the chemical and biological versatility of riboflavin.pdf

8/11/2019 The chemical and biological versatility of Riboflavin.pdf

http://slidepdf.com/reader/full/the-chemical-and-biological-versatility-of-riboflavinpdf 1/14

The Chemical and Biological

Versatility

of

Riboflavin

V.

Massey'

Department of Biological Chemistry, University of Michigan, Ann Arb or, MI

48 109-0606

U.S.A.

Abstract

Since their discovery and chemical characteriz-

ation in the 193Os lavins have been recognized as

being capable of both one- and two-electron

transfer processes, and as playing a pivotal role in

coupling the two-electron oxidation of most or-

ganic substrates to the one-electron transfers of

the respiratory chain. In addition, they are now

known as versatile compounds that can function as

electrophiles and nucleophiles, with covalent

intermediates of flavin and substrate frequently

being involved in catalysis. Flavins are thought to

contribute to oxidative stress through their ability

to produce superoxide, but at the same time flavins

are frequently involved in the reduction of hydro-

peroxides, products of oxygen-derived radical

reactions. Flavoproteins play an important role in

soil detoxification processes via the hydroxylation

of many aromatic compounds, and a simple flavo-

protein in liver microsomes catalyses many re-

actions similar to those carried out by cytochrome

P450 enzymes. Flavins are involved in the pro-

duction of light in bioluminescent bacteria, and

are intimately connected with light-initiated re-

actions such as plant phototrapiam and nucleic

acid repair processes. Recent reports also link

them to programmed cell death. The chemical

versatility of flavoproteins is clearly controlled by

specific interactions with the proteins with vdhich

they are bound. One of the main thrusts of current

research is to try to define the nature of these

interactions, and to understand in chemical terms

the various steps involved in catalysis by flavo-

protein enzymes.

Introdu ctieamd history

Long before riboflavin was recognized as a vitamin

and was characterized chemically, some of its

more spectacular manifestations drew attention.

Key

words:

elec tron transfer, FAD, flavins, flavoproteins,

FMN.

Abbreviation used:

ETF,

electron-transfer flavoprotein.

'e-mail [email protected]

283

Delivered at King's College, London, on I9 Nove mber

1999 and at the University

of

Leeds, on

22 November I999

VINCENT MASSEY

One of these is described in graphic terms [l] in

the account of the historic voyage of the HMS

Challenger, the first systematic exploration of the

world's oceans, begin-

in

1874

:

'The re was

no

moon, and although the night

was

perfectly clear and the stars shone

brightly, the lustre of the heavens was fairly

eclipsed by that of the sea. The unbroken part

of the surface appeared pitch black, but

wherever there was the least ripple the whole

line h o k e into a brilliant crest of clear white

light. Near the ship the black interspaces

predominated, but as the distance increased

000

Biochemical

Society

n

P



Biochemical Society Transactions 2000) Volume

28,

part

4

the glittering ridges looked closer, until

toward the horizon, as far as the eye could

reach, they seemed to run together and melt

into one continuous sea of light.

’

What was described was, of course, the

bioluminescence of marine bacteria originating

from the enzyme luciferase, acting on reduced

flavin and oxygen.

I

will deal later with the

chemistry of reduced flavins and oxygen, but let us

begin with a brief history of flavins and their

chemical and physical properties.

Approximately 110 years ago an English

chemist by the name of A. Wynter Blyth reported

in the Transactions of the Chemical Society his

work on the chemical composition of cow’s milk

[2]. One of the components that he isolated was a

bright yellow pigment, which he called lacto-

chrome, which was later shown to be the com-

pound that we now know as riboflavin.

After Wynter Blyth’s report, almost half a

century passed before any significant new de-

velopment occurred with his yellow pigment.

The n, in the late 1920sand early 193Os, a tremen-

dous flurry of research took place. Yellow pig-

ments with bright greenish fluorescence were

isolated from a wide variety of sources. Interest in

them became intense when the yellow pigment

was recognized to be a constituent of the vitamin B

complex, and isolation of the vitamin was helped

enormously by the realization that the potency of

the vitamin was correlated with the green fluor-

escence. Some of the foremost chemists of the

time, Richard Kuhn in Heidelberg and Paul

Karrer in Zurich, engaged in what became a bitter

race to determine the structure and prove it by

chemical synthesis. They both succeeded almost

concurrently [3,4], and the name riboflavin was

given to replace the variety of previous names,

such as lactoflavin and ovoflavin, which were

merely descriptive of the source from which it had

been isolated. The name, of course, derives from

the ribityl side chain and the yellow colour of the

conjugated ring system (Figure 1 .

Quite concurrently with this chemical work,

Otto Warburg in Berlin had been carrying out his

pioneering work on the mechanism of biological

respiration, the process by which small substrates

derived from food are converted by living cells

into usable energy, employing molecular oxygen

as the final oxidant in a chain of reactions. In the

process of these studies, Warburg isolated a yellow

protein from yeast which was shown to catalyse

the oxidation of another vitamin cofactor, now

known as NADPH, using oxygen as the second

substrate [S].

Soon after its discovery, Hugo Theorell, a

Swedish biochemist, found that if the enzyme was

precipitated with ammonium sulphate at a pH

around 2, the protein precipitate was white and the

supernatant was yellow. Neither the colourless

apoprotein nor the yellow flavin could catalyse the

oxidation of NADPH ; however, when they were

mixed together at the appropriate pH, activity was

reconstituted. The yellow pigment isolated from

the enzyme was in its spectral and fluorescence

properties indistinguishable from riboflavin, yet

riboflavin would not reconstitute enzyme activity.

Theorell quickly found that the difference was due

to the enzyme flavin having a phosphate residue in

ester linkage at the terminal hydroxy group of the

ribityl side chain [6]. Th is form is known as flavin

mononucleotide, FMN.

Theorell’s work was an important historical

development, as it showed the biochemical basis

for the necessity of riboflavin as a vitamin, i.e. as a

cofactor in enzyme catalysis. In the intervening

years it has of course been shown that most of the

Figure

I

Structures of riboflavin,

FMN

and

FAD

Hd

OH

IroclrYouPnr Rin9Syrt .m

000

Biochemical Society

84



Chemical and Biological Versatility of Riboflavin

water-soluble vitamins function in a similar

fashion, as prosthetic groups of coenzymes for a

wide variety of enzymes. A few years later, Hans

Krebs recognized the existence of another flavo-

protein, D-amino acid oxidase [7], and Warburg

and Christian, in 1938, showed that this protein

could be separated into apoprotein and flavin, in

the same way as had been done with the yeast

yellow enzyme [8]. Again, enzyme activity could

be regained on mixing the apoprotein and the

isolated flavin, but not with riboflavin or FMN.

