The Reduction of Membrane-bound Dopamine (cid:98) -Monooxygenase in Resealed Chromaffin Granule Ghosts IS INTRAGRANULAR ASCORBIC ACID A MEDIATOR FOR EXTRAGRANULAR REDUCING EQUIVALENTS?*

The role of internal and external reductants in the dopamine (cid:98) -monooxygenase (D (cid:98) M)-catalyzed conversion of dopamine to norepinephrine has been investi- gated in resealed chromaffin granule ghosts. The rate of norepinephrine production was not affected by the ex- clusion of internal ascorbate. The omission of ascorbate from the external medium drastically reduced the nor- epinephrine production without affecting the net rate of dopamine uptake. In the presence of the external reduc- tant, the internal ascorbate levels were constant throughout the incubation period. The rate of norepi- nephrine production was not affected when ghosts were resealed to contain the D (cid:98) M reduction site inhibitor, imino- D -glucoascorbate. Ghosts incubated with external imino- D -glucoascorbate reduced the norepinephrine production. The weak D (cid:98) M reductant, 6-amino- L -ascor-bic acid, was found to be a good external reductant for granule ghosts. The outcome of the above experiments was not altered when dopamine was replaced with the reductively inactive D (cid:98) M substrate, tyramine. These results and the known topology of membrane-bound D (cid:98) M disfavor the direct reduction of the enzyme by the ex- ternal reductant. Our observations are consistent with the hypothesis that external ascorbate is the sole source of reducing equivalents for D (cid:98) M monooxygenation and 2 and (powder) Pharmacia Inc., and catalase Boehringer Mannheim. Protein assay rea-gent and bovine serum albumin were from Bio-Rad Laboratories. Imi- no- D -glucoascorbic acid (IGA) was synthesized according to literature procedures (34) and characterized as reported previously (36). 6-Amino- L -ascorbic acid (6-AAscH 2 ) and 2-amino- L -ascorbic acid (2-AAscH 2 ) were synthesized in our laboratory (35, 36) and were kindly provided by S. Dharmasena. All other chemicals and solvents were of the highest grade obtainable. Tissue pellets were homogenized using glass-Teflon Potter-Elvejhem homogenizers. Low speed centrifugations in the granule isolation procedure were performed using IEC 870 and 877 rotors in a refrigerated IEC B-20A centrifuge, while the high speed centrifugation step and the ghost preparations were performed using Sorvall A-641 and T-1270 rotors, respectively, in a refrigerated Sorvall OTD-75 ultracentrifuge. HPLC-EC analyses were performed using a Kontron Instruments Model 420 HPLC pump and an ESA Coulochem Model 5100A electrochemical detector coupled to a Spectra Physics Chromjet integrator. UV-visible spectroscopic measurements were carried out using an HP-8452A diode array spectrophotometer. Isolation of soluble D (cid:98) M and membrane-bound D (cid:98) M were accomplished using a Pharmacia Mono Q (HR 10/10) column controlled by a Pharmacia FPLC system. Initial rates of steady state oxygen consumption by the D (cid:98) M reaction were measured using a Yellow Springs Model YSI 5300 polarographic oxygen monitor.

Dopamine ␤-monooxygenase (D␤M; 1 EC 1.14.17.1) catalyzes the conversion of dopamine (DA) to the neurotransmitter, norepinephrine (NE), within the neurosecretory vesicles of the adrenal medullae and the large dense-cored synaptic vesicles of the sympathetic nervous system. D␤M exists in both soluble and membrane-bound forms in these tissues (1)(2)(3). The reducing equivalents required for in vivo D␤M monooxygenation is believed to be provided by ascorbic acid (AscH 2 ) which is present in high concentrations (20 mM) within the D␤M-containing neurosecretory vesicles (4,5). Despite the fact that AscH 2 can function as an overall two-electron reductant, it has been shown to function as a single-electron reductant producing semidehydroascorbic acid as the immediate product during the D␤M turnover (6 -8). Semidehydroascorbic acid is a relatively nonreactive radical which spontaneously disproportionates to produce AscH 2 and the oxidized form, dehydroascorbic acid, under normal conditions. Although dehydroascorbic acid is reduced back to AscH 2 by a glutathione-dependent enzyme system in plants (9,10), no mechanism is available to regenerate AscH 2 from dehydroascorbic acid in neuroendocrine cells.
Numerous detailed studies have demonstrated that neuroendocrine secretory vesicles of the adrenal medullae or other endocrine glands do not transfer AscH 2 across the vesicle membrane at a detectable rate (11)(12)(13)(14). Based on these observations and the fact that the concentration of AscH 2 in the catecholamine secretory vesicles is approximately 20 mM and the catecholamine concentration is 500 mM (4,5), a transmembrane electron carrier which shuttles electrons from the cytosolic pool of AscH 2 (2 mM (15)) to the intercellular matrix to regenerate AscH 2 from semidehydroascorbic acid has been proposed. The chromaffin granule-transmembrane hemoprotein, cytochrome b 561 (cyt b 561 ) has been identified as the most likely candidate to serve as an electron carrier, since it has a favorable midpoint potential of ϩ140 mV (16 -18). According to this model, semidehydroascorbic acid generated by the D␤M monooxygenation reaction within the granule matrix is efficiently reduced to AscH 2 by the reduced cyt b 561 , preventing the radical disproportionation and thereby maintaining a constant pool of AscH 2 within the granule matrix (19 -22). Oxidized cyt b 561 is subsequently reduced by cytoplasmic AscH 2 , and depletion of cytoplasmic AscH 2 is prevented by a mitochondrial NADH:semidehydroascorbic acid oxidoreductase (19 -22).
