Báo cáo khoa học: On the reaction of D-amino acid oxidase with dioxygen: O2 diffusion pathways and enhancement of reactivity docx

On the reaction of D-amino acid oxidase with dioxygen:
O2 diffusion pathways and enhancement of reactivity
Elena Rosini, Gianluca Molla, Sandro Ghisla and Loredano Pollegioni
`
Dipartimento di Biotecnologie e Scienze Molecolari, Universita degli Studi dell’Insubria, and The Protein Factory, Centro Interuniversitario di
`
Biotecnologie Proteiche, Politecnico di Milano and Universita degli Studi dell’Insubria, Varese, Italy

Keywords
flavoproteins; mutagenesis; oxidases;
oxygen diffusion; oxygen reactivity
Correspondence
L. Pollegioni, Dipartimento di Biotecnologie
`
e Scienze Molecolari, Universita degli Studi
dell’Insubria, Varese, Italy
Fax: +39 332 421500
Tel: +39 332 421506
E-mail:
(Received 19 July 2010, revised 2
November 2010, accepted 20 November
2010)
doi:10.1111/j.1742-4658.2010.07969.x

Evidence is accumulating that oxygen access in proteins is guided and controlled. We also have recently described channels that might allow access
of oxygen to pockets at the active site of the flavoprotein D-amino acid oxidase (DAAO) that have a high affinity for dioxygen and are in close proximity to the flavin. With the goal of enhancing the reactivity of DAAO
with oxygen, we have performed site-saturation mutagenesis at three positions that flank the putative oxygen channels and high-affinity sites. The
most interesting variants at positions 50, 201 and 225 were identified by a
screening procedure at low oxygen concentration. The biochemical properties of these variants have been studied and compared with those of wildtype DAAO, with emphasis on the reactivity of the reduced enzyme species
with dioxygen. The substitutions at positions 50 and 225 do not enhance
this reaction, but mainly affect the protein conformation and stability.

However, the T201L variant shows an up to a threefold increase in the rate
constant for reaction of O2 with reduced flavin, together with a fivefold
decrease in the Km for dioxygen. This effect was not observed when a
valine is located at position 201, and is thus attributed to a specific alteration in the micro-environment of one high-affinity site for dioxygen
(site B) close to the flavin that plays an important role in the storage of
oxygen. The increase in O2 reactivity observed for T201L DAAO is of
great interest for designing new flavoenzymes for biotechnological applications.

Introduction
Flavins are highly versatile co-factors of flavoproteins
that catalyze a wide array of chemical and photochemical processes [1–3]. As a prominent member of this family, d-amino acid oxidase (DAAO, EC 1.4.3.3) is a
homodimeric enzyme found in all eukaryotic cells,
where it fulfils various roles [4]. It is the archetype of the
oxidase ⁄ dehydrogenase class of flavoproteins [1]; each
subunit contains one non-covalently bound FAD mole˚
cule > 10 A below the surface [5]. DAAO catalyzes net
hydride transfer from the aC–H bond of neutral

d-amino acids (and of basic d-amino acids, but with
lower efficiency) to FAD (on the Re side) in the reductive half-reaction (Scheme 1a) and oxidation of reduced
co-factor (FADH)) by O2 in the oxidative half-reaction,
forming H2O2 as a product (Scheme 1b,c) [6,7].
With some notable exceptions [8], research into
the mechanistic details of the reaction of (reduced)
flavoprotein oxidases with dioxygen has long been
neglected. The reasons for this were mainly due to
experimental limits, such as conversion of the species in

Abbreviations
DAAO, D-amino acid oxidase (EC 1.4.3.3); E-Flox, oxidized enzyme form; E-Flred, reduced enzyme form; E-Flred–IA, reduced enzyme–imino

acid complex; IA, imino acid.

482

FEBS Journal 278 (2011) 482–492 ª 2010 The Authors Journal compilation ª 2010 FEBS


Modulation of O2 reactivity in D-amino acid oxidase

E. Rosini et al.

k1
(a) E-Flox + AA

k–1

(b) E-Flred + O2

(c) E-Flred–IA + O2

k6

k5

k2
E-Flox–AA

E-Flred–IA

k–2

k–5

E-Flred + IA

E-Flox
k3

E-Flox–IA

k4

E-Flox–IA

(d) et/v = Φ0 + ΦD-Ala / [D-Ala] + ΦO2 / [O2] + ΦD-ALa,O2 / [D-Ala]•[O2]

et/v = [(k2 + k4)/(k2•k4)] + [(k–1 + k2)/(k1•k2•[D-Ala])] + [(k2 + k–2)/(k2•k3•[O2])] + [(k–1 + k–2)/(k1•k2•k3•[D-Ala]•[O2])]
where: kcat = 1/Φ0; Km,D-Ala = ΦD-Ala/Φ0; Km,O2 = ΦO2/Φ0
For wild-type DAAO (k4»k2), the expressions can be simplified:
kcat = [k2•k4/k2 + k4)] ≈ k2
Km,O2 = [k4•(k2 + k–2)]/[k3•(k2 + k4)] ≈ k2/k3
Km,D-Ala = [k4•(k–1 + k2)]/[k1•(k2 + k4)] ≈ (k–1 + k2)/k1
Scheme 1. Kinetic steps in the catalytic cycle proposed for DAAO [6,7]: (a) reductive half-reaction, (b,c) oxidative half-reaction, and (d) correlation between steady-state kinetic parameters and single rate constants, and their reduction for wild-type DAAO [7].

