1

Immobilization of LccC laccase from Aspergillus nidulans on hard surfaces using fungal 1 hydrophobins 2 3 Oleksandra Fokina#, Alex Fenchel, Lex Winandy, Reinhard Fischer# 4 5 Institute for Applied Biosciences - Department of Microbiology, Karlsruhe Institute of 6 Technology (KIT), Karlsruhe, Germany 7 8 Running title: Laccase immobilization on surfaces using hydrophobins 9 10 #Address correspondence to Oleksandra Fokina, oleksandra.fokina@kit.edu, or Reinhard 11 Fischer, reinhard.fischer@kit.edu. 12 13 14

AEM Accepted Manuscript Posted Online 26 August 2016Appl. Environ. Microbiol. doi:10.1128/AEM.01413-16Copyright © 2016, American Society for Microbiology. All Rights Reserved.

on Decem

ber 11, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

2

ABSTRACT 15 Fungal hydrophobins are small amphiphilic proteins that can be used for coatings of 16 hydrophilic and hydrophobic surfaces. Through formation of monolayers they change 17 hydrophobicity of a given surface. Especially the class I hydrophobins are interesting for 18 biotechnology, because their layers are stable at high temperatures and can only be removed 19 with strong solvents. These proteins self-assemble into monolayers under physiological 20 conditions and undergo conformational changes that stabilize the layer structure. Several 21 studies demonstrate, how fusion of hydrophobins with short peptides allows to specifically 22 modify the properties of a given surface or increase the protein production levels through 23 controlled localization of hydrophobin molecules inside the cell. 24

Here we fused the Aspergillus nidulans laccase LccC to the class I hydrophobins 25 DewA and DewB and used the fusion proteins to functionalize surfaces with immobilized 26 enzymes. In contrast to previous studies with enzymes fused to class II hydrophobins, the 27 DewA-LccC fusion protein is secreted into the culture medium. The crude culture supernatant 28 was directly used for coatings of glass and polystyrene without additional purification steps. 29 The highest laccase surface activity was achieved after protein immobilization on modified 30 hydrophilic polystyrene at pH 7. This study presents an easy-to-use alternative to classical 31 enzyme immobilization techniques and can be applied not only for laccases, but also for other 32 biotechnologically relevant enzymes. 33 34 IMPORTANCE 35 Although fusion with small peptides to modify the hydrophobin properties have already been 36 performed in several studies, fusion with an enzyme presented a more challenging task. Both 37 protein partners needed to remain in active form, so that the hydrophobins could interact with 38 one another and form layers and the enzyme (e.g. laccase) would remain active at the same 39 time. Also because of the amphiphilic nature of hydrophobins their production and 40

on Decem

ber 11, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

3

purification remains challenging so far and often includes steps that would irreversibly disrupt 41 most enzymes. In our study we present the first functional fusion proteins of class I 42 hydrophobins from A. nidulans with a laccase. The resulting fusion enzyme is directly 43 secreted into the culture medium by the fungus and can be used for functionalization of hard 44 surfaces. 45 46 47 INTRODUCTION 48 Immobilization of enzymes is of increasing importance in biotechnology. It provides various 49 advantages compared to the application of free enzymes in solution, like increased stability, 50 easy recovery and reuse of the enzymes (1, 2). In some cases binding to certain surfaces even 51 improved enzyme activity (3). Several methods of immobilization are distinguished and are 52 based on chemical and physical interactions between the enzyme and the surface, each having 53 its own advantages and disadvantages (1, 2). Basic parameters like maintaining high enzyme 54 activity, prevention of enzyme leaching and contamination of the product are relevant for 55 choosing the right method depending on the reaction system. Special surface materials, 56 chemical treatments or spacer molecules are often required to ensure binding of the enzyme in 57 an active form (2). These specifications do not only limit the method application, but also 58 increase the procedure complexity and costs. An ability of some proteins, like for example 59 fungal hydrophobins, to self-assemble in stable layers under physiological conditions presents 60 a clear advantage in development of enzyme immobilization systems. 61

Hydrophobins are small amphiphilic proteins that spontaneously form monolayers on 62 hydrophilic and hydrophobic surfaces, changing their characteristics (4-6). In fungi 63 hydrophobins are secreted to reduce surface tension at the medium-air interface during hyphal 64 growth and are responsible for the hydrophobicity of the aerial structures such as aerial 65 hyphae, conidiophores, fruiting bodies and spores (6). Depending on their structural 66

on Decem

ber 11, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

4

characteristics, hydrophobins are divided into two classes, with class I protein aggregates 67 binding strongly to surfaces and resisting detergents and high temperatures (5, 7). In contrast, 68 layers of class II hydrophobin can be easily dissolved by pressure, detergents or ethanol. The 69 assembly of class I hydrophobins in highly stable monolayers is associated with the formation 70 of amyloid fibrillar structures and includes conformational changes of the protein molecules 71 upon interaction (8, 9). 72

The development of recombinantly produced hydrophobins in E. coli opened the 73 possibility to test their application in various systems (10-12). Class I hydrophobins can be 74 used to disperse hydrophobic substances in water, immobilize molecules on solid surfaces and 75 in antifouling (6). The filamentous fungus Aspergillus nidulans produces several class I 76 hydrophobins, including a well-studied protein DewA that already showed a great industrial 77 potential and appears in several patent applications (13, 14). DewA has been used as an 78 emulsion stabilizer, in optimization of biliary stents, production of microcapsules and was 79 even fused to peptides for selective enhancement of human cells adhesion to surfaces (15-17). 80 A DewA-enzyme fusion presents a tempting alternative to conventional methods of highly 81 stable surface functionalization with an enzyme. So far, a functional enzyme fusion was 82 achieved to class II hydrophobins to enhance the activity of cutinases in solution and to create 83 self-organized membranes with glucose oxidase on solid surface (18-20) Extracellular fungal 84 laccases present a suitable target for fusion with hydrophobins for immobilization due to their 85 monomeric structure, high stability and great biotechnological potential (21, 22). Different 86 conventional immobilization methods have already been reported for laccases with 87 application in dye decolorization, waste degradation and in biological fuel cells (23). 88