The dynamics and various mechanistic aspects of NE biosynthesis has been mostly studied by using resealed chromaffin granule ghosts as a model system for adrenergic neurotransmitter storage vesicles (19 -27). Several granule membrane proteins, including membrane-bound D␤M, ATPase, cyt b 561 , and monoamine transporter, have been well characterized, especially in relation to catecholamine neurotransmitter biosynthesis. The transmembrane proton-translocating ATPase has been shown to be responsible for maintaining a low intravesicular pH of 5.5 relative to the neutral cytosolic pH (28,29), which is proposed to be mandatory for the proper functioning of chromaffin granule ghosts. In accordance with the chemiosmotic hypothesis (30), the vectorial transport of protons across the chromaffin granule membrane by proton-translocating ATPase creates an electrochemical (⌬) and a pH gradient (⌬pH) across the granule membrane. It has been hypothesized that the electrochemical and proton gradients across the granule membrane are important for the cyt b 561 -mediated electron transport process, since the pH gradient will shift the reduction potential of the AscH 2 /semidehydroascorbic acid couple from ϩ70 mV (outside (pH 7.0)) to ϩ220 mV (inside (pH 5.5)), facil-itating the electron flux from cytosol to the interior of the granule ghosts through cyt b 561 (19 -22, 31, 32).
However, in contrast to the above generally accepted mechanism for the reduction of D␤M in catecholamine secretory vesicles, several recent reports appear to indicate that the external AscH 2 is the kinetically preferred reductant for the membrane-bound D␤M in resealed chromaffin granule ghosts (14,33). In the present study, we have examined the role of internal and external electron donors in the D␤M-catalyzed conversion of DA to NE and the net rate of DA uptake in resealed chromaffin granule ghosts using three new AscH 2 analogs, 6-amino-L-ascorbic acid (6-AAscH 2 ), 2-amino-L-ascorbic acid (2-AAscH 2 ), and imino-D-glucoascorbic acid (IGA) as probes (Scheme 1). The results presented in this report suggest that the external reductant, AscH 2 , is the sole source of the reducing equivalents for the membrane-bound D␤M-catalyzed monooxygenation in resealed chromaffin granule ghosts. More importantly, our results also suggest that internal soluble AscH 2 (or DA) may not directly reduce or mediate the reduction of membrane-bound D␤M in chromaffin granule ghosts. The implications of these findings on the mechanism of the reduction of membrane-bound and soluble D␤M in relation to in vivo NE biosynthesis is discussed.

Materials
L-Norepinephrine hydrochloride, DL-epinephrine bitartrate, dopamine hydrochloride, Mg-ATP, HEPES, Trizma (Tris base), ascorbate oxidase, and polyoxyethylene 10 tridecyl ether (Emulphogene) were obtained from Sigma. Ascorbic acid (AscH 2 ) and sodium fumarate were from Aldrich. Ficoll (powder) was obtained from Pharmacia Biotech Inc., and catalase was from Boehringer Mannheim. Protein assay reagent and bovine serum albumin were from Bio-Rad Laboratories. Imino-D-glucoascorbic acid (IGA) was synthesized according to literature procedures (34) and characterized as reported previously (36). 6-Amino-L-ascorbic acid (6-AAscH 2 ) and 2-amino-L-ascorbic acid (2-AAscH 2 ) were synthesized in our laboratory (35,36) and were kindly provided by S. Dharmasena. All other chemicals and solvents were of the highest grade obtainable. Tissue pellets were homogenized using glass-Teflon Potter-Elvejhem homogenizers. Low speed centrifugations in the granule isolation procedure were performed using IEC 870 and 877 rotors in a refrigerated IEC B-20A centrifuge, while the high speed centrifugation step and the ghost preparations were performed using Sorvall A-641 and T-1270 rotors, respectively, in a refrigerated Sorvall OTD-75 ultracentrifuge. HPLC-EC analyses were performed using a Kontron Instruments Model 420 HPLC pump and an ESA Coulochem Model 5100A electrochemical detector coupled to a Spectra Physics Chromjet integrator. UV-visible spectroscopic measurements were carried out using an HP-8452A diode array spectrophotometer. Isolation of soluble D␤M and membrane-bound D␤M were accomplished using a Pharmacia Mono Q (HR 10/10) column controlled by a Pharmacia FPLC system. Initial rates of steady state oxygen consumption by the D␤M reaction were measured using a Yellow Springs Model YSI 5300 polarographic oxygen monitor.

Methods
HPLC-EC Analyses-The catecholamines, NE, E, and DA, and AscH 2 levels of chromaffin granule ghost incubates were quantitated using ion-paired HPLC with electrochemical detection. Perchloric acid extracts of ghosts were separated by a C 18 reversed phase column (ESA, HR-80) pre-equilibrated with a mobile phase composed of 50 mM NaH 2 PO 4 , 0.22 mM sodium EDTA, 1.24 mM octanesulfonic acid, pH 2.6, with 6% CH 3 OH at a flow rate of 1 ml/min. All four analytes were detected by electrochemical oxidation at 300 mV. When tyramine was the substrate, octopamine was separated using a mobile phase composed of 75 mM NaH 2 PO 4 , 10 mM trichloroacetic acid, 0.52 mM sodium dodecyl sulfate, pH 2.8, with 14% CH 3 CN. The octopamine levels were detected by electrochemical oxidation at 800 mV. The E and NE levels of these incubates were quantitated under standard conditions by a second injection of the same sample. Sample peak areas were quantitated by comparison to standard curves which were linear over the range of sample sizes encountered.
