Scholarly article on topic 'A Lytic Polysaccharide Monooxygenase with Broad Xyloglucan Specificity from the Brown-Rot Fungus Gloeophyllum trabeum and Its Action on Cellulose-Xyloglucan Complexes'

A Lytic Polysaccharide Monooxygenase with Broad Xyloglucan Specificity from the Brown-Rot Fungus Gloeophyllum trabeum and Its Action on Cellulose-Xyloglucan Complexes Academic research paper on "Biological sciences"

Share paper
OECD Field of science

Academic research paper on topic "A Lytic Polysaccharide Monooxygenase with Broad Xyloglucan Specificity from the Brown-Rot Fungus Gloeophyllum trabeum and Its Action on Cellulose-Xyloglucan Complexes"

AEM Accepted Manuscript Posted Online 2 September 2016 Appl. Environ. Microbiol. doi:10.1128/AEM.01768-16 Copyright © 2016 Kojima et al.

This is an open-access article distributed under the terms of the Creative Commons Attribution 4.0 International license.

1 Characterization of an LPMO from the brown-rot fungus Gloeophyllum

2 trabeum with broad xyloglucan specificity, and its action on cellulose-

3 xyloglucan complexes

5 Yuka Kojima3'*, Aniko Varnaib'*, Takuya Ishidac, Naoki Sunagawac, Dejan M. Petrovicb,

6 Kiyohiko Igarashic'd', Jody Jellisone, Barry Goodellf, Gry Alfredsen6, Bj0rge Westerengb,

7 Vincent G.H. Eijsinkb't,#' Makoto Yoshida^*

9 aDepartment of Environmental and Natural Resource Science, Tokyo University of

10 Agriculture and Technology, Fuchu, Tokyo, Japan

11 bDepartment of Chemistry, Biotechnology and Food Science, Norwegian University of Life

12 Sciences (NMBU), As, Norway

13 cDepartment of Biomaterial Sciences, Graduate School of Agriculture and Life Sciences, The

14 University of Tokyo, Bunkyo-ku, Tokyo, Japan

15 dVTT Technical Research Centre of Finland, Espoo, Finland

16 eCenter for Agriculture, Food and the Environment, University of Massachusetts, Amherst,

17 Massachusetts, USA

18 'Department of Sustainable Biomaterials, Virginia Polytechnic Institute and State University,

19 Blacksburg, Virginia, USA

20 gNorwegian Institute of Bioeconomy Research, As, Norway

03 Q. 0 Q.

03 C/J


o cr 0

—5 0

22 Running title: Enzymatic properties of a brown-rot LPMO

24 Section: enzymology and protein engineering

26 Keywords: brown-rot, LPMO, AA9, viscosity, xyloglucan, cellulose

28 #To whom correspondence should be addressed: V.G.H. Eijsink,;

29 M. Yoshida,

30 *,f: These authors contributed equally to the manuscript.

CD Q. 0 Q.


O o o cr 0

—5 0

33 Abstract

34 Fungi secrete a set of glycoside hydrolases and lytic polysaccharide monooxygenases

35 (LPMOs) to degrade plant polysaccharides. Brown-rot fungi, such as Gloeophyllum trabeum,

36 tend to have few LPMOs and information on these enzymes is scarce. The genome of G.

37 trabeum encodes four AA9 LPMOs, whose coding sequences were amplified from cDNA. Due

38 to alternative splicing, two variants of GtLPMO9A seem to be produced, a single domain

39 variant, GtLPMO9A-1, and a longer variant, GtLPMO9A-2, which contains a C-terminal

40 domain comprising approximately 55 residues without a predicted function. We have

41 overexpressed the phylogenetically distinct GtLPMO9A-2 in Pichia pastoris and investigated

42 its properties. Standard analyses, using HPAEC-PAD and MS, showed that GtLPMO9A-2 is

43 active on cellulose, carboxymethylcellulose and xyloglucan. Importantly, compared to other

44 known xyloglucan-active LPMOs, GtLPMO9A-2 has broad specificity, cleaving at any position

45 along the P-glucan backbone of xyloglucan, regardless of substitutions. Using dynamic

46 viscosity measurements to compare the hemicellulolytic action of GtLPMO9A-2 to that of a

47 well-characterized hemicellulolytic LPMO, NcLPMO9C from Neurospora crassa, revealed that

48 GtLPMO9A-2 is more efficient in depolymerizing xyloglucan. These measurments also

49 revealed minor activity on glucomannan that could not be detected by the analysis of

50 soluble products by HPAEC-PAD and MS and that was lower than the activity of NcLPMO9C.

51 Experiments with co-polymeric substrates showed an inhibitory effect of hemicellulose-

52 coating on cellulolytic LPMO activity and did not reveal additional activities of GtLPMO9A-2.

53 These results provide insight into the LPMO-potential of G. trabeum and provide a novel

54 sensitive method, measurement of dynamic viscosity, for monitoring LPMO activity.


03 C/J


o cr CD

55 Importance

56 Currently, there are only a few methods available to analyze end-products of lytic

57 polysaccharide monooxygenase (LPMO) activity, the most common ones being liquid

58 chromatography and mass spectrometry. Here we present an alternative and sensitive

59 method based on measurement of dynamic viscosity, for real-time continuous monitoring of

60 LPMO activity in the presence of water-soluble hemicelluloses such as xyloglucan. We have

61 used both this novel and existing analytical methods to characterize a xyloglucan-active

62 LPMO from a brown rot fungus. This enzyme, GtLPMO9A-2, differs from previously

63 characterized LPMOs, in having broad substrate specificity, enabling almost random

64 cleavage of the xyloglucan backbone. GtLPMO9A-2 acts preferentially on free xyloglucan,

65 suggesting a preference for xyloglucan chains that tether cellulose fibres together. The

66 xyloglucan-degrading potential of GtLPMO9A-2 suggests a role in decreasing wood strength

67 at the initial stage of brown-rot, through degradation of the primary cell wall.

69 Introduction

70 For decades, the enzymatic degradation of cellulose by filamentous fungi was considered

71 to proceed through the hydrolytic action of cellulases. As early as in 1950, the involvement

72 of additional factors in cellulose conversion was proposed (1), and in 1974, Eriksson et al.

73 showed that enzymatic hydrolysis of cellulose by a crude fungal enzyme mixture is promoted

74 by the presence of molecular oxygen, suggesting a role for redox reactions (2). The

75 explanation for the observations by Eriksson et al. came almost four decades later, when

76 copper-dependent lytic polysaccharide monooxygenases (LPMOs), which are currently

77 classified into auxiliary activities (AA) family 9, 10, 11, and 13 in the CAZy database (3, 4),

Ш Q. CD Q.



о cr CD

CD i О Ш

78 were first described (5-7). Since their discovery, LPMOs received much attention and are

79 currently considered as one of the key enzymes in fungal cellulose degradation (8). LPMOs

80 employ molecular oxygen and an external electron donor (5, 9-12) to carry out oxidative

81 cleavage of the p-1,4-glucosidic bonds in cellulose. Some LPMOs exclusively oxidize C1,

82 others exclusively oxidize C4, whereas a third type of LPMOs yields a mixture of C1- and C4-

83 oxidized products. Recently, LPMOs have been found to show oxidative activity against

84 various other plant polysaccharides, including xyloglucan, glucomannan, xylan and starch

85 (13-19). Thus, the physiological roles of LPMOs in biomass decomposition are likely to be

86 more complex and varied than initially assumed.

87 Brown-rot fungi are a group of wood-rotting basidiomycetous fungi and represent the

88 dominant wood decay fungi in northern coniferous forest ecosystems. They are able to

89 remove plant cell wall polysaccharides, such as cellulose and hemicelluloses together with

90 extensive lignin depolymerization and modification, but without lignin metabolism (20-22).

91 Brown-rot fungi employ a unique system of wood degradation, in which they combine

92 enzymatic and chemical mechanisms. The chemical mechanism implies the formation of

93 hydroxyl radicals (*OH) through a chelator-mediated Fenton (CMF) reaction that randomly

94 attack the wood cell wall components (22-26). This process causes structural and chemical

95 changes in the wood cell wall, potentially providing carbohydrate-active enzymes with

96 greater access to the substrate by generating holes large enough for the enzymes to

97 infiltrate the cell wall (21). Compared to soft-rot and white-rot fungi, the majority of brown-

98 rot fungi seem to have an incomplete enzymatic system for cellulose degradation. In

99 particular, they do not produce processive cellobiohydrolases, which are key enzymes in the

100 degradation of crystalline cellulose (27-29). Brown-rot fungi also often lack cellobiose

101 dehydrogenase (CDH), which is considered to be a natural electron donor of fungal LPMOs

03 Q. CD Q.

03 C/J


o cr CD

102 (9, 10, 12), albeit not the only one (11, 30, 31). While several cellulases and hemicellulases

103 from brown-rot fungi have been characterized, little is known about brown-rot LPMOs (32).

104 In general, the physiological function of oxidative enzyme systems in wood degradation by

105 brown-rot fungi is still unclear.

106 Currently, detection of LPMO activity on polymeric substrates is based on measurment of

107 the production of soluble oligomeric products which are released as a result of LPMO

108 cleavage near polymer chain ends or cleavage of the same polymer chain twice at nearby

109 positions. Such products can be detected using various types of chromatography (33, 34) or

110 mass spectrometry (5). The activity of strictly C4-oxidizing LPMOs may also be quantified

111 directly using a standard method for detection of the newly generated reducing ends (13).

112 Recently, a semiquantitative, high-throughput method for screening activity towards water-

113 soluble substrates has been reported (35), which is suitable for a wide variety of

114 carbohydrate-acting enzymes and can be adapted for LPMOs (13). LPMO-action on solid

115 substrates may also be quantified using confocal laser scanning microscopy after labeling C1-

116 oxidized cellulose chain ends with a fluorescence dye that is specific for carboxylic acids (36).

117 Unlike for endo-acting glycoside hydrolases, the use of gel permeation chromatography or

118 viscosity measurements for detecting LPMO-generated decreases in the molecular weight of

119 carbohydrate polymers has not yet been reported. Such analytical tools could potentially

120 reveal activities that cannot be detected by the other available methods and are also of

121 industrial relevance because they relate to (reducing) biomass viscosity.

122 In the present study, we cloned five cDNAs encoding putative LPMO9s (GtLPMO9s) from

123 Gloeophyllum trabeum, which is one of the most studied brown-rot fungi. We

124 heterologously expressed the one LPMO that, by phylogenetic analysis, seemed distant from

125 most other known LPMOs, namely GtLPMO9A-2. This enzyme has the longer amino acid

03 Q. CD Q.

03 C/J


o cr CD

CD i O' 03

126 sequence of two naturally occurring variants of GtLPMO9A (see below) and was expressed in

127 Pichia pastoris. As part of the in-depth characterization of GtLPMO9A-2, we developed a

128 method for monitoring LPMO-activity on hemicelluloses by measuring reduction in viscosity.

129 We show that GtLPMO9A-2 is a promiscuous LPMO with a unique ability to cleave

130 xyloglucan regardless of the substitutions of the p-1,4-linked glucan backbone. Using

131 dynamic viscosity measurements we were able to detect activity on additional

132 hemicelluloses that remained undetected in the standard chromatographic and mass

133 spectrometry analyses.

