THEJOURNAL

OF BIOLOGICAL
CHEMISTRY Vol. 262, No. 4, Issue of Fehruary 5, pp. 1485-1492. 1YX7
B 1987 by The American Society of Biological Chemists, Inc. Printed in 1J.S.A

Activation of Dopamine 8-Monooxygenase by External and Internal

Electron Donors in Resealed Chromaffin Granule Ghosts*
(Received for publication, March 10, 1986)

Natalie G . Ah& and JudithP. Klinmant

From the Department of Chemistry, University of California, Berkeley, California 94720

Membrane ghosts derived from chromaffin vesicles which is driven by a chemiosmotic proton gradient supported
of bovine adrenal medullas have been usedto examine by ATP hydrolysis, prior to hydroxylation, Equation 1:
the mechanism of reduction of dopamine B-monooxy-
genase in its compartmentalized state. The rate of the
dopamine 8-monooxygenase-catalyzed conversion of +Pi
dopamine to norepinephrine is greatly stimulated by
the presence of ATP, reflecting substrate hydroxyla-
tion on the ghost interior subsequent to the active
transport of dopamine. We demonstratea 2-3-fold
increase in the turnover rate forghosts resealed with ’ DA+Oz
I

0.2-2 mM potassium ferrocyanide, conditions leading

to a slight decrease in the rate of dopamine transport. 1
2e-,2H+ DPM
These data provide the first evidence that an intrave- NE + HZ0
sicular pool of reductant can activate dopamine 8-mon-
ooxygenase, as required by models in which vesicular where DA,’ NE, and DBM represent dopamine, norepineph-
ascorbatebehaves as enzyme reductant. Although rine, and dopamineP-monooxygenase, respectively.
there issufficient catecholamine (endogenousplus sub- Extensive studies haveelucidated many chemical details of
strate) to keep internal ferrocyanide reduced in these the dopamine P-monooxygenase mechanism (1-6). However,
experiments, an additional 2-3-fold increase in turn- thesestudies employ the solubleform of DPM and basic
over occurs in the presence of 0.2-2 mM ascorbate on features of the mechanismof dopamine hydroxylation within
the ghost exterior. The magnitudeof this activation is the chromaffin vesicle remain unanswered.A fundamental
found to be constant at all concentrations of internal question concerns the mechanismwhereby DPM obtains re-
ferrocyanide (both below and above saturation), imply- ducing equivalents in its compartmentalized state.
ing that reductantson opposite sides of the membrane Ascorbic acid is the best electron donor for DPM in vitro
behave independently. Replacement of ascorbate by (7). Inaddition, 10-20 mM ascorbatehas beenmeasured
potassium ferrocyanide as external reductant leads to within bovine chromaffin vesicles (8, 9), implying that it
almost identical results, andwe are able to rule out an functions as thein vivo electron donor aswell. Recent studies
inward transport of dehydroascorbate as the source of have shown that a membrane-bound cytochrome beGIin ghost
activation by external ascorbate. We conclude that vesicles can mediate transfer of electrons from ascorbate in
externalreductants are capable of reducing mem- the ghost interior to the electron acceptorsferricyanide and
brane-bound dopamine 8-monooxygenase from the ex- ferric cytochrome c in the external medium (10-12). Accord-
terior face of the vesicle, either by direct reduction or ingly, a model for DPM reduction hasbeen proposed in which
through a membrane-bound mediator. It appears that ascorbate within the vesicle is regenerated by external elec-
two viable modes for reduction of dopamine B-monoox- trondonors via transmembraneelectrontransferthrough
ygenase may exist in vivo, involving the reduction of cytochrome bM1.
membrane-bound enzyme by cytosolic ascorbate as In an important initial study, Grouselle and Phillips (13)
well as the reduction of soluble enzyme by the pool of reportedtheassay of DPM inchromaffin vesicle ghosts.
intravesicular ascorbate present in chromaffin vesi- Although this study indicated a role for external reductants
cles. in the activation of DPM, the authors failed to see a role for
either ascorbate orferrocyanide located in the ghost interior.
To clarify this issue, the ATP-dependent hydroxylation of
dopamine hasbeen studied in ghosts which are preloaded with
Chromaffin granule ghostvesicles derived from bovine ad- varying concentrationsof potassium ferrocyanide and subse-
renal cells contain all proteinsnecessary to convert dopamine quently exposed to either ascorbate or ferrocyanide in the
to norepinephrine. Theprocess involves uptake of dopamine, assay medium. The results presented herein demonstrate first
that both internally and externally localized reductants acti-
* This work was supported by National Institutes of Health Grant vate hydroxylation and second that they appear to act inde-
GM 25765. The costs of publication of this article were defrayed in
part by the payment of page charges. This article must therefore be The abbreviations used are: DA, dopamine; NE, norepinephrine;
hereby marked “aduertisement” in accordance with 18 U.S.C. Section DBM, dopamine P-monooxygenase; sDBM, soluble dopamine P-mon-
1734 solely to indicate this fact. ooxygenase; mbDBM, membrane-bound dopamine 0-monooxygenase;
$ Present address: Dept. of Medicine RG-26, University of Wash- Hepes, 4-(2-hydroxyethyl)-l-piperazineethanesulfonic acid; Mes, 4-
ington, Seattle, WA 98195. morpholineethanesulfonic acid; SDS, sodium dodecyl sulfate; HPLC,
To whom correspondence should be addressed. high performance liquid chromatography.

