New membrane-associated and soluble peptide methionine sulfoxidereductases in Escherichia coli

Daniel Spector,a Frantzy Etienne,a Nathan Brot,b and Herbert Weissbacha,*

a Center for Molecular Biology and Biotechnology, Florida Atlantic University, 777 Glades Road, Boca Raton, FL 33431, USAb Department of Microbiology and Immunology, Hospital for Special Surgery, Weill Medical College of Cornell University, New York, NY, USA

Received 15 January 2003

Abstract

It is known that reactive oxygen species can oxidize methionine residues in proteins in a non-stereospecific manner, and cells have

mechanisms to reverse this damage. MsrA and MsrB are members of the methionine sulfoxide family of enzymes that specifically

reduce the S and R forms, respectively, of methionine sulfoxide in proteins. However, in Escherichia coli the level of MsrB activity is

very low which suggested that there may be other enzymes capable of reducing the R epimer of methionine sulfoxide in proteins.

Employing a msrA/B double mutant, a new peptide methionine sulfoxide reductase activity has been found associated with

membrane vesicles from E. coli. Both the R and S forms of N -acetylmethionine sulfoxide, DD-ala-met(o)-enkephalin and methionine

sulfoxide, are reduced by this membrane associated activity. The reaction requires NADPH and may explain, in part, how the R

form of methionine sulfoxide in proteins is reduced in E. coli. In addition, a new soluble Msr activity was also detected in the soluble

extracts of the double mutant that specifically reduces the S epimer of met(o) in proteins.

� 2003 Elsevier Science (USA). All rights reserved.

Keywords: Methionine; Methionine sulfoxide; Methionine sulfoxide reductase; Oxidation; Enzyme

Proteins and other macromolecules are highly sus-

ceptible to oxidation by reactive oxygen species (ROS)

produced by cells as a result of oxidative metabolism

[1]. The methionine (met) residues of proteins are par-

ticularly vulnerable to oxidation, forming methionine

sulfoxide (met(o)). While much of the cellular antioxi-dant machinery, including superoxide dismutases,

catalases, and peroxidases, is directed at destroying

these radicals, there are also important activities, which

repair oxidative damage. The met(o) residues resulting

from met oxidation can be reduced back to met by

members of the methionine sulfoxide reductase (Msr)

family of enzymes, using electrons derived from thior-

edoxin [2]. The reduction of met(o) in proteins can re-store the function of proteins in which essential met

residues have been oxidized [3]. In addition to this re-

pair function, the oxidation/reduction of met residues in

proteins might also be an important mechanism to

scavenge ROS [4,5].

There is considerable evidence to suggest that the Msr

system is an important cellular mechanism to protect

against oxidative damage. Escherichia coli and yeast

lacking MsrA, one of the Msr genes, are more sensitiveto oxidative stress than wild type strains [6–8], and an-

imal cells that overexpress MsrA, in vitro, are more re-

sistant to oxidative stress [7]. Additionally, when bovine

MsrA is overexpressed in fruit flies there is close to a

doubling of their life span [9]. Conversely, the deletion

of MsrA in mice results in a 30% decrease in life span

and these animals exhibit neurological defects [10]. Fi-

nally, there is recent evidence that in humans andprobably other animals, one form of MsrA is associated

with mitochondria [11], consistent with the notion that it

is required to counteract endogenously produced ROS

generated by mitochondrial respiration.

The asymmetric sulfur atom produced by chemical

oxidation of met results in two epimers of met(o),

referred to as met-R-(o) and met-S-(o). MsrA is a

Biochemical and Biophysical Research Communications 302 (2003) 284–289

www.elsevier.com/locate/ybbrc

BBRC

* Corresponding author. Fax: 1-561-297-2594.

E-mail address: hweissba@fau.edu (H. Weissbach).

