Báo cáo khoa học: Role of the C-terminal extension in a bacterial tyrosinase Michael Fairhead and Linda Thony-Meyer doc

With respect to its overall fold and active site architecture, the bacterial enzyme is strongly similar to the related enzyme catechol Keywords C-terminal domain; melanin; tyrosinase; Ve

Michael Fairhead and Linda Tho¨ny-Meyer

EMPA, Swiss Federal Laboratories for Materials Testing and Research, Laboratory for Biomaterials, St Gallen, Switzerland

Introduction

Tyrosinases and the related catechol oxidases

(collec-tively termed polyphenol oxidases) comprise a family

of binuclear copper enzymes found in many species

of animals, plants, fungi and bacteria that use

phe-nol-like starting materials to produce a variety of

biologically important compounds, such as melanin

and other polyphenolic compounds [1–3] These

type III copper proteins are capable of two activities:

monophenolase or cresolase activity (EC 1.14.18.1)

and diphenolase or catecholase activity (EC 1.10.3.1)

Both activities result in the formation of reactive

quinones, and these species are important

intermedi-ates in the biosynthesis of compounds such as

melanin

Given the ability of tyrosinases to react with phenols

and its di-copper redox centres, they have been

proposed for use in a variety of biotechnological,

biosensor and biocatalysis applications [2] One exam-ple includes tyrosinase immobilization as an electro-chemical biosensor for a range of phenolic compounds [4] The enzyme can also react with tyrosine found on polypeptides, and the reactive quinones formed allow for protein cross-linking to chitosan films as well as protein-protein cross-linking [5,6]

The only available crystal structure of the tyrosin-ases comes from the secreted enzyme of Streptomyces castaneoglobisporus [7] tyrosinase The structure shows the enzyme in complex with its accessory caddie protein (see below) The tyrosinase is predominately a-helical in structure and contains six histidine residues co-ordinating the two copper atoms that form the active site of the enzyme With respect to its overall fold and active site architecture, the bacterial enzyme

is strongly similar to the related enzyme catechol

Keywords

C-terminal domain; melanin; tyrosinase;

Verrucomicrobium spinosum; zymogen

Correspondence

L Tho¨ny-Meyer, EMPA, Swiss Federal

Laboratories for Materials Testing and

Research, Laboratory for Biomaterials,

Lerchenfeldstrasse 5, St Gallen, CH-9014,

Switzerland

Fax: +41 44 071 274 7788

Tel: +41 44 071 274 7792

E-mail: linda.thoeny@empa.ch

(Received 22 October 2009, revised

13 January 2010, accepted 22 February

2010)

doi:10.1111/j.1742-4658.2010.07621.x

The well studied bacterial tyrosinases from the Streptomyces sp bacteria are distinguishable from their eukaryotic counterparts by the absence of a C-terminal extension In the present study, we report that the tyrosinase from the bacterium Verrucomicrobium spinosum also has such a C-terminal extension, thus making it distinct from the Streptomyces enzymes The entire tyrosinase gene from V spinosum codes for a 57 kDa protein (full-length unprocessed form), which has a twin arginine translocase type signal peptide, the two copper-binding motifs typical of the tyrosinase protein family and the aforementioned C-terminal extension We expressed various mutants of the recombinant enzyme in Escherichia coli and found that removal of the C-terminal extension by genetic engineering or limited tryp-sin digest of the pro-form results in a more active enzyme (i.e 30–100-fold increase in monophenolase and diphenolase activities) Further studies also revealed the importance of a phenylalanine residue in this C-terminal domain These results demonstrate that the V spinosum tyrosinase is a new example of this interesting family of enzymes In addition, we show that this enzyme can be readily overproduced and purified and that it will prove useful in furthering the understanding of these enzymes, as well as their biotechnological application

Abbreviations

L -DOPA, L -3,4-dihydroxyphenylalanine; TAT, twin arginine translocase.

oxidase from sweet potato [8]; however, the plant

enzyme is only capable of the diphenolase reaction

(EC 1.10.3.1)

The major distinguishing feature of the

Strepto-myces sp enzyme is the requirement for an accessory

protein that is necessary for copper incorporation [1]

Several mutagenesis studies, as well as the crystal

structure, have demonstrated the importance of this

accessory ‘caddie protein’ for copper incorporation

into the Streptomyces tyrosinase [7,9] and the

expres-sion of active Streptomyces tyrosinase in either

Escherichia coli or its native host requires the

co-expression of the gene encoding this caddie

pro-tein This arrangement is entirely different from that

of the eukaryotic enzymes, which are not known to

require such a caddie protein and also have a

C-ter-minal extension, the removal of which usually leads

to a marked increase in activity [10] Indeed, it is

esti-mated that approximately 98% of the the tyrosinase

present in mushrooms occurs in such a latent form

[11] However, the Streptomyces sp tyrosinases may

not be wholly representative of the bacterial form of

these enzymes because the Rhizobium etli tyrosinase

has been reported not to require a copper chaperone

for activity [12]