1485
1486 Dopamine P-Monooxygenase Activation
in
pendently of one another. A mechanism invoking a transfer
of electrons from an external donor to an internal mediator
is unnecessary to explain our data, which instead suggest that
electron donors may reduce membrane-bound enzyme from
the external membrane surface.
EXPERIMENTALPROCEDURES
Materials
Reagents were analytical grade wherever possible and were pur-
chased from the following sources:disodium fumarate, vanadium-free
disodium ATP, Hepes, Mes, n-octyl-p-D-glucopyranoside, dopamine
. HC1, and (-)norepinephrine bitartrate, Sigma; ascorbic acid, British
Drug House; potassium ferrocyanide, Mallinckrodt; CuSOl. 5Hz0, J.
T. Baker Chemical Co.; bovine liver catalase, Boehringer Mannheim;
bovine serum albumin, Behring Diagnostics; Ficoll and Sephadex G-
25, Pharmacia; Surfactol-100, Westchem; sodium dodecyl sulfate
(SDS), P-mercaptoethanol, and polyacrylamide gel reagents, Bio-Rad.
[1-“CIDopamine and 3H20 were from Amersham Corp. and [“C]
methylamine, [14C]thiocyanate, and [“Cldextran were from New
England Nuclear.
Methods
Absorbance measurements were carried out on a Cary 118 spectro-
photometer and pH measurements on a Radiometer pH meter 26.
HPLC separations utilized a Beckman Model 332 system and scintil-
lation counting a Beckman LS8000 spectrometer. The scintillation
mixture contained 2.85 g/liter 2,5-diphenyloxazoleand 35.7 mg/liter
1,4-bis[2-(5-phenyloxazolyl)]benzene dissolved in toluene with 30%
Surfactol-100 detergent. For measurements of oxygen consumption,
a Yellow Springs Instrument Model 53 biological oxygenmonitor was
employed. Protein concentrations were determined by the method of
Bradford (14) using the Bio-Rad protein assay with bovine serum
albumin as a standard. This method has been reported to show little
interference by contaminating catecholamines (15).
SDS-polyacrylamide gel electrophoresis was carried out according
to Laemmli (16) using a 10% running gel and a 5% stacking gel. Gels
were stained overnight with a solution of 0.1% (w/v) Coomassie Blue,
25% (v/v) isopropyl alcohol, 10% (v/v) acetic acid, and 0.1% (w/v)
cupric acetate. Protein samples were prepared by boiling for 3 min in
1%SDS, 10%glycerol, 2.5% p-mercaptoethanol, and 35 mM Tris, pH
6.8. Intensity of staining inselected bands was quantitated at 530 nm
using a Kratos Model 5D3000 Spectrodensitometer with a Kratos
SDC Density Computer and a Hewlett-Packard 3380A Integrator.
Preparation of Chromaffin Granule Ghosts
Our ghost preparations use techniques borrowed from many re-
searchers in the chromaffin granule field (17-20). About one pound
of bovine adrenal glands was collected from a local slaughterhouse.
Medullas were dissected out within 2 h of slaughter and homogenized
in 0.3 M sucrose. Fragments of medullas and intact nuclei were
removed bylow speed centrifugation (620 X g for 10 min) and the
supernatant was centrifuged at higher speed (10,400 X g for 30 min)
to concentrate the organelles. Chromaffin granules in this fraction
were purified away from mitochondria, lysosomes, and microsomes
by centrifugation through a 1.6 M sucrose density step gradient
(58,400 X g for 90 min). The granules were then hypoosmotically
lysed by >100-fold dilution into a buffer containing 5 mM Hepes, pH
7, 5 mM fumarate, 20 pg/ml catalase, 4 p M cuso4, and varying
concentrations of K,Fe(CN)+ These preparationswere allowed to sit
at 4 “C for 60 min. Lysed granule membranes were collected by
centrifugation (18,000 X g for 30 min) and resuspended into approx-
imately 2 ml of the same buffer. The granules were then loaded onto
a Sephadex G-25 column (1.2 X 20 cm) which was pre-equilibrated
with a buffer containing 5 mM Hepes, pH 7,5mM fumarate, 150 mM
KCl, 20 pg/ml catalase, 4 p M CuSO,, and varying concentrations of
K4Fe(CN)6and allowed to flow by gravity. It has been reported that
high salts in the medium lead to resealing of the vesicles in which the
original sidedness is maintained (17). During this process, catechol-
amine and other low molecular weight constituents rapidly separated
from the membranes, which were collected in the void volume frac-
tion. The membranes were allowed to sit in the isoosmotic buffer for
60 min and collected by centrifugation. As a final wash, pellets were
resuspended into a buffer containing 5 mM Hepes, pH 7, and 150 mM
KCl. They were then collected by centrifugation onto a 10 mM Hepes,
10% (w/v) Ficoll D20pad and resuspended to a final proteinconcen-
Chromaffin Granule Ghosts
tration of about 5 mg/ml with the final wash buffer. The entire
isolation procedure lasts 12-14 h from the time of slaughter and yields
approximately 20 mg of ghost membrane protein. Experiments with
ghosts were performed immediately and generally completed within
9-15 h after isolation, during which time the membranes were stored
on ice.
Protein determinations were found to be artifactually low unless
the membranes were solubilized hence, octyl glucoside was added to
each sample and standard to3.3% (w/v), prior to addition of the dye
reagent (final detergent concentration, 0.1% (w/v)). Catecholamine
(norepinephrine + epinephrine) concentrations inthe resealed ghosts
were estimated from the integration of absorbance peaks on HPLC
traces and found to be in the range of 60-100 nmol/mg of protein.
Measurement of ApH, A#, and Intravesicular Water Space
Standard techniques were used for measuring ApH, A$, and intra-
vesicular water space (21) which wereapplied to chromaffin granules
by Johnson et al. (17) and Johnson and Scarpa (22).
(a) Intravesicular water space was estimated using [“Cldextran as
a membrane-impermeant label and 3Hz0 asa permeant radiolabel.
Chromaffin granule ghosts were incubated in the presence of both
labels for 10 min and concentrated by centrifugation, after which
aliquots of both supernatant and pellet were counted. Internal vesicle
volume per milligram of protein is given as:
vesicle vol (pl) = Vol, (pl) X
I
’Hpllet l4cpeuet
-
’H,,
“C,,
--
(b) Measurements of ApH and A$ were performed the same way,
1
replacing [“Cldextran by [’4C]methylamine(MA) for measuring ApH
and by [“C]thiocyanate (SCN) for measuring A$. (“C/3H)p~~et ratios
were corrected for the presence of extravesicular water in the pellet:
= RT/F logJS + ( S - 1)(X/1 - X ) ] = RT/F In S - X/1 - X
R=
where X = ([“C]dextran/3H20)~~~e,/([14C]dextran~Hz0)~w~mt,
([“ClMA/3H~O),~~,~/([14ClMA/3H~O)~~~~~mt,
and S = ([“ClSCN/
3HzO)p,~,/([14ClSCN/3Hz0)~upsmamt.
Assays for Dopamine Transport and Turnover
In all assays using radiolabeled dopamine, chromaffin granule
ghosts were incubated for approximately 10 min at 35 “C ina medium
containing 50 mM Hepes, pH 7, 150 mMKC1, 6 mM MgS04, +6 mM
ATP, and varying concentrations of external reductant. Final mem-
brane protein concentration was about 0.5 mg/ml. At t = 0, [“C]
dopamine was added to a final concentrationof 100 p~ (-10 Ci/mol)
and transportor turnover assayed in thefollowing ways.
Dopamine Uptake-At various times, 50-111 aliquots (-25pg of
protein) were removed,added to 0.45-pm cellulosenitrate filters over
a vacuum manifold, and washed with 4-6 ml of cold 0.3 M sucrose
buffered with 10 mM Hepes, pH 7. Filters were dried and counted in
5 ml of scintillation fluid. Standards of each reaction mixture were
added onto filters inscintillation vials (without filtration orwashing)
and counted to measure the apparent specific activity of [“Cldopa-
mine. In measurements of substrate accumulation a t equilibrium,
time points were taken from 0 to 60 min. Net counts incorporated
into ghosts typically reached a maximum value at 10-20 min (Fig.
U).In measuring initial rates of transport, aliquots were sampled
within 100 s, after which the curves became nonlinear (Figure 1B).
To obtain these time points, multiple 0.12-ml reaction mixture ali-
quots were used to sample measurements at 10 and 60 s, 20 and 80 s,
and 40 and 100 s.
Turnover of Labeled Dopamine-After initiating the reaction, 100-
p1 aliquots were quenched at 1-6 min, +ATP, and 6 min, -ATP into
a mixture containing 2 ml of 0.33 M HClO,, 81 mM Tris, pH 8.4, 10.6
mM EDTA, and 3.25 mgof sodium bisulfite (23). Catecholamines
were extracted from the quenched mixture by adsorption onto basic
alumina at pH 8.4 and elution in 0.2 N acetic acid, which proceeds
with about 90% recovery of counts. Catecholamines were separated
by reversed-phase HPLC (Altex C-18 Ultrasphere column) with a
 Dopamine /3-Monooxygenase Activation in Chromaffin Granule Ghosts