0006-291X/03/$ - see front matter � 2003 Elsevier Science (USA). All rights reserved.

doi:10.1016/S0006-291X(03)00163-3

ubiquitous enzyme that specifically reduces met-S-(o),whether as free met(o) or met(o) in proteins [12]. Re-

cently, another enzyme, MsrB, that specifically reduces

met-R-(o) in proteins has been identified in E. coli and

other organisms, based on its homology to the PilB

protein of Neisseria gonorrhoeae [13–16]. Although the

msrB gene has been identified in E. coli and the active

recombinant E. coli protein isolated [13], we have not

been able to detect any significant MsrB activity inE. coli extracts (unpublished data). In addition E. coli

mutants lacking MsrB show no apparent phenotype

(unpublished data) as compared to E. colimsrA mutants

which are sensitive to oxidative stress [6,8]. Recently

Moskovitz et al. [17] reported MsrB activity in E. coli

extracts when large amounts of the extract were used.

Despite this report, all of our results have questioned the

in vivo significance of MsrB in E. coli and pose thequestion as to how E. coli cells reduce the met-R-(o) in

proteins. This is important in order to prevent the ac-

cumulation of met-R-(o) in proteins which could be

deleterious to the cells. Recently we reported the pres-

ence of an enzyme (fRMsr) in E. coli that reduces free

met-R-(o), but not met-R-(o) in proteins [18]. In the

present study we have identified Msr activity(s) in the

membrane fraction of E. coli capable of reducing boththe R and S forms of met(o) in peptide linkage, as well

as free met(o). In addition, a new soluble Msr activity

has been identified that reduces the S epimer of met(o)

in peptide linkage, but not free met(o).

Materials and methods

Wild type (wt) E.coli strain MC1061 was used for the construction

of the msrB mutant by transposon insertion as described previously for

the msrA mutant [6]. The msrA mutant was used to prepare the msrA/

B double mutant. The msrA gene was disrupted using a kanomycin

resistance marker, while the msrB gene was disrupted using a chl-

oramphenicol marker. Western blot analysis confirmed that the mu-

tants were lacking the individual gene products. The cells were grown

in an LB broth to an A600 of 0.8–0.9, pelleted, and suspended in 1.5

volumes of a buffer containing 10mM Tris–Cl, pH 7.4, 10mMMgCl2,

and 10mM NH4Cl (buffer A) per gm of cells.

The cell extracts were prepared by sonication of the cells (5� 15 s)

followed by a low speed (8000g) centrifugation step which pelleted the

cell debris and membranes (membrane fraction). The membrane

fraction was then washed with buffer A to remove any remaining su-

pernatant. The 8000g supernatant was centrifuged at 100,000g for onehour to obtain an S-100 fraction. Membrane vesicles were prepared

from permeable cells as described previously by Kaback [19].

The routine assay to detect peptide methionine sulfoxide reductase

activity used radiolabeled N -acetylmet-R,S-(o) as substrate [20]. Thissubstrate has been shown to mimic protein bound methionine, since

the amino group is blocked [20]. A typical reaction (30 ll) contained50mM Tris–Cl, pH 7.4, 6lg E. coli thioredoxin, 0.5lg E. coli thior-edoxin reductase, 40 nmol NADPH, 600nmol glucose-6-phosphate,

150 ng glucose-6-phosphate dehydrogenase, 5 nmol 3H-N -acetylmet-R/S-(o), and enzyme as indicated. In some experiments, where indicated,

15mM DTT was used in place of the thioredoxin/NADPH/glucose-

6-phosphate system. Incubations were at 37 �C for up to one hour. The

radioactive N -acetylmet formed was extracted into ethyl acetate and

the radioactivity was determined as described previously [20]. This

assay is highly sensitive and can detect as little as 10 picomol of

product. Previously, using MsrA or MsrB, it was shown that either

dithiothreitol (DTT) or the thioredoxin system could be used as an

electron source for MsrA and MsrB activity.

A modified form of the above assay was also used to determine the

stereospecificity of the reaction. In this assay, saturating amounts of

MsrA or MsrB were first incubated with the N -acetylmet-R,S-(o) for60min to completely reduce either the S or R epimer, respectively.

After this first incubation was complete, the enzyme fraction con-

taining the Msr activity (either membrane fraction or soluble extract)

was then added, along with additional reducing system components.

