activity PG

30 25 20 15 10 5 0 10

- Relative activity (%) activity Relative

0 Absorbance at 280 nm PG PG

50 exo

150 Protein PG PG 100 250 350 200

0.2 300 Protein content (mg/mL) content Protein 40 8 0 350 50 150 100 250 350 200 300 6 350

20 0.1 4

2 300 300

0 0.0 0 0 1000 2000 3000 4000 5000 6000 25 20 15 10 5 0 250 Elution volume (mL) 250 Figure 6‎ .7 Separation of PG enzyme from A. sojae extract by anion-exchange chromatography on DEAE-Sepharose Fast Flow resin with 125 mL cV: () exo-PG activity; () relative activity; (|) protein4000 200 content; (—) absorbance at 280 nm; (—) elution by a linear NaCl gradient (0.0 – 0.5 M NaCl in 40 cV). 200 3500

In comparison150 to the experiment performed at smaller scale approximately 22 % less of the total recovered PG activity was obtained utilizing the 25 times bigger column 3000 volume. Regarding the peak of fractions with highest enzyme activity, the 150 100 purification fold was 1.8 times lower in the bigger column. However, the obtained chromatogram presented in Figure 6.7‎ was similar to the elution profile obtained at2500 50 the smaller scale (Figure 6.6‎ ). 100 After assuring the reproducibility of the chromatographic profile at larger scale, the 2000 0 elution of PG enzymes bound to the anion exchanger was changed to a step-wise 50 procedure (Figure 6.8‎ ). 1500 1000 0 500 0

146 Chapter 6 Protein

mAU 0.16 PG PG 25 0.14 200 Protein PG 0.12 PG 20 25

150 0.10

20 15

0.08

100 Protein PG PG

PG activity(U/mL), PG

15 - Absorbance at 280 nm Absorbance 280 at

10 (%) activityRelative 0.06

Protein PG PG exo

Protein PG PG Protein content (mg/mL) contentProtein

10 50 0.04 5

5 0.02

0

F2 A2A3A4A5A6B1B2B3B4B5C1C2C3C4C5C6C7D2F2D4D5D6D7 Waste 0 0.00

0 0 500 1000 1500 2000 2500 3000 3500 4000 25 20 15 10 5 0 25 20 15 10 5 0

0 500 1000 1500 2000 2500 3000 3500 4000 Elution volume (mL) 25 20 15 10 5 0 Fractions: w2 3 4 5 6 7 4000 Figure 6‎ .8 Fractionation of PG enzyme from A. sojae by anion-exchange chromatography on DEAE- 4000 Sepharose Fast Flow resin with 125 mL cV: () exo-PG activity; () relative activity; (|) protein 4000 content; (—) absorbance at 280 nm; (—) elution by a step-wise increase of the NaCl concentration; 3500 (—) conductivity. 3500 3500

Therefore, about 75 mg protein with a specific enzyme activity of 96.6 U/mg was 3000 3000 loaded onto the DEAE-Sepharose Fast Flow resin (GE Healthcare). Majority of PG activity was eluted between 0.11 to 0.13 M NaCl concentration by applying the linear 2500 2500

gradient for the elution (Figure 6.7‎ ). Hence, during the step-wise procedure the 3000 washing step started with a concentration of 0.05 M NaCl within the buffer system 2000 2000 for the removal of more impurities within the washing fraction (Figure 6.8‎ ).

Moreover, the washing step was extended to 20 cV, since flushing of more proteins 2500 1500 1500 within this fraction was expected. The washing fraction was collected in two parts, one part containing the third 2000 1000 protein peak detected at 280 nm absorbance eluted between 1060 and 1110 mL of 1000 elution volume and the remaining washing volume (131 to 1060 mL and 1110 to 2630 500 mL elution volume) was collected in another fraction. As can be seen from Figure 500 1500 6.8‎ about 21 % of loaded PG activity were eluted during the washing with 0.05 M 0 0 NaCl, whereof the majority of 20 % was recovered in the larger washing fraction. From the absorbance profile detected at 280 nm it can be assumed that most of the 1000 activity was eluted at the beginning of the washing step. Due to the large volume of the bigger washing fraction accounted the exo-PG activity only for 0.6 U/mL. 500

147 0 Chapter 6

Nevertheless, the eluted specific activity of this fraction amounted to approximately 108 U/mg. The second washing fraction containing the third protein peak eluted during this step accounted only for 1 % of loaded enzyme activity. Also, the lower specific activity of about 11 U/mg indicated the elution of more proteins within this peak, which are not involved in the degradation of pectic acid. The step-wise elution of proteins was achieved with increasing NaCl concentration to 0.15 M and 0.3 M NaCl and holding for 4 and 3 cV, respectively. The PG-activity peak obtained at the elution with 0.15 M NaCl concentration contained about 58 % of the loaded PG activity. Moreover, the specific activity of this peak amounted to approximately 191.4 U/mg. Hence, indicating a 2 fold purification of PG enzyme. These fractions were pooled together and concentrated by freeze-drying for application studies (chapter 7). Further increasing the salt concentration to 0.3 M NaCl eluted a protein peak containing about 11 % of loaded PG activity with a specific activity of 36.7 U/mg. This peak accounted for 29 % of the loaded protein content. At the end a flushing with 1.0 M NaCl for 5 cV was performed for the removal of remaining material. The last peak eluted at 1.0 M NaCl concentration contained only about 1 % of loaded PG activity and 5 % of the loaded protein content. Fractions taken at various steps during the chromatographic process were desalted utilizing pre-packed PD-10 desalting columns (Amersham Biosciences AB), containing Sephadex G-25 medium, according to the instructions given by the provider [33] and concentrated to 1 mg/mL protein concentration by freeze-drying. Proteins in concentrated fractions were further separated by electrophoresis on a 12.5 % SDS-PAGE (Figure 6.9‎ ). As can be seen from Figure 6.9‎ , each fraction contained a mixture of proteins of various molecular weights. The two washing fractions mainly differ in proteins of lower molecular weight. Fraction 4 composing highest exo-PG activity and a high specific activity of 309.2 U/mg contained three major bands. Comparing these bands with proteins present in fraction 3 (which was also eluted at 0.15 M NaCl, but composed lower exo-PG activity) the band with a molecular weight of about 40 kDa and the one with a molecular weight above 80 kDa were dominant in fraction 4. PGs isolated from different microbial sources differ markedly in their biochemical and physicochemical properties, and molecular weights of reported PGs vary between 35 and 79 kDa [2]. The molecular weight of a homogenous PG derived from solid-state cultivation of A. niger was determined with 36 kDa by Dinu et al. [13]. Kester & Visser [34] isolated two most abundant PGs with molecular masses of 38 and 55 kDa from a commercial enzyme preparation (Pectinase K2B 078, Rapidase) derived from A. niger. Another PG from a commercial A. niger pectinase preparation (Shaxian Enzyme Factory, China) was purified by Guo et al. [15] with a

148 Chapter 6 molecular weight of 40.4 kDa. Furthermore, two exo-PGs with molecular masses of 56 and 82 kDa were isolated from another commercial A. niger enzyme preparation (Pectinex AR, Novozymes, France). Semenova et al. [14] isolated two PGs with molecular masses of 38 and 65 kDa from the culture liquid of A. japonicus. Pedrolli & Carmona [35] purified an exo-PG with a molecular weight of 69.7 kDa from submerged culture of A. giganteus. Dogan & Tari [36] obtained two bands with the molecular weights of 36 and 53 kDa after three-phase partitioning of A. sojae extract produced in SmF. With regard to the reported values the dominant band in fraction 4 with a molecular mass of approximately 40 kDa might be the potential PG. Nevertheless, this is just a speculation as long as the desired protein is not completely purified. Hence, further purification steps were needed (section 6.3.3.3).‎

Load Marker IEXC w1 w2 3 4 5 6 7 Mw (kDa) 175

80

58

46

30

23

Figure 6‎ .9 Analysis of IEXC fractions by 12.5 % SDS-PAGE showing from left to right: ColorPlus prestained protein marker (New England BioLabs); PG extract loaded onto IEXC column; w1 = larger fraction of the washing step; w2 = smaller washing fraction including third protein peak detected at 280 nm; 3 & 4 = fractions eluted at 0.15 M NaCl; 5 & 6 = fractions eluted at 0.3 M NaCl; 7 = fraction eluted at 1.0 M NaCl.

Besides the elution procedure described above, also the effect of a step gradient including more steps was explored for fractionation of PG extract. Similar to the previously presented step gradient started the IEXC procedure with the washing of unbound material at 0.05 M NaCl concentration for 20 cV. Stepwise elution of proteins was performed with increasing NaCl concentration to 0.1 M NaCl for 4 cV, 0.15 M NaCl for 4 cV, 0.2 M NaCl for 3 cV, 0.3 M NaCl for 3 cV and finally 1.0 M NaCl for 5 cV. About 111 mg protein with a specific enzyme activity of 104.8 U/mg was

149 Chapter 6 loaded onto the DEAE-Sepharose Fast Flow resin (GE Healthcare) and the elution profile is presented in Figure 6.10‎ .

0.16

30 0.14

0.12 25

0.10 20

0.08

15 Protein PG PG

PG activity(U/mL), PG -

Absorbance at 280 nm 280 at Absorbance 0.06 Relative activity (%) activityRelative Protein PG PG

Protein PG PG exo

Protein content (mg/mL) contentProtein 10 0.04

5

0.02

0.00 0 0 1000 2000 3000 4000 5000 25 20 15 10 5 0 Elution volume (mL) 25 20 15 10 5 0 Fractions: 1 2 3 4 5 6 7 8 9 25 20 15 10 5 0

Figure 6‎ .10 Purification of PG enzyme from A. sojae by anion-exchange chromatography on DEAE- 4000 4000 Sepharose Fast Flow resin with 125 mL cV: () exo-PG activity; () relative activity; (|) protein content; (—) absorbance at 280 nm; (—) elution by a step-wise rising of the NaCl concentration; (—) 4000 3500 conductivity. 3500

The washing fraction was divided into several parts. The major part consisted of the 3500 3000 3000 elution volume ranging from 126 to 900 mL and from 1050 to 2500 mL. The other part, containing the third protein peak detected at 280 nm absorbance, which was 2500 2500 collected between 900 and 1050 mL of elution volume, was divided into three 3000 fractions each of 50 mL volume. The major part of the washing fraction contained 2000 30 % of loaded PG activity and had a specific activity of 269.5 U/mg. From the2000 2500 absorbance profile detected at 280 nm can be assumed that most of the activity 1500 was eluted at the beginning of the washing step. Due to the large volume of this1500 washing fraction accounted the exo-PG activity only for 1.5 U/mL. Fractions of the 2000 1000 washing peak eluted between 900 to 1050 mL of elution volume contained about1000 15 % of the loaded protein concentration, but in total only 9 % of the loaded PG 500 500 activity. 1500 The step-wise elution of proteins started at 0.1 M NaCl concentration, which yielded 0 0 in the recovery of 24 % of the loaded PG activity with a specific activity of

354.5 U/mg. Hence, PG enzyme was purified 3.4 fold within this elution peak. 1000 Fractions eluted at 0.1 M NaCl concentration between 2733 and 3133 mL of the

500 150 0 Chapter 6 elution volume, were pooled together and concentrated by freeze-drying for application studies (chapter 7). Further elution of proteins was achieved by rising NaCl concentration to 0.15 M. This protein peak contained 12 % PG activity and the specific activity amounted to 88.4 U/mg. The next group of proteins was eluted with 0.2 M NaCl concentration, which resulted in the recovery of 5 % exo-PG activity with a specific activity of 48.6 U/mg. Further rising of the salt concentration to 0.3 M NaCl eluted another protein peak containing 4 % of exo-PG activity and a specific activity of 46.4 U/mg. Within the last peak, which was eluted at 1.0 M NaCl concentration, were only 1 % PG activity recovered and it contained 5 % of the loaded protein content. Protein patterns were studied of various fractions collected during the chromatographic process. These fractions were treated like the fractions of the aforementioned process and concentrated to 1 mg/mL protein concentration before analyzing by SDS-PAGE (Figure 6‎ .11).

Load Load Marker IEXC 1 2 3 4 6 8 9 Marker IEXC 2 3 4 5 6 7 Mw Mw (kDa) (kDa) 175 175

80 80

58 58

30 30

23 23

Figure 6‎ .11 Protein profile of IEXC fractions on 12.5 % SDS-PAGE showing: Marker = ColorPlus prestained protein marker (New England BioLabs); Load IEXC = PG extract loaded onto IEXC column; 1 & 2 = fractions eluted at 0.10 M NaCl; 3 – 5 = fractions eluted at 0.15 M NaCl; 6 & 7 = fractions eluted at 0.2 M NaCl; 8 = fraction eluted at 0.3 M NaCl; 9 = fraction eluted at 1.0 M NaCl.

In Figure 6‎ .10 are the eluted fractions marked which were analyzed by electrophoresis and presented in Figure 6‎ .11. All fractions contained a mixture of protein bands. Comparing the fractions containing highest PG activity collected during both step-wise elution procedures, the proteins of these fractions showed a distinct pattern. While fraction 4 of the aforementioned elution procedure contained three dominant bands (Figure 6‎ .9), only two dominant bands are present in fraction 2 of the currently discussed elution procedure. Furthermore, these two bands in fraction 2 presented molecular weights of approximately 50 and 70 kDa. Hence, they differ from the previously narrowed proteins down. However, exo-PG

151 Chapter 6 activity in fraction 4 was about 1.5 times higher than in fraction 2, but fraction 2 contained a slightly higher specific activity due to the elution with a lower salt concentration. In order to identify the molecular weight of PG enzyme further purification steps were needed.

6.3.3.3 Size exclusion chromatography Several other names for this method have been put forward, including gel filtration, gel exclusion and molecular sieving. Columns are packed with a porous gel matrix in bead form, consisting of an open, cross-linked, three-dimensional molecular network. The pores within the beads are of such sizes that some are not accessible by larger molecules, but smaller molecules can penetrate all pores. Thus, the separation of proteins is based on their size and shape. Proteins are usually eluted from SEC column in the order of decreasing molecular weight [19, 32]. Gel filtration studies were performed for further isolation of PG enzyme from partially purified IEXC fractions. Therefore, analogous fractions obtained by repeated IEXC processes containing highest PG activity were pooled together (section 6.3.3.1)‎ and concentrated by freeze-drying. Sample was rehydrated in 0.05 M sodium acetate buffer (pH 5.0) composing 0.2 M L-arginine hydrochloride up to a concentration factor of about 40. Concentrated sample of 2.39 mg protein was loaded onto Superdex 200 10/300 GL (GE Healthcare) column as described in section 6.2.5.3.‎ The SEC elution profile is shown in Figure 6.12‎ . The chromatogram of gel filtration contained several peaks detected at 280 nm absorbance. However, exo-PG activity was only detected in three fractions, 5 – 7, of which fraction 6 composed the majority of 92 % activity. This fraction was eluted after 16 mL elution volume and the specific activity of this fraction amounted to 2384.0 U/mg, which represented 36.2 fold purification of the original PG extract obtained by membrane filtration processing. Regarding the loaded partially purified IEXC fraction a purification of 4.5 fold was achieved by SEC. Jacob et al. [11] purified an exo-PG with a molecular weight of 43 kDa produced by Streptomyces lydicus by ultrafiltration and a combination of IEXC and SEC with a recovery yield of 57.1 % and 54.9 fold purification of PG enzyme. Considering only the gel filtration step, they obtained 1.7 fold purification of loaded partially purified extract utilizing a Sephadex G-100 column.

152 Chapter 6

2011 11 24 crude UF SEC 2001:1_UV 2011 11 24 crude UF SEC 2001:1_Cond 2011 11 24 crude UF SEC 2001:1_Fractions 2011 11 24 crude UF SEC 2001:1_Inject 2011 11 24 crude UF SEC 2001:1_Logbook

mAU 100 1200 90 700 1100

1000 80 600 900 70

500

800 );

60 g/mL)

700 μ 400 50

600 30 25 20 15 10 5 0

PG PG

300 500 40

PG activity (U/mL activity PG

Absorbance at 280 nmAbsorbance 280 at -

30 25 20 15 10 5 0 Relative activity activity (%) Relative 0

50 rotein content ( content rotein PG PG 150 100

250 exo 350 200 300 Protein PG PG

400 p 30 200 0 350 50 150 100 250 350 200 300 300

20 350

100 200 10 100 300 0 300

D8 D9 D10 D11 D12 E1 E2 E3 E4 E5 E6 E7 E8 E9 E10 E11 E12 F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11 F12 G1 G2 G3 G4 G5 G6 G7 G8 G9 G10 G11 G12 H1 H2 0 0 0.0 5.0 10.0 15.0 20.0 25.0 30.0 35.0 40.0 ml 0 10 20 30 40

Elution volume (mL) 25 20 15 10 5 0 Fraction: 1 2 3 4 5 6 7 250

250 Figure 6‎ .12 Gel filtration of concentrated IEXC fractions; () exo-PG activity; () relative activity; (|) protein content; (—) absorbance at 280 nm; (—) conductivity. 4000 200

Repetition of the SEC process produced almost identical chromatograms (data not 200 3500 shown). In order to analyze the proteins present in the collected SEC fractions, 150 which were marked in Figure 6.13‎ , analogous fractions 1 – 7 of the repetitions were 3000 150 pooled together and concentrated to 1 mg/mL protein content by evaporation of 100 liquid samples in the concentrator 5301 (Eppendorf, Germany) at 30 °C. 2500

Concentrated samples were loaded on 12.5 % SDS-PAGE (Figure 6.13‎ ). 50 100 As expected, proteins were eluted by gel filtration according to their molecular 2000

weight starting with proteins of higher molecular weight. Moreover, the most 0

active fraction number 6 obtained by SEC still contained four dominant bands, 50 1500 indicating that PG enzyme was only partially purified from the crude extract by combination of IEXC and SEC. However, in comparison to previous active fractions 1000 obtained by IEXC and analyzed by SDS-PAGE composed the fraction obtained by SEC 0 only protein bands with a molecular weight below 50 kDa. With regard to the fact 500 that only fraction number 5 to 7 contained exo-PG activity and highest activity was present in fraction number 6, while fraction number 7 contained only very low 0 amount of enzyme activity, matching proteins of fraction number 5 and 6 had molecular weights of about 36 and 48 kDa. Since the specific activity of fraction 6 was 5.8 times higher than in fraction number 5, one could speculate that the predominant band of fraction number 6 at approximately 36 kDa might be the desired PG enzyme.

153 Chapter 6

Load 1 2 3 4 5 6 7 SEC Marker Mw (kDa) 175

80

58

46

30

23

Figure 6‎ .13 Protein profile of SEC fractions on 12.5 % SDS-PAGE showing from left to right SEC fraction 1 – 7, Load SEC = partially purified PG loaded onto Superdex 200 10/300 GL column, and Marker = ColorPlus prestained protein marker (New England BioLabs).

6.3.3.4 Hydrophobic interaction chromatography HIC separates proteins by exploiting differing degrees of the proteins surface hydrophobicity, which depends on the occurrence of hydrophobic interactions between the hydrophobic patches on the protein surface and hydrophobic groups covalently attached to a suitable matrix [19]. Preliminary tests were investigated on the screening of binding and elution conditions for protein purification by HIC, using HiTrap (GE Healthcare) columns pre-packed with sepharose. During this screening two ligands, phenyl (HiTrap Phenyl FF High Sub, GE Healthcare) and butyl (HiTrap Butyl FF, GE Healthcare), were explored for further fractionation of concentrated IEXC samples. These experiments resulted in deficient PG purification from concentrated IEXC samples (data not shown). Nevertheless, since combination of IEXC and SEC yielded only in partial purification of PG enzyme from SSF extract, the partially purified sample was further subjected to HIC using phenyl sepharose FF resin for separation of PG enzyme from other proteins (Figure 6.14‎ ). Keller et al. [37] developed a purification scheme for separation of PG from commercial pectinase source. Incorporating HIC (using Phenyl-Sepharose resin) as a final step after gel filtration and IEXC resulted in highest purification.

154 Chapter 6

HIC 2011 01 25 SEC B3 Sample001:1_UV HIC 2011 01 25 SEC B3 Sample001:1_Cond HIC 2011 01 25 SEC B3 Sample001:1_Cond% HIC 2011 01 25 SEC B3 Sample001:1_Conc HIC 2011 01 25 SEC B3 Sample001:1_Fractions HIC 2011 01 25 SEC B3 Sample001:1_Inject HIC 2011 01 25 SEC B3 Sample001:1_Logbook mAU 80 50

50.0 70

40 60

40.0 )

50 g/mL) μ 30 30.0

40

20.0 content ( content 20

30

PG activity (U/mL activity PG -

30 25 20 15 10 5 0 Absorbancenm 280 at

10.0 rotein Protein PG PG

PG PG exo

20 P

10 0 50 150 100 250 350 200 300 0.0 10 350

A1 A2 A3 A4 A5 A6 A7 A8 A9 A10 A11 A12 B1 B2 B3 B4 B5 B6 B7 B8 B9 B10 B12 C1 C2 C3 C4 C5 C6 C7 C8 C9 C10 C12 D1 D2 D3 D4 D5 D6 D7 D8 D9 D10 D12 E1 E2 E3 E4 E5 E6 E7 E8 Waste 0 0 0 20 40 60 80 100 120 ml 0 20 40 60 80 100 120

Elution volume (mL) 300 Figure 6‎ .14 Elution profile obtained by HIC; () exo-PG activity; () relative activity; (|) protein

content; (—) absorbance at 280 nm; (—) elution by a linear decrease of ammonium 25 sulfate20 15 10 5 0 concentration; (—) & (—) conductivity. 250 4000 The HIC elution profile presented in Figure 6‎ .14 showed the separation of two 200 protein peaks, which was detected at 280 nm absorbance during the 3500 chromatographic run. Analysis of HIC fractions determined that exo-PG activity was 150 only present in the first elution peak, while higher protein content was eluted within 3000 the second elution peak. The peak eluted between 53 and 73 mL of elution volume 100 contained about 65 % of the loaded PG activity. In order to analyze the fractions for 2500 purity, the active fractions composing 65 % of the PG activity were pooled together 50 and dialyzed against distilled water for the removal of salts. Dialyzed fraction 2000 mixture was concentrated by freeze-drying and loaded on a 12.5 % SDS-PAGE to 0 check the homogeneity of the enzyme and to determine the molecular weight 1500 (Figure 6‎ .15). Purification of PG from A. sojae utilizing a combination of IEXC, SEC and HIC 1000 provided apparently homogenous preparation of the enzyme, which showed a single protein band on SDS-PAGE with a molecular weight of about 40 kDa. 500 Determination of the molecular weight by analysis of the SDS-PAGE just gave an estimated value. Hence, comparing the band in Figure 6.15‎ with the protein pattern 0 presented in Figure 6.13‎ , one could speculate that the single band is equivalent to the band in SEC fraction 6, which was located between the marker with a molecular weight of 30 and 46 kDa.

155 Chapter 6

HIC Marker B9 –C6 Mw (kDa) 175

80 58

46

30

23

Figure 6‎ .15 SDS-polyacrylamide gel electrophoresis on a 12.5 % gel of the purified PG enzyme showing from left to right: Marker = ColorPlus prestained protein marker (New England BioLabs); HIC B9 – C6 = pooled active fractions obtained by HIC.

As previously mentioned PGs derived from various microbial sources with diverse molecular weights have been purified (section 6.3.3.2).‎ A summary of reported microbial polygalacturonases with similar molecular weights is presented in Table 6.5‎ . For instance, Zhang et al. [38] isolated PG with the molecular weight of 41 kDa from Penicillium oxalicum cultured on sugar beet pulp. Dogan & Tari [36] reported about a partially purified PG extract produced by A. sojae ATCC 20235 in liquid culture, which contained two bands with the molecular weights of 36 and 53 kDa. This result is close to the reported value of the present study of purified PG enzyme utilizing the same microorganism at solid-state culture. In order to confirm the presence of PG enzyme in A. sojae extract the identification of the proteins had to be performed (chapter 7). Furthermore, Table 7.1 in chapter 7 contains a summary of the biochemical characterization of crude and three-phase partitioned exo-PG produced by A. sojae ATCC 20235 in SmF, which was published by Tari et al. [39] and Dogan & Tari [36].

156 Chapter 6

Table 6‎ .5 Summary of biochemical properties of certain microbial polygalacturonases.

Cultivation Molecular Refer- Organism method / Nature weight Other characteristics ence source (kDa) number pH optimum 4.0; A. carbonarius SSF Exo-PG 61 (PG I) [40] optimum temperature 55 °C pH optimum 4.1; Exo-PG 42 (PG II) optimum temperature 50 °C pH optimum 4.3; Exo-PG 47 (PG III) optimum temperature 55 °C

A. carbonarius SSF PG* 42 - [41]

pI 5.13; pH optimum 4.6; A. niger SSF Exo-PG 36.3 [13] optimum temperature 40 °C pI 5.6; maximal activity [14] A. japonicus SmF Endo-PG 38 (PG I) at pH 4.0–5.5 pI 3.3; maximal activity Endo-PG 65 (PG II) at pH 4.0–5.5 A. niger Pectinase pI in the range 4.6 – 5.9; [34] Endo-PG 55 (PG I) K2B 078 pH optimum 4.8 (Rapidase) pI in the range 3.2 – 3.5; Endo-PG 38 (PG II) pH optimum 4.9 Pectinase (Shaxian pH optimum 5.0; A. niger PG* 40.4 [15] Enzyme optimum temperature 36 °C Factory) Penicillium maximal activity SSF PG* 41 [38] oxalicum at pH 5.0 – 5.5 Streptomyces pH optimum 6.0; SSF Exo-PG 43 [11] lydicus optimum temperature 50 °C Trichoderma SmF Exo-PG 31 pI 4.5; pH optimum 5.0 [42] harzianum Sclerotium SmF PG* 39.5 (PG I) pI 6.5 [43] rolfsii

PG* 38 (PG II) pI 5.4

optimum temperature 45 °C; Mucor flavus SmF PG* 40 [44] pH optimum 3.5 – 5.5 * Nature of the enzyme was not specified or reported as endo-PG, but the determination of enzyme activity based on the measurement of reducing groups released from polygalacturonic acid.

6.3.4 Summary of chromatographic performance Purification of PG enzyme secreted by A. sojae is summarized in Table 6‎ .6 showing the performance of each purification procedure.

157 Chapter 6

Table 6‎ .6 Summary of the purification of PG enzyme from A. sojae.

Total Specific Purification PG activity Purification Recovery Sample protein activity step (U) fold* (%)* (mg) (U/mg)

PG extract 2835.6 43.0 65.9 IEXC 6.2 68 IEXC fraction 1941.3 4.7 410.1

Load SEC 1263.5 2.4 528.7 SEC 4.5 92 SEC fraction 1164.2 0.5 2384.0

Load HIC 709.2 HIC 65 HIC fraction 460.9

PG extract 26142.6 352.1 74.2 IEXC scale up 3.4 56 (linear) IEXC fraction 14689.1 57.5 255.4

IEXC scale up PG extract 7218.9 74.7 96.6 (step 0.15 M 2.0 58 IEXC fraction 4222.3 22.1 191.1 NaCl)

IEXC scale up PG extract 11596.8 110.7 104.8 (step 0.1 M 3.4 24 IEXC fraction 2767.0 7.8 354.7 NaCl)

* Presented values of purification fold and recovery yield were calculated for each individual step.