Frc. 1. [“C]Dopamine accumulation

into chromaffin granule ghosts verslls
time (O-60 min) is shown in A. Ghosts
were resealed with 2.0 mM K,Fe(CN),
and 4 PM CuSOa, and uptake assays were
performed in the presence of 2.0 mM
ascorbate and 100 pM [Wldopamine, +
6 mM ATP (0) or in its absence (A).
[‘“C]Dopamine accumulation into chro-
maffin granule ghosts versus time (O-100
s) is shown in B. Preparation of ghosts
and assay conditions are as described for
A.

TimeCmin) Time (s)

mobile phase of methanol, 1% acetic acid (15:85, v/v). One-minute (1 dopamine to a final concentration of 10 mM. Solubilization of mem-
ml) fractions were collected and counted in 5 ml of the scintillation brane ghosts by octyl glucoside results in expression of latent activity,
fluid described above. Fractional turnover of dopamine is calculated representing DBM located in correctly sealed vesicles. This activity
as: fractional turnover = counts/minute (NE)/counts/minute (NE) is 60-70 % of the total and is maximally expressed with 0.5% octyl
+ counts/minute (DA). This assay is sensitive to a fractional turnover glucoside. In the comparison of kinetic parameters and deute-
of 0.01, which typically ranged from 0.01 to 0.4. ATP-dependent rium isotope effects with ascorbate versus KIFe(CN)G, initial veloci-
norepinephrine formation (Fig. 2) shows a small lag due to substrate ties were measured at a fixed oxygen concentration (0.2 mM) and
transport prior to hydroxylation, and rate measurements generally varying concentrations of either [2-‘Hz]- or [2-‘H.Jdopamine as de-
utilize data taken after 1 min. For the purposes of this experiment, scribed by Ahn and Klinman (2).
ATP-dependent hydroxylation was calculated as Au = v(+ATP) -
u(-ATP), where v(-ATP) is derived from a single time point at 6 RESULTS
min.
Magnitudes of transport and turnover rates typically vary between Chromaffin granules were lysed and resealed as described
ghost preparations even when prepared under identical conditions. under “Methods.” Although ascorbate is believed to be the
In order to compare data derived from various preparations, the
following correction procedure was used. Turnover rates were meas- redox mediator in uiuo, we found that resealing granule mem-
ured at both 0.2 mM KdFe(CNJG(in) and 2.0 mM K,Fe(CN)a(in) with branes in the presence of ascorbate results in reduced and
2 mM ascorbate in the external medium, conditions common to every irreproducible rates of dopamine transport and that this de-
preparation. The observed rates under these conditions were then structive effect is accentuated by the addition of copper (Table
compared to standard values from a single representative preparation I). Copper was added in accordance with previous findings
of 10.5 nmol/min.mg for 0.2 mM &Fe(CN)G(in) and 9.3 nmol/min that micromolar concentrations of the metal are required for
mg for 2.0 m&t KIFe(CN)&n). The ratio of standard to observed
values was then averaged and yielded a correction factor for each optimal soluble dopamine &monooxygenase (sDBM) activity
preparation by which all other rates (both transport and turnover) (3, 4). Samuni et ul. (24) have described the damaging effects
were multiplied. of ascorbate and Cu2+ on T7 and bacteriophage through
Assay of Solubilized, Membrane-bound Dopamine &Morwoxygenme
(mbD,YM)-The assay of detergent-solubilized mbD@M activity
in- TABLE I
volved the continuous monitoring of oxygen consumption using a Measurements of [‘Qdopamine transport in ghosts prepared under
polarographic oxygen electrode. In the study of the additivity of varying conditions
ascorbate and K9e(CN)e as reductants, ghosts (-80 &g/ml) were
incubated in media containing 50 mM Mes, pH 6.0, 150 rnM KCl, 5 Assay conditions were as described under “Methods,” with 2 mM
mM fumarate, 4 PM CuSO,, 0.5% octyl glucoside, and varying concen- ascorbate, 6 mM MgATP, and 100 FM (‘4C]dopamine in the external
trations of reductant. Reactions were initiated with the addition of medium.
Components of lysis and Transport of [“Cldopamine
resealing buffers Initial velocity Net uptake (6 min)
nmol/min mg nmol/mg
-Reductant
-CL?+, pH 7 28.1 110
4 /JM cu’+, pH 7 6.9 30
5 mM ascorbate
-Cu’+, pH 5.5” 5.3 36
3 PM cu*+, pH 5.5 1.4 13
10 Pi CL?+, pH 5.5 0.5 6.2
2 rnM K,Fe(CN),
-Cd+, pH 7 21.1 90
4 /LM CL?+, pH 7 16.3 68
Time (m1n1 10 FM CL?+, pH 7 11.5 55
FIG. 2. [“ClNorepinephrine production versus time (O-6 4 PM cu2+, DH 5.5 7.6 49
min). Preparation of ghosts and assay conditions were as described “Preparations with ascorbate were conducted at pH 5.5 using
for Fig. 1. nitrogen-bubbled buffers in an effort to minimize ascorbate oxidation.
1488 Dopamine
&Monooxygenme
Kinetic parameters and deuterium isotope effects withdetergent-solubilized mbDpM in the presence of saturating
ascorbate or varying concentrations of K a e ( c N ) ~
DV-b
Reductant VW:
nmol/min.rng
Ascorbate, 5 mM 1030 -C 20
K4Fe(CN)6,0.02 mM 132 2 9
K4Fe(CN)G,0.2 mM 379 2 20
K4Fe(CN)6,2.0 mM 362 k 18
Vmaris the velocity in the limit of saturating substrates and V / K D Athe velocity in the limit of zero substrate
concentration. These values are apparent, having been measured at a single 0 2 concentration, as described under
"Methods."
* "Vmarrratio of V,,, for [l-'H?]- and [l-2H2]dopamine,respectively; "( V/KDA),ratio of V / K D Afor [l-'Hz]- and
Il-'H,ldopamine.
ND, not determined.
formation of oxygen-derived radical species. Since several
hours elapse between the exposure of ghosts to ascorbate/
Cu2+/02 and transportassays, we attribute the observed in-
hibition to a nonspecific damage of membrane protein and
ghost integrity. It is most likely that the magnitude of initial
transport reflects the fitness of all membrane proteins present
in theghost, rather than thecarrier alone, providing an assay
for vesicle viability.
As result
a of the destructive effects of ascorbate,
K.,Fe(CNI6,an alternate electron donor for DPM, wasused as