the absence of observable intermediates. In oxidases,
the reaction of reduced flavin with O2 proceeds through
an electron-transfer step that generates a caged radical
pair. This is generally thought to be rate-limiting.
Reduced flavoprotein oxidases (and monooxygenases)
react rapidly with dioxygen (exhibiting bimolecular rate

constants up to 106 m)1Ỉs)1) and show no saturation
with O2 (i.e. the Kd value for oxygen is much larger
than the maximal O2 concentration in solution) [6–9].
These enzymes appear to ‘consume’ O2 without highaffinity binding.
Because of the scarcity of information concerning
the reaction of flavoproteins with molecular oxygen,
we previously used a directed evolution approach to
enhance the affinity for dioxygen of DAAO, specifically to generate optimized enzyme variants for use in
biocatalysis or medical applications [10]. The
S19G ⁄ S120P ⁄ Q144R ⁄ K321M ⁄ A345V DAAO variant
has increased activity at low O2 concentrations, resulting from a 10-fold lower Km;O2 value, although the
rate constant for reduced flavin re-oxidation was only
marginally affected [10]. Recently, however, important
advances have been made in the field: as with hemedependent enzymes [11–13], specific paths within the
protein matrix of flavoprotein oxidases have been
identified that serve to channel O2 to its destination
[14,15]. We have described funnels that lead to pockets at the active site of DAAO, in particular two
regions (sites A and B) that have the highest affinity
for dioxygen inside the protein and are in close prox-

imity to the Si side of the flavin, and determined the
most likely diffusion pathways (Fig. 1) [16]. The
energy required to place an O2 molecule at site A or B
is  15 kJỈmol)1 lower than the corresponding energy
required to place O2 in the solvent; thus, the probability of finding a molecule of O2 at these sites is correspondingly higher. This corresponds to a virtual [O2]
that is  1000-fold higher than that in an equivalent
solvent volume. The local lower dielectric constant
might play a main role in the apparent [O2] increase.
Site A is in close proximity with the C(4a) position of
˚

the isoalloxazine ring of FAD (at  3.5 A), an ideal
location for efficient oxygen reactivity. Site B is located
˚
 5 A from the xylene ring of the flavin and is thus
also suitable for electron transfer. The importance of
the predicted site A was tested experimentally by
mutating Gly52, i.e. by partially filling the space
expected to be occupied by O2: the G52V DAAO variant shows a 100-fold lower oxygen reactivity [16].
The present study represents an extension of these
previous studies with the goal of enhancing the
reactivity of DAAO with dioxygen. It is based on the
predictions of the implicit ligand sampling analysis
[16], which has identified several residues in the oxygen channels. The most interesting substitutions at
positions 50, 201 and 225 were identified by a screening procedure at low oxygen concentration on mutant
libraries prepared by site-saturation mutagenesis.
These studies identified a DAAO variant at position
201 that reacted more efficiently with dioxygen and

FEBS Journal 278 (2011) 482–492 ª 2010 The Authors Journal compilation ª 2010 FEBS

483


Modulation of O2 reactivity in D-amino acid oxidase

E. Rosini et al.

A

B

Fig. 1. (A) O2 channel connecting bulk solvent to the Si face of FAD at the DAAO active site [16]. Mutated residues described in the text, the
product imino acid (IA) and FAD are shown using CPK representation. Water molecules filling the channels (green) are shown using VdW
representation (50% VdW radius). (B) Position of the mutated residues with respect to the O2 high-affinity sites A (green) and B (blue). The
product (imino pyruvate, IA) and the water molecules were modeled into the DAAO structure (PDB code 1c0p) as described previously [16].

provided further insight into the oxygen reactivity of
DAAO.

Results
Identification of residues possibly involved in
oxygen binding
Recent computational implicit ligand sampling studies
have discovered a small channel (filled by several water
molecules) that leads from the bulk solvent to the Si
face of the flavin at the active site of yeast DAAO
(PDB code 1c0p; Fig. 1A) and is fundamental for O2
access to the active site during turnover [16]. The W50
and T201 residues are aligned along this channel, and,
in particular, the side chain of residue W50 is oriented
toward the outer part of the channel toward the bulk
solvent (Fig. 1B). We have hypothesized that the modifying size and polarity of this side chain could result in
major alterations in O2 accessibility through the proposed channel. A further important residue, T201, is
located in the inner part of the channel, close to the
˚
benzene ring of FAD (3.7 A) and in close proximity to
the two regions with the highest affinity for oxygen
(sites A and B in Fig. 1) [16]. T201 is thus a further
candidate as a target for the study of O2 interactions.
The third such residue is I225, the side chain of which,

together with the side chain of M213, forms a large
part of the DAAO active site roof; it is located at a
˚
distance of  4.5 A from the FAD C(4)=O locus.
However, it is possible that mutations at this site could
also affect the activity ⁄ substrate specificity of the
enzyme [16–18].
Libraries of DAAO variants at positions 50, 201
or 225 were generated by site-saturation mutagenesis,
484

and two or three variants for each single position
showing altered oxygen reactivity were selected by a
screening procedure performed at low oxygen concentration (2.5%). The W50F, T201L and T201V
variants were identified because of higher activities
than the wild-type DAAO, the I225F and I225V
variants were identified as the most active clones at
this position 225, while the W50R and W50P variants were isolated because they exhibit no activity.
General properties of DAAO variants
All the purified recombinant DAAO variants are
homodimeric 80 kDa holoenzymes, as determined by
gel-permeation chromatography and spectral analyses,
and in the oxidized state show the typical spectrum of
FAD-containing flavoproteins (i.e. absorbance maxima
at  455 and 375 nm, an e455 nm of  12 600 m)1Ỉcm)1,
and an A274 nm ⁄ A455 nm ratio of  8.2; Appendix S1
and Fig. S1). Free FAD is not found in all purified
enzyme preparations, indicating preservation of the
strong interaction between the co-factor and the apoprotein moiety. The substitutions introduced at position 50 alter the tertiary structure of DAAO (Fig. S2)
as well as the protein stability: the W50 DAAO variants are less thermostable than wild-type enzyme

(Table S1). The redox properties of the flavin co-factor
are also altered by the W50P substitution: an Em of
) 207 ± 6 mV was determined for the W50P variant
versus ) 109 mV for wild-type DAAO [19]. The conformation and flavin properties of DAAO variants at
position 201 and 225 are not significantly affected by
the substitutions introduced, with only a slight alteration in the ability to stabilize the flavin semiquinone
species (Appendix S1).