135 Materials and Methods

136 Strains and enzymes

137 Gloeophyllum trabeum strain NBRC 6430 was used as a source of LPMO genes.

138 Escherichia coli strain JM109 (Takara Bio, Shiga, Japan) and Pichia pastoris strain KM71H

139 (Invitrogen, Carlsbad, CA) were used as hosts for subcloning experiment and heterologous

140 production of recombinant GtLPMO9A-2, respectively.

141 Recombinant NcLPMO9C from Neurospora crassa (UniProt:Q7SHI8) was prepared

142 according to Kittl et al. (37). Endoglucanases AfCel12A from Aspergillus fumigatus

143 (UniProt:Q8TG26) and TaCel5A from Thermoascus aurantiacus (UniProt:Q8TG26) were

144 produced and purified as previously described (38).

145 Sequence analysis


03 C/J


o cr CD

146 Multiple sequence alignments were generated using MAFFT, version 7.295 (39), as

147 available at the European Bioinformatics Institute website. For phylogenetic analysis of

148 LPMO9s, amino acid sequences were aligned using MAFFT and then manually edited using

149 Sea View (40). A phylogenetic tree was generated from this alignment by the neighbor-

150 joining method (41) in ClustalX software (42), with 1,000 bootstraps as previously described

151 (43).

152 Cloning of genes encoding GtLPMO9s

153 G. trabeum was cultivated in Highley's medium (44) containing 0.5% glucose (Wako,

154 Osaka, Japan) as the sole carbon source, and without thiamine hydrochloride. After 7 days of

155 cultivation at 23°C on 120 rpm, total RNA was extracted using the RNeasy Plant Mini Kit

156 (Qiagen, Venlo, The Netherlands). First-strand cDNA was then synthesized using reverse

157 transcriptase (SuperScript III, Invitrogen) with 3' rapid amplification of the cDNA ends

158 (3'RACE) using the GeneRacer Kit (Invitrogen) according to the manufacturer's instructions;

159 the 3'RACE primers used are listed in Table 1. The regions encoding the mature LPMOs were

160 sequenced after PCR amplification from first-strand cDNA using the ORF primers. The

161 nucleotide sequences of the genes (cDNA) encoding GtLPMO9A-1, GtLPMO9A-2, GtLPMO9C,

162 and GtLPMO9D have been deposited in the DDBJ database under accession numbers

163 LC157847, LC157848, LC157849, and LC157850, respectively. The cDNA sequence of

164 GtLPMO9B has already been described by Jung et al. (32) and has been deposited in the

165 GenBank database under accession number AEJ35168 as a endo-p-1,4-glucanase (named

166 Cel61G in the CAZy database). For expression, the insertion fragments of the GtLPMO9s,

167 except for GtLPMO9A-2, were amplified by expression primers appending the XhoI site and

168 the Kex2 cleavage site at the 5' end and the NotI site at the 3' end of the coding sequences

169 without signal peptide (Table 1) and cloned into pPICZa (Invitrogen) by restriction digestion

170 and ligation. After amplification with expression primers, the insertion fragment of

03 Q. CD Q.

03 C/J


o cr CD

171 GtLPMO9A-2 was inserted into pPICZa (Invitrogen) using the In-Fusion HD Cloning Kit

172 (Takara Bio); this strategy was chosen because the ORF encoding GtLPMO9A-2 contains a

173 NotI cleavage site.

174 Heterologous expression of the GtLPMO9s and purification of GtLPMO9A-2

175 Approximately 10 |g of pPICZa expression plasmid DNA was linearized with Bpu1102I

176 (Takara Bio) prior to transformation into P. pastoris. Electroporation, selection of

177 transformants, and production of recombinant protein were carried out according to the

178 instruction manual of the EasySelect Pichia expression kit (Invitrogen). After cultivating for 4

179 days in YP media using methanol as the carbon source, the cells were removed by

180 centrifugation (30 min at 10,000 g), and then the supernatants were applied for SDS-PAGE

181 analysis to evaluate the production of recombinant proteins.

182 For purification of GtLPMO9A-2, all steps were carried out at 4 °C, unless indicated

183 otherwise. The P. pastoris strain expressing GtLPMO9A-2 was grown in 4 liter YP medium

184 containing 1% glycerol for 1 day at 30 °C at 180 rpm. The cells were collected by

185 centrifugation at 1,500 g for 5 min and resuspended in 400 ml YP medium containing 1%

186 (v/v) methanol to induce expression and incubated further at 30 °C. Every 24 hours,

187 methanol was added to the culture, to a final concentration of 1% (v/v). After 4 days, the

188 cell-free supernatant was harvested by centrifugation at 10,000 g for 30 min. Ammonium

189 sulfate was added to the cell-free broth, to a final concentration of 50% (w/w). After

190 removing the precipitate by centrifugation (30 min at 10,000 g), the supernatant was diluted

191 approximately 2.2 times with 20 mM sodium acetate buffer (pH 4.5), to a final concentration

192 of approximately 1 M ammonium sulfate, and applied to a Toyopearl Phenyl-650S column

193 (020 mm by 160 mm; TOSOH, Tokyo, Japan) equilibrated with 20 mM sodium acetate buffer

03 Q. 0 Q.

03 C/J


o cr 0

—5 0

0 i o'

194 (pH 4.5) containing 1 M ammonium sulfate. The protein was eluted by applying a 100-mL

195 linear reverse-gradient to 20 mM sodium acetate buffer (pH 4.5). Fractions containing the

196 recombinant protein were collected and pooled, and the buffer was changed to 20 mM

197 sodium acetate buffer (pH 4.0) using Vivaspin 20 centrifugal concentrator tubes with a 10

198 kDa MWCO membrane (Sartorius AG, Gottingen, Germany). Then, the sample was applied to

199 a Toyopearl SP-650S column (020 mm by 250 mm; TOSOH) equilibrated with the same

200 buffer. The flow-through, containing the recombinant protein, was collected, buffer

201 exchanged to 20 mM sodium acetate buffer (pH 5.5) with Vivaspin 20 centrifugal

202 concentrator tubes, and ca. 13.1 mg proteins were treated with 1 ^L Endo H (500 Units; New

203 England Biolabs, Ipswich, MA) at 30 °C overnight. The protein solution was then buffer-

204 exchanged to 20 mM Tris-HCl buffer (pH 7.0) with Vivaspin 20 centrifugal concentrator tubes

205 and applied to a Toyopearl DEAE-650S column (020 mm by 250 mm; TOSOH) equilibrated

206 with the same buffer. The recombinant protein was eluted from the column with a linear

207 NaCl gradient, and the fractions eluting from 28 to 52 mM NaCl were collected. Protein

208 purity was analyzed by SDS-PAGE analysis with 12% polyacrylamide gels, and the N-terminal

209 amino acid sequence was determined with a protein sequencer (model 491 cLC; Applied

210 Biosystems, Foster City, CA).

211 Substrates

Ш Q. CD Q.


о t-о

К) О

212 The following substrates were used to characterize the substrate specificity of

213 GttPMO9A-2: phosphoric acid swollen cellulose (PASC), carboxymethylcellulose (CMC),

214 cellohexaose, cellopentaose, tamarind xyloglucan - partially arabinosylated (XG), xyloglucan

215 oligosaccharide (XG-oligomers), xyloglucan heptasaccharide (XG7), konjac glucomannan

216 (GM), ivory nut mannan, wheat arabinoxylan, oat flour mixed-linkage (ß-1,3-1,4) glucan, oat

CD i О Ш

217 spelt xylan, beech wood xylan and birchwood xylan. PASC was prepared as described earlier

218 (45); birchwood xylan was purchased from Carl Roth GmbH (Karlsruhe, Germany). Konjac

219 glucomannan used in the dynamic viscosity experiments was purchased from Wako; konjac

220 glucomannan used in the coating experiments was purchased from Megazyme (Wicklow,

221 Ireland). Oat spelt xylan was purchased from Serva Electrophoresis GmbH (Heidelberg,

222 Germany); beech wood xylan was purchased from Sigma-Aldrich (St. Louis, MO). All other

223 substrates were obtained from Megazyme.


224 LPMO-activity measurements

225 Dynamic viscosity experiments

226 Dynamic viscosity of the reaction solution was determined using an AMVn Automated

227 Micro Viscometer (Anton Paar, Graz, Austria), which is a falling ball-type viscometer

228 measuring the rolling time of a ball through a liquid in a capillary. Reaction mixtures of 1 mL

229 in total contained 1 ^M copper-saturated enzyme (see below), 0.5 mM dithiothreitol (DTT),

230 and substrate, in 50 mM sodium acetate buffer (pH 5.0). The concentration of each

231 substrate was adjusted, based on its intrinsic viscosity, so that the viscosity falls within the

232 optimal range of the viscometer. The concentrations of konjac glucomannan, tamarind

233 xyloglucan, arabinoxylan, and CMC were 0.05%, 0.15%, 0.2%, and 0.5% (w/v), respectively.

234 The reactions were carried out inside a glass capillary (01.6 mm) containing a steel ball (01.5

235 mm), thermostated at 30 °C. The capillary was positioned at a 50-degree angle relative to

236 horizontal (causing the ball to move through the reaction mixture along the capillary) and

237 was automatically inverted when the ball reached its lower end. The rolling time of the steel

238 ball was continuously recorded with the viscometer while the capillary was inverted 2000

239 times. End-point samples (after 16 hours of incubation, when the change in viscosity had

03 C/J


o cr CD

240 levelled off) were analyzed by HPAEC-PAD and MALDI-ToF MS to check for the formation of

241 soluble oxidized oligosaccharides, as described below.

242 Dynamic viscosity was calculated as K-(p - Po)-T, where K is the calibration constant of the

243 system (mPa-cmVg), p is the density of the ball (g/cm3), p0 is the density of the reaction

244 mixture (g/cm3), and T is the rolling time of ball (sec). The calibration constant, K was

245 experimentally determined to be 0.00845 mPa-cm3/g using water as a standard.

246 The dynamic viscosity data were fit to an exponential decay formula (y=a-e(b'x)+c) with

247 DeltaGraph (SPSS Inc., CA), where y is the dynamic viscosity, x is the time, and a, b, and c are

248 constants. The constant c is the final viscosity of the substrate. The initial decline rate of

249 dynamic viscosity was calculated by differentiating the formula with respect to x=0. In case

250 of arabinoxylan and CMC, the correlation between dynamic viscosity and time was linear.

251 Endoglucanase treatment

252 To convert oxidized products into short oligosaccharides, the end-point samples of the

253 dynamic viscosity experiments were subjected to endoglucanase treatment and the resulting

254 oligomer mixtures were analyzed by HPAEC-PAD and MALDI-ToF MS (see below). Purified

255 endoglucanases, AfCel12A and IaCel5A, were diluted to a concentration of 100 |M, and 1 |l

256 AfCel12A or IaCel5A solution was added to 50 |l of the reaction mixtures containing LPMO-

257 treated xyloglucan or glucomannan. Subsequently, the reactions were incubated for 20

258 hours at 30 °C.