the internal reductant. As shown in Table I, ferrocyanide has
a minor effect on transport and furthermore appearsto pro-
tect against the inactivation seen by copper in the absence of
reductant. The use of K4Fe(CN), in place of ascorbate in these
studies appearswell justified by the similar magnitudes of the
limiting parameter, VmaX,V/Ko,, and V / K D ~and , isotope
effects on these parameters for sDPM.2 In addition, the use
of K4Fe(CN)6with solubilized mbDPM leads to respectable
rates and deuterium isotope effects analogous to ascorbate,
under conditions of atmospheric oxygen (0.2 mM) and 4 pM
Cu2+(Table 11).
The effect of internal K4Fe(CN16on the magnitudes of the
ApH and A$ generated by ATP hydrolysis in the presence of
C1- is given inTable 111. As predicted from the work of
Johnson et al. (17), dopamine uptakein the presence of
permeant anions is driven primarily by the transmembrane
pH gradient, about -1.0 f 0.2 (120 mV), with a negligible
contribution from the transmembrane potential (910 mV).
Accordingly, the observed ratios of [DA]i,/[DA],ut at steady
state are predicted from the ApH within the range of error.
Although there is a small variation in ApH and [DA]i,/
[DAlOutat varying concentrations of K,Fe(CN),(in) C ascor-
bate(out), these parameters are significantly reduced in the
absence of ferrocyanide. This is similar to the data in Table
I, which indicate a reduction in theinitial rate of transport in
the presence of Cuz+upon omission of internal reductantfrom
the lysis buffer.
Since a direct comparison of the rates of transport and
turnover +. reductant would require a correction for the frac-
tion of viablevesicles (Table 111), the effect of internal
K4Fe(CP& on DPM activity was examined by varying its
concentration in the lysis medium from 0.02 t o 2.0 mM. As
shown in Fig. 3A, K4Pe(CN)&n) clearly activates hydroxyla-
tion 2-3-fold over the base-line rate. Activation appears sat-
urabie with an apparent K,,, cz 0.1 mM. Because the rate of
dopamine hydroxylation has been corrected by the (small)
rate measured in the absence of ATP, these data reflect
L. C. Stewart and J. P. Klinman, manuscript submitted for
publication.
Activation in Chromaffin Granule Ghosts
TABLEI1
VIKDA." D(~ / K D A ) ~
nrnollmin.mg mM
I
1.82 ? 0.09 562 +- 20 3.02 Ifr 0.24
ND' 106 k 18 ND
1.54 0.10 254 +- 13 2.77 If: 0.38
*
1.79 0.18 *
315 39 5.91 -+ 1.06
dopamine turnover occurring exclusively in the vesicle ghost
interior! Upon addition of 2 mM ascorbate to the external
medium, a similar 2-3-fold activation by internal K4Fe(CNj6
is observed(Fig. 3B). Although thedata in Fig. 3 show
experimental scatter, a statistical comparison (cf. Ref. 32) of
the points at 0.02 to 2.0 mM internal ferrocyanide indicates a
very significant increase, with the probability that these Val-
ues are thesame, being t0.01 without external reductant(Fig.
3A) and t0.005with 2.0 mM ascorbate(out) (Fig. 3B). Turning
to a comparison of the data in Fig. 3A to 3B, it can be seen
that external ascorbate increases the rate of hydroxylation to
a similar extent at all internal K4Fe(CN)6 concentrations.
Importantly, DPM activity continues to undergo an -%fold
activation by externalreductant in the region of internal
reductant Concentration which is saturating. This behavior
is not observed in the corresponding transport measure-
ments (Fig. 4), where both increasing concentrations of
K,Fe(CN)&n) and ascorbate(out) decrease initial transport
velocities and, to a small extent, the steady state levels of
uptake (Table 111).The reason for this rate inhibition is not
well understood, but may be related to theslight inhibition in
turnover observed at high internal ferrocyanide in the pres-
ence of 2 mM external ascorbate?
It is known that high copper concentrations (>2 mol of
Cu2+/mo1of enzyme subunit) lead to reversible inhibition of
sDPM activity in the presence of either ascorbate (3) or
ferrocyanide (25). Since copper andK4Fe(CN)6 appear to
interact in a competitive manner (25), we wished to demon-
strate thatactivation by K4Fe(cN)6(in) in ghosts is not caused
by removal of copper inhibition. Therefore, measurements
were made at varying concentrations of added CuS04, 0-10
p ~Fig.. 5, A and B, shows that increasing copper across this
range activates turnover in both the absence and presence of
external reductant, despite a small inhibitory effect on trans-
port (Table I). The greater activation by copper with ascor-
bate(out) may be related to the different pools of enzyme
undergoing reduction, as described later. Significantly, the
data in Fig. 5 indicate that K,Fe(CN), activation is not due
to a competitive interaction with inhibitory copper.
In an effort to examine moreclosely whether internal
K4Fe(CN), is required for activation of turnover by external
ascorbate, the concentration of ascorbate(out) was varied
simultaneously with K4Fe(CN)6(in).Analogous to Fig. 3, we
Reports of reversible ADP activation of purified sDBM and
mbDPM suggest the possibility that ATP hydrolysis might lead to
Dj3M activation (31). However, under our assay conditions, up to 6
mM ADP does not appear to activate purified sDPM from 0.1-10 mM
dopamine.
Since the extent of inhibition by external reductant increases as
the vesicle ghost preparations age, we believe this inhibition results
from free radical damage byactive oxygen species.
 Dopamine P-Monooxygenase Activation in Chromaffin GranuleGhosts 1489
TABLE111
Measurements of 4pH, A$ and [DA]i,,/[DA]oucratios in ghosts prepared with vatying amounts of Kze(CN),
Ghosts were prepared as described under “Methods” with5 mM Hepes, pH 7 , 5 mM fumarate, 150 mM KC], 20
pg/mI catalase, 4 p~ CuS04.The number of determinations aregiven in parentheses. Measurementswere conducted
using the assay procedures described under “Methods”in the presence of 6 mM ATP. For ApH and A$, dopamine
was omitted from the external buffer.