From Table 6.6‎ can be seen that IEXC yielded in 68 % recovery of PG enzyme with 6.2 fold purification. The next step of SEC yielded in 92 % recovery of PG with 4.5 fold purification from the loaded IEXC fraction. With respect to the PG extract, these two purification steps led to 36.2 fold purification of PG enzyme with a recovery of 41 %. Regarding the purification by HIC, values of protein concentration were not obtained due to low sample volume and interference of ammonium sulfate with the reagent for protein determination. However, it can be seen that this step yielded in 65 % enzyme recovery. Separation of PG enzyme by IEXC in the scaled up process applying the linear gradient for enzyme elution reduced the purification fold by factor 1.8 in comparison to the process at the smaller column volume. Nevertheless, 56 % of PG enzyme was recovered within this step. Changing the elution procedure to a step gradient starting the enzyme elution with 0.15 M NaCl, slightly decreased the purification fold to 2.0 with 58 % of PG recovery. Whereas, applying a step-wise

158 Chapter 6 elution procedure starting with 0.1 M NaCl decreased the recovery of PG to 24 % within this elution step. This can be explained by further elution of PG enzyme at higher NaCl concentration as presented in Figure 6.10‎ . The elution procedure starting with 0.1 M NaCl was conducted for the removal of other proteins at this concentration in order to achieve a higher purity of PG enzyme within the next elution step, but most of the activity eluted already at 0.1 M NaCl. Comparing the enzyme recovery achieved by IEXC processes applying the stepwise elution procedures, the process starting with 0.15 M NaCl concentration is more efficient for recovery of PG. Further comparison of the IEXC fractions, which were obtained by stepwise elution procedure, in terms of application testing is given in chapter 7.

6.4 Conclusions PG enzyme produced by A. sojae in SSF process was successfully separated from the biomass and isolated to apparently homogeneity by combination of IEXC, SEC and HIC. The enzyme leaching results showed that investigating and optimizing the leaching conditions is a simple way of concentration of the desired enzyme in the crude extract. Moreover, the selection of an appropriate solvent and optimization of the extraction conditions could be a useful tool for a more selective extraction of the desired bioproduct from the fermented solids. The PG purification by combination of several chromatographic methods yielded in sufficient enzyme recovery and isolation of a single protein band in the range of about 40 kDa. Further studies on A. sojae extract characterization and application investigations are presented in chapter 7.

References [1] Kashyap, DR, Vohra, PK, Chopra, S, Tewari, R (2001). Applications of pectinases in the commercial sector: a review. Bioresource Technology 77, 215-227. [2] Jayani, RS, Saxena, S, Gupta, R (2005). Microbial pectinolytic enzymes: A review. Process Biochemistry 40, 2931-2944. [3] Lang, C, Dörnenburg, H (2000). Perspectivies in the biological function and the technological application of polygalacturonases. Applied Micriobiology and Biotechnology 53, 366-375. [4] Singh, SA, Ramakrishna, M, Appu Rao, AG (1999). Optimisation of downstream processing parameters for the recovery of pectinase from the fermented bran of Aspergillus carbonarius. Process Biochemistry 35, 411-417.

159 Chapter 6

[5] Patil, SR, Dayanand, A (2006). Optimization of process for the production of fungal pectinases from deseeded sunflower hesd in submerged and solid- state conditions. Bioresource Technology 97, 2340-2344. [6] Díaz-Godínez, G, Soriano-Santos, J, Augur, C, Viniegra-González, G (2001). Exopectinases produced by Aspergillus niger in solid-state and submerged fermentation: a comparative study. J. Ind. Microbiol. Biot. 26, 271-275. [7] Solís-Pereira, S, Favela-Torres, E, Viniegra-González, G, Gutiérrez-Rojas, M (1993). Effects of different carbon sources on the synthesis of pectinase by Aspergillus niger in submerged and solid state fermentations. Applied Micriobiology and Biotechnology 39, 36-41. [8] Pandey, A, Soccol, CR, Mitchell, D (2000). New developments in solid state fermentation: I-bioprocesses and products. Process Biochemistry 35, (10) 1153-1169. [9] Lonsane, BK, Kriahnaiah, MM (1992). Product leaching and downstream processing. In: H.W. Doelle; D.A. Mitchell; C.E. Rolz, eds. Solid substrate cultivation. Elsevier Applied Science: England. [10] Lonsane, BK, Saucedo-Castaneda, G, Raimbault, M, Roussos, S, Viniegra- González, G, Ghildyal, NP, Ramakrishna, M, Krishnaiah, MM (1992). Scale-up strategies for solid state fermentation systems. Process Biochemistry 27, 259-273. [11] Jacob, N, Poorna, CA, Prema, P (2008). Purification and partial characerization of polygalacturonase from Streptomyces lydicus. Bioresource Technology 99, 6697-6701. [12] Gummadi, SN, Panda, T (2003). Purification and biochemical properties of microbial pecinases - a review. Process Biochemistry 38, 987-996. [13] Dinu, D, Nechifor, MT, Stoian, G, Costache, M, Dinischiotu, A (2007). Enzymes with new biochemical properties in the pectinolytic complex produced by Aspergillus niger MIUG 16. Journal of Biotechnology 131, (2) 128- 137. [14] Semenova, MV, Grishutin, SG, Gusakov, AV, Okunev, ON, Sinitsyn, AP (2003). Isolation and properties of pectinases from the fungus Aspergillus japonicus. Biochemistry (Moscow) 68, (5) 559-569. [15] Guo, C-T, Xue, W-M, Chen, T-B, Deng, W-H, Rao, P-F (2002). Purification and partial characterization of an endo-polygalacturonase from Aspergillus niger. Journal of Food Biochemistry 26, 253-265. [16] Celestino, SM, de Freitas, SM, Medrano, FJ, de Sousa, MV, Filho, EXF (2006). Purification and characterization of a novel pectinase from Acrophialophora nainiana with emphasis on its physicochemical properties. Journal of Biotechnology 123, 33-42. [17] Sharma, A, Gupta, MN (2001). Purification of pectinase by three-phase partitioning. Biotechnology Letters 23, 1625-1627.

160 Chapter 6

[18] Camperi, SA, Auday, RM, Del Cañizo, AN, Cascone, O (1996). Study of variables involved in fungal pectic enzyme fractionation by immobilized metal ion affinity chromatography. Process Biochemistry 31, (1) 81-87. [19] Walsh, G, Headon, DR (1994). Protein Biotechnology. John Wiley & Sons Ltd: Chichester, England. [20] GE Healthcare, BA (2010). Gel filtration Principles and Methods. Uppsala, Sweden. [21] Panda, T, Naidu, GSN, Sinha, J (1999). Multipleresponse analysis of pectinolytic enzymes by Aspergillus niger: a statistical view. Process Biochemistry 35, 187-195. [22] Bradford, MM (1976). A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Analytical Biochemistry 72, 248-254. [23] Laemmli, UK (1970). Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680-685. [24] AmershamBiosciencesAB (1999). Protein electrophoresis. Technical manual. Uppsala, Sweden. [25] Neuhoff, V, Arnold, N, Taube, D, Ehrhardt, W (1988). Improved staining of proteins in polyacrylamide gels including isoelectric focusing gels with clear background at nanogram sensitivity using Coomassie Brillant Blue G-250 and R-250. Electrophoresis 9, 255-262. [26] Scheper, A (2000). New products and new areas of bioprocess engineering. Springer-Verlag: Heidelberg. [27] Doelle, HW, Mitchell, DA, Rolz, CE (1992). Solid substrate cultivation. Elsevier Applied Science: England. [28] Castilho, LR, Alves, TLM, Medronho, RA (1999). Recovery of pectolytic enzymes produced by solid state culture of Aspergillus niger. Process Biochemistry 34, 181-186. [29] Ghildyal, NP, Ramakrishna, M, Lonsane, BK, Karanth, NG (1991). Efficient and simple extraction of mouldy bran in a pulsed column extractor for recovery of amyloglucosidase in concentrated form. Process Biochemistry 26, 235-141. [30] Castilho, LR, Medronho, RA, Alves, TLM (2000). Production and extraction of pectinase obtained by solid state fermentation of agroindustrial residues with Aspergillus niger. Bioresource Technology 71, 45-50. [31] Mill, PJ (1966). The pecic enzymes of Aspergillus niger. A mercury-activated exopolygalacturonase. Biochemical Journal 99, (3) 557-561. [32] Scopes, RK (1993). Protein purification: principles and practise. Springer- Verlag New York, Inc.: New York. [33] AmershamBiosciencesAB (2003). Instructions: PD-10 desalting column. Sweden.

161 Chapter 6

[34] Kester, HC, Visser, J (1990). Purification and characterization of polygalacturonases produced by the hyphal fungus Aspergillus niger. Biotechnology and Applied Biochemistry 12, (2) 150-160. [35] Pedrolli, DB, Carmona, EC (2010). Purification and characterization of the exopolygalacturonase produced by Aspergillus giganteus in submerged cultures. Journal of Industrial Microbiology & Biotechnology 37, 567-573. [36] Dogan, N, Tari, C (2008). Characterization of three-phase partitioned exo- polygalacturonase from Aspergillus sojae with unique properties. Biochemical Engineering Journal 39, 43-50. [37] Keller, SE, Jien, JJ, Brunner, JR (1982). Purification of commercial pectinase by hydrophobic chromatography. Journal of Food Science 47, 2076-2077. [38] Zhang, C-h, Li, Z-m, Peng, X-w, Jia, Y, Zahng, H-x, Bai, Z-h (2009). Separation, purification and chatracterization of three endo- polygalacturonases from newly isolated Penicillium oxalicum. The Chinese Journal of Process Engineering 9, (2) 242-249. [39] Tari, C, Dogan, N, Gogus, N (2008). Biochemical and thermal characterization of crude exo-polygalacturonase produced by Aspergillus sojae. Food Chemistry 111, 824-829. [40] Devi, NA, Rao, AG (1996). Fractionation, purification, and preliminary characterization of polygalacturonases produced by Aspergillus carbonarius. Enzyme and Microbial Technology 18, 59-65. [41] Nakkeeran, E, Umesh-Kumar, S, Subramanian, R (2011). Aspergillus carbonarius polygalacturonases purified by integrated membrane process and affinity precipitation for apple juice production. Bioresource Technology 102, 3293-3297. [42] Mohamed, SA, Farid, NM, Hossiny, EN, Bassuiny, RI (2006). Biochemical characterization of an extracellular polygalacuronase from Trichoderma harzianum. Journal of Biotechnology 127, 54-64. [43] Schnitzenhofer, W, Weber, H-J, Vrsanská, M, Biely, P, Cavaco-Paulo, A, Guebitz, GM (2007). Purification and mechanistic characterisation of two polygalacturonases from Sclerotium rolfsii. Enzyme and Microbial Technology 40, 1739-1747. [44] Gadre, RV, van Driessche, G, van Beeumen, J, Bhat, MK (2003). Purification, characterization and mode of action of an endo-polygalacturonase from the psychrophilic fungus Mucor flavus. Enzyme and Microbial Technology 32, (2) 321-330.

162 Chapter 7

Chapter 7.

Characterization of Aspergillus sojae enzyme extracts and application studies

Abstract

Pectinolytic enzyme preparations used in industrial applications related to the food industry, e.g. in fruit juice production and winemaking, are usually mixtures of pectinolytic enzymes associated with cellulytic, proteolytic and other species of enzymes apart from the main pectinases. The mass spectrometric characterization of proteins derived from solid-state culture of Aspergillus sojae ATCC 20235 identified a broad spectrum of carbohydrate-active enzymes. The presence of a wide range of different types of enzymes is essential for the degradation of complex plant polysaccharide molecules. A. sojae produced enzymes involved in the degradation of various plant polysaccharides including cellulose, pectin, xylan and starch. For instance, A. sojae produced glucoamylase and α-amylase which are involved in the degradation of starch. Rhamnogalacturonan lyase was identified, which cleaves within the hairy regions of pectin. Additionally, xylanase was determined which acts on xylan. A comparative study on pectinolytic enzymes produced by A. sojae was performed in order to characterize the pectinases on the basis of their substrate preference and degrading mode, which revealed the production of various pectinolytic enzymes by the filamentous fungus under solid- state conditions. These results were compared to commercial pectinase preparations for the identification of specific pectinolytic enzyme profiles with regard to their applications. The mass spectrometric characterization and the pectinolytic enzyme profiling revealed that A. sojae co-produced other enzymes along with polygalacturonase in solid-state fermentation, which might have beneficial effects in terms of certain applications. Moreover the conducted application tests using A. sojae extracts in processes related to the food industry demonstrated the improvement of the process performances by the enzyme treatment in comparison to the blank experiments.

163 Chapter 7

7.1 Introduction The degradation of plant polysaccharides by utilization of enzymes is needed in various industrial sectors, such as within the paper, food or feed industry. Plant cell wall polysaccharides are mainly composed of cellulose, hemicelluloses and pectin. In nature, filamentous fungi secrete a broad spectrum of carbohydrate-active enzymes, which play a central role in the degradation of plant biomass. Carbohydrate-active enzymes can be organized in different families based on amino acid sequence of the structurally related catalytic modules. Majority of these carbohydrate-active enzymes are hydrolytic enzymes and play important roles in nutrition and releasing of carbon and nitrogen locked in insoluble macromolecules. The carbohydrate-active enzymes are assigned to at least 35 glycosidic hydrolase families. Moreover, fungal enzymes involved in the degradation of plant polysaccharides are also assigned to three carbohydrate esterase families and six polysaccharide lyase families. [1, 2] The variety of these enzyme sets differs between fungal species and often corresponds to the requirements of its habitat. For instance, Aspergillus species produce a large number of enzymes particularly involved in the degradation of pectic substances [2]. The fungal biodiversity with respect to plant cell wall degradation has industrial importance for utilization of desired enzyme sets as needed for certain applications. Pectinolytic enzymes are applied in various industrial processes, such as scouring of cotton, degumming of plant fibers, waste water treatment, vegetable oil extraction, tea and coffee fermentations, and in the alcoholic beverages and food industries [3]. Fungal acidic pectic enzymes are extensively used in food industry, e.g. for the extraction and clarification of fruit juices or in winemaking [4]. The enzyme mixtures used in food industry contain generally other cell wall degrading enzymes in addition to pectinases, e.g. to obtain greater juice yield or clarity in juice production. The composition of these enzyme sets differs significantly between fungal species and this is also observed for the subset of pectinolytic enzymes [5]. A correlation between the composition of enzyme sets and degradation of structural plant cell wall polysaccharides supports the development of tailor-made enzyme preparations utilized for specific applications. Analysis of samples containing multi-protein complexes in order to elucidate a biological function at the molecular level can be achieved by mass spectrometry (MS). MS is used to analyze and possibly identify molecules and compounds based on accurate determination of their masses. Combined with fragmentation of such compounds and determining the resultant fragment masses, a reliable identification scheme can be developed

164 Chapter 7 and used in a wide range of applications. The development of the two ionization methods electrospray (ES) and matrix-assisted laser desorption ionization (MALDI) allowed the routine analysis of biopolymers, which made MS to a viable technique in protein analysis. MALDI has some advantages over ES, such as most of generated ions are detected and the method is more tolerant towards salts and detergents. In MALDI MS a large excess of matrix material is co-precipitated with the molecule to be analyzed and irradiated by nanosecond laser pulses. The matrix impart energy to the biomolecule during desorption and ionization resulting in a degree of fragmentation. Mass spectrometers measure the mass-to-charge ratio of analytes. In MALDI TOF mass spectrometers the principle of mass separation is achieved on separation on the basis of time-of-flight (TOF). After the generation of a short burst of ions by irradiation by the laser beam, the ions are accelerated to a fixed amount of kinetic energy and travel down a flight tube. The mass-to-charge ratio is related to the time it takes an ion to reach the detector. Ions are detected by a channeltron electron multiplier. The smaller ions have a higher velocity and are recorded on the detector first, producing the TOF spectrum. [6, 7] MALDI is routinely used to identify gel-separated proteins in proteomic studies and several databases have accumulated over the years, giving a wealth of reliable data to compare individual findings with. Furthermore, with the use of tandem MALDI MS/MS (such as TOF/TOF MS) the instrument allows generation of peptide mass fingerprint from just one single step of sample preparation, making MALDI a key analytical technique in proteomics [7]. Proteins produced during cultivation of A. sojae were gel-separated and subjected to in-gel digestion for mass spectrometric characterization. Furthermore, the current study provided a pectinolytic enzyme profiling based on the activities determined in the extract of A. sojae, which was compared to the enzyme activity profiles of various commercial pectinase preparations used in food industry. Moreover, the efficiency of A. sojae enzymes was tested and compared to the commercial enzyme preparations in several applications related to food industry, such as maceration and clarification fruit juices.

7.2 Materials and Methods

7.2.1 Materials Chemicals were mainly purchased from AppliChem GmbH (Darmstadt, Germany), except the substrate for detection of exo-PG activity, polygalacturonic acid, and the chemical sodium arsenate dibasic heptahydrate, as well as Pectinex 3XL® (a product of Novozyme Corp.) and pure cellulase (A. niger) and α-amylase (A. oryzae)

165 Chapter 7 preparations were obtained from Sigma-Aldrich Chemie GmbH (Steinheim, Germany). Matrix, α-cyano-4-hydroxy cinnamic acid (HCCA), and HCCA peptide calibration standard for MALDI-MS analysis were purchased from Bruker Daltonics (Bremen, Germany). Trypsin (sequencing grade modified trypsin) utilized for in-gel digestion was obtained from Promega GmbH (Mannheim, Germany). Trifluoro acetic acid (TFA) was ordered from Iris Biotech GmbH (Marktredwitz, Germany). Acetonitrile (ACN), dithiothreitol (DTT), and ready-to-use gel solutions Rotiphorese® Gel 40 (19:1) and Rotiphorese® Gel 40 (37.5:1) were obtained from Carl Roth GmbH & CO. KG (Karlsruhe, Germany). Commercial enzyme preparations Panzym Fino G, Panzym Clair Rapide G, Panzym Pro Clear, Panzym First Yield and Panzym YieldMASH were kindly provided by E. Begerow GmbH & Co. (Langenlonsheim, Germany). Fructozym® P was obtained from ERBSLÖH Geisenheim AG (Geisenheim, Germany).

7.2.2 A. sojae enzyme extracts SSF extracts derived from A. sojae ATCC 20235, which was purchased from Procochem Inc (Teddington, United Kingdom), and its mutant M3 (chapter 4). Both fungal strains were cultivated on wheat bran and sugar beet pulp in the ratio 70:30, wetted at 160 % with 0.2 M HCl, at 30 °C for enzyme production as described in chapter 5. Fermented substrate was harvested after 8 days for enzyme leaching as described in chapter 6. Moreover, some SSF extracts of A. sojae ATCC 20235 were partially purified by ion exchange chromatography (IEXC) and combination of IEXC and size exclusion chromatography (SEC) prior utilization for characterization and application studies (chapter 6). SmF extract utilized in the application studies derived from mutant DH56, which is also descending from A. sojae ATCC 20235 (chapter 4). Submerged culture was performed as described in chapter 4.

7.2.3 Electrophoresis SDS-PAGE was performed according to the method of Laemmli [8] following the procedure described in the technical manual for protein electrophoresis [9]. Briefly: SDS-polyacrylamide gel with an approximately 2 cm stacking buffer zone was cast and samples run in constant current mode at 20 mA/gel, at room temperature. Samples were mixed with 2x treatment buffer in the ratio 2:1. Sample load add up to 10 µL per lane. Protein bands were visualized, using colloidal Coomassie (G-250) staining [10].

166 Chapter 7

7.2.4 Isoelectric focusing (IEF) and 2D - polyacrylamide gel electrophoresis (PAGE) conditions 2D-PAGE was performed primarily based on the methods developed by Görg et al. [11], with slight modifications. Briefly: The first dimension IEF was performed using Immobiline DryStrip gels (IPG strips) pH 3-10 or pH 4-7 (liner gradient) strips (GE Healthcare) which were rehydrated over night at room temperature in rehydration buffer (8 M urea, 1 % CHAPS, 0.4 % DTT and 0.5 % pharmalyte pH 3-10). 100 µL sample, containing concentrated partially purified (by IEXC and SEC) A. sojae protein solution and rehydration buffer in the ratio 1:3, was loaded per strip using anodic cup loading and IEF was performed utilizing a Multiphor II system (GE Lifesciences) and running conditions were followed according to the instructions given by GE Healthcare [12]. After focusing, the gels were reduced (1 % DTT) for 15 min at room temperature and then alkylated (4 % iodoacetamide) for 15 min at room temperature in equilibration buffer (6 M urea, 30 % glycerol, 2 % SDS, 0.05 M Tris-HCl buffer pH 8.8, 0.1 % bromophenol blue). The second dimension, SDS-PAGE, was performed by embedding the equilibrated strips using hot 1 % low-melting agarose solution in SDS tank buffer (0.0025 M Tris, 0.192 M glycine, 0.1 % SDS) containing 0.1 % bromophenol blue on top of handcast 12.5 % polyacrylamide gels containing no stacking gel. The second dimension separation was carried out on the Mini-PROTEAN Tetra cell electrophoresis system (Bio-Rad Laboratories, Inc.) at constant current (20 mA/gel). The gels were fixed and stained following the colloidal Coomassie (G-250) staining procedure [10].

7.2.5 Combination of native PAGE/ zymogram and separation by SDS-PAGE under denaturing conditions for the identification of active zones A method developed in the working group of Prof. Muskhelishvili (Jacobs University Bremen GmbH, Germany), that facilitated the SDS-PAGE under denaturing conditions of samples which were previously separated by native PAGE, was adapted with modifications for the identification of active zones present in the native PAGE. Therefore, proteins were first separated by native PAGE excluding SDS and DTT from the electrophoresis protocol [9]. In addition, a native polyacrylamide gel was prepared by mixing ready-to-use gel solutions Rotiphorese® Gel 40 (19:1) and Rotiphorese® Gel 40 (37.5:1) (Carl Roth GmbH & CO. KG, Germany) in order to achieve the ratio of 32:1 of acrylamide to bisacrylamide. Electrophoresis of both native gel types was performed at constant current (15 mA/gel), at 4°C. Subsequently the “sandwich” method was applied to detect the activity of pectinases acting on polygalacturonic acid sodium salt as substrate [13], starting

167 Chapter 7 with the incubation of the gels in 0.1 M sodium acetate buffer (pH 4.8) for 20 min. Afterwards, the polyacrylamide gels were contacted with an (solid) agar substrate containing 0.25% (w/v) polygalacturonic acid sodium salt for 15 min at 30°C (80% humidity chamber). The agar plates were then treated with 1% (w/v) cetyltrimethylammonium bromide which precipitated the substrate and revealed pectinases activity as translucent bands on an opaque background. The polyacrylamide gels were rinsed with distilled water and placed on a plate above the corresponding zymogram. The parts on the native PAGEs above the active zones, which were visualized in the zymograms, were cut out with a scalpel and the gel pieces were transferred into tubes to follow the procedure provided by the working group of Prof. Muskhelishvili. Therefore, 500 µL SDS-sample buffer pH 6.8 (containing per 100 mL: tris (0.151 g), 86 % glycerol (23.26 mL), bromophenol blue (0.02 g), SDS (4 g)) and 45 µL β-mercaptoethanol were added to the gel pieces and incubated for 10 min at room temperature. The mixture was heated in the microwave for 20 sec at medium stage (600 W) and incubated at room temperature for 15 min. The gel pieces were embedded on top of a handcast 12.5 % SDS- polyacrylamide gel using hot 0.5 % low-melting agarose solution in SDS tank buffer (0.0025 M Tris, 0.192 M glycine, 0.1 % SDS). Electrophoresis was performed at constant current (20 mA/gel). The gels were fixed and stained following the colloidal Coomassie (G-250) staining procedure [10].

7.2.6 MALDI-TOF mass spectrometry

7.2.6.1 Sample preparation – In-gel digestion Once visualization on SDS-polyacrylamide gels was done by colloidal Coomassie staining, spots and bands of interest were manually excised and enzymatically cleaved into peptides following the optimized in-gel digestion protocol of Shevchenko et al. [14], with slight modifications. The detailed protocol of the performed in-gel digestion, which based on the procedure of Shevchenko et al.[14], is given in Appendix A.

7.2.6.2 MALDI-TOF MS A 2-µL drop of sample was spotted onto ground steel target (Bruker Daltonics). After drying down at room temperature, 1.2 µL of matrix solution (0.7 mg/mL α- cyano-4-hydroxy cinnamic acid (HCCA) in TA85 solvent (85:15 (v/v) acetonitrile : 0.1 % trifluoro acetic acid in water)) was added on top of the dried spot according to the HCCA dried droplet protocol (Bruker Daltonics). MALDI-TOF MS and tandem MS (MS/MS) sequencing analyses were performed on an Autoflex II TOF/TOF mass spectrometer (Bruker Daltonics). Calibration of the instrument was performed with

168 Chapter 7 a HCCA peptide calibration standard (Bruker, Daltonics). Analysis of spectra was done using Bruker Daltonics Flex Analysis and BioTools software.

7.2.6.3 Database search The peptide mass fingerprinting and peptide fragment ion data obtained from MALDI-TOF and MS/MS analyses were used to search for protein candidates in the NCBI database. The correlation of mass spectrometric data with the sequence database was achieved by using Mascot (v. 2.0 Matrix Sciences, UK) database searching software.

7.2.7 Application studies

7.2.7.1 Enzymatic extraction of apple juice Enzymatic extraction of apple juice was performed following the procedure described by Nakkeeran et al. [15], with slight modifications. Briefly: Boskoop apples purchased from local market were washed, peeled, core removal was done and apples were cut into small pieces. Apple pulp was mashed using a hand-held blender with addition of ascorbic acid (0.5 g/kg). Enzymatic treatment was done by adding enzyme solutions to the mashed apple pulp in the ratio 1:10 and incubation of apple mash at 50 °C for 2 h. All utilized enzyme solutions were adjusted to 5 U exo-PG activity per g of apple mash. After enzymatic treatment, the mash was heated to 85 °C for 10 min in a water bath for enzyme inactivation. Juice was separated from mash by vacuum filtration for 3 min using filter paper discs grade 292 (87 g/m²) (Sartorius AG, Göttingen, Germany) and analyzed for yield (formula 1) and viscosity (section 7.2.7.5).‎ Blank was incorporated by adding distilled water instead of enzyme solution to the mash. Laboratory treatments were performed in duplicate.

7.2.7.2 Enzymatic extraction of sloe juice Enzymatic juice extraction from stone fruits was performed with regard to the method described by Nakkeeran et al. [15], with slight modifications. Briefly: Sloes were harvested from local shrubs in autumn, washed and frozen at -20 °C for storage until use. Frozen fruits were defrosted and mashed with a hand-held blender. Enzymatic treatment was done by adding enzyme solutions to the mashed blackthorns in the ratio 1:10 and incubation at 40 °C for 2 h. All utilized enzyme solutions were adjusted to 5 U exo-PG activity per g of mash. Blank was incorporated by adding distilled water instead of enzyme solution to the mash. After enzymatic treatment, the mash was heated to 85 °C for 10 min in a water bath for enzyme inactivation. Juice was separated from mash by centrifugation at 3220 ×

169 Chapter 7 g, at 20 °C for 10 min. Yield was determined by measuring the weight and calculated according to the formula 1. Experiments were performed in duplicate.

