Structural Insights into the Substrate Promiscuity of a Laboratory

Subscriber access provided by University of Sunderland

Article

Structural insights into the substrate promiscuity of a laboratory-evolved peroxygenase Mercedes Ramirez-Escudero, Patricia Molina-Espeja, Patricia Gomez de Santos, Martin Hofrichter, Julia Sanz-Aparicio, and Miguel Alcalde ACS Chem. Biol., Just Accepted Manuscript • DOI: 10.1021/acschembio.8b00500 • Publication Date (Web): 30 Oct 2018 Downloaded from http://pubs.acs.org on October 31, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41

ACS Chemical Biology

Evolved UPO

Native UPO F311

I248

L311 V248

V75 I75 FeIII

FeIII

V57 A57

F67

L67

ACS Paragon Plus Environment

ACS Chemical Biology 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 31

1 1 2 3 4 5

Ref nº cb-2018-00500k-Revised version ACS-Chemical Biology Draft#25 01/10/2018

6

Structural insights into the substrate promiscuity of a

7

laboratory-evolved peroxygenase

8 9 10

Mercedes Ramirez-Escudero,1ǂ Patricia Molina-Espeja,2ǂ Patricia Gomez de Santos,2 Martin Hofrichter3, Julia Sanz-Aparicio1* and Miguel Alcalde2*

11 12

1Department

13

"Rocasolano", CSIC, Madrid, Spain.

14

2Department

of Biocatalysis, Institute of Catalysis, CSIC, Madrid, Spain.

15

3Department

of Bio- and Environmental Sciences, TU Dresden, International Institute

16

Zittau, Mark 23, 02763 Zittau, Germany.

of Crystallography & Structural Biology, Institute of Physical Chemistry

17 18

ǂThese

19

* Corresponding authors:

20

Miguel Alcalde, Institute of Catalysis, CSIC, Marie Curie 2 L10 28049, Madrid, Spain. E-

21

mail: [email protected];

22

Julia Sanz-Aparicio, Institute of Physical Chemistry "Rocasolano", CSIC, Serrano 117

23

28006, Madrid, Spain. E-mail: [email protected].

two authors contributed equally to this work

24 25

Keywords: peroxygenase, directed evolution, substrate promiscuity, peroxidative

26

activity, peroxygenative activity, heme channel.


Page 3 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


2 27

ABSTRACT

28

Due to their minimal requirements, substrate promiscuity and product selectivity,

29

fungal peroxygenases are now considered to be the jewel in the crown of C-H

30

oxyfunctionalization biocatalysts. In this work, the crystal structure of the first laboratory

31

evolved peroxygenase expressed by yeast was solved at a resolution of 1.5 Å. Notable

32

differences were detected between evolved and native peroxygenase from Agrocybe

33

aegerita, including the presence of a full N-terminus and a broader heme access channel

34

due to the mutations accumulated through directed evolution. Further mutagenesis and

35

soaking experiments with a palette of peroxygenative and peroxidative substrates

36

suggested dynamic trafficking through the heme channel as the main driving force for

37

the exceptional substrate promiscuity of peroxygenase. Accordingly, this study provides

38

the first structural evidences at an atomic level regarding the mode of substrate binding

39

for this versatile biocatalyst, which are discussed within a biological and chemical

40

context.

41 42 43 44 45 46 47 48 49 50 51



Page 4 of 31

3 52

INTRODUCTION

53

Selective oxyfunctionalizations of organic molecules are among the most

54

attractive transformations in synthetic chemistry (1). Given that the use of transition metal

55

catalysts to activate inert C-H bonds is strongly associated with poor selectivity and harsh

56

conditions (2), there remains hope of performing such reactions in milder conditions,

57

encouraged by the fact that such processes are carried out, as a matter of routine, within

58

biological networks. Indeed, for more than 30 years cytochrome P450 monooxygenases

59

have been considered as promising biocatalysts to achieve such a milestone in

60

contemporary chemistry (3,4). However, and despite their exquisite regio-, chemo- and

61

enantio-selectivity, the use of isolated P450s on an industrial scale is littered with

62

hurdles, which range from their dependence on expensive redox cofactors (e.g. NADPH)

63

and auxiliary flavoproteins, to a lack of stability in the face of the demanding industrial

64

conditions and their common association to membranes. More significantly, P450s suffer

65

of uncoupling reactions (i.e. the reducing power is diverted from the target reaction to

66

futile reduction reactions, the so-called oxygen dilemma which is linked to their inherent

67

catalytic mechanism) (5-8). Different enterprises have been undertaken to overcome

68

these limitations, engineering more efficient and self-sufficient P450s, for example

69

working through the peroxide shunt pathway by using H2O2 as the main oxygen source

70

and the final electron acceptor, just like an “artificial” peroxygenase (3,9) or more

71

recently, to lead their promiscuity towards non-natural chemistry (10). Besides, several

72

studies are arising with natural P450 peroxygenases (i.e. fueled by H2O2) that catalyze

73

the -selective hydroxylation of long fatty acids (11-13). Meanwhile, the first extracellular

74

natural peroxygenase was identified and isolated from the fungal kingdom in 2004,

75

specifically from the basidiomycete Agrocybe aegerita (14) and it was characterized at

76

the genome level in 2009 (15). Less than one decade later, over 4000 peroxygenase-

77

like gene sequences have been deposited in genomic databases, and fungal

78

peroxygenases are now considered by many scientists as the ideal biocatalyst for


Page 5 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


4 79

oxyfunctionalization chemistry (16-18). Because of its ‘chimeric’ properties (bringing

80

together

81

chloroperoxidase from Caldariomyces fumago -CPO-), it has been referred to as

82

unspecific peroxygenase (UPO) and recognized as the first member of a new sub-

83

subclass of oxidoreductases (EC 1.11.2.1). Through a hybrid catalytic mechanism, UPO

84

shows both peroxidative activity (one-electron oxidation, e.g. of phenolics) and more

85

notably, peroxygenative activity (oxygen-transferring two-electron oxidations with

86

peroxides as oxygen source) (Supplementary Scheme 1). The latter drives the

87

promiscuity of UPOs in a wide portfolio of transformations that includes aromatic, alkylic,

88

(cyclo)aliphatic and heterocyclic hydroxylations, aromatic and aliphatic olefin

89

epoxidations, ether cleavages (O-dealkylations), N-dealkylations, sulfoxidations, N-

90

oxidations, deacylations (C-C bond cleavages) and halide oxidations/halogenations (18–

91

21).

the

catalytic

attributes

of

P450s,

classical

heme-peroxidases

and

92

To adjust the catalytic properties of this enzyme and satisfy the needs of industry,

93

heterologous production of UPOs is essential. As such, we achieved the functional

94

expression (secretion) of an UPO from A. aegerita (AaeUPO) in Saccharomyces

95

cerevisiae and Pichia pastoris through several rounds of laboratory evolution (8,22,23).