Predicted Observed
mV
-Reductant(out)
0 mM -0.61 f 0.32 (2) 2.9 f 1.6 (2) 18.5 71 -+ 8 (2)
0.02 mM -0.96 (1) 7.0 (1) 108 195 f 60 (2)
0.2 mM -0.90 f 0.15 (2) 10.8 f 3.9 (2) 95 206 k 46 (5)
2.0 mM -1.07 f 0.26 (3) 4.9 f 4.1 (3) 166 153 -+ 62 (6)
Averagesb -1.00 6.8 129 180
+2 mM ascorbate(out)
0 mM -0.66 (1) 14.0 (1) 36 69 _t 55 (3)
0.02 mM -0.67 (1) 6.5 (1) 26 136 -+ 5 (3)
0.2 mM -0.92 -+ 0.08 (2) 11.1 k 3.0 (2) 105 181 f 104 (8)
2.0 mM -1.18 f 0.35 (2) 12.8 k 5.9 (2) 371 129 f 51 (10)
Averagesb -0.97 10.8 131 150
’[DA]i. is calculated using the average vesicular volume of 9.31 pl/mg (n = 12). Predicted ratios of [DAIi,/
[DA], are calculated accordingto Johnson et al. (17):

Averages of measurements using ghosts prepared with

0.02, 0.2, and 2.0 mM K9e(CN)6.

I I I

FIG. 3. ATP-dependent rate of hydroxylation in resealed ghosts

uersus internal K4Fe(CN)G concentration is shown in A . Ghosts were
resealed with 4 ~ L MCuS04and varying concentrationsof K,Fe(CN),,
and assayswere performed inthe absence of reductant in the external
medium. ATP-dependent rate of hydroxylationinresealedghosts
versus internal K4Fe(Cb06concentration is shown in B . Ghosts were
resealed in 4 ~ L MCuS04 and varying concentrations of K4Fe(CN),
OO
- 5
Ascorbote ( o u t ) ( m M )
8

IO

FIG. 4. Initial rateof [“Cldopamine transport into resealed

ghosts versus external ascorbate concentration. Ghostswere
resealed with 4 p~ CuSO, and 0.05 (O),0.2 (V),and 2.0 mM (m)
and assays were performed with 2.0 mM ascorbate in the external K4Fe(CN)6.Assays were performed with 6 mM ATP in the external
medium. medium.

observe that ascorbate(out) in the range of 0.05-1 mM acti-

vates turnover at all concentrations of K4Fe(CN)Jin), witha
maximal effect at 0.2 mM ascorbate(out) (data not shown).
We also observe an inhibition by external ascorbate at con-
centrations greater than 2 mM. Because of the complication
of inhibition by 4 ~ L MCu2+ (Table I), externalascorbate
activation is difficult to demonstrate in thecomplete absence
of internal K4Fe(CN)+Although omission of copper from the
lysis media leads to somewhatreduced turnover rates (Fig. 5),
transport rates remainhigh (Table I). Therefore, we examined
D@Mturnover in the absenceof both K,Fe(CN)6 and copper,
observing the same 3-fold activation of enzyme by external
ascorbate as in the presence of internal reductant (Fig. 6).
As shown (Figs. 3 and 61, the maximal degree of enzyme
activation by ascorbate appears constant as the internal con-
centration of ferrocyanide is varied. This behavior is not
predicted by a model in which K4Fe(CN)&n) is the sole
reductant for DPM. If ascorbate(out) were behaving only as
an electron source for the reduction of internal ferrocyanide,
activation by ascorbate(out) would not be observed either in
the absence of K,Fe(CN),(in) or a t saturating levels of inter-
nal reductant. Instead, K,Fe(CN),(in) does not affect activa-
tion of dopamine turnover by ascorbate(out), andconversely,
ascorbate(out) does not affect activation by KBe(CN)e(in).
These data lead us to suggest that ascorbate(out) reduces
DPM from the external membraneface.
Alternatively, ascorbatemightactasanelectrondonor
1490 Dopamine
P-Monooxygenme ActivaItion i n Chromaffin Granule Ghosts
I I I I
IO A 0