7.2.7.3 Enzymatic extraction of grape must Enzymatic extraction of grape must was performed according to the preparation of grapes used for microvinification described by Mojsov et al. [16], with slight modifications. Briefly: Grapes of the variety Michele Palieri with origin in Puglia, Italy, were bought at wholesale. Red grapes were washed, destemmed, weighed into beakers and crushed. Dosage for enzyme treatment was done according to the specification of the commercial enzyme preparation using 2 g/hl of “Panzym Clair Rapide G” (E. Begerow GmbH & Co.). This dosage corresponded to an exo-PG activity of 74 U/kg grape mash. Enzymatic treatment was performed at 20 °C over night (12.5 h). Blank was incorporated by adding distilled water instead of enzyme solution to the mash. Must was separated from mash by centrifugation at 3220 × g, 20 °C for 15 min. Yield was determined by measuring the weight and calculated according to the formula 1. Laboratory experiments were performed in triplicate.

7.2.7.4 Estimation of juice yield The juice yield was determined by measuring the weight and using the following equation: Final juice weight (g) Juice yield (%) = × 100 (7.1) Initial pulp weight (g)

7.2.7.5 Determination of viscosity Viscosity was measured using a Micro-Ostwald viscometer (SI Analytics GmbH, Mainz, Germany) at a constant water bath temperature of 25 °C.

7.2.7.6 Clarifying cloudy pure apple juice Cloudy pure apple juice was purchased from local market (beckers bester GmbH, Lütgenrode, Germany). Enzymatic treatment was performed in a water bath at 50 °C applying 5 U exo-PG activity per mL of cloudy juice [15].

7.2.7.7 Vinification Wines were prepared following the method described by Arauner [17], with slight modifications. Briefly: Red grapes of the variety Michele Palieri with origin in Puglia, Italy, bought at wholesale were washed, destemmed, weighed into beakers and crushed. 0.5 kg of mashed grapes was filled into 1L-Schott flasks. Laboratory treatments were performed in duplicate. Potassium metabisulfite was added at a concentration of 0.1 g/kg of mashed grapes. Mashes were macerated for 12.5 h at

170 Chapter 7

20 °C with addition of 74 U exo-PG activity per kg red grape mash. Blank was incorporated by adding distilled water instead of enzyme solution to the mash. Alcoholic fermentations were conducted using Bordeaux Kitzinger pure yeast culture (Paul Arauner GmbH & Co. KG, Kitzingen-Main, Germany). 0.5 kg grape mashes were inoculated with 1.6 mL of pure yeast culture. Treatments were incubated at 23 °C. Two control flasks with partially purified enzyme extract by IEXC were incorporated to monitor the sugar and ethanol concentration during alcoholic fermentation. When the reducing sugar content was below 4 g/L, wines were separated from solids by filtration [18].

7.2.7.8 Ethanol determination Ethanol in wine samples was determined using the ethanol test kit (R-Biopharm AG, Darmstadt, Germany). This UV-method based on an indirect measurement of ethanol by determination of NADH. First ethanol is oxidized to acetaldehyde by nicotinamide adenine dinucleotide (NAD+) in the presence of alcohol dehydrogenase (ADH). Acetaldehyde is further quantitatively oxidized to acetic acid in the presence of aldehyde dehydrogenase. The complete oxidation of ethanol is taking place under alkaline conditions. The generated NADH was determined by means of its light absorbance at 340 nm. The method was modified for application in microtiter plates [19].

7.2.7.9 Determination of sugar concentration in grape must and wine The sugar concentration in grape must and wine samples was determined by measuring of reducing groups using the DNS method [20], which was adapted to test tube scale [21]. The corresponding sugar concentration was determined from absorbance measurements at 575 nm and the standard glucose (0.0 – 1.0 mg/mL) calibration curve (r² = 0.9756).

7.2.7.10 Standard color measurements of red wine Measurements of chromatic characteristics were made using a Shimatzu spectro- photometer (UV-1700 Pharma Spec) at room temperature.

7.2.7.10.1 Color intensity The color intensity (CI) of wines is the sum of absorbance at 420 nm, 520 nm and 620 nm [16].

171 Chapter 7

7.2.7.10.2 Color composition The color composition of wines, expressed as percentage, was calculated as proportion of yellow (% Ye) (λ = 420 nm), red (% Rd) (λ = 520 nm) and blue (% Bl) (λ = 620 nm) dividing the respective absorbance by the color intensity [22].

7.2.7.10.3 Wine color hue

Wine color hue or tint is the ratio of A420 nm to A520 nm [22].

7.2.7.11 Total phenol content Total phenol content was determined spectrophotometrically according to the Folin-Ciocalteu colorimetric method adapted for wine analysis to a micro scale [23]. In brief, 0.79 mL of distilled water was mixed with 0.01 mL sample and 0.05 mL Folin-Ciocalteu reagent. After 1 min, 0.15 mL of 20 % (w/v) sodium carbonate was added and mixed. The mixture was allowed to stand in obscurity at room temperature for 2 h. Absorbance was measured at 750 nm, and total phenol content was calculated from a calibration curve (r² = 0.9985), using gallic acid as a standard. Results were expressed as mg/L gallic acid equivalents.

7.2.8 Exo-polygalacturonase assay Exo-PG activity was assayed according to the procedure of Panda et al. [24], which was further optimized as described in Appendix A.

7.3 Results and Discussion Many fungi secrete plenty of proteins under solid-state culture conditions [25]. Proteins secreted by A. sojae under solid-state conditions were characterized by isoelectric focusing (IEF), gel electrophoresis and mass spectrometry (MS) analysis. Fungal pectinolytic enzymes are extensively used for the extraction and clarification of fruit juices and in winemaking [4]. In the present work extracts derived from A. sojae were characterized for the pectinolytic enzyme profiling and the efficiency of A. sojae enzymes was investigated in various applications related to food industry in comparison to commercial enzyme preparations.

7.3.1 2D-gel electrophoresis 2D-PAGE constitutes a combination of isoelectric focusing and SDS-polyacrylamide gel electrophoresis and thus, the separation of proteins is based on the isoelectric point (pI) in the first dimension and on size in the second dimension. The SSF enzyme extract of A. sojae was subjected to IEXC and SEC for partial purification of PG enzymes (chapter 6). The partially purified PG extract of A. sojae was analyzed for characterization and identification of secreted proteins.

172 Chapter 7

Therefore, active fractions eluted at 0.15 M NaCl concentration of a scaled up IEXC process, applying a three step elution procedure starting with the elution of proteins at 0.15 M NaCl as described in chapter 6 in section 6.3.3.2, were pooled together, concentrated by UF and further fractionated by SEC as described in chapter 6 in section 6.3.3.3. The active fraction obtained by SEC was desalted utilizing pre-packed PD-10 desalting columns (Amersham Biosciences AB) containing Sephadex G-25 medium. This sample was 3.75 times concentrated by freeze-drying and used for separation of proteins by 2D-PAGE (Figure 7‎ .1).

Mw IEF pH 3 – 10 (kDa) 175

80 58

46

30

23

Figure 7‎ .1 2D–PAGE showing separation of partially purified proteins derived from A. sojae cultured under solid-state conditions. Protein extract was electrophoresed in an IPG of pH 3 to 10 (7 cm) in the first dimension and a 12.5 % SDS-polyacrylamide gel in the second dimension.

Proteins in a solution with pH lower than their pI are positively charged, while they are negatively charged in solutions with higher pH than their pI. Applied to an electric field like during IEF, proteins migrate according to their charge in the pH gradient until they reach the pH of their pI, where they become uncharged and stop migrating in the field. As expected, the separation of proteins by 2D-PAGE showed that the proteins present in the partially purified extract had acidic isoelectric points (Figure 7‎ .1). The purification by IEXC utilizing a positively charged anion exchanger results in binding of negatively charged proteins. Since the IEXC process separates proteins on the basis of charge, proteins co-eluting in the same fraction should have similar pIs. Hence, it was expected that the protein spots were not spread over the whole pH range. In order to achieve a better separation of A. sojae proteins, the separation by 2D-PAGE was repeated utilizing IPG strips of pH 4 – 7 (Figure 7‎ .2).

173 Chapter 7

SEC IEF pH 4 – 7 Marker fraction Mw Mw (kDa) (kDa) 175 175 80

80 58

58 46

46

30

30

23

23

Figure 7‎ .2 Separation of A. sojae proteins, which were partially purified by IEXC and SEC prior to 2D- PAGE. Left: Separation on a 12.5 % SDS-polyacrylamide gel; Marker = ColorPlus prestained protein marker (New England BioLabs). Right: 2D–PAGE showing separation of proteins. IEF was performed on an IPG of pH 4 to 7 (7 cm) in the first dimension and separation based on the molecular weights was done on 12.5 % SDS-polyacrylamide gel in the second dimension.

Besides IEXC the sample applied to 2D-PAGE was also partially purified by SEC. Hence, proteins within this sample were expected to have similar pI and also similar molecular weights. However, analyzing these proteins by SDS- and 2D-PAGE presented acidic proteins with a broad molecular weight range (predominantly between 30 – 80 kDa) (Figure 7‎ .2). Various protein spots were excised and subsequently enzymatically digested for mass spectrometric characterization (section 7.3.2.2‎ ).

7.3.2 Identification of proteins by mass spectrometry The protein purification started with crude extract obtained by solid-state culture of A. sojae and ended with a gel-separated protein band or spot. Mass spectrometry (MS) analysis was carried out on peptides obtained after enzymatic degradation of the gel-separated proteins. The combination of separation of proteins by 2D-PAGE or SDS-PAGE and quantification of individual proteins with mass spectrometry and database searching was performed for the identification of the separated proteins present in the partially purified A. sojae extract.

7.3.2.1 MS analysis of SDS-polyacrylamide gel-separated protein bands Partially purified extract produced by A. sojae was subjected to SDS-PAGE for separation of proteins according to their molecular weights. Visualized protein

174 Chapter 7 bands were excised from the gel and subjected to the tryptic in-gel digestion procedure for mass spectrometric characterization of proteins (Figure 7‎ .3 and Figure 7‎ .4).

Load Mw SEC Marker B4 B5 (kDa) 175

βGlcA 80 Gta 58

Taa

βMan 30

23

Figure 7‎ .3 12.5% SDS-PAGE analysis of PG extract proteins partially purified by IEXC (Load SEC) and subsequently fractionation of most active fractions by SEC (B4 & B5). Predominant bands were excised and identified by MS analysis. Identified bands were labeled by the abbreviations of the proteins as described in Table 7‎ .2.

As presented in Figure 7‎ .3 only a limited number of predominant bands of partially purified proteins were identified by MS analysis. In order to achieve a better separation of proteins with higher molecular weights SEC fractions were loaded on gels, where the percentage of acrylamide in the resolving gel was decreased to 10 and 7.5 % (Figure 7‎ .4).

175 Chapter 7

Marker B1 B2 B1 B2 B6 Mw (kDa) Marker B2 175 Mw (kDa) βGlcA αFc 175 80 RglB Gta

58 βGlB Gla Taa 46 80 βMan

30 58

46

23 A B Figure 7‎ .4 SDS-PAGE analysis of PG extract proteins partially purified by combination of IEXC and SEC. Various bands were excised and identified by MS analysis. Identified bands were labeled by the abbreviations of the proteins as described in Table 7‎ .2. A: Analysis of SEC fractions on 10% SDS- polyacrylamide gel. B: Analysis of SEC fraction on 7.5% SDS-polyacrylamide gel.

The decrease of acrylamide in the SDS-polyacrylamide gel resulted in a better separation of protein with higher molecular weight, which yielded in more identified proteins. Majority of the proteins could not be identified by MALDI-MS. Hence, proteins were identified by tandem mass spectrometry (MALDI-TOF MS/MS). Proteins secreted from filamentous fungi are usually highly glycosylated. The glycosylation supports the resistance to proteolysis and makes extracellular proteins stable and soluble in culture medium [1]. The resistance to proteolysis hinders the identification by peptide mass fingerprinting [26]. Oda et al. [26] suggested in-gel deglycosylation for the elimination of sugar chains of fungal proteins to remove their protease resistance and thus, facilitates identifying the extracellular proteins of fungi. Furthermore, they claim to overcome problems resulted from the presence of O-linked oligosaccharides, which results in a diversity of peptide masses, by obtaining peptide amino acid sequences utilizing tandem mass spectrometry (e.g. MALDI-TOF MS/MS). This might be a possible explanation that most of identifications of A. sojae proteins were achieved by MALDI-TOF MS/MS. However, the PG enzyme was not among the identified proteins. Therefore, two commercial pectinolytic enzyme preparations, which were previously tested for the presence of polygalacturonic acid degrading

176 Chapter 7 activity by the enzyme assay (section 7.2.8‎ ), were analyzed by SDS-PAGE and further subjected to mass spectrometric characterization of proteins (Figure 7.5‎ ).

Yield Clair Mw Marker MASH Mw Marker Rapide G (kDa) (kDa) 175 175

80 80

58 58 αAf PGA PGB 46 46

con 09045 PG PmE 30 30

23 23

A B Figure 7‎ .5 Comparative protein analysis of commercial pectinolytic enzyme preparations by 12.5% SDS-PAGE. Predominant bands were excised and identified by MS analysis. Identified bands were labeled by the abbreviations of the proteins as described in Table 7‎ .2. A: Panzym YieldMASH (Begerow); B: Panzym Clair Rapide G (Begerow).

MALDI-TOF MS/MS analysis yielded in the successful identification of PG enzymes in commercial pectinase preparations. The protein pattern of the commercial preparations showed the presence of less protein bands compared to the partially purified sample derived from A. sojae (Figure 7‎ .3). However, there was still a multiplicity of bands in the partially purified extract of A. sojae which could not be identified. A possible reason for no clear identification of analyzed bands might be the presence of more than one protein within the excised band. Thus, 2D-PAGE was applied to separate proteins with similar molecular weights prior to MS analysis.

7.3.2.2 MS analysis of 2D-polyacrylamide gel-separated protein spots Proteins present in A. sojae extract partially purified by IEXC and SEC were further separated on the basis of their pI in the first dimension and according to their molecular weights in the second dimension by subjecting to 2D-PAGE. Various

177 Chapter 7 protein spots were excised and used for mass spectrometric characterization (Figure 7‎ .6).

IEF pH 4 – 7 Mw (kDa) 175

βGlcA 80

58 FDA Gla Gla 46 Taa-G1 Taa-G1 Cbh

30 Lap

23

Figure 7‎ .6 2D-PAGE analysis of PG extract proteins partially purified by combination of IEXC and SEC. IEF was performed on an IPG of pH 4 to 7 (7 cm) in the first dimension and separation based on the molecular weights was done on 12.5 % SDS-polyacrylamide gel in the second dimension. Various spots were excised and identified by MS analysis. Identified spots were labeled by the abbreviations of the proteins as described in Table 7‎ .2.

The separation of proteins by 2D-PAGE resulted in appearance of protein spots of similar molecular weights but differing pIs. However, only in a limited number of protein spots were identified by mass spectrometric characterization and the PG enzyme was not among the identified protein spots (Table 7.2‎ ). Moreover, there were some protein spots of the same molecular weight, but with differing pIs identified for the same protein, which is probably due to post-translational modifications such as oxidation, isoenzyme variation or protein charges [27].

7.3.2.3 MS analysis of proteins separated in two stages combining native PAGE/ zymogram and separation by SDS-PAGE under denaturing conditions In order to identify proteins present in the active zones of degrading polygalacturonic acid sodium salt, which were visualized in a zymogram, a separation by native PAGE was conducted in the first step. This was followed by separation of proteins present in the active zones by SDS-PAGE under denaturing

178 Chapter 7 conditions, which was performed prior to MS analysis (Figure 7.7‎ and Figure 7.8‎ ). The method for SDS-PAGE analysis of proteins that were previously separated by native PAGE, was developed in the working group of Prof. Muskhelishvili (Jacobs University Bremen GmbH, Germany). This method was combined with the identification of polygalacturonic acid degrading enzymes by the “sandwich” method (zymogram).

Z1 Z2 Z3 Z4 Z5

A Marker Z1 Z2 Z3 Z4 Z5 Mw (kDa) 80 58

βMan βMan 30 Lap βXyn βXyn

23

B

Figure 7‎ .7 Analysis of proteins present in the “active zones” of a native PAGE gel. A: Zymogram for visualization of active zones. B: Analysis of active zones on 12.5 % SDS-polyacrylamide gel under denaturing conditions.

The substrate agar plates revealed the activity of polygalacturonic acid degrading enzymes, which were present in both types of native polyacrylamide gels (section

179 Chapter 7

7.2.5‎ ). However, the active zones on the substrate agar plate in Figure 7.7‎ -A indicated a longer migration of substrate degrading proteins within the native PAGE gel prepared according to the procedure described in the technical manual of protein electrophoresis by omitting SDS from the gel [9]. Active zones within the native PAGE gel prepared by mixing ready-to-use gel solutions Rotiphorese® Gel 40 (19:1) and Rotiphorese® Gel 40 (37.5:1) (Carl Roth GmbH & CO. KG, Germany) in order to achieve the ratio of 32:1 of acrylamide to bisacrylamide resulted in a lower extend of protein migration (Figure 7.8‎ -A), and thus might have yielded in lesser separation of proteins.

Z1 Z2 Z3 Z4 Z5

A Marker Z1 Z2 Z3 Z4 Z5 Mw (kDa) 80 58 Gla

30 Taa βXyn βXyn

23

B Figure 7‎ .8 Analysis of proteins present in the “active zones” of 32:1 native polyacrylamide gel. A: Zymogram for visualization of active zones. B: Analysis of active zones on 12.5 % SDS-polyacrylamide gel under denaturing conditions.

180 Chapter 7

As can be seen from Figure 7‎ .7-B and Figure 7‎ .8-B the method for analyzing proteins present on native polyacrylamide gels by further subjecting to SDS-PAGE under denaturing conditions still needs to be improved. Nevertheless, some proteins were successfully extracted from the native polyacrylamide gel and further separated under denaturing conditions. Moreover, the aforementioned assumption that native PAGE utilizing ready-to-use Rotiphorese® gel solutions resulted in lower protein separation was supported by the presence of more stained proteins in Figure 7‎ .8-B. Various proteins were identified by tandem mass spectrometry, which were present within the active zone causing the degradation of polygalacturonic acid sodium salt on the substrate agar plate (Table 7‎ .2). However, there was no PG enzyme among the identified proteins of A. sojae. Oda et al. [26] provided a comparative proteomic analysis of extracellular proteins from A. oryzae grown under submerged and solid-state conditions facing similar low identification rates of proteins by mass fingerprinting. They assumed that proteins in the gel might be resistant to trypsin digestion so that a mass fingerprint could not be obtained. Majority of A. oryzae proteins had acidic pIs, indicating the presence of few basic lysine and arginine residues, which are specific sites of trypsin digestion. Hence, Oda et al. [26] suggested the use of other peptidases, such as asparaginylendopeptidase, which could be more successful at digesting the protein and would than allow identification by peptide-mass fingerprinting. All analyzed proteins derived from A. sojae had acidic pIs (section 7.3.1‎ ). Hence, this could be a possible explanation that PG enzymes could not be identified due to a resistance to trypsin digestion. Nevertheless, a broad range of carbohydrate-active enzymes, including glycoside hydrolases and polysaccharide lyase, were identified in the extract of A. sojae (Table 7.2‎ ). For instance, A. sojae produced cellobiohydrolase and β-glucosidase, which are proteins acting on cellulose [28]. Furthermore, β- glucosidase as well as β-galactosidase and α-fucosidase were produced, which are enzymes degrading xyloglucan [2, 28]. Moreover, xylanases and β-galactosidase were identified, which are acting on xylan, whereas β-galactosidase is also involved in the degradation of galactomannan, similar as β-1,4-mannosidase [2, 28]. Xylanases are utilized in baking industry, e.g. for modification of cereal flours [29], whereas β- mannosidases are applied in the extraction of vegetable oils, as well as during the manufacture of instant coffee [30, 31]. A. sojae extract contained also glucoamylase and α-amylase, which are acting on starch [2]. Glucoamylases are industrial important biocatalysts and are applied in the production of crystalline glucose or glucose syrup [32]. Alpha amylases are also industrial important enzymes used in the starch saccharification [33]. Moreover, rhamnogalacturonan lyases were

181 Chapter 7 identified, which cleave within the hairy regions of pectin, and pectin structures xylogalacturonen and rhamnogalacturonan also require accessory enzymes to remove the side chains, such as β-galactosidases that are produced by A. sojae [2]. The production of these enzymes by A. sojae was performed on wheat bran and sugar beet pulp. These crude plant compounds are cost-efficient substrates that induced the production of a large number of carbohydrate-active enzymes which was reflected by the identification of various enzyme involved in the degradation of different substrates. Plant cell wall degrading enzymes have a broad spectrum of industrial applications which is expected to further expand and a growing world market to $2 billion is expected [29]. Tari et al. [34] claimed that crude polygalacturonase extracts can have great applications in the industry given that other pectinolytic enzymes and different enzymes can exist in the extract. Moreover, Tari et al. [34] and Dogan & Tari [35] provided a biochemical and thermal characterization of crude and three-phase partitioned exo-PG produced by A. sojae ATCC 20235 in submerged culture, which is summarized in Table 7.1‎ .

Table 7‎ .1 Biochemical properties of exo-PGs derived from cultivation of A. sojae ATCC 20235 in SmF [34, 35].

Parameter Crude exo-PG Three-phase partitioned exo-PG

pH optimum 5.0 4.0

Retained more than 60 % of its Retained over 73 % of its activity in the pH stability activity in the pH range 3.0 – pH range of 4.0 – 9.0 7.0

Temperature optimum 55 °C 55 °C

Temperature stability Residual activity of 65 – 80 % at Incubation at temperatures of 25 – 65 incubation temperatures of 37, °C for 10 – 60 min revealed an increase 45, 55 and 65 °C in stability with increasing temperature and time until 55°C and 20 min

Kinetic constant Km 0.424 g/L 0.751 g/L

Kinetic constant vmax 80 µmol/min 1.139 µmol/min

Inhibition studies - Enzyme activity was induced in the presence of 1 mM of K+, Na+, Ca2+ ions and completely inhibited in the presence of Mn2+ as well as in the presence of 0.1 % SDS.

182 Chapter 7

The optimum pH and temperatures of crude and three-phase partitioned exo-PG were very close. However, the thermodynamic study revealed that the crude PG was more stable, who could be explained by stabilizing effects of compounds present in the crude extract [36]. Tari et al. [34] reported a thermal stability of crude PG at 75 °C and its easy inactivation at high temperatures, e.g. 3.53 lower half-life at 85 °C. This observation is essential with regard to industrial application of PG enzyme in fruit juice production, e.g. the polishing process in industrial apple juice production for the clarification is performed at 50 °C, which is followed by heat treatment at 85 °C for enzyme inactivation (Fahner Frucht GmbH, Gierstädt, Germany). These findings confirm the claim of Tari et al. [34] that crude polygalacturonase extracts can have great applications in the industry, especially in combination with other pectinolytic enzymes and different enzymes that might exist in the extract. However, their SDS-PAGE profiling revealed only the presence of three major bands in the SmF extract of A. sojae ATCC 20235. The solid-state cultivation of this strain presented in this study yielded in the production of a broad enzyme spectrum, which might be promising in industrial application with regard to the findings of Tari et al. [34]. Moreover, the presence of other pectinolytic activities in the SSF extract, was already demonstrated in the previous chapters. Nevertheless, in order to identify the presence of further pectinases a characterization of A. sojae SSF extracts was performed on the basis of pectinolytic enzyme activities.

183 Chapter 7

Table 7‎ .2 Summary of proteins identified by MS analysis.

Sample and band/ spota Protein Species NCBI database Mw (kDa) pI Scoreb

Identified A. sojae protein spots and bands Gla Glucoamylase Aspergillus oryzae gil 2980896 52.4 4.68 63 RglB Rhamnogalacturonate lyase B Aspergillus oryzae gil 317136669 73.7 4.87 213 βGlcA Beta-glucosidase A Aspergillus oryzae gil 169764719 93.4 4.86 77 Gta Glutaminase Aspergillus oryzae gil 9954449 76.2 4.88 77 Taa α-Amylase Aspergillus sojae gil 393659848 54.6 4.54 69 βMan Endo-1,4-beta-mannosidase Aspergillus oryzae gil 391866252 41.9 4.6 90 αFc Alpha-fucosidase Aspergillus oryzae gil 317138010 86.4 5.34 57 βGlB Beta-galactosidase B Aspergillus oryzae gil 169769625 108.6 5.0 60 Taa-G1 Taka-amylase (Taa-G1) precursor Aspergillus oryzae gil 166531 54.7 4.55 118 Cbh Cellobiohydrolase, putative Aspergillus flavus gil 238498612 43.2 4.33 134 FDA FDA binding domain protein Aspergillus oryzae gil 169765087 61 4.93 106 Lap Leucine amnopeptidase 1 Aspergillus oryzae gil 169782566 41.1 5.03 131

A. sojae proteins present in active zones of zymogram βXyn Beta-1,4-xylanase Aspergillus oryzae gil 391864184 33.6 4.62 96 Taa α-Amylase Aspergillus sojae gil 393659848 54.6 4.54 112 Gla Glucoamylase Aspergillus oryzae gil 2980896 52.4 4.68 281 Lap Leucine amnopeptidase 1 Aspergillus oryzae gil 169782566 41.1 5.03 72 βMan Endo-1,4-beta-mannosidase Aspergillus oryzae gil 391866252 41.9 4.6 184

184 Chapter 7

Commercial enzyme preparations PGA Endopolygalacturonase A Aspergillus kawachii gil 74664131 38.7 3.99 70 con 09045 Hypothetical protein RO3G 09045 Rhizopus delemar PE Pectinesterase Aspergillus tubingensis gil 130345 35.7 4.21 73 αAf Alpha-N-arabinofuranosidase B precursor Aspergillus kawachii gil 74620984 52.6 4.27 134 PGB Exopolygalacturonase B Aspergillus oryzae gil 83776451 48.4 4.91 54 PG 2 hits: Polygalacturonase, chain A Aspergillus aculeatus gil 15988279 34.6 4.7 201 Endopolygalacturonase I Aspergillus aculeatus gil 61214424 38.5 4.83 201 a Band and spot names correspond to the abbreviations for the proteins in Figure 7.3‎ – Figure 7.8‎ . b Only results for proteins that had scores of greater than 50, matching with the molecular weights and were reproducibly identified are shown.

185 Chapter 7

7.3.3 Characterization of A. sojae crude extracts on the basis of pectinolytic enzyme activities A comparative study on pectinolytic enzymes produced by A. sojae was performed in order to characterize the enzymes on the basis of their substrate degrading mode. Results were compared to commercially available pectinase preparations for the identification of specific pectinolytic enzyme profiles with regard to their applications. Therefore, crude extract derived from A. sojae ATCC 20235, as well as crude extract produced by its mutant M3 (chapter 4) was analyzed for pectinase activities (Figure 7‎ .9). The values of the pectinolytic enzyme profile activities were obtained within a co-operation with the ITSM (Monterrey, México) and were kindly provided by Marco Mata (guest scientist of ITSM). The corresponding enzyme assays performed besides the assay for detection of exo-PG activity are described in Appendix A, in section 1.1.4.

Figure 7‎ .9 Pectinolytic enzyme activities produced after 8 days SSF under optimized culture conditions for PG production:  A. sojae ATCC 20235;  mutant M3.