The structure of the new flavin was correctly

deduced to be that of a condensation product of

FMN and AMP, and finally proved in

1954

by

Todd and his group in Cambridge by total

synthesis [9]. This form is known as flavin adenine

dinucleotide, FAD.

When I started working in the flavin field

some 45 years ago, there were only a handful of

flavoprotein enzymes known. Now they number

in the hundreds, and new ones are reported every

month. Mostly they contain non-covalently bound

FA D or F MN , and are specific for binding either

of the two flavin forms, whatever Nature initially

provided them with. This is readily under-

standable from the more than 40 flavoprotein

crystal structu res now available, which reveal that

the majority of the flavin-protein interactions are

with the N-10 side chain, i.e. the ribityl side chain

of FM N or FAD.

Figure 2

Artificial flavins that have been used as flavin re-

placements in flavoprot eins

etc.

Thus it is a general phenomenon that, pro-

vided one can remove the original flavin under

conditions that avoid denaturing the protein, most

flavoproteins will accept into their active sites

chemically modified flavin ring structures, pro-

vided that the N-10 side chain is the appropriate

one. A large number of chemically modified flavins

are available, with different substituents at various

positions all around the isoalloxazine ring system,

as illustrated in Figure 2.

The property of reconstitution with modified

flavins has proved to be of enormous value in the

study of flavoproteins, since those with chemically

reactive substituents, such as sulphur-containing

ones, or photoreactive ones, such as

6-

and 8-

azidoflavins, can be used as probes of the flavin

environment at that particular position, i.e. to

assess whether the particular position is accessible

to solvent or buried in the protein

[lo].

In this

respect they serve as valuable reporters of protein

or flavin dynamics, especially in conjunction with

knowledge from more static crystal structures.

Some of them, as I will illustrate later, are also

good reporters of the overall charge of the active

site. In addition, there is a beautiful correlation of

the redox potential of the flavin and the electron-

withdrawing or electron-donating properties of

the substituents, making possible determination

of mechanism through linear free energy cor-

relations with activity, substrate dissociation con-

stants and other kinetic and thermodynamic

properties (see [l l-141 for recent examples).

Biological roles

of

flavins

Because of their chemical versatility, flavins are

involved in a host of biological phenomena. They

play a central role in aerobic metabolism through

their ability to catalyse two-electron dehydro-

genations of numerous substrates and to par-

ticipate in one-electron transfers t o various metal

centres through their free radical states. In this

capacity they frequently form parts of multi-

redox-centre enzymes, such as the succinate and

NADH dehydrogenases, xanthine oxidasel

dehydrogenase, cytochrome P450 systems and the

more recently recognized nitric oxide synthase.

A riboflavin-binding protein is involved in

the development of chicken and mammalian

fetuses [15,16], and flavoproteins have recently

been implicated as playing signal transduction

roles in programmed cell death

[

171 and regulation

of biological clocks, such as that involved in

adjustment to jet lag

[18].

They are intimately

involved in soil detoxification of aromatic

285 000

Biochemical Society



Biochemical Society Transactions 2000) Volume

28,

part

4

pollutants [19] and in several light-dependent

processes, such as photosynthesis [20] and light-

dependent repair of DNA damage

-

he photo-

reduction of DNA dimers

[21].

They have also

been shown recently to be the blue-light receptors

long known to be involved in plant phototropism

-

he directional response of growing plants

towards the source of light [22].

Properties

of

flavins

Before we consider the chemistry of some of these

flavin-linked processes, I want to spend a little

time describing some of their fundamental

properties. I have already mentioned their ability

to participate in one-electron transfer reactions,

which automatically implies the existence of semi-

quinone oxidation states.

In free solution, i.e. when not enzyme-bound,

a mixture of oxidized and reduced Aavin very

rapidly sets up an equilibrium in which a certain

amount of flavin radical is formed (Scheme

1).

With free flavins the equilibrium is very much to

the left, so that at pH 7 only about 5 % radical is

stabilized in an equimolar mixture of oxidized and

reduced flavin. The semiquinone can exist in a

S c h e m e I

Flavin semiquinone equilibria

FI, flavin: ox, oxidized ; red, reduced.

HH

H -

+ H+

F i g u r e 3

Spectra of glucose oxidase in the oxidized, semi-

quino ne and

fully

redu ced states

Data are

from [23].

Neutral etniquinone

...

._

'..,

neutral or anionic form, with a p K of - 8.5. On

binding to a specific protein, this equilibrium can

change dramatically. Some enzymes show essen-

tially zero stabilization of semiquinone, while

others give almost

100

% stabilization. With such

cases, the protein may stabilize the neutral radical

species over the whole range of pH values a t which

the enzyme is stable, i.e. the p K is shifted up

significantly from 8 5 In other cases, it is the

semiquinone anion that is stabilized, i.e. the p K is

decreased significantly. Fortunately, with some

enzymes, of which glucose oxidase was the first

example

[23],

the enzyme shows a pK in the

observable range, permitting the identification of

the spectral properties of both forms (Figure

3).

With such large spectral differences between

the various flavin oxidation states, it is clearly

possible to monitor the events occurring in

catalysis using the flavin itself as a reporter. All

flavoprotein reactions involve two separate half-

reactions, which can be monitored separately by

rapid reaction techniques (Scheme

2).

Th e majority of flavoprotein-reducing sub -

strates are dehydrogenated in a two-electron re-

duction step. The resulting reduced flavin is then

re-oxidized by its oxidizing substrate, either in a

two-electron step, as shown in Scheme 2, or in

single one-electron steps,

in which the flavin

semiquinone would be observed as an intermedi-

ate. In some enzymes, molecular oxygen is the

physiological substrate. Because of the general

reactivity

of

reduced flavins with

0

the reductive

half-reaction is studied under anaerobic con-

ditions by mixing enzyme from one syringe of a

stopped-flow instrument with the substrate, AH,,

from the other syringe. In this way, reaction

intermediates can be detected and rate constants of

individual steps determined. The same thing can

be done for the oxidative half-reaction. In this case

the enzyme is pre-reduced by a stoichiometric

amount of substrate, or some chemical reductant

S c h e m e

2

Reductive and o xidat ive half-reactions of flavoproteins

FI, flavin; ox, oxid ized; red, reduced. See the te xt for details.

AH2Y'ox VBH2




such as dithionite, and then mixed in the stopped-

flow instrument with the oxidizing substrate,

oxygen or whatever it may be B in Scheme 2).