Cu+‘ ( p M ) Cu+‘ (pM)

FIG. 5. ATP-dependent rate of hydroxylation in resealed
ghosts uersus internal CuS04.Ghosts were resealed with 0.02 (A),
0.05 (e),0.2 (V),and 2.0 mM (W) K,Fe(CN),,andassayswere
performed in the absence ( A ) or presence ( B ) of 2 mM ascorbate in
the externalmedium.
.-C
\
E p
:
0‘
I I I I I I
- 1.0 2.0 1.0 2.0

- K,Fe(CN)e(in)(mM) K 4 Fe (CN)6(in)
(mM)
FIG. 7. ATP-dependent rate of hydroxylation in resealed
ghosts uersus internal K4Fe(CN)B concentration. Ghosts were
resealed in 4 p~ CuSO4and varying concentrations of K4Fe(CN)6
and assays were performed as a function of &Fe(CN)6in the external
medium: A, 0 K4Fe(CN)6(out); B, 0.05 mM IdFe(CN)&(out); c, 0.2
mM K4Fe(CN)6(out); and D,2.0 mM K4Fe(CN)6(out).
-
As noted under “Methods,” lysis of chromaffin vesicles
reduces endogenous catecholamines from -0.5 M to 5 mM. A
corresponding decrease in endogenous ascorbate is expected
0- I I to lead to -0.1 mM in resealed ghosts. It is therefore conceiv-
0 5 IO
able that activation by both external ascorbate and ferrocya-
Ascorbate(out) (mM1
nide is the result of electron transfer to a small endogenous
FIG. 6 . ATP-dependent rate of hydroxylation in resealed
ghosts uersus external ascorbate as reductant. Ghostswere pool of ascorbate. To explore this possibility, the extent of
resealed in the absenceof any added CuSO,or reductant. enzyme activation by ascorbate was examined at varying
levels of K4Fe(CN), using detergent-solubilized mbDPM (Fig.
from the membrane interior. Although results of previous 8). Under these conditions, enzyme is subject to competitive
investigations eliminate the possibility that ascorbate is trans- activation by K4Fe(CN)fi andascorbate, such that the acti-
ported in its reduced form (10, 26, 27) membrane-permeable vation by ascorbate decreases with increasing ferrocyanide
dehydroascorbate might diffuse into the ghost and become concentrations. Thus, if the concentration of internal ascor-
reduced by way of the hypothesized transmembrane electron bate falls to zero in theabsence of external reductants but is
transfer cytochrome. To clarify this point, K4Fe(CN), was maintained at itsinitial level, -0.1 mM, in their presence, we
used in place of ascorbate as theexternal electron donor, since would expect activation by external reductants to decrease
it ishighly unlikely that K4Fe(CN), or K3Fe(CN), could from 200-300% at 0.02 mM K,Fe(CN),(in) to <20% at 0.2
diffuse through the ghost membrane. In any case, this occur- mM K,Fe(cN)&n). We therefore conclude that reduction of
rence would only elevate the internalferrocyanide concentra- a small pool of endogenous dehydroascorbate by external
tion, yielding no further increase in turnover ratesat saturat- reducant is inconsistent with the patternsof activation shown
ing K4Fe(CN)6(in).Fig. 7 shows that, like ascorbate(out), in Figs. 3, 6, and 7.
K,Fe(CN)fi(out) activates hydroxylation 2-3-fold at all con- Overall, the datapresented herein imply that reductants on
centrations of K4Fe(CN),(in).Analogous to Fig. 3, a statistical opposite sides of the membrane activate DPM independently.
analysis (32) of the datain Fig. 7C indicates alow probability Possible causes of this behavior include more efficient reduc-
(C0.025) that the values at 0.05 and 2.0 mM internal ferro- tion of mbDPM from the external face of the membrane or
cyanide are the same. K,Fe(CN),(out) is also observed to the presence of more than one form of enzyme within vesicle
activateturnoverin the absence of internalreductant, ghosts. With regard to the latterpossibility, it is conceivable
whereas in corresponding transport measurements no in- that K4Fe(CN),(in) reduces residual soluble enzyme, whereas
crease in rate occurs (data not shown). Thus, the observed electron donors on the ghost exterior reduce mbDBM. SDS-
activation by either ferrocyanide or ascorbate as external polyacrylamide gel electrophoresis was performed on chro-
electron donor cannot be rationalized by their interaction maffin granule ghosts to estimate the fraction of sDPM.
with DPM from the interior membrane face. Several cycles of freezing and thawing in H 2 0 released a M,
 Dopamine P-MonooxygenuseActivation in Chromaffin Granule Ghosts 1491
experiment using ascorbate-loaded phospholipid vesicles that
had been reconstituted with purified cytochrome b561, showing
that Fe3+cytochrome c could be reduced rapidly (>9.5 s-').
Based on these results, a model for DPM reduction has
been proposed in which a soluble redox mediator, presumed
to be ascorbate in vivo, interacts directly with mbDBM and
sDPM within the chromaffin granule interior. Support of
hydroxylation by an external reductantoccurs through trans-
membrane electron transport via cytochrome b561 to the in-
ternal mediator, which must be regenerated with each turn-
over. Although the experimentally measured electron flow has
been in theopposite direction from that hypothesized in uivo,
the dataprovide a reasonable mechanism for the regeneration
of compartmentalized ascorbate. The model also explains the
fact that no enzymatic oxidoreductase activity has yet been
clearly demonstrated in chromaffin granules.
A fundamental feature of the above model is the require-
ment for intravesicular reductants which support DPM-cata-
lyzed hydroxylation. In earlier experiments, Grouselle and
' OO .Ascorbate
I. 0 I 2.o
(mM1 1 Phillips (13) demonstrated stimulationof an ATP-dependent
turnover of tyramine by external reductants, but failed to
observe a role for internally localized K4Fe(CNI6.This result
FIG. 8. Activity of detergent-solubilizedmbD@Hversus re- contrasts with the data presented herein, where