The values presented in Figure 7‎ .9 confirmed previous results that the applied medium composition and fermentation condition, which were optimized for enhanced exo-PG titers, yielded in predominantly exo-PG enzyme production (chapter 3). Moreover, these results also approved the successful strain improvement for enhanced PG production (chapter 4). However, it can be also seen that other pectinolytic enzyme activities were detected in the extract of A. sojae, even though the corresponding enzymes could not be identified by MS analysis.

186 Chapter 7

In order to achieve comparative pectinolytic enzyme profiles, the enzymes produced by A. sojae, as well as commercially available enzyme preparations used in wine making and fruit juice industry, were transferred into 0.1 M sodium acetate buffer (pH 5) and adjusted to 100 U/mL exo-PG activity (Table 7‎ .3).

Table 7‎ .3 Summary of pectinolytic enzyme activities*.

Sample exo-PG endo-PG exo-PMG PL PME

A. sojae ATCC 20235 97.8 38.0 12.0 14.7 0.27 Mutant M3 104.9 40.2 11.0 16.5 0.26

Commercial enzyme preparations Fino G 99.3 10.1 14.5 21.3 5.8

Clair Rapide G 107.0 29.5 25.8 35.7 5.9

Pro Clear 94.9 31.4 13.1 333.5 21.0

First Yield 97.2 25.5 35.9 491.0 24.3

YieldMASH 108.8 0.6 8.9 107.7 27.1

Fructozym P 91.9 7.3 15.3 495.9 38.6 * Values correspond to the average of 3 tests.

Even though mutant M3 produced approximately double endo-PG activity compared to the wild strain, the process of transferring the enzymes in sodium acetate buffer and adjusting exo-PG concentration to 100 U/mL, which included the steps of dialysis, lyophilization, rehydration in buffer and desalting in PD-10 columns prior to analysis, resulted in similar enzyme profiles in the extracts of both strains (Table 7‎ .3). Hence, only the extract of A. sojae ATCC 20235 was considered for the comparison to commercially available enzyme preparations presented in the radar graphs (Figure 7‎ .10 and Figure 7‎ .11).

187 Chapter 7

endo-PG (0 – 40 U/mL)

PME exo-PMG (0 – 40 U/mL) (0 – 40 U/mL)

PL (0 – 40 U/mL) Figure 7‎ .10 Comparison of A. sojae pectinolytic enzyme profile to commercial enzyme preparations used in wine making. All extracts were adjusted to 100 U/mL exo-PG activity; — A. sojae ATCC 20235; — Fino G (E. Begerow GmbH & Co.); — Clair Rapide G (E. Begerow GmbH & Co.).

Comparing the extract of A. sojae to pectinolytic enzyme preparations used in wine making revealed a higher endo-PG present in the extract of A. sojae (Figure 7‎ .10). Other pectinolytic activities of A. sojae were similar to the enzymatic profile of the commercial preparation “Fino G” (E. Begerow GmbH & Co., Germany). This preparation is an enzymatic mixture containing pectinases and β-glucanases produced by A. niger and Trichoderma harzianum under submerged conditions. It is used for the degradation of pectic and glucan substances in grape must. The enzyme preparation “Clair Rapide G” (E. Begerow GmbH & Co., Germany) contained higher exo-polymethylgalacturonase (exo-PMG) and pectinlyase (PL) activities. Moreover, both commercial enzyme preparations used in wine making contained a higher pectin methylesterase (PME) contend, while the A. sojae extract contained only low amount of PME. The activity of PME leads to the release of methanol. Hence the almost absence of this enzyme activity in the extract of A. sojae might be an advantage, since no additional purification step is needed for the removal of this enzyme.

188 Chapter 7

A. sojae ATCC 20235

Pro Clear

Fructozym P

First Yield

YieldMASH

Figure 7‎ .11 Comparison of A. sojae pectinolytic enzyme profile to commercial enzyme preparations used in fruit juice industry.

The comparison of the A. sojae extract to commercial preparations which are used in the fruit juice industry (e.g. for maceration and clarification) showed a significant higher amount of PL present in commercial enzyme preparations (Figure 7.11‎ ). PL degrades pectin directly by β-elimination mechanism which leads to the formation of 4,5-unsaturated oligogalacturonides, while other pectinases have to sequently act in combination for the complete degradation of the pectin molecule [37]. Hence, this enzyme has broad applications, e.g. easy clarification of apple juice was reported with the application of pure PL [4]. Similar to the aforementioned comparison contained the extract of A. sojae higher endo-PG activity than the commercial preparations, which might conduce to viscosity reduction. Furthermore, contained the A. sojae extract the lowest amount of PME activity, which might be beneficial as described before. The exo-PMG activity was close to three commercial preparations (Fructozym P, Pro Clear and YieldMASH). “YieldMASH” is used for the maceration of stone fruits and “Pro Clear” is used for the degradation of pectic substances in apple and pear juices. Whereas, “Fructozym P” is applied for the maceration and clarification of stone fruits, as well as for apple juice processing. The commercial enzyme preparation “First Yield”, which is utilized for maceration of stone fruits, contained almost three times more exo-PMG activity compared to the extract of A. sojae.

189 Chapter 7

7.3.4 Application testing of A. sojae extract The MS analysis and pectinolytic enzyme profiling showed that A. sojae co-produced other enzymes along with PG in SSF, which might have beneficial effects in terms of certain applications. Hence, A. sojae crude extract and partially purified extracts were tested in comparison to commercial enzyme preparations in several applications related to food industry.

7.3.4.1 Application of A. sojae extract for juice extraction studies

7.3.4.1.1 Extraction of apple juice Addition of pectinases to fruits prior to crushing and pressing may yield in a substantial increase in juice recovery [38]. The efficiency of crude and partially purified enzyme extracts of A. sojae was tested for extraction of apple juice in comparison to commercial available pectinolytic enzyme preparations. Therefore, apple mash made of apples of the variety Boskoop, that were purchased from local market (Bremen, Germany), was enzymatically treated with different enzyme extracts as described in section 7.2.7.1‎ applying 5 U exo-PG activity per g of apple mash. Extracted juice was analyzed for yield and viscosity (Table 7.4‎ ).

Table 7‎ .4 Enzymatic extraction and effect on juice viscosity of various enzyme solutions.

Viscosity Viscosity Sample Yield (%) (mm2/s) reduction (%) Blank 47.4 10.88 A. sojae ATCC 20235 crude extract 56.4 4.78 56.1 Partially purified extract (IEXC, 0.15 M NaCl)1 55.8 5.68 47.8 Partially purified extract (IEXC, 0.10 M NaCl)1 58.6 5.17 52.5 Mutant 3UV crude extract 54.1 6.81 37.4 Fructozym P 72.3 1.44 86.8 Pectinex 3XL 76.8 1.39 87.2 First Yield 75.8 1.33 87.8 YieldMASH 68.4 1.53 85.9 Pro Clear 77.4 1.35 87.6 Cellulase² 72.5 1.67 84.6 α-amylase² 39.2 15.82 - 1 Partial purification of PG enzyme by IEXC as described in chapter 6 in section 6.3.3.2 ² Pure enzyme preparations (Sigma-Aldrich Chemie GmbH; Steinheim, Germany)

190 Chapter 7

The enzymatic treatment of apple mash with crude and partially purified PG extracts of A. sojae and commercial pectinase preparations resulted in a greater juice recovery compared to the control. However, the obtained amounts by utilization of A. sojae extract were lower than recovery yields of commercial pectinolytic enzyme preparations. Furthermore, the viscosity of apple juices could be decreased of about 50 % applying A. sojae extracts. Nevertheless, commercial enzyme mixtures were more efficient in viscosity reduction. The partially purification of PG extracts by IEXC did not significantly influence juice recovery yields or viscosity. Besides A. sojae extracts and commercial pectinolytic enzyme preparations, the efficiency of pure cellulase (A. niger) and α-amylase (A. oryzae) preparations was tested for apple juice extraction. Utilizing cellulase enzyme yielded in similar juice recovery like commercial pectinase preparations. Also the reduction in viscosity was comparable to pectinolytic enzyme preparations. Commercial pectinolytic enzyme preparations usually contain mixtures of pectinases associated with cellulytic, proteolytic and other species of enzymes apart from the main pectinases [39], and the observed results during applications cannot easily attributed to the action of a single enzyme [38]. However, according to the present observation is the presence of cellulase significantly affecting the process of maceration, and thus influencing the juice extraction. The application of pure α-amylase resulted in lower apple juice extraction with higher viscosity in comparison to the control.

7.3.4.1.2 Extraction of sloe juice The commercial pectinolytic enzyme preparation “Panzym YieldMASH”, is praised for high PG activity that supports maceration of stone fruits. Due to the identification of PG enzyme in “Panzym YieldMASH” and expected presence of PG enzyme in A. sojae extracts, which is based on the enzyme assay, A. sojae extracts were also applied for juice recovery from stone fruits. Enzymatic extraction of juice from drupes was tested using blackthorn (Prunus spinosa). Blackthorn or “sloe”, which is the common name of the fruit, is a round drupe of approximately 10 mm in diameter with purplish-black color of the fruit skin. Enzymatic treatment was done by applying 5 U exo-PG activity per g of sloe mash and the yield of recovered juice was determined (Table 7‎ .5). The enzymatic treatment of sloe mash with crude and partially purified PG extract of A. sojae and commercial pectinase preparations yielded in a greater juice recovery compared to the control. In case of “Panzym YieldMASH”, the obtained juice yield was only 6 % higher in comparison to the yield obtained with the A. sojae

191 Chapter 7 crude extract. Regarding the fact that “Panzym YieldMASH” is a pectinolytic enzyme preparation produced by Aspergillus sp. in submerged culture, the generally cost-efficient solid-state production of the A. sojae enzyme mixture might be a potential alternative for the production of enzyme preparations used for maceration of stone fruits. However, the juice recovery obtained with A. sojae crude extract was slightly higher than using the partially purified extract. A possible explanation might be the separation of enzymes during the process of IEXC, which supported the maceration of blackthorns.

Table 7‎ .5 Blackthorn juice yield after enzymatic extraction.

Sample Yield (%) Blank 10.0 YieldMASH 33.3 Fructozym P 39.7 A. sojae ATCC 20235 crude extract 26.9 Partially purified extract (IEXC, 0.15 M NaCl) 22.2

7.3.4.1.3 Extraction of grape must The presence of PG enzymes was also detected in a commercial preparation that is used in wine making for the degradation of pectic substances in grape must (section 7.3.2.1‎ ). Hence, the efficiency of crude and partially purified enzyme extracts of A. sojae was tested for the extraction of grape must, too (Table 7.6‎ ).

Table 7‎ .6 Juice recovery after enzymatic grape maceration.

Sample Yield (%) Blank 42.8 Clair Rapide G 58.8 A. sojae ATCC 20235 SSF crude extract 36.1 Partially purified extract (IEXC, 0.15 M NaCl) 39.5 Mutant DH56 SmF crude extract 40.5

The enzymatic treatment applying the same dosage of exo-PG activity resulted only in increased juice recovery with the commercial enzyme preparation “Panzym Clair Rapide G”. In comparison to the control experiment, the treatment with extracts of A. sojae did not enhance the grape juice extraction. Nevertheless, commercial pectinolytic enzyme preparations used in wine making are not only applied for

192 Chapter 7 increased must recovery rather than they are also utilized to improve wine characteristics, e.g. for efficient extraction of desirable red grape pigments and other phenol compounds [16, 40]. Therefore, extracts 0f A. sojae were also tested for their application in wine making (section 7.3‎ .4.3).

7.3.4.2 Application of A. sojae extract for apple juice clarification studies Fruit juices are naturally cloudy due to the presence of suspended solids. High concentration of pectin leads to colloid formation. For the production of clear juice the suspended particles will be removed by precipitation prior to filtration. The precipitation step is called clarification. The enzymatic depectinization of fruit juices leads to efficient reduction of turbidity. Pectinases are the major types of enzymes used in apple juice processing, e.g. clarification of apple juice could be obtained by a mixture of PG and PME [4]. A. sojae enzyme extract was tested for clarification efficiency of cloudy pure apple juice in comparison to a commercial available pectinolytic enzyme preparation. The enzyme concentration was adjusted to 5 U exo-PG activity per mL of cloudy juice in all experiments. After 1.5 h incubation at 50 °C a visual clarifying effect was already observed in the juice treated with the commercial pectinase preparation “Fructozym P”, while juice treated with A. sojae extract was only partially clarified and remained turbid (Figure 7.12‎ -A). Further incubation of the juice treated with A. sojae extract over night at 21 °C could not further enhance the clarification effect, while application of the commercial preparation produced a clear juice (Figure 7.12‎ - B). In comparison to the blank caused the treatment with A. sojae extract a partial precipitation, but the juice remained unclear.

FructozymP A. sojae FructozymP A. sojae Blank Cellulase Blank SSF extract SSF extract (Sigma)

A B C Figure 7‎ .12 Clarifying effect of enzyme preparations; A: After 90 min incubation at 50 °C; B: Additional 17.5 h incubation at 21 °C; C: Control and juice treated with cellulase for 90 min at 50 °C and additional 17.5 h at 21 °C.

193 Chapter 7

An addition to the pectinase preparations was also the application of pure cellulase (derived from A. niger; Sigma-Aldrich Chemie GmbH) for apple juice clarification tested, which resulted also in partially clarification of the juice (Figure 7.12‎ -C). However, the juice treated with cellulase looked more clarified than the one treated with the A. sojae extract. There is usually a large number of enzymes present in commercial preparations and the observed results cannot easily attributed to the action of a single enzyme [38]. However, according to this observation, the presence of cellulase would support the clarifying effect of the enzymatic treatment. Factors influencing clarification are pH, temperature, contact time and enzyme concentration [4]. The temperature was fixed at 50 °C following the conditions applied during the polishing process in industrial apple juice production (Fahner Frucht GmbH, Gierstädt, Germany). It resulted in clarification applying the commercial enzyme preparation. Generally, the clarification increases with increasing temperature as long as the temperature is below denaturation temperature for the enzyme [4]. Tari et al. [34] claimed that A. sojae polygalacturonase produced in submerged culture is thermostable up to 65 °C and thus, would have great potential for the application in fruit juice industry. According to the findings of Tari et al. [34], one could assume the applied temperature of 50 °C should be below the denaturation temperature, which would eliminate the speculation of low activity due to enzyme inactivation. The contact time of 1.5 h at 50 °C was sufficient for juice clarification by the commercial enzyme preparation. Nevertheless, even further incubation over night (17.5 h) at 21 °C did not further enhance the clarification effect of A. sojae extract. The applied enzyme concentration was adjusted to 5 U exo-PG activity per mL of cloudy juice in all experimental setups, which yielded in clear juice with commercial pectinase. However, as aforementioned contained the commercial preparation a couple of other pectinolytic enzyme activities, such as a higher PME content. The combination of PG and PME was successfully applied for clarification of apple juice [4]. However, the almost absence of PME activity in extracts derived from A. sojae might have negatively influenced the clarification effect. Moreover, pectinolytic enzyme preparations utilized in fruit juice industry contained a significant high amount of PL activity, which was very low in the extract produced by A. sojae (section 7.3.3‎ ). PL is capable of degrading highly esterified pectins and it was reported about the efficiency of PL in juice clarification [4, 37]. Hence, the low PL production by A. sojae might be the main reason for the low productivity in clarifying of apple juice.

194 Chapter 7

7.3.4.3 Application testing of A. sojae extract in red winemaking Pectic enzyme preparations are used for a more efficient extraction of desirable red grape pigments and other phenol compounds bound in plant cells. Moreover, they are used to shorten the time of maceration, settling and filtration [40]. During this study the effect of crude and partially purified enzyme extracts derived from A. sojae was tested in winemaking and compared to a commercial enzyme preparation.

7.3.4.3.1 Maceration and alcoholic fermentation of red wines Enzymes play a fundamental role in the process of winemaking by breaking down the cellular structure of the grape skin and thus, supporting the release of compounds which improve the wine characteristics [16]. Enzymatic treatment in the form of maceration of crushed grapes was started prior alcoholic fermentation with the addition of enzyme extracts with 74 U exo-PG activity per kg red grape mash. The enzymatic treatment in the grape musts resulted in slightly higher sugar release by addition of commercial enzyme preparation “Panzym Clair Rapide G” and the extract obtained by mutant DH56 in SmF in comparison to the blank (Table 7.7‎ ). Extracts obtained by A. sojae under solid-state conditions did not show any significant effect on grape must compared to the blank experiment. The alcoholic fermentation of the grape must was started with the addition of yeast as described in section 7.2.7.7‎ . Alcoholic fermentation is the transformation of sugars into ethanol and carbon dioxide by yeast under anaerobic conditions. Many biochemical, chemical and physicochemical reactions take place during this process to turn grape must into wine. Several species of yeasts may be present in the grape mash depending on several factors, like grape variety, ripening stage, and antifungal treatments [41]. The addition of sulfur dioxide and inoculation of a selected yeast culture should help to prevent undesirable yeast developing. The carbohydrate consumption and ethanol production by Bordeaux Kitzinger pure yeast culture during the process of alcoholic fermentation was monitored in two control flasks. Enzymatic treatment of control flasks was achieved with partially purified enzymes derived from A. sojae. The end of the alcoholic fermentation was taken as lowering of sugar concentration to < 4 g/L (Figure 7.13‎ ).

195 Chapter 7

200

180 50

160 Ethanol concentration (g/L) Ethanol 40 140

120 30 100

80 20

60 Sugar concentration (g/L) concentration Sugar

40 10

20

0 0 0 1 2 3 4 5 6 7 Time (d)

Figure 7‎ .13 Alcoholic fermentation of grape mash in control flasks: (——) sugar concentration, (——) ethanol concentration.

After 7 days cultivation the sugar level was below 4 g/L. During the time course of fermentation a consistent increase of ethanol concentration was observed (Figure 7.13‎ ). At the end of alcoholic fermentation, wines were filtered through cotton cloths and analyzed for sugar and ethanol concentration (Table 7.7‎ ).

Table 7‎ .7 Sugar and ethanol concentration of grape musts and wines.

Sugar Sugar Ethanol Sample concentration concentration concentration must (g/L) wine (g/L) wine (g/L)

Blank 163.6 ± 2.2 2.8 ± 0.5 43.3 ± 1.1

Clair Rapide G 182.2 ± 1.5 2.4 ± 0.0 40.5 ± 0.3

A. sojae ATCC 20235 SSF crude extract 159.9 ± 5.3 2.5 ± 0.4 42.8 ± 0.8

Partially purified extract (IEXC, 0.15 M NaCl) 163.6 ± 9.6 3.0 ± 0.5 42.9 ± 1.5

Mutant DH56 SmF crude extract 171.8 ± 1.0 1.6 ± 0.1 41.6 ± 0.6

Within all experimental approaches the sugar concentration was below 4 g/L and the resulting ethanol concentration was above 40 g/L after 7 days of fermentation. There was no significant difference between the blank and enzymatic treated experiments.

196 Chapter 7

7.3.4.3.2 Enzymatic treatment for improvement of wine characteristics In assessing the quality of a wine some sensory characteristics are used which develop during the winemaking process through biochemical reactions. Most of these reactions are catalyzed by different enzymes. Exogenous enzymes are an important component of modern winemaking. The extent of certain enzymatic reactions may determine the efficiency of specific steps during winemaking, e.g. wine clarification, filterability, and color extraction [42]. The color of a red wine provides an important initial impression. It significantly influences the judgment about its quality. The red color in grapes and young red wines results mainly from the pigments that are present, the anthocyanins [43]. A summary of the chromatic characteristics of the young red wines is given in Table 7.8‎ . There was no increase in color intensity detected in the wines which were subjected to enzymatic treatment prior alcoholic fermentation. Furthermore, the hue values of the wines which were subjected to enzymatic treatment were also lower in comparison to the blank, except for the wines treated with extract obtained by SmF. Generally, the hue of young wines is lower and increases from 0.4 – 0.5 to 0.8 – 0.9 in aged wines [43]. It is describing a shift from purple red to brown tones of the wine color during aging. Hence, lower hue values were desired since it represents a degree of brownness.

Table 7‎ .8 Summary of chromatic characteristics of red wines*.

CIa Hue Ye/%b Rd/%c Bl/%d Enzymatic treatment Blank 2.82 0.80 39.2 49.0 11.8

Clair Rapide G 2.23 0.62 34.5 55.4 10.1

A. sojae ATCC 20235 SSF crude extract 2.28 0.67 34.5 52.9 12.5

Partially purified extract (IEXC) 2.26 0.76 38.2 50.5 11.3

Mutant DH56 SmF crude extract 2.66 0.84 39.9 48.1 12.0 * The values are average from two replicates. a CI: color intensity b Ye%: percentage of yellow contribution c Rd%: percentage of red contribution d Bl%: percentage of blue contribution

Evaluation of the % Rd indicated that the highest value of red component was achieved utilizing the commercial enzyme preparation. SSF extracts of A. sojae presented only a slight increase in % Rd, while the extract obtained by SmF yielded even in lower value of the red component in comparison to the blank. The higher

197 Chapter 7 color intensity of the blank resulted from the higher % yellow, which is also indicated by the higher hue of the blank. The same applies for wines treated with the extract derived from SmF. The optimal ratio between the components of red wine color was considered to be Ye:35%, Rd:55%, and Bl:10% [22]. A similar component profile was obtained in the wines which were treated with the commercial preparation “Clair Rapide G”, and with the SSF extract obtained by A. sojae during enzymatic maceration. All other wines had a higher yellow and a lower red component. In summary, the A. sojae enzyme extracts produced in SSF only slightly improved the characteristics of the young red wines. Revilla & González-San José [18] claimed that utilization of pectolytic enzymes maintained chromatic intensity and lower loss of red over a storage period of 2 years. Hence, in order to evaluate the stability, the wines produced in the present study have to be stored and re-analyzed for the chromatic characteristics after the storage period. According to the results of Revilla & González-San José [18], this would lead to improved characteristics of red wines treated enzymatically. This application was not tested within the present study due to the long storage period of two years. However the findings of Revilla & González-San José [18] seemed to be a potential approach for application of A. sojae enzymes. Besides the visual characteristics of wines also the effect of enzymatic treatment on total polyphenol content was investigated (Figure 7.14‎ ). Polyphenols are characterized by the presence of more than one phenol group per molecule. They are suggested to be powerful antioxidants that can provide potential health benefits. Phenolic substances are naturally present in plant material. For instance, common polyphenols found in red grapes include the colored anthocyanins (which can be extracted from the skin), and flavonols (which mainly derived from seeds) [43, 44]. Pinelo et al. [44] reviewed on the positive effect of pectinases during maceration on enhancing the phenolic content of wine during processing and conservation. The enzyme extracts utilized in the present study seemed to have no impact on the extraction of polyphenol (Figure 7.14‎ ). A slight increase in the phenol content was observed in all samples during the process of alcoholic fermentation. However, about the impact of the enzymatic treatment on the conservation can be only speculated, since this application of A. sojae extracts stilled needs to be proved.

198 Chapter 7

800

700

600

(mg/L) 500

400 olyphenols

p 300

Total Total 200

100

0 Blank ClairClair Rapide Rapide G A.A. sojae PartiallyIEXC purified ExtractSmF crude extract mutant DH56 extract (IEXC) (SmF) A 800

700

600

(mg/L) 500

400 olyphenols

p 300

Total Total 200

100

0 Blank ClairClair RapideRapide GG A. sojae PartiallyIEXC purified ExtractSmF crude extract mutant DH56 extract (IEXC) (SmF) B Figure 7‎ .14 Total polyphenol content; A: Polyphenol concentration in grape musts; B: Polyphenol concentration in wines.

7.4 Conclusions Mass spectrometric characterization of A. sojae proteins led to the identification of a broad enzyme spectrum. However, the present study did not bring the mass spectrometric characterization of PG enzyme among the A. sojae proteins. Oda et al. [26] assumed the resistance of proteins in the gel towards trypsin digestion as possible explanation for low identification rate of proteins by mass fingerprinting. Thus, the use of other peptidases might be helpful for mass spectrometric characterization of PG derived from A. sojae. Nevertheless, various carbohydrate- active enzymes were identified, which are involved in the degradation of different

199 Chapter 7 plant polysaccharides including cellulose, xylan, pectin, starch, etc. The presence of a wide range of different types of enzymes is essential for the degradation of complex plant cell wall polysaccharide molecules. Furthermore, the pectinolytic enzyme profiling revealed the presence of various pectic enzyme activities besides the exo-PG activity in the extract of A. sojae, which are involved in the degradation of pectin – one of the major components of the primary cell wall of plants. The enzymes produced by A. sojae have been studied for their biotechnological potential. Pectinolytic enzymes are utilized in various production processes, e.g. in juice and wine production pectinases have been utilized to improve the yield, decrease the viscosity, clarify the juices and make them more stable [4]. Thus, the pectinolytic enzyme profile derived from A. sojae was compared with pectinase activities of commercial preparations utilized in juice and wine making, which provided an overview on the enzyme spectrums. Moreover, the comparison revealed the low PL activity in the extract of A. sojae, which was detected in a significant higher amount in commercial enzyme preparations used in juice processing. Comparing the different application tests, enzymes of A. sojae achieved only partial satisfying application results. However, the enzyme treatment improved performances comparing to the blank experiments. In comparison to commercial preparations the extracts derived from A. sojae showed lower performances. Commercial enzyme preparations are often mixtures of enzyme sets that derive from various species, such as “Panzym Fino G” which is a preparation used in wine making that contains enzymes derived from A. niger and Trichoderma harzianum. Moreover, “Panzym Pro Clear” is a commercial preparation used for degradation of pectic substances in apple juice production and derives from Aspergillus species in submerged and solid-state culture. The combination of A. sojae extract with an extract derived from another fungal strain or even another species could cover a broad spectrum of enzyme activities, which is beneficial for the degradation of plant material. For instance, Jackisch & Cordes applied for a patent describing the olive oil extraction by a pectinase mixture composed of exo-PG, endo-PG and PME activities in the ratio 1.0 : 2.3 : 0.3 [45]. The extract of A. sojae could be utilized to provide the PG activities. However, this extract has to be mixed with another extract supplying PME. Nevertheless, it has to be recalled that the extract of A. sojae derived from a production process that was optimized for enhanced PG production. The variety of the fungal enzyme set often corresponds to the requirements of its habitat [2]. With regard to the broad pectinolytic enzyme spectrum produced by A. sojae (section 7.3.3‎ ), variation of the cultivation conditions and media design could

200 Chapter 7 increase production levels of the other enzymes, which might yield in a more beneficial enzyme extract for certain applications. Thus, A. sojae ATCC 20235 is particularly useful for the production of enzymes, because of its broad enzyme set and this study showed that the fungal strain is capable of secreting high levels of proteins after optimizing the culture conditions for the specific enzyme production (chapter 3). Moreover, the good fermentation properties allowed easy large-scale production of the desired enzyme (chapter 5) and classical strain improvement further enhanced enzyme titers (chapter 4). Regarding the fact that several products from this species have “generally regarded as safe” (GRAS) status, allowing them to be used in food and feed applications, establishes a wide range of application fields for the exo- and endo-acting enzymes that are involved in the degradation of plant cell wall polysaccharides.