Any enzymatic activity in the second incubation was due to reduction

of the epimer that remained after the first incubation with MsrA or

MsrB. As an example, if the first incubation contained MsrA, which

reduced all of the N -acetyl-met-S-(o), any activity of the enzyme

fractions in the second incubation would be due to reduction of

N -acetylmet-R-(o). The opposite would be true if the first incubation

contained MsrB, which is specific for the R form.

Met-R-(o) and met-S-(o) were prepared from the racemic mixture

by the method of Lavine [21]. The met-R-(o) had about a 5% con-

tamination with met-S-(o), whereas the met-S-(o) was essentially free

of met-R-(o). A typical reaction mixture for free met-(o) reduction to

met contained, in a total volume of 200ll, 100mM KH2PO4, pH 6.9,

4 nmol glucose-6-phosphate, 1lg glucose-6-phosphate dehydrogenase,15 lg E. coli thioredoxin, 1 lg E.coli thioredoxin reductase, 100 nmol

NADPH, 1 lmol met(o) (either R or S), and the enzyme fraction as

indicated. Incubations were at 37 �C for up to one hour. The met

synthesized was assayed colorimetrically using nitroprusside reagent as

described previously [22].

DD-Ala2, DD-Met5 enkephalin (met-enkephalin) was purchased from

Sigma, oxidized with H2O2 [20], dried in a speed vac, and resuspended

in water. This oxidized peptide, (met(o)-enkephalin) containing met-

R,S-(o), was also tested as substrate for the membrane Msr using in-

cubation conditions similar to those described above for the reduction

of N -acetylmet-(o). The incubations were carried out in a total volumeof 50 ll and contained 53 nmol met(o)-enkephalin and 140lg of the

membrane fraction protein. After incubation, samples were heated to

100 �C for 1min, centrifuged, and 40 ll was injected onto a C-18

HPLC column. The column was eluted with a sodium acetate gradient

in methanol. The two met(o)-enkephalin epimers eluted at 70min for

the R form and 70.5min for the S form, while the product, met-en-

kephalin, was well separated at 83min. The met-enkephalin peptide

used in these studies contained two DD-amino acids, Ala and Met, which

prevented destruction of the peptide by proteolytic activity present in

the crude E. coli fractions. Whether a DD or LL-met isomer is used should

not affect the Msr assay, since the well-characterized enzymes, MsrA

and MsrB, are active with both the DD and LL-met(o) enkephalin sub-

strates (data not shown).

Results and discussion

The msrA/B double mutant was used initially to

search for Msr activity that reduced met(o) in peptide

linkage since the background level of Msr activity was

reduced in these cells. A low speed pellet and S-100

fraction were prepared from broken cells as described in

Materials and methods. Significant Msr activity, usingN -acetylmet(o) as a model substrate, was found in the

low speed pellet as well as in the S-100 fraction, pre-

pared from extracts of these cells, using NADPH as the

reducing agent.

D. Spector et al. / Biochemical and Biophysical Research Communications 302 (2003) 284–289 285

Membrane associated Msr

To determine if the Msr activity in the low speed

pellet was associated with the cell membrane, membrane

vesicles were prepared using the protocol developed by

Kaback [19]. The Msr activity originally in the low

speed pellet was essentially all recovered in the mem-

brane vesicles, indicating that it is either an integral

membrane protein or a tightly bound membrane asso-ciated protein. Fig. 1 shows the effect of membrane

protein concentration and time on the reduction of N -acetylmet (o). The reaction is not linear with protein

amounts below 5 lg in the incubations (Fig. 1A), sug-

gesting that this is a multi-component system. Using

8 lg of membrane protein the reaction rate also shows a

lag of about 20min after which the rate is linear, for

about 10min as shown in Fig. 1B. Table 1 shows thatthe reaction requires NADPH and can be stimulated

about twofold by the addition of the thioredoxin re-

ducing system (thioredoxin and thioredoxin reductase).

At the level of the membrane protein (9 lg) used in

Table 1, DTT could not substitute for NADPH. At

higher concentrations of membrane protein (20–30 lg)DTT did give significant activity in the absence of

NADPH (data not shown). It should be noted that thelack of Msr activity at low membrane protein concen-

trations (Fig. 1A) was not due to limiting amounts of the

thioredoxin system since the addition of the thioredoxin

reducing system had no effect. The lack of a complete

dependency on thioredoxin suggests that the membranesare contaminated with thioredoxin and /or contain an-

other reducing system for this Msr activity.