Scheme 3 shows the chemical reactions that

occur in the reaction of reduced flavin with oxygen

[24-281. The initial reaction is a one-electron

reduction of

0

by the reduced flavin to yield a

caged radical pair

of

neutral flavin radical and

superoxide. This radical pair now has several

alternative routes. I t can collapse into a flavin C4a

peroxide, a nucleophile, which on protonation

becomes the electrophilic hydroperoxide. The

peroxide species may eliminate hydrogen peroxide

to yield oxidized flavin, or there may be a second

one-electron transfer from the radical pair to give

the same products. T he flavin peroxide species are

involved in hydroxylation reactions, as we will see

later. Th e third alternative route is the dissociation

of the radical pair into its components, flavin

radical and superoxide. T he superoxide produced

can react with peroxide to form hydroxyl radicals

[29],

and with nitric oxide to form peroxynitrite

[30]. Production of hydroxyl radicals is commonly

believed to be one of the major sources of oxidative

stress and tissue damage, largely by reaction with

lipids to produce lipid hydroperoxides. Recent

work also suggests that peroxynitrite is an active

agent in apotosis [31] .

Now I would like to return to some examples

of

the information gained by replacing the native

flavin with artificial flavins. A particularly good

example is that provided by 8-chloroflavins and

their reaction products.

The 8-chloro substituent is readily eliminated

by sulphur nucleophiles (RS-), yielding 8-SR-

flavin and chloride ion

[32].

Huge spectral changes

accompany these reactions, which serve to make

8-C1-flavins very valuable probes of the protein

environment around the flavin. Most 8-SR-flavins

have absorption maxima in the 470 nm region,

with molar absorption coefficients of approx.

25000

M-'.cm-'. Thus , with enzymes where the

dimethylbenzene ring of the flavin is exposed to

solvent, those enzymes in which the native flavin is

replaced by 8-C1-flavin react rapidly with thio-

phenol to give 8-thiophenyl-flavin enzymes,

whereas those enzymes in which the flavin is

buried in the protein fail to react, even over long

periods [33]. Chloride is also displaced from 8-C1-

flavins by Na,S, yielding 8-mercaptoflavins.

Scheme 3

Reactions

of

reduced flavin with oxygen

R R R

~y---+ OZ)y OH aeyer-Villiger reactions

VI I V H 0

0-

'OH

287

0

2000

Biochemical Society



Biochemical Society Transactions (2000) Volume

28, part

4

These flavins exist in several easily distinguishable

forms, the neutral 8-SH-flavin, with a spectrum

similar to those of 8-SR-flavins A, 450 nm;

20000 M- * cm-'), and the

8s-

thiolate anion form

Amax

520-530 nm ; 30000 M-lacm-'), with a pKa

for the transition of 3.8 [32]. On binding to many

flavoproteins (in place of the native flavin), the

spectrum of the 8-mercaptoflavin is shifted dra-

matically

Amax

590-610 nm;

E

30000 M-'*cm-')

[33]. Thi s form is due to the protein stabilizing a

benzoquinoid resonance form, in which the nega-

tive charge is localized in the N l)C 2)

=

0 locus

of the flavin, as shown in Scheme 4.

Most flavoproteins of the oxidase class, which

react rapidly with 0 to give H202s product,

stabilize 8-mercaptoflavin in this benzoquinoid

resonance form, and subsequent crystal structures

of such enzymes have indeed confirmed the exist-

ence of positive charge in this region of the bound

flavin. Such proteins also stabilize the flavin

semiquinone anion, and many also facilitate the

formation of flavin N(S)-sulphite adducts. A

further correlative of protein stabilization of

anionic flavin forms (including that of the reduced

flavin) is to increase the redox potential of the

bound flavin compared with that in free solution.

A dramatic demonstration of the utility of

such flavin replacement studies comes from work

with xanthine oxidoreductase. This enzyme has

been known for many years to exist in two

interconvertible forms [34]

:

xanthine dehydro-

genase, which oxidizes xanthine at the expense of

reduction of NAD+, and xanthine oxidase, which

cannot use NAD+ as an electron acceptor, but

instead employs molecular oxygen, forming both

H202and superoxide as products [35]. The de-

hydrogenase form is converted into the oxidase

form by oxidation of specific cysteine residues to

cystine residues, and the oxidase form can be

reconverted into the dehydrogenase by incubation

with thiols such as dithiothreitol [36,37]. The

redox potentials of the molybdopterin and iron-

sulphur centres of the two enzyme forms are

similar, but the flavin potential is lowered sub-

stantially in the dehydrogenase form, due prin-

cipally to stabilization by the protein of the neutral

flavin semiquinone. On replacement of the native

FAD by 8-mercapto-FAD, dramatic differences

in the spectral properties of the flavin are found

between the two forms. In the oxidase form the 8-

mercapto-FAD

is

clearly stabilized as the benzo-

quinoid resonance form, while in the dehydro-

genase the neutral 8-SH-flavin form is found. By

pH titration, the pK, of the mercaptoflavin was

found to be - 5.5 in the oxidase form, only slightly

perturbed from the free solution pKa of 3.8.

However, in the dehydrogenase form, the pK, was

increased to - 9.0 [37]. Thus i t was clear that, in

the oxidase form, the electrostatic potential of the

protein around the flavin must be positive,

whereas in the dehydrogenase it is probably largely

negative. This prediction has been very nicely

confirmed by the recent determination of the

crystal structures of the two enzyme forms [38].

Catalytic versatility of flavoproteins

Flavoprotein enzymes catalyse a large variety of

different types of reactions. Many attempts have

been made to achieve a rational classification of the

different types of flavoproteins, depending on the

type of chemical reaction catalysed, the nature of

the reducing and oxidizing substrates, the physi-

cochemical properties of the enzymes and, more

recently, their structural motifs as determined

by X-ray crystallography. None of these attempts

has been entirely satisfactory. Nevertheless, it is

clear that enzymes catalysing similar chemical

reactions tend to have common characteristics

which are particular to that group. T he field is too

vast to try to give a comprehensive review here, so

I

will concentrate on describing some enzyme

groups with which I have had experience and

where common properties seem to exist. See [39]

for a more comprehensive review.

Scheme

4

Thiolate and benzoquinoid resonance forms of 8-mercaptoflavins

R

h x

530 nm

hmax

- 600 nm

0

2000

Biochemical Society

288



Chemical and Biological Versatility

of

Riboflavin

Flavoproteins catalysing oxidation of

a-hydroxyacids and a-amino acids

These enzymes bring about dehydrogenation at

the a-carbon atom of the substrate, yielding

respectively the 2-0x0 (a-keto) acid or a-imino

acid as primary product. The enzymes involved

share many common characteristics, although

there are also many differences, illustrating the

difficulty of making a rational classification of

flavoprotein enzymes. In addition to the similarity

of the reactions catalysed, both groups have also

been shown to catalyse the anaerobic elimination

of

chloride from /?-chloro a-amino acids or

/?-

chloro a-hydroxyacids

[40 41]

nd, in the case

of

D-aminO acid oxidase and /?-C1-a-aminobutyrate,

without any sign of flavin reduction during the

reaction [42].Such reactions, and the charac-

terization of a flavin N(5)-glycollyl adduct in the

reaction of L-lactate mono-oxygenase with glycol-

ate as substrate

[43]

ed to the widespread ac-

ceptance

of

the catalytic reaction proceeding via

a carbanion mechanism, in which the primary

step involves abstraction

of

the a-proton of the

substrate by an enzyme base, as illustrated in

Scheme 5.