Given their interesting properties and the wide

poten-tial of these enzymes, there are few successful examples

of recombinant production systems that provide high

yields of pure enzyme, with most studies using the native

Streptomyces sp [13,14], Neurospora crassa [15] and

Agaricus bisporus [2] enzymes To cover this shortfall,

we have cloned several uncharacterized tyrosinase

genes from different bacterial species with the aim of

identifying enzymes that have suitable characteristics

for structure⁄ function studies, as well as

biotechnologi-cal applications In the present study, we report

the results obtained with the tyrosinase gene from

Verrucomicrobium spinosum

Verrucomicrobium spinosum is part of the ubiquitous

Verrucomicrobia phylum These bacteria are found in

a wide range of aquatic and terrestrial habitats

[16,17] Verrucomicrobium spinosum in particular is

found in fresh water eutrophic (nutrient rich, oxygen

poor) habitats and is capable of both aerobic and

fer-mentative metabolism This Gram-negative,

yellow-pigmented bacterium is somewhat unusual as a result

of the presence of numerous wart-like prosthecae

appendages on its surface [17,18] and its

compartmen-talized cytoplasm [19] This bacterium is not known to

normally produce melanin, and thus the presence of a

tyrosinase gene in its genome was somewhat surprising

because such genes are usually associated with

black pigment formation in various bacterial and fungal species [20]

Results and Discussion

Analysis of the V spinosum tyrosinase gene region

The V spinosum tyrosinase gene is preceded upstream

by a gene encoding a predicted laccase and followed downstream by a gene encoding a predicted b-sheet-rich protein for which we could find no obvious func-tion or homologue (Fig 1A) This differs from the Streptomyces tyrosinase gene arrangement, where the tyrosinase is typically preceded by a gene encoding an accessory protein required for copper incorporation [1] Given the absence of such a caddie protein gene upstream or downstream of the V spinosum tyrosinase gene, we drew the conclusion that the V spinosum tyrosinase does not require such a protein for copper insertion The V spinosum tyrosinase may therefore be similar to the aforementioned R etli tyrosinase, which also has been reported not to require a copper chaper-one [12] The presence of another multicopper oxidase-like laccase gene upstream of the tyrosinase gene

is also interesting because laccases are known to be capable of synthesizing melanin, albeit usually from diphenols such as epinephrine and l-3,4-dihydroxy-phenylalanine (l-DOPA) [21]

Also present in the surrounding DNA sequence are several regions with homologies to the binding sites of

E coli RpoS and RpoD sigma factors, which are known to be involved in transcriptional regulation [22] The predicted b-sheet-rich protein gene is fol-lowed by a region with a high probability of leading to

an RNA secondary structure in the transcript, indica-tive of a site of transcription termination The presence

of these features may indicate that the tyrosinase gene

is part of an operon

As stated in the Introduction, V spinosum is not known to produce melanin under normal growth conditions The laccase and⁄ or tyrosinase are there-fore probably only synthesized under a specific set of circumstances or serve some alternative function to melanin production We attempted to induce melanin synthesis by cultivating the V spinosum bacterium on solid or in liquid media supplemented with excess copper or amino acids in an attempt to mimic con-ditions known to induce Streptomyces species tyro-sinases [23] However, these experiments did not yield any detectable tyrosinase activities, as indicated

by the lack of formation of any black pigments or

monophenolase⁄ diphenolase activities in bacterial

extracts (data not shown)

Features of the amino acid sequence of the

V spinosum tyrosinase

The amino acid sequence of the full-length V

spino-sum pre-pro-tyrosinase (Fig S1) can be divided

approximately into three domains: a twin arginine

translocase (TAT) signal peptide, a core domain

con-taining the two copper-binding motifs and a

C-termi-nal extension (Fig 1B) The presence of a predicted

TAT signal peptide at the N-terminus (amino acids

1–36) would suggest that the protein is exported to the

periplasmic space of V spinosum in an already folded

form, as often found for metal-containing periplasmic

proteins [24] The presence of this signal peptide is in

agreement with the fact that the Streptomyces

tyrosin-ases are also secreted via the TAT secretion pathway

[25]

Also present in the sequence are the two copper

A (amino acids 86–96) and copper B (amino acids

258–294) binding motifs common to most tyrosinase

sequences [3] that contain five of the six

copper-bind-ing histidine ligands The sixth histidine ligand found

in tyrosinases typically occurs before the copper A

motif From sequence alignments, we suggest that this

ligand is most likely histidine 80 in the V spinosum

tyrosinase Another motif, which is present not only in

tyrosinases, but also in the oxygen transporting haemocyanin proteins, is the PYWDW (amino acids 118–122) and has been hypothesized to be involved in oxygen binding [26]

Previous sequence analysis in other studies has dem-onstrated the presence of a conserved Yx(Y⁄ F) motif

in the C-terminal domains of both the Streptomyces type tyrosinases and processed eukaryotic tyrosinases and haemocyanins [10] This motif can also be seen to

be present in the V spinosum tyrosinase (Figs 1B and S1) It has been hypothesized, with support from the crystal structure of catechol oxidase, that the tyrosine residue(s) in this motif form a hydrogen-bonding network to a conserved arginine residue close to the N-terminus that stabilizes the mature, processed form

of polyphenol oxidases [8,10] A homologue of this arginine residue is also present in V spinosum tyrosinase (Arg40) (Figs 1 and S1)