The substrate and stereospecificity of the membrane

Msr activity is shown in Table 2. Lines 1 and 2 of Table

2 are controls showing that MsrA and MsrB reduce

specifically the S and R epimers of N -acetylmet(o), re-spectively. The membrane preparation, however, re-

duces both of the epimers of N -acetylmet-(o) (Line 3,Table 2). Table 3 shows that the membrane vesicles can

also reduce both free met-S-(o) and free met-R-(o). At

this point we cannot be certain whether one or more

reductases are involved in the reduction of the R and S

epimers of both free and N -blocked met-(o).

Although N -acetylmet-(o) mimics a peptide bound

methionine sulfoxide, a penta-peptide, met(o)-enkeph-

alin was also used as substrate. The R and S stereoi-somers of this peptide could be resolved on HPLC as

shown in Fig. 2A. After incubation with membrane

vesicles both the R and S species of the oxidized pep-

tide were reduced (Fig. 2B) and the product met-

enkephalin formed. The membrane vesicles always

appeared to be more active with the R epimer than the

S, as evidenced by the greater decrease in the R peak in

Fig. 2B. These results confirm that the membrane ves-icle Msr can utilize peptide bound met(o) as substrate.

The membrane associated Msr activity described here

was also present in wild type extracts of E. coli (data

not shown).

Table 2

Stereospecificity of the membrane vesicle Msr activity using the R or S

epimers of N -acetylmet(o) as substrate

Enzyme Substrate

N -acetylmet-R-(o)(pmol)

N -acetylmet-S-(o)(pmol)

MsrA <10 439

MsrB 220 <10

Membrane vesicles 401 473

The R and S epimers were prepared enzymatically from the racemic

mixture as described in Materials and methods. MsrA(2lg) and

MsrB(4lg) were used in control incubations with the R and S sub-

strates to check the purity of the R and S epimers. Twenty micrograms

of membrane vesicle proteins was used in the experiments.

Table 3

Stereospecificity of the membrane vesicle Msr activity with free

met-R-(o) and met-S-(o) as substrates

Membrane protein (lg) Substrate

Met-R-(o) (nmol) Met-S-(o)

(nmol)

180 310 188

270 530 400

The met(o) substrates were prepared as described previously [21].

Details of the incubations are described in Materials and methods and

the nitroprusside assay has been described previously [22].

Fig. 1. Membrane vesicle Msr activity with N -acetylmet-R,S-(o) assubstrate. The incubations contained NADPH as the reducing agent

and other details are described in the Materials and methods. (A)

Effect of membrane protein concentration using a 60min incubation.

(B) Time course using 8 lg of protein.

Table 1

Reducing system requirement for the reduction of N -acetylmet-R,S-(o)by membrane vesicles

Reducing system N -acetylmet (pmol)

NADPH 59

None <10

NADPH+ thioredoxin system 108

DTT <10

The reaction mixture contained 9 lg of membrane protein and

the incubations were for 60min. See Materials and methods for

details.

286 D. Spector et al. / Biochemical and Biophysical Research Communications 302 (2003) 284–289

Soluble Msr activity

In the course of these studies using extracts from an

MsrA/B double mutant we recently found that the sol-

uble fraction also contained Msr activity that reduced

free met-R-(o) [18]. In addition the S-100 supernatant

fraction of the MsrA/B double mutant also contains aweak Msr activity that reduces N -acetylmet(o). As

shown in Table 4, Lines 1 and 2, the N -acetylmet(o)reducing activity in the S-100 fraction has an absolute

requirement for NADPH. The Msr activity in this crude

extract was not stimulated by the thioredoxin reducing

system and DTT could only partially replace NADPH

(data not shown). As also shown in Table 4, Lines 3 and

4, the soluble activity displays a stereospecificity for theS epimer, similar to MsrA. This MsrA like activity,

which we refer to as MsrA1, was very likely not detected

in an earlier study with an MsrA mutant [6], because of

the low activity in the extracts compared to MsrA, es-

pecially in the presence of DTT, which was used in the

previous studies. MsrA1 not only differs from MsrA in

that DTT is not a good reductant, but also in its sub-

strate specificity, since this new activity does not reducefree met(o) (not shown). This E. coli peptide met-S-(o)

activity may be similar to a Msr activity recently de-

scribed in Staphylococcus aureus [17,23].