The anaerobic elimination of chloride could

also be explained simply by this mechanism, as

shown in Scheme

6.

The carbanion mechanism received support

from the crystal structures

of

two representative

a-hydroxyacid-oxidizing enzymes, yeast L-lactate

dehydrogenase (flavocytochrome b,

[44]

and

spinach glycolate oxidase

[45]

which show a

striking similarity

of

amino acid residues sur-

rounding the flavin, and a histidine residue which

Scheme 6

Chloride elimination from /?-chloro a-amino acids

or

P-chlo ro hyd roxyacids by a carbanion mechanism

+-

Scheme 5

Possible carbanion mechanism fo r the dehyd rog enation of a-hydroxyacids

X

289

0

2000

Biochemical Society



Biochemical Society Transactions (2000)

Volume 28,

part 4

could serve as the active-site base required for

a

carbanion mechanism. Remarkably, all the other

known a-hydroxyacid-oxidizing enzymes have the

same conserved amino acid residues as those of

flavocytochrome b, and glycolate oxidase, and

rapid-reaction kinetic experiments are consistent

with them all proceeding by similar mechanisms.

The determination of the crystal structu res of

the D-aminO acid oxidases of pig kidney [46,47]

and the yeast, Rhodotorula gracilis [48], has

sparked a re-evaluation of the carbanion mech-

anism. Since these crystal structures show that

there is no base present in the active site, with this

enzyme at least, direct hydride transfer to the

flavin is favoured and, by implication, the same

mechanism also for the a-hydroxyacid oxidases

[46,48]. While this may indeed be true, i t is as well

to point out that, so far, there has been no

satisfactory explanation for the chloride elimin-

ations in terms of a hydride transfer mechanism. It

should also be emphasized that t he two groups of

enzymes differ in many ways. The hydroxyacid

enzymes all contain F M N as prosthetic group,

while the amino acid oxidases have F A D . The

F M N enzymes bind substrate on the flavin si-face,

while the F A D enzymes interact with substrate on

the opposite re-face [49]. Th e F M N enzymes have

the flavin aromatic ring well protected from

solvent, while this portion of the

F A D

is quite

accessible to solvent in D-amino acid oxidase [SO].

Common characteristics are the stabilization of

anionic flavin forms, such as the anionic semi-

quinone,

N (

5)-sulphite adducts and the benzo-

quinoid form of 8-mercaptoflavin [SO]. As an

illustration of the difficulties in classification, all

flavoprotein oxidases, not only those catalysing the

oxidation of a-hydroxyacids, share these same

characteristics [lo]. But

so

also does yeast

D-

lactate dehydrogenase, flavocytochrome

b,,

which

re-oxidizes the reduced flavin by single-electron

transfer to a protein-bound haem rather than to

0

[511.

Flavoprotein disulphide reductases

This is an ever-growing family of enzymes which

contain an active-site disulphide in addition to the

F A D prosthetic group, and involve a pyridine

nucleotide as one substrate and a disulphide or

dithiol as the other substrate (or alternatively

mercuric ion liganded with thiols in the case of

mercuric reductase). They share many structural

and mechanistic similarities, and a wealth of

information is available (see [S2] for a compre-

hensive review). In some cases, such as lipoyl

dehydrogenase, the reaction is freely reversible,

and physiologically the enzyme functions as part

of an 0x0 acid oxidase multienzyme complex to re-

oxidize protein-bound dihydrolipoic acid at the

expense of reduction of

NAD

to

N A D H .

In the

case of glutathione reductase the physiological

reaction (and thermodynamic equilibrium) is the

utilization of N A D P H to reduce oxidized gluta-

thione. Although differing in detail from one

enzyme to another, the same basic mechanism

appears to apply. Th e reductive half of the reaction

is illustrated in Scheme 7.

In a typical reaction, the reduced pyridine

nucleotide binds rapidly and positions itself over

the re-face of the flavin,

so

that hydride transfer

from the pro(S) position of

N A D ( P ) H

occurs

efficiently. The reduced flavin reacts rapidly with

the active-site disu lphide t o form the flavin C4a-

cysteinyl adduct and a thiol from the other sulphur

of the active-site disulphide. T hi s species, like the

reduced flavin, is rarely observed experimentally,

except when trapped by alkylation of one of the

thiols [53] or by removal of the second thiol by

mutagenesis

[54].

The next step is the elimination

of thiol from the cysteinyl adduct,

Scheme

7

Reductive half-reaction o f a typical flavopr otein disulphide reductase

Em

0

2000

Biochemical Society

290

re-forming

Yo



Chemical and Biological Versatility

of

Riboflavin

oxidized flavin and resulting in t he charge-transfer

complex of active-site thiolate and oxidized flavin

characteristic of this group of flavoproteins. The

oxidative half-reaction of the catalytic cycle does

not involve the flavin, but consists of a series of

thiol-disulphide interchange reactions with the

second substrate, a low-molecular-mass disul-

phide such as lipoic acid or oxidized glutathione,

or a disulphide-containing protein such as thio-

redoxin. In the case of mercuric reductase this

involves an intramolecular reaction with a C-

terminal dithiol liganded with mercuric ion [55].

Flavoprotein mono-oxygenases

Here again we have a broad group of enzymes

which share many properties and mechanistic

features. The major common property is use of

NADH or NADPH to reduce the enzyme flavin

(always FAD) and the formation of a flavin C4a

peroxide on reaction of the reduced enzyme with

molecular oxygen ; this is the reactive oxygen

species that is responsible for oxygenation of the

substrate. The flavin peroxide in its protonated

form is a potent electrophile and it is this form that

is employed in the subgroup of aromatic hydroxyl-

ases, where a second hydroxy group is introduced

into the aromatic subs trate already containing one

hydroxy group [56]. On the other hand, the

peroxide anion, formed initially in the reaction of

the reduced flavin with 0 (see Scheme

3),

is a

good nucleophile, and is employed by the second

subgroup of mono-oxygenases in oxygen-in-

sertion reactions, such as those catalysed by

bacterial luciferase [57] nd cyclohexanone mono-

oxygenase [58]. The two subgroups each have

distinctive common properties. In the aromatic

hydroxylases there is an exquisite control mech-

anism to ensure that NAD(P)H is used only when

substrate is present to be hydroxylated. A complex

of oxidized enzyme and aromatic substrate is

required for rapid reduction of the flavin by

NAD(P)H

;

differences of as much as 100000-fold

in rate constants have been observed in the absence

and presence of substrate. Th e nucleophilic mono-

oxygenases do not display thi s control mechanism,

but have developed an alternative one whereby the

C4a peroxide formed on reaction with 0 is

somehow stabilized by the protein, to such an

extent that it can be isolated by low-temperature

chromatography in the case of bacterial luciferase

[59] nd is reactive only in the presence of the

substrate for oxygen transfer. This feature is

distinct from the situation with the aromatic

hydroxylases, where the hydroperoxide is very

29

unstable in the absence of the substrate to be

hydroxylated.