Another notable feature of the V spinosum tyrosi-nase sequence is the presence of the proteins only cys-teine residue at position 84 A cyscys-teine at this position

is also found in some other eukaryotic tyrosinases and plant catechol oxidases This cysteine may be of functional importance because it has been shown to form a novel alkane-thiol bond to one of the copper ligand histidine residues in the structure of the related sweet potato catechol oxidase [8] The equivalent cysteine and bond are absent in the structure of S cas-taneoglobisporus tyrosinase [7] Indeed, Streptomyces

A

TAT signal peptide 1–36

Pre-pro-tyrosinase 518 amino acids

Core domain 37–357

C-terminal extension amino acids 358–518

Pro-tyrosinase 481 amino acids

Core domain 36–357

C-terminal extension amino acids 358–518

Core tyrosinase 320 amino acids

Core domain 36–357

Trypsinisedpro-tyrosinase 332 amino acids

Core domain 36–370

Lys370 Ala36

Ala36 Ala36

Val357

Phe518

B

Fig 1 Overview of the tyrosinase gene

and surrounding genes in the genome of

V spinosum (A) Showing the tyrosinase

gene and those in its immediate vicinity in

the V spinosum genome Triangles indicate

regions with homology to the binding sites

of the E coli RpoS and RpoD regulatory

proteins; the octagon shows the position of

a region predicted to have a high probability

of RNA secondary structure, which is

indicative of a termination transcript.

(B) An overview of the pre-pro-tyrosinase,

pro-tyrosinase and core-tyrosinase

constructs and their notable features.

sp tyrosinases contain no cysteine residues at all [27].

However, experimental evidence does demonstrate the

presence of such a bond in N crassa tyrosinase [15]

and in molluscan haemocyanins [28]

The arrangement of a core tyrosinase domain

followed by a C-terminal extension (Fig 1B) is similar

in design to mushroom tyrosinase and plant polyphenol

oxidases [10] The mushroom C-terminal domain can be

removed by proteolysis or denatured by SDS, leading to

an activation of the enzyme [11,29] By contrast, the

Streptomycestype tyrosinases have no such C-terminal

extension after the core tyrosinase domain [1]

One proposed function of the C-terminal extension

in plant and fungal polyphenol oxidases is a role in

membrane binding, making them similar to the

mam-malian tyrosinases, which have a single transmembrane

domain [27] However, it is considered that the plant

forms are not integral membrane proteins because they

can be released in an active form from the membrane

by sonication, proteolysis or treatment with mild

deter-gents [30,31] Thus, whether the C-terminal domain in

the plant and fungal enzymes has a purely inhibitory

function and⁄ or a role in membrane binding is unclear

at present With regard to V spinosum pro-tyrosinase,

sequence analysis of the C-terminal domain, and

indeed of the entire sequence, suggested that no

trans-membrane helices were present, as also demonstrated

by the fact the enzyme is produced in a soluble form

in E coli

Recombinant expression of V spinosum tyrosinase in E coli

To study the properties of the V spinosum tyrosinase,

we created a range of constructs (Table 1) for recombi-nant expression of the pre-tyrosinase, the pro-tyrosinase and the core pro-tyrosinase (Fig 1B) It can be seen from Fig 2 that E coli cells transformed with plasmids containing either the pre-pro-tyrosinase or the pro-tyrosinase tyrosinase constructs (Fig 2B, C) produced a black pigment when streaked onto M9 agar plates containing tyrosine and copper, whereas a strain lacking a tyrosinase construct remained white (Fig 2A)

The activity observed on the M9 agar plates was found to correlate with over-expression of the various proteins in liquid media It can be seen from the gel presented in Fig 3A that bands are present in samples

of lysate of E coli cells transformed with plasmids encoding the different tyrosinase variants These bands correspond to the calculated molecular masses of the respective polypeptides (Table 1), namely 57 kDa for pre-pro-tyrosinase (lane 4) and 53.4 kDa for pro-tyros-inase (lane 3) The different constructs were expressed

at different levels, with an increase in expression occur-ring when the putative N-terminal TAT signal peptide was removed (Fig 3A, lanes 3 and 4)

We found it necessary to express all the tyrosinase constructs in an apo-form, by growing and inducing

Table 1 List of active constructs produced in the work and their features ND, not determined; NA, not applicable.

Name (plasmid)

Mutations or modifications

Calculated molecular mass (kDa) a

Determined molecular

Extinction coefficient

280 nm (m M )1Æcm)1)a Purpose Pre-pro-tyrosinase

(pMFvppt)

gene from

V spinosum Pro-tyrosinase

(pMFvpt)

Amino acids 36–518 with non-original methionone start codon

pepetide from pro-tyrosinase gene for cytosolic expression Trypsinized

pro-tyrosinase (NA)

extension via trypsin for improved activtiy Core tyrosinase

(pMFvct)

Amino acids 36–357 with non-original methionone start codon

extension for improved activity

Pro-tyrosinase

F453A (pMFvptf2a)

Pro-tyrosinase with phenylalanine 453 mutated to alanine

residue performs a

‘gatekeeper’ function at the tyrosinase active site a

Values calculated using PROTPARAM (24).bMolecular mass determined by MS.