ductant. Activities were measured by monitoring oxygen uptake at
varying concentrationsof ascorbate in the presence of 0 (O),0.02 (A), K4Fe(CN)&n) clearly activates DBM 2-3-fold at saturating
0.1 (m), and 0.2 m M (V)K4Fe(CN)6. conditions of reductant (Figs. 3 and 7). In all studies of
ferrocyanide activation, we have observed the initial rate of
75,000 polypeptide into the supernatant. No further extrac- product formation to be linear after a short lag period. As
tion of protein was observed upon incubating ghost mem- seen in Fig. 2, product synthesis reached 40% in 6 min,
branes in 250 mM NaC1, 5 mM EDTA, whereas the bulk of corresponding to 80 nmol of norepinephrine/mg of protein.
the M, 75,000 polypeptide extractedintoSDS(datanot Since the vesicle volumeis small relative to thetotal (-0.5%),
shown). These results suggest that ghost preparations may the amount of internal K4Fe(CN)6(-20 nmol/mg of protein)
indeed be contaminated by a DPM form which is not tightly is insufficient to account for the product formed and should
membrane-associated, although we cannot be certain to what not yield linear rates of turnover. However, as noted under
extent this putative sDPM originates from a small pool of "Methods," in the absence of substrate uptake, our vesicle
intact chromaffin granules which fail to lyse in the course of preparations already contain -5 mM catecholamines. Fur-
the ghost preparation verus residual sDpM which does not thermore, Stewart and Klinman' have shown that catechol-
separate from the lysed membranes before resealing occurs. amines rapidly reduce 2 mM K,Fe(CN), leading to initial
Quantitation of hydroxylase activity in the sDpM uersus rates with sDPM which are comparable to those measured
mbDPM pools by gel densitometer scanning through the M, with 2 mM K.,Fe(CN),. Inallinstances the reduction of
75,000 band indicated 30% sDPM, whereas activity measure- enzyme by catecholamine plus ferricyanide is much faster
ments in 1%octyl glucoside lead to a value of 15%. Since than catecholamine alone, implicating ferrocyanide as the
chromogranin A may co-migrate with sDPM (28), the activity enzymatic mediator. Thus we conclude that resealed vesicles
estimates areconsidered more reliable and lead to anestimate contain sufficient catecholamine to maintain ferrocyanide in
of sDPM to solubilized mbDBMof0.15:0.85. Comparative the reduced form. Overall, there aresufficient reducing equiv-
studies of octyl glucoside solubilized mbDPM versus mbDPM alents in our ghost preparations to support complete conver-
in vesicle ghosts generally indicate an increase in latent sion of substrate to product and hence, no participation by an
activity: so that following correction for differing activities external reductant need be invoked under our experimental
in solubilized versus vesicular mbDpM we estimate conditions.
sD@M:mbDpMas 0.3~0.7.Although these considerations sup- However, our results clearly implicate a role for external
port a role for sDPM, we note that theestimated contribution electron donors in turnover, demonstrated by an activation
of contaminating sDBM is -30% and as such is unlikely to of enzyme which is independent of the presence of internal
be the exclusive source of internal K4Fe(CN)6activation. reductant (Figs. 3, 6, and 7). Since the same pattern is
observed with either external K,Fe(CN), or ascorbate, it is
DISCUSSION not caused by dehydroascorbate uptake and subsequent re-
Evidence to date strongly suggests that chromaffin granules duction. Furthermore, the behavior of competing reductants
have the capability to transfer electrons from one side of the for DPM (Fig. 8) makes it unlikely that low levels of oxidized,
membrane tothe other, mediated by a membrane-bound endogenous ascorbate are undergoing reduction by the exter-
protein which can interact directly with soluble redox com- nal pool of reductant. We conclude that reductants on oppo-
ponents on both sides. Using chromaffin granule ghosts pre- site sides of the membrane act independently of one another
loaded with 100 mM ascorbate, pH 7, Njus et al. (10) and and contribute additively to the total rate. This behavior is
Harnadek et al. (12) demonstrated the reduction of both not predicted by a model involving transmembrane electron
Fe(CN)z3 (-+Fe(CN)g4) and Fe3' cytochrome c (-+Fe2+cyto- flow to a soluble internal mediator as the exclusive source of
chrome c), concomitant with the appearance of a transmem- reducing equivalents. It is proposed that mbDPM may accept
brane potential A$ > 0. Strivastava et al. (11)repeated this electrons from the membrane exterior without the participa-
tion of an internal mediator. This behavior might occur either
N. Ahn, B. Huyghe, and J. P. Klinman, unpublished results. by direct interaction of the external reductant with the en-
1492 Dopamine
P-Monooxygenase
Activation
zyme or, alternatively, through the intervention of another
mediator, e.g. the membrane-localized cytochrome bSG1.
Finally, we note that the rate of ATP-dependent hydrox-
ylation (1-2 nmol/min. mg) is non-zero in the compIete ab-
sence of added reductant. Eitherdopamine added as substrate
or endogenous norepinephrine and epinephrine could function
as electron donors. We were unable to detect labeled dopa-
quinone by HPLC in [I4C]dopamineturnover assay eliminat-
ing a major contribution by dopamine. Furthermore, Grouselle
and Phillips (13) observed atime-dependent oxidation of
endogenous catecholamine concomitant with the hydroxyla-
tion of tyramine. Using [I4C]tyramine as substrate, we also