The reaction mechanism of p-hydroxy-

benzoate hydroxylase has been investigated in

great detail, and is summarized in Scheme 8. The

basic mechanism shown appears to be followed by

all members of this class [60]. These enzymes are

delightful ones to work with experimentally, since

practically every species in the catalytic cycle can

be distinguished on the basis of characteristic

absorbance and fluorescence properties, permit-

ting in the case of p-hydroxybenzoate hydroxylase

the estimation of the values of each of the rate

constants shown in Scheme 8. Thu s the step k

results in a long-wavelength-absorbing charge-

transfer complex with NADPH as donor and

oxidized flavin as acceptor. The step k involves

reduction of the flavin and is accompanied by large

changes in absorbance and fluorescence, and

results in another distinctive long-wavelength

charge-transfer complex, in which reduced flavin

is the donor and NADP+ the acceptor. Th e decay

of this characteristic absorbance permits the de-

termination of k the release of NADP+ to yield

the reduced enzyme substrate complex. It is this

species that reacts with 0 to give intermediate (I) ,

a complex of substrate and the flavin C4a-hydro-

peroxide, which has a distinctive spectrum with a

wavelength maximum at 380 nm, and is formed

with an observed pseudo-first-order rate constant

directly proportional to the oxygen concentration,

yielding the value of k . It is from intermediate ( I)

that the oxygen transfer reaction to substrate

occurs, yielding intermediate (11), a complex of

enzyme 4a-hydroxyflavin and the non-aromatic

dienone form of the product. With some sub-

strates, such as 2,4-dihydroxybenzoate and

p-

aminobenzoate, this species is sufficiently long-

lived to permi t characterization of its absorbance

properties and determination of the rate constant

k . With these substrates the re-aromatization to

give the hydroxylated product is sufficiently slow

to determine the value of k . With p-hydroxy-

benzoate as substrate the non-aromatic inte rmedi -

ate is not observed, presumably because it re-

aromatizes faster than it is formed. In this case

only the rate constant of the hydroxylation step,

k can be determined. The final intermediate

(111),a complex involving the C4a-hydroxyflavin

and product, then undergoes dehydration to give

the oxidized flavin enzyme, ready for the next

cycle of catalysis.

An important finding from structural studies

of p-hydroxybenzoate hydroxylase is that the

0

2000

Biochemical Society



Biochemical Society Transactions

2000)

Volume 28, part 4

flavin exists in two different conformations :one in

which the flavin is mostly buried within the

protein, and positioned ideally for hydroxylation

of the substrate, and the other in which the flavin

ring system is swung out by

30-40”

towards a

solvent-exposed position where it could not

hydroxylate the substrate [61,62]. This inherent

mobility of the flavin has been confirmed in

solution studies by replacement of the native flavin

by the photoreactive 6-azido-FAD

[63],

and

appears to be an important component of catalysis,

the movement of the flavin to the ‘out

’

position

being required for reduction by NADPH and

movement back to the ‘i n’ position being required

for the hydroxylation reaction [64]. The possibility

of this being a general phenomenon is raised by the

recent report that the flavin

of

the related phenol

hydroxylase, the reaction mechanism of which

is very similar to that of p-hydroxybenzoate

hydroxylase [65] s also found crystallographically

in two similar buried and exposed conformations.

Reductases, dehydrogenases and elect ron

transferases

Th is is a somewhat unsatisfactory categorization,

since many enzymes that function in one-electron

transfers also dehydrogenate an organic substrate,

and similarly ones that reduce olefinic bonds also

involve dehydrogenation of the primary reducing

substrate, which in both cases is frequently a

reduced pyridine nucleotide. There are many

different enzymes in this broad grouping; I will

limit discussion to a few individual cases.

Acyl-CoA dehydrogenases

These are widely occurring enzymes involved in

the oxidation of fatty acids, which oxidize acyl-

CoA thioesters to the corresponding enoyl-CoA

esters. Th e reduced flavoprotein

so

formed is re-

oxidized by successive one-electron transfers to

another flavoprotein, the electron-transfer Aavo-

protein (ETF), which in mammals is in turn re-

Scheme

8

Reaction mechanism

of

p-hydro xybenzoate hydroxylase

E-FAD,,

pH

R

.

NADPH

HOH

”

0

2000

Biochemical Society

k

I

R

NADP+

H o

292




-FAD NADPH NADP

-FUN

\f

oxidized by a third flavoprotein, ETF-ubiquinone

reductase

[66].

Crystal structures of both mam-

malian and bacterial acyl-CoA dehydrogenases are

available, and permit an authoritative evaluation

of the reaction mechanism, which involves re-

moval of the substrate a -pro ton by an active-site

aspartate residue and a concerted hydride transfer

from the /?-position to the flavin N-5 position

(Scheme

9) [66].

Cytochrome P450 reductase

Th is enzyme is responsible for reducing the haem

moiety of the widespread cytochromes of the P450

family. It is unusual

in containing equimolar

amounts of FAD and FMN, which were both

shown to be stabilized as the neutral semiquinone

on reductive titration, but with very different

redox potentials

[67].

Studies on the enzyme in

which the FM N was removed, but leaving the

FAD intact

[68],

and in which the FMN was

-FADH2

S c h e m e

9

Oxidation of substrate by acyl-CoA dehydrogenases

R

do

replaced by a series of artificial FMNs of different

redox potentials

[69],

showed that it

is

the FAD

that is reduced by NADPH. Rapid internal elec-

tron transfer occurs between the two flavins, with

the equilibrium distribution of flavin redox states

being determined by their relative redox potentials

[70], in a fashion similar to that which had been

found for xanthine oxidase [71]. Reduction of

cytochrome P450 is by electron transfer from the

fully reduced F MN , and during steady-state turn-

over the enzyme cycles predominantly between

the one- and three-electron reduced states, with

the FAD acting as an electron buffer, as shown in

Scheme 10.