References [1] Peberdy JF. Protein secretion in filamentous fungi - trying to understand a highly productive black box. Trends in Biotechnology 1994; 12:50-57. [2] van den Brink J, de Vries RP. Fungal enzyme sets for plant polysaccharide degradation. Applied Micriobiology and Biotechnology 2011; 91:1477-1492. [3] Jayani RS, Saxena S, Gupta R. Microbial pectinolytic enzymes: A review. Process Biochemistry 2005; 40:2931-2944. [4] Kashyap DR, Vohra PK, Chopra S, Tewari R. Applications of pectinases in the commercial sector: a review. Bioresource Technology 2001; 77:215-227. [5] Benoit I, Coutinho PM, Schols HA, Gerlach JP, Henrissat B, de Vries RP. Degradation of different pectins by fungi: correlations and contrasts between the pectinolytic enzyme sets identified in genomes and the growth on pectins of different origin. BMC Genomics 2012; 13:321. [6] Mann M, Hendrickson RC, Pandey A. Analysis of proteins and proteomes by mass spectrometry. Annual Review of Biochemistry 2001; 70:437-473. [7] Rappsilber J, Moniate M, Nielsen ML, Podtelejnikov AV, Mann M. Experiences and perspectives of MALDI MS and MS/MS in proteomic research. International Journal of Mass Spectrometry 2003; 226:223-237. [8] Laemmli UK. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 1970; 227:680-685. [9] AmershamBiosciencesAB. Protein electrophoresis. Technical manual. 1999. [10] Neuhoff V, Arnold N, Taube D, Ehrhardt W. Improved staining of proteins in polyacrylamide gels including isoelectric focusing gels with clear background at nanogram sensitivity using Coomassie Brillant Blue G-250 and R-250. Electrophoresis 1988; 9:255-262.

201 Chapter 7

[11] Görg A, Klaus A, Lück C, Weiland F, Weiss W. Two-dimensional electrophoresis with immobilized pH gradients for proteomic analysis - A laboratory manual. 2007. [12] GE Healthcare BA. Instructions 28-9537-55 AA: Immobiline DryStrip. 2009. [13] Manchenko GP. Handbook of Detection of Enzymes on Electrophoresis Gels. 1994. [14] Shevchenko A, Tomas H, Havlis J, Olsen JV, Mann M. In-gel digestion for mass spectrometric characterization of proteins and proteomes. NATURE PROTOCOLS 2006; 6:2856-2860. [15] Nakkeeran E, Umesh-Kumar S, Subramanian R. Aspergillus carbonarius polygalacturonases purified by integrated membrane process and affinity precipitation for apple juice production. Bioresource Technology 2011; 102:3293-3297. [16] Mojsov K, Ziberoski J, Bozinovic Z, Petreska M. A comparison of effects of three commercial pectinolytic enzyme preparations in red winemaking. International Journal of Pure and Applied Sciences and Technology 2010; 1(2):127-136. [17] Arauner P. Weine und Säfte, Liköre und Sekt selbstgemacht. 1995. [18] Revilla I, González-San José ML. Addition of pectolytic enzymes: an enological practice which improves the chromaticity and stability of red wines. International Journal of Food Science and Technology 2003; 38:29-36. [19] Diercks-Horn S, Fernandez-Lahore M (2012) Manual of the advanced lab course fermentation technology. Jacobs University Bremen gGmbH [20] Miller GL. Use of dinitrosalicylic acid reagent for determnation of reducing sugar. Analytical Chemistry 1959; 31:426-428. [21] Binner S, Cabrera R, Fernandez-Lahore M. First-year natural science lab biochemical engineering II - 560112. 2009. [22] Kelebek H, Canbas A, Cabaroglu T, Selli S. Improvement of anthocyanin content in the cv. Öküzgözü wines by using pectolytic enzymes. Food Chemistry 2007; 105:334-339. [23] Arnous A, Makris DP, Kefalas P. Correlation of pigment and flavanol content with antioxidant properties in selected aged regional wines from Greece. Journal of Food Composition and Analysis 2002; 15:655-665. [24] Panda T, Naidu GSN, Sinha J. Multipleresponse analysis of pectinolytic enzymes by Aspergillus niger: a statistical view. Process Biochemistry 1999; 35:187-195. [25] Díaz-Godínez G, Soriano-Santos J, Augur C, Viniegra-González G. Exopectinases produced by Aspergillus niger in solid-state and submerged fermentation: a comparative study. J. Ind. Microbiol. Biot. 2001; 26:271-275. [26] Oda K, Kakizono D, Yamada O, Iefuji H, Akita O, Iwashita K. Proteomic analysis of extracellular proteins from Aspergillus oryzae grown under submerged and

202 Chapter 7

solid-state culture conditions. Applied and Environmental Microbiology 2006; 72(5):3448-3457. [27] Zhao G, Hou L, Yao Y, Wang C, Cao X. Comparative proteome analysis of Aspegillus oryzae 3.042 and A. oryzae 100-8 strains: Towards the production of different soy sauce flavors. Journal of Proteomics 2012; 75:3914-3924. [28] de Vries RP. Regulation of Aspergillus genes encoding plant cell wall polysaccharide-degrading enzymes; relevance for industrial production. Applied Micriobiology and Biotechnology 2003; 61:10-20. [29] Rose SH, van Zyl WH. Exploitation of Aspergillus niger for heterologous production of cellulases and hemicellulases. The Open Biotechnology Journal 2008; 2:167-175. [30] van Zyl PJ, Moodley SH, Rose SH, Roth RL, van Zyl WH. Production of the Aspergillus aculeatus endo-1,4-ß-mannanase in A. niger. Journal of Industrial Microbiology & Biotechnology 2009; 36:611-617. [31] Sachslehner A, Foidl G, Foidl N, Gübitz G, Haltrich D. Hydrolysis of isolated coffee mannan and coffee extract by mannanases of Sclerotium rolfsii. Journal of Biotechnology 2000; 80(2):127-134. [32] Norouzian D, Akbarzadeh A, Scharer JM, Young MM. Fungal glucoamylases. Biotechnology Advances 2006; 24(1):80-85. [33] Ramachandran S, Patel AK, Nampoothiri KM, Chandran S, Szakacs G, Soccol CR, Pandey A. Alpha amylase from a fungal culture grown on oil cakes and its properties. Brazilian Archives of Biology and Technology 2004; 47(2):309-317. [34] Tari C, Dogan N, Gogus N. Biochemical and thermal characterization of crude exo-polygalacturonase produced by Aspergillus sojae. Food Chemistry 2008; 111:824-829. [35] Dogan N, Tari C. Characterization of three-phase partitioned exo- polygalacturonase from Aspergillus sojae with unique properties. Biochemical Engineering Journal 2008; 39:43-50. [36] Jacob N, Poorna CA, Prema P. Purification and partial characerization of polygalacturonase from Streptomyces lydicus. Bioresource Technology 2008; 99:6697-6701. [37] Yadav S, Yadav PK, Yadav D, Yadav KDS. Pectin lyase: A review. Process Biochemistry 2009; 44:1-10. [38] Whitaker JR. Pectic substances, pectic enzymes and haze formation in fruit juices. Enzyme and Microbial Technology 1984; 6:341-349. [39] Del Cañizo AN, Hours RA, Miranda MV, Cascone O. Fractionation of fungal pectic enzymes by immobilized metal ion affinity chromatography. J. Sci. Food Agric. 1994; 64:527-531. [40] Capounova D, Drdak M. Comparison of some commercial pectic enzyme preparations applicable in wine technology. Czech Journal of Food Sciences 2002; 20(4):131-134.

203 Chapter 7

[41] Zamora F. Biochemistry of alcoholic fermentation. In: Moreno-Arribas MVPolo MC. Wine chemistry and biochemistry. New York: Springer New York; 2009. 3- 26. [42] Ugliano M. Enzymes in winemaking. In: Moreno-Arribas MVPolo MC. Wine chemistry and biochemistry. New York: Springer New York; 2009. 103-126. [43] Birse MJ (2007) The colour of red wine. The University of Adelaide [44] Pinelo M, Arnous A, Meyer AS. Upgrading of grape skins: Significance of plant cell-wall structural components and extraction techniques for phenol release. Trends in Food Science & Technology 2006; 17:579-590. [45] Jacksch B-O, Cordes A. Enzymatische Behandlung einer Masse aus Oliven oder Olivenbestandteilen. 2005; (patent 10339010).

204 Chapter 8

Chapter 8.

Concluding remarks and future prospects

Pectinolytic enzyme production occupies about 10 % of the worldwide manufacturing of enzyme preparations [1]. Pectinolytic enzymes have been exploited for many industrial applications [2], e.g. the elimination of pectic substances is an essential step in many food processing and beverage industries. Moreover, pectinases have a share of 25 % in the global sales of food enzymes [3]. The largest industrial application of these enzymes is in juice and wine production. Polygalacturonases (PGs) belong to the pectinases and are an inherent part of commercial enzyme preparations used for food processing [4]. The majority of industrial microbial pectinases production derived from filamentous fungi, especially A. niger [5]. Recent articles reported about the potential of A. sojae ATCC 20235 to produce PG enzyme in SmF and via surface cultivation methods [6, 7]. A. sojae has a long history of safe use as koji mold and many of the products are used in the food industry, which favors this fungus as potential production organism for the synthesis of enzymes used in food industry. The microbial screening for pectinolytic enzyme production, which was performed at the beginning of this investigation, revealed the potential of A. sojae ATCC 20235 as pectinase producer under solid-state culture conditions [8]. This strain, which is still deposited as A. sojae at the American Type of Culture collection (ATCC), did not meet the requirements to be classified as A. sojae on the basis on morphological parameters [9], and has been recently reclassified as A. oryzae based on the alpA restriction fragment length polymorphism [10]. However, A. oryzae, similar like A. sojae, belongs to the non-aflatoxin producing species within the section Flavi. Both fungal species have a long history of safe use in the production of oriental fermented food and several Aspergillus-derived food additive products had already obtained a GRAS (generally recognized as safe) status from regulatory authorities [10, 11]. Moreover, Wei & Jong [12] studied A. sojae ATCC 20235 for aflatoxin- producing abilities in rice, peanut and semi-synthetic medium, and published that this strain did not produce detectable levels of aflatoxins. Thus, A. sojae as well as A. oryzae are potential production organisms for enzymes used in the food industries. Besides A. sojae ATCC 20235, two other A. sojae strains, A. sojae CBS 100928 and A. sojae IMI 191303, were also identified as pectinase producers in SSF

205 Chapter 8 during the course of this investigation. However, highest production yields of all explored enzymes were achieved by cultivation of A. sojae ATCC 20235. Hence, A. sojae ATCC 20235 was chosen for optimization studies in this work in order to increase PG enzyme titers. The optimization of process conditions resulted with 909.5 ± 2.7 U/g exo-PG activity in 10.9 time increased enzyme production, which is superior to the highest reported exo-PG activity produced by filamentous fungi [13]. Moreover, utilization of agricultural and agro-industrial by-products as cultivation medium provided the establishment of an economical process for enzyme production. This study presented that the presence of a pectinase inducer is essential for pectinolytic enzyme production. However, the optimized amount of this inducer substrate significantly effects the enzyme production. Besides A. sojae ATCC 20235, one of the newly identified pectinase producing A. sojae strains, A. sojae CBS 100928, was also investigated for optimization studies applying advanced statistical designs, which resulted in the identification of similar process conditions for improved PG production. A. sojae CBS 100928 produced under optimized conditions an exo-PG activity of 131.9 ± 6.9 U/g. However, PG yields obtained by A. sojae ATCC 20235 were considerably higher. Nevertheless, the optimization of process conditions resulted in significantly increased pectinolytic enzyme production by both strains. Simultaneously, an enhancement in PG production was traced by microbial strain improvement. Therefore, a classical mutation and selection strategy was developed in order to generate mutants showing increased exo-PG activity. Mutation was induced by ultraviolet rays employing a 254-nm-wavelength germicidal lamp, and via the chemical agent N-methyl-N’-nitro-N-nitrosoguanidine (NTG). The selection of desired mutants based on a three-step strategy. The first pre-screening focused on morphological parameters regarding the sporulation. It was performed to assure the production of sufficient amount of spores, which could be used as inoculum for SSF cultivations. The second screening based on the selection of “zone mutants”, which enabled the detection of desired mutants with enhanced pectinase activity measured as clear zones on an agar screening medium containing polygalacturonic acid (PGA) sodium salt as substrate. This screening method allowed the identification of mutants with enhanced substrate (PGA) degrading properties. However, it was not sufficient for the differentiation between increased enzyme production in SmF or SSF systems. Barrios-González [14] reported that over- producing strains, generated for SmF, generally do not perform well in SSF. This creates the need to generate over-producing strains particularly suited for the specific cultivation system. Results of the present study confirmed this observation and thus, it was necessary to perform a third screening of the desired “zone

206 Chapter 8 mutants” within a SSF system for the detection of enhanced exo-PG activity under solid-state conditions. In case of A. sojae ATCC 20235, the repeated mutagenesis by UV irradiation resulted in the generation of stable mutants, whereas repeated cycles of NTG treatment generated mutants where the exo-PG activity decreased with each new generation cycle. However, the combination of both methods within sequential mutagenesis created also stable mutants. Nevertheless, highest exo-PG activity of 1087.2 ± 151.9 U/g was achieved by repeated cycles of UV treatment, which created an increase of 72 % after 3 cycles, whereas the combination of the methods generated mutants with slightly higher activity on pectin degradation in SSF. In contrast to A. sojae ATCC 20235 produced A. sojae CBS 100928 stable mutants by repeated cycles of NTG treatment, which resulted with 203.7 ± 17.9 exo-PG activity in an increase of 53 % in comparison to the wild strain. However, exo-PG activity produced by mutant M3, which was generated from A. sojae ATCC 20235 by three cycles UV irradiation, was 5.3 times higher compared to the exo-PG activity produced by the mutant descending from A. sojae CBS 100928 applying repeated NTG treatment. Subjecting A. sojae CBS 100928 to three repeated cycles of UV irradiation generated a mutant which yielded with 19.4 ± 0.1 U/mL in 83 % enhanced exo-PG activity in SmF. Further UV treatment of mutant M3 by two more cycles generated mutant DH56 showing with 98.8 ± 8.7 U/mL a 2.4 fold enhanced PG production in SmF in comparison to the wild type. This mutant was already successfully applied for optimization studies in culture flask experiments yielding in 145 U/mL exo-PG activity, and high enzymes yields were also obtained in the bioreactor [15]. This mutant represents a promising alternative for PG enzyme production in submerged culture. Thus, A. sojae ATCC 20235 and its mutants produced also significant higher levels of pectinolytic enzymes under submerged fermentation conditions in comparison to A. sojae CBS 100928 and its mutants. This observation indicated once more the potential of A. sojae ATCC 20235 and its mutant as promising microorganism for enzyme production. Moreover, comparing the overall enzyme units produced by mutant M3 and mutant DH56 in the respective fermentation system clearly indicated that PG yields obtained in SSF were superior to those achieved under submerged conditions. Namely, mutant DH56 produced 98.8 U/mL in shaking flask culture utilizing 30 mL medium, which amounts to a maximal possible enzyme yield of 2964 U. Mutant M3 produced 1087.2 U/g dry weight of substrate utilizing 10 g medium, which amounts to a maximal possible enzyme yield of 10872 U within the fermentation system. 10 g substrate wetted at 160 % corresponds to 26 g wetted medium. Thus, a comparable amount of medium was

207 Chapter 8 utilized in both fermentation systems. Hence, results of this study favored PG enzyme production by A. sojae under solid-state conditions. Optimization of SSF process conditions for enhanced PG production by mutant M3 resulted in the same optimized conditions in comparison to the wild strain. Changing the substrate pretreatment with regard to SSF scale up studies caused a significant reduce in exo-PG activity obtained by A. sojae ATCC 20235 in culture flask studies. On the contrary, enzyme yield (1009.4 U/g exo-PG activity) obtained by mutant M3 did not seem to be affected by the substrate pretreatment at culture flask level. The PG enzyme production was successfully transferred into a rotating drum type bioreactor at a scaling ratio of 100. Similar exo-PG activity (1194.2 U per g dried solids) was obtained by mutant M3 at bioreactor level in comparison to the culture flask level. Enzyme production by A. sojae ATCC 20235 at bioreactor level, applying forced aeration, increased 1.9 times to 829.7 U per g dried solids in comparison to the culture flask level. Despite several potential advantages of SSF such as higher enzyme production yields, in contrast to SmF which is well established, SSF is still in evolutionary state in terms of engineering aspects and under intensive research [16, 17]. The main obstacles are the low amenability of the process to regulation, the strong heterogeneous fermentation conditions and the often unfeasible biomass determination, which is essential for the kinetic studies [18]. Biomass is determined by measuring indirect indicators, such as glucosamine, ergosterol or protein content, dry weight changes and CO2 evaluation, but all of them have their own weakness. Moreover, even examining the same constituent, various modifications of detection methods exist, which complicates comparison to literature values. Hence, standardization of biomass determination of SSF samples is needed. After selection of a biomass detection method, it will be used to give an estimate of biomass in sample removed over the course of fermentation for the construction of a kinetic profile. However, there is also a great variation in the units that have been used to construct the profiles, e.g. biomass was calculated as “g of biomass per g fresh sample”, “g biomass per g dry sample”, “g biomass per g initial fresh sample” and “g biomass per g initial dry sample”; which also complicates the comparison to reported values [19]. In this study, biomass determination was performed by measuring the glucosamine content. Glucosamine content of fungal biomass in liquid cultures is not applicable to fungal biomass in solid-state culture [20]. It was impossible to separate fungal biomass from solid medium. Thus, the amount of glucosamine was used for representing biomass in this study, without converting it

208 Chapter 8 into mycelia dry weight. Moreover, it was expressed as “mg per g dried solids” to eliminate the impact of the moisture content. Besides the enzyme production focus was also set on efficient enzyme recovery and PG purification. In contrast to SmF, the downstream processing starts with a leaching step to separate the product from the fermented solids in SSF systems. Results of this study indicated that the selection of an appropriate solvent and optimization of extraction conditions could be a simple way of selective product extraction and concentration of the desired enzyme in the crude extract. In terms of PG enzyme leaching, highest specific activity of 207.1 ± 10.6 U/mg in combination with high exo-PG activity was obtained utilizing water as leaching solvent. Since water represented the most cost-efficient solvent, the solvent selection positively affected the process economics. The aim was the maximal PG recovery within one enzyme leaching step. Nevertheless, studies of Castilho et al. [21] indicated the potential of increasing enzyme yields by the application of counter-current extraction strategies for PG enzyme leaching. Thus, applying the optimized leaching conditions in a counter-current extraction might be a promising strategy to improve process economics. PG purification was traced by combination of various chromatographic techniques, which is based on unique combinations of PG characteristics, the so-called chromatographic “fingerprint”. However, as few as possible chromatographic steps should be applied, since these steps significantly increase the downstream processing costs. Most purification systems employ three to five high-resolution chromatographic steps [22]. Apparently purity of PG enzyme by means of a single band on SDS-polyacrylamide gel was achieved after separating A. sojae proteins on the basis of their charge by ion exchange chromatography, followed by separation on the basis of their size and shape (size exclusion chromatography), and finally on the degree of hydrophobicity (hydrophobic interaction chromatography). Hence, three steps were sufficient for the purification of PG enzyme from A. sojae crude extract. Even though the separation of a single band on SDS-polyacrylamide gel was achieved, PG enzyme derived from A. sojae ATCC 20235 could not be mass spectrometric characterized by the method described in chapter 7. Oda et al. [23] assumed the resistance of proteins in the gel towards trypsin digestion as possible explanation for low identification rate of proteins by mass fingerprinting. Thus, the use of other peptidases might be helpful for mass spectrometric characterization of PG derived from A. sojae. However, mass spectrometric characterization of A. sojae proteins led to the identification of a broad spectrum of carbohydrate-active enzymes, which are involved in the degradation of different plant polysaccharides

209 Chapter 8 including cellulose, xylan, pectin, starch, etc. The presence of a wide range of different types of enzymes is essential for the degradation of complex plant cell wall polysaccharide molecules. For instance, A. sojae produced glucoamylase and α- amylase which are involved in the degradation of starch. Rhamnogalacturonan lyase was identified, which cleaves within the hairy regions of pectin, and xylanase was determined which acts on xylan. Moreover, PG as well as various other pectinolytic enzymes, were characterized in the extract of A. sojae ATCC 2023 on the basis of their substrate degrading mode. One has to recall that the commercial available pectinolytic enzyme preparations used in food industry are usually mixtures of pectinolytic enzymes associated with cellulytic, proteolytic and other species of enzymes apart from the main pectinases [24, 25]. Thus, application studies related to fruit juice production and wine making where enzymes mixtures containing PG are extensively used [4], have been performed applying A. sojae crude extracts or only partially purified enzyme extracts. Comparing the different application tests, enzymes treatments utilizing extracts of A. sojae improved performances comparing to the blank experiments. However, in comparison to commercial preparations the extracts derived from A. sojae showed lower performances. Commercial enzyme preparations are often mixtures of enzyme sets that derive from various species, such as “Panzym Fino G” which is a preparation used in wine making that contains enzymes derived from A. niger and Trichoderma harzianum. Moreover, “Panzym Pro Clear” is a commercial preparation used for degradation of pectic substances in apple juice production and derives from Aspergillus species in submerged and solid-state culture. Thus, the combination of A. sojae extract containing high exo-PG and endo-PG activities, as well as sufficient amounts of exo- PMG activity, with an extract derived from another microorganism could cover a broad spectrum of enzyme activities, which might be beneficial certain applications. Moreover, it has to be recalled that the extract of A. sojae derived from a production process that was optimized for enhanced PG production. The variety of the fungal enzyme set often corresponds to the requirements of its habitat [26]. With regard to the broad carbohydrate-active enzyme spectrum produced by A. sojae, variation of the cultivation conditions and media design could increase production levels of the other important enzymes. Thus, A. sojae ATCC 20235 is particularly useful for the production of enzymes in SSF, because of its broad enzyme set and this study showed that the fungal strain is capable of secreting high levels of proteins after optimizing the culture conditions for the specific enzyme production. With regard to the project goals, the work for this thesis provided the successful completion of all milestones related to PG production by A. sojae ATCC 20235 in SSF.

210 Chapter 8

Research highlights . Microbial screening identified A. sojae ATCC 20235 as potential pectinase producer under solid-state culture conditions. Furthermore, two strains of the species A. sojae, A. sojae CBS 100928 and A. sojae IMI 191303, were newly identified as pectinase producers.

. Enzyme production under optimized conditions yielded in 10.9 times increased PG production, and thus in the highest reported exo-PG activity produced by filamentous fungi at laboratory scale [13]. Highest enzyme yield (909.5 ± 2.7 U/g) was obtained by A. sojae ATCC 20235 after 8 days SSF at 30 °C applying 30% sugar beet pulp as inducer substrate in combination with wheat bran as medium wetted at 160% by 0.2 M HCl.

. Generation of mutant M3 by three cycles of repeated UV irradiation yielded in 72 % increased exo-PG activity in SSF. Moreover, mutant DH56 was generated in 5 cycles repeated UV treatment, which produced 145 % increased exo-PG activity in submerged culture.

. SSF process for PG enzyme production was successfully transferred into a rotating drum type bioreactor at a scaling ratio of 100. At bioreactor level an exo-PG activity of 1194.2 U/g dried solids was produced by mutant M3, which was similar to the enzyme yields obtained at culture flask level.

. PG enzyme was successfully recovered applying optimized enzyme leaching conditions. Moreover, combination of IEXC, SEC and HIC resulted in the isolation of PG enzyme as a single band on SDS-polyacrylamide gel in the range of about 40 kDa.

. Mass spectrometric characterization identified a broad spectrum of carbohydrate-active enzymes produced by A. sojae ATCC 20235 under solid- state conditions.

Future prospects The results of this study presented the potential of A. sojae ATCC 20235 as production organism under solid-state conditions. It was shown that the fungal strain is capable of secreting high levels of proteins after optimizing the culture conditions for the specific enzyme production. However, it was also shown that this strain secreted a broad enzyme set. The variety of the fungal enzyme set often corresponds to the requirements of its habitat [26]. Thus, this microorganism might be also a promising candidate for the production of various other enzymes.

211 Chapter 8

Microbial strain improvement by classical mutation and screening procedure resulted in successful enhanced PG production. However, the combination of NTG and UV treatment together with a screening procedure utilizing pectin as screening substrate might result in the improvement of the overall pectinolytic enzyme yields. Besides that, protoplast fusion of mutant M3 or mutant DH56 with another fungal strain showing other beneficial enzyme activities, such as high PL activity, might generate a mutant producing an enzyme preparation in SSF or SmF, that could be directly applied in fruit juice processing. In terms of application testing, the extract derived from A. sojae ATCC 20235 could be also applied in orange juice production. Since, during the production of orange juices pectinolytic enzyme preparations with high levels of polygalacturonase activity have been utilized for the stabilization of clouds [29], this might be a promising application field. Moreover, the combination of A. sojae extracts with enzyme extracts derived from other microorganisms could be applied for tailor-made applications. Nevertheless, the identification of PG enzyme derived from A. sojae ATCC 20235 by mass spectrometric characterization should be pursued as well as the biochemical characterization of the enzyme for the determination of optimal temperature and pH values, as well as enzyme stability and inactivation studies, which might be essential for the formulation. During the course of this investigation it has been figured out that there exists a high interlaboratory variability in pectinolytic enzyme assays, which complicates the comparison of activity values to reported enzyme activities. This high variability might be caused by the multiplicity of pectinases [27] and thus, much variability in the physicochemical properties of pectinases. Gummadi & Panda [28] summarized some biochemical properties of different polygalacturonases, such enzymes had optimal activities at temperatures varying from 35 to 50 °C and pH values between

3.75 and 7.0. They also differ in their molecular weights and Km values. In order to achieve meaningful comparisons, the activity values obtained during this course of investigation was compared to reported enzyme activities that were achieved utilizing the same substrate and basing on the same detection method. However, differences in the “standard assay conditions”, such as different incubation temperatures, pH values, time or reaction volume might still exist. Those who continue the work on polygalacturonase production by A. sojae should pay attention to use the same assay conditions when comparing enzyme activities or at least present comparative values/ transformation factors.