the transformed cells in media prepared using Milli-Q

water (Millipore, Billerica, MA, USA) and lacking

added copper This was necessary because, otherwise,

a black pigment was produced during incubation This

pigment was found to inhibit the growth of E coli and

to foul protein purification columns, both of which

resulted in a low protein yield This problem was

par-ticularly acute with the highly active core tyrosinase

The formation of a black pigment (presumably

mela-nin) was most likely a result of the action of the

expressed tyrosinase on the tyrosine present in the

pep-tone or N-Z-amine (Sigma-Aldrich, Buchs,

Switzer-land) that was added to the expression medium as an

external source of amino acids to aid recombinant

protein production Provided the precaution of not

supplying copper to the medium was taken, we found

that soluble protein could be obtained for all the

described constructs

In experiments with the pre-pro-tyrosinase construct,

we did not obtain sufficient amounts of protein for

purification We also attempted to isolate the protein

from the E coli periplasm but could not find any

evi-dence of activity, indicating a lack of export of the

protein It could be that the E coli TAT system is

unable to recognize the V spinosum export signal

peptide

When designing tyrosinase constructs without the

predicted N-terminal signal peptide (amino acids

1–36), we retained amino acid 36, an alanine, rather

than using amino acid 37, a lysine, because it is

known that, after a post-translational processing of the

N-terminal methionine, which often occurs for proteins expressed in E coli, according to the N-end rule, a newly-created N-terminal lysine would result in a very short protein half-life, whereas an N-terminal alanine would be fine [32]

The recombinant pro-tyrosinase was expressed and purified with final yields of approximately 20 mgÆL)1

of pure protein Subsequent analytical gel filtration of the purified pro-tyrosinase showed a single peak corre-sponding to a monomer (Fig S2) The mass of the purified protein determined via MS (53 501 kDa) corresponded closely to the expected full-length pro-tyrosinase (53 500 kDa) assuming the removal of the N-terminal methionine

Reconstitution of recombinat V spinosum tyrosinase with copper

The holo-forms of tyrosinase were obtained after puri-fication by adding copper to a three-fold molar excess, and samples were subsequently exhaustively dialysed in

an attempt to remove any nonspecifically bound cop-per The final copper content of the dialysed samples was then determined (Table 2) Although pro-tyrosi-nase was found to be nearly fully loaded with copper using this method (1.8 molar equivalents), the core tyrosinase and pro-tyrosinase F453A mutant were found to be significantly under-loaded (1.4 and 1.2 molar equivalents respectively) It is possible that the protocol used was not optimal for copper incorporation into these variants (see Experimental

A

B

Fig 2 Melanin formation on tyrosine

con-taining solid media by E coli cells

express-ing V spinosum tyrosinase constructs (A)

Escherichia coli transformed with vector

containing no insert (pQE-60); (B) E coli

transformed with pMFvppt

(pre-pro-tyrosi-nase); (C) E coli transformed with pMFvpt

(pro-tyrosinase); (D) E coli transformed with

pMFvct (core tyrosinase); (E) E coli

trans-formed with pMFvptf2a (pro-tyrosinase

F453A).

procedures) and, indeed, it has been reported that incubation at pH 6 may result in higher levels of cop-per reconstitution than at pH 8 [33,34] We are cur-rently investigating this possibility

In addition, despite extensive dialysis of reconsti-tuted samples, it cannot be excluded that some of the copper is nonspecifically bound to the protein We have found, however, that attempts to remove any such copper ions with low concentrations of the chelat-ing agent EDTA (100 lm) resulted in a complete loss

of activity and detectable copper As an alternative to copper reconstitution of the purified proteins, we also attempted to grow bacteria in minimal media contain-ing copper as a means of produccontain-ing holo protein directly However, we found that the omission of an external amino acid source such as N-Z-amine led to very low levels of tyrosinase expression, as well as low cell densities, meaning that the purification of holo protein in this way was impracticable

C-terminal processing by trypsin

As noted above, the C-terminal extension found in the latent form of mushroom tyrosinase has been shown

to be inhibitory to activity, and its removal by serine proteases such as subtisilin results in an activation of the enzyme, similar to the protease zymogen system found for many digestive enzymes, such as trypsin [11] The related plant catechol oxidase enzymes also have similar C-terminal extensions [10] Sequence analysis suggested that this may also be the case for the

V spinosum enzyme (see above) We therefore used trypsin digestion to determine whether a smaller, more active fragment could be produced from purified pro-tyrosinase The gel in Fig 3C shows that trypsin diges-tion indeed yielded a smaller stable fragment, which was subsequently found to be far more catalytically active than the original pro-tyrosinase (Table 3) The stability of the smaller trypsinized fragment, even after

24 h of incubation with trypsin, suggests that this is a highly ordered domain with no accessible cleavage sites for trypsin This interpretation corresponds to the pro-posal that the C-terminal extension of eukaryotic poly-phenol oxidases (i.e tyrosinase and plant catechol oxidases) is highly disordered [10] compared to the corresponding core oxidase domains containing the two copper-binding motifs These disordered domains would thus be more susceptible to proteolysis than the more ordered stable core domains of the enzymes High levels of disorder in the pro-domain are also present in zymogens such as in procathepsin K [35] and probably represent an important feature in the activation mechanism of these enzymes The fact that

A

B

C

Fig 3 (A) SDS-PAGE of cells expressing the tyrosinase constructs.