'Old Yellow Enzyme'

Finally, I want to return to Warburg's 'Old

Yellow Enzyme

,

whose physiological function in

yeast still remains unknown. Despite lack of

knowledge of its true function, a lot is known

abou t this enzyme. One of its most characteristic

properties is its ability to form beautiful charge-

transfer complexes with aromatic and hetero-

aromatic compounds containing an ionizable

hydroxy group [72]. Th e enzyme from brewer's

bottom yeast, the source used by Warburg, and

the similar enzyme from baker's yeast,

Succhuro-

myces cerevisiue,

were shown to be a mixture of

isoenzymes

[73]

derived by subunit association

from two separate genes

[74].

The expression of

the protein from a single gene from brewer's

S c h e m e 10

Catalytic turno ver with N AD PH-cy toc hro me P450 reductase

-

e

-FAD

FMNK

NADPH

Priming

293

Turnover

000

Biochemical Society



Biochemical Society Transactions

2000)

Volume

28,

part 4

bottom yeast (OYEl) permitted the isolation of

high-quality crystals and structural de termination

of the protein [75]. The wave number maximum

of the long-wavelength band formed with different

p-substituted phenols is correlated well with the

Hammett 6p substituent of the phenol [76], and

with the one-electron redox potential of the

enzyme-bound flavin when the native FM N was

replaced with a series

of

artificial Aavins [77].

These correlations were historically very import-

ant in establishing the existence of charge-transfer

complexes in flavoproteins. The crystal structure

of the enzyme with bound phenol clearly shows

the parallel stacking of phenol and flavin required

for the observed charge-transfer interaction. The

ligand-binding site is largely hydrophobic, but

His-191 and Asn-194 form hydrogen bonds with

the phenolic oxygen and thus contribute to the

lowering of the phenolic pK, and its binding in the

phenolate anion form.

A breakthrough in determination of the func-

tion of Old Yellow Enzyme came from the find-

ing that it catalysed efficiently an NADPH-

cyclohexenone reductase activity, in which the

olefinic linkage, not the carbonyl function, was

reduced [74]. It quickly became evident that the

enzyme was capable of reduction

of

a large number

of a,/?-unsaturated aldehydes and ketones, and

additionally that it could catalyse dismutation

reactions, such

as

that with cyclohexenone, where

the product of anaerobic incubation of the enzyme

with cyclohexenone was an equimolar mixture of

cyclohexanone and phenol [78] (Scheme 11). The

ability of the enzyme to be reduced by a saturated

aldehyde or ketone is clearly dependent on the

redox potential of the bound flavin, as shown

in studies where the native flavin was replaced

Scheme I I

Cyclohexenone dismutase reaction catalysed by the

Ol d Yellow Enzyme

E, enzyme,

FI,

flavin; ox, oxidized; red, reduced.

with the high-potential 8-cyano-FMN, which

effectively converts the enzyme from an NADPH-

enone reductase into an oxygen-dependent de-

saturase [79].

The above reactions have been found to be

highly stereospecific. The reduction of unsatur-

ated carbonyl compounds, such as cinnam-

aldehyde, was shown to involve a trans-addition

across the double bond, with the P-hydrogen being

derived from the pro-R-hydrogen of NADPH [via

solvent-protected flavin N-( S)], and the a -hyd ro-

gen being derived from solvent [78]. Thi s stereo-

specificity is conserved in the oxidation of satur-

ated aldehydes and ketones by 8-C N-FM N-OY E

The crystal structure suggested a possible

candidate as a carrier for the hydrogen that is

placed at the a-carbon atom of the saturated

product, i.e. Tyr-196, which is ideally located to

serve as a solvent-exchangeable acid for proton-

ation of the a-carbon. This was confirmed by

changing this residue

to

phenyalanine, whereby

the reactivity of the mutant enzyme with cyclo-

hexenone was decreased by a factor of 5

x lo5,

and

that with cinnamaldehyde by a factor of 1.1

x

lo3

[80]. Modelling of cyclohexenone into the active

site of OYE, with the carbonyl oxygen hydrogen-

bonded t o His-191 and Asn-194 in the same way

as with phenolates, places the /?-carbon in an ideal

position to receive a hydride from the flavin

N-5,

and the a-carbon in the correct position to receive

a proton from Tyr-196 in a trans-addition reaction

across the olefinic bond [80].

Th e role of Tyr-196 as proton donor to the a-

carbon has also been demonstrated in a striking

fashion in the NADPH-dependent reduction of

1791.

Scheme I 2

Reduction

of

nitrocyclohexene by the Old Yellow

Enzyme

R

4v0-

R

0

2000 Biochemical Society

294




nitrocyclohexene. Nitrocyclohexene is a very good

substrate for the enzyme, with the nitro group

functioning to activate the olefinic bond in an

analogous fashion to that with the carbonyl group

of aldehydes and ketones. In this case, however, a

discrete intermediate, identified by its spectral

properties as the nitronate form of nitrocyclo-

hexane, is formed and released from the enzyme

[81] (Scheme 12). Th e nitronate is a stable species

and in free solution is only slowly protonated at

the a-position t o give nitrocyclohexane. Th e

NADPH-dependent reduction of nitrocyclo-

hexene proceeds just as fast with the Tyr-196

Phe mutant enzyme as with wild-type enzyme,

with the rapid accumulation of the nitronate

product. The wild-type enzyme also catalyses

efficiently the protonation of the nitronate, but the

Tyr-196 Phe enzyme is incapable of catalysing

the conversion, thus confirming convincingly the

role of Tyr-196 as an active-site acid in the overall

reaction.

While the physiological substrate (or sub-

strates) of

OYE

has not yet been identified,

it

seems very likely that it is an a -unsaturated

compound. Since the cloning of the O Y E 1 gene

from brewer's bottom yeast and of the

0

Y E 2 and

0

Y E 3 genes from

S

cerevisiae [74,82],

a

number

of

related enzymes have been identified from

plant and bacterial species. One of these, from

Pseudomonas putida functions as a morphinone

reductase [83], and ones from plant sources as a

12-oxophytodienoic acid reductase [84,85]. Both

of

these reactions involve NAD(P)H-linked re-

duction of a,&unsaturated carbonyl compounds.

All members of the family contain FMN as the

prosthetic group, and all catalyse the reduction of

cyclohexenone.

References

I

2

3

4

5

6

7

8

9

10

I I

Thompson, C.

W.

( 1877) Voyage o f the Challenger, vol. 2,

p. 85, Macmillan Co., Lon don

Blyth. A.

W.

( I 879) J. Chem. SOC. 35,530-539

Kuhn,

R,

Reinemund,

K.

and Weygand,

F (I

934) Ber. 67,

1460- I463

Karrer, P., Schopp,

K.

and Benz,

F.