212 Chapter 8

References [1] Semenova MV, Sinitsyna OA, Morozova VV, Fedorova EA, Gusakov AV, Okunev ON, Sokolova LM, Koshelev AV, Bubnova TV, Vinetskii YP, Sinitsyn AP. Use of preparation from fungal pectin lyase in the food industry. Applied Biochemistry and Microbiology 2006; 42(6):598-602. [2] Kashyap DR, Vohra PK, Chopra S, Tewari R. Applications of pectinases in the commercial sector: a review. Bioresource Technology 2001; 77:215-227. [3] Jayani RS, Saxena S, Gupta R. Microbial pectinolytic enzymes: A review. Process Biochemistry 2005; 40:2931-2944. [4] Lang C, Dörnenburg H. Perspectivies in the biological function and the technological application of polygalacturonases. Applied Micriobiology and Biotechnology 2000; 53:366-375. [5] Naidu GSN, Panda T. Production of pectolytic enzymes - a review. Bioprocess Eng. 1998; 19:355-361. [6] Ustok FI, Tari C, Gogus N. Solid-state production of polygalacturonase by Aspergillus sojae ATCC 20235. Journal of Biotechnology 2007; 127:322-334. [7] Tari C, Gögus N, Tokatli F. Optimization of biomass, pellet size and polygalacturonase production by Aspergillus sojae ATCC 20235 using response surface methodology. Enzyme and Microbial Technology 2007; 40:1108-1116. [8] Heerd D, Yegin S, Tari C, Fernandez-Lahore M. Petinase enzyme-complex production by Aspergillus spp in sold-state fermentation: A comparative study. Food and Bioproducts Processing 2012; 90:102-110. [9] Ushijima S, Hayashi K, Murakami H. The current taxonomic status of Aspergillus sojae used in Shoyu fermentation. Agric. Biol. Chem. 1982; 46:2365- 2367. [10] Heerikhuisen M, Van den Hondel C, Punt P. Aspergillus sojae. In: Gellissen G. Production of recombinant proteins. Novel microbial and eukaryotic expression systems. Weinheim: WILEY-VCH; 2005. 191 - 214. [11] Jørgensen TR. Identification and toxigenic potential of the industrially important fungi, Aspergillus oryzae and Aspergillus sojae. Journal of Food Protection 2007; 70(12):2916-2934. [12] Wei D-L, Jong S-C. Production of aflatoxins by strains of Aspergillus flavus group maintained in ATCC. Mycopathologia 1986; 93:19-24. [13] Jacob N. Pectinolytic Enzymes. In: Nigam PSPandey A. Biotechnology for agro- industrial residues utilisation: Springer Netherlands; 2009. 383-396. [14] Barrios-González J. Solid-state fermentation: Physiology of solid medium, its molecular basis and applications. Process Biochemistry 2012; 47:175-185. [15] Buyukkileci AO, Tari C, Fernandez-Lahore M. Enhanced production of exo- polygalacturonase from agro-based products by Aspergillus sojae. BioResources 2011; 6(3):3452-3468.

213 Chapter 8

[16] Prabhakar A, Krishnaiah K, Janaun J, Bono A. An overview of engineering aspects of solid-state fermentation. Malaysian Journal of Microbiology 2005; 1(2):10-16. [17] Singhania RR, Patel AK, Soccol CR, Pandey A. Recent advances in solid-state fermentation. Biochemical Engineering Journal 2009; 44:13-18. [18] Hölker U, Lenz J. Solid-state fermentation - are there any biotechnological advantages? Current Opinion in Microbiology 2005; 8:301-306. [19] Mitchell DA, Krieger N, Berovic M. Solid-state fermentation bioreactors. 2006. [20] Sardjono, Zhu Y, Knol W. Comparison of fermentation profiles between lupine and soybean by Aspergillus oryzae and Aspergillus sojae in solid-state culture systems. J. Agr. Food Chem. 1998; 46:3376-3380. [21] Castilho LR, Alves TLM, Medronho RA. Recovery of pectolytic enzymes produced by solid state culture of Aspergillus niger. Process Biochemistry 1999; 34:181-186. [22] Walsh G, Headon DR. Protein Biotechnology. 1994. [23] Oda K, Kakizono D, Yamada O, Iefuji H, Akita O, Iwashita K. Proteomic analysis of extracellular proteins from Aspergillus oryzae grown under submerged and solid-state culture conditions. Applied and Environmental Microbiology 2006; 72(5):3448-3457. [24] Pedrolli DB, Monteiro AC, Gomes E, Carmona EC. Pectin and pectinases: production, characterization and industrial application of microbial pectinolytic enzymes. The Open Biotechnology Journal 2009;3:9-18. [25] Del Cañizo AN, Hours RA, Miranda MV, Cascone O. Fractionation of fungal pectic enzymes by immobilized metal ion affinity chromatography. J. Sci. Food Agric. 1994; 64:527-531. [26] van den Brink J, de Vries RP. Fungal enzyme sets for plant polysaccharide degradation. Applied Micriobiology and Biotechnology 2011; 91:1477-1492. [27] Stratilová E, Markovic O, Skrovinová D, Rexová-Benková L, Jörnvall H. Pectinase Aspergillus sp.polygalacturonase: Multiplicity, Divergence, and structural patterns linking fungal, bacterial, and plat polygalacturonases. Journal of Protein Chemistry 1993; 12(1):15-22. [28] Gummadi SN, Panda T. Purification and biochemical properties of microbial pecinases - a review. Process Biochemistry 2003; 38:987-996. [29] Whitaker JR. Pectic substances, pectic enzymes and haze formation in fruit juices. Enzyme and Microbial Technology 1984; 6:341-349.

214 Appendix A

Appendix A

General Materials and Methods

This Appendix describes the standard materials and methods used in most experiments presented in this work. Variations of these standard procedures are described in each chapter.

1.1 Analytical methods

1.1.1 Total Protein determination Total soluble protein was measured in the supernatant according to modified Bradford method [1], using Coomassie Plus™ Protein Assay Kit (Pierce, Fischer scientific, Schwerte, Germany) [2]. The assay was performed in a microplate by determining the absorbance at 595 nm using bovine serum albumin (BSA) as a standard. Determinations were performed in duplicate.

1.1.2 Total soluble carbohydrate assay Soluble carbohydrates in supernatants were determined by the phenol-sulfuric acid method [3], using D-glucose as standard. For the assay, 1 mL of 5 % (w/v) phenol solution and 5 mL 96 % H2SO4 were added to 1 mL of an appropriate solution of the sample. Mixtures were incubated at room temperature for 20 min. The absorbance of each sample was spectrophotometrically determined at 490 nm using a Shimatzu spectrophotometer (UV-1700 Pharma Spec) at room temperature.

1.1.3 Polygalacturonase assay Exo-PG activity was assayed according to the procedure provided by Panda et al. [4] with slight modifications. In brief, samples of 0.086 mL containing appropriate diluted PG enzyme were mixed with 0.4 mL of 2.4 g/L polygalacturonic acid solution dissolved in 0.1 M acetate buffer (pH 4.8). This mixture was incubated at 40 °C for 10 min. The reducing sugar released was measured by using the Nelson-Somogyi method [5] as adapted by Panda et al.[4]. First reaction was terminated by adding 0.5 mL copper reagent and placing the mixture in boiling water for 10 min. After cooling down, 1 mL of arsenomolybdate reagent was added, followed by intensive vortexing and centrifugation at 3220 × g at 22 °C for 5 min. The absorbance of the supernatant was read on a Shimatzu spectrophotometer (UV-1700 Pharma Spec) at 500 nm. Blanks were in-cooperated containing all the reagents and the enzyme, but the enzyme was not allowed to react with the substrate. Standard solutions of

215 Appendix A galacturonic acid were used for calibration. One unit of exo-PG activity was defined as the amount of enzyme that catalyzes the release of 1 µmol of galacturonic acid per unit volume of sample per unit time under the standard assay conditions mentioned above.

1.1.4 Enzyme assays for pectinolytic activity profiling Exo-polymethylgalacturonase (exo-PMG) activity was assayed according to the procedure provided by Panda et al. [4] with slight modifications. In brief, samples of 0.086 mL containing appropriate diluted PMG enzyme were mixed with 0.4 mL of 1 % pectin solution dissolved in 0.1 M acetate buffer (pH 4.8). This mixture was incubated at 40 °C for 10 min. The reducing sugar released was measured by using the Nelson-Somogyi method [5] as adapted by Panda et al.[4]. First reaction was terminated by adding 0.4 mL copper reagent and placing the mixture in boiling water for 10 min. After cooling down, 1 mL of arsenomolybdate reagent was added, followed by intensive vortexing and centrifugation at 2000 × g for 5 min. The absorbance of the supernatant was read on a spectrophotometer at 500 nm. Blanks were in-cooperated containing all the reagents and the enzyme, but the enzyme was not allowed to react with the substrate. Standard solutions of galacturonic acid were used for calibration. One unit of exo-PG activity was defined as the amount of enzyme that catalyzes the release of 1 µmol of reducing sugars per unit volume of sample per minute under the standard assay conditions mentioned above. Endo-polygalacturonase (Endo-PG) activity was estimated by measuring the hydrolyzed substrate after the reaction by calculating the remaining polygalacturonic acid as described by Torres et al. [6]. The mixture reaction was prepared by the addition of 90 µL of polygalacturonic acid (0.2 % w/v in 80 mM acetate buffer pH 5.5) to 10 µL of the enzyme. After the incubation of the enzymatic reaction, at 40° C per 10 min, the mixture was filled up to a volume of 3mL with distilled water, then 40 µL of a solution of red ruthenium (4 mg/mL prepared in distilled water) were added and the mixture was homogenized. The mixture was adjusted to a final volume of 6 mL with distilled water, homogenized and centrifuged at 2000 × g for 5 min at room temperature and the absorbance of supernatant was measured at 535 nm using a Shimatzu spectrophotometer (UV- 1700 Pharma Spec). Control mixtures without enzyme and without substrate were also incubated. One unit of enzyme was defined as the amount of enzyme required to hydrolyze 1 mg of polygalacturonic acid per minute under the standard assay conditions. Pectin lyase (PL) assay was carried out by the determination of the unsaturated galacturonides with 2-thiobarbituric acid (2-TBA) according to the methodology described by Nedjma et al. [7]. The reaction mixture contained 250 µL of the crude

216 Appendix A extract and 250 µL of a 1% pectin solution prepared in acetate buffer (pH 4.8). The tubes containing the reaction mixture were incubated in water bath at 30° C for 30 min. After the addition of 50 µL of 1 M NaOH, the mixture was heated at 80° C for 5 min and cooled down in an ice-water bath. Then, 600 µL of 1 M HCl were added to the mixture followed by the addition of 500 µL of 0.04 M 2-TBA. The tubes were incubated again at 80° C for 5 min. The solution was briefly cooled in an ice-water bath, filtered through a 0.45 µm syringe filter and absorbance was measured at 550 nm using 1-cm quartz cells. Blanks were prepared by the addition of acetate buffer instead of the enzyme solution. One unit of enzyme activity was defined as the quantity of enzyme required to increase the absorbance by 0.01. Pectin methylesterase (PME) assay was measured by the determination of the released methanol as described by Klavons & Bennett [8]. The method is based on the oxidation of methanol to aldehyde by the alcohol oxidase, and the subsequent reaction of aldehyde with acetylacetone. The reaction mixture was composed of 10 µL enzyme extract and 90 µL of a pectin solution (0.5% w/v) prepared in 50 mM acetate buffer (pH 4.8). The mixture was incubated at 25° C for 10 min, heated to 100° C for 10 min and the tubes were cooled down in an ice-water bath. Afterwards, 100 µL of alcohol oxidase from Pichia pastoris (1 U/mL) were added and the mixture was incubated at 25° C in a water bath for 15 min. Finally, 200 µL of an acetylacetone/ammonium acetate solution (41 µL of acetylacetone, 3.08 g of ammonium acetate and 49 µL of acetic acid in a final volume of 20 mL) were added to the mixture. The mixture was incubated at 60 °C for 15 min, cooled down and adjusted to a final volume of 1 mL. The absorbance was measured at 412 nm. One unit of enzyme activity was defined as amount of enzyme required to release 1 µmol of methanol per min at 25 °C.

1.2 Chemicals for mutagenesis

1.2.1 PBS buffer – Phosphate buffered saline  137 mM NaCl  2.7 mM KCl

 10 mM Na2HPO4 × 2 H2O

 2 mM KH2PO4  PBS buffer was adjusted to pH 7.4 with diluted HCl

1.2.2 N-methyl-N’-nitro-N-nitrosoguanidine stock solution  0.3 % NTG was dissolved in PBS buffer (pH 7.4)  sterile filtration of NTG stock solution was performed through a 0.2 µm syringe filter

217 Appendix A

 storage of aliquots was done in dark tubes at -20 °C

1.3 Protocol: Tryptic in-gel digestion The following procedure of in-gel digestion is based on the procedure given by Shevchenko et al. [9].

1.3.1 Materials

(1) Washing Solution: 0.1% trifluoroacetic acid (TFA) in 1:1 acetonitrile (ACN) : ddH2O

(2) 100 mM NH4HCO3 (prepare fresh daily)

(3) 1:1 100 mM NH4HCO3 : ACN (4) Neat ACN

(5) Trypsin buffer (13 ng/µL trypsin in 10 mM NH4HCO3 containing 10% (v/v) ACN)

1.3.2 Provisions 1.5 mL tubes, PCR tubes and glass plates were pre-rinsed with solution (1). In order to avoid the release of polymers nitrile gloves were worn at all time. Furthermore, only autoclaved tips have been used to pipette solutions.

1.3.3 Excision of protein bands

Polyacrylamide gel was rinsed with ddH2O for few hours prior to the excision of protein bands. The gel was placed on a cleaned glass plate and the complete band or protein spot of interest was cut out and chopped into cubes of 0.5 – 1 mm³ size. The cubes were transferred into the cleaned 1.5 mL tubes and the content of each tube was spun down.

1.3.4 Destaining of gel pieces 100 µL of solution (3) were added to the gel cubes in each tube and the mixture was incubated for 30 min with occasional vortexing. The supernatant was removed and this procedure was repeated once more in case if there was still much stain present in the gel cubes. Otherwise 500 µL of solution (4) were added into each tube and the mixture was incubated for 10 min at room temperature with occasional vortexing. The content was spun down when the gel cubes shrank into white pellets and supernatant was removed.

1.3.5 Trypsin saturation & digestion

Solution (5) was added into each tube to cover the dry gel pieces completely (~ 20 µL). Samples were kept on ice for 30 min. In order to assure complete coverage of gel pieces by trypsin buffer, additional solution (5) was added into the tubes (max 100 µL), if it was absorbed by the gel cubes. The mixture was incubated

218 Appendix A for 90 min on ice for saturation. About 15 µL of solution (2) (max 50 µL) were added to cover the gel pieces and to keep them wet during the process of trypsin digestion. Trypsin digestion was performed at 37°C over night (min 12 h; max 16 h). Samples were chilled to room temperature and the content of each tube was spun down. The supernatant was withdrawn and concentrated in vacuum centrifuge concentrator 5301 (Eppendorf, Germany) at room temperature. Concentrated samples were re-dissolved in 5 µL of 0.1 % TFA for subsequent MALDI-MS analysis.

References [1] Bradford MM. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Analytical Biochemistry 1976; 72:248-254. [2] Pierce Biotechnology. Coomassie plus protein assay reagent kit. Instructions 2002. [3] Dubois M, Gilles KA, Hamilton JK, Rebers PA, Smith F. Colorimtric method for determination of sugars and related substances. Analytical Chemistry 1955; 28:350-356. [4] Panda T, Naidu GSN, Sinha J. Multipleresponse analysis of pectinolytic enzymes by Aspergillus niger: a statistical view. Process Biochemistry 1999; 35:187-195. [5] Nelson N. A photometric adaption of the somogyi method for the determination of glucose. Journal of Biological Chemistry 1944; 153(2):375-380. [6] Torres S, Sayago JE, Ordoñez RM, Isla MI. A colorimetric method to quantify endo-polygalacturonase activity. Enzyme and Microbial Technology 2011; 48(2):123-128. [7] Nedjma M, Hoffmann N, Belarbi A. Selective and sensitive detection of pectin lyase activity using a colorimetric test: application to the screening of microorganisms possessing pectin lyase activity. Analytical Biochemistry 2001; 291(2):290-296. [8] Klavons J, Bennett RD. Determination of methanol using alcohol oxidase and its application to methyl ester content of pectins. Journal of Agricultural and Food Chemistry 1986; 34(4):597-599. [9] Shevchenko A, Tomas H, Havlis J, Olsen JV, Mann M. In-gel digestion for mass spectrometric characterization of proteins and proteomes. NATURE PROTOCOLS 2006; 6:2856-2860.

219 Appendix B

Appendix B

Bioreactor studies

This Appendix describes the experiments which were conducted to study exo-PG production at laboratory bioreactor level. The effect of various process parameters such as aeration rate and mixing rate as well as medium sterilization on the pectinolytic enzyme production by A. sojae was investigated. The aim of this study was the development of a solid-state process yielding in efficient exo-PG activity.

1.1 Materials and Methods

1.1.1 Materials All chemicals were purchased from AppliChem GmbH (Darmstadt, Germany), except the substrate for detection of exo-PG activity, polygalacturonic acid, and the chemical sodium arsenate dibasic heptahadrate was obtained from Sigma-Aldrich Chemie GmbH (Steinheim, Germany). Microbial substrates like wheat bran and sugar beet pulp pellets were obtained from local suppliers (Bremer Rolandmühle Erling GmbH & Co. KG, Bremen, Germany; Nordzucker AG, Uelzen, Germany).

1.1.2 Microorganism A. sojae ATCC 20235 was purchased from Procochem Inc (Teddington, United Kingdom). Mutant M3, an enhanced producer of PG, was generated by repeated mutagenesis applying UV irradiation. This mutant derived from the wild strain A. sojae ATCC 20235 (chapter 4). Fungal strains were propagated on agar plates according to the specifications given in Heerd et al. [1]. Spores from these plates were used as inoculum for molasses agar slants containing: glycerol (45 g/L), molasses (45 g/L), peptone (18 g/L), NaCl (5 g/L), KCl

(0.5 g/L), FeSO4·7H2O (15 mg/L), KH2PO4 (60 mg/L), MgSO4 (50 mg/L), CuSO4·5H2O

(12 mg/L), MnSO4·H2O (15 mg/L) and agar (20 g/L). High amount of spore production was achieved by cultivation on molasses agar slants at 30 °C for 1 week. Spores suspensions obtained from harvesting molasses agar slants with sterile Tween 80 water (0.02%) were adjusted to a spore concentration of 4×107 spores/mL by counting spores of the suspension in a Thoma counting chamber and diluting with sterile distilled water. 50 mL of this spore suspension were used to inoculate a solid- state culture utilizing 1 kg dry substrate. This amount was chosen in order to keep

220 Appendix B the ratio of previous utilized amounts in flask cultures, applying 2×106 spores per gram dry substrate.

1.1.3 Medium For all cultivation a simple and nutrient rich medium of wheat bran and sugar beet pulp pellets was used. Following findings obtained during media design and optimization studies for enhanced PG production in culture flask level (chapter 3), the optimized medium composition of wheat bran and sugar beet pulp in the ratio 70:30 was also utilized in bioreactor studies. The total amount of dry substrate used in bioreactor studies was 1 kg. However, the pretreatment of the inducer substrate, sugar beet pulp pellets, changed from grinding to soaking in the bioreactor level. Hence, sugar beet pulp pellets soaked in tap water and after swelling sugar beet pulp was mixed with wheat bran and 0.5 M HCl to adjust desired moisture level at the final HCl concentration of 0.2 M prior sterilization. The moisture level of medium, which was separately sterilized in the autoclave, was adjusted to a final moisture content of 160 %. The moisture level of medium that was used for in-situ sterilization had to be optimized (section 1.2.1‎ ). The adjusted moisture level prior sterilization was calculated as dry basis moisture content according to the following equation:

Moisture content (%) = (weightwet – weightdry) / weightdry *100 (B.1) where weightwet is the weight of solid media together with tap water and 0.5 M HCl, and weightdry means weight of solid media (natural moisture content in solid media has not been considered for calculation).

1.1.4 Bioreactor Bioreactor studies for PG production were done using Terrafors-IS, a solid-state, in- situ sterilizable bioreactor (Infors HT, Switzerland). The bioreactor was in-situ sterilized via a double jacket and direct steam injection, controlled via a pulse valve. Hence, the amount of steam fed through the air supply line via a pulse valve can be set from 10 % to 100 %. This valve has a duty cycle of 10 seconds. This means that a setting of e.g. 10% opens the valve for 1 second and then closes it for 9 seconds, and so on. The vessel is a cylindrical stainless steel drum (total volume, 15 L), which could rotate clockwise and anticlockwise between 0.1 rpm up to 10 rpm. Temperature control was achieved with electrical heating and cooling valve. Operating temperature was maintained at 30 °C passing water through the jacket of the fermenter. According to the bioreactor design forced aeration was employed by passing compressed air through a sterile filter (Novasip™ capsules, Pall GmbH,

221 Appendix B

Dreieich, Germany) before entering the vessel via a sparger (pore size, Ø = 1 mm). Aeration was supplied according to the standard aeration procedure used throughout this study (section 1.2.4‎ ) increasing the air flow rate from 2 L/min to 5 L/min at the first day of SSF. Variations of this standard aeration procedure are described in the respective chapters and in the figure legends. Air inlet and outlet were placed at opposite positions in the vessel, supporting better aeration of the SSF process (Figure 1).

B

C A

Figure 1 Vessel of the SSF bioreactor; A: Process air is fed into the sparger and thus through the pores into the vessel [2]; B: The exit gas leaves the vessel via the exit gas pipe [2]; C: Vessel filled with medium illustrating the forced air supply directly into the substrate and placement of air inlet and outlet at opposite position.

An exit gas cooler minimized evaporation losses. During growth, the evaporated water was condensed by an exit gas cooler. The evaporated and condensed water was re-circulated to the air inlet with a peristaltic pump. Levels of oxygen and carbon dioxide in the exit gas stream were detected by an exit gas analyzer. Data logging of on-line parameters was done with Iris V5 control software.

1.1.4.1 Effect of in-situ sterilization on exo-PG production The effect of in-situ sterilization on exo-PG production was studied at an intermittent mixed process applying an agitation rate of 10 rpm for 10 min clockwise

222 Appendix B followed by 10 min mixing anticlockwise twice at the day of inoculation and at the first day of cultivation. The following cultivation period was performed under static conditions. Aeration was supplied by the standard aeration procedure.

1.1.4.2 Comparison of exo-PG production in laboratory bioreactor and in culture flasks Comparison experiments at bioreactor level were performed utilizing separately sterilized medium (section 1.1.3‎ ) at an intermittent mixed process applying an agitation rate of 10 rpm for 10 min clockwise followed by 10 min mixing anticlockwise twice at the day of inoculation and at the first day of cultivation, as well as in a continuous mixed process with 0.1 rpm rotation rate. Aeration was supplied by the standard aeration procedure. Cultivation at culture flask level was done as described in section 1.1.5‎ .

1.1.4.3 Effect of intermittent and continuous mixing on PG production The effect of mixing rate was studied in intermittent mixed SSF processes applying two times mixing at the day of inoculation and at the first day of cultivation, and also in continuous mixed systems. Aeration was supplied by the standard aeration procedure. Agitation experiments were performed utilizing separately sterilized substrate (section 1.1.3‎ ).

1.1.4.4 Effect of aeration rate on PG production The effect of aeration rate was studied in intermittent mixed processes applying an agitation rate of 1 rpm for 10 min clockwise followed by 10 min mixing anticlockwise, twice at the day of inoculation and at the first day of cultivation, as well as in continuous mixed processes with 0.1 rpm rotation rate. Aeration was supplied by the standard aeration procedure and also using a constant flow rate. Aeration experiments were performed utilizing separately sterilized substrate (section 1.1.3‎ ).

1.1.4.5 PG production in SSF with intermittent agitation at repeated fed-batch mode Cultivation was started as a batch culture utilizing separately sterilized medium (section 1.1.3‎ ) in an intermittent mixed process applying an agitation rate of 1 rpm for 10 min clockwise followed by 10 min mixing anticlockwise, twice at the day of inoculation and at the first day of cultivation. Aeration was supplied by the standard aeration procedure with a shift in the aeration rate from 2 L/min to 5 L/min at the first day of SSF. After 6 days SSF the biomass was harvested and 5 % (130 g) of biomass was used to inoculate 2.6 kg separately sterilized moist medium. This was repeated twice. After each addition, the vessel contents were mixed by 10 min rotation clockwise followed by 10 min mixing anticlockwise at 1 rpm, which was

223 Appendix B repeated after 10 h at the inoculation day. After the first shift from 2 L/min to 5 L/min the aeration rate was kept constant at 5 L/min over the whole SSF process.

1.1.4.6 Comparison of PG production by A. sojae ATCC 20235 and its mutant M3 at laboratory bioreactor level Comparison experiments of exo-PG production at bioreactor level were performed in an in-situ sterilized process applying media wetted at 100 % (section 1.2.1‎ ), as well as in a solid-state process utilizing separately sterilized medium (section 1.1.3‎ ). In both processes intermittent mixing at a rotation rate of 1 rpm for 10 min clockwise followed by 10 min mixing anticlockwise, twice at the day of inoculation and at the first day of cultivation was used. Aeration was supplied by the standard aeration procedure with a shift in the aeration rate from 2 L/min to 5 L/min at the first day of SSF.

1.1.4.7 Sampling At 24 h intervals samples of approximately 100 g were taken and divided into aliquotes. Some aliquots of biomass were used for enzyme leaching. Enzyme leaching was performed utilizing 80 mL distilled water for 20 g biomass, in order to keep the ratio 1:4 of biomass to eluent (chapter 6). The mixture was incubated for 20 min at 250 rpm at 25 °C. Supernatant was separated from biomass by centrifugation for 30 min at 3220 × g at 4 °C. Exo-PG activity (section 1.1.6‎ ) and soluble protein content (section 1.1.7‎ ) were determined in the supernatant.

1.1.5 Solid-state fermentation in culture flasks and enzyme leaching SSF was performed in 250-mL Erlenmeyer flasks containing 26 g moist medium, which was filled under aseptic conditions into sterilized flasks after autoclaving the medium used for bioreactor studies. The medium was prepared as described in section 1.1.3‎ . The amount was chosen to perform the cultivation under similar conditions like previous optimized SSF process utilizing 10 g dry medium wetted at 160 % with 0.2 M HCl (chapter 3). Flasks were inoculated with the total amount of 2 × 107 spores, corresponding to approximately 770,000 spores per gram moist medium. Incubation was done at 30 °C with manually shaking twice at the day of inoculation and the first cultivation day. Cultivation was performed for periods of 3 to 6 and 6 to 8 days with harvesting of flasks at 24 h intervals during this time. At the end of cultivation, enzyme leaching was done by adding 80 mL water into the flasks and incubating for 20 min at 250 rpm agitation and 25 °C. Enzyme activity (section 1.1.6‎ ) and total protein content (section 1.1.7‎ ) were determined in the supernatant obtained after centrifugation at 4 °C, 3220 × g, for 20 min.

224 Appendix B

1.1.6 Exo-polygalacturonase assay Exo-PG activity was assayed according to the procedure of Panda et al. [3], which was further optimized as described in Appendix A. Exo-PG activity values were calculated per g biomass.

1.1.7 Protein content Soluble protein content was determined according to the modified Bradford’s method [4] as described in Appendix A. Soluble protein content was expressed as mg per g biomass.

1.2 Results and Discussion Bioreactor studies targeted the scale-up of the previously optimized SSF process for exo-PG production by A. sojae. The aim was to develop a SSF process for pectinolytic enzyme production at laboratory bioreactor level and demonstrate the promising potential of A. sojae for PG production in scale-up studies.