Protein Assays-Protein contents of the ghost preparations were determined by the method of Bradford (37) using the Bio-Rad protein assay with bovine serum albumin as the standard.
Tissue Preparations-Chromaffin granules were prepared from bovine adrenal medullae obtained within 1 h postmortem from a local meat packing plant as described previously (26,27) using the original methods of Kirshner (38) as modified by Njus and Radda (39), except that the granules were purified by a discontinuous sucrose density gradient at 58,700 ϫ g for 90 min at 4°C (40). The granules were homogenized in 0.2 M Tris phosphate, pH 7.0, containing 100 g/ml catalase and lysed by the addition of 1/8 volume of a glycerol solution (glycerol, 0.2 M Tris phosphate, pH 7.0, 3:7 (v/v)). The lysate was stored in 1.5-ml aliquots at Ϫ78°C. Chromaffin granule ghost membranes were isolated from the stores on the day of each experiment by diluting with water containing 20 mM AscH 2 and 100 g/ml catalase, leaving for 10 min at 4°C and centrifugation at 36,000 ϫ g for 20 min at 4°C. The pellet was homogenized and resuspended in 2 ml of a solution containing 20 mM Tris phosphate, 100 mM KCl, 150 mM sucrose, 10 mM sodium fumarate, 100 g/ml catalase, and the desired amounts of AscH 2 , IGA, 6-AAscH 2 , 2-AAscH 2 , or ascorbate oxidase, as indicated in the figure legends, and the pH of the final solution was adjusted to 7.0. The ghost membranes were allowed to reseal by incubation for 20 min at room temperature, diluted to 5 ml with the same solution placed inside (but never containing AscH 2 , ascorbate derivative, or catalase), and layered over 2.0 ml of 15% Ficoll, 0.3 M sucrose, 10 mM HEPES, pH 7.0, and 3.0 ml of 0.4 M sucrose, 10 mM HEPES, pH 7.0, and centrifuged for 30 min at 90,000 ϫ g at 4°C. The resealed ghosts which separate as a band at the 0.4 M sucrose-HEPES/Ficoll interface was drawn out using a disposable syringe, diluted with 5 ml of 0.3 M sucrose, 10 mM HEPES, pH 7.0, homogenized, and pelleted by centrifugation at 36,000 ϫ g for 20 min at 4°C. The supernatant was removed, and the pellet was gently washed with the same buffer, homogenized, and resuspended in 1.0 ml of 0.3 M sucrose, 10 mM HEPES, pH 7.0, and stored at 0°C. Samples were taken for protein analysis, and catalase was added at 100 g/ml to the rest of the sample.
Uptake and Conversion Experiments-To aliquots of ghosts prepared as described above were added 5 mM Mg-ATP, 5 mM MgSO 4 , 100 g/ml catalase, and various concentrations of AscH 2 , ascorbate derivative, or ascorbate oxidase (as indicated in the figure legends), and diluted with 0.3 M sucrose, 10 mM HEPES, pH 7.0, to a final volume of 2.4 ml. This mixture was incubated for 10 min at 30°C, and the reactions were initiated by adding DA to a final concentration of 200 M and incubated further for 20 -25 min at 30°C. At 5-min time intervals, 400-l aliquots were withdrawn and diluted into 5.0 ml of ice-cold 0.4 M sucrose, 10 mM HEPES, pH 7.0 and stored at 0°C until the incubation period was completed. These samples were then centrifuged at 36,000 ϫ g for 25 min at 4°C, the supernatants were removed, the pellets were washed gently three times with 0.4 M sucrose, 10 mM HEPES, pH 7.0, and the tubes were swabbed dry. Then, 500 l of 0.1 M HClO 4 was added, the pellets were homogenized, and the extraction was allowed to proceed for 20 min at room temperature. After low speed centrifugation to remove coagulated protein, 25 l of the acidic extracts were analyzed for catecholamines and AscH 2 , by reversed phase HPLC-EC. The same protocol was followed for the experiments carried out using tyramine as the substrate, except that tyramine was used at a final concentration of 30 M, and the aliquots were withdrawn every 4 min for 20 min.
Isolation of Membrane-bound D␤M-Membrane-bound D␤M was isolated from chromaffin granule membranes which were stored at Ϫ78°C in 20 mM potassium phosphate buffer, pH 7.0, containing 100 g/ml catalase. The frozen membranes were thawed and centrifuged at 48,500 SCHEME 1.