(I 935) Helv. Chim. Acta

I 8 , 42& 429

Warburg,0 nd Christian, W. ( 1933) Biochem. Z. 266,

3 7 7 4 1

I

Theorell,

H.

( I 935) Biochem. Z. 275, 344-346

Krebs, H. A. ( I 935) Biochem. J. 29, 1620-1 625

Warburg. 0 nd Ch ristian,

W. ( I

938) Biochem. Z. 298,

150-155

Christie, S. M. H., Kenner, G. W. and Tod d, A. R. I 954)

J.

Chem. SOC. 46-52

Ghisla,

S.

and Massey, V.

(

1986) Biochem. J. 239,

I-

2

Eckstein,

J. W.,

Hastings,

J. W.

and Ghisla, S.

(

1993)

Biochemistry 25, 8095-8 I0 2

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

Miller, J.R. and Edmondson,

D.

E. (I 999) J. Biol. Chem . 274,

Ortiz-Maldonado. M., Ballou, D. P. and Massey, V. ( 1999)

Biochemistry 38,8 124-8 I3 7

Yorita, K., Misaki, M., Palfey, 6. A. and Massey, V. (2000)

Proc. Natl. Acad. Sci. U.S.A. 97, 24 8G 24 85

Ramana Murty, C.

V.

and Adiga . P. R

( I

982) Science 216,

191-193

White, H. B. and Me nill, A. H. ( I 988) Annu. Rev. Nutr . 8,

279-299

Susin. S.A., Lorenzo, H. K., Zamzami, N., Matzo, I.,Snow,

6.

E., Brothers G. M., Mangion,J.,Jacotot,E., Con stantini, P.,

Loeffler,

M.

et al.

I

999) Nature (London) 397,44 I4 4 5

Ishiwaka,

T.. Todo , T.. Ma tsum oto. A,, Tan imuta, T.. Togashi,

S. and Ueda, R ( 1999) in Flavins and Flavoproteins I999

(Ghisla, S., Kro neck P.. Macheroux. P. and Sund, H., eds),

pp. 3 17-320, Rudolf We be r Agency for Scientific

Publications, Berlin

Dagley,

S.

I 987) Annu. Rev. Microbiol. 4 I, 1-23

Zanetti, G. and Aliverti, A.

( I

99

I )

in Chem istry and

Biochemistry

of

Flavoenzymes, vol

II

(Muller,

F.,

ed.),

pp.

305-35

I ,

CRC Press, Boca Raton

Joms,M. S. Wang, 6. and Jordan,S. P.

( I

987) Biochemistry

Briggs,W. R., Christie, J. M., Knieb, E. and Salomon, M.

I 999) in Flavins and Flavoproteins I999 (Ghisla, S.,

Kroneck P., Macheroux, P. and Sund, H.,eds), pp. 299-308,

Rudolf W eb er Agency for Scientific Publications, Berlin

Massey, V. and Palmer, G.

I 966) Biochemistry 5,

Massey. V., Palmer, G. and Ballou, D. P.

(

1973) in Oxidases

and Related Systems (King,T.

E.,

Mason, H.S. and Morrison,

M., eds). pp. 2 5 4 3 . U niversity Park Press, Baltimore

Kemal, C., Chan, T.

W.

and Bruice. T. C. ( I 977) J. Am.

Chem. SOC. 99 272-7286

Bruice,

T.

C.

( I

984)

Isr.

J.

C hem. 24 , 5 66

Palfey, B. A., Ballou. D. P. and Massey, V. ( 1995) in R eactive

Oxygen Species in Biochemistry (Valentine,

J.

S., Foote, C. S.,

Liebman.

J.

and Greenberg, A,, eds), pp. 37-83, Blackie

Academic, Londo n

Halliwell. B. ( 1995) in Reactive Oxygen Species in

Biochemistry (Valentine,

J.

S.. Foote, C. S. iebman,J. and

Greenberg, A., eds). pp.

3

13-335, Blackie Academ ic,

London

Huie , R.

E.

and Padmaja. S.

(

1993) Free Radical Res.

Commun. 18. 195-199

23515-23525

26, 68 10-68 I 6

3 I8 1-3 I89

30 Emster, L. ( 1993) in Active Oxygens, Lipid Peroxides and

Antioxid ants (Yagi, K., ed.), pp. 1-38, Japan Scientific

Societies Press, Yokyo

3

I Cassina, A. M., Hodara, R., Souza, J. M., Thomson, L., Castro,

L., Ischiropoulos,

H.,

Freeman, B. A. and Radi,

R.

(2000)

J. Biol. Chern. 275, in the press

32 Moore, E. G., Ghisla, S and M assey, V. ( I 979) J. Biol. Chem.

33

Massey, V., Ghisla, S. and Moore, E. G. (1979)

J.

Biol. Chem.

34 Batteli, M. G., Lorenzoni,

E.

and Stirpe, F.

( I

973) Biochem.

1.

35 Hille, R. and Nishino,

T. I

995) FASEB

J.

9, 995- I003

36 Saito,

T.

and Nishino,

T. (

1989)

J.

Biol. Chern. 264,

37 Hunt, J. and Massey, V. (I 992)

J.

Biol. Chem . 267,

254, 8 173-8

I78

254,9640-9650

131, 191-198

10015-10022

21479-21485

295

0

2000

Biochemical Society



Biochemical Society Transactions 2000) Volume 28, part 4

38 Enroth, C., Eger,

6.

T., Pai, E. F., Okamoto, K., Iwasaka,

T.,

Nishino, T., Hori, H. and Nishino, T.

( I

999) in Flavins and

Flavoproteins I999 Ghisla,

S.

Kroneck P., Macheroux,P.

and Sund, H., eds), pp. 783-786, Rudolf Weber Agency for

Scientific Publications, Berlin

39 Palfey, 6. A. and Massey, V. (1996) in Comprehensive

Biochemical Catalysis Sinnott, M., ed.), pp. 83- 154,

Academic Press, New York

40 Walsh, C. T., Schonbtunn, A. and Abeles, R H. (I97I)

J. Biol. Chem. 246, 6855-6866

41 Walsh, C. T., Lockridge, O., Massey, V. and Abeles,

R H.

(I973)

J.

Biol. Chem. 248, 7049-7054

42 Walsh, C. T., Krodel, E., Massey, V. and Abeles, R H.

( I

973)

J. Biol. Chem. 248, 19461955

43 Ghisla, S. and Massey, V. (I 980) J. Biol. Chem. 255,

44

Xia,

Z.

X. and Mathews, F.

S.

(1990)J. Mol. Biol. 2 12,

837463

45 Lindqvist, Y. and Bidnden, C. I. (I989) J. Biol. Chem. 264,

3624-3628

46 Mattevi,A, Vanoni, M. A., Todone, F., Rini, M., Teplyakov,

A., Coda,

A,

Bolognesi, M. and Curti, B.