259 Screening for substrate specificity

260 After Cu(II)-saturation (46), GtLPMO9A-2 (or NcLPMO9C) was incubated with various

261 substrates in a 100 |L total reaction volume containing 1 |M enzyme, 0.5 mM dithiothreitol

262 (DTT) or 1 mM ascorbic acid, and substrate in 50 mM sodium acetate buffer (pH 5.0). The

03 Q. CD Q.

03 C/J


o cr CD

263 concentrations of substrates were as follows: 0.2% (w/v) PASC or Avicel, 0.2% (w/v) CMC,

264 0.05% (w/v) GM, 0.15% (w/v) XG, 0.2 g/l cellopentaose, cellohexaose, XG7 or XG-oligomer,

265 0.5% (w/v) mannan, xylan, or mixed-linkage glucan. Reactions were incubated at 30 °C in an

266 Eppendorf Thermomixer C (Eppendorf AG, Hamburg, Germany) with shaking at 1,000 rpm

267 for 16 hours. The reactions were run in at least duplicates, the reaction mixtures were boiled

268 for 10 minutes after sampling, and the solids were separated by centrifugation at 10 000 g

269 for 5 min. The supernatants were analyzed by HPAEC-PAD and MALDI-ToF MS (see below) to

270 check for the formation of soluble oxidized oligosaccharides.

271 Activity on complex substrates

272 In order to evaluate LPMO activity on complex substrates, hemicellulose was coated on

273 cellulose by pre-mixing aqueous solution of 1% (w/v) tamarind XG or 1% (w/v) konjac GM

274 with 1% (w/v) PASC in a 1:1 (v/v) ratio for 15 minutes (at room temperature without shaking,

275 in 20 mM Na-acetate buffer, pH 5.0) prior to enzyme addition. After the pre-incubation

276 reactions, reaction mixtures were prepared in 20 mM Na-acetate buffer (pH 5.0) containing

277 0.2% (w/v) PASC and 0.2% (w/v) XG or GM. Reaction mixtures containing 0.2% (w/v) of XG,

278 GM, or PASC as single substrate were also prepared. Further, the reaction mixtures

279 contained 1 ^M GtLPMO9A-2 or WcLPMO9C (copper-saturated according to Loose et al. (46))

280 and 1 mM ascorbic acid, and they were incubated at 30 °C, for 24 hours. The reactions were

281 run in at least duplicates, the reaction mixtures were boiled for 10 minutes after sampling,

282 and the solids were separated by centrifugation at 10 000 g for 5 min. Products were

283 analyzed using HPAEC-PAD as described below. Control experiments were performed

284 without electron donor or enzyme.

03 Q. 0 Q.

03 C/J


o cr 0

—5 0

0 i o'

285 Detection of oxidized oligosaccharides

286 The oligosaccharides released by LPMO action and sequential LPMO-endoglucanase

287 treatment were analyzed by high-performance anion exchange chromatography (HPAEC) on

288 a Dionex ICS3000 system equipped with pulsed amperometric detection (PAD), using a 50289 minute gradient (33) for cellulosic and a 75-minute gradient (13) for hemicellulosic

290 substrates.

291 The oligosaccharides were further analyzed using MALDI-ToF mass spectrometry (MALDI-

292 ToF MS). The analysis was carried out on an Ultraflex MALDI-ToF/ToF instrument (Bruker

293 Daltonics, Bremen, Germany) equipped with nitrogen 337 nm laser beam as described

294 earlier (5). Samples (2 |l) were applied to an MTP 384 ground steel target plate TF (Bruker

295 Daltonics) together with 4.5 mg 2,5-dihydroxybenzoic acid (DHB) matrix dissolved in 0.5 ml

296 30% acetonitrile. Data were collected with the lowest laser energy necessary to obtain

297 sufficient quality spectra, using Bruker's flexControl software. Spectra were analyzed using

298 Bruker's flexAnalysis software. All samples analyzed in this study contained 50 mM sodium

299 acetate, which suppressed the formation of potassium adducts.

301 Results

302 Nucleotide and amino acid sequences

303 The publicly available genome database of G. trabeum shows four genes encoding AA9

304 enzymes (29), which we have named lmpo9A (location in the G. trabeum genome database

305 v1.0: scaffold_00011:177323-178510), lmpo9B (scaffold_00011:179858-181138), lmpo9C

306 (scaffold_00002:3232078-3233061), and lmpo9D (scaffold_00001:4158823-4160504). So far,

03 Q. CD Q.

03 C/J


o cr CD

307 only one of these genes, lmpo9B, has been cloned (called Cel61G in the CAZy database);

308 studies of the recombinant protein have suggested that it acts on cellulose but the enzyme

309 remains largely uncharacterized (32). In the present study, the cDNA species transcribed

310 from the lpmo genes were cloned by RT-PCR using total RNA extracted from mycelium of G.

311 trabeum. The lmpo9A gene was transcribed into two different mRNAs (named lmpo9A-1 and

312 -2), the first 706 nucleotides of which were identical. The first 706 nucleotides were followed

313 by another 53 nucleotides, corresponding to an exon for the shorter splicing variant

314 (GtLPMO9A-1) or an intron for the longer splicing variant (GtLPMO9A-2). The latter, longer

315 transcript continued with another 407 nucleotides (Fig. S1).

316 The coding transcripts of lmpo9A-1, A-2, B, C, and D consist of 756, 1110, 756, 768, and

317 1032 bp, respectively, excluding the stop codon. Excluding the N-terminal signal peptide

318 from the translated amino acid sequences, as predicted by the SignalP program (47), the

319 mature GtLPMO9A-2 protein consists of 351 amino acids with a molecular weight (MW) of

320 36.0 kDa and a predicted pI of 4.56. For the other LPMOs these values are: GtLPMO9A-1, 233

321 amino acids, MW 25.1 kDa, pI 4.94; GtLPMO9B, 232 amino acids, MW 24.5 kDa, pI 6.19;

322 GtLPMO9C, 239 amino acids, MW 25.8 kDa, pI 4.68; GtLPMO9D, 321 amino acids, MW 33.1

323 kDa, pI 5.84.

324 In a phylogenetic tree of LPMO9s, GtLPMO9A-1, A-2, and B cluster in the same group,

325 whereas GtLPMO9C and D ended up in a different cluster each (Fig. 1). GtLPMO9A-1, A-2,

326 and B are closely related to a C1/C4 oxidizing LPMO from Neurospora crassa (NcLPMO9M or

327 NCU-07898; (48, 49)). Among characterized LPMOs, the ones being closest to GtLPMO9C are

328 the strictly C4-oxidizing LPMOs NcLPMO9A (NCU-02240; (49)), NcLPMO9C (a well-

329 characterized LPMO from Neurospora crassa with broad P-1,4-glucan-degrading abilities;

330 NCU-02916; (13, 15, 50)), and NcLPMO9D (NCU-01050; (48, 49)). GtLPMO9D is part of a

03 Q. CD Q.

03 C/J


o cr CD

CD i O' 03

331 clade that seems quite distinct from all LPMO9 enzymes characterized so far. Fig. 2 shows a

332 multiple sequence alignment including the GtLPMO9s. The catalytic domains of all four

333 LPMOs contain the two characteristic copper-binding histidines (His1 and His86 in

334 GtLPMO9A-2) as well as a conserved tyrosine that occupies one of the axial positions in the

335 copper coordination sphere (Tyr175 in GtLPMO9A-2) (6).

336 GtLPMO9A-2 and D have an additional C-terminal domain that starts with a region of low

337 sequence complexity that could possibly be a linker (Fig. S2). We were not able to detect

338 sequence similarity with known carbohydrate-binding domains (CBMs) or with other

339 domains of known function. For more detailed sequence analysis of the C-terminal domains,

340 see the Supplementary material (including Figs. S2 - S4). Notably, the sequence alignments

341 of Fig. S3 & S4 show that the C-terminal domains of GtLPMO9A-2 and GtLPMO9D also occur

342 in other LPMOs. Furthermore, the alignment shows that the C-terminal domain of

343 GtLPMO9A-2 has features not unlike those of well-known cellulose-binding domains.

344 As a first step towards mapping the ability of G. trabeum to oxidatively cleave plant

345 polysaccharides, we have overexpressed and characterized GtLPMO9A-2. This enzyme was

346 selected as a representative for three (GtLPMO9A-1, A-2, and B) of the five G. trabeum

347 LPMOs that are closely related (Fig. 1) and that cluster in an area of the phylogenetic tree

348 with little available functional data. Also, the coding sequence of GtLPMO9A-2 has not been

349 reported before. GtLPMO9D was also considered interesting, but we were not able to

350 express this enzyme in sufficient amounts.

351 Physical properties of recombinant GtLPMO9A-2

352 GtLPMO9A-2 was expressed in Pichia pastoris and purified to homogeneity by 3 steps of

353 column chromatography (Fig. S5). Purified GtLPMO9A-2 appeared to have a significantly


03 C/J


o cr CD

354 higher MW (ca. 60 kDa) than that estimated from the amino acid sequence (36.0 kDa).

355 Treatment with EndoH led to a small decrease in MW (to approx. 57 kDa), indicating N-

356 glycosylation of the protein. The remaining mass difference suggests that the recombinant

357 enzyme was also O-glycosylated, perhaps in the serine-rich linker region. Correct processing

358 of the purified protein was confirmed by determination of N-terminal amino acid sequence

359 of the purified enzyme, yielding HGYVDQVTIG, which is identical to the amino acid sequence

360 deduced from the nucleotide sequence. As observed previously for LPMO9s expressed in P.

361 pastoris, the N-terminal histidine was not methylated (14, 15, 51).

362 Cellulolytic activity of recombinant GtLPMO9A-2


03 C/J

363 GtLPMO9A-2 was active on phosphoric acid-swollen cellulose (PASC) (Fig. 3), similar to all

364 LPMO9s characterized so far (52), and Avicel (data not shown). HPAEC analysis showed

365 diagnostic product patterns that reflect a mixed C1/C4 oxidation pattern, similar to the

366 phylogenetically closest characterized LPMO (Fig. 1), WcLPMO9M (NCU-07898; (49)). The

367 type of electron donor, ascorbic acid or DTT, had no effect on the product profile of the

368 LPMO (Fig. S6) but the reaction was slower when using DTT. GtLPMO9A-2 was inactive on

369 shorter cello-oligosaccharides, such as cellopentaose (Fig. S7A) and cellohexaose (not

370 shown), compared to WcLPMO9C. Neither GtLPMO9A-2 nor WcLPMO9C showed activity on

371 XG7 (not shown). Interestingly, GtLPMO9A-2 was active on carboxymethylcellulose (CMC)

372 (Fig. S8), suggesting that the enzyme can act on soluble polymeric ß-glucans and is not

373 restricted by substitutions on the D-glucose backbone.

374 Hemicellulolytic activity of recombinant GtLPMO9A-2


o cr CD

The activity of GtLPMO9A-2 against hemicellulosic substrates was assayed using dynamic

376 viscosity analysis, using NcLPMO9C, with known activity towards several p-1,4-glucans (13),

377 as positive control. As expected, significant decreases in viscosity were observed when

378 xyloglucan (XG, Fig. 4b) or glucomannan (GM, Fig. 4d) was incubated with NcLPMO9C,

379 whereas no significant change of viscosity was detected in the reaction with arabinoxylan

380 (Fig. 4f). Similarly to NcLPMO9C, GtLPMO9A-2 also caused a decrease in viscosity in the

381 reactions with XG (Fig. 4a) and GM (Fig. 4c) but not with arabinoxylan (Fig. 4e). Both

382 enzymes also reduced slightly the viscosity in reactions with CMC (Fig. 4g,h). The dynamic

383 viscosity remained unchanged in reactions without the LPMO or an electron donor, showing

384 that the observed effects are due to oxidative cleavage by the LPMOs and not to e.g.