Reduction of D␤M in Chromaffin Granule Ghosts
ϫ g, the pellet obtained was homogenized in 20 mM potassium phosphate buffer, pH 7.0, and recentrifuged to obtain a pellet free of contaminating soluble D␤M. This pellet was homogenized in 20 mM potassium phosphate buffer, pH 7.2, containing 0.25% Emulphogene and allowed to stand at 4°C for 30 min to solubilize the membranes. The homogenate was centrifuged at 88,000 ϫ g for 60 min at 4°C, and the supernatant was concentrated to about 8 ml using an Amicon ultrafiltration cell with a YM-30 membrane and dialyzed against 3 ϫ 200 ml of 20 mM potassium phosphate buffer, pH 7.2, containing 0.25% Emulphogene. The dialysate was centrifuged at 98,000 ϫ g for 60 min at 4°C, and the supernatant was introduced into a Mono Q (HR 10/10) column pre-equilibrated with 20 mM potassium phosphate buffer, pH 7.2, containing 0.25% Emulphogene controlled by a Pharmacia FPLC system. The column was washed with the same buffer until the 280 nm absorbance returned to baseline and then eluted with a KCl gradient of 0 -300 mM over a 120-min period using a buffer system containing 20 mM potassium phosphate, pH 7.2, and 0.25% Emulphogene. The fractions with the highest D␤M activity were combined and concentrated using an Amicon ultrafiltration device using a YM-30 membrane. The protein content of the concentrated enzyme was determined by the absorbance at 280 nm (⑀ 1% ϭ 12.4 (41)). The specific activity of purified membranebound D␤M was 3.5 mol/mg/min.
Membrane-bound D␤M from the above preparation was used in the estimation of the inhibition potency of IGA and the substrate activity of 6-AAscH 2 using the oxygen monitor assay as described previously (35).

RESULTS
The resealed chromaffin granule ghosts prepared to contain 20 mM AscH 2 , 20 mM Tris phosphate, 100 mM KCl, 150 mM sucrose, 10 mM sodium fumarate, and 100 g/ml catalase according to the above procedures (standard resealed granule ghosts), actively accumulate DA and convert it to NE in a time-dependent manner in the presence of 5 mM ATP, 5 mM MgSO 4 , and 20 mM AscH 2 in the external incubation medium (standard uptake and conversion incubation conditions). Appropriate control experiments with the standard resealed granule ghosts revealed that the exclusion of ATP or inclusion of 2 M reserpine in the external incubation medium completely abolishes the DA accumulation and NE production under the above conditions as expected (data not shown).
A series of control experiments were carried out with standard resealed granule ghosts in order to quantify the experimental variability of the rate of DA uptake and its conversion to NE and to examine the viability of the standard ghost preparations during the time course of the experiment under the standard uptake and conversion incubation conditions. Under these conditions, the average internal E levels of the standard resealed granule ghosts varies from 23.0 to 15.7 nmol/mg within a 25-min incubation period (Fig. 1a) probably due to the slow lysis or the leakage of the internal contents of the resealed granule ghosts (see "Discussion"). The data standardized against average constant indigenous E levels show that the average internal AscH 2 (reduced) levels were 23.5 nmol/mg (8 -12 mM depending on 2-or 3-l internal volume of granule ghosts) under the above experimental conditions (Fig. 1b) demonstrating that the trapping efficiencies are in the range of 50% (resealing solution contained 20 mM AscH 2 ). Since only the levels of reduced AscH 2 were measured by electrochemical detection methods, and reduced AscH 2 is susceptible to slow auto-oxidation during the experiment, the trapping efficiencies of these experiments may be higher than these values. The maximum experimental variability of the standardized rate of NE production was less than 30% (an average of five different experiments; see Fig. 2a, vertical bars) for these preparations suggesting a good reproducibility of the results regardless of the complexity of the system. However, this experimental variability has been taken into account in the analysis of all the rates of DA accumulation and NE production under various experimental conditions and was assumed to be similar to the standard resealed granule ghosts, if the rates were in the above range. The above experiments also revealed that the rate of NE production is in the range of 1.7-2.4 nmol/mg/min (average rate 2.0 nmol/mg/min; Fig. 1b), and the rate of DA accumulation varies in the range of 0.2-0.7 nmol/mg/min (average rate 0.5 nmol/mg/min; Fig. 1b) for different standard resealed granule ghost preparations under the standard uptake and conversion incubation conditions. In addition, the time courses of both NE production and DA accumulation are relatively linear (Fig. 2, a  and b; vertical bars), and the internal AscH 2 levels were found to be relatively constant throughout the incubation period in all these experiments (Fig. 1, a and b).
The rates of NE production or DA uptake under standard uptake and conversion incubation conditions were not affected FIG. 1. The time courses of NE production and DA uptake in standard resealed chromaffin granule ghosts under standard uptake and conversion conditions. a, ghosts were resealed to contain 20 mM Tris phosphate, 100 mM KCl, 150 mM sucrose, 10 mM sodium fumarate, 100 g/ml catalase, and 20 mM AscH 2 (pH 7.0) and incubated in a medium (2.5 ml total volume) containing 0.3 M sucrose, 10 mM HEPES, 5 mM Mg-ATP, 5 mM MgSO 4 , 20 mM AscH 2 , and 100 g/ml catalase (pH 7.0) at 30°C for 10 min. At zero time, the uptake and conversion reaction was initiated by adding DA to a final concentration of 200 M, and, at 5-min time intervals, 400-l aliquots of the incubate were withdrawn, and HPLC-EC quantitation was performed as noted under "Experimental Procedures." The data presented are the averages of five different experiments. q, NE levels; E, E levels; Ⅺ, AscH 2 levels; f, DA levels. b, same as above, but the catecholamine and AscH 2 levels are standardized with respect to the average indigenous E levels.