Laccases (EC 1.10.3.2) belong to a group of blue oxidoreductases that can oxidize 89 various aromatic and non-aromatic compounds (21, 24). Their redox potential depends on the 90 coordination of copper ions in the catalytic center and serves as a criterion to divide them into 91 three classes (25). A. nidulans produces several low redox potential laccases with low 92

on Decem

ber 11, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

5

expression levels in vegetative hyphae (26-28). Since no laccase activity can be detected in 93 supernatants of A. nidulans culture under normal growth conditions, no interference with the 94 heterologously expressed laccases occurs. The overexpressed LccC laccase showed higher 95 activity levels in culture supernatant towards the commonly used artificial substrate ABTS 96 (2,2’-azino-bis-[3-ethylthiazoline-6-sulfonate]), compared to the other characterized A. 97 nidulans laccases (26). 98

In this study we created fusion proteins of the LccC laccase with class I hydrophobins 99 DewA and DewB in A. nidulans. The proteins were directly applied for surface 100 functionalization in form of cell-free crude culture supernatant without additional purification 101 steps. We also tested different hydrophobic/hydrophilic surfaces and experimental settings to 102 determine optimal conditions for the production and immobilization of hydrophobin-fused 103 laccase. 104 105 106 MATERIALS AND METHODS 107 Strains and growth conditions. The Aspergillus nidulans strains GR5 (pyrG89; wA3; 108 pyroA4; veA1) (29) and FGSCA4 (FGSC Missouri) were cultivated in supplemented minimal 109 (MM) or yeast extract-agar-glucose (YAG) media (30). Standard cloning and transformation 110 procedures for A. nidulans were used (31-33). Top10 Escherichia coli strain (Invitrogen, 111 Karlsruhe), used for molecular biology techniques. For laccase and fusion protein production 112 A. nidulans culture was incubated in liquid YAG-Medium with 1% glucose or 2% straw (0.1 113 – 1 cm pieces) as carbon source at 28°C, 120 rpm for 2 days. 114

Construction of hydrophobin-laccase fusion proteins. Hydrophobin genes dewA 115 and dewB (AspGD ID 1837) from A. nidulans were inserted into the A. nidulans lccC gene 116 after the signal peptide sequence. For this purpose, three separately amplified DNA fragments 117 were merged using fusion PCR technique with primers listed in Table 1 as described 118

on Decem

ber 11, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

6

previously (34). The construct was then cloned into the pMCB17 vector (35) under the 119 control of the constitutive yeast gpd promoter with AscI and PacI restriction enzymes. As a 120 control the lccC gene was also cloned into the pMCB17 vector under the gpd promoter. AscI 121 restriction site in the lccC gene was removed using quick change PCR. Clones were checked 122 by sequencing. A. nidulans GR5 strain was used as the recipient strain for all plasmids. 123

RNA isolation and quantitative real-time PCR. Conidia (107 spores) were 124 inoculated in 50 ml liquid medium with glucose or straw as carbon source in 100 ml 125 Erlenmeyer flasks. After 48 h of incubation the culture was filtered through miracloth (Merck 126 KGaA, Darmstadt, Germany) and grinded in liquid nitrogen. RNA was isolated with the 127 E.Z.N.A. Fungal RNA Mini Kit (Omega Biotek, Norcross, USA) DNA was digested with 128 TURBO DNA-freeTM Kit (Thermo Fisher Scientific, Waltham, USA) and diluted to 50 ng/μl. 129 Quantitative real-time PCR was performed using SensiFAST™ SYBR® & Fluorescein One-130 Step Kit (Bioline, Lueckenwalde, Germany) on an iCycler (Bio-Rad, Munich, Germany). 131 Each reaction contained 0.2 µM primers and 100 ng RNA in 25 µl total volume. 132 Oligonucleotides are listed in Table 1. The program included 10 min reverse transcription at 133 45°C for cDNA synthesis, followed by 2.5 min inactivation of reverse transcriptase at 95°C 134 and 40 PCR cycles (10 s at 95°C, 30 s 58°C). Melting curve analysis was performed to assess 135 the specific amplification of DNA. Results for each sample from cultures grown in the 136 presence of straw were normalized to the corresponding results with h2b gene and to the 137 normalized sample obtained from cultures grown in the presence of glucose. Each expression 138 level result is the average of five independent experiments. 139

Laccase activity assay. To measure laccase activity in crude culture supernatant 107 140 conidia spores were inoculated in 50 ml liquid YAG medium with glucose or straw as carbon 141 source. The cultures were harvested 48 h after inoculation by filtering through miracloth 142 (Merck KGaA, Darmstadt, Germany) and 0.45 µm membrane filter. The laccase activity in 143 culture supernatant was assayed using 1 mM ABTS (2,2’-azino-bis-[3-ethylthiazoline-6-144

on Decem

ber 11, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

7

sulfonate]) in 50 mM acetate buffer pH 5 (36). The change in absorbance was detected at 420 145 nm over a period of 10 minutes at 25°C with an Ultrospec III Spectrophotometer (Pharmacia) 146 or EnSpire Multimode Plate Reader (PerkinElmer, Rodgau, Germany). The laccase activity in 147 solution was calculated using the molar absorption coefficient of ABTS (ε420 = 36000 l mol-1 148 cm-1), with one unit of laccase catalyzing the oxidation of 1 µmol ABTS per minute. The 149 laccase activity on surface was calculated using the molar absorption coefficient of ABTS in 150 units per cm2 of the solid surface with immobilized enzyme. One-way and two-way ANOVA 151 statistical analysis was performed using StatPlus:mac LE program (AnalystSoft, Walnut, 152 USA) to compare the effect of pH and surface hydrophobicity on laccase activity after the 153 immobilization on microtiter plate and glass surfaces. An alpha level of 0.05 was used for all 154 statistical tests. 155