FEBS Journal 278 (2011) 482–492 ª 2010 The Authors Journal compilation ª 2010 FEBS


Modulation of O2 reactivity in D-amino acid oxidase

E. Rosini et al.

The apparent kinetic parameters of DAAO variants
were determined using d-alanine as the substrate by an
oxygen consumption assay at 21% oxygen saturation
and 25 °C (Appendix S1 and Table S2). In comparison
to wild-type DAAO, the most significant changes were
apparent for the W50P (lower kcat,app and higher
Km,app), W50R (lower d-alanine affinity) and I225F
(lower kcat,app) variants. A comparison of the substrate
specificity of the I225F variant with respect to the
wild-type DAAO is shown in Table S3: these results
support the conclusion that the side chain at this position contributes to the d-amino acid preference of
yeast DAAO.
Kinetic mechanism of DAAO variants
Steady-state kinetics
Dependence of the catalytic activity of DAAO variants

on the oxygen and d-alanine concentrations was
assessed using the enzyme-monitored turnover method
[6,7,20]. Air-saturated solutions of DAAO and d-alanine were mixed in a stopped-flow instrument, and
absorbance spectra were recorded continuously in the
300–700 nm range at 15 °C [16]. During turnover, all
DAAO variants are largely present in the oxidized
form, and the spectrum of the reduced enzyme is
observed only at the end of the observation time, i.e.
when the O2 concentration becomes very low (Fig. 2A
and Fig. S3). This is consistent with the steps involving
oxidation of reduced DAAO by oxygen being faster
than those involved in reduction, as observed for wildtype and other variants of DAAO [7,10,16]. The results
of these steady-state measurements (at saturating O2
and d-alanine concentrations; Scheme 1 and Table 1)
show a decrease in kcat for all the variants with the
exception of the T201V DAAO; the variant with the
lowest kcat is the W50P DAAO (compare traces in
Fig. S3A). The W50P and W50R variants show a
lower affinity for d-alanine compared to wild-type
DAAO, and a decrease in the Km;O2 value is apparent
for W50P, W50R, T201L and I225V enzymes. For
the I225V DAAO, although the corresponding
Lineweaver–Burk (double-reciprocal) plots show a
set of parallel lines (not shown), secondary plots of
the reciprocals of the x and y intercepts from the
Lineweaver–Burk plot (apparent kcat and Km;O2 , respectively) against [d-alanine] show an unprecedented sigmoidal dependence on d-amino acid concentration
(Fig. 2B,C). The steady-state parameters for I225V
DAAO determined by using the extreme kinetic Hill
coefficient for cooperativity (h = 2.2) [21] are listed in
Table 1. In contrast, the I225F variant does not show

any sigmoidal behavior, but an  20-fold decrease

Fig. 2. Steady-state kinetics of the I225V variant of DAAO. (A) The
kinetic data were determined by the enzyme-monitored turnover
method (Fig. S3), using D-alanine (at 0.4, 0.5, 0.8 and 1.0 mM) and
0.25 mM oxygen [20], by monitoring the time course of the flavin
oxidation state based on its absorbance at 455 nm [6,7] at pH 8.5
and 15 °C. (B,C) Tertiary plots of the reciprocal of the y intercepts
(B) and the x intercepts (C) as calculated from Lineweaver–Burk
plots obtained from the experimental traces shown in (A).
Experimental values were fitted using a hyperbolic fit using a Hill
coefficient of 2.2 [21].

FEBS Journal 278 (2011) 482–492 ª 2010 The Authors Journal compilation ª 2010 FEBS

485


Modulation of O2 reactivity in D-amino acid oxidase

E. Rosini et al.

Table 1. Comparison of steady-state kinetic parameters for wild-type and variants of DAAO using D-alanine as substrate and at 15 °C. Data
were obtained in 50 mM sodium pyrophosphate buffer, pH 8.5, 1% glycerol and 0.25 mM 2-mercaptoethanol. The steady-state F parameters
are defined in Scheme 1.
Lineweaver–Burk
plot behavior
Wild-type [10]
W50F
W50P

W50R
T201L
T201V
I225F
I225V

G52V [16]

Convergent
Parallel
Parallel
Parallel
Parallel
Parallel
Parallel
Parallel
(sigmoidal
n = 2.2)
Convergent

kcat (s)1)

FD-alanine (MỈs)1)

330 ±
190 ±
 4.0
62 ±
170 ±
365 ±

15 ±
98 ±

(0.8
(2.0
(3.6
(1.7
(3.6
(1.7
(8.6
(8.0

30
20
6
15
50
2.1
6

0.33 ± 0.03

±
±
±
±
±
±
±
±

0.1)
0.3)
0.3)
0.2)
0.5)
0.3)
0.4)
0.4)

·
·
·
·
·
·
·
·

10)5
10)5
10)3
10)3
10)5
10)5
10)5
10)5

(1.2 ± 0.2) · 10)5

in kcat compared to wild-type DAAO is apparent
(Fig. S3C).

Km,D-alanine (mM)
2.6
3.9
14.4
10.4
6.2
6.9
1.2
0.9

±
±
±
±
±
±
±
±

0.4
0.7
3.4
1.5
0.8
0.7
0.1
0.1

0.036 ± 0.005

FO2 (MỈs)1)
(5.0
(7.7
(50
(8.8
(2.3
(9.4
(85
(11

±
±
±
±
±
±
±
±

0.1) · 10)6
0.4) · 10)6
4) · 10)6
1.7) · 10)6
0.05) · 10)6
0.1) · 10)6
4) · 10)6
1) · 10)6

(4.4 ± 1.0) · 10)4

Km;O2 (mM)
1.9
1.4
0.16
0.54
0.40
3.9
1.2
0.8

±
±
±
±
±
±
±
±

FD-alanine,O2
(M2Ỉs)1 · 10)9)

0.1
0.5
0.04
0.18
0.05

0.1
0.4
0.1

3.0 ± 0.2

0.15 ± 0.03

12.1 ± 1.2

A

Reductive half-reaction
The reductive half-reaction (Scheme 1a) was studied
for wild-type and variants of DAAO, using d-alanine
as the substrate under anaerobic conditions at 15 °C,
and monitoring the absorbance changes [7,15]. In all
cases, the oxidized form of the enzyme is rapidly
converted to the reduced enzyme–imino acid complex
(E-Flred–IA in Scheme 1; phase 1, kobs1), an intermediate that is then converted at a slower rate into free
fully reduced enzyme (phase 2, kobs2) (Fig. 3). For all
DAAO variants, the dependence of kobs1 values on
[d-alanine] shows curvature: there is ample evidence
for a hyperbolic dependence of kobs1 on [d-alanine] for
various DAAOs, and it represents a second-order process (formation of an initial enzyme–substrate complex), followed by a first-order reaction as shown in
Scheme 1a [6,7,22]. As the data are satisfactorily fitted
by a rectangular hyperbola that intersects close to the
origin, the reduction step is practically irreversible (k)2
of  0). A significant change in the rate constant of
flavin reduction was observed only for W50P DAAO