when the granule ghosts were resealed to contain no added AscH 2 and a high concentration of active ascorbate oxidase (Fig. 2, a and b). The rate of NE production in this preparation was in the normal range in comparison with that of the control standard resealed granule ghosts suggesting a high rate of DA to NE conversion in the absence of internal AscH 2 . On the other hand, when standard resealed granule ghosts were incubated in a medium in the absence of external AscH 2 , but in the presence of a high concentration of external ascorbate oxidase, the rate of DA accumulation was increased by about 3-fold, and NE production was decreased by about 3-fold in comparison with the standard resealed granule ghosts (Fig. 2, a and b). Furthermore, when AscH 2 was excluded from the interior and the external incubation medium (by the addition of ascorbate oxidase into both media), the rate of NE production was again reduced by about 5-fold without significantly altering the net rate of DA uptake (data not shown).
The effect of external and/or internal IGA, which is a potent inhibitor for the AscH 2 site of both soluble (K i ϭ 26.4 Ϯ 5.2 M(36)) and membrane-bound D␤M, 2 on the rate of DA accumulation and NE production in resealed chromaffin granule ghosts was examined in a series of experiments. When granule ghosts were resealed to contain 10 mM IGA together with 20 mM AscH 2 and incubated under standard uptake and conversion conditions, the rates of either the DA accumulation or NE production were not significantly altered compared to identical controls without IGA (data not shown). In fact, the rate of NE production under these conditions was in the high normal range in comparison to the control standard resealed granule ghosts. In addition, the rates of both DA accumulation and NE production in ghosts resealed to contain either 1 or 10 mM IGA in the presence (5 mM or 20 mM) or absence of added AscH 2 were also in the normal range under standard uptake and turnover conditions (Fig. 3, a and b). On the other hand, the rate of NE production was significantly decreased when standard resealed granule ghosts were incubated under uptake and conversion conditions in the presence of 1 mM IGA and 5 mM AscH 2 with respect to a parallel identical control without IGA. The rate of DA accumulation in the IGA-containing incubates was significantly higher than the rate of the controls suggesting a slow rate of DA to NE conversion without changing the net rate of DA uptake as expected (Fig. 3b). Furthermore, when ghosts were prepared to contain either AscH 2 , IGA, or ascorbate oxidase and incubated in the presence of external IGA (note that IGA is not a substrate or inhibitor for ascorbate oxidase (42)) and ascorbate oxidase, the rate of NE production dropped to the background level while the rate of DA accumulation increased to a very high level confirming that external IGA does not support the D␤M monooxygenation reaction in resealed granule ghosts. Control experiments revealed that IGA is not actively or passively accumulated into granule ghosts under the experimental conditions (data not shown).
The specificity of the extragranular reduction site of the chromaffin granule membrane was examined by using the AscH 2 analog, 6-AAscH 2 , which is a very weak reductant for purified soluble (36) as well as membrane-bound D␤M (less than 10% activity at 20 mM concentration in comparison to AscH 2 ). The data presented in Fig. 4a clearly demonstrate that in contrast to its inability to reduce soluble or membranebound purified D␤M efficiently, under standard experimental conditions, external 6-AAscH 2 supports the NE production in standard resealed granule ghost with the efficiency similar to that of AscH 2 itself. The data also show that external AscH 2 supports NE production in 6-AAscH 2 -loaded ghosts with the efficiency similar to that of AscH 2 -loaded ghosts. Furthermore, the rate of DA accumulation is in the normal range for both of these sets of experiments suggesting that external or internal 6-AAscH 2 has no significant effect on the net rate of DA uptake (Fig. 4, a and b). Although 6-AAscH 2 has a remarkable structural similarity to DA, control experiments revealed that it is 2 In control experiments, 1 mM concentration of IGA was found to inhibit 90% of membrane-bound D␤M activity in the presence of 5 mM AscH 2 ; a similar inhibition was also observed with lysed membrane fragments under similar conditions. not accumulated into granule ghosts through the DA uptake mechanism in an ATP-and time-dependent manner, under standard uptake and conversion incubation conditions. Other control experiments indicated that 6-AAscH 2 had no significant effect on the viability of resealed ghosts under the experimental conditions.
The data presented in Fig. 5 demonstrate that the chro-mophoric D␤M reductant, 2-AAscH 2 (35), also supports the conversion of DA to NE in standard resealed granule ghosts with an efficiency similar to that of AscH 2 (control experiments indicate that 2-AAscH 2 does not permeate the granule membrane at a detectable rate although it is somewhat less polar than AscH 2 due to the alteration of the pK a of the 3Ј-OH group). Although quantitative studies have not been completed yet, we have observed the formation of the chromophoric oxidized product of 2-AAscH 2 , the red pigment, in a time-dependent manner in these experiments when 2-AAscH 2 is the external electron donor (we believe this reaction could be adopted to measure the rate of electron flux into granule ghosts; detailed experiments are in progress). On the other hand, the chromophoric non-

FIG. 3. The effect of internal and/or external imino-D-glucoascorbic acid (IGA) on the time courses of NE production (a) and DA accumulation (b) in resealed chromaffin granule ghosts.