Surface coating. To test the immobilization of fusion proteins on hydrophobic and 156 hydrophilic microtiter plate surfaces, 96 well non-modified polystyrene microtiter plates 157 (Greiner Bio-One, Frickenhausen, Germany) and cell culture plates with hydrophilic standard 158 growth surface (Sarstedt, Nümbrecht, Germany) were used. Coating of glass surfaces was 159 performed on high precision cover glasses (Carl Roth, Karlsruhe, Germany) and 0.4-0.85 mm 160 glass beads (Weissker, Greiz, Germany). To compare the immobilization on glass surface 161 with different characteristics, cover glasses were also treated with Sigmacote (Sigma-Aldrich, 162 Taufkirchen, Germany) to generate hydrophobic surface prior to fusion protein 163 immobilization. The coating procedure was adapted from previous studies (15). Filtered crude 164 culture supernatants were diluted to 0.1 U/ml laccase activity (appr. 0.1 mg/ml total protein), 165 directly applied on glass or polystyrene surfaces and incubated over night at 37°C in 50 mM 166 sodium acetate buffer pH 5 or sodium phosphate buffer pH 7 with 1 mM CaCl2. For the 167 immobilization in the presence of DewA recombinant H*proteinB containing YAAD-DewA-168 His (BASF-SE, Ludwigshafen, Germany) was mixed with 0.1 mg/ml laccase-containing 169 culture supernatant in the molar ratios 1:0, 1:0.5, 1:2.5 and 1:5 DewA-LccC:DewA. After 170

on Decem

ber 11, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

8

incubation the non-bound proteins were removed, the surface was washed 5 times for 10 min 171 with 50 mM sodium acetate buffer pH 5 and the laccase assay was performed under standard 172 conditions. 173

Determination of water contact angles. The static water contact angles of uncoated 174 and coated glass and polystyrene surfaces were measured with an OCA20 and the software 175 SCA 202 v3.12.11 (DataPhysics Instruments GmbH, Filderstadt, Germany). 4 µl of millipore 176 water were put on the surfaces by the “hanging drop” method and imaged with a CCD 177 camera. An ellipse fit was chosen to approach the droplet form, followed by the determination 178 of the contact angles. 179 Homology modeling. PDB file of the LccC laccase model was generated by SWISS-180 MODEL web server (http://swissmodel.expasy.org) by homology modeling method (37-39). 181 A high resolution structure (1.7 A) of a laccase from Botrytis aclada (PBD ID 3SQR) with 182 49,3 % sequence identity to LccC was chosen as template. Figures were generated using 183 PyMOL (www.pymol.org). 184 185 186 RESULTS 187 Design of fusion proteins. Two class I hydrophobins from A. nidulans were chosen for 188 fusion with the LccC laccase. Formation of monolayers on surfaces with the DewA protein is 189 well characterized, and the protein has been previously modified with peptides without 190 impairment of its coating ability (13, 15). Another hydrophobin from A. nidulans, DewB has 191 the typical class I structure, but also possesses a GPI anchor for immobilization on the spore 192 surface, where it contributes to its hydrophobicity (14). In our constructs the hydrophobin 193 genes were inserted into the lccC gene after the signal peptide sequence (Fig. 1a). This 194 expression strategy was chosen, because fusion of the laccase to the C-terminus of DewB 195 would mask the GPI anchor recognition site and prevent immobilization of DewB on the 196

on Decem

ber 11, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

9

spore surface. This fact could increase the probability to obtain a soluble fusion protein 197 outside the cells. Also the C-terminus of a laccase from Botrytis aclada, which shows the 198 highest sequence identity to the LccC laccase, compared to other published crystal structures, 199 lies in the direct vicinity of the catalytic center inside the protein molecule (40). Therefore, a 200 fusion of the hydrophobins to the C-terminus of the LccC laccase could impair the three-201 dimensional structure of the enzyme and affect its activity. 202

In the resulting fusion proteins the hydrophobin parts would be responsible for the 203 assembly into a monolayer on the surface, leaving the laccase freely exposed to the 204 surrounding medium (Fig. 1b). Due to the size difference between the LccC laccase and the 205 hydrophobins, we expected that additional native hydrophobin could be added to the coating 206 solution to act as a spacer. These small protein molecules would contribute to the formation 207 of the monolayer and prevent steric hindrance between laccase molecules. 208

Production of hydrophobin-fused laccase. The fusion constructs were transformed 209 into the A. nidulans GR5 strain. As shown in Fig. 2a, almost no laccase was present in culture 210 supernatants of transformed strains grown in submerged cultures on glucose (DewA-LccC – 211 0.1 ± 0.1 U/L, DewB-LccC – 2.9 ± 0.1 U/L). It is known that glucose represses secretion 212 whereas enzyme secretion is induced by polymers such as lignocellulose in filamentous fungi 213 (41). Straw was used as single carbon source in submerged culture, which resulted in the 214 highest activity of hydrophobin-fused enzymes two days after inoculation (Fig. 2a). The 215 activity of DewB-LccC protein (520.0 ± 146 U/L) in culture supernatant was 4.5 times higher 216 than that of the DewA-LccC (114.7 ± 10 U/L), which could be explained by the absence of 217 the GPI anchor and artificial solubilization of DewB. DewA, however, doesn’t have the GPI 218 anchor and therefore no increased secretion of the soluble protein compared to wild-type was 219 expected. As observed in the previous studies (26), no laccase activity was detected in wild-220 type stain culture supernatant. The LccC-producing strain showed moderate laccase activity in 221

on Decem

ber 11, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

10

culture supernatant in the presence of both glucose (52.0 ± 0.1 U/L) and straw (67.7 ± 11.5 222 U/L). 223