385 contaminating hydrolases. The fact that treatment with the LPMOs led to a drop in viscosity

386 in the presence (but not in the absence) of an electron donor clearly indicates that both

387 enzymes have an endo-type of action on XG and GM.

388 Comparative studies revealed that GtLPMO9A-2 reduced the viscosity of XG almost two-

389 fold faster than NcLPMO9C, whereas the difference in final viscosity was neglectable (Table

390 2). The viscosity data indicated that the Gloeophyllum enzyme was slightly less active than

391 the Neurospora enzyme on GM, both in terms of initial rate and final viscosity. Both

392 GtLPMO9A-2 and NcLPMO9C were 5-10-fold slower in reducing the viscosity of a CMC

393 solution (0.5%, w/v) as compared with XG (0.15%, w/v).

394 We used HPAEC-PAD and MALDI-ToF MS to analyse water-soluble oligosaccharides

395 potentially released by GtLPMO9A-2 from a selection of hemicellulosic substrates. No

396 products could be detected upon incubating the LPMO (+ electron donor) with ivory nut

397 mannan (a linear P-1,4-linked mannan), wheat arabinoxylan, oat spelt xylan, beech wood

398 xylan, and birchwood xylan (heteropolymers with P-1,4-linked xylan backbone) or mixed

399 linked p-glucan (with a P-1,3 linkage at every third linkage in the p-1,4-linked glucan

400 backbone) (data not shown). Data for XG and GM are discussed below.

03 Q. CD Q.

03 C/J


o cr CD

CD i O' 03

401 Identification of oligosaccharides released from xyloglucan and glucomannan

402 To confirm LPMO-action on hemicelluloses, which eventually would lead to the formation

403 of oligosaccharides, we subjected the end-point samples from the dynamic viscosity

404 experiments to HPAEC-PAD analysis. The HPAEC-PAD-profiles show that GtLPMO9A-2

405 released a broad range of xyloglucan oligosaccharides eluting from 25 to 60 minutes,

406 whereas the major products generated by NcLPMO9C eluted between 45 and 60 minutes

407 (Fig. 5A). No products could be detected when either of the LPMOs or the reducing agent

408 was incubated with the substrate alone. To simplify the product profiles, we treated the

409 samples with AfCel12A, a XG-active endoglucanase, which depolymerizes XG into its

410 repeating oligomeric units by cleaving next to non-substitued glucose units. The known

411 dominant products of AfCel12A are XXXG, XXLG/XLXG, XLLG (G, glucose; X, glucose carrying a

412 xylose substitution at C6; L, X with a further galactose appended to the C2 of the xylose).

413 Note that the basic repeating unit in tamarind xyloglucan is XXXG carrying galactosylations at

414 various positions (13, 53). In addition, the tamarind xyloglucan was partially arabinosylated,

415 carrying 0.27 arabinosyl unit per repeating (XXXG) unit (compositional data provided by the

416 manufacturer, Megazyme).

417 The resulting chromatograms (Fig. 5B) are dominated by standard (non-oxidized) XG

418 breakdown products that were released by the endoglucanase from the still mainly

419 polymeric XG remaining after the LPMO treatment. The dominating products with a four

420 glucan backbone (XXXG fragments with varying degrees of galactosylation) elute between 30

421 and 40 minutes, whereas less abundant longer products elute between 45 and 55 minutes in

422 discrete peaks (Fig. 5B, black line). The latter products are likely to be a variety of oligomers

423 of DP 12-18, some of which contain arabinosyl units (54). Interestingly, some of these

424 oligomeric products (marked with grey dashed line in Fig. 5B) were absent in samples that

03 Q. CD Q.

03 C/J


o cr CD

425 had been treated with LPMO + reductant (compare reactions without, light blue and red,

426 and with, dark blue and red lines, reducing agent in Fig. 5B). Next to these standard

427 products, the product profiles from reactions that had been pretreated with an active LPMO

428 showed additional peaks, and the patterns of these additional peaks reveal differences

429 between the two enzymes. Endoglucanase treatment of the product mixture generated by

430 GtLPMO9A-2 yielded a wide variety of additional products, with elution times ranging from

431 20 to 60 minutes and including products eluting earlier than XXXG (XG7; Fig. 5B, green line).

432 In contrast, endoglucanase treatment of NcLPMO9C-generated material yielded a less varied

433 collection of additional products, which almost exclusively eluted at 45 minutes or later and

434 which included a cluster of products eluting at 61-66 minutes that is lacking from the

435 reactions with GtLPMO9A-2. The product profile obtained here with NcLPMO9C is similar to

436 what has been observed before for this C4-oxidizing enzyme (13), whereas the product

437 profile for GtLPMO9A-2 is clearly different. Based on data presented by Bennati-Granier et

438 al. (14), the peak eluting at 43 min that is unique for GtLPMO9A-2 could be C1-oxidized XG7.

439 HPAEC-PAD chromatograms for the reactions with konjac GM showed clear effects of

440 NcLPMO9C whereas the effects of GtLPMO9A-2 on GM were less pronounced (Fig. S9A).

441 Treatment of GM with NcLPMO9C led to a clear shift in the polymer peak, and small

442 amounts of oligmeric products were also observed. For the reaction with GtLPMO9A-2, only

443 a minor shift in the polymeric peak was observed. Subsequent endoglucanase treatment

444 with TaCel5A led to the accummulation of chromatographically distinct products in case of

445 NcLPMO9C, while similar products were not observed after the combined action of

446 GtLPMO9C and TaCel5A (Fig. S9B).

447 MALDI-ToF MS analysis of the reaction mixtures obtained after the dynamic viscosity

448 experiments confirmed major differences in the product profiles generated by GtLPMO9A-2


03 C/J


o cr CD

449 and NcLPMO9C (Fig. 6). While NcLPMO9C released clusters of oxidized xyloglucan fragments

450 with the number of xylose residues being a multiple of three, GtLPMO9A released a much

451 broader range of oxidized oligosaccharides. The results for NcLPMO9C are in accordance

452 with previous work (13) showing that the enzyme cleaves XG at the non-reducing end of the

453 non-substituted glucosyl residue in the (XXXG)n backbone of xyloglucan, forming C4-oxidized

454 XG fragments. On the other hand, the data for GtLPMO9A-2 suggest that this LPMO can

455 cleave almost anywhere in the XG chain, since reactions with this enzyme yield oxidized XG

456 fragments with a number of xylose residues not being only a multiple of three. Addition of

457 LPMO in the absence of reducing agent did not have an impact on the product profiles

458 obtained from XG after endoglucanase treatment. This indicates that the GtLPMO9A-2 and

459 NcLPMO9C enzyme preparations were not contaminated with background activity of an

460 enzyme that would cleave off xylosyl units of XG, and that all XG fragments in Fig. 6 were

461 products of LPMO-treatment. Upon treating GM with GtLPMO9A-2, we could not detect

462 oxidized oligosaccharides with MALDI-ToF MS.

463 Closer inspection of Fig. 6 (see Fig. S10) showed that the labeled signals correspond to

464 clusters containing signals for the native, the non-hydrated oxidized, and the hydrated

465 oxidized oligosaccharide, with the former two signals dominating. Signals for the sodium salt

466 of the aldonic acid and double-oxidized species, which would be indicative of C1 oxidation,

467 were not observed. Although it is not certain that oxidized xyloglucan oligosaccharides

468 behave like oxidized cello-oligomers under these analytical conditions (see legend to Fig. 3),

469 the data suggest that C4-oxidation dominated during the degradation of XG.

470 In addition to the expected masses originating from galactosylated XG fragments, the

471 MALDI-ToF spectra for the reactions with GtLPMO9A-2 contained a few species with more

472 than three pentose units per repeating unit (4 hexoses), and hence are predicted to carry an


03 C/J


o cr CD

473 arabinosylated X-unit (Fig. 6; (54)). Since the arabinosyl and xylosyl groups (both pentoses)

474 cannot be distinguished by MS, we cannot exclude that some of the other fragments

475 generated by GtLPMO9A-2 also contain an arabinosyl group replacing a xylosyl unit (see

476 proposed structures by Niemann et al. (54)).

477 The chromatogram of Fig. 5B shows that incubation of the LPMO-treated xyloglucan

478 sample with AfCel12A resulted in production of shorter xyloglucan oligosaccharides,

479 producing mostly native oligosaccharides with DP 7-9 and a wide variety of putatively

480 oxidized products, particularly in the case of GtLPMO9A-2 (Fig. 5B). MALDI-ToF MS (results

481 not shown) showed that many of the higher DP species visible in Figs. 5B and 6A were no

482 longer present after combined endoglucanase-LPMO action. The combined action of

483 GtLPMO9A-2 and AfCel12A on XG resulted in the formation of shorter fragments (in

484 particular, m/z 494.1 (oxidized XG/GX or L) and 791.4 (native XXG/XGX/GXX or XL/LX)). These

485 shorter native species may correspond to the compounds eluting before the XG7 peak, at

486 20-30 min, in Fig. 5B. It is noteworthy that the native species observed with HPAEC may also

487 originate from C4-oxidized products as it has recently been shown that C4-oxidized products

488 are unstable and tend to loose the oxidized monosugar under the chromatographic

489 conditions used here (34).

490 LPMO activity on complex substrates

CD Q. 0 Q.

03 C/J


o cr 0

—5 0

491 In order to evaluate possible additional differences between GtLPMO9A-2 and WcLPMO9C

492 and to study LPMO action on more natural substrates, we then carried out experiments with

493 co-polymeric substrates generated by mixing xyloglucan or glucomannan with PASC, using

494 ascorbic acid as reducing agent (Fig. 7; note that the reaction is faster when using ascorbic

495 acid, explaining the apparent differences between Fig. 7 and Fig. 5A). These experiments

496 were inspired by the findings by Frommhagen et al. (17), who demonstrated LPMO activity

497 on xylan, but only if the xylan was grafted onto cellulose. In general, hemicellulose coating

498 hindered LPMO-activity on cellulose, reducing the amount of oxidized cello-oligosaccharides

499 released (compare blue and purple lines in Figs. 7 and S10). Interestingly, coating XG on

500 cellulose also seemed to have a negative effect on the activity of both LPMOs towards XG

501 itself (compare red and purple lines in Fig. 7). Experiments with GM-coated PASC and

502 GtLPMO9A-2 showed that coating with GM completely abolished GtLPMO9A-2 activity on

503 PASC, whereas soluble GM products were not detected by the HPAEC-PAD analysis (Fig.

504 S11). Interestingly, control reactions with WcLPMO9C, with much higher activity on GM,

505 showed that this LPMO degraded the GM and that, consequently, degradation of PASC was

506 still observed after coating with GM. We carried out several experiments with birchwood

507 xylan, alone or mixed with PASC, without detecting any activity on xylan.