(Table 2), in agreement with the observed decrease in
the kcat value. For the I225F and I225V variants, the
k2 rate constant was two- to threefold slower than for
wild-type DAAO (Table 2): importantly, no indication
of sigmoidal behavior is evident.
The rate for the observed second-phase kobs2, corresponding to product dissociation from E-Flred–IA (k5 in
Scheme 1a) does not depend on [d-alanine], and its
value is  1.3 ± 0.5 s)1 for wild-type and variants of
DAAO. The main exception is the W50P variant, for
which the k5 value is estimated to be £ 0.1 s)1 (Table 2).
486

B

Fig. 3. Reductive half-reaction of wild-type and variants of DAAO.
Comparison of time courses of flavin reduction followed at 455 nm
(vertical bars = experimental data points) for W50 variants (A) and
T201L ⁄ V and I225V variants (B) versus wild-type DAAO. The
enzymes ( 8 lM) were reacted under anaerobic conditions with
0.25 mM D-alanine at pH 8.5 and 15 °C (D-alanine concentration of
5 mM for W50P). The rate constants were obtained by fitting using
a double exponential equation (continuous line). The rates are listed
in Table 2.

FEBS Journal 278 (2011) 482–492 ª 2010 The Authors Journal compilation ª 2010 FEBS


Modulation of O2 reactivity in D-amino acid oxidase

E. Rosini et al.

Table 2. Rate constants for the reductive and the oxidative half-reaction of wild-type and variants of DAAO estimated from rapid reaction
methods at 15 °C. For the reductive half-reaction, the parameters were obtained using D-alanine as substrate; for the oxidative half-reaction,
the re-oxidation was started from the free reduced enzyme species or the imino acid complex (Scheme 1b,c). The rate constants refer to
those defined in Scheme 1. Data were obtained in 50 mM sodium pyrophosphate buffer, pH 8.5, 1% glycerol and 0.25 mM 2-mercaptoethanol.
Reductive half-reaction

Oxidative half-reaction
a

)1

kobs1 ( k2) (s )
Wild-type
W50F
W50P
W50R
T201L
T201V
I225F
I225V

c

G52V

‡ 250
170 ±
3.5 ±
130 ±

250 ±
‡ 250
79 ±
160 ±

30
0.4
20
15
4
21

‡ 550

)1

Kd (k)1 ⁄ k1) (mM)
2.1
1.2
14.6
2.3
3.1
2.6
0.5
1.7

±
±
±
±

±
±
±
±

0.5
0.3
2.6
0.3
0.2
0.4
0.1
0.3

kobs2 ( k5) (s )
1.2
1.7
£ 0.1
0.7
1.8
0.8
1.3
1.4

2

± 0.2
± 0.2
±
±

±
±
±

0.2
0.5
0.2
0.2
0.2

1.6 ± 0.2

k3 from E-Flred–IA
(M)1Ỉs)1) · 105

2.4 ±
2.3 ±
n.f.
n.f.
4.2 ±
1.4 ±
1.4 ±
1.0 ±

0.3 [3.0 · 104]
0.1 [2.9 · 104]

0.4 [9.0 · 104]
0.2 [3.8 · 104]
0.1 [0.7 · 104]

0.1

NF (0.024 ± 0.008)

b

k6 from E-Flred
(M)1Ỉs)1) · 104
3.6 ± 0.5
3.8 ± 0.2
4.0 ± 0.5
2.1 ± 0.3
11.0 ± 0.5
2.9 ± 0.2
1.2 ± 0.2
Biphasic: £ 3 s)1
(60% amplitude);
2.9 ± 0.1 (40%
amplitude)
0.046 ± 0.002

a

Buffer as above containing 20 mM glucose, 20 mM pyruvate and 400 mM NH4Cl. The rate constants of the second (slower) phase of flavin
re-oxidation observed in the presence of imino acid and corresponding to re-oxidation of the free reduced enzyme form (k6 in Scheme 1b)
are shown in parentheses.
b
Buffer as above containing 20 mM glucose.
c
The value estimated from simulation of the steady-state kinetics is shown in parentheses [16].

NF, not feasible (flavin re-oxidation following imino acid addition).

Oxidative half-reaction
The (re)oxidation of reduced DAAO variants by dioxygen (Scheme 1b,c) was also studied using the stoppedflow apparatus. For this, anaerobic solutions of free
reduced enzyme were reacted with buffer solutions
equilibrated at various O2 concentrations (Scheme 1b),
and re-oxidation was monitored by following the
(re)appearance of absorption of the oxidized flavin species. The experimental traces at 455 nm (conversion of
the reduced enzyme form E-Flred into the oxidized
enzyme form E-Flox) are close to those of wild-type
for the W50F, W50P and T201V variants, slightly
slower for W50R and I225F variants, and appreciably
faster for T201L DAAO (Fig. 4). The time course of
(re)oxidation is monophasic with the exception of
I225V. The kobs values obtained for re-oxidation are
reported as a function of [O2], yielding a line that does
not indicate saturation with [O2] (data not shown); this
behaviour is assumed to reflect a second-order process.
The slope of this linear fit yields the k6 rate constant,
which is approximately two- to threefold lower for
I225F and W50R DAAOs and threefold higher for the
T201L variant compared with wild-type DAAO (Fig. 4
and Table 2). For the I225V variant, the experimental
traces of re-oxidation at 455 nm are better fitted using

a two-exponential equation (Fig. 4C): a fast phase is
followed by a second slower phase, with rate £ 3 s)1,
and for which the value and amplitude ( 60% of the
overall absorbance change) do not depend on oxygen
concentration. The kobs values for the faster phase of

re-oxidation of the I225V variant show a linear dependence on [O2], with no indication of oxygen saturation:
the k6 bimolecular rate constant is unchanged for
I225V variants compared to wild-type DAAO
(Table 2).
A similar experiment was performed using the
reduced enzyme E-Flred–IA complex (i.e. the form
present at high concentrations of ammonia and pyruvate; Scheme 1c): in this case, the time course of
re-oxidation is clearly biphasic. A fast phase with an
amplitude corresponding to  50% of the overall
absorbance change at 455 nm is followed by a slower
one, whose rate corresponds to that observed with free
reduced DAAO at the same [O2] (Fig. 4B). From this
we deduce that the first fast phase corresponds to
(re)oxidation of the E-Flred–IA complex present at
equilibrium (Scheme 1c) and the second phase corresponds to the re-oxidation of uncomplexed E-Flred
(Scheme 1b). DAAO variants behave similarly to the
wild-type enzyme: the re-oxidation is still faster for
T201L than for T201V, W50F or wild-type DAAO,