The expression levels of dewA and dewB genes in wild-type A. nidulans strain, as well 224 as the expression levels of artificial dewA::lccC and dewB::lccC genes were compared in 225 submerged culture with straw and glucose (Fig. 2b). The housekeeping gene h2b was used for 226 the normalization. Both dewA and dewB genes were expressed at higher levels in the wild-227 type strain cultures grown in the presence of straw and, although the fusion proteins were 228 expressed under the control of a constitutive promoter, their expression levels were increased 229 as well. Using straw in submerged cultures also prevented the formation of larger mycelium 230 agglomerates and increased the active surface of cultures, probably contributing to the higher 231 enzyme levels in culture supernatants. 232

Functionalization of polystyrene surface. Hydrophobin DewA with N-terminal or 233 C-terminal peptide fusions have previously been used to coat wells of 96-well microtiter 234 plates (15). Since a fusion to a large enzyme like laccase could influence the binding ability of 235 hydrophobins, different coating conditions and surfaces were tested. Non-modified 236 polystyrene microtiter plates were used as hydrophobic and cell culture plates with standard 237 growth surface as hydrophilic surfaces. The highest activity of 2.5 ± 0.1 x 10-4 U/cm2 was 238 achieved on hydrophilic surface at pH 7 with DewA-LccC protein (Fig. 3a). The activity on 239 coated hydrophobic surfaces at pH 7 was slightly lower, indicating that despite the fusion 240 with the laccase, immobilization on polystyrene was not impaired. After immobilization at pH 241 5 the laccase activity on the surface was lower, indicating that neutral pH was more preferable 242 for coating. A two-way analysis of variance showed that the effect of pH on DewA-LccC 243 surface activity was significant, F(1, 8) = 17.6, p = 0.003. The effect of microtiter plate 244 surface hydrophobicity, however, was not significant, F(1, 8) = 1.06, p = 0.33, as was not the 245 combination of these two factors, F(1, 8) = 0.4, p = 0.55. The maximal laccase activity on 246 surfaces coated with DewB-LccC was much lower, 0.1 ± 0.01 x 10-4 U/cm2 (Fig. 3a). The 247

on Decem

ber 11, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

11

two-way analysis of variance showed the significance of the effect of both pH (F(1, 8) = 248 10,25, p = 0.01) and surface hydrophobicity (F(1, 8) = 15.2, p = 0.004). The interaction of 249 these two factors was not significant. However, due to the low surface activity, the DewB-250 LccC fusion protein is not suitable for surface functionalization. Similarly, the activity levels 251 achieved with LccC alone were low (0.1 ± 0.1 x 10-4 U/cm2). The statistical analysis showed 252 that the effect of both tested factors was not significant in this case. Probably, low amounts of 253 laccase itself can stick to the surface and perform the substrate oxidation. 254

Laccase is a relatively large protein compared to the 13 kDa DewA hydrophobin. The 255 immobilization with the fusion protein alone would likely cause a reduction in the laccase 256 activity due to the steric hindrance between single LccC molecules. Previous studies with the 257 class II hydrophobin HFBI fused to a glucose oxidase revealed that the molar ratio between 258 1:1 and 1:19 GOx-HFBI to HFBI allowed the highest enzyme activity on surfaces (18). 259 Therefore, we tested different ratios between DewA and the DewA-LccC fusion protein (Fig. 260 3b). Non-modified DewA had a negative effect on the surface activity under all coating 261 conditions. This effect was proportional to the increasing DewA amounts, indicating that 262 DewA substituted the fusion protein. Since the immobilization was performed with crude 263 culture supernatant containing native exoproteins from A. nidulans, including DewA, it can be 264 assumed that the natural production increase of hydrophobins in the presence of straw (Fig. 265 2b) was sufficient for effective coating with laccase-fused DewA and did not require 266 additional supplements, as it was the case for heterologically produced and purified GOx-267 HFBI (18). 268

Functionalization of glass surfaces. The ability of laccase-fused hydrophobins to 269 coat glass was tested with non-modified and siliconized glass to compare hydrophilic and 270 hydrophobic surfaces. The effect of both pH and surface hydrophobicity on DewA-LccC 271 surface activity was significant, F(1, 8) = 18.97, p = 0.002 for hydrophobicity and F(1, 8) = 272 20.5, p = 0.002 for pH, as was the interaction of these two factors, F(1, 8) = 17.26, p = 0.003. 273

on Decem

ber 11, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

12

The highest laccase activity was achieved after coating of non-modified hydrophilic 274 microscopy cover slips at pH 7, similarly as with polystyrene microtiter plates (Fig. 4a). 275 However, the activity levels were three times lower, compared to the polystyrene coating 276 under the same conditions. The laccase activity after the coating under other conditions was 277 more than four times lower compared to non-modified glass at pH 7. The activity levels on 278 the surfaces with DewB-LccC were comparable with those achieved on polystyrene. 279 However, the laccase surface activity did not significantly depend on immobilization 280 conditions, F(1, 8) = 0.9, p = 0.37 for hydrophobicity and F(1, 8) = 4.67, p = 0.06 for pH. 281 Similarly, both factors had no significant effect on the surface activity of LccC alone, F(1, 8) 282 = 3.85, p = 0.09 for hydrophobicity and F(1, 8) = 0.38, p = 0.55 for pH. 283

Coating of non-modified glass beads resulted in the highest laccase surface activity 284 with DewA-LccC at pH 7 (Fig. 4b). The effect of pH on laccase activity was significant for 285 all tested proteins, as determined by the one-way analysis of variance: DewA-LccC – F(1, 4) 286 = 26.08, p = 0.007, DewB-LccC – F(1, 4) = 40.0, p = 0.0004, LccC – F(1, 4) = 23.14, p = 287 0.009. 288