observe base-line rates of turnover in the absence of added
reductant (data notshown). Therefore, it is likely that endog-
enous catecholamines representathird pool of reductant
capable of supporting hydroxylation. However, it is unlikeIy
that this pool contributes significantly to turnover rates in
the presence of added internal or external reductants, since
catecholamines are very poor reductants for DPM.*
Given the demonstrationof transmembrane electron trans-
fer with ascorbate loadedvesicles (10-12) anda role for
internal reductant in DPM activity (this study), it appears
reasonable that vesicular ascorbate should function as an in
uiuo reductant. In addition, the data reported herein indicate
a pathway for enzyme reduction from the exterior face of the
membrane. This important resultimplies either that mbDPM
is reduced more rapidly from the outer than the inner face of
2AL+2Ht 2AH'
ADP
mb DBM
2 AH' 2A"+ 2H'
SCHEME I. Proposed model for DBM reduction in intact ves-
icles, where AH- and A" are the ionized forms of ascorbate
and semidehydryascorbate, respectively. As shown, sDpM re-
duction occurs by intravesicular ascorbate, which i s recycled from
cytoplasmic ascorbate. In addition, mbDBM can undergo reduction
from the exterior face of the vesicle. Although direct reduction of
mbDOM is drawn for illustrative purposes, this reduction may be
mediated by a membrane-bound protein such as cytochrome b5G1.
in Chromaffin
Granule
Ghosts
the vesicle membrane, or that different forms of DPM are
kinetically apparent in this system. With regard to the latter
possibility, gel electrophoresis and activity analyses of ghost
preparations do support the presence of -30% sDPM, sug-
gesting that the observed kinetic patterns may reflect differ-
ent mechanisms for the reduction of sDPM and mbDPM. We
therefore propose a model for DPM in intact vesicles (Scheme
I). As illustrated, cytoplasmic ascorbic acid (-1 mM (29))
serves as the ultimate reductant pool for both sDPM and
mbDPM (30). Inthe case of sDpM, reduction occurs by
intravesicular ascorbate, which is subsequently recycled by
cytoplasmic ascorbate. Since the active sites of both sDPM
and mbDPMface the vesicle interior, mbDPMmay also
undergo some reduction by this mode. In addition, however,
mbDPM can undergo efficient reduction from the exterior
face of the vesicle. As noted above and in the legend to Scheme
I, the precise mechanism of this newmode of reduction is
currently unclear. It will be of particular interest to ascertain
whether mbDpM can interact directly with ascorbate in the
absence of a mediator, since such a mechanism would imply
a long-range intramolecular electron transfer from the site of
enzyme reduction (cytosolic) to that of substrate hydroxyla-
tion (vesicular). In closing, we also note that the demonstra-
tion of two viable modes for reduction of DPM in vivo may
provide a functional explanation for the presence of multiple
forms of DPM in chromaffin vesicles.
REFERENCES
1. Miller, S. M., and Klinman, J. P.(1983)Biochemistry 22,3091-3096
2. Ahn, N., and Klinman, J. P. (1983)Biochemist 22,3096-3106
3. Klinman, J. P., Krueger, M., Brenner, M., and gknondson,D. E. (1984)J .
Bwl. Chem. 259,3399-3402
4. Ash, D. E., Papadopoulus, N. J., Colombo, G., and Villafranca, J. J. (1984)
J. Biol. Chem. 259,3395-3398
5. Miller, S. M., and Klinman, J. P. (1985)Biochemistry 24,2114-2127
6. Fitzpatrick, P. F., Flory, D. R., Jr., and Villafranca, J. J. (1985)Biochemistry
24, 2108-2114
7 . Levin, E. Y., Levenberg, B., and Kaufman, S. (1960)J. Biol. Chem. 235,
-"-" -"-"
vnQn-vnw
8. Ingebretsen, 0. C., Terland, O., and Flatmark, T. (1980)Biochim. Biophys.
Acta 628,182-189
9. Terland, 0.. and Flatmark, T. (1975)FEBS Lett. 59,52-56
10. NJUS,D., Knoth, J., Cook, C., and Kelly, P. M. (1983)J. Biol. Chern. 258,
27-21)
11. St&a&va, M., Duong, L. T., and Fleming, P. J. (1984)J. Biol. Chem.
259,8072-8075
12. Harnadek, G. J., Ries, E. A., and Njus, D. (1985)Biochemisty 24, 2640-
?Add
13. Gri&ke, M., and Phillips, J. H.(1982)Biochem. J. 202, 759-770
14. Bradford, M. M. (1976)Anal. Biochem. 72,248-254
15. Pollard, H.B., Menard, R.,,Brandt,H. A., Pazoles, C. J., Creutz, C. E., and
Ramu, A. (1978)Anal. Blochem. 86,761-763
16. Laemmli, U. K. (1970)Nature 227,680-685
17. Johnson, R. G., Pfister, D., Carty, S. E., and Scarpa, A. (1979)J. Biol.
Ch.mn - - -,in~m-10972
-. .-. .. 254. . .. - ... -
18. Phillips J. H. (1974)Biochem J. 144,311-318
19. Bartletc S. F., and Smith, A. D. (1974)Methods Enzymol. 31,379-389
20. Helle. K.. Flatmark. T.. Serck-Hansen,. G... andLonnina.
-. S. (1971)Bimhirn.
Bidphys. Acta 226,l-8
21. Rottenber H (1979)Methods Enzymol. 5 5 , 547-572
22. Johnson #. G:, and Scarpa, A. (1976)J. Gen. Pbysiol. 68 601-631
23. Weil-Milherbe, H. (1971)Methods Biochem. Anal. P. Suipl., 119-152
24. Samuni, A., Aronovitch, J., Godinger, D., Chevion, M., and Czapski, G.
(1983)Eur. J. Biochem. 137,119-124
25. Rosenherg, R. C., Gimble, J. C., and Lovenberg, W. (1980) Biochim.
Bio h s Acta 613,62-72
26. Tirreif, f.,'and Westhead, E.(1979)Neuroscience 4,181-186
27. Levine, M.,Morita, K.. and Pollard, H. B. (1985)J. Biof. Chem. 260,
12942-12947
28. Winkler, H.(1976)Neuroscience 1,65-80
29. Daniels, A. J., Dean, G., Viveros, 0. H., and Diliberto, E. J., Jr. (1982)
Science 216, 737-779
30. Levine, M., Morita, K., Heldman, E., and Pollard, H. B. (1985)J. Bid.
Chem. 260,15598-15603
31. Ohuchi, T., Tachikawa, E., Morita, K., Ishimura, Y., and Oka, M. (1982)
Adu. Biosci. 36,201-208
32. Snedecor, G., and Cochran, W. (1980)Statistical Methods, 7th Ed,, pp. 83-
93,Iowa University Press, Ames, IA