FEBS Journal 278 (2011) 482–492 ª 2010 The Authors Journal compilation ª 2010 FEBS

487


Modulation of O2 reactivity in D-amino acid oxidase

E. Rosini et al.

after adding ammonia and pyruvate to enzyme that
had been anaerobically reduced using an up to 10-fold

molar excess of d-alanine or because flavin was re-oxidized when the photoreduced W50R DAAO was
mixed with the imino acid solution or with classical
inhibitors, such as benzoate and anthranilate. The
reoxidation of E-Flred–IA was also not feasible using
G52 DAAO variants [16].
Therefore, the T201L substitution increases the oxygen reactivity of yeast DAAO, a change that does not
modify the kcat value (i.e. the maximal activity at saturating concentration of both substrates) but does affect
activity at low oxygen concentrations. These results
also indicate that two forms of the free reduced
enzyme species exist for I225V DAAO, and that these
forms affect turnover at low substrate concentrations,
i.e. when there is competition between E-Flred and the
E-Flred–IA complex for O2-induced re-oxidation
(Scheme 1).

A

B

C

Discussion

Fig. 4. Oxidative half-reaction of wild-type and variants of DAAO.
(A) Comparison of time courses of the (re)oxidation of reduced
wild-type and W50 variants of DAAO followed at 455 nm upon mixing of  10 lM reduced enzyme with 153 lM oxygen (vertical
bars = experimental data points). Conditions were as described in
Experimental procedures. The mono-exponential fit of the experimental data is shown as a continuous line. (B) Comparison of
re-oxidation of E-Flred (vertical bars) and the E-Flred–IA complex
(crosses; obtained by adding 20 mM pyruvate and 400 mM ammonium chloride) wild-type and T201L DAAOs by 153 lM oxygen. A

mono-exponential fit was used for the re-oxidation of E-Flred and a
bi-exponential equation was used for re-oxidation of the E-Flred–IA
complex. (C) Comparison of re-oxidation of E-Flred wild-type and
I225V DAAOs by 0.6 mM oxygen. The experimental data points for
I225V were fitted using a mono-exponential equation (dashed line)
or a bi-exponential equation (solid line): the latter gave a better
reproduction. The rates are listed in Table 2.

and is approximately twofold slower for I225F and
I225V variants (Fig. 4B and Table 2). However, the
same experiment was not feasible using W50P or
W50R DAAO variants because flavin was re-oxidized
488

The biochemical (and structural) basis of the capacity of
flavoenzymes to react with dioxygen is still poorly
understood, but represents a very interesting issue in
flavoenzymology. Trajectories and sites of high affinity
for O2 in the yeast DAAO have recently been identified
by molecular dynamics simulations and implicit ligand
sampling methods [16]: a specific dynamic channel for
O2 diffusion leads from the solvent to the flavin Si side
(i.e. the opposite side with respect to the substrate ⁄ product binding site; Fig. 1). In a previous study, we investigated the role of the residue G52: in the G52V variant,
the valine side chain occupies the site that has the highest O2 affinity in wild-type DAAO (site A in Fig. 1),
and the reactivity of reduced G52V DAAO with O2 is
considerably decreased, as well as the turnover number
[16]. Here we have focused on three additional residues
that are potentially involved in oxygen migration as
they flank the putative O2 high-affinity sites.
The substitution of W50 (a residue close to the tunnel entrance; Fig. 1A) with R or P significantly destabilizes DAAO, as is evident from the  10 °C lower

melting temperatures (Table S1), and modifies the CD
and fluorescence spectra (Fig. S2). Unpredictably (as
this residue is distant from the isoalloxazine ring of the
flavin), the redox properties of the W50P variant are
also significantly altered, for example there is no stabilization of the flavin anionic semiquinone and the midpoint potential is  100 mV more negative than in
wild-type DAAO. This substitution also significantly
alters the kinetic properties of the flavo-oxidase:
compared to wild-type DAAO, the W50P variant has

FEBS Journal 278 (2011) 482–492 ª 2010 The Authors Journal compilation ª 2010 FEBS


E. Rosini et al.

a much lower ( 100-fold) maximal activity, an
 12-fold lower Km for oxygen and an increased Km
for d-alanine; the binding of the competitive inhibitor
benzoate is also negatively affected by this substitution
(Table S2). The change in turnover number for W50P
DAAO is mainly due to the slower rate of the reductive half-reaction (the rate constant for flavin reduction
is decreased  100-fold; Table 2), but the rate of flavin
re-oxidation is slightly decreased for W50R DAAO
only ( 1.8-fold slower; Table 2), indicating a change
in the rate-limiting step for catalysis in this latter
variant.
The substitutions introduced at position 225
decreased the enzyme activity (kcat  20-fold lower for
the I225F variant; Table 1), but Km;O2 and the rate
constants for the reduced flavin re-oxidation are only
modified to a limited extent (Tables 1 and 2). The

I225V variant shows an unprecedented sigmoidal
behavior in the kcat versus [d-alanine] and Km;O2 versus
[d-alanine] plots (Fig. 2B,C). The biphasic re-oxidation
process observed using the free reduced enzyme species
(Fig. 4C) possibly explains this, suggesting the existence of two alternative configurations (whose ratio is
close to 1, based on the amplitude of the observed
phases of re-oxidation), one of which shows a very
slow reactivity with dioxygen (£ 3 s)1; Table 2). Similar behavior was previously observed for the F359W
variant of Streptomyces cholesterol oxidase [23], and
was related to kinetic cooperativity known as the mnemonic model [21]. This occurs when the free enzyme
exists in at least two conformations that can react with
the substrate at different rates: the mnemonic model
for a two-substrate, two-product reaction sequence displays kinetic cooperativity with respect to the first substrate but no cooperativity with respect to the other
substrate. In fact, this unusual behavior of I225V
DAAO is lost in the presence of the imino acid product. The reaction of the reduced enzyme forms with
oxygen is negatively affected by substitution of I225:
the most significant change is an approximately threefold decrease in k3 as observed for the I225F variant
versus wild-type DAAO.
With regard to position 201 (close to both highaffinity sites A and B; Fig. 1), the activity of the
T201V variant resembles that of wild-type DAAO (at
both 21% and saturating oxygen concentrations;
Table 1 and Table S2). However, introduction of leucine results in a slight decrease in kcat and k2 values,
and, more interestingly, an approximately fivefold
decrease in Km;O2 , which is accompanied by faster
re-oxidation of the corresponding reduced enzyme
species, up to threefold higher using the E-Flred–IA
complex (Fig. 4B and Tables 1 and 2).