In previous studies, the coating of glass with recombinantly produced DewA showed 289 lower protein immobilization efficiency at room temperature than at 80°C (13). Due to the 290 structural characteristics of glass surfaces, the hydrophobin immobilization could be impaired 291 at lower temperatures compared to microtiter plates and would require fixation through 292 application of higher temperatures during the incubation. 293

Determination of water contact angles. We anticipated that coating with DewA 294 would render hydrophilic surfaces hydrophobic and that coating with the DewA-laccase 295 fusion protein would change the polarity to hydrophilic due to the exposure of the hydrophilic 296 enzyme. To test this hypothesis, we performed contact angle measurements on different 297 surfaces. Therefore, 4 µl millipore water was spotted onto uncoated and coated glass and 298 polystyrene surfaces and visualized with a CCD camera (Fig. 5a). Contact angles of the 299

on Decem

ber 11, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

13

droplets were measured (Fig. 5b). Coating with DewA generated slightly more hydrophobic 300 surfaces on non-modified glass and polystyrene. On siliconized glass, DewA coating lowered 301 the hydrophobicity. This result can be explained by the amphiphilic nature of hydrophobin. 302 The protein-surface interaction mechanism is different for hydrophobic and hydrophilic 303 surfaces, leaving different parts of hydrophobin exposed. All surfaces coated with DewA-304 LccC showed a reduction in hydrophobicity compared to the uncoated surfaces, because of 305 the exposed laccase. This effect was strongest on polystyrene, which also showed the highest 306 enzymatic activity. 307 308 309 DISCUSSION 310 Class I hydrophobins are suited for establishment of highly stable monolayers on surfaces (5). 311 Compared to previously published studies that presented combinations of peptides and tags 312 bound to hydrophobins for improved production, detection or surface functionalization, 313 fusion with enzymes is more problematic (11, 15). Not only the ability of the hydrophobin to 314 self-assemble into monolayers could be impaired, but also the enzyme could lose its activity 315 due to the conformational changes caused by the fusion. Using bioinformatic tools we came 316 to the conclusion that fusions of the laccase LccC to the C-terminus of the hydrophobins 317 should allow the functionality of both protein parts of the resulting fusion product. 318 Unfortunately, due to the vicinity of the laccase C-terminus to the catalytic center, no C-319 terminal fusion of a detection tag to the enzyme was possible. However, laccase activity 320 assays could easily substitute for immunodetection assays to show immobilization and 321 functionality of the fusion proteins. 322

The choice of A. nidulans as a production strains also had several advantages and 323 disadvantages. Heterologous production of hydrophobins, especially DewA, has already been 324 established in E. coli, however it includes protein purification procedure that would 325

on Decem

ber 11, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

14

denaturate enzymes (11). Also, E. coli is not suited for the production of highly glycosylated 326 proteins like fungal laccases. A. nidulans on the other hand produced and secreted the fusion 327 proteins in the amounts suitable for direct application on surfaces. The lack of additional 328 purification steps reduces the cost and effort for protein recovery. Since all hydrophobins 329 from A. nidulans are present on the spore surface in the wild-type strain (14), an additional 330 class I hydrophobin DewB was also chosen for the fusion with the LccC laccase. The GPI 331 anchor that is normally present at the C-terminus of this hydrophobin in its native form and 332 responsible for its immobilization on spores was substituted by LccC. As expected, the 333 resulting DewB-LccC protein was present in culture supernatants at much higher 334 concentration than DewA-LccC, probably due to the disturbance of its immobilization 335 mechanism. Unfortunately, however, the DewB-LccC protein failed in creating stable 336 coatings on the tested surfaces. The concentration of DewA-LccC in crude culture supernatant 337 was, however, sufficient for conducting immobilization experiments. Cultivation of the 338 fungus with straw under conditions that stimulate protein secretion and increased expression 339 of both hydrophobins and fusion protein genes, contributed to the high protein yield. A. 340 nidulans also provided an unexpected advantage as a homologues hydrophobin producer. 341 Probably due to the presence of native DewA in the culture supernatant, no addition of 342 recombinant DewA protein was needed to achieve maximum laccase activity on the surfaces. 343

The immobilization experiments on different surfaces showed that despite the C-344 terminal fusion to a laccase, DewA retained its coating abilities on both hydrophilic and 345 hydrophobic surfaces. The coating of hydrophilic glass and polystyrene surfaces showed the 346 highest laccase activity after immobilization. This results of enzymatic activity assays are in 347 agreement with the results obtained by contact angle measurements. The reduction in 348 hydrophobicity of DewA-LccC coated surfaces, compared to uncoated or with recombinant 349 DewA coated surfaces, are probably caused by the highly glycosylated laccase. The highest 350 decrease in hydrophobicity and the highest laccase activity were both observed on the 351