03 Q. 0 Q.

03 C/J

509 Discussion

510 The present data show that GtLPMO9A-2 is a C1/C4-oxidizing LPMO with high activity on

511 cellulose and xyloglucan, negligible activity on cellodextrins and a minor activity on

512 glucomannan. The enzyme has the exceptional capability of cleaving any Glc-Glc bond in

513 xyloglucan, regardless of substitutions, thus functionally expanding the arsenal of

514 hemicellulolytic LPMOs so far described (13, 14, 16, 17). Xylosyl substitutions at the C6

515 position in XG seemed to be less inhibitory than carboxymethyl substitutions in CMC, which

516 may be located at the C2, C3, or C6 positions on the glucan backbone. The fact that

517 GtLPMO9A-2 was not active on cello-oligosaccharides, nor on mixed-linked glucan, suggests

518 that GtLPMO9A-2, in contrast to WcLPMO9C, requires longer stretches of p-1,4-glucosidic

—5 0

0 i o'

519 linkages in the polysaccharide backbone for substrate recognition. Perhaps, the minor

520 depolymerizing activity that we observed with konjac glucomannan could reflect the

521 occasional presence of longer stretches of p-1,4-linked glucose units. The fact that

522 GtLPMO9A-2 was inactive towards XG7 (consisting of a cellotetraose backbone),

523 cellopentaose, and cellohexaose may be taken to imply that the enzyme requires at least

524 seven glucosyl units for productive substrate binding.

525 So far, only three xyloglucan-active LPMOs have been studied in detail, NcLPMO9C from

526 Neurospora crassa (13), PaLPMO9H from Podospora anserina (14), and An3046 from

527 Aspergillus nidulans (16). All these LPMOs cleave xyloglucan next to the non-substituted

528 glucosyl residue (i.e. G-unit), yielding a clustered product profile on MALDI-ToF MS (as in Fig.

529 6B). In contrast, the oxidized products released by GtLPMO9A-2 had a wide variety of

530 masses (Fig. 6A), showing that GtLPMO9A-2 activity is unaffected by substitution of the

531 glycosyl units at the C6 position and that GtLPMO9A-2 can cleave everywhere in the p-

532 glucan main chain. These observations are further supported by the wide range of products

533 observed upon the combined action of GtLPMO9A-2 and AfCel12A (which only cleaves at

534 non-substituted G's), including substituted products with dimeric or trimeric backbones that

535 only can emerge upon cleavage in between two substituted main chain sugars.

536 The oxidative regioselectivities of xyloglucan-active LPMOs on cellulose show that all

537 variants occur, C1 (An3046; literature data are not conclusive; C1/C4 activity is also possible),

538 C4 (NcLPMO9C) and C1/C4 (PaLPMO9H and GtLPMO9A-2). However, there is little data on

539 the oxidative regioselectivity of these enzymes on xyloglucan. Our own previous data for

540 NcLPMO9C and the present data for GtLPMO9A-2 indicate that these enzymes almost

541 exclusively oxidize C4 when acting on XG.


03 C/J


o cr CD

542 The sequence alignment of Fig. 2 shows that GtLPMO9A-2 has an extended L2 loop

543 (compared to NcLPMO9C and PoLPMO9H), while it lacks the L3 loop and also has a deletion

544 (relative to NcLPMO9C and PoLPMO9H) in the so-called LC-loop, very close to a conserved

545 surface-exposed aromatic residue (Tyr204 in NcLPMO9C and Tyr215 in GtLPMO9A-2). The

546 same applies to An3046, although this was not noted by the authors (16). The L3 loop has

547 been proposed to be a structural determinant of xyloglucan activity (14), and a recent NMR

548 study of enzyme-substrate interactions in NcLPMO9C showed that the L3 loop indeed

549 interacts with xyloglucan (55). The present data, and data for An3046, show that xyloglucan

550 cleavage can also be achieved by LPMOs lacking the L3 loop. The extension of the L2 loop

551 may compensate for the deletion of the neighboring L3 loop and it is interesting to note that

552 the extension and deletion are of the same length, 14 amino acids. Lack of the L3 loop and

553 the seemingly correlated deletion near Tyr204/215 may have effects on the specificity of

554 xyloglucan cleavage and/or the ability to cleave shorter substrates. For example, only the

555 two xyloglucan-active LPMOs with the L3 loop and with no deletion in the LC loop, i.e.

556 NcLPMO9C and PoLPMO9H, have been shown to cleave soluble cellodextrins (14, 50).

557 Viscosity measurements have occasionally been used to assess activity of endoglucanases

558 on water-soluble polysaccharides (56-58). Here we show that dynamic viscosity

559 measurements provide an alternative and sensitive method for assessing and comparing

560 LPMO-activity on water-soluble polysaccharides. This simple method proved to be a good

561 choice for both characterizing substrate specificity and comparing depolymerization rates. In

562 fact, by relying solely on the standard methods (HPAEC-PAD and MALDI-ToF MS) that are

563 based on monitoring released oligomeric products, we would have overlooked the low

564 activity of GtLPMO9A-2 on glucomannan. These findings show the need for using

Ш Q. CD Q.


о »-


CD i О Ш

565 complementary methods, including methods capable of monitoring the polymeric fraction,

566 when characterizing LPMOs.

567 Furthermore, while quantification of LPMO activity is complex, due to product

568 heterogeneity and lack of product solubility, the dynamic viscosity measurements provide a

569 direct measurement of the number of cuts introduced in the polysaccharide substrate. This

570 method is independent of oxidative regioselectivity and the cleavage pattern. In the present

571 study, the dynamic viscosity experiments showed that WcLPMO9C is more active on

572 glucomannan compared to GtLPMO9A-2, whereas the latter depolymerized xyloglucan at a

573 higher rate. The higher activity of GtLPMO9A-2 on xyloglucan could be due to its broader

574 cleavage specifity, which implies that there are more productive binding sites on the

575 substrate.

576 The experiments with co-polymeric substrates yielded somewhat surprising results. We

577 show that hemicellulose coating of PASC reduces LPMO activity on PASC dramatically. This

578 provides a rationale for hemicellulolytic activity among LMPOs, as exemplified by the studies

579 with glucomannan. WcLPMO9C, with its glucomannan activity was able to degrade GM-

580 coated PASC (degrading both GM and PASC), wheras GtLPMO9A-2 with only very weak GM

581 activity, could not degrade the mixture of these two polymers. On the other hand, in

582 mixtures of cellulose and XG, the activity on both polymers was strongly reduced, for both

583 tested LPMOs (see below for further discussion).

584 To date, there is limited information on LPMOs from brown-rot fungi. Vanden

585 Wymelenberg et al. showed low expression levels of LPMO genes in the brown-rot fungus

586 Postia placenta even when cultivated on aspen wood (59). Expression of LPMOs was

587 undetectable in proteome analyses for the brown-rot fungi Fomitopsis pinicola and

588 Wolfiporia cocos growing on aspen wood (60). These results indicate that LPMOs may not


03 C/J


o cr CD

CD i O' 03

589 constitute a major enzyme component in the wood degradation system of brown-rot fungi.

590 However, recently, Jung et al. reported that saccharification of oak and kenaf pretreated

591 with popping was promoted by a homologue of GtLPMO9B, suggesting the importance of

592 this enzyme in the enzymatic degradation of plant biomass by brown-rot fungi (32).

593 Moreover, we were able to amplify all G. trabeum LPMOs from mRNA, meaning that the

594 genes are expressed, and we show that GtLPMO9A-2 is active on biomass.

595 Brown-rot fungi, such as G. trabeum, preferentially invade softwood. In softwood, the

596 main hemicelluloses are glucomannan and glucuronoxylan, which occur in the secondary cell

597 wall. The secondary cell wall is surrounded by a primary cell wall, where the most abundant

598 hemicelluloses are xyloglucan and pectin (61-63). The genome of G. trabeum encodes sets of

599 CAZymes that target xyloglucan (GH12s and GH74) and pectin (10 GH28s, GH43s, 2 CE8s, and

600 CE15), besides cellulose (e.g. GH3s, GH5s and GH12s), glucomannan (GH3s and GH5s), and

601 xylan (3 GH10s, GH43s, CE1, and 6 CE16s) (29). Due to its substrate promiscuity, GtLPMO9A-

602 2 could play a role in the depolymerization of not only cellulose but also xyloglucan during

603 biomass-degradation. Perhaps the enzyme has additional capabilities that remain

604 undiscovered, perhaps activities on natural co-polymeric plant cell walls. It could be of

605 interest to test other electron donors, since there are indications that the electron donor

606 affects LPMO performance, including substrate specificity (11, 31). By degrading xyloglucan,

607 GtLPMO9A-2 may loosen the adhesion between cellulose bundles, thus creating access for

608 various CAZymes. The fact that GtLPMO9A-2 (and also WcLPMO9C) acted preferentially on

609 free XG as compared to cellulose-bound XG suggests that GtLPMO9A-2 may preferentially

610 cut XG chains that tether cellulose fibres together as opposed to XG chains that adhere to

611 cellulose fiber surfaces.

03 Q. CD Q.

03 C/J


o cr CD

612 It is well known that wood strength rapidly decreases at the initial stage of brown-rot

613 even before the loss of weight in wood (64). As the fungus invades the wood cells from the

614 lumen, depolymerization of cell wall components using a low molecular weight decay system

615 facilitates enzyme penetration into the cell wall and enables enzymatic degradation of the

616 outer primary cell wall concurrent with the inner secondary cell wall layers (22). The rapid

617 decrease in cellulose DP throughout the wood cell wall and the degradation of the primary

618 cell wall are likely to be primary causes of strength reduction at this decay stage (22). It is

619 tempting to speculate that GtLPMO9A-2 plays a role in this process, i.e. primary cell wall

620 degradation at the initial stage of brown-rot.


03 C/J

622 Funding information

623 The work was carried out as part of the Biomim project funded by the Norwegian

624 Research Council through grant 243663, and partly co-funded by the Bilateral Joint Research

625 Projects programme of JSPS in Japan (to M.Y.), and JSPS KAKENHI (grant number 16J03946

626 to Y.K., 15H04526 to M.Y., and 16K14952 to M.Y.).

628 References


—5 0

629 1. Reese ET, Siu RG, Levinson HS. 1950. The biological degradation of soluble cellulose

630 derivatives and its relationship to the mechanism of cellulose hydrolysis. J Bacteriol

631 59:485-497.

632 2. Eriksson KE, Pettersson B, Westermark U. 1974. Oxidation: an important enzyme

633 reaction in fungal degradation of cellulose. FEBS Lett 49:282-285.

634 3. Cantarel BL, Coutinho PM, Rancurel C, Bernard T, Lombard V, Henrissat B. 2009. The

635 Carbohydrate-Active EnZymes database (CAZy): an expert resource for

636 Glycogenomics. Nucleic Acids Res 37:D233-D238.

637 4. Levasseur A, Drula E, Lombard V, Coutinho PM, Henrissat B. 2013. Expansion of the

638 enzymatic repertoire of the CAZy database to integrate auxiliary redox enzymes.

639 Biotechnol Biofuels 6:41.

640 5. Vaaje-Kolstad G, Westereng B, Horn SJ, Liu Z, Zhai H, Sorlie M, Eijsink VG. 2010. An

641 oxidative enzyme boosting the enzymatic conversion of recalcitrant polysaccharides.