Modulation of O2 reactivity in D-amino acid oxidase

Functional data on flavo-oxidase variants designed
on the basis of molecular dynamic simulations of O2
diffusion have been reported recently for alditol oxi˚
dase [14]. Its 3D structure was determined at 1.1 A
resolution [24], and indicated five putative pathways
that bring O2 molecules in front of reduced FAD
co-factor and a small cavity that may contain O2: symmetrically to that observed for DAAO, this site is
located on the Re side of the FAD co-factor, while
substrate ⁄ product exchange occurs on the Si side. The
changes in kinetic parameters for the A105G variant
of alditol oxidase were minor (see below). Site-directed
mutagenesis designed to block individual routes had
little effect on the kcat ⁄ Km ratio in copper-containing
amine oxidase [13], but significantly affected cholesterol oxidases [23,25]. These results suggest that multiple pathways are employed by dioxygen to reach the
active site, as also suggested by our results for DAAO.
Of the residues modified in DAAO, G52 appeared to
play the major role in O2 reactivity (up to a 100-fold
decrease in k3 and k6 rate constants for G52V compared to the wild-type enzyme; Table 2), while W50
and I225 appeared to mainly affect the conformation
of the flavoprotein. On the other hand, an increase in
the rate constant for reduced flavin re-oxidation was
observed for the T201L DAAO variant. Sites A and B
˚
are in close proximity ( 8 A apart), and are connected through a pathway that has a low activation
energy barrier [16]: site B appears to increase the effective dioxygen concentration in the proximity of site A.
From a structural point of view, the T201L mutation
results in substitution of a small and polar side chain
by a large hydrophobic one whose d-methyl groups are
very close to site B and the FAD xylene ring
(Fig. 5A,B). Local alteration of the hydrophobicity

close to site B could increase the affinity of this site for
O2, and, as a consequence, the reactivity of the T201L
DAAO variant with molecular oxygen. A similar effect
was not observed with the T201V variant, as the distance between the valine side chain c-methyl groups
and site B is expected to be larger than that in T201L
˚
˚
DAAO ( 3.5 A versus  2.0 A; Fig. 5B,C).
The threefold increase in oxygen reactivity observed
for the T201L DAAO is a meaningful difference, as
improvements in oxygen reactivity of a similar extent
for ‘efficient flavo-oxidases’ are uncommon. Limited
changes have been observed for various flavo-oxidases
(e.g. a 1.5-fold increase for the A105G variant of alditol
oxidase and a 2.2-fold increase for the E475Q variant of
cholesterol oxidase) [14,25]. The opposite alterations in
O2 reactivity for variants at position 52 and 201 confirm
the importance of inferred oxygen channels in DAAO,
and are in agreement with the ‘storage’ role proposed

FEBS Journal 278 (2011) 482–492 ª 2010 The Authors Journal compilation ª 2010 FEBS

489


Modulation of O2 reactivity in D-amino acid oxidase

A

E. Rosini et al.

B

C

Fig. 5. Detail of the residues surrounding oxygen high-affinity site B. (A) Wild-type DAAO, (B) T201V DAAO, and (C) T201L DAAO. Steric
hindrance of atoms is shown as molecular surface colored by atom type. The distance between the residue 201 side chain and the center
of site B is shown as a dotted line. Residue 201 is labeled in bold. IA: imino acid product (imino pyruvate) modeled into the DAAO active site
(PDB code 1c0p) as described previously [16]. Models of DAAO mutants were prepared using VMD [31]. Molecular surfaces were calculated
using the MSMS program [32].

for site B. Our results give direction to enhance the
robustness of existing O2-consuming flavo-oxidases in
order to design new catalysts for novel biotechnological
applications, e.g. for evolution of a flavo-oxidase useful
in enzyme pro-drug cancer therapy [10].

type). A significantly lower yield was achieved for the
W50P variant DAAO (0.6 mg per g cell paste and 2 mgỈL)1
of fermentation broth). The expression of DAAO mutants
at positions 201 and 225 was similar to that of the wildtype DAAO ( 2 mg enzyme per g cell paste and
 11 mgỈL)1 of fermentation broth).

Experimental procedures

Spectral properties

Site-saturation mutagenesis and enzyme
expression and purification

Extinction coefficients of the oxidized form of DAAO variants were determined by heat denaturation of the enzymes
(at 95 °C for 10 min) and using the absorption coefficient
for free FAD of 11.3 mm)1Ỉcm)1. Semiquinone formation
was achieved by light irradiation (using a 250 W lamp at a
distance of  20 cm) with anaerobic enzyme solutions
( 10 lm) containing 5 mm EDTA and 0.5 lm 5-deazaflavin [26]. The amount of the thermodynamically stable semiquinone form was evaluated after incubation for 24 h at
4 °C or after adding 5 lm benzyl viologen to the enzyme
solution [27]. Redox potentials were estimated by the dye
equilibration method [19,28]. The dissociation constants for
sulfite and benzoate binding to DAAO ( 10 lm) were
assessed spectrophotometrically by following the changes in
absorbance at 455 nm and  497 nm, respectively, that
accompany complex formation: Kd values were estimated
based on [29].
Protein fluorescence measurements were obtained
between 300 and 400 nm, with excitation at 280 nm; flavin
emission spectra were recorded from 475 to 600 nm, with
excitation at 450 nm. Fluorescence measurements were
performed using a Jasco FP-750 instrument (Cremello,
Italy) at 15 °C and 0.1 mgỈmL)1 protein concentration, and
corrected for buffer contributions. Temperature-ramp
experiments were performed as reported previously [30]
using a software-driven, Peltier-equipped fluorometer in
which a temperature gradient could be reproduced
(0.5 °CỈmin)1). Circular dichroism (CD) spectra were
recorded at 15 °C using a Jasco J-810 spectropolarimeter
and analyzed by means of Jasco software. The cell path