on Decem

ber 11, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

15

polystyrene surface. Therefore, polystyrene has proved to be the best surface for 352 immobilization of the fusion protein. The pH conditions during coating were modified in 353 comparison to previous studies, performed at pH 8 (13, 15), due to the acidic pH optimum 354 profile of fungal laccases (21). Though pH 5 had a clear negative effect on the 355 immobilization, laccases are generally stable at neutral pH and pH 7 was suitable for the 356 assembly of the hydrophobin layer. The polystyrene surface proved to be more suitable for 357 coating at low temperatures. Fusion of DewA to thermostable enzymes would provide a 358 possibility for fixing the hydrophobin layer on glass at higher temperatures. Since DewA 359 forms stable coatings after 16 hours incubation at 80°C (13), it is perfectly suited for fusion 360 with thermostable enzymes. 361 The LccC laccase itself is a low redox potential laccase with relatively low activity 362 (26) and is not suited for industrial application. Due to a conserved three-dimensional 363 structure and a conserved reaction mechanism, its fusion to a class I hydrophobin and 364 successful coating serves as proof of principle for a possibility of enzyme immobilization 365 using this system. LccC can be substituted by other laccases that are of interest in for example 366 dye decolorization, phenolic waste degradation or biofuel cells (22). Also other monomeric 367 enzymes like lipases, dehydrogenases or decarboxylases can be fused to hydrophobins for 368 surface functionalization. Class I hydrophobins like DewA provide not only a stable binding 369 to the surface, they create an ordered protein monolayer with even enzyme exposure to the 370 surroundings. Therefore, the overloading of the surface, which often leads to the inhibition of 371 enzyme activity in other immobilization techniques, is prevented. Hydrophobins also do not 372 require additional treatment, because their coating ability is caused by a natural process of 373 transformation from the soluble form to monolayer aggregates at the air/liquid and 374 liquid/surface interfaces. 375 376 ACKNOWLEDGMENTS 377

on Decem

ber 11, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

16

We would like to thank E. Wohlmann for technical assistance. 378 We declare that we have no competing interests. 379 380 381 FUNDING INFORMATION 382 This work was supported by the Federal Ministry of Education and Research (BMBF) 383 through the program “BioProFi” (FKZ: 03SF0424). 384 The funders had no role in study design, data collection and interpretation, or the decision to 385 submit the work for publication. 386 387 388 REFERENCES 389 1. Sheldon RA. 2007. Enzyme immobilization: the quest for optimum performance. 390 Advanced Synthesis & Catalysis 349:1289-1307. 391 2. Hanefeld U, Gardossi L, Magner E. 2009. Understanding enzyme 392 immobilisation. Chem Soc Rev 38:453-468. 393 3. Pierre A. 2004. The sol-gel encapsulation of enzymes. Biocatalysis and 394 Biotransformation 22:145-170. 395 4. Wosten H, De Vries O, Wessels J. 1993. Interfacial Self-Assembly of a Fungal 396 Hydrophobin into a Hydrophobic Rodlet Layer. Plant Cell 5:1567-1574. 397 5. Kershaw MJ, Talbot NJ. 1998. Hydrophobins and repellents: proteins with 398 fundamental roles in fungal morphogenesis. Fungal Genet Biol 23:18-33. 399 6. Wosten HA, Scholtmeijer K. 2015. Applications of hydrophobins: current state 400 and perspectives. Appl Microbiol Biotechnol 99:1587-1597. 401 7. Scholtmeijer K, Wessels JG, Wosten HA. 2001. Fungal hydrophobins in medical 402 and technical applications. Appl Microbiol Biotechnol 56:1-8. 403

on Decem

ber 11, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

17

8. Gebbink MF, Claessen D, Bouma B, Dijkhuizen L, Wosten HA. 2005. Amyloids-404 -a functional coat for microorganisms. Nat Rev Microbiol 3:333-341. 405 9. Morris VK, Ren Q, Macindoe I, Kwan AH, Byrne N, Sunde M. 2011. 406 Recruitment of class I hydrophobins to the air:water interface initiates a multi-407 step process of functional amyloid formation. J Biol Chem 286:15955-15963. 408 10. Kwan AH, Winefield RD, Sunde M, Matthews JM, Haverkamp RG, Templeton 409 MD, Mackay JP. 2006. Structural basis for rodlet assembly in fungal 410 hydrophobins. Proc Natl Acad Sci U S A 103:3621-3626. 411 11. Wohlleben W, Subkowski T, Bollschweiler C, von Vacano B, Liu Y, Schrepp 412 W, Baus U. 2010. Recombinantly produced hydrophobins from fungal analogues 413 as highly surface-active performance proteins. Eur Biophys J 39:457-468. 414 12. Kirkland BH, Keyhani NO. 2011. Expression and purification of a functionally 415 active class I fungal hydrophobin from the entomopathogenic fungus Beauveria 416 bassiana in E. coli. J Ind Microbiol Biotechnol 38:327-335. 417 13. Rieder A, Ladnorg T, Woll C, Obst U, Fischer R, Schwartz T. 2011. The impact 418 of recombinant fusion-hydrophobin coated surfaces on E. coli and natural mixed 419 culture biofilm formation. Biofouling 27:1073-1085. 420 14. Grunbacher A, Throm T, Seidel C, Gutt B, Rohrig J, Strunk T, Vincze P, 421 Walheim S, Schimmel T, Wenzel W, Fischer R. 2014. Six hydrophobins are 422 involved in hydrophobin rodlet formation in Aspergillus nidulans and contribute 423 to hydrophobicity of the spore surface. PLoS One 9:e94546. 424 15. Boeuf S, Throm T, Gutt B, Strunk T, Hoffmann M, Seebach E, Muhlberg L, 425 Brocher J, Gotterbarm T, Wenzel W, Fischer R, Richter W. 2012. Engineering 426 hydrophobin DewA to generate surfaces that enhance adhesion of human but not 427 bacterial cells. Acta Biomater 8:1037-1047. 428