96

With expression levels of 8 mg/L in S. cerevisiae and above 200 mg/L in P. pastoris in

97

bioreactor, the highly active and stable AaeUPO secretion variant (called PaDa-I) was

98

very recently subjected to new directed evolution enterprises aimed at producing, with

99

high efficiency, agrochemicals and relevant pharma compounds (24,25). Within the

100

same framework, the crystal structure of the native AaeUPO was published at a

101

resolution of 2.2 Å (26). It is a compact monomeric protein comprised of 10 -helices

102

and 5 short -sheets, and with a catalytic pocket carved into the center of the structure

103

through a cone-shaped channel (frustum) that is formed by hydrophobic-aromatic amino

104

acids. Buried at its apex, this channel has a protoporphyrin IX that carries a ferric iron,

105

constituting the heme prosthetic group. AaeUPO also has a cysteine residue (Cys36) as



Page 6 of 31

5 106

the axial (fifth) ligand of the heme, which situates UPOs along with CPO and P450s in

107

the category of the heme-thiolate enzymes (17).

108

Understanding the mode of substrate binding in UPO will help answer key

109

questions regarding the biological and chemical activity of this versatile biocatalyst. In

110

turn, this information will open the way to design “smarter” mutant libraries, combining

111

computational and experimental tools in a drive towards the generation of more efficient,

112

industrially competent peroxygenases (27). Here we present a thorough study of the

113

mode of substrate binding used by UPO with a complete set of peroxidative and

114

peroxygenative substrates. First, the evolved PaDa-I mutant was crystallized and its high

115

resolution structure was compared to that of the native AaeUPO. Significant differences

116

between the two templates were found, which were rationalized in further mutagenesis

117

experiments. Subsequently, a broad palette of peroxidative and peroxygenative

118

compounds was complexed with the PaDa-I mutant in order to obtain the first

119

experimental evidence of the patterns of substrate binding for this multipurpose

120

biocatalyst.

121

RESULTS

122

Structural analysis of the evolved peroxygenase

123

i) Crystallization of ligand-free mutant and general structure

124

The PaDa-I mutant is the outcome of 5 rounds of directed evolution to enhance

125

its activity and its heterologous functional expression in S. cerevisiae (22). Carrying 9

126

mutations (F12Y-A14V-R15G-A21D-V57A-L67F-V75I-I248V-F311L, the underlined

127

mutations lying in the signal peptide), this evolved recombinant peroxygenase is also

128

readily secreted by P. pastoris under the control of the AOX1 promoter in a highly stable

129

and active form, and at a good titer while preserving the properties obtained after

130

laboratory evolution in S. cerevisiae (23). PaDa-I has a strong degree of glycosylation

131

(30%, MW 51.1 kDa), which makes it difficult to obtain high quality crystals. In order to


Page 7 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


6 132

reduce the heterogeneity of the sample, the mutant produced by P. pastoris in a

133

bioreactor and purified to homogeneity (Reinheitszahl value, RzA418/A2802) was treated

134

with Endo H, and its MW decreased to 37 kDa (Supplementary Figure 1[a-b]).

135

Preliminary crystals of deglycosylated peroxygenase appeared with 20% PEG3350 and

136

0.2 M NH4NO3 (pH 6.3), and they were further optimized by seeding and buffering the

137

solution in 0.1 M sodium acetate (pH 4.0). These crystals diffracted to a maximum

138

resolution of 1.4 Å but they had an acetate anion bound to the active site, as also

139

occurred in the crystal of native AaeUPO in sodium acetate buffer (26). Since acetate is

140

a peroxygenase inhibitor, subsequent crystallization trials were carried out with

141

cacodylate, citrate and phosphate buffers, although nitrate was then found associated

142

with the active site. Finally, high quality peroxygenase crystals without any molecule

143

attached to the active site were obtained in 1.5-1.8 M Na/K phosphate buffer pH 5.6 with

144

3-10% MPD. Under such conditions, crystals of the enzyme with a ligand-free active site

145

were obtained at the highest resolution reported so far for a fungal peroxygenase, 1.5 Å

146

(a full description of the experimental and structural measurements are provided in Table

147

1, and in the Supporting Information, -Material and Methods section-).

148

The crystals belonged to the P21 space group, with one molecule per asymmetric

149

unit. Endo H cleaves oligosaccharide moieties to leave single N-acetylglucosamine

150

(GlcNAc) units, although poorly accessible glycosylation sites may have remained

151

partially glycosylated. Recombinant PaDa-I expressed in P. pastoris is more strongly

152

glycosylated (23) than native AaeUPO (30% vs. 22%, respectively), with six potential

153

glycosylation sites (as determined for the wild type enzyme). The electron density maps

154

allowed modeling of the the GlcNAc units and they were attached to Asn11, Asn141,

155

Asn161, Asn182, Asn286 and Asn295. We measured the kinetic parameters of PaDa-I

156

from P. pastoris (before and after deglycosylation (Supplementary Figure 1c) and we

157

found that the deglycosylated enzyme conserved its activity on both peroxidative and

158

peroxygenative substrates.



Page 8 of 31

7 159

The final structure of PaDa-I with a ligand-free active site is formed by the

160

polypeptide chain Glu1-Arg327 (harbouring the mutational backbone V57A-L67F-V75I-

161

I248V-F311L) and it has an overall structural arrangement similar to that described for

162

native AaeUPO, (Figure 1). PaDa-I is composed of 10 -helices linked by long loops.

163

According to SCOP2 (28), PaDa-I belongs to the all  class of common heme containing

164

proteins, with the heme cofactor deeply embedded in the helices and surrounded by

165

loops. The evolved peroxygenase includes the protoporphyrin IX (heme group) with the

166

iron hexacoordinated, with its fifth position associated to the characteristic proximal

167

Cys36 (axial) ligand and the sixth position (distal ligand) to a water molecule. The

168

Glu196-Arg189 forms the distinctive acid-base pair required for binding and heterolytic

169

cleavage of peroxide, which leads to enzyme activation via consecutive formation of the

170

catalytic intermediates Compound 0 (ferric peroxo-complex) and Compound I (ferryl oxo-

171

complex), respectively. Compound I is he reactive key intermediate that can follow two

172

different pathways depending on the particular substrate: i) ‘two-electron oxidation’ of

173

one substrate molecule along with oxygen atom transfer (peroxygenative activity) or ii)

174

two ‘one-electron oxidations’ of two substrate molecules resulting in the formation the of

175

the corresponding substrate radicals (e.g. peroxy radicals; peroxidative activity) (17),

176

(Supplementary Scheme 1). The heme access channel is upholstered by up to ten

177

aromatic residues (nine Phe and one Tyr), of which Phe69, Phe121 and Phe199 conform

178

a triad that orients the substrate towards the heme, whereas Phe76 and Phe191 delimit

179

the entrance of the channel (26). As in the native AaeUPO, the heme channel is a 17 Å

180

long deep funnel (conical frustum), with outer and inner diameters of 10 and 8.5 Å,

181

respectively. PaDa-I has a structural magnesium (Mg2+, green sphere) – that may

182

stabilize the porphyrin ring system – with the same coordination as that reported

183

previously for native AaeUPO (26). Also, similar to wild-type AaeUPO and

184

chloroperoxidase (CPO, (30)), PaDa-I possesses at least one halide binding site near

185

the enzyme's channel entrance where, for example, chloride can bind (Figure 1b, red

186

sphere). A phosphate anion can also be found within a positively charged region of the


Page 9 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


8 187

PaDa-I crystals, like the sulphate found in crystals of the wild type enzyme. Furthermore,

188

several glycerol molecules from the cryoprotectant solution were bound to different

189

positions of the molecular surface.

190

ii) Structural differences between the evolved and native peroxygenases

191

As in the native AaeUPO, the C-terminus of the evolved PaDa-I is stabilized by

192

the Cys278-Cys319 disulphide bridge, while the N-terminus is markedly different: the

193

PaDa-I crystal maintains the complete N-terminal region, which, in the wild type enzyme,

194

undergoes proteolysis at Gly3-Leu4 (EPG/LPP) (26) (Fig. 1 [a-b]). This observation was

195

confirmed by N-terminal sequencing of PaDa-I expressed by S. cerevisiae and P.

196

pastoris, each with an identical N-terminus (EPGLPPPGPL). The lack of proteolysis at

197

the N-terminus of the recombinant PaDa-I may be due to the introduction of the V57A

198

and V75I mutations by directed evolution, although we cannot rule out the absence of

199

efficient proteases in the heterologous hosts. These substitutions are located in the inner

200

part of two helices that flank the N-terminal Pro5-Pro6 segment, and they produce a shift

201

of 1 Å at Pro5 that may modify the existing hydrophobic contacts, possibly impeding the

202

N-terminal processing.

203

Among other mutations added by directed evolution, the L67F mutation is buried

204

in a hydrophobic pocket defined by Phe204, Phe222 and Phe232, and it is located in the

205

vicinity of the heme group, making hydrophobic contact with one of its methylene groups.

206

While Phe67 retains the same pattern of interactions with the surrounding aromatic

207

residues, producing no evidence of structural changes, we cannot rule out electronic

208

effects in the protoporphyrin ring caused by the substitution. Finally, the I248V and F311L

209

mutations are placed in positions with the side-chains oriented towards the heme

210

channel, shaping the active site for ligand binding. Ile248 is in a buried hydrophobic

211

pocket far from the heme and its replacement by Val did not result in structural

212

modifications. By contrast, the F311L mutation produced the most significant changes in



Page 10 of 31

9 213

the structure of PaDa-I, along with the aforementioned complete N-terminus, (Figure

214

[2a-d]).

215

Phe311 lies 3.6 Å from Phe76, establishing a hydrophobic contact (as shown in

216

Figure 2b). Both Phe76 and Phe191 form a pair of protruding aromatic residues at the

217

entrance of the heme access channel that are supposed to guide the substrate to the

218

active site, also establishing hydrophobic contacts in the native AaeUPO. The change of

219

a bulky Phe to a smaller Leu at position 311 shifts the Phe76 side-chain but it preserves

220

the interaction with Leu311, while weakening the contact with Phe191 (Figure 2c). This

221

change possibly leaves the Phe191 more exposed to the solvent, with a dual

222

conformation observed for this residue (Figure 2a), as also evident in the soaking

223

experiments, (see below). Indeed, the modification provoked a marked switch in the

224

position of the Phe191 side-chain that reorients to establish a stabilizing interaction with

225

Phe274 at a distance of 3.7 Å (Figure 2c). The consequence is a larger entrance to the

226

heme channel, with a distance between Phe191 and Phe76 of 4.1 and 7.9 Å in native

227

AaeUPO and PaDa-I, respectively (Figure [2a, 2d]).

228

In order to assess the role played by F311L, we reverted this mutation by site-

229

directed mutagenesis and we characterized the purified reverted PaDa-I-rev mutant

230

biochemically. The kinetic thermostability of this mutant was not compromised, with the

231

same T50 value as PaDa-I (58 ºC). When the apparent steady-state kinetic parameters

232

for its peroxidative (2,2´-azino-bis(3-ethylbenzothiazoline-6-sulfonic acid) [ABTS] and

233

2,6-dimethoxyphenol [DMP]) and peroxygenative (veratryl alcohol [VA] and 5-nitro-1,3-

234

benzodioxole [NBD]) substrates were measured, PaDa-I-rev showed a 4- and 3-fold

235

enhancement in the kcat/Km for ABTS and VA, respectively, whereas the catalytic

236

efficiency for DMP and NBD were practically conserved (Table 2). Strikingly, the

237

secretion level was reduced by roughly 70% relative to PaDa-I, such that the F311L

238

mutation necessarily exerts a beneficial effect on folding and expression, while its


Page 11 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


10 239

influence on activity and selectivity may vary depending on the substrate and the reaction

240

catalyzed.

241

Given the plasticity of the heme access channel observed, with a dual

242

conformational state for Phe191 and a broader entrance after F311L mutation, we

243

explored other positions involved in such effects. The Ala316 at the entrance of the

244

funnel (Figure 2d) is located in a flexible loop (G314-G318) that also shows distinct

245

conformational states upon substrate binding (see below) and this seems to compromise

246

accessibility to the heme. Accordingly, this residue was studied by saturation

247

mutagenesis, and screening of the corresponding mutant library with ABTS and NBD

248

unveiled the JEd-I variant (carrying the A316P mutation). The kinetics of the purified JEd-

249

I were enhanced for all the substrates tested, with an improvement in the kcat/Km that

250

ranged from 6- to 1.5-fold, while its thermostability was conserved (Table 2). The change

251

of Ala to Pro would be expected to decrease the flexibility of this loop, although rigidity

252

is not usually related to improved activity. While the precise effect of this substitution may

253

be difficult to explain, these results do address the ease with which the malleability of the

254

heme access channel can be manipulated and its direct implication on the broad UPO

255

substrate specificity.

256

Substrate binding mode

257

PaDa-I mutant crystals were examined in soaking experiments with panels of

258

peroxidative (ABTS, DMP, 1-naphthol and 2-naphthol) and peroxygenative substrates

259

(naphthalene, propranolol, diclofenac, acetanilide, styrene, NBD, VA, benzyl alcohol -

260

BA-, and palmitic acid). Productive crystal complexes were obtained with many of the

261

substrates (DMP, 1-naphthol, naphthalene, propranolol, styrene and VA) at a catalytic

262

distance of 4 Å to the heme (Supplementary Figure 2).

263

crystallography of PaDa-I complexes can be found in the Supporting Information

264

Material and Methods section and in Table 1. The other substrates were associated

265

with low occupancy or a disordered active site, and they were not included in the


Full details on the


Page 12 of 31

11 266

analysis. As mentioned above, the F311L modification in the PaDa-I mutant enhanced

267

the accessibility of the substrates to the heme catalytic site by increasing the size of the

268

outer entrance to the channel from 8 to 12 Å (Figure [2a, 2d]). The switch in the Phe191

269

side-chain found in PaDa-I was also evident in all the complexes obtained, in some cases

270

with the dual conformation observed in the ligand-free PaDa-I crystal (Figures [2a, 3a,

271

3e]).

272

All the complexes obtained, including those with substrates for the peroxidase

273

reaction, showed a clear electron density within the heme channel that situates ligand

274

binding in van der Waals contact with the heme group, indicating that catalysis is mostly

275

performed by direct interaction with the Fe protoporphyrin IX cofactor (Figure [3a-i]).

276

Moreover, a water molecule is bound between the heme and the substrate, in a position

277

similar to that reported for the distal oxygen of the ferric hydroperoxo intermediate,

278

denoted as Compound 0 and representing the first step in the reaction pathway (29).

279

Consequently, the complexes provided features relevant to substrate recognition that

280

are necessary for the reaction to occur. Although the residues defining the channel are

281

mostly hydrophobic, the funnel is occupied by an extended net of water molecules that

282

is ordered through hydrogen bonds with the main chain of different residues, thereby

283

shaping the channel. In the soaking experiments, the ligands displaced some of these

284

water molecules upon binding. The solved structure of the complexes revealed that all

285

the substrates sandwich their aromatic ring between Phe121 and Phe199, mostly in a T-

286

interaction mode, also contacting the hydrophobic Phe69.

287

Figure 3a and 3b shows the bound acetate ion and DMSO, two strong

288

peroxygenase inhibitors, trapped from the medium and blocking the access of the

289

substrates to the active site. This phenomenon was reported previously for acetate

290

bound to wild type AaeUPO (26), and also for CPO (30) and the dimeric Marasmius

291

rotula peroxygenase (MroUPO, PDB code 5FUJ, unpublished, Piontek, K., Strittmatter,

292

E., Plattner, D.A., Institute of Organic Chemistry, University of Freiburg, Germany).


Page 13 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


12 293

Moreover, a glycerol molecule trapped from the cryoprotectant solution showed a similar

294

mode of binding in the crystals soaked into the beta blocker propranolol (Figure 3c). The

295

substrate was found at two positions within the channel, both far from the heme site.

296

In the structure deposited of MroUPO complexed with propranolol (PDB code

297

5FUK, unpublished, Piontek, K., Strittmatter, E., Plattner, D.A., Institute of Organic

298

Chemistry, University of Freiburg, Germany), the ligand is also bound at a position far

299

from the heme. However, in that case the substrate competes with a partially occupied

300

palmitic acid molecule that spans the catalytic site. Nevertheless, our crystals reveal that

301

UPO can allocate the substrate propranolol to ordered positions within the channel,

302

emulating the entrance/exit of the pathway towards the heme catalytic site. In this

303

respect, it is worth noting that we recently reported the in vitro evolution of UPO for the

304

selective hydroxylation of propranolol and, by computational analysis, we defined the

305

gradual diffusion of propranolol through the heme channel as the main factor for a correct

306

binding at a distance of 4 Å from the catalytic heme-ferryl oxygen (25).

307

The peroxygenative substrate naphthalene and its hydroxylation product 1-

308

naphthol (a peroxidative substrate itself) also bound at a position situated 4 Å from the

309

heme, with a water molecule in the middle that mimics the heme oxygen as indicated

310

above (Figure [3d-f]). The bicyclic aromatic ring is sandwiched between Phe121 and

311

Phe199 in a T-packing mode, with an additional strong hydrophobic interaction to Phe69

312

that maintains a highly conserved orientation with respect to the heme plane in all the

313

crystals analysed. However, some disorder is observed in the electron density of many

314

crystals at the second fused ring and in the positions of the hydroxyl groups, possibly

315

indicating dynamic non-specific binding. In this regard, when we consider two

316

independent complexes obtained after soaking in 1-naphthol for different times, one of

317

them contains two molecules of the substrate passing along the heme channel with a

318

well-defined density (Figure 3e, 1-naphthol complex I). In the other crystal there were

319

five molecules of the substrate occupying the channel (Figure 3f, 1-naphthol, complex



Page 14 of 31

13 320

II), although the position of the hydroxyl group could not be unambiguously assigned in

321

the molecule next to the heme, probably due to its dynamic binding. A similar binding

322

pattern was observed in DMP-soaked crystals (Figure 3g), which had the phenolic ring

323

orientated similarly by the three Phe residues. It should be noted that the complex of the

324

dye-decolorizing peroxidase (DyP) with this peroxidative substrate reported previously

325

had the DMP molecule bound at a position far from the heme site, in which electron

326

transfer occurs through a net of water molecules connecting the substrate to the heme

327

ion (31). By contrast, in our complex, the aromatic ring was clearly located 4 Å from the

328

heme and included the ferric water, which may indicate the alternative nature of DMP

329

both as a peroxidative and peroxygenative substrate. Indeed, it is reasonable to think

330

that, besides phenoxy radical formation, the ‘aromatic double bond’ between C4 and C5

331

in DMP is epoxidized, leading, after spontaneous rearrangement and re-aromatization,

332

to 2,6-dimethoxyhydroquinone (Figure 3g). Moreover, it should be noted that two

333

conformations of the catalytic Glu196 of the acid-base pair have been detected in this

334

complex, one of them showing steric hindrance with one of the DMP methoxy groups

335

(Figure 3g). The same dual conformation was detected in the 1-naphthol (complex I)

336

and in the VA complex crystal (Figure [3e, 3h]). This feature is possibly related to the

337

highly dynamic binding indicated above. Finally, the complexes with VA and styrene

338

displayed a similar binding mode, with the primary alcohol group and the vinyl tail located

339

close to the iron and substituting the heme water respectively (Figure [3h, 3i]). In an

340

earlier study, several crystals of cytochrome P450 BM3 treated with styrene presented

341

different modes of binding for this substrate depending on the soaking time (32). The

342

position observed in this study was similar to that previously considered as a low-affinity

343

productive binding mode. Consequently, these complexes might be mimicking the

344

corresponding products of the hydroxylation (oxidation of primary alcohols by UPO

345

proceeds via initial hydroxylation to form gem-diols that are in equilibrium with the

346

corresponding free aldehydes) and epoxidation reactions.


Page 15 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


14 347

DISCUSSION

348

Until very recently, peroxygenases were considered as a ‘hidden treasure trove’

349

within the fungal kingdom. Due to their exquisite C-H oxyfunctionalization chemistry,

350

these heme-thiolate enzymes naturally insert oxygen into unactivated aliphatic and

351

aromatic hydrocarbons, fed simply with hydrogen peroxide. As such, the last decade has

352

witnessed intense research into these enzymes, mostly focusing on AaeUPO, the first

353

enzyme of this clade identified. This extensive portfolio of reactions has recently been

354

opened up to industrial needs by evolving the first fungal laboratory peroxygenase. To

355

aid such developments, it is paramount to understand the pattern of UPO substrate

356

binding in order to be capable of exploring its substrate and catalytic promiscuity through

357

computational and directed evolution methods. Here, we investigated the substrate

358

binding of UPO by crystallizing, mutating and complexing our laboratory-evolved

359

peroxygenase with a palette of substrates of different chemical natures.

360

The PaDa-I variant possesses a unique combination of mutations (V57A-L67F-

361

V75I-I248V-F311L) that improves the functional expression in yeast roughly 1100-fold,

362

also provoking notable differences in activity (22). The distribution of such substitutions

363

in different regions was largely conservative (i.e. equivalent residues in terms of polarity

364

and charge) and they helped to maintain the general structure of the native enzyme.

365

Nevertheless, the presence of this mutational backbone meant that some relevant

366

differences were observed when comparing the crystals of PaDa-I to the wild type

367

AaeUPO, including a complete N-terminus, and a highly dynamic and broader heme

368

access channel.

369

Although PaDa-I shows rather unspecific and dynamic substrate trafficking, it

370

maintains a common mode of binding to the heme in the different crystal complexes

371

(Figure 4a). Irrespective of the ligands’ nature, they were all unequivocally sandwiched

372

between Phe121 and Phe199, also interacting strongly with Phe69. As such, and despite

373

the low specificity of these hydrophobic interactions, the triad of phenylalanines



Page 16 of 31

15 374

participate in the direct binding of the substrate to the heme. The dual conformation of

375

the catalytic Glu196 observed in some of the complexes (with DMP, 1-naphthol and VA)

376

may indicate a dynamic catalytic mechanism (with distinct induced fit and plasticity of the

377

UPO protein), although further experimental evidence will be needed to confirm this.

378

Small-chain residues like Ala73 and Thr192 can be found in the neighborhood of the Phe

379

triad, these accommodating the methoxy or hydroxyl groups of some of the ligands

380

through additional hydrophobic or polar contacts.

381

In terms of the highly dynamic substrate entry and trafficking observed, some of

382

the 1-naphthol and propranolol molecules were trapped within the heme channel in our

383

soaking experiments, and their similar and ordered positions suggest relevant

384

information regarding the diffusion of the substrates towards the heme. The side chains

385

of Phe76 and Phe191 protrude within the channel, and they are involved in strong

386

aromatic interactions that are essential to guide the substrate towards the catalytic site,

387

(Figure 4b). Hence, the dual conformational state observed in Phe191 seems to be

388

important in this process. Indeed, in a recent study of a peroxygenase mutant evolved to

389

synthesize human drug metabolites from propranolol, we also observed this striking

390

switch of Phe191 in molecular dynamics simulations, which displaced the -helix hosting

391

this residue and in turn, controlled substrate access to the heme (25). Conversely, it is

392

reasonable to think that this flexible residue may act as a “molecular hinge”, which

393

regulates the trafficking and residence time of substrates and products within the

394

channel, along with Phe76, an effect that seems to be paramount to control the enzyme’s

395

peroxygenative and peroxidative activity. It is also worth noting that there are a few more

396

hydrophobic residues exposed to the solvent at the outer entrance to the channel (i.e.

397

Val244, Phe274 and Pro277), which precisely position the trapped substrates through

398

direct hydrophobic interactions, (Figure 4c).

399

As indicated above, and despite its marked hydrophobicity, the channel is filled

400

with water molecules that are displaced by the ligands upon binding. These water


Page 17 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


16 401

molecules are mostly hydrogen bound to main-chain atoms, involving two regions of the

402

peroxygenase in particular: the Ser240-Asp245 and Gly314-Gly318 loops. The former

403

seems to be implicated in the pattern of substrate binding, since the G241D substitution

404

(from a previous directed evolution campaign aimed at naphthalene hydroxylation)

405

induced a shift in the Ser240-Asp245 loop, apparently displacing the -helix that harbors

406

the catalytic acid-base pair Glu196-Arg189 (24). The latter motif (314GVAAG318) shapes

407

the heme funnel and it is very flexible, with different conformations observed in the crystal

408

complexes (Figure 4c). After saturating Ala316 in this loop, a noticeable improvement

409

of both peroxidative and peroxygenative activities was observed in the JEd-I variant,

410

again reflecting the plasticity of this loop (Table 2). In summary, the strongly malleable

411

loops at the outer entrance of the heme channel, along with the dual conformation

412

observed for both the Phe191 side chain and the catalytic Glu196, highlight how dynamic

413

the overall binding is, the main driving force behind the substrate promiscuity of UPO

414

with simultaneous selectivity of these reactions.

415

CONCLUSIONS AND OUTLOOK

416

After more than ten years of research, we might now consider it inappropriate to

417

refer to fungal peroxygenase as an “unspecific” peroxygenase (UPO), particularly given

418

its strong selectivity for C-H oxyfunctionalization (in many cases more enantio- than

419

regio-selective). Indeed, this biocatalyst might now be best referred to as a

420

“promiscuous” peroxygenase. Although only six fungal UPOs have been isolated to date

421

and at least partially characterized -from Agrocybe aegerita (14), Coprinellus radians

422

(33), Marasmius rotula (34), Coprinopsis cinerea (recombinant UPO expressed in

423

Aspergillus oryzae) (35), Chaetomium globosum (36), and Marasmius wettsteinii (37),

424

new peroxygenases and peroxygenase-like-proteins are rapidly emerging from different

425

sources that open the way to engage in “peroxygenase-based” synthetic chemistry.

426

Protein engineering is a valuable technique to adapt peroxygenases to current needs.

427

Accordingly, a multidisciplinary approach that combines directed evolution with structural



Page 18 of 31

17 428

and computational analysis will help to develop ad-hoc peroxygenases for biotechnology

429

applications. However, the oxidative damage caused by peroxide in peroxygenases on

430

a preparative scale is a serious drawback and several approaches are being followed to

431

limit this effect. These include the use of enzyme cascade reactions with methanol as

432

the sacrificial electron donor for the reductive activation of O2, or more recently, the

433

combination of evolved UPOs with inorganic photocatalysts (38-40).

434

ACKNOWLEDGEMENTS

435

This work was supported by the European Union [FP7-KBBE-2013-7-613549-

436

INDOX;

437

201580E042], and the Spanish Ministry of Economy, Industry and Competitiveness

438

[projects LIGNOLUTION, BIO2013-48779-C4-2-R and BIO2016-76601-C3-3-R]. We

439

also thank the Synchrotron Radiation Source at Alba (Barcelona, Spain) for assistance

440

at BL13-XALOC beamline.

441

SUPPORTING INFORMATION:

442

Material and Methods, Figures S1, S2 and Scheme S1 can be found free of charge online

443

at http://pubs.acs.org.

444 445

REFERENCES

446 447

H2020-BBI-PPP-2015-2-720297-ENZOX2],

the

CSIC

[project

PIE-

1. Bollinger, J.-M.-Jr., and Broderick, J.-B. (2009) Frontiers in enzymatic C-H bond activation, Curr. Opin. Chem. Biol. 13, 51-57.

448

2. Roduner, E., Kaim, W., Sarkar, B., Urlacher, V.-B., Pleiss, J., Gläser, R.,

449

Einicke, W.-D., Sprenger, G.A., Beifuβ, U., Klemm., E., Liebner, C., Hieronymus, H., Hsu,

450

S.-F., Plietker B., Laschat, S. (2013) Selective catalytic oxidation of C-H bonds with

451

molecular oxygen, ChemCatChem 5, 82-112.

452 453

3. Fasan, R. (2012) Tuning P450 enzymes as oxidation catalysts, ACS Catal. 2, 647-666.


Page 19 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


18 454 455 456 457

4. Urlacher, V.-B., and Girhard, M. (2012) Cytochrome P450 monooxygenases: an update on perspectives for synthetic applications, Trends Biotechnol. 30, 26-36. 5. Sakaki, T. (2012) Practical application of cytochrome P450, Biol. Pharm. Bull. 35, 844-849.

458

6. Bernhardt, R., and Urlacher, V.-B. (2014) Cytochrome P450 as promising

459

catalysts for biotechnological applications: chances and limitations, Appl. Microbiol.

460

Biotechnol. 98, 6185-6203.

461 462

7. Holtmann, D. and Hollmann, F. (2016) The oxygen dilemma: a severe challenge for the application of monooxygenases?, ChemBioChem 17, 1391-1398.

463

8. Molina-Espeja, P., Gomez de Santos, P., and Alcalde, M. (2017) Directed

464

evolution of unspecific peroxygenase, in Directed enzyme evolution: Advances and

465

applications (Alcalde, M., Ed.), pp 127-143, Springer International Publishing,

466

Switzerland.

467

9. Hrycay, E.-G., and Bandiera, S.-M. (2012) The monooxygenase, peroxidase,

468

and peroxygenase properties of cytochrome P450, Arch. Biochem. Biophys. 522, 71-89.

469

10. Arnold, F.-H. (2017) Directed evolution: Bringing new chemistry to life,

470

Angew. Chem. Int. Ed. 57, 4143-4148.

471

11. Xu, H., Ning, L., Yang, W., Fang, B., Wang, C. Wang, Y., Xu, J., Collin, S.,

472

Laeuffer, F., Fourage, L., and Li., S. (2017) In vitro oxidative decarboxylation of free fatty

473

acids to terminal alkenes by two new P450 peroxygenases, Biotechnol. Biofuels 10, 208.

474

12. Munro, A.-W., McLean, K.-J., Grant, J.-L., and Makris, T.-M. (2018) Structure

475

and function of the cytochrome P450 peroxygenase enzymes, Biochem. Soc. Trans. 46,

476

183-196.



Page 20 of 31

19 477

13. Onoda, H., Shoji, O., Suzuki, K., Sugimoto, H., Shiro, Y., and Watanabe, Y.

478

(2018) -Oxidative decarboxylation of fatty acids catalyzed by cytochrome P450

479

peroxygenases yielding shorter -alkyl-chain fatty acids, Catal. Sci. Technol. 8, 434-442.

480

14. Ullrich, R., Nüske, J., Scheibner, K., Spantzel, J., and Hofrichter, M. (2004)

481

Novel haloperoxidase from the agaric basidiomycete Agrocybe aegerita oxidizes aryl

482

alcohols and aldehydes, Appl. Environ. Microbiol. 70, 4575-4581.

483

15. Pecyna, M.-J., Ullrich, R., Bittner, B., Clemens, A, Scheibner, K., Schubert,

484

R., and Hofrichter, M. (2009) Molecular characterization of aromatic peroxygenase from

485

Agrocybe aegerita, Appl. Microbiol. Biotechnol. 84, 885-897.

486

16. Hofrichter, M., Ullrich, R., Pecyna, M.-J., Liers, C., and Lundell, T. (2010) New

487

and classic families of secreted fungal heme peroxidases, Appl. Microbiol. Biotechnol.

488

87, 871-897.

489

17. Hofrichter, M., Kellner, H., Pecyna, M.-J., and Ullrich, R. (2015) Fungal

490

unspecific peroxygenases: Heme-thiolate proteins that combine peroxidase and

491

cytochrome P450 properties, in Advances in Experimental Medicine and Biology

492

(Hrycay, E.-G., and Bandiera, S.-M. Eds.), 851, pp 341-368, Springer International

493

Publishing, Switzerland.

494

18. Wang, Y., Lan, D., Durrani, R., and Hollmann, F. (2017) Peroxygenases en

495

route to becoming dream catalysts. What are the opportunities and challenges?, Curr.

496

Opin. Chem. Biol. 37, 1-9.

497 498

19. Hofrichter, M., and Ullrich, R. (2014) Oxidations catalyzed by fungal peroxygenases, Curr. Opin. Chem. Biol. 19,116-125.

499

20. Bormann, S., Gomez Baraibar, A., Ni, Y., Holtmann, D., and Hollmann, F.

500

(2015) Specific oxyfunctionalisations catalysed by peroxygenases: opportunities,

501

challenges and solutions, Catal. Sci. Technol. 5, 2038-2052.


Page 21 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


20 502

21. Ullrich, R., Poraj-Kobielska, M., Scholze, S., Halbout, C., Sandvoss, M.,

503

Pecyna, M.-J., Scheibner, K. and Hofrichter, M. (2018) Side chain removal from

504

corticosteroids by unspecific peroxygenase, J. Inorg. Biochem 183, 84-93.

505

22. Molina-Espeja, P., Garcia-Ruiz, E., Gonzalez-Perez, D., Ullrich, R.,

506

Hofrichter, M., and Alcalde, M. (2014) Directed evolution of unspecific peroxygenase

507

from Agrocybe aegerita, Appl. Environ. Microbiol. 80, 3496-3507.

508

23. Molina-Espeja, P., Ma, S., Mate, D.-M., Ludwig, R., and Alcalde, M. (2015)

509

Tandem-yeast

510

peroxygenase, Enzyme Microb. Technol. 73-74, 29-33.

expression

system

for

engineering

and

producing

unspecific

511

24. Molina-Espeja, P., Cañellas, M., Plou, F.-J., Hofrichter, M., Lucas, F., Guallar,

512

V., and Alcalde, M. (2016) Synthesis of 1-naphthol by a natural peroxygenase

513

engineered by directed evolution, ChemBioChem 17, 341-349.

514

25. Gomez de Santos, P., Cañellas, M., Tieves, F., Younes, S.-H.-H., Molina-

515

Espeja, P., Hofrichter, M., Hollmann, F., Guallar, V., and Alcalde, M. (2018) Selective

516

synthesis of the human drug metabolite 5´-hydroxypropranolol by and evolved self-

517

sufficient peroxygenase, ACS Catal. 8, 4789-4799.

518

26. Piontek, K., Strittmatter, E., Ullrich, R., Gröbe, G., Pecyna, M.-J., Kluge, M.,

519

Scheibner, K., Hofrichter, M., and Plattner, D.-A. (2013) Structural basis of substrate

520

conversion in a new aromatic peroxygenase cytochrome P450 functionality with benefits,

521

J. Biol. Chem. 288, 34767-34776.

522

27. Molina-Espeja, P., Plou F.J., Gomez de Santos, P., and Alcalde M. (2017)

523

Mutants of unspecific peroxygenase with high monooxygenase activity and uses thereof.

524

Patent nº WO/2017/081355.

525

28. Andreeva, A., Howorth, D., Chothia, C., Kulesha, E., and Murzin, A.-G. (2014)

526

SCOP2 prototype: a new approach to protein structure mining, Nucleic Acids Res. 42,

527

D310-D314.



Page 22 of 31

21 528

29. Kühnel, K., Derat, E., Terner, J., Shaik, S., and Schlichting, I. (2007) Structure

529

and quantum chemical characterization of chloroperoxidase compound 0, a common

530

reaction intermediate of diverse heme enzymes, Proc. Natl. Acad. Sci. U.S.A. 104, 99-

531

104.

532

30. Kühnel, K., Blankenfeldt, W., James Terner, W.-B., and Schlichting, I. (2006)

533

Crystal Structures of Chloroperoxidase with Its Bound Substrates and Complexed with

534

Formate, Acetate, and Nitrate. J. Biol. Chem. 281, 23990-23998.

535

31. Yoshida, T., Tsuge, H., Hisabori, T., and Sugano, Y. (2012) Crystal structures

536

of dye-decolorizing peroxidase with ascorbic acid and 2,6-dimethoxyphenol, FEBS Lett.

537

586, 4351-4356.

538

32. Shehzad, A., Panneerselvam, S., Linow, M., Bocola, M., Roccatano, D.,

539

Mueller-Dieckmann, J., Wilmanns, M., and Schwaneberg, U. (2013) P450 BM3 crystal

540

structures reveal the role of the charged surface residue Lys/Arg184 in inversion of

541

enantioselective styrene epoxidation. Chem Commun 49, 4694-4696.

542

33. Anh, D.-H., Ullrich, R., Benndorf, D., Svatoś, A., Muck, A., and Hofrichter, M.

543

(2007) The coprophilous mushroom Coprinus radians secretes a haloperoxidase that

544

catalyzes aromatic peroxygenation, Appl. Environ. Microbiol. 73, 5477–5485.

545

34. Gröbe, G., Ullrich, R., Pecyna, M.-J., Kapturska, D., Friedrich, S., Hofrichter,

546

M., and Scheibner, K. (2011) High-yield production of aromatic peroxygenase by the

547

agaric fungus Marasmius rotula, AMB Express 1, 31.

548

35. Babot, E.-D., del Rio, J.-C., Kalum, L., Martinez, A.-T., and Gutierrez, A.

549

(2013) Oxyfunctionalization of aliphatic compounds by a recombinant peroxygenase

550

from Coprinopsis cinerea, Biotechnol. Bioeng. 110, 2323-2332.

551

36. Kiebist, J., Schmidtke, K.-U., Zimmermann, J., Kellner, H., Jehmlich, N.,

552

Ullrich, R., Zänder, D., Hofrichter, M., and Scheibner, K. (2017) A peroxygenase from


Page 23 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


22 553

Chaetomium

554

ChemBioChem 18, 563-569.

globosum

catalyzes

the

selective

oxygenation

of

testosterone,

555

37. Ullrich, R., Poraj-Kobielska, M., Scholze, S., Halbout, C., Sandvoss, M.,

556

Pecyna, M.-J., Scheibner, K. and Hofrichter, M. (2018) Side chain removal from

557

corticosteroids by unspecific peroxygenase, J. Inorg. Biochem. 183, 84-93.

558

38. Ni, Y., Fernández-Fueyo, E., Gomez Baraibar, A., Ullrich, R., Hofrichter, M.,

559

Yanase, H., Alcalde, M., van Berkel, W.-J.-H., and Hollmann, F. (2016) Peroxygenase-

560

catalyzed oxyfunctionalization reactions promoted by the complete oxidation of

561

methanol, Angew. Chem. Int. Ed. 55, 798-801.

562

39. Zhang, W., Burek, B.-O., Fernandez-Fueyo, E., Alcalde, M., Bloh, J.-Z., and

563

Hollmann, F. (2017) Selective activation of C-H bonds by cascading photochemistry with

564

biocatalysis, Angew. Chem. Int. Ed. 56, 15451-15455.

565

40. Zhang, W., Fernandez-Fueyo, E., Ni, Y., van Schie, M., Gacs, J., Renirie, R.,

566

Wever, R., Mutti, F.-G., Rother, D., Alcalde, M., and Hollmann, F. (2018) Selective

567

aerobic oxidation reactions using a combination of photocatalytic water oxidation and

568

enzymatic oxyfunctionalizations, Nat. Catal. 1, 55-62.

569 570 571 572 573 574 575 576 577



Page 24 of 31

23 578

FIGURE LEGENDS

579

Figure 1 General view of the evolved peroxygenase structure at 1.5 Å resolution with a

580

ligand-free active site. (a) The cartoon represents the overall structure of native (light

581

orange) vs. PaDa-I (white), showing the magnesium ion (Mg2+, green sphere), the heme

582

group (dark green sticks), the chloride ion (red sphere), and the different glycan chains

583

as sticks. (b) The residues mutated in the evolved peroxygenase are highlighted in purple

584

sticks. PDB code 5OXU, see also Table 1.

585

Figure 2 Active site difference of evolved vs. native peroxygenase. (a) The F311L

586

change (wild type AaeUPO, yellow; PaDa-I, purple) produces a shift in the position of

587

Phe76 that affects the Phe191 side chain, enlarging the entrance to the heme (green)

588

channel. Details of the atomic interactions between relevant residues in the wild type

589

AaeUPO (b) or the recombinant PaDa-I (c) crystal structures. (d) The surface

590

representation of PaDa-I illustrates the funnel that provides access to the heme group.

591

The equivalent positions of Phe76 and Phe191 in the wild type AaeUPO structure (yellow

592

spheres) show the expanded entrance in PaDa-I. Ala316 is highlighted as a violet stick.

593

PaDa-I PDB code 5OXU; AaeUPO PDB code 2YOR, see also Table 1.

594

Figure 3 Structural peroxygenase complexes. PaDa-I crystals complexed with: acetate,

595

Ac (a); DMSO (b); propranolol, Prl (with glycerol, Gly), (c); naphthalene, NAF (d); 1-

596

naphthol (complex I, 1NaI) (e); 1-naphthol (complex II, 1NaII) (f); DMP (g); VA (h); and

597

styrene, Sty (i). Key residues for binding are represented as sticks, water molecules as

598

red spheres and the heme group with green sticks. The 2Fo-Fc electronic density at the

599

ligands has been contoured at a rmsd of 0.9-1 . See also Table 1.

600

Figure 4 Molecular basis of PaDa-I specificity. (a) Structural superimposition of the

601

different molecules bound at the heme catalytic site reported in this study, represented

602

as stereopair, showing relevant interacting protein residues. (b) Comparison of bound

603

propranolol and 1-naphthol (complex II) in the corresponding crystal complexes to

604

illustrate the putative transitory binding sites of ligands to the heme site. (c) Cartoon (left)


Page 25 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


24 605

and surface (right) representation of a 90º rotation view of (b), as seen from the outer

606

entrance to the funnel, showing the residues that establish atomic interactions with the

607

ligands at the outer entrance. Residues from loops S240-S245 and G314-G318

608

contribute to substrate binding by mostly main-chain contacts. The red loop indicates an

609

alternative conformation of loop G314-G318 observed in some complexes and the yellow

610

sticks the disulfide bridge stabilizing the C-terminus.

611


ACS Chemical Biology 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41

a

b


Fig. 1

Page 26 of 31

Page 27 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41

a

c


b

d


Fig. 2

Fig. 3

ACS Chemical Biology 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41

Page 28 of 31

a

b

c

d

e

f

g

h

i


Page 29 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56


a

b

c


Fig. 4

ACS Chemical Biology 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47

Page 30 of 31

!"#$%&'(!"#$%&'(()*#'+,-.!%&'&-%&-.%!/0'(12%!-3!+'#23&,2%2%!'#2!4)#!&,2!,-*,!#2%)(1&-)3!%,2((56! ! )*+,-"$&."-"!

789:8';'.2&'&2!

789:8';'
789:8';'
789:8';'
789:8';'
789:8';'
789:8';'
789:8';'
?)!(-*'3@!

C:?'+,&,)(!/ 789:8';'! !DC"

B+'.2!*#)1+!

!DC!

!DC!

!DC!

!DC"

!DC"

!DC"

!DC"

!DC"

!DC"

73-&!.2((!+'#'E2&2#%!

!

!

!

!

!

!

!

!

!

!

'!/F5!

GH6I!

HC6IJ!

HC6IK!

HC6CG!

HC6CL!

HC6IL!

HC6CK!

HC6IM!

HC6DM!

HC6IJ!

N!/F5!

LL6CH!

HJ6LO!

HK6IJ!

HK6ID!

HJ6JO!

HJ6KI!

HJ6LJ!

HJ6KK!

HJ6JM!

HJ6LK!

.!/F5!

KM6KL!

GC6CI!

GI6KC!

GI6KH!

GC6CI!

GI6GL!

GI6LK!

GI6KL!

GC6ICK!

GI6LD!

P!/Q5!

CIM6DJ!

CIL6MG!

CIL6LO!

CIL6GH!

CIL6OG!

CIL6KL!

CCI6CD!

CCI6IL!

CIL6JI!

CIL6KH!

/"-"&01$$%0-213!

!

!

!

!

!

!

!

!

!

!

R2'E(-32!

S>T9"!/>TR>5!

S>T9"!/>TR>5!

S>T9"!/>TR>5!

S>T9"!/>TR>5!

S>T9"!/>TR>5!

S>T9"!/>TR>5!

S>T9"!/>TR>5!

S>T9"!/>TR>5!

S>T9"!/>TR>5!

S>T9"!/>TR>5!

U'V2(23*&,!/F5!

I6LJLH!

I6LJLH!

I6LJKJ!

I6LJLM!

I6LJLH!

I6LJLD!

I6LJLD!

I6LJLD!

I6LJLH!

I6LJLM!

OH6DM:C6OD!

OK6ID:C6OM! /C6OM:!C6OH5!

OK6CO:C6MG! /C6MG:C6MK5!

OK6DG:C6OM! /C6OM:C6OH5!

OK6IO:C6CM! /C6CM:C6CO5!

OK6IG:C6DM! /C6DM:C6DH5!

OJ6LD:C6IK! /C6IK:C6CI5!

OH6CL:C6GH! /C6GH:C6GK5!

OK6IO:C6HM! /C6HM:C6HG5!

!

!

!

!

!

!

!

!

/C6OD:C6OO5!

OK6CL:C6OJ! /C6OJ:C6HI5!

/"-"&4*10%,,235!

!

!

X)&'(!#24(2.&-)3%!

CDLCGCD! /OJLCC5!

CDKLHC!

73-Y12!#24(2.&-)3%!

CLICKO!/KGLO5!

A1(&-+(-.-&$!

G6K!/H6H5!

")E+(2&232%%!/Z5!

LJ6L!/LI6O5!

A2'3!#=[!/#5! $%&'(&)"/Z5!

W2%)(1&-)3!/F5!

CHGLJC!

OCHMJJ!/DIODD5!

OKHDDO!/DMCMD5!

DLJMKH!/CCMKC5!

JMHKKK!/CKKML5!

HKOGIH!/CLCCC5!

KOIDLL!/COHLD5!

HMDMG!/DGDH5!

GCKMG!/MIOG5!

JCKLJ!/MHJL5!

GCIDL!/DJKM5!

CDOHDJ!/HOHK5!

LGKKC!/OHKG5!

CDJKGD!/MDOG5!

MKGJM!/CLOO5!

D6O!/D6D5!

G6J!/G6J5!

G6J!/G6H5!

O6L!/O6C5!

H6L!/M6H5!

G6I!/O6D5!

G6G!/O6H5!

O6C!/M6L5!

H6O!/H6H5!

LD6L!/LM6O5!

CII6I!/CII6I5!

LL6K!/LL6C5!

LK6H!/LC6I5!

LL6I!/KJ6K5!

LL6J!/LH6C5!

LI6I!/KG6M5!

LH6J!/LH6O5!

LL6J!/LJ6J5!

CL6D!/G6K5!

J6J!/D6H5!

CH6D!/D6K5!

CO6L!/M6C5!

CI6K!/D6I5!

CM6C!/D6I5!

CM6G!/D6C5!

CO6L!/D6I5!

CI6G!/D6M5!

CC6J!/M6C5!

G6D!/CK6C5!

L6G!/GH6I5!

K6O!/GJ6D5!

K6O!/GL6H5!

K6L!/GH6C5!

G6G!/GI6H5!

G6D!/GC6O5!

H6J!/GI6O5!

L6C!/GO6H5!

L6C!/GI6K5!

$*+%))"/Z5!

D6G!/K6O5!

H6I!/MD6I5!

M6H!/DJ6L5!

M6H!/DL6M5!

O6M!/MO6L5!

D6K!/MG6L5!

D6J!/MM6O5!

D6M!/MI6K5!

O6L!/MH6I5!

O6D!/DJ6J5!

A)(2.1(2%!+2#!>B7!

M!

C!

C!

C!

C!

C!

C!

C!

C!

C!

6%723%8%3-!

!

!

!

!

!

!

!

!

!

!

W\)#]!=!W4#22)))!/Z5!

CH6I=CG6O!

CG6O=CJ6K!

CO6K=CH6D!

CG6I=CJ6I!

CH6H=CK6I!

CG6G=CJ6H!

CG6I=CG6L!

CH6J=CG6M!

CH6O=CJ6O!

CH6J=CG6J!

9:&17&"-18,&;& "3*(2%!/Q5!

C6MMC!

C6DLK!

C6DLD!

C6MCK!

C6OIH!

C6MHC!

C6MHL!

C6ODL!

C6MID!

C6DKK!

A/=&"00%,,213& 01.%,&

H9SX!

H9S7!

H9_C!

H9_D!

GàU!

GàS!

Gà_!

Gàb!

G`TI!

G`TO!

! ! c

WE2#*2!d!e,](!e-!f!

Structural Insights into the Substrate Promiscuity of a Laboratory

Recommend Documents