1.2.1 Medium sterilization The sterilization of medium on larger scale poses problems such as temperature profiles, physicochemical alterations in the medium and thermal degradation of nutrients due to the effect of scale factors. The heating time to attain and hold the desired temperature of sterilization and the post-sterilization cooling to incubation temperature are scale dependent and pose problems in batch sterilization processes [5]. In-situ sterilization was performed via a double jacket and direct steam injection. Hence, the setting of the pulse valve influences the amount of water, in the form of condensed steam, entering the vessel during sterilization (section 1.1.4‎ ). Thus, it is critical for optimizing the moisture content of the substrate. In order to keep the additional water entry as small as possible, the pulse valve was set at 10 % during sterilization. This setting resulted in the accumulation of approximately 100 mL water in the empty vessel. Hence, two experimental approaches were tested for in-situ sterilization of the medium. On the one hand, the medium was prepared as described in section 1.1.3‎ and on the other hand, 100 mL less water was used for soaking of sugar beet pulp pellets during medium preparation. In both cases was the medium sterilized in the vessel with the pulse valve set at 10 % and a sterilization temperature set at 121 °C for 15 min holding time. Rotation speed of the drum was set at 5 rpm during sterilization. Prior sterilization the medium consisted of loose particles. After sterilization physic-chemical changes in the medium of both experimental approaches were observed resulting in the formation of a heterogeneous mixture of sticky clumps (Figure 2). The nature of the

225 Appendix B in-situ sterilized substrate was not comparable with the one obtained by sterilization in the autoclave.

A B C

Figure 2 Effect of in-situ sterilization on the medium texture; A: Medium before in-situ sterilization; B: Optimized medium as described in section 1.1.3‎ after sterilization; C: Medium containing 100 mL less water after sterilization.

The changes after in-situ sterilization in the texture of the substrate, such as compact mass formation and adhesion of the sticky clumps to the walls of the fermenter should not be neglected since it directly affects important factors, such as interparticle space containing the gaseous phase and thus influencing fungal growth. In order to prevent this compact mass formation during in-situ sterilization the effect of reduced moisture content was investigated. Hence, the moisture level of the fermentation medium was decreased to 100 % obtained at a final HCl concentration of 0.2 M. This composition was used for in-situ sterilization performed under the conditions as described above. The resulting medium texture is presented in Figure 3. The decreased moisture level in the medium composition yielded in the formation of almost uniform small pellets. This texture seemed to be suitable with regard to the oxygen transfer and thus, was utilized for PG production by A. sojae ATCC 20235 in SSF.

226 Appendix B

Figure 3 Medium wetted at 100 % obtained after in-situ sterilization.

1.2.1.1 Effect of in-situ sterilization on exo-PG production A comparative study on PG production was performed utilizing in-situ sterilized medium, which was wetted at 100 % prior sterilization and medium, prepared as described in section 1.1.3‎ , which was separately sterilized in the autoclave. Both cultivations were done in duplicate, while other process parameters were kept constant (Figure 4). Hence, aeration was provided by the standard aeration procedure with a change from 2 L/min to 5 L/min at the first day of cultivation. Taking the total bioreactor volume of 15 L into account the aeration rates defined according to Mitchell et al. [6] (vvm, m³-air m-3-total-bioreator-volume min-1) are changing from 0.13 vvm to 0.33 vvm at the first day of cultivation. However, considering the amount of used substrate the applied aeration rate utilizing 1 kg dry substrate wetted at 100 % (2 kg moist mass) for in-situ sterilization corresponded to 1 L/min/kg and 2.5 L/min/kg, respectively. Utilizing the separately sterilized medium wetted at 160 % (2.6 kg moist mass) would yield in aeration rates of 0.77 L/min/kg and 1.92 L/min/kg, respectively. Hence, the aerations rates of the in-situ sterilized substrate would be higher due to the lower moisture content.

227 Appendix B

in-situ In -in-situsitu sterilized medium; SSF 23.06.2011 In in-situ-situ sterilized medium; SSF 08.11.2011 in-situ 300 300 Separately P5 sterilized medium; SSF 03.05.2011 P5 P5 Separately P5 sterilized medium; SSF 14.07.2011 250 250

200 200

150 150

100 100

exo-PG activity (U/g) exo-PGactivity exo-PG activity (U/g) exo-PGactivity 50 50

0 0 0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7 Time (d) Time (d)

Figure 4 Effect of in-situ medium sterilization on PG production by A. sojae ATCC 20235: ( & ) PG production in two cultivations utilizing in-situ sterilized medium wetted at 100 %; ( & ) PG production in two fermentations using medium sterilized in the autoclave.

Following the PG production profiles presented in Figure 4 similar results were obtained applying in-situ sterilized medium wetted at 100 % prior sterilization in comparison to medium wetted at 160 % prior to sterilization in the autoclave. Hence, the decrease in moisture content and the formation of almost uniform small pellets during in-situ sterilization (Figure 3) did not affect the exo-PG production by A. sojae ATCC 20235. This allows the performance of the enzyme production in SSF at bioreactor level under sterile conditions without losses in product yields.

1.2.2 Comparison of exo-PG production in laboratory bioreactor and in culture flasks The comparison of exo-PG production in culture flasks and at laboratory bioreactor level was performed utilizing the same substrate in both systems for a better comparison. Hence, separately sterilized substrate was used, which was filled in sterile flasks and in the in-situ sterilized vessel of the bioreactor after autoclaving. Both systems were inoculated with approximately 770,000 spores per gram moist medium and incubated at 30°C. In the first experimental approach SSF at bioreactor level was intermittent mixed twice at the day of inoculation and the first day of cultivation similar to the manual shaking of the flask culture during this time. Both systems were kept under static conditions for the following cultivation period. Total soluble protein content and enzyme yields obtained during this SSF are presented in Figure 5.

228 Appendix B

3,033,0.0 exo-PG exo-PG acivity acivity in in bioreactor bioreactor 250250 exo-PG exo-PG acivity acivity in in flasks flasks Protein Protein content content in in bioreactor bioreactor 2,52.52,5 Protein Protein content content in in flasks flasks 200200 2,02.02,0

150150 1,51.51,5

100100

1,01.01,0

exo-PG activity (U/g) exo-PGactivity

exo-PG activity (U/g) exo-PGactivity

Protein content (mg/g) Protein content (mg/g) 5050 0,50.50,5

00 0,00.00,0 00 11 22 33 44 55 66 77 88 TimeTime (d) (d)

Figure 5 SSF with intermittent mixing in bioreactor in comparison to flask cultivation utilizing the same medium; (——) exo-PG yield in bioreactor; (——) Protein content in bioreactor; () exo- PG activity in culture flasks; () Protein content in culture flasks.

Previous optimization studies of PG production at culture flask level (chapter 3) yielded in maximum enzyme activity at the 8th day of SSF. Hence, the maximum PG production of 223.8 ± 18.8 U/g obtained after 7.7 days of SSF could be considered as maximal possible exo-PG activity. Comparing the highest enzyme activities achieved in culture flasks and at bioreactor level, the enzyme yield at bioreactor level of 248.0 U/g is higher. Furthermore it was produced after 6 days SSF and thus, resulted also in a higher productivity of 41.3 U/g/d. At the same day the productivity in flask culture corresponded to 26.3 U/g/d, which slightly increased to 29.1 U/g/d after 7.7 days of SSF. Highest specific activity of 231.5 U/mg was obtained after 4 days SSF at bioreactor level, which decreased to 163.1 U/mg after 6 days. In flask culture a similar specific activity of 169.8 ± 5.9 U/mg was obtained after 6 days, which slightly decreased to 165.4 ± 6.8 U/mg after 7.7 days SSF. Similar specific activities at the 6th day of SSF and higher PG yield at bioreactor level promoted enzyme production in the bioreactor. Since it was already demonstrated that a better PG production was obtained at bioreactor level using similar agitation conditions (intermittent mixing) another comparative study was investigated utilizing continuous mixing at a rotation rate of 0.1 rpm at bioreactor level. Since this was not performable in flask cultivation the flask culture was incubated as described aforementioned with harvesting from day 3 to day 5.6 of SSF. Total soluble protein content and PG yields obtained during this SSF are presented in Figure 6.

229 Appendix B

3,03.0 exo-PG exo-PG activity activity inin bioreactorbioreactor 250250 exo-PG exo-PG activity activity inin flasksflasks Protein Protein content content inin bioreactorbioreactor 2,52.5 Protein Protein content content inin flasksflasks 200200 2,02.0

150150 1,51.5

100100

1,01.0

exo-PG activity (U/g) exo-PGactivity

exo-PG activity (U/g) exo-PGactivity

Protein content (mg/g) Protein content (mg/g) 5050 0,50.5

00 0,00.0 00 11 22 33 4 5 6 TimeTime (d)(d)

Figure 6 SSF in bioreactor with continuous mixing in comparison to flask cultivation with intermittent mixing utilizing the same medium; (——) exo-PG yield in bioreactor; (——) protein content in bioreactor; () exo-PG activity in culture flasks; () protein content in culture flasks.

Highest exo-PG activity of 232.1 U/g was achieved at bioreactor level after 3.9 days SSF. This corresponded to a productivity of 59.5 U/g/d. At the same day a specific activity of 284.6 U/mg was obtained. The maximum PG activity was slightly below the maximum enzyme activity produced in the intermittent mixed process, but it was achieved in a shorter cultivation time. In comparison to the flask cultivation the exo-PG activity was 1.7 times higher at bioreactor level after 3.9 days SSF. Thus, exo- PG production by A. sojae in SSF seemed to be promoted by the use of a mixing system.

1.2.3 Effect of intermittent and continuous mixing on PG production Agitation in fermentation processes ensures homogeneity by distribution, e.g. uniform distribution of nutrients, and thus promotes effective distribution of inoculum and growth on individual substrate particles, prevents aggregation and counter the gradients formation, supports gas transfer and heat exchange. However, agitation also has adverse effects such as disruption of fungal attachment on solid particles, damage of fungal mycelia and compacting of substrate particles coming along with decreasing the substrate porosity [5].

1.2.3.1 Effect of mixing rate in SSF with intermittent mixing In accordance to flask cultures, which were mixed by manually shaking at the day of inoculation and during the first day of SSF, PG production at bioreactor level was also mixed by agitation during these periods of fermentation. According to Lonsane

230 Appendix B

et al. [5] scale-up of agitation systems in SSF should not pose problems due to the 350350 quite low agitation ratesEquation used in most processes. Nevertheless, the influence of Adj. R-Square

300 mix 300 mix 350 rotation rate on exo-EquationPG production at intermittent mixed SSF processes was mix mix Adj.mix R-Square 250250 studied (Figure 7). 300 mix mix mix mix mix Fit Curve 11 200200 250 SSF 03.05.2011 03.05.2011 SSFmix 14.07.2011 14.07.2011 150150 350 Fit Curve 1 Equation 200 Fit curve: exo-PG = - 14.206 + 235.582 * exp(-0.5 * ((t – 5.931)/2.282)²); R²03.05.2011 = 0.876 Adj. R-Square 100 14.07.2011 mix

100 300 mix exo-PG activity (U/g) exo-PGactivity exo-PG activity (U/g) exo-PGactivity 150 mix mix 5050 250 100 exo-PG activity (U/g) exo-PGactivity mix Fit Curve 1 00 200 50 03.05.2011 0 1 2 3 4 5 6 7 8 0 1 2 3 4 5 6 7 8 14.07.2011 150 0 TimeTime (d) (d) 0 1 2 3 4 5 6 7 8 100 Time (d) (U/g) exo-PGactivity 50

0 0 1 2 3 4 5 6 7 8

mix Time (d) A FitFit CurveCurve 11 350350350 SSF 01.11.201101.11.2011 01.11.2011 SSF 22.11.201122.11.2011 22.11.2011 300300300 SSFmix 16.01.201216.01.2012 16.01.2012 350 Equation Fit Curve 1 Fit curve: exo-PG = 293.31 + (-3.514 – 293.31) / (1 + (t / 3.096)^2.653); Equation Adj. R-Square 350 Adj. R-SquareR-Square 01.11.2011 Adj. R-Square 250250 R² = 0.883 250 mix mix 22.11.2011 mix 300 mix mix mix mix 300 mix 16.01.2012 mix mix 200200200 Equationmix mix 250 Adj. R-Square 250 150 mix 150150 mix 200 mix 200 mix 100

100100

exo-PG activity (U/g) exo-PGactivity exo-PG activity (U/g) exo-PGactivity exo-PG activity (U/g) exo-PGactivity 150 150 50 5050 100

100 exo-PG activity (U/g) activity exo-PG

exo-PG activity (U/g) exo-PGactivity 0 0 0 0 1 2 3 4 5 6 7 8 50 50 0 0 1 1 2 2 3 3 44 55 66 77 88 Time (d) Time (d) 0 0 Time (d) 0 1 2 3 4 5 6 7 8 0 1 2 3 4 5 6 7 8 Time (d) Time (d)B Figure 7 PG production under solid-state conditions with intermittent mixing twice at day of inoculation and twice at the first day of cultivation; A: Mixing at 10 rpm for 10 min clockwise and for 10 min anticlockwise during the mixing periods; B: Mixing at 1 rpm for 10 min clockwise and for 10 min anticlockwise during the mixing periods.

PG production profile presented in Figure 7-A was obtained at maximum agitation rate of 10 rpm. Hence the drum was rotating approximately 200 times during the mixing period. Enzyme production shown in Figure 7-B was performed at a mixing rate of 1 rpm and thus, the number of rotations during the same mixing period was ten times less. Both profiles show a similar trend until the 6 day of SSF. Further cultivation of biomass, which was mixed intermittent at a rotation rate of 10 rpm during the initial fermentation state yielded in a decrease in exo-PG activity.

231 Appendix B

Whereas, cultivation of the SSF process applying lower agitation rate further increased in enzyme activity. Even if the enzyme production seemed to be not directly affected by the rotation rate first, because of revealing similar profiles at the period of intermittent mixing, higher shear forces may have adverse effects of fungal attachment to the solids and damage to fungal mycelia. With regard to process economics the process at lower agitation rate seemed to be preferable due to lower energy introduced into the system.

1.2.3.2 Effect of mixing rate in SSF with continuous mixing Lonsane et al. [5] suggested intermittent rather than continuous agitation to prevent damage to mycelia. However, continuous rotation in SSF might counter concentration gradients within the system. Moreover, it is relatively simple to add solutions uniform to the bed in continuous mixed systems, e.g. water to prevent drying, which is not practical in a packed bed. In the present study the effect of a continuous mixed system on exo-PG production by A. sojae ATCC 20235 was investigated under solid-state conditions. Previous studies demonstrated the formation of smaller and more uniform substrate pellets at lower rotation rates (data not shown), further discussion of this topic is given below (Figure 9). Hence, maximum agitation rate was set at 1 rpm. The minimal agitation rate provided by the bioreactor design of 0.1 rpm and also a mixing rate of 0.3 rpm continuous drum rotation were applied for PG enzyme production under solid-state conditions. 0.1 rpm 350 0.3 rpm 1.00.1 0.1 rpm rpm rpm mixing rate 0.1 rpm Fit 0.3 curve:Curve rpm exo 1 -PG = 283.439 + (-3.018 – 283.439)/(1 + (t/2.725)^3.262); 350 0.3 rpm 350300 1.0 rpm FitR² 0.1 = Curve 0.975 rpm 2 1.00.1 Fit rpm Curve 1 300 Fit0.3 0.3 rpmCurve rpm mixing 3 rate 250350 Fit0.3 Fit Curve rpm Curve 1 2 300350 350 1.0 rpm 1.0Fit Fit0.1 curve:Curve rpm Curve rpm exo 2 -3PG = 241.771 + (0.043 – 241.771)/(1 + (t/3.691)^4.478); FitR² Fit= Curve 0.999 Curve 3 1 250300350 Fit 0.3 Curve rpm 1 250300200 300 Fit Curve 2 Fit1.0 1.0 rpmCurve rpm mixing 2 rate Fit Curve 3 Fit Fit curve:Curve Curve exo 3 -1PG = 48.650 + (-0.809 – 48.650)/(1 + (t/3.503)^2.6); 200300250 200250150 250 R² Fit= 0.957 Curve 2 Fit Curve 3 150200250

150200100 200 exo-PG activity (U/g) exo-PGactivity 100

exo-PG activity (U/g) exo-PGactivity 200150

10015050 150 exo-PG activity (U/g) exo-PGactivity 50 100100150

exo-PG activity (U/g) exo-PGactivity 50 0 100

exo-PG activity (U/g) exo-PGactivity exo-PG activity (U/g) exo-PGactivity 0 0 1 2 3 4 5 6 5010050

exo-PG activity (U/g) exo-PGactivity 50 0 0 1 2 Time3 (d) 4 5 6 0 1 2 3 4 5 6 Time (d) 5000 0 Time (d) 00 1 1 2 2 3 3 4 4 5 5 6 6 0 1 2 3 4 5 6 0 TimeTime (d) (d) Time (d) 0 1 2 3 4 5 6 Time (d) Figure 8 Effect of PG production under solid-state conditions with continuous mixing at different mixing rates: () 0.1 rpm, () 0.3 rpm and () 1.0 rpm.

232 Appendix B

PG production profiles obtained at several agitation rates are presented in Figure 8. Rotation rate in the continuous mixed system significantly influenced enzyme production. Maximum exo-PG production was obtained at minimum agitation rate of 0.1 rpm. An increase in mixing rate resulted in decreased enzyme yield. Furthermore, with increased rotation rate also an increase in pellet size formed by the rotary motion was noticed (Figure 9).

A B C

Figure 9 Fermented mass after 1 day SSF in continuous mixed system; A: 0.1 rpm rotation rate; B: 0.3 rpm rotation rate; C: 1.0 rpm rotation rate.

The rate of rotation consequently influenced the medium texture and thus the whole SSF process by affecting the surface area. At the lowest rotation rate of 0.1 rpm more uniform and smaller pellets were obtained and thus providing a bigger surface area for fungal growth. The bigger surface area promotes also other important factors, such as aeration and heat removal. Low rotation rate minimized the adverse effect of compacting the substrate particles and might also have minimized the damage of fungal mycelia in the continuous mixed SSF. Dhillon et al. [7] noticed similar findings utilizing apple pomace as substrate for citric acid production in a continuous mixed process at 2 rpm , which caused the formation of clumps (1 – 2 cm balls). Thus, the process for citric acid production was changed to an intermittent mixed process, which yielded also in a higher product concentration. Comparing exo-PG production achieved at 0.1 rpm rotation rate in the continuous mixed process with the intermittent mixed processes (section 1.2.3.1‎ ) similar exo-PG activity values were obtained. The continuous mixing at low rotation rate (0.1 rpm) did not significantly affect the process productivity of exo-PG production by A. sojae ATCC 20235 in comparison to the intermittent mixed process at 1 rpm rotation rate. With regard to process economics the process with intermittent mixing at the beginning of SSF followed by static cultivation conditions seemed to be preferable due to lower energy introduced into the system.

233 Appendix B

1.2.4 Effect of aeration rate on PG production Aeration provides on one hand oxygen for metabolic processes and growth and on the other hand it removes carbon dioxide and heat from the culture. Many process parameters and medium characteristics affect the oxygen transfer rates, such as flow rate, porosity of moist particles and the bed depth. The requirement of aeration at laboratory culture flask scale is met by discrete circulation of air and optional manually mixing, while in large scale processes forced aeration is employed. Generally, forced aeration is advantageous as it achieves higher efficiency by combining aeration with agitation [5]. However, besides removal of heat and carbon dioxide also evaporated water is removed by forced aeration. Thus, aeration rate influences the moisture level in SSF systems and promotes the formation of a moisture gradient within the substrate bed. Evaporation losses were minimized in the present study by using an exit gas cooler as standard. Condensed water was re-circulated into the sparger for aeration presented in Figure 1-A. In the present study the aeration rate was changed from 2 L/min (0.13 vvm; 0.77 L/min/kg) to 5 L/min (0.33 vvm; 1.92 L/min/kg) at the first day of cultivation due to the increased oxygen demand during the first and second day of SSF. Oxygen and carbon dioxide concentration were measured in the exhaust gas analyzer and their trend in batch processes is presented in chapter 5 in Figure 5.4-A and Figure 5.5-A as blue and violet line for oxygen and carbon dioxide concentration, respectively. Oxygen and carbon dioxide concentrations in exhaust gas of a repeated fed-batch process are presented as blue and violet line in Figure 11. The aeration procedure was used to minimize evaporation losses using a lower aeration rate at the initial phase, the so-called lag-phase and to provide sufficient amount of oxygen during the growth phase using an increased aeration rate, which also promotes heat removal generated by metabolic processes. The effect of a change from this standard aeration procedure to a constant aeration rate over the whole fermentation time was studied with regard to PG production yield (Figure 10). The PG production profiles presented in Figure 10-A were obtained in SSF with intermittent mixing during the day of inoculation and at the first day of cultivation. At the beginning a similar trend in exo-PG was observed, while over longer incubation times a significant lower enzyme yield was produced after 5 to 6 days SSF by applying the constant flow rate of 5 L/min. At that time both processes were aerated with the same flow rate of 5 L/min and thus, the difference in enzyme production had to originate from the initial aeration period, which might be arisen by the loss of substrate moisture with high aeration rate and thus leaded to drying of solids.

234 Appendix B 2-5 L/min 5 L/min Aeration 2-5 L/min rate: 2 L/min  5 L/min 250 Aeration5 L/min rate: 5 L/min 350

300 200

250 150

200 100

150 exo-PG activity (U/g) exo-PGactivity 50

100 exo-PG activity (U/g) exo-PGactivity

50 0 0 1 2 3 4 5 6 0 Time (d) 0 1 2 3 4 5 6 TimeA (d) 2-5 L/min 3 L/min Aeration 2-5 L/min rate: 2 L/min  5 L/min 250 Aeration 3 L/min rate: 3 L/min 250 200 200 150 150 100

100 exo-PG activity (U/g) exo-PGactivity

50 exo-PG activity (U/g) exo-PGactivity 50 0 0 0 1 2 3 4 5 6 0 1 2 3 4 5 6 Time (d) TimeB (d)

Figure 10 Effect of aeration rate on PG production; A: Intermittent mixed SSF process () with change in aeration rate from 2 L/min to 5 L/min at the first cultivation day and () continuous aeration rate at 5 L/min; B: Continuous mixed SSF process at 0.1 rpm () with change in aeration rate from 2 L/min to 5 L/min at the first cultivation day and () continuous aeration rate at 3 L/min.

Furthermore, it has to be noted, that the operation of an intermittent mixed process which remained static for longer periods as aforementioned might require an aeration system, which aerates the bed evenly during the static period. The actual bioreactor design in which the air line passed through the bed in a readily curved metal rod, which had several small holes in the horizontal section that passed through the center of the bed (Figure 1) did not promote an evenly aeration and caused gradients of moisture level distribution within the forcefully-aerated bed. With regard to the continuous mixed process, the uniform distribution of the air might not be so crucial. The application of a constant flow rate was also tested in a continuous mixed process (Figure 10-B). Continuous agitation promotes oxygen supply and heat removal, hence a lower constant flow rate of 3 L/min (0.2 vvm; 1.15 L/min/kg) was tested to minimize evaporation losses. Comparing the aeration methods at a continuous mixed process, similar exo-PG activities were obtained. Nevertheless,

235 Appendix B higher enzyme yields were achieved applying the standard aeration procedure. Hence, this procedure including a shift from 2 L/min to 5 L/min at the first day of cultivation was fixed for all following cultivations in the bioreactor.

1.2.5 PG production in SSF with intermittent agitation at repeated fed-batch mode Batch processes are most commonly used in SSF. Nevertheless, also fed-batch, repeated fed-batch and continuous modes of bioreactor operation have been utilized in SSF processes [8]. In contrast to SmF processes the addition of fresh, uninoculated solid substrate particles requires interparticle colonization, which is a relative slow process. Thus, fed-batch and continuous operation modes require mixing devices to promote the distribution process. Furthermore, in many SSF processes the operation after medium sterilization is not performed under aseptic conditions, which is increasing the risk of contamination. Employed strategies of minimizing this risk are such as using a high ratio of inoculum, low moisture content and pH level of the substrate [5]. PG production by A. sojae was also studied in repeated fed-batch mode to explore the effect of inoculum and enzyme production over longer time under not aseptic conditions. For this purpose the vessel was harvested after 6 days SSF and 5 % of the fermented solids were used to inoculate 2.6 kg of fresh substrate particles, which were sterilized in the autoclave before mixing with the fermented mass. This procedure was repeated once more and obtained exo-PG and protein yields are presented in Figure 11. Inoculation applying a spore suspension of 2×109 total spores (approximately 770,000 spores per gram moist medium) resulted in a lag phase of circa 0.5 days before a decrease of oxygen or increase of carbon dioxide concentration in the exhaust gas was observed (Figure 11). This time was reduced by the use of fermented mass in the repeated fed-batch process. Thus, the productivity in exo-PG production obtained one day after inoculation was increased from 2.83 U/g/d using spore suspension as inoculum to 31.23 U/g/d in the first repeated and 17.64 U/g/d in the second repeated batch applying 5 % fermented mass as inoculum. Explanation of this change over time might be given by the use of a non-dormant inoculum, while a dormant form of inoculum such as the fungal spore suspension used at the beginning would introduce a lag phase due the time required for germination.

236 Appendix B PG B 300 4 C Protein D 250 E 3 200

150 2

100

PG Protein exo-PG activity (U/g) exo-PGactivity

PG B C Protein D E

1 Protein content (mg/g)

50

Protein content (mg/g) content Protein 0 0 (mg/g) content Protein 4 3 2 1 0

0 2 4 6 8 10 12 14 16 18 4 3 2 1 0

exo-PG activity (U/g) activity exo-PG exo-PG activity (U/g) activity exo-PG 0

Time (d) 0 50

50 150 100 250 150 200 100 300 250 200 300 18 18 Figure 11 Repeated fed-batch process of intermittent mixed SSF: (——) exo-PG production; (——) protein content. 16 16

Usually, a high inoculum ratio is applied to produce the desired product level of a 14 14 secondary metabolite in a short period and to prevent contamination. In SSF the use 12 of spore inoculum is preferred due to the ease of uniform mixing of spores with 12

moist solid substrate. The formation of spores requires 6 – 7 days in the case of 10 10 fungal cultures. At large scale processes more subculturings might be needed for 8 Time (d) Time 8 the inoculum preparation, which may lead to culture degenerations. The spore (d) Time 6

formation during inoculum development and their germination can increase the 6 mutation frequency. Thus, minimum number of subculturing is essential in the 4 inoculum development [5]. 4 2

In the present study the SSF processes was initiated with the inoculation of a spore 2 suspension similar like in culture flask level. At this laboratory bioreactor level the 0 use of the inoculum type was just applicable, while at a larger scale the inoculum 0 had to be provided by a subculturing step. The PG production in a solid-state fed- batch process seemed to be preferable in economizing the operation of inoculum preparation. Moreover, it was demonstrated that the adaption phase could be reduced. In the present study 5 % of fermented solids were used as inoculum in the repeated fed-batch process, which is quite low. The increase of the inoculum ratio might even further enhance the productivity. Abdullah et al. [9] compared protein production by Chaetomium cellulolyticum in batch, fed-batch and repeated fed- batch culture on wheat straw. 20 % higher protein production was achieved in repeated fed-batch culture where one half of the fermented material was removed

237 Appendix B at three day intervals and replaced by fresh substrate. In this mode protein productivity was maintained for 12 days at a steady state. A similar trend in exo-PG production was observed in the repeated fed-batch process. The trend of constant enzyme production is illustrated in Figure 12. PG PG1 700 PG2 PG3 600 B C 500

400

300

200 exo-PG activity (U/g) exo-PGactivity

100

0 0 2 4 6 8 10 12 14 16 18 Time (d)

Figure 12 Bioprocess kinetics of the exo-PG production by A. sojae ATCC 20235 in a repeated fed- batch process. Time courses of (——) exo-PG activity and additive exo-PG activity (——) are shown. Arrows () indicate the replacing of fermented solids with new substrate keeping 5 % of biomass as inoculum.