(I

996) Proc. Natl.

Acad. Sci. U.S.A. 93,7496-750

I

Setoyama, C. and Mium

R (I

996)

J.

Biochem.

(Tokyo)

120,

14-17

48 Pollegioni. L,Ghisla,

S.

Umhau,

S.

Molla G., Hams, C. M.

and Pilone, M.

S. (I

999) in Flavins and Flavoprotein I999

Ghisla,

S.

Kroneck P., Macheroux, P. and Sund. H. eds),

pp. 55 1-558, Rudolf Weber Agency for Scientific Publications,

Berlin

49 Manstein,

D.,

Massey, V., Ghida, S. and Pai, E. F. 1988)

Biochemistry 27,230C-2305

50 Schopfer, L M., Massey,V. and Claibome, A. 198

I )

J Biol.

Chem. 256,7329-7337

5 I Lederer, F. ( I99I) in Chemistry and Biochemistry

of

Flavoenzymes Muller,

F.,

ed.), vol. II,pp. 153-242, CRC

Press.

Boca Raton

52 Williams, C. H.

( I

992) in Chemistry and Biochemistry

of

Flavoenzymes Muller, F., ed.), vol. Ill,pp. I21-2

I I,

CRC

Press, Boca Raton

53 Thorpe, C. and Williams, C. H. (I98

I)

Biochemistry20,

I507- I5 I3

54 Miller,

S.

M., Massey,

V.,

Ballou, D.P., Williams, C.

H.,

Distefano, M.

D.,

Moore, M.J.and Walsh, C. T. I 990)

Biochemistry 29,283 1-284

I

55 Miller,

S.

M., Massey, V., Williams. C. H., Ballou, D.

P.

and

Walsh, C. T. (I99I) Biochemistry 30,260C-26 I2

56 Entsch, B., Ballou,

D.

P. and Massey, V. (I976) J. Biol. Chem.

25I,2550-2563

57 Macheroux, P.. Ghisla

S.

and Hastings,

J. W. (

1993)

Biochemistry 32, I4183- I4 86

58 Ryenon. C. C., Ballou, D. P. and Walsh, C. T. 1982)

Biochemistry 2I,2644-2655

59 Hastings,

J.

W.,

Balny, C., LePeuch, C. and Douzou, P.

(I973) Proc. Natl. Acad. Sci. USA 70, 3468-3472

60 Entsch, 6. and van Berkel.

W.

J. H. (I995) FASEBJ. 9,

476483

5688-5696

47 Mizutani, H., Miyaham I., Hirotsu, K., Nishina, Y., Shiga, K.,

61 Schreuder, H. A,, Mattevi, A., Obmolova, G., Kalk K. H., Hol,

W.G.J., Van der Bolt, F.

J.

T. and van Berkel,

W.

J. H.

( I

994)

Biochemistry33,

I0

6I-

0

70

Ballou,

D.

P.

and Ludwig, M. L. (1994) Science 266,

I

I

C-

I

I4

Biochemistry 36, I57I

3-

I5723

Massey, V.

(

1999) Biochemistry 38,

I

153-

I

I58

62 Gatti, D. L., Palfey, B. A., Lah, M.

S.

Entsch,

B.,

Massey, V.,

63 Palfey,

6.

A., Ballou, D.P. and Massey, V.

( I

997)

64 Palfey,

6.

A., Moran, G.R., Entsch,

B.,

Ballou,D.

P.

and

65 Enroth, C., Neujahr, H., Schneider, G. and Lindqvist, Y.

66 Thorpe, C. and Kim,

J.

-1. P. (1995) FASEB J. 9,7 18-725

67 lyanagi,

T.,

Makino,

N.

and Mason, H.

S.

(

1974) Biochemistry

68 Vermilion,

J

L and Coon, M. J. ( I 978)

J.

Biol. Chem. 253,

69 Vermilion,

J.

L, Ballou,

D. P.,

Massey, V. and Coon, M.

J.

( I

98

I )

J. Biol. Chem. 256, 266-277

70 Oprian,

D. D.

and Coon, M. J.

(I

982) J. Biol. Chem. 257,

8935-8944

71 Olson,

J.

S.,

Ballou,

D.

P.,

Palmer, G. and Massey, V. 1974)

J.

Biol. Chem. 249, 4363-4382

72 Abramovitz, A S. and Massey, V. (I976)

J,

Biol. Chem. 25

I,

73 Mium

R,

Yamano, T. and Miyake,

Y.

(I986)

J.

Biochem.

74 Stott, K., Saito,

K.,

Thiele,

D.J.

and Massey, V.

( I

993)

J

Biol.

(

1998) Sttucture 6, 605-6 I7

13,

1701-1710

88 12-88 I9

5327-5336

(Tokyo) 99,90 1-906

Chem. 268,6097-6 I06

75 FOX,

K.

M. and K~@ Js ,P.A. (1994) Structure 2, 1089- I I05

76 Massey, V., Schopfer, L. M. and Dunham,

W. R (I

984) in

Flavins and Flavoproteins Btay, R C.,Engel, P. C. and

Mayhew,

5.

G., eds), pp. I9 1-2 10,Walter de Gruyter, Berlin

77

Stewart R C. and Massey, V. (1985) J. Biol. Chem. 260,

78 Vaz,

A

D.

N.,

Chaktabotty,S. and Massey, V.

( I

995)

79 Murthy, Y.

V. S. N.,

Meah, Y. and Massey, V.

( I

999) J. Am.

80 Kohli, R M. and Massey, V.

( I

998) J. Biol. Chem. 273,

8

I

Massey,V.. Meah,Y.. Xu, D.and Brown, B.

J. (

1999) in

Flavins and Flavoproteins I999 Ghisla,

S..

Kroneck

P..

Macheroux P. and Sund, H., eds), pp. 645-654, Rudolf

Weber Agency for Scientific Publications, Berlin

82 Niino, Y. S., Chakrabotty,

S. Brown, 6.

J. and Massey, V.

(I

995)

J

Biol. Chem. 270, 1983- I99

I

83 French, C.

E.

and Bruce,

N.

C.

( I

995) Biochem.

J 3

12,

67 1-678

84 Schaller, F. and Weiler, E.

W. ( I

997)

J

Biol. Chem. 272,

28066-28072

85 Strassner,

J.,

Furholz, A., Macheroux, P., Amrhein.

N.

and

Schaller,A (1999) J. Biol. Chem. 274, 35067-35073

13639-1 3647

Biochemistry 34,42464256

Chem. SOC. I 2 , 5344-5345

32763-32770

Received2 May 2000

0

2000

Biochemical Society

296

the chemical and biological versatility of riboflavin.pdf

Documents