on Decem

ber 11, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

18

16. Weickert U, Wiesend F, Subkowski T, Eickhoff A, Reiss G. 2011. Optimizing 429 biliary stent patency by coating with hydrophobin alone or hydrophobin and 430 antibiotics or heparin: an in vitro proof of principle study. Adv Med Sci 56:138-431 144. 432 17. Schulz A, Liebeck BM, John D, Heiss A, Subkowski T, Böker A. 2011. Protein–433 mineral hybrid capsules from emulsions stabilized with an amphiphilic protein. 434 Journal of Materials Chemistry 21:9731-9736. 435 18. Takatsuji Y, Yamasaki R, Iwanaga A, Lienemann M, Linder MB, Haruyama T. 436 2013. Solid-support immobilization of a "swing" fusion protein for enhanced 437 glucose oxidase catalytic activity. Colloids Surf B Biointerfaces 112:186-191. 438 19. Ribitsch D, Herrero Acero E, Przylucka A, Zitzenbacher S, Marold A, 439 Gamerith C, Tscheliessnig R, Jungbauer A, Rennhofer H, Lichtenegger H, 440 Amenitsch H, Bonazza K, Kubicek CP, Druzhinina IS, Guebitz GM. 2015. 441 Enhanced cutinase-catalyzed hydrolysis of polyethylene terephthalate by 442 covalent fusion to hydrophobins. Appl Environ Microbiol 81:3586-3592. 443 20. Espino-Rammer L, Ribitsch D, Przylucka A, Marold A, Greimel KJ, Herrero 444 Acero E, Guebitz GM, Kubicek CP, Druzhinina IS. 2013. Two novel class II 445 hydrophobins from Trichoderma spp. stimulate enzymatic hydrolysis of 446 poly(ethylene terephthalate) when expressed as fusion proteins. Appl Environ 447 Microbiol 79:4230-4238. 448 21. Baldrian P. 2006. Fungal laccases - occurrence and properties. FEMS Microbiol 449 Rev 30:215-242. 450 22. Kunamneni A, Ballesteros A, Plou FJ, Alcalde M. 2007. Fungal laccase—a 451 versatile enzyme for biotechnological applications. Communicating current 452 research and educational topics and trends in applied microbiology 1:233-245. 453

on Decem

ber 11, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

19

23. Fernandez-Fernandez M, Sanroman MA, Moldes D. 2013. Recent 454 developments and applications of immobilized laccase. Biotechnol Adv 31:1808-455 1825. 456 24. Thurston CF. 1994. The structure and function of fungal laccases. Microbiology 457 140:19-26. 458 25. Eggert C, LaFayette PR, Temp U, Eriksson KE, Dean JF. 1998. Molecular 459 analysis of a laccase gene from the white rot fungus Pycnoporus cinnabarinus. 460 Appl Environ Microbiol 64:1766-1772. 461 26. Mander GJ, Wang H, Bodie E, Wagner J, Vienken K, Vinuesa C, Foster C, 462 Leeder AC, Allen G, Hamill V, Janssen GG, Dunn-Coleman N, Karos M, 463 Lemaire HG, Subkowski T, Bollschweiler C, Turner G, Nusslein B, Fischer R. 464 2006. Use of laccase as a novel, versatile reporter system in filamentous fungi. 465 Appl Environ Microbiol 72:5020-5026. 466 27. Aramayo R, Timberlake WE. 1990. Sequence and molecular structure of the 467 Aspergillus nidulans yA (laccase I) gene. Nucleic Acids Res 18:3415. 468 28. Scherer M, Fischer R. 2001. Molecular characterization of a blue-copper laccase, 469 TILA, of Aspergillus nidulans. FEMS Microbiol Lett 199:207-213. 470 29. Waring RB, May GS, Morris NR. 1989. Characterization of an inducible 471 expression system in Aspergillus nidulans using alcA and tubulin-coding genes. 472 Gene 79:119-130. 473 30. Barratt RW, Johnson GB, Ogata WN. 1965. Wild-type and mutant stocks of 474 Aspergillus nidulans. Genetics 52:233-246. 475 31. Sambrook J, Russel DW. 1999. Molecular Cloning: A laboratory manual. Cold 476 Spring Harbor Laboratory Press. 477

on Decem

ber 11, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

20

32. Kafer E. 1977. Meiotic and mitotic recombination in Aspergillus and its 478 chromosomal aberrations. Adv Genet 19:33-131. 479 33. Yelton MM, Hamer JE, Timberlake WE. 1984. Transformation of Aspergillus 480 nidulans by using a trpC plasmid. Proc Natl Acad Sci U S A 81:1470-1474. 481 34. Szewczyk E, Nayak T, Oakley CE, Edgerton H, Xiong Y, Taheri-Talesh N, 482 Osmani SA, Oakley BR. 2006. Fusion PCR and gene targeting in Aspergillus 483 nidulans. Nat Protoc 1:3111-3120. 484 35. Fernandez-Abalos JM, Fox H, Pitt C, Wells B, Doonan JH. 1998. Plant-adapted 485 green fluorescent protein is a versatile vital reporter for gene expression, protein 486 localization and mitosis in the filamentous fungus, Aspergillus nidulans. Mol 487 Microbiol 27:121-130. 488 36. Eggert C, Temp U, Eriksson KE. 1996. The ligninolytic system of the white rot 489 fungus Pycnoporus cinnabarinus: purification and characterization of the laccase. 490 Appl Environ Microbiol 62:1151-1158. 491 37. Arnold K, Bordoli L, Kopp J, Schwede T. 2006. The SWISS-MODEL workspace: a 492 web-based environment for protein structure homology modelling. 493 Bioinformatics 22:195-201. 494 38. Benkert P, Biasini M, Schwede T. 2011. Toward the estimation of the absolute 495 quality of individual protein structure models. Bioinformatics 27:343-350. 496 39. Biasini M, Bienert S, Waterhouse A, Arnold K, Studer G, Schmidt T, Kiefer F, 497 Gallo Cassarino T, Bertoni M, Bordoli L, Schwede T. 2014. SWISS-MODEL: 498 modelling protein tertiary and quaternary structure using evolutionary 499 information. Nucleic Acids Res 42:W252-258. 500 40. Osipov E, Polyakov K, Kittl R, Shleev S, Dorovatovsky P, Tikhonova T, Hann 501 S, Ludwig R, Popov V. 2014. Effect of the L499M mutation of the ascomycetous 502

on Decem

ber 11, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

21

Botrytis aclada laccase on redox potential and catalytic properties. Acta 503 Crystallogr D Biol Crystallogr 70:2913-2923. 504 41. Bouws H, Wattenberg A, Zorn H. 2008. Fungal secretomes--nature's toolbox for 505 white biotechnology. Appl Microbiol Biotechnol 80:381-388. 506 507