Site-saturation mutagenesis at positions 50, 201 and 225
was performed on DAAO cDNA subcloned into pT7-HisDAAO as template [18] using a QuikChange site-directed

mutagenesis kit (Stratagene, La Jolla, CA, USA) and a set
of degenerate synthetic oligonucleotides. The PCR products
were used to transform JM109 Escherichia coli cells, and
then the recombinant plasmids were transferred into
BL21(DE3)pLysS E. coli cells; these clones were used for
the screening procedure.
DAAO variants with an altered enzymatic activity at low
O2 concentration (2.5% = 30 lm) were identified using the
screening procedure described previously [10] and the
AtmosBag incubation system (Sigma Aldrich, St Louis,
MO, USA) on  250 clones for each position. Introduction
of the mutations was confirmed by automated DNA
sequencing. Recombinant clones encoding DAAO variants
selected using the screening procedure were grown and
purified as described previously [10]. The purified DAAO
preparations were then equilibrated with 50 mm potassium
phosphate buffer, pH 7.5, 10% glycerol, 2 mm EDTA and
5 mm 2-mercaptoethanol. As with wild-type DAAO, the
purified variants were > 90% pure according to SDS ⁄ PAGE
analysis (data not shown).
The expression of W50F and W50R variants was similar
to that of the wild-type DAAO ( 1.5 mg enzyme per g cell
paste), but with a lower volumetric yield ( 6 mg protein
per liter of fermentation broth versus 11 mg for the wild-

490

FEBS Journal 278 (2011) 482–492 ª 2010 The Authors Journal compilation ª 2010 FEBS

E. Rosini et al.

length was 1 cm for measurements above 250 nm and
0.1 cm for measurements in the 190–250 nm region [30].

Activity assays and stopped-flow measurements
DAAO activity was assayed using an oxygen electrode at
pH 8.5 and 25 °C with 28 mm d-alanine and air saturation
([O2] = 0.253 mm) [10]. One DAAO unit is defined as the
amount of enzyme that converts 1 lmol of d-alanine per
minute at 25 °C.
Steady-state and pre-steady-state stopped-flow experiments were performed in 50 mm sodium pyrophosphate,
pH 8.5, containing 1% v ⁄ v glycerol and 0.25 mm 2-mercaptoethanol, at 15 °C in a BioLogic SFM-300 instrument
equipped with a J&M diode array detector (BioLogic,
Grenoble, France) as described previously [7,10]. Steadystate kinetic parameters were determined by the enzymemonitored turnover technique, mixing equal volumes of
 8–10 lm air-saturated enzyme with an air-saturated
solution of d-alanine. The time courses at 455 nm reflect
the conversion of oxidized into reduced enzyme species,
and indicate the rate of catalysis as a continuous function
of oxygen concentration (the limiting substrate) [20].
For reductive half-reaction experiments, the oxidized
enzyme form was reacted with increasing d-alanine concentrations in the absence of dioxygen (the final solutions contained 100 mm glucose, 0.1 lm glucose oxidase and 30 nm
catalase). For study of the oxidative half-reaction, reduced
enzyme forms were reacted with solutions of appropriate O2
concentrations. Two reduced enzyme forms were used: (a)
free reduced DAAO (E-Flred, generated by reacting oxidized
DAAO with a small excess of d-alanine), and (b) the reduced
DAAO–imino acid complex (E-Flred–IA generated as above
but in the presence of 400 mm NH4Cl and 20 mm pyruvate
to generate imino pyruvate). Reaction rates for both the

reductive and oxidative half-reaction (Scheme 1) were estimated from traces extracted at 455 and 530 nm by fitting
using the application Biokine32 (BioLogic) and one or
two exponential terms (e.g. for a bi-exponential fit:
y = A e)k1t + B e)k2t + C, where A and B are amplitudes
and C is an initial absorbance value). The determined rate
constants were used to simulate the half-reactions and to
estimate the kinetic steps that cannot be detected experimentally using Specfit software (Spectrum Software Associates,
Chapel Hill, NC).

Acknowledgements
This work was supported by grants from Fondo di
Ateneo per la Ricerca (University of Insubria, Varese,
Italy). We are grateful for the support of the Consorzio Interuniversitario per le Biotecnologie, and the
Centro di Ricerca in Biotecnologie per la Salute
Umana (University of Insubria).

Modulation of O2 reactivity in D-amino acid oxidase

References
1 Massey V & Hemmerich P (1980) Active-site probes of
flavoproteins. Biochem Soc Trans 8, 246–257.
2 Massey V (1995) Introduction: flavoprotein structure
and mechanism. FASEB J 9, 473–475.
3 Bornemann S (2002) Flavoenzymes that catalyse reactions with no net redox change. Nat Prod Rep 19, 761–
772.
4 Pollegioni L, Piubelli L, Sacchi S, Pilone MS & Molla
G (2007) Physiological functions of d-amino acid
oxidases: from yeast to humans. Cell Mol Life Sci 64,
1373–1394.
5 Umhau S, Pollegioni L, Molla G, Diederichs K, Welte

W, Pilone MS & Ghisla S (2000) The X-ray structure of
d-amino acid oxidase at very high resolution identifies
the chemical mechanism of flavin-dependent substrate
dehydrogenation. Proc Natl Acad Sci USA 97, 12463–
12468.
6 Porter DJ, Voet JG & Bright HJ (1977) Mechanistic
features of the d-amino acid oxidase reaction studied by
double stopped flow spectrophotometry. J Biol Chem
252, 4464–4473.
7 Pollegioni L, Langkau B, Tischer W, Ghisla S & Pilone
MS (1993) Kinetic mechanism of d-amino acid oxidases
from Rhodotorula gracilis and Trigonopsis variabilis.
J Biol Chem 268, 13850–13857.
8 Klinman JP (2007) How do enzymes activate oxygen
without inactivating themselves? Acc Chem Res 40,
325–333.
9 Mattevi A (2006) To be or not to be an oxidase: challenging the oxygen reactivity of flavoenzymes. Trends
Biochem Sci 31, 276–283.
10 Rosini E, Pollegioni L, Ghisla S, Orru R & Molla G
(2009) Optimization of d-amino acid oxidase for low
substrate concentrations – towards a cancer enzyme
therapy. FEBS J 276, 4921–4932.
11 Furse KE, Pratt DA, Schneider C, Brash AR, Porter
NA & Lybrand TP (2006) Molecular dynamics simulations of arachidonic acid-derived pentadienyl radical
intermediate complexes with COX-1 and COX-2:
insights into oxygenation regio- and stereoselectivity.
Biochemistry 45, 3206–3218.
12 Hiromoto T, Fujiwara S, Hosokawa K & Yamaguchi
H (2006) Crystal structure of 3-hydroxybenzoate
hydroxylase from Comamonas testosteroni has a large

tunnel for substrate and oxygen access to the active site.
J Mol Biol 364, 878–896.
13 Johnson BJ, Cohen J, Welford RW, Pearson AR,
Schulten K, Klinman JP & Wilmot CM (2007) Exploring molecular oxygen pathways in Hansenula polymorpha copper-containing amine oxidase. J Biol Chem 282,
17767–17776.
14 Baron R, Riley C, Chenprakhon P, Thotsaporn K,
Winter RT, Alfieri A, Forneris F, van Berkel WJH,