642 Science 330:219-222. D o

643 6. Quinlan RJ, Sweeney MD, Lo Leggio L, Otten H, Poulsen JC, Johansen KS, Krogh KB, w -i

644 Jorgensen CI, Tovborg M, Anthonsen A, Tryfona T, Walter CP, Dupree P, Xu F, —i o

645 Davies GJ, Walton PH. 2011. Insights into the oxidative degradation of cellulose by a 03 CL

646 copper metalloenzyme that exploits biomass components. Proc Natl Acad Sci U S A CD CL

647 108:15079-15084. ^ —5

648 7. Beeson WT, Phillips CM, Cate JH, Marletta MA. 2012. Oxidative cleavage of cellulose o

649 by fungal copper-dependent polysaccharide monooxygenases. J Am Chem Soc ZJ h

650 134:890-892. r—t-r—t-

651 8. Johansen KS. 2016. Discovery and industrial applications of lytic polysaccharide

652 mono-oxygenases. Biochem Soc Trans 44:143-149. CD

653 9. Phillips CM, Beeson WT, Cate JH, Marletta MA. 2011. Cellobiose dehydrogenase and

654 a copper-dependent polysaccharide monooxygenase potentiate cellulose CO

655 degradation by Neurospora crassa. ACS Chem Biol 6:1399-1406.

656 10. Langston JA, Shaghasi T, Abbate E, Xu F, Vlasenko E, Sweeney MD. 2011. o —i

657 Oxidoreductive cellulose depolymerization by the enzymes cellobiose dehydrogenase /

658 and glycoside hydrolase 61. Appl Environ Microbiol 77:7007-7015. o

659 11. Kracher D, Scheiblbrandner S, Felice AK, Breslmayr E, Preims M, Ludwicka K, O

660 Haltrich D, Eijsink VG, Ludwig R. 2016. Extracellular electron transfer systems fuel o r—t-

661 cellulose oxidative degradation. Science 352:1098-1101. o CT

662 12. Garajova S, Mathieu Y, Beccia MR, Bennati-Granier C, Biaso F, Fanuel M, Ropartz D, CD —i

663 Guigliarelli B, Record E, Rogniaux H, Henrissat B, Berrin JG. 2016. Single-domain —

664 flavoenzymes trigger lytic polysaccharide monooxygenases for oxidative degradation 2 o

665 of cellulose. Sci Rep 6:28276. — b

666 13. Agger JW, Isaksen T, Varnai A, Vidal-Melgosa S, Willats WGT, Ludwig R, Horn SJ,

667 Eijsink VGH, Westereng B. 2014. Discovery of LPMO activity on hemicelluloses shows -

668 the importance of oxidative processes in plant cell wall degradation. Proc Natl Acad CD

669 Sci U S A 111:6287-6292.

670 14. Bennati-Granier C, Garajova S, Champion C, Grisel S, Haon M, Zhou S, Fanuel M, O

671 Ropartz D, Rogniaux H, Gimbert I, Record E, Berrin JG. 2015. Substrate specificity o —i

672 and regioselectivity of fungal AA9 lytic polysaccharide monooxygenases secreted by C

673 Podospora anserina. Biotechnol Biofuels 8:90. M

674 15. Borisova AS, Isaksen T, Dimarogona M, Kognole AA, Mathiesen G, Varnai A, R0hr CD n

675 AK, Payne CM, Sorlie M, Sandgren M, Eijsink VG. 2015. Structural and functional O

676 characterization of a lytic polysaccharide monooxygenase with broad substrate C

677 specificity. J Biol Chem 290:22955-22969. i b

678 16. Jagadeeswaran G, Gainey L, Prade R, Mort AJ. 2016. A family of AA9 lytic VJ 0)

679 polysaccharide monooxygenases in Aspergillus nidulans is differentially regulated by y

680 multiple substrates and at least one is active on cellulose and xyloglucan. Appl

681 Microbiol Biotechnol 100:4535-4547.

682 17. Frommhagen M, Sforza S, Westphal AH, Visser J, Hinz SW, Koetsier MJ, van Berkel

683 WJ, Gruppen H, Kabel MA. 2015. Discovery of the combined oxidative cleavage of

plant xylan and cellulose by a new fungal polysaccharide monooxygenase. Biotechnol Biofuels 8:101.

Vu VV, Beeson WT, Span EA, Farquhar ER, Marletta MA. 2014. A family of starch-active polysaccharide monooxygenases. Proc Natl Acad Sci U S A 111:13822-13827. Lo Leggio L, Simmons TJ, Poulsen JC, Frandsen KE, Hemsworth GR, Stringer MA, von Freiesleben P, Tovborg M, Johansen KS, De Maria L, Harris PV, Soong CL, Dupree P, Tryfona T, Lenfant N, Henrissat B, Davies GJ, Walton PH. 2015. Structure and boosting activity of a starch-degrading lytic polysaccharide monooxygenase. Nat Commun 6:5961.

Yelle DJ, Ralph J, Lu F, Hammel KE. 2008. Evidence for cleavage of lignin by a brown

rot basidiomycete. Environ Microbiol 10:1844-1849.

Arantes V, Jellison J, Goodell B. 2012. Peculiarities of brown-rot fungi and

biochemical Fenton reaction with regard to their potential as a model for

bioprocessing biomass. Appl Microbiol Biotechnol 94:323-338.

Arantes V, Goodell B. 2014. Current understanding of brown-rot fungal

biodegradation mechanisms: a review, p 3-21, Deterioration and Protection of

Sustainable Biomaterials, vol 1158. American Chemical Society.

Halliwell G. 1965. Catalytic decomposition of cellulose under biological conditions.

Biochem J 95:35-40.

Koenigs JW. 1974. Hydrogen peroxide and iron: a proposed system for decomposition of wood by brown-rot basidiomycetes. Wood Fiber Sci 6:66-80. Goodell B, Jellison J, Liu J, Daniel G, Paszczynski A, Fekete F, Krishnamurthy S, Jun L,

Xu G. 1997. Low molecular weight chelators and phenolic compounds isolated from wood decay fungi and their role in the fungal biodegradation of wood. J Biotechnol 53:133-162.

Hammel KE, Kapich AN, Jensen Jr KA, Ryan ZC. 2002. Reactive oxygen species as agents of wood decay by fungi. Enzyme Microb Technol 30:445-453. Horn SJ, Sikorski P, Cederkvist JB, Vaaje-Kolstad G, Sorlie M, Synstad B, Vriend G, Varum KM, Eijsink VG. 2006. Costs and benefits of processivity in enzymatic degradation of recalcitrant polysaccharides. Proc Natl Acad Sci U S A 103:1808918094.

Eastwood DC, Floudas D, Binder M, Majcherczyk A, Schneider P, Aerts A, Asiegbu FO, Baker SE, Barry K, Bendiksby M, Blumentritt M, Coutinho PM, Cullen D, de Vries RP, Gathman A, Goodell B, Henrissat B, Ihrmark K, Kauserud H, Kohler A, LaButti K, Lapidus A, Lavin JL, Lee YH, Lindquist E, Lilly W, Lucas S, Morin E, Murat C, Oguiza JA, Park J, Pisabarro AG, Riley R, Rosling A, Salamov A, Schmidt O, Schmutz J, Skrede I, Stenlid J, Wiebenga A, Xie X, Kues U, Hibbett DS, Hoffmeister D, Hogberg N, Martin F, Grigoriev IV, Watkinson SC. 2011. The plant cell wall-decomposing machinery underlies the functional diversity of forest fungi. Science 333:762-765. Floudas D, Binder M, Riley R, Barry K, Blanchette RA, Henrissat B, Martinez AT, Otillar R, Spatafora JW, Yadav JS, Aerts A, Benoit I, Boyd A, Carlson A, Copeland A, Coutinho PM, de Vries RP, Ferreira P, Findley K, Foster B, Gaskell J, Glotzer D, Gorecki P, Heitman J, Hesse C, Hori C, Igarashi K, Jurgens JA, Kallen N, Kersten P, Kohler A, Kues U, Kumar TK, Kuo A, LaButti K, Larrondo LF, Lindquist E, Ling A, Lombard V, Lucas S, Lundell T, Martin R, McLaughlin DJ, Morgenstern I, Morin E, Murat C, Nagy LG, Nolan M, Ohm RA, Patyshakuliyeva A, et al. 2012. The Paleozoic


03 C/J

o »o cr CD

CD i O' 03

730 origin of enzymatic lignin decomposition reconstructed from 31 fungal genomes.

731 Science 336:1715-1719.

732 30. Westereng B, Cannella D, Wittrup Agger J, J0rgensen H, Larsen Andersen M, Eijsink

733 VGH, Felby C. 2015. Enzymatic cellulose oxidation is linked to lignin by long-range

734 electron transfer. Sci Rep 5:18561.

735 31. Cannella D, Mollers KB, Frigaard NU, Jensen PE, Bjerrum MJ, Johansen KS, Felby C. D o

736 2016. Light-driven oxidation of polysaccharides by photosynthetic pigments and a w

737 metalloenzyme. Nat Commun 7:11134. o

738 32. Jung S, Song Y, Kim HM, Bae HJ. 2015. Enhanced lignocellulosic biomass hydrolysis 03 Q.

739 by oxidative lytic polysaccharide monooxygenases (LPMOs) GH61 from Gloeophyllum CD CL

740 trabeum. Enzyme Microb Technol 77:38-45. ^ —i

741 33. Westereng B, Agger JW, Horn SJ, Vaaje-Kolstad G, Aachmann FL, Stenstrom YH, O

742 Eijsink VG. 2013. Efficient separation of oxidized cello-oligosaccharides generated by h

743 cellulose degrading lytic polysaccharide monooxygenases. J Chromatogr A 1271:144- r—t-r—t- "O

744 152.

745 34. Westereng B, Arntzen MO, Aachmann FL, Varnai A, Eijsink VG, Agger JW. 2016. CD

746 Simultaneous analysis of C1 and C4 oxidized oligosaccharides, the products of lytic 3

747 polysaccharide monooxygenases acting on cellulose. J Chromatogr A 1445:46-54.