In the present study an E. coliMsrA/B double mutant

has been used to identify new Msr activities in this or-

ganism. The impetus for this study was to understand

how the R form of met (o) in proteins was reduced, since

our previous studies suggested that MsrB activity in E.

coli was very low. Both membrane associated and sol-

uble Msr activities were detected. The membrane asso-

ciated activity(s) has a broad specificity since it reduces

both the R and S forms of met(o), either free or in

peptide linkage. It is present in isolated membrane

vesicles and although not purified there is preliminary

evidence that it can be partially solubilized with 1,2-

diheptanoyl-sn-glycero-3-phosphocholine. At this pointit is not clear whether the Msr activity observed with the

free and peptide bound epimers of met(o) is due to one

or several enzymes. Competition experiments using free

met-R,S-(o) suggest that the reductase active with the R

epimer of the free and peptide bound met(o) might be

the same enzyme. The eventual purification of the Msr

activities in the membrane vesicles will answer this and

other questions. What is clear is that the Msr activity inthe membranes requires NADPH and is stimulated by

the addition of reduced thioredoxin, which is the bio-

logical reducing system for the other Msr activities [2]

that have been described. In addition, DTT, which ef-

ficiently replaces reduced thioredoxin with MsrA and E.

coli MsrB [13,20], is only weakly active with the mem-

brane Msr. It is possible that the membrane vesicles

contain another reducing system that may be active withthe membrane Msr activity, which can be partially re-

placed by the thioredoxin system. Alternatively, the

membrane fraction may contain limiting amounts of

thioredoxin accounting for the partial stimulation. Once

the membrane Msr activity(s) has been purified, the

Fig. 2. Met(o)-enkephalin is a substrate for the membrane vesicle Msr activity. Details of the procedure to separate the R and S forms of the met(o)-

enkephalin pentapeptide are described in Materials and methods. (A) Separation of the R and S forms of the met(o)-enkephalin in incubations minus

enzyme. (B) Reduction of both the R and S forms of met(o)-enkephalin by the membrane vesicle preparation. Formation of the product is evident by

the peak of the met-enkephalin at 83min. See Materials and methods for details.

Table 4

Stereospecificity and NADPH requirement for the Msr activity in the

S-100 fraction

System Substrate N -acetylmet (pmol)

Complete N -acetylmet-R,S-(o) 283

Minus NADPH N -acetylmet-R,S-(o) <10

Complete N -acetylmet-S-(o) 207

Complete N -acetylmet-R-(o) <10

One hundred micrograms of S-100 protein was used and other

details of the incubations are described in Materials and methods.

D. Spector et al. / Biochemical and Biophysical Research Communications 302 (2003) 284–289 287

physiological reducing system can be identified, and therole of thioredoxin, if any, can be elucidated.

The new soluble activity described here, MsrA1, is

similar to MsrA, in that it reduces met-S-(o) in proteins.

It was very likely not detected in earlier studies with the

msrA mutant [6], because the level of MsrA1 activity in

extracts, using reduced thioredoxin, is <20% of MsrA,

and this enzyme has much less activity compared to

MsrA when DTT is used as the reductant. The earlierexperiments with the msrA mutant used DTT to measure

reductase activity in the extracts [6]. One major difference

between MsrA1 and MsrA is that free met-(o) is not a

substrate forMsrA1. Recently, anMsrA like activity was

identified in S. aureus, which may be similar to the ac-

tivity described here [17,23]. The reducing system for the

soluble Msr described here has not yet been determined.

NADPH is required but since the enzyme fraction iscrude, the lack of a thioredoxin effect is not meaningful.