FEBS Journal 278 (2011) 482–492 ª 2010 The Authors Journal compilation ª 2010 FEBS

491


Modulation of O2 reactivity in D-amino acid oxidase

15

16

17

18

19

20

21

22

23

24

25

26

492

E. Rosini et al.

Chaiyen P, Fraaije MW et al. (2009) Multiple pathways
guide oxygen diffusion into flavoenzymes active sites.
Proc Natl Acad Sci USA 106, 10603–10608.
Vrielink A & Ghisla S (2009) Cholesterol oxidase:
biochemistry and structural features. FEBS J 276,
6826–6843.
Saam J, Rosini E, Molla G, Schulten K, Pollegioni L &
Ghisla S (2010) O2-reactivity of flavoproteins: dynamic
access of dioxygen to the active site and role of a H+
relay system in d-amino acid oxidase. J Biol Chem 285,
24439–24446.
Sacchi S, Lorenzi S, Molla G, Pilone MS, Rossetti C &
Pollegioni L (2002) Engineering the substrate specificity
of d-amino-acid oxidase. J Biol Chem 277, 27510–
27516.
Sacchi S, Rosini E, Molla G, Pilone MS, Ghisla S &
Pollegioni L (2004) Modulating d-amino acid oxidase

substrate specificity: production of an enzyme for
analytical determination of all d-amino acids by
directed evolution. Protein Eng Des Sel 17, 517–
525.
Pollegioni L, Porrini D, Molla G & Pilone MS (2000)
Redox potentials and their pH dependence of d-amino
acid oxidase of Rhodotorula gracilis and Trigonopsis variabilis. Eur J Biochem 267, 6624–6632.
Gibson QH, Swoboda BE & Massey V (1964) Kinetics
and mechanism of action of glucose oxidase. J Biol
Chem 239, 3927–3934.
Ricard J & Cornish-Bowden A (1987) Co-operative and
allosteric enzymes: 20 years on. Eur J Biochem 166,
255–272.
Strickland S, Palmer G & Massey V (1975) Determination of dissociation constants and specific rate constants
of enzyme–substrate (or protein–ligand) interactions
from rapid reaction kinetic data. J Biol Chem 250,
4048–4052.
Chen L, Lyubimov AY, Brammer L, Vrielink A &
Sampson NS (2008) The binding and release of oxygen
and hydrogen peroxide are directed by a hydrophobic
tunnel in cholesterol oxidase. Biochemistry 47, 5368–
5377.
Forneris F, Heuts DP, Delvecchio M, Rovida S,
Fraaije MW & Mattevi A (2008) Structural analysis of
the catalytic mechanism and stereoselectivity in
Streptomyces coelicolor alditol oxidase. Biochemistry
47, 978–985.
Piubelli L, Pedotti M, Molla G, Feindler-Boeckh S,
Ghisla S, Pilone MS & Pollegioni L (2008) On the oxygen reactivity of flavoprotein oxidases: an oxygen access
tunnel and gate in Brevibacterium sterolicum cholesterol

oxidase. J Biol Chem 283, 24738–24747.
Massey V & Hemmerich P (1977) A photochemical procedure for reduction of oxidation–reduction proteins
employing deazariboflavin as catalyst. J Biol Chem 252,
5612–5614.

27 Molla G, Porrini D, Job V, Motteran V, Vegezzi C,
Campaner S, Pilone MS & Pollegioni L (2000) Role of
arginine 285 at the active site of Rhodotorula gracilis
d-amino acid oxidase. A site-directed mutagenesis
study. J Biol Chem 275, 24715–24721.
28 Massey V (1991) A simple method for the determination
of redox potentials. In Flavins and Flavoproteins (Curti
B, Ronchi S & Zanetti G, eds), pp. 59–66. Walter de
Gruyter, Berlin.
29 Harris CM, Ghisla S & Pollegioni L (2001) pH and
kinetic effects in d-amino acid oxidase catalysis.
Evidence for a concerted mechanism in substrate
dehydrogenation via hydride transfer. Eur J Biochem
268, 5504–5520.
30 Caldinelli L, Iametti S, Barbiroli A, Bonomi F, Ferranti
P, Pilone MS & Pollegioni L (2004) Unfolding of the
peroxisomal flavoprotein d-amino acid oxidase. J Biol
Chem 279, 28426–28434.
31 Humphrey W, Dalke A & Schulten K (1996) VMD:
visual molecular dynamics. J Mol Graph 14, 33–38.
32 Sanner M, Olson AJ & Spehner JC (1996) Reduced
surface: an efficient way to compute molecular surfaces.
Biopolymers 38, 305–320.

Supporting information

The following supplementary material is available:
Fig. S1. Absorbance spectrum of wild-type and variants of DAAO in the oxidized state.
Fig. S2. Comparison of the spectral properties related
to protein conformation for W50 variants of DAAO.
Fig. S3. Steady-state measurements of oxygen consumption by DAAO variants.
Table S1. Comparison of melting temperatures as
determined by various approaches for the wild-type
and W50 variants of DAAO.
Table S2. Comparison of the ligand-binding properties
and apparent steady-state kinetic parameters on
d-alanine as substrate determined for wild–type and
variants of DAAO.
Table S3. Comparison of the substrate specificity of
wild-type and I225F DAAOs.
Appendix S1. Biochemical properties of DAAO variants.
This supplementary material can be found in the
online version of this article.
Please note: As a service to our authors and readers,
this journal provides supporting information supplied
by the authors. Such materials are peer-reviewed and
may be re-organized for online delivery, but are not
copy-edited or typeset. Technical support issues arising
from supporting information (other than missing files)
should be addressed to the authors.

FEBS Journal 278 (2011) 482–492 ª 2010 The Authors Journal compilation ª 2010 FEBS