Lane 1, lysate from cells transformed with pMFvptf2a

(pro-tyrosi-nase F453A); lane 2, lysate from cells transformed with pMFvct

(core tyrosinase); lane 3, lysate from cells transformed with pMFvpt

(pro-tyrosinase); lane 4, lysate from cells transformed with pMFvppt

(pre-pro-tyrosinase); lane 5, lysate from control cells transformed

with pQE-60 containing no insert (B) SDS-PAGE of purified and

trypsinized tyrosinases Lane 1, purified pro-tyrosinase; lane 2,

puri-fied core tyrosinase; lane 3, purifed pro-tyrosinase F453A mutant;

lane 4, trypsinized pro-tyrosinase; lane 5, trypsinized core

tyrosi-nase (C) SDS-PAGE showing time course of proteolysis of

pro-tyrosinase by trypsin Lane 1, pro-tyrosinase after 24 h of

incu-bation at room temperature; lane 2, trypsin after 24 h of incuincu-bation

at room temperature; lane 3, pro-tyrosinase plus trypsin after 0 h

at room temperature; lane 4, pro-tyrosinase plus trypsin after 1 h at

room temperature; lane 5, pro-tyrosinase plus trypsin after 4 h

at room temperature; lane 6, pro-tyrosinase plus trypsin after

24 h at room temperature M, Molecular mass markers.

the pro-tyrosinase exhibits some low levels of catalytic

activity also suggests some mobility between the core

tyrosinase domain and the C-terminal extension

(Table 3)

Recombinant core tyrosinase

To further asses the functional importance of the

C-terminal extension, we created a shortened form of

the V spinosum tyrosinase, using the presence of the

conserved YX(Y⁄ F) motif as a guide The resulting

construct was readily overexpressed in the E coli

cytoplasm (Fig 3A, lane 4) and found to be highly

active after loading with copper compared to the

pro-tyrosinase form (Table 3)

We also treated the mature (i.e copper-containing)

tyrosinase with trypsin and found that the trypsinized

recombinant core tyrosinase (Fig 3B, lane 5) exhibited

no apparent size difference compared to the

un-trypsi-nized preparation (Fig 3B, lane 2) but appeared to be

smaller than the trypsinized pro-tyrosinase (Fig 3B,

lane 4) Determination of the mass of the proteins by

MS revealed masses of 36 506 kDa (recombinant core

tyrosinase) and 37 874 kDa (trypsinized

pro-tyrosi-nase) corresponding to a C-terminal amino acid of

Val357 and Lys370, respectively Gel filtration revealed

that both proteins also exist in solution, similar to

pro-tyrosinase, as monomers (Fig S2)

The results obtained in the present study suggest

that the C-terminal extension has no role in copper

insertion like the Streptomyces sp ‘caddie’ protein

because the recombinant core tyrosinase enzyme was

found to be readily reconstituted with copper, as

indi-cated by its high activity and subsequent analysis of its

copper content (Table 2) This correlates with the

results obtained using apo-forms of mature tyrosinase

from both N crassa [36] and A bisporus [37], which

could also be readily reconstituted with copper This is

in contrast to the results obtained with the

Streptomy-cessp enzyme [38,39], which has an absolute

require-ment for the accessory caddie protein for copper

incorporation Furthermore, the results obtained in the

present study are in agreement with the previously

noted finding that, in the gene region around the

V spinosumtyrosinase, no gene encoding a caddie-like

protein is present (Fig 1A)

Because the pro-tyrosinase form contains no

predicted transmembrane helices and is indeed fully

soluble in E coli (see above), we suggest that the

C-terminal extension in this case has a purely inhibitory

function and neither a significant role in stabilizing the

enzyme, nor a chaperone-like function during folding,

as has been proposed for other N-terminal⁄ C-terminal

zymogen-like systems [40] It remains to be determined whether this is also the case for other pro-tyrosinase forms

Stability of the tyrosinase forms to chemical denaturation

To characterize the domain structure of the V spino-sum tyrosinase in more detail, we determined protein stability by recording protein unfolding via fluores-cence spectroscopy when increasing amounts of guani-dine hydrochloride (GdnCl) were present The determined unfolding curves (Fig S3) appeared to show two apparent transitions for holo pro-tyrosinase and one for either holo trypsinized pro-tyrosinase or the holo recombinant core tyrosinase However, the unfolded proteins were not found to refold once

Table 2 Stability and determined copper content of the tyrosinase enzymes ND, not determined.

Enzyme

GdnCl concentration ( M )

at 50% unfolded a

Molar equivalents

of copper

Holo trypsinized pro-tyrosinase 3.3 b 1.8 b ⁄ 1.5 c Apo tyrpsinized pro-tyrosinase 2.0 0.4

a Protein solutions (0.1 mgÆmL)1) were incubated for 24 h at room temperature in 10 m M Tris-HCl (pH 8) containing 0–6 M GdnCl before measurements were made (for details, see Experimental procedures) b Copper content and stability determined with trypsi-nized holo pro-tyrosinase c Copper content determined by reconsti-tuting trypsinized apo pro-tyrosinase.

Table 3 Monophenolase and diphenolase activities of the tyrosi-nase enzymes Activity of the various constructs ⁄ mutants towards the model substrates L -tyrosine and L -DOPA (n = 3 for all determi-nations).