748 35. Vidal-Melgosa S, Pedersen HL, Schuckel J, Arnal G, Dumon C, Amby DB, Monrad RN, 3

749 Westereng B, Willats WG. 2015. A new versatile microarray-based method for high o —i

750 throughput screening of carbohydrate-active enzymes. J Biol Chem 290:9020-9036. /

751 36. Eibinger M, Ganner T, Bubner P, Rosker S, Kracher D, Haltrich D, Ludwig R, Plank H, o

752 Nidetzky B. 2014. Cellulose surface degradation by a lytic polysaccharide O

753 monooxygenase and its effect on cellulase hydrolytic efficiency. J Biol Chem o r—t-

754 289:35929-35938. o cr

755 37. Kittl R, Kracher D, Burgstaller D, Haltrich D, Ludwig R. 2012. Production of four CD —i

756 Neurospora crassa lytic polysaccharide monooxygenases in Pichia pastoris monitored C

757 by a fluorimetric assay. Biotechnol Biofuels 5:79. 2 o

758 38. Varnai A, Tang C, Bengtsson O, Atterton A, Mathiesen G, Eijsink VG. 2014. CD b

759 Expression of endoglucanases in Pichia pastoris under control of the GAP promoter.

760 Microb Cell Fact 13:57. C

761 39. Katoh K, Misawa K, Kuma K, Miyata T. 2002. MAFFT: a novel method for rapid l

762 multiple sequence alignment based on fast Fourier transform. Nucleic Acids Res

763 30:3059-3066. O

764 40. Galtier N, Gouy M, Gautier C. 1996. SEAVIEW and PHYLO_WIN: two graphic tools for o —i

765 sequence alignment and molecular phylogeny. Comput Appl Biosci 12:543-548. l

766 41. Saitou N, Nei M. 1987. The neighbor-joining method: a new method for M

767 reconstructing phylogenetic trees. Mol Biol Evol 4:406-425. CD o

768 42. Thompson JD, Gibson TJ, Plewniak F, Jeanmougin F, Higgins DG. 1997. The O

769 CLUSTAL_X windows interface: flexible strategies for multiple sequence alignment l

770 aided by quality analysis tools. Nucleic Acids Res 25:4876-4882. i b

771 43. Umezawa K, Takeda K, Ishida T, Sunagawa N, Makabe A, Isobe K, Koba K, Ohno H, w 0)

772 Samejima M, Nakamura N, Igarashi K, Yoshida M. 2015. A novel pyrroloquinoline y

773 quinone-dependent 2-keto-D-glucose dehydrogenase from Pseudomonas

774 aureofaciens. J Bacteriol 197:1322-1329.

775 44. Highley TL. 1973. Influence of carbon source on cellulase activity of white-rot and

776 brown-rot fungi. Wood Fiber Sci 5:50-58.

777 45. Wood TM. 1988. Preparation of crystalline, amorphous, and dyed cellulase

778 substrates, p 19-25, Methods in Enzymology, vol 160. Academic Press.

779 46. Loose JS, Forsberg Z, Fraaije MW, Eijsink VG, Vaaje-Kolstad G. 2014. A rapid

780 quantitative activity assay shows that the Vibrio cholerae colonization factor GbpA is

781 an active lytic polysaccharide monooxygenase. FEBS Lett 588:3435-3440.

782 47. Petersen TN, Brunak S, von Heijne G, Nielsen H. 2011. SignalP 4.0: discriminating O o

783 signal peptides from transmembrane regions. Nat Methods 8:785-786. w -i

784 48. Li X, Beeson WTt, Phillips CM, Marletta MA, Cate JH. 2012. Structural basis for —i o

785 substrate targeting and catalysis by fungal polysaccharide monooxygenases. 0) Q.

786 Structure 20:1051-1061. CD o

787 49. Vu VV, Beeson WT, Phillips CM, Cate JHD, Marletta MA. 2014. Determinants of ^ —i

788 regioselective hydroxylation in the fungal polysaccharide monooxygenases. J Am O - h

789 Chem Soc 136:562-565.

790 50. Isaksen T, Westereng B, Aachmann FL, Agger JW, Kracher D, Kittl R, Ludwig R, »t r—t--jj

791 Haltrich D, Eijsink VG, Horn SJ. 2014. A C4-oxidizing lytic polysaccharide

792 monooxygenase cleaving both cellulose and cello-oligosaccharides. J Biol Chem CD

793 289:2632-2642.

794 51. Wu M, Beckham GT, Larsson AM, Ishida T, Kim S, Payne CM, Himmel ME, Crowley CO

795 MF, Horn SJ, Westereng B, Igarashi K, Samejima M, Stahlberg J, Eijsink VG,

796 Sandgren M. 2013. Crystal structure and computational characterization of the lytic o —i

797 polysaccharide monooxygenase GH61D from the Basidiomycota fungus /

798 Phanerochaete chrysosporium. J Biol Chem 288:12828-12839. o

799 52. Hemsworth GR, Johnston EM, Davies GJ, Walton PH. 2015. Lytic Polysaccharide O

800 Monooxygenases in Biomass Conversion. Trends Biotechnol 33:747-761. o r—t-

801 53. York WS, van Halbeek H, Darvill AG, Albersheim P. 1990. Structural analysis of o CT

802 xyloglucan oligosaccharides by 1H-n.m.r. spectroscopy and fast-atom-bombardment CD -

803 mass spectrometry. Carbohydr Res 200:9-31. J-1,

804 54. Niemann C, Carpita NC, Whistler RL. 1997. Arabinose-containing oligosaccharides 2 o

805 from Tamarind xyloglucan. Starke 49:154-159. — b

806 55. Courtade G, Wimmer R, Rohr AK, Preims M, Felice AK, Dimarogona M, Vaaje-

807 Kolstad G, Sorlie M, Sandgren M, Ludwig R, Eijsink VG, Aachmann FL. 2016. -

808 Interactions of a fungal lytic polysaccharide monooxygenase with beta-glucan CD.

809 substrates and cellobiose dehydrogenase. Proc Natl Acad Sci U S A

810 doi:10.1073/pnas.1602566113. O

811 56. Karlsson J, Saloheimo M, Siika-Aho M, Tenkanen M, Penttila M, Tjerneld F. 2001. o —i

812 Homologous expression and characterization of Cel61A (EG IV) of Trichoderma reesei. C

813 Eur J Biochem 268:6498-6507. M

814 57. Shepherd MG, Tong CC, Cole AL. 1981. Substrate specificity and mode of action of e n

815 the cellulases from the thermophilic fungus Thermoascus aurantiacus. Biochem J O

816 193:67-74. C

817 58. Spier VC, Sierakowski MR, Ibrahim AT, Scholze Baum JC, Silveira JL, de Freitas RA. i b

818 2015. Time-dependent viscometry study of endoglucanase action on xyloglucan: A VJ 0)

819 real-time approach. Int J Biol Macromol 81:461-466. y

820 59. Vanden Wymelenberg A, Gaskell J, Mozuch M, Sabat G, Ralph J, Skyba O, Mansfield

821 SD, Blanchette RA, Martinez D, Grigoriev I, Kersten PJ, Cullen D. 2010. Comparative

822 transcriptome and secretome analysis of wood decay fungi Postia placenta and

823 Phanerochaete chrysosporium. Appl Environ Microbiol 76:3599-3610.

Hori C, Gaskell J, Igarashi K, Samejima M, Hibbett D, Henrissat B, Cullen D. 2013. Genomewide analysis of polysaccharides degrading enzymes in 11 white- and brown-rot Polyporales provides insight into mechanisms of wood decay. Mycologia 105:1412-1427.

Cosgrove DJ. 2005. Growth of the plant cell wall. Nat Rev Mol Cell Biol 6:850-861. Donaldson LA, Knox JP. 2012. Localization of cell wall polysaccharides in normal and compression wood of radiata pine: relationships with lignification and microfibril orientation. Plant Physiol 158:642-653.

Park YB, Cosgrove DJ. 2015. Xyloglucan and its interactions with other components of the growing cell wall. Plant Cell Physiol 56:180-194.

Wilcox WW. 1978. Review of literature on the effects of early stages of decay on wood strength. Wood Fiber Sci 9:252-257.

Westereng B, Ishida T, Vaaje-Kolstad G, Wu M, Eijsink VG, Igarashi K, Samejima M, Stahlberg J, Horn SJ, Sandgren M. 2011. The putative endoglucanase PcGH61D from Phanerochaete chrysosporium is a metal-dependent oxidative enzyme that cleaves cellulose. PLoS One 6:e27807.

Forsberg Z, Mackenzie AK, Sorlie M, Rohr AK, Helland R, Arvai AS, Vaaje-Kolstad G, Eijsink VG. 2014. Structural and functional characterization of a conserved pair of bacterial cellulose-oxidizing lytic polysaccharide monooxygenases. Proc Natl Acad Sci U S A 111:8446-8451.

03 Q. CD Q.

03 C/J


o cr CD

846 Figure legends

847 Figure 1. Position of G. trabeum LPMOs in the phylogenetic tree of LPMO9s. The

848 phylogenetic tree was created from catalytic domains of LPMO9s by using the neighbor-

849 joining method in ClustalX (v. 2.1). Several LPMOs are labeled and their cleavage specificities

850 are indicated (C1, C4, or both, C1/C4; N. D. means not determined). For all labeled non-

851 GtLPMOs, activity on cellulose has been demonstrated. An * indicates that additional activity

852 on xyloglucan has been detected. The functional data for GtLPMO9A-2 stem from the

853 present report; other functional data come from references (13, 14, 16, 49, 65).

855 Figure 2. Multiple sequence alignment of the catalytic domains of selected LPMO9s. Fully

856 conserved residues are printed in white on a black background. Active site histidines (black-

857 filled triangles) and a tyrosine (gray-filled triangle) involved in copper coordination are

858 indicated. The labeled bars over the sequences indicate known variable regions in LPMO9s

859 (15, 51). Note that GtLPMO9A-2 and GtLPMO9D have C-terminal extensions; for more

860 information see Figs. S2-S4. This alignment was generated using MAFFT, version 7.295 (39),

861 available at the European Bioinformatics Institute website

862 (

864 Figure 3. Products generated by GtLPMO9A-2 or NcLPMO9C on PASC. (A) HPAEC-PAD

865 chromatograms showing cello-oligosaccharides released by GtLPMO9A-2 (blue) and

866 WcLPMO9C (red) from PASC. Peaks were assigned based on previous assignments by Isaksen

867 et al. (50); native cello-oligosaccharides are labeled as Glcn where n is the degree of

868 polymerization (DP). Note that it has recently been shown that C4-oxidized products are

Ш Q. CD Q.




CD i О Ш

869 unstable under these chromatographic conditions and that the peaks labeled C4 are, in fact,

870 diagnostic degradation products (34). The fractions of native products are high because C4-

871 oxidized products tend to loose the oxidized monosugar under these chromatographic

872 conditions. (B) MALDI-ToF spectrum of cello-oligosaccharides released by GtLPMO9A-2 from

873 PASC, where the inset shows details for the heptamer ion cluster (sodium adducts only).

874 Possible products in these clusters are the native Glc7 (m/z 1175.8), the Cl-oxidized lactone

875 or C4-oxidized ketoaldose (anhydrated species, m/z 1173.8), the Cl-oxidized aldonic acid or

876 C4-oxidized gemdiol (hydrated species, m/z 1191.8), and the sodium adduct of the aldonic

877 acid sodium salt (1213.8). The double-oxidized species corresponds to m/z 1171.8

878 (anhydrated form, lactone-ketoaldose species), 1189.8 (hydrated form), and 1211.8 (the

879 sodium salt of the sodium adduct). The presence of sodium salts is diagnostic for C1-

880 oxidation, since only this oxidation yields an aldonic acid. The strong signals for dehydrated

881 oxidized species in MALDI-ToF are diagnostic for C4-oxidation. The 4-keto to gemdiol

882 equilibrium is less skewed towards the hydrated form compared to the lactone - aldonic

883 acid equilibrium, and, besides, C4-oxidized products are more efficiently dehydrated during

884 spotting on MALDI sample plates (see Forsberg et al. (66) for further discussion).