It is surprising that in E. coli there appear to be at

least 6 members of the Msr family of proteins. These

include MsrA, MsrB, fSMsr, the membrane associated

Msr reductase, and MsrA1 described here and a recently

identified free met-R-(o) reductase [18]. If the membrane

associated Msr activity is composed of more than 1

enzyme, as it seems likely, then the number of Msr ac-tivities in E. coli is even higher. Previously we presented

preliminary evidence, suggesting the presence of a free

met-(o) epimerase in E. coli [2], but attempts to purify

this activity to verify its existence have been unsuccess-

ful. Although much less is known about the family of

Msrs in higher organisms there are at least 2 MsrB like

activities in human cells [16,24], a mitochondrial form of

MsrA in humans [11] and a soluble MsrA have beendetected in a variety of animal tissues [25]. Whether the

mitochondrial and soluble forms of MsrA are due to

alternative splicing of the MsrA transcript or separate

genes is not known. A free met-S-(o) reductase has also

been demonstrated in yeast cells [7].

From the previous genetic studies and the number of

members in the Msr family one must conclude that the

reduction of met(o), both free and in peptide linkage, isan important mechanism that cells use to protect against

oxidative damage. Knowledge on the properties of each

of these different Msr proteins could be of great value in

understanding the role of this system in normal and

pathological conditions.

Acknowledgment

The authors thank Dr. H. Ronald Kaback for many helpful dis-

cussions during the course of this study.

References

[1] J.L. Farber, Mechanisms of cell injury by activated oxygen

species, Environ. Health Perspect. 102 (1994) 17–24.

[2] H. Weissbach, F. Etienne, T. Hoshi, S.H. Heinemann, W.T.

Lowther, B. Matthews, G. St. John, C. Nathan, N. Brot, Peptide

methionine sulfoxide reductase: structure mechanism of action

and biological function, Arch. Biochem. Biophys. 397 (2002) 172–

178.

[3] N. Brot, H. Weissbach, Peptide methionine sulfoxide reductase:

biochemistry and physiological role, Biopolymers 55 (2000) 288–

296.

[4] R.L. Levine, J. Moskovitz, E.R. Stadtman, Oxidation of methi-

onine in proteins: roles in antioxidant defense and cellular

regulation, IUBMB Life 50 (2000) 301–307.

[5] R.L. Levine, B.S. Berlett, J. Moskovitz, L. Mososni, E.R.

Stadtman, Methionine residues may protect proteins from critical

oxidative damage, Mech. Ageing Dev. 107 (1999) 323–332.

[6] J. Moskovitz, M.A. Rahman, J. Straman, S.O. Yancey, S.R.

Kushner, N. Brot, H. Weissbach, Escherichia coli peptide methi-

onine sulfoxide reductase gene: its regulation of expression and

role in protecting against oxidative damage, J. Bacteriol. 177

(1995) 502–507.

[7] J. Moskovitz, E. Flesher, B.S. Berlett, J. Azare, J.M. Poston, E.R.

Stadtman, Overexpression of peptide methionine sulfoxide reduc-

tase in Saccharomyces cerevisiae and human T-cells provides them

with high resistance to oxidative stress, Proc. Natl. Acad. Sci.

USA 95 (1998) 14071–14075.

[8] G. St. John, N. Brot, J. Ruan, H. Erdjument-Bromage, P. Tempst,

H. Weissbach, C. Nathan, Peptide methionine sulfoxide reductase

from Escherichia coli and Mycobacterium tuberculosis protects

bacteria against oxidative damage from reactive nitrogen inter-

mediates, Proc. Natl. Acad. Sci. USA 98 (2002) 9901–9906.

[9] H. Ruan, X.D. Tang, M.L. Chen, M.A. Joiner, G. Sun, N. Brot,

H. Weissbach, S.H. Heinemann, L. Iverson, C.-F. Wu, T. Hoshi,

High-quality life extension by the enzyme peptide methionine

sulfoxide reductase, Proc. Natl. Acad. Sci. USA 99 (2002) 2748–

2753.

[10] J. Moskovitz, S. Bar-Noy, W.M. Williams, J. Requena, B.S.

Berlett, E.R. Stadtman, Methionine sulfoxide reductase (MsrA) is

a regulator of antioxidant defense and lifespan in mammals, Proc.

Natl. Acad. Sci. USA 98 (2001) 12920–12925.