on Decem

ber 11, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

22

FIGURE LEGENDS 508 FIG 1 Schematic representation of hydrophobin-laccase constructs and their immobilization 509 on surfaces. (a) DewA-LccC and DewB-LccC constructs. Hydrophobin genes were inserted 510 into the lccC gene after the signal peptide sequence (SP) to allow fusion to the laccase. (b) 511 Schematic representation of solid surface coated with hydrophobin-laccase fusion protein. 512 Red triangles represent enzyme substrate, orange rhombs – product. Magnification shows the 513 structure of DewA hydrophobin (PDB ID 2LSH), colored in blue, and LccC laccase model, 514 colored in green, as strings and ribbons. 515 516 FIG 2 Production of the hydrophobin-fused LccC laccase. (a) Laccase activity in culture 517 supernatant, measured with an ABTS based assay. The cultures were incubated with glucose 518 or straw, as indicated. (b) Expression of dewA, dewB genes in wild-type and dewA::lccC, 519 dewB::lccC constructs measured with real-time PCR. The results obtained with straw were 520 normalized to the signals obtained from cultures grown in the presence of glucose. Error bars 521 represent standard deviations from five experiments. 522 523 FIG 3 Coating of polystyrene microtiter plates. (a) Crude culture supernatants containing 524 LccC, DewA-LccC and DewB-LccC proteins with 0.1 U/ml laccase activity were used for 525 coating 96-well microtiter plates with hydrophobic/hydrophilic surface characteristics at pH 5 526 and 7. Laccase surface activity was measured using the ABTS assay. Uncoated plates were 527 used as control. (b) Different DewA-LccC:DewA molar ratios were used to determine optimal 528 coating conditions on hydrophobic/hydrophilic surface at pH 5 and 7. Recombinantly 529 produced H*proteinB (DewA) was mixed with DewA-LccC-containing culture supernatant in 530 rations 1:0, 1:0.5, 1:2.5, 1:5, as indicated. Standard deviations from three independent 531 experiments for each data point are indicated by error bars. 532 533

on Decem

ber 11, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

23

FIG 4 Coating of glass. (a) Microscopy cover slips were coated using culture supernatants 534 containing LccC, DewA-LccC and DewB-LccC with 0.1 U/ml laccase activity at pH 5 and 7. 535 Hydrophobic glass surface was generated though siliconization. Laccase activity on surfaces 536 was measured using the ABTS assay. (b) Glass beads with 0.4-0.85 mm diameter were coated 537 with laccase-containing culture supernatants at pH 5 and 7. Standard deviations from three 538 independent experiments for each data point are indicated by error bars. 539 540 FIG 5 Water contact angle measurements. (a) 4 µl of millipore water were put on untreated 541 glass, siliconized glass and polystyrene. Surfaces were tested uncoated, coated with DewA 542 and coated with DewA-LccC. (b) The angle (degree) between the surface and the droplet was 543 measured. The mean of 10 measurements is displayed, and the error bars represent the 544 standard deviation. 545

on Decem

ber 11, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

TABLES

TABLE 1 Oligonucleotides used for cloning and real-time PCR

Oligo name Sequence (5’-3’)

gpd(p)fw TATAGAATTCAATCATCCTTATTCGTTGACC

gpd(p)rev TATAGGCGCGCCTGTGATGTCTGCTCAAGC

P1SP(lccC)fw TCTGGTTTCTTCTTTCTATGGG

P2SPdewABlccCfw TATAGGCGCGCCATGCTGCGTTCTTCCTTTCT

P3dewAF1fw ATGCGCTTCATCGTCTCTC

P3dewBF2fw GACAAGTTCCCCGTCCCC

P5SP(lccC)F1rev AGAGAGACGATGAAGCGCATGGCAGAAGCATAGAGTGCATA

P5SP(lccC)F2rev GGGACGGGGAACTTGTCGGCAGAAGCATAGAGTGCAT

P6dewAF1rev CTCAGCCTTGGTACCGGC

P6dewBF2rev CAGAATGGAGCCAAGGGC

P7SPdewABlccCrev TATATTAATTAACTAGACACCCGAATCATACTG

P8lccCrev TTTCACACAGGAAACAGCTATG

lccCfw TAGGCGCGCCATGCTGCGTTCTTCCTTTCT

lccCrev TCTTAATTAACTAGACACCCGAATCATACTG

QC1LccCAscIdel GCTGGCGAGCCCAGCAGTACGGGACAAC

QC2LccCAscIdel CTGGGCTCGCCAGCGGTACGTGTACTC

rtDewAfw CTCTCGGGCAACACTGG

rtDewArev GGCAACACAGTTGGTGGTTC

rtDewBfw GAGTGACCTGCTCGGTG

rtDewBrev CCACTCTTCGCACAGCAAG

rtDewALccCfw GAACCACCAACTGTGTTGCC

rtDewBLccCfw CTTGCTGTGCGAAGAGTGG

on Decem

ber 11, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

rtDewABLccCrev CGAGACATCGGTCCCAAAG

rtH2Bfw TGCCGAGAAGAAGCCTAGCA

rtH2Brev GAGTAGGTCTCCTTCCTGGT

gpdfwSeq CCCAGTCACGACGTTGTA

lccCfwSeq GATACAAACTACCACGAC

revSeq GCATGCCTGCAGGTCGAC

on Decem

ber 11, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

on Decem

ber 11, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

on Decem

ber 11, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

on Decem

ber 11, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

on Decem

ber 11, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

on Decem

ber 11, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

on Decem

ber 11, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

on Decem

ber 11, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

on Decem

ber 11, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

on Decem

ber 11, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

on Decem

ber 11, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from