Enzyme

V maxa K m (l M ) V maxa K m (m M ) Pro-tyrosinase 5.8 ± 0.6 421 ± 43 4.7 ± 0.3 7.0 ± 0.7 Trypsinized

pro-tyrosinaseb

325 ± 8 258 ± 6 565 ± 20 7.9 ± 0.5 Core tyrosinase 148 ± 4 280 ± 15 230 ± 7 7.6 ± 0.3 Pro-tyrosinase

F453A

16 ± 0.9 808 ± 66 14 ± 0.2 6.4 ± 0.4

a Units = lmol dopachromeÆmin)1ÆmgÆprotein)1 b Values deter-mined for trypsinized holo pro-tyrosinase.

denatured and, thus, the apparent shapes of the

unfolding curves should not be over interpreted The

use of the concentration of GdnCl at 50% unfolded as

a simple measure of the change in stability between the

various tyrosinase forms allows some conclusions to be

drawn (Table 2) The values show that the

incorpora-tion of copper into either pro-tyrosinase, trypsinized

pro-tyrosinase or recombinant core tyrosinase

signifi-cantly increases the overall stability of the protein It

was also apparent that the C-terminal extension of

pro-tyrosinase reduces its overall stability in either the

holo- or apo-forms of the enzyme The negative effect

on stability as a result of C-terminal extension would

suggest this domain is less stable than the core domain

of the enzyme, which correlates with the results

obtained with trypsin digestion It can also be seen

from Table 2 that the recombinant core domain

tyrosi-nase appears to be less stable than the trypsinized

pro-tyrosinase; this could be a result of its reduced copper

content Alternatively, it may be that the recombinant

core tyrosinase C-terminal extension is slightly too

short for optimal stability and that residues after the

YX(Y⁄ F) motif also play a role in protein stability

Mono- and diphenolase activities of the

recombinant tyrosinases

When we measured activities towards either l-tyrosine

or l-DOPA of pro-tyrosinase, a major increase in

activity upon removal of the C-terminal extension by

trypsin was found, namely an approximately 50-fold

increase in mono- and a 100-fold increase in

dipheno-lase activitiy (Table 3) There was also a less significant

lowering in the Km value for l-tyrosine upon removal

of the C-terminal extension (i.e from 421 to 258 lm)

The activities of the trypsinized pro-tyrosinase

towards l-tyrosine or l-DOPA was found to be

approximately twice that of the recombinant core

tyrosinase, although the Km for both substrates is

almost identical The increased level of activity is

prob-ably a result of the higher copper content of the

trypsi-nized pro-tyrosinase (Table 2) The actual activities of

the trypsinized pro-tyrosinase and recombinant core

tyrosinase towards l-DOPA (i.e 565 and 230 lmol

dopachromeÆmin)1Æmg protein)1, respectively) compare

favourably with the activities reported for

Strepto-myces antibioticus tyrosinase, which are 1000

dopa-chromeÆmin)1Æmg protein)1 [41] The Km values for

these two preparations towards l-DOPA (7.9 and

7.6 mm, respectively) are also similar to those

report-ed for the S castaneoglobisporus enzyme (8.1 mm)

but substantially higher than that reported for the

A bisporus enzyme (0.8 mm) [42] However, the Km

values for l-tyrosine (258 and 280 lm, respectively) were similar to that of the A bisporus enzyme (270 lm) [42]

Role of Phe453 in the pro-tyrosinase C-terminal The inhibitory effect of the C-terminal extension found

in some plant polyphenol oxidases has been hypothe-sized to be a result of the presence of an amino acid that occludes the active site This idea has been proposed because of similarities in the structures of the C-terminals of the related family of haemocyanins to plant polyphenol oxidases [3] The crystal structure of octopus haemocyanin shows that a leucine (Leu2830) residue is present near the active site and acts as a

‘blocking residue’ [43] This ‘blocking residue’ prevents substrate molecules from entering the active site, although oxygen can freely diffuse in and out, allowing oxygen transport to be the primary function of this protein However, upon denaturation with SDS or proteolysis, it has been observed that tyrosinase-like activities can be introduced into haemocyanins and this has been proposed to occur via movement of the

‘blocking residue’ [44] A leucine or similar hydropho-bic residue in an equivalent position has also been demonstrated to be present by sequence alignments

of plant polyphenol oxidases [3] In the case of the catechol oxidase from Ipomea, molecular modelling of the C-terminal domains was used to propose Leu439

as the ‘blocking residue’ [45]

Using a similar process of sequence alignment, we hypothesized that the functional equivalent of this blocking residue in V spinosum pro-tyrosinase is Phe453 Thus, we constructed a pro-tyrosinase mutant carrying an alanine at this position, F453A Curi-ously, an increase in protein expression was obtained for this mutant tyrosinase similar to that obtained when the entire C-terminal extension was removed (i.e that of the core tyrosinase; Fig 3B, lanes 1–3) It can be seen from the results shown in Table 3 that this variant had a higher activity than wild-type pro-tyrosinase, as would be expected if the amino acid residue at this position has the aforementioned block-ing function However, the level of increase is very modest (approximately three-fold) compared to a vari-ant in which the C-terminal domain was removed completely by trypsin digest (50- to 100-fold) How-ever, it should be noted that copper analysis revealed that this mutant was very underloaded with copper (only 1.2 equivalents per mole rather than the expected 2) It could be reasonably expected that a higher level of loading would allow much greater levels of activity