Ghosts were resealed to contain 20 mM Tris phosphate, 100 mM KCl, 150 mM sucrose, 10 mM fumarate, and 100 g/ml catalase with AscH 2 , IGA, ascorbate oxidase, or a combination of the above (pH 7.0) and incubated in a medium (2.5 ml total volume) containing 0.3 M sucrose, 10 mM HEPES, 5 mM Mg-ATP, 5 mM MgSO 4 , and 100 g/ml catalase with AscH 2 , IGA, ascorbate oxidase, or a combination of the above (pH 7.0) at 30°C for 10 min. At zero time, the uptake and conversion reaction was initiated by adding DA to a final concentration of 200 M, and, at 5-min time intervals, 400-l aliquots of the incubate were withdrawn, and HPLC-EC quantitation was performed as noted under "Experimental Procedures." q, 5 mM AsCH 2 inside, 20 mM AscH 2 outside; E, 20 mM AscH 2 inside, 5 mM AscH 2 outside; f, 5 mM AscH 2 and 1 mM IGA inside, 20 mM AscH 2 outside; Ⅺ, 20 mM AscH 2 inside, 5 mM The possibility that DA may function as a redox mediator between the extragranular reduction site and membranebound D␤M was investigated by using reductively inactive tyramine as the D␤M substrate. In these experiments, the rates of intragranular octopamine formation were accurately quantified by reversed phase HPLC-EC as described under "Experimental Procedures." The data presented in Fig. 6 clearly demonstrate that standard resealed granule ghosts uptake tyramine and convert it to the D␤M hydroxylation product, octopamine, in a time-dependent manner with a rate somewhat slower than that of the DA to NE conversion under standard uptake and turnover conditions (see Fig. 6, vertical bars; however, note that the external tyramine concentrations in these experiments were 30 M, whereas DA concentrations in analogous experiments were 200 M). The exclusion of internal AscH 2 by resealing the granule ghosts to contain a high level of highly active ascorbate oxidase 3 does not significantly change the rate of octopamine production in comparison to the standard controls under identical conditions (Fig. 6). The rate of octopamine production was not inhibited when ghosts were resealed to contain 5 mM IGA and no added AscH 2 and incubated under standard uptake and turnover conditions (Fig. 6). In contrast, a slight increase in the rate of octopamine production was observed compared to the standard controls when internal IGA was present, similar to that observed for the DA to NE conversion. In addition, external 20 mM 6-AAscH 2 supports the tyramine to octopamine conversion in 20 mM AscH 2loaded granule ghosts with a rate slightly lower than that of the standard controls. DISCUSSION The optimum conditions and procedures to examine the DA uptake and NE production in resealed chromaffin granule ghosts under various experimental conditions with non-radiolabeled substrate have been reported previously (26,27). The key steps in this procedure were the resealing of the granule membranes under optimum conditions necessary for maximum D␤M turnover, the termination of the uptake of DA and its conversion to NE at a given period of time by diluting the incubates with an ice-cold osmotically balanced solution, the separation and reisolation of the intact granule ghosts from the external incubation medium without contamination, and the quantification of the internal contents of the granule ghost incubates by HPLC-EC. The ghosts prepared, incubated, and analyzed according to the above procedures were found to actively uptake DA and convert it to NE in an ATP-and time-dependent and reserpine-sensitive manner giving consistent and reproducible results. Therefore, we have used the same protocols and procedures in the present study without significant alterations or modifications.
The resealed granule ghosts prepared according to the above 3 In a previous study, Huyghe and Klinman (14) have reported that the rate of conversion of tyramine to octopamine was increased by about 2.7-fold when granule ghosts were resealed to contain a high concentration of AscH 2 inside and incubated in an external medium containing 2 mM AscH 2 . However, we have not observed such a significant change in the rate of DA to NE or tyramine to octopamine conversion when the internal AscH 2 levels of resealed granule ghosts were changed from none (no added AscH 2 and with ascorbate oxidase) to 20 mM AscH 2 and incubated in a medium containing 20 mM AscH 2 (Figs. 2, a and b, and  6). We believe this apparent discrepancy may at least be partly due to the differences in the experimental conditions used in the two studies. procedures contained relatively constant levels of internal E which were found to slowly decline during the incubation time period (Fig. 1a). Since the depletion of E was observed even when excess AscH 2 was present in both the internal matrix and in the external incubation medium, this could not be due to the auto-oxidation of the catechol moiety of the molecule and must be due to the slow lysis or the leakage of the internal contents of the ghosts during the incubation period or reisolation procedure. Therefore, any loss or leakage of internal contents from resealed granule ghosts during the experiments could accurately be corrected (Fig. 1b) by standardizing against the average indigenous E levels since NE is not converted to E within resealed granule ghosts due to the lack of the essential enzyme, phenethylamine N-methyltransferase (5). Furthermore, external contamination or incomplete resealing of ghost preparations prior to the incubation could also be estimated by measuring the DA content at the time 0 point which is usually 0 under the above experimental conditions (Fig. 1, a and b). The normal rates of NE production and DA accumulation in standard resealed granule ghosts (Figs. 1 and 2) were in good agreement with the rates previously reported under similar experimental conditions (26,27,43). The observed low ratio of the net rate of DA uptake to NE production (in the range of 1.1-1.3) together with the lack of lag or burst periods in both of these time courses under the standard experimental conditions (Fig.  2, a and b), suggest that both the rate of DA uptake and D␤M turnover contribute to the net rate of NE production in resealed granule ghosts which is also consistent with the previous observations (44).