[11] A. Hansel, L. Kuschel, S. Hehl, C. Lemke, H.J. Agricola, T.

Hoshi, S.H. Heinemann, Mitochondrial targeting of the human

peptide methionine sulfoxide reductase (MsrA) an enzyme

involved in the repair of oxidized proteins, FASEB 16 (2002)

911–913.

[12] J. Moskovitz, H. Weissbach, N. Brot, Cloning and expression of a

mammalian gene involved in the reduction of methionine sulfox-

ide residues in proteins, Proc. Natl. Acad. Sci. USA 93 (1996)

2095–2099.

[13] R. Grimaud, B. Ezraty, J.K. Mitchell, D. Lafitte, C. Briand, P.L.

Derrick, F. Barras, Repair of oxidized proteins. Identification of a

new methionine sulfoxide reductase, J. Biol. Chem. 276 (2001)

48915–48920.

[14] W.T. Lowther, H. Weissbach, F. Etienne, N. Brot, B.W.

Matthews, The mirrored methionine sulfoxide reductases of

Neisseria gonorrhoeae pilB, Nat. Struct. Biol. 9 (2002) 348–

352.

[15] A. Orly, S. Boshi-Muller, M. Marraud, S. Sanglier-Cianferani, A.

Van Dorsselear, G. Branlant, Characterization of the methionine

sulfoxide reductase activities of PilB, a probable virulence factor

from Neisseria meningitidis, J. Biol. Chem. 277 (2002) 12016–

12022.

[16] W. Huang, J. Escribano, M. Sarfarazi, M. Coca Prados, Identi-

fication expression and chromosome localization of a human gene

encoding a novel protein with similarity to the pilB family of

transcriptional factors (pilin) and to the bacterial peptide methi-

onine sulfoxide reductases, Gene 233 (1999) 233–240.

[17] J. Moskovitz, V.K. Singh, J. Requena, B.J. Wilkinson, R.K.

Jayaswal, E.R. Stadtman, Purification and characterization of

288 D. Spector et al. / Biochemical and Biophysical Research Communications 302 (2003) 284–289

methionine sulfoxide reductases from mouse and Staphylococcus

aureus and their substrate stereospecificity, Biochem. Biophys.

Res. Commun. 290 (2002) 62–65.

[18] F. Etienne, D. Spector, N. Brot, H. Weissbach, A methionine

sulfoxide reductase in Escherichia coli that reduces the R epimer of

methionine sulfoxide, Biochem. Biophys. Res. Commun. (2003)

300 (2003) 378–382.

[19] H.R. Kaback, Bacterial membranes, Methods Enzymol. 22 (1971)

99–120.

[20] N. Brot, J. Werth, D. Koster, H. Weissbach, Reduction of N -acetyl methionine sulfoxide: a simple assay for peptide methionine

sulfoxide reductase, Anal. Biochem. 122 (1982) 291–294.

[21] F.T. Lavine, The formation resolution and optical properties of

the diastereoisometric sulfoxides derived from LL-methionine,

J. Biol. Chem. 169 (1947) 477–491.

[22] S.-I. Ejiri, H. Weissbach, N. Brot, The purification of methionine

sulfoxide reductase from Escherichia coli, J. Bacteriol. 139 (1979)

161–164.

[23] V.K. Singh, R.K. Jayaswal, B.J. Wilkinson, Cell wall-active

antibiotic-induced proteins of Staphylococcus aureus identified

using a proteomic approach, FEMS Microbiol. Lett. 199 (2001)

79–84.

[24] A. Lescure, D. Gautheret, P. Carbon, A. Krol, Novel selenopro-

teins identified in silico and in vivo by using a conserved RNA

structural motif, J. Biol. Chem. 274 (1999) 38147–38154.

[25] J. Moskovitz, N. Jenkins, D.J. Gilbert, N.G. Copeland, F. Jursky,

H. Weissbach, N. Brot, Chromosomal localization of the mam-

malian peptide-methionine sulfoxide reductase gene and its

differential expression in various tissues, Proc. Natl. Acad. Sci.

USA 93 (1996) 3205–3206.

D. Spector et al. / Biochemical and Biophysical Research Communications 302 (2003) 284–289 289