Graphical presentation of the additive exo-PG activity is showing an increasing enzyme activity over the time course. The additive exo-PG activity rose to a final value of 682.4 U/g after about 17.3 days. This corresponds to a productivity of 39.4 U/g/d. Nevertheless, it has to be also noted that the productivity decreased from batch to batch within the repeated fed-batch process. During the first batch a productivity of 48.4 U/g/d was obtained, which decreased to 35.6 U/g/d in the second and to 33.2 U/g/d in the third batch. The fermentation operation beyond autoclaving of the medium was not under aseptic conditions in the fed-batch process. Thus, with each opening of the vessel, e.g. to harvest biomass and replace it with new substrate particles, there was always the risk of contamination. After harvesting the third batch, the vessel was filled again with new substrate, but this time 20 % of produced biomass was used as inoculum to enhance productivity (data not shown). This cultivation had to be aborted due to a contamination, which significant affected fungal growth and enzyme production. This contamination might have been already present during the second or third batch, which could explain the lower PG production. With the use of a bigger inoculum size and thus, if a contamination was already present, with more contaminated material, the slower

238 Appendix B

growing fungal culture was repressed after adding new substrate. However, this showed the need for aseptic culture conditions in repeated fed-batch or continuous processes.

1.2.6 Comparison of PG production by A. sojae ATCC 20235 and its mutant M3 at laboratory bioreactor level In order to explore the exo-PG production of a previous improved strain (chapter 4) at bioreactor level, a comparative study on pectinolytic enzyme production was performed using mutant M3 descending from A. sojae ATCC 20235 and the wild strain. In the first experimental approach in-situ sterilized medium was utilized, which was wetted at 100 % prior sterilization. The profiles of PG production are presented in Figure 13.

in-situ 250 Mutant in-situ M3 in-situ 250 A. in-situ sojae ATCC 20235

200 200

150 150

100

100 exo-PG activity (U/g) exo-PG activity (U/g) 50 50

0 0 0 1 2 3 4 0 1 2 3 4 Time (d) Time (d) Figure 13 PG production in SSF utilizing in-situ sterilized medium wetted at 100%; () exo-PG activity produced by mutant M3; () exo-PG activity produced by A. sojae ATCC 20235.

Mutant M3 showed an enhanced PG enzyme production starting at the second cultivation day at bioreactor level, which resulted in a higher productivity. The exo- PG productivity was nearly twice as fast with 62.1 U/g/d over about 3.6 days in comparison to the wild strain, which produced 34.3 U/g/d over about 4.2 days. This result supported previous findings (chapter 4) indicating the potential of this strain as production organism for industrial applications. Furthermore, it demonstrated the use of mutant M3 for exo-PG production in an in-situ sterilized process, which might be relevant with regard to large scale applications. Nevertheless, to achieve a better comparison to culture flask experiments a comparative enzyme production at bioreactor level was also performed utilizing

239 Appendix B

separately sterilized medium. While exploring PG yield obtained by mutant M3 in the bioreactor it was observed, that an increase of the setpoint entry to 50 (%) positively affect the enzyme production (data not shown). The peristaltic pump “Feed1” is used for feeding condensate back into the vessel during fermentation. For automated return of the condensate the parameter “Feed1” should have a setpoint entry of 10 (%) [2]. However, it was discovered that a higher setpoint entry enhanced PG production (data not shown). The setpoint entry of 50 (%) resulted in an effective process value of about 33 (%). Further increase of the setpoint did not rise the effective process value. Therefore, a comparative study between A. sojae ATCC 20235 and its mutant M3 was performed fixing the setpoint entry at 50 (%) of the parameter “Feed1” M3 in the SSF process for PG enzyme production by both wild type 600 strains. Profiles of exo M3-FitPG Curve production 1 are presented in Figure 14. wildFit Curve type 1 M3 600 M3FitM3 Curve 1 wild type Mutant FitM3 Curve M3; 1 SSF 18.09.2012 500 600 wild type Fit Curve 1 Mutant FitM3 Curve M3; 1 SSF 31.07.2012 600 Equation Fit Curve 1 500 Fit FitM3 curve: Curve exo 1-PG = 439.524 + (-9.438 – 439.524)/(1 + (t/4.349)^2.997); Adj. R-Square M3 400 R² = 0.915 Equation M3 M3 500 MIX M3 M3 MIX M3 500 Adj. R-Square A.wildM3 wild sojae type type ATCC 20235 400 MIX M3 MIX M3 MIX M3 FitFit Fit curve: Curve Curve exo 1 1-PG = - 11.563 + 282.649 * exp(-0.5 * ((t – 9.026)/3.451)²); Equation 600300600 Equation MIX M3R² Fit = 0.956 Curve 1 Equation Fit Curve 1 Adj. R-Square 400 Adj.MIX R-Square M3 400 MIX M3 M3 300 Adj. R-Square M3 MIX M3 MIX M3 500 Equation M3 500200 MIX28.08.2012 M3 M3 MIX M3 MIX28.08.2012 M3 MIX M3 Adj. R-Square MIX28.08.2012 M3 MIX M3 exo-PG activity (U/g) exo-PGactivity 300 EquationEquation 300 28.08.2012 200 28.08.2012 Adj.EquationAdj. R-Square R-Square 400100400 28.08.2012 Equation 28.08.2012 exo-PG activity (U/g) exo-PGactivity MIXAdj.MIX M3R-Square M3 28.08.2012 MIXMIX M3 M3 Adj. R-Square 200 MIX M3 100 MIX28.08.2012 M3 200 MIX28.08.2012MIX M3 M3 28.08.2012 300300 28.08.2012 exo-PG activity (U/g) exo-PGactivity 0 28.08.2012

Equation28.08.2012Equation 28.08.2012 exo-PG activity (U/g) exo-PGactivity 0 1 2 3 4 5 6 7 8 9 28.08.2012 1000 Adj.Adj. R-Square R-Square 200200 100 28.08.2012 Time (d) 28.08.2012 0 1 2 3 4 5 6 7 8 9 28.08.201228.08.2012 28.08.2012

exo-PG activity (U/g) exo-PGactivity 28.08.2012 exo-PG activity (U/g) exo-PGactivity 0 Time (d) 28.08.201228.08.2012 100100 0 1 2 3 4 5 6 7 8 9 0 Time (d) 0 1 2 3 4 5 6 7 8 9 0 0 00 1 1 2 2 33 44 55 66 77 88 99 Time (d) TimeTime (d) (d) Figure 14 PG production under solid-state condition in an intermittent mixed process and a water recirculation rate of “Feed1” set at 50 %, utilizing a separately sterilized medium with a moisture content of 160 % prior autoclaving; ( & ) mutant M3; () A. sojae ATCC 20235.

Comparing the exo-PG activities, mutant M3 produced a higher enzyme yield than the wild strain. Similar to the results obtained utilizing in-situ sterilized medium (aforementioned), a higher productivity was also achieved by mutant M3 using the separately sterilized medium wetted at 160 %. A deepening comparison in PG enzyme production between A. sojae ATCC 20235 and its mutant M3 at bioreactor level is given in chapter 5. Furthermore, the comparison to previous culture flask experiments is also presented in chapter 5.

240 Appendix B

References [1] Heerd, D, Yegin, S, Tari, C, Fernandez-Lahore, M (2012). Petinase enzyme- complex production by Aspergillus spp in sold-state fermentation: A comparative study. Food and Bioproducts Processing 90, 102-110. [2] InforsHT (2011). Operating manual: Terrafors-IS in situ sterilisable solid state bioreactor: Bottmingen, Swiss. [3] Panda, T, Naidu, GSN, Sinha, J (1999). Multipleresponse analysis of pectinolytic enzymes by Aspergillus niger: a statistical view. Process Biochemistry 35, 187-195. [4] Bradford, MM (1976). A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Analytical Biochemistry 72, 248-254. [5] Lonsane, BK, Saucedo-Castaneda, G, Raimbault, M, Roussos, S, Viniegra- González, G, Ghildyal, NP, Ramakrishna, M, Krishnaiah, MM (1992). Scale-up strategies for solid state fermentation systems. Process Biochemistry 27, 259-273. [6] Mitchell, DA, Krieger, N, Berovic, M (2006). Solid-state fermentation bioreactors. Springer-Verlag Berlin Heidelberg. [7] Dhillon, GS, Brar, SK, Kaur, S, Verma, M (2013). Bioproduction and extraction optimization of citric acid from Aspergillus niger by rotating drum type solid-state bioreactor. Industrial Crops and Products 41, 78-84. [8] Scheper, A (2000). New products and new areas of bioprocess engineering. Springer-Verlag: Heidelberg. [9] Abdullah, AL, Tengerdy, RP, Murphy, VG (1985). Optimization of solid substrate fermentation of wheat straw. Biotechnology and Bioengineering 27, 20-27.

241 Appendix C

Appendix C

Utilization of NW-Q fiber for PG purification

This Appendix describes the experiments which were conducted to explore the potential of non-woven fibers functionalized with quaternary ammonium ligands (NW-Q fiber) for purification of A. sojae PG enzyme by ion-exchange chromatography (IEXC). The effect of colored material, which was an impurity present in the crude extract, on the chromatographic process and an approach to overcome the problem of this impurity was investigated.

1.1 Materials and Methods

1.1.1 Materials All chemicals were purchased from AppliChem GmbH (Darmstadt, Germany), except the substrate for detection of exo-PG activity, polygalacturonic acid, and the chemical sodium arsenate dibasic heptahydrate was obtained from Sigma-Aldrich Chemie GmbH (Steinheim, Germany). Furthermore, the two anion exchange resins Amberlite® IRA-96 and Amberlite® IRA-67, as well as the cation exchange resin Amberlite® IRC-50 were also purchased from Sigma-Aldrich Chemie GmbH (Steinheim, Germany). The fibrous material, NW-Q fibers, was a proprietary of Jacobs University Bremen gGmbH and was provided by Poondi Rajesh Gavara [1]. Fibers were made of pure cotton (TN & Platex S.A., Argentina) with a density of 1.3 g/cm³. The strong anionic fibers, NW-Q fibers, were synthesized by cationic surface functionalization with quaternary ammonium ligands (Q).

1.1.2 PG extract A mixture of crude extracts obtained by several cultivations of A. sojae ATCC 20235 at laboratory bioreactor level (chapter 5) and leaching out of PG enzyme from fermented solids as described in chapter 6, was pooled together and centrifuged at 15,000×g for 40 min at 4 °C for removal of solids. The supernatant containing PG enzyme was subjected to ultrafiltration (UF) and diafiltration (DF) utilizing hollow fiber cartridge (UFP-10-C-6A, GE Healthcare, Life Sciences) to concentrate the sample and transfer the enzyme in 0.02 M acetate buffer (pH 5.0) as described in chapter 6.

242 Appendix C

1.1.3 PG purification with NW-Q fiber Experiments were performed on ÄKTA FPLC system (GE Healthcare) applying a volumetric flow rate of 2 mL/min. PG extract of 10 mL in acetate buffer (0.02 M, pH 5.0) was loaded to the column containing the NW-Q fibers as solid phase. Therefore, 1 g of fiber was packed in a Pharmacia C 10/10 column (GE Healthcare) taking a column volume of about 5 mL. Washing was performed by running 20 column volumes (cV) of the loading buffer (0.02 M acetate buffer, pH 5.0). This was followed by elution using a linear NaCl gradient (0.0 – 0.5 M) in 40 column volumes. Samples were collected in 2.5 mL fractions and analyzed for exo-PG recovery (section 1.1.6‎ ) and protein content (section 1.1.7‎ ).

1.1.4 Ion exchange resins for color removal Two anion exchange resins, Amberlite® IRA-96 and Amberlite® IRA-67, and a cation exchange resin, Amberlite® IRC-50, were utilized for the removal of colored matter from the PG extract. Therefore, Amberlite® ion exchange resins were twice washed with distilled water and twice incubated in 0.1 M acetic acid, each for 1 h. This was followed by two times washing the resin in 0.5 M acetate buffer (pH 5.0), each for 1 h. The resins were separated from the washing buffer by vacuum filtration and subsequently contacted with PG extract for color reduction. For this purpose were 5 g Amberlite® ion exchange resin mixed with 20 mL of PG extract and incubated for 2.5 h at 4 °C on a rotary incubator. The supernatant was collected by vacuum filtration and the resins subjected to elution with 10 mL of 0.5 M NaCl in the first step and 10 mL of 1.0 M NaCl in the second step.

1.1.5 Desalting and buffer exchange of crude extract Desalting of crude extract and buffer exchange was performed, utilizing pre-packed PD-10 desalting columns (Amersham Biosciences AB) containing Sephadex G-25 medium, according to the instructions given by the provider [2].

1.1.6 Exo-polygalacturonase assay Exo-PG activity was assayed according to the procedure of Panda et al. [3], which was further optimized as described in Appendix A.

1.1.7 Protein content Soluble protein content was determined according to the modified Bradford’s method [4] as described in Appendix A.

243 Appendix C

1.1.8 Estimation of color in the extract Absorbance at 350 nm where the contribution of protein/enzyme was minimal was used for comparative purpose for the measurement of colored matter in the extract [5].

1.2 Results and Discussion Purification studies targeted the demonstration of PG enzyme purification by NW-Q fiber in IEXC. The aim was the development of an alternative chromatographic approach utilizing NW-Q fibers as solid phase instead of commercial beads for protein capture and thus, demonstrating the application of these fibers for pectinase purification from SSF crude extracts.

1.2.1 Purification of PG extract with NW-Q fiber UF and DF of SSF extract obtained by cultivation of A. sojae on wheat bran and sugar beet pulp did not remove contaminating colored compounds from the crude extract and thus, a concentration of PG extract by UF resulted also in an increase of these colored impurities. The color of concentrated PG extract is presented in Figure 2. Nevertheless, the increase of dark brown color did not significantly increase the viscosity and the sample was applied to commercial DEAE-Sepharose Fast Flow resins without noticing an increase in pressure during the chromatographic process (chapter 6). Hence, the PG extract obtained after UF and DF was also applied to NW-Q fibers for protein separation under the same conditions applied utilizing commercial beads. The load of the colorful sample resulted in the adsorption of the colored impurities to the fibers at the column inlet and caused a blocking of the flow through the column. This led to an increase of the pressure within the chromatographic system and compressed the NW-Q fiber as presented in Figure 1-B. Hence, application of NW-Q fiber for purification of proteins from fermentation crude extracts had to be coupled with sample pretreatment in order to remove colorful compounds before loading to the NW-Q fiber.

244 Appendix C

A B

Figure 1 IEXC using NW-Q fibers as solid phase in chromatographic column; A: NW-Q fiber after equilibration with 0.02 M acetate buffer (pH 5.0); B: NW-Q fiber after sample load and washing step.

1.2.2 Pretreatment of PG extract with Amberlite® ion exchange resins The pretreatment of PG extract was performed in order to remove colored impurities from the enzyme extract. Previous experiments applying activated charcoal for removal of these compounds resulted in insufficient removal (data not shown) and thus, Amberlite® resins were explored as an alternative approach for the removal of contaminating colored material from the concentrated PG extract. Therefore, two anion exchange resins, Amberlite® IRA-67 and Amberlite® IRA-96, as well as a cation exchange resin, Amberlite® IRC-50, were tested for reduction of the brownish color in the PG extract (Figure 2). Contacting 5 g resins with 20 mL PG extract leaded to a significant reduction of brownish color in the extracts which were incubated with the anion exchange resins for 1 h (Figure 2). While the use of the cation exchange resin did not significantly contribute to the reduction of colored matter in the extract. This was also confirmed by the absorbance values measured in the supernatant at 350 nm (Table 1).

245 Appendix C

AMBERLITE® PG extract IRA - 96 IRA - 67 IRC - 50

Figure 2 Effect of Amberlite® resins on the removal of colored impurities in PG extract after 1 h incubation at 6 °C on a rotary incubator.

Table 1 Treatment of PG extract with several Amberlite® resins.

Protein content exo-PG activity Absorbance Sample pH mg/mL %* U/mL %* 350 nm PG extract 2.79 100 287.3 100 3.7 4.9

IRA-96 0.26 9 0.4 0 0.3 7.0

IRA-67 1.04 37 1.09 0 0.9 8.6

IRC-50 2.24 80 191.2 67 3.1 4.4

Elution with 0.5 M NaCl

IRA-96 1.37 25 1.5 0 1.8 8.6

IRA-67 0.29 5 0.2 0 2.8 9.8

IRC-50 0.33 6 40.1 7 0.7 3.5

Elution with 1.0 M NaCl

IRA-96 0.45 8.0 0.6 0 1.6 8.8

IRA-67 0.10 1.7 0.1 0 1.7 9.8

IRC-50 0.05 0.9 7.7 1 0.2 3.5

* calculated from total amount of mg protein and units exo-PG activity

However, besides removal of colored matter also a reduction in protein content as well as in exo-PG activity was observed in the extracts incubated with the anion exchange resins. Thus, a two step elution using sodium chloride solution was performed. As it can be seen from absorbance values at 350 nm in Table 1 the elution resulted in a release of the colored compounds from the anion exchange resins, but except for the cation exchange resin there was no enzyme activity

246 Appendix C detected in the elution fractions, which pointed in the direction of an inactivation of PG enzyme. A possible explanation was found in the pH values of the extracts that were contacted with the anion exchange resins and which increased the pH of the sample in 0.02 M acetate buffer from 4.9 to 7.0 and 8.6. Previous experiments transferring PG enzyme from 0.1 M acetate buffer (pH 5.0) in 0.02 M phosphate buffer (pH 7.5) and 0.05 M tris-HCl buffer (pH 7.4) leaded to 52 and 49 % reduction in enzyme activity, respectively (data not shown). The buffer exchange was performed utilizing PD-10 desalting column, Sephadex G-25 (Amersham Biosciences). The desalting of the sample with distilled water using the Sephadex G-25 column did not reduce exo-PG, which indicated an enzyme inactivation at higher pH values. This corresponded to findings of Tari et al. [6], they noticed about 60 % lost of exo-PG activity at pH 8.0 exploring PG enzyme stability produced by A. sojae in SmF. The washing step of the resins was extended prior incubation based on the observation of an increase of pH value resulting from the contact of sample with the anion exchange resins, following the procedure described in section 1.1.4‎ . This led to the removal of color matter in PG extract without significant loss in enzyme activity (Table 2). The incubation of PG extract with pretreated Amberlite® IRA-96 and Amberlite® IRC-50 caused a negligible increase of pH value to pH 5.1 and 5.2, respectively, while the incubation with pretreated Amberlite® IRA-67 resin increased the pH value from 4.9 to pH 6.2. Nevertheless, highest exo-PG activity was measured in the extract incubated with this resin. The total amount of PG activity in the extract incubated with Amberlite® IRA-67 resin was reduced by 15 % through a binding to the resin, which was recovered in the elution fractions. However, besides PG enzyme also other proteins were adsorbed and thus, the specific activity in the extract contacted with this resin increased from 106.4 to 143.7 U/mg, which represented a 1.4 fold purification by this pretreatment. Moreover, PG activity in the extract incubated with the cation exchange resin Amberlite® IRC-50 was also relatively high, but similar to previous experiments the incubation did not lead to a significant removal of contaminating colored impurities from the PG extract (Figure 2). Hence, the cation exchange resin seemed not to be suitable for the reduction of colored matter in the extract. Utilization of the anion exchange resin Amberlite® IRA-96 yielded in efficient removal of colored matter (Figure 2), but resulted in increased binding of the target protein to the resin. Thus, PG activity largely decreased in the extract. This would favor the use of the resin for capturing the enzyme, but therefore the elution had to

247 Appendix C be improved and on the other hand besides the enzyme also the contaminating colored impurities were co-eluted (Table 1). Promising results of pretreatment with ion exchange resin for the removal of colored impurities from PG extract was obtained utilizing Amberlite® IRA-67. The efficiency of this pretreatment step could be optimized by varying several factors such as contact time and resin to extract ratio. Nevertheless, these experiments proved the application of the resin as an approach for the removal of colored matter.

Table 2 Treatment of PG extract with pretreated Amberlite® resins.

Protein content exo-PG activity Sample mg/mL %* U/mL %* PG extract 2.77 100 294.8 100

IRA-96 0.39 14 14.5 5

IRA-67 1.75 63 251.5 85

IRC-50 2.04 74 236.2 80

Elution with 0.5 M NaCl

IRA-96 1.90 34 321.8 55

IRA-67 0.51 9 61.5 10

IRC-50 0.35 6 47.8 8

Elution with 1.0 M NaCl

IRA-96 0.85 15 99.6 17

IRA-67 0.19 4 19.1 3

IRC-50 0.09 2 11.3 2

* calculated from total amount of mg protein and units exo-PG activity

1.2.3 Purification of pretreated PG extract with NW-Q fiber Pretreated PG extract, which was incubated with the anion exchange resin Amberlite® IRA-67 as described above for the removal of colored matter, was loaded to functionalized NW-Q fiber for protein purification by IEXC. Loading of pretreated PG extract did not cause a blocking or compressing of fibers. In fact, the IEXC process was performed under the same conditions as utilizing commercial DEAE-Sepharose beads (chapter 6) without appearance of technical difficulties (Figure 3).

248 Appendix C

Figure 3 IEXC using NW-Q fibers as solid phase in chromatographic column showing the NW-Q fibers during the washing step after the sample load.

The staining of fibers at the column inlet presented in Figure 3 was caused by the adsorption of remaining colored matter in the extract, but it did not create difficulties in the performance of the chromatographic process. Moreover, the staining was completely removed by washing the fibers with 1 M sodium chloride solution for several column volumes after performing the run of ion-exchange chromatography. The elution profile for 21.5 mg loaded proteins of pretreated PG extract is presented in Figure 4. A partially separation of PG enzyme from other proteins was achieved collecting the peak of approximately 65 % of relative enzyme activity with a specific activity of 244.5 U/mg in the fractions between 150 and 180 mL of elution volume. However, a part of enzyme activity did not bind to the fibers and was lost during the washing step using 0.02 M acetate buffer at pH 5.0. Hence, the purification of PG enzyme utilizing NW-Q fibers as solid phase in IEXC might be improved varying process conditions such as buffer system and pH value. Furthermore, the binding capacity of PG enzyme has to be determined and hence, the maximum loading.

249 Appendix C

2011 10 11 NWQ001:1_UV 2011 10 11 NWQ001:1_Cond 2011 10 11 NWQ001:1_Cond% 2011 10 11 NWQ001:1_Conc 2011 10 11 NWQ001:1_Fractions 2011 10 11 NWQ001:1_Inject 2011 10 11 NWQ001:1_Logbook 0,350.35 80 20 mAU

600

18 70

0,300.30

20 18 16 14 12 10 8 6 4 2 0

16 500

60 20 18 16 14 12 10 8 6 4 2 0 20 18 16 14 12 10 8 6 4 2 0

0.250,25

80 70 60 50 40 30 20 10 0 14 350 80 70 60 50 40 30 20 10 0

80 70 60 50 400 40 30 20 10 0 50

12 350 0.200,20 350

40 10 300

300

300

0,150.15 300

E PG PG PG activity (U/mL) activity PG

-

8 (%) activity elative

30 R E PG PG

250 exo

E PG PG Protein content (mg/mL) Proteincontent 200

6 250 0,100.10 250 20

4 200 100

0,050.05 200

200 10 2 150 0 0.000,00 A1 A2 A3 A4 A5 A6 A7 A8Waste A12 B2 B4 B6 B8 B10 C1 C3 C5 C7 C9 C11 D1 D3 D5 D7 D9 D11 E1 E3 E5 E7 E9 E11 F1 F3 F5 F7 F9 F11 G1 G3 G5 G7 G9 G11 H1 H3 H5 Waste 0 0 0 50 100 150 200 250 300 ml 150 0 150 50 100 150 200 250 300 350 Elution volume (mL)

100 100

Figure 4 Purification100 of PG enzyme from pretreated extract by anion-exchange chromatography on   | NW-Q fibers: ( ) exo-PG activity; ( ) relative activity; ( ) protein content; (—) absorbance at 280 nm; 50 (—) elution by a linear NaCl gradient (0.0 – 0.5 M in 40 cV and washing with 1 M NaCl at the end); (—) 50 conductivity. 50 0

The present study was conducted to explore the potential of NW-Q fibers for PG 0 0 0,15 0,10 0,25 0,35 0,20 0,30 0,05 enzyme purification produced by A. sojae in SSF. In contrast to the studies with NW- 0,00 0,15 0,10 0,25 0,35 0,20 0,30 0,05 Q fibers, the purification of A. sojae PG enzyme by IEXC was also performed utilizing 0,00 0,15 0,10 0,25 0,35 0,20 0,30 0,05 0,00 commercial DEAE-Sepharose Fast Flow resin (chapter 6). Both processes were performed under the same conditions, e.g. 0.02 M sodium acetate buffer (pH 5.0), elution by linear NaCl gradient (0.0 – 0.5 M) at a volumetric flow rate of 2 mL/min, etc. Therefore, 1 g of NW-Q fibers was packed in a Pharmacia C 10/10 column taking a column volume of about 5 mL, whereas the same column volume was filled with commercial beads in the process utilizing commercial resin. Table 3 presents a summary of IEXC performance comparing fiber vs. commercial beads.

250 Appendix C

Table 3 Comparison of PG purification by IEXC utilizing NW-Q fiber and commercial DEAE-Sepharose resin.

PG Total Specific Purification Recovery Matrix Sample activity protein activity fold (%) (U) (mg) (U/mg)

Pretreated PG extract 1751.1 21.5 81.4 NW-Q 3 65 fiber IEXC fraction 1134.0 4.6 244.5

PG extract 2835.6 43.0 65.9 DEAE- 6.2 68 Sepharose IEXC fraction 1941.3 4.7 410.1

It has to be recalled that the PG extract loaded to the fiber was pretreated by Amberlite® ion exchange resin for the removal of colored impurities. Thus, specific activity of the pretreated PG extract is higher in comparison to the untreated PG extract loaded to commercial resin, since the treatment resulted also in the removal of proteins from the extract (Table 2). This might also explain the difference in the purification fold. However, the achieved recovery yields are similar. As stated above, the binding capacity and maximum loading still needs to be determined. Nevertheless, capturing 65 % of loaded enzyme activity seemed to be a promising start point for further studies. The invention of a cartridge containing the fiber with several meshes for partition might even enhance the process performance applying untreated PG extract containing a higher amount of colored matter or reducing the process time by increasing the volumetric flow rate.

References [1] Gavara, PR (2012). Fibrous adsorbents as novel chromatography matrices for enhancing industrial downstream processing of bioproducts. Jacobs University Bremen gGmbH, Bremen. [2] Amersham Biosciences AB (2003). Instructions: PD-10 desalting column. Sweden. [3] Panda, T, Naidu, GSN, Sinha, J (1999). Multipleresponse analysis of pectinolytic enzymes by Aspergillus niger: a statistical view. Process Biochemistry 35, 187-195. [4] Bradford, MM (1976). A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Analytical Biochemistry 72, 248-254.

251 Appendix C

[5] Singh, SA, Ramakrishna, M, Appu Rao, AG (1999). Optimisation of downstream processing parameters for the recovery of pectinase from the fermented bran of Aspergillus carbonarius. Process Biochemistry 35, 411-417. [6] Tari, C, Dogan, N, Gogus, N (2008). Biochemical and thermal characterization of crude exo-polygalacturonase produced by Aspergillus sojae. Food Chemistry 111, 824-829.

252