In the experiments where removal of internal AscH 2 contaminants was necessary, ghosts were resealed to contain high concentrations of active ascorbate oxidase. The interior contents of the ghosts reisolated from these preparations, after incubation under standard uptake and conversion conditions, were found to consistently contain an average of 6 -7 nmol/mg of AscH 2 regardless of the presence of high levels of intragranular ascorbate oxidase (similar observations have been reported previously (14)). However, these AscH 2 levels were not increased in a time-and/or ATP-dependent manner during the incubation period suggesting that external AscH 2 was not actively taken up into the ghosts (data not shown) and were much less than the external AscH 2 concentration of the incubation medium, ruling out the possibility that external AscH 2 was passively transported and equilibrated during the incubation period. Furthermore, since ascorbate oxidase is a very efficient enzyme (k cat ϭ 3,000 -11,000 s Ϫ1 (46)) with very high affinity toward AscH 2 (K m is in the M range (46)), the observed levels of AscH 2 could not possibly be in the interior soluble matrix of the resealed granule ghosts and not be accessible to ascorbate oxidase. Therefore, a possible simple explanation for the above observation is that AscH 2 nonspecifically and tightly interacts with the external phase of the resealed granule ghost membrane. Since the amino acid sequence and tertiary structure model proposed for cyt b 561 contains clusters of positively charged amino acid side chains in the cytoplasmic domain of the protein (47), it is possible that negatively charged AscH 2 molecules may specifically and strongly interact with these regions of the protein in order to facilitate the subsequent reaction with the buried heme. Alternatively, it is possible that a small pool of permanently inner or outer membrane-associated externally reducible AscH 2 exists in resealed chromaffin granule ghosts (for example, see Ref. 45) which may not be accessible to the intragranular ascorbate oxidase. However, the present results do not distinguish between these possibilities.
The internal AscH 2 levels in standard resealed granule ghosts remain relatively constant (average of 23.5 nmol/mg of protein of detectable reduced AscH 2 ) throughout the incubation period (25-30 min) under standard uptake and conversion incubation conditions (Fig. 1, a and b) suggesting that there is no net consumption of internal AscH 2 by the D␤M monoxygenation reaction under these conditions (44). Since it is generally accepted (our results also confirm) that AscH 2 is not actively or passively transported into the interior of the granule ghosts at a detectable rate (11)(12)(13)(14), the reducing equivalents essential for the D␤M reaction must be provided from the cytosolic pool of the reductant. Alternatively, although the contaminating intragranular catecholamines present in the system could provide necessary reducing equivalents for the D␤M-catalyzed DA to NE conversion in theory (for example, see Ref. 45), the rate of depletion of the catecholamine concentration during the incubation period was much slower (about 0.2 to 0.3 nmol/mg/ min) and was not sufficient (Fig. 1, a and b) to account for the observed rate of NE production. On the other hand, the data presented in Fig. 2a clearly demonstrate that external AscH 2 is essential for the efficient conversion of DA to NE in standard resealed granule ghosts. 4 However, the internal or external AscH 2 has no significant effect on the net rate of DA uptake into resealed granule ghosts (i.e. the combined rates of DA accumulation and NE production; Fig. 2b) under the same experimental conditions. Therefore, it can be concluded that while external AscH 2 is essential for the efficient conversion of DA to NE in resealed chromaffin granule ghosts under standard uptake and conversion incubation conditions, internal AscH 2 has no significant effect.
Internal IGA (an efficient reductant site-competitive inhibitor of D␤M) concentrations of up to 10 mM has no significant effect on either the rate of NE production or the rate of DA accumulation under standard uptake and conversion incubation conditions in the presence or even in the absence of externally added AscH 2 in the internal matrix of resealed granule ghosts (Fig. 3, a and b). In fact, close examination of the results indicate that the rate of NE production in the presence of internal IGA was slightly higher in comparison with identical controls without IGA. Therefore, it is clear that although IGA is an excellent inhibitor for both soluble and membrane-bound forms of D␤M and effectively interacts with their reduction sites (AscH 2 sites), it does not effectively interact with the reduction site of membrane-bound D␤M in resealed granule ghosts from the internal phase of the granule membrane. In addition, the lack of expected competition between internal IGA and AscH 2 for the reduction site of the enzyme strongly suggests that internal AscH 2 also may not directly interact with the reduction site of D␤M in the resealed chromaffin granule ghost membrane. On the other hand, externally applied IGA was found to inhibit the intragranular DA to NE conversion efficiently (Fig. 3, a and b) without significantly affecting the net rate of DA uptake (Fig. 3, a and b) suggesting that IGA effectively interacts with the external reduction site of the granule membrane. These results support the proposal that, while intragranular AscH 2 may not directly reduce or mediate the reduction of membrane-bound D␤M in resealed chromaffin granule ghosts, the external reductant may be the exclusive reductant for the enzyme under these experimental 4 An argument was made that the lack of efficient DA to NE conversion in these experiments is simply due to the external ascorbate oxidase-catalyzed fast depletion of internal AscH 2 and/or dissolved oxygen. However, since the exclusion of ascorbate oxidase from the external medium does not change the results of these experiments, as previously shown (see Ref. 26), these results could not be due to the depletion of internal AscH 2 or oxygen. Furthermore, in these experiments, ascorbate oxidase was added to the incubation medium as a precaution, and, under the experimental conditions, the contaminating external AscH 2 levels were very low.