885 Experiments with Avicel yielded similar product patterns (data not shown).

887 Figure 4. Assessing LPMO-activity on hemicelluloses with dynamic viscosity experiments.

888 (a, c, e, g) GtLPMO9A-2 or (b, d, f, h) WcLPMO9C were incubated with (a, b) 0.15% (w/v)

889 xyloglucan, (c, d) 0.05% (w/v) glucomannan, (e, f) 0.2% (w/v) arabinoxylan, and (g, h) 0.5%

890 (w/v) carboxymethylcellulose in the presence (red lines) or absence (green lines) of DTT as a

891 reducing agent. Reactions with only DTT and no LPMO are shown by blue lines.

03 Q. 0 Q.

03 C/J

o »o cr 0

—5 0

0 i o'

893 Figure 5. HPAEC-PAD analysis of reaction products generated from xyloglucan (XG) by

894 GtLPMO9A-2 and WcLPMO9C in the dynamic viscosity experiments. Samples were taken

895 after 16 hours of incubation and the chromatograms show the product profiles before (A)

896 and after (B) a subsequent treatment with AfCel12A. Green line: XG7 standard (XXXG); grey

897 line: XG-oligo standard (this is a mixture of shorter XG oligomers with DP in the range of 14898 27). Black line: XG incubated with DTT only. Red lines: products generated from XG with

899 WcLPMO9C in the presence (dark red) or absence (light red) of DTT. Blue lines: products

900 generated from XG with GtLPMO9A-2 in the presence (dark blue) or absence (light blue) of

901 DTT. Grey dashed lines indicate oligosaccharides the concentration of which was lower as a

902 result of LPMO activity. The reaction conditions are specified in the Materials and Methods

903 section. Nb. Fig. 7 provides an even clearer example of the difference between WcLPMO9C

904 and GtLPMO9A-2 with respect to the activity on xyloglucan.

906 Figure 6. MALDI-ToF MS analysis of the reaction products generated by (A) GtLPMO9A-2 and

907 (B) WcLPMO9C during the dynamic viscosity experiments with tamarind xyloglucan (XG) in

908 the presence of reducing agent (DTT). The samples analyzed were the end-point samples (16

909 hours). The reaction conditions are specified in the Materials and Methods and were similar

910 to the conditions used for generating Fig. 5. GtLPMO9A-2 generated a wide range of

911 oligosaccharides with all possible combination of hexose (Hex) and pentose (Pen) units,

912 whereas WcLPMO9C generated clusters of oligosaccharides with the number of pentose

913 residues being a multiple of three. In both spectra, oxidized species (-2, as compared with

914 the mass of native species) are marked with #; species with masses indicating

915 arabinosylation are labeled green (see text). Note that most labeled peaks in fact are a

Q. CD Q.




К) О

cluster of signals; see Fig. S10 and text. All labeled species are Na+-adducts. No species were detected below m/z 1000.

Figure 7. HPAEC-PAD analysis of the reaction products generated by (A) GtLPMO9A-2 or (B) WcLPMO9C on xyloglucan-coated PASC. Green line: native cello-oligosaccharides with DP 26; yellow line: XG7 standard (XXXG); grey line: XG-oligo standard with DP 14-27. Brown line: mixture of PASC and XG incubated with LPMO only. Black line: mixture of PASC and XG incubated with ascorbic acid (ASC) only. Blue line: products generated from PASC with LPMO in the presence of ASC. Red line: products generated from XG with LPMO in the presence of ASC. Purple line: products generated from xyloglucan-coated PASC with LPMO in the presence of ASC. The reaction conditions are specified in the Materials and Methods section. Note that the chromatographic conditions were similar to those used in Fig. 5 but different from those used in Fig. 3. This explains why the retention times of cello-oligomers differ between Fig. 3 and Fig. 5.

CD Q. 0 Q.

03 C/J

o »o cr 0

—5 0

0 i o'

931 Tables

933 Table 1. Oligonucleotide primers.


Sequences (5'-sequence-3'



Gt LPMO9C_3'RACE_forward



Gt LPMO9A_O RF_forward











GtLPMO9B_expression_forwa rd
































The restriction sites Xhol (CTCGAG; forward expression primers) and NotI (GCGGCCGC; reverse expression primers; only partial in the GtLPMO9A-2 primer) and the sequence encoding the cleavage site of Kex2 (AAAAGA, encoding Lys-Arg; forward primers) are printed in italics. ORF sequences are underlined. Expression primers were designed with a

CD Q. 0 Q.



o cr 0

—5 0

protecting AA sequence at the 5' end. GtLPMO9A-2 expression primers, for In-Fusion cloning, were designed to contain 15 extra bases at the 5' end, overlapping with the Kex2 (forward) or NotI restriction cleavage site (reverse) of pPICZa.



O f—t-O cr CD

—5 0

Table 2. Parameters obtained after fitting the dynamic viscosity curves. The correlation between dynamic viscosity and time (data from Fig. 4) was fitted to an exponential (Exp.) curve (y=a-e(b'x)+c) or to a linear (Lin.) equation (y=a\x+b), where y is the dynamic viscosity (in mP), x is the time (in ks), and a, b, and c are constants. The constant c is the final viscosity of substrate after completion of LPMO9 action. The initial depolymerization rate (IDR) was calculated as the first derivative for the formula at x=0. The high R-square values (>0.97) reflect a good fit of the model.


Substrate Enzyme Fit Parameters a (mP) b (ks-1) c (mP) R-square IDR (mP^ks-1)

XG GtLPMO9A-2 Exp. 0.5748 -0.1514 0.9812 0.9927 0.0870

NcLPMO9C Exp. 0.4633 -0.1067 0.9947 0.9950 0.0495

GM GtLPMO9A-2 Exp. 0.7115 -0.07416 1.0496 0.9966 0.0528

NcLPMO9C Exp. 0.3907 -0.1537 0.9367 0.9832 0.0600

WAX GtLPMO9A-2 Lin. -0.001454 1.7106 - 0.9936 0.001454

NcLPMO9C Lin. -0.001654 1.663 - 0.9731 0.001653

CMC GtLPMO9A-2 Exp. 0.3713 -0.01999 1.7082 0.9998 0.0074

NcLPMO9C Exp. 0.4455 -0.02732 1.6661 0.9991 0.0122

03 C/J

o »o cr CD

CD i O' 03

Л/с1_РМ09Р, С4 A/CLPM09A, С4

GfLPM09C, N.D.


PaLPM09H, С1/С4*

NcLPM09C, C4*

AN3046, N.D.*


PcLPM09D, Cl

A/CLPM09M, C1/C4

NCU07760, C1/C4

Gfl_PM09A-l, N.D. GfLPM09A-2, C1/C4* GtLPM09B, N.D.


















































t i/D-QGQLKG-----------V! \D-QGQLKG-----------1-

J 3E-TTPNQY-----------V-














¡l/TDVMS--------SDI] !




5 gy----

! MA----

5 M TN---

i MA----


/32 ß3



j AR ¡AK J ¡SI' A0D



















VnßgJlD V]?¡SID

V] i IÍ

LS 9 j SI

AK ?j SID KV ? ; Ян Le ! i 3Vi LK ? j aVi 106 GD ?j Я1Н 100 QlIiSlY

113 113 113 106 105

106 105




ь \L 3AETYPC Иоиа

¡I М. 3AETYPC 3qf


¡I W <AETYP( 3qf

¡I M. -GMTEGC Je f

¡I M 5AYSYPC 3qf

L rL 5ASVRG/ 3qf


fcv jAHD-GC iEF


176 176 176


16 8--------------p^eJ

16 9--------------Pi



17 0--------------Yl










j GV-TYDAYQAA------------





GtLPM09A-l 231 INS---



GtLPM09C 229 —NYTV l "1PAEG—


NcLPM09M 211 —TYTI j i PAVFTC-



AN3046 211 -KDYPM j i PAVYTC-





О сг CD

Applied and Environmental à


Applied and Environmental Microbiology

ccepted Mârïïjscrrpt HostecTOnimë

Relative intensity

00 o o

H -Glcfi

- Glc7

Relative intensity

- GICq

00 o o


/ 1173.8 X 1175.8

1189.8 X 1191.8 1211.8 1213.8

Signal (nC)

I—4 I—4 H* NJ

Ni>®(»0loj>01[»0 ooooooooooo

■t* o

■li U1

- Glc2

- Glc3 -Glc4

- Glee

AjBjqn |B0|p9|/\| ||9UJ0Q |||9AA Äq 9|,02 'l Jeqopo uo /ßJ0WSBW9B//:dHi| luoj^ p9pB0|UM0Q

со CL Е

со CL

сл ■>

Е со с

со CL

со о

о ■>

Е со с

1.6 1.4 1.2 1

2.4 2.2 2 1.8 1.6 1.4 1.2 1

10 20 30 40 50 Incubation time (ksec)

10 20 30 40 Incubation time (ksec)

0 10 20 30 40 50 Incubation time (ksec)

10 20 30 40 50 Incubation time (ksec)

ю о о

ел ■>

Е со с

со Q_

со о о

(п ■>

Е со с

1.8 г» 1.6^ 1.4 1.2 1

2.4 2.2 2 1.8 1.6 1.4 1.2 1

10 20 30 40 Incubation time (ksec)

10 20 30 40 Incubation time (ksec)

0 10 20 30 40 Incubation time (ksec)

10 20 30 40 Incubation time (ksec)



О f—t-О er CD

Longer Polymer peak

Time (min)

Longer XG-oligosaccharides

Time (min)

XG + StLPM09A-2 + DTT


XG + GtLPM09A-2




XG + GtLPM09A-2 + DTT + d/Cell2A ■ XG + NCLPM09C + DTT + AfCe I12A XG + GtLPM09A-2 + AfCe\12A XG + NcLPM09C + AJCe I12A XG + DTT + /\/Cell2A XG-oligo XG7


03 C/J

O f—t-o cr CD

—5 0

150 -,

!" 100

** ^ LT) иэ

m с с С с 1Л с

с ai ф Ф CL) с LT) ш ф

ф о. CL Û_ Q_ CD с с Q_

Û_ г^ 00 00 CD û_ at cu tH

из X X X X ai û_ о ü_ о тЧ

X ai ai a) ai X X

Ф X X X X Ф T-i fH Ф


о о ^H ГМ ф Ф

CT) it rvj ^ Ю (X) ГМ X X LO

r^ О Ю СП (N 00

о w с Г-s 00 СП о LO т-i

LT) CD т—1 tH тН ГМ TH о (N KD

i-H С o_ (N гм LH..... fN

XG + GfLPM09A-2 + DTT

> 50 -

lO ID CT1 *H

rr> m 00 oo ел гм о *н

гм IN

XG + /VcLPM09C + DTT




150 П

r-. ni 00 о


Cl гм r\i

<N 1Л ■tf


U5 ID rv r->

мммА -1

u c 250

"5 200

OD 150

u c 250

"5 200

00 150

Longer XG-oligosaccharides

Cl-oxidized cello-


C4-oxidized cello-oligosaccharides

Time (min)

Time (min)

PASC + XG + GtLPM09A-2 + ASC

XG + GtLPM09A-2 + ASC



PASC + XG + GtLPM09A-2


Cello-oligos DP 2-6


XG + A/cLPM09C + ASC





Cello-oligos DP 2-6


0) C/J

O o o er CD

CD i O' 0)