The importance of Phe453 in pro-tyrosinase is also

indicated by the fact that we could not induce

wild-type pro-tyrosinase to form its active oxy complex, as

indicated by its absorbance spectrum, whereas the

F453A mutant, similar to the recombinant core

tyrosi-nase, readily formed this complex (Fig S4) These

results suggest that the Phe453 residue is in the close

vicinity of the enzyme active site and plays some role

in oxygen binding

Nonfunctional tyrosinase mutants

To further investigate the function of various other

amino acids in V spinosum tyrosinase, we also

con-structed two further mutants The importance of

Arg40 as a potential residue interacting with Tyr347

and Tyr348 was tested by changing the arginine to an

alanine However, the mutation abolished the

expres-sion of recombinant protein completely (not shown),

which could indicate this residue is vital for protein

stability

We also attemted to test whether Cys84 of the

V spinosumtyrosinase has a similar role in forming an

alkane thiol bond, as has been shown for the sweet

potato catechol oxidase [8] or the N crassa tyrosinase

enzyme [15]; therefore, this residue was mutated to a

serine in the pro-tyrosinase Unlike in A oryzae, where

a similar mutation resulted in a loss of activity but not

of expression [46], we found that this mutation resulted

in a complete loss of detectable protein This suggests

that the residue is essential for correct folding and

expression of the enzyme This appeared to contradict

the results obtained with the A oryzae enzyme;

how-ever, it should be noted that this is a unique tyrosinase

that has a novel acid-induced self-activation

mecha-nism [47] Furthermore, it has been shown to change

from a tetramer in the pro-form to a disulfide-linked

dimer in the mature form Because the V spinosum

pro-tyrosinase, its trypsinized form and the recombinat

core domain were all found to be monomeric, they are

probably not directly comparable to the A oryzae

enzyme (Fig S2)

Verrucomicrobium spinosum tyrosinase as an

alternative model bacterial enzyme

In summary, we present a system that allows the

expression of high levels of a novel bacterial

tyrosi-nase This system has the advantage of an accessory

copper chaperone not needing to be expressed for

copper reconstitution because the protein can be

expressed in the apo-form and reconstituted after

purification The expression and purification of the

apo-form prevents melanin formation during culture growth, which greatly simplifies downstream process-ing and improves protein yields The resultprocess-ing enzyme preparations have been demonstrated to have high lev-els of tyrosinase activities provided the inhibitory C-terminal domain is removed either by proteolysis or recombinant expression The recombinant V spinosum tyrosinase constructs should prove useful for the investigation of non-streptomyces type tyrosinases and may also allow the determination of a crystal structure

of a tyrosinase in its low activity pro-form as well as the solution of a structure that is not in complex with

a caddie protein

Experimental procedures

Materials Chemicals and proteins were purchased from Sigma-Aldrich, molecular biology reagents from Fermentas GmbH (Le Mont-sur-Lausanne, Switzerland) and oligonucleotides from Microsynth AG (Balgach, Switzerland) Chromatogra-phy resins and columns were purchased from GE Health-care Europe GmbH (Bjo¨rkgatan, Sweden)

Molecular biology and molecular cloning Verrucomicrobium spinosum(strain No 4136) was obtained from DSMZ GmbH (Braunschweig, Germany) and cul-tured under the recommended conditions [48] Primers Ver-rucFP01 and VerrucRP01 were used to amplify the tyrosinase gene (Pubmed Locus Tag VspiD_010100001190) and were designed using the draft genome from TIGR (Project ID: 10620) The full-length gene was then cloned into the BamHI and HindIII sites of pUC18 and the sequence verified using the Synergene Biotech GmbH (Zurich, Switzerland) sequencing service Mutants were made using standard PCR techniques or Quikchange (Stratagene, La Jolla, CA, USA) mutagenesis using the primers listed in Table S1

Protein expression For protein expression, the full-length tyrosinase insert or mutants thereof were sub-cloned into the EcoRI and HindIII sites of the pQE60 vector (Qiagen AG, Hom-brechtikon, Switzerland) using the VerrucRBSFP01⁄ VerrucRP01 or VerrucRBSFP02⁄ VerrucRP01 primer pairs The resulting plasmids (Table 1) were transformed into

E coli strain DH5a Constructs were tested for melanizing activity by streaking transformed cells onto M9-agar plates [19] containing 100 lm CuSO4, 1 mm isopropyl thio-b-d-galactoside, 100 lgÆmL)1 ampicillin, 1% glycerol and 0.5 mgÆmL)1 (2.76 mm) l-tyrosine The plates were then

incubated at 37C overnight and visually checked the next

day for the formation of melanin

For 1 L scale expression of the pro-tyrosinase and its

F453A mutant, a 30 mL overnight culture was grown from

a single transformant in LB [49] + 1% glucose +

ampicil-lin 100 lgÆmL)1 The overnight culture was used to

inocu-late (1 : 50) 2· 500 mL of M9 + medium, containing: M9

salts, 1% peptone, 1% glycerol and 1% glucose, 100 lm

calcium (CaCl2), 2 mm magnesium (MgSO4), 100 lm

thia-mine and 100 lgÆmL)1 ampicillin in 2· 2 L Erlenmeyer

flasks This culture was grown at 37C with shaking at

180 r.p.m for 4–5 h, D600 0.5, then 1 mm isopropyl

thio-b-d-galactoside and 100 lgÆmL)1 ampicillin was added and

growth continued for another 20 h

Expression of the recombinant core domain of tyrosinase

was performed using modified autoinduction media [50]