conditions.
In our previous studies (36) we have shown that 6-AAscH 2 was not an efficient reductant for purified soluble D␤M. We have now shown that 6-AAscH 2 is a very weak substrate for membrane-bound D␤M as well (see "Results"). However, 6-AAscH 2 was found to be a good external reductant for membrane-bound D␤M in standard resealed granule ghosts and supports the DA to NE conversion with an efficiency only slightly lower than that of AscH 2 (Fig. 4). In addition, although 6-AAscH 2 is structurally remarkably similar to DA, these results clearly demonstrate that it is not actively or passively taken up into resealed granule ghosts. These results strongly suggest that the specificity of the reduction site of purified D␤M must be significantly different from the specificity of the external reduction site of the resealed granule ghost membrane. Therefore, it could be concluded that although membrane-bound D␤M in resealed granule ghosts appear to be exclusively reduced by the external reductant, the external reductant may not directly interact with the reduction site of the enzyme from the cytosolic phase of the granule. This conclusion is further substantiated by our preliminary observation that the efficient artificial chromophoric D␤M reductant, N,Ndimethyl-1,4-phenylenediamine (35), does not support the DA to NE conversion in standard resealed granule ghosts from the external phase.
Since DA is a known weak reductant for purified D␤M (48), it is possible that the high rate of DA to NE conversion in the absence of internal AscH 2 may simply be due to the ability of internal DA to act as a redox mediator between the external reduction site and membrane-bound D␤M. On the other hand, our observation that internal IGA does not competitively inhibit the conversion of DA to NE in resealed granule ghosts in the presence or absence of internal AscH 2 appears to rule out this possibility because if DA is an effective redox mediator between the external reduction site and membrane-bound D␤M, this process must also be competitively inhibited by IGA, since IGA has been shown to compete with DA for the reduction site of membrane-bound D␤M (see "Results"). However, this possibility was further examined by replacing DA in several key experiments with the reductively inactive well characterized alternate D␤M substrate, tyramine. The remarkable similar outcome of these key experiments to the corresponding experiments with DA clearly rules out the possibility that under internal AscH 2 -depleted conditions DA acts as a redox mediator between the external reduction site and membranebound D␤M in resealed granule ghosts.
Taken together, the above results clearly demonstrate that the reducing equivalents for the membrane-bound D␤M monooxygenation reaction in resealed chromaffin granule ghosts are exclusively provided by the external reductant. These results are in general agreement with the previous proposal of Huyghe and Klinman (14). More interestingly, in contrast to the generally accepted notion, the results presented here also strongly suggest that the intravesicular solution AscH 2 (or DA) may not directly reduce or mediate the reduction of membrane-bound D␤M in resealed chromaffin granule ghosts. Since the specificity of the reduction site of membrane-bound D␤M appears to be different from the specificity of the extragranular reduction site in the membrane and the fact that membrane-bound D␤M is anchored to the internal surface of the granule membrane through a short signal sequence (49), it is clear that membranebound D␤M could not be directly reduced from the external phase. Therefore, we conclude that there must be an electron acceptor site in the external phase of the granule membrane which is capable of accepting and transferring the external reducing equivalents directly or indirectly to the reduction site of the membrane-bound D␤M without using intragranular AscH 2 (or DA) as an intermediate.
Numerous previous studies have demonstrated that the transmembrane hemoprotein, cyt b 561 , is capable of transferring external reducing equivalents through the membrane to the interior of the granule (19 -22). Furthermore, it has been proposed that the cyt b 561 -mediated reduction of D␤M in chromaffin granules requires intragranular AscH 2 as a mediator (50,51). However, the above results strongly suggest that internal solution AscH 2 (or DA) is not involved in the mediation or the direct reduction of membrane-bound D␤M in resealed chromaffin granule ghosts. Therefore, if cyt b 561 is the external reducing equivalent acceptor, then the reducing equivalents must be transferred to the reduction site of membrane-bound D␤M either directly or through an unidentified intermediate electron carrier without using intragranular solution AscH 2 (or DA) as an intermediate. Since this electron transfer process appears to be highly efficient and requires tight functional coupling between cyt b 561 and the reduction site of membranebound D␤M, it is very unlikely that the soluble D␤M in the granule is also reduced by this same cyt b 561 -mediated electron transport system. Therefore, it is possible that while membrane-bound D␤M in chromaffin granules may be reduced through an unidentified external reduction site in the membrane, the soluble enzyme is reduced by intragranular AscH 2 , and constant internal AscH 2 levels are maintained through the cyt b 561 -mediated AscH 2 regenerating system as previously proposed (19 -22). An attractive alternate possibility would be that the membrane-bound D␤M is the only physiologically functional form which is reduced through cyt b 561 directly or using a mediator other than solution AscH 2 (or DA), and the soluble form is only designated as a disposal form of the enzyme (not recovered and disposed through exocytosis) which might not be functional under physiological conditions. 5 The physiological function of intragranular AscH 2 may be to provide a reducing environment inside the chromaffin granules to protect catecholamines from auto-oxidation. While further experimental evidence is certainly necessary to distinguish between these possibilities, the understanding of the topological arrangement of the various proteins and co-factors in the granule membrane, physiological function(s) of other redox-active proteins or small molecules in the granule matrix, and the role of soluble D␤M in intact chromaffin granules are of prime importance.