A 30 mL overnight culture was grown from a single

trans-formant in LB + 1% glucose + ampicillin 100 lgÆmL)1

The overnight culture was used to inoculate (1 : 50)

2· 500 mL of auto induction media: 1% N-Z-amine, 0.5%

yeast extract, 25 mm Na2HPO4, 25 mm KH2PO4, 50 mm

NH4Cl, 5 mm Na2SO4, 1% glycerol, 0.4% lactose, 0.5%

glucose, 100 lm CaCl2, 2 mm MgSO4, 100 lm thiamine and

100 lgÆmL)1ampicillin in 2· 2.5 L full baffle Tunair flasks

(Shelton Scientific, Shelton, CT, USA) This culture was

grown at 37C with shaking at 160 r.p.m for 24 h

Protein purification

Cells were harvested by centrifugation and washed in 0.1 m

Tris-HCl (pH 8) The washed cell pellet was resuspended

using 2 mL of 0.1 m Tris-HCl (pH 8) per gram wet weight

of cells, to which lysozyme was added to 1 mgÆmL)1 Cells

were incubated for 1 h on ice and then frozen at )80 C

Cells were then thawed and sonicated with a Branson

soni-fier cell disruptor (Branson Ultrasonics Corp., Danbury,

CT, USA), equipped with a 13 mm tip on 50% power

using five 20 s bursts The sample was then centrifuged at

50 000 g for 30 min To the soluble fraction, 0.6 g⁄ mL of

NH4SO4 was then added and the sample centrifuged at

50 000 g for 30 min The resulting pellet was dissolved in

20 mL of 0.1 m Tris-HCl (pH 8) and dialysed against 5 L

of 10 mm Tris-HCl (pH 8) for 2 h, at which point the

buf-fer was exchanged and dialysis continued overnight The

dialysed sample was then centrifuged at 50 000 g for

30 min The desalted sample was then passed over a

160 mL bed volume Q-Sepharose fast flow column (GE

Healthcare Europe GmBH) and the unbound fraction

con-taining tyrosinase collected, running 10 mm Tris-HCl buffer

(pH 8) Tyrosinase containing fractions were then pooled

and concentrated to 5 mL and loaded onto Superdex 75

16⁄ 60 gel filtration column (GE Healthcare Europe

GmBH), 120 mL bed volume, running 10 mm

Tris-HCl + 0.1 m NaCl buffer (pH 8) Tyrosinase containing

fractions were then pooled concentrated to 10 mgÆmL)1

and stored at –80C in 100 lL aliquots All purification steps were performed using an A¨KTA purifier 100 FPLC (GE Healthcare Europe GmbH) The calculated extinction coefficients at 280 nm were used to measure the concentra-tion of the purified proteins (Table 1)

Size determination For analytical gel filtration, a Superdex 75 16⁄ 60 column was used, 120 mL bed volume, running 10 mm Tris-HCl + 0.1 m NaCl buffer (pH 8) A calibration curve for size determination was made using blue dextran (2 MDa) and the proteins: horse heart cytochrome c (12.4 kDa), horse heart myoglobin (17 kDa), bovine b-lactoglobulin (35 kDa), ovalbumin (44.3 kDa) and bovine serum albumin (67 kDa) (Fig S2) The sizes of purified proteins was also determined using the mass MS service of the ETH func-tional genomics centre Zurich (http://www.fgcz.ethz.ch/)

Enzyme assay Kinetic characterization of l-tyrosine and l-DOPA oxidation was measured by dopachrome formation [51] at 475 nm using

a molar extinction coefficient of 3600 M)1Æcm)1at 25C in

3 mL of 0.1 m potassium phosphate buffer (pH 6.8) using a stirred Peltier assembly, with the spectra being monitored on

a Cary 50 bio UV⁄ visible spectrophotometer (Varian Inc., Zug, Swizerland) Kinetic parameters were calculated using prism5(GraphPad Software Inc., San Diego, CA, USA)

Bioinformatics The molecular mass and theoretical extinction coefficient of the various proteins were calculated using the protparam tool available through the ExPasy server (http://www.exp-asy.ch/tools/protparam.html) [52] The signalP server was used for signal peptide prediction (http://www.cbs.dtu.dk/ services/SignalP/) [53]

Copper reconstitution Purified apo-tyrosinase was reconstituted with copper by mixing an aliquot of protein ( 10 mg) with an equal vol-ume of 10 mm Tris-HCl (pH 8), containing a three-fold molar excess of CuSO4in a final volume of 1 mL The sam-ple was incubated on ice for 1 h and then dialysed twice against 1 L of 10 mm Tris-HCl buffer (pH 8)

Copper analysis The copper concentration of the protein samples was mea-sured using a slight modification of the biquinoline method [54] Briefly 100 lL of 10 mgÆmL)1 protein sample was added to 0.2 mL of 0.1